SRIE LIVROS 28

PORIFERA RESEARCH BIODIVERSITY, INNOVATION AND SUSTAINABILITY

Editors Mrcio Reis Custdio Gisele Lbo-Hajdu Eduardo Hajdu Guilherme Muricy

RIO DE JANEIRO MUSEU NACIONAL 2007

Universidade Federal do Rio de Janeiro Reitor Alosio Teixeira Museu Nacional Diretor Srgio Alex K. Azevedo Comisso de Publicaes do Museu Nacional Editores Miguel Angel Monn Barrios, Ulisses Caramaschi

ISBN 978-85-7427-023-4

Editores de rea Adriano Brilhante Kury, Alexander Wilhelm Armin Kellner, Andrea Ferreira da Costa, Ctia Antunes de Mello Patiu, Ciro Alexandre vila, Dbora de Oliveira Pires, Guilherme Ramos da Silva Muricy, Izabel Cristina Alves Dias, Joo Alves de Oliveira, Joo Wagner de Alencar Castro, Marcela Laura Monn Freire, Marcelo de Arajo Carvalho, Marcos Raposo, Maria Dulce Barcellos Gaspar de Oliveira, Marlia Lopes da Costa Fac Soares, Rita Scheel Ybert, Vnia Gonalves Loureno Esteves Normalizao Vera de Figueiredo Barbosa MUSEU NACIONAL Universidade Federal do Rio de Janeiro Quinta da Boa Vista, So Cristvo, 20940-040 Rio de Janeiro, RJ, Brasil Reviso Fernando Moraes, Raphael Augusto Sims Belleza, Gisele Lbo-Hajdu, Guilherme Muricy Diagramao e arte-final Mrcio Reis Custdio e Beatriz Waller Capa Beatriz Waller Comisso Editorial do Volume Mrcio Reis Custdio Gisele Lbo-Hajdu Eduardo Hajdu Guilherme Muricy Patrocinadores FAPERJ, CNPq, CAPES Apoio Universidade Federal do Rio de Janeiro, Universidade do Estado do Rio de Janeiro, Universidade de So Paulo, PETROBRAS Impresso: IMOS Grfica e Editora Impresso no Brasil / Printed in Brazil

Ficha catalogrfica P836 Porifera research : biodiversity, innovation and sustainability / editors Mrcio Reis Custdio ... [et al.]. Rio de Janeiro : Museu Nacional, 2007 694 p. ; il. ; 28 cm. (Srie Livros ; 28) Inclui bibliografia ISBN 978-85-7427-023-4 1. Esponjas. I. Custdio, Mrcio Reis. II. Museu Nacional (Brasil). III. Srie. CDD 593.4

Preface
This book began to be assembled in the frame of the 7th International Sponge Symposium, held in Armao dos Bzios (Rio de Janeiro, Brazil) in May 2006. Under different names, and with a history almost four decades long now, this series of meetings started in London (1968), followed by Paris (1978), Woods Hole (1985), Amsterdam (1993), Brisbane (1998), Rapallo (2002) and Armao dos Bzios. These are the worlds main international scientific events centered on Porifera. The 7th ISS was the first of this series held in Latin America, and the Museu Nacional of the Universidade Federal do Rio de Janeiro (MN-UFRJ) and the Sociedade dos Amigos do Museu Nacional (SAMN) were honored to organize it. In what seems to be a welcome trend in these meetings, this seventh edition was also the largest so far. Almost 260 participants from 35 countries contributed with 308 presentations, distributed in 14 sessions, and covering all aspects of sponge basic and applied research. But we sincerely hope that these numbers will be superseded in 2010, when the sponge research community is expected to meet again in Spain. Following the lead of our fellow Italians, Maurizio Pansini, Roberto Pronzato, Giorgio Bavestrello and Renata Manconi, editors of the previous volume Sponge Science in the New Millenium (2002), we decided not to divide the book by subject matters. This organization reflects the current tendency of multidisciplinary work in biological sciences, in which ever more studies use different approaches, drawn from several disciplines, to address their hypotheses. Like in all previous volumes, the data presented here will be a valuable, updated reference of the present knowledge for those working with this still largely unknown and fascinating group of animals. Titles bridging a variety of disciplines tend to be unattractive to those conducting sharply focused research. Nevertheless, the focus of ISS meetings has to adjust to a time where borders between disciplines become more and more blurred, not to speak of borders between Phyla! Nonetheless, Porifera constituted the magnet for all the contributions presented in the 7th ISS, and so it is with those published in the following pages of this book. Porifera Research alone would not convey the excitement of organizing the meeting and editing the book, neither would it be fair to all of you who packed your back-packs and suitcases on every continent on Earth to travel to the far meeting ground at Bzios. To reflect this, and in respect to Brazils nearly synonymous significance, first subtitle emerged Biodiversity. But, natures treasures alone are no guarantee of wealth to any nation, and among the many biodiverse countries represented by those who participated in the 7th ISS, most are developing and struggle to generate wealth to their peoples. Accordingly, an unavoidable target became second subtitle - Innovation. Hoping we can all adjust to an era of growing environmental concern, partly as a consequence of fear of the global consequences of increased warming of the planet, our third subtitle wishes to convey the ideals of the editors, as far as exploration of natural resources are concerned Sustainability. In this way, we reached our motto Biodiversity, innovation and sustainability. The book Porifera research: biodiversity, innovation and sustainability begins with a series of twelve invited contributions, not meant to match the books motto, and spanning most fields of research on Porifera, from the paleontology of old Pre-Cambrian rocks to DNA barcoding of recent sponges and its potential effects on the classification of the Phylum. Following, 61 articles are listed in alphabetical order of first authors family name, again spanning a vast spectrum of disciplines. Differently from the other books in the series, this is not strictly a Proceedings volume. We decided to open the possibility for those who were unable to attend the 7th ISS, as well as those who participated, to publish results other than those presented in the meeting. In this way, we expect to present a broad perspective of the contemporary knowledge and future research trends in the group. The 73 manuscripts published in this book contain the work of over 230 coauthors, and were evaluated by more than 100 anonymous peer reviewers, in a process that took over a year. During these procedures, we attempted to assure that contrasting ideas and opinions could be published. To provide a wide distribution of all articles, the whole book is available in PDF format for download without restrictions from the Porifera Brasil website: www.poriferabrasil.mn.ufrj.br. It would have been impossible to organize this book without the help of many persons and sponsors. Our deepest acknowledgements go to the authors of the articles published along these pages, for their scientific contributions, which are the heart of the book. We are also grateful to all reviewers, who spent their time and experience correcting or making clearer the rationales (and often the language) of submitted manuscripts. Their work greatly improved the quality of the book, and we thank you all for your cooperation and patience, and hope that you enjoy the final product as we did. The financial support by some sponsors was essential for the realization of the 7th ISS and publication of this book. Special thanks go to FAPERJ, CAPES, CNPq and CENPES/ PETROBRAS. The logistic support of some institutions was also pivotal. Thanks here go to Museu Nacional (Universidade Federal do Rio de Janeiro), Universidade do Estado do Rio de Janeiro (UERJ), Universidade de So Paulo (USP), Sociedade dos Amigos do Museu Nacional, Hotel Prola Bzios and ABVTUR. We also warmly thank all members of the steering committee: Antnio Sol-Cava, Beatriz Mothes, Carla Silva, Carla Zilberberg, Ceclia Volkmer-Ribeiro, Cla Lerner, Cristiano Coutinho, Michelle Klautau, Radovan

Borojevic, Roberto Berlinck and Solange Peixinho; as well as Andrezj Pisera, Marinella Laport and Sally Leys for additional support. Finally, we could not possibly forget the several volunteers who helped us in preparing and running the conference: Andr Rossi, Barbara Andrea, Carla Zilberberg, Daniela Batista, Daniela Lopes, Emiliano Calderon, Emlio Lanna, Fernanda Azevedo, Fernanda Cavalcanti, Fernando de Moraes, Guilherme Maia, Gustavo da Silva, Karina Hajdu, Leandro Monteiro, Mara de Oliveira, Mariana Carvalho, Maurcio de Campos, Suzi Ribeiro and Viviane Santos. Looking back in time, two moments were crucial for the making of the 7th ISS, and consequently for the publication of this book. First of these, the gathering in Rio de Janeiro, back in 1987, of Professors Nicole Boury-Esnault, Solange

Peixinho, Radovan Borojevic and Antnio Sol-Cava, to teach a course on sponge biology to a group of young undergraduates of Universidade Federal do Rio de Janeiro. Among these, three editors of this book. Remembering these first steps is a much deserved honor we are obliged to render. Secondly, the many demands for a meeting in Rio de Janeiro, starting in Amsterdam, at the occasion of the 4th ISS in 1993, a time when the four editors of this volume were still half way in their PhDs. We hope that this book will not only provide an update of achievements in most fields of inquiry regarding sponges, but also be a fertile ground for the birth of new questions, debates and ideas for future endeavours on Porifera Research spread worldwide.

The Editors
Mrcio Reis Custdio Gisele Lbo-Hajdu Eduardo Hajdu Guilherme Muricy

Table of contents

Invited Articles
Pedro M. Alcolado Reading the code of coral reef sponge community composition and structure for environmental bio-monitoring: some experiences from Cuba ....................................................... 3-10 Matilde Sylvia Beresi Fossil sponges of Argentina: a review ...........................................................................................11-21 Nicole Boury-Esnault, Chantal Bzac Morphological and cytological descriptions of a new Polymastia species (Hadromerida, Demospongiae from the North-West Mediterranean Sea ............................................................. 23-30 Maria Cristina Diaz, Robert W. Thacker, Klaus Rtzler, Carla Piantoni Two new haplosclerid sponges from Caribbean Panama with symbiotic filamentous cyanobacteria, and an overview of sponge-cyanobacteria associations ....................................... 31-39 Alexander V. Ereskovsky Sponge embryology: the past, the present and the future ............................................................. 41-52 Sally P. Leys Sponge coordination, tissues, and the evolution of gastrulation .................................................. 53-59 Renata Manconi, Roberto Pronzato Gemmules as a key structure for the adaptive radiation of freshwater sponges: a morphofunctional and biogeographical study ........................................................................................... 61-77 Gradimir N. Misevic, Camille Ripoll, Jonathan Norris, Vic Norris, Yann Guerardel, Emmanuel Maes, Gerard Strecker, Pascal Ballet, Yannis Karamanos, Lazar T. Sumanovski, Octavian Popescu, Nikola Misevic Evolution of multicellularity in Porifera via self-assembly of glyconectin carbohydrates .......... 79-88 Werner E.G. Mller, Isabel M. Mller Porifera: an enigmatic taxon disclosed by molecular biology/cell biology ............................... 89-106 Jean Vacelet Diversity and evolution of deep-sea carnivorous sponges ........................................................107-115 Ceclia Volkmer-Ribeiro South American continental sponges: state of the art of the research .......................................117-121 Gert Wrheide, Dirk Erpenbeck, Christian Menke The Sponge Barcoding Project: aiding in the identification and description of poriferan taxa ........................................................................................................................................... 123-128

ii

Research Articles
Brbara R. Andra, Daniela Batista, Cludio L.S. Sampaio, Guilherme Muricy Spongivory by juvenile angelfish (Pomacanthidae) in Salvador, Bahia State, Brazil .............. 131-137 William C. Austin, Kim W. Conway, J. Vaughn Barrie, Manfred Krautter Growth and morphology of a reef-forming glass sponge, Aphrocallistes vastus (Hexactinellida), and implications for recovery from widespread trawl damage .................... 139-145 Enrique vila, Jos Lus Carballo, Jos Antonio Cruz-Barraza Symbiotic relationships between sponges and other organisms from the Sea of Cortes (Mexican Pacific coast): same problems, same solutions ........................................................ 147-156 Francesca Azzini, Barbara Calcinai, Carlo Cerrano, Giorgio Bavestrello, Maurizio Pansini Sponges of the marine karst lakes and of the coast of the islands of Ha Long Bay (North Vietnam)........................................................................................................................ 157-164 Kristina Bayer, Susanne Schmitt, Ute Hentschel Microbial nitrification in Mediterranean sponges: possible involvement of ammoniaoxidizing Betaproteobacteria .................................................................................................... 165-171 Leontine E. Becking, Yoichi Nakao, Nicole J. de Voogd, Rob W.M. van Soest, Nobuhiro Fusetani, Shigeki Matsunaga Perplexing distribution of 3-alkylpyridines in haplosclerid sponges ....................................... 173-178 Sergey I. Belikov, Oksana V. Kaluzhnaya, Heinz C. Schrder, Isabel M. Mller, Werner E.G. Mller Lake Baikal endemic sponge Lubomirskia baikalensis: structure and organization of the gene family of silicatein and its role in morphogenesis ........................................................... 179-188 Marco Bertolino, Laura Schejter, Barbara Calcinai, Carlo Cerrano, Claudia Bremec Sponges from a submarine canyon of the Argentine Sea ......................................................... 189-201 Barbara Calcinai, Francesca Azzini, Giorgio Bavestrello, Laura Gaggero, Carlo Cerrano Excavating rates and boring pattern of Cliona albimarginata (Porifera: Clionaidae) in different substrata ..................................................................................................................... 203-210 Emiliano Nicolas Calderon, Carla Zilberberg, Paulo Csar de Paiva The possible role of Echinometra lucunter (Echinodermata: Echinoidea) in the local distribution of Darwinella sp. (Porifera: Dendroceratida) in Arraial do Cabo, Rio de Janeiro State, Brazil ...................................................................................................................211-217 Maurcio Campos, Beatriz Mothes, Cla Lerner, Joo Lus Carraro, Inga Ludmila Veitenheimer-Mendes Sponges (Porifera, Demospongiae) from Bransfield strait, off Joinville Island, collected by Brazilian Antarctic Program - PROANTAR ....................................................................... 219-232 Victor Ribeiro Cedro, Eduardo Hajdu, Hilda Helena Sovierzosky, Monica Dorigo Correia Demospongiae (Porifera) of the shallow coral reefs of Macei, Alagoas State, Brazil ........... 233-237

iii

Carlo Cerrano, Barbara Calcinai, Cristina Gioia Di Camillo, Laura Valisano, Giorgio Bavestrello How and why do sponges incorporate foreign material? Strategies in Porifera ...................... 239-246 Andia Chaves-Fonnegra, Sven Zea Observations on reef coral undermining by the Caribbean excavating sponge Cliona delitrix (Demospongiae, Hadromerida) .................................................................................... 247-254 Mark Chiappone, Leanne M. Rutten, Steven L. Miller, Dione W. Swanson Large-scale distributional patterns of the encrusting and excavating sponge Cliona delitrix Pang on Florida Keys coral substrates ......................................................................... 255-263 Steve de C. Cook Clarification of dictyoceratid taxonomic characters, and the determination of genera ............ 265-274 Bruno Cosme, Solange Peixinho A new species of Stelletta (Astrophorida: Demospongiae) with a redescription and distribution range expansion for Stelletta kallitetilla in the Southwestern Atlantic Region .... 275-280 Cristiano C. Coutinho, Guilherme de Azevedo Maia Mesenchymal cells in ancestral spongiomorph urmetazoa could be the mesodermal precursor before gastrulation origin ......................................................................................... 281-295 Alan R. Duckworth, Carsten Wolff, Elizabeth Evans-Illidge Developing methods for commercially farming bath sponges in tropical Australia ................ 297-302 Hermann Ehrlich, Hartmut Worch Sponges as natural composites: from biomimetic potential to development of new biomaterials .............................................................................................................................. 303-312 Elthon G. Ferreira, Diego V. Wilke, Paula C. Jimenez, Tiago A. Portela, Edilberto R. Silveira, Eduardo Hajdu, Cludia Pessoa, Manoel O. de Moraes, Letcia V. Costa-Lotufo Cytotoxic activity of hydroethanolic extracts of sponges (Porifera) collected at Pedra da Risca do Meio Marine State Park, Cear State, Brazil ............................................................. 313-318 Christopher J. Freeman, Daniel F. Gleason, Rob Ruzicka, Rob W.M. van Soest, Alan W. Harvey, Greg McFall A biogeographic comparison of sponge fauna from Grays Reef National Marine Sanctuary and other hard-bottom reefs of coastal Georgia, U.S.A. ......................................... 319-325 Elena I. Gerasimova, Alexander V. Ereskovsky Reproduction of two species of Halichondria (Demospongiae: Halichondriidae) in the White Sea.................................................................................................................................. 327-333 Deborah J. Gochfeld, Carmen Schlder, Robert W. Thacker Sponge community structure and disease prevalence on coral reefs in Bocas del Toro, Panama ..................................................................................................................................... 335-343 Elisaveta L. Gonobobleva Basal apparatus formation in external flagellated cells of Halisarca dujardini larvae (Demospongiae: Halisarcida) in the course of embryonic development.................................. 345-351

iv

Eduardo Hajdu, Daniela A. Lopes Checklist of Brazilian deep-sea sponges .................................................................................. 353-359 Isabel Heim, Michael Nickel, Franz Brmmer Molecular markers for species discrimination in poriferans: a case study on species of the genus Aplysina .......................................................................................................................... 361-371 Isabel Heim, Jrg U. Hammel, Michael Nickel, Franz Brmmer Salting sponges: a reliable non-toxic and cost-effective method to preserve poriferans in the field for subsequent DNA-work ......................................................................................... 373-377 Friederike Hoffmann, Eberhard Sauter, Oliver Sachs, Hans Ry, Michael Klages Oxygen distribution in Tentorium semisuberites and in its habitat in the Arctic deep sea ....... 379-382 Valeria Itskovich, Sergey Belikov, Sofia Efremova, Yoshiki Masuda, Thierry Perez, Eliane Alivon, Carole Borchiellini, Nicole Boury-Esnault Phylogenetic relationships between freshwater and marine Haplosclerida (Porifera, Demospongiae) based on the full length 18S rRNA and partial COXI gene sequences .......... 383-391 Michelle Kelly, Michael Ellwood, Lincoln Tubbs, John Buckeridge The Lithistid Demospongiae in New Zealand waters: species composition and distribution................................................................................................................................ 393-404 Anne Kuusksalu, Madis Metsis, Tnu Reintamm, Merike Kelve Construction and characterization of a cDNA library from the marine sponge Chondrosia reniformis .............................................................................................................. 405-412 Emilio Lanna, Leandro C. Monteiro, Michelle Klautau Life cycle of Paraleucilla magna Klautau, Monteiro and Borojevic, 2004 (Porifera, Calcarea) ................................................................................................................................... 413-418 Nathan Lemoine, Nicole Buell, April Hill, Malcolm Hill Assessing the utility of sponge microbial symbiont communities as models to study global climate change: a case study with Halichondria bowerbanki ....................................... 419-425 Daniela Marques, Marise Almeida, Joana Xavier, Madalena Humanes Biomarkers in marine sponges: acetylcholinesterase in the sponge Cliona celata .................. 427-432 Diana M. Mrquez, Sara M. Robledo, Alejandro Martnez Antileishmanial epidioxysterols from extracted sterols of the Colombian marine sponge Ircinia campana ........................................................................................................................ 433-437 Mirna Mazzoli-Dias, Suzi M. Ribeiro, Patricia Oliveira-Silva Foraminifera associated to the sponge Mycale microsigmatosa in Rio de Janeiro State, southeastern Brazil - an initial approach .................................................................................. 439-442 Elizabeth L. McLean, Paul M. Yoshioka Associations and interactions between gorgonians and sponges.............................................. 443-448 Larisa L. Menshenina, Konstantin R. Tabachnick, Daniela A. Lopes, Eduardo Hajdu Revision of Calycosoma Schulze, 1899 and finding of Lophocalyx Schulze, 1887 (six new species) in the Atlantic Ocean (Hexactinellida, Rossellidae) ........................................... 449-465

Fernando Moraes, Guilherme Muricy A new species of Erylus (Geodiidae, Demospongiae) from Brazilian oceanic islands............ 467-475 Beatriz Mothes, Manuel Maldonado, Rafael Eckert, Cla Lerner, Maurcio Campos, Joo Lus Carraro A new species of Characella (Demospongiae, Astrophorida, Pachastrellidae) from the south Brazilian continental shelf .............................................................................................. 477-482 Yulia I. Mukhina, Vadim V. Kumeiko, Olga I. Podgornaya, Sofia M. Efremova The events of metamorphosis in the demosponge Halisarca dujardini Johnston, 1842, studied with immunocytochemical method .............................................................................. 483-490 Marisa H. Nicolai, Ana Esteves, Marise Almeida, Madalena Humanes Haloperoxidase from the marine sponge Erylus discophorus (Schmidt, 1862) ....................... 491-496 Ronald Osinga, Michiel Kotterman Ferric iron promotes the formation of oscules: observations on sponges in aquaria ............... 497-502 Lorenzo Parma, Dario Fassini, Giorgio Bavestrello, Iain C. Wilkie, Francesco Bonasoro, Daniela Candia Carnevali Ecology and physiology of mesohyl creep in Chondrosia reniformis ..................................... 503-508 Solange Peixinho, Jlio Fernandez, Mara V. Oliveira, Simone Cares, Eduardo Hajdu Description of two new species of Acanthotetilla Burton, 1959 from NE Brazil, Southwestern Atlantic (Tetillidae, Spirophorida, Demospongiae) ........................................... 509-515 Henry M. Reiswig, Welton L. Lee A new species of Cladorhiza (Porifera: Cladorhizidae) from S. California (USA)................... 517-523 Pilar Ros, Javier Cristobo Sponges of genus Myxilla Schmidt, 1862, collected in Antarctic waters by Spanish Antarctic expeditions ................................................................................................................ 525-546 Pablo R.D. Rodriguez, Guilherme Muricy A new species of Cinachyra (Demospongiae: Tetillidae) collected by Project REVIZEE off Esprito Santo State, SE Brazil ........................................................................................... 547-553 Adriana Salgado, Thomz Vieiralves, Flvia R.M. Lamaro, Leonardo L.M. Assumpo, Dbora Gomes, Lia Jascone, Ana Luiza Valado, Rodolpho M. Albano, Gisele Lbo-Hajdu Field preservation and optimization of a DNA extraction method for Porifera ....................... 555-560 Susanne Schmitt, Markus Wehrl, Niels Lindquist, Jeremy B. Weisz, Ute Hentschel Morphological and molecular analyses of microorganisms in Caribbean reef adult sponges and in corresponding reproductive material ............................................................................. 561-568 Christine H.L. Schnberg, Ryota Suwa Why bioeroding sponges may be better hosts for symbiotic dinoflagellates than many corals......................................................................................................................................... 569-580

vi

Heinz C. Schrder, Anatoli Krasko, David Brandt, Matthias Wiens, Muhammad Nawaz Tahir, Wolfgang Tremel, Werner E.G. Mller Silicateins, silicase and spicule-associated proteins: synthesis of demosponge silica skeleton and nanobiotechnological applications ...................................................................... 581-592 Carla M.M. da Silva, Meiryelen V. da Silva, Bruno Cosme Redescription of the Brazilian endemic sponge Geodia glariosa (Demospongiae: Geodiidae), with new records on its geographic and bathymetric distribution ........................ 593-602 Antonio M. Sol-Cava, Gert Wrheide The perils and merits (or The Good, the Bad and the Ugly) of DNA barcoding of sponges a controversial discussion ........................................................................................ 603-612 Frank Spetland, Hans Tore Rapp, Friederike Hoffmann, Ole Secher Tendal Sexual reproduction of Geodia barretti Bowerbank, 1858 (Porifera, Astrophorida) in two Scandinavian fjords ....................................................................................................... 613-620 Robert W. Thacker, Maria Cristina Diaz, Klaus Rtzler, Patrick M. Erwin, Steven J.A. Kimble, Melissa J. Pierce, Sandra L. Dillard Phylogenetic relationships among the filamentous cyanobacterial symbionts of Caribbean sponges and a comparison of photosynthetic production between sponges hosting filamentous and unicellular cyanobacteria ............................................................................... 621-626 Carsten Thoms, Peter J. Schupp Chemical defense strategies in sponges: a review .................................................................... 627-637 Laura Valisano, Attilio Arillo, Giorgio Bavestrello, Marco Giovine, Carlo Cerrano Influence of temperature on primmorph production in Petrosia ficiformis (Porifera: Demospongiae) ........................................................................................................ 639-643 Rob W.M. van Soest, Fleur C. van Duyl, Connie Maier, Marc S.S. Lavaleye, Elly J. Beglinger, Konstantin R. Tabachnick Mass occurrence of Rossella nodastrella Topsent on bathyal coral reefs of Rockall Bank, W of Ireland (Lyssacinosida, Hexactinellida) ................................................................ 645-652 Eduardo Vilanova, Carla Zilberberg, Michele Kochem, Mrcio R. Custdio, Paulo A.S. Mouro A novel biochemical method to distinguish cryptic species of genus Chondrilla (Chondrosida, Demospongiae) based on its sulfated polysaccharides ..................................... 653-659 Author index.................................................................................................................................. 661-663 Subject index ................................................................................................................................. 665-679 Participants list ............................................................................................................................. 680-684

INVITED ARTICLES

Porifera research: Biodiversity, innovation and sustainaBility - 2007

Reading the code of coral reef sponge community composition and structure for environmental biomonitoring: some experiences from Cuba
Pedro M. Alcolado
Instituto de Oceanologa, Ave. 1ra, No. 18406, Playa, La Habana, Cuba. alcolado@ama.cu Abstract: The structure of exposed (non-cryptic) coral reef sponge communities could be considered as a potentially readable coded message reflecting their physical environment. The present paper describes explorations in Cuba of the potential use of sponge communities as bio-indicators. Clathria venosa is the sponge that most consistently has proved to be a bioindicator of urban based pollution in Cuban coral reefs due to its stenotopic character with regard to this stress source. Iotrochota birotulata forma musciformis was abundant close to the polluted Havana Bay, but not in other polluted sites, making it inconsistent as indicator. It has been quite rare in non-polluted waters. Cliona delitrix was represented in an area with great sewage influence. However it did not appear in some polluted sites probably because corals were extremely scarce and small. Scopalina ruetzleri was well represented close to bays with different degrees of urban based pollution. Cliona varians was well represented only in one polluted place. Multivariate analyses (cluster analysis, non-metric multidimensional scaled analysis) have proved to be very useful tools to clearly segregate sites with regard to level of pollution, and to identify factors and interactions determining community structure and composition. Abundance or dominance of Tectitethya crypta and Cliona vesparia (alpha stage) were typical of heavy sedimentation conditions; while Aplysina cauliformis tended to dominate in sites affected by both hurricanes and sedimentation (abundance increased by fragmentation). Meta-analysis of Shannons heterogeneity index H and Pielous equitability index J is proposed as a useful tool to classify and compare sites with regard to the way that sponges interpret their environment (degree of severity and predictability). Meta-analysis by means of a scatter graph with ranges of H at different depths provides a spatial framework for comparing and classifying sponge communities with regard to environment severity. Keywords: sponges, bio-indicators, coral reefs, Cuba

Introduction
Many papers have dealt with the factors and interactions that determine sponge distribution and community characteristics (partly reviewed by Sar and Vacelet 1973, Bergquist 1978, Wulff 2006), but few have been explicitly devoted to exploring the potential usefulness of sponge communities as bio-indicators for environmental bio-monitoring purposes. In the last few decades, the search for bio-indicators has become an urgent need in a world environment that is changing at an unprecedented rate. According to Alcolado (1984; with some added arguments), sessile taxa are suitable as potential environmental bio-indicators because: - They must be adapted to the environment due to their immobility. Thus, their abundance or their presence (or even absence) must reflect the average ecological conditions, or very recent strong stressful events. - Their composition and community structure are not affected by migrations or local displacements. - The exposed (non-cryptic) sponge communities, having passed the fish predation filter thanks to deterrence (Wulff 1997), are influenced more by the physical environment

than by ecological interactions within themselves (sensu Bradbury 1977). Cooperation rather than competition seems to be the rule among sponge populations (Sar 1970) and, according to Rtzler (1970), sponges are able to solve competition by entering into complex epizoic relationships, without detriment to their pumping and filtering activities. Reiswig (1973) adds that small sponge individuals (during the first year after settlement) are subject to severe mortality by competition with other sessile organisms, but when sponges reach greater volume competitors have little further effect. On the other hand, sponges overgrow corals much more frequently than the reverse, although when the reverse occurs, the sponge tissue shows no adverse effect (Jackson and Buss 1975). - The absence of food partitioning mechanisms influencing community structure. Such features favor sponges over many other zoological groups as potential indicators. That does not mean that there could not be some degree of influence of biological interactions, but apparently to a much lower extent than the physical environment (light, waves, sediments, pollution) in building up the community structure and composition. This

also makes community structure and composition easier to analyze and to understand in a bio-monitoring context. For these reasons, the structure of exposed (non-cryptic) coral reef sponge communities could be considered as a potentially readable coded message reflecting how sponges interpret their physical environment. Indeed, sponges have been suggested as potential environmental bio-indicators by Alcolado (1984, 1985, 1990, 1992, 1994, 1999), Alcolado and Herrera (1987), Muricy (1989, 1991), Zea (1994), Alcolado et al. (1994), Carballo et al. (1994, 1996), Carballo and Naranjo (2001), and Vilanova et al. (2004). Some attempts and successes in Cuba and other countries exploring the potential use of sponge communities as simpler, faster and lower cost bio-indicators (from a sponge life perspective) are discussed below. The results compiled in this review come from a great number of coral reef sites sampled around Cuba since 1976.

Discussion Indicator species

A few sponge species have been found to be associated with polluted or relatively unpolluted conditions in coral reefs (Table 1). Particularly, Clathria venosa (Alcolado, 1984) and Iotrochota birotulata forma musciformis (Duchassaing and Michelotti, 1864) have only been observed dominating in fore-reefs (10-20 m deep) affected by organic pollution (Alcolado and Herrera 1987) (Table 1; Fig. 1). The first species appeared to be markedly stenotopic of enriched inshore and coral reef waters and its occurrence has been very consistent in all polluted reefs evaluated or visited in northwestern Cuba (close to Havana Bay, Almendares and Quib rivers, and the town of Santa F), and according to Zea (1994), also in Santa Marta, Colombia. This species was previously reported by Hechtel (1965) on shells and piling in the enriched waters of Port Royal (southern shore of Kingston Harbor), Jamaica (as Microciona microchela n. sp.); and by van Soest (1984) in the fouling community

on dead corals and gorgonians at the bay and Hilton Hotel Landing of Curaao (as Rhaphidophlus raraechelae n. sp.). It was also found in the fouling communities of the concrete dock of Marina Barlovento (organically enriched site) and the seawall of a small organically polluted cove (Rada del Instituto de Oceanologa), both in the western Havana City, Cuba. However, I. birotulata forma musciformis was not consistently dominant or abundant in the visited Cuban polluted sites. Mycale microsigmatosa Arndt, 1927, which has been found dominating under domestic sewage stress in Brazil (Muricy 1989), was also found in a very polluted coastal lagoon at Jaimanitas Town, west of Havana city (muddy/ algal bottom) together with well developed Suberites aurantiaca (Duchassaing and Michelotti, 1964), Chondrilla aff. nucula Schmidt, 1862 and Halichondria melanadocia de Laubenfels, 1936. Both S. aurantiaca and C. aff. nucula are bacteriosponges (Rtzler 2002), which could explain their abundance in this lagoon. Holmes (1997, 2000), Holmes et al. (2000) and Rtzler (2002), comment on the increased abundance and activity of boring sponges in areas affected by urban based pollution. Indeed, Cliona delitrix Pang, 1973, a species reported as abundant in areas submitted to sewage pollution (Rose and Risk 1985, Chvez-Fonnegra and Zea 2006), was observed during four years by Marcos and Alcolado (unpublished observations) with significant relative abundance (%) at a fore-reef site close to both the polluted Quib River and a nearby sewage outfall (western Havana City). However, it was not found by Alcolado and Herrera (1987) at stations near Havana Bay, maybe due to the scarcity and small size of corals (dominated by Siderastrea radians Pallas, 1766). Another boring sponge, Cliona varians (Duchassaing and Michelotti, 1864), was well represented only in a polluted fore-reef close to both the Quib River and a nearby sewage outfall at western Havana City (Marcos and Alcolado unpublished observations). However, it was also common

Table 1: Potential indicator species and their respective inferred condition according to authors. D and M = Duchassaing and Michelotti. Dominant or abundant species Clathria venosa (Alcolado, 1984) Iotrochota birotulata f. musciformis (D. and M., 1864) Scopalina ruetzleri (Wiedenmayer, 1977) Cliona delitrix Pang, 1973 Mycale microsigmatosa Arndt, 1927 Amphimedon viridis D. and M., 1864 Aplysina fistularis (Pallas, 1766) Cliona caribbaea D. and M., 1864 Cliona vesparia (Lamarck, 1815) (alpha stage) Tectitethya crypta (de Laubenfels, 1949) Aplysina fulva (Pallas, 1766) Aplysina cauliformis (Carter, 1882) Indicated condition Organic pollution Organic pollution Moderate organic pollution Sewage pollution Sewage (bacterial) pollution Sewage pollution Sewage pollution Comparatively non-polluted Comparatively non-polluted Sedimentation plus wave stress Sedimentation stress Strong waves Eventual strong waves and sedimentation Author Alcolado and Herrera (1987) Alcolado and Herrera (1987) Alcolado and Herrera (1987); Muricy (1989); Zea (1994) Muricy (1989) Rose and Risk (1985) Muricy (1989) Muricy (1989) Alcolado (present paper, Fig. 1) Lpez-Victoria and Zea (2004) Alcolado (present paper) Alcolado and Gotera (1985) Wulff (1995) Alcolado (present paper)

Fig. 1: Relative abundances of potential pollution bioindicators species, presented as percentages of total sponge abundance (number of individuals), at stations located at different distances from two main pollution sources in the north-western Cuba (Havana Bay and Almendares River). Mariel Bay is not significantly polluted.

in non-polluted reef areas, which makes it inconsistent as a potential bio-indicator. Lpez-Victoria and Zea (2004) showed that the abundance of Cliona caribbaea is not related to pollution in San Andrs Archipelago, Colombia. Indeed, this species did not occur at sites close to the organically polluted Quib River and the nearby sewage outfall, but only in more distant sites (Marcos and Alcolado, unpublished observations). Other sponge species have been associated with factors other than pollution, namely sedimentation and wave stress (Table 1). Particularly, Aplysina cauliformis is apparently tolerant to strong waves, as can be deduced from its dominance in coral reef sites exposed to more frequent tropical storms (keys Juan Garca and Cantiles, southwestern Cuba). This can be due to its branching morphology, flexibility and elasticity, similar to what was suggested by Wulff (1995) for Aplysina fulva, also branching and with rather similar consistency. The suggested usefulness of the presence or abundance of some sponges as environmental indicators has been based much on expert observation and on inferences related to distance from known pollution sources, wave and wind exposure, visual evidence of varying intensity of sedimentation, etc. For that reason, to validate these results and make further progress, more evidence is necessary, obtained both from well designed experiments and from multivariate analysis in which factors are directly measured on appropriate temporal and spatial scales. Additionally, more sites in the Wider Caribbean, suffering various degrees of pollution, tropical storm frequency, exposure to waves and dominant winds, etc., are worth being researched to test the generality of the mentioned findings. It would be of particular

interest to determine if Clathria venosa feeds on bacteria with emphasis on enteric taxa, as does Clathria prolifera (Ellis and Solander, 1786) according to Claus et al. (1967).

Community indices
In agreement with other authors (Muricy 1989, Carballo et al. 1996, among others), Alcolado and Herrera (1987) found that species richness and Shannons heterogeneity index H were lower at more polluted sites (Fig. 2). Pielous equitability index J was also lower in the more polluted sites close to the mouth of Almendares River (Fig. 2). Given that a condition of significant stress can be inferred only when the dominance of some of the mentioned indicator species (Table 1) is coupled with low values of species richness or species heterogeneity (Alcolado et al. 1994), these univariate indices have to be taken into account as an important complement for environmental monitoring. The summing up of the numerical percentages of individuals belonging to species that are tolerant to the same kind of stressor (e.g., pollution, sedimentation, turbulence, etc.) could be useful as another potential community index for monitoring purposes, as done by Alcolado (1981) with gorgonians to infer relative turbulence intensity, and by Herrera-Moreno (1991), also with gorgonians, to infer relative organic pollution level. The usefulness and conceptual validity of diversity indices has been controversial (Hurlbert 1971, Peet 1974). However (without disregarding potential pitfalls), the herein explored diversity indices can be used and tested pragmatically and heuristically for bio-monitoring purposes in the context of environmental management, not specifically for advancing

Fig. 2: H and J in stations located at different distances from two main pollution sources in the north-western Cuba (Havana Bay and Almendares River). Mariel Bay is not significantly polluted.

science (but being increasingly supported by scientific research). The same validation and progress efforts can be applied to community indices.

Environmental severity and predictability inference graph

Preston and Preston (1975) deserve the credit for integrating and applying theoretical criteria from classic community ecology in a simple and practical scheme to infer the environmental severity and predictability (constancy) for comparative purposes. According to Margalef (1963, 1968) and Odum (1969), high species diversity and high equitability are generally associated with mature, late successional stages. On the other hand, Sanders (1969) and Slobodkin and Sanders (1969) hypothesize that severe environments generally permit less diversity to develop than favorable ones do. As suggested by Slobodkin and Sanders (1969) and supported by Preston and Preston (1975), the degree of stress is determined primarily by the degree of temporal predictability of environmental conditions and the degree of physiological stress imposed by the physical environment. After Preston and Preston (1975), by means of the values of H and J three different ecological situations can be inferred: high values of both indices suggest a favorable and predictable environment; a low H coupled to a high J indicates a constantly severe environment; and low values of both indices reflect an unpredictably severe environment. Due to the fact that this scheme excessively reduces the real variety of situations and leads to misinterpretation of

the specific case of extremely low values of both indices, it was modified by Alcolado (1992) for sponges (Fig. 3). This modification consisted of an inference diagram obtained from a scatter graph of pairs of H and J values from 112 sites. The resulting scatter area was divided into 11 environmental inference zones or classes reflecting corresponding ecological situations instead of Preston and Prestons (1975) original three zones or classes (constantly favorable, constant or temporally predictable stress and unpredictable stress). Except for the class 1 of the scale, which is a qualitative addition to the original scheme, the remaining classes resulted from subdividing the original overly inclusive classes. This resulted in a finer grain of environmental situations to infer, matching more closely the relatively wide range of withinclass environmental variability perceived in the field by the author within the original three classes (e.g., sponge size, coral size and cover, wind and wave exposure, etc.). The number of subdivisions preferred among different persons would certainly vary, as happens with the different temperature measurement scales (Celsius, Fahrenheit and Kelvin). The new scale comprises the following environmental severitypredictability classes: 1 = environment extremely stressed by both, a constant basal level of disturbance and intermittent unpredictable strong events (H = 0-1.3 natural bells; J = 00.5); 2 = very severe and unpredictable environment (H = 0-1.3 natural bells; J = 0.5-0.69); 3 = severe and unpredictable environment (H = 1.3-2.0 natural bells; J = 0.5-0.69); 4 = almost constantly severe environment (H = 1.3-2.0 natural bells; J = 0.7-0.8); 5 = constantly severe environment

(e.g., Cliona aprica, among sponges, Gorgonia flabellum Linnaeus, 1758, among gorgonians, and Acropora palmata Lamarck, 1816, among scleractinians). For that reason, both J an H show extremely low values. This combination, within Preston and Prestons (1976) original scheme, would suggest an unpredictable environment (with its constant component omitted). Alcolados (1992) inference diagram also differentiates the very favorable and constant environments of the deep reefs (e.g., at 20-30 m) within the rank 11, from those that are simply favorable and quasi-constant (rank 10). Nevertheless, it is advisable to be aware of specific situations of very longterm environmental stability where some species can escape from demographic control and become excessively dominant, and consequently diminishing H and J. This situation is common at reef sites deeper than 25 m. This phenomenon of community senescence is not contemplated in either of the two mentioned inference methods and has to be taken into account in supposedly extremely constant environments (e.g., deep reef zones, and reefs where hurricanes are very rare, as those of Bonaire and Tobago). The authors scale is proposed as an alternative reference (among other possible ones) and could be tested and improved with further research. More sites across the Wider Caribbean should be studied and included in the scatter graph to refine its spatial contour.

Scatter graphs for comparing community indices at different depths

Fig. 3: Inference diagram reflecting eleven ways in which sponges interpret their physical environment, derived from a meta-analysis with 112 coral reef stations. 1 = extremely severe with mixture of constant and unpredictable environment; 2 = very severe and unpredictable; 3 = severe and unpredictable; 4 = quasiconstantly severe; 5 = constantly severe; 6 = moderately severe and unpredictable; 7 = moderately severe and quasi-constant; 8 = moderately severe and constant; 9 = favorable and quasi-constant; 10 = constantly favorable; and 11 = very favorable and constant.

(H = 1.3-2.0 natural bells; J = 0.8-1); 6 = moderately and unpredictably severe environment (H = 2.0-2.5 natural bells; J = 0.5-0.69); 7 = moderately and almost constantly severe environment (H = 2.0-2.5 natural bells; J = 0.7-0.8); 8 = moderately and constantly severe environment (H = 2.0-2.5 natural bells; J = 0.8-1); 9 = favorable and almost constant environment (H = 2.5-2.9 natural bells; J = 0.7-0.8); 10 = favorable and constant environment (H = 2.5-2.9 natural bells; J = 0.8-1); and 11 = very favorable and constant environment (H >2.9 natural bells; J = 0.8-1). Rank 1 shows a (qualitative) situation that is not considered by Preston and Preston (1975). This is the case of the surf zones of some Cuban reefs, which have constant average (basal or chronic) conditions of fairly strong wave action, but which are also unpredictably affected by severe impacts of tropical storms. Under these circumstances, there is only one predominant species within each sessile taxon

Scatter graphs of variability of sponge diversity indices, population density and cover with regard to depth were obtained for many Cuban reef sites (Alcolado 1994, 1999). These graphs, which display the area (range) of variation of those indices with regard to depth, can be used as a reference pattern to infer in a comparative way the community condition within stress gradients, taking into account site depth, given that such indices do not necessarily behave in the same way along depth gradients. What is normally a moderate value of H for a given depth could be considered a high value for a lower depth. The upper border of the variation area (an ascending convex line with a slight diminution at depths greater than 25 m) reflects the best conditions registered at different depths for sponge species richness and species heterogeneity (Fig. 4), while the lower border (an asymptotically ascending curve) shows the worst environmental conditions at different depths (Alcolado 1994). Care must be taken at deep reef stations (about 30 m depth or more), as lower diversities can be caused by extremely constant and favorable conditions that lead to the dominance of competitively stronger species, and not by any stressor. The same recommendations given for the environmental severity and predictability inference graph are applicable here.

Classification and ordination

Classification (Fig. 5) and ordination (Fig. 6) analyses have proved to be useful when using sponge communities to separate sites with regard to degree of pollution and

Fig. 4: Example of meta-analysis as a scatter graph of H values at different depths, with classification bands of inferred environmental conditions (from 112 coral reef sites of Cuba). Care must be taken at stations about 30 m depth or more, as lower diversities can be determined by extremely constant and favorable conditions leading to dominance of very competitive species, and not by severe environmental conditions. Arrows indicate more polluted stations (10 m depth) and stations affected by sediments (deeper stations).

Fig. 6: MDS analysis segregating stations located at different distances from two main pollution sources (Havana Bay and Almendares River) with regard to degree of pollution (from left to right: very polluted, polluted, and little polluted). This analysis was done with quadratic transformation of sponge densities and Bray Curtis similarity Index.

Fig. 5: Cluster analysis segregating stations located at different distances from two main pollution sources (Havana Bay and Almendares River) with regard to degree of pollution (going downwards: very polluted, polluted, and little polluted). This analysis was done with quadratic transformation of sponge densities, Bray Curtis similarity Index, and un-weighted paired average clustering.

serve also to strengthen the validation and to reduce the pitfalls of potential indicator species and ecological indices that have been proposed, to a great extent based on observational and inference approaches. In the context of the application of the suggested biomonitoring methods, an aspect that deserves future effort is to assess the convenience of using sponge cover instead of sponge density, both from practical and scientific points of view. Finally, another matter of concern could be the need of sponge taxonomy skills for the implementation of the proposed bio-monitoring methods. In this sense, the potential indicator species are easy to identify in situ, and sampling for calculation of community indices would only require differentiation of species, and not necessarily identification to species level. With some practice, the identification of most common species can be learned and the sampling work can become even easier.

Acknowledgements
I would like to thank the National Museum of Rio de Janeiro, PETROBRAS, the UNDP/GEF Project SabanaCamagey and Dr. Robert N. Ginsburg (Ocean Research and Education Foundation) for making possible my participation in the 7th International Sponge Symposium. I am grateful to Dr. Janie Wulff, Dr. Georgina Bustamante and Marta Rivero for their valuable comments to the manuscript.

to explore the factors impinging on their structure and composition (Alcolado and Herrera 1987, Muricy 1989, 1991, Carballo et al. 1994, 1996, Carballo and Naranjo 2001, Bell and Barnes 2003, Vilanova et al. 2004, Marcos and Alcolado unpublished observations). Particularly, the Multidimensional Scaled analysis (MDS) has provided clear results (PRIMER version 5). Multivariate techniques are useful tools for identifying factors and interactions, and as such have to be applied complementarily with simpler, faster and lower cost univariate inference approaches in environmental monitoring. Multivariate analyses have to

References
Alcolado PM (1981) Zonacin de los gorgonceos someros de Cuba y su posible uso como indicadores comparativos de tensin

hidrodinmica sobre los organismos del bentos. Inf Cient-Tc Inst Oceanol Acad Cien Cuba 187: 1-43. Alcolado PM (1984) Utilidad de algunos ndices ecolgicos estructurales en el estudio de comunidades marinas de Cuba. Cien Biol 11: 61-77 Alcolado PM (1985) Estructura ecolgica de las comunidades de esponjas de Punta del Este, Cuba. Rep Invest Inst Oceanol Acad Cien Cuba 38: 1-65 Alcolado PM (1990) General features of Cuban sponge communities. In: Rtzler K (ed). New perspectives in sponge biology. Smithsonian Institution Press, Washington DC. pp. 351-357 Alcolado PM (1992) Sobre la interpretacin del medio marino mediante el empleo de los ndices de diversidad y equitatividad. Cien Biol 24: 124-127 Alcolado PM (1994) General trends in coral reef sponge communities of Cuba. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 251-255 Alcolado PM (1999) Comunidades de esponjas de los arrecifes del Archipilago Sabana-Camagey, Cuba. Bol Invest Mar Cost 28: 95-124 Alcolado PM, Gotera GG (1985) Estructura de las comunidades de esponjas en arrecifes cubanos. Contrib Simp Cien Mar, VII Jornada Cient Inst Oceanol, XX Aniv 1: 11-15 Alcolado PM, Herrera A (1987) Efectos de la contaminacin sobre las comunidades de esponjas en el Litoral de La Habana, Cuba. Rep Invest Inst Oceanol Acad Cien Cuba 68: 1-17 Alcolado PM, Herrera-Moreno A, Martnez-Estalella N (1994) Sessile communities as environmental bio-monitors in Cuban coral reefs. In: Ginsburg RN (ed). Proceeding of the Colloquium on Global aspects of coral reefs: health hazards and history, 1993 RSMAS, University of Miami. pp. 27-33 Bell JJ, Barnes DK (2003) Effect of disturbance on assemblages: an example using Porifera. Biol Bull 3(1): 144-159 Bergquist PR (1978) Sponges. Hutchinson & Co., London Bradbury RH (1977) Independent lies and holistic truths: towards a theory of coral reefs communities as complex systems. Proc 3rd Int Coral Reef Symp, Miami 1: 2-17 Carballo JL, Snchez-Moyano JE, Garca-Gmez JC (1994) Taxonomic and ecological remarks on boring sponges (Clionidae) from the Strait of Gibraltar (southern Spain): tentative bioindicators? Zool J Linnean Soc 112: 407-424 Carballo JL, Naranjo SA, Garca-Gmez JC (1996) Use of marine sponges as stress indicators in marine ecosystems at Algeciras Bay (southern Iberian Peninsula). Mar Ecol Progr Ser 135: 109-122 Carballo JL, Naranjo S (2001) Environmental assessment of a large industrial marine complex based on a community of benthic filter feeders. Mar Poll Bull 44: 605-610 Claus G, Madri P, Kunen S (1967) Removal of microbial pollutants from waste effluents by the redbeard sponge. Nature 216: 712714 Chvez-Fonnegra A, Zea S (2006) Observation of reef coral undermining by the Caribbean excavating sponge Cliona delitrix (Demospongiae, Hadromerida). In: Custdio MR, Lbo-Hajdu G, Hajdu E, Muricy G (eds). 7th International Sponge Symposium Book of Abstracts (Armao dos Bzios, Brazil). Museu Nacional, Srie Livros, vol. 16. p. 324 Hechtel G (1965) A systematic study of the Demospongiae of Port Royal, Jamaica. Bull Peabody Mus nat Hist 20: 1-94

Herrera-Moreno A (1991) Efectos de la contaminacin sobre la estructura ecolgica de los arrecifes coralinos en el litoral Habanero. PhD Thesis. Instituto de Oceanologa, La Habana Holmes KE (1997) Eutrophication and its effect on bioeroding sponge communities. Proc 8th Int Coral Reef Symp, Balboa 2: 1411-1416 Holmes KE (2000) Effects of eutrophication on bioeroding sponge communities, with description of new West Indian sponges Cliona spp. (Porifera, Hadromerida: Clionaidae). Invert Biol 119: 125138 Holmes KE, Edinger EN, Hariyadi, Limmon GV, Risk MJ (2000) Bioerosion of Live Massive Corals and Branching Coral Rubble on Indonesian Coral Reefs. Mar Poll Bull 40(7): 606-617 Hurlbert SH (1971) The nonconcept of species diversity: A critique and alternative parameters. Ecology 52: 577-586 Jackson JBC, Buss L (1975) Allelopathy and spatial competition among coral reef invertebrates. Proc Natl Acad Sci USA 72: 51605163 Lpez-Victoria M, Zea S (2004) Current trends of space occupation by encrusting excavating sponges on Colombian coral reefs. Mar Ecol 26: 33-41 Margalef R (1963) On certain unifying principles in ecology. Am Nat 97: 357-374 Margalef R (1968) Perspectives in ecological theory. University of Chicago Press, Chicago Muricy G (1989) Sponges as pollution bio-monitors at Arraial do Cabo, Southeastern Brazil. Rev Bras Biol 49(2): 347-354 Muricy G (1991) Structure des peuplements de spongiaires autour de lgout de Cortiou (Marseille, France). Vie et Milieu 41(4): 205221 Odum EP (1969) The strategy ecosystem development. Sciences 164: 262-270 Peet RK (1974) The measurement of species diversity. Ann Rev Ecol Syst 5: 285-307 Preston EM, Preston JL (1975) Ecological structure in a West Indian gorgonian fauna. Bull Mar Sci 252: 248-258 Reiswig HM (1973) Population dynamics of three Jamaican Demospongiae. Bull Mar Sci 23: 191-226 Rose CS, Risk MG (1985) Increase in Cliona delitrix infestation of Montastraea cavernosa heads on an organically polluted portion of the Grand Cayman fringing reef. Mar Ecol 6(4): 345-363 Rtzler K (1970) Spatial competition among Porifera: solution by epizoism. Oecologia 5: 85-95 Rtzler K (2002) Impact of crustose clionid sponges on Caribbean reef corals. Acta Geo Hisp 37(1): 61-72 Sanders HL (1969) Benthic marine diversity and the stability-time hypothesis. Brookhaven Symp Biol 22: 71-81 Sar M (1970) Competition and cooperation in sponge populations. Symp Zool Soc Lond 25: 273-284 Sar M, Vacelet J (1973) cologie des Demosponges. In: Grass PP (ed). Trait de Zoologie, Vol. 3, Masson et Cie, Paris. pp. 462-576 Slobodkin LB, Sanders HL (1969) On the distribution of environmental predictability to species diversity. Brookhaven Symp Biol 22: 82-93 van Soest RWM (1984) Marine sponges from Curacao and other Caribbean localities. Part. III. Poecilosclerida. Stud Fauna Curacao Caribb Isl 66(199): 1-167 Vilanova E, Mayer-Pinto M, Curbelo-Fernandez MP, Gonalves da Silva SH (2004) The impact of a nuclear power plant discharge on the sponge community of a tropical bay (SE Brazil). In: Pansini M,

10

Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Boll Mus Ist Biol Univ Genova 68: 647-644 Wulff JL (1995) Hurricane effects on survival and orientation of large erect coral reef sponges. Coral Reefs 14: 55-61 Wulff JL (1997) Parrotfish predation on cryptic sponges of Caribbean coral reefs. Mar Biol 129(1): 41-52

Wulff JL (2006) Ecological interactions of marine sponges. Can J Zool 84(2): 146-166 Zea S (1994) Patterns of coral and sponge abundance in degraded versus still healthy coral reefs at Santa Marta Colombian Caribbean. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 257-264

Porifera research: Biodiversity, innovation and sustainaBility - 2007

11

Fossil sponges of Argentina: a review

Matilde Sylvia Beresi
CONICET-CRICYT: Ianigla, Dto. de Geologa y Paleontologa, Avda Ruiz Leal s/n, 5500 Mendoza, Argentina. mberesi@lab.cricyt.edu.ar Abstract: This is a review on fossil sponges and sponge spicules reported from several regions in Argentina and in strata ranging in age from Early Cambrian to Tertiary. Sponges have been collected from marine sediments of the Puna, Cordillera Oriental and Sierras Subandinas basins, northern Argentina; Famatina Range; Precordillera terrane, San Rafael block, Neuqun basin and from lacustrine deposits of the North Patagonian Massif. Knowledge of the sponge fossil record is based on whole relatively rigid skeletons, fragments of skeletal nets and spicules seen in thin sections or recovered from acetic acid residues. Early to Middle Cambrian Porifera and Chancelloriids are known from the carbonate platform and slope facies of the Precordillera terrane. Specimens with body preservation of Protospongia, Diagoniella, Kiwetinokia, fragments of hexactinellid, and anthaspidellid sponges and sclerites of Chancelloria had been reported from Cambrian of the Precordillera. Remains of hexactinellid sponges, Pelicaspongiidae and Protospongiidae, have been found in Ordovician rocks of the Puna and of the Famatina System, western margin of Gondwana. Protospongia sp. and hexactinellid mesh were reported from Upper Cambrian-Lower Ordovician siliciclastic sediments in the Cordillera Oriental and Sierras Subandinas. The most significant fossil record of Lower-Middle Ordovician sponge faunas is from the carbonate platform of the San Juan Precordillera. Sponge faunas are dominated by orchoclad lithistid demosponge genera, although hexactinellids are known from loose spicules and root tufts, and calcareous heteractinid sponges are known from isolated octactine spicules and only one genus. Hexactinellid, calcarean and demosponge spicules were reported from diverse localities of the Precordillera. A Late Jurassic (Oxfordian) carbonate complex of the Neuqun Basin, west-central Argentina, contains siliceous sponges dominated by hexactinellids (Hexactinosa and Lyssakinosa). Palaeospongilla chubutensis, a fresh water sponge, was described from lacustrine Cretaceous deposits of the Chubut River valley. Oxeas and strongyles, belonging to the Family Spongillidae, have been mentioned from Tertiary sediments of the Paran basin, northeastern Argentina. Keywords: biostratigraphy, fossil sponges, Argentina.

Introduction
The actual knowledge of the record fossil in Argentina is based on relatively rigid whole skeletons, fragments of skeletal nets and spicules seen in thin sections or recovered from acid residues for obtaining conodonts. The fossil record of sponges comes from several geological provinces of Argentina with different lithologic, paleontologic and environmental characteristics. Sponge faunas were collected from marine sediments of the Puna, Cordillera Oriental, Sierras Subandinas basins of northern Argentina; Famatina Range, Precordillera Terrane, San Rafael Block, Neuqun Basin (western Argentina), from lacustrine deposits of the North Patagonian Massif, and from the Chaco-Paranaense Basin. Occurrences of sponge faunas have been reported from the Lower Paleozoic (upper-Lower Cambrian) to the Cenozoic (Tertiary). Only a few publications have dealt with sponge and sponge spicules in Argentina. Most of them concern the fauna from the Cambrian and Ordovician rocks of the Precordillera terrane. Associated with protospongiid spicules, well-preserved chancelloriid sclerites occur in the Cambrian carbonate platform of the Precordillera. Chancelloriids are Cambrian

enigmatic organisms constituting the monophyletic taxon Coeloscleritophora (Bengtson and Missarzhevsky 1981). Sclerites of chancelloriids (Family Chancelloridae) were first described from the Burgess Shale by Walcott (1920), who interpreted them as heteractinid sponges. This traditional view of the fossil group as sponges was accepted for more than 50 years. In the Precordillera, chancelloriid sclerites associated with spicules are a common and distinguishing features of the Lower to Middle Cambrian fossil fauna. Protospongia and Chancelloria crucensis (Rusconi, 1955) were the first Cambrian species of the Mendoza Precordillera. Cambrian Protospongia Salter, 1864 and hexactinellid spicules were mentioned by Pernas (1964), Devizia (1973), Bordonaro and Martos (1985), Heredia et al. (1987) and Beresi and Heredia (1995) from the San Isidro area in the Precordillera of Mendoza. Afterwards, assemblages tentatively identified as Kiwetinokia utahensis Walcott, 1920, protospongiid skeletal nets and Chancelloria eros Walcott, 1920 were reported from early-Middle Cambrian carbonates blocks of San Juan and Mendoza, Precordillera, by Beresi and Rigby (1994). Two small associated specimens assigned to Protospongia sp. occur in the western Precordillera of San

12

Juan (Beresi and Banchig 1997). A synthesis of the Cambrian sponge occurrence in the Argentine Precordillera was given by Beresi (2003a). Silicified sponge spicules from residues of conodonts from diverse Cambrian and Ordovician sections in the Precordillera, were described by Mehl and Lehnert (1997). Anthaspidellid skeletal fragments from the Middle Cambrian rocks of the Precordillera, in the province of San Juan, document the only known occurrence from South America (Beresi and Rigby 1994). Protachileum kayseri Zittel, 1877 from the San Juan province, was the first report of a Precordilleran sponge. Taxonomic studies have been concentrated in the rich sponge fauna of the warm carbonate platform (San Juan Formation) from the Lower-Middle Ordovician of the Precordillera basin (Beresi and Rigby 1993, Carrera 1996a, 1996b). The purpose of this paper is to review the occurrence of fossil sponges and their biostratigraphic distribution in the diverse geological provinces of Argentina.

Cambrian and Ordovician fauna from the Northwest Argentinian region

Siliciclastic sediments with minor carbonates dominate the Cambrian and Ordovician of Northern Argentina and the Central Andean Basin of South America. Cambrian and Ordovician sponges and spicules from Northwest Argentina provide additional paleontological data from the siliciclastic platforms of western Gondwana.

Puna
In the Puna region (Fig. 1A-B), a single specimen (Fig. 2M-N) of a complete Ordovician hexactinellid sponge was discovered. It has been collected from volcaniclastic rocks of the Las Vicuas Formation (Tremadocian) in Lari Creek, southwest of the Salar del Rincon area, Salta province, Argentina. The material was assigned to the new genus and species Larispongia magdalenae (Carrera, 1998), that belong to the family Pelicaspongiidae, and it is the first record of the family in western Gondwana.

Subandean Ranges
The first report on the occurrence of one completely preserved hexactinellid sponge is a part and counterpart of a round small sponge described as Diagoniella sp. (Beresi et al. 2006). This sponge (Fig. 2H) was reported for the first time from the Upper Cambrian-Lower Ordovician siliciclastic rocks of the Orcomato Formation, in the Candelaria Range, Sub-Andean Ranges, in Salta Province (Fig. 1B). The fossiliferous record of the unit suggests a CambrianOrdovician age. The sponge is preserved in greenishyellowish shales, as dark stained flattened rounded bodies (somewhat deformed and fragmentary).

Fig. 1: A. South America showing the position of Argentina. B. Location of the Basins of Argentina with sponge fauna: (Pu) Puna; (CO) Cordillera Oriental, (SS) Sierras Subandinas, (F) Famatina, (Pa) Paran basin, (P) Precordillera, (SR) San Rafael Block, (N) Neuqun basin, (MP) North Patagonian Massif. C. Map of Argentine showing the San Juan and Mendoza provinces. D. Sponge localities within the Precordillera Precordillera of San Juan and northern Mendoza provinces.

Eastern Cordillera
Hexactinellid meshes of Protospongia (Beresi et al. 2006) were recently reported from Lower Tremadocian siliciclastic sediments in the Eastern Cordillera, Salta Province (Fig. 1). The hexactinellid spicules were recovered from the lower

13

Fig. 2: A. Cambrian anthaspidellid sponge fragment, genus and species indeterminant from San Isidro, Mendoza. B-G. Coeloscleritophora: Chancelloria eros Walcott, 1920, sclerites from the upper LowerMiddle Cambrian of Zonda Range and San Isidro area. H. Diagoniella sp. from Salta province, SubAndean ranges. I. Cambrian fragment of bioclastic limestone showing sclerites of Chancelloria Walcott, 1920, San Isidro. J. Root tuft, large monaxons or monactine-like spicules, in bundle with small stauracts or hexacts (Tontal Range). K. Protospongia sp. showing stauractine-based skeleton and long rayed hexactines forming marginal spines along the left margin (Tontal Range). L. Protospongia sp. Esteban and Rigby, 1998, specimen PIL 14.192 from Pea Negra section, Famatina region. M-N. Larispongia magdalenae Carrera, 1998, holotype CEGH-UNC 17365 from Lari Creek, Puna region. N. Detail of the same specimen showing dermal hexactines surrounding major gaps.

levels of the Saladillo Formation, at the Angosto de la Quesera section, Eastern Cordillera. Spicules of Protospongia (Fig. 2K) are preserved on upper surfaces of yellow-brownish shales and sandstones belonging to the lower section of the unit, sharing the stratigraphic position with Tremadocian graptolites. The sedimentation of the greenish and yellowish shales and sandstones of the Saladillo Formation indicates a transition to an upper offshore-lower offshore environmental setting.

which have been described by Aceolaza and Toselli (1977) for the Chaschuil region. The material appears dispersed in carbonate concretions of the Suri Formation (Arenig).

La Rioja Province
The sponge material came from black siliceous graptolitic shales of the upper part of the Volcancito Formation of Lower Tremadocian age, outcropping in the Pea Negra location in the Famatina range (Fig. 1B). Fragments of a reticulated skeletal net of Protospongia species (A and B) were described by Esteban and Rigby (1998), in the siliclastic Famatina basin, western margin of Gondwana (Fig. 2L). The sponges are associated with planktonic graptolites and this level can be assigned to the Lower Tremadoc (Esteban and Gutierrez-Marco 1997).

Famatina system Catamarca Province

The only discovery of Ordovician sponges in this province corresponds to isolated hexactinellid spicules,

14

Precordillera (Cuyania) Terrane, Western Argentina

The Argentine Precordillera is situated along the forefront of the high Andes at approximately 28 to 37 S, and it is a major geologic province in northwestern Argentina. It contains a complete thick sequence of Early Paleozoic rocks. The Precordillera, as part of the Cuyania Terrane, was formed during the Andean (Tertiary) crustal shortening. This distinctive terrane can be recognized mainly on the basis of its key stratigraphic composition, involving biostratigraphic, sedimentary and magmatic events; its boundaries with adjacent geologic regions are abrupt (Ramos et al. 1986). In accordance with the Terrane concept, the present Precordillera, plus the San Rafael Block and San Jorge Limestones, integrate a unique geologic entity, the so-called Precordillera Terrane or Cuyania Terrane. Two hypotheses exist regarding the origin of the Precordillera: 1) the Precordillera represents a terrane of Laurentian origin that became attached to Gondwana (western Argentina) already during Ordovician times (Thomas and Astini 2003). It includes either the classical Precordillera as well as the San Rafael Block, to the south in the province of Mendoza, and the San Jorge Limestones cropping out in the Province of La Pampa, within the Sierras Pampeanas structural setting as an allochthonous terrane Cuyania, accreted to Gondwana during the lower Paleozoic. 2) the Precordillera is considered as an autochthonous Gondwanan fragment (Baldis et al. 1989, Aceolaza et al. 1999, 2002) displaced by simple transcurrence mechanics, from a hypothetical intermediate sector between South America, Africa, and Antarctica. Recently, U-Pb geochronology of detrital zircons indicated a Gondwanan provenance for Lower Cambrian and Upper Ordovician sandstones of the Precordillera of western Argentina, supporting the autochthonous Gondwanan nature of the Precordillera Terrane (Finney et al. 2003a, 2003b).

the Precordillera of San Juan and Mendoza. Well preserved sclerites are associated with protospongiid spicules (Fig. 2BG).

Cambrian assemblages
Two spicule assemblages occur in the Cambrian facies (Beresi and Rigby 1994, Beresi 2003a). The autochthonous assemblage corresponding to material collected from the upper Lower to Middle Cambrian platform sequence of the eastern Precordillera of San Juan. This assemblage consists of a variety of stauractines and sclerites of Chancelloria eros (Walcott, 1920). The Protospongiidae are represented by triradiate prodianes, pentactines and hexactines, all belonging to Kiwetinokia utahensis Walcott, 1920, Protospongia and anthaspidellid fragments. This fossil fauna represents the oldest assemblages known of Argentina. The allochthonous assemblage proceeds from the diverse Cambrian carbonate olistoliths of slope sequences of the western San Juan Precordillera and from the classical area of San Isidro, Mendoza Province. The assemblages consist of the first precordilleran Protospongia with body preservation, Diagoniella Rauff, 1894, Kiwetinokia Walcott, 1920 and Chancelloria and skeletal net with hexactines and monaxons (Beresi and Banchig 1997). Demosponges have a limited record in the Cambrian of the Precordillera. Typically anthaspidellid fragments with dendroclones (Fig. 2A) have been reported from the carbonate platform and slope sequences of the San Juan and Mendoza Precordillera (Beresi and Rigby 1994).

Ordovician Porifera
Deposits of Ordovician carbonate basins occur in the central and eastern Precordillera. The Lower-Middle Ordovician sediments of the Precordillera represent a drowning carbonate platform with a diverse and relatively complete fossil record. Well-preserved and diverse faunas of sponges have been collected from limestones of the San Juan Formation (Upper Tremadoc-Early Llanvirn) in the Precordillera basin of San Juan province (Fig. 1C-D). This fauna represents the most significant Ordovician sponge fauna known from South America and provides the first extensive record of sponges derived from a stable carbonate platform, constituting one of the most important Early Ordovician sponge associations of the world. Precordilleran sponge faunas are dominated by orchoclad lithistid demosponge genera, although hexactinellids are known from loose spicules and root tufts, and calcareous heteractinid sponges are known from isolated octactine spicules and only one genus. Spicules assemblages were reported from diverse localities of the Precordillera (Beresi and Esteban 2003, Carrera 2003). The San Juan Formation was deposited on an open carbonate shelf, bounded to the west by continental slope and oceanic basin deposits. The diverse marine fauna and the lack of specific structures indicative for shallow water

Cambrian Porifera
The Cambrian system of the Argentinian Precordillera is represented by a carbonate platform, in the east, and of a continental slope, in the west. Cambrian platform and slope facies, containing spicules and chancelloriids sclerites are located in the Precordillera of San Juan and northern Mendoza provinces, western Argentina (Fig. 1C-D). Cambrian sponges are known mainly from fragments of skeletal nets and dissociated spicules from the shallow carbonate platform sequences of the upper-Lower to Middle Cambrian in the eastern and central belts, and from the slope olistostromic sequences with allochthonous blocks in the western part of the Precordillera of San Juan Province. The spicules from the Upper Cambrian were collected in Tontal Range, San Juan Precordillera (Beresi and Banchig 1997) and in the La Cruz Olistolith, San Isidro area (Beresi and Heredia 1995), Precordillera of Mendoza Province. Sclerites of Chancelloria (Coeloscleritophora) occur in shallow carbonate platforms and allochthonous blocks in

15

conditions point to low energy, subtidal conditions within an open-platform environment during the entire sedimentation interval. The age of the San Juan Formation is well constrained by conodonts (Sarmiento 1985, Albanesi and Ortega 2002) spanning from the late Tremadocian (Paltodus deltifer Zone) to the early Darriwilian (Lenodus variabilis Eoplacognathus suecicus zones). The first report of a Precordilleran sponge was Protachileum kayseri Zittel, 1877 from the Talacasto Gulch, Precordillera of San Juan Province (Fig. 3J). Anthaspidellid genera first appear in the basal beds of the San Juan Formation (Upper Tremadoc) at the Niquivil Range, Eastern Precordillera of San Juan, associated with

reef-mound (Caas and Carrera 1993). The cosmopolitan Archaeoscyphia Hinde, 1889 and Rhopalocoelia Raymond and Okulitch, 1940 are the predominant genera in this sponge-algal association (Fig. 3H, 3E). Diverse and well-preserved sponge faunas are from the middle and upper part of the San Juan Formation (ArenigLower Llanvirn). In this carbonate platform the fauna is dominated by orchocladine lithistid demosponges. Their first taxonomic study was made by Beresi and Rigby (1993) and afterwards by Carrera (1996a, 1996b, 1998). Apart from orchocladine demosponges there are hexactine spicules, hexactinellid root tufts, and isolated octactine spicules that document the presence of the Heteractinida.

Fig. 3: Early and Middle Ordovician orchoclad lithistid sponges from the Precordillera of San Juan. (A-B-C-D-G-K; Beresi and Rigby 1993). A. Talacastonia chela Ianigla PI T-2. B. Tangential thin section. C. Anthaspidella annulata Ianigla PI T-49. D. Calycocoelia perforata Ianigla PI VI-2. E. Rhopalocoelia clarkii Raymond and Okulitch, 1940, Ianigla PI T-22. F. Hudsonospongia cyclostoma Raymond and Okulitch, 1940, Ianigla PI VI-2. G. Hudsonospongia talacastensis. Ianigla PI T-32. H. Archaeoscyphia minganensis Billings, 1859 Ianigla PI T-47. I. Incrassospongia ramis Carrera, 1996b, CEGH-UNC 9308. J. Protachileum kayseri Zittel, 1877, Ianigla PI H-43. K. Aulocopium sanjuanensis Ianigla PI VI-13.

16

The greatest generic and specific diversity of lithistid sponges occurs during the Lower Llanvirnian (Darriwilian) in the upper most part of the San Juan Formation. A total of 15 demospongiid genera and 20 species are described. This fauna shows a variety of external morphologies and body plan. Many new Ordovician species were described for the San Juan Formation by Beresi and Rigby (1993): Anthaspidella inornata, A. annulata (Fig. 3C), and A. alveola, Archaeoscyphia nana, Aulocopium sanjuanensis (Fig. 3K), Calycocoelia perforata (Fig. 3D), Hudsonospongia talacastensis (Fig. 3G), H. cyclostoma (Fig. 3F), Patellispongia robusta, Psarodictyum magna, Rhopalocoelia rama, among other species. The new species Talacastonia chela (Fig. 3A-B) was described from the classical Ordovician Talacasto section, Central Precordillera (Fig. 1B). Root tufts of hexactinellids also occur (see Table 1). New megamorinid genera as Rugospongia viejoensis (family Saccospongidae) (Carrera, 1996a) and the tricranocladine sponge Eoscheiella concave (Family Hindiidae Rauff, 1893) have been recovered from the top of the San Juan Formation in the Cerro Viejo, Huaco (Carrera 2007). Nexospongia sillaensis (family Nexospongiidae Carrera, 1996a) was described from the Cerro La Silla, Niquivil Range, Eastern Precordillera. In the upper levels of the San Juan Formation from the Early Llanvirnian, at the Cerro La Chilca section, the calcareous heteractinid Chilcaia bimuralis (Carrera, 1994) and a lithistid species Incrassospongia ramis (Carrera, 1996a) were described (Fig. 3I). Endemic genera such as Protachileum and Talacastonia Beresi and Rigby, 1993, from the Talacasto Gulch and Rugospongia Carrera, 1996 and Chilcaia from different localities of the San Juan Formation occur in the Precordillera. Ordovician sponges from the Precordillera show changes from algal-sponge (reef ecosystems) in the Early Arenig to stromatoporid associations in the Middle Arenig to anthaspidellid demosponge dominated associations in the Upper Arenig to Lower Llanvirn. From the Llanvirn up to the Upper Ordovician, the effects of diverse abiotic factors such as volcanic activity, sea level fluctuations and finally the global climatic cooling, could have contributed to the decrease of the sponges diversity. The diversification of the orchoclad demosponges in the Lower Ordovician carbonate platform of the Precordillera was similar to worldwide radiation pattern (Carrera and Rigby 1999).

assemblages were documented by Gnoli and Serpagli (1980) in the Pachaco section, western Precordillera. Calcarean and demosponge spicules assemblages recovered from residues of conodont samples of several Lower to Middle Ordovician sections of the San Juan Precordillera were described by Mehl and Lehnert (1997). The well preserved silicified spicules include: Polyactinellidae, Heteractinellidae (Calcarea) and hexactinellid and demosponge spicules. The species Dodecaactinella oncera Mehl and Lehnert, 1997, Sardospongia cynodonta Mehl and Lehnert, 1997, Praephobetractinia sp. and Eiffelia sp. were reported from Lower Ordovician (Arenig) strata of the San Juan Formation (Fig. 4A-E). These spicule assemblages were collected from reef-mound horizons and biostromes with sponges, stromatoporoids and receptaculitids of the San Juan Formation (Lower Arenig-Lower Llanvirn) and from the Gualcamayo and Las Aguaditas formations (Lower Llanvirn to Caradocian).

San Rafael Block

Sponge spicules are derived from residues of conodont samples from Middle Ordovician strata, in the geological province of the San Rafael Block, southern Mendoza Province, Argentina (Fig. 1C-D). Spicules (Fig. 4F-Q) have been recovered from the Ponn Trehu Formation, a clasticcarbonate sequence. Poriferan taxa (Beresi and Heredia 2000) include two spicule assemblages: 1) associations of exclusively heteractinellid spicules (sexiradiates) restricted to Arenigian allochthonous blocks of the Oepikodus evae Zone (Heredia 2001) from the shallow platform of the San Juan Formation; and 2) associations of hexactinelliid spicules, calcarean triaene and monaxons, from Upper Llanvirnian autochthonous limestones and carbonate sandstones of the Pigodus serra Zone and the P. anserinus Zone (Heredia 2001) from the outer platform and slope. The spicule associations of the Ponn Trehu Formation represent the most austral Ordovician assemblage described in the context of the Precordillera (Cuyania) Terrane.

Jurassic sponges from the Neuqun Basin

A late Jurasic (Oxfordian) carbonate complex was developed on the foreland side of the Neuqun Basin (Fig. 1B), west- central Argentina and form part of the Cordillera Principal. Shelf carbonates facies are exposed throughout Mendoza and Neuqun provinces. One of these facies consists of small siliceous sponge buildups of the La Manga Formation (Plicatilis Zone), well developed at the Ro Poti Malal section, in southern Mendoza Province. The siliceous sponges with moderate diversity are fossilized in their original shape and exhibit calcareous preservation. Sponge fauna is dominated by hexactinellids (Hexactinosida and Lyssacinosida, 95%) and lithistid demosponges (5%). Up to now, approximately 20% of the material has been preliminary determinated (Beresi 2003b).

Ordovician sponge spicules Precordillera of San Juan Province

The oldest spicule assemblage comes from the Lower Ordovician limestones (Oepikodus intermedius Zone) of the San Juan Formation. The Arenigian silicified spicule

17

Table 1: Biogeographic distribution of sponge taxa from the Cambrian and Ordovician Argentine Basins. Precordillera San Juan Mendoza Paleoenvironment Carbonate platform X X X X X X X X X X X X X X X X X X X X X X X X X X Carbonate platform Carbonate platform Carbonate platform Marine volcaniclastic X X Silicoclastic platform Silicoclastic platform Carbonate platform Carbonate platform Carbonate platform Carbonate platform Carbonate platform X X X X X Famatina System Northern Region X X X

Sponge taxa

DEMOSPONGIAE Family Anthaspidellidae Allosacus sp. Carrera, 1994 Anthaspidella alveola Beresi and Rigby, 1993 Anthaspidella annulata Beresi and Rigby, 1993 Anthaspidella inornata Beresi and Rigby, 1993 Archaeoscyphia minganensis Beresi and Rigby, 1993 Archaeoscyphia nan Beresi and Rigby, 1993 Archaeoscyphia pulchra Bassler, 1941 Aulocopium sanjuanensis Beresi and Rigby, 1993 Calycocoelia perforata Beresi and Rigby, 1993 Hudsonospongia cyclostoma Raymond and Okulitch, 1940 Hudsonospongia talacastensis Beresi and Rigby, 1993 Incrassospongia ramis Carrera, 1996 Patellispongia argentina Carrera, 1994 Patellispongia occulata Bassler, 1941 Patellispongia robusta Beresi and Rigby, 1993 Patellispongia magna sp. Protachilleum kayseri Zittel, 1877 Psarodictium magna Beresi and Rigby, 1993 Rhopalocoelia clarikii Raymond and Okulitch, 1940 Rhopalocoelia rama Beresi and Rigby, 1993 Rhopalocoelia regularis Raymond and Okulitch, 1940 Rhopalocoelia tenuis Carrera, 1994 Talacastonia chela Beresi and Rigby, 1993 Family Hindiidae Eoscheiella concave Carrera, 2007 Family Nexospongiidae Nexospongia sillaensis Carrera, 1996 Family Saccospongiidae Rugospongia viejoensis Carrera, 1996 HEXACTINELLIDA Family Pelicaspongiidae Larispongia magdalenae Carrera, 1998 Family Protospongiidae Protospongia sp. A Protospongia sp. B Diagoniella sp. Protospongia sp. Kiwetinokia utahensis Walcott, 1920 Family uncertain Root tuft HETERACTINIDA Chilcaia bimuralis Carrera, 1994 Dodecaactinella oncera Mehl and Lehnert, 1997 Sardospongia cynodonta Mehl and Lehnert, 1997 Praephobetractinia sp. Eiffelia sp. Octactine spicules Triactine spicules

X X X X

X X X

X X X

18

Fig. 4: Ordovician sponge spicules. A-E. Calcarean and demospongid spicules from the Lower-Middle Ordovician of the San Juan Precordillera (Mehl and Lehnert 1997). CEGHUNC 15951. A. Sardospongia cynodonta. B. Calcarean triactine. C. Demospongid oxea. D-E. Dodecaactinella oncera. F-Q. Spicules from the Ponn Trehu Formation, San Rafael Block (Beresi and Heredia 2000). F. Octactine spicule shows the distal and two of the tangential rays broken. G. Octactine spicule. H. The proximal-distal vertical ray shows a prominent node. I. Monaxon spicule appears to have been sheared diagonally by diagenetic processes. J. Calcarean triactine. K-Q. Scanning electron microscope (SEM) photomicrographs. K, M, N, P. Octactine spicules. L. monaxon spicule shows the central circular canal. O, Q. Hexactine spicules.

The following species have been identified: Laocoetis sp., Laocoetis procumbens and Laocoetis parallela (Hexactinosida Schrammen, 1903, Family Craticulariidae Rauff, 1893) (Fig. 5C-J); Cribrospongia sp., Cribrospongia clathrata and Cribrospongia cucullata (family Cribrospongiidae Roemer, 1864) (Fig. 5A-B) and Polygonatium sp. (Lyssacinosida Zittel, 1877). Sponges belonging to the Family Cribrospongiidae are cup-shaped (Cribrospongia reticulata), tubular and conical. Only a few specimens are triangular in shape and compressed (Cribrospongia cucculata). Fragments of cylindrical to

tubular sponges belong to the genus Laocoetis (=Craticularia Zittel, 1877; emend. Schrammen, 1937).

Cretaceous freshwater sponge from the North Patagonian Massif

The only freshwater sponge was described from Lower Cretaceous lacustrine sediments of the Chubut River Valley, in the Chubut Province, North Patagonian Massif (Fig. 1B). The sponge was determined as Palaeospongilla chubutensis by Ott and Volkheimer (1972). The encrusting sponge belongs to the monogeneric family Palaeospongillidae Volkmer-

19

Fig. 5: Oxfordian hexactinellid sponges at the Ro Potimalal section, Neuqun Basin (Beresi 1997) A. Cribrospongia cuccullata Quenstedt, 1878, lateral view. B. Dermal surface with craticulariid diplorhysis. C. Laocoetis procumbens Goldfuss, 1826, lateral view. D. Upper view of the osculum sponge showing in H. E. Upper view of the narrow osculum and folded wall in a cribrospongiid sponge. F. Lateral view of a cylindrical sponge. G. Laocoetis clathrata Goldfuss, 1833, lateral view. H. Longitudinal section of a tubular cribrospongiid sponge. I. Longitudinal section of a tubular cribrospongiid sponge. J. Upper view of the same sponge, showed in I. K-M. Palaeospongilla chubutensis Ott and Volkheimer, 1972. K. Megasclere with central canal. L. Gemmules. M. Gemmule and spicular texture of the sponge skelton. N-O. Tertiary megascleres, Paran basin.

Ribeiro and Reitner, 1991. It is characterized by acanthoxeas to acanthostrongyles gemmoscleres (Fig. 5K-O).

Remarks
Fossil sponges are known from several geologic basins with different lithologic, sedimentologic and environmental characteristics (Table 2). There are sponges and loose spicules representative of the Classes: Hexactinellida, Demospongida and Calcarea. The oldest sponge fauna known in Argentina is from the upper Lower Cambrian of the Precordillera, the youngest one occurs in the Middle Tertiary of the Paran Basin. Protospongiids characterized the western old Gondwana continent and are known from the Cambrian and Lower

Tertiary spicules of the Paran basin

Isolated oxeas and strongyles, possibly belonging to the species Trochospongilla repens (family Spongillidae) were collected from Tertiary (Miocene) pelitic sediments of the Paran basin, northeast of Argentina (Fig. 1). Borings of Cliona entrerriana and C. ameghinoi on calcareous shells of Ostrea patagonica have also been found in Tertiary sediments.

20

Table 2: Sponge taxa from the Mesozoic and Tertiary Argentine Basins. Taxon sponges and spicules types DEMOSPONGIAE Family Spongilliidae ? Trochospongilla repens Bonetto and Ezcurra de Drago, 1969 Family Palaeospongilliidae Paleoespongilla chubutensis Ott and Volkheimer, 1972 HEXACTINELLIDA Order Hexactinosida Family Craticulariidae Laocoetis paradoxa Goldfuss, 1833 Laocoetis procumbens Goldfuss, 1826 Laocoetis parallela Goldfuss, 1826 Laocoetis clathrata Goldfuss, 1826 Family Cribrospongiidae Cribrospongia sp. Cribrospongia cucullata Quenstedt,1878 Order Lyssacinosida Polygonatium sp.; Schrammen, 1937 Paran Agrio Chaco-Paranense North Patagonian Massif Middle Tertiary Cretaceous: Hauterivian Fluviomarine Lacustrine Formation Basin Age Environment

La Manga

Neuqun

Upper Jurassic Oxfordian

Marine, carbonate bioherms

Ordovician of the Puna, Eastern Cordillera and Sub-Andean Ranges of northern Argentina and the Famatina System, occurring also in Cambrian rocks of the Precordillera. Middle Ordovician sponge faunas are mostly dominated by demosponges, whereas calcareans and hexactinellids do not occur frequently in the warm carbonate platform of the Precordillera. Within the range of the Oxfordian Plicatilis Zone Hexactinellid sponges are common in the carbonate facies of the Neuqun basin, similarly to the Oxfordian European sponge facies. There is still a long way to go concerning the fossil sponges of Argentina, especially in the Cambrian of the Precordillera and northern regions, the Jurassic of Neuqun Basin and freshwater sponges from Mesozoic and Cenozoic lacustrine sediments.

Acknowledgements
The author thanks the Scientific Committee of the 7th International Sponge Symposium for the invitation to participate as Invited Lecturer in the session Palaeontology, and Dr. W. Volkheimer (Conicet-Ianigla, Mendoza) for critic reading of the manuscript. This paper is a contribution to the CONICET-Project: PIP 5222, Argentina.

References
Aceolaza FG, Toselli AJ (1977) Observaciones geolgicas y paleontolgicas sobre el Ordovcico de la zona de Chaschuil, Provincia de Catamarca. Acta Geol Lilloana 14: 55-81

Aceolaza FG, Toselli AJ (1999) Argentine Precordillera: allochthonous or autochthonous Gondwanic? Zentralbl Geol Palontol Teil I 7/8: 743-756 Aceolaza FG, Miller H, Toselli AJ (2002) Proterozoic-Early Paleozoic evolution in western South America - a discussion. Tectonophysics 354: 121-137 Albanesi GL, Ortega G (2002) Advances on conodont-graptolite biostratigraphy of the Ordovician system of Argentina. In: Aceolaza FG (ed). Aspects of the Ordovician System in Argentina, Serie Correlacin Geolgica 16. INSUGEO, Tucumn. pp. 143165 Baldis BA, Bordonaro OL, Armella C, Beresi MS, Cabaleri N, Peralta SH, Bastias H (1989) La cuenca Paleozoica inferior de la Precordillera Argentina. Correl Geol 6: 101-121 Bassler RS (1941) The Nevada early Ordovician (Pogonip) sponge fauna. US Natl Mus Proc 91(3126): 91-102 Bengston S, Missarzhevski VV (1981) Coeloscleritophora - a major group of enigmatic Cambrian metazoans. US Geological Survey, Open-file report 81(743): 19-21 Beresi MS (2003a) Cambrian sponge spicules and chancelloriids of the Argentine precordillera: a review. Geol Acta 1(1): 73-84 Beresi MS (2003b) Oxfordian sponge association from the Neuqun basin, Mendoza, west-central Argentina. J S Am Earth Sci 16:105112 Beresi MS, Aceolaza G, Nievas S (2006) Cambro-Ordovician sponges and spicule assemblages from Northwest Argentina: additional data from the siliciclastic platforms of western Gondwana. Neues Jahrbuch Geol Palontol Monatsh 7: 403-420 Beresi MS, Banchig AL (1997) First hexactinnellid spicules from the Late Cambrian olistolith of the Los Sombreros Formation, Tontal Range, San Juan Precordillera, Argentina. Rev Esp Paleontol 12(2): 141-150

21

Beresi MS, Esteban SB (2003) Sponge fauna and spicules assemblages from the Ordovician of Argentina: a review. In: Aceolaza FG (ed). Aspects of the Ordovician System in Argentina, Serie Correlacin Geolgica 16. INSUGEO, Tucumn. pp. 71-85 Beresi MS, Heredia SE (1995) Asociacin de espculas de porferos cmbricos en la Precordillera de Mendoza. Ameghiniana 32: 401405 Beresi MS, Heredia SE (2000) Sponge spicule assemblages from the Middle Ordovician of Ponn Trehu, southern Mendoza, Argentina. Rev Esp Paleontol 15(1): 37-48 Beresi MS, Rigby JK (1993) The Lower Ordovician sponges of San Juan, Argentina. Geol Stud 39: 1-64 Beresi MS, Rigby JK (1994) Sponges and Chancelloriids from the Cambrian of western Argentina. J Paleontol 68(2): 208-217 Beresi MS, Banchig AL (1997) First hexactinellid spicules of the Late Cambrian Olistolith, of the Los Sombreros Formation, Tontal Range, San Juan Precordillera, Argentina. Rev Esp Paleontol 12(2): 141-150 Bonetto AA, Ezcurra de Drago LD (1969) Systematics notes about the genus Uruguaya Carter (Porifera, Spongillidae). Physis 28(77): 351-358 Bordonaro OL, Martos L (1985) El gnero Chancelloria en el Cmbrico de San Juan. Reunin de Comunicaciones Paleontolgicas, Asociacin Paleontolgica Argentina, San Juan: 6-7 Caas F, Carrera MG (1993) Early Ordovician microbial-spongereceptaculitid bioherms of the Precordillera basin, Western Argentina. Facies 29: 169-178 Carrera MG (1994) An Ordovician sponge fauna from San Juan Formation, Precordillera basin, Western Argentina. Neues Jahrb Geol Palontol Abhand 191(2): 201-220 Carrera MG (1996a) Ordovician Megamorinid sponges from San Juan Formation, Precordillera, Western Argentina. Geobios 29(6): 643-650 Carrera MG (1996b) Nuevos porferos de la Formacin San Juan (Ordovcico), Precordillera Argentina. Ameghiniana 33(3): 335342 Carrera MG (1998) First ordovician sponge from the Puna region, northwestern Argentina. Ameghiniana 35(2): 205-210 Carrera MG (2007) The oldest hindiid demosponge from the Darriwilian (Middle Ordovician) of the Argentine Precordillera: evolutionary implications for the Tricranocladines. J Paleontol 81(4): 754-759 Carrera MG, Rigby JK (1999) Biogeography of the Ordovician sponges. J Paleontol 73: 26-37 Devizia C (1973) Estudio geolgico del sector de San Isidro, Departamento Las Heras Mendoza. PhD Thesis, Museo de La Plata. La Plata Esteban SB, Gutirrez-Marco JC (1997) Graptolitos del Tremadoc del Sistema de Famatina (Argentina). V Reunin Internacional Proyecto 351 PICG Paleozoico Inferior del Noroeste de Gondwana, Libro de Resmenes. pp. 59-63 Esteban SB, Rigby JK (1998) Hexactinellid sponges from the Lower Tremadocian Volcancito Formation, Famatina Range, Northwestern Argentina. Geol Stud 43: 1-7 Finney SC, Gleason J, Gehrels G, Peralta S, Aceolaza G (2003) Early Gondwanan connection for three Argentine Precordillera terrane. Earth Planet Sci Lett 205: 349-359 Finney SC, Gleason J, Gehrels GE, Peralta SH, Vervoort JD (2003b) U/Pb geochronology of detrital zircons from Upper Ordovician

Las Vacas, La Cantera, and Empozada formations, NW Argentina. In: Albanesi GL, Beresi MS, Peralta SH (eds). Ordovician from the Andes. Serie Correlacin Geolgica 17. INSUGEO, Tucumn. pp. 191-196 Gnoli M, Serpagli E (1980) A sponge spicula assemblage from Lower Ordovician of Precordilleran Argentina. Riv Ital Paleontol 86: 267-272 Goldfuss A (1826-1833) Petrefacta Germaniae tam ea, quae in museum Universitatis Regiae Borussicae Fredericae Wilhelminae Rhenanae servantur quam ali quaecunque in Museis Hoeninghausiano Muensteriano aliisque extant. Iconibuset Descriptionibus Illustrata Erster Theil. Arnz & Comp, Dsseldorf Heredia SE (2001) Late Llanvirnian conodonts from the Ponn Trehu Formation, Mendoza, Argentina. GAIA 16: 101-117 Mehl D, Lehnert O (1997) Cambro-Ordovician sponge spicules assemblages in the Ordovician of the Argentine Precordillera and paleoenvironmental ties. Neues Jahrbuch Geol Palaontol Abhand 204: 204-246 Ott E, Volkheimer W (1972) Palaeospongilla chubutensis n.g et n.sp. ein Ssswasserschwamm der Kreide Patagoniens. Neues Jahrbuch Mineral Geol Palontol 140(1): 49-63 Pernas RD (1964) El gnero Protospongia en el Cmbrico de Mendoza. Miscelnea 1, Consejo Investigaciones Cientficas: 3-4. La Plata Quenstedt FA (1876-1878) Die Schwmme In: Petrefactenkunde Deutschlands. Der Ersten Abtheilung, Fnfter Band. Korallen (Schwmmen). Leipzig Rauff H (1893-1894) Palaeospongiologie. Palaeontographica 40: 1-346 Ramos VA, Jordan TA, Allmendinger R, Mpodozis C, Kay SM, Corts J, Palma M (1986) Paleozoic terranes of the Central Argentine-Chilean Andes. Tectonics 5(6): 855-880 Raymond PE, Okulitch VJ (1940) Some Chazyan sponges. Bull Mus Comp Zool 86(5): 197-214 Roemer FA (1864) Die Spongitarien des norddeutschen Kreigebirges. Palaeontographica 8: 1-64 Rusconi C (1955) Fsiles cmbricos y ordovcicos al oeste de San Isidro, Mendoza. Rev Mus Hist Nat Mendoza 8: 3-64 Sarmiento GN (1985) La biozona de Amorphognathus variabilisEoplacognathus pseudoplanus (Conodonta), Llanvirniano inferior, en el flanco oriental de la Sierra de Villicm. 1 Jornadas sobre Geologa de la Precordillera, Actas 1: 119-123 Schrammen A (1903) Zur Systematik der Kieselspongien. Mittelungen aus dem Roemer-Museum, Hildescheim 19: 1-21 Schrammen A (1937) Die Kieselspongien des Oberen Jura von Sddeutschland, B, Besonderer teil. Palaeontographica 85: 1-114 Thomas WA, Astini RA (2003) Ordovician accretion of the Argentine Precordillera terrane to Gondwana: a review: J S Am Earth Sci 16: 67-79 Volkmer-Ribeiro C, Reitner J (1991) Renewed study of the type material of Palaeospongilla chubutensis Ott and Volkheimer (1972). In: Reitner J, Keupp H (eds). Fossil and recent sponges. Springer, Berlin. pp.121-133 Walcott CD (1920) Middle Cambrian Spongiae. Cambrian geology and paleontology. Smithsonian Misc Collect 67(6): 261-364 Zittel KA (1877) Protachileum Kayseri: in Uber Primordiale und untersilurische fossilien aus der Argentine Republik. Beitrge Geologie und Palontologie der Argentine Republik, 2: 22-23

Porifera research: Biodiversity, innovation and sustainaBility - 2007

23

Morphological and cytological descriptions of a new Polymastia species (Hadromerida, Demospongiae) from the North-West Mediterranean Sea
Nicole Boury-Esnault(1*), Chantal Bzac(2)
Aix-Marseille Universit, CNRS UMR-6540 DIMAR, Centre dOcanologie de Marseille. Station Marine dEndoume. Rue Batterie des Lions, 13007. Marseille, France. nicole.boury_esnault@univmed.fr (2) Aix-Marseille Universit, CNRS UMS-2196, Centre dOcanologie de Marseille. Station Marine dEndoume. Rue Batterie des Lions, 13007. Marseille, France
(1)

Abstract: A new species of Polymastia (Hadromerida, Polymastiidae), P. harmelini is described from the coast of Provence (NW Mediterranean). Although this region has been intensively studied, new species are regularly found there. Its description includes morphological, anatomical and cytological features and the species is compared to the Polymastia species from the Atlanto-Mediterranean area. Keywords: Hadromerida, Polymastia, skeleton, anatomy, cytology, taxonomy

Introduction
Even in the new era of bar-coding precise morphological and anatomical descriptions of organisms are still necessary for an unassailable systematics. As Jenner (2006) stressed The study of morphology needs no excuse. It is the uncontested and irreplaceable documentation of lifes diversity. This is particularly essential in animals such as sponges. The framework of the existing classification for Porifera has been recently revisited in a collective book Systema Porifera (Hooper and van Soest 2002). In the last 20 years the taxonomy of the genus Polymastia Bowerbank, 1864 (Demospongiae, Hadromerida, Polymastiidae) has been improved by taking into account the precise organisation of the skeleton in the main body and in the papilla (BouryEsnault 1987, Kelly-Borges and Bergquist 1997, Morrow and Boury-Esnault 2000, Boury-Esnault 2002). It has been shown also the importance of cytological criteria as discriminating characters (Boury-Esnault 1974, Boury-Esnault et al. 1994). In a survey of the sponge fauna from the caves of the NW Mediterranean coast, Jean-Georges Harmelin has discovered at the entrance of the 3PP cave a new species belonging to the genus Polymastia (Fig. 1).

skeleton in HNO3 was done using routine procedures (BouryEsnault and Rtzler 1997), and then mounted on a slide in a drop of epoxy resin. For the skeleton thin sections were made after inclusion in epoxy resin of small pieces of the specimen following Boury-Esnault et al. (2002). Sections of about 1 mm were made with an 11-1180 Isomet low speed saw (Buehler). The sections were then adhered to a slide, ground and polished with a polisher (ESCIL 200 GTL) to a thickness of 15 m. The finishing touches were done by hand with abrasive papers n 600 and n 1200, and 8 and 3 m alumina powder. The thin section was then coloured with toluidine blue under heat for several seconds. A coverslip with a small drop of resin was placed on the thin slides for observation. For cytology in light and transmission electron microscopy (TEM), the specimens were fixed in glutaraldehyde 2.5% in a mixture of 0.4 M cacodylate buffer and seawater (4 vol: 5 vol) (Boury-Esnault et al. 1984). They were postfixed during 1h in 2% osmium tetroxide in seawater, dehydrated through an alcohol series, and embedded in Araldite. Semi-thin sections were stained with toluidine blue. Thin sections, contrasted with uranyl acetate and lead citrate, were observed under a ZEISS EM912 transmission electron microscope.

Materials and methods

The specimens were collected by SCUBA diving during a survey of La Ciotat 3PP cave (4309N, 536E). The 3PP cave (Vacelet et al. 1994) is a long term biodiversity research focal site of the NW Mediterranean (Warwick et al. 2003). The specimens were collected in 1999, 2002 and 2004 fixed in buffered formalin 4% and then transferred to alcohol 70%. To study the shape and size of spicules, dissociation of siliceous

Results Polymastia Bowerbank, 1864

Polymastia Bowerbank, 1864: p. 177; type species Halichondria (Spongia) mamillaris by original designation. Pencillaria Gray, 1867: p. 527; type species Spongia mamillaris by original designation

24

Fig. 1: A. Polymastia harmelini sp. nov. Living specimen photographed in situ. The specimen was covered by sediments, scale bar: 0.8 cm. B. Detail of an exhalant papilla of P. harmelini sp. nov. Note the dark ring below the oscule, scale bar: 0.4 cm. (photos Roland Graille). C. Type specimen in alcohol, scale bar: 0.5 cm.

Rinalda Schmidt, 1870: p. 51; type species Rinalda uberrima by original designation. Diagnosis: Polymastiidae, thickly encrusting, spherical or cushion-shaped, always with papillae. Skeleton composed of radial tracts of principal spicules with free spicules scattered in between. Cortex composed of at least two layers, the superficial layer is a palisade of small tylostyles, the lower layer is made of intermediary spicules, tangential, semi-tangential or perpendicular to the surface. Exotyles echinating the surface may be present. The principal spicules can be tylostyles, subtylostyles, styles, or strongyloxeas, intermediary spicules are most often tylostyles, and ectosomal spicules are always tylostyles.

Museum national dHistoire naturelle de Paris (MNHNDNBE.1562). Type locality: on the threshold of the 3PP cave (4309N/536E) at the basis of the west wall of the entrance at about 18 m deep.

Description
External characters (Fig. 1): The specimens are cushion shaped and cover a surface of about 100 cm2. The thickness is about 3-5 mm. In situ (Fig. 1A) the papillae only are visible. The body is covered by sediments and particles trapped by the hispid surface. The colour of the papillae is brown, as well as the surface. The choanosome has a deep yellow colour in life. The cortex is difficult to tear but it is easily detachable from the choanosome. There are about 28 inhalant papillae and 1 exhalant papilla bearing an oscule per specimen. A dark ring followed by a white one surrounds the oscule (Fig. 1B). The length of the inhalant papilla is 4-10 x 1.5-3 mm and that of the exhalant ones is 8-12 x 4 mm.

Polymastia harmelini sp. nov.

Material examined: 5.08.1999 type specimen (Fig. 1C); 13.09.1999; 6.11.2002; threshold of the 3PP cave in La Ciotat (Provence coast). Type specimen deposited in the

25

Fig. 2: Polymastia harmelini sp. nov. Organisation of the main body skeleton. A. General view, scale bar: 175 m. B. Detail of the cortex, scale bar: 115 m. C. Detail of the cortex, scale bar 115 m. D. detail of the base, scale bar: 115 m.

Skeleton (Fig. 2-3): The ectosomal skeleton is about 320370 m thick and composed of three layers: the upper one is a dense palisade (150-170 m) of tylostyles which lie on a layer of collagen (90-120 m) (Fig. 2A). The basal layer of the cortex is a tangential layer (50-80 m) made of intermediary spicules (Figs 2A-C). Right below the surface is a layer of cells with granular inclusions (25 m) which is responsible for the brown colour of the ectosome (Fig. 2B-C). The basal part in contact with the substratum is constituted by the tangential layer of intermediary spicules (Figs 2A and 2D). The palisade is absent and the sponge is fixed to the substratum by a collagen layer (Fig. 2D). Choanosomal tracts of principal spicules can reach 340 m in thickness at the basis. These tracts are divided into two or three smaller ones (170 m) below the ectosome (Fig. 2A). They cross the ectosome and echinate the surface at distances of approximately 400-500 m (Fig. 2C). Ectosomal and intermediary spicules are scattered between the choanosomal tracts (Fig. 2). The skeleton of the papilla consists of ascending multispicular tracts running through the length of the papillae (Fig. 3). About 25 to 35 tracts are present in a papilla and each tract has a diameter of 50-100 m. The central exhalant canal is about 160 m in diameter. It is surrounded in the exhalant papilla by about 10 inhalant canals 80 to 150 m in diameter. The septa between the canals are strengthened by intermediary spicules (Fig. 3). The ectosomal skeleton of the papilla is about 260-300 m thick and composed of two layers (Fig. 3). Towards the periphery there is a layer of tangentially arranged intermediary spicules (50 m) and followed by a palisade of ectosomal spicules (180-290 m). Towards the surface, the extremities of the ectosomal spicules form a regular hispidation of about 100 m in height. Below the cell surface there is a layer of spherulous cells of about 50 m thick. Spicules (Fig. 4): Ectosomal spicules are tylostyles with a well-marked head 122-239 x 1.7-5.2 m (mean = 193 x 2.8 m) straight or slightly bent (Fig. 4C). Intermediary spicules

Fig. 3: Polymastia harmelini sp. nov. Organisation of the papilla skeleton. A. Exhalant papilla, scale bar: 100 m. B. Detail of the inhalant part of a papilla. The arrow indicates inhalant opening, scale bar: 100 m. Abbreviations: C: cortex; E: exhalant canal; F: transversal section of fascicle of principal spicules. I: inhalant canal.

26

Fig. 4: Polymastia harmelini sp. nov. SEM views of spicules. A. Principal spicules, scale bar: 48 m. B. Intermediary spicules, scale bar: 38 m. C. Ectosomal spicules, scale bar: 15 m. Abbreviations: i: intermediary spicules; e: ectosomal spicules.

Fig. 5: A Polymastia harmelini sp. nov. anatomy of the main body, semithin sections. A. General view, scale bar: 100 m. B. Detail of the limit between ectosome and choanosome, scale bar: 50 m. C. Detail of the upper part of the ectosome, scale bar: 25 m. D. Detail of the choanosome, scale bar: 25 m. Abbreviations: C: cortex; Ch: choanosome; cc: choanocyte chamber; S: location of spicules.

are styles, subtylostyles or tylostyles straight 370-583 x 5.311 m (mean = 456 x 6.5 m) (Fig. 4B). Principal spicules are styles, subtylostyles or tylostyles straight 646-837 x 8-16 m (mean = 745 x 11 m) (Fig. 4A). Anatomy (Fig. 5-6): The cortex, 430-600 m thick, is collagenous with few cells present except close to the surface (Fig. 5). The choanosome has a higher density of cells. On semi-thin sections, choanocyte chambers have a diameter of about 15-25 m which correspond to an estimated volume of 1750-8120 m3. The number of choanocytes is 8-18 on a section of a choanocyte chamber. Using the indirect method of Rasmont and Rozenfeld (1981) the estimated number of choanocytes is 75-120 per chamber. The choanocytes have a periflagellar sleeve between the flagellum and the collar of microvillies. Oocytes are visible in the semithin sections in the specimen collected in August 1999. The oocytes are ovoid or spherical in shape. The size is about 32 x 12 m and the nucleus 8.5 x 5 m. They have a homogeneous content. The papillae have a higher cell density than the cortex especially close to the surface where cells with inclusions constitute a layer of about 30 m (Fig. 6A-B). The exhalant

canal is surrounded by a sphincter of contractile cells, absent around the inhalant canals (Fig. 6A and 6C). Cytology (Figs. 7-9): The most abundant cells are the cells with granular inclusions, which constitute a layer close to the surface (Figs. 2, 5A, 6A, 7) and which confer a brown colour to the cortex. These cells are dispersed in the mesohyl. They have an ovoid to spherical shape (Fig. 7A-B) and the size is about 6.3-11.6 x 2.8-7.9 m. The cytoplasm is reduced to small strands and the nucleus is distorted by the abundance of granular inclusions the diameter of which varies from 0.8 to 4.9 m (Fig. 7A-B). In some cells the inclusions seem to have completely fused and the cell has a granular appearance (Fig. 7B). The distorted nucleus has a diameter from 1.6 to 2.6 m and is often smaller than the inclusions. Cells with a cytoplasmic paracrystalline inclusion are present in the mesohyl (Fig. 8). The cells are ovoid in shape and are 5.6-6.9 m in length and 3.2-5.9 m in thickness. The cytoplasm is reduced to thin strands due to the presence of vacuoles (0.9-1.7 m in diameter) with a heterogenous content and a paracrystalline rod of 2.4-5.6 m in length to 1.6-2.4 m in thickness (Fig. 8A). The crystalline structure

27

has a mesh of 0.03 m in diameter (Fig. 8B). The nucleus is 1.7-1.9 m in diameter. Spiculocytes are present in the mesohyl of the choanosome. They are elongated cells which contain a vacuole with a triangular axial filament around which a spicule is secreted. The nucleus is often nucleolated and numerous mitochondria are present in the cytoplasm. The contractile cells present around the exhalant canals and the oscule have a length which can reach more than 20 m for a thickness of about 3 m at the level of the nucleus (Fig. 9A). The nucleus is ovoid (3 x 1.6 m). All along the length of the cell, contractile filaments of about 0.025 m thick are aligned (Fig. 9B). Archaeocytes are present in the mesohyl (Fig. 9C). They are ovoid cells about 6 x 3.5 m and a nucleus of 3.2 x 2.6 m. The nucleus is nucleolated and the diameter of the nucleolus is about 0.5 m. A very active Golgi apparatus is always present. A variable number of phagosomes (about 1 m in diameter) is observed in the cytoplasm. Rare glycocytes which possess small osmiophilic inclusions and rosettes of glycogen are also present in the mesohyl (Fig. 9D). They measure 5.4-8.6 x 2-4.6 m; the nucleus is about 2 m in diameter and the osmiophilic inclusions about 0.2-0.3 m.

Discussion
Fig. 6: Polymastia harmelini sp. nov. Anatomy of the papillae, semithin section. A. General view, scale bar: 70 m. B. Detail of the external part of the ectosome, scale bar: 25 m. C. Detail of the internal part and of the sphincter of contractile cells around the exhalant canal, scale bar: 25 m. Abbreviations: E: exhalant canal; Ec: ectosome; I: inhalant canal.

In the Atlanto-Mediterranean area three Polymastia species have a cortex made of three layers: an external palisade of tylostyles, an intermediary layer of collagen, and an internal layer of tangential intermediary spicules: P. mamillaris (Mller, 1806), P. arctica (Merejkowsky, 1878) and P. grimaldi (Topsent, 1913) and these species have been often mixed up

Table 1: Comparison of P. harmelini sp. nov. with the three species of the Atlantic area sharing a cortex of three layers as recently redescribed in Boury-Esnault (1987) for P. grimaldi, Morrow and Boury-Esnault (2000) for P. mamillaris and Plotkin and Boury-Esnault (2004) for P. arctica (measures in m). Characters Cortex Thickness Number of layers Palisade layer Collagenous layer Tangential layer Subcortical layer of free spicules Free spicule type Number/specimen Budding Ectosomal Size Intermediary Size Principal Size Exotyles Size P. harmelini sp. nov 350 3 170 100 80 absent ectosomal and intermediary >10 absent tylostyles 190 x 3 tylostyles 456 x 6.5 tylostyles 745 x 11 absent NW Mediterranean 18 m P. mamillaris 400 3 300 20 80 500 ectosomal >10 absent fusiform tylostyles 170 x 12 subtylostyles 445 x 13 fusiform strongyloxea 1052 x 24 absent Swedish west coast 76-225 m P. arctica 560 3 235 130 200 560 ectosomal >100 present fusiform tylostyles 170 x 5 styles 410 x 10 fusiform tylostyles 800 x 14 absent White and Barrents Sea 4-109 m P. grimaldi 650 3 250 150 250 absent ectosomal >100 absent tylostyle 220 x 7 fusiform tylostyles 440 x 14 fusiform strongyloxea 1800 x 23 present 4000 x 10 Boreal Atlantic 70-650 m

Choanosome

Papillae Spicules

Distribution Depth range

28

Fig. 7: Polymastia harmelini sp. nov. TEM micrographs of cells with inclusions. A. Cell with individualized granular inclusions, scale bar: 1.5 m. B. Cell with fused granular inclusions, scale bar: 1.5 m. Abbreviations: n: nucleus; g: granular inclusion.

Fig. 8: Polymastia harmelini sp. nov. TEM micrographs. A. Cell with a paracrystalline inclusion, cell with granular inclusion, scale bar: 1.6 m B. Detail of a paracrystalline inclusion, scale bar: 0.3 m. Abbreviations: c: paracrystalline inclusion; g: granular inclusion; n: nucleus.

Table 2: Comparison of the cytology of P. harmelini sp. n with the three species of Polymastia for which we have cytological data. Cell types Exopinacocytes Cells with intranuclear paracrystalline inclusions Cells with paracrystalline inclusions in the cytoplasm Spherulous cells Vacuolar cells Bacteriocyte Glycocytes Contractile cells around exhalant canals and oscule Periflagellar sleeve P. harmelini sp. nov T-shaped + + + + P. penicillus T-shaped endopinacocyte + + + + + P. robusta T-shaped several vacuoles + + + + P. janeirensis T-shaped collencytes, glycocytes 1 vacuole + and at the limit ectosome/ choanosome +

until the recent redescription of their type specimens (Table 1) (Boury-Esnault 1987, Morrow and Boury-Esnault 2000, Plotkin and Boury-Esnault 2004). Polymastia harmelini sp. nov. shares with these three species a cortex constituted by three layers. Polymastia grimaldi differs from the three other species by the presence of a fringe of exotyles at the

limit between the upper and the lower surface. Polymastia mamillaris (type species of the genus) differs from the other species by the shape and size of ectosomal spicules, and the thinness of the collagenous layer. Polymastia mamillaris and P. arctica share the presence of a layer of groups of ectosomal spicules at the limit of the choanosome (Morrow and Boury-

29

Fig. 9: Polymastia harmelini sp. nov. TEM micrographs. A. Elongated contractile cells in the sphincter around the exhalant canal, scale bar: 3 m. B. Detail of the contractile filaments of a contractile cell, scale bar: 0.3 m. C. Archaeocyte with a large nucleolated nucleus, scale bar: 0.7 m. D. Glycocyte with small osmiophilic inclusions and cell with a paracrystalline inclusion, scale bar: 1 m. Abbreviations: c: paracrystalline inclusion; co: collagen; f: contractile filaments; i: osmiophilic inclusion; n: nucleus.

Esnault 2000, Plotkin and Boury-Esnault 2004) absent in P. harmelini sp. nov. and P. grimaldi. Polymastia arctica is the only species of Polymastia known so far which show buds at the extremity of the papillae. The cytology is known only in three species: P. penicillus (Montagu, 1818) [under the name P. mamillaris], P. robusta (Bowerbank, 1861) [Boury-Esnault 1974, Boury-Esnault 1976] and P. janeirensis (Boury-Esnault, 1973) [BouryEsnault et al. 1994]. The four species show identical cytological characters such as T-shaped exopinacocytes as it is general in Demospongiae, the presence of contractile cells around exhalant canals and oscules and, at the limit of ectosome and choanosome in P. janeirensis, of a periflagellar sleeve around the flagella, a character of Hadromerida. The volume of the choanocyte chamber is in the same range as that known for P. janeirensis (3400-7800 m3) and more generally in Hadromerida (Boury-Esnault 2006). Glycocytes are present in the four species even if they are less abundant in P. harmelini sp. nov. The cells with inclusions are the most characteristic features of the four species. Spherulous cells are present in P. penicillus, cells with paracrystalline inclusions in the cytoplasm and cells with granular inclusions in P. harmelini sp. nov., endopinacocytes with intranuclear paracrystalline inclusion in P. penicillus and collencytes with

intranuclear paracrystalline inclusion in P. janeirensis and vacuolar cells in P. robusta and P. janeirensis.

Biogeography
In the Mediterranean Sea six Polymastia species have been recorded: P. mamillaris, P. robusta, P. inflata Cabioch, 1968, P. polytylota Vacelet, 1969, P. tissieri (Vacelet, 1961) [Uriz and Rosell 1990], and P. sola Pulitzer-Finali, 1983. The specimens under the name P. mamillaris are probably P. penicillus (Morrow and Boury-Esnault 2000). Thanks to the precise drawing it is possible to reassign the specimens collected by Uriz (1983) to P. penicillus but such a reassignment is difficult in many other cases (Sar 1958, Carballo and Gmez 1994). The Polymastia species collected in the Mediterranean Sea so far are bathyal or circalittoral species and are also present in the nearby Atlantic (Boury-Esnault et al. 1994). Polymastia harmelini sp. nov. has been collected on the threshold of a cave at 18 m. Sar (1958) has collected a P. mamillaris from littoral cave of the Italian coast. However the description of Sar is not sufficiently precise to understand to which species the specimens collected belong. Carballo and Garcia-Gmez (1994) have also collected specimens of Polymastia in a littoral cave of the Gibraltar strait. There is

30

no description in the paper and it is impossible to understand to which species the specimens belong. Polymastia sola is insufficiently described and the type specimen is not available. In conclusion with this new species six species have been found in NW Mediterranean: P. penicillus [under the name P. mammillaris], P. robusta, P. inflata, P. tissieri, P. polytylota and this new species P. harmelini sp. nov. which is for the time being the only Mediterranean endemic species.

Acknowledgements
We would like to thank our friend Jo Harmelin who discovered the first specimen of this sponge, Roland Graille for in situ photographs and Christian Marschal for his ability to make the sections of the skeleton. We thank also the service de microscopie lectronique of the IBDM and Jean-Paul Chauvin to have given access to the TEM.

References
Boury-Esnault N (1973) Campagnes de la Calypso au large des ctes atlantiques de lAmrique du Sud (1961-1962). 29. Spongiaires. Rs Sci Camp Calypso 10: 263-295 Boury-Esnault N (1974) Structure et ultrastructure des papilles dponges du genre Polymastia Bowerbank. Arch Zool exp gn 115: 141-165 Boury-Esnault N (1976) Morphognse exprimentale des papilles inhalantes de lponge Polymastia mamillaris (Mller). Arch Zool exp gn 117 : 181-196 Boury-Esnault N (1987) The Polymastia species (Demosponges, Hadromerida) of the Atlantic area. In: Vacelet J, Boury-Esnault N (eds). Taxonomy of Porifera from the N.E. Atlantic and Mediterranean Sea, vol. 13. Springer-Verlag, Berlin Heidelberg. pp. 29-66 Boury-Esnault N (2002) Family Polymastiidae Gray, 1867. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges, vol. 1. Kluwer Academic/Plenum Publishers, New York. pp. 201-219 Boury-Esnault N (2006) Systematics and evolution of Demospongiae. Can J Zool 84: 205-224 Boury-Esnault N, Rtzler K (1997) Thesaurus of sponge morphology. Smithsonian Press, Washington Boury-Esnault N, De Vos L, Donadey C, Vacelet J (1984) Comparative study of the choanosome of Porifera: I. The Homoscleromorpha. J Morphol 180: 3-17 Boury-Esnault N, Hajdu E, Klautau M, Custdio M, Borojevic R (1994) The value of cytological criteria in distinguishing sponges at the species level: the example of the genus Polymastia. Can J Zool 72: 795-804 Boury-Esnault N, Marschal C, Kornprobst J-M, Barnathan G (2002) A new species of Axinyssa Lendenfeld, 1897 (Porifera, Demospongiae, Halichondrida) from the Senegalese coast. Zootaxa 117: 1-8 Bowerbank JS (1861) List of British sponges. In: McAndrew R (ed). List of the British marine invertebrate fauna. Rep Br Ass Advmt Sci 30: 235-236 Bowerbank JS (1864) A monograph of the British Spongiadae. Robert Hardwicke, London Cabioch L (1968) Contribution la connaissance de la faune des Spongiaires de la Manche occidentale. Dmosponges de la rgion de Roscoff. Cah Biol mar 9: 211-246

Carballo JL, Garca-Gmez JC (1994) Esponjas del Estrecho de Gibraltar y reas prximas, con nuevas aportaciones para la fauna iberica. Cah Biol mar 35: 192-211 Gray JE (1867) Notes on the arrangement of sponges, with the description of some new genera. Proc Zool Soc London 1867(2): 492-558 Hooper JNA, van Soest RWM (2002) Systema Porifera: a guide to the classification of sponges, vol. 1. Kluwer Academic/Plenum Publishers, New York Jenner RA (2006) Challenging received wisdoms: Some contributions of the new microscopy to the new animal phylogeny. Integr Comp Biol 46: 93-103 Kelly-Borges M, Bergquist PR (1997) Revision of Southwest Pacific Polymastiidae (Porifera: Demospongiae: Hadromerida) with descriptions of new species of Polymastia Bowerbank, Tylexocladus Topsent, and Acanthopolymastia gen. nov. from New Zealand and the Norfolk Ridge, New Caledonia. N Z J Mar Freshwat Res 31: 367-402 Merejkowsky C de (1878) Les ponges de la mer Blanche. Mm Acad imp Sci, St Petersburg 26: 1-51 Montagu G (1818) An essay on sponges, with descriptions of all the species that have been discovered on the Coast of Great Britain. Mem Werner Nat Hist Soc 2: 67-121 Morrow CC, Boury-Esnault N (2000) Redescription of the type species of the genus Polymastia Bowerbank, 1864 (Porifera, Demospongiae, Hadromerida). Zoosystema 22: 327-335 Mller OF (1806) Zoologia danica. Havniae, Copenhagen Plotkin AS, Boury-Esnault N (2004) Alleged cosmopolitanism in sponges: the example of a common Arctic Polymastia (Porifera, Demospongiae, Hadromerida). Zoosystema 26: 13-20 Pulitzer-Finali G (1983) A collection of Mediterranean Demospongiae (Porifera) with, in appendix, a list of the Demospongiae hitherto recorded from the Mediterranean Sea. Ann Mus Civ Stor Nat Genova 84: 445-621 Rasmont R, Rozenfeld F (1981) Etude micro-cinmatographique de la formation des chambres choanocytaires chez une ponge deau douce. Ann Soc R Zool Belg 111: 33-44 Sar M (1958) Studio sui Poriferi di una grotta di marea del Golfo di Napoli. Arch Zool Ital 43: 203-280 Schmidt O (1870) Grundzge einer Spongien Fauna des Atlantischen Gebietes. Wilhelm Engelmann, Leipzig Topsent E (1913) Spongiaires provenant des campagnes scientifiques de la Princesse Alice dans les mers du Nord (1898-1899 - 19061907). Imprimerie de Monaco, Monaco Uriz MJ (1983) Contribucin a la fauna de esponjas (Demospongia) de Catalunya. An Sec Cienc Col Univ Gerona 7: 1-220 Uriz MJ, Rosell D (1990) Sponges from bathyal depths (10001750m) in the western Mediterranean Sea. J Nat Hist 24: 373-391 Vacelet J (1961) Quelques ponges remarquables de Mditerrane. Rec Trav Inst Pches maritimes 25: 351-354 Vacelet J (1969) Eponges de la Roche du Large et de ltage bathyal de Mditerrane (rcoltes de la Soucoupe plongeante Cousteau et dragages). Mm Mus natl Hist nat 59: 145-219 Vacelet J, Boury-Esnault N, Harmelin J-G (1994) Hexactinellid cave, a unique deep-sea habitat in the scuba zone. Deep-Sea Res 41: 965-973 Warwick RM, Emblow C, Fral J-P, Hummel H, Avesaath P van, Heip C (2003) European marine biodiversity research sites. NIOOCEME, Yerseke, the Netherlands

Porifera research: Biodiversity, innovation and sustainaBility - 2007

31

Two new haplosclerid sponges from Caribbean Panama with symbiotic filamentous cyanobacteria, and an overview of sponge-cyanobacteria associations
Maria Cristina Diaz(1,2*), Robert W. Thacker(3), Klaus Rtzler(1), Carla Piantoni(1)
Invertebrate Zoology, National Museum of Natural History, Smithsonian Institution, Washington, D.C. 20560-0163, USA. ruetzler@si.edu (2) Museo Marino de Margarita, Blvd. El Paseo, Boca del Ro, Margarita, Edo. Nueva Esparta, Venezuela. crisdiaz@ix.netcom.com (3) Department of Biology, University of Alabama at Birmingham, Birmingham, AL 35294-1170, USA. thacker@uab.edu
(1)

Abstract: Two new species of the order Haplosclerida from open reef and mangrove habitats in the Bocas del Toro region (Panama) have an encrusting growth form (a few mm thick), grow copiously on shallow reef environments, and are of dark purple color from dense populations of the cyanobacterial symbiont Oscillatoria spongeliae. Haliclona (Soestella) walentinae sp. nov. (Chalinidae) is dark purple outside and tan inside, and can be distinguished by its small oscules with radial, transparent canals. The interior is tan, while the consistency is soft and elastic. The species thrives on some shallow reefs, profusely overgrowing fire corals (Millepora spp.), soft corals, scleractinians, and coral rubble. Xestospongia bocatorensis sp. nov. (Petrosiidae) is dark purple, inside and outside, and its oscules are on top of small, volcano-shaped mounds and lack radial canals. The sponge is crumbly and brittle. It is found on live coral and coral rubble on reefs, and occasionally on mangrove roots. The two species have three characteristics that make them unique among the families Chalinidae and Petrosiidae: filamentous, multicellular cyanobacterial symbionts rather than unicellular species; high propensity to overgrow other reef organisms and, because of their symbionts, high rate of photosynthetic production. These are the first descriptions of West Atlantic haplosclerid species associated with an Oscillatoria-type symbiont; all previous records of haploscleridcyanobacteria associations were of symbioses with unicellular cyanobacteria. High rates of photosynthetic production of Oscillatoria spongeliae could explain the abundance and overgrowth capability of the two host sponges in the regions reef environments. An overview of associations between sponges and cyanobacteria is presented. Keywords: Haplosclerida, new species, cyanobacteria, Panama

Introduction
The marine subtidal habitats of the Bocas del Toro region (coral reef, mangrove, and sea grasses) are abundantly colonized by marine sponges. A recent survey of sponges from non-cryptic habitats in this region reports 120 described species (Diaz 2005). The Haplosclerida Topsent, 1928 represent the most diverse sponge order at Bocas del Toro, with thirty five species spread across five sponge families: Chalinidae (12 spp.), Petrosiidae (8 spp.), Niphatidae (7 spp.), Callyspongiidae (4 spp.), and Phloeodictydae (4 spp.). Two undescribed species were encountered during this survey, one belonging to Haliclona Grant 1835, sub-genus Soestella de Weerdt, 2000, family Chalinidae Gray, 1867, and the second one to Xestospongia de Laubenfels, 1932, family Petrosiidae van Soest, 1980. Both are thin to thickly encrusting species copiously packed with filamentous cyanobacteria identified as Oscillatoria spongeliae (Thacker et al. 2007). The presence of filamentous cyanobacteria as symbionts in these sponges

constitutes a unique occurrence, both phylogenetically and geographically. To date, 100 sponge species in 29 families are known to harbor cyanobacteria (Table 1). The order Haplosclerida contains the highest number of species with this type of association (25, in 11 genera). Of these, 24 species support unicellular cyanobacteria, while only one undescribed Caribbean species in the family Niphatiidae van Soest, 1980 is reported to have filamentous symbionts (Diaz 1996). This unique association seems to have two striking ecological consequences: a competitive advantage over other reef organisms through overgrowth, including even aggressive reef species such as Millepora (Hydrozoa, Cnidaria) and Neofibularia Hechtel, 1965 (Demospongeae, Porifera), and high photosynthetic rates which characterize these two species as phototrophic sponges (Thacker et al. 2007). The present paper describes the morphology and ecological features of both new species and discusses their systematic affinities with close relatives in the Caribbean.

32

Table 1: Sponge species with cyanobacterial symbionts, modified from Diaz (1996). Families assigned to orders: 1. Homoclerophorida; 2. Astrophorida; 3. Halichondrida (sensu van Soest et al., 1989); 4. Poecilosclerida; 5. Lithistida; 6. Hadromerida; 7. Haplosclerida (sensu de Weerdt, 1985); 8. Dictyoceratida; 9. Dendroceratida; 10. Verongida; 11. Clathrinida; 12. Leucettida; 13. Sycettida; 14. Spirophorida. Symbionts (SYM) include the unicellular Aphanocapsa feldmanni-like (Af), A. raspaigella-like (Ar), Prochloron spp. (Pro), Synechococcus spongiarum (S.spo), Synechocystis trididemni-like (St), Synechocystis spp.-like (Sy), the filamentous Oscillatoria spp.-like (O.sp), Oscillatoria spongeliae-like (O.spo), ? = uncertain status and * = only cyanobacterial pigments detected with thin layer chromatography. Some species have more than one described symbiont; others may contain synonymous Aphanocapsa and Synechococcus symbionts. The regions surveyed: Australia (AUS), Bahamas (BAH), Belize (BEL), Great Barrier Reef (GBR), Guam (GU), Japan (JP) Mediterranean (MED), North and South Baja California (NBC, SBC), Palau (PAL), Papua New Guinea (PNG), Puerto Rico (PR), Red Sea (RS), Sulawesi (SUL) and Zanzibar (ZZ). Family Plakinidae1 Plakinidae1 Ancorinidae2 Ancorinidae2 Ancorinidae2 Ancorinidae2 Ancorinidae2 Geodiidae2 Geodiidae2 Geodiidae2 Axinellidae3 Axinellidae3 Halichondriidae3 Halichondriidae3 Halichondriidae3 Dictyonellidae3 Dictyonellidae3 Desmacellidae4 Desmacellidae4 Chondropsidae4 Crambeidae4 Hymedesmiidae4 Isodictyidae4 Microcionidae4 Mycalidae4 Rhabderemidae4 Rhabderemidae4 Theonellidae5 Theonellidae5 Theonellidae5 Theonellidae5 Theonellidae5 Theonellidae5 Theonellidae5 Siphonidiidae5 Alectonidae6 Chondrosiidae6 Chondrosiidae6 Clionaidae6 Clionaidae6 Clionaidae6 Tethyidae6 Spirastrellidae6 Latrunculiidae6 Callyspongiidae7 Callyspongiidae7 Taxa Oscarella sp. Placinolopha mirabilis Penares aff. schulzei Jaspis stellifera Stelletta clavosa Stelletta kallitetilla Stelletta pudica Geodia papyracea Geodia neptuni Geodia sp. 1 Cymbastela sp. Pseudaxinella tubulosa Axinyssa aplysinoides Halichondria sp. Pseudaxinyssa sp. Dictyonella funicularis Svenzea zeai Neofibularia irata Neofibularia notilangere Batzella melanos Crambe sp. Phorbas sp. Coelocarteria singaporense Clathria sp. Mycale hentscheli Rhabderemia sorokinae Rhabderemia sp. Discodermia dissoluta Theonella conica Theonella swinhoei Theonella swinhoei Theonella sp. 1 Theonella sp. 2 Theonella sp. 3 Leiodermatium sp. Neamphius huxleyi Chondrilla australiensis Chondrilla nucula Spheciospongia florida Spheciospongia sp. Cliona sp. Tethya sp. Spirastrella sp. Latrunculia sp. Callyspongia sp. Siphonochalina sp. Sym Ar O.spo Af Af Af S.spo S.spo Af Af Af ? S.spo ? Ar Sy Ar S.spo Af Af St Af Af, Ar ? Af Sy * Af Af Af, O.sp, S.spo Pro Af S.spo Af Af O.sp * Af S.spo Af, S.spo S.spo Af Af O.sp St Af Af Af Region MED SUL SUL GBR PNG BAH BAH BEL BEL BEL PNG BAH PNG MED GBR BEL BAH GBR BEL GBR MED MED PNG MED NZ PNG SUL BEL SUL ZZ JP RS SUL SUL SUL PNG PNG, SUL AUS MED ZZ BEL MED MED GBR SUL GBR RS Source Wilkinson 1980 Daz 1996 Daz 1996 Wilkinson 1979 Daz 1996 Steindler et al. 2005 Steindler et al. 2005 Rtzler 1990 Rtzler 1990 Rtzler 1990 Daz 1996 Steindler et al. 2005 Daz 1996 Wilkinson 1980 Larkum et al. 1988 Rtzler 1981 Steindler et al. 2005 Wilkinson 1980 Rtzler 1990 Larkum et al. 1988 Wilkinson 1980 Wilkinson 1980 Daz 1996 Wilkinson 1980 Webb and Maas 2002 Daz 1996 Daz 1996 Daz 1996 Daz 1996 Steindler et al. 2005 Hentschel et al. 2002 Wilkinson 1978 Steindler et al. 2005 Daz 1996 Daz 1996 Daz 1996 Daz 1996 Daz 1996 Usher et al. 2004a, 2004b Sar 1966 Steindler et al. 2005 Rtzler 1990 Sar 1966 Sar 1966 Cox et al. 1985 Daz 1996 Wilkinson 1980 Wilkinson 1978

33 Table 1 (cont.)

Chalinidae7 Chalinidae7 Niphatidae7 Niphatidae7 Niphatidae7 Niphatidae7 Niphatidae7 Petrosiidae7 Petrosiidae7 Petrosiidae7 Petrosiidae7 Petrosiidae7 Petrosiidae7 Petrosiidae7 Petrosiidae7 Petrosiidae7 Petrosiidae7 Petrosiidae7 Phloeodictyidae7 Phloeodictyidae7 Phloeodictyidae7 Phloeodictyidae7 Dysideidae8 Dysideidae8 Dysideidae8 Dysideidae8 Dysideidae8 Dysideidae8 Dysideidae8 Irciniidae8 Irciniidae8 Irciniidae8 Irciniidae8 Irciniidae8 Spongiidae8 Spongiidae8 Spongiidae8 Spongiidae8 Spongiidae8 Thorectidae8 Thorectidae8 Thorectidae8 Thorectidae8 Thorectidae8 Thorectidae8 Thorectidae8 Aplysillidae9 Darwinellidae9 Aplysinellidae10 Aplysinellidae10 Aplysinidae10 Aplysinidae10

Haliclona sp. Haliclona (Reniera) sp. Amphimedon sp. 1 Amphimedon sp. 2 Cribrochalina dura Cribrochalina vasculum Niphates sp. Neopetrosia exigua Neopetrosia subtriangularis Petrosia ficiformis Petrosia pellasarca Petrosia sp. Xestospongia muta Xestospongia proxima Xestospongia rosariensis Xestospongia sp. Xestospongia testudinaria Xestospongia wiedenmayeri Calyx podatypa Oceanapia sp. Oceanapia ambionensis Pellina semitubulosa Dysidea granulosa Dysidea sp. Dysidea sp. 1 Dysidea sp. 2 Dysidea sp. 3 Lamellodysidea chlorea Lamellodysidea herbacea Ircinia campana Ircinia felix Ircinia ramosa Ircinia variabilis Psammocinia sp. Coscinoderma sp. Phyllospongia alcicornis Phyllospongia foliacens Phyllospongia papyracea Spongia sp. Carteriospongia foliascens Carteriospongia sp. Carteriospongia sp. Dactylospongia elegans Hyrtios violaceus Lendenfeldia frondosa Lendenfeldia dendyi Aplysilla sp. Darwinella sp. 1 Suberea azteca Suberea mollis Aplysina aerophoba Aplysina archeri

* Ar Ar Af Af Af O.sp Af, S.spo Af, S.spo Af Af S.spo Af, S.spo S.spo Af Af Af Af Af * Ar Af O.spo O.spo O.spo O.spo O.sp O.spo O.spo Af, S.spo Af, S.spo * Af, Ar, S.spo * Af Af Af Af ? S.spo Af Af * O.spo Ar Pro, O.spo Ar Af Af O.sp Af, S.spo Af, S.spo

RS MED SUL SUL BEL BEL BAH, BEL PNG, SUL, PAL BEL MED PR ZZ BEL BAH PR SUL PNG, SUL BEL BEL PNG, SUL SUL MED GU GBR PNG, SUL PNG, SUL PNG, SUL PNG, SUL GBR BEL BEL GBR MED PNG GBR GBR GBR GBR MED ZZ SUL, PNG GBR PNG, SUL BEL SUL, PNG ZZ MED SUL SBC RS MED BEL

Wilkinson 1978 Wilkinson 1978 Daz 1996 Daz 1996 Rtzler 1990 Rtzler 1990 Daz 1996 Daz 1996 Thacker 2005 Rtzler 1990 Sar 1966 Vicente 1990 Steindler et al. 2005 Rtzler 1990 Steindler et al. 2005 Steindler et al. 2005 Vicente 1990 Daz 1996 Daz 1996 Rtzler 1990 Rtzler 1990 Daz 1996 Daz 1996 Sar 1966 Thacker and Starnes 2003 Larkum et al. 1987 Daz 1996 Daz 1996 Daz 1996 Daz 1996 Larkum et al. 1987 Rtzler 1990 Steindler et al. 2005 Rtzler 1990 Steindler et al. 2005 Wilkinson 1983 Sar 1971 Steindler et al. 2005 Daz 1996 Wilkinson 1980 Wilkinson 1992 Wilkinson 1978 Wilkinson 1992 Wilkinson 1980 Steindler et al. 2005 Daz 1996 Wilkinson 1992 Daz 1996 Rtzler 1990 Daz 1996 Steindler et al. 2005 Wilkinson 1980 Daz 1996 Daz 1996 Wilkinson 1978 Sar 1966 Hentschel et al. 2002 Rtzler 1990 Steindler et al. 2005

34 Table 1 (cont.)

Aplysinidae10 Aplysinidae10 Aplysinidae10 Aplysinidae10 Aplysinidae10 Aplysinidae10 Aplysinidae10 Aplysinidae10 Aplysinidae10 Clathrinidae11 Leucettidae12 Leucettidae12 Sycettidae13 Tetillidae14 Tetillidae14

Aplysina cauliformis Aplysina fistularis Aplysina fulva Aplysina gerardogreeni Aplysina lacunosa Aplysina sp. Verongula rigida Verongula gigantea Verongula reiswigi Clathrina sp. Pericharax heteroraphis Leucetta sp. Sycon sp. Cinachyrella australiensis Tetilla arb

Af, S.spo Af, S.spo Af, S.spo Af, S.spo Af, S.spo Af Af Af Af Ar Af ? Ar * Af

BEL BEL BEL SBC BEL BEL BEL BEL BEL MED GBR PNG MED PNG BCN

Rtzler 1990 Steindler et al. 2005 Rtzler 1990 Steindler et al. 2005 Rtzler 1990 Steindler et al. 2005 Daz 1996 Steindler et al. 2005 Rtzler 1990 Steindler et al. 2005 Daz 1996 Rtzler 1990 Rtzler 1990 Rtzler 1990 Wilkinson 1980 Wilkinson 1979 Daz 1996 Feldmann 1933 Daz 1996 Daz 1996

Materials and methods

Specimens were collected during field work in 2003 and 2005, using snorkel and SCUBA equipment while exploring two reefs (Swan Cay and Crawl Cay Canal) between 0-15 m deep in the Bocas del Toro region. Sponges were fixed in 10% formalin in seawater and preserved in 70% ethanol. Skeletal and histological preparations for light microscopy and scanning electron microsopy (SEM) followed standard methodology (Rtzler 1978). The skeletal arrangement was described, and the length and width of each spicule type were measured in each specimen. Type material is deposited in the Porifera collection of the Smithsonian Institutions National Museum of Natural History, Washington, DC (USNH), and in the Snithsonian Tropical Research Institute (STRI) laboratory at Bocas del Toro, Panama (BT).

Paratypes: USNM 1106221, Crawl Cay Canal (9o15050N, 82o07631W), 5 m, on top and along sides of Acropora cervicornis, collectors M.C. Daz and R. Thacker, 21-06-05. BT-045, Swan Cay (9o27198N, 82o18024W), 5 m deep, between fire coral (Millepora sp), and lettuce coral (Agaricia sp.) on a shallow reef with strong surge and currents, collector M.C. Daz, 08-2003.

Description
Shape and size: Thin encrusting sheets (1-2 mm thick) covering patches ranging from five to a few hundred cm2 (Fig. 1A, B). Surface: Smooth to irregularly rugose to the naked eye, porous under a microscope. Small oscules (1-2 mm in diameter) with transparent membranes, regularly distributed over the sponge surface. Radial canals converging toward oscules. Spicule tracts piercing through the skin (ectosome) create a microhispid appearance, only visible under a microscope (Fig. 1B). Colour: In live, deep dark- brown to purple outside, tan inside. Cream to white in alcohol. External color due to the photosynthetic cyanobacteria.
Fig. 1: In situ morphology and skeleton arrangement of the new species: A. Haliclona walentinae sp. nov. habit (scale: 6 cm); B. detail showing bumpy surface, radial canals, and oscules (1-2 mm) with white oscular membranes (scale: 5 mm); C. cross section through the choanosome with Soestella-type arrangement of subanisotropic choanosomal skeleton of ill defined paucispicular primary lines connected by paucispicular secondary ones (scale: 100 m); D. Xestospongia bocatorensis sp. nov. habit (scale: 2 cm); F. isotropic unispicular to paucispicular reticulation (2-3 spicules across) forming polygonal-shaped meshes (scale: 120 m).

Results Systematic descriptions

Class Demospongiae Sollas, 1885. Order Haplosclerida Topsent, 1928 Family Chalinidae Gray, 1867 Genus Haliclona Grant, 1835 Sub-Genus Soestella de Weerdt, 2000 Haliclona (Soestella) walentinae sp. nov. (Figs. 1-3; Table 2) Material examined. Holotype: USNM 1106220, Crawl Cay Canal (9o15050N, 82o07631W), 5-10 m deep, covering top and sides of Acropora cervicornis on a shallow reef where Millepora, and Porites were the dominant coral species, collectors M.C. Daz and R. Thacker, 21-06-05.

35

36

Table 2: Spicule measurements of specimens of Haliclona walentinae sp. nov. [max.-min. length (meanSD) x max.-min. width (meanSD)] in m. Material studied USNM 1106220 USNM 1106221 BT-045 Oxea 130161 (1409.3) x 69 (7.60.9) 130160 (1409.2) x 39 (4.81.6) 100180 (13219) x 38 (51)

Haliclona walentinae a very interesting subject for both ecological and evolutionary studies. Etymology: The species is named after Dr. Walentina de Weerdt (University of Amsterdam) whose work with the Haplosclerida has been essential in our understanding of the group. Family Petrosiidae van Soest, 1980 Genus Xestospongia de Laubenfels, 1932 Xestospongia bocatorensis sp. nov. (Figs. 1-3; Table 3) Material examined. Holotype: USNM 1106222, Crawl Cay Canal (9o15050N, 82o07631W), 12 m, top of Acropora cervicornis on a shallow reef where Millepora and Porites were the dominant coral species, collectors M.C. Daz and R. Thacker, 21-06-05. Paratypes: BT-019, Crawl Cay Canal (9o15050N, 82o07631W), 6 m, between fire coral, and Agaricia spp. colonies, on a shallow reef, collector: M.C. Daz, 08-2003; BT-163, same data as holotype.

Consistency: soft, compressible, and resilient, easy to peel off the substrate. Ectosomal skeleton: Poorly developed, some paucispicular spicule tracts and loosely strewn spicules (Fig. 1C). Ectosome not peelable. The ectosome on the underside of the sponge accumulates sand. Choanosomal skeleton: Paucispicular, loosely organized primary skeleton tracts (20-40 m in diameter), and mostly unispicular tracts or single spicules connecting them. Spicule tracts densely enveloped by filamentous cyanobacteria (Fig. 3A, B). Spongin scarce, barely discernable. Spicules: Hastate to fusiform oxea, straight or slightly curved (100-180 x 3-9 m). (Table 2, Fig. 2A). Ecology: The species was found thriving on a shallow reef, profusely overgrowing fire corals (Millepora spp.), soft corals, scleractinians, and other sponges, such a Neofibularia nolitangere (Duchassaing and Michelotti, 1864). It appeared to be a rather aggressive species, dominating all neighboring sessile invertebrates. Remarks: This species is here assigned to the subgenus Soestella, following the definition by de Weerdt (2000) where ill defined paucispicular primary lines, irregularlly connected by unispicular secondary lines characterize the skeletal architecture. Eight additional species in this subgenus occur in the Caribbean: H. (Soestella) caerulea (Hechtel, 1965), H. (S.) lehnerti de Weerdt (2000), H. (S.) luciencis de Weerdt (2000), H. (S.) melana Muricy and Ribeiro (1999), H. (S.) piscaderaensis (van Soest, 1980), H. (S.) smithsae de Weerdt (2000), H. (S.) twincayensis de Weerdt et al. (1991), and H. (S.) vermeuleni de Weerdt (2000). Four of these, H. (S). caerulea, H. (S.) piscaderaensis, H. (S.) twincayensis, and H. (S.) vermeuleni are among common species in the region of Bocas del Toro (Diaz 2005). None of these, nor any other species of Chalinidae, are known to be associated with cyanobacteria (Table 1). Two species in Soestella (melana, and luciencis) are black to dark brown color, but only darkly pigmented cells are reported, at least for the former (de Weerdt 2000). Distinct morphological and ecological differences separate H. (S.) walentinae from the other Haliclona (Soestella) Caribbean species. Among them a thinly encrusting growth habit, soft but resilient consistency, characteristic oscule morphology, and possession of cyanobacterial symbionts. The filamentous cyanobacteria turn out to be a branch of Oscillatoria spongeliae, with genetic affinities to certain Pacific sponge symbionts (Thacker et al. 2007), making

Description
Shape and size: Thinly encrusting species (2-5 mm thick), in patches from five to a few hundred cm2 . Surface: Smooth. Oscules (1-2 mm diameter) on top of low volcano-shaped mounds (1-2 mm of height). Consistency: Crumbly and brittle. Colour: In live, dark purple, inside and out (Fig. 1D). Cream to white in alcohol. Ectosomal skeleton: No organization, spicules strewn tangentially (Fig. 1E). Choanosomal skeleton: Isotropic unispicular to paucispicular reticulation forming polygonal meshes (100-320 m in diameter), and paucispicular primaries (2-3 spicules across) 200-300 m apart. Filamentous cyanobacteria densely packed around the skeleton (Fig. 3C, D). Spicules: Fusiform to slightly hastate stout oxeas in one size class (230-320 x 8-15 m) (Table 3), with pointed ends. Sigmas, c-shaped (10-26 x <1 m) (Fig. 2B). Ecology: The species was found growing in small patches on mangrove roots, empty shells, or coral rubble, and occasionally profusely overgrowing live corals. Also found on shallow reefs (0-15 m deep) growing over coral and bare rock substrates. Remarks: The predominance of unispicular over paucispicular reticulation, and the relatively light spicule density, compared to other Xestospongia, makes this species slightly atypical for the genus. However, two other Caribbean congeners, X. arenosa van Soest and de Weerdt, 2001, and X. wiedenmayeri van Soest, 1980 are precedents for similar skeleton structure. The assignment to Xestospongia is based on spicule size, skeleton structure (more petrosiid than chalinid), and petrosiid consistency (brittle and crumbly); it was first suggested by Dr. Walentina de Weerdt who examined our

37

Fig. 2: SEM photomicrographs of spicules: A. Haliclona walentinae sp. nov. (USNM 1106220) oxeas; B. Xestospongia bocatorensis sp. nov. (USNM 1106222), oxeas and sigmas.

Table 3: Spicule measurements of specimens of Xestospongia bocatorensis sp. nov. [max-min. length (meanSD-) x max.- min. width (mean SD)] in m. Material studied USNM 1106222 BT-019 BT-163 Oxea 280320 (30211.5) x 1215 (130.9) 230260 (24811.2) x 812 (11.80.7) 270305 (29312.4) x 812 (10.61.2) Sigma (length in m) 2025 (221.6) 1012 (11.80.7) 1026 (191.22)

material. Seven other Xestospongia species are recognized in the Caribbean: X. arenosa van Soest and de Weerdt (2001), X. caminata Pulitzer-Finali (1986), X. deweerdtae Lehnert and van Soest (1999), X. muta (Schmidt, 1870), X. portoricensis van Soest (1980), X. proxima (Duchassaing and Michelotti, 1864), X. rosariensis Zea and Rtzler (1983), none of these has the thinly encrusting morphology of X. bocatorensis sp. nov. Three are very common inhabitants of Bocas del Toro reefs: X. proxima, X. muta, X. rosariensis. Even though all of these species harbor symbiotic cyanobacteria, Xestospongia bocatorensis sp. nov. is unique for its possession of a hostspecific clade of filamentous Oscillatoria spongeliae, rather than the more typical unicellular symbionts, Candidatus Synechococcus spongiarum (Usher et al. 2004a, 2004b, Thacker et al. 2007). Etymology: The species is named after the Bocas del Toro region, an extensive system of islands with well developed

mangrove communities and patchy reefs in northeastern Panama where the new species was found.

Discussion and conclusions

To evaluate the relative frequency of associations between cyanobacterial symbionts and marine sponges, we compiled data from morphological and phylogenetic studies of sponges and their symbionts (Table 1, Diaz 1996, Steindler et al. 2005). Prior to genetic studies, many unicellular cyanobacterial symbionts were classified as Aphanocapsa feldmanni Fremy, 1933; some of these have subsequently been recognized as members of the genus Synechococcus (Usher et al. 2006), including a proposed species of sponge-specific unicellular cyanobacteria, Candidatus Synechococcus spongiarum Usher, 2004. Here, we present symbiont names as given by the authors of each study, and recognize that some of

38

Fig. 3: Filamentous cyanobacteria (Oscillatoria spongeliae) and choanocyte chambers shown in sections of the new species: A, B. Haliclona walentinae sp. nov.; C, D. Xestospongia bocatorensis sp. nov.

these names may be synonyms (Table 1). Clearly, combined morphological and genetic studies are needed to resolve some of these issues. Symbiosis of sponges and filamentous (Oscillatoriatype) cyanobacteria is a common occurrence in the IndoPacific region where at least 10 common species are known for this association. The families concerned are Plakinidae (Homosclerophorida), Theonellidae (Lithistida), Dysideidae and Spongiidae (Dictyoceratida), and Aplysinidae (Verongida). In the much better studied Mediterranean Sea, only one Tethya (Tethyidae, Hadromerida) is known with this kind of symbiont. Until our discovery of Haliclona walentinae sp. nov. and Xestospongia bocatorensis sp. nov., only two records of sponges with Oscillatoria-type symbionts existed in the tropical western Atlantic. One is the common bleeder sponge Hyrtios violaceus (Duchassaing and Michelotti, 1864) (Thorectidae, Dictyoceratida), of which the symbiont fine-structure was studied (Rtzler 1990). The other is an undescribed species of Niphates (Niphatidae, Haplosclerida),

which was recorded from the Bahamas and Belize (Diaz 1996). The phototrophic properties of the new species, the nature of the cyanobacterial symbionts, and the phylogenetic affinities of the symbionts to those hosted by Pacific sponges (Thacker et al. 2007) lends these biological assemblages a unique ecological and evolutionary significance. An unsolved issue remains about the origin of the two new species: are they systematic and ecological oddities among Caribbean sponges, or are they invasive species that originated in the tropical Pacific?

Acknowledgments
We thank Dr. Walentina (Wallie) de Weerdt (Amsterdam) who kindly examined fragments of the specimens and commented on their identification. This is Caribbean Coral Reef Ecosystems (CCRE) contribution number 798, supported in part by the Hunterdon Oceanographic Research Fund.

39

References
Cox GC, Hiller RG, Larkum AWD (1985) An unusual cyanophyte containing phycourobilin and symbiotic with ascidians and sponges. Mar Biol 89: 149-163 de Weerdt WH (2000) A monograph of the shallow-water Chalinidae (Porifera, Haplosclerida) of the Caribbean. Beaufortia 50: 1-67 de Weerdt WH, Ruetzler K, Smith KP (1991) The Chalinidae (Porifera) of Twin Cays, Belize, and adjacent waters. Proc Biol Soc Wash 104(1): 189-205 Diaz MC (1996) Molecular and ecological studies of spongemicrobial associations. PhD Thesis. University of California, Santa Cruz Diaz MC (2005) Common sponges from shallow marine habitats from Bocas del Toro region, Panama. Carib J Sci 41(3): 466-475 Duchassaing de Fonbressin P, Michelotti G (1864) Spongiaires de la mer Caraibe. Natuurk Verh Holl Mij Wetensch Haarlem (2) 21(3): 1-124 Feldmann J (1933) Sur quelques cyanophyces vivant dans le tissue des ponges. Arch Zool Exp Gn 75: 331-404 Hechtel GJ (1965) A systematic study of the Demospongiae of Port Royal, Jamaica. Bull Peobody Mus Nat Hist 20: 1-94 Hentschel U, Hopke J, Horn M, Friedrich AB, Wagner M, Hacker J, Moore BS (2002) Molecular evidence for a uniform microbial community in sponges from different oceans. Appl Environ Microbiol 68: 4431-4440 Larkum AWD, Cox GC, Hiller RG, Dibbayawan TP (1988) Prokaryotic algal symbionts of coral reef sponges. Proc 6th Int Coral Reef Symp, Townsville 3: 163-169 Larkum AWD, Cox GC, Hiller RG, Parry DL, Dibbayawan TP (1987) Filamentous cyanophytes containing phycouribilin and in symbiosis with sponges and an ascidian of coral reefs. Mar Biol 95: 1-13 Lehnert H, van Soest RWM (1999) More North Jamaican deep forereef sponges. Beaufortia 49(12): 141-169 Muricy G, Ribeiro SM (1999) Shallow water Haplosclerida (Porifera, Demospongiae) from Rio de Janeiro State, Brazil (southwestern Atlantic). Beaufortia 49(9): 83-108 Pullitzer-Finalli G (1986) A collection of West Indian Demospongiae (Porifera). In appendix, a list of Demospongiae hitherto recorded from the West Indies. Ann Mus Civ Storia Nat Genova 86: 65-216 Rtzler K (1978) Sponges in coral reefs. In: Stoddart DR, Johannes RF (eds). Coral reefs: research methods. Monographs on Oceanographic Methodologies (UNESCO) 5: 299-314 Rtzler K (1981) An unusual bluegreen alga symbiotic with two new species of Ulosa (Porifera:Hymeniacidonidae) from Carrie Bow Cay, Belize. Mar Ecol 2: 35-50 Rtzler K (1990) Associations between Caribbean sponges and photosynthetic organisms. In: Rtzler K (ed). New perspectives in sponge biology. Smithsonian Institution Press, Washington DC. pp. 455-466 Sar M (1966) Associazioni fra Porifera e alghe in acque superficiali del litorale marino. Ric Sci 36: 277-282 Sar M (1971) Ultrastructural aspects of the symbiosis between two species of the genus Aphanocapsa (Cyanophyceae) and Ircinia variabilis (Demospongiae). Mar Biol 11: 214-221 Schmidt O (1870) Grundzge einer Spongien-Fauna des Atlantischen Gebietes. Wilhelm Engelmann, Leipzig

Steindler L, Huchon D, Avni A, Ilan M (2005) 16S rRNA phylogeny of sponge-associated cyanobacteria. Appl Environ Microbiol 71(7): 4127-4131 Thacker RW (2005) Impacts of shading on sponge-cyanobacteria symbioses: a comparison between host-specific and generalist associations. Int Comp Biol 45: 69-376 Thacker RW, Diaz MC, Rtzler K, Erwin PM, Kimble SJA, Pierce MJ, Dillard SL (2007) Phylogenetic relationships among the filamentous cyanobacterial symbionts of Caribbean sponges and a comparison of photosynthetic production between sponges hosting filamentous and unicellular cyanobacteria. In: Custdio MR, Lbo-Hajdu G, Hajdu E, Muricy G (eds). Porifera research: biodiversity, innovation and sustainability. Srie Livros 28. Museu Nacional, Rio de Janeiro. pp. 621-626 Thacker RW, Starnes S (2003) Host specificity of the symbiotic cyanobacterium Oscillatoria spongeliae in marine sponges, Dysidea spp. Mar Biol 142: 643-648 Usher KM, Fromont J, Sutton DC, Toze S (2004a) The biogeography and phylogeny of unicellular cyanobacterial symbionts in sponges from Australia and the Mediterranean. Microb Ecol 48: 167-177 Usher KM, Kuo J, Fromont J, Toze S, Sutton DC (2006) Comparative morphology of five species of symbiotic and non-symbiotic coccoid cyanobacteria. Eur J Phycol 41(2): 179-188 Usher KM, Toze S, Fromont J, Kuo J, Sutton DC (2004b) A new species of cyanobacterial symbiont from the marine sponge Chondrilla nucula. Symbiosis 36: 183-192 van Soest RWM (1980) Marine sponges from Curaao and other Caribbean localities. Part II: Haplosclerida. Stud fauna Curacao Caribb Isl 62(191): 1-173 van Soest RWM, de Weerdt WH (2001) New records of Xestospongia species (Haplosclerida: Petrosiidae) from the Curaao reefs, with a description of a new species. Beaufortia 51: 109-117 Vicente V (1990) Response of sponges with autotrophic symbionts during the coral bleaching episode in Puerto Rico. Coral Reefs 8(1): 99-202 Webb VL, Maas EW (2002) Sequence analysis of 16S rRNA gene of cyanobacteria associated with the marine sponge Mycale (Carmia) hentscheli. FEMS Microbiol Lett 207(1): 43-47 Wilkinson CR (1978) Microbial associations in sponges. I. Ecology, physiology, and microbial populations of coral reefs sponges. Mar Biol 49: 161-167 Wilkinson CR (1979) Nutrient translocation from symbiotic cyanobacteria to coral reef sponges. In: Lvi C, Boury-Esnault N (eds). Biologie des spongiaires. Colloques Internationaux du CNRS, vol. 291. ditions du CNRS, Paris. pp. 373-380 Wilkinson CR (1980) Cyanobacteria symbiotic in marine sponges. In: Schwemmler W, Schenck HEA (eds). Endocytobiology, Endosymbiosis and Cell Biology. De Gruyter, Berlin. pp. 9931002 Wilkinson CR (1983) Net primary productivity in coral reef sponges. Science 219: 410-412 Wilkinson CR (1992) Symbiotic interactions between marine sponges and algae. In: Reiser W (ed.) Algae and symbioses. Biopress Ltd, Bristol. pp.111-151 Zea S and K Rtzler (1983) A new species of Xestospongia (Porifera, Demospongiae), from the Colombian Caribbean. Caldasia 13: 817-831

Porifera research: Biodiversity, innovation and sustainaBility - 2007

41

Sponge embryology: the past, the present and the future

Alexander V. Ereskovsky
Aix-Marseille Universit, CNRS UMR-6540 DIMAR, Centre dOcanologie de Marseille. Station Marine dEndoume. Rue Batterie des Lions, 13007. Marseille, France (corresponding address); and Department of Embryology, Biological Faculty, Saint-Petersburg State University, Universitetskaya nab. 7/9, 199034 Saint-Petersburg, Russia. aereskovsky@mail.ru Abstract: Developmental biology of sponges has a 140-year-old history. It made important contributions to spongiology in general: the creation of the subkingdom Enantiozoa, the separation of the Calcinea and the Calcaronea in the Calcarea, and the separation of the Tetractinomorpha and the Ceractinomorpha within the Demospongiae. Nevertheless, embryonic development has been studied in only 93 sponge species. This review must be restricted primarily to embryonic development and metamorphosis of sponges, because full modern information on its normal development is limited to only a few studies. Mechanisms of morphogenetic movements emerged in the course of evolution prior to the separation of the covering cell layer as ectoderm and the internal digestive cell mass as endoderm. Therefore, it is incorrect to apply the term gastrulation to sponge development. It is difficult to use comparative embryological data on sponges for phylogenetic interpretations because their development is highly polymorphic. The same cleavage pattern and blastula type may be characteristic of different larval types. On the other hand, the same larval types may develop from different cleavage patterns and types of morphogenesis. However, embryological data do indicate groups or types of sponge development. The observed variety of developmental patterns indicates that no linear ways of developmental evolution are common for all Porifera. It testifies to an early divergence of sponge macrogroups or, probably, paraphyly and their long parallel evolution. The basal phylogenetic position of the Porifera among the Metazoa and its suggested paraphyly make new investigations on embryology and larvae especially important. Relatively few homologues of developmental genes are known in the Porifera. Comparison of transcription factors that regulate genes expression during sponge morphogenesis will provide an evolutionary perspective to relationships among basal metazoan phyla. Keywords: development, evolution, morphogenesis, phylogeny, sponges

The past
Developmental studies of sponges have a 140-year-old history. Ever since the work of Ernst Haeckel (1866) (Fig. 1), they have inspired, enriched and modified evolutionary thought. Altogether, about 540 articles concerning sponge embryology have been published. They involve approximately 36 species of Calcarea, 140 species of Demospongiae, and as little as 3 Hexactinellida species. The first period of sponge development studies falls on the last third of the 19th century. It was the Golden Age of sponge embryology. About 110 articles on this topic were published (Fig. 2). Uncontested leadership in this research field belonged to German zoologists: Schulze, Maas, Schmidt, Keller (Fig. 3) and others. The basis of sponge comparative embryology was laid at that time. Haeckel (1874) admitted that embryological studies of calcareous sponges (Haeckel 1872) were the starting point for his ideas about the origin of Metazoa, later formalized as the Gastraea theory of ontogeny recapitulating phylogeny, in which the gastrula is viewed as the recapitulation of a gastraean ancestor that evolved via selection on a simple,

planktonic hollow ball of cells to develop the capacity to feed (Haeckel 1874). On the basis of comparative embryological data of some demosponges, Delage (1892, 1899) (Fig. 4) discovered that during a metamorphosis of parenchymella larvae external flagellated cells migrate inward to form the choanoderm of the adult sponge. These observations have allowed Delage to propose a hypothesis of inversion of the germ layers. Being based on this hypothesis, he separated sponges from Metazoa into a special group, Enantiozoa that signified inside out animals. Bidder (1898), following Minchin (1896), subdivided Calcarea into two subclasses, Calcinea and Calcaronea, distinguished deep embryological differences (e.g., coeloblastula in Calcinea, amphiblastula in Calcaronea), and the position of the nucleus in the choanocytes (with nucleus basal in choanocyte independent of flagellum in Calcinea and with nucleus apical in choanocyte linked to the flagellum in Calcaronea). The second period falls on the first half of the 20th century (1900-1960), when interest in sponge development declined (Fig. 2). Almost the only active researches were made in Belgium by Brien (Fig. 5), Meewis, Leveaux, and in France by Tuzet, Duboscq and Lvi (Fig. 6). Claude Lvi

42

Fig. 1: Ernst Haeckel (1834-1919) and Nicolas Miklucho-Maclay (1846-1888) during the expedition to the Red Sea in 1866 (From. I.I. Kanaev, 1966).

in his famous work (1956) was the first to use embryological characters of sponges in systematics. However, this period was marked by the emergence of a major branch in developmental biology of sponges. Wilson

(1907) pioneered the use of sponges as model animals for cell adhesion research. He described species-specific reaggregation of mechanically dissociated sponge cells. His works provided an impulse for studies of behaviour of separate cells and regeneration in sponges. The third period started with the application of electron microscopy and new optical and experimental methods to sponge studies (Fig. 2). Spermatogenesis and oogenesis, fertilization (in Calcaronea) and larval structure (in all poriferan classes) were investigated ultrastructurally. The results of these works were extensively applied to evolutionary and phylogenetic constructions concerning both Porifera and Metazoa in general. At the same time, complete development from egg to juvenile was investigated at the ultrastructural level only in some species, including some Spongillidae (see: Weissenfels 1989), Halisarca dujardini Johnston, 1842 (Demospongiae, Halisarcida) (Ereskovsky and Gonobobleva 2000, Ereskovsky 2002, Gonobobleva and Ereskovsky 2004a, 2004b, Ereskovsky et al. 2005, 2007a, Mukhina et al. 2006), some species of Oscarella (Ereskovsky and Boury-Esnault 2002, Boury-Esnault et al. 2003, Ereskovsky 2005, Ereskovsky et al. 2007b) and Amphimedon queenslandica Hooper and van Soest, 2006 (as Reniera sp.) (Leys and Degnan 2002, Degnan et al. 2005, Larroux et al. 2006). Looking back, we can see that out of the 540 articles on sponge embryology, only 93 are devoted to embryonic development in the strict sense. They deal with 21 species of Calcarea, 2 species of Hexactinellida and about 70 species of Demospongiae (Fig. 7). Strangely enough, there is only two publications (Hill et al. 2004, Laroux et al. 2006) on the development of sponges during sexual reproduction where molecular-biological methods were used. Researchers only start to decode complex embryonic morphogenesis at the ultrastructural level (Boury-Esnault et al. 1999, 2003, Ereskovsky and Gonobobleva 2000, Ereskovsky and Boury-

Fig. 2: The trend in general sponge development publications between 1870 and 2006.

43

Fig. 3: German spongiologs portraits. A. Frans Eilhard Schulze (1840-1921) ZM B IX-608; B. Otto Maas (1867-1916); C. Oscar Schmidt (1823-1886); D. Conrad Keller (1848-1930) ZM B I/1753. (The photos are kindly given by C. Eckert).

Esnault 2002, Leys and Degnan 2002, Gonobobleva and Ereskovsky 2004a, Usher and Ereskovsky 2005, Leys et al. 2006) and investigate developmental genes expression during embryonic development (Hill et al. 2004, Laroux et al. 2006).

The present Gastrulation: verbal or real problem in sponges?

Applicability of the term gastrulation to sponge development is one of the sore points in our discussion (see: Efremova 1997, Leys 2004, Ereskovsky and Dondua 2006). There are two principal definitions of this term. The first is used by most, but not all developmental biologists: Gastrulation is the process in embryonic development in the course of which three primary germ layers are formed and the gut is formed through complex

cell migrations (Technau and Scholz 2003, Stern 2004, Keller 2005, Martindale 2005). The second definition is rare: Gastrulation is the process that results in a multilayered organism during embryonic development (Efremova 1997, Leys and Degnan 2002, Maldonado 2004, Leys 2004). According to these authors, the formation of a multilayered embryo during embryogenesis in sponges should be considered as gastrulation, since mechanisms of cell reorganization in the blastula are similar with those recognized as gastrulation in cnidarians. This contradiction stems from the absence of a generally accepted point of view on the homology of embryonic processes and their derivatives in sponges and other animals. Despite recent impressive progress in morphogenetic research in general, works on sponge embryonic morphogenesis are very rare. Therefore, investigations of mechanisms of sponge embryonic development are currently

44

Fig. 5: Paul Brien (1894-1975), Brussels, 1968. (The photo is kindly given by Ph. Willenz). Fig. 4: Yves Delage (1854-1920) at the Roscoff Marine laboratory, 1905 (From: Beetscen and Fischer, 2004).

much more important than terminological discussions. Morphogenesis is the mechanism responsible for creation of body plan during embryonic development, metamorphosis, asexual reproduction and regeneration. Morphogenetic investigations are a promising branch in developmental biology of sponges. Formation of multilayer embryos in Metazoa is achieved either by the migration of individual cells, or, by movements of cell sheets (Keller et al. 2003, Keller 2005). The former morphogenetic movements are known as mesenchymal morphogenesis or epithelialmesenchymal transitions. One such example is multipolar ingression (Shook and Keller 2003). The latter is epithelial morphogenesis and invagination is such an example (Keller et al. 2003, Gilbert 2003). Morphogenetic cell movements are determined by complex and specific gene systems. Their origin and evolution resulted in the diversity of metazoan developmental types. Apparently, they are involved in multicellular embryos formation in all animals, including sponges. For instance, formation of sponge larvae is accompanied by almost all types of cell movements, characteristic of Eumetazoa (Efremova 1997, Leys 2004, Maldonado 2004, Ereskovsky 2005, Leys and Ereskovsky 2006, Ereskovsky and Dondua 2006): cell delamination (Hexactinellida Oopsacas minuta Topsent, 1927) (Fig. 8A), morula delamination (Demospongiae: Dendroceratida, Dictyoceratida, Halichondrida, Haplosclerida) (Fig. 8B), invagination,

unipolar and multipolar ingression (Demospongiae: Halisarcida Halisarca dujardini) (Fig. 8C, D). At the same time, some unique morphogeneses, not found in other multicellular animals, have been described in sponges. They are, for example, multipolar egression in Homoscleromorpha (Demospongiae) (Fig. 8E), polarized delamination (Demospongiae: Poecilosclerida and Halichondrida) (Fig. 8F), excurvation in Calcaronea (Calcarea) (Fig. 8G), formation of blastula (pseudoblastula) by means of ingression of maternal cells into the embryo in Chondrosia reniformis Nardo, 1833 (Demospongiae: Chondrosida) (Fig. 8H), and unipolar proliferation (Demospongiae: Verticillitida Vaceletia crypta (Vacelet, 1977) (Fig. 8I). According to comparative embryological data on Porifera and Cnidaria, ancestors of Metazoa must have been able to form epithelial layers and to disaggregate these layers into individual cells. They were capable of epithelial morphogenesis and also had regulatory mechanisms controlling cell ingression and ensuring directed movement of cell masses. It may be therefore concluded that mechanisms of morphogenetic movements emerged in the course of evolution prior to the separation of the covering cell layer as ectoderm and the internal digestive cell mass as endoderm. This testifies to the independence of processes of spatial distribution of cells and their specification in the forming embryo as ectodermal and endodermal tissues (Ereskovsky and Dondua 2006). So, it is incorrect to apply the term gastrulation to sponge development.

45

larvae have a strongly pronounced anterior-posterior polarity, distinct photoreception and other kinds of taxis (Maldonado 2004). Some demosponge larvae have desmosome-like cell junctions (Fig. 9). Finally, it has been shown that parenchymella of A. queenslandica possesses some of the transcription factor genes that appear to be characteristic of Metazoa. They are expressed during the development of this species (Larroux et al. 2006).

Cellular and molecular basis of embryonic morphogenesis in sponges Cellular basis of embryonic morphogenesis
During metazoan embryonic development, the cells can undergo changes either autonomously or in conjunction with their neighbors to form an embryo. Most of morphogenetic movements require that a subset of cells detach from their neighbors and acquire properties allowing them to migrate to new position. Obviously, the consequences of changes in cell shape and motility will be quite different if cells are joined in an epithelium or if they are unconstrained by neighbors. Cell motility is generated by contractile elements of the cytoskeleton. The following question requiring an answer arises: What is the cytoskeleton dynamics during embryonic morphogenesis in sponges?

Cell-extracellular matrix adhesion

One of the main molecules that mediate cell anchorage to the substratum during the morphogenesis is integrin, which are key molecules during early animal development (Darribere et al. 2000). (Integrins, members of the transmembrane linker proteins family, traverse the cell membrane, anchoring the actin microfilaments on the inside and may bind to the fibronectin and in other extracellular matrix proteins). Integrins were shown to be present in some adult demosponges: Ophlitaspongia tenuis and Microciona prolifera Ellis and Solander, 1786 (Brower et al. 1997, Kuhns et al. 2001, Sabella et al. 2004), Geodia cydonium (Jameson, 1811) (Pancer et al. 1997, Mller 1997) and in Suberites domuncula (Olivi, 1792) (Wimmer et al. 1999). The following questions arise: Are integrins involved in embryonic development of sponges? Is their morphogenetic role the same in sponges and in other animals?

Intercellular adhesion
Fig. 6: A. Odette Tuzet (1903-1976) and O. Duboscq (1868-1943), Banuls-sur-Mer Marine laboratory, 1937; B. Claude Lvi, Paris, 2000 (The photo is kindly given by J. Vacelet).

Evolutionary importance of larvae

Evolutionary importance has been attached to larvae of Bilateria since A. Kowalevskys studies on ascidia (1866). This idea has recently received molecular-biological support (Raff 1994, Peterson and Davidson 2000). Indeed, sponge

Cell-cell interactions are also important for tissue formation during development. A remarkable feature of sponges is that when dissociated to single cells they can undergo species-specific reaggregation (Wilson 1907). This is mediated by an extracellular proteoglycan complex, known as aggregation factor (AF) that acts as a bridge between receptor proteins on neighboring cells (Schutze et al. 2001). The AF receptor also possesses an RGD (Arg-GlyAsp attachment site) integrin-binding motif. RGD containing peptides will block AF-mediated aggregation. Both the RGD peptide and AF stimulate a range of intracellular responses (Wimmer et al. 1999). It was proposed that binding of AF promotes interaction between the RGD of the AF receptor

46

Fig. 7: The trend in sponge embryonic development publications between 1870 and 2006.

Fig. 8: Different types of morphogenesis in sponges resulting in larva formation: A. Cell delamination (Hexactinellida Oopsacas minuta); B. Morula delamination (Demospongiae: Dendroceratida, Dictyoceratida, Halichondrida, Haplosclerida); C. Invagination (Halisarca dujardini, Demospongiae); D. Multipolar ingression (H. dujardini, Demospongiae); E. Multipolar egression (Homoscleromorpha, Demospongiae); F. Polarized delamination (Demospongiae: Poecilosclerida and Halichondrida); G. Excurvation (Calcaronea, Calcarea); H. Formation of blastula (pseudoblastula) by means of ingression of maternal cells into the embryo (Chondrosia reniformis, Demospongiae: Chondrosida); I. Unipolar proliferation (Demospongiae: Verticillitida Vaceletia crypta). (From: Ereskovsky and Dondua 2006).

47

Fig. 9: Semi-thin micrographs of demosponges larvae with the desmosom-like cell junctions (insets). A. Parenchymella of Ircinia oros (Dictyoceratida); B. Disphaerula of Halisarca dujardini (Halisarcida); C. Cinctoblastula of Corticium candelabrum Schmidt, 1862 (Homoscleromorpha); D. Paren-chymella of Pleraplysilla spinifera (Schulze, 1879) (Dictyoceratida). Abbreviations: AP anterior pole, PP posterior pole. Scale bar, A 100 m; Inset 0,2 m; B - 50 m; Inset 25 nm; C - 50 m; Inset 0,2 m; D 50 m; Inset 0,2 m.

and sponge integrin proteins (Harwood and Coates 2004). Many excellent studies dealt with the fine mechanisms of cell-cell interactions in sponge cell cultures (see: FernandezBusquets et al. 2002, Misevic et al. 2004). Nevertheless, no cadherin, catenin or related proteins have been identified in sponges (Harwood and Coates 2004). However, there is not a single work demonstrating either specific or differential cellular adhesiveness in sponge embryonic development.

Developmental genes in sponges

The presence of metazoan developmental genes in demosponge genomes has been shown (e.g., Degnan et al. 1993, 1995, Coutinho et al. 1994, 2003, Seimiya et al. 1994, 1997, Hoshiyama et al. 1998, Richelle-Maurer et al. 1998, Manuel and Le Parco 2000, Adell et al. 2003, Perovic et al. 2003, Wiens et al. 2003a, b, Adell and Muller 2004, Hill et al. 2004, Manuel et al. 2004, Richelle-Maurer et al. 2006, Larroux et al. 2006, 2007). However, their roles in

embryogenesis and metamorphosis are unknown. To date, our understanding of sponge gene expression is restricted chiefly to asexual reproductive processes, such as gemmule germination, cell aggregation, and primorphs formation. Hill et al. (2004) followed the expression of non-Hox Antp-class Bar-/Bsh-like gene during larva releasing, larva swimming and metamorphosis. Laroux et al. (2006) demonstrated that an extensive range of metazoan transcription factor genes, including members of the ANTP class (outside Hox, ParaHox, and extended-Hox clades), Pax, POU, LIM-HD, Sox, nuclear receptor (NR), Fox (forkhead), T-box, Mef2, and Ets gene classes are expressed during A. queenslandica (Haplosclerida) development. These data combined with developmental gene expression patterns from other animals suggest that these genes may have played important regulatory roles in the embryos of the first metazoans. These works will probably now trigger an explosion of studies on the role of developmental genes in sponge development.

48

Fig. 10: The basal-apical and posterior-anterior axis of sponges on different stage of its life cycle. A. Haliclona sp. from White Sea. B. Diagram of demosponges organization. C, D. Semi-thin micrograph (C) and diagram (D) of Halisarca dujardini (Halisarcida) rhagon. E, F. SEM micrograph (E) and diagram (F) of Halisarca dujardini (Halisarcida) disphaerula larva. The arrows indicate the basal-apical (A D) and posterior-anterior (E, F) axis. Abbreviation: AP anterior pole, CC choanocyte chamber; Ep exopinacoderm; O osculum; PP posterior pole. A 2 cm; B 250 m; C - 50 m; D - 50 m; E - 50 m; F - 50 m.

Axis formation
An important characteristic distinguishing sponges from higher metazoans is the nature of body symmetry. Higher animals have two obvious body axes, anterior-posterior and dorsal-ventral, and are therefore bilaterally symmetrical (Bilateria). All young (rhagon or olynthus) and monooscular sponges, by contrast, have a single overt axis (apical-basal) defined by the presence of an osculum at one end (Fig. 10 AD). The question is: Does the apical-basal axis of an adult sponge correspond to the posterior-anterior axis of higher animals? Since the larvae of all the sponges investigated also possess an apical-basal axis, the answer to this question may be yes (Ereskovsky 2005) (Fig. 10E, F).

Development: phylogeny and evolution

The following question is very important in this respect: Can embryological data be applied to sponge phylogeny and evolution? According to the paraphyletic hypotheses, based on molecular data, Porifera consists of four groups: Calcarea, Hexactinellida, Demospongiae and Homoscleromorpha (Borchiellini et al. 2001, 2004). I proposed seven

developmental types in recent Porifera: I - trichimella (Hexactinellida); II - calciblastula (Calcinea); III - amphiblastula (Calcaronea); IV - cinctoblastula (Homoscleromorpha); V - disphaerula (Halisarcida); VI - direct development (Tetilla, Spirophorida); VII parenchymella (Ereskovsky 2004). There are four principal cleavage patterns in sponges: incurvational, polyaxial, radial and chaotic (Fig. 11: 1-4). They result in formation of three main types of blastula: stomoblastula, coeloblastula and stereoblastula (Fig. 11: 57). The latter two types are derived from different cleavage patterns. Different embryonic morphogeneses lead to 8 or 9 larval types (Fig. 11: 8-22). Difficulties of using embryological data for phylogenetic interpretation of Porifera are associated with a high degree of polymorphism of their development. The same cleavage pattern and blastula type may be characteristic of several different larval types. For example, radial cleavage, resulting in coeloblastula and stereoblastula, leads to parenchymella, trichimella, direct development, and coeloblastula (Fig. 11: 3-62-13, 14, 15; 71-16, 17, 18). On the other hand, the same larval type may be the result of different cleavage patterns and modes of morphogenesis.

49

Fig. 11: Diagram of sponges cleavage and morphogenesis, leading to the larvae. 14 Cleavage patterns in sponges: incurvational (1), polyaxial (2), radial (3), and chaotic (4). Three main form of sponges blastula: stomoblastula (5), coeloblastula (6), and stereoblastula (7). Different larval types of sponges: amphiblastula of Calcaronea (Calcarea) (8); calciblastula of Calcinea (Calcarea) (9); coeloblastula (10), parenchymella (11), and disphaerula (12) of Halisarca (Halisarcida); parenchymella of Vaceletia crypta (Verticillitida) (13); pseudoblastula of Chondrosia reniformis (Chondrosida) (14); trichimella of Oopsacas minuta (Hexactinellida) (15); juvenile of Tetilla under direct development (16); parenchymella of Tethya aurantium (Pallas, 1766; Hadromerida) (17); coeloblastula of Polymastia robusta Bowerbank, 1866 (Hadromerida) (18); parenchymella of Dictyoceratida (19); parenchymella of freshwater Haplosclerida (20); parenchymella of Poecilosclerida (21); cinctoblastula of Homoscleromorpha (22).

50

For example, parenchymella may originate from polyaxial, radial and chaotic cleavage, using different modes of morphogenesis (Fig. 11: 11, 13, 17, 19-21). Coeloblastula larva can result from polyaxial (Halisarcida and Calcinea) as well as radial cleavage (Hadromerida), through coeloblastula or stereoblastula stages (Fig. 11: 9, 10, 18). The fact that similar characters can result from different developmental pathways means that ontogenetically earlier stages can be evolutionarily altered. The opposite case showing early similarity with later occurring differences is more common. However, both aspects taken together reveal that in the course of evolution developmental stages may be altered at all levels, from the molecular to the morphogenetic, regardless of whether a stage occurs early or late during the ontogenetic process. Results of the comparative analysis of the cleavage and embryonic morphogenesis testify that these characters taken separately cannot form a basis for phylogenetic constructions within the Porifera. For example, Calcinea (Calcarea) embryogenesis (Fig. 11: 2-61-9) is much closer to the development of Halisarcida (Demospongiae) (Fig. 11: 2-6110) than to Calcaronea (Fig. 11: 1-5-8). Thus, a variety of cleavage patterns, types of blastulae and morphogenesis, leading to larvae formation in sponges, does not allow making a conclusion about certain linear ways of developmental evolution for all Porifera. It testifies to an early divergence of sponge macrogroups or, more likely, paraphyly and their long parallel evolution.

Acknowledgments
I thank Prof. Galina Korotkova, Prof. Archil Dondua (SaintPetersburg State University), Dr. Nicole Boury-Esnault, Dr. Jean Vacelet, Dr. Carole Borchiellini, Dr. Thierry Perez (Centre dOcanologie de Marseille, France) for helpful discussions, Carsten Eckert (Museum fr Naturkunde, Zentralinstitut der Humboldt, Universitt zu Berlin, Germany) and Dr. Philippe Willenz (Royal Belgian Institute of Natural Sciences, Brussels) and J. Vacelet for providing spongiologs portraits, Daria Tokina (Zoological Institute of RAS, Saint-Petersburg, Russia) for technical assistance, and Natalia Lentsman for improving the English. This work was funded by the program RFBR N 06-04-48660 and 06-04-58573.

References
Adell T, Grebenjuk VA, Wiens M, Mller WEG (2003) Isolation and characterization of two T-box genes from sponges, the phylogenetically oldest metazoan taxon. Dev Genes Evol 213: 421-434 Adell T, Mller WEG (2004) Cloning and characterization of T-box and forkhead transcription factors from Porifera. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Bull Mus Ist Biol Univ Genova 68: 163-173 Bidder GP (1898) The skeleton and classification of calcareous sponge. Proc Roy Soc 64: 61-76 Borchiellini C, Manuel M, Alivon E, Boury-Esnault N, Vacelet J. Le Parco Y (2001) Sponge paraphyly and the origin of Metazoa. J Evol Biol 14(1): 171-179 Borchiellini C, Chombard C, Manuel M, Alivon E, Boury-Esnault N, Vacelet J (2004) Molecular phylogeny of Demospongiae: implications for classification and scenarios of character evolution. Mol Phyl Evol 32: 823-837 Boury-Esnault N, Efremova SM, Bzak C, Vacelet J (1999) Reproduction of a hexactinellid sponge: first description of gastrulation by cellular delamination in the Porifera. Invertebr Rep Dev 35: 187-201 Boury-Esnault N, Ereskovsky AV, Bezac C, Tokina D (2003) Larval development in Homoscleromorpha (Porifera, Demospongiae) first evidence of basal membrane in sponge larvae. Invertebr Biol 122: 187-202 Brower DL, Brower SM, Hayward DC, Ball EE (1997) Molecular evolution of integrins - genes encoding integrin beta subunits from a coral and a sponge. Proc Natl Acad Sci 94(17): 9182-9187 Coutinho CC, Vissers S, van de Vyver G (1994) Evidence of homeobox genes in the freshwater sponge Ephydatia fluviatilis. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 385-388 Coutinho CC, Fonseca RN, Mansure JJ, Borojevic R (2003) Early steps in the evolution of multicellularity: deep structural and functional homologies among homeobox genes in sponges and higher metazoans. Mech Dev 120: 429-440 Darribere T, Skalski M, Cousin H, Gaultier A, Montmory C, Alfandari D (2000) Integrins: regulators of embryogenesis. Biol Cell 92: 5-25 Degnan BM, Degnan SM, Naganuma T, Morse DE (1993) The ets multigene family is conserved throughout the Metazoa. Nucl Acids Res 21(15): 3479-3484 Degnan BM, Degnan SM, Giusti A, Morse DE (1995) A Hox/Hom homeobox gene in sponges. Gene 155(2): 175-177

The future
We are now on the threshold of the fourth period of sponge developmental studies. To enhance our knowledge on this topic, the following steps are currently necessary: - To select some model sponge species with different types of development; - To investigate their development from egg to juvenile at ultrastructural level; - To decode morphogenetic mechanisms of development using ultrastructural and molecular methods. Studies of cellular and molecular basis of embryonic morphogenesis in sponges will provide answers to the following important questions: - What is the role of intercellular contacts and cytoskeleton dynamics in embryonic and postembryonic morphogenesis? - What is the role of cell-cell and cell-extacellular matrix interactions? - Are integrins, laminins, and signaling molecules involved in development the same in sponges and other metazoans? - Which developmental genes work during sponge embryonic development? - Does the apical-basal axis of adult sponge correspond to the anterior-posterior axis of higher animals?

51

Degnan BM, Leys SP, Larroux C (2005) Sponge development and antiquity of animal pattern formation. Integr Comp Biol 45: 335341 Delage Y (1892) Embryogenie des ponges. Arch Zool Expr Gn 10: 345-498 Delage Y (1899) On the position of sponges in the Animal Kingdom. In: Sedgwick A (ed). Proceedings of the 4th International Congress of Zoology. Clay & Sons, Cambridge. pp 57-62 Efremova SM (1997) Once more on the position among Metazoa - Gastrulation and germinal layers of sponges. In: Ereskovsky AV, Keupp H, Kohring R (eds). Modern problems of poriferan biology. Freie Universitat, Berlin. pp. 7-15 Ereskovsky AV, Gonobobleva EL (2000) New data on embryonic development of Halisarca dujardini Johnston, 1842 (Demospongiae: Halisarcida). Zoosystema 22: 355-368 Ereskovsky AV, Boury-Esnault N (2002) Cleavage pattern in Oscarella species (Porifera, Demospongiae, Homoscleromorpha), transmission of maternal cells and symbiotic bacteria. J Nat Hist 36: 1761-1775 Ereskovsky AV (2002) Polyaxial cleavage in sponges (Porifera): a new pattern of Metazoan cleavage. Dokl Biol Sci 386: 472-474 Ereskovsky AV (2004) Comparative embryology of sponges and its application for poriferan phylogeny. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Bull Mus Ist Biol Univ Genova 68: 301-318 Ereskovsky AV (2005) Comparative embryology of sponges. St. Petersburg State Univ. Press. St. Petersburg Ereskovsky AV, Gonobobleva EL, Vishnyakov AE (2005) Morphological evidence for vertical transmission of symbiotic bacteria in the viviparous sponge Halisarca dujardini Johnston (Porifera, Demospongiae, Halisarcida). Mar Biol 146(5): 869-875 Ereskovsky AV, Dondua AK (2006) The problem of germ layers in sponges (Porifera) and some issues concerning early metazoan Evolution. Zool Anzeiger 245: 65-76 Ereskovsky AV, Konjukov P, Willenz P (2007a) Experimental metamorphosis of Halisarca dujardini larvae (Demospongiae, Halisarcida): evidence of flagellated cell totipotentiality. J Morphol 268(6): 529-536 Ereskovsky AV, Tokina DB, Bzac C, Boury-Esnault N (2007b) Metamorphosis of cinctoblastula larvae (Homoscleromorpha, Porifera). J Morphol 268(6): 518-528 Fernandez-Busquets XK, William J, Simpson TL, Ho M, Gerosa D, Grob M, Burger MM (2002) Cell adhesion-related proteins as specific markers of sponge cell types involved in allogeneic recognition. Dev Comp Immunol 26(4): 313-323 Gilbert SF (2003) Developmental biology. Sinauer Associates Inc. Sunderland Gonobobleva EL, Ereskovsky AV (2004a) Polymorphism in free-swimming larva of Halisarca dujardini (Demospongiae, Halisarcida). In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Bull Mus Ist Biol Univ Genova 68: 349-356 Gonobobleva EL, Ereskovsky AV (2004b) Metamorphosis of the larva of Halisarca dujardini (Demospongiae, Halisarcida). Bull Inst roy Sci Nat Belgique Biol 74: 101-115 Haeckel E (1866) Generelle Morphologie der Organismen. Berlin, George Reismer Haeckel E (1872) Die Kalksschwmme. Eine Monographie in zwei Bnden. Georg Reimer, Berlin

Haeckel E (1874) Die Gastrae Theorie, die phylogenetische Classification des Thierreichs und die Homologie der Keimblatter. Jena Ztschr Naturwiss 8: 1-55 Harwood A, Coates JC (2004) A prehistory of cell adhesion. Cur Opinion Cell Biol 16: 470-476 Hill A, Tetrault J, Hill M (2004) Isolation and expression analysis of a poriferan Antp-class Bar-/Bsh-like homeobox gene. Dev Genes Evol 214: 515-523 Hoshiyama D, Suga H, Iwabe N, Koyanagi M, Nikoh N, Kuma K, Matsuda F, Honjo T, Miyata T (1998) Sponge pax cDNA related to pax-2/5/8 and ancient gene duplications in the pax family. J Mol Evol 47(6): 640-648 Keller R, Davidson LA, Shook DR (2003) How we are shaped: the biomechanics of gastrulation. Differentiation 71: 171-205 Keller R (2005) Cell migration during gastrulation. Curr Opin Cell Biol 17: 533-541 Kowalevsky A.O. (1866) Entwiklungsgeschichte der einfachen Ascidien. Mm Acad Sci St Petersbourg 10(15): 1-16 Kuhns WJ, Rusciano D, Kaltenbach J, Ho M, Burger MM, FernandezBusquets X (2001) Up-regulation of integrins alpha(3) beta(1) in sulfate-starved marine sponge cells: functional correlates. Biol Bull 201(2): 238-239 Larroux C, Fahey B, Liubicich D, Hinman VF, Gauthier M, Gongora M, Green K, Wrheide G, Leys SP, Degnan BM (2006) Developmental expression of transcription factor genes in a demosponge: insights into the origin of metazoan multicellularity. Evol Dev 8: 150-173 Larroux C, Fahey B, Degnan SM, Adamski M, Rokhsar DS, Degnan BM (2007) The NK homeobox gene cluster predates the origin of Hox genes. Curr Biol 17: 1-5 Lvi C (1956) tude des Halisarca de Roscoff. Embryologie et systematique des Dmosponges. Trav Stat Biol Roscoff NS 7: 3181 Leys SP (2004) Gastrulation in sponges. In: Stern CD (ed). Gastrulation. From cells to embryo. Cold Spring Harbor Lab Press, New York. pp. 23-32 Leys SP, Degnan BM (2002) Embryogenesis and metamorphosis in a haplosclerid demosponge: gastrulation and transdifferentiation of larval ciliated cells to choanocytes. Invertebr Biol 121: 171-189 Leys SP, Cheung E, Boury-Esnault N (2006) Embryogenesis in the glass sponge Oopsacas minuta: formation of syncytia by fusion of blastomeres. Integr Compar Biol 46(2): 104-117 Leys SP, Ereskovsky AV (2006) Embryogenesis and larval differentiation in sponges. Can J Zool 84: 262-287 Maldonado M (2004) Choanoflagellates, choanocytes, and animal multicellularity. Invertebr Biol 123: 1-22 Manuel M, Le Parco Y (2000) Homeobox gene diversification in the Calcareous sponge, Sycon raphanus. Mol Phylog Evol 17: 97-107 Manuel M, Le Parco Y, Borchiellini C (2004) Comparative analysis of Brachyury T-domains, with the characterization of two new sponge sequences, from a hexactinellid and a calcisponge. Gene 340: 291-301 Martindale MQ (2005) The evolution of metazoan axial properties. Nature Rev Genetics 6: 917-927 Minchin EA (1896) Suggestions for a natural classification of the Asconidae. Ann Mag Nat Hist 18: 349-362 Misevic GN, Guerardel Y, Sumanovski LT, Slomianny MC, Demarty M, Karamanos Y, Maes E, Popescu O, Strecker G (2004) Molecular recognition between glyconectins as an adhesion self-assembly pathway to multicellularity. J Biol Chem 279: 15579-15590

52

Mukhina YI, Kumeiko VV, Podgornaya OI, Efremova SM (2006) The fate of larval flagellated cells during metamorphosis of the sponge Halisarca dujardini. Int J Dev Biol 50: 533-541 Mller WEG (1997) Origin of metazoan adhesion molecules and adhesion receptors as deduced from cDNA analyses in the marine sponge Geodia cydonium: a review. Cell Tissue Res 289(3): 383395 Pancer Z, Kruse M, Mller I, Mller WEG (1997) On the origin of metazoan adhesion receptors: cloning of integrin alpha subunit from the sponge Geodia cydonium. Mol Biol Evol 14: 391-398 Pavans de Ceccatty M (1986) Cytosceletal organisation and tissue patterns of epithelia in the sponge Ephydatia muelleri. J Morphol 189: 45-65 Perovic S, Schrder HC, Sudek S, Grebenjuk VA, Batel R, Mauro Tifani IMM, Mller WEG (2003) Expression of one sponge Iroquois homeobox gene in primmorphs from Suberites domuncula during canal formation. Evol Devel 5(3): 240-250 Peterson KJ, Davidson EH (2000) Regulatory evolution and the origin of the bilaterians. Proc Natl Acad Sci 97: 4430-4433 Raff R (1994) Developmental mechanisms in the evolution of animal form. In: Bengdson S (ed). Early life of earth. Nobel symposium. N 84. Columbia University Press, New York. pp. 489-500 Richelle-Maurer E, van de Vyver G, Vissers S, Coutinho CC (1998) Homeobox-containing genes in freshwater sponges: characterization, expression, and phylogeny. In: Mller WEG (ed). Molecular evolution: evidence for monophyly of Metazoa. Springer Verlag, vol. 19. Berlin. pp. 157-175 Richelle-Maurer E, Boury-Esnault N, Itskovich VB, Manuel M, Pomponi SA, van de Vyver G (2006) Conservation and phylogeny of a novel family of non-Hox genes of the Antp class in Demospongiae (Porifera). J Mol Evol 63: 222-230 Sabella CA, Faszewski EE, Kaltenbach JC, Kuhns WJ, Burger MM, Fernandez-Busquets X (2004) Immunocytochemical detection of integrins alpha(3) and beta(1) in allografts of the marine sponge, Microciona prolifera. Biol Bull 207: 162 Schutze J, Krasko A, Diehl-Seifert B, Mller WEG (2001) Cloning and expression of the putative aggregation factor from the marine sponge Geodia cydonium. J Cell Sci 114: 3189-3198

Seimiya M, Ishiguro H, Miura K, Watanabe Y, Kurosawa Y (1994) Homeobox-containing genes in the most primitive Metazoa, the sponges. Europ J Biochem 221: 219-225 Seimiya M, Watanabe Y, Kurosawa Y (1997) Identification of POUclass homeobox genes in a freshwater sponge and the specific expression of these genes during differentiation. Europ J Biochem 243: 27-31 Shook D, Keller R (2003) Mechanisms, mechanics and function of epithelialmesenchymal transitions in early development. Mech Devel 120: 1351-1383 Stern C (ed) (2004). Gastrulation: from cells to embryo. Cold Spring Harbor Press, Cold Spring Harbor Technau U, Scholz CB (2003) Origin and evolution of endoderm and mesoderm. Int J Dev Biol 47: 531-539 Usher K, Ereskovsky AV (2005) Larval development, ultrastructure and metamorphosis in Chondrilla australiensis Carter, 1873 (Demospongiae, Chondrosida, Chondrillidae). Invert Rep Dev 47: 51-62 Weissenfels N (1989) Biologie und Mikroskopische Anatomie der Swasserschwmme (Spongillidae). Gustav Fischer Verlag. Stuttgart, New York Wiens M, Mangoni A, DEsposito M, Fattorusso E, Korchagina N, Schroder HC, Grebenjuk VA, Krasko A, Batel R, Mller IM, Mller WEG (2003) The molecular basis for the evolution of the metazoan bodyplan: extracellular matrix-mediated morphogenesis in marine demosponges. J Mol Evol 57: S60-S75 Wiens M, Batel R, Korzhev M Mller WEG (2003) Retinoid X receptor and retinoic acid response in the marine sponge Suberites domuncula. J Exp Biol 206: 3261-3271 Wilson HV (1907) On some phenomena of coalescence and regeneration in sponges. J Exp Zool 5: 245-258 Wimmer W, Perovic S, Kruse M, Schroder HC, Krasko A, Batel R, Mller WEG (1999) Origin of the integrin-mediated signal transduction -functional studies with cell cultures from the sponge Suberites domuncula. Europ J Biochem 260: 156-165

Porifera research: Biodiversity, innovation and sustainaBility - 2007

53

Sponge coordination, tissues, and the evolution of gastrulation

Sally P. Leys
University of Alberta. CW 405, Biological Sciences. Edmonton, Alberta, T6G 2E9. Phone: (780) 492-6629. Fax: (780) 4929234. sleys@ualberta.ca Abstract: One of the unifying features of animals is that they carry out rapid, coordinated movement. This ability results from the early evolution of tissues that can both conduct signals and contract. The origin of tissues is thus intimately tied to the origin of nerves and muscle, which have long been considered the major innovations of cnidarians (anemones and jellyfish). However, the hypothesis that muscle may have conferred important selective advantages in preying and escape in the cnidarian ancestor suggests that this ancestor also had nerves, or at least the ability to coordinate contractions of its muscle. These ideas are supported by a considerable body of work showing that the genes which in triploblasts are involved in muscle specification and differentiation are expressed during the development of medusa and polyp in three cnidarian model species. Recent data suggest that sponges are paraphyletic, implying that cnidarians and sponges share a common ancestor that had a sponge-like body plan. Here I re-examine evidence for a coordinated contraction system in modern cellular sponges as evidence for functional tissues. I suggest that coordination of the animal is evidence that sponges do possess tissues which arise by gastrulation-like processes during embryogenesis as is the case in other metazoans. The sponge peristaltic contractile system may represent the foundations of coordinating tissues and have set the stage for the innovation of nerves and muscle in later animals. Keywords: evolution of nerves, coordination, evolution of tissues, Porifera

Introduction:
Although the Porifera have long been regarded as an evolutionary side-branch of metazoans (Parazoans), recent molecular phylogenies propose that sponges are paraphyletic that calcareous sponges (Calcispongia) and possibly Homoscleromorphs (Peterson, personal communication) are more closely related to cnidarians and other metazoans than they are to other sponges (Borchiellini et al. 2001, Medina et al. 2001). A paraphyletic Porifera suggests that the ancestral metazoan was a sponge-like organism, a suggestion that may not be as revolutionary as it first appears. The sponge body plan is only difficult to grasp in the light of work which suggests that sponges are colonies of a few types of cells organized around a system of water canals (Simpson 1984), animals that lack true epithelia (Mackie 1984, Tyler 2003), and the ability to coordinate activity at the level of the whole organism (Mackie 1979). In contrast, recent work shows that sponge epithelia are sealed units (Gonobobleva and Ereskovsky 2004) with tight junction proteins (Adell et al. 2004) and basement membranes (Boute et al. 1996, Boury-Esnault et al. 2003, Maldonado 2004). Sponges lack neurons (Pavans de Ceccatty 1989) but syncytial sponge tissues are nonetheless excitable and propagate electrical signals that control the feeding current (Leys and Mackie 1997). In cellular sponges external stimuli (such as contact by amphipods) trigger waves of contraction (Nickel 2004), and similar contractile waves are known to be widespread among

all cellular sponges. Though slow, the contractions in cellular sponges illustrate the components of peristaltic waves seen in animals with a nervous system. Peristalsis is an efficient mechanism for controlling fluid movement through a tube, and typically consists of a series of motor patterns that control relaxation in front of and contraction behind the object being moved by the fluid. The same pulsating movement occurs in the hearts of the invertebrate chordate Amphioxus (Holland et al. 2003) and ascidians, as well as in the gut and circulatory systems of insects and molluscs. In the sea pansy Renilla, a member of the most basal cnidarian group the Pennatulacea, peristaltic contractions of the gastrovascular cavity (GVC) a tube-like gut that pervades the entire animal control the movement of fluid for feeding and respiration as well as gamete release (Mechawar and Anctil 1997). I suggest that the sponge body represents the first elaboration of a peristaltic contractile system in Metazoa, a system that was later adapted for locomotion, digestive and circulatory activity, and which gave rise to the hydrostatic skeleton. It is likely that elements of signaling used in these activities in higher animals may be found in extant sponges. I further suggest that coordinated contractions in sponges is evidence that these animals do possess tissues, and that these must have arisen during embryogenesis via gastrulation-like processes as in other animals. It is our challenge to understand what elements of tissues known in higher animals are used in sponges.

54

Propagated contractions in sponges

Cellular sponges have long been known to contract inhalant and exhalent openings (ostia and oscula), and portions of their canal system (Mackie 1979). The slow rates of propagation (4-400 m s-1) and difficulty of observation in large animals have led to the conclusion that events are local and decremental. But at least three species show propagated contractions that are involved in expulsion of sediment (Nickel 2004) or gametes (Reiswig 1970), and rhythmic, diurnal pulses that may assist water flow through the animal (Weissenfels 1990, Nickel 2004). Given the slowness of the events, electrical signaling is not likely to be involved. Although glass sponges can propagate electrical signals, they do this through an uninterrupted giant syncytium (Leys and Mackie 1997). So far there is no convincing evidence of gap junctions (or other communicating junctions) in cellular sponges that would allow equivalent, rapid signalling (see Leys and Meech 2006). The slowness of the events suggests a slower mechanism of signaling via extracellular molecules is possible. Neurotransmitter molecules have been localized in tissues of calcareous sponges (Lentz 1966) and demosponge larvae (Weyrer et al. 1999), and many of these chemicals have been shown to affect the contraction of ostia and oscula (Parker 1910, Prosser et al. 1962, Emson 1966, Prosser 1967). Fascinating videos showing rhythmic contractions of the asconoid calcareous sponge Leucosolenia (C. Bond, personal communication) and time-lapse photographs of contractions in leuconoid calcareous sponges (Gaino et al. 1991) suggest the habit is widespread among cellular sponges.

Mechanism(s) of coordination of contractions

Video microscopy and image analysis of contractile behaviour show that both marine and freshwater sponges (de Vos and van de Vyver 1981, Weissenfels 1990, Nickel 2004) control the movement of water through a single aquiferous system (one osculum) in a manner similar to peristaltic contractions in the pennatulacean anthozoan Renilla. However, whereas most adult sponges are opaque to microscopy, juvenile demosponges are usually transparent so that individual cells crawling and cells forming epithelial linings to canals can be observed in vivo. Freshwater sponges have the added advantage that they can be readily grown in vitro from gemmules (overwintering cysts), thereby providing a relatively easy preparation for analysis of the mechanism of signaling. Experiments can be carried out on 7 day-old juvenile sponges hatched at room temperature using wellestablished protocols (de Vos 1971, de Vos and van de Vyver 1981, Francis and Poirrier 1986, Elliott 2004). Contractions begin after mechanical or chemical stimuli, and kinetics of contractions can be determined by computer assisted image analysis (1 image/10s) (Fig. 1, Elliott and Leys 2007). After each treatment the sponge carries out a stereotypical behaviour involving dilation and contraction, effectively expelling water (and any particulates) from the canal system. How are contractions propagated? The problem has been considered in depth by Jones (1962) who suggested the following possibilities: local changes in pressure that induce contractions some distance away; stretch receptors acting

sequentially in adjacent cells; release of an aqueous hormone into the water or of a chemical messenger into the mesohyl; and local (non-propagating) action potentials that function to excite adjacent cells. Although recent work has focused on the secreted hormone hypothesis (Ellwanger et al. 2004, Ellwanger and Nickel 2006), it is most likely that several mechanisms interact. For example, a local change in pressure could activate stretch receptors, which in turn could trigger the release of a locally acting messenger. In Ephydatia muelleri responses to mechanical stimuli involve a peristaltic-like wave of dilation and contraction (Leys and Meech 2006), yet spasms also occur simultaneously on either side of a single sponge (Ellwanger et al. 2004, Elliott and Leys 2007) how are these triggered? Perhaps minute changes in pressure stretch cell membranes at a distant location and trigger an apparently simultaneous contraction. Waves of contraction can also be seen to travel down (along) a canal and at the same time across canals. While a pressure wave could precede the contraction in both directions, evidence that amoeboid cells in the mesohyl cease crawling as the contraction passes (Fig. 2) point to a secreted messenger. Many molecules have been shown to trigger oscular and ostia closure, and to intiate contractions (reviewed by Jones 1962, Lawn 1982, Leys and Meech 2006). Widespread evidence for glutamate in signaling in plants (Demidchik et al. 2004), Paramecium (Yang et al. 1997) and astrocytes (Nedergaard et al. 2002), and evidence for metabotropic Glu/GABA receptors in sponges (Perovic et al. 1999), makes this molecule an especially good candidate for a signaling molecule. Indeed, recent experiments in both Tethya (Ellwanger and Nickel 2006) and Ephydatia (Elliott and Leys 2007) suggest that contractions can be triggered by application of glutamate in a concentration dependent manner. The working hypothesis is that contractions propagate via calcium waves as in mammalian astrocytes (Nedergaard 1994) and mast cells (Osipchuk and Cahalan 1992), either via the direct action of stretch receptors or by the release of locally acting chemicals, much as outlined by Jones (1962): a stimulus (pressure/mechanical) causes stretch receptors to trigger a rise in intracellular calcium, causing contraction of the cell, tension on adjacent cells, and stimulating release of a secreted messenger (such as glutamate), which in turn stimulates contraction of nearby cells, and so on. The slow rates of contraction around 20 m s-1 lend support to this hypothesis. However, some contractions, like that which travels up the osculum of E. muelleri at up to 375 m s-1 (McNair, 1923), are much faster. Since gap junction coupling is enhanced in the presence of glutamate (Enkvist and McCarthy 1994), it is possible that cells may be coupled by almost gap junctions, which connect cells as glutamate levels rise. During peristalsis in cnidarians and higher animals, nitric oxide signaling allows relaxation prior to contraction (Moroz et al. 2004, Anctil et al. 2005). Preliminary results show that nitric oxide (NO) synthase staining in E. muelleri tissues fixed for NADPH-diaphorase (Elliott and Leys 2007), and experiments by Ellewanger and Nickel (2006) suggest that NO modulates contractions in Tethya. Thus waves of contraction along the aquiferous canals in sponges may be modulated much in the same way they are in cnidarians

55

Fig. 1: Dilation and contraction of the aquiferous system of a 7-day old juvenile Ephydatia muelleri. White arrows point to the osculum, which increases in diameter by frame C as the canals dilate in A and B, and contract in C and D. The white regions between the dilated exhalent canals in (B) are the compressed inhalant canals (Elliott and Leys 2007).

Fig. 2: Cells crawling through the mesohyl stop when a wave of contraction passes by. A, B. India ink added to a dish with a sandwich preparation is taken into the chambers (black oblique lines). When all chambers are filled a wave of contraction propagates along the canals (white arrow pointing left) and across canals (white arrow pointing across). There is a slight delay before the contraction is seen moving along the second canal (2). C. Plot of the track (diamond) of a single cell crawling in the mesohyl of the first canal during the first contraction. As the canal diameter narrows during a contraction (10 minutes, 600 s after addition of dye), the cell ceases forward movement (long arrow and bar). Crawling commences once the canal is fully contracted (10 minutes later). Forward motion is slowed a second time (small arrows) when the incurrent canal (1) begins to dilate once more (Modified from Elliott and Leys, 2007).

even though the events occur at a much slower rate than allowed by nerves and true muscle.

Histology and ontogeny of contractile and signaling tissues

Coordination of contractions in cellular sponges remains a difficult problem because the concept of the sponge as an animal (Eumetazoa) is controversial. If sponges have a cellular level of organization (Parazoa) as is traditionally thought, there are no tissues: thus along what structure do contractions propagate and how does the animal maintain integrity among the cells during contractions? Most modern texts suggest that during the evolution of basal metazoans there was a graded acquisition of structured

tissues (Gilbert 2003), culminating in the invention of mesoderm by primitive bilaterians. Cnidarians are usually regarded as diploblastic animals with only endoderm and ectoderm; however genes whose bilaterian homologs are implicated in mesodermal specification and differentiation are expressed during development of the anthozoan cnidarian Nematostella (Martindale et al. 2004) as well as in a tissue that gives rise to striated muscle in the hydrozoan jellyfish Podocoryne (Spring et al. 2002). It has therefore been suggested that mesoderm in triploblasts may have arisen from the endoderm of diploblastic animals, or, alternatively, that cnidarians arose from a triploblastic ancestor and that diploblasty is a secondary simplification (Martindale et al. 2004, Siepel and Schmid 2005).

56

Tissues
Under the hypothesis of sponge paraphyly (see above), a sponge-like animal is presumed to have given rise to early cnidarians. The leap is large. Sponges are filter feeders that lack nerves and muscle. Furthermore, sponges are not generally considered to have tissues, yet to some extent this last point may be a problem of definition. For example, transport strands in the tropical sponge Aplysina function as a distinct tissue, carrying phagocytosed material to the tip of the sponge presumably for growth (Leys and Reiswig 1998); regional concentrations of symbiont-containing cells that provide nutrition to sponges may do the same (e.g., Yahel et al. 2003). Tissues arise from germ layers during ontogeny and are mesenchymal or epithelial in nature. While current theory holds that epithelia preceded mesenchyme both during evolution and in development (Hay 1968, Prez-Pomares and Muoz-Chpuli 2002, Price and Patel 2004), extant sponges are considered to be largely mesenchymal animals. The idea that their covering layer is not a true epithelium comes from the absence of a well-defined basement membrane (Woollacott and Pinto 1995). Strictly speaking, epithelia are considered to be sheets of cells with apical-basal polarity, cell-cell junctions and an extracellular matrix ECM, cuticle or equivalent apically, and basement membrane basally that maintains cell polarity (Tyler 2003, Cereijido et al. 2004). Other authors however, consider epithelia as physical barriers between two different extracellular environments, with or without the presence of a basal lamina (Prez-Pomares and Muoz-Chpuli 2002). Athough homoscleromorph sponges are considered to have a true basement membrane, the condensation of ECM below choanocyte chambers is extremely slight (Boury-Esnault et al. 2003), and readily can be confused with the mucous coat on the apical surface of cells. An image showing similar condensation of ECM under the epithelium of the larva of the demosponge Crambe crambe (Maldonado 2004), coupled with evidence that other sponges have a type of collagen functionally equivalent to Type IV collagen (a typical constituent of basement membranes) (Aouacheria et al., 2006) suggests that a comprehensive morphological study of sponge epithelia with a focus on seeking structures equivalent to a basal lamina is well-warranted. One practical tool for studying sponge epithelia is simple labeling of the actin cytoskeleton in epithelia (Pavans de Ceccatty 1986). Using modern fluorescent labels for actin microfilaments (Bodipy-phallacidin fluorescein, Molecular Probes, OR) the surface epithelium of 6-day old juvenile freshwater sponges shows remarkably extensive tracts of actin (Fig. 3). Bundles run from cell to cell forming continuous paths from the choanosome to the edge of the animal, a distance of several hundred microns. In fact, continuous tracts can traverse over 1 mm through the apical pinacoderm of these sponges. Where cells adjoin one-another there are dense plaques of actin reminiscent of adhesion plaques in smooth muscle, as shown in freeze fracture and thin section by Pavans de Ceccatty (1986) (Fig. 3C, D). The apical pinacoderm is a tri-layered structure formed by two epithelial sheets that sandwich a very thin collagenous (and

cellular) mesohyl (Elliott and Leys 2007). Like the transport pathways in Aplysina (Leys and Reiswig 1998), the apical pinacoderm of the juvenile sponge is designed to function as an epithelium.

Gastrulation
If sponges can be considered to possess functional tissues, how do these tissues arise? The problem of the origin of tissues and germ layers in sponges has been a longstanding debate with growing disagreement as to whether gastrulation-like processes occur (Efremova 1997, Leys 2004, Ereskovsky and Dondua 2006, reviewed in Leys and Ereskovsky 2006). Most studies of sponge development are largely descriptive. Given that the vast majority of sponges are ovoviviparous (brood their young), development cannot be readily followed in vitro as it can in many other phyla. Furthermore, mechanisms of development (cleavage patterns and segregation of cells to form layers) are diverse (reviewed in Leys and Ereskovsky 2006), and the lack of uniformity has led to a great disparity in views on comparative development. Experimental data is now needed to test the hypothesis that sponges undergo gastrulation-like processes during ontogeny (Leys and Degnan 2002, Leys 2004, Leys and EerkesMedrano 2005). Homologs of genes known to be involved in the formation of mesoderm in other animals have been sequenced from a calcareous sponge (Manuel et al. 2004) and from demosponges (Adell et al. 2003, Bebenek et al. 2004, Hill et al. 2004), but expression patterns in early gastrula-like stages are only just being examined (Larroux et al. 2006). Of specific interest should be the expression pattern in calcareous sponges, a group in which embryogenesis involves distinctly epithelial movements and invagination of the larva to form a blastopore-like structure (Leys and Eerkes-Medrano 2005).

Speculations and considerations

It is valuable to remember that all is possible with sponges (Boury-Esnault 2006). These animals are specialized for suspension feeding on bacteria and ultraplankton (Pile et al. 1996, 1997, Ribes et al. 1999), but this poses certain problems: the filter may occasionally become clogged, and the flow bringing in food may not be sufficient for gas exchange in certain habitats. The idea that peristaltic-like contractions (condensation contractions) may have arisen to assist the filtering mechanism in sponges is not new (Weissenfels 1990). However, with increasing knowledge of the physiology and development of other basal metazoans, it is now interesting to note how similar the rhythmic contractions in sponges are to those in pennatulaceans such as Renilla (Anctil 1989, 1991). Apparently the absence of nerves and true muscle in sponges is no handicap given the power of a hydrostatic skeleton. Seen in this light, the sponge body plan with its aquiferous system could be considered to contain the underpinnings for the evolution of peristaltic contractile systems found in later animals.

57

Fig. 3: A, B. The actin cytoskeleton in the apical pinacoderm (exopinacoderm) of a 6-day old juvenile sponge (E. muelleri). In A, arrows indicate actin tracts labeled with Bodipy phallacidin-fluorescein, and arrowheads mark points of contact of individual cells. In B, arrows indicate dense plaques of actin at points of contact of cells. C. Freeze fracture preparation showing a desmosome-like region on the basal epithelial membrane in E. muelleri. D. Thin section (TEM) showing actin microfilaments (mf) and a dense plaque at the point of contact between two cells (C and D from Pavans de Ceccatty, 1986).

Acknowledgements
I thank G. Elliott and G. Tompkins-MacDonald for interesting discussions and G. Elliott and I. Ijieke for contributions to figures 1 and 2. The foundations for these ideas derive from the research of M. Pavans de Cecatty and G.O. Mackie. Funding was kindly provided by NSERC.

References
Adell T, Gamulin V, Perovic-Ottstadt S, Wiens M, Korzhev M, Mller IM, Mller WEG (2004) Evolution of metazoan cell junction proteins: the scaffold protein MAGI and the transmembrane receptor tetraspanin in the demosponge Suberites domuncula. J Mol Evol 59(1): 41-50 Adell T, Grebenjuk VA, Wiens M, Mller WEG (2003) Isolation and characterization of two T-box genes from sponges, the phylogenetically oldest metazoan taxon. Dev Genes Evol 213: 421-434 Anctil M (1989) Modulation of a rhythmic activity by serotonin via cyclic AMP in the coelenterate Renilla koellikeri. J Comp Physiol B 159: 491-500

Anctil M (1991) Modulation of rhythmic contractions by melatonin via cyclic GMP in the coelenterate Renilla koellikeri. J Comp Physiol B 161: 569-575 Anctil M, Poulain I, Pelletier C (2005) Nitric oxide modulates peristaltic muscle activity associated with fluid circulation in the sea pansy Renilla koellikeri. J Exp Biol 208: 2005-2017 Aouacheria, A, Geourjon, C, Aghajari, N, Navratil, V, Delage, G, Lethias, C Exposito, J-Y (2006) Insights into early extracellular matrix evolution: spongin short chain collagen-related proteins are homologous to basement membrane Type IV collagens and form a novel family widely distributed in invertebrates. Mol Biol Evol 23: 2288-2302 Bebenek I, Gates R, Morris J, Hartenstein V, Jacobs D (2004) sine oculis in basal Metazoa. Dev Genes Evol 214: 342-351 Borchiellini C, Manuel M, Alivon E, Boury-Esnault N, Vacelet J, Le Parco Y (2001) Sponge paraphyly and the origin of Metazoa. J Evol Biol 14 (1): 171-179 Boury-Esnault N (2006) Systematics and evolution of Demospongiae. Can J Zool 84: 205-224

58

Boury-Esnault N, Ereskovsky A, Bezac C, Tokina D (2003) Larval development in the Homoscleromorpha (Porifera, Demospongiae). Invertebr Biol 122 (3): 187-202 Boute N, Exposito JY, Boury-Esnault N, Vacelet J, Nor N, Miyazaki K, Yoshizato K, Garrone R (1996) Type IV collagen in sponges, the missing link in basement membrane ubiquity. Biol Cell 88: 3744 Cereijido M, Contreras RG, Shoshani L (2004) Cell adhesion, polarity, and epithelia in the dawn of metazoans. Physiol Rev 84 (4): 1229-1262 de Vos L (1971) Etude ultrastructurale de la gemmulogense chez Ephydatia fluviatilis. J Microsc 20: 15-20 de Vos L, van de Vyver G (1981) Etude de la contraction spontane chez lponge deau douce Ephydatia fluviatilis cultive in vitro. Ann Soc R Zool Belg 111: 21-31 Demidchik V, Adobea Essah P, Tester M (2004) Glutamate activates cation currents in the plasma membrane of Arabidopsis root cells. Planta 219: 167-175 Efremova SM (1997) Once more on the position among Metazoa Gastrulation and germinal layers of sponges. Berl geowiss Abhand 20: 7-15 Elliott GRD (2004) The contractile apparatus of the sneezing freshwater sponge Ephydatia muelleri. Newsletter Am Microscr Soc November: 5-7 Elliott GRD, Leys SP (2007) Coordinated contractions effectively expel water from the aquiferous system of a freshwater sponge. J Exp Biol (in press) Ellwanger K, Brmmer F, Nickel M (2004) Glutamate, GABA and serotonin induce contractions in the sponge Tethya wilhelma (Porifera: Demospongiae). In: Kinzelbach R (ed). Jahrestagung der Deutschen Zoologischen Gesellschaft, Abstractband, Rostock. pp. 157 Ellwanger K, Nickel M (2006). Neuroactive substances specifically modulate rhythmic body contractions in the nerveless metazoon Tethya wilhelma (Demospongiae, Porifera). Front Zool 3: 7 Emson RH (1966) The reactions of the sponge Cliona celata to applied stimuli. Comp Biochem Physiol 18: 805-827 Enkvist MOK, McCarthy KD (1994) Astroglial gap junction communication is increased by treatment with either glutamate or high K+ concentration. J Neurochem 62: 489-495 Ereskovsky AV, Dondua AK (2006) The problem of germ layers in sponges (Porifera) and some issues concerning early metazoan evolution. Zoologischer Anzeiger 245: 65-76 Francis JC, Poirrier MA (1986) Particle uptake in two fresh-water sponge species, Ephydatia fluviatilis and Spongilla alba (Porifera: Spongillidae). Trans Am Microsc Soc 105(1): 11-20 Gaino E, Pansini M, Pronzato R, Cicogna F (1991) Morphological and structural variations in Clathrina clathrus (Porifera, Calcispongiae). In: Reitner J, Keupp H (eds). Fossil and recent sponges. Springer-Verlag, Berlin. pp. 360-371 Gilbert R (2003) Developmental Biology. 8th Edition. Sinauer Associates Inc. Sunderland Gonobobleva EL, Ereskovsky AV (2004) Metamorphosis of the larva of Halisarca dujardini (Demospongiae, Halisarcida). Bull Inst R Sci Nat Belg 74: 101-115 Hay ED (1968) Organization and fine structure of epithelium and mesenchyme in the developing chick embryo. In: Fleischmajer R, Billingham R (eds). Epithelial-mesenchyme interactions. Williams and Wilkins, Baltimore. pp. 31-55

Hill A, Tetrault J, Hill M (2004) Isolation and expression analysis of a poriferan Antp-class Bar-/Bsh-like homeobox gene. Dev Genes Evol 214: 515-523 Hille B (2002) Ionic channels of excitable membranes. Sinauer, Sunderland Holland ND, Venkatesh TV, Holland LZ, Jacobs DK, Bodmer R (2003) AmphiNK2-tin, an amphioxus homeobox gene expressed in myocardial progenitors: insights into evolution of the vertebrate heart. Dev Biol 255: 128-137 Jones WC (1962) Is there a nervous system in sponges? Biol Rev Camb Philos Soc 37: 1-50 Larroux C, Fahey B, Liubicich D, Hinman V, Gauthier M, Gongora M, Green CR, Woerheide G, Leys SP, Degnan BM (2006) Developmental expression of transcription factor genes in a demosponge: insights into the origin of metazoan multicellularity. Evol Dev 8(2): 150-173 Lawn ID (1982) Porifera. In: Shelton GAB (ed). Electrical conduction and behaviour in simple invertebrates. Clarendon Press, Oxford. pp. 49-72 Lentz TL (1966) Histochemical localization of neurohumors in a sponge. J Exp Zool 162: 171-180 Leys SP (2004) Gastrulation in sponges. In: Stern C (ed). Gastrulation: from cells to embryo. Cold Spring Harbor Press. pp. 23-31 Leys SP, Degnan BM (2002) Embryogenesis and metamorphosis in a haplosclerid demosponge: gastrulation and transdifferentiation of larval ciliated cells to choanocytes. Invertebr Biol 121 (3): 171189 Leys SP, Eerkes-Medrano DI (2005) Gastrulation in calcareous sponges: in search of Haeckels gastraea. Integr Comp Biol 45(2): 342-351 Leys SP, Ereskovsky A (2006) Embryogenesis and larval differentiation in sponges. Can J Zool 84: 262-287 Leys SP, Meech RW (2006) Physiology of coordination in sponges. Can J Zool 84: 288-306 Leys SP, Mackie GO (1997) Electrical recording from a glass sponge. Nature 387: 29-30 Leys SP, Reiswig HM (1998) Nutrient transport pathways in the neotropical sponge Aplysina. Biol Bull 195: 30-42 Mackie GO (1979) Is there a conduction system in sponges? In: Lvi C, Boury-Esnault N (eds). Biologie des spongiaires. Colloques Internationaux du CNRS, vol. 291. ditions du CNRS, Paris. pp 145-152 Mackie GO (1984) Introduction to the diploblastic level. In: Bereiter-Hahn J, Matoltsy JAG, Richards KS (eds). Biology of the integument. Springer-Verlag, Berlin. pp. 43-46 Mackie GO (1990) The elementary nervous system revisited. Amer Zool 30: 907-920 Maldonado M (2004) Choanoflagellates, choanocytes, and animal multicellularity. Invertebr Biol 123: 1-22 Manuel M, Le Parco Y, Borchiellini C (2004) Comparative analysis of Brachyury T-domains, with the characterization of two new sequences, from a hexactinellid and a calcisponge. Gene 340: 291301 Martindale MQ, Pang K, Finnerty JR (2004) Investigating the origins of triploblasty: mesodermal gene expression in a diploblastic animal, the sea anemone Nematostella vectensis (phylum, Cnidaria; class, Anthozoa). Development 131: 2463-2474 McNair GT (1923) Motor reactions of the fresh-water sponge Ephydatia fluviatilis. Biol Bull 44: 153-166

59

Mechawar N, Anctil M (1997) Melatonin in a primitive metazoan: seasonal changes in levels and immunohistochemical visualization in neurons. J Comp Neurol 387: 243-254 Medina M, Collins AG, Silberman J, Sogin ML (2001) Evaluating hypotheses of basal animal phylogeny using complete sequences of large and small subunit rRNA. Proc Natl Acad Sci USA 98: 9707-9712 Moroz LL, Meech RW, Sweedler JV, Mackie GO (2004) Nitric oxide regulates swimming in the jellyfish Aglantha digitale. J Comp Neurol 471: 26-36 Nedergaard M (1994) Direct signaling from astrocytes to neurons in cultures of mammalian brain cells. Science 263: 1768-1771 Nedergaard M, Takano T, Hansen AJ (2002) Beyond the role of glutamate as a neurotransmitter. Nat Rev Neurosci 3: 748-755 Nickel M (2004) Kinetics and rhythm of body contractions in the sponge Tethya wilhelma (Porifera: Demospongiae). J Exp Biol 207: 4515-4524 Osipchuk Y, Cahalan M (1992) Cell-to-cell spread of calcium signals mediated by ATP receptors in mast cells. Nature 359: 241-246 Parker GH (1910) The reactions of sponges with a consideration of the origin of the nervous system. J Exp Zool 8: 765-805 Pavans de Ceccatty M (1986) Cytoskeletal organization and tissue patterns of epithelia in the sponge Ephydatia mulleri. J Morphol 189: 45-65 Pavans de Ceccatty M (1989) Les ponges, laube des communications cellulaires. Science 142: 64-72 Prez-Pomares JM, Muoz-Chpuli R (2002) Epithelialmesenchymal transitions: a mesodermal cell strategy for evolutive innovation in metazoans. Anat Rec 268: 343-351 Perovic S, Krasko A, Prokic I, Mller IM, Mller WEG (1999) Origin of neuronal-like receptors in Metazoa: cloning of a metabotropic glutamate/GABA-like receptor from the marine sponge Geodia cydonium. Cell Tissue Res 296: 395-404 Pile AJ, Patterson MR, Savarese M, Chernykh VI, Fialkov VA (1997) Trophic effects of sponge feeding with lake Baikals littoral zone. 2. Sponge abundance, diet, feeding efficiency, and carbon flux. Limnol Oceanogr 42: 178-184 Pile AJ, Patterson MR, Witman JD (1996) In situ grazing on plankton <10 m by the boreal sponge Mycale lingua. Mar Ecol Prog Ser 141: 95-102

Price AL, Patel NH (2004) The evolution of gastrulation: cellular and molecular aspects. In: Stern C (ed). Gastrulation. From cells to embryos. Cold Spring Harbor Press, Cold Spring Harbor. pp. 695-701 Prosser CL (1967) Ionic analysis and effects of ions on contractions of sponge tissues. Z vergl Physiol 54: 109-120 Prosser CL, Nagai T, Nystrom RA (1962) Oscular contractions in sponges. Comp Biochem Physiol 6: 69-74 Reiswig HM (1970) Porifera: sudden sperm release by tropical demosponge. Science 170: 538-539 Ribes M, Coma R, Gili J-M (1999) Natural diet and grazing rate of the temperate sponge Dysidea avara (Demospongiae, Dendroceratida) throughout an annual cycle. Mar Ecol Progr Ser 176: 179-190 Siepel K, Schmid V (2005) Evolution of striated muscle: Jellyfish and the origin of triploblasty. Dev Biol 282: 14-26 Simpson TL (1984) The cell biology of sponges. Springer Verlag, New York Spring J, Yanze N, Josch C, Middel AM, Winninger B, Schmid V (2002) Conservation of Brachyury, Mef2, and Snail in the myogenic lineage of jellfish: a connection to the mesoderm of Bilateria. Dev Biol 244: 372-384 Tyler S (2003) Epithelium - The primary building block for metazoan complexity. Integr Comp Biol 43 (1): 55-63 Weissenfels N (1990) Condensation rhythm of fresh-water sponges (Spongillidae, Porifera). Eur J Cell Biol 53: 373-383 Weyrer S, Rtzler K, Rieger R (1999) Serotonin in Porifera? Evidence from developing Tedania ignis, the caribbean fire sponge (Demospongiae). Memoir Queensl Mus 44: 659-665 Woollacott RM, Pinto RL (1995) Flagellar basal apparatus and its utility in phylogenetic analyses of the porifera. J Morphol 226: 247-265 Yahel G, Sharp JH, Marie D, Hase C, Genin A (2003) In situ feeding and element removal in the symbiont-bearing sponge Theonella swinhoei: bulk DOC is the major source for carbon. Limnol Oceanogr 48: 141-149 Yang WQ, Braun C, Plattner H, Purvee J, Van Houten JL (1997) Cyclic nucleotides in glutamate chemosensory signal transduction of Paramecium. J Cell Sci 110: 2567-2572

Porifera research: Biodiversity, innovation and sustainaBility - 2007

61

Gemmules as a key structure for the adaptive radiation of freshwater sponges: a morphofunctional and biogeographical study
Renata Manconi(1), Roberto Pronzato(2)
Dipartimento di Zoologia e Genetica Evoluzionistica. Universit di Sassari. Via Muroni 25, I-07100 Sassari, Italy. r.manconi@uniss.it (2) Dipartimento per lo Studio del Territorio e delle sue Risorse. Universit di Genova. Corso Europa 26, I-16132 Genova, Italy. pronzato@dipteris.unige.it
(1)

Abstract: This paper concerns an excursus on morphology, diversity and biogeography of freshwater sponges to test if gemmules could be considered a strategic successful device in the natural history of Spongillina. Taxa richness of the six families belonging to the suborder Spongillina (Demospongiae, Haplosclerida) is notably high if compared to the other sessile filter feeder benthic taxa living in freshwater such as Cnidaria and Bryozoa. Although extremely conservative for some characters the plastic bauplan of Porifera favoured the adaptive radiation of Spongillina in inland waters worldwide and produced both structural and functional evolutionary novelties mainly at the level of resting bodies (gemmules). Clonation, modular architecture and cryptobiosis represent successful adaptive strategies to support survival and dispersal of freshwater sponges at all latitudes from the Arctic Circle to Patagonia in a wide variety of extremely discontinuous freshwater habitats. A comparative analysis on morpho-functional diversity of asexual resting bodies based on literature data vs. new investigations by scanning electron microscopy highlighted some trends in the evolution of the genera belonging to Metaniidae, Potamolepidae, and Spongillidae that display typical gemmular architectures. Other endemic sponges from ancient lakes, namely Lubomirskiidae, Malawispongiidae and Metschnikowiidae, are exclusively sexual and share this reproductive strategy by swimming larvae (parenchymellas) with the gemmule-producing families. The evolutionary success of Spongillina is not easy to be interpreted. In some cases a dispersalist model explains better the natural history of the taxon mainly for species till now considered cosmopolitan as in the case of Ephydatia fluviatilis. In other cases, e.g. the genus Corvospongilla, the high level of speciosity and endemicity seems to match the vicariance model. The distributive pattern of genera/families from ancient lakes represents a peculiar case; these taxa share both morphological (spiculation, skeletal architecture, absence of gemmules) and biological (perennial life cycle, absence of cryptobiosis) traits in spite of their extremely disjunct geographic distribution and high levels of endemism. Keywords: Biodiversity, Porifera, taxonomic richness, morpho-functional traits, cryptobiosis, distribution, ecology, discontinuous habitats

Introduction
The wide adaptive radiation of sponges has resulted in the colonization of continental waters at all latitudes from cold deserts to equatorial rainforests and hot deserts; from the coast line to high plains and mountains and subterranean environments. An extremely wide variety of lotic and lentic habitats have been colonised ranging from springs, water courses, and brackish waters in estuaries and enclosed seas, to thermal vents, caldera lakes, alpine lakes, and caves. Ephemeral pools, marshes, swamps, billabongs, oueds and pans, are also suitable habitats together with oceanic islands and man-made basins, such as pools in gardens and archaeological sites, reservoirs, water tanks and pipelines (see Manconi and Pronzato 2002, Pronzato and Manconi 2002). Recorded bathymetric distribution ranges from hundreds of meters in some lakes (Crane 1991, de Ronde et al. 2002)

to the surface exposed to sunlight during low-water levels (Manconi and Pronzato 1994, 2002). Freshwater sponges are able to survive extreme environmental conditions ranging from ice, hot waters, dry-up, anoxy, eutrophy, high levels of chemicals, hydrocarbons and heavy metals (Old 1932, Jewell 1935, 1939, Sar and Vacelet 1973, Harrison 1974, 1977, Rader 1984, Rader and Winget 1985, van Soest and Velikonja 1986, Willenz et al. 1986, Francis and Harrison 1988, Ricciardi and Reiswig 1993, Richelle-Maurer et al. 1994a, 1994b, Vacelet 1994, de Ronde et al. 2002, Rota and Manconi 2004). Body shape of sponges from inland waters range from thin whitish crusts to dark brown massive cushion, brilliant green branching or erect growth forms (Manconi and Pronzato 1991). In some seasons, according to the climate, most freshwater sponges are represented on the substratum only by small spherules or resting stages, known as gemmules

62

(Simpson and Fell 1974, Manconi and Pronzato 1991, 1994, 2002, Corriero et al. 1993, Pronzato et al. 1993, Pronzato and Manconi 1994, 1995, 2002). Sponges represent a natural resource for their functional role in auto-depurative processes of water bodies (Manconi and Desqueyroux-Faundez 1999, Manconi et al. 1999) playing a key-role in the re-cycling of organic matter. Their contribution to the energetic equilibrium of freshwater ecosystems by pumping rate is conspicuous and a fingersized Spongilla lacustris can filter more than 125 litres per day (Frost 1978, 1980, 1991). Sponges are also centres of biological association, representing a suitable but selective refuge microhabitat. They host a notably diverse assemblage of organisms ranging from animals and protists to bacteria and algae related by endocellular symbiosis, inquinilism, commensalisms, and highly specialized predation as in the case of Insecta belonging to the orders Neuroptera, Tricoptera, and Diptera. Most freshwater invertebrate taxa have been recorded in sponges, namely Hydrozoa, Nematoda, Oligochaeta, Polychaeta, Gastropoda, Bivalvia, Isopoda, Amphipoda, Ostracoda, Hydracarina, and Bryozoa, to several families of insects encompassing the typical and exclusive spongillaflies Sisyridae (Berg 1948, Brown 1952, Brnsted and Brnsted 1953, Brnsted and Loevtrup 1953, Parfin and Gurney 1956, Volkmer-Ribeiro and de Rosa-Barbosa 1974, Resh 1976, Resh et al. 1976, Steffen 1967, Frost and Williamson 1980, Kahl and Konopacha 1981, Konopacha and Socinski 1985, Kamaltynov et al. 1993, Weissmair and Mildner 1995, Gugel 1996, Oliveira Roque et al. 2004, Rota and Manconi 2004, Weinberg et al. 2004). Sponges host fishes and amphibians also to nest fertilized eggs (Kilian and Campos 1969, Manconi and Pronzato 2002), represent one of the food items for fishes, freshwater turtles and ducks (Dominey and Snyder 1988, McCauley and Longcore 1988, Seigel and Brauman 1994, Kennett and Tory 1996, Armstrong and Booth 2005), and are reported as a refuge for millipedes during inundation of Amazonian floodplains (Adis 1992). Freshwater sponges have been used since ancient times also by humans, and it is known that some African and Amazonian populations produce ceramics strengthened by sponge spicules (Linn 1925, Serrano 1933, Brissaud and Houdayer 1986, Adamson et al. 1987, McIntosh and MacDonald 1989). Other practical uses are in the field of cosmesis as in the case of dried spongillids used in the 19th century by Russian young ladies to scrub their faces to have rosy cheeks (Kuznetzow 1898), and at present some cosmetics are based on the action of siliceous spicule powder. In the 17th century Samuel Hahnemann enclosed freshwater sponges in his Materia Medica as a homeopathic remedy for psoriasis with the common pre-linnean Russian name of Badiaga (Allen 1986) although a notable confusion exists on the identification at the genus and species level of the used material, reported as Spongilla fluviatilis and Spongia palustris. Freshwater sponges are also useful indicators of palaeoenvironmental changes by the analysis of spicular remains in sediments (Harrison et al. 1979, Harrison and Warner 1986, Harrison 1988, Volkmer-Ribeiro and Turcq 1996, Candido et al. 2000).

No sponges are currently listed on the IUCN Red List, although official threatened species lists of few countries report on freshwater sponges (e.g. Brazil, Norway); in some cases they are indirectly protected being sympatric with umbrella species such as fishes and amphibians. Manconi and Desqueyroux-Faundez (1999) suggest that conservation of freshwater sponge fauna would represent a sustainable approach to maintain biodiversity and to improve the rational management of freshwater natural resources. It is proved also that freshwater sponges could control the presence of the bivalve Dreissena polymorpha (Ricciardi et al. 1995, Swierczynski 1996, Lauer and Spacie 2004).

Recent freshwater sponges: gemmular structure and biogeography

Recent freshwater sponges are ascribed to 45 genera in 6 families, namely Lubomirskiidae, Malawispongiidae, Metaniidae, Metschnikowiidae, Potamolepidae, and Spongillidae. Species richness, 217 species (Table 1), is high if compared to that of the other freshwater sessile invertebrates belonging to Cnidaria and Bryozoa (Manconi and Pronzato, in press a). Freshwater sponges occur worldwide, except for the Antarctic region and the North Pole. Geographic distribution is related of course to the geological and climatic history of the continents and to the long-term dynamics of hydrographic basins. The highest taxonomic diversity at the biogeographic scale is recorded from the Neotropical (63 species), Palaearctic (59 species), and Afrotropical (49 species) regions. Species richness is lower in the other zoogeographic regions: Oriental (37 species), Australasian (33 species), Nearctic (32 species), and Pacific Oceanic Islands (5 species) (Fig. 1; Table 1) (Manconi and Pronzato, in press a). The oldest fossils of sponges from inland water are very few and dated back to the Mesozoic Era (Ott and Volkheimer 1972, Dunagan 1999, Richter and Wuttke 1999). Eospongilla morrisonensis Dunagan, 1999 is known from the Colorado Upper Jurassic, while Palaeospongilla chubutensis Ott and Volkheimer, 1972 was recorded together with Spongilla patagonica Volkmer-Ribeiro and Reitner, 1991 from the Patagonian Lower Cretaceous in the Chubut Valley. More recent taxa are known from the European Eocene as in the case of Lutetiospongilla heily Richter and Wuttke, 1999. The presence of resistant bodies sharing most traits with recent Spongillidae in one of the best preserved fossil, Palaeospongilla chubutensis, indicates that gemmules have been extremely conservative structures since the Cretaceous (Manconi and Pronzato 2002). The process of freshwater colonization seems to be strictly related to the cryptobiosis phenomenon and to the evolutionary novelty represented by resistant bodies, the gemmules. Different approaches exist to the problem of inland water colonization and phylogeny of freshwater sponges (Brien 1966, 1969, Volkmer-Ribeiro 1979, 1986, 1990, Volkmer-Ribeiro and de Rosa Barbosa 1979, VolkmerRibeiro and Watanabe 1983, Pronzato and Manconi 2002). Our opinion is that only Haplosclerida, belonging to the suborder Spongillina, colonised inland waters. At present, we do not know, however, how many colonization processes

63

Table 1: Worldwide checklist of freshwater sponges (no fossil species). Zoogeographic Regions are reported as Nearctic (NA), Palaearctic (PA), Neotropical (NT), Afrotropical (AT), Oriental (OL), Australasian (AU), Pacific Oceanic Islands (PAC). Suborder SPONGILLINA Manconi and Pronzato, 2002 PA-NA-AT-OL-AU-NT-PAC (6 families, 45 genera, 217 species) Family Spongillidae Gray, 1867 PA-NA-AT-OL-AU-NT-PAC (21 genera, 146 species) Anheteromeyenia Schrder, 1927 (3 species) NA-NT A. argyrosperma (Potts, 1880) (Canada to Florida) type species NA A. ornata (Bonetto and Ezcurra de Drago, 1970) (Brazil, Argentina) NT A. cheguevarai Manconi and Pronzato, 2005 (Cuba) NT Corvoheteromeyenia Ezcurra de Drago, 1979 (2 species) NT C. australis (Bonetto and Ezcurra de Drago, 1966) (Neotropical) type species NT C. heterosclera (Bonetto and Ezcurra de Drago, 1966) (Argentina, Brazil, Venezuela, Curacao) NT Corvospongilla Annandale, 1911 (18 species) PA-NA-NT-AT-OL C. loricata (Weltner, 1895) (lower Nile basin) type species PA C. bhavnagarensis Soota, Pattanayak and Safena, 1984 (Gujarat, India) OL C. becki Poirrier, 1978 (Louisiana) NA C. boehmii (Hilgendorf, 1883) (Equatorial Africa, Brazil) AT-NT C. burmanica (Kirkpatrick, 1908) (Burma, India) OL C. caunteri Annandale, 1911 (India) OL C. lapidosa (Annandale, 1908) (India) OL C. mesopotamica Manconi and Pronzato, 2004 (Iraq) PA C. micramphidiscoides Weltner, 1913 (R. Congo) AT C. novaeterrae (Potts, 1886) (NE Nearctic) NA C. scabrispiculis Annandale, 1913 (Lower R. Nile) PA C. seckti Bonetto and Ezcurra de Drago, 1966 (Paran, R. Paraguay, R. Uruguay, Argentina, Brazil) NT C. sodenia Brien, 1969 (Cameroon) AT C. thysi Brien, 1968 (L. Kumba, Cameroon) AT C. ultima (Annandale, 1910) (S-India) OL C. zambesiana (Kirkpatrick, 1906) (R. Zambezi) AT C. victoriae Annandale, 1914 (Zambesi, Zambia) AT C. volkmeri (de Rosa-Barbosa, 1988) (Brazil) NT Dosilia Gray, 1867 (5 species) OL-AT-NA-NT D. plumosa (Carter, 1849) (India, Philippines) type species OL D. brouni (Kirkpatrick, 1906) (R. Nile, L. Baringo, E- Africa) AT D. palmeri (Potts, 1885) (Louisiana, Florida, Texas, Mexico, Central America) NA-NT D. pydanieli Volkmer-Ribeiro, 1992 (Brazil) NT D. radiospiculata (Mills, 1888) (Louisiana, Mexico) NA Duosclera Reiswig and Ricciardi, 1993 (1 species) NA D. mackayi (Carter, 1885) (NE-America) monotypic NA Ephydatia Lamouroux, 1816 (10 species) PA-NA-AT-OL-AU-NT-PAC E. fluviatilis (Linnaeus, 1759) (cosmopolitan) type species PA-NA-AT-OL-AU E. meyeni (Carter, 1849) (India, China) OL E. millsii (Potts, 1887) (Florida) NA E. muelleri (Lieberkuhn, 1855) (North Hemisphere) PA-NA E. facunda Weltner, 1895 (Central and South America) NT E. fortis Weltner, 1895 (Indonesia, Philippines, Japan, Vanuatu) OL-PA-PAC E. japonica (Hilgendorf, 1882) (USA, Manchuria, Japan) NA-PA E. ramsayi (Haswell, 1882) (Australia, New Zealand, New Guinea?) AU E. robusta (Potts, 1887) (E-USA, Mexico, California) NA E. syriaca Topsent, 1910 (Syria) PA Eunapius Gray, 1867 (16 species) PA-AT-OL-AU-NA-NT E. carteri (Bowerbank, 1863) (S-Asia, East Europe, Africa) type species OL-PA-AT E. aetheriae (Annandale, 1913) (Lower R. Nile) PA E. ambiguus (Annandale, 1909) (KwaZulu-Natal) AT E. calcuttanus (Annandale, 1911) (India) OL E. conifer (Annandale, 1916) (China, Korea) PA E. crassissimus (Annandale, 1907) (India, Australia, tropical SE Asia) OL-AU E. fragilis (Leidy, 1851) (Cosmopolitan) PA-NA-AT-NT-OL-AU E. geminus (Annandale, 1911) (India) OL E. geei (Annandale, 1918) (China) PA E. michaelseni (Annandale, 1914) (Central Africa) AT E. nitens (Carter, 1881) (Africa) AT E. potamolepis (Annandale, 1918) (Thailand) OL E. ryuensis (Sasaki, 1970) (Japan, Korea) PA E. sinensis (Annandale, 1910) (China, Manchuria, Korea, Russia, Australia) PA-AU E. subterraneus Sket and Velikonja, 1984 (Croatia) PA E. tinei (Gee, 1932) (Philippines) OL

64 Table 1 (cont.)

Heteromeyenia Potts, 1881 (7 species) NA-PA-NT-AU H. baileyi (Bowerbank, 1863) (North and Central America, Argentina, Europe) type species NA-PA-NT H. horsti Ezcurra de Drago, 1988 (Argentina) NT H. insignis Weltner, 1895 (Brazil) NT H. latitenta (Potts, 1881) (NE-USA, Mexico) NA H. stepanowii (Dybowsky, 1884) (Europe, Russia, China, Korea, Japan, Australia, Brazil, Argentina) PA-AU-NT H. tentasperma (Potts, 1880) (NE-USA) NA H. tubisperma (Potts, 1881) (NE-America) NA Heterorotula Penney and Racek, 1968 (7 species) AU-NT H. capewelli (Bowerbank, 1863) (Australia) type species AU H. contraversa (Racek, 1969) (E-Central Australia) AU H. fistula Volkmer and Motta 1995 (South America) NT H. kakauensis (Traxler, 1896) (New Zealand) AU H. multidentata (Weltner, 1895) (E-Australia, Tasmania) AU H. multiformis (Weltner, 1910) (W-Australia) AU H. nigra (von Lendenfeld, 1887) (E-Australia) AU Nudospongilla Annandale, 1918 (6 species) PA-AT-AU N. coggini (Annandale, 1910) (Yunnan Lakes, W-China ) type species PA N. cunningtoni (Kirkpatrick, 1906) (L. Tanganyika, Africa) AT N. ehraiensis Lizhen, 1998 (Yunnan, W-China) PA N. moorei (Evans, 1899) (L. Tanganyika) AT N. vasta (Weltner, 1901) (Sulawesi, Indonesia) AU N. yunnanensis (Annandale, 1910) (Yunnan, W-China) PA Pachyrotula Volkmer-Ribeiro and Rtzler, 1997 (1 species) PAC P. raceki (Rtzler, 1968) (New Caledonia) monotypic PAC Pectispongilla Annandale, 1909 (4 species) OL-AU-PA P. aurea Annandale, 1909 (India) type species OL P. botryoides Haswell, 1882 (Australia) AU P. stellifera Annandale, 1915 (S-India) OL P. subspinosa Annandale, 1911 (India, Japan, Korea) OL-PA Racekiela Bass and Volkmer-Ribeiro, 1998 (4 species) PA-NA-NT R. ryderi (Potts, 1882) (Amphiatlantic in the N-Hemisphere, Belize) type species PA-NA-NT R. biceps (Lindenschmidt, 1950) (Michigan, USA) NA R. pictovensis (Potts, 1885) (from E-Canada to New York) NA R. sheilae (Volkmer-Ribeiro, de Rosa Barbosa and Tavares, 1988) (South America) NT Radiospongilla Penney and Racek, 1968 (16 species) AU-PAC-NT-PA-AT-OL-NA R. sceptroides (Haswell, 1882) (E-Australia, New Zealand, New Guinea, New Caledonia) type species AU-PAC R. amazonensis Volkmer-Ribeiro and Maciel, 1983 (Brazil) NT R. cantonensis (Gee, 1929) (China) PA R. cerebellata (Bowerbank, 1863) (trop. Africa, Ind.-Pak., Indon., Philipp., Japan, N. Guinea, China, Russia?, SE-Eur.) AU-AT-OL-PA R. cinerea (Carter, 1849) (Bombay, Himalayas, Pakistan) OL R. crateriformis (Potts, 1882) (USA, Can., Mexico, W. Indies, Suriname, Brazil, China, Japan, S-Asia, Australia) NA-NT-PA-AU-OL R. hemephydatia (Annandale, 1909) (India, New Guinea, E-Australia) OL-AU R. hispidula (Racek, 1969) (Australia) AU R. hozawai (Sasaki, 1936) (Japan) PA R. indica (Annandale, 1907) (India, Indonesia, Philippines, New Guinea?) OL-AU R. multispinifera (Gee, 1933) (E-Australia) AU R. philippinensis (Annandale, 1909) (Philippines to N-Australia) OL-AU R. sansibarica (Weltner, 1895) (Zanzibar, L. Upemba, L. Sonfon, Sierra Leone, Zambia, Congo basin) AT R. sendai (Sasaki, 1936) (Japan, Korea) PA R. sinoica (Racek, 1969) (Australia) AU R. streptasteriformis Stanisic, 1978 (N. Territory, Australia) AU Sanidastra Volkmer-Ribeiro and Watanabe, 1983 (1 species) PA S. yokotonensis Volkmer-Ribeiro and Watanabe, 1983 (Japan, Sardinia and Corsica) monotypic PA Saturnospongilla Volkmer-Ribeiro, 1976 (1 species) NT S. carvalhoi Volkmer-Ribeiro, 1976 (R. Juru, Brazil, Venezuela) monotypic NT Spongilla Lamarck, 1816 (15 species) PA-NA-OL-AT-AU-NT S. lacustris (Linnaeus, 1759) (Palaearctic, Nearctic) type species PA-NA S. alba Carter, 1849 (Ind., SE-Asia, Turk., Afr., Madag., Australia, Papua-N. Guin., USA, S. Salvador, S. Am.) OL-AT-AU-NT-PA-NA S. aspinosa Potts, 1880 (E-Canada, USA, China) NA-PA S. cenota Penney and Racek, 1968 (Yucatan, Costa Rica, Florida) NT-NA S. chaohuensis Cheng, 1991 (Hebei, N-China) PA S. heterosclerifera Smith, 1918 (N-America) NA S. inarmata Annandale, 1918 (Japan) PA S. jiujiangensis Cheng, 1991 (Hebei, N-China) PA S. mucronata Topsent, 1932 ( Dienn-Timbuct, R. Niger) AT

65 Table 1 (cont.)

S. permixta Weltner, 1885 (Bibisande, central Africa) AT S. prespensis Hadzische, 1953 (L. Prespa, Macedonia) PA S. shikaribensis Sasaki, 1934 (Japan) PA S. spoliata Volkmer-Ribeiro and Maciel, 1983 (Venezuela, Brazil) NT S. stankovici Arndt, 1938 (L. Ohrid, Macedonia) PA S. wagneri Potts, 1889 (SE-USA, Florida, South Carolina) NA Stratospongilla Annandale, 1909 (9 species) OL-AT-PA-AU-NA S. bombayensis (Carter, 1882) (Mumbay, Natal) type species OL-AT S. africana Annandale, 1914 (Victoria Falls, Zambesi) AT S. akanensis (Sasaki, 1934) (Japan) PA S. clementis (Annandale, 1909) (Philippines, China, Japan, tropical W-Africa) OL-PA-AT S. gravelyi (Annandale, 1912) (Mumbay, India) OL S. indica (Annandale, 1908) (Thailand, India, Africa) OL-AT S. lanei Racek, 1969 (Australia) AU S. penney (Harrison, 1979) (Florida) NA S. sumatrana (Weber, 1890) (Indonesia, India, Africa) OL-AT Trochospongilla Vejdowsky, 1883 (14 species) PA-NA-NT-OL-AU-AT T. horrida (Weltner, 1893) (Holarctic) type species PA-NA T. delicata Bonetto and de Drago, 1967 (Argentina, Brazil) NT T. gregaria Bowerbank, 1863 (Venezuela) NT T. lanzamirandai Bonetto and Ezcurra de Drago, 1964 (Brazil) NT T. latouchiana Annandale, 1907 (India, Burma, China, SE-Australia, SW-Africa) OL-PA-AU-AT T. leidii (Bowerbank, 1863) (Florida, Panama) NA-NT T. minuta (Potts, 1881) (Argentina, Bolivia, Venezuela, E-Brazil) NT T. paulula (Bowerbank, 1863) (R. Amazon, Venezuela, Suriname, Argentina) NT T. pennsylvanica (Potts, 1882) (North America) NA T. petrophila Racek, 1969 (E-Australia) AU T. philottiana Annandale, 1907 (India, S-China, Philippines, Africa) OL-AT T. tanganyikae Evans, 1899 (L. Tanganyika) AT T. singpuensis Chen, 1991 ((Hebei, N-China) PA T. variabilis Bonetto and Ezcurra de Drago, 1973 (Argentina, Brazil) NT Umborotula Penney and Racek, 1968 (1 species) PA-OL-AU U. bogorensis (Weber, 1890) (N-China, Korea, SE-Asia, Andaman Islands, E-Australia) monotypic PA-OL-AU Uruguayella Bonetto and Ezcurra de Drago, 1969 (5 species) NT U. repens (Hinde, 1888) (R. Uruguay, upper R. Paran, Argentina) type species NT U. amazonica (Weltner, 1895) (R. Amazon) NT U. macandrewi (Hinde, 1888) (R. Paraguay, R. Paran) NT U. pygmea (Hinde, 1888) (R. Paraguay, R. Uruguay) NT U. ringueleti (Bonetto and Ezcurra de Drago, 1969) (upper R. Paran, R. Uruguay) NT Family lubomirSkiidae Rezvoi, 1936 (4 genera, 10 species) PA Baikalospongia Annandale, 1914 (4 species) PA B. bacillifera Dybowsky, 1880 (L. Baikal) type species PA B. intermedia Dybowsky, 1880 (L. Baikal) PA B. dzhegatajensis Rezvoi, 1927 (L. Djegataj Kul, Urianhajskaja Region) PA B. erecta Efremova, 2004 (L. Baikal) PA Lubomirskia Dybowsky, 1880 (4 species) PA L. baikalensis (Pallas, 1776) (L. Baikal) type species PA L. abietina Swartschewsky 1901 (L. Baikal) PA L. fusifera Soukatschoff, 1895 (L. Baikal) PA L. incrustans Efremova, 2004 (L. Baikal) PA Rezinkovia Efremova, 2004 (1 species) PA R. echinata Efremova, 2004 (L. Baikal) monotypic PA Swartschewskia Makushok, 1927 (1 species) PA S. papiracea (Dybowsky, 1880) (L. Baikal) monotypic PA Family malawiSpongiidae Manconi and Pronzato, 2002 (5 genera, 6 species) PA-AT-AU Cortispongilla (Annandale, 1918 (1 species) PA C. barroisi (Topsent, 1892) (L. Tiberiade, R. Jordan) monotypic PA Malawispongia Brien, 1972 (1 species) AT M. echinoides Brien, 1972 (L. Malawi) monotypic AT Ochridaspongia Arndt, 1937 (2 species) PA O. rotunda Arndt, 1937 (L. Ohrid, Macedonia) type species PA O. interlithonis Gilbert and Hadzische, 1984 (L. Ohrid) PA

66 Table 1 (cont.)

Pachydictyum Weltner, 1901 (1 species) AU P. globosum Weltner, 1901 (L. Posso, Sulawesi) monotypic AU Spinospongilla Brien, 1974 (1 species) AT S. polli Brien, 1974 (L. Tanganyika) monotypic AT Family metaniidae Volkmer-Ribeiro, 1986 (5 genera, 22 species) NA-NT-AT-OL-AU Acalle Gray, 1867 (1 species) NT A. recurvata (Bowerbank, 1863) (Brazil) monotypic NT Corvomeyenia Weltner, 1913 (4 species) NA-NT C. everetti (Mills, 1884) (NE-USA, S-Canada) type species NA C. carolinensis Harrison 1971 (South Carolina, NE-USA) NA C. epilithosa Volkmer-Ribeiro, de Rosa Barbosa and Machado, 2005 (Brazil) NT C. thumi (Traxler, 1895) (Brazil) NT Drulia Gray, 1867 (5 species) NT D. browni (Bowerbank, 1863) (R. Amazon, Rio Negro in Brazil, Rio Beni in Bolivia, Venezuela) Type species NT D. conifera Bonetto and Ezcurra de Drago, 1973 (Rio Orinoco, Venezuela) NT D. cristata (Weltner, 1895) (R. Amazon, Tapajos) NT D. ctenosclera Volkmer-Ribeiro and Mothes de Moraes, 1981 (Rio Negro, Amazonia) NT D. uruguayensis Bonetto and Ezcurra de Drago, 1969 (R. Uruguay, R. Paran, Argentina, Suriname) NT Houssayella Bonetto and Ezcurra de Drago, 1966 (1 species) NT H. iguazuensis Bonetto and Ezcurra de Drago, 1966 (R. Iguaz) monotypic NT Metania Gray, 1867 (11 species) NT-AT-AU-OL M. reticulata (Bowerbank, 1863) (Brazilian and Venezuelan Amazonian basin) type species NT M. fittkaui Volkmer-Ribeiro, 1979 (Amazonian Basin) NT M. godeauxi Brien, 1968 (Central Africa) AT M. kiliani Volkmer-Ribeiro and Costa, 1992 (Brazil) NT M. ovogemata Stanisic, 1979 (N-Australia) AU M. pottsi (Weltner, 1895) (Congo basin, Angola, Borneo, Indonesia) AT-OL M. rhodesiana Burton, 1938 (SE-Africa, Congo basin) AT M. spinata (Carter, 1881) (Amazonian basin) NT M. subtilis Volkmer-Ribeiro, 1979 (Amazonian basin) NT M. vesparia (von Martens, 1868) (Borneo, Indonesia, Australia) OL-AU M. vesparioides (Annandale, 1908) (Tenasserim, Burma, Australia) OL-AU Family metSchnikowiidae Czerniawsky, 1880 (1 genus, 1 species) PA Metschnikowia Grimm, 1876 (1 species) PA M. tuberculata Grimm, 1876 (Caspian Sea) monotypic PA Family potamolepidae Brien, 1967 (6 genera, 29 species) NT-AT-PAC Echinospongilla Manconi and Pronzato, 2002 (1 species) AT E. brichardi Brien, 1974 (L. Tanganyika) monotypic AT Oncosclera Volkmer-Ribeiro, 1970 (14 species) NT-PAC-AT O. jewelli Volkmer-Ribeiro, 1963 (Rio Grande do Sul, Brazil) type species NT O. atrata (Bonetto and Ezcurra de Drago, 1973) (R. Apur, Argentina) NT O. diahoti (Rtzler, 1968) (New Caledonia) PAC O. gilsoni (Topsent, 1912) (Fiji) PAC O. intermedia (Bonetto and Ezcurra de Drago, 1973) (R. Orinoco, Venezuela) NT O. navicella (Carter, 1881) (Brazil, Argentina) NT O. petricola (Bonetto and Ezcurra de Drago, 1973) (R. Uruguay, Argentina) NT O. ponsi (Bonetto and Ezcurra de Drago, 1973) (R. Uruguay, Argentina) NT O. rousseletii (Kirkpatrick, 1906) (R. Zambezi, Africa) AT O. schubarti (Bonetto and Ezcurra de Drago, 1973) (R. Uruguay, Argentina, Brazil) NT O. schubotzi Weltner, 1913 (Aruwimi, Congo basin) AT O. spinifera (Bonetto and Ezcurra de Drago, 1973) (R. Orinoco, Venezuela) NT O. stolonifera (Bonetto and Ezcurra de Drago, 1973) (R. Pir, Argentina) NT O. tonolli (Bonetto and Ezcurra de Drago, 1968) (Parami, Brazil) NT Potamolepis Marshall, 1883 (7 specie) AT P. leubnitziae Marshall, 1883 (Congo basin, L. Mweru, R. Niger, L. Tanganyika) type species AT P. belingana Lvi, 1965 (Cameroon, R. Ivindo Gabon) AT P. chartaria Marshall, 1883 (Isangila Congo basin, R. Luapula, L. Tanganyika, R. Niger) AT P. marshalli Burton, 1938 (Matadi Congo basin) AT P. micropora Burton, 1938 (Matadi Congo basin) AT P. pechuelii Marshall, 1883 (Matadi-Matemba Congo basin, L. Tanganyika) AT P. weltneri Moore, 1903 (L. Tanganyika, Zimbabwe) AT

67

Table 1 (cont.) Potamophloios Brien, 1970 (5 species) AT P. stendelli (Jaffe, 1916) (L. Mweru, L. Luapula, L. Tanganyika) type species AT P. gilberti Brien, 1969 (L. Mweru, R. Luapula) AT P. hispida Brien, 1969 (L. Mweru, R. Luapula) AT P. songoloensis Brien, 1969 (L. Mweru, R. Luapula) AT P. symoensi (Brien, 1967) (Luapula basin) AT Sterrastrolepis Volkmer- Ribeiro and de Rosa Barbosa, 1978 (1 species) NT S. brasiliensis Volkmer-Ribeiro and de Rosa Barbosa, 1978 (R. Turvo, Brazil) monotypic NT Uruguaya Carter, 1881 (1 species) NT U. corallioides (Bowerbank, 1863) (Brazil, Uruguay, Venezuela) monotypic NT incertae SediS (3 genera, 3 species) NT-AT-PA Balliviaspongia Boury-Esnault and Volkmer-Ribeiro, 1992 (1 species) NT B. wirrmanni Boury-Esnault andVolkmer-Ribeiro, 1992 (L. Titicaca, Peru-Bolivia) monotypic NT Makedia Manconi, Cubeddu and Pronzato, 1999 (1 species) AT M. tanensis Manconi, Cubeddu and Pronzato, 1999 (L. Tana, Ethiopia) monotypic AT Ohridospongilla Gilbert and Hadzische, 1984 (1 species) PA O. stankovici Gilbert and Hadzische, 1984 (L. Ohrid, Macedonia) monotypic PA

Fig. 1: Taxonomic diversity of freshwater sponges at the zoogeographic scale recorded from the Neotropical (63 species, 22 genera, 3 families), Palaearctic (59 species, 21 genera, 4 families), Afrotropical (49 species, 18 genera, 4 families), Oriental (37 species, 11 genera, 2 families), Australasian (33 species, 13 genera, 3 families), Nearctic (32 species, 13 genera, 2 families), and Oceanic Pacific Islands (5 species, 4 genera, 2 families).

occurred in continental waters. It is proved that Spongillina share the general structure of both eggs and sperms, and the larval stage parenchymella type III sensu Ereskovsky (1999, 2004) with the other Haplosclerida suborders; the parenchymella of Spongillina shows however exclusive traits such as choanocyte chambers and canals. The key evolutionary novelty, represented by peculiar resistant bodies is known among marine Haplosclerina in the genus Haliclona (de Weert 2002), while resistant bodies are absent in the suborder Petrosina (Desqueyroux-Faundez and Valentine 2002). Among the recent Spongillina, the trait resistant body is shared only by the families Metaniidae, Potamolepidae, and Spongillidae, but it is absent in the other families Lubomirskiidae, Malawispongiidae, and Metschnikowiidae (Manconi and Pronzato 2002) (Fig. 2).

Freshwater sponges producing gemmules display a pluriannual life cycle characterised by four steps: vegetative growth phase, gemmulation/sexual reproduction, cryptobiosis, hatching of gemmules and regeneration (Fig. 3). The low metabolism of gemmules allows sponges to survive extreme environmental conditions and to re-establish an active sponge by the rapid proliferation of totipotent cells (Weissenfels 1989, Pronzato and Manconi 1995). The production of gemmules is a trait shared by most freshwater sponges but gemmules display different levels of morphological complexity. The functional role of gemmules is double as propagules and as resting bodies. These evidences suggest testing the working hypothesis the evolutionary success of freshwater sponges at the level of geographic range, species richness, and abundance is related to the efficiency of gemmules as dispersal devices. Moreover the

68

Fig. 2: Schematic reconstruction of the morpho-functional evolutionary trend of dormant bodies in sponges: a. Totipotent cell (several Porifera); b. Gemmule with a simple structure (few Haplosclerina* and few Spongillina**); c. Gemmule with tri-layered pneumatic theca (several Spongillina**); d. Spiny gemmulosclere radially arranged (several Spongillina**). All sponge species are characterized by the presence of totipotent cells able to regenerate or recover the sponge body. Only few species of Haplosclerina*, belonging to the genus Haliclona, are able to produce over wintering gemmules. Metaniidae and Spongillidae, among Spongillina**, bear complex gemmules with multilayered theca, pneumatic layer and spiny gemmuloscleres frequently radially arranged.

Fig. 3: Life cycle phases of Ephydatia fluviatilis recorded in a Sardinian stream by means of time lapse photography: a. Hatching of gemmules and growth; b. Vegetative phase; c. Sexual reproduction, gemmulation and degeneration; d. Cryptobiosis. The presence of the active sponge and gemmules are indicated by dotted and solid strips respectively.

structure of gemmules and their morphological traits are, at present, diagnostic at the family, genus and species level. The families Lubomirskiidae and Metschnikowiidae show an extremely restricted geographic range in the Baikal Lake and the Caspian Sea, respectively (Fig. 4). A peculiar case is represented by Malawispongiidae with a discontinuous distribution in ancient lakes from the Great African Rift Valley (Tanganyika and Malawi) to the Syrian-Palestinian Jordan Rift Valley (Lake Tiberias), and from the Balkanian

area (Ohrid Lake), to the extremely distant Wallacea in the Sulawesi Island (Poso Lake) (Fig. 4). These three families share the trait absence of gemmules being their reproductive mode exclusively sexual or by fragmentation (Manconi and Pronzato 2002). Potamolepidae show a circumtropical range of Gondwanian origin, in rainforests with 29 species (Fig. 4). Their reproductive mode is both sexual and asexual and resistant bodies are gemmules with a simple architecture (Fig. 5).

69

Fig. 4: Geographic distribution of the seven recent families belonging to the suborder Spongillina. The total number of species and genera are reported for each family on maps.

Spongillidae are cosmopolitan with 146 species, while the Gondwanian Metaniidae are Circumtropical with 22 species (Fig. 4). The two families share the traits sexual and asexual reproductive mode and gemmules with a complex architecture (Fig. 5). These complex gemmules with an armed gemmular theca to protect totipotent cells, seems to be perfect dispersal propagules (Fig. 5). Several species of Spongillidae are characterised by a gemmule with a well developed pneumatic layer to float and to perform dispersal downstream, spicules from tangentially to radially arranged to strengthen the gemmular theca, and spiny spicules are able to hook efficiently onto the carrier. In some genera, as Stratospongilla, a double spicular layer protects the gemmule more efficiently (Fig. 5),

and the genus Saturnospongilla shows the displacement of the pneumatic layer with the result of a ring-shaped pneumatic layer possibly to increase the performances for anemophilous-hydrochorous dispersal (Fig. 5). A comparative analysis of the gemmular architecture clearly shows that several species of the two families Metaniidae and Spongillidae share most gemmular diagnostic traits. Spicules are radially arranged in the gemmular theca, the pneumatic layer is well developed, gemmuloscleres are spiny, and a cage of megascleres protects the gemmule (Fig. 5). A first evidence of this analysis is that the working hypothesis freshwater sponge success is strictly related to the efficiency of gemmules as dispersal devices does not work. In fragmented-discontinuous habitats it appears that

70

Fig. 5: Gemmular structure in different families of freshwater sponges. a-b. Gemmules of Potamolepidae (a, gemmular theca of Potamophloios stendelli; b, Gemmular cage of Uruguaya corallioides). df. Gemmules of Metaniidae are more complex, with a pneumatic layer almost always present, and radially arranged gemmuloscleres bearing hooks in most species (c-e, Metania reticulata); in some speciess an external cage adds protection to the gemmular theca (f, Drulia browni). g-j. The general structure of gemmules is shared by Spongillidae and Metaniidae: developed pneumatic layer, spiny gemmuloscleres radially arranged (g-i, Anheteromeyenia argyrosperma), outer spicular cage (j, Stratospongilla bombayensis). k. Saturnospongilla carvalhoi, among Spongillidae, shows a gemmular structure with a peculiar ring-shaped pneumatic layer (arrows).

71

Fig. 6: The gemmule structure of Spongillidae. Umborotula bogorensis shows spiny gemmuloscleres radially arranged (a-c) within the pneumatic layer (c). Species belonging to the genus Eunapius display a peculiar pneumatic layer (d, E. carteri) with spiny gemmuloscleres not radially arranged (e, E. nitens). Some gemmules of Spongilla lacustris lack gemmuloscleres (f) and pneumatic layer (g).

the absence of gemmules determines the absence of dispersal at a large-scale and therefore a condition of endemism as in sponges from ancient lakes (Fig. 4). On the other hand, dispersal power would be low for simple gemmules (absence of spiny gemmuloscleres, absence of pneumatic layer) with consequently a relatively restricted geographic range as in the case of Potamolepidae (Fig. 4-5). Finally, taxa bearing complex gemmules would show a high dispersal power with a tendency to cosmopolitism. This might happen for Metaniidae and Spongillidae.

Biogeographic data on Metaniidae show however that a highly complex gemmular architecture does not support their relatively restricted distribution and low species richness. Although the gemmular complexity in Spongillidae and Metaniidae is comparable, a notable divergence exists between the worldwide distribution of the former and the Circumtropical pattern of the latter. The geographic range and the species richness of Metaniidae are, on the other hand, similar to that of Potamolepidae characterised by a simple gemmule (Figs 4-5).

72

pneumatic layer is absent (Figs 6-7) suggesting that the trait complexity of gemmular architecture does not support the relatively restricted distribution of the genus Umborotula vs. Eunapius and Spongilla. At the species level within the genus Spongilla, different biogeographic patterns are evident, some species being very common and widespread - as in the case of the Holarctic Spongilla lacustris - while other species are rare and monotopic - as for Spongilla prespensis Hadzische, 1953 and Spongilla stankovici Arndt, 1938 from the Balkanian area, respectively endemic to Lake Prespa and Lake Ochrid (Pronzato and Manconi 2002) (Fig. 8). It is consequently evident that it is not true that taxa with complex gemmules as perfect dispersal devices show a wider geographic range; this statement is valid at the family, the genus and the species level. We can consider a further condition among Spongillidae. Spongilla lacustris displays two gemmular types (Manconi and Desqueyroux-Faundez 1999) and its widespread distribution, with a Holarctic range, could be related to this peculiar condition. Also, species belonging to the genus Corvospongilla, one of the most speciose within Spongillidae, display two gemmular types diverging in their morpho-functional roles. A first type, the sessile gemmule, is strictly adherent to the substratum, with absence of pneumatic layer and well developed spicular cage; the supposed functional role of this heavy gemmule is to persist in situ. A second type is a free gemmule in the sponge skeleton, with a well developed pneumatic layer and without spicular cage; the supposed functional role of this light gemmule is dispersal (Fig. 9) (Manconi et al. 2004). The genus Corvospongilla shows at the biogeographic scale a notably disjunct range with 17 species extremely rare and known only from restricted areas. This biogeographic pattern matches better the vicariance model versus the dispersal one. The evolutionary novelty of two diverging gemmular architectures seems not to favour the wide spreading of the genus Corvospongilla, being most species endemic and rare (Manconi and Pronzato 2004).

Approaching conclusions
Different possible problems could affect our working hypothesis; they could be: i) errors in the systematics; ii) a high richness in species complexes with the result of misleading geographic ranges; iii) a diversified physiological control of gemmular dormancy. Another approach to explain the relationships between the structure of gemmules and the distribution of freshwater sponge taxa could be to consider the old axiom of Biogeography: species distribution and richness is frequently strictly related to the distribution and number of taxonomists. If we focus on the species richness of African freshwater sponges, since the first record by Carter in 1881, the temporal trend shows clearly that taxonomic richness increased slowly in more than one century, with a present maximum of 56 species at the continental scale (other 6 taxa are reported at the genus, family and suborder level) out of 175 findings (Manconi and Pronzato in press b, unpublished). Moreover endemicity s.s. is notably high (78.5%), 23 species are known

Fig. 7: Geographic distribution of the genera Umborotula, Eunapius and Spongilla. The latter two genera are cosmopolitan although their gemmules are simple (see Fig. 6), while the former is restricted to the Oriental Region.

The analysis of gemmular complexity at the genus level within Spongillidae indicates that the genus Umborotula, with a highly complex gemmule (Fig. 6), is monotypic (only 1 species) and restricted to E-Asia and Australia (Manconi and Pronzato 2002) (Fig. 7). On the other hand the genus Eunapius is cosmopolitan with a moderately complex gemmular architecture (spicules not radially arranged, spicules without hooks) (Figs 6-7). This condition is also shared by the genus Spongilla, where sometimes the

73

Fig. 8: Pattern of geographic distribution in the genus Spongilla. Holarctic range of S. lacustris vs. S. prespensis and S. stankovici each strictly endemic to two different Balkanian lakes.

only from the holotype, and the knowledge on some areas is scattered and fragmentary, if not lacking, as for Madagascar. All these data strongly suggest that taxonomic richness of African freshwater sponges is underestimated and species distribution is probably mislead (Manconi and Pronzato in press b). A historical analysis of the whole knowledge on freshwater sponges in the field of taxonomy and biogeography - based on the number of papers per year - shows that only 3 species have been described before the erection of the genus Spongilla by Lamarck (1816). The following 60 years are characterised by long-lasting periods of inactivity (no papers) or very low activity (1-2 paper/year). Successively a positive trend is evident, with peaks corresponding both to the most productive authors, and to the periodical maxima of taxonomic richness increase (Fig. 10). Annandale, for example, described 29 new species in 10 years (1907-1918), Potts described 16 species in 10 years (1880-1889), Weltner described 15 species in 21 years (1893-1913), Bonetto and Ezcurra de Drago described 17 species in 8 years (1966-1973) always together, VolkmerRibeiro described 13 species in 43 years (1963-2005) with

different co-authors, Brien described 11 species in 8 years (1967-1974) while Bowerbank described 11 species in a single year (1863). We can generalize that species richness of freshwater sponges seems, at present, to be underestimated and destined to increase with further researches based on a critical analysis and a synthesis on all materials in collections and on all taxonomic data from the literature, to have a strong morphological basis to compare and support the results of molecular biology. New sampling campaigns are also needed, as suggested by the recent new findings on the freshwaters sponge fauna. For example, one out of five samplings in the Australian Kakadu park resulted in the discovery of new morpho-traits for the genus Pectispongilla, and additional ones for the whole Spongillidae (Manconi et al. 2006). If we move to the Caribbean area, 12 samplings in the West Indies resulted in the discovery of a notably rich Spongillino-fauna in the island of Cuba (Manconi and Pronzato 2005). Finally, from a total of 209 contributions reported in the VIth Sponge Conference abstract book, ten concern freshwater sponges but only two focus on the field of systematics

74

Fig. 9: Corvospongilla mesopotamica is characterized by the presence of two gemmular morpha. The sessile heavier gemmule, strictly adherent to the substratum, lacks pneumatic layer and is protected by a reinforced spicular cage (a); the free lighter gemmule shows a well developed pneumatic layer (b) and few gemmuloscleres; c. Schematic reconstruction of the life cycle of C. mesopotamica to evidence the double role of gemmules diverging in their morpho-functional role. Circles indicate dispersal (light gemmules), squares indicate persistence (heavy gemmules).

Fig. 10: Yearly production of taxonomic and biogeographic papers on freshwater sponges after their actual separation from other sponge higher taxa. Data source: Zoological Record.

75

(Pansini and Pronzato 2002). This trend is confirmed by contents of the 7th International Sponge Symposium abstract book: only 16 contributions, out of 308, concern Spongillina (10 on systematics, faunistic and distribution) (Custdio et al. 2006). At present the number of taxonomists on freshwater sponges is dramatically low (less than ten) and without absolutely necessary new recruitments they are on the brink of extinction.

Acknowledgements
This paper is dedicated to the memory of the Brazilian specialist of freshwater sponges Rosaria de Rosa-Barbosa. R. Manconi is grateful to the organizers of the 7th International Sponge Symposium for their kind invitation and financial support. Research supported by the Italian Ministero dellIstruzione, dellUniversit e della Ricerca Scientifica e Tecnologica (MIUR-PRIN 2004057217 Zoogeography of Mediterranean-Southern African disjunct distributions by a multimethod approach), the European program INTERREG Sardinia-Corsica-Tuscany on Biodiversity, the Universit di Sassari and Universit di Genova.

References
Adamson DA, Clark JD, Williams MAJ (1987) Pottery tempered with sponge from the White Nile, Sudan. Afr Arch Rev 5: 115-127 Adis J (1992) On the survival strategy of Mesostoma hylaeicum Jeekel, a millipede from Central Amazonian floodplains. Ber Natmed Ver Innsbruck, suppl. 10: 183-187 Allen TF (1986) Encyclopaedia of pure materia medica, Vol. II, B. Jain Publishers, New Delhi Armstrong G, Booth DT (2005) Dietary ecology of the Australian freshwater turtle (Elseya sp.: Chelonia: Chelidae) in the Burnett River, Queensland. Wildlife Res 32: 349-353 Berg K (1948) Biological studies on the river Susaa. Faunistic and biological investigations. Porifera. Folia Limnol Scand 4: 10-23 Brien P (1966) Le poliphiletisme des ponges deau douce. Lembryognse et la larve chez Potamolepis stendelli (Jaff). C R Acad Sci Paris 263: 649-652 Brien P (1969) Les Potamolepides africaines. Polyphyletisme des ponges deau douce. Arch Zool Exp Gen 110: 527-561 Brissaud I, Houdayer A (1986) Sponge spicules as a characteristic of ancient African pottery from Mali. J Fld Arch 13: 357-358 Brnsted A, Brnsted HV (1953) The effect of symbiotic zoochlorellae on the germination rate of gemmules of Spongilla lacustris (L.). Vidensk Medd Dan Naturhist Foren 115: 133-144 Brnsted HV, Louvtrup E (1953) The respiration of sponge gemmules without and with symbiotic unicellular algae. Vidensk Medd Dan Naturhist Foren 115: 145-157 Brown HP (1952) The life history of Climacia areolaris (Hagen), neuropterous parasite of freshwater sponges. Amer Mid Nat 47: 130-160 Candido JL, Volkmer-Ribeiro C, Filho FLS, Turcq BJ, Desjardins T, Chauvel A (2000) Microsclere variations of Dosilia pydanieli (Porifera, Spongillidae) in Caracaran lake (Roraima - Brazil). Palaeoenvironmental implication. Biociencias 8(2): 77-92 Carter HJ (1881) History and classification of the known species of Spongilla. Ann Mag Nat Hist 7: 77-107 Corriero G, Vaccaro P, Manconi R, Pronzato R (1993) Life strategies of Ephydatia fluviatilis (L., 1759) in two different environments.

In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space. Balkema, Rotterdam. pp. 321-326 Crane K (1991) Hydrothermal vents in Lake Baikal. Nature 350(6316): 281 Custdio MR, Lbo-Hajdu G, Hajdu E, Muricy G (eds.) (2006) 7th International Sponge Symposium - Book of Abstracts (Armao dos Bzios, Brazil). Museu Nacional, Srie Livros 16, Rio de Janeiro Desqueyroux-Faundez R, Valentine C (2002) Phloeodictyidae, Petrosiidae. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges. Kluwer Academic/Plenum Publishers, New York. pp. 893-917 Dominey J, Snyder AM (1988) Kleptoparasitism of freshwater crabs by cichlid fishes endemic to Lake Barombi Mbo, Cameroon, West Africa. Env Biol Fish 22(2): 155-160 Dunagan SP (1999) A North American freshwater sponge (Eospongilla morrisonensis new genus and species) from the Morrison Formation (Upper Jurassic), Colorado. J Paleontol 73(3): 389-393 Ereskovsky AV (1999) Development of sponges of the order Haplosclerida. Russian J Mar Biol 25(5): 361-371 Ereskovsky AV (2004) Comparative embryology of sponges and its application for poriferan phylogeny. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Boll Mus Ist Biol Univ Genova 68: 301-318 Francis JC, Harrison FW (1988) Copper and zinc toxicity in Ephydatia fluviatilis (Porifera: Spongillidae). Trans Am Microsc Soc 107(1): 67-78 Frost TM (1978) In situ measurements of clearance rates for the freshwater sponge Spongilla lacustris. Limnol Oceanogr 23(5): 1034-1039 Frost TM (1980) Clearance rate determinations for the freshwater sponge Spongilla lacustris: effect of temperature, particle type and concentration, and sponge size. Arch Hydrobiol 90(3): 330-356 Frost TM (1991) Porifera. In: Thorp JH, Covich AP (eds). Ecology and classification of North American freshwater invertebrates. Academic Press, New York. pp. 95-124 Frost TM, Williamson CE (1980) In situ determination of the effect of symbiotic algae on the growth of the fresh-water sponge Spongilla lacustris. Ecology 61: 1361-1370 Gugel J (1996) The sponge mite Unionicola minor (Acari, Hydrachnellae, Unionicolidae) in a pond at the spring of the Darmbach near Darmstadt. Hessische Faunistische Briefe 15(1): 11-16 Harrison FW (1974) Porifera. In: Hart CW, Fullers SLH (eds). Pollution ecology of freshwater invertebrates. Academic Press, New York. pp. 29-66 Harrison FW (1977) The taxonomic and ecological states of the environmentally restricted spongillid species of North America. III Corvomeyenia carolinensis Harrison, 1971. Hydrobiologia 56(2): 187-190 Harrison FW (1988) Utilization of freshwater sponges in paleolimnological studies. In: Gray J (ed). Paleolimnology: aspects of freshwater paleoecology and biogeography. Elsevier, Amsterdam. pp. 387-397 Harrison FW, Warner BG (1986) Fossil freshwater sponges (Porifera: Spongillidae) from Western Canada: an overlooked group of Quaternary paleoecological indicators. Trans Am Microsc Soc 105(2): 110-120

76

Harrison FW, Gleason PJ, Stone PA (1979) Paleolimnology of Lake Okeechobee, Florida: an analysis utilizing spicular components of freshwater sponges (Porifera: Spongillidae). Not Nat Acad Nat Sci Philadelphia 454: 1-6 Jewell ME (1935) An ecological study of the fresh-water sponges of northeastern Wisconsin. Ecol Monogr 5: 461-504 Jewell ME (1939) An ecological study of the freshwater sponges of Wisconsin. II. The influence of calcium. Ecology 20: 11-28 Kahl K, Konopacka A (1981) Oligochaeta inhabiting the colonies of the sponge Spongilla lacustris (L.) in the River Gac. Acta Hydrobiol 23(3): 243-249 Kamaltynov RM, Chernykh VI, Slugina ZV, Karabanov EB (1993) The consortium of the sponge Lubomirskia baikalensis in Lake Baikal, east Siberia. Hydrobiologia 271(3): 179-189 Kennett R, Tory O (1996) Diet of two Freshwater turtles, Chelodina rugosa and Elseya dentata (Testudines: Chelidae) from the wetdry tropics of Northern Australia. Copeia 2: 409-419 Kilian EF, Campos H (1969) Freshwater sponge as sites for hatching of the brood of a fish. Naturwissenschaften 5(6): 333 Konopacka A, Socinski J (1985) Macrofauna inhabiting the colonies of the sponge Spongilla lacustris (L.) in the River Gac. XIII. Ecology of aquatic organisms. 3. Animals. Verh Int Verein Limnol 22: 2968-2973 Kuznetzow J (1898) Fischerei und thierebentung in den Gewassern Russlands. Ministerium fur landwirthschaft und reichsolomanen. Department fur landwirthschaft, St. Petersburg Lamarck JBPA (1816) Histoire naturelle des animaux sans vertbres. Verdire, Paris Lauer TE, Spacie A (2004) Space as a limiting resource in freshwater systems: competition between zebra mussels (Dreissena polymorpha) and freshwater sponges (Porifera). Hydrobiologia 517(1): 137-145 Linn S (1925) The technique of the South American ceramics. Gteb K Vetensk Vitter Hets-Samh-Handl Sjatte Foljden Ser. B 29(5): 1-199 Manconi R, Desqueyroux-Faundez R (1999) Freshwater sponges (Porifera, Spongillidae) from the Lake of Geneva, Switzerland. Rev Suisse Zool 106(3): 571-580 Manconi R, Pronzato R (1991) Life cycle of Spongilla lacustris (Porifera, Spongillidae): a cue for environmental dependent phenotype. Hydrobiologia 220: 155-160 Manconi R, Pronzato R (1994) Spongillids of Mediterranean islands. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space. Balkema, Rotterdam. pp. 333-340 Manconi R, Pronzato R (2002) Spongillina subord. nov. Lubomirskiidae, Malawispongiidae n. fam., Metaniidae, Metschnikowiidae, Potamolepidae, Spongillidae. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges. Kluwer Academic/Plenum Publishers, New York. pp. 921-1019 Manconi R, Pronzato R (2004) The genus Corvospongilla Annandale (Haplosclerida, Spongillina, Spongillidae) with description of a new species from eastern Mesopotamia, Iraq. Arch Hydrobiol suppl Mon Stud 151(1-2): 161-189 Manconi R, Pronzato R (2005) Freshwater sponges of the West Indies: discovery of Spongillidae (Haplosclerida, Spongillina) from Cuba with biogeographic notes and a checklist for the Caribbean area. J Nat Hist 39(36): 3235-3253 Manconi R, Pronzato R (in press a) Global diversity of sponges (Porifera: Spongillina) in freshwater. In: Balian E, Martens

K, Leveque C (eds). Freshwater animal diversity assessment. Hydrobiologia Manconi R, Pronzato R (in press b) Sponges from African inland waters: the genus Eunapius (Haplosclerida, Spongillina, Spongillidae). Arch Hydrobiol suppl Mon Stud Manconi R, Cubeddu T, Corriero G, Pronzato R (1999) Commercial sponges farming as natural control of floating cages pollution. In: Enne G, Greppi F (eds). New species for Mediterranean aquaculture. Biofutur Elsevier, Paris. pp 269-274 Manconi R, Saadalla HAA, Cubeddu T, Ferretti C, Pronzato R (2004) Resting stages of Corvospongilla sp. (Demospongiae, Spongillidae) from Mesopotamia, Iraq. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds) Sponge science in the new millennium. Boll Mus Ist Biol Univ Genova 68: 441-448 Manconi R, Cubeddu T, Ferretti C, Pronzato R (2006) A new morphological trait for northern Australian pectispongillas. In: Custdio MR, Lbo-Hajdu G, Hajdu E, Muricy G (eds). 7th International Sponge Symposium - Book of Abstracts (Armao dos Bzios, Brazil). Museu Nacional, Srie Livros 16, Rio de Janeiro. pp. 225 McCauley DG, Longcore JR (1988) Foods of juvenile ring-necked ducks: relationship to wetland pH. J Wildl Manage 52(2): 177185 McIntosh SK, MacDonald KC (1989) Sponge spicules in pottery: new data from Mali. J Fld Arch 16: 489-494 Old MC (1932) Environmental selection of the fresh-water sponges (Spongillidae) of Michigan. Trans Amer Microsc Soc 51: 129-136 Oliveira Roque de F, Trivinho-Strixino S, Marques Couceiro SR, Hamada N, Volkmer-Ribeiro C, Messias MC (2004) Species of Oukuriella Epler (Diptera, Chironomidae) inside freshwater sponges in Brazil. Rev Bras Entomol 48(2): 291-292 Ott E, Volkheimer W (1972) Palaeospongilla chubutensis n.g. et n.sp. ein susswasserschwamm aus der Kreide Patagoniens. N Jb Geol Palaont Abh 140: 49-63 Pansini M, Pronzato R (eds) (2002) VIth International Sponge Conference, Book of Abstracts. Boll Mus Ist Biol Univ Genova 66-67: 1-706 Parfin SI, Gurney AB (1956) The spongillaflies with special reference to those of the Western Hemisphere (Sisyridae, Neuroptera). Proc Unit States Natl Mus 105: 421-529 Pronzato R, Manconi R (1994) Life history of Ephydatia fluviatilis: a model for adaptive strategies in discontinuous habitats. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space. Balkema, Rotterdam. pp. 327-331 Pronzato R, Manconi R (1995) Long term dynamics of a freshwater sponge population. Freshwat Biol 33: 485-495 Pronzato R., Manconi R (2002) Atlas of European freshwater sponges. Ann Mus Civ St Nat Ferrara 4: 3-64 Pronzato R, Manconi R, Corriero G (1993) Biorhythm and environmental control in the life history of Ephydatia fluviatilis (Demospongiae Spongillidae). Boll Zool 60: 63-67 Rader RB (1984) Factors affecting the distribution of a freshwater sponge. Freshwat Invertebr Biol 3: 86-97 Rader RB, Winget RN (1985) Seasonal growth rate and population dynamics of a freshwater sponge. Hydrobiologia 123: 171-176 Resh MA (1976) Life cycles of invertebrate predators of fresh-water sponge. In: Harrison FW, Cowden RR (eds). Aspects of sponge biology. Academic Press, New York. pp. 299-354

77

Resh VH, Morse JC, Wallace JD (1976) The evolution of the sponge feeding habit in the caddis fly genus Ceraclea (Trichoptera: Leptoceridae). Ann Entomol Soc Amer 69(5): 937-941 Ricciardi A, Reiswig HM (1993) Freshwater sponges (Porifera, Spongillidae) of eastern Canada: Taxonomy, distribution, and ecology. Can J Zool 71(4): 665-682 Ricciardi A, FL Snyder, DO Kelch, HM Reiswig (1995) Lethal and sublethal effects of sponge overgrowth on introduced dreissenid mussels in the Great Lakes - St. Lawrence River system. Can J Fish Aquat Sci 52: 2695-2703 Richelle-Maurer E, Degoudenne Y, van de Vyver G, Dejinghe L (1994a) Some aspects of the ecology of Belgian freshwater sponges. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space. Balkema, Rotterdam. pp. 341350 Richelle-Maurer E, Degoudenne Y, van de Vyver G, Dejinghe L (1994b) Some aspects of heavy metal tolerance in freshwater sponges. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space. Balkema, Rotterdam. pp. 351354 Richter G, Wuttke M (1999) Lutetiospongilla heili n. gen. n. sp. und die eozne Spongillidenfauna von Messel. Cour Forschungsinst Senckenb 216: 183-195 Ronde De CEJ, Stoffers P, Garbe-Schonberg D, Jones B, Manconi R, Christenson BW, Browne PRL., Hissman K, Botz R, Davy BW, Schmitt M, Battershill CN (2002) Discovery of active hydrothermal venting in Lake Taupo, New Zealand. J Volcan Geotherm Res 115: 257-275 Rota E, Manconi R (2004) Taxonomy and ecology of spongeassociate Marionina spp. (Clitellata, Enchytraeidae) from the Horomatangi Geothermal System of Lake Taupo, New Zealand. Int Rev Hydrobiol 89: 58-67 Sar M, Vacelet J (1973) cologie des Dmosponges: influence des facteur physicochimiques. In: Grass PP (ed). Trait de Zoologie, Anatomie, Systematique, Biologie. 3(l) Masson et Cie, Paris. pp. 462-576 Serrano A (1933) Observacines sobre la lfareria de los-medanos de Coln. Mem Mus Parana 6: 1-2 Seigel RA, Brauman RJ (1994) Food habits of the yellow-blotched map turtle (Graptemys flavimaculata). Mus Tech Rep No. 28, Mississippi Museum of Natural Science, Jackson Simpson TL, Fell PE (1974) Dormancy among the Porifera: gemmule formation and germination in fresh-water and marine sponges. Trans Am Microsc Soc 93(4): 544-577 Steffen AW (1967) Ectosymbiosis in aquatic insects. Symbiosis 2: 207-209 Swierczynski M (1996) A cooccurrence of sponges (Porifera) and Dreissena polymorpha (Pall.) Przegl Zool 40(1-2): 119-121

Vacelet J (1994) Porifera. In: Juberthie C, Decu V (eds). Encyclopaedia Biospeologica, Socit Internationale de Biospologie. Pub. 1. Moulis, Bucarest. pp. 35-38 van Soest RWM, Velikonja M (1986) Porifera. In: Botosaneanu L (ed). Stygofauna Mundi. Brill, Leiden Volkmer-Ribeiro C (1979) Evolutionary study of the freshwater sponge genus Metania Gray, 1867: I. The new species. Amazoniana 6(4): 639-649 Volkmer-Ribeiro C (1986) Evolutionary study of the freshwater sponge genus Metania Gray, 1867: III. Metaniidae, new family. Amazoniana 9(4): 493-509 Volkmer-Ribeiro C (1990) A new insight into the systematics, evolution and taxonomy of freshwater sponges. In: Rtzler K (ed). New perspectives in sponge biology. Smithsonian Institution Press, Washington DC. pp. 323-33 l Volkmer-Ribeiro C, Turcq BJ (1996) SEM analysis of siliceous spicules of a freshwater sponge indicates paleoenvironmental changes. Acta Microsc 5B: 186-187 Volkmer-Ribeiro C, de Rosa-Barbosa R (1974) A freshwater spongemollusk association in Amazonian waters. Amazoniana 5(2): 285291 Volkmer-Ribeiro C, de Rosa-Barbosa R (1979) Neotropical freshwater sponges of the family Potamolepidae Brien, 1967. In: Lvi C, Boury-Esnault N (eds). Biologie des spongiaires. Coll Int CNRS 291. pp. 503-511 Volkmer-Ribeiro C, Watanabe Y (1983) Sanidastra yokotonensis, n. gen. and n. sp. of freshwater sponge from Japan. Bull Nat Sci Mus Tokyo Ser A 9(4): 151-159 Weinberg I, Glyzina O, Weinberg E, Kravtsova L, Rozhkova N, Sheveleva N, Natyaganova A, Bonse D, Janussen D (2004) Types of interactions in consortia of Baikalian sponges. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds) Sponge science in the new millennium. Boll Mus Ist Biol Univ Genova 68: 655-663 Weert de WH (2002) Chalinidae. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges. Kluwer Academic/Plenum Publishers, New York. pp. 852-873 Wells A (1995) New Caledonian Hydroptilidae (Trichoptera), with new records, descriptions of larvae and a new species. Aquat Insects 17(4): 223-239 Weissenfels N (1989) Biologie und mikroskopische anatomie der susswasserschwamme (Spongillidae). Fischer, Stuttgart, New York Weissmair W, Mildner P (1995) On the study of Sisyridae (Neuroptera), their hosts and their aquatic habitat in Carinthia. Carinthia II 105(2): 535-552 Willenz P, Vray B, Maillard MP, van de Vyver G (1986) A quantitative study of the retention of radioactively labeled E. coli by the freshwater sponge Ephydatia fluviatilis. Physiol Zool 59(5): 495-504

Porifera research: Biodiversity, innovation and sustainaBility - 2007

79

Evolution of multicellularity in Porifera via selfassembly of glyconectin carbohydrates

Gradimir N. Misevic(1*), Camille Ripoll(1), Jonathan Norris(1), Vic Norris(1), Yann Guerardel(2), Emmanuel Maes(2), Gerard Strecker(2), Pascal Ballet(3), Yannis Karamanos(4), Lazar T. Sumanovski(5), Octavian Popescu(6), Nikola Misevic(7)
Laboratoire Assemblages Molculaires: Modlisation Imagerie et SIMS, FRE CNRS 2829, Facult des Sciences de lUniversit de Rouen, 76 821 Mont Saint Aignan Cedex, France, gradimir@gradimir.com (2) Unit de Glycobiologie Structurale et Fonctionnelle, Universit des Sciences et Technologies de Lille, UMR 8576 CNRS, 59655 Villeneuve DAscq, France (3) University of Brest, Brest, France (4) Laboratoire de Biochimie Molculaire et Cellulaire, Universit dArtois, Facult J Perrin, rue J. Souvraz, SP18, 62307 Lens, France (5) Department of Research, University Hospital of Basel, CH-4058 Basel, Switzerland, (6) Molecular Biology Center and Institute for Interdisciplinary Experimental Research, Babes-Bolyai University, 400006 Cluj-Napoca, Rumania (7) University of Bremen, 28359 Bremen, Germany
(1)

This work is dedicated to the memory of Maurice Demarty Abstract: Research done in the last century on Porifera has provided insights into the molecular mechanism of the biological processes of cell adhesion, innate immunity, and self-recognition. Evidence that this mechanism is based on glyconectin selfassembly is shown by the structure to function relationships deduced from studies of carbohydrates isolated from three different sponge species. The structural studies were performed on purified glyconectin carbohydrates from Microciona prolifera, Halichondria panicea and Cliona celata using nuclear magnetic resonance and mass spectrometry. Seventeen novel, speciesspecific carbohydrate sequences were revealed that belong to the Porifera glyconectin family. The functional, cell recognition analyses of carbohydrate self-association were performed by measuring binding forces between individual glycan molecules under physiological conditions; the results show that the association strength between homotypic pairs of glycans (400 pN) are higher than those between heterotypic pairs (20 pN). This difference is sufficient to explain the species-specific separation of glycan-coated beads in vitro and the sorting of cells in nature. We propose that the glyconectin carbohydrates, which are the constituents on the cell surface that are the most exposed to the environment, were responsible for the molecular recognition processes that underpinned the emergence of multi-cellularity. Keywords: evolution of multicellularity, Porifera, cell recognition and adhesion, glyconectin carbohydrates and proteoglycans, atomic force microscopy

An approach to study the evolution of multicellularity Why?

Changing patterns of matter and energy achieved in multicellular life their most sophisticated form so far known. During biological evolution, two essential steps occurred: the first was the emergence of cells and the second subsequent one was the development of multicellular assemblies. In an effort to understand the emergence and evolution of the complex and versatile multicellular life forms present today on earth, as well as to satisfy our curiosity of whether other kinds of similarly hierarchically organized life could exist in

the universe, whether it be in the past, present or future, two common, related questions may be formulated in the following way: what were the molecular mechanisms that enabled the creation and persistence of multicellular (or multi unit) life? And, can the similar or alternative mechanisms be predictable in both time and space? In order to devise experiments that could generate data to provide, in part, the answer to these questions, we begin with the following logical statement: the evolution of multicellular forms of life required the emergence of cellular self-non-self discrimination and adhesion, where the sensor molecules guiding such recognition and adhesion should be present at the outermost cell surface (although many processes based on novel as yet imaginary physico-

80

chemical principles might be invoked in extraterrestrial life forms). By gaining recognition and adhesion properties, the primordial multicellular organisms could preserve functional and morphological identity throughout their life cycles.

Structural analyses of glyconectin glycan self-nonself recognition molecules

Structural analyses of glyconectins (GNs) isolated from three sponge species Microciona prolifera (GN1), Halichondria panicea (GN2) and Cliona celata (GN3), used as a experimental model system to study evolution of the multicellularity, will be reviewed here. General approach, depicted in Fig. 1, consists of four steps ranging from the release of glycans of their protein core towards glycan fragmentation, fingerprinting and sequencing using nuclear magnetic resonance (NMR) and mass spectrometry (MS). In the first step protein free polysaccharide chains from purified GNs were prepared by extensive pronase digestion of the protein part (Misevic et al. 1982, Misevic and Burger 1993). This was followed by separation of intact glycans from free amino acids by gel filtration and ionexchange chromatography. In the second step glycans were separated and isolated by column chromatography and/or gel electrophoresis (Fig 1). GN1 was found to have two major glycans with molar masses 200 x 103 and 6 x 103, as previously reported (Misevic et al. 1982, Misevic and Burger 1986, 1990a, 1990b, 1993, Spillmann et al. 1993, Spillmann et al. 1995), GN2 had one major glycan with molar mass of 180 x 103 kD, representing more then 60% of the total carbohydrate content, and GN3 contained also one major glycan species (50% of total carbohydrates) with molar mass of 110 x 103 kD (Fig. 1). The third step in our analyses was chemical and enzymatic fragmentation of GN glycans. The results obtained revealed that each species has its own fingerprint signature. The last step was sequencing of each GN glycan fragment by combination of two dimensional COSY90 high resolution NMR and three types of MS: EI-MS - Electronic Impact Mass Spectrometry, CI-MS - Chemical Ionization Mass Spectrometry and MALDI-TOF MS - Matrix Assisted Laser Desorption Ionization Time of Flight Mass Spectrometry (Fig 1). Complex analyzes of NMR and MS fingerprinting data revealed that four large glycans g200 and g6, g180, and g110 of GN1, GN2 and GN3 respectively, are built by novel repetitive units (Misevic et al. 1982, 1987, Misevic 1989, Spillmann et al. 1993, 1995, Misevic and Burger 1986, 1990a, 1990b, 1993, Popescu and Misevic 1997) (Fig. 2). As shown in Table 1 four short sulfated and one pyruvilated unit in GN1, eight larger and branched pyruvilated oligosaccharides in GN2 which represent heterogeneous but related family of structures, and four sulfated units in GN3 were sequenced (Guerardel et al. 2004, Misevic et al. 2004). We propose two possible models of organization for GN1 and GN3 carbohydrate moieties within the glycan chain (Fig. 2A and B). The first model represents a high molecular weight, linear, acid sensitive polysaccharide connected to an acid resistant domain. This polysaccharide may be composed of either heterogeneous short repetitive units or a large homogeneous repetitive unit comprising acidic labile glycosidic bonds (Fuc/Ara). The actual size of a homogeneous repetitive unit is difficult to assess since most Fuc- and Araglycosidic linkages are cleaved in mild acidic conditions. The second model represents a mixed ramified polysaccharide composed of an acid resistant core connected through Fuc/

How?
In our experimental study of the molecular bases guiding the evolution of multicellularity, it was imperative to select the most appropriate model systems and organisms. We have chosen xenogeneic cell recognition and adhesion of several sponge species from the phylum Porifera because they represent today the simplest multicellular life forms, closest in terms of evolution to primordial multicellular life. In consequence, the molecular mechanism of self-non-self discrimination and adhesion in sponges should be most similar to the mechanism that operated during the emergence of multicellularity. At the beginning of the last century, fundamental phenomenological experiments showed that the xenogeneic re-aggregation of dissociated sponge cells is promoted by extracts made from the surfaces of their cells (Wilson 1907, Galstoff 1925, Curtis 1962, Moscona 1968). Numerous repetitions and variations of these experiments and characterization of aggregationpromoting extracts, called at the time aggregation factors, have been done (Humphreys 1963, Cauldwell et al. 1973, Muller and Zahn 1973, Jumblatt et al. 1980). More sophisticated purification and characterization of aggregation-promoting extracts have shown that they contain a new class of large cell surface proteoglycan-like molecules, heavily covered by long glycan chains, named by G.N. Misevic glyconectins (GNs, derived from glyco connecting, connecting cells via glycans; Papakonstantinou and Misevic 1993, Dammer et al. 1995, Misevic and Popescu 1995, Guerardel et al. 2004, Misevic et al. 2004). Large GN glycans are the outermost macromolecules on the cell surface and display an enormous variability despite their similar structures (Fig. 1). These glycans should therefore be considered as the primary candidates for sensing the environment and performing the self/non-self recognition and adhesion essential for the evolution of the multicellularity. In the first part of this report we describe the key experiments performed to identify, isolate and sequence functional glyconectin glycan molecules. In the second part we explain our novel experimental approach and show quantitative measurements of recognition and adhesion phenomenon on the molecular and cellular level. Finally, in the last part we present a few thoughts and facts about the value of the Porifera model system to basic research.

Emergence of complexity
The emergence of more complex multicellular organisms was based on the appearance of higher degrees of complexity and the multistep nature of cell recognition and adhesion systems. These can be related to 1) allogeneic self-non-self discrimination in the divergence of species and 2) syngeneic organ and tissue specificity during morphogenesis.

81

Fig. 1: The fist panel shows EM, AFM and X-ray images of glycans dimensions at cellular, molecular and atomic level. EM; the Electron Microscope image of cells stained for acidic polysaccharides. These glycans are the most peripheral molecules (over 200 nm) from the cell surface with very high density and abundance. AFM; Atomic Force microscope image of GN1 with g200 glycan arms of 180 nm. X-ray; model of protein on plasma membrane in blue with small glycans in yellow and large glycan in green which is an order of magnitude longer then presented if the real length of g200 glycan would be taken in account. In the second panel as the example of the second step of structural analyses, a polyacrylamide gel electrophoresis of purified glyconectin glycan fraction is presented. Electrophoresis of glyconectin glycans was performed on a polyacrylamide gradient gel (7.5-15%). Gels were stained with 0.3% alcian blue in 3% acetic acid in aqueous 25% isopropanol. Lane a, 20 g of GN1 glycans; lane b, 20 g of GN2 glycans; lane c, 20 g of GN3 glycans. The third panel shows the third step of structural analyses of glycans by fingerprinting with trifluoroacetic acid hydrolyses. TLC analysis of hydrolyzed fractions of GN1 and GN2 stained by sulfuric orcinol. Lane 1, standard Glc degrees of polymerization (DP); lane 2, 0.1 M trifluoroacetic acid hydrolysis of GN1; lane 3, 0.1 M trifluoroacetic acid hydrolysis of GN2; lane 4, 1 M hydrolysis of GN2. The forth step in structural analyses using NMR and MS sequencing are shown in the forth panel. Complex sequencing procedure in combined NMR and MS complementary approach requires sophisticated instrumentation and high skills.

Fig. 2: A and B. Two putative models of ultrastructural organization of GN1 and GN3 glycans with linear and/or ramified repetitive units (blue circles symbolize Hexose and/or GlcNAc, blue striped squares Fucose). C. Model of GN2 highly ramified repetitive glycan structure (blue circle symbolize Hexose and/or GlcNAc, yellow triangles Py(4,6)Gal), blue striped squares Fucose.

Ara residues to small oligosaccharides that are released by mild acidic hydrolysis. In contrast with GN1 and GN3, mild hydrolysis of GN2 released large oligosaccharides that were further fragmented in smaller units using stronger acidic conditions. Analysis

of both fractions revealed that the acid labile carbohydrate moiety of GN2 comprised a highly ramified polysaccharide backbone. It is constituted by an extremely heterogeneous mixture of hexose (mannose and galactose) oligomeres all terminated by Py(4,6)Gal residues and randomly interrupted

82

Table 1: Glycan structures obtained by NMR and MS after mild hydrolyses of isolated GN1, GN2 and GN3 polysaccharides.

GN1 GlcNAc-Fuc-(SO3)GlcNAc-Fuc GlcNAc-(SO3)Gal-Fuc Gal-(SO3)Gal-GlcNAc-Fuc Py(4,6)Gal-GlcNAc-Fuc (SO3)GlcNAc-[Fuc]Fuc

GN2 Py(4,6)Gal-(Hex)0-1-Fuc Py(4,6)Gal-(Hex)0-3-GlcNAc Py(4,6)Gal-(Hex)1-5 Py(4,6)Gal-(Hex)0-2[Hex]Hex Py(4,6)Gal-[Hex]Hex-Hex Py(4,6)Gal-Hex-[Py(4,6)Gal]Hex (Hex)4-[Py(4,6)Gal]Hex PyGal-Hex-[Hex]HexNAc

GN3 HexNAc-(SO3)Ara/Fuc-Fuc Hex-HexNAc-(SO3)Ara/Fuc-HexNAc-Fuc

by Fuc and GlcNAc residues. The observed heterogeneity of released oligosaccharides did not permit definitive conclusions about the ultrastructural organization of repetitive motifs. In conclusion, structural analyses of GN1, GN 2 and GN3 isolated from three different sponge species revealed that their carbohydrate content ranges between 4060% of their total mass thus characterized them as a heavily glycosilated macromolecules. The physico-chemical properties of each of four major GN glycan (GM1 g200 and g6, GN2 g189 and GN3 g110) such as size, composition, high content of anionic groups (carboxyl and/or pyruvate and/or sulfate), resistance to most glycosaminoglycan degrading enzymes, monoclonal antibodies mapping and their highly repetitive new type of sequences characterized them as novel class of acidic glyconectin type of glycans. Using the above interdisciplinary approach and technologies we have found that also higher invertebrates like sea urchins as well as vertebrates like mammals (rodents and humans) have similar type of glyconectin glycan structures (Papakonstantinou and Misevic 1993, Misevic and Popescu 1995). Therefore, glyconectin carbohydrates can be considered as a new family of species-specific glycans containing different classes of molecules present in Metazoans.

AFM measurements of intermolecular binding strength

Intermolecular binding forces between cell surface molecules are keeping cells together in multicellular organisms. To provide direct and quantitative evidence that glyconectin carbohydrates can indeed support cell adhesion, in 1993 we have developed a novel technology based on AFM measurements of binding strength between glyconectin carbohydrates under the physiological conditions (Dammer et al. 1995). Interactions between individual adhesion molecules (immunoglobulin, selectin, cadherin, integrins and extracellular matrix adhesions) were usually investigated by kinetic binding studies, calorimetric methods, x-ray diffraction, nuclear magnetic resonance and other spectroscopic analyses. These methods do not provide direct measurement of the intermolecular binding forces, which are fundamental for ligan-receptor association related to cell adhesion and recognition. To measure glyconectin to glyconectin interaction forces, we covalently attached glyconectins via their protein part to an AFM sensor tip and a flat mica surface (Fig. 3). The attachment process involved only glyconectin proteins but did not modify functional carbohydrate adhesion sites. As shown in the schematic presentation in Fig. 3 the cantilever tip having attached glyconectin molecules was carefully moved toward the substrate surface and a series of approach-andretract cycles were collected in physiological liquid medium. GN-GN binding was characterized by measuring both the force necessary to separate the GN-functionalized sensor tip from the GN surface (final jump-off) and the percentage of interaction events under different ionic conditions (Dammer et al. 1995). These two indicators of GN activity varied reversibly with the Ca2+ concentration, in agreement with GN-promoted cell adhesion and GN-coated bead aggregation (shown in the following section of functional analyses). At a Ca2+ concentration of 10 mM, the average force between GNs was 125 pN, ranging up to 400 pN, with high probability of binding (60 10%). At a Ca2+ concentration of 2 mM, cell adhesion and GN bead aggregation were sharply reduced, and the force (40 15 pN) and probability of binding (12 5%) were also reduced (Dammer et al. 1995). The interaction between GNs is Ca2+-selective, as reported with a cell aggregation assay. Indeed, 10 mM Mg2+ could not replace Ca2+ in AFM experiments or in adhesion of glyconectincoated beads (Dammer et al. 1995).

Functional measurements of glyconectin glycans self-non-self recognition properties by Atomic Force Microscopy and color coded cell-bead adhesion
For complete understanding of molecular mechanisms guiding evolution of multicellularity, it is necessary to complement structural studies on glyconectin glycans with quantitative functional measurements of cell adhesion and recognition. Such structure to function relationship in Porifera experimental model system was established by taking two complementary approaches. The first one was Atomic Force Microscopy (AFM) measurements of intermolecular binding strength between individual glyconectin glycans under physiological conditions (Dammer et al. 1995). The second one was quantification of color coded cell-cell, bead-bead and bead-cell recognition and adhesion mediated by glyconectin glycans (Popescu and Misevic 1997, Misevic et al. 2004).

83

Fig. 3: Schematic presentation of AFM measurements of intermolecular binding strength between glyconectin 1 carbohydrates.

Use of the monoclonal antibody (mAb) block 2 provided a third line of evidence that the AFM-measured interactions originate from binding between glyconectin glycans. This antibody recognizes a carbohydrate epitope of GN1 (see Table 1) and specifically inhibits GN1-promoted cell adhesion and GN1-coated bead aggregation (Dammer et al. 1995). In AFM experiments, block 2 Fab fragments in 10 mM Ca2+ SWT (Sea Water buffered with 20 mM Tris pH 7.4) reduced the interactive force to approximately the level measured at 2 mM Ca2+. A control mAb did not prevent GN1-GN1 binding under equivalent conditions. Thus, during AFM measurements in all tested experimental conditions, GN1-GN1 interactions resemble cell-cell adhesion events observed in vivo. The shape of the approach-and-retract cycles shows that string-like structures were responsible for GN1-GN1 interactions. These strings are likely to be the GN1 arms composed of glycans with a relative molecular weight of 200 x 103 (g200), which have been shown to mediate polyvalent GN1-GN1 binding (Fig. 3). This possibility is further supported by the fact that the length (180 nm) and the number (20 copies) of the g200 glycan per GN1 molecule are similar to the length and number of GN1 arms as measured by AFM and electron microscopy (Dammer et al. 1995). Finally, the inhibitory mAb block 2 is directed against a selfassociation epitope located on the g200 glycan. The shape of the approach-and-retract curves between glyconectins

suggested the presence of long-range interactions, interpreted as the lifting and extension of string-like glyconectin glycans, followed by further stretching until the elastic force of the cantilever equals the strength of the binding and the lever jumps off. At a physiological Ca2+ concentration of 10 mM in seawater multiple jump-offs were frequently observed, indicating polyvalent binding with an average adhesive force of 40 15 pN per release (Dammer et al. 1995). AFM measurements of intermolecular binding between homotypic pairs of GN2 and GN3 showed that intermolecular binding forces per pair of molecules are, as in GN1, in the range of 400 pN. Heterotypic combination of GNs revealed intermolecular binding strength of 20 pN (detailed results to be published). Similar results were obtained with purified GN glycans confirming the results of intact GNs where only carbohydrate moieties were available for interaction while the protein part was used for immobilization to surfaces (see Fig. 3). Therefore, carbohydrate to carbohydrate interaction is responsible for GN-GN self-non-self recognition and adhesion. In conclusion, measurement of binding forces intrinsic to adhesion molecules is necessary to assess their contribution to the maintenance of the anatomical integrity of multicellular organisms. Our atomic force microscopy results of the binding strength between cell adhesion glyconectin glycans under physiological conditions showed that homotypic adhesive force of 400 piconewtons per molecular

84

Fig. 4: Glyconectin glycoconjugates are cell adhesion and recognition molecules. Ca2+-dependent glyconectin to glyconectin interactions mediate species-specific cellcell recognition and adhesion. A. M. prolifera, B. H. panicea, and C. C. celata living sponges. Shown are self- and non-self-discrimination and adhesion in the suspension of mixed M. prolifera (orange), H. panicea (white), and C. celata (brown) live cells bearing glyconectins. D and E) ACMFSW (Artificial Calcium and Magnesium Free Sea Water) without 10 mM Ca2+ (D) and ACMFSW with 10 mM Ca2+ at 0C after 20 min of rotation (E). The microscopically observed color of the cells is somewhat different from that of the whole sponge. Early cell sorting experiments were usually done with binary sponge combinations at room temperature without rotation. The sorting is thus dependent on the presence of recognition molecules at the cell surface, cell motility, and speed of new synthesis and/or secretion of additional recognition molecules. Our rotary assays using either metabolically attenuated or fixed cells reduce the number of variable parameters.

85

pair could hold the weight of 1600 cells assuring the integrity of the multicellular sponge organism (Dammer et al. 1995). Interaction forces between heterotypic molecules were 20 times lower and are thus not sufficient to sustain existence of heterotypic aggregates under physiological hydrodynamic conditions of natural sea environment. Furthermore, this data also explain why small and loose unspecific aggregation was sometimes observed during the initial stage of heterotypic mixing under mild agitation.

Glyconectin glycans mediate color coded cell and bead adhesion

The cell adhesive function of three sponge glyconectins purified from Microciona prolifera (GN1), Halichondria panicea (GN2) and Cliona celata (GN3) was tested in a rotary reaggregation assay with live metabolically attenuated and/or fixed cells depleted of endogenous GNs. All three glyconectins, at concentrations mimicking in vivo conditions, mediated cell adhesion in the presence of physiological sea water with 10 mM CaCl2, and not below 1 mM CaCl2 (Guerardel et al. 2004, Misevic et al. 2004). In the absence of GNs, independently of CaCl2 concentration, no aggregation could be observed. Magnesium ions could not replace Ca2 titration experiments of Ca2+ concentration dependence of sponge glyconectin self-interactions revealed a transition at 5mM and 100% interactions at physiological 10 mM CaCl2, identical to that of Ca2+ dependent glyconectin promoted cell adhesion (Jumblatt et al. 1980, Dammer et al. 1995). These experiments indicated that a Ca2+ dependent glyconectin to glyconectin interactions play a pivotal role in cell adhesion of the three selected marine sponge species. The specificity of adhesion of GNs bearing cells was tested in a trinary species combination (Microciona prolifera, Halichondria panicea and Cliona celata) with living dissociated and metabolically attenuated cells in artificial sea water at 0C in the presence, and absence of 10 mM Ca2+ (physiological concentration in seawater). In a rotary assay, species-specific recognition and adhesion occurred only with 10 mM Ca2+ within 5-15 min. (Fig. 4). Upon removal of GNs from cell surface by repetitive washing none of the three species displayed aggregation in the presence of 10 mM Ca2+. Adding back the purified GNs to the same live cells at 0C completely restored species-selective cohesion. Similar results were obtained with non-living fixed cells. These experiments indicated that glyconectins and Ca2+ mediate the initial steps of xenogeneic cell recognition and adhesion of the three selected sponge species and were extending previously reported phenomenological and biochemical studies about the role of proteoglycan-like glycoconjugates in binary assays of dissociated sponge cells (Wilson 1907, Galstoff 1925 , Curtis 1962, Humphreys 1963, Moscona 1968, Cauldwell et al. 1973 , Mller and Zahn 1973, Jumblatt et al. 1980). In the second type of recognition assay, we reconstituted the observed cell recognition by using artificial system of glyconectin color coated beads. Glyconectin 1 was attached via its protein part to fluorescent pink, glyconectin 2 to fluorescent green, and glyconectin 3 to fluorescent blue latexamidine beads leaving glycan molecules free for interactions. Unlabeled glyconectins were immobilized on a nitrocellulose

Fig. 5: Simultaneous species-specific glyconectin to glyconectin recognition in suspension and blotting assay. Letters were drawn using 4 l of 1.5 mg/ml glyconectins on a Hybond-C extra nitrocellulose membrane (Amersham Biosciences) and probed in SWT with pink, green, and blue fluorescent beads coated with glyconectin 1, 2, and 3, respectively. A. SWT without 10 mM Ca2+. B. SWT with 10 mM Ca2+. All photographs were taken after 30 min of mixing.

membrane in such a manner that the three molecules were used to draw the subsequent letters of the words GLYCONECTIN RECOGNITION. The three bead types were mixed and added to the coated membrane in the presence of 10 mM CaCl2 or absence of calcium ions. As shown in Fig. 5, within 5-15 min of constant rotation species-specific bead-bead aggregation and homophilic recognition between membrane-bound and beadbound glyconectins were observed through three separate colored aggregates and selective staining of each letter only with 10 mM CaCl2. Both processes occurred at apparently similar rates for each of the three glyconectins. In control experiments with glyconectin 1 separately attached to pink, yellow and white beads, as expected, mixed color aggregates were formed upon addition of 10 mM CaCl2. In the absence of 10 mM CaCl2, bead aggregation did not occur either in the mixture of three glyconectins or of one glyconectin coated to three color beads. Ex vivo color coded cell-bead experiments

86

Fig. 6: Species-specific glyconectin to glyconectin interactions mediate bead-cell recognition and adhesion. Xenogeneic glyconectin selfrecognition in a mixture of glutaraldehyde-fixed cells and glyconectin-coated beads in SWT in the presence of 10 mM Ca2+. M. prolifera cells bearing glyconectin 1 were incubated with: glyconectin 1 (pink beads) (A), glyconectin 2 (yellow beads) (D), and glyconectin 3 (white beads) (G). H. panicea cells bearing glyconectin 2 were incubated with: glyconectin 1 (B), glyconectin 2 (E), and glyconectin 3 (H) color-coded beads. C. celata cells bearing glyconectin 3 were incubated with: glyconectin 1 (C), glyconectin 2 (F), and glyconectin 3 (I) color-coded beads (glutaraldehyde fixation changes cell colors, i.e. M. prolifera, orange to yellowish white; H. panicea, white to yellowish brown; and C. celata, brown to brownish orange. We did not observe differences in adhesion properties between fixed and live metabolically attenuated cells in a rotary assay.

showed that artificial beads covered with glyconectin glycans will co-aggregating in the species-specific manner only with homotypic cells having same glyconectin glycans (Fig. 6). Similar types of experiments were also done with purified glycans from all three species. Results obtained confirmed that self-non-self discrimination of GNs is based on selective carbohydrate to carbohydrate self-assembly (Misevic et al. 1987, Misevic and Burger 1993) which represents a novel mechanism complementary to well studied protein to protein and protein to carbohydrate interactions of adhesion and recognition molecules. Color coded bead experiment was also performed by overlaying agarose gel containing electrophoretically separated three glyconectins with color coated glyconectin beads. As shown in Fig. 7, after overnight incubation at room temperature, in the presence of 10 mM CaCl2 under gentle agitation, species specific staining of gel glyconectin bands identical to ones stained with Toluidine blue and Amido black showed that glyconectin to glyconectin interactions are highly species-specific (Guerardel et al. 2004, Misevic et al. 2004).

The combinations of the above described experiments demonstrate species-specific molecular self-recognition of glyconectins in an elementary reconstituted bead adhesion system which fully resembles glyconectin mediated cellcell recognition and adhesion. Thus, glyconectin glycans mediate self and non-self discrimination via selective glycan to glycan assembly in the initial step of sponge cell adhesion and xenogeneic recognition.

Evolution of the Porifera model system in research

In the second part of the past century, zoology and ecology research on Porifera was highly considered. Unfortunately, the same sponge model system was often neglected in the field of biochemistry and molecular biology. This research was put to the bottom of the list of priory and was classified as risky, marginal and not serious (e.g. comments that if possible this research should be avoided for the sake of the scientists and institutions involved). Fortunately, and in

87

References
Cauldwell CB, Henkart P, Humphreys T (1973) Physical properties of sponge aggregation factor. A unique proteoglycan complex. Biochemistry 12: 3051-3055 Curtis ASG (1962) Pattern and mechanism in the reaggregation of sponges. Nature 196: 245-248 Dammer U, Popescu O, Wagner P, Anselmetti D, Guntherodt HJ, Misevic GN (1995) Binding strength between cell adhesion proteoglycans measured by atomic force microscopy. Science 267: 1173-1175 Galstoff PS (1925) Regeneration after dissociation (an experimental study on sponges) I. Behavior of dissociated cells of Microciona prolifera under normal and altered conditions. J Exp Zool 42: 183221 Guerardel Y, Czeszak X, Sumanovski LT, Karamanos Y, Popescu O, Strecker G, Misevic GN (2004) Molecular fingerprinting of carbohydrate structure phenotypes of three porifera proteoglycanlike glyconectins, J Biol Chem 279: 15591-15603 Humphreys T (1963) Cell aggregation: chemical dissolution and in vitro reconstruction of sponge cell adhesions: I. Isolation and functional demonstration of the components involved. Dev Biol 8: 27-47 Jumblatt JE, Schlup V, Burger MM (1980) Cell-cell recognition: specific binding of Microciona sponge aggregation factor to homotypic cells and the role of calcium ions. Biochemistry 19: 1038-1042 Misevic GN (1989) Immunoblotting and immunobinding of acidic polysaccharides separated by gel electrophoresis. Methods Enzymol 179: 95-104 Misevic GN, Burger, MM (1986) Reconstitution of high cell binding affinity of a marine sponge aggregation factor by cross-linking of small low affinity fragments into a large polyvalent polymer. J Biol Chem 261: 2853-2859 Misevic GN, Burger MM (1990a) Involvement of a highly polyvalent glycan in the cell-binding of the aggregation factor from the marine sponge Microciona prolifera. J Cell Biochem 43: 307-314 Misevic GN, Burger MM (1990b) The species-specific cell-binding site of the aggregation factor from the sponge Microciona prolifera is a highly repetitive novel glycan containing glucuronic acid, fucose, and mannose. J Biol Chem 265: 20577-20584 Misevic GN, Burger, MM (1993) Carbohydrate-carbohydrate interactions of a novel acidic glycan can mediate sponge cell adhesion, J Biol Chem 268: 4922-4929 Misevic GN, Popescu, O (1995) A novel class of embryonic cell adhesion glycan epitopes is expressed in human colon carcinomas. J Mol Recognit. 8: 100-105 Misevic GN, Finne J, Burger MM (1987) Involvement of carbohydrates as multiple low affinity interaction sites in the self-association of the aggregation factor from the marine sponge Microciona prolifera. J Biol Chem 262: 5870-5877 Misevic GN, Guerardel Y, Sumanovski LT, Slomianny MC, Demarty M, Ripoll C, Karamanos Y, Maes E, Popescu O, Strecker G (2004) Molecular recognition between glyconectins as an adhesion selfassembly pathway to multicellularity. J Biol Chem 279: 1557915590 Misevic GN, Jumblatt JE, Burger MM (1982) Cell binding fragments from a sponge proteoglycan-like aggregation factor. J Biol Chem 257: 6931-6936

Fig. 7: Electrophoretic separation of sponge glyconectins. A. 0.75% agarose gel stained with 0.02% toluidine blue followed by 0.1% Amido Black 10B. a-c, GNs from M. prolifera GN1, H. panicea GN2, and C. celata GN3, respectively (10 g each). B. 0.75% agarose gel stained with color-coded fluorescent beads coated with GN1 (pink) (a), GN2 (green) (b), and GN3 (blue) (c) in the presence of SWT with 10 mM CaCl2.

contrast to the expectations of the official representatives of the scientific community, molecular- and cellular-oriented fundamental research on sponges - exemplified in this article by evolution of multicellularity, as well as by other reports in this book - have generated a vast body of knowledge of new structures, novel molecular mechanisms and new nanotechnologies. Consequently this interdisciplinary research on sponges, which integrates biology, physics, chemistry and mathematics, starts to gain deserved respect as measured by the appearance of publications in journals with high impact factors and the citations of these papers, and the level of attendance at the international conferences in the now clearly established interdisciplinary sponge field. In conclusion we are arguing that any model system is valuable if competent scientists can use it to develop and test original ideas to help solve fundamental scientific questions.

Acknowledgments
This work is supported by private funds of Gradimir Misevic, Swiss National Science Foundation, European Union 6th Framework Network of Excellence Nanobeams, CNRS France, University of Rouen France, Rgion Normandy France, University of Lille France, and region Nord pas de Calais.

88

Moscona AA (1968) Cell aggregation: properties of specific cellligands and their role in the formation of multicellular systems. Dev Biol 18: 250-277 Muller WEG, Zahn RK (1973) Purification and characterization of a species-specific aggregation factor in sponges. Exp Cell Res 80: 95-104 Papakonstantinou E, Misevic GN (1993) Isolation and characterization of a new class of acidic glycans implicated in sea urchin embryonal cell adhesion. J Cell Biochem 53: 98-113 Popescu O, Misevic GN (1997) Self-recognition by proteoglycans. Nature 386: 231-232

Spillmann D, Hard K, Thomas-Oates J, Vliegenthart JF, Misevic G, Burger MM, Finne J (1993) Characterization of a novel pyruvylated carbohydrate unit implicated in the cell aggregation of the marine sponge Microciona prolifera. J Biol Chem 268: 13378-13387 Spillmann D, Thomas-Oates JE, van Kuik JA, Vliegenthart JF, Misevic G, Burger MM, Finne J (1995) Characterization of a novel sulfated carbohydrate unit implicated in the carbohydratecarbohydrate-mediated cell aggregation of the marine sponge Microciona prolifera. J Biol Chem 270: 5089-5097 Wilson HV (1907) On some phenomena of coalescence and regeneration in sponges. J Exp Zool 5: 245-258

Porifera research: Biodiversity, innovation and sustainaBility - 2007

89

Porifera: an enigmatic taxon disclosed by molecular biology/cell biology

Werner E.G. Mller(*), Isabel M. Mller
Institut fr Physiologische Chemie, Abteilung Angewandte Molekularbiologie, Universitt, Duesbergweg 6, D-55099 Mainz; Germany. tel.: +49-6131-39-25910; fax.: +49-6131-39-25243. wmueller@uni-mainz.de Abstract: It was a long and painful scientific process to position the most enigmatic metazoan phylum, the Porifera, into the correct phylogenetic place among the eukaryotes in general and the multicellular animals in particular. This well studied taxon, the sponges, was first grouped to the animal-plants or plant-animals then to the Zoophyta or Mesozoa and finally to the Parazoa. Only after molecular biological techniques became available, was it possible to place the Porifera monophyletically with the other metazoan phyla, justifying a unification of all multicellular animals to only one kingdom, the Metazoa. The first strong support came from the discovery that cell-cell and cell-matrix adhesion molecules cloned from sponges (the main work was performed with the demosponges Suberites domuncula and Geodia cydonium) and subsequently expressed, share high sequence similarity/functional identity with the corresponding molecules of other metazoans. Besides these evolutionary novelties for Metazoa, the sponges possess also morphogens and transcription factors in common with the other metazoan phyla. Surprising was the fact that even those elements exist in Porifera which are characteristic for pattern and axis formation (e.g. Frizzled receptor). The cell culture system from sponges, the primmorphs, provided a suitable tool for the understanding of morphogenetic events. Furthermore, stem cell marker genes could be isolated, underscoring that sponge cells have the capacity to differentiate. Hence, we can conclude that the body plan of the Urmetazoa must have been constructed by similar genomic regulatory systems as found now in higher Metazoa. In a relatively short period, within approximately 200 million years, from 700 to 500 million years ago, the basic pathways are established which allowed the transition to complex, individual metazoan systems through the formation of adhesion molecules; based on the development of a complex immune system and of the apoptotic machinery an integrated system, the Urmetazoa, could be reached. Hence, from the insights into the genome repertoire of some representative species within the Porifera it became overt that the ancestor of all multicellular animals, the Urmetazoa, was provided with a surprisingly high complexity of structural and regulatory molecules. These data caused a paradigmatic change; the Porifera are complex and simple but by far not primitive. The different hurdles which had to be taken to state that the Porifera are not Parazoa [alongside animals] or Mesozoa [middle animals], but Metazoa are reviewed here with special emphasis on the contributions coming from molecular biological approaches. Keywords: Evolution; Metazoa; molecular phylogeny; Urmetazoa

Introduction
Sponges are sessile filter-feeding organisms with an extremely effective and complex network of water-conducting channels and choanocyte chambers lined with flagellated choanocyte cells. Until not too long ago the ground space, the mesohyl, between the external pinacoderm and the internal choanoderm [endopinacoderm], the two cell layers that surround sponges, was thought to consist of mostly functionally independent cells (Pechenik 2000). Such a set up would result in the formation of amorphous, unorganized creatures (Pechenik 2000). However, during the last few years the existence of cell surface-bound receptors and their extracellular as well as intracellular segments could be proved and it was concluded that the sponges possess molecules that allow the establishment of a distinct body plan. The discovery of the metazoan novelties first developed during evolution in sponges which comprise the cell-cell/-matrix, signal transduction-, immune-, neuronal- and morphogenetic

molecules helped to overcome the long-standing debate whether sponges are specialized protists or true Metazoa (Hyman 1940). The phylum Porifera is subdivided into three classes, the Hexactinellida, the Demospongiae and the Calcarea. Until very recently, the phylogenetic positions of these classes remained unresolved. Like any other metazoan also sponges have a defined Bauplan; this has most artistically been illustrated by Haeckel (1872). Unlike other Metazoa adult sponges are considered to have no pronounced anterior/posterior polarity; surely they do not have a dorsal ventral axis. In higher metazoans the patterning along the anterior-posterior axis is regulated among others by the famous family of homeobox genes. Homeobox-related genes that have been identified in sponges display, however, a more general function as transcription factors active in all sponge cells (Seimiya et al. 1998). The body plan of metazoans is defined and its orientation is fixed by the inorganic skeleton. In most sponges this

90

solid support, the spicules, is composed of hydrated, amorphous, noncrystalline silica [SiO2/H2O] as in the classes of Demospongiae and Hexactinellida, or of calcium carbonate [CaCO3] as in the class of Calcarea. The secretion of spicules occurs in specialized cells, the sclerocytes. While in Demospongiae silica is deposited around an organic filament, no organic axial structure is found in the spicules from Calcarea. Exciting new data on spicule formation and the enzymatic silicatein-mediated biosynthesis of silica has been summarized recently (Mller et al. 2005, 2006). In the following, the different steps which led to an understanding of the genetic repertoire of sponges and to the tuned expression of structural and morphogenetic molecules that form the basis for sponge individuals and their body plans (Bauplan) are discussed.

Position of the Porifera among the metazoans

The presently used phylum name Porifera dates back to 1836 (Grant 1836) and was originally Poriphera/Poriphora. The impact of sponges in biology was always remarkable, because they are the (major) taxon which can shed light on the evolution and the origin of the Metazoa. There was, however, one obstacle; before 1850 hardly any scientist had seen sponges in their natural environment, especially not in the ocean. After the first extensive expeditions at the end of the 19th century, e.g. the Challenger expedition, this situation changed. In 1876 Campbell (Campbell 1876), wrote in his log-letters, those beautiful glass-rope sponges, Hyalonema etc., have been found by our researchers to be the most characteristic inhabitants of the great depths all over the world, and with them ordinary siliceous sponges, some of which rival Hyalospongiae in beauty. Then Sollas (1888), who worked on the demosponge collection from the Challenger expedition wrote Now that von Lendenfeld has pointed it out nothing can be clearer, and no one, as he remarks, will raise any objection to the statement that the sponges are evidently Metazoa. But nevertheless many spongiologists were still under the impression of earlier works, which claimed that sponges are separate independent entities in systematics and should be looked upon as e.g. plant-animals (Esper 1794), animal-plants (Pallas, 1787) or Mesozoa (DeLage and Hrouard 1899). Today it should be accepted that this phylum must be grouped to the Metazoa, being qualitatively identical to the higher Metazoa [or, if pressed, to the Eumetazoa] and having only quantitatively different characters as other metazoan phyla (Mller 2001). In the Proceeding of the Second Sponge Symposium from 1978, Rasmont (1979) wrote At least five main differences can be recognized between the development of sponges and that of other Metazoa; among them morphogenetic interactions between sponge cells depend largely on long-ranged chemical messengers and the genetic individuality appears to be very different of that of other Metazoa. Around the year 1990 it became obvious that a further clarification of the phylogenetic position and the (potential) richness of the sponge morphological and functional characters would not be possible unless the basic genomic regulatory systems were known. As a consequence, sequencing of ribosomal genes was introduced to obtain more phylogenetic information. The 5S rRNA proved to be little reliable (Doolittle

and Brown 1994), while data from 18S rRNA as well as 28S rRNA molecules allowed phylogenetic inference (Halanych 1991). However, this approach revealed conflicting results. Based on sequence data of 18S rRNA alone both a polyphyletic (Field et al. 1988) and a monophyletic origin of Metazoa could be concluded (Christen et al. 1991). As reasons for this controversy, the following possibilities were claimed; (i) the molecular phylogenetic methods, basing on sequence data of rRNA, have reached their limits, (ii) there is a hidden paralogy, and/or (iii) lateral gene transfer is responsible. The introduction of molecular cloning of genes coding for informational proteins, increased in a rapid, self-accelerating manner our knowledge on the evolution to the Porifera; it furthermore supported the subdivision of this phylum into the three classes, Hexactinellida, Demospongiae and Calcarea. It is now clear that the Porifera must be grouped together with the other metazoan phyla into one monophyletic unit (Mller et al. 1994a, Mller 1995). Consequently the hypothetical ancestor of all metazoan phyla was named Urmetazoa (Mller 2001, Mller et al. 2001). Even though the present-day data on the strikingly high similarity/homology of sponge proteins with related molecules from evolutionary younger metazoan phyla are overwhelming, the earlier view of the sponge origin can still be found in the literature (e.g. Pechenik 2000). In the following, the different steps in the elucidation of the beauty of the sponge body plan, based on a complex genetic network, with the main focus on demosponges (mainly by using Suberites domuncula (Olivi, 1792) and Geodia cydonium (Jameson, 1811)) are outlined. The discussion is restricted primarily to results which are based on molecular and cell biological studies, because data obtained with cytological/ morphological techniques proved to be insufficient. The basic turning point to the modern integrating view of sponge biology/ body plan started with the paper of Mller et al. (1994a [Fourth Sponge Symposium in 1993]), describing the molecule [lectin] that for the first time indicated a monophyly of all metazoan animals, including the Porifera. Another publication of Mller et al. (1994b [3rd Brazilian Symposium on Extracellular Matrix, Angra dos Reis (Brazil), 1994]) formulated this hypothesis.

Cell-cell and cell-matrix adhesion in sponges (1973)

Sponges have long been used as classic model for basic studies to understand metazoan cell-cell adhesion. Wilson (1907) introduced this system to experimental biology where it then became a traditional model to study both cell-cell and cell-matrix adhesion (reviewed in: Burger and Jumblatt 1977, Mller 1982, Mller et al. 1988). The marine demosponges Microciona prolifera Bowerbank, 1862, G. cydonium and S. domuncula have been most thoroughly studied. In 1973 the first extracellular particle, termed aggregation factor [AF], which promotes the species-specific aggregation of sponge cells, had been described from G. cydonium (Mller and Zahn 1973) (Fig. 1). Shortly after the second AF was identified in M. prolifera (Henkart et al. 1973). The AF was characterized as high-molecular weight complexes with a sedimentation coefficient of 90S which are assembled from a series of proteins that are covalently and noncovalently bound to the core structure (summarized in: Mller 1982) (Fig. 1A and B). In its native form the AF appears as a sphere with

91

localized on the cell periphery. These systems are prerequisites for the establishment and stabilization of the functional arrangement of cells in the organism (Mller 1982).

Monophyletic origin of all metazoan phyla (1993)

Phylogenetic relationships especially with regard to the phylum Porifera had been formulated, but often uncertainties remained because of difficulties to distinguish between convergent or divergent features. During the 4th International Porifera Congress in 1993, we described the first protein-coding sequence, a galectin from the sponge G. cydonium; galectins exist only in Metazoa (Mller at al. 1994a). In parallel, it was stated/reported that sequencing of the 18S rRNA gene is not suitable to resolve deep branches in the phylogenetic tree and the data (rRNA-sequencing data) fail to support the monophyly of sponges and other Metazoa stronger than the monophyly of sponges and plants (Rodrigo et al. 1994). Our approach relied on amino acid [aa] sequence data obtained from genes identified mainly from the marine sponges G. cydonium and S. domuncula. We have selected mainly genes that code for proteins which are features of multicellularity, such as (i) extracellular adhesion molecules (galectins) with their typical carbohydrate binding site [LH(F) NPR~(G)~V~NT~(G)~W~(T)E~FPF], found in all vertebrate S-type lectins, and also found in the sponge lectin (Pfeifer et al. 1993), (ii) cell surface receptors, like integrin (Pancer et al. 1997) and speract receptors Scavenger Receptor, CysteineRich domain (SRCR) containing receptors (Blumbach et al. 1998), (iii) signal transduction molecules, like the receptor tyrosine kinase [RTK] (Schcke et al. 1994a), the serine/ threonine [Ser/Thr] kinases (Kruse et al. 1997, 1998), or (iv) transcription factors, the serum response factor (Scheffer et al. 1997). These molecular biological and cell biological findings have been taken as one major clue for the, now established, view that Metazoa, including Porifera, are of monophyletic origin, with the Urmetazoa as the hypothetical ancestor (Fig. 3). Based on these sequence data it was reasonable to place the Porifera into the kingdom Animalia together with the (Eu)Metazoa (Gamulin et al. 1994, Mller et al. 1994b, Mller 1995, Mller [ed.] 1998a, 1998b, Mller 1998c). This monophyletic origin of all metazoan phyla was subsequently often confirmed. Especially by taking sponge genes that code for RTK it is now established that modular proteins that were composed by exon-shuffling, are common to all metazoan phyla.

Fig. 1: Cell adhesion system in sponges. The aggregation factor [AF] from G. cydonium. A. Electron micrograph of the native AF; preparation shadowed with platinum. B. Electron micrograph of the core structure AF; shadowed with platinum. C. Working model in 1976 for the AF-aggregation receptor (AR)-mediated cell recognition in G. cydonium. The sunburst structure comprises the core of the AF (after Mller 1976).

a diameter of 1000 and a concave cup structure (Fig. 1A). Treatment of the AF with detergents yields a core structure that appears as sunburst with a circular center (diameter 1,000 ) and 25 radiating arms (Fig. 1B). In the presence of Ca2+ the AF interacts with a membrane component, termed aggregation receptor [AR]. A 140 kDa polypeptide was found to participate in the AF-mediated reaggregation process. This polypeptide interacts with a galectin that links individual AF molecules to the AR at the plasma membrane and consequently bridges two cells together (Fig. 1C). The major obstacle to the identification of the molecules participating in the very complex and dynamic cell-cell and cell-matrix recognition in sponges was the fact that the underlying molecules involved had not been obtained by molecular cloning. Even until 1994 it remained uncertain if the sponge adhesion molecules display high sequence relationship to functionally related molecules in higher Metazoa (Gamulin et al. 1994). With cloning of a galectin sequence (Pfeifer et al. 1993, Mller et al. 1997) as the first cell-cell adhesion molecule, and the integrin sequence as the first cell-matrix adhesion receptor in G. cydonium (Pancer et al. 1997, Wimmer et al. 1999a), it became overt that sponges contain molecules highly related to those known to also promote adhesion in Protostomia and Deuterostomia (Fig. 2). These molecular biological data confirmed earlier cell biological observations that sponges have developed a number of recognition systems

Evolutionary age of the Porifera (1994)

Also in 1994 (Schcke et al. 1994b) it was possible, based on the number of non-synonymous substitutions in sponge sequences, to estimate the time when the sponges diverged from the common ancestor of all metazoans; approximately 650-665 million years ago (MYA). Although the Porifera are the oldest Metazoa by fossil documentation, their earliest record is only from the Vendian about 600 MYA (reviewed in: Mehl et al. 1998). Hence, it still leaves us with a large stratigraphical gap down to the assumed age of the monophylum Metazoa,

92

Fig. 2: Different stages of morphogenesis in sponges (schematic model based on studies with the demosponges G. cydonium and S. domuncula). First, cell-cell and second, cell-matrix adhesion molecules allow the mechanical interaction between adjacent cells or cells and extracellular matrix molecules. AF in the extracellular space interacts with the AR and mediate the first mechanical contact. The AF-mediated cell-cell recognition is species-specific and very likely controlled by the complex SRCR/SCR-AR, that is embedded in the plasma membrane. Third, after this primary AF-AR-AF contact intracellular signal transduction pathways are activated, resulting in a selective gene expression. New insertion of adhesion receptors into the plasma membrane follows; e.g. integrin and other receptors, involved in tissue and skeleton formation. Together with these receptors their ligands are synthesized, e.g. collagen, molecules containing the fibronectin FN3 modules, or mucin-like molecules, which establish the cell-matrix adhesion system. During this phase growth factors are synthesized, e.g. the precursor and the mature epidermal growth factor (EGF), which interact with the newly synthesized receptors. Fourth, solute molecules are released which initiate axis formation. Fifth, after completion of these phases pattern formation can start, a process which is controlled by morphogenetic cell-surface receptors, like Frizzled, and by transcription factors, e.g. Forkhead. Also LIM-class homeodomain transcription factors are activated.

800 MYA, from which the Porifera are supposed to be the first group that split off. In molecular evolution an unsolved question is the validity of the molecular clock hypothesis which implies that the rate of molecular evolution is nearly constant per year among different evolutionary lineages. Consequently, this rate should be linked with the mutation rate and hence is closer correlated with the number of generations per unit time than to time itself. However, the standard value for the average rate of nonsynonymous substitutions in DNA can vary between 0.6 and

0.9 x 10-9 per site and year. Therefore, the calculated time of divergence of the different phyla might vary as well. Applying these rates of non-synonymous substitutions to the conserved amino acid (aa) stretches in the tyrosine kinase domains of the RTK and the sequences of two sponge S-type lectins, it could be estimated that the sponge molecules diverged from the common ancestral gene approximately 800 MYA. This figure is in good agreement with the generally accepted time of divergence found by traditional methods (Reitner and Wrheide 2002).

93

Fig. 3: Phylogenetic position of the Porifera between the Urmetazoa and the Urbilateria. The major evolutionary novelties which have to be attributed to the Urmetazoa are those molecules which mediate apoptosis and control morphogenesis, the immune molecules and primarily the cell adhesion molecules. The siliceous sponges with the two classes Hexactinellida and Demospongiae emerged first and finally the Calcarea, which possess a calcareous skeleton, appeared. These three classes of Porifera are living fossils that provide a reservoir for molecular biological studies. The Archaeocyatha, sponge related animals with a calcareous skeleton, became extinct. The Calcarea are very likely a sister group of the Cnidaria. From the latter phylum the Ctenophora evolved which comprise not only an oral/aboral polarity but also a biradial symmetry. Finally the Urbilateria emerged from which the Protostomia and the Deuterostomia originated. Very likely the Urmetazoa emerged between the two major snowball earth events, the Sturtian glaciation (710 to 680 MYA) and the Varanger-Marinoan ice ages (605 to 585 MYA). In the two poriferan classes Hexactinellida and Demospongiae the skeleton is composed of amorphous and hydrated silica, while the spicules of Calcarea are composed of Ca-carbonate. The latter biomineral is also prevalent in Protostomia and also in Deuterostomia. In vertebrates the bones are composed of Ca-phosphate [apatite]. The autapomorphic character for the Demospongiae is the spicule-synthesizing enzyme silicatein.

Intracellular signal transductions (2003)

With the increasing awareness of monophyly of all metazoans several pressing questions arose. The major enigma in development is pattern formation. Detailed investigations led to an understanding of the genetic networks constructing and controlling body plan formation in metazoan crown species and recently also answers for the body plan of sponges could be given (see: Mller 2005); most of these studies were performed with G. cydonium and S. domuncula. The structural and functional molecules were cloned and expressed. The cellcell and cell-matrix adhesion molecules found in sponges share amazingly high sequence and functional similarity to those of higher Metazoa (Fig. 2). The extracellular binding sites to

the ligands as well as the intracellular domains of these cell membrane receptors remained conserved throughout the animal kingdom. Functional studies proved that the receptors are provided with the properties of outside-in signaling (Wimmer et al. 1999b). One major step forward was the development of the primmorph system, a technique to grow sponge cells in culture (Mller et al. 1999a). With application of this system, solute morphogenic factors (e.g. myotrophin; Schrder et al. 2000), or secreted molecules (e.g. epidermal growth factor; Perovi-Ottstadt et al. 2004), as well as their receptors, which are involved in axis formation (Frizzled receptor; Adell et al. 2003b), and transcription factors that are required for polarity formation (e.g. the organizer-specific factor LIM homeobox

94

protein; Wiens et al. 2003b, or Forkhead; Adell et al. 2003a) were discovered.

The phylogenetic relationship of the sponge classes (1997)

It is generally agreed that multicellularity in Plantae, Fungi and Metazoa arose in the Proterozoic approximately 1,000 MYA. As the earliest Metazoa the Porifera evolved with the major taxa Hexactinellida, Demospongiae and Calcarea. The skeletal elements of the sponges, the spicules are composed in Demospongiae and Hexactinellida of hydrated, amorphous, noncrystalline silica while those of the class Calcarea are formed from calcium carbonate. The Demospongiae and Hexactinellida are the only two taxa within the kingdom of Metazoa which utilize silica instead of calcium in their mineral skeleton; calcium is otherwise the dominant inorganic skeletal component. According to fossil records, the Hexactinellida appeared first while Demospongiae and Calcarea developed later (Reitner and Mehl 1995). It can be argued that Porifera might not have been the first metazoan phylum which evolved; however, they are the only still extant witnesses of an evolutionary step that occurred during the maturation of the Metazoa, near the Proterozoic-Phanerozoic boundary close to 800 MYA. In this respect they have been considered as living fossils (Mller 1998c). Our group has analyzed genes of sponges, the Demospongiae S. domuncula and G. cydonium, the Calcarea, Sycon raphanus (Schmidt, 1862), as well as from the Hexactinellida, Aphrocallistes vastus (Schulze, 1887) and Rhabdocalyptus dawsoni (Lambe, 1892) in order to obtain an insight into the genome organization as well as the function of genes coding for functional proteins (Fig. 3). Hexactinellids are syncytial rather than cellular; approximately 75% of the tissue forms a multinucleate syncytium while the remaining tissue consists of uninucleate cells which are connected to the syncytium by perforated (aqueous) plugged junctions (Leys 1995). In two early approaches to resolve the phylogenic relationships of the three sponge classes (Schulze 1887, von Lendenfeld 1889) the Calcarea were considered to form the basis of the sponge taxon. While Schulze (1887) suggested that the Hexactinellida represent together with the Calcarea the earliest classes, von Lendenfeld (1889) proposed that the Demospongiae group together with the Calcarea and the Hexactinellida appeared later in evolution. More recent cladistic analyses including morphological data suggested that Porifera are monophyletic. However, for respect to the phylogenetic relationships among the sponge classes two contrary views have been presented. One considers cellular sponges, the Calcarea and the Demospongiae, as the sister group of the syncytial sponges, the Hexactinellida (Mehl and Reiswig 1991), while in the opposite view the siliceous sponges, Hexactinellida and Demospongiae, form the sister group to the calcareous sponges, the Calcarea (Bger 1988). Again, the sequencing data from rDNA were not conclusive enough to solve this question (Medina et al. 2001). On the other hand, studies on several sequences of informative molecules from the three poriferan classes, as well as detailed analyses of housekeeping proteins, e.g. heat shock

protein (Koziol et al. 1997) or -tubulin (Schtze et al. 1999), and of proteins involved in signal transduction, e.g. Ser/Thr kinases (Kruse et al. 1997 and 1998) or calmodulin (Schtze et al. 1999), revealed that among the three classes of Porifera the Hexactinellida are the phylogenetically oldest taxon, while the Calcarea is the class closest related to higher metazoan phyla (Fig. 4). Furthermore, data especially from studies with a series of Ser/ Thr kinases supported the position of Calcarea as sister group to higher metazoan phyla (Mller et al. 1998). The branching order originating from ancestral unicellular eukaryotes via Viridiplantae-Fungi to Porifera, the simplest metazoans, follows both the published fossil data and the sequence data obtained (Fig. 4). Since the Hexactinellida are syncytial animals and since our analyses indicate that they branched off from a common ancestor earlier than other sponges, this could imply that multicellularity came about by the division of a multinucleate syncytium rather than by aggregation of formerly single cells. However, it is also possible that the syncytial nature of the Hexactinellida reflects a reduction of a previously multicellular stage by fusion to form a syncytium. Additional support for an early origin of the Hexactinellida comes from paleontological evidence, which shows that the Hexactinellida were present in the Early Cambrian while the Calcarea and the Demospongiae arose later, in the Middle Cambrian (Reitner and Mehl 1995). The oldest sponge spicules found to date (approximately 600 MYA) come from China, and are mainly monaxonal spicules, but some also have definite crosses, which are typical of triaxones from hexactinellids. These data show that within the phylum Porifera, the class Hexactinellida diverged first from a common ancestor; while Calcarea and Demospongiae appeared later; the Calcarea are hence the sister group to the Cnidaria [paraphyletic relationship].

Primmorphs the sponge in vitro cell culture system (1998)

Sponge cells can be cultivated in vitro as primmorphs (Custdio et al. 1998, Mller et al. 1999a); they can be defined as a special form of in vitro cell culture, which allows the formation of a three-dimensional organization of aggregates that contain proliferating and differentiating cells. We have focused on the formation of primmorphs from the demosponges S. domuncula, Dysidea avara (Schmidt, 1862) (Mller et al. 2000) and occasionally G. cydonium. For this technique, sponge tissue samples are separated into single cells under shaking in Ca2+ and Mg2+-free artificial seawater containing EDTA. The cells are transferred after washing into medium, supplemented with Ca2+, Mg2+ and antibiotics. Immediately after transfer, the single cells form small ~20 cells containing aggregates which grow in size during the subsequent three days to 1,000 m large cell clumps. After usually five days primmorphs are formed. As the basal medium natural seawater is enriched to 0.2% with RPMI 1640medium. This innovative step has been patented (granted US patent 6,664,106). Primmorphs are characterized by the presence of proliferating cells and a distinct histology. The amount of DNA-synthesizing/ proliferating cells present in primmorphs reaches 20% to 30% depending on the age of the primmorphs. The diameter of the

95

Also the iron concentration, as Fe(+++), should be increased to 30 M for optimal growth conditions of primmorphs. One further growth promoting protein has been isolated from S. domuncula which was shown to stimulate proliferation of sponge cells; the myotrophin-like polypeptide (Schrder et al. 2000). Recombinant sponge myotrophin stimulated the overall protein synthesis by 5-fold (Schrder et al. 2000). Besides these chemical factors physical factors such as water current are critically important for primmorph survival and cell growth (reviewed in: Schrder et al. 2003). Incubation of the primmorphs under such conditions resulted in the formation of canals in the primmorphs and the expression of the homeobox gene Iroquois (Perovi et al. 2003).

Immune system (1998)

Fig. 4: Branching order of the three poriferan taxa at the basis of the Metazoa. During the transition from the common ancestor with the Fungi, the Urmetazoa evolved from which the Hexactinellida branched off first, followed by the Demospongiae. The third sponge taxon, the Calcarea, appeared which forms the sister group to the Urbilateria. This branching order, now well established, was first elucidated on molecular terms using representative protein sequences encoding the protein kinase C (PKC; the catalytic domain has been used for the analysis; Kruse et al. 1997). Sequences from the following organisms have been used: Metazoa cPKC from the deuterostomes Xenopus laevis [frog - cPKC_XL] and Lytechinus pictus [sea urchin - cPKC_ LP], from the protostomes cPKC from Drosophila melanogaster [fruit fly - cPKC_DM] and Aplysia californica [mollusc, cPKC_AC], also from Caenorhabditis elegans [cPKC_CE] and those from the sponges of the classes (i) Demospongiae, G. cydonium [cPKC_GC] and S. domuncula [cPKC_SD], (ii) Calcarea, S. raphanus [cPKC_SR], and (iii) the Hexactinellida, R. dawsoni [cPKC_RD] as well as from the yeast Saccharomyces cerevisiae [PKC_SC]. The phylogenetic tree was constructed on the basis of aa sequence alignment. Computing of the sequences by using the procedure of neighbour-joining applying the Neighbor program from the PHYLIP package PROTPARS [Protein-Parsinomy] (Felsenstein 1993) as described (Kruse et al. 1997).

cell aggregates increases steadily after approximately three days incubation; after a total treatment/incubation for five days primmorphs are formed from cell aggregates. During the phase of primmorph formation the aggregates contract to round 1 to 5 mm large bodies, leaving behind detritus and dead cells. In the initial phase the primmorphs remain round-shaped but after incubation of longer than three to four weeks many of them attach to the bottom of the culture dish. Microscopic analysis of the sections through primmorphs revealed that the cells present in the interior are surrounded by an almost complete single-cell layer of epithelial-like cells. The cells that compose the squamous epithelium of the primmorphs are pinacocytes as judged from their flattened, fusiform extensions and their prominent nucleus. The cells inside the primmorphs are primarily spherulous cells while the others may be termed amoebocytes and archaeocytes. Growth conditions could be optimized by supplementing the natural seawater/0.2% of RPMI1640medium with silicate. Natural seawater contains < 5 M silicate; however, the optimal concentration of silicate for cell proliferation and spicule formation is 60 M (Krasko et al. 2000, 2002).

The important contribution of Metchnikoff (1892) was the description of the phagocytotic activity of sponge cells, archaeocytes, as a mechanism to eliminate non-self particles and, even more advanced, to encapsulate the foreign material within cell aggregates of the sponge prior to the elimination by ablation. These abilities of sponges were discussed by Metchnikoff in the context of inflammation processes, proceeding in Metazoa during infection. The major step in the elucidation of the cellular mechanisms by which the sponges eliminate non-self and accept self came from elegant experimental transplantation studies. Smith and Hildemann in their extensive review (Smith and Hildemann 1986) have grouped sponge alloimmune responses into two major rejection processes. Some species may form barriers to separate from non-self tissue; e.g. the marine sponge Axinella verrucosa (Esper, 1794) (Buscema and van de Vyver 1983) or the freshwater sponge Ephydatia muelleri Lieberkhn, 1855, while others may react by cytotoxic factors which destroy the transplant; e.g. the marine sponges Callyspongia diffusa Sollas, 1885 (Hildemann et al. 1979) or G. cydonium (Pfeifer et al. 1992). The breakthrough in the discovery that immune mechanisms in sponges are highly similar to those, found in other metazoan phyla, came again after application of molecular biological techniques (see: Mller et al. 1999b). Defense against microbes/parasites: Almost all marine demosponges contain bacteria. Until now no conclusive data are available to say which bacterial strains are symbiotic and which are parasitic. One report at least suggests that the number of bacterial strains that are symbiotic or commensalic is limited (Althoff et al. 1998). All specimens of the marine demosponge Halichondria panicea (Pallas, 1766) collected from the Baltic Sea, the North Sea, as well as in the Mediterranean Sea were found to harbor one defined bacterium which belongs to the taxon of Roseobacter/Rhodobacter. Based on this finding it was suggested that this bacterial strain accepts at least a commensal relationship with the host. First data on the molecular mechanism by which the host (sponge) might discriminate between symbiotic or commensalic and parasitic bacteria have been obtained. We could demonstrate that defined bacterial strains can be engulfed by specific sponge cells, the bacteriocytes (Bhm et al. 2001). Furthermore, protein synthesis in tissue from S. domuncula is inhibited after incubation with the bacterial endotoxin lipopolysaccharide (LPS; Bhm et al. 2001). Since

96

Ser/Thr directed mitogen-activated protein (MAP) kinases are essential components of the LPS-mediated pathway, evidence of activation of these kinases in response to LPS was sought. Molecular biological and immunological studies confirmed that these pathways also exist for the Porifera, indicating that such defense pathways are highly conserved between sponges and humans (Bhm et al. 2001). One powerful tool to eliminate microbes is intracellular digestion. This cellular defense mechanism against foreign invaders is well developed from Porifera to insects and humans. Sponges possess specialized amoeboid cells, the archaeocytes (Metchnikoff 1892), which have in the past been considered as macrophages of sponges. Mammalian macrophages are the first cells to encounter non-self material. They express several receptors, scavenger receptors, that bind to bacteria or their constituents, and hence act as key molecules in innate immunity. Among them is the type I macrophage scavenger receptor which comprises highly conserved SRCR domains. With regard to sponges, molecules comprising SRCR domains have been discovered first in G. cydonium (Blumbach et al. 1998). Data strongly suggest that sponges comprise SRCRdomain containing cell-surface molecules which might be involved in the recognition of bacteria. In addition, it is likely that the ingested non-self bacteria are killed by both an oxidative and a nonoxidative (enzymatic) mechanism. Several cDNAs coding for lysosomal enzymes, have been isolated from sponges as well. Very recently, a further (putative) defense system against invading bacteria and/or viruses has been detected in Demospongiae: the (2-5)oligoadenylate synthetase [(2-5)A synthetase] system. In mammalian organisms, the (2-5)A synthetase(s) catalyzes the synthesis of a series of 2-5-linked oligoadenylates, termed (2-5)A [= pppA(2p5A)n [pnAn]] from ATP (Hovanessian 1991). In turn, (2-5)A acts as an allosteric activator of a latent endoribonuclease, the RNase L, which degrades single-stranded viral or cellular RNA. In mammalian organisms the (2-5)A system is activated by interferons. The first sponge species studied that was found to display higher levels of (2-5)A oligoadenylate synthetase and its products than vertebrate cells (Kuusksalu et al. 1995) was G. cydonium. The sponge (2-5)A synthetase was cloned (Wiens et al. 1999). This enzyme as well as its products are present in sponges and in the deuterostome lineage, but absent in protostomes. Recently, functional assays elucidated the role of the (2-5)A synthetase in sponges, especially with respect to a potential infection with foreign, pathogenic microorganisms. The sponge cellular system, which proved to be suitable for this approach, are the sponge primmorphs. The experiments showed that primmorphs synthesized (2-5)A in larger amounts if they were incubated with LPS, suggesting an activation of the synthetase through a LPS-initiated pathway (Grebenjuk et al. 2002). Histo(in)compatibility responses in sponges on tissue level: Studies of histo(in)compatibility response in sponges have been performed for 30 years. Initially it was reported that sponges have only a low capacity for allorecognition (Moscona 1963). However, after defining the system, it became apparent that sponges have a very high degree of precision when discriminating between self/self and self/nonself. We applied two transplantation techniques for our studies:

the insertion technique for G. cydonium and the parabiosis method. From G. cydonium tissue, pieces were removed with a cork borer from one specimen and were inserted into holes in the recipients (insertion technique) (Pancer et al. 1996; reviewed in: Mller and Mller 2003). All autografts fused and eventually no boundary line was seen; in contrast allografts initially fused together, but after approximately 3 to 5 days the rejected graft tissue formed a pronounced demarcation boundary and underwent apoptotic/necrotic degeneration and finally resorption. Histo(in)compatibility responses in sponges on cellular level: A cellular assay was developed to allow analysis of the histo(in)compatibility reactions on a cellular level (Mller et al. 2002). The basis of the assay was developed following the establishment of the primmorph system. In the mixed sponge cell reaction (MSCR) assay dissociated cells either from the same individual (autogeneic MSCR) or from different individuals (allogeneic MSCR) were mixed at equal cell concentrations. If cells from the same individual were mixed, autogeneic MSCR, 2 mm large aggregates were formed during the initial two days of incubation, which finally became 5 to 10 mm large primmorphs. In assays using cells from different specimens, they did not form single primmorphs but separated after two days, indicating that during the allogeneic MSCR the cells recognize non-self and form individual-specific aggregates. Molecules involved in histocompatibility response of sponges: Using transplantation models from both G. cydonium and S. domuncula (Mller et al. 1999b) it was established that macrophage-derived cytokine-like molecules are activated during allograft rejection. Among those sponge cytokines activated is the allograft inflammatory factor 1, a factor which has been described in rats and was identified as a cytokineresponsive macrophage molecule. The cDNA encoding the putative allograft inflammatory factor 1 (AIF-1) like molecule from S. domuncula has been cloned (Kruse et al. 1999). A strong upregulation has been determined in the rejection zone from allografts (Kruse et al. 1999). In parallel with this change in expression, a second characteristic molecule was identified which resulted in increased expression of the Tcf-like transcription factor (TCF) after transplantation in S. domuncula (Mller et al. 2006). Also the sponge TCF polypeptide shares highest similarity to those protostome and deuterostome transcription factors that are involved in diverse developmental processes. Further molecules/factors very likely involved in histo(in)compatibility reactions are the glutathione peroxidase and the endothelial-monocyte-activating polypeptide (EMAP). In vertebrates EMAP (type II) causes cell activation and expression of adhesion molecules in endothelial cells as well as in monocytes and granulocytes from human and mouse resulting in angiogenesis. The putative EMAP-related polypeptide was cloned from the marine sponge G. cydonium; it has a deduced molecular mass of 16 kDa and shows high sequence similarity (again) to the human and murine EMAP (Pahler et al. 1998). The glutathione peroxidase (GPX) is activated in humans/ vertebrates during the early phases of inflammation that occur during graft recognition or during wound healing in mammals when reactive oxygen species (ROS) are formed. It is the

97

major enzyme involved in the detoxification of ROS during these processes. The cDNA encoding the putative sponge GPX is known from S. domuncula (Kruse et al. 1999). As in the previous experiments using the AIF-1 like molecule from S. domuncula, the expression of the gene encoding the GPXrelated protein was low in the controls. However in the zones between grafts (the attachment zones), the expression of the S. domuncula GPX increased gradually with time, and reached a maximal level of 6.5-fold. This finding suggested again that during graft fusion and rejection in sponges, ROS are generated which amplify the immune response, as they do in cytokineactivated macrophages in vertebrates. Finally a pre-B-cell colony-enhancing factor was found in S. domuncula (Mller et al. 1999b). In the primmorph system of S. domuncula, the expression of the gene encoding this cytokine-like molecule increased after exposure of the cells to membranes from another species, such as those from G. cydonium. This indicated further that sponges have a molecular mechanism for the recognition of non-self. Molecules in sponges comprising polymorphic Ig-like domains: The most striking similarity between molecules involved in the human adaptive immunity and sequences isolated from G. cydonium are among those which contain immunoglobulin (Ig)-like domains, the receptor tyrosine kinase (RTK) and the sponge adhesion molecules (SAM). The G. cydonium RTK molecule possesses in the deduced polypeptide structure two complete Ig-like domains (Mller and Schcke 1996). Two other SAM species have been found which do not encode a tyrosine kinase but also contain in the extracellular part two Ig-like domains (GC-SAM) (Blumbach et al. 1999). The Ig-like domains found in GC-SAM long form (L) and GC-SAM short form (S) as well as in the RTK display high sequence similarity to the V domain of mammalian immunoglobulin domains (Blumbach et al. 1999). Studies with the two G. cydonium genes GC-SAML and GC-SAMS were performed during auto- and allografting; the results revealed that those genes undergo a differential expression (Blumbach et al. 1999).

have been identified in G. cydonium and/or S. domuncula (Wiens et al. 2002, 2003b).

Stem cells (1966/2003)

By definition, stem cells (i) have the capacity of selfreplication and (ii) give rise to more than one type of mature daughter cells. In general, three levels of stem cells can be distinguished, (i) totipotent stem cells that can rise to an intact organism, including the germinal tissues, (ii) pluripotent stem cells are capable to give rise to cells, derived from the germ layers, and (iii) multipotent stem cells that give rise to a single organ/tissue system. In sponges, the plasticity of the differentiation states of cells dominates over a pointof-no-return differentiation. While Borojevic (1966) provided experimental data indicating that the origin of the differentiation paths to the terminally differentiated cells starts with archaeocytes, Diaz (1977) proposed that the choanocytes have the capacity to differentiate to the archaeocytes. The presently accepted hypothesis is that the archaeocytes, present in early embryos or gemmules, are the totipotent stem cells. In the freshwater sponge Ephydatia fluviatilis Linn 1758 the skeletal cells, sclerocytes, very likely originate from archaeocytes (Weissenfels 1989). The fate of the sclerocytes is unknown; degeneration as well as a movement, away from the spicule, has been discussed (Bergquist 1978, Simpson 1984). For a series of sponge species it could be shown that oocytes originate from archaeocytes (see: Witte and Barthel 1994). These findings imply that both types of cells, the archaeocytes and the germ cells have the same stem cell propensity; they are totipotent stem cells. For this reason we group here the sponge cells only to totipotent stem cells, including the pluripotent stem cells, and the multipotent progenitor cells. Through the multipotent stem cell stages the sponge cells proceed to the terminally differentiated, somatic cells (see: Fig. 5). The terminally differentiated cells, the collencytes/sclerocytes and the myocytes, very likely undergo cell death, after formation of the structural elements they produce the spicules and the fibrils (Mller et al. 2005). At present the study of embryonic stem cells in sponges is limited, since no technique to induce mass production of embryos under controlled conditions has yet been successful. As a substitution, the three-dimensional cell culture has been established for S. domuncula. Under suitable conditions dissociated, single cells form special types of cell aggregates, the primmorphs. They contain cells of high proliferation and differentiation capacity. Gene expression pattern of archaeocytes (stem cells); reproductive cells: Until recently it was not possible to define cell types in sponges in a strict manner. Now molecular markers have been worked out, allowing also the distinction of the different levels of stem cells. Stem cells are self-renewing populations of cells that undergo symmetric and/or asymmetric divisions either to self-renew or to differentiate. This minimal definition does not allow a clear distinction of stem cells from other dividing and differentiating cells. Recently genetic expression markers have been identified, which can be applied for the identification of embryonic cells and tissue in sponges.

Apoptosis (2000)
Apoptosis represents the morphological manifestation of programmed cell death and paradoxically it is a prerequisite for metazoan life. Thus, apoptosis is responsible for the demise of cells during many physiological processes. Obviously apoptosis must be regulated by a complex network of various molecular signaling pathways. Research during the past 20 years has led to the identification of major functional groups of molecules involved in apoptotic pathways. These include members of the Bcl-2 superfamily, members of the TNF family, caspases and their activators. Yet, the evolutionary conservation of those elements of the apoptotic machinery was only established from nematode to man. Recently the present day knowledge on apoptosis in sponge was compiled (Wiens and Mller 2006). The key molecules are: The poriferan Bcl-2 homologues (Wiens et al. 2000a, 2001a, 2004), the death domain proteins, which are inducers of apoptotic cell death, belonging to the family of Fas, TRADD (Wiens et al. 2000b) as well as the the poriferan caspases that

98

Fig. 5: Sponge stem cell system. Schematic outline of the postulated development of the toti-/pluri-/multipotent sponge embryonic stem cells, the archaeocytes, to the germ cells on one side and to the three major differentiated cell types, the epithelial-, the contractile- and the skeletal cells. It is indicated that during these transitions progenitor cells characteristic for these lineages have to be passed. The (potential) factors, e.g. noggin and the mesenchymal stem cell-like protein (MSCP) on the path to the skeletal cells, which trigger the differentiation are shown. In addition it is outlined that committed progenitor cells are formed which respond to the silicate/Fe(+++) stimulus through differentiation to skeletal cells, the sclerocytes (=skeletal cells). Embryonal stem cells are present in the blastocyst embryo, the blastomeres, and also in the gemmules, there as the thesocytes.

Like in any other metazoan, also in sponges fertilized eggs develop during a series of cell divisions to morulae, blastulae and perhaps even to gastrulae. Subsequently the embryos develop to mature larvae, which can be classified into several types (Leys 2003). The first study, using molecular markers to determine the restriction of gene expression during embryogenesis in a sponge appeared recently (Perovi-Ottstadt et al. 2004). We found that in oocytes, morulae and blastulae/ larvae from S. domuncula, distinct genes are expressed, among them a sponge-specific receptor tyrosine kinase (RTKvs). In addition, the sex-determining protein FEM1 and the sperm associated antigen-related protein are highly expressed; in adult animals the levels of expression of these genes are very low (Perovi-Ottstadt et al. 2004) (Fig. 6). Marker genes for totipotent cells: Using the biological systems of larvae and gemmules, the question for molecular markers of stem cells according to the stringent definition could be approached by screening for genes which are expressed in dormant germ cells and in dormant cells of gemmules. Two candidate genes have been identified, which are highly expressed either in oocytes or in cells of gemmules, the receptor tyrosine kinase RTKvs (oocytes and early larvae; Perovi-Ottstadt et al. 2004) and the embryonic development

protein EED (gemmules; Mller 2006). In tissue of adult S. domuncula both genes are expressed only in scattered cells of the pinacoderm. In situ hybridization demonstrated that RTKvs_SUBDO is highly expressed in eggs and early stages of embryos in S. domuncula. In the adults RTKvs_SUBDO-expressing cells are detectable only rarely around the aquiferous canals. EED2_ SUBDO is highly expressed in gemmules, while only a few isolated cells are found in the other tissue. In view of these data, we suggest that the cells expressing these two genes represent archaeocytes, which are in the functional state in either fertilized eggs or cells constituting early embryos (as in RTKvs_SUBDO), or form the gemmules (EED2_SUBDO). Future studies must show if precursor archaeocyte-cells exist which express RTKvs_SUBDO and EED2_SUBDO. To date we propose that these two genes are expressed in primordial, perhaps totipotent, cells of S. domuncula. Gene expression pattern of archaeocytes (stem cells) sclerocyte lineage [skeletal cells]: Sclerocytes are the cells which produce the siliceous spicules, the skeletal elements of sponges. During the progression from the totipotent archaeocytes, via the multipotent cells to the terminally differentiated somatic sclerocytes, marker genes are expressed. The first cDNAs

99

identified in S. domuncula whose deduced proteins share sequence similarity to mammalian stem cell markers were the mesenchymal stem cell-like protein (MSCP-l) and noggin (Mller et al. 2003a). They can be considered as marker genes for multipotent stem cells. The mesenchymal stem cell-like protein (MSCP) is expressed in vertebrates in mesenchymal stem cells (van den Bos et al. unpublished [see: accession number MN_016647]). These authors provided evidence that MSCP is expressed in osteogenic mesenchymal stem cells. The S. domuncula gene MSCP encoding the mesenchymal stem cell-like protein was isolated by PCR (Mller et al. 2003a). Functional studies revealed that the expression of this gene is under positive control of the morphogenetic inorganic elements silicon and ferric iron (Krasko et al. 2002). The other potential gene involved in the differentiation of stem cells in sponges is noggin. Noggin is a glycoprotein that binds bone morphogenetic proteins selectively and opposes their effects. The noggin-like protein from S. domuncula was deduced from the SDNOGG-l cDNA (Mller et al. 2003b). Expression studies revealed a higher level of the SDNOGG-l transcripts in primmorphs in the presence than in the absence of silicate/Fe(+++). Finally, a deduced protein with similarity to the vertebrate glia maturation factor should be mentioned. This protein is restricted in vertebrates to the nervous system. The sponge related protein is again closer related to the human molecule than to the corresponding molecules from D. melanogaster or S. cerevisiae (Mller et al. 2003b). Northern blot studies, supported by in situ hybridization analyses, revealed that the expression of these genes follows a sequential order, after exposure of the animals/primmorphs to silicate/Fe(+++). At first noggin is expressed, followed by silicatein and finally the glia maturation factor-like molecule. At present no molecules have been identified, which might act as cis- or trans regulators for this sequential expression, and control the temporal and spatial expression of these genes. Gene expression pattern of archaeocytes (stem cells) - pinacocyte lineage [epithelial layer]: The surface layer constituted of pinacocytes can be looked upon as an epithelium. One marker gene for the differentiation of the archaeocytes/stem cells to the pinacocytes has been isolated; the Iroquois (marker gene for the pinacocyte lineage) that codes for a putative homeobox gene has been isolated from S. domuncula (Perovic et al. 2003). Expression of the putative Iroquois transcription factor was found in these cells adjacent to the canal system; it is upregulated in primmorphs which are cultivated in strong water current (Perovic et al. 2003). The restriction of the Iroquois expression to a specific tissue region, here the epithelial layer of the aquiferous system, adds a further piece to the understanding of the complexity of tissue organization in sponges. Gene expression pattern of archaeocytes (stem cells) - myocyte lineage: Myocytes in sponges are functionally characterized as cells which synthesize the organic skeletal elements, e.g. collagen. During the progress of archaeocytes to myocytes, myotrophin is expressed in S. domuncula (Schrder et al. 2000). Myotrophin was first found in mammalian systems; in cardiac myocytes myotrophin stimulates protein biosynthesis, suggesting a crucial role in the formation of cardiac hypertrophy. The closest similarity of the sponge

Fig. 6: Sequential expression of (putative) stem cell marker genes in S. domuncula. In primmorphs as well as in germ cells a high expression two genes can be identified, the sponge-specific receptor tyrosine kinase (RTKvs) and the embryonic ectoderm development protein (EED). They might be considered as markers for totipotent stem cells. At exposure of the primmorphs to a water current, the transcription factor Iroquois is expressed; this process is seen primarily in epithelial cells. Noggin as well as silicatein gene expression is provoked after addition of silicate/Fe(+++) to the culture medium; the expression is prominent in the skeletal (spicule)-forming cell lineage. In contractile cells (myocytes), myotrophin is expressed. As underlay of the bars, early drawings of a larva of Aplysilla sulfurea Schulze, 1878. (above; DeLage 1892) and a cross section through an entire sponge (Craniella schmidtii Sollas, 1886), showing embryos within the parent (Sollas 1888), are given.

molecule is with the human sequence (identity 50% / similarity 72%; Schrder et al. 2000). Recombinant sponge myotrophin stimulates protein synthesis by 5-fold (Schrder et al. 2000). Since myotrophin is neither expressed during embryogenesis nor in gemmules it might be characterized as a marker gene for the myocyte lineage. After incubation of single cells with myotrophin the primmorphs show an unusual elongated, oval shape. Furthermore, in the presence of myotrophin sponge cells up-regulate collagen gene expression. We assume that the sponge myotrophin causes in homologous cells the same/ similar effect as the cardiac myotrophin in mammalian cells, where it is also involved in initiation of cardial ventricular hypertrophy. Focusing on the stem cell system in sponges the main lessons are; (i) sponge cells progress from a primordial stage to terminally differentiated stages, (ii) they contain totipotent stem cells, (iii) during the progression from stem cells to differentiated cells genes are expressed, among which some

100

share high sequence similarity to those identified in vertebrates (Fig. 6). At present it is the notion that the plasticity of stem cells is high, because of the high regeneration/repair capacity of somatic sponge cells.

Axis formation in sponges (2003)

Pattern formation requires the definition of the main axes (reviewed in: Mller 2005). One basic requirement for a metazoan body plan is the close interactions between adjacent cells via junctions. Our screening for a gene encoding a tight junction scaffold protein from a sponge, here S. domuncula, was successful; the scaffold protein membrane-associated guanylate kinase with inverted arrangement (MAGI) had been identified (Adell et al. 2004). The sponge MAGI scaffold protein comprises the characteristic six PDZ domains that are involved in protein:protein interaction, two WW domains that bind to proline-rich peptide motifs and the conserved guanylate kinase motif. In addition, the existence of one tetraspan receptor, tetraspanin, in S. domuncula has been reported (Mller et al. 1999d). The tetraspanins belong to a group of hydrophobic proteins, comprising four transmembrane domains with a series of conserved aa residues in the extracellular loops (Fig. 7B). By in situ hybridization it is shown that the MAGI gene is highly expressed in the epithelial cell layer and in the cells surrounding the canals. Focusing of the axis formation, diploblastic animals, Porifera and Cnidaria, are characterized by one main body axis [apical/oscular-basal], while the triploblasts have two axes [in addition to the antero-posterior axis, the dorsal-ventral]. Many signaling pathways are involved in those polarity forming processes. Two large groups can be distinguished; pathways which originate from secreted signaling molecules, and those which are controlled by transcription factors. Signaling molecules: Wnt pathway: The Wnt signaling pathway is a cell communication system which regulates cellfate decisions, tissue polarity and morphogenesis (Fig. 7A). The Frizzled protein is the membrane receptor for the Wnt secreted glycoproteins. Through the canonical Wnt signaling pathway, the activated Frizzled binds to Dishevelled (Dsh), which leads to the stabilization and accumulation of -catenin in the nucleus, where it activates the TCF/lymphoid enhancer factor (LEF) transcription factor. Besides this canonical Wnt signaling pathway two further related pathways have been identified. A signaling downstream of Dsh which includes the Rho and JNK cascade, the non canonical Wnt signaling pathway. And, the Wnt/calcium pathway that stimulates intracellular calcium release in a G-protein-dependent manner. In vertebrates the canonical Wnt signaling pathway is involved in axis specification, non canonical Wnt signaling in the formation of cell polarity and convergent extension, and the Wnt/calcium pathway in tissue separation. Recent studies demonstrated that genes encoding the Frizzled receptor (Adell et al. 2003a), as well as genes expressing proteins downstream of this receptor are present already in sponges. In situ hybridization analysis in adult S. domuncula specimens showed expression in the cortex region and in the epithelial layer of the aquiferous canals. Furthermore, Northern blot analysis revealed an upregulation of its levels of expression during the formation of sponge primmorphs.

Transcription factors: T-box Forkhead: During development sets of genes, most of them transcription factors, that are responsible for cell fate and pattern determination, are expressed. Among them are T-box (Adell et al. 2002, 2003b), Forkhead and Homeobox gene families; they have been found to be extremely conserved on sequence and functional level in all metazoans. Members of the T-box family, e.g. Brachyury, are involved in the formation and differentiation of the third germ layer, the mesoderm, in triploblastic animals. Recently, two T-box genes have been isolated from the sponge S. domuncula; a Brachyury gene and a homologue of the Tbx2-34-5 genes from chordates, which in the latter taxa are involved in the formation of the limbs. Since the larvae of S. domuncula cannot be routinely cultivated it is at present not possible to conduct experiments with them. Therefore, studies are restricted to adult animals and cultured sponge primmorphs. Interestingly it could be demonstrated that the expression of the Brachyury gene is upregulated in differentiating sponge cells during formation of canal-like structures, suggesting that already in Porifera the primordial axis is genetically fixed. This assumption was subsequently confirmed by the isolation and phylogenetic characterization of five members of the winged-helix/Forkhead gene family from the sponge S. domuncula. Forkhead proteins form a subfamily within the large group of helix-turn-helix proteins. They are responsible for a wide range of functions and key roles in early developmental processes, during organogenesis and also for the function of the major organs and tissues in the adult animals. HNF3, the founding member of this gene family, is responsible for the formation of terminal structures that develop into the gut. The expression patterns of Forkhead and T-box/Brachyury genes during late blastulae and early gastrulae, support the assumption that these two sets of genes are required for the morphogenetic movements occurring during processes identical or phylogenetically preceding gastrulation. Moreover, the overexpression of Brachyury is seen from Porifera and Coelenterata onwards to the triploblasts in distinct regions of the body, usually adjacent to the organizers. In sponges the water enters the animals through many porocytes on their surface, the inhalant openings, into the canals formed by the endopinacoderm and then passes to the choanocyte chambers from where the water is driven to the central atrium and finally pressed through the oscule back into the environment. Generally, the number of oscules is restricted and many species comprise only one major oscule. Several lines of morphological/cytological and molecular biological evidence indicate that the aquiferous canal system in sponges represents the organizational center of the sponges. During embryogenesis amphiblastula or parenchymella larvae are formed (see: Leys 2003) that have in their interior cavity choanocyte chambers; these chambers are the central organlike structure which directs and controls the water current. Subsequently canals are formed which finally fuse with the outer pinacoderm layer. 3D-cell culture experiments likewise revealed that an increase of the water current in the culture fluid results in canal formation, controlled by the homeodomain protein Iroquois (Perovi et al. 2003). In order to determine if the oscule, the morphologically most prominent region in the sponge acts also as an organizer-like region ablation/

101

Fig. 7: A. Wnt signaling pathway. The extracellular Wingless (Wg)/Wnt ligand binds to the Frizzled receptor (Fz) which regulates cell-fate decisions through the Dishevelled (Dsh) molecule that is composed of three functional domains (DIX, PDZ, DEP). From there two pathways either lead to the activation of the TCF/LEF transcription factor or to the JNK kinase cascade. The resulting selective gene expression causes a planar cell polarity. Those molecules which have been already identified in sponges are highlighted in bold. B. Tight junction proteins in S. domuncula; schematic representation of the molecules involved. Tight junctions are sealing the epithelial layer of metazoan organisms to control the lateral-extracellular transport in the aqueous milieu. By the formation of tight junctions cells undergo a polarization. The tight junctional cell membrane-spanning receptors, here tetraspanin, associate with the scaffold protein, the PDZ protein MAGI. The scaffold protein MAGI plays a crucial role in the organization of the membrane receptor molecules and the effector molecules; the latter compose the cytoskeleton and the signal transduction molecules.

transplantation studies were performed with S. domuncula; the specimens comprise only one oscule and histological sections show a large atrium in which the exhaling canals end. The regeneration capacity of the oscule region is different from that of other regions of the body. After removal of the oscule this area regenerates and is sealed only by an intact epithelium with a double pinacoderm layer but no new oscule is formed even at other parts of the surface. If however, the oscule is excised and transplanted to another site an intact oscule with an atrium is formed (reviewed in: Mller 2005). It could also be demonstrated that the oscule region comprises the highest level of expressed genes indicative for organizer regions. Among the overexpressed genes in the tissue surrounding the oscule are the LIM-homeodomain protein, Brachyury, Frizzled receptor and the secreted Frizzled-related proteins, noggin or Iroquois. These data are strong arguments for the assumption that regionalized organizer centers are present from Porifera to the crown triploblastic species and are localized close to the openings to the body cavity (see: Mller 2005). After accepting the monophyletic origin of all metazoans, including Porifera, and that all animals have common basic elements of the immune system and the body plan, it was not

too surprising that also ancestral homeobox genes are present in sponges. The developmental processes resulting in the formation of a body axis require a head center; e.g. in bilaterians, the Spemanns organizer (Spemann 1936). The genes which are involved in the establishment of the head organizer during embryogenesis have been grouped into three classes of homeobox genes, the Paired-class, the Antennapedia-class and the LIM-class genes. In this area a rapid progress was made in sponges in the last few years. A paired-class (Pax-2/5/8)-gene had been isolated from the freshwater sponge E. fluviatilis, which encodes a complete although substantially degenerated homeodomain (Seimiya et al. 1998). In S. domuncula a cDNA encoding a LIM/homeobox protein has been isolated which comprises high sequence similarity to the related LIM homeodomain proteins in other animals (Wiens et al. 2003a): its potential function was elucidated in the primmorph model. If they are cultured on a homologous galectin matrix on which they form canal-like structures, morphogenetic processes are triggered that also involve the LIM/homeobox protein/ transcription factor. In addition, retinoic acid plays an important role in local signaling of homeodomain factor-mediated vertebrate development. As the first metazoan nuclear hormone receptor

102

the retinoid X receptor (RXR) was identified in Porifera and its role was studied in the primmorph system. Like in vertebrates or in Drosophila also in S. domuncula retinoic acid is formed from -carotene via cleavage by the ,-carotene-15,15dioxygenase to retinal and further oxidization to retinoic acid. Retinoic acid (9-cis-retinoic acid) binds to RXR resulting in a regulation of transcriptional activity of morphogenetic genes, including also a homeobox gene (Wiens et al. 2003c). In Porifera the zygote increases in size and develops flagellae after fertilization. Depending on the taxon morphologically slightly different types of larvae are formed and released from the maternal body. During this phase the embryo polarizes, a process which is morphologically primarily obvious in the localization of the cilia and the formation of the body cavity. Then gastrulation takes place driven by an asymmetric and tangentially arranged cleavage. The epithelia within the body cavity invaginate under formation of the choanocyte chambers. The embryos become sessile and the young sponge forms an oscule, through which the body cavity opens to the external milieu; with this process the young sponge forms a oscular/ apical-basal axis. The oscule region as mentioned above can be considered as an organizer, since there the characteristic vertebrate organizer genes are expressed. In addition, this region is provided with a regeneration capacity which is distinguished from the rest of the animal. There are especially the organizer-specific genes, Lim-homeodomain, Brachyury, Frizzled receptor, noggin and Iroquois, which are expressed along the aquiferous system and the oscule. This water canal system is the combined route for feeding, secretion and gas exchange. Strong evidence has been collected in the last two years which indicate that these organizer genes are not only expressed around the canals and the oscule but also display morphogenetic activity. X-ray analyses of the skeleton of the Lake Baikal sponge Lubomirskia baikalensis (Pallas, 1776) reveal that the architecture of the specimens is supported by a highly ordered arrangement of the spicules within the body (Fig. 8); Kaluzhnaya et al. (2005a, 2005b). Lubomirskia baikalensis follows an organized growing pattern, reflecting a radiate accretive growth process. This pattern is maintained throughout the body of the animals and can be seen at the tips, the growing parts of the specimens, and at the basis. During growth new layers/rings of new longitudinal spicule bundles are added at the tip of the branches. A similar highly regular arrangement of the spicules can also be seen in the hexactinellid Euplectella aspergillum (Owen, 1841). There is no doubt that such highly ordered growth processes in the Porifera are genetically controlled. The discovery that organizer-specific genes are present at the tips of the branches, around the oscules, and the fact that growth along the oscular/ apical-basal axis proceeds in a radiate accretive pattern, suggest that at the growth lines, which are readily seen between the preceding and the succeeding growth layers, genes are expressed which form serial modules along this oscular/apical axis (Fig. 8). Such serial modules should, however, not be termed segments, because this would imply that they were built by a more complex genetic network, as seen in insects.

The power of genomics

There was never a faster progress in the understanding of the differentiation capacity of sponge cells than during the last 10 years when the power of molecular biology was developed and applied. In the past biochemical data were not sufficient to provide solid evidence for the phylogenetic position of the Porifera within the multicellular animals. With the application of molecular biological techniques it became clear that during the transition from the fungal state to the colonial state of multicellular animals two sets of innovative molecules had to be developed, the cell-cell and the cell-matrix adhesion systems. Surprisingly the receptors and ligands involved in these processes and so far identified from the Porifera to the crown taxa of Metazoa share no similarity to the Fungi or the Plantae. These evolutionary novelties allowed the cells to form complex aggregates which made interaction in a tuned manner possible. The basis for a specialized integration of the cells within these aggregates was an efficient signal transduction that allowed extracellular ligands to cause a modulation of the cell metabolism after complex interaction with membraneassociated receptors. These processes affected either the intermediate metabolism or the gene replication/expression machinery. Only operationally separated from these adhesion/signal transduction systems are the two further qualities of the multicellular animals systems, the immune system and the programmed cell death/apoptosis. These achievements allowed the aggregates to form individuals which have the capability to form functional units composed of cells with different properties. Having reached this evolutionary step the multicellular animals could distinguish between self-self and self/non-self and were able to adjust their cellular homeostasis. In addition, data suggest that the transcription factors already existed in the Urmetazoa which allowed to establish a simple form of body axis. These evolutionary novelties had to be created within a period of approximately only 200 million years, from the common ancestor of Fungi and Metazoa to the appearance of the Urmetazoa as the common ancestor of all metazoan phyla. This time period is comparatively short with respect to the subsequent 600 million years of evolution of Metazoa. The finding that the basic molecular strategies to form the integrated Urmetazoa have not changed qualitatively during this time, but were only prone to a processes of specialization is surprising. Until 10 years ago the opposite view was dominating, stating that the Porifera are only Parazoa [alongside animals] or Mesozoa [middle animals]. The latter terms imply qualitative assessments that in view of the molecular data compiled proved not to be helpful for a further understanding of the evolution of Metazoa. Taken together, the hypothetical ancestor of Metazoa, the Urmetazoa (Mller 2001, Mller et al. 2001), was already provided with the basic regulatory gene repertoire, like cell adhesion molecules [aggregation factor or galectin] and cell differentiation factors [involved in collagen synthesis] as well as with a highly elaborated immune system (Mller et al. 1999b) which allowed a pattern formation due to an expression of morphogens, a Pax-A like homeodomain protein, Hox-like molecules or a Lim-class homeodomain protein. In addition, recent studies in our laboratory demonstrated that

103

Fig. 8: Schematic view of pattern formation in Porifera. During embryogenesis the fertilized egg develops to a larva with a central cavity (ca). During and after settlement of the larva first choanocyte chambers (cc) are formed and the body cavity opens up to the external milieu through an oscule (os). The growth of the adult proceeds along the apical-basal axis in a compartmental manner; next to the photo of a branch, a schematic outline is given. The data suggest that the growth of those sponges which display an arborescent morphology, like the freshwater sponge L. baikalensis, proceeds by addition of serial modules along the body axis. The expression levels of those genes that code for structural proteins (the four silicateins and the MBL) as well as the mago nashi follow an apical-basal gradient; while the expressions of the catabolic enzyme cathepsin L and the house-keeping gene tubulin follows an opposite direction or remains constant. Interestingly enough, the inhibitor of the Wnt-pathway, the soluble frizzled molecule, comprises a higher level of expression at the basis, compared to the top of the branches. Finally, it is proposed that the body axis is controlled by paired-class and LIM-class homeoproteins (HP).

S. domuncula expresses Forkhead/T-box genes (Adell et al. 2003a) and also a retinoic acid receptor (Wiens et al. 2003c). Furthermore, apoptotic genes are expressed (e.g. Wiens et al. 2001) which are considered to be involved in the formation of the sponge body cavities. Based on these data it can be deduced that sponges are provided with amazingly rich and diversified regulatory molecules that allow pattern formation.

References
Adell T, Wiens M, Mller WEG (2002) Sd-Bra and SD-Tbx: first T-box transcription factors from Porifera. Boll Mus Ist Biol Univ Genova 66-67: 5 Adell T, Nefkens I, Mller WEG (2003a) Polarity factor Frizzled in the demosponge Suberites domuncula: identification, expression and localization of the receptor in the epithelium/pinacoderm. FEBS Letters 554: 363-368 Adell T, Grebenjuk VA, Wiens M, Mller WEG (2003b) Isolation and characterization of two T-box genes from sponges, the phylogenetically oldest metazoan taxon. Dev Genes Evol 213: 421-434 Adell T, Gamulin V, Perovi-Ottstadt S, Wiens M, Korzhev M, Mller IM, Mller WEG (2004) Evolution of metazoan cell junction proteins: The scaffold protein MAGI and the transmembrane receptor tetraspanin in the demosponge Suberites domuncula. J Mol Evol 59: 41-50

Althoff K, Schtt C, Krasko A, Steffen R, Batel R, Mller WEG (1998) Evidence for a symbiosis between bacteria of the genus Rhodobacter and the marine sponge Halichondria panicea: harbor also for putatively-toxic bacteria? Mar Biol 130: 529-536 Bergquist PR (1978) Sponges. Hutchinson, London Blumbach B, Pancer Z, Diehl-Seifert B, Steffen R, Mnkner J, Mller I, Mller WEG (1998) The putative sponge aggregation receptor: isolation and characterization of a molecule composed of scavenger receptor cysteine-rich domains and short consensus repeats. J Cell Sci 111: 2635-2644 Blumbach B, Diehl-Seifert B, Seack J, Steffen R, Mller IM, Mller WEG (1999) Cloning and expression of new receptors belonging to the immunoglobulin superfamily from the marine sponge Geodia cydonium. Immunogenetics 49: 751-763 Bger H (1988) Versuch ber das phylogenetische System der Porifera. Meyniana 40: 67-90 Bhm M, Hentschel U, Friedrich A, Fieseler L, Steffen R, Gamulin V, Mller IM, Mller WEG (2001) Molecular response of the sponge Suberites domuncula to bacterial infection. Mar Biol 139: 1037-1045 Borojevic R (1966) tude exprimentale de la diffrentiation des cellules de l ponge au cours de son dveloppement. Dev Biol 14: 130-153 Burger MM, Jumblatt J (1977) Membrane involvement in cellcell interactions: a two component model system for cellular recognition that does not require live cells. In: Lash JW, Burger

104

MM (eds). Cell and tissue interactions. Raven Press, New York. pp. 155-172 Buscema M, van de Vyver G (1983) Variability of allograft rejection processes in Axinella verrucosa. Dev Comp Immunol 7: 613-616 Campbell LG (1876) Log Letters from The Challenger. MacMillan, London Christen R, Ratto A, Baroin A, Perasso R, Grell KG, Adoutte A (1991) An analysis of the origin of metazoans, using comparisons of partial sequences of the 28S RNA, reveals an early emergence of triploblasts. EMBO J 10: 499-503 Custdio MR, Prokic I, Steffen R, Koziol C, Borojevic R, Brmmer F, Nickel M, Mller, WEG (1998) Primmorphs generated from dissociated cells of the sponge Suberites domuncula: A model system for studies of cell proliferation and cell death. Mech Ageing Develop 105: 45-59 Delage Y (1892) Embryognie des ponges. Arch Zool Exp 2: 345498 DeLage Y, Hrouard E (1899) Trait de zoologie concrte. Le Soudier, Paris Diaz JP (1977) Transformation histologiques et cytologiques posttraumatiques chez la demosponge Suberites massa Nardo. Bull Mus Natl Hist Nat (Paris) 445: 375-396 Doolittle WF, Brown JR (1994) Tempo, mode, the progenote, and the universal root. Proc Natl Acad Sci USA 91: 6735-6742 Esper EJC (1794) Die Pflanzentiere. Raspesche Buchhandlung. Nrnberg Felsenstein J (1993) PHYLIP, ver. 3.5. University of Washington, Seattle Field KG, Olsen GJ, Lane DJ, Giovannoni SJ, Ghiselin MT, Raff EC, Pace NR, Raff RA(1988) Molecular phylogeny of the animal kingdom. Science 239: 748-753 Gamulin V, Rinkevich B, Schcke H, Kruse M, Mller IM, Mller WEG (1994) Cell adhesion receptors and nuclear receptors are higly conserved from the lowest Metazoa (marine sponges) to vertebrates. Biol Chem Hoppe-Seyler 375: 583-588 Grant RE (1836) Animal kingdom. In: Todd RB (ed). The cyclopaedia of anatomy and physiology, vol 1. Sherwood-GilbertPiper, London Grebenjuk VA, Kuusksalu A, Kelve M, Schtze J, Schrder HC, Mller WEG (2002) Induction of (2-5)oligoadenylate synthetase in the marine sponges Suberites domuncula and Geodia cydonium by the bacterial endotoxin lipopolysaccharide. Europ J Biochem 269: 1382-1392 Haeckel E (1872). Atlas der Kalkschwmme. Verlag von Georg Reimer, Berlin Halanych KM (1991) 5S ribosomal RNA sequences inappropriate for phylogenetic reconstruction. Mol Biol Evol 8: 249-253 Henkart P, Humphreys S, Humphreys T (1973) Characterization of sponge aggregation factor. A unique proteoglycan complex. Biochemistry 12: 3045-3050 Hildemann WH, Johnston IS, Jokiel PL (1979) Immunocompetence in the lowest metazoan phylum: transplantation immunity in sponges. Science 204: 420-422 Hovanessian AG (1991) Interferon-induced and double-stranded RNA-activated enzymes: a specific protein kinase and 2,5oligoadenylate synthetases. J Interferon Res 11: 199-205 Hyman LH (1940) The Invertebrates; vol. 1. McGraw Hill, New York Koziol C, Leys SP, Mller IM, Mller WEG (1997) Cloning of Hsp70 genes from the marine sponges Sycon raphanus (Calcarea)

and Rhabdocalyptus dawsoni (Hexactinellida). An approach to solve the phylogeny of sponges. Biol J Linn Soc 62: 581-592 Kaluzhnaya OV, Belikov SI, Schrder HC, Rothenberger M, Zapf S, Kaandorp JA, Boreiko A, Mller IM, Mller WEG (2005a) Dynamics of skeleton formation in the Lake Baikal sponge Lubomirskia baicalensis. Part I. Biological and biochemical studies. Naturwissenschaften 92: 128-133 Kaluzhnaya OV, Belikov SI, Schrder HC, Wiens M, Giovine M, Krasko A, Mller IM, Mller WEG (2005b) Dynamics of skeleton formation in the lake baikal sponge Lubomirskia baicalensis. Part II. Molecular biological studies. Naturwissenschaften 92: 134138 Krasko A, Batel R, Schrder HC, Mller IM, Mller WEG (2000) Expression of silicatein and collagen genes in the marine sponge Suberites domuncula is controlled by silicate and myotrophin. Europ J Biochem 267: 4878-4887 Krasko A, Schrder HC, Batel R, Grebenjuk VA, Steffen R, Mller IM, Mller WEG (2002) Iron induces proliferation and morphogenesis in primmorphs from the marine sponge Suberites domuncula. DNA Cell Biol 21: 67-80 Kruse M, Mller IM, Mller WEG (1997) Early evolution of Metazoan serine/threonine- and tyrosine kinases: Identification of selected kinases in marine sponges. Mol Biol Evol 14: 1326-1334 Kruse M, Leys SP, Mller IM, Mller WEG (1998) Phylogenetic position of the Hexactinellida within the phylum Porifera based on amino acid sequence of the protein kinase C from Rhabdocalyptus dawsoni. J Mol Evol 46: 721-728 Kruse M, Steffen R, Batel R, Mller IM, Mller WEG (1999) Differential expression of allograft inflammatory factor 1 and of glutathione peroxidase during auto- and allograft response in marine sponges. J Cell Sci 112: 4305-4313 Kuusksalu A, Pihlak A, Mller WEG, Kelve M (1995) The (2-5) oligoadenylate synthetase is present in the lowest multicellular organisms, the marine sponges: demonstration of the existence and identification of its reaction products. Eur J Biochem 232: 351357 Leys SP (1995) Cytoskeletal architecture and organelle transport in giant syncytia formed by fusion of hexactinellid sponge tissues. Biol Bull 188: 241-254 Leys SP (2003) Comparative study of spiculogenesis in demosponge and hexactinellid larvae. Micr Res Tech 62: 300-311 Medina M, Collins AG, Silberman JD, Sogin ML (2001) Evaluating hypothesis of basal animal phylogeny using complete sequences of large and small subunit rRNA. Proc Natl Acad Sci USA 98: 97079712 Mehl D, Reiswig HM (1991) The presence of flagellar vanes in choanomeres of Porifera and their possible phylogenetic implications. Z zool Syst Evolut-Forsch 29: 312-319 Mehl D, Mller I, Mller WEG (1998) Molecular biological and palaeontological evidence that Eumetazoa, including Porifera (sponges), are of monophyletic origin. In: Watanabe Y, Fusetani N (eds). Sponge science - multidisciplinary perspectives. SpringerVerlag, Tokyo. pp 133-156 Metchnikoff (1892) Leons sur la pathologie compare de inflammation. G. Masson, Paris Moscona AA (1963) Studies on cell aggregation: demonstration of materials with selective cell-binding activity. Proc Natl Acad Sci USA 49: 742-747 Mller WEG (1982) Cell membranes in sponges. Intern Rev Cytol 77: 129-181

105

Mller WEG (1995) Molecular phylogeny of Metazoa (animals): monophyletic origin. Naturwissenschaften 82: 321-329 Mller WEG (editor) (1998a) Molecular evolution: evidence for monophyly of Metazoa. Progr Molec Subcell Biol 19. Springer Verlag, Berlin-Heidelberg Mller WEG (editor) (1998b) Molecular evolution: towards the origin of Metazoa. Progr Molec Subcell Biol 21. Springer Verlag, Berlin-Heidelberg Mller WEG (1998c) Origin of Metazoa: sponges as living fossils. Naturwissenschaften 85: 11-25 Mller WEG (2001) How was metazoan threshold crossed: the hypothetical Urmetazoa. Comp Biochem Physiol 129: 433-460 Mller WEG (2005) Spatial and temporal expression patterns in animals. In: Meyers RA (ed). Encyclopedia of molecular cell biology and molecular medicine. vol. 13. Wiley-VCH, Weinheim. pp. 269-309 Mller WEG (2006) Review: the stem cell concept in sponges (Porifera): metazoan traits. Semin Cell Dev Biol. 17(4): 481-491 Mller WEG, Zahn RK (1973) Purification and characterization of a species-specific aggregation factor in sponges. Exp Cell Res 80: 95-104 Mller WEG, Schcke H (1996) Characterization of the receptor protein-tyrosine kinase gene from the marine sponge Geodia cydonium. Prog Molec Subcell Biol 17: 183-208 Mller WEG, Mller IM (2003) Origin of the metazoan immune system: identification of the molecules and their functions in sponges. Integr Comp Biol 43: 281-292 Mller WEG, Mller I, Zahn RK, Kurelec B (1976) Species-specific aggregation factor in sponges. VI. Aggregation receptor from the cell surface. J Cell Sci 21: 227-241 Mller WEG, Diehl-Seifert B, Gramzow M, Friese U, Renneisen K, Schrder HC (1988) Interrelation between extracellular adhesion proteins and extracellular matrix in reaggregation of dissociated sponge cells. Int Rev Cytol 111: 211-229 Mller WEG, Spreitzer I, Milanovic S, Schrder HC, Rinkevich B, Gamulin V, Bretting H (1994a) Identification and characterization of S-type lectins from the marine sponge Geodia cydonium. In: van Soest RWM, van Kempen TMG, Braeckman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 389-393 Mller WEG, Mller IM, Gamulin V (1994b) On the monophyletic evolution of the Metazoa. Brazil J Med Biol Res 27: 2083-2096 Mller WEG, Blumbach B, Wagner-Hlsmann C, Lessel U (1997) Galectins in the phylogenetically oldest metazoa, the sponges [Porifera]. Trends Glycosci Glycotechnol 9: 123-130 Mller WEG, Kruse M, Koziol C, Leys SP (1998) Evolution of early Metazoa: phylogenetic status of the Hexactinellida within the phylum of Porifera [sponges]. Progr Molec Subcell Biol 21: 141-156 Mller WEG, Wiens M, Batel R, Steffen R, Borojevic R, Custdio MR (1999a) Establishment of a primary cell culture from a sponge: primmorphs from Suberites domuncula. Mar Ecol Progr Ser 178: 205-219 Mller WEG, Blumbach B, Mller IM (1999b) Evolution of the innate and adaptive immune systems: relationships between potential immune molecules in the lowest metazoan phylum [Porifera] and those in vertebrates. Transplantation 68: 1215-1227 Mller WEG, Bhm M, Batel R, de Rosa S, Tommonaro G, Mller IM, Schrder HC (2000) Application of cell culture for the production of bioactive compounds from sponges: Synthesis of

avarol by primmorphs from Dysidea avara. J Nat Prod 63: 10771081 Mller WEG, Schrder HC, Skorokhod A, Bnz C, Mller IM, Grebenjuk VA (2001) Contribution of sponge genes to unravel the genome of the hypothetical ancestor of Metazoa (Urmetazoa). Gene 276: 161-173 Mller WEG, Krasko A, Skorokhod A, Bnz C, Grebenjuk VA, Steffen R, Batel R, Schrder HC (2002) Histocompatibility reaction in tissue and cells of the marine sponge Suberites domuncula in vitro and in vivo: Central role of the allograft inflammatory factor 1. Immunogenetics 54: 48-58 Mller WEG, Wiens M, Mller IM, Schrder HC (2003a) The chemokine networks in sponges: potential roles in morphogenesis, immunity and stem cell formation. Progr Molec Subcell Biol 34: 103-143 Mller WEG, Korzhev M, Le Pennec G, Mller IM, Schrder HC (2003b) Origin of metazoan stem cell system in sponges: first approach to establish the model (Suberites domuncula). Biomol Engineer 20: 369-379 Mller WEG, Rothenberger M, Boreiko A, Tremel W, Reiber A, Schrder HC (2005) Formation of siliceous spicules in the marine demosponge Suberites domuncula. Cell Tissue Res 321: 285-297 Mller WEG, Belikov SI, Tremel W, Perry CC, Gieskes WWC, Boreiko A, Schrder HC (2006) Siliceous spicules in marine demosponges (example Suberites domuncula). Micron 37: 107120 Pahler S, Krasko A, Schtze J, Mller IM, Mller WEG (1998) Isolation and characterization of a cDNA encoding a potential morphogen from the marine sponge Geodia cydonium, that is conserved in higher metazoans. Proc R Soc Lond B 265:421-425 Pallas PS (1787) Charakteristik der Thierpflanzen. Raspesche Buchhandlung, Nrnberg Pancer Z, Kruse M, Schcke H, Scheffer U, Steffen R, Kovcs P, Mller WEG (1996) Polymorphism in the immunoglobulin-like domains of the receptor tyrosine kinase from the sponge Geodia cydonium. Cell Adhes Commun 4: 327-339 Pancer Z, Kruse M, Mller I, Mller WEG (1997) On the origin of adhesion receptors of metazoa: cloning of the integrin subunit cDNA from the sponge Geodia cydonium. Molec Biol Evol 14: 391-398 Pechenik JA (2000) Biology of the invertebrates. McGraw Hill, Boston Perovi S, Schrder HC, Sudek S, Grebenjuk VA, Batel R, tifanic M, Mller IM, Mller WEG (2003) Expression of one sponge Iroquois homeobox gene in primmorphs from Suberites domuncula during canal formation. Evol Develop 5: 240250 Perovi-Ottstadt S, etkovi H, Gamulin V, Schrder HC, Kropf K, Moss C, Korzhev M, Diehl-Seifert B, Mller IM, Mller WEG (2004) Molecular markers for germ cell differentiation in the demosponge Suberites domuncula. Int J Dev Biol 48: 293-305 Pfeifer K, Schrder HC, Rinkevich B, Uhlenbruck G, Hanisch FG, Kurelec B, Scholz P, Mller WEG (1992) Immunological and biological identification of tumor necrosis factor in sponges: role of this factor in the formation of necrosis in xenografts. Cytokine 4: 161-169 Pfeifer K, Haasemann M, Gamulin V, Bretting H, Fahrenholz F, Mller WEG (1993) S-type lectins occur also in invertebrates: high conservation of the carbohydrate recognition domain in the lectin genes from the marine sponge Geodia cydonium. Glycobiology 3: 179-184

106

Rasmont R (1979) Les ponge: des metazoaires et des societes de cellules. In: Lvi C, Boury-Esnault N (eds). Biologie des spongiaires. Centre National de la Recherche Scientifique, Paris. pp. 21-29 Reitner J, Mehl D (1995) Early Paleozoic diversification of sponges: new data and evidences. Geol Palont Mitt Innsbruck 20: 335347 Reitner J, Wrheide G (2002) Non-lithistid Demospongiae origins of their palaeobiodiversity and highlights in history of preservation. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges. Kluwer Academic/Plenum Publishers, New York. pp 52-70 Rodrigo AG, Bergquist PR, Bergquist PL, Reeves RA (1994) Are sponges animals? An investigation into the vagaries of phylogenetic inference. In: van Soest RWM, van Kempen TMG, Braeckman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 389-393 Schcke H, Schrder HC, Gamulin V, Rinkevich B, Mller IM, Mller WEG (1994a) Molecular cloning of a receptor tyrosine kinase from the marine sponge Geodia cydonium: a new member of the receptor tyrosine kinase class II family in invertebrates. Mol Membr Biol 11: 101-107 Schcke H, Mller IM, Mller WEG (1994b) Tyrosine kinase from the marine sponge Geodia cydonium. In: Mller WEG (ed). Use of aquatic invertebrates as tools for monitoring of environmental hazards. Gustav Fischer verlag, Stuttgart. pp 201-211 Scheffer U, Krasko A, Pancer Z, Mller WEG (1997) High conservation of the serum response factor within metazoa: cDNA from the sponge Geodia cydonium. Biol J Linn Soc 61: 127-137 Schrder HC, Krasko A, Batel R, Skorokhod A, Pahler S, Kruse M, Mller IM, Mller WEG (2000) Stimulation of protein (collagen) synthesis in sponge cells by a cardiac myotrophin-related molecule from Suberites domuncula. FASEB J 14: 2022-2031 Schrder HC, Brmmer F, Fattorusso E, Aiello A, de Rosa S, Ammar MSA, Mller WEG (2003) Sustainable production of bioactive compounds from sponges: primmorphs as bioreactors. In: Mller WEG (ed). Marine molecular biotechnology, vol. 1: sponges (Porifera). Springer, Berlin. pp. 163-197 Schulze FE (1887) Zur Stammesgeschichte der Hexactinelliden. Akademie der Wissenschaften, Berlin Schtze J, Custdio MR, Efremova SM, Mller IM, Mller WEG (1999) Evolutionary relationship of Metazoa within the eukaryotes based on molecular data from Porifera. Proc R Society Lond 266: 63-73 Seimiya M, Naito M, Watanabe Y, Kurosawa Y (1998) Homeobox genes in the freshwater sponge Ephydatia fluviatilis. Progr Molec Subcell Biol 19: 132-155 Simpson TL (1984) The cell biology of sponges. Springer-Verlag, New York Smith LC, Hildemann W H (1986) Allogeneic cell interactions during graft rejection in Callyspongia diffusa (Porifera; Demospongia); a study with monoclonal antibodies. Proc R Soc Lond 226: 465-377 Sollas WJ (1888) Report on the Tetractinellida. Rep Sci Res Voy H.M.S. Challenger, Zool 25(63): 1-458 Spemann H (1936) Experimentelle Beitrge zu einer Theorie der Entwicklung. Springer Verlag, Berlin von Lendenfeld R (1889) A monograph of the horny sponges. Royal Society, London

Weissenfels N (1989) Biologie und Mikroskopische Anatomie der Swasserschwmme (Spongillidae). Gustav Fischer Verlag, Stuttgart Wiens M, Mller WEG (2006) Cell death in Porifera: molecular players in the game of apoptotic cell death in living fossils. Can J Zool 84: 307-321 Wiens M, Kuusksalu A, Kelve M, Mller WEG (1999) Origin of the interferon-inducible (2-5)oligoadenylate synthetases: cloning of the (2-5)oligoadenylate synthetase from the marine sponge Geodia cydonium. FEBS Letters 462: 12-18 Wiens M, Krasko A, Mller CI, Mller WEG (2000a) Molecular evolution of apoptotic pathways: cloning of key domains from sponges (Bcl-2 homology domains and death domains) and their phylogenetic relationships. J Mol Evol 50: 520-531 Wiens M, Krasko A, Mller IM, Mller WEG (2000b) Increased expression of the potential proapoptotic molecule DD2 and increased synthesis of leukotriene B4 during allograft rejection in a marine sponge. Cell Death Diff 7: 461-469 Wiens M, Diehl-Seifert B, Mller WEG (2001) Sponge Bcl-2 homologous protein (BHP2-GC) confers selected stress resistance to human HEK-293 cells. Cell Death Diff 8: 887-898 Wiens M, Luckas B, Brmmer F, Ammar MSA, Steffen R, Batel R, Diehl-Seifert B, Schrder HC, Mller WEG (2002) Okadaic acid: a potential defense molecule for the sponge Suberites domuncula. Mar Biol 142: 213-223 Wiens M, Mangoni A, DEsposito M, Fattorusso E, Korchagina N, Schrder HC, Grebenjuk VA, Krasko A, Batel R, Mller IM, Mller WEG (2003a) The molecular basis for the evolution of the metazoan bodyplan: extracellular matrix-mediated morphogenesis in marine demosponges. J Mol Evol 57: 1-16 Wiens M, Krasko A, Perovic S, Mller WEG (2003b) Caspasemediated apoptosis in sponges: cloning and function of the phylogenetic oldest apoptotic proteases from metazoa. Biochim Biophys Acta 1593: 179-189 Wiens M, Batel R, Korzhev M, Mller WEG (2003c) Retinoid X receptor and retinoic acid response in the marine sponge Suberites domuncula. J Exp Biol 206: 3261-3271 Wiens M, Perovic-Ottstadt S, Mller IM, Mller WEG (2004) Allograft rejection in the mixed cell reaction system of the demosponge Suberites domuncula is controlled by differential expression of apoptotic genes. Immunogenetics 56: 597-610 Wilson HV (1907) On some phenomena of coalescence and regeneration in sponges. J Exp Zool 5: 245-258 Wimmer W, Blumbach B, Diehl-Seifert B, Koziol C, Batel R, Steffen R, Mller IM, Mller WEG (1999a) Increased expression of integrin and receptor tyrosine kinase genes during autograft fusion in the sponge Geodia cydonium. Cell Adhes Commun 7: 111-124 Wimmer W, Perovic S, Kruse M, Krasko A, Batel R, Mller WEG (1999b) Origin of the integrin-mediated signal transduction: functional studies with cell cultures from the sponge Suberites domuncula. Europ J Biochem 178: 156-165 Witte U, Barthel D (1994) Reproductive cycle and oogenesis of Halichondria panicea (Pallas) in Kiel bight. In: van Soest RWM, van Kempen TMG, Braeckman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 297-305

Porifera research: Biodiversity, innovation and sustainaBility - 2007

107

Diversity and evolution of deep-sea carnivorous sponges

Jean Vacelet
Aix-Marseille Universit, CNRS UMR-6540 DIMAR, Centre dOcanologie de Marseille. Station Marine dEndoume. Rue Batterie des Lions, 13007. Marseille, France. jean.vacelet@univmed.fr Abstract: The carnivorous habit of feeding that has been discovered in a cavernicolous species of Cladorhizidae is probably general for all the representatives of this deep-sea family, which numbered approximately 90 species at the end of the 20th century. Recent reports have shown that the number of species is actually considerably higher and that carnivory probably also occurs in several representatives of other Poecilosclerida families. A few specimens collected by trawling in the Pacific and Atlantic oceans have been described as new species. A larger sample collected from manned submersibles on rocky substrates near active hydrothermal sites in the south Pacific has provided a remarkably high proportion of new species. However, it is at present difficult to determine whether the abundance and diversity of carnivorous sponges in this collection is linked to the vicinity of hydrothermal sites, which provides solid substrata and general organic enrichment, and also stimulates a special sampling effort by direct methods. Carnivorous sponges cannot be considered as true members of the hydrothermal fauna, as they are apparently absent from the rich animal communities that thrive in the immediate environment of active smokers. The new species from the South Pacific include several representatives of Abyssocladia, previously synonymized with Phelloderma (Myxillina), increasing the microsclere heterogeneity of carnivorous sponges. Moreover, some other deepsea poecilosclerids, Euchelipluma spp. (Guitarridae) and some Esperiopsis spp. (Esperiopsidae) also appear to be carnivorous. This may suggest that carnivory appeared independently in several evolutionary lines of poecilosclerids. Conversely, however, the polyphyly of carnivorous sponges is not supported by a number of shared characters. The two hypotheses are discussed, but it is suggested, given the important morphological adaptations of these sponges, their ambiguous relationships with extant families of poecilosclerids and our rapidly increasing knowledge regarding their diversity, that it would be premature to drastically change the classification before having more information, especially of reproduction and molecular characters. Keywords: Cladorhizidae, Carnivorous sponges, Evolution, Deep Sea, Poecilosclerida

Introduction
The discovery that a representative of the Cladorhizidae, Asbestopluma hypogea Vacelet and Boury-Esnault, 1996 was carnivorous (Vacelet and Boury-Esnault 1995) has led to renewed interest in these strange deep-sea sponges that were previously only studied from a taxonomic point of view. Several lines of evidence suggest that this unexpected feeding habit is general in the Cladorhizidae as defined in Systema Porifera (Hajdu and Vacelet 2002) with three valid genera, Cladorhiza Sars, 1872, Asbestopluma Topsent, 1901 and Chondrocladia Thompson, 1873. It has been shown that in addition to A. hypogea several cladorhizids contain crustacean debris in the course of being digested (Kbler and Barthel 1999, Vacelet and Boury-Esnault 2002, Reiswig and Lee 2007). Furthermore, all the cladorhizids display morphological characters that are seemingly related to carnivory. They display a peculiar symmetrical shape, generally stipitate with lateral processes lined by hook-shaped microscleres. Most of them seem to be devoid of the sponge diagnostic attributes, i.e. an aquiferous system with canals, ostia, osculum and choanocyte chambers. An aquiferous system is present only in the genus Chondrocladia, in which, however, it is modified

and apparently not used for water filtration, but for the inflation of turgescent spheres at the surface of which prey capture is performed. Furthermore, recent observations suggest that this feeding regime also occurs in some other poecilosclerids that may have other family level affinities. These carnivorous sponges, which do not concur with the conventional definition of the phylum as given by Bergquist (1978) a sedentary, filter-feeding metazoan which utilizes a single layer of flagellated cells (choanocytes) to pump a unidirectional water current through its body, have developed an organization that is unique in the Metazoa, feeding on macro-prey by cells acting individually, without any digestive cavity (Vacelet and Duport 2004). The evolution, most likely from normal sponges, biology, ecology and diversity of such a remarkable derivation from a taxon that is considered as the most basal in the evolution of Metazoa, are fascinating new topics of research. The aim of this paper is to examine what is known to date of the diversity, classification and ecology of the carnivorous poecilosclerid sponges. How many taxa there are, whether they are monophyletic, or polyphyletic as carnivorous plants, and whether they are significant components of the deep-sea ecosystems, will be the main questions addressed.

108

Biodiversity of carnivorous sponges

At the end of the 20st century the Cladorhizidae numbered approximately 90 species. Most of them were described without any reference to their histological organization. This lack of information could be due to the poor preservation usual for deep-sea animals collected by dredging or trawling. It may be stressed too that the describers of cladorhizids were sponge taxonomists expecting a system of apertures, canals and choanocyte chambers, which in this case was absent or significantly modified. Several of them, however, such as Lundbeck (1905, p. 47), expressed their surprise that neither pores nor oscula have ever been mentioned. Careful observers such as Ridley and Dendy (1887) wrote The Crinorhiza forms appear to be without oscula and pores, nor have we succeeded in finding flagellated chambers, although some of the specimens were in very fair condition. It seems just possible, therefore, that, as originally suggested by Sars in the case of the first known species of the genus, Cladorhiza abyssicola, these sponges have some method of obtaining their supplies of nutriment which is quite different from that found in other sponges; this is, however, extremely unlikely. The same authors interpreted the lining of hook-like microscleres, which are now understood to be trapping devices for prey, as an efficient protection against parasites and other enemies. Recent observations on species that are definitely carnivorous, Asbestopluma hypogea and Chondrocladia gigantea (Hansen, 1885) have provided new information on their histology and organization. Moreover, a few recent taxonomic studies and studies in progress indicate that the diversity of cladorhizid sponges, and more generally of carnivorous sponges, in the deep sea is much higher than previously assumed. Since the beginning of the 21st century, 15 species have been described as new (Vacelet and Boury-Esnault 2002, Cristobo et al. 2005, Lehnert et al. 2005, Vacelet 2006, Reiswig and Lee 2007), increasing significantly the number of known species and resurrecting the genus Abyssocladia Lvi, 1964 which has been tentatively transferred from Phellodermidae to Cladorhizidae. The new species were collected in the deep south Pacific and south Atlantic, with a very high ratio of new species in the various collections, indicating that these large areas certainly still contain a large number of undescribed species. The study that I recently published on specimens from the deep Pacific (Vacelet 2006) is particularly indicative of the poor knowledge that we have of this fauna. This collection includes 9 species, of which 9 are new, although this area admittedly very large has been explored for sponges by the Challenger, Vitiaz and Galathea expeditions (Ridley and Dendy 1887, Koltun 1958, 1959, 1970, Lvi 1964). Deep-sea sponges have often been described from very few specimens, so that their intraspecific variability is poorly known, possibly wrongly resulting in an exaggerated splitting of species. This does not appear to be the case, as the new species are significantly different from any known species, and variability is low when several specimens are present. An example is Abyssocladia agglutinans Vacelet, 2006, which is known by two specimens that are exactly similar although distant by 557 km. In this study, such a high proportion of new species could be ascribed to the collection of the specimens by manned

submersibles on submarine ridges, in deep-sea environments where active hydrothermal vents favour general fauna enrichment and where hard substrates are relatively common. This could provide a higher diversity than the methods that were used in the deep Pacific by the Challenger, Vitiaz and Galathea expeditions during which the specimens were collected by blind trawling generally on mud bottoms. However, a preliminary study of cladorhizids collected by trawling off New Zealand (Kelly and Vacelet, in progress), including the Kermadec Trench which has been previously thoroughly explored by the Galathea expedition (Lvi 1964), also reveals a high ratio of undescribed species. Similar results were obtained by Cristobo et al. (2005) for the genus Chondrocladia in the south Atlantic. A collection presently under study (Fourt, Vacelet and Boury-Esnault, unpublished report to IFREMER) from the deep North Atlantic, an area which has been more thoroughly explored than the deep Pacific, also contains an unexpectedly high proportion of new species of Cladorhizidae, although not as high as in the Pacific. It is thus obvious that the diversity of cladorhizid sponges in the deep sea is far higher than is known to date and that many species, possibly genera, have yet to be discovered. It must be stressed, however, that caution must be exercised when describing such fragile sponges that are most often incomplete, in which some spicule categories are often precisely located, and which often collect pieces of other sponges due to the adhesive properties of their prey-trapping surfaces. None of the new species described in the genera Asbestopluma, Cladorhiza and Abyssocladia display any trace of aquiferous system or choanocyte chamber. The best preserved specimens display a regular arrangement of the microscleres that line the lateral filaments, with the alae of chelae or the teeth of sigmancistras outwardly directed, an arrangement which allows the capture of the thin setae of crustacean prey. Moreover, a few specimens contain crustacean debris, especially clear in Abyssocladia huitzilopochtli Vacelet, 2006, also found in two new Chondrocladia spp. These facts confirm that all the sponges presently classified in Cladorhizidae are very likely carnivorous. Indisputable general evidence, however, is difficult to provide in these fragile deep-sea animals, which easily lose the lateral filaments or appendages on which the prey are trapped and on which experimentation is not easy. Even in the best preserved specimens, prey are rarely visible, which is not surprising, considering that carnivorous sponges are sit-andwait predators that very likely do not eat frequently in the oligotrophic deep sea. Furthermore, it appears that Asbestopluma, Chondrocladia and Cladorhiza spp. are not the only carnivorous sponges. Several deep-sea poecilosclerids that are, or were, classified in diverse families due to their microsclere spicules, but that display morphology similar to that of Cladorhizidae, may also be carnivorous. As already pointed out, Abyssocladia, previously synonymized with Phelloderma Ridley and Dendy, 1886 (Phellodermidae, Myxillina), is now reconsidered as a valid genus of Cladorhizidae with seven species (Vacelet 2006) and several new species in the course of description. A carnivorous feeding habit is highly likely in Euchelipluma Topsent, 1909, which has been classified in the Guitarridae

109

due to the presence of placochelae similar to those highly diagnostic of filter-feeding sponges belonging to Guitarra Carter, 1874. The genus contains three species, E. pristina Topsent, 1909, E. (Desmatiderma) arbuscula Topsent, 1928 and E. elongata Lehnert et al., 2006. All display a pinnate shape, with regularly arranged lateral filaments lined by microscleres with the alae and teeth outwardly directed. This skeleton organization, the shape of the sponge and the seemingly absence of aquiferous system suggest a carnivorous feeding habit, which has been confirmed by the observation of crustaceans debris in some specimens of E. pristina (Vacelet 1999, Vacelet and Segonzac 2006). Another example may be found in the Esperiopsidae, where deepsea Esperiopsis spp. of the group of E. villosa Carter, 1874, including E. symmetrica Ridley and Dendy, 1886 and E. desmophora Hooper and Lvi, 1989, have a morphology that is highly suggestive of carnivorous feeding. The similarities in shape and skeleton arrangement, including a reinforcement by desmas which is rather unsual in poecilosclerids, between Esperiopsis desmophora and the Ordovician sponge Saccospongia baccata Bassler suggest that carnivory may be very ancient in poecilosclerid sponges. It thus appears that there is in fact a very high biodiversity of carnivorous sponges in the deep sea. There are now more than one hundred described representatives of Cladorhizidae which very likely have this feeding habit, but this number is certainly much higher, and it appears too that carnivory has been developed in some representatives of other poecilosclerid families as construed in the present consensual classification. So far, however, this feeding habit is restricted to the order Poecilosclerida. A role of exotyles present in diverse orders of demosponges in trapping large particles has been suggested by Hajdu (1994), but this does not indicate a carnivorous feeding habit in sponges possessing exotyles, in which an aquiferous system has generally been reported. The high specific diversity and the biogeographical distribution of the carnivorous sponges are difficult to correlate with peculiarities in the reproduction mechanisms and to the dispersal ability, as the reproduction of these species is poorly known. In the Mediterranean, Asbestopluma hypogea has been able to colonize several littoral caves most probably from deep-sea canyons (Bakran-Petricioli et al. 2007), suggesting relatively high dispersal ability.

rhizoids as an anchoring base, or on rocky bottom where their diversity has certainly been more seriously underestimated. However, it remains unknown which is the more favourable type of bottom. In a few cases, it seems that the same species, or very close species, may live on both types of substratum, developing either an enlarged fixation base on rocks or rhizoids in mud. The high diversity found in collections taken by manned submersibles from rocky bottom near active hydrothermal vents might suggest that this previously poorly sampled environment is their preferred habitat. It would appear that carnivorous sponges, as generally filter-feeding sponges, do not take part in the rich oases of life thriving in the immediate proximity of active hydrothermal vents. One exception is Cladorhiza methanophila (Vacelet et al. 1995, 1996) which constitutes unusually large populations near methane sources of a mud volcano in the Barbados because of symbiosis with methanotroph bacteria. The number and diversity of carnivorous sponges, however, could be enhanced at a certain distance from such sites both by the unusual prevalence of rocky substrates due to volcanic activity, and by a general enrichment of the deep-sea ecosystem. No quantitative data are available to date in relation to this question. The abundance of Chondrocladia lampadiglobus Vacelet, 2006 on the East Pacific Ridge between 2586 and 2684 m depth has been estimated at 1-2.6 individuals per km of path by the manned submersible Nautile, but this estimation was made some hundreds of meters from active vents and we do not know if this is general on the East Pacific Ridge. So far hydrothermal sites have been the favorite targets of exploration from manned submersibles and ROVs, introducing an evident bias. In a Pacific abyssal plain rich in polymetallic nodules, the density of Cladorhizidae has been estimated respectively at 16, 4 and 5 individuals per hectare for Chondrocladia, Cladorhiza and Asbestopluma (Tilot 1992). Given their relatively small number and small size, it would not appear at present that carnivorous sponges play an important role in the deep-sea food chains.

Evolution and classification

The diversity of microscleres and of organization in carnivorous poecilosclerids raises a puzzling problem of evolution and classification. It has been pointed out by Hajdu and Vacelet (2002) that the family Cladorhizidae as construed in Systema Porifera in the suborder Mycalina lacks a clear synapomorphy and that it could be polyphyletic. However, this has to be reexamined, as several shared characters of Cladorhizidae were also found in other sponges, such as Euchelipluma, Abyssocladia and Esperiopsis spp. that now appear to belong to a set of carnivorous sponges. The three cladorhizid genera Asbestopluma, Cladorhiza and Chondrocladia display a special shape, generaly stipitate, pinnate or branching (Table 1), and their megascleres, although differing in size and shape according to their localization in the sponge, are referable to a single category (mycalostyles) with the same skeleton arrangement. Asbestopluma and Cladorhiza share the general organization, with a stipitate, often pinnate shape and absence of aquiferous system, but not Chondrocladia, which has kept the sponge aquiferous system although its function and organization are

Ecology
All carnivorous sponges are deep-sea species that were previously considered as well adapted to the most foodpoor mid-basin areas (Gage and Tyler 1991). They may be considered as sit-and-wait predators, spending a minimal amount of energy during long periods between rare feeding opportunities. Three species of Cladorhizidae have been found at more than 8000 m depth and Asbestopluma occidentalis Lambe, 1883 is the deepest known sponge, living in hadal depth at 8840 m (Koltun 1970). A few cladorhizids, however, are able to live at only 100 m depth in high latitudes, and the Mediterranean cavernicolous species, Asbestopluma hypogea, lives at a few meters depth in a cold-water littoral cave, but most likely colonizes this habitat from a deep-sea population. They live either on muddy bottom, where they often develop

110

Table 1: Characters of the genera of carnivorous sponges (including some unpublished data). +: present in all species. : present in some species only. * present in a single species. Abyssocladia Stipitate, pinnate or branching Stipitate, with inflated spheres Rhizoids Aquiferous system Mycalostyles Basal substrongyles Basal desmas Spinose tylostyles or oxeas Microtylostyles Surface lining by microscleres Palmate anisochelae Anchorate anisochelae Anchorate isochelae Palmate isochelae Arcuate isochelae Abyssochelae Placochelae Sigmas Sigmancistras Forceps + Asbestopluma Chondrocladia + + + + + + + Cladorhiza + + + * * + + + + + * + + ? + * + Esperiopsis (pars) + Euchelipluma +

+ + * + +

+ + + *

significantly modified. The chelae microscleres of the three genera, however, are different, being palmate anisochelae in Asbestopluma, anchorate/unguiferate anisochelae in Cladorhiza and anchorate isochelae in Chondrocladia, indicating possible polyphyly according to the present interpretation of chelae (Hajdu et al. 1994). The value of microscleres, and especially of cheloids in sponge classification has often been the subject of debate. However, the most common opinion today for Poecilosclerida is that summarized by van Soest (2002, p. 518): Because of complex morphology, chelae are considered to reflect phylogenetic relationships both at the family and genus level. This seemingly phyletic heterogeneity of sponges conventionally classified in Cladorhizidae is even more evident for the whole set of carnivorous sponges with the recent additions. Carnivorous sponges now very likely include: (i) three species of Euchelipluma, presently classified in Guitarridae due to the presence of the diagnostic placochelae; (ii) several species of Esperiopsis of the villosa group, currently classified in the Esperiopsidae, with palmate isochelae (Vacelet 2006); (iii) seven described species of Abyssocladia and several new species under study from New Zealand and the North Atlantic, currently classified in the Cladorhizidae although they have palmate or arcuate isochelae. The case of Abyssocladia is particularly puzzling because the genus was previously synonymized, on the basis of possession of special isochelae (abyssochelae), with Phelloderma Ridley and Dendy, 1886 in family Phellodermidae van Soest and Hajdu, 2002 in the suborder Myxillina in which the chelae are not palmate, but arcuate. However, it now appears that the type of isochelae, arcuate or palmate, is rather uncertain in Abyssocladia as the genus stands now with the inclusion of the newly described

species and of species in the process of description, as will be explained below in greater detail. The heterogeneity of the Cladorhizidae, and more generally of carnivorous poecilosclerids, could be interpreted in two different ways: (i) carnivory has developed relatively recently in several different lines of evolution of Poecilosclerida, some or all of the characters that they share being homoplasies due to their special mode of life, with the consequence that the various genera of carnivorous sponges are to be classified in these different lines, matching up several families of Mycalina, possibly of Myxillina; (ii) carnivory developed very early in Poecilosclerida, possibly before the divergence of Mycalina and Myxillina, the shared characters being symplesiomorphic, with the consequence that they are to be classified in a single high level taxon, possibly a distinct suborder. The present set of data, summarized in Table 1, with the shared and distinctive characters of carnivorous sponges, will be examined and discussed below.
Fig. 1: Representatives of four genera of carnivorous sponges illustrating the diversity of microscleres in sponges with a similar morphology and organization. All have similar megascleres (mycalostyles) with or without addition of strongyles. A. Abyssocladia naudur Vacelet, 2006, paratype, abyssochelae, sigma and sigmancistra. B. Asbestopluma agglutinans Vacelet, 2006, holotype and paratype, anisochelae 1, anisochelae 2, and sigmancistra. C. Cladorhiza segonzaci Vacelet, 2006, holotype and paratypes, anchorate anisochelae, sigma and sigmancistra. D. Euchelipluma pristina Topsent, 1919, specimen from Barbados, isochela, placochela and sigmancistra. A, B and C from Vacelet (2006). D from Vacelet and Segonzac (2006).

111

112

All the presumed carnivorous sponges share a certain number of morphological characters. They display a special outer morphology, absence or significant modification of the aquiferous system, an unusual microsclere arrangement and the same type of megasclere skeleton. A stipitate shape with symmetrical lateral expansions is found in all genera. For instance, the feather-like shape with a more or less laterally compressed axis bearing symmetrical laterals filaments lined by prey-trapping microscleres is found in all genera except Chondrocladia, and species classified in different genera or families may be remarkably similar in shape (Fig.1). The absence of canal system and choanocyte chambers also appears to be general, again with the exception of Chondrocladia. The main skeleton is always composed of more or less modified mycalostyles, frequently strongly fusiform, building longitudinal axes, showing only size differentiation according to their position in the main or secondary axes, without an ectosomal differentiation. In several species of the various genera, the base of the main axis is reinforced by mycalostyles modified in short strongyles. In three species, Euchelipluma arbuscula, Asbestopluma (Helophloeina) stylivarians (Topsent, 1928), Esperiopsis desmophora, again belonging to various genera and families, these strongyles are themselves modified in desmas. As far as may be inferred from the description of often poorly preserved specimens, most, possibly all species have on their lateral extensions a lining of sigmoid or cheloid microscleres arranged with teeth and alae outwardly directed. Carnivorous sponges are the only poecilosclerids possessing true sigmancistras, which are sometimes difficult to differentiate from sigmas but most often clearly distinct. Sigmancistras are recorded in all genera except Esperiopsis, although their occurrence is not general in all the species. They have been reported for only two Asbestopluma spp. and four Chondrocladia spp., but they are more frequent in Cladorhiza and occur in all species of Abyssocladia and Euchelipluma. Sigmancistras have been supposed by Hajdu (1994) to be the primitive condition of development of diverse poecilosclerid microscleres (cyrtancistra, diancistra, clavidisc), an hypothesis which may underline the possible antiquity of carnivory in Poecilosclerida. An interesting issue for other shared characters of carnivorous sponges could be the reproductive phenomena. Reproduction is very poorly known in these deep-sea sponges, although large embryos have been reported fairly often (for instance by Lundbeck 1905), but without precise description. These reports and preliminary observations suggest that carnivorous sponges could share some peculiarities. In several genera the embryos have been described as large, including fascicles of megascleres and a variety of microscleres, and with a special envelope which suggested to Topsent (1909) that they were gemmules rather than embryos in Cladorhiza spp. According to personal unpublished observations in Asbestopluma hypogea, young embryos have multiflagellated cells (Fig. 2), a character which is very unusual in sponges (multiflagellated cells are known only in the trichimella larva of Hexactinellida). This is confirmed by preliminary observations in another species of Asbestopluma (Leys, pers. comm.). The spermatogenesis of A. hypogea also appears very unusual (Vacelet 1996) (Fig. 2), possibly in relation with the absence of choanocyte chambers from which the sperm

cells of sponges generally derive. Spermatocysts develop in the body, and then migrate towards the end of the lateral processes, where the mature cysts become free. In the mature cyst, sperm cells are surrounded by two envelopes, the inner one unicellular, the outer one made by closely intertwined cells. Two tufts of forceps are diametrically protruding on the mature cyst, and may serve either as flotation devices for dispersal of the whole cyst or for capture by the Velcro-like lining of the filaments of another individual. Preliminary light microscope observations suggest that similar phenomena may occur in Cladorhiza methanophila (Vacelet and BouryEsnault 2002), in Chondrocladia gigantea (Hansen, 1885) (Kbler and Barthel 1999), and in Euchelipluma pristina (unpublished). These shared characters of carnivorous sponges, however, are in contrast with the characters of the chelae microscleres, which is an important character for the suborder classification of Poecilosclerida. The chelae are arcuate in the suborder Myxillina, and palmate in the suborder Mycalina (Hajdu et al. 1994, Hooper and van Soest 2002). The presence of palmate anisochelae, anchorate anisochelae and anchorate isochelae in the three genera of Cladorhizidae as defined in Systema Porifera, was already rather puzzling. The addition of Euchelipluma, Abyssocladia and some Esperiopsis spp., in which are present placochelae and isochelae grading from palmate to arcuate, greatly increases the microsclere diversity of carnivorous poecilosclerids. Moreover, most of the abyssochelae and isochelae of Abyssocladia, as well as the isochelae of Euchelipluma are difficult to assign precisely to the arcuate or palmate type. In several species, isochelae and abyssochelae may be palmate to arcuate, or clearly palmate, or clearly arcuate (confirmed by Hajdu, pers. comm.). In a few other Abyssocladia and Asbestopluma, some microsclere characters do not agree with the present distinction between microscleres of Mycalina and Myxillina. Abyssocladia dominalba Vacelet, 2006 has small anisochelae with one end palmate and the other arcuate, in addition to arcuate isochelae and palmate to arcuate abyssochelae. An Asbestopluma sp. from New Zealand, in the course of description by Kelly and Vacelet, has arcuate, possibly anchorate, anisochelae in addition to the normal palmate anisochelae. An Abyssocladia sp., which will also be described from New Zealand, has isochelae intermediate between the arcuate and the anchorate types. This supports the hypothesis of polyphyly of carnivorous poecilosclerids, which would mean that several shared characters were obtained by homoplasic evolution in several lines of Mycalina, in relationship with carnivorous feeding. Is that likely? This feeding mode does not in fact need any aquiferous system and could induce a certain number of common features. A plumose or pinnate shape, which is most propitious for the passive capture of swimming prey, needs strong axes of fusiform spicules, ideally reinforced at the base of the stalk by intermingled spicules such as a cover of short vermiform strongyles or of desmas. Prey capture also requires a lining of the lateral expansions by hook-like cheloid microscleres, able to trap the setae or appendages of invertebrate prey. Diverse shapes of chelae and sigmas are propitious for such a role, and chelae, sigmancistras and placochelae appear particularly well adapted. Possible similarities in reproduction processes could

113

Fig. 2: Reproductive stages in Asbestopluma hypogea. A. Semi-thin section through an embryo, showing internal cells with yolk inclusions, an outer layer of multiflagellated cells (arrow) and a maternal envelope. B. TEM view of a multiflagellated cell of the same embryo. C. Mature sperm cyst at the tip of a filament, showing sperm cells (Sp), an inner envelope (En) and an outer envelope (Oe) made of closely intertwined cells, and two tufts of forceps (desilicified) within their sclerocytes (Fo).

also be induced by the absence of aquiferous system and life in the deep sea. Such a hypothesis of convergent evolution of different characters linked to a carnivorous feeding habit from different families of Poecilosclerida cannot be ruled out. However, it is rather unlikely that the microscleres appeared

independently in several lines of evolution, for instance the placochelae in Euchelipluma and in Guitarridae. Moreover, this interpretation would mean the allocation of each genus to a precise family, which in fact appears difficult. The affinities of Asbestopluma with Mycalidae, of Esperiopsis

114

with Esperiopsidae and of Euchelipluma with Guitarridae rely only on a single category of microscleres and on similar megascleres which, however, are very differently arranged in filter-feeding representatives of these families. Furthermore, Cladorhiza, Chondrocladia and Abyssocladia would remain with uncertain affinities in the present classification system. The second hypothesis is more in agreement with the characters shared by carnivorous poecilosclerids, such as special morphology, megasclere nature and arrangement, microsclere arrangement, and the absence or modification of aquiferous system. Only the genus Chondrocladia would appear to be distinctive by its morphology and the preservation of an aquiferous system. However, a taxon including all the carnivorous sponges, even excluding Chondrocladia, would not be in agreement with the present classification of Poecilosclerida. The highly diverse microsclere complement found in the various genera would match more or less adequately several families of Mycalina, possibly also of Myxillina. This could mean that the present basis of the classification of Poecilosclerida in suborder is not appropriate. It could also mean that carnivory is very ancient in sponges with a possible support from the similarities between the Ordovician Saccospongia baccata and Esperiopsis desmophora and appeared in an early ancestor of Poecilosclerida with mycalostyle megascleres and very diverse cheloid microscleres, to be considered as symplesiomorphies. These microscleres could subsequently have evolved differently in the diverse lines of Poecilosclerida, developing a clearer distinction between palmate, arcuate and anchorate types of chelae. It is interesting to note that in carnivorous poecilosclerids the cheloid microscleres have an obvious function, for which both their shape and arrangement are well designed, while in filter-feeding poecilosclerids they have no apparent function. Was this function lost in filterfeeding sponges, or was it relatively recently designed by carnivorous sponges from preadapted microscleres of presently unknown usefulness? I consider that there is presently not enough strong evidence for one or another of these hypotheses, which rest on morphological and spicule characters. As suggested by the surprising specific diversity found in the recent collections, we may in the near future expect the discovery in the deep sea of a large number of undescribed species, which will bring new information. The data on reproduction are still very incomplete and poorly understood, but the uniqueness of preliminary observations in a few species suggests that they will provide significant information. Furthermore, indications from gene sequences will certainly become available soon and will bring new data on the molecular phylogeny of Poecilosclerida, which is at present insufficiently achieved and includes very few exploitable data for carnivorous sponges. My feeling is that it would be preferable to wait for this new information before offering an interpretation of the phylogeny or a formal proposal for the classification of carnivorous Poecilosclerida. However, in predictive mode, the present set of data appears to me more in agreement with the hypothesis of monophyly or paraphyly of carnivorous poecilosclerids, possibly to be considered as a distinct suborder of Poecilosclerida, rather than with a polyphyletic interpretation.

Acknowledgements
I am grateful to Michel Segonzac (Ifremer) and Michelle Kelly (NIWA) for entrusting me with the study of deep-sea specimens. I also acknowledge the technical help of Chantal Bzac, Centre dOcanologie de Marseille. This paper benefited from fruitful discussions and comments from Nicole Boury-Esnault and Eduardo Hajdu.

References
Bakran-Petricioli T, Vacelet J, Zibrowius H, Petricioli D, Chevaldonn P, Rada T (2007) New data on the distribution of the deep-sea sponges Asbestopluma hypogea and Oopsacas minuta in the Mediterranean Sea. Mar Ecol Evol Persp 28(suppl. 1): 1023 Bergquist PR (1978) Sponges. Hutchinson & Co, London Cristobo FJ, Urgorri V, Rios P (2005) Three new species of carnivorous deep-sea sponges from the DIVA-1 expedition in the Angola Basin (South Atlantic). Organisms Diversity & Evolution 5: 203-213 Gage JD, Tyler PA (1991) Deep-sea biology. Cambridge University Press, Cambridge. Hajdu E (1994) A phylogenetic interpretation of hamacanthids (Demospongiae, Porifera), with the redescription of Hamacantha popana. J Zool 232: 61-77 Hajdu E, Vacelet J (2002) Family Cladorhizidae. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges, vol. 1. Kluwer Academic/Plenum Publishers, New York. pp. 636-641 Hajdu E, van Soest RWM, Hooper JNA (1994) Proposal for a phylogenetic subordinal classification of poecilosclerid sponges. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 123-140 Hooper JNA, Lvi C (1989) Esperiopsis desmophora n. sp. (Porifera: Demospongiae): a desma-bearing Poecilosclerida. Memoir Queensl Mus 27: 437-441 Hooper JNA, van Soest RWM (2002) Order Poecilosclerida Topsent, 1928. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges, vol. 1. Kluwer Academic/ Plenum Publishers, New York. pp. 403-408 Koltun VM (1958) Cornacuspongida of sea waters washing the South Sakhalin and the South Kurile Islands region. Issled dalvenost Mor SSSR 5: 42-77 Koltun VM (1959) Corneosiliceous sponges of the northern and far eastern seas of the USSR. Zoological Institute of the Academy of Sciences of the USSR, Moscow-Leningrad Koltun VM (1970) Sponge fauna of the north-western Pacific from the shallows to the ultra-abyssal depths. Inst Oceanol Acad Sci USSR 86: 165-221 Kbler B, Barthel D (1999) A carnivorous sponge, Chondrocladia gigantea (Porifera: Demospongiae: Cladorhizidae), the giant deepsea clubsponge from the Norwegian trench. Memoir Queensl Mus 44: 289-298 Lehnert H, Watling L, Stone R (2005) Cladorhiza corona sp. nov. (Porifera: Demospongiae: Cladorhizidae) from the Aleutian Islands (Alaska). J Mar Biol Assoc UK 85: 1359-1366 Lvi C (1964) Spongiaires des zones bathyale, abyssale et hadale. Galathea Rep 7: 63-112

115

Lundbeck W (1905) Porifera. Pars II: Desmacidonidae (pars). Danish Ingolf-Exp 6(2): 1-219 Reiswig HM, Lee WL (2007) A new species of Cladorhiza (Porifera: Cladorhizidae) from S. California (USA). In: Custdio MR, LboHajdu G, Hajdu E, Muricy G (eds). Porifera research: biodiversity, innovation, sustainability. Srie Livros 28. Museu Nacional, Rio de Janeiro. pp. 517-523 Ridley OS, Dendy A (1887) Report on the Monaxonida collected by H.M.S. Challenger during the years 1873-1876. Rep Sci Res Voy H.M.S. Challenger, Zool 20(59): 1-275 Tilot V (1992) La structure des assemblages mgabenthiques dune province nodules polymtalliques de lOcan Pacifique tropical Est. Thesis, Universit de Bretagne occidentale, Brest Topsent E (1909) tude sur quelques Cladorhiza et sur Euchelipluma pristina n. g. et n. sp. Bull Inst ocanogr Monaco 151: 1-23 Topsent E (1928) Une mycaline productrice de desmes Desmatiderma arbuscula n. g., n.sp. Bull Inst ocanogr Monaco 519: 1-8 Vacelet J (1996) Deep-sea sponges in a Mediterranean cave. In: Uiblein F, Ott J, Stachowitsch M (eds). Deep-sea and extreme shallow-water habitats: affinities and adaptations, 11. Austrian Academy of Sciences, Vienna. pp. 299-312 Vacelet J (1999) Outlook to the future of sponges. Memoir Queensl Mus 44: 27-32 Vacelet J (2006) New carnivorous sponges (Porifera, Poecilosclerida) collected from manned submersibles in the deep Pacific. Zool J Linn Soc 148: 553-584

Vacelet J, Boury-Esnault N (1995) Carnivorous sponges. Nature 373: 333-335 Vacelet J, Boury-Esnault N (2002) A new species of carnivorous deep-sea sponge (Demospongiae: Cladorhizidae) associated with methanotrophic bacteria. Cahiers Biol Mar 43: 141-148 Vacelet J, Boury-Esnault N, Fiala-Mdioni A, Fisher CR (1995) A methanotrophic carnivorous sponge. Nature 377: 296 Vacelet J, Duport (2004) Prey capture and digestion in the carnivorous sponge Asbestopluma hypogea (Porifera: Demospongiae). Zoomorphology 123: 179-190 Vacelet J, Fiala-Mdioni A, Fisher CR, Boury-Esnault N (1996) Symbiosis between methane-oxidizing bacteria and a deep-sea carnivorous cladorhizid sponge. Mar Ecol Prog Ser 145: 77-85 Vacelet J, Segonzac M (2006) Porifera. In: Desbruyres D, Segonzac M, Bright M (eds). Handbook of deep-sea hydrothermal vent fauna, 18. Denisia. pp. 35-46 van Soest RWM, Hajdu E (2002) Family Phellodermidae fam. nov. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges, vol. 1. Kluwer Academic/Plenum Publishers, New York. pp. 621-624 van Soest RWM (2002) Suborder Myxillina Hajdu, van Soest & Hooper, 1994. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges, vol. 1. Kluwer Academic/Plenum Publishers, New York. pp. 515-520

Porifera research: Biodiversity, innovation and sustainaBility - 2007

117

South American continental sponges: state of the art of the research

Ceclia Volkmer-Ribeiro
Museu de Cincias Naturais, Fundao Zoobotnica do Rio Grande do Sul. Rua Dr. Salvador Frana, 1427. 90690-000, Porto Alegre, RS, Brazil. Research fellow of CNPq. cvolkmer@fzb.rs.gov.br Abstract: An intensive survey and study of the South American freshwater sponges started around four decades ago. A large number of results came out which now allows a first overview. An outstanding number of new species of freshwater sponges and even new genera were described. The new materials presented remarkable differences when compared with species from the other continents. This new sponge continental fauna has ever since been subjected to constant checking due to continued and extensive surveys, redescriptions and habitat descriptions. The diversity of the South American sponge fauna is remarkable being possibly the richest in the world. All the review efforts dedicated to this branch of Demosponges are enhancing confidence in the perception of how thrustworthy some characters are and how much their degree of variation is linked to specific habitats, thus enabling the application of this taxonomic knowledge. Keywords: Continental sponges, research, South America, state of the art

Build and rebuild your software

Evolution has been the main word leading the research of all of us as biologists and naturalists in our efforts to understand how living beings have organized themselves and how living processes of all sorts play around us. Lately, with the growing success of replicating life in our labs, evolution includes the idea of how we are setting up life around the world and even around the planetary system. One but not the least of the resulting consequences is that taxonomists are being urged to come up with applied propositions stemming from their taxonomic research. This taxonomic perfectioning takes place and weight as our scientific life goes by and particularly if it stays long enough as that of a large number of us has been staying. Looking back and forth becomes a daily routine with special emphasis on ideas, hypotheses and conclusions, challenged by every new evidence, findings and reading. One starts with grain-size evidences, cements them together, builds up his/her initial operating system and from then on builds and rebuilds oneselfs rationale. Plain common sense, but a painful process And here it is when Science turns into Art and Philosophy. Into art because a scientists always present inquiring impulses are driven by a desire to grasp and reproduce the perfection offered by Nature. Into philosophy because only this endeavour will heal the human mind in its failure to reach that goal.

The gains of the continental approach

A whole new continental fauna of Demosponges, that of the South American continent, was disclosed mainly over the last half of the 20th century. This continental fauna may be comparable in importance to the findings of fossil marine

sponge faunas preserved in rocky outcrops which lately emerged as continental areas. The main difference between these lies on the fact that those marine faunas are a snapshot of the past and not an ongoing movie as the present continental sponge faunas are. And, as such, research on continental sponge faunas benefits from playbacks whenever collecting sites are re-visited to check hypotheses, improve descriptions and study niches (niche sensu Hutchinson 1967). But more than offering easy access to previous collecting sites, continental sponge faunas offer the chance to study biological/ ecological (adaptation),versus geologic (phylogeny) evolution, when continental drift is considered and continental plates are understood as analogous to huge islands. A large isolation degree in respect to time and space causing easier determination and follow up of geographic barriers and vicariance effects (Nelson and Platnick 1981) appears then as important aspects to drive the search for evolving characters in these continental Demosponge faunas aiming to the perfection of species identification, the detection of endemisms and the estimation of the time consumed along all these processes. All such aspects are obviously harder to detect when marine demosponges are considered. Three main realms should be in fact considered in what respects sponges: the marine, the epicontinental and the continental one. The first and third ones need no further explanation. The second one has to do with those seas in the process of continental enclosure, so turning into brackish and then fresh water. This epicontinental marine fraction of the present world waters has seldom been surveyed for sponges, in spite of the fact that other phyla were studied in detail and seen to pass by a drastic reduction in biodiversity, like, for instance the Echinoderms in the Baltic Sea (Hutchinson 1967). The presently known rich marine versus poor freshwater

118

poriferan biodiversity (Hooper and van Soest 2002) certainly allows for the expectation that the present epicontinental sponge faunas may also offer very interesting and intriguing selection processes and local extinctions prior to adapting to freshwater.

Continental surveying of an unknown sponge fauna

The first descriptions of South American continental sponges were produced in the 19th century and were based on a few specimens gathered in the Orinoco, Amazon and Uruguay Rivers by foreign explorers and deposited mainly in English and German Museums. The next main taxonomic efforts date to the middle of the last century, when Bonetto and Ezcurra de Drago started to survey the Argentinean continental sponges (Ezcurra de Drago 1971) and VolkmerRibeiro the Brazilian ones (Volkmer-Ribeiro 1981) resulting at present in a number of twenty nine new species, six new genera and one new extant family described, plus one new fossil species and family defined. At this time significant freshwater sponge collections were initiated by these authors at respectively the Instituto Nacional de Limnologia - INALI, Santa F, Argentina and the Museu de Cincias Naturais, Fundao Zoobotnica do Rio Grande do Sul, Rio Grande do Sul, Brazil. Bonetto and Ezcurra de Drago as well as Volkmer-Ribeiro initially faced a confusing situation in what respected the taxonomy of the worlds freshwater sponges long ago split by Carters taxonomic proposals (Jewell 1952). The problem was however overcome due to the remarkable differences that the new materials presented when compared with the descriptions available for the species already known. In that way an outstanding number of new species was described which is resisting the continued studies and surveys that are being carried out until the present. The appearance of Penney and Raceks (1968) comprehensive revision of the worlds freshwater sponges offered next a sound basis for revisional studies of species and genera and proposition of new genera. The master lines established by Penney and Racek were followed by both of the authors, Volkmer-Ribeiro and Ezcurra de Drago, resulting in the validation and enlightening of prior proposals for diverse continental sponge genera by Gray (1867), according to detailed redescriptions (VolkmerRibeiro and De Rosa-Barbosa 1972, Volkmer-Ribeiro 1984, Volkmer-Ribeiro and Costa 1992, Volkmer-Ribeiro and Tavares 1995, Tavares and Volkmer-Ribeiro 1997) of new materials with South American species, which had been only briefly described before.

South American continental plates. In this way genera with exclusive Nearctic-Neotropical distribution came into light such as Corvomeyenia Weltner, 1913 (Volkmer-Ribeiro et al. 2005) and Anheteromeyenia Schrder, 1927 (VolkmerRibeiro 1986a), or with predominant occurrence in these two continents like Racekiela Bass and Volkmer-Ribeiro, 1998 (Bass and Volkmer-Ribeiro 1998).

One genus and five continental plates drifted apart

The continued concern with the search for trustworthy diagnostic specific characters present in dried preserved materials (so that comparative studies could keep on encompassing old preserved materials) was deepened next with the revision of Metania Gray 1867, which has a tropical distribution. Such an effort sprung from the large number of specimens of this genus collected particularly in the Brazilian and Venezuelan Amazonia. At this time the search for what would become trustworthy diagnostic characters for species identification was centered on the large isolated sponge faunas at the plates of South America, Africa, Australia, India and Indonesia split from Gondwana and drifted apart. The resulting study brought to attention the existence of a gondwanian fauna of freshwater sponges (Volkmer-Ribeiro 1986b, Volkmer-Ribeiro and Costa 1992, 1993, Silva and Volkmer-Ribeiro 2001), which Volkmer-Ribeiro and De Rosa-Barbosa (1979) had already indicated by extending the occurrence of the family Potamolepidae from the Ethiopian to other gondwanian plates. An array of characteristics came to light again which confirmed Grays (1867) insight of the value of gemmoscleres, microscleres and the gemmule structure for the definition of genera and species.

Timing the birth of a continental sponge fauna

Also, from the studies on Metania it was possible for the first time to envision the time elapsed in attaining speciation and generic diversification for a group of continental sponges i.e. the Cretaceous drifting apart of the Gondwanian plates. Given the established paradigm that geologic time scales are those required to produce measurable change, particularly of gemmoscleres and microscleres shape and size, as well as on number of spicule categories present, a continued revisional effort was extended to all genera of South American continental sponges, allowing a more confident redefinition of monospecific genera (Acalle, Volkmer-Ribeiro and De Rosa-Barbosa 1972, Uruguaya and Sterrastrolepis, VolkmerRibeiro and De Rosa-Barbosa 1979, and the recognition of new genera (Oncosclera Volkmer-Ribeiro, 1970, Saturnospongilla Volkmer-Ribeiro, 1976, Corvoheteromeyenia, Ezcurra de Drago, 1979, Racekiela Bass and Volkmer-Ribeiro, 1998) as well as the description of new species, among which those where monospecific genera descriptions had been based on: Saturnospongilla carvalhoi, Sterrastrolepis brasiliensis. At present the largest number of records in the South American freshwater sponges are from Brazil and Argentina but reports have also been produced for Suriname (Ezcurra de Drago 1975), Venezuela (Bonetto and Ezcurra de Drago 1973, Volkmer-Ribeiro and Pauls 2000), Chile (Ezcurra de Drago 1974, Kilian and Wintermann-Kilian 1976) and Uruguay

Two neighboring continental plates

A milestone mark at this point was the study of the Edward Potts collection of type materials of the species he described for the United States and Canada (Potts 1887) and deposited at the Academy of Natural Sciences of Philadelphia (VolkmerRibeiro and Traveset 1987). That collection had not been taken in consideration by Penney and Racek (op. cit.). The idea was that comparative descriptions of species restricted to the Nearctic/Neotropical regions would favor any future vicariance studies due to the vicinity of the North and the

119

(Berroa Beln 1968). The South American continental plate has so been crossed from north to south and east to west leading to the idea that a large number of habitats has yet to be surveyed for continental sponges in this remarkably diverse continent. Now, quite a different picture has emerged of this South American fauna showing that it is one of the richest, if not the richest in the world.

Applied sponge taxonomy

Continuous checking of habitat characteristics versus species occurrence is a tool never discarded by specialists in their search for the confirmation of a species status and the description of ecomorphic variations of characters (VolkmerRibeiro 1973, Poirrier 1974). The performance of the abovementioned procedures, besides submitting species and genera to constant revisional efforts generates confidence in species definitions and provides a series of applications for this taxonomic knowledge. The first concerns the monitoring of freshwater habitats with respect to the integrity of their biodiversity or their restoration with environmental recovering practices. The knowledge now available encompasses the detection of several sponge assemblages in some typical South American habitats, such as Coastal ponds, lakes and lagoons (VolkmerRibeiro and Machado 2007), Cerrado (Savannah) ponds where an assemblage of five species thrive or where they formed spongillite deposits in the past (Volkmer-Ribeiro et al. 1998). Also, particular sponge assemblages have been detected in large South American rivers, as for instance the rocky bottoms of the middle Uruguay river (Ezcurra de Drago and Bonetto 1969), the rocky tributaries of the middle Paran river, as well as in the macrophyte stands of its middle floodplain (Ezcurra de Drago 1993, 2003), in Amazonian river rocky bottoms and in their marginal seasonally flooded forests (Batista et al. 2003). The application of this taxonomic tool is also proving to be rewarding onto environmental and climatic paleointerpretations, following the identification of sponge species based on the spicules detected in columns of recovered lake sediments of quaternary age (Siffeddine et al. 1994, Volkmer-Ribeiro and Turcq 1996, Turcq et al. 1998, Cndido et al. 2000, Volkmer-Ribeiro et al. 2007, Parolin et al. 2007). While the surveys are being extended across the continent some restricted local endemisms remain unchallenged. As a result the first official State and National recognitions and red listings of sponge species under threat were attained, upon the following of IUCN standard procedures. Oncosclera jewelli, Anheteromeyenia ornata and Drulia browni integrate the red list of endangered fauna of Rio Grande do Sul State, Brazil (Volkmer-Ribeiro 2003). Oncosclera jewelli, A. ornata, Uruguaya corallioides, Sterrastrolepis brasiliensis, Corvoheteromeyenia australis, Corvoheteromeyenia heterosclera, Corvospongilla volkmeri, Heteromeyenia insignis, Houssayella iguazuensis, Racekiela sheilae and Metania kiliani are listed with the Brazilian endangered freshwater invertebrates and fishes (Brasil 2004). This fact offers support to national and regional policies aiming at the

preservation of particular freshwater habitats and the species they contain. Lately, surveys for the South American continental sponges have also focused on river, lake, lagoonal and pond waters contained in preserved areas such as State and National Parks and Ecological Stations with the aim of establishing parameters for biomonitoring and bioindication, at the same time improving the knowledge of the species they contain and the use of such preserved areas as biodiversity banks (Volkmer-Ribeiro et al. 1988, 1999, 2005, Tavares et al. 2005, Volkmer-Ribeiro and Almeida 2005). A further application of this taxonomic tool is within the context of archeological studies. Spicules (cauxi), present in archeological Amazonian pottery are revealing unsuspected native technologies and histories of the sustainable management of natural resources (in this case biosilica produced by the sponges), besides allowing the tracing of cultural trends and past native population migrations within the continent (Volkmer-Ribeiro and Gomes 2006, VolkmerRibeiro and Viana 2006).

What next?
Homo sapiens may be producing a more extensive modification of the Earths surface than any other animal species did before. One such profound environmental change is the damming of large rivers in order to produce hydroelectric power, particularly throughout the Tropical and Sub-tropical realms. South America is a continent where the damming of large rivers has boomed over the last thirty years. Huge lakes have been formed in areas where this permanent freshwater habitat was previously absent, such as in the Amazonian Region, famous for its seasonal vrzea lakes. In regard to the rich Amazonian sponge fauna, surveys have extended to this new habitat in order to detect the invasion by sponges and the exclusion/adaptation forces in action. Results have shown that colonization is being carried by some species previously detected in the riverine rocky bottoms when the prior Impact Assessment surveys were done. All harder substrates located in the lake waters (excluding the anoxic ones), including the trunks of the forest flooded by the lake, are being used by those sponges that had occupied more extensively the original river bottoms (Volkmer-Ribeiro and Hatanaka 1991). The monitoring of the occupation of these dammed waters by sponges is being continued bearing in mind to offer taxonomic substrate for further research purposes encompassing from basic sponge biology and ecology to the production of biosilica or biocompounds by sponges. The mapping of substrates occupied by sponges in these dammed waters, allied to their continued recruitment as a consequence of the permanent ingression of upstream gemmules, renders these living stocks ideal for continued observation/monitoring and experimentation. These natural systems are better than laboratory aquaria, where freshwater sponge species other than those belonging to Ephydatia are barely kept alive for a few days. Another area of research being pursued based on the taxonomic knowledge currently available is the area of medicine, as there are several historical and some current records of dermal diseases caused by contact with sponge

120

spicules, particularly in the Amazonian Region. The discovery of freshwater sponge spicules acting as agents of ocular pathology in the Araguaia River area (Brazilian Amazonia) has only recently been reported (Volkmer-Ribeiro et al. 2006, Volkmer-Ribeiro and Batista 2007). Other pathologies related to freshwater sponge spicules inferred from archeological work have also been compiled and discussed in the aforementioned publications. The spreading of the geographic surveys of the South American continental sponges is obviously an ongoing process. Hopefully at the same speed as global economic enterprises are reaching them and their habitats. Renewed efforts should, from now on, focus on the aquatic habitats contained in preserved areas which, as a rule, benefit of previous selections aiming the protection of continental biomes. The invertebrate faunas of such biomes are yet poorly known and their study will certainly come up with the detection of new species.

Acknowledgments
The author is indebted to the organizers of the 7th International Sponge Symposium (Armao dos Bzios, RJ) for the invitation to present this opening speech as well as to two anonymous referees for the suggestions presented. She heartily thanks Dr. Eduardo Hajdu for a minutious reading of the MS and valuable improvements indicated. She acknowledges the continued support CNPq. has provided to the research projects proposed along the last three decades.

References
Bass D, Volkmer-Ribeiro C (1998) Radiospongilla crateriformis (Porifera, Spongillidae) in the West Indies and taxonomic notes. Iheringia Sr Zool 85: 123-128 Batista TCA, Volkmer-Ribeiro C, Darwich A, Alves LF (2003) Freshwater sponges as indicators of floodplain lake environments and of river rocky bottoms in Central Amazonia. Amazoniana 17(3/4): 525-549 Berroa Beln C (1968) Nomina de las esponjas dulceacucolas de la fauna del rio Uruguay, Amrica Del Sur. Physis 27(75): 285-289 Bonetto AA, Ezcurra de Drago I (1973) Aportes al conocimiento de las esponjas del Orinoco. Physis 32(84): 19-27 Brasil (2004) Instruo Normativa n 5, de 21 de maio de 2004. Dirio Oficial da Republica Federativa do Brasil, Braslia 28 de maio de 2004, Seo 1: 136-142 Candido JL, Volkmer-Ribeiro C, Simes Filho FL, Turcq BJ, Chauvel A (2000) Microsclere variations of Dosilia pydanieli (Porifera, Spongillidae) in Caracaran lake (Roraima - Brazil). Palaeoenvironmental implication. Biocincias 8(2): 77-92 Ezcurra de Drago I (1971) Porifera. In: Hulbert SH (ed). Biota acutica de Sudamrica Austral. San Diego State University, San Diego. pp. 57-61 Ezcurra de Drago I (1974). La presencia de Spongilla (Eunapius) fragilis Leidy en Chile (Porifera, Spongillidae). Physis B 33(87): 249-252 Ezcurra de Drago I (1975) Freshwater sponges of Suriname. Stud Fauna Suriname 15: 175-183 Ezcurra de Drago I (1979) Um nuevo gnero sudamericano de esponjas: Corvoheteromeyenia gen. nov. (Porifera Spongillidae). Neotropica 25(74): 109-118

Ezcurra de Drago I (1993) Distribucin geogrfica de las esponjas argentinas (Porifera: Spongillidae, Potamolepidae y Metaniidae). Relaciones zoogeogrficas, vias de poblamiento. In: Boltovskoy A, Lpez HL (eds). Conferencias de Limnologia, Instituto de Limnologia Dr. R. A. Ringuelet. La Plata, Buenos Aires. pp. 115-125 Ezcurra de Drago I (2003) Biodiversidad de Porifera en el litoral argentino. Grado de competncia com el bivalvo invasor Limnoperna fortunei (Dunker, 1857) (Bivalvia, Mytilidae). In: Aceolaza FG (ed). Temas de la Biodiversidad del Litoral argentino, INSUGEO, Miscelneas 12, Tucumn. pp. 5-12 Ezcurra de Drago I, Bonetto AA (1969) Algunas caracteristicas del bentos en los saltos del rio Uruguay, con especial referencia a la Ecologia de los Porferos. Physis 28(77): 359-369 Gray JE (1867) Notes on the arrangement of sponges, with the description of some new genera. Proc Zool Soc London 1867: 492558 Hooper JNA, van Soest RWM (2002) Systema Porifera: a guide to the classification of sponges. Kluwer Academic/Plenum Publishers, New York Hutchinson GE (1967) A treatise on limnology. Volume II. Introduction to lake biology and the limnoplancton. John Wiley & Sons, New York Jewell ME (1952) The genera of North American freshwater sponges; Parameyenia, new genus. Trans Kansas Acad Sci 55: 445-457 Kilian E, Wintermann-Kilian G (1976) Die Spongilliden Sdamericas derzeitiger Stand der Kenntniss iher Verbreitung. In: Descimon H (ed). Biogeographie et evolution en Amerique tropicale. Publications du Laboratoire de Zoologie de lcole Normale Superieure 9: 75-97 Nelson G, Platnik NI (1981) Systematics and biogeography: cladistics and vicariance. Columbia University Press, New York Parolin M, Volkmer-Ribeiro C, Stevaux JC (2007) Sponge spicules in peaty sediments as paleoenvironmental indicators of the holocene in the Upper Paran River, Brazil. Rev Bras Paleontol 10(1): 17-26 Penney JT, Racek AA (1968) Comprehensive revision of a worldwide collection of freshwater sponges (Porifera: Spongillidae). Bull U S natn Mus 272: 1-184 Poirrier M (1974) Ecomorphic variation in gemmoscleres of Ephydatia fluviatilis Linnaeus (Porifera: Spongillidae) with comments upon its systematics and ecology. Hydrobiologia 44(4): 337-347 Potts E (1887) Contributions towards a synopsis of the American forms of freshwater sponges with descriptions of those named by other authors and from all parts of the world. Proc Acad Nat Sci Philadelphia 39: 158-270 Sifeddine A, Frhlich F, Fournier M, Martin L, Servant M, Soubis F, Turcq B, Suguio K, Volkmer-Ribeiro C (1994) La sedimentation lacustre indicateur de changements des paloenvironments au cours des 30.000 dernires annes (Carajs, Amazonie, Brsil). C R Acad Sci Paris 318(II): 1645-1652 Silva CMM, Volkmer-Ribeiro C (2001) Key to the Ethiopian species of the genus Metania Gray, 1867 (Porifera: Metaniidae) with redescription of Metania rhodesiana and M. godeauxi, comb. n. Bull Inst r Sci nat Belg Biologie 71: 127-138 Tavares MCM, Volkmer-Ribeiro C (1997) Redescrio das esponjas de gua doce Oncosclera navicella (Carter, 1881) (Potamolepidae) e Spongilla spoliata Volkmer-Ribeiro & Maciel, 1983 (Spongillidae). Biocincias 5(1): 97-111

121

Tavares MCM, Volkmer-Ribeiro C, Hermany G (2005) Seasonal abundance in a sponge assembly at a Southern Neotropical inner delta. J Coast Res 42: 335-342 Turcq B, Sifeddine A, Martin L, Absy ML, Soubies F, Suguio K, Volkmer-Ribeiro C (1998) Amazon Forest fires: a lacustrine report of 7.000 years. Ambio 27(2): 139-142 Volkmer-Ribeiro C (1973) Redescription and ecomorphic variations of the freshwater sponge Trochospongilla minuta (Potts, 1887). Proc Acad Nat Sci Philadelphia 125(8): 137-144 Volkmer-Ribeiro C (1970) Oncosclera - a new genus of freshwater sponges (Porifera-Spongillidae) with redescription of two species. Amazoniana 2(4): 435-442 Volkmer-Ribeiro C (1976) A new monotipic genus of neotropical freshwater sponges (Porifera-Spongillidae) and evidence of a speciation via hybridism. Hydrobiologia 50(3): 271-281 Volkmer-Ribeiro C (1981) Porifera. In: Hurlbert SH, Rodrigues G, Santos ND (eds). Aquatic biota of tropical South America. Part 2: Anarthropoda. San Diego State University, San Diego. pp. 86-95 Volkmer-Ribeiro C (1984) Evolutionary study of the genus Metania Gray, 1867 (Porifera: Spongillidae): II. Redescription of two Neotropical species. Amazoniana 8(4): 541-553 Volkmer-Ribeiro C (1986a) Acanthodiscus new genus and genus Anheteromeyenia redefined (Porifera, Spongillidae). Iheringia, Sr Zool 81: 31-43 Volkmer-Ribeiro C (1986b) Evolutionary study of the freshwater sponge genus Metania Gray, 1867: III. Metaniidae, new family. Amazoniana 9(4): 493-509 Volkmer-Ribeiro C (2003) Porferos. In: Fontana CS, Bencke GA, Reis RE (orgs). Livro vermelho da fauna ameaada de extino no Rio Grande do Sul. Edipucrs, Porto Alegre. pp. 43-48 Volkmer-Ribeiro C, Almeida FB de (2005) As esponjas do Lago Tup. In: Santos-Silva E, Aprile FM, Scudeller VV, Melo S (orgs). BioTup: meio fsico, diversidade biolgica e sociocultural do Baixo Rio Negro, Amaznia Central. INPA, Manaus. pp.123-134 Volkmer-Ribeiro C, Batista TCA (2007) Levantamento de cauxi (Porifera, Demospongiae), provvel agente etiolgico de doena ocular em humanos, Araguatins, Rio Araguaia, Estado do Tocantins, Brasil. Rev Bras Zool 24(1): 133-143 Volkmer-Ribeiro C, Correia MMF, Brenha SLA, Mendona MA (1999) Freshwater sponges from a Neotropical sand dune area. Mem Qld Mus 44: 643-649 Volkmer-Ribeiro C, Costa PR (1992) On Metania spinata (Carter, 1881) and Metania kiliani n. sp.: Porifera, Metaniidae VolkmerRibeiro, 1986. Amazoniana 7(1): 7-16 Volkmer-Ribeiro C, Costa PR (1993) Redescription of the Oriental and Australian species of the genus Metania Gray, 1867 (Porifera: Metaniidae). Iheringia, Sr Zool 74: 81-101 Volkmer-Ribeiro C, De Rosa-Barbosa R (1972) On Acalle recurvata (Bowerbank, 1863) and an associated fauna of other freshwater sponges. Rev Bras Biol 32(3): 303-317 Volkmer-Ribeiro C, De Rosa-Barbosa R (1978) A new genus and species of Neotropical freshwater sponges. Iheringia, Sr Zool 52: 103-107 Volkmer-Ribeiro C, De Rosa-Barbosa R (1979) Neotropical freshwater sponges of the Family Potamolepidae Brien, 1967. In:

Lvi C, Boury-Esnault N (eds) Biologie des spongiaires. Centre National de la Rechereche Scientifique, Paris. pp: 503-511 Volkmer-Ribeiro C, De Rosa-Barbosa R, Machado VS (2005) Corvomeyenia epilithosa sp. nov. (Porifera, Metaniidae) no Parque Nacional da Serra Geral, Rio Grande do Sul, Brasil. Rev Bras Zool 22(4): 844-852 Volkmer-Ribeiro C, Ezcurra de Drago I, Parolin M (2007) Spicules of the freshwater sponge Ephydatia facunda indicate lagoonal paleoenvironments at the Pampas of Buenos Aires Province, Argentina. J Coast Res 50: 449-452 Volkmer-Ribeiro C, Gomes DMC (2006) Ferraz Egreja: implicaes zooarqueolgicas no estudo do antiplstico cermico. In: Vialou AV (org.) Pr-histria do Mato Grosso, Volume 2: Cidade de Pedra. EDUSP, So Paulo. pp. 203-206 Volkmer-Ribeiro C, Hatanaka T (1991) Nota Cientifica: composio especfica e substrato da espongofauna (Porfera) no Lago da Usina Hidroeltrica-Tucuru, Par, Brasil. Iheringia, Sr Zool 71: 177-178 Volkmer-Ribeiro C, Lenzi HL, Orfice F, Pelajo-Machado M, Alencar LM de, Fonseca CF, Batista TCA, Manso PPA, Coelho J, Machado M (2006) Freshwater sponge spicules: a new agent of ocular pathololgy. Mem Inst Oswaldo Cruz 101(8): 899-206 Volkmer-Ribeiro C, Machado VS (2007) Freshwater sponges (Porifera, Demospongiae) indicators of some coastal habitats in South America: redescriptions and key to identification. Iheringia, Sr Zool 97(2): 157-167 Volkmer-Ribeiro C, Motta JFM, Callegaro VLM (1998) Taxonomy and distribution of Brazilian spongillites. In: Watanabe Y, Fusetani N (eds). Sponge sciences: multidisciplinary perspectives. SpringerVerlag, Tokyo. pp. 271-278 Volkmer-Ribeiro C, Mansur MCD, Mera PAS, Ross SM (1988) Biological indicators in the aquatic habitats of the Ilha de Maraca. In: Milliken W, Ratter JA (eds). Marac The biodiversity and environment of an Amazonian rainforest. Chichester, John Willey & Sons Volkmer-Ribeiro C, Parolin M (2005) Segundo registro de Sterrastrolepis brasiliensis Volkmer-Ribeiro & De Rosa-Barbosa (Demospongiae, Potamolepidae) com descrio do habitat e de assemblia, Bacia do Rio Paran, Brasil. Rev Bras Zool 22(4): 1003-1013 Volkmer-Ribeiro C, Pauls SM (2000) Esponjas de agua dulce (Porifera, Demospongiae) de Venezuela. Acta Biol Venez 20(1): 1-28 Volkmer-Ribeiro C, Tavares MCM (1995) Redescrio de Drulia uruguayensis Bonetto & Ezcurra de Drago, 1968 com redefinio do gnero Drulia Gray, 1867 (Porifera: Metaniidae). Biocincias 3(1): 183-205 Volkmer-Ribeiro C, Traveset A (1987) Annotated catalog of the type specimens of Potts species of freshwater sponges. Proc Acad Nat Sci Philadelphia 139: 223-242 Volkmer-Ribeiro C, Turcq B (1996) SEM analysis of siliceous spicules of a freshwater sponge indicate paleoenvironmental changes. Acta Microscop 5(B): 186-187 Volkmer-Ribeiro C, Viana SA (2006) Cermica arqueolgica com cauixi. In: Viana SA (coord). Pr-Histria no Vale do Rio Manso/ MT. Universidade Catlica de Gois, Goinia. pp. 309-327

Porifera research: Biodiversity, innovation and sustainaBility - 2007

123

The Sponge Barcoding Project: aiding in the identification and description of poriferan taxa
Gert Wrheide(1*), Dirk Erpenbeck(1,2), Christian Menke(1)
Abteilung Geobiologie, Geowissenschaftliches Zentrum, Georg-August Universitt Gttingen, Goldschmidtstr. 3, D37077 Gttingen, Germany. gert.woerheide@geo.uni-goettingen.de, derpenb@gwdg.de, cmenke@davion.de (2) Biodiversity Program, Queensland Museum, South Brisbane, Queensland, Australia
(1)

Abstract: Sponges are among the most ancestral metazoans and are often notoriously difficult to identify due to their depauperate suite of complex morphological characters. However, as a group they are highly diverse, ecologically important and of significant commercial importance to the pharmaceutical and biomaterials industry. Therefore, means of unambiguous identification are urgently needed. Sponge barcodes promise a set of indispensable tools to aid taxonomists and ecologists in confirmation of identification of sponge species, and have the potential to enhance the discovery of drug-producing species. Here, we introduce the Sponge Barcoding Project (SBP) and present the structure of the Sponge Barcoding Database (SBD). Keywords: DNA barcoding, DNA taxonomy, sponges

Introduction
Sponges are among the most ancestral metazoans (e.g., Medina et al. 2001) and may hold many clues to our understanding of the evolution of early animal and developmental processes (Martindale 2005). They are highly diverse, abundant in nearly every aquatic habitat, some freshwater and most marine, and play numerous important ecological roles, e.g. in nutrient cycling (Lesser 2006) or as bioeroding organisms in coral reefs (Lopez-Victoria and Zea 2005). Their significant commercial importance to the pharmaceutical and biomaterials industry is increasingly being recognized, e.g. as producers of highly potent secondary metabolites (reviewed in e.g. Faulkner 2000) useful for drug development (Munro et al. 1994). Many sponge species are notoriously difficult to identify, often even by taxonomic experts, because morphological characters for comparative morphology are scarce and prone to homoplasies, highly variable or otherwise unsuitable for unambiguous identification. In addition, many sponges discovered in large scale biodiversity surveys remain undescribed (Hooper and Ekins 2005), partly also due to the lack of skilled taxonomists. As a result of uncertainties in morphological systematics, sponge species have frequently been regarded as widely distributed (cosmopolitan). However, genetic approaches, mostly using allozymes, have clearly shown that such cosmopolitan sponge species are rare and appear to result from over-conservative systematics, lumping morphologically similar but evolutionary distinct lineages into one widely distributed morpho-species (e.g. Klautau et al. 1999). The question of how to describe and distinguish such genetically distinct and reproductively isolated lineages remains, due to the difficulty of relating those genetic differences to morphological delineation of species.

Secondly, however, what is a species in sponges? While the use of fixed differences in diagnostic morphological characters (e.g. spicules and architecture) is practical and has served reasonably well to catalogue diversity, it is doubtful that such a typological system reflects the real biological diversity. Sponge alpha-taxonomy still is a quite artificial system solely based on morphological differences without considering evolutionary history and/or reproductive isolation. However, this issue will not be discussed in depth here. For further discussion see Sol-Cava and Wrheide (2007). Nonetheless, correctly identifying reproductively isolated and evolutionary distinct lineages of sponges remains pertinent for understanding a broad range of subjects such as marine ecology, biodiversity, dispersal, animal evolution and discovery of pharmaceutically / biotechnologically valuable taxa. Nonetheless, conventional morphological taxonomy alone clearly is at its limit with the task of distinguishing closely related but evolutionary distinct sponge lineages, especially in character poor taxa such as e.g. Halichondrida. The utilization of additional characters, such as informative signature DNA sequences (also known as DNA barcodes, Hebert et al. 2003a, 2003b), and the establishment of a DNA sequence-aided taxonomic system might provide an opportunity to overcome these shortcomings and aid our future endeavours to strive for more comprehensive species discoveries and descriptions as well as the deeper understanding of evolutionary factors that shape species distributions in space and time. In any case, a DNA sequence-based taxonomic system should by no means replace but rather complement conventional taxonomy based on comparative morphology the DNA sequences will be regarded as additional characters to described morphological (and biochemical) features.

124

The pitfalls and conceptual weaknesses of DNA barcoding have been discussed extensively in the literature (e.g. Moritz and Cicero 2004, Meyer and Paulay 2005, Hickerson et al. 2006) and shall not be repeated here as they are also summarized by Sol-Cava and Wrheide (2007). Despite all acknowledged shortcomings of DNA barcoding, a group of sponge biologists convened after a round table discussion at the 7th International Sponge Conference and agreed that time was ripe to act and initiated the Sponge Barcoding Project. This project was proposed by the first author (GW) during his talk, with an international steering committee consisting of Gert Wrheide (project coordinator, Germany), Andrea Blanquer (Spain), Paco Cardenas (Norway), Christina Maria Diaz (Venezuela), Sandra Duran (Spain), Dirk Erpenbeck (Australia, Germany), Dennis Lavrov (USA), Jose Lopez (USA), Grace McCormack (Ireland), Shirley Pomponi (USA) and Bob Thacker (USA) (in alphabetical order). The aim of the present contribution is to briefly outline the Sponge Barcoding Project (SBP), its recently launched website (www.spongebarcoding.org) and the Sponge Barcoding Database (SBD).

Scope
Phylum Porifera (sponges) consists of more than 8,000 described species, with an estimated species number of 15,000 (Hooper and van Soest 2002). The Sponge Barcoding Project (SBP) aims at establishing a DNA sequence-based reference system to aid future species discovery and description. It will work towards covering species from all sponge taxa, from classes Demospongiae, Hexactinellida, and Calcarea, ranging in habitat from the marine intertidal to the deep-sea, as well as freshwater, and from different biogeographic regions. In the long term the SBP intends to sequence DNA signature sequences of about 8,000 taxa, with an initial phase of 3 years focusing on 2,000 species covering all genera. This will provide a platform from which more extensive sampling can be directed. To establish a solid taxonomic framework, the SBP will start with recently described type specimens curated in associated museums and supplement these with unequivocally identifiable species. Fresh material of such taxa will be collected by individual groups involved in the SBP and will be taxonomically identified by an expert before making DNA signature sequences publicly available via the projects website and database. Ongoing pilot studies will evaluate the efficiacy of the methods using closely related taxa and investigate the relationship between intra- vs. interspecific diversity (i.e., the Barcoding gap, Meyer and Paulay 2006).

Background and approach

This is the first worldwide barcoding project on any diploblast taxon, and intends to cover the complete taxonomic range of Porifera. Several smaller pilot studies have recently been conducted independently, with various levels of resolution and success (Duran and Rtzler 2006, Wrheide 2006). The standard mtDNA cytochrome oxidase subunit 1 (COI) barcoding fragment, which is used for almost all current (eukaryotic) barcoding initiatives, spans over

a ca. 650 nucleotide region close to the 5 end (Erpenbeck et al. 2006). This mitochondrial protein displays sufficient variability in most bilaterian species. However, in diploblastic animals mitochondrial proteins display a lower evolutionary rate (Wrheide et al. 2000, Shearer et al. 2002) and it has been shown, that frequently co-occurring, congeneric sibling sponge species are difficult to separate with COI fragments (Wrheide 2006) due to very low variability. However, a more variable downstream fragment appears to bear adequate resolution (Erpenbeck et al. 2006). Therefore, a concerted effort is now needed to evaluate the usefulness of DNA signature sequences for poriferan species discovery and description, and warrants comprehensive, phylum-wide coverage. Due to the fact that the highly conserved COI-barcoding primers are prone to amplify sponge commensals and/or symbionts, the primary task in the first phase of the project is to optimize sponge-specific primer design. Additionally, to resolve closely related species, we will supplement the standard ca. 650 bp fragment with 440 bp of downstream sequence. The addition of an unlinked marker such as either rDNA ITS or the C2D2 region of the 28S rDNA gene might prove pivotal to accomplish the projects aims. The second task will then be, after sufficient initial data have been gathered, to evaluate the potential of those DNA signature sequences for species distinction, i.e. the error rate associated with certain thresholds of genetic distances commonly used for species designation (see also Meyer and Paulay 2005, Hickerson et al. 2006). It is imaginable that once a sufficiently and densely covered reference system has been established and evaluated, identification of any given specimen, using the standard barcoding marker, at least to genus level should be possible. From there, species designation would be contingent on more variable signature sequences such as rDNA ITS or a fragment of the 28S rDNA. However, all this will take place in combination with conventional comparative morphology. Therefore, we will focus the initial phase of the SBP on appropriately identified and curated type specimens to build a taxonomically sound and solid backbone for a DNA sequenceaided taxonomy. Samples will be obtained from associated partners e.g. at the Zoological Museum in Amsterdam/ Netherlands (Dr. Rob van Soest - >18,000 specimens) and from the Queensland Museum in Brisbane/Australia (Prof. John Hooper - >34,000 specimens), or from collections of SBP partners such as the Harbor Branch Oceanographic Institution (Dr. S. Pomponi - >34,000 specimens) in Florida, USA. Initial assessments of DNA quality from these collections indicate they are of an adequate suitability. DNA sequencing will be carried out either by individual SBP partners (all of whom have significant prior expertise in the field and appropriate capacities), or can be pooled at certain DNA sequencing facilities (e.g. at the coordinating institution, the Gttingen Centre for Biodiversity and Ecology, Germany). All efforts will be undertaken to support developing countries in their efforts to produce DNA barcodes from specimens of their sponge fauna. Sequences and associated data (voucher and taxonomic information) will be made publicly available at the projects database (see below) on its website (www.spongebarcoding.org) and submitted to the Barcode of Life Data Systems and Genbank/EMBL databases.

125

Such a system could also enable a reverse taxonomic system, in that in large-scale biodiversity surveys, all collected samples are genotyped for DNA signature sequences, followed by pooling all specimens with identical or highly similar sequences. This will enable focused morphological work on distinct genetic lineages.

point of entry for the Sponge Barcoding Database, which is outlined in more detail below.

Technical concept and implementation of the Sponge Barcoding Database

The Sponge Barcoding Database (SBD) is developed with the aim to function as the primary access point for DNA signature sequences together with providing information on conventional morphological taxonomic characters to aid species discovery, description and characterization. The unique combination of sponge-specific conventional taxonomic information and DNA signature sequences is the distinguishing feature, in which the SBD differs from other database systems, such as Genbank (www.ncbi.nlm.nih.gov/) or the Barcode of Life Data Systems (www.barcodinglife.com/). While records of the SBD will be

The Sponge Barcoding Website (www.spongebarcoding.org)

The Sponge Barcoding Website (Fig. 1) has been set up to serve the projects aims and provide a centralized platform for data exchange. Information about the approach, progress and the people involved can be found on separate pages, as well as a list of local (taxonomic or geographic) campaigns, as well as a guide on how to get involved in the project (Fig. 1). The central place for data access is the Data page, the

Fig. 1: The entry page of the Sponge Barcoding Projects website (www.spongebarcoding.org), accessed on 31 May 2007.

126

Fig. 2: The structure of the Sponge Barcoding Database. Rectangles represent tables in the database with ovals connected to them denoting their attributes (i.e. table columns), diamonds denote relationships between the tables and their multiplicities given as numbering on each line (1 or n). For example, for any given species, there can be arbitrarily many specimen records, but each specimen is of exactly one species. For each specimen, arbitrarily many sequences can be saved, but each sequence will be obtained from only one specimen.

linked with both databases, both do not provide the desired flexibility and have the desired options available for the SBD, e.g. they do not provide fields to store more detailed (morphological) taxonomic descriptions. An additional backbone for nomenclatorial and taxonomical entries is the cross-linking to the World Porifera Database (WPD, www.marinespecies.org/porifera/). This data base is edited by Rob van Soest, Nicole Boury-Esnault, Dorte Janussen and John Hooper and has gone online in 2005. The WPD will provide the ultimate taxonomic authority with regards to accepted species names. The structure of the SBD (Fig. 2) was developed with flexibility and avoidance of redundancy in mind. For example, if multiple DNA sequences will be provided for one specimen, then the specimen information is saved only once. To achieve this, the database consists of multiple tables (Fig. 2): - A species table containing species names and an ID for each species. The Species ID (WPD-ID) will directly be obtained and linked to the record in the World Porifera Database, which can also directly be accessed through the SBP website. - A specimen table containing all the information related to one collected specimen, e.g. collection date, location and a morphological description. Each specimen record has its own, unique record number in the SBD (SBD-ID). Specimen records are linked to the species table by their species IDs. This table also contains fields for filenames of associated pictures. The associated images themselves are saved outside of the database in the file system.

- A sequence table containing sequence strings and related information like sequence type and the Genbank Accession Number. Sequence records are linked to the specimen table by the SBD-ID. This table also stores filenames of sequencing chromatograms and quality values, which will have to be submitted along with the sequence string to allow quality assessment. Chromatograms can be viewed from the query results page. - A separate table linked to the specimen table that contains relevant literature references e.g. for molecular applications of the sequence. (References to original descriptions and/or revisions are found in the WPD). - A user table and an edit history table. For any change made to a specimen record, an entry containing the specimen record number, the ID of the user that committed the change and a comment is inserted into the edit history. This allows tracking of any changes made to the specimen data. The splitting of the database into multiple tables increases its flexibility. A specimen record can have arbitrarily many associated sequences and reference entries without any redundant saving of information. A field submitted as in the specimen table contains the genus and species name as valid at the time of entry into the database. Additionally, the WPD-ID will be stored, and records displayed to the user after queries will dynamically load the current valid genus and species names from the WPD and display the currently accepted name in addition to the name as submitted. This allows records to still be found by their original name even after a name has been synonymized e.g. after a taxonomic

127

Fig. 3: Record #167 from the Sponge Barcoding Database, accessible via the data button on www.spongebarcoding.org. This is an example of a Reference record that contains all neccessary information (see text for details). Several subcategories (e.g., morphological description, reference, and associated DNA sequences can be enlarged to display their full content by clicking on show/hide). Accessed on 31 May 2007.

128

revision. The database currently displays two categories of records to the end user (additional maintenance categories are available to the editors): REFERENCE: records from described species with a full taxonomic description, DNA signature sequence(s), and verification of voucher material by a recognized taxonomic expert (Fig. 3). SUBMITTED: records from described species that either lack full taxonomic description or verification by a taxonomic expert, or DNA signature sequences from as yet undescribed and unverified species. Categorical Submitted records should only be used for comparative purposes and NOT for species identification. To avoid database inconsistency, the deletion of records needs to cascade through the database. If a specimen record is deleted, all associated reference and sequences must in turn also be deleted. The database does this automatically. Likewise, any image or chromatogram files are automatically deleted from the file system upon removal of their corresponding database entry. The databases user interface is written in PHP as a website. Web enabled data viewing and editing allows for flexible access to the database from many different platforms. The PHP interface also takes care of user authentication and error checking.

Acknowledgments
We acknowledge funding by the German Research Foundation (DFG). D.E. acknowledges financial support of the European Union under a Marie-Curie outgoing fellowship (MOIF-CT-2004 Contract No 2882).

References
Duran S, Rtzler K (2006) Ecological speciation in a Caribbean marine sponge. Mol Phylogenet Evol 40: 292-297 Erpenbeck D, Hooper JNA, Wrheide G (2006) CO1 phylogenies in diploblasts and the Barcoding of Life - are we sequencing a suboptimal partition? Mol Ecol Notes 6: 550553. Faulkner DJ (2000) Highlights of marine natural products chemistry (1972-1999). Nat Prod Rep 17:1-6 Hebert PD, Cywinska A, Ball SL, deWaard JR (2003b) Biological identifications through DNA barcodes. Proc Roy Soc Lond B Biol Sci 270: 313-321 Hebert PD, Ratnasingham S, deWaard JR (2003a) Barcoding animal life: cytochrome c oxidase subunit 1 divergences among closely related species. Proc Roy Soc Lond B 270(Suppl 1): S96-9

Hickerson M, Meyer C, Moritz C (2006) DNA barcoding will often fail to discover new animal species over broad parameter space. Syst Biol 55: 729-739 Hooper JNA, Ekins M (2005) Collation and validation of museum collection databases related to the distribution of marine sponges in northern Australia. Report to the National Oceans Office, Australia (Contract number C2004/020). pp. 235 Hooper JNA, van Soest RWM (2002) Systema Porifera: a guide to the classification of sponges. Kluwer Academic/Plenum Publishers, New York Klautau M, Russo CAM, Lazoski C, Boury-Esnault N, Thorpe JP, Sol-Cava AM (1999) Does cosmopolitanism result from overconservative systematics? A case study using the marine sponge Chondrilla nucula. Evolution 53: 1414-1422 Lesser MP (2006) Benthic-pelagic coupling on coral reefs: Feeding and growth of Caribbean sponges. J Exp Mar Biol Ecol 328: 277288 Lopez-Victoria M, Zea S (2005) Current trends of space occupation by encrusting excavating sponges on Columbian coral reefs. Mar Ecol 26: 33-41 Martindale MQ (2005) The evolution of metazoan axial properties. Nat Rev Genet 6: 917-927 Medina M, Collins AG, Silberman JD, Sogin ML (2001) Evaluating hypotheses of basal animal phylogeny using complete sequences of large and small subunit rRNA. Proc Natl Acad Sci USA 98: 9707-9712 Meyer CP, Paulay G (2005) DNA barcoding: error rates based on comprehensive sampling. PLoS Biology 3: e422 Moritz C, Cicero C (2004) DNA Barcoding: promise and pitfalls. PLoS Biology 2: e354 Munro MHG, Blunt JW, Lake RJ, Litaudon M, Battershill CN, Page MJ (1994) From seabed to sickbed: what are the prospects? In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 473-484. Shearer TL, Van Oppen MJH, Romano SL, Worheide G (2002) Slow mitochondrial DNA sequence evolution in the Anthozoa (Cnidaria). Mol Ecol 11: 2475-2487 Wrheide G (2006) Low variation in partial cytochrome oxidase subunit I (COI) mitochondrial sequences in the coralline demosponge Astrosclera willeyana across the Indo-Pacific. Mar Biol 148: 907-912 Wrheide G, Degnan BM, Hooper JNA (2000) Population phylogenetics of the common coral reef sponges Leucetta spp. and Pericharax spp. (Porfera: Calcarea) from the Great Barrier Reef and Vanuatu. In: Hopley D, Hopley P, Tamelander J, Done T (eds). Ninth International Coral Reef Symposium, Bali, vol. Abstracts, p. 21

RESEARCH ARTICLES

Porifera research: Biodiversity, innovation and sustainaBility - 2007

131

Spongivory by juvenile angelfish (Pomacanthidae) in Salvador, Bahia State, Brazil

Brbara R. Andra(1), Daniela Batista(1,2), Cludio L.S. Sampaio(3), Guilherme Muricy(1*)
Departamento de Invertebrados, Museu Nacional, Universidade Federal do Rio de Janeiro. Quinta da Boa Vista, s/no., So Cristvo. 20940-040 Rio de Janeiro, RJ, Brasil. muricy@acd.ufrj.br (2) Departamento de Biologia Marinha, Universidade Federal Fluminense, Niteri, Rio de Janeiro, Brasil (3) Departamento de Sistemtica e Ecologia, Universidade Federal da Paraba, Joo Pessoa, Brasil.
(1)

Abstract: Adult angelfish of the genera Pomacanthus and Holacanthus (Family Pomacanthidae) are considered the most important spongivorous fishes of the Caribbean, with sponges comprising more than 70% of their gut contents. However, despite their commercial importance as ornamental fish, little is known about the diet of juvenile angelfish, which are generally considered to be cleaners. The goal of this study was to identify through gut content analysis the sponge species eaten by juveniles of the angelfish Pomacanthus paru, Holacanthus ciliaris and Holacanthus tricolor in Salvador, Bahia state, Brazil. We also estimated the frequency of occurrence of each sponge species in the diet of juvenile angelfish, and tested the correlation between fish size and number of sponge species preyed. In Salvador, 34 species of sponges were found in the gut contents of 14 out of 16 specimens of juvenile angelfish. Twenty-two species of sponges were eaten by Holacanthus tricolor, 15 by H. ciliaris, and 14 by Pomacanthus paru. There was a significant positive correlation between fish size and number of preyed sponge species, but the coefficient of determination was low and even the smallest fishes had sponges in their gut contents. These findings indicate that juveniles of all three species of angelfish are generalists in the consumption of sponges in Brazil, a diet similar to that of the adults. The most frequent sponges in juvenile angelfish gut contents were Tedania ignis, Mycale sp., and Spirastrella sp., all with 37.5% of frequency in the 16 stomachs analyzed. Tedania ignis was not consumed by H. ciliaris in Salvador, but it was consumed by the other fish species, being the most frequent prey of H. tricolor and P. paru. Juvenile angelfish probably adapt to a cleaning behavior or to benthic feeding according to local environmental conditions. The development of sponge-based artificial foods may allow longer maintenance and reproduction of angelfish in aquaria, thus helping to protect their natural populations. Keywords: Holacanthus, Pomacanthus, Porifera, Predation, Southwestern Atlantic

Introduction
Sponges are important structural and functional components of coral reefs, where they participate in numerous ecological relationships such as competition, commensalism, symbiosis, bioerosion and predation (reviewed by Diaz and Rtzler 2001, Wulff 2001, 2006a). Although reef sponges are abundant and many species live exposed to predators, few animals are known to feed on them, probably due to their physical (spicules) and chemical (secondary metabolites) defenses (Pawlik et al. 1995, Hill et al. 2005, Wulff 2006a, 2006b). Among invertebrates, spongivory is shown by opisthobranchs (e.g., Glossodoris pallida, Archidoris montereyensis, Peltodoris atromaculata, Tylodina perversa), sea stars (e.g., Oreaster reticulatus, Echinaster echinophorus), sea urchins (e.g., Eucidaris tribuloides and Lytechinus variegatus) and hermit crabs (Reiswig 1973, Wulff 1995, 2006b, Becerro and Paul 1998, Wadell and Pawlik 2000, Santos et al. 2002, Gemballa and Schermutzki 2004). Some associated copepods, amphipods, isopods and alpheid shrimps also consume their host sponges (Pawlik 1983, Ros and Duffy 1999, Mariani and Uriz 2001). The main vertebrate sponge-feeders are the

sea turtle Erethmochelys imbricata (Meyland 1988) and reef fishes belonging to the families Pomacanthidae, Ostraciidae, Tetraodontidae, Ephippidae and Monacanthidae. Among these, the family Pomacanthidae contains the most important sponge-feeding fishes in the Caribbean, particularly in the genera Pomacanthus and Holacanthus (Randall and Hartman 1968, Wulff 1994). The family Pomacanthidae includes 88 species distributed in all tropical seas (Allen et al. 1998, Debelius et al. 2003). Within the family there is a diverse range of feeding specializations, including herbivory, planktivory, cleaning activity, and omnivory (Bhlke and Chaplin 1968, Hourigan et al. 1989, Allen et al. 1998, Deloach 1999, Bellwood et al. 2004). Ontogenetic variation in diet contents has been observed in many pomacanthids (Thresher 1980, Deloach 1999) as well as in other reef fish families such as Blenniidae, Kyphosidae and Scaridae (Bellwood 1988, Sturm and Horn 1998, Muoz and Ojeda 2000). In the Caribbean, sponges comprise over 70% of the diet of adults of the common pomacanthid species Pomacanthus paru (French angelfish), Pomacanthus arcuatus (Gray angelfish), Holacanthus ciliaris (Queen angelfish) and Holacanthus tricolor (Rock beauty)

132

(Randall and Hartman 1968, Dunlap and Pawlik 1996). In Brazil, adults of Pomacanthus paru also feed mostly on sponges and algae, in variable proportions according to the locality studied (Batista 2006). The feeding behavior of juveniles, however, appears to be different from that of adults in these species. Juveniles of H. ciliaris, P. paru and P. arcuatus may act as cleaners until they have approximately 10-15 cm in total length (Feder 1966, Thresher 1980, Deloach 1999, Sazima et al. 1999). The primary food of P. paru and P. arcuatus juveniles was reported to be filamentous algae, with copepods picked from client fishes and few free-living copepods making up to 25% of their diet (Deloach 1999). In Abrolhos Archipelago, Brazil, juveniles of the French angelfish P. paru removed ectoparasites of 31 species of reef fishes (Sazima et al. 1999). Food items in their stomachs included caligid (15-30%) and harpaticoid copepods (5-10%), together with both red and green algae (30-70%). Juveniles of H. ciliaris also feed on algae until they reach sexual maturity (Deloach 1999). The diet of H. tricolor juveniles is largely unknown, but it is assumed that it is a combination of drifting plankton, small benthic invertebrates, and possibly the mucus and parasites of larger fishes (Thresher 1979, 1980, Gasparini and Floeter 2001). Brazil is one of the five leading exporting countries of tropical aquarium fishes in the world, and the interest in marine ornamental organisms has increased substantially from mid- to late 1990s (Gasparini et al. 2005, Floeter et al. 2006). Juvenile angelfish are preferred over the adults for ornamental purposes due to their smaller size and beautiful color patterns. In Cear State, NE Brazil, angelfish juveniles have been captured and exported in large numbers: 43,730 specimens of H. ciliaris, 22,969 of P. paru, and 8,757 specimens of H. tricolor between 1995 and 2000 (Monteiro-Neto et al. 2003), although these values may be overestimated (Gasparini et al. 2005). Cleaner species play an important ecological role in reef habitats, and their removal may negatively affect other fish species, including commercially important ones (Sazima et al. 1999, Monteiro-Neto et al. 2003). Therefore, knowledge about the feeding behavior of juvenile angelfish is important for the conservation of angelfish species and of coral reef communities. In this study we tested the hypothesis that sponges are an important part of the diet of juvenile angelfish, and not only of the adults. We identified the sponge species found in the gut contents of juveniles of Pomacanthus paru, Holacanthus ciliaris and H. tricolor in Salvador, Bahia state, Brazil. We also estimated the frequency of each sponge species in the gut contents of juvenile angelfish, and tested the correlation between fish size and number of prey species. Our results may aid in the development of a better diet for the growth and reproduction of angelfish in captivity, thus helping to protect angelfish species which are currently threatened by the increase of marine ornamental fish trade in Brazil.

Fig. 1: Location of the study area.

Materials and methods

Sixteen specimens belonging to three species of angelfish were studied: Pomacanthus paru (n=5); Holacanthus ciliaris (n=6) and Holacanthus tricolor (n=5). Fish were collected in four locations in Salvador, Bahia state, Brazil (Fig. 1): Praia

da Ribeira, in Todos os Santos Bay (1254S3829W), Barra Grande, in Itaparica Island (1303S3838W), Rio Vermelho (1300S3830W), and Praia da Pituba (1300S 3830W). The specimens were collected from 18/II/2003 to 10/III/2004 by commercial ornamental fisheries (Ax Online Ltd.) in shallow reefs and rocky bottoms. Only juveniles were caught, easily recognized by their special color patterns, which change after sexual maturity is reached (Bhlke and Chaplin 1968, Thresher 1980). Specimen size varied between 5.0-14.5 cm in H. ciliaris, 5.0-14.0 cm in H. tricolor, and 3.0-14.5 cm in P. paru. The stomachs were removed and their contents were separated based on color, texture and consistency. Only the sponge fragments were identified to lower rank taxons, and no attempts were made to quantify the abundance of each species consumed in the gut contents of the 16 angelfish specimens analyzed. The sponge fragments were dehydrated in an alcohol series (50-100%) with a final xylene step and included in paraffin. Transverse sections were mounted on microscope slides for identification. Free spicules were not considered as evidence of predation, only fragments large enough to be sectioned and in which the skeleton could be observed. Sponges were deposited in the sponge collection of Museu Nacional, Universidade Federal do Rio de Janeiro, Brazil. A linear regression between fish size (standard length SL) and number of prey species was calculated online (http:// faculty.vassar.edu/lowry/VassarStats.htm).

133

Results
Sponges comprised more than 90% of the gut contents of juvenile angelfish in Salvador, together with a few unidentified filamentous algae (< 10%). The guts of two juveniles of H. ciliaris were empty. A total of 34 sponge species were found in the diet of angelfish in Salvador (Table 1). The greatest species richness was observed in gut contents of Holacanthus tricolor (22 sponge species), followed by H. ciliaris and Pomacanthus paru (15 and 14 sponge species, respectively). Twelve sponge species were only predated by H. tricolor: Acanthanchora sp., Coelosphaeridae unidentified, Desmapsamma anchorata, Lissodendoryx sp., Microcionidae unidentified 1, Microcionidae unidentified 2, Monanchora sp., Myxillina unidentified, Pachastrellidae unidentified, Phellodermidae unidentified, Plakinastrella sp., and Timea sp. Only three species were predated exclusively by Holacanthus ciliaris: Artemisina sp., Chalinidae unidentified, and Tethya sp. Six species were eaten exclusively by P. paru: Chondrilla nucula complex, Cyamon sp., Mycale laxissima, Niphatidae unidentified, Raspaciona sp., and Tetillidae unidentified (Table 1). As a whole, the sponge species most frequently predated

by pomacanthids in Salvador were Tedania ignis, Mycale sp., and Spirastrella sp., all present in 37.5% of the 16 specimens analyzed. Although Tedania ignis was not consumed by H. ciliaris, it was very frequent in the gut contents of P. paru (80% of the specimens examined) and common in H. tricolor (40%; Table 1). The number of sponge species per stomach varied from 4-9 in H. tricolor, from 1-7 in P. paru, and from 0-8 in H. ciliaris. The curves of accumulated richness of sponge species in angelfish diet did not reach stabilization with the small sample sizes studied here, neither with each fish species considered separately nor when they were considered together (Fig. 2). This indicates that the real number of sponge species in the diet of juvenile angelfish in Salvador is probably much higher than that shown by our results. There was a significant positive correlation (p<0.05) between fish size and number of sponge species in gut contents, although with a low coefficient of determination (r2 = 0.465; Fig. 3). The single smallest fish (3 cm SL) had only one sponge species in its stomach, and three of the five smallest fishes (3.0-5.5 cm SL) fed on few sponge species (15 species per stomach).

Table 1: Frequency of occurrence of sponge species in gut contents of juvenile angelfish in Salvador, Brazil (in %). Combined = frequency in all species pooled together (n=16). Sponge species Acanthanchora sp. Acervochalina sp. Artemisina sp. Chalinidae unidentified Chondrilla nucula complex Clathria sp. 1 Clathria sp. 2 Coelosphaeridae unidentified Cyamon sp. Desmapsamma anchorata Halichondriidae unidentified Haplosclerida unidentified Lissodendoryx sp. Microcionidae unidentified 1 Microcionidae unidentified 2 Mycale laxissima Mycale sp. Mycalidae unidentified Myxillina unidentified Monanchora sp. Niphatidae unidentified Pachastrellidae unidentified Phellodermidae unidentified Plakinastrella sp. Raspaciona sp. Spirastrella sp. Stelleta sp. 1 Stelleta sp. 2 Strongylacidon sp. Suberitidae unidentified Tedania ignis Tethya sp. Tetillidae unidentified Timea sp. Species richness P. paru 20 20 40 20 20 40 H. tricolor 20 H. ciliaris 17 17 17 33 17 Combined 6.25 12.50 6.25 6.25 6.25 25.00 25.00 6.25 6.25 12.50 12.50 31.25 6.25 12.50 6.25 6.25 37.50 25.00 12.50 12.50 6.25 6.25 6.25 6.25 6.25 37.50 12.50 12.50 12.50 12.50 37.50 6.25 6.25 6.25 34

40 20 20 40 20 20 20 20 20 60 40 40 20 20 20 20 20 20 20 40 20 22

50 17 50 17

20 40

20

20 60 20

80 20 14

33 17 17 17 17 17

15

134

Fig. 2: Curves of accumulated richness of sponge prey species in juvenile angelfish gut contents. A, Holacanthus tricolor; B, Holacanthus ciliaris; C, Pomacanthus paru; D, all species pooled together.

Discussion
Juvenile angelfish fed largely on sponges in Salvador (> 90%), and only a few filamentous algae were also found in their gut contents. These figures are similar to those of adult angelfish diet in the Caribbean, where sponges comprised 74.8% to 97.1% of the diet of Pomacanthus paru and Holacanthus ciliaris, respectively (Randall and Hartman 1968). Algae were also the second most common item in the diet of adult pomacanthids in the Caribbean, varying from 0.8% to 13.4% of the diet of H. tricolor and P. paru, respectively (Randall and Hartman 1968). Other food items common in Caribbean adults such as tunicates, hydroids, zoantharians, and bryozoans were not found in juveniles from Salvador. Fourteen out of 16 fishes (87.5%) collected in one year had sponge remains in their stomachs, indicating that spongivory is not occasional, but frequent and widespread in the juvenile angelfish population in Salvador. Adult angelfish appear to have a more varied diet, at least in the Caribbean, where sponges were found in only 12 of 23 stomachs of P. paru (52.1%) and 18 out of 34 specimens of P. arcuatus (52.2%). Adults of H. tricolor are more specialized in sponges, with sponge remains in 22 of 24 specimens (91.7%; Randall and Hartman 1968). The diet of juvenile angelfish in Salvador includes at least 34 sponge species. All three angelfish species studied here, Pomacanthus paru, Holacanthus ciliaris, and H. tricolor

were generalists in the consumption of sponges, eating 14 to 22 different species each. The real number of sponge species consumed by these angelfish is probably higher than that, as indicated by the shape of the accumulated richness curves (Fig. 2). Although there was a significant positive correlation between fish size and number of preyed sponge species (Fig. 3), the coefficient of determination was low and even the smallest fishes had sponges in their gut contents. Part of this correlation may reflect greater gut volume in larger fish individuals, allowing them to have more bites (and therefore more species) in their guts at the time they were collected. The diet of pomacanthid juveniles in Salvador is very similar to that of the adults, both in the Caribbean and in Brazil: they are smorgasbord-feeding sponge specialists, consuming a variety of benthic invertebrates and algae, apparently chosen according partly to prey palatability and partly to prey availability in the environment (Randall and Hartman 1968, Hourigan et al. 1989, Wulff 2006a, Batista 2006). This is at odds with some studies which suggest that juveniles of H. ciliaris and P. paru are cleaners until they have approximately 10-15 cm in total length, when they change to benthic feeding of a mainly sponge and algae diet (Feder 1966, Bhlke and Chaplin 1968, Hourigan et al. 1989, Sazima et al. 1999). Other studies however found little cleaning activity in juveniles of P. paru (Wicksten 1995, 1998). Among Tropical Western Atlantic angelfish, juveniles of P. paru seem to be the most specialized as cleaners, with P. arcuatus and H. ciliaris having a mixed diet composed mostly of algae and detritus

135

Fig. 3: Linear regression between fish size (standard length in cm) and the number of sponge species in gut contents of juvenile angelfish. Slope = 1.1173, intercept = 5.0347, Standard error of the estimate = 3.169.

and being only occasional cleaners (Thresher 1980, Deloach 1999). The diet of H. tricolor juveniles was hitherto poorly known, although cleaning activity has been reported in a few other studies (Thresher 1979, 1980, Gasparini and Floeter 2001). The method of study employed by each author is important to explain such disagreements. When only direct underwater observation is considered (e.g., counting the bites given in each prey species), there is a trend to increase the relative importance of algae and of cleaning activity in the diet of pomacanthids (Bhlke and Chaplin 1968, Thresher 1980, Hourigan et al. 1989, Sazima et al. 1999, Batista 2006). Juveniles of angelfish such as H. tricolor dwell mostly in cryptic habitats inaccessible for divers (Thresher 1980). These habitats are often dominated by sponges (Sar and Vacelet 1973). Other species of angelfish, including adult specimens, also use small caves and reef crevices for sheltering and foraging, biasing the direct observation of preyed items towards more exposed organisms such as algae and gorgonians (Thresher 1980, Batista 2006). When only gut contents are analyzed, the relative importance of spongivory seems to increase, both in proportion to algae and benthic invertebrates and in number of species (e.g., Randall and Hartman 1968, Deloach 1999). It is possible that sponge fragments remain recognizable in the guts of angelfish longer than copepods and filamentous algae, due to their high collagen and spicule content (Chanas and Pawlik 1995). This would artificially increase the proportion of spongivory in studies based only in gut contents and without direct observations, such as the present one (see also Randall and Hartman 1968). The complementary use of both field observations and gut content analysis allows a better understanding of pomacanthid feeding habits (Hourigan et al. 1989, Sazima et al. 1999, Batista 2006, Wulff 1994, 2006a). It is possible therefore that juvenile angelfish in Salvador are not almost exclusively sponge-eaters as shown by our results,

but might also consume other benthic organisms and act as cleaners occasionally. This issue can only be solved through systematic direct observations on juvenile angelfish, which could not be carried out in this study. Another explanation for the variation in the diet of pomacanthid juveniles in different studies is a high dietary plasticity of pomacanthid species. Allopatric angelfish populations may specialize to use different resources, depending on their availability in each locality. For example, the abundance of reef fish in general is greater in the Abrolhos National Marine Park, which is a protected area, than in Salvador, which has been subject to heavy fishing (Gasparini et al. 2005, Floeter et al. 2006). It may thus be easier for juveniles of pomacanthids such as P. paru to establish a cleaning station in Abrolhos than in Salvador, where they have to adapt their diet to the available resources, mainly sponges and algae. The presence of a suitable habitat for juveniles close to a reef fish community also helps to explain the intense cleaning activity of P. paru juveniles in Abrolhos reefs (Sazima et al. 1999). Although always composed mostly of sponges, the specific composition of the diet of adult angelfish varied strongly in different localities such as the Bahamas, Virgin Islands, and Panama in the Caribbean (Randall and Hartman 1968, Feddern 1968, Hourigan et al. 1989, Wulff 1994), and Atol das Rocas, Abrolhos and Ilha Grande in Brazil (Batista 2006), irrespective of the study method. In all these locations and in Salvador, each species of angelfish fed on a diverse array of sponge species, ranging from 12 in P. paru to 40 in H. ciliaris, both in the U.S. Virgin Islands (Randall and Hartman 1968, Hourigan et al. 1989). Together, Pomacanthus paru and P. arcuatus ate 64 sponge species in Panama (Wulff 1994). The only exception was H. tricolor, which was observed feeding on a single sponge species in Panama and was therefore considered a possible specialist (Wulff 1994). However, adults of H. tricolor are known to fed on 14 to 28 different sponge species in the U.S. Virgin Islands and Puerto Rico (Randall and Hartman 1968, Hourigan et al. 1989), and juveniles in Salvador fed on 22 sponge species. This indicates that H. tricolor is also a generalist in the consumption of sponges, like most other pomacanthids. The number of sponge species eaten by each individual fish in Salvador varied from 0-9 (average 4.5). This tolerance to predate upon diverse sponge species may be a strategy of angelfish to overcome the effects of sponge toxins by eating small amounts of each toxic substance from many different sponge species (Wulff 1994). It has also be suggested as an adaptation of benthic fish species with pelagic larvae such as angelfish to increase their chances of survival in areas with different availability of specific food items (Hourigan et al. 1989). The similarity between gut contents of different but co-specific specimens was low in Salvador (maximum 0.273 between two specimens of H. tricolor; data not shown), indicating that individual fish choose their food independently. This may be related to the very small home range of angelfish juveniles (around 1-2 m2), much smaller than that of the adults (up to 2,300 m2 in P. paru; Thresher 1980, Hourigan et al. 1989). All specimens studied here came from the same general region in Salvador, but dietary similarity between the three species was low (maximum 0.296 between H. ciliaris and H.

136

tricolor; data not shown) and in the same range of intraspecific similarity. Most of the interspecific variation found in angelfish diet in Salvador thus appears to be random, and due to the sum of independent choices of individual fishes of the three different species studied. As most fishes, spongivores have great plasticity in their feeding behavior. This variability could have influence from food availability, fish size, species and other temporal and spatial factors. However, part of the differences in dietary composition between sympatric species is probably due to an active choice of preys, possibly restrained by the physiological tolerance of each fish species to the preys toxins. A certain degree of choice in pomacanthids is supported by observations of predation upon relatively rare sponge species (Hourigan et al. 1989, Wulff 1994). Holacanthus passer often feeds on plankton and fish faeces in the water column in the eastern Pacific (Aburto-Oropeza et al. 2000); there are few exposed sponges on coral reefs there, but when cryptic sponges were exposed the angelfish quickly swam down out of the water column to feed on them (Wulff 1997). Caribbean angelfish also prefer to eat mangrove or cryptic sponge species whenever they are made available in the reefs (Dunlap and Pawlik 1996, Wulff 2005). In general, spongivorous fish tend to choose the most palatable or undefended prey species available in a given locality (Dunlap and Pawlik 1996), although which prey is more palatable varies according to the predator species (Wulff 2006a). For instance, the fire-sponge Tedania ignis is commonly eaten by H. tricolor and P. paru, but apparently not by H. ciliaris, both in Brazil and in the Caribbean (Table 1; see also Randall and Hartman 1968). The starfish Oreaster reticulatus also feeds on Tedania ignis in Belize, but not on the sibling species T. klausi (Wulff 2006b). Also, the dietary similarity between congeneric species of both Pomacanthus and Holacanthus is greater than between species from different genera (Table 1; see also Hourigan et al. 1989). Whether this truly represents a species-specific choice of prey by pomacanthids remains to be experimentally demonstrated. Both adult and juvenile angelfish present high plasticity, tolerance, and a certain degree of choice in their smorgasbord sponge-feeding specialized strategy. This appears to be a highly derived feeding strategy, which is coupled with morpho-functional modifications in the teeth, jaws, and gill rakers that make pomacanthids (especially Holacanthus and Pomacanthus) more apt to feed on structurally resilient and firmly attached benthic prey such as sponges, gorgonians and tunicates (Hourigan et al. 1989, Bellwood et al. 2004; Konow and Bellwood 2005). Sponges have been evolving during at least 580 MY (Li et al. 1998), and they developed chemical and physical defenses so effective that only very specialized predators such as angelfish and opisthobranchs are able to feed on them. This may help to explain why there are so few predators of sponges, despite their great abundance in coral reefs. Angelfish are threatened by marine ornamental fish trade in Brazil, and juveniles are the targets preferred by aquarists (Gasparini et al. 2005, Floeter et al. 2006). The maintenance of angelfish in aquaria is difficult due to their aggressiveness and their diet made up mostly of sponges (Thresher 1980). The development of artificial food based on the sponge species found in this study to be eaten by angelfish juveniles in the

field might help their reproduction and survival in captivity, allowing aquarists to enjoy angelfish in their saltwater tanks without eliminating them from coral reefs, where they play important ecological roles.

Acknowledgements
We thank Jos de Anchieta Nunes, Camilo Ferreira and Ericka Coni for laboratory assistance. We are also grateful to Samuele Clerici of Ax Online Fishes for the kind donation of specimens for study. We thank the two anonymous reviewers for their comments, which greatly improved the manuscript. This study was supported by grants and fellowships from Conselho Nacional de Desenvolvimento Cientfico e Tecnolgico (CNPq) and Fundao Carlos Chagas Filho de Amparo Pesquisa do Estado do Rio de Janeiro (FAPERJ).

References
Aburto-Oropeza O, Sala E, Snchez-Ortiz C (2000) Feeding behaviour, habitat use, and abundance of the angelfish Holacanthus passer (Pomacanthidae) in the southern Sea of Corts. Environ Biol Fish 57: 435-442 Allen GR, Steene R, Allen M (1998) A guide to angelfishes and butterflyfishes. Odyssey Publishing, Perth Batista D (2006) Espongivoria por Pomacanthus paru Bloch, 1787 (Perciformes: Pomacanthidae) na costa brasileira. MSc Thesis, Universidade Federal Fluminense, Niteri Becerro MA, Paul VJ (1998) Intra-colonial variation in chemical defenses of the sponge Cacospongia sp. and its consequences on generalist fish predators and the specialist nudibranch predators Glossodoris pallida. Mar Ecol Prog Ser 168: 187-196 Bellwood DR (1988) Ontogenetic changes in the diet of early postsettlement Scarus species (Pisces: Scaridae). J Fish Biol 33(2): 213-219 Bellwood DR, Herwerden LV, Konow N (2004) Evolution and biogeography of marine angelfishes (Pisces: Pomacanthidae). Mol Phylog Evol 33: 140-155 Bhlke JE, Chaplin CCG (1968) Fishes of the Bahamas and adjacent tropical waters. Livingston Publications, Wynnewood Chanas B, Pawlik JR (1995) Defenses of Caribbean sponges against predatory reef fish. II. Spicules, tissue toughness, and nutritional quality. Mar Ecol Prog Ser 127: 195-211 Debelius H, Tanaka H, Kuiter RH (2003) Angelfishes, a comprehensive guide to Pomacanthidae. TMC Publishing, Chorley Deloach N (1999) Reef fish behavior Florida, Caribbean, Bahamas. New World Publications, Jacksonville Diaz MC, Rtzler K (2001) Sponges: an essential component of Caribbean coral reefs. Bull Mar Sci 69(2): 535-546 Dunlap M, Pawlik JR (1996) Video-monitored predation by Caribbean reef fishes on an array of mangrove and reef sponges. Mar Biol 126: 117-123 Feddern HA (1968) Systematics and ecology of Western Atlantic angelfishes, family Chaetodontidae, with an analysis of hybridization in Holacanthus. PhD Thesis, University of Miami, Coral Gables Feder HM (1966) Cleaning symbiosis in the marine environment. In: Henry SM (ed). Symbiosis. Vol. 1. Academic Press, New York. pp. 327-380 Floeter SR, Halpern BS, Ferreira CEL (2006) Effects of fishing and protection on Brazilian reef fishes. Biol Conserv 128: 391-402

137

Gasparini JL, Floeter S R (2001) The shore fishes of Trindade Island, southwestern Atlantic. J Nat Hist 35 (11): 1639-1656 Gasparini JL, Ferreira CEL, Floeter SR, Sazima I (2005) Marine ornamental trade in Brazil. Biodiv Conserv 14: 2883-2899 Gemballa S, Schermutzki F (2004) Cytotoxic haplosclerid sponges preferred: a field study on the diet of the dotted sea slug Peltodoris atromaculata (Doroidoidea: Nudibranchia). Mar Biol 144: 12131222 Hill MS, Lopez NA, Young KA (2005) Anti-predator defences in Western North Atlantic sponges with evidence of enhanced defence through interactions between spicules and chemicals. Mar Ecol Prog Ser 291: 93-102 Hourigan TF, Stantion FG, Motta PJ, Kelley CD, Carlson B (1989) The feeding ecology of three species of Caribbean angelfishes. Environ Biol Fishes 24(2): 105-116 Konow N, Bellwood DR (2005) Prey-capture in Pomacanthus semicirculatus (Teleostei, Pomacanthidae): functional implications of intramandibular joints in marine angelfishes. J Exp Biol 208: 1421-1433 Li CW, Chen JY, Hua TE (1998) Precambrian sponges with cellular structures. Science 279: 879-882 Mariani S, Uriz MJ (2001) Copepods of the genus Asterocheres (Copepoda: Siphonostomatoida) feeding on sponges: behavioral and ecological traits. Invertebr Biol 120: 269-277 Meyland A (1988) Spongivory in hawksbill turtles: a diet of glass. Science 239: 393-395 Monteiro-Neto C, Cunha FEA, Nottingham MC, Arajo ME, Rosa IL, Barros GML (2003) Analysis of the marine ornamental fish trade at Cear sate, northeast Brazil. Biol Conserv 12: 1287-1295 Muoz AA, Ojeda FP (2000) Ontogenetic changes in the diet of the herbivorous Scartichthys viridis in a rocky intertidal zone in central Chile. J Fish Biol 56: 986-998 Pawlik JR (1983) A sponge-eating worm from Bermuda: Branchiosyllis oculata (Polychaeta, Syllidae). Mar Ecol 4: 65-79 Pawlik JR, Chanas B, Toonen RJ, Fenical W (1995) Defenses of Caribbean sponges against predatory reef fish I. Chemical deterrency. Mar Ecol Prog Ser 127: 183-194 Randall JE, Hartman WD (1968) Sponge-feeding fishes of the West Indies. Mar Biol 1(3): 216-225 Reiswig HM (1973) Population dynamics of three Jamaican Demospongiae. Bull Mar Sci 23: 191-226 Ros R, Duffy JE (1999) Description of Synalpheus williamsi: a new species of sponge-dwelling shrimp (Crustacea: Decapoda: Alpheidae), with remarks on its first larval stage. Proc Biol Soc Wash 112: 541-552

Santos CP, Coutinho AB, Hajdu E (2002) Spongivory by Eucidaris tribuloides from Salvador, Bahia. J Mar Biol Assoc UK 82: 295297 Sar M, Vacelet J (1973) cologie des dmosponges. In Grass PP (ed). Trait de Zoologie, vol. 3, Spongiaires. Masson, Paris. pp. 462-576 Sazima I, Moura RL, Sazima C (1999) Cleaning activity of juvenile angelfish, Pomacanthus paru, on the reef of the Abrolhos Archipelago, western South Atlantic. Environ Biol Fishes 56: 399407 Sturm EA, Horn MH (1998) Food habits, gut morphology and pH, and assimilation efficiency of the zebra-perch Hermosilla azurea, an herbivorous kyphosid fish of temperate marine waters. Mar Biol 132: 515-522 Thresher RE (1979) Possible mucophagy by juvenile Holacanthus tricolor (Pisces: Pomacanthidae). Copeia 1979: 160-162 Thresher RE (1980) Reef fish: behavior and ecology on the reef and in the aquarium. John Bartholomew and Sons, Edinburg Waddell B, Pawlik JR (2000) Defenses of Caribbean sponges against invertebrate predators. II. Assays with sea stars. Mar Ecol Prog Ser 195: 133-144 Wicksten MK (1995) Associations of fishes and their cleaners on coral reefs of Bonaire, Netherland Antilles. Copeia 1995: 477481. Wicksten MK (1998) Behavior of cleaners and their client fishes at Bonaire, Netherland Antilles. J Nat Hist 32: 13-30 Wulff JL (1994) Sponge feeding by Caribbean angelfishes, trunkfishes and filefishes. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 265-271 Wulff JL (1995) Sponge-feeding by the starfish Oreaster reticulatus. Mar Biol 123: 313-325 Wulff JL (1997) Causes and consequences of differences in sponge diversity and abundance between the Caribbean and eastern Pacific at Panama. Proc 8th Int Coral Reef Symp, Balboa 2: 1377-1382 Wulff JL (2001). Assessing and monitoring coral reef sponges: why and how? Bull Mar Sci 69(2): 831-846 Wulff JL (2005) Trade-offs in resistance to competitors and predators, and their effects on the diversity of tropical marine sponges. J Anim Ecol 74: 313-321 Wulff JL (2006a). Ecological interactions of marine sponges. Can J Zool 84: 146-166 Wulff JL (2006b) Sponge systematics by starfish: predators distinguish cryptic sympatric species of Caribbean fire sponges, Tedania ignis and Tedania klausi n. sp. (Demospongiae, Poecilosclerida). Biol Bull 211: 83-94

Porifera research: Biodiversity, innovation and sustainaBility - 2007

139

Growth and morphology of a reef-forming glass sponge, Aphrocallistes vastus (Hexactinellida), and implications for recovery from widespread trawl damage
William C. Austin(1*), Kim W. Conway(2), J. Vaughn Barrie(2), Manfred Krautter(3)
Marine Ecology Centre & Khoyatan Marine Laboratory, 9835 Seaport Place, Sidney, B.C., Canada. baustin@mareco.org Pacific Geoscience Centre, Geology Survey of Canada, P.O. Box 6000, Sidney, B.C., Canada. KConway@nrcan.gc.ca, vbarrie@nrcan.gc.ca (3) Institut fr Geologie und Palontologie der Universitt Stuttgart, Herdweg 51, D-70174, Stuttgart, Germany. Manfred.krautter@geologie.uni-stuttgart.de
(1) (2)

Abstract: Living hexactinellid reefs, known only on the western Canadian shelf, are being damaged by dredging and trawling. Recovery of damaged or destroyed hexactinellid reefs depend on many interrelated factors including sponge larval settlement and survival, sponge growth rates and the balance between suspended sediment trapping by the sponges and smothering by sedimentation. In this paper we present our work on one species found on the reefs, Aphrocallistes vastus Schulze 1887, measuring growth rates (approximately 300 cm2 yr-1 in surface area), the sizes of larger sponges (up to 3.4 m long x 1.1 m high x 0.5 m wide), indicators of successful recruitment (low based on occurrence of only one small individual in study site), and the form sponges can take under various environmental conditions. A. vastus sponges are very fragile and one was observed to die after breaking in two. Broken sponges have been observed on trawled reefs. Keywords: Aphrocallistes, form, growth, Hexactinellida

Introduction
Hexactinellid sponge reefs are found widely distributed on the western Canadian shelf in both Georgia and Queen Charlotte Basins of British Columbia. Multibeam surveying has provided accurate maps of sponge reef distribution in water depths of 100 to 240 m. The largest reef complex is roughly 40 kilometers long and has been growing for about 9,000 years with bioherms commonly up to 15 m in height (Conway et al. 1991). These reefs form a stable and complex habitat for many species of invertebrates and fish. However, bottom trawling has damaged reefs in most areas (Conway et al. 2000, Krautter et al. 2001). Aphrocallistes vastus is a dominant species in all sponge reefs surveyed (Conway et al. 1991, Conway et al. 2000). However, it also regularly occurs in BC fjords at depths as shallow as 25 m (Leys et al. 2004) and in one location at 5 m (Austin 1999, 2003). Determination of predicted tide levels has shown that some of these sponges actually occurred at only 2 m below 0 datum (Austin 2007). A. vastus together with two other species (Heterochone calyx, Farrea occa) can be considered as foundation species in the sense that they are the reef formers and generate a complex hard substrate habitat supporting a diverse biota in contrast to the surrounding soft level bottom substrate (Conway et al. 1991, Austin pers. observ.). Given their role as foundation species, and the demonstrated impacts of

trawling over the reefs (Conway et al. 2001), there has been much interest in assessing growth rates and recruitment as important determinants of recovery rates. However, estimates of minimum growth rates have only been inferred from observed sizes of A. vastus on a pipeline, cables, and a sunken vessel available for settlement for a known number of years. (see discussion for details). Shallow and accessible populations of Aphrocallistes vastus are present in Saanich Inlet, a fjord near Victoria, BC (Austin 1984, Leys et al. 2004) allowing these sponges to be studied as proxies for the sponges of the vast, but remote, northern reefs. Divers could directly observe, measure and photograph sponges over time. Here we present the results to date on growth, form, size, and recruitment at Senanus Reef in Saanich Inlet.

Material and methods Growth

Increase in mass of the sponge wall would be the most direct measure of growth but is not feasible for in situ measurements. However, the wall thickness is nearly constant for most of the sponge so that increase of surface area can approximate growth. Weighted, numbered floats were placed adjacent to sponges selected for subsequent growth measurements.

140

Initially five sponges were selected with a surveyors meter rod over time to assess growth; the very irregular form of four of these precluded accurate measurements of surface area. One small individual was ideal for measurement. It included a solid non-growing phallus-like cylinder and a growing fairly flat lobe. Multiple photos were taken while looking at a photo taken previously so that photos could be taken from the same position and distance relative to the sponge. A surveyors rod was held at the same position adjacent to the sponge. The sponge was photographed on eight dates over a period of four years. The outline of the sponge was drawn on transparent acetate which was slightly enlarged or reduced as necessary to fit the outline of the non-growing phallus portion as well as the decimeter spacing on the surveyors rod. The acetate was overlain on graph paper and the area was calculated by summing the squares within the outline. The area was also calculated from digitized images.

total surface area is, therefore, approximately 33,840 cm2 or 3.38 m2 ([948.4 g/0.55 g] x 19.625 cm2). Divers measured some of the larger sponges (Fig. 4, Table 1). Most of the sponges in the Senanus Reef population are on the order of one meter in size and typically have a very convoluted surface which amplifies the surface area considerably. Between 30 and 60 such sponges would be seen on a dive. Divers also looked for sponges smaller than approximately 5 cm in height or width. Six divers were able to find only one small sponge over the course of their dives (about 20 minutes) in March and September 2003.

Form
Aphrocallistes vastus individuals can take many forms in addition to the moderately dense mitten form dominating Senanus Reef (Fig. 5A). In an area of very high currents (Seymour Narrows) A. vastus has a very dense compact form (Fig. 5B). On Senanus Reef a few sponges form a tall narrow cylinder with sparse lateral mittens (Fig. 5C). This growth form dominates on some sponge reefs along with chalices (Fig 5D). Tubes with no mittens (Fig 5E) infrequently occur in fjords. Divers made counts of the number of sponges with living and dead bases. Seventy nine percent of 28 sponges over 30 cm in height had dead bases (Fig. 6).

Surface area of large sponge

Divers harvested one moderate size sponge to determine surface area of the sponge wall. In A. vastus the wall thickness remains constant throughout the body except for the edges of the mittens and of the oscula. Given that the thickness of the sponge wall remains constant, the ratio of surface area to weight should be approximately the same throughout the sponge. This sponge measured 0.65 m long x 0.48 m in diameter and occupied a space of about 0.12 m3. The sponge was air dried for 4 months.

Discussion Growth
We consider that surface area increases are a good measure of biomass changes as the thickness of the fused sponge wall is approximately the same throughout the sponge. The soft parts where growth occurred were limited to the ends of branches or of the main tube. Wulff (1990) reported that growth was only at branch apices in 3 demosponges she studied. The measurements are a relatively small underestimate as the rounded edges would be directed obliquely at the camera. An additional underestimate is likely for the surface area of the sponge taken at year 3 as there are some slight bulbous lateral expansions on the edges. The limited data, if reasonably accurate, indicate a modest increase in growth rate with increased size. If the growth rate prior to year 0 approximates that measured for year 0 to year 1, then extrapolating back (dotted line in Fig. 3) the sponge first settled about 1 years earlier. The straight lines on the graph are for visualization of overall growth rate. The lines would actually be slightly curved upward. At this time we only have growth rates for one sponge and for the first 5 years of age. Changes in a linear dimension (height or width) are often used as a measure of growth. Given that this dimension may bear little relation to increase in mass (see e.g. the different growth forms in Fig. 5), for purposes of comparison we included the length along the greatest axis when we first measured the sponge on year 0 and when we measured it again 4 years later. This was a difference of 39.8 cm or an average increase in size along one dimension of 10 cm year-1. Austin (2003) reported on the height of A. vastus sponges that had settled on a gas pipeline that had been in place for 9 years in the Strait of

Form
The forms of A. vastus sponges were observed and photographed from a range of habitats.

Results Growth
The appearance of the measured sponge over a 4 year period is shown in Fig. 1. The area of the A. vastus sponge when first measured, with additions one year later and 3 years later are shown in Fig. 2. In A. vastus the framework spicules are fused together along the main stem and at least on the sides of mitten shaped lateral projections. The edges of the mittens may be soft as are the edges of the oscula. Increase in surface area only occurred in the soft unfused areas of the sponge. The surface area (2 sides) was 369 cm2 when first measured; 661 cm2 one year later and 1,631 cm2 3 years later (Fig. 3). No attempt was made to assess surface area in the sponge photographed after 4 years as much of the surface was directed obliquely at the camera. However, the greatest linear dimension increased from 10.2 cm in year 0 to 50.0 cm in year 4

Surface area of large sponge

The large sponge collected by the divers weighed 948.4 g. Six 5 cm diam. wall plugs taken from different regions of the sponge were weighed. Weights ranged from 0.44 g to 0.74 g (mean 0.55 g for plugs with a surface area of 19.625 cm2. The

141

Fig. 1: Size of phallus form of A. vastus Sept. 12, 2002 (A), Sept. 13, 2003 (B), Feb. 18, 2006 (C), and Sept. 10, 2006 (D).

Georgia (Secret Cove). The largest was 64 cm tall, equivalent to an average growth rate of 7 cm year-1 if the sponge settled on the pipe when it was first installed. If this sponge had a linear growth rate comparable to that measured in our tagged sponge (10 cm year-1) and our tagged sponge was 1 year old, then the largest pipeline sponge could have settled 3 years after the pipeline was installed, and was 6 years old when measured. Leys et al. (2007) report the linear growth rate of a small Aphrocallistes vastus of approximately 1-3 cm year-1 based on photographs of the same sponge taken 1 year apart. They report massive changes in shape with growth including the flanges or projections and the location of the osculum. Our observations as noted above are that no changes occurred in the fused portions except for repair of damage to small areas of the surface. We suggest that some of the changes seen between the two images in Leys et al. (2007) are due to differences in perspective as the two images appear to be rotated about 45o apart when viewed from above the sponges.

Fig. 2: Area of A. vastus sponge when first measured, with additions one year later and 3 years later.

142

Fig. 3: Growth rate of A. vastus sponge over 3 years (solid line) with estimate of time when first settled (dotted line).

Table 1: Sizes of large A. vastus Length 3.4m 2.7m 2.1m 1.4m 1.4m Height 1.1m 1.0m 1.0m 2.0m 0.8m Width 0.5m 1.4m 0.4m 0.6m 0.8m

Fig. 4: Typical form of a giant A. vastus sponge.

There have been a few studies reported on growth rates in other species of hexactinellids. Dayton (1979) monitored Antarctic populations of 3 hexactinellid species over periods of 3 to 10 years. Scolymastra joubini and Rosella nuda showed no evidence of growth except for slight growth of 2-3 cm over 3-10 years in 3 individuals. However, small individuals of Rossella racovitzae increased in volume up to 292% over 3 years and larger individuals of the same species increased in length by 11 to 16 cm over 10 years. Marliave (1992) measured the length of Rhabdocalyptus dawsoni (an approximately cylindrical hexactinellid) over a period of about 6 months at the entrance to a British Columbia fjord (Howe Sound). He found that large (30-100 cm) sponges grew less than 20% in length while small (2-3.5 cm) sponges grew up to 71% in length. Leys and Lauzon (1998) measured the change in length and in volume of Rhabdocalyptus dawsoni in the same fjord as that used in the present study. They found the average growth in length over 3 years was 1.98 cm year-1 with a minimum of 0.76 cm year-1 and a maximum of 5.7 cm

year-1. Gatti (2002) estimated the age of Rossella spp. based on respiration rates. She used these rates to model growth in AMIGO [Advanced modeling of invertebrate growth from Oxygen consumption]. The average size Rossella spp. were 186 years old and the largest was 1515 years old based on the model. The hexactinellid sponges studied by the above authors belong to a different order from that of Aphrocallistes, have a quite different type of skeleton and generally have a cylindrical rather than a branching growth form. Given these differences, the growth rate in greatest dimension for our individual Aphrocallistes vastus averaging 10 cm year -1 is significantly greater than that found in other hexactinellids to date. The total surface area (3.68 m2) in the harvested sponge is about 20 times that of our measured sponge when estimated to be 5 years old. If the growth rate increase in our measured sponges holds true for large individuals, the harvested sponge would be about 1 century old. The largest sponge measured was 30 times the size of our harvested sponge based on overall height, width and depth. We will not speculate on the age of this sponge, but hope to obtain some age estimates in the coming year. A 1 m high Rhabdocalyptus dawsoni, in the same inlet was estimated to be 220 years old based on

143

Fig. 5: Various forms of A. vastus sponge. Typical in fjord (A); in area of high water currents (B); atypical skeleton sponge (C); on sponge reef (D); atypical form in fjord (E).

144

Fig. 6: Living A. vastus sponge with dead base.

measured volume increases for small individuals (Leys and Lauzon 1998). Given that divers were able to find only one small (< 5cm) sponges after a total of 2 hours searching, we conclude that recruitment was nil or very low over the period of the study. Divers report large numbers of small individuals at another site about 9 km away. Leys et al. (2007) report that of the many specimens of NE Pacific Aphrocallistes vastus collected since the early 1980s, developing embryos were found only once.

above the substrate so is less subject to clogging from buildup or re-suspension of sediment. We speculate that, perhaps, rather than living material dying at the base, the syncytium might move apically. Syncytial streaming has been demonstrated in microscopic preparations of Rhabdocalyptus dawsoni tissue sandwiched between slide and coverslip (Leys and Mackie 1994). The rate of streaming was 2 m sec-1, which is equivalent to 7 mm hour-1. Cytoplasmic streaming at a macroscopic level might be inferred from the opening and closing of spaces in the soft portions of the osculum noted by Austin (2003). The rate or frequency of apical streaming could be in response to impacts from sedimentation. Hence, sponge reefs subject to moderate sedimentation might grow up more slowly than those with high sedimentation. Such a mechanism could help explain how sponges avoid burial but also do not grow high so fast that they become unstable. For such a strategy to be successful the base and its attachment must be structurally sound during the life of the sponge. A. vastus may live, at least, many decades. Maldonado et al. (2005) have demonstrated that unlike diatoms, the silica in sponge spicules (including one hexactinellid species) showed little or no dissolution after acid cleaning and submersing in water low in silicates over an 8 month period. There is some support for very slow dissolution in the field from observations of hexactinellid sponge bases attached to the wall of Saanich Inlet. These are at a depth where the water has been anoxic for at least many decades (Levings et al. 1983).

Form
Aphrocallistes vastus is very fragile. Most of the wall has a texture and friability of a thin slice of toast. An erect growth form is suitable in the deep sea, and in most areas of a fjord where currents are minimal. The low compact form shown in Fig. 5B is likely an adaptation to the strong currents in the area. The mittens in Fig. 5A amplify the surface area and their largely vertical orientation minimizes clogging by sediment as discussed elsewhere (Austin 2003). The tall thin sponge (Fig. 5C) has been dubbed the skeleton sponge by divers. It occurs in the same habitat as the more typical sponges. Simple tubes (Fig. 5E) and chalices (Fig. 5D) may be a response to some environmental factor. For example, the entire inhalant surface would be facing obliquely down in a chalice form where sediment would not accumulate. We can only surmise that sediment would be blown off the upper exhalent surface. Some demosponges vary their form during growth. Halichondria panicea, for example has a low encrusting form which becomes stiffer and stronger in high wave/current energy environments, while in low current energy environments it has an erect ramose form and a more pliant, weaker skeleton (Palumbi 1986, Barthel 1991). The number of moderate sized sponges with dead bases (79%) was surprising. One interpretation of the data is that those with dead bases are dying which, if correct, indicates a major die-off of sponges at our study site. Another interpretation is that concentrating growth above a dead base has adaptive value. The filtering portion of the sponge is

Conclusions
What are the implications of our observations on impacts of trawling over A. vastus populations? The fragile nature of these sponges and their high profile dictates that they would most certainly be broken by a trawl and such damage has been observed on the reefs (e.g., Conway et al. 2001). Once broken, several lines of evidence indicate that they would likely not survive. When small areas are broken or removed by a hole saw, regeneration repairs the damage (Austin 2003). However, divers found a sponge sliced in two (likely by fishing line) (Austin et al. unpublished), and it subsequently died. If a sponge were knocked on its side by a trawl the broad surface of the mittens would be horizontal resulting in accumulation of sediment on the upper surface. Similarly, the tubular form common in sponge reefs, if knocked over, would be buried in sediment on its lower side and subject to sedimentation on its upper side. If the growth rate measured in our study is representative of growth rates on a sponge reef, then juveniles could reach a moderate size in a decade or two. However, many decades to perhaps a century or more would elapse before an old growth reef was fully developed. Recruitment of juveniles is also a key factor. No small (<5 cm) sponges have been found on Senanus reef while they may be common at a site about 9 km away. Finally, A. vastus is one of two (Strait of Georgia reefs) or three (northern reefs) reef forming sponges. Nothing is known about the growth of the other species. However, their fragility is comparable to that of A. vastus (Austin pers. observ).

145

Acknowledgements
Our thanks to Paula Romagosa, who worked on the graphics and who with Katya Austin reviewed the manuscript for spelling, grammatical and typographical errors, Sherry Ward, Coastal & Ocean Resources Inc., who helped on digitizing images, Jonathan Grant & Fred Holmes vessel operators, and the key contributions by members of the Victoria Dive Club: Mike Miles, Neil Lake, Tom Dakin, Doug Bifford, Parris Champoise, Joe Doiron, Carole Valkenier Pope, Ian Pope, Al Truby, Andy Murch, Sandie Hankewich, Doug Campbell, Mike Kalina, James Dranchuk, Mark Gottfried, Fred Peters, and David Willis. Also, our thanks to two anonymous reviewers for their helpful suggestions.

References
Austin WC (1984) Underwater birdwatching. In: Juniper SK, Brinkurst RO (eds). Proceedings of a multidisciplinary symposium on Saanich Inlet. Canadian Tech Rept Hydrogr Ocean Science 38. pp 104 Austin WC (1999) The relationship of silicate levels to the shallow water distribution of hexactinellids in British Columbia. Memoir Queensl Mus (abstract) 44: 44 Austin WC (2003) Sponge gardens. A hidden treasure in British Columbia. http://mareco.org/khoyatan/spongegardens (accessed on September 23, 2006) Austin WC (2007) Sponge gardens update. A hidden treasure in British Columbia. http://mareco.org/khoyatan/spongegardens Barthel D (1991) Influence of different current regimes on the growth form of Halilchondria panicea Pallas. In: Reitner J, Keupp H (eds). Fossil and recent sponges. Springer-Verlag, Berlin. pp. 387-394 Conway KW, Barrie JV, Austin WC, Luternauer JL (1991) Holocene sponge bioherms on the western Canadian continental shelf. Cont Shelf Res 11: 771-790 Conway KW, Krautter M, Barrie JV, Austin WC, Neuweiler M (2000) Extant hexactinellid sponge reefs: our endangered seafloor heritage. Abstracts GeoCanada, May 29-June 2, 2000. Calgary Conway KW, Krautter M, Barrie JV, Neuweiler M (2001) Hexactinellid sponge reefs on the Canadian continental shelf: a unique living fossil. Geoscience Canada 28(2): 71-78

Dayton PK (1979) Observations of growth, dispersal and population dynamics of some sponges in McMurdo Sound, Antarctica. In: Lvi C, Boury-Esnault N (eds). Biologie des spongiaires. Colloques Internationaux du CNRS, vol. 291. ditions du CNRS, Paris. pp. 271-282 Gatti S. (2002) The role of sponges in high-Antarctic carbon and silicon cycling a modeling approach. Ber Polarforsch Meeresforsch 434: 1-102 Krautter M, Conway KW, Barrie JV, Neuweiler M (2001) Discovery of a living dinosaur; globally unique modern hexactinellid sponge reefs off British Columbia, Canada. Facies 44: 265-282 Levings CD, Foreman RE, Tunnicliffe VJ (1983) Review of the benthos of the Strait of Georgia and contiguous fjords. Can J Fish Aq Sci 40: 1120-1141 Leys SP, Lauzon NRJ (1998) Hexactinellid sponge ecology: growth rates and seasonality in deep water sponges. J Exp Mar Biol Ecol 230: 111-129 Leys SP, Mackie GO (1994) Cytoplasmic streaming in the hexactinellid sponge Rhabdocalyptus dawsoni (Lambe 1873). In: van Soest RWM, van Kempen TMG, Braeckman J (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 417-423 Leys SP, Wilson K, Holeton C, Reiswig HM, Austin WC, Tunnicliffe VJ (2004) Patterns of glass sponge (Porifera, Hexactinellida) distribution in coastal waters of British Columbia. Canada. Mar Eco Prog Ser 283: 133-149 Leys SP, Mackie GO, Reiswig HM (2007) The biology of glass sponges. Adv Mar Biol 52: 1-145 Marliave JB (1992) Environmental monitoring through natural history research. Canadian Tech Rpt Fish Aq Sci 1879: 199-209 Maldonado M, Carmona M, Velsquez Z, Puig A, Cruzado A, Lpez A, Young CM (2005) Siliceous sponges as a silicon sink: an overlooked aspect of benthopelagic coupling in the marine silicon cycle. Limnol Oceanogr 50(3): 799-809 Palumbi SR (1986) How body plans limit acclimation responses of a demosponge to wave force. Ecology 67(1): 208-214 Wulff JL (1990) Patterns and processes of size change in Caribbean demosponges of branching morphology. In: Rtzler K (ed). New perspectives in sponge biology. Smithsonian Institution Press, Washington. pp. 425-434

Porifera research: Biodiversity, innovation and sustainaBility - 2007

147

Symbiotic relationships between sponges and other organisms from the Sea of Cortes (Mexican Pacific coast): same problems, same solutions
Enrique vila(1,2*), Jos Lus Carballo(1), Jos Antonio Cruz-Barraza(1,2)
Laboratorio de Ecologa del Bentos, Instituto de Ciencias del Mar y Limnologa. Universidad Nacional Autnoma de Mxico, Estacin Mazatln, Avenida Joel Montes Camarena S/N, Apartado Postal 811, Mazatln 82000, Mxico. Tel. +52 669 985 28 45. Fax: +52 669 982 61 33. kike@ola.icmyl.unam.mx (2) Posgrado en Ciencias del Mar y Limnologa, UNAM, Mazatln, Mxico
(1)

Abstract: This study provides a morphological description of three symbiotic associations between sponges (Haplosclerida) and other macroorganisms from the Sea of Cortes (Mexican Pacific Ocean). These associations include: (1) a two-sponge association (Haliclona sonorensis/Geodia media), (2) a sponge-red macroalga association (Haliclona caerulea/Jania adherens) and (3), a sponge-coral association (Chalinula nematifera/Pocillopora spp.). So far these interactions seem to be obligatory for the sponges (Haliclona spp. and C. nematifera), since we have never found them living in isolation. Interestingly, similar associations have been described from other places around the world. Associations quite similar to (1) have been described from the Caribbean, and associations (2) and (3) are comparable to others described from the Western Pacific. Instead of comparing these associations with their sibling associations worldwide, we discussed the ability of haplosclerid sponges to form symbiotic associations with other organisms, since these sponges pertain to the group of which the most associations have been described. Keywords: Sea of Cortes, sponge-alga, sponge-coral, symbiotic associations, two-sponge

Introduction
Sponges are one of the major phyla found in the hardsubstrate marine benthos (Sar and Vacelet 1973). One of their more interesting characteristics is that they are able to establish a great diversity of relationships (mutualism, commensalism and parasitism) with unicellular and multicellular organisms (Palumbi 1985, Rtzler 1990, Magnino and Gaino 1998, Ilan et al. 1999, Wulff 1999, Thakur and Mller 2004). Most of these relationships have been reported in the Caribbean (West 1979, Rtzler 1990, Wulff 1997a, 1997b, 1999, Wilcox et al. 2002), Mediterranean Sea (Uriz et al. 1992, Gaino and Sar 1994), Red Sea (Meroz and Ilan 1995, Ilan et al. 1999), Indo-Pacific region (Steindler et al. 2002, Calcinai et al. 2004) and Western Pacific (Vacelet 1981, Trautman et al. 2000). Surprisingly, in a large number of cases, the same species are involved in similar interactions in different oceans (Ilan et al. 1999, Wulff 2006). For example, the sponges Haliclona caerulea Hechtel, 1965, Haliclona cymaeformis Esper, 1794 and Dysidea janiae (Duchassaing and Michelotti, 1864) are species found to establish similar interactions with red macroalgae (Vacelet 1981, Rtzler 1990, Trautman et al. 2000, Carballo and vila 2004). However, it is unknown whether this group of cosmopolitan associations occurs in the East Pacific Ocean.

The order Haplosclerida (Demospongiae) is the group in which the most of the symbiotic relationships have been registered with the family Chalinidae Gray, 1867 accounting for more than 50% of the associations described in this order (see discussion). Sponges of this order can establish associations with microorganisms such as bacteria (Vacelet et al. 2001), cyanobacteria (Steindler et al. 2002), dinoflagellates (Garson et al. 1998) and zoochlorellae (Frost and Williamson 1980), and with macroorganisms such as polychaetes (Dauer 1974), macroalgae (Vacelet 1981), mangroves (Ellison et al. 1996), barnacles (Ilan et al. 1999), hydrozoans (Schuchert 2003), anthozoans (West 1979), ophiurids (Henkel and Pawlik 2005), insects (Gaino et al. 2004) and other sponges (Wilcox et al. 2002). In the present study, we describe three interactions involving haplosclerid sponges from the Sea of Cortes (Mexican Pacific Ocean). They are a two-sponge association (Haliclona sonorensis-Geodia media), a sponge-alga association (Haliclona caerulea-Jania adherens) and a sponge-coral association (Chalinula nematifera-Pocillopora spp.). We discuss the surprising parallelism that exists worldwide, given that very similar interactions occur in different oceans. In addition, we comment on the characteristics that this group in which more symbiotic relationships have been registered has in order to establish these symbiotic associations.

148

Material and methods

Specimens of the three associations were collected by SCUBA diving from 2 to 12 m depth, in four localities from the Sea of Cortes (Eastern Pacific Ocean, Mexico) (Fig. 1). We followed the techniques described by Rtzler (1974) for spicule and tissue preparations for light microscopy. Crosssections of the specimens were washed in distilled water and then dried on a cover glass and coated with gold for scanning electron microscope (SEM) observations. A number of 20 to 50 spicules chosen randomly were measured (length x width) from each of the specimens studied. The number between brackets in each description is the average. After the description, the specimens were fixed in formaldehyde 4% and after 24 h they were transferred to 70% alcohol for their preservation. All the specimens were deposited in the Coleccin de Esponjas of the Instituto de Ciencias del Mar y Limnologa, UNAM (LEB-ICML-UNAM), in Mazatln (Mexico). In the Haliclona sonorensis/Geodia media association, the surface of G. media covered by H. sonorensis was estimated. First we took photographs of the specimens, then we determined the covered area (%) using the computer program Coral Point Count with Excel extensions (CPCe) (Kohler and Gill 2006). The frequency of the C. nematifera/Pocillopora spp. association was determined in three transects of 50 m length at a depth between 4 to 6 m. In each one of these transects we chose 20 colonies of coral at random, and determined the percentage of these containing sponge. In the same area, we also checked if the sponge was on another type of substratum. For each sample of the association we determined the species of coral and estimated the percentage of the colony overgrown by the sponge as number of branches with sponge of the total of branches.

Results
Association Haliclona sonorensis Cruz-Barraza and Carballo, 2006 Geodia media Bowerbank, 1873 Material examined. Eleven specimens of the association were collected between 2 and 5 m depth in two localities from the northern Sea of Cortes: Punta Cazn (Baha Kino, Sonora, 285220 N, 1120201 W), and Punta Pinta (La Choya, Puerto Peasco, Sonora, 311805 N, 1135911 W) (Fig. 1), from August 2000 to April 2005. Description of the species involved in the association. The epizoic sponge was identified as Haliclona sonorensis CruzBarraza and Carballo, 2006, which is a cushion-shaped sponge from 2 to 5 mm in thickness (Fig. 2A). Consistency is soft, compressible, but fragile and brittle. The surface is smooth and the ectosomic layer is not easily detachable. The color is pinkish violet in life and ocre or light brown in alcohol. The oscules are scarce, and circular or oval-shaped (from 0.5 to 1 mm in diameter), situated at the summits of volcano-shaped elevations. The skeletal material is constituted by oxeas that measure: 77-(112)-150 x 2-(5.6)-10 m. The supporting sponge was identified as Geodia media Bowerbank, 1873. This is a massive-incrusting to massive amorphous sponge (from 3.5 to 8 cm thick). The surface is

irregular, smooth to the naked eye, but finely rough to the touch. The natural color of the surface is from dark-brown to white. The choanosome is yellowish or beige. Small ostialpores from 150 to 300 m are regularly distributed on the surface. The oscules are contained in several small, circular or oval-shaped well-defined flattened sieves (containing from 7 to more than 100 oscules). The oscules measure from 0.22 to 2.5 mm in diameter. Consistency of the ectosome is very hard due to the cortex of sterrasters. The choanosome is cavernous and very densely spiculated, with a firm and slightly compressible consistency. The skeletal material is constituted by megascleres: oxeas, 620-(1430)-1950 x 10(31)-42 m; large styles, 620-(1077)-1260 x 22-(36)-45 m; strongyloxeas, 150-(197.2)-292 x 2.5-(4.9)-7.5 m; plagiotriaenes, 550-(1078)-1700 m rabdome length; and anatriaenes, 1120-(1410)-2040 m rabdome length, and microscleres: sterrasters, 25-(62.8)-90 m oxyasters, 20(27.2)-45 m; oxyspherasters, 6.3-(9.5)-13 m. Description of the association. The specimens of the association were found attached to the rocky substrata, covering areas from 8 x 6.5 to 20 x 15 cm, approximately. H. sonorensis forms a thin layer that covers up to 57 % of the surface of G. media, while the surface that is not covered by the sponge is occupied by other epibionts (green and red algae, bryozoans, polychaetes and bivalves). Only the oscular areas (from 1 to 4 cm in diameter) are free of these epibionts (Fig. 2A). In some cases, more than one individual of H. sonorensis was observed on a same specimen of Geodia, which was evident by their different tonality of coloration. Despite being interwoven, Haliclona was unattached in some areas, where we observed that the external tissue of Geodia did not seem to be damaged by the epizoic sponge (Fig. 2B, C). The area of G. media lacking epibionts has a rough texture due to the external layer of sterrasters (Fig. 2C, D), but the SEM showed that the megascleres (triaenes and oxeas) of G. media protrude the surface in the areas covered by Haliclona (Fig. 2C, F) penetrating in the Haliclona sonorensis tissue (Fig. 2D, E). There were spicules (oxeas) of H. sonorensis inside the ostias and embedded in the choanosome of G. media, and there were also sterrasters of G. media in the choanosome of H. sonorensis. Haliclona sonorensis has been invariably found living on the surface of G. media which suggests that this species needs to live in association with G. media. Association Chalinula nematifera (de Laubenfels, 1954) Pocillopora spp. Lamarck, 1816 Material examined. A total of ten specimens of this association were collected between 3 and 12 m depth in three sites from Isabel Island (Baha Tiburones, Playa Las Monas and Playa Iguanas), Nayarit, Mexico (285220 N, 1120201 W) (Fig. 1), from December 2003 to July 2006. Morphological description of the species in the association. The epizootic sponge has been identified as Chalinula nematifera (de Laubenfels, 1954). This is an encrusting sponge of violet color (1-4 mm thickness). This sponge grows only on live corals found in Isabel Island (Fig. 3). The surface is smooth to the naked eye, but it is punctated and shaggy in some places. Oscules are abundant, circular, from 4 to 6 mm

149

Fig. 1: Sampling localities (letters). The numbers indicate the site where specimens of each sponge association were collected: (1) two-sponge association, (2) sponge-alga association and (3) sponge-coral association.

in diameter, and regularly distributed on the surface. They are situated at the summits of volcano-shaped elevations. Consistency is soft and spongy, somewhat elastic and slimy. The skeleton consists of oxeas: 87-(98)-112.5 x 2.5-(4.4)-5 m. Our specimens also show the characteristic pale threads through the body as described by de Laubenfels (1954), which presumably is a symbiotic fungus (WF Prudhomme van Reine, comments in de Weerdt 2002). The coral species on which C. nematifera was found were identified as Pocillopora damicornis Linnaeus, 1758, P. meandrina Dana, 1846, P. capitata Verrill, 1864 and P. verrucosa Ellis and Solander, 1786. In general, these coral species have characteristic shape because they form densely ramified colonies. They have calices crowded together over regularly-spaced wart-like projections (verrucae) (P. capitata and P. verrucosa). Description of the association. C. nematifera was found always on live ramified corals that live in areas exposed to light, most frequently with P. verrucosa (67%), and never on another type of substratum. Nevertheless, it is possible to find these species of coral (mentioned above) without the sponge.

Approximately 17% of the Pocillopora colonies studied had C. nematifera in association. In these sponge/coral interactions, C. nematifera can cover branches partially or totally (55.12% of the total of branches of the colony), and in all these cases we observed that the surface of the covered coral has no polyps. This sponge adhered firmly to the coral and is not easy to detach it from the substratum without breaking it. In fact, through the SEM images, we observed that the skeletal structure of the sponge seems to be cemented to the coral septa by spongin layers (Fig. 3C). Association Haliclona caerulea Hechtel, 1965 Jania adherens Lamouroux, 1816 Material examined. Sixteen specimens of the sponge-alga association were collected from ten sites from the Mazatln Bay (231349 N, 1062743 W), Sinaloa, Mexico (Fig. 1), between 2 to 6 m depth, from November 1997 until October 2003. Morphological description of the species in the association. Haliclona (Gellius) caerulea is a massive sponge (from

150

Fig. 2: Haliclona sonorensis Geodia media association. A. two-sponge association containing several epibionts on its surface except on the osculate area. B. cross section of a specimen showing the Geodia surface almost totally covered by Haliclona sonorensis. D. SEM image showing the surface contact of the two interacting sponges, and C, E, F. the megascleres of Geodia protruding its ectosome, which are used as anchorage for the external sponge. The arrow in F shows an ostium of the internal sponge. Scale bars: A and B= 2 cm, C= 5 mm, D= 500 m, E= 200 m, F= 500 m.

3 to 13 cm high), white or beige in life and very brittle. The skeleton is constituted by oxeas (82.5-(177.3)-210 m) and sigmas (17.5-(21.6)-30 m). The sponge has an unispicular ectosomal skeleton, formed by an isotropic tangential reticulation of oxeas, and the choanosomal skeleton is a somewhat confused reticulation of uni-multispicular primary and secondary lines that are difficult to appreciate because of their association with the alga (Fig. 4B). The alga was identified as Jania adherens Lamouroux, 1816, which is an articulated erect red macroalga. The alga is

pink with white joints, repeatedly branched, with a calcified thallus (from 0.4 to 0.5 mm diameter) except at the genicula. Description of the association. This sponge lives in intimate association with the red calcareous alga Jania adherens. The association consists of a massive and compact form where the sponge completely fills the spaces between the algal branches (Fig. 4A). The sponge generally covers the alga, and the algal branches very rarely protrude beyond the association surface. The morphology of the association is derived from the growth form of both organisms. Specimens of this association are lo-

151

Fig. 3: A. Chalinula nematifera Pocillopora spp. association. B. C. nematifera tissue in close contact with the coral polyps (arrows). C. SEM image showing the skeletal structure of the sponge cemented to the coral septae (arrows). Scale bars A= 1 cm, B= 2000 m, C = 100 m.

cally abundant in rocky substrate environments (2-6 m depth), in areas of strong wave force. However, in these places, it is not possible to find the two species living in isolation, even though J. adherens lives in isolation in the intertidal zone. Jania adherens forms compact turfs approximately 2 cm high in the intertidal zone. However, in the subtidal zone (in association with H. caerulea) it reaches up to 13 cm in height. In the images obtained by SEM, we observed that the spicules of H. caerulea adhere firmly to the stalks of J. adherens by means of a fine spongin layer (Fig. 4C). We also observed that the primary lines of the sponge are partially replaced by the macroalgal thallus.

Discussion Two-sponge association

Interactions among two or more sponges of different species, including cases of mutualism (Wilcox et al. 2002), parasitism (Wulff 1999) or space competition (Rtzler 1970, Wulff 2006) have been previously recorded. A two-sponge association quite similar to ours was described recently from the Florida Keys (Wilcox et al. 2002). In both cases, as it has also been documented for other two-sponge associations (Sim 1998, Wilcox et al. 2002),

152

Fig. 4: A. Haliclona caerulea Jania adherens association. B. SEM images showing the skeletal structure of H. caerulea, which are partially substituted by the J. adherens branches in the association, and C. a close up of a spicule secondary line adhered to the algal thallus by means of a fine spongine layer. Scale bars A= 1 cm, B= 300 m, C= 60 m.

Haliclona sp. seems to be anchored on the protruding spicules of Geodia spp. There are studies that suggest that growth and survival increases when certain sponges of different species live in association, because they have different susceptibility to factors such as burial by sediment, fragmentation, predation and pathogens (Wulff 1997a, 1999, Engel and Pawlik 2005, Wulff 2006). However, in the association from the Florida Keys, it was suggested that it could be a mutualistic and obligatory symbiosis, where the external sponge protects its host from the predation, while Geodia sp. offers it a sure substratum for its survival (Wilcox et al. 2002). However, it is interesting to comment that the Caribbean sponge Geodia gibberosa is chemically undefended against turtle and fish predation (Dunlap and Pawlik 1996, 1998), and uses its chemical products to attract fouling organisms. This could resemble Geodia sp. in association with Haliclona sp. (Wilcox et al. 2002), since one of the benefits acquired by the sponge Geodia cydonium fouled by the red alga Rytyphlea tinctoria is the protection against the adverse effects of ultraviolet radiation, allowing the specimens to live under illuminated habitats (Mercurio et al. 2006). It is possible that an obligatory relationship also exists between H. sonorensis and G. media, at least for H. sonorensis, since it has not been found in isolation within the environment where the associations are encountered. In fact, the high specificity of Haliclona to live with Geodia is very interesting because this association does not originate in a highly space-competitive environment like reef ecosystems, and therefore it could not be a simple case of epizoism due to space limitation (Rtzler 1970). In addition, it is important to mention that there is no evidence of tissue damage on the surface of G. media covered by H. sonorensis, as it has been documented in other two-sponge interactions (Thacker et al.

1998). Although we did not make experiments to test whether a benefit exists between G. media and H. sonorensis, bioactive natural products have been described in G. media from the Sea of Cortes (Pettit et al. 1981, Pettit et al. 1990), that could attract fouling organisms, similar to Geodia gibberosa (Engel and Pawlik 2000, Engel and Pawlik 2005). In the association described here, one or more specimens of H. sonorensis are found attached to the same G. media specimen, establishing contact but without fusing. These observations support the idea that this sponge most likely colonizes G. media through larval settlement, in order to obtain an appropriate substratum for its survival (Wulff 1997b, Wilcox et al. 2002).

Sponge-coral association
Sponge-coral interactions are common in coral reefs where strong competition for space exists and where it is frequent that aggressive sponges overgrow the coral (Aerts 1998). The sponge Chalinula nematifera has been described previously overgrowing hard corals in reefs from the WestCentral Pacific (de Laubenfels 1954). Surprisingly, this species also seems to be specifically attracted to pocilloporid corals from the Sea of Cortes, since it has not been found colonizing other types of substratum. Nevertheless, the species of coral associated with this sponge can be found in an isolated form. This suggests that it could be an obligatory relationship for the sponge, and facultative for the host (Pocillopora spp.). We also suggest that this relationship can have negative outcome for the coral, since the sponge seems to be killing the coral tissue, similar to what has been documented for several sponge/coral interactions (Jackson and Buss 1975, PlucerRosario 1987, Macintyre et al. 2000, Aerts 2000, Rtzler 2002, de Voogd et al. 2004, Coles and Bolick 2006, Gochfeld

153

et al. 2006). However, we do not know if C. nematifera uses chemical products to kill the coral polyps (Nishiyama and Bakus 1999) or if it simply smothers them (Wulff 1999). Some authors have suggested that the ability of sponges to overgrow corals appears to depend on their growth rate and form (Aerts 2000). For example, in the sponge Terpios hoshinota, it was suggested that its success to overgrow large extensions of live corals is due to a fast spreading rate, aided by fast asexual propagation (fragmentation) (Plucer-Rosario 1987, Rtzler and Muzik 1993). Chalinula nematifera possibly finds a substratum that offers multiple protected microhabitats in pocilloporid corals, since they are densely ramified species which are often used as microhabitats by several organisms (Patton 1974). Thus, C. nematifera could likely benefit by finding an appropriate substratum that offers protection against environmental factors (e.g. UV radiation, sedimentation and/or predation). This is an advantage that it could not find with the other non-branched coral species (Porites, Psammocora, Pavona and Fungia) that inhabit the same site as the C. nematifera/ Pocillopora spp. association, most likely because they do not offer this kind of physical protection. Although in most of the sponge-coral interactions that have been described, it has been found that these relationships seem to be negative for the coral (de Voogd et al. 2004, Coles and Bolick 2006, Gochfeld et al. 2006, Lpez-Victoria et al. 2006), there are also documented cases of mutualisms; for example, the sponge Mycale laevis Carter 1882 which is found encrusting on the lower surfaces of flattened reef corals (mainly on Montastrea annularis) in the Caribbean Sea. M. laevis benefits by colonizing a substrate that is free from other sessile organisms. In turn, the coral benefits from an increased feeding efficiency as a result of water currents produced by the sponge and it is protected from invasion by boring sponges (Goreau and Hartman 1966). In the case of sponge/octocoral associations, the sponge generally obtains structural support from its partner and the octocoral obtains protection against predation (van Soest 1987, Calcinai et al. 2004). In addition, it has been documented that both organisms also benefit by increasing their dispersal capacity (Calcinai et al. 2004).

and vila 2004, Carballo et al. 2006). However, in spite of the similarities of these two interactions, the H. caerulea/ J. adherens association does not live in an oligotrophic environment in the Bay of Mazatln (see Carballo 2006). Our investigations into the H. caerulea-Jania adherens association suggests that this is an obligatory and mutualistic association, where both organisms benefit by increased growth, widespread their distribution area toward an environment where none can inhabit by itself, and by the acquisition of a structural support that gives them bigger rigidity than their free-living form (vila and Carballo 2004, Carballo and vila 2004, Carballo et al. 2006). H. caerulea can also substitute part of its skeletal structure (mainly primary lines) with the branches of J. adherens, reducing the investment in spicule production (Carballo et al. 2006), as it has been documented in other sponge symbioses (de Laubenfels 1950, Vacelet 1981, Rtzler 1990, Uriz et al. 1992). In addition, it has been demonstrated that H. caerulea in association with J. adherens (from the Pacific coast of Panama) acquires the benefit of being protected against fish predation (Wulff 1997a). On the other hand, H. caerulea has never been found in isolated form at the depth range where the sponge-alga association lives in the Bay of Mazatln, because it has a very fragile structure, and it is unable to live in that environment, which is characterized by strong water movement (Carballo et al. 2006). Although this association reproduces mainly by fragmentation, larval settlements of H. caerulea have been registered in the intertidal zone where J. adherens inhabits in isolated form (vila and Carballo 2006). In fact, laboratory experiments demonstrated that the larvae of H. caerulea can select J. adherens as substratum while rejecting others (vila and Carballo 2006), probably because the alga fronds offer a shady and tangled microrefuge, which could increase the post-settlement survival (Buss 1979, Maldonado and Uriz 1998). On the other hand, it is important to add that this association has also been found in other localities in Mexico, such as Tuxpan, Veracruz (in the Gulf of Mexico) and Manzanillo, Colima and Huatulco, Oaxaca (in the Mexican Pacific) (personal observations).

Sponge-alga association
Associations between sponges and photosynthetic organisms are important not only for the partners, but also for the ecosystem they inhabit, because they can contribute significantly to the primary productivity, mainly in oligothrophic ecosystems (Wilkinson 1983, Steindler et al. 2002). One of the most-studied sponge/macroalga symbioses is the Haliclona cymaeformis/Ceratodictyon spongiosum association (from the Great Barrier Reef, Australia) (Trautman et al. 2000, Trautman and Hinde 2002, Davy et al. 2002), which we could consider as a sibling association of the H. caerulea/J. adherens association described here. In both cases the mutualistic association derives similar benefits for the partners (such as protection against physical disturbances, structural support, high dispersal capacity through fragmentation), and in both cases the sponge is a species of Haliclona that lives associated with a red macroalga (Trautman et al. 2000, Trautman and Hinde 2002, Carballo

The diversity of associations among the haplosclerid sponges

The three symbiotic sponges (H. sonorensis, C. nematifera and H. caerulea) described in this study belong to the order Haplosclerida, which is one of the most diverse groups of the phylum Porifera (Hooper and van Soest 2002). Upon revising the literature, we have found that most of the sponges (21% of 248 cases) that have been previously recorded establishing associations with other taxa (see introduction), belong to the order Haplosclerida, and mostly to the family Chalinidae (50% of 51 registrations) (Fig. 5) (Dauer 1974, Vacelet 1981, Ellison et al. 1996, Garson et al. 1998, Ilan et al. 1999, Vacelet et al. 2001, Steindler et al. 2002, Wilcox et al. 2002, among others). These associations have been recorded mainly in shallow tropical and subtropical environments (such as coral reefs), which are characterized by a high competition for space and food by the organisms that inhabit there. It seems that some sponges have probably learned to associate with

154

Fig. 5: A. Symbiotic associations between sponges and other organisms registered in the different orders of the class Demospongiae and B. in the order Haplosclerida.

other organisms that offer them some kind of protection against environmental factors (such as fish predation, sedimentation, UV radiation, tissue resistance). For example, in the Geodia sp./Haliclona sp. association, it has been suggested that Geodia sp. is chemically defended against fish predation and protected from sedimentation by the sponge partner (Wilcox et al. 2002). In another haploclerid species such as Haliclona cymaeformis and Haliclona implexiformis that live associated with photosynthetic organisms (macroalga and mangrove respectively), it has been demonstrated that nutrient translocation also exists between the sponge and its host (Ellison et al. 1996, Davy et al. 2002). In the case of species that shelter cyanobacteria (e.g. Adocia atra and Haliclona debilis) in their tissue, it has been suggested that they can benefit by protecting their surface from UV-radiation and/or obtain an alternative source of food (Rtzler 1990, Steindler et al. 2002). It has also been suggested that they reinforce their skeletal structure using their partner to avoid being broken into fragments by the current, as is the case of H. caerulea (Carballo et al. 2006) and H. cymaeformis (Trautman and Hinde 2002). On the other hand, there are many bioactive compounds known in haplosclerid sponges (Frincke and Faulkner 1982, Isaacs and Kashman 1992, van Soest and Braekman 1999), which can also be used for different purposes such as antifoulants (Sera et al. 2002), as antimicrobial compounds (Xue et al. 2004, Ely et al. 2004) or in competition for space (Nishiyama and Bakus 1999).

Conclusions
In general, it seems that in the sponge/sponge and sponge/macroalga associations the relationships are usually mutualistic, or at least no type of damage has been evidenced among the associated organisms. In contrast, in most of the non-boring sponges associated to ramified corals that have been described (including C. nematifera/Pocillopora spp. association), the relationship seems to be negative for the coral species. Nevertheless, it is necessary to do more experimental and population dynamics studies of these associations in order to understand more about the complexity of their life history and their ecological importance in the ecosystem they inhabit.

Acknowledgements
We are grateful to the following sources of funding: CONABIO FB666/S019/99, CONABIO FB789/AA004/02, CONABIO DJ007/26 and CONACYT SEP-2003- C02-42550. This work was carried out under permission of SAGARPA (Permit numbers: DGOPA.02476.220306.0985 and DGOPA.06648.140807.3121). We thank Clara Ramrez, Pedro Allende and Victoria Montes for help with the literature and images, German Ramrez and Carlos Surez for their computer assistance, Cayetano Robles, Gonzalo Prez and Cristina Vega for their assistance in the field sampling, and the Director Parque Nacional Isla Isabel Jorge Castrejn for the availability and the permission conferred for the collection of the samples in Isabel island.

155

References
Aerts LAM (1998) Sponge/coral interactions in Caribbean reefs: analysis of overgrowth patterns in relation to species identity and cover. Mar Ecol Prog Ser 175:241-249 Aerts LAM (2000) Dynamics behind standoff interactions in three reef sponge species and the coral Montastraea cavernosa. PSZN Mar Ecol 21: 191-204 vila E, Carballo JL (2004) Growth and standing stock biomass of a mutualistic association between the sponge Haliclona caerulea and the red alga Jania adherens. Symbiosis 36(3): 225-244 vila E, Carballo JL (2006) Habitat selection by larvae of the symbiotic sponge Haliclona caerulea (Hechtel, 1965) (Demospongiae, Haplosclerida). Symbiosis 41: 21-29 Buss LW (1979) Bryozoan overgrowth interactions-the interdependence of competition for space and food. Nature 281: 475-477 Calcinai B, Bavestrello G, Cerrano C (2004) Dispersal and association of two alien species in the Indonesian coral reefs: the octocoral Carijoa riisei and the demosponge Desmapsamma anchorata. J Mar Biol Assoc UK 84: 937-941 Carballo JL (2006) Effect of natural sedimentation on the structure of tropical rocky sponge assemblages. Ecoscience 13: 119-130 Carballo JL, vila E (2004) Population dynamics of a mutualistic interaction between the sponge Haliclona caerulea, and the red alga Jania adherens. Mar Ecol Prog Ser 279: 93-104 Carballo JL, vila E, Enrquez S, Camacho L (2006) Phenotypic plasticity in a mutualistic association between the sponge Haliclona caerulea and the calcareous macroalga Jania adherens induced by transplanting experiments. I: morphological responses of the sponge. Mar Biol 148: 467-478 Coles SL, Bolick H (2006) Assessment of invasiveness of the orange keyhole sponge Mycale armata in Kneohe Bay, Oahu, Hawaii. 7th Int Sponge Symp (abstract) pp. 110 Cruz-Barraza JA, Carballo JL (2006) A new species of Haliclona (Demospongiae: Haplosclerida) living in association with Geodia media Bowerbank (Mexican Pacific coast). Zootaxa 1343: 43-54 Davy SK, Trautman D, Borowitzka MA, Hinde R (2002) Ammonium excretion by a sponge and its ecological importance in a spongerhodophyte symbiosis. J Exp Biol 205: 3505-3511 Dauer DM (1974) Polychaete fauna associated with Gulf of Mexico sponges. Florida Scientist 36: 192-196 de Laubenfels MW (1950) The sponges of Kaneohe Bay, Oahu. Pac Sci 4(1): 3-36 de Laubenfels MW (1954) The sponges of the west-central Pacific. Oreg St Monogr Zool 7: 1-306 de Voogd N, Becking LE, Hoeksema BW, Noor A, van Soest RWM (2004) Sponge interactions with spatial competitors in the Spermonde Archipelago. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Boll Mus Ist Biol Univ Genova 68: 253-261 de Weerdt WH (2002) Family Chalinidae Gray. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges. Kluwer Academic/Plenum Publishers, New York. pp. 852-873 Dunlap M, Pawlik JR (1996) Video-monitored predation by Caribbean reef fishes on an array of mangrove and reef sponges. Mar Biol 126: 117-123 Dunlap M, Pawlik JR (1998) Spongivory by parrotfish in Florida mangrove and reef habitats. PSZN Mar Ecol 19: 325-337

Ellison AM, Farnsworth EJ, Twilley RR (1996) Facultative mutualism between red mangroves and root-fouling sponges in Belizean mangal. Ecology 77(8): 2431-2444 Ely R, Supriya T, Naik CG (2004) Antimicrobial activity of marine organisms collected off the coast of South East. India. J Exp Mar Biol Ecol 309: 121-127 Engel S, Pawlik JR (2000) Allelopathic activities of sponge extracts. Mar Ecol Prog Ser 207: 273-281 Engel S, Pawlik JR (2005) Interactions among Florida sponges. II. Mangrove habitats. Mar Ecol Prog Ser 303: 145-152 Frincke JM, Faulkner DJ (1982) Antimicrobial metabolites of the sponge Reniera sp. J Am Chem Soc 104: 265-269 Frost TM, Williamson CE (1980) In situ determination of the effect of symbiotic algae on the growth of the freshwater sponge Spongilla lacustris. Ecology 61: 1361-1370 Gaino E, Lancioni T, La Porta G, Todini B (2004) The consortium of the sponge Ephydatia fluviatilis (L.) living on the common reed Phragmites australis in Lake Piediluco (central Italy). Hydrobiologia 520: 175-178 Gaino E, Sar M (1994) Siliceous spicules of Tethya seychellensis (Porifera) support the growth of a green alga: A possible light conducting system. Mar Ecol Prog Ser 108: 147-151 Garson MJ, Flowers AE, Webb RI, Charan RD, McCaffrey EJ (1998) A sponge/dinoflagellate association in the haplosclerid sponge Haliclona sp.: cellular origin of cytotoxic alkaloids by Percoll density gradient fractionation. Cell Tissue Res 293: 365-373 Gochfeld D, Harrison L, Olson J, Lesser MP, Ankisetty S, Slattery M (2006) Allelopathic effects of the Caribbean sponge Svenzea zeai on the coral Montastrea annularis. In: Custdio MR, Lbo-Hajdu G, Hajdu E, Muricy G (eds). 7th International Sponge Symposium Book of Abstracts (Armao dos Bzios, Brazil). Museu Nacional, Srie Livros, vol. 16. pp. 116 Goreau TF, Hartman WD (1966) Sponge: effect on the form of reef corals. Science 151: 343-344 Henkel TO, Pawlik JR (2005) Habitat use by sponge-dwelling brittlestars. Mar Biol 146: 301-313 Hooper NA, van Soest RWM (2002) Systema Porifera: a guide to the classification of sponges. Kluwer Academic, Plenum Publishers, New York Ilan M, Loya Y, Kolbasov GA, Brickner I (1999) Sponge inhabiting barnacles on coral reefs. Mar Biol 133: 709-716 Isaacs S, Kasbman Y (1992) Shaagrockol B and C; two hexaprenylhydroquinone disulfates from the Red Sea sponge Toxiclona toxius. Tetrahedron Lett 33: 2227-2230 Jackson JBC, Buss L (1975) Allelopathy and spatial competition among coral reef invertebrates. Proc Natl Acad Sci 72: 5160-5163 Kohler KE, Gill SM (2006) Coral Point Count with Excel extensions (CPCe): A Visual Basic program for the determination of coral and substrate coverage using random point count methodology. Comp Geosci 32: 1259-1269 Lpez-Victoria M, Zea S, Weil E (2006) Competition for space between encrusting excavating Caribbean sponges and other coral reef organisms. Mar Ecol Prog Ser 312: 113-121 Macintyre IG, Precht WF, Aronson RB (2000) Origin of the Pelican Cays ponds, Belize. Atoll Res Bull 466: 1-11 Magnino G, Gaino E (1998) Haplosyllis spongicola (Grbe) (Polychaeta, Syllidae) associated with two species of sponges from East Africa (Tanzania, Indian Ocean). PSZN Mar Ecol 19: 77-87

156

Maldonado M, Uriz MJ (1998) Microrefuge exploitation by subtidal encrusting sponges: patterns of settlement and post-settlement survival. Mar Ecol Prog Ser 174: 141-150. Mercurio M, Corriero G, Gaino E (2006) Sessile and non-sessile morphs of Geodia cydonium (Jameson) (Porifera, Demospongiae) in two semi-enclosed Mediterranean bays. Mar Biol 148: 489-501 Meroz E, Ilan M (1995) Cohabitation of a coral reef sponge and a scyphozoan. Mar Biol 124: 453-459 Nishiyama GK, Bakus GK (1999) Release of allelochemicals by three tropical sponges (Demospongiae) and their toxic effects on coral substrate competitors. Memoir Queensl Mus 44: 411-417 Palumbi SR (1985) Spatial variation in an algal-sponge commensalism and the evolution of ecological interactions. Am Nat 126: 267-274 Patton WK (1974) Community structure among the animals inhabiting the coral Pocillopora damicornis at Heron Island, Australia. Bull Mar Sci 14(2): 212-243 Pettit GR, Rideout JA, Hasler JA (1981) Isolation of geodiastatins 1 and 2 from the marine sponge Geodia mesotriaena. J Nat Prod 44(5): 588-92 Pettit GR, Rideout JA, Hasler JA (1990) Isolation of geodiatoxins 2 and 4 from the marine sponge Geodia mesotriaena. Comp Bio Physiol Part C 96(2): 305-6 Plucer-Rosario G (1987) The effect of substratum on the growth of Terpios, an encrusting sponge which kills corals. Coral Reefs 5(4): 197-200 Rtzler K (1970) Spatial competition among Porifera: solution by epizoism. Oecologia 5: 85-95 Rtzler K (1974) The burrowing sponges of Bermuda. Smithson Contrib Zool 165:1-32 Rtzler K (1990) Associations between Caribbean sponges and photosynthetic organisms. In: Rtzler K (ed). New perspectives in sponge biology. Smithsonian Institution Press, Washington DC. pp. 455-466 Rtzler K (2002) Impact of crustose clionid sponges on Caribbean Reef corals. Acta Geol Hisp 37: 61-72 Rtzler K, Muzik K (1993) Terpios hoshinota, a new cyanobacteriosponge threatening Pacific reefs. In: Uriz MJ, Rtzler K (eds). Recent Advances in Ecology and Systematics of Sponges. Sci Mar 57(4): 395-403 Sar M, Vacelet (1973) cologie des Dmosponges. In: Grass PP (ed). Spongiaires. Anatomie, Physiologie, Systmatique, cologie. Masson de Cie, Paris. pp. 462-576 Schuchert P (2003) Hydroids (Cnidaria, Hydrozoa) of the Danish expedition to the Kei Islands. Steenstrupia 27: 137-256 Sera Y, Adachi K, Fujii K, Shizuri Y (2002) Isolation of Haliclonamides: New peptides as antifouling substances from a marine sponge species, Haliclona. Mar Biotechnol 3: 441-446

Sim CJ (1998) Two-sponge association from Komun Island, Korea. In: Watanabe Y, Fusetani, N (eds). Sponge sciences: multidisciplinary perspectives. Springer-Verlag, Tokyo. pp 109117 Steindler L, Beer S, Ilan M (2002) Photosymbiosis in intertidal and subtidal tropical sponges. Symbiosis 33:263-273 Thacker RW, Becerro MA, Lumbang WA, Paula VJ (1998) Allelopathic interactions between sponges on a tropical reef. Ecology 79(5): 1740-1750 Thakur NL, Mller WEG (2004) Biotechnological potential of marine sponges. Curr Sci 86(11): 1506-1512 Trautman DA, Hinde R, Borowitzka MA (2000) Population dynamics of an association between a coral reef sponge and a red macroalga. J Exp Mar Biol Ecol 244: 87-105 Trautman DA, Hinde R (2002) Sponge/algal symbioses: a diversity of associations. In: Seckbach J (ed). Symbiosis: Mechanisms and Model Systems. Kluwer Academic Publishers, Netherlands, 796 pp. Uriz MJ, Rosell D, Maldonado M (1992) Parasitism, comensalism or mutualism? The case of Scyphozoa (Coronatae) and horny sponges. Mar Ecol Prog Ser 81: 247-255 Vacelet J (1981) Algal-sponge symbioses in the coral reefs of New Caledonia: a morphological study. Proc 4th Int Coral Reef Symp, Manila 2: 713-719 Vacelet J, Sofyani AA, Lihaibi SA, Kornprobst JM (2001) A new haplosclerid sponge species from the Red Sea. J Mar Biol Assoc UK 81: 943-948 van Soest RWM, Braekman JC (1999) Chemosystematics of Porifera: a review. Memoir Queensl Mus 44: 569-589 West DA (1979) Symbiotic zoanthids (Anthozoa: Cnidaria) of Puerto Rico. Bull mar Sci 29: 253-271. Wilcox TP, Hill M, DeMeo K (2002) Observations on a new twosponge symbiosis from the Florida Keys. Coral Reefs 21: 198204 Wilkinson CR (1983) Net primary productivity of coral reef sponges. Science 219: 410-412 Wulff JL (1997a) Causes and consequences of differences in sponges diversity and abundance between the Caribbean and eastern Pacific at Panama. Proc 8th Int Coral Reef Symp, Panama 2: 1377-1382 Wulff JL (1997b) Mutualisms among species of coral reef sponges. Ecology 78: 146-159 Wulff JL (1999) A sponge that cheats on diffuse mutualism among other sponge species. Mem Qld Mus 44: 686 Wulff JL (2006) Ecological interactions of marine sponges. Can J Zool 84: 146-166 Xue S, Zhanga HT, Wua PC, Zhanga W, Yuan Q (2004) Study on bioactivity of extracts from marine sponges in Chinese Sea. J Exp Mar Biol Ecol 298: 71-78

Porifera research: Biodiversity, innovation and sustainaBility - 2007

157

Sponges of the marine karst lakes and of the coast of the islands of Ha Long Bay (North Vietnam)
Francesca Azzini(1), Barbara Calcinai(1), Carlo Cerrano(2), Giorgio Bavestrello(1), Maurizio Pansini(2*)
Dipartimento di Scienze del Mare, Universit Politecnica delle Marche, Via Brecce Bianche, 60131 Ancona, Italy (2) Dip.Te.Ris. Universit di Genova, Corso Europa 26, 16132 Genova, Italy. mpansini@dipteris.unige.it
(1)

Abstract: The purpose of this paper is to describe the sponge assemblages of the marine karst lakes and some coastal sites in the islands of Ha Long Bay (Vietnam). These lakes are very shallow basins with a bottom covered by mud and vegetable debris. Patches of mangroves were seldom observed, while isolated colonies of massive corals are frequently found. Some lakes show a connection with the surrounding sea, evidenced by the flow of tidal streams; others are apparently closed but connected to the sea through the cavities of the karst system. The coast of Ha Long Bay islands are characterized by shallow depth, murky waters, patchy and low diversity reefs, and very low water-movement. A total of 63 demosponge species were identified: 46 species were recorded in the marine lakes, while 23 of them were never found outside the lakes. Extreme variations in environmental conditions occur yearly in the lakes, due to heavy monsoon rains that cause a stratification of the water column. A thermal crisis was recorded at the end of the summer in the Hang Du I lake, with bottom temperatures as high as 36C. Sponges can withstand these conditions but undergo important rearrangements, with a late summer degeneration followed by a very quick fall and winter growth. Keywords: Porifera, anchialine lakes, karst islands, environmental stress, Tonkin Gulf

Introduction
The Ha Long Bay (Vietnam) is located in the northern part of the Tonkin Gulf, in a shallow water area of the South China Sea where more than 3000 karst islands of different sizes are present. Karstic processes, enhanced by the tropical conditions, carved out of the limestone many depressions, leading to the formation of shallow salt water lakes similar to the meromictic marine lakes of Palau (Hamner and Hamner 1998) and to the anchialine lakes of East Kalimantan (Indonesia), where the only specific study on the sponge fauna of these special environments is in progress (de Voogd et al. 2006). The scope of this study was to investigate the species composition of the sponge fauna of the lakes compared with that of the coastal areas of the isles of the bay. The sponges of the coast of Vietnam are still scarcely studied. In fact one can find only a paper by Lindgren (1898), an inventory by Dawydoff (1952), a study of a collection from Nha Trang (Lvi 1961) and a recent paper devoted to boring sponges (Calcinai et al. 2006) among the literature. A checklist of sponges recorded from the South China Sea by Hooper et al. (2000) reports as many as 161 demosponges (106 of which identified at species level) from the coast of Vietnam.

Materials and methods Description of the study area

The climatic conditions of the study area are tropical, with a mean annual air temperature of 25C, warm and wet summer monsoons (up to 2000 mm/year of rain) from May to October and relatively cold and dry winter monsoons from November to April (Tang 2001). The marine lakes are remarkably different from one another in the degree of isolation from the surrounding sea. In some lakes, connection is detectable by the flow of tidal streams. Sometimes artificial canals have been built to enhance this water exchange. However, some lakes are apparently closed, being connected to the sea only through the cavities of the karst system. They are no more in pristine conditions, having been exploited for fishery, mollusc harvesting and aquaculture by people dwelling in boat villages located around the islands. In the surrounding sea, owing to the presence of thousands of islands, water is calm and turbid. The mean tidal excursion ranges between 3 and 4 m, salinity ranges around 32, and sea-water temperature varies between a minimum of 23C at the end of winter and 29-30C in late summer (Tang 2001).

Sampling
Three field campaigns were performed in April 2003, September 2003 and April 2004, in a joint venture between

158

the Universit Politecnica delle Marche and the Hai Phong Institute of Oceanology (HIO) for the study of biodiversity and conservation in a coastal area of Vietnam. Samples were collected either through SCUBA diving or snorkelling in 15 sites located in Ha Long Bay Islands (Fig. 1). Eight of these sites are salt lakes, while seven are located along the coast surrounding the islands. The latter were chosen among those periodically surveyed for biodiversity assessments by the team of HIO. Sponges were collected from the rocky shores of both the lakes and the islands down to a depth of about 3 m and from the small reefs along the islands, which are only slightly deeper (7-8 m). Lakes were considered as open (Hang Du II, Hang Tham, Hang Luong), semi-enclosed (Dau Be, Cat Ba, Me Cung) and enclosed (Hang Du I, Bui Xam) when they were connected to the sea by large canals, small conduits or through the karst system, respectively. No quantitative sampling was taken, but in each sampling station all the discrete sponge species were collected by three divers during 45 minutes of dive. The small-encrusting and cryptic species were not collected. Whenever possible, sponge specimens were photographed

in the field, or on board after collection; they were fixed in 4% formaldehyde solution in sea water and preserved in 60% ethanol. Some specimens were dried.

Results Distribution
Sixty three demosponges have been identified (36 to species level), out of 182 specimens collected from the marine lakes and the coasts of the isles of Ha Long Bay (Table 1). Forty six species were found in the lakes and 40 in the bay; 23 species were found only in the lakes and 17 only in the bay (Table 1). According to presence/absence data, the most common species in the studied area, including lakes and coastal sites, were Dysidea cinerea Keller, 1889 and Haliclona (Haliclona) sp. 2 (present in 60% of the collecting stations), Tethya seychellensis (Wright, 1881) (53%), Haliclona (Gellius) cymaeformis (Esper, 1794) and Spheciospongia tentorioides (Dendy, 1905) (46.6%), and Mycale philippensis (Dendy, 1896) (40%). However, in the absence of quantitative

Fig. 1: Location of the marine lakes and coastal sites surveyed in six island groups (Bo Hon, Hang Trai, Dau Be, Cat Ba, Conf and Cong Do) in the western part of Ha Long Bay (Vietnam): 1-Hang Luong Lake; 2-Me Cung Lake; 3-Bui Xam Lake; 4-Hang Du I Lake; 5-Hang Du II Lake; 6-Dau Be Lake; 7-Cat Ba Lake; 8-Hang Tham Lake; 9-Coastal site I; 10Coastal site II; 11-Coastal site III; 12-Coastal site IV; 13-Coastal site V; 14-Hang Toi Dark Cave; 15Coastal site VI.

159

Table 1: List of the Demosponge species collected from some marine lakes and coastal sites of Ha Long Bay: 1-Hang Luong Lake; 2-Me Cung Lake; 3-Bui Xam Lake; 4-Hang Du I Lake; 5-Hang Du II Lake; 6-Dau Be Lake; 7-Cat Ba Lake; 8-Hang Tham Lake; 9-Coastal site I; 10-Coastal site II; 11-Coastal site III; 12-Coastal site IV; 13-Coastal site V; 14-Hang Toi Dark Cave; 15- Coastal site VI. The species that are new records for Vietnam are marked by *. Species Aaptos cf. pernucleata (Carter, 1870) * Acanthella hispida Pulitzer-Finali, 1980 * Aka mucosa (Bergquist, 1965) * Amorphinopsis excavans Carter, 1887 * Amphimedon sp. Aplysina sp. Biemna megalosigma Hentschel, 1912 * Bubaris sp. Callyspongia sp. Chondrilla australiensis Carter, 1873 Cinachyrella australiensis (Carter, 1886) * Cladocroce sp. Clathria sp. Cliona aurivilli (Lindgren, 1897) Cliona celata Grant, 1826 Cliona orientalis Thiele, 1900 * Cliona sp. Cliothosa hancocki (Topsent, 1888) Desmanthus incrustans (Topsent, 1889) * Dictyonella sp. 1 Dictyonella sp. 2 Dysidea cinerea Keller, 1889 * Dysidea cf. fragilis (Montagu, 1818) Echinodictyum asperum Ridley and Dendy, 1886 Eurypon sp. Gelliodes fibulatus (Carter, 1881) Halichondria sp. Haliclona (Gellius) cymaeformis (Esper, 1794) Haliclona (Haliclona) sp. 1 Haliclona (Haliclona) sp. 2 Haliclona (Haliclona) sp. 3 Haliclona (Haliclona) sp. 4 Haliclona (Haliclona) sp. 5 Haliclona (Haliclona) sp. 6 Haliclona (Reniera) sp. 1 Haliclona (Reniera) sp. 2 Haliclona (Reniera) sp. 3 Haliclona (Gellius) sp. Hyattella intestinalis (Lamarck, 1814) * Ircinia echinata (Keller, 1889) * Ircinia sp. Mycale (Mycale) crassissima (Dendy, 1905) Mycale (Zygomycale) parishi (Bowerbank, 1875) * Mycale (Mycale) philippensis (Dendy, 1896) Mycale (Mycale) sp. Penares cf. sollasi Thiele, 1900 * Petrosia (Petrosia) nigricans Lindgren, 1897 * Pione carpenteri (Hancock, 1867) * Suberites sp. 1 Suberites sp. 2 Protosuberites sp. 1 Protosuberites sp. 2 Spheciospongia solida Ridley and Dendy, 1886 Spheciospongia tentorioides (Dendy, 1905) * Spirastrella cf. cunctatrix Schmidt, 1868 * Spirastrella decumbens Ridley, 1884 * Spongia irregularis (von Lendenfeld, 1885) Stelletta aruensis Hentschel, 1912 * Tedania (Tedania) brevispiculata Thiele, 1903 Terpios cruciata Dendy, 1905 * Tethya seychellensis (Wright, 1881) * Topsentia cavernosa (Topsent, 1897) * Xestospongia cf. testudinaria (Lamarck, 1815) * 1 2 Marine Lakes 3 4 5 6 7 8 9 10 Coastal sites 11 12 13 14 15

160

data but based on field observations, Haliclona (Gellius) cymaeformis, living in symbiotic association with the rodophyte alga Ceratodictyon spongiosum Zanardini, 1878, was the most abundant species, thriving with an impressive number of specimens on all the horizontal substrates, even in very shallow waters where it might be exposed to air at low tide. It was dark green in colour and may vary in shape from thickly encrusting to massive or, more often, bushy (Fig. 2A). A remarkable number of boring sponges i.e. Aka mucosa (Bergquist, 1965), Cliona celata Grant, 1826, C. orientalis Thiele, 1900, Cliothosa hancocki (Topsent, 1888), Spheciospongia tentorioides, and Cliona aurivilli (Lindgren, 1897) the latter according to Calcinai et al. (2006) were found both in the lakes and in the surrounding sea, while the distribution of Pione carpenteri (Hancock, 1867) and Spirastrella decumbens Ridley, 1884 seems to be restricted to the lakes (Table 1). The bay In the costal stations of the bay we have observed that sponges settled on two types of substrates: The calcareous rocky shores of the islands, extending down to 3-4 m in depth and the coral gardens which proliferate on the horizontal surfaces to a depth of about 8 m. On the coast, the suitable substrate for sponge settlement was very scarce, due to the dense belt of bivalve molluscs. Horny sponges such as Dysidea cinerea, Dysidea cf. fragilis (Montagu, 1818), Spongia irregularis (von Lendenfeld, 1885), as well as several species of Haplosclerida (Haliclona spp., Cladocroce sp.) and Tethya seychellensis were the most common species on the rocky shores. In the coral gardens, sponges behave as opportunistic species, dwelling on corals [e.g. Gelliodes fibulatus (Carter, 1881), Amphimedon sp., Ircinia echinata (Keller, 1889)], on pockets of sediment in between corals (Biemna megalosigma Hentschel, 1912), and on coral rubble [Acanthella hispida Pulitzer-Finali, 1980, Spheciospongia tentorioides, Xestospongia cf. testudinaria (Lamarck, 1815)]. These last two species were particularly abundant and seem to contribute to consolidate the coral fragments. Along the coast of the islands one can find some peculiar habitats represented by semi-dark environments, such as a rocky tunnel with a steady current, open at both ends, leading to Xang Luong cove in the Island of Bo Hon and Hang Toi Cave, a cavity with a depth of 1-1.5 m, in the Island of Cong Do (Fig. 1). In these habitats, sponges were abundant and diverse. Cinachyrella australiensis (Carter, 1886), Penares cf. sollasi Thiele, 1900, and Stelletta aruensis Hentschel, 1912 have been recorded in Hang Toi Cave only. In the above mentioned tunnel steady water movement supported numerous colonies of the octocoral Carijoa riisei (Duchassaing and Michelotti, 1860). Several sponge species [Callyspongia sp., Mycale philippensis, Mycale (Mycale) crassissima (Dendy, 1905), Spirastrella cf. cunctatrix Schmidt, 1868, Tedania (Tedania) brevispiculata Thiele, 1903 and some unidentified Haliclona] were epizoic on Carijoa colonies in the Ha Long Bay. They initially use the octocoral skeleton as support and subsequently overgrow it completely. Haliclona (Haliclona) sp. 2 (Fig. 2C) has been observed on a colony already covered by Mycale

philippensis. Carijoa appears to be unharmed by the epizoic sponges because its anthocodia are free to expand and retract (Fig. 2B). The lakes Among the pool of 23 species recorded only in the lakes, strong differences among the different basins were observed. The only species recorded in 50% of the lakes is Haliclona (Gellius) sp.; four other species i.e. Pione carpenteri (Hancock, 1867), Suberites sp. 1, Protosuberites sp. 1, Spirastrella cf. cunctatrix, were present in 25% of the lakes. Each of the other 18 species was recorded only in one lake. The highest number of species (28) was found in the enclosed lake of Bui Xam, which has no detectable connection to the sea. 70% of these species were found both in the lakes and the coastal stations. Three species, Clathria sp., Eurypon sp. and Spirastrella decumbens were found only in this lake. In the enclosed lake of Hang Du I, ten species were identified, 7.5% of which were in common with the coastal stations. However, Aaptos cf. pernucleata (Carter, 1870), Dictyonella sp. 1, Haliclona (Haliclona) sp. 3 and Haliclona (Reniera) sp. 2 were found only in this lake. Suberites sp. 1, a very abundant species as regards the number and size of specimens, was found just in Hang Du I and in Dau Be, a lake connected to the sea by small conduits. Pione carpenteri and Spirastrella decumbens seem to be restricted to the enclosed lakes of Hang Du I and Bui Xam. In three lakes, Hang Luong, Me Cung and Cat Ba, the sponge fauna was completely composed of species that were also present in the bay, while in Hang Du II, Dau Be and Hang Tham the overlapping with the bay fauna ranged between 55 and 75%.

Ecology
Hang Du I lake was studied in detail because of the extreme variability of its environmental conditions directly affecting the sponge fauna. Direct observations of Suberites sp. 1 showed that during the dry season, in winter and early spring, numerous healthy specimens thrived close to the lake surface (Fig. 2D). In late spring and summer, morphological rearrangements and degeneration phenomena due to the combined effect of rainwater stratification with high temperatures were observed in this upper zone (Fig. 2E). Particularly evident was the case of Protosuberites sp. 1 (Fig. 2F, G). As soon as water cooled again, very quick fall and winter growth followed, but the original conditions did not seem to be restored after a single season.
Fig. 2: A. Haliclona (Gellius) cymaeformis: a bushy specimen. B. Sponges associated with colonies of the octocoral Carijoa riisei. Polyps continue their filtering activity; the worm-like organisms are synaptid holothurians. C. Epizoic sponges on Carijoa colonies: Haliclona (Haliclona) sp. 2 (green) and Mycale (Mycale) philippensis (red). Morphological rearrangements of two sponge species of the Hang Du I lake in spring and late summer: D-E. Suberites sp. 1; F-G. Protosuberites sp. 1.

161

162

Biogeography
Twenty-three species, corresponding to 63.8% of the 36 fully identified sponges of our collection, are new records for the coast of Vietnam (Table 1). Most of these species (32 out of 36) are distributed in the Indian Ocean, including the Red Sea, and in the West Pacific Ocean, including Australia. Three of them (Cliona celata, Dysidea fragilis, Tethya seychellensis) are considered cosmopolitan, while a single species, Aaptos pernucleata, is known from the West Indies only.

Discussion
The recorded data suggest that the sponge fauna of the Tonkin Gulf is similar to that of the adjacent tropical areas of Indonesia and Northern Australia. Both the peculiar geomorphology and oceanography of the bay may represent important factors negatively affecting the northern expansion of sponges. The number of species found in the lakes (46) is comparable to that found until now in four anchialine lakes of East Kalimantan which were thought to represent a lagoonal reef of a former barrier reef complex (de Voogd et al. 2006). A species particularly adapted and restricted to the lakes appears to be Suberites sp. 1, very likely cospecific with the Suberites sp. reported from Lake Satonda (Palau) and East Kalimantan Lakes (Indonesia) (de Voogd et al. 2006). The species composition of sponge fauna is very different in the studied lakes without any apparent relation with the observed degree of connection to the open sea. This fact results from the comparison of the number of sponge species, 28 and 10, respectively, recorded in the two enclosed lakes of Bui Xam and Hang Du I. The occurrence of scattered colonies of massive corals which are present in Bui Xam and absent in Hang Du I suggests the presence of an undetected, large connection between the first lake and the open sea. Conversely, the high degree of isolation and the peculiar character of the above mentioned lake are confirmed by the presence of a dense population of the non-stinging jellyfish (Rhizostoma sp.) (Cerrano et al. 2006) and four species of sponges out of a total of 10 which are absent elsewhere. The different degree of isolation from the open sea of these two lakes is demonstrated by the severe water stratification occurring in Hang Du I, while in the Bui Xam lake stratification phenomena are less intense. Due to the stillness of this sheltered basin, a light and cool rainwater layer (salinity < 7 ) as thick as 150 cm (on a maximum lake depth of about 6 m) stratifies on the lake surface thus preventing the normal mixing of the water column. This resulted in an abnormal rising of the bottom temperature, which was as high as 36C in September 2003 (Fig. 3). This thermal crisis produced a surplus of organic debris, coming from vegetable and animal decay, that deposits on the lakes muddy bottom that became anoxic (Cerrano et al. 2006). The spring conditions, recorded in April 2003 and 2004, were each similar and presumably bound to the climatic pattern of the year (Fig. 3). Seasonal variations in temperature and salinity such as those recorded in Hang Du I remarkably affect the sponge fauna. The response to the environmental stress seems to

be related to the sponge position on the bottom, because Clionaid species and Tethya seychellensis, living in sheltered habitats, appear unaffected, whereas Suberites sp. 1 and Pseudosuberites sp. 1, living in exposed locations, show evident regression. Rapid growth rates as those observed after the late summer crisis were already recorded in sponges after negative events (Ayling 1983). These growth rates may be also supported by increased food abundance (Duckworth et al. 2003) resulting from the restoration of the normal conditions in the water column. Summer temperatures in temperate regions normally cause positive growth rates in sponges (Turon et al. 1998, Tanaka 2002) whereas sponge shrinking was associated to colder temperatures (Duckworth and Battershill 2001). However, persistence of high water temperatures in late summer may stress benthic organisms, sometimes triggering mass casualties (Cerrano et al. 2000). Regression and reorganization of adult sponges apparently neither related to environmental stress nor to a seasonal cycle were observed in natural conditions (Pansini and Pronzato 1990, Bell and Smith 2004). Besides salinity variations, another physical factor affecting sponge fauna both in lakes and coastal sites appears to be the high water turbidity. The impact of sedimentation on sponges is well known (Sar and Vacelet 1973, Bell and Smith 2004) as is the defensive reaction set up by sponges in order to avoid clogging of their aquiferous system (Reiswig 1971, Bell 2004, Cerrano et al. 2004). In a shallow water area such as Ha Long Bay, tidal range and competition with bivalve molluscs reduce the space available for sponge settlement on vertical substrates and overhangs, which are protected from sediment deposition. This may cause an overall reduction of the specific richness of porifera but may also select taxa producing fistules (e.g. Biemna megalosigma), adapted to live partially buried by the sediment on horizontal substrates. Symbiosis may be the clue for explaining the great abundance of Haliclona (Gellius) cymaeformis, associated with the rodophyte Ceratodyction spongiosum, in the intertidal and subtidal of the bay. Steindler et al. (2002) suggest that the photosynthetic activity of the algae may fulfill the energetic needs of the sponges when they stop filtering, being partially emerged at low tide, and that symbionts protect them from UV radiation, particularly intense in shallow water. In addition, Pile et al. (2003) showed that the sponge, feeding on nitrogen-rich bacteria and protozoans of the ultraplankton, can meet the nitrogen demand of both partners of the symbiotic association. Therefore, several positive factors could allow Haliclona (Gellius) cymaeformis to thrive in the shallow water environment even in the presence of rather low water transparency. Until the present, marine lake biota is poorly known, but they very likely host many species new to science as the taxonomy of sponges here suggests. As highlighted by Dawson and Hamner (2005) marine lakes offer the possibility to study the founder effect at different stages, in relation to their age and level of isolation. The lakes of North Vietnam are smaller than Palau and Indonesian ones and host a fauna very interesting in relation to speciation processes and to its physiology, being strongly adapted to sudden modifications of the environmental parameters. These aspects may at least partially explain the large variation in qualitative composition

163

Fig. 3: Environmental variations in Hang Du I lake.

of the sponge fauna between different lakes, and between the lakes and the surrounding marine areas. It is important to increase the knowledge of these unique habitats before human activities and climate change would irreparably damage them.

Acknowledgements
We are indebted to Do Co Thung and his team of the Hai Phong Institute of Oceanology for supporting us in the field and to Massimo Sarti of the Universit Politecnica delle Marche for his geological advice.

References
Ayling AL (1983) Growth and regeneration rates in thinly encrusting Demospongiae from temperate waters. Biol Bull 165: 343-352 Bell JJ (2004) Evidence for morphology-induced sediment settlement prevention on the tubular sponge Haliclona urceolus. Mar Biol 146(1): 29-38 Bell JJ, Smith D (2004) Ecology of sponge assemblages (Porifera) in the Wakatobi region, south-east Sulawesi, Indonesia: richness and abundance. J Mar Biol Assoc UK 84: 581-591 Calcinai B, Azzini F, Bavestrello G, Cerrano C, Pansini M, Thung DC (2006) Boring sponges from the Ha Long Bay (Tonkin Gulf, Vietnam). Zool Stud 45(2): 201-212 Cerrano C, Bavestrello G, Bianchi CN, Cattaneo-Vietti R, Bava S, Morganti C, Morri C, Picco P, Sara G, Schiaparelli S, Siccardi A, Sponga F (2000) A catastrophic mass-mortality episode of gorgonians and other organisms in the Ligurian Sea (north-western Mediterranean), summer 1999. Ecol Lett 3: 284-293 Cerrano C, Pansini M, Valisano L, Calcinai B, Bavestrello G, Sar M (2004) Lagoon sponges from Carrie Bow Cay (Belize): ecological benefits of selective sediment incorporation. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge Science in the new Millenium. Boll Mus Ist Biol Univ Genova 68: 239-252

Cerrano C, Azzini F, Bavestrello G, Calcinai B, Pansini M, Sarti M, Thung DC (2006) Marine lakes of karst islands in Ha Long Bay (Vietnam). Chem Ecol 22(6): 489-500 Dawydoff C (1952) Inventaire des animaux benthique rcolts par moi dans le domaine maritime Indochinois. Porifres. Suppl Bull Biol France Belgique 37: 46-51 Dawson M, Hamner WN (2005) Rapid evolutionary radiation of marine zooplankton in peripheral environments. Proc Natl Acad Sci USA 102 (26): 9235-9240 de Voogd N, de Weerdt WH, van Soest RWM (2006) The sponge fauna of the anchialine lakes of Kakaban and Maratua (East, Kalimantan, Indonesia). In: Custdio MR, Lbo-Hajdu G, Hajdu E, Muricy G (eds). 7th International Sponge Symposium - Book of Abstracts (Armao dos Bzios, Brazil). Museu Nacional, Srie Livros, vol. 16. p. 242 Duckworth AR, Battershill CN (2001) Population dynamics and chemical ecology of New Zealand Demospongiae Latrunculia sp. nov. and Polymastia croceus (Poecilosclerida: Latrunculiidae: Polymastiidae). N Zealand J Mar Freshw Res 35: 935-949 Duckworth AR, Samples GA, Wright AE, Pomponi SA (2003) In vitro culture of the tropical sponge Axinella corrugata (Demospongiae): effect of food cell concentration on growth, clearance rate, and biosynthesis of stevensine. Mar Biotechnol 5(6): 519-527 Hamner WM, Hamner PP (1998) Stratified marine lakes of Palau (Western Caroline Island). Phys Geogr 19: 175-220 Hooper JNA, Kennedy JA, van Soest RWM (2000) Annotated checklist of sponges (Porifera) of the South China Sea region. The Raffles Bull Zool suppl 8: 125-207 Lvi C (1961) ponges intercotidales de Nha Trang (Viet Nam). Arch Zool Exp Gn 100: 127-150 Lindgren NG (1898) Beitrag zur kenntniss der Spongienfauna des Malayischen Archipels und der Chinesischen Meere. Zool Jahrb Abt Syst Geogr Biol Thiere 11: 283-378 Pansini M, Pronzato R (1990) Observations on the dynamics of a Mediterranean sponge community. In: Rtzler K, Macintyre VV, Smith KP (eds). New Perspectives in Sponge Biology. Smithsonian Institution Press, Washington DC. pp. 404-415

164

Pile AJ, Grant A, Hinde R, Borowitzka MA (2003) Heterotrophy on ultraplankton communities is an important source of nitrogen for a spongerhodophyte symbiosis J Exp Biol 206: 4533-4538 Reiswig HM (1971) In situ pumping activities of tropical Demospongiae Mar Biol 9: 38-50 Sar M, Vacelet J (1973) cologie des Dmosponges. In: Grass PP (ed). Trait de Zoologie, Vol. 3, Part. 1. Masson, Paris. pp. 462576 Steindler L, Beer S, Ilan M (2002) Photosymbiosis in intertidal and subtidal tropical sponges. Symbiosis 33: 263-273

Tanaka K (2002) Growth dynamics and mortality of the intertidal encrusting sponge Halichondria okadai (Demospongiae, Halichondrida). Mar Biol 140(2): 383-389 Tang VT (2001) The Eastern Sea Resources and Environment. The Gioi Publishers, Hanoi Turon x, Tarjuelo I, Uriz, MJ (1998) Growth dynamics and mortality of the encrusting sponge Crambe crambe (Poecilosclerida) in contrasting habitats: correlation with population structure and investment in defence. Funct Ecol 12(4): 631-639

Porifera research: Biodiversity, innovation and sustainaBility - 2007

165

Microbial nitrification in Mediterranean sponges: possible involvement of ammonia-oxidizing Betaproteobacteria

Kristina Bayer, Susanne Schmitt, Ute Hentschel(*)
Research Center for Infectious Diseases, University of Wuerzburg. Roentgenring 11, D-97070 Wuerzburg, Germany. Tel.: +49-931-312581. Fax: +49-931-312578. kristina.bayer@mail.uni-wuerzburg.de, susanne.schmitt@mail.uni-wuerzburg.de, ute.hentschel@mail.uni-wuerzburg.de Abstract: The aim of this study was to assess the potential for nitrification in Aplysina aerophoba Schmidt 1862 using a combined physiological and molecular approach. Whole animals were incubated in experimental aquaria and the concentrations of ammonia, nitrate and nitrite were determined in the incubation water using colorimetric assays. Nitrate excretion rates reached values of up to 3.6 mol g-1 fresh weight day-1 (equivalent to 830 nmol g-1 dry weight h-1) and were matched by ammonia excretion rates of up to 0.56 0.09 mol g-1 fresh weight day-1. An accumulation of nitrite was not detected in any of the experiments. Control experiments without sponges showed no variation in nitrogen species in the incubation water. A slight increase in ammonia excretion was observed over 11 days of maintenance in holding tanks that were constantly supplied with fresh, untreated Mediterranean seawater. Other sponges from the same habitat (Dysidea avara Schmidt 1862, Tethya sp., Chondrosia reniformis Nardo 1847) showed high rates of ammonia excretion but nitrate excretion was significantly reduced or absent. Using specific PCR primers, 16S rRNA genes of the betaproteobacterial clade of the Nitrosospira cluster 1 were recovered from A. aerophoba, D. avara and Tethya sp. tissues. In conclusion, this study provides physiological and molecular evidence for the presence of nitrifying bacteria in A. aerophoba while the potential for nitrification in the other sponges remains to be investigated. Keywords: Ammonia-oxidizing Betaproteobacteria, microbial consortia, nitrification, 16S rRNA gene, sponge

Introduction
Sponges (Porifera) are evolutionarily ancient metazoans with a fossil record dating back nearly 600 million years in time (Li et al. 1998). Today, an estimated 13,000 species, classified in three classes (Demospongiae, Calcarea, Hexactinellida) populate virtually all benthic marine and freshwater habitats (Hooper and van Soest 2002). Sponges have a primitive morphology lacking true organs or tissues. Most metabolic functions are carried out by totipotent, amoeboid cells, termed archaeocytes that move freely through the mesohyl matrix. Inhalant and exhalant canals build an aquiferous system through which water is actively pumped by flagellated choanocytes (Brusca and Brusca 1990). As filter-feeders, sponges efficiently take up nutrients like organic particles and microorganisms from the seawater, leaving the expelled water essentially sterile (Reiswig 1974, Pile 1997, Wehrl et al. 2007). Despite the fact that sponges feed on microorganisms, large amounts of extracellular microorganisms populate the mesohyl matrix of many demosponges (for recent reviews, see Hentschel et al. 2003, 2006, Imhoff and Sthr 2003, Hill 2004). Bacterial numbers may constitute as much as

40 - 60 % of the total biomass exceeding concentrations of seawater by two to four orders of magnitude. Molecular diversity analyses showed that the sponge microbiota is phylogenetically complex, yet highly sponge-specific. Members of eight eubacterial phyla [Proteobacteria (Alpha-, Gamma-, Deltaproteobacteria), Acidobacteria, Actinobacteria, Bacteroidetes, Chloroflexi, Cyanobacteria, Gemmatimonadetes and Nitrospira], members of the recently discovered candidate phylum Poribacteria (Fieseler et al. 2004), and of one archaeal phylum (Crenarchaeota) are numerically abundant and metabolically active in sponges. As none of these sponge-specific microorganisms have been obtained in pure culture, their function, metabolism, and possibly nutritional interactions with the host sponge are virtually unknown. In this study, we aimed to investigate the process of microbial nitrification in Mediterranean demosponges. Nitrification describes the oxidation of ammonia (NH3) to nitrite (NO2-) by ammonia-oxidizing bacteria (AOB) and archaea (AOA) and subsequently to nitrate (NO3-) by nitrite-oxidizing bacteria (NOB) for energy purposes (Kowalchuk and Stephen 2001). Several lines of evidence suggest that marine sponges are indeed a reservoir for

166

nitrifying microorganisms. Firstly, sponges and many other marine invertebrates release ammonia as a metabolic waste product (Wang and Douglas 1998, Davy et al. 2002) and nitrate excretion has already been documented in Caribbean (Corredor et al. 1988, Pile 1996, Diaz and Ward 1997) and Mediterranean (Jimenez and Ribes 2007) sponges. Secondly, 16S rRNA gene sequences of several clades of ammoniaoxidizing Beta- and Gammaproteobacteria and nitriteoxidizing Nitrospina were recovered from sponge tissues, making microbial nitrification a likely scenario (Hentschel et al. 2002, Diaz et al. 2004). In the present study, a combination of physiological and phylogenetic approaches was employed to explore the potential of microbial nitrification in Aplysina aerophoba Schmidt 1862 and several other Mediterranean demosponges.

DNA extraction and PCR

For the amplification of 16S rRNA genes from ammoniaoxidizing Betaproteobacteria, the primers AOB189f (5GGA GAA AAG CAG GGG ATC G-3) and AOB1224r (5-CGC CAT TGT ATT ACG TGT GA-3) were used, that originally had been designed as the FISH probes NSO190 and NSO1224, respectively (Loy et al. 2003). The PCR reaction mix contained 1 x PCR reaction buffer (Qiagen), 2 mM of each primer, 0.2 mM dNTPs (Sigma) and 1.25 U Taq Polymerase (Qiagen) in a final volume of 50 l. The PCR protocol was as follows: 1 min initial denaturation at 94C followed by 30 cycles of denaturation at 94C for 30 sec, primer annealing at 56C for 30 sec and elongation at 72C for 50 sec. The PCR was terminated with a final elongation step at 72C for 5 min.

Materials and methods Animal collection

Whole, intact colonies of the sponges A. aerophoba, Dysidea avara Schmidt 1862, Tethya sp. and Chondrosia reniformis Nardo 1847 were collected by SCUBA diving offshore Banyuls-sur-Mer (France) (4243N, 1008E) and from Rovinj (Croatia) (4505N, 1338E) at depths from 220 m. The animals were in the range of 30-50 g wet weight (about 15 g for D. avara). Small tissue pieces were removed from freshly collected animals, immediately frozen in liquid nitrogen and stored at -80C until use. Whole, intact animals were maintained in > 1000 L volume, flow-through holding tanks that were constantly supplied with fresh, untreated Mediterranean seawater prior to the experiments.

Cloning, RFLP-Analysis and Sequencing

Purified PCR products (PCR purification kit, Qiagen) were ligated into the pGEMT-easy vector (Promega) and transformed by electroporation into competent E. coli XL 1Blue cells. The enzymes Msp I and Hae III were used for restriction fragment length polymorphism (RFLP) analysis. Plasmid DNA was isolated from selected clones by standard miniprep procedures (Sambrook et al. 1989) and sequencing was performed on an ABI 377XL automated sequencer (Applied Biosystems).

Phylogenetic analysis
Sequences obtained in this study were checked for chimeras with the program Pintail. Sponge sequences together with reference sequences [received from GenBank using BLAST (http://www.ncbi.nlm.nih.gov/BLAST)] were aligned automatically with ClustalX and the alignment was subsequently corrected manually in Align. Neighbor Joining (with Jukes-Cantor correction) and Maximum Parsimony trees were constructed using the ARB software package (Ludwig et al. 2004).

Sponge incubations
Individual specimens were placed into aquaria containing three liters of fresh, untreated Mediterranean seawater. A constant water current was generated by small aquarium pumps (Vita Tech 300, Vitakraft, Germany). Only sponges that were in good physiological condition as judged by their regular filtration activity were chosen for the experiments. The experiments were performed in triplicate while a fourth aquarium without a sponge served as a control. In time intervals, 10 ml aliquots were removed from each aquarium, placed on ice and frozen at -20C until use.

Results and discussion In vivo sponge incubations

For A. aerophoba, nitrate excretion rates of 3.6 0.27 mol g-1 fresh weight day-1 (equivalent to 830 nmol g-1 dry weight h-1) were determined which corresponded to an ammonia excretion of 0.56 0.09 mol g-1 fresh weight day-1 (n = 4 S.E.) (Fig. 1A). The nitrate excretion rate was about six fold higher than the ammonia excretion rate. Ammonia and nitrate did not appear in the incubation water in sponge-free control aquaria. Nitrite was not detected in any of the incubations. In order to measure ammonia uptake rates, 100 or 200 M NH4Cl final concentrations were added to the incubation water. Aplysina aerophoba was capable of ammonia uptake which corresponded to a nitrate excretion rate of 9.2 and 5.0 mol g-1 fresh weight day-1 (Fig. 1B). Aplysina aerophoba was not capable of taking up nitrate which was tested at a concentration of 100 M (data not shown).

Determination of ammonia, nitrate and nitrite concentrations

The ammonia concentration was determined with the Indol-phenol-blue reaction (Parsons et al. 1984). The concentration of nitrite (NO2-) was determined by the Griess reaction (Parsons et al. 1984). Nitrate (NO3-) concentrations were measured indirectly after conversion to nitrite by the nirmutant E. coli strain JBC 606 as described in Pospesel et al. (1998). Ammonia and nitrate standard curves ranging from 0-100 M were performed for each measurement series and fresh standards were prepared on a weekly basis.

167

Fig. 1: A. Ammonia and nitrate excretion by A. aerophoba (n = 4 S.E.). B. Ammonia uptake and nitrate excretion (n = 2). Symbols represent ammonia () and nitrate () concentrations.

Ammonia and nitrate excretion rates of A. aerophoba were determined in correlation to the maintenance time in holding tanks (Fig. 2). After one day of maintenance, the ammonia excretion rate was 0.54 0.07 mol g-1 fresh weight day-1. After six and 11 days of maintenance, the ammonia excretion rates were slightly increased (1.04 0.05 and 1.11 0.05 mol g-1 fresh weight day-1, respectively, Krustal-Wallis Test p=0.051). Nitrate excretion rates were similar over time of maintenance, ranging from 0.74 0.15 (Fig. 2A), to 0.94 0.36 (Fig. 2B) and 0.87 0.43 mol g-1 fresh weight day-1 (Fig. 2C). Interestingly, a correlation between ammonia and nitrate excretion and sponge pumping activity was evident. While ammonia was excreted at almost double rates in non pumping sponges, nitrate was not excreted in sponges whose osculi were closed as judged by visual inspection (data not shown). The observation that the mesohyl of non-pumping sponges becomes anaerobic within 15 min (Hoffmann et al. unpublished) would be consistent with an inhibition of the aerobic process of nitrification. However, possible differences in the diffusion process of ammonia and nitrate in non pumping sponges cannot be excluded.

Fig. 2: Ammonia and nitrate excretion by A. aerophoba over time of maintenance in flow-through holding tanks (n = 3 S.E.). The rates were measured after one day (A), six days (B) and 11 days (C) after collection. Symbols represent ammonia () and nitrate () concentrations.

With respect to microbial loads in the mesohyl tissue, sponges are characterized as high and low microbial abundance sponges (Hentschel et al. 2006). Accordingly, ammonia and nitrate excretion rates were determined in the Mediterranean

168

weight day-1 (Fig. 3C) respectively, nitrate excretion was reduced or not detectable in any of the three sponge species investigated. Furthermore, Tethya sp. and C. reniformis were not able to take up ammonia (data not shown).

Phylogenetic analysis
In total, 200 clones were compared by restriction fragment length polymorphism analysis and four major restriction patterns were detected. After removal of five chimeras, nine, one, and two sequences from A. aerophoba, D. avara and Tethya sp. libraries, respectively, were used for phylogenetic tree construction (Fig. 4). All twelve sequences fell into the marine Nitrosospira cluster 1 of the Betaproteobacteria together with marine seawater and sediment sequences. Except Aplysina aerophoba (F) clone 5, the sequences obtained in this study build a subcluster with a high in-cluster similarity (98.5-99.9%). It is noteworthy that sequences were also recovered from the bacteria-free sponge D. avara as well as Tethya sp. whose mesohyl shows moderate amounts of microorganisms (Thiel et al. 2007). While it cannot be excluded that the cloned AOB sequences represent seawater bacteria rather than true sponge symbionts, the high nitrate excretion rates, at least for A. aerophoba, would argue for a specific and probably symbiotic association. Although primers used in this study also match Nitrosomonas species, no such bacteria could be detected in the clone libraries. Our findings expand those by Diaz (1997) and Diaz et al. (2004) who had reported on the identification of members of the Nitrosomonas eutropha/europae lineage (Betaproteobacteria) in five tropical sponges.

Modelling nitrogen fluxes in the sponge-microbe association

The following scenario, depicted in Fig. 5, is proposed based on this and other studies. The sponge host excretes ammonia as a metabolic waste product, which in turn, is oxidized to nitrite by ammonia-oxidizing bacteria (AOB), such as Nitrosospira (this study), Nitrosococcus (Hentschel et al. 2002) or members of the Nitrosomonas eutropha/ europaea l lineage (Diaz et al. 2004). Nitrite is further oxidized to nitrate by nitrite-oxidizing bacteria (NOB), such as Nitrospina or members of the phylum Nitrospira (Hentschel et al. 2002). The coordinated action of members of these two groups might then be responsible for the conversion of ammonia to nitrate in A. aerophoba and possibly also in other sponges. In addition to eubacteria, the involvement of archaea should be considered in future studies. Recent literature illustrates

Fig. 3: Ammonia and nitrate excretion by D. avara (A), Tethya sp. (B), and C. reniformis (C), (each species, n = 3 S.E.). Symbols represent ammonia () and nitrate () concentrations.

low microbial abundance sponges D. avara and Tethya sp. and the high microbial abundance sponge C. reniformis (Fig. 3). While ammonia was excreted at similar rates of 3.0 0.39 mol g-1 fresh weight day -1 (Fig. 3A), 6.95 0.08 mol g1 fresh weight day-1 (Fig. 3B) and 5.8 0.9 mol g-1 fresh

Fig. 4: Distance 16S rRNA (1053 bp) gene phylogeny showing the relationships between ammonia-oxidizing Betaproteobacteria using the ARB program package. Sponge derived sequences are shown in bold. Neighbor-joining and maximum parsimony (100 pseudoreplicates) bootstrap values are indicated. Arrow to outgroup (Nitrosococcus oceani Nc1 (AJ298727)). Designation of clusters is adapted from Freitag and Prosser (2003). Scale bar indicates 1% sequence divergence.

169

170

Fig. 5: Schematic diagram showing the inferred nitrogen-fluxes resulting from nitrification in the sponge-microbe-association.

that archaea, rather than bacteria, might in fact be involved in nitrification in marine and terrestrial ecosystems (Leininger et al. 2006, Wuchter et al. 2006). In fact, close relatives of the Cenarchaeum symbiosum lineage that are also present in sponges have recently been isolated and were shown to be capable of nitrification (Knneke et al. 2005). Additionally, it needs to be investigated whether nitrate serves as an energy substrate for denitrifying microorganisms under anaerobic conditions. In conclusion, this study contributes to an ongoing effort to link microbial diversity with function in these phylogenetically highly diverse, elusive and so far uncultivated marine sponge-associated microbial communities.

References
Brusca RC, Brusca GJ (1990) Phylum Porifera: the sponges. In: Sinauer AD (ed). Invertebrates. Sinauer Press, Sunderland. pp. 181-210 Corredor JE, Wilkinson CR, Vicente VP, Morell JM, Otero E (1988) Nitrate release by Caribbean reef sponges. Limnol Oceanogr 33(1): 114-120 Davy SK, Trautman DA, Borowitzka MA, Hinde R (2002) Ammonium excretion by a symbiotic sponge supplies the nitrogen requirements of the rhodophyte partner. J Exp Biol 205: 35053511 Diaz MC (1997) Molecular detection and characterization of specific bacterial groups associated to tropical sponges. Proc 8th Int Coral Reef Symp, Balboa 2: 1399-1402 Diaz MC, Ward BB (1997) Sponge-mediated nitrification in tropical benthic communities. Mar Ecol Progr Ser 156: 97-107 Diaz MC, Akob D, Cary CS (2004) Denaturing gradient gel electrophoresis of nitrifying microbes associated with tropical sponges. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Boll Mus Ist Biol Univ Genova 68: 279-289 Fieseler L, Horn M, Wagner M, Hentschel U (2004) Discovery of a novel candidate phylum Poribacteria in marine sponges. Appl Environ Microbiol 70: 3724-3732 Freitag TE, Prosser JI (2003) Community structure of ammoniaoxidizing bacteria within anoxic marine sediments. Appl Environ Microbiol 69(3): 1359-1371

Acknowledgements
We thank Prof. F. Brmmer (University of Stuttgart), Prof. W.E.G. Mller (University of Mainz), Dr. R. Batel (Institute Rudjer Boskovic, Croatia) and all BiotecMarin colleagues as well as Dr. F. Hoffmann (MPI Bremen, Germany) for sampling support and many interesting discussions. Additionally, we thank C. Gernert for excellent laboratory assistance (University of Wuerzburg). Financial support was provided by bmb+f Center of Competence BiotecMarin (FKZ 03F0414E) and Sonderforschungsbereich SFB 567 - TPC3 grant to U. H.

171

Hentschel U, Fieseler L, Wehrl M, Gernert C, Steinert M, Horn M, Hacker J (2003) Microbial diversity of marine sponges. In: Mller WEG (ed). Molecular biology of sponges. Springer-Verlag, Heidelberg. pp. 59-88 Hentschel U, Hopke J, Horn M, Friedrich AB, Wagner M, Hacker J, Moore BS (2002) Molecular evidence for a uniform microbial community in sponges from different oceans. Appl Environ Microbiol 68: 4431-40 Hentschel U, Usher KM, Taylor MW (2006) Marine sponges as microbial fermenters. FEMS Microbiol Ecol 55(2): 167-177 Hill RT (2004) Microbes from marine sponges: a treasure trove of biodiversity for natural products discovery. In: Bull AT (ed). Microbial diversity and bioprospecting. ASM Press, Washington DC. pp 177-190 Hooper JNA, van Soest RWM (2002) Systema Porifera: a guide to the classification of sponges, vol. 1. Kluwer Academic/Plenum Publishers, New York Imhoff JF, Sthr R (2003) Sponge-associated bacteria: general overview and special aspects of bacteria associated with Halichondria panicea. In: Mller WEG (ed). Molecular biology of sponges. Springer Verlag, Heidelberg. pp. 35-56 Jimnez E, Ribes M (2007) Sponges as a source of dissolved inorganic nitrogen: nitrification mediated by temperate sponges. Limnol Oceanogr 52(3): 948-958 Kowalchuk GA, Stephen JR. (2001) Ammonia-oxidizing bacteria: a model for molecular microbial ecology. Annu Rev Microbiol 55: 485-529 Knneke M, Bernhard AE, de la Torre JR, Walker CB, Waterbury JB, Stahl DA (2005) Isolation of an autotrophic ammonia-oxidizing marine archaeon. Nature 437(7058): 543-546 Leininger S, Urich T, Schloter M, Schwark L, Qi J, Nicol GW, Prosser JI, Schuster SC, Schleper C (2006) Archaea predominate among ammonia-oxidizing prokaryotes in soils. Nature 442(7104): 806-809 Li CW, Chen JY, Hua TE (1998) Precambrian sponges with cellular structures. Science 279: 879-82 Loy A, Horn M, Wagner M (2003) ProbeBase - an online resource for rRNA-targeted oligonucleotide probes. Nucleic Acids Res 31: 514-516 (http://www.microbial-ecology.net/probebase/)

Ludwig W, Strunk O, Westram R, Richter L, Meier H, Yadhukumar, Buchner A, Lai T, Steppi S, Jobb G, Forster W, Brettske I, Gerber S, Ginhart AW, Gross O, Grumann S, Hermann S, Jost R, Konig A, Liss T, Lussmann R, May M, Nonhoff B, Reichel B, Strehlow R, Stamatakis A, Stuckmann N, Vilbig A, Lenke M, Ludwig T, Bode A, Schleifer KH (2004) ARB: a software environment for sequence data. Nucleic Acids Res 32: 1363-1371 Parsons TR, Maita Y, Lalli CM (1984) Manual of chemical and biological methods for seawater analysis. Pergamon Press, New York Pile AJ (1996) The role of microbial food webs in benthic-pelagic coupling in freshwater and marine ecosystems. PhD thesis. College of William and Mary, Williamsburg Pile AJ (1997) Finding Reiswigs missing carbon: quantification of sponge feeding using dual-beam flow cytometry. Proc 8th Intern Coral Reef Symp, Balboa 2: 1403-1410 Pospesel M, Hentschel U, Felbeck H (1998) Determination of nitrate in the blood of the hydrothermal vent tubeworm Riftia pachyptila using nitrate reduction assay. Pergamon Deep Sea Res I (45): 2189-2200 Reiswig H (1974) Water transport, respiration and energetics of three tropical marine sponges. J Exp Mar Biol Ecol 14: 231-249 Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: A laboratory manual (second edition). Cold Spring Harbor Press, Cold Spring Harbor, New York Thiel V, Neulinger SC, Staufenberger T, Schmalljohann R, Imhoff JF (2007) Spatial distribution of sponge-associated bacteria in the Mediterranean sponge Tethya aurantium. FEMS Microbiol Ecol 59: 47-63 Wang JT, Douglas AE (1998) Nitrogen recycling or nitrogen conservation in an alga-invertebrate symbiosis? J Exp Biol 201: 2445-2453 Wehrl M, Steinert M, Hentschel U (2007) Bacterial uptake by the marine sponge Aplysina aerophoba. Microb Ecol 53(2): 355-365 Wuchter C, Abbas B, Coolen MJ, Herfort L, van Bleijswijk J, Timmers P, Strous M, Teira E, Herndl GJ, Middelburg JJ, Schouten S, Sinninghe Damste JS (2006) Archaeal nitrification in the ocean. Proc Natl Acad Sci USA 103(33): 12317-12322

Porifera research: Biodiversity, innovation and sustainaBility - 2007

173

Perplexing distribution of 3-alkylpyridines in haplosclerid sponges

Leontine E. Becking(1*), Yoichi Nakao(2), Nicole J. de Voogd(1), Rob W.M. van Soest(3), Nobuhiro Fusetani(2), Shigeki Matsunaga(2)
National Museum of Natural History Naturalis, Dept. Zoology, P.O. Box 9517, 2300 RA Leiden, The Netherlands. becking@naturalis.nnm.nl (2) University of Tokyo, Laboratory of Aquatic Natural Products Chemistry, The University of Tokyo, Bunkyo-ku, Tokyo 113-8657, Japan (3) University of Amsterdam, Zoological Museum, P.O. Box 94766, 1090 GT Amsterdam, The Netherlands
(1)

Abstract: In this study we reviewed the natural product literature for the distribution of 3-alkylpyridines among sponge taxa. In parallel, we traced selected 3-alkylpyridines, amphitoxins, in three haplosclerid genera (Amphimedon, Callyspongia, Haliclona) in order to establish the utility of such compounds as genuine chemotaxonomic markers. We confirmed that this group of compounds had been almost solely extracted from sponges of the order Haplosclerida. Three groups of compounds within the 3-alkylpyridine derivatives were noteworthy, as they appear to be concentrated in restricted taxonomic units within the order Haplosclerida: 1) polymers, 2) cyclic dimers, and 3) bicyclic dimers. There was a concentration of the polymer amphitoxin in the families Niphatidae and Callyspongiidae of the suborder Haplosclerina, and more particularly in the genera Amphimedon and Callyspongia. Our experimental results reconfirmed the presence of amphitoxin in Callyspongia (Euplacella) biru, but we were unable to trace any amphitoxins or 3-alkylpyridines of any kind in Callyspongia (Callyspongia) truncata, Haliclona sp., and Amphimedon aff. queenslandica. Assuming that these classifications are correct, our results diminish the value of both amphitoxins and 3-alkylpyridines as monophyletic markers. Keywords: 3-alkylpyridines, amphitoxin, chemotaxonomy, Order Haplosclerida, secondary metabolites

Introduction
Sponges have proven to be a magnificent source of numerous highly bioactive compounds with pharmaceutical potential (Faulkner 2000, Sipkema et al. 2005). Since the 1960s natural product chemists have been active in extracting these compounds from sponges, which were collected directly and randomly from the seas. For the past thirty decades several authors have been considering sponge compounds as candidates for additional markers that could potentially be useful in verifying or aiding the standing phylogeny of the Porifera. (e.g. Bergquist and Hartman 1969, van Soest and Braekman 1999, Erpenbeck and van Soest 2007). The present classification of sponges is mainly based on external and internal morphological characters and some life history characteristics (Hooper and van Soest 2002). This classification of sponges is not infrequently disputed amongst sponge taxonomists and in fact the phylogenetic tree has been altered several times over the past decades. The field of chemotaxonomy in its turn has, however, yet to prove its utility. The bulk of published data on natural products from marine organisms has been archived in a database called MarinLit (Munro and Blunt 2004) that is updated annually. Andersen et al. (1996) and van Soest and Braekman (1999) reviewed the MarinLit database to discuss the utility of secondary

metabolites of sponges as systematic tools. One conclusion from these studies was that 3-alkylpyridine derivatives might be useful markers for the Order Haplosclerida. Since the publication of the reviews of Andersen et al. (1996) and van Soest and Braekman (1999), numerous additional compounds have been isolated from sponges. We, therefore, once again reviewed the distribution of 3-alkylpyridines in sponge genera. A well-reported problem with allocating taxonomic relevance to a compound based on the results of literature reviews is the strong bias of natural product laboratories to publish only novel compounds, thereby confounding any conclusions on the absence of compounds in certain taxonomic groups. To overcome this problem we selected particular 3-alkylpyridine derivatives, amphitoxins, and traced their distribution in representatives of three genera to establish whether such compounds could be viewed as genuine chemotaxonomic markers.

Material and methods Literature review

MarinLit (Munro and Blunt, 2004), Web of Science and scholar.google.com were reviewed for publications reporting extraction of 3-alkylpyridine derivatives from sponges. Keywords of all known names within this compound

174

group were submitted in the databases, as well as the names of the genera within the Order Haplosclerida.

Results Literature review

We identified 49 records reporting extraction of 3alkylpyridine alkaloids from sponges. The majority of the records were identified to genus-level and belonged to the Order Haplosclerida. There were three records of 3-alkylpyridines extracted from sponges of other orders: Theonella swinhoei (Lithistida, from Kobayashi et al. 1989), Batzella sp. (Poecilosclerida, from Segraves and Crews 2005), and Halichondria sp. (Halichondrida, from Chill et al. 2002). Within the Order Haplosclerida, 3-alkylpyridines have been recorded from the 5 of the 6 main marine haplosclerid families, but have not been recorded from all genera belonging to these families. In the suborder Haplosclerina the families represented were Callyspongiidae (2 genera, 5 species), Chalinidae (2 genera, 3 species), Niphatidae (3 genera, 6 species), and in the suborder Petrosina the families were Petrosiidae (2 genera, 3 species), and Phloedictyidae (1 genus, 1 species). Three groups of compounds appeared to be concentrated in small taxonomic units (see Table 1A-C): 1. Polymers (hali/amphitoxins) 2. Cyclic dimers (cyclostelletamines/haliclamines) 3. Bicyclic dimers (haliclonacyclamines/halicyclamines/ arenosclerins)

Collection, extraction and fractionation

Four tropical reef sponge species were collected at depths ranging from 2-10 m: Callyspongia (Euplacella) biru de Voogd 2004 (Sulawesi, Indonesia), Callyspongia (Callyspongia) truncata (von Lendenfeld, 1887) (Shizuoka, Japan), Amphimedon aff. queenslandica Hooper and van Soest 2006 (Okinawa, Japan), Haliclona sp. (Kagoshima, Japan), (see Fig. 1A-D). Voucher specimens have been deposited at the Zoological Museum, University of Amsterdam, the Netherlands (C. (E.) biru collection number POR.15222), and at the National Museum of Natural History Naturalis, Leiden The Netherlands (C. (C.) truncata, A. aff. queenslandica, Haliclona sp. collection numbers respectively: POR.2971, POR.2972, POR.2973). Two individuals of each species were screened for amphitoxins. Amphitoxins are known to be mainly responsible for the bioactivity of the crude extracts of C. (E.) biru (Dubut 2000 as Callyspongia sp. blue; de Voogd et al. 2005). To date no amphitoxins have been reported from C. (C.) truncata,while a large number of bioactive polyacetylenes have been extracted from this species (Nakao et al. 2002). Each sponge was extracted with MeOH (500 ml x 3) and the extract was partitioned between water and CHCl3. The water-soluble fraction was extracted with n-butanol. The organic phase was partitioned between MeOH/H2O (9:1) and n-hexane. The aqueous MeOH layer was then partitioned between MeOH/H2O (6:4) and CHCl3. Finally the n-butanol fraction and MeOH/H2O (6:4) fraction were combined. All fractions were analyzed by TLC on silica gel with CHCl3/ MeOH/H2O (7:3:0.5) and different mixtures of CHCl3/ MeOH. Tertiary or quaternary bases were visualized by spraying with Dragendorff reagent. The fractions exhibiting orange/red spots were subsequently fractionated by ODS flash chromatography using stepwise elution of aqueous MeOH (0 % to 100 % MeOH), followed by elution with CHCl3/MeOH/H2O (7:3:0.5). The fractions showing spots positive to Draggendorff on TLC were further separated by ODS HPLC (Column: COMOSIL-5C18-ARII 10 x 250 mm; flow rate 2 ml/min.; UV 266 nm) using gradient elution from 35 % MeCN/0.05 % TFA to 80 % MeCN/0.05 % TFA. The peaks positive to Draggendorff were analyzed by 1H NMR spectroscopy to examine if typical proton signals for amphitoxin were present. As C.(E.) biru has previously been shown to contain amphitoxins (Dubut 2000, de Voogd et al. 2005), the above procedure was first performed on this species to obtain a standard reference for amphitoxin. After a standard was obtained, the extracts of the other sponges (after solvent partitioning and ODS flash) were first checked for the presence of a similar compound by comparing Rf values on TLC with that of amphitoxin, before subjected to HPLC.

Extraction and fractionation

Amphitoxin was successfully extracted from the C. (E.) biru specimens, and was present in the 20-100% elution by ODS chromatography of the combined n-butanol and MeOH/ H2O (6:4) fractions. We were unable to trace any amphitoxins or 3-alkylpyridines from the extracts of the C. (C.) truncata, Haliclona sp. or A. aff. queenslandica sp. specimens.

Discussion
We confirmed that 3-alkylpyridines have almost exclusively been isolated from sponges of the order Haplosclerida, after having made the necessary amendments on sponge taxonomy and nomenclature. There are, however, three reports of 3alkylpyridines from sponges of other orders: Theonella swinhoei is a probable case of mislaid labels or undetected haplosclerid overgrowth (cf. van Soest and Braekman 1999), and Batzella sp. and Halichondria sp. remain unresolved for the time being. It must be noted, that these genera share a single diactinal spicule type with Haplosclerida, making it possible that they have been misidentified Haplosclerida. In our view the use of more specific compound structures/ groups, rather than the relatively broad nominator 3alkylpyridine derivatives, ought to be the next step in supporting classifications based on compounds. In previous reviews linear 3-alkylpiperidines appeared to be concentrated in the families Callyspongiidae and Niphatidae, while cyclic 3-alkylpiperidines were concentrated in Chalinidae and Petrosiidae (Andersen et al.

175

Fig. 1A-D: Sponges examined in this study: A. Callyspongia (Euplacella) biru (photo: B. W. Hoeksema), B. Callyspongia (Callyspongia) truncata (photo: S. Hoshino), C. Amphimedon aff. queenslandica (photo: L. E. Becking), D. Haliclona sp. (photo: Y. Nakao).

1996, van Soest and Braekman 1999). Records of extracted compounds published since that review (notably the cyclic 3-alkylpiperidines found in Amphimedon sp. by Matsunaga et al. 2004), however, indicate that a taxonomic distinction based on the presence of cyclic or linear 3-alkylpiperidine derivatives can most likely no longer be upheld. Three groups of compounds within the 3-alkylpyridines were noteworthy in that they have structures with presumed similar biogenetic pathways and appear to be concentrated in small taxonomic units (see Table 1A-C). Based on the common presence of these specific compounds, the close relationship between Callyspongiidae, Chalinidae, Niphatidae, and Petrosiidae is confirmed. Previous works suggested that the closely related polymers halitoxin and amphitoxin might be markers for the family Niphatidae. Our review showed a concentration of these types of compounds not only in Niphatidae, but also in Callyspongiidae, and more particularly in the genera Amphimedon and Callyspongia (Table 1A).

Our laboratory work reconfirmed the presence of amphitoxin in C. (E.) biru, but we were unable to trace any amphitoxins from C. (C.) truncata, nor from the Haliclona sp. and A. aff. queenslandica. In fact, the situation is rather more complex as we did not extract 3-alkylpyridines of any kind from these three specimens. The sponge specimens that did not contain 3-alkylpyridines were collected from Japan, but we do not suspect this is a case of geographic variation of compound distribution as these compounds have been isolated from Japanese sponges before (Fusetani et al. 1989, 1994, Matsunaga et al. 2004). Causes of these results may lie in the frequently reported high natural variation in compound production, which in turn may be due to physical and environmental variation (e.g. Thacker et al. 1998, de Voogd et al. 2004, de Voogd 2007) or to symbionts that may be the true producers of the compounds (e.g. Jadulco et al. 2002, Becerro and Paul 2004). Sponges generally harbor vast amounts of symbionts, among which there are not only the specialists, which may

176

Table 1A-C: Three compound groups that appear to be concentrated in restricted taxonomic units (compound figures reproduced from Andersen et al. 1996): A. polymers, B. cyclic dimers, C. bicyclic dimers.

A. Polymers

Family Callyspongidae Callyspongidae Callyspongidae Niphatidae Niphatidae Niphatidae Niphatidae Niphatidae Niphatidae Niphatidae Chalinidae Chondropsidae

Species Callyspongia (Euplacella) biru Callyspongia (Cladochalina) fibrosa Callyspongia ridleyi Amphimedon compressa Amphimedon erina Amphimedon viridis Amphimedon viridis Amphimedon viridis Amphimedon compressa Amphimedon paraviridis Haliclona (Rhizoniera) sarai Batzella sp.

Compound Amphitoxin Halitoxin Halitoxin Halitoxin Halitoxin Halitoxin Halitoxin Hali/Amphitoxin Amphitoxin Amphitoxin Halitoxin Halitoxin

Location Indonesia Micronesia unknown Caribbean Caribbean Caribbean Brazil Red Sea Bahamas Indonesia Adriatic Sea Madagascar

Identification de Voogd van Soest unknown Hartman Hartman Hartman Hajdu unknown Genoa Museum de Voogd Vacelet Diaz

Reference de Voogd et al. 2005 Davies-Coleman et al. 1993 Scott et al. 2000 Schmitz et al. 1978 Schmitz et al. 1978 Schmitz et al. 1978 Berlinck et al. 1996 Kelman et al. 2001 Albrizio et al. 1995 de Voogd et al 2005 Sepi et al. 1997 Segraves and Crews 2005

B. Cyclic dimers

Family Chalinidae Chalinidae Chalinidae Chalinidae Niphatidae Petrosiidae

Species Haliclona sp. Haliclona (Rhizoniera) viscosa Haliclona sp. Haliclona (Rhizoniera) viscosa Pachychalina sp. Xestospongia sp.

Compound Haliclamines Haliclamines Cyclostellettamines Cyclostellettamines Cyclostellettamines Cyclostellettamines

Location Japan Arctic Japan Arctic Brazil Japan

Identification Watanabe de Weerdt van Soest de Weerdt Hajdu van Soest

Reference Fusetani et al. 1989 Volk et al. 2004 Fusetani et al. 1994 Volk and Kck 2004 de Oliviera et al. 2004 Oku et al. 2004

C. Bicyclic dimers

Family Niphatidae Niphatidae Niphatidae Callyspongiidae Callyspongiidae Chalinidae Chalinidae Petrosiidae Halichondriidae

Species Amphimedon sp. Amphimedon sp. Amphimedon sp. Arenosclera braziliensis Arenosclera braziliensis Haliclona sp. Haliclona sp. Xestospongia sp. Halichondria sp.

Compound Tetradehydrohalicyclamine 22-hydroxyhalicyclamine Tetrahydrohalicyclamine Arenosclerins Haliclonacyclamines Haliclonacyclamines Halicyclamine Halicyclamine Halichondramine

Location Japan Japan Japan Brazil Brazil Australia Indonesia Indonesia Eritrea

Identification van Soest van Soest van Soest Hajdu Hajdu Hooper unknown Diaz van Soest

Reference Matsunaga et al. 2004 Matsunaga et al. 2004 Matsunaga et al. 2004 Torres et al. 2002 Torres et al. 2002 Clark et al. 1998, Charan et al. 1996 Jaspars et al. 1994 Harrison et al. 1996 Chill et al. 2002

177

have co-evolved with the host sponge, but also the generalist microbes which can be found in a wide array of sponges and some of which are even widely present in seawater (Lee et al. 2001, Taylor et al. 2004). When identical or very similar compounds are found in species from different orders or genera, microbial producers are suspected (e.g. bacteria and fungi). If these generalist symbionts play any role in the production of the toxic compounds extracted from the sponges, this could mystify the classifications based on chemotaxonomy. Erpenbeck and van Soest (2007) provide a full review of the various pitfalls in the use of compounds in chemotaxonomy. They illustrate that the major problems are not only related to the great natural variation in compound production and ambiguous origins of compounds, but also the unknown homology of compounds, misidentifications of sponge species and the lack of reporting by natural products chemists of presence/absence of known compounds. To conclude, our results diminish the value of both 3alkylpyridines and amphitoxins as monophyletic markers assuming that the classifications are correct. The classification of haplosclerid families and genera is presently under siege from molecular studies (e.g. McCormack et al. 2002, Nichols 2005). Thus, this study may prove of some value in the ongoing debate on the classification of the poriferan phylum. Any definate conclusions of the utility of 3-alkylpyridines as chemotaxonomic markers would, however, be presumptuous before a more comprehensive systematic study is performed, particularly of genera that have not been examined previously, and the presence in the other orders is unequivocally ruled out.

Acknowledgements
The field- and laboratory-work was financed by NWO-WOTRO and the Japan Prizewinners Programme administered by the Dutch Ministry of Education, Culture and Science. The first author is also grateful for financial support received by the Scientific Committee of the 7th International Sponge Symposium and the Jan Joost ter Pelkwijk Fonds, which made attending the symposium possible. We would like to thank Dr. Hoeksema and Dr. Hoshino for providing photographs and Yamashita-san for all his friendly help in the lab.

References
Albrizio S, Cimiello P, Fattorusso E, Magno S, Pawlik JR (1995) Amphitoxin, a new high molecular weight antifeedent pyridinium salt from the Caribbean sponge Amphimedon compressa. J Nat Prod 58: 647-652 Andersen RJ, van Soest RWM, Kong F (1996) 3-alkylpiperidine alkaloids isolated from marine sponges in the order Haplosclerida. In: Pelletier SW (ed). Alkaloids: Chemical and Biological Perspectives, Vol 10. Pergamon, Oxford, pp 301-356 Becerro MA and Paul VJ (2004) Effects of depth and light on secondary metabolites and cyanobacterial symbionts of the sponge Dysidea granulosa. Mar Ecol Prog Ser 280: 115-128 Bergquist PR, Hartman WD (1969) Free amino acid patterns and the classification of the Demospongiae. Mar Biol 3: 247-268 Berlinck RGS, Ogawa CA, Almeida AMP, Sanchez MAA, Malpezzi ELA, Costa LV, Hajdu E, Freitas de JC (1996) Chemical and Pharmacological characterization of halitoxin from Amphimedon

viridis (Porifera) from the Southeastern Brazilian Coast. Comp Biochem Phys 115C: 155-163 Charan RD, Garson MJ, Brereton IM, Willis AC, Hooper JNA (1996) Haliclonacyclamines A and B, cytotoxic alkaloids from the tropical marine sponge Haliclona sp. Tetrahedron 52: 9111-9120 Chill L, Miroz A, Kashman Y (2000) Haliclonyne, a new highly oxygenated polyacetylene from the marine sponge Haliclona species. J Nat Prod 63: 523-526 Clark RJ, Field KL, Charan RD, Garson MJ, Brereton IM, Willis AC (1998) The Haliclonacyclamines, cytotoxic tertiary alkaloids from the tropical marine sponge Haliclona sp. Tetrahedron 54 (30): 8811-8826 Davies-Coleman MT, Faulkner DJ, Dubowchik GM, Roth GP, Polson C, Fairchild C (1993) A new EGF-active polymeric pyridinium alkaloid from the sponge Callyspongia fibrosa. J Org Chem 58: 5925-5930 de Voogd NJ (2004) Callyspongia (Euplacella) biru spec. nov. (Porifera: Demospongiae: Haplosclerida) from Indonesia. Zool Meded 78: 477-483 de Voogd NJ, Becking LE, Noor A, Hoeksema BW, van Soest RWM (2004) Sponge interactions with spatial competitors in S.W. Sulawesi. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Boll Mus Ist Biol Univ Genova 68: 253-261 de Voogd NJ, Haftka JJH, Hoeksema BW (2005) Evaluation of the ecological function of amphitoxin in the reef-dwelling sponge Callyspongia (Euplacella) biru (Haplosclerida: Callyspongiidae) at southwest Sulawesi, Indonesia. Contr Zool 74: 51-59 de Voogd NJ (2007) The mariculture potential of the Indonesian reef-dwelling sponge Callyspongia (Euplacella) biru: growth, survival and bioactive compounds. Aquaculture 262:54-64 Dubut D (2000) Isolment et dtermination de structure de toxins dponges du genre Callyspongia. Student Report, Universit Libre de Bruxelles, Belgium Erpenbeck E, van Soest RWM (2007) Status and perspective of sponge chemosystematics. Mar Biotech 9: 2-19 Faulkner DJ (2000) Highlights of marine natural products chemistry. Nat Prod Rep 17: 1-6 Fusetani N, Yasumuro K, Matsunaga S, Hirota H (1989) Haliclamines A and B, cytotoxic macrocyclic alkaloids from a sponge of the genus Haliclona. Tetrahedron Lett 30: 6891 Fusetani N, Asai N, Honda K, Yasumuro K (1994) Cyclostellettamines A-F, pyridine alkaloids which inhibit binding of methyl quinuclidinyl benzylate (QB) to muscarine acetylcholine receptors, from the marine sponge Stelletta maxima. Tetrahedron Lett 35: 3967-3970 Harrison B, Talapatra S, Lobkovsky E, Clardy J, Crews P (1996) The structure and biogenetic origin of (-) Halicyclamine B from a Xestospongia sponge. Tetrahedron Lett 37: 9151-9154 Hooper JNA, van Soest RWM (2002) Systema Porifera: a guide to the classification of sponges. Kluwer Academic/Plenum Publishers, New York Hooper JNA, van Soest RWM (2006) A new species of Amphimedon (Porifera, Demospongiae, Haplosclerida, Niphatidae) from the Capricorn-Bunker Group of Islands, Great Barrier Reef, Australia: target species for the sponge genome project. Zootaxa 1314: 3139 Jadulco R, Brauers G, Angelie R, Ebel R, Wray V, Sudarsono, Proksch P (2002) New metabolites from sponge-derived fungi

178

Curvularia lunata and Cladosporium herbarum. J Nat Prod 65: 730-733 Jaspars M, Pasupathy V, Crews P (1994) A tetracyclic diamine alkaloid, halicyclamine A, from the marine sponge Haliclona sp. J Org Chem 59: 3253-3255 Kelman D, Kashman Y, Rosenberg E, Ilan M, Ifrach I, Loya Y (2001) Antimicrobial activity of the reef sponge Amphimedon viridis from the Red Sea: evidence for selective toxicity. Aquat Microb Ecol 24: 9-16 Kobayashi J, Murayama T, Ohizumi Y, Sasaki T, Ohta T, Nozoe S (1989) Theonelladins A-D, novel antineoplastic pyridine alkaloids from the Okinawa marine sponge Theonella swinhoei. Tetrahedron Lett 30 (22): 2963-2966 Lee YK, Lee JH, Lee HK (2001) Microbial symbiosis in marine sponges. J Microbiol 39(4): 254-264 Matsunaga S, Shinoda K, Fusetani N (1993) Cribrochalinamine oxides A and B, antifungal beta-substituted pyridines with an azomethine N-oxide from the marine sponge Cribrochalina sp. Tetrahedron Lett 34: 5953 Matsunaga S, Miyata Y, van Soest RWM, Fusetani N (2004) Tetradehydrohalicyclamine A and 22-hydroxyhalicyclamine A, new cytotoxic bis-piperidine alkaloids from a marine sponge Amphimedon sp. J Nat Prod 67: 1758-1760 McCormack GP, Erpenbeck D, van Soest RWM (2002) Major discrepancy between phylogenetic hypotheses based on molecular and morphological criteria within the Order Haplosclerida (Phylum Porifera: Class Demospongiae). J Zool Syst Evol Res 40: 237-240 Munro MGH, Blunt JW (2004) MarinLit. A database of the literature on marine natural products for the use of macintosh computers prepared and maintained by the Marine Chemistry Group. Department of Chemistry, University of Canterbury, New Zealand Nakao Y, Uehara T, Matsunaga S, Fusetani N, van Soest RWM (2002) Callyspongic acid, a polyacetylenic acid which inhibits glucosidase, from the marine sponge Callyspongia truncata. J Nat Prod 65: 922-924 Nichols SA (2005) An evaluation of support for order-level monophyly and interrelationships within the class Demospongiae using partial data from the large subunit rDNA and cytochrome oxidase subunit I. Mol Phylogenet Evol 34: 81-96 Oku N, Nagai K, Shindoh N, Tetrada Y, van Soest RWM, Matsunaga S, Fusetani N (2004) Three new cyclostellettamines, which

inhibit histone deacetylase, from a marine sponge of the genus Xestospongia. Bioorg Med Chem Lett 14: 2617-2620 Oliveira JHHL de, Grube A, Kock M, Berlinck RGS, Macedo ML, Ferreira AG, Hajdu E (2004) Ingenamine G and Cyclostelettamines G-I, K and L from the new Brazilian species of marine sponge Pachychalina sp. J Nat Prod 67: 685-1689 Schmitz FJ, Hollenbeak KH, Campbell DC (1978) Marine natural products: halitoxin, toxic complex of several marine sponges of the genus Haliclona. J Org Chem 43: 3316-3822 Scott RH, Whyment AD, Foster A, Gordon KH, Milne BF, Jaspars M (2000) Analysis of the structure and electrophysiological actions of halitoxins: 1,3-alkylpyridinium salts from Callyspongia ridleyi. J Membr Biol 176: 119-131 Segraves NL, Crews P (2005) A Madagascar sponge Batzella sp. as a source of alkylated iminosugars. J Nat Prod 68: 118-121 Sepi K, Guella G, Mancini I, Pietra F, Dalla Serra M, Menestrina G, Tubbs K, Macek P, Turk T (1997) Characterization of anticholinesterase-active 3-alkylpyridinium polymers from the marine sponge Reniera sarai in aqueous solutions. J Nat Prod 60: 991-996 Sipkema D, Franssen MCR, Osinga R, Tramper J, Wijffels RH (2005) Marine Sponges as Pharmacy. Mar Biotech 7: 142-162 Taylor MW, Schupp PJ, Dahllof I, Kjelleberg S, Steinberg PD (2004) Host specificity in marine sponge-associated bacteria, and potential implications for marine microbial diversity. Environ Microbiol 6: 121-130 Thacker RW, Becerro MA, Lumbang WA, Paul VJ (1998) Allelopathic interactions between sponges on a tropical reef. Ecology 79: 1740-1750 Torres YR, Berlinck RGS, Nascimento GGF, Fortier SC, Pessoa C, de Moraes MO (2002) Antibacterial activity against resistant bacteria and cytotoxicity of four alkaloid toxins isolated from the marine sponge Arenosclera brasiliensis. Toxicon 40: 885-891 van Soest RWM, Braekman J-C (1999) Chemosystematics of Porifera: a review. Memoir Queensl Mus 44: 569-589 Volk CA, Lippert H, Lichte E, Kck M (2004) Two new haliclamines from the Arctic sponge Haliclona viscosa. Eur J Org Chem 14: 3154-3156 Volk CA, Kck M (2004) Viscosaline: new 3-alkyl pyridinium alkaloid from the arctic sponge Haliclona viscosa. Org Biomol Chem 2: 1827-1830 Web of Science: http://portal.isiknowledge.com

Porifera research: Biodiversity, innovation and sustainaBility - 2007

179

Lake Baikal endemic sponge Lubomirskia baikalensis: structure and organization of the gene family of silicatein and its role in morphogenesis
Sergey I. Belikov(1,2), Oksana V. Kaluzhnaya(1,2,3), Heinz C. Schrder(3,2), Isabel M. Mller(3), Werner E.G. Mller(2,3*)
Limnological Institute of the Siberian Branch of Russian Academy of Sciences, Ulan-Batorskaya 3, RUS-664033 Irkutsk, Russia (2) Limnological Institute of the Siberian Branch of Russian Academy of Sciences, Joint Russian-German Laboratory for Biology of Sponges, Ulan-Batorskaya 3, RUS-664033 Irkutsk, Russia (3) Institut fr Physiologische Chemie, Abteilung Angewandte Molekularbiologie, Universitt, Duesbergweg 6, D-55099 Mainz, Germany. tel.: +49-6131-392-5910; fax.: +49-6131-392-5243. wmueller@uni-mainz.de
(1)

Abstract: Lake Baikal is known for its abundant endemic fauna and flora. The photosymbiotic sponges are the most numerous sessile animals in the Baikals littoral zone. These endemic sponges are grouped to the family Lubomirskiidae and, based on molecular data, they are separated from the cosmopolitan family Spongillidae. The endemic sponges are monophyletic and originate from a common ancestor with the freshwater sponges Ephydatia fluviatilis / Spongilla lacustris. They are grouped to the class of Demospongiae having spicules that are composed of hydrated, amorphous, noncrystalline silica. With the Baikalian sponge Lubomirskia baikalensis it could be shown that silica is deposited around an organic filament. A major step forward to elucidate the formation of the siliceous spicules on molecular level was the finding that the axial organic filament of siliceous spicules is an enzyme, silicatein, which mediates the apposition of amorphous silica and hence the formation of spicules. The formation of siliceous spicules is certainly genetically controlled; this process initiates the morphogenesis phase and involves primarily silicatein. In the present study the existence of silicatein- genes in the fresh-water Lake Baikal sponge L. baikalensis is demonstrated. The intron-exon structure of the full-size silicatein- gene was determined. A comprehensive phylogenetic analysis with new sequences sheds light upon the evolution of the cathepsin L and silicatein families which occurred at the base of the metazoan phyla. It is concluded, that in parallel with the emergence of these enzymes at first the number of introns increased, especially in the coding region of the mature enzyme. Later in evolution the number of introns decreased again. We postulate that modification of the catalytic triad, especially of its first amino acid, is a suitable target for a chemical modulation of enzyme function of the silicateins/cathepsins L. Keywords: cathepsin, intron/exon structure, Lubomirskia baikalensis, silicatein, Suberites domuncula

Introduction Sponges in the Neo-Proterozoic Eon

Sponges were designated as living fossils (Mller 1998) because they represent the evolutionary oldest, still extant taxon which testifies the developmental level of animals living in the Neo-Proterozoic Eon (1000 to 542 million years ago [MYA]). This is important to note since two major snowball earth events (Walker 2003) occurred during this period, the Sturtian glaciation (710 to 680 MYA) and the Varanger-Marinoan ice ages (605 to 585 MYA), which very likely resulted in a continuous ice cover of the earth that supposedly caused extinction of most organisms on Earth at that time (Hoffman et al. 1998). The primordial earth surface comprised initially insoluble silicates, carbonates, and also phosphates. During the cycle of silicate weathering and carbonate precipitation, prior or

simultaneously with the glaciations, a dissolution of surface rocks composed of insoluble silicates [CaSiO3] resulted in the formation of soluble calcium carbonate [CaCO3] and soluble silica [SiO2], under consumption of atmospheric CO2 (Walker 2003). These soluble minerals leached subsequently out to the oceans, rivers and lakes and there again led to a reprecipitation of the dissolved minerals to new compositions as part of the sedimentary rocks. Such processes are dependent upon temperature, pH and atmospheric carbon dioxide; passively, the minerals are transformed diagenetically to secondary structures. In contrast to passive re-precipitation, biogenic deposition of minerals by metazoans is first seen in sponges. The oldest sponge fossils (Hexactinellida) have been described from Mongolia and were assessed to have lived coeval with the diverse Ediacara fauna of Namibia more than 540 MYA (Brasier et al. 1997). Hence, the Hexactinellida are the oldest

180

group of sponges as documented there and later in fossil records of the Sansha section in Hunan (Early Cambrian; China; Steiner et al. 1993), where more or less completely preserved sponge fossils, e.g. Solactiniella plumata (Mehl et al. 1998), have been found. The oil-shales of the Messel pit, near Darmstadt (Germany), are very rich in fossil freshwater sponges; among them is Spongilla gutenbergiana from the Middle Eocene (Lutetian), approximately 50 MYA (Mller et al. 1982). Based on sequencing data of informative genes, which code for structural and functional proteins, it had been calculated that the sponges diverged from the common metazoan ancestor approximately 650 MYA (Schcke et al. 1994). This calculation is in close accordance with fossil records and implies that the sponges evolved between the two glaciations, Sturtian and Varanger-Marinoan. The existence of a large genetic repertoire in the Porifera, the basis for the establishment of complex metabolic and morphogenetic pathways, may have contributed to the rapid evolution of sponges occurred between the snowball periods 710 to 680 MYA and 605 to 585 MYA. At present it cannot be ruled out that other animal phyla evolved simultaneously with the Porifera, but became extinct during the last ice age.

Sponges in the Quaternary Eon at Lake Baikal

An exceptional freshwater sponge fauna has evolved, as endemic taxa, in the Lake Baikal in the last few million years. Interestingly enough these species branched off from a common ancestor with cosmopolitan freshwater sponges during ice periods. Fossil sponge spicules have been described from the Pliocene, 3.1 to 2.9 MYA (Weinberg et al. 2003). Paleoclimatic records over the period of 5 million years revealed two cold episodes, each approximately 300,000 years long, at the time intervals 2.8-2.5 MYA and 1.8-1.5 MYA (Kashiwaya et al. 2000). Therefore, it can be assumed/ postulated that the radiation of the freshwater sponges started in Lake Baikal more than 3 MYA. This view is also consistent with the finding of Weinberg et al. (2003) who described fossil sponge spicules from the Academician Ridge. At present, this ridge forms a submerged elevation of 400 m, while the maximal depth of the lake is 1,473 m. During the Upper Pliocene-Eopleistocene period (3.5 MYA) the ridge might have formed a land barrier between the Northern Basin and the Central Basin of the lake (Ovchinnikova 2005). During those ice periods the average temperature in the lake dropped from 20C (Eocene; 70-35 MYA) to <10C (Upper Pliocene; 3.5 MYA). Todays average water temperature is 35C, but changes considerably; in January the temperature is 0 to 4C while during the short summer period from June to August/September it rises to 12C (Scientific Council 1993). Nevertheless the metabolic activities, the pumping rates (Savarese et al. 1997) and the feeding efficiencies (Pile et al. 1997) of the Baikalian sponges are identical to those found in species in the marine tropic environment. Baikalian species are generally larger in size than other freshwater species, they grow up to 1.5 m high. The three most abundant species in the lake are Lubomirskia baikalensis Dybowski 1880 (Fig. 1AB), followed by Baikalospongia bacillifera Dybowski, 1880 and Baikalospongia intermedia Dybowski, 1880 (Pile et al.

1997). The habitus of the Baikalian sponges is highly variable; even though the basic Bauplan remains constant (Belikov et al. 2005, Mller et al. 2006b). In depths between 2 and 4 m L. baikalensis is an irregular crust. However, below 4 m an arborescent growth pattern is seen; from the 2 cm thick crusts, arborescent branches protrude with a modular arrangement of the tissue units. At larger depths fusion of the branches starts (> 7 m). Under certain environmental conditions the shape of the branches changes from cylindrical to flattened. Related species, e.g. Baikalospongia intermedia profundalis Rezvoj, 1936 or Baikalospongia fungiformis Makuschok, 1927 are deep-water sponges (they occur between 40-140 m [B. fungiformis] and up to 540 m [B.media profundalis]); a fact which is of prime ecological importance (Efremova 2001). One most remarkable chemical component of Lake Baikal is the relatively high level of dissolved silicic acid. On average the lake water contains 30-100 mol L-1 of silicic acid (Grachev 2002). In the marine coastal areas the mean silicon concentration at the surface is less than 3 mol L-1 (Maldonado et al. 1999); while it gradually increases with depth (Trguer et al. 1995). The high content of silicic acid in Lake Baikal is caused by the heavy influx of silicon from the rivers (Bukharov and Fialkov 2001). The turnover rate of silicic acid in the lake is elevated; more than 70% of the mineral is cycled during the year through biological uptakebiomineralization-sedimentation-redissolving as well as by upwelling and diffusion back to the lake.

Monophyly of sponges in Lake Baikal

Variability of populations, adaptive changes in populations, geographic variation and speciation are all processes that are described by the term microevolution, while macroevolution refers to processes that occur above the species level (Mayr 2001). The microevolutionary events that took place during the last several hundreds to a few million years can be studied exemplarily in the populations of endemic animal species in Lake Baikal and in some surrounding lakes. According to geological records that go back to 24 million years Lake Baikal is the oldest and deepest (1,637 m) of the earths ancient lakes; its water mass contributes to one fifth of the worlds unfrozen freshwater. More than 1,500 endemic species inhabit Lake Baikal (Timoshkin 1997) with the highly diverse sponges being the dominant animals in the littoral zone (Kozhov 1972); it is assumed that there are up to 18-20 species, all belonging to the taxon Spongillidae/Spongillina (Efremova 2004). Also other ancient lakes have a similar rich endemic sponge fauna, e.g. Malawispongiidae Manconi and Pronzato, 2002, Metschnikowiidae Czerniavsky, 1880 or Potamolepidae Brien, 1967 (see: Manconi and Pronzato 2000).

Molecular phylogeny: cytochrome oxidase-tubulinsilicatein

Nucleotide sequence data obtained from protein-coding sponge genes, such as the mitochondrial cytochrome oxidase subunit I (COI) gene and the exon/intron sequences framing intron-2 of the tubulin gene evidenced the evolution of the different sponge species in Lake Baikal from a common

181

Fig. 1: Lubomirskia baikalensis. A. The first illustration of this species by Dybowski (1880). B. Bush-like growth form of L. baikalensis. Specimens from a depth of 10 m, showing the transition from the crust growth form to the arborescent pattern organization. The photograph was taken in March 2006 from animals growing under the ice cover of 1 m. They appear in bright green, due to abundance of the dinoflagellates, highly related to Gymnodinium sanguineum.

ancestor which was also the origin for the ubiquitous freshwater sponge Spongilla lacustris Linnaeus, 1758 (Spongillidae Gray, 1867; Schrder et al. 2003). The degree of nucleotide substitutions within the 18S rDNA region of the different Baikalian sponges proved not sufficiently high to draw unequivocal conclusions (Itskovich et al. 1999). Even the degree of sequence identity in the COI gene of the different Baikalian sponges required additional support from analyses of intron-2 of the tubulin genes (Schrder et al. 2003). A phylogenetic tree (Fig. 2B) could be constructed on the ground of the different lengths of the introns (Fig. 2A) and their building blocks. The sequence of intron-2 from tubulin from the marine demosponge Suberites domuncula (Olivi, 1792) (Hadromerida Topsent, 1894: Suberitidae Schmidt, 1870) served as outgroup. The cosmopolitan freshwater sponge S. lacustris formed the next similar branch from wich the Baikalian sponges, all belonging to the family Lubomirskiidae Rezvoj, 1936, branch off; first, the member of the genus Swartschewskia Makuschok, 1927, S. papyracea Dybowski, 1880, followed by the three species among the Baikalospongia Annandale, 1914, B. recta Efremova, 2001, B. intermedia Dybowski, 1880 and B. bacillifera Dybowski, 1880 (Fig. 2B) and finally the species Lubomirskia baikalensis (Pallas, 1776). This branching pattern was supported also by the COI sequence data, even though their statistical significance was low. Paleontological data also indicate that Swartschewskia-species appeared before Baikalospongiaand Lubomirskia-species in Lake Baikal (Weinberg et al. 2003). Analysis of the protein sequence of the gene encoding silicatein, the major protein in the axial filament of spicules from Demospongiae, further supported this phylogeny. The organic filament in the central canal of the spicules, composed

of a cathepsin L-related enzyme, was termed silicatein by the group of Morse (Shimizu et al. 1998). They cloned two of the proposed three isoforms of silicateins, the - and -form, from the demosponge Tethya aurantium Pallas, 1766 (Cha et al. 1999). Later silcateins were also cloned from other sponges, among them S. domuncula and L. baikalensis (Krasko et al. 2000, 2002, Schrder et al. 2004, Kaluzhnaya et al. 2005b, Wiens et al. 2006, Mller et al. 2006a). For the studies of the relationship between the endemic Baikalian sponges and the cosmopolitan freshwater sponges, S. lacustris and Ephydatia fluviatilis Lamarck, 1816, the silicatein cDNAs from these two species were isolated. Phylogenetic analysis was performed with the silicateins from the demosponges L. baikalensis, T. aurantium and S. domuncula as well as with the cathepsins L from Metazoa and the papain cysteine peptidase from the plant species Arabidopsis thaliana. The phylogenetic tree showed that the cathepsin L sequences form the basic branch from which the silicatein sequences diverge (see: Wiens et al. 2006) indicating that the silicateins have a common ancestor with the cathepsin L sequences from the marine hexactinellids such as Aphrocallistes vastus Schulze, 1899 and demosponges, here from S. domuncula. Among the silicateins those from the cosmopolitan species S. lacustris and E. fluviatilis form the basal branch for the polypeptides of the Baikalian sponges L. baikalensis (Fig. 3; [Mller et al. 2006c]). From the results obtained by three different molecular biological approaches (mitochondrial genes [Schrder et al. 2003]; introns of tubulin [Fig. 2; Schrder et al. 2003]; silicatein genes [Fig. 3 and review: Mller et al. 2006c]) it may be concluded that the endemic sponges in Lake Baikal evolved monophyletically from the cosmopolitan freshwater sponges. Studies are in progress to investigate the silicatein

182

Fig. 2: Phylogenetic relationship of the sequences of intron 2 in tubulin genes of freshwater sponges and the marine sponge S. domuncula. A. Intron/exon organization of the human tubulin beta 2 gene (accession number X02344); coding region of the tubulin gene interspersed with (at least) three introns. The intron numbers, their sizes (in nucleotides) as well as their positions within the human tubulin are indicated; the numbers refer to the corresponding cDNA (accession number NM_006088; length 1338 nts). B. Phylogenetic tree (slanted cladogram), constructed after alignment of the sequences for intron-2 from the tubulin genes of the freshwater sponges [FW-sponge] S. lacustris (SLAC$TUI2), L. baikalensis (LBAI$TUI2), B. intermedia (BINT$TUI2), B. recta (BREC$TUI2), B. bacillifera (BBAC$TUI2), S. papyracea (SPAP$TUI2) and the corresponding sequence from the marine demosponge S. domuncula (SDOM$TUI2), which was used as outgroup. The analysis was performed by neighbour-joining. The numbers at the nodes indicate the level of confidence (in percent) for the branches after 1000 bootstrap replicates. Modified after Schrder et al. (2003a) and Mller et al. (2006c).

surfaces of the branches and have diameters of 3-4 mm. The spicules of L. baikalensis are slightly curved amphioxea which are covered by many spines (Fig. 4A and B); they measure 150-210 m in length and 8-15 m in diameter (Dybowski 1880, Masuda et al. 1997). The spicules of Lubomirskia are other than in the related genus Baikalospongia embedded into a horny sheet (Annandale 1914). This organic support allows a regular construction of the bundles within the sponge tissue. X-ray analysis revealed a highly ordered arrangement of the 150 m to 220 m long spicules within the body into longitudinal bundles. These structures are to some extent interrupted by transversal lines which demarcate 1 cm strong growth annuli (Kaluzhnaya et al. 2005a). Longitudinal cross-sections through the skeleton of L. baikalensis show the pronounced apical-basal organization of the branches into approximately 1 cm long modules (Wiens et al. 2006) determined by the arrangement of spicule bundles. The ascending, longitudinal bundles of one module originate from the annulus, a concavely curved demarcation zone between modules which is formed by a dense fusiform arrangement of the bundles. Annuli are composed of a ramified network of longitudinal and traverse bundles which appear very bright in cross-sections (Fig. 5A and B). Within the branches, the spicules are embedded in an organic matrix which fixes them in the ordered skeleton (Fig. 5C). At the uppermost tips of the branches the spicules protrude freely into the environment (Fig. 5D).

Enzymatic activities of L. baikalensis axial filaments

According to earlier studies (Cha et al. 1999, Krasko et al. 2000) the axial filament is composed of silicatein which causes a polymerization of silica from monomeric forms, like from tetraethylorthosilicate (TEOS). The filaments from L. baikalensis were incubated with TEOS (Belikov et al. 2005, Mller et al. 2006d); after a reaction period of 180 min the filaments were washed and analyzed by SEM. The images revealed that the smooth surfaces from the rhomboid to cylindrical axial filaments (Fig. 4C-a) found immediately after isolation from the spicules, changed after incubation with TEOS and were decorated with silica lumps (Fig. 4Cb). In a second approach to demonstrate that the axial filaments catalyze the formation of biosilica from monomeric TEOS, the filament samples were stained with Rhodamine 123 and inspected by fluorescence microscopy. At time zero, only the filaments are stained (Fig. 4D), while after 30 min (Fig. 4E) or after 180 min the filaments are surrounded with silica depositions which are intensively stained (Fig. 4F). It is known (see: Sumerel and Morse 2003, Mller et al. 2003) that the silicatein proteins share high sequence similarity with the cathepsins L. Therefore, we investigated by an in situ detection assay for proteinases whether silicatein displays in addition to its ability to synthesize polymeric biosilica also proteolytic activity (see: Mller et al. 2006d). At the beginning of the incubation period with a synthetic substrate no staining is seen, while after an incubation for 30 min a distinct color reaction proceeds which is first seen at the surfaces of the filaments and later more diffusely also

sequences from other endemic sponge species of Lake Baikal, especially those living in the deep water, below 30 m.

Morphological characteristics of L. baikalensis

The Lubomirskiidae are one major family of endemic sponges in Lake Baikal (Masuda et al. 1997). The species L. baikalensis is found on every hard bottom down to at least 15 m (Efremova 2001); as reported by Savarese et al. (1997) it is one of the most abundant species in the littoral-zone. The initial description of this species came from Pallas (1776), with the precise description by Dybowski (1880) (Fig. 1A). In some locations L. baikalensis grows to more than 120 cm high specimens with dichotomous branches between 1 and 3 cm in diameter (Fig. 1B). The oscules are located at the lateral

183

Fig. 3: Phylogenetic relationship of the silicateins, the enzyme which catalyzes the polymerization process of biosilica in the sponge spicules. Four deduced silicatein sequences of the isoform silicatein- [-1, -2, -3 and -4] from L. baikalensis (SILICAa1_LUBAI, accession number AJ872183; SILICAa2_LUBAI, AJ968945; SILICAa3_LUBAI, AJ968946; SILICAa4_LUBAI, AJ968947) and the two cathepsin L sequences (CATL1_LUBAI, AJ96849; CATL2_LUBAI, AJ968951) were aligned with silicatein- from S. domuncula (SILICAa_SUBDO; CAC03737.1) from Tethya aurantia [T. aurantium] (SILICAa_TETYA, AAD23951) and with the -isoenzymes from S. domuncula (SILICAb_SUBDO, AJ547635.1) and T. aurantia (SILICAb_TETYA, AF098670), as well as with the cathepsin L sequences from sponges S. domuncula (CATL_SUBDO, AJ784224), G. cydonium (CATL_GEOCY, Y10527) and Aphrocallistes vastus (CATL_APHRVAS, AJ968951); and the related papain-like cysteine peptidase XBCP3 from Arabidopsis thaliana (PAPAIN_ARATH, AAK71314) [outgroup]. Additionally the deduced silicateins from the cosmopolitan freshwater sponges E. fluviatilis (SILCA1_EPHYDAT and SILCA2_EPHYDAT) and S. lacustris (SILCA_SPONGILLA) are included in this analysis. The numbers at the nodes are an indication of the level of confidence for the branches as determined by bootstrap analysis [1000 bootstrap replicates]. The slanted phylogenetic tree was constructed after the alignment of these sequences and had been rooted with the plant enzyme (papain from A. thaliana) as an outgroup.

as patches around them. Prolonged incubation for 120 min results in a strong increase of the color reaction. In parallel experiments it was established that the axial filaments of L. baikalensis show no proteolytic activity if the substrate for elastase was applied.

Intron/exon structures of sponge silicatein and cathepsin L genes

The cDNAs from the sponge silicateins and cathepsins L were used to construct the primers to clone the complete genes. Interestingly the intron/exon structures of the genes for these two groups of proteins are very similar among the sponges and characteristically different from other metazoan proteins. The two genes from L. baikalensis for silicatein and cathepsin L are shown here (Fig. 6A); the other new genes are in the data base.

nts (accession number AJ872183). Six introns which are delimited by characteristic donor splice sites and acceptor splice sites are located between nt274 to nt344 (intron-1; phase 0), nt498 to nt820 (intron-2; 0), nt936 to nt1313 (intron-3; 1), nt1424 to nt1501 (intron-4; 0), nt1655 to nt1728 (intron-5; 0) and nt1848 to nt1930 (intron-6; 2).

The cathepsin L gene from L. baikalensis

The 2,149 nts long sequence, which encompasses the ORF of LBCATL (accession number AJ872184) comprises the following introns: between nt82 to nt298 (intron-1; phase 0), nt443 to nt619 (intron-2; 0), nt773 to nt988 (intron-3; 0), nt1104 to nt1187 (intron-4; 1), nt1298 to nt1553 (intron-5; 0), nt1710 to nt1830 (intron-6; 0) and nt1950 to nt2043 (intron-7; 2).

The silicatein- gene from L. baikalensis

The full-length sequence comprising the ORF (nt43 to nt2027) of the corresponding cDNA, LBSILICA, has 2,040

The silicatein- gene from S. domuncula

The ORF is located from nt63 to nt2271 within the SDSILICAa gene. The six introns are found between nt303 to nt510 (intron1; phase 0), nt667 to nt832 (intron-2; 0), nt948 to nt1085 (intron-3;

184

Fig. 4: Spicules and axial filament/silicatein from L. baikalensis. A. SEM micrograph of amphioxea. B. Broken spicules show the axial canal (ac); SEM analysis. C. Axial filaments which are composed of silicatein (SEM). One axial filament is shown immediately after isolation from a spicule (C-a), or after incubation with TEOS for 180 min (C-b). The time-dependent formation of biosilica from TEOS and mediated by silicatein is shown after staining with Rhodamine 123 (D - F). The silica production was analyzed with the dye Rhodamine 123 at time zero (D), or after the incubation period of 15 min (E) or 180 min with (F) with TEOS. The reaction products are visualized by immunofluorescence microscopy; they are deposited around the axial filaments.

1), nt1196 to nt1368 (intron-4; 0), nt1522 to nt1896 (intron-5; 0) and nt2016 to nt2171 (intron-6; 2).

The silicatein- gene from S. domuncula

The 1,740 long gene includes the complete ORF which is located from nt63 to nt1725 within the SDSILICAb gene. The five introns are located as follows: between nt390 to nt469 (intron-1; phase 0), nt683 to nt737 (intron-2; 0), nt853 to nt907 (intron-3; 1), nt1033 to nt1289 (intron-4; 0) and nt1562 to nt1628 (intron-5; 2).

The cathepsin L gene from S. domuncula

The intron/exon borders of cathepsin L exist in the mature enzyme from S. domuncula between aa130 and aa131 (phase 0), within aa165 (phase 1), between aa201 and aa202 (phase 0), aa252 and aa253 (phase 0) and within aa292 (phase 2).

Conservation of the intron/exon borders of sponge silicatein and cathepsin L genes

The comparison of the intron/exon borders of sponge silicateins with those of cathepsins L from sponges and other Metazoa and with the plant cysteine protease from A. thaliana shows a characteristic distribution/conservation within the

mature deduced enzymes. While in the sponge enzymes, both in the silicateins- and -, five introns are found, the human, insect and nematode cathepsin L genes have less introns. In Fig. 6 the introns within the mature enzyme region in all sequences are indicated consecutively by small letters; e.g. intron-a corresponds to intron-2 in the LBSILICA gene, or to intron-3 in LBCATL gene. It is obvious that intron-a in all sponge silicatein and cathepsin genes is of phase 0, intronb of phase 1, intron-c of phase 0, intron-d of phase 0 and intron-e of phase 2 and they are all located at the same sites within the deduced coding segments (Fig. 6A). In the human cathepsin L gene intron-b is missing, while all other introns exist at the same site within the gene. In D. melanogaster none of the introns exists within the gene region encoding the mature cathepsin L. The only existing unusual intron (Le Boulay et al. 1998) in this region of the cathepsin L gene does not match with any other intron discussed here (Fig. 6A and B). Interestingly enough intron-a, intron-c and intron-d exist at the same sites in sponges as the corresponding introns in the human sequence. This comparison underlines once more that the intron/exon borders and even their phases are highly conserved among the Metazoa. The number of introns is lower in the protostomians D. melanogaster and C. elegans in comparison to sponges and human (Fig. 6B). Moreover

185

Fig. 5: Microscopical analysis of the skeleton of L. baikalensis. SEM images through a longitudinal section of a branch. A. The lower magnification shows the two modules (mo) which are separated from each other by an annulus (an). (B to D) At higher magnification it is seen that the longitudinal bundles (lo) are fortified by traverse bundles (tr). The top of one branch is marked (to). Size bars are given. Modified according to Wiens et al. (2006).

the presence of the comparatively high number of the introns in the plant sequence underscores the phylogenetically conserved relationship of the cysteine proteases.

Consequence: Joint Russian-German Laboratory

Furthermore, the lesson of history has been learnt and transformed to a sustainable collaboration between scientists of two nations by the foundation of the Joint RussianGerman Laboratory for Biology of Sponges in Irkutsk. Through the interlinking of the expertise of the two groups headed by S.I. Belikov, O.V. Kaluzhnaya [Irkutsk] and W.E.G. Mller, H.C. Schrder [Mainz] under the umbrella of academician M.A. Grachev a tradition, which had been started by German scientists e.g. Gmelin, Pallas and von Humboldt, is continued and will contribute to a sustainable social and commercial progress in an under-populated area, which is so extremely rich in natural resources. Focusing on the scientific biotechnological issue, biosilica will perhaps even dominate the present day high economical trade factor, e.g. oil/gas or metals. The treasure of Lake Baikal is larger, it is a unique place where (i) evolution in action can be studied, (ii) a unique and conserved climate weather situation exists, which will provide us with early warning markers for the potential present day global warming process, (iii) solid methane, a powerful greenhouse gas that is also a valuable fuel to generate mechanical and electrical energy. (iv) More biotechnological innovations will emerge, e.g. those coming from other biosilica forming organisms, e.g. diatoms (Popovskaya et al. 2002). Needless to say, that the Joint

General conclusion
In summary, sponge silicatein, here with the example L. baikalensis silicatein, catalyzes biosilica formation from monomeric silicon alkoxides. The in situ analyses showed an impressive deposition of biosilica. Even though the catalytic triad of the enzyme is changed from cysteine to serine (in the first amino acid of the triad) the molecule exhibits also proteolytic activity towards a substrate which is specific for cathepsin L. This biochemical result is supported by cDNA analysis and the elucidation of the intron/exon structure of the genes, showing the high sequence similarity between these two groups of enzymes. The data also show that in parallel with the emergence of the silicateins the number of introns increased in Porifera, the oldest phylum which branched off from the common metazoan ancestor, the Urmetazoa. This study will also contribute to the development of new strategies to chemically modify the active sites of the silicateins/cathepsins in the direction to change their enzymic properties.

186

Fig. 6: A. Comparison of the intron/exon borders (}{In-1 to 5) of the silicatein genes from the sponges L. baikalensis (silicatein- from L. baikalensis (SILCAa_LUBAI, truncated from aa 1-110; accession number AJ877018) and S. domuncula (silicatein- (SILCAa_SUBDO, aa1-114; CAC03737.1 and silicatein- SILCAb_SUBDO, aa 1-162), with the related cathepsin L genes from L. baikalensis (CATL_LUBAI, aa 1-108; AJ877019) and S. domuncula (CATL_SUBDO, aa 1-108) and also from human (CATL_HUMAN, aa 1-114; HGNC:2537MIM:116880), D. melanogaster (CATL_DROME, aa 1-124; CG6692-FBgn0013770) and C. elegans (CATL_CAEEL, aa 1-120; 2B354K02E7.10). From the protein sequences shown, the propeptides have been truncated. The characteristic sites of the catalytic triad and, the serine cluster and the cleavage site of the signal peptide are marked. The sites of the introns are marked and numbered consecutively with small letters (In-a to In-e). B. For the slanted phylogenetic tree the above sequences were aligned with the cysteine protease from A. thaliana (CyP_ARATH; BAB08269.1). The sponge cathepsins (CATL) and the silicateins (SILCA) are in gray. If present, the introns with the respective numbers within the range of the mature polypeptides are given.

187

Laboratory is open for the scientific community and surely will soon attract also the commercially directed institutions. At present the focus of the Laboratory centers around the topic sponge biosilica. The proof on concept of biosilica for the application in nano-biotechnology applicable in biomedicine has been documented. The rich presence of silicatein genes in Lake Baikal endemic sponges, with L. baikalensis as a model, has been recognized, their genes analyzed and the product of this key molecule [silicatein] has been secured. Biosilica can be synthesized by recombinant technologies, and the material will have pronounced impact in biomedicine, since it is biocompatible, biodegradable and mechanically stabile. To close with von Humboldt (1844): analyze the richness of Siberia to a global view and exploit those for the technological development and hence for the prosperity of the population.

Acknowledgements
This work was supported by grants from the European Commission, the Deutsche Forschungsgemeinschaft, the Bundesministerium fr Bildung und Forschung Germany [project: Center of Excellence BIOTECmarin], WTZ Germany - Russia (German-Russian cooperation through the BMBF [Founding of the Joint RussianGerman Laboratory for Biology of Sponges, Irkutsk]) and the International Human Frontier Science Program, as well as by a grant from the Presidium of the Russian Academy of Science (no. 25.5) and from RFBR (no. 03-04-4985).

References
Annandale N (1913) Notes on some sponges from Lake Baikal in the collection of the Imperial Academy of Science [St. Petersburg]. Ann Mus Zool Acad Sci St. Petersburg 18:18-101 Belikov SI, Kaluzhnaya OV, Schrder HC, Krasko A, Mller IM, Mller WEG (2005) Metabolism of spicules from the Baikalian sponge Lubomirskia baicalensis. Cell Biol Intern 29: 943-951 Brasier M, Green O, Shields G (1997) Ediacarian sponge spicule clusters from southwest Mongolia and the origins of the Cambrian fauna. Geology 25: 303-306 Bukharov AA, Fialkov VA (2001) Baikal in numbers. Irkutsk: Baikal Museum Cha JN, Shimizu K, Zhou Y, Christianssen SC, Chmelka BF, Stucky GD, Morse DE (1999) Silicatein filaments and subunits from a marine sponge direct the polymerization of silica and silicones in vitro. Proc Natl Acad Sci USA 96: 361-365 Dybowski W (1880) Studien ber die Spongien des russischen Reiches mit besonderer Bercksichtigung der Spongien-Fauna des Baikal-Sees. Mm Acad Sci St Petersburg sr 7, 27:1-71 Efremova SM (2001) Porifera. Index of animal species inhabiting Lake Baikal and its catchment area. Novosibirsk: Nauka, - Vol. 1: Lake Baikal, Book 1; pp. 182-192 Efremova SM (2004) New genus and new species of sponges from family Lubomirskiidae Rezvoy, 1936. Index of Animal species inhabiting Lake Baikal and its catchment area. - Novosibirsk: Nauka, - Vol. 1: Lake Baikal, Book 2; p.p.1261-1278 Grachev MA (2002) On the current state of Baikal ecosystem. Novosibirsk: Publishing House SD RAS Hoffman PF, Kaufman AJ, Halverson GP, Schrag DP (1998) A Neoproterozoic snowball earth. Science 281: 1342-1346

Itskovich VB, Belikov SI, Efremova SM, Masuda Y (1999) Phylogenetic relationships between Lubomirskiidae, Spongillidae and some marine sponges according partial sequences of 18S rDNA. Memoir Queensl Mus 44: 275-280 Kaluzhnaya OV, Belikov SI, Schrder HC, Rothenberger M, Zapf S, Kaandorp JA, Borejko A, Mller IM, Mller WEG (2005a) Dynamics of skeletal formation in the Lake Baikal sponge Lubomirskia baicalensis. Part I: Biological and biochemical studies. Naturwissenschaften 92 128133 Kaluzhnaya OV, Belikov SI, Schrder HC, Rothenberger M, Zapf S, Kaandorp JA, Borejko A, Mller IM, Mller WEG (2005b) Dynamics of skeletal formation in the Lake Baikal sponge Lubomirskia baicalensis. Part II: Molecular biological studies. Naturwissenschaften 92: 134-138 Kashiwaya K, Tanaka A, Sakai H, Kawai T (2000) Paleoclimatic signals printed in Lake Baikal sediments. In: Minoura K (ed). Lake Baikal. Elsevier, Amsterdam. pp 53-70 Kozhov MM (1972) Essays on Baikalology. East-Siberian Publishers, Irkutsk (In Russian) Krasko A, Batel R, Schrder HC, Mller IM, Mller WEG (2000) Expression of silicatein and collagen genes in the marine sponge Suberites domuncula is controlled by silicate and myotrophin. Europ J Biochem 267: 4878-4887 Krasko A, Schrder HC, Batel R, Grebenjuk VA, Steffen R, Mller IM, Mller WEG (2002) Iron induces proliferation and morphogenesis in primmorphs from the marine sponge Suberites domuncula. DNA Cell Biol 21: 67-80 Le Boulay C, Sellos D, Van Wormhoudt A (1998) Cathepsin L gene organization in crustaceans. Gene 218: 77-84 Maldonado M, Carmona MC, Uriz MJ, Cruzado A (1999) Decline in Mesozoic reef-building sponges explained by silicon limitations. Nature 401: 785-788 Manconi R, Pronzato R (2000) Suborder Spongillina subord. nov.: freshwater sponges. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges. Kluwer Academic/Plenum Publishers, New York. pp 921-1019 Masuda Y, Itskovich VB, Veinberg EV, Efremova SM (1997) Studies on the taxonomy and distribution of freshwater sponges in the Lake Baikal. In: Miyzaki N (ed). Animal community, environment and phylogeny in Lake Baikal. Otsuchi Marine Center/Ocean Research Institute, Tokyo. pp 21-41 Mayr E (2001) What evolution is. Basic Books, Tokyo Mehl D, Mller I, Mller WEG (1998) Molecular biological and palaeontological evidence that Eumetazoa, including Porifera (sponges), are of monophyletic origin. In: Watanabe Y, Fusetani N (eds). Sponge Science - Multidisciplinary Perspectives. SpringerVerlag, Tokyo. pp 133-156 Mller WEG, Zahn RK, Maidhof A (1982) Spongilla gutenbergiana n.sp., ein Swasserschwamm aus dem Mittel-Eozn von Messel. Senckenbergiana lethaea 63: 465-472 Mller WEG (1998) Origin of Metazoa: sponges as living fossils. Naturwissenschaften 85: 11-25 Mller WEG, Krasko A, Le Pennec G, Steffen R, Ammar MSA, Wiens M, Mller IM, Schrder HC (2003) Molecular mechanism of spicule formation in the demosponge Suberites domuncula: silicatein - collagen - myotrophin. Progr Molec Subcell Biol 33: 195-221 Mller WEG, Belikov S I, Tremel W, Perry CC, Gieskes WWC, Boreiko A, Schrder HC (2006a) Siliceous spicules in marine

188

demosponges (example Suberites domuncula). Micron 37: 107120 Mller WEG, Kaluzhnaya OV, Belikov SI, Rothenberger M, Schrder HC. Reiber A, Kaandorp JA, Manz B, Mietchen D, Volke F (2006b) Magnetic resonance imaging of the siliceous skeleton of the demosponge Lubomirskia baicalensis. J Struct Biol 153: 3141 Mller WEG, Schrder HC, Wrede P, Kaluzhnaya OV, Belikov SI (2006c) Speciation of sponges in Baikal-Tuva region (an outline). J Zool Syst Evol Res 44: 105-117 Mller WEG, Gamulin V, Kaluzhnaya OV, Schrder HC, Grebenjuk VA, Borejko A, Mller IM, Krasko A, Belikov SI (2006d) Silicateins, the major biosilica forming enzymes present in demosponges: protein analysis and phylogenetic relationship. Gene 395:62-71. Ovchinnikova NS (ed). (2005) Atlas of Lake Baikal: past-presentfuture. Omskaya kartographicheskaya Fabr, Omsk Pallas PS (1771-1776) Reise durch die verschiedenen Provinzen des russischen Reiches. Theil 3-Buch 2. Kayserliche Academie der Wissenschaften, St. Petersburg Pile AJ, Patterson MR, Savarese M, Chernykh VI, Fialkov VA (1997) Trophic effects of sponge feeding within Lake Baikals littoral zone. 2. Sponge abundance, diet, feeding efficiency, and carbon flux. Limnol Oceanogr 42: 178-184 Popovskaya GI, Genkal SI, Likhoshway YV (2002) Diatoms of the plankton of Lake Baikal. Nauka, Novosibirsk Savarese M, Patterson MR, Chernykh VI, Fialkov VA (1997) Trophic effects of sponge feeding within Lake Baikals littoral zone. 1. In situ pumping rate. Limnol Oceanogr 42: 171-178 Schcke H, Mller IM, Mller WEG (1994) Tyrosine kinase from the marine sponge Geodia cydonium: the oldest member belonging to the receptor tyrosine kinase class II family. In: Mller WEG (ed). Use of aquatic invertebrates as tools for monitoring of environmental hazards. Gustav Fischer Verlag, Stuttgart, New York. pp 201-211 Schrder HC, Efremova SM, Itskovich V B, Belikov S, Masuda Y, Krasko A, Mller IM, Mller WEG (2003) Molecular phylogeny of

the freshwater sponges in Lake Baikal. J Zool Syst Evol Research 41: 80-86 Schrder HC, Perovi-Ottstadt S, Wiens M, Batel R, Mller IM, Mller WEG (2004) Differentiation capacity of the epithelial cells in the sponge Suberites domuncula. Cell Tissue Res 316: 271-280 Schulze FE (1904) Hexactinellida. Wissenschaftliche Ergebnisse der Deutschen Tiefsee-Expedition; vol 4. Fischer-Verlag, Jena Scientific Council (1993) Atlas of Baikal. Federal Agency of Geodesy and Maps of Russia, Moscow Shimizu K, Cha J, Stucky GD, Morse DE (1998) Silicatein alpha: cathepsin L-like protein in sponge biosilica. Proc Natl Acad Sci USA 95: 6234-6238 Steiner M, Mehl D, Reitner J, Erdtmann BD (1993) Oldest entirely preserved sponges and other fossils from the Lowermost Cambrian and a new facies reconstruction of the Yangtze Platform (China). Berl Geowiss Abhand (E) 9: 293-329 Sumerel JL, Morse D (2003) Biotechnological advances in biosilicification. Progr Molec Subcell Biol 33: 225-247 Timoshkin OA (1997) Biodiversity of Baikal fauna: state-of-theart (preliminary analysis). In: Wada E, Timoshkin OA, Fujita N, Tanida K (eds). New scope on boreal ecosystems in east Siberia. DIWPA series vol. 2; Siberian Branch of the Russian Academy of Science, Novosibirsk. pp 35-76 Trguer P, Nelson DM, van Bennekom AJ, DeMaster DJ, Leynaert A, Quguiner B (1995) The silica balance in the world ocean: a reestimate. Science 268: 375-379 Walker G (2003) Snowball Earth. Bloomsbury, London Weinberg E, Weinberg I, Efremova S, Tanichev A, Masuda Y (2003) Late Pliocene spongial fauna in Lake Baikal (from material from the deep drilling core BDP-96-1). In: Kashiwaya K (ed). Long continental records from Lake Baikal. Springer Verlag, Tokyo. pp 283-293 Wiens M, Belikov SI, Kaluzhnaya OV, Krasko A, Schrder HC, Perovic-Ottstadt S, Mller WEG (2006) Molecular control of serial module formation along the apical-basal axis in the sponge Lubomirskia baicalensis: silicateins, mannose-binding lectin and mago nashi. Dev Genes Evol 216: 229-242

Porifera research: Biodiversity, innovation and sustainaBility - 2007

189

Sponges from a submarine canyon of the Argentine Sea

Marco Bertolino(1), Laura Schejter(2), Barbara Calcinai(3*), Carlo Cerrano(1), Claudia Bremec(2)
Dipartimento per lo studio del Territorio e delle sue Risorse, C.so Europa, 26, 16132, Genova, Italy. marco.bertolino75@libero.it, cerrano@dipteris.unige.it (2) Laboratorio de Bentos, Instituto Nacional de Investigacin y Desarrollo Pesquero, Paseo Victoria Ocampo 1, (B7602HSA) Mar del Plata, Argentina schejter@inidep.edu.ar, cbremec@inidep.edu.ar (3) Dipartimento di Scienze del Mare, Via Brecce Bianche, 60131, Ancona, Italy. b.calcinai@univpm.it
(1)

Abstract: During a research cruise to assess the abundance and distribution of Patagonian scallop Zygochlamys patagonica (King and Broderip, 1832), a submarine canyon was discovered using a multibeam SIMRAD EM1002 sonar. The canyon is positioned at 4335S and 5933W, close to the southern commercial scallop beds in the Argentine Sea. The existence of submarine canyons on the continental shelf of Argentina was already known, but their edaphic and biotic conditions remain unstudied. A sample of the benthic community was collected at the head of the canyon at 360 m depth. In total nine species of demosponges were identified; two of them represent new records for the Argentine Sea and two are new to science. Keywords: Argentina, new records, new species, Porifera, submarine canyon

Introduction
The existence of submarine canyons on the continental shelf of Argentina was known (Parker et al. 1997), but their exact location, number, edaphic and biotic conditions remain unstudied. During a research to assess the distribution and abundance of the Patagonian scallop Zygochlamys patagonica (King and Broderip, 1832), a submarine canyon was discovered. The Patagonian scallop is distributed in the Magellanic Biogeographical Province and its exploitation results in the destruction of associated species. Porifera are particularly affected (Schejter et al. 2006). The Argentinian sponge fauna is relatively well known; with the earliest studies commencing in the 19th century (Ridley and Dendy 1886, 1887, Sollas 1886, 1888, Schulze 1887). More recent contributions include Boury-Esnault (1973), Sar (1978), Mothes de Moraes and Pauls (1979), and particularly Cuartas (1986, 1991, 1992a, 1992b, 1992c, 1995). Despite our knowledge of the Argentinian sponge fauna, a recent paper on the Porifera associated with commercial Patagonian scallop beds reported four new records for the area (Schejter et al. 2006). Here we describe 9 species of demosponges recorded in a canyon close to scallop beds. In particular Pseudosuberites cf. antarcticus Carter, 1876 and Guitarra dendyi (Kirkpatrick, 1907) were previously known only for the Antarctic area, representing new records for the Argentine Sea, while two (Stelodoryx argentinae sp. nov. and Tedania (Tedaniopsis) sarai sp. nov.) are new to science.

Materials and methods

The canyon was discovered using a multibeam SIMRAD EM1002 sonar, during a research cruise (R/V Oca Balda, INIDEP, April 2005) for assessment of the Patagonian scallop (Zygochlamys patagonica). The canyon is positioned at 4335S and 5933W, close to the southern commercial scallop beds, in the Argentine Sea (Fig. 1). Sampling was carried out on board the R/V Oca Balda (INIDEP, National Fisheries Research and Development Institute) during April 2005, at 360 m depth. The samples were labelled OB-4-05 with the indication of the ship (Oca Balda) and the date of collection, and with progressive numbers (e.g. OB-4-05 caon 14). Specimens were frozen upon collection and fixed in 5% formaldehyde in sea water and then preserved in alcohol 70%, or dried, in the laboratory. For the study of spicules, small fragments of sponge tissue were heat-dissolved in nitric acid, rinsed in water, and dehydrated in ethanol; then spicules were mounted on microscope slides. Spicule dimensions are given as range of lengths and of widths and as average (in brackets); they were obtained measuring 20 to 30 spicules per category. Sections were cut by hand, perpendicularly and tangentially to the sponge surface, using a razor blade. For SEM analyses spicule dissociations were transferred onto stubs and sputter-coated with gold. SEM studies were carried out on a Philips XL 20 scanning electron microscope. Photographs of the specimens were taken using a Nikon Coolpix 4500.

190

Class Demospongiae Sollas, 1885 Order Spirophorida Bergquist and Hogg, 1969 Family Tetillidae Sollas, 1886 Genus Craniella Schmidt, 1870 Craniella leptoderma (Sollas, 1886) Examined material: OB-4-05: caon 14, caon 15 Sponge round or elongate, egg-shaped (Fig. 2A, B). Surface hispid, conulose. Several small oscules are located on the tip of the sponge conules (Fig. 2A). In caon 15 the surface is more regular (without conules) and more hispid (Fig. 2B). The consistency is hard. The colour is dirty white, or brown-pinkish in alcohol. Skeleton: Spicule tracts radiate from the centre of the sponge towards the surface (Fig. 2C). Spicules: Megascleres are large oxeas, fusiform; they are frequently broken. Small oxeas (Fig. 2D), straight or slightly curved, frequently centrotylote. They measure 540 (913.6) 1500 x 5 (25) 30 m. Anatriaenes 1, with thin clads (Fig. 2E). The rhabdomes are frequently broken; the clads measure 100 (150) 200 x 15 (18) 20 m. Anatriaenes 2, with thick and short clads 70 (105) 120 x 20 (32.5) 40 m (Fig. 2F). Rhabdomes are frequently broken. Protriaenes 1, with rhabdomes that reach more than 8 cm in length, while the clads are 40 (140.5) 200 m long (Fig. 2G). Protriaenes 2 are filiform, up to 1 cm long (Fig. 2H) and with a clad longer than the others; the long clads are 26 (32.5) 44 m long and the short ones 10 m long. Microscleres are spinispires (Fig. 2I); they measure 5 (7.8) 10 m. Distribution: Antarctic shores, South Georgia, Straits of Magellan, Malvinas, Kerguelen and Heard Islands (Sar et al. 1992), South Shetland Islands (Ros et al. 2004), Chile, Atlantic coast of South America (mouth of the Rio de la Plata) (Desqueyroux-Fandez 1989). Remarks: This species has a highly variable habitus; the body is round, elongate egg-shaped (Koltun 1964) or massive, spherical (Desqueyroux-Fandez 1989); the surface varies from even and smooth, to more or less bristly (Koltun 1964), to strongly hispid (Desqueyroux-Fandez 1989). Order Hadromerida Topsent, 1894 Family Suberitidae Schmidt, 1870 Genus Pseudosuberites Topsent, 1896 Pseudosuberites cf. antarcticus Carter, 1876 Examined material: OB-4-05: caon 6, caon 10 Massive sponge (Fig. 3A) with a cavernous structure covered by remains of a thin membrane easily detachable from the sponge body. The consistency is soft. The colour is beige-grey in alcohol. The sample caon 6 hosted numerous samples of the bivalve Hiatella solida (Sowerby, 1834). Skeleton: The ectosomal skeleton is made of tylostyles tangentially arranged (Fig. 3B); the choanosomal skeleton is made of well defined spicule tracts running towards the surface (Fig. 3C). Spicules: Tylostyles and subtylostyles (Fig. 3D-H), often slightly curved. They have well formed heads (Fig. 3H),

Fig. 1: Location of the canyon and of the sampling station in the Atlantic Ocean, Argentine Sea, Argentina.

The present collection is preserved at the Laboratorio de Bentos, Instituto Nacional de Investigacin y Desarrollo Pesquero (INIDEP). The type materials of the new species here described are deposited at the Museo Civico di Storia Naturale G. Doria of Genoa (MSNG).

Results
In total nine species were identified. Craniella leptoderma (Sollas, 1886), Myxilla (Myxilla) mollis Ridley and Dendy, 1886, Tedania (Tedaniopsis) charcoti Topsent, 1907, Tedania (Tedaniopsis) massa Ridley and Dendy, 1886 and Tedania (Trachytedania) mucosa Thiele, 1905 are already known for the area. Pseudosuberites cf. antarcticus Carter, 1876 and Guitarra dendyi (Kirkpatrick, 1907) were previously known only for the Antarctic area and so they represent new records for the Argentine Sea. Stelodoryx argentinae sp. nov. and Tedania (Tedaniopsis) sarai sp. nov. are new for science. All species dealt with are described below.

191

Fig. 2: A, B. Specimens of Craniella leptoderma (Sollas, 1886); C. Radial skeleton; D. Small oxeas; the arrow points the central tyle; E. Anatriaene 1; F. Anatriaene 2; G. Protriaene 1; H. Protriaene 2; I. Spinispires.

sometimes sub-terminal (Fig. 3G) and acerate tips (Fig. 3E). They measure 350 (1063.6) 1350 x 5 (15.6) 25 m. Distribution: Antarctic shores, Heard and Kerguelen Islands (Koltun 1964). Remarks: This specimen is easily recognisable as Pseudosuberites antarcticus Carter, 1876 in spicule characteristics (comparable size and shape), but it differs from the holotype as well as from other records of the species, in the habitus. Pseudosuberites antarcticus was previously reported as erect and ramified (Ridley and Dendy 1887, Topsent 1902). This represents a new record for the Argentine Sea that enlarges the known distribution of this species northwards.

Order Poecilosclerida Topsent, 1894 Suborder Myxillina Hajdu, van Soest and Hooper, 1994 Family Myxillidae Dendy, 1922 Genus Myxilla Schmidt, 1862 Myxilla (Myxilla) mollis Ridley and Dendy, 1886 Examined material: OB-4-05: caon 3 Massive sponge with an irregular surface and an irregular system of cavities (Fig. 4A). No dermal membrane was present. The sponge is compressible and elastic. The colour is beige in alcohol. Skeleton: The ectosomal skeleton is absent. The choanosome is a loose reticulation of smooth styles and anisotylotes (Fig.

192

Fig. 3: A. Specimen of Pseudosuberites cf. antarcticus Carter, 1876; the arrow points the thin membrane; B. Ectosomal skeleton; C. Choanosomal skeleton; D, F, H. Heads of tylostyles; E. Tip of a tylostyle; G. Subtylostyle.

4B). Microscleres are abundantly scattered all over the choanosome. Spicules: Megascleres: Smooth, slightly curved styles (Fig. 4C). Rare subtylostyles. They measure 325 (407.5) 437.5 x 12 (14) 15 m. Anisotylotes are straight or slightly sinuous, with swollen and microspined extremities (Fig. 4D). They measure 225 (250) 275 x 6 m. Microscleres: Spatuliferous anchorate isochelas 1, with three teeth, slightly curved (Fig. 4E). They measure 24 (35.4) 40 m. Spatuliferous anchorate

isochelas 2 of similar shape (Fig. 4F). They measure about 20 m. Sigmas 1 are C- or S- shaped (Fig. 4G, H). They measure 40 (51) 60 m. Sigmas 2 are C- shaped (Fig. 4F). They measure 15 (25) 30 m. Distribution: West and East coast of South America; Malvinas Islands (Desqueyroux-Fandez and van Soest 1996); Antarctic shores; Strait of Magellan, Kerguelen Island (Sar et al. 1992).

193

Fig. 4: A. Specimen of Myxilla (Myxilla) mollis Ridley and Dendy, 1886; B. Choanosomal skeleton; C. Smooth style; D. Anisotylote; E. Spatuliferous anchorate isochela 1; F. Spatuliferous anchorate isochela 2; G. Sigma 1; H. Sigma 1 and 2 (arrow).

Remarks: This specimen has entirely smooth styles as originally described by Ridley and Dendy (1886) and two categories of sigmas and isochelas as reported by Desqueyroux-Fandez and van Soest (1996). Genus Stelodoryx Topsent, 1904 Stelodoryx argentinae sp. nov. Examined material: OB-4-05: caon 4 Holotype MSNG 54057 Comparative material: holotype of Myxilla cribrigera Ridley and Dendy 1886, Natural History Museum, London (Challenger coll. BMNH: 87.5.2.138.) The species consists of a single specimen about 6.5 cm long, massive, with an irregular surface (Fig. 5A). Some oscules, slightly elevated (0.5-1 mm), 1-2 mm in diameter are visible. The consistency is soft and elastic when alive, fragile and friable in the dried state. The colour is black in alcohol and dried. The sponge includes a large amount of sand. Skeleton: The ectosomal skeleton consists of brushes of anisostrongyles with spined ends (Fig. 5B). A thin tangential, dermal membrane is present (Fig. 5B). The choanosomal skeleton (Fig. 5C) is a paucispicular reticulum of main styles, thin styles and anchorate chelae. Spicules: Megascleres are smooth styles (Fig. 5D), straight or slightly curved with acerate or conical tips; they measure 287.5 (351) 412.5 x 10 (13) 15 m. Straight, thin styles are 188.7 (220.5) 260 x 2.6 m (Fig. 5E). Straight anisostrongyles (Fig. 5F), with finely spined extremities (Fig, 5G, H); the

diameter of the spicule decreases from an extremity to the other. They measure 209 (240) 262.5 x 5 (7.8) 10 m. Microscleres are polydentate, spatuliferous isochelae (Fig. 5I) with five teeth. They measure 40.8 (52.4) 65 m. A few unguiferous anchorate isochelae are present (Fig. 5J). Etymology: The name of this species refers to the sea of origin. Remarks: Among the 11 known species of Stelodoryx, S. cribrigera (Ridley and Dendy, 1886), reported also for the Malvinas Islands is very close to the specimen described here. The study of the holotype of S. cribrigera has shown some differences: the species of Ridley and Dendy is characterized by tylotornotes with slight swelling on the tips. Our species has anisostrongyles with different tips, never swollen. Moreover the species of Ridley and Dendy has larger styles (650 x 25 m) and lacks thin styles. In the skeleton preparation of the holotype an ectosomal layer of isochelas is evident (also reported by Desqueyroux-Fandez and van Soest (1996) from additional material), but it is not present in our species. Family Tedaniidae Ridley and Dendy, 1886 Genus Tedania Gray, 1867 Subgenus Tedaniopsis Dendy, 1924 Tedania (Tedaniopsis) charcoti Topsent, 1907 Examined material: OB-4-05: caon 1 Massive sponge with conulose surface, uneven and with numerous oscules evident. The sponge is soft and brownish in alcohol (Fig. 6A).

194

Fig. 5: A. Stelodoryx argentinae sp. nov.: Holotype; B. Ectosomal skeleton; the arrow points the thin dermal membrane; C. Choanosomal skeleton; D. Smooth style; E. Thin style; F. Straight anisostrongyle; G, H. tips of an anisostrongyle; I. Spatuliferous anchorate isochela 1; J. Unguiferous anchorate isochela 2.

Skeleton: Ectosomal skeleton consists of tangentially disposed anisotornotes which form a loose reticulum. The choanosome is an irregular and confused reticulation of longitudinal tracts of styles, and free onychaetes (Fig. 6B).

Spicules: Megascleres are smooth styles, slightly curved (Fig. 6C). They measure 412.5 (432) 462.5 x 10 m. Smooth, mucronate, straight anisotornotes (Fig. 6D). They measure 262.5 (292) 312.5 x 5 m. Microscleres (Fig. 6E): Onychaetes

195

Fig. 6: A. Specimen of Tedania (Tedaniopsis) charcoti Topsent, 1907; B. Choanosomal skeleton; C. Smooth style; D. Anisotornotes; E. Onychaetes 1 and onychaetes 2.

1 measure 200 (278.6) 462.5 x 2 m. Onychaetes 2 measure 50 (56.6) 60 m. Distribution: Antarctic shores, South Georgia, South Sandwich, South Orkney, Strait of Magellan; Kerguelen and Malvinas Islands (Sar et al. 1992). Argentine Sea (Mar del Plata), Chile (Desqueyroux and Moyano 1987, Cuartas 1992b). South Shetland Islands (Ros et al. 2004). Remarks: This is a very common species frequently collected in the area. In Desqueyroux-Fandez and van Soest (1996), this species is well described. Our material fits very well with the description provided by these authors. Tedania (Tedaniopsis) massa Ridley and Dendy, 1886 Examined material: OB-4-05: caon 11, caon 13 The specimen consists of two small fragments of a massive, lobose sponge (Fig. 7A). The surface is irregular and minutely hispid. The consistency is soft. The colour is beige.

Skeleton: In the ectosome tornotes project towards the surface in divergent brushes (Fig. 7B). The choanosome is a loose reticulum of styles and fibres of onychaetes. These run to the surface, anastomosing and connecting to secondary fibres. Spicules: Megascleres are smooth styles slightly to strongly curved (Fig. 7C). They measure 400 (440) 500 x 15 (16) 20 m. Mucronate anisotornotes are 287.5 (373) 500 x 10 m (Fig. 7D). Microscleres (Fig. 7E): Onychaetes 1 measure 437.5 (480) 660 m. Onychaetes 2 measure 45 (67.8) 85 m. Distribution: South Atlantic Ocean: from Uruguay to the Strait of Magellan, Argentina (Mar del Plata) (Mothes and Pauls 1979, Cuartas 1992b, 1992c). Antarctic shores, South Georgia, Malvinas Islands (Sar et al. 1992). Tedania (Tedaniopsis) sarai sp. nov. Examined material: OB-4-05: caon 8; Holotype MSNG 54058

196

Fig. 7: A. Fragments of Tedania (Tedaniopsis) massa Ridley and Dendy, 1886; B. Ectosomal skeleton made of tornotes in divergent brushes (arrow) and loose reticulum of styles of the choanosome; C. Smooth style; D. Mucronate anisotornotes; E. Onychaetes 1 and onychaetes 2.

Comparative material: holotype of Tedania armata Sar, 1978, Museo Civico di Storia Naturale G. Doria of Genoa (Ant. 3). Specimen of Tedania lanceta Koltun, 1964, identified by V. M. Koltun; pictures of spicules of the specimen labelled NN 6767 and NN 6751 as the type material is missing. This species is massive, cavernous with a smooth, uneven surface (Fig. 8A). The consistency is hard. The colour is brown in alcohol, dirty grey in the dried state, clearer in the interior. Skeleton: In the ectosome anisotornotes make tangential surface tracts. The choanosome is a reticulation of smooth styles (Fig. 8B). Spicules: Megascleres: Curved, often slightly flexuous styles (Fig. 8C), with frequent lanceolate or abruptly pointed tips (Fig. 8D). The heads of the styles are elongated, blunt (Fig. 8E). The thickness of the shaft is often reduced below the elongated extremity of the style. Measurements 387.6 (430.9) 469 x 10.4 (12.3) 13 m. Straight or slightly curved, anisotornotes, with lanceolate, slightly inflated extremities (Fig. 8F, G). They measure 275.4 (319.3) 375 x 5.2 (6.8) 7.8 m. Microscleres: onychaetes 1 (Fig. 8H) with short spines (Fig. 8I); they measure 387.6 (434.4) 489.6 x 2.6 m; onychaetes 2 with different extremities (Fig. 8J) and longer spines (Fig. 8K); they measure 44 (63.7) 122 x 1 m. Etymology: Named after the late Prof. Michele Sar for his contribution to the knowledge of Porifera in general,

including an important contribution towards Argentine sponge taxonomy. Remarks: This species is very close to T. (Tedaniopsis) lanceta Koltun, 1964 and to T. armata Sar, 1978. Koltuns species differs in having stouter styles (400-480 m long and 16-22 m wide) that are not flexuous and have lanceolate tips. In the description of Koltun (1964) the anisotornotes are thicker (360-400 m long and 14-16 m wide), and moreover onychates 1 are shorter (270-320 m long). The holotype is missing so any direct comparison is not possible. Thanks to Dr. Alexander Ereskovsky (St. Petersburg State University, Russia) we had the possibility to compare the spicule complement of this species with some pictures of spicules of T. lanceta Koltun, 1964, determined by Koltun. This comparison confirmed the previous observed differences. Moreover the shape of the spicules of this specimen seems different from our species: the extremities of the anisotornotes are often asymmetrical and bent. Tedania armata Sar, 1978 was synonymised with T. charcoti by Desqueyroux-Fandez and van Soest (1996), but the authors gave no reason for this decision. We suggest the synonymy between T. armata and T. charcoti should be reconsidered on the basis of the examined material and on our experience dealing with antarctic and subantarctic sponges. Tedania charcoti has in fact true styles while in T. armata these principal spicules are subtylostyles with long and acuminate tips (Fig. 9), and also the tornotes are different from those of T. charcoti that have mucronate tips (see e.g.

197

Fig. 8: Tedania (Tedaniopsis) sarai sp. nov.: Holotype; B. Choanosomal skeleton; C. Style; D. Magnification of an abruptly pointed tip of the style; E. Magnification of the blunt head of the style; F. Lanceolate anisotornotes; G. Magnification of the lanceolate tip of the anisotornotes; H. Onychaetes 1; I. Magnification of onychaetes 1; J. Onychaetes 2; K. Magnification of onychaetes 2.

Desqueyroux-Fandez and van Soest (1996), Figs. 107-110, page 57). Comparison with the holotype of T. armata (Fig. 9A) illustrates the primary differences between this species and T. sarai sp. nov.: in T. armata the principal spicules are subtylostyles with rounded heads (Fig. 9B, C, D), while the tips are similar to those of T. sarai sp. nov. (Fig. 9B). In T. armata numerous styles modified to oxeas (Fig. 9D) and

expanded just before the distal tip, rendering the spicule lancelike (lanceolated, Fig. 9B) are common. The dimensions of the styles are also different: in T. armata these spicules are shorter and thinner (300-350 x 6-8 m, Sar (1978); 308-372 x 8 m, Desqueyroux-Fandez and van Soest (1996)). Tornotes have rounded, and mucronate tips in T. armata (Fig. 9E-H), while in our species they have lanceolate tips. The dimensions of

198

Fig. 9: A. Tedania armata. Sar, 1978: Holotype (Ant 3); B. Subtylostyle; C. Heads of subtylostyles; D. Style modified in oxea; E, F. Tornotes with rounded tips; G. Magnification of a tip of a tornote; H. Anisotornote with mucronate and rounded tips; I. Onychaete.

Fig. 10: A. Specimen of Tedania (Trachytedania) mucosa Thiele, 1905; B. Loose reticulation of tracts of styles and abundant onychaetes of the choanosomal skeleton; C. Smooth style; D. Mucronate tornote; E. Onychaetes 1 and onychaetes 2.

199

the tornotes in T. armata are also smaller. Onychaetes 1 in T. sarai sp. nov. are longer: 150-180 m (Fig. 9I). Subgenus Trachytedania Ridley, 1881 Tedania (Trachytedania) mucosa Thiele, 1905 Examined material: OB-4-05: caon 5 The sponge is massive, irregularly elongate. Surface uneven with scattered oscules (Fig. 10A). The consistency is hard, the colour light brown. Skeleton: The ectosome is a perpendicular palisade of densely arranged mucronate tornotes. The choanosomal skeleton is a loose reticulation of tracts of styles and abundant onychaetes (Fig. 10B). Spicules: Megascleres are smooth, slightly curved styles (Fig. 10C). They measure 250 (266.5) 287.5 x 12 (13) 15 m. Mucronate tornotes (Fig. 9D) measure 200 (215) 245 x 6 m. Microscleres (Fig. 10E): Onychaetes 1 measure 135 (186.5) 210 m; onychaetes 2 measure 40.8 (54) 71 m. Distribution: Malvinas Islands, Chilean coast (Desqueyroux 1972). Argentina, Mar del Plata (Cuartas 1992b, DesqueyrouxFandez and van Soest 1996); Strait of Magellan (Sar et al. 1992). Remarks: Two kinds of onychaetes are present as reported by Desqueyroux-Fandez and van Soest (1996). Suborder Mycalina Hajdu, van Soest and Hooper, 1994 Family Guitarridae Dendy, 1924 Genus Guitarra Carter, 1874 Guitarra dendyi (Kirkpatrick, 1907) Examined material: OB-4-05: caon 2 Massive, cushion-shaped sponge with uneven surface (Fig. 11A). The consistency is soft and in alcohol the colour is brick red. Skeleton: In the ectosome the skeleton is made of exotyles arranged in bouquets with apices pointing toward the surface of the body and scattered sigmas. In the choanosome strongyles are irregularly arranged. Spicules: Megascleres are rare exotyles with a spherical, wrinkled apex and a rounded base (Fig. 11B). They measure 210 (300) 400 x 10 (15) 20 m; head diameter measures 50 (65) 80 m. Straight anisostrongyles (Fig. 11C). They measure 375 (454.5) 512.5 x 6 (8) 9 m. Microscleres: placochelae (Fig. 11C-F); they measure 75 (84.5) 90 m. C-shaped sigma. They measure 10 (12) 15 m. Distribution: Antarctic shores and South Shetland Islands (Ros et al. 2004). Remarks: This is the first record of Guitarra dendyi for the Argentine Sea. The distribution of this species was limited to the Antarctic shores (Wilhem II Coast, Banzare Coast, Victoria Land) (Ros et al. 2004) and to the South Shetland

Islands, therefore its geographical range is considerably extended northwards.

Discussion
Among the nine species collected, only one (Tedania charcoti) was previously found associated with Patagonian scallop beds (Schejter et al. 2006). The present findings, including two species new for science, suggest the importance to continue the study of these deep areas, still widely unexplored. Our results extend the geographical range northwards for Pseudosuberites cf. antarcticus and Guitarra dendyi, for which species these are the first records outside the Antarctic sea. The lack of data on deep-water faunas around Antarctica has been recently highlighted by Brandt et al. (2007). These Authors evidenced how the abyssal Antarctic fauna has strong links with others oceans, mainly Atlantic, but only when taxa are good dispersers. For example, isopods, ostracods and nematodes include many species known exclusively in Antarctica differently from the foraminifera. In this way our results suggest once again the importance to study these environments to better evaluate also the relative importance of dispersal by larvae or by floating propagules (as suggested by Burton 1932) to account for a possible relationship between sponge distribution and oceanic current systems. Although not as deep as the community described by Brandt et al. (2007), there are just few studies of the Argentinian Sea waters of subantarctic origin in the continental shelf, the shelf break and submarine canyons. The Argentinian side of the Magellanic Biogeographic Province is influenced by the Malvinas Current, a relatively fresh and cold branch of the Circumpolar Current, strongly flowing northward along the continental shelf of Argentina (Garzoli 1993, Piola and Rivas 1997, Vivier and Provost 1999). However, based on evidence from faunal analysis done in the study area and also considering the shelters and dead shells found by Bremec et al. (2006), it is possible that important fluxes from the shelves to deeper oceanic waters were performed throughout submarine canyons. Fishing effort can produce several ecological consequences, for instance changes in species richness and biodiversity, loss of erect and fragile epifauna, widely damaging epibiotic community (Turner et al. 1999, Coleman and Williams 2002, Thrush and Dayton 2002). Sponges are frequently collected in the invertebrate by-catch of the Patagonian scallop fishery in the neighbouring shelf of the study area and represented approximately 510% of total community biomass (Bremec et al. 1998, 2000, Bremec and Lasta 2002, Schejter et al. 2006). Porifera biomass at Patagonian scallop beds in the Argentine Sea decreased between 1995 and 1998 in exploited areas (Bremec et al. 2000). Between 1998 and 2001, the sponge contribution in these areas represented an average of 0.3 kg / 100 m2 (wet weight) (Schejter 2004). It will be interesting to know if a long term consequence of the fishing activity will

200

Fig. 11: A. Specimen of Guitarra dendyi (Kirkpatrick, 1907); B. Light microscopy observation of an exotyle; C. Anisostrongyle and placochela (arrow); D, E, F. Placochelae.

be, owing to sponge fragmentation caused by trawling, the increase in the distribution of some sponge species in respect to other species on intermediately deep Argentinian bottoms.

Ereskovsky (St. Petersburg State University, Russia) for sending us V.M. Koltuns material for comparison.

References
Boury-Esnault N (1973) Campagne de la Calypso au large des ctes atlantiques de lAmrique du Sud (1961-1962). 29. Spongiaires. Rs Sci Camp Calypso 10: 263-295 Brandt A, Gooday AJ, Brando SN, Brix S, Brkeland W, Cedhagen T, Choudhury M, Cornelius N, Danis B, De Mesel I, Diaz RJ,

Acknowledgments
We thank A. Madirolas and G. Alvarez Colombo (Hydroacoustics Lab., I.N.I.D.E.P. Argentina) for the information and images of the Canyon and Maurizio Pansini (Dip. Te. Ris, Genoa University) for his helpful suggestions. The authors are indebted to Dr. Alexander

201

Gillan DC, Ebbe B, Howe JA, Janussen D, Kaiser S, Linse K, Malyutina M, Pawlowski J, Raupach M, Vanreusel A (2007) First insights into the biodiversity and biogeography of the Southern Ocean deep sea. Nature 447: 307-311 Bremec CS, Lasta ML (2002) Epibenthic assemblage associated with scallop (Zygochlamys patagonica) beds in the Argentinian shelf. Bull Mar Sci 70(1): 89-105 Bremec C, Lasta M, Lucifora L, Valero J (1998) Anlisis de la captura incidental asociada a la pesquera de vieira patagnica (Zygochlamys patagonica (King and Broderip 1832)). Inf Tcn INIDEP 22: 1-18 Bremec C, Brey T, Lasta M, Valero J, Lucifora L (2000) Zygochlamys patagonica beds on the Argentinian shelf: Part I: Energy flow through the scallop bed community. Arch Fish Mar Res 48(3): 295-303 Bremec C, Schejter L, Madirolas A, Tripode M (2006) Comunidades de aguas profundas: macrofauna bentnica de un can submarino localizado en la plataforma patagnica (4335S, 5933W). VI Jornadas Nacionales de Ciencias del Mar, Puerto Madryn, Argentina. Book of abstracts. p. 128 Burton M (1932) Sponges. Discov Rep 6: 237-392 Carter HJ (1876) Descriptions and figures of deep-sea sponge and their spicules from the Atlantic ocean, dredged up on board HMS Porcupina Chiefly in 1869 (concluded). Ann Mag Nat Hist, ser. 4, 20: 38-42 Coleman FC, Williams SL (2002) Overexploiting marine ecosystem engineers: potencial consequences for biodiversity. Trends Ecol Evol 17: 40-44 Cuartas EI (1986) Porferos de la Provincia Biogeogrfica Argentina. Physis A 44 (106): 37-41 Cuartas EI (1991) Demospongiae (Porifera) de Mar del Plata (Argentina) con la descripcin de Cliona lisa sp. n. y Plicatellopsis reptans sp. n. Neritica 6(1-2): 43-63 Cuartas EI (1992a) Porferos intermareales de San Antonio Oeste, provincia de Ro Negro, Argentina (Porifera: Demospongiae). Neotropica 38(100): 111-118 Cuartas EI (1992b) Porferos de la Provincia Biogeogrfica Argentina. III. Poecilosclerida (Demospongiae) del litoral marplatense. Physis A 47 (113): 73-88 Cuartas EI (1992c) Algunas Demospongiae (Porifera) de Mar del Plata, Argentina, con descripcin de Axociella marplatensis sp. n. Iheringia Sr Zool 73: 3-12 Cuartas EI (1995) Redescripcin de Clathria burtoni nomen novum de C. prolifera Burton, 1940. (Porifera: Demospongiae). Ann Mus Civ St Nat G. Doria XC: 571-576 Desqueyroux R (1972) Demospongiae (Porifera) de la costa de Chile. Gayana Zool 20: 1-71 Desqueyroux-Fandez R (1989) Demospongiae (Porfera) del litoral chileno antrtico. Ser Cient Inach 39: 97-158 Desqueyroux-Fandez R, Moyano HI (1987) Zoogeografa de Demospongias chilenas. Bol Soc Biol Concepcin (Chile) 58: 3966 Desqueyroux-Fandez R, van Soest RWM (1996) A review of Iophonidae, Myxillidae and Tedaniidae occurring in the south east Pacific (Porifera: Poecilosclerida). Rev Suisse Zool 103: 3-79 Garzoli SL (1993) Geostrophic velocity and transport variability in the Brazil-Malvinas confluence. Deep Sea Res 40: 1379-1403

Koltun VM (1964) Sponges of the Antarctic. I. Tetraxonida and Cornacuspongida. In: Pavlovskii EP, Andriyashev AP, Ushakov PV (eds). Biol Rep Soviet Antarct Exped (1955-1958), Akademya Nauk SSSR [English translation, 1966, Israel Program for Scientific Translations]. Vol. 2 (10). S. Monson, Jerusalem. pp. 6133, 443-448 Mothes-de-Moraes B, Pauls SM (1979) Algunas esponjas monaxonidas (Porifera: Demospongiae) do litoral sul do Brasil, Uruguay e Argentina. Iheringia Sr Zool 54: 57-66 Parker G, Paterlini MC, Violante RA (1997) El fondo marino. In: Boschi E (ed). El mar argentino y sus recursos pesqueros, I. antecedentes histricos de las exploraciones en el mar y las caractersticas ambientales. pp. 65-87 Piola AR, Rivas AL (1991) Corrientes en la plataforma continental. Mar Arg Rec Pesq 1: 119-132 Ridley S, Dendy A (1886) Preliminary report on the Monaxonida collected by H.M.S. Challenger. Part I and II. Ann Mag Nat Hist 18(5): 325-351, 470-493 Ridley S, Dendy A (1887) Report on the monaxonida collected by H.M.S. Challenger during the year 1873-1876. Rep Sci Res Voy H.M.S. Challenger, Zool 20: 1-275 Ros P, Cristobo FJ, Urgorri V (2004) Poecilosclerida (Porifera, Demospongiae) collected by the Spanish Antarctic expedition BENTART-94. Cah Biol Mar 45: 97-119 Sar M (1978) Demospongie di acque superficiali della Terra del Fuoco (Spedizioni A.M.F.Mares-G.R.S.T.S. e S.A.I). Boll Mus I Biol Univ Genova 46: 7-117 Sar M, Balduzzi A, Barbieri M, Bavestrello G, Burlando B (1992) Biogeographic traits and checklist of Antarctic demosponges. Polar Biol 12: 559-585 Schejter L (2004) Estructura comunitaria en un banco de vieira patagnica Zygochlamys patagonica (Mollusca: Bivalvia: Pectinidae) sujeto a arrastres pesqueros, Mar Argentino. MSc Thesis. Universidad Internacional de Andaluca, Baeza Schejter L, Calcinai B, Cerrano C, Bertolino M, Pansini M, Giberto D, Bremec C (2006) Porifera from the Argentina Sea: diversity in Patagonian scallop beds. Ital J Zool 73 (4): 373-385 Schulze FE (1887) Report on the Hexactinellida collected by H.M.S. Challenger during the years 1873-76. Rep Sci Res Voy H.M.S. Challenger, Zool 21: 1-513 Sollas WJ (1886) Preliminary account of the Tetractinellid sponges dredged by H.M.S. Challenger 1872-76. Part I. The Choristida. Scient Proc R Dubl Soc 5: 177-199 Sollas WJ (1888) Report on the Tetractinellida collected by H.M.S. Challenger during the years 1873-76. Rep Sci Res Voy H.M.S. Challenger, Zool 25: 1-458 Thrush SF, Dayton PK (2002) Disturbance to marine benthic habitats by trawling and dredging: implications for marine biodiversity. Ann Rev Ecol Syst 33: 449-473 Topsent E (1902) Spongiaires. Expd. antarct. belge. Rs Voy S Y Belgica, 1897-1899: 1-54 Turner SJ, Thrush ST, Hewitt JE, Cumming UJ, Funnell G (1999) Fishing impacts and the degradation or loss of habitat structure. Fish Manag Ecol 6: 401-420 Vivier F, Provost C (1999) Direct velocity measurements in the Malvinas Current. J Geophys Res 104: 21083-21103

Porifera research: Biodiversity, innovation and sustainaBility - 2007

203

Excavating rates and boring pattern of Cliona albimarginata (Porifera: Clionaidae) in different substrata
Barbara Calcinai(1), Francesca Azzini(1), Giorgio Bavestrello(1), Laura Gaggero(2), Carlo Cerrano(2*)
Dipartimento di Scienze del Mare, Universit Politecnica delle Marche, Via Brecce Bianche I-60131 Ancona, Italy. b.calcinai@univpm.it, cianfry77@hotmail.com, g.bavestrello@univpm.it (2) Dipartimento per lo studio del Territorio e delle sue Risorse, Universit di Genova, Corso Europa 26 I-16100 Genova, Italy. gaggero@dipteris.unige.it, cerrano@dipteris.unige.it
(1)

Abstract: Eroding sponges create a series of connected chambers and galleries into calcareous substrata where they live. While it is well known that only calcium carbonate is etched by sponge activity, no comparative data are available regarding the different forms of carbonate. In this work we investigate the erosion rates and erosion pattern of the tropical boring sponge Cliona albimarginata in different biogenic and non-biogenic calcareous rocks. In particular, we tested portions of the shell of the large bivalve Hippopus sp. and of the branches of the stony coral Acropora sp. together with different kinds of carbonatic stones such as the Carrara marble, the Majolica of the Conero Promontory, the Finale medium-grained calcarenite, the Prun fine-to medium-grained limestone and the homogeneously fine-grained Vicenza limestone. The dissolution rates of the sponge on the different kinds of carbonate are highly variable and these differences are discussed in terms of crystal shape and aggregation, the rock fabric and the presence of other minerals. Keywords: boring pattern, Cliona albimarginata, excavating rates, Indonesia, Porifera

Introduction
Excavating sponges are able to live in carbonatic substrata, perforated by mechanical and chemical activity of specialised etching cells (Rtzler and Rieger 1973, Pomponi 1980). The contact of sponge canal system with water is ensured by portions of the sponge, called papillae, protruding from its substratum surface. This growing form known as -stage is, in some species, substituted by a complete removal of the substratum such that the sponge becomes a free-living organism (-stage). In other cases the epilithic portion continues to develop until the papillae are connected by a more or less thick crust of sponge (-stage). The activity of boring sponges has been documented especially in tropical waters, and several authors have shown their ecological role in coral reefs such as their influence on the balance of calcium carbonate stored in the reef (Hutchings 1986), the production of fine sediments (Ftterer 1974), their influence on reef morphology (Goreau and Hartman 1963), and their influence on coral asexual reproduction (Tunnicliffe 1981). Sponges are able to excavate into both inorganic and biogenic calcareous materials. In some biogenic structures like mollusc shells, organic components, such as conchiolin, are digested by acid phosphatase released by the etching cell (Pomponi 1980) while the periostracum layer may prevent sponge erosion (Mao Che et al. 1996, Kaehler and McQuaid 1999). In the same way living tissue protects corals from

sponge erosion as very few species are able to attack living tissue (Tunnicliffe 1979, 1981, MacKenna 1997, Schnberg and Wilkinson 2001, Lpez-Victoria et al. 2003, LpezVictoria and Zea 2005). Very few comparative data are available regarding the differences in boring sponge activity in the substrate with different textures. Neumann (1966) suggested that mineralogy of the carbonates does not affect the process while the density of the substrate improves penetration (Highsmith 1982, Rose and Risk 1985, Schnberg 2002). Other physical and biological factors may also affect erosion, such as nutrient availability, temperature, symbiosis with zooxanthellae (Hill 1996), etc. In this study we investigated the influence of the substrata with similar carbonatic composition but different microtexture and mineralogy composition on the erosion pattern and rate of the coral reef species Cliona albimarginata Calcinai et al. 2005. In particular we tested: i) metamorphic rocks with medium and regular grain and isorientation of calcitic grains; ii) sedimentary rocks with fine and homogeneous grain; iii) sedimentary rocks with fine and irregular grains; iv) sedimentary rocks with medium and irregular grain and a detrital component.

Materials and methods

Our experiment was carried out in the Bunaken Marine Park (North Sulawesi, Indonesia). Carbonatic blocks of

204

different mineral composition, texture and porosity were fixed, using wires and nails, on a single large specimen of C. albimarginata (about 15 m2) living 5 m deep on the edge of the coral reef (Fig. 1A, B). In this way, avoiding sponge grafts commonly used in this kind of experiment (Neumann 1966, Rtzler 1975, Schnberg and Wilkinson 2001), we limited the stress to the sponge due to the handling procedure. As both light and current enhance boring activity in sponges with zooxanthellae (Hill 1996) we chose a single specimen with the same light exposure and uniform current. The following materials were tested: 15 blocks of coral coming from branches of dead colonies of Acropora sp; 15 blocks obtained from the umbo portion of large shells of Hippopus sp.; 15 blocks of Carrara marble; 4 blocks of Finale fine-medium grained calcarenite; 5 blocks of Conero fine grained mudstone (Majolica); 4 blocks of homogeneously fine grained Vicenza limestone; 4 blocks of Prun fine-to medium-grained limestone. As control, for each kind of material 5 blocks were placed out of the sponge to evaluate the excavation due to microborers. Before the test, the blocks were dried and weighted. Each block surface was calculated using program Image-J v1.37. After 200 days the blocks were removed from the sponge, cleaned in hydrogen peroxide (120 vol.), dried and weighed again. Encrusting organisms present on the blocks were manually removed. Sponge erosion rates (Kg/m2/y) were calculated as the difference in weight of the blocks, before and after the experiment, as microerosion due to microboring organisms was negligible. To study the penetration of sponges together with rock microtexture some blocks were consolidated by epoxy resin, then processed to 30 m thick section. The mineralogical and petrographic analysis of the biogenic or inorganic substrata was carried out by stereoscopic and transmitted light microscopy. The mineral composition of substrata was analysed by Xray diffraction (XRPD) using a Philips PW1140-X-CHANGE

diffractometer (CuK radiation; current 30 mA, voltage 40 kV, scan speed, 0.5 2/min; scan interval, 3-70 2) and interfaced with PC-APD software for data acquisition and processing.

Results
The boring rates of Cliona albimarginata into the different kinds of carbonates were highly variable (Fig. 2): 29.5 2.2 Kg/m2 /y for the Carrara marble; 24.2 2.3 Kg/m2/y for the Finale calcarenite; 24.0 2.6 Kg/m2/y for Hippopus sp. (umbo); 15.6 4.7 Kg/m2/y for the Acropora (branch); 12.6 2.9 Kg/m2/y for the Conero majolica; 11.01 1.0 Kg/m2/ y for the Vicenza limestone and 2.9 1.2 Kg/m2/y for the Prun limestone. The dissolution rate in all the blocks placed outside the sponge, considered as control, was negligible. Plotting the erosion rate vs specific gravity of the different materials it was possible to observe that for four materials (Vicenza limestone, Acropora, Hippopus, and Marble) the rate of erosion is directly related to the density of the different rocks while for the Finale calcarenite, the Majolica and the Prun limestone no obvious relationships could be observed (Fig. 3). The maximum vertical penetration was also variable in different substrata, with the highest penetration in the biogenic substrata and in the marble, while the limestones were harder to penetrate and the Prun limestone was attacked only on the surface (Fig. 4). The boring pattern produced by C. albimarginata was different in the tested substrata (Fig. 5). In the marble the sponge produced vertically elongated excavations, oval in section and tidily organised (Fig. 5A). In the compact Hippopus umbo, the sponge produced similar boring patterns in every direction (Fig. 5B). In the porous Acropora the sponge largely penetrated the pre-existing canals, producing a fine spongious pattern of erosion (Fig. 5C). In the Majolica the boring activity produced circular tunnels running in an

Fig. 1: Images of the field experiment. A. General view of the experimental set with Carrara marble. Scale bar: 5 cm B. A detail of experimental set with Acropora sp. Scale bar: 2 cm.

205

Fig. 2: Dissolution rates of Cliona albimarginata in the tested substrata.

Fig. 3: Dissolution rate vs. specific gravity of the tested substrata.

indefinite directions (Fig. 5D). The Prun limestone was only slightly etched with superficial erosion marks (Fig. 5E). The Finale calcarenite was heavily affected by the sponge action that had detached the large biogenic elements (Fig. 5F). Also at the level of rock microtexture, the activity of etching cells produced different results in the tested materials. Hippopus shell has crystals organized in layers that are deformed as kinks. The erosion pattern developed in the laminated layers of the shell produced rounded excavations

0.3 -1 mm that fused into each other to form lobated cavities (Fig. 6A, B). Acropora is made of large rare aragonitic crystals (more than 0.2 mm in diameter). From this compact structure, sheaves with a centrifuge growth pattern of calcite fibres (about 0.002-4 mm long) had originated. The growth of coral was visible because of the parallel arrangement of fibrous radiating structure. The erosion pattern was affected by the isoorientation of the crystals. In fact the sponge produced

206

Fig. 4: The maximum vertical penetration of Cliona albimarginata in the tested substrata.

elongated excavations (0.15-1.5 mm) that followed the orientation of the fibrous crystals (Fig. 6C, D). In the Carrara marble (Fig. 6E, F) the average crystal size is 0.2 mm and the grains are regularly arranged in a mosaic texture. The sponge produced excavations that were regularly spaced following the regular orientation of the crystals. Under a transmitted polarised light, the boundary of the cavity was clearly thinned, and the sites of more pervasive etching exhibited micro-lobes along the external portion of the calcite grain. In fact, excavations can merge but the regular pattern of corrosion is maintained. The Prun mudstone displays an inhomogeneous texture characterized by fine-grained calcite with patches of coarser sparry calcite and scarce detrital fraction made by chlorite. The rock showed submillimetric fractures filled by clay minerals and hydroxides and clusters of forams occurring as biointraclasts. The erosion pattern was evident only where the sponge met an embedded shell that the sponge reached through a fracture filled by coarse grained spatic calcite (Fig. 6G, H). The Vicenza limestone is a biodetritic rock with the grain size varying irregularly between 0.004 and 0.1 mm. It contains forams, bivalve shells and sea urchin fragments, together with iron oxides and hydroxides. The primary porosity was distributed irregularly. The erosion pattern of the sponge was very irregular (Fig. 6I, J). The Finale calcarenite is composed of fine-mediumgrained detrital limestone, rich in detrital fossiliferous content and terrigenous minerals (quartz, lithic fragments) in sparry cement. As a consequence the texture within recrystallized areas showed irregular granulometry. Calcite grains present in the cement were between 0.01 and 0.2 mm of diameter. The sponge avoided the cement. The erosion pattern showed merging lobate excavations that led to irregular and wide galleries (Fig. 6K, L).

The Conero majolica mudstone is fine-grained (< 0.01 mm) and pervasively recrystallized. It includes nannofossils and scarce impurities of detrital quartz. The calcite crystals are fine but coarser than those found in the Prun mudstone. The erosion pattern produced lobate cavities with stochastic directions (Fig. 6M, N).

Discussion
Even if Cliona albimarginata exclusively perforates corals in the field, it is able to bore a wide variety of both mineral and biogenic substrates. The sponge had excavated all the substrata used with different intensity. The microtexture of rock substrata affects the microscopic pattern of erosion. Observation by optical techniques (transmitted light microscopy) reveals that the erosion pattern of sponge erosion may be affected by the mineral setting (i.e. rock fabric) of the substrate. In fact, the sponge produces excavations that follow the preferred orientation of fibrous calcite crystals (Acropora), or the parallel lamination within the Hippopus sp. shell. Conversely, the anisotropic granoblastic and mosaic texture of Carrara marble turns out to be etched with preference along the flow schistosity. Within a fine-grained, virtually isotropic material such as limestones (Conero Majolica or in Vicenza limestone), the general behaviour of the sponge is to etch the rock with irregular, randomly oriented, erosion patterns. The boring pattern is also affected at macroscopic scale by the characteristic of the substrata such as crystal preferred orientation (e.g. marble), pre-existing cavities (e.g. Acropora) or the presence of silicatic fragments (e.g. Prun limestone). For example in the compositionally homogeneous, massive, flow oriented, medium grained marble, the sponge produces vertically elongated excavations, oval in section and regularly organised. This suggests that dissolution is driven by crystal organisation. In the Acropora skeleton the sponge widely uses

207

Fig. 5: Macroscopic boring pattern of Cliona albimarginata in the tested substrata. A. Vertical, tidy organised excavations in the compact marble. B. A boring pattern, similar in both directions is produced in Hippopus sp. umbo. C. A fine spongious pattern produced by C. albimarginata that uses in Acropora the pre-existing canals of the coral. D. Tunnels, circular in section and running in indefinite direction in the Majolica. E. Superficial erosion marks are produced by C. albimarginata in the Prun stone. F. C. albimarginata detaches the large biogenic elements in the Finale stone, that is strongly eroded. Scale bars: 1 cm.

the pre-existing canals producing a fine spongious pattern around the principal tunnels (Fig. 5C). The regular pattern is lost in the Finale stone because the sponge detaches the large biogenic elements. In this way the substratum appears to be widely destroyed. In the Prun stone, chambers or tunnels are not detectable because the rock is only slightly etched with superficial erosion due to the presence of silicatic fragments in the stone that are not excavated by the sponge. Some authors (Hoeksema 1983, Bromley and DAlessandro 1984) report that the macroscopic pattern of excavation, produced by a sponge species, may be different because of its substrate dimension or its age. Other authors (Rtzler

1974, Schnberg 2000) have demonstrated that this character might be useful to differentiate among various species. Our data demonstrate how the previous cited characteristics of the substrata may affect the macroscopic erosion pattern produced by the boring sponge. In this way the use of the erosion pattern as a taxonomic tool requires some caution. Also the erosion rates are affected by the characteristic of the substrata. C. albimarginata is a highly destructive species. It may erode 300 - 400 kg per year of corals supplying a corresponding amount of fine sediments to the bottom. These values are similar to those of a few other excavating sponges (Schnberg 2002). Neumann (1966) studied the

208

Fig. 6: Polarized light microphotographs. A. Longitudinal section of perforated Hippopus sp. The texture is microcrystalline; the finegrained crystals are optically discontinuous and define a layered fabric of the shell, having kinked or sheaf textured appearance inside the laminae. The erosion proceed from the outer (large cavities) to inner layers (finer cavities) of the shell, following the shape of calcite grains. B. Longitudinal section of perforated Hippopus sp. The stratified structure is evidenced by the contrasting orientation of crystals. Inside the cavities a detrital fraction is preserved, most likely deriving from the mechanical action of the sponge. C. Acropora sp. exhibits an inner fabric characterized by fibrous calcite (about 0.1 mm long) elongated towards the top of the organism. The erosion pattern is sinuous and cavities are oblate and generally parallel to the fibre length. In the cavities, a fine grained detrital fraction, most likely deriving from the mechanical erosion, is preserved. D. Open cavity in Acropora sp.: the overall shape is elongated with sinuous with lobate boundaries. E. Carrara marble has homogeneous granoblastic mosaic texture. The cavities are rounded with apparently lobate boundaries like bites. F. Close up of the rock-hole area in Carrara marble: the calcite grain is thinned and corroded with lobate geometry. A detrital fraction is preserved in the cavity. G. The Prun fine-grained limestone is a mudstone (bioclasts < 10%). H. The erosion on the Prun limestone was restricted to coarse grained bioclasts (bivalve shell fragment), made of sparry (i.e. coarser than matrix) calcite. I. The Vicenza limestone is characterized by abundant bio- and intraclasts cemented by sparry calcite. The excavations are regularly dispersed, rounded to sub-rounded and flattened when two rounded holes merge. J. Detail of an excavation in the Vicenza limestone: the boundaries are finely lobated, with selective etching at the expence of sparry calcite. K. Microtexture of the Finale calcarenite. The rock is mostly formed of sparry cement including the terrigenous fraction. The excavation is developed in a patch of sparry calcite. L. Detail of an excavation etched in an organic fragment, made of sparry calcite. M. The Conero Majolica is made of nannofossils in a very fine-grained micrite matrix (mudstone). N. Erosion pattern in the Conero Majolica. Subrounded excavations tend to merge. Top, centre: a relic mudstone septum is almost isolated within a cavity. Scale bars A-D, G-N: 1 mm; E-F: 0.5 mm.

209

210

boring activity of Cliona lampa on substrata with a different mineralogy (calcite or aragonite) concluding that the density of the substrata is important to determine the value of the erosion rates. This is due to the fact that the sponge in a porous material first occupies the available spaces before excavating resulting in reduced erosion rates. Schnberg (2002) came to a similar conclusion for C. orientalis. Our data suggest a more complex scenario where density is only one of the factors affecting erosion rate. Here we have shown that there is a group of substrata where the erosion rate is directly related to substratum density. Nevertheless, other characteristics of the substratum are involved in this phenomenon. High values of erosion rates are obtained for the Finale calcarenite in spite of its low density due to the removal of entire large clasts of the biogenic fraction during the erosion activity. On the contrary, the presence of silicatic fragments in the substrate, that are not excavated by the sponge reduces the erosion rate in the Prun stone. Also the size of the crystals affects the erosion rate. In fact the low erosion rate showed by the Majolica in spite of its high density (Fig. 3) is quite likely related to the randomly oriented very small grains.

Acknowledgements
This manuscript is dedicated to Prof. L. Cortesogno who greatly stimulated this research. The project was partially funded by the Executive program of scientific co-operation between Italy and Indonesia (project 2004-07 STE-16).

References
Bromley RG, DAlessandro A (1984) The ichnogenus Entobia from the Miocene, Pliocene and Pleistocene of Southern Italy. Riv Ital Paleo Strat 90: 227-296 Calcinai B, Bavestrello G, Cerrano C (2005) Excavating sponge species from the Indo-Pacific Ocean. Zool Stud 44(1): 5-18 Ftterer DK (1974) Significance of the boring sponge Cliona for the origin of fine grained material of carbonate sediments. J Sediment Petrol 44: 79-84 Goreau TF Hartman WD (1963) Boring sponges as controlling factors in the formation and maintenance of coral reefs. Publ Am Assoc Advan Sci 75: 25-54 Highsmith RC (1982) Reproduction by fragmentation in corals. Mar Ecol Prog Ser 7: 207226 Hill MS (1996) Symbiotic zooxanthellae enhance boring and growth rates of the tropical sponge Anthosigmella varians forma varians. Mar Biol 125: 649-654 Hoeksema BW (1983) Excavation patterns and spiculae dimensions of the boring sponge Cliona celata from the SW Netherlands. Senckenberg Marit 15: 55-85

Hutchings PA (1986) Biological destruction of coral reefs. Coral Reefs 4: 239-252 Kaehler S, McQuaid CD (1999) Lethal and sub-lethal effects of phototrophic endoliths attacking the shell of the intertidal mussel Perna perna. Mar Biol 135: 497-503 Lpez-Victoria M, Zea S, Weil E (2003) New aspects on the biology of the encrusting excavating sponges complex Cliona caribbea - Cliona langae - Cliona aprica. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Boll Mus Ist Biol Univ Genova 68: 425-432 Lpez-Victoria M, Zea S (2005) Current trends of space occupation by encrusting excavating sponges on Colombian coral reefs. Mar Ecol 26 (1): 33-41 MacKenna SA (1997) Interactions between the boring sponge, Cliona lampa and two hermatypic corals from Bermuda. Proc 8th Int Coral Reef Symp, Balboa 2: 1369-1374 Mao Che L, Le Campion-Alsumard T, Boury-Esnault N, Payri C, Golubic S, BZac C (1996) Biodegradation of shells of the black pearl oyster, Pinctada margaritifera var. cumingii, by microborers and sponges of French Polynesia. Mar Biol 126: 509-519 Neumann AC (1966) Observations on coastal erosion in Bermuda and measurements of the boring rate of the sponge Cliona lampa. Limnol Oceanogr 11: 92-108 Pomponi SA (1980) Cytological mechanisms of calcium carbonate excavation by boring sponges. Int Rev Cytol 65: 301-319 Rose CS, Risk MJ (1985) Increase in Cliona delitrix infestation of Montastrea cavernosa heads on an organically polluted portion of the Grand Cayman. PSZN Mar Ecol 6: 345-363 Rtzler K (1974) The burrowing sponges of Bermuda. Smithsonian Contrib Zool 165: 1-32 Rtzler K (1975) The role of burrowing sponge in bioerosion. Oecologia 19: 203-216 Rtzler K, Rieger G (1973) Sponge burrowing: fine structure of Cliona lampa penetrating calcareous substrata. Mar Biol 21: 144162 Schnberg CHL (2000) Bioeroding sponges common to the Central Australian Great Barrier Reef: description of three new species, two new records, and additions to two previously described species. Senckenberg Marit 30: 161-221 Schnberg CHL (2002) Substrate effects on the bioeroding Demosponge Cliona orientalis. 1. Bioerosion rates. PSZN Mar Ecol 23: 313326 Schnberg CHL, Wilkinson CR (2001) Induced colonization of corals by a clionid bioeroding sponge. Coral Reefs 20: 6976 Tunnicliffe V (1979) The role of boring sponges in coral fracture. In: Lvi C, Boury-Esnault N (eds). Biologie des spongiaires. Coll Int CNRS 291: 309-315 Tunnicliffe V (1981) Breakage and propagation of the stony coral Acropora cervicornis. Proc Nat Acad Sci USA 78(4): 2427-2431

Porifera research: Biodiversity, innovation and sustainaBility - 2007

211

The possible role of Echinometra lucunter (Echinodermata: Echinoidea) in the local distribution of Darwinella sp. (Porifera: Dendroceratida) in Arraial do Cabo, Rio de Janeiro State, Brazil
Emiliano Nicolas Calderon(1*), Carla Zilberberg(2), Paulo Csar de Paiva(1)
Universidade Federal do Rio de Janeiro, Instituto de Biologia, Departamento de Zoologia, Laboratrio de Polychaeta. CCS. Bloco A, Ilha do Fundo, s/n, 21940-590, Rio de Janeiro, RJ, Brazil. encalderon@yahoo.com.br, ppoliqueta@hotmail.com (2) Departamento de Biologia Molecular e Gentica, Instituto de Biologia Roberto Alcntara Gomes, Universidade do Estado do Rio de Janeiro. Rua So Francisco Xavier 524 PHLC, sala 205. 20550-013. Rio de Janeiro (RJ), Brazil. carlazilber@ yahoo.com.br
(1)

Abstract: Sea urchins are known to control algal populations, leading to an increase in substrate accessibility for many organisms, thus having a fundamental role in the maintenance of diversity in those habitats. Darwinella sp. is a common sponge at several sites in Arraial do Cabo, Rio de Janeiro State, Brazil, but factors determining its distributional patterns are still unknown. The goal of this study was to establish the relationship among Echinometra lucunter density, algal and Darwinella sp. cover. The study was conducted at Saco do Cherne (SC) and Porcos Island (PI), Arraial do Cabo. At both sites, the density of E. lucunter and the percent cover of Darwinella sp. and algae were quantified. Additionally, Darwinella sp. percent cover was compared between portions of high and low E. lucunters density. Spearman Rank correlation analyses were performed between E. lucunter density and algae cover, E. lucunter density and Darwinella sp. cover, and between Darwinella sp. and algae cover. Results show that areas with high urchin density support a significantly higher cover of Darwinella sp. compared to areas with low density of urchins. Correlation analyses demonstrate that, in general, algae percent cover was negatively related to the density of E. lucunter. The same pattern was found between algae and Darwinella sp. percent covers. On the other hand, the relationship between E. lucunter density and Darwinella sp. cover was weakly positive. The results from this study suggest that algae is possibly competing with Darwinella sp. for space, but the presence of E. lucunter is probably mediating this interaction by decreasing the percent cover of algae. However, the weak relationship observed between E. lucunters density and Darwinella sp. cover suggests that other factors besides competition with algae are affecting the percent cover and distribution of the sponge Darwinella sp. at the studied sites. Keywords: Algae, herbivory, sea urchin, sponge, competition

Introduction
Rocky marine substrates are characterized by the presence of clonal sessile invertebrates, such as sponges, corals and ascidians (Jackson 1977, Buss 1979, McKinney 1992). In these habitats, competition for space and predation act as major biotic determinants of sessile invertebrates distribution (Connell 1961, 1973, Sutherland 1976, Jackson 1977, Buss 1980). Competition for space is particularly strong in taxa that grow horizontally along the substrate, since the fitness of these organisms is often related to their size (Jackson 1977, Buss 1990). In shallow hard substratum communities, competition with algae seems to negatively affect sessile invertebrate distribution by settlement, growth inhibition or overgrowth (Sammarco 1980, Sammarco 1982, Jernakoff 1985, Jompa and McCook 2002). Algae, for example, can inhibit invertebrate larvae

settlement by chemical or physical means (Sammarco 1980, Jernakoff 1985, Jompa and McCook 2002). Additionally, algae have greater advantage over invertebrates by their extremely fast growth rates at shallower depths (Lobban and Harison 1994), which allow them to overgrow most benthic invertebrates (McCook et al. 2001). This negative effect has been demonstrated, for example, by Jompa and McCook (2002), where the presence and consequent overgrowth of the red algae Lobophora variegata (Lamouroux) Womersley ex. Oliveira 1977 caused significant colony tissue mortality on the coral Porites cylindrica Dana, 1846. Herbivores, such as sea urchins and many fishes, are known to control algal populations leading to an increase in substrate accessibility for many organisms (Paine 1966, Jackson and Winston 1982, Ferreira et al. 1998, Tuya et al. 2004). Therefore, the presence of herbivores has a fundamental role

212

in the maintenance of diversity in marine hard substratum communities (Paine and Vadas 1969, Carpenter 1986, Jompa and McCook 2002). One remarkable example of a herbivore impact in a marine community was the mass mortality of the sea urchin Diadema antillarum Philippi, 1845 around the Caribbean region in 1983/1984 (Carpenter 1990). The decrease in 95-99% of D. antillarum population in Caribbean reefs had a large negative impact on scleractinian coral cover, changing entire reef systems from coral to algal dominated communities (Hughes 1994). Only with the slow increase in D. antillarium densities algal cover began to decrease and coral cover consequently increase (Edmunds and Carpenter 2001). Therefore, it is clear that the presence of sea urchins have a positive effect on marine hard substratum communities by the reduction of algal cover and, consequently, a greater diversity of invertebrate fauna. Sponges are often competitively superior to most benthic invertebrates (Jackson and Winston 1982, Bell and Barnes 2003). However, algae might be competitively superior to sponges as demonstrated in a study of competition between a coralline algae (Corallina vancouveriensis Yendo, 1901) and the sponge Halichondria panicea (Pallas, 1766) in temperate seas (Palumbi 1985). If sponges are inferior competitors to algae, their distribution in rock bottom communities may be negatively affected by algal competition, particularly in the absence of an effective herbivore (Bell and Barnes 2003). The negative association between sponges and macroalgae in temperate rocky subtidal communities has been suggested to be caused by competition between these two groups of organisms (Witman and Sebens 1990, Bell 2002). Conversely, it has been argued that the observed macroalgae and sponge distribution in those habitats might be, instead, the result of abiotic factors, such as depth and substratum inclination (Preciado and Maldonado 2005). What remains unknown is whether in habitats with similar abiotic factors competition with algae has a negative effect on sponge distribution. Wulff (2005) has argued that in the absence of abiotic factors, that could prevent a species from occurring in a particular habitat, biotic factors, such as competition, can have a large effect on sponge distribution. Consequently, if sponges are outcompeted by algae, the presence of an efficient herbivore is important to mediate these interactions and avoid the competitive exclusion of sponges in a particular habitat (Palumbi 1985). Darwinella sp. is a demosponge that fifteen years ago was rarely seen around the Arraial do Cabo region (Rio de Janeiro State, Brazil) (Muricy et al. 1991). Currently though, Darwinella sp. can be commonly seen at some localities around Arraial do Cabo (ENC, personal observation). Although frequent, Darwinella sp. distribution in a scale of meters to kilometers seems patchy at shallower depths (0-8 m; ENC, personal observation), however, the factors affecting the distribution of this species remains unknown. In Arraial do Cabo, the sea urchin Echinometra lucunter (Linnaeus, 1758) is the most conspicuous benthic herbivore in shallow waters (Castro et al. 1995), although herbivore fishes in the area are also common (Ferreira et al. 1998). It has been shown that 98% of E. lucunters diet is constituted by algae, while the remaining 2% is composed of invertebrates that are probably ingested accidentally (Oliveira 1991). Therefore, one possible

biotic factor determining the distribution of Darwinella sp. around Arraial do Cabo might be the relative abundance of algae, which should be regulated by E. lucunter. The goal of the present study was to evaluate the relationship among E. lucunter density, algal and Darwinella sp. cover in Arraial do Cabo, Rio de Janeiro State, Brazil.

Materials and methods Study area

The study was conducted at Porcos Island (PI) and Saco do Cherne (SC), in Arraial do Cabo, RJ, Brazil (Fig. 1). PI is characterized by low wave action, sheltered from the northeastern winds that are predominant in the area. The substratum morphology is characterized by rocky walls with variable inclination interspaced by small portions of vertical walls. SC is a small inlet, characterized by high wave action, exposed to the predominant northeastern winds. Substratum morphology is characterized by a single rocky wall predominantly vertical. On both sites PI and SC, the interface between the rocky wall and the bottom happens abruptly in approximately 90o angle.

Sampling
Eight 5 m portions of the rocky substrate were haphazardly chosen at PI and four at SC, where every portion was located at the interface between the bottom and the vertical wall (4-6 m depth), where the urchin E. lucunter was most abundant. Within each portion, five to six vertical rectangles (60 x 40

Fig. 1: Map showing the study area in Arraial do Cabo, Rio de Janeiro state, Brazil. The two studied localities were at Porcos Island (PI) and Saco do Cherne (SC).

213

Data analyses
Percent covers were arcsin transformed prior to all statistical analyses (Sokal and Rohlf 1995). Nonparametric tests were performed when data did not conform to assumptions of normality and homoscedasticity (Sokal and Rohlf 1995). A Mann-Whitney test compared the percent cover of Darwinella sp. between areas of high and low densities of E. lucunter. To choose areas with significant differences in sea urchin densities (i.e., high and low densities), at PI, out of the eight sampled portions of the rocky substratum, two with the highest and two with the lowest average densities of E. lucunter were chosen to be used in the comparison of Darwinellas percent cover (Fig. 3). A one way ANOVA followed by a Tukey post hoc test, established whether there were significant differences in sea urchins density within and between high and low density portions. No significant differences in urchin densities were found within low (F = 2.87; df = 3, 19; P > 0.5) or high (F = 0.000; df = 3, 19; P > 0.5) urchin density portions, while portions of high density were significantly different form portions of low urchin density (F = 14.20; df = 3, 19; P < 0.005). Therefore, the two portions of low density, as well as the two of high density were pooled for the comparison of the percent cover of Darwinella sp. To establish the relationship among E. lucunter densities, algal and Darwinella sp. cover, three pair-wise Spearman Rank correlation analyses were performed: 1) E. lucunter density and Darwinellas cover; 2) E. lucunter density and algal cover; 3) algal cover and Darwinellas cover. All three correlations were performed separately for each locality and afterwards by pooling data from both PI and SC. In all tests, the portions along the rocky bottom were pooled at each locality and the photographed rectangles were used as replicates.

Fig. 2: Schematic representation of the sampling area at PI showing the disposition of the photographed rectangles (Rt) on the substratum.

cm) were randomly chosen and photographed (Fig. 2) using a digital camera (SONY Cyber-Shot DSC-P41). To estimate the percentage cover of Darwinella sp., algae, and other sessile invertebrates within each portion of the rocky substrate, all digital images were analyzed with the CPCe V.3.3 software (National Coral Reef Institute/New Southeastern University) using a 60 point grid system. Preliminary tests showed no difference in percent cover when estimated with a 60 or 100 point system. Density of E. lucunter was quantified by counting all sea urchins that had some portion of its body within each sampled rectangle. In the present study algae was defined by the combination of algal turfs (represented primarily by coralline filamentous algae: Steneck and Dethier 1994) and a few other macroalgae, such as Codium spp. Stackh, 1797, Sargassum furcatum Ktzing 1843 and Dictyota spp. Lamouroux 1809, that occur in lower densities but in conjunction with the predominant algal turfs (Yoneshigue and Valentin 1988).

Results
The density of Echinometra lucunter varied greatly among portions of the rocky bottom at each locality (Fig. 3). At PI, sea urchin density varied between 3.33 1.56 and 65.28 8.95 individuals/m2, while density at SC varied between 6.25 2.57 and 32.64 4.15 individuals/m2 (mean + SD; Fig. 3). Darwinella sp.s percent cover varied between 0% and 17.69 7.30% at PI and 2.68 3.16% and 6.47 2.66% at SC

Fig. 3: Density of Echinometra lucunter within sampled portions of the rocky substratum at Porcos Island (PI) and Saco do Cherne (SC). The portions of High (H) and Low (L) urchin density are shown above bars (see Material and methods for details). Densities are number of individuals/m2. Mean + SD are shown.

214

Fig. 4: Density of Echinometra lucunter (individuals/m2), Darwinella sp. (% cover), algae (% cover) and other invertebrates (% cover) at Porcos Island (PI; N=46) and Saco do Cherne (SC; N=24). Mean + SE are shown.

with averages of 3.92% and 7.53% (PI and SC, respectively; Fig. 4). Algal cover was extremely high at both localities, reaching up to 90.33% at PI and 74.64% at SC (Fig. 4). All other invertebrates together reached maximum densities of 26.66% and 78.33% at PI and SC, correspondingly. Darwinella sp. cover was significantly different between portions of high (62.5 + 7.43 individuals/m2) and low (9.85 + 3.24 individuals/m2; mean + SE; Fig. 3) density of sea urchins (U = 121.00; N =12; P < 0.0001; Fig. 5). The percent cover of Darwinella sp. was more than ten-fold higher in areas of high (11.92 + 2.73%; mean + SE) compared to areas with low sea urchin densities (0.08 + 0.08%; mean + SE; Fig. 5). In general, algal cover was negatively related to the density of E. lucunter (R = -0.807; N = 70; P < 0.0001; Fig. 6A). A similar relationship was found between the percent cover of algae and Darwinella sp. (R = -0.649; N = 70; P < 0.0001; Fig. 6B). Conversely, the relationship between sea urchin density and Darwinella sp. cover, although significant, was weakly positive (R = 0.470; N = 70; P < 0.001; Fig. 6C). The same pattern was found when analyses were performed solely for PI (Fig. 7A, 7B, 7C). At SC, most relationships were weak (Fig. 7D, 7E, 7F), with the strongest relationship being between E. lucunters density and algal cover (R = 0.430; N = 24; P < 0.05). From the total number of Darwinellas quantified in the present study (N = 52), 75% of them were in total or partial contact with algal turfs. Less than 20% were in partial contact with other invertebrates, particularly the bryozoan Schizoporella errata (Walters, 1878). In most partial contacts between Darwinella sp. and other invertebrates, only 10% of the sponge surface was actually in contact with the invertebrate.

Fig. 5: Darwinella sp.s percent cover in areas of low and high urchin densities (see Material and methods for details). The mean density of Echinometra lucunter is also shown (individuals/m2). Mean + SE are shown (N=12).

Discussion
The present study demonstrates that areas with high density of the sea urchin Echinometra lucunter have a significantly higher cover of the sponge Darwinella sp. than areas with low sea urchin density. This observation, in addition with the negative correlations found between E. lucunters density and algal cover, and also between Darwinella sp. and algal cover suggest that Darwinella sp. and algae are

Fig. 6: Sperman Rank correlation analyses including both localities (PI and SC; N=70). A. Echinometra lucunter density and Darwinella sp. percent cover; B. Echinometra lucunter density and algae percent cover. C. Algae and Darwinella sp. percent cover. All percent covers were arcsin transformed.

215

Fig. 7: Sperman Rank correlation analyses performed by localities. PI = A, B, C (N=46); SC = D, E, F (N=24); A. Echinometra lucunter density and Darwinella sp. percent cover; B. Echinometra lucunter density and Darwinella sp. percent cover; C. Echinometra lucunter density and algae percent cover; D. Echinometra lucunter density and algae percent cover; E. Algae and Darwinella sp. percent cover; F. Algae and Darwinella sp. percent cover.

possibly competing for space and that the sea urchin might be controlling algal population. This result is corroborated by the frequent observation of contacts between Darwinella sp. and algae, however, manipulative experiments would be required to confirm this hypothesis. On the other hand, the weak relationship found between E. lucunters density and Darwinella sp. cover also suggests that there might be other factors besides competition with algae that is affecting the density and distribution of Darwinella sp. in Arraial do Cabo. Species of the genus Echinometra usually possess strong homing behavior and are often distributed in an aggregated pattern (McClanahan and Murtiga 2001). In the present study, the large variation in sea urchin density found among portions of the rocky substratum at both localities suggests that E. lucunter is distributed in an aggregated manner (in a scale of less than 0.5 meters), as observed in other species of this genus (McClanahan and Kurtis 1991). Therefore, it is expected that areas of high sea urchin density will support a low algal cover

and vice versa, since this urchins diet is mainly composed of algae (Oliveira 1991). The negative correlation between sea urchin density and algal cover, in the present study, supports Oliveiras (1991) findings that algae is E. lucunters main diet. However, it has been argued that in the absence of algae, E. lucunter can also feed on benthic invertebrates, such as sponges and cnidarians (McClintock et al. 1982, Oliveira 1991, McClanahan and Murtiga 2001). During the course of this study it was observed that Darwinella sp. that was found close to sea urchin aggregations presented irregular shapes that visually appeared to be caused by sea urchin removal (e.g., predation or abrasion). Thus, the weak relationship found between urchin density and Darwinella sp. cover could be due to the sea urchins grazing activity (predation) or by the accidental removal of sponge tissue (abrasion) when sea urchins move along the substratum. Although few studies have focused on sponge/algae competitive interactions (but see Palumbi 1985), studies on the negative effect of algae on the distribution of

216

other invertebrates are abundant (Quinn 1982, Jernakoff 1985, McCook et al. 2001). Algae are often the dominant competitive organism owing to their relative fast growth rates and the presence of secondary metabolites (Duffy and Hay 1990), although, sponges are also known to produce large quantities of chemical compounds (Thacker et al. 1998). The negative correlation between algae and Darwinella sp. cover in the present study suggests that, if competition is occurring, algae are competitively superior to sponges. If that is the case then, it would be interesting to determine what factors make algae competitively superior to sponges, and how general is this phenomenon. In general, the relationships among sea urchin density, Darwinella sp. and algal cover were similar at both localities, although they were weaker at SC compared to PI. The weak relationships at SC might be due to a lower sample size compared to PI, or it can possibly be due to site differences, such as fauna and flora composition or abiotic factors. At SC there was a higher abundance of benthic invertebrates, other than Darwinella sp., when compared to PI (Fig. 4). The high cover of other invertebrates at SC could have a negative effect on the distribution of Darwinella sp., through competition, which could have led to the observed weak relationship between algae and Darwinella sp. and also between Darwinella sp. and E. lucunter. The algal composition was also different between localities, with PI being dominated primarily by algal turfs, while SC had a larger abundance of macroalgae such as Codium sp. and Dictyota spp. (ENC, personal observation). Besides their greater palatability, macroalgae have a relatively higher biomass than algal turfs (Duffy and Hay 1990). Therefore, sea urchins at SC might have smaller grazing patches than at PI, since it has been demonstrated that patch size and grazing patterns are dependent on food availability (Russo 1977, McClanahan and Murtiga 2001). Many authors in the past have suggested that sponge distribution was intimately related to competition (reviewed by Wulff 2006). However, recent studies have pointed out the importance of abiotic factors as determinants of sponge distribution (Preciado and Maldonado 2005). What has recently been argued is that abiotic factors could actually exclude species from a locality (Alcolado 1994, Wulff 2005), but if a species is not inhibited by these factors, competitive interactions may greatly affect species distribution (Wulff 2005). In the present study, abiotic factors such as substratum inclination and depth were similar; however, water flow regimes were different with SC being an exposed site and PI a sheltered one (Fig. 1). This difference in water flow regimes might affect the distribution of Darwinella sp. and its relationship with algae and other invertebrates. To conclude, the results of this study point to a possible competition for space between algae and Darwinella sp., which is probably mediated by the sea urchin E. lucunter. Nevertheless, manipulative studies are needed in order to confirm the negative effect of algal cover on the cover of Darwinella sp, and the positive impact of the sea urchin on Darwinella sp. distribution.

Acknowledgements:
We thank Isaac Zilberberg for providing logistic support in Arraial do Cabo (boat and lodge); Mariana Melo and Fernanda Oliveira for help during the field work. ENC thanks Renato Ventura for insightful conversations that greatly improved the contents of this manuscript. This manuscript was submitted as partial fulfillment of the PhD degree to ENC at Universidade Federal do Rio de Janeiro, Brazil. This study was supported by a fellowship from CAPES (Coordenao de Aperfeioamento de Pessoal de Nvel Superior) to ENC.

References
Alcolado PM (1994) General trends in coral reef sponge communities in Cuba. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 251-256 Bell JJ (2002) The sponge community in a semi-submerged temperate sea cave: density, diversity and richness. Mar Ecol 23: 297-311 Bell JJ, Barnes DKA (2003) The importance of competitor identity, morphology and ranking methodology to outcomes in interference competition between sponges. Mar Biol 143: 415-426 Buss LW (1979) Habitat selection, directional growth and spatial refuges: why colonial animals have more hiding places. In: Coulson JC, London DF (eds). Biology and systematics of colonial organisms. Academic Press, New York. pp. 459-497 Buss LW (1980) Competitive intransitivity and size-frequency distributions of interacting populations. Proc Natl Acad Sci USA 77(9): 5355-5359 Buss LW (1990) Competition within and between encrusting clonal invertebrates. Trends Ecol Evol 5(11): 352-356 Carpenter RC (1986) Partitioning herbivory and effects on coral reef algal communities. Ecol Monog 56(4): 345-363 Carpenter RC (1990) Mass mortality of Diadema antillarum. I. Long-term effects on sea urchin population-dynamics and coral reef algal communities. Mar Biol 104: 67-77 Castro CB, Echeverria CA, Pires DO, Mascarenhas MJ, Freitas SG (1995) Distribuio de Cnidaria e Echinodermata no infralitoral de costes rochosos de Arraial do Cabo, Rio de Janeiro, Brasil. Rev Bras Biol 55(3): 471-480 Connell JH (1961) The influence of interspecific competition and other factors on the distribution of the barnacle Chthamalus stellatus. Ecology 42: 710-723 Connell JH (1973) Population ecology of reef-building corals. In: Jones OA, Endean R (eds). Biology and geology of coral reefs. Academic Press, New York. pp. 205-245 Duffy JE, Hay ME (1990) Seaweed adaptations to herbivory: chemical, structural, and morphological defenses are often adjusted to spatial or temporal patterns of attack. BioScience 40(5): 368375 Edmunds P, Carpenter RC (2001) Recovery of Diadema antillarum reduces macroalgal cover and increase abundance of juvenile corals on a Caribbean reef. Proc Natl Acad Sci 98: 5067-5071 Ferreira CEL, Gonalves JEA, Coutinho R, Peret AC (1998) Herbivory by the Dusky Damselfish Stegaster fuscus (Cuvier, 1830) in a tropical rocky shore: effects on the benthic community. J Exp Mar Biol Ecol 229: 241-264 Hughes TP (1994) Catastrophes, phase shifts, and large-scale degradation of a Caribbean coral reef. Science 265: 15471551

217

Jackson J, Winston JE (1982) Ecology of cryptic coral reef communities. I. Distribution and abundance of major groups of encrusting organisms. J Exp Mar Biol Ecol 57: 135-147 Jackson J (1977) Competition on marine hard substrata: the adaptive significance of solitary and colonial strategies. Am Nat 111(980): 743-767 Jernakoff P (1985) The effect of overgrowth by algae on the survival of the intertidal barnacle Tesseropora rosea Krauss. J Exp Mar Biol Ecol 94: 89-97 Jompa J, McCook LJ (2002) Effects of competition and herbivory on interactions between a hard coral and brown alga. J Exp Mar Biol Ecol 271: 25-39 Lobban CS, Harison PJ (1994) Seaweed ecology and physiology. Cambridge University Press, Cambridge McClanahan TR, Kurtis JD (1991) Population regulation of the rock-boring sea urchin Echinometra mathaei (de Blainville). J Exp Mar Biol Ecol 147: 121-146 McClanaham TR, Muthiga N (2001) The ecology of Echinometra. In: Lawrence JM (ed). Edible sea urchins: biology and ecology. Developments in Aquaculture and Fisheries Science, vol. 32. Elsevier Science, Amsterdam. pp. 225-243 McClintock JB, Klinger TS, Lawrence JM (1982) Feeding preferences of echinoids for plant and animal food models. Bull Mar Sci 32: 365-369 McCook LJ, Jompa J, Diaz-Pulido G (2001) Competition between corals and algae on coral reefs: a review of evidence and mechanisms. Coral Reefs 19: 400-417 McKinney FK (1992) Competitive interactions between related clades: evolutionary implications of overgrowth interactions between encrusting cyclostome and cheilostome bryozoans. Mar Biol 114: 645-652 Muricy G, Hajdu E, Custdio MR, Klautau M, Russo C, Peixinho S (1991) Sponge distribution at Arraial do Cabo, SE Brazil. In: Magoon OT, Converse H, Tippie V, Tobie LT, Clark D (eds). Proc VIIth Symp Coast Ocean Manag, Long Beach. ASCE Publ. pp. 1183-1195 Oliveira MC (1991) Survival of seaweeds ingested by three species of tropical sea urchins from Brazil. Hydrobiologia 222: 13-17 Paine RT (1966) Food web complexity and species diversity. Am Nat 100: 65-75 Paine RT, Vadas R (1969) The effects of grazing by sea urchins, Strongylocentrotus spp., on benthic algal populations. Limnol Oceanogr 14: 710-719

Palumbi SR (1985) Spatial variation in an alga-sponge commensalismand the evolution of ecological interactions. Am Nat 126: 267274 Preciado I, Maldonado M (2005) Reassessing the spatial relationship between sponges and macroalgae in sublittoral rocky bottoms: a descriptive approach. Helgol Mar Res 59: 141-150 Quinn JF (1982) Competitive hierarchies in marine benthic communities. Oecologia 54: 129-135 Russo AR (1977) Water flow and the distribution and abundance of echinoids (genus Echinometra) on a Hawaiian Reef. Aust J Mar Freshw Res 28: 693-702 Sammarco PW (1980) Diadema and its relationship to coral spat mortality: grazing, competition, and biological disturbance. J Exp Mar Biol Ecol 45: 245-272 Sammarco PW (1982) Echinoid grazing as a structuring force in coral communities: whole reef manipulations. Exp Mar Biol Ecol 61: 31-35 Sokal RR, Rohlf FJ (1995) Biometry, the principles and practices of statistics in biological research (3th ed). WH Freeman and Company, San Francisco Steneck RS, Dethier MN (1994) A functional group approach to the structure of algal-dominated communities. Oikos 69: 476-498 Sutherland J (1976) Species diversity gradients: synthesis of the roles of predation, competition, and temporal heterogeneity. Am Nat 110(973): 351-369 Thacker RW, Becerro MA, Lumbang WA, Paul VJ (1998) Allelopathic interactions between sponges on a tropical reef. Ecology 79(5): 1740-1750 Tuya F, Boyra A, Sanchez-Jerez, Barbera C, Haroun R (2004) Can one species determine the structure of the benthic community on a temperate rocky reef? The case of the long-spined sea-urchin Diadema antillarum (Echinodermata: Echinoidea) in the eastern Atlantic. Hydrobiologia 519: 211-214 Witman JD, Sebens KP (1990) Distribution and ecology of sponges at a subtidal rock ledge in the Central Gulf of Maine. In: Rutzler K (ed). New perspectives in sponge biology. Smithsonian Institution Press, Washington DC. pp. 391-396 Wulff JL (2005) Trade-offs in resistance to competitors and predators, and their effects on the diversity of tropical marine sponges. J Anim Ecol 74: 313-321 Wulff JL (2006) Ecological interactions of marine sponges. Can J Zool 84: 146-166 Yoneshigue Y, Valentin JL (1988) Comunidades algais fotfilas de infralitoral de Cabo Frio, Rio de Janeiro, Brasil. Gayana Bot 45(14): 61-75

Porifera research: Biodiversity, innovation and sustainaBility - 2007

219

Sponges (Porifera, Demospongiae) from Bransfield strait, off Joinville Island, collected by Brazilian Antarctic Program - PROANTAR
Maurcio Campos(1,2*), Beatriz Mothes(1), Cla Lerner(1), Joo Lus Carraro(1,3), Inga Ludmila VeitenheimerMendes(2)
Museu de Cincias Naturais, Fundao Zoobotnica do Rio Grande do Sul. Rua Salvador Frana 1427, 90690-000. Porto Alegre-RS, Brazil. mrcpoa@hotmail.com, bmothes@fzb.rs.gov.br, cblerner@fzb.rs.gov.br, jlc_rs@hotmail.com (2) Programa de Ps-Graduao em Biologia Animal, Universidade Federal do Rio Grande do Sul. Av. Bento Gonalves 9500, 91501-970. Porto Alegre-RS, Brazil. inga.mendes@ufrgs.br (3) Programa de Ps-Graduao em Ecologia, Universidade Federal do Rio Grande do Sul. Av. Bento Gonalves 9500, 91501-970. Porto Alegre-RS, Brazil
(1)

Abstract: This is the first taxonomic study of the sponges of Joinville Island (Bransfield Strait, Antarctica). In total, 10 species were identified, viz. Iophon terranovae, Artemisina apollinis, Myxilla (Ectyomyxilla) mariana, Mycale (Oxymycale) acerata, Isodictya erinacea, Haliclona (Gellius) rudis, Haliclona (Rhizoniera) dancoi, Haliclonissa verrucosa, Microxina benedeni and Microxina phakelloides. Iophon terranovae and M. (E.) mariana are recorded for the first time for this Antarctic region; I. terranovae, M. (E.) mariana, H. verrucosa and M. phakelloides had their bathymetric ranges extended. Keywords: Antarctica, Demospongiae, distribution, taxonomy, PROANTAR

Introduction
The taxonomy of Antarctic sponges was studied by many authors, who described over 300 species collected through several oceanographic expeditions undertaken in the past 110 years. Highlights are the works of Topsent (1901, 1908, 1913, 1917), von Lendenfeld (1907), Kirkpatrick (1908), Hentschel (1914), Burton (1929, 1932, 1934, 1938) and Koltun (1964). More recently, new records were made by Desqueyroux-Fandez (1989), Barthel et al. (1990, 1997), Pansini et al. (1994), Gutt and Koltun (1995), Mothes and Lerner (1995), Calcinai and Pansini (2000), and Ros et al. (2004). Additionally, important contributions were made on the taxonomy of sponges from the subantartic region, which belongs to the Antarctic Faunistic Complex (Sar et al. 1992), by Ridley (1881), Ridley and Dendy (1887), Sollas (1888), Thiele (1905), Burton (1940), Sar (1978), Boury-Esnault and van Beveren (1982) and Desqueyroux and Moyano (1987). In spite of the many studies conducted, some areas are still unsatisfactorily sampled, such as the South Atlantic Ocean islands, the South Shetland Islands and neighboring areas (Ros et al. 2004). The conduction of new faunistic surveys in the Antarctic continent will be of great importance in order to correlate abundance with environmental factors, to improve understanding of yearly changes and also to extend geographic and bathymetric distributions (Desqueyroux-

Fandez 1989), besides describing new species. The present study aims to increase the knowledge of the sponge fauna from Antarctica, and also to provide a complete illustration of all the identified species.

Materials and methods

The sponges studied here were collected with a beamtrawl during the IVth expedition of the Brazilian Antarctic Program, near Joinville Island (6253S-5627W / 6301S5449W; Fig. 1), between 82 and 274 m depth. The specimens are deposited in the Porifera collection of Museu de Cincias Naturais, Fundao Zoobotnica do Rio Grande do Sul, Brazil. Taxonomic identification was based on dissociated spicules mounts and thick sections of skeletal architecture, following the techniques of Mothes-de-Moraes (1978) and Mothes et al. (2004); preparations for SEM study were done according to Mothes and Silva (2002). Abbreviations used are BMNH (The Natural History Museum, London, England); MCNPOR (Porifera Collection, Museu de Cincias Naturais, Fundao Zoobotnica do Rio Grande do Sul, Brazil); MSNG (Museo Civico di Storia Naturale Giacomo Doria, Genova, Italy); ZMB (Zoologische Museum fr Naturkunde an der Universitt Humboldt zu Berlin, Berlin, Germany).

220

Spicules: Megascleres: styles I: 350-450-550 / 12-18-21 m (Figs. 2C-D); styles II: 330-410-520 / 2.5-4.8-9.0 m (Fig. 2E); anisochelae: 48-55-62 m (Fig. 2F); bipocilli: 7.5-10-12 m (Fig. 2G). Remarks: Slightly malformed styles I, albeit seen in comparative material, were not observed in the specimen from Joinville Island. Distribution: Antarctica (Victoria Land, Bransfield Strait, Joinville I.). Bathymetry: 82-135 m. Family Microcionidae Artemisina apollinis (Ridley and Dendy, 1886) (Figs. 3A-J) Amphilectus apollinis Ridley and Dendy, 1886: 350. Artemisina apollinis; Koltun, 1976: 188; DesqueyrouxFandez, 1989: 125, figs. 22a-e; Ros et al., 2004: 103, figs. 5A-H. Further synonymy see Desqueyroux-Fandez (1989). Material: MCNPOR 1999, Est. 4866: 6253 S - 5627 W, 194 m, 03.II.1986.
Fig. 1: Map showing the studied area; marked points indicate the specific collecting places.

Comparative material: BMNH 1908.2.5.166d - Artemisina apollinis (Ridley and Dendy, 1886). Description: Massive specimen (Fig. 3A), dimensions 7.5 x 4.5 x 3.2 cm; surface hispid to the touch with ramified conules, oscules 0.3-0.4 cm diameter. Preserved material fragile, with brittle consistency, colour light brown. Skeleton: (Fig. 3B) Ectosome without specialization. Choanosome formed by multispicular directionless tracts, whereas closer to the surface such tracts are perpendicularly arranged, producing the surface hispidation. Isochelae dispersed throughout the skeleton. Spicules: Megascleres: styles I: 550-663-730 / 30-34-38 m (Figs. 3C-D); styles II: 320-412-530 / 5.0-6.0-9.0 m (Figs. 3E-F); isochelae: 12-15-16 m (Fig. 3G); toxas I: 340-520740 m (Fig. 3H-I); toxas II: 118-157-195 m (Fig. 3J). Remarks: Spicule sizes reported for the species are varied, but nevertheless measurements obtained for the sample from Joinville Island are slightly different. The comparative material analysed has thinner styles I and smaller toxas, not separable in different size classes (remeasured, in m: styles I: 503-582-665 / 13-15-17; styles II: 295-348-485 / 4.0-5.07.0; isochelas: 14-16-18; toxas: 80-186-370). Samples from Kerguelen (Koltun 1976, Boury-Esnault and van Beveren 1982) have smaller styles I, and toxas separable in distinct size classes only in some descriptions (Topsent 1908, BouryEsnault and van Beveren 1982, Desqueyroux-Fandez 1989).

Results
Order Poecilosclerida Suborder Microcionina Family Acarnidae Iophon terranovae Calcinai and Pansini, 2000 (Figs. 2A-G) Iophon terranovae Calcinai and Pansini, 2000: 371, figs. 2AJ, figs. 3A-F Material: MCNPOR 1951, Est. 4865: 6255 S - 5516 W, 82 m, 03.II.1986. Comparative material: MSNG 31 - Iophon terranovae Calcinai and Pansini, 2000. Description: Cylindrical specimen, incomplete (Fig. 2A), dimensions 6.7 x 3.2 x 3.4 cm; surface smooth, oscules not observed. Preserved material extremely friable consistency, colour dark brown. Skeleton: (Fig. 2B) Ectosome with styles I, perpendicular to the surface and protruding beyond it. Choanosome a confused reticulation, styles I and II arranged in multispicular tracts or dispersed. Anisochelae and bipocilli occur throughout the skeleton.

221

Fig. 2: Iophon terranovae Calcinai and PansinI. 2000. A. Specimen. B. Skeleton. C. Styles I. D. Extremities of style I. E. Styles II. F. Anisochela. G. Bipocilli.

Distribution: Subantarctic region (Kerguelen I.), South Atlantic (South Georgia), Antarctica (Wilhelm II Land, Graham Land, Victoria Land, South Shetland Is., Bransfield Strait, Joinville I.). Bathymetry: 7-380 m. Suborder Myxillina Family Myxillidae Myxilla (Ectyomyxilla) mariana Ridley and Dendy, 1886 (Figs. 4A-J) Myxilla mariana Ridley and Dendy, 1886: 472. Ectyomyxilla mariana; Koltun, 1964: 76. Further synonymy see Koltun (1964). Material: MCNPOR 1962, Est. 4865: 6255 S - 5516 W, 82 m, 03.II.1986. Comparative material: BMNH 1887.5.2.108 - Myxilla mariana Ridley and Dendy, 1886. Description: (Fig. 4A) Fragment, 7.7 x 4.5 x 3.8 cm in dimensions; surface rugose, with ridges and grooves; several oscules scattered on the surface (< 0.1-0.2 cm diameter).

Preserved material slightly compressible and fragile, colour light brown. Skeleton: (Fig. 4B) Ectosome formed by megascleres in confusion. Choanosome bearing an inconspicuous reticulation which sometimes forms triangular meshes, with multispicular tracts transversed by free spicules, without any evident orientation. Spicules: Megascleres: acanthostyles I: 342-438-494 / 1316-20 m (Figs. 4C-D); acanthostyles II: 76-104-142 / 4.05.0-6.0 m (Figs. 4E-F); tylotes: 238-271-304 / 7.0-9.0-12 m (Figs. 4G-H); isochelae: 20-25-30 m (Fig. 4I); sigmas: 41-50-71 m (Fig. 4J). Remarks: Strongyloid tylotes were observed in the studied samples, some of which without spines. Isochelae are smaller in comparison to the measurements supplied by Ridley and Dendy (1887) and Hentschel (1914). Another unexpected observation made here were the very small numbers of acanthostyles II. The comparative material studied has spicules identical to those observed in the specimen from Joinville Island, with only a few significant differences (remeasured, in m: acanthostyles I: 332-385-418 / 13-14-16; acanthostyles II: 103-139-193 / 8.0-10-14; tylotes: 209-244-275 / 6.0-7.08.0; isochelae: 17-29-49; sigmas: 34-44-57).

222

Fig. 3: Artemisina apollinis (Ridley and Dendy, 1886). A. Specimen. B. Skeleton. C. Styles I. D. Extremities of style I. E. Style II. F. Extremities of styles II. G. Isochela. H. Toxa I. I. Detail of extremity of toxa I. J. Toxa II.

Distribution: Subantarctic region (Marion I.), South America (Chile), Antarctica (Wilhelm II Land, Queen Mary Land, Joinville I.). Bathymetry: 82-385 m. Suborder Mycalina Family Mycalidae Mycale (Oxymycale) acerata (Kirkpatrick, 1907) (Figs. 5A-H) Mycale acerata Kirkpatrick, 1907: 281; Burton, 1934: 23, pl. VIII, figs. 1-4; 1938: 11; 1940: 103; Koltun, 1976: 169; Desqueyroux and Moyano, 1987: 48; Desqueyroux-Fandez, 1989: 111, pl. II, figs. 9a-e, pl. VII, fig. 43, pl. VIII, figs. 4447; Barthel et al., 1990: 122; 1997: 48; Pansini et al., 1994: 70; Cattaneo-Vietti et al., 1999: 540; Ros et al., 2004: 110, text-fig. 9A-E

Oxymycale acerata; Sar et al., 1990: 253; Gutt and Koltun, 1995: 231. Further synonymy see Desqueyroux-Fandez (1989). Material: MCNPOR 1983, Est. 4864: 6301 S - 5449 W, 275 m, 02.II.1986. Comparative material: BMNH 1908.2.5.171b - Mycale acerata Kirkpatrick, 1907. Description: (Fig. 5A) Erect and ramified specimen, dimensions: 9.0 x 7.0 x 9.5 cm; surface partially destroyed. Preserved material bears hard consistency and little flexibility, colour mostly white, with light brown shades in some regions. Skeleton: Ectosome a tangential reticulation composed by multispicular tracts, forming triangular meshes (Fig. 5B). Choanosome with thick and compact multispicular tracts, connected by secondary tracts (Fig. 5C). Both types of anisochelae, as well as the raphids, are distributed along the entire tracts.

223

Fig. 4: Myxilla (Ectyomyxilla) mariana Ridley and Dendy, 1886. A. Specimen. B. Skeleton. C. Acanthostyle I. D. Extremities of acanthostyle I. E. Acanthostyle II. F. Extremities of acanthostyle II. G. Tylote. H. Extremities of tylote. I. Isochela. J. Sigmas.

Spicules: Megascleres: oxeas: 650-806.4-890 / 12.5-17.1-20 m (Figs. 5D-E); raphids: 25-31-35 m (Fig. 5F); anisochelae I: 87.5-104.6-117.5 m (Fig. 5G); anisochelae II: 30-44.852.5 m (Fig. 5H). Remarks: The species is widely distributed in the Antarctic and subantarctic regions. Comparative material analysed only differs by the presence of larger raphids and absence of smaller anisochelae (remeasured, in m: oxeas: 760-830-920 / 17-19-24; raphids: 62-76-90; anisochelas: 90-102-117). Distribution: Subantarctic region (Macquarie I., Kerguelen I.), South America (Chile, Argentina, Falkland Is.), South Atlantic Ocean (Shag Rocks, South Georgia, South Orkneys), Antarctica (Victoria Land, Graham Land, Adelie Land, Wilhelm II Land, Banzare Land; McRobertson Land; Princess

Ragnhild Land, Enderby Land, Weddell Sea, South Shetland Is., Joinville I.). Bathymetry: 0-731 m. Family Isodictyidae Isodictya erinacea (Topsent, 1916) (Figs. 6A-E) Homoeodictya erinacea Topsent, 1916: 169 Isodictya erinacea; Koltun, 1964: 40, pl. VIII, figs. 4-7; 1976: 171; Desqueyroux, 1972: 52; 1975: 59, pl. II, figs. 18-20; Vacelet and Arnaud, 1972: 15; Desqueyroux-Fandez, 1989: 114, pl. III, figs. 12a-c, pl. IX, figs. 53-55; Barthel et al., 1990:

224

Fig. 5: Mycale (Oxymycale) acerata (Kirkpatrick, 1907). A. Specimen. B. Ectosome. C. Choanosome. D. Oxea. E. Extremities of oxea. F. Raphid. G. Anisochelas I, H. Anisochelas II.

122; 1997: 48; Sar et al., 1990: 253; Pansini et al., 1994: 72; Gutt and Koltun, 1995: 230; Cattaneo-Vietti et al., 1999: 540; Ros et al., 2004: 116, figs. 14a-g. Further synonymy see Koltun (1964). Material: MCNPOR 1961, Est. 4865: 6255 S - 5516 W, 82 m, 02.II.1986. Description: (Fig. 6A) Ramose fragment, dimensions: 14 x 3.5 x 2.0 cm; spiny surface, branching from the central axis. Preserved material of stiff consistency, colour light brown. Skeleton: (Fig. 6B) Ectosome absent. Choanosome composed of thick longitudinal multispicular tracts (400-820 m thickness), irregularly connected by megascleres in crisscross arrangement, forming rounded to polygonal meshes (370-810 m diameter). Isochelae dispersed along the tracts.

Spicules: Megascleres: oxeas: 600-718.4-800 / 18.8-27.131.3 m (Figs. 6C-D); isochelae: 42.5-52.7-57.5 m (Fig. 6E). Remarks: Desqueyroux-Fandez (1989) found a second category of smaller isochelae in her samples. In the samples studied by Ros et al. (2004), as well as in the present study, such spicules were also seen, but considered to be growth stages of the larger ones. Distribution: South Atlantic Ocean (South Georgia, Burdwood Bank), Antarctica (Graham Land, Palmer Archipelago, Victoria, Banzare Land; McRobertson Land, Enderby Land, Adelie Land, Weddell Sea, Joinville I., South Shetland Is., Bransfield Strait). Bathymetry: 20-920 m.

225

Fig. 6: Isodictya erinacea (Topsent, 1916). A. Specimen. B. Skeleton. C. Oxeas. D. Extremities of oxea. E. Isochela.

Order Haplosclerida Suborder Haplosclerina Family Chalinidae Haliclona (Gellius) rudis (Topsent, 1901) (Figs. 7A-G) Gellius rudis Topsent, 1901: 14, pl. I, fig. 9, pl. III, fig. 4; Desqueyroux-Fandez, 1989: 127, pl. IV, figs. 24a-b, pl. XV, fig. 86; Barthel et al., 1990: 123; Pansini et al., 1994: 80; Cattaneo-Vietti et al., 1999: 540. Further synonymy see Desqueyroux-Fandez (1989).

Material: MCNPOR 1984, Est. 4866: 6253 S - 5627 W, 194 m, 03.II.1986. Description: (Fig. 7A) Digitiform specimen, dimensions: 8.0 x 5.5 cm; surface hispid to the touch, with ridges and grooves; oscules 0.1-0.2 cm diameter, positioned on top of conules; small pores observed on surface (< 0.1 cm diameter). Preserved material with friable consistency, colour light brown. Skeleton: (Fig. 7B) Ectosome without specialization. Choanosome formed by a dense arrangement of multispicular tracts, interconnected by isolated megascleres. Part of the skeleton is halichondrioid, confused and irregular, with tracts oriented in several directions. Sigmas are seen at the nodes of megascleres.

226

Fig. 7: Haliclona (Gellius) rudis (Topsent, 1901). A. Specimen. B. Skeleton. C. Oxea I. D. Extremities of oxea I. E. Oxea II. F. Extremities of oxea II. G. Sigmas.

Spicules: Megascleres: oxeas I: 330-439.6-530 / 12.5-17.420 m (Figs. 7C-D); oxeas II: 240-322-420 / 2.5-5.5-8.8 m (Figs. 7E-F); sigmas: 17.5-32.5-55 m (Fig. 7G). Remarks: The spicules in the present material differ from those reported in most previous studies. In the present study two size classes of oxeas were found, a feature previously reported only by Boury-Esnault and van Beveren (1982). Distribution: Subantarctic region (Kerguelen I.), Antarctica (Bellingshausen Sea, Graham Land, Victoria Land, Weddell Sea, Joinville I., South Shetland Is., Bransfield Strait). Bathymetry: 20-500 m. Haliclona (Rhizoniera) dancoi (Topsent, 1901) (Figs. 8A-D) Reniera dancoi Topsent, 1901: 12, pl. II, fig. 1, pl. III, fig. 3.

Haliclona dancoi; Koltun, 1964: 95, pl. XV, figs. 5-6; 1976: 196; Barthel et al., 1990: 123; Gutt and Koltun, 1995: 231; Cattaneo-Vietti et al., 1999: 540. Further synonymy see Koltun (1964). Material: MCNPOR 1986, Est. 4867: 6257 S - 5650 W, 95 m, 03.II.1986. Description: (Fig. 8A) Partially broken specimen, arborescent, dimensions: 5.2 x 1.2 x 1.4 cm; surface hispid to the touch, with protruding spicules; oscules 0.1-0.2 cm in diameter. Preserved material with friable consistency, colour beige. Skeleton: (Fig. 8B) Ectosome formed by the ends of primary tracts, partially arranged in discrete bouquets. Choanosomal network composed by multispicular primary tracts (65-140 m thickness), connected by uni to paucispicular secondary tracts, forming polygonal to triangular meshes. Spicules: Megascleres: oxeas: 380-468.2-590 / 16.3-23.6-30 m (Figs. 8C-D).

227

Fig. 8: Haliclona (Rhizoniera) dancoi (Topsent, 1901). A. Specimen. B. Skeleton. C. Oxea. D. Extremities of oxea.

Remarks: Comparing the measurements of spicules from previous records (Topsent 1901, 1908, Kirkpatrick 1908, Hentschel 1914, Koltun 1964, 1976) with measurements obtained in the present study, some variation was observed; however this particularity is interpreted as intraespecific variation. Distribution: South Atlantic Ocean (South Orkneys), Antarctica (Bellingshausen Sea, Graham Land, Victoria Land, Wilhelm II Land, Princess Elisabeth Land, McRobertson Land, Enderby Land, Adelie Land, Sabrina Land, Weddell Sea, Joinville I., South Shetland Is.). Bathymetry: 18-2267 m.

Comparative material: BMNH Haliclonissa verrucosa Burton, 1932.

1928.2.15.723a

Description: (Fig. 9A) Cylindrical sponge; dimensions: 5.0 x 1.9 x 1.2 cm; surface verrucose and hispid to the touch. Oscules 0.1-0.3 cm in diameter. Preserved material showing very friable consistency, colour beige. Skeleton: (Fig. 9B) Ectosome formed by the ends of choanosomal tracts, in varied positions. Choanosome composed by longitudinal multispicular tracts, which protrude through the surface, irregularly connected by secondary tracts. Both types of oxeas form the tracts, although few oxeas II are present in the specimens studied. Spicules: Megascleres: oxeas I: 351.5-412.1-503.5 / 10.412.4-15 m (Figs. 9C-D); oxeas II: 256.5-288.2-323 / 2.54.3-6.9 m (Figs. 9E-F). Remarks: Desqueyroux-Fandez and Valentine (2002) added new information concerning the spicular content of this species, recording a second category of oxeas which were also observed in the present study. Distribution: South America (Uruguay, Argentina), Antarctica (Palmer Archipelago, Victoria Land, Weddell Sea, Joinville I., South Shetland Is.). Bathymetry: 25-194 m.

Family Niphatidae Haliclonissa verrucosa Burton, 1932 (Figs. 9A-F) Haliclonissa verrucosa Burton, 1932: 270, pl. LI, fig. 3, textfig. 8; 1940: 100; Koltun, 1964: 102; Desqueyroux, 1972: 54; Barthel et al., 1990: 123. Material: MCNPOR 1990, Est. 4866: 6253 S - 5627 W, 194 m, 03.II.1986.

228

Fig. 9: Haliclonissa verrucosa Burton, 1932. A. Specimen. B. Skeleton. C. Oxea I. D. Extremities of oxea I. E. Oxea II. F. Extremities of oxea II.

Microxina benedeni (Topsent, 1901) (Figs. 10A-E) Gelliodes benedeni Topsent, 1901: 16, pl. II, fig. 3, pl. III, fig. 5. Microxina benedeni; Burton, 1934: 11; Koltun, 1976: 196; Desqueyroux, 1975: 73, pl. IV, figs. 55-57; Barthel et al., 1990: 123; 1997: 49; Pansini et al., 1994: 80; Gutt and Koltun, 1995: 230; Cattaneo-Vietti et al., 1999: 540. Further synonymy see Desqueyroux (1975).

Material: MCNPOR 1982, Est. 4866: 6253 S - 5627 W, 194 m, 03.II.1986. Description: (Fig. 10A) Cylindrical specimen; dimensions: 7.3 x 2.2 x 2.4 cm; surface densely spiny due to the presence of stiff conules; oscules 0.1-0.3 cm diameter. Preserved material very firm and incompressible in consistency, colour light brown. Skeleton: (Fig. 10B) Ectosome without tangential specialization. Choanosome composed by longitudinal multispicular tracts (320-700 m thickness), forming tufts

229

Fig. 10: Microxina benedeni Topsent, 1901. A. Specimen. B. Skeleton. C. Oxea. D. Extremities of oxea. E. Sigma.

which characterize the superficial texture; between the tracts the spicules occur in an irregular arrangement which can bear irregular meshes. Spicules: Megascleres: oxeas: 408.5-804.1-902.5 / 27.5-30.355 m (Figs. 10C-D); sigmas: 20-30.4-55 m (Fig. 10E). Remarks: The present sample has only sigmas as microscleres, which occur in low frequency. Distribution: South America (Falkland Is.), South Atlantic Ocean (South Georgia), Antarctica (Bellingshausen Sea, Graham Land, Victoria Land, Palmer Archipelago, Banzare Land; Princess Elisabeth Land, McRobertson Land, Enderby Land, Weddell Sea, Joinville I., South Shetland Is.). Bathymetry: 81-1266 m. Microxina phakelloides (Kirkpatrick, 1907) (Figs. 11A-F) Sigmaxynissa phakelloides Kirkpatrick, 1907: 272. Gellius phakelloides; Barthel et al., 1990: 123.

Haliclona phakelloides; Koltun, 1964: 101, pl. XIV, figs. 1113. Further synonymy see Koltun (1964). Material: MCNPOR 2046, Est. 4865: 6255 S - 5516 W, 82 m, 03.II.1986. Description: (Fig. 11A) Massive and amorphous specimen; dimensions: 7.8 x 6.3 x 1.1 cm; surface conulose; oscules 0.1 cm diameter. Preserved material with friable consistency, colour light brown. Skeleton: (Fig. 11B) Ectosome formed by thick multispicular tracts, perpendicular to the surface, where megascleres are positioned in tufts which render the surface hispid. Choanosome with uni- to paucispicular tracts forming polygonal meshes, diameter 300-380 m. Sigmas and toxas between the meshes. Spicules: Megascleres: oxeas: 627-707-779 / 25.3-32.7-36.8 m (Figs. 11C-D); sigmas: 52.9-83.9-128.8 / 2.5-3.9-5.0 m (Fig. 11E); toxas: 94.3-130-184 / 2.5-3.9-5.0 m (Fig. 11F). Remarks: The spicules of the species are very characteristic, with oxeas of great dimensions, sigmas with remarkable

230

Fig. 11: Microxina phakelloides (Kirkpatrick. 1907). A. Specimen. B. Skeleton. C. Oxea. D. Extremities of oxea. E. Sigmas. F. Toxas.

deformation in its contour and centroangulate toxas. The values registered by Kirkpatrick (1908), Hentschel (1914) and Koltun (1964) for the spicular meaurements are very similar to the samples of the present study. Distribution: Antarctica (Victoria Land, Knox Land, Banzare Land, Wilhelm II Land, Weddell Sea, South Shetland Is., Joinville I., Bransfield Strait). Bathymetry: 66-550 m.

Concluding remarks
The species dealt with here are all first records for Joinville Island. With the new occurrences of Iophon terranovae, Myxilla (Ectyomyxilla) mariana and Haliclona (Rhizoniera) dancoi for this island, their distribution is extended. The known bathymetric ranges for I. terranovae, M. (Ectyomyxilla)

mariana, Haliclonissa verrucosa and Microxina phakelloides were also extended in the present contribution. This new panorama of the sponges in Antarctica corroborates the ideas of Desqueyroux-Fandez (1989) and Ros et al. (2004), in revealing the necessity for new collections, mainly in the region comprising the Graham Land and Palmer Archipelago, along with South Shetland Is., South Orkneys and the vicinity of South Sandwich. Accomplishment of this task would permit a better understanding of the real geographic and bathymetric distribution of the species belonging to the Antarctic complex. All the species recorded for Jonville I. until the present also occur at continental antarctic areas, and the majority generally extends their occurrence for the southernmost tip of South America, to

231

South Atlantic localities (South Georgia and South Orkneys) and to Kerguelen I. in Subantarctic region.

Acknowledgements
We thank Clare Valentine (BMNH), Dr. Barbara Calcinai (Dipartimento di Scienze del Mare, Universit di Ancona, Ancona, Italy) and Dr. Carsten Lter (ZMB) for the loan of comparative material; Dr. Eduardo Hajdu (Museu Nacional, Universidade Federal do Rio de Janeiro, Brazil) and Dr. Ruth Desqueyroux-Fandez (Museum dHistoire Naturelle de Genve, Geneva, Switzerland) for help with bibliography. We also thank CAPES and CNPq (Brazil) for research grants.

References
Barthel D, Tendal O, Panzer K (1990) Ecology and taxonomy of sponges in the eastern Weddell Sea shelf and slope communities. Ber Polarforsch 68: 120-130 Barthel D, Tendal O, Gatti S (1997) The sponge fauna of the Weddell Sea and its integration in benthic processes. Ber Polarforsch 249: 44-52 Boury-Esnault N, van Beveren M (1982) Les dmosponges du plateau continental de Kerguelen-Heard. Com Nat Fr Rech Antarct 52: 1-175 Burton M (1929) Porifera. Part II. Antarctic sponges. Nat Hist Rep Brit Antarct Terra Nova Exped, 1910, Zool 6(4): 393-458 Burton M (1932) Sponges. Discovery Rep 6: 237-392 Burton M (1934) Sponges. Further Zool Res Swedish Antarctic Exped 1901-1903 3(2): 1-58 Burton M (1938) Non-calcareous sponges. Australas Antarct Exp 1911-1914, Sci Rep Ser C - Zool Bot 9(5): 3-22 Burton M (1940) Las esponjas marinas del Museo Argentino de Ciencias Naturales. An Mus Argent Cienc Nat Bernardino Rivadavia 40: 95-121 Calcinai B, Pansini M (2000) Four new demosponge species from Terra Nova Bay (Ross Sea, Antarctica). Zoosystema 22(2): 369381 Cattaneo-Vietti R, Bavestrello G, Cerrano C, Gaino E, Mazzella L, Pansini M, Sar M (1999) The role of sponges in the Terra Nova Bay ecosystem. In: Faranda F (ed). Benthic ecology of Terra Nova bay. Springer-Verlag, Berlin. pp. 539-549 Desqueyroux R (1972) Gubki (Porifera) sobranie u Iuynix Shetlandskix Ostrovof y Antarktichescovo poluostrovo. Issledovania Fauni Morei 11(19): 49-55 Desqueyroux R (1975). Esponjas (Porifera) de la region antartica chilena. Cah Biol Mar 16: 47-82 Desqueyroux R, Moyano H (1987) Zoogeografia de demospongias chilenas. Bol Soc Biol Concepcin 58: 39-66 Desqueyroux-Fandez R (1989) Demospongiae (Porifera) del litoral chileno antrtico. Ser Cient Inst Antrt Chil 39: 97-158 Desqueyroux-Fandez R, Valentine C (2002). Family Niphatidae. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges. Kluwer Academic, New York. pp. 874-889

Gutt J, Koltun WM (1995) Sponges of the Lazarev and Weddell Sea, Antarctica: explanations for their patchy occurrence. Antarct Sci 7 (3): 227-234 Hentschel E (1914) Monaxone Kieselschwmme und Hornschwmme der Deutschen Sdpolar-Expedition 1901-1903. Deutsche Sdpolar-Expedition, 1901-1903 15(1): 35-141 Kirkpatrick R (1907) Preliminary report on the Monaxonellida of the National Antarctic Expedition. Ann Mag Nat Hist (7) 20(117): 271-291 Kirkpatrick R (1908) Porifera (Sponges). II. Tetraxonida, Dendy. Nat Antarct Exped 1901-1904 4: 1-56 Koltun WM (1964) Sponges of the Antarctic. I. Tetraxonida and Cornacuspongida. Biological Reports of the Soviet Antarctic Expedition (1955-1958). Issledovania Fauni Morei 2: 6-116 Koltun WM (1976) Porifera - Part I: Antarctic Sponges. B.A.N.Z. Antarct Res Exped Rep - Ser B Zool Bot 9(4): 147-198 Mothes-de-Moraes B (1978) Esponjas tetraxonidas do litoral sulbrasileiro: II. Material coletado pelo N/Oc. Prof. W. Besnard durante o programa RS. Bol Inst Oceanogr 27(2): 57-78 Mothes B, Lerner CB (1995) Ectyonanchora ruthae sp. n. (Myxillidae) e outras esponjas detectadas na 1 Expedio Antrtica Brasileira (Porifera: Hexactinellida e Demospongiae). Biocincias 3(1): 155-171 Mothes B, Silva CMM (2002) Stelletta ruetzleri sp. nov., a new ancorinid from the Southwestern Atlantic (Porifera, Astrophorida). Sci Mar 66(1): 69-75 Mothes B, Campos MA, Lerner CB, Ferreira-Correia, MM (2004) Esponjas (Demospongiae, Halichondrida) da costa do Maranho, Brasil. Iheringia, Sr Zool 94(2): 149-154 Pansini M, Calcinai B, Cattaneo-Vietti R, Sar M (1994) Demosponges from Terra Nova bay (Ross Sea, Antarctica): 1987/88 and 1989/89 P.N.R.A. Expeditions. Nat Sci Com Antarct 3: 67-100 Ridley SO (1881) XI. Spongida. Horny and siliceous sponges of Magellan Straits, S.W. Chili, and Atlantic off S.W. Brazil. Proc Zool Soc London, 1881: 107-137 Ridley SO, Dendy A (1886) Preliminary report on the Monaxonida collected by H.M.S. Challenger. Ann Mag Nat Hist (5) 18(107): 325-351, 470-493 Ridley SO, Dendy A (1887) Report on the Monaxonida collected by H.M.S. Challenger during the years 1873-1876. Rep Sci Res Voy H.M.S. Challenger, Zool 20: 1-279 Ros P, Cristobo FJ, Urgorri V (2004) Poecilosclerida (Porifera, Demospongiae) collected by the spanish antarctic expedition BENTART-94. Cah Biol Mar 45: 97-119 Sar M (1978) Demospongiae di acque superficiale della Terra del Fuoco (Spedizioni AMF Mares - GRSTS e SAI). Bol Mus Ist Biol Univ Genova 46: 7-117 Sar M, Balduzzi A, Barbieri M, Bavestrello G, Bianchi CN, Boero F, Cattaneo-Vietti R, Corriero G, Morri C, Pansini M (1990) Hard bottom zoobenthos: an analysis of its composition, distribution and of the adaptative strategies of the species. Nat Sci Com Antarct 2: 249-260 Sar M, Balduzzi A, Barbieri M, Bavestrello G, Burlando B (1992) Biogeographic traits and checklist of Antarctic demosponges. Polar Biol 12: 559-585

232

Sollas WJ (1888) Report on the Tatractinellida collected by H.M.S. Challenger during the years 1873-1876. Rep Sci Res Voy H.M.S. Challenger, Zool 25: 1-458 Thiele J (1905) Die Kiesel- und Hornschwmme der Sammlung Plate. Zool Jahrb 6: 407-496 Topsent E (1901) Spongiaires. Rs Voy S.Y. Belgica 1897-1898-189, 6-9: 1-54 Topsent E (1908) Spongiaires. Exp Antarct Fr (1903-1905) comm Dr. Jean Charcot 4: 1-37 Topsent E (1913) Spongiaires de lExpdition Antarctique Nationale cossaise. Trans Roy Soci Edinburgh 49(3): 579-643

Topsent E (1916) Diagnoses dponges recueillies dans lAntarctique par le Pourquoi-Pas? Bull Mus Nat Hist Nat (1) 22(3): 163-172 Topsent E (1917) Spongiaires. In: Joubin L (ed). Deuxime Expdition Antarctique Franaise (1908-1910) commande par le Dr. Jean Charcot. Sciences Physiques: Documents Scientifiques (Paris). 4. Masson & Cie, Paris. pp. 1-88 Vacelet J, Arnaud F (1972) Invertbrs marins des XIIme et XVme expditions antarctiques franaises en Terre Adlie. 2. Dmosponges. Tethys Suppl 4: 9-24 von Lendenfeld R (1907) Tetraxonia der Deutschen SdpolarExpedition 1901-1903. Deutsche Sdpolar-Expedition, 1901-1903 9(1): 303-342

Porifera research: Biodiversity, innovation and sustainaBility - 2007

233

Demospongiae (Porifera) of the shallow coral reefs of Macei, Alagoas State, Brazil
Victor Ribeiro Cedro(1), Eduardo Hajdu(2), Hilda Helena Sovierzosky(1), Monica Dorigo Correia(1*)
Setor de Comunidades Bentnicas, Instituto de Cincias Biolgicas e da Sade, Universidade Federal de Alagoas. vrcedro@gmail.com, mdc@fapeal.br (2) Museu Nacional, Departamento de Invertebrados, Universidade Federal do Rio de Janeiro. hajdu@acd.ufrj.br
(1)

Abstract: Sponges occurring at Alagoas State reefs, in north-eastern Brazil, are poorly known. This work reports on the sponges of Macei coral reefs, the states capital. A total of 29 species were identified so far. These were Agelas dispar, Amphimedon aff. complanata, A. viridis, Biemna microacanthosigma, Chalinula molitba, Chondrilla aff. nucula, Chondrosia collectrix, Cinachyrella apion, C. alloclada, Cliona aff. celata, C. varians, Dragmacidon reticulatum, Dysidea etheria, Echinodictyum dendroides, Geodia corticostylifera, G. papyracea, Haliclona curacaoensis, H. manglaris, H. melana, Iotrochota birotulata, Ircinia strobilina, Mycale diversisigmata, Niphates erecta, Placospongia aff. melobesioides, Scopalina ruetzleri, Spirastrella coccinea, S. hartmani, Tedania ignis and Tethya sp. The numbers of species in each station varied from 16 to 24. Mycale diversisigmata is a first record for the southern Atlantic. These preliminary results indicate the occurrence of a moderately rich sponge fauna at the Macei reefs area, arguing for stricter control on human impact on these urban reefs. Keywords: Alagoas, Demospongiae, coral reef, SW Atlantic

Introduction
Sponges are among the most important components of coral reef benthic communities, frequently exceeding hermatypic corals in diversity and biomass. Each species abundance and distribution within a coral reef sponge community is related to biotic and abiotic parameters such as recruitment, spatial competition, luminosity, sedimentation and substrate type (Wiedenmayer 1977, Zea 1987, Daz and Rtzler 2001, Rtzler 2002, Valderrama and Zea 2003). Brazil has a large coastline, with a faunistic component which is still poorly studied in general, and Porifera stands out as one of the least studied taxa. Very few coastal states in Brazil have a well-known sponge fauna (e.g. Hajdu et al. 1996, 1999, Muricy and Silva 1999), and comprehensive knowledge of species distributions, habitat requirements, reproduction and all sorts of ecological interactions are virtually unknown for the entire coast. The taxonomy of marine sponges of north-eastern Brazil (geographic north-east coastal states Bahia to Cear; not regional, geopolitical north-east coastal states Bahia to Maranho) has been the focus of a very few studies, and even fewer described species from Alagoas, mostly from deep-waters. The latter date back to the H.M.S. Challenger expedition, three collecting stations having been set off the state of Alagoas (st. 122b, 122c, 123). The species reported from this material were Cacospongia levis (Poljaeff, 1884) and Stelospongus longispinus Duchassaing and Michelotti, 1864 [= Ircinia strobilina (Lamarck, 1816)], reported by Poljaeff (1884); Pheronema carpenteri Schulze, 1887, reported by Schulze (1887); and Geodia neptuni (Sollas, 1886)

[= Geodia vosmaeri (Sollas, 1886)] and Thenea fenestrata (Schmidt, 1880), reported by Sollas (1888). Further studies reporting on sponges of the north-eastern Brazilian coast were those of de Laubenfels (1956; without descriptions), Johnson (1971; essentially based on beach worn material), Boury-Esnault (1973), Hechtel (1976; without descriptions, 1983) and Sarmento and Correia (2002; without descriptions), among others. The high diversity of sponges on the Brazilian north-eastern coast has been best illustrated by Boury-Esnault (1973), Muricy and Moraes (1998) and Muricy et al. (2006), and of north-eastern Brazilian oceanic islands by Mothes and Bastian (1993) and Moraes et al. (2006). Sarmento and Correia (2002) reported the finding of 17 species of demosponges on Ponta Verde coral reef (Macei, AL), and correlated these to different habitat types. The present study reports on sponge assemblages found in four shallow urban reef areas of Macei (Alagoas State, north-eastern Brazil).

Materials and methods

Four sampling areas were chosen: (1) Pajuara reef (94100.27S / 354319.78W; Fig. 1A, 2A), (2) Piscina dos Amores (94009.24S / 354214.16W; Fig. 1B, 2B), (3) Ponta Verde (93956.87S / 354144.98W; Fig. 1C, 2C) and (4) Jatica reef (93914.56S / 354150.05W; Fig. 1D, 2D), in a SW - NE sequence. Distances were 2300 m between (1) and (2), 2000 m between (2) and (3) and 1200 m between (3) and (4). Collections were made in the years 2004 and 2005 by wading at low tide and snorkelling. Live specimens were observed underwater, and whenever possible photographed. Each specimen was individually

234

Results and discussion

A total of twenty-nine sponge species were identified (Table 1), comprising ten orders and nineteen families. The Haplosclerida, with seven species, was the richest order, followed by the Hadromerida, with six, and the Poecilosclerida, with five. The numbers of species found in each station varied from 16 in the Piscina dos Amores to 24 at Ponta Verde reef. This difference is likely to reflect the more extensive sampling at Ponta Verde, an easily accessible reef, coupled with its possession of a large sciaphilic and slightly eutrophic environment, where sponges constitute the dominant benthic taxon. Amphimedon viridis (Fig. 2E) was the sole species observed to be very common on all four sampled areas, thus confirming once more its abundance along the Brazilian coastline. The species has been previously reported as common along the south-eastern Brazilian coast (Muricy et al. 1991, Hajdu et al. 1999). Cinachyrella alloclada (Fig. 2F) was very common on three areas, and Chondrilla aff. nucula, Cliona aff. celata, C. varians, Haliclona manglaris (Fig. 2H), H. melana, Tedania ignis (Fig. 2G) and Tethya sp. on two stations only. Of these, all but H. manglaris were previously known to be common on distinct sectors of the Brazilian coastline (Muricy et al. 1991, Klautau et al. 1999, Muricy and Ribeiro 1999, Lazoski et al. 2001). Cinachyrella apion, Iotrochota birotulata and Spirastrella coccinea were all rare and each species was found in a single station. Mycale diversisigmata is a first record for the southern Atlantic. Tethya sp. is probably a new species and needs a formal description. Twenty-one out of the 28 species found are distributed in the Caribbean area too (e.g. Pulitzer-Finali 1986, van Soest et al. 2007), which indicates a marked Tropical western Atlantic affinity of the Macei reefs sponge fauna. Only three species found are provisional Brazilian endemics, viz. Echinodictyum dendroides, Biemna microacanthosigma and the likely new species of Tethya. Four species have widespread disjunct distributions and entitle for a morphogenetic revision study (Chondrilla aff. nucula, Chondrosia collectrix, Cliona aff. celata, Placospongia aff. melobesioides). Only 12 species reported here were listed for the neighbouring state of Pernambuco by Muricy and Moraes (1998), viz. Agelas dispar, Amphimedon viridis, Chondrilla aff. nucula (as Chodrilla nucula), Chondrosia collectrix, Cliona varians (as Anthosigmella varians), Dragmacidon reticulatum (as Pseudaxinella reticulata), Geodia corticostylifera, Geodia papyracea, Ircinia strobilina, Scopalina ruetzleri, Spirastrella coccinea and Tedania ignis. The similarity has not been found to be higher, probably because the reef strata sampled were

Fig. 1: Map showing the location of Macei (north-eastern Brazil) and the collecting stations analysed in this study. A. Pajuara coral reefs. B. Piscina dos Amores coral reef. C. Ponta Verde coral reef. D. Jatica coral reef. Scales: 400 km in map, 10 km in inserts. Maceis urban area is shaded.

packed and fixed in ethanol 70% soon afterwards. Some specimens collected earlier (2001) and deposited in the MNRJ (Museu Nacional/Universidade Federal do Rio de Janeiro Porifera collection) and UFALPOR (Setor de Comunidades Bentnicas/Universidade Federal de Alagoas Porifera collection) collections were also analyzed. Voucher specimens for each species studied are listed in Table 1, although not for every locality sampled. A complete list of localities sampled and collected specimens is available from the authors upon request. A semi-quantitative estimation of sponge abundance has been applied to each station considered, and consisted of the exhaustive compilation of visual records achieved on at least five full low-tide periods (between 2001 and 2006). These were written down immediately after leaving the sampling grounds when incoming rising tide forced a complete stop in the observations. Sponges were classified into dominant, common and rare. A fourth category, absent, has been added after comparison of data compiled from the four stations, and includes only those species found in at least one station. Specimens were identified by field observations and morphometry of the skeletal architecture. Microscopic slides of dissociated spicules and thick sections were made following the standard methods described by Hooper and van Soest (2002). A few specimens were further studied through scanning electron microscopy.

Fig. 2: Aerial views of the collecting localities, showing the reefs at low or nearly low tide, and some of the commonest species of demosponges found. A. Pajuara coral reefs. B. Piscina dos Amores coral reef. C. Ponta Verde coral reef. D. Jatica coral reef. E-H. Ponta Verde scyaphilic habitat. E. Amphimedon viridis. F. Cinachyrella alloclada (predominantly). G. Tedania ignis. H. Haliclona manglaris. Scales: 2 cm (E, G, H) and 4 cm (F).

235

236

Table 1: Occurrence and semi-quantitative estimation of abundance of the Macei reefs sponge fauna. Legend: (Dominant); (Common); (Rare); (Absent). Pajuara 17 Coral reefs Piscina dos Ponta Verde Amores 16 24 Jatica 24

Taxa Agelas dispar Duchassaing & Michelotti, 1864 (MNRJ 10294) - Agelasida Amphimedon viridis Duchassaing & Michelotti, 1864 (MNRJ 9033 ; UFAL/ POR 0027) - Haplosclerida Amphimedon aff. complanata Duchassaing, 1850 (MNRJ 4713) - Haplosclerida Biemna microacanthosigma Mothes, Hajdu, Lerner & van Soest, 2004 (UFAL/ POR 0043) - Poecilosclerida Chalinula molitba (de Laubenfels, 1949) (UFAL/POR 0042) - Haplosclerida Chondrilla aff. nucula Schmidt, 1862 (MNRJ 3146; UFAL/POR 0049) Chondrosida Chondrosia collectrix (Schmidt, 1870) (MNRJ 10279) - Chondrosida Cinachyrella alloclada (Uliczka, 1929) (MNRJ 10292; UFAL/POR 0004) Spirophorida Cinachyrella apion (Uliczka, 1929) (MNRJ 10290) - Spirophorida Cliona aff. celata Grant, 1826 (MNRJ 4650) - Hadromerida Cliona varians (Duchassaing & Michelotti, 1864) (MNRJ 3151) - Hadromerida Dragmacidon reticulatum (Ridley & Dendy, 1886) (MNRJ 10283;UFAL/POR 0053) - Halichondrida Dysidea etheria de Laubenfels, 1936 (UFAL/POR 0021) - Dictyoceratida Echinodictyum dendroides Hechtel, 1983 (MNRJ 4711; UFAL/POR 0025) Poecilosclerida Geodia corticostilyfera Hajdu, Muricy, Custdio, Russo & Peixinho, 2002 (MNRJ 10274) - Astrophorida Geodia papyracea Hechtel, 1965 (MNRJ 10285;UFAL/POR 0046) Astrophorida Haliclona curacaoensis van Soest, 1980 (MNRJ 10280; UFAL/POR 0040) - Haplosclerida Haliclona manglaris Alcolado, 1984 (MNRJ 10289) - Haplosclerida Haliclona melana Muricy & Ribeiro, 1999 (MNRJ 10277; UFAL/POR 0041) - Haplosclerida Iotrochota birotulata (Higgin, 1877) (MNRJ 10291) - Poeciloclerida Ircinia strobilina Lamarck, 1816 (MNRJ 4717; UFAL/POR 0017) Dictyoceratida Mycale diversisigmata van Soest, 1984 (MNRJ 4639) - Poecilosclerida Niphates erecta Duchassaing. & Michelotti, 1864 (MNRJ 10287) Haplosclerida Placospongia aff. melobesioides Gray, 1867 (MNRJ 4724) - Hadromerida Scopalina ruetzleri (Wiedenmayer, 1977) (MNRJ 10281) - Halichondrida Spirastrella coccinea (Duchassaing & Michelotti, 1864) (MNRJ 4629) Hadromerida Spirastrella hartmani Boury-Esnault, Klautau, Bzac, Wulff & Sol-Cava, 1999 (MNRJ 10286; UFAL/POR 0002) - Hadromerida Tedania ignis (Duchassaing. & Michelotti, 1864) (MNRJ 10293; UFAL/POR 0031) - Poecilosclerida Tethya sp. (MNRJ 4712) - Hadromerida Total of species

different. Muricy and Moraes (1998) focused mainly on the subtidal, while the data presented here were gathered in the intertidal. No sponge species has been hitherto recorded from the coast of Sergipe State, the adjacent state on the south, which is a clear sampling gap. On the other hand, most of the

species reported upon here also occur on Brazilian oceanic islands, as well as in the State of Bahia, further south (Moraes et al. 2006, Peixinho et al. unpubl. res.). These results, albeit preliminary, indicate the occurrence of a moderately rich demosponge assemblage in the Macei

237

intertidal coral reef areas. The new records found on these shallow urban reefs demonstrate the lack of comprehensive faunistic and taxonomic surveys of the benthic fauna of Alagoas coral reefs. Macei has a population quickly approaching 106 citizens. Findings presented here advise for stricter control on human impact on these shallow water urban reefs, some of which are usually visited at low tide by hundreds of people (local collectors of sea-food and tourists). The collected data increase significantly the knowledge concerning the sponge richness of Alagoas state coral reefs.

Acknowledgements
Authors are grateful to Alvaro Borba Junior (LABMAR/UFAL) for helping in the field trips. Marcia Atthias and Noemia Gonalves (Instituto de Biofsica Carlos Chagas Filho/UFRJ) kindly provided access to, and technical assistance on SEM operation. CNPq, FAPEAL and FAPERJ are thanked for the provision of grants and/or fellowships.

References
Boury-Esnault N (1973) Campagne de la Calypso au large des ctes atlantiques de lAmrique du Sud (1961-1962). I, 29. Spongiaires. Rs Scient Camp. Calypso, Paris, 10: 263-295 de Laubenfels, MW (1956) Preliminary discussion of the sponges of Brazil. Contr Avulsas Inst Oceanogr Univ So Paulo Oceanogr Biol 1: 1-4 Daz MC, Rtzler K (2001) Sponges: an essential component of Caribbean coral reefs. Bull Mar Sci 69(2): 535-546 Hajdu E, Berlinck RGS, Freitas JC de (1999) Porifera. In: Migotto A, Tiago CG (eds). Biodiversidade do Estado de So Paulo: sntese do conhecimento ao final do sculo XX (ser. eds. Joly CA, Bicudo CEM), vol. 3. Invertebrados Marinhos. Fapesp, So Paulo. pp. 20-30 Hajdu E, Muricy G, Berlinck RGS, Freitas JC (1996) Marine poriferan diversity in Brazil: through knowledge to management. In: Bicudo CE, Menezes NA (eds) Biodiversity in Brazil: a first approach. So Paulo: Conselho Nacional de Desenvolvimento Cientfico e Tecnolgico. pp. 157-172 Hechtel GJ (1976) Zoogeography of Brazilian marine Demospongiae. In: Harrison FW, Cowden RR (eds), Aspects of sponge biology. Academic Press, New York. pp. 237260 Hechtel GJ (1983) New species of marine Demospongiae from Brazil. Iheringia, Zool, 63: 58-89 Hooper JNA, van Soest RWM (eds.) (2002). Systema Porifera: a guide to the classification of sponges, vol. 1. Kluwer Academic/ Plenum Publishers, New York Johnson MF (1971) Some marine sponges of northeast Brazil. Arq Cienc Mar 11(2): 103-116 Klautau M, Russo CAM, Lazoski C, Boury-Esnault N, Thorpe J, Sol-Cava A (1999) Does cosmopolitanism result from overconservative systematics? A case study using the marine sponge Chondrilla nucula. Evolution 53: 1414-1422

Lazoski C, Sol-Cava AM, Boury-Esnault, N, Klautau M, Russo CAM (2001) Cryptic speciation in a high gene flow scenario in the oviparous marine sponge Chondrosia reniformis. Mar Biol 139: 421-429 Moraes FC, Oliveira MV, Klautau M, Hajdu E, Muricy G (2006) Biodiversidade de esponjas das ilhas ocenicas brasileiras. In: Alves RJV, de Alencar Castro JW. (Orgs.). Ilhas Ocenicas Brasileiras, da Pesquisa ao Manejo. Braslia: Ministrio do Meio Ambiente. pp. 147-177 Mothes B, Bastian MCKA (1993) Esponjas do Arquiplago de Fernando de Noronha (Porifera, Demospongiae). Iheringia Sr Zool 75: 15-31 Muricy G, Hajdu E, Custdio MR, Klautau M, Russo C, Peixinho S (1991) Sponge distribution at Arraial do Cabo, SE. Brazil. In: Magoon OT, Converse H, Tippie V, Tobie LT, Clark D (eds). Proc 7th Symp Coast Ocean Manag (Long Beach, USA). ASCE Publ. 2: 1183-1196 Muricy G, Moraes FC (1998) Marine sponges of Pernambuco State, NE Brazil. Rev Bras Oceanogr 46(2): 213-217 Muricy G, Ribeiro SM (1999). Shallow-water Haplosclerida (Porifera, Demospongiae) from Rio de Janeiro state, Brazil (Southwestern Atlantic). Beaufortia, 49(6): 47-60 Muricy G, Silva OC (1999) Esponjas marinhas do Estado do Rio de Janeiro: um recurso renovvel inexplorado. In: Silva SHG, Lavrado HP (eds). Ecologia dos Ambientes Costeiros do Estado do Rio de Janeiro. Sr Oecol Bras, vol. 2. PPGE-UFRJ. pp. 155-178 Poljaeff N (1884) Report on the Keratosa collected by H.M.S. Challenger during the years 18731876. Rep Sci Res Voy H.M.S. Challenger, Zool 11: 1-88 Pulitzer-Finali G (1986) A collection of West Indian Demospongiae (Porifera). In appendix, a list of the Demospongiae hitherto recorded from the West Indies. Ann Mus civ sto nat Giacomo Doria 86: 65-216 Rtzler K (2002) Impact of crustose sponges on Caribbean reef corals. Acta Geo Hisp 31(1): 61-72 Sarmento FJQ, Correia MD (2002) Descrio de parmetros ecolgicos e morfolgicos externos dos porferos no recife de coral da Ponta Verde, Macei, Alagoas, Brasil. Rev Bras Zooc 4(1): 215-226 Schulze FE (1887) Report on the Hexactinellida collected by H.M.S. Challenger during the years 18731876. Rep Sci Res Voy H.M.S. Challenger, Zool 21: 1-514 Sollas WJ (1888) Report on the Tetractinellida collected by H.M.S. Challenger, during the years 18731876. Rep Sci Res Voy H.M.S. Challenger, Zool 25(63): 1-458 Valderrama D, Zea S (2003) Esquemas de distribuicin de esponjas arrecifales (Porifera) del Noroccidente del Golfo de Urab, Caribe Sur, Colmbia. Bol Invest Mar Cost 32: 37-56 van Soest RWM, Boury-Esnault N, Janussen D, Hooper J (2007) World Porifera database. http://www.marinespecies.org/porifera/. Accessed on 2007-06-23 Wiedenmayer F (1977) Shallow-water sponges of the western Bahamas. Experientia Suppl 28: 1-287 Zea S (1987) Esponjas del Caribe Colombiano. Catlogo Cientifico, Bogot

Porifera research: Biodiversity, innovation and sustainaBility - 2007

239

How and why do sponges incorporate foreign material? Strategies in Porifera

Carlo Cerrano(1*), Barbara Calcinai(2), Cristina Gioia Di Camillo(2), Laura Valisano(1), Giorgio Bavestrello(2)
Dipartimento per lo studio del Territorio e delle sue Risorse, C.so Europa, 26, 16132, Genova, Italy. cerrano@dipteris.unige.it, valisano@dipteris.unige.it (2) Dipartimento di Scienze del Mare, Via Brecce Bianche, 60131, Ancona, Italy. b.calcinai@univpm.it, g.bavestrello@univpm.it, c.dicamillo@univpm.it
(1)

Abstract: The selection and incorporation of foreign materials in sponges is a complex phenomenon: it involves both a system of recognition of pinacocytes versus sand grain mineralogy and a system of coordination among cells, which transport and engulf particles in specific areas of the sponge surface. Concerning the mineralogical characteristic of the incorporated particles, it seems that quartz particles, when incorporated, could play an important role in collagen production. Among incorporating species, two different modalities can be defined, depending on the habit of the species: i) soft-bottom species (e.g. genera Oceanapia, Tectitethya, Cliona) engulf particles mainly from the base of their body and select mainly the size of particles independently from their mineralogical characteristics; engulfed particles, due to their weight, help the sponge to stabilize and to anchor to the soft substrate; ii) in hard-bottom species (e.g. genera Chondrosia and Ircinia) ectosome pinacocytes select particles, in relation to their size and mineralogy, and may incorporate them differently in some areas of their body according to their skeletal arrangement. Keywords: Porifera, selectivity, sediment incorporation, foreign inclusions, mineralogy

Introduction
Marine organisms, particularly in benthic environments, have to coexist with a continuous sediment rain and have adapted to this phenomenon in several ways (Miller et al. 2002). They can react cleaning their surface, more or less actively, or by trying to exploit sediments to feed or to build protective and/or structural elements. Several organisms like protists (Takahashi and Ling 1984), sponges (Teragawa 1986, Cerrano et al. 1999a), cnidarians (Haywick and Mueller 1997) annelids (Wilson 1974, Main and Nelson 1988), molluscs (Min-Da 1984), crustaceans (Dixon and Moore 1997, Krasnow and Taghon 1997), echinoderms (Massin and Doumen 1986), and tunicates (Kott 2006) are able to use foreign material as a cover, to protect or mask their body, building thecae, coats, tubes or other structures. Among Porifera and Cnidaria there are examples of species able to incorporate particles into their body. Its generally assumed that this strategy is performed to strengthen the skeleton but the real meaning of it and the related mechanisms are often unclear (Teragawa 1986). One of the most debated problems regarding this phenomenon is if organisms are able to select foreign bodies or if they utilise every kind of particle available in the surrounding environment. Generally this behaviour is regulated by the ability of organisms to handle particles so that a physical limit related to the particle size has to be always considered. The most intriguing aspect is the ability

of some species to recognise the mineral characteristics of the particles and therefore to select them (Bavestrello et al. 1996). The aim of this paper is to review the incorporation of foreign bodies in sponges, comparing the strategies of species living on soft and hard substrates and suggesting possible physical and biological explanations for this intriguing behaviour.

Soft-bottom sponges
Even if sponges typically live on hard substrates, there are several species more or less adapted to soft bottom environments. These may live loose on the sediment often partially or completely buried in it and survive well thanks to some very special adaptations, which limit sponge rolling and occlusion of aquiferous system by sand. On soft substrates it is possible to observe sponge fragments which occasionally may fall from coral or rocky reefs due to the production of asexual reproductive bodies and/or fragments, breakage during storms, localised infections by pathogens, or predator bites (Wulff 1985, Battershill and Bergquist 1990). The survival of unattached fragments depends on their ability to re-settle in a short span of time to avoid clogging by sediments (Ilan and Abelson 1995). For these fragments the incorporation of large amount of foreign bodies is crucial to assume a gravimetric polarity which allows them to stabilize and reorganize their aquiferous system.

240

We can classify soft bottom sponges into three main groups according to their adaptative strategies: i) sponges living on the sediment surface, ii) sponges partially buried, iii) sponges with the body completely buried, with particular anatomic adaptations.

Sponges on the sediment surface

Here we consider sponges that do not live exclusively on soft substrates but that can easily survive on soft substrates, fully regaining their vital functions. These non-sessile specimens have been generally found in shallow sub-littoral environments (Mercurio et al. 2006, Bell and Barnes 2002) and in the deep sea (Barthel and Tendal 1993). Examples of this habit can be found in lagoon environments (Ise et al. 2004, Mercurio et al. 2006), where several sponge fragments, often from species living typically on hard substrates, can be present. In studies performed in the Caribbean (Cerrano et al. 2004) and the Indonesian (Cerrano et al. 2002) lagoons, the comparison between environmental sediments and the particles incorporated by several sponge species shows that they mainly contain the fraction larger than 5 mm. Only a few species use the fractions available in the surrounding substrates without size preferences. The percentage of incorporated sediments can be highly variable, between 5 and 99% per sponge dry weight. A particular case concerns the gamma stage of Cliona nigricans, an excavating Atlanto-Mediterranean sponge living symbiotically with zooxanthellae that can grow with different shapes: endolithic, into coralligenous accretions, and massive, laid on detritic sediments (Fig. 1A). This species can engulf from the base (Fig. 1B) huge amounts of foreign material, up to 99% of its dry weight, being also able to store the fraction of sediment larger than 5 mm (Calcinai et al. 1999). Moreover, experimental data indicated that in this species the mineralogical features of the engulfed particles can affect morphogenetic processes, in particular quartz negatively affects the growth of C. nigricans specimens limiting the development of the oscula in the basal portion of the sponge that is in direct contact with the grains. On the contrary, oscula have been observed in specimens living on calcareous sand (Cerrano et al. 2007), highlighting once again the importance of substrate chemical composition on benthic organism distribution and development (Cerrano et al. 1999b, Bavestrello et al. 2003). In massive specimens of C. nigricans, the aquiferous system opens on the sediment using a water expulsion mechanisms similar to the one described for Spheciospongia cuspidifera in Belize (Rtzler 1997).

This species incorporates all the granulometric size classes of nearby benthic sediments, using them in different ways. In the choanosome, sediments are sorted and distributed according to their size: fine sediments (40-60 m) are densely aggregated in the choanosome, whereas coarse particles are more evenly distributed in the lower portion of the body were they contribute to the stability of the sponge (Fig. 1C). Qualitatively, the choanosomal aggregations of fine sediment contain more siliceous material than the ambient sediment of the same size class. Microscopical analysis of the particles shows that this species selects and incorporates allocthonous sponge spicules, radiolarians and diatoms (Cerrano et al. 2004). Another interesting species is Biemna fortis, living in tropical lagoons in North Sulawesi (Indonesia). This sponge displays two different growth patterns depending on the thickness of unconsolidated sediments: when the sediment layer is thick, the sponge assumes a cylindrical form and incorporation is low; when there is a thin sediment layer the sponge adheres to the basal coral rock, developing a massive buried portion that is generally rich in embedded particles (Cerrano et al. 2002).

Sponges specialised to psammobiontic habit

All the known species of the genus Oceanapia live on soft substrates thanks to the ability of producing long fistules that anchor the sponge body to the loose substrate and discharge waste-water deep into the sediments (Werding and Sanchez 1991, Bavestrello et al. 2002). The specialisation of this genus to soft substrates is evidenced also by the differential production of secondary metabolites used as antipredatory that are synthesised exclusively in the exposed portions in O. sagittaria, suggesting that sediments are not just a mere substrate where sponges can live with low competition but also a refuge from potential predators (Schupp et al. 1999, Salomon et al. 2001). In lagoons O. amboinensis lives buried in unconsolidated sediments among sea grasses. The sponge develops a massive body and emerges from the sediment through numerous closed fistules. The sponge body is whitish, while the portions protruding from the sediment take an olive green colour. The buried portion of the body incorporates a high quantity of foreign materials, selecting particles larger than 2 mm, throughout the pinacoderm. Only exhalant areas, of 1-4 cm2, do not participate in this process (Cerrano et al. 2002, Bavestrello et al. 2002). Oceanapia fistulosa lives from 15-20 m depth down to at least 80 m, grows partially buried in detritic sediment (Fig. 1D). The globular sponge body, 5-15 cm in diameter, bears on its upper side several closed cylindrical fistules that emerge from the sediment generally covered by epibionts. On the other side, other buried closed fistules are strongly rooted in the sediments. The buried portion of the sponge incorporates a lot of foreign material such as sand, coral and shell fragments, particularly on the rooted fistules, which can reach a length of 15-20 cm and 1 cm in diameter, depending on the thickness and the granulometry of the unconsolidated sediments. In fine sediments this species, to

Sponges partially buried in sediments

Several species can live on soft substrates even without morphological adaptations to this environment. Tectitethya crypta is a massive, shallow-water sponge common in the Caribbean and frequently covered by a sediment and/ or algal coat, both on hard and soft bottoms. In lagoon environments this sponge can occur either loose or anchored, significantly varying its morphology (Cerrano et al. 2004).

241

Fig. 1: Examples from soft-bottoms sponges. A. Specimen of Cliona nigricans living on detritic substrates. B. Detail of the lower face of the sponge with several rocks having highly variable sizes. The fraction bigger than 5 mm is more abundant in the sponge than in the ambient sediments. Arrows indicate oscular openings. C. Half cut specimen of Tectitethya crypta. White arrows indicate aggregations of fine sediments, black arrow indicates coarse sediments. D. Drawing of Oceanapia fistulosa with the buried body mass covered by sand grains.

242

get stabilization, produces more rooted fistules, smaller in diameter but longer than those in coarse sediments.

Hard-bottom sponges
Burial/smothering, scour/abrasion, and changes in the physical characteristics of the substrate surface are the three main mechanisms by which sediments may affect benthic assemblages (Airoldi 2003). On hard substrates sedimentation is partly due to particles suspended in the water column and partly due to the detritus that rolls down vertical cliffs (Bavestrello et al. 1995a). This may cause mechanical damages in sponges and other benthic filter feeders, especially by clogging the aquiferous system impeding filtration. A solution to avoid sedimentation is generally represented by the colonisation of substrates under overhangs, but in this situation sponges have to compete with many other sessile animals that share the same strategy. Other organisms choose to grow vertically, limiting the surface available for sediments as happens for several species of the genera Axinella or Dysidea (Fig. 2A). Other species can clean their pinacoderm using superficial cellular movements (Bond 1992) that can easily either remove or take up several kinds of particles transforming the problem of sediments into an opportunity to providing a physical support to the skeletal development (Fig. 2B). According to Teragawa (1986) the sediment that settles on the surface of Dysidea etheria may follow different pathways being i) inhalated through ostia, ii) eliminated by transport or through dermal membrane oscillations or mucus sloughing, iii) incorporated into primary fibres, and iv) engulfed into secondary fibres in case the sponge is overloaded by sediments. When sediments are incorporated into spongin fibres it is possible to consider this localization as definitive but, on the contrary, a turnover was described for the sediments engulfed in the cortex of Chondrosia reniformis (Cerrano et al. 1999a). This species, presenting a collagenous structure and lacking its own spicules and spongin fibres, when anchored to a substrate, is able to incorporate foreign material, discerning from crystalline quartz sand grains and amorphous siliceous opaline spicules (Bavestrello et al. 1998a, 1998b). Laboratory experiments have shown that the cells of the sponge ectosome play a key role in the selection processes: quartz particles are incorporated while carbonatic particles are agglutinated and drop out from the sponge ectosome. In C. reniformis specimens anchored to the substrate, the upper ectosome can distinguish between silica and carbonates, ability lost in free, non-attached individuals, which incorporate both. This behavior indicates that specific receptors are present and can distinguish among the different mineralogical features of the sediment. Depending on the environmental conditions this mechanism can be switched on or off (Bavestrello et al. 1998b). The turnover of particles inside the body of C. reniformis is due to the ability of this sponge to dissolve quartz crystals releasing silicate (Bavestrello et al. 1995b). In C. reniformis the amount of incorporated sediment was used as a character to separate different species (Wiedenmayer 1977) while in dictyoceratid sponges the presence of sediment

in fibres or as a dermal crust is considered as a character to distinguish between different genera (Vacelet 1959). Nevertheless, Pronzato et al. (2004) considered the amount of mineral granules a specific character to distinguish Ircinia felix from I. variabilis. The evidence of a specific and fine tuned mechanism to select particles, according to their mineralogical features, suggests that a mineralogical and granulometric analysis of incorporated sediments may represent a tool for the classification of problematic taxonomic groups. The genus Ircinia is characterised by spongin fibres cored with foreign debris. In an unpublished investigation we have compared the foreign bodies incorporated by two sympatric species of Ircinia (I. variabilis and I. retidermata) inhabiting two different areas. Results show that part of the sediment is included into growing fibres, probably definitively, while another portion is incorporated into the choanosomal tissue where it is subjected to a quick turn over. In both species, the material incorporated into fibers and the one engulfed in the choanosomal tissue are different. Ircinia variabilis incorporated sponge spicules and sand grains in the same proportion both in the ectosome and in the mesohyl (here considered as choanosome excluding fibres). Spongin fibres include almost only sand grains (Fig. 2C-F). Ircinia retidermata has a more homogeneous ectosomal coat of quartz grains. The amounts of ectosomal sediments allowed the determination of interspecific differences while choanosomal ones (not considering spongin fibres) are affected by local sedimentation rates, so that differences at intraspecific and interspecific level can be similar and not useful for species classification.

Discussion
Sediment incorporation is a widespread aptitude in sponges and is observed in species belonging to different not-related groups (Fig. 3). On the contrary in cnidarians the incorporation of sediment occurs only in the order Zoanthidaea. With this exception, several other metazoans use foreign bodies to build external protective structures, but no one is able to incorporate foreign bodies in their tissues. In sponges and cnidarians particles are embedded in the collagenous mesohyl/mesoglea, and spongin fibres and their incorporation is mediated by the interaction with dermal cells. In the incorporation processes the most intriguing aspects relate to the ability of selecting the mineral features and the size of the foreign bodies and their transport to definite areas. Sediment selection based on mineral composition is not exclusive of hard-bottom sponges but can occur after stable anchoring also in sponges living on soft bottoms as described in Tectitethya crypta. On hard bottoms, Ircinia retidermata
Fig. 2: Examples from hard-bottoms sponges. A. Dysidea avara specimen in its natural environment. B. Detail of the ectosome with accumulation of sediments on the tip of conules. C. Detail of the ectosome of Ircinia variabilis. D. Drawing of a section of I. variabilis. E. Drawing of a single conule of I. variabilis with detail of a primary fibre and sediment coat. F. Electron micrograph of a conule of I. variabilis.

243

244

Fig. 3: Sponges living both on hard and soft bottoms can incorporate sediment in a selective way or not. The soft bottom sponges that are specialised to live in unconsolidated sediments (ex. Oceanapia spp.) get true stabilization via peculiar morphological adaptations. In this case incorporation can be selective towards particles size. Unanchored sponges incorporate without selection until they stabilize, then their behaviour can become selective. Fixed hard bottom sponges have a selective behaviour both vs. particles mineralogy and/or size. Unselective behaviour is generally related to stabilization, selection towards particles size can be related to stabilization and/or skeletal growth, selection towards mineralogy may be related to some biological need.

selects particles, organising quartz grains with homogeneous size in the ectosome. Although there is some evidence that in Cliona nigricans the incorporated quartz particles negatively affect the sponge growth, the sponge incorporates these particles indiscriminately if they are present in the surrounding sediments. In Chondrosia reniformis the mineral discrimination of the upper ectosome may be switched on by the adhesion of the sponge to the substrate. When attached, the upper side collects quartz and silicates while the lower bottom side specifically engulfs the calcareous particles, thus helping the sponge to attach to the substrate. When unattached, the sponge does not select and incorporates with modalities that resemble those described for soft bottom species, being its priority the stabilisation and a new polarity. This is a puzzling behaviour because it is not easy to understand why a sponge selects and engulfs particles to dissolve them. A possible explanation was suggested by the

evidence that the expression of the gene for collagen was found to be dependent on the silicate concentration (Krasko et al. 2000, Nickel and Brmmer 2003). In this way the induction of collagen production by quartz dissolution may be hypothesised (Bavestrello et al. 2003). In soft-substrates specimens, selection is mainly dimensional. Several sponge species living on soft substrates select from the environmental sediment mainly the larger granulometric fractions. Incorporation happens in two ways: i) pinacocytes may recognise and incorporate only the larger particles fraction; ii) the pinacoderm may engulf sediment of all available size classes and subsequently sort them within the mesohyl, selecting the larger particles and expelling the smaller ones. In both cases a selection mechanism at the cellular level has to be hypothesised. This ability is particularly evident in Tectitethya crypta because of the presence of two even more different ways to handle fine

245

and coarse sediments. Fine sediments are concentrated in the nuclei within the sponge body while coarse grains are moved to the base of the sponge to anchor and stabilize by gravity. The incorporated sediment is used for very different purposes in different species. Soft-bottom sponges generally use foreign bodies to anchor and to gain a gravimetric polarity. This stability allows the sponge to (re)organise its aquiferous system in the most efficient way. Hard-bottom sponges use foreign bodies to reinforce their skeletal structure but this structural use is not the only possible. The case of Chondrosia indicates that this species is able to metabolise quartz, with possible effects on the metabolism of collagen. These data can lead to the hypothesis that in sponges with a skeleton structured by a spongine net the inclusion of particles in the growing fibres could stimulate the production of spongine. Particle selection and handling has to be related to the self/ non-self recognition mechanisms in sponges. Even if sponges lack a specific immune system, several cellular processes can enable discrimination between true symbionts from potential pathogenic microorganisms (Steindler et al. 2007) or develop a sort of primitive short-term immune memory, as evidenced by allografts (Bigger et al. 1982). Recognition mechanisms are modulated by sponge condition, attached or unattached to a substrate (Bavestrello et al. 1998b). Several studies suggest that the allorecognition system may change during ontogeny and this aspect is generally considered in the case of chimeric sponges (Maldonado 1998, McGhee 2006). The fusion among different sponge species in adult phase may help in stabilizing rolling sponges (Cerrano et al. 2004), and could be related to the loss of selectivity evidenced in unanchored specimens. In conclusion the use of sediments depends on the habit of the species and can be selective (when sponges are stable on the substrate) or not (when sponges are not stable). Moreover, the mineralogical composition of particles can affect sponge growth, in particular quartz that, depending on the species, can enhance or limit this process.

Acknowledgements
Authors are indebted with Marzia Sidri (Porifarma, Wageningen) and two anonymous referees for helpful comments. This paper comes from a lecture hold in the framework of the Biologisches Kolloquium Wintersemester at the Universitaet of Stuttgart.

References
Airoldi L (2003) The effects of sedimentation on rocky coast assemblages. Oceanogr Mar Biol Annu Rev 41: 161-236 Barthel D, Tendal OS (1993) The sponge association of the abyssal NorwegianGreenland Sea: species composition, substrate relationships and distribution. Sarsia 78: 83-96 Battershill CN, Bergquist PR (1990) The influence of storms on asexual reproduction, recruitment and survivorship of sponges. In: Rtzler K (ed). New perspectives in sponge biology. Third International Conference on Biology of Sponges, Smithsonian Inst. Press, Washington, D.C. pp. 397-403 Bavestrello G, Cattaneo-Vietti R, Cerrano C, Danovaro R, Fabiano M (1995a) Annual sedimentation rates and role of the resuspension

processes along a vertical cliff (Ligurian Sea, Italy). J Coast Res 11(3): 690-696 Bavestrello G, Arillo A, Benatti U, Cattaneo-Vietti R, Cerrano C, Cortesogno L, Gaggero L, Giovine M, Tonetti M, Sar M (1995b) Quartz dissolution by the sponge Chondrosia reniformis (Porifera, Demospongiae). Nature 378: 374-376 Bavestrello G, Cerrano C, Cattaneo-Vietti R, Calabria F, Cortesogno L, Sar M (1996) Selective incorporation of foreign material in Chondrosia reniformis. Ital J Zool 63: 215-220 Bavestrello G, Arillo A, Calcinai B, Cerrano C, Lanza S, CattaneoVietti R, Gaino E, Sar M (1998a) Siliceous particles incorporation in Chondrosia reniformis (Porifera, Demospongiae). Ital J Zool 65: 343-348 Bavestrello G, Benatti U, Calcinai B, Cattaneo-Vietti R, Cerrano C, Favre A, Giovine M, Lanza S, Pronzato R, Sar M (1998b) Body polarity and mineral selectivity in the demosponge Chondrosia reniformis. Biol Bull 195(2): 120-125 Bavestrello G, Calcinai B, Boyer M, Cerrano C, Pansini M (2002) The aquiferous system of two Oceanapia species (Porifera, Demospongiae) studied by corrosion casts. Zoomorphology 121: 195-201 Bavestrello G, Benatti U, Cattaneo-Vietti R, Cerrano C, Giovine M (2003) Sponge cell reactivity to various forms of silica. Micros Res Tech 62: 327-335 Bell JJ, Barnes DKA (2002) The relationship between sedimentation, flow rates, depth and time at Lough Hyne Marine Nature Reserve. Ir Nat J 27:106-116 Bigger CH, Hildemann WH, Johnston IS (1982) Characterization of alloimmune memory in a sponge. J Immunol 129: 1570-1572 Bond C (1992) Continuous cell movements rearrange anatomical structures in intact sponges. J Exp Zool 263: 284-302 Calcinai B, Bavestrello G, Cerrano C, Sar M (1999) Biology of the massive symbiotic sponge Cliona nigricans (Porifera, Demospongiae) in the Ligurian Sea. Memoir Queensl Mus 44: 7783 Cerrano C, Arillo A, Bavestrello G, Benatti U, Calcinai B, CattaneoVietti R, Cortesogno L, Gaggero L, Giovine M, Puce S, Sar M (1999a) Organism-quartz interactions in structuring benthic communities: towards a marine bio-mineralogy? Ecol Lett 2: 1-3 Cerrano C, Bavestrello G, Cattaneo-Vietti R, Giovine M, Benatti U, Sar M (1999b) Incorporation of inorganic matter in Chondrosia reniformis (Porifera, Demospongiae): the role of water turbulence. Memoir Queensl Mus 44: 85-90 Cerrano C, Bavestrello G, Boyer M, Calcinai B, Lalamentik LTX, Pansini M (2002) Psammobiontic sponges from the Bunaken Marine Park (North Sulawesi, Indonesia): interactions with sediments. Proc 9th Int Coral Reef Symp, Bali. pp. 279-282 Cerrano C, Pansini M, Valisano L, Calcinai B, Sar M, Bavestrello G (2004) Lagoon sponges from Carrie Bow cay (Belize): ecological benefits of selective sediment incorporation. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge Science in the new Millenium. Boll Mus Ist Biol Univ Genova 68: 239-252 Cerrano C, Sambolino P, Calcinai B, Azzini F, Bavestrello G (2007) Growth of the massive morph of Cliona nigricans (Schmidt, 1862) (Porifera, Clionaidae). Ital J Zool 74: 13-19 Dixon IMT, Moore PG (1997) A comparative study on the tubes and feeding behaviour of eight species of corophioid Amphipoda and their bearing on phylogenetic relationships within the Corophioidea. Philosoph Trans Royal Soc London B 352: 93-112

246

Haywick DW, Mueller EM (1997) Sediment retention in encrusting Palythoa spp. a biological twist to a geological process. Coral Reefs 16: 39-46 Ilan M, Abelson A (1995) The life of a sponge in a sandy lagoon. Biol Bull 189: 363-369 Ise Y, Takeda M, Watanabe Y (2004) Psammobiontic Clionaidae (Demospongiae: Hadromerida) in lagoons of the Ryukyu Islands, southwestern Japan. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millenium. Boll Mus Ist Biol Univ Genova 68: 381-389 Kott P (2006) Observations on non-didemnid ascidians from Australian waters. J Nat Hist 40: 169-234 Krasnow LD, Taghon GL (1997) Rate of tube building and sediment particle size selection during tube construction by the tanaid crustacean, Leptochelia dubia. Estuaries 20: 534-546 Krasko A, Batel R, Schrder HC, Mller IM, Mller WEG (2000) Expression of silicatein and collagen genes in the marine sponge Suberites domuncula is controlled by silicate and myothropin. Eur J Biochem 267: 4878-4887 Main MB, Nelson WG (1988) Sedimentary characteristics of sabellariid worm reefs (Phragmatopoma lapidosa Kinberg). Estuar Coast Shelf Sci 26: 105-109 Maldonado M (1998) Do chimeric sponges have improved survival? Mar Ecol Prog Ser 164: 301-306 Massin C, Doumen D (1986) Distribution and feeding of epibenthic holothuroids on the reef flat of Laing Island (Papua New Guinea). Mar Ecol Prog Ser 31: 185-195 McGhee KE (2006) The importance od life-history stage and individual variation in the allorecognition system of a marine sponge. J Exp Mar Biol Ecol 333: 241-250 Mercurio M, Corriero G, Gaino E (2006) Sessile and non-sessile morphs of Geodia cydonium (Jameson) (Porifera, Demospongiae) in two semi-enclosed Mediterranean bays. Mar Biol 148: 489-501 Miller DC, Muir CL, Hauser OA (2002) Detrimental effects of sedimentation on marine benthos: what can be learned from natural processes and rates? Ecol Eng 19: 211-232 Min-Da Z (1984) Le phnomne dagglutination dans le genre Xenophora. Nouv arch Mus hist nat Lyon 22: 3-51

Nickel M, Brmmer F. (2003) In vitro sponge fragment culture of Chondrosia reniformis (Nardo, 1847). J Biotech 100: 147-159 Pronzato R, Malva R, Manconi R (2004) The taxonomic status of Ircinia fasciculata, Ircinia felix, and Ircinia variabilis (Dictyoceratida, Irciniidae). In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millenium. Boll Mus Ist Biol Univ Genova. 68: 553-563 Rtzler K (1997) The role of psammobiontic sponges in the reef community. Proc 8th Int Coral Reef Symp, Balboa 2: 1393-1398 Salomon CE, Deerinck T, Ellisman MH, Faulkner DJ (2001) The cellular localization of dercitamide in the Palauan sponge Oceanapia sagittaria. Mar Biol 139: 313-319 Schupp P, Eder C, Paul V, Proksch P (1999) Distribution of secondary metabolites in the sponge Oceanapia sp. and its ecological implications. Mar Biol 135: 573-580 Steindler L, Schuster S, Ilan M, Avni A, Cerrano C, Beer S (2007) Differential gene expression in a marine sponge in relation to its symbiotic state. Mar Biotech DOI: 10.1007/s10126-007-9024-2 Takahashi K, Ling YI (1984) Particle selectivity of pelagic tintinnid agglutination. Mar Micropaleontol 9: 87-92 Teragawa CK (1986) Particle transport and incorporation during skeleton formation in a keratose sponge Dysidea etheria. Biol Bull 170: 321-334 Vacelet J (1959) Rpartition general des ponges et systmatique des ponges cornes de la region de Marseille et de quelque stations mditerranennes. Rec Trav St mar Endoume 16(26): 39-101 Werding B, Sanchez H (1991) Life habits and functional morphology of the sediment infaunal sponges Oceanapia oleracea and Oceanapia peltata (Porifera, Haplosclerida). Zoomorphology 110: 203-208 Wiedenmayer F (1977) Shallow-water sponges of the western Bahamas. Experientia Suppl 28: 1-287 Wilson DP (1974) Sabellaria colonies at Duckpool, North Cornwall, 1971-1972, with a note for May 1973. J Mar Biol Assoc UK 54: 393-436 Wulff J (1985) Dispersal and survival of fragments of coral reef sponges. Proc 5th Int Coral Reef Symp, Tahiti 5: 119-124

Porifera research: Biodiversity, innovation and sustainaBility - 2007

247

Observations on reef coral undermining by the Caribbean excavating sponge Cliona delitrix (Demospongiae, Hadromerida)
Andia Chaves-Fonnegra, Sven Zea(*)
Departamento de Biologa and Centro de Estudios en Ciencias del Mar CECIMAR, Universidad Nacional de Colombia, INVEMAR, Cerro Punta de Betn, AA 10-16, Santa Marta, Colombia. andiachaves@gmail.com, szea@invemar.org.co Abstract: Sponges which simultaneously encrust and excavate calcareous substratum are strong space competitors in coral reefs, actively undermining and displacing live coral tissue. On Caribbean reefs, Cliona delitrix colonizes massive corals, encrusting, deeply excavating and aggressively killing entire coral heads. To establish the details of the process of colonization, excavation, undermining and death of corals by this sponge, we carried out observations on sponge-colonized corals at San Andrs Island (SW Caribbean Colombia), and obtained samples for microscopical observation. As it spreads sideward, C. delitrix removed the upper few mm of the coral skeleton, maintaining its surface slightly lower than the surrounding coral, following the curved outline of the coral head. Internal excavation resulted in a solid outer supporting frame and a strongly eroded lace-like internal network. The outer frame was perforated below inhalant papillae by narrow vertical tunnels and below the large oscules by wide and deep spaces. A band of dying or dead coral surrounded the sponge. The sponge sent out a front of tissue using pioneering filaments projecting underneath the coral polyps. Long filaments may surface farther off, forming new bodies that later fuse. From microscopical observations, physical detachment of polyps was ruled out as the cause of coral death, because coral tissue displacement occurred before significant erosion of the polypar skeletal support had taken place. When sponge tissue reached underneath coral tissue, the latter remained healthy when still separated by thin skeletal barriers, but appeared as debris when barriers were broken. Live sponge and coral tissue could occur in direct contact, in which case there was accumulation of granulous cells in the coral tissue. We hypothesize that the mechanism of coral death involves close-range tissue, cell and/or biochemical interactions rather than fluid- or mucus-borne allelochemicals. Keywords: Cliona delitrix, excavation, corals, colonization, cell-cell interactions, Caribbean reefs.

Introduction
Sessile colonial organisms living on hard substrata often compete for space by overtopping or directly overgrowing their neighbors. Takeover of space could be achieved by fast lateral growth that smothers and kills the overgrown tissue, but the neighbors tissue death could precede or parallel space takeover through deployment of aggressive appendages at the boundary and/or release of allelochemical substances (Jackson and Buss 1975, Suchanek and Green 1981, Lang and Chornesky 1990). Upright growing organisms with a limited holdfast area avoid competing for space with more massive or encrusting ones. Similarly, organisms able to penetrate and excavate into the substratum limit competition with organisms dwelling over the surface of the substratum to those points of their body that are exposed (Jackson 1979, Woodin and Jackson 1979). Excavating (burrowing, boring) sponges live in cavities that they themselves bore into calcium carbonate (corals and coralline substratum, mollusk shells, polychaete tubes, calcareous algae) (Goreau and Hartman 1963, Rtzler 1975). Excavation is achieved by both mechanical and chemical etching and removal of coarse silt-size grains through filopodial

extensions of the basal epithelium cellular membrane (see Rtzler and Rieger 1973, Pomponi 1977). Many species only expose papillae and oscula through which food and oxygen are taken and wastes eliminated. However, several species also encrust the excavated substratum and aggressively compete for space with corals and other reef organisms. Some may overgrow live tissue of neighboring organisms (e.g. Vicente 1978). Others, however, avoiding external defense mechanisms of their neighbors, bore directly under them, sending excavating tissue fronts preceded by pioneering tissue filaments, making neighbors detach, apparently by eroding their support (Ward and Risk 1977, Schnberg and Wilkinson 2001, Rtzler 2002, Lpez-Victoria et al. 2003, 2006). Some papillated sponges also kill surrounding coral tissue by releasing mucus presumably laden with allelopatic compounds (Sullivan et al. 1983, Sullivan and Faulkner 1990). The encrusting and excavating sponge Cliona delitrix Pang, 1973 (lat. delitrix: a destroyer; Hadromerida: Clionaidae) is considered one of the most destructive bioeroders in reef corals. It appears as an encrusting layer of bright scarlet tissue of closely spaced papillae and interspersed large, high collared oscules in which deep exhalant canals open (Pang 1973). It

248

penetrates deeply (down to 10-12 cm or more) into the coral skeleton, filling existing and newly eroded spaces with tissue. The lateral and vertical expansion of the sponge eventually leads to the overpowering of entire coral heads, sometimes as large as 1 m in diameter (Pang 1973, Rose and Risk 1985). The sponge is often surrounded by a band of dead coral from where live coral tissue has been eliminated (Pang 1973, Rose and Risk 1985), but the exact mechanism by which this is achieved is not known. Cliona delitrix has been reported to increase in abundance in recent decades, at the expense of live corals, in various reef areas subjected to nutrient runoff and organic pollution (Rose and Risk 1985, Ward-Paige et al. 2005, Chaves-Fonnegra et al. in press). With the aim of obtaining a better understanding how encrusting excavating sponges in general, and C. delitrix in particular, confront, erode, and kill reef corals, we undertook detailed field observations on the sponge and its host corals at San Andrs Island (SW Caribbean, Colombia), and histological analyses of samples from the sponge-coral boundary, to establish: (1) substratum preferences and growth forms, (2) overall excavation processes and, (3) details on the undermining and killing of live coral tissue.

Fig. 1: Cross-section diagram of a massive coral colonized by Cliona delitrix, showing how measurements of the difference in depth between the sponge surface and the surrounding coral skeleton (step), and the approximate depth of excavation (bicycle ray) were measured.

Materials and methods

San Andrs (1232N, 8143W) is an oceanic island of coralline origin, located in the SW Caribbean and surrounded by a calcareous platform with a windward barrier reef, a lagoon with patch reefs, and leeward and windward fore reef terraces with coral carpets (Daz et al. 1995, 1996). Observation and sampling took place throughout the leeward, western margin of the island. There, individuals of Cliona delitrix were photographed and observed in detail, noting its host coral species or substratum (pavement, old dead coral), surface characteristics, aspect of the zone of interaction with live coral, etc. To understand why Cliona delitrix tended to grow at a slightly deeper level than the surrounding substratum, we made detailed observations and measured the depth of the sponge at the boundary step using the butt of a caliper (Fig. 1). To study excavation patterns, several live sponges were fragmented with hammer and chisel. By bathing in commercial bleach remains of dead sponges and small coral heads completely colonized by a live sponge, intact clean skeletons were obtained. To obtain a crude estimate of the depth of sponge excavation inside the coral skeleton, a bicycle steel ray was forcefully driven perpendicularly through one sponge papilla, approximately at the center of the sponge (Fig. 1). This was done only in sponges whose host coral was still alive. The ray was then retrieved and the depth of penetration measured with a caliper. The size of the sponge, measured as projected surface area, was statistically correlated to the depth of the ray penetration using Piersons product-moment correlation coefficient (Sokal and Rohlf 1981). Surface area was obtained from digital photos using the Coral Count Point Program with Excel extension (NCRI-NOVA). To understand the way in which Cliona delitrix excavates under live coral tissue, portions of the coral at the sponge border were detached with hammer and chisel. Structures were measured with a caliper and sketches of the sections

were made. Some fragments were fixed for histology in seawater 10% formalin neutralized with methenamine (20 gL-1) for three days, after which they were stored in 70% ethanol. In the laboratory samples were trimmed and cut into 2 mm thick slabs with a petrographic low speed circular diamond saw (Isomet, Buehler, Chicago). They were then dehydrated, stained (acid fucsin and crystal violet), and embedded in Spurrs low viscosity resin (ERL 4206, Electron Microscopy Sciences). Resin blocks were cut in two and each cut section was glued onto microscope slides using fresh resin. Each section was then ground in a low speed Polisher/Grinder (Minimet 1000, Buehler, Chicago) with diamond-coated grinding paper of increasingly finer grain (79-9 m), and then polished by hand with carburendum 1500 grit paper and metal polishing paste (for details see Rtzler 1974, Willenz and Pomponi 1996, Lpez-Victoria 2003).

Results Substratum preferences and growth forms

Most individuals of Cliona delitrix were found growing on live massive corals or on elevated substratum (old dead coral); none were colonizing branching or thin foliaceous corals and only a few were dwelling directly on the flat calcareous pavement (Fig. 2A, B). There was no mucus on the surface nor was it produced when the sponge was handled. The few small sponge individuals seen were always on dead areas of corals, and consisted generally of an oscule with surrounding papillae, separated or fused (Fig. 2C); individuals larger than 3-5 cm always completely encrusted the surface of the excavated area (Fig. 2D). Sponges had inhalant papillae throughout their surfaces (Fig. 3A), but often the external border of a given colony was comprised of a belt of smooth and level tissue, lacking papillae (Fig. 3B).

General observations on the excavation process

The surface of Cliona delitrix colonies was 0.1 to 1.7 cm lower than the surrounding substratum (from 164 measuring

249

Fig. 2: Underwater photographs of Cliona delitrix. A. On coral Siderastrea siderea. B. On pavement. C. Young individual on dead coral. D. Small individual growing on coral Siderastrea siderea.

points in 30 sponges), being variable within coral species (mean1 standard deviation; Diploria labyrinthiformis 0.50.2 cm, n=4; Porites astreoides 0.90.8 cm, n=2; Siderastrea siderea 0.50.5 cm, n=24; statistical tests were not carried out because of low sample sizes of all but one coral species). The surfaces of the sponges that were found within about 0.5-1 cm of live coral tissue were only a few mm deeper than that of the coral (Fig. 2D). At greater distances the surrounding substratum could be higher, forming a step, because while the sponge eroded the outer coral skeleton to about the same elevation, the surrounding live coral (or dead coral covered by crustose coralline algae) had grown further upwards (Fig. 3B, C). As the sponge spread, it apparently carved the wall of the step by undermining its base, often with tissue fingers and papillae appearing beyond the wall margin (Fig. 3B; 4A, D). When live coral was at some distance from the sponge, its new upward growth formed a second step (Fig. 3B). Apart from oscular collars and papillae, the sponge did not elevate its tissue above the initial level. A single individual of Cliona delitrix could overtake and completely encrust a medium size (up to ca. 30-50 cm

in diameter) coral head (Fig. 3D). In larger coral heads the sponge was frequently characterized by several contiguous mounds, which we assumed were formerly surviving islands of coral that grew upward until the sponge covered them (Fig. 3C, E). Fragmentation of a few whole coral colonies revealed that the sponge could send tissue filaments of about 1 mm in diameter that would extend into the coral skeleton as far as 10 cm from the main sponge body, surfacing to form new sponge bodies, which would later fuse with the main body. From inspection of dead sponges and bleached fragments we found that just below the dermis of the sponge there remained a rather solid frame, 3 to 6 mm thick, perforated by vertical tunnels located below each papilla, and by ample and deep cavities below oscules (Fig. 3F). Vertical tunnels were made directly into each coral calyx, eroding continuously downwards (allowing the insertion of bicycle rays to measure approximate depth of excavation). Below the frame, the skeleton was strongly eroded to a lace-like thinner network, thicker towards the periphery of the sponge, thinner towards the center, traversed by tunnels of the exhalant canals, which converged into oscular cavities. The bicycle rays penetrated

250

Fig. 3: Underwater photographs of Cliona delitrix highlighting: A. Dying and dead coral strands at the sponge-coral border, underlined by sponge excavating tissue filaments. B. The difference in level between the sponge surface and its surrounding substratum, product of further growth of the coral. A first step (left arrow) between the sponge border and the surrounding dead coral band, and a second step (right arrow) formed as the live coral grew upwards further. C. Another example of the step between the sponge and the coral, product of further coral growth, but in this case the coral finally died and has been covered by crustose coralline; the sponge maintains approximately the same curved outline of the original level where it started the excavation. D. An individual completely colonizing a medium size (ca. 30 cm) coral head. E. One individual forming various contiguous mounds, almost completely colonizing a large coral head (ca. 1 m). F. A dying sponge individual showing the naked underlying coral skeleton it had eroded. Notice the rather intact outer frame perforated at the calices under papillae, and deeply eroded areas underneath oscules.

up to 10.1 cm inside sponges that had not yet covered a given coral head. For these sponges, penetration depth was rather similar between coral species: Diploria labyrinthiformis, 3.71.4 cm, n=3; Montastraea faveolata, 4.72.1 cm, n=2; Porites astreoides, 5.21.0 cm; n=2; Siderastrea siderea, 4.71.6 cm, n=30 (again, no statistical comparisons were carried out because of small sample sizes in most corals). Sponge surface tissue area did not correlate statistically with depth of bicycle ray penetration in Siderastrea siderea (r=0.36, p=0.063, n=28). For sponges larger than 25 cm2, depth of ray penetration varied between 3.5 cm and 6.8 cm, while in smaller sponges it varied between 2 and 3.5 cm. In S. siderea colonies of about 30 cm in diameter by 20-25 cm in height, completely overtaken by C. delitrix, we noticed (but did not measure) that they were excavated almost to their base.

Undermining by the sponge vs. coral tissue death

Every Cliona delitrix confronting a live coral, regardless of its size, was surrounded by a band of dead or dying coral. When coral tissue was dead, the calices appeared clean (coral tissue was just removed), or already colonized by turf, frondose or crustose algae. Often the band of dead coral appeared rasped by the long-spined urchin Diadema antillarum Philippi 1845, algae being partly removed and coral calices leveled further. Fish bite marks were rare on the edge of coral tissue confronting the sponge. In coral tissue close 5 mm to the sponge, 1-2 mm to 1-2 cm-wide unhealthy or dead strips of coral tissue were sometimes seen spreading radially outwards from the sponge border (Fig. 3A). Dislodging of coral tissue with hammer and chisel showed that these affected coral areas were undermined by shallow excavating sponge pioneering tissue filaments, apparently involved in coral

251

death. In relatively narrow bands of dead coral (<2.5 cm), an excavating tissue front 1-3 cm thick advanced 0.5-2 cm below the surface, often penetrating beyond the edge of live coral tissue, preceded by tissue filaments, reaching upwards towards coral polyps or extending horizontally farther ahead (Fig. 4A, B, C). In wider bands of dead coral, the front of sponge tissue reached only a few mm beyond the edge of the sponge, but filaments advanced several cm ahead, sometimes into live coral (Fig. 4D). Thick tissue bridges extended beneath adjacent sponge bodies of the same individual (Fig. 4E). The etching marks characteristic of excavating sponge erosion were evident in ground and polished sections of the coral skeletons inhabited by Cliona delitrix, including the encrusted surface. Beneath live coral tissue, sponge tissue filaments penetrated through existing skeletal spaces, expanding them, proceeding upwards through the columellar space of the calices, breaking through dissepiments, until reaching the base of the polyp. Throughout the filaments there were abundant granulous, naturally pigmented goldenbrown cells of varied shapes, 12-15 m in diameter (Fig. 5), also common in the outer surface tissue and reminiscent of spherulous cells (see Pang 1973). In several places healthy sponge and coral tissue were separated by still

intact dissepiments or walls. But empty calyx spaces where the sponge had partially penetrated often contained what appeared to be dead coral tissue debris. In the few instances where actual sponge and coral tissue contact was seen, coral granulous cells accumulated (Fig. 5, pers. comm. by Esther Peters, July 27 2005).

Discussion
Our observations have confirmed that Cliona delitrix is a highly aggressive and destructive reef sponge, and added details about its growth form and how it confronts live coral, allowing us to formulate hypotheses on the mechanism by which it kills coral tissue. The few small, young Cliona delitrix observed were growing on dead coral substratum, indicating that the larva, possibly planktonic (see Mariani et al. 2000) settles preferentially on old dead coral. This behavior has been described for several excavating sponges (see reviews by Goreau and Hartman 1963, McKenna 1997, Schnberg and Wilkinson 2001). But for slightly larger individuals, which were completely surrounded by live coral tissue, one may wonder if larval settlement could have occurred directly on live coral tissue. In fact, Rtzler (1971) hypothesized that

Fig. 4: Schematic drawings showing how Cliona delitrix excavating tissue fronts and pioneering filaments reach out below live coral (A to D), and how two bodies of the same individual are connected under the surface (E). The leftover coral skeleton within the sponge tissue was not drawn.

252

Fig. 5: Photomicrograph of a ground and polished section of the lower part of a live polyp of the coral Siderastrea siderea, being undermined by the sponge Cliona delitrix. It shows the carbonate calyx skeleton (Sk) and the coral tissue (C), traversed by a sponge tissue filament (S, outlined for clarity). Where the sponge and coral tissues touch, there is an accumulation of golden-colored sponge cells (sg) and coral granular cells (cg, slightly lighter gray near the sponge tissue, distinguishable by color under microscope).

mucus-laden larva of the excavating sponges of the genus Aka (as Siphonodictyon), which release abundant mucus, might be able smother and kill live coral polyps, then taking root and expanding. But C. delitrix does not release mucus, and recent studies have shown that healthy corals prevent settlement of other organisms on them (Daz-Pulido and McCook 2004), perhaps through the use of allelochemicals (Fearon and Cameron 1997). As most other encrusting and excavating sponges, Cliona delitrix tends to grow at a level slightly lower than the surrounding substratum (see Acker and Risk 1985, Schnberg and Wilkinson 2001, Rtzler 2002, Lpez-Victoria et al. 2003; but see Vicente 1978 for an exception). Rather than directly overgrowing the adjacent substratum, the lateral undermining by these sponges tends to remove its upper layer (elevated septa in corals, crustose and turf algae in fouled substratum). Differences in height between C. delitrix and the surrounding substratum were quite variable within a coral species, showing that other factors apart from the density and texture of the outer skeleton play a role. By biting the edge of the live coral that confronts other encrusting and excavating sponges, corallivorous fish may sometimes be responsible for initiating and maintaining the difference (see Lpez-Victoria et al. 2006). But fish bite marks were rare on coral tissue confronting C. delitrix in San Andrs. Perhaps grazing of the band of dead coral by sea urchins play a similar role, but we could not clearly ascertain this. Overall, for C. delitrix we believe that two opposing factors are responsible: (1) while the sponge maintains its curved surface at the same level of the coral skeleton, upward growth of adjacent coral tissue or of crustose coralline algae increases the height of the substratum around it and, (2) rasping of the dead coral area by sea-urchins decreases the height of dead coral areas. Further shrinkage

of the substratum under the live sponge by bioerosion and subsequent lowering of the sponge surface may also occur (Acker and Risk 1985). Microscopical observations on shallow excavating species (Rtzler 1971, Zea and Weil 2003) and in C. delitrix (this study) revealed characteristic etching marks on the surface of the external skeleton, showing that some vertical reduction of the substrate level indeed occurs. However, significant substratum shrinking is unlikely to exist in the case of C. delitrix, because its outer skeletal frame remains uniformly thick throughout the sponge. Further substratum shrinkage would occur only if directly eaten and scoured by fish or urchins, which does not seem to be happening, or from mechanical abrasion during storms. In fact, marked individuals and fragments of shallow excavating sponges did not shrink within 1 to 5 years of observation (Lpez-Victoria el al. 2003, S.Z. pers. observations 2005). The curved surface of Cliona delitrix may be maintained as it grows, when the lateral erosion and advance is made preferentially following the same coral skeletal growth bands, which are laid cyclically in different densities in various coral species (e.g., Macintyre and Smith 1974, Huston 1985). Unfortunately, from our skeletal fragments and skeletal sections we could not ascertain whether this was the case. The extent of vertical penetration of Cliona delitrix into a coral while the sponge is still spreading does not seem to be affected by density or texture of the host coral species skeleton, and does not increase significantly as the sponge increases in size. Perhaps it penetrates only deeper once it has taken over the entire colony. Similar values for vertical extension of C. delitrix have been reported in other studies (5 cm, Rose and Risk 1985; 10-12 cm, Pang 1973; vs. 10 cm in our study). Vertical penetrations of the substratum as deep as 8 cm occur in other species as well: in Pione lampa

253

de Laubenfels 1950 (see Rtzler 1974, as Cliona) and in Aka coralliphaga (Rtzler, 1971) (see Glynn 1997). The outer advancing tissue front of Cliona delitrix in massive cerioid corals is similar to the string of beads morphology of excavating tissues in non-encrusting excavating sponges such as Cliona vermifera, in which vertical lobes of tissue located within coral calices are interconnected by horizontal excavating filaments of varying diameters; smaller filaments penetrate through the thecal wall, invading another calyx; the calyx is then filled with tissue and eroded vertically and horizontally cutting dissepiments and septae (Ward and Risk 1977). This may be a general way by which excavating sponges erode corals, first filling existing spaces and then eroding walls and obstacles (see review en Schnberg 2003), subsequently varying widely within and among species in further erosion of the coral skeleton. The vertical component of calix erosion is emphasized in C. delitrix in comparison to other shallow excavating encrusting sponges. When they become large enough, some excavating sponge species add tissue and supporting siliceous skeleton above the substratum, becoming massive, and often erode all skeletal connections to the substratum, becoming free-living, in both cases conforming what is called a gamma stage (e.g., Topsent 1888, Vosmaer 1933, Rtzler 1971, Vicente 1978, Calcinai et al. 1999). Even shallow encrusting excavating species that spread horizontally in the upper few cm of substratum slightly thicken their tissue when they run out of space (Lpez-Victoria et al. 2003). Owing to its rather soft tissue Cliona delitrix is perhaps not able to grow further upwards. It may compensate its lack of upward growth with deep vertical penetration into the substratum. Its excavation pattern of an outer solid shell and an inner cavernous core provides enough space for the tissue and water-circulation system, while giving stability and support to the sponge. The growth of Cliona delitrix with far-reaching ramifications inside the coral skeleton allows it to simultaneously weaken coral tissue in various portions of the colony. In a similar way, but from a completely eroded central chamber filled with tissue, some species of the genus Aka erode tunnels that reach out to several separated places of the surface of a coral; further expansion of the outgrowths and the central chamber ends up taking over the entire coral head (Rtzler 1971). Not penetrating deeply, shallow encrusting and excavating sponges resort to shallow-lying tissue fronts and closer-range pioneering filaments that penetrate directly beneath live tissue coral at their borders (Ward and Risk 1977, Schnberg and Wilkinson 2001, Schnberg 2001, Rtzler 2002, Lpez-Victoria et al. 2003, 2006). Our histological observations of the zone of interaction between Cliona delitrix and coral tissues show that coral tissue death occurs before erosion of the calyces is enough to induce coral polyp detachment. Indeed, coral tissue and its skeletal support are so strongly interwoven that polyp detachment seems hardly possible if the skeleton is not thoroughly eroded. This purely physical detachment has been assumed in other works with encrusting and excavating sponges (e.g., Lpez-Victoria et al. 2003, 2006). Coral death in close vicinity of C. delitrix could be the consequences of the release of allelochemicals that can kill coral tissue, and which are known to be present in its organic extract (Chaves-Fonnegra

et al. 2005). Similar effects have been observed in the fistulate excavating sponges of the genus Aka (Rtzler 1971), and have been attributed to the release of abundant mucus assumed to be carrying known allelopathic compounds (Sullivan et al. 1983, Sullivan and Faulkner 1990). The mucus is a medium that helps spreading and retaining compounds on the surface of corals, facilitating their entrance into tissue (Hay et al. 1998). However, as C. delitrix does not produce mucus, there must be some other means, if any, for the deployment of the potentially harmful substances it produces. Whether they are exuded directly to the water remains to be determined. At any rate, our histological observations show healthy sponge and coral tissues within the coral skeleton just separated by thin carbonate walls, as well as sites of direct contact where the two tissues are still alive. This appears to indicate that chemical substances are not being released into the fluids contained in the coral skeleton. Moreover, the observed accumulation of cells in coral tissue in direct contact with sponge tissue points towards close-range cellular or biochemical processes involved in coral tissue death, which require further study.

Acknowledgments
This paper is part of the M.Sc. thesis of A.Ch.-F., Marine Biology Program, Universidad Nacional de Colombia at Instituto de Investigaciones Marinas y Costeras INVEMAR. Supported by the Colombian Science Fund COLCIENCIAS (grant 110109-13544 to C. Duque and S. Zea), Universidad Nacional de Colombia, and INVEMAR. We are grateful to M. Lpez-Victoria, E. Peters and J. Reyes for their help with histological techniques and interpretation, and to J. C. Mrquez, L. Castellanos and Banda Dive Shop for help during field work. Contribution 979 of INVEMAR and 301 of CECIMAR and the Graduate Program in Marine Biology of the Universidad Nacional de Colombia, Faculty of Sciences. Two anonymous reviewers helped greatly to improve the manuscript.

References
Acker KL, Risk MJ (1985) Substrate destruction and sediment production by the boring sponge Cliona caribbaea on Grand Caiman Island. J Sedim Petrol 55(5): 705-711 Calcinai B, Cerrano C, Bavestrello G, Sar M (1999) Biology of the massive symbiotic sponge Cliona nigricans (Porifera: Demospongiae) in the Ligurian Sea. Mem Qld Mus 44: 77-83 Chaves-Fonnegra A (2006). Mecanismos de agresin de esponjas excavadoras incrustantes y consecuencias sobre corales arrecifales en el Caribe colombiano. M.Sc. Thesis, Universidad Nacional de Colombia - INVEMAR, Santa Marta Chaves-Fonnegra A, Zea S, Gmez ML (in press) Abundance of the excavating sponge Cliona delitrix in relation to sewage discharge at San Andrs Island, SW Caribbean, Colombia. Bol Invest Mar Cost 36 Chaves-Fonnegra A, Lpez-Victoria M, Parra-Velandia F, Zea S (2005) Ecologa qumica de las esponjas excavadoras Cliona aprica, C. caribbaea, C. delitrix y C. tenuis. Bol Invest Mar Cost 34: 43-67 de Laubenfels MW (1950) The Porifera of the Bermuda Archipelago. Trans Zool Soc London 27: 1-201 Daz JM, Garzn-Ferreira J, Zea S (1995) Los arrecifes coralinos de la isla de San Andrs, Colombia: estado actual y perspectivas

254

para su conservacin. Coleccin Jorge Alvarez LLeras 7, Acad Col Cienc Exact Fis Nat, Bogot Daz JM, Daz-Pulido G, Garzn-Ferreira J, Geister J, Snchez JA, Zea S (1996) Atlas de los arrecifes coralinos del Caribe colombiano. I. Complejos arrecifales ocenicos. Serie Publicaciones Especiales 2, INVEMAR, Santa Marta Daz-Pulido G, McCook L. J (2004) Effects of live coral, epilithic algal communities and substrate type on algal recruitment. Coral Reefs 23: 225-233 Fearon RJ, Cameron AM (1997) Preliminary evidence supporting the ability of hermatypic corals to affect adversely larvae and early settlement stages of hard coral competitors. J Chem Ecol 23(7): 1769-1780 Glynn P (1997) Bioerosion and coral reef growth: a dynamic balance. In: Birkeland C (ed). Life and death of coral reefs. Chapman & Hall, New York. pp. 68-95 Goreau TF, Hartman WD (1963) Boring sponges as controlling factors in the formation and maintenance of coral reefs. In: Sognnaes RF (ed). Mechanisms of hard tissue destruction. American Association for the Advancement of Science, Washington DC. pp. 25-54 Hay ME, Stachowicz JJ, Cruz-Rivera E, Bullard S, Deal MS, Lindquist N (1998) Bioassays with marine and freshwater macroorganisms. In: Haynes KF, Millar JG (eds). Methods in chemical ecology: biossay methods. Chapman and Hall. New York. pp. 32-122 Huston M (1985) Variation in coral growth rates with depth at Discovery Bay, Jamaica. Coral Reefs 4: 19-25 Jackson JBC (1979) Morphological strategies of sessile animals. In: Larwood G, Rosen BR (eds). Biology and systematics of colonial organisms. Academic Press, London. pp. 499-555 Jackson JBC, Buss LW (1975) Allelopathy and spatial competition among coral reef invertebrates. Proc Natl Acad Sci USA 72: 51605163 Lang JC, Chornesky EH (1990) Competition between scleractinian reef corals a review of mechanisms and effects. In: Dubinsky Z (ed). Ecosystems of the world 25: Coral reefs. Elsevier, Amsterdam. pp. 209-252 Lpez-Victoria M (2003) Interacciones entre esponjas excavadoras del complejo Cliona aprica - C. langae - C. caribbaea y corales ptreos en el Caribe colombiano. M.Sc. Thesis, Marine Biology, Universidad Nacional de Colombia - INVEMAR, Santa Marta Lpez-Victoria M, Zea S, Weil E (2003) New aspects on the biology of the excavating sponge complex Cliona caribbaea - C. langae - C. aprica. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Boll Mus Ist Biol Univ Genova 68: 425-432 Lpez-Victoria M, Zea S, Weil E (2006) Competition for space between encrusting excavating Caribbean sponges and other coral reef organisms. Mar Ecol Progr Ser 312: 113-121 Macintyre IG, Smith SV (1974) X-radiographic studies of skeletal development in coral colonies. Proc 2nd Int Coral Reef Symp, Brisbane 2: 277-287 Mariani S, Uriz MJ, Turon X (2000). Larval bloom of the oviparous sponge Cliona viridis: coupling of larval abundance and adult distribution. Mar Biol 137: 783-790 McKenna SA (1997) Interactions between the boring sponge, Cliona lampa and two hermatypic corals from Bermuda. Proc 8th Int Coral Reef Symp, Balboa 2: 1369-1374 Pang RK (1973) The systematics of some Jamaican excavating sponges. Postilla 161: 1-75

Pomponi SA (1977) Etching cells of boring sponges: an ultraestructural analisis. Proc 3rd Int Coral Reef Symp, Miami 2: 485-490 Rose CS, Risk MJ (1985) Increase in Cliona delitrix infestation of Montastrea cavernosa heads on an organically polluted portion of the Grand Cayman fringing reef. PSZN Mar Ecol 6(4): 345-363 Rtzler K (1971) Bredin-Archbold Smithsonian Biological Survey of Dominica: burrowing sponges, genus Siphonodictyon Bergquist, from the Caribbean. Smithson Contrib Zool 77: 1-37 Rtzler K, Rieger G (1973) Sponge burrowing: fine structure of Cliona lampa penetrating calcareous substrata. Mar Biol 21: 144162 Rtzler K (1974) The burrowing sponges of Bermuda. Smithson Contrib Zool 165: 1-32 Rtzler K (1975) The role of burrowing sponges in bioerosion. Oecologia (Berl.) 19: 203-216 Rtzler K (2002) Impact of crustose clionid sponges on Caribbean reef corals. Acta Geol Hisp 37: 61-72 Schnberg CHL (2001). Estimating the extent of endolithic tissue of a Great Barrier Reef clionid sponge. Senckenberg Marit 31(1): 29-39 Schnberg CHL, Wilkinson CR (2001) Induced colonization of corals by a clionid bioeroding sponge. Coral Reefs 20: 69-76 Schnberg CHL (2003) Substrate effects on the bioeroding Demosponge Cliona orientalis. 2. Substrate colonization and tissue growth. PSZN Mar Ecol 24(1): 59-74 Sokal R.R, Rohlf FJ (1981) Biometry: The principles and practice of statistics in biological research, 2nd edition. W.H. Freeman and Company, San Francisco Suchanek TH, Green DJ (1981) Interspecific competition between Palythoa caribaeorum and other sessile invertebrates on St. Croix reef, U.S. Virgin Islands. Proc 4th Int Coral Reef Symp, Manila 2: 679-684 Sullivan BW, Faulkner DJ (1990) Chemical studies of the burrowing sponge Siphonodictyon coralliphagum. In: Rtzler K (ed). New perspectives in sponge biology. Smithsonian Institution Press, Washington DC. pp. 45-50 Sullivan BW, Faulkner DJ, Webb L (1983) Siphonodictidine, a metabolite of the burrowing sponge Siphonodictyon sp. that inhibits coral growth. Science 221: 1175-1176 Topsent, E (1888) Contribution letude des clionides. Arch Zool Exp Gn, Ser 2, 5(bis): 1-165 Vicente VP (1978) An ecological evaluation of the West Indian Demosponge Anthosigmella varians (Hadromerida: Spirastrellidae). Bull Mar Sci 28: 771-777 Vosmaer, GC J (1933) The sponges of the Bay of Naples: Porifera incalcaria. Martinus Nijhoff, The Hague, 828 pp Ward P, Risk M (1977) Boring pattern of the sponge Cliona vermifera in the coral Montastrea annularis. J Paleontol 51 (3): 520-526 Ward-Paige CA, Risk MJ, Sherwood OA, Jaap WC (2005) Clionid sponge surveys on the Florida Reef Tract suggest land-based nutrient inputs. Mar Poll Bull 51: 570-579 Willenz P, Pomponi S (1996) A new deep sea coralline sponge from Turks and Caicos Islands: Willardia caicosensis gen. et sp. nov. (Demospongiae: Hadromerida). Bull Inst Royal Sci Nat Belg Biol 66 Suppl: 205-218 Woodin SA, Jackson JBC (1979) Interphyletic competition among marine benthos. Amer Zool 19: 1029-1043 Zea S, Weil E (2003) Taxonomy of the Caribbean excavating sponge species complex Cliona caribbaea - C. aprica - C. langae (Porifera, Hadromerida, Clionaidae). Caribb J Sci 39(3): 348-370

Porifera research: Biodiversity, innovation and sustainaBility - 2007

255

Large-scale distributional patterns of the encrusting and excavating sponge Cliona delitrix Pang on Florida Keys coral substrates
Mark Chiappone(1*), Leanne M. Rutten(1), Steven L. Miller(1), Dione W. Swanson(2)
Center for Marine Science, University of North Carolina at Wilmington, 515 Caribbean Drive, Key Largo, FL 33037, USA. chiapponem@uncw.edu, ruttenl@uncw.edu, millers@uncw.edu (2) Division of Marine Biology and Fisheries, Rosenstiel School of Marine and Atmospheric Science, University of Miami, 4600 Rickenbacker Causeway, Miami, FL 33149, USA. dswanson@rsmas.miami.edu
(1)

Abstract: While encrusting and excavating sponges of the genus Cliona are important bioeroders on Caribbean reefs, relatively little information exists on large-scale distributional patterns and environmental correlates. A survey encompassing 181 Florida Keys coral reef and hard-bottom sites sampled C. delitrix Pang density and size to explore relationships with habitat-related characteristics such as depth, distance from shore, and coral density and cover. Nine habitats were surveyed across the continental shelf representing a range in depth (1-27 m), cross-shelf position, topographic complexity, and coral abundance. A stratified random sampling design used replicate 8-m x 1-m transects per site to survey 2,896 m2 of coral reef and hard-bottom habitat. Sixty of the 181 sites yielded 189 individuals of C. delitrix. The distribution of C. delitrix was not proportional to the sampling effort, with four of the nine habitats accounting for ~83% of all sightings. Mean sponge density (no. per m2) ranged from 0.01 to 0.24 and differed significantly among habitats. Shallower (< 6 m) and more wave-exposed habitats on the platform margin had statistically lower C. delitrix densities, while densities on the deeper fore reef slope were up to 18 times greater. Cliona delitrix area per m2 and mean size were similarly patterned to density, with significantly greater mean values on patch reef and deeper (> 10 m) fore reef slope habitats. Regression analysis indicated that greater C. delitrix densities were found at greater depths, while larger C. delitrix sponges were found closer to shore in areas of higher coral abundance. Patterns of coral colonization relative to coral availability suggest that C. delitrix occupied ~23% of the coral taxa preferentially relative to availability, especially coral species confined to relatively few habitats or those most abundant on patch reefs or the deeper fore reef slope. Keywords: Abundance, Cliona, corals, delitrix, Florida Keys, sponges

Introduction
The condition and stability of coral reefs represents a balance between processes that contribute to calcium carbonate deposition and those forces that work to erode the reef framework (Hallock and Schlager 1986). A variety of physical and biological factors affect carbonate accretion and erosion. Several invertebrate and vertebrate taxa are involved in eroding carbonate substrates, including bivalves, sipunculid worms, sea urchins, parrotfishes, and sponges, principally from the Cliona and Aka (= Siphonodictyon) genera (Bak 1976, Hudson 1977, Tribollet and Golubic 2005). Many bioeroders feed on components of the plankton in both the larval and adult stages, and thus abundance and distribution patterns can be affected by both benthic and water column processes (Hallock and Schlager 1986, Kiene and Hutchings 1994, Glynn 1997). Bioerosion or boring contributes to the silt fraction of reef sediments, weakens the reef framework, and can thus render reefs more susceptible to storms (MacGeachy 1977, Macdonald and Perry 2003).

Coral reef sponges are frequently very important components of the sessile benthic fauna in terms of their biomass, species richness, and ecological roles in nutrient recycling (Reiswig 1974, Wilkinson 1983), bioerosion (Hallock and Schlager 1986), consolidation of the reef framework (Wulff and Buss 1979), and enhancing recovery of reefs from disturbances, among others (reviewed in Diaz and Rtzler 2001). At least 36 different Caribbean sponges, 20 of which are from the genus Cliona, are involved in bioerosion and are among the most effective at eroding coral substrates and producing fine carbonate sediments (Hudson 1977, Diaz and Rtzler 2001, Zea and Weil 2003). Sponge bioerosion can account for up to 50% to 90% of the calcium carbonate removed from coral skeletons (Bak 1976, MacGeachy and Stearn 1976, Highsmith et al. 1983, Sammarco and Risk 1990), with rates of calcium carbonate erosion as high as 23 kg/m2/yr (Neumann 1966), but usually lower (Hein and Risk 1975, Bak 1976, Reis and Leo 2002). Several studies have explored potential factors affecting Cliona distribution and abundance patterns, including habitat and depth, coral growth morphology, coral growth rate, the growth of encrusting organisms, nutrient

256

input, and temperature (Risk and MacGeachy 1978, Holmes 1997, Lopez-Victoria and Zea 2005, Marquez et al. 2006). The Florida Keys coral reef ecosystem has experienced dramatic decadal changes in coral reef community structure and condition, as evidenced by declines in live coral cover, concurrent increases in algal cover, increases in the incidence of coral disease, mass mortality of the sea urchin Diadema antillarum, and a general pattern of poor recovery relative to baseline observations in the 1960s and 1970s (Chiappone et al. 1997, Miller et al. 2002, Porter and Porter 2002). The proximate causes for these declines have been and continue to be the subject of much debate and include a variety of potential causes such as climate change, diseases of generally unknown etiologies, physical impacts, water quality degradation, and intensive fishing (Bohnsack and Ault 1996, NOAA 1996). Of particular interest is whether reefs in this region have suffered such dramatic changes that reef accretion is now outpaced by processes eroding carbonate substrates. There is some evidence that clionaid sponges have increased in abundance in the Florida Keys and elsewhere in the wider Caribbean and that this pattern may in some cases be related to anthropogenic changes to water quality (Zea and Weil 2003, Ward-Paige et al. 2005). While large-scale coral reef assessments and longer term monitoring exist, few studies have assessed even the basic spatial patterns of excavating or boring sponges and the temporal patterns of infestation of corals (Schmahl 1990, Calahan 2005). The present study addressed three questions concerning the distribution and abundance of Cliona delitrix in the Florida Keys. First, what are habitat-related patterns in the density and size of C. delitrix that correspond to differences in cross-shelf position and depth? Second, what are the relationships, if any, between C. delitrix density and size and coral abundance such as cover and colony density? Finally, does the distribution of C. delitrix provide evidence for preferential colonization of particular coral species, or are coral species colonized in proportion to their abundance?

Materials and methods

Coral reef development in the continental U.S. primarily occurs from 25 m to 13 km offshore of the Florida Keys archipelago, a chain of more than 1,700 Pleistocene islands consisting of Key Largo Limestone and Miami Oolite (FDEP 1998, Lidz et al. 2006). The Florida Keys reef tract extends approximately 360 km along the south Florida shelf and includes a discontinuous series of offshore bank reefs paralleling the islands in a general southwesterly arc. The Pleistocene islands are separated from offshore reefs by Hawk Channel, an elongated basin 5 m to 12 m in depth dominated by sand, seagrasses, and patch reefs (FDEP 1998). Several authors (Marszalek et al. 1977, Shinn et al. 1989) have classified the Florida Keys by regional sectors (upper, middle, and lower Keys) based upon geographic variations in reef development and distribution, the influence of estuarine and marine water masses, and the underlying Pleistocene topography. Offshore bank reefs with high-relief spur and groove topography and patch reefs are most numerous in the upper and lower Keys regions, where reefs occur preferentially seaward of the islands (Marszalek et al. 1977, FDEP 1998).

Florida Keys coral reef and hard-bottom habitats have been classified into four principal community types: live bottom (octocoral-dominated hard grounds, low-relief hard-bottom), patch reefs (individual, aggregate), transitional reefs, and bank reefs (Marszalek et al. 1977, Jaap 1984). In subsequent mapping efforts, these types were divided further (FDEP 1998, Lidz et al. 2006). Nine habitat types were selected for sampling C. delitrix based upon habitat area coverage and distribution in the study area (Table 1). Habitats relatively close to shore included mid-channel and offshore patch reefs distributed from Hawk Channel to the shoreward edge of the platform margin and consisted of relatively small (10-25 m) dome-shaped or linear-shaped patches with clusters of coral heads (FDEP 1998). The mid-channel and offshore patch reefs included four of the five patch reef categories (individual, aggregation, halo, aggregation with halo) used in previous mapping efforts (FDEP 1998). Along the outer platform margin or reef tract, four depth zones were surveyed: 1) spur and groove topography and lowrelief hard-bottom from 1-6 m depth, 2) low-relief hard-bottom and low-relief spur and groove from 6-12 m depth, 3) low-relief spur and groove and rocky outcrops from 15-19 m depth, and 4) rocky outcrops, terrace, and low-relief spur and groove from 22-27 m depth. To quantify habitat distribution patterns of C. delitrix density and size, a two-stage stratified random sampling design was employed during May-October 2005 that incorporated unique habitat types across the southeast Florida shelf and regional sectors hypothesized to represent spatial variations in water quality regimes (Cochran 1977). Using a Geographic Information System (GIS), the Florida Keys sampling domain was overlaid with a grid of sites (each 200 m x 200 m) that were the primary sampling units. Each site that contained coral reef or hard-bottom habitat, as determined from the FDEP (1998) benthic habitat map, was assigned a unique number. Sites were then randomly selected for sampling from a discrete uniform probability distribution to ensure that each site had equal selection probability distribution. The 181 sites were located in the northern terminus of the reef tract in Biscayne National Park, upper Keys, middle Keys, and lower Keys west to the Marquesas Keys (Fig. 1). At each site, two randomized, pre-determined GPS points were used to haphazardly orient single 8-m x 1-m belt transects at each GPS point to quantify C. delitrix density, size (area) per individual, live coral cover, and coral density. Cliona delitrix individuals were defined as continuous patches of live sponge tissue on the substratum surface, whether on dead or live coral. Using a 0.5 PVC stick, a 0.5-m length on each side of an 8 m transect was carefully surveyed for the presence of C. delitrix, yielding a total sample area of 8 m2 per transect and 16 m2 per site (Table 2). Measurements using calipers or a plastic ruler were made of C. delitrix dimensions (diameter, length, width), depending on the approximate shape of the individual sponge patch, to estimate surface area coverage. On the same two transects, all scleractinian coral colonies greater than 4 cm in maximum diameter were counted to estimate colony densities for each species. On each transect, 100 points equally spaced on along the transect tape were surveyed for benthic cover using the planar-point intercept technique to derive estimates of scleractinian coral cover and other benthic types (Ohlhorst et al. 1988, Miller et al. 2002).

257

Table 1: Survey sampling effort and characteristics of coral reef and hard-bottom habitats surveyed for C. delitrix density and size in the Florida Keys during 2005. Habitat types are arranged from inshore to offshore. Shore distance and survey depth represent the mean (range) in linear distance to the nearest shoreline and surveyed transect depth, respectively. Habitat type / abbreviation Mid-channel patch reef (MPR) Offshore patch reef (OPR) Inner line reef tract (IRT) Low-relief hard-bottom 6 m (LHB6) High-relief spur and groove (HSG) Low-relief hard-bottom (LHB10) Low-relief spur and groove (LSG10) Fore reef slope (FRS18) Fore reef slope (FRS24) Total No. sites (% total effort) 51 (28.2%) 27 (14.9%) 5 (2.8%) 15 (8.3%) 19 (10.5%) 23 (12.7%) 16 (8.8%) 16 (8.8%) 9 (5.0%) 181 (100%) No. 8 m2 transects 102 54 10 30 38 46 32 32 18 362 Shore distance (km) 4.6 (1.6-7.5) 6.9 (4.8-9.7) 7.0 (6.9-7.1) 7.5 (5.6-9.5) 8.6 (6.5-10.2) 8.1 (4.6-10.4) 8.4 (6.7-10.3) 8.5 (7.1-10.6) 8.6 (6.4-10.3) Transect depth (m) 4.9 (0.9-9.9) 6.0 (3.0-11.1) 4.0 (2.1-5.7) 4.7 (2.7-6.3) 4.4 (1.2-7.2) 8.0 (5.7-11.4) 10.1 (7.8-12.0) 16.9 (15.0-19.2) 24.6 (21.6-27.0)

Fig. 1: C. delitrix survey locations in Biscayne National Park (BNP) and the Florida Keys National Marine Sanctuary (FKNMS) during May-October 2005.

The survey effort encompassed 73 field days of SCUBA surveys between 8 May and 2 October 2006. Mean density (no. individuals per m2), area coverage (cm2/m2), and mean sponge size (cm2/individual) for C. delitrix were computed for each of the nine habitat types. Statistical comparisons of means were conducted by calculating confidence intervals (CI) based upon the equation CI = mean t[, df] * SE, with SE (standard error) estimated by the two-stage stratified design (Cochran, 1977). Confidence intervals were adjusted for multiple comparisons using the Bonferroni procedure (Miller 1981). While this adjustment was made for relatively conservative statistical testing, it reduced the probability of spurious significant pair-wise comparisons. The experiment-wise error rate was held at

= 0.05, and the comparison-wise error rate was adjusted based on the number of multiple comparisons (comparisonwise error rate = /c, where c = k (k-1)/2 and k = number of habitat types compared). Relationships between C. delitrix density and size were explored with respect to distance from shore, depth, scleractinian coral cover, and scleractinian coral density using correlation and regression (linear and non-linear) analyses (Zar 1996). The pattern of stony coral colonization by C. delitrix was compared against the expected distribution, calculated by multiplying the total number of C. delitrix sponges colonizing corals by the proportional availability of stony coral species. Cliona delitrix patterns of coral colonization were assessed using Ivlevs index of electivity

258

(e) = (ri Pi)/(ri + Pi), where ri is the proportion of coral species i colonized and Pi is the proportion of coral species i available (Manly et al. 1993). The index rates coral species colonization from 1 to +1, with a value of 1 indicating total rejection, 0 indicating those corals colonized in proportion to their abundance, and +1 indicating a preference for coral species to the exclusion of others by C. delitrix.

Results
Cliona delitrix density and size data for the nine coral reef and hard-bottom habitat types representing 181 Florida Keys sites are summarized in Table 2. As a percentage of sites where C. delitrix was encountered, the deepest habitat types on the fore reef slope yielded the highest site frequencies (4478%). A total of 189 C. delitrix individuals on the substratum surface were recorded, with four habitats (mid-channel patch reefs, 6-15 m low-relief spur and groove, 18 m fore reef slope, and 24 m fore reef slope) accounting for ~82% of all C. delitrix encountered, despite only representing ~51% of the total substratum area surveyed. The distribution of individuals among habitats was not proportional to the sampling effort by habitat, with fewer C. delitrix than expected recorded from inner line reef tract (IRT), high-relief spur and groove (HSG), and shallow (< 6 m) low-relief hard-bottom (LHB6), all of which are shallow, relatively wave-exposed habitats (Table 2). Cliona delitrix was most frequently encountered on deeper fore reef slope habitats (LSG10, FRS18, FRS24) from 8-28 m depth, where individuals were encountered along transects at more than 75% of sites. Mean C. delitrix densities (no. sponges per m2) among habitats ranged from 0.013 per m2 on inner line reef tract (IRT) and shallow hard-bottom (LHB6) to more than 0.115 individuals per m2 on low-relief spur and groove and deeper fore reef slope habitats (Table 2). At the 0.0014 significance level, statistical differences in mean density were detected between deeper fore reef slope (FRS18 and FRS24) habitats and several of the shallower, more wave exposed habitats (IRT, HSG, LHB6) (Fig. 2A). Mean densities on the deeper fore reef slope were nine to 18 times greater than on the shallower platform margin habitats, while patch reefs yielded intermediate mean C. delitrix density values compared to the shallower and deeper fore reef slope (Table 2).

Comparisons of C. delitrix densities with habitat characteristics indicated that depth was most strongly correlated with sponge densities (r = 0.887, P < 0.002). Of the environmental factors considered, depth accounted for 75.7% (adjusted R2) of the total variability in density in the linear regression model (y = 0.0102x 0.259), which was highly significant (F = 25.874, P < 0.001). Upon further analysis, it was found that the relationship between C. delitrix density and depth was best modeled (R2 = 0.973) using a third-degree polynomial function (F = 58.952, P < 0.001). This pattern reflected the greater sponge densities with increasing depth to a maximum level in the deeper fore reef slope habitat at 1519 m depth (FRS18), then decreasing on the deepest portion of the fore reef slope (FRS24) sampled (Fig. 3A). Cliona delitrix densities were not significantly correlated with site distance from shore (r = 0.432, P > 0.20), scleractinian coral cover (r = 0.083, P > 0.50), or scleractinian colony density (r = 0.293, P > 0.20). Habitat variations in mean C. delitrix area per m2 and mean size per individual are summarized in Table 2 and Fig. 2. Mean sponge area per m2 ranged from a low of 0.04-0.05 cm2/ m2 on high-relief spur and groove and shallow hard-bottom (LHB6) to 10.4 cm2/m2 on the deep fore reef slope (FRS24) (Table 2). Mean sponge area per m2 followed a similar habitat pattern as sponge density, with significantly greater coverage (P < 0.0014) on patch reefs and deeper fore reef slope habitats compared to shallower, more wave exposed habitats. Mean C. delitrix area coverage on high-relief spur and groove and shallow hard-bottom (LHB6) was significantly lower than on patch reefs and fore reef slope habitats 10+ m in depth (Fig. 2B). Mean C. delitrix area per sponge also exhibited significant differences (P < 0.0014) among the habitats sampled (Table 2), with mean sizes ranging from 128 to 139 cm2 per individual on mid-channel and offshore patch reefs, respectively, to less than 3.5 cm2 per individual on relatively wave-exposed high-relief spur and groove and shallow hard-bottom (LHB6) habitats (Fig. 2C). In contrast to sponge density, mean C. delitrix size was greatest in habitats closest to shore, intermediate on the deeper fore reef slope, and lowest on the shallow, wave-exposed platform margin. Comparisons of C. delitrix size (area per sponge) with the habitat characteristics considered indicated a negative correlation with distance

Table 2: Mean ( 1 SE) density, area (cm2) per transect, and size (cm2) of C. delitrix sampled in the Florida Keys. See Table 1 for habitat abbreviations. Habitat (sites) (% of total effort) MPR (51) (28.2%) OPR (27) (14.9%) IRT (5) (2.8%) LHB6 (15) (8.3%) HSG (19) (10.5%) LHB10 (23) (12.7%) LSG10 (16) (8.8%) FRS18 (16) (8.8%) FRS24 (9) (5.0%) Site frequency (%) 31.4 18.5 20.0 13.3 10.5 30.4 43.8 81.3 77.8 No. individuals (% of total) 37 (19.6) 14 (7.4) 1 (0.5) 3 (1.6) 5 (2.6) 11 (5.8) 30 (15.9) 61 (32.3) 27 (14.3) No. individuals per m2 0.045 0.020 0.032 0.020 0.013 0.013 0.013 0.009 0.016 0.013 0.030 0.012 0.117 0.052 0.238 0.068 0.188 0.042 Sponge area (cm2) per m2 5.78 3.48 4.49 3.09 0.36 0.36 0.04 0.03 0.05 0.05 2.16 0.85 7.34 4.34 7.56 2.98 10.36 5.97 Size (cm2)/sponge 127.5 50.5 138.6 70.9 28.3 3.3 1.4 3.2 0.8 72.1 22.6 63.5 25.8 31.7 10.9 55.2 17.9

259

Fig. 2: A. Mean (+ 1 SE) density (no. individuals per m2). B. Total area per transect (cm2 per m2). C. Size (cm2 per individual) of Cliona delitrix in the Florida Keys. Habitat type abbreviations are listed in Table 1. Numbers in parentheses for A and B show the number of sites sampled per habitat type, and for C are the number of individuals sampled for size. Darkened bars indicate those habitat types that were significantly different (P < 0.05) from at least one other habitat type.

total variability, respectively, in mean sponge size according to linear regression modeling. However, both linear models of C. delitrix size versus distance from shore (F = 4.136, P = 0.081) and versus coral cover (F = 5.306, P = 0.055) were not significant. Upon further analysis, the relationship between C. delitrix size and coral cover was best modeled (R2 = 0.484) using a third-degree polynomial function (Fig. 3C), but this non-linear relationship was not significant (P > 0.05). The overall pattern of C. delitrix size per sponge indicated that the largest individuals generally occurred on patch reefs characterized by a combination of closer proximity to shore (4.6-6.9 km) and higher coral cover (8-16%) compared to other habitats. Patterns of C. delitrix occupancy of scleractinian coral colonies relative to colony availability are summarized by species in Table 3. Of the 22,006 scleractinian coral colonies sampled representing 47 taxa, 387 colonies (1.8%) among 20 taxa (43%) had one or more C. delitrix individuals that were visibile on colony skeletons. The proportion or percent colonization of coral colonies with C. delitrix ranged from 0.61% to 5.43% among the nine habitat types, with the greatest percent colonization on high-relief spur and groove (HSG, 5.43%), the deeper fore reef slope (FRS18, 4.67%), and inner line reef tract (IRT, 2.98%) habitats. Percent colonization of corals by C. delitrix was relatively low (0.61-0.63%) on midchannel and offshore patch reefs that were characterized by close proximity to shore, high coral cover, and intermediate to high topographic complexity, as well as on shallower (< 6 m) and deeper (> 6 m) low-relief hard-bottom (0.611.11%). Of the 47 coral taxa encountered during the study, 27 taxa (57.5%) were not found with C. delitrix and thus had electivity index values of -1 (Table 3). An additional nine coral taxa had negative electivity index values, while 11 taxa (23.4%) had positive values, indicating perhaps preferential colonization of particular coral species relative to colony availability. Of the corals with the most positive electivity index values, Acropora palmata was relatively rare in the study area and restricted to HSG and IRT habitats. Agaricia lamarcki was only found in FRS18 and FRS24 habitats, while Solenastrea bournoni was most common on patch reefs. Both Montastraea franksii (all habitat types except HSG, IRT, LHB6) and M. cavernosa were relatively abundant in all habitat types except those on the shallow platform margin (IRT, LHB6, HSG) (Table 3).

Discussion
The abundance of bioeroding organisms and rates of bioerosion in coral reef areas can vary according to the coral host and other factors such as depth and cross-shelf position (Sammarco and Risk 1990, Kiene and Hutchings 1994, Tribollet and Golubic 2005), water energy regime (MacGeachy and Stearn 1976), nutrient levels (Risk and MacGeachy 1978, Hallock 1988, Hallock et al. 1993), and the degree of anthropogenic pollution (Holmes 1997, Lopez-Victoria and Zea 2005, Ward-Paige et al. 2005). Characteristics of the substratum that affect bioeroding organisms can involve the type of reef framework (Bromley 1978), substrate porosity (Neumann 1966, Highsmith et al. 1983, Sammarco and Risk 1990), coral structure, colony age, growth rate, and skeletal

from shore (r = -0.611, 0.10 > P > 0.05) and a positive correlation with scleractinian coral cover (r = 0.658, 0.10 > P > 0.05). Both distance from shore and scleractinian coral cover were themselves highly correlated (r = -0.774, P < 0.02), with significant linear (R2 = 0.600, F = 10.496, P < 0.02) and non-linear relationships (R2 = 0.919, F = 10.301, P < 0.02) (Fig. 3B). Of the environmental factors considered, depth and coral cover accounted for 37.4% and 43.3% of the

260

Table 3: Stony coral colonization (ri) by C. delitrix relative to stony coral availability (Pi) in the Florida Keys, based upon coral colony counts from 181 sites and nine habitat types. Selectivity was calculated using Ivlevs electivity index (Manly et al. 1993). fa = frequency of availability, fc = frequency of colonization, ri = proportion of stony corals colonized by C. delitrix, Pi = proportion of stony corals available. Coral species Acropora cervicornis A. palmata Agaricia agaricites A. fragilis A. humilis A. lamarcki Cladocora arbuscula Colpophyllia natans Dendrogyra cylindrus Dichocoenia stokesi Diploria clivosa D. labyrinthiformis D. strigosa Eusmilia fastigiata Favia fragum Isophyllastrea rigida Isophyllia sinuosa Leptoseris cucullata Madracis decactis M. formosa M. mirabilis M. senaria Manicina areolata Meandrina meandrites Montastraea spp. M. annularis M. cavernosa M. faveolata M. franksii Mussa angulosa Mycetophyllia spp. Mycetophyllia aliciae M. danaana M. ferox M. lamarckiana Oculina diffusa Porites astreoides P. branneri P. porites f. divaricata P. porites f. furcata P. porites f. porites Scolymia spp. Siderastrea radians S. siderea Solenastrea bournoni Stephanocoenia michelini Unidentified hard coral All species Coral availability fa Pi 99 16 2,680 106 38 29 3 368 1 538 23 88 200 194 8 2 8 71 388 24 19 178 23 80 98 67 1,244 378 69 25 73 27 86 11 2 144 3,911 1 407 645 1,122 51 369 4,946 127 3,018 1 22,006 0.0045 0.0007 0.1218 0.0048 0.0017 0.0013 0.0001 0.0167 0.0001 0.0244 0.0010 0.0040 0.0091 0.0088 0.0004 0.0001 0.0004 0.0032 0.0176 0.0011 0.0009 0.0081 0.0010 0.0036 0.0045 0.0030 0.0565 0.0172 0.0031 0.0011 0.0033 0.0012 0.0039 0.0005 0.0001 0.0065 0.1777 0.0001 0.0185 0.0293 0.0510 0.0023 0.0168 0.2248 0.0058 0.1371 0.0001 1.0000 C. delitrix colonization fc ri 0 6 20 0 0 1 0 3 0 2 0 2 3 0 0 0 0 0 4 0 0 4 0 1 2 0 40 11 3 0 0 0 0 0 0 0 109 0 0 0 5 0 2 132 4 32 0 386 0 0.0155 0.058 0 0 0.0026 0 0.0078 0 0.0052 0 0.0052 0.0078 0 0 0 0 0 0.0104 0 0 0.0104 0 0.0026 0.0052 0 0.1036 0.0285 0.0078 0 0 0 0 0 0 0 0.2824 0 0 0 0.0130 0 0.0052 0.3420 0.0104 0.0829 0 1.0000 Electivity Index (ri Pi)/(ri + Pi) -1.00 +0.91 -0.35 -1.00 -1.00 +0.33 -1.00 -0.36 -1.00 -0.65 -1.00 +0.13 -0.08 -1.00 -1.00 -1.00 -1.00 -1.00 -0.26 -1.00 -1.00 +0.12 -1.00 -0.16 +0.07 -1.00 +0.29 +0.25 +0.43 -1.00 -1.00 -1.00 -1.00 -1.00 -1.00 -1.00 +0.23 -1.00 -1.00 -1.00 -0.59 -1.00 -0.53 +0.21 +0.28 -0.25 -1.00

density (Kiene 1988, Reis and Leo 2002). Most excavating sponges such as C. delitrix prefer to bore into and excavate dead skeleton (MacGeachy 1977), so the factors that lead to coral tissue mortality such as disease and bleaching could also be responsible for influencing the distribution of this and other bioeroding sponges (Lopez-Victoria and Zea 2005).

This large-scale study of C. delitrix distribution, abundance and size in the Florida Keys illustrated complex patterns related to the physical structure of coral reef and hard-bottom habitat, in turn related to cross-shelf position, depth, and coral composition. Results from this study indicated that the greatest C. delitrix densities occurred on the deeper fore

261

Fig. 3: Relationships between mean (A) Cliona delitrix density and transect depth, (B) scleractinian coral cover and site distance from shore, and (C) C. delitrix average size and scleractinian coral cover in the Florida Keys. Error bars represent 1 SE for the independent and dependent variables. Included are the best-fit non-linear regression lines with corresponding equations and coefficients of determination (R2).

reef slope and were lowest on the shallow, relatively waveexposed platform margin. Lopez-Victoria and Zea (2005) found a similar pattern in the Caribbean and suggested that clionaid sponges may prefer recently dead coral skeletons compared to heavily encrusted substrata, especially with high relief. In the Florida Keys study area, the lowest densities and sizes of C. delitrix were associated with relatively shallow, high wave energy habitats such as hard-bottom with relatively small corals or in topographically complex spur and groove habitat dominated by algae and gorgonians. In contrast to C. delitrix densities in the Florida Keys, the largest sponge sizes were associated with patch reefs closer to shore with relatively high coral cover. Most C. delitrix encountered were relatively small in size (< 50 cm2),

confirming earlier assessments at 40 permanent monitoring stations in the Florida Keys (Calahan 2005). For the majority of the coral taxa encountered, there did not appear to be preferential colonization of particular species by C. delitrix, although there were some exceptions. In an earlier largescale study assessing Florida Keys sponge distribution patterns, Chiappone et al. (in press) observed that while C. delitrix was encountered in most coral reef and hard-bottom habitat types across the shallow (< 20 m) south Florida shelf, this species tended to be infrequently recorded from shallow (< 6 m), topographically complex spur and groove reefs compared to patch reefs and deeper low-relief spur and groove reefs, where massive frame-building corals (Diploria, Montastraea, Siderastrea) are the largest and most abundant. Calahan (2005) observed from monitoring of 40 sites in the Florida Keys that Montastraea annularis, M. cavernosa, and Siderastrea siderea were the most frequently and extensively invaded by clionaid sponges, as did a similar study (WardPaige et al. 2005). Several studies have suggested or documented bioerosion rates and density patterns of C. delitrix and other bioeroding organisms in terms of anthropogenic nutrient pollution (Alcolado 1990, Zea 1994, Holmes 1997). These studies illustrate that sponges react to nutrient enrichment, especially in terms of abundance and biomass, relative to the enrichment source (Wilkinson and Cheshire 1988, Zea 1994), but some cases are not as clear (Lopez-Victoria and Zea 2005), presumably due to greater sedimentation and hence reduction of sponge pumping (Macdonald and Perry 2003). Risk and MacGeachy (1978) predicted that bioerosion rates would increase on reefs subjected to eutrophication and found that population densities of C. delitrix were greater in areas subjected to raw sewage inputs on Grand Cayman reefs. On Barbados reefs, Holmes (1997) found that clionaid boring on Porites coral rubble was greatest at sites most impacted by coastal eutrophication. In the Florida Keys, sites with the greatest cover and size of C. delitrix and C. lampa were in areas with the highest levels of water column total nitrogen, ammonium, and 15N (Ward-Paige et al. 2005). Calahan (2005) found that clionaid area and abundance on patch reefs were significantly correlated with several water quality paramaters indicating higher nutrient flux and food resources. A clear inshore-to-offshore pattern in in C. delitrix densities in our study was not as evident, as densities, but not individual sponge sizes, were greatest on the deeper fore reef slope and only intermediate on patch reefs closer to shore. However, larger clionaid sizes were found on patch reefs closer to shore, confirming previous observations (WardPaige et al. 2005). Although our study did not attempt to correlate C. delitrix abundance with available water quality data, it is also possible that the potential availability of recently dead tissue on larger coral skeletons in the patch reef environment allows for greater clionaid sizes (LopezVictoria and Zea 2005). Moreover, as coral cover exhibits a strong direct relationship with distance from shore in the Florida Keys, it is possible that differences in water quality (e.g. more food) allow for larger and more abundant corals in the patch reef environment. However, once space becomes available due to coral mortality, clionaid sponges and other benthic organisms may colonize the available space more

262

rapidly than coral colonies can repair themselves or new coral recruits can become established. Previous studies of sponge bioerosion found that the amount of calcium carbonate material removed by boring sponges was greater on deeper reefs compared to shallower reefs (MacGeachy 1977). However, this depth-related pattern is not universal (Kiene and Hutchings 1994) and may be related to nutrient levels in the water column and the availability of coral substrates to settle and bore into. There was some indication in the Florida Keys that C. delitrix preferentially occupied the skeletons of some species, especially some of the more massive, mound-shaped corals from the genera Montastraea, Siderastrea, and Solenastrea, similar to previous observations in the Florida Keys and Caribbean (Calahan 2005, LopezVictoria and Zea 2005, Ward-Paige et al. 2005). On a fringing coral reef off the Venezuelan coast, C. delitrix was dominant between 9 m and 20 m depth in areas with larger coral colonies (Alvarez et al. 1990). Indeed, the dead basal areas or cavities of larger coral colonies compared to smaller colonies are highly susceptible to sponge bioerosion (MacGeachy 1977), suggesting that there is a direct relationship between coral substrate availability and the abundance of C. delitrix (Alvarez et al. 1990). Results from this study elucidate some of the basic spatial patterns of C. delitrix density and size among the majority of shallow-water (< 30 m) hard-bottom and coral reef habitat types in the Florida Keys. The distribution and abundance of this sponge are clearly not related to just one factor such as proximity to shore, but instead are potentially related to a suite of factors such as habitat depth and the abundance of corals. There is a continuing interest in exploring the factors affecting rates of clionaid infestation and calcium carbonate removal for this and other wider Caribbean reef systems, as there is concern that diseases and thermallyinduced bleaching may lead to the continued decline of corals in favor of other organisms (Marquez et al. 2006). Other studies of clionaid distribution indicate that reefs with high coral mortality are more susceptible to colonization and space monopolization by clionaid sponges (Lopez-Victoria and Zea 2005). Comparisons of clionaid prevalence patterns with existing water quality information are clearly of interest (Calahan 2005, Ward-Paige et al. 2005), as well as further in depth-investigations into the size and condition of corals that may render them more susceptible to bioerosion.

References
Alcolado PM (1990) General features of Cuban sponge communities. In: Rtzler K (ed). New perspectives in sponge biology. Smithsonian Institution Press, Washington, DC. pp. 351-357 Alvarez B, Diaz MC, Laughlin RA (1990) The sponge fauna on a fringing coral reef in Venezuela, I: composition, distribution, and abundance. In: Rtzler K (ed). New perspectives in sponge biology. Smithsonian Institution Press, Washington, DC. pp. 358-366 Bak RPM (1976) The growth of coral colonies and the importance of crustose coralline algae and burrowing sponges in relation with carbonate accumulation. Netherl J Sea Res 10: 285-337 Bohnsack JA, Ault JS (1996) Management strategies to conserve marine biodiversity. Oceanography 9: 73-82 Bromley RG (1978) Bioerosion of Bermuda reefs. Palaeoceanogr Palaeoclimatol Palaeoecol 23: 169-197 Calahan MK (2005) Distribution of clionid sponges in the Florida Keys National Marine Sanctuary. MS Thesis, University of South Florida, St. Petersburg Chiappone M, Rutten L, Swanson DW, Miller SL (in press) Spatial patterns of benthic coral reef organisms in the Florida Keys. III. Sponge species richness and coverage. Coral Reefs Chiappone M, Sullivan KM (1997) Rapid assessment of reefs in the Florida Keys: results from a synoptic survey. Proc 8th Int Coral Reef Symp, Balboa 2: 1509-1514 Cochran WG (1977) Sampling techniques. 3rd edition. Wiley, New York Diaz MC, Rtzler K (2001) Sponges: an essential component of Caribbean coral reefs. Bull Mar Sci 69: 535-546 FDEP (Florida Department of Environmental Protection) (1998) Benthic habitats of the Florida Keys. FMRI Technical Report TR4, St. Petersburg Glynn PW (1997) Bioerosion and coral reef growth: a dynamic balance. In: Birkeland C (ed). Life and death of coral reefs. Chapman & Hall, New York. pp 69-95 Hallock P (1988) The role of nutrient availability in bioerosion: consequences to carbonate buildups. Palaeoceanogr Palaeoclimatol Palaeoecol 63: 275-291 Hallock P, Mller-Karger FE, Halas JC (1993) Coral reef decline. Nat Geogr Res Expl 9: 358-378 Hallock P, Schlager W (1986) Nutrient excess and the demise of coral reefs and carbonate platforms. Palaios 1: 389-398 Hein FJ, Risk MJ (1975) Bioerosion of coral heads: inner patch reefs, Florida reef tract. Bull Mar Sci 25: 133-138 Highsmith RC, Lueptow RL, Schonberg SC (1983) Growth and bioerosion of three massive corals on the Belize barrier reef. Mar Ecol Progr Ser 13: 261-271 Holmes KE (1997) Eutrophication and its effect on bioeroding sponge communities. Proc 8th Int Coral Reef Symp, Balboa 2: 1411-1416 Hudson JH (1977) Long-term bioerosion rates on a Florida reef: A new method. Proc 3rd Int Coral Reef Symp, Miami 2: 491-497 Jaap WC (1984) The ecology of the south Florida coral reefs: a community profile. FWS/OBS-82/08, MMS 84-0038, Washington, DC Kiene WE (1988) A model of bioerosion on the Great Barrier Reef. Proc 6th Int Coral Reef Symp, Townsville 3: 449-454 Kiene WE, Hutchings PA (1994) Bioerosion experiments at Lizard Island, Great Barrier Reef. Coral Reefs 13: 91-98

Acknowledgements
Grants from the Florida Keys National Marine Sanctuary Program, Emerson Associates International, and NOAAs National Undersea Research Program-University of North Carolina at Wilmington supported this research. The staff of NURC, Biscayne National Park and the R/V Expedition II provided logistical support. J. Ault and S. Smith provided statistical guidance and B. Keller provided program management assistance. Permission to conducted research in the National Marine Sanctuary was conducted under permit FKNMS074-98.

263

Lidz BH, Reich CD, Peterson RL, Shinn EA (2006) New maps, new information: coral reefs of the Florida Keys. J Coast Res 22: 260282 Lopez-Victoria M, Zea S (2005) Current trends of space occupation by encrusting excavating sponges on Colombian coral reefs. Mar Ecol 26: 33-41 Macdonald IA, Perry CT (2003) Biological degradation of coral framework in a turbid lagoon environment, Discovery Bay, north Jamaica. Coral Reefs 22: 523-535 MacGeachy JK (1977) Factors controlling sponge boring in Barbados reef corals. Proc 3rd Int Coral Reef Symp, Miami 2: 477-483 MacGeachy JK, Stearn CW (1976) Boring by macro-organisms in the coral Montastrea annularis on Barbados reefs. Int Rev Ges Hydrobiol 61: 715-745 Manly BFJ, McDonald LL, Thomas DL (1993) Resource selection by animals: statistical design and analysis for field studies. Chapman & Hall, New York Marquez JC, Zea S, Lopez-Victoria M (2006) Is competition for space between the encrusting and excavating sponge Cliona tenuis and corals influenced by higher-than-normal temperatures? Bol Invest Mar Cost 35: 259-265 Marszalek DS, Babashoff G, Noel MR, Worley DR (1977) Reef distribution in south Florida. Proc 3rd Int Coral Reef Symp, Miami 2: 223-229 Miller RG (1981) Simultaneous statistical inference. SpringerVerlag, New York Miller SM, Swanson DW, Chiappone M (2002) Multiple spatial scale assessment of coral reef and hard-bottom community structure in the Florida Keys National Marine Sanctuary. Proc 9th Int Coral Reef Symp, Bali 1: 69-74 Neumann AC (1966) Observations on coral erosion in Bermuda and measurements of the boring rate of the sponge Cliona lampa. Limnol Oceanogr 11: 92-108 NOAA (National Oceanic and Atmospheric Administration) (1996) Final management plan/environmental impact statement. Volume II: Development of the management plan: environmental impact statement. NOS/SRD, Silver Spring Ohlhorst SL, Liddell WD, Taylor RJ, Taylor JM (1988) Evaluation of reef census techniques. Proc 6th Int Coral Reef Symp, Townsville 2: 791-796 Porter JW, Porter KG (2002) The Everglades, Florida Bay and Coral Reefs of the Florida Keys: an ecosystem sourcebook. CRC Press, Boca Raton

Reis MAC, Leo ZMAN (2002) Bioerosion rate of the sponge Cliona celata (Grant 1826) from reefs in turbid waters, north Bahia, Brazil. Proc 9th Int Coral Reef Symp, Bali 1: 273-278 Reiswig HM (1974) Water transport, respiration and energetics of three tropical marine sponges. J Exp Mar Biol Ecol 14: 231-249 Risk MJ, MacGeachy (1978) Aspects of bioerosion of modern Caribbean reefs. Rev Biol Trop 26: 85-105 Sammarco PW, Risk MJ (1990) Large-scale patterns in internal bioerosion of Porites: cross continental shelf trends on the Great Barrier Reef. Mar Ecol Progr Ser 59: 145-156 Schmahl GP (1990) Community structure and ecology of sponges associated with four south Florida coral reefs. In: Rtzler K (ed). New perspectives in sponge biology. Smithsonian Institution Press, Washington, DC. pp. 376-383 Shinn EA, Lidz BH, Halley RB, Hudson JH, Kindinger JL (1989) Reefs of Florida and the Dry Tortugas: field trip guidebook. T176, American Geophysical Union, Washington, DC Tribollet A, Golubic S (2005) Cross-shelf differences in the pattern and pace of bioerosion of experimental carbonate substrates exposed for 3 years on the northern Great Barrier Reef, Australia. Coral Reefs 24: 422-434 Ward-Paige CA, Risk MJ, Sherwood OW, Jaap WC (2005) Clionid sponge surveys on the Florida Reef Tract suggest land-based nutrient inputs. Mar Pollut Bull 51: 570-579 Wilkinson CR (1983) Net primary productivity in coral reef sponges. Science 219: 410-412 Wilkinson CR, Cheshire AC (1988) Cross-shelf variations in coral reef structure and function: Influences of land and ocean. Proc 6th Int Coral Reef Symp, Townsville 1: 227-233 Wulff JL, Buss LW (1979) Do sponge help hold coral reefs together? Nature 281: 474-475 Zar JH (1996) Biostatistical analysis. 3rd edition. Prentice Hall, Upper Saddle River Zea S (1994) Patterns of coral and sponge abundance in stressed coral reefs at Santa Marta, Colombian Caribbean. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 257-264 Zea S, Weil E (2003) Taxonomy of the Caribbean excavating sponge species complex Cliona caribbea C. aprica C. langae (Porifera, Hadromerida, Clionaidae). Carib J Sci 39: 348-370

Porifera research: Biodiversity, innovation and sustainaBility - 2007

265

Clarification of dictyoceratid taxonomic characters, and the determination of genera

Steve de C. Cook
Earth & Oceanic Sciences Research Institute, Auckland University of Technology, 24 St Paul Street, Private Bag 92006, Auckland, New Zealand. steve.cook@aut.ac.nz(*) Abstract: Distinguishing between dictyoceratid families is relatively straight forward, but distinguishing genera and subgenera within some of those families can be difficult. Dictyoceratid genera are characterised by the presence, structure and organisation of skeletal elements, surface characters, choanocyte chambers, collagen, cortical armour and by their general form. There are published instances of misunderstood characters leading to unnecessary and erroneous reassignments or generic mis-diagnoses. There are also situations where authors have redefined genera to accommodate new species, to the point where generic descriptions become too broad, making the character boundaries of the genus indistinct. These factors have led to confusion when attempting to allocate dictyoceratid specimens to some genera. Previously published work by the author attempted to clarify some of these problems, but the aim here is to consolidate those clarifications of dictyoceratid characters into a single paper. Keys and descriptions of dictyoceratid genera are provided. Keywords: Characters, Dictyoceratida, genera, keys, taxonomy

Introduction
Dictyoceratid families are relatively easy to distinguish, with fine collagenous filaments in the irciniids, the homogeneous skeletal fibres of spongiids, eurypylous choanocyte chambers in dysideids, and the opposite of these characters in thorectids, i.e. diplodal choanocyte chambers, pithed and laminated fibres, and an absence of fine filaments. However, distinguishing genera and subgenera within some of these families can be difficult. Genera are defined in terms of their skeletal architecture, mucus production, and whether or not they are armoured. Skeletal characters include the presence of primary, secondary or tertiary fibres, fascicular fibres, foreign coring, fibre diameter, skeletal density, collagen deposition, and general skeletal morphology and distribution. Published instances of misunderstood characters, e.g. cortical armour, have led to incorrect generic reassignments or mis-diagnoses, and situations where authors have massaged generic definitions to the point where they become ill-defined and largely meaningless within the context of the appropriate family, e.g. the thorectid Cacospongia. Taxonomic and systematic research based on New Zealand Dictyoceratida encountered a number of ambiguities in morphological characters used to distinguish between some taxa. The following three pairs of genera, one from each of three different families, provide examples where ambiguities in generic diagnoses had resulted in species being incorrectly assigned to genera, or where genera have become a catchall for species that approximate the generic characters.
* present email address: cooknz@bigfoot.com

Cacospongia vs Scalarispongia
Cacospongia, as historically diagnosed, admitted any thorectid sponge with cored primary fibres, uncored secondary fibres, and an unarmoured and finely conulose surface (Bergquist 1980, Desqueyroux-Fandez and van Soest 1997). While the New Zealand sponges could be pigeon-holed into Cacospongia, they did not quite fit. Closer inspection of Cacospongia found that the diagnosis was too loose, and that species displayed two different skeletal morphologies a well-developed secondary fibre skeleton (Fig. 1), in contrast to a simpler, more ladder-like skeleton (Fig. 2), epitomised respectively by Cacospongia mollior Schmidt, 1862 (type species of Cacospongia) and Cacospongia scalaris Schmidt, 1862. Consequently, Cacospongia was redefined and amended to provide a less ambiguous taxon, which more closely conformed to the morphology of the type species Cacospongia mollior Schmidt, 1862. A new genus, Scalarispongia, was established (Cook and Bergquist 2000) for sponges previously assigned to Cacospongia, but with a more ladder-like skeleton, with Cacospongia scalaris Schmidt, 1862 designated as the type species.

Spongia vs Hippospongia
Hippospongia was traditionally distinguished from the very similar Spongia by the relative rarity of primary fibres, the presence of large canals, subdermal lacunae and vestibules within the mesohyl, and the firmly attached dermis (de Laubenfels and Storr 1958, Vacelet 1959, Wiedenmayer 1977, van Soest 1978, Bergquist 1980). Historically, the definition of Hippospongia became too loose, largely because of the

266

Fig. 1: Cored primary fibres and well-developed secondary fibre reticulum of Cacospongia mollior (Mediterranean).

Fig. 2: Cored primary fibres and the more sparse, ladder-like skeleton of Scalarispongia scalaris (Mediterranean).

published works of von Lendenfeld, who focused closely on the presence of subdermal lacunae as the distinguishing feature of Hippospongia (see Cook and Bergquist 2001 for a full discussion). Examination of dry, macerated specimens, and published photos of Spongia (Fig. 3) and Hippospongia (Fig. 4) species, led to an emended diagnosis of Hippospongia that expands and emphasises the diagnostic characters within specified limits (Cook and Bergquist 2001).

Ircinia vs Psammocinia
Within the Irciniidae, distinguishing between Ircinia and Psammocinia can be difficult while dermal armouring characterises Psammocinia, specimens of Ircinia can have a dermal dusting of foreign debris that has been interpreted as armouring. There has been a tendency to diagnose Psammocinia solely on the basis of fine filaments and apparent armouring, without due regard for other relevant characters. This has led to some confusion, and historically some authors have chosen to ignore Psammocinia or to synonymise it with Ircinia (e.g. Wiedenmayer, 1977). For specimens that are difficult to place in either Ircinia and Psammocinia, the presence and magnitude of fascicular primary fibres is a useful character. Psammocinia species have simple primary fibres, sometimes showing moderate fasciculation (Fig. 5), whereas Ircinia species typically have massive, sometimes spectacular fascicular fibres (Fig. 6). If a specimen has a significant amount of sand in its ectosome, but does not appear to form a distinct armoured crust, and has heavily fascicular fibres, it should be classified as Ircinia. Using this method, all New Zealand specimens so far encountered have been able to be placed within one genus or the other without difficulty.

Fascicles are defined as a bundle or bunch, and the term is used to describe a fibre, usually primary, that shows multiple diverging and converging tracts within a single fibre axis; fascicles can range in size and complexity, from minor to massive; fascicles also vary in form, from a well-defined meshwork to a tangled mass (Figs. 5 and 6, 9 and 10). Secondary webs are broad, web-like structures stretched between primary fibres where secondary fibres would normally be seen, that are distinct from a fascicle; secondary webs resemble stretched plastic, may be entire or perforate, and may also be observed in the V-shaped space created where a single primary fibre bifurcates into two diverging primary fibres (Figs. 9 and 10).

Skeletal fibre types

Dictyoceratids have an anastomosing fibre skeleton, usually organised into a hierarchy reflecting size and orientation: Primary fibres are typically orientated at right angles to the surface; distally they usually terminate at the sponge surface, and support conules in those species that have them; they may be simple, coalescing or fascicular, and they may be axially to fully cored with foreign inclusions, e.g. sand, spicules (Figs. 7-10, 12). Secondary fibres interconnect primary fibres; they are simple or compound, and at their simplest, they resemble the rungs of a ladder; they may be axially to fully cored with foreign inclusions (Figs. 7-9, 11-12). Tertiary fibres typically interconnect secondary fibres; they are uncored, usually very fine, and are recognised as fibres of very small diameter in relation to secondary fibres, usually forming a fine mesh-work within the meshes of the secondary reticulum, e.g. as seen in Luffariella (Fig. 11); in one genus, Carteriospongia, they are more vermiform and wandering. Pseudo-tertiary fibres are finer than secondary fibres, but are not as fine as those typically called tertiary fibres. While this name is inadequate, the authors (Cook and Bergquist 2001) considered it best to use an interim term, until such time as

Clarification of dictyoceratid characters Fibre forms

Simple fibres are essentially undivided fibres (Fig. 7). Coalescing fibres are created where two or more fibres converge and coalesce into a single, often larger fibre (Fig. 8).

267

Fig. 3: Surface of macerated Spongia officinalis.

Fig. 4: Surface of macerated Hippospongia communis.

Fig. 5: Lightly fasciculated fibre of Psammocinia hawere (New Zealand).

Fig. 6: Heavily fasciculated fibre of Ircinia irregularis (Torres Strait).

Fig. 7: Simple cored primary fibres and uncored secondary fibres of Thorecta reticulata (New Zealand).

Fig. 8: Coalescing primary fibre of Spongia gorgonocephalus (New Zealand); sand in dermis is inconsistent over sponge surface.

268

Fig. 9: Cored primary fibre fascicles and perforate secondary webbing of Ircinia subaspera (New Zealand).

Fig. 10: Cored primary fibre fascicles and secondary webbing of Ircinia aucklandensis (New Zealand).

Fig. 11: Secondary and fine tertiary fibre network of Luffariella variabilis (Gt Barrier Reef).

Fig. 12: Cored primary fibre, secondary fibre network and thinner pseudo-tertiary fibres of Spongia cristata (New Zealand).

their diagnostic significance within the relevant family can be more accurately determined (Fig. 12). Fine filaments are not skeletal fibres, but they add support and strength to irciniid sponges. They are long, thin collagenous threads that may be relatively sparse, to forming thick bundles within the mesohyl. Filaments are easily recognisable by their swollen tips (Figs. 13 and 14).

Fibre construction
The construction of skeletal fibres is used as a familial character, principally, whether or not fibres are laminated and pithed. Primary and secondary fibres are laminated (Fig. 15) in the Dysideidae, Irciniidae and the Thorectidae. The primary fibres of these sponges are also pithed, though this can be difficult to see, particularly in those genera with a core of foreign material. Unlike the verongids, the pith is somewhat diffuse, and blends with the surrounding fibre,

rather than having a distinct boundary between pith and fibre. Note however that in some specimens fibre laminations can be difficult to see. When this occurs, look carefully for evidence of laminations and pith in the primary fibres, particularly near the sponge surface, and if necessary where there is a gap in foreign coring material (Fig. 15). In some cases, e.g. Psammocinia halmiformis (Fig. 16), this may not be possible and other characters have to be employed. This highlights the importance of knowing the morphological range of taxonomic characters. In spongiids, the skeletal fibres are not laminated when viewed under a light microscope, i.e. they are homogeneous, and are unpithed.

Dermal armour
There are clearly some genera that are able to produce an organised armoured dermal layer (Fig. 16). There is evidence that some species exercise active selection of particles for

269

Fig. 13: Fine filaments with swollen tips (arrowed), characteristic of the Irciniidae, from Psammocinia beresfordae (New Zealand).

Fig. 14: Detail of fine filaments from Ircinia aucklandensis (New Zealand), with swollen tips arrowed. The granular appearance of filaments is due to a coating of lepidocrocite granules.

their armour. For example, Psammocinia halmiformis (Fig. 16) and Coscinoderma spp. (Fig. 17) have a uniformly finegrained surface armour, whereas other species consistently use coarse particles. The transport of foreign particles in sponges has been studied in the dysideid Dysidea etheria (Teregawa 1986a, 1986b). These studies demonstrated active transport of foreign particles to specific sites on the surface and within the sponge body. Observations suggested that coordinated migration of groups of mesohyl cells control particle transport to conules and that patterns of cell migration are associated with the structural organization of the dermal membrane (Teregawa 1986b). However, newly grown areas may have thinner or poorly-developed sand crusts that may confound diagnosing the presence of a dermal armour. This remains the subject of a separate study. Note also the work of Cerrano et al. (2007), on how and why sponges incorporate foreign material.

Mesohylar collagen
The mesohyl of dictyoceratid genera includes collagen in varying degrees of density. Thin section microscopy, with a suitable stain, e.g. Mallory-Heidenhain, can be used to determine the relative volume of collagen that occurs in a specimen. This is used as a character in some genera, e.g. Aplysinopsis, or to assist in distinguishing between some otherwise similar genera, e.g. Scalarispongia and Semitaspongia (Fig. 18).

Fig. 15: Laminated primary fibre of Thorecta reticulata (New Zealand), showing a section of pith where there is a break in foreign coring.

type of choanocyte chambers, i.e. whether they are diplodal or eurypylous. Important: the keys below are not suitable for use on their own and should, at a bare minimum, be used in conjunction with the generic descriptors that follow.
Key to dictyoceratid families 1. Fine filaments absent ...................................................................2 Fine filaments present .................................................... Irciniidae 2. Skeletal fibres concentrically laminated ......................................3 Skeletal fibres homogeneous, without laminations .............................................................. Spongiidae 3. Choanocyte chambers small and spherical (diplodal) ................................................................ Thorectidae

Determination of dictyoceratid families and genera

Dictyoceratida incorporates four families that are relatively easy to distinguish from one another. When producing histological slides of specimens, it is useful to take thin sections (12 m is a good thickness), as well as thicker ones (often hand sections), to assist in the diagnosis. Stained with Mallory-Heidenhain, or similar, these thin sections facilitate determining the proportion of collagen in the sponge, and the

270

Fig. 16: Heavily armoured dermis and cored primary fibre of Psammocinia halmiformis (New South Wales).

Fig. 17: Armoured dermis and cored primary fibre of Coscinoderma sp. (Chuuk, FSM).

Choanocyte chambers large and oval (eurypylous)..............................................................Dysideidae

Dysideidae Dictyoceratida with laminated skeletal fibres and eurypylous choanocyte chambers. Irciniidae Dictyoceratida with unique fine collagenous filaments in the mesohyl, and diplodal choanocyte chambers. Spongiidae Dictyoceratida with homogeneous skeletal fibres, i.e. without distinct laminations, a skeleton dominated by sub-primary fibres, and diplodal choanocyte chambers. Thorectidae Dictyoceratida with laminated skeletal fibres and diplodal choanocyte chambers.

Family Irciniidae
This family is characterised by the presence of fine collagenous filaments within the body of the sponge. Filaments occur in varying density and organisation, from scattered through the mesohyl, to forming thick bundles. In the past, the presence of very thin filaments (0.5-5 m) has been used as a diagnostic character for Sarcotragus (Vacelet 1959, Bergquist 1980), but subsequently, very thin filaments have also been observed in some species of Psammocinia from New Zealand, Australia and South Korea (Cook and Bergquist 1996, 1998, Sim and Lee 1998).
Key to irciniid genera 1. Dermis unarmoured .....................................................................2 Dermis armoured with an organised layer ..............Psammocinia 2. Primary fibres form minor fascicles, lightly cored or uncored .......................................................................3 Primary fibres form massive fascicles, cored with foreign debris...........................................................Ircinia 3. Primary fibres form minor fascicles, lightly cored or uncored ..................................................... Sarcotragus Primary fibres form minor uncored fascicles, with secondary web ................................................. Bergquistia Fig. 18: Stained thin section of Semitaspongia incompta (New Zealand), with heavy collagen deposition, showing several small, spherical diplodal choanocyte chambers (arrowed). Collagen is a lot easier to observe when selectively stained (and in colour)

Bergquistia unarmoured irciniids, with slightly fascicular and uncored primary fibres, though there may be some coring near the surface. In some areas, primary fibres are connected by secondary webs (Sim and Lee 2002). The distinction between this genus and Sarcotragus requires clarification or revision. Ircinia unarmoured irciniids with massive fascicular primary fibres, that are usually cored with foreign debris. Primary fibres may be connected by secondary webs. Psammocinia armoured irciniids, with minor fascicles, usually near the surface. Primary fibres are cored and secondary fibre coring is variable, from uncored to fully cored. Primary fibres may be connected by secondary webs. Sarcotragus unarmoured irciniids in which the primary fibres are simple or form relatively minor fascicles, compared

271

to Ircinia, and are uncored or have only light, intermittent coring.

species may also have a superficial fibre net supporting the pinacoderm.

Family Spongiidae
Spongiids have a well-developed skeleton of primary, secondary, and in a group of Australasian sponges, distinct fine secondary or pseudo-tertiary fibres. Primary fibres can be sparse. All fibres are unpithed, and are homogeneous when viewed with light microscopy, i.e. they show little or no sign of concentric laminations within the fibres. Sometimes very fine laminations can be seen in the fibres, typically at stress points, e.g. fibre junctions, but they are tightly adherent, and not as obvious as seen in other dictyoceratid families. Spongiids are characterised by a dense secondary fibre reticulum that dominates the skeleton, though this character is not unique to spongiids. Choanocyte chambers are diplodal, and spherical to oval in shape. In some species, the mesohyl and ectosome are supported by heavy deposits of collagen.
Key to spongiid genera 1. Dermis armoured .........................................................................2 Dermis unarmoured .....................................................................3 2. Dense secondary skeleton of thick, branching secondary fibres ...........................................................Leiosella Dense secondary skeleton of very fine, intertwined fibre network ................................... Coscinoderma 3. Sponge body or ectoderm lacunose .............................................4 Sponge not lacunose ....................................................................5 4. Primary fibres common ..................................................Hyattella Primary fibres uncommon to rare ...........................Hippospongia 5. Primary fibres form long, minor fascicles ...............Rhopaloeides Primary fibres are simple, not forming fascicles ............. Spongia

Family Thorectidae
Dictyoceratida with an anastomosing skeleton of concentrically laminated primary, secondary, and sometimes tertiary fibres. Primary fibres have an axial, granular pith that may extend into secondary fibres. Foreign material is often incorporated as a core within the fibres, obscuring the pith when coring material is abundant. The anastomosing fibre skeleton is regular and varies in arrangement from rectangular to disorganised. Fibres range in form, from simple fibres to strong, complex fascicles. Primary fibres may be reduced, and are not apparent in one genus. Zones of disjunction between successive fibrous layers remain tightly adherent, producing an overall solid structure with visible contiguous laminae, in contrast to the skeletal fibres of some dendroceratids, where fibre laminations are not tightly adherent and may gape slightly in histological sections. Thorectids have small, spherical and diplodal choanocyte chambers. The cortex may be armoured with foreign material, but when not armoured, the surface is always coarsely to micro-conulose. There are two subfamilies in the Thorectidae, Thorectinae and Phyllospongiinae.
Key to thorectid subfamilies 1. Foliose, thin lamellate or folio-digitate; tertiary fibres present (except in one genus) ...............Phyllospongiinae Variable form; tertiary fibres in only two non-folio-lamellate genera ..................................... Thorectinae

Coscinoderma thinly armoured spongiid, with a dense skeleton of simple, cored primary fibres and dominated by characteristic very fine, meandering, uncored secondary fibres. Hippospongia these unarmoured spongiids have large diameter oscules and associated canals, rendering the sponge body lacunose. Primary fibres are uncommon to rare, and surface conules are usually produced by a tuft of emergent fibres. Hyattella unarmoured spongiid, with a lacunose body. The fibre skeleton consists of primary fibres, which are cored with foreign material, a dense regular network of uncored secondary fibres, plus a fine, dermal fibre network. Leiosella lightly armoured spongiids, with lightly cored primary fibres, and a dense secondary skeleton of characteristically thick, uncored fibres. Rhopaloeides massive, upright sponges, with an unarmoured surface, bearing thick tuberculate conules. These sponges could easily be mistaken for Spongia species, but are distinguished by the presence of fascicular primary fibres, particularly near the sponge surface. Spongia unarmoured spongiids, with relatively few simple, cored primary fibres and uncored secondary fibres. Some

Phyllospongiinae foliose, thin lamellate or folio-digitate, sponges, with tertiary fibres in the skeleton (except in one genus). Thorectinae variable form. Fibre skeleton comprised of primary and secondary fibres, except for Luffariella and Fenestraspongia, which also have tertiary fibres but are not foliose, lamellate or folio-digitate.
Key to Thorectinae genera 1. Dermis armoured .........................................................................2 Dermis unarmoured .....................................................................6 2. Armour moderate to heavy and consistent over whole sponge ............................................................................3 Armour light, patchy, restricted to specific areas (may form crust, not armour) ...................................................4 3. Large diameter fibres; excess mucus when live ......Thorectandra Fibres not of large diameter; without excess mucus....... Thorecta 4. Massive forms, not lamello-digitate or foliose ............................5 Thin-walled lamello-digitate or foliose sponges; upper part of sponge may have sand crust; primary fibres parallel to surface; surface fibre network present .................................................... Collospongia 5. Hard and incompressible; dense secondary skeleton..............................................................Petrosaspongia Firm, compressible and collagenous ........................ Aplysinopsis 6. Fine tertiary fibres supplement fibre skeleton .............................7 Without tertiary fibres ..................................................................8

272

7. Fenestrate or ridged surface; strongly fascicular primary fibres ..................................................Fenestraspongia Conulose surface; primary fibres simple (vague fascicles may be seen) .............................................. Luffariella 8. Uncored primary and secondary fibres ........................................9 Primary or secondary fibres with core of foreign debris ...........12 9. Dense, branching fibre network .................................................10 Axially-concentrated skeleton ................................... Thorectaxia 10. Surface with coarse irregular conules ......................................11 Surface microconulose, though may appear smooth ................................................................... Narrabeena 11. No distinction between primary and secondary fibres ................................................................ Dactylospongia Ascending primary fibres and a network of secondary fibres................................................. Smenospongia 12. Secondary fibres uncored .........................................................13 Secondary fibres heavily cored ........................................Hyrtios 13. Primary fibres strongly fascicular throughout sponge..............14 Primary fibres not strongly fascicular (slight subsurface fascicles possible) ...............................................15 14. Very thick fibres; collagenous throughout mesohyl ......................................................... Fascaplysinopsis Thick fibres; prominent central exhalant canals and subdermal lacunae; heavy dermal collagen ............................................................. Fasciospongia 15. Low forms ................................................................................16 Upright form, on a basal stalk ........................................ Taonura 16. Primary and secondary fibres in approximately equal proportion ....................................................................17 Secondary fibre skeleton well-developed............... Cacospongia 17. Heavily collagenous mesohyl; irregular skeleton ............................................................. Semitaspongia Low to moderate mesohyl collagen; very regular skeleton ................................................ Scalarispongia

Aplysinopsis thinly armoured thorectid, with a relatively sparse fibre skeleton of simple, cored primary fibres and an irregular network of uncored secondary fibres. The mesohyl of these species is characteristically collagenous. Cacospongia unarmoured thorectid, with a well-developed branching uncored secondary fibre skeleton, cored primary fibres, and low to moderate collagen deposition. Collospongia unarmoured thorectids, but the surface may be encrusted with sand patches. The skeletal characters of this genus are unique within the Dictyoceratida, particularly the arrangement of the primary fibres, and the presence of a tangential dermal reticulum. Lightly cored primary fibres are arranged parallel to the surface, within the lamellae of the sponge, with primary fibre fascicles and secondary fibres curving out towards both surfaces. Near the surface, fascicles divide into finer fibres, forming a tangled network that supports a regular dermal fibre network. Dactylospongia unarmoured thorectids, with a relatively dense skeleton of uncored fibres and no clear distinction between primary and secondary fibres. Fascaplysinopsis unarmoured thorectids, with a sparse skeleton of relatively large diameter, cored primary and uncored

secondary fibres, producing coarse, widely-spaced conules at the surface. Primary fibres can be either fasciculate or laddered. The mesohyl is characteristically gelatinous or fleshy. Fasciospongia unarmoured, and with strong central or subdermal exhalant canals. The skeleton consists of relatively large diameter primary and secondary fibres. Primary fibres are cored and fascicular, secondary fibres are uncored, and there are heavy deposits of ectosomal collagen, usually including a dermal band. Fenestraspongia unarmoured thorectids, with a fenestrate surface. The skeletal reticulum is comprised of heavy primary and secondary fibres, and very fine tertiary fibres. Primary fibres are cored with foreign material and fascicular, while secondary and tertiary fibres are uncored and form a relatively dense reticulum. Hyrtios unarmoured thorectids with heavily cored primary and secondary fibres. Luffariella unarmoured thorectids with a skeletal reticulum of simple cored primary (though coring thin in one species), and uncored secondary and fine tertiary fibres. Narrabeena unarmoured thorectids with a well-developed skeletal reticulum of uncored fibres, with no clear distinction between primary and secondary elements. These sponges also have a micro-conulose surface with a slimy texture, and light collagen deposition throughout the sponge. Petrosaspongia unarmoured thorectids, with a dense secondary reticulum of uncored fibres, rendering these sponges hard and incompressible. Primary fibres are limited in number, but are present in the ectosome where they are formed by converging secondary fibres. Primary fibres merge into a fenestrated spongin plate, from which primary elements extend to the surface. Scalarispongia unarmoured thorectids, with a regular, rectangular fibre skeleton of simple cored primary fibres and uncored secondary fibres that occasionally form light webbing. There is moderate collagen deposition in the mesohyl. Semitaspongia unarmoured thorectids, with an irregular to regular skeletal reticulum, slightly fascicular, cored primary fibres and uncored secondary fibres. There is moderate to abundant collagen deposition in the mesohyl. Smenospongia unarmoured, with a characteristic honeycombed surface. The relatively dense secondary fibre skeleton may sometimes obscure the primary fibres. There is only minor collagen deposition in the mesohyl. Taonura soft, unarmoured thorectids, upright in form with a basal stalk. They possess a regular skeleton of cored primary and uncored secondary fibres. Thorecta armoured thorectids, with a regular skeleton of simple cored primary and uncored secondary fibres, fine to moderate in diameter. Thorectandra armoured thorectids, with simple, large diameter, cored primary and uncored secondary fibres. This genus characteristically exudes copious amounts of mucus when collected.

273

Thorectaxia unarmoured thorectids, with a characteristic axially concentrated skeleton, and uncored, loosely laminated fibres.
Key to Phyllospongiinae genera 1. Dermis moderately to heavily armoured .....................................2 Dermis unarmoured, or with patchy sand crust (not armoured) ..........................................................................4 2. Tertiary fibres supplement fibre skeleton.....................................3 Without tertiary fibres; bright white ................... Candidaspongia 3. Stellate pattern of exhalant canals around each oscule ....................................................... Strepsichordaia Without stellate pattern of exhalant canals ......... Carteriospongia 4. Sand crust (not armour) on one or both faces; primary fibres perpendicular to surface; tough and flexible ............................................... Phyllospongia Fasciculate primary fibres; primary fibres meandering; soft and fleshy...................................Lendenfeldia

3. Encrusting, massive or branching, very soft and collapsible ....................................................... Euryspongia Lamellate or digitate, dense, soft and pliable .................. Citronia 4. Sponge thin, encrusting, fragile, with secondary skeleton sparse or absent .......................................Pleraplysilla Sponge massive, with secondary fibre skeleton well-developed ............................................................. Dysidea

Candidaspongia heavily armoured thorectids, that are characteristically brilliant white in colour. These sponges have simple, cored primary fibres and uncored secondary fibres, but do not have tertiary fibres. Carteriospongia heavily armoured thorectids, with an undulating surface. The fibre skeleton is relatively dense, and is comprised of heavily cored primary fibres, cored or uncored secondary fibres, that can be difficult to identify, and a tangled network of vermiform tertiary fibres. Lendenfeldia unarmoured thorectids, producing lamellate forms. These sponges have tertiary skeletal fibres, and have a soft, fleshy consistency. Phyllospongia unarmoured Phyllospongiinae, but with a sand crust. Regular fibre skeleton of primary, secondary and vermiform tertiary fibres. Strepsichordaia heavily armoured thorectids, of lamellate or foliose forms, with a characteristic stellate exhalant and oscular system. The skeleton is dominated by tertiary fibres, and these sponges are firm, tough and flexible.

Citronia Dysideidae with cored primary fibres and uncored secondary fibres. Consistency is soft, but dense and pliable. Dysidea Dysideidae in which primary and secondary fibres are cored with foreign material. Euryspongia Dysideidae in which the primary fibres are cored and the secondaries are clear of foreign coring. Secondary fibres form a well-developed reticulum, which has been likened in appearance to the development seen in Spongia. Lamellodysidea massive, lamellate to digitate dysideids, with a thin, encrusting basal plate; the skeleton is irregular, without clear distinction between primary and secondary fibres; all skeletal fibres are cored with foreign material. Pleraplysilla encrusting Dysideidae with cored fibres, and a secondary skeleton that, where present, is weak.

Discussion
When determining dictyoceratid genera, the easiest way to minimise errors and avoid confusion is to ensure that all diagnostic characters required for generic diagnosis are present, not just the obvious ones if they are absent, determine why. If in doubt, do your homework understand the character ranges for any given taxa, as some genera or species can be problematic to distinguish accurately, and a successful diagnosis can be confounded by the observer not being fully aware of character ranges. While these suggestions may seem like stating the obvious to some, examples of these errors entering the literature still occur. Not intended as an example of the errors mentioned above, but as a taxonomic update, in recent research Schmitt et al. (2005) proposed that Smenospongia be reassigned to the Aplysinidae. The genus Smenospongia Wiedenmayer, 1977 was originally described to accommodate Aplysina aurea Hyatt, 1875 within the Aplysinidae. Smenospongia has characters that align it with both the Thorectidae (Dictyoceratida) and the Aplysinidae (Verongida), with an anastomosing fibre skeleton of primary and secondary fibres, and displaying a pronounced very dark colour change on death or exposure to air, respectively. Van Soest (1978) considered that Smenospongia was an aplysinid, but with remarkable similarities to some dictyoceratid genera. Subsequently, Bergquist (1980) placed Smenospongia in the Thorectidae, based on the presence of secondary metabolites identical to those found in other thorectids, and which had never been found in members of the Verongida. Also, all tested verongids yielded bromotyrosine derivatives, and were consequently considered to characterise the order, but they had not been recorded from any other taxa, including Smenospongia. However, recent research suggests reassigning Smenospongia back to the Aplysinidae, as proposed by

Family Dysideidae
Dictyoceratida with an anastomosing skeleton of concentrically laminated primary and secondary fibres. The skeletal fibres are pithed, and may be uncored to fully cored with foreign material, the latter situation obscuring details within the fibres. Choanocyte chambers are large, usually oval, and eurypylous. These sponges are typically soft and compressible, though are rendered firmer by the presence of foreign interstitial material, e.g. sand.
Key to dysideid genera 1. Primary and secondary fibres cored with foreign debris .............2 Primary fibres cored, secondary fibres clear................................3 2. Skeleton regular, with primary fibres perpendicular to surface ..................................................................................4 Skeleton irregular, without a clear hierarchy of primary and secondary fibres ........................... Lamellodysidea

274

Schmitt et al. (2005), based on molecular research. In addition, bromotyrosine derivative compounds have now been found in other non-keratose sponge taxa, e.g. Agelas spp. (van Soest and Braekman 1999) and Jaspis spp. (Kim et al. 1999), and are no longer considered exclusive to the verongids, as evidenced by the review of sponge chemosystematics by Erpenbeck and van Soest (2006). Note also that the new species of Smenospongia described from Korea by Lee and Sim (2005) is clearly not a member of this genus. The published images show a pale alcoholpreserved specimen, where Smenospongia would have turned a very dark colour, and the fibre skeleton bears a very strong resemblance to that of Cacospongia (see Fig. 1 above). Given the difficulty associated with dictyoceratid taxonomy, and the very real possibility of specimen contamination, mislabeling, and other confounding issues, it is suggested that Smenospongia remain within the Thorectidae, until such time as all Smenospongia species are thoroughly reviewed morphologically, and verified as species of Smenospongia.

Acknowledgements
To the Auckland University of Technology (AUT) and the Faculty of Health & Environmental Sciences Travel Grant, for assistance with travel expenses to the Seventh International Sponge Symposium (Bzios, Brazil); Brid Lorigan and Olive Sykes (AUT) for assistance with numerous mundane requests, and Debbie Blake, Claire Stockman and Steve Henry for access to microscopy equipment; Brent Beaumont (Department of Anatomy, University of Auckland) for micrographs and Emma Beatson (AUT) for the underwater photographs, used in the conference presentation. This manuscript was improved by the valuable comments of two anonymous referees.

References
Bergquist PR (1980) A revision of the supraspecific classification of the orders Dictyoceratida, Dendroceratida and Verongida (class Demospongiae). NZ J Zool 7: 443-503 Cerrano C, Calcinai B, Di Camillo CG, Valisano L, Bavestrello G (2007) How and why do sponges incorporate foreign material? Strategies in Porifera. In: Custdio MR, Lbo-Hajdu G, Hajdu E, Muricy G (eds). Porifera research: biodiversity, innovation and sustainability. Srie Livros 28. Museu Nacional, Rio de Janeiro. pp. 239-246 Cook S de C, Bergquist PR (1996) New species of dictyoceratid sponges (Porifera: Demospongiae: Dictyoceratida) from New Zealand. NZ J Mar Freshw Res 30: 19-34

Cook S de C, Bergquist PR (1998) Revision of the genus Psammocinia (Porifera: Demospongiae: Dictyoceratida), with six new species from New Zealand. NZ J Mar Freshw Res 32: 399-426 Cook S de C, Bergquist PR (2000) Two new genera and five new species of the Cacospongia group (Porifera: Demospongiae: Dictyoceratida) from New Zealand, and a proposed subgeneric structure. Zoosystema 22(2): 383-400 Cook S de C, Bergquist PR (2001) New species of Spongia (Porifera: Demospongiae: Dictyoceratida) from New Zealand, and a proposed subgeneric structure. NZ J Mar Freshw Res 35: 33-58 de Laubenfels MW, Storr JF (1958) The taxonomy of American commercial sponges. Bull Mar Sci Gulf Caribb 8(2): 99-117 Desqueyroux-Fandez R, van Soest RWM (1997) Shallow water demosponges of the Galpagos Islands. Rev Suisse Zool 104(2): 379-467 Erpenbeck D, van Soest RWM (2006) Status and perspective of sponge chemosystematics. Mar Biotech 9(1): 2-19 Kim DY, Lee IS, Jung JH, Yang SI (1999) Psammaplin A, a natural bromotyrosine derivative from a sponge, possesses the antibacterial activity against methicillin-resistant Staphylococcus aureus and the DNA gyrase-inhibitory activity. Arch Pharmacol Res 22:25-29 Lee KJ, Sim JS (2005) A new sponge of the genus Smenospongia (Dictyoceratida: Thorectidae) from Gageodo Island, Korea. Integr Biosc 9: 9-11 Schmitt S, Hentschel U, Zea S, Dandekar T, Wolf M (2005) ITS2 and 18S rRNA gene phylogeny of Aplysinidae (Verongida, Demospongiae). J Mol Evol 60: 327-336 Sim CJ, Lee KJ (1998) New species of two Psammocinia horny sponges (Dictyoceratida: Irciniidae) from Korea. Korean J Syst Zool 14(4): 335-340 Sim CJ, Lee KJ (2002) A new genus in the family Irciniidae (Demospongiae: Dictyoceratida). Korean J Biol Sci 6(4): 283-285 Teregawa C (1986a) Particle transport and incorporation during skeleton formation in a keratose sponge: Dysidea etheria. Biol Bull 170: 321-334 Teregawa C (1986b) Sponge dermal membrane morphology: histology of cell-mediated particle transport during skeletal growth. J Morph 190: 335-347 Vacelet J (1959) Rpartition gnrale des ponges et systmatique des ponges cornes de la rgion de Marseille et de quelques stations mditerranennes. Rec Trav Sta mar Endoume 16(26): 39101 van Soest RWM (1978) Marine sponges from Curacao and other Caribbean localities. Part 1. Keratosa. Stud Fauna Curacao Caribb Isl 56(179): 1-94 van Soest RWM, Braekman JC (1999) Chemosystematics of Porifera: a review. Memoir Queensl Mus 44: 569-589 Wiedenmayer F (1977) Shallow-water sponges of the western Bahamas. Experientia Suppl 28: 1-287

Porifera research: Biodiversity, innovation and sustainaBility - 2007

275

A new species of Stelletta (Astrophorida: Demospongiae) with a redescription and distribution range expansion for Stelletta kallitetilla in the Southwestern Atlantic Region
Bruno Cosme(1*), Solange Peixinho(2)
(1) (2)

Museu Nacional, UFRJ. Quinta da Boa Vista, s/n, 20940-040, Rio de Janeiro, RJ, Brazil. brunoignis@yahoo.com.br Universidade Federal da Bahia, Departamento de Zoologia, Campus Universitrio de Ondina. Salvador, Bahia, Brazil

Abstract: This study describes Stelletta soteropolitana sp. nov and redescribes Stelletta kallitetilla (de Laubenfels, 1936) from Northeastern Brazil. Unlike other Stelletta, the new species has two categories of oxeas, one type of protriaenes, and one of ortho- to plagiotriaenes. The distribution range of Stelletta kallitetilla is amplified to the Southwestern Atlantic and a key for the Brazilian species of Stelletta is provided. Keywords: Ancorinidae, Porifera, Bahia, Stelletta soteropolitana sp. nov., Stelletta kallitetilla

Introduction
The genus Stelletta Schmidt, 1862 is represented in the Brazilian coast by seven species, with four from the Northeastern region: Stelletta anancora (Sollas, 1886), S. crassispicula (Sollas, 1888) and S. anasteria Esteves and Muricy, 2005 from Bahia, and S. gigas (Sollas, 1886) from St. Paul Rocks. Stelletta beae Hajdu and Carvalho, 2003 occurs in the Southeastern region, S. hajdui Lerner and Mothes, 1999 is the single species from deep waters, and Stelletta ruetzleri Mothes and Silva, 2002 has been found in the Brazilian Southern region. Stelletta purpurea Ridley, 1884 described from Australia shallow waters was recorded by Mothes-de-Moraes (1985) and Mothes and Lerner (1994) in Rio de Janeiro and Santa Catarina states. This species is now considered by Hajdu and Carvalho (2003) as conspecific with Stelletta beae. In the present study we describe a new species, Stelletta soteropolitana sp. nov., and redescribe a new record, Stelletta kallitetilla (de Laubenfels, 1936), both from Bahia state in Northeastern Brazil.

maximum (SD=standard deviation) (N=number of spicules measured). For the triaenes the data are always presented in the order: cladi (length, measured of base to point) and width; rhabdome (length and width) and asters (diameter). Abbreviations used in this text are: UFBA (Universidade Federal da Bahia), POR (Porifera Collection).

Results
Class Demospongiae Sollas, 1885 Order Astrophorida Sollas, 1888 Familly Ancorinidae Schmidt, 1870 Genus Stelletta Schmidt, 1862 Stelletta soteropolitana sp. nov. (Figs. 2A, 3, 5A) Holotype: UFBA500-POR, Mont-Serrat (1255.62 S 3831.19 W), Salvador, Bahia state, Brazil, <10 m depth, collector M. Viotto, 7.X.1983. Diagnosis: Stelletta with two categories of oxeas, one type of protriaenes and ortho- to plagiotriaenes. Description: The specimen is an irregular piece with 46.5 x 39 x 19.5 mm in dimensions, beige in color when alive and dark salmon when preserved in ethanol. Sponge of hard consistency, only slightly compressible. Hispid surface. Oscules and pores not observed. Skeleton: Ectosome about 500 m in thickness, formed by the cladomes of triaenes and lined in its inner portion by many subectosomal aquiferous canals. Choanosome composed of

Material and Methods

The samples examined in this study were collected in Mont Serrat in 1983 and the Port Authority breakwater (Capitania dos Portos, Fig. 1) in 1988, from a depth of approximately 10 m. These places are located in Todos os Santos Bay (Salvador, Bahia, Brazil), where rocky coasts and mangroves predominate. Dissociated spicules and thick section mounts were obtained according to the protocols suggested by Mothes-de-Moraes (1985). The micrometric data, length and width, are presented in the following format: minimum-mean-

276

Fig. 1: Map of Todos os Santos Bay showing the location of the collecting sites. A. Mont Serrat, type locality of Stelletta soteropolitana sp. nov. B. Port Authority breakwater (Capitania dos Portos), Salvador, collecting site of S. kallitetilla.

Fig. 2: Specimens studied. A. UFBA500-POR Holotype of Stelletta soteropolitana sp. nov., scale bar 1 cm. B. UFBA949-POR S. kallitetilla, scale bar 1 cm.

277

Fig. 3: Megasclere set of Stelletta soteropolitana sp. nov. A. Oxea 1. B. Oxea 2. C. Orthotriaene. D. Plagiotriaene (variation). E. Protriaene. Scale bar 100 m.

(N=30). Rhabdome variable in size, usually with thin pointed tip, rarely strongyloid 1538-1811.5-2143 (SD=231.8) (N=30) / 34-41.4-49 (SD=10) (N=30). Protriaenes Cladome and rhabdome extremely variable in size, the first one thick; both sharply pointed. Cladome 136-153.4-170 (SD=13.5) (N=30) / 51-64-77 (SD=11) (N=30). Rhabdome 780-990-1222 (SD=177.4) (N=30) / 26-28.9-34.1 (SD=4) (N=30). Microscleres. Acanthoxyasters minute spicules, with thin, spined rays 8-12-17 (SD=0.8) (N=30). Geographic Distribution: Brazil (Bahia state). Etimology: The name soteropolitana (Soterion, greek, refers to Saviour; Polis, greek, refers to city), means from the city of the Saviour and is derived from the species occurrence, so far restricted to Salvador, Bahia state. Discussion: This species was collected from the shallow littoral region in the early 1980s and was not found again. Its spicule composition consists of two types of oxeas, differentiated by their form and thickness. The sample was compared with Stelletta species from Western Atlantic and West coast of Africa. This species differs from Stelletta carolinensis (Wells, Wells and Gray, 1960), S. digitifera (Lvi, 1959), S. fibrosa (Schmidt, 1870), S. globulariformis (Wilson, 1902), S. grubei Schmidt, 1862, S. hispida sensu Lvi, 1960, S. incrustata (Uliczka, 1929), S. kallitetilla (de Laubenfels, 1936), S. paucistellata (Lvi, 1952), S. pudica (Wiedenmayer, 1977), S. pumex (Nardo, 1847), S. stenospiculata Ulickza, 1929 and S. tenuispicula (Sollas, 1886) because it has two categories of oxeas. When the protriane and orthotriane/plagiotriane combination is compared, Stelletta soteropolitana sp. nov. differs from S. anancora (Sollas, 1886), S. carolinensis, S. crassispicula (Sollas, 1886), S. debilis (Thiele, 1900), S. digitifera (Lvi, 1959), S. fibrosa, S. gigas (Sollas, 1886), S. globulariformis, S. grubei, S. hispida (Buccich, 1886), S. kallitetilla, S. paucistellata, S. pudica, S. pumex, S. stenospiculata, S. tenuispicula. Only one species, S. incrustata, has a similar set of triaenes; nevertheless, it differs from the new species by the presence of a single category of oxeas, measuring 990-1250/ 17.5-28 m. Therefore, we propose to consider the specimen described here as a new species, based on the presence of two categories of oxeas, one type of protriaene and of ortho- to plagiotriaene, clearly differentiated from other species of Stelletta known to occur in related biogeographic provinces. Stelletta kallitetilla (de Laubenfels, 1936) (Figs. 2B, 4, 5B) Myriastra kallitetilla de Laubenfels, 1936, p.169. Material: UFBA949-POR, Capitania dos Portos (1258.18 S - 3831.10 W), Salvador, Bahia state, <10 m depth, collector E. Hajdu, 2.XI.1988. Description: Massive irregular, small sponge, with some prominences in the external surface and measuring 21.6 mm by 17.7 mm wide and 6.5 mm thick. Specimen of white color with yellowish portions when alive and in ethanol.

radiating bundles of oxeas, triaenes and occasional styles. Choanosomal oxyasters scattered. Spicules (measurements in m): Megascleres. Oxea 1 Thin, fusiform, straight, or slightly to marked curved, eventually sinuous, with pointed ends 591-725.6-931.2 (SD=113.7) (N=30) / 2.9-5.4-9.7 (SD=2.3) (N=30). Oxea 2 Mostly straight, extremely variable in length with thin, occasionally rounded tips 534-1164.31630.6 (SD=272.7) (N=30) / 6.3-20.1-29.9 (SD=7.7) (N=30). Ortothriaenes Varying in size and shape, occasionally to plagiotriaenes. Cladome thin, rarely with strongyloid ends 256-295.3-324.4 (SD=25.3) (N=30) / 77-81-85.4 (SD=6)

278

12.7-31.8-53.8 (SD=11.2) (N=30) / 5.4-11.3-20.1 (SD=3.8) (N=30). Microscleres. Acanthotylasters Minute, rare scleres with centrum conspicuous and extremely slender rays exhibiting a distal microspined crown 14.2 -11.2-46.6 (SD=2.5) (N=30). Ecology: The sponge was collected in a depth less than 10 m, in a rocky shore at Salvador in Todos os Santos Bay. There are many other sponges and invertebrates in the collection site, like the abundant sponge, Desmapsamma anchorata (Carter, 1882), colonial cnidarians, ascidians and crustaceans. Stelletta kallitetilla inhabits mangrove environments, as was also reported by Wiedenmayer (1977) in the Caribbean region. Geographic Distribution: Central America [Caribbean region]: Dry Tortugas (de Laubenfels 1936); Bahamas (Wiedenmayer 1977); Cuba (Alcolado 1980); Southwestern Atlantic - Brazil, Bahia State, Salvador (present paper). Discussion: Stelletta kallitetilla is for the first time recorded in the Southwestern Atlantic. The specimen here described is in agreement with the original skeleton description and with Wiedenmayer`s (1977), except for the oxeas, which are abundant only in our material. There is a difference too in the aspect of the external surface, because in our specimen it is not so evident the cauliflower-like aspect reported as typical of this species.
Key for the Brazilian species of Stelletta 1A. Without euasters ...................................................... S. anasteria 1B. With euasters .............................................................................2 2A. Euasters in a single category .....................................................3 2B. Euasters in two categories .........................................................7 Fig. 4: Megasclere set of Stelletta kallitetilla. A. Oxea. B. Anatriaene. C. Detail of the anatriaene (cladome). D. Plagiotriaene. Scale bars A, B, D, 100 m; C, 20 m. 3A. With tylasters and anatriaenes ...................................................4 3B. Without tylasters ........................................................................5 4A. Anatriaenes with reduced cladome plus orthotriaenes .... S. beae 4B. Anatriaenes and protrianes ..................................... S. kallitetilla

Compressible to firm consistency; microhispid surface. There are no visible oscules. Pores are concentrated in regions nearly 1 mm2. Skeleton: The ectosome is formed by the cladomes of the triaenes side-by-side. The choanosome is composed by dense, multispicular bundles of triaenes and oxeas; there are rare bundles formed exclusively by anatriaenes. Tylaster microscleres are visible but scarce in the entire skeleton, even more so in the ectosome. Spicules (measurements in m): Megascleres. Oxeas straight, sharply pointed 436.4635.3-876.6 (SD=92.7) (N=30) / 6.2-11.9-18.1 (SD=2.6) (N=30). Anatriaenes Long cladome with thin ends 16.623.1-29.3 (SD=3.4) (N=30) / 3.3-5.4-7.2 (SD=0.9) (N=30). Rhabdome slender, straight and abruptly pointed 381-425.4495.3 (SD=31.1) (N=30) / 5.2-6.6-7.9 (SD=0.9) (N=30). Plagiotriaenes Thick, straight rhabdome, with abruptly pointed ends 362.3-494.3-652.9 (SD=66.2) (N=30) / 7.4-13.523.9 (SD=4.6) (N=30). Cladome with short and straight cladi

5A. Only orthotriaenes and sphaeroxyasters ...................................6 5B. Orthotrianes, plagiotriaenes and acanthoxyasters..................................... S. soteropolitana sp. nov. 6A. One category of oxeas ............................................. S. anancora 6B. Two categories of oxeas .................................................S. gigas 7A. With dichotrianes, acantho- and sphaeroxyasters ..... S. ruetzleri 7B. Without dichotriaenes ................................................................8 8A. Plagiotriaenes, oxyasters and sphaeroxyasters............. S. hajdui 8B. Orthotriaenes and two categories of sphaeroxyasters.................................................... S. crassispicula

Acknowledgements
The authors are grateful to Dr. Elizabeth Neves and Dr. Rodrigo Johnsson for their suggestions, especially regarding the new species. We are also grateful to Dr. Eduardo Hajdu for collection and MEV of microscleres, to Marinette Viottto for collection, to MSc. Mariana

279

Fig. 5: Microscleres of the studied species. A. Acanthoxyaster of new species Stelletta soteropolitana sp. nov. B. Acanthotylaster of Stelletta kallitetilla. Scale bar 5 m.

Carvalho for the chart used to compare species of Stelletta, and to Dr. Marlene Peso de Aguiar for permission of use the image analysis equipment to photograph and measure the spicules. Lastly, the authors thank the Conselho Nacional de Desenvolvimento Cientfico e Tecnolgico (PIBIC/CNPQ ) for a fellowship to BC and to Fundao de Amparo Pesquisa do Estado da Bahia (FAPESB) for a grant in support of the project.

References
Alcolado PM (1980) Lista de nuevos registros de Poriferos para Cuba. Inst Oceanol Acad Cienc Cuba, Ser Oceanol 36: 1-11 Buccich G (1886) Alcune spugne dellAdriatico sconosciute en nuove. Bol Soc Adriat Sci Nat Trieste 9: 222-225 Carter HJ (1882) Some sponges from the West Indies and Acapulco in the Liverpool Free Museum described, with general and classificatory remarks. Ann Mag Nat Hist 9(5): 266-301, 346-368 de Laubenfels MW (1936) A discussion of the sponge fauna of the Dry Tortugas in particular, and the West Indies in general, with material for a revision of the families and orders of the Porifera. Tortugas Lab Pap 467: 1-225 Esteves EL, Muricy G (2005) A new species of Stelletta (Demospongiae, Astrophorida) without microscleres from Abrolhos Archipelago, northeastern Brazil. Zootaxa. 1006: 43-52 Hajdu ECM, Carvalho MS (2003) A new species of Stelletta (Porifera, Demospongiae) from the southwestern Atlantic. Arq Mus Nac 1(61): 3-12 Lerner C, Mothes B (1999) Stelletta hajdui, a new species from the Southwestern Atlantic (Porifera, Choristida, Ancorinidae). Bul Zol Mus Univ Amsterdam 16(12): 85-88 Lvi C (1952) Spongiaires de la cte du Sngal. Bull Inst fr Afrique Noire 14: 36-59 Lvi C (1959) Campagne de la Calypso: Golfe de Guine. 5: Spongiaires. Ann Inst Ocanogr 37: 115-141 Lvi C (1960) Spongiaires des ctes occidentales africaines. Bull Inst fr Afrique Noire 22: 743-769

Mothes B, Lerner CB (1994) Esponjas marinhas do infralitoral de Bombinhas (Santa Catarina, Brasil) com descrio de trs espcies novas (Porifera: Calcarea e Demospongiae). Biocincias 2(1): 4762 Mothes B, Silva CMM (2002) Stelletta ruetzleri sp. nov., a new ancorinid from the Southwestern Atlantic (Porifera: Astrophorida). Sci Mar 66(1): 69-75 Mothes-de-Moraes B (1985) Primeiro registro de Myriastra purpurea (Ridley, 1884) para a costa brasileira (Porifera: Demospongiae). Rev Bras Zool 2(26): 321-326 Nardo GD (1847) Prospetto della fauna marina volgare del Veneto estuario con cenni sulle principali specie commestibili dell Adriatico, sulle venete pesche, sulle valli, ecc. In Venezia e la sua lagune. 113-156 (1-45 in reprint) Ridley SO (1884) Spongiida. In: Report on the zoological collections made in the Indo-Pacific Ocean during the voyage of H.M.S. Alert, 18812. British Museum (Natural History), London. pp. 366-482, 582-630 Schmidt O (1862) Die Spongien des adriatischen Meeres. Wilhelm Engelmann, Leipzig Schmidt O (1870) Grundzge einer Spongien-Fauna des atlantischen Gebietes. Wilhelm Engelmann, Leipzig Sollas WJ (1885) A classification of the sponges. Sci Proc Roy Dublin Soci 5: 112 Sollas WJ (1886) Preliminary account of the Tetraxinellid sponges dredged by H.M.S. Challenger 187276. Part I. The Choristida. Sci Proc Roy Dublin Soc 5: 177-199 Sollas WJ (1888) Report on the Tetractinellida collected by H.M.S. Challenger, during the years 1873-1876. Rep Sci Res Voy H.M.S. Challenger, Zool 25:1-458 Thiele J (1900) Kieselschwmme von Ternate. I. Abhandlungen herausgegeben von der Senckenbergischen Naturforschenden Esellschaft 25, Frankfurt. pp. 19-80 Uliczka E (1929) Die tetraxonen Schwmme Westindiens (auf Grund der Ergebnisse der Reise Kkenthal-Hartmeyer). In: Kkenthal W, Hartmeyer R (eds). Ergebnisse einer zoologischen

280

Forschungsreise nach Westindien. Zoologische Jahrbcher. Abteilung fr Systematik, Geographie und Biologie der Thiere, suppl 16, Frankfurt. pp. 35-62 Wells HW, Wells MJ, Gray IE (1960) Marine sponges of North Carolina. J Elisha Mitchell Sci Soc 76(2): 200-245

Wiedenmayer F (1977) Shallow-water sponges of the Western Bahamas. Experientia Suppl 28: 1- 287 Wilson HV (1902) The sponges collected in Porto Rico in 1899 by the U.S. Fish Commission Steamer Fish Hawk. Bull US Fish Comm 2: 375-411

Porifera research: Biodiversity, innovation and sustainaBility - 2007

281

Mesenchymal cells in ancestral spongiomorph urmetazoa could be the mesodermal precursor before gastrulation origin
Cristiano C. Coutinho(*), Guilherme de Azevedo Maia
Deptartamento de Histologia e Embriologia, Instituto de Cincias Biomdicas, CCS, B-25. Universidade Federal do Rio de Janeiro. Caixa Postal 68021, CEP 21941-970. Brasil. ccoutinho@histo.ufrj.br Abstract: This review proposes a possible scenario for the origin and early evolution of the gastrulation process based on sponge cell organization and developmental biology. We first assume that modern sponges are metazoans derived from a common urmetazoan ancestor which contained primitive multicellular organization and spongiomorph structure. Then, the concept of gastrulation is contrasted among sponge classes and other animals, probably indicating that developmental processes in sponges are mainly based on cellular terms and not in terms of germ layers and gastrulation process. Tissue organization, most likely a cnidarian innovation, could be the base for germ layer body plan organization, but is absent in sponges. A coherent alternative is proposed, in which their organization would be based on mesenchymal and mesenthelial cell differentiation programs. Finally, sponge developmental genes of the homeobox family are reviewed and the NKL group is further discussed, since it is involved in patterning of mesenchymal cells derived from mesoderm in Drosophila, as well as in mesenchymal cell programs in mouse mesoderm. As a general conclusion, the proposed scenario would have primitive spongiomorph mesohyl (mesenchyma) as the precursor for all metazoan germ layers and the Tlx gene family (NKL) as a primary molecular control for mesenchymal lineage, regulating proliferation and differentiation. Gastrulation would thus be originated in the cnidarian lineage, generating the three germ layers, body plan patterning by Hox complex, true epithelium, and three stem cell systems keeping homeostasis, one for each germ layer. Keywords: Gastrulation, homeobox, Porifera, urmetazoa, mesoderm

Introduction
Transition from unicellular to multicellular grade of organization was a major evolutionary step that occurred relatively late in evolution. While fungi and several plant groups have independently reached multicellularity, the question of monophyletic versus polyphyletic evolution of metazoans has raised extensive debate. Up to the middle of the 20th century, organisms without typical tissues, such as sponges, were separated in Parazoa, which were considered as a side offshoot of the major evolutionary tree, comprising all the other Eumetazoa (Hyman 1940). However, recent comparative studies on gene structure, protein primary sequences and ribosomal RNA have gathered a broad set of data, giving a strong support to the monophyly of Metazoa (reviewed in Mller 2001). Molecular studies strongly suggest that Porifera is the oldest metazoan group, and that choanoflagellates are the parent group among Protista, being thus close to the evolutionary root of multicellular animals (Brooke and Holland 2003, King et al. 2003). Multicellular grade of organization, namely the programmed cell proliferation and differentiation, the spatial order and functional integration of differentiated cells, the maintenance of cell population homeostasis, ordered growth,

as well as the response to injury by controlled regeneration and self-recognition, have all appeared during the evolution of spongiomorph urmetazoan, from which modern Porifera and other metazoans diverged. It is thus interesting to address the question of which mechanisms underlie the multicellular organization in sponges, and govern the cell differentiation and their spatial organization. Since this spatial order is established early in embryogenesis, during the separation and positioning of germ layers, sponge type of gastrulation is discussed at morphologic level and in the light of the recent molecular data, shedding perspectives on the evolution of tissues, germ layers and gastrulation. Gastrulation is known to have responded to changes in the environment and egg architecture during evolution, indicating the plasticity of these developmental processes. Leptin (2005) was able to create a complete and up-to-date definition despite the wide diversity of gastrulation types, the period during the early development of animals when major cell and tissue movements remodel an initially unstructured group of cells, requiring coordinated control of different types of cellular activities in different cell populations. A hierarchy of genetic control mechanisms, involving cell signaling and transcriptional regulation, sets up the embryonic axes and specify the territories of the future germ layers. According to

282

this definition gastrulation would be more than integrated cell movements and generation of multilayered organisms. The intimate correlation among axes patterning and gastrulation was also clearly defined in Martindale (2005): The developmental events that generate axial organization during the early cleavage stages lead to the site of gastrulation, and therefore determine the onset of differential gene activity that is responsible for the specification of distinct mesoderm and endodermal germ-layer fates. Gastrulation in deuterostome includes a body-organizing center by which cells move and receive inductive cues in order to differentiate according to axes patterning (Willmer 1994, Holland 2000, Leptin 2005, Martindale 2005). Spiralians, which include the protostomes: molluscs, annelids, polyclad flatworms, sipunculids, echiurans, and nemerteans, display a common pattern of embryogenesis referred to as spiral cleavage, which specify cell fates (Willmer 1994, Martindale 2005). During spiralian gastrulation, the specified blastomeres are placed according to the body axes (Technau and Scholz 2003). Basal metazoans, such as cnidarians and ctenophores also have cell lineages allocated during gastrulation (Martindale 2005). As a conclusion, in both Protostome and Deuterostome the not terminally differentiated blastomeres are allocated according to their designated differentiation program and embryonic axes, thus forming multilayered animals with specific tissue organization, which will be the base for adult body plan (Holland 2000, Technau and Scholz 2003, Leptin 2005). If gastrulation definition is based only on integrated cell movements and generation of axial patterned organisms we should be careful, and keep in mind that it is a very primitive feature also observed in colonial protists, such as Dictyostelium (amoeba). Gastrulation definition should not be based on primitive features also observed in non Metazoa clade. This unicellular amoeba does not gastrulate, but the colony has integrated cell movements that generate a patterned axis without cell layers. They experiment social life in stressed conditions, joining together to form moving streams of cells that converge at a central point where they culminate in a fruiting body that releases spores. The chemotaxic cAMP is secreted mostly in the aggregation central and works as a morphogen that attracts and patterns the axis of the fruit body (Ginsburg and Kimmel 1997). As discussed later, the sponge larvae have antero/posterior axis and probably a chemioattractant patterning it (Degnan et al. 2005, Leys and Ereskovsky 2006), but these primitive features should not be sufficient to assume gastrulation in Porifera. Nevertheless, a distinguishing feature is the cell layer organization in sponge larvae. It is acquired just after cell movements in the blastula and this cell layer organization is not observed in Dictyostelium. In conclusion, comparative analysis of deuterostome, protostome, sponges and amoeba development indicates that integrated cell movements in sponge larvae formation is slightly more complex than Dictyostelium multicellular organization.

consensual, of Porifera into four classes: Demospongiae, Homoscleromorpha, Hexactinellida and Calcispongia (Borchiellini et al. 2004, Ereskovsky 2004, Boury-Esnault 2006). As expected for an old group with a long evolutionary history, different sponge taxa have a great variety of embryological processes, and this diversity generated old controversies in the literature (Ereskovsky 2004). As reviewed by Leys (2004), one group classifies sponges as Parazoa, because they postulate that sponges do not undergo gastrulation, arising from a separate unicellular ancestor (Rasmont 1979, Ereskovsky and Korotkova 1997). Another group believes that gastrulation occurs at metamorphosis, when intensive cell movements completely rearrange the cell layers and give rise to the definitive sponge tissues (Brien 1937, Lvi 1963, Tuzet 1963, Brien 1967, Fell 1974, Simpson 1984). A third group recognizes gastrulation cell movements of ingressions, epiboly, delamination and invagination just after blastula stage (Lvi 1956, Efremova 1997, BouryEsnault et al. 1999). Here sponges are regarded as Metazoa, and personal views on the available data are proposed. Sponge embryogenesis and adult organization should be understood in cellular terms and not in terms of gastrulation and germ layers, stressing a complementary differentiation of flagellated and amoeboid cells, and their mutual interaction in three-dimensional structures.

Demospongiae
The vast majority of sponges belongs to the Demospongiae. As reviewed by Ereskovsky (2004) and Leys and Ereskovsky (2006), there are three main types of development in this class. The most common generates the parenchymella larvae, another type generates disphaerula larvae, represented by Halisarcida, and the direct development is represented by Tetilla. As reviewed by Borojevic (1970), total chaotic cleavage is followed by formation of a morula, which generates parenchymella larvae either by delamination or by micromeres appearing and centrifugal migrating to the periphery. In both cases micromeres and macromeres are segregated in external and internal layers of a solid swimming larva. Subsequently, micromeres differentiate into the flagellated larval cell layer that covers either the whole larva or only the anterior part. During metamorphosis, the flagellated epithelium disintegrates and micromeres migrate inward to become choanocytes or, sometimes, are partially or completely phagocyted during metamorphosis. In this case the adult is generated directly from a group of non-flagellate amoeboid cells. Nevertheless, the use of a tracer indicated that ciliated cells of the haplosclerid larva may form the future flagellated cells of the adult, the choanocytes (Leys and Degnan 2002). Maldonado (2004) recognized, in this case, gastrulation movements as mixed delamination. Such cell movement without a clear establishment of cell lineage according to the body plan is here assumed not sufficient to be defined as gastrulation. For instance, flagellated choanocyte can be generated either by amoeboid archeocytes or flagellated micromeres. Moreover, all adult cells and structures can be generated from non-flagellate amoeboid cells without

Sponge embryology
Morphological and molecular data place the extant sponges in phylogenetically related groups. Here we follow the innovative division, and not completely

283

the maintenance of an organization previously established during larva formation (Borojevic 1966a). Cells dissociated from a parenchymella larva re-aggregate directly to form juvenile sponge rather than a larva like organism (Borojevic and Lvi 1965). This property of generating all sponge cell types is unique to amoeboid mesenchymal stem cells only, whether derived from the larva or from the adult organism, since differentiated flagellated cells from the external layer of the parenchymella and from adult choanoderm cannot do so (Huxley 1911, Borojevic 1963, 1966a). Moreover, other studies with cell tracer would reveal the contribution of ciliated/flagellated cell lineage for the adult generation, as previously observed by Leys and Degnan (2002). Extensive studies done by Degnan et al. (2005) in Haliclona development (as Reniera; Demosponge, parenchymella), suggested that gastrulation occurs after late blastula, followed by antero-posterior axis patterning. Sponge homologues of metazoan developmental transcription factors seems to be expressed in regions, layers and cell types of polarized Haliclona larva, as expected after gastrulation (Larroux et al. 2006). These authors observed fertilization followed by a period of cell division yielding distinct cell populations which, through a gastrulation-like process, become allocated into different cell layers and patterned within these layers. There are some divergences in Haliclona gastrulation and we suggest keeping the status of gastrulation-like. For instance, already determined cells such as sclerocytes (and probably pigmented cells too) sort out to external embryo layer, become fully differentiated during this process and come back to internal layer to be positioned posteriorly, a process of fully differentiated cell migration among germ layers not common in metazoan gastrulation. Moreover, all adult cells and structures can be generated from non-flagellate amoeboid cells without the maintenance of an organization previously established during larva formation (Borojevic 1966a), and sponge type of gastrulation seems to be unnecessary to specify the body plan of adult sponges. As already suggested, the body structure of metazoans is more similar to larvae than adult Haliclona and it can be used as evidence for the hypothesis of larval neotenic evolution during the transition from the spongiomorph urmetazoan to multilayered animals with patterned axis (Larroux et al. 2006, Leys and Ereskovsky 2006). In this case, the sponge gastrulation-like process could be proposed as homologous to other metazoan gastrulation. Larva must adhere on the substrate and undergo metamorphosis to form porocytes and channels, thus completing the development of the functional aquiferous system (Borojevic 1971). This extrinsic induction to undergo metamorphosis and generate sponge body plan is quite different from the intrinsic signals of the gastrulation. This argument is valid only for the group that believes that gastrulation occurs during metamorphosis. The simplest development, albeit not the most primitive one is that of Tetillidae (Spirophorida), in which there is no free larval stage. The substrate-adherent egg generates, by equal divisions, an amoeboid cell mass, which differentiates directly into an adult sponge (Watanabe 1978). This is an indication of the cellular organization of sponges, contrary to gastrulation positioning cell types in an ordered body plan.

This is repeated in asexual reproduction from gemmules or involution bodies, as well as from re-aggregated cells in sponge regeneration from dissociated cells (Wilson 1907). In Polymastia (Hadromerida), gastrulation is also missing. Eggs are retained at the substrate by mucus produced by the mother-sponge during egg-laying (Borojevic 1966b). After equal divisions, they give rise to a single-layered flat blastula containing only one cell type with short flagella, and only a virtual central lumen (Maldonado 2004). These creeping benthic larvae have an antero-posterior axis, since they are larger at the front (clavoblastula) and no dorso-ventral difference could be observed. After a long free life, they settle and convert directly to clumps of amoeboid cells, giving rise to the adult sponge in a process similar to the hatching of the gemmule (Borojevic 1966b). The most striking example of Demospongiae cell movement similar to gastrulation is the larva formation in some species of the highly derived group Halisarcida (disphaerula larvae type), pointed by Maldonado (2004). As reviewed by Ereskovsky, (2004) and Leys and Ereskovsky (2006), Halisarcida does not follow parenchymella larva formation but the disphaerula type. Before larval differentiation, few cells migrate from the epithelium of the coeloblastula into the blastocoel and remain peripheral and close to the overlying epithelium without proliferating. The posterior-lateral ciliated blastoderm then invaginates and reorganizes into a monolayered internally ciliated tube suspended in a fluid-filled blastocoel. The lumen of the internal tube and the blastocoel are transitory cavities, since they are filled by proliferation of the internal cells as well as by late cell migration. Because the adult form does not preserve the primitive archenteron structure, and because other external cells migrate inward after gastrulation and generate internal structures, it is problematic to assume gastrulation strictu sensu during Halisarcida larval differentiation (Lvi 1956, Harrison and de Vos 1991, Ereskovsky and Gonobobleva 2000). We could not recognize here a process that should define the layers from which all adult structures are generated.

Hexactinellida
In the early development of Hexactinellida, the least studied class of sponges, a hollow blastula is formed, the coeloblastula (reviewed in Boury-Esnault et al. 1999, Leys 2003, Ereskovsky 2004, Leys and Ereskovsky 2006). Cleavage is total, equal and pseudospiral, and generates a hollow blastula. Subsequent delamination renders a twolayered stereoblastula. The external layer gives rise to ciliated micromeres, sclerocytes, choanocytes (which later become collar buds) and spherulous cells. Internal macromeres envelop everything and gives rise to the internal, mostly syncytial, reticular tissue. Antero-posterior polarity of the larva (called trichimella) is evident already during early embryogenesis. The swimming trichimella is devoid of larval skeleton and has no functional filter feeding choanocyte chambers. Such generalized, and not regionalized, delamination around Hexactinellida coeloblastula is not common in bilaterians. Since metamorphosis has not yet been described in Hexactinellida, it is difficult to affirm

284

that delamination in hexactinellids is devoid of a clear establishment of functional cell lineages, as happens in protostomes for instance. Moreover, it is not strange to assume that, as in demosponges, such morphogenesis based the origin of gastrulation.

the transient cavity that is formed by invagination is not the future gut. Calcinea As reviewed by Ereskovsky (2004) and Leys and Ereskovsky (2006), Calcinea has a typical simple coeloblastula, sometimes called the calciblastula. Such larvae are found in the simplest asconoid Calcinea, which explain why it was considered to be the simplest sponges. At the end of total and equal cleavages, blastomeres become flagellated, but a few large ones may remain without flagella and are located at the posterior pole. Antero-posterior polarity becomes expressed towards the end of larva formation. These non-flagellated cells have been interpreted as founders of the endoderm after their polar ingression into the blastocoel, although there is no clear morphological evidence. Such process has generated many speculations on the evolutionary significance of these cells. Moreover, not all larvae of Calcinea have these non-flagellate cells; some species have a coeloblastula with only flagellated cells. The cell ingression site is not a conditio sine qua non to generate Calcinea bauplan, contrasting with gastrulation in bilaterians. The equatorial section of the larva gives anterior and posterior halves that are equally able to generate a new sponge, regardless of the presence of the posterior nonflagellate blastomeres in only one of them. Also, while the inner cell mass is indeed preferentially formed at the posterior pole of the larva by polar ingression, all the flagellate cells eventually lose their flagella during larval maturation. At the end of the free life, the whole larva thus collapses in a solid cell mass that settles to the bottom, and is converted into a small mass of amoeboid cells that generate a new sponge (Borojevic 1968). Further sponge development, settlement and metamorphosis are dependent upon conversion of larval flagellate cells into amoeboid cells that can proliferate and generate the necessary cell mass. The larval flagellate layer is thus a temporary embryonic structure that has to disintegrate before the progression to larval metamorphosis. We find again here the general pattern of cell differentiation and morphogenesis already discussed for Demospongiae, where a clump of substrate-adherent amoeboid cells is necessary and sufficient to generate a new sponge. Hence, the progressive filling of the blastocoel does not correspond to the generation of a second embryological cell layer, but a conversion from flagellate cell types responsible for movement into amoeboid cells that can proliferate. Calcaronea This group has a divergent and most complex development compared to other sponges and its earlier stages were recently reviewed (Eerkes-Medrano and Leys 2006). It is somewhat similar to the development of Volvox. In these viviparous sponges a single cell layer hollow blastula, the stomoblastula, is formed after total and unequal cleavage. It already has the two major cell lineages arranged in opposite poles, amoeboid and flagellated, corresponding to macromeres and micromeres respectively (Amano and Hori 1993, Ereskovsky 2004, Leys and Ereskovsky 2006). Micromeres can have inwardly directed flagella forming a sheet that turns inside

Calcispongia
The embryogenesis and larval morphology of the two subgroups of Calcispongia, Calcinea and Calcaronea are so different that, considering differences in anatomy, cytology and spicule morphology, it has been questioned whether they belong to the same evolutionary lineage (Borojevic 1970). These are also quite distinct from sponges with siliceous or organic skeletons. Molecular analyses have settled this issue, and it is now considered that they do represent a well defined and monophyletic group (Manuel 2006). Molecular studies indicated that they may be closer to other Metazoa (Ctenophora, Cnidaria) than to other sponges (CavalierSmith et al. 1996, Collins 1998, Kruse et al. 1998, Zrzavy et al. 1998, Schutze et al. 1999, Borchiellini et al. 2001, Medina et al. 2001), although this proposal requires further analyses (Manuel et al. 2003, Manuel 2006). Following an extensive study of calcareous sponges, Haeckel (1872) proposed that the simple tubular ascons are the most primitive sponges. The main argument was the structure of tubular calcareous sponges (ascons), which is constituted essentially by two layers (diploblast): the internal flagellated choanoderm and the external pinacoderm, corresponding to endoderm and ectoderm respectively. This larva was described as a blastula, and the following metamorphosis was assumed to be a gastrulation process. At this time an invagination of the non-flagellate cells in one pole forms a gastraea, a two-layered ovoid larva provided with a single mouth, whose illustration was provided in the monograph Die Kalkshwmme (Haeckel 1872). After settlement, this larva would give rise directly to a twolayered ascon with a single apical opening, and this theory of the origin of multicellularity was named the Gastraea theory (Haeckel 1874). This formation of the second germ layer by invagination of the hollow blastula is still frequently considered to be the primitive form of gastrulation (Wolpert 1992). Only Hammer (1908) and recently Leys and EerkesMedrano (2005) were able to capture a stage that represented Haeckels gastrula, a larva invaginating to form a ciliated gut, but not Metschnikoff (1879) and Borojevic (1968). Leys and Eerkes-Medrano (2005) observed such epithelial invagination occurring prior to and during attachment. The authors suggested that this event is difficult to capture because metamorphosis in calcaronean sponges takes place very rapidly. An alternative interpretation for such difficulty to observe is a rare and aberrant development related with substrate adhesion failure (R. Borojevic, personal communication). The idea of gastrulation during metamorphosis in calcaronean sponges is not well accepted by modern spongiologists and the main argument is that the hole is engulfed by the newly forming basal epithelium, and the ciliated cells dedifferentiate to form an inner cell mass. Thus

285

out (excurvation) through the opening between macromeres, the mouth of the stomoblastula, and generates an amphiblastula with outwardly directed flagellated cells at the anterior pole. During the long free life of the larva, posterior pole macromeres continue dividing slowly until larval settlement and metamorphosis. At this stage non-flagellated cells cover the flagellated external layer by a process similar to epiboly, thus generating the pinacoderm. Meanwhile, flagellated cells regress their flagellum, submerge inside and generate the other sponge cell types. It could be regarded as the establishment of cell lineages but these lineages are not respected during adult life. The adult mesenchymal stem cell, the archeocyte, generates all cell types, including pinacocytes. The conversion from flagellated cells responsible for larval movement to the proliferating totipotent amoeboid cells is similar to what is observed in choanoflagellates, as discussed latter. The functional organization of sponge larvae by cell sheet inversion and flagellar-mesenchymal conversion during metamorphosis are both cell movements commonly found in colonial protozoans or algae and probably they are not enough to define gastrulation.

germ layer, we believe that true epithelium is important for ectoderm and endoderm origin, as discussed further on. If is assumed that metamorphosis is part of the gastrulation, thus Homoscleromorpha has gastrulation aspects more similar to other metazoans, as preservation of the epithelium. Ancestral homoscleromorpha would have aspects in its metamorphosis, such as invaginating epithelium, that could be evolutionary committed to originate gastrulation later on.

Germ layers in sponges: Enantiozoa

The assumption of germ layers in sponges is problematic, since the contribution of endoderm and ectoderm to adult sponge tissues is quite different from that in other basal metazoans, and it is commonly adopted the concept of gastrodermis and epidermis (Hyman 1940) to avoid embryological implications, i.e., the gastrodermis need not necessarily be endoderm throughout. These issues have direct implications in defining the nature of sponge gastrulation as discussed here. The anterior flagellated pole of swimming sponge larvae (amphiblastula, cinctoblastula and parenchymella), which putatively corresponds to the animal pole of other larvae, gives origin to the internal choanoderm, while amoeboid cells at the posterior (vegetative) pole generate the outer layer of the sponge body (Borojevic 1970). Delage (1892) has proposed classifying sponges as Enantiozoa, the inverted animals, with an internal ectoderm (choanoderm) and external endoderm (pinacoderm). The intense phagocytic activity of the internal and external pinacoderm layer was associated with food ingestion, a feature of the endoderm in other metazoans (Willenz and van de Vyver 1982), and this would be an additional argument in support of the Enantiozoa concept. An alternative approach to this issue was proposed by Willmer (1970) and Lvi (1970) who considered that sponge embryogenesis and adult organization should be understood in cellular terms and not in terms of germ layers, stressing the complementary differentiation of flagellated and amoeboid cells, and their mutual interaction in threedimensional structures. This is similar to the proposal of Morris (1993), who considered that the common feature of metazoan development and probably sponge development is the mechanism controlling the multicellular grade of organization, involving complex interactions among motile mesenchymal cells, epithelial cells and extracellular matrix.

Homoscleromorpha: primitive gastrulation with primitive epithelium

Homoscleromorpha have mostly free swimming, hollow flagellated blastulae, originated from combined process of total, equal and chaotic cleavage, histolysis and outward migration of micromeres (multipolar egression) (BouryEsnault et al. 2003, Ereskovsky 2004, Leys and Ereskovsky 2006). This unusual outward migration of micromeres is only comparable with the centrifugal cell movements generating parenchymella larvae in true Demosponges. There is no such cell movement during gastrulation of other metazoans. This larva is called the cinctoblastula, and contains one-layered flagellated epithelium, basal lamina and belt desmosomes (Boury-Esnault et al. 2003). After settlement, cells of the anterior pole give rise to the large choanocytes typical of this group, and the posterior ones to pinacocytes, which are also flagellated, with the exception of those belonging to the basopinacoderm. During larval metamorphosis, the structure of the flagellated layer of the anterior pole is preserved, folding and invaginating by quite complex cellsheet movements. After fragmentation it gives rise directly to several choanocyte chambers of the rhagon, a small leuconoid sponge with a simple aquiferous system containing a group of choanocyte chambers arranged around the central exhalant aquiferous lacuna and the osculum (N. Boury-Esnault, personal communication). Homoscleromorpha have a basal lamina under both larval and adult cell layers, with its typical components such as the collagen IV (Boute et al. 1996, Boury-Esnault et al. 2003). We believe that it may stabilize flat cellular sheets, equivalent to epithelia in other metazoans. We discuss further the similarities and differences on epithelia from Homoscleromorpha, others sponges, Dictiostelium and clades with true epithelium. Now it is just assumed that the existence of basal lamina and desmosome contribute significantly to the cell-sheet movements, permanence and stability of cell lineages and the larval and adult epithelial layers. Since gastrulation places cells in their definitive

Germ layers in sponges: mesoderm, ectoderm or endoderm derived epithelium?

As in any textbook, mesenchyme is here defined as being formed by elongated cells with abundant cytoplasmic extensions, immersed in an abundant, viscous extracellular matrix with few fibers. Mesenchymal cell morphology is commonly present in mesoderm (embryonic or extraembryonic) and in the evolutionarily recent neural crest. Sponges have a similar connective tissue layer between covering cell layers, which was named mesohyl (Borojevic et al. 1967). The authors concluded that only morphological data was not sufficient to associate the sponge mesenchyma with mesoderm.

286

Even a simple cover layer of cells can be considered an epithelium (Allaby 1985). Nevertheless, there is a clear distinction between simple cell layer and true epithelial tissue composed by polyhedral and juxtaposed cells with only a small amount of extracellular substance, firmly attached by intercellular junctions, and both morphologically and functionally polarized by attachment to the basal lamina. These features are the basis for true functional epithelial covering layers on all external and internal surfaces of metazoans (excluding articular cartilage). The three germ layers may form epithelium. The mesodermderived epithelium is called mesothelium when covering the peritoneum, pleura and pericardium, and endothelium when covering blood and lymphatic vessels. Junctional complexes are less well developed in epithelia derived from mesoderm (except the renal duct with complete junctional structures) than in epithelia derived from ectoderm and endoderm. For instance, tight junctions, adherens junctions and gap junctions are less ordered in endothelia than in the epithelia derived from ectoderm and endoderm (Imhof and Aurrand-Lions 2004). It thus causes endothelial cells more dynamics and mobile. Despite the fact that sponges (at least the most simple asconoid ones) are considered classical diploblasts, i.e. composed of ectoderm and endoderm only, the working hypothesis here is that no sponge adult tissue has a typical epithelial structure, with the possible exception of the Homoscleromorpha. Basal lamina has not been observed in sponges (excluding Homoscleromorpha), in spite of the presence of a dermal membrane in a number of adult sponges and larvae (reviewed in Maldonado 2004). Sponges possess a fibrillar component secreted by pinacocytes, consisting of collagen fibrils scattered between pinacocytes, which resembles the lamina reticularis zone of the vertebrate epithelial basal lamina (reviewed in Harrison and de Vos 1991). The classical desmosome and macula adherens have not been described in sponges (excluding Homoscleromorpha), but localized electron-dense deposits exhibiting tonofilament insertions have been reported in two distinct Demospongiae, suggesting the presence of specializations in pinacocyte membranes comparable to the macula adherens and desmosome (reviewed in Harrison and de Vos 1991). Pinacocytes associate through interdigitating membranes (Adell et al. 2004), but apposed membrane junctions are often bordered by gaps of 100-300 nm between cells. This is not so different from what is observed in colonial amoeba Dictyostelium, which does not form epithelium, but has adherens junctions connecting cells through their actin cytoskeleton (Grimson et al. 2000). These details are morphological evidences supporting that the simple epithelial structures of sponges are not sufficient to characterize them as true epithelium, but were committed to originate it later on along evolution. Nevertheless, some authors recognize true tissues in non Homoscleromorpha Demosponges (reviewed in Harrison and de Vos 1991). Pinacocytes of some Demospongiae and Calcispongia have a large part of the cell body embedded in the mesohyl, and readily migrate into it where they convert to typical collencytes with fibroblastoid morphology and collagen secretor activity (Faur-Frmiet 1932, Harrison 1972). It was suggested that the ability to become motile is related to

the general absence of specialized desmosomal junctions in sponge epithelia (reviewed in Harrison and de Vos 1991). The absence of solid adhesion and the ability to become mobile would attest against gastrulation in sponges because there is no conservation of cells into a specific germ layer, as introduced previously. Loose attachment and the mesenchymal nature of pinacocytes are not characteristics of epithelia derived from ectoderm and endoderm, and it is here hypothesized having mesodermal like nature, as previously suggested by Bagby (1970). Assuming early spongiomorphs at the base of metazoan origins, early pinacoderm would be ancestral mesothelium-like/endothelium-like tissue. Afterwards, when Porifera phylum branched off from the main metazoans lineage, the common ancestry of Cnidarians and Bilateria is here hypothesized to originate ectoderm and endoderm derived epithelia. This hypothesis is different from the classical view of ectoderm and endoderm being more ancient than mesoderm. Moreover, increasing number of authors argument in favor of mesoderm features among cnidarians, and this is in agreement with the hypothesis of ancient origin of mesoderm fulfilling internal structures and making volume for 3D structure, as discussed later on (Spring et al. 2000, 2002, Scholz and Technau 2003, Hayward et al. 2004, Martindale et al. 2004, Galle et al. 2005, Seipel and Schmid 2005). The choanoderm is a flagellated flat cell layer, facing the choanocoel of the aquiferous system. Such simple filtering devices with cells containing an apical flagellum surrounded by a collar of microvilli are observed in protonephridia (flame cells), and were suggested to have existed in the ancestor of Bilateria (Dewel 2000). Protonephridia has a terminal flame cell, or crytocyte, and a proximal tubule. It occurs in most of the lower Metazoa, and has long been regarded as the primitive excretory and osmoregulatory apparatus in animals. Despite the traditional view deriving protonephridia from ectoderm (Willmer 1994), Dewel (2000) further suggested that all Bilaterian nephridial systems have been derived ultimately from the mesothelium of coelomic cavities, which is derived from mesoderm. The hypothesis of mesoderm derived protonephridia was also suggested by Younossi-Hartenstein and Hartenstein (2000), who reported the development of these cells in polyclad flatworms. There is a morphological analogy between choanocytes and flame cells, and they also osmoregulate the multi-cellular structure. Only after molecular studies address these analogies it will be possible to distinguish if there is some degree of homology among them. Actually, we conclude that it is possible to think that sponges may have a mesenchymal/mesentheliallike nature, a possible precursor for mesoderm evolutionary origin. The structural similarity between choanoflagellates and sponge choanocyte does not necessarily mean that choanoflagellate-like organisms were ancestors of the extant sponges and the remaining metazoans. An alternative hypothesis regards choanoflagellates as derived from evolutionary simplification of sponges. Choanocyte would be an end branch of the animal evolution with no homolog flagellated collar cell identified in metazoans (Maldonado 2004). Thus, choanocytes-lacking larvae would be

287

architecturally closer to the remaining metazoans than adult sponges. Reviewing histological organization of adult sponge and larva, Maldonado (2004) identified in larva, but not in adult sponges, several evidences of more complex junction structures and external cell layer with lamina reticularis zone. These data were considered as being consistent with the hypothesis of neotenic evolution of metazoans from primitive spongiomorph larva. Nevertheless, evolutionary simplification is much more common in parasitic organisms than in free life style ones and Maldonado hypothesis needs more evidences to be credited.

Germ layers and stem cell systems

Differentiation of all cell lineages from a single progenitor with unlimited proliferation capacity and plasticity could be a hallmark of cell differentiation programs in primitive metazoans (Harrison and de Vos 1991). Sponges have a single totipotent stem-cell system (Borojevic 1966a, Borojevic 1970, Funayama et al. 2005) and it is based on archaeocytes. Its capacity for long-term proliferation is probably explained by keeping high telomerase activity, which prevents cell senescence (Koziol et al. 1998). Archeocytes have mesenchymal morphology and are also responsible for phagocytosis, including self-recognition in a primitive immune system (Willmer 1970). In contrast to sponges, Hydra has three stem-cell systems, two of which produce cells for maintenance and growth of ectodermal and endodermal derived tissues. The third one is responsible for the mesenchymal interstitial cell lineages, which forms a thin mesoglea between epithelia (Bode 1996). In contrast to the autonomy of sponge mesenchymal amoeboid cells, Hydra only succeeds to regenerate from a small clump of cells if there is simultaneous presence of at least some tissues of the two epithelia in the regenerative blastema (Bekkum 2004). A different result was observed in marine hydroid Hydractinia (Mller et al. 2004). Despite the presence of three stem-cell types in normal condition (for ectodermal, endodermal and interstitial lineages respectively), interstitial stem cells can give rise to all cell types during stressful conditions. Moreover, the epithelial microenvironment had to be preserved in order to generate epithelial cell types from interstitial stem cell system. Despite the lack of evidence for a third interstitial mesenchymal stem-cell system in Cnidaria other than hydroids, it is possible to recognize a broad gap spearing the stem-cell systems of Porifera and hydroids, potentially representing one of the major steps in the evolution of multicellularity. The emergence of a true epithelium in the common ancestor of Cnidaria/Bilateria, with a basal lamina and its own stem cell system, had pivotal importance. Epithelial cells do not usually cross the basal lamina to colonize tissues derived from other germ layers, and thus a stem-cell system is required for maintenance of each germ layer-derived tissue. It is possible that gastrulation coevolved with the basal lamina and stemcell systems, since gastrulation would place cell layers and their microenvironment in a defined axis, as in Cnidaria and Bilateria. The totipotency of Hydractinia interstitial stem cells is observed only under stressful condition and is

dependent of the epithelial microenvironment, which was already positioned by gastrulation. It could be viewed as an intermediary condition from one totipotent mesenchymal stem cell system in Porifera, to isolated germ layers with their own stem cell system in most Bilateria. This is not a rule, but a tendency, and exceptions are expected. For instance, ectoderm may switch for mesoderm lineage during axolotl tail regeneration (Echeverri and Tanaka 2002). Planarian have only one totipotent stem cell system, the neoblasts, despite the presence of three germ layers (Reddien and Alvarado 2004). This divergent stem cell plasticity would be responsible for the uncommon capacity for body regeneration. Even presenting spectacular capacity for regeneration, a preexisting ordered body is needed for perfect regeneration, and this order was built during planarian or axolotl gastrulation. This is different from sponges, which regenerate from masses of mesenchymal cells without any influence of preexisting gastrulation order and no true epithelium separating germ layers derivatives.

Gastrulation: amoeboid (mesenchymal) and flagellated cells?

Since sponges cells have sufficient autonomy to generate a functional body from a clump of amoeboid (mesenchymal) cells adherent to the substrate, the major requirement for their multicellular integration is the control of cell proliferation versus cell differentiation. This has already been proposed to be the major requirement for formation of multicellular bodies from flagellated Protozoa (Buss 1983). Mitosis requires the use of the microtubule organizing center and interruption of flagellar activity, and it hampers the movement of the swimming flagellated colony. Cell divisions have to be restricted to amoeboid cells. They have to follow a temporal program in which the functional flagellated state is interspersed with dividing cells in the amoeboid state. For swimming organisms it may be advantageous to segregate amoeboid cells inside, as in swimming sponge larvae or colonial choanoflagellates (King 2004). In contrast to Buss (1983), we do not consider that this segregation should be considered as true gastrulation, since it may be caused simply by a proliferating state, as in colonial choanoflagellates. The essence of being a sponge is thus the coordinated and cooperative opposition of the amoeboid and flagellated state of their cells, as in a choanoflagellate colony (Lvi 1970). In this context, the genetic control of the self-renewal of dividing amoeboid cells versus differentiation of flagellated cells is the first requirement towards integration of the multicellular body, independently of gastrulation.

How old is mesoderm?

The formation of mesoderm in jellyfish has already been suggested (Spring et al. 2002, Mller et al. 2003, Seipel and Schmid 2005) and is in accordance with the hypothesis of a spongiomorph mesenchymal structure originating mesoderm in the common ancestor of Urbilateria and Cnidaria. Actually, during medusa bud development, cells detach from the epithelial outer layer and ingress to form the entocodon, a proliferating cell mass separated from the ectoderm and

288

endoderm by an extracellular matrix (Boelsterli 1977). Cells from the entocodon proliferate and migrate to give rise to new tissues, such as the striated (including nonmyoepithelial cell layers) and smooth muscle tissues of the developing medusoid (Spring et al. 2002, Mller et al. 2003, Seipel and Schmid 2005). The authors suggested the existence of a motile tri-layered cnidarians ancestor and a monophyletic descent of striated muscle in Cnidaria and Bilateria. The expression of mesoderm and myogenic cell line-specifying genes was assessed to substantiate that the entocodon in cnidarians is homologous to the mesoderm of bilaterians. These data come from JellyD1, related to an ancestral MyoD gene (Mller et al. 2003) brachyury, Mef2, Snail (Spring et al. 2002) and Twist (Spring et al. 2000). Based on mesoderm markers, gene expression, and morphological features, the authors suggested that gastrulation would be initiated at the blastulalarva transition, interrupted at the larva and polyp stages, and continued during entocodon formation. Following the authors, diploblasty evolved secondarily in Cnidarian larvae and polyps (Seipel and Schmid 2005). The attached polyps do not feed and would be an evolutionary simplification of a more complex form. The medusoid form of life is faced with locomotory and other complex behaviors, and these would be the evolutionary pressure to retain genetic programs of other cell types. These data are in agreement with the early origin of mesoderm features in spongiomorph urmetazoans (maybe not a true mesoderm), before the origin of true gastrulation processes. In the next section this hypothesis will be confronted with new genetic evidences.

Protein localization of Sd-Bra (sponge brachyury homologous) showed a granular pattern in the cytoplasm of some cells dispersed in the sponge tissue. Only in a few cells was the signal seen both in the cytoplasm and the nucleus (Adell and Mller 2005). Sd-Bra is also located in the cytoplasm of cultured sponge cells. Low amounts of Sd-Bra were found in almost all the cells that adhered to the plastic after 24 hours of culture; later the expression is restricted to a few cells of the already formed primmorphs, the sponge cells aggregate. Brachyury expression in sponges is not restricted to embryo or larval development, but includes adult mesenchymal cells undergoing cell movements and specification. It is possible that the original feature of brachyury in urmetazoan spongiomorphs was to regulate genes involved in morphogenetic movements and differentiation of mesenchymal cells, and was afterwards committed to gastrulation cell movements in other metazoans. Again, gastrulation in sponges seems to be very primitive but further data on sponge brachyury expression and function would shed light on this question.

Hox complex: genetic programs for antero-posterior patterning, downstream gastrulation

The processes of gastrulation are very flexible and appear to have changed rapidly during evolution in response to environmental changes and architecture of the egg. Nevertheless, it is possible to recognize homology among protostome and deuterostome gastrulation (Holland 2000). Gastrulation is based on common genetic interactions and Antp-Hox genes are part of these conserved gene interactions (Martindale 2005). Their expression is set up during gastrulation and they function as transcriptional regulators arranged in clusters along the chromosome (Hom complex in invertebrates and Hox complex in vertebrates), whose genomic organization reflects their central roles in patterning along the anterior/posterior (A/P) axis (Duboule and Dolle 1989, Graham et al. 1989, Mcginnis and Krumlauf 1992, Duboule 1994, Ferrier and Minguillon 2003). If Cnidaria has true gastrulation with tissue movements remodeling an initially unstructured group of cells, with a hierarchy of genetic control mechanisms that sets up the embryonic axes and specifies the territories of the future germ layers, one should expect a Cnidaria Hox system associated with gastrulation. In a recent publication Kamm et al. (2006) concluded that axial patterning and diversification in the Cnidaria predate the Hox system, since there is no equivalent Hox cluster in Cnidaria. According to the authors, the cnidarian Hox genes are expressed in patterns that are inconsistent with the Hox paradigm. Nevertheless, another recent publication has divergent conclusions. Chourrout et al. (2006) characterized the full Hox/ParaHox gene complements and genomic organization in two cnidarian species, and suggested an ancestral ProtoHox cluster consisted of only two anterior genes. Non-anterior genes could have appeared independently in the Hox cluster after the separation of bilaterians and cnidarians. These conclusions were partially supported by the gene linkages of five genes (HoxC/HoxDa/HoxDb/Evx/HoxA) in a tandem

Genetic evidences
Among all cloned sponge genes, the most significant to the discussion on sponge gastrulation and germ layers is the brachyury family of T-domain containing transcription factor (Manuel et al. 2004, Adell and Mller 2005). A consensual expression for brachyury in the blastopore and subsets of mesodermal cells in most metazoans has now emerged (Technau 2001). In embryos of the sea anemone Nematostella, a basal Cnidaria, brachyury is expressed around the blastopore and its derivatives, the endodermal mesenteries (Scholz and Technau 2003). In Hydra and jellyfish polyps, expression is found in the endoderm of the mouth anlage (Technau and Bode 1999, Spring et al. 2002). The blastopore has a key function in the gastrulation movements of all metazoans, but the cell types it gives rise to, in a brachyury-dependent manner, vary (Marcellini et al. 2003). It specifies the cells in which it is expressed. Marcellini et al. (2003) observed after heterologous experimentations that brachyury ortologous from different phyla, including Cnidaria, were able to induce the specification of mesoderm and/or endoderm lineages in competent Xenopus tissue, and distinguished both ancestral and derived functions of brachyury proteins. Therefore, their observation could be directly linked to an ancestral function of brachyury in mesoderm specification and blastopore formation. A derived function is proposed for insects and tunicates, with no circumblastoporal expression, lost of the N-terminal peptide and inductive activity, in heterologous assay, for both endoderm and mesoderm.

289

array over about 50 kilobases (kb) and two other homeobox genes (Mnx and Rough) about 200 kb downstream. Mnx is also present in the neighborhood of chordate Hox and the Drosophila Rough is on the same chromosome arm, 3R, as the ANT-C and BX-C Hom complexes. These new data on gene structure, genomic organization and Cnox2-Pc (anterior Hox gene) expression during the establishment of an anteriorposterior axis (Masuda-Nakagawa et al. 2000) support true gastrulation in Cnidaria. Identification of Hom gene clusters in sponges would be surprising. It would be a strong evidence supporting homology among bilaterians and sponges axial patterning and probably its regulation during gastrulation. After Seimiya et al. (1994), Coutinho et al. (1994) and Kruse et al. (1994), several other authors have tried to identify and clone sponge Antp-Hox genes. The chosen methodology was the use of heterologous PCR primers designed for conserved regions of homeobox genes of the Antp-Hox family (Seimiya et al. 1994, 1997, Coutinho et al. 1994, Kruse et al. 1994, Degnan et al. 1995, Hoshiyama et al. 1998, Manuel and Le Parco 2000, Wiens et al. 2003, Perovic et al. 2003, Hill et al. 2004, Larroux et al. 2006). None of these authors succeeded, and the sponge Antp-Hox genes reported by Degnan et al. (1995) (SpoxH1 and SpoxH2) were isolated from an ascidian contaminant DNA, as already indicated by Manuel and Le Parco (2000). We can not draw any conclusion until the sequencing of the sponge genome has been completed, but all the failures to identify a sponge Antp-Hox gene, using a precise methodology, could be an evidence that the sponge type of gastrulation, or its multicellular grade of organization, precedes the origin of Antp-Hox gene interactions and Bilateria type of antero-posterior axis patterning. If gastrulation does occur in sponges, it would be very simple and primitive when compared with that in other metazoans that already contain these genetic toolkits.

(Seimiya et al. 1994), SpoxTA1 (Degnan et al. 1995) and EmH-3 (Richelle-Maurer et al. 1998). Coutinho et al. (2003) published a possible scheme for classifying these related and problematic sponge homeobox genes into the Lbx/Tlx gene family. This result was also confirmed by Gauchat et al. (2000), Wiens et al. (2003), Hill et al. (2004) and Larroux et al. (2006). Actually, Tlx and Lbx are closely related homeobox gene families (Pollard and Holland 2000, Jagla et al. 2001, Coutinho et al. 2003), suggesting a common evolutionary origin.

Ancestral NKL-like genes co-originating with spongiomorphs: genetic programs for mesenchymal cells?
The close homology among Tlx and Lbx gene family extends to human, mouse, amphioxus, Anopheles and Drosophila genomic organization, raising the possibility of gene duplication followed by evolutionarily conserved co-regulation. Tlx and Lbx are linked in the vertebrate and Drosophila genomes, and they are believed to have evolved from an ancient gene cluster common in ancestral Bilateria (NKL-like), which has been secondarily split in the chordate ancestry (Pollard and Holland 2000, Jagla et al. 2001, Luke et al. 2003). From the ancestral NK gene cluster, only the TlxLbx and NK3NK4 linkages have been retained in chordates. No split has occurred in the 93DE (NKL) gene cluster of Drosophila or Anopheles, which retained the ancestral physical linkage of slouch, bap, tin, 93Bal/C15/311 (Tlx), ldl and lbe (Jagla et al. 2001). This is the opposite of the splitting of the Hox cluster in Drosophila and its conservation in chordates. The Drosophila 93DE homeobox gene cluster (NKL) appears to participate in a network of gene interactions that governs progressive cell fate decisions during Drosophila mesoderm patterning, just downstream from gastrulation (Jagla et al. 2001). Since it is now established that sponges have homeobox genes classified in the Nk3 and Tlx/Lbx families, that form the 93DE gene complex in Drosophila, it is reasonable to suppose that these are ancient homeobox gene families from which the 93DE/NKL homeobox gene complex evolved, but which was further split in the chordate NKL. Because homeobox gene families Nk3 and Tlx/Lbx have only been found in animal lineages, as far as we know, we assume that these genes appeared in the early stages of metazoan evolution. Alternatively, but less probably, these gene families could have existed in other phyla (prokaryotes, fungi and plants), but were subsequently lost. The grouping of Ctenophora, Porifera and Nematoda Tlx genes in a phylogenetic tree is in agreement with the early origin of Tlx gene family in basal metazoans (Martinelli and Spring 2005). We do not know the role and genetic map of Nk3 and Tlx/ Lbx in sponges. Further investigations in these directions are key points to determine if the ancestral NKL-like genes cooriginated with ancestral spongiomorphs. It will also address its probable co-evolution with mesoderm development, controlling proliferation and differentiation of mesenchymal cells. Recently, the NKL gene complex was linked to mesoderm origin and evolution (Garcia-Fernandez 2005).

Phylogenetic relationship among sponges homeobox genes

Many homeobox genes have been identified in sponges, as reviewed by Manuel and Le Parco (2000), Gauchat et al. (2000), Wiens et al. (2003) and Larroux et al. (2006). Unlike these authors, we considered the Antp-nonHox NK3/ Bap gene family as a separate sub-group of the NKL family, as do Pollard and Holland (2000) and Jagla et al. (2001). Contrastingly to the analysis done by Gauchat and coworkers, who did not included NK3/Bap genes, it is generally assumed that Seimiya et al. (1994) was the first to identify sponge Antp-nonHox genes of the NK-3/Bapx (prox1) and Msx (prox3) families. Another Antp-nonHox gene (BarBshHb and RenBsh) was later added to this list, and they were classified into the Bsh/Bar homeobox gene family (Hill et al. 2004, Larroux et al. 2006). The nonAntp homeobox gene families, such as Iroquois (SUBDOIRX-a), POU (spou1 and spou-2) and PAX (sPax2/5/8) were further identified (Seimiya et al. 1997, Hoshiyama et al. 1998, Perovic et al. 2003, Larroux et al. 2006). Manuel and Le Parco (2000) could not classify several other closely related sponge genes into an obvious orthologous group of homeobox genes. This related group contained EfH-1 (Coutinho et al. 1994), prox2

290

The Drosophila 93DE homeobox gene complex is under the control of BMP2/4-dpp signaling during gastrulation. Acting as morphogen, the concentration of dpp (in a functional state) along the dorso-ventral axis of the Drosophila mesoderm determines the 93DE gene that will be expressed. Since NKL was split in the chordate lineage, we do not expect extended evolutionary conservation during mesoderm patterning in protostomes and deuterostomes, but analyzing each gene individually it is possible to recognize, in some aspects, similar functions inducing cell proliferation and differentiation control of mesodermally derived cells (Jagla et al. 1995, Newman et al. 1997, Park et al. 1998, Rovescalli et al. 2000, Lettice et al. 2001, Holland 2003). Genes of the NKL/93DE complex are not mesodermal markers, and they participate in other developmental processes, commonly controlling cell proliferation and differentiation. Since Nk3 (Bap) and Tlx gene families were described in sponges, further functional analysis of these sponge homeobox genes would distinguish deep homology among the Nk3 (Bap), Tlx and by consequence, the NKL complex. These functional analyses would thus address the hypothesis wether sponges are organized by mesenchymal cell programs which originated mesoderm germ layer.

Are Tlx genes deeply homologous?

The mouse Tlx-2/Hox11L1 promoter is an established model to test the presence of BMP2/4, as Tlx-2 is a direct target for the BMP2/4 signaling pathway (Guo et al. 2001). Tlx-2/Hox11L1 is expressed at the highest concentration of BMP in extraembryonic mesoderm of the ventral yolk sac (extraembrionary hematopoietic tissue) of the mouse embryo (Tang et al. 1998). This is similar to the expression of the homologous C15 gene in the homologous BMP rich dorsal amnioserosa of the Drosophila gastrula (Jagla et al. 2001). This comparative expression analysis indicated evidences for homologously conserved gene interaction during dorsoventral patterning, since the BMP rich ventral region of arthropods is homologous to the BMP rich dorsal region in vertebrates. The comparison analysis can be extended to the functional level and new evidences of deep homology among Tlx genes were discussed. Results from hox11L1/tlx-2 knockout mice from diverse groups are different. First, postnatal lethality was associated with neurological problems (Hatano et al. 1997, Shirasawa et al. 1997). Later, Tang et al. (1998) observed embryonic lethality caused by proliferation failure in embryonic and extra-embryonic (BMP rich primitive hematopoiesis in yolk sac) mesoderm during gastrulation. This result is in agreement with Tlx expression in the BMP rich region, a hematopoietic region in Drosophila and vertebrates. Further analyses of functional and expression data suggest the involvement of the Tlx/Hox11 gene family in hematopoiesis control. The human Hox11/tcl3/Tlx1 (homeobox gene/T-cell leukemia) is a proto-oncogene first described in leukemic T cell patients with accumulation of blast cells that fail to differentiate normally (Dube et al. 1991, Hatano et al. 1991, Kennedy et al. 1991). Its oncogenic activity has been confirmed in myeloid lineages by bone marrow transfection with retroviruses containing Hox11 (Hawley et al. 1994). After comparative expression analysis, clinical cases, knock-

out and enforced expression, there is enough data suggesting that Tlx expression could be associated with proliferation of immature cells, mainly the BMP rich hematopoietic region, and with the delay or abrogation of their terminal differentiation. Additional data is in agreement with this hypothesis, as seen in the following sentences. Erythroblasts already committed to the differentiation program lose all differentiation markers and become mesenchymal adherent cells after enforced expression of Hox11/Tlx1 (Greene et al. 2002). Moretti et al. (1994) detected by RT-PCR the expression of the Hox11/Tlx1 gene in very uncommitted human bone marrow CD34+ cells, which are target cells for BMP signaling (Tang et al. 1998). Retrovirus-transfected ES cells expressing the HOX11 gene became representative of early stages of primitive yolk sac hematopoiesis, and contain primitive erythroblast and monocyte cell lineages (Keller et al. 1998). The cell lineage K562, representing yolk sac primitive hematopoiesis, expresses a Tlx gene, but only during the proliferative immature state (Coutinho et al. 2003). Hox11 enforced expression successfully immortalized cell lineages from yolk sac, which rendered 26 lineages characterized as primitive hematopoietic and eight as hemangioblasts (Yu et al. 2002). These data link Tlx gene family with induction of cell proliferation and abrogation of differentiation progress, or even programming cells to become primitive-hematopoietic cells of the yolk sac extraembryonic mesoderm. If the role of Tlx gene family is deeply conserved, from sponges to vertebrates, then the hypothesis of an early origin of mesoderm will be supported. Knowledge of genetic controls in sponge stem cell biology is scarce. One of the most studied transcription factors known to be specifically expressed by sponge mesenchymal stem cells, the archaeocytes, is EmH-3 (Tlx homeobox gene family) (Richelle-Maurer et al. 1998, Richelle-Maurer and van de Vyver 1999). RT-PCR analysis of EmH-3 expression indicated that it is correlated with archaeocyte proliferation after gemmule hatching. It is down regulated during differentiation of a functional aquiferous system, which is the major morphogenetic phenomenon during the development of a functional sponge. The temporal expression of an EmH-3 homologous (Renprox2) in Haliclona demosponge development was recently published by Larroux et al. (2006). The RT-PCR results showed that this Tlx gene family is weakly expressed during early stages of development, when cleavage is the main way for cell proliferation. The mRNA level increases significantly during metamorphosis and juvenile form of adult sponge, a period of much higher mitosis activity. These expression data seem to be similar to what is observed in Bilaterian Tlx genes. Actually, increased Tlx/EmH3 expression is associated with mitoses of immature cells, and with delay or abrogation of their terminal differentiation. Regulatory similarities between sponge EmH-3 and human Tlx promoters were proposed (Coutinho et al. 1998, 2003). Using a reporter-gene strategy, the EmH-3 promoter was shown to be operational in mouse 3T3 cell lineage and in the self-renewal and differentiation of the human cell line K562, a representative of human yolk sac hematopoiesis. The latter cells express the endogenous Tlx gene, but down regulates it when induced to differentiate with sodium butyrate. The

291

EmH-3 promoter was active in K562 undifferentiated cells and down regulated during their differentiation. This is in agreement with the expression pattern of the endogenous Tlx gene of K562 cells, and with the high expression of the Emh-3 gene in the sponge archaeocytes followed by its down regulation after differentiation (Richelle-Maurer and van de Vyver 1999). This heterologous model provided evolutionary evidence of a deep conservation of Tlx expression, associating Tlx/EmH3 expression with proliferation and delay or abrogation of terminal differentiation of immature sponge and yolk sac metazoan precursor cells (K562). Genes originally performing a defined function during development can be co-opted during evolution to play secondary roles in different embryonic processes. It has happened with the Tlx gene family, which is also expressed in certain regions of the developing central nervous system, the mesenchymal cranial neural crest, and the proliferative fibroblastoid cell mass that gives rise to the spleen (Roberts et al. 1994, Logan et al. 1998). These are apparent secondary functions, since C15 (Drosophyla) is not expressed in neurons. Considering the control of proliferation and differentiation of extra embryonic mesoderm as the primary function of Tlx gene family in mouse and Tlx deep conservation, at least in expression regulation, the most plausible scenario for metazoan origin is a urmetazoan spongiomorph with mesenchymal stem cell lineages under the control of primitive NKL homeobox genes (TLX for instance) for proliferation initiation and abrogation of terminal differentiating program. Following the discussion on sponge development presented before it would be the base for the later origin of mesoderm. Since the sponge genome contains TGF (BMP) receptors (Suga et al. 1999), a further speculative scenario would hypothesize that BMP signaling and Tlx genes formed a tool kit that originated in spongiomorph urmetazoan and was further evolutionarily conserved in archaeocytes, the sponge stem cell, and extra-embryonic mesoderm in vertebrates. These are very speculative thoughts, and additional functional studies with sponge and vertebrate homeobox genes would bring new evidence which would corroborate, or not, these hypotheses.

gastrulation, the preservation of cells in their adult tissue, and new epithelial stem-cell systems for endoderm and ectoderm, in addition to the earliest mesenchymal stem-cell system. The apparent absence of Hox genes in sponges and the absence of this gene interaction system for antero-posterior axis patterning were here interpreted as evidence against the concept of a sponge gastrulation strictu sensu. The presence of NKL genes in sponges would suggest ancestral NKLlike genes co-originating with ancestral spongiomorphs. Moreover, Nk3 (Bap) and Tlx would be the precursor of the Nkl complex. If functional conservation is expected among sponges and bilaterian NKL genes, the primary function in the ancestral spongiomorph would be associated with individual programs for mesenthelial and mesenchymal cell types, and would later contribute to the origin of mesoderm genetic programs. Expression and heterologous data corroborated the hypothesis of deep homology among Tlx genes. However, more data at the functional level could bring more solid evidences for it. If so, this genetic interaction would be one of the bases for the origin of mesoderm in triploblasts. According to von Baers laws, the general feature of a large group of animals appears earlier in development than do the specialized features of a smaller group and less general characters develop from the more general until the most specialized appear. Bilaterians do not cross a diploblast phase during their development, but they cross a period with mesenchymal cells in BMP rich vegetal pole, which is homologous to vertebrate yolk sac extraembryonic mesoderm. Considering sponges at the base of monophyletic evolution of animals, and if our hypotheses are further confirmed by additional studies, a general and earlier feature among animal development is the mesenchymal cell mass which evolved to BMP rich mesoderm in invertebrates and to BMP rich yolk sac extraembryonic mesoderm in vertebrates.

Acknowledgments
Many thanks to Dr. Radovan Borojevic who revised the manuscript, Dr Adrian T. Sumner for English revision and FAPERJ for financial support (Grant: Pronex-FAPERJ number: E-26/171213/2003).

Conclusion
It was concluded from the analysis of sponge embryology that the developmental processes in these organisms are mainly based on cellular terms and not in terms of germ layers and gastrulation process. Sponge type of primitive gastrulation was here considered to be an individual sequence of complementary cell differentiation events converting flagellated cells into amoeboid morphology, as happened, in a simpler way, in choanoflagellate colonies. Based on functional and morphological data, amoeboid and flagellated sponge cells apparently belong to the mesenthelial and mesenchymal type rather than to the true epithelial type derived from ectoderm and endoderm. It is thus suggested that sponge would have precursors for mesoderm origin. The appearance of a true epithelium with a basal lamina would contribute to the major difference between the multicellular grade of organization of Porifera and Cnidaria, with the appearance of an ordered placement of tissues as a result of

References
Adell T, Gamulin V, Perovic-Ottstadt S, Wiens M, Korzhev M, Mller I, Mller WEG (2004) Evolution of metazoan cell junction proteins: the scaffold protein MAGI and the transmembrane receptor tetraspanin in the demosponge Suberites domuncula. J Mol Evol 59: 41-50 Adell T, Mller WEG (2005) Expression pattern of the Brachyury and Tbx2 homologs from the sponge Suberites domuncula. Biol Cell 97: 641-650 Allaby M (ed) (1985) The Oxford dictionary of natural history. Oxford University Press, Oxford Amano S, Hori I (1993) Metamorphosis of calcareous sponges II. Cell rearrangement and differentiation in metamorphosis. Invertebr Reprod Dev 24: 13-26 Bagby RM (1970) The fine structure of pinacocytes in the marine sponge Microciona prolifera (Elis and Solander). Zeit Zellforsch 105: 579-594

292

Bekkum DWV (2004) Phylogenetic aspects of tissue regeneration: role of stem cells. A concise overview. Blood Cells Mol Dis 32: 11-16 Bode HR (1996) The interstitial cell lineage of hydra: a stem cell system that arose early in evolution. J Cell Sci 109: 1155-1164 Boelsterli U (1977) An electron microscopic study of early developmental stages, myogenesis, oogenesis and cnidogenesis in the anthomedusa, Podocoryne carnea M. Sars. J Morphol 154: 259-289 Borchiellini C, Manuel M, Alivon E, Boury-Esnault N, Vacelet J, Le Parco Y (2001) Sponge paraphyly and the origin of Metazoa. J Evol Biol 14: 171-179 Borchiellini C, Chombard C, Manuel M, Alivon E, Vacelet J, Boury-Esnault N (2004) Molecular phylogeny of Demospongiae: implications for classification and scenarios of character evolution. Mol Phylogenet Evol 32: 823-837 Borojevic R (1963) Essai de reconstitution du choanoderme partir dune suspension choanocytaire dponge acalcaire. Comp Rend Acad Sci Paris III: 961-962 Borojevic R (1966a) Experimental study of the cell differentiation of the sponge during the course of its development. Dev Biol 14: 130-153 Borojevic R (1966b) La ponte et le dveloppement de Polymastia robusta (Dmosponges). Cah Biol Mar 8: 1-6 Borojevic R (1968) tude du developpement et de la diffrenciation cellulaire dponges calcaries calcinennes (genres Clathrina et Ascandra). Ann Embryol Morphogen 2: 15-36 Borojevic R (1970) Diffrenciation cellulaire dans l`embryogense et la morphogense chez les Spogiaires. In: Fry WG (ed). The biology of the Porifera. Academic Press, London. pp. 467-490 Borojevic R (1971) Le comportement des cellules dponge lors de processus morphognetiques. Anne Biol 10: 533-545 Borojevic R, Fry WG, Jones WC, Lvi C, Rasmont R, Sar M, Vacelet J (1967) Mise au point actuelle de la terminologie des ponges (A reassessment of the terminology for sponges). Bull Mus Natl Hist Nat 2: 1224-1235 Borojevic R, Lvi C (1965) Morphogense exprimentale dune ponge partir des cellules de la larve nageante dissocie. Zeit Zellforsch 68: 57-69 Boury-Esnault N, Efremova S, Bzac C, Vacelet J (1999) Reproduction of a hexactinellid sponge: first description of gastrulation by cellular delamination in the Porifera. Invertebr Reprod Dev 35: 187-201 Boury-Esnault N, Ereskovsky A, Bzac C, Tokina. D (2003) Larval development in the Homoscleromorpha (Porifera, Demospongiae). Invertebr Biol 122: 187-202 Boury-Esnault N (2006) Systematics and evolution of Demospongiae. Can J Zool 84: 205-224 Boute N, Exposito JY, Boury-Esnault N, Vacelet J, Noro N, Miyazaki K, Yoshizato K, Garrone R (1996) Type IV collagen in sponges, the missing link in basement membrane ubiquity. Biol Cell 88: 37-44 Brien P (1937) La rorganisation de lponge aprs dissociation par filtration et phnomnes dinvolution chez Ephydatia fluviatilis. Arch Biol 48: 185-268 Brien P (1967) Les ponges. Leur nature metazoire, leur gastrulation, leur tat colonial. Ann Soc R Zool Belg 97: 197-235 Brooke NM, Holland PW (2003) The evolution of multicellularity and early animal genomes. Curr Opin Genet Dev 13: 599-603 Buss LW (1983) Evolution, development, and the units of selection. Proc Natl Acad Sci USA 80: 1387-1391

Cavalier-Smith T, Allsopp MTEP, Chao EE, Boury-Esnault N, Vacelet J (1996) Sponge phylogeny, animal monophyly, and the origin of the nervous system: 18S rRNA evidence. Can J Zool 74: 2031-2045 Chourrout D, Delsuc F, Chourrout P, Edvardsen RB, Rentzsch F, Renfer E, Jensen MF, Zhu B, Jong PD, Steele RE, Technau U (2006) Minimal ProtoHox cluster inferred from bilaterian and cnidarian Hox complements. Nature 442: 684 - 687 Collins AG (1998) Evaluating multiple alternative hypotheses for the origin of Bilateria: an analysis of 18S rRNA molecular evidence. Proc Natl Acad Sci USA 95: 15458-15463 Coutinho CC, Vissers S, van de Vyver G (1994) Evidence of homeobox genes in the freshwater sponge Ephydatia fluviatilis. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 385-388 Coutinho CC, Seack J, van de Vyver G, Borojevic R, Mller WEG (1998) Origin of the metazoan bodyplan: characterization and functional testing of the promoter of the homeobox gene EmH3 from the freshwater sponge Ephydatia muelleri in mouse 3T3 cells. Biol Chem 379: 1243-1251 Coutinho CC, Fonseca RN, Mansure JJ, Borojevic R (2003) Early steps in the evolution of multicellularity: deep structural and functional homologies among homeobox genes in sponges and higher metazoans. Mech Dev 120: 429-440 Degnan BM, Degnan SM, Giusti A, Morse DE (1995) A hox/hom homeobox gene in sponges. Gene 155: 175-177 Degnan BM, Leys SP, Larroux C (2005) Sponge development and antiquity of animal pattern formation. Integr Comp Biol 45: 335341 Delage (1892) Embriogense des ponges siliceuses. Arch Zool Exp Gen 10: 345-498 Dewel RA (2000) Colonial origin for Emetazoa: major morphological transitions and the origin of bilaterian complexity. J Morphol 243: 35-74 Dube ID, Kamel-Reid S, Yuan CC, Lu M, Wu X, Corpus G, Raimondi SC, Crist WM, Carroll AJ, Minowada J, Baker JB (1991) A novel human homeobox gene lies at the chromosome 10 breakpoint in lymphoid neoplasias with chromosomal translocation t(10;14). Blood 78: 2996-3003 Duboule D, Dolle P (1989) The structural and functional organization of the murine HOX gene family resembles that of Drosophila homeotic genes. EMBO J 8: 1497-1505 Duboule D (1994) Temporal colinearity and the phylotypic progression: a basis for the stability of a vertebrate Bauplan and the evolution of morphologies through heterochrony. Dev Suppl: 1994: 135-142 Echeverri K, Tanaka EM (2002) Ectoderm to mesoderm lineage switching during axolotl tail regeneration. Science 298: 19931996 Eerkes-Medrano DI, Leys SP (2006) Ultrastructure and embryonic development of a syconoid calcareous sponge. Invertebr Biol 125(3): 177-194 Efremova SM (1997) Once more on the position among Metazoa - Gastrulation and germinal layers of sponges. Berl Geowiss Abh 20: 7-15 Ereskovsky AV, Korotkova GP (1997) The reasons of sponge sexual morphogenesis peculiarities. Berl Geowiss Abh 20: 25-33

293

Ereskovsky AV, Gonobobleva EL (2000) New data on embryonic development of Halisarca dujardini Johnston, 1842 (Demospongiae: Halisarcida). Zoosystema 22: 355-368 Ereskovsky AV (2004) Comparative embryology of sponges and its application for poriferan phylogeny. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millenium. Boll Mus Inst Biol Univ Genova 68: 301-318 Faur-Fremiet ME (1932) Morphogense exprimentale chez Ficulina ficus L. Archs Anat Microsc 28: 1-80 Fell PE (1974) Porifera. In: Giese AC, Pearse JS (eds). Reproduction of marine invertebrates. Academic Press, New York. pp. 51-132 Ferrier DE, Minguillon C (2003) Evolution of the Hox/ParaHox gene clusters. Int J Dev Biol 47: 605-611 Funayama N, Nakatsukasa M, Hayashi T, Agata K (2005) Isolation of the choanocyte in the fresh water sponge, Ephydatia fluviatilis and its lineage marker, Ef annexin. Develop Growth Differ 47: 243-253 Galle S, Yanze N, Seipel K (2005) The homeobox gene Msx in development and trans-differentiation of the jellyfish striated muscle. Int J Dev Biol 49: 961-967 Garcia-Fernandez J (2005) The genesis and evolution of homeobox gene clusters. Nat Rev Genet 6: 881-892 Gauchat D, Mazet F, Berney C, Schummer M, Kreger S, Pawlowski J, Galliot B (2000) Evolution of Antp-class genes and differential expression of Hydra Hox/paraHox genes in anterior patterning. Proc Natl Acad Sci USA 97: 4493-4498 Ginsburg GT, Kimmel AR (1997) Autonomous and nonautonomous regulation of axis formation by antagonistic signaling via 7-span cAMP receptors and GSK3 in Dyctyostelium. Gene Dev 11: 21122123 Graham A, Papalopulu N, Krumlauf R (1989) The murine and Drosophila homeobox gene complexes have common features of organization and expression. Cell 57: 367-378 Greene WK, Ford J, Dixon D, Tilbrook PA, Watt PM, Klinken SP, Kees UR (2002) Enforced expression of HOX11 is associated with an immature phenotype in J2E erythroid cells. Br J Haematol 118: 909-917 Grimson MJ, Coates JC, Reynolds JP, Shipman M, Blanton RL, Harwood AJ (2000) Adherens junctions and beta-catenin-mediated cell signaling in a non-metazoan organism. Nature 408: 727-731 Guo X, Lin Y, Horbinski C, Drahushuk KM, Kim IJ, Kaplan PL, Lein P, Wang T, Higgins D (2001) Dendritic growth induced by BMP-7 requires Smad1 and proteasome activity. J Neurobiol 48: 120-130 Haeckel E (1872) Die Kalkschwmme. Eine Monographie in Zwei Bnden Text und einem Atlas mit 60 Taflen Abbildungen. Georg Reimer, Berlin Haeckel E (1874) Die Gastraea-Theorie, die phylogenetische classification des Thierreichs und die homologie der keimblatter. Jenaische Ztschr, Naturw 8: 1-55 Hammer E (1908) Neu beitrage zur kenntnis der histologie und entwicklung von Sycon raphanus. Arch Biontol 2: 289-334 Harrison FW (1972) The nature and the role of the basal pinacoderm of Corvomeyenia carolinensis Harrison (Porifera: Spongillidae). Hydrobiologia 39: 495-508 Harrison FW, de Vos L (1991) Porifera. In: Harrison FW; Westfall JA (eds). Microscopic anatomy of invertebrates: Placozoa, Porifera, Cnidaria, and Ctenophora, Vol. 2. Wiley-Liss, New York. pp 2989

Hatano M, Roberts CW, Minden M, Crist WM, Korsmeyer SJ (1991) Deregulation of a homeobox gene, HOX11, by the t(10;14) in T cell leukemia. Science 253: 79-82 Hatano M, Aoki T, Dezawa M, Yusa S, Iitsuka Y, Koseki H, Taniguchi M, Tokuhisa T (1997) A novel pathogenesis of megacolon in Ncx/ Hox11L.1 deficient mice. J Clin Invest 100: 795-801 Hawley RG, Fong AZ, Lu M, Hawley TS (1994) The HOX11 homeobox-containing gene of human leukemia immortalizes murine hematopoietic precursors. Oncoge 9: 1-12 Hayward DC, Miller DJ, Ball EE (2004) Snail expression during embryonic development of the coral Acropora: blurring the diploblast/triploblast divide? Dev Genes Evol 214: 257-260 Hill A, Tetrault J, Hill M (2004) Isolation and expression analysis of a poriferan Antp-class Bar-/Bsh-like homeobox gene. Dev Genes Evol 214: 515-523 Holland LZ (2000) Body-plan evolution in the Bilateria: early anteroposterior patterning and the deuterostome-protostome dichotomy. Curr Opin Genet Dev 10: 434-442 Holland PW (2003) More genes in vertebrates? J Struct Funct Genom 3: 75-84 Hoshiyama D, Suga H, Iwabe N, Koyanagi M, Nikoh N, Kuma K, Matsuda F, Honjo T, Miyata T (1998) Sponge Pax cDNA related to Pax-2/5/8 and ancient gene duplications in the Pax family. J Mol Evol 47: 640-648 Huxley J (1911) Some phenomena of regeneration in Sycon. Phil Trans R Soc 202: 165-189 Hyman LH (1940) The invertebrates: Protozoa through Ctenophora. McGraw-Hill, New York Imhof BA, Aurrand-Lions M (2004) Adhesion mechanisms regulating the migration of monocytes. Nat Rev Immunol 4: 432444 Jagla K, Dolle P, Mattei MG, Jagla T, Schuhbaur B, Dretzen G, Bellard F, Bellard M (1995) Mouse Lbx1 and human LBX1 define a novel mammalian homeobox gene family related to the Drosophila lady bird genes. Mech Dev 53: 345-356 Jagla K, Bellard M, Frasch M (2001) A cluster of Drosophila homeobox genes involved in mesoderm differentiation programs. BioEssays 23: 125-133 Kamm K, Schierwater B, Jakob W, Dellaporta SL, Miller DJ (2006) Axial patterning and diversification in the cnidaria predate the Hox system. Curr Biol 16: 920-926 Keller G, Wall C, Fong AZ, Hawley TS, Hawley RG (1998) Overexpression of HOX11 leads to the immortalization of embryonic precursors with both primitive and definitive hematopoietic potential. Blood 92: 877-887 Kennedy MA, Gonzalez-Sarmiento R, Kees UR, Lampert F, Dear N, Boehm T, Rabbitts TH (1991) HOX11, a homeobox-containing T-cell oncogene on human chromosome 10q24. Proc Natl Acad Sci USA 88: 8900-8904 King N, Hittinger CT, Carroll SB (2003) Evolution of key cell signaling and adhesion protein families predates animal origins. Science 301: 361-363 King N (2004) The unicellular ancestry of animal development. Dev Cell 7: 313-325 Koziol C, Borojevic R, Steffen R, Mller WEG (1998) Sponges (Porifera) model systems to study the shift from immortal to senescent somatic cells: the telomerase activity in somatic cells. Mech Ageing Dev 100, 107-20 Kruse M, Mikoc A, Cetkovic H, Gamulin V, Rinkevich B, Mller IM, Mller WEG (1994) Molecular evidence for the presence of

294

a developmental gene in the lowest animals: identification of a homeobox-like gene in the marine sponge Geodia cydonium. Mech Ageing Dev 77: 43-54 Kruse M, Leys SP, Mller IM, Mller WEG (1998) Phylogenetic position of the Hexactinellida within the phylum Porifera based on the amino acid sequence of the protein kinase C from Rhabdocalyptus dawsoni. J Mol Evol 46: 721-728 Larroux C, Fahey B, Liubicich D, Hinman VF, Gauthier M, Gongora M, Green K, Worheide G, Leys SP, Degnan BM (2006) Developmental expression of transcription factor genes in a demosponge: insights into the origin of metazoan multicellularity. Evol Dev 8: 150-173 Leptin M (2005) Gastrulation movements: the logic and the nuts and bolts. Dev Cell 8: 305-320 Lettice L, Hecksher-Sorensen J, Hill R (2001) The role of Bapx1 (Nkx3.2) in the development and evolution of the axial skeleton. J Anat 199: 181-187 Lvi C (1956) tude des Halisarca de Roscoff. Embryologie et systmatique des dmosponges. Arch Zool Exp Gn 93: 1-184 Lvi C (1963) Gastrulation and larval phylogeny in sponges. In: Dougherty EC (ed). The lower metazoa: comparative biology and phylogeny. University of California Press, Los Angeles. pp. 375382 Lvi C (1970) Les cellules des ponges. In: Fry WG (ed). Symposia of the Zoological Society. Academic Press, London. pp. 353-364 Leys SP, Degnan BM (2002) Embryogenesis and metamorphosis in a haplosclerid demosponge: gastrulation and transdifferentiation of larval ciliated cells to choanocytes. Invertebr Biol 121: 171-189 Leys SP (2003) Comparative study of spiculogenesis in demosponge and hexactinellid larvae. Microsc Res Tech 62: 300-311 Leys SP (2004) Gastrulation in sponges. In: Stern C (ed). Gastrulation. Cold Spring Harbor Laboratory Press, Cold Spring Harbor. pp. 23-31 Leys SP, Eerkes-Medrano D (2005) Gastrulation in calcareous sponges: in search of Haeckels gastraea. Comp Biol 45: 342-351 Leys SP, Ereskovsky AV (2006) Embryogenesis and larval differentiation in sponges. Can J Zool 84: 262-287 Logan C, Wingate RJ, Kay IJM, Lumsden A (1998) Tlx-1 and Tlx-3 homeobox gene expression in cranial sensory ganglia and hindbrain of the chick embryo: markers of patterned connectivity. J Neurosci 18: 5389-5402 Luke GN, Castro LF, McLay K, Bird C, Coulson A, Holland PW (2003) Dispersal of NK homeobox gene clusters in amphioxus and humans. Proc Natl Acad Sci USA 100: 5292-5295 Maldonado M (2004) Choanoflagellates, choanocytes, and animal multicellularity. Invertebr Biol 123: 1-22 Manuel M, Le Parco Y (2000) Homeobox gene diversification in the calcareous sponge, Sycon raphanus. Mol Phylog Evol 17: 97-107 Manuel M, Borchiellini C, Alivon E, Le Parco Y, Vacelet J, BouryEsnault N (2003) Phylogeny and evolution of calcareous sponges: monophyly of Calcinea and Calcaronea, high level of morphological homoplasy, and the primitive nature of axial symmetry. Syst Biol 52: 311-333 Manuel M, Parco YL, Borchiellini C (2004) Comparative analysis of Brachyury T- domains, with the characterization of two new sponge sequences, from a hexactinellid and a calcisponge. Gene 340: 291-301 Manuel M (2006) Phylogeny and evolution of calcareous sponges. Can J Zool 84: 225-241

Marcellini S, Technau U, Smith JC, Lemaire P (2003) Evolution of brachyury proteins: identification of a novel regulatory domain conserved within Bilateria. Dev Biol 260: 352-361 Martindale MQ, Pang K, Finnerty JR (2004) Investigating the origins of triploblasty: mesodermal gene expression in a diploblastic animal, the sea anemone Nematostella vectensis (phylum Cnidaria; class Anthozoa). Development 131: 2463-2474 Martindale MQ (2005) The evolution of metazoan axial properties. Nat Rev Gen 6: 917-927 Martinelli C, Spring J (2005) T-box and homeobox genes from the ctenophore Pleurobrachia pileus: comparison of brachyury, Tbx2/3 and Tlx in basal metazoans and bilaterians. FEBS Lett 579: 5024-5028 Masuda-Nakagawa LM, Groer H, Aerne BL, Schmid V (2000) The HOX-like gene Cnox2-Pc is expressed at the anterior region in all life cycle stages of the jellyfish Podocoryne carnea. Dev Genes Evol 210: 151-156 McGinnis W, Krumlauf R (1992) Homeobox genes and axial patterning. Cell 68: 283-302 Medina M, Collins AG, Silberman J, LSogin M (2001) Evaluating hypotheses of basal animal phylogeny using complete sequences of large and small subunit rRNA. Proc Natl Acad Sci 98: 97079712 Metschnikoff E (1879) Spongiolische Studien. III. Entwicklungsgeschichtliches ber die Kalkschwmme. Z wiss Zool 32: 362-371 Moretti P, Simmons P, Thomas P, Haylock D, Rathjen P, Vadas M, DAndrea R (1994) Identification of homeobox genes expressed in human haemopoietic progenitor cells. Gene 144: 213-219 Morris PJ (1993) The developmental role of the extracellular matrix suggests a monophyletic origin of the kingdom animalia. Evolution 47: 152-165 Mller P, Seipel K, Yanze N, Reber-Mller S, Streitwolf-Engel R, Stierwald M, Spring JVS (2003) Evolutionary aspects of developmentally regulated helix-loop-helix transcription factors in striated muscle of jellyfish. Dev Biol 255: 216-229 Mller WA, Teo R, Frank U (2004) Totipotent migratory stem cells in a hydroid. Dev Biol 275: 215-224 Mller WEG (2001) How was metazoan threshold crossed? The hypothetical Urmetazoa. Comp Biochem Physiol A 129: 433-460 Newman CS, Grow MW, Cleaver O, Chia F, Krieg P (1997) Xbap, a vertebrate gene related to bagpipe, is expressed in developing craniofacial structures and in anterior gut muscle. Dev Biol 181: 223-233 Park M, Lewis C, Turbay D, Chung A, Chen JN, Evans S, Breitbart RE, Fishman MC, Izumo S, Bodmer R (1998) Differential rescue of visceral and cardiac defects in Drosophila by vertebrate tinmanrelated genes. Proc Natl Acad Sci USA 95: 9366-9371, 13348 Perovic S, Schroder HC, Sudek S, Grebenjuk VA, Batel R, Stifanic M, Mller IM, Mller WEG (2003) Expression of one sponge Iroquois homeobox gene in primmorphs from Suberites domuncula during canal formation. Evol Dev 5: 240-250 Pollard SL, Holland PW (2000) Evidence for 14 homeobox gene clusters in human genome ancestry. Curr Biol 10: 1059-1062 Rasmont R (1979) Les ponges: des metazoaires et des societes de celules. In: Lvi C, Boury-Esnault N (eds). Biologie des spongiaires. Colloques Internationaux du CNRS, vol. 291. ditions du CNRS, Paris. pp. 21-29 Reddien PW, Alvarado AS (2004) Fundamentals of planarian regeneration. Annu Rev Cell Dev Biol 20: 725-757

295

Richelle-Maurer E, van de Vyver G (1999) Temporal and spatial expression of EmH-3, a homeobox-containing gene isolated from the freshwater sponge Ephydatia muelleri. Mech Ageing Dev 109: 203-219 Richelle-Maurer E, van de Vyver G, Vissers S, Coutinho CC (1998) Homeobox-containing genes in freshwater sponges: characterization, expression, and phylogeny. In: Mller WEG (ed). Progress in molecular and subcellular biology. Vol. 19. SpringerVerlag, Berlin. pp 157-175 Roberts CW, Shutter JR, Korsmeyer SJ (1994) Hox11 controls the genesis of the spleen. Nature 368: 747-749 Rovescalli AC, Cinquanta M, Ferrante J, Kozak CA, Nirenberg M (2000) The mouse Nkx-1.2 homeobox gene: alternative RNA splicing at canonical and noncanonical splice sites. Proc Natl Acad Sci USA 97: 1982-1987 Scholz CB, Technau U (2003) The ancestral role of Brachyury: expression of NemBra1 in the basal cnidarian Nematostella vectensis (Anthozoa). Dev Genes Evol 212: 563-570 Schutze J, Krasko A, Custdio MR, Efremova SM, Mller IM, Mller WEG (1999) Evolutionary relationships of Metazoa within the eukaryotes based on molecular data from Porifera. Proc R Soc Lond B Biol Sci 266: 63-73 Seimiya M, Ishiguro H, Miura K, Watanabe Y, Kurosawa Y (1994) Homeobox-containing genes in the most primitive metazoa, the sponges. Eur J Biochem 221: 219-225 Seimiya M, Watanabe Y, Kurosawa Y (1997) Identification of POUclass homeobox genes in a freshwater sponge and the specific expression of these genes during differentiation. Eur J Biochem 243: 27-31 Seipel K, Schmid V (2005) Evolution of striated muscle: jellyfish and the origin of triploblasty. Dev Biol 282: 14-26 Shirasawa S, Yunker AM, Roth KA, Brown GA, Horning S, Korsmeyer SJ (1997) Enx (Hox11L1)-deficient mice develop myenteric neuronal hyperplasia and megacolon. Nat Med 3: 646650 Simpson TL (1984) The cell biology of sponges. Springer-Verlag, New York Spring J, Yanze N, Middel AM, Stierwald M, Groger H, Schmid Y (2000) The mesoderm specification factor twist in the life cycle of jellyfish. Dev Biol 228: 363-375 Spring J, Yanze N, Josch C, Middel AM, Winninger B, Schmid V (2002) Conservation of Brachyury, Mef2, and Snail in the myogenic lineage of jellyfish: a connection to the mesoderm of Bilateria. Dev Biol 244: 372-384

Suga H, Ono K, Miyata T (1999) Multiple TGF-beta receptor related genes in sponge and ancient gene duplications before the parazoaneumetazoan split. FEBS Lett 453: 346-350 Tang SJ, Hoodless PA, Lu Z, Breitman ML, McInnes RR, Wrana JL, Buchwald M (1998) The Tlx-2 homeobox gene is a downstream target of BMP signalling and is required for mouse mesoderm development. Development 125: 1877-1887 Technau U, Bode HR (1999) HyBra1, a brachyury homologue, acts during head formation in Hydra. Development 126: 999-1010 Technau U (2001) Brachyury, the blastopore and the evolution of the mesoderm. Bioessays 23: 788-794 Technau U, Scholz CB (2003) Origin and evolution of endoderm and mesoderm. Int J Dev Biol 47: 531-539 Tuzet O (1963) The phylogeny of sponges according to embryological, histological and serological data, and their affinities with the Protozoa and the Cnidaria. In: Dougherty EC (ed). The lower metazoan. University of California Press, Los Angeles. pp. 129-148 Watanabe Y (1978) The development of two species of Tetilla (Demosponge). Nat Sci Rep Ochanomizu Univ 29: 71-106 Wiens M, Mangoni A, DEsposito M, Fattorusso E, Korchagina N, Schroder HC, Grebenjuk VA, Krasko A, Batel R, Mller IM, Mller WEG (2003) The molecular basis for the evolution of the metazoan bodyplan: extracellular matrix-mediated morphogenesis in marine demosponges. J Mol Evol 57: S60-75 Willenz P, van de Vyver G (1982) Endocytosis of latex beads by the exopinacoderm in the fresh water sponge Ephydatia fluviatilis: An in vitro and in situ study in SEM and TEM. J Ultrastr Res 79: 294-306 Willmer EN (1970) Cytology and evolution. Academic Press, New York - London Willmer P (1994) Invertebrate relationships: patterns in animal evolution. Cambridge University Press, Cambridge Wilson H (1907) On some phenomena of coalescence and regeneration in sponges. J Exp Zool 5: 245-258 Wolpert L (1992) Gastrulation and the evolution of development. Dev Suppl: 7-13 Younossi-Hartenstein A, Hartenstein V (2000) The embryonic development of the polyclad flatworm Imogine mcgrathi. Dev Genes Evol 210: 383-398 Yu WM, Hawley TS, Hawley RG, Qu CK (2002) Immortalization of yolk sac-derived precursor cells. Blood 100: 3828-3831 Zrzavy J, Mihulka S, Kepka P, Bezdek A, Tietz D (1998) Phylogeny of the Metazoa based on morphological and 18S ribosomal DNA evidence. Cladistics 14: 249-285

Porifera research: Biodiversity, innovation and sustainaBility - 2007

297

Developing methods for commercially farming bath sponges in tropical Australia

Alan R. Duckworth(1, 2*), Carsten Wolff(1, 2), Elizabeth Evans-Illidge(1, 2)
(1) (2)

Australian Institute of Marine Science, PMB 3 Townsville, QLD 4810, Australia. AIMS@JCU, Post Office James Cook University, QLD 4811, Australia. a.duckworth@aims.gov.au

Abstract: For aquaculture to supply sufficient and sustainable quantities of bath sponges, suitable farming methods need to be developed. In this study, we developed farming methods for Coscinoderma sp. and Rhopaloiedes odorabile, two Australian species with potential commercial value. In one experiment we tested 11 farming techniques, grouped into four general categories or methods based on how sponges are supported: threaded line, cut-explant (explants cut half way through then line placed into the cut), spike (hard plastic pushed through) and mesh. For each general category, we trialled different material types, such as nylon and polyethylene, or varied compartment sizes. After 1 year, most Coscinoderma sp. farmed in mesh survived, probably because this general farming category is less damaging to the explant. In contrast, survival in the other general categories was poor, with many explants dying after 2 months. Unlike survival data, growth of Coscinoderma sp. was similar among the farming techniques with most explants almost tripling in size in 1 year. Survival of R. odorabile varied among farming treatments, being highest when farmed in mesh or on threaded lines. Growth of R. odorabile was also highest using these general farming categories, with most explants doubling in size over one year. In a separate experiment, we examined whether survival of Coscinoderma sp. and R. odorabile farmed on threaded lines could be improved by letting the cut explants heal first inside mesh for several months (nursery period) before being threaded. For both sponge species, the nursery period did not improve final explant growth or survival. Overall, both Coscinoderma sp. and R. odorabile grew and survived well when farmed in mesh panels. This farming method will soon be used to commercially culture both bath sponge species in tropical Australia. Keywords: Bath sponges, farming, growth, methods, survival

Introduction
Global demand for bath sponges, for cosmetic, bathroom and industrial use, far exceeds supply (Pronzato 1999). Commercial bath sponges, species from the order Dictyoceratida that have a high quality spongin skeleton, have traditionally been harvested from natural populations in the Mediterranean Sea, Florida and the Caribbean (Stevely and Sweat 1994, Pronzato 1999). Over-harvesting and periodic disease outbreaks have unfortunately greatly reduced these natural populations and severely limited the yield of bath sponges (reviewed in Pronzato 1999), thus alternative supply methods need to be developed. In-sea aquaculture is one method that could supply sufficient and sustainable quantities of bath sponges to meet market demand. Sponges have been experimentally farmed in the sea for over 100 years, with several methods or support structures being tested. The most widely used and one of the earliest methods involved threading sponges onto thin line so that they hung in mid water (Moore 1910, Crawshay 1939). More recently, the threaded line method has been successfully trialled using modern materials (plastics) with several Mediterranean bath sponge species (Verdenal and Vacelet 1990, Pronzato et al. 1999, Corriero et al. 2004) and is employed in Micronesia for small scale commercial farming

(MacMillan 1996). This method is not appropriate for all sponges, however, with some species growing away and dislodging themselves from threaded line (Duckworth et al. 1997, Duckworth and Battershill 2003a). Damage incurred when the line is threaded through the sponges may elicit this response, as could inappropriate line material promoting an antagonistic tissue response (Duckworth and Battershill 2003a). Sponges farmed on threaded line can also be torn off when cultured in areas of strong water flow (Duckworth and Battershill 2003b). One farming method that eliminates these problems is the mesh method, which involves farming sponges in mesh bags or containers (Duckworth et al. 1997, Duckworth and Battershill 2003b, van Treeck et al. 2003, Kelly et al. 2004). Although sponges farmed inside mesh have high survival because tissue damage is minimal, low growth rates may result from the mesh strands reducing water flow and thus food supply (Duckworth and Battershill 2003b). This could potentially be overcome by having a nursery period, whereby sponges are placed inside mesh until they have healed their cut surfaces and reorganised their canal system, and then threaded onto line. It is likely that the farming method for sponge culture will vary among locations or countries, depending on the species requirements and local farming environment.

298

However, regardless of the farming method chosen it must be inexpensive, easy to use and deploy, and promote high sponge survival and growth. In this study we ran two experiments to help develop a good farming method for commercially culturing Coscinoderma sp. and Rhopaloeides odorabile, two species common to the central Great Barrier Reef, Australia. Market analysis has determined that both Coscinoderma sp. and R. odorabile have commercial grade spongin. The first experiment compared growth and survival of sponges cultured using four general farming categories or methods: threaded line, cut-explant, spike and mesh. Each general category additionally trialled different material such as nylon and polyethylene or compartment sizes. Sponges farmed using the threaded line, cut-explant and spike methods are directly exposed to the environment, while the mesh methods are less damaging to the sponge but may restrict water flow to the explant. In the second experiment, we examined the importance of a nursery period for Coscinoderma sp. and R. odorabile, to determine whether growth and survival are higher on threaded line after explants have first healed in mesh. The nursery period of several months allows each explant sufficient time to reorganise their canal system and become a fully-functional sponge, thus possibly better able to withstand the threading process.

Fig. 1: Schematic diagram of the four general methods: A. Threaded line method; B. Cut-explant method, with the cotton thread stitching the two cut sides together shown in white; C. Spike method; D. Mesh method.

Materials and methods Developing a good farming method

In this experiment we tested 11 farming techniques, grouped into four general categories or methods: threadedline, cut-explant, spike and mesh methods. For the threadedline methods, a hollow needle was used to thread line through the middle of each explant (Fig. 1A). For the cut-explant methods, the explant was first cut halfway through, allowing the line to be pushed down into the slot, then a thin cotton thread was used to sew the cut ends of the explant together (Fig. 1B). The cotton would disintegrate within several weeks by which time the two cut sides would rejoin. To determine the importance of line material for both farming categories we tested nylon, polyethylene and polypropylene lines. All lines were 4 mm thick, about 20 cm long, with one explant situated at mid-length. For the spike-methods, sponges were farmed on cross shaped pieces of PVC or polyethylene plastic that were sharpened at one end (Fig. 1C). An explant was carefully pushed onto the spike so that it rested on the plastic base. Each spike was 15 cm in length and about 0.5 cm in width. For the remaining farming category sponges were farmed inside mesh (Fig. 1D), comparing bags, aquapurses (Tooltech PTY Limited) and panels (Australian NetMakers). Both the aquapurse and mesh panels are used to commercially farm oysters and for statistical independence they were broken down into small containers each holding one explant. The mesh panel and aquapurse were approximately 15 x 20 cm in height and length but differed in width, with the aquapurse being wider (~10 cm) than the mesh panel (~5 cm). Mesh size was larger for the mesh panel than for the aquapurse being 4 cm and 1.5 cm, respectively. Strand thickness was about 3 mm for both methods. The mesh bags were made from 1

mm thick nylon strands with mesh size of 3 cm and each contained one explant. For both Coscinoderma sp. and R. odorabile, we deployed 2 x 2 m plastic grids to test the 11 farming techniques; each suspended upright 3 m off the substrate at a mean depth of 12 m. Three plastic grids were used per sponge species, each holding 5 replicates of each treatment in a randomised block pattern. For Coscinoderma sp. and R. odorabile, each farming treatment had 15 explant replicates, thus 165 explants were experimentally farmed per species. To obtain sufficient explants, approximately 20 sponges of each species were partially collected, whereby 1/3 of each sponge was left behind to regrow. At all times, sponges were kept underwater. The sponges were experimentally farmed for 1 year.

Importance of a nursery period

In this experiment, 3 farming treatments or methods were trialled: threaded line, mesh and mesh-line. For the threaded line method, explants were threaded onto 4 mm nylon line for the duration of the experiment. In the mesh method, explants were farmed in mesh panels for the entire length of the experiment, representing a threaded line control. For the mesh-line method, explants were placed in mesh for 4 months allowing sufficient time to fully heal (nursery period), and were then threaded onto nylon line. For Coscinoderma sp. and R. odorabile, each method had 5 replicate lines or mesh strips placed in a randomised block pattern. The ropes or mesh strips were 1 m apart, in an upright position at a depth of approximately 12 m. Each method replicate had 5 explants approximately 15 cm apart, thus 25 explants were experimentally farmed per method per bath sponge species. To obtain sufficient explants approximately five sponges of Coscinoderma sp. and R. odorabile were partially collected, leaving a portion of the sponge on the seafloor to regrow. The explants were experimentally farmed for 9 months.

Data analysis
For both experiments, One Way ANOVAs were used to test whether growth or survival varied significantly among the

299

farming treatments for Coscinoderma sp. and R. odorabile. Growth data was log transformed and survival data was arcsine transformed to meet the assumptions of ANOVA (Zar 1999). Statistical analysis of survival compared explant survival among grids for experiment 1 and among the upright lines and mesh strips for experiment 2.

Results Developing a good farming method

Coscinoderma sp. survival varied significantly among the farming treatments (Fdf(10,22) = 16.29, P < 0.0001). Survival was highest for explants farmed in mesh, with 90% of the explants cultured in mesh bags and panels alive after 1 year (Fig. 2A). Final survival of Coscinoderma sp. farmed on threaded line was considerably lower, with less than 50% of the explants surviving regardless of line material. Final explant survival in the remaining methods was very poor, with only 7% surviving on the spike methods and none surviving in the cut-explant methods (Fig. 2A). Monitoring 2 months after the experiment started showed that most explants farmed in the spike and cut-explant methods had died. In contrast to survival, final size of Coscinoderma sp. explants did not vary significantly among the farming treatments (Fdf(7,50) = 1.54, P = 0.17). On average, Coscinoderma sp. explants grew from an initial size of 83 cm3 to a final size of 224 cm3. Therefore, explants almost tripled in size, on average, within 1 year. Although not significantly different among treatments, explant growth was highest overall in mesh panels and polypropylene spikes (Fig. 2B), although the later method had few surviving replicates. Explant growth was lowest overall in mesh bags (Fig. 2B). Many explants farmed on threaded line or on spikes did not attach to the farming material, instead forming a donut shape. In contrast, explants farmed in mesh were generally round in shape. Similar to Coscinoderma sp., final survival of R. odorabile explants differed significantly among the farming treatments (Fdf(10,22) = 3.06, P = 0.014). Survival was highest overall for explants farmed in mesh bags and panels and on nylon and polypropylene threaded lines, with approximately 90% surviving after 1 year (Fig. 3A). Unlike the other mesh methods, survival was relatively poor for explants in mesh aquapurses. Final survival of R. odorabile in the spike and cut-explant methods ranged from 30-70%. Overall, 63% of R. odorabile survived after 1 year. Growth of R. odorabile explants also varied significantly among the farming treatments (Fdf(10,94) = 2.49, P = 0.011). After 1 year of culture explants had more than doubled in size, on average, in the polyethylene threaded-line, polypropylene cutexplant and three mesh treatments (Fig. 3B). Explant growth was lowest in the polyethylene cut-explant and PVC spike methods. Similar to Coscinoderma sp., many explants of R. odorabile cultured on spikes or with line through the explants formed a donut shape by the end of the experiment. Growth rates of R. odorabile were lower, overall, than recorded for Coscinoderma sp. Throughout the farming experiment, biofouling on the farming treatments and structures was generally minimal. From the numerous scrape marks on the farming structure

Fig. 2: Final survival and growth of Coscinoderma sp. explants experimentally farmed in the different methods. Note that all explants died in the cut-explant methods. Dashed line in (B) represents initial explant size. Error bars represent variation between plastic grids for survival and between explants for final size.

Fig. 3: Final survival and growth of R. odorabile explants experimentally farmed in the different methods. Dashed line in (B) represents initial explant size. Error bars represent variation between plastic grids for survival and between explants for final size.

300

it appeared that fish, probably from the family Siganidae, were removing most of the biofouling. The few biofouling organisms that escaped predation were ascidian and hydroid species, which were ephemeral and disappeared after a few months. The one exception where biofouling was a problem was inside the aquapurse mesh containers, which at times become filled with bivalve species.

Importance of a nursery period

For Coscinoderma sp., final survival varied significantly among the three treatments (Fdf(2,12) = 37.59, P < 0.0001), highest for explants farmed continuously in mesh with 88% surviving (Fig. 4A). In comparison, final survival of explants farmed in mesh for 4 months and then threaded with line (nursery treatment) was 52%. None of the sponges threaded with line at the start of the experiment were alive after 9 months. Final growth did not differ greatly between the mesh and nursery explants (Fdf(1,33) = 0.05, P = 0.818), with most explants almost tripling in size over 9 months (Fig. 4B). Similar to Coscinoderma sp., final survival of R. odorabile also varied among the three treatments (Fdf(2,12) = 4.83, P = 0.029), highest overall for explants farmed in mesh where all 25 explants survived (Fig. 5A). Final survival of explants in the nursery treatment was also high, being 92%. For explants farmed on threaded line, survival was 76% after 9 months. Growth also varied significantly among the treatments (Fdf(2,66) = 3.58, P = 0.033), with final size greatest for explants farmed continuously in mesh or on threaded line (Fig. 5B). These explants more than doubled in size on average over 9 months.

Discussion
Final survival of Coscinoderma sp. and R. odorabile varied greatly among the farming methods, thus clearly indicating the importance of selecting an appropriate method to guarantee commercial success. Coscinoderma sp. had highest survival when farmed in mesh, particularly mesh bags and panels with 90% of the explants alive after 1 year. Final survival of 90% is considered essential for commercial bath sponge culture (Verdenal and Vacelet 1990). In contrast, farming methods that involved additional tissue damage, from threading for example, had poor survival. Survival of Coscinoderma sp. on threaded line can be improved if a nursery period (in mesh) is used, however, final survival will still be lower than farming explants continuously in mesh. Poor survival on threaded line suggests that explants of Coscinoderma sp. either reject the threaded line as a suitable substrate for attachment and dislodge themselves or were torn off during periods of high water flow. For R. odorabile, final survival was also high (90%) when farmed in mesh bags or panels. Differing from Coscinoderma sp., survival of R. odorabile was also high when farmed on threaded lines. Interspecific differences in explant survival are common among sponge farming studies (Verdenal and Vacelet 1990, Duckworth and Battershill 2003a) and may result from variation in regenerative ability, susceptibility to infection after cutting and general hardiness. Although survival results on threaded line were promising, many R. odorabile did not attach to the line, instead forming a

Fig. 4: Final survival and growth of Coscinoderma sp. explants farmed in the nursery experiment. Note that all explants died in the line method. Dashed line in (B) represents initial explant size. Error bars represent variation between ropes or mesh strips for survival and between explants for final size.

donut shape, which is not marketable. In contrast, R. odorabile farmed in mesh retained a good shape. Explants of R. odorabile that did not attach to the farming support material in the threaded line, cut-explant and spike methods, generally had poor growth. Verdenal and Vacelet (1990) also reported poor growth of Mediterranean bath sponges that did not attach to the threaded line. In contrast, most Coscinoderma sp. farmed using the cut-explant, threaded line and spike methods did not attach to the farming material but still had good growth. These results further highlight that farming responses can differ among species. For R. odorabile, final explant size varied among the farming

301

Fig. 5: Final survival and growth of R. odorabile explants farmed in the nursery experiment. Dashed line in (B) represents initial explant size. Error bars represent variation between ropes or mesh strips for survival and between explants for final size.

methods. For example, final explant size was high when R. odorabile was farmed in mesh or on polyethylene threadedline, but low when grown in the polyethylene cut-explant and PVC spike methods. Comparing the great difference in growth of R. odorabile farmed using polyethylene line between the threaded line and cut-explant methods suggests that the culture method is more important than the material used. Coscinoderma sp. and R. odorabile explants farmed in mesh in both experiments had some of the highest growth rates recorded, with explants doubling and tripling in size on average in one year. Previous sponge farming studies have

found that it generally takes 2 years or more for bath sponges to double or triple in size (e.g. Moore 1910, Crawshay 1939, Verdenal and Vacelet 1990, Kelly et al. 2004). In the present study, mesh panels, in particular, promoted consistent high growth. Similar high growth of sponges farmed in mesh and on line was not expected. Duckworth and Battershill (2003b) found that after 9 months sponges grown on threaded line were at least twice as large as sponges cultured in mesh, possibly because the mesh strands reduce water flow and thus food availability to the farmed sponges. Mesh panels in the present study consisted of thin strands and large mesh, which would likely have offered minimal resistance to water flow. Duckworth and Battershill (2003b) also suggested that sponge growth is further reduced if explants are unable to grow out through the mesh strands, restricting their final size to the size of the mesh pocket. This likely explains the relatively poor growth of Coscinoderma sp. farmed in mesh bags. Mesh panels, in contrast, contain but do not smother explants, allowing for better growth. Lastly, the relatively large surface area of mesh structures can result in high levels of fouling by other filter-feeding organisms, which can reduce food abundance (Duckworth and Battershill 2003b) Apart from inside mesh aquapurses, biofouling was not a problem in the present study probably due to the feeding activities of Siganidae fish. These various factors help explain the high growth of Coscinoderma sp. and R. odorabile farmed in mesh panels. For bath sponge culture to be commercially viable, explants should at least double in size each year (Crawshay 1939, Verdenal and Vacelet 1990). After 1 year in the farming method experiment, R. odorabile more than doubled in size, on average, in several farming methods, greatest for threaded polyethylene line and mesh panels where explants were > 270% of their initial size. Growth of Coscinoderma sp. was even better, with final percent size of explants farmed in mesh panels and on polypropylene spikes being 350% and 460%, respectively. These high growth rates also occurred in the nursery experiment and show the promising potential of farming Coscinoderma sp. and R. odorabile in tropical Australian waters. Because bath sponge culture does not require large community infrastructure (MacMillan 1996) and processing and storage requirements are minimal, it is likely that commercial sponge culture will be a profitable industry for some remote coastal indigenous communities in Australia. This study indicates that both Coscinoderma sp. and R. odorabile should be commercially cultured in mesh panels and that a nursery period is not warranted.

Acknowledgements
We would like to thank Sarah Lowe, Stephan Whalan, Raymond Bannister and the skippers and crews of the RV Cape Ferguson and RV Lady Basten for help and support during the field work. This project was part of the sponge aquaculture program of AIMS@JCU, which receives funding and in-kind support from the Australian Institute of Marine Science, James Cook University Research Advancement Program, Great Barrier Reef Research Foundation, Coolgaree Aboriginal Corporation, Queensland Department of State Development Innovation and Trade, Queensland Department

302

of Primary Industries and Fisheries, and the Commonwealth Department of Transport and Regional Services.

References
Corriero G, Longo C, Mercurio M, Marzano CN, Lembo G, Spedicato MT (2004) Rearing performance of Spongia officinalis on suspended ropes off the Southern Italian Coast (Central Mediterranean Sea). Aquaculture 238: 195-205 Crawshay LR (1939) Studies in the market sponges I. Growth from the planted cutting. J Mar Biol Assoc UK 23: 553-574 Duckworth AR, Battershill CN, Bergquist PR (1997) Influence of explant procedures and environmental factors on culture success of three sponges. Aquaculture 156: 251-267 Duckworth AR, Battershill CN (2003a) Developing farming structures for production of biologically active sponge metabolites. Aquaculture 217: 139-156 Duckworth AR, Battershill CN (2003b) Sponge aquaculture for the production of biologically active metabolites: the influence of farming protocols and environment. Aquaculture 221: 311-329 Kelly M, Handley S, Page M, Butterfield P, Hartill B, Kelly S (2004) Aquaculture trials of the New Zealand bath-sponge Spongia (Heterofibria) manipulatus using lanterns. New Zeal J Mar Fresh 38: 231-241

MacMillan SM (1996) Starting a successful commercial sponge aquaculture farm. Center for Tropical and Subtropical Aquaculture, University of Hawaii, Publication no. 120 Moore HF (1910) A practical method of sponge culture. Bull US Bur Fish 28(1908, Pt. I): 545-585 Pronzato R (1999) Sponge-fishing, disease and farming in the Mediterranean Sea. Aquat Conserv 9: 485-493 Pronzato R, Bavestrello G, Cerrano C, Magnino G, Manconi R, Pantelis J, Sara A, Sidri M (1999) Sponge farming in the Mediterranean Sea: new perspectives. Memoir Queensl Mus 44: 485-491 Stevely JM, Sweat DE (1994) A preliminary evaluation of the commercial sponge resources of Belize with reference to the location of the Turneffe Islands sponge farm. Atoll Res Bull 424: 1-21 van Treeck P, Eisinger M, Mller J, Paster M, Schuhmacher H (2003) Mariculture trials with Mediterranean sponge species: the exploitation of an old natural resource with sustainable and novel methods. Aquaculture 218: 439-455 Verdenal B, Vacelet J (1990) Sponge culture on vertical ropes in the Northwestern Mediterranean Sea. In: Rtzler K (ed). New perspectives in sponge biology. Smithsonian Institution Press, Washington DC. pp. 416-424 Zar JH (1999) Biostatistical analysis. Prentice-Hall, New Jersey

Porifera research: Biodiversity, innovation and sustainaBility - 2007

303

Sponges as natural composites: from biomimetic potential to development of new biomaterials

Hermann Ehrlich(*), Hartmut Worch
Max Bergmann Center of Biomaterials, Institute of Materials Science. University of Technology, Budapester Str. 27, D01069 Dresden, Germany. hermann.ehrlich@mailbox.tu-dresden.de Abstract: Investigations of the compositions and the microstructures of the sponge skeletons as examples for natural structural biomaterials are of fundamental scientific relevance. Recently, we show that some demosponges (Verongula gigantea, Aplysina sp.) and glass sponges (Farrea occa) possess chitin as a component of their skeletons. The main practical approach we used for chitin isolation was based on alkali treatment of corresponding exoskeletal sponge material with the aim of obtaining alkaliresistant compounds for detailed analysis. Here, we present a detailed study of the structural and physico-chemical properties of skeletal fragments of the glass sponge Euplectella aspergillum. The structural similarity of chitin derived from this sponge to invertebrate alpha-chitin has been confirmed by us unambiguously using physico-chemical and biochemical methods. This is the first report of a silica-chitin composite biomaterial found in Euplectella species. Finally, the present work includes a discussion relating to strategies for the practical application of sponges as biomaterials. A comprehensive understanding of collagen- and chitin-based sponge skeletons with respect to chemical composition and structure may prove to be a novel model for biomimetic synthesis of three-dimensional collagen- and chitin-based composites with specific mechanical, optical and bioactive properties for applications in different modern technologies, including materials science and biomedicine. Keywords: biocomposites, biosilicification, chitin, collagen, sponges

Introduction
Biological structures are a constant source of inspiration for finding solutions to a variety of technical challenges in bionics, architecture, aerodynamics and mechanical engineering, as well as materials science (reviewed in Fratzl 2007). Sponges are fascinating research objects because of the hierarchical organisation of their fibrous skeletons (Demospongiae) and mineralized spicules containing amorphous silica (Demospongiae and Hexactinellida) or calcium carbonate (Calcarea). Thus, skeletal formations of sponges are examples of natural rigid glass-based or calcium carbonatebased composites. However, sponges are presently gaining increased scientific attention because of their secondary metabolites and biotechnological applications and not as templates for biomaterials research. The biotechnological potential of marine sponges as a goldmine to chemists and pharmacologists is well known (Thakur and Mller 2004). Unique and innovative structural leads have been discovered with cytotoxic, antifouling, antitumoral, antibiotic, antiviral or cytoprotective, enzyme-inhibitory, anti-inflammatory and anti-Alzheimer activities (Faulkner 2001). In contrast to the examples listed above, only five main aspects relating to sponges as biomaterials are recently described as follows: - Hexactinellid spicules as natural glass-based composites with specific mechanical properties (Mayer 2005, Walter et al. 2007)

- Skeleton of Euplectella sp. (Hexactinellida) as a hierarchical natural structural material of remarkable design (Aizenberg et al. 2005, Weaver et al. 2007) - Basal spicules of Hexactinellida as biological glass fibers with specific optical properties (Cattaneo-Vietti et al. 1996, Aizenberg et al. 2004, Mller et al. 2006) - Silicatein-based biocatalitic formation of nanocomposite materials (reviewed in Schrder et al. 2007) - Biomimetically inspired hybrid materials based on silicified sponge collagen (Ehrlich and Worch 2007, Heinemann et al. 2007a, 2007b) The key features of natural structural composites include durable interfaces between hard and soft materials and excellent bonding, a desirable combination of toughness and strength, good fatigue resistance and resiliency, and a dramatic influence of the presence of water, which can have notable effects on mechanical and material properties (Mayer and Sarikaya 2002). In case of hexactinellid spicules, it is reported that they are highly flexible and tough, possibly because of their layered structure and the hydrated nature of the silica (Sarikaya et al. 2001). In Euplectella species, the skeleton comprises an elaborate cylindrical lattice-like structure with at least six hierarchical levels spanning the length scale from nanometers to centimeters. The basic building blocks are laminated spicules that consist of a central proteinaceous axial filament surrounded by alternating concentric domains

304

of consolidated silica nanoparticles and organic interlayers (Weaver et al. 2007). There is no doubt that some of the basal spicules of hexactinellid sponges or fibrous skeletons of demosponges are remarkable for their size, durability, flexibility and their optical properties, a set of features rendering them of interest as biocomposites with high biomimetic potential. Of course, the materials science aspects of sponges can be studied by model systems, and utilized for biomimetic engineering. However, we cannot mimic nature with a view to designing novel biomaterials without knowledge of the nature and origin of the organic matrices of corresponding natural biocomposites which are present in sponges. Therefore, the biggest shortcoming common to all publications relating to mechanical, structural and optical properties of sponge skeletal formations is a lack of real information regarding the chemical nature of corresponding organic matrices. The finding of collagen within basal spicules of two representatives of Hexactinellida, Hyalonema sieboldi and Monorhaphis chuni (Ehrlich et al. 2005a), as well as the occurrence of chitin within the framework skeleton of Farrea occa (Ehrlich et al. 2007a), as revealed by gentle desilicification in alkali, stimulated further attempts to search for materials of organic nature in other glass sponges. Consequently, the objective of the current study was to test the hypothesis that chitin is an essential component of the silica spicules of Euplectella aspergillum as well, and if so, to unravel its involvement in the mechanical behaviour of these spicules, which was recently well investigated (Walter et al. 2007, Weaver et al. 2007). We decided to re-examine also the results of some previously reported studies concerning the presence of polysaccharides within silica-containing spicules of this sponge. For example, Travis et al. (1967) reported the presence of parallel-oriented cellulose-like filaments with an average width of 1.9 nm observed in organic matrix material after HF-based desilicification of the spicules of hexactinellid Euplectella sp. These matrices also contained considerable amounts of hexosamine. Finally, the present work includes a discussion relating to strategies for the practical application of sponges as biomaterials.

Standard Methods 4500-Si E using Silicat (Kieselsure)-Test (Merck). Skeletons of E. aspergillum were treated according to the following procedure. Sponge material of E. aspergillum was stored for several days in fresh sea-water. The sponge was dried afterwards for 4 days at 45C. Finally the sponge skeleton was cleaned in 10% H2O2 and dried again at 45C. Tissue-free dried sponge material was washed three times in distilled water, cut into 3 cm long pieces and placed in a solution containing purified Clostridium histolyticum collagenase (Sigma) to digest any possible collagen contamination of exogenous nature. After incubation for 24h at 15C, the pieces of glass sponge skeleton were again washed three times in distilled water, dried and placed in a 15 ml vessel containing chitinase solution (as described below) to digest any possible exogenous chitin contaminations. After incubation for 48h at 25C, fragments of skeleton were again washed, dried and placed in 10 ml plastic vessel containing 8 ml 2.5 M NaOH solution. The vessel was covered and placed under thermostatic conditions at 37C without shaking. The effectiveness of the alkali etching was also monitored using optical and scanning electron microscopy (SEM) at different locations along the length of the spicular material and within the cross-sectional area. The colourless alkaliinsoluble material obtained after alkali treatment of the glass sponge samples was washed with distilled water five times and finally dialysed against deionized water on Roth (Germany) membranes with a MWCO of 14 kDa. Dialysis was performed for 5 days at 4C. The dialyzed material was dried at room temperature and used for staining and analytical investigations.

Staining and detection of chitin

We used Calcofluor White (Fluorescent Brightener M2R, Sigma) which shows enhanced fluorescence when it binds to chitin (Pringle 1991). The pieces of natural glass sponge skeleton samples and those which were subjected to alkali treatment were placed in 0.1 M Tris-HCl, pH 8.5 for 30 min. After this procedure they were stained using 0.1% Calcofluor White solution for 30 min in darkness, rinsed three times with distilled water, dried at room temperature and finally investigated using Wide Field Fluorescence as well as confocal Laser Scanning Microscopy (LSM) (Zeiss LSM 510 META). For LSM, the fluorescence of Calcofluor White was excited by a NIR pulsed laser (770 nm) using two-photonexcitation. This corresponds to a single-photon-excitation at approximately 380 nm that yields a fluorescence emission at 440 nm.

Materials and methods Chemical etching of glass sponge skeletons

The object of our study was Euplectella aspergillum (Hexactinellida: Porifera), collected in 2005 in the Pacific Ocean (Philippines). Additionally, the following representatives of Hexactinellida sponges: Hyalonema sieboldi, Monorhaphis chuni, Aphrocallistes vastus and Farrea occa were investigated as comparisons to obtain preliminary results relating to differences in the resistance of these species to alkali treatment. Skeletons of these sponges were treated similarly and according to the procedure that follows as described below for E. aspergillum. The duration of the desilicification period was the time required to dissolve 50% of the silica. Silicon concentrations were determined by the silicomolybdate method (Iler 1979) according to US

FTIR spectroscopy
IR spectra were recorded with a Perkin Elmer FTIR Spectrometer Spectrum 2000, equipped with an AutoImage Microscope using the FT-IRRAS technique (Fourier Transform Infrared Reflection Absorption Spectroscopy).

305

X-ray analysis
X-ray diffraction measurements were performed by a STOE-STUDIP-MP diffractometer with Ge-monochromator at Cuk1 wave length.

Preparation of -Chitin
Alpha-Chitin was prepared from a commercially available crab shell chitin (Fluka). The material was purified with aqueous 1 M HCl for 2 h at 25C and then refluxed in 2 M NaOH for 48h at 25C. The resulting -chitin was washed in deionized water by several centrifugations until neutrality was reached. The whole procedure was repeated twice.

Transmission electron microscopy (TEM)

Conventional transmission electron microscopy was performed with a Philips CM200 FEG\ST Lorentz electron microscope at an acceleration voltage of 200 kV. For electron microscopy, a drop of the water suspension containing the sample was placed onto the electron microscopy grid. After one minute, the excess was removed using blotting paper and thereafter dried in air. The electron microscopy grids (Plano, Germany) were covered with a holey carbon film.

Preparation of colloidal chitin

Ten grams of -chitin (Fluka) was mixed with 500 ml of 85% phosphoric acid and stirred for 24h at 4C. The suspension was poured into 5 litres of distilled water (DW) and centrifuged (15000 x g for 15 min). The resulting precipitate was washed with DW until the pH reached 5.0 and then neutralized by addition of 6 N NaOH. The suspension was centrifuged (15000 x g for 15 min) and washed with 3 litres of DW for desalting. The resulting precipitate was suspended in DW and dialyzed. The chitin content in the suspension was determined by drying a sample.

Fluorescence and confocal laser scanning microscopy

For both methods an upright light microscope Axioscope 2 FS mot was used. For LSM, it was equipped with a Zeiss LSM 510 META scanning head. Fluorescence was excited either by a mercury arc lamp HBO50 or, in the case of LSM, by Ar+ ion- (488 nm), He/Ne- (546 nm) and Titanium/Sapphire-NIR (770 nm) lasers. The spectra were recorded by the META spectrograph inside the scanning head.

Silicification of colloidal chitin

In the first step, 0.93 g of colloidal chitin previously suspended in 34.6 g of methanol was added to 51.8 ml of deionized water. The suspension pH was then raised above 10 with the addition of 100 L of 1 N NaOH solution. Finally, 169 L of tetramethylorthosilicate (TMOS, 99 wt%) were added and the solution stirred at room temperature. After 1 h the suspension was filtered and the recovered precipitate rinsed with water, then with methanol and finally air-dried.

SEM analysis
The samples were fixed in a sample holder and covered with carbon, or with a gold layer for 1 min using an Edwards S150B sputter coater. The samples were then placed in an ESEM XL 30 Philips or LEO DSM 982 Gemini scanning electron microscope.

Results and discussion

Species of the class Hexactinellida are found in all oceans, and live at depths from 30 to 6235 m (Reiswig 2004). The glass sponges of the families Hyalonematidae, Monorhaptidae, Pheronematidae and, to a certain extent, Euplectellidae inhabit loose muddy substrates. One of the strategies of survival under such conditions is the formation of root like structures that prevent the body of the animal from sinking into the ground (Tabachnik 1991). In contrast to demosponges, where the cytoskeleton is organised on the basis of individual cells, in hexactinellids it provides a supporting framework and transport pathways within vast, multinucleate tissue masses (Leys 1995). It was suggested that hexactinellid sponges may have evolved from cellular sponges and that syncytial tissues are not a primitive trait of the Metazoa (Leys 2002). Due to their preferred deep-sea habitat, the Hexactinnellida have been poorly investigated with respect to their general biology (Reiswig and Mehl 1991) and the nature of organic components which build their skeletal structures. It was generally accepted that their skeletons are composed of amorphous hydrated silica (de La Rocha 2003) deposited around a proteinaceous axial filament (Uriz et al. 2003, Weaver and Morse 2003, Weaver et al. 2007). The resolution of the question: how to isolate the organic matrix from the silica containing sponge skeleton, is absolutely dependent on the desilicification method. Up

Chitinase digestion and test

Dried 20 mg samples of sponge skeleton, previously pulverized to a fine powder in an agat mortar, were suspended in 400 l of 0.2 M phosphate buffer at pH 6.5. Positive control was prepared by solubilizing 0.3% colloidal chitin in the same buffer. Equal amounts of 1 mg/ml of three chitinases (EC 3.2.1.14 and EC 3.2.1.30): N-acetyl--glucosaminidase from Trichoderma viride, Sigma, No. C-8241, and two Poly (1,4-[2-acetamido-2-deoxy-D-glucoside])glycanohydrolases from Serratia marcescens, Sigma, No. C-7809 and Streptomyces griseus, Sigma, No. C-6137 respectively, were suspended in 100 mM sodium phosphate buffer at pH 6.0. Digestion was started by mixing 400 l of the samples and 400 l of the chitinase-mix. Incubation was performed at 37C and stopped after 114 h by adding 400 l of 1% NaOH, followed by boiling for 5 min. The effectiveness of the enzymatic degradation was monitored using optical microscopy (Zeis, Axiovert). The Morgan-Elson assay was used to quantify the N-acetylglucosamie released after chitinase treatment as described previously (Boden et al. 1985). The sample which contains chitinase solution without substrate was used as a control.

306

to now, the common technique for the demineralization of glass sponge spicules and skeletons was based on hydrogen fluoride solutions (Uriz et al. 2003). However, the HF-based silica dissolution procedure was criticised as being a rather aggressive chemical procedure which could drastically change the structure of the organic matrix (Travis et al. 1967, Croce et al. 2004, Ehrlich et al. 2005b, 2005c). To overcome this obstacle, we developed novel, slow-etching methods which use a solution of 2.5 M NaOH at 37C and take 14 days or more (Ehrlich et al. 2005a, 2005b, 2005c, 2006). This type of biosilica dissolution is practically impossible in well-crystallized silica such as quartz, but is much stronger in amorphous silica because of the increased surface area and disordered, open lattice structure. Thus the main practical approach we used was based on alkali treatment of the corresponding exoskeletal glass sponge material with the aim of obtaining alkali-resistant material for detailed analyses with respect to its identification.

Table 1: Comparison of desilicification duration relating to different representatives of Hexactinellida. Species of Hexactinellida Monorhaphis chuni Hyalonema sieboldi Euplectella aspergillum Aphrocallistes vastus Farrea occa Duration of desilicification 3-10 days 14-30 days 4-6 months 6-8 months 6-12 months

Chitin within skeletal formations of E. aspergillum

The finding of collagen within basal spicules of H. sieboldi and M. chuni (Ehrlich et al. 2005a, 2005b, 2005c, 2006) after their desilicification using alkali treatment during 7-30 days, stimulated our attempts to find materials of organic nature in other species of glass sponges. Thus, we started similar demineralization experiments with skeletons of E. aspergillum (Fig. 1), A. vastus and F. occa. However, we have not observed any visible signs of demineralization of these materials using optical microscopy and SEM after the same time elapsed and at the similar experimental conditions as in the study on H. sieboldi and M. chuni. On the contrary, fragments of skeletons of E. aspergillum, A. vastus and F. occa show high resistance to alkali treatment even after some months of

demineralization (Table 1). This phenomenon led us to the assumption that siliceous skeletons of investigated sponges possess a biomaterial which protects amorphous silica from dissolution in alkali, and is highly resistant to alkali digestion. It is well known that chitin in alkali is stable with respect to degradation (Einbu et al. 2004). Correspondingly, in our experiments, chitin was the first candidate for a biomaterial with this property. Scanning electron microscopy (SEM) investigations of the natural tissue-free skeleton of E. aspergillum (Fig. 1A) confirmed their framework structure, typical for Euplectella species (Weaver et al. 2007). Alkali treatment of the same skeleton samples led to loss of integrity. Figs. 2A and 2B show a typical view of a partially demineralised glass sponge skeleton obtained after demineralization. Desilicification of E. aspergillum skeletons was definitively effective within this structural formation: alkali treatment leads to loss of the interior layers of the spicule, causing the spicules to become twisted (Fig. 2B). In contrast to alkali-resistant spicules, alkali-etched junction areas (Fig. 2C) show that the silica-based cement-like lattice

Fig. 1: A. Micrograph of E. aspergillum siliceous tissue-free skeleton. B. SEM image of spicules which build the square-grid lattice of vertical and horizontal struts typical for the skeleton of this hexactinellid.

307

Fig. 2: SEM images: A. E. aspergillum skeleton after alkali treatment (6 months). B. Desilicification leads to loss of the interior layers of the spicule, causing it to become twisted. C, D. Alkali-etched junctions show that the silica-based cement-like lattice is strongly corroded in contrast to alkali-resistant spicules.

is corroded (Fig. 2D). Corrosion and subsequent dissolution of silica-containing cement might be the main reason for the collapse of the mechanically stable and hierarchically constructed skeleton of Euplectella, for which the intact sponge is well-known. To test our hypothesis that alkali-insoluble residues of E. aspergillum skeletons are of a chitinous nature, we stained them with Calcoflour White and investigated them using fluorescence microscopy. Wide fluorescence microscopy images of natural glass sponge skeleton treated with alkali and stained with Calcofluor White revealed that chitin is distributed on and within the outermost layer of these skeletal structures (Fig. 3A). The characteristic blue fluorescence of stained chitin within partially desilicified glass sponge skeletons (Fig. 3B) led us to the assumption that it might originally be involved in the phenomenon of biosilification and act as a center of silica nanoparticle nucleation. Chitin, or poly [(14)-2-acetamido-2-deoxy-Dglucopyranose], is crystalline in its native state (Minke and

Blackwell 1969) in contrast to amorphous silicon dioxide of biogenic origin. This property determined corresponding structural investigations. The overview TEM micrograph of Fig. 4A shows a chitin residue after demineralization of the skeleton sample. In the Fourier Transformation of the high-resolution micrographs the spacings of 4.85 , 4.42 , 2.95 , 2.54 could be detected corresponding to (100), (200), (105) and (106) (Fig. 4B) reflections proving the orthorhombic structure typical for -chitin as described in detail by Carlstrm (1957). The results of the additional physico-chemical analysis performed using FTIR, Raman and XRD (data not shown) were similar to those from F. occa (Ehrlich et al. 2007a) and from keratose sponges (Ehrlich et al. 2007b). Therefore, all analysis listed above indicate without ambiguity that the material isolated from E. aspergillum skeleton is -chitin, further confirming our earlier observations that chitins in marine sponges appear to be consistently in the alpha modification.

308

Fig. 3: A. Wide field fluorescence microscopy image exhibiting strong evidence of the presence of chitin after Calcofluor White staining of an alkali-resistant twisted spicule which lost its content during desilicification. B. 3D reconstruction of a confocal LSM image stack (70 single images) showing Calcofluor White fluorescence in the outermost region of the twisted spicule.

The analysis of chitin within glass sponge skeletal material using enzymatic methods is quite difficult, which is underlined by the fact that chitin is bound to silica and that the corresponding enzyme must gain sufficient access to attack the surface of the substrate. Therefore, corresponding skeleton samples of E. aspergillum were mechanically disrupted before enzymatic treatment started. We used a chitinase digestion test for chitin identification on and within the investigated hexactinellid millimeter-large skeleton fragments. It is known that chitinolytic systems in nature comprise of an endochitinase, chitobiase and an exochitinase, whose actions may be synergistic and consecutive in the degradation of chitin to free N-acetyl glucosamine (Gohel et al. 2006). To quantify chitin in our samples, we measured the amount of Nacethyl-glucosamine released by chitinases using a MorganElson colorimetric assay (Boden et al. 1985), which is the most reliable method for the identification of alkali-insoluble chitin because of its specificity (Bulawa 1993). We detected 25.2 1.5 g N-acetyl-glucosamine per mg of alkali-resistant skeleton residues of E .aspergillum. We believe that chitin is acting as an organic template for silica mineralization in Euplectella species in a highly similar fashion as in Farrea occa (Ehrlich et al. 2007a). The finding of silica-chitin natural composites as the main component of the E. aspergillum skeleton is in good agreement with results of in vitro experiments on silicification of a -chitincontaining cuttlebone-derived organic matrix (Ogasawara et al. 2000). The cuttlebone of Sepia officinalis is a highly organized internal shell structure constructed from aragonite (CaCO3) in association with a -chitin organic framework. Corresponding pieces of the -chitin mineral-free cuttlebone matrix were soaked in sodium silicate solution (pH 11.5) at room temperature, then removed and immersed in an ethanol/ water mixture. The authors suggest that silicate ions and silica oligomers preferentially interact with glycopyranose rings

exposed at the -chitin surface, presumably by polar and Hbonding interactions. There are also no doubts that investigations into the organic matrix estimation in demosponges and hexactinellids skeletons are of great scientific importance not only for materials science, but also for evolution research and systematics of sponges. Biogenic silica exhibit diversity in structure, density and composition, and can exist in several structural forms. A diverse range of siliceous structures including internal and external skeletons, scales, spines, bristles, cell walls, cyst walls and loricae were described for flagellated protists (Anderson 1994, Preisig 1994). The presence of chitin as a structural component of these Protozoa was also reported (Herth 1980), however never with respect to chitin-silica composite materials. The results of our study suggest that silica-chitin biocomposites might be identified also in choanoflagellatelike protists as ancestral organisms which are architecturally close to sponge larvae (Maldonado 2004). Our results suggest that the chitin system has never been lost in the lineage leading from protostomes to deuterostomes. The evolution, localization and functions of chitin within sponge structural formations could be re-examined. The question of chitin synthesis among both keratose and glass sponges should gain importance as a result of our findings. We show that chitin is present as a structural component in skeletons of both poriferan classes, Hexactinellida and Demospongia. This most intriguing finding led us to a better understanding of sponge evolution and gives a new impulse to discussions about the mono- or diphyletic origin of the Metazoa. The remarkable differences between Hexactinellida and Demospongia/Calcarea lead Bergquist (1985) and Borchiellini et al. (2001) to seriously consider the Hexactinellida as a distinct phylum, independently evolved from other Porifera. However, other authors (Reiswig

309

Fig. 4: A. TEM micrograph of demineralized E. aspergillum skeleton sample. B. Fast Fourier Transformation of Fig. A displaying the orthorhombic crystal structure of alpha-chitin with denoted (100) and (105) reflections.

and Mehl 1991) argue that the Metazoa constitute a well established monophylum. Experiments on chitin identification in Calcarea sponges are currently in progress.

Strategies for practical application of collagen- and chitin-based biocomposites of poriferan origin
Collagen and chitin are most investigated materials of biological origin with wide fields of applications in biomedicine because of their unique multifunctional engineering mechanical properties and biocompatibility (reviewed in Stenzel et al. 1974, Rinaudo 2006). In addition to the well-known example of mammalian bone, which consists of collagen and hydroxyapatite, a natural hybrid material based on silicified collagen which is evolutionary much older was recently found by us within the highly flexible basal spicules of some glass sponges. As collagen also serves as a template for calcium phosphate and carbonate deposition in bone, this suggests that the evolution of silica and bone skeletons share a common origin with respect to collagen as a unified template for biomineralization (Ehrlich and Worch 2007). Understanding the composition, hierarchical structure and resulting properties of glass sponge spicules gives impetus for the development of equivalents designed in vitro. Only recently, we showed for the first time that the silica skeletons of hexactinellids represent examples of biological materials in which a collagenous (Ehrlich et al. 2005a, Ehrlich et al. 2006, Ehrlich and Worch 2007) or chitinous (Ehrlich et al. 2007a) organic matrix serves as a scaffold for the deposition of a reinforcing mineral phase in the form of silica or calcium carbonate (Ehrlich et al. 2007b). These findings allow us to discard different speculations about materials, which have previously been defined as

organic structures (layers, filaments, surfaces) of unknown nature, and open the way for detailed studies on sponge skeletons and spicules as collagen- and/or chitin-based biocomposites. Also, the definition of spongin as the main organic component of sponges must now be re-examined. Crookewitt first pointed out in 1843 that the endoskeleton of the common bath sponge is derived from the dermal layer and called it spongin (Block and Bolling 1939). Till now, spongin (named also fibrous skeleton, pseudokeratin, neurokeratin, horny protein, collagen-like protein, scleroprotein) (reviewed in Ehrlich et al. 2003) has no clear chemical definition. Contrary to the postulate that silicateins, as the major biosilica-forming enzymes present in demosponges (Mller et al. 2007), are responsible for the formation of silica-based structures in all sponges, we suggested that silicateins are associated with collagen (Ehrlich and Worch 2007). From our point of view, silicateins resemble cathepsins, which are known to be collagenolytic and capable of attacking the triple helix of fibrillar collagens. Therefore, it is not unreasonable to hypothesize that silicateins are proteins responsible for the reconstruction of collagen to form templates necessary for the subsequent silica formation. Recently, we confirmed experimentally that silicification of sponge collagen in vitro occurred via self-assembling, non-enzymatic mechanisms (Heinemann et al. 2007a). Bridging the nano- and microlevel, we use different techniques to create a wide spectrum of macroscopic silica-collagen-based hybrid materials (Fig. 5) which are potentially useful for technical and biomedical applications. Now, we developed an advanced procedure for the biomimetically inspired production of monolithic silica-collagen hybrid xerogels (Heinemann et al. 2007b). The disc-like samples showed convincing homogeneity and

310

Fig. 5: Three photographs of 3D hybrid scaffolds ( 10 mm) consisting of silica and Chondrosia reniformis collagen (left). Light micrograph of the 5 mm in diameter silica-chitin sphere obtained using sol-gel techniques (right).

mechanical stability, enabling cell culture experiments for the first time on such materials. We demonstrated that the silicacollagen hybrid materials exhibit proper biocompatibility, by supporting the adhesion, proliferation and osteogenic differentiation of human mesenchymal stem cells. A comprehensive understanding of silica-chitin based sponge skeletons with respect to chemical composition and structure may prove to be a novel model for the biomimetic synthesis of sponge-like three dimensional chitin-based composites (Rinaudo 2006) analogous to well established chitosan-silica hybrid materials (Shchipunov et al. 2005, Shirosaki et al. 2005) with specific optical and bioactive properties for applications in different modern technologies. To test our hypothesis that also chitin could be used as substrate for silicification, we obtained silica-chitin based materials in the form of rods or spheres (Fig. 5) using TMOS and sol-gel techniques in vitro as described in Materials and Methods. The diameter of these spheres could be varied between 2 and 10 mm. Silica-based and highly flexible hexactinellid spicules offer bioinspired lessons for potential biomimetic design of optical fibers with long-term durability that could potentially be fabricated at room temperature in aqueous solutions (Sarikaya et al. 2001). From a biomaterial point of view, it is worth noting that although the peculiar fibre-optical features of sponge spicules have attracted the attention of a large part of the scientific community only recently, the first report on their optical properties was published by Ehrenberg in 1848 (Schultze 1860). He observed that silica spicules of the hexactinellid Hyalonema exhibit double light reflection properties of unknown origin whenever deposited on a thin film of organic substances. In several recent publications (Cattaneo-Vietti et al. 1996, Aizenberg et al. 2004, Mller et al. 2006) it was demonstrated that spicules of different glass sponge species are capable of transmitting light very efficiently, following a mechanism remarkably similar to that of commercial silica optical fibers. However, none of these authors reported that the fraction of light lost within the 15 centimeters of a spicule is greater than that which is lost through a kilometre of ordinary commercial fiber. Moreover, since sponge spicules contain lots of water, they would automatically block the infrared light most commonly used for telecommunications. However, we suggest that silica-

chitin skeletons of glass sponges could be also investigated now as first examples of very ancient natural photonic crystals in comparison with other chitin-based photonic structures reported previously (Parker et al. 2001). Photonic band structures soon may find commercial application in devices that entail the inhibition of spontaneous emission, such as laser diodes and high-efficiency light-emitting diodes, as well as in integrated optics components, such as waveguides and wavelength multiplexers for optical telecommunications (Vukusic and Sambles 2003). For the purpose of developing biomaterials, not only is chitin in the form of biopolymers isolated from sponges of a large interest for practical use, but also chitin-based fibrous skeletons recently isolated from some keratose sponges and described by us (Ehrlich et al. 2003, Ehrlich et al. 2007b). The practical value of similar sponge skeletons is due to their large internal surface area estimated at between 25 and 34 m2 for a 3-to 4-gram skeleton, which enables considerable liquid absorption to take place by capillary attraction (Garrone 1978). This phenomenon is the key principle for application of 3D chitinous networks of the sponge origin as reservoirs for different kinds of liquids and gel-forming mediums which correspondingly could contain biotechnologically useful cells, bacteria or yeast, or electrolyte solutions for subsequent mineralization or metallization of the fibrous surfaces (Fig.6). Because fiber skeletons of marine demosponges (Spongia sp.) have recently been used as biomimetic scaffolds for human osteoprogenitor cell attachment, growth and differentiation, showing specific elastomeric and bioactive properties for potential applications in biomedicine and material sciences (Green et al. 2003), we are striving to obtain better understanding of the synthesis, chemical composition, and structure of these fiber-based natural constructions. We suggest that chitin sponge scaffolds with large internal surface area (Fig. 6) isolated after demineralization of native demosponges skeletons are highly optimized biocompatible structures that would support and organize functional tissues if applied in tissue engineering of bone and cartilage replacements.

311

Fig. 6: Schematic view: biomimetical potential of chitinbased natural sponge scaffolds obtained after demineralization of fibrous skeletons (Verongida).

Acknowledgements
We thank K. Tabachnik for sponge samples and discussions, T. Hanke, R. Born, P. Simon, O. Trommer, T. Douglas, S. Heinemann, G. Richter, H. Meissner, H. Zimmermann, A. Mensch, C. Fischer for their technical assistance.

References
Aizenberg J, Weaver JC, Thanawala MS, Sundar VC, Morse DE, Fratzl P (2005) Skeleton of Euplectella sp.: structural hierarchy from the nanoscale to the macroscale. Science 309(5732): 275278 Aizenberg J, Sundar VC, Yablon AD, Weaver JC, Chen G (2004) Biological glass fibers: Correlation between optical and structural properties. Proc Natl Acad Sci USA 101: 3358-3363 Anderson OR (1994) Cytoplasmic origin and surface deposition of siliceous structures in Sarcodina. Protoplasma 181: 61-77 Bergquist PR (1985) Porifera relationships. In: Morris SC, George JD, Gibson R, Platt HM (eds). The origins and relationships of lower invertebrates. Clarendon Press, Oxford. pp. 14-17 Block RJ, Bolling D (1939) The amino acid composition of keratins. The composition of gorgonin, spongin, turtle scutes and other keratins. J Biol Chem 127: 685-693 Boden N, Sommer U, Spindler K-D (1985) Demonstration and characterization of chitinases in the Drosophila Kc cell line. Insect Biochem 15: 19-23 Borchiellini C, Manuel M, Alivon E, Boury-Esnault N, Vacelet J, Le Parco Y (2001) Sponge paraphyly and the origin of Metazoa. J Evol Biol 14(1): 171-179 Bulawa CE (1993) Genetics and molecular biology of chitin synthesis in Fungi. Annu Rev Microbiol 47: 505-534 Carlstrm D (1957) The crystal structure of alpha-chitin (poly-Nacetyl-D-glucosamine). J Biophys Biochem Cytol 3(5): 669-683 Croce G, Frache A, Milanesio M, Marchese L, Causa M, Viterbo D, Barbaglia A, Bolis V, Bavestrello G, Cerrano C, Benatti U, Pozzolini M, Giovine M, Amenitsch H (2004) Structural characterization of

siliceous spicules from marine sponges. Biophys J 86(1 Pt 1): 526534 Cattaneo-Vietti R, Bavestrello G, Cerrano C, Sara M, Benatti U, Giovine M, Gaino E (1996) Optical fibres in an Antarctic sponge. Nature 383: 397-398 De La Rocha CL (2003) Silicon isotope fractionation by marine sponges and the reconstruction of the silicon isotope composition of ancient deep water. Geology 31: 423-426 Ehrlich H, Maldonado M, Hanke T, Meissner H, Born R, Scharnweber D, Worch H (2003) Spongins: nanostructural investigations and development of biomimetic material model. VDI Berichte 1803: 287-292 Ehrlich H, Worch H, Eckert C, Hanke T, Heinemann S, Meissner H (2005a) Glasschwammkollagen, Verfahren zu dessen Herstellung und Verwendung. DE-PA 10 2005 041 414 A1 2007.03.01 Ehrlich H, Hanke T, Simon P, Goebel C, Heinemann S, Born R, Worch H (2005b) Demineralization of natural silica based biomaterials: new strategy for the isolation of organic frameworks. BIOmaterialien 6: 297-302 Ehrlich H, Hanke T, Meissner H, Richter G, Born R, Heinemann S, Ereskovsky A, Krylova D, Worch H (2005c) Nanoimagery and the biomimetic potential of marine glass sponge Hyalonema sieboldi (Porifera). VDI Berichte 1920: 163-166 Ehrlich H, Ereskovsky A, Drozdov AL, Krylova D, Hanke T, Meissner H, Heinemann S, Worch H (2006) A modern approach to demineralization of spicules in glass sponges (Porifera: Hexactinellida) for the purpose of extraction and examination of the protein matrix. Russ J Mar Biol 32(3): 186-193 Ehrlich H, Worch H (2007) Collagen, a huge matrix in glasssponge flexible spicules of the meter-long Hyalonema sieboldi. In: Buerlein E (ed). Handbook of Biomineralization. Vol 1. Wiley VCH, Weinheim. pp. 23-41 Ehrlich H, Krautter M, Hanke T, Simon P, Knieb C, Heinemann S, Worch H (2007a) First evidence of the presence of chitin in skeletons of marine sponges. Part II. Glass sponges (Hexactinellida: Porifera). J Exp Zool (Mol Dev Evol) 308B: 473-483 Ehrlich H, Maldonado M, Spindler K-D, Eckert C, Hanke T, Born R, Goebel C, Simon P, Heinemann S, Worch H (2007b) First evidence

312

of chitin as a component of the skeletal fibers of marine sponges. Part I. Verongidae (Demospongia: Porifera). J Exp Zool (Mol Dev Evol) 308B: 347-356 Einbu A, Naess SN, Elgsaeter A, Varum KM (2004) Solution properties of chitin in alkali. Biomacromolecules 5(5): 2048-2054 Faulkner J (2001) Marine natural products. Nat Prod Rep 19: 1-48 Fratzl P (2007) Biomimetic materials research: what can we really learn from natures structural materials? J R Soc Interface 4: 637642 Garrone R (1978) Phylogenesis of connective tissue. Karger Press, Basel Gohel V, Singh A, Vimal M, Ashwini P, Chhatpar HS (2006) Bioprospecting and antifungal potential of chitinolytic microorganisms. Afr J Biotechnol 5: 54-72 Green D, Howard D, Yang X, Kelly M, Oreffo RO (2003) Natural marine sponge fiber skeleton: a biomimetic scaffold for human osteoprogenitor cell attachment, growth, and differentiation. Tissue Eng 9(6): 1159-1166 Heinemann S, Ehrlich H, Knieb C, Hanke T (2007a) Biomimetically inspired hybrid materials based on silicified collagen. Int J Mat Res 98: 603-608 Heinemann S, Knieb C, Ehrlich H, Meyer M, Baltzer H, Worch H, Hanke T (2007b) A novel biomimetic hybrid material made of silicified collagen: perspectives for bone replacement. Adv Eng Mater 9 (in press) Herth W (1980) Calcofluor White and Congo Red inhibit chitin microfibril assembly of Poterioochromonas: evidence for a gap betqween polymerization and microfibrill formation. J Cell Biol 87: 442-450 Iler RK (1979) The chemistry of silica. Wiley, New York Leys SP (1995) Cytoskeletal architecture and organelle transport in giant syncytia formed by fusion of hexactinellid sponge tissues. Biol Bull 188: 241-254 Leys SP (2002) The significance of syncytial tissues for the position of the Hexactinellida in the Metazoa. Interg Comp Biol 43: 19-27 Maldonado M (2004) Choanoflagellates, choanoxcytes, and animal multicellularity. Invert Biol 123: 1-22 Mayer G, Sarikaya M (2002) Rigid biological composite materials: structural examples for biomimetic design. Exp Mech 42: 395403 Mayer G (2005) Rigid biological systems as models for synthetic composites. Science 310: 1144-1147 Minke R, Blackwell J (1969) The structure of alpha-chitin. J Mol Biol 120: 167-181 Mller WEG, Wendt K, Geppert C, Wiens M, Reiber A, Schrder HC (2006) Novel photoreception system in sponges? Unique transmission properties of the stalk spicules from the hexactinellid Hyalonema sieboldi. Biosens Bioelectron 21: 1149-1155 Mller WEG, Boreiko A, Wang X, Belikov SI, Wiens M, Grebenjuk VA, Schlomacher U, Schrder HC (2007) Silicateins, the major biosilica forming enzymes present in demosponges: Protein analysis and phylogenetic relationship. Gene 395: 62-71 Ogasawara W, Shenton W, Davis SA, Mann S (2000) Template mineralization of ordered macroporous chitin-silica composites using a cuttlebone-derived organic matrix. Chem Mater 12: 28352837

Parker AR, McPhedran RC, McKenzie DR, Botten LC, Nicorovici NAP (2001) Aphrodites iridescence. Nature 409: 36-37 Preisig HR (1994) Siliceous structures and silicification in flagellated protists. Protoplasma 181: 29-42 Pringle JR (1991) Staining of bud scars and other cell wall chitin with calcofluor. Methods Enzymol 194:732-735 Reiswig HM (2004) Hexactinellida after 132 years of study - what new? Boll Mus Ist Biol 68: 71-84. Reiswig HM, Mehl D (1991) Tissue organization of Farrea occa (Porifera, Hexactinellida). Zoomorphology 110: 301-311 Rinaudo M (2006) Chitin and chitosan. Properties and applications. Prog Polym Sci 31: 603-632 Sarikaya M, Fong H, Sunderland BD, Flinn BD, Mayer G, Mescher A, Gaino E (2001) Biomimetic model of a sponge-spicular optical fiber-mechanical properties and structure. JMR 16: 1420-1428 Schrder HC, Brandt D, Schlomacher U, Wang X, Tahit MN, Tremel W, Belikov SI, Mller WEG (2007) Enzymatic production of biosilica glass using enzymes from sponges: basic aspects and application in nanobiotechnology (material sciences and medicine). Naturwissenschaften 94: 339-359 Schultze M (1860) Die Hyalonemen. Adolph Marcus, Bonn Shchipunov YA, Karpenko TY, Krekoten AV, Postnova IV (2005) Gelling of otherwise nongelable polysaccharides. J Colloid Interface Sci 287(2): 373-378 Shirosaki Y, Tsuru K, Hayakawa S, Osaka A, Lopes MA, Santos JD, Fernandes MH (2005) In vitro cytocompatibility of MG63 cells on chitosan-organosiloxane hybrid membranes. Biomaterials 26(5): 485-493 Stenzel KH, Miyata T, Rubin AL (1974) Collagen as a biomaterial. Annu Rev Biophys Bioeng 3: 231-253 Tabachnik K (1991) Adaptation of the hexactinellid sponges to deepsea fife. In: Reitner J, Keupp H (eds). Fossil and Recent Sponges, Berlin. pp.378-386 Thakur NL, Mller WEG (2004) Biotechnological potential of marine sponges. Current Sciences 86: 1506-1512 Travis DF, Francois CJ, Bonar LC, Glimcher MJ (1967) Comparative studies of the organic matrices of invertebrate mineralized tissues. J Ultrastruct Res 18(5): 519-550 Uriz MJ, Turon X, Becerro MA, Agell G (2003) Siliceous spicules and skeleton frameworks in sponges: origin, diversity, ultrastructural patterns, and biological functions. Microsc Res Tech 62(4): 279-299 Vukusic P, Sambles JR (2003) Photonic structures in biology. Nature 424: 852-855 Walter SL, Flinn BD, Mayer G (2007) Mechanisms of toughening of a natural rigid composite. Mater Sci Eng C 27: 570-574 Weaver JC, Morse DE (2003) Molecular biology of demosponge axial filaments and their roles in biosilicification. Microsc Res Tech 62(4): 356-367 Weaver JC, Aizenberg J, Fantner GE, Kisailus D, Woesz A, Allen P, Fields K, Porter MJ, Zok FW, Hansma PK, Fratzl P, Morse DE (2007) Hierarchical assembly of the siliceous skeletal lattice of the hexactinellid sponge Euplectella aspergillum. J Struct Biol 158: 93-106

Porifera research: Biodiversity, innovation and sustainaBility - 2007

313

Cytotoxic activity of hydroethanolic extracts of sponges (Porifera) collected at Pedra da Risca do Meio Marine State Park, Cear State, Brazil
Elthon G. Ferreira(1), Diego V. Wilke(1), Paula C. Jimenez(1), Tiago A. Portela(2), Edilberto R. Silveira(2), Eduardo Hajdu(3), Cludia Pessoa(1), Manoel O. de Moraes(1), Letcia V. Costa-Lotufo(1,4*)
Departamento de Fisiologia e Farmacologia, Faculdade de Medicina, Universidade Federal do Cear, Coronel Nunes de Melo, 1127, Rodolfo Tefilo, 60.430-270, Fortaleza, CE, Brasil. lvcosta@secrel.com.br (2) Departamento de Qumica Orgnica e Inorgnica, Universidade Federal do Cear, Fortaleza, CE, Brasil (3) Museu Nacional. Universidade Federal do Rio de Janeiro, RJ, Brasil (4) Instituto de Cincias do Mar, Universidade Federal do Cear, Fortaleza, CE, Brasil
(1)

Abstract: Marine sponges are well-known as a rich source of bioactive substances. This study aimed on screening the hydroethanolic extracts of 22 sponge species colleted at the Pedra da Risca do Meio Marine State Park, Cear, off the Northeast coast of Brazil, for antiproliferative, antimitotic and hemolytic activity. After collection, the species were immediately immersed in ethanol and stored at low temperatures (-20C) until use. The hydroethanolic extracts (HEEs) were obtained by evaporation of the immersion solvent and its residue was reconstituted at a concentration of 20 mg/mL in sterile dimethylsulphoxide (DMSO) to be used in the bioassays. Geodia corticostylifera and Monanchora arbuscula were highly active on all performed assays, followed by Amphimedon compressa, which showed a strong inhibitory activity on cultured cell growth and sea-urchin eggs division, but exhibited only a moderate hemolytic activity. Keywords: Brazil, Cear, cytotoxicity, Pedra da Risca do Meio Marine State Park, sponges

Introduction
In opposition to terrestrial natural products knowledge, the studies with marine natural products are quite young. However, despite the short time, this field has been conquering an important status among chemists and pharmacologists. Studies with marine natural products showed a variety of organic compounds derived from marine species with known and with novel biological activities. Sponges are among the most studied zoological groups by marine chemists and pharmacologists, while showing the highest rates of cytotoxic molecules. Several studies also describe antitumor activity (Osinga et al. 1998, Faulkner 2000, Prado et al. 2004). Life began in the oceans over 3.5 billion years ago, therefore when compared with terrestrial natural products, nature has had much more time to develop chemical armamentarium in marine organisms. It is strongly suggested that these cytotoxic compounds are produced or stored by invertebrates as a defense mechanism, such as antifouling agents against parasites or natural predators (Jimeno et al. 2004). On the other hand, the advent of high throughput screening (HTS) was very important to increase the number of samples tested. In addition to the exceptional marine biodiversity, the probability to find a bioactive compound on a random screening with marine collections is higher than in any other source (Munro et al. 1999).

Seas and oceans cover about 70% of the Earths surface and those are now viewed by the scientific community as the last great frontier for natural source of bioactive compounds (Cart 1996, Fenical 1998, Munro et al. 1999). The northeastern region of Brazil has the largest tropical coast extension of the country and remains virtually unexplored by research groups that study natural products. Nevertheless, two previous reports describing a screening for bioactive extracts of marine invertebrates collected on the coast of Cear can be acknowledged (Jimenez et al. 2003, Jimenez et al. 2004). In fact, one of the mentioned studies evaluates the cytotoxic and anti-microbial activities of hydroalcoholic extracts obtained from the most abundant sponge species of Flexeiras Beach (located about 125km west from Fortaleza, the State Capital) (Jimenez et al. 2004). It is worth mentioning that, among the 8 species screened on 4 different assays, 7 presented some kind of biological activity, whether stronger or weaker, while two species showed a very promising bioactive profile. This study describes a screening for cytotoxicity of hydroethanolic extracts derived from twenty-two sponge species collected at the Pedra da Risca do Meio Marine State Park, off the coast of Cear. Antiproliferative, antimitotic and hemolytic activities were evaluated in order to improve the knowledge on the pharmacological potential of the sponge fauna from the northeast of Brazil.

314

Materials and methods Species collection and extracts

The sponge species studied in this work were collected through SCUBA diving at Pedra da Risca do Meio Marine State Park (situated at 10 nautical miles from Mucuripe Harbor, at A, 333.800 S and 3826.000 W; B, 336.000 S and 3826.000 W; C, 336.000 S and 3821.600 W; D, 333.800 S and 3821.600 W) at depths around 20 m. Most samples were photographed in situ for better species characterization. The samples were immersed in ethanol and stored at low temperatures (-20C) until use. Voucher samples of each species were deposited at the Porifera Collection of Museu Nacional, of Universidade Federal do Rio de Janeiro (MNRJ Table 1). The hydroethanolic extracts (HEEs) were obtained by evaporation of the immersion solvent and its residue was reconstituted at a concentration of 20 mg/mL in sterile dimethylsulphoxide (DMSO) to be used in the assays. Table 1 lists the twenty two species studied on this report.

fertilization membrane under light microscope. The assay was carried out in 24-multiwell plates. The extracts were added immediately after fertilization (within 2 min) at a single concentration of 500 g/mL (N= 3) in a final volume of 2 mL/well. At appropriate intervals, aliquots of 200 L were fixed with the same volume of 10% formaldehyde to obtain first cleavage and blastulae stage. One hundred embryos from each well were counted to obtain the percentage of normal embryos.

MTT assay
The cytotoxicity activity of HEEs was tested against HL-60 (human leukemia), HCT-8 (human colon carcinoma), MDAMB435 (human breast carcinoma) and SF-295 (glioblastoma) cell lines obtained from National Cancer Institute-USA. Cells were grown in RPMI-1640 medium supplemented with 10% fetal bovine serum, 2 mM glutamine, 100 g/mL streptomycin and 100 U/mL penicillin, and incubated at 37C with 5% CO2 atmosphere. For experiments, the cells were plated in 96-well plates (1.0 x 105 cells/well for adherent cells or 0.5 105 cells/ well for suspended cells in 100 L of medium) and the HEEs (1.56 to 100.0 g/mL) were added to each well (final volume = 200 L) and incubated for 72h. Control groups received the same amount of sterile DMSO. Tumor cell growth was quantified by the ability of living cells to reduce the yellow dye 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium (MTT) to a purple formazan product. At the end of incubation period, plates were centrifuged, and the medium was replaced by 150 L of fresh medium containing 0.5 mg/mL of MTT and reincubated for 3h. The formazan was dissolved in 150 L DMSO and the absorbance was measured using a multiplate

Sea-urchin eggs development assay

The assay was performed following the method described in Jimenez et al. (2003). Adult sea-urchins, Lytechinus variegatus, were collected at Lagoinha Beach (Paraipaba District), Cear State, Brazil. Gamete elimination was induced by injecting 3.0 mL of a 0.5 M KCl solution into the sea-urchins coelomic cavity. For fertilization, 1 mL of sperm suspension (0.05 mL of concentrated sperm in 2.45 mL of filtered sea water) was added to each 50 mL of egg solution. Fertilization was assured by the observation of the

Table 1: List of the sponge species collected at Pedra da Risca do Meio State Park, Cear, Brazil. Species Agelas clathrodes (Schmidt, 1870) Agelas dispar Duchassaing and Michelotti, 1864 Agelas sp. Aiolochroia crassa (Hyatt, 1875) Amphimedon compressa Duchassaing and Michelotti, 1864 Aplysina fistularis (Pallas, 1766) Aplysina fulva (Pallas, 1766) Aplysina lactuca Pinheiro, Hajdu and Custdio, 2007 Aplysina muricyana Pinheiro, Hajdu and Custdio, 2007 Aplysina solangeae Pinheiro, Hajdu and Custdio, 2007 Aplysinidae (indet.) Callyspongia vaginalis (Lamarck, 1814) Dictyonella sp. Dragmacidon reticulatum (Ridley and Dendy, 1886) Ectyoplasia ferox (Duchassaing and Michelotti, 1864) Geodia corticostylifera Hajdu, Muricy, Custdio, Russo and Peixinho, 1992 Hyattella intestinalis (Lamarck, 1814) Ircinia strobilina (Lamarck, 1816) Monanchora arbuscula (Duchassaing and Michelotti, 1864) Niphates sp. Petromica ciocalyptoides (van Soest and Zea, 1986) Topsentia ophiraphidites (de Laubenfels, 1934) Family, Order Agelasidae, Agelasida Agelasidae, Agelasida Agelasidae, Agelasida Aplysinidae, Verongida Niphatidae, Haplosclerida Aplysinidae, Verongida Aplysinidae, Verongida Aplysinidae, Verongida Aplysinidae, Verongida Aplysinidae, Verongida Aplysinidae, Verongida Callyspongiidae, Haplosclerida Dictyonellidae, Halichondrida Axinellidae, Halichondrida Raspailiidae, Poecilosclerida Geodiidae, Astrophorida Spongiidae, Dictyoceratida Irciniidae, Dictyoceratida Crambeidae, Poecilosclerida Niphatidae, Haplosclerida Desmanthidae, Lithistids Halichondriidae, Halichondrida Voucher(s) not available MNRJ 8682, 8691 MNRJ 8678 MNRJ 8458 MNRJ 8677, 8693, 8694 MNRJ 8686 MNRJ 8683, 8690 MNRJ 8675 MNRJ 8697, 8700 MNRJ 8702 MNRJ 8698 MNRJ 8671 MNRJ 8673, 8689 MNRJ 8701 MNRJ 8695 MNRJ 8692 MNRJ 8685 MNRJ 8679 MNRJ 8670, 8674 MNRJ 8676 MNRJ 8681 MNRJ 8680

315

reader (Multimode Detector DTX 880, Beckman Coulter) (Berridge and Tan, 1993, Mosmann, 1983). HEEs effect was quantified as the percentage of control absorbance of reduced dye at 595 nm.

Results
As shown on Table 2 the extracts derived from Agelas clathrodes, Agelas sp., Dictyonella sp. and Hyattella intestinalis showed a moderate antiproliferative activity against all the cell lines tested, while Aplysina muricyana only inhibited the growth of the HL-60 cells. Amphimedon compressa, Geodia corticostylifera and Monanchora arbuscula showed a potent growth inhibition effect, with IC50s ranging from 3 to under 1.5 g/mL. Still on Table 2, the hemolytic activity of the extracts can be observed. Once more, Geodia corticostylifera and Monanchora arbuscula are responsible for the most active extracts, with their respective EC50 being 55.10 and 37.54 g/mL. The effects of the extracts on the sea-urchin eggs development were evaluated with a single concentration treatment of 500 g/mL and the results are presented on Table 3. Most of the extracts induced a significant disruption on the cell division, but different patterns can be observed. The species Amphimedon compressa, Monanchora arbuscula, Agelas dispar, A. clathrodes, Agelas sp., Aplysina muricyana, Aiolochroia crassa, Ircinia strobilina and Hyattella intestinalis showed a more pronounced inhibitory effect of the cell division, as the treated cells were mostly undivided. The species Geodia corticostylifera, Topsentia ophiraphidites, Aplysina fulva, A. fistularis, A. solangeae, Callyspongia vaginalis and Aplysinidae (indet) incited abnormal cell

Hemolytic assay
This test was performed in 96-well plates using a 1% mice erythrocytes (Mus musculus swiss) suspension in 0.85% NaCl containing 10 mM CaCl2, following the method described in Jimenez et al. (2003). The extracts were assayed at concentrations ranging from 3.9 to 1000 g/mL. After 1h incubation, the plates were centrifuged and the supernatant containing hemoglobin was measured spectrophotometrically for the absorbance at 450 nm (Multimode Detector DTX 880, Beckman Coulter).

Statistical analysis
Data for the sea-urchin eggs assay are presented as the mean of the percentage of inhibition. Data for the MTT and hemolytic assays are presented, respectively, as IC50 (mean inhibitory concentration) and EC50 (mean effective concentration). IC50, EC50 and their respective confidence intervals were obtained by nonlinear regression using the GRAPHPAD program v5.0 (Intuitive Software for Science, San Diego, USA).

Table 2: Cytotoxic and hemolytic activity of the hydro-ethanolic extracts of sponge species on tumor cell lines and mice erythrocytes. Data are presented as IC50 and EC50 values with their 95% confidence intervals by non-linear regression. Experiments were performed in triplicate. IC50 (g/mL) CI 95% Tumor cell lines HL-60 48.51 36.82 to 63.92 62.48 45.68 to 85.47 1.60 1.40 to 1.81 30.20 17.47 to 52.20 39.03 27.65 to 55.09 > 100 < 1.56 16.99 14.67 to 19.68 N.D. > 100 HCT-8 85.46 40.55 to 180.1 63.29 50.95 to 78.62 1.63 1.38 to 1.92 > 100.00 23.80 18.13 to 31.24 > 100 < 1.56 18.97 11.44 to 31.46 2.032 1.696 to 2.434 > 100 MDA-MB435 58.01 24.37 to 138.1 52.88 31.59 to 88.50 1.69 1.04 to 2.72 > 100.00 54.61 26.19 to 113.8 > 100 < 1.56 37.82 24.79 to 57.70 1.640 ----------> 100 SF-295 62.36 46.95 to 82.81 59.39 47.43 to 74.36 2.96 2.31 to 3.80 > 100.00 33.07 26.63 to 41.06 > 100 < 1.56 23.46 15.67 to 35.10 1.601 1.168 to 2.196 > 100 EC50 (g/mL) CI 95% Erythrocytes 714.0 672.9 to 757.7 883.10 854.2 to 913.1 255.1 239.5 to 271.7 > 1000.00 > 1000.00 522.60 499.4 to 546.7 55.10 46.74 to 64.95 107.50 83.23 to 38.9 37.54 29.99 to 47.00 295,1 269.0 to 323.7

Species Agelas clathrodes Agelas sp. Amphimedon compressa Aplysina muricyana Dictyonella sp. Ectyoplasia ferox Geodia corticostylifera Hyattella intestinalis Monanchora arbuscula Petromica ciocalyptoides
N.D.: not determinated.

316

Table 3: Antimitotic activity of the hydro-ethanolic extracts of sponge species at 500 g/mL on sea-urchin (Lytechinus variegatus) embryos (1st cleavage and blastulae stage). The results are showed as means of percentage of inhibition of the development and their standard deviation (S.D.). Experiments were performed in triplicates. Samples Negative control Agelas clathrodes Agelas dispar Agelas sp. Aiolochroia crassa Amphimedon compressa Aplysina fistularis Aplysina fulva Aplysina lactuca Aplysina muricyana Aplysina solangeae Aplysinidae (indet.) Callyspongia vaginalis Dictyonella sp. Dragmacidon reticulatum Geodia corticostylifera Hyattella intestinalis Ircinia strobilina Monanchora arbuscula Niphates sp. Topsentia ophiraphidites
N.D.: not determinated.

Inhibition of the development % S.D. 1st cleavage 5.000 1.000 100.0 100.0 100.0 97.00 2.646 100.0 95.67 1.528 100.0 99.00 1.732 100.0 94.67 3.512 N.D. 98.00 1.000 18.00 5.196 41.33 8.083 100.0 99.67 0.5774 100.0 100.0 45.50 2.121 99.33 1.155 Blastulae 8.667 2.082 100.0 100.0 100.0 100.0 100.0 100.0 100.0 99.67 0.5774 100.0 100.0 100.0 100.0 6.333 3.215 38.00 5.196 100.0 100.0 100.0 99.33 1.155 41.00 1.732 100.0

divisions, as the treated cells did not present regular sizes of their blast cells, or even a regular cell division pattern. Niphates sp., Dictyonella sp. and Dragmacidon reticulatum had little or no effect on the sea-urchin cell division.

Discussion
The studies conducted with marine natural products during the last decades have uncovered many substances with biomedical potential, which raised the interest of many research groups towards these ecosystems as a source of new drugs (Munro et al. 1999). Sponges are among the most promising groups, and compounds with cytotoxic and antitumor activity are the most frequently found in these organisms (Faulkner 2000). Brazil has the second most extensive coast line in the world, with over 8000 km in length. Because the main focus of Brazilian natural products chemistry has been directed to the study of plant-derived compounds, to date, mostly limited bioassay-directed screenings of extracts derived from invertebrates have been reported (Rangel et al. 2001, Monks et al. 2002, Jimenez et al. 2003, Berlinck et al. 2004, Silva et al. 2006). Information concerning the biomedical properties of the marine fauna from the north-eastern Brazilian coast remains

indeed scanty. The State of Cear, roughly located between latitudes 3 and 4o S, has a coast line of 573 km, which remains virtually unexplored by natural products chemists and pharmacologists. Pedra da Risca do Meio State Park, 10 nautical miles away from Mucuripe Harbor, in Fortaleza the States capital was created in 1997 with the goal of preserving the richness of local fauna and flora which were already suffering with environmental stress. The marine park comprises an area of 3,320 hectares with depths ranging from 17 to 30 meters and a yearly mean temperature of 27oC. Sponges are the dominant macrobenthic organisms in the park (Salani et al. 2006), as well as in most of Cears continental shelf. Some 30 sponge species were either collected or identified from underwater photos taken in four dives aiming at bioprospecting and inventorying the sponge diversity of the park. These are mostly of Tropical western Atlantic affinity, being widely distributed along the Tropical Brazilian coast line, as well as in the Caribbean region. But an important component of provisional Brazilian endemics is also present, to which additional ones may be added once the taxonomic study of this collection is concluded. The present study evaluated the cytotoxicity of 22 crude sponge extracts collected at Pedra da Risca do Meio Marine State Park on the following bioassays: 1) antiproliferative activity on cultured tumor cell lines; 2) hemolytic effect on mice erythrocytes and 3) anti-mitotic activity on sea-urchin eggs. The extract derived from G. corticostylifera was strongly active on all tested bioassays. It showed the highest inhibitory activity on the tumor cell growth and the second highest hemolytic effect on the mice erythrocytes. On the sea-urchin eggs, this extract inhibited the normal progression of the embryos development, while inducing a high degree of cell abnormalities. Previous reports have recognized the extract of G. corticostylifera as possessing cytotoxic, neurotoxic and antimicrobial activities (Muricy et al. 1993, Rangel et al. 2001). A bioassay-guided pharmacological screening study with sponge-derived extracts from specimens collected off the coast of So Paulo State, on the south-eastern region of Brazil, also found the organic extracts of G. corticostylifera to be hemolytic on mice erythrocytes and toxic to sea-urchin eggs (Rangel et al. 2001). The same study reported, as well, that the aqueous extract was highly neurotoxic on a nerve preparation of the blue crab, and, in fact, G. corticostilyfera was one of the species showing strongest pharmacological activities from among the 24 screened. Further studies with this extract revealed it to be hemolytic on frog erythrocytes, while intraperitoneal injections in mice caused death by respiratory failure (LD50 = 18.4 mg/kg) (Rangel et al. 2005). An association of the hemolytic and neurotoxic activity was suggested. Depsipeptides isolated from G. corticostylifera, the geodiamolides A, B, H and I, showed anti-proliferative activity on human breast cancer cells MCF-7 and T47D, and studies on microtubule assembly of these cells undergone treatment with geodiamolides A, B, H and I recognized that these compounds act by disorganizing actin filaments, while keeping the normal microtubule organization. Interestingly

317

enough, normal cells were not affected by the geodiamolides (Rangel et al. 2006). Monanchora arbuscula and Amphimedon compressa were equally active against the tumor cells tested, but differed on the intensity of their lytic effect. While M. arbuscula had the lowest EC50 on mice erythrocytes, A. compressa showed only a moderate activity. A similar activity profile among these extracts can also be noticed on the sea-urchin eggs, where both inhibited by nearly 100% the progression of cell division, at 500 g/mL. A M. arbuscula specimen colleted in Bahia (Northeast Brazil), yielded the isolation of crambescidin 800, a previously known compound from the MediterraneanAtlantic sponge Crambe crambe. Crambescidin 800 has cytotoxic and antiviral effects (Jares-Erijman et al. 1991, Tavares et al. 1994, 1995). On the evaluation of tumor cell growth inhibitory potential, we found 3 extracts to be highly active, while other 4 showed a moderate to weak activity and the last 15 showed no activity at all at 100 g/mL. As mentioned above, the previous screening on the cytotoxic potential of organic extracts obtained from 8 sponge species collected on the coast of Cear revealed 3 active extracts against human tumor cell lines. Another screening for cytotoxicity on sponge-derived extracts from Santa Catarina State (southern Brazil) was carried out using human tumor cell lines. Four out of 10 organic extracts were found to be active on a pre-test and showed a moderate IC50 towards the 3 cell lines tested later (Monks et al. 2002). Lytic compounds isolated from marine animals are not uncommon (Fusetani 1987). On the present study, eight species presented hemolytic activity to some extent: rather weak (3 species), moderate (3 species) or strong (2 species). The previous screening with species from Cear did not find any sponge-derived extract with hemolytic activity under 1 mg/mL (Jimenez et al. 2004). Rangel and coworkers (2001) screening of 24 sponge species from So Paulo State found 42% of the extracts to have a moderate to strong hemolytic effect on mice erythrocytes, mainly in non-polar fractions. Rangel et al. (2001) also studied the anti-mitotic potential of the 24 sponges collected in So Paulo, as did Jimenez et al. (2004) with the eight species collected in Cear. The former work found 30% of the extracts to be moderate to strong inhibitors of the embryos development (considering an IC50 between 60 and 500 g/mL), while the later found that over 60% of the extracts were active on this assay. On our study, we found that about 85% (17/20) of the extracts assayed on the sea-urchin eggs were able to disturb the normal cell division on nearly 100% of the embryos, either by inhibiting the cleavages or by inducing abnormalities, at 500 g/mL. Only three sponge-derived extracts did not have a pronounced activity on the sea-urchin egg development, suggesting that, perhaps, 500 g/mL is already a high concentration to be used alone on screenings of this nature. Other screening studies considering antimicrobial, antiviral, antichemotactic, neurotoxic, and more specifically, on microtubule integrity and cell cycle progression with Brazilian sponge-derived extracts are also available (Muricy et al. 1993, Rangel et al. 2001, Monks et al. 2002, Jimenez et al. 2004, Prado et al. 2004, Silva et al. 2006). Berlinck and coworkers (2004) summarized the results of a screening of over 300 marine-derived extracts obtained from

sponges, ascidians, bryozoans and octocorals against cancer cell lines (MCF-7, HCT-8 and B-16), resistant and non-resistant strains of Staphilococcus aureus and other microorganisms, virulent strain of Mycobacterium tuberculosis (H37Rv) and the inhibition of adenosine phosphoribosyl trasferase (LAPRT) from Leishmania major, and found marine sponges to afford the highest number of active extracts. This recent review discusses the status of Brazilian marine natural products chemistry and pharmacology. The article focuses on isolation, structure elucidation and evaluation of biological activities of natural products, highlighting the importance of bioassay-directed screenings for selection of target species and the great value of a multidisciplinary approach on studies of this nature. Finally, we bring to the attention that this is the first scientific report of any nature on species collected from Pedra da Risca do Meio Marine State Park. Few studies have been conducted at this site and very little is known about the local fauna, especially when concerning the invertebrates populations. This study is part of a more comprehensive project, which focuses on the pharmacological potential of the yet poorly explored coast of Cear. Further steps for this work have already been taken and deeper studies on chemical and pharmacological aspects of the most interesting species are already in progress.

Acknowledgements
The authors would like to acknowledge the financial support in the form of grants and/or fellowships from the following Brazilian agencies: Conselho Nacional de Desenvolvimento Cientfico e Tecnolgico (CNPq), Financiadora de Estudos e Projetos (FINEP), Fundao de Amparo Pesquisa do Estado do Rio de Janeiro (FAPERJ) and Instituto Claude Bernard (InCB). Further, we are thankful to Dr. Tito Monteiro da Cruz Lotufo, M.Sc. Luis Ernesto Arruda Bezerra and M.Sc. Sabine Schwientek for helping with the collection and sorting of samples, to B.Sc. Sula Salani Mota for assisting with voucher registration, and to the crew of Projeto Netuno for logistic support during the field trips.

References
Amador ML, Jimeno J, Paz-Ares L, Cortes-Funes H, Hidalgo M (2003) Progress in the development and acquisition of anticancer agents from marine sources. Ann Oncol 14: 1607-1615 Bergman W, Feeney RJ (1951) Nucleosides of sponges. J Org Chem 16: 981-987 Berlinck RGS, Hajdu E, Rocha RM, Oliveira JHHL, Hernndez ILC, Seleghim MHR, Granato AC, Almeida EVR, Nuez CV, Muricy G, Peixinho S, Pessoa C, Moraes MO, Cavalcante BC, Nascimento GGF, Thiemann OM, Silva M, Souza AO, Silva CL, Minarini PRR (2004) Challenges and rewards of research in marine natural products chemistry in Brazil. J Nat Prod 67: 510-522 Berridge MV, Tan AS (1993) Characterization of the cellular reduction of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolim bromide (MTT): subcellular localization, substrate dependence, and involvement of mitochondrial electron transport in MTT reduction. Arch Biochem Biophys 290: 57-65 Cart BK (1996) Biomedical potential of marine natural products. BioScience 46(4): 271-286

318

Faulkner DJ (2000) Marine natural products. Nat Prod Rep 17: 755 Fenical W (1998) New pharmaceuticals from marine organisms. Trends Biotechnol 15: 339-341 Fusetani N (1987) Marine metabolites which inhibit development of echinoderm embryos. In: Scheur PJ (ed). Bioorganic marine chemistry, vol. 1. Springer-Verlarg, Berlin, pp. 61-92. Jares-Erijman EA, Sakai R, Rinehart KL (1991) Crambescidins: new antiviral and cytotoxic compounds from the sponge Crambe crambe. J Org Chem 56: 5712-5715 Jimenez PC, Fortier SC, Lotufo TMC, Pessoa C, Moraes MEA, Moraes MO, Costa-Lotufo LV (2003) Biological activity in extracts of ascidians (Tunicata Ascidiacea) from the northeastern Brazilian coast. J Exp Mar Biol Ecol 4040: 1-9 Jimenez PC, Teixeira GLS, Wilke DV, Nogueira NAP, Hajdu E, Pessoa C, Moraes MO, Costa-Lotufo LV (2004) Cytotoxic and antimicrobial activities in hydromethanolic extracts of sponges (Porifera) from the northeastern Brazilian Coast. Arq Cien Mar 37: 85-91 Jimeno J, Faircloth G, Fernndez Sousa-Faro JM, Scheuer P, Rinehart K (2004) New marine derived anticancer therapeutics- A jouney from the sea to clinical trials. Mar Drugs 2:14-29 Monks NR, Lerner C, Henriques AT, Farias FM, Schaopoval EES, Suyenaga ES, Rocha AB, Schwartsmann G, Mothes B (2002) Anticancer, antichemotactic and antimicrobial activities of marine sponges collected off the coast of Santa Catarina, southern Brazil. J Exp Mar Biol Ecol 281: 1-12 Mosmann T (1983) Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J Immunol Methods 16: 55-63 Munro MHG, Blunt JW, Dumdei EJ, Hickford SJH, Lill RE, Li S, Battershill CN, Duckworth AR (1999) The discovery and development of marine compounds with pharmaceutical potential. J Biotechnol 70: 15-25 Muricy G, Hajdu E, Araujo FV, Hagler NA (1993) Antimicrobial activity of southwestern Atlantic shallow-water marine sponges (Porifera). Sci Mar 57: 427-432

Osinga R, Tramper J, Wijiffels RH (1998) Cultivation of marines sponges for metabolites production: applications for biotechnology? Trends Biotechnol 16: 130-134 Prado MP, Torres YR, Berlinck RGS, Desider C, Sanchez MA, Craveiro MV, Hajdu E, da Rocha RM, Machado-Santelli GM (2004) Effects of marine organisms extracts on microtubule integrity and cell cycle in cultured cells. J Exp Mar Biol Ecol 313: 125-137 Rangel M, Konno K, Brunaldi K, Procopio J, Freitas JC (2005) Neurotoxic activity induced by a hemolytic substance in the extract of the marine sponge Geodia corticostylifera. Comp Biochem Physio C 141(2): 207-215 Rangel M, Prado MP, Konno K, Naoki H, Freitas JC, MachadoSantelli GM (2006) Cytoeskeleton alterations induced by Geodia corticostylifera depsipeptides in breast cancer cells. Peptides 29(2): 2047-2057 Rangel M, Sanctis B, Freitas JC, Polatto JM, Granato AC, Berlinck RG, Hajdu E (2001) Cytotoxic and neurotoxic activities in extracts of marine sponges (Porifera) from southeastern Brazilian coast. J Exp Mar Biol Ecol 262: 31-40 Salani S, Lotufo TMC, Hajdu E (2006) Sigmaxinella cearense sp. nov. from sandstone reefs off Fortaleza (Cear State, Brazil) (Desmacellidae, Mycalina, Poecilosclerida, Demospongiae). Zootaxa 1369: 43-53. Silva AC, Kratz JM, Farias FM, Henriques AT, Santos J P, Leonel RMV, Lerner C, Mothes B, Barardi CRM, Simes CMO (2006) In vitro antiviral acticity of marine sponges collected off Brazilian coast. Biol Pharmaceut Bull 29(1): 135-140 Tavares R, Daloze D, Braekman JC, Hajdu E, Muricy G, van Soest RWM (1994) Isolation of crambescidin 800 from Monanchora arbuscula (Porifera). Biochem Systemat Ecol 22(6): 645-646 Tavares R, Daloze D, Braekman JC, Hajdu E, van Soest RWM (1995) 8b-hydroxiptilocaulin, a new guanidine alkaloid from the sponge Monanchora arbuscula. J Nat Prod 58(7): 1139-1142

Porifera research: Biodiversity, innovation and sustainaBility - 2007

319

A biogeographic comparison of sponge fauna from Grays Reef National Marine Sanctuary and other hard-bottom reefs of coastal Georgia, U.S.A.
Christopher J. Freeman(1*), Daniel F. Gleason(1), Rob Ruzicka(1,2), Rob W.M. van Soest(3), Alan W. Harvey(1), Greg McFall(4)
Department of Biology, Georgia Southern University, P.O. Box 8042, Statesboro, GA 30460-8042, USA. cjfre_freeman@yahoo.com, dgleason@georgiasouthern.edu, aharvey@georgiasouthern.edu (2) Department of Biological Sciences, Florida International University, 11200 SW 8th Street, Miami, FL 33199. rruzi001@fiu.edu (3) Section Invertebrates, Zoological Museum, University of Amsterdam, P.O. Box 94766, 1090 GT, Amsterdam, The Netherlands. soest@science.uva.nl (4) Grays Reef National Marine Sanctuary, 10 Ocean Science Circle, Savannah, GA 31411. greg.mcfall@noaa.gov
(1)

Abstract: Grays Reef National Marine Sanctuary and other hard-bottom habitats off the coast of Georgia in the south-eastern USA provide habitat for a diverse assemblage of tropical and temperate benthic organisms. These limestone, sandstone, or relic scallop shell reefs are characterized by hard-bottom ledges and escarpments of moderate relief (1-2 m above the bottom) and raised, sandy plateaus in 13 to 30 m of water. The objectives of this investigation were to 1) provide the first comprehensive list of sponge species found at these north-western Atlantic sites as well as an indication of growth forms and patterns and general habitats and 2) assess whether the sponge fauna of Georgia reefs supports the hypothesis that the Carolinian province represents a transition between temperate and tropical regions. To date, we have found 52 species of sponges, two of which are thought to be undescribed species and 15 of which are new records for the area, from eight reef habitats in this region. Published distributional records indicated that nine of the 48 taxa we could identify to species were previously reported exclusively from tropical habitats, eight only from temperate areas, and 31 from both temperate and tropical locations. This equal mix of temperate and tropical sponge species supports the contention that this area represents a biogeographic transition zone for faunas from disparate oceanic regions. In addition to supporting a biogeographically diverse sponge fauna, the ledge, plateau, and cryptic habitats off Georgia provide the topographic complexity to sustain a variety of growth forms. Keywords: biogeography, habitat, South Atlantic Bight, sponge morphology, temperate and tropical sponges

Introduction
Sponges are a dominant component of many benthic communities in tropical and temperate regions and are commonly observed on both hard and soft substrata (Reiswig 1973, Sar and Vacelet 1973, Rtzler 1978, Wenner et al. 1983, Targett and Schmahl 1984). The abundance, distribution, and diversity of sponges is relatively well documented in tropical Florida, the Caribbean, and Bermuda as well as in some temperate locations off the east coast of the United States from North Carolina to Cape Cod (George and Wilson 1919, Hartman 1964, Sterrer 1986, Alcolado 1990, Schmahl 1990, Diaz 2005, Engel and Pawlik 2005a, 2005b). In contrast, knowledge of sponge communities is more limited for the southern portions of the South Atlantic Bight (SAB), a region of the temperate northwestern Atlantic that includes coastal Georgia (SCWMRD 1982a, 1982b, Wenner et al. 1983). The SAB represents an area extending from Cape Hatteras, NC to Cape Canaveral, FL. These boundaries correspond

closely to those of the Carolinian biogeographic province (cf. Gosner 1971). Approximately 30% of the seafloor in this area is composed of hard-bottom areas of lithified limestone or sandstone embedded with fossilized scallop shells or other organisms (Harding and Henry 1994, Erv Garrison pers. comm.). Reefs in the SAB off the coast of Georgia, including those located within Grays Reef National Marine Sanctuary (GRNMS), are continuous to patchy ledge systems that vary in depth from 13-30 m and are characterized by two distinct habitats: 1) hard-bottom ridges and ledges of moderate relief (1 to 2 m above the seafloor); and 2) sandy plateaus or valleys separating adjacent ledges (Hunt 1974). Benthic invertebrates inhabiting these ledge systems in the SAB, especially those off Georgia, have received little attention. Most of our knowledge regarding diversity of benthic invertebrates in this area is contained within two large scale investigations carried out more than 25 years ago (SCWMRD 1982a, b). These studies used dredge and trawl collections to provide a description of benthic and nektonic

320

organisms at a limited number of reef sites througout the SAB, including one site within GRNMS. Whether the Carolinian province represents a distinct temperate biogeographic province or a transitional region between the temperate Virginian and tropical West Indian provinces has been controversial (Engle and Summers 1999, 2000). Despite a paucity of descriptive information, the reefs of coastal Georgia are ideally situated to study this question. The proximity of these reefs to the warm waters and tropical recruits of the Gulf Stream suggest this area is a likely habitat for a biogeographically diverse mix of benthic organisms, including sponges. Thus, the objectives of this study were to 1) survey the sponges found on SAB reefs of coastal Georgia, including GRNMS, with an emphasis on the growth forms and general habitats occupied by the species present, and 2) evaluate the extent to which the sponges we found are also known from adjacent temperate or tropical regions.

Material and methods

We surveyed sponges at eight sites off Georgia between the summers of 2002 and 2006 (Fig. 1). These sites included hard-bottom ledges within GRNMS (GRNMS Monitoring Site, Station #16, and Patch Reef #1), a neighboring lithified scallop-shell reef outside of the boundaries of the sanctuary (J Reef), three hard-bottom reefs of unknown substrate composition (Anchor Ledge, R2 Live-bottom, and Cabretta Banks), and one artificial substrate (R2 Navy Tower) (Table 1). We qualitatively estimated sponge species present at these sites by swimming the length of the ledge, across the plateau, or along the substrata looking for both common and rare, as well as cryptic species. We photographed and collected small fragments from sponges for identification in the laboratory. Samples were preserved in 70% ethanol. Skeletal structure was determined from dried thin sections that were cleared and embedded in Permount. Spicule types were determined after dissolving a fragment in bleach (5% sodium hypochlorite). Voucher specimens for each species are kept at the Department of Biology at Georgia Southern University and the Zoological Museum at the University of Amsterdam. We restricted our biogeographic comparisons to the list of sponge species we collected and identified ourselves. To assess whether the sponges of Georgia reefs represent a temperate, tropical, or transitional fauna, we broadly categorized each species as either tropical or temperate based on zoogeographic provinces that have been observed for the Atlantic coast of the United States (reviewed in Engle and Summers 1999). Specifically, sponge species reported from the Caribbean and southern Florida (south of latitude 26 N) are from the West Indian province and were designated in our study as tropical. Sponges from Bermuda were included in this tropical group based on the close proximity of this island to the Gulf Stream and the documented presence of marine flora and fauna that is characteristically tropical (Sterrer 1986). Sponge species reported from Atlantic coast locations in the United States that are north of the West Indian

Fig. 1: Map of the 8 sites included in this study. Abbreviations for the sites are as follows: JR= J Reef, AL= Anchor Ledge, CB= Cabretta Banks, MS= GRNMS Monitoring Site, P1= Patch Reef 1, St. 16= Station 16, R2 LB= R2 live-bottom, and R2 T= R2 Tower.

province up to Cape Cod, MA were designated as temperate. This designation combined records from the Carolinian (Palm Beach, Florida to Cape Hatteras, NC; approximately 26 to 35 N latitude) and Virginian (north of Cape Hatteras, NC to Cape Cod, MA; approximately 35 to 41 N latitude) provinces, but was suitable for our purposes. In our description of growth form, we placed the sponges we observed into eight categories (Fig. 2). We classified arborescent species that either grew upright or as repent branches along the substrate as branching sponges (Fig. 2C). Massive sponges were either classified as amorphous (displaying upright growth with no branching or predictable shape; Fig. 2F) or vase (exhibiting a pronounced and deep depression in the center; Fig. 2B). Encrusting sponges displayed little vertical growth and generally took on the shape of the substrata (Fig. 2G), digitate sponges were partially buried under sand with only their small digitate projections visible (Fig. 2D), and globular sponges were more or less spherical (Fig. 2A). Pedunculate sponges were upright fan or beard-shape sponges (Fig. 2E), and the clathrate growth form described sponges with a characteristic flat cushion of small (1 mm diameter) tubes (Fig. 2H).

321

Table 1: Sites in the coastal Georgia SAB surveyed for sponge fauna between 2002 and 2006 with GPS coordinates, depth ranges (due to tides and depth differences of ledge and plateau), and general topographic characteristics. Site (abbreviation) J Reef (JR) Anchor Ledge (AL) GRNMS Monitoring Site (MS) Patch Reef #1 (P1) Cabretta Banks (CB) Station 16 (St. 16) R2 Live-bottom (R2 LB) R2 Tower (R2 T) GPS coordinates 31 36.056 N 80 47.431 W 31 37.688 N 80 34.662 W 31 23.815 N 80 53.461 W 31 24.340 N 80 51.983 W 31 22.382 N 81 04.039 W 31 23.791 N 80 53.419 W 31 24.305 N 80 35.490 W 31 22.300 N 80 34.010 W Depth range (m) 18-20 25-30 14-22 14-22 13 14-22 25-30 25-30 General characteristics Sandstone and lithified scallop shell ledge/plateau Sandstone and limestone ledge/plateau Sandstone and limestone ledge/plateau Patchy hard-bottom area without defined ledge or plateau Thin veneer of sand over limestone substrate Sandstone and limestone ledge/plateau Patchy hard-bottom areas without defined ledge or plateau Artificial substrate provided by pilings of navy tower

Results
We encountered 52 species of sponges from GRNMS and neighboring hard-bottom reefs (Table 2), 48 of which we could identify to species. Two of the four species identified only to genus (Raspailia sp. and Coelosphaera sp.) are thought to be new to science. Nine of the 48 species we identified have been reported previously only from tropical regions, eight only from temperate regions, and 31 from both of these regions (Table 2). Of these 48 species, 15 are new records for the Carolinian province, two are endemic to this region, and 31 species have been either previously found in this area or have a distribution beyond this region (Table 2). Twenty-five of the 52 species from GRNMS and neighboring reefs were found predominantly on the hardbottom areas provided by the scarp, ledge, and rocky outcroppings around the ledge. Eight of these 25 species were primarily or exclusively cryptic and were located under ledges, between cracks and crevices on the scarp and between or under other sponges, gorgonians, tunicates, and bivalves (Table 3). On the other hand, none of the 12 species found predominantly on the sandy bottom around the ledges or on the plateau were observed in cryptic locations. Of the rare sponges encountered in our surveys (found only 1-2 times), ten species were observed exclusively in cryptic locations on the reef either under the rocks or ledges (Tethya sp., Callyspongia (Callyspongia) fallax, Chalinula molitba) or surrounded and partially covered by other organisms (Coelosphaera sp. nov., Mycale (Carmia) fibrexilis, Leucandra sp., Geodia gibberosa, Spheciospongia vesparium, Aulospongus pearsi, Clathrina canariensis). The remaining five species of sponges were found in cryptic locations on the metal substrate of the R2 tower (Igernella notabilis and Phorbas aff. amaranthus), or were common

on both the hard bottom, scarp region and the sandy plateau (Cliona celata, Halichondria bowerbanki, Smenospongia cerebriformis). The two dominant sponge growth forms were encrusting (40% of species) and amorphous/massive (25% of species), followed by branching, pedunculate, and digitate species. The scarp habitat, with its hard substrata, was heavily colonized by encrusting (36% of species) or amorphous/massive (32% of species) sponges. On the other hand, 66% of the species present on the sparsely colonized, sandy plateau were either digitate (Raspailia sp. nov., Ciocalypta gibbsi, Aulospongus samariensis, Axinyssa ambrosia) or pedunculate (Clathria (Clathria) carteri, Axinella waltonsmithi, Axinella bookhouti, Higginsia strigilata, and Clathria (Clathria) prolifera) (Table 3).

Discussion
Our results show that reefs in the SAB off coastal Georgia are characterized by three major habitat types, each with a distinctive set of sponge species and sponge growth forms. While a combination of biotic (predation and competition) and abiotic (sedimentation and current regime) factors likely maintain the differences in sponge species that we observed between scarp and plateau habitats, we have yet to conduct extensive investigations to determine which of these factors are most important in structuring this sponge community. However, initial observations indicating higher densities of spongivorous fish predators on scarp habitats (Ruzicka 2005) and greater sediment stress on the plateau (Gleason, pers. obs.) allow generation of hypotheses for future studies. The third major habitat type, the cryptic region, was either the predominant or sole habitat for many of the rare sponges we encountered. Again, we have not determined why these species appear to be relegated to these hidden locations, but

322

Fig. 2: Examples of sponge growth forms. A. Globular (Cinachyrella alloclada); B. Vase (Ircinia campana); C. Branching (Axinella pomponiae); D. Digitate (Axinyssa ambrosia); E. Pedunculate (Axinella waltonsmithi); F. Amorphous (Ircinia felix); G. Encrusting (Chondrosia collectrix complex); H. Clathrate (Clathrina coriacea complex). Photographs by Rob Ruzicka (A, C, D, G), Greg McFall (B, E, F), and Bernard Picton and Christine Morrow (H).

feeding by spongivorous fish and invertebrate predators or competitive exclusion by faster growing, open reef species might play a role (Meesters et al. 1991, Wulff 1997). As might be expected given the scarcity of published literature on sponges in this region, a relatively large number of our records represent major range extensions, particularily for tropical sponges, of which nine species are newly reported in temperate regions. This study also extends the southern range of Mycale fibrexilis, which was previously known only

from the Cape Cod region (Hartman 1964). In addition, from these Georgia reefs, we have identified fifteen sponge species that represent new records for the Carolinian province and two species that are considered endemic to this area. The diverse and balanced assortment of temperate and tropical sponge species found on Georgia reefs supports the contention that the Carolinian province is a true biogeographic transition zone between temperate and tropical Atlantic waters. This is consistent with recent generic-level analyses of benthic

323

Table 2: List of sponge species observed in surveys of eight coastal Georgia reefs. For each of the species observed, an * in one of the distribution columns indicates the region or regions where this species has been reported prior to this investigation. ** indicates that this species is a new record for the Carolinian province and *** indicates that this species is endemic to the Carolinian province. The tropical region includes Caribbean locations, Southern Florida, and Bermuda. Temperate refers to locations from Georgia and North Carolina to Cape Cod and both refers to species found in both tropical and temperate localities. The following references (Rf.) were used in compiling these data: 1. George and Wilson (1919); 2. de Laubenfels (1953); 3. Wells et al. (1960); 4. Hartman (1964); 5. Wiedenmayer (1977); 6. van Soest (1978); 7. van Soest (1980); 8. SCWMRD (1982a, b); 9. van Soest (1984); 10. Sterrer (1986); 11. Bibiloni et al. (1989); 12. Alcolado (1990); 13. Alvarez et al. (1990); 14. Schmahl (1990); 15. van Soest et al. (1990); 16. Alvarez et al. (1998); 17. Diaz (2005); 18. Engel and Pawlik (2005a, b); and 19. van Soest et al. (2005). Species Aiolochroia crassa (Hyatt, 1875) Aplysilla longispina George and Wilson, 1919 Aplysina fulva (Pallas, 1776) Aulospongus pearsi (Wells, Wells and Gray, 1960) Aulospongus samariensis Hooper, Lehnert and Zea, 1999 Axinella bookhouti Wells, Wells and Gray, 1960 Axinella pomponiae Alvarez, van Soest and Rtzler, 1998 Axinella waltonsmithi (de Laubenfels, 1953) Axinyssa ambrosia (de Laubenfels, 1934) Callyspongia (Callyspongia) fallax (Duchassaing and Michelotti, 1864) Chalinula molitba (de Laubenfels, 1949) Chondrilla nucula complex Schmidt, 1862 Chondrosia collectrix complex Schmidt, 1862 Chondrosia reniformis complex Nardo, 1847 Cinachyrella alloclada (Uliczka, 1929) Ciocalypta gibbsi (Wells, Wells and Gray,1960 ) Clathria (Clathria) carteri Topsent, 1889 Clathria (Clathria) prolifera (Ellis and Solander, 1786) Clathria (Thalysias) schoenus (de Laubenfels, 1936) Clathrina canariensis (Miklucho-Maclay, 1868) Clathrina coriacea complex (Montagu, 1818) Cliona caribbaea Carter, 1882 Cliona celata complex Grant, 1826 Coelosphaera sp. nov. Coscinoderma lanuga de Laubenfels, 1936 Desmapsamma anchorata (Carter, 1882) Dragmacidon reticulatum (Ridley and Dendy, 1886) Dysidea fragilis complex (Montagu, 1818) Geodia gibberosa Lamarck, 1815 Halichondria bowerbanki Burton, 1930 Higginsia strigilata (Lamarck,1814) Hyrtios violaceus (Duchassaing and Michelotti, 1864) Igernella notabilis (Duchassaing and Michelotti, 1864) Ircinia campana (Lamarck, 1816) Ircinia felix (Duchassaing and Michelotti, 1864) Leucandra sp. Leucetta imberbis (Duchassaing and Michelotti, 1864) Lissodendoryx (Anomodoryx) sigmata (de Laubenfels, 1946) Mycale (Carmia) fibrexilis Wilson, 1891 Niphates erecta Duchassaing and Michelotti, 1864 Phorbas aff. amaranthus Duchassaing and Michelotti, 1864 Ptilocaulis walpersi (Alvarez et al. 1998) Raspailia sp. nov. Scopalina ruetzleri (Wiedenmayer, 1977) Smenospongia cerebriformis (Duchassaing and Michelotti, 1864) Spheciospongia vesparium (Lamarck, 1815) Spirastrella coccinea (Duchassaing and Michelotti, 1868) Spirastrella mollis Verill, 1907 Spongia graminea Hyatt, 1877 Spongia (Spongia) tubulifera (Lamarck, 1814) Stelletta carolinensis (Wells, Wells and Gray, 1960) Tethya sp. References 17, 18a, 5, 10, 6, 14, 12, 13 10, 1 17, 12, 18a, 6, 5 3, 8a, 8b 19 3, 8a, 8b, 2 16 16, 8a, 8b, 2 15 17, 14, 7, 5, 8a, 13 17, 7, 5, 10, 14 17, 14, 12, 18a, 5, 10, 18b, 8a,b 17, 14, 5, 10, 8a 5 17, 14, 5, 10, 8a, 8b, 18a, 13 3, 15, 8a 3 9, 4, 1, 8a, 8b, 4 17, 12, 9 3, 18b, 8a 14, 12, 5, 10, 8a, 8b 17, 14, 3, 10, 8a, 14, 2 3, 1, 4, 8a 19, 3 17, 9 17 14, 11, 16, 3, 5, 2 12, 3, 5, 10, 8a, 8b, 18b, 2 3, 15, 8a, 4 3, 5, 1, 8a, 2 5 6 17, 14, 13, 18a, 6, 3, 8a, 8b, 2 17, 14, 12, 13, 1, 8a, 8b, 18a, 6, 5, 10 19, 3 9, 5, 8a 4 17, 14, 12, 13, 18a, 7, 3, 5, 10, 8a 18a, 8b, 9, 14 17, 16, 18a, 15 17, 12, 18a, 5, 10, 14, 13 18a, 8a 17, 14, 11, 3, 5, 1, 8a, 2 17, 14, 12, 18a, 3, 5, 8a 17, 10 3, 2 17, 6, 5, 8b 3 17, 14, 12, 5, 10, 2, 8b Distribution Temperate

Tropical

Both ** *

** **

*** * * * * ** * * ** * * * * * * * * * ** * * * * * * * * * ** * * * * *

**

**

**

* ** **

** ** *

** ***

324

Table 3: List of sponge species observed in surveys of eight coastal Georgia reefs along with their habitat(s) and their dominant growth form(s). The habitat column refers to the general environment where this species was predominantly found: H = hard substrate of the scarp, ledge, or rocky outcroppings, S = sandy substrate around ledges and on top of plateau, Cr = cryptic locations in crevices, under rocks, ledges, and other organisms, As = artificial substrate of R2 tower. Growth forms are characterized in the following categories: A = amorphous, B = branching, C = clathrate, D = digitate, E = encrusting, G = globular, P = pedunculate, and V = vase. Species Aiolochroia crassa Aplysilla longispina Aplysina fulva Aulospongus pearsi Aulospongus samariensis Axinella bookhouti Axinella pomponiae Axinella waltonsmithi Axinyssa ambrosia Callyspongia (Callyspongia) fallax Chalinula molitba Chondrilla nucula complex Chondrosia collectrix complex Chondrosia reniformis complex Cinachyrella alloclada Ciocalypta gibbsi Clathria (Clathria) carteri Clathria (Clathria) prolifera Clathria (Thalysias) schoenus Clathrina canariensis Clathrina coriacea complex Cliona caribbaea Cliona celata complex Coelosphaera sp. nov. Coscinoderma lanuga Desmapsamma anchorata Dragmacidon reticulatum Dysidea fragilis complex Geodia gibberosa Halichondria bowerbanki Higginsia strigilata Hyrtios violaceus Igernella notabilis Ircinia campana Ircinia felix Leucandra sp. Leucetta imberbis Lissodendoryx (Anomodoryx) sigmata Mycale (Carmia) fibrexilis Niphates erecta Phorbas aff. amaranthus Ptilocaulis walpersi Raspailia sp. nov. Scopalina ruetzleri Smenospongia cerebriformis Spheciospongia vesparium Spirastrella coccinea Spirastrella mollis Spongia graminea Spongia (Spongia) tubulifera Stelletta carolinensis Tethya sp. Habitat H H H Cr S S S S S Cr Cr H H H/Cr S S S H H/Cr Cr H/Cr H S/H Cr H H H/Cr H Cr S/H S H Cr/As H H Cr H/Cr S Cr H As/Cr S S H S/H Cr H H H/Cr H/Cr H/Cr Cr Growth form E/A E B A D P B P D E E E E E G D P P E C E/C E E/V D/E A B A A E E P A A V A A A A E B E B D E A E/G E E A E/A G G

estuarine macroinvertebrates (Engle and Summers 1999, 2000). Curiously, our survey of the sponges of the reefs in and around GRNMS revealed a dramatically different result from that of the last major faunal survey of the area, done a quarter century earlier by the South Carolina Wildlife and Marine Resources Department (1982a). That investigation, surveying GRNMS and neighboring areas, identified 61 and 77 sponges to species when collecting by dredge or trawl, respectively, but only about 20 of these species were also encountered in our surveys. Thus, we failed to find almost two-thirds of the sponges that they reported from the area, and likewise they did not report nearly two-thirds of the species that we found. This discrepancy is as yet unexplained, and may be due to any number of factors. For example, our extensive diver surveys of scarp, plateau, and cryptic sponge populations may have allowed us to find sponges restricted to the plateau and scarp, which are usually not captured in dredge and trawls. Alternatively, the discrepancies may reflect developments in sponge taxonomy and diagnostic tools, or real changes in the composition of the sponge fauna over the last 25 years. The results of this study, although still preliminary, present the first comprehensive list of the sponge fauna from coastal Georgia waters, thereby providing data on the habitats and dominant growth forms of this biogeographically and taxonomically diverse collection of sponges. Data from this study support the contention that this area represents an important zone of convergence for sponge faunas from disparate oceanic regions. In addition to the species documented above, we anticipate that the number of sponge species reported will continue to increase as we explore other sites in this region and more closely survey existing sites. Finally, this report is part of a larger project creating a field guide and web site designed to document the benthic invertebrate fauna and cryptic fishes in this area (see http:// www.bio.georgiasouthern.edu/gr-inverts/index.html). These tools are providing an important database for the scientific community, the marine sanctuaries program, and recreational divers in this region.

Acknowledgements
We thank the staff of the Grays Reef National Marine Sanctuary and NOAA for providing boats and other facilities to support our work. We especially thank Peter Fischel, Keith Golden, and Scott Fowler for the myriad of services they provided for our trips offshore. Lauren Wagner, Leslie Bates, Sarah Mock, Leslie Sutton, Hampton Harbin, and the crew of the NOAA ship NANCY FOSTER provided assistance with field work. The comments of two anonymous reviewers improved the manuscript greatly. Funding was provided by NOAAs Grays Reef National Marine Sanctuary, NOAA, and the National Undersea Research Center at the University of North Carolina at Wilmington (Award# NA030AR4300088). Collections of sponges in Grays Reef were made under permit numbers GRNMS-2003-002 and GRNMS-2005-002.

325

References
Alcolado PM (1990) General features of Cuban sponge communities. In: Rtzler K (ed.). New perspectives in sponge biology. Smithsonian Institution Press, Washington DC. pp. 351-357 Alvarez B, Diaz MC, Laughlin R (1990) The sponge fauna on a fringing coral reef in Venezuela, I: composition, distribution, and abundance. In: Rtzler K (ed.). New perspectives in sponge biology. Smithsonian Institution Press, Washington, DC. pp. 358366 Alvarez B, van Soest RWM, Rtzler K (1998) A revision of Axinellidae (Porifera: Demospongiae) of the central West Atlantic region. Smith Contrib Zool 598: 1-45 Bibiloni MA, Uriz MJ, Gili JM (1989) Sponge communities in 3 submarine caves of the Balearic Islands (Western Mediterranean) adaptations and faunistic composition. PSZN Mar Ecol 10(4): 317-334 de Laubenfels MW (1953) Sponges from the Gulf of Mexico. Bull Mar Sci Gulf Carib 2(3): 511-557 Diaz MC (2005) Common sponges from shallow marine habitats from Bocas del Toro region, Panama. Caribb J Sci 41(3): 465-475 Engel S, Pawlik JR (2005a) Interactions among Florida sponges. I. Reef habitats. Mar Ecol Progr Ser 303: 133-144 Engel S, Pawlik JR (2005b) Interactions among Florida sponges. II. Mangrove habitats. Mar Ecol Progr Ser 303: 145-152 Engel VD, Summers JK (1999) Latitudinal gradients in benthic community composition in Western Atlantic estuaries. J Biogeogr 26: 1007-1023 Engel VD, Summers JK (2000) Biogeography of benthic macroinvertebrates in estuaries along the Gulf of Mexico and western Atlantic coasts. Hydrobiologia 436: 17-23 George WC, Wilson HV (1919) Sponges of Beaufort (N.C.) harbor and vicinity. Bull Bur Fish 36: 130-179 Gosner KL (1971) Guide to identification of marine and estuarine invertebrates. Wiley-Interscience, New York Harding JL, Henry VJ Jr. (1994) Geological history of Grays Reef National Marine Sanctuary: a final report to the National Oceanographic Atmospheric Administration, Marine Estuarine Management Division under cooperative agreement NA-87-AAH-CZ033. pp. 58 Hartman WD (1964) Chapter 1: Phylum Porifera. In: Smith RI (ed). Keys to marine invertebrates of the Woods Hole region: a manual for the identification of the more common invertebrates. Marine Biological Laboratory, Woods Hole. pp. 1-7 Hunt JL (1974) The geology and origin of Grays Reef, Georgia Continental Shelf. PhD Thesis, University of Georgia, Athens Meesters E, Knijn R, Willemsen P, Pennartz R, Roebers G, van Soest RWM (1991) Sub-rubble communities of Curacao and Bonaire coral reefs. Coral Reefs 10: 189-197 Reiswig HM (1973) Population dynamics of three Jamaican Demospongia. Bull Mar Sci 23: 191-226 Rtzler K (1978) Sponges in coral reefs. In: Stoddart DE, Johannes JE (eds). Coral reefs: research methods. Monographs on oceanographic methodology, vol. 5. UNESCO, Paris. pp. 299-313

Ruzicka R (2005) Sponge community structure and anti-predator defenses on temperate reefs of the South Atlantic Bight. MSc Thesis, Georgia Southern University, Statesboro Sar M, Vacelet J (1973) cologie des Dmosponges. In: Grass PP (ed). Trait de zoologie (anatomie, systmatique, biologie). Paris, Masson. pp. 462-576 Schmahl GP (1990) Community structure and ecology of sponges associated with four southern Florida coral reefs. In: Rtzler K (ed). New perspectives in sponge biology. Smithsonian Institution Press, Washington DC. pp. 376-383 SCWMRD - South Carolina Wildlife and Marine Resources Department (1982a) South Atlantic OC area living marine resources study: Year II-Volume I: An Investigation of live-bottom habitats off South Carolina and Georgia. Washington DC. pp. 1189 SCWMRD - South Carolina Wildlife and Marine Resources Department (1982b) South Atlantic OC area living marine resources study: Year II-Volume II: An investigation of live-bottom habitats off North Carolina. Washington DC. pp. 1-143 Sterrer W (1986) Marine fauna and flora of Bermuda. John Wiley and Sons, New York Targett NM, Schmahl GP (1984) Chemical ecology and distribution of sponges in the Salt River Canyon, St. Croix, U.S.V.I. NOAA Technical Memorandum. pp. 1-29 van Soest RWM (1978) Marine sponges from Curacao and other Caribbean localities. Part I. Keratosa. Stud Fauna Curacao Caribb Isl 56(179): 1-94 van Soest RWM (1980) Marine sponges from Curacao and other Caribbean localities. Part II. Haplosclerida. Stud Fauna Curacao Caribb Isl 62(191): 1-173 van Soest RWM (1984) Marine sponges from Curacao and other Caribbean localities. Part III. Poecilosclerida. Stud Fauna Curacao Caribb Isl 66(199): 1-167 van Soest RWM, Diaz MC, Pomponi SA (1990) Phylogenetic classification of the Halichondrids (Porifera, Demospongiae). Beaufortia 40: 15-62 van Soest RWM, Boury-Esnault N, Janussen D, Hooper J (2005) World Porifera Database. http://www.marinespecies.org/porifera. Accessed September 1st to September 20th, 2006 Wells HM, Wells MJ, Gray IE (1960) Marine sponges of North Carolina. J Elisha Mitchell Sci Soc 76(2): 200-245 Wenner EL, Knott DM, Dolah RF van, Burrell VG (1983) Invertebrate communities associated with hard bottom habitats in the South Atlantic Bight. Estuar Coast Mar Sci 17: 143-158 Wiedenmayer F (1977) Shallow water sponges of the Western Bahamas. Experientia Suppl 28: 1-287 Wulff JL (1997) Parrotfish predation on cryptic sponges of Caribbean coral reefs. Mar Biol 129: 41-52 Zea S (1993) Cover of sponges and other sessile organisms in rocky and coral reef habitats of Santa Marta, Colombian Caribbean Sea. Carib J Sci 29(1-2): 75-88

Porifera research: Biodiversity, innovation and sustainaBility - 2007

327

Reproduction of two species of Halichondria (Demospongiae: Halichondriidae) in the White Sea

Elena I. Gerasimova(1*), Alexander V. Ereskovsky(1,2)
Faculty of Biology and Soil Sciences, Saint-Petersburg State University, Universitetskaya Embankment 7/9, 199034 Saint-Petersburg, Russia. eigerasimova@mail.ru (2) Aix-Marseille Universit, CNRS UMR-6540 DIMAR, Centre dOcanologie de Marseille. Station Marine dEndoume. Rue Batterie des Lions, 13007. Marseille, France
(1)

Abstract: Life history investigations of Arctic sponges allow one to observe the reproductive adaptations of these primitive metazoans to an extreme climate environment. The present study focuses on the life histories of two sympatric species Halichondria panicea and Halichondria sitiens in the White Sea. Both species appeared to be ovoviviparous and are characterized by asynchronism of gametogenesis and embryogenesis within the populations and within individuals, which agrees well with the data on halichondriids from other species and regions. On the contrary there are considerable differences in sexuality, which can be classified as successive hermaphroditism in the White Sea sponges but which varies from contemporaneous hermaphroditism to gonochorism in halichondriids from other regions. Substantial differences between halichondriid species and populations in terms of reproductive stages should also be emphasized. Oogenesis in the White Sea H. sitiens began at the same time as H. panicea but was of longer duration such that the last stages of gamete maturation and embryogenesis occurred later. This difference may be partially explained by the 2-3 week delay in warming of the deeper water layer inhabited by H. sitiens in contrast to H. panicea. A more significant reason may be related to the physiological differences between these species. Thus, the differences in sponge reproductive patterns may be caused either by environmental differences between the habitats occupied by the species or by physiological distinctions between the different taxa. Keywords: Halichondria panicea, Halichondria sitiens, sympatric species, reproduction

Introduction
Life history investigations of Arctic Porifera increase our knowledge of the reproductive adaptations of these primitive metazoans to an extreme climate. At present, even basic information concerning the reproductive biology of sponges in the Russian Arctic Region is lacking. Only two studies dealing with their life histories have been conducted (Ivanova 1981, Ereskovsky 2000). This is especially surprising since sponges are known to dominate some common benthic communities in this region (Ereskovsky 1995). The present study focuses on the family Halichondriidae, which contains sponges that are relatively common and abundant in high latitudes, dominating several hard bottom communities of the Barents and White Sea (Propp 1971). Therefore, the first aim of our research was to obtain baseline life history data on the White Sea halichondriids for the purpose of future long-term population monitoring. However, the taxonomy of the family Halichondriidae, one of the pivotal demosponge taxa, still remains difficult and confused (Erpenbeck and van Soest 2002). The use of traditional taxonomic characters in the classification of halichondriids is hampered due to considerable variation in their skeletons, which are made up entirely of smooth monaxonal megascleres. The need for additional

discriminating characters calls for the application of genetic, biochemical and life history data. In particular, a comparative embryological approach has been successfully used for the discrimination of several sympatric halichondriid species (Vethaak et al. 1982, Wapstra and van Soest 1987, Hoshino et al. 2004). Consequently, the second aim of our research was to look for any differences between the life histories of two sympatric species, Halichondria (Halichondria) panicea (Pallas, 1766) and Halichondria (Eumastia) sitiens (Schmidt, 1870). It should be emphasized that there are many data on the reproduction of the former species (Fell 1974, Ivanova 1981, Vethaak et al. 1982, Wapstra and van Soest 1987, Witte and Barthel 1994), whereas no life history studies had been undertaken on the latter species prior the present study.

Material and methods Study area and its hydrological conditions

This investigation was conducted in the area of Keret Archipelago in the Kandalaksha Bay of the White Sea (Fig. 1). The study area is characterized by a complicated bottom relief including vertical rock cliffs, large stones, sandy plains and silted trenches. The average depth is about 20 m, while the maximum depth can reach 65 m. Brown algae occur from the

328

Fig. 1: Map of the study area.

Fig. 2: Water temperatures in the study area in 2004.

intertidal to 5-8 m, while the depth range from 5-8 to 10-12 m is occupied by red algae. The hydrological conditions of the Kandalaksha Bay are characterized by considerable seasonal fluctuations of water temperature and salinity (Babkov and Golikov 1984). In addition, the water is stratified most of the year, and the amplitude of the seasonal fluctuations decreases from the surface to deeper depths. The temperature regime is determined by a long severe winter and short and relatively warm summer. From December to mid-May the coastal zone of Kandalaksha Bay is covered with ice. According to the data obtained in 2004-2005, the temperature was below zero from the beginning of December to the end of May (annual minimum -1.4 to -1.5C) in the depth range 5-15 m, which corresponded to hydrological winter (Fig. 2). In June the temperature rose rapidly, which corresponded to hydrological spring. Hydrological summer lasted from the end of June to the beginning of September. The annual temperature peak measured 15-17C depending on depth. According to the 2004-2005 data for the 5-15 m depth range of the Chupa Inlet, a salinity minimum of 22-23 was registered at the end of May. In summer the salinity fluctuated between 23-26. Hydrological winter was characterized by higher salinity values exceeding 27.

of approximately 0.5 cm3 were cut from different parts of the sponges and fixed in Bouin fluid. Further processing of sponge fragments followed standard histological techniques (Ereskovksy 2000), which included spicule dissolution in 20 % fluoride acid, dehydration in an ethanol series, celloidincastor oil mixture and chloroform, embedding in paraffin, and sectioning to 6 m thickness. The resulting sections were cleared of paraffin, stained with Mayers hematoxylin, eosin and mounted on slides. Five sections of each sponge were examined with light microscopy.

Results Halichondria panicea (Fig. 3)

Halichondria panicea is ovoviviparous. Gametogenesis and embryogenesis appear to be asynchronous both within the population studied and within individuals. Of the eight sponges sampled in June, one contained no reproductive elements, one contained both male and female gametes, and the ratio between females and males was 2:1 for the remainder of the specimens. No males were found from July to September. The ratio between reproductive and non reproductive specimens was 1:1 in July while no reproductive sponges were found in August, and in September only one sponge contained reproductive elements. Spermatogenesis occurred in June. Spherical spermatocysts of about 70 m in diameter were scattered throughout the choanosome and were most abundant in the middle and basal areas of the sponge body (Fig. 4A). The development of gametes within each spermatocyst was synchronous. All male sponges contained spermatocysts at various stages of maturation, from spermatogonia to mature sperm. The single hermaphroditic sponge possessed both spermatocysts and previtellogenous oocytes.

Sampling and processing of the material

Sponge sampling was performed by SCUBA diving from mid-June to mid-September of 2004 at approximately two week intervals. At each sampling time 5-8 individuals of each species were collected. Altogether 34 specimens of H. panicea and 42 specimens of H. sitiens were sampled. Individuals of H. panicea were collected between 3-6 m in the brown algae Laminaria sp. zone from rock substrata or from the algae thalli. H. sitiens sponges were sampled from 8-12 m depth in the red alga zone from rocks or ascidians Styela rustica. Fragments

329

Fig. 3: Growth forms of Halichondria panicea, underwater photographs.

Females containing previtellogenous oocytes were found from the middle of June. The youngest oocytes were characterized by an irregular or oval shape, homogenous cytoplasm with rare small inclusions, an homogenous nucleus and prominent nucleolus (Fig. 4B). They measured 15-24 m in diameter with nuclei of 6-11 m and nucleoli of 2.8-4.3 m. Vitellogenesis took place most actively in the second part of June. During vitellogenesis oocytes were gradually being surrounded by aggregates of ameboid cells. (Fig. 4C, D). These aggregates became a single layer of flat pinacocyte-like cells coating the mature oocytes and developing embryos. The mature oocytes had a spherical or slightly oval shape and measured 170-200 m. By the end of June the oocytes were at all developmental stages within an individual sponge. Cleavage and larval formation occurred in July. Cleavage was total, equal and chaotic, resulting in a stereoblastula (Fig. 4E). Additionally vitellogenous and previtellogenous oocytes were found among the groups of cleaving embryos within some sponges in the middle of July. At the same time, the first prelarvae were also observed. The prelarvae were typical parenchymellae with an outer layer of ciliated cells and spicules between the internal cell masses. The larvae were elongate and measured up to 225x115 m (Fig. 4F). Just prior to larval release the larval surface became folded. Release occurred in the last third of July. In the mesohyl of one sponge collected in September were a few early oocytes.

Halichondria sitiens (Fig. 5)

Halichondria sitiens is ovoviviparous. Gametogenesis and embryogenesis of this species appears to be asynchronous both within the population studied and within individuals. The ratio between reproductive and non reproductive specimens was 2:3 in June, 8:1 in July and 6:1 in August, while in September all studied sponges contained reproductive elements. In June only females were found and the sex ratio between females and males was respectively 4:1 in July, 1:2 in August and 2:1 in September. Also in August one individual contained both male and female gametes. Spermatogenesis was found from the middle of July to the middle of September. Spermatocysts were spherical and measured up to 80 m (Fig. 6A). As recorded in H. panicea the development of sperm of H. sitiens was synchronous within a spermatocyst, but each male could possess spermatocysts at different stages of maturation. The single hermaphroditic august sponge contained both spermatocysts and previtellogenous oocytes. Young previtellogenous oocytes were found from the middle of June. They measured 15-40 m in diameter with nuclei of 7-24 m, and nucleoli of 3-8 m (Fig. 6B). In July and August most of the female specimens contained both previtellogenous and vitellogenous oocytes while some sponges possessed exclusively previtellogenous oocytes. In the middle of August some oocytes reached maturity. They had a spherical or slightly oval shape and measured 180-220 m (Fig. 6C). In the middle of September all female sponges studied contained cleaving embryos but some also contained

330

Fig. 4: Reproductive elements of Halichondria panicea. A. different stages of spermatogenesis within spermatocysts (Sp). B. Previtellogenous oocyte. C. Oocyte at the early vitellogenic stage. D. Oocytes (O) at the late stage of vitellogenesis, surrounded by an aggregate of ameboid cells (CA). E. Cleaving embryos (E). F. Embryos on stereoblastula stage (E), prelarvae (Pl) and larvae (L) before release. Abbreviations: N nucleus; Nu nucleolus.

vitellogenous and previtellogenous oocytes. Cleavage was total, equal and chaotic, resulting in a stereoblastula (Fig. 6D).

Discussion
Halichondria panicea and H. sitiens in the present study appear to be ovoviviparous and are characterized by asynchronous gametogenesis and embryogenesis both within the populations and within individuals. These life history features agree well with previous data on all Halichondria species investigated to date (Sar 1993, Ereskovsky 2005). Gamete morphology and cleavage pattern observed by light

microscopy are also similar to reports of halichondriids from other regions (Ivanova 1981, Barthel and Detmer 1990, Witte and Barthel 1994). However some differences in sexuality should be emphasized. It is possible to suppose that the sexuality of both species studied is successive hermaphroditism, when the sexes are mainly separated but very a few hermaphrodites exist. The same type of sexuality was registered in the Barents Sea population of H. panicea (Ivanova 1981). Meanwhile, H. panicea and H. bowerbanki from the SW coast of the Netherlands were demonstrating contemporaneous hermaphroditism (Wapstra and van Soest 1987), while H. panicea from the Kiel Bight of the Baltic Sea was reported

331

Fig. 5: Growth forms of Halichondria sitiens, underwater photographs.

Fig. 6: Reproductive elements of Halichondria sitiens. A. Different stages of spermatogenesis within spermatocysts. B. Previtellogenous oocytes. C. Vitellogenous oocytes (O) at different developmental stages, surrounded by an aggregate of ameboid cells (CA). D. Cleaving embryos in the mesohyl. Nuclei of blastomeres are indicated by arrows. Abbreviations: N - nucleus; Nu - nucleolus; Sp spermatocyst; Sz - spermatocyst with the spermatozoa.

332

to be gonochoristic (Witte and Barthel 1994). Thus, it can be concluded that the species within the genus Halichondria as well as the populations within the species H. panicea are characterized by a labile sexuality. This finding is a result of differences in environmental conditions rather than to taxonomic differences. There are also substantial differences between halichondriid species and populations in terms of reproductive stages (Table 1). In the White Sea subtidal population of H. panicea studied here gametogenesis went active in June early July when the local water temperature was rapidly rising from 0C to 10-13C, and larvae were released in late July early August when the temperature reached its annual peak of 15-17C. In the Baltic Sea population of this species the period of larval release varied in different years from the April to August depending on temperature dynamics (Barthel 1986, 1988, Witte and Barthel 1994). Similar variability in timing of larval release was observed in the Netherlands population of H. panicea. Vethaak et al. (1982) registered mature oocytes and embryos from mid-May to mid-August whereas Wapstra and van Soest (1987) observed embryos and larvae from May to September. The Barents Sea intertidal population of H. panicea examined by Ivanova (1981) was reproductively active mainly in JulyAugust, when the local water temperature reached the annual peak of 7-8C and larvae were released in September-October, when the temperature fell to 4-5C. These differences between the three populations of H. panicea in terms of reproduction may be explained by the environmental preferences of this Atlantic borealArctic species. The Kiel Bight of the Baltic Sea where Witte and Barthel (1994) conducted their studies is located approximately in the middle of the biogeographical range inhabited by H. panicea and is characterized by a mild climate with relatively high annual temperatures and early spring warming. On the contrary the White and Barents Seas regions are the extreme of the distributional ranges of H. panicea and these areas are characterized by severe climates with late spring warming. This may cause the delay in reproductive activity of local sponges. More significant differences can be found if one compares the reproductive timing of different species of Halichondria. In H. bowerbanki populations inhabiting the Netherlands waters, mature oocytes and embryos were recorded at either of two times, from early August to mid-October (Vethaak et al. 1982) or from June to November (Wapstra and van Soest 1987). The latter authors suppose that in its reaction to water temperatures H. panicea exhibits a cooler thermal range for existence and reproduction than H. bowerbanki, which requires warmer conditions in the Oosterschelde area. The reduced tolerance to lower temperature causes H. bowerbanki to reproduce later in the year. It appears that there is little or no geographical variation in the reproductive periods: H. panicea invariably breeds within a rising temperature range, H. bowerbanki in a stable or decreasing temperature range (Wapstra and van Soest 1987). In the White Sea population of H. sitiens in this study the last stages of gametogenesis and embryogenesis occurred about 8 weeks later in comparison with H. panicea of the same region. It may be partially explained by the 2-3 weeks delay of increasing

Wapstra and van Soest, 1987 JulyNovember Wapstra and van Soest, 1987 AprilJune Witte and Barthel, 1994 late Julyearly August present study MaySeptember JuneNovember NovemberMay JuneJuly AugustNovember JuneSeptember H. panicea H. panicea MarchMay JuneJuly MarchMay July AprilNovember Year-round JuneNovember H. bowerbanki JuneNovember

Ivanova, 1981

late JanuaryOctober

Previtellogenesis

Table 1: Reproductive periods of Halichondria species at different localities.

Spermatogenesis

Species

H. panicea

H. panicea

H. sitiens

Murman Coast, Barents sea Oosterschelde area, North sea Oosterschelde area, North sea Kiel Bight, Baltic sea Kandalaksha Bay, White sea Kandalaksha Bay, White sea

Region

JuneSeptember

JuneOctober

JuneSeptember

JulySeptember

Vitellogenesis

JulyAugust

late AugustSeptember

Embryogenesis

JulyAugust

? late September October

late Augustearly October JuneSeptember

Larvae

present study

Reference

333

water temperature at 8-12 m depth from which H. sitiens was sampled. In comparison the sampling site of H. panicea was 3-6 m depth and increasing temperature occurred earlier. An additional more likely reason for the stated difference in reproductive onset between the two species may be the result of the physiological differences between the species. The similar situation was observed by Ereskovksy (2000) in the White Sea populations of Halisarca dujardini, Myxilla incrustans and Iophon piceus. The release of larvae in the former species sampled from the depth range 1,5-5 m took place in July, whereas in two other species collected from 15-25 m the larvae were released in September-October. We propose two explanations for the later reproductive onset of H. sitiens compared to H. panicea. The first explanation applies to the low tolerance to low temperatures and subsequent delay of the most energy-consuming stage of gametogenesis, i.e. vitellogenesis, until a warmer period, as emphasized by Wapstra and van Soest (1987) for H. bowerbanki. The second explanation may concern the fact that the released larvae of H. sitiens are probably adapted for a rather narrow temperature range. In contrast to H. panicea, H. sitiens is a high boreal-Arctic species know to occur as far north as the Greenland and the Kara Sea (Koltun 1966) and is evidently well adapted to severe climatic conditions. The White Sea is situated in the middle of its range and the local climate allows for the extension of gametogenesis, and the delay of larval release until the autumn temperature decrease. It can be concluded that the differences in reproductive patterns of halichondriid sponges may be caused either by environmental differences within the geographical range occupied by a species, or by physiological distinctions between different taxa.

Acknowledgements
We thank Alexander Plotkin and Michael Fedjuk for diving assistance, and Natalia Lentsman for improving the English. This work was financially supported partially by the programs Universities of Russia UR. 07.01.325, Russian Foundation for Basic Research grants 06-04-48660, 06-04-58573, and by grant of Marie Curie IIF MIF1-CT-2006-040065.

References
Babkov AI, Golikov AN (1984) Hydrobiocomplexes of the White Sea. Editions of the Zoological Institute, Academy of Sciences of USSR. Leningrad Barthel D (1986) On the ecophysiology of the sponge Halichondria panicea (Pallas) in Kiel Bight. I. Substrate specifity, growth and reproduction. Mar Ecol Prog Ser 32: 291-298 Barthel D (1988) On ecophysiology of the sponge Halichondria panicea (Pallas) in Kiel Bight. II. Biomass, production, energy

budget and integration in environmental processes. Mar Ecol Prog Ser 43: 87-93 Barthel D, Detmer A (1990) The spermatogenesis of Halichondria panicea (Porifera, Demospongiae). Zoomorpology 110: 9-15 Ereskovsky AV (1995) Materials to the faunistic study of the White and Barents seas sponges. 5. Quantitative distribution. Berliner geowiss Abh (A) 16: 709-714 Ereskovsky AV (2000) Reproduction cycles and strategies of the coldwater sponges Halisarca dujardini (Demospongiae, Halisarcida), Myxilla incrustans and Iophon piceus (Demospongiae, Poecilosclerida) from the White Sea. Biol Bull 198(1): 77-87 Ereskovsky AV (2005) Comparative embryology of sponges (Porifera). Saint-Petersburg University Press, Saint-Petersburg Erpenbeck D, van Soest RWM (2002) Family Halichondriidae Gray, 1867. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges, vol. 1. Kluwer Academic/ Plenum Publishers, New York. pp. 787-815 Fell PE (1974) Porifera. In: Giese AC, Pearce JS (eds). Reproduction of marine invertebrates, vol.1. Academic Press, New York. pp. 51132 Hoshino S, Takeda M, Watanabe Y (2004) Systematic status of Halichondria japonica (Kadota) (Demospongiae, Halichondrida) from Japan. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Bull Mus Ist Biol Univ Genova 68: 373-379 Ivanova LV (1981) The life cycle of Halichondria panicea (Pallas). In: Korotkova GP (ed). Morphogenesis in sponges. Trud Biol Nauch Issled Inst, vol. 33. Publ. Leningrad University, Leningrad. pp. 59-73 Koltun VM (1966) Tetraxonid sponges of the North and the Far East Seas of the USSR. Nauka, Moscow-Leningrad Propp MV (1971) Ecology of coastal benthic communities of the Murman coast of the Barents Sea. Nauka, Leningrad Sar M (1993) 1. Porifera. Reproductive Biology of Invertebrates. In: Adiyodi KG, Adiyodi RG (eds). Sexual differentitation and behaviour. Oxford & IBH Publishing Co., New Delhi. pp. 1-29 Vethaak AD, Cronie RJA, van Soest RWM (1982) Ecology and distribution of two sympatric, closely related sponge species, Halichondria panicea (Pallas, 1766) and H. bowerbanki Burton, 1930 (Porifera, Demospongiae), with remarks on their speciation. Bijdr Dierk 52(2): 82-102 Wapstra M, van Soest RWM (1987) Sexual reproduction, larval morphology and bechaviour in Demosponges from the southwest of the Netherlands. In: Boury-Esnault N, Vacelet J (eds). Taxonomy of Porifera. NATO ASI Ser. vol. 13. Springer-Verlag, Berlin. pp. 281-307 Witte U, Barthel D (1994) Reproductive cycle and oogenesis of Halichondria panicea (Pallas) in Kiel Bight. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 297-306

Porifera research: Biodiversity, innovation and sustainaBility - 2007

335

Sponge community structure and disease prevalence on coral reefs in Bocas del Toro, Panama
Deborah J. Gochfeld(1*), Carmen Schlder(2), Robert W. Thacker(3)
National Center for Natural Products Research, University of Mississippi, P.O. Box 1848, University, MS 38677-1848, USA. gochfeld@olemiss.edu (2) Smithsonian Tropical Research Institute, Unit 0948, APO AA, 34002, USA. schloederc@si.edu (3) Department of Biology, University of Alabama at Birmingham, 109 Campbell Hall, 1300 University Blvd., Birmingham, AL 35294-1170, USA. thacker@uab.edu
(1)

Abstract: Sponges are sessile, filter-feeding organisms that are sensitive to both biotic and abiotic components of their environment and are therefore likely to be impacted by environmental stressors. For this reason, sponges are useful as bioindicators of changing environmental conditions. The present study characterized sponge diversity, abundance and disease prevalence on three reefs in Bocas del Toro, Panama. The reefs were similar in general characteristics, however, one site was located just offshore of a village where anedoctal reports suggest that black water outflow (sewage, road pollution, and solid waste dumping) occurs. Overall, 51 species and 2532 individual sponges were identified. Analysis of similarity indicated significant differences in sponge community structure between the sites. The site nearest Saigon village had significantly fewer species per quadrat, although total number of individuals and number of individuals per quadrat were similar between this site and the more distant, upstream site (Punta Caracol). Evenness (J) and diversity (H) were significantly reduced at Saigon, as was the slope of the species-area curve. Dominant species also differed among sites, with the most abundant species at Saigon considered rare at the two upstream sites. Only Niphates erecta was among the five most dominant species at all three sites; Aplysina fulva, which dominated the upstream sites and is known to be sensitive to stress, was rare at Saigon, and Chondrilla nucula, another stress-intolerant species, was only dominant at Casa Blanca. Hymeniacidon sp., on the other hand, dominated the reef only at Saigon. Aplysina red band syndrome (ARBS) was present at all three sites, but prevalence was higher and more variable at Saigon. Differences in sponge community structure and disease prevalence at these three sites in Bocas del Toro, Panama, may be indicative of differences in environmental conditions on these reefs. Keywords: biodiversity, disease, disturbance, sponge community structure

Introduction
Caribbean coral reefs have undergone dramatic changes over the past several decades, resulting in shifts from coral- to macroalgal-dominated communities (Porter and Meier 1992, Hughes 1994, Jackson 1997, Hughes et al. 2003, Aronson et al. 2004). The cause of these changes is presumably multi-faceted, and both natural and anthropogenic factors are clearly important (Hughes et al. 2003, Mumby et al. 2006). Corals have been the primary focus of most studies that have catalogued these shifts in community structure. Although corals are clearly susceptible to the effects of nutrient enrichment, pollution, turbidity and sedimentation (Dubinsky and Stambler 1996, Fabricius 2005, Kuntz 2005), other benthic invertebrates may exhibit differential responses to these potential stressors. Next to corals, sponges are the most important macrofauna on Caribbean coral reefs (Wulff 2001, Diaz and Rtzler 2001, Rtzler 2004). Sponges are sessile, filter-feeding organisms that are sensitive to both biotic and abiotic components of

their environment and are therefore likely to be impacted by environmental stressors. An increasing number of studies has documented variations in sponge community structure correlated with water quality parameters. In some cases, sponge biomass may increase with proximity to runoff (Zea 1994); however, it appears that increasing amounts of stress (e.g. increasing concentrations of organic pollutants) result in reduced sponge diversity, leaving only specialists behind (Alcolado and Herrera 1987, Muricy 1989, 1991, Muricy et al. 1991, Alcolado 1994, Rtzler 2004, Vilanova et al. 2004). For this reason, sponges, both as a community and as individual organisms, have been considered useful bioindicators of changing environmental conditions. Diseases of marine organisms, notably corals, have been reported with increasing frequency over the past few decades, particularly on Caribbean reefs (Harvell et al. 1999, Rosenberg and Loya 2004, Weil 2004). While the specific causes of this increase have not been fully elucidated, several studies have identified correlations between disease prevalence and localscale processes such as pollution (Green and Bruckner 2000,

336

Sutherland et al. 2004, Kaczmarsky et al. 2005). Although diseases of sponges are less well known, these too have been reported more frequently in recent years (Rtzler 1988, Vacelet et al. 1994, Pronzato et al. 1999, Webster et al. 2002, Olson et al. 2006, Wulff 2006). The few earlier descriptions of disease among Caribbean sponges typically described outbreak conditions, such as those affecting the commercial sponge industry in the 1940s (reviewed by Lauckner 1980). Recently, Olson et al. (2006) described a new condition, Aplysina red band syndrome (ARBS), affecting approximately 10% of rope sponges in the Aplysina cauliformis-A. fulva complex on reefs in the Bahamas. The disease manifests as a red band composed of filamentous cyanobacteria surrounding a necrotic lesion that becomes colonized with algae. ARBS weakens the sponge skeleton, which often breaks at the site of the lesion. To date, the prevalence of ARBS in areas other than the Bahamas is unknown, and the relationship between this syndrome and environmental variables has not yet been characterized. The Bocas del Toro Archipelago consists of a complex network of rain-forested islands, mangrove cays and peninsulas surrounding shallow bays on the Caribbean side of the isthmus of Panama (Collin et al. 2005). The region has complex and poorly described current patterns and is subjected to episodic severe rainfall of 3-5 m annually (Collin et al. 2005), resulting in high sedimentation and turbidity. Although historically home to small communities of indigenous people, there has been a long history of deforestation for agriculture (Carruthers et al. 2005) and, more recently, larger human settlements. Coral reefs in Baha Almirante are well developed and highly diverse (Collin et al. 2005). The present study investigated sponge community structure on a coral reef offshore of a village and on two reefs upstream from this site. We examined sponge diversity, abundance and disease prevalence at all three sites, and assessed whether observed variations in these parameters were associated with variations in environmental conditions.

Fig. 1: Map depicting location of the three study sites at Bocas del Toro, Panama. PC = Punta Caracol, CB = Casa Blanca, S = Saigon.

at 5-7 m depth. Due to severe episodic pulses in rainfall, these reefs are subjected to variations in temperature, salinity, sedimentation and turbidity (Kaufmann and Thompson 2005). Visibility at all three sites during the survey dives was less than 10 m.

Surveys
At each site, a 30 m transect line was laid along the general axis of the reef at 5 m depth. Sponge diversity and abundance were characterized from 15 1-m2 quadrats placed on alternating sides at every other meter along the transect line. Every individual sponge within each quadrat was identified to the lowest possible taxonomic level and counted. For sponges that occurred as multiple independent ramets, even if they appeared to be part of a larger genet, each ramet was counted as a separate individual. If sponges could not be identified with certainty in situ, underwater photographs and/or small vouchers were collected for subsequent identification. Prevalence (percent of affected sponges) of ARBS (Olson et al. 2006) was assessed from three band transects (10 x 1 m) paralleling the transect line on each reef. Additional syndromes observed and the identities of their host sponges were also noted. Surveys were performed on 23-24 August 2005.

Materials and methods Study sites

Three reefs were surveyed in Baha Almirante along the south shore of Isla Colon in Bocas del Toro, Panama (Fig. 1). The three reefs were located just offshore of the village Saigon (92035N / 821525W), approximately 2 km upstream from Saigon (Casa Blanca, 92132N / 821631W) and approximately 6 km upstream from Saigon (Punta Caracol, 92238N / 821810W) (Figure 1). There are anecdotal reports of black water outflow (sewage, road pollution and solid waste dumping) off of Saigon (C. Schlder, pers.obs). Within 1 km of that reef site are 200 houses, an airport and several businesses. In contrast, Casa Blanca has six houses and Punta Caracol has four houses and an eight-room hotel (Google Earth 2006). Punta Caracol is at the mouth of Big Bight, a large drainage basin, and a prevailing long-shore current flows west to east (C. Schlder, pers. obs.), with Saigon being the easternmost of these sites. All three reefs were similar in general characteristics, consisting of a high diversity of both sponges and corals on a relatively flat slope

Environmental variables
In an attempt to characterize differences in water quality parameters among the sites, water samples were collected on 15 November 2005 for analyses of inorganic nutrients and

337

on 13 July 2006 for analyses of inorganic nutrients, fecal coliform bacteria and polycyclic aromatic hydrocarbons (PAHs). There were many different parameters that could have been measured; however, these were chosen to represent potential inputs from sewage and road runoff. Although these samples were not collected concurrently with the community surveys, the relative levels of these indicators among sites is still likely to be indicative of putative risks to these coral reef communities. Nutrient and fecal coliform analyses, and preliminary processing of samples for PAHs were performed at the Smithsonian Tropical Research Institutes Bocas del Toro Research Station on Isla Colon. Concentrations of inorganic nutrients (micromolar: M) in three replicate water samples from each site were determined in November using a chemical titration method (nitrate, nitrite, phosphate; DCroz et al. 2005), and in July using a Hach DR/890 Colorimeter (nitrate, phosphate, ammonia). For the fecal coliform analyses, three replicate 1 L samples of water from each site were collected and filtered through sterile 0.45 m nitrocellulose membrane filters. Filters were transferred onto Petri plates containing absorbent pads to which 2 ml of mFC media (Millipore) was added. Plates were incubated at 37C for 24h, photographed and counted. Colonies with a blue coloration were identified as fecal coliform bacteria. Additional 1 L samples (n=3 per site) of water were collected for PAH analysis. These samples were filtered through preconditioned Waters Oasis HLB cartridges (20 cc/1 g) until dry. Cartridges were kept cold prior to and during transport to the University of Mississippi, where they were rinsed with 10 ml ddH2O, then eluted with methylene chloride (16 ml) into amber vials. Extracts were evaporated to dryness, weighed, and solubilized in 200 l isooctane. Samples were run on an Agilent gas chromatograph coupled to a mass spectrometer, using the method described by Wade et al. (1993). PAHs within our samples were quantified against a standard curve comprising 16 standards in solution (UltraScientific, Cat. No. US-106N-4) from 0.05 to 0.5 ppm.

two-way non-parametric ANOVAs on rank-transformed data. Counts of fecal coliform bacteria and concentrations of individual PAHs among the three sites were compared using Kruskal-Wallis tests.

Results Surveys
Overall, 51 species (Table 1) and 2532 individual sponges were identified in the 45 quadrats. Analysis of similarity (ANOSIM) indicated significant differences in sponge community structure among the three sites (Fig. 2; ANOSIM, R=0.705, P<0.001). Punta Caracol and Casa Blanca appeared to be more similar to each other than to Saigon, however, pair-wise comparisons indicated significant differences among all three sites (P<0.001 for each pair). Most species (41.2%) were observed at all three sites; however, 33.3% of species were observed from only a single site. Of the species found at two sites (25.5%), a greater number (6) were seen at Punta Caracol and Casa Blanca than were seen at either of those sites and Saigon (4 and 3 for Punta Caracol and Casa Blanca, respectively). Saigon had significantly fewer species per quadrat than Punta Caracol, which had significantly fewer species per quadrat than Casa Blanca. The total number of individuals per site and the number of individuals per quadrat were similar between Saigon and the more distant upstream site, Punta Caracol (Table 2), and were significantly lower than at Casa Blanca. Evenness (J) was significantly reduced at Saigon, compared to the other two sites (Table 2), and diversity (H) differed significantly at all three sites, with Saigon having the lowest diversity and Casa Blanca the highest. Rarefaction (species-area) curves also indicate that the total number of species at Saigon is lower than at the other two sites (Fig. 3). Dominant species also differed among sites, with one of the five most abundant species at Saigon (Hymeniacidon sp.) considered rare at the upstream sites (Fig. 4). Only Niphates erecta was among the five most dominant species at all three sites. Aplysina fulva, which dominated the reefs at Casa Blanca and Punta Caracol, was rare at Saigon. Chondrilla nucula and Neopetrosia subtriangularis, two of the dominant species at Casa Blanca, were rare or completely absent at the other two sites Lesions on sponges were observed at all three sites. ARBS was present at all three sites, but prevalence was higher and more variable at Saigon (Fig. 5). However, the quadrat data indicate that the number of Aplysina rope sponges at Saigon (n=8) was negligible compared to the other two sites, where they were among the dominant species (n=127 and 272 at Punta Caracol and Casa Blanca, respectively). Other types of lesions were observed on Iotrochota birotulata (Punta Caracol), Plakortis angulospiculatus (Punta Caracol and Casa Blanca), Amphimedon compressa (Casa Blanca) and Neofibularia nolitangere (Saigon).

Data analysis
Species abundance data were arranged in a species by quadrat matrix and analyzed using the PRIMER software package (Plymouth Routines in Multivariate Ecological Research, version 5.1.2; Clarke and Warwick 1994). A similarity matrix was calculated using the fourth-root of the Bray-Curtis similarity index. Analyses of similarity (ANOSIM) were used to compare similarity within and among sites. Non-metric multi-dimensional scaling was used to generate a two-dimensional ordination of the communities. For each quadrat, the number of species, the number of individual sponges, the Shannon index of diversity (H), and Pielous measure of evenness (J) were calculated (Magurran 1988). These four indices were compared among the three sites using Kruskal-Wallis tests. Post-hoc means comparisons were performed using Mann-Whitney U tests. Rarefaction (species accumulation) curves were generated using the EstimateS software package (Colwell 2004). Nutrient concentrations at the three sites on each date were compared using Kruskal-Wallis tests. The effects of date and site on concentrations of nitrate and phosphate were analyzed using

Environmental variables
Comparisons of nutrient concentrations among sites at each date are shown in Table 3. Nitrate concentrations varied greatly between the two sampling periods, but did not vary

338

Table 1: Distribution and abundance of sponge species identified in the quadrats from three sites in Bocas del Toro, Panama. Numbers indicate number of individual sponges (or ramets) in 15 1-m2 quadrats at each site. Species Aiolochroia crassa (Hyatt, 1875) Amphimedon compressa Duchassaing and Michelotti, 1864 Aplysina cauliformis (Carter, 1882) Aplysina fulva (Pallas, 1766) Aplysina lacunosa (Pallas, 1766) Chondrilla cf. nucula Schmidt, 1862 Cinachyrella alloclada (Uliczka, 1929) Cliona aprica Pang, 1973 Cliona delitrix Pang, 1973 Dragmacidon reticulatum (Ridley and Dendy, 1886) Ectyoplasia ferox (Duchassaing and Michelotti, 1864) Haliclona (Halichoclona) vansoesti de Weerdt, de Kluijver and Gomez, 1999 Haliclona (Rhizoniera) curacaoensis (van Soest, 1980) Haliclona sp. Haliclona sp. (unidentified) Halisarca caerulea Vacelet and Donadey, 1987 Hymeniacidon sp. (unidentified) Hyrtios proteus Duchassaing and Michelotti, 1864 Iotrochota birotulata (Higgin, 1876) Ircinia campana (Lamarck, 1814) Ircinia felix (Duchassaing and Michelotti, 1864) Ircinia strobilina (Lamarck, 1816) Ircinia sp. (undescribed) Lissodendoryx (Lissodendoryx) colombiensis Zea and van Soest, 1986 Monanchora arbuscula (Duchassaing and Michelotti, 1864) Mycale (Arenochalina) laxissima (Duchassaing and Michelotti, 1864) Mycale (Carmia) microsigmatosa Arndt, 1927 Mycale (Mycale) laevis (Carter, 1882) Neofibularia nolitangere (Duchassaing and Michelotti, 1864) Neopetrosia carbonaria (Lamarck, 1814) Neopetrosia subtriangularis (Duchassaing, 1850) Niphates caycedoi (Zea and van Soest, 1986) Niphates erecta Duchassaing and Michelotti, 1864 Oceanapia nodosa (George and Wilson, 1919) Petrosia (Petrosia) pellasarca (de Laubenfels, 1934) Placospongia intermedia Sollas, 1888 Plakortis angulospiculatus (Carter, 1882) Plakortis halichondrioides (Wilson, 1902) Spirastrella hartmani Boury-Esnault et al. 2000 Spirastrella sp. (unidentified) Spongia (Spongia) pertusa Hyatt, 1877 Svenzea zeai (Alvarez, van Soest and Rtzler, 1998) Tedania (Tedania) ignis (Duchassaing and Michelotti, 1864) Verongula reiswigi Alcolado, 1984 Verongula rigida (Esper, 1794) Xestospongia sp. Xestospongia proxima (Duchassaing and Michelotti, 1864) Xestospongia rosariensis Zea and Rtzler, 1983 Unidentified sp. 1 Unidentified sp. 2 Unidentified sp. 3 Punta Caracol 23 34 33 94 8 9 9 21 3 6 0 9 0 0 1 3 16 0 10 1 8 1 1 2 21 6 1 26 6 0 0 4 52 1 3 1 10 0 0 75 1 0 4 0 11 86 0 0 0 0 0 Casa Blanca 11 36 92 180 7 265 0 35 5 7 22 0 4 90 0 42 8 3 1 1 10 1 0 3 30 0 0 68 0 15 98 0 98 5 2 8 13 6 40 61 0 0 9 1 26 6 8 11 0 1 12 Saigon 1 11 0 8 2 0 4 12 3 5 0 0 2 0 0 1 303 0 6 0 8 3 0 0 4 0 4 88 0 1 0 16 28 3 2 0 9 0 0 22 1 1 0 0 2 4 0 18 1 0 0

significantly among sites (Table 3; Two way non-parametric ANOVA, df = 1, F = 66.441, P < 0.0001 for date, df = 2, F = 0.099, P = 0.9067 for site). Phosphate concentrations did not vary significantly over time but there was a trend towards Punta Caracol having higher concentrations than the other two sites (Two way non-parametric ANOVA, df = 1, F = 2.059, P = 0.1768 for date, df = 2, F = 3.629, P = 0.0585 for site). When analyzed separately, phosphate concentrations varied significantly among sites in November, but not in July (Table

3). Nitrite concentrations also varied significantly among sites in November (Table 3). For both phosphate and nitrite concentrations in November, lower concentrations were found at Casa Blanca that at Punta Caracol and Saigon, which were similar to each other. Fecal coliform counts were highly variable within and among sites; however, a statistically significant difference between coliform numbers was not detected (Table 3). Ten PAHs were detectable in our samples at low concentrations that were highly variable even among

339

Fig. 2: Non-metric multi-dimensional scaling ordination of sponge diversity in 15 1-m2 quadrats at each of the three sites in Bocas del Toro, Panama; stress = 0.2. Each symbol represents the community composition based upon a single quadrat. Closer symbols indicate more similar assemblages.

Fig. 3: Rarefaction (species-area) curves based on species counts in 1-m2 quadrats at the three sites in Bocas del Toro, Panama.

replicates within sites (Table 3). Only for phenanthrene and anthracene was there a trend towards differences among sites (Kruskal-Wallis, df = 2, H = 5.600, P = 0.0608 for both), with Punta Caracol having significantly higher concentrations of both compounds than Casa Blanca, and Saigon being intermediate (Table 3).

Discussion Site
The reefs of Baha Almirante in Bocas del Toro, Panama, are well-developed and host high biodiversity (Collin et al. 2005). However, the region receives a significant amount of rainfall, usually in severe episodic downpours, resulting in excessive runoff of terrestrial sediment and potentially agricultural and other forms of land-based pollution. These inputs, combined with low flushing from the Caribbean Sea (Carruthers et al. 2005), result in prolonged residence times and exposures of resident organisms to sediments, nutrients and potential pollutants. Most of the towns and villages in the region are small, but they are built directly beside the water, with little or no treatment for sewage or other inputs.

Fig. 4: Abundance of the five most dominant sponge species at the three sites in Bocas del Toro, Panama. Values are total numbers of individuals in 15 1-m2 quadrats at each site.

Table 2: Comparison of sponge diversity, abundance and community indices from 15 1-m2 quadrats at each of the three sites in Bocas del Toro, Panama. Values are means with standard errors in parentheses. Degrees of freedom (df), H and P values are from Kruskal-Wallis tests. Superscripts indicate groups that differ significantly from each other, based on post-hoc pairwise Mann-Whitney U tests at P < 0.05. Punta Caracol Number of species/quadrat Total species Number of individuals/quadrat Total individuals J (evenness) H (diversity) 10.8 (3.5)a 36 40 (23.8) a 600 0.85 (0.05) a 1.99 (0.27) a Casa Blanca 14.6 (4.1) b 39 90.7 (39.5) b 1360 0.84 (0.04) a 2.21 (0.31) b Saigon 7.8 (2.6) c 31 38.1 (19.2) a 572 0.68 (0.15) b 1.39 (0.44) c df 2 2 2 2 H 17.937 17.267 15.723 24.230 P 0.0001 0.0002 0.0004 < 0.0001

340

Table 3: Comparison of nutrient concentrations (in micromolar = M), counts of fecal coliform colonies on culture plates, and PAH concentrations (in parts per million = ppm) from replicate water samples (n = 3) at the three sites in Bocas del Toro, Panama. Values are means with standard errors in parentheses. n.d. = not detectable. Degrees of freedom (df), H and P values are from Kruskal-Wallis tests. Superscripts indicate groups that differ significantly from each other, based on post-hoc pairwise Mann-Whitney U tests at P < 0.05. Punta Caracol Nutrients (M) Nitrate (November) Nitrite (November) Phosphate (November) Nitrate (July) Ammonia (July) Phosphate (July) Fecal Coliform Counts PAHs (ppm) phenanthrene anthracene benz[a]anthracene chrysene benzo[b]fluoranthene benzo[k]fluoranthene benzo[a]pyrene indeno[1,2,3-cd]pyrene dibenzo[a,h]anthracene benzo[ghi]perylene 0.40 (0.013) 0.25 (0.012)a 0.61 (0.017)a 0.22 (0.053) n.d. 0.81 (0.21) 12.67 (7.22) 0.025 (0)a 0.035 (0)a 0.018 (0.018) 0.015 (0.015) n.d. n.d. n.d. 0.02 (0.02) n.d. n.d. Casa Blanca 0.43 (0) 0.16 (0.012)b 0.52 (0.015)b 0.16 (0) n.d. 1.33 (1.33) 1.75 (1.75) n.d.b n.d.b 0.023 (0.023) 0.02 (0.02) 0.018 (0.018) 0.023 (0.023) 0.027 (0.027) 0.027 (0.027) 0.003 (0.003) 0.005 (0.005) Saigon 0.41 (0.017) 0.25 (0.009)a 0.56 (0.01)ab 0.22 (0.053) 0.057 (0.057) 0.14 (0.14) 8.0 (4.16) 0.008 (0.008)ab 0.012 (0.012)ab 0.018 (0.018) 0.015 (0.015) n.d. n.d. n.d. n.d. 0.02 (0.02) 0.01 (0.01) df 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 H 1.787 5.744 6.330 1.143 2.000 3.840 2.617 5.600 5.600 0.095 0.095 2.000 2.000 2.000 1.167 1.167 1.167 P .4093 .0506 .0422 .5647 .3379 .1566 .2702 .0608 .0608 .9535 .9535 .3679 .3679 .3679 .5580 .5580 .5580

region at present are probably adapted to a combination of these natural and anthropogenic environmental stressors.

Surveys
The sponge fauna on the reefs and mangrove habitats in Bocas del Toro, Panama is estimated at approximately 120 species (Diaz 2005). We found 51 species of sponges in our quadrats at three coral reef sites within this region. In general, the species found in Bocas del Toro and in our quadrats are widely distributed throughout the Caribbean (Diaz 2005). As seen in other studies (Diaz 2005), while many species were found at all three sites, one third of all species were observed at only a single site. Niphates erecta, for example, was one of the five most dominant species at all of our sites, and Diaz (2005) also found this species to be common at all 14 sites surveyed in both reef and mangrove habitats in Bocas del Toro. By contrast, Hymeniacidon sp. was dominant at Saigon but virtually absent from our quadrats at the other two sites, whereas Chondrilla nucula and Neopetrosia subtriangularis were only dominant at Casa Blanca. Saigon had lower numbers of sponge species per quadrat and overall, as well as lower community evenness and diversity compared to Punta Caracol and Casa Blanca. There is a widely-observed trend towards species-rich, highly diverse communities at environmentally healthy sites (Birkeland 1997, Bertness et al. 2001), suggesting that Saigon may be exposed to a higher degree of stress than the other two sites. Organic and inorganic pollution have been demonstrated to reduce sponge species diversity in Cuba (Alcolado and Herrera 1987), Brazil (Muricy et al. 1991, Monteiro and Muricy 2004) and France (Muricy 1991). By contrast, organic pollution increased cover and abundance of sponges in Colombia (Zea 1994), possibly because sponges are more

Fig. 5: Prevalence (percent of sponges with lesions) of Aplysina Red Band Syndrome on Aplysina cauliformis and A. fulva at each of the three sites in Bocas del Toro, Panama. Points represent mean ( standard error) prevalence along three 10 x 1 m transects. Total number of Aplysina spp. sponges surveyed along the transects is shown in parentheses.

Carruthers et al. (2005) examined lagoonal scale processes in Bocas del Toro, and although Baha Almirante had less terriginous input than nearby Laguna de Chiriqui, the entire region was found to have high levels of nutrients and sediment. Aronson et al. (2004) reported a phase shift from Poritesdominated to Agaricia-dominated reefs in Baha Almirante since the 1970s, and attributed this primarily to changes in water quality. Thus, the coral reef communities found in this

341

tolerant of high turbidity than other reef macro-organisms. Muricy (1989) found reduced numbers of sponge species, species abundance and dominance, percent cover, density of individuals and species diversity indices at an organically polluted bay in Brazil, as compared to two nearby clean sites, and Alcolado (1994) found that organically polluted sites near Havana, Cuba, had lower heterogeneity and diversity than unpolluted sites. Similar trends were obtained at the site of a nuclear power plant discharge in Brazil (Vilanova et al. 2004), where the main stressors included thermal pollution, chlorine and high water flow. In the present study, certain community characters, such as overall number of sponges and number of sponges per quadrat, were similar between Saigon and Punta Caracol. These data suggest either that community indices alone cannot be used to indicate ecological stress or that environmental conditions at Saigon and Punta Caracol are more similar to each other than at Casa Blanca, even though the latter site is located in between the other two. In addition to changes in diversity and abundance, sponge community composition and species dominance often differ in response to environmental stress. In the present study, species composition was more similar among the two upstream sites than between those two sites and Saigon, further suggesting that Saigon is exposed to different environmental factors than the other two sites. In addition, species that were dominant at Saigon were rare or absent from the other two sites and vice versa. Among these, Chondrilla nucula, is known to be sensitive to stress (Alcolado and Herrera 1987, Muricy 1989, Vilanova et al. 2004), and was completely absent from the reef at Saigon. However, C. nucula was also not common at Punta Caracol and could easily have been missed in surveys. The absence of this species may not be sufficient to indicate a stressed site, but its presence likely indicates cleaner water. Neopetrosia subtriangularis was only found at Casa Blanca and may also be indicative of more favorable water conditions. Aplysina cauliformis and A. fulva were common at both upstream sites and absent or rare at Saigon. Preliminary experiments have indicated that A. cauliformis is highly sensitive to nutrient enrichment (D. Gochfeld, unpubl. data: Bahamas 2006), suggesting at least one possible stressor that could account for the differences among sites. In contrast, Hymeniacidon sp. occurred predominantly at Saigon. Previous studies have demonstrated that the degree of dominance by particular sponge species can also differ among sites, with stressed sites typically dominated by fewer species (Alcolado and Herrera 1987, Muricy 1991, Alcolado 1994, Monteiro and Muricy 2004). At Saigon, the sponge community was dominated by Hymeniacidon sp., with Mycale laevis a distant second and other species dropping off numerically thereafter. By contrast, although Chondrilla nucula was a clear dominant species at Casa Blanca, four other species were each represented by > 90 individuals at that site. The sponge community at Punta Caracol was much more even, with the five most dominant species each represented by fewer than 100 individuals. Syndromes or diseases of sponges were present at all three sites. In addition to ARBS, other types of lesions were observed on Iotrochota birotulata, Amphimedon compressa, Plakortis angulospiculatus and on one large colony of Neofibularia nolitangere. It is not possible to identify the causes of these

lesions in the absence of any active signs of disease without detailed microbiological and microscopic study. Aside from these observations, ARBS was the main syndrome observed. ARBS was present at all three sites, but disease prevalence was highest at Saigon, which also had the lowest sponge diversity. However, this result is confounded by the fact that the affected sponges, Aplysina cauliformis and A. fulva were very rare on the reef at Saigon, indicating the need for further study on the dynamics of this sponge disease.

Environmental parameters
All three sites may be affected by specific stressors that have not yet been identified. We hypothesized that nutrients and fecal coliform bacteria from sewage, and PAHs from road runoff could be likely candidate pollutants at these sites, but these have not yet been fully characterized. Sewage effluent can result in turbidity, eutrophication, pathogenic microbes and pharmaceuticals in the environment, and is known to cause changes in the structure and distribution of many biological communities, including those dominated by sponges (Rose and Risk 1985, Muricy 1989, Ward-Paige et al. 2005). In an effort to characterize the mechanisms by which sewage effluent affects these organisms, Roberts et al. (2006) found declines in growth, reproduction and chlorophyll a concentrations in the phototrophic sponge Cymbastela concentrica when exposed to shade, silt and salinity gradients, but not to nutrients alone. Muricy (1989) found a significant correlation between several pollution parameters (water transparency, oil and coliform levels) and community indices, identifying Mycale microsigmatosa as a species that is tolerant of variable conditions, Ulosa ruetzleri and Amphimedon viridis as low-sensitivity species, and Aplysina fistularis, Tedania ignis, Chondrilla nucula and Polymastia sp. as highly sensitive species due to their total or near absence at highly polluted sites. Certain species, such as Scopalina ruetzleri exhibited an increase in the proximity of runoff in Colombia (Zea 1994), but were less abundant at polluted sites in other studies (Alcolado and Herrera 1987, Muricy 1989), a situation that may reflect the variable physiology of these sponges or differences in overall levels or types of stress (Muricy et al. 1991, Zea 1994). Sponges are known to exhibit specific responses to other types of pollution. For example, Spongia officinalis can metabolize certain polychlorobiphenyl contaminants (Perez et al. 2003). Sponges can also concentrate metals from their environment (Cebrian et al. 2003, Perez et al. 2004), and have been found to produce metallothionein-like proteins that may serve as useful biomarkers (Berthet et al. 2005). To date, no studies have elucidated specific sponge responses to PAHs, but PAHs are known to induce phototoxicity (Peachey and Crosby 1995) and reduced reproduction (Guzman and Holst 1993) in corals. Since we observed significant differences in sponge community structure among sites that appear to have similar, and relatively low, abundances of corals and algae, we believe that sponges may serve as sensitive bioindicators of natural and anthropogenic impacts on these reef communities. Based upon their restricted distributions, several sponge species are promising candidates as bioindicators in the Bocas del Toro

342

region. Our measurements of potential pollutants represent a snapshot in time at sites that exhibit a high degree of temporal variability in nutrients and other water quality parameters (P. Gondola, pers. comm.). The combination of elevated nutrient levels, fecal coliform bacteria and several PAH compounds, indicates that pollution from storm-water runoff is a concern at all three of the sites studied. Clearly, a further analysis of natural and anthropogenic stressors would aid in identifying the specific causes of the differences in sponge community structure and disease prevalence observed among these sites. Nonetheless, in this system, sponges appear to be more sensitive bioindicators than other components of these coral reefs.

Acknowledgements
This project was undertaken as part of the Taxonomy and Ecology of Caribbean Sponges course at the Smithsonian Tropical Research Institutes Bocas del Toro Research Station in 2005. We thank Maria Cristina Diaz and Guilherme Muricy for help with sponge identification and Rachel Collin and Gabriel Jacome for facilitating this research. Plinio Gondola performed the initial nutrient analyses. Julie Olson provided advice on fecal coliform assays. Lauren Wheeler, Kristie Willett and Cammi Hickman assisted with the PAH analyses. Marc Slattery and two anonymous reviewers offered helpful comments on the manuscript. Funding for D.J.G. was provided by the National Institute of Undersea Science and Technology under Grant Number NA16RU1496. Funding for R.W.T. was provided by the Smithsonian Tropical Research Institute and by the National Science Foundation under Grant Number 0209329.

References
Alcolado P (1994) General trends in coral reef sponge communities of Cuba. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 251-255 Alcolado P, Herrera A (1987) Effectos de la contaminacion sobre las comunidades de esponjas en el litoral de la Habana, Cuba. Rep Inv Inst Oceanol 68: 1-17 Aronson RB, Macintyre IG, Wapnick CM, ONeill MW (2004) Phase shifts, alternative states, and the unprecedented convergence of two reef systems. Ecology 85: 1876-1891 Bertness, MD, Gaines SD, Hay ME (2001) Marine community ecology. Sinauer Associates, Sunderland Berthet B, Mouneyrac C, Perez T, Amiard-Triquet C (2005) Metallothionein concentration in sponges (Spongia officionalis) as a biomarker of metal contamination. Comp Biochem Physiol C 141: 306-313 Birkeland C (1997) Life and death of coral reefs. Chapman and Hall, New York. Cebrian E, Marti R, Uriz JM, Turon X (2003) Sublethal effects of contamination on the Mediterranean sponge Crambe crambe: metal accumulation and biological responses. Mar Pollut Bull 46: 1273-1284 Carruthers TJB, Barnes PAG, Jacome GE, Fourqurean JW (2005) Lagoon scale processes in a coastally influenced Caribbean system: implications for the seagrass Thalassia testudinum. Carib J Sci 41: 441-455

Clarke KR, Warwick RM (1994) Change in marine communities: an approach to statistical analysis and interpretation. Natural Environment Research Council, Plymouth Collin R, Diaz MC, Norenburg J, Rocha RM, Sanchez JA, Schulze A, Schwartz M, Valdes A (2005) Photographic identification guide to some common marine invertebrates of Bocas Del Toro, Panama. Carib J Sci 41: 638-707 Colwell RK (2004) EstimateS: Statistical estimation of species richness and shared species from samples. Version 7. Users Guide and application published at: http: //purl.oclc.org/estimates DCroz L, Del Rosario JB, Gondola P (2005) The effect of fresh water runoff on the distribution of dissolved inorganic nutrients and plankton in the Bocas del Toro Archipelago, Caribbean Panama. Carib J Sci 41: 414-429 Diaz MC (2005) Common sponges from shallow marine habitats from Bocas del Toro region, Panama. Carib J Sci 41: 465-475 Diaz MC, Rtzler K (2001) Sponges: an essential component of Caribbean coral reefs. Bull Mar Sci 69: 535-546 Dubinsky Z, Stambler N (1996) Marine pollution and coral reefs. Global Change Biol 2: 511-526 Fabricius KE (2005) Effects of terrestrial runoff on the ecology of corals and coral reefs: review and synthesis. Mar Poll Bull 50: 125-146 Google Earth (2006) http://earth.google.com (accessed on 18 September 2006) Green EP, Bruckner AW (2000) The significance of coral disease epizootiology for coral reef conservation. Biol Conserv 96: 347361 Guzman HM, Holst I (1993) Effects of chronic oil-sediment pollution on the reproduction of the Caribbean reef coral Siderastrea siderea. Mar Pollut Bull 26: 276-282 Harvell CD, Kim K, Burkholder JM, Colwell RR, Epstein PR, Grimes DJ, Hofman EE, Lipp EK, Osterhaus ADME, Overstreet RM, Porter JW, Smith GW, Vasta GR (1999) Emerging marine diseases: climate links and anthropogenic factors. Science 285: 1505-1510 Hughes TP (1994) Catastrophes, phase shifts and large scale degradation of a Caribbean coral reef. Science 265: 1547-1549 Hughes TP, Baird AH, Bellwood DR, Card M, Connolly SR, Folke C, Grosberg R, Hoegh-Guldberg O, Jackson JBC, Kleypas J, Lough JM, Marshall P, Nystrm N, Palumbi SR, Pandolfi JM, Rosen B, Roughgarden J (2003) Human impacts and the resilience of coral reefs. Science 301: 929-933 Jackson JBC (1997) Reefs since Columbus. Coral Reefs 16(Suppl): 23-32 Kaczmarsky LT, Draud M, Williams EH (2005) Is there a relationship between proximity to sewage effluent and the prevalence of coral disease? Carib J Sci 41: 124-137 Kaufmann KW, Thompson RC (2005) Water temperature variation and the meteorological and hydrographic environment of Bocas del Toro, Panama. Carib J Sci 41: 392-413 Kuntz NM, Kline DI, Sandin SA, Rohwer F (2005) Pathologies and mortality rates caused by organic carbon and nutrient stressors in three Caribbean coral species. Mar Ecol Prog Ser 294: 173-180 Lauckner G (1980) Diseases of Porifera. In: Kinne O (ed). Diseases of marine animals. Wiley & Sons, Chichester. pp. 139-165 Magurran AE (1988) Ecological diversity and its measurement. Princeton University Press, New Jersey

343

Monteiro LC, Muricy G (2004) Patterns of sponge distribution in Cagarras Archipelago, Rio de Janeiro, Brazil. J Mar Biol Ass UK 84: 681-687 Mumby PJ, Hedley JD, Zychaluk K, Harborne AR, Blackwell PG (2006) Revisiting the catastrophic die-off of the urchin Diadema antillarum on Caribbean coral reefs: fresh insights on resilience from a simulation model. Ecol Model 196: 131-148 Muricy G (1989) Sponges as pollution-biomonitors at Arraial do Cabo, Southeastern Brazil. Rev Bras Biol 49(2): 347-354 Muricy G (1991) Structure of sponge communities around the sewage outfall at Cortiou (Marseille, France). Vie Milieu 41: 205221 Muricy G, Hajdu E, Custdio MR, Klautau M, Russo C, Peixinho S (1991) Sponge distribution at Arraial do Cabo, SE Brazil. In: Magoon OT, Converse H, Tippie V, Tobin LT, Clark D (eds). Coastal Zone 91. Proc VII Symp Coast Ocean Manag. ASCE Publ., Long Beach. 2: 1183-1196 Olson JB, Gochfeld DJ, Slattery M (2006) Aplysina red band syndrome: a new threat to Caribbean sponges. Dis Aquat Org 71: 163-168 Peachey RL, Crosby DG (1995) Phototoxicity in a coral reef flat community. In: Gulko D, Jokiel PL (eds). Ultraviolet radiation and coral reefs. Hawaii Institute of Marine Biology, Honolulu. pp. 193-200 Perez T, Wafo E, Fourt M, Vacelet J (2003) Marine sponges as biomonitor of polychlorobiphenyl contamination: concentration and fate of 24 congeners. Environ Sci Technol 37: 2152-2158 Perez T, Vacelet J, Rebouillon P (2004) In situ comparative study of several Mediterranean sponges as potential biomonitors of heavy metals. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Bull Mus Ist Biol Univ Genova 68: 517-525 Porter JW, Meier OW (1992) Quantification of loss and change in Floridian reef coral populations. Am Zool 32: 625-640 Pronzato R, Bavestrello G, Cerrano C, Magnino G, Manconi R, Pantelis J, Sara A, Sidri M (1999) Sponge farming in the Mediterranean Sea: new perspectives. Memoir Queensl Mus 44: 485-491 Roberts DE, Davis AR, Cummins SP (2006) Experimental manipulation of shade, silt, nutrients and salinity on the temperate reef sponge Cymbastela concentrica. Mar Ecol Prog Ser 307: 143154 Rose CS, Risk MJ (1985) Increase in Cliona delitrix infestation of Montastrea cavernosa heads on an organically polluted portion of the Grand Cayman fringing reef. PSZN Mar Ecol 6: 345-363 Rosenberg E, Loya Y (2004) Coral health and disease. Springer, Berlin

Rtzler K (1988) Mangrove sponge disease induced by cyanobacterial symbionts: failure of a primitive immune system? Dis Aquat Org 5: 143-149 Rtzler K (2004) Sponges on coral reefs: a community shaped by competitive cooperation. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Bull Mus Ist Biol Univ Genova 68: 85-148 Sokal RR, Rohlf FJ (1995) Biometry, 3rd ed. WH Freeman & Company, New York Sutherland KP, Porter JW, Torres C (2004) Disease and immunity in Caribbean and Indo-Pacific zooxanthellate corals. Mar Ecol Prog Ser 266: 273-302 Vacelet J, Vacelet E, Gaino E, Gallissian MG (1994) Bacterial attack of sponge skeleton during the 1986-1990 Mediterranean sponge disease. In: van Soest RWM, van Kempen TMG, Braekman JCV (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 355-362 Vilanova E, Mayer-Pinto M, Curbelo-Fernandez MP, Goncalves da Silva SH (2004) The impact of a nuclear power plant discharge on the sponge community of a tropical bay (SE Brazil). In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Bull Mus Ist Biol Univ Genova 68: 647-654 Wade TL, Brooks JM, Kennicutt II MC, McDonald TJ, Sericano JL, Jackson TJ (2003) GERG trace organics contaminant analytical techniques. In: Lauenstein GG, Cantillo AY (eds). Sampling and Analytical Methods of the National Status and Trends Program National Benthic Surveillance and Mussel Watch Projects 19841992. Volume IV. NOAA Technical Memorandum NOS ORCA 71 Ward-Paige CA, Risk MJ, Sherwood OW, Jaap WC (2005) Clionid sponge surveys on the Florida Reef Tract suggest land-based nutrient inputs. Mar Pollut Bull 51: 570-579 Webster NS, Negri AP, Webb RI, Hill RT (2002) A spongin-boring a-proteobacterium is the etiological agent of disease in the Great Barrier Reef sponge Rhopaloeides odorabile. Mar Ecol Prog Ser 232: 305-309 Weil, E (2004) Coral reef diseases in the wider Caribbean. In: Rosenberg E, Loya Y (eds). Coral Health and Disease. Springer, Berlin. pp. 35-68 Wulff JL (2001) Assessing and monitoring coral reef sponges: why and how? Bull Mar Sci 69: 831-846 Wulff JL (2006) A simple model of growth form-dependent recovery from disease in coral reef sponges, and implications for monitoring. Coral Reefs 25: 419-426 Zea S (1994) Patterns of sponge and coral abundance in stressed coral reefs at Santa Marta, Columbian Caribbean. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 257-264

Porifera research: Biodiversity, innovation and sustainaBility - 2007

345

Basal apparatus formation in external flagellated cells of Halisarca dujardini larvae (Demospongiae: Halisarcida) in the course of embryonic development
Elisaveta L. Gonobobleva
Department of Embryology, Biological Faculty, St. Petersburg State University, Universitetskaya nab. 7/9, St. Petersburg 199034, Russia. gonobol@pochtamt.ru Abstract: The basal apparatus in external flagellated cells of embryos of the sponge Halisarca dujardini (Demospongiae, Halisarcida) was examined by transmission electron microscopy and compared with the basal apparatus of external flagellated cells in swimming larva. Formation of the basal apparatus begins in the embryos consisting of ~128 cells. At this stage of development nascent basal body localized in the apical part of a cell and is in contact with a nucleus and an external membrane. Flagella, rootlets and basal foot are absent. Golgi complex is in the perinuclear cytoplasm but its association with a nascent basal body is not observed. Flagella appear in the embryos consisting of ~500 cells. From this stage and during all subsequent embryonic development the basal apparatus consists from the basal body and accessory centriole; the pair short cross-striated rootlets descend from the basal body. On a lateral surface of the basal body there is a stalked basal foot. The Golgi complex is in association with the basal apparatus. In the external flagellated cells of swimming larvae the basal apparatus structure changes. From the basal body descend long fibrous rootlets. Simple basal foot arises from the lateral part of the basal body. The basal apparatus structure differs in flagellated cells of the embryos and those of swimming larva. It can be connected with the proliferate activity of the embryonic cells and a degree of a cellular differentiation. Presence of cross-striation rootlets in embryonic flagellated cells is considered as a functional attribute. Keywords: Sponge, Halisarca, embryonic development, larva, flagellated cells, basal apparatus

Introduction
In the flagellated epithelia of metazoans, flagella are anchored to the cell surface by a basal apparatus. It consists of the basal body and several accessory structures. Transitional fibrils serve to tether the basal body into the apical cell surface. The basal foot is oriented in the direction of the effective stroke of flagella beating and also serves as microtubule organizing centre (Hagiwara et al. 2000). Although the precise function of the longitudinal or oblique rootlets remain largely unknown they are thought to anchor the basal body and flagella in the cell by interaction with cytosceletal components and form structural basis for effective and coordinate work of flagella. Rootlets also can serve as nucleus-basal body connectors (Holley 1984, Anstrom 1992, Hagiwara et al. 2000). Besides locomotion, the basal body and so the accessory centriole take part in a cell division as microtubule organizing centers. Formation of the flagella basal apparatus during epithelial cells differentiation is divided into the following four stages: (1) duplication of centrioles; (2) migration of centrioles to the apical cell surface to become basal bodies; (3) elongation of flagella containing the axoneme; (4)

formation of accessory structures of basal bodies (Hagiwara et al. 1997). Flagella apparatus formation in unicellular and multicellular organisms is associated with cell proliferation and differentiation. Detailed studies on unicellular flagellates show that in an actively proliferating population elements of the basal apparatus may be formed incompletely (Beech et al. 1991). Differences in rootlet structure at various stages of cell differentiation and epithelium proliferation were also shown in multicellular animals during ciliogenesis (Hagiwara et al. 1997, Domaratskii and Onishchenko 2002). These differences may be associated with specific relations between the cytoskeletal elements of the basal apparatus and the organelles at various stages of cell differentiation (Hagiwara et al. 1997). In the present paper, we describe the ultrastructure of the basal apparatus of the external flagellated cells of embryos at different stages of embryonic development and swimming larva. Segregation and differentiation of the external cells layer occur in the course of embryonic development. In Halisarca dujardini, segregation of the external cells layer takes place during regular cleavage and is determined by the position of the blastomeres and the orientation of cleavage

346

Fig. 1: Schematic diagram of the successive stages of Halisarca dujardini development. A. Cleavage (~16 blastomeres). B. An embryo (~128 - 256 cells). C. An embryo (~1000 cells). D. Prelarva. E. Larva (disphaerula). External cells of the embryos and larva are allocated.

spindles (Ereskovsky and Gonobobleva 2000, Gonobobleva and Ereskovsky 2004, Ereskovsky 2004). After segregation of the external layer, its cells proceeds to divide mitotically throughout the embryogenesis, cleavage furrows of external cells being radial (see Fig. 1). Flagellated cells are differentiated within the integrated and polarized layer. According to our data, the flagella basal apparatus starts to form in embryos consisting of 64-128 cells. Apicobasal polarity of the external cells is established during this period. At this stage of embryogenesis the nascent basal body present in apical part of cytoplasm. At the subsequent stages of the development flagellum and accessory structures of the basal body are formed (transition zone, basal foot and striated rootlets). Basal apparatus of flagellated cells of larvae differs from that in embryonic cells. Difference occurs in structure of oblique rootlets and basal foot. The ultrastructure of the basal apparatus is variable and is traditionally used in phylogenetic and systematic constructions (Karpov 1990, Woollacott and Pinto 1995, ). In the same time, the structure of flagella basal apparatus and associated organelles depends upon a functional condition of a cell or epithelium and can vary depending on type of epithelium and the nature of medium they propel (water or mucus) (Holley 1984, Willmer 1991). In H. dujardini definitive formation of the basal apparatus coincides with the completion of embryonic development and a release of the larva from the maternal sponge, when flagellated cells starts to execute the locomotion. It specifies that the difference in basal apparatus ultrastructure in embryonic and larval flagellated cells of H. dujardini have a functional explanation.

embedded in Epon-Araldite. Semi-thin sections were stained with methylene blue-borax. Ultrathin sections were contrasted with uranyl acetate and lead citrate and were examined with a JEM-100CX electron microscope.

Results
During cleavage (in embryos consisting of 2-32 cells) the nucleus is located centrally in the blastomeres, and the centriole is located in the perinuclear cytoplasm (Fig. 2A). In embryos consisting of 64-128-256 cells, the organelles of the external blastomeres are being gradually distributed along the cell axis. Finally, the nucleus appears in the apical portion of cytoplasm. The nascent basal body is situated between the nucleus and the apical plasma membrane perpendicular to the latter (Fig. 2B, C). The proximal and the distal end of the nascent basal body contact, respectively, the nucleus and the plasma membrane. Alar sheets are formed in the site of the contact with the cell membrane (Fig. 2C). Basal foot is observed on the lateral surface of the nascent basal body. Thin fibrous rootlets descend from the proximal end of the basal body connecting it with the apical surface of the nucleus (Fig. 2C). Golgi complexes are present in the perinuclear cytoplasm, mainly in its apical part (Fig. 2B, C). From the stage of about 1000 cells flagella are present. The basal apparatus has the following ultrastructural characteristics (Fig. 3A, B). The basal body is situated between the nucleus and the apical plasma membrane perpendicular to the latter. The accessory centriole may be oriented perpendicular to the basal body or with angle about 45. Mutual orientation of the accessory centriole was not investigated specially; we can
Fig. 2: Early stages embryonic development. A. Cleaving embryo (~16 cells). The nucleus occupies a central position in the blastomere, the centriole is located in the perinuclear cytoplasm. B. Longitudinal section of the nascent basal body demonstrates formation of the accessory structures (alar sheets and basal foot). C. Embryo at the stage of 128-256 cells. Longitudinal section through the apical region of external cell. Nascent basal body is located between the nucleus and the apical cell membrane. Short rootlets appear in proximal part of the basal body and contact with nucleus. Golgi apparatus are present in perinuclear cytoplasm (ag Golgi apparatus, as alar sheets, bb basal body, bf basal foot, c centriole, ec embryonic capsule, n nucleus, nu nucleolus, r rootlet, yg yolk granule. Scale bars: A, B 0.20 m; C 0.40 m).

Material and methods

Reproducing individuals of Halisarca dujardini Johnston, 1842 (Demospongiae, Halisarcida) were collected in the Chupa Inlet near the Srednii Island (Kandalaksha bay, the White Sea) from a depth of 1.5-5 m in June-July 2000. For transmission electron microscopy, tissue fragments and larvae were prefixed in 1% OsO4 for 10 min and fixed in 2.5 % glutaraldehyde in phosphate buffer (pH 7.4) at room temperature for 1 h. After fixation, the tissue fragments ant larvae were washed in the phosphate buffer (pH 7.4) and postfixed in 1% OsO4 in phosphate buffer for 1 h. Samples were dehydrated through a graded ethanol series and

347

348

only say that it varies. In the ~1000 cells embryos the nucleus of the external flagellated cell has a spherical or ellipsoidal form. One Golgi complex is located in the apical part of perinuclear cytoplasm. There is a alar sheet at the border of the axoneme and the basal body. The basal foot, ~0.12 m length, is at a right angle to the lateral surface of the basal body. It is shaped like a mushroom with a stalk ~0.08 m in length and spherical cap ~0.05 m in diameter (Fig. 4A, B). A couple of oblique rootlets start from the proximal end of the basal body Their length varies in different cells from 0.3 to 0.7 m. The oblique rootlets are cross-striated with a period of 0.05 m. Their distal parts are in contact with the apical part of the nucleus (Fig. 3; Fig. 4). At the prelarva stage, the oblique rootlets have a higher electron density and a less pronounced cross-striation. One Golgi complex envelops apical part of the pyriform nucleus (Fig. 4A, B). In the external flagellated cells of swimming larvae, the basal body and the accessory centriole are oriented at an angle of about 45 (Fig. 5; Fig. 6A). Proximal part of accessory centriole is in contact with fibrous rootlet. The basal body is 0.5 m in length and 0.17 m in diameter. Alar sheets radiate from the basal body to the adjacent membrane. Spherically shaped basal foot is situated in the lateral surface on the posterior side of the basal body (Fig. 6A, C). Oblique rootlets, about 2.4 m in length, start from the proximal end of the basal body. They consist of a dense fibrillar material (Fig. 6A, B). They course along the nucleus surface between the nuclear membrane and cis-face of the Golgi complex. Nucleus is pyriform.

Discussion
We have shown that formation of the basal apparatus during embryonic development of Halisarca dujardini (Demospongiae, Halisarcida) includes following stages. 1 - the displacement of the centriole to the apical part of cytoplasm during the external cells polarization; 2 - transformation of the centriole into the basal body with the establishment of its contact both with the nucleus and the cell membrane and the appearance of accessory structures; 3 - flagella formation. The sequence of the events of the basal apparatus formation in H. dujardini is the same as in other multicellular animals (Hagiwara et al. 1997). Data on the structure of oblique rootlets in H. dujardini embryos and larvae show that only their proximal, crossstriated part is formed during embryogenesis. After the cells pass into the prolonged interphase and become functionally active, the proximal part of the rootlet is altered and its distal
Fig. 3: , . Embryo consisting of 1000-2000 cells. Longitudinal sections through the apical region of the external flagellated cells illustrate basal body with its alar sheets, striated rootlets and accessory centriole (ac accessory centriole, as alar sheet, bb basal body, n nucleus, nu nucleolus, r rootlets. Scale bars: A, B 0.20 m).

349

Fig. 4: A, B. Prelarva. Longitudinal sections of apical part of the external flagellated cells show the striated rootlets and stalked basal foot (ag Golgi apparatus, as alar sheet, bb basal body, bf basal foot, n nucleus, r rootlets. Scale bars: A, B 0.20 m).

part is formed. Differences in a rootlets structure at various stages of the cell differentiation and epithelium proliferation were also shown in multicellular animals during ciliogenesis (Hagiwara et al. 1997, Domaratskii and Onishchenko 2002). A short interphase period and basal apparatus participation in the formation of the microtubular cleavage spindle may cause an incomplete formation of the basal apparatus elements in H. dujardini embryos. It is unexpected, but the structure of the basal foot also differs in the embryonic and larval flagellated cells. On preliminary data, in flagellated cells of embryos and larvae of H. dujardini, the stalked basal foot can serve as the microtubule organizing center, and a simple basal foot can be connected with the fibrous component. These differences may be associated with specific relations between the cytoskeletal elements of the basal apparatus and the organelles at various stages of cell differentiation. There is a close association between basal apparatus, pyriform nucleus and a Golgi complex in flagellated cells of H. dujardini embryos and larvae. The rootlets course along one face of the nucleus between the nuclear membrane and the cis-face of the Golgi stack of cisternae. Position of a nucleus in apical part of a cell (in case of length of the nucleus is less than a half of long axis of a cell)

and this type of the association between the basal apparatus, Golgi complex and a nucleus is characteristic for the larval flagellated cells of the calcareous sponges (Galissian and Vacelet 1992, Amano and Hori 1992, Amano and Hori 2001), the Homoscleromorpha (flagellated cells of the posteriorlateral and the posterior zones of the larvae (Boury-Esnault et al. 2003)), and the oviparous sponges (Lvi 1956, Usher and Ereskovsky 2005). In all cases, the nucleus is situated immediately subjacent to the basal body. Cross-striated rootlets have been described in flagellated cells of the Calcarea larvae (Gallissian and Vacelet 1992, Amano and Hori 1992, Amano and Hori 2001), Homoscleromorphas larvae (Boury-Esnault et al. 2003) and embryonic cells of H. dujardini (present paper). It has been shown that cross-striated centrin-containing rootlets (system II type fibrous roots, according to Andersen et al. 1991) are a nucleus-basal body connector and in green algae they polarize the mitotic spindle (Beech et al. 1991, Wolfrum 1991, Brugerolle and Mignot 2003). Morphological data reveal significant similarity in an association of the basal apparatus with striated rootlets, nucleus and Golgi complex in larval flagellated cell of sponges and green algae (at which this structure is investigated most in details). Unfortunately,

350

Fig. 5: Diagram of the basal apparatus of the external flagellated cell of Halisarca dujardini larva (ac accessory centriole, ag Golgi apparatus, as alar sheets, bb basal body, bf basal foot, f flagella, g glycocalix, m mitochondria, n - nucleus).

Fig. 6: , B. Longitudinal sections in different planes of the external flagellated cells of the anterior hemisphere of swimming larva illustrate structure of the basal body with fibrous rootlets, simple basal foot and accessory centriole. C. Transverse section of the apical part of the external flagellated cells of swimming larva (ag Golgi apparatus, ac accessory centriole, as alar sheet, bb basal body, bf basal foot, f flagella, m mitochondria, n nucleus, r rootlets. Scale bars: A 0.4 m; B 0.15 m; C 0.20 m).

351

the data on the presence of calcium-modulated contractile protein centrin in striated rootlets of sponges are absent. The data on the basal apparatus in some sponge larvae demonstrate a high diversity of this structure within the class Demospongiae (Woollacott and Pinto 1995). Structure of basal apparatus in H. dujardini larval flagellated cells has its own characteristic features: two fibrous rootlets, simple basal foot, lateral and oblique orientation of accessory centriole. This structure can be considered as a new one in the flagellated cells of the sponge larvae.

Acknowledgments
I thank Dr S. M. Efremova and two anonymous referees for critical comments, which greatly improved the original manuscript, and Dr. V. V. Semenov for advice and help in TEM. This work was funded by the program RFBR 07-04-01703 and INTAS CIG 5721.

References
Andersen RA, Barr DJS, Lynn OH, Melconian M, Moestrup O, Sleigh MA (1991) Terminology and nomenclature of the cytoskeletal elements associated with the flagellar/ciliary apparatus in protists. Protoplasma 164: 1-8 Amano S, Hori I (1994) Metamorphosis of a demosponge. I. Cells and structure of swimming larva. Invertebr Reprod Dev 25: 193204 Amano S, Hori I (2001) Metamorphosis of coeloblastula perfomed by multipotential larval flagellated cells in the calcareous sponge Leucosolenia laxa. Biol Bull 200: 20-32 Anstrom JA (1992) Organization of the ciliary basal apparatus in embryonic cells of the sea urchin, Lytechinus pictus. Cell Tissue Res 269: 305-313 Beech PL, Heimann K, Melkonian M (1991) Development of the flagellar apparatus during the cell cycle in unicellular algae. Protoplasma 164: 23-37 Boury-Esnault N, Ereskovsky A, Bzac C, Tokina D (2003) Larval development in the Homoscleromorpha (Porifera, Demospongiae). Invertebr Biol 122: 187-202 Brugerolle G, Mignot JP (2003) The rhizoplast of chrysomonads, a basal body-nucleus connector that polarises the dividing spindle. Protoplasma 222: 13-21 Domaratskii KE, Onishchenko GE (2002) Formation of basal bodies in ciliary epithelium of molluscs Buccinum undatum L. and Lymnaea stagnalis L. Russ J Dev Biol 33(3): 151-157 Ereskovsky AV (2004) Polyaxial cleavage in sponges (Porifera): A new pattern of metazoan cleavage. Dokl Biol Sci 386(1-6): 472474

Ereskovsky AV, Gonobobleva EL (2000) New data on embryonic development of Halisarca dujardini Johnston, 1842 (Demospongiae, Halisarcida). Zoosystema 22: 355-368 Ereskovsky AV, Tokina DB (2004) Morphology and fine structure of the swimming larvae of Ircinia oros (Porifera, Demospongiae, Dictyoceratida). Invertebr Reprod Dev 45(2): 137-150 Gallissian M-F, Vacelet J (1992) Ultrastructure of the oocyte and embryo of the calcified sponge, Petrobiona massiliana (Porifera, Calcarea). Zoomorphology 112: 133-141 Gonobobleva E, Ereskovsky A (2004) Polymorphism in freeswimming larvae of Halisarca dujardini (Demospongiae, Halisarcida). In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Boll Mus Ist Biol Univ Genova 68: 349-356 Hagiwara H, Aoki T, Ohwada N, Fujimoto T (1997) Development of striated rootlets during ciliogenesis in the human oviduct epithelium. Cell Tissue Res 290: 39-42 Hagiwara H, Kano A, Aoki T, Ohwada N (2000) Immunocytochemistry of the striated rootlets associated with solitary cilia in human oviductal secretory cells. Histochem Cell Biol 114: 205-212 Holley MC (1984) The ciliary basal apparatus is adapted to the structure and mechanics of the epithelium. Tissue Cell 16(2): 287310 Lvi C (1956) tude des Halisarca de Roscoff. Embryologie et systematique des Dmosponges. Arch Zool Exp Gen 93: 3-181 Lvi C (1964) Ultrastructure de la larve parenchymella de Dmosponge. I. Mycale contarenii (Martens). Cah Biol Mar 5: 97-104 Karpov SA (1990) System of Protista. OSPI Press, Omsk Maldonado M (2004) Choanoflagellates, choanocytes, and animal multicellularity. Invertebr Biol 123(1): 1-22 Sukhodolskaya AN, Ivanova LV (1988) Fine structural investigation of the fresh-water sponge Spongilla lacustris swimming larvae. Tsitologiya 30(12): 1409-1417 Usher KM, Ereskovsky AV (2005) Larval development, ultrastructure and metamorphosis in Chondrilla australensis Carter, 1873 (Demospongiae, Chondrosida, Chondrillidae). Invertebr Reprod Dev 47: 51-62 Willmer P (1991) Invertebrate relationships. Patterns in animal evolution. Cambridge University Press, Cambridge Wolfrum U (1991) Centrin- and -actinin- like immunoreactivity in the ciliary rootlets of insect sensilla. Cell Tissue Res 266(2): 231-238 Woollacott RM, Pinto RL (1995) Flagellar basal apparatus and its utility in phylogenetic analyses of the Porifera. J Morphol 226: 247-265

Porifera research: Biodiversity, innovation and sustainaBility - 2007

353

Checklist of Brazilian deep-sea sponges

Eduardo Hajdu, Daniela A. Lopes
Museu Nacional, Departamento de Invertebrados, Universidade Federal do Rio de Janeiro, Quinta da Boa Vista, s/n, 20940040, Rio de Janeiro, RJ, Brazil. hajdu@acd.ufrj.br Abstract: A comprehensive revision of literature dealing with Brazilian marine sponges has been undertaken. One hundred and ten sponges were found to occur deeper than 100 m, and are here considered deep-sea sponges. These species were reported upon in 18 papers, starting from Polejaeff (1884) on the Keratosa collected by the H.M.S. Challenger. For each deep-sea species, its distribution in terms of Brazilian states off which it has been found, its depth range, original literature and expedition, are given. Keywords: Brazil, REVIZEE

Introduction
The inventory of the Brazilian marine sponge fauna has been a concern solely of foreign expeditions and researchers until not too long ago (Hajdu et al. 1996). In the past two decades this scenario changed dramatically, and study of Brazilian marine sponge taxonomy is now a concern mainly of Brazilian scientists. Most Brazilian sponges reported upon by foreign researchers had been collected by dredging conducted by oceanographic expeditions, but nevertheless, deep sea sponges were not the main focus of these (e.g. Boury-Esnault 1973, Collette and Rtzler 1977). A few exceptions, although numerically not so important, were the Challenger and Albatross expeditions, which reported on 11 (Polejaff 1884, Ridley and Dendy 1887, Schulze 1887, Sollas 1888) and on one (Schulze 1899) species, respectively. The Challenger expedition is important for being for over a century the only effort at collecting Brazilian sponges deeper than 2000 m. Recently, the REVIZEE Programme (Programme for the Evaluation of the Sustainable Potential of Life Resources in the Brazilian EEZ) established a few collecting stations deeper than 2000 m, where some sponges were collected (Lavrado 2006, Muricy et al.2006 and unpubl. res.). This material still awaits taxonomic description. A checklist is an important first step to map what is known and where the main gaps are. From what is presented below, it is clear that gaps still predominate, and the 18 articles found to contain reports on Brazilian deep sea sponges are a small effort to change this picture. Noteworthy, an important proportion of recently found records still await increased taxonomic effort prior to reaching full species identifications. Such is the case of the large collections gathered by Programme REVIZEE, where from several family level morphotype identifications were left out of this compilation.

Materials and methods

A comprehensive review of all the literature dealing with marine sponges recorded from the Brazilian EEZ has been undertaken, and those records made for sponges collected deeper than 100 m listed in alphabetical order of accepted nomenclature. Where more than one deep sea record existed for the same species, these were compiled in separate and linked to the respective literature. The compiled list is also presented classified according to the Systema Porifera taxonomic scheme (Hooper and van Soest 2002), and grouped according to a latitudinal Brazilian states gradient. Abbreviations used are those of Brazilian states, viz. (from S to N) RS Rio Grande do Sul, SC Santa Catarina, SP So Paulo, RJ Rio de Janeiro, ES Esprito Santo, BA Bahia, AL Alagoas, PE Pernambuco, RN (ASPSP) Rio Grande do Norte (Saint Peters and Saint Pauls Archipelago) and CE Cear. The original specimens were not checked upon.

Results and discussion

One hundred and ten marine sponge species were found to have been reported so far from deeper than 100 m and from within the Brazilian EEZ (Tables 1 and 2). These were reported upon by Polejaff (1884, three species); Ridley and Dendy (1887, two species); Schulze (1887, two species; 1899, one species); Sollas (1888, four species); Boury-Esnault (1973, four species); Mothes (1977, one species); Hajdu and Desqueyroux-Fandez (1994, one species); Santos et al. (1999, eight species); Vacelet (1999, one species); Silva and Mothes (2000, three species); Hajdu et al. (2004 38 species); Lopes and Hajdu (2004, two species); Lopes et al. (2005, two species); Mothes et al. (2004, 19 species); Oliveira and Hajdu (2005, three species); Muricy et al. (2006, 29 species) and Menshenina et al. (2007, two species). The largest addition was that made by Programme REVIZEE, with 88 records of marine sponges for the Brazilian deep sea (Santos et al. 1999, Hajdu et al. 2004, Mothes et al. 2004, Lopes and Hajdu 2004,

354

Table 1: List of Brazilian deep-sea sponges in alphabetical order of currently accepted nomenclature. Species 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 Aaptos sp. Agelas clathrodes Schmidt, 1870 Agelas dispar Duchassing and Michelotti, 1864 Agelas schmidti Wilson, 1902 Agelas sp. Aiolochroia crassa (Hyatt, 1975) Alectona mesatlantica Vacelet, 1999 Aphrocallistes beatrix Gray, 1858 Aplysina archeri (Higgins, 1875) Aplysina cauliformis Carter, 1882 Aplysina cf. fulva (Pallas, 1766) Aplysina fulva (Pallas, 1766) Aplysina lacunosa (Pallas, 1766) Aplysina sp. Aplysina sp. nov. Aulospongus sp. Axinella sp. Bubaris sp. Cacospongia levis (Polejaeff, 1884) Characella aspera Sollas, 1888 Cinachyra sp. Cinachyra sp. nov. Cinachyrella aff. alloclada (Uliczka, 1929) Cinachyrella aff. apion (Uliczka, 1929) Cinachyrella aff. kuekenthali (Uliczka, 1929) Cinachyrella kuekenthali (Uliczka, 1929) Clathria (Clathria) sp. Cliona aff. celata Grant, 1826 Cliona sp. Corallistes typus (Schmidt, 1870) Craniella sp. Crella (Yvesia) sp. Crellomyxilla chilensis (Thiele, 1905) Dactylocalyx pumiceus (Stutchbury, 1814) Desmacella annexa (Schmidt, 1870) Desmacella aff. pumilio Schmidt, 1870 Desmacella sp. Dragmacidon sp. Erylus diminutus Mothes et al., 1999 Erylus soesti (Mothes and Lerner, 2001) Erylus sp. Esperiopsis bathyalis Lopes and Hajdu, 2004 Euplectella suberea Thomson, 1877 Forcepia sp. Gastrophanella sp. Geodia australis Silva and Mothes, 2000 Geodia neptuni (Sollas, 1886) Geodia riograndensis Silva and Mothes, 2000 Geodia splendida Silva and Mothes, 2000 Geodia sp. Grantia sp. Halichondria sp. Haliclona (Halichoclona) sp. Haliclona (Gellius) sp. Haliclona sp. Distribution
ES/108m ES/110m CE/116m ES/108m RJ/270m, ES/108m RJ/270m, ES/108, 125m, BA/100m RN (ASPSP)/2030m RJ/640m, ES/500m AL/731m RJ/270m ES/108-110m RJ/270m, CE/166m ES/108m ES/108m, BA/100m RJ/270m SC/360-366m RJ/270m RS/299, 360m, SP/153m AL/731m AL/640m SP/505m RJ/270m, ES/500m BA/100m BA/100m RJ/250-500m, BA/100m RJ/270m SP/157m SP/100m SP/157m AL/640m ES/773m SP/380m SP/100m RS/165-188m, ES/552m SP/153, 167m SP/133m SP/153m, RJ/114m, SC/350, 380m, RS/145-350m RJ/270m SC/366m, RS/145m SC/366m RJ/100, 270m, ES/125, 360m SP/808m PE/2928m SP/258m SP/167m RS/207m AL/634m RS/200-300m RS/520m SP/153m SP/153, 167m CE/166m, SP/153m RS/120-220m SP/380m CE/166m

Literature source/expedition or project Muricy et al. (2006)/ REVIZEE Muricy et al. (2006)/ REVIZEE Santos et al. (1999)/ REVIZEE Muricy et al. (2006)/ REVIZEE Muricy et al. (2006)/ REVIZEE Muricy et al. (2006)/ REVIZEE Vacelet (1999)/ Saint Paul Lopes et al. (2005), Muricy et al. (2006)/ REVIZEE Polejaff (1884, como A. tenuissima)/ Challenger Muricy et al. (2006)/ REVIZEE Muricy et al. (2006)/ REVIZEE Muricy et al. (2006), Santos et al. (1999)/ REVIZEE Muricy et al. (2006)/ REVIZEE Muricy et al. (2006)/ REVIZEE Muricy et al. (2006)/ REVIZEE Hajdu et al. (2004)/ REVIZEE Muricy et al. (2006)/ REVIZEE Hajdu et al. (2004), Mothes et al. (2004)/ REVIZEE Polejaff (1884)/ Challenger Sollas (1888)/ Challenger Hajdu et al. (2004)/ REVIZEE, Muricy et al. (2006)/ REVIZEE Muricy et al. (2006)/ REVIZEE Muricy et al. (2006)/ REVIZEE Muricy et al. (2006)/ REVIZEE Muricy et al. (2006)/ REVIZEE Hajdu et al. (2004)/ REVIZEE Boury-Esnault (1973)/ Calypso Hajdu et al. (2004)/ REVIZEE Sollas (1888)/ Challenger Muricy et al. (2006)/ REVIZEE Hajdu et al. (2004)/ REVIZEE Boury-Esnault (1973)/ Calypso Mothes-de-Moraes (1977), Lopes et al. (2005)/ REVIZEE Hajdu et al. (2004)/ REVIZEE Hajdu et al. (2004)/ REVIZEE Hajdu et al. (2004), Mothes et al. (2004), Muricy et al. (2006)/ REVIZEE Muricy et al. (2006)/ REVIZEE Hajdu et al. (2004), Mothes et al. (2004)/ REVIZEE Hajdu et al. (2004)/ REVIZEE Muricy et al. (2006)/ REVIZEE Lopes and Hajdu (2004)/ REVIZEE Schulze (1887)/ Challenger Hajdu et al. (2004)/ REVIZEE Hajdu et al. (2004)/ REVIZEE Silva and Mothes (2000)/ Projeto Talude Sollas (1888)/ Challenger Silva and Mothes (2000)/ Projeto Talude Silva and Mothes (2000)/ Projeto Talude Hajdu et al. (2004)/ REVIZEE Hajdu et al. (2004)/ REVIZEE Santos et al. (1999), Hajdu et al. (2004)/ REVIZEE Mothes et al. (2004)/ REVIZEE Hajdu et al. (2004)/ REVIZEE Santos et al. (1999)

355 Table 1 (cont.)

56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110

Halicometes minuta Sar and Rosa-Barbosa, 1995 Hamacantha microxifera Lopes and Hajdu, 2004 Hamacantha sp. 2 Hamacantha sp. 3 Hyalonema schmidti Schulze, 1899 Hyattella sp. Hymedesmia sp. 1 Hymedesmia sp. 2 Hymedesmia sp. 3 Hymedesmia sp. 4 Ircinia sp. Ircinia strobilina (Lamarck, 1816) Jaspis sp. Latrunculia sp. Leucosolenia sp. Lissodendoryx (Lissodendoryx) sp. Mycale beatrizae Hajdu and DesqueyrouxFandez, 1994 Lophocalyx sp. nov. 1 Lophocalyx sp. nov. 2 Mycale sp. 1 Mycale sp. 2 Myxilla (Ectyomyxilla) tenuissima (Thiele, 1905) Neofibularia sp. Niphates sp. Oceanapia sp. Pachastrella monilifera Schmidt, 1868 Pachastrissa sp. Pachypellina sp. Petromica sp. Phakellia connexiva Ridley and Dendy, 1887 Pheronema carpenteri Thomson, 1869 Plakina sp. Plakinastrella sp. Poecillastra sollasi (Topsent,1892) Polymastia corticata Ridley and Dendy, 1886 Polymastia sp. Raspaciona sp. Raspailia sp. Raspailia (Parasyringella) sp. Raspailia (Raspaxilla) phakellina Topsent, 1913 Rhabderemia besnardi Oliveira and Hajdu, 2005 Rhabderemia itajai Oliveira and Hajdu, 2005 Rhabderemia sp. Rhabderemia uruguaiensis van Soest and Hooper, 1993 Sceptrella sp. Sphaerotylus sp. Spongosorites sp. Stelletta sp. Suberites caminatus (Ridley and Dendy, 1886) Tedania (Tedaniopsis) vanhoffeni (Hentschel, 1914) Thenea fenestrata Sollas, 1886 Thrombus sp. Timea sp. Topsentia sp. Vulcanella sp.
* Family or lower level identifications were omitted

SP/133, 153m SP/167m SP/166m SP/167m CE/763m CE/166m SC/350m RS/165m SC/350m RS/165m CE/166m, SC/420m, RS/200m AL/731m RS/145, 299m, SC/366m SC/420m SP/380m SP/167m SP/136m ES/597-610m BA/1717m RS/280m SC/350m SP/500m RS/120, 145m CE/166m RJ/270m, ES/108, 110m SP/258m RJ/270m SP/258m RJ/270m AL/2196m AL/2928m ES/108m CE/166m, RJ/270m SP/417m PE/2196m SP/168m SP/153, 168m SC/144, 420m, RS/120-296m SP/167m SP/167, 380m SP/153m SC/380m RJ/270m RS/133-296m, SP/153, 167m SP/167m SC/420m RJ/270m SP/167m SC/100m SC/420m AL/3138m SP/417m SP/147m, RS/120, 145m SP/147m SP/167, 380m

Hajdu et al. (2004)/ REVIZEE Lopes and Hajdu (2004)/ REVIZEE Hajdu et al. (2004)/ REVIZEE Hajdu et al. (2004)/ REVIZEE Schulze (1899)/ Albatross Santos et al. (1999)/ REVIZEE Mothes et al. (2004)/ REVIZEE Mothes et al. (2004)/ REVIZEE Mothes et al. (2004)/ REVIZEE Mothes et al. (2004)/ REVIZEE Santos et al. (1999), Mothes et al. (2004)/ REVIZEE Polejaff (1884)/ Challenger Hajdu et al. (2004), Mothes et al. (2004)/ REVIZEE Mothes et al. (2004)/ REVIZEE Hajdu et al. (2004)/ REVIZEE Hajdu et al. (2004)/ REVIZEE Hajdu and Desqueyroux-Fandez (1994) Menshenina et al. (2007)/ Marion Dufresne Menshenina et al. (2007)/ REVIZEE Mothes et al. (2004)/ REVIZEE Mothes et al. (2004)/ REVIZEE Hajdu et al. (2004)/ REVIZEE Mothes et al. (2004)/ REVIZEE Santos et al. (1999)/ REVIZEE Muricy et al. (2006)/ REVIZEE Hajdu et al. (2004)/ REVIZEE Muricy et al. (2006)/ REVIZEE Hajdu et al. (2004)/ REVIZEE Muricy et al. (2006)/ REVIZEE Ridley and Dendy (1887)/ Challenger Schulze (1887)/ Challenger Muricy et al. (2006)/ REVIZEE Santos et al. (1999), Muricy et al. (2006)/ REVIZEE Hajdu et al. (2004)/ REVIZEE Ridley and Dendy (1887)/ Challenger Hajdu et al. (2004)/ REVIZEE Hajdu et al. (2004) / REVIZEE Mothes et al. (2004)/ REVIZEE Hajdu et al. (2004) / REVIZEE Hajdu et al. (2004) / REVIZEE Oliveira and Hajdu (2005) / REVIZEE Oliveira and Hajdu (2005) / REVIZEE Muricy et al. (2006)/ REVIZEE Hajdu et al. (2004), Mothes et al. (2004) , Oliveira and Hajdu (2005)/ REVIZEE Hajdu et al. (2004) / REVIZEE Mothes et al. (2004)/ REVIZEE Muricy et al. (2006)/ REVIZEE Hajdu et al. (2004) / REVIZEE Boury-Esnault (1973)/ Calypso Mothes et al. (2004)/ REVIZEE Sollas (1888)/ Challenger Hajdu et al. (2004) / REVIZEE Hajdu et al. (2004), Mothes et al. (2004)/ REVIZEE Hajdu et al. (2004) / REVIZEE Hajdu et al. (2004) / REVIZEE

356

Table 2: Classification of Porifera species reported for the Brazilian outer shelf (below 100 m depth) and slope. Phylum Porifera Grant, 1836 Class Calcarea Bowerbank, 1864 Order Leucosolenida Hartman, 1958 Family Leucosoleniidae Minchin, 1900 Grantia sp. Leucosolenia sp. Class Demospongiae Sollas, 1885 Order Homosclerophorida Dendy, 1905 Family Plakinidae Schulze, 1880 Plakina sp. Plakinastrella sp. Order Astrophorida Sollas, 1888 Family Ancorinidae Schmidt, 1870 Jaspis sp. Stelletta sp. Family Calthropellidae Lendenfeld, 1907 Pachastrissa sp. Family Geodiidae Gray, 1867 Erylus diminutus Mothes et al., 1999 Erylus soesti (Mothes and Lerner, 2001) Erylus sp. Geodia australis Silva and Mothes, 2000 Geodia neptuni (Sollas, 1886) Geodia riograndensis Silva and Mothes, 2000 Geodia splendida Silva and Mothes, 2000 Geodia sp. Family Pachastrellidae Carter, 1875 Characella aspera Sollas, 1888 Pachastrella monilifera Schmidt, 1868 Poecillastra sollasi (Topsent,1892) Thenea fenestrata Sollas, 1886 Thrombus sp. Vulcanella sp. Lithistida Family Corallistidae Sollas, 1888 Corallistes typus (Schmidt, 1870) Family Desmanthidae Topsent, 1894 Petromica sp. Family Siphonidiidae Lendenfeld, 1903 Gastrophanella sp. Order Spirophorida Bergquist and Hogg, 1969 Family Tetillidae Sollas, 1886 Cinachyra sp. Cinachyra sp. nov. Cinachyrella aff. alloclada (Uliczka, 1929) Cinachyrella aff. apion (Uliczka, 1929) Cinachyrella aff. kuekenthali (Uliczka, 1929) Cinachyrella kuekenthali (Uliczka, 1929) Craniella sp. Order Hadromerida Topsent, 1894 Family Alectonidae Rosell, 1996 Alectona mesatlantica Vacelet, 1999 Family Clionaidae dOrbigny, 1851 Cliona aff. celata Grant, 1826 Cliona sp. Family Polymastiidae Gray, 1867 Polymastia corticata Ridley and Dendy, 1886 Polymastia sp. Sphaerotylus sp. Family Suberitidae Schmidt, 1870 Aaptos sp. Suberites caminatus (Ridley and Dendy, 1886)

Family Tethyidae Gray, 1867 Halicometes minuta Sar and Rosa-Barbosa, 1995 Family Timeidae Topsent, 1928 Timea sp. Order Halichondrida Vosmaer, 1887 Family Axinellidae Carter, 1875 Axinella sp. Dragmacidon sp. Phakellia connexiva Ridley and Dendy, 1887 Family Bubaridae Topsent, 1894 Bubaris sp. Family Desmoxyidae Hallmann, 1917 Spongosorites sp. Family Halichondriidae Gray, 1867 Halichondria sp. Topsentia sp. Order Agelasida Hartman, 1980 Family Agelasidae Verril, 1907 Agelas clathrodes (Schmidt, 1870) Agelas dispar Duchassaing and Michelotti, 1864 Agelas schmidti Wilson, 1902 Agelas sp. Order Poecilosclerida Topsent, 1928 Family Coelosphaeridae Dendy, 1922 Forcepia sp. Lissodendoryx (Lissodendoryx) sp. Family Crellidae Dendy, 1922 Crella (Yvesia) sp. Family Desmacellidae Ridley and Dendy, 1886 Desmacella annexa (Schmidt, 1870) Desmacella aff. pumilio Schmidt, 1870 Desmacella sp. Neofibularia sp. Family Esperiopsidae Hentschel, 1923 Esperiopsis bathyalis Lopes and Hajdu, 2004 Family Hamacanthidae Gray, 1872 Hamacantha microxifera Lopes and Hajdu, 2004 Hamacantha sp. 2 Hamacantha sp. 3 Family Hymedesmiidae Topsent, 1928 Hymedesmia sp. 1 Hymedesmia sp. 2 Hymedesmia sp. 3 Hymedesmia sp. 4 Family Latrunculiidae Topsent, 1922 Latrunculia sp. Sceptrella sp. Family Microcionidae Carter, 1875 Clathria (Clathria) sp. Family Mycalidae Lundbeck, 1905 Mycale beatrizae Hajdu and Desqueyroux-Fandez, 1994 Mycale sp. 1 Mycale sp. 2 Family Myxillidae Dendy, 1922 Crellomyxilla chilensis (Thiele, 1905) Myxilla (Ectyomyxilla) tenuissima (Thiele, 1905) Family Raspailiidae Hentschel, 1923 Aulospongus sp. Raspaciona sp. Raspailia sp. Raspailia (Parasyringella) sp. Raspailia (Raspaxilla) phakellina Topsent, 1913 Family Rhabderemiidae Topsent, 1928 Rhabderemia besnardi Oliveira and Hajdu, 2005 Rhabderemia itajai Oliveira and Hajdu, 2005 Rhabderemia sp. Rhabderemia uruguaiensis van Soest and Hooper, 1993

357

Family Tedaniidae Ridley and Dendy, 1886 Tedania (Tedaniopsis) vanhoffeni (Hentschel, 1914) Order Haplosclerida Topsent, 1928 Family Chalinidae Gray, 1867 Haliclona (Gellius) sp. Haliclona (Halichoclona) sp. Haliclona sp. Pachypellina sp. Family Niphatidae van Soest, 1980 Niphates sp. Family Phloeodictyidae Carter, 1882 Oceanapia sp. Order Dictyoceratida Minchin, 1900 Family Irciniidae Gray, 1867 Ircinia sp. Ircinia strobilina (Lamarck, 1816) Family Spongiidae Gray, 1867 Hyattella sp. Family Thorectidae Bergquist, 1968 Cacospongia levis (Polejaff, 1884) Order Verongida Bergquist, 1978 Family Aplysinidae Carter, 1875 Aiolochroia crassa (Hyatt, 1875) Aplysina archeri (Higgins, 1875) ? Aplysina cauliformis Carter, 1882 Aplysina cf. fulva (Pallas, 1766) Aplysina fulva (Pallas, 1766) Aplysina lacunosa (Pallas, 1766) Aplysina sp. Aplysina sp. nov. Class Hexactinellida Schmidt, 1870 Order Hexactinosida Schrammen, 1903 Family Aphrocallistidae Gray, 1867 Aphrocallistes beatrix Gray, 1858 Family Dactylocalycidae Gray, 1867 Dactylocalyx pumiceus (Stutchbury, 1814) Order Lyssacinosida Zittel, 1877 Family Euplectellidae Gray, 1867 Euplectella suberea Thomson, 1877 Family Rossellidae Schulze, 1885 Lophocalyx sp. nov. 1 Lophocalyx sp. nov. 2 Order Amphidiscosida Schrammen, 1924 Family Hyalonematidae Gray, 1867 Hyalonema schmidti Schulze, 1889 Family Pheronematidae Gray, 1870 Pheronema carpenteri Thomson, 1869

Fig. 1: Bar graph illustrating the distribution of records of Brazilian deep-sea species per depth zone (depth in meters).

Lopes et al. 2005, Oliveira and Hajdu 2005, Muricy et al. 2006, Menshenina et al. 2007), and these are only a fraction of what sits on shelves awaiting deeper taxonomic study. It can be safely stated that the Brazilian deep sea sponge fauna was virtually unknown before this project. On the other hand, the observation that from the 127 accepted families of extant sponges (Hooper and van Soest 2002), only 47 (one of Calcarea, 40 of Demospongiae, six of Hexactinellida) were hitherto found among Brazilian deep sea sponges, is highly suggestive of the still fragmentary nature of the inventory. Figure 1 shows how skewed knowledge about deep-sea Brazilian sponges is. Nearly half the records made this far

originate from the outer continental shelf (100-200 m depth). Shelf break is usually shallower in the Brazilian NE, but the numbers of deep-sea species known from the area are small, so that the whole picture is not changed. The other half comprises a much larger proportion of species from the continental slope (63 spp.; 200-2000 m depth), and only 6 which may have originated from the continental rise (> 2000 m). The knowledge of the bathymetric distribution of Brazilian deep-sea sponges is rather meager, as revealed by the fact that among fully identified species, only four have ranges larger than 200 m (e.g. Cinachyrella aff. kuekenthali, 400 m depth range; Dactylocalyx pumiceus, 387 m depth range; Erylus diminutus, 221 m depth range; Raspailia phakellina, 213 m depth range). On top of this, proper descriptions are available only for 27 out of 110 species recorded, which is further evidence of the fragmentary nature of the knowledge available. Table 3 shows the 110 species compiled, organized according to their distribution in a latitudinal Brazilian states gradient. So Paulo state holds the largest number of published records of deep sea sponges, 40, 37 of which reported from material collected by Programme REVIZEE (Hajdu et al. 2004, Lopes and Hajdu 2004, Oliveira and Hajdu 2004). Next is a series of southern and south-eastern Brazilian states, viz. Rio Grande do Sul, Santa Catarina, Rio de Janeiro and Esprito Santo, all with comparable numbers of species (15-19 spp). The shallow waters of these and of So Paulo state belong in Palacios (1982) concept of a transitional biogeographic unit, the Paulista Province. This proposal is not universally accepted (e.g. Floeter and Soares-Gomes 1999), but the apparent large numbers of endemic shallow water sponge species seem to favor Palacios view (e.g. Lerner and Hajdu 2002). A provisional deep sea sponge endemic to this proposed province (including also Paran and Santa Catarina states) was known for some time already, Mycale beatrizae Hajdu and Desqueyroux-Fandez 1994. Many more are being described on the basis of recently collected material, viz. Esperiopsis bathyalis Lopes and Hajdu 2004, Hamacantha microxifera Lopes and Hajdu 2004, Rhabderemia besnardi Oliveira and Hajdu 2005 and R. itajai Oliveira and Hajdu 2005. From the little data available on deep sea sponge biodiversity in the area it is judged premature to postulate the existence of a transitional biogeographic unit also in the deep

358

Table 3: List of Brazilian deep-sea sponges known from each state in alphabetical order of currently accepted nomenclature. Brazilian states RS Deep sea sponges hitherto recorded Bubaris sp., Dactylocalyx pumiceus, Desmacella sp., Erylus diminutus, Geodia australis, Geodia riograndensis, Geodia splendida, Haliclona (Halichoclona), Hymedesmia sp. 2, Hymedesmia sp. 4, Ircinia sp., Jaspis sp., Mycale sp. 1, Neofibularia sp., Raspailia sp., Rhabderemia uruguaiensis, Timea sp. Aulospongus sp., Desmacella sp., Erylus diminutus, Erylus soesti, Hymedesmia sp. 1, Hymedesmia sp. 3, Ircinia sp., Jaspis sp., Latrunculia sp., Mycale sp. 2, Raspailia sp., Rhabderemia itajai, Sphaerotylus sp., Suberites caminatus, Tedania (Tedaniopsis) vanhoffeni Bubaris sp., Cinachyra sp., Clathria (Clathria) sp., Cliona aff. celata, Cliona sp., Crella (Yvesia) sp., Crellomyxilla chilensis, Desmacella aff. pumilio, Desmacella annexa, Desmacella sp., Esperiopsis bathyalis, Forcepia sp., Gastrophanella sp., Geodia sp., Grantia sp., Halichondria sp., Haliclona (Gellius) sp., Halicometes minuta, Hamacantha microxifera, Hamacantha sp. 2, Hamacantha sp. 3, Leucosolenia sp., Lissodendoryx (Lissodendoryx) sp., Mycale beatrizae, Myxilla (Ectyomyxilla) tenuissima, Pachastrella monilifera, Pachypellina sp., Poecillastra sollasi, Polymastia sp., Raspaciona sp., Raspailia (Parasyringella) sp., Raspailia (Raspaxilla) phakellina, Rhabderemia besnardi, Rhabderemia uruguaiensis, Sceptrella sp., Stelletta sp., Thrombus sp., Timea sp., Topsentia sp., Vulcanella sp. Agelas sp., Aiolochroia crassa, Aphrocallistes beatrix, Aplysina cauliformis, Aplysina fulva, Aplysina sp. nov., Axinella sp., Cinachyra sp. nov., Cinachyrella aff. kuekenthali, Cinachyrella kuekenthali, Desmacella sp., Dragmacidon sp., Erylus sp., Oceanapia sp., Pachastrissa sp., Petromica sp., Plakinastrella sp., Rhabderemia sp., Spongosorites sp. Aaptos sp., Agelas clathrodes, Agelas schmidti, Agelas sp., Aiolochroia crassa, Aphrocallistes beatrix, Aplysina cf. fulva, Aplysina lacunosa, Aplysina sp., Craniella sp., Cinachyra sp. nov., Dactylocalyx pumiceus, Erylus sp., Lophocalyx sp. nov. 1, Oceanapia sp., Plakina sp. Aiolochroia crassa, Aplysina sp. Cinachyrella aff. alloclada, Cinachyrella aff. apion, Cinachyrella aff. kuekenthali, Lophocalyx sp. nov. 2 Aplysina archeri ?, Cacospongia levis, Characella aspera, Corallistes typus, Geodia neptuni, Ircinia strobilina, Phakellia connexiva, Pheronema carpenteri, Thenea fenestrata Euplectella suberea, Polymastia corticata Alectona mesatlantica Agelas dispar, Aplysina fulva, Halichondria sp., Haliclona sp., Hyalonema schmidti, Hyattella sp., Ircinia sp., Niphates sp., Plakinastrella sp. Number of species records 17

SC SP

15 40

RJ

19

ES BA AL PE RN (ASPSP) CE

16 6 9 2 1 9

sea of south/south-eastern Brazil, but ongoing inventories may bring exciting new discoveries in this respect.

Acknowledgements
Authors are grateful to CAPES, CENPES/PETROBRAS, CNPq and FAPERJ for grants and/or fellowships. Two reviewers contributed for greater clarity and comprehensiveness of this manuscript.

References
Boury-Esnault N (1973) Rsultats scientifiques des campagnes de la Calypso. Campagne de la Calypso au large des ctes atlantiques de lAmrique du Sud (19611962). I. 29. Spongiaires. Ann Inst ocanogr 49 (Supp 10): 263-295 Collette BB, Rtzler K (1977) Reef fishes over sponge bottoms off the mouth of Amazon River. Proc 3rd Int Coral Reef Symp 1: 305309 Floeter SR, Soares-Gomes A (1999) Biogeographic and species richness patterns of Gastropoda on the southwestern Atlantic. Braz J Biol 59: 567-575

Hajdu E, Desqueyroux-Fandez R (1994) A synopsis of South American Mycale (Mycale) (Poecilosclerida, Demospongiae), with description of three new species and a cladistic analysis of Mycalidae. Rev suisse Zool 101: 563-600 Hajdu E, Muricy G, Berlinck RGS, Freitas JC (1996) Marine poriferan diversity in Brazil. Through knowledge to management, In: Bicudo CEM, Menezes N (eds). Biodiversity in Brasil. A first approach. CNPq, So Paulo. pp. 157-171 Hajdu E, Santos CP, Lopes DA, Oliveira MV, Moreira MCF, Carvalho MS, Klautau M (2004) Filo Porifera. In: Amaral ACZ, Rossi-Wongtschowski CLLD (orgs.), Biodiversidade bentnica das regies sudeste e sul do Brasil - Plataforma externa e talude superior. Instituto Oceanogrfico, So Paulo. pp. 49-56 Hooper JNA, van Soest RWM (2002) Systema Porifera: a guide to the classification of sponges. Kluwer Academic/Plenum Publishers, New York Lavrado HP (2006) Caracterizao do ambiente e da comunidade bentnica. In: Lavrado HP, Igncio BL (orgs.). Biodiversidade bentnica da regio central da Zona Econmica Exclusiva brasileira. Museu Nacional, Srie Livros 18, Rio de Janeiro. pp. 19-64

359

Lerner C, Hajdu E (2002) Two new Mycale (Naviculina) Gray (Mycalidae, Poecilosclerida, Demospongiae) from the Paulista Biogeographic Province (Southwestern Atlantic). Revta bras Zool 19: 109-122 Lopes DA, Hajdu E (2004) Two new Mycalina from the south-eastern Brazilian shelf and slope collected by Programme REVIZEE (Poecilosclerida: Demospongiae). J Mar Biol Assoc UK 84: 25-28 Lopes DA, Hajdu E, Reiswig HM (2005) Redescription of two Hexactinosida (Porifera, Hexactinellida) from the southwestern Atlantic, collected by Programme REVIZEE. Zootaxa 1066: 4356 Menshenina LL, Tabachnick KR, Lopes DA, Hajdu E (2007) Revision of Calycosoma Schulze, 1899 and finding of Lophocalyx Schulze, 1887 (six new species) in the Atlantic Ocean (Hexactinellida, Rossellidae). In: Custdio MR, Lbo-Hajdu G, Hajdu E, Muricy G (eds). Porifera research: biodiversity, innovation and sustainability. Srie Livros 28. Museu Nacional, Rio de Janeiro. pp. 449-465 Mothes B (1977) Ocorrncia de Dactylocalyx pumiceus Stutchbury, 1841 no litoral do Rio Grande do Sul (Porifera, Hexactinellida). Iheringia 50: 41-49 Mothes B, Capitoli RR, Lerner, CB, Campos MA (2004) Filo Porifera. Regio Sul. In: Amaral ACZ, Rossi-Wongtschowski CLLD (orgs.), Biodiversidade bentnica das regies sudeste e sul do Brasil - Plataforma externa e talude superior. Instituto Oceanogrfico, So Paulo. pp. 57-63 Muricy G, Santos CP, Batista D, Lopes DA, Pagnoncelli D, Monteiro LC, Oliveira MV, Moreira MCF, Carvalho M de S, Melo M, Klautau M, Rodriguez PRD, Costa RN, Silvano RG, Schwientek S, Ribeiro SM, Pinheiro US, Hajdu E (2006) Filo Porifera. In: Lavrado HP, Igncio BL (orgs.). Biodiversidade bentnica da regio central da Zona Econmica Exclusiva brasileira. Museu Nacional, Srie Livros 18, Rio de Janeiro. pp. 109-145

Oliveira MV, Hajdu E (2005) Taxonomy of Rhabderemia Topsent, 1890 collected from the south-eastern Brazilian continental shelf and slope by Programme REVIZEE (Rhabderemiidae, Poecilosclerida, Demospongiae), with description of two new species. Zootaxa 844:1-12 Palacio FJ (1982) Revisin zoogeogrfica marina del sur del Brasil. Bol Inst Oceanogr Univ So Paulo 31: 69-92 Polejaff N (1884) Report on the Keratosa collected by H.M.S.Challenger during the years 18731876. Rep Sci Res Voy H.M.S. Challenger, Zool 11: 1-88 Ridley SO, Dendy A (1887) Report on the Monaxonida collected by H.M.S. Challenger during the years 18731876. Rep Sci Res Voy H.M.S. Challenger, Zool 20(59): 1-275 Santos JP, Mothes B, Tenrio DO, Cantarelli J (1999) Porifera (Demospongiae, Calcarea) entre os estados do Cear e Pernambuco, Brasil. Taxonomia e distribuio. Trab Oceanogr Univ Fed PE 27(2): 49-60 Schulze FE (1887) Report on the Hexactinellida collected by H.M.S. Challenger during the years 18731876. Rep Sci Res Voy H.M.S. Challenger, Zool 21: 1-514 Schulze FE (1899) Amerikanische Hexactinelliden, nach dem Materiale der Albatross-Expedition. (Fischer: Jena): 1-126 Silva CMM, Mothes B (2000) Three new species of Geodia Lamarck, 1815 (Porifera, Demospongiae) from the bathyal depths off Brazilian coast, southwestern Atlantic. Rev suisse Zool 107: 31-48 Sollas WJ (1888) Report on the Tetractinellida collected by H.M.S. Challenger, during the years 18731876. Rep Sci Res Voy H.M.S. Challenger, Zool 25(63): 1-458 Vacelet J (1999) Planktonic armoured propagules of the excavating sponge Alectona (Porifera: Demospongiae) are larvae: evidence from Alectona wallichii and A. mesatlantica sp. nov. Memoir Queensl Mus 44: 627-642

Porifera research: Biodiversity, innovation and sustainaBility - 2007

361

Molecular markers for species discrimination in poriferans: a case study on species of the genus Aplysina
Isabel Heim(*), Michael Nickel, Franz Brmmer
Universitt Stuttgart, Biologisches Institut, Abteilung Zoologie, Pfaffenwaldring 57, 70569 Stuttgart, Germany.() isabel.heim@bio.uni-stuttgart.de, m.nickel@uni-jena.de, franz.bruemmer@bio.uni-stuttgart.de Abstract: We tested different molecular markers for their utility as species discriminators in Porifera: internal transcribed spacer-1 and -2 (ITS-1, ITS-2) rDNA, mitochondrial 12S, 16S, and cytochrome oxidase subunit I (COI). The study was performed on specimens of the genus Aplysina from different locations in the Mediterranean Sea, East, and West Atlantic. This genus is widespread in tropical and subtropical waters and rich in natural substances. From the Mediterranean Sea, two different species are known: A. aerophoba (Schmidt, 1862) and A. cavernicola (Vacelet, 1959). We intended to find an adequate molecular marker to differentiate the Mediterranean species A. aerophoba and A. cavernicola. However, there was a high degree of intra-individual polymorphism within the markers ITS-1 and ITS-2. Consequently, these markers were not adequate for species differentiation of A. aerophoba and A. cavernicola for technical reasons discussed here. In contrast, the mitochondrial 12S and 16S are highly conserved and no differences among species were observed. Only the COI showed a low variability in the seven analysed Aplysina species. Based on COI, it was possible to classify the specimens COI in two clades. One clade is represented by A. aerophoba and A. cavernicola which could be distinguished from the Caribbean Aplysina species. However, the resolution was low in the second clade consisting of Western Atlantic Aplysina species, suggesting a recent radiation event. Our results suggest that a general molecular markers for species discrimination does not exist. Hence, the choice of a suitable marker strongly depends on the evolutionary context of each single taxon and will have to be tested accordingly. Keywords: Aplysina, COI, ITS, mt-rDNA, species discrimination

Introduction
Sponges of the genus Aplysina, Nardo 1834 are abundant in the subtropical and tropical waters of the Mediterranean Sea (Boury-Esnault 1971, Kreuter et al. 1992), Pacific Ocean (Carney and Rinehart 1995, Betancourt-Lozano et al. 1998), and Atlantic Ocean (Pinheiro and Hajdu 2001, Saeki et al. 2002). Only two species of this genus are described for the Mediterranean Sea: A. aerophoba (Schmidt, 1862) and A. cavernicola (Vacelet 1959). A. aerophoba lives in shallow water from 1 30 m on rocks and tolerates high variation of temperature, insolation and density (Kreuter et al. 1992, Vacelet 1971). In contrast A. cavernicola favours caves and shadowy areas in depth from 7 130 m (Vacelet 1959, 1969). A. aerophoba and A. cavernicola can be differentiated in the Western Mediterranean Sea in morphology and biochemistry (Vacelet 1959, Ciminiello 1997, Heim 2003). In the Limski kanal north of Rovinj/Croatia, both species occur in the same
() M. Nickel present address: Friedrich-Schiller-Universitt Jena, Institut fr Spezielle Zoologie und Evolutionsbiologie mit Phyletischem Museum, Erbertstr. 1, 07743 Jena, Germany

habitat. In addition, species differentiation is difficult due to high intra-specific variation of form and colour (Heim 2003; Fig. 1). The same problem was reported from the Aegean Sea, which gave rise to the question whether or not A. cavernicola is a true species of its own or just an ecological variant of A. aerophoba, living in caves (Voultsiadou-Koukoura 1987). In addition to morphological characters, the biochemistry of secondary metabolites was recently investigated in order to differentiate the two Mediterranean species. However, it turned out that biochemical profiles are no valuable character since specimens were found in the Limski kanal in Croatia, which combine the secondary metabolite profile of A. aerophoba and A. cavernicola or do not display any of the characteristic substances at all (Heim 2003, Thoms 2004). For this reason, a molecular marker for species discrimination would be useful in the case of the Mediterranean Aplysina species. Especially for the differentiation of the species found in the Adriatic Sea. For this purpose different regions from the nuclear or mitochondrial genome are potential markers. In the last few years, several studies were performed on sponges and corals using the one or the other candidate marker. The first and the second internal transcribed spacer regions (ITS-1 and ITS-2) between the 18S, 5,8S and 28S ribosomal

362

RNA genes are preferably used for intra- and interspecific phylogeographic relationships (Wrheide 1998, King et al. 1999, van Oppen et al. 2002, Wrheide et al. 2002b, Duran et al. 2004a, Schmitt et al. 2005). The ITS-regions evolve rapidly and, hence, are useable as high resolution marker in populations genetics (van Oppen et al. 2002, Wrheide et al. 2002a). However, in some cases polymorphisms have been detected in these non-coding regions (Wrheide et al. 2004, Nichols and Barnes 2005). When intragenomic variations are found, the ITS region is not useful for analyses at populationlevel (Nichols and Barnes 2005). On this account it is necessary to screen the investigated taxon for the presence and extent of intragenomic polymorphisms to adapt the molecular methods (Wrheide et al. 2004). In other invertebrate taxa the mitochondrial 12S rDNA as well as the 16S rDNA were used for phylogeny and phylogeography analysis, including scleractinian corals and hydrozoans (Chen et al. 2002, Govindarajan et al. 2005). Both studies reported a slow divergence rate in Cnidaria which is probably triggered by a mismatch repair system homologue to the bacterial MutSLH system (Pont-Kingdon et al. 1998). The same could be occurring in sponges because Duran et al. (2004b) reported about a low genetic variation in the cytochrome oxidase subunit I for the species Crambe crambe, but there are no studies yet on mitochondrial 12S and 16S in sponges. Another region of the mitochondrial genome is often used in marine invertebrates. For population genetics the cytochrome c oxidase subunit I (COI) was used for mussels and ascidians (Avise et al. 1987, King et al. 1999, Tarjuelo et al. 2001). In Crambe crambe as well as in Astrosclera willeyana the COI sequences were tested for its use as an intraspecific genetic marker among populations of sponge species (Duran et al. 2004b, Wrheide 2005), but in both cases the marker was too conserved for population studies, as also shown for cnidarians (Shearer et al. 2002, van Oppen et al. 2002). Although the COI has been found to be too conserved for population studies in cnidarians and sponges, it could be possible to use the marker for species discrimination in sponges. Recent studies demonstrated that the COI is a valuable marker in higher taxa like birds, fishes and butterflies to differentiate species (Hebert et al. 2004, Ward et al. 2005). In the present study we tested five different molecular markers (ITS-1, ITS-2, 12S, 16S and COI) for their usefulness in species discrimination within the genus Aplysina.

paper. The Aplysina sp. specimen has secondary metabolites that are present in A. aerophoba as well as in A. cavernicola (Thoms 2004). Sampling was undertaken by SCUBA diving. Pseudocertina sp. was taken from the public aquarium of the zoological and botanical gardens Stuttgart (Wilhelma) and used as outgroup. The cortex of the samples was cut off before frozen in liquid nitrogen for storage.

DNA extraction, amplification and sequencing

Whole cellular DNA extraction for all organisms was performed with Guanidinethiocyanate-buffer (25 mg tissue per ml; 5 M Guanidinethiocyanate, 20 mM EDTA dissolved at 65C, cooled down and than add 10% N-Laurylasarcosinin) followed by phenol - chloroform extraction. Polymerase chain reaction (PCR) amplifications were performed in a total volume of 50 l using Genaxxon polymerase (Genaxxon Biosciences, Biberach, Germany). For the PCR of the ITS-region (including ITS-1, 5.8S and ITS-2), the primers ITS1 and RA2 (Wrheide 1998) were used. After an initial denaturation step at 96C for 3 min, rDNA was amplified during 30 cycles of 95C for 1min, 55C for 30 s and 72C for 1 min and a final extension at 72C for 7 min. In the case of the mitochondrial 12S region, the primers 12S-For and 12S-Rev (Chen and Yu 2000) were utilised. The amplification started with an initial denaturation at 95C for 4 min, rDNA was amplified during 4 cycles of 94C for 1 min, 50C for 30 s and 72C for 3 min and 30 cycles of 94C for 30 s, 60C for 1 min and 72C for 3 min. For the mitochondrial 16S region, the primers 16S1 and 16S2 were used (Bridge et al. 1995). After an initial denaturation step at 95C for 4 min, the rDNA was amplified during 4 cycles of 94C for 1 min, 50C for 30 s and 72C for 3 min and 30 cycles of 94C for 30 s, 52C for 1 min and 72C for 3 min. A part of the cytochrome oxidase subunit I (COI) was amplified using the primers COI-For and COI-Rev (Folmer et al. 1994). The amplification started with an initial denaturation step at 94C for 2 min followed by 35 cycles of 94C for 50 s, 40C for 55 s and 72C for 1 min and a final extension at 72C for 7 min. All of the obtained PCR-products (ITS-1, ITS-2, 12S, 16S and COI) were cleaned with NucleoSpin Extract Kit (Machery-Nagel, Dren, Germany) before they were ligated into pCR4-TOPO vector (Invitrogen, Carlsbad, Canada) and transformed by heat-shock into competent E. coli One ShotTOP10 (Invitrogen, Carlsbad, Canada). Plasmid DNA was isolated with NucleoSpin Plasmid Quick Pure (MacheryFig. 1: Different form and colour variations of Aplysina specimens in the Mediterranean Sea. A. A. cavernicola (Giglio, Secca II, Italy); B. A. cavernicola, arrow tags colour variation within one individual (Giglio, Fenaio, Italy); C. A. aerophoba (Banyuls-surmer, France); D. A. aerophoba (Banyuls-sur-mer, France); E. Different variations of A. aerophoba and A. cavernicola (Limski kanal, Rovinj, Croatia); F. A. cavernicola (Limski kanal, Rovinj, Croatia); G. A. aerophoba (San Giovanni, Rovinj, Croatia); H. A. aerophoba (Rovinj, Croatia) and I. Aplysina sp. (Limski kanal, Rovinj, Croatia).

Material and methods Sampling

Live sample organisms of the genus Aplysina were collected during fieldwork within the research project BIOTECmarin: Aplysina aerophoba (CRO), (FRA), (SPA) and (POR); A. cavernicola (CRO), (FRA) and (ITA); Aplysina. sp. (CRO); A. fistularis (Pallas, 1766) from Cat Island/Bahamas; A. archeri (Higgin, 1875), A. cauliformis (Carter, 1882), A. insularis (Duchassing and Michelotti, 1864) from Little San Salvador/Bahamas and A. fulva (Pallas, 1766) from Sweetings Cay/Bahamas (Fig. 2). In Table 1 the processed sponge specimens are listed with their abbreviations used in this

363

364

Fig. 2: Map showing the localities localities in Europe (large map) and the Bahamas area (inset map) where Aplysina individuals were sampled (1 = A. aerophoba, Madeira; 2 = A. aerophoba, Cadaqus, Spain; 3 = A. aerophoba, Banyuls-sur-mer, France; 4 = A. cavernicola, Marseille, France; 5 = A. cavernicola, Fenaio, Giglio, Italy; 6 = A. aerophoba, A. cavernicola and Aplysina sp., Rovinj, Croatia; 7 = A. fulva, Sweetings Cay, Bahamas; 8 = A. archeri, A. cauliformis and A. insularis, Little San Salvador Island, Bahamas; 9 = A. fistularis, Cat Island, Bahamas).

Table 1: Listing of the analysed Aplysina species with origin and the abbreviations used in this paper. Species A. aerophoba A. aerophoba A. aerophoba A. aerophoba A. cavernicola A. cavernicola A. cavernicola Aplysina sp. A. fistularis A. archeri A. cauliformis A. insularis A. fulva Origin Limski kanal, Croatia Banyuls-sur-mer, France Cadaquz, Spain Madeira, Portugal Limski kanal, Croatia Marseille, France Isola del Giglio, Italy Limski kanal, Croatia Cat Island, Bahamas Little San Salvador, Bahamas Little San Salvador, Bahamas Little San Salvador, Bahamas Sweetings Cay, Bahamas Abbreviation A. aerophoba (CRO) A. aerophoba (FRA) A. aerophoba (SPA) A. aerophoba (POR) A. cavernicola (CRO) A. cavernicola (FRA) A. cavernicola (ITA) Aplysina sp. (CRO) A. fistularis A. archeri A. cauliformis A. insularis A. fulva

Nagel). The correct insert size was verified by using agarose gel electrophoresis following an insert check with the internal primer and the M13 forward. The sequencing was performed by AGOWA GmbH (Berlin, Germany). The sequencing reactions were done with the M13 forward or M13 reverse primers. Three clones per individual were picked and sequenced to make a consensus sequence expect of the 12S and 16S cloning. An overview of the tested molecular markers and the Aplysina species we used is given in Table 2. Sequences were edited and manipulated using BioEdit (Hall 1999). Sequence alignment was performed using the multiple sequence alignment program CLUSTAL X (Thompson et al. 1997). A similarity search (BLAST) was performed to confirm that sequences were from sponge origin and not from other possible contaminants such as symbionts. The nucleotide

sequence data reported in this paper have been deposited in the GenBank nucleotide sequence database with accession numbers EF043343 to EF043378.

Phylogenetic analysis
For the COI we obtained 654 bp of four A. aerophoba specimens (SPA, POR, FRA and CRO), three A. cavernicola specimens (FRA, CRO and ITA), one Aplysina sp. (CRO), one specimen respectively of A. fulva, A. archeri, A. cauliformis, A. fistularis, A. insularis and Pseudoceratina sp. from the public aquarium in Stuttgart. These sequences were utilised for the following phylogenetic analysis. Aiolochroia crassa (AJ843885) from GenBank and Pseudoceratina sp. were used as outgroup in the calculations.

365

Table 2: Molecular markers analysed in the present study for species of the genus Aplysina. Sample/Origin A. aerophoba (CRO) Aplysina sp. (CRO) A. cavernicola (CRO) A. cavernicola (ITA) A. cavernicola (FRA) A. aerophoba (FRA) A. aerophoba (SPA) A. fistularis A. archeri A. cauliformis A. insularis A. fulva Pseudoceratina sp. ITS-1 ITS-2 12S 16S COI

Neighbour-Joining analysis
Accessorily to the Bayesian approach we carried out a neighbour-joining analysis (NJ) using PAUP*4.0b10. To find the best model of DNA substitution we used Modeltest 3.7. The HKY+I was the best-fit model and the parameters of this model were used for the subsequent neighbour-joining analysis. All trees were rooted by the outgroup Pseudoceratina sp. and displayed by using TreeView 1.6.6 (Page 1996).

Results ITS-1
For the ITS-1 region we obtained a total of 270 bp. Two deletions are present in clone 2 between position 42 to 47 with six bp as well as in position 133 to 163 with 31 bp and one insertion in position 216 to 219 in clone three. Altogether 8 bp are exchanged in all three clones (Fig. 3).

Bayesian analysis
The hierarchical Akaike information criterion (AIC), which is implemented in MrModeltest 2.2, was used for calculation of models of nucleotide substitution (Posada and Crandall 1998). The AIC was chosen instead of the hierarchical likelihood ratio test (hLRT) because it imposes a disadvantage for model complexity resulting in models with better predictive accuracy (Sober 2002). The HKY model was chosen as the best fit model and the parameters were used for the calculation in MrBayes 3.1.2. Four Markov chains were run for one million generations and sampled every 100 generations to generate a posterior probability distribution of 10.001 trees. Posterior probabilities were calculated by constructing a 50% majority rule consensus tree of the stationary trees (i.e. trees saved after burn-in trees are excluded).

ITS-2
For the ITS-2 region we obtained a sequences length between 209 and 274 bp for A. aerophoba (CRO), A. aerophoba (FRA), A. cavernicola (CRO) and A. cavernicola (FRA) specimens. In the case of A. aerophoba (FRA), A. cavernicola (FRA) and A. cavernicola (CRO), we detected intra-individual polymorphism at the ITS-2 region within one individual. Clone 2 of A. cavernicola (FRA) displays an insert of 61 bp at position 124 to 184 (Fig. 4A), which does not occur in the other two clones. In the case of A. aerophoba (FRA) a deletion of 25 bp is present (Fig. 4B). Furthermore, clone 3 shows another deletion of 6 bp at position 117 to 123 (Fig. 4B). For A. cavernicola (CRO), we found a total of 12 base pair exchanges and one deletion of six base pairs in the poly-C region at position 152 to 157 in the clone 3 (Fig. 4C). The specimen A. aerophoba (CRO) displays only one deletion of two base pairs in the poly-C region of the clone 3 (Fig. 4D).

Fig. 3: Alignment of the three sequenced ITS-1 clones of A. cavernicola, Giglio.

366

367

12S
In case of the mitochondrial 12S rRNA we obtained a partial sequence with 966 bp for A. aerophoba (FRA), A. aerophoba (CRO) and A. cavernicola (CRO). No base pair exchanges are present in this region (data not shown, see GenBank entries).

in contrast to the A. aerophoba individuals analysed here: The specimens of A. aerophoba from Croatia and Spain group together, while the specimens of A. aerophoba from Madeira and Banyuls-sur-mer are separated from the others and even display differences among each other.

Discussion
This is the first time that different molecular markers are analysed for their usefulness in species discrimination of sponges within a genus. For this purpose, seven species of the genus Aplysina were collected in the Mediterranean Sea, East, and West Atlantic Ocean spanning a distance of around 9.000 km apart. In the case of the internal transcribed spacer regions (ITS1 and ITS-2) our results for the genus Aplysina showed a high intra-individual variability. The ITS-1 and ITS-2 region has been frequently used for intra- and interspecific relationships in corals and sponges (van Oppen et al. 2002, Wrheide et al. 2002b, Duran et al. 2004a). For the Mediterranean species of the genus Aplysina (A. aerophoba and A. cavernicola) we found in both regions intragenomic variations like Wrheide et al. (2004) have described. Inserts as well as deletions occur in both species. The largest insert was found in A. cavernicola from Marseille with 61 bp. In the specimen from Marseille, a deletion with 21 bp occurred. There is no general pattern behind the inserts and deletions, so it is not possible to differentiate the Mediterranean Aplysina species with the ITS-1 and ITS-2 region. Principally, this polymorphism does not exclude ITS sequences from the list of suitable genetic markers for species discrimination. However, in order to accurately use such a polymorphic marker, it would be necessary to sequence a high number of clones from every single specimen, in order to reach a saturation of all present sequences. Only then, the corresponding sequences could be used for phylogenetic analyses. This is a very expensive and timely endeavour. We also tested part of the mitochondrial 12S and 16S rDNA for the purpose of species discrimination. In contrast to the ITS sequences, both coding rDNA regions are highly conserved and no basepair exchanges have been found in the two Aplysina species analysed. Therefore, as shown the mitochondrial 12S and 16S regions are also not adequate for species differentiation. However, there could be the possibility for using these markers at the family or genus level like it has been used in damselfishes and shrimps (Jang-Liaw et al. 2002, Quan et al. 2004, respectively). A more promising marker for species discrimination in sponges seems to be COI. Its usefulness for population and biogeography studies had tested only in two sponge species so far: Crambe crambe and Astrosclera willeyana. In both cases, the variability was not sufficiently high enough for population and phylogeographical analyses (Duran et al. 2004b, Wrheide 2005). Recent studies in higher taxa like birds, fishes and butterflies show that the differences between closely related species is 18 time higher than within the species (Hebert et al. 2004, Ward et al. 2005, Hajibabaei et al. 2006). In addition, a recent study on the sponge genus Tethya, provided promising results with clear species discrimination (Heim et al. 2007).

16S
Additionally, we analysed the mitochondrial 16S rRNA and obtained a partial sequence of 707 bp for A. aerophoba (FRA), A. aerophoba (CRO) and A. cavernicola (CRO). All three sequences are identical (data not shown, see GenBank entry).

COI
A BLAST search for the COI sequences in GenBank revealed A. fistularis, Aiolochroia crassa, Axos cliftoni and Chondrosia sp. sequences as closest matches. No contaminating sequences were found. Analysis of sequence variations within the COI gene among the seven Aplysina species displayed six phylogenetically informative sites. The transition to transversion substitution ratio was 1.0 (3/3). All six informative sites represent substitutions in the third position of the respective codon. None of these base pair exchanges resulted in an amino acid substitution (Table 3: upper section). All samples analysed of the genus Aplysina, regardless of their species have the same amino acid sequence (alignment not shown). Table 3 (lower section) displays pairwise base pair exchanges between all the species (including pairwise comparisons between specimens of the same species but from different locations). The most exchanges are present between A. archeri and A. cavernicola (CRO, ITA, and FRA) with 5 bp. This represents an exchange rate of 0.75% between these species. The lowest substitution rate is found between the Caribbean species A. fistularis, A. cauliformis, A. fulva and A. insularis with 0 bp. The exchange rate between the Mediterranean species A. aerophoba (CRO, SPA) and A. cavernicola (CRO, FRA, ITA) is 1 bp with a substitution rate of 0.15%. Between the A. aerophoba specimens from Madeira and Banyuls-sur-mer occurs 1 bp respectively 2 bp differences to the A. aerophoba individuals from Croatia and Spain. This conforms an exchange rate of 0.15% and 0.30%. The phylogenetic trees calculated with the Bayesian approach and the neighbour-joining analyse display identical topologies (Fig.5A, B). A. archeri represents a basal branch within the genus and is separated from the other Caribbean species A. fistularis, A. fulva, A. insularis and A. cauliformis. There are no sequence differences between the later four species. In the case of the Mediterranean species A. aerophoba and A. cavernicola, the situation is complicated. All three A. cavernicola specimens (CRO, ITA, FRA) form a single clade
Fig. 4: Alignment of the three sequenced ITS-2 clones of A. aerophoba, Limski kanal, Croatia (A), A. aerophoba, Banyulssur-mer, France (B), A. cavernicola, Limski kanal, Croatia (C) and A. cavernicola, Marseille, France (D).

368

Table 3: Amino acid exchanges (upper section) and base pair exchanges (lower section) of the COI between two different Aplysina species. A. cavernicola, FRA A. cavernicola, ITA A. cavernicola, CRO A. aerophoba, CRO Aplysina sp., CRO A. aerophoba, SPA A. aerophoba, POR A. aerophoba, FRA

A. cauliformis

A. fistularis

A. insularis

A. cavernicola, FRA A. cavernicola, ITA A. cavernicola, CRO A. aerophoba, CRO Aplysina sp., CRO A. aerophoba, SPA A. aerophoba, FRA A. aerophoba, POR A. fistularis A. fulva A. cauliformis A. insularis A. archeri 1 3 2 3 3 3 3 5

0 0

0 0 0

0 0 0 0

0 0 0 0 0

0 0 0 0 0 0

0 0 0 0 0 0 0

A. archeri 0 0 0 0 0 0 0 0 Amino acid exchanges

2 1 2 2 2 2 4 1 1 1 1 1 4

1 1 1 1 3 0 0 0 2

A. fulva 0 0 2

0 2 2

Nucleotide base pair exchanges

Clades within the phylogenetic tree of Aplysina specimens partly correspond to geographical locations A. archeri is basal to all other species analysed here. The Western Atlantic species A. insularis, A. cauliformis, A. fulva and A. fistularis display identical sequences, even though they represent, based on morphological characters, clearly identifiable species. In the context of the clear species discrimination in the genus Tethya, we may conclude that the radiation of these four Aplysina species started only recently. As a consequence, the morphology changed, but no base pair exchange in the COI occurred up to now. The West Atlantic Aplysina species are relatively young and seem to represent sexually compatible taxa (Schmitt et al. 2005). They also mentioned a low resolution in the COI within the genus Aplysina. In contrast, the situation is more difficult amongst he European specimens of Aplysina. Despite all sequence similarities, the present differences in the COI sequence are sufficient to separate at least A. cavernicola from specimens which are regarded as A. aerophoba. This is maintained through sequencing different clones from every individual and we can find the same base pair exchange in the position 607 in different geographical localities (Rovinj/Croatia, Marseille/France, and Giglio/Italy). They also show all the same transversion. This supports the hypothesis that A. cavernicola is a true species (Vacelet 1959) and not an ecological variant of A. aerophoba (VoultsiadouKoukoura 1987). Both trees support the possibility to identify A. cavernicola regardless of geographic origin using COI sequences.

However, this seems not to be the case for specimens regarded as A. aerophoba. This species came out to be problematic. The species A. aerophoba sensu morphology does not form its own genetic clade of itself, but two separate groups, of which the specimens from Croatia and Spain form a branch together with A. cavernicola. The A. aerophoba specimens from France and Portugal branch out more basally. There are two possible scenarios: either all A. aerophoba analysed here represent one species. In this case, A. cavernicola would consequently be no valid species as suggested earlier in the literature (Voultsiadou-Koukoura 1987). However, due to the clear identity of A. cavernicola sensu morphology and sensu COI, a second scenario seems to be more likely: A. aerophoba sensu morphology is not a single valid species, but a cluster of species. Aplysina sp. (CRO), which displays secondary substances both from A. aerophoba and A. cavernicola, shares its complete COI sequence with the A. aerophoba CRO and SPA. It had been speculated before whether it is a hybrid of both species because in the Limski kanal near Rovinj/ Croatia both A. aerophoba and A. cavernicola overlap in the inhabiting environment. However, this is not verifiable with the COI sequence. This group of three identical genotypes would represent a sister species of A. cavernicola. In contrast, the samples identified as A. aerophoba POR and FRA most likely represent one or even two distinct species. This cannot be decided from the present dataset. By any means, our result calls for a complete revision of the European Aplysina species, to resolve the emerging

369

Fig. 5: A. 50% majority rule consensus phylogram of the stationary trees obtained from the Bayesian analysis of COI sequences. Posterior probabilities are shown below branches. B. Neighbour-joining tree calculated with PAUP*4.0b10. Bootstrap values higher than 50% are plotted above branches.

problem of A. aerophoba. Beside careful morphological and histological studies, the application of bulk ITS-sequencing might help, if a saturation of the intra-individual alleles is reached as discussed above. The same might be true for the West Atlantic group. All base pair exchanges found in COI took place at the third codon position. Consequently, no mutations are present in the amino acid sequences of any species we analysed from West and East Atlantic Ocean and Mediterranean Sea. It is not clear yet why most sponge genera display such a low mutation rate in the COI. Contrary to the Aplysina species we have found for the genus Tethya a good resolution for species discrimination with the COI (Heim et al. 2007). Studies on octocorals show also a low mutation rate in the mitochondrial DNA (van Oppen et al. 1999) but this is probably caused by a mismatch repair system homologue to the bacterial system (Pont-Kingdon et al. 1998). Although possible, until now there is no evidence for such a mismatch repair system in sponges

(Lavrov et al. 2005, Lavrov and Lang 2005). Furthermore the low mutation rate could be avowed through the variable environmental conditions Aplysina species living in, because mitochondrial evolution is advantaged by relaxed selection pressure (Quesada et al. 1998). To conclude, it was not possible to differentiate the Mediterranean Aplysina species with the ITSs-regions without massively increasing the number of sequencings per individual. Also, for the 12S and 16S, a species differentiation is not probable, because of a high genetic similarity between the species. But maybe they are useful at the family or genus level. Nevertheless, for COI the discrimination between A. aerophoba and A. cavernicola is feasible, despite lower differences in their sequences in comparison to those between the species of the genus Tethya (Heim et al. 2007). It is always necessary to check the COI for its usefulness in species discrimination. Additionally, the same is to be considered for the other tested molecular markers.

370

Acknowledgements
We would like to thank Renato Batel (Institut Ruder Bokovi, Rovinj), Ute Hentschel (University Wrzburg), Isabel Koch (Wilhelma Stuttgart) and Wolfgang Zucht (University Stuttgart) for providing sponge specimens. Gisela Fritz (Universitt Stuttgart, Germany) for critical reading of our manuscript and two anonymous reviewers greatly helped to improve the manuscript. This study was supported by the project BIOTECmarin (03F0414D) funded by the German Federal Ministry of Education and Research and the University of Stuttgart.

References
Avise JC, Arnold J, Ball RM, Bermingham E, Lamb T, Neigel JE, Reeb CA, Saunders NC (1987) Intraspecific phylogeography: the mitochondrial DNA bridge between population genetics and systematics. Annu Rev Ecol Syst 18: 489-522 Betancourt-Lozano M, Gonzales-Farias F, Ganzales-Acosta B, Garcia-Gasca A, Bastida-Zavala JR (1998) Variation of antimicrobial activity of the sponge Aplysina fistularis (Pallas, 1766) and its relation to associated fauna. J Exp Mar Biol Ecol 223: 1-18 Boury-Esnault N (1971) Spongiaires de la zone rocheuse littorale de Banyuls-sur-mer. I. - cologie et rpartition. Vie Milieu 22 159192 Bridge D, Cunningham CW, DeSalle R, Buss LW (1995) Class-level relationships in the phylum Cnidaria: molecular and morphological evidence. Mol Biol Evol 12(4): 679-689 Carney JR, Rinehart KL (1995) Biosynthesis of brominated tyrosine metabolites by Aplysina fistularis. J Nat Prod 58(7): 971-985 Chen CA, Wallace CC, Wolstenholme J (2002) Analysis of the mitochondrial 12S rRNA gene supports a two-clade hypothesis of the evolutionary history of scleractinian corals. Mol Phylogenet Evol 23: 137-149 Chen CA, Yu JK (2000) Universal primers for amplification of mitochondrial small subunit ribosomal RNA-encoding gene in scleractinian corals. Mar Biotech 2: 146-153 Ciminiello P, Fattorusso E, Forino M, Magno, S (1997) Chemistry of Verongida sponges VIII - Bromocompounds from the Mediterranean sponges Aplysina aerophoba and Aplysina cavernicola. Tetrahedron 53(18): 6565-6572 Duran S, Giribet G, Turon X (2004a) Phylogeographical history of the sponge Crambe crambe (Porifera, Poecilosclerida): range expansion and recent invasion of the Macaronesian islands from the Mediterranean Sea. Mol Ecol 13 109-122 Duran S, Pascual M, Turon X (2004b) Low levels of genetic variation mtDNA sequences over the western Mediterranean and Atlantic range of the sponge Crambe crambe (Poecilosclerida). Mar Biol 144: 31-35 Folmer O, Black M, Hoch W, Lutz R, Vrijenhoek R (1994) DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol Mar Biol Biotechnol 3: 294-299 Govindarajan AF, Halanych KM, Cunningham CW (2005) Mitochondrial evolution and phylogeography in the hydrozoan Obelia geniculata (Cnidaria). Mar Biol 146: 213-222 Hajibabaei M, Janzen DH, Burns JM, Hallwachs W, Hebert PDN (2006) DNA barcodes distinguish species of tropical Lepidoptera. Proc Natl Acad Sci USA 103(4): 968-971

Hall TA (1999) BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symp 41: 95-98 Hebert PDN, Penton EH, Burns JM, Janzen DH, Hallwachs W (2004) Ten species in one: DNA barcoding reveals cryptic species in the neotropical skipper butterfly Astraptes fulgerator. Proc Natl Acad Sci USA 101(41): 14812-14817 Heim I (2003) Synthetische Taxonomie zur Gattung Aplysina Nardo, 1834. Diploma Thesis. Fakultt fr Bio- und Geowissenschaften. Universitt Stuttgart, Stuttgart Heim I, Nickel M, Picton B, Brmmer F (2007) Description and molecular phylogeny of Tethya hibernica sp. nov. (Porifera, Demospongiae) from Northern Ireland with remarks on the European species of the genus Tethya. Zootaxa 1595: 1-15 Jang-Liaw NH, Tang KL, Hui CF, Shao KT (2002) Molecular phylogeny of 48 species of damselfishes (Perciformes: Pomacentridae) using 12S mtDNA sequences. Mol Phylogenet Evol 25(3): 445-454 King TL, Eackles MS, Gjetvaj B, Hoeh WR (1999) Intraspecific phylogeography of Lasmigona subviridis (Bivalvia: Unionidae): conservation implications of range discontinuity. Mol Ecol 8: S65S78 Kreuter MH, Robitzki A, Chang S, Steffen R, Michaelis M, Kljajic Z, Bachmann M, Schrder HC, Mller WEG (1992) Production of the cytostatic agent aeroplysinin by the sponge Verongia aerophoba in in vitro culture. Comp Biochem Physiol B 101(1): 183-187 Lavrov DV, Forget L, Kelly M, Lang F (2005) Mitochondrial genomes of two demosponges provide insights into an early stage of animal evolution. Mol Phylogenet Evol 22: 1231-1239 Lavrov DV, Lang F (2005) Transfer RNA gene recruitment in mitochondrial DNA. Trends Genet 21(3): 129-133 Nichols S, Barnes PAG (2005) A molecular phylogeny and historical biogeography of the marine sponge genus Placospongia (Phylum Porifera) indicate low dispersal capabilities and widespread crypsis. J Exp Mar Biol Ecol 323: 1-15 Page RDM (1996) TreeView: an application to display phylogenetic trees on personal computers. Comp Applic Biosci 12(4): 357-358 Pinheiro US, Hajdu E (2001) Shallow-water Aplysina Nardo (Aplysinidae, Verongida, Demospongiae) from the So Sebastio Channel and its environs (Tropical southwestern Atlantic), with description of a new species and a literature review of other Brazilian records of the genus. Rev Bras Zool 18(1): 143-160 Pont-Kingdon G, Okada NA, Macfarlane JL, Beagley CT, WatkinsSims CD, Cavalier-Smith T, Clark-Walker GD, Wolstenholme DR (1998) Mitochondrial DNA of the coral Sarcophyton glaucum contains a gene for a homologue of bacterial MutS: a possible case of gene transfer from the nucleus to the mitochondrion. J Mol Evol 46: 419-431 Posada D, Crandall KA (1998) MODELTEST: testing the model of DNA substitution. Bioinformatics 14(9): 817-818 Quan J, Zhuang Z, Deng J, Dai J, Zhang YP (2004) Phylogenetic relationships of 12 Penaeoidea shrimp species deduced from mitochondrial DNA sequences. Biochem Genet 42(9/10): 331-345 Quesada H, Warren M, Skibinski DOF (1998) Nonneutral evolution and differential mutation rate of gender-associated mitochondrial DNA lineages in the marine mussel Mytilus. Genetics 149: 15111526 Saeki BM, Granato AC, Berlinck RGS, Magalhaes A, Schefer AB, Ferreira AG, Pinheiro US, Hajdu E (2002) Two unprecedented

371

dibromotyrosine-derived alkaloids from the Brazilian endemic marine sponge Aplysina caissara. J Nat Prod 65: 796-799 Schmidt O (1862) Die Spongien des adriatischen Meeres. Wilhelm Engelmann Verlag, Leipzig Schmitt S, Hentschel U, Zea S, Dandekar T, Wolf M (2005) ITS2 and 18S rRNA gene phylogeny of Aplysinidae (Verongida, Demospongiae). J Mol Evol 60: 327-336 Shearer TL, van Oppen MJH, Romanos SL, Wrheide G (2002) Slow mitochondrial DNA sequence evolution in Anthozoa (Cnidaria). Mol Ecol 11: 2475-2487 Sober E (2002) Instrumentalism, parsimony, and the Akaike framework. Philos Sci 69 112-123 Tarjuelo I, Posada D, Crandall KA, Pascual M, Turon X (2001) Cryptic species of Clavellina (Ascidiacea) in two different habitats: harbours and rocky littoral zones in the northwestern Mediterranean. Mar Biol 139: 455-462 Thompson JD, Gibson TJ, Plewniak F, Jeanmaougin F, Higgins DG (1997) The Clustal X Windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res 24: 4876-4882 Thoms C (2004) Alkaloide in Schwmmen der Gattung Aplysina chemische kologie und assoziierte Bakterien. PhD Thesis. Pharmazeutische Biologie. Heinrich-Heine-Universitt, Dsseldorf Vacelet J (1959) Rpartition gnrale des ponges et systmatique des ponges cornes de la rgion de Marseille et de quelques stations mditerranennes. Rec Trav Sta mar Endoume 16(26): 39101 Vacelet J (1969) Rle des populations phytoplanctoniques et bactriennes dans le cycle du phospore et de lazote en mer et dans les flaques supralittorales du golfe de Marseille. Tethys 1(1): 5118 Vacelet J (1971) tude en microscopie lectronique de lassociation entre une cyanophyce chroococcale et une ponge du genre Verongia. J Microsc 12(3): 363-380

van Oppen MJH, Willis BL, Miller DJ (1999) Atypically low rate of cytochrome b evolution in the scleractinian coral genus Acropora. Proc Roy Soc Lond B 266: 179-183 van Oppen MJH, Wrheide G, Takahashi M (2002) Nuclear markers in evolutionary and population genetic studies of scleractinian corals and sponges. Proc 9th Int Coral Reef Symp, Bali 1: 131-138 Voultsiadou-Koukoura E (1987) Some remarks on the Mediterranean species of the genus Aplysina (Demospongiae, Verongida). In: Vacelet J, Boury-Esnault N (eds) Taxonomy of Porifera: From the N.E. Atlantic and Mediterranean Sea. Springer, Berlin, pp. 332 Ward RD, Zemlak TS, Innes BH, Last PR, Hebert PDN (2005) DNA barcoding Australias fish species. Phil Trans Roy Soc Lond B 360: 1847-1857 Wrheide G (1998) The reef cave dwelling ultraconservative coralline demosponge Astrosclera willeyana Lister 1900 from the Indo-Pacific. Micromorphology, ultrastructure, biocalcification, isotope record, taxonomy, biogeography, phylogeny. FACIES 38: 1-88 Wrheide G (2005) Low variation in partial cytochrome oxidase subunit I (COI) mitochondrial sequences in the coralline demosponge Astrosclera willeyana across the Indo-Pacific. Mar Biol 148(5): 907-912 Wrheide G, Degnan BM, Hooper JNA, Reitner J (2002a) Phylogeography and taxonomy of the Indo-Pacific reef cave dwelling coralline demosponge Astrosclera willeyana - new data from nuclear internal transcribed spacer sequences. Proc 9th Int Coral Reef Symp Bali 1: 339-346 Wrheide G, Hooper JNA, Degnan DM (2002b) Phylogeography of western Pacific Leucetta chagosensis (Porifera: Calcarea) from ribosomal DNA sequences: implications for population history and conservation of the Great Barrier Reef World Heritage Area (Australia). Mol Ecol 11: 1753-1768 Wrheide G, Nichols S, Goldberg J (2004) Intragenomic variation of the rDNA internal transcribed spacers in sponges (Phylum Porifera): implications for phylogenetic studies. Mol Phylogenet Evol 33(3): 816-830

Porifera research: Biodiversity, innovation and sustainaBility - 2007

373

Salting sponges: a reliable non-toxic and costeffective method to preserve poriferans in the field for subsequent DNA-work
Isabel Heim(*), Jrg U. Hammel, Michael Nickel, Franz Brmmer
Universitt Stuttgart, Biologisches Institut, Abteilung Zoologie, Pfaffenwaldring 57, 70569 Stuttgart, Germany.() isabel.heim@bio.uni-stuttgart.de, hammel@porifera.net, nickel@porifera.net, franz.bruemmer@bio.uni-stuttgart.de Abstract: For preservation of sponge material, ethanol between 70% and 99.8% is used in most cases. The advantage of using ethanol is that, even after several years, subsequent taxonomic studies as well DNA isolation is possible. In recent years, safety regulations of the International Air Transport Association (IATA) were gradually increased as countermeasure to security issues on airplanes. Since 70% ethanol has a flash point < 23C, it belongs to dangerous goods and underlies special regulations. For shipping, it is dictated to pack samples in ethanol by special trained personal, which is circumstantial and expensive. Consequently, the aim of this study was to develop a non-toxic and cost-effective method for sponge tissue and DNA preservation for short-term storage, e.g. for transportation purpose. Among biological conservation liquids used for terrestrial arthropod sampling, high concentrated NaCl solutions are in use. We adapted this method for poriferans and implemented a case study on Tethya aurantium. We tested different storage condition in the salt solution, subsequently isolated DNA, amplified a part of cytochrome oxidase subunit I (COI) and sequenced specific PCR products for verification. Our results demonstrate that high concentrated NaCl solutions conserve DNA as well as overall morphologic features of sponges. It is non-toxic and not conflicting recent IATA regulations. In addition, it is inexpensive and always on-hand to transport sponge material, even in the most remote sampling locations. Thus, we recommend this method for standard short-term shipping and transportation of sponge material. Keywords: IATA safety regulations, NaCl conservation, sample transportation, shipping, sponge preservation

Introduction
Due to increasing problems with safety issue on airplanes, the International Air Transport Association (IATA) continually adjusted and strengthened regulations over the past years. Consequently, strict regulations were enacted on how to pack flammable or toxic liquids for transport in airplanes (IATA 2006: Section 5). Following the strict rules for sending biological samples in 70% ethanol, specially trained personal is required which is time-consuming and expensive (IATA 2006: Section 1.5). Organizations, which violate these regulations, will be blacklisted (Dr. Michael Rannenberg, personal communication). Obviously, this leads to severe problems for airmail shipping as well as personal transport of various biological samples, which are indented for taxonomic, and/or DNA studies. Facing these major problems, it also has to be kept in mind that today even surface mail within international and national postal services rely on air transportation.
() J.U. Hammel and M. Nickel present address: Friedrich-Schiller-Universitt Jena, Institut fr Spezielle Zoologie und Evolutionsbiologie mit Phyletischem Museum, Erbertstr. 1, 07743 Jena, Germany

For transportation of sponge specimens for DNA isolation and taxonomic studies, different methods exist. The most common method is to preserve them in 80% to 99.8% ethanol (Borchiellini et al. 1998, Chombard et al. 1998, Duran et al. 2004, Nichols 2005), which directly relates to IATA safety regulations. Another procedure is to freeze samples in liquid nitrogen and to transport either with liquid nitrogen or with dry ice (Chombard et al. 1998, Schmitt et al. 2005). Meanwhile, this method is also subject to restrictions in airplanes. Furthermore, for DNA isolation it is possible to use silica gel or 8 M guanidinium chloride solution (Lavrov and Lang 2005, Wrheide 2005). However, for histological studies it is necessary to preserve another part in ethanol, because e.g. guanidinium chloride solution degrades the sponge tissue. Another possibility is to transport sponges alive in seawater to the lab, but this procedure is only useful for transfers up to two days (Perovic et al. 1999). For shipping sponge material at room temperature, we searched for a non-toxic, non-inflammable, and cost-saving method. From another field of research, we came across a possible solution. For catching spiders and harvestmen in winter, it is common practice to capture them with ground traps filled with a high concentrated salt solution (Teichmann 1994). The animals do not freeze and the NaCl solution does not evaporate like ethanol. Moreover, a positive secondary

374

effect was achieved, as the DNA isolation was equally or even more effective than with 70% ethanol (Axel Schnhofer, personal communication). Here we showed, by using PCR tests and sequencing of cytochrome oxidase subunit I (COI) of Tethya aurantium (Pallas, 1766) after varying storing conditions and times, that the NaCl-preservation method is also suitable and highly effective for sponges. The procedure is easy and NaCl is always on-hand. Moreover, after the transport it is possible to isolate directly DNA or to put the samples in ethanol for later works.

Cloning
PCR-products were cleaned with NucleoSpin Extract Kit (Machery-Nagel, Dren, Germany) before they were ligated into pCR4-TOPO vector (Invitrogen, Carlsbad, Canada) and transformed by heat-shock into competent E. coli One ShotTOP10 cells (Invitrogen). Plasmid DNA was isolated with NucleoSpin Plasmid Quick Pure Kit (Machery-Nagel). The correct insert size was verified by using agarose gel electrophoresis following an insert check with internal and M13-forward primers.

Material and Methods Sampling

Sponge samples of T. aurantium were collected during fieldwork within the research project BIOTECmarin by SCUBA diving following the C.M.A.S. (Confdration Mondiale des Activits Subaquatiques; World Underwater Federation) safety rules avoiding decompression dives in the Limski kanal near Rovinj/Croatia. The specimens were deposited in the BIOTECmarin database under ID numbers 553 and 875.

DNA sequencing
AGOWA GmbH (Berlin, Germany) performed the sequencing. Sequencing reactions were done with the M13 forward or M13 reverse primers. The nucleotide sequence data reported in this paper have been deposited in the GenBank nucleotide sequence database with accession number EF093529 to EF093531. We performed BLAST searches in GeneBank for all sequences. The sequences matched best with T. actinia (AY320033). No COI sequences of T. aurantium were available in GenBank prior to our study. Sequences were edited using BioEdit (Hall 1999). Sequence alignments were performed using the multiple sequence alignment program CLUSTAL X (Thompson et al. 1997).

Storage conditions
One individual of T. aurantium (ID 553) was cut into fragments. The samples were stored under different conditions: one fragment of T. aurantium was stored in 80% Ethanol, one in NaCl solution (300 g/l solved in ddH2O) at room temperature, one in NaCl solution at 4C, and one in a combination of NaCl solution at room temperature for one year followed by 70% Ethanol. Another specimen (ID 875) was used for the control experiment.

Results
In our experiments, we tested different storage conditions for the sponge species T. aurantium: 80% EtOH at room temperature, NaCl solution (300 g/l solved in ddH2O) at room temperature and 4C, and a combined storage of 340 days in NaCl solution at 4C with subsequent transfer to 70% EtOH. For every experimental approach, we accomplished DNA isolation and amplification of the cytochrome oxidase subunit I (Tab. 1 and Fig. 1). As a control experiment, we took living sponge material, which was maintained for 7 weeks in the aquarium. For the sponge material in 80% EtOH as well as for the NaCl solution at room temperature, we measured DNA concentrations between 3 ng/l and 82 ng/l. The amplification of the COI was successful for all DNA preparations instead of the 80% EtOH material after 14 days, but this could be a technical mistake. For long-term storage, we isolated the DNA for the 80% EtOH sample after 349 days, for the NaCl solution at 4C after 343 days and for the NaCl solution at RT after 342 days. COI amplification was successful in all cases. In combined storage, we conserved the sponge material for 340 days in NaCl solution at 4C and room temperature and subsequently transferred it in 70% EtOH. After that, we isolated DNA and amplified COI from the sponge material stored for 7 and 14 days in 70% EtOH. We measured DNA concentrations between 1.3 ng/l and 9.3 ng/l. Amplification of COI was successful in both experiments. As a control experiment, we sequenced the PCR products for the living material, short-term storage at day 76 and for the material stored for 343 days in the NaCl solution at 4C.

DNA extraction and amplification

Prior to starting DNA isolation, the specimens were transferred from salt solution into 70% ethanol (EtOH) to rinse out NaCl. We completed an increasing ethanol series: three times for 15 minutes in 70% EtOH, one time for 15 minutes in 80% EtOH, one time for 15 minutes in 90% EtOH and one time for 15 minutes in 99,8% EtOH. To eliminate the complete ethanol samples were frozen in liquid nitrogen. Afterwards DNA was extracted using the Qiagen DNeasy tissue extraction kit (Qiagen, Hilden, Germany) according to the manufacturers instructions. DNA concentration was measured with NanoDrop Spectrophotometer ND-1000 (Peqlab, Erlangen, Germany). For the amplification of a part of the cytochrome oxidase subunit I gene (COI) the primers COI-For and COI-Rev were used (Folmer et al. 1994). Amplification of the poriferan mtDNA was performed in 50 l total reaction volume, with 1 l of each primer (10 M), 4 l dNTPs (2.5 mM) , 5 l 10x buffer containing 15 mM MgCl2, 2 U Taq polymerase (TaKaRa BIO INC., Shiga, Japan) and 4 10 l template DNA. An initial denaturation at 94C for 2 min was followed by 35 cycles (94C 50 s, 40C for 55 s and 72C for 1 min), and a final extension at 72C for 7 min.

375

Fig. 1: Cytochrome oxidase subunit I (COI) PCR amplifications. a. Control, b. short-term storage after 76 days in 300 g/l NaCl at RT and c. long-term storage after 343 days in 300 g/l NaCl at 4C. Sequences obtained from these PCR products are shown in Fig. 2. An overview of experimental conditions is given in Table 1. Markers: HL I: HyperLadder I (Bioline, Luckenwalde, Germany); HL II: HyperLadder II (Bioline); GL: GeneLadder 1kbp (Genaxxon, Biberach, Germany).

The alignment of all sequences is shown in Fig. 2. For all samples, we got the identical sequence demonstrating that DNA is preserved well in all experimental setups.

Discussion
The aim of this study was to find a cost-saving and easy way to transport sponge material from the sea to the laboratory. For taxonomy studies as well as DNA isolation, it

is standard to preserve them in 70% ethanol. Another method to conserve tissue material of marine invertebrates is DMSONaCl or CTAB-NaCl. Both solution are saturated with NaCl, but DMSO-NaCl have the widest range in preserve marine invertebrate tissue and to get an amplifiable high molecular weight DNA (Dawson et al. 1998). Nevertheless, in the last year the laws of the International Air Transport Association (IATA) have been more rigorous due to upcoming safety issues. The Dangerous goods regulations of IATA stipulate that toxic and flammable liquids like ethanol and phenol have to be packed in a special way for transport in planes (IATA 2006: Section 5). The packing of the samples on this account is time-consuming and expensive. Therefore, we turned our attention to a method for short- and middle-termed transport at room temperature to transfer sponge samples to the laboratory. The high concentrated NaCl-solution is easy to obtain, low priced, non-flammable and non-dangerous according to IATA regulations (IATA 2006). Over one year we could observe a decrease of DNA concentration in the salt solution at room temperature as well as at 4C. However, in fresh as well as in ethanol preserved sponge material variations in DNA concentration was detected, too. Nevertheless, in all cases it was possible to amplify the cytochrome oxidase subunit I and to get identical sequence information of the cloned PCR products. In addition, the sample of T. aurantium, which has been stored in high NaCl for nearly one year, preserved his live colour, which is not the case in ethanol. Furthermore, we observed that high NaCl storage even eases DNA extraction: proteinase K digest seems to be faster in comparison to living or ethanol preserved material (data not shown). Eventually, the NaCl solution damaged the membranes of the sponge cells due to osmotic effects. Another positive effect is that a high concentration of NaCl inhibits nucleases (Seutin et al.

Table 1: Experimental settings and results of short- and long-term storage under four different conditions, concerning DNA extraction and PCR. Control sequencing of PCR products are indicated, amplifications are shown in Fig. 1, for sequencing results refer to Fig. 2. Experiment Control experiments Short term storage Conditions living material 80% EtOH, RT 300 g/l NaCl, RT Storage (days) 0 6 14 8 14 33 76 90 349 343 342 340+7 340+14 340+7 340+14 DNA-Conc. (ng/l) 14 69 29 82 12 5.3 13 3 83 8 4 7.4 1.3 9.3 3.3 PCR (Fig. 1) + (a) + + + + + (b) + + + (c) + + + + + Seq. Fig.2 Fig.2 Fig.2 -

Long term storage

Combined storage

80% EtOH, RT 300 g/l NaCl, 4C 300 g/l NaCl, RT 340 d in 300 g/l NaCL, 4C, afterwards 70% EtOH 340 d 300 g/l NaCL, RT, afterwards 70% EtOH

376

Fig. 2: Alignment of the three sequenced PCR products of T. aurantium: control, short-term storage after 76 days in 300 g/l NaCl at RT and long-term storage after 343 days in 300 g/l NaCl at 4C.

1991). After the transport in the salt solution, it is possible to transfer the sponge material in 70% ethanol for museum collections. Nevertheless, it is also feasible to isolate DNA and to make a positive PCR. In relation to our starting initial research related to spider preservation in pitfall traps and our results on sponges presented here, we assume that high NaCl storage could be suitable for all kind of marine invertebrates. In conclusion, high NaCl storage is a straightforward and cheap way to store sponges for transport without raising safety issues. It was designed particularly to facilitate transport of freshly collected material from the field to the lab for taxonomic and DNA studies. Nevertheless, even though it has not been tested yet, we assume that short term NaCl preservation will also work for material preserved and stored in ethanol beforehand, therefore it should be suitable for museum material, too. Modifications of the method, e.g. increasing NaCl concentrations up to 350 g/l or adding antifungal and antibacterial agents, may be necessary for other groups of sponges or other groups of marine invertebrates, to suit special transportation and storage conditions. Care has to be taken that no new safety issues are raised by these additives. Finally, even if the high NaCl method allows for relatively easy transport and shipping of sponge material it has to be assured that no regulations of the Convention on

Biological Diversity are violated (Brmmer and Nickel 2002, UNEP 2000).

Acknowledgements
We would like to thank Renato Batel and his co-workers (Center of Marine Research Rovinj; Institut Ruder Bokovi, Zagreb) and Anne Klppel (Universitt Stuttgart, Germany) for their support collecting sponge specimens. Michael Rannenberg (Universitt Stuttgart, Germany) for information on IATA dangerous goods regulations. Axel Schnhofer (Universitt Mainz, Germany) for information on harvestman pitfall traps and Hans-Dieter Grtz (Universitt Stuttgart, Germany) for support of this work. Gisela Fritz (Universitt Stuttgart, Germany) for critical reading of our manuscript. This study was supported by the project BIOTECmarin (03F0414D) funded by the German Federal Ministry of Education and Research and the University of Stuttgart.

References
Borchiellini C, Boury-Esnault N, Vacelet J, Le Parco Y (1998) Phylogenetic analysis of the Hsp70 sequences reveals the monophyly of Metazoa and specific phylogenetic relationships between animals and fungi. Mol Biol Evol 15(6): 647-655 Brmmer F, Nickel M (2002) Nachhaltige Nutzung mariner Schwmme. In: Krn H, Feit U (eds). Treffpunkt Biologische Vielfalt II. Interdisziplinrer Forschungsaustausch im Rahmen

377

des bereinkommens ber die biologische Vielfalt. Bundesamt fr Naturschutz, Bonn, pp. 183-189 Chombard C, Boury-Esnault N, Tillier S (1998) Reassessment of homology of morphological characters in Tetractinellid sponges based on molecular data. Systematic Biol 47(3): 351-366 Dawson MN, Raskoff KA, Jacobs DK (1998) Field preservation of marine invertebrate tissue for DNA analyses. Mol Mar Biol Biotechnol 7(2): 145-152 Duran S, Giribet G, Turon X (2004) Phylogeographical history of the sponge Crambe crambe (Porifera, Poecilosclerida): range expansion and recent invasion of the Macaronesian islands from the Mediterranean Sea. Mol Ecol 13: 109-122 Folmer O, Black M, Hoch W, Lutz R, Vrijenhoek R (1994) DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol Mar Biol Biotechnol 3: 294-299 Hall TA (1999) BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucl Acids Symp Ser 41: 95-98 IATA (2006) Dangerous Goods Regulations (IATA-Resolution 618 Attachment A). Effective 1 January - 31 December 2007. Montreal Geneva Lavrov DV, Lang F (2005) Transfer RNA gene recruitment in mitochondrial DNA. Trends Genet 21(3): 129-133 Nichols S (2005) An evaluation of support for order-level monophyly and interrelationships within the class Demospongiae using partial

data from the large subunit rDNA and cytochrome oxidase subunit I. Mol Phylogenet Evol 34: 81-96 Perovic S, Krasko A, Prokic I, Mller I, Mller WEG (1999) Origin of neuronal-like receptors in Metazoa: Cloning of a metabotropic glutamate/GABA-like receptor from the marine sponge Geodia cydonium. Cell Tissue Res 296(2): 395-404 Schmitt S, Hentschel U, Zea S, Dandekar T, Wolf M (2005) ITS2 and 18S rRNA gene phylogeny of Aplysinidae (Verongida, Demospongiae). J Mol Evol 60: 327-336 Seutin G, White BN, Boag PT (1991) Preservation of avian blood and tissue samples for DNA analyses. Can J Zool 69: 82-90 Teichmann B (1994) Eine wenig bekannte Konservierungsflssigkeit fr Bodenfallen. Entomol Nachr Ber 38: 2530 Thompson JD, Gibson TJ, Plewniak F, Jeanmaougin F, Higgins DG (1997) The Clustal X Windows interface: flexible strategies for multible sequence alignment aided by quality analysis tools. Nucleic Acids Res 24: 4876-4882 UNEP (2000) Sustaining live on earth. How the Convention on Biological Diversity promotes nature and human well-being. Secretariat of the Convention on Biological Diversity, New York, pp. 21 Wrheide G (2005) Low variation in partial cytochrome oxidase subunit I (COI) mitochondrial sequences in the coralline demosponge Astrosclera willeyana across the Indo-Pacific. Mar Biol 148(5): 907-912

Porifera research: Biodiversity, innovation and sustainaBility - 2007

379

Oxygen distribution in Tentorium semisuberites and in its habitat in the Arctic deep sea
Friederike Hoffmann(1*), Eberhard Sauter(2), Oliver Sachs(2), Hans Ry(1), Michael Klages(2)
Max Planck Institute for Marine Microbiology, Celsiusstr. 1, 28359 Bremen, Germany. fhoffman@mpi-bremen.de, hroey@mpi-bremen.de (2) Alfred Wegener Institute for Polar and Marine Research, Columbusstrae, 27568 Bremerhaven, Germany. esauter@awi-bremerhaven.de, osachs@awi-bremerhaven.de, mklages@awi-bremerhaven.de
(1)

Abstract: The arctic deep-sea morphotype of the hadromerid sponge Tentorium semisuberites is common in the Arctic deep sea, where it lives partially buried in soft bottom sediments. To investigate the chemical microenvironment of sponge cells and associated microbes, oxygen gradients in sponge tissue were measured with Clark-type microelectrodes. Profiles with step resolutions between 100 and 500 m were measured vertically through the sponge body (4 mm). Similar profiles were measured in sediments of the sponge sampling sites. The characteristic shape of the profiles showed that sponges were alive and pumping during profiling. Oxygen concentrations were highest at the sponge surface and decreased towards the centre of the sponge, which vertically coincided with the sediment-water interface. Below, oxygen was found to increase again. The lowest oxygen concentration measured in T. semisuberites was 53 M (15% of surrounding water). Oxygen concentrations from 2 mm above the sediment surface until 2 mm into the sediment ranged from 300 270 M. A part of the water filtered by this species is presumably sediment pore water, which is only slightly lower in oxygen than the overlaying bottom water. The sponge tissue thus provides an oxic to hypoxic habitat for the associated Archaea and Bacteria. Keywords: Arctic deep sea, oxygen microelectrodes, microenvironments, sponge microbes, Tentorium semisuberites

Introduction
Sponges amount to a large part of the macrobenthos in the Arctic deep sea. The arctic deep-sea morphotype of Tentorium semisuberites (Hadromerida, Demospongiae) is among the most common sponge species of Arctic deep sea soft bottom sediments (Barthel and Tendal 1993). It lives partially buried in the sediment, using tiny stones as substrate (e.g Fram Strait, west off Spitzbergen), or anchoring with long root-like spicules in the sediment (e.g. central Greenland Sea). The arctic deep-sea morphotype of T. semisuberites is cone shaped, 2-5 mm across and 4-5 mm high (Fig. 1A). In shallower waters (ca. 30-600 m) along the Norwegian coast T. semisuberites grows on hard substrate and reaches more than twice the size of arctic specimens. In a recent study combining microbial lipid biomarker analysis and differential fluorescence in situ hybridisation (FISH) on sponge sections (Pape et al. 2006), we showed that Archaea provide a major and Bacteria a minor part of the microbial endobiont community of T. semisuberites specimens from the Hausgarten region. Archaea and Bacteria were evenly distributed throughout the entire sponge body. Indications for Archaea as endobionts have also been reported for several species of the Demospongiae (Preston et al. 1996, Webster et al. 2001, Margot et al. 2002, Lee et al. 2003) and, furthermore, for the Hexactinellida (Thiel et al. 2002, Pape et al. 2004). To investigate the chemical microenvironment of sponge cells and associated microbes,

we measured the oxygen distribution in the tissue of T. semisuberites as well as in the sediment and bottom water of its habitat.

Material and Methods

During Expedition AWI-ATL with R/V LAtalante (September 2005), sponges were sampled in the Fram Strait on the continental rise off Svalbard (7904.34 N, 408.2 E) at 2440 m depth (Fig. 2). This site is part of the long-term monitoring program Hausgarten of the Alfred Wegener Institute Bremerhaven, Germany, and is revisited annually (Soltwedel et al. 2005). Samples were taken by Slurp Gun operated by the ROV Victor 6000 (Fig. 1B). Sponges were kept submersed in ambient bottom water during the rest of the ROV dive. After retrieval of the ROV, sponge specimens were immediately placed in an aquarium filled with bottom water from the sampling site which was kept at in situ temperature (0 to -1 C). Oxygen gradients over the surface and into the tissue of T. semisuberites were measured with two specimens of T. semisuberites using Clark-type oxygen microelectrodes (tip diameters 18-30 m) as described (Schnberg et al. 2004). Adaptation time of sponges in the aquarium prior to lab measurements was 7 hours. All measurements were taken within 40 hours after sampling. Profiles with step-resolutions between 100 and 500 m were measured vertically through the sponge body (4 mm).

380

Fig. 2: Sampling site in the Fram Strait on the continental rise off Svalbard, Arctic Sea at 2240 m.

Fig. 1: A. Tentorium semisuberites specimens sampled at 2440 m depth. B. Non-intrusive and precise sampling with a slurp gun operated by the ROV Victor 6000 at 2440 m depth. Tentorium semisuberites (arrow) lives partially buried in the sediment and is difficult to detect (Copyright: Ifremer, France).

Oxygen gradients in the sediment were measured in situ at 2440 m depth in direct proximity of the sponge sampling sites with the in situ microprofiler MIC, operated by the ROV Victor 6000 (Sauter et al. 2004, Soltwedel et al. 2005, DeBeer et al. 2006). To measure in situ oxygen gradients in sponge tissue was not possible due to optical and mechanical limitations.

Results
In total, 21 oxygen profiles were measured over the surface and into the tissue of the two sponge specimens, some of them through the entire sponge body from the top to the basis (see Fig. 3). Though the water current created by sponge pumping activity was too low to be visualised with dye, it was obvious

from oxygen profiles that sponges were alive and pumping. Non-pumping sponges show diffusive oxygen profiles, with decreased oxygen concentrations above their surface due to diffusive oxygen fluxes over the boundary layer, and a steep oxygen decrease down to zero in the first millimeter of the tissue (Hoffmann et al. 2005a, Hoffmann et al. 2005b, Schlppy et al. in press). In our investigation, water was oxygen saturated until the sponge surface, and concentrations decreased gradually into the tissue. This is a typical pattern for pumping sponges (Hoffmann et al. 2005b, Schlppy et al. in press). Typically, oxygen concentrations decreased towards the centre of the sponge, reaching a minimum 2-3 mm into the sponge tissue. The lowest concentration measured was 53 M, which is 15% of the oxygen concentration of the surrounding water. This trend was visible in all profiles measured, though the actual oxygen concentrations could vary at the same position when there were several hours between the measurements. Profiles measured in direct sequence, however, were nicely reproducible. Figure 3 shows a series of replicated oxygen profiles measured vertically through the sponge body. Two parallel profiles each were measured at three different positions ranging from close to the outer wall until close to the osculum. Pores for incurrent water were found

381

Fig. 3: Left: 3x2 replicated oxygen profiles measured vertically through the entire body of T. semisuberites kept in the aquarium under in situ conditions. Sketch shows positioning of the profiles. Arrows show direction of water flow through the sponge body. Oxygen concentrations are highest at the sponge surface, where oxygen-rich water enters the sponge from both the top and the side. Oxygen concentration decreases towards the center of the sponge which vertically coincided with the sediment-water interface. Below, oxygen was found to increase again. Right: Three replicated oxygen profiles at the sediment-water interface from 2 mm above the sediment surface until 2 mm into the sediment. Oxygen concentrations ranged from 300 to 270 M (88-79% oxygen concentration compared to that of bottom water used for the aquarium experiments). Therefore, in situ oxygen concentrations in sponge tissue can be expected to be only slightly lower than those measured in aquarium experiments.

both at the porous surface next to the oscule, and to a lower amount also at the side surfaces of the sponge by microscopic investigation. Tissue oxygen concentrations were highest at the sponge surface, where oxygen-rich water entered the sponge from both the top and the side, as indicated by arrows. Oxygen concentration decreased towards the middle section of the sponge, which vertically coincided with the sedimentwater interface. Below, oxygen was found to increase again. In contrast to the gradients measured within the sponge body, the pore water oxygen concentration in the ambient sediment only decreased slightly from 300 M above the sediment-water interface to 270 M in 2 mm sediment depth. The oxygen penetration depth in the sediment was not reached by the in situ measurements which were performed down to 90 mm sediment depth.

Discussion
The oxygen minimum (53 M) is most likely due to higher respiration rate or lower pumping activity in the middle section of the sponge, the area where the sponge intersects the sediment-water interface. The higher oxygen concentrations towards the sponge bottom may be explained by the conical shape; when inserted perpendicular to the upper surface, the electrode slightly approaches the side surface, which is reflected by higher oxygen concentrations in the profiles. An alternative explanation is a higher filtration activity in the bottom region of the sponge. Tentorium semisuberites usually

lives half buried in the sediment (see Fig. 1B). If it filtrates over its entire surface, as both microscopic investigation and oxygen profiles indicate, or even increases its pumping activity in the lower part, a large part of water filtered by the sponge is actually sediment pore water, which usually is richer in nutrients and organic matter than overlaying bottom water. With a minimum of 270 M (79% saturation) the oxygen concentration below the sediment-water interface where T. semisuberites lives is only slightly lower than the bottom water concentration (300 M; 88% saturation - Fig. 3). Assuming a similar metabolism (similar pumping rates) in situ and in the laboratory, the tissue of T. semisuberites will then show oxygen concentrations in situ only slightly lower than those we measured in the aquarium, where the sponge specimens were entirely surrounded by bottom water with 340 M oxygen. From these results, it seems that Bacteria and Archaea associated with Tentorium semisuberites usually live in oxic to hypoxic conditions.

Acknowledgements
The team from GENAVIR/IFREMER is acknowledged for the expert operation with ROV Victor. We thank the crew and officers of R/V LAtalante for their excellent support, Uta Wiesner for performing Winkler titration, and the technicians of the microsensor group at MPI for preparation of superb microsensors. Thomas Soltwedel is kindly acknowledged for spotting tiny sponges at 2440 m depth. This study was supported by the EU - project HERMES and by the

382

Max Planck Society. F.H. and O.S. were funded by the Deutsche Forschungsgemeinschaft (DFG, Project No. HO 3293/1-1 and SA 1030/1-3, respectively).

References
Barthel D, Tendal OS (1993) The sponge association of the abyssal Norwegian-Greenland Sea: species composition, substrate relationships and distribution. Sarsia 78: 83-96 DeBeer D, Sauter E, Niemann H, Kaul N, Foucher J, Witte U, Schlter M, Boetius A (2006) In situ fluxes and zonation of microbial activity in surface sediments of the Hkon Mosby Mud Volcano. Limnol Oceanogr 51: 1315-1331 Hoffmann F, Larsen O, Rapp HT, Osinga R (2005a) Oxygen dynamics in choanosomal sponge explants. Mar Biol Res 1: 160163 Hoffmann F, Larsen O, Thiel V, Rapp HT, Pape T, Michaelis W, Reitner J (2005b) An anaerobic world in sponges. Geomicrobiol J 22: 1-10 Lee E-Y, Lee HK, Lee YK, Sim CJ, Lee J-H (2003) Diversity of symbiotic archaeal communities in marine sponges from Korea. Biomol Eng 20: 299-304 Margot H, Acebal C, Toril E, Amils R, Puentes JLF (2002) Consistent association of crenarchaeal Archaea with sponges of the genus Axinella. Mar Biol 140: 739-745 Pape T, Blumenberg M, Thiel V, Michaelis W (2004) Biphytanes as biomarkers for sponge-associated Archaea. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Bull Mus Ist Biol Univ Genova 68: 509-515 Pape T, Hoffmann F, Queric N-V, Juterzenka Kv, Reitner J, Michaelis W (2006) Dense populations of Archaea associated with

the hadromerid demosponge Tentorium semisuberites from Arctic deep waters. Polar Biol 29: 662-667 Preston CM, Wu KY, Molinski TF, DeLong EF (1996) A psychrophilic crenarcheon inhabits a marine sponge: Cenarchaeum symbiosum gen. nov., sp. nov. Proc Natl Acad Sci USA 93: 6241-6246 Sauter E, Baumann L, Wegner J, Delius J (2004) Geochemical and hydrodynamic investigations at the sediment-water interface. Rep Polar Mar Res 488: 233-236 Schlppy M-L, Hoffmann F, Ry H, Wijffels RH, Mendola D, Sidri M, Beer Dd (in press) Oxygen dynamics and flow patterns of Dysidea avara (Porifera, Demospongiae). J Mar Biol Assoc UK Schnberg CHL, Hoffmann F, Gatti S (2004) Using microsensors to measure sponge physiology. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Bull Mus Ist Biol Univ Genova 68: 593-604 Soltwedel T, Bauernfeind E, Bergmann M, Budaeva N, Hoste E, Jaeckisch N, Juterzenka Kv, Matthiessen J, Mokievsky V, Nthig E-M, Queric NV, Sablotny B, Sauter E, Schewe I, UrbanMalinga B, Wegner J, Wlodarska-Kowalczuk M, Klages M (2005) Hausgarten - Multidisciplinary investigations at a deep-sea, longterm observatory in the Arctic Ocean. Oceanography 18: 47-61 Thiel V, Blumenberg M, Hefter J, Pape T, Pomponi S, Reed J, Reitner J, Wrheide G, Michaelis W (2002) A chemical view of the most ancient metazoa - biomarker chemotaxonomy of hexactinellid sponges. Naturwissenschaften 89: 60-66 Webster NS, Watts JEM, Hill RT (2001) Detection and phylogenetic analysis of novel Crenarcheote and Euryarcheote 16S Ribosomal RNA gene sequence from a Great Barrier Reef sponge. Mar Biotechnol 3: 600-608

Porifera research: Biodiversity, innovation and sustainaBility - 2007

383

Phylogenetic relationships between freshwater and marine Haplosclerida (Porifera, Demospongiae) based on the full length 18S rRNA and partial COXI gene sequences
Valeria Itskovich(1*), Sergey Belikov(1), Sofia Efremova(2), Yoshiki Masuda(3), Thierry Perez(4), Eliane Alivon(4), Carole Borchiellini(4), Nicole Boury-Esnault(4)
Limnological Institute of the Siberian Branch of Russian Academy of Sciences, Ulan-Batorskaya 3, RUS-664033. Irkutsk, Russia. itskovich@mail.ru (2) Laboratory of Ontogenesis, Biological Research Institute, St. Petersburg State University, Oranienbaumskoe sch. 2, Stary Peterhoff, 198 904. St. Peterburg, Russia (3) Department of Biology, Kawasaki Medical School, Kurashiki, 701-0192. Japan (4) Aix-Marseille Universit, CNRS UMR-6540 DIMAR, Centre dOcanologie de Marseille. Station Marine dEndoume. Rue Batterie des Lions, 13007. Marseille, France
(1)

Abstract: The order Haplosclerida comprises five marine sponge families and all freshwater sponge families. Taxonomy within this order remains unclear. To study phylogenetic relationships within the order Haplosclerida and to investigate origin of freshwater sponges, sequences from the partial COXI gene and of full length 18S rRNA of 10 sponge species were obtained. The new sequences and available sequences of other Porifera from GenBank were used for phylogenetic analyses. Results based on 18S rRNA data indicate that the order Haplosclerida is not monophyletic. All freshwater sponge species clustered on the phylogenetic tree together in one monophyletic group. The obtained phylogenetic trees based on the COXI data do not support the monophyly of the Haplosclerida. The obtained phylogenetic trees based on the 18S rRNA and COXI gene data are congruent, and contradict the existing hypothesis that freshwater sponges are polyphyletic and support the monophyly of Spongillina, whereas the monophyly of the suborders Haplosclerina and Petrosina was not supported. Keywords: Spongillina, Lubomirskiidae, molecular phylogeny, cytochrome oxidase, 18S ribosomal RNA

Introduction
Demosponges inhabit marine and freshwater habitats. Seven freshwater sponge families and five marine sponge families constitute the order Haplosclerida Topsent, 1928 (Hooper and van Soest 2002). The definition of that order following Hooper and van Soest (2002) is: Demospongiae in which the main skeleton is partially or entirely composed of an isodictyal-isotropic or anisotropic, occasionally alveolate reticulation of spongin fibres and/or spicules, with uni- to multispicular tracts of diactinal spicules forming triangular, rectangular or polygonal meshes. The history of classification in this order is very complicated (see Bergquist 1980, de Weerdt 1985, Manconi and Pronzato 2002). Bergquist (1980) proposed to exclude from the order Haplosclerida a group of genera (among which Petrosia, Xestospongia, Calyx, Strongylophora, Oceanapia etc) on the basis of their heavy spicule content, reproductive strategy (oviparity) and biochemical characteristics. Berquist erected for this group of genera a new order called Nepheliospongida. This new order was not accepted by de Weerdt (1985) and de Weerdt

and van Soest (1986), although de Weerdt (1989) admitted a sister relationship between the families Oceanapiidae and Petrosiidae within the order Haplosclerida. Furthermore, the study of sterol chemistry (Fromont et al. 1994) did not support the two orders. In Systema Porifera (Hooper and van Soest 2002) three suborders are recognized within the order Haplosclerida: Haplosclerina, Petrosina (=Nepheliospongida sensu Bergquist) and Spongillina (freshwater sponges). Several molecular biology studies based on 18S rRNA, 28S rRNA and cytochrome oxidase subunit I (COXI) sequences (McCormack et al. 2002, Borchiellini et al. 2004, Erpenbeck et al. 2004, Nichols 2005, Redmond et al. 2007) found that the order Haplosclerida could be polyphyletic and therefore its taxonomy should be revised. For that, a larger number of species should be analyzed and many Haplosclerida genera have to be included into the molecular taxonomy works. In a recent revision of sponge taxonomy, freshwater sponges were divided into 7 families within the suborder Spongillina: Spongillidae (21 genera), Potamolepidae (6 genera), Lubomirskiidae (4 genera), Malawispongiidae (5 genera), Metschnikowiidae (1 genus), Metaniidae (5 genera)

384

and Palaeospongillidae (fossil family) (Manconi and Pronzato 2002). The phylogenetic relationships within Spongillina are unclear. The origin of freshwater sponges and their phylogenetic relationships with other haplosclerid or demosponge families are questions which remain unresolved. Do the freshwater species share a common ancestor? With what other clades do they share a more recent common ancestor? Brien (1970) and Volkmer-Ribeiro and de Rosa-Barbosa (1979) observed that the freshwater family Potamolepidae shares several characteristics with Hadromerida. Volkmer-Ribeiro (1990) hypothesized that the freshwater family Metaniidae shares several characters with Acarnidae (Poecilosclerida). If the hypothesis of Brien and Volkmer-Ribeiro is correct, then freshwater sponges are polyphyletic. From more than 200 existing species of freshwater sponges, only about 10 species, belonging to 7 genera representing half of the existing families were used in the molecular systematics studies so far. In the phylogenetic trees produced with this group of species they appear as monophyletic (Itskovich et al. 1999, Efremova et al. 2002, Schrder et al. 2003, Addis and Peterson 2005, Itskovich et al. 2006, Redmond et al. 2007). The endemic family Lubomirskiidae is restricted to the ancient Lake Baikal (Siberia). This lake is the deepest (1647m) and the oldest (about 30 million years old) lake in the world (Mats 1993, Timoshkin 1995). Lake Baikal is famous by its high level of endemism (70% species are endemics) and is a hot spot of biodiversity (Timoshkin 1995). Fourteen endemic sponge species are known from this area and this represent a high level of biodiversity among ancient lakes (Masuda 1999, Efremova 2001, 2004). Phylogeny of Lubomirskiidae is unknown. In this work we used 18S and COX1 sequences from GenBank, complemented with new sequences of Haplosclerida, including species of families Lubomirskiidae, Potamolepidae and marine Haplosclerida to try to address the phylogeny of freshwater sponges. Our results confirm the monophyly of Spongillina and do not support the monophyly of Haplosclerida. The results show also that Lubomirskiidae and Spongillidae are not monophyletic.

Total genomic DNA was extracted with the Genomic-tip 100 G Kit (QIAGEN). For amplification of the 676 bp of the 5 end of the COXI gene universal primers (Folmer et al. 1994) were used. For amplification of the full-length 18S rRNA gene (1800 bp), the following primers were designed based on the alignment of sponge 18S rRNA sequences from GenBank: forward primer 5-GTCAATTGTCATGGCAAATCAGGT3, and reverse primer 5- GGTTTATGGTACCGGTCAACT3. For the sequencing one internal primer was also used: 5ATCGTGCATATAGGTCCCATCGTC-3. Each 25 l PCR reaction mix contained 2.5 l of 10xPCR Buffer (Promega), 3 l of MgCl2 (25 mM), 0.5 l of each primers (10 pmol/ l), 1 l of dNTP mix (100mM each ), 1 l of DNA (20 ng), 1 Unit of Taq DNA polymerase and double distilled water to 25 l. Thermocycling parameters were: initial denaturation at 94oC for 120s, followed by 40 cycles of denaturation at 94oC for 60s, annealing at 55oC for 60s, and extension at 72oC for 60s, followed by a final extension of 8 min. at 72oC. Each PCR reaction was purified through a QIAquick Spin column (QIAGEN) and cloned into pGEM-T (Promega). At least two independent clones were sequenced on both strands from each specimen. The new sequences were aligned with sponge sequences from GenBank (Table 1) using the CLUSTALW program (Thompson et al. 1994). For maximum likelihood (ML) analysis, a general time reversible (GTR+G) model incorporating estimates of the alpha (Rodriguez et al. 1990) was chosen as the best model both for 18S rRNA and COXI genes using the MODELTEST program (Posada and Crandall 1998). Neighbor joining (NJ) analysis was performed using Kimura (1980) two-parameters distance. Maximum parsimony (MP), ML and NJ analyses were done using the programs MEGA 1.01 (Kumar et al. 2001), Treeconw (van de Peer et al. 1993) and TREEPUZLE (Schmidt et al. 2000). Bootstrap values for each tree were calculated from 500 replicates. The sequences of Aplysina fistularis (Verongida) and Aplysilla sulfurea (Dendroceratida) were used as outgroups.

Results 18S rRNA

Complete sequences of 18S rRNA gene are available now for 40 sponge species. The 18S rRNA sequences of Echinospongilla brichardi, Swartschewskia papyracea, Baikalospongia bacillifera, B. fungiformis and B. intermedia were obtained. In the dataset used for phylogenetic tree reconstruction all available sequences for clades G3 and G4 (following Borchiellini et al. 2004) from GenBank were included. The sequence of Aplysilla sulfurea (Aplysillidae, Dendroceratida, G1 clade) was used as outgroup to root the 18S rRNA tree. Sequences of Baikalospongia intermedia and Baikalospongia fungiformis were identical to that of B. bacillifera and were excluded from the analysis. Ephydatia fluviatilis and E. cooperensis also had identical 18S sequences. The final alignment is 1490 bp long and has 425 variable sites from which 257 were parsimony informative. Trees obtained by ML, NJ and MP analyses had a similar topology (Fig. 1). Results based on 18S rRNA data

Material and methods

Species of four marine families (Callyspongiidae, Petrosiidae, Chalinidae and Niphatidae) from the Mediterranean sea and the Caribbean area, one cosmopolitan freshwater sponge family (Spongillidae) from the lake Biwa (Japan), one endemic freshwater sponge family (Lubomirskiidae) from lake Baikal (Russia) and one species of Potamolepidae from lake Tanganyika (Zambia), were included in the analyses. Sponge samples were collected by SCUBA diving. All specimens were photographed alive. Samples were snapfrozen in liquid nitrogen and used later for DNA extraction. Part of each sample was fixed in 70% ethanol for taxonomic identification. Spicules and skeleton preparations were made as described previously (Efremova 2001). Sampling included species belonging to the families Chalinidae, Niphatidae, Callyspongiidae, Petrosiidae, Spongillidae, Potamolepidae, Lubomirskiidae.

385

indicate that the order Haplosclerida is not monophyletic. On the tree, marine haplosclerid species Haliclona sp., H. oculata, H. mediterranea (Chalinidae, Haplosclerina) and Xestospongia muta (Petrosiidae, Petrosina) are situated at the base of a clade including freshwater Haplosclerida, some other haplosclerid species and all the species belonging to Hadromerida, Halichondrida, Poecilosclerida and Tetractinellida (Astrophorida + Spirophorida) (Fig. 1). This clade is supported by 100% bootstrap and included species from the Chalinidae and Petrosiidae families. All Spongillina species (Echinospongilla brichardi, Swartschewskia papyracea, Baikalospongia bacillifera, B. robusta and B. intermedia), Calyx podatypa (Phloeodictyidae, Petrosina) and Vetulina stalactites (incertas sedis lithistid) constitute a second clade well supported in all analyses (92% NJ, 66% MP, 95% ML of bootstrap). Calyx podatypa and Vetulina stalactites constitutes a clade with a bootstrap value 77% NJ, 56% MP, 70% ML. Even though those bootstrap values are not very high, it is an indication on possible relationships of the incertae sedis lithistid (Vetulina stalactites) with Phloeodictyidae. The Spongillina clade has a high bootstrap support (100%), but the relationships within that clade are not well resolved. Echinospongilla brichardi (Potamolepidae) and Trochospongilla horrida (Spongillidae) are combined in one group with 61-65% bootstrap (NJ, ML) which is not supported by MP analyses. Sponges of the endemic family Lubomirskiidae were situated in a group with Spongillidae species and had unresolved relationships with the species of the genera Ephydatia, Spongilla and Eunapius.

bootstrap) or with other clades of Demospongiae which means that the COXI gene is not sufficiently informative.

Discussion Phylogeny of Haplosclerida

Our results of analyses of 18S rRNA and COXI sequences are congruent with each other and do not support the monophyly of Haplosclerida. The data confirm the previous results based on partial 28S rRNA sequences (McCormack et al. 2002) and 18S rRNA sequences (Redmond et al. 2007) that reveal polyphyletic relationships within the order, as well as in some of its families and genera. In a previous study based on COXI sequences a clade (Haplosclerina Petrosina) was supported if species from the suborder Spongillina were excluded (Nichols 2005). However, our 18S rRNA phylogenetic tree does not support the monophyly of Petrosina. Xestospongia muta (Petrosiidae) and Calyx podatypa (Phloeodictyidae) belonging to two different clades. A clade including Xestospongia muta (Petrosiidae) and three species of Haliclona (Chalinidae) belonging to Haplosclerina and Petrosina is well supported (100% bootstrap). Our analyses of 18S rRNA do not support the monophyly of Petrosina but support the monophyly of Spongillina. On the COXI tree there is no resolution at all between the different haplosclerid taxa. Clearly, the rate of nucleotide substitution on the COXI gene is much higher than for the 18S rRNA gene, which must have resulted on high levels of homoplasy in the former. From our data with the COXI gene the monophyly of Spongillina is well supported while the monophyly of marine Haplosclerida and of Haplosclerina and Petrosina is not supported.

COXI
The dataset of sponge COXI sequences in the GeneBank is growing and as of November 2006 there were 53 sequenced species. We have obtained partial sequences of the COXI gene of Callyspongia vaginalis, Xestospongia muta, Haliclona aquaeductus, Niphates digitalis, Niphates sp. and Halichondria panicea. The alignment produced was 455 bp long, of which 237 nucleotide positions were variable and 208 parsimony informative. A sequence of Aplysina fistularis (Verongida, Aplysinidae, G2 clade) was used as an outgroup. Sequences of all Baikalospongia species, Lubomirskia baikalensis and Ephydatia muelleri were identical to that of Baikalospongia bacillifera and were excluded from alignment. Since between freshwater Haplosclerida sequences there are very few aminoacid substitutions we used nucleotide sequences for tree reconstruction. Trees obtained by ML, NJ, MP analyses had similar topology. In the phylogenetic tree based on the COXI data (Fig. 2) the monophyly of the Haplosclerida is not recovered. Monophyly of the Spongillina is supported with 100-90%. Corvomeyenia sp. (family Metaniidae) forms a basal branch within the Spongillina clade on the COXI tree, whereas it has unresolved relationships on 18S tree. The internal relationships between the other species are unresolved the bootstrap value being very low. Monophyly of the endemic Lubomirskiidae is not supported either the monophyly of the Spongillidae. Most marine haplosclerid species have unresolved relationships either between them (with the exception of Callyspongia vaginalis and Xestospongia muta supported with 100%

Phylogeny and origin of Spongillina

The phylogenetic relationships of Spongillina with other Demospongiae are always an open question. On the 18S rRNA tree a clade [Calyx podatypa - Vetulina stalactites] is the sister group of a Spongillina clade (Fig. 1). This suggests the hypothesis that Spongillina could share a common ancestor with Phloeodictyidae (Calyx podatypa) taxa and that the incertae sedis lithisitd Vetulina stalactites could also have a close relationships with the Phloeodictyidae. On the COXI tree there is no resolution between haplosclerid taxa and unfortunately we could not obtain sequences for Calyx podatypa or Vetulina stalactites (Fig. 2). On both phylogenetic trees the Spongillina (four families represented) form a monophyletic, well supported clade. We confirmed the results obtained with 3 families by Addis and Peterson (2005) and Redmond et al. (2007). However, the monophyly of Spongillina has yet to be confirmed using species from the Metschnikowiidae and Malawispongiidae families. Although the markers used did not allow a good resolution of the phylogenetic relationships within Spongillina, they indicate that the Metaniidae is at the basis of the Spongillina clade. Species of this family are widely distributed and can produce gemmules. Gemmules are also produced by some species of the other families. Therefore, it is possible that the common ancestor of Spongillina may have been able to

Table 1: List of species following classification of Sytema Porifera and references to the sequences used in this work. Species References (18S, COX) AJ972398 AY561980 AY737636 AY348878 AY737635 AF246618 AY561981 AY561974 AY561977 AY561964 AY734440 AY561962 AY320032 AY561961 18S rRNA, GenBank No COXI, GenBank No

386

Order

Family

Agelasida Astrophorida

Astroscleridae Ancorinidae Corallistidae Geodiidae

Dendroceratida Darwinellidae Hadromerida Clionaidae Hemiasterellidae

Placospongiidae Spirastrellidae Suberitidae

AF100947

Tethyidae AY737637 AY348887 U43190 AY734442 AJ843894 AY737638 AY348880 AY561976 AY561972 AY737639 EF095183 AY348885 EF095182 AJ843892 EF095186 AY348879

Halichondrida

Axinellidae

Astrosclera willeyana Ecionemia sp. Corallistes sp Geodia cydonium Geodia media Geodia (Sidonops) neptuni Geodia papyracea Aplysilla sulfurea Pione velans Axos cliftoni Hemiasterella sp. Placospongia sp. Spheciospongia vesparium Prosuberites laughlini Protosuberites sp. Rhizaxinella sp. Suberites domuncula Suberites ficus Suberites sp. Tethya actinia Tethya californiana Axinella corrugata Axinella damicornis Axinella polypoides Pseudaxinella reticulata AY561960 AY561979 AY561983 AJ843891 AY561983 AY320033 AY561978 AY791693

Dictyonellidae

Halichondriidae

Haplosclerida

Callyspongiidae Chalinidae

Ptilocaulis gracilis Dictyonella incisa Scopalina ruetzleri Didiscus sp. Halichondria melanodocia Halichondria panicea Spongosorites genitrix Callyspongia vaginalis Haliclona amphioxa Haliclona aquaeductus Haliclona mediterranea

Worheide 2006 Nichols 2005 Richelle-Maurer and van de Vyver 2005, unpublished Borchiellini et al. 2004 Nichols 2005 Richelle-Maurer and van de Vyver 2005, unpublished, Lavrov et al. 2005 Nichols 2005 Borchiellini and Le Parco 2001 Nichols 2005 Nichols 2005 Nichols 2005 Nichols 2005 Richelle-Maurer and van de Vyver 2005, unpublished Nichols 2005 Nichols 2005 Nichols 2005 Schrder et al. 2003 Collins 1998, Hess et al. 2004, unpublished Nichols 2005 Lavrov et al. 2005 Nichols 2005 Richelle-Maurer and van de Vyver 2005, unpublished, Lavrov and Lang 2005 Borchiellini et al. 2004 Cavalier-Smith et al. 1996 Richelle-Maurer and van de Vyver 2005, unpublished (under the name P. lunaecharta), Hess et al. 2005, unpublished Richelle-Maurer and van de Vyver 2005, unpublished Borchiellini et al. 2004 Nichols 2005 Nichols 2005 Richelle-Maurer and van de Vyver 2005, unpublished This article Borchiellini et al. 2004 This article Hess et al. 2005, unpublished This article Borchiellini et al. 2004

Table 1 (cont.)

Lubomirskiidae

Haliclona oculata Haliclona sp. Baikalospongia bacillifera AY734450 AY734444 DQ176775, EF095191 EF095188 EF095190 EU000570 DQ176776 EF095189 DQ176774 Richelle-Maurer and van de Vyver 2005, unpublished Richelle-Maurer and van de Vyver 2005, unpublished Addis and Peterson 2005, this article, Schrder et al. 2003

Metaniidae Niphatidae AY621510 AY734447 EF095192 AF140354 AY578146 AF121110 AF121111 AF121112 AY609320 DQ087503 AJ224648 AY348882 AY348881 AY737641 L10825 AF100946

Petrosiidae Phloeodictyidae

EU000567 EU000569 EU000568 EU000571 DQ176781 EF095187 EF095184 EF095185

Potamolepidae Spongillidae

Baikalospongia fungiformis Baikalospongia intermedia Baikalospongia recta Lubomirskia baikalensis Swartschewskia papyracea Corvomeyenia sp. Niphates digitalis Niphates sp. Xestospongia muta Calyx podatypa Oceanapia sp. Echinospongilla brichardi Ephydatia cooperensis Ephydatia fluviatilis Ephydatia muelleri Eunapius fragilis Spongilla lacustris

This article This article, Schrder et al. 2003 Schrder et al. 2003 Addis and Peterson 2005, Schrder et al. 2003 This article, Schrder et al. 2003 Addis and Peterson 2005, Addis and Peterson 2005 This article This article Richelle-Maurer and van de Vyver 2005, unpublished, this article Richelle-Maurer and van de Vyver 2005, unpublished Nichols 2005 This paper, Itskovich et al. 2006 Addis and Peterson 2005, Peterson and Butterfield 2005 Richelle-Maurer et al. 2005, unpublished, Addis and Peterson 2005 Addis and Peterson 2005, Addis and Peterson 2005, Hess et al. 2005, unpublished Addis and Peterson 2005, Hess et al. 2005, unpublished Addis and Peterson 2005, Hess et al. 2005, unpublished, Itskovich et al. 2006 Richelle-Maurer and van de Vyver 2005, unpublished Peterson and Butterfield 2005

AY561967 EU000573 DQ087505 DQ176777 DQ176778, AJ843884 DQ176779, AJ843882 AJ843883, EU000572

AY737642 AY734439 D15067

AY561963 DQ087475 AJ843890 AJ843889 AY561975 AJ843893 DQ133905 AJ843895

Trochospongilla horida Trochospongilla pennsylvanica Lithistid Vetulinidae Vetulina stalactites Poecilosclerida Crellidae Crella elegans Hymedesmiidae Phorbas tenacior Iotrochotidae Iotrochota birotulata Microcionidae Microciona prolifera Mycalidae Mycale fibrexilis Phoriospongiidae Strongylacidon bermudae Podospongiidae Diacarnus spinipoculum Raspaillidae Eurypon clavatum Tedaniidae Tedania ignis Spirophorida Tetillidae Cinachyrella apion Cinachyrella sp. Tetilla japonica Verongida Aplysinidae Aplysina fistularis AY561987

McInerney et al. 1999 Borchiellini et al. 2004 Borchiellini et al. 2004 Richelle-Maurer and van de Vyver 2005, unpublished, Nichols 2005 Wainright et al. 1993, Peterson et al. 2005. Collins 1998, Hess et al. 2004, unpublished Hess et al. 2004, unpubl Nichols 2005 Hess et al. 2005, unpublished Richelle-Maurer and van de Vyver 2005, unpublished, Wulff 2006 Hess et al. 2005, unpublished Richelle-Maurer and van de Vyver 2005, unpublished Kobayashi et al. 1999, unpublished Nichols 2005

387

388

Fig. 1: Neighbour-joining phylogenetic tree based on 18S rRNA sequences. Maximum parsimony and maximum likelihood trees have similar topology. The numbers under the nodes are bootstrap values for NJ, MP, ML analyses (from top to bottom). The tree is rooted on Aplysilla sulfurea (Aplysillidae, Dendroceratida). Poec - Poecilosclerida, Hadr Hadromerida, Halic - Halichondrida, Agel - Agelasida, Spir - Spirophorida, Astr - Astrophorida, Lith - Lithistid, Dendr - Dendroceratida, Petr Petrosina (Haplosclerida), Spon - Spongillina(Haplosclerida), Hapl - Haplosclerina (Haplosclerida). Species of Haplosclerida are indicated in bold.

produce gemmules, and that trait may have been lost in some taxa. Our results do not support the polyphyly of Spongillina as hypothesized by Brien (1970) and Volkmer-Ribeiro (1979, 1990). Echinospongilla (Potamolepidae) and Corvomeyenia (Metaniidae) are clearly part of the clade Spongillina and do not seem to bear any relationships with the Hadromerida or Poecilosclerida. Baikalospongia, Lubomirskia and Swartschewskia (Lubomirskiidae) from Lake Baikal and Echinospongilla (Potamolepidae) from Lake Tanganyika constitute with the Spongillidae taxa a clade well supported in both trees. Baikalospongia, Lubomirskia and Swartschewskia (Lubomirskiidae) from Lake Baikal constitute with the

Ephydatia, Spongilla, Eunapius (Spongillidae) a clade in the 18SrDNA tree (52-65%) and in the COXI tree (60%).

Phylogeny of Spongillidae and Lubomirskiidae

The classification of Spongillina species is mainly based on shape and size of megascleres (oxeas and/or strongyles), of microscleres and especially of gemmoscleres. The presence of gemmules is often linked to ecological conditions and it has been shown that freshwater sponges living in great lakes do not have gemmules. The Spongillidae species living in Lake Baikal do not produce gemmules (Masuda 1999, Efremova 2001) and in consequence there are no gemmoscleres.

389

Fig. 2: Neighbour-joining phylogenetic tree based on partial COXI sequences sequences. Maximum parsimony and maximum likelihood trees have similar topology. The numbers under the nodes are bootstrap values for NJ, MP, ML analyses (from top to bottom). The tree is rooted on Aplysina fistularis (Verongida, Aplysinidae). Poec Poecilosclerida, Hadr - Hadromerida, Halic -Halichondrida, Spir Spirophorida, Astr - Astrophorida, Veron - Verongida, Petr - Petrosina (Haplosclerida), Spon - Spongillina (Haplosclerida), Hapl - Haplosclerina (Haplosclerida). Species of Haplosclerida are indicated in bold.

Spongillidae with 21 genera represents the largest family within Spongillina (Manconi and Pronzato 2002). Racek and Harrison (1974) on paleontological data suggested the monophyly of Spongillidae with Radiospongilla as an ancestral genus. The monophyly of Spongillidae is supported if Echinospongilla (Potamolepidae) is included within Spongillidae both in 18S and COXI trees (Fig. 1-2).

Lubomirskiidae according to the present classification has four genera: Lubomirskia, Baikalospongia, Swartschewskia and Rezinkovia (Efremova 2001, 2004). Sequences of both genes of one species of Lubomirskia and four species of Baikalospongia were identical to each other, whereas sequences of Swartschewskia papyracea seem to be more divergent from those of the other Lubomirskiidae. This species

390

is also the only species within Lubomirskiidae that can be easily distinguish from the others both by body shape and spicule morphology. Swartschewskia papyracea also differs from the other species in the family in its ecology: it always grows on overhangs or under the stones and never contains symbiotic algae. Therefore, the morphological, ecological and genetic data all agree in indicating the distinctness of S. papyracea within the Lubomirskiidae.

Biodiversity of sponges in Lake Baikal.

Several hypotheses have been proposed about the time divergence of Baikalian sponges classified in the Lubomirskiidae. Baikalian sponges could be relicts of ancient marine or freshwater sponge fauna older than Lake Baikal (Berg 1910, Vereshagin 1940). Lubomirskiidae could be representatives of mesolimnic fauna formed in the Paleogene (Martinson 1940, Starobogatov 1970). Efremova and Goureeva (1989) based on embryological data, considers that Lubomirskiidae and Spongillidae have a common ancestor and that they lost the ability to produce gemmules due to the constant ecological condition in lake. The molecular analyses show a low level of genetic divergence within the species classified in the Lubomirskiidae and also between the Lubomirskiidae and Spongillidae. This could indicate a recent burst of diversification within Baikal Lake and that the Spongillidae and Lubomirskiidae share a common ancestor. This result would confirm the hypothesis proposed by Efremova and Goureeva (1989) which would mean that the family Lubomirskiidae is not monophyletic. Consequently, Lubomirskia, Baikalispongia, Rezinkovia and Swartschewskia would have to be transferred to family Spongillidae. The study of the phylogenetic history of Mollusca and part of Oligochaeta from Lake Baikal based on analyses of COXII sequences indicate that a rapid species diversification occurred, within those groups, around 3.5 million years ago (Zubakov 1997, Kaigorodova 2000, Sherbakov 1999). The biodiversity of Baikalian sponge fauna is high. However, the sequences analyzed so far are not sufficiently informative to help to make hypothesis on their phylogenetic relationships. More sequence data are necessary on more species of Spongillidae, Lubomirskiidae and also Potamolepiidae, Metaniidae or Malawispongiidae before the internal relationships among Spongillina can be more adequately understood.

Acknowledgments
We thank Jean Vacelet for help in identification of samples and Christian Marshal for technical assistance. This work was supported by INTAS-YSF 2002-0382 from the European Commission. We also thank Antonio Sol-Cava for his help with the first version of the text.

References
Addis JS, Peterson KJ (2005) Phylogenetic relationships of freshwater sponges (Porifera, Spongillina) inferred from analyses of 18S rDNA, COI mtDNA, and ITS2 rDNA sequences. Zool Scr 34: 549-557 Berg LS (1910) Baikalian fauna and its origin. Biol J 1(1): 10-45 (in russ.)

Bergquist PR, Warne KP (1980) The marine fauna of New Zealand Porifera, Demospongiae, Part I (Haplosclerida and Nepheliospongida). NZ Oceanogr Inst Mem 87: 1-78 Brien P (1970) Les Potamolepides africaines nouvelles du Luapula et du Lac Moero. Symp Zool Soc London 25: 163-186 Borchiellini C, Chombard C, Manuel M, Alivon E, Vacelet J, Boury-Esnault N (2004) Molecular phylogeny of Demospongiae: implications for classification and scenarios of character evolution. Mol Phylogenet Evol 32: 823-837 de Weerdt WH (1985) A systematic revision of the North Eastern Atlantic shallow-water Haplosclerida (Porifera, Demospongiae), part I. Introduction, Oceanapiidae and Petrosiidae. Beaufortia 35: 61-92 de Weerdt WH, van Soest RWM (1986) Marine shallow-water Haplosclerida (Porifera) from the south-eastern part of the North Atlantic Ocean. Zool Ver 225: 3-47 de Weerdt WH (1989) Phylogeny and vicariance biogeography of North Atlantic Chalinidae (Haplosclerida, Demospongiae). Beaufortia 39: 55-88 Efremova SM (2001) Sponges (Porifera). In: Timoshkin OA (ed). Index of animal species inhabiting lake Baikal and its catchment area, vol.1, Lake Baikal, Book 1. Nauka, Novosibirsk. pp. 182192 (in russ.) Efremova SM (2004) New genus and new species of sponges from family Lubomirskiidae Rezvoj, 1936. In: Timoshkin OA (ed). Index of animal species inhabiting lake Baikal and its catchment area, vol.1, Lake Baikal, Book 2. Nauka, Novosibirsk. pp. 12611278 Efremova SM, Goureeva MA (1989) The problem of the origin and evolution of Baikalian sponges. The 1st Vereshchagin Baikal Int Conf Abstracts. Irkutsk, 1989: 21-22 Efremova SM, Itskovich VB, Parfenova V, Drucker VV, Mller WEG, Schrder HC. (2002) Lake Baikal: a unique place to study evolution of sponges and their stress response in an environment nearly unimpaired by anthropogenic perturbation. Cell Mol Biol 48(4): 359-371 Erpenbeck D, McCormack GP, Breeuwer JA, van Soest RWM (2004) Order level differences in the structure of partial LSU across demosponges (Porifera): new insights into an old taxon. Mol Phylogenet Evol 32(1): 388-95 Folmer O, Black M, Hoeh W, Lutz R, Vrijenhoek R (1994) DNA primer for amplification of mitochondrial cytochrome C Mol Mar Biotech 3: 294-299 Fromont J, Kerr S, Kerr R, Riddle M, Murphy P (1994) Chemotaxonomic relationships within, and comparisons between, the orders Haplosclerida and Petrosida (Porifera, Demospongiae) using sterol complements. Biochem Syst Ecol 22: 735-752 Hooper JNA, van Soest RWM (2002) Class Demospongiae Sollas, 1885. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges, vol. 1. Kluwer Academic/ Plenum Publishers, New York. pp. 15-18 Itskovich VB, Belikov SI, Efremova SM, Masuda Y (1999) Phylogenetic relationship between Lubomirskiidae, Spongillidae and some marine sponges according partial sequences of 18S rDNA. Memoir Queensl Mus 44: 275-280 Itskovich VB, Belikov SI, Efremova SM, Masuda Y, Krasko A, Schrder HC, Mller WEG (2006) Monophyletic origin of freshwater sponges in ancient lakes based on partial structures of COXI gene. Hydrobiologia 568(1): 155-159

391

Kaygorodova IA, Sherbakov DY, Verheyen E (2000) Comparative molecular phylogenetic analysis of two populations of Rhinchelmis alyonae from Lake Baikal. Biodivers Dynam Ecosyst North Eurasia 5(1-2): 177-179 Kimura M (1980) A simple method for estimating evolutionary rate of base substitutions through comparative studies of nucleotide sequences. J Mol Evol 16: 111-120 Kumar S, Tamura K, Jakobsen IB, Nei M (2001) MEGA2: molecular evolutionary genetics analysis software. Bioinformatics 17(12): 1244-1245 Manconi R, Pronzato R (2002) Suborder Spongillina subord. nov.: freshwater sponges. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges, vol. 1. Kluwer Academic/Plenum Publishers, New York. pp. 921-1019 Martinson GG (1940) Materials for the study of the fossil fauna of the Baikal area. Trudy Baikalskoi Limnolicheskoi Stantsii Academy of Sciences USSR 10: 425-451 (in russ) Masuda Y, Itskovich V, Veinberg EV, Efremova SM (1999) Perspective studies of freshwater sponges in Lake Baikal. Berliner geowiss Abh Berlin 30: 329-332 Mats VD (1993) The structure and development of the Baikal rift depression. Earth-Sci Rev 34: 81-118 McCormack GP, Erpenbeck D, van Soest RWM (2002) Major discrepancy between phylogenetic hypotheses based on molecular and morphological criteria within the Order Haplosclerida (Phylum Porifera: Class Demospongiae). J Zool Syst Evol Res 40(4): 237240 Nichols SA (2005) An evaluation of support for order-level monophyly and interrelationships within the class Demospongiae using partial data from the large subunit rDNA and cytochrome oxidase subunit I. Mol Phylogenet Evol 34 (1): 81-96 Posada D, Crandall KA (1998) MODELTEST: testing the model of DNA substitution. Bioinformatics 14(9): 817-818 Racek AA, Harrison FW (1974) The systematic and phylogenetic position of Palaeospongilla chubutensis (Porifera, Spongillidae. Proc Linn Soc New South Wales 99(3): 157-165 Redmond NE, van Soest RWM, Kelly M, Raleigh J, Travers SAA, McCormack GP (2007) Reassessment of the classification of the Order Haplosclerida (Class Demospongiae, Phylum Porifera) using 18S rRNA gene sequence data. Mol Phylogenet Evol 43: 344-352

Rodriguez F, Oliver JF, Marin A, Medina JR (1990) The general stochastic model of nucleotide substitution. J Theor Biol 142: 485501 Schmidt HA, Strimmer K, Vingron M, Haeseler von A (2000) TREE-PUZZLE: maximum likelihood phylogenetic analysis using quartets and parallel computing. Bioinformatics 18(3): 502-4 Schrder HC, Efremova SM, Itskovich VB, Belikov SI, Masuda Y, Krasko A, Mller IM, Mller WEG (2003) Molecular phylogeny of the freshwater sponges in Lake Baikal. J Zool Syst Evol Research 41: 80-86 Sherbakov DY (1999) Molecular phylogenetic studies on the origin of biodiversity in Lake Baikal. Trends Ecol Evol 14: 92-94 Starobogatov YI (1970) Molluscs fauna and zoogeographical demarcation of the continental waterbodies. Nauka Publ, Leningrad, 372p. (in russ) Thompson JD, Higgins DG, Gibson TJ (1994) CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, positions-specific gap penalties and weight matrix choice. Nucleic Acids Res 22: 46734680 Timoshkin OA (ed) (1995) Index of animal species inhabiting lake Baikal and its catchement area, vol.1, Lake Baikal, Book 2. Nauka, Novosibirsk. pp. 1261-1278 Vereshchagin GY (1940) The origin and the history of Baikal, its fauna and flora. Trudi Baikalskoy Limnologicheskoy Stanzcii AS USSR. 10: 73-239 (in russ.) van de Peer Y, de Wachter R (1993) TREECON: a software package for the construction and drawing of evolutionary trees. Comput Applic Biosci 9: 177-182 Volkmer-Ribeiro C, de Rosa-Barbosa R (1979) Neotropical freshwater sponges of the family Potamolepidae Brien, 1967. In: Lvi C, Boury-Esnault N (eds). Biologie des spongiaires. Colloques Internationaux du CNRS, vol. 291. ditions du CNRS, Paris. pp. 503-511 Volkmeir-Ribeiro C (1990) A new insight into the systematics, evolution and taxonomy of freshwater sponges. In: Rtzler K (ed). New perspectives in sponge biology. Smithsonian Institution Press: Washington DC. pp. 323-331 Zubakov DI, Sherbakov DY, Sitnikova TI (1997) Analysis of phylogeny of endemic mollusca of family Baicaliidae, Clessin 1878 (Gastropoda, Pectinibranchia) from Baikal lake using fragments of nucleotide sequences of the mitochondrial gene CO1. Mol Biol (Mosk) 31(6): 1092-1097

Porifera research: Biodiversity, innovation and sustainaBility - 2007

393

The lithistid Demospongiae in New Zealand waters: species composition and distribution
Michelle Kelly(1*), Michael Ellwood(2), Lincoln Tubbs(3), John Buckeridge(4)
National Centre for Aquatic Biodiversity and Biosecurity, National Institute of Water and Atmospheric Research (NIWA) Ltd, Newmarket, Auckland, New Zealand. m.kelly@niwa.co.nz (2) Department of Earth and Marine Sciences and Research School of Earth Sciences, The Australia National University, Canberra, ACT 0200, Australia. michael.ellwood@anu.edu.au (3) National Institute of Water and Atmospheric Research (NIWA) Ltd, Newmarket, Auckland, New Zealand. l.tubbs@niwa.co.nz (4) School of Civil, Environmental and Chemical Engineering, RMIT University, Melbourne, Victoria, Australia. john.buckeridge@rmit.edu.au
(1)

Abstract: The lithistid Demospongiae fauna of New Zealand is reviewed here following the extensive inventory, documentation and revision of the fauna over the period 1991 to 2006. Examination of almost 300 specimens led to the discovery of 29 species (13 of which were new to science) in 9 families. New species of the poorly known phymatellid genera Neoaulaxinia Pisera and Lvi, 2002 and Neosiphonia Sollas, 1888 were described, and a new corallistid genus, Awhiowhio Kelly, 2007, was recognised. All specimens were dredged from between 80 and 1700 m, but were most common between c. 300-1100 m. With the exception of several specimens dredged from the Challenger Plateau to the west of New Zealand, the majority of specimens were found north of the Chatham Rise to the east, and in northern waters, contrasting markedly with what we know of the distribution of these sponges in the past. The availability of hard substrate, ocean circulation patterns, silica availability, and past and present water temperature, are key factors constraining the present day distribution of lithistid sponges in New Zealand. Keywords: Porifera, lithistid Demospongiae, sponges, taxonomy, New Zealand

Introduction
Lithistid demosponges are a polyphyletic group (de Laubenfels 1936, Reid 1963, 1970, Kelly-Borges and Pomponi 1994), comprising 13 extant families; 36 genera are included in the most recent classification (Pisera and Lvi 2002) but 5 genera remain of uncertain status. They differ from other demosponges in that the dominant structural spicules (desmas) are articulated, forming in most species a solid, rigid, heavily siliceous skeleton. These desmas are highly diverse morphologically; the overall architecture of the desma, the ornamentation of the desma surface, and the pattern of articulation with adjacent spicules, is diagnostically important (Schrammen 1910; see Kelly 2000a, Pisera and Lvi 2002). An indication of the polyphyletic nature of lithistid sponges is revealed in the wide range of microscleres and ectosomal megascleres, and desma axial geometries, which include tetraxial, monaxial, polyaxial and anaxial forms. A study of lithistid demosponges is of particular relevance in New Zealand because of their abundance in seamount and deep-sea faunas, and their high level of endemism at the regional and local scale (Kelly 2007). The past and present distributions of lithistid sponges in New Zealand waters is disjunct, with lithistid species reported from what are now

southern locations, in the late Paleocene to early Oligocene (Kelly et al. 2003, Kelly and Buckeridge 2005). Two major regional faunas are known worldwide: the continental shelf and slope fauna of the tropical western Atlantic region (Schmidt 1870, 1880, van Soest and Stentoft 1988, Kelly-Borges et al. 1994, Kelly-Borges and Pomponi 1994, Lehnert and van Soest 1996, Pisera 1999, Pomponi et al. 2001), and the seamount fauna of the southwest Pacific including the seamounts of the New Caledonian Norfolk Ridge (Lvi and Lvi 1983, 1988, Lvi 1991, 1993, Schlacher-Hoenlinger et al. 2005). In both locations lithistid sponges dominate the fauna (Lvi 1991, Reed and Pomponi 1997, Richer de Forges et al. 2000), with Astrophorida, Halichondrida and Hexactinellida between 150-1800 m, but the structure and taxonomic composition of the communities differ. Lithistid sponges are also common throughout the tropics in relatively shallow waters, and a significant number of species are known from South African waters (Kirkpatrick 1902). Recently, Kelly (2007) recorded the first polar lithistid, Neoschrammeniella antarctica Kelly, 2007.

Historical records
The earliest record of lithistid sponges in New Zealand is to be found in the study of a species-rich assemblage

394

identified from siliceous microfossil spicules embedded in marine diatomaceous sediments from near Oamaru in North Otago (Hinde and Holmes 1892). This formation is from the early Runangan (late Eocene) horizon within the Oamaru Diatomite Member of the basaltic Waireka Volcanics (c. 35 million years ago) (Suggate et al. 1978, Edwards 1991). Hinde and Holmes (1892) illustrated what appeared to be lithistid desmas amongst the numerous megascleres that were clearly of astrophorid, haplosclerid, halichondrid, poecilosclerid, and hexactinellid species origin. These included illustrations of choanosomal desmas and ectosomal triaenes of what was thought to represent species of Lyidium Schmidt, 1870, Corallistes Schmidt, 1870, Discodermia du Bocage, 1869 and Theonella Gray, 1868. While Lyidium (= Pleroma Sollas, 1888) and Discodermia are known from New Zealand waters today (Kelly 2007), Corallistes and Theonella are restricted predominantly to tropical Pacific and southeast Asian waters. A fuller understanding of the present day New Zealand lithistid fauna enabled Kelly (2003), Kelly et al. (2003), and Kelly (2007) to increase the diversity of taxa illustrated as spicules in Hinde and Holmes (1892). Spicules identified as being from species of Lyidium, were compared with Pleroma turbinatum Sollas, 1888 and P. menoui Lvi and Lvi, 1983, and the illustration of a discotriaene attributed to Discodermia sinuosa Carter, 1881 from the Indian Ocean by Hinde and Holmes (1892) was considered to be more like Macandrewia spinifoliata Lvi and Lvi, 1983 present today in the Bay of Plenty. Illustrations of short-shafted dichotriaenes attributed to species of Corallistes by Hinde and Holmes (1892) were compared by Kelly et al. (2003) to an undescribed species of the phymatellid genus Neosiphonia Sollas, 1888. Kelly (2007) expanded this possibility to include closely related genus Neoaulaxinia Pisera and Lvi, 2002 and corallistid Awhiowhio Kelly, 2007, which also occur on the Chatham Rise today. Hinde and Holmes (1892) also described a new species of Vetulina Schmidt, 1879, V. oamaruensis Hinde and Holmes, 1892 based upon what they identified as sphaerocladine desmas. However, these desmas strongly resemble those found in a species of Crambe Vosmaer, 1880 from Spirits Bay, Northland; thus their conspecificity with this latter species cannot be discounted (Kelly et al. 2003). Although lithistid sponges have been of considerable interest to pharmacology over the last 50 years or so, due to the broad range of biological activities that their chemical extracts possess (see Bewley et al. 1998, Munro et al. 1999, Pomponi 2001, Mayer and Hamann 2004, Piel et al. 2004), only a single paper refers to a New Zealand species. Crist et al. (1983) reported the first example of double bioalkylation of the sterol side chain at position 26 in New Zealand endemic Aciculites pulchra Dendy, 1924. New Caledonian specimens of Pleroma menoui, Reidispongia coerulea Lvi and Lvi, 1988 and Neosiphonia superstes Sollas, 1888 from New Caledonia, have been the subject of considerable attention for their production of bromindoles (Guella et al. 1989), and antiviral (Laille et al. 1998) and antifungal (DAuria et al. 1995) cytotoxic macrolides (DAuria et al. 1996, Zampella et al. 1997, Carbonelli et al. 1999, Bassarello et al. 2000). Over 40

papers document biologically active compounds and their synthesis from southwest Pacific genera Aciculites Schmidt, 1879, Pleroma, Callipelta Sollas, 1888, Microscleroderma Kirkpatrick, 1903, Reidispongia Lvi and Lvi, 1988, Discodermia, Neosiphonia, and Scleritoderma Sollas, 1888, present in New Zealand, representing considerable potential for biotechnological discovery and subsequent development.

Records post 1990

For many years Aciculites pulchra and Lepidothenea incrustans (Dendy, 1924), were the only lithistid sponges known from New Zealand, both recorded during the nonAntarctic phase of the British Antarctic (Terra Nova) Expedition of 1910 (Dendy 1924). While Bergquist (1968) expanded locality records for A. pulchra and confirmed de Laubenfels (1936) correction of the original occupied name of Lepiospongia incrustans to Lepidothenea incrustans, no further species were added. By the late 1990s many of the New Caledonian lithistid species first discovered by Lvi and Lvi (1983, 1988) and Lvi (1993), and several new species, were progressively discovered in New Zealand waters (Kelly et al. 1999, Kelly 2000b, Kelly 2001a, 2001b, Pomponi et al. 2001, Kelly 2003, Kelly et al. 2003, 2004, Kelly and Buckeridge 2005). In 2000, Homophymia stipitata Kelly, 2000 was described from carbonate banks on the western continental shelf off Northland (Kelly 2000a). This species was the second described within the genus Homophymia Vacelet and Vasseur, 1971, known previously only from Madagascar and La Runion in the Western Indian Ocean. In 2003, a further lithistid sponge endemic to New Zealand was described from the same location; Pleroma aotea Kelly, 2003 was only the third species in this genus to be described (Kelly 2003). A comprehensive inventory, redescription, and revision of the New Zealand lithistid Demospongiae, commenced in 2004. Kelly (2007) recognised 9 families, 18 genera and 29 species (Table 1). Lithistid sponge specimens were recovered from collections resulting from several hundred New Zealand Oceanographic Institute (NZOI) and National Institute of Water and Atmospheric Research (NIWA) voyages around the New Zealand EEZ (Gordon 2000, Kelly 2007) (Fig. 1). The area covered extends from 24 to 57S and 155E to 170W, covering parts of Lord Howe Seamount Chain, Norfolk Ridge to the north, Challenger Plateau to the southwest, Kermadec and Colville Ridges to the north east, the Bay of Plenty and East Cape region to the east, and the Chatham Rise and Subantarctic slope in the southeast. The systematics scheme used in Kelly (2007) followed the recent revisions of Andrzej Pisera and Claude Lvi of the polyphyletic group Lithistid Demospongiae (Pisera and Lvi 2002, and subsequent references on this group). The reader is referred to this major publication for full family and genuslevel diagnoses and histories, as none will be given here.

The New Zealand lithistid sponge fauna

In the 1980s Professor Claude Lvi described numerous sponges from the seamounts of the Norfolk Ridge south of

395

Fig. 1: New Zealand and the southwest Pacific region. A. Black dots represent NZOI (New Zealand Oceanographic Institute) collection localities from which lithistid sponges were recorded; B. Geographic names of the major marine ridges, seamount chains, basins, and plateaus in the New Zealand region. Depth contours are 500 m, 1000 m, 2000 m, and 4000 m.

New Caledonia, hailed as refuges for invertebrates whose nearest relatives were Mesozoic fossils (c. 260-60 million years ago) (Bourseau et al. 1987, Richer de Forges et al. 2000). The discovery of almost 20 lithistid species in almost as many genera by Lvi and Lvi (1983, 1988) and Lvi (1991), coupled with the observation that over 60% of New Zealand lithistid genera are also represented by only a single species (Kelly 2007), confirm the relict nature of these southwest Pacific lithistid sponge species. The majority of non-lithistid demosponge genera are represented by numerous species (Kelly et al. 2007). Several southwest Pacific lithistid genera with species in New Caledonia (Anaderma Lvi and Lvi, 1983, Corallistes, Isabella Schlacher-Hoenlinger et al. 2005, Jereicopsis Lvi and Lvi, 1983 and Neophrissospongia Pisera and Lvi, 2002) are not found in New Zealand, but are found elsewhere in southeast Asia and the west central Atlantic. Herengeria Lvi and Lvi, 1988, Neoschrammeniella Pisera and Lvi, 2002 (Family Corallistidae), and Neoaulaxinia, Neosiphonia, and Reidispongia (Family Phymatellidae) are endemic to the southwest Pacific region, and Lepidothenea de Laubenfels, 1936 and Awhiowhio are endemic to New Zealand (Table 2). Scleritoderma flabelliformis Sollas, 1888 was found to be common to New Zealand, southeast Asia and the west central Pacific (Table 2). Despite this being the only species with such a distribution, the species and megasclere morphology (and ornamentation) were considered to be so similar that a new species could not be justified under those characters. There is some doubt, however, that these specimens comprise the same species, due to the lack of overall characters for differentiating species within Scleritoderma (Kelly 2007). Only a single specimen was found on the Three Kings Ridge

and probably represents the southernmost reaches of the species range. Neosiphona superstes and Pleroma turbinatum were first recorded from the Fiji Islands (Sollas 1888) (Table 2), but are now known to occur as far south as the south New Caledonia slope and New Zealand (Lvi and Lvi 1983, Kelly 2003, 2007, Schlacher-Hoenlinger et al. 2005). The New Zealand region shares 13 lithistid species with the New Caledonian southern slope and northern Norfolk Ridge seamounts (Table 2, 3). The majority of these were found north of East Cape, but 60% were found in the Bay of Plenty, 54% in the Three Kings region, and 40% in the Cavalli Seamount region (Fig. 1).

Endemic New Zealand lithistid species

New Zealand waters harbour 15 endemic species of lithistid sponges and 2 endemic genera (Table 2, 3). Awhiowhio is closely related to the corallistid genus Herengeria, with three new species confirming the integrity of the genus (Kelly 2007). Unfortunately A. sepulchrum Kelly, 2007 was described from an incomplete and macerated specimen, and requires further study when fresh material is obtained. The most common of the three species, A. osheai Kelly, 2007, has not been recorded outside the Bay of Plenty. Lepidothenea incrustans was only recorded once from the North Cape and has not been seen since. The description of two new species of the genus Leiodermatium Schmidt, 1870 from the southwest Pacific region in Kelly (2007) provided the first record of this genus further south than the Philippines; Leiodermatium is typically present in the west central Pacific, extending into southeast Asia, but is also present on both sides of the tropical Atlantic

396

Table 1: Classification and checklist of lithistid Demospongiae species from the New Zealand region (after Kelly 2007).* Class Demospongiae Sollas, 1885 Lithistid Demospongiae (formerly Order Lithistida Schmidt, 1870) Family Theonellidae von Lendenfeld, 1903 Discodermia proliferans Lvi and Lvi, 1983 Family Phymatellidae Schrammen, 1910 Neoaulaxinia clavata (Lvi and Lvi, 1988) Neoaulaxinia zingiberadix Kelly, 2007 Neoaulaxinia persicum Kelly, 2007 Neosiphonia superstes Sollas, 1888 Neosiphonia motukawanui Kelly, 2007 Reidispongia coerulea Lvi and Lvi, 1988 Family Corallistidae Sollas, 1888 Neoschrammeniella fulvodesmus (Lvi and Lvi, 1983) Herengeria auriculata Lvi and Lvi, 1988 Herengeria vasiformis Schlacher-Hoenlinger et al. 2005 Awhiowhio osheai Kelly, 2007 Awhiowhio unda Kelly, 2007 Awhiowhio sepulchrum Kelly, 2007 Family Neopeltidae Sollas, 1888 Homophymia stipitata Kelly, 2000 Callipelta punctata Lvi and Lvi, 1983 Neopelta pulvinus Kelly, 2007 Family Macandrewiidae Schrammen, 1924 Macandrewia spinifoliata Lvi and Lvi, 1983 Family Pleromidae Sollas, 1888 Pleroma turbinatum Sollas, 1888 Pleroma menoui Lvi and Lvi, 1983 Pleroma aotea Kelly, 2003 Family Isoraphiniidae Schrammen, 1910 Costifer wilsoni Lvi, 1993 Family Scleritodermiidae Sollas, 1888 Microscleroderma novaezelandiae Kelly, 2007 Scleritoderma flabelliformis Sollas, 1888 Aciculites pulchra Dendy, 1924 Aciculites manawatawhi Kelly, 2007 Aciculites sulcus Kelly, 2007 Family Azoriciidae Leiodermatium dampieri Kelly, 2007 Leiodermatium linea Kelly, 2007 Lithistid Demospongiae incertae sedis Lepidothenea incrustans (Dendy, 1924)
*The systematics scheme used in Kelly (2007) followed the recent revisions of Andrzej Pisera and Claude Lvi of the polyphyletic group Lithistid Demospongiae in the Systema Porifera: a guide to the classification of sponges (Pisera 2002 and subsequent references on this group). The reader is referred to this major publication for full family and genus-level diagnoses and histories, as none will be given here.

Ocean. Two additional tropical west central Pacific and southeast Asian species, L. intermedia Sollas, 1888, and L. colini Kelly, 2007, were recognised as part of the revision of this group (Kelly 2007) (Table 3). Kelly (2007) also recognised and described several new species of the poorly known phymatellid genus Neoaulaxinia and Neosiphonia (Table 1) previously only known from the type species Neoaulaxinia clavata (Lvi and Lvi, 1988) and Neosiphonia superstes, respectively. In the genus Neoaulaxinia, the species zingiberadix Kelly, 2007 was described from a single specimen dredged off the western continental slope of Northland, while N. persicum Kelly, 2007 is very common and widely spread from the Three Kings region in the north to the northern edge of the Chatham Rise in the southeast. The second known species of Neosiphonia, N. motukawanui Kelly, 2007, is only known from the West Cavalli Seamount (Table 3). Significant levels of endemism in lithistid sponges occur on the continental slope and carbonate banks west of North Cape (80%), the Bay of Plenty region (43%), the Cavalli Seamounts region (44%), West Norfolk Ridge (40%), Kermadec Ridge (40%), and the Three Kings Ridge (36%).

Geographic distribution
The majority of New Zealand lithistid demosponge species are found in two relatively distinct regions off northern and northeastern New Zealand (Table 3). The first region spans northern New Zealand from the Lord Howe Seamount Chain and the Challenger Plateau in the north and west, across the West Norfolk Ridge east to the North Cape of New Zealand (Fig. 1). This region includes the carbonate banks and seamounts to the north and west of the Three Kings Islands and contains 61% of New Zealands lithistid species. Although the area of greatest diversity in this region is the Three Kings Ridge and seamounts region, with 40% of New Zealand lithistid species, only 1 species is endemic (Lepidothenea incrustans). Three species are endemic to the West Norfolk Ridge and slope region (Homophymia stipitata, Neoaulaxinia zingiberadix, Aciculites manawatawhi Kelly, 2007) and one to the Lord Howe Seamount Chain (Leidermatium dampieri Kelly, 2007). The second region that harbours a majority of New Zealand lithistid sponges is to the northeast of New Zealand, from the Kermadec Ridge, south to the Cavalli Seamount region off the northeast coast of Northland, and east to the volcanically active seamounts of the Bay of Plenty (Table 3, Fig. 1). This region contains 64% of New Zealands lithistid species. It is noteworthy that five species are endemic to this general region and three are unique to the Bay of Plenty (Aciculites sulcus Kelly, 2007, Awhiowhio osheai, and Leidermatium linea Kelly, 2007). The southeast New Zealand region (Raukumara Plain, Hikurangi Plateau, south to the Chatham Rise) is depauperate of lithistid demosponges compared to the northern regions; only 21% of New Zealand total lithistid fauna (6 species) were recorded from this region (Table 3, Fig. 1). Lithistid sponges were not recorded south of the Graveyard Seamount complex on the northern edge of the Chatham Rise. Only two species were recorded from the Graveyard seamount

397

Table 2: Checklist and geographic distribution of southwest Pacific Lithistid Demospongiae species. Indian Ocean Scleritoderma flabelliformis Sollas, 1888 Aciculites orientalis Dendy, 1905 Microscleroderma herdmani (Dendy, 1905) Neosiphonia superstes Sollas, 1888 Pleroma turbinatum Sollas, 1888 Aciculites oxytylota Lvi and Lvi, 1983 Aciculites papillata Lvi and Lvi, 1983 Anaderma rancureli Lvi and Lvi, 1983 Corallistes australis Schlacher-Hoenlinger et al. 2005 Corallistes multituberculatus Lvi and Lvi, 1983 Corallistes undulatus Lvi and Lvi, 1983 Homophymia pollubrum Schlacher-Hoenlinger et al. 2005 Isabella mirabilis Schlacher-Hoenlinger et al. 2005 Jereicopsis graphidophora Lvi and Lvi, 1983 Microscleroderma stonae Lvi and Lvi, 1983 Neopelta plinthosellina Lvi and Lvi, 1988 Neophrissospongia microstylifer Lvi and Lvi, 1983 Neoschrammeniella castrum Schlacher-Hoenlinger et al. 2005 Neoschrammeniella moreti (Lvi and Lvi, 1988) Neoschrammeniella norfolkii Schlacher-Hoenlinger et al. 2005 Scleritoderma camusi Lvi and Lvi, 1983 Callipelta punctata Lvi and Lvi, 1983 Costifer wilsoni Lvi, 1993 Discodermia proliferans Lvi and Lvi, 1983 Herengeria auriculata Lvi and Lvi, 1988 Herengeria vasiformis Schlacher-Hoenlinger et al. 2005 Macandrewia spinifoliata Lvi and Lvi, 1983 Neoaulaxinia clavata Lvi and Lvi, 1988 Neoschrammeniella fulvodesmus (Lvi and Lvi, 1983) Neosiphonia motukawanui Kelly, 2007 Pleroma menoui Lvi and Lvi, 1983 Reidispongia coerulea Lvi and Lvi, 1988 Aciculites manawatawhi Kelly, 2007 Aciculites pulchra Dendy, 1924 Aciculites sulcus Kelly, 2007 Awhiowhio osheai Kelly, 2007 Awhiowhio sepulchrum Kelly, 2007 Awhiowhio unda Kelly, 2007 Homophymia stipitata Kelly, 2000 Leiodermatium dampieri Kelly, 2007 Leiodermatium linea Kelly, 2007 Lepidothenea incrustans (Dendy, 1924) Microscleroderma novaezealandiae Kelly, 2007 Neoaulaxinia persicum Kelly, 2007 Neoaulaxinia zingiberadix Kelly, 2007 Neopelta pulvinus Kelly, 2007 Pleroma aotea Kelly, 2003 X X X Southeast Asia X X X Fiji New New Islands Caledonia Zealand X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X

X X X X X X X X X X X X X X X X X X X X X X X X X X

complex, one of which is endemic to the locality (Awhiowhio sepulchrum), and only known from the type specimen. Interestingly, the other species, Neoaulaxinia persicum, was found in considerable numbers and many of the specimens were quite small probably representing a recent settlement from more northern sites.

Depth and temperature range

Lithistid sponges are found in most temperate and tropical oceans in the world but are generally restricted to depths greater than 150 m and are rarely collected below 1700 m (Pomponi et al. 2001). The New Zealand lithistid sponges

were collected between 80-1700 m, the water temperatures at these sites ranging from less than 4C to c. 18C (Fig. 2A). The deepest sites (>1000 m) were on seamounts in the Cavalli and Bay of Plenty regions (1040-1540 m), the North Chatham Rise, the South Norfolk Basin (1100-1680 m), Ngatoro Ridge off East Cape (1050 m) and on the Louisville Seamount Chain (1050 m). Species found at the greatest depths and thus experienced the lowest temperatures of below 6C, include Pleroma aotea in the Family Pleromidae, Neoaulaxinia persicum, N. clavata and Reidispongia coerulea in the Family Phymatellidae, and Neoschrammeniella fulvodesmus, and Herengeria vasiformis in the Family Corallistidae.

398

Table 3: Geographic distribution of endemic lithistid Demospongiae species and those common to the New Zealand and New Caledonian Norfolk Ridge region. X = New Zealand endemic, = common species. Shaded columns link closely adjacent biodiverse regions.

Lord Howe Seamount Chain

Leiodermatium dampieri Kelly, 2007 Pleroma turbinatum Sollas, 1888 Reidispongia coerulea Lvi and Lvi, 1988 Neoschrammeniella fulvodesmus (Lvi and Lvi, 1983) Homophymia stipitata Kelly, 2000 Costifer wilsoni Lvi, 1993 Pleroma aotea Kelly, 2003 Herengeria vasiformis Schlacher-Hoenlinger et al. 2005 Pleroma menoui Lvi and Lvi, 1983 Aciculites pulchra Dendy, 1924 Neoaulaxinia zingiberadix Kelly, 2007 Aciculites manawatawhi Kelly, 2007 Neoaulaxinia persicum Kelly, 2007 Discodermia proliferans Lvi and Lvi, 1983 Herengeria auriculata Lvi and Lvi, 1988 Lepidothenea incrustans (Dendy, 1924) Scleritoderma flabelliformis Sollas, 1888 Awhiowhio unda Kelly, 2007 Neopelta pulvinus Kelly, 2007 Neosiphonia superstes Sollas, 1888 Neoaulaxinia clavata Lvi and Lvi, 1988 Neosiphonia motukawanui Kelly, 2007 Aciculites sulcus Kelly, 2007 Awhiowhio osheai Kelly, 2007 Callipelta punctata Lvi and Lvi, 1983 Leiodermatium linea Kelly, 2007 Macandrewia spinifoliata Lvi and Lvi, 1983 Awhiowhio sepulchrum Kelly, 2007 Total number of species

X X X X X X X X X X X X X X X X X X X

X X X X

X X X X 1 2 5 2 5 11 2 1 5 9 14 1 1 1 2 2

Some of the shallower sites were volcanically active seamounts and knolls in the Bay of Plenty (Rungapapa and Tuatoru Knolls), and on the Cavalli Seamount Ridge (140200 m). The shallowest sites were just off Ngunguru Bay (80 m) and on Cavalli Seamount (103 m), Northland. Over 60% of species were found in waters shallower than c. 300 m (experiencing water around c. 14-18C). While the majority of specimens were collected between 200 and 1100 m (Fig. 2A), many of these species had

considerable depth and thus temperature ranges. Aciculites pulchra was found between 80 and 1100 m (5-17C), and Neoschrammeniella fulvodesmus (Lvi and Lvi, 1983) had the greatest depth range of 100-1680 m (5-15C) (Fig. 2). Species displaying very narrow depth and temperature ranges, such as Neoaulaxinia zingiberadix, A. sulcus, Leiodermatium dampieri, L. linea, and Lepidothenea incrustans, were only recorded from a single location.

Total number of species 1 3 2 5 2 3 4 5 8 8 1 1 4 2 1 1 1 1 1 3 1 1 1 1 1 1 1 1

Louisville Seamount Chain

Western Continental Slope

Cavalli Seamount region

South Norfolk Basin

West Norfolk Ridge

Three Kings Ridge

Challenger Plateau

Hikurangi Plateau

Raukumara Plain

Kermadec Ridge

South Fiji Basin

Colville Ridge

Chatham Rise

Bay of Plenty

East Cape

399

Discussion
Lithistid demosponges are absent south of the Subtropical Convergence just below the Chatham Rise (Shackelton and Kennett 1975, Nelson and Cooke 2001), and restricted to warmer waters in northern New Zealand, extending from the Lord Howe Seamount Chain in the west through the West Norfolk Ridge, western continental margin off North Cape, Three Kings Ridge, east to the Cavalli Seamounts, Kermadec volcanic arc, and Bay of Plenty. Several sites within the latter three regions have relatively high diversity and abundance of lithistid sponges; these sites are relatively shallow (100-200 m) compared to the majority of sites, where lithistid sponges typically occur much deeper. What are some of the factors that effect the distribution of lithistid sponges today? Availability of hard substrate Lithistid sponges require hard rocky substrate for settlement (Pomponi et al. 2001). With the exceptions of free-living Herengeria auriculata Lvi and Lvi, 1988, and Discodermia proliferans Lvi and Lvi, 1983, the vast majority of lithistids are found attached to rock, e.g., carbonate reef rubble, mudstone, or basalt, in temperate and tropical regions. As for most sponges and many other invertebrates, larval settlement is facilitated by hard surfaces devoid of macroinvertebrate competitors and heavy sediment loads. Lithistid sponges are thus restricted to deep and shallow sites that have an abundance of hard sediment-free substrate such as on steep-sided seamounts, and ledges on continental margins. This is clearly illustrated in the restriction of lithistid sponges to the Graveyard Seamount complex on the North Chatham Rise; no lithistids have been recorded at other sites due, we believe, to their sediment loading. The sediments on the Chatham Rise are a combination of relic authigenic sediment, whereas towards the south the seafloor is draped with a fine-grained combination of pelagic, hemipelagic sediment (Carter et al. 2000). Carbonate banks such as Pandora and Wanganella to the west of North Cape also provide an abundance of clean carbonate rubble for settlement of lithistid sponges and numerous other invertebrates (Kelly 2000a). The extent of settlement success of lithistid sponges in some locations where hard substrate is available was illustrated by Kelly et al. (2003, Fig. 10), in which the basalt substrate, viewed using an epibenthic sled, was closely covered with immature sponges. The site was relatively shallow at 160 m and was situated on Rungapapa Knoll in the Bay of Plenty. This settlement success also appears in at least one population of lithistid sponges living on a shallow (50-150 m) volcanic tuff cone (Lee et al. 1997) off the South Island in the late Eocene (35-33 million years ago) (Kelly et al. 2003). The spongebearing horizon contained extremely numerous sponge body fossils with a normal bell-shaped size frequency distribution population curve (Kelly et al. 2003: Fig. 8). The stratigraphy and palaeoecology of the horizon indicated that the sponges were attached to volcanically derived Mineral Breccia during their life, and the uniform distribution of the fossils throughout the sponge-bearing horizon support the hypothesis that the live sponges were simultaneously detached by a submarine flow down the flanks of the tuff cone, transported down slope and rapidly buried.

Ocean circulation While the distribution of lithistid sponges coincides with the distribution of major hard grounds such as seamount chains and banks across northern New Zealand, they are absent from the hard grounds on seamounts and continental margins south of the Subtropical Convergence below the Chatham Rise. What factors contribute to this disjunct distribution? The major current systems around New Zealand (Fig. 3) coincide with areas of high diversity of lithistid sponges, and other sponges and invertebrates; the East Auckland Current, fed by the Tasman Front (Stanton 1969, Carter et al. 1998), is reknowned for its contribution of subtropical taxa to eastern New Zealand locations (Squires 1964, Yaldwyn 1968, Russell and Ayling 1976, Powell 1976, Ayling 1982, see also Kelly 1983) and may explain the rather high incidence of New Caledonian lithistid species on the West Norfolk Ridge, Three Kings Ridge and Kermadec Ridge. Major eddies and gyres from the East Auckland current dominate the Kermadec and Colville Ridges, sustaining the high population levels in these regions. The major current systems on the west coast of New Zealand are the West Auckland Current in the north and the Westland Current to the south. The West Auckland Current, which diverges from the East Auckland Current to move south along the west coast of the North Island, may however explain the occurrence of a few rare specimens on the northern edge of the Challenger Plateau (Fig. 1). The East Cape Current and Wairarapa Eddy, which touch the northern boundary of Chatham Rise, may be a source of larvae of the single dominant lithistid species, Neoaulaxinia persicum, on the isolated Graveyard Seamount complex situated on the northern edge of the Chatham Rise, In this location specimens are as numerous as in the Bay of Plenty and Cavalli Seamount region further north and east. The second species known from the Graveyard Seamounts is only known from a single specimen, which was dead on collection. Finally, the Chatham Rise constrains the Subtropical Convergence thereby forming a strong barrier to the transport of larvae southwards over the Chatham Rise, perhaps contributing to the absence of sponges below the Subtropical Convergence. Historical patterns of distribution As noted previously, microfossil spicules very similar to those of living lithistid sponges were recorded further south than the present day Subtropical Convergence. Hinde and Holmes (1892) recorded diverse spicules in the Oamaru Diatomite (c. 35 million years ago) and Kelly and Buckeridge (2005) recorded microfossil lithistid desmas in the Tutuiri Greensand of Chatham Island (c. 56-53 million years ago). Sponge body fossils attributed to the extant lithistid Pleroma aotea were recorded from the Ototara Limestone of Kakanui (c. 33.7 million years ago). Campbell et al. (1993) contend that water temperatures in the southern parts of the New Zealand region were sufficiently warm during periods of the Cainozoic to permit the establishment of sponges, cirripedes (Buckeridge 1993, 1996, 1999), deep-sea bryozoans (Gordon 1984), brachiopods (Lee et al. 1997), crinoids (Lee 1987), and molluscs (Beu and Maxwell 1990), that today are generally found in waters north of the Chatham Rise.

400

Fig. 2: Box and Whisker plots of depth distribution and bottom-water temperatures for New Zealand lithistid species. A. Depth range for combined specimens collected within each lithistid species from stations listed by Kelly et al. (2007); B. Estimated bottom-water temperature at sponge collection station. Bottom water temperatures were estimated from nearby WOCE stations where standard hydrographic data have been collected (www.clivar.org). Upper and lower bound of box represents 75% and 25% of range, mid line in box is the median, upper and lower bounds of whiskers are maxima and minima.

It is thought that around the end of the Eocene major changes in Southern Ocean circulation occurred as the Drake Passage between South America and Antarctica opened (Flower et al. 1997). The formation of the Circum Antarctic Current lead to cooling of the Southern Ocean and the

development of cold bottom waters (Shackleton and Kennett 1975). The amount of biologically-available silica in shallow environments also decreased markedly after the Eocene, because of the increase in diatom diversity and abundance (Maliva et al. 1989); only the deepest waters around New

401

Fig. 3: Modern-day circulation patterns within the New Zealand region. Abbreviations for fronts features: TF = Tasman, STF = Subtropical, SAF = Subantarctic; Abbreviations for surface currents: WAUC = West Auckland, EAUC = East Auckland, ECC = East Cape, DC = DUrville, WC = Westland, SC = Southland, ACC = Antarctic Circumpolar Current. This figure taken from Carter (2001), but based on Carter et al. (1998).

Zealand are silica-rich today (Vincent et al. 1991, Frew and Hunter 1995). These events contribute to our understanding of the extinction of lithistid sponges and some other invertebrate species from southern New Zealand waters after the Eocene, as evidenced by Gordon (1984), Hinde and Holmes (1892), Lee et al. (1997), Buckeridge (1996), Kelly et al. (2003), and Kelly and Buckeridge (2005), and the present day restriction of lithistids to northern warmer, deeper waters. Silica availability Several sites within the Cavalli Seamounts region (NZOI I014: 103 m), the Bay of Plenty (Rungapapa and Tuatoru Knolls: 136+ m), on Kermadec Ridge Volcano L: 154+ m), and on carbonate banks off North Cape (192+ m) have relatively high diversity and abundance of lithistid sponges; these sites are relatively shallow compared to the majority of sites where lithistid sponges typically occur much

deeper (Kelly 2007) (Fig. 1). The broad Bay of Plenty region covers less than 15% of the marine region studied yet it harbours over half of New Zealand lithistid species recorded and has several endemic species. What factors facilitate higher settlement rates and greater diversity in these relatively shallow sites? The Bay of Plenty and southern Kermadec Volcanic Arc regions are hydrothermally active with a number of submarine hydrothermal plumes releasing hydrothermal fluids into the water column. Basic chemistry measurements of these fluids indicate that they are rich in silica and heavy metals (Pantin and Wright 1994, Tarasov 2006). Primary phytoplankton production in the waters near fluid discharges was found to be significantly greater than in surrounding oceanic waters, thereby indicating release of increased nutrients to fuel

402

primary and secondary production (Sorokin et al. 1998). Lithistid sponges living in the waters near these vents would benefit greatly from the increased silicon concentration within the water column and the increases in food supply linked to increased primary production. A recent geochemical study has shown considerable hydrothermal fluid release from a number of submarine volcanoes along the southern Kermadec Ridge system (de Ronde et al. 2001). Such fluid releases are rich in heavy metals and are likely, although no silicon measurements were made, to enrich in silica which would explain shallow depth distribution of lithistid sponges along the Kermadec volcano system. The carbonate banks west of North Cape are an interesting exception as they are not presently hydrothermally active. Wanganella Bank may have had localised phases of volcanism, however, with hydrothermal activity associated with older plate boundary positions to the north of New Zealand (Herzer and Mascle 1996, Balance 1999). However, these banks are in regions of strong upwelling (Stanton 1976) and primary and secondary productivity are elevated as a result (Bradford and Roberts 1978). We conclude that the disjunct distribution of lithistid sponges, both geographically and in depth, is due to multiple factors. The most important of these factors include major changes in Southern Ocean circulation patterns coupled with global reduction in bioavailable silica in shallow water since the Palaeogene, overlayed by changing patterns of active submarine volcanism, and maintained in the present day by ocean currents, high primary productivity in areas of upwelling and hydrothermal activity, and determined overall by the availability of hard substrate.

Acknowledgements
Research support and fieldwork for aspects of the research leading to this review were funded by Biological and Biotechnological Research Council (BBSRC), UK, and British Airways Conservation Travel, while the senior author was employed at the Natural History Museum, London (1992-1996). Support in New Zealand was given by a New Zealand Lotteries Grants Board Repatriation Fellowship to MK (1997-1999), a Royal Society of New Zealand Marsden Fund Grant to MK (2000-2002), and a French Foreign Affair Ministry travel grant to work with IRD (Institut de Recherche pour le Developement) in New Caledonia (1999). In later years this work was supported by the New Zealand Foundation for Research Science and Technology (Contracts C01X0028 (Seamounts) and C01X0219 (Biodiversity)) to NIWA.

References
Ayling AM (1982) Collins guide to the sea fishes of New Zealand. Collins, Auckland Balance PF (1999) Simplification of the southwest Pacific neogene arcs: inherited complexity and control by a retreating pole of rotation. Spec Publ Geol Soc Lond 164: 7-19 Bassarello C, Napolitano A, Bifulco G, Bruno I, Casapullo A, Randazzo A, Zampella A, Riccio R (2000) Bioactive macrolides and cyclopeptides from the South Pacific sponges. Rev Latinuam Quim 28(Supl.): 59-62

Bergquist PR (1968) The marine fauna of New Zealand: Porifera, Demospongiae, Part 1 (Tetractinomorpha and Lithistida). Mem New Zeal Oceanogr Inst 37: 1-106 Beu A, Maxwell PA (1990) Cenozoic Mollusca of New Zealand. New Zeal Geol Surv Paleontol Bull 58: 1-518 Bewley CA, Faulkner D, Angew J (1998) Lithistid sponges - star performers or hosts to the stars [review]. Angewanta Chemi International Edition English 37: 2163-2178 Bourseau JP, Ameziane-Cominardi N, Roux M (1987) Un Crinode pdoncul nouveau (Echinodermes) reprsentant actuel de la famille jurassique de Hemicrinidae: Gymnocrinus richeri nov. sp. des fonds bathyaux de Nouvelle-Caldonie (SW Pacific). C R Hebd Acad Sci Paris 305(3): 595-599 Bradford JM, Roberts P (1978) Distribution of reactive phosphorus and plankton in relation to upwelling and surface circulation around New Zealand. NZ J Mar Fr Res 12(1): 1-15 Buckeridge JS (1983) The fossil barnacles (Cirripedia: Thoracica) of New Zealand and Australia. New Zeal Geol Surv Paleontol Bull 50: 1-151 Buckeridge JS (1996) Phylogeny and biogeography of the primitive Sessilia and a consideration of a Tethyan origin for the group. Crustacean Issues 10: 225-267 Buckeridge JS (1999) Post Cretaceous biotic recovery: a case study on (Crustacea: Cirripedia) from the Chatham Islands, New Zealand. Rec Cant Mus 13: 43-51 Campbell HJ, Andrews PB, Beu AG, Maxwell PA, Edwards AR, Laird MG, Hornibrook NdeB, Mildenhall DC, Watters WA, Buckeridge JS, Lee DE, Strong CP, Wilson GJ, Hayward BW (1993) Cretaceous-Cainozoic geology and biostratigraphy of the Chatham Islands. New Zeal Inst Geol Nucl Sci Monogr 2: 1-240 Carbonelli S, Zampella A, Randazzo A, Debitus C, Gomez-Paloma L (1999) Sphinxolides E-G and reidispongiolide C: four new cytotoxic macrolides from the New Caledonian lithistida sponges N. superstes and R. coerulea. Tetrahedron 55: 14665-14674 Carter L (2001) Currents of change: the ocean flow in a changing world. Water and Atmosphere 9: 15-17 Carter L, Garlick RG, Sutton P, Chiswell SM, Oien NA, Stanton BR (1998) Ocean circulation New Zealand. NIWA Chart. Miscellaneous Series 76 Carter L, Neil HL, Mccave IN (2000) Glacial to interglacial changes in non-carbonate and carbonate accumulation in the SW Pacific Ocean, New Zealand. Palaeogeography Palaeoclimatology Palaeoecology 162: 333-356 Crist BV, Li X, Bergquist PR, Djerassi C (1983) Sterols of marine invertebrates. 44. Isolation, structure elucidation, partial synthesis, and determination of absolute configuration of pulchrasterol. The first example of double bioalkylation of the sterol side chain at position 26. J Org Chem 48: 4472-4479 Dauria MV, Gomez-Paloma L, Minale L, Zampella A, Debitus C, Perez J (1995) Neosiphoniamolide A, a novel cyclodepsipeptide, with antifungal activity from the marine sponge Neosiphonia superstes. J Nat Prod 58: 121-123 Dauria MV, Zampella A, Gomez-Paloma L, Minale L, Debitus C, Roussakis C, Le Bert V (1996) Callipeltins B and C; bioactive peptides from a marine Lithistida sponge Callipelta sp. Tetrahedron 52: 9589-9596 de Laubenfels MW (1936) A discussion of the sponge fauna of the Dry Tortugas in particular and the West Indies in general, with material for the revision of the families and orders of the Porifera.

403

Pub Carnegie Inst 467, Papers from the Tortugas Laboratory 30: 1-224 Dendy AO (1924) Porifera. Part I. Non-Antarctic Sponges. Natural History Report. British Antarctic (Terra Nova) Expedition, 1910, Zoology 6(3): 269-392 Edwards AR (1991) The Oamaru diatomite. New Zeal Geol Surv Paleontol Bull 64: 1-260 Flower B, Zachos J, Martin E (1997) Latest Oligocene through early Miocene isotopic stratigraphy and deep-water paleoceanography of the Western Equatorial Atlantic: Sites 926 and 929. In: Shackleton N. Curry W, Richter C, Bralower T (eds). Proceedings of the Ocean Drilling Program, Scientific Results 154(30): 451-46 Frew RD, Hunter KA (1995) Cadmium-phosphorus cycling at the subtropical convergence south of New Zealand. Mar Chem 51: 223-237 Gordon DP (1984) The marine fauna of New Zealand: Bryozoa: Gymnolaemata from the Kermadec Ridge. New Zeal Oceanogr Inst Mem 91: 1-198 Gordon DP (2000) The Pacific Ocean and global OBIS: a New Zealand perspective. Oceanography 13(3): 41-47 Guella G, Mancini I, Duhet D, Richer De Forges B, Pietra F (1989) Ethyl 6 bromo-3-indolcarboxylate and 3-hydroxyacetal6-bromoindole, novel bromoindoles from the sponge Pleroma menoui of the Coral Sea. Z Naturforsch C Biosci 44: 914-916 Herzer RH, Mascle J (1996) Anatomy of a continent backarc transform the Vening Meinsez Fracture Zone, northwest of New Zealand. Mar Geophys Res 18: 401-427 Hinde GJ, Holmes WM (1892) On the sponge remains in the Lower Tertiary strata near Oamaru, Otago, New Zealand. J Linn Soc Lond Zool 24: 177-262 Kelly M (1983) A bibliography and literature review for the Poor Knights Islands Marine Reserve. Unpublished report prepared for the Poor Knights Islands Marine Reserve Management committee in the Zoology Department of the University of Auckland, and the Administration division of the Ministry of Agriculture and Fisheries, Auckland, 126 pp Kelly M (2000a) Description of a new lithistid sponge from northeastern New Zealand, and consideration of the phylogenetic affinities of families Corallistidae and Neopeltidae. Zoosystema 22(2): 265-283 Kelly M (2000b) Seamount refuges: sponges show the way. Water and Atmosphere 8(3): 7 Kelly M (2001a) Rock sponges living fossils. Marsden Fund Update 16: 5 Kelly M (2001b) Rock sponges on seamounts: relicts from the past. Biodiversity Update 4: 3 Kelly M (2003) Revision of the sponge genus Pleroma Sollas (Lithistida: Megamorina: Pleromidae) from New Zealand and New Caledonia, and description of a new species. New Zeal J Mar Freshwat Res 37:113-127 Kelly M (2004) News: The oceans own tuatara discovered. Water and Atmosphere 11(4): 5 Kelly M (2007) The marine fauna of New Zealand: Porifera: lithistid Demospongiae (Rock sponges). NIWA Biodiversity Memoir 121: 1-100 Kelly M, Bell LJ, Buckeridge JS (1999) Treasures from the deep: Living fossil sponge finds in northern New Zealand are a key to the evolution of the demosponges. Water and Atmosphere 7: 11-12

Kelly M, Buckeridge JS (2005) An early Paleogene sponge fauna, Chatham Island, New Zealand. New Zeal J Mar Freshwat Res 39(4): 899-914 Kelly M, Edwards AR, Wilkinson MR, Alvarez B, Cook SdeC, Bergquist PR, Buckeridge JS, Campbell H, Reiswig HM, Valentine C (in press) Phylum Porifera Sponges. In: Gordon DP (ed). The New Zealand inventory of biodiversity, Volume 1. Kingdom Animalia: Radiata, Lophotrochozoa, and Deuterostomia. Canterbury University Press, Christchurch Kelly M, Lee D, Kelly S, Buckeridge JS (2003) A Recent sponge, Pleroma aotea Kelly (Order Lithistida: Family Pleromidae), in the Late Eocene Ototara Limestone of Otago, New Zealand. New Zeal J Mar Freshwat Res 37: 129-148 Kelly-Borges M, Pomponi SA (1994) Phylogeny and classification of lithistid sponges (Porifera: Demospongiae): a preliminary assessment using ribosomal DNA sequence comparisons. Mol Mar Biol Biotech 3: 87-103 Kelly-Borges M, Robinson EV, Gunasekera S, Gunasekera M, Gulavita N, Pomponi SA (1994) Species differentiation in the marine sponge genus Discodermia (Demospongiae, Lithistida): The utility of secondary metabolites as species-specific markers. Biochem Systemat Ecol 22(4): 353-365 Kirkpatrick R (1902) Descriptions of South African sponges. Part II. Marine investigations in South Africa 2(14): 171-180 Laille M, Gerald F, Debitus C (1998) In vitro antiviral activity on dengue virus of marine natural products. Cell Mol Life Sci 1998(54): 167-170 Lee DE (1987) Cenozoic and recent Inarticulate brachiopods of New Zealand: Discinisca, Pelagodiscus and Neocrania. J Roy Soc New Zeal 17(1): 49-72 Lee DE, Scholz J, Gordon DP (1997) Paleoecology of a late Eocene mobile rockground biota from north Otago. Palaios 5(12): 568581 Lehnert H, van Soest RWM (1996) North Jamaican deep fore-reef sponges. Beaufortia 46: 53-81 Lvi C (1991) Lithistid sponges from the Norfolk Rise: recent and Mesozoic genera. In: Reitner J, Keupp H (eds). Fossil and recent sponges. Springer-Verlag Publishers, Berlin Lvi C (1993) Porifera Demospongiae: Spongiaires bathyaux de Nouvelle-Caldonie, rcolts par le Jean Charcot. Campagne BIOCAL, 1985, pp 9-87 In: Crosnier A (ed). Rsultats des campagnes Musorstom, Vol 11. Mem Mus Natl Hist Nat (A, Zoologie) 158: 1-426 Lvi C, Lvi P (1983) ponges Tetractinellides et Lithistides bathyales de Nouvelle Caldonie. Bull Mus natl Hist nat Paris (4)5 (A,1): 101-168 Lvi C, Lvi P (1988) Nouveaux spongiaires ponges Lithistides bathyaux affinitis crtaces de la Nouvelle Caldonie. Bull Mus natl Hist nat Paris (4)10 (A,2): 241-263 Maliva RG, Knoll AH, Siever R (1989) Secular changes in chert distribution: a reflection of evolving biological participation in the silica cycle. Palaios 4: 519-532 Mayer AMS, Hamann MT (2004) Marine Pharmacology in 2000: Marine compounds with antibacterial, anticoagulant, antifungal, anti-inflammatory, antimalarial, antiplatelet, antituberculosis, and antiviral activities; affecting the cardiovascular, immune and nervous systems and other miscellaneous mechanisms of action. Mar Biotec 6: 37-52 Munro MHG, Blunt JW, Dumdei EJ, Hickford SJH, Lill RE, Li S, Battershill CN, Duckworth AR (1999) The discovery and

404

development of marine compounds with pharmaceutical potential. J Biotechnol 70: 15-25 Nelson CS, Cooke PJ (2001) History of oceanic front development in the New Zealand sector of the Southern Ocean during the Cenozoic a synthesis. New Zeal J Geol Geophys 44 (4): 535-553 Pantin HM, Wright IC (1994) Submarine hydrothermal activity within the offshore Taupo Volcanic Zone, Bay of Plenty continentalshelf, New Zealand. Continent Shelf Res 14(13-14): 1411-1438 Piel J, Hui D, Wen G, Butzke D, Platzer M, Fusetani N, Matsunaga S (2004) Antitumor polyketide biosynthesis by an uncultivated bacterial symbiont of the marine sponge Theonella swinhoei. Proc Natl Acad Sci USA 101(46): 16222-16227 Pisera A (1999) Lithistid sponge Setidium obtectum Schmidt, 1879, rediscovered. Memoir Queensl Mus 44: 473-477 Pisera A, Lvi C (2002) Lithistid Demospongiae. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges. Kluwer Academic/Plenum Publishers, New York. pp. 299-301 Pomponi SA (2001) The oceans and human health: the discovery and development of marine-derived drugs. Oceanography 14(1): 78-87 Pomponi SA, Kelly M, Reed J, Wright A (2001) Diversity and bathymetric distribution of lithistid sponges in the tropical western Atlantic region. Bull Biol Soc Wash 10: 344-353 Powell AWB (1976) On the considerable influx of warm-water molluscs that have invaded northern New Zealand waters within recent years. Records of the Auckland Institute and Museum 13: 141-166 Reed JK, Pomponi SA (1997) Biodiversity and distribution of deep and shallow water sponges in the Bahamas. Proc 8th Int Coral Reef Symp, Balboa 2: 1387-1392 Reid REH (1963) A classification of the Demospongia. Neues Jahrbuch Geol Palaontol Monatsh 4: 196-207 Reid REH (1970) Tetraxons and demosponge phylogeny. In: Fry WG (ed). The biology of the Porifera. Symposia of the Zoological Society of London 25: 63-89 Richer De Forges B, Koslow J-A, Poore GCB (2000) Diversity and endemism of the benthic seamount macrofauna in the southwest Pacific. Nature 405: 944-947 Ronde CEJ de, Baker ET, Massoth GJ, Lupton JE, Wright IC, Feely RA, Greene RA (2001) Intra-oceanic subduction-related hydrothermal venting, Kermadec volcanic arc, New Zealand. Earth Planet Sci Lett 193: 359-369 Russell BC, Ayling AM (1976) New records of fishes from northeastern New Zealand. J Roy Soc New Zeal 6(3): 277-286 Schlacher-Hoenlinger MA, Pisera A, Hooper JNA (2005) Deep-sea lithistid assemblages from the Norfolk Ridge (New Caledonia),

with description of seven new species and a new genus (Porifera, Demospongiae). Zoosystema 27(4): 649-698 Schmidt O (1870) Gundzge einer Spongien Fauna des atlantischen Gebietes. Wilhelm Engelman, Leipzig Schmidt O (1880) Die Spongien des Meerbusens von Mexico und des Caraibischen Meeres. Gustav Fischer, Jena Schrammen A (1910) Die Kieselspongien der oberen Kreide von Nordwestdeutschland. Palaeontographica Suppl. 5. I, Tetraxonia, Monaxonia und Silicea. Grundzge einer Spongien fauna des Atlantischen Gebietes. Wilhemlm Engelmann, Leipzig Shackleton N, Kennett J (1975) Paleotemperature history of the Cenozoic and the initiation of Antarctic glaciation: oxygen and carbon isotope analyses in DSDP Sites 277, 279 and 281. In: Kennett, J, Houtz, R (eds). Initial Reports of the Deep Sea Drilling Program, Washington 29: 743-755 Sorokin YI, Sorokin PY, Zakuskina OY (1998) Microplankton and its functional activity in zones of shallow hydrotherms in the Western Pacific. J Plankton Res 20(6): 1015-1031 Stanton BR (1969) Hydrological observations across the tropical convergence north of New Zealand. New Zeal J Mar Freshwat Res 3: 124-146 Stanton BR (1976) An oceanic front near the Norfolk Ridge north west of New Zealand. Deep Sea Res 23: 821-824 Sollas WJ (1888) Report on the Tetractinellida collected by H.M.S. Challenger, during the years 1873-1876. Rep Sci Res Voy H.M.S. Challenger, Zool 25(63): 1-458 Squires DF (1964) New stony corals (Scleractinia) from northeastern New Zealand. Records of the Auckland Institute and Museum 6(1): 1-10 Suggate RP, Stevens GR Te Punga MT (eds) (1978) The geology of New Zealand. Government Printer, Wellington Tarasov VG (2006) Effects of shallow-water hydrothermal venting on biological communities of coastal marine ecosystems of the western Pacific. Adv Mar Biol 50: 267-421 van Soest RWM, Stentoft N (1988) Barbados deep-water sponges. Stud Fauna Curacao Caribb Isl 70: 1-175 Vincent WF, Howard-Williams C, Tildesley PC, Butler ECV (1991) Distribution and biological properties of ocean water masses around the South Island, New Zealand. New Zeal J Mar Freshwat Res 25(1): 21-42 Yaldwyn JC (1968) Records and observations on the coral shrimp Stenopus in Australia, New Zealand and the southwest Pacific. Aust Zool 14(3): 277-289 Zampella A, Dauria MV, Minale L, Debitus C (1997) Callipeltosides B and C, two novel cytotoxic glycoside macrolides from a marine lithistida sponge Callipelta sp. Tetrahedron 53: 3243-3248

Porifera research: Biodiversity, innovation and sustainaBility - 2007

405

Construction and characterization of a cDNA library from the marine sponge Chondrosia reniformis
Anne Kuusksalu(1), Madis Metsis(2), Tnu Reintamm(1), Merike Kelve(1, 2*)
National Institute of Chemical Physics and Biophysics, Akadeemia tee 23, 12618 Tallinn, Estonia. kuusksal@kbfi.ee, tonu@kbfi.ee (2) Department of Gene Technology, Tallinn University of Technology, Akadeemia tee 15, 12618 Tallinn, Estonia. madis.metsis@ttu.ee, merike.kelve@ttu.ee
(1)

Abstract: A cDNA library from a marine sponge Chondrosia reniformis was constructed with an initial aim to isolate cDNA for 2,5-oligoadenylate (2-5A) synthetase. The library consisted of about 100 000 clones with an insert size ranging from 0.7 to 3.5 kb. 96 cDNAs were end-sequenced which resulted in a total of 83 000 bp DNA sequences. The analysis revealed a high similarity of cDNA sequences from C. reniformis to those from evolutionarily higher organisms, mostly vertebrates, including primates and human. No cDNA for 2-5A synthetase was found probably due to its low homology with a specific probe used for screening the 2-5A synthetase cDNA from another marine sponge, Geodia cydonium. Statistical analysis of dinucleotide frequencies in the sequenced cDNA fragments from C. reniformis showed some deviations from their common distribution in other sponges. Keywords: cDNA library, Chondrosia reniformis, marine sponge

Introduction
Sponges are evolutionarily the most ancient and simpliest multicellular animals. Though molecular studies in recent years have demonstrated a surprising similarity between sponges and higher organisms, especially mammals (Gamulin et al. 2000, Cetkovic et al. 2007), the knowledge about the sponge genome is still limited. We have previously demonstrated the presence of the 2-5A synthetase enzymatic activity, characteristic for mammals, in a variety of marine sponges, including C. reniformis (Kuusksalu et al. 1995, Kelve et al. 2003, Reintamm et al. 2003). In higher vertebrates this enzyme is part of the innate immunity and controls a regulated RNA decay pathway involved in the antiviral and growth inhibitory effects of the interferons (rev. by Sarkar et al. 2004). 2-5A synthetases and their genomic structures have been well characterized in mammals (rev. by Justesen et al. 2000) and birds (Tatsumi et al. 2000). Contrary to the wide occurrence of the 2-5A synthetases in higher vertebrates, this enzyme has been found neither in fish nor in lower organisms (Grebenjuk et al. 2002, Mashimo et al. 2003). 2-5A synthetase cDNAs which have been cloned from two marine sponges, Geodia cydonium Fleming, 1828 and Suberites domuncula Olivi, 1792 (Wiens et al. 1999, Grebenjuk et al. 2002), have a limited homology to each other and particularly to those from vertebrates. In addition to 2-5A synthetases, many mammalian genes, present in sponges and corals, are known to be missing or highly diverged in nematode and insect genomes (Gamulin et al. 2000, Kortschak et al. 2003).

The marine sponge Chondrosia reniformis Nardo, 1847 is one of the most widespread sponge species in the Mediterranean Sea (Hooper and van Soest 2002) known for its high collagen content (Imhoff and Garrone 1983). Several studies have pointed to the relative ease of cultivating C. reniformis and to the perspective of using it as a model organism to develop feeding strategies and to evaluate the biotechnological potential of sponge cultivation (Nickel et al. 2003, Sipkema et al. 2006). The current study focused on the construction and characterization of a cDNA expression library from the marine sponge C. reniformis and screening it with a 2-5A synthetase cDNA probe from G. cydonium. This is the first molecular characterization of the genome of C. reniformis at the EST level.

Materials and methods Sponge material

The specimen of Chondrosia reniformis Nardo, 1847 (Porifera, Demospongiae, Tetractinomorpha, Hadromerida, Chondrillidae) was collected in the Aegean Sea near Kalymnos Island, transported in seawater into the lab, frozen in liquid nitrogen and stored at -70oC.

RNA isolation
Ten pieces from different parts of the sponge body were combined to obtain a comprehensive mRNA pattern of the animal. The frozen tissue was mechanically broken in liquid

406

nitrogen and total RNA was extracted using RNAwiz reagent (Ambion), followed by mRNA isolation with Oligotex mRNA Midi kit (Qiagen) according to the manufacturers instructions. The purity of the total RNA was confirmed by absorbance ratio at 260/280 nm 1.8 and its integrity was demonstrated by electrophoresis in 1.2% agarose gel containing formaldehyde stained with ethidium bromide. The amount of eukaryotic and bacterial material in the preparations was estimated from fluorescence intensities of different rRNA forms using Kodak Digital Science 1DTM software (Kodak). The enrichment of polyA+ RNA through the oligo(dT) resin was performed twice.

DNA sequencing
Sequencing was performed with the ABI PRISM gene analyzer (PerkinElmer) using Dyenamics ET (Amersham) or BigDye (Applied Biosystems) sequencing reagents.

Sequence data analysis

The similarities of obtained ESTs to known sequences were compared using Basic Local Alignment Search Tool (BLAST) blastn and tblastx searches against September 8, 2006 version of the complete nucleotide database downloaded from the NCBI FTP server (http://www.ncbi.nlm.nih.gov/ Ftp/). The EST database for Reniera sp. was downloaded from EnsEMBL (http://www.ensembl.org) and ESTs for Oscarella carmela were downloaded from NCBI Entrez (http://www. ncbi.nlm.nih.gov). The sequence motifs were analyzed with the widely used MEME motif discovery tool (http://meme.sdsc.edu/meme/), that finds one or more motifs in a collection of DNA or protein sequences by using the technique of expectation maximization to fit a two-component finite mixture model to the set of sequences (Bailey et al. 1994). Motifs with a maximum length of 20 bp (-minw 20) with possible locations on both strands of DNA (-revcomp) present at least once in sequences (-oops) were searched. To find motifs that are similar to the 2-5A synthetase of G. cydonium, its cDNA was included in the input sequence set. Only the motifs also found in the 2-5A synthetase cDNA of G. cydonium were selected as true ones. For assessing the dinucleotide frequency bias the odds ratio XY = fXY/fXfY, where fX denotes the frequency of the nucleotide X and fXY denotes the frequency of the dinucleotide XY, was used (Karlin and Mrazek 1997). For easier visual appearance and for comparison of different dinucleotide frequencies the theoretical dinucleotide frequency (fXfY) was set equal to 1 for each dinucleotide and declination (dXY) was calculated as dXY = XY - 1.

Construction of the cDNA library

The cDNA library was constructed using ZAP Express cDNA Synthesis Kit (Stratagene) according to the manufacturers manual. The reverse transcription products shorter than 500 bp were excluded from the preparation by size-fractionation of the cDNA. ZAP Express lambda phage library was packaged using Gigapack III Gold Packaging Extract (Stratagene) according to the manufacturers instructions.

Northern blot analysis

An amount of 10 g of total RNA was electrophoresed through the 1.2% formaldehyde/agarose gel and blotted onto a Hybond N membrane (Amersham). Hybridization was performed either with a 1.1 kb part of the G. cydonium 2-5A synthetase cDNA (EMBL Accession No. Y18497) or 1.0 kb part of S. domuncula (EMBL Accession No. AJ301653) 25A synthetase cDNA as described earlier (Wiens et al. 1999). The probe was labelled radioactively using DecalabelTM DNA Labeling Kit (Fermentas).

Hybridization and screening of the cDNA library

The plaque lifts were prepared using a Hybond N membrane according to the protocol for the ZAP Express System. The hybridizations and washing steps were carried out in low stringency conditions (42oC, 30% formamide, 6x SSC, 5x Denhardts solution, 0.5% SDS and 50 mM Tris-HCl, pH 7.0) and 32P-labelled 1.1 kb DNA fragment of 2-5A synthetase from G. cydonium was used as a probe. The plaques which were positive also in the second screen were subjected to in vivo excision protocol using ExAssist helper phage together with the XLOLR bacterial strain in accordance with the instructions provided by the manufacturer (Stratagene).

Results cDNA library construction

A comprehensive mRNA pattern for C. reniformis was obtained from a combined sample of different sponge body parts. According to the band intensities of the ribosomal RNAs, it was shown that 2/3 of the material was of bacterial origin (data not shown). The yield of the RNA extracted from the sponge tissue was comparatively low as compared to the expected yield in higher animals - approximately 30 g per 100 mg of wet tissue, supposedly due to the relative abundance of the extracellular matrix in these animals. At each step of mRNA purification, its integrity was analysed by electrophoresis. After two rounds of polyA+ enrichment the RNA still contained approximately 10% of ribosomal RNA. The amount of mRNA of the total eukaryotic material in the RNA preparations was approximately one percent. For the maximal enrichment of the cDNA library with full-length cDNA clones, the oligo(dT) was used for priming the first strand cDNA synthesis. The primary cDNA library was estimated to contain approximately 100 000 individual

Insert length
The cDNA insert size was determined by PCR using two universal primers (T3 and T7; 0.5 M each). The reaction mixture contained 0.2 mM each dNTP, 2.5 mM MgCl2, ~1 ng of pBK-CMW phagemid DNA and Taq DNA polymerase (Fermentas). The amplification was performed at 25 cycles of 94oC/45s, 55oC/45s and 72oC/90s in the GeneAmp 2400 thermal cycler (Applied Biosystems). PCR products were analysed by electrophoresis in 1% agarose gels.

407

clones. The insert size ranged from 0.7 to 3.5 kb with an average size of 1.8 kb. The sequencing of the 5 ends revealed the presence of ATG start codon in locations, homologous to those of orthologs (as demonstrated by BLAST analysis). We concluded that the majority of cDNAs in the library contained complete reading frames for encoded proteins.

EST sequencing
Two alternative approaches were used to select the clones for sequencing. As the original aim of the study was to isolate a cDNA encoding for 2-5A synthetase, we performed a low stringency screening with G. cydonium 2-5A synthetase cDNA, which had been shown to recognize a distinct 1.5 kb band in the Northern blot analysis (data not shown). The cDNA probe from S. domuncula did not give a hybridization signal in the Northern blot analysis. In addition to the clones originating from the screening approach we also picked a set of random phage clones. The end-sequences were obtained from 96 clones, approximately 83 000 basepairs in total. As a result, 46 continuous sequences (either from one or both ends) and 50 discontinuos sequences from longer inserts were obtained. The examples of the sequenced ESTs are given in Table 1.

Analysis of Chondrosia ESTs

To find out the possible common sequence motif(s) for the set of clones with positive signals obtained in screening of the cDNA library, we performed the MEME motif analysis (Bailey et al.1994). Only one major sequence motif was revealed. It was found in all sequences derived from a low stringency screening, as well as in the 3-terminal part of the probe sequence from G. cydonium. The sequence of this unusual Gcontaining triplet repeat is GVWGRHGRWGDWGRWGRA (presented in the IUPAC nucleotide code). We conclude that the presence of this sequence was most probably the basis for the hybridization of the selected clones. The comparison of the EST sequences from C. reniformis with all known nucleotide sequences (the NCBI nucleotide database) revealed a significant similarity distance between them: only 24% of ESTs were matched to the nucleotide database entries with a significant similarity (with e-value < 10-4). However, application of the tblastx algorithm against the NCBI nucleotide database allowed the identification of ortholog sequences for most of the isolated sequences (Table 1). The sequences from eight clones were not assigned to any known sequences. The best match of 48 (57%) of the sequences assigned to their vertebrate orthologs. For 22 sequences (26%) the best hits were found among the proteins from lower animals including sponges, 7 sequences from insects, two from nematodes and one from plants. One sequence gave the best match against a bacterial sequence. Recently the EST data for a marine sponge Oscarella carmela (Muricy and Pearse 2004) and both EST and genomic sequence data for another sponge species, Reniera sp. Schmidt 1862, have been made available. To estimate the similarity of our EST sequences to those from O. carmela and Reniera sp. and to compare them to sequences presented in the entire NCBI nucleotide database, we calculated similarity

and processivity of similar sequence regions for those three sets of sequences based upon blastn and tblastx results. Only the best hits from blast searches that corresponded to the criteria presented above (e-value < 10-4) were included (Table 2). The data in Table 2 demonstrate that the similarity of nucleotide sequences from C. reniformis compared to those from two other sponges and to the entire NCBI nucleotide database were in the same range. Another picture was drawn from the tblastx search. As expected, the sequences of C. reniformis were more similar to those of the two other sponges than to the entire NCBI database: the orthologous sequences of the higher organisms present in the entire database had a lower median similarity. The higher quality and better representation of sequence data in the whole NCBI database provided, however, longer median alignment lengths. One of the molecular characteristics of an organism suggested by Karlin and Ladunga (1994) is the dinucleotide composition of its genome. To characterize the sequence information of C. reniformis against that of O. carmela and Reniera sp., we calculated frequencies of dinucleotides for those three sponges and for the set of cDNA sequences which gave the best-scoring maches (from tblastx searches against the entire NCBI database). Figure 1 represents the declinations of the relative dinucleotide frequencies for each dinucleotide. The datasets of C. reniformis and the best-hit sequences demonstrated rather similar dinucleotide frequencies, including a low CpG and relatively high GpA and TpG contents. The CpG dinucleotide content of Reniera sp. was almost twice higher than that of C. reniformis. In general, the dinucleotide frequencies were more similar between C. reniformis and the best-hit datasets than between the three represented sponge datasets. In the case of the dinucleotide frequency declinations, the theoretical values for Reniera sp. were generally smaller than those of C. reniformis and O. carmela. This observation cannot be explained by the different C/G contents in these animals, as this characteristic was rather similar in all the used sequence datasets (42-45%).

Discussion
The widespread Mediterranean demosponge Chondrosia reniformis was selected for the construction of a cDNA library as one of the suggested model sponges for whole-genome sequencing. For its perspective utility in biotechnology for its high content of potential nanoparticle carrier material, collagen (Swatschek et al. 2002), it is currently used for development of cultivation conditions of the animal (Nickel et al. 2003, Sipkema et al. 2006). Sponges are hosting a variety of micro-organisms, mostly prokaryotes, as symbionts, commensals or a diet (Hooper and van Soest 2002, Hentschel et al. 2006). Indeed, two thirds of the RNA extracted from C. reniformis appeared to be of bacterial origin. The high bacterial content in sponge samples hampers the high quality cDNA preparation. Our sequence analysis provides clear evidence that carefully performed double oligo(dT) selection for mRNAs is sufficient to overcome this problem. The yield of mRNA in the eukaryotic

408

Table 1: Examples of the clones from the cDNA library of C. reniformis. The data for the longest continous streches from tblastx search are given. Clone No. 74-2-1 98-9 92-1 95-3 95-1 101-1-1 88-3 89-3 88-9 81-1-1 80-1-1 94-3 104-1-1 p79-7 89-9 94-2 93-9 74-2-2 p70-6 p66-2 98-6 p70-9 98-7 p70-1 90-1-2 94-9 73-1-1 p84-1 98-8 93-2 83-4-1 92-2 98-2 98-5 89-7 p79-2 p79-1 p66-6 T10 p66-8 p78-4 87-1-2 95-9 p84-5 86-1-2 91-1-2 88-2 89-6 92-6 96-3-1 Insert Sequenced (kb) (nt) 1.4 1.75 0.8 1.8 2.5 0.9 2.3 1.5 2.0 1.2 1.75 2.5 1.9 1.5 1.3 1.1 1.3 1.5 1.3 1.0 2.0 2.3 0.8 1.2 2.9 1.3 2.3 2.5 1.3 1.5 2.1 2.2 2.2 1.8 2.1 1.8 2.4 1.2 1.6 2.4 3.0 2.0 1.3 2.0 2.5 1.0 1.4 2.0 1.2 1.4 686 1181 693 626 1124 709 686 605 681 780 785 1252 1068 938 698 698 1018 1107 1148 470 572 620 573 1013 1167 625 1241 998 627 1100 778 670 592 1180 1300 888 580 906 415 910 592 1220 689 911 1207 821 1180 1155 616 1550 Prediction (best match) cytoplasmic actin sterol carrier protein 2 small GTPase Rac1 heat shock protein 70 (hsp70) actinin (F-actin cross-linking protein) ferritin (fer gene) heat-shock protein 70 protein phosphatase 2A catalytic subunit- aldehyde dehydrogenase malate/L-lactate dehydrogenase acyl-CoA synthetase long-chain family member 5 actinin alpha 2 zinc finger CCHC-type and RNA binding motif 1 major vault protein ribosome biogenesis protein Brix advillin Dullard homolog glutamine: fructose-6-phosphate amidotransferase protein phosphatase 1B, magnesium dependent similar to ribosomal protein L7 scavenger receptor cysteine-rich protein type 12 SNF related kinase (SNRK) ferritin-like protein mRNA for zinc finger protein Ser/Thr protein kinase Rsk-2 cortactin eukaryotic translation initiation factor 4 scavenger receptor cysteine-rich protein type 12 precursor similar to thyroid hormone receptor interactor 12 signal transducer and activator of transcription poly A binding protein actinin, alpha 1 (Actn1) similar to aldehyde dehydrogenase 4 family member A1 TBC1 domain family member 22A eIF4-related protein NAT1 aconitase (iron response element binding protein/IRE) heat-shock protein 90- protein disulphide isomerase cysteine dioxygenase apoptosis gene MA3 regulator of G-protein signaling 1 cytochrome P450 nucleoporin (NUP153) similar to lipidosin; very long-chain acyl-CoA synthetase collagen protein chromatin modifying protein 2B dynactin 2 (p50) full-length cDNA (unknown function ) myosin heavy chain from catch muscle (catchin) collagen Species Oikopleura longicauda Rattus norvegicus Xenopus tropicalis Anopheles albimanus Drosophila melanogaster Crassostrea gigas Cotesia rubecula Homo sapiens Homo sapiens Caenorhabditis elegans Homo sapiens Rattus norvegicus Homo sapiens Strongylocentrotus purpuratus Gallus gallus Heliocidaris erythrogramma Mus musculus Gallus gallus Rattus norvegicus Apis mellifera Strongylocentrotus purpuratus Homo sapiens Pinctada fucata Ciona intestinalis Xenopus laevis Suberites domuncula Homo sapiens Strongylocentrotus purpuratus Strongylocentrotus purpuratus Apis mellifera Danio rerio Mus musculus Gallus gallus Homo sapiens Xenopus laevis Gallus gallus Strongylocentrotus purpuratus Arabidopsis thaliana Gillichthys mirabilis Suberites domuncula Gallus gallus Suberites domuncula Takifugu rubripes Gallus gallus Suberites domuncula Homo sapiens Xenopus laevis Tetraodon nigroviridis Mytilus galloprovincialis Suberites domuncula Identities Positives e-value 111/116 82/102 124/138 128/160 107/131 99/167 90/103 75/98 88/139 70/134 61/128 37/56 28/47 49/51 76/118 44/73 48/111 47/61 79/127 72/107 66/136 74/119 64/93 63/87 69/82 45/83 36/69 25/40 38/80 32/59 37/50 46/58 53/115 47/88 49/92 52/66 56/64 48/95 27/54 43/91 34/76 29/74 28/52 20/36 27/51 18/30 31/94 25/64 35/106 18/41 115/116 92/102 131/138 148/160 118/131 134/167 97/103 86/98 111/139 91/134 87/128 45/56 39/47 51/51 90/118 57/73 65/111 55/61 98/127 85/107 87/136 89/119 80/93 74/87 73/82 60/83 50/69 23/48 55/80 42/59 43/50 52/58 79/115 64/88 68/92 60/66 60/64 70/95 39/54 38/88 50/76 39/74 35/52 27/36 33/51 25/30 59/94 37/64 62/106 26/41 1e-129 1e-89 1e-83 1e-82 6e-69 5e-67 2e-57 1e-55 3e-55 5e-55 6e-53 2e-51 3e-50 1e-49 3e-48 1e-47 3e-47 2e-46 2e-45 1e-45 6e-43 7e-42 1e-42 5e-41 5e-38 2e-38 5e-37 3e-37 7e-31 3e-30 5e-29 1e-29 1e-29 3e-27 4e-27 3e-27 4e-27 4e-25 1e-24 1e-23 1e-22 4e-21 1e-21 9e-21 3e-18 5e-17 e-15 1e-15 2e-14 3e-14

409 Table 1 (cont.)

89-4 p70-10 97-2-1 p66-7 p70-4 p84-3 p79-5 p70-8 p79-8 93-8 92-5 T8 93-4 92-3 85-4-3

2.8 2.0 1.7 2.3 1.8 2.2 2.2 0.6 3.0 1.4 2.5 1.1 1.0 2.0

1234 1260 1175 1253 612 624 995 571 692 599 629 750 868 901 1232

mitofilin (mitochondrial inner membrane protein) steroid hydroxylase CLIP associating protein 2 (Clasp2) DEAD/DEAH box helicase containing protein collagen type I archain 1 transglutaminase troponin C gelsolin (gels gene) arginine-tRNA-protein transferase smooth muscle myosin phosphatase LPS-binding protein CCAAT/enhancer binding protein gamma Bax inhibitor-1(testis enhanced gene transcript) carbohydrate (chondroitin) synthase

Strongylocentrotus purpuratus Homo sapiens Mus musculus Mus musculus Rana catesbeiana Homo sapiens Penaeus monodon Danio rerio Suberites domuncula Caenorhabditis elegans Gallus gallus Suberites domuncula Danio rerio Sus scrofa Mus musculus

35/93 74/119 32/72 35/70 20/42 30/37 15/35 23/53 23/50 19/42 25/62 15/40 26/68 13/28 22/76

56/93 89/119 49/72 42/70 23/42 32/37 26/35 30/53 33/50 29/42 34/62 23/40 38/68 19/28 39/76

2e-13 3e-13 7e-13 7e-12 5e-11 2e-09 4e-09 2e-07 2e-07 9e-06 2e-05 2e-05 4e-05 6e-05 2e-04

Table 2: The similarity of EST sequences from Chondrosia reniformis with those from Reniera sp. and Oscarella carmela. Number of hits used Blastn Reniera sp. O. carmela NCBI nt Tblastx Reniera sp. O. carmela NCBI nt Median of alignment homology SD (%) Median lengths of identical elements in alignment 108 108 107 41 43 63 Median lengths of alignment

18 14 25 61 49 69

87 5 88 6 89 6 61 16 56 18 50 15

126 108 123 66 74 127

RNA pool of a sample was close to that in higher animals (15% depending on the source). The initial aim of the study was to screen the cDNA library for 2-5A synthetase. Two 2-5A synthetase cDNA probes from G. cydonium and S. domuncula were used but only the probe from G. cydonium gave positive results in the Northern blot analysis. It is not surprising as the homology between orthologs from various sponge species may be rather low (Funayama et al. 2005). The 2-5A synthetases from G. cydonium and S. domuncula share only 28% of identity and 48% of similarity at the amino acid level (Wiens et al. 1999, Grebenjuk et al. 2002). The preliminary analysis of the end-sequencing data did not assain any of the sequenced ESTs to the 2-5A synthetase family, probably because of the low homology between the 2-5A synthetase sequences. Considering the specific product pattern of 2-5A synthetase from C. reniformis, rather different from that of any other species examined so far (Reintamm et al. 2003), this result may reflect significant differences in the primary structure of this particular 2-5A synthetase. On the other hand, the fact that this sequence is missing in the cDNA library from C. reniformis could be explained by the low abundance of this particular mRNA. Our analysis of the available sponge

EST databases revealed the presence of four copies of 2-5A synthetase out of 11 176 ESTs (5588 cDNA clones) of O. carmela but no similar sequence was found among 83 040 ESTs (about 30 000 cDNA clones) of Reniera sp. The low stringency screening of the cDNA library resulted in a set of clones with positive signals (altogether 86 clones) in which the common sequence motif was assigned to a G-starting triplet repeats of 18 nucleotides, also present in the probe sequence of G. cydonium. This sequence motif could be responsible for the numerous false positive results in low stringency hybridization conditions. In addition to the hybridized clones a set of randomly picked clones was subjected to DNA sequencing. Since the sequence analysis of all isolated cDNAs revealed their comparatively random origin from the transcriptome of C. reniformis (Table 1), we considered them as random samples of EST sequences to characterize the organism at the molecular level in our further analysis. For most sequences the orthologs were identified in the NCBI database. Only eight clones were not assigned to any known sequences and could be regarded as sponge specific ones. The relatively low abundance of the proteins characteristic exclusively for sponges was also observed by

410

Fig. 1: The declinations of the relative dinucleotide frequences for each dinucleotide from the predicted theoretical dinucleotide frequencies. Cho C. reniformis, Ren Reniera sp., Osc O. carmela, BH best hit.

Gamulin et al. (2000) during their analysis of almost 300 cDNAs, mostly from S. domuncula. The best match of the majority of the clones assigned to the corresponding sequences from vertebrates. These results are in agreement with the data asserting that genes and proteins from these simplest multicellular animals, sponges, share a higher degree of conservation with the orthologs from vertebrates than with those from insects and nematodes or yeast and plants (Gamulin et al. 2000, Perina et al. 2006, Cetkovic et al. 2007). Among the identified sequences in the cDNAs of C. reniformis, those coding for structural and cytoskeleton proteins such as actin, collagen, cortactin, actinin, dynactin and myosin, had the best matches with their orthologs (Table 1). Indeed, the proteins of the extracellular matrix are known to be highly conserved in diverse species (Har-El and Tanzer 1993, Nichols et al. 2006). The most abundant proteins in the extracellular matrix, collagens (rev. Exposito et al. 2002), were represented by three individual cDNAs. The cytoskeletonrelated cDNAs were represented in the C. reniformis cDNA library with several proteins involved in cytoskeleton formation, remodelling, adhesion and cell motility. The examples are actin, the F-actin cross-linking protein actinin, cortactin and a member of Rho-family GTPase Rac1 which induces cortactin to localize the cell membrane (Weed et al. 2000). Another set of highly conserved sequences fall into the group of ubiquitously expressed molecular chaperones (rev. by Robert 2003), represented by 70 kDa and 90 kDa heat-shock proteins. An essential part (24%) of the cDNAs

coded for proteins known to have enzymatic activities. Most prominent among them were protein phosphatases, aldehyde dehydrogenases, protein kinases and the acyl-CoA synthetase. The nucleotide sequences form C. reniformis were in the same range of similarity both with their orthologs from O. carmela and Reniera sp. and with the entire NCBI database (Table 2) but the deduced amino acid sequences were in general more close to two other sponges. We will not draw further conclusions in this respect as the dataset of C. reniformis used in the current study was relatively small. It still seems that sequence differences between sponge species may be as big as between any sponge and any other organism. This finding may be a reflection of the overall time of accumulation of evolutionary mutations which has been the same for any species. Dinucleotide sequences as the shortest elements of genetic information are the most stable units of information of a genome and therefore they have the lowest mutational frequencies throughout the evolution (Karlin and Burge 1995). Based on this fact the analysis of dinucleotide frequencies has been used to analyze evolutional proximity of both the prokaryotes (Nakashima et al. 1998) and higher organisms (Karlin and Mrazek 1997). This analysis has demonstrated that species with close or overlapping dinucleotide distributions are evolutionarily closer (Campbell et al. 1999). From the results of our dinucleotide frequency analysis of the EST sequences from three sponge species we could conclude that Oscarella carmela from the subclass Homoscleromorpha

411

could be evolutionarily closer to Chondrosia reniformis (the subclass Tetractinomorpha) than to Reniera sp. (the subclass Ceractinomorpha). However, at the moment it is difficult to fit our data about the reperesentatives of different sponge subclasses into available evolutionary trees built upon the analysis of ribosomal rDNA or cytochrome oxidase subunit I DNA sequences (Borchiellini et al. 2004, Nichols 2005). Certainly the situation would be clearer if a more detailed and extensive analysis of EST datasets from various sponge species become publicly available. Presently our data represent only the coding regions of the genome of C. reniformis. A more detailed analysis of the genomes of the animals under discussion and the comparison of the information obtained using data from other species will be our future goal.

Conclusion
The sequencing of 96 individual clones from the cDNA library of a marine sponge C. reniformis indicated that they contained full-length cDNA sequences as compared to their orthologs from higher animals. Presently assigned cDNA clones represent evolutionarily conserved sequences coding for various functional classes of proteins. Thus the preliminary analysis of ESTs isolated from the cDNA library of C. reniformis has provided the first molecular characterization of this organism. Characterization of the Chondrosia reniformis cDNA library has provided evidence for its high quality, low presence of bacterial sequences and high content of full-length cDNAs. The library is available to all interested researchers who should contact the relevant author.

Acknowledgements
This study was supported by the European Comission (project COOP-CT-2005, contract No. 017800) and the Estonian Science Foundation (grant No. 5932). The authors gratefully acknowledge the Genomic Toolbox LLC (Estonia) for bioinformatics support.

References
Bailey TL, Elkan C (1994) Fitting a mixture model by expectation maximization to discover motifs in biopolymers. Proc 2nd Int Conf Intell Syst Mol Biol AAAI Press, Menlo Park. pp. 28-36 Borchiellini C, Chombard C, Manuel M, Alivon E, Vacelet J, Boury-Esnault N (2004) Molecular phylogeny of demospongiae: implications for classification and scenarios of character evolution. Mol Phylogenet Evol 32: 823-837 Campbell A, Mrazek J, Karlin S (1999) Genome signature comparisons among prokaryote, plasmid, and mitochondrial DNA. Proc Natl Acad Sci USA 96(16): 9184-9189 Cetkovic H, Mikoc A, Mller WEG, Gamulin V (2007) Ras-like small GTPases form a large family of proteins in the marine sponge Suberites domuncula. J Mol Evol 64(3): 332-341 Exposito JY, Cluzel C, Garrone R, Lethias C (2002) Evolution of collagens. Anat Rec 268(3): 302-316 Funayama N, Nakatsukasa, M, Kuraku, S, Takechi, K, Dohi, M, Iwabe, N, Miyata, T, Agata, K (2005) Isolation of Ef silicatein and Ef lectin as molecular markers for sclerosytes and cells involved

in innate immunity in the freshwater sponge Ephydatia fluviatilis. Zool Sci 22: 1113-1122 Gamulin V, Mller I, Mller WEG (2000) Sponge proteins are more similar to those of Homo sapiens than to Caernohabditis elegans. Biol J Linn Soc 71: 821-828 Grebenjuk VA, Kuusksalu A, Kelve M, Schutze J, Schroder HC, Mller WEG (2002) Induction of (2-5)oligoadenylate synthetase in the marine sponges Suberites domuncula and Geodia cydonium by the bacterial endotoxin lipopolysaccharide. Eur J Biochem 269(5): 1382-1392 Har-El R, Tanzer M L (1993) Extracellular matrix 3: evolution of the extracellular matrix in invertebrates. FASEB J 7: 1115-1122 Hentschel U, Usher KM, Taylor MW (2006) Marine sponges as microbial fermenters. FEMS Microbiol Ecol 55(2): 167-177 Hooper JNA, van Soest RWM (2002) Systema Porifera: a guide to the classification of sponges. Kluwer Academic/Plenum Publishers, New York Imhoff JM, Garrone R (1983) Solubilization and characterization of Chondrosia reniformis sponge collagen. Connect Tissue Res 11(23): 193-197 Justesen J, Hartmann R, Kjeldgaard NO (2000) Gene structure and function of the 2-5-oligoadenylate synthetase family. Cell Mol Life Sci 57(11): 1593-1612 Karlin S, Burge C (1995) Dinucleotide relative abundance extremes: a genomic signature. Genetics 11(7): 283-290 Karlin S, Ladunga I (1994) Comparisons of eukaryotic genomic sequences. Proc Natl Acad Sci USA 91(26): 12832-12836 Karlin S, Mrazek J (1997) Compositional differences within and between eukaryotic genomes. Proc Natl Acad Sci USA 94(19): 10227-10232 Kelve M, Kuusksalu A, Lopp A, Reintamm T (2003) Sponge (2,5)oligoadenylate synthetase activity in the whole sponge organism and in a primary cell culture. J Biotechnol 100(2): 177180 Kortschak RD, Samuel G, Saint R, Miller DJ (2003) EST analysis of the cnidarian Acropora millepora reveals extensive gene loss and rapid sequence divergence in the model invertebrates. Curr Biol 13(24): 2190-2195 Kuusksalu A, Pihlak A, Mller WEG, Kelve M (1995) The (25)oligoadenylate synthetase is present in the lowest multicellular organisms, the marine sponges. Demonstration of the existence and identification of its reaction products. Eur J Biochem 232(2): 351-357 Mashimo T, Glaser P, Lucas M, Simon-Chazottes D, Ceccaldi PE, Montagutelli X, Despres P, Guenet JL (2003) Structural and functional genomics and evolutionary relationships in the cluster of genes encoding murine 2,5-oligoadenylate synthetases. Genomics 82(5): 537-552 Muricy G, Pearse JS (2004) A new species of Oscarella (Demospongiae: Plakinidae) from California. Proc Calif Acad Sci 55: 600-614 Nakashima H, Ota M, Nishikawa K, Ooi T (1998) Genes from nine genomes are separated into their organisms in the dinucleotide composition space. DNA Res 5(5): 251-259 Nichols SA (2005) An evaluation of support for order-level monophyly and interrelationship within the class Demospongiae using partial data from the large subunit rDNA and cytochrome oxidase subunit I. Mol Phylogenet Evol 34: 81-96

412

Nichols SA, Dirks W, Pearse JS, King N (2006) Early evolution of animal cell signaling and adhesion genes. Proc Natl Acad Sci USA 103(33): 12451-12456 Nickel M, Brummer F (2003) In vitro sponge fragment culture of Chondrosia reniformis (Nardo, 1847). J Biotechnol 100(2): 147159 Perina D, Cetkovic H, Harcet M, Premzl M, Lukic-Bilela L, Mller WEG, Gamulin V (2006) The complete set of ribosomal proteins from the marine sponge Suberites domuncula. Gene 366(2): 275284 Reintamm T, Lopp A, Kuusksalu A, Subbi J, Kelve M (2003) Qualitative and quantitative aspects of 2-5A synthesizing capacity of different marine sponges. Biomol Eng 20(4-6): 389-399 Robert J (2003) Evolution of heat shock protein and immunity. Dev Comp Immunol 27(6-7): 449-64 Sarkar SN, Sen GC (2004) Novel functions of proteins encoded by viral stress-inducible genes. Pharmacol Ther 103(3): 245-259 Sipkema D, Yosef NA, Adamczewski M, Osinga R, Mendola D, Tramper J, Wijffels RH (2006) Hypothesized kinetic models for

describing the growth of globular and encrusting demosponges. Mar Biotechnol 8(1): 40-51 Swatschek D, Shatton W, Kellermann J, Mller WEG, Keuter J (2002) Marine sponge collagen: isolation, characterization and effects on the skin parameters surface-pH, moisture and sebum. Eur J Pharm 53(1): 107-113 Tatsumi R, Hamada K, Sekiya S, Wakamatsu M, Namikawa T, Mizutani M, Sokawa Y (2000) 2,5-oligoadenylate synthetase gene in chicken: gene structure, distribution of alleles and their expression. Biochim Biophys Acta 1494(3): 263-268 Wiens M, Kuusksalu A, Kelve M, Mller WEG (1999) Origin of the interferon-inducible (2-5)oligoadenylate synthetases: cloning of the (2-5)oligoadenylate synthetase from the marine sponge Geodia cydonium. FEBS Lett 462(1-2): 12-18 Weed SA, Karginov AV, Schafer DA, Weaver AM, Kinley AW, Cooper JA, Parsons JT (2000) Cortactin localization to sites of actin assembly in lamellipodia requires interactions with F-actin and the Arp2/3 complex. J Cell Biol 151(1): 29-40

Porifera research: Biodiversity, innovation and sustainaBility - 2007

413

Life cycle of Paraleucilla magna Klautau, Monteiro and Borojevic, 2004 (Porifera, Calcarea)
Emilio Lanna(1*), Leandro C. Monteiro(2), Michelle Klautau(1)
Departamento de Zoologia, Instituto de Biologia, Universidade Federal do Rio de Janeiro, Rio de Janeiro, RJ, Brazil. emiliolanna@gmail.com, mklautau@biologia.ufrj.br (2) Departamento de Invertebrados, Museu Nacional, Universidade Federal do Rio de Janeiro, Rio de Janeiro, RJ, Brazil. lcmonteiro@mn.ufrj.br
(1)

Abstract: In the present study, we describe the life cycle of a calcaronean species, focusing on its reproductive period, the relationship between size and reproduction, and on its reproductive elements (oocytes and larvae). This work is the first study to describe reproductive aspects of a Calcarea in the Southern Hemisphere. For two years and three months, a population of Paraleucilla magna from Rio de Janeiro, Brazil, was sampled monthly. The reproductive period occurred during the summer (from January to March) as reported for Calcaronea sponges in the Northern Hemisphere. The first oocytes were observed in January, while in February most of the specimens had oocytes and larvae. In March larvae were more abundant than oocytes. Additionally, reproductive activity was not related to sponge size, as both small and large specimens showed similar reproductive elements. The oocytes of P. magna were similar to those of other studied calcareous sponges. Oocytes were spherical or globular and had a large central nucleus with a marked nucleolus. P. magna incubated amphiblastula larvae. Keywords: Brazil, Calcarea, life cycle, Paraleucilla magna, reproduction

Introduction
The class Calcarea comprises about 9.5% of all known sponge species (Lanna et al. in press). They are usually small and uncommon sponges that often live in cryptic habitats, which leaves them neglected by marine biologists and sometimes even by spongiologists (Manuel 2006). This is one of the reasons why we find large gaps in all fields of knowledge on calcareous sponges. One of the largest gaps is on the reproductive biology of the class. Studies on the reproduction of calcareous sponges began with Haeckel (1872, 1874), when he described his Gastraea theory. This theory raised the interest of many scientists to conduct research on the development of this class (e.g. Schulze 1875, Hammer 1908). Those studies were abundant throughout the end of the 19th century and reached their peak in the first half of the 20th century with the French researchers O. Duboscq and O. Tuzet. Duboscq and Tuzet published several studies on the oogenesis and development of calcareous sponges (e.g. Duboscq and Tuzet 1933, 1937, 1941, 1942, 1944), but little has been published on the subject since that period. In the 1970s Johnson published studies on the life cycle, gametogenesis and recruitment of two calcinean species: Clathrina aff. coriacea and Guancha blanca (Johnson 1978, 1979a, 1979b). In the 1980s some studies were carried out on the development and fertilization of Grantia compressa, mainly using ultrastructural analysis (Galissian 1983, Galissian and Vacelet 1985). The oogenesis and development of Sycon ciliatum were studied in two different works (Gaino et al. 1987, Franzen, 1988), and the

life history of Clathrina cerebrum was investigated by Gaino et al. (1996). Recently, the metamorphosis of a calcinean and the development of a calcaronean species were investigated (Amano and Hori 2001, Leys and Eerkes-Mendrano 2005, respectively). In the present study, we describe for the first time the life cycle of a calcareous species from the Southern Hemisphere, Paraleucilla magna Klautau, Monteiro and Borojevic, 2004 (family Amphoriscidae, subclass Calcaronea), describing the species reproductive period, reproductive elements and determining if there is any relationship between size and reproduction.

Materials and methods

This study was undertaken at Vermelha Beach, Rio de Janeiro, Brazil (225718S - 430942W) (Fig. 1). This beach is in a semi-enclosed bay, bounded by two large rocky shores, with low freshwater input, limited to rain water runoff, and sewage water drainage. Depending on the abundance of sponges, three to twelve whole specimens were collected monthly from January 2004 to March 2006, from 0 to 7 m of depth. The specimens were fixed in absolute ethanol and preserved in 70% ethanol. In the laboratory, each specimen was examined under a stereoscopic microscope to the analyses of the external morphology. The volume of each sponge was measured by fluid displacement in a graduated cylinder (e.g. Ereskovsky 2000). Then, fragments of the sponge were removed in order to prepare skeletal, spicule and histological slides.

414

Fig. 1: Study area (Source: Online Map Creator www.aquarius. geomar.de/omc - accessed in August, 04 2006).

Skeleton and spicule slides were prepared following standard procedures (Wrheide and Hooper 1999, Klautau and Valentine 2003). Skeletal slides were observed with light microscopy to analyse skeletal organization and composition, and also to search for reproductive elements. If reproductive elements were found, another fragment of the specimen was taken and decalcified in a solution of EDTA 10% pH 7 for 72 h. Subsequently, the fragment was washed in distilled water, dehydrated in ethanol and cleared in xylene. The fragment was embedded in paraffin, sectioned with a microtome (7 m thick), and the sections were stained with Harris Haematoxylin and Eosin solutions. All photographs were taken with a digital camera mounted on a Zeiss Axioscop II microscope at the Laboratrio de Captura de Imagens (PROIN) in the Biology Institute (UFRJ). We carried out the non-parametric Mann-Whitney test to determine if sponge volume differed between reproductive and non-reproductive specimens. To determine if sponge volume varied significantly among months we performed the non-parametric Kruskal-Wallis test. For all statistical analyses the GraphPad Prism program was used (GraphPad Softwares, Inc.).

Results
At Vermelha Beach, Paraleucilla magna was found associated with calcareous algae, but usually attached to rocks We observed photophilous as well as sciaphilous specimens. The external morphology of the sponges varied considerably, and could not be associated with seasonal variation. However, the majority of specimens consisted of massive body with a smooth surface and numerous oscula. Sponge colour was always white, except for some light pink specimens collected in March 2005. The main reproductive period of P. magna was from January to March (summer). In January, the specimens examined contained the primary stages of oocyte, which

measured 10 to 15 m diameter. Oocytes became larger (20 to 30 m) and more abundant in February, when embryos in different developmental stages and larvae were found. In March, the amount of reproductive elements decreased and larvae became more abundant than oocytes. Some isolated reproductive events were observed in June 2004 and 2005, but only in a reduced number of specimens (Fig. 2). Sponge volume varied significantly throughout the year, ranging from 0.1 cm3 (June 2005) to a maximum of 37 cm3 (January 2005) (KS = 67.51; df = 12; P < 0.001) (Fig. 3). However, reproductive activity had no relationship to volume, since small specimens (0.8 cm3) as well as large ones (37 cm3) showed reproductive elements (U = 1184; df = 1; p = 0.2440). The skeleton did not change its composition and organization throughout the year. Cortical triactines and tetractines, subatrial triactines and tetractines, a disorganized zone with scattered subatrial tetractines, and atrial triactines were always present (Fig. 4). As described above, Paraleucilla magna had abundant oocytes from January to April 2004, January to March 2005 and December 2005 to February 2006. Older oocytes were large ovoid to spherical cells (30 m). They were found in the mesohyl in contact with choanocyte chambers (Fig. 5A). These gametes had a large and granulated nucleus that was ovoid to elongate, in the centre of the cell. The nucleolus was easily observed and centrally located, and the cytoplasm was granulated (Fig. 5B). Nurse-cells were observed surrounding the oocytes. They were ovoid to elongate and were not observed inside the gametes. Spermatozoids or spermatic cysts were not found, consequently we cannot confirm whether P. magna is hermaphroditic or gonochoristic. Larvae were typically amphiblastula which in their later stages of development were 50 m in length. Unfortunately, only some of the developmental stages were observed. Larvae were ovoid, hollow, and had a smooth surface. Two main types of cells were observed (macromeres and micromeres). Micromeres were more numerous and occupied a large part of the larvae, while macromeres were positioned at the opposite pole and were less numerous (Fig. 5C). The nucleus of macromeres was large and central, while the nucleus of micromeres was small and apical. Larvae were spread throughout the body near choanocyte chambers, and were always surrounded by a protective membrane composed of elongate, thin cells. This membrane was in close contact with the larva at the point where the membrane contacted the choanocyte chambers (Fig. 5D).

Discussion
The reproductive season of Paraleucilla magna (summer) complies with the results of other studies on calcareous sponges in the Northern Hemisphere (Duboscq and Tuzet 1944, Johnson 1978, Ilan and Vacelet 1993, Gaino et al. 1996, Amano and Hori 2001, Leys and Eerkes-Mendrano 2005) (Fig. 6). Reproduction in spring and/or summer is also common in demosponges, such as Oscarella lobularis (Gaino et al. 1986), Halisarca dujardini, Myxilla incrustans and Iophon piceus (Ereskovsky 2000), and Ephydatia fluviatilis (Gaino et al. 2003).

415

Fig. 2: Reproduction effort of Paraleucilla magna. Oocyte and larval production from January 2004 to March 2006.

Fig. 3: Box plots of the volume of Paraleucilla magna in 2005. Each box displays the median, upper and lower quartiles of the distribution of sponge volume per month. Box whiskers represent the maximum and minimum range.

Fig. 4: Cross-sections of Paraleucilla magna. A. Non-reproducing specimen (June 2005); B. Reproducing specimen (March 2005). Circles surround larvae. Atrium (at) and cortex (cx).

This pattern of reproduction in spring/summer has been explained mainly by an increase in water temperature, which would appear to be the key factor that triggers reproduction in sponges (Fromont 1994, Fromont and Bergquist 1994, Ereskovsky 2000). However, variation in water temperature is less likely to be the key factor inducing reproduction of P. magna in Rio de Janeiro, Brazil, where the sea water undergoes only minor changes throughout the year, ranging from 21C to 26C, with a few sporadic upwelling events decreasing sea surface temperature to 16C (Paranhos et al. 2001, Alves et al. 2002). A possible environmental cue for reproductive activity in P. magna could be the increase in primary production in summer. Sevrin-Reyssac et al. (1979) observed an increment

of microphytoplankton abundance on the Rio de Janeiro coast during summer, which they correlated to the rainy season (summer) increasing the runoff of nutrients into the sea water. More detailed data, including temperature, primary productivity and rainfall throughout the year are necessary to propose environmental factors that could be triggering the reproduction of P. magna. Another interesting finding about the reproductive activity of calcareous sponges is the need for a minimum adult size. Sar (1955) observed a minimum size for reproduction in two calcareous sponges in Italy (Clathrina coriacea and Guancha blanca), and Gaino et al. (1996) found regression of the body and reorganization of the anastomosis of the tubes in Clathrina cerebrum after the reproductive period. In contrast,

416

Fig. 5: Photomicrographs of the reproductive elements of Paraleucilla magna: A. Oocyte (O) lying in the periphery of a choanocyte chamber (CC): mesohyl (Me), nucleus (N); B. Oocyte: cytoplasm (C), nucleus (N), nucleolus (Nu); C. Longitudinal section of a larva showing macromeres (Ma), micromeres (Mi), and the larval cavity (Ca) with amoeboid cells; D. Larva surrounded by the protective membrane (PM), near a choanocyte chamber (CC). Arrow indicates where the larva attaches to the protective membrane.

no relationship between volume and reproduction could be established for P. magna. We believe that this difference is related to the aquiferous system, which is asconoid in Clathrina and Guancha and leuconoid in P. magna. In asconoid sponges the atrium is completely layered by choanocytes and these cells become gametes (see e.g. Franzen 1988), hence the body would be reduced after reproduction. On the other hand, in leuconoid sponges the transformation of some choanocytes into gametes would not significantly change the body of the sponge. Results that support our hypothesis are found in studies on demosponges, which have only leuconoid aquiferous system, and where no correlation between size and reproduction has been found until now. In the Great Barrier Reefs species belonging to orders Haplosclerida and Petrosida did not show relation between size and maturity (Fromont 1994, Fromont and Bergquist 1994).

Despite the possibility of shrinkage and underestimation of size and shape of the cells promoted by the use of ethanol as fixative, the oocytes of P. magna were similar to those found in other calcaronean species, e.g. Leucosolenia botryoides, Grantia compressa and Leucandra aspera (Duboscq and Tuzet 1933, 1942). Additionally, the oocytes lack internal nurse cells them and nutrition appears to be by cytoplasmic bridges, a result also observed by Duboscq and Tuzet (1942) for other calcaronean species. Spermatic cysts or spermatozoids were not found in P. magna, hence, we do not know if this species is hermaphroditic or gonochoristic. In fact, very few studies have detected the presence of male gametes in calcareous sponges. Haeckel (1872) observed free spermatozoids of Sycortis quadrangulata and Poljaeff (1883) called some structures he found in four species spermospores (spermatic

417

Fig. 6: Reproductive period of different species of the class Calcarea: P. magna is the only species described from the Southern hemisphere. The species C. aff. coriacea, C. cerebrum and G. blanca belong to the subclass Calcinea and the other species belongs to the subclass Calcaronea.

cysts?). Duboscq and Tuzet (1933) studied the fecundation of Grantia compressa and Sycon ciliatum, but did not describe spermatozoids. Galissian and Vacelet (1983) observed spermiocysts in Petrobiona massiliana and, more recently, Anakina and Drozdov (2001) described free spermatozoids in Leucosolenia complicata. Both studies were carried out with transmission microscopy. Further studies on male gametogenesis are still required. As expected, P. magna had amphiblastula larvae with micromeres, macromeres and amoeboid cells. Only crosscells (cellules en croix) were not observed. Cross-cells were first described in Sycon and were linked to photoreception (Duboscq and Tuzet 1941). The absence of cross-cells has also been observed in Grantia compressa, which only has these cells in the early larval stages (Galissian 1983). Nonetheless, the absence of this cell type in P. magna may be associated to the fixation methodology utilized in the present work. The follicle found surrounding the larvae was originally called a placental membrane by Lufty (1957). The origin of this follicle probably occurs in the development of the egg, with the conversion of some blastomeres. Lufty (1957) considered that this follicle would be responsible for nutrition and protection of the embryo. Recent works, however, suggest that it helps larval inversion (Leys and Eerkes-Mendrano 2005, Leys and Ereskovsky 2006). Nonetheless, further

studies are necessary to understand the role of this membrane in amphiblastula development. This new knowledge of the reproductive period of P. magna will provide a basis for future studies on calcareous reproductive biology (as gametogenesis and embryology), and particularly for comparative studies in other areas in the southern hemisphere.

Acknowledgements
We thank Andr Rossi and Emiliano Caldern for their help in the collection of the sponges and Paulo Paiva for support. We are grateful to Andrea Junqueira, Antonio Sol-Cava, Carla Zilberberg and Guilherme Muricy, for their comments on this work. Grants and fellowships were provided by FAPERJ and CAPES.

References
Alves SLS, Pereira AD, Ventura CRR (2002) Sexual and asexual reproduction of Coscinasterias tenuispina (Echinodermata: Asteroidea) from Rio de Janeiro, Brazil. Mar Biol 140: 95-101 Amano S, Hori I (2001) Metamorphosis of coeloblastula performed by multipotential larval flagellated cells in the calcareous sponge Leucosolenia laxa. Biol Bull 200: 1- 20

418

Anakina RP, Drozdov AL (2001) Gamete structure and fertilization in the Barents Sea sponge Leucosolenia complicata. Russ J Mar Biol 27: 143-150 Duboscq O, Tuzet O (1933) Sur lovognse et la fcondation des ponges calcaires: Grantia compressa pennigera Haeckel et Sycon ciliatum Lieberkhn. Arch Zool Exp Gn 73(2): 45-56 Duboscq O, Tuzet O (1937) Lovognse, la fcondation et les premiers stades du dveloppement des ponges calcaires. Arch Zool Exp Gn 79: 157-316 Duboscq O, Tuzet O (1941) Sur les cellules en croix des Sycon (Sycon ciliatum Fabr., Sycon coronatum Ellis et Sol., Sycon elegans Bower.) et leur signification. Arch Zool Exp Gn 81(4): 151-163 Duboscq O, Tuzet O (1942) Recherches complmentaires sur lovogense, la fcondation et les premiers stades du dveloppement des ponges calcaires. Arch Zool Exp Gn 81: 395466 Duboscq O, Tuzet O (1944) Lovognse, la fcondation et les premiers stades du dveloppement de Sycon elegans Bower. Arch Zool Exp Gn 83: 445-459 Ereskovsky AV (2000) Reproduction cycles and strategies of the cold-water sponges Halisarca dujardini (Demospongiae, Halisarcida), Myxilla incrustans and Iophon piceus (Demospongiae, Poecilosclerida) from the White Sea. Biol Bull 198: 77-87 Franzen W (1988) Oogenesis and larval development of Scypha ciliata (Porifera, Calcarea). Zoomorphology (Berl.) 107: 349-357 Fromont J (1994) Reproductive development and timing of tropical sponges (Order Haplosclerida) from the Great Barrier Reef, Australia. Coral Reefs 13: 127-133 Fromont J, Bergquist PR (1994) Reproductive biology of three sponge species of the genus Xestospongia (Porifera: Demospongiae: Petrosiida) from the Great Barrier Reef. Coral Reefs 13: 119-126 Gaino E, Burlando B, Buffa P, Sar M (1986) Ultrastructural study of spermatogenesis in Oscarella lobularis (Porifera, Demospongiae). Invertebr Reprod Dev 10: 297-305 Gaino, E, Burlando, B, Buffa, P (1987) Ultrastructural study of oogenesis and fertilization in Sycon ciliatum (Porifera, Calcispongiae). Int J Invertebr Reprod Dev 11: 73-82 Gaino E, Bavestrello G, Cerrano C, Sar M (1996) Survival of the calcareous sponge Clathrina cerebrum (Haeckel, 1872) on a vertical cliff during the summer crisis. Ital J Zool 63 (1): 41-46 Gaino E, Rebora M, Corallini C, Lancioni T (2003) The life-cycle of the sponge Ephydatia fluviatilis (L.) living on the reed Phragmites australis in an artificially regulated lake. Hydrobiologia 495: 127142 Galissian MF (1983) tude ultrastructurale du developpement embryonnaire chez Grantia compressa F. (Porifera, Calcarea). Arch Anat Microsc 72: 59-75 Galissian MF, Vacelet J (1985) Fertilization and nutrition of the oocyte in the calcified sponge Petrobiona massiliana. In: Rtzler, K (ed). New perspectives in sponge biology. Smithsonian Institution Press, Washington D.C. pp.175-181 Haeckel E (1872) Die Kalkschwmme, Eine Monographie, vols 1-3. Verlag von Georg Reimer, Berlin Haeckel E (1874) Die Gastraea-Theorie, die phylogenetische Classification des Thierreichs und die Homologie der Keimbltter. J Zeitschr Naturwiss 8: 1-55

Hammer, E (1908) Neu beitrage zur kenntnis der histologie und entwicklung von Sycon raphanus. Arch Biontol 2: 289-334 Ilan M, Vacelet J (1993) Kebira uteoides (Porifera, Calcarea) a recent Pharetronid sponge from coral reefs. Ophelia 38(2): 107-116 Johnson MF (1978) Studies on the reproductive cycles of the calcareous sponges Clathrina coriacea and C. blanca. Mar Biol 50: 73-79 Johnson MF (1979a) Recruitment, growth, mortality and seasonal variations in the calcareous sponges Clathrina coriacea (Montagu) and C. blanca (Miklucho-Maclay) from Santa Catalina Island, California. In: Lvi C, Boury-Esnault N (eds). Biologie des spongiaires. Colloques Internationaux du CNRS, vol. 291. ditions du CNRS, Paris. pp. 325-334 Johnson MF (1979b) Gametogenesis and embryonic development in the calcareous sponges Clathrina coriacea and Clathrina blanca from Santa Catalina Island, California. Bull South Calif Acad Sci 78(3): 183-191 Klautau M, Valentine C (2003) Revision of the Genus Clathrina (Porifera, Calcarea). Zool J Linn Soc 139: 1-62 Klautau M, Monteiro L, Borojevic R (2004) First occurrence of the genus Paraleucilla (Calcarea, Porifera) in the Atlantic Ocean: P. magna sp. nov. Zootaxa 710: 1-8 Lanna E, Rossi AL, Cavalcanti FF, Hajdu E, Klautau M (in press) Calcareous sponges from So Paulo state, Brazil (Porifera: Calcarea: Calcinea) with the description of two new species. J Mar Biol Assoc UK Leys SP, Eerkes-Mendrano D (2005) Gastrulation in calcareous sponges: In search of Haeckels Gastraea. Integr Comp Biol 45: 342-351 Leys SP, Ereskovsky AV (2006) Embryogenesis and larval differentiation in sponges. Can J Zool 84: 262-287 Lufty RG (1957) On the placental membrane of calcareous sponges. Cellule 58: 231-236 Manuel M (2006) Phylogeny and evolution of calcareous sponges. Can J Zool 84: 225-241 Paranhos R, Andrade L, Mendona-Hagler LC, Pfeiffer WC (2001) Coupling bacterial abundance with production in a polluted bay. In: Faria BM, Farjalla VF, Esteves FA (eds). Aquatic microbial ecology in Brazil. Series Oecol Brasiliensis, vol. IX. PPGE-UFRJ, Rio de Janeiro, Brazil. pp. 117-132 Poljaeff MA (1883) Reports on the Calcarea dredged by H.M.S. Challenger during the years 1873-1876. Rep Sci Res Voy H.M.S. Challenger, Zool 8(2): 1-76 Sar M (1955) La nutrizione dellovocita in Calcispongie Omoceli. Ann Ist Mus Zool Univ Napoli 7(2): 1-30 Schulze FE (1875) Ueber den bau und die entwicklung von Sycandra raphanus Haeckel. Zeitschr Wiss Zool 25: 247280 Sevrin-Reyssac J, Machado MC, Schtze MLM, Bibas SG, de Lima IC, Lima CA, Esteves CP (1979) Biomasse et production du phytoplancton de la baie de Guanabara (tat de Rio de Janeiro, Brsil) et du secteur ocanique adjacent. Variations de mai juillet 1978. Bull Mus Natl Hist Nat Paris 4e ser., 1, sect B, no 4: 329354 Wrheide G, Hooper JNA (1999) Calcarea from the Great Barrier Reef. 1: Cryptic Calcinea from Heron Island and Wistari Reef (Capricorn-Bunker Group). Memoir Queensl Mus 43: 859-891

Porifera research: Biodiversity, innovation and sustainaBility - 2007

419

Assessing the utility of sponge microbial symbiont communities as models to study global climate change: a case study with Halichondria bowerbanki
Nathan Lemoine(1,2), Nicole Buell(1), April Hill(1), Malcolm Hill(1*)
Department of Biology, University of Richmond, Richmond, VA 23173, USA. nlemoine@disl.org, nicole.buell@richmond.edu, ahill2@richmond.edu, mhill2@richmond.edu (2) Dauphin Island Sea Lab. 101 Bienville Blvd., Dauphin Island, AL 36528. USA
(1)

Abstract: We examined the stability of sponge microbial communities in Halichondria bowerbanki, a common sponge in the Chesapeake Bay (Virginia, USA), in response to potentially stressful thermal environments. Sponges were reared under three thermal regimes that corresponded to current thermal maxima and realistic projections of SST increases over the next 50 and 100 years. No obvious changes in the density or diversity of microbial populations were observed when electron micrographs were compared among treatments. However, denaturing gradient gel electrophoresis (DGGE) uncovered consistent changes among the three treatments. Some bands obtained via DGGE disappeared when sponges were reared under thermally stressful temperatures indicating loss or significant reduction in population size of some species. Some bands were present only in the warmest treatment, which may indicate that a rare species increased in relative frequency in the microbial community. Sponge microbial communities offer numerous opportunities to explore hypotheses generated by current community ecological theory. We are only beginning to understand the significance of these sponge-microbe associations in the context of global warming. Keywords: community stability, global warming, microbial symbionts

Introduction
Among eukaryotes harboring prokaryotic symbionts, some marine sponges stand apart for harboring extraordinarily dense and diverse communities that are composed, at least in part, of a derived, sponge-specific microflora (e.g., Lee et al. 2001, Hentschel et al. 2003, Hill 2004, Hill et al. 2006). It is not uncommon for the symbiont community to occupy a larger volume than that of the host cells (Santavy et al. 1990, Brantley et al. 1995). The symbionts appear to perform vital functions for their hosts and perhaps the entire benthic community (e.g., nutrient cycling (Corredor et al. 1988) and production of secondary metabolites (Piel 2004)). These ancient symbioses can involve prokaryotic and/or eukaryotic partners, and can occur in intra- and extracellular locations (Wilkinson 1987). Despite the evolutionary and ecological significance, we know relatively little about how these associations are structured and how stable these communities are in the face of environmental change (but see Vicente 1990, Cerrano et al. 2001). While there is a wealth of information on the response of corals and their zooxanthellar symbionts to stressful conditions (e.g., Hoegh-Guldberg 1999), only a handful of studies have examined the stability of symbiont communities in sponges as they respond to environmental factors (e.g., Vicente 1990, Friedrich et al. 2001, Thoms et al. 2003, Taylor et al. 2004). There are often few clear phenotypic markers that

could be used for visual assessment of association health for sponges that lack symbiotic phototrophs. Thus, we have only limited ideas about what kinds of changes may be occurring between sponges and their symbionts during stress inducing events. This ignorance represents an unfortunate state of knowledge because changes may occur that have important consequences for sponge health and ultimately community structure. Survey-based analyses have found that some spongebacterial symbioses appear to be stable across large spatial scales despite including hosts that belong to different orders and occupy highly dissimilar habitats (e.g., Hentschel et al. 2002, Taylor et al. 2004, Webster et al. 2004, Olson and McCarthy 2005, Hill et al. 2006). Furthermore, some laboratory and field experiments have found sponge-bacterial symbioses to be remarkably stable. Friedrich et al. (2001) found that antibiotics and starvation had little effect on the symbiont communities present in Aplysina aerophoba reared in aquaria. Thoms et al. (2003) transplanted Mediterranean A. cavernicola from 40 m to depths as shallow as 7 m and found that bacterial communities harbored by the sponge were largely unaffected by the novel environmental conditions. Not all sponges, however, harbor immutable bacterial communities. Cymbastela concentrica exhibited limited, short-term changes in microbial communities over small geographic distances. Furthermore, the microbial community present in temperate populations of C. concentrica was distinct

420

from the community existing in tropical populations of this species (Taylor et al. 2004, 2005). These studies highlight the need for additional work in the analysis of sponge-associated microbial diversity. This is especially true for temperate sponges, which remain very poorly studied (but see Wichels et al. 2006) and will also face significant increases in sea surface temperature (SST - McCarthy et al. 2001). The Third Intergovernmental Panel on Climate Change (IPCC) found that the Earths average temperature had increased by 0.8C since the start of the 20th century and may exceed an increase of 2C by 2050 (McCarthy et al. 2001, Thomas et al. 2004). In marine systems, projected increases in SSTs over the next 100 years indicate that stressful conditions will intensify in many marine habitats and that habitat deterioration will become more apparent (e.g., Hoegh-Guldberg 1999, Knowlton 2001, Walther et al. 2002, Bellwood et al. 2004, Sheppard and Rioja-Nieto 2005). For example, relatively small ( 0.1C) increases in average SST correlate strongly with increases in the amount of bleached coral cover while mass bleaching events occur when SST anomalies exceed only 0.2C (McWilliams et al. 2005). Models indicate that SST will increase globally up to 1.2C over the next 50 years and 2-4C over the next 100 years (McCarthy et al. 2001, Wigley and Raper 2001). These temperature increases are not confined to the surface but can penetrate to > 500 m (Barnett et al. 2005). The consequences of 1-2C increases for marine fauna of non-tropical regions can be quite severe. In 1999, the Mediterranean experienced temperature increases in this range that lead to mass mortality of gorgonian and sponge populations in the Ligurian Sea (Cerrano et al. 2000, 2001, Perez et al. 2000). Our ignorance of the ecological effects of climate change of the sort predicted may have substantial ecological consequences. The goal of this work was to begin to assess changes in the bacterial community harbored by the temperate sponge Halichondria bowerbanki when exposed to elevated temperatures of extended duration. We argue that this system could serve as a model for marine ecosystems for a number of reasons. The complexity of the symbiont community provides opportunities to explore whether relatively small changes in environment can modify ecological relationships among competing species (e.g., Jiang and Morin 2004). The often important physiological role of symbiont species provides an opportunity to examine the consequences of environmental change on symbiont, and therefore host, performance. Finally, the sponge-symbiont association provides a tractable system and therefore permits experimentation in a manner that is exceedingly difficult if not impossible for many other communities.

to the laboratory in aerated containers and processed within 4 h.

Experimental design
Sponges were reared in separate 1-liter containers that were part of a re-circulating seawater system with a 100 L reservoir tank ( 125 L total system volume). Three samples ( 1 cm3) from each sponge were placed in each container. Temperature conditions for the containers were assigned randomly. Control containers (n = 7) were set at the thermal maximum experienced at the collection site ( 29C). This temperature was achieved by heating the reservoir tank to the desired temperature via a submersible heater. After passing through the heated system, the water was diverted to a chiller that cooled the water to room temperature before it returned to the reservoir tank. Treatment chamber temperatures were either 1C (n = 9) or 2C (n = 7) above the thermal maximum. Individual heaters were placed in the containers to achieve the desired temperatures. Temperature profiles were stabilized before sponges were placed in the system. During the 14 d experiment, approximately one third of the volume of the entire system ( 40 liters) was exchanged with fresh Bay water every other day to regulate salinity, remove waste products, and provide sponges with natural bacterioplankton communities. Water temperature, salinity and overall sponge health were monitored twice daily for the 14 days of the experiment. Small adjustments to heaters were made as required to keep temperatures at the desired level.

Molecular analysis
At the conclusion of the experiment, two samples from each replicate container were immediately frozen at -80C for subsequent molecular work. DNA was isolated from sponge samples that had both choanoderm and pinacoderm using a standard CTAB isolation protocol (Hill et al. 2004). DNA was quantified and diluted to 50 ng l-1. Regions of small subunit rDNA were amplified using two sets of primers. The first set of primers (1055f and 1406r; Webster et al. 2004) amplified a 350bp region conserved in the domain Bacteria. The second set of primers (PRBA338f and PRUN518r) were taken from vres et al. (1997) and amplified approximately 180bp of a more variable region and were used to sample a wider taxonomic range of sponge-associated microbes. Promegas GoTaq reagents were used for all PCR reactions. Ten pmol of each primer and 50 ng of DNA were used in each reaction. The GoTaq master mix yielded final concentrations of 1.5 mM and 2.5 mM for MgCl2 and dNTPs respectively when diluted with ultra-pure water. We typically ran 20 l final volume reactions. The PCR program included an initial denaturation run of 2 min at 95C. This was followed by 35 cycles of the following profile: 1 min at 95C, 30 s at 57C (for 1055f and 1406r) or 30 s at 67C (for PRBA338f and PRUN518r), and 45 s at 72C. This was followed by a final extension of 6 min at 72C. PCR products were run on a 1% agarose gel to examine the quality of amplification. To compare profiles of the bacteria found in sponges reared at different temperatures, the Bio-Rad Dcode Universal

Materials and methods Sponge collection

Halichondria bowerbanki individuals (n = 3) were collected via snorkeling from a pier at the mouth of the York River in the Chesapeake Bay (Lat. +37.246, Long. -76.500). Collections were made near the mean low water line during the end of July 2005. Sponges were immediately transported

421

Mutation Detection System was used for Denaturing Gradient Gel Electrophoresis (DGGE). A 1 mm thick, 10% (w/v) polyacrylamide gel with a 30-60% formamide:urea denaturing gradient was poured following manufacturers instructions. Electrophoresis was performed in a 1X TAE buffer solution at 33V for 16 hours for PCR products derived using primers 1055f and 1406r. For products obtained using primers PRBA338f and PRUN518r, gels were run at 44V for 18 hours. All gels were stained with ethidium bromide and images were captured using UV transillumination on the Kodak Gel Logic camera system.

Microscopy
For electron microscopy, sponge samples from each treatment were fixed in 2.5% glutaraldehyde and sterile filtered seawater. Samples were washed in 0.2 M cacodylate buffer (pH 7.4) and postfixed with 1% OsO4 and 1% Uranyl acetate. Samples were then dehydrated in an ethanol series, infiltrated in propylene oxide, and embedded in Embed 812 plastic resin. After polymerization, 1 mm sections were cut and treated in 4% hydrofluoric acid:76% ethanol to dissolve spicules. These sections were then dehydrated, infiltrated, and embedded again following the protocol described above. Ultrathin sections were stained with uranyl acetate and quick lead. Micrographs were taken using a Joel transmission electron microscope.

Fig. 1: Treatment temperatures were measured frequently for the two weeks of this experiment. Significant differences were maintained among the treatment and control replicates and variability was maintained, with few exceptions, within 20% of the mean (see text for absolute values).

Results
Average (SEM) temperatures during the course of the experiment were as follows: control (28.6 0.04 C), +1C treatment (29.8 0.04 C), and +2C treatment (31.3 0.04 C). Temperature variability observed during the course of the experiment arose primarily as a function of fluctuations in building temperature and daily values for each treatment are shown in Fig. 1. Regression analysis indicated that there were no differences in temperature among days within treatments (i.e., slope of the line was 0, Fig. 1). Analysis of variance indicated significant differences among treatments (p < 0.0001; F2,226 = 926.06) and post hoc Tukeys HSD indicated that each level was significantly different from the others. There were no significant differences among days within the controls or +2C treatment (p > 0.28, F9,60 = 1.58 and p > 0.14, F9,60 = 1.58 respectively). Although a significant difference was found among days within the +1C treatment (p = 0.025, F9,79 = 2.29), post hoc Tukeys HSD found no significant differences among days. Implementing Hsus Best MCB post hoc test indicated that days 2 and 6 were significantly different than days 1, 8, and 13. A fraction of the bacteria associated with Halichondria bowerbanki was found to be sensitive to the temperature changes experienced in these experiments. DGGE banding patterns differed between treatment and control sponges for the PCR reactions that used primers 1055f and 1406r (Fig. 2). At least three bands present in the control treatments were consistently lost in the +1C and +2C treatment levels (arrowheads on the left, Fig. 2). Two bands discovered in the +2C treatment were not present in the control or in the +1C treatment (arrowheads on the right, Fig. 2). Comparisons

Fig. 2: PCR-Denaturing Gradient Gel Electrophoresis performed with primers 1055f and 1406r. The arrows on the left identify species that may be sensitive to the temperature treatments employed in these experiments and were subsequently lost from the symbiont community. The arrows on the right identify two bands that were only present in the warmest temperature treatment.

between the sample sponges and samples of the source sponge taken at the time of collection were not performed for this primer set. Because the PRBA338f and PRUN518r primers amplify a more variable region of 16S rDNA, the banding patterns obtained were more complex than those observed above. However, as in Fig. 2, we found evidence of the loss and gain of bands in the +2C treatment when compared to the other temperature conditions (see white and black arrowheads in

422

Fig. 3). We also found some evidence of aquarium effects because one band was only present in samples taken from the source sponges at the time of collection (star on the left, Fig. 3). Another band was found in the majority of the heated treatments, but was also found in one control sponge (star on the right, Fig. 3). Bacterial densities were found to be low in H. bowerbanki (Fig. 4A-C). However, five distinct phenotypic types were identified in this study. The first type (Fig. 4D) had thin projections that extended in some cases > 0.5 m from the surface of the bacterium. Another type (Fig. 4E), had projections that resembled small bumps on the surface of the bacterium, which were uniformly spaced at an interval of 0.16 m. Another commonly observed microbe was cylindrical to ovoid in shape and had a non-uniformly electro-dense cytoplasm (Fig 4F). A rod shaped type was also observed that had electro-dense inclusions in its cytoplasm (Fig. 4G). Finally, a cylindrical to ovoid bacterial type was identified that appeared to be encircled by a collagen layer and was observed to have a nuclear region that was indistinct and fibrous (Fig. 4H).

Discussion
This research represents one of the first attempts to study changes in sponge microbial symbiont communities that might be expected if sea surface temperatures (SSTs) increase as predicted under conditions of global warming. The treatment levels employed in these experiments were chosen to reflect conditions under conservative models of SST warming. During the experiment, our goal was to keep temperatures below 20% of the mean and we largely achieved this for the +2C treatment and the control. One of the problems we faced in these experiments was maintaining the +1C treatment levels so that they did not overlap with either the control or the +2C treatment. In one instance, a replicate in the +1C treatment approached the mean temperature for the controls. In another instance, a replicate in the +1C treatment exceeded the average temperature for the +2C treatment; two other replicates also came close to that average. The significant differences recorded using Hsus MCB can be explained by the latter data. Nonetheless, significant differences recorded among treatment levels, and lack of overlap within the 95% confidence interval, indicate that we largely achieved the desired temperature separations. However, in future experiments, we will work to control the precision within and separation between treatments. Due to their location in shallow water habitats, however, daily fluctuations in temperature are likely to be as extreme as was encountered in this experiment. Results indicate that the sponge symbiont community changes in response to conditions that mimic 50-100 year projections of sea surface temperature increases. Analysis of DGGE gels revealed segments of the symbiont community that are negatively affected by increases in temperature (Figs. 2 and 3). These bacteria may be lost from the community (a local extinction) or may decrease to such low densities that we were unable to amplify their DNA. Regardless, the loss of species under the types of controlled conditions described here would provide an important opportunity to study in detail

Fig. 3: PCR-Denaturing Gradient Gel Electrophoresis performed with primers PRBA338f and PRUN518r. White arrows identify species that appear to be sensitive to the temperature treatments and were subsequently lost from the symbiont community. Black arrows label bands that appear to be tolerant of warmer temperatures. The star on the left is located near a band that was only found in the source sponges and was thus lost from control and treatment sponges alike. The star on the right indicates bands that were predominantly found in +1C and +2C treatments and were absent from source sponges and most control sponges.

ecological consequences of reductions in species richness that are predicted under various global warming scenarios. Furthermore, our data indicate that some species of bacteria harbored by sponges may increase in frequency under the highest temperatures (Figs. 2 and 3). It is important to determine whether these bacteria begin at lower frequencies (below the threshold easily detected with DGGE) and then gain a competitive advantage under the new temperature environment or are acquired from the environment. Given limitations in DGGE analysis to monitor changes in relative frequencies, we are exploring other options (e.g., FISH) to follow community level parameters of species richness and relative abundance. We are also beginning to obtain sequences from the variable bands so that we may begin to identify the species that are susceptible to this type of environmental change. As has been found in other studies examining changes in sponge symbiont microbial communities, morphological characteristics proved of limited value in attempts to chart
Fig. 4: Electron micrographs of symbionts from H. bowerbanki. A and B. Images of bacterial types found in the sponge mesohyl. C. A bacterium in a vacuole after phagocytosis. D-H. The five bacterial types identified in the H. bowerbanki (see text for general descriptions).

423

424

changes in species richness. The morphological diversity discovered in H. bowerbanki appears similar to that identified in other temperate sponges. While it was impossible to determine which bacterial types were affected by the thermal conditions devised in these experiments, some interesting types were uncovered. We are particularly interested in identifying species with projections (Fig 4D and E) as these types have been observed in other sponges (M. Maldonado, pers. comm.). Greater work needs to be done to more fully understand the consequences of warming sea surface temperatures in the context of sponge microbial symbiont communities. The work presented here demonstrates that controlled experimental conditions can be created to mimic proposed increases in seawater temperatures and community profiles can be monitored using molecular tools. What is needed is a more precise determination of which species of symbiont are susceptible to temperature changes and the community-wide consequences of these shifts in community composition and structure. We are currently working to answer these types of questions.

Acknowledgements
We thank the Virginia Institute of Marine Science, Gloucester Point, Virginia for permission to collect on their pier. Carolyn Marks guidance with electron microscopy is greatly appreciated. Comments from two reviewers were especially helpful.

References
Ahn YB, Rhee SK, Fennell DE, Kerkhof LJ, Hentschel U, Hggblom MM (2003) Reductive dehalogenation of brominated phenolic compounds by microorganisms associated with the marine sponge Aplysina aerophoba. Appl Environ Microbiol 60: 4159-4166 Barnett TP, Pierce DW, AchutaRao KM, Gleckler PJ, Santer BD, Gregory JM, Washington WM (2005) Penetration of humaninduced warming into the worlds oceans. Science 309: 284-287 Bellwood DR, Hughes TP, Folke C, Nystrom M (2004) Confronting the coral reef crisis. Nature 429: 827-833 Brantley SE, Molinski TF, Preston CM, DeLong EF (1995) Brominated acetylenic fatty acids from Xestospongia sp., a marine sponge-bacterial association. Tetrahedron 51: 7667-7672 Burja AM, Hill RT (2001) Microbial symbionts of the Australian Great Barrier Reef sponge, Candidaspongia flabellata. Hydrobiologia 461: 41-47 Cerrano C, Bavestrello G, Bianchi CN, Cattaneo-Vietti R, Bava S, Morganti C, Morri C, Picco P, Sara G, Schiaparelli S, Siccardi A, Sponga F (2000) A catastrophic mass-mortality episode of gorgonians and other organisms in the Ligurian Sea (North-western Mediterranean), summer 1999. Ecol Letters 3: 284-293 Cerrano C, Magnino G, Sara A, Bavestrello G, Gaino E (2001) Necrosis in a population of Petrosia ficiformis (Porifera, Demospongiae) in relations with environmental stress. Ital J Zool 68: 131-136 Corredor JE, Wilkinson CR, Vicente VP, Morell JM, Otero E (1988) Nitrate release by Caribbean reef sponges. Limnol Oceanogr 33: 114 -120 Duineveld BM, Kowalchuck GA, Keijzer A, Dirk van Elsas J, van Veen JA (2001) Analysis of bacterial communities in the rhizosphere

of Chrysanthemum via denaturing gradient gel electrophoresis of PCR-amplified 16S rRNA as well as DNA fragments coding for 16S rRNA. Appl Environ Microbiol 67: 172-178 Ferris MJ, Ward DM (1996) Denaturing gradient gel electrophoresis profiles of 16S rRNA-defined populations inhabiting a hot spring microbial mat community. Appl Environ Microbiol 62: 340-346 Friedrich AB, Fischer I, Proksch P, Hacker J, Hentschel U (2001) Temporal variation of the microbial community associated with the Mediterranean sponge Aplysina aerophoba. FEMS Microbiol Ecol 38: 105-113 Griffiths RI, Whiteley AS, ODonnell AG, Bailey MJ (2003) Physiological and community responses of established grassland bacterial populations to water stress. Appl Environ Microbiol 69: 6961-6968 Hentschel U, Fieseler L, Wehrl M, Gernert C, Steinert M, Horn M, Hacker J (2003) Microbial diversity of marine sponges. In: Mller WEG (ed). Molecular marine biology of sponges. Springer-Verlag, Heidelberg. pp. 59-88 Hentschel U, Hopke J, Horn M, Friedrich AB, Wagner M, Hacker J, Moore BS (2002) Molecular evidence for a uniform microbial community in sponges from different oceans. Appl Environ Microbiol 68: 4431-4440 Hill AL, Tetrault J, Hill M (2004) Isolation and expression analysis of a poriferan Antp-class Bar-/Bsh-like homeobox gene. Dev Genes Evol 214: 515-523 Hill MS, Hill AL, Lopez N, Harriott O (2006) Sponge-specific bacterial symbionts in the Caribbean sponge, Chondrilla nucula (Demospongiae, Chondrosida). Mar Biol 148: 1221-1230 Hill RT (2004) Microbes from marine sponges: A treasure trove of biodiversity for natural products discovery. In: Bull AT (ed). Microbial diversity and biopropecting. ASM Press, Washington DC. pp. 177-190 Hoegh-Guldberg O (1999) Climate change, coral bleaching and the future of the worlds coral reefs. Mar Freshw Res 50: 839-866 Imhoff JF, Sthr R (2003) Sponge-associated bacteria: general overview and special aspects of bacteria associated with Halichondria panicea. In: Mller WEG (ed). Marine molecular biotechnology, vol. 1: sponges (Porifera). Springer, Berlin. pp. 3557 Imhoff JF, Trper HG (1976) Marine sponges as habitats of anaerobic phototrophic bacteria. Microbial Ecol 3: 1-9 Jiang L, Morin PJ (2004) Temperature-dependent interactions explain unexpected responses to environmental warming in communities of competitors. J Anim Ecol 73: 569-576 Knowlton N (2001) The future of coral reefs. P Natl Acad Sci USA 98: 5419-5425 Lee YK, Lee JH, Lee HK (2001) Microbial symbiosis in marine sponges. J Microbiol 39: 254-265 Lovell CR, Piceno YM, Quattro JM, Bagwell CE (2000) Molecular analysis of diazotroph diversity in the rhizosphere of the smooth cordgrass, Spartina alterniflora. Appl Environ Microbiol 66: 38143822 McCarthy JJ, Canziani OF, Leary NA, Dokken DJ, White KS (2001) Climate Change 2001: Impacts, Adaptation, and Vulnerability; Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change (IPCC). Cambridge University Press, UK McWilliams JP, Ct IM, Gill JA, Sutherland WJ, Watkinson AR (2005) Accelerating impacts of temperature-induced coral bleaching in the Caribbean. Ecology 86: 2055-2060

425

Olson JB, McCarthy PJ (2005) Associated bacterial communities of two deep-water sponges. Aquat Microb Ecol 39: 47-55 vres L, Forney L, Daae FL, Torsvik V (1997) Distribution of bacterioplankton in meromictic Lake Slenvannet, as determined by denaturing gradient gel electrophoresis of PCR-amplified gene fragments coding for 16s rRNA. Appl Environ Microbiol 63: 33673373 Perez T, Garrabou J, Sartoretto S, Harmelin JG, Francour P, Vacelet J (2000) Mortalit massive dinvertbrats marins: un vnement sans precedent en Mditerrane nord-occidentale. CR Acad Sci Ser 3 Sci Vie/Life Sci 323: 853-865 Piel J, Hul D, Wen G, Butzke D, Platzer Matthias, Fusetani N, Matsunaga S (2004) Antitumor polyketide biosynthesis by an uncultivated bacterial symbiont of the marine sponge Theonella swimhoei. P Natl Acad Sci USA 101: 16222-16227 Santavy DL, Willenz P, Colwell RR (1990) Phenotypic study of bacteria associated with the Caribbean sclerosponge Ceratoporella nicholsoni. Appl Environ Microb 56: 1750-1762 Sheppard CRC, Rioja-Nieto R (2005) Sea surface temperature 1871-2099 in 38 cells in the Caribbean region. Mar Environ Res 60: 389-396 Taylor MW, Schupp PJ, Dahllf I, Kjelleberg S, Steinberg PD (2004) Host specificity in marine sponge-associated bacteria, and potential implications for marine microbial diversity. Environ Microbiol 6: 121-130 Taylor MW, Schupp PJ, de Nys R, Kjelleberg S, Steinberg PD (2005) Biogeography of bacteria associated with the marine sponge Cymbastela concentrica. Environ Microbiol 7: 419-433 Thomas CD, Cameron A, Green RE, Bakkenes M, Beaumont LJ, Collingham YC, Erasmus BFN, de Siqueira MF, Grainger A,

Hannah L, Hughes L, Huntley B, van Jaarsveld AS, Midgley GF, Miles L, Ortega-Huerta MA, Peterson AT, Phillips OL, Williams SE (2004) Extinction risk from climate change. Nature 427: 145148 Thoms C, Horn M, Wagner M, Hentschel U, Proksch P (2003) Monitoring microbial diversity and natural product profiles of the sponge Aplysina cavernicola following transplantation. Mar Biol 142: 685-692 Vicente VP (1990) Response of sponges with autotrophic endosymbionts during the coral-bleaching episode in Puerto Rico. Coral Reefs 9: 199-202 Walther GR, Post E, Convey P, Menzel A, Parmesan C, Beebee TJC, Fromentin JM, Hoegh-Guldberg O, Bairlein F (2002) Ecological responses to recent climate change. Nature 416: 389-395 Webster NS, Negri AP, Munro MMHG, Battershill CN (2004) Diverse microbial communities inhabit Antarctic sponges. Environ Microbiol 6: 288-300 Webster NS, Wilson KJ, Blackall LL, Hill RT (2001) Phylogenetic Diversity of Bacteria Associated with the Marine Sponge Rhopaloeides odorabile. Appl Environ Microb 67: 434-444 Wichels A, Wrtz S, Dpke H, Schtt, Gerdts G (2006) Bacterial diversity in the breadcrumb sponge Halichondria panicea (Pallas). FEMS Microbiol Ecol 56: 102-111 Wigley TML, Raper SCB (2001) Interpretations of high projections for global-mean warming. Science 293: 451-454 Wilkinson CR (1987) Significance of microbial symbionts in sponge evolution and ecology. Symbiosis 4: 135-146 Woollacott RM, Pinto RL (1995) Flagellar basal apparatus and its utility in phylogenetic analyses of the Porifera. J Morphol 226: 247-265

Porifera research: Biodiversity, innovation and sustainaBility - 2007

427

Biomarkers in marine sponges: acetylcholinesterase in the sponge Cliona celata

Daniela Marques(1), Marise Almeida(1,2), Joana Xavier(3), Madalena Humanes(1*)
Centro de Qumica e Bioqumica - Departamento de Qumica e Bioqumica da Faculdade de Cincias da Universidade de Lisboa, Edifcio C8, Campo Grande, 1749-016 Lisboa, Portugal. mmhumanes@fc.ul.pt (2) Laboratrio de Biomateriais, Faculdade de Medicina Dentria da Universidade de Lisboa, Cidade Universitria, 1649-003 Lisboa, Portugal. marise.almeida@fmd.ul.pt (3) Institute for Biodiversity and Ecosystem Dynamics, University of Amsterdam, Mauritskade 57, 1092 AD Amsterdam, The Netherlands and CIBIO - Plo Aores, Centro de Investigao em Biodiversidade e Recursos Genticos, Departamento de Biologia, Universidade dos Aores, 9501-801 Ponta Delgada, Portugal. xavier@science.uva.nl
(1)

Abstract: Sponges, ubiquitously occurring organisms, have many characteristics that indicate their potential as biomonitors. Since it is very important to find the most wide range of organisms that can act as early warners of pollution, we investigated the presence of the biomarker acetylcholinesterase in the cosmopolitan sponge Cliona celata Grant, 1826. Specimens of C. celata were collected from pre-selected sites along the western coast of Iberian Peninsula considered undisturbed in terms of pollution, although some anthropogenic activities may occur. A standard analytical method was used production of the anion 5-thio-2-nitrobenzoate for the determination of acetylcholinesterase. Our goal was to check the presence of this enzyme in this particular sponge, to validate the method used for these organisms, to determine the normal interval range of acetylcholinesterase activity and to check for eventual differences in the collection sites. Acetylcholinesterase activity was detected in all samples with two exceptions - one from Arrbida and another from Berlengas, both from the Portuguese coast. The mean value for the remaining set was 2.87x10-8 U/ mg prot /g of fresh sponge and the variation interval was (0-11.010-8 U/ mg prot /g of fresh sponge). No significant differences have been found between collection sites but the samples preserved for a longer time displayed smaller values. Keywords: Acetylcholinesterase, Biomarker, Cliona celata, Northeastern Atlantic, Sponges

Introduction
The guarantee of a sustainable use of the sea can only be achieved by monitoring the environmental quality of the marine media. Responses to pollution at the various levels of biological organization are widely used in environmental monitoring studies since, by definition, pollution implies hazard to living resources. The biomarkers represent the changes that may occur, from the molecular to the organism level, due to the toxic effects of exposure to chemical contaminants. In addition, biomarker responses occur prior to alterations at the population and community levels, so they can be predictive and anticipatory. Thus, biomarkers have the ability to diagnose causes and may act as early warning signals of ecosystem-level damage (Tsangaris et al. 2006). Acetylcholinesterase (AChE) is an enzyme essential to the correct transmission of nerve impulses. An inhibition of this enzyme has been used to detect and measure the biological effects of organophosphorus and carbamates pesticides in the marine environment (Fulton and Key 2001). Moreover, AChE may be also inhibited by heavy metals and surfactants (Bocquen et al.1997, Hamza-Chaffai et al.1998, Guilhermino et al. 1998). The suitability of this biomarker

for use in freshwater and marine environments that are not, at least apparently, contaminated by pesticides, was also demonstrated (Galgani et al. 1992, Payne et al. 1996). The main objective of this study was to investigate the variation of AChE in C. celata collected in three pre-selected sites, along the Iberian Western coast, considered undisturbed in terms of pollution, although some anthropogenic activities can occur. Cliona celata is a well-known yellow sponge, representative of the family Clionaidae (DOrbigny, 1851) widely distributed in the North East Atlantic sub littoral rocky shores, characterized by its bioerosive properties (Marques et al. 2006). This sponge has been reported as possessing a great adaptive plasticity in its relationship to environmental variables and being present in areas subjected to high environmental stress (Carballo et al. 1996, Bell et al. 2002). Furthermore, this species infests often marine cultures of mussels and oysters, emerging (with Cliona viridis) as the single excavating species trying to colonize the particular calcareous substrate provided by oyster shells, in the north-western Mediterranean Sea (Rossell et al. 1999). Mussels and oysters are considered key species in biomonitoring programs. Being filter feeding animals, the presence of sponges associated with these

428

commercially important molluscs maybe important in terms of ecosystem functioning. Although some compounds that inhibit the acetylcholinesterase activity have been extracted from sponges (Sepcic et al. 1998, Philp 1999, Sepcic et al. 1999, Sepcic et al. 2001, Bunc et al. 2002, Aoki et al.2003), only one paper reports the presence of cholinesterase-like activity in a species (Spongia officinalis), suggesting a particular esterase form also fitted for hydrolyzing choline esters with a low catalytic efficiency (Talesa et al. 1996).

Materials and methods Chemicals

5,5 dithiobis-2-nitrobenzoate (DTNB) and acetylthiocholine (ATC) iodide were purchased from Sigma. Phosphate salts for buffer preparation were purchased from Merck. All other reagents used were analytical grade products from various sources and all solutions were made in bidistilled water.

Biological specimens
Specimens of C. celata (n = 21) were collected by scuba diving, in several areas of two Portuguese Marine Natural Reserves (Arrbida and Berlengas), and in Graa Ferrol, Spain, from 1998 to 2004 (Fig. 1). After collection, the specimens were transported to the laboratory in isothermal containers, frozen and stored at -20 C. Vouchers specimens have been kept in 96% (v/v) ethanol, for taxonomical analysis. The samples were named according to the respective collection site: B for Berlengas, A for Arrbida and F for Ferrol, followed by a number that corresponds to the order of collection.

Fig. 1: Location of the sponge collection sites.

Sample and extract preparation

Since marine sponges possess several organisms associated with them, prior to any study, the biological material was cleaned from any macro organisms (such as small algae or crustacean) or even alien bodies, like sediments or rocks. After that, the samples were homogenized with a blender in ten volumes of ice-cold 0.2 M Tris-SO4 (pH 8.3) buffer, for two minutes, followed by a thirty minutes extraction process at 4C. The homogenates were centrifuged at 5000 g for 35 min at 4C. The filtered supernatants, were then ready to use.

104 M-1cm-1), between 2 and 4 minutes after the reaction start. TNB production by the thiol materials from sponge cells is determined by a similar assay, substituting the ATC by equal volume of phosphate buffer. Three experiments were done for each determination and the value presented is the mean value. One enzyme unit was defined as the amount of enzyme which catalyses the hydrolysis of 1 pmol of substrate per minute.

Protein determination in sponge extracts

Protein content was determined by the modified Lowry method (Bensadoun et al. 1976)

Acetylcholinesterase activity determinations

AChE activity measurements were carried out at 20C, using ATC as substrate, according to Brown (Brown et al. 2004), representing a slight modification of the method previously described by Ellman (Ellman et al. 1961). This method was used since it is extremely sensitive and small quantities of sample can be used. The assay mixture was composed of 2.6 mL of 0.1 M Na-phosphate buffer, pH 8.0, 0.1 mL of DTNB 0.01M with NaHCO3 1.5 mg/mL in Na phosphate buffer 0.1M, pH 7.0, 0.02 mL of ATC 0.075 M in Na phosphate buffer 0.1 M, pH 7.0, and 0.4 mL of enzyme solution (extract). The convertion of DTNB to 5thionitrobenzoate (TNB) was monitored at 412nm (= 1.36 x

Results
The Ellmans method for acetylcholinesterase activity determination is based on the rate of production of thiocholine, as a result of the hydrolysis of acetylthiocholine, a satisfactory substitute of the natural substrate acetylcholine. This reaction is monitored through the reaction of the produced thiocholine with the DTNB ion, producing the yellow anion 5-thionitrobenzoate (Fig. 2) (TNB) (Ellman, et al. 1961). The production of TNB can be spectrophotometrically monitored at 412 nm. The reaction with the thiol is fast enough, and therefore is not the rate limiting step in the determination of the enzymatic activity and also, at the used concentration range, does not inhibit the enzymatic hydrolysis.

429

Fig. 2: Chemical equation for the production of 5-thionitrobenzoate.

This method is extremely efficient for AChE activity determinations, in comparison with other previously developed methods determinations are made in the visible region of the spectrum, the molar absorptivity of the 5thionitrobenzoate ion is 13600 M-1cm-1, twice the value for acetylthiocholine (5140 M-1cm-1), allowing a considerable increase of the sensivity. Another important feature lies in the absence of a homogenate treatment prior to the execution of the method. Reproducibility and repeatability were also checked, before determining the AChE specific activity values for the samples. In order to perform these studies one sponge specimen was randomly chosen (B33). From this sample, 100 mL of homogenate was produced and divided into five vials. Four vials were kept at -80 C to perform reproducibility studies and the remaining vial was used for the repeatability studies. To achieve the repeatability conditions, we must have the same measuring procedure, the same person, the same instrument, the same place and all the repetition experiments should be done in a short period of time. Ten experiments were performed for enzymatic activity determination. The values obtained are shown in Table 1. To check the reproducibility of the results, we used the crude extracts kept at -80C and the activity was determined changing the parameter time varying the number of days with the sample kept at -80C. Results are presented in Table 2. The AChE activities of the 21 samples of C. celata are shown in Fig. 3. These results can be used to determine the expected variability of AChE activity levels in this species. The mean value found, i.e., the central location of this set, is 2,865 x 10-8 U/mg of protein/g of fresh sponge. Table 3 presents the AChE specific activity values for all the sponge samples, for the three different locations. A statistical treatment, using non-parametric methods, show that the distribution of AChE values is asymmetrical and similar to the probability density distribution for the 2 function (Fig. 4). There is a clear spread of values to the right that can be due to outliers (elements outside the data general pattern). To check if these values are truly outliers we calculate the peripheral barriers they should contain, for normal populations, all the elements of the sample. The expected variability for the AChE specific activity is [0; 11.0 x 10-8] U/ mg prot/ g of

Table 1: Repeatability study for the AChE specific activity determination in fresh B33 sponge extract. Experiments 1 2 3 4 5 6 7 8 9 10 Mean value Standard deviation Standard uncertainty U/ mg prot/ g of fresh sponge 2.32 x 10-9 1.16 x 10-9 2.32 x 10-9 2.32 x 10-9 3.47 x 10-9 3.47 x 10-9 2.32 x 10-9 2.32 x 10-9 2.32 x 10-9 2.32 x 10-9 2.43 x 10-9 0,66 x 10-9 0,21 x 10-9

Table 2: Reproducibility study for the AChE specific activity determination in fresh B33 sponge extract. Experiments Day 0 Day 1 Day 2 Day 3 Day 4 Mean value Standard deviation Standard uncertainty U/ mg prot/ g of fresh sponge 0.243 x 10-8 0.425 x 10-8 0.656 x 10-8 0.656 x 10-8 0.579 x 10-8 0.512 x 10-8 0.177 x 10-8 0.079 x 10-8

fresh sponge considering this set of C. celata samples. All the values determined for the 21 samples are within this interval; this means that there are no suspicious values in the examined samples.

Discussion
Marine sponges are interesting organisms for biomonitoring. The lack of tissue differentiation in these

430

Fig. 3: Histogram representing the distribution form of AChE levels in the 21 C. celata samples studied.

organisms represents here an advantage, since it facilitates the obtention of homogenates, simplifying the laboratory procedures. C. celata was chosen because it is a cosmopolitan species, with a very extensive geographical distribution along the Portuguese coast, occuring predominantly in the massive (or ) form (Xavier, personal observations), making the biomass collection easier. This study evaluated the presence of acetylcholinesterase in C. celata, its normal range values and the variation interval in this species. To our knowledge, this is the first set of data for acetylcholinesterase activity in sponges. The only reference found, is a previous paper that describes the partial purification of AChE from Spongia officinalis. To get a normal range interval, is just the first step to assess the viability of AChE as a biomarker in sponges. The Ellman method for acetylcholinesterase activity determinations is simple, easy to apply and presents a quite acceptable uncertainty, specially because the order of the levels of AChE activities ( 10-9). The tests of reproducibility and repeatability allowed us to evaluate the proximity of the values, obtained either with the same analytical conditions or introducing small variations and can give us an idea of the expectable dispersion of the results, solely due to the analytical method. In this case, the most important uncertainty componentis the reproducibility; for this reason it is necessary to pay particular attention to the assay conditions for instance, the acetylthiocholine should be kept in an ice bath, during the experiments, in order to avoid its spontaneous degradation. The interval range determined for AChE activity in the sponge C. celata is [0; 11.0 10-8] U/ mg prot./g of fresh sponge. This interval means that the values found between zero and 11 x 10-8 indicate that there is not a reason to suspect of a problem inhibiting the acetylcholinesterase activity. The values determined are summarized in Table 3, according to the stations of collection. The samples collected in Arrabida Natural Park present the highest variation - one of them did not show any activity at all (Table 3). These results may suggest that some environmental influence can occur in some of the samples for instance, the specimen A03/75 collected in Arflor, a site that can be

Table 3: AChE specific activity (in U/mg prot/g of fresh sponge) for the C. celata samples. Specimen A03/6 A03/7 A03/8 A03/70 A03/71 A03/75 B33 B109 B124 B173 B277 B341 B350 B357 B373 B400 B401 B418 FE01 FE03 FE04 Collection site Rampa da Secil (Arrabida) Collecting date Deep (m) AChE

15/03/2003 10 - 20 3.7510-8 4.7410-8 7.7310-9 Ponta da Barragem 16/07/2003 7 0.3510-8 (Cabo Espichel) 2.3210-8 (Arrabida) Arflor 28/07/2003 7 0.00100 (Arrabida) Furado Grande 04/07/1998 4 0.51210-8 (Berlenga) Estela 11/08/1998 13 0,00100 (Berlenga) Furado Grande 12/08/1998 2 0.42810-8 (Berlenga) Furado Grande 10/08/1999 ? 3.5110-8 (Berlenga) Furado Grande 22/07/2000 4 3.2110-8 (Berlenga) Furado Grande 30/06/2001 5 2.9810-8 (Berlenga) Furado Grande 30/06/2001 5 2.8810-8 (Berlenga) Furado Grande 30/06/2001 5 2.8510-8 (Berlenga) Furado Grande 30/06/2001 5 5.4010-8 (Berlenga) Farilhes 22/07/2003 29 3.0210-8 (Berlenga) Farilhes 22/07/2003 25 2.8510-8 (Berlenga) Furado Grande 22/07/2003 4 9.4810-8 (Berlenga) Ria de Ferrol 22/05/2003 15-18 4.6210-8 Ria de Ferrol 23/05/2003 15-18 8.1910-8 Ria de Ferrol 23/05/2003 15-18 3.2810-8

431

as a biomarker, since the proposed stressors (carbamates, pesticides) should inhibit this activity. The low activity values we found for normal situations, do not allow a significant response to pollutants.

Acknowledgements
We wish to thank Reserva Natural das Berlengas and Instituto da Conservao da Natureza for all the facilities granted and Dr. G. Calado and Dr. H. Gaspar for collection of some samples. This work was supported by Fundao para a Cincia e a Tecnologia, Project: POCTI/ QUI/45670/2002.

References
Fig. 4: Values of AChE specific activity for the C. celata samples expressed in activity units per milligram of protein per gram of fresh sponge. Aoki S, Wei H, Matsui K, Rachmat R, Kobayashi M (2003) Pyridoacridine alkaloids inducing neuronal differentiation in a neuroblastoma cell line, from marine sponge Biemna fortis. Bioorg Med Chem 11(9): 1969-73 Bell JJ, Barnes DKA, Turner JR (2002) The importance of micro and macro morphological variation in the adaptation of a sublittoral demosponge to current extremes. Mar Biol 140: 75-81 Bensadoun A, Weinstein D (1976) Assay of proteins in the presence of interfering materials. Anal Biochem 70: 241-250 Bocquen G, Roig A, Fournier D (1997) Cholinesterases from the common oyster (Crassostrea gigas). FEBS Letters 14: 1-6 Brown M, Davies IM, Moffat CF, Redshaw J, Craft JA (2004) Characterization of choline esterases and their tissue and subcellular distribution in mussel (Mytilus edulis). Mar Environ Res 57: 155-169 Bunc M, Strupi-Suput J, Vodovnik A, Suput D (2002) Toxic effects of head-to-tail 3-alkylpyridinium polymers isolated from the marine sponge Reniera sarai in rat. Toxicon 40(7): 843-849 Carballo JL, Naranjo S, Garcia-Gomez J (1996) Use of marine sponges as stress indicators in marine ecosystems at Algeciras Bay (southern Iberian Peninsula). Mar Ecol Prog Series 135: 109-122 Ellman GL, Courtney D, Andreas V, Featherstone RM (1961) A new and rapid colorimetric determination of acetylcholinesterase activity. Biochem Pharm 7: 88-95 Fulton MH, Key PB (2001) Acetylcholinesterase inhibition in estuarine fish and invertebrates as an indicator of organophosphorus insecticide exposure and effects. Environ Toxicol Chem 20(1): 3745 Galgani F, Bocquene G, Cadiou Y (1992) Evidence of variation in cholinesterase activity in fish along a pollution gradient in the north-sea. Mar Ecol Progr Ser 91: 77-82 Guilhermino L, Barros P, Silva MC, Soares AMVM (1998) Should the use of inhibition of cholinesterases as a specific biomarker for organophosphate and carbamate pesticides be questioned? Biomarkers 3: 157-163 Hamza-Chaffai A, Romeo M, Gnassia-Barel, El Abed A. (1998) Effect of copper and lindane on some biomarkers measured in the clam Ruditapes decussatus. Bull Env Cont Tox 61: 397- 404 Marques D, Esteves AI, Xavier J, Almeida M, Humanes M (2006) Marine Sponge Cliona celata as a potential bioindicator In: Alpoim MC, Morais PV, Santos MA, Cristovao AJ, Centeno J, Collery P, Libbey J (eds). Metal ions in biology and medicine, vol. 9. Eurotext, Paris. pp. 451- 456 Payne JF, Mathieu A, Melvin W, Fancey LL (1996) Acethylcholinesterase, an old biomarker with a new future? Field

subjected to some anthropogenic influences from urban wastes of a nearby town (Setbal) and a neighbour cement factory, did not present any AChE activity. The low values are in line with the inhibitory effect of AChE by organic pollutants. We can also observe that one of the samples for the Berlengas Natural Park, an archipelago composed by one small island with several tiny islands, located in the Atlantic Ocean (10/15 km from the mainland), did not show any AChE activity , as well. Even considering the upwelling of this region, we do not believe that this is a true indication of an environmental problem, since the other samples collected in the same area, in the same period present measurable values, of the same magnitude of the values obtained for the other regions. This is really a good example how these values can be misleading if the number of data is not sufficiently high. We suspect that the time of sample preservation can contribute, as well, to the variability of the AChE specific activity values. This was observed for the specimens collected at Berlengas islands, but further studies must be performed to confirm or to dismiss this observation. Ra de Ferrol, Galicia, is the only station that is not a Natural Reserve. The values obtained there are within the interval we propose for normal AChE activity. AChE activity is, usually, related to the transmission of nerve impulses. Sponges do not possess a nervous system as such, but some form of nervous transmission must occur, at least to promote contractions. According to a previous paper (Talesa et al. 1996), low values for AChE activity were also found in Spongia officinalis and the authors proposed that these low values are due to acetylthiocholine not being the right substrate for the enzyme found in sponges, but it can also be the case the enzyme is not required for so many functions as in higher organisms. This is a quite interesting question to be rexplored. This study clearly shows that a form of AChE is present in sponges and its value is usually low. We found that for C. celata a value between zero and 11x10-8 U/mg prot./g of fresh sponge should be considered normal. With this set of results, the variation found seems to exclude the use of AChE

432

trials in association with two urban rivers and a paper mill in Newfoundland. Mar Pol Bull 32: 225-231 Philp RB (1999) Cadmium content of the marine sponge Microciona prolifera, other sponges, water and sediment from the eastern Florida panhandle: possible effects on Microciona cell aggregation and potential roles of low pH and low salinity. Comp Biochem Physiol C 124(1): 41- 49 Rossell D, Uriz M-J, Martin D (1999) Infestation by excavating sponges on the oyster (Ostrea edulis) populations of the Blanes littoral zone (North-Western Mediterranean Sea). J Mar Biol Assoc UK 79(3): 409-413 Sepcic K, Marcel V, Klaebe A, Turk T, Suput D, Fournier D (1998) Inhibition of acetylcholinesterase by an alkylpyridinium polymer from the marine sponge, Reniera sarai. Biochim Biophys Acta 1387(1-2): 217-225

Sepcic K, Mancini I, Vidic I, Franssanito R, Pietra F, Macek P, Turk T (2001) Antibacterial and anticholinesterase activities of aplysamine-4, a bromotyrosine-derived metabolite of a Red Sea marine sponge. J Nat Toxins 10(3): 181-191 Sepcic K, Poklar N, Vesnaver G, Fournier D, Turk T, Macek P (1999) Interaction of 3-alkylpyridinium polymers from the sea sponge Reniera sarai with insect acetylcholinesterase. J Protein Chem 18(3): 251-257 Talesa V, Romani R, Rosi G, Giovannini E (1996) Presence of an acetylcholinesterase in the Cnidarian Actinia equina (Anthozoa: Actiniaria) and of a thiocholine ester-hydrolyzing esterase in the sponge Spongia officinallis (Demospongiae: Keratosa). J Exp Zool 276: 102-111 Tsangaris C, Papathanasiou E, Cotou E (2006) Assessment of the impact of heavy metal pollution from a ferro-nickel smelting plant using biomarkers. Ecotoxicol Environ Saf 66(2): 232-43

Porifera research: Biodiversity, innovation and sustainaBility - 2007

433

Antileishmanial epidioxysterols from extracted sterols of the Colombian marine sponge Ircinia campana
Diana M. Mrquez F.(1), Sara M. Robledo R.(2), Alejandro Martnez M.(1*)
Grupo de Productos Naturales Marinos, Facultad de Qumica Farmacutica, Universidad de Antioquia. Medelln, Colombia. dmarquez@farmacia.udea.edu.co, amart@farmacia.udea.edu.co (2) Facultad de Medicina, Universidad de Antioquia. Calle 67 N 53108 - Apartado Areo 1226. Medelln, Colombia. srobledo@guajiros.udea.edu.co
(1)

Abstract: Sterols with 5,7-3-hidroxyandrostadiene nuclei were isolated from the Colombian marine sponge Ircinia campana (Lamarck, 1814), through bioassay-guided fractioning. The sterols fraction was oxidized under light and air controlled conditions to produce the 5,8-epidioxysterols which were identified by GC/MS. Nine compounds were identified: (1) 5,8-epidioxy-24-norcholesta-6,22-dien-3-ol; (2) 5,8-epidioxy-cholesta-6,22-dien-3-ol; (3) 5,8-epidioxy-(24)methylcholesta-6-en-3-ol (epimer 1); (4) 5,8-epidioxy-cholesta-6-en-3-ol; (5) 5,8-epidioxy-(24)-methylcholesta-6en-3-ol (epimer 2); (6) 5, 8-epidioxy-(24)-ethylcholesta-6,22-dien-3-ol; (7) 5,8-epidioxy-cholesta-6,9-dien-3-ol; (8) 5,8-epidioxy-cholesta-6,9,22-trien-3-ol, and (9) a compound 5,8-epidioxysterol with molecular formula C27H42O4. This fraction showed antileishmanial activity against amastigotes of the Leishmania (Viannia) panamensis parasite. Keywords: Antileishmanial activity, epidioxysterols, Ircinia campana, marine sponges

Introduction
Marine organisms are source of substances with potential use for the pharmaceutical, cosmetic and food industries (Faulkner 2002). Many studies related with the chemistry and biology of the substances produced by this organisms have been conducted as it has been found they present various and interesting biological activities. Marine bioactive compounds have diverse structural nuclei and some of them have steroid nature. Epidioxysterols have shown different biological activities such as: anti-tumor (Gauvin et al. 2000, Sheu et al. 2000, Petrichtcheva et al. 2001, Iwashima et al. 2002), repellent activity against marine mollusks (Sera et al. 1999), antileishmanial (Martnez et al. 2001), antitubercular (Saludes et al. 2002), antimicrobial (Abourriche et al. 2000), some plant germination stimulating effect (Macas et al. 1997), TPA inflammation inhibition (Yasukawa et al. 1996), anticomplement activity over the classic pathway and antiviral activity against the influenza virus (Casteel 1999). This work shows the results of the chemical and antileishmanial study on the activity of the fraction 5, 8epidioxysterols of the Colombian marine sponge Ircinia campana (Lamarck, 1814) (class Demospongiae, order Dictyoceratida, family Irciniidae) as this fraction showed activity against promastigotes of the parasite Leishmania panamensis on a previous study (Martnez et al. 2001).

Materials and methods

The methanol, chloroform, dichloromethane and ethyl acetate solvents used on the extraction and oxidation process were of pure reactive quality. The infrared analyses were made on a Fourier transform infrared spectrometer (FTIR Spectrum I), Perkin Elmer. The sample was read on a zinc selenide cell. The nuclear magnetic resonance analyses were made on a Bruker AMX 300 (300 MHz) nuclear magnetic resonance equipment using pre deuterated chloroform as solvent. The GC/MS analyses were made on an Agilent 5973 equipment with a gas chromatographer series 6890, which consists on an automatic injector. One l of sample diluted with dichloromethane was injected on mode splitless using helium as grabbing gas at 0.9 ml/min. The injector temperature was 270C. A HP5-MS column (30 m x 0.25 mm; 0.25 ) was used with a furnace programming from 200C until 290C at a rate of 5C/min. The mass range was from 40 until 500 Daltons. The compound separation was made by an Agilent 1100 system with ultraviolet detection (monitored at 210 nm), on a reverse phase column RP-18 Merck LiChrospher 100 (10 m x 250 mm; 5 ) using acetonitrile:methanol:water (8:1:1) as mobile phase and a flow rate of 1 ml/min. All thin layer chromatography analyses were made with chromatoplates of Silica gel 60 F254 (DC-Alufolien Kieselgel 60 F254, 0.2 mm). An UV 254/366 nm (115 volts, 60Hz and 0.16 amperes) model UVGL-58 MINERALIGHT light multi-band lamp and a phosphomolybdic acid solution (Merck) at 5% in

434

ethanol were used as developers. The column chromatography analyses were made using Silica gel 40.

Animal material
The sponge Ircinia campana was collected and identified by professor Sven Zea in March and November 2001, in Punta Betn, Santa Marta at 6 to 12 m depth. A voucher of the specimen was deposited under the code INV-POR 0022 in the sponge collection at INVEMAR, Santa Marta, Colombia. The samples of the sponge Ircinia campana were kept at 4C in a freezer immediately after the collection and until their later extraction.

The plates were incubated for an additional 30 min under agitation at room temperature. Production of formazan was measured at an optical density of 570 nm using an ELISA plate reader (Bio-Rad Laboratories, Hercules, California). Cells cultivated in absence of treatment but maintained under the same conditions were used as control. Three independent experiments in triplicate were performed for the determination of toxicity of each extract or fraction. Results were expressed as LD50 calculated by Probit (Finney 1982).

In vitro assay for antileishmanial activity

Effect of extracts and isolated fractions was evaluated on intracellular amastigotes as described in Robledo et al. (1999). Antileishmanial activity was measured on intracellular amastigotes from Leishmania (Viannia) panamensis (M/ HOM//87/UA140 strain), a strain isolated at the Programa de Estudio y Control de Enfermedades Tropicales PECET, Universidad de Antioquia, from one patient having localized cutaneous Leishmaniasis and criopreserved in liquid nitrogen until use. To maintain the virulence of parasite and therefore to obtain a good in vitro infection, parasites were maintained by passage in golden hamsters (Messocrycetus aureatus) (Rey et al. 1990). Periodically, lesions were aspirated and the material obtained was cultivated in NNN medium (Novy, Nicolle and McNeal) until stationary phase growth of promastigotes. After 48h of growth, the U-937 cells were washed twice with Dulbeccos phosphate buffer saline (DPBS) (Gibco BRL). One hundred thousand cells were added to each well of a 24-well plate (Falcon) containing sterile round 12-mm glass cover slip (Fisher Scientific, Pittsburg, Pennsylvania) and were exposed to stationary phase growth promastigotes at a ratio of 25:1 parasite:cell. Infected cell cultures were incubated for 2h with 5% CO2 at 34oC. Free parasites were removed by washing twice with warm DPBS. After 24h of incubation with 5% CO2 at 34oC in RPMI 1640 medium containing 10% FCS, the medium was replaced with complete RPMI 1640 medium containing the corresponding concentration of extract or fraction. The range of concentration varied between 0.1 and 10 mg/ml, depending of the LD50 for each extract or fraction. Thereafter, the medium was renewed every 2 days. After 96h of incubation in the presence of extract or fraction, cells were washed, fixed with methanol (Fisher) for 20 min and stained with Giemsa (Sigma-Aldrich Chemical Co, St Louis, MO). Similarly, infected cells cultured in absence of extract or fraction served as control of infection. Three independent experiments in triplicate were performed for the determination of leishmanicidal activity of each extract or fraction. All assays were evaluated blindly. For each test 200 cells/well were examined at random; the numbers of infected and uninfected cells were recorded. Percentage of infection was calculated by dividing the number of infected cells obtained in presence of each extract or fraction by the number of infected cells obtained in absence of treatment. Results were expressed as 50% Effective Dose (ED50) which was calculated by Probit analysis (Finney 1982).

Extraction and fractioning

The frozen sponge was cut into small pieces, and then it was lyophilized and ground for its extraction. The sponge (255.94 g) was exhaustively extracted with methanol at room temperature. The methanol extract (6.36 g) was dried at a 45C temperature with a constant rotation and reduced pressure. The dried residuum was extracted again with ethyl acetate and the sterol fraction (1526.3 mg) was separated and purified with TLC and CC on Silica gel F254 and Silica gel 40 respectively, using n-hexane : ethyl acetate (2:1) as mobile phase. The purified sterol fraction 5,7 was dissolved with CHCl3 and placed into an open tide with constant agitation and direct light from an halogen lamp (50 W, 120 V), Philips Master N-Flood 30, during 24h. The advance of the reaction was verified using the same conditions for the TLC and was compared with a sample of non oxidized sterols. Later, the oxidized sample was fractioned with CC and the solvent system CCl3 -methanol (95:5, 80:20 y 60:40) was used. Two fractions (A and B) were obtained and analyzed by HPLC, 1 H-NMR and GC/MS.

Cytotoxic activity
Cytotoxicity of Caribbean sponge was evaluated on the human promonocytic U-937 cell line (Sundstrom and Nilsson 1976). To estimate 50% lethal doses (LD50), the 3(4,5-dimethyithiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) enzymatic micromethod was used (Sereno and Lemesre 1997). Briefly, the U-937 cells were cultured in suspension in complete RPMI 1640 medium (Gibco BRL, Grand Island, New York) containing 10% heat-inactivated fetal calf serum (FCS) in the presence of 5% CO2 at 37oC. Medium was renewed at 2-day intervals. Cells were harvested and washed by centrifuging for 10 min at 400 g, then counted and adjusted to a final concentration of 1 x 106 cell/ml. One hundred l were seeded in 96-well-flat-bottom micro plates (Falcon, Becton Dickinson, Franklin Lakes, New Jersey). One hundred ml of corresponding fraction were added at a concentration ranged from 3 to 100 mg/ml. Cells were then incubated at 37oC with 5% CO2. After 48h of incubation medium was changed and cells were incubated again in presence of the same concentrations of extract or fraction. After 96h of incubation, 10 ml of MTT (10 mg/ml) was added to each well. Plates were further incubated for 3h. The enzymatic reaction was then stopped by addition of 100 ml of 50% isopropanol 10% sodium dodecyl sulfate solution.

435

Results
Spectrum 1H-NMR analysis of the oxidized fractions showed characteristic signals of compounds with the nuclei of epidioxysterols. Methylic proton signals between 0.6 and 1.0 ppm were observed, a multiplet at 3.9 ppm which corresponds to C3 proton, a doublet signal at 6.24 ppm (d, J=8.5 Hz) which corresponds to the olefinic H-7 proton, and a doublet signal at 6.50 ppm (d, J=8.5 Hz) corresponding to olefinic H-6 proton. In further analysis of the fractions with GC/MS, a mixture of compounds was observed and the following were identified: 5,8-epidioxy-24-norcholesta6,22-dien-3-ol (1), 5,8-epidioxy-cholesta-6,22-dien3-ol (2), 5,8-epidioxi-(24)-methylcholesta-6-en-3-ol (epimer 1) (3), 5,8-epidioxy-cholesta-6-en-3-ol (4), 5,8-epidioxy-(24)-methylcholesta-6-en-3-ol (epimer 2) (5), 5,8-epidioxy-(24)-ethylcholesta-6,22-dien-3-ol (6), 5,8-epidioxy-cholesta-6,9-dien-3-ol (7), 5,8epidioxy-cholesta-6,9,22-trien-3-ol (8) and a compound 5,8-epidioxysterol with molecular formula C27H42O4 (9) (Fig. 1).

On the other hand, results of the antileishmanial activity show both fractions A and B have good performance against amastigotes of the parasite Leishmania panamensis. In Table 1 is shown that CE50 values obtained by both fractions of epidioxysterols are higher than the positive control (Glucantime) values.

Discussion
The results show the epidioxysterols identified in the sponge Ircinia campana are oxidation appliances of the sterols 5,7 natives of the sponge and not new compounds made by the animal. This fact was suggested by Martnez (1996), who explained the oxidation reaction for the epidioxysterols of the sponge Ircinia campana when he accidentaly observed the presence of these compounds during the analysis of a sponge sample (Fig. 2). Aditionally, results suggest epidioxysterols reported by Caldern (1982) for the marine sponges Ircinia campana and Ircinia fasciculata were not naturaly synthetized products by these sponges but appliances of the photochemical oxidation of their native sterols. From the epidioxysterols identified in this work, compounds 1-8 are

Fig. 1: Structures of 5, 8epidioxysterols from the marine sponge Ircinia campana.

436

Table 1: In vitro cytotoxicity and antileishmanial activity of fractions A and B against amastigotes of Leishmania V. panamensis. Compound Fraction A Fraction B Glucantime (control) CL 50 (g / ml) 25.5 39.7 416.4 CE 50 (g / ml) 30 2.8 25.7 7.4 6.7 IS 1.2 0.6 62.1

Fig. 2: Oxidation reaction of native sterols-5,7 from Ircinia campana sponge.

known. It is suspected that the compound 9 could be new as, opposite to the other previously reported compounds in literature, this one has four atoms of oxigen. On the other hand, only the compounds (2) and (4) have been previously reported in the sponge Ircinia campana (Calderon et al. 1982), which means this is the first report on the presence of the compounds (1, 3 and 5-9) in this sponge. Fractions A and B of the 5, 8- epidioxysterols showed good antileishmanial activity but were toxic as well. It would be interesting to compare the toxicity of the individual components with the toxicity shown by their mixture, aiming to stablish if this is a synergistic effect and stablish the individual antileishmanial strength of each compound. Besides, from the biological activity point of view, the citotoxicity shown by both fractions is interesting if it is wished to continue with studies related with compounds against cancer. Finally, this work shows an easy, fast and economic methodology for converting compound without biological activity such as sterols with nuclei 5,7-3-hidroxyandrostadiene, into bio-active compounds such as 5, 8-epidioxyesterols.

Compounds isolated from Ircinia campana

5,8-epidioxy-24-norcholesta-6,22-dien-3-ol (1). RMN1 H (CDCl3, 300MHz): 0.6-1.0 (3H, s, C-18, C-19, C-21, C-25, C-26), 3.90 (1H, m, H-3), 6.24 (1H, d, J=8.5 Hz, H-7), 6.50 (1H, d, J=8.5 Hz, H-6). EMIE m/z 400 [M], 368, 353, 349. Retention time by GC, 15.08 min. These data are consistent with a epidioxysterol of molecular formula C26H40O3. 5,8-epidioxy-cholesta-6,22-dien-3-ol (2). RMN-1H (CDCl3, 300MHz): 0.6-1.0 (3H, s, C-18, C-19, C-21, C26, C-27), 3.90 (1H, m, H-3), 6.24 (1H, d, J=8.5 Hz, H7), 6.50 (1H, d, J=8.5 Hz, H-6). EMIE m/z 414 [M] , 382, 396, 364, 349. Retention time by GC, 16.21 min. These data

are consistent with a epidioxysterol of molecular formula C27H42O3. 5,8-epidioxy-(24)-methylcholesta-6-en-3-ol (3) Epimer 1. RMN-1H (CDCl3, 300MHz): 0.6-1.0 (3H, s, C18, C-19, C-21, C-26, C-27, C-28), 3.90 (1H, m, H-3), 6.24 (1H, d, J=8.5 Hz, H-7), 6.50 (1H, d, J=8.5 Hz, H-6). EMIE m/z 430 [M] 412, 398, 380, 365, 353. Retention time by GC, 17.90 min. These data are consistent with a epidioxysterol of molecular formula C28H46O3. 5,8-epidioxy-cholesta-6-en-3-ol (4). RMN-1H (CDCl3, 300MHz): 0.6-1.0 (3H, s, C-18, C-19, C-21, C-26, C-27), 3.90 (1H, m, H-3), 6.24 (1H, d, J=8.5 Hz, H-7), 6.50 (1H, d, J=8.5 Hz, H-6). EMIE m/z 416 [M], 412, 398, 366, 353, 351. Retention time by GC, 19.07 min. These data are consistent with a epidioxysterol of molecular formula C27H44O3. 5,8-epidioxy-(24)-methylcholesta-6-en-3-ol (5) Epimer 2. RMN-1H (CDCl3, 300MHz): 0.6-1.0 (3H, s, C-18, C-19, C-21, C-26, C-27, C-28), 3.90 (1H, m, H-3), 6.24 (1H, d, J=8.5 Hz, H-7), 6.50 (1H, d, J=8.5 Hz, H6). EMIE m/z 430 [M], 412, 398, 380, 365, 353. Retention time by GC, 19.40 min. These data are consistent with a epidioxysterol of molecular formula C28H46O3. 5,8-epidioxy-(24)-ethylcholesta-6,22-dien-3-ol (6). RMN-1H (CDCl3, 300MHz): 0.6-1.0 (3H, s, C-18, C-19, C-21, C-26, C-27, C-29), 3.90 (1H, m, H-3), 6.24 (1H, d, J=8.5 Hz, H-7), 6.50 (1H, d, J=8.5 Hz, H-6). EMIE m/ z 442 [M], 424, 410, 392, 377, 253. Retention time by GC, 20.64 min. These data are consistent with a epidioxysterol of molecular formula C29H46O3. 5,8-epidioxy-cholesta-6,9(11)-dien-3-ol (7). RMN-1H (CDCl3, 300MHz): 0.6-1.0 (3H, s, C-18, C-19, C-21, C-26, C-27), 3.90 (1H, m, H-3), 6.24 (1H, d, J=8.5 Hz, H-7), 6.50 (1H, d, J=8.5 Hz, H-6). EMIE m/z 414 [M], 396, 382, 364, 349, 353. Retention time by GC, 21.44 min. These data are consistent with a epidioxysterol of molecular formula C27H42O3. Compound 8. RMN-1H (CDCl3, 300MHz): 0.6-1.0 (3H, s, C-18, C-19, C-21, C-26, C-27), 3.90 (1H, m, H-3), 6.24 (1H, d, J=8.5 Hz, H-7), 6.50 (1H, d, J=8.5 Hz, H-6). EMIE m/z 430 [M], 412, 398, 365, 253. Retention time by GC, 21.79 min. These data are consistent with a epidioxysterol of molecular formula C27H42O4. 5,8-epidioxy-cholesta-6,9(11),22-trien-3-ol (9). RMN1 H (CDCl3, 300MHz): 0.6-1.0 (3H, s, C-18, C-19, C-21, C-26, C-27), 3.90 (1H, m, H-3), 6.24 (1H, d, J=8.5 Hz, H-7), 6.50 (1H, d, J=8.5 Hz, H-6). EMIE m/z 412 [M], 394, 380, 365, 253. Retention time by GC, 23.45 min. These data are consistent with a epidioxysterol of molecular formula C27H40O3.

Acknowledgments
The authors wish to thank professor Sven Zea from the Universidad Nacional de Colombia and member of INVEMAR for collecting and identifying the sponge, and the Comit para el Desarrollo de la Investigacin (CODI) of the Universidad de Antioquia for sponsoring this research.

437

References
Abourriche A, Charrouf M, Chaib N, Bennamara A, Bontemps N, Francisco C (2000) Isolation and bioactivities of epidioxysterol from the tunicate Cynthia savignyi. Il Farmaco 55(6-7): 492-494 Caldern JS, Quijano L, Gmez F, Acosta F, Ros T (1982) Estudios preliminares sobre algunas esponjas del litoral Mexicano. I. Rev Latinoam Qum 13: 49-52 Casteel DA (1999) Peroxy natural products. Nat Prod Rep 16: 5573 Faulkner DJ (2002) Marine natural products. Nat Prod Rep 19: 148 Finney DJ (1982) The questioning statistician. Stat Med 1: 5-13 Gauvin A, Smadja J, Aknin M, Faure R, Gaydou E-M (2000) Isolation of bioactive 5, 8 epidioxy sterols from the marine sponge Luffariella cf. variabilis. Can J Chem 78(7): 986-992 Iwashima M, Terada I, Iguchi K, Yamori T (2002) New biologically active marine sesquiterpenoid and steroid from the Okinawan sponge of the genus Axinyssa. Chem Pharm Bull 50(9): 12861289 Macas FA, Simonet AM Galindo JCG (1997) Bioactive steroids and triterpenes from Melilotus messanensis and their allelopathic potential. J Chem Ecol 2: 1781-1803 Martnez MA, Robledo RSM, Muoz HDL, Blair TS, Higuita DE, Echeverri PE, Bedoya JJ, Zea S, Vlez BID (2001) Antiparasitic activity of methanol extracts and isolated fractions from Caribbean sponges. Vitae 8(1,2): 71-81 Martnez A (1996) Estudio qumico y de actividad biolgica de esponjas marinas Colombianas del gnero Ircinia. PhD Thesis. Universidad Nacional de Colombia, Santa Fe de Bogot

Petrichtcheva NV, Duque C, Fujimoto Y (2001) Ambrosinosterol: un nuevo 5,8-epidioxiesterol citotxico aislado de la esponja marina Axinyssa ambrosia. Rev Acad Colomb Cienc 25(97): 569577 Rey-Ladino JA, Travi BL, Valencia AZ, Saravia NG (1990) Infectivity of the subspecies of the Leishmania braziliensis complex in vivo and in vitro. Am J Trop Med Hygiene 43: 623-631 Robledo RSM, Valencia AZ, Saravia NG (1999) Sensitivity to glucantime of Leishmania viannia isolated from patients prior to treatment. J Parasitol 85: 360-366 Saludes JP, Garson MJ, Franzblau SG, Aguinaldo AM (2002) Antitubercular constituents from de hexane fraction of Morinda citrifolia Linn. (Rubiaceae). Phytoter Res16: 683-685 Sera Y, Adachi K, Shizuri Y (1999) A new epidioxy sterol as an antifouling substance from a palauan marine sponge, Lendenfeldia chondrodes. J Nat Prod 62(1): 152-154 Sereno D, Lemestre J (1997) Use of an enzymatic micromethod to quantify amastigotes stages of Leishmania amazonensis in vitro. Parasitol Res 83: 401-403 Sheu JH, Chang KC, Duh CY (2000) A cytotoxic 5alpha,8alphaepidioxysterol from a soft coral Sinularia species. J Nat Prod 63(1): 149-51 Sundstrom C, Nilsson K (1976) Establishment and characterization of a human histiocytic lymphoma cell line (U-937). Int J Canc 17: 565-577 Yasukawa K, Akihisa T, Kanno H, Kaminaga T, Izumida M, Sakoh T, Tamura T, Takido M (1996) Inhibitory effects of sterols isolated from Chlorella vulgaris on 12-0-tetradecanoylphorbol-13-acetate induced inflammation and tumor promotion in mouse skin. Biol Pharm Bull 19(4): 573-576

Porifera research: Biodiversity, innovation and sustainaBility - 2007

439

Foraminifera associated to the sponge Mycale microsigmatosa in Rio de Janeiro State, southeastern Brazil - an initial approach
Mirna Mazzoli-Dias(1), Suzi M. Ribeiro(2*), Patricia Oliveira-Silva(3)
Faculdades Integradas Maria Thereza, Niteri - Rio de Janeiro, Brazil. Programa de Ps-graduao em Biologia Marinha - Universidade Federal Fluminense, Niteri, Rio de Janeiro, Brazil. suzimr@yahoo.com.br (3) Departamento de Geoqumica Ambiental - Universidade Federal Fluminense, Niteri, Rio de Janeiro, Brazil. patriciaols@geoq.uff.br
(1) (2)

Abstract: Ecological relationships (e.g. mutualism, commensalism and parasitism) between sponges and polychaetes, crustaceans, algae, protozoans, and other marine organisms are common. However, there are few studies on the ecological relations between sponges and foraminifera. The main goal of this study was to relate the occurrence of benthic foraminifera associated to ten individuals of the sponge Mycale (Carmia) microsigmatosa Arndt, 1927 collected at Itaipu Beach, Rio de Janeiro State, Brazil. A total of 26 foraminifera species were found living inside M. (C.) microsigmatosa. Among them, Lobatula aff. lobatula, Rosalina globularis, and Rosalina floridana were the most frequent species found. The latter was the most representative species in abundance and frequency. Foraminifera species inside the sponges were found in various ontogenetic stages, suggesting the occurrence of commensalism as they use this microhabitat for nutrition and reproduction purposes. Keywords: associated fauna, ecology, foraminifera, Mycale, SW Atlantic

Introduction
Porifera is one of the most abundant invertebrate groups on hard substrates (Sar and Vacelet 1973, Vacelet 1979, Hooper and Lvi 1994, van Soest 1994). The presence of inner channels, a continuous water flux and a soft body allow the occurrence of a great diversity of marine organisms inside them (crustacean, mollusks, echinoderms, protists, etc.; Cuartas and Excoffon 1993, Duarte and Nalesso 1996, Ribeiro et al. 2003, Neves and Omena 2004, Gaino et al. 2004). Foraminifera are marine protists which use pseudopods for nutrition, reproduction and locomotion (Boersma 1998) and are also a food resource for benthic macroinvertebrates (Murray 1991, Hannah and Rogerson 1997). Foraminifera could have some advantages in living within sponges, for example, obtaining facilities to get food through inhalant currents and life protection against predators, although sponges seem to have at first sight no advantages in this relationship. Foraminiferans are also found associated to red algae (Boltovskoy et al. 1976), diatoms, bryozoans, gastropods and nematods in different kinds of relationships such as epibiosis, epizoism, commensalism, parasitism, and symbiosis (Hallock 1984, Murray 1991, Hallock et al. 2003).

The presence of foraminifera living inside sponges is common in Brazil. However, it is a poorly documented and studied relationship. Interactions between sponges and foraminifera were observed even in deep sea samples (Guilbault et al. 2006). Associations of recent foraminifera with sponges still remain misunderstood as most studies deal with this purpose in the past geological time, which suggests an ancient relationship between these groups (Bromley and Heinberg 2006). Recently, some works have been published dealing with the possible parallel evolution of single-celled eukaryotes as foraminifera and metazoans, which may indicate similarities of evolutionary response to environmental changes (Lipps 2006). The main goal of this study was to report the occurrence of benthic foraminifera associated with the shallow-water sponge Mycale (Carmia) microsigmatosa Arndt, 1927 in Rio de Janeiro State, Brazil.

Material and methods Sampling

Ten individuals of M. (C.) microsigmatosa (labeled from E1 to E10) were collected by free diving in Itaip rocky shores, Niteri City (Rio de Janeiro State, Brazil; Fig. 1) in

440

January 2004. According to Salvador and Silva (2002), the Itaip soft bottom is covered by quartz-rich sand, followed by pebbles of biological detritus, and silt and clay fractions may occur only locally. During collection, the sponges were covered with a plastic bag to avoid the escape of foraminifera. The bagged seawater was filtered in a sieve with a mesh size of 0.062 mm, and the sponges with their associated fauna were stained with Rose Bengal, in order to colour living foraminifera individuals, and preserved in 70% ethanol.

Laboratorial procedures
In the laboratory, the sponges were macerated carefully under a stereomicroscope to remove the associated foraminifera which were identified until the specific level using specialized literature (Boltovskoy et al. 1980, Loeblich and Tappan 1988, Hottinger et al. 1993). The number of individuals belonging to each foraminifera species were recorded for each collected sponge.

Results
A total of 26 foraminifera species, representing one arenaceous and 25 calcareous taxa, were found living inside Mycale (Carmia) microsigmatosa (Table 1), including sessile (Cibicides spp), clinging (Rosalina spp.) and free species (Quinqueloculina spp., Triloculina spp.) according to Murray (1991). The number of individuals varied from 21 to 332 in the samples E1 and E4, respectively. The maximum value to species richness was 13 in E4 and the minimum was 5 in the samples E3, E8, E10. It was observed a great abundance of Lobatula aff. lobatula, Rosalina floridana and Rosalina globularis in all sponge samples (Table 1). On the other hand, there were few individuals of Textularia gramen, Triloculina sp., Bulimina sp., what can suggest that their lifestyles do not adapt to the interior of the sponge body, even though they are abundant in the sediments in Itaip (data not shown).

Fig. 1: Guanabara Bay. Arrow A indicates Itaip Beach, Niteri, Rio de Janeiro State (Brazil).

Discussion
The occurrence of one arenaceous and 25 calcareous taxa found living in our samples against 40 and 53 respectively found dead and living on modern sponges on the continental shelf off British Columbia (Guilbault et al. 2006) may be correlated to the sampling design, as our study dealt with samples collected near the shore and the latter was done with samples deeper than 100 meters, which presents a slightly different foraminiferal fauna. Foraminifera species inside the sponges were found in various ontogenetic stages, suggesting the occurrence of commensalism as they seem to use this microhabitat for nutrition and reproductive purposes. According to Alexander and DeLaca (1987), Cibicides refulgens living epifaunally on scallops can use direct interception of particles by the stick pseudopodia, which would form a large part of the filtering process. Lutze and Thiel (1989) corroborate this idea by saying that Cibicidoides wuellerstorfi and Planulina

ariminensis prefer an elevated position above the sedimentwater interface for a better chance to catch food particles from slightly streaming water, a behavior which seems to be similar to those forams in Brazil living inside sponges under turbulent rocky-shore conditions. Besides that, Lobatula lobatula is a suspension feeder foraminifera (Murray 1991, Guilbault et al. 2006) found in great abundance on the samples studied here, which makes us believe that it is intercepting its food on the sponges inhalant currents. It seems that sponges may act as an optional ecological niche to living foraminifera since the species found inside of them are the same observed in 2004 in Itaip beach sediments (data not shown). Therefore, we conclude that foraminifera really live inside Mycale microsigmatosa at Itaip beach, as samples were free of sand grains indicating that they were not transported by currents to the sponge canal system. Consequently, they are not a result of postmortem colonization, since they were alive. In conclusion, we suggest that these are common sponge-dwelling foraminifera fauna which lives inside M. (C.) microsigmatosa in Brazil. This work is the first attempt to study the occurrence and relationship between modern foraminifera and sponges in South Atlantic tropical waters. Apart from the results found, it is clear that much more data is needed in order to understand and clarify this association in waters that are submitted to high sedimentation rates as noted in this region in Brazil.

441

Table 1: Species of Foraminifera associated to Mycale (C.) microsigmatosa. Species/Samples Bulimina sp. Cibicides lobatulus Cibicides refulgens Cibicides sp. Discorbis cf. bertheloti Discorbis sp. 1 Discorbis sp. 2 Discorbis mamilla Elphidium sp. Eponides sp. Lobatula aff. lobatula Miliolinella cf. lutea Nonion sp. Poroeponides lateralis Quinqueloculina cf. fusca Quinqueloculina elegans Q. seminulum Quinqueloculina sp. 1 Quinqueloculina sp. 2 Quinqueloculina sp. 3 Rosalina floridana Rosalina globularis Textularia gramen Triloculina oblonga Triloculina sp. Triloculina trigonula Total individuals Richness E1 0 0 1 2 3 0 0 0 0 0 2 0 0 0 0 0 0 0 0 1 10 2 0 0 0 0 21 7 E2 0 0 1 0 0 0 0 0 0 1 14 0 0 0 0 0 5 2 0 0 68 2 0 1 1 0 95 9 E3 0 1 0 0 0 0 0 0 0 0 6 0 0 0 0 0 0 1 0 0 49 2 0 0 0 0 59 5 E4 1 0 6 0 0 4 1 1 0 1 78 1 0 1 0 0 0 1 0 0 196 31 0 10 0 0 332 13 E5 1 0 1 0 0 0 0 0 0 0 12 1 0 1 0 0 0 1 0 0 58 14 1 1 0 1 92 11 E6 0 1 0 0 0 0 0 0 1 1 7 0 0 0 0 0 1 1 1 0 29 4 0 0 0 0 46 9 E7 0 2 1 0 0 0 0 0 0 1 17 0 0 0 1 1 0 2 0 0 50 6 1 7 0 1 90 12 E8 0 0 2 0 0 0 0 0 0 0 3 2 0 0 0 0 0 0 0 0 151 0 0 0 0 3 161 5 E9 0 1 5 0 0 1 0 0 0 0 10 4 0 0 0 0 9 1 1 3 57 3 0 9 0 0 104 12 E10 0 0 0 0 0 0 0 0 0 0 5 0 1 0 0 0 15 1 0 0 70 0 0 0 0 0 92 5

Acknowledgements
We thank Dr. Guilherme Muricy for reviewing the final work from which this study is part; Guilherme Martins for English revision; Cristiane Fiori and Gabriela Benkendorfer for helping on fieldwork.

References
Alexander SP, DeLaca TE (1987) Feeding adaptions of the foraminiferan Cibicides refulgens living epizonically and parasitically on the antarctic scallop Adamussium colbecke. Biol Bull 173: 136-159 Boersma A (1998) Foraminifera. In: Haq BU, Boersma A (eds). Introduction to marine micropaleontology. Elsevier, Amsterdam. pp. 19-77 Boltovskoy E, Giussani G, Watanabe S, Wright R (eds) (1980) Atlas of benthic shelf foraminifera of the Southwest Atlantic. W Junk Publishers, The Hage Boltovskoy E, Lena H, Asens A (1976) Algae as a substrate for foraminifera in the Puerto Deseado area (Patagonia). J Mar Biol Assoc India 18(1): 140-148 Bromley RG, Heinberg C (2006) Attachment strategies of organisms on hard substrates: a palaeontological view. Palaeogeogr Palaeoclimatol Palaeoecol 232: 429-453

Cuartas EI, Excoffon AC (1993) La fauna acompaante de Hymeniacidon sanguinea (Grant,1827) (Porifera: Demospongiae). Neotropica 39: 3-10 Duarte LFL, Nalesso RC (1996) The sponge Zygomycale parishii (Bowerbank) and its endobiotic fauna. Estuar Coast Shelf Sci 42: 139-151 Gaino E, Lancioni T, La Porta G, Todini B (2004) The consortium of the sponge Ephydatia fluviatilis (L.) living on the common reed Phragmites australis in Lake Piediluco (central Italy). Hydrobiologia 520(1-3): 165-178 Guilbault J, Krautter M, Conway KW, Barrie JV (2006) Modern foraminifera attached to Hexactinellid sponge meshwork on the West Canadian Shelf: comparison with Jurassic counterparts from Europe. Palaeontol Electr 9(1): 3a (http://palaeo-electronica.org; accessed in october 19, 2006) Hallock P (1984) Distribution of selected species of living algal symbiont-bearing foraminifera on two pacific coral reefs. J Foram Res 14(4): 250-261 Hallock P, Lidz BH, Cockey-Burkhard EM, Donnely KB (2003) Foraminifera as bioindicators in coral reef assessment and monitoring: the foram index. Environ Monit Assess 81: 221-238 Hannah F, Rogerson A (1997) The temporal and spatial distribution of foraminiferans in marine benthic sediments of the clyde sea area, Scotland. Estuar Coast Shelf Sci 44: 377-383

442

Hooper JNA, Lvi C (1994) Biogeography of Indo-west Pacific sponges: Microcionidae, Raspailiidae, Axinellidae. In: van Soest RWM, van Kempen TMG , Braekman JC (eds). Sponges in time and space. Balkema, Rotterdam. pp.191-212 Hottinger L, Halicz E, Reiss Z (eds) (1993) Recent Foraminiferida from the Gulf of Aqaba, Red Sea. Slovenska Akadenija Znamosti in Umetnosti, Ljublyana Langer MR (1993) Epiphytic foraminifera. Mar Micropaleontol 20: 235-265 Lipps JH (2006) Major features of protistan evolution: controversies, problems and a few answers. An Inst Geoc UFRJ 29(1): 55-80 Loeblich AR, Tappan H (eds) (1988) Foraminiferal genera and their classification. Van Nostrand Reinhold, New York Lutze GF, Thiel H (1989) Epibenthic foraminifera from elevated microhabitats: Cibicidoides wuellerstorfi and Planulina ariminensis. J Foram Res 19: 153-158 Murray JW (1991) Ecology and paleoecology of benthic Foraminifera. John Wiley and Sons, New York

Neves G, Omena EP (2003) Influence of sponge morphology on the composition of the polychaete associated fauna from Rocas Atoll, northeast Brazil. Coral Reefs 22: 123-129 Ribeiro SM, Omena EP, Muricy G (2003) Macrofauna associated to Mycale microsigmatosa (Porifera, Demospongiae) in Rio de Janeiro State, SE Brazil. Estuar Coast Shelf Sci 57: 1-9 Salvador SVM, Silva MAM (2002) Morphology and sedimentology of the Itaip Embayment Niteri/RJ. An Acad Bras Cinc 74(1): 127-134 Sar M, Vacelet J (1973) cologie des dmosponges. In: Grass, PP (ed). Trait de Zoologie, III - Spongiaires. Masson et Cie, Paris. pp. 462-576 van Soest RWM (1994) Demosponge distribution patterns. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space. Balkema, Rotterdam. pp. 213-223 Vacelet J (1979) La place des spongiaires dans les systmes trophiques marins. In: Lvi C, Boury-Esnault N (eds). Biologie des spongiaires. ditions du CNRS, Paris. pp. 259-270

Porifera research: Biodiversity, innovation and sustainaBility - 2007

443

Associations and interactions between gorgonians and sponges

Elizabeth L. McLean(*), Paul M. Yoshioka
Department of Marine Sciences at the University of Puerto Rico Mayagez , PO Box 9013, Lajas, PR 00667-3264. elmclean@sbcglobal.net, p_yoshioka@cima.uprm.edu Abstract: The biodiversity of sessile marine invertebrates in coral reefs may be highly dependent upon associations and interactions between species. In this study we describe the associations (i.e. physical contact between organisms) and subsequent interactions (e.g. competitive overgrowth, mutualistic effects) between sponges and gorgonians in La Parguera, southwest coast of Puerto Rico. Nearly half (15/32) of the sponge species were epibiotic on other sessile organisms. The number of gorgonian colonies associated with sponges was positively correlated with the relative abundances of the gorgonian taxa indicating that associations with sponges are generally random in nature. For instance, Pseudopterogorgia spp. was the most abundant gorgonian taxa and was the most often associated with sponges. Ecological interactions often involved overgrowth of gorgonians including Pseudopterogorgia americana, Gorgonia ventalina and Pseudoplexaura spp. by sponges such as Desmapsamma anchorata, Iotrochota birotulata, Dysidea janiae and Monanchora barbadensis. Of the 14 most abundant gorgonian taxa, three were only mildly affected by sponges, and tissues of the remaining were killed by overgrowth. In contrast to negative effects, field observations suggest that Briareum asbestinum has mutualistic interactions with sponges. These results indicate that epibiotic associations may be a major factor for the high biodiversity of sponges in Caribbean reefs. As indicated by their relationships with gorgonians, these associations are largely random. However, interactions with the associated organisms appear to be species-specific. Keywords: biodiversity, coral reef, gorgonians, sponges, species associations

Introduction
Space on marine hard substratum is a limiting resource for numerous sessile organisms in coral reefs (Jackson 1977, Connell 1978) and is a major factor affecting biodiversity in this habitat. Organisms growing in close proximity will often compete for suitable substrate by overgrowing, crowding, undercutting, digesting, overshadowing, or poisoning each other (Bakus et al. 1986). These interactions can affect the organization and function of reef communities (Buss and Jackson 1979, Wulff 2005). Stebbing (1973) and Wahle (1980) observed that competitive interactions that occur among sessile organisms are often intense because of their inability to change locations. These competitive interactions may be influenced by many factors including habitat conditions as well as by competitive hierarchies or networks (Buss and Jackson 1979). In addition to negative (competitive) effects, interactions with scleractinian corals may have beneficial (mutualistic) effects by increasing coral survival (Goreau and Hartman 1963, Wulff and Buss 1979). Sponges may also enhance biodiversity by harbouring numerous invertebrate species (Neves and Omena 2003, Ribeiro et al. 2003). Sponges also clear up the water column, increase primary productivity by recycling nutrients and serve as food for other organisms (Reiswig 1971, Diaz and Rtzler 2001, Wulff 2001). In this study we describe the associations and interactions of sponges with other sessile organisms with special attention

to the gorgonians. To avoid misunderstanding, it should be noted that associations and interactions are not synonymous. Associations refer to the physical contact between sponges and gorgonians; whereas interactions refer to the post-contact ecological outcome of the association. Interactions often involves overgrowth, defined here as the elevation of the growing edge of one individual over the edge of another one (Stebbing 1973). Overgrowth generally involves the death of the overgrown tissue although recovery can occur in some instances (Buss and Jackson 1979, Russ 1981). Specific overgrowth mechanisms include differential growth rates (Buss and Jackson 1979, see also Lang 1973) and allelochemicals (Buss and Jackson 1979). Jackson and Buss (1975), Suchanek et al. (1983) and Aerts and van Soest (1996) suggest that these mechanisms are generally welldeveloped in sponges enabling them to outcompete corals for space. Other interactions include standoff situations (Aerts 2000), where growth of both organisms ceases at the contact boundary and intermingled growth which may benefit both species by increasing structural support and predator defenses (eg., Wulff 1997).

Materail and methods

The study site is located at a depth of 6.7 m off of Media Luna reef (17 56.2 N, 67 3.2W) in La Parguera, southwest coast of Puerto Rico (see Yoshioka and Yoshioka 1989).

444

This site has no emergent reefs, is characterized by a low topographic relief, and is exposed to wave action generated by trade winds. Gorgonians are the visually dominant sessile macroinvertebrate taxa at this study site. Sponges and their associations with gorgonians were surveyed during December 2005 February 2006 in a 1 x 20 m belt transect. The nature of interaction between gorgonian and sponges (eg, overgrowth of the gorgonian) was inferred from five monitoring surveys between October 2003 January 2005 at approximately four month intervals. Gorgonians in this transect have been monitored since 1983 (Yoshioka and Yoshioka 1989, Yoshioka 1998, 2005). To determine whether associations were positive, random or negative we compared the observed frequencies of sponge gorgonian associations to expected values based on the relative abundances of gorgonian species. For instance, the association would be positive if the observed frequency of a gorgonian species occurring with sponges is greater than expected value based on its relative abundance. Photographs were also taken for reference purposes.

Results
A total of 32 demosponge species were found in the study area. Of these, 18 (56%) were growing on other sessile organisms (Table 1). The vast majority (15/18) of these sponges were associated with gorgonians, probably because gorgonians constituted the greatest proportion of biotic substrate (at least in terms of their visual dominance) at the study site. Sponges were associated with 4.1% (71/1730) of the gorgonian colonies. Desmapsamma anchorata, Dysidea janiae, and Iotrochota birotulata were the sponges most commonly found on gorgonians. The associations between gorgonian and sponge species are given in Table 2. Gorgonians most commonly associated with sponges were also the most abundant gorgonian taxa in the transesct: Pseudopterogorgia spp., Pseudoplexaura spp., and Gorgonia ventalina. In terms of quantitative relationships, the proportions of gorgonian colonies associated with sponges were significantly correlated (r2 = 0.899, p < 0.01) with the relative abundances of the gorgonian taxa (Fig. 1). The 95% CL of the regression slope is not significantly different from 1.00 ( 0.72 1.02) indicating that proportions of sponge associated gorgonian

Table 1: Sponge species and their substrates in the study area. Abiotic substrates include rock, rubble and sediment. Three additional unidentified sponges (likely pertaining to order Dendroceratida and Order Halisarcida) were found only on abiotic substrates. Species Amphimedon compressa Duch. and Mich., 1864 A. viridis Duch. and Mich., 1864 Cliona varians (Duch. and Mich., 1864) Aplysina spp. Nardo, 1834 Artemisina melana Vosmaer, 1885 Callyspongia vaginalis Lamarck, 1814 C. plicifera Lamarck, 1814 Chondrilla nucula Schmidt,1862 Cinachyra spp. Sollas, 1886 Cliona delitrix Pang, 1974 C. caribbaea (Langae) Carter, 1882 (Pang, 1973) Desmapsamma anchorata Carter, 1882 Dysidea janiae Duch. and Mich., 1864 Ectyoplasia ferox Duch. and Mich., 1864 Haliclona implexiformis Hechtel, 1966 Iotrochota birotulata Ridley, 1884 Ircinia strobilina Lamarck, 1816 Monanchora barbadensis Hechtel, 1969 M. unguifera van Soest, 1984 Mycale laevis Carter, 1882 M. carmigropila Hajdu and Rtzler, 1998 M. laxissima Duch. and Mich., 1864 Neofibularia notilangere Duch. and Mich., 1864 Niphates erecta Duch. and Mich., 1864 Niphates caycedoi (Zea and van Soest 1986) Plakortis spp. Schulze, 1880 Scopalina ruetzleri Wiedenmeyer, 1977 Verongula rigida Esper, 1794 Xestospongia proxima Duch. and Mich., 1864 Substrate Abiotic Epibiotic yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes yes no yes no yes no no no yes yes yes yes no no yes yes yes yes yes yes no no yes yes no yes no yes Comments On rocks, sediments, gorgonians and other sponges On rocks, sediments and gorgonians On rocks and rubble On rocks and gorgonians On sediments On rocks and gorgonians. Often with zoanthids and brittle stars on its surface On rocks. Often with zoanthids and brittle stars on its surface On sediments and coral rubble. Often binds coral rubble On sediments. Often covered with brown algae and sand Overgrows and bores dead/live corals. Often with zoanthids and brittle stars Overgrows and bores dead/live corals and hydrocorals On rocks, sediments, buoy lines, gorgonians, hydrocorals, zoanthids and other sponges. On gorgonians and very common on buoy lines On sediments, rocks and rubble On rocks and coral rubble On sediments, rocks and gorgonians. Often with zoanthids and brittle stars on its surface On sediments, rocks, rubble and gorgonians On coral rubble and gorgonians On rocks, rubble and gorgonians On numerous hard coral species On rocks and gorgonians On sediments, rocks and rubble On sediments and dead coral. Harbors numerous invertebrates On sediments, gorgonians and other sponges. Often with zoanthids and brittle stars On sediments, rocks, rubble and gorgonians On sediments, rocks and rubble On sediments, corals, gorgonians, hydrocorals and other sponges On sediments, rocks and rubble On sediments, rocks, rubble and gorgonians

445

Table 2: Associations of sponges and gorgonians. The values represent the number of observed incidences. See gorgonian taxa code given in Fig. 1. Gorgonian species Sponge species Amphimedon compressa Amphimedon erina Callyspongia vaginalis Chondrilla nucula Desmapsamma anchorata Dysidea janiae Iotrochota birotulata Ircinia strobilina Monanchora unguifera Mycale laevis Mycale laxisima Mycale carmigropila Niphates erecta Niphates caycedoi Scopalina ruetzleri Xestospongia proxima Total Br 1 Ecal Emam Esuc/ Elax 1 5 7 1 1 1 6 2 1 3 1 2 1 14 18 2 18 1 11 2 2 4 Gv Mur Mflv Pflx Ph/ Phk Psx 2 1 1 1 9 2 1 3 2 8 1 Pa 2 3 0 0 5 26 6 7 8 1 0 0 6 0 1 2 67 Pte 1 Plla Total 7 11 1 20 30 43 16 15 2 0 18 1 5 2 1 3 171 % 0.075 0.118 0.000 0.011 0.215 0.323 0.462 0.172 0.161 0.022 0.000 0.194 0.011 0.054 0.022

1 1 4 1 2 1 1 1

1 2 3 1 1 1 1 2 12 15

9 3

1 1

1 1

21

Fig. 1: Relative abundances (percentages) of colonies of gorgonian species versus the relative abundances associated with sponges. Gorgonian abundances derived from Yoshioka and Yoshioka (1989). Br = Briareum asbestinum was excluded from the regression analysis. Taxa code gorgonians in decreasing order of abundance are Pa = Pseudopterogorgia americana Gmelin, 1791 and Pseudopterogorgia acerosa Pallas, 1766; Ecal = Eunicea calyculata Ellis and Solander, 1786 and E. tourneforti Milne Edwards and Haime, 1857; Psx = Pseudoplexaura spp. Houttuyn 1772; Esuc = Eunicea succinea Pallas, 1766, Elax = E. laxispica Lamarck, 1815; Gv = Gorgonia ventalina Linnaeus, 1758; Pflx = Plexaura flexuosa Lamouroux, 1821; Ph = Plexaura homomalla Esper, 1792. The less abundant unlabeled taxa in the figure include: Mur = Muricea elongata Lamouroux, 1821 - M. muricata Pallas, 1766; Mflv = Muriceopsis flavida Lamarck, 1815; Plla = Plexaurella spp.; E sp. 7 = Eunicea sp. 7, Emam = Eunicea mammosa Lamouroux, 1816; Elac = Eunicea laciniata Duchassaing and Michelotti, 1860; Efus = Eunicea fusca Duchassaing and Michelotti, 1860; E sp. 2 = Eunicea sp. 2; Pte = Pterogorgia anceps Pallas, 1766.

446

species are equivalent to their relative abundances in the study site. This result indicates that the number of colonies of the gorgonian species associated with sponges are mostly determined by the abundances of the gorgonian species. Associations between sponges and gorgonian species can be described as random in this respect. A notable exception to this overall random pattern was Briareum, which was found highly associated with sponges relative to its low abundance. Most interactions with sponges had negative effects on gorgonians as observed in the overgrowth (smothering) of various gorgonian species by sponges such as Desmapsamma anchorata, Dysidea janiae and Iotrochota birotulata. Smothering could be direct in the sense that the sponge adhered directly to the tissue/skeleton of the overgrown organism (i.e., direct smothering, Fig. 2) or indirect where lethal or non lethal effects occurred without the adherence of the sponge tissue (i.e., indirect smothering). These and other interactions are given in Table 3. These interactions were often species-specific. For instance, Desmapsamma anchorata usually outcompeted Pseudopterogorgia spp. by direct overgrowth but tissue discoloration of Gorgonia ventalina adjacent to the growing edge of Desmapsamma (Fig. 3) indicated the presence of allelochemicals. A notable exception to the negative interactions with gorgonians was observed between Briareum and Desmapsamma anchorata, Amphimedon compressa, and Mycale carmigropila. In these cases, Briareum and sponge colonies were intertwined and no indications of negative interactions were observed.

Discussion
We stress that the results of this study are preliminary in the sense that future observations will increase the number of sponges and associated species as well as species-specific interactions. Nevertheless, some general inferences can be made about Caribbean sponges and the biodiversity of coral reefs. As in the case of tropical rainforests where the tree canopy accounts for much of the biodiversity, the gorgonian canopy serves as a habitat for many sponge species. Our results thus indicate that the high biodiversity of sponges is partially attributable to their ability to circumvent substrate space limitation in coral reefs (Connell 1978) by growing epibiotically on other sessile organisms (Rtzler 1970). In turn, sponges harbor numerous invertebrate species (Neves and Omena 2003, Ribeiro et al. 2003). The associations and interactions of sponges with other species also affect the biodiversity of sponges. Our results suggest that, with the exception of Briareum, the relative abundance of gorgonian taxa largely determines their associations with sponges. This pattern represents a random association in the sense that the probability of a gorgonian colony being associated with sponges is equal among individuals of all gorgonian species. In contrast to patterns of associations, interactions between sponges and gorgonians are often species-specific. Depending upon the species involved negative interactions may involve direct or indirect smothering or allelochemicals. Also, in addition to negative interactions, the intertwined growth morphology observed between Briareum and various sponges may be indicative

Fig. 2: Adherence of Dysidea janiae to Pseudoplexaura spp. Direct smothering of the tissue of Pseudoplexaura spp. caused by the overgrowing Dysidea janiae.

Table 3: Sponges associated with gorgonians and their modes of interaction. Species Amphimedon compressa Amphimedon erina Callyspongia vaginalis Desmapsamma anchorata Dysidea janiae Iotrochota birotulata Ircinia strobilina Monanchora arbuscula Monanchora unguifera Mycale laevis Mycale microsigmatosa Niphates erecta Scopalina ruetzleri Xestospongia caycedoi Interactions overgrowth, smothering and intertwined growth smothering smothering overgrowth, smothering and intertwined growth overgrowth, smothering overgrowth, smothering overgrowth, smothering overgrowth smothering smothering intertwined growth overgrowth, smothering smothering smothering

447

Fig. 3: Overgrowth of a Gorgonia ventalina colony by Desmapsamma anchorata. A. May 2004; B. January 2005. Note the dying/ dead tissue (arrow) in front of the Desmapsamma anchorata colony indicating allelopathic effects.

of a mutualistic interaction. Wulff (1997) documented that intertwined growth of sponge species has mutualistic effects (see also Calcinai et al. 2004). Jackson (1977) suggests that the nature of interactions is related to various biological characteristics, such as differences in growth rates and morphologies of the sponge and associated organism. The inferred allelochemical effects of Desmapsamma anchorata on Gorgonia ventalina are consistent with the observations of Buss and Jackson (1979). Combined with possible mutualistic effects of Desmapsamma anchorata with Briareum and negative (overgrowth) interactions with other gorgonians, these results indicate that variations in species-specific interactions may be a major factor underlying the high biodiversity of coral reef systems.

of Marine Sciences of the University of Puerto Rico Mayagez for all their help and support. Funding was provided by NOAA CRES program (NA 17OP2919).

References
Aerts LAM (2000) Dynamics behind standoff interactions in three reef sponge species and the coral Montastraea cavernosa. Mar Ecol 21: 191-204 Aerts LAM, van Soest RWM (1996) Quantification of sponge/coral interactions in a physically stressed reef community, NE Colombia. Mar Ecol Progr Ser 148: 125-134 Bakus GJ, Targett NM, Schulte B (1986) Chemical ecology of marine organisms: an overview. J Chem Ecol 12: 951-987 Buss LW, Jackson JBC (1979) Competitive networks: nontransitive competitive relationships in cryptic coral reef environment. Am Nat 113: 223-234 Calcinai B, Bavestrello G, Cerrano C (2004) Dispersal and association of two alies species in the Indonesian coral reefs: the octocoral Carijoa riisei and the demosponge Desmapsamma anchorata. J Mar Biol Assoc UK 84: 937-941

Acknowledgements
We are grateful to Sven Zea for his assistance on the identification of sponge species and to Klaus Rtzler for revision and providing insights on the manuscript. We thank E. Rodriguez, C. Prada, and A. Mercado for their assistance in the field, the staff of the Department

448

Connell JH (1978) Diversity in tropical rain forest and coral reefs. Science 199: 1302-1309 Diaz MC, Rtzler K (2001) Sponges: an essential component of Caribbean Coral Reefs. Bull Mar Sci 69(2): 535-546 Goreau TF, Hartman WD (1963) Boring sponges as controlling factors in the formation and maintenance of coral reefs. Am Assoc Adv Sci Publ 75: 25-54 Jackson JBC (1977) Competition on marine hard substrata: the adaptive significance of solitary and colonial strategies. Am Nat 111: 743-767 Jackson JBC, Buss LW (1975) Allelopathy and spatial competition among coral reef invertebrates. Proc Nat Acad Sci USA 72: 51605163 Lang JC (1973) Interspecific aggression by scleractinian corals. II. Why the race is not only to the swift. Bull Mar Sci 23: 260-279 Neves G, Omena E (2003) Influence of sponge morphology on the composition of the polychaete associated fauna from Rocas Atoll, northeast Brazil. Coral Reefs 22: 123-129 Reiswig HM (1971) In situ pumping activity of tropical demospongiae. Mar Biol 9: 38-50 Ribero S, Omena E, Muricy G (2003) Macrofauna associated to Mycale microsigmatosa (Porifera, Demospongiae) in Rio de Janeiro State, SE Brazil. Estuar Coast Shelf Sci 57: 951-959 Russ GR (1981) Effects of predation by fishes, competition and structural complexity if substrates on establishment of a marine epifaunal community. J Exp Mar Biol Ecol 42: 55-69

Rtzler K (1970) Spatial competition among Porifera: solution by epizoism. Oecologia (Berlin) 5: 85-95 Stebbing ARD (1973) Competition for space between epiphytes of Fucus serratus L. J Mar Biol Assoc UK 53: 247-261 Suchanek TH, Carpenter RC, Witman JD, Harvell CD (1983) Sponges as important space competitors in the deep Caribbean coral reef communities. Contr W Indies Lab Symp Ser 109: 55-60 Wahle CM (1980) Detection, pursuit, and overgrowth of tropical gorgonians by milleporid hydrocorals: Perseus and Medusa revisited. Science 209: 689-691 Wulff JL (1997) Mutualism among species of coral reef sponges. Ecology 78: 146-159 Wulff JL (2005) Trade-offs in resistance to competitors and predators, and their effects on the diversity of tropical marine sponges. J Anim Ecol 74: 313-321 Wulff JL, Buss LW (1979) Do sponges help hold coral reefs together? Nature 281: 474-475 Yoshioka PM (1998) Are large colonies a key factor in the dynamics of gorgonian populations? Rev Biol Trop 46: 137-143 Yoshioka PM (2005) Biotic neighborhoods of shallow water gorgonians of Puerto Rico. Bull Mar Sci 76:625-636 Yoshioka PM, Yoshioka B (1989) A multispecies, multiscale analysis of spatial pattern and its application to shallow-water gorgonian community. Mar Ecol Progr Ser 54: 257-264

Porifera research: Biodiversity, innovation and sustainaBility - 2007

449

Revision of Calycosoma Schulze, 1899 and finding of Lophocalyx Schulze, 1887 (six new species) in the Atlantic Ocean (Hexactinellida, Rossellidae)
Larisa L. Menshenina(1), Konstantin R. Tabachnick(2), Daniela A. Lopes(3), Eduardo Hajdu(3)
Biophysical Department, Physical Faculty, MSU 2, b.2, Moscow State University, Moscow, 119992, Russia P.P. Shirshov Institute of Oceanology, Russian Academy of Sciences, Nahimovsky pr. 36, 117997, Moscow, Russia. tabachnick@mail.ru (3) Museu Nacional, Departamento de Invertebrados, Universidade Federal do Rio de Janeiro, Quinta da Boa Vista, s/n, 20940-040, Rio de Janeiro, RJ, Brazil. hajdu@acd.ufrj.br
(1) (2)

Abstract: New data are brought on the taxonomy of Calycosoma and Lophocalyx, two closely related rossellid genera. A specimen from the Antarctic, previously identified as Calycosoma validum, is transferred here to Lophocalyx sp. Six new species of Lophocalyx are described, found in different regions of the Atlantic Ocean, between 500 and 2860 m depth. The diagnoses of these two genera are revised, and their likely affinities are discussed. Keywords: New species, Hexactinellida, Atlantic Ocean, taxonomy

Introduction
The monotypic genus Calycosoma was described by Schulze (1899) on the basis of a specimen collected off the East Coast of North America (C. validum Schulze, 1899). Schulze placed it within Asconematidae. Later, however, Ijima (1903, 1904, 1927) transferred it into Rossellidae-Lanuginellinae. Schulze (1903) described another species of Calycosoma, C. gracile Schulze, 1903 off the Indonesian Archipelago. Ijima (1927) moved it to Sympagella Schmidt, 1870. Later, when Caulophacidae was abolished and synonymized with Rossellidae, Sympagella too was classified within subfamily Lanuginellinae (Tabachnick 1999, 2002). Topsent (1910, 1913) reported on C. validum finding in the Antarctic Ocean. This report was cited by Ijima (1927) and followed by Barthel and Tendal (1994), but analysis of Topsents data was not made, in spite of his description of anchorate pentactines in the Antarctic specimen, spicules which are absent from the holotype. The examination of sponges collected by Brazilian, French, Norwegian, Russian and United States oceanographic expeditions in the Atlantic Ocean showed that Lophocalyx is widely represented in this ocean, with forms which are close to Calycosoma but for the single feature respectively, presence or absence of anchorate pentactines. The description of these materials is given below.

sites in the Atlantic Ocean, where all the material studied here has been gathered. The methods employed were the usual ones for preparation microscope slides of dissociated spicules and thick sections, as described elsewhere (Janussen et al. 2004, Lopes et al. 2005), with occasional substitution of nitric acid or sulfochromic solution. Abbreviations used here are: MZUB - Museum of Zoology - University of Bergen (Norway), BMNH - Natural History Museum (London), IORAS - Institute of Oceanology of Russian Academy of Sciences (Moscow), MNHN Musum national dHistoire naturelle (Paris), MNRJ - Museu Nacional - Universidade Federal do Rio de Janeiro (Rio de Janeiro), NMS - National Museum of Scotland (Edinbourg), REVIZEE (Evaluation of the Sustainable Potential of Life Resources in the Economic Exclusive Zone), USNM National Museum of Natural History - Smithsonian Institution (Washington D.C.). Calycosoma Schulze, 1899 Synonymy: Calycosoma Schulze 1899: 27. Not C. gracile Schulze 1903: 14 and not C. validum from Antarctic Topsent 1910: 17; 1913: 606; Barthel and Tendal 1994: 111. Type species: Calycosoma validum Schulze, 1899 (by monotypy) Definition: Basiphytous, often pedunculate Lanuginellinae with dermalia consisting of pinular pentactines and rare hexactines; hypodermalia are of one type (orthotropal pentactines); with prostalia lateralia consisting of diactines;

Materials and methods

The materials studied here were collected by several oceanographic expeditions. Figure 1 shows the collecting

450

.028

.029

.021

.015

.012 .072 .100 15 .010 .119

.002 .020 .016 15 .002 .018 .015 15 .002 .011 .011

.182

.129

.205

USNM (kt1458) avg min max

.046

.137

.053

.156

.098

.169

.099

15

15

.019

.012

.015

.013

.129

.099

.160

.122

.133

USNM (kt1123) avg min max

.068

.053

.091

.061

.101

.099

.079

.133

.091

15

15

15

15

15

.022

.021

.009

.003 .097 .094 .095 2 .004 .094 .083

.167

.122

.104

.022

Table 1: Some measures of spicules of Calycosoma validum Schulze, 1899 (in mm).

oxyoidal microscleres and strobiloplumicomes (slightly modified from Tabachnick 2002). Remarks: An emphasis has been put on the lack of a second category of hypodermal pentactines. As argued below, this is the main feature separating Calycosoma from Lophocalyx. Calycosoma validum Schulze, 1899 (Fig. 1; Table 1) Synonymy: Calycosoma validum Schulze 1899: 27. Not C. validum from Antarctic - Topsent 1910: 17; 1913: 606; Barthel and Tendal 1994: 111. Material examined: Holotype. USNM 4761 - USFCS Albatross, stn. 2573, 40o34.18 N 66o09.00 W (eastern North American continental slope, N of the New England seamounts), 3186 m. A fragment of holotype. BMNH 1908.09.24.029. Other materials. USNM (kt3005; kt522.1; kt522.2; kt523; kt524.1; kt527.1; kt528; kt529.1; kt529.2; kt531; kt533; kt534; kt535.1; kt535.2; kt536; kt537) - USFCS Albatross, stn. 2117, 15o24.40 N 63o31.30 W (Aves Ridge, W of the Lesser Antilles), 1250 m. USNM (kt1123; kt1124; kt1126; kt1128; kt1130; kt1131; kt1132; kt1133; kt1134) USFCS Albatross, stn. 2126, 13o17.45 N 70o01.00 W, 3111 m. USNM (kt135; kt200; kt202; kt204; kt205.1; kt205.2) USFCS Albatross, stn. 2127, 19o4500 N 75o04.00 W, 2997 m. USNM (kt1405; kt1406; kt1407; kt1409; kt1420; kt1421) - R.V. Bartlett 1301-82, stn. 58, 13o48.30 N 67o49.12 W (N of the Netherlands Antilles), 5008 m. USNM (kt1306) R.V. Bartlett 1301-82, stn. 90, 13o27.4 N 64o43.9 W (N

USNM (kt1306) avg min max

.084

.038

.068

.014

.120

.080

.091

.017

12

13

15

15

.014

.022

.020

.010

.002

USNM 4761 - holotype avg min max std

.091

.091

.129

.108

.030

.053

.053

.053

.072

.005

.011 6 .014

.018
L - length, D - diameter, d - diameter of primary rosette.

.114

.072

.070

.091

.073

.090

.009

.011

20

25

25

25

L dermal pentactine distal ray L dermal pentactine tangential ray L atrial pentactine proximal ray L atrial pentactine tangential ray D oxyhemihexaster and oxyhexaster d oxyhemihexaster and oxyhexaster D oxyhexactine D strobiloplumicome d strobiloplumicome

.089

.003

.011

.015

.014

.018

.002

Fig. 1: Distribution of Calycosoma and Atlantic species of Lophocalyx. A. C. validum. B. L. pseudovalida sp. nov. C. L. biogasi sp. nov. D. L. oregoni sp. nov. E. L. brasiliensis sp. nov. F. L. atlantiensis sp. nov. G. L. reiswigi sp. nov.

std

15 15

.050 .016

.116

.040 .014

.061 .018

.006 .002

std

15 15 15

.102 .044 .015

.083 .027 .011

.114

.115 .068 .022

.114

.011 .011 .004

std

451

of Isla Margarita), 3422-3464 m. USNM (kt1458; kt1460) - R.V. Alaminos, stn.68 A3-5B (Gulf of Mexico), 3840 m. USNM (kt 328) - R.V. Oregon, stn. 2199, 24o42 N 92o18 W, 3658 m. IORAS 5/2/414; 5/2/415 - R.V. Akademik Mstislav Keldysh- 1, stn. 54, 21o04.5 N 82o27.6 W (S of Cuba), 4350-4370 m. We do not present the description of these specimens as they are given in the Family Rossellidae chapter of the Systema Porifera (Tabachnick 2002). Only a table (Table 1) with spicule micrometric data for several specimens is given below. Lophocalyx Schulze, 1887 Synonymy: Lophocalyx Schulze 1887: 514. Part of Rossella - R. philippinensis Gray 1872: 137; 1873a: 234; 1873b: 361; Carter 1875: 118; Marshall and Meyer 1877: 261. Polylophus (preoccupied name) Schulze 1885 (nomen nudum): 451; 1886: 47: 1887: 132. Calycosoma (part) - C. validum Topsent 1910: 17; 1913: 606; Barthel and Tendal 1994: 111. Type species: Rossella philippinensis Gray, 1872 (by monotypy) Definition: Lophophytous or basiphytous Lanuginellinae with hypodermal pentactines of two types: with normal, straight tangential rays and with short-toothed anchorate rays, microscleres with discoidal outer ends are rare microdiscohexasters (modified from Tabachnick 2002). Remarks: The definition of Tabachnick (2002) has been slightly modified for greater clarity. It is unclear how to classify the large hexactines seen in Lophocalyx spp.: choanosomal or hypoatrial. The radial rays in these spicules are different; the distal ray is pinular or carries some spines, while the tangential rays are similar in size and shape to those in hypodermal pentactines. Such spicules are known from several genera of Rossellidae. Lophocalyx pseudovalida sp. nov. (Fig. 1, 2, 3; Table 2) Etymology: The species is named pseudovalida because of its resemblance to Calycosoma validum. Material examined: Holotype - IORAS 5/2/3111 - R.V. Akademik Mstislav Keldysh- 43, stn. 3988, submarine Mir, 44o57,40 N 28o00,90 W (Mid Atlantic Ridge, N of the Azores), 2800 m. Paratype - IORAS 5/2/3112 - Ibid. Description: Body. The sponge is tubular with relatively small osculum, basiphitous with conules of prostalia lateralia of diactines gathered in tufts. The holotype is 55 mm high, about 30 mm in diameter, the walls are about 5 mm thick, the osculum is 11 x 16 mm in diameter, the conules are 2-3 mm in length, the largest prostalia lateralia protrude 5 mm. The paratype is smaller with broken upper part: 40 mm high, about 25 mm in diameter, the walls are about 3 mm in thickness. Spicules: The choanosomal skeleton is composed of diactines from 1.5 to several mm long and 0.006-0.060 mm in diameter.

Fig. 2: Lophocalyx pseudovalida sp. nov., holotype (IORAS 5/2/3111). External shape (scale 20 mm).

The smaller ones of these have a widening in the middle; the larger ones are stout; their outer ends conically pointed, rounded, rarely clavate or lanceolate, smooth or rough. The diactines that serve as prostalia lateralia are about 18 mm long and 0.03-0.04 mm in diameter with rough surface. Hypodermal pentactines are of two types for Rossellidae, normal and rare anchorate, which serve as prostalia lateralia. The normal hypodermal pentactines have tangential rays 0.41.0/0.023-0.038 mm (here and bellow measures offered are smallest length largest length/smallest diameter close to the centre largest diameter close to the centre); the proximal ray is 0.9-1.2 mm long. They are smooth with conically pointed outer ends. The anchorate pentactines have short tangential rays 0.2-0.4/0.042-0.053 mm; the proximal ray is about 8 mm long, it is smooth at base and rough in proximal part. Hypoatrialia are mostly hexactines, sometimes diactines and rarely pentactines and stauractines. The choanosomal hexactines have proximal ray 0.14-0.38 mm long, tangential rays are 0.17-0.38 mm long, distal ray is 0.17-0.53 mm long, and their diameter is 0.009-0.076 mm. These rays are smooth, with conically pointed outer ends either smooth or

452

Table 2: Some mesures of spicules of Lophocalyx pseudovalida sp. nov. (in mm). n L dermal hexactine distal ray L dermal hexactine tangential ray L dermal hexactine proximal ray L atrial hexactine proximal ray L atrial hexactine tangential ray L atrial hexactine distal ray D oxyhemihexaster and oxyhexaster d oxyhemihexaster and oxyhexaster D strobiloplumicome d strobiloplumicome 25 25 25 25 25 25 25 25 25 25 IORAS 5/2/3111 - holotype avg min max .106 .088 .078 .103 .078 .069 .124 .013 .044 .016 .068 .061 .053 .068 .061 .046 .076 .009 .032 .009 .152 .144 .099 .137 .110 .084 .173 .018 .052 .020 std .022 .015 .011 .019 .011 .011 .020 .002 .004 .003 n 25 25 25 25 25 25 25 25 25 25 IORAS 5/2/3112 - paratype avg min max .115 .091 .183 .117 .084 .078 .128 .014 .043 .017 .076 .076 .061 .076 .068 .053 .079 .011 .032 .014 .160 .114 .106 .190 .114 .137 .157 .018 .050 .018 std .023 .012 .012 .027 .012 .020 .014 .002 .005 .001

L - length, D - diameter, d - diameter of primary rosette.

rough; the outer ends of proximal rays are often tuberculated. The hypodermal/hypoatrial pentactines and choanosomal hexactines have similar dimensions. The diactines located in the vicinity of the atrial surface have rays similar in shape to those of hexactines. These diactines are 0.6-1.7 mm long and 0.023-0.038 mm thick. Dermalia are pinular hexactines, rarely pentactines with the unpaired ray distally directed. The distal ray of dermal hexactines is clavate with short spines, it is 0.068-0.160 mm in length, the tangential rays are 0.061-0.114 mm long, and the proximal ray is 0.053-0.106 mm long. The rays of dermal spicules, except the dermal one, are rough with conically pointed outer ends; they are 0.007-0.009 mm in diameter. Atrialia are mostly hexactines, sometimes pentactines. The proximal ray of atrial hexactines is 0.068-0.190 mm long, tangential rays are 0.061-0.114 mm long, the distal one is 0.046-0.137 mm long. They are 0.005 mm in diameter, the proximal ray is nearly equal to the others (which are rough) but it has short spines. Microscleres: The microscleres comprise oxyhemihexasters, some oxyhexasters and strobiloplumicomes. The oxyhemihexasters and oxyhexasters are 0.076-0.173 mm in diameter; their primary rosettes are 0.009-0.018 mm in diameter. Their secondary rays are slightly rough 0.004-0.005 mm in diameter. The strobiloplumicomes are 0.032-0.052 mm in diameter, with primary rosette 0.009-0.020 mm in diameter; their primary rays have a terminal spine. Remarks: The new species is distinguished on the basis of specific dermal spicules, mostly hexactines, and presence of hypoatrial hexactines. The rare anchorate pentactines and composition of microscleres permits referring this sponge to Lophocalyx. Lophocalyx biogasi sp. nov. (Fig. 1, 4; Table 3) Etymology: The name of the species is derived from Biogas, the name of the French expedition during which it was collected.

Material examined (slope between the Celtic Sea and Biscay Plain): Holotype - MNHN HCL 590 - Biogas V, stn. CW 40, 47o33.10 N 9o01.90 W, 2860 m. Paratypes: MNHN HCL 591; HCL 592; HCL 593 - Ibid. MNHN HCL 594; HCL 595- Byocyan II, stn. Pl. 18, submarine Cyana, 47o32.05 N 8o27.55 W, 2000 m. MNHN HCL 596 - EPI I, R.V. Suroit, stn. CP 39, 47o32 N 8o38 W, 2100 m. MNHN HCL 597 Biogas VII, CP 26, 47o32.80 N 8o33.50 W, 2115 m. Description: Body. Various fragments represent this species: tubular or lamellate. The holotype is 90 mm high, 170 x 120 mm in diameter; the walls are about 2 mm in thickness. The paratype MNHN HCL 594 is 22 mm in length, 20 x 40 mm in diameter; the walls are 4 mm thick. The paratype MNHN HCL 595 is 150 mm long, 60 mm in diameter; the walls are about 4 mm thick. The paratype MNHN HCL 597 is 210 mm long, 70 mm in diameter; the walls are 2-3 mm thick. The other fragments are lamellate, some as large as 300 x 300 mm (MNHN HCL 591), 2-5 mm in thickness. Hence the sponges may be large. Spicules: The choanosomal skeleton is composed of diactines with rounded or conically pointed rough outer ends. The smaller diactines have a widening in the middle; the large spicules are stout. Some of the largest diactines have traces of synapticular junctions. The diactines are up to 20 mm in length, 0.004-0.099 mm in diameter. Hypodermal pentactines are of two types: normal and rare anchorate. The normal hypodermal pentactines have tangential rays 0.5-0.7/0.019-0.038 mm; the proximal ray is 1.0-1.5 mm long. These rays are smooth with rounded, rough outer ends. The anchorate pentactines have curved tangential rays 0.23-0.38/0.038 mm. Hypoatrialia are represented by a few pentactines. Dermalia are pinular hexactines and rare pentactines. The distal ray of dermal hexactines is clavate with short spines, it is 0.044-0.229 mm long, the tangential rays are 0.052-0.155 mm long, and the proximal ray is 0.044-0.126 mm long. The rays of dermal spicules, except the distal one, are rough with conically pointed or rounded outer ends; they are 0.005-0.008 mm in diameter. The rare dermal pentactines are similar to hexactines but they have a short rudimentary tubercle instead of the proximal ray. The distal ray of these pentactines is 0.070-0.170 mm long, the tangential rays are about 0.074 mm

453

Fig. 3: Spicules of the holotype of Lophocalyx pseudovalida sp. nov. (IORAS 5/2/3111). A. dermal hexactine. B-C. atrial hexactines. D. atrial pentactine. E-F. hypoatrial pentactines. G-H. choanosomal hexactines. I. anchorate pentactine. J. hypodermal pentactine. K. hypoatrial diactine. L-O. fragments of large choanosomal diactines. PQ. fragments of small choanosomal diactines. R-S. oxyhemihexasters. T. strobiloplumicome.

long, and the reduced proximal ray is 0.011-0.015 mm long. Atrialia are pentactines and hexactines occurring in similar proportions. The proximal pinular ray of these spicules is conically pointed, its outer end protrudes far beyond the last spines, and the spines are sparse and longer than those of dermal spicules are. The other rays are rough. The proximal ray of atrial hexactines is 0.089-0.185 mm long, tangential rays are 0.044-0.137 mm in length, distal one is 0.044-0.126 mm long, and these rays are 0.005-0.008 mm in diameter. The proximal ray of atrial pentactines is 0.052-0.159 mm

in length; tangential rays are 0.063-0.141 mm long, distal rudimental one is 0.009-0.030 mm long. Microscleres: Oxyhemihexasters, oxyhexactines and their derivatives with reduced rays, rare oxyhexasters and strobiloplumicomes represent the microscleres. The oxyhemihexasters and oxyhexasters are 0.058-0.115 mm in diameter; their primary rosettes are 0.005-0.014 mm in diameter. The oxyhexactines and their derivatives are 0.054-0.108 mm in diameter. The oxyoidal spicules have

454

Table 3: Some measures of spicules of Lophocalyx biogasi sp. nov. (in mm). n 15 15 15 1 1 1 .157 .107 .092 .128 .105 .017 .085 .008 .088 .027 .014 .079 .022 .011 .104 .032 .018 .007 .005 .003 12 2 6 .088 .027 .015 .076 .025 .011 .101 .029 .018 .006 .011 .001 15 .008 .005 .011 .058 .097 .010 15 .080 .065 .097 .008 .001 .008 .003 .002 .011 .019 .003 12 .016 .011 .026 .004 25 15 15 2 10 12 .085 .126 .014 12 .107 .085 .130 .016 25 .052 .152 .034 12 .119 .074 .141 .022 25 .116 .099 .015 .089 .009 .085 .029 .015 .044 .126 .020 14 .084 .067 .096 .009 6 .078 .067 .130 .017 14 .098 .070 .137 .020 6 .093 .044 .056 .081 .067 .009 .072 .007 .079 .025 .013 .115 .185 .020 14 .134 .111 .148 .013 6 .118 .100 .148 .118 .096 .159 .141 .030 .115 .014 .090 .036 .018 .011 .011 .011 .019 .027 .016 .018 .017 .002 .011 .002 .008 .004 .001 .074 .074 .074 .170 .170 .170 1 1 1 17 17 17 12 12 12 10 10 25 11 11 .085 .067 .100 .010 11 .085 .059 .104 .014 25 .087 .063 .111 .014 11 .091 .070 .111 .013 25 .075 .069 .070 .074 .015 .109 .083 .069 .107 .090 .017 .075 .007 .079 .028 .012 .160 .096 .204 .028 11 .106 .044 .159 .036 25 .118 MNHN HCL 595 paratype avg min max std n n .081 .052 .044 .070 .074 .015 .089 .063 .044 .093 .063 .011 .061 .005 .054 .025 .011 MNHN HCL 597 paratype avg min max std MNHN HCL 594 paratype avg min max std .167 .093 .096 .070 .074 .015 .130 .111 .100 .133 .107 .022 .090 .011 .104 .032 .014 .014 .012 .013 .012 .012 .003 .009 .001 .013 .002 .002 .022 .011 .011

n .167 .107 .096 .078 .118 .013 .081 .137 .015 .104 .218 .036

MNHN HCL 590 - holotype avg min max std

15

15

15

15

15

15

15

L dermal hexactine distal ray L dermal hexactine tangential ray L dermal hexactine proximal ray L dermal pentactine distal ray L dermal pentactine tangential ray L dermal pentactine rudimental ray L atrial hexactine proximal ray L atrial hexactine tangential ray L atrial hexactine distal ray L atrial pentactine proximal ray L atrial pentactine tangential ray L atrial pentactine distal rudimental ray D oxyhemihexaster and oxyhexaster d oxyhemihexaster and oxyhexaster D oxyhexactine D strobiloplumicome d strobiloplumicome

15

9 4 4

L - length, D - diameter, d - diameter of primary rosette

455

Fig. 4: Spicules of Lophocalyx biogasi sp. nov.: A-G, J, LP, S, holotype (MNHN HCL 590); H, I, Q, paratype (MNHN HCL 591); K, paratype (MNHN HCL 597); R, paratype (MNHN HCL 594). A. dermal hexactine. B. dermal pentactine. C. atrial pentactine. D. atrial hexactine. E. hypodermal pentactine. F. hypodermal stauractine. G. hypodermal tauactine. HI. anchorate pentactine. J-K. fragments of large choanosomal diactines. L. small choanosomal diactine. M. oxyhexaster. N. oxyhemihexaster. O oxyhexactine. P and R. abnormal oxyoidal spicules. Q. oxystauractine. S. strobiloplumicome.

nearly smooth rays about 0.001 mm in diameter. The strobiloplumicomes are 0.022-0.036 mm in diameter, with primary rosette 0.011-0.018 mm in diameter; their primary rays have each a terminal spine. Remarks: Lophocalyx biogasi sp. nov. differs from L. pseudovalida sp. nov. by the following features: (1) the dimensions of oxyoidal microscleres (0.054-0.115 mm in diameter in L. biogasi sp. nov., 0.076-0.173 mm in L. pseudovalida sp. nov.), (2) the rays of oxyoidal microscleres (thinner in L. biogasi sp. nov., about 0.001 mm in diameter; vs. 0.004-0.005 mm in L. pseudovalida sp. nov.), (3) rarity of oxyhexasters and predominance of derivatives of oxyhexactines and oxyhemihexasters with reduced ray number in L. biogasi sp. nov., (4) smaller size of strobiloplumicomes (0.022-0.036 mm in diameter in L. biogasi sp. nov., 0.0320.052 mm in L. pseudovalida sp. nov.). Very similar specimens to this new species were collected off Peru by the French expedition NAVTIPERC-2, R.V. Naudir.

Lophocalyx oregoni sp. nov. (Fig. 1, 5; Table 4) Etymology: The name of the species, oregoni, is derived from the name of the United States research ship R.V. Oregon from which all the type series was collected. Material examined: Holotype - USNM (kt 286) - R.V. Oregon, stn. 11136, 2427 N 8738 W (Gulf of Mexico), 500 m. Paratypes: USNM (kt 287; kt 288) - Ibid. Description: Body. Lamellate fragments, all of which likely belong to a single specimen, represent this sponge. The holotype is about 1.5 mm in thickness with notable prostalia, probably basalia. The other fragments are 2-4 mm thick without the notable prostalia. It is likely that this sponge is lophophitous. Spicules: The choanosomal skeleton is composed of diactines 0.9-11.0 mm long and 0.007-0.068 mm thick, with conically pointed, rounded or clavate, rough outer ends. The large diactines are stout and the small ones have a widening in the middle, some of them have traces of synapticular junctions.

456

Fig. 5: Spicules of the holotype of Lophocalyx oregoni sp. nov. (USNM kt 286). A. dermal hexactine. B. atrial pentactine. C. atrial hexactine. D. hypodermal pentactine. EF. anchorate pentactines. G-H. large choanosomal diactines. I. small choanosomal diactine. J. oxyhexaster. K. oxyhexactine. L. oxypentactine. M-N. oxystauractines. O. oxydiactine. P. abnormal oxyoidal spicule. Q. strobiloplumicome.

Table 4: Some measures of spicules of Lophocalyx oregoni sp. nov. (in mm). n L dermal hexactine distal ray L dermal hexactine tangential ray L dermal hexactine proximal ray L atrial hexactine proximal ray L atrial hexactin tangential ray L atrial hexactine distal ray L atrial pentactine proximal ray L atrial pentactine tangential ray L atrial pentactine rudimental ray D oxyhemihexaster and oxyhexaster d oxyhemihexaster and oxyhexaster D oxyhexactine D strobiloplumicome d strobiloplumicome 25 25 25 4 4 4 17 17 17 1 1 25 25 25 USNM(kt286) - holotype avg min max .110 .103 .087 .111 .101 .096 .122 .098 .013 .072 .007 .078 .028 .012 .070 .081 .063 .093 .093 .081 .074 .081 .011 .072 .007 .050 .022 .009 .133 .122 .133 .126 .107 .111 .133 .118 .019 .072 .007 .108 .040 .016 std .017 .014 .016 .014 .006 .015 .015 .009 .002 .014 .005 .002 n 15 15 15 2 2 2 15 15 15 1 1 15 15 15 USNM(kt288) paratype avg min max .100 .104 .086 .120 .085 .102 .104 .101 .012 .061 .007 .083 .029 .012 .070 .085 .063 .111 .074 .081 .078 .074 .007 .061 .007 .058 .025 .011 .122 .130 .118 .130 .096 .122 .130 .126 .015 .061 .007 .108 .036 .014 std .017 .011 .015 .013 .016 .029 .015 .015 .002 .013 .003 .001

L - length, D - diameter, d - diameter of primary rosette

Hypodermal pentactines are of two types: normal and anchorate, the latter of which act as prostalia. The normal hypodermal pentactines have tangential rays 0.44-0.99/0.0150.027 mm; the proximal ray is about 1.5 times the length of the tangential ones. The rays are smooth with conically

pointed rough outer ends. The anchorate pentactines have short tangential rays 0.23-0.34/0.019-0.068 mm; the proximal ray is 0.9-32.0 mm long. Some of the anchorate pentactines near the dermal surface (mostly with short proximal ray) have hook-like tangential rays and smooth proximal rays.

457

The largest pentactines located above the dermal surface and likely functioning as basalia have straight tangential rays bent proximally (the tangential rays have dimensions similar to the hook-like ones) and shafts which become rough at a certain distance from the distal part of the spicule. Hypoatrialia are likely to be absent; some rare orthotropal pentactines are very similar to dermal ones and may be allochthonous. Dermalia are pinular hexactines, rarely pentactines with the unpaired ray distally directed and a rudimental tubercle as a reduced proximal ray. The distal ray of dermal hexactines is 0.070-0.133 mm long, it is clavate with relatively long spines, the tangential rays are 0.081-0.130 mm long, and the proximal ray is 0.063-0.133 mm long. The rays of the dermal spicules, except the dermal one, are rough with conically pointed outer ends; they are 0.007-0.009 mm in diameter. Atrialia are mostly pinular pentactines, sometimes hexactines. Their pinular ray is conically pointed; less spiny and thinner than those of dermal spicules, the reduced distal ray in pentactines is represented by a short tubercle; the tangential rays are rough with conically pointed outer ends. The proximal ray of atrial pentactines is 0.074-0.133 mm in length, tangential rays are 0.074-0.126 mm long, and the reduced distal ray is 0.007-0.019 mm long. The distal ray of atrial hexactines is 0.081-0.122 mm in length. These rays are 0.005-0.006 mm in diameter. Microscleres: The microscleres are abnormal oxyoidal spicules (with some reduced rays) derived from oxyhexactinic forms and oxyhemihexasters, oxyhexactinal microscleres are rare oxyhexasters and strobiloplumicomes. Most of the oxyoidal spicules have hook-like outer ends, rarely straight; their rays are slightly rough. The oxypentactines, oxystauractines and oxyhexactines prevail over the other forms. Sometimes they have small tubercle-like rudiments in place of the absent ray. These spicules are 0.050-0.108 mm in diameter. Their rays are rough and 0.001-0.002 mm in diameter. The strobiloplumicomes are 0.022-0.040 mm in diameter, with a primary rosette 0.009-0.016 mm in diameter; the primary rays each have a short terminal spine. Remarks: Lophocalyx oregoni sp. nov. is distinguished by its peculiar hook-like oxyoidal microscleres mostly comprising abnormal forms of oxyhexactines with some reduced rays; specific dermal spicules are mostly hexactines, and atrial spicules are mostly pentactines. The anchorate pentactines with straight tangential rays in addition to curved ones are known also in L. philippinensis (e.g. Tabachnick, 2002) but in L. oregoni sp. nov. these spicules prevail in the likely basalia, while the hook-like ones do not protrude far and act as prostalia lateralia. Lophocalyx brasiliensis sp. nov. (Fig. 1, 6; Table 5) Etymology: The name of the species is derived from its type locality, off the southeastern Brazilian Coast. Material examined: The holotype is distinguished formally as a fragment which together with the other fragments (paratypes), probably belong to a single specimen. Holotype - MNHN HCL 598 - R.V. Marion Dufresne- MD 55, sta. 64

CB 105, 23o46 S 42o9 W (southeastern Brazilian continental slope), 597-610 m. Paratypes: MNHN HCL 599-605 - Ibid. Description: Body. The studied material is composed of relatively small fragments. The holotype was chosen on the basis of its possession of both dermal and atrial spicules in the same fragment, a single conule with a tuft of prostalia lateralia 15 mm in length, 12 x 3 mm in diameter. The other fragments (paratypes) have a similar shape. Prostalia lateralia protrude about 20 mm. Spicules: The choanosomal skeleton is composed of diactines from 1.8 mm to several mm in length; their diameter is 0.0060.038 mm. The diactines, of the prostalia are longer (about 25 mm long). The choanosomal diactines are conically pointed, rarely clavate, with rough or rarely smooth outer ends. The smaller diactines have a widening in the middle, the larger ones are stout. Hypodermal pentactines are of two types: normal and anchorate, both as prostalia. Some of the normal hypodermal pentactines have smooth rays with conically pointed rough outer ends; their tangential rays are 0.44-0.68/0.014-0.038 mm, the proximal ray is 2.3-3.0 mm long. The other normal pentactines with regular orthotropal tangential rays have all the rays rough with conically pointed outer ends. Their tangential rays are smaller and thinner, 0.17-0.43 mm long and 0.004-0.011 mm thick, than those in smooth pentactines, the proximal ray is usually about 1.5 times longer. The anchorate pentactines are very rare; rays are smooth, with straight or slightly bent tangential ones. The tangential rays are 0.1-0.2 mm long; the proximal ray is very long, being 0.015-0.023 mm in diameter at base and 0.030-0.053 mm in diameter in some distance. Hypoatrialia seem to be similar to the hypodermal spicules, but anchorate forms are lacking and hexactines with short proximal ray and rounded rough outer end are present. The proximal ray of these spicules is 0.130.21 mm long and 0.009-0.027 mm thick; tangential rays are 0.21-0.47 mm long, the distal ray is 2.1-4.6 mm long and is equal in diameter to the other rays at its base, but thicker (about 0.04 mm in diameter) at some distance from the base. Dermalia are pinular hexactines with slightly clavate pinular distal ray bearing relatively long spines; all other rays are rough. The distal ray is 0.061-0.144/0.006-0.007 mm, the tangential rays are 0.068-0.160 mm long, and the proximal ray is 0.068-0.099 mm long. Atrialia are hexactines with spiny or rough rays. Their proximal ray is 0.099-0.175/0.0040.038 mm, the tangential rays are 0.076-0.160 mm long, and the distal ray is 0.046-0.129 mm long. Microscleres: The microscleres comprise oxyoidal, asterous or nearly asterous (with very short primary rays) spicules derived from oxyhemihexasters, oxyhexactines and strobiloplumicomes. Most oxyoidal spicules have minute spines, but sometimes a few rays may have long spines. These spicules are 0.065-0.176 mm in diameter, and morphotypes with and without secondary rays are present. The central portions of these microscleres vary from minute to slightly spherical, 0.009-0.014 mm in diameter. The rays of these spicules are 0.002-0.005 mm in diameter. The strobiloplumicomes are 0.025-0.047 mm in diameter, with primary rosette 0.013-0.022 mm in diameter; the primary rays have each a short terminal spine.

458

Fig. 6: Spicules of Lophocalyx brasiliensis sp. nov.: A-G, I-W, holotype (MNHN HCL 598); H, paratype (MNHN HCL 605). A-B. dermal hexactines. C-D. atrial hexactines. E. choanosomal hexactine. F. hypodermal rough pentactine. G-I. hypodermal pentactine. J-L. large choanosomal diactines. M. small choanosomal diactine. N-Q. asters derived from oxyhemihexasters. R. oxystauractine. S-T. oxyoidal microscleres derived from hexactines. U-V. outer ends of oxyoidal microscleres. W. strobiloplumicome.

Remarks: Lophocalyx brasiliensis sp. nov. is distinguished from other species in the genus by its characteristic oxyoidal microscleres which have asterose forms with thick secondary rays 0.002-0.005 mm in diameter. These microscleres are derived mostly from oxyhemihexasters. Lophocalyx atlantiensis sp. nov. (Fig. 1, 7, 8; Table 6) Etymology: The name of the species indicates its central north Atlantic type locality. Material examined (Mid Atlantic Ridge, Charlie-Gibbs Fracture Zone): Holotype - MZUB n14822 - Mar-Eco,

superstation 70, local station 385, 5258.540 - 5257.950 N, 3452.150 - 3451.910 W, 1860-2165 m. Paratype MZUB n15318 - Mar-Eco, superstation 60, local station 380, 5155.080 -5156.140 N 3025.020 - 3024.440 W, 1911-1830 m. Description: Body. The holotype is ovoid, 23 mm high, its diameter is 19 mm in the middle, the osculum is about 4 mm in diameter, several tufts of prostaslia lateralia protrude about 5 mm above the body. The paratype is a small fragment. Spicules: Choanosomal spicules are diactines 0.75-2.1 mm long and 0.006-0.014 mm thick, with a widening in the middle portion, surrounded by four rudimentary tubercles; their outer ends being conically pointed, rough.

459

Table 5: Some measures of spicules of Lophocalyx brasiliensis sp. nov. (in mm). n L dermal hexactine distal ray L dermal hexactine tangential ray L dermal hexactine proximal ray L atrial hexactine proximal ray L atrial hexactine tangential ray L atrial hexactine distal ray D oxyhemihexaster and oxyhexaster d oxyhemihexaster and oxyhexaster D strobiloplumicome d strobiloplumicome 25 24 18 25 25 25 25 25 5 5 MNHN HCL 598 - holotype avg min max std .090 .111 .073 .138 .109 .087 .121 .011 .037 .017 .061 .068 .038 .099 .076 .046 .065 .009 .025 .013 .144 .160 .099 .175 .160 .129 .176 .014 .047 .022 .018 .027 .016 .025 .019 .019 .022 .002 .008 .003

L - length, D - diameter, d - diameter of primary rosette

Prostalia in tufts are mostly diactines about 7 mm long and thick 0.05-0.08 mm. Hypodermal pentactines are of two types: normal and anchorate. The anchorate pentactines have tangential rays about 0.12 mm long, the proximal ray is more than 3 mm long, their diameter is 0.04 mm, and the tangential rays are usually smooth. The normal pentactines have tangential rays 0.25-0.50 mm long, the proximal ray is 0.6-1.10 mm long, their diameter is 0.010-0.020 mm, the rays and outer ends of these spicules are usually smooth, rarely rough. Choanosomal hexactines were found in large amounts in the paratype, their distal ray is rough while all the other rays are smooth. The distal ray in hypodermal hexactines is about 0.30 mm long, other rays are about 0.5 mm long, and their diameter is about 0.015 mm. Dermalia are pinular hexactines, their pinular ray is stout, 0.043-0.123 mm long, tangential rays are 0.050-0.121 mm long, the proximal ray is 0.011-0.092 mm long, their diameter is 0.005-0.008 mm. Atrialia are also pinular hexactines, their proximal ray is 0.047-0.182 mm long, tangential rays are 0.058-0.126 mm long, distal ray is 0.007-0.104 mm long,

Fig. 7: Lophocalyx atlantiensis sp. nov., holotype (MZUB n14822). External shape (scale 10 mm).

their diameter is 0.004-0.008 mm. The rays of dermal and atrial spicules are rough or spiny and have conically pointed outer ends. Microscleres: Oxyoidal microscleres are mostly oxyhexactines and a few oxyhemihexasters, the latter have one ray branching into two, rarely three secondary rays. The diameter of oxyhexactines and oxyhemihexasters is 0.1060.169 mm; the primary ray in the oxyhemihexasters is 0.0030.008 mm. The strobiloplumicomes are 0.025-0.068 mm in diameter with primary rosette 0.012-0.043 mm in diameter; the primary rays have each a short, conical terminal spine.

Table 6: Some measures of spicules of Lophocalyx atlantiensis sp. nov. (in mm). n L dermal hexactine distal ray L dermal hexactine tangential ray L dermal hexactine proximal ray L atrial hexactine proximal ray L atrial hexactine tangential ray L atrial hexactine distal ray D oxyhexactine or oxyhemihexaster D strobiloplumicome d strobiloplumicome 32 31 27 25 25 24 25 25 25 MZUB n14822 - holotype avg min max .081 .088 .064 .111 .087 .068 .146 .046 .019 .043 .050 .011 .047 .058 .007 .122 .029 .012 .109 .121 .090 .182 .126 .104 .169 .068 .043 std .016 .016 .018 .028 .016 .022 .012 .008 .006 n 25 25 25 MZUB n15318 - paratype avg min max .094 .079 .072 .062 .053 .042 .123 .106 .092 std .015 .013 .011

34 13 13

.128 .035 .018

.106 .025 .014

.151 .047 .022

.011 .006 .002

L - length, D - diameter, d - diameter of primary rosette

460

Fig. 8: Spicules of Lophocalyx atlantiensis sp. nov.: A-E, GN, holotype (MZUB n14822); F, paratype (MZUB n15318). A-B. dermal hexactines. C. atrial hexactine. D. anchorate pentactine. E. hypoatrial pentactine. F. choanosomal hexactine. G. prostalia lateralia (diactine). H-J. choanosomal diactines (an outer end and central parts). K. oxyhexactine. L-M. oxyhemihexasters. N. strobiloplumicome.

Remarks: The specific features of this new species are that oxyoidal microscleres are in the majority oxyhexactines; oxyhemihexasters (with one ray branching only) are only rarely seen, and the pinular ray of dermal hexactines is stout. Lophocalyx reiswigi sp. nov. (Fig. 9, 10; Table 7, 8) Etymology: The specific epithet honours Dr. Henry M. Reiswig (Royal British Columbia Museum, Victoria, Canada),

one of the greatest specialists on hexactinellid systematics, who has just turned 70. Material examined: The holotype is distinguished formally as a fragment, which together with the others fragments (paratypes), probably belong to a single specimen. Holotype - MNRJ 3339C R.V. Thalassa, Programme REVIZEE Bahia II, stn.0496, 07.VI.2000, 1317.580 S 3817.599 W (off Bahia State, Brazil), 1717 m. Paratypes - MNRJ 3339A, MNRJ 3339B, MNRJ 3339D, MNRJ 3339E, MNRJ 3339G and MNRJ 3339I Ibid.

461

Fig. 9: Lophocalyx reiswigi sp. nov.: A-B. holotype (MNRJ 3339C); C-F. paratypes (MNRJ 3339A, MNRJ 3339B, MNRJ3339D, MNRJ 3339E, respectively). External shape (scale 50 mm).

462

Fig. 10: Spicules of Lophocalyx reiswigi sp. nov.: A-O, holotype (MNRJ 3339C); P, paratype (MNRJ 3339B). A-B. dermal hexactines. C-F. atrial hexactines. G. anchorate pentactine. H-I. hypodermal pentactines. G. prostalia lateralia diactine. K. choanosomal diactine. L. oxyhemihexaster. M. oxyhexactine. N. secondary ray of oxyoidal microsclere. O. strobiloplumicome. P. choanosomal hexactine.

Table 7: Some measures of spicules of Lophocalyx reiswigi sp. nov. (in mm). n L dermal hexactine distal ray L dermal hexactine tangential ray L dermal hexactine proximal ray L atrial hexactine proximal ray L atrial hexactine tangential ray L atrial hexactine distal ray D oxyhexactine or oxyhemihexaster D strobiloplumicome d strobiloplumicome 25 25 20 25 25 21 25 8 MNRJ 3339 A paratype avg min max std .143 .101 .090 .113 .075 .072 .112 .016 .081 .073 .065 .078 .063 .057 .091 .013 .196 .154 .104 .183 .117 .099 .138 .021 .029 .015 .010 .028 .017 .011 .010 .003 n 18 23 5 25 25 25 25 2 10 MNRJ 3339 B paratype avg min max std .113 .075 .069 .158 .113 .096 .111 .013 .070 .050 .047 .117 .097 .081 .094 .021 .013 .170 .177 .091 .222 .144 .120 .125 .042 .016 .026 .025 .016 .026 .011 .008 .008 .001 n 25 25 25 25 25 25 25 10 MNRJ 3339 C - holotype avg min max std .147 .099 .091 .159 .105 .090 .113 .014 .089 .057 .070 .104 .068 .065 .102 .013 .201 .123 .117 .211 .123 .115 .130 .016 .031 .017 .011 .026 .014 .012 .007 .001

L - length, D - diameter, d - diameter of primary rosette

463

Table 8: Some measures of spicules of new species of Lophocalyx (in mm). L. pseudovalida sp. nov. min max L dermal hexactine distal ray L dermal hexactine tangential ray L dermal hexactine proximal ray L atrial hexactine proximal ray L atrial hexactine tangential ray L atrial hexactine distal ray D oxyoidal microsclere D strobiloplumicome .068 .061 .053 .068 .061 .046 .076 .032 .160 .114 .106 .190 .114 .137 .173 .052 L. biogasi sp. nov. min max .044 .052 .044 .089 .044 .044 .054 .022 .218 .137 .118 .185 .137 .126 .115 .036 L. oregoni sp. nov. min max .070 .081 .063 .093 .074 .081 .050 .022 .133 .130 .133 .130 .107 .122 .108 .040 L. brasiliensis sp. nov. min max .061 .068 .038 .099 .076 .046 .065 .025 .144 .160 .099 .175 .160 .129 .176 .047 L. atlantiensis sp. nov. min max .043 .050 .011 .047 .058 .007 .106 .025 .123 .121 .092 .182 .126 .104 .169 .068 L. reiswigi sp. nov. min max .070 .050 .047 .078 .063 .057 .091 .021 .201 .177 .117 .222 .144 .120 .138 .042

L - length, D - diameter, d - diameter of primary rosette

Description: Body. The lamellate holotype is a fragment 200 mm long, 145 mm in diameter; the walls are 1-4 mm thick, and with tufts of prostalia marginalia piercing the surface for 9 mm at most. Other specimens are also lamellate fragments, and these could all possibly belong to a single specimen. Both surfaces of the holotype are clearly reticulated to the naked eye, with meshes 1-5 mm in diameter. The paratype MNRJ 3339A is the largest fragment, with 295 x 135 mm in area. MNRJ 3339B has thick walls (10-19 mm), and a detachable atrial surface membrane. MNRJ 3339D is harder than the other specimens. MNRJ 3339I comprises several fragments, one of them with tufts of prostalia marginalia protruding for 15 mm over the dermal surface, and meshes of up to 7 mm in diameter. Spicules: The choanosomal skeleton is composed of diactines with microspined terminations which can be rounded or conically pointed. The diactines of the prostalia lateralia are the stouter ones (1-3/0.015-0.088 mm), while the choanosomal diactines are slender, slightly curved or straight, and have four rudimentary tubercles in the middle (up to at least 19 mm in length and 0.013-0.028 mm in thickness). Many choanosomal diactines are fused by secondary synapticular junctions. Hypodermal pentactines occur in two types: normal and rare anchorate. The normal hypodermal pentactines have short tangential rays 0.5-1/0.01-0.025 mm and proximal rays from 1 mm up to at least 4 mm long, 0.018-0.035 mm in diameter; the rays have conical terminations, and can be spined or rough, all over, or on a variably long distal section. Anchorate pentactines have smooth tangential rays, 0.160.26/0.033-0.043 mm. Hypoatrial pentactines are equal to dermal orthotropal ones. The choanosomal hexactines have slightly curved, rough rays (at least in the distal and proximal rays) and gradually tapering ends. The distal ray in the choanosomal hexactines is about 0.54-0.65/0.013-0.018 mm. Dermalia are pinular hexactines with slender pinular distal rays (sometimes reduced) bearing short spines; all other rays are rough or microspinouse. The distal rays are 0.070-0.201/0.02-0.03 mm, the tangential rays are 0.0500.1770.008-0.015 mm, and the proximal rays are 0.0470.117/0.008-0.013 mm. Atrialia are pinular hexactines with slender or slightly clavate distal rays bearing short spines; all other rays are rough. The distal rays are 0.078-0.222/0.0180.028 mm, the tangential rays are 0.063-0.144/0.008-0.013

mm, and the proximal rays are 0.057-0.120/ 0.008-0.013 mm. Microscleres: Thin rayed oxyoidal microscleres are oxyhexactines and oxyhemihexasters 0.091-0.138 mm in diameter. The primary rays of the latter may branch into 2 secondary rays. The strobiloplumicomes are about 0.0210.040 mm in diameter with primary rosettes 0.013-0.021 mm in diameter. Remarks: Lophocalyx reiswigi sp. nov. is closely related to L. atlantiensis sp. nov. on the basis of the marked similarity of their spicule complement. Main points of distinction are the tangential rays of the anchorate hypodermal pentactines, bent, becoming parallel to the main axis in L. reiswigi sp. nov. and flat, perpendicular to the main axis in L. atlantiensis sp. nov. The normal pentactines are always rough in the former species but only rarely so in the latter species; the pinular ray is larger in the dermal hexactines of L. reiswigi sp. nov. (0.0700.201 mm vs. 0.043-0.123 mm); and the oxyhemihexasters may have more than one branching ray. Lanuginellinae indet. An interesting representative of this subfamily was captured together with L. pseudovalida sp. nov. (R.V. Akademik Mstislav Keldysh- 43, stn. 3988). Unfortunately it is a strongly damaged fragment which contains many allochthonous spicules of other Hexactinellida collected simultaneously, and which is not possible to assign confidently to a genus. Its spicules strongly differ from all known species of the genera Lophocalyx and Calycosoma to one of which this fragment should be referred on the basis of its microsclere complement. The microscleres of this fragment resemble much those of L. biogasi sp. nov. by having many oxyhexactines, while the hypodermal and/or hypoatrial pentactines are mostly large, rough spicules, similar in shape to the small, rough pentactines of L. brasiliensis sp. nov. The possibly dermal and atrial spicules of this sponge have no peculiar features, but are clearly different from those of L. biogasi sp. nov. proximal rays of the likely atrial spicules of this fragmented sponge are entirely covered by spines.

464

Discussion Remarks on the new taxa of Lophocalyx

The species of Lophocalyx may be divided into three groups by characteristics of their dermal skeleton. 1. Sponges with dermal stauractines: L. philippinensis (e.g. Carter, 1875, Schulze, 1886) and L. suluanus (Ijima, 1927) (e.g. Ijima 1927, Lvi 1964). 2. Species where the stauractines are supplemented by pentactines with distally directed unpaired ray: L. spinosa (Schulze, 1900) (e.g. Schulze 1900, 1902). 3. Species with predominance of dermal hexactines (the pentactines are rare): L. moscalevia Tabachnick, 1988 (e.g. Tabachnick 1988, Tabachnick and Lvi 2004), L. sp. from the Antarctic (Topsent 1910, 1913, Barthel and Tendal 1994) and all species of Lophocalyx from the Atlantic Ocean described in this paper. Lophocalyx sp. from the Antarctic [probably a new species; Calycosoma validum sensu Topsent (1910)] is easily distinguished from the other species of the 3 above groups by peculiar, very large dermal hexactines with pinular distal ray covered by spines almost everywhere (Topsent 1910, 1913). Other specimens of Lophocalyx from the Antarctic collected by the R.V. Eltanin (deposited in the USNM, reexamined by KT) are very similar to the specimen of Topsent after analysis of his material stored in the NMS. They all seem to have large hypoatrial hexactines and their interpretation by Topsent as dermal may be erroneous. Unlike the hexactines described above, those in L. pseudovalida sp. nov. have spines or tubercles situated close to the upper end. Lophocalyx moscalevia is more similar to the Atlantic species but its dermal and atrial spicules differ considerably in dimensions from each other. The other differences in the microscleres composition seem to be less important. Lophocalyx moscalevia has oxyhemihexasters, oxyhexactines and derivatives of the latter with reduction of ray number. The features, which allow recognition of the Atlantic species of Lophocalyx, are given in the remarks after their descriptions. Some microdiscohexasters were reported by Ijima (1927) for one of three specimens of L. suluanus. Some discohexasters were found in several specimens of L. philippinensis (Tabachnick 2002). But these discoidal spicules are absent in all the Atlantic species. All the new species of Lophocalyx from the Atlantic Ocean have many common features: dermalia are hexactines, the spicules measures mostly overlap between all these species (see Table 8). The features which permit their differentiation are the shape of dermal pinular rays and the type and shape of oxyoidal microscleres.

Remarks on the affinities of Calycosoma and Lophocalyx

The genus Lophocalyx (then Polylophus) was created by Schulze (1887) for a single specimen, named L. philippinensis (Gray, 1872). Ijima (1927) later mostly accepted his diagnosis. Several species of Lophocalyx have been described since and its diagnosis required clarification.

The genus Calycosoma never had a diagnosis (before Tabachnick 2002), being differentiated in keys from allied Lophocalyx by their dermal skeleton construction. Schulze (1904) proposed to differentiate both genera on the basis of presence of hexactines and pentactines in the former, and stauractines in the latter. At the same time, Ijima (1904) proposed the basiphytous habit, prostalia diactines, dermalia pentactines and hexactines as diagnostic for Calycosoma; while the lophophitous habit, pentactine anchors, dermalia stauractines and pentactines as diagnostic for Lophocalyx. If the criteria by Schulze were the only ones available, Calycosoma would better be regarded as a junior synonym of Lophocalyx through newly found transitional forms: L. moscalevia Tabachnick, 1988 and the new species described above. Furthermore, the criteria by Ijima are not strong enough since a basyphitous form of Lophocalyx is already known. Lophocalyx moscalevia Tabachnick, 1988 is fixed to a dead Hexactinosida directly by the base, but numerous anchorate hypodermal pentactins are also present. Besides it is also possible to imagine lophophytous Calycosoma fixed by diactines only (without anchorate spicules) to soft substrata. Thus, the most important feature, which still allows the distinction of these two genera, is the presence in Lophocalyx of a second category of hypodermal pentactines, which are extended and serve as prostalia anchorate spicules, and are absent in Calycosoma. Additionally, Calycosoma has prostalia diactines gathered in tufts. Calycosoma validum described from the Antarctic by Topsent (1910, 1913) should be transferred to Lophocalyx, as it has these anchorate spicules and also due to other differences from C. validum from the type location. The Antarctic material possibly belongs to a new species, the description of which depends on re-analysis of C. validums type specimen and other Antarctic materials. After description of the new species of Lophocalyx with hexactines in dermalia and especially the new species from the Atlantic Ocean, the definition of Calycosoma also required a revision. The new species of Lophocalyx described here approach closely the spicule set known from C. validum; most species of Lophocalyx from the Atlantic have few anchorate pentactines and walls which are smooth (without prostalia lateralia) or with tufts of prostalia lateralia which are mostly diactines (L. atlantiensis sp. nov., L. brasiliensis sp. nov., L. pseudovalida sp. nov.) as it is known for Calycosoma. But the external shape of the body and the shape of the hypodermal pentactines seem to be important for generic definitions within the Lanuginellinae (Tabachnick 2002). The presence of a peduncle defines three doubtless basiphytous genera: Calycosoma (very short peduncle), Lanugonychia von Lendenfeld, 1915 and Sympagella Schmidt, 1870. Lanuginella Schmidt, 1870 is a basiphytous sponge but with tendency to form a veil of outwardly protruded hypodermal pentactines in some specimens and it may thus have lophophytous forms. Most Atlantic species of Lophocalyx (except L. oregoni sp. nov.) seem to be basiphytous. The shape of the common hypodermal pentactines seems to be an important feature for the recognition of Mellonympha Schulze, 1897: it has hypodermal pentactines with paratropal tangential rays, besides anchorate ones (Tabachnick 2002). The only remaining Lanuginellinae genus, Dochonestes Topsent, 1928,

465

is a monotypic genus known from a single fragment, which is clearly distinct from the other genera mentioned above. It is basyphitous, and has exclusively diactines in the dermalia. As a conclusion, it is worth keeping Calycosoma, albeit monotypic, separate from Lophocalyx in view of the absence of anchorate pentactines in the former. Study of the collections of the Smithsonian Institution (Washington, DC) showed that only specimens collected in the North Atlantic and Caribbean should be related to C. validum.

Acknowledgements
We are thankful to colleagues Drs. J. Vacelet, N. Boury-Esnault (Station Marine dEndoume, Marseille), C. Lvi (Musum National dHistoire Naturelle, Paris), K. Rtzler, K. Smith (National Museum of Natural History, Washington D.C.), C. Valentine (The Natural History Museum, London), A.V. Gebruk, S.A. Evseenko (Institute of Oceanology of Academy of Sciences of Russia, Moscow); F. Ware, S. Chambers (National Museums of Scotland, Edinburgh), Dr. J. Kongsrud and Dr. A. Willassen (Museum of Zoology, University of Bergen), H.P. Lavrado (Instituto de Biologia, Universidade Federal do Rio de Janeiro, Rio de Janeiro) for donation of the materials and others, for their help in investigations of sponge collections and participation in the collecting of the described materials. This work was an element of MAR-ECO, a field project under the Census of Marine Life programme. CENPES/PETROBRAS (Centro de Pesquisas e Desenvolvimento Leopoldo Amrico Miguez de Mello/Petrleo Brasileiro S.A.) is acknowledged for covering the costs of KTs visit to EHs lab, when joint work on L. reiswigi sp. nov. has been possible. DAL and EH are thankful for grants and/or fellowships provided by CNPq (Conselho Nacional de Desenvolvimento Cientfico e Tecnolgico), FAPERJ (Fundao de Amparo Pesquisa do Estado do Rio de Janeiro) and CAPES (Coordenao de Aperfeioamento de Pessoal de Nvel Superior), all from the Brazilian government. This work greatly benefited from the comments of two anonymous reviewers.

References
Barthel D, Tendal OS (1994) Antarctic Hexactinellida. Theses zoologicae. Volume 23. In: Wgele JW, Sieg J (eds). Synopses of the Antarctic benthos. Koeltz Scientific Books, Champaign. pp. 1154 Carter HJ (1875) On the genus Rossella (a hexactinellid sponge) with descriptions of three species. Ann Mag nat Hist (4) 15: 113122 Gray JE (1872) On a new genus of hexaradiate and other sponges discovered in the Philippine Islands by Dr. A.B.Meyer. Ann Mag nat Hist (4) 10(56): 134-139 Ijima I (1903) Studies on the Hexactinellida. Contribution III. (Placosoma, a new Euplectellid: Leucopsacidae and Caulophacidae). J Sci Imp Coll Univ Tokyo 18(1): 1-124 Ijima I (1904) Studies on the Hexactinellida. Contribution IV. (Rossellidae). J Sci Imp Coll Univ Tokyo 18(7): 1-307 Ijima I (1927) The Hexactinellida of the Siboga Expedition. In: Weber M (ed). Siboga-Expeditie. Uitkomsten op zologisch, botanisch,

oceanographisch en geologisch gebied versameld in Nedelandsch Oost-Indi 1899-1900 aan boord H.M. Siboga onder co mmando van Luitenant ter zee le kl. G.F. Tydeman 106 (Monographie VI). E.J. Brill: Leiden. pp. 1-383 Janussen D, Tabachnick KR, Tendal OS (2004) Deep-sea Hexactinellida (Porifera) of the Weddell Sea. Deep-Sea Res II 51: 1857-1882. Lvi C (1964) Spongiaires des zones bathyale, abyssale et hadale. Galathea Rep 7: 63-112 Lopes DA, Hajdu E, Reiswig HM (2005) Redescription of two Hexactinosida (Porifera, Hexactinellida) from the southwestern Atlantic, collected by Programme REVIZEE. Zootaxa 1066: 4356 Marshall W, Meyer AB (1877) ber einige neue und wenig bekannte Philippinishe Hexactinelliden. Mitth Zool Mus Dreseden 2: 261279 Schulze FE (1885) The Hexactinellida. Rep Sci Res Voy H.M.S. Challenger, 18731876. Narrative 1(1): 437-451 Schulze FE (1886) eber den Bau und das System der Hexactinelliden. Abh Knigl Akad Wiss Berlin (Physik-Mathemat Cl) 1886: 1-97 Schulze FE (1887) Report on the Hexactinellida collected by H.M.S. Challenger during the years 1873-1876. Rep Sci Res Voy H.M.S. Challenger, Zool 21: 1-514 Schulze FE (1899) Amerikanische Hexactinelliden, nach dem Materiale der Albatross-Expedition. Fischer, Jena Schulze FE (1900) Hexactinelliden des Indischen Oceanes. III Theil. Abh Knigl Preuss Akad Wiss. Berlin 1900: 1-46 Schulze FE (1902) An Account of the Indian Triaxonida collected by the Royal Indian Marine Survey Ship Investigator. Indian Museum, Calcutta Schulze FE (1903) Caulophacus arcticus (Armauer Hansen) und Calycosoma gracile F.E. Sch. nov. spec. Abh Knigl Akad Wiss Berlin: 1-22 Schulze FE (1904) Hexactinellida. Wiss Ergebn Deu Tiefsee-Exp Dam Valdivia 1898-1899 4: 1-266 Tabachnick KR (1988) Hexactinellid sponges from the mountains of West Pacific. In: Shirshov PP (ed). Structural and functional researches of the marine benthos. Acad Sci USSR, Moscow. pp. 49-64 Tabachnick KR (1999) Abolishment of the family Caulophacidae (Porifera: Hexactinellida). Memoir Queensl Mus 44: 603-605 Tabachnick KR (2002) Family Rossellidae Schulze, 1885. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to classification of sponges. Kluwer Academic/Plenum Publishers, New York. pp. 1441-1502 Tabachnick KR and Lvi C (2004) Lyssacinosa du Pacifique sudouest (Porifera: Hexactinellida). Mem Mus Natl Hist Nat 191: 1171 Topsent E (1910) Les Hexasterophora recueillies par la Scotia dans lAntarctique. Bull Inst ocanogr Monaco 166: 1-18 Topsent E (1913) Spongiaires de lExpdition Antarctique Nationale Ecossaise. Trans r Soc Edinburgh 49(3): 579-643

Porifera research: Biodiversity, innovation and sustainaBility - 2007

467

A new species of Erylus (Geodiidae, Demospongiae) from Brazilian oceanic islands

Fernando Moraes(*), Guilherme Muricy
Departamento de Invertebrados, Museu Nacional - Universidade Federal do Rio de Janeiro. Quinta da Boa Vista, s/no. So Cristvo, Rio de Janeiro - RJ, Brazil 20940-040. fmoraes@mn.ufrj.br, muricy@acd.ufrj.br Abstract: Despite its biogeographical importance, the sponge fauna of Brazilian oceanic islands has been poorly studied. In this study we describe a new species of Erylus Gray (Demospongiae, Astrophorida) from two of the most isolated Brazilian oceanic islands, viz., Trindade Island and So Pedro e So Paulo Archipelago (formerly Saint Pauls Rocks). The new species is characterized by the presence of short-shafted plagiotriaenes, elongate aspidasters, two categories of oxyasters, and one category of strongylasters. This is the first description of a sponge from Trindade Island, and it increases the number of valid species of Erylus in Brazil to seven: E. formosus, E. corneus, E. transiens, E. diminutus, E. toxiformis, E. soesti, and Erylus latens sp. nov. Keywords: Brazil, Erylus latens sp. nov., Geodiidae, oceanic islands, Porifera, taxonomy

Introduction
Due to their isolation and comparatively small size, oceanic islands are interesting areas for taxonomic, ecological, biogeographical, evolutionary, and conservation biology studies, both of terrestrial and of marine infralitoral habitats (e.g., Briggs 1966, MacArthur and Wilson 1967, Case and Cody 1987, Paulay 1994). In Brazil, there are five groups of oceanic islands: Fernando de Noronha Archipelago, Atol das Rocas, So Pedro e So Paulo Archipelago (formerly Saint Pauls Rocks), Trindade Island, and Martin Vaz Archipelago. Despite the great environmental, biogeographic, economic, and strategic importance of these islands, their sponge fauna has received little attention (Hyatt 1877, Carter 1890, Edwards and Lubbock 1983, Mothes and Bastian 1993, Esteves et al. 2002, Moraes et al. 2003, Moraes and Muricy 2003). Recent studies demonstrated that the sponge fauna of these islands is very rich and diverse, with a high percentage of new and endemic species (Moraes et al. 2006). Most of these studies, however, had faunistic or ecological approaches and contain only species lists, not descriptions. Furthermore, such lists contain many species identified only to genus or family levels, making it difficult to estimate precisely endemism rates and biogeographical affinities (e.g. Muricy et al. 2006, Moraes et al. 2006). It is therefore important to describe the sponges from Brazilian oceanic islands, particularly the species new to science, to have a better knowledge of their diversity and biogeography. In an extensive survey of the sponge fauna of Brazilian oceanic islands, Moraes et al. (2006) listed 138 species, including an undescribed species of Erylus Gray, 1867, which is the subject of this study. The genus Erylus is characterized by the combination of ortho- or plagiotriaenes with microscleres including more-or-less flattened sterrasters (aspidasters)

and centrotylote microrhabds together with small euasters (oxyasters, strongylasters, tylasters) in one or more categories. Both inhalant and exhalant orifices are uniporal (Uriz 2002). Several species of Erylus produce compounds with interesting pharmacological activities, such as cytotoxic, antitumoral, antifungal, inhibitory of neuraminidase, thrombin receptor antagonist, and inhibitory of human platelet aggregation in vitro (Carmely et al. 1989, Gulasavita et al. 1994, Stead et al. 2000, Takada et al. 2002, van Altena et al. 2003, Sandler et al. 2005, Okada et al. 2006). Some of these substances also show ecological importance, such as the triterpene glycosides produced by Erylus formosus, which deterred fish predation, microbial attachment, and fouling by invertebrates and algae (Kubanek et al. 2000, 2002). The genus Erylus contains approximately 60 valid species, 17 of which occur in the Atlantic and Caribbean (Adams and Hooper 2001, Mothes and Lerner 2001, Lehnert et al. 2006). So far, six species were described from Brazil: E. formosus Sollas, 1886, E. alleni de Laubenfels, 1934, E. corneus Boury-Esnault, 1973, E. diminutus Mothes et al., 1999, E. toxiformis Mothes and Lerner, 1999, and E. soesti Mothes and Lerner, 2001 (Sollas 1886, 1888, Boury-Esnault 1973, Mothes-de-Moraes 1978, Sol-Cava et al. 1981, Mothes and Bastian 1993, Mothes and Lerner 1999, 2001, Mothes et al. 1999, 2004). The record of E. topsenti von Lendenfeld, 1903 by Mothes-de-Moraes (1981) was synonymyzed with E. soesti by Mothes and Lerner (2001), and that of E. oxyaster von Lendenfeld, 1910 by Mothes-de-Moraes (1978) was synonymyzed with E. diminutus by Mothes et al. (1999). Erylus alleni was considered a junior synonym of E. transiens (Weltner, 1882) by van Soest and Stentoft (1988), but not by Mothes et al. (1999), based on the presence of one versus two size categories of oxyasters. The distinction of microsclere size categories is often very subtle in sponges, and therefore

468

we agree with van Soest and Stentoft (1988) that the two species are synonymous, with priority to the older name E. transiens. In this study, we describe a new species of Erylus from the oceanic islands of Trindade and So Pedro e So Paulo Archipelago, Brazil. The new species increases to seven the number of species of Erylus described from Brazil.

Material and methods Study area

So Pedro e So Paulo Archipelago is located on the So Paulo Fracture Zone (0o55N-29o21W), 1,010 km NE from the city of Natal, Rio Grande do Norte State, NE Brazil (Fig. 1). So Pedro e So Paulo is highly isolated from other shallow areas, lying in the middle of the Atlantic basin, which ranges from 2,000-4,000 m depth. With only 400 m across and 20 m of maximum height, it is one of the smallest isolated archipelagos of the world (Figs. 2A, 3A). In contrast to other Atlantic islands, its origin is plutonic and not volcanic, with ultrabasic rocks resulting from the uplift of the upper mantle (Tilley 1947, Melson et al. 1972). Six sites were sampled, but the new species was found in only two (Figs. 2A, 3B): Cove: a small bay, relatively sheltered, ranging from 318 m depth, with rock and rubble bottoms dominated by the green alga Caulerpa racemosa (see Villaa et al. 2006), the zoanthid Palythoa sp., and sponges; and Vertical Wall of Belmonte Island: a deep vertical wall > 100 m depth with many crevices, on the western side of the archipelago. Trindade Island (20o30S-29o20W) is located at the eastern edge of the VitriaTrindade Chain, 1,140 km E off Vitria, Brazil (Fig. 1). It has an area of approximately 8 km2, with sandy beaches, rocky coasts and tide pools (Castro and Antonello 2006). Nine sites were sampled, but the new species was found in only three (Figs. 2B, 3C, D): Ponta do Paredo: rocky

Fig. 2: So Pedro e So Paulo Archipelago (A) and Trindade Island (B), showing the location of the collection sites: 1, Cove; 2, Vertical Wall of Belmonte Island; 3, Ponta do Paredo; 4, Ilha do Sul; 5, Ponta dos Farrilhes.

vertical wall, with many small caves, 30 m depth; Ilha do Sul: large boulders forming small caves close to the bottom, 25 m depth; and Ponta dos Farrilhes: rocky vertical wall, with small caves close to the bottom, 30 m depth.
Fig. 3: Collection sites and external morphology of Erylus latens sp. nov. A. So Pedro e So Paulo Archipelago, East Shore; B. Cove in So Pedro e So Paulo Archipelago; C-D. Trindade Island, North Shore; E-F. Living specimens of Erylus latens sp. nov. (greyish-brown) from So Pedro e So Paulo Archipelago, partially covered by green and brown algae; G-H. Living specimens of Erylus latens sp. nov. (brown) from Trindade Island, partially covered by other sponges.

Fig. 1: Location of So Pedro e So Paulo Archipelago (SPSPA) and Trindade Island (TI).

469

470

471

Collection and identification

Sponge samples were collected by snorkeling and SCUBA diving, from 0-25 m depth, in four expeditions: October 2000, August 2001 (So Pedro e So Paulo Archipelago), February 2003, and August 2003 (Trindade Island). The specimens were preserved in 70% ethanol and deposited in the Porifera collection of the Museu Nacional, Universidade Federal do Rio de Janeiro, Brazil (MNRJ). In situ photographs were taken with a Nikonos V camera with 35 mm and close-up lenses. Photographs were digitalized using a Nikon Coolscan IV ED scanner. Spicule slides were prepared by dissociation of a small fragment of sponge in boiling nitric acid. Thick sections of the skeleton were observed under light microscope. Depending on spicule abundance, 5-20 spicules of each type were measured per sponge specimen. Measurements are given as minimummean-maximum length x minimum-mean-maximum width (in m). For identification, specimens were compared with the other described species of Erylus in the literature and with museum specimens whenever possible. Other abbreviations used: MNHN, Musum National dHistoire Naturelle, Paris; BMNH, The Natural History Museum, London.

Systematics
Class Demospongiae Sollas, 1885 Order Astrophorida Sollas, 1888 Family Geodiidae Gray, 1867 Genus Erylus Gray, 1867 Definition: Geodiidae with short-shafted triaenes (ortho- or plagiotriaenes); sterrasters usually more-or-less flattened (aspidasters). The somal microsclere is a centrotylote microrhabd. Uniporal inhalant and exhalant orifices (Uriz 2002). Erylus latens sp. nov. (Figs. 3, 4) Synonyms: Erylus cf. formosus sensu Edwards and Lubbock 1983: 63 (non: Erylus formosus Sollas, 1886 and all other authors). Erylus sp. nov., Moraes et al. 2006: 167. Diagnosis: Erylus with short-shafted plagiotriaenes, large diactines ranging from oxeas to strongyles, centrotylote microxeas, elongate aspidasters, two categories of oxyasters, and strongylasters.
Fig. 4: Morphology of Erylus latens sp. nov. A. preserved specimen (Holotype, MNRJ 7397); B. transverse section showing the cortex and the choanosome; C. large diactines; D. short-shafted plagiotriaene; E. aspidaster, strongylaster, and centrotylote microxea; F. detail of aspidaster showing the starshaped spines; G. oxyaster 1; H. strongylaster (left) and oxyaster 2 (right) (E-H, MEV).

Etymology: The species name refers to the algae often overgrowing this species, which makes the specimens difficult to see (from Latin: latens = hidden). Material examined: Trindade Island, Brazil (20o30S29o20W): Holotype MNRJ 7397, 17/VIII/2003, 15 m depth, Ilha Sul, coll. F. Moraes and G. Muricy; Paratypes MNRJ 7375, 17/VIII/2003, 25 m depth, Ponta do Paredo, coll. G. Muricy; MNRJ 7399, 17/VIII/2003, 20 m depth, Ponta do Paredo, coll. G. Muricy; MNRJ 7407, 17/VIII/2003, 17 m depth, Ponta do Paredo, coll. F. Moraes; MNRJ 7409, 17/ VIII/2003, 15 m depth, Ilha Sul, coll. G. Muricy; MNRJ 7419, 18/VIII/2003, 22 m depth, Ponta dos Farrilhes, coll. F. Moraes. So Pedro e So Paulo Archipelago, Brazil (0o55N - 29o21W): Paratypes MNRJ 3572, 26/X/2000, Cove, 3 m depth; MNRJ 3571, 27/X/2000, Vertical Wall of Belmonte Island, 16 m depth; MNRJ 4742, 15/VIII/2001, Cove, 5 m depth MNRJ 4748, 28/VIII/2001, Cove, 13 m depth; all coll. F. Moraes. Comparative material: Erylus corneus Boury-Esnault, 1973: MNHN LBIM-NBE 975 (schyzotype), Brazil, coll. R.V. Calypso; Erylus formosus Sollas, 1886, BMNH 1889.1.1.77, Brazil, coll. H.M.S. Challenger. Description: massive or subspherical sponge, up to 12 x 6 x 8 cm (Figs. 3E-H, 4A). Color brown, greyish brown, or dark grey to almost black externally and beige internally, both in vivo and in 70% ethanol. Surface uneven, rough, microhispid, often covered by algae, hydroids, and other sponges. Oscules circular, 0.5-5.0 mm in diameter, flush, frequently in clusters. Consistency hard, inelastic. Skeleton: cortex dense, 210-392-500 m thick, formed by abundant aspidasters and microxeas vaguely perpendicular to the surface (Fig. 4B). Choanosome with multispicular tracts of large diactines, 100-141-175 m thick, vaguely radial, which expand and become plumose below the cortex (Fig. 4B); the cladomes of the triaenes form a very sparse tangential layer in the subcortical region. Dispersed diactines are common between the tracts. Oxyasters and strongylasters are randomly dispersed in the choanosome; microxeas are restricted to the cortex. Spicules: large diactines ranging from oxeas to strongyles, straight or slightly curved: 282-518-720 x 2-10-17 m (Fig. 4C). Plagiotriaenes (Fig. 4D) short-shafted, rare to absent, with a reduced, conical, often telescopic rhabdome (100-155180 x 5-7-10 m), and sinuous clads which may be unequal or irregularly bifurcated (90-118-150 x 5-9-12 m). Microxeas centrotylote, smooth, with acerate endings: 39-56-70 x 1-3-5 m (Fig. 4E). Aspidasters elongate, with rounded endings and star-shaped spines: 160-222-302 x 20-44-70 m (Fig. 4E, F). Oxyasters 1 rare to absent, with thin, apparently smooth rays, but with very small spines, ray tips blunt or acerate (Fig. 4G): 17-24-40 m in diameter. Oxyasters 2 with a small centrum and spined, acerate rays, with spines larger at the distal end (Fig. 4H): 8-13-25 m in diameter. Strongylasters, with rays spined and with rounded endings: 24-35-50 m in diameter (Fig. 4H). Ecology: Erylus latens sp. nov. was found between 325 m depth, usually on vertical hard substrate, either exposed to light, in crevices, or under shaded overhangs. Several species

472

Table 1: Spicular characteristics of Brazilian Erylus. All measurements are in micrometers.

Characters

E. formosus

E. corneus

E. transiens

E. diminutus

E. toxiformis

E. soesti

E. latens sp. nov.

Triaene type

Rhabdome

Cladome Diactine type 494-680/ 8-19.5 smooth microxea 27.6-57/ 1-3.5 Elliptical 30-71/ 1-7 Elliptical or diskshaped

long-shafted orthotriaenes 180-625/ 9-24 171-446 Oxea

short-shafted orthotriaenes 126-380/ 11.5 238-428 Oxea

short-shafted orthotriaenes 171-665/ 4.6-5.6 119-617 Oxea

short-shafted dichotriaenes 256-304/ 38-57 684-855 Strongyles

dichotriaenes and rare plagiotriaenes 805-1380/ 47-95 713-1058 Oxea

Diactine size

Centrotylote microdiactine

475-989/ 7-28 smooth microstrongyles

437-950/ 4.6-21 smooth microxea, rarely blunt

short-shafted orthotriaenes 199-389/ 11-23 361-636 Oxea with rare strongyles 897-1817/ 9-25 smooth microstrongyles

short-shafted plagiotriaenes 100-180/ 5-10 90-150 Oxeas and strongyles 282-720/ 2-17 smooth microxea 39-56-70/ 1-3-5 Elongate, with rounded endings

Microdiactine size 39-83/ 2.3-4.6 Aspidaster Digitiform

460-920/ 9.5-24 smooth microstrongyles, rare microxeas 39-48-59/ 3.5-6.9 Elliptical or diskshaped, irregular

Aspidaster size

Oxyaster 1 Oxyaster 2 Strongylaster Diactinal aster References

95-305/ 11-55 16-64 7-23 Sollas 1886, BouryEsnault 1973, SolCava et al. 1981, Mothes et al. 1999

119-153/ 69-87 9-23 Boury-Esnault 1973, Mothes et al. 1999

35-145/ 50-114 23-60 7-27.6 de Laubenfels 1934, van Soest and Stentoft 1988, Mothes et al. 1999

159-228/ 105-151 24-54 Mothes-de-Moraes 1978, Mothes et al. 1999

50-97/ 2-7 disk-shaped or elliptical , sterrasterlike 207-506/ 184-414 34-97 73-103 Mothes and Lerner 1999

2093-3220/ 33-57 microspined microstrongyles, rarely smooth 37-76/ 4.6-9.2 Variable, diskshaped to lobate, very irregular 46-128/ 39-92 9-21 (spheroxyaster) Mothes-de-Moraes 1981, Mothes and Lerner 2001

160-302/ 20-70 17-40 8-25 24-50 Present study

473

of algae and other sponges were found on the surface of most specimens studied, making it difficult to locate and identify the specimens in the field (Fig. 3E-H). Distribution: Endemic from Brazil: So Pedro e So Paulo Archipelago and Trindade Island (Moraes et al. 2006).

Discussion
Six valid species of Erylus have been previously described from Brazil: E. formosus, E. corneus, E. transiens (as E. alleni, by Mothes et al. 1999), E. diminutus, E. toxiformis, and E. soesti (Sollas 1886, 1888, Boury-Esnault 1973, Mothesde-Moraes 1978, Sol-Cava et al. 1981, Mothes and Bastian 1993, Mothes and Lerner 1999, 2001, Mothes et al. 1999, 2004). Most of these species are known only from one or a few museum specimens collected by dredging, and sometimes only by fragments (e.g., Erylus toxiformis and E. soesti; Mothes and Lerner 1999, 2001). The only exceptions are E. formosus and E. latens sp. nov., which have been collected through SCUBA diving (Sol-Cava et al. 1981; present study). External morphological characters, particularly color in vivo and oscular characteristics, are therefore of little usefulness to discriminate among Brazilian species. The skeletal architecture of the choanosome is similar in all species, with radial bundles or isolated oxeas and triaenes whose cladomes form a tangential subcortical layer; microscleres are randomly dispersed between the megasclere bundles. The ectosome is always a cortex of microscleres, with centrotylote microrhabds in the external layer and densely packed aspidasters in the internal layer, but the orientation of the microrhabds varies from tangential to oblique or perpendicular, thus representing a good taxonomic character. Spicule composition and details of their ornamentation are however the best characters to identify Brazilian species of Erylus (Table 1). Erylus latens sp. nov. shares the short-shafted triaenes with E. corneus, E. transiens and E. toxiformis, and the elongate aspidasters with E. formosus. Erylus corneus however has orthotriaenes instead of plagiotriaenes, elliptical aspidasters, only one category of oxyasters, and no strongylasters (BouryEsnault 1973, Mothes et al. 1999). Erylus transiens also has no strongylasters, but it has only one or two categories of oxyasters; furthermore, its aspidasters are disk-shaped and some specimens have dichotriaenes in variable abundance in addition to the short-shafted orthotriaenes (Weltner 1882, de Laubenfels 1934, van Soest and Stentoft 1988, Mothes et al. 1999). The record of E. transiens from Azores (Topsent 1892) probably belongs to a different species: its shape is pedunculate, ramose; its tetraxons are exclusively dichotriaenes (short-shafted plagio- or orthotriaenes are absent); and its microstrongyles are shorter (up to 23 m) and rarely centrotylote. Erylus toxiformis differs from E. latens sp. nov. by having orthotriaenes instead of plagiotriaenes, sterraster-like aspidasters, and the peculiar toxiform asters diagnostic of the species (Mothes and Lerner 1999). Erylus formosus can be easily distinguished from the new species by its large, long-shafted orthotriaenes, smaller strongylasters, and by the presence of a single large oxyaster category, as opposed to smaller short-shafted plagiotriaenes, larger strongylasters, and two smaller categories of oxyasters in the new species (Table 1). Edwards and Lubbock (1983)

recorded Erylus cf. formosus from So Pedro e So Paulo Archipelago; although the specimens studied by Edwards and Lubbock (1983) were not reexamined, extensive collections in the archipelago failed to find Erylus formosus, reveiling instead a relatively great abundance of Erylus latens sp. nov. The record of Edwards and Lubbock (1983) of E. formosus is thus here synonymyzed with the new species. Erylus latens sp. nov. differs from all other Brazilian species of the genus by its short-shafted plagiotriaenes together with three categories of asters (two of oxyasters and one of strongylasters). Nine species of Erylus were recorded from the Caribbean (Pulitzer-Finali 1986), of which only four have shortshafted plagiotriaenes: Erylus transiens, E. ministrongilus Hechtel, 1965, E. clavatus Pulitzer-Finali, 1986 (probably a junior synonym of E. formosus; cf. van Soest et al. 2005), and E. trisphaera (de Laubenfels, 1953, as Unimia). Erylus ministrongilus differs from the new species by the smaller and thinner aspidasters and by the absence of strongylasters. Erylus clavatus has larger oxeas (930-1230/14-28 m), tylasters, and its calthrops are orthotriaenes instead of plagiotriaenes (Pulitzer-Finali 1986). Erylus trisphaera has exclusive trilobate aspidasters. Other three species of Erylus are known from the Atlantic: E. granularis Topsent, 1904, E. expletus Topsent, 1927, and E. pappilatus Topsent, 1928 (see also Adams and Hooper 2001); all of them differ from Erylus latens sp. nov. by their oval or rounded aspidasters. Brazilian species of Erylus were collected mostly by dredging along the continental shelf (E. formosus, E. corneus, E. transiens, E. diminutus, and E. toxiformis Boury-Esnault 1973, Mothes and Lerner 1999, Mothes et al. 1999, 2003, 2004) and slope (E. soesti Mothes and Lerner 2001), with only E. formosus also occuring in littoral areas (Sol-Cava et al. 1981) and in oceanic islands such as Fernando de Noronha (Mothes and Bastian 1993, Muricy and Moraes 1998) and Atol das Rocas (Moraes et al. 2003). The new species is only the second species of Erylus described from Brazilian oceanic islands, and the first description of a sponge from Trindade Island. Beyond the seven species of Erylus described so far from Brazil, at least six other records, as yet unidentified and undescribed, are also known (Erylus spp. 1-5 in Muricy et al. 2006, Erylus sp. 1 in Moraes et al. 2006). The diversity of the genus Erylus in Brazil is therefore probably greater than the current estimations.
Key to Brazilian species of Erylus (modified from Mothes and Lerner 2001) 1A. Tetractinal megascleres include dichotriaenes ..........................2 1B. Tetractinal megascleres include only orthotriaenes or plagiotriaenes, dichotriaenes absent ............................................4 2A. Tetractinal megascleres dichotriaenes only; ortho- and plagiotriaenes absent ...................................................................3 2B. Dichotriaenes, when present, occur together with plagiotriaenes; aspidasters disk-shaped, oxyasters in one or two recognizable size categories ................ E. transiens 3A. Dichotriaenes with short rhabdome (256-304 m long); strongyles varying to strongyloxeas (460920 m long); aspidasters flattened, with slightly irregular outline (159-229 m long) ......................... E. diminutus

474

3B. Dichotriaenes with long rhabdome (8051380 m long); oxeas long (2093-3220 m long); aspidasters not flattened, with strongly irregular outline (46-129 m long) .................................. E. soesti 4A. Aspidasters atypical, sterraster-like, diskshaped, not flattened (207-506 m long); reduced toxa-like oxyasters present (74-104 m long) ..........E. toxiformis 4B. Aspidasters typical, flattened; toxa-like oxyasters absent ...........................................................................5 5A. Aspidasters elliptical or disk-shaped .........................................6 5B. Aspidasters finger-shaped or elongate (95305 m long) ...............................................................................7 6A. Oxyasters in a single category (9-23 m) ................. E. corneus 6B. Oxyasters in two size categories (23-57 and 8-27 m) .................................................................... E. transiens 7A. Long-shafted orthotriaenes; oxyasters and strongylasters in one size category ............................ E. formosus 7B. Short-shafted plagiotriaenes; oxyasters in two shape categories and strongylasters in one size category ........................................... E. latens sp. nov.

Acknowledgements
We thank Diogo Pagnoncelli, Brbara Rustum Andra, Bertran Feitoza, Zaira Matheus and Claudio Moraes for laboratory and/ or field help. We also thank Shirley Stone (The Natural History Museum, London) and Dr. Claude Lvi (Musm National dHistoire Naturelle, Paris) for the kind loan of specimens for comparison. We are grateful to Dr. Marcia Attias and Nomia Rodrigues (Laboratrio de Ultraestrutura Celular Hertha Meyer, Instituto de Biofsica Carlos Chagas Filho, UFRJ) for their help in the use of SEM. The comments of two anonymous reviewers greatly improved the manuscript. Special thanks to Rob van Soest and Beatriz Mothes for the kind help to obtain relevant literature. Fundao O Boticrio de Proteo Natureza, Conselho Nacional de Desenvolvimento Cientfico e Tecnolgico (CNPq), Fundao Carlos Chagas Filho de Apoio Pesquisa do Estado do Rio de Janeiro (FAPERJ), Secretaria da Comisso Interministerial para os Recursos do Mar (SECIRM), and Marinha do Brasil provided financial and logistic support.

References
Adams CL, Hooper JNA (2001) A revision of Australian Erylus (Porifera: Demospongiae: Astrophorida: Geodiidae) with a tabular review of worldwide species. Invert Taxon 15: 319-340 Boury-Esnault N (1973) Campagne de la Calypso au large des ctes atlantiques de lAmrique du Sud (1961-1962). Spongiaires. Rs Sci Camp Calypso, Paris 10: 263-295 Briggs J (1966) Oceanic islands, endemism, and marine paleotemperatures. Syst Zool 15: 153-163 Carmely S, Roll M, Loya Y, Kashman Y (1989) The structure of eryloside-A, a new antitumor and antifungal 4-methylated steroidal glycoside. J Nat Prod 52(1): 161-170 Carter HJ (1890) Porifera Notes on the zoology of Fernando de Noronha. J Linn Soc 20: 564-569 Case TJ, Cody ML (1987) Testing theories of island biogeography. Am Sci 75(4): 402-411

Castro JWA, Antonello LL (2006) Geologia das ilhas ocenicas brasileiras. In: Alves RJV, Castro JWA (eds). Ilhas ocenicas brasileiras: da pesquisa ao manejo. Ministrio do Meio Ambiente, Braslia. pp. 27-57 de Laubenfels MW (1934) New sponges from Puerto Rican Deep. Smithson Misc Coll 91(17): 1-28 de Laubenfels MW (1953) Sponges from the gulf of Mexico. Bull Mar Sci Gulf Carib 2(3): 511-557 Edwards A, Lubbock R (1983) Marine zoogeography of St. Pauls Rocks. J Biogeogr 10: 65-72 Esteves E, Moraes F, Muricy G, Amaral F (2002) Duas novas ocorrncias da ordem Hadromerida (Porifera, Demospongiae) para o Arquiplago de So Pedro e So Paulo, Brasil. Bol Mus Nac, Zool 488: 1-12 Gray JE (1867) Notes on the arrangement of sponges, with descriptions of some new genera. Proc Zool Soc London 2: 492558. Gulavita NK, Wright AE, Kelly-Borges M, Longley RE, Yarwood D, Sills MA (1994) Eryloside E from an Atlantic sponge Erylus goffrilleri. Tetrahedr Lett 35(25): 4299-4302 Hechtel GJ (1965) A systematic study of the Demospongiae of Port Royal, Jamaica. Bull Peabody Mus Nat Hist 20: 1-103 Hyatt A (1877) Revision of the North American Porifera, with remarks upon foreign species. Part II. Mem Boston Soc Nat Hist 2(5): 481-554 Kubanek J, Pawlik JR, Eve1 TM, Fenical W (2000) Triterpene glycosides defend the Caribbean reef sponge Erylus formosus from predatory fishes. Mar Ecol Progr Ser 207: 69-77 Kubanek J, Whalen KE, Engel S, Kelly SR, Henkel TP, Fenical W, Pawlik JR (2002) Multiple defensive roles for triterpene glycosides from two Caribbean sponges. Oecologia (Berlin) 131(1): 125-136 Lehnert H, Stone R, Heimler W (2006) Erylus aleuticus sp. nov. (Porifera: Demospongiae: Astrophorida: Geodiidae) from the Aleutian Islands, Alaska, USA. J Mar Biol Assoc UK 86: 971-975 MacArthur RH, Wilson EO (1967) The theory of island biogeography. Princeton University Press, Princeton. Melson WG, Hart SR, Thompson G (1972) St. Pauls Rocks, Equatorial Atlantic: petrogenesis, radiometric ages, and implications on sea-floor spreading. Mem Geol Soc Am 132: 241272 Moraes FC, Muricy G (2003) Taxonomy of Plakortis and Plakinastrella from oceanic islands off north-eastern Brazil, with description of three new species. J Mar Biol Assoc UK 83: 385397 Moraes FC, Ventura M, Klautau M, Hajdu E, Muricy G (2006) Biodiversidade de esponjas das ilhas ocenicas brasileiras. In: Alves RJV, Castro, JWA (eds). Ilhas ocenicas brasileiras da pesquisa ao manejo. Ministrio do Meio Ambiente, Braslia. pp. 147-178 Moraes FC, Vilanova EP, Muricy G (2003) Distribuio das esponjas (Porifera) na Reserva Biolgica do Atol das Rocas, Nordeste do Brasil. Arq Mus Nac, Rio de Janeiro 61(1): 13-22 Mothes B, Bastian MCKA (1993) Esponjas do Arquiplago de Fernando de Noronha, Brasil (Porifera, Demospongiae). Iheringia, Zool 75: 15-31 Mothes B, Captoli RR, Lerner C, Campos MA (2004) Filo Porifera Regio Sul. In: Amaral CZA, Rossi-Wongtschowski CLB (eds). Biodiversidade bentnica da regio sudeste-sul do Brasil plataforma externa e talude superior. Instituto Oceanogrfico USP (srie Documentos REVIZEE score sul), So Paulo. Pp. 57-63

475

Mothes B, Lerner C (1999) Erylus toxiformis (Porifera, Geodiidae), a new species from the Southwestern Atlantic. Beaufortia 49(4): 29-33 Mothes B, Lerner C (2001) A new species of Erylus Gray, 1867 (Porifera, Geodiidae) from the Southeastern coast of Brazil. Beaufortia 51(4): 83-89 Mothes B, Lerner C, Silva CM (2003) Guia ilustrado das esponjas marinhas da costa sul brasileira. USEB, Porto Alegre Mothes B, Lerner CB, Silva CMM (1999). Revision of Brazilian Erylus (Porifera: Asthrophorida: Demospongiae) with description of a new species. Memoir Queensl Mus 44: 369-380 Mothes-de-Moraes B (1978) Esponjas tetraxonidas do litoral sulbrasileiro: II- Material coletado pelo N/Oc. Prof. W. Besnard durante o Programa Rio Grande do Sul. Bolm Inst Oceanogr So Paulo 27(2): 57-78 Mothes-de-Moraes B (1981) Ocorrncia de Erylus topsenti Lendenfeld, 1903 na costa do Rio de Janeiro (Porifera, Demospongiae). Iheringia, Zool 57: 105-111 Muricy G, Moraes FC (1998) Marine sponges of Pernambuco State, NE Brazil. Rev Bras Oceanogr 46(2): 213-217 Muricy G, Santos CP, Batista D, Lopes DA, Pagnoncelli D, Monteiro LC, Oliveira MV, Moreira MCF, Carvalho MS, Melo M, Klautau M, Rodriguez PRD, Costa RN, Silvano RG, Schwientek S, Ribeiro SM, Pinheiro US, Hajdu E (2006) Captulo 3. Porifera. In: Lavrado HP, Ignacio BL (eds). Biodiversidade bentnica da regio central da Zona Econmica Exclusiva brasileira. Srie Livros 18, Museu Nacional, Rio de Janeiro. Pp. 109-145 Okada Y, Matsunaga S, van Soest RWM, Fusetani N (2006) Sokodosides, steroid glycosides with an isopropyl side chain, from the marine sponge Erylus placenta. J Org Chem 71(13): 48844888 Paulay G (1994) Biodiversity on oceanic islands: its origin and extinction. Am Zool 34: 134-144 Pulitzer-Finali G (1986) A collection of West Indian Demospongiae (Porifera). In appendix, a list of the Demospongiae hitherto recorded from the West Indies. Ann Mus Civ Stor Nat Genova 86: 1-216 Sandler JS, Forsburg SL, Faulkner J (2005) Bioactive steroidal glycosides from the marine sponge Erylus lendenfeldi. Tetrahedron 61(5): 1199-1206 Sol-Cava AM, Kelecom A, Kannengiesser GJ (1981) Study of some sponges (Porifera, Demospongiae) from the infralitoral of Guarapari, Esprito Santo, Brazil. Iheringia, Zool 60: 125-150 Sollas WJ (1885) A classification of the sponges. Ann Mag Nat Hist 5(16): 395 Sollas WJ (1886) Preliminary account of the Tetractinellid sponges dredged by the H.M.S. Challenger during the years 1873-1876. Part I. The Choristida. Sci Proc Roy Dublin Soc 5: 177-199

Sollas WJ (1888) Report on the Tetractinellida collected by H.M.S. Challenger during the years 1873-1876. Rep Sci Res Voy H.M.S Challenger, Zool 25(63): 1-458 Stead P, Hiscox S, Robinson PS, Pike NB, Sidebottom PJ, Roberts AD, Taylor NL, Wright AE, Pomponi SA, Langley D (2000) Eryloside F, a novel pentasterol disaccharide possessing potent thrombin receptor antagonist activity. Bioorg Med Chem Lett 10(7): 661-664 Takada K, Nakao Y, Matsunaga S, van Soest RWM, Fusetani N (2002) Nobiloside, a new neuraminidase inhibitory triterpenoidal saponin from the marine sponge Erylus nobilis. J Nat Prod 65(3): 411-413 Tilley CE (1947) The Dunite-Milonytes of St. Pauls Rocks (Atlantic). Am J Sci 245(8): 483-491 Topsent E (1892) Contribution ltude des spongiaires de lAtlantique Nord. Rs Camp Sci Accomp Prince Albert 1er de Monaco 2: 1-165 Topsent E (1904) Spongiaires des Aores. Rs Camp Sci Accomp Prince Albert 1er de Monaco 25: 1-218 Topsent E (1927) Diagnoses dponges nouvelles recueillies par le Prince Albert 1er de Monaco. Bull Inst Ocanogr Monaco 502: 119 Topsent E (1928) Spongiaires de lAtlantique e de la Mediterrane. Rs Camp Sci Accomp Prince Albert 1er de Monaco 74: 1-376 Uriz MJ (2002) Family Geodiidae Gray, 1867. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges. Kluwer Academic/Plenum Publishers, New York. pp. 134-140 van Altena I, van Soest RWM, Roberge M, Andersen RJ (2003) Trisphaerolide A, a novel polyketide from the Dominican sponge Erylus trisphaerus. J Nat Prod 66(4): 561-563 van Soest RWM, Boury-Esnault N, Janussen D, Hooper J (2005) World Porifera Database. http://www.marinespecies.org/porifera/. Accessed on 2007-06-19 van Soest RWM, Stentoft N (1988) Barbabos deep-water sponges. Stud Fauna Curaao Caribb Isl 70: 1-175 Villaa RC, Pedrini AG, Pereira SMG, Figueiredo MAO (2006) Flora marinha bentnica das ilhas ocenicas brasileiras. In: Alves RJV, Castro JWA (eds). Ilhas ocenicas brasileiras da pesquisa ao manejo. Ministrio do Meio Ambiente, Braslia. pp. 105-146 von Lendenfeld R (1903) Porifera. Tetraxonia. Das Tierreich 19: 1168 von Lendenfeld R (1910) The sponges. 2. The Erylidae. Report on the scientific results of the expedition to the Eastern Tropical Pacific carried out by Alexander Agassiz on the US Fisheries Comission Steamer Albatross in 1904-1905. Mem Mus Comp Zool Harvard Coll 41(2): 261-324 Weltner W (1882) Beitrge zur Kenntniss der Spongien. Inaugural Dissertation, Freiburg

Porifera research: Biodiversity, innovation and sustainaBility - 2007

477

A new species of Characella (Demospongiae, Astrophorida, Pachastrellidae) from the south Brazilian continental shelf
Beatriz Mothes(1*), Manuel Maldonado(2), Rafael Eckert(1), Cla Lerner(1), Maurcio Campos(3), Joo Lus Carraro(3)
Museu de Cincias Naturais, Fundao Zoobotnica do Rio Grande do Sul. Rua Salvador Frana 1427, 90690-000 Porto Alegre-RS, Brazil. bmothes@fzb.rs.gov.br, rafael_eckert@hotmail.com, cblerner@fzb.rs.gov.br (2) Centro de Estudios Avanzados de Blanes (CSIC). Acceso Cala St. Francesc 14, Blanes 17300, Girona, Spain. maldonado@ceab.csic.es (3) Programa de Ps-Graduao, Universidade Federal do Rio Grande do Sul. Av. Bento Gonalves 9500, 91501-970 Porto Alegre-RS, Brazil. mrcpoa@hotmail.com, jlc.fzb@terra.com.br
(1)

Abstract: This work deals with the taxonomic description of Characella capitolii sp. nov., providing the second record of the genus Characella Sollas, 1886 for the southern Brazilian coast (Rio Grande do Sul State; 170-173 m depth). The new species is characterized by two types of oxeas, one being large and abundant, the other being small and rare; by having all tetraxons transformed into three-rayed forms; spiny microxeas in two sizes; and streptasters with straight axis (plesiasters and metasters transitional to amphiasters). Keywords: Pachastrellidae, South Brazilian coast, Taxonomy, Characella capitolii sp. nov.

Introduction
The genus Characella Sollas, 1886 was erected for a large massive sponge collected from deep Brazilian waters (640 m) by the Challenger Expedition and described as Characella aspera Sollas, 1886. Ever since no other species of Characella has formally been reported from Brazilian waters. Likewise, the current knowledge of the family Pachastrellidae, in which Characella is currently contained (Maldonado 2002), indicates that very few representatives of this group are known from the Brazilian coast and slope. The previous knowledge of the family in the area can be summarized (in chronological order) as it follows: 1) Characella aspera Sollas, 1886 from Bahia State (Sollas 1888); 2) P. monilifera Schmidt, 1868 from Rio Grande do Sul State (Mothes-de-Moraes 1978); 3) Pachastrella monilifera, Poecillastra sollasi Topsent, 1904, and Vulcanella sp. from So Paulo State (Hajdu et al. 2004); 4) Unidentified Pachastrellidae from Rio Grande do Sul State (Mothes et al. 2004); 5) Pachataxa lutea PulitzerFinali, 1986 from Atol das Rocas (Moraes et al. 2006); 6) Stoeba sp. and a unidentified Pachastrellidae from Esprito Santo and Rio de Janeiro State (Muricy et al. 2006). Because most of the above records belong to relatively recent work, we suspect that pachastrellid sponges may be more common in Brazilian waters than it is suggested by the available literature. Subsequent explorations of the deep shelf and upper slope,

the preferred habitat of this group, are expected to bring to light more new species of this demosponge family. In this study we report on a new species of the genus Characella collected off the coast of Rio Grande do Sul State (Brazilian Atlantic coast) by the oceanographic program REVIZEE Score Sul.

Material and methods

A single individual of the new species was collected by trawling on the continental shelf (170-173 m), at 310886 S - 493204 W, off the coast of Rio Grande do Sul State (Fig. 1), during the Federal Government Oceanographic cruise Programa Recursos Vivos da Zona Econmica Exclusiva (REVIZEE), by R/V Atlntico Sul, in 2001. General information on the sponge fauna and geomorphological features of the studied area can be found elsewhere (Silva and Mothes 2000, Mothes et al. 2004). The studied material was fixed in formalin, then preserved in 96 GL alcohol, and stored in the Museu de Cincias Naturais - Porifera Collection (MCNPOR) Fundao Zoobotnica do Rio Grande do Sul, Porto Alegre, RS, Brazil. The light microscocopy study of the skeletal features followed the standard procedures and methodology described elsewhere (e.g., Mothes-de-Moraes 1978, Mothes et al. 2004). The Scanning Electron Microscopy (SEM) study followed the procedures outlined in Silva and Mothes (1996). Spicule size

478

was described by minimum, mean and maximum values of lenght and (/) width, and given in microns.

Systematic description
Class Demospongiae Sollas, 1885 Order Astrophorida Sollas, 1888 Family Pachastrellidae Carter, 1875 Genus Characella Sollas, 1886 Definition: Pachastrellidae whose megascleres consist of abundant oxeas and scarce calthrops (and/or short-shafted triaenes), mostly restricted to subectosomal regions. Microscleres are spiny or smooth microxeas-microstrongyles in at least two size categories and streptasters with a straight central axis (never spirasters); streptasters may be very scarce in some species. Anatriaene with cladomes that protrude the sponge surface and blunt rhabdomes embedded in the choanosome occur in some species (Maldonado 2002). Characella capitolii sp. nov. (Figs. 2A-H, 3, 4, 5A-C) Material studied: Holotype, Brazil, off Rio Grande do Sul State, 310886 S - 493204 W, MCNPOR 6926, 170173 m depth, REVIZEE - Sul, R/V Atlntico Sul leg., 02.XI.2001. Distribution: Rio Grande do Sul State (only known from the holotype description). Diagnosis: Characella capitolii sp. nov. is characterized by the presence of two types of megascleric oxeas, the largest ones being relatively short and slender when compared with the stout, conical oxeas of the remaining members of the genus; the smallest being hastate oxeas completely atypical for the genus. It is also remarkable the fact that all short-shafted triaenes are consistentely reduced to three-rayed forms. Streptasters are metasters and plesiasters, a combination atypical among the remaining Characella species. Macroscopic features (Fig. 2A): Small massive sponge (2.5 x 1.5 x 1.5 cm), being probably a young individual. Irregular, hispid surface, forming some ridges and folds. A single oscule (0.1 cm in diameter) occurred, being slightly elliptical. Small openings, possibly ostia (<0.1 cm in diameter), occurred sparsedly at some areas over the sponge surface. In spirit, the external colour was light beige, being the choanosome withish. The sponge consistency was hard but friable, as typically in the genus. Skeleton (Fig. 2B): The ectosomal skeleton consisted of a feltwork of microxeas, streptaster, and small oxeas. This pseudocortex was internally supported by large oxeas in confuse arrangement and triactinal calthrops, which may protrude the sponge surface. In deeper regions of the sponge, the skeletal organization was similar, characterized by triactinal calthrops irregularly scattered and large oxeas arranged more or less radially through the body. Microxeas were also abundant in the choanosome, along with streptasters.

Fig. 1: Location of the holotype collection ().

Spicules: 1) Oxeas I: not very abundant, though well represented in the slides; mostly occurring through the choanosome. They were relatively slender, hardly fusiform and nearly isodiametric, slightly curved in the middle or somewhat sinuous, with one or both blunt ends (Fig. 2C), measuring 700-1166.7-2400 / 7.5-8.3-20.8 m (n=12). 2) Oxeas II: mostly located at the pseudocortex. They were rare in the slides, being slightly curved in the middle, with acerate ends. These ectosomal spicules were very similar to the hastate oxeas of the genus Haliclona (Fig. 2D). They measured 230274.4-310 / 5.0-9.5-12.5 m (n=25). 3) Triactinal calthrops: (Fig. 2E) abundant in the slides, occurring in a large size range, with actines measuring 123.5-350.1-598.5 / 9.2-20.5-36.8 m (n=150). They never showed tetraxonid appearance, since one of the actines was consistently underdeveloped and reduced to a small protuberance, which became more obvious for the small spicules (Fig. 2F). In most cases, the remaining three actines were similar in dimensions and showed equiangular distribution. Nevertheless, in very few cases, the actines were also noticed to be bifurcated (Fig. 2G). Some spicules showed two and even three of their actines malformed and reduced to protuberances (Fig. 2H). 4) Microxeas (Fig. 3) were very abundant in the slides. They appeared to be distributed in two size categories connected by some transitional spicules (graph Fig. 4). The small category was characterized by uniformly microspined spicules, curved or centroangulated, rarely sinuous, with acerate ends. They were markedly fusiform, in many cases being centrotylote and more rarely with swellings occurring along the shaft, typically measuring 27.6.-69.280 m. The large category was characterized by uniformly microspined spicules, softly curved, never centroangulated,
Fig. 2: Poecillastra capitolii sp. nov.: A. Holotype MCNPOR 6926. B. Transverse section of the body showing the skeletal architecture. C. Choanosomal oxeas. D. Ectosomal hastate oxeas, along with the representatives of the two categories of microxeas. E. SEM of a triactinal calthrop. F. Small calthrop with an underdeveloped actine reduced to a small protuberance. G-H. Malformed calthrop showing a malformed and bifurcated actines.

479

480

Fig. 4: Length frequency distribution (counts; N= 320 spicules; size interval = 5 m) of microspiny microxeas and smooth hastate oxeas. Microxeas distribute in two size populations connected by some transitional spicules occurring at low abundance. Hastate oxeas occur in a relatively wider size range and are uncommon when compared to the microspiny microxeas.

Fig. 3: SEM micrographs showing shape and microornamentation of microxeas in both the small and large size categories, as well as in between-category intermediate stages.

hardly fusiform, very rarely centrotylote, typically measuring from 90-109.3-170 m. 5) Streptasters were moderately abundant in the slides, displaying two morphological types, pleasiasters and metasters, both characterized by a straight central axis and microspiny actines (Fig. 5A-C). Plesiasters were a bit larger (17 to 25 m in total diameter) than metasters, having a very short axis with few (4 to 7) relatively long (up to 10 m) actines (Fig. 5A, C). Metasters measured 12 to 16 m in maximum total diameter, having 7 to 11 actines, thinner and shorter (5 to 8 m) than those of the plesiasters (Fig. 5B, C). Because many actines are inserted on a short, straight central axis, the metaster axis often looked bumpy. Both plesiasters and metasters showed occasional forms that were transitional to amphiasters, i.e., with the central portion of the axis lacking actines (Fig. 5C). Etymology: This species is named after Ricardo Captoli, who has enriched the Porifera Collection (MCNPOR) of the Museu de Cincias Naturais with relevant material collected from the continental shelf and slope off the coast of the Rio Grande do Sul State.

Discussion
The general skeletal features of the new species suggest assignation to the genus Characella. This taxonomic allocation was largely based on the shape and size of the two categories of microspiny microxeas in combination with the shape and moderated abundance of the two categories

of streptasters, a set of features strongly similar to those described from other Characella species (e.g., Maldonado 2002). Occurrence of at least two categories of microxeas in combination with small streptasters characterized by a short, central axis is a key diagnostic character for Characella. In the studied specimen, we found some transitional stages between the two categories of microxeas (Fig. 3), which at first sight could suggest that there was only one size category rather than two. Nevertheless, the probability function of spicule size rendered by a higher number of measurements (N= 320) revealed two major populations of microxeas despite occurrence of some transitional spicules (Fig. 4). Indeed, such scarce transitional spicules between the size categories of microxeas also occur in other species of Characella (see Maldonado 2002), including the type species Characella aspera and C. connectens (Schmidt, 1870). The metasters of the studied specimen were very similar in shape and size to those described in other Characella. By contrast, the small accompanying pleasiasters were a characteristic trait of the new species. The plesiaster is a streptaster type often found in species of the sister genus Poecillastra Sollas, 1888. Nevertheless, the plesiasters of the new species, which were comparatively small -up to 24 m in maximum diameter- and characterized by 4 to 7 actines, are clearly distinguishable from those found in Poecillastra species, which are typically larger (often larger than 30 m) and usually have 2 to 5 actines only. In addition, the newly described sponge could not be allocated into Poecillastra, since it lacks streptasters with twisted axis (i.e., spirasters), one of the diagnostic characters of the genus Poecillastra (Maldonado 2002). The new species Characella capitolii is clearly distinguishable from previously known Characella species by not only its small pleasiasters, but also its small megascleric

481

Fig. 5: Streptasters. A. SEM micrograph of a plesiaster. B. SEM micrograph of a metaster. C. Light microscopy micrographs showing comparative views of metasters and streptasters.

oxeas (230-274.4-310 / 5.0-9.5-12.5 m) and its tetraxons consistently reduced to triactinal forms. To our knowledge, the small ectosomal oxeas of hastate appearance occurring in C. capitolii have never been reported in other Characella species. Nevertheless, they are known to occur in some members of at least other pachastrellid genus, Pachastrella Schmidt, 1868 (e.g., Maldonado 1996, 2002). These small oxeas of C. capitolii have tentatively been interpreted as megascleres because they are smooth (i.e. lack microornamentation) and can grow to a thickness (5-12 m) that is unconventional for pachastrellid microxeas. However, we cannot discard the possibility that they are a third category of microxeas. It should be kept in mind that Characella aspera, the type species of the genus and also from Brazilian waters, is characterized by two categories of smooth microxeas, and that those in the largest category are of length (150300 m) similar to that of the putative megascleric oxeas of the new species, but thinner (4-5 m). It is unlikely that these hastate oxeas, which are very similar to those of many haplosclerid demosponges, are exogoneous to the sponges. They certainly occur in low abundance compared to the microspiny microxeas (Fig. 3, 4), but are too well represented to be contaminating spicules. A definitive interpretation on the taxonomic status and value of these hastate oxeas may require future descriptions of new specimens from different locations. An additional striking feature of the studied sponge is that all their tetraxons spicules had at least one of their actines reduced to a small protuberance. Occurrence of three-rayed spicules by development of a dwarf actine rather than a full grown actine is not uncommon in some species of the genus Characella, such as C. connectens. Yet, C. capitolii appear to be the only species in the genus in which all tetraxons are affected by such a phenomenon. At this stage, it is impossible to ascertain whether such an actine reduction was a malformation caused in the studied individual by ecological or physiological factors or is a feature genetically fixed and representative of this new species. Among the pachastrellids, occurrence of some calthrops and short-shafted triaenes with aberrant and reduced actines is common in some species of the genera both Characella and Poecillastra Sollas, 1888, but consistent reduction of all tetraxons to triactinal and diactinal forms was only previously reported in Ancorella paulini von Lendenfeld, 1906, from the Pacific coast of Chile. Despite sharing the complete reduction of tetraxons to triactinal forms, the newly described material could never be classified into the genus Ancorella von Lendenfeld, 1906, since the current diagnosis of this monotypic genus is defined to include pachastrellids lacking streptasters (Maldonado 2002). The newly described species Characella capitolii provides the second record of the genus in Brazilian waters, which was only previously known from the description of the type species Characella aspera. In this regard, there is an additional confusing report indicating occurrence of pachastrellid material collected from So Paulo State that could be related to the genus Characella. It has never been described skeletally, but just mentioned under the name Poecillastra sollasi Topsent, 1904 in a check list elaborated by Hadju et al. (2004). Several problems arise from such a report. First of all, the authority of the species name is mistaken, since

482

the original description of the species, as Characella sollasi, was due to Topsent, but in 1892 rather than in 1904. More importantly, Topsent himself acknowledged in subsequent revisions of the Characella sollasi type that the material he used to erect the species C. sollasi was conspecific with that of Characella pachastrelloides (Carter, 1876) and that, consequently, C. sollasi was not longer a valid species, but a junior synonym of C. pachastrelloides (see, Topsent 1902, Topsent 1904). Therefore, the definitive taxonomic assignation of the material referred to as Poecillastra sollasi by Hadju and co-workers (2004) is pending on further examination. If it is finally proved to belong to C. pachastrelloides, it will make the third record of the genus in this biogeographical region.

Acknowledgments
The authors are grateful to Dr. Ricardo Captoli (Fundao Universidade Federal do Rio Grande) for donation of the studied specimen, and Dr. Francisco Jos Kiss (Universidade Federal do Rio Grande do Sul) for SEM photographs. The authors thanks CNPq, CAPES and FAPERGS (Brazil) and the Spanish Ministry of Science and Education (MEC- CTM2005-05366/MAR) for financial support.

References
Carter HJ (1876) Descriptions and figures of deep-sea sponges and their spicules, from the Atlantic Ocean, dredged up on board H.M.S. Porcupine, chiefly in 1869 (concluded). Ann Mag Nat Hist (4) 18(105): 226-240; (106):307-324; (107):388-410; (108):458-479 Hajdu B, Santos CP, Lopes DA, Oliveira MV, Moreira MCF, Carvalho MS, Klautau M (2004) Filo Porifera. In: Amaral ACZ, Rossi-Wongtschowski CLDB (eds). Biodiversidade bentnica da regio sudeste-sul do Brasil - Plataforma externa e talude superior. Instituto Oceanogrfico - USP, So Paulo. pp. 49-56 Maldonado M (1996) On the presence of anatrienes in Pachastrellidae (Porifera, Demospongiae): evidence for a new phylogenetic family concept. J Nat Hist 30: 389-405

Maldonado M (2002) Family Pachastrellidae. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges. Kluwer Academic/Plenum Publishers, New York. pp. 141-162 Moraes F, Ventura M, Klautau M, Hajdu E, Muricy G (2006) Biodiversidade de esponjas das ilhas ocenicas brasileiras. In: Alves RJV, Castro JWA (eds). Ilhas Ocenicas Brasileiras - da Pesquisa ao Manejo. MMA - SBF, Braslia. pp. 147-177 Mothes-de-Moraes B (1978) Esponjas tetraxonidas do litoral sulbrasileiro: II. Material coletado pelo N/Oc. Prof. W. Besnard durante o Programa RS. Bol Inst Oceanogr 27(2): 57-78 Mothes B, Captoli RR, Lerner C, Campos MA (2004) Filo Porifera - Regio Sul. In: Amaral ACZ, Rossi-Wongtschowski CLDB (eds). Biodiversidade bentnica da regio sudeste-sul do Brasil - Plataforma externa e talude superior. Instituto Oceanogrfico USP, So Paulo. pp. 57-63 Muricy G, Santos CP, Batista D, Lopes DA, Pagnoncelli D, Monteiro LC, Oliveira MV, Carvalho MS, Melo M, Moreira MCF, Klautau M, Rodriguez PRD, Costa RN, Silvano RG, Schwientek S, Ribeiro SM, Pinheiro US, Hajdu E. (2006) Captulo 3. Filo Porifera. In: Lavrado HP, Ignacio BL (eds). Biodiversidade bentnica da regio central da Zona Econmica Exclusiva brasileira. Museu Nacional, Srie Livros 18, Rio de Janeiro. pp. 109-145 Silva CMM, Mothes B (1996) SEM analysis: an important instrument in the study of marine sponges biodiversity. Acta Microscopica 5(B): 188-189 Silva CMM, Mothes B (2000) Three new species of Geodia Lamarck, 1815 (Porifera, Demospongiae) from the bathyal depths off Brazilian coast, Southwestern Atlantic. Rev Suisse Zool 107(1): 31-48 Sollas WJ (1886) Preliminary account of the Tetractinellid sponges dredged by H.M.S. Challenger 1872-76. Part I. The Choristida. Sci Proc Roy Dublin Soc 5:177-199 Sollas WJ (1888) Report on the Tetractinellida collected by H.M.S. Challenger during the years 1873-1876. Rep Sci Res Voy H.M.S. Challenger 1873-1876, Zool 25: 1-458 Topsent E (1902) Les Asterostreptidae. Bull Soc Sci Md Ouest 11(2): 1-18 Topsent E (1904) Spongiaires des Aores. Rs Camp Sci Prince Albert I0 Monaco 25: 1-280

Porifera research: Biodiversity, innovation and sustainaBility - 2007

483

The events of metamorphosis in the demosponge Halisarca dujardini Johnston, 1842, studied with immunocytochemical method
Yulia I. Mukhina(1*), Vadim V. Kumeiko(2), Olga I. Podgornaya(3), Sofia M. Efremova(1)
Biological Institute of St. Petersburg State University, St. Petersburg, Oranienbaumskoye sch., 2 Stary Peterhof, S. Petersburg, 198504, Russia. juliamuchina@mail.ru, smefremova@mail.ru (2) Institute of Marine Biology FEB RAS, Far Eastern National University, Vladivostok, Russia.vkumeiko@yandex.ru (3) Institute of Cytology RAS, St. Petersburg, Russia. podg@IM1632.spb.edu
(1)

Abstract: The behaviour and metamorphosis of larvae in Halisarca dujardini from the White Sea were studied using immunocytochemical methods. Larvae released from the parent sponges body and moved actively, demonstrating positive phototaxis. All external cells had flagella immediately after larva released. Before the settlement, posterior flagellated cells lost flagella, while anterior-lateral flagellated cells retain them up to the end of the free-swimming period. Flagellated cells were directly involved in metamorphosis. Just after settlement the narrow monolayer of lateral flagellated cells attached to the substrate was formed around the flattened larva. Their flagella were internalised and surrounded by flagellated cells cytoplasm. Gradually antibodies (AB) S (against the major protein S (68 kDa) of flagellated cells of larva) label spread inward the metamorphosing larva; both AB S and AB T (commercial anti-tubulin antibodies) staining were confined to internal cell mass. In three days post settlement most of the former flagellated cells composed the choanocyte-chambers. AB S labels granules in the cell body, as it did in premetamorphic flagellated cells. Diffusive AB S and AB T staining is visible else in T-shaped dermal pinacocytes on sponge surface. Still, these cells do contain both antigen characteristics of flagellated cells. A newly-formed adult sponge develops a definitive network of the choanocyte chambers within one week after settlement. At that time AB S label is seen in choanocytes but it disappears in the pinacoderm cells (T-shaped pinacocytes). Protein S is concentrated at cells apical part, exactly in the collars. Amoebocytes the other cell type visualized in the sections display no AB S staining. Thus in the development of H. dujardini the larval flagellated cells participate directly in construction of definitive organs such as the choanocyte chambers and the upper pinacoderm. Keywords: Halisarca dujardini, larvae, metamorphosis, protein marker, sponge

Introduction
Metamorphosis is a short stage in post-embryonic development. It is a time of crucial changes in the morphology and physiology of larva as it turns into a juvenile adult. The metamorphosis of sponges, as well as that of the other multicellular animals with sedentary or slow-moving adult forms, involves global changes in the body plan occurring when a mobile larva passes to sedentary life. The process normally comprises several stages, substrate choice and settlement, contact reaction of larval attachment to the substrate, followed by morphogenesis per se, including the development of choanoderm, coverings, and aquiferous canals. A successful completion of each stage is crucial for the proper initiation of the next one. The first stage of metamorphosis, larval settlement and attachment to the substrate, is critical for further development as it is only after this stage when metamorphosis begins. How does a sponge larva, which has neither statocysts nor other sensory organs, choose the proper substrate for settlement, and how does the settlement occur? Some authors proposed

that the direction and speed of larval movement depend on long flagella of the external cells and on pigmented cells associated with flagellated cells, both situated at the posterior pole (Woollacott and Hadfield 1989, Woollacott 1993, Maldonado and Young 1996, Maldonado 2006). Recently, it was shown that long flagella of posterior pole cells in tufted parenchymella larvae are extremely sensitive to sharp changes in light intensity (Leys and Degnan 2001, Jackson et al. 2002, Maldonado 2006). Leys and Degnan (2001) proposed that changes in light intensity affect coordinated work of long flagella and force a larva to swim off brightly illuminated areas thereby choosing shaded areas for settlement. Another important peculiarity of early metamorphosis is larval attachment to the substrate. The settlement can theoretically occur only after larvae have attained a threshold of physiological and morphological maturity that renders them competent for settlement (Maldonado 2006). In some sponge larvae, initial adhesion to the substrate is achieved via mucous secretion produced by flagellated anterior pole cells, while eventual attachment occurs after the differentiation of collagen-secreting cells (Borojevic and Lvi 1965, Boury-

484

Esnault 1976, Bergquist and Green 1977, Evans 1977, Bergquist et al. 1979, Kaye and Reiswig 1991, Leys and Degnan 2002, Maldonado 2006). However, some authors point out that attachment does not involve secretions from specialized cells and that initial attachment may consist of tenuous point-adhesions of the basal lamella (Kaye and Reiswig 1991). To date, the mechanism of larval adhesion to the substrate is still obscure. Undoubtedly, the most interesting and really intriguing point of metamorphosis is the mechanism of morphogenetic movements sensu stricto, resulting in the formation of main definitive structures. In 1892, Y. Delage proposed the theory of inversion of the germ layers in sponges (Delage 1892). He described the inward migration of the superficial larval flagellated cells (ectoderm) during metamorphosis in some freshwater and marine Demospongiae. The migrated cells formed the choanoderm, internal compartment of the adult sponge responsible for animal feeding. Based on this phenomenon Delage named the sponges Enantiozoa (turned inside out) opposing them to the other Metazoa. Since that time, Delages observations have been confirmed and disproved several times, and the fate of flagellar cells and formation of definitive layers in sponges are topical questions of developmental and evolutionary biology (see Efremova 1979, Leys 2004, Maldonado 2004, Ereskovsky and Dondua 2006). One view holds that the flagellar cells are transformed into choanocytes as is the case in juvenile sponges Mycale contarenii (Borojevic and Lvi 1965), Hamigera hamigera (Boury-Esnault 1976), Haliclona permollis (Amano and Hori 1996), Reniera sp. (Leys and Degnan 2002) but the opposing view contends that they are lost by exfoliation or phagocytosis during metamorphosis as it was shown in Spongilla lacustris (Brien and Meewis 1938), Ulosa sp., Halichondria moorei and Microciona rubens (Bergquist and Green 1977, Bergquist and Glasgow 1986), Microciona prolifera (Misevic et al. 1990) and commercial sponges Hippospongia and Spongia (Kaye and Reswig 1991). The metamorphosis of the sponge Halisarca dujardini was described using light (Lvi 1953, 1956, Gonobobleva and Ereskovsky 2004) and electron microscopy (Gonobobleva and Ereskovsky 2004). However, given that the morphology of flagellar cells changes substantially in metamorphosis, it is obvious that to precisely trace their fate, one needs special approaches using molecular markers specific for superficial flagellated cells. We isolated a protein (68, protein S) from superficial flagellar cells of H. dujardini, which is a specific marker for these cells. Earlier, we showed that antibodies against this protein can be used for specific labeling of superficial cells (Mukhina et al. 2004, 2006). In the present work, we used this approach to study in detail the metamorphosis of H. dujardini with an emphasis on morphogenetic movements involving flagellated cells and to trace the fate of these cells.

Inlet, Kandalaksha Bay, White Sea (6620N, 3340E), at a depth of 0.5 5 m in July 2001- 2004. The specimens were transported to the laboratory of White Sea Biological Station of Zoological Institute RAS and maintained in the aquarium with aerated seawater at 14oC. The released larvae were placed in plastic Petri dishes, or in Petri dishes coated with 2% bactoagar (DiFCo Laboratories, USA) and containing sterilefiltered seawater (FSW) with 0.003% sodium cefazoline added. Two to six days after that, free-swimming larvae attached to the substrate and flattened. Experiments on larval phototaxis were run in the darkroom at 14oC. Light from a cold light source (I-32D) was passed through a diffuser on one side of aquarium, where larvae were maintained.

Light microscopy
For whole mount preparations, free-swimming and metamorphosing larvae were fixed in 4% paraformaldehyde in 0.02 M phosphate buffer (pH 7.4). Sucrose was added to the fixative mixtures to reach isoosmotic condition of 625 mOsM. For paraffin sections, juvenile sponges and the fragments of adult sponges were fixed in Bouins solution (5 ml glacial acetic acid, 25 ml formalin, and 75 ml saturated solution of picric acid) for 4 hours, dehydrated through an ethanol series, transferred into xylene and 1:1 xylene:paraffin mixture and embedded in paraffin.

Immunofluorescence
Twelve micrometer thick paraffin sections were placed on poly-L-lysine coated slides, paraffin was removed with o-xylene treatment and after stepwise hydration in ethanol series (from 96% to 30%) the slides were put in TBS-Tw. In order to block non-specific reactions the slides were pretreated with 5% bovine serum albumin solubilized in TBS. We used AB S against the major 68-kDa protein S from larval flagellated cells (about antibody production see (Mukhina et al. 2006)) and commercially anti-, -tubulin (AB T) (raised in mouse, Sigma, USA). Anti-tubulin antibodies were chosen as an accessory marker for visualizing the flagellum and submembrane cytoskeleton structures. Fluorescein isothiocyanate (FITC) or Texas redconjugated goat antiguinea pig or goat anti-mouse antibodies were used as secondary antibodies (Sigma, USA). The final dilutions of secondary antibodies corresponded to recommendations provided. All the procedures were done at room temperature. Sections were mounted in Moviol 4-88 media (Calbiochem) and analyzed under a LEICA fluorescent microscope and LSM 5 PASCAL confocal laser scanning microscope (Carl Zeiss, Germany), equipped with Ar (488 nm) and He-Ne (543 nm) laser sets. The images were processed by the LEICA Q-FISH or the LSM 5 Image Browser. For control, some sections were incubated in non-immune serum and processed otherwise as above. They were free of label. Fixed preparation of swimming and metamorphosing larvae were washed three times in TBS (30 min each time), permeabilized with 0.2% Triton X-100 in TBS for 5 min at room temperature to prepare the whole mount. Non-specific binding sites were blocked with 5% bovine serum albumin in TBS for 1 hour. Primary antibodies against major protein

Materials and methods Collection of sponges and larvae cultivation

The sponges Halisarca dujardini were collected from the alga Fucus vesiculosus in the bays of Cape Kartesh, Chupa

485

of flagellated cells (AB S, final dilution 1:500) and anti-, -tubulin (AB T) were incubated with larvae on poly-Llysine coated multiwell slides at 4C during 12 hours. The samples were washed three times in TBS (30 min each time) and stained with secondary antibodies as above, the whole procedure taking 2 hours. Adobe Photoshop CS2 software was used for image processing for publication.

Results Larval release, swimming behavior and vital observations

Larvae release from a body of a parent sponge begins in the end of June and proceeds about two weeks. One - two days before a release, larvae start to rotate slowly inside the sponge body that is clearly visible through a transparent layer of pinacoderm. Carried away with the water current the larvae quit a sponge through an osculum or, less often, through the breaks in exopinacoderm. Released larvae proceed their free swimming period. They move forward concurrently rotating around the anteriorposterior axis in a clockwise direction. Larvae have rounded shape and are about 120 - 130 m in diameter. External cells of the larva bear flagella about 12 m in length though the length of flagella of posterior pole cells is a little bit greater and amounts to 15 m. Newly released larvae move actively in water under illumination of an aquarium by a bunch of light and concentrate near to the shined surface, so they demonstrated a positive phototaxis response. No geotaxis was detected; larvae remained active at all depths in the aquarium illuminated by an ambient light source. Sometimes larvae fall on the bottom, swim near the substratum, and then turn away toward the water surface. In 12-24 hours these larvae swim slowly along the substrate and remain indifferent about light. Then they settle on a substratum by their anterior pole and continue to rotate simultaneously with varying velocity. Sometimes they temporarily stop rotation, and then recommence the movement. Such larval behavior could be observed from 2 hours to several days after release from a parent sponge. Finally, larvae become motionless, and flatten slightly. The transparent layer of external flagellated cells attached to the substrate appears around the flat larva body (Fig. 1). Process of an attachment of larva to a substratum needs about 40 minutes. A few larvae settle to the air-water interface, attach to it and undergo metamorphosis up to the stage of an osculum formation. Later, such juvenile sponges sink on the bottom, attach to a substratum and continue to develop.

Protein markers and the fate of flagellated cells

It is well known that just after attachment to and spreading on the substratum parenchymella initiates metamorphosis. Vital observation described above give us some evidences for inactivation of flagellar motors after settlement and for disappearing of flagella from larva periphery at the first stage of metamorphosis. Further investigation of superficial cell fate is actually impossible without using special markers,

which help to trace cell behavior and definitive pattern formation through a metamorphosis. As it was shown earlier we have found a specific protein marker (p68, S protein) of the sponge flagellated cells after larval cell separation with density gradient fractionation (Mukhina et al. 2004, Mukhina et al. 2006). By means of double immunofluorescence using both the anti-p68 (AB S) and commercial anti-tubulin (AB T) antibodies, we studied the distribution of protein S in larval superficial cells and its fate at successive stages of metamorphosis. Present data focused on the details of each stage of the process from superficial flagella inactivation to choanocyte differentiation with flagellated chamber formation and growth. Just before the settlement, posterior flagellated cells loose flagella, anterior-lateral flagellated cells keep them up to the end of free swimming period (Fig. 2A-C). Flagella visualized with AB T are seen all over the larval surface except for posterior pole, which is naked and lack AB T fluorescence (Fig. 2B). Such larvae settle on the substratum and flattened soon. In the narrow transparent monolayer of lateral flagellated cells attached to the substrate (Fig. 2D-F) their flagella appear to be submerged into flagellated cells cytoplasm Thus, flagella were internalized (Fig. 2G-I). Flagellated cells were directly involved into metamorphosis. On the optical section in the middle of the larva of metamorphosing larvae (Fig. 2KM), both AB S and AB T staining were confined to internal cell mass. This implies that former external cells (Fig. 2AC) rearrange to form flagellated chambers or line aquiferous canal anlages. The paraffin sections of the sponge of 3 days postsettlement were double stained (Fig. 3A). As it is seen, most of the flagellated cells composed the newly formed choanocyte chambers. AB S label granules in the cell body of the choanocytes, while their shape is not in full conformity with those of typical terminal-differentiated choanocyte of adult sponge (Fig. 3A, insert a). AB S and AB T staining is visible also in T-shaped dermal pinacocytes on sponge surface (Fig. 3A, insert b). Still, these cells do contain both antigen characteristics of flagellated cells. A newly-formed sponge develops the definitive choanocyte chambers within 6 days after the settlement. At that time AB S label is seen in the choanocytes, but it disappears in the pinacoderm cells (the T-shaped pinacocytes) (Fig. 3B). Amoebocytes the other cell type visualized in the sections display no AB S staining (Fig. 3A, B). Choanocyte protein S is concentrated at cells apical part and visualized as rings or curves outlining apical border. (Fig. 3B, insert). A peculiar AB S staining pattern is caused by cutting apical cell parts in the vicinity of collars at different angles. Hence, flagellated chambers acquire specific organization with cellular polarity and particular choanocyte features.

Discussion Larval behaviour

Prior to leaving the parental sponge, a larva slowly rotates around its longitudinal axis inside the embryonic capsule for a short time. At that time, the cavities of the embryonic capsules merge with the spongocoel opening outside through

486

Fig. 1: H. dujardini early metamorphosing larva. Phase contrast microscopy. Attaching the narrow monolayer of lateral flagellated cells to the substrate, 30 min after start of attachment. Abbreviations: fl, flagella of external flagellated cell; the narrow monolayer of lateral flagellated cells cytoplasm (arrowheads). Scale bar: 20 m.

the osculum (Ereskovsky 2005), which facilitates larvae leaving the sponge. The free-swimming period averages about two days. In the presence of light, the larvae generally swim forward continuously rotating clockwise (as observed from the posterior pole of the larva), with occasional bursts of acceleration for periods of several seconds. These observations are consistent with the data reported by the researchers who studied the larval behavior of H. dujardini (Lvi 1956, Bergquist et al. 1979, Ereskovsky and Gonobobleva 2000). H. dujardini larvae were found to display positive phototaxis (Bergquist et al. 1979). Unlike the planulae of the Hydrozoa, sponge larvae lack neurons (Leys and Degnan 2001). Nonetheless, they display prominent light-induced behavior (Warburton 1966, Bergquist and Sinclair 1968, Bergquist et al. 1970, Wapstra and van Soest 1987, Woollacott 1990, 1993). Some researchers suppose that in tufted Demospongiae larvae the role of photoreceptors may be played by pigmented cells localized close to the posterior pole (Leys and Degnan 2001, Maldonado 2006). Halisarca dujardini larvae seem to lack pigmented cells. However, the posterior pole area in newly released H. dujardini larvae is covered with a layer of flagellated cells, with flagella longer than those of cells

in the anterior and lateral sides of the larva. These cells may participate in photoreception. Furthermore, posterior pole cells of the older larvae lose their flagella before settlement. Our data confirm other researchers conclusions that H. dujardini larvae have no flagella on the posterior pole (Lvi 1956, Bergquist et al. 1979, Bergquist 1980). An opinion that posterior pole cells of H. dujardini larvae are ciliated (Ereskovsky 2005) only holds for young larvae. Evidently the morphology and consequently the function of the posterior pole cells change during the free swimming period of the larvae (Mukhina et al. 2004, 2006) that can be connected
Fig. 2: Halisarca dujardini free-swimming and metamorphosing larva double stained with AB S and T. Laser confocal microscopy of the whole mount larva. A, D, G. K. The merged image; B, E, L. AB T (green) only; C, F, M. AB S (red) only. A-C. Freeswimming larva; top optical section and midsection (insert). D-I. Larva 40 min after attachment to the substrate. Optical section, 4.84 m above the glass substrate; D, E, F. The whole larva; G, H, I. The narrow monolayer of lateral flagellated cells attached to the substrate. K, L, M. Six hours post-settlement. Optical section in the middle of the metamorphosing larva demonstrates the AB S label in the internal cell mass. Abbreviations: pp, posterior pole; fl, flagellum. Scale bar: 20 m.

487

488

Fig. 3: Laser confocal microscopy of paraffin section (optical midsection) of H. dujardini juvenile double stained with anti-tubulin and anti-p68 antibodies. A. 3 days post-settlement; AB S labels granules in the cell body of choanocytes (insert a) and pinacocytes (insert b). B. 6 days post-settlement; AB S labels are seen only in the choanocytes; abbreviations: fch, flagellated chambers; pinacocytes (arrowheads). Scale bar: 10 m.

with physiological and morphological maturity of larva before settlement (Maldonado 2006). By that time, larvae cease swimming and active substrate inspecting, sink to the bottom and attach to the substrate. The attachment to the substrate is an important stage in the sponge life cycle essential for the occurrence of other events in the metamorphosis of sponge (Kaye and Reiswig 1991). In some sponge larvae, the initial adhesion to the substrate is achieved owing to the mucous secretion produced by flagellated cells of the anterior pole, while the definitive attachment occurs after the differentiation of collagensecreting cells (Borojevic and Lvi 1965, Boury-Esnault 1976, Bergquist and Green 1977, Evans 1977, Bergquist et al. 1979, Kaye and Reiswig 1991, Leys and Degnan 2002). Some authors point out, however, that attachment does not involve secretions from specialized cells and the initial attachment may be realized by point-adhesions of the basal lamella (Kaye and Reiswig 1991).

Gonobobleva and Ereskovsky (2004) suppose that primary adhesion of H. dujardini larva to the substratum is carried out by means of secretion of a slimy substance by the external flagellated cells of its anterior hemisphere. However, H. dujardini larvae lack the specialized secretory cells. According to our data the narrow monolayer of lateral flagellated cells appears on the substrate around the larvae just after the settlement with their flagella immersed into the cytoplasm. Apparently the initial adhesion to the substrate involves the formation of pseudopodia by the apical cytoplasm of flagellated cells resulting in the appearance of focal contacts and subsequent formation of the basal plate.

Cell fate and definitive pattern formation

As we have shown the flagellated cells are directly involved in metamorphosis by formation of flagellated chambers and pinacoderm. So the larval flagellated cells are the source for

489

both internal and surface cells in juvenile H. dujardini. This conclusion is not in full agreement with Y. Delages classical description of inversion of germ layers (Delage 1892). The external cells of the larva transdifferentiate into the collar cells which are the inherent part of the internal environment, but the inner cells of the larva do not form the upper pinacoderm. According to Ivanova (2002), even in fresh-water sponges of the family Spongillidae, whose larvae have prematurely developed choanocyte chambers, external cells of a larva may transform into choanocytes and pinacocytes of a juvenile animal. A treatment of the haplosclerid sponge Reniera sp. larvae with fluorescein-containing marker CMFDA, labeling only surface, flagellated cells revealed a similar flagellated cells behaviour (Leys and Degnan 2002). In a juvenile sponge three days after metamorphosis, the marker was localized mainly in choanocytes, but some elongated cells on the sponge periphery were also labeled. One can not exclude that these cells belong to pinacocyte category. Transdifferentiation of larval flagellated cells into choanocytes was shown by TEM in the demosponge Haliclona permolis (Amano and Hori 1996). In calcareous sponges, the transformation of the larval flagellated cells into the choanocytes occurs during metamorphosis of both amphiblastula and coeloblastula larvae (Amano and Hori 1993, 2001) in spite of the paraphyly of the sponges demonstrated recently by Borchiellini et al. (2001). Larval flagellated cells and the choanocytes of the adult sponge differ significantly in morphology and organization of the basal apparatus of the flagellum (see Simpson 1984, Woollacott 1993). The distribution of protein S in larval cells differs from that in choanocytes. In flagellated cells, it is spread all over the cytoplasm as granules, while in definitive choanocytes it is confined to the apical part of the cell, exactly, to its collar which composed of a row of microvilli. The very unusual polypeptide composition of larval external cells with few prominent major proteins might suggest their terminal differentiation. We therefore believe that the unique protein composition of flagellated cells rather suggest their preparation for rapid and substantial changes evoked by transdifferentiation. If it is the case, the presence of protein S in flagellated cells may be accounted for by its anticipatory expression and storage for subsequent use in the formation of choanocytes apical structures. We carried out the preliminary identification of protein S by MALDI (matrix-assisted laser desorption/ionization mass spectroscopy) procedure. The control sample of the 42 kDa zone was identified as -actin with highest scores of similarity of invertebrate actin. At present, the protein S spectrum obtained cannot be identified unambiguously. The identification of protein S and the study of its fine localization in the cells are now in progress.

References
Amano S, Hori I (1993) Metamorphosis of calcareous sponges: II. Cell rearrangement and differentiation in metamorphosis. Invertebr Reprod Dev 24: 13-26 Amano S, Hori I (2001) Metamorphosis of coeloblastula performed by multipotential larval flagellated cells in the calcareous sponge Leucosolenia laxa. Biol Bull 200: 2032 Amano S, Hori I (1996) Transdifferentiation of larval flagellated cells to choanocytes in the metamorphosis of the demosponge Haliclona permollis. Biol Bull 190: 161-172 Bergquist PR (1980) A revision of the supraspecific classification of the orders Dictyoceratida, Dendroceratida, and Verongida (class Demospongiae). NZ J Zool 7: 443-503 Bergquist PR, Glassgow K (1986) Developmental potential of ciliated cells of ceractinomorph sponge larvae. J Exp Biol 45: 111122 Bergquist PR, Green C (1977) An ultrastructural study of settlement and metamorphosis in sponge larvae. Cah Biol Mar 18: 289-302 Bergquist PR, Sinclair ME, Green CR, Silyn-Roberts H (1979) Comparative morphology and behavior of larvae of Demospongiae. In: Lv C, Boury-Esnault N (eds). Biologie des Spongiaires. Coll Int CNRS 291: 103-111 Bergquist PR, Sinclair ME (1968) The morphology and behavior of larvae of some intertidal sponges. N Z J Mar Freshw Res 2: 426-437 Bergquist PR, Sinclair ME, Hogg JJ (1970) Adaptation to intertidal existence: reproductive cycles and larval behavior in Demospongiae. Zool Soc Lond 25: 247271 Borojevic R, Lvi C (1965) Morphogense exprimentale dune ponge partir de cellules de la larve nageante dissocie. Zeitsch Zellforsch 68: 57-69 Borchiellini C, Manuel M, Alion E, Boury-Esnault N, Vacelet J, Le Parco Y (2001) Sponge paraphyly and the origin of Metazoa. J Evol Biol 14: 171-179 Boury-Esnault N (1976) Ultrastructure de la larve parenchymella d`Hamigera hamigera (Schmidt) (Demosponge, Poecilosclerida). Origine des cellules grises. Cah Biol Mar 17: 9-20 Brien P, Meewis H (1938) Contribution ltude de lembryogense des Spongillidae. Arch Biol 49: 177-250 Delage Y (1892) Embryognie des ponges. Dveloppement postlarvaire des ponges siliceuses et fibreuses marines et deau douce. Arch Zool Exp Gn 2(10): 345-498 Efremova SM (1997) Once more on the position among Metazoa Gastrulation and germinal layers in sponges. In: Ereskovsky AV, Keupp HK, Kohring R (eds). Modern problems of poriferan biology. Berliner geowiss Abh E 20: 7-15 Ereskovsky AV (2005) Comparative embryology of sponges (Porifera). Saint-Petersburg State University, Saint-Petersburg Ereskovsky AV, Dondua AK (2006) The problem of germ layers in sponges (Porifera) and some issues concerning early metazoan evolution. Zool Anz 245: 65-76 Ereskovsky AV, Gonobobleva EL (2000) New data on embryonic development of Halisarca dujardini Johnston, 1842 (Demospongiae: Halisarcida). Zoosystema 22: 355-368 Evans CW (1977) The ultrastructure of larvae from the marine sponge Halichondria moorei Bergquist ( Porifera, Demospongiae) Cal Biol Mar 18: 427-433 Gonobobleva EL, Ereskovsky AV (2004) Polymorphism in freeswimming larvae of Halisarca dujardini (Demospongiae,

Acknowledgements
The present work was supported by Russian Foundation for Basic Research (grant 04-04-48990, 07-04-01703) and by the DOE Human Genome Program Grant (USA). We are grateful to Prof. Dr. V. Ya. Berger, the director of White Sea Biological Station of Zoological Institute RAS, for an opportunity to accomplish the field experiments. We used the equipment of Cores Facilities Chromas of Biological Institute of St. Petersburg State University and Materials and diagnostics in high technology of Institute of Cytology RAS, St. Petersburg.

490

Halisarcida). In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Bull Mus Ist Biol Univ Genova 68: 349-356 Ivanova LV (2002) Multiple variations of the fate of the flagellated cells during the metamorphosis of the parenchimella larvae in freshwater and marine demosponges. Boll Mus Ist Biol Univ Genova 66-67: 99 Jackson D, Leys SP, Hinman VF, Woods R, Lavin M, Degnan B (2002) Ecological regulation of development: induction of marine invertebrate metamorphosis. Int J Dev Biol 46: 679-686 Kaye HR, Reiswig HM (1991) Sexual reproduction in four Caribbean commercial sponges. III. Larval behaviour, settlement and metamorphosis. Invertebr Reprod Dev 19: 25-35 Lvi C (1953) Structure de la larve et du rhagon de lponge Halisarca dujardini (Johnston). Compt Rend Acad Sci Paris 234: 1710-1712 Lvi C (1956) tude des Halisarca de Roscoff. Embriologie et systmatique des Dmosponges. Trav Stat Biol Roscoff N S 7: 3181 Leys SP (2004) Gastrulation in sponges. In: Stern S (ed). Gastrulation. Cold Spring Harbor Laboratory Press, New York. pp. 23-31 Leys SP, Degnan BM (2001) Cytological basis of photoresponsive behavior in a sponge larva. Biol Bul 201: 323-338 Leys SP, Degnan BM (2002) Embryogenesis and metamorphosis in a gaplosclerid demosponge: gastrulation and transdifferentiation of larval ciliated cells to choanocytes. Invertebr Biol 121(3): 171189 Maldonado M (2004) Choanoflagellates, choanocytes, and animal multicellularity. Invertebr Biol 123: 1-22 Maldonado M (2006) The ecology of the sponge larva. Can J Zool 84: 175-194

Maldonado M, Young CM (1996) Effects of physical factors on larval behavior, settlement and recruitment of four tropical demosponges. Mar Ecol Progr Ser 138: 169-180 Misevic G, Schlup V, Burger M (1990) Larval metamorphosis of Microciona prolifera: evidence against the reversal of layers. In: Rtzler K (ed). New perspective in sponge biology. Smithsonian Inst. Press, Washington. pp. 182-187 Mukhina YI, Kumeiko VV, Podgornaya OI, Efremova SM (2006) The fate of larval flagellated cells during metamorphosis of the sponge Halisarca dujardini. Int J Dev Biol 50: 533-541 Mukhina YI, Podgornaya OI, Efremova SM (2004) Peculiarities of sponge Halisarca dujardini (Demospongiae, Halisarcida) larval cells. I. Separation of cells and their aggregation properties. Tchitologia 46(6): 483-491 Simpson TL (1984) The cell biology of sponges. Springer-Verlag. New York Wapstra M, van Soest RWM (1987) Sexual reproduction, larval morphology and behavior in demosponges from the southwest of the Netherlands. In: Vacelet J, Boury-Esnault N (eds). Taxonomy of Porifera. Springer-Verlag, Berlin. pp. 281-307 Warburton FE (1966) The behavior of sponge larvae. Ecology 47: 672674 Woollacott RM (1990) Structure and swimming behavior of the larva of Halichondria melanadocia (Porifera: Demospongiae). J Morphol 205: 135-145 Woollacott RM (1993) Structure and swimming behavior of the larva of Haliclona tubifera (Porifera: Demospongiae). J Morphol 218: 301-321 Woollacott RM, Hadfield MG (1996) Induction of metamorphosis in larvae of a sponge. Invertebr Biol 115: 257-262

Porifera research: Biodiversity, innovation and sustainaBility - 2007

491

Haloperoxidase from the marine sponge Erylus discophorus (Schmidt, 1862)

Marisa H. Nicolai(1), Ana Esteves(1), Marise Almeida(1, 2), Madalena Humanes(1*)
Centro de Qumica e Bioqumica - Departamento de Qumica e Bioqumica da Faculdade de Cincias da Universidade de Lisboa, Edifcio C8, Campo Grande, 1749-016 Lisboa, Portugal. mmhumanes@fc.ul.pt (2) Laboratrio de Biomateriais, Faculdade de Medicina Dentria da Universidade de Lisboa, 1649-003 Lisboa, Portugal. marise.almeida@fmd.ul.pt
(1)

Abstract: The presence of halogenated compounds is well documented in the phylum Porifera, although little is known about the biosynthetic pathways leading to their production. Studies with sponge enzymes are scarce, in particular those involving halogenating enzymes. Preliminary results with sponge crude extracts revealed haloperoxidase activity in some sponges. From Erylus discophorus (Schmidt, 1862) we extracted and partially purified a haloperoxidase with iodo and bromoperoxidase activities. This very unstable enzyme is a glycoprotein with a molecular mass higher than 200 kDa, as determined by gel filtration. A UV-visible spectrum with a shoulder in the Soret region, and the observed inhibition by azide ions, makes the hypothesis of a heme protein more likely. Keywords: Bromoperoxidase, Erylus discophorus, halogenation reactions, Northeastern Atlantic, Porifera

Introduction
The oceans are the single largest source of biogenic organohalogens, which are biosynthesized by a myriad of seaweeds, sponges, corals, tunicates, bacteria, and other marine life (Gribble 2003). Due to their habitat, sponges rely heavily on chemicals for survival, and many of these compounds contain halogen(s). Examples of sponge halogenated compounds include fatty acid derivatives (Pham et al. 1999), pyrroles (Cafieri et al. 1998), indoles (Qureshi et al. 1999), phenol derivatives (Utkina et al. 2005), tyrosine derivatives (Nicholas et al. 2001, Saeki et al. 2002, Kijjoa et al. 2002), terpenes (Miyaoka et al. 2006), diphenyl ethers (Vetter and Jun 2003, Hanif et al. 2007), and even dioxins (Utkina et al. 2001). The formation of these compounds probably involves an enzymatic halogenation. Haloperoxidases (HPO) catalyse, in the presence of hydrogen peroxide, the oxidation of halides (X-: iodide, bromide or chloride) to their corresponding hypohalous acids or to a related electron-oxidised halogenating intermediates such as OX-, X3- and X+. Most peroxidases are able to catalyze the formation of a carbon-halogen bond with a suitable halide, in the presence of a halide acceptor, such as tyrosine (Henderson and Heinecke 2003). Concerning the cofactor nature, three major classes of HPO are known: HPO containing no prosthetic group, the heme-containing HPO and the vanadium-dependent HPO (VHPO) that binds a vanadate ion (HVO42-). Enzymes representing these two latter classes differ, at least, in two aspects - catalytic mechanism and stability. Heme-containing HPO catalyse the formation of hypohalous acid by a redox mechanism, whereas in VHPO the vanadate group does not change its redox state but may function as a

Lewis acid. Unlike heme-containing HPO, VHPO exhibit a high level of thermostability, and good tolerance to organic solvents, as well as to high concentrations of their substrates and products (Almeida et al. 2001). Haloperoxidases from a variety of plant, animal and microbial sources have been identified and studied. These include haloperoxidases from horseradish, milk, saliva, tears, red and white blood cells as well as fungi, marine algae, and invertebrates (Alexander 1959, Shaw and Hager 1959, Archer et al. 1965, Ilgner and Woods 1985). As far as sponges are concerned, just one report was found: a labile chloroperoxidase-like activity in a subtropical marine sponge (Baden and Corbett 1979). In this paper we present a survey of haloperoxidase activities detected in sponges from the Northeast Atlantic and the extraction and partial purification of a haloperoxidase from the sponge Erylus discophorus (Schmidt, 1862).

Material and methods

Chemicals: Analytical grade inorganic salts were purchased from Merck. Diethylaminoethyl (DEAE)-Sephacel, Sephacryl S-300, hydrogen peroxide, high molecular weight gel filtration and Sodium Dodecyl Sulphate-Polyacrylamide Gel Electrophoresis, (SDS-PAGE) standards, orcinol, bovine serum albumin (BSA), glucose, 2-(N-morpholino)ethanesul fonic acid (MES) and Coomassie brilliant blue G-250 were purchased from Sigma. Biological specimens: Sponge specimens were collected by SCUBA diving in several areas of two Portuguese marine natural reserves (Arrbida and Berlengas) in the period May-

492

July 2003, at Ferrol (Spain) in April 2004, and at Gorringe (a large seamount located off the south west coast of Portugal) in June 2006 (Fig. 1). After collection, the specimens were transported to the laboratory in refrigerated containers and frozen at -20C until required. From all the samples, a voucher was preserved in absolute ethanol for identification purposes. The samples were named according to the respective collection site: B for Berlengas, A for Arrbida, F for Ferrol and G for Gorringe, followed by a number that corresponds to the order of collection. Detection of halogenating activity: A small volume of sponge (circa 2.5 mL) was homogenized in 30 mL of 0.2 M Tris-SO4 buffer (pH 8.3). After centrifugation, to remove cell debris, this crude extract was used to test the triiodide formation from I- and H2O2, catalysed by a iodoperoxidase (IPO) and followed at 350 nm, (Bjrkstn 1968). If the sample showed positive activity (+), then the consumption of monochlorodimedone (the assay to test bromoperoxidase (BrPO) activity) will follow (Hager et al.1966). If this test was positive (++), bromide was replaced by chloride and the crude extract tested again, now for the presence of chloroperoxidase (ClPO) activity and assign (+++), if positive. For quantitative purposes, one enzyme unit was defined as the amount of enzyme which catalyses the formation of 1 mol of product per minute. Enzyme extraction and purification: Fresh sponge of E. discophorus, specimen FE05 (10g) was triturated and homogenized with 40 mL of 0.2 M Tris-SO4 buffer (pH 8.3), for 30 minutes, at 4C, followed by centrifugation (35 minutes, 5500 g, 4C), to remove cell debris and clarification. This crude extract was then applied to a DEAE-Sephacel column equilibrated with 0.2 M Tris-SO4 buffer (pH 8.3). The column was eluted with 200 mL of 0.2 M Tris-SO4 (pH 8.3) buffer, followed by 400 mL of a linear gradient 02 M NaCl in 0.2 M Tris-SO4 buffer (pH 8.3), and finally eluted with 100 mL of 2 M NaCl in 0.2 M Tris-SO4 buffer (pH 8.3), at 1 mL.min-1 flow. The fractions presenting the highest iodoperoxidase activity were pooled and concentrated by ultrafiltration using a 10 kDa membrane cut-off. This concentrated fraction was then introduced into a gel filtration column (Sephacryl S300) equilibrated and eluted with 165 mL of 0.2 M NaCl in 0.2 M Tris-SO4 buffer (pH 8.3), using a 0.5 mL.min-1 flow. The fraction showing the highest iodoperoxidase activity was pooled and stored at 4C. In the purification procedures, all the activity determination were done using the IPO assay, since this determination gives the highest absolute values and allow us to monitor the purification with more precision. Protein determination: Protein quantification was based on the Bradford microassay method, using bovine serum albumin (BSA) as a standard (Bradford 1976). Total glycid determination: Glycidic content was determined by the orcinol method using glucose as a standard (White and Kennedy 1986). The pH dependence of iodoperoxidase activity: Iodoperoxidase activity tests were performed using the following buffer solutions: 0.1 M sodium acetate (pH 4.0,

Fig. 1: Location of the sponge collection sites.

4.5, 5.0 and 5.5), MES (pH 5.0 and 5.5) and 0.2 M citratephosphate (pH 6.2), at 25C. Temperature dependence of iodoperoxidase activity: The iodoperoxidase activity tests were performed in 0.1 M sodium acetate buffer (pH 5.0) after incubation of the samples, for 30 minutes, at 4, 15, 25, 30, 35, 40 and 45C.

Results Haloperoxidase survey

The survey conducted on the presence of haloperoxidase activity in the sponges of the Portuguese coast (Berlengas and Arrbida) revealed that this activity could be found in many sponge species (Table 1). In line with these results, we choose the sponge E. discophorus to extract and purify the haloperoxidase, since it was the sponge that presented the highest activity values and also is quite abundant in some of the collection sites. The results of the activity tests are presented in Table 2 for the nine examined samples of E. discophorus, expressed as specific activity, i.e., the number of activity units per mg of protein in the sample.

Purification of Erylus discophorus haloperoxidase

For enzyme extraction, several conditions were tested. The most efficient conditions were those described in the materials and methods section. Once the crude extract was obtained, it

493

Table 1: Presence of haloperoxidase activity in marine sponges. Species Stelletta grubii Schmidt, 1862 Erylus discophorus (Schmidt, 1862) Cliona celata Grant, 1826 Cliona viridis (Schmidt, 1862) Suberites carnosus (Johnston, 1842) Ciocalypta penicillus Bowerbank, 1862 Axinyssa aurantiaca (Schmidt, 1864) Halichondria (Halichondria) panicea (Pallas, 1766) Myxilla (Myxilla) rosacea (Lieberkhn, 1859) Crambe sp. Hymedesmia sp. Phorbas fictitius (Bowerbank, 1866) Chalinula molitba (de Laubenfels, 1949) Haliclona (Reniera) cinerea (Grant, 1826) Spongia sp. Sarcotragus spinosulus Schmidt, 1862 Ircinia sp. Dysidea fragilis (Montagu, 1818) Aplysina sp. Clathrina cerebrum (Haeckel, 1870) Clathrina contorta Minchin, 1905 Specimens Haloperoxidase examined activity 3 7 5 3 1 1 1 2 10 1 1 3 2 3 3 3 2 1 1 1 1 + ++ + ++ + ++ + ++ ++ + + + + -

Table 2: Haloperoxidase specific activities (enzymatic activity units per mg of protein) for the species E. discophorus. Sample reference B125 B172 B206 B329 B351 B358 B397 FE05 G06.01* IPO (U/mg) 5.080 3.275 8.643 1.969 3.179 10.535 8.750 26.550 11.524 BrPO (U/mg) ClPO (U/mg) 0.065 0.217 0.528 0.032 0.076 0.249 0.463 0.570 0.594 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000

IPO - iodoperoxidase specific activity BrPO - Bromoperoxidase specific activity ClPO - Chloroperoxidase specific activity * this specimen has been identified only to the genus level (Erylus sp.)

+ positive for iodoperoxidase (IPO) activity ++ positive for iodoperoxidase (IPO) and bromoperoxidase (BrPO) activity

was subjected to several chromatographic steps. Fig. 2 shows a representative chromatogram, obtained from a weak anionic exchange chromatography, with DEAE-Sephacel. The chromatogram shows three different regions concerning the 280 nm absorption characteristics of the eluted fractions. For each region the total glycid and protein content was determined as well as it IPO specific activity (see insert in Fig. 2). These results clearly show that the eluted fractions corresponding to region 1 are composed essentially by sugars, fractions in region 2 contained, besides sugar and proteins, the iodoperoxidase fraction and the remaining region 3, is composed also by sugar and proteins. This chromatographic step removed sugars and some contaminant proteins from the crude extract but the specific activity was just slightly increased. However, even keeping the fractions with IPO activity at 4C, their activities were quickly lost (in a period of 24 hours activity decreases

approximately 50%). We tried to freeze the fractions at -80C, but the activity decreasing was even steeper, which may be due to thawing conditions. The fractions with IPO activity were concentrated by ultrafiltration, using a 10 kDa membrane cut-off and further applied into a strong anionic exchange chromatography (Mono-Q), but, unfortunately none of the collected fractions presented IPO activity. So, the DEAE-Sephacel IPO containing fractions were, therefore, subject to a gel filtration chromatography with Sephacryl S-300. The results are shown in Fig. 3. In this chromatogram we observe 4 regions which included two peaks (named 2 and 3), with some degree of overlap. Only the first peak presented IPO activity, but solely around 70% of the initial IPO specific activity could be recovered. This fraction loses activity very fast. After 24 hours, IPO activity decreases to 50% and after 48 hours the IPO activity is completely lost. As for the weak anionic chromatography step, the glycid and protein content and IPO specific activity was determined for the several regions in the chromatogram (see insert in Fig. 3). For molecular weight determination, this sample was subject to a gel filtration chromatography in a Sephacryl S300 column, which had been previously calibrated using a wide range of molecular weight standards (6.5-669 kDa), and using dextran blue to determine the void volume of the column. However, the protein was eluted within this void volume meaning that its molecular mass should be between 700 and 2000 kDa (results not shown). The SDS-PAGE of these fractions did not show any band staining with Coomassie brilliant blue or silver nitrate but, when the gel was stained for HPO activity with o-dianisidine, one well defined band was observed (Almeida et al., 2001). Also, a very faint band, just near the entrance well, was observed for the glycoprotein staining (results not shown). These results are in agreement with the lower 280 nm

494

Fig. 2: Elution profile from DEAESephacel chromatography for specimen FE05. 1, 2 and 3, stands for the different regions of the chromatogram. Insert: glycid and protein content and IPO specific activity values for the several regions of the chromatogramm.

Fig. 3: Elution profile from Sephacryl S-300 gel filtration chromatography for fractions from region 2 of DEAE-Sephacel chromatography. 1, 2, 3 and 4 stands for the different regions of the chromatogram. Insert: glycid and protein content and IPO specific activity values for the several regions of the chromatogram.

absorption values (due to lower protein content), observed in all chromatographic steps and with a high specific IPO activity values also observed for the IPO containing fractions. The UV-visible absorbance of the partially purified enzyme in 0.2 M Tris-SO4 buffer (pH 8.3) was determined from 250600 nm (Fig. 4). The spectrum showed a slight absorption at 410 nm (Soret region). An optimum pH of 5.0, for IPO activity, was determined for this partially purified enzyme, as shown in Fig. 5. It is interesting to notice the absence of activity in MES buffer at pH higher than 6.0. The enzyme presents the same activity from 4 to 25C, but after 25C quickly loses activity; at 45C, 80% of activity is lost (Fig. 6). Addition of 0.3 mM of sodium azide completely inhibits the iodoperoxidase activity.

Discussion
Preliminary studies of haloperoxidase activity on sponges collected in Arrbida and Berlengas Archipelago, allowed us to identify some sponges from which haloperoxidase enzymes could be detected. It is interesting to notice a high percentage (ca. 50% of the samples examined) of sponge samples with haloperoxidase activity. Also, it was, the first time that sponges from the Calcarea class, have been screened for haloperoxidase activity. Both Calcarea samples belong to the Clathrina Gray, 1867 genus and, in one of them, haloperoxidase activity was detected. The remaining specimens belong to several orders of the class Demospongiae. Along with E. discophorus, other species also showed haloperoxidase activity, such as Myxilla (Myxilla) rosacea (Lieberkhn, 1859), Cliona viridis

495

Fig. 4: UV-visible spectrum of the partially purified haloperoxidase from Erylus discophorus.

Fig. 5: Effect of pH on IPO specific activity at 25C ( 0.1 M acetate buffer pH 4.0, 4.5, 5.0 and 5.5; 0.1 M MES buffer pH 5.5, 6.0 and 6.5; 0.2 M citrate-phosphate buffer pH 6.2).

samples of E. discophorus presented haloperoxidase activity, independently of the collection site. Samples of C. celata did not present any haloperoxidase activity at all. The studies on the partial purification and preliminary characterization of the haloperoxidase from the sponge E. discophorus revealed a complex system to deal with. The rapid decrease in the activity during the purification steps was the major point that hampered this study. Several reasons may be ascertain for this fact, for example the possible involvement of an ion or molecule in the maintenance of the three dimensional structure of the enzyme that was removed during purification procedures. These same stability problems were also observed in an old work on the tropical marine sponge Iotrochota birotulata (Higgin, 1877) (Baden and Corbertt 1979). Neverthless, due to the high specific activity values showed by this enzyme a preliminary characterization could be performed. This sponge contains a haloperoxidase that shows iodo- and bromoperoxidase activities. The very faint shoulder in the Soret region of the UV-Vis spectrum, the inhibition by azide and the thermal instability may provide evidences that we are in the presence of a heme haloperoxidase. This fact would be also in close agreement with the known fact that in animal kingdom only heme haloperoxidases were found. However, other type of haloperoxidases may not be discharged. This study, though still preliminary in its results, provides a background for the screening and study of the haloperoxidases in marine sponges and we hope that will stimulate more studies. Since marine sponges are a rich source of halometabolites, with diverse biological activities, the knowledge of the enzymes involved in their production in vivo might be of major importance not only from a biological point of view but also from a pharmaceutical approach.

Acknowledgments
We thank Helena Gaspar, Gonalo Calado, Joana Xavier and F.J. Cristobo for the collection and identification of sponge specimens. The authors wish to thank all the support from Reserva Natural da Berlenga Instituto da Conservao da Natureza. This work was supported by Fundao para a Cincia e a Tecnologia, Project: POCTI/ QUI/45670/2002.

References
Fig. 6: Effect of temperature on IPO specific activity, using 0.1 M sodium acetate buffer (pH 5.0). Alexander NM (1959) Iodide peroxidase in rat thyroid and salivary glands and its inhibition by antithyroid compounds. J Biol Chem 234: 1530-1533 Almeida M, Filipe S, Humanes M, Maia MF, Melo R, Severino N, Silva JL, Frasto da Silva JJR, Wever R (2001) Vanadium haloperoxidases from brown algae of the Laminariceae family. Phytochemistry 57(5): 633-642 Archer JT, Jackas M, Morell DB (1965) Studies on rat eosinophil peroxidase. Biochem Biophys Acta 99: 96-101 Baden DG, Corbett MD (1979) Peroxidases produced by the marine sponge Iotrochota birotulata. Comp Biochem Physiol 64B: 279283 Bjrkstn F (1968) A kinetic study of the horse-radish peroxidasecatalyzed oxidation of iodide. Eur J Biochem 5: 133-142

(Schmidt, 1862) and Haliclona (Reniera) cinerea (Grant, 1826). Considering that these results could be influenced by some form of consortia or association with other organisms, in particular microorganisms, we tried to collect the same species in several locations, geographically apart. Unfortunately, only two species (E. discophorus and Cliona celata Grant, 1826) could be collected in more than one site. All the examined

496

Bradford MM (1976) A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72: 248-254 Cafieri F, Fattorusso E, Taglialatela-Scafati O (1998) Novel bromopyrrole alkaloids from the sponge Agelas dispar. J Nat Prod 61(1): 122-125 Gribble GW (2003) The diversity of naturally produced organohalogens. Chemosphere 52: 289-297 Hager LP, Morris DR, Brown, FS, Eberwein H (1966) Chloroperoxidase II - utilization of halogen anions. J Biol Chem 241: 1769-1777 Hanif N, Tanaka J, Setiawan A, Trianto A, de Voogd NJ, Murni A, Tanaka C, Higa T (2007) Polybrominated diphenyl ethers from the Indonesian sponge Lamellodysidea herbacea. J Nat Prod 70(3): 432-435 Henderson JP, Heinecke JW (2003) Myeloperoxidase and eosinophil peroxidase: phagocyte enzymes for halogenation in humans. In: Hutzinger O (ed). Natural production of organohalogen compounds The handbook of environmental chemistry, vol 3P. Springer Berlin/Heidelberg. pp. 201-214 Ilgner RH, Woods AE (1985) Purification, physical properties and kinetics of peroxidases from freshwater crayfish (genus Orconectes). Comp Biochem Physiol 82B: 433-440 Kijjoa A, Watanadilokc R, Sonchaeng P, Sawangwong P, Nascimento MSJ, Silva AMS, Eaton G, Herzg W (2002) Further halotyrosine derivatives from the marine sponge Suberea aff. praetensa. Z Naturforsch C 57(7-8): 732-738 Miyaoka H, Yamanishi M, Mitome H (2006) PLA2 inhibitory activity of marine sesterterpenoids cladocorans, their diastereomers and analogues. Chem Pharm Bull (Tokyo) 54(2): 268-270

Nicholas GM, Newton GL, Fahey RC, Bewley CA (2001) Novel bromotyrosine alkaloids: inhibitors of mycothiol S-conjugate amidase. Org Lett 3(10): 1543-1545 Pham NB, Butler MS, Hooper JN, Moni RW, Quinn RJ (1999) Isolation of xestosterol esters of brominated acetylenic fatty acids from the marine sponge Xestospongia testudinaria. J Nat Prod 62(10): 1439-1442 Qureshi A, Stevenson CS, Albert CL, Jacobs RS, Faulkner DJ (1999) Homo-and nor-plakotenin, new carboxylic acids from the Palauan sponge Plakortis lita. J Nat Prod 62(8): 1205-1207 Saeki BM, Granato AC, Berlinck RG, Magalhaes A, Schefer AB, Ferreira AG, Pinheiro US, Hajdu E (2002) Two unprecedented dibromotyrosine-derived alkaloids from the Brazilian endemic marine sponge Aplysina caissara. J Nat Prod 65(5): 796-799 Shaw PO, Hager LP (1959) An enzymatic chlorination reaction. J Am Chem Soc 81: 1011-1012 Utkina NK, Denisenko VA, Scholokova OV, Virovaya MV, Gerasimenko AV, Popov DY, Krasokhin VB, Popov AM (2001) Spongiadioxins A and B, two new polybrominated dibenzo-pdioxins from an Australian marine sponge Dysidea dendyi. J Nat Prod 64(2):151-153 Utkina NK, Makarchenko AE, Denisenko VA (2005) Zyzzyanones B-D, dipyrroloquinones from the marine sponge Zyzzya fuliginosa. J Nat Prod 68(9): 1424-1427 Vetter W, Jun W (2003) Non-polar halogenated natural products bioaccumulated in marine samples II - Brominated and mixed halogenated compounds. Chemosphere 52(2): 423-431 White CA, Kennedy JF (1986) Oligosaccharides. In: Chapin MF, Kennedy JF (eds). Carbohydrate analysis a practical approach. IRL Press, Oxford. pp. 37-38

Porifera research: Biodiversity, innovation and sustainaBility - 2007

497

Ferric iron promotes the formation of oscules: observations on sponges in aquaria

Ronald Osinga(1*), Michiel Kotterman(2)
(1) (2)

Porifarma BV, Poelbos 3, 6718 HT, Ede, The Netherlands. ronald.osinga@porifarma.com Institute for Marine Resources and Ecological Studies (IMARES), Haringkade 1, 1976 CP, IJmuiden, The Netherlands

Abstract: Three species of Mediterranean sponges, Dysidea avara, Chondrosia reniformis and Acanthella acuta were maintained for more than a year in indoor cultivation systems. Different regimes for feeding were tested, which consisted predominantly of microalgae. Recently, it was demonstrated that iron (particularly Fe3+ ions) plays an important role in sponge morphogenesis and physiology. Hence, we added different concentrations of ferric citrate to the seawater in the culture systems and compared pulsed administration of iron with continuous administration. None of the experiments resulted in a significant increase in biomass of the sponges. However, continuous addition of 3 M Fe3+ in combination with deep-frozen Phaeodactylum tricornotum (a marine diatom) as a food source resulted in a positive physiological response of the sponges: all species increased their pumping activity under these conditions. The effect was most clear on C. reniformis. No oscula were observed on explants of this species. After the addition of ferric iron, the explants started to develop new oscula that showed a clear pumping activity. We conclude that continuous administration of Fe3+ is important to maintain pumping activity of sponges in tanks. Keywords: aquiferous system, cultivation, ferric iron, Porifera, sponges

Introduction
Controlled cultivation of marine sponges in land-based systems has been a persisting challenge to marine biologists and biotechnologists. For decades, sponges have been regarded as extremely difficult to culture ex-situ (Arndt 1933, Kinne 1977, Osinga et al. 1999). Osinga et al. (1999) emphasized that the design of an appropriate feeding regime is very likely to be a key factor determining culture success. In nature, sponges feed on a broad spectrum of particulate and dissolved organic matter. Hence, for successful culturing, the sponges should be provided a diet that appropriately mimics the highly diverse natural diet. An example of such an approach is a study by Duckworth and Pomponi (2004) on the Caribbean sponge Halichondria melanadocia. These authors mimicked the food availability in the coastal waters of Florida by supplying the sponges with mixtures of cultures of different species of microalgae, bacteria and yeast. Several other studies have reported positive culture results (Thomassen and Riisgrd 1995, Belarbi et al. 2003, De Caralt et al. 2003, Nickel and Brummer 2003, Osinga et al. 2003, Sipkema et al. 2006). Two of these (Thomassen and Riisgrd 1995, Osinga et al. 2003) demonstrated that some sponge species can be grown even on a single food species diet. Despite these positive reports, sponge culturists still experience many problems: average growth rates are often low, intra-specific variation is high and the reproducibility of the results is low. Hence, the importance of factors other than feeding may have been neglected. Important in this respect

may be the effects that physical and chemical properties of the water surrounding the sponges exhibit on sponge physiology. For instance, it is known that sudden changes in temperature and salinity can have profound effects on sponges (Storr 1964, McMillan 1996). However, these factors can be controlled relatively easy in laboratory systems. Only few studies have addressed the importance of chemical properties of (sea)water in relation to sponge physiology. Francis et al. (1990) reported that the freshwater sponge Ephydatia fluviatilis required concentrations of the ions Cu2+, Zn2+ and Co2+ between 10-9 and 10-8 M for optimal growth in culture, indicating that the trace element composition of the surrounding water in a sponge culture system needs to be clearly defined. More recently, Krasko et al. (2002) discovered that ferric iron (Fe3+) plays an important role in sponge morphogenesis. Addition of Fe3+ to primmorph cultures of the Mediterranean sponge Suberites domuncula resulted in an up-regulation of genes encoding silicatein (an enzyme involved in spicule formation), ferritin (iron binding protein) and septin, which is associated with cell proliferation). Hence, the availability of iron (in particular Fe3+) may be a key factor in determining sponge growth. In this study, we tested different feeding regimes on three species of Mediterranean sponges, Dysidea avara Schmidt, 1862, Chondrosia reniformis Nardo, 1833 and Acanthella acuta Schmidt, 1862. In addition, we supplemented the cultivation systems with different concentrations of ferric iron. It was shown that iron promoted the formation of oscules in all sponges when administered appropriately.

498

Materials and methods Sponges and sponge culture systems

Specimens of three species of sponges (Dysidea avara, Chondrosia reniformis and Acanthella acuta) were collected by SCUBA divers in the Limski Fjord (Croatia, Mediterranean) in spring 2003. They were transported to the laboratory in Wageningen, The Netherlands, in filtered cool-containers as described by Sipkema et al. (2006). There, sponges were maintained in a 1500 liter tank equipped with a novel DYMICOTM filtration system (EcoDeco BV, The Netherlands; patent WO 02069701) at a temperature of 15C and a salinity of 35. This culture system will be referred to as System I (Fig. 1A). After six months, most of the sponges were transported to another system at the Dutch Fishery Institute (RIVO, IJmuiden, The Netherlands). Here, the sponges were positioned in 30 liter aquaria that were integrated within an 80,000 liter recirculating aquaculture system in which a brood stock of sole (Solea solea) is being maintained. This system is operated with natural seawater obtained from the North Sea. The water flow rate through sponge tanks was 70 ml/min (giving a residence time of ~7 hours). We kept the sponges at a temperature of 18-19C and a salinity of 32. A schematic drawing (Fig. 1B) shows the position of the sponge tanks within the system. This culture system will be referred to as System II.

Feeding experiments and supplementation with iron

In System I, sponges were fed with the marine diatom Phaeodactylum tricornotum, which were cultured according to Osinga et al. (2003). The algae were frozen immediately after cultivation. Every day, a batch of frozen algae was thawed and diluted with seawater from the system. For the execution of these feeding experiments, a 3 liter culture vessel was integrated into System I: it received a continuous input of water from the main system to which the algae feed was added. The effluent of this feeding tank was led back into the main tank. The diluted algae suspension was added slowly using a peristaltic pump, thus creating a continuous input of algae food to the sponges in the feeding tank that was equivalent to an end concentration of ~2 x 105 cells/ml-1. This concentration was reported to promote growth of the tropical Demosponge Pseudosuberites andrewsi (Osinga et al. 2003). Ferric iron (Fe3+) was added to the main tank of System I as ferric citrate, hereby applying a final concentration of 30 M, which is the concentration reported by Krasko et al. (2002) to stimulate morphogenesis in sponge primmorphs. The total iron concentration (Fe2+ and Fe3+) in System I was measured monthly using ICP-OES/ICP-MS. The iron concentration was re-adjusted to 30 M by adding ferric citrate when the measured values indicated a lower concentration. In System II, two feeding methods were applied. Firstly, it was attempted to feed the sponges by operating the feeding tank preceding the sponge tank (see Fig. 1) as a continuous photobioreactor: nitrate (added as KNO3; final concentration: 500 M) and phosphate (added as NaH2PO4; final concentration: 50 M) were continuously added to this bioreactor to enhance the growth of phototropic
Fig. 1: A. Schematic drawing of System I (including feeding system). Water is re-circulated continuously through the sand bed (direction of water flow is indicated by small arrows), in which a DYMICOTM reactor has been mounted. B. Schematic drawing of System II (see text for explanation).

microorganisms present in the system. Secondly, we applied a food regime based on Phaeodactylum tricornotum as described for System I. In this trial, iron was supplied by continuous addition of a stock solution of ferric citrate to the sponge tank using a peristaltic pump. Initially, iron was added up to a final concentration of 30 M in the sponge tank. Two days later, the final concentration was lowered tenfold to 3 M. Due to the dimensions of System II, these additions had no detectable effect on the iron concentration in the entire system throughout the experimental period. The biomass of D. avara and C. reniformis in System II was monitored by measuring buoyant weight of sponge explants (Osinga et al. 2001). Explants were prepared by cutting fragments of the sponges using razor knifes. The fragments were positioned onto perspex slides using nylon fishing line. Most fragments attached to their supports within one week.

Results
Continuous growth of sponges could not be obtained (Fig. 2). In System 1 (Dysidea avara only), the sponges retained

499

outer appearance; all explants of D. avara had active oscules. Continuous enrichment of the system seawater in System II with nitrate and phosphate resulted in a feed that consisted for almost 100% of the marine prasinophyte Tetraselmis suecica. This feed appeared to have a negative impact on the sponges. After switching from the Tetraselmis-dominated feed to frozen Phaeodactylum, most sponges got a much healthier outer appearance (a clearer surface, more fleshy tissue and, in the case of D. avara: a switch from non-pumping to pumping oscules). These observations are nicely illustrated in Fig. 3, showing five consecutive images of a specimen of A. acuta, photographed throughout the entire experimental period (both in System I and System II). The sponge had a nice outer appearance in System I (Figs. 3A, B), was in a stage of deterioration when fed with Tetraselmis (Fig. 3D) and recovered completely when fed with Phaeodactylum tricornotum (Fig. 3E). Although addition of ferric iron did not promote growth, visual observations on the sponges in System II showed a strong effect of Fe3+ on their physiology: addition of 30 M Fe3+ had a negative impact on D. avara, which ceased its pumping activity immediately after the addition of the ions. After lowering the Fe3+ concentration tenfold, all explants of D. avara retained their pumping activity, oscules were clearly visible. Most explants of C. reniformis also developed actively pumping oscules (Fig 4A-B) after lowering the Fe3+ addition from 30 M to 3 M. In nature, this species exhibits clear oscules (Fig. 4C), a phenomenon that we have never observed in our tank systems. The third species studied (A. acuta) also showed active oscules after changing the Fe3+ concentration from 30 M to 3 M (Fig. 3E). Discussion We conclude that ferric iron promotes the formation and activity of the aquiferous system in C. reniformis, D. avara, and A. acuta. Our data confirm observations on primmorphs described by LePennec et al. (2003) who reported that in addition to spiculogenesis and cell proliferation, Fe3+ also induces channel formation in primmorphs. However, the concentration of Fe3+ causing the positive effect on functional sponges in this study is tenfold lower than the concentration that was applied on primmorphs (Krasko et al. 2002, LePennec et al. 2003). In those experiments, a very strong upregulation of gene expression was found. The rapid change in our experiment, from a very low (undetectable) concentration to 30 M has probably caused an over-expression of ferric iron-regulated genes, which may not have been in agreement to other conditions in the culture system, such as food availability, or the presence of other trace elements. Such an imbalance in nutrients may negatively affect the physiology of an organism: Mandalam and Palsson (1998) proposed that a similar imbalance might inhibit the growth of the green microalga Chlorella vulgaris. Hence, we assume that an iron up-shock caused the sponges in System II to shut down. The Fe3+ levels in System I exhibited strong fluctuations (Table 1), which may have had similar negative effects on the sponge explants. It can be concluded that Fe3+ has a beneficial effect on sponges when administered appropriately, and that rapid changes in concentration should be prevented.

Fig. 2: Development of sponge biomass (expressed as mg or g buoyant weight) in System I (2A: Dysidea. avara) and System II (2B: Dysidea avara; 2C: Chondrosia reniformis).

their original weight throughout a period of four months. In System II, all explants lost some weight. Hence, the feeding regimes applied and the addition of Fe3+ did not promote growth. However, differences in response to the various treatments of the sponges were observed. Notwithstanding the lack of growth, sponges in System I exhibited a reasonably healthy

500

Fig. 3: Change in outer appearance of a specimen of Acanthella acuta. A. upon arrival from the Mediterranean. B. after three months of cultivation in System I. C. Upon arrival in System II after having been in System I for six months. D. After three months of cultivation in System II (feeding with Tetraselmis-dominated food). E. After nine months of cultivation in System II (feeding with frozen Phaeodactylum tricornotum, addition of 3 M Fe3+). Arrows indicate the position of oscules.

Fig. 4: Chondrosia reniformis. A. In nature (Mediterranean). B. Explant before the addition of iron. C. Explant after the addition of iron (the arrow indicates the position of the oscule).

In our experiments, iron was added as ferric citrate, because the chelating effect of the citrate moiety increases the bioavailability of the ferric ion. Other potential chelating agents have been reported in relation to sponge physiology: Belas et al. (1992) found that low molecular weight humic substances stimulated the growth of the freshwater sponge Ephydatia fluviatilis. Our results may shed a new light on these findings:

by acting as a chelating agent, these compounds will increase the bioavailability of Fe3+ and will thus stimulate growth in general and formation of channel structures in particular. A similar mechanism has been proposed for collagen production in sponges (Bavestrello et al. 2003): ascorbic acid increases the bioavailability of silicon, which promotes the expression of genes involved in collagen biosynthesis.

501

Table 1: Concentrations of iron in System I. Date 14 August 2003 2 October 2003 20 January 2004 8 March 2004 7 May 2004 7 July 2004 Concentration (M) 19.9 0.3 0.0 0.0 15.6 18.6

Table 2: Iron concentrations in natural seawater samples. Location Blanes (Spain), Mediterranean Limski Fjord (Croatia), Mediterranean Eastern Scheldt (Netherlands) Florida Keys (USA) Concentration (M) 16 22 16 72

In natural seawater, iron concentrations are usually low, some oceanic concentrations reported are 1 nM (Kennish 1994) and 36 nM (Brown et al. 1993). This is much lower than the concentrations used by Krasko et al. (2002) and in the present study. However, iron concentrations in coastal waters may often be considerably higher than the values measured in the oligotrophic ocean (see for instance the data in Table 2: iron concentrations measured in seawater samples from four near-shore locations). The potential role of ferric iron as a driver for population density and size distribution of sponges in nature is a subject that deserves further study. Our results show that controlled production of sponges in closed aquaria remains a challenge for marine biologists and biotechnologists, although progress towards a better understanding of the requirements of sponges in aquaria has been achieved. The observed preference of sponges for Phaeodactylum tricornotum as a food source is in agreement with previous works (Osinga et al. 2003, Sipkema et al. 2006). Appropriate addition of ferric iron to the cultivation system in combination with Phaeodactylum tricornotum as a food source seems a good starting point for further studies. In addition to iron and silicon, the role of other trace elements should be investigated. Another factor that has been neglected is the physical environment around a sponge, such as the effect of hydrodynamics on food uptake. Experiments in which these factors are included are currently in progress.

Acknowledgements
RO was supported by The Dutch Ministry of Economic Affairs (BIOPARTNER program).

References
Arndt W (1933) Haltung und Aufsucht von Meeresschwmmen. In: Abderhalden E (ed). Handbuch der Biologischen Arbeitsmethoden. Vol. I: Methoden der Meeresbiologie, Abteilung IV. Urban & Schwarzenberg, Berlin, pp. 443-464 Bavestrello G, Benatti U, Cattaneo-Vietti R, Cerrano C, Giovine M (2003) Sponge cell reactivity to various forms of silica. Micr Res Tech 62: 327-335 Belarbi EH, Domnguez MR, Cern Garca MC, Contreras Gmez A, Garca Camacho F, Molina Grima E (2003) Cultivation of explants of the marine sponge Crambe crambe in closed systems. Biomol Eng 20: 333-337

Belas FJ, Francis JC, Poirrier MA (1992) Significance of small organic chelators in laboratory cultures of Ephydatia fluviatilis (Porifera: Spongillidae). Trans Am Microsc Soc 11: 169-179 Brown J, Colling A, Park D, Phillips J, Rothery D, Wright J (1993) Seawater: its composition, properties and behaviour. Pergamon Press, Oxford de Caralt S, Agell G, Uriz M-J (2003) Long-term culture of sponge explants: conditions enhancing survival and growth, and assessment of bioactivity. Biomol Eng 20: 339-347 Duckworth AR, Pomponi SA (2004) Relative importance of bacteria, microalgae and yeast for growth of the sponge Halichondria melanadocia (de Laubenfels, 1936): A laboratory study. J Exp Mar Biol Ecol 323: 151-159 Francis JC, Bart L, Poirrier MA (1990) Effect of medium pH on the growth rate of Ephydatia fluviatilis in laboratory culture. In: Rtzler K (ed). New Perspectives in sponge biology. Smithsonian Institution Press, Washington, DC, pp. 485-490 Kennish MJ (1994) Practical Handbook of Marine Science. CRC press, Boca Raton, FL Kinne O (1977) Cultivation. 3. Porifera. In: Kinne O (ed). Marine Ecology. Vol 3, part 2, pp. 627-641 Krasko A, Schrder HC, Batel R, Grebenjuk VA, Steffen R, Mller IM, Mller WEG (2002) Iron induces proliferation and morphogenesis in primmorphs from the marine sponge Suberites domuncula. DNA Cell Biol 21: 67-80 Le Pennec G, Perovic S, Ammar MSA, Grebenjuk VA, Steffen R, Brmmer F, Mller WEG (2003) Cultivation of primmorphs from the marine sponge Suberites domuncula: morphogenetic potential of silicon and iron. J Biotechnol 100: 93-108 Mandalam RK, Palsson BO (1998) Elemental balancing of biomass and medium composition enhances growth capacity in high-density Chlorella vulgaris cultures. Biotechnol Bioeng 59: 605-611 McMillan SM (1996) Starting a successful commercial sponge aquaculture farm. Publication No. 120. Center for Tropical and Subtropical Aquaculture, Waimanalo Nickel M, Brmmer F (2003) In vitro sponge fragment culture of Chondrosia reniformis (Nardo, 1847). J Biotechnol 100: 147-159. Osinga R, Tramper J, Wijffels RH (1999) Cultivation of marine sponges. Mar Biotechnol 1: 509-532 Osinga R, Kleijn R, Groenendijk E, Niesink P, Tramper J, Wijffels RH (2001) Development of in vivo sponge cultures: particle feeding by the tropical sponge Pseudosuberites aff. andrewsi. Mar Biotechnol 3: 544-554

502

Osinga R, Belarbi EH, Molina Grima E, Tramper J, Wijffels RH (2003) Progress towards a controlled culture of the marine sponge Pseudosuberites andrewsi in a bioreactor. J Biotechnol 100: 141146 Sipkema D, Yosef NAM, Adamczewski M, Osinga R, Mendola D, Tramper J, Wijffels RH (2006) Hypothesized kinetic models for

describing the growth of globular and encrusting demosponges. Mar Biotechnol 8: 40-51 Storr JF (1964) Ecology of the Gulf of Mexico commercial sponges and its relation to the fishery. Special Scientific Report - Fisheries no. 466, United States Department of the Interior: 1-73 Thomassen S, Riisgrd HU (1995). Growth and energetics of the sponge Halichondria panicea. Mar Ecol Prog Ser 128: 239-246

Porifera research: Biodiversity, innovation and sustainaBility - 2007

503

Ecology and physiology of mesohyl creep in Chondrosia reniformis

Lorenzo Parma(1*), Dario Fassini(1), Giorgio Bavestrello(2), Iain C. Wilkie(3), Francesco Bonasoro(1), Daniela Candia Carnevali(1)
Dipartimento di Biologia Luigi Gorini, Universit degli Studi di Milano, via Celoria 26, 20133 Milano. Italia. lorenzo.parma@unimi.it (2) Istituto di Scienze del Mare, Universit Politecnica delle Marche, via Brecce Bianche, 60131, Ancona. Italia (3) Department of Biological and Biomedical Sciences, Glasgow Caledonian University, 70 Cowcaddens road, Glasgow G4 0BA, Scotland
(1)

Abstract: Chondrosia reniformis is a marine demosponge which consists mainly of a collagenous tissue known as mesohyl. Mesohyl is a very adaptable material: for example, it reacts to mechanical stimulation by stiffening. It has also been observed in nature that parts of sponges undergo slow elongation and attenuation resulting in the formation of propagules, a process which provides a means of asexual reproduction and dispersion. This phenomenon of mesohyl creep (slow progressive deformation) appears to be initiated by the fragmentation of the substratum to which the sponge is attached, and can be interpreted as a possible response to gravity. The aim of the present work was to provide more information on the creeping phenomenon of C. reniformis specimens in nature and to investigate the possible control of the phenomenon by the sponge itself. These aspects have been addressed using an integrated approach which consists of: 1) a field survey; 2) an experimental field study; 3) and an experimental laboratory study. Field survey: the phenomenon was explored in parallel in three different regions of the Italian coasts. The behaviour of specimens from the different areas was correlated with the nature of the substratum, the water temperature and the trophic conditions. Experimental field study: the creeping phenomenon was induced experimentally in the field by attaching weights (2 to 40 g) to the sponges. The consequent elongation, which was similar to the natural phenomenon, demonstrated that gravity is involved in creep. Experimental laboratory study: the effect on mesohyl tensility of changing parameters (temperature, salinity) was tested in physiological experiments. The results confirmed a close relationship between water temperature and mesohyl stiffness. Keywords: collagenous tissue, creep, mechanical properties, mesohyl, propagule

Introduction
Chondrosia reniformis Nardo 1847 is a common demosponge (Lazoski et al. 2001) which lives on shaded rocky cliffs or caves at a depth of 1 to 50 m, and lacks siliceous spicules or the reinforcing spongin fibres present in many other members of the phylum Porifera (Garrone et al. 1975, Harrison and De Vos 1991). Its generic name is due to the cartilage-like consistency of its thick and dense collagenous cortex (ectosome) (Garrone et al. 1975), which is crossed only by branched inhalant channels (Bavestrello et al.1988). The sponge can generate long, attenuated outgrowths which extend from the parental body for up to 3 m (normal sponges being on average 5-20 cm in their maximal diameter) then detach and form a new sponge (propagule). These outgrowths can extend downwards, as if under the force of gravity. Less frequently small attenuated portions extend horizontally (personal observation). This capacity for creep (slow and progressive deformation) and elongation is known also in other species as Oscarella lobularis (Schmidt, 1862) (Sar and Vacelet 1973) and Chondrilla nucula Schmidt, 1862

(Gaino and Pronzato 1983), all of which lack an organised endoskeleton of either spicules or spongin fibres. This creeping phenomenon has been interpreted in different ways: a) It has been regarded as a form of asexual reproduction (Connes 1967, Gaino and Pronzato 1983, Bavestrello et al. 1998). Asexual reproduction by budding occurs in several sponges (see Simpson 1984 for a detailed review), and sometimes is accompanied by the development of long retractile filaments which usually do not detach from the parent body (Fell 1994). b) Bond and Harris (1988) suggested that these dynamic deformations can represent a sort of localised locomotion by parts of the sponge, possibly preceding asexual reproduction. c) They may be responses to environmental changes: the whole sponge body can slowly flatten and slide under compressive stress (Garrone et al. 1975) or stretch out under tensile stress such as develops when part of the substratum to which the sponge is adhering breaks off (Sar and Vacelet 1973). Recent studies demonstrated that Chondrosia reniformis is able to react to mechanical stimulation by stiffening its collagenous body. In fact when previously undisturbed specimens of C. reniformis in the sea or laboratory aquaria

504

are touched repeatedly, they feel softer the first time they are touched than on second and subsequent occasions. Morphological studies of C. reniformis had not revealed the presence of enough potentially contractile cells to account for this phenomenon (Bonasoro et al. 2001, unpublished observations), so the response to mechanical stimulation is not due to contractile structures, but, as evidenced in other studies, is due to changing in mechanical properties of the extracellular matrix which are under cellular control (Wilkie et al. 2006, Wilkie et al. 2004a). A well studied model showing a similar feature is the mutable collagenous tissue (MCT) of echinoderms, the variable tensility of which is neurally modulated and which is involved throughout the phylum in the energy-sparing maintenance of posture and in the rapid detachment of anatomical structures at autotomy (Trotter et al. 2000, Wilkie 2001, 2005). Recent evidence suggests that connective tissue mutability is not a unique feature of echinoderms but is an adaptive strategy present also in primitive animals such as sponges (Wilkie et al. 2004b). The aim of the present work is to provide more information on the creeping phenomenon of C. reniformis specimens in nature in order to identify underlying mechanisms and how these are controlled. These aspects are being investigated by means of an integrated approach, which consists of: 1) a field survey; 2) an experimental field study; 3) and an experimental laboratory study.

Fig. 1: Inducing creep in intact sponges. Photograph of sponge whose attachment point to substratum was cut through under central part, allowing a line (arrow) to be passed round it and tied. A weight or float was then attached to cord (arrowhead; label identifying specimen VE).

Materials and methods

Field survey: The general morphology of the specimens of C. reiniformis and the creeping phenomena in natural condition were observed at two Italian locations (Trave Central Adriatic sea, Paraggi Eastern Ligurian sea) characterized by different trophic and edaphic conditions. In these places several specimens were photographed in different periods using a digital photocotocamera (Nikon Coolpix 8400 in Ikelite housing). The area and the outline of the specimens were calculated. Experiments on intact animals: in the same areas and in Bergeggi, Western Ligurian sea, the creeping phenomenon was artificially induced by attaching either lead weights (540 g), (see Fig. 1 for attaching methods) or floats (about 4 g of buoyancy, rather spherical fishing float), or by dislodging part of the substratum to which the sponge was attached. Experiments on isolated tissue samples: specimens of C. reniformis were collected by SCUBA divers at Bergeggi on the Italian Ligurian coast, then transported to the University of Milan and maintained in 50 l tanks of artificial seawater at 14-16oC. Beam-shaped samples 2.5 x 2.5 x 15 mm in size were cut from both the ectosome and choanosome regions and attached to a glass coverslip using cyanoacrylate cement, with exactly 10 mm projecting from the edge of the coverslip. A lead pellet (weighing 0.056 g) was attached to the free end of some samples using cyanoacrylate cement. The samples were left untouched overnight, then placed vertically in different test solutions, so that the lead pellet subjected them to a tensile force (Fig. 2).

Fig. 2: Inducing creep in isolated tissue samples. A Beam shaped samples (2.5 x 2.5 x 15 mm) of both ectosome and choanosome were excised. B. Each sample was attached by means of cyanoacrylate cement to a coverslip with exactly 10 mm protruding. C. A 0.056 g lead pellet was fixed to sample with cyanoacrylate cement. After an overnight destiffening the samples were vertically placed in the test solutions.

Results
Specimens at Paraggi (Ligurian sea) live on a compact rocky cliff (maximum depth 20 m) and have compact and regular forms each covering an area ranging from 3 to 65 cm2 (30,2320,18 cm2; n=7). During the observation period they showed insignificant variation in either shape and size. Specimens at Trave (Adriatic sea) live on a substratum mostly constituted by a bed of mussel shells (maximum

505

depth 8.5 m) and have complex, lobated shapes and cover areas from 200 to 400 cm2 (mean 337,17 152,66 cm2; n=5), during the experimental time they showed a great changing in shape, often making difficult to recognize specimens from one month to the next. Specimens at Bergeggi live on a substratum constituted of both solid rock and a more organic and loose material (maximum depth 5 m). Their sizes and shapes are similar to those of the Paraggi specimens (covering area 16,7 to 62 cm2; mean 40,117,14; n=7). Though the variation in both shape and size during the experimental time was insignificant, they showed more dynamicity than the Paraggi specimens, 3 to 7 new creeping phenomena being observed every months. All ongoing creeping phenomena slowed down as water temperature fell and attenuated regions of sponges shortened by 50 - 66% by the time the temperature reached the minimum value (Fig. 3). During Spring, as the water temperature rose, all the attenuated regions started to elongate again.

Fig. 3: Attenuated region of one specimen. The creeping phenomenon seemed to vary with sea temperature.

Experiments on intact animals

In Paraggi part of the substratum to which the sponges were attached was dislodged, leaving 5 specimens with a part of the body subjected to the force of gravity. All 5 specimens underwent the creeping phenomenon, though at different rates (from 11 cm elongation in 3 months to 95 cm in 1 month). The application of a lead weight always induced creep. The phenomenon was influenced by the weight of the lead pellet, the distance between the two attachment points of the sponge to the substratum and the diameter of the line. Results very similar to the natural creeping phenomenon were observed using a small weight (about 5 g) and an elastic strip (instead of a line) with a width of about 1 cm. The application of a weight accelerated the creeping process. The formation of the attenuated region and the final detachment of the propagule occurred generally over a few days (Fig. 4). A weight greater than 15 g caused the line to pass through the sponge body without, however, separating the sponge into two parts, since the wound closed up soon after the passage of the line, leaving a visible scar (Fig. 5). The application of a float induced the same creeping phenomenon but directed upwards.

Fig. 4: Timelapse photographs of induced creep and propagule detachment.

Experiments on isolated tissue samples

The behaviour of the isolated samples was comparable to that of whole animals (Fig. 6). Samples without an attached lead pellet did not change in length over 3 days, whereas weighted samples elongated by up to 3 times their original length (all samples reached the maximum value in the experimental apparatus (Fig. 7). Weighted samples treated with distilled water showed no elongation (Fig. 8). There was a positive correlation between the amount of elongation of weighted samples and temperature (Fig. 9).Removal of the lead pellet from weighted samples that had undergone elongation was followed by reshortening of the samples.

Fig. 5: Scar (arrow) left after cord has passed through body.

506

Fig. 6: Creeping of isolated samples weighted with 0.056 g lead pellets. A. time 0; B. 24h.

Fig. 7: Mean elongation after 3 days of isolated samples (n=5). A. samples weighted with 0.056 g lead pellets; B. samples with no attached weight; ec, ectosome; ch, choanosome.

Fig. 8: Mean elongation after 24 h of isolated samples weighted with 0.056 g lead pellets (n = 5; bars = standard deviations). asw, artificial seawater; dw, distilled water; ec, ectosome; ch, choanosome.

Discussion
Specimens of C. reniformis from Trave and Paraggi showed completely different characteristics, both in terms of size and shape, and in terms of frequency of natural creeping. Specimens at Paraggi are smaller and show a more compact shape while those observed at Trave are larger and lobate. During the observation time the former showed no

dynamicity, whereas the latter showed much changing in shape, often making difficult to recognize specimens from one month to the next. In these two regions the type of substratum is different. At Paraggi the cliff is solid and compact, while at Trave the substratum consists mostly of mussel shells. The natural instability of the latter type of substratum could at least

507

Fig. 9: Mean elongation after 24h of weighted (0.056 g) samples at different temperatures (n = 5; bars = standard deviations). asw, artificial seawater; ec, ectosome; ch, choanosome.

facilitate the frequent creeping shown by the Trave specimens. Regarding differences in size, it may be relevant that Paraggi has an oligotrophic environment with a low sedimentation rate, while at Trave the environment is eutrophic with a high sedimentation rate. The observation that the large sponges of Trave are more dynamic than the small sponges of Paraggi suggests that the size of the sponge can influence the probability that creeping will take place. It appears that the Paraggi specimens are not intrinsically less dynamic than those at Trave, since fast creeping was observed in 5 Paraggi specimens after dislodgement of the substratum beneath them. It is feasible that the adaptive significance of creeping is that it reduces the size of large animals, large size being a disadvantage possibly because it increases the chance that a sponge will inhale its own waste products, which could have a deleterious effect on its metabolism. Such waste products could conceivably be the signal that triggers the onset of creeping in large animals (Fry 1979). However, in Bergeggi, where the substratum comprises both solid and loose rocks, and there were frequent creeping events (3 to 7 new per month), there was no correlation between the size of the sponge and its dynamicity. Most of the sponges at this site, which occur at a depths down to 3 m, are exposed to strong wave action which frequently causes the detachment of part of the sponge from the substratum. Thus the frequency of the creeping phenomenon may be more dependent on environmental factors than on intrinsic factors, such as size. Both detachment of part of the substratum under a sponge, which imposed a tensile force on the freed part of the sponge, and the attachment of weights or floats, which also applied tension, resulted in creeping phenomena. Though such creep could be an entirely passive response to extraordinary tensile forces, we suspect that the animal may be able to exert some level of control over the phenomenon (such as modulation of the rate of creep), since there is strong evidence that the stiffness of the collagenous mesohyl of C. reniformis is variable and under direct cellular control (Wilkie et al. 2004a, 2006). We provided evidence for the latter in these experiments, since distilled water completely blocked

the elongation of weighted tissue samples. This treatment kills cells by osmotic lysis, possibly resulting in the release of a factor that directly stiffens the collagenous mesohyl (Wilkie et al. 2006). Our observations are consistent with the idea that temperature is positively correlated with the rate of creep. This may explain the evidence that creeping events in C. reniformis are more frequent in summer than at other times of the year (personal observation). We do not know yet whether water temperature affects directly the mechanical properties of the extracellular matrix or if the observed differences are due to a cell-mediated mechanism. We also showed that removal of the lead pellet from weighted tissue samples that had elongated resulted in reshortening of the samples. This is also shown by the attenuated filament of intact sponges once the propagule is detached. Since there are unlikely to be enough contractile cells in the mesohyl to explain such tissue retraction (Bonasoro et al. 2001), it appears to represent a passive elastic mechanism the molecular basis of which needs to be investigated, although the possible active involvement of cells cannot be dismissed completely. The creeping phenomenon is just one manifestation of the mechanical adaptability of the collagenous mesohyl, which makes up the bulk of the body of C. reniformis. This tissue can also undergo rapid and reversible changes in stiffness, which depend on the direct cellular control of interactions between components of the extracellular matrix of the mesohyl, such as the collagen fibrils and the molecules that interconnect these fibrils (Wilkie et al. 2004a, 2006). At this stage the relationship between creeping and variable stiffness is not clear. Both, however, may depend on what is likely to be a primitive connective tissue feature, which is the absence of strong chemical bonds between the molecules responsible for cohesion between the collagen fibrils, a feature that the mesohyl shares with the mutable collagenous tissue of echinoderms (Wilkie et al. 2004b).

References
Bavestrello G, Burlando B, Sar M (1988) The architecture of canal system of Petrosia ficiformis and Chondrosia reniformis studied by corrosion casts (Porifera, Demospongiae). Zoomorphology 108: 161-166 Bavestrello G, Cerrano C, Calcinai B, Benatti U, Cattaneo-Vietti R, Favre A, Giovine M, Lanza S, Pronzato R, Sar M (1998) Body polarity and mineral selectivity in the demosponge Chondrosia reniformis. Biol Bull 195:120125 Bonasoro F, Wilkie IC, Bavestrello G, Cerrano C, Candia Carnevali MD (2001) Dynamic structure of the mesohyl in the sponge Chondrosia reniformis (Porifera, Demospongiae). Zoomorphology 121: 109-121 Bond C, Harris AK (1988) Locomotion of sponges and its physical mechanism. J Exp Zool 246:271-284 Connes R (1967) Structure et devloppement des bourgeons chez leponge siliceuse Tethya lyncurium Lamarck: recherches exprimentales et cytologiques. Arch Zool Exp Gen 108:157-195 Fell PE (1994) Porifera. In: Adiyodi KG, Adiyodi RG (eds). Reproductive biology of invertebrates, vol VI B. Asexual propagation and reproductive strategies. Oxford and IBH Publishing, New Delhi, pp 144

508

Fry WG (1979) Taxonomy, the individual and the sponge. In: Larwood G, Rosen BR (eds). Biology and Systematic of colonial organism. Academic Press, London. pp 49-80 Gaino E, Pronzato R (1983) tude en microscopie lectronique du filament des formes tires chez Chondrilla nucula Schmidt (Porifera, Demospongiae). Ann Sci Nat Zool Paris 5:221-234 Garrone R, Huc A, Junqua S (1975) Fine structure and physicochemical studies on the collagen of the marine sponge Chondrosia reniformis Nardo. J Ultrastructure Res 52: 261-275 Harrison FW, de Vos L (1991) Porifera. In: Harrison FW, Ruppert EE (eds). Microscopic anatomy of invertebrates, vol. 2. Wiley Liss, New York. pp 29-89 Lazoski C, Sol-Cava AM, Boury-Esnault N, Klautau M, Russo CAM, 2001. Cryptic speciation in a high flow scenario in the oviparous marine sponge Chondrosia reniformis. Mar Biol 139: 421-429 Sar M, Vacelet J (1973) Ecologie des Dmosponges. In: Grass PP (ed). Trait de zoologie. Anatomie, systmatique, biologie: Spongiaires. Masson, Paris, 3: 462-576 Trotter JA, Tipper J, Lyons-Levy G, Chino K, Heuer AH, Liu Z, Mrksich M, Hodneland C, Dillmore WS, Koob TJ, Koob-Emunds MM, Kadler K, Holmes D (2000) Towards a fibrous composite

with dynamically controlled stiffness: lessons from echinoderms. Biochem Soc Trans 28: 357-362 Wilkie IC (2001) Autotomy as a prelude to regeneration in echinoderms. Microsc Res Tech 55: 369-396 Wilkie IC, Bonasoro F, Bavestrello G, Cerrano C, Candia Carnevali MD (2004a) Mechanical properties of the collagenous mesohyl of Chondrosia reniformis: evidence for physiological control. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Boll Mus Ist Biol Univ Genova 68: 665-672 Wilkie IC, Candia Carnevali MD, Trotter JA (2004b) Mutable collagenous tissue: recent progress and an evolutionary perspective. In: Heinzeller T, Nebelsick JH (eds), Echinoderms: Mnchen. Taylor & Francis, London. pp. 371-378 Wilkie IC (2005) Mutable collagenous tissue: overview and biotechnological perspective. In: Matranga V (ed). Echinodermata. Progress in molecular and subcellular biology, vol. 39. Springer, Berlin. pp. 219-248 Wilkie IC, Parma L, Bonasoro F, Bavestrello G, Cerrano C, Candia Candia Carnevali MD (2006) Mechanical adaptability of a sponge extracellular matrix: evidence for cellular control of mesohyl stiffness in Chondrosia reniformis Nardo. J Exp Biol 209: 44364443

Porifera research: Biodiversity, innovation and sustainaBility - 2007

509

Description of two new species of Acanthotetilla Burton, 1959 from NE Brazil, Southwestern Atlantic (Tetillidae, Spirophorida, Demospongiae)
Solange Peixinho(1*), Jlio Fernandez(1), Mara V. Oliveira(2), Simone Cares(2), Eduardo Hajdu(2)
Departamento de Zoologia, Instituto de Biologia, Universidade Federal da Bahia, Campus de Ondina, 40210 - 170, Salvador, BA, Brazil. peixinho@ufba.br (2) Museu Nacional, Universidade Federal do Rio de Janeiro, Quinta da Boa Vista s/ n, 20940 - 040, Rio de Janeiro, RJ, Brazil
(1)

Abstract: The genus Acanthotetilla Burton, 1959 is recorded for the first time in the South Atlantic, with two new species. Acanthotetilla rocasensis sp. nov. is described from a single specimen obtained at das Rocas Atoll, North-eastern Brazil, and is characterized by an incrusting habit, possession of two categories of oxeas and presence of small acanthoxeas (192 - 238 m). Acanthotetilla walteri sp. nov. is described from thirty three specimens collected at the northern sector of Bahia states coast, and is characterized by an incrusting habit, and possession of a second category of protriaenes, as well as a second category of rare, stout acanthoxeas. Seven species of Acanthotetilla are known now, and A. walteri sp. nov. is the only species known from a series of individuals. Key words: Brazil, Porifera shallow-water, Taxonomy, Acanthotetilla, A. rocasensis sp. nov., A. walteri sp. nov.

Introduction
Species of Acanthotetilla Burton, 1959 generally have porocalyces on the surface, an ectosome with bundles of oxeas and triaenes reinforced by acanthoxeas, and sigmaspires, usually abundant in the choanosome. Van Soest (1994) listed Acanthotetilla as an example of a sponge genus with discontinuous distribution. Only five species of Acanthotetilla Burton, 1959 were previously known - A. hemisphaerica Burton, 1959 (South Arabian Coast); A. enigmatica (Lvi, 1964; Inhaca Island); A. seychellensis (Thomas, 1973; Seychelles), from the Indian Ocean; A. gorgonosclera van Soest, 1977 (Barbados), from the Caribbean; and A. celebensis de Voogd and van Soest, 2007 (Sulawesi), from Indonesia. The knowledge of Brazils marine sponge biodiversity has increased steadily over the past two decades thanks to additional specialists working on this taxonomic group and the widespread use of direct collecting techniques, which greatly contributed to a marked expansion of existing poriferan collections (Hajdu et al. 1996, Muricy et al. 2006). This work reports on the first occurrences of Acanthotetilla for the South Atlantic Ocean, as part of an effort to describe the large quantities of material gathered through this increased taxonomic effort. Two new species from the Brazilian northeastern coast are described below.

Materials and methods

A single specimen was collected at Das Rocas Atoll. This atoll is located 260 km N-NE of Natal, and is only 5.5 km2

in area (Kikuchi and Leo 1997). Two sandy islands stay above water at high tide, Farol Island and Cemitrio Island. The specimen studied here was collected at Cemitrio tide pool, south of Cemitrio Island (Fig. 1A) in August 2002, at 3 m depth, during an inventory of Das Rocas Atoll marine sponges (Moraes et al. 2003, 2006). Northeastern Brazil has a straight and plain continental shelf 15 - 75 km wide and with maximum depth of 70 m (Lana et al. 1996). The largest part of the studied material originated from the northern sector of Bahia States continental platform (between rivers Jacupe and Joanes, in the Camaari municipality, approximately at 12o44.210 - 12o53.568 S and 38o04.092 - 38o16.110 W), and was collected during an integrated monitoring programme (1993 - 2005) of submarine outfalls belonging to two chemical industries. Twenty-four sampling stations were established (2000 - 2005) in this area. Dissociated spicules and thick section mounts were obtaining according to the protocols suggested by Mothes de Moraes (1985) for light microscopy and following Hajdu (1994) for scanning electron microscopy. Spicule data are based on 20 measurements for each dimension unless otherwise noted; means are underlined. Abbreviations used in this text are: UFBA (Universidade Federal da Bahia), POR (Porifera Collection), MNRJ (Museu Nacional, Universidade Federal do Rio de Janeiro), SEM (Scanning Electron Microscopy).

510

Fig. 1: Map showing the type localities for Acanthotetilla rocasensis sp. nov. (A) and Acanthotetilla walteri sp. nov. (B). A. Das Rocas Atoll, RN. Colour legend: black, sandy island; dark gray, tide pool and grooves; medium gray, algal crest; light gray, sandy deposits (lagoon); white, sandstone/algal reefs within the perimeter of the atoll; B. Northern sector of Bahia state littoral, Camaari, BA. I and II: areas under influence of two industrial submarine outfalls. Scale bar = 100 km.

Systematics
Class Demospongiae Order Spirophorida Bergquist and Hogg, 1969 Family Tetillidae Sollas, 1886 Genus Acanthotetilla Burton, 1959 Diagnosis: Tetillidae with megacanthoxeas as auxiliary megascleres (van Soest and Rtzler, 2002). Acanthotetilla rocasensis sp. nov. (Figs. 2 - 4) Material type: Holotype - MNRJ 6359, Cemitrio tide pool (351.928 S - 3349.097 W, das Rocas Atoll, Rio Grande do Norte state, Brazil), 3 m depth, coll. E. Hajdu, U. Pinheiro and M. V. Oliveira, 26.viii.2002. Diagnosis: Acanthotetilla with encrusting, possibly endolithic habit, two categories of oxeas thinner than 10 m, and acanthoxeas only 192 - 238 m long by 4.8 - 12 m thick. Description: The specimen is encrusting (2 x 2 cm by 0.5 2.0 mm thick) over and within calcareous sediment substrate. The consistency is soft and fragile, the surface bears papillae and is hispid, due to spicules projecting at least 0.5 mm beyond it. The color in vivo was yellow turning to light-beige after preservation in ethanol. Oscules and porocalices not apparent.

Fig. 2: Acanthotetilla rocasensis sp. nov. preserved specimen (MNRJ 6359, arrow). Holotype. Scale bar = 1 cm.

Skeleton: The skeletal architecture is radial, composed of bundles of oxeas and protriaenes, crossing the choanosome and piercing the surface. The subectosomal skeleton is reinforced by abundant acanthoxeas and sigmaspires spread between the bundles; the microscleres are considerably less common in the choanosome. In the latter, developing stages of the acanthoxeas can be seen.

511

Fig. 3: Acanthotetilla rocasensis sp. nov. A. Schematic representation of the skeletal architecture; B-G. Schematic representation of the spicules: protriane I (B), oxea I (C), oxea II (D), acanthoxea (E), young acanthoxea (F), sigmaspire (G). Scale bars= I - 200 m (A); II - 100 m (B-C); III - 50 m (DF); IV - 10 m (G).

Spicules (measurements in m): Megascleres: Protriaenes with fusiform rhabdome, tapering gradually until the distal end, length 582 - 829.0 - 1018 and thickness 2.4; cladome with clads of equal length 19.4 - 22.8 - 38.8 and thickness of 1.2 - 1.6 - 2.4 (n = 40). Oxeas I, large, slender, abundant, tapering gradually to a very sharp apex, length 689 - 939.3 - 1193 and thickness 1.2 - 9. Oxeas II, smaller, slender, lighty curved at the central region and tapering gradually to a very sharp apex, length 120 - 172.4 - 218 and thickness 2.4 - 3.2 4.8 . Acanthoxeas lightly curved at central region, with sharp ends, length 192 - 216.0 - 238 and thickness 4.8 - 9.2 - 12; spines sharp, 2 - 4 m high, robust and pointing to the center of the spicule. Growth forms are smooth and/or bear thin, straight spines with an expanded central region. Microscleres: Sigmaspires, length 10 - 11.6 - 15. Ecology and distribution: Das Rocas Atoll is under the influence of sub-equatorial surface waters which render its water temperature warm throughout the year, around 27C. Cemitrio tide pool is one of the largest pools in the atoll, being over 50 m maximum diameter, and reaching over 4 m depth at certain points. Overall, the temperature in this pool does not rise considerably at low tides, with the exception of the smallest and shallowest branches. Over 500 sponges were collected in this Marine Biological Reserve already, but only a single specimen of A. rocasensis sp. nov., has been found, which suggests the species may be very rare. It is unclear whether or not the species was perforating the substrate. The studied specimen encrusts a piece of calcareous algalic substrate, but at certain points it is seen inside little crevices which might have been excavated by this or by another organism. As a consequence, the species may

Fig. 4: Acanthotetilla rocasensis sp. nov. SEM of the microscleres of the holotype: (A) acanthoxea, B) detail of spines of acanthoxeas, (C) sigmaspires. Scale bars = 50 m (A), 10 m (B), 5 m (C).

512

Table 1: Spicule measurements in m (means underlined) of all known species of Acanthotetilla.

be inconspicuous, and perhaps not so rare as indicated by the scarce material available for study. Etymology: The name rocasensis is derived from its typelocality, das Rocas Atoll. Remarks: Acanthotetilla rocasensis sp. nov., differs from all other species of Acanthotetilla by its possession of two categories of oxeas, the smallest of which is 120 - 218 m in length, and by its small acanthoxeas in a single size category, 192 - 237.6 m long. Acanthotetilla seychellensis is the only other species of Acanthotetilla with a second category of smaller oxeas. Both species can be easily distinguished by the oxeas II, with are larger in A. seychellensis (740 - 1260 m) than in the new species (120 - 218 m; Table 1). The new species differs from A. gorgonosclera by external morphology, which in the latter resembles that of a tuber, and by much larger protriaenes. The distinction between the new species and A. enigmatica is clear from the latters considerably larger oxeas and protriaenes, and possession of anatriaenes. Additionally, Lvi (1964) reported A. enigmatica to be irregularly semiglobular, and to bear porocalices, two features not seen in the new species. The new species differs from A. hemisphaerica by the latters considerably larger oxeas and protriaenes, and possession of anatriaenes. Further, A. hemisphaericas acanthoxeas are the largest so far found in the genus. Acanthotetilla celebensis is known from a single, large, semiglobular/hemispherical individual (10 x 12 cm), which apart from this remarkably distinct habit, still possesses only a single category of large oxeote megascleres and two size categories of acanthoxeas (de Voogd and van Soest 2007). It differs considerably from the new species reported upon here. Comparison to the second new species described will be offered below. Acanthotetilla walteri sp. nov. (Figs. 5 - 7) Type material: Holotype - UFBA 1902 - POR Camaari (12o47.083 S - 38o06.640 W, Bahia State, Brazil), 26 m depth, coll. W. Andrade, vii.2005; Paratypes - Camaari, Bahia State, Brazil, coll. W. Andrade - UFBA 1893 - POR (1247.333 S - 3806.167 W), 35 m depth, ii.1993; UFBA 1894 - POR, UFBA 1906 - POR, UFBA 2020 - POR (1250.383 S 3811.368 W), 23 m depth, vii.2003; UFBA 1895 - POR, UFBA 2021 - POR, UFBA 2031 - POR (1245.827 S 3806.568 W), 22 m depth; UFBA 2022 - POR (1250.012 S - 3810.112 W), 31 m depth, ii.2004; UFBA 1896 - 1897 - POR, UFBA 1901 - POR, UFBA 2023 - 2030 - POR, UFBA 2032 - 2035 - POR (1250.383 S - 3811.368 W), 23 m depth, iii.2005; UFBA 1898 - POR (1247.970 S - 3807.355 W), 28 m depth, iii.2005; UFBA 1899 - POR, UFBA 1900 - POR (1244.995 S - 3804.092 W), 28 m depth, iii.2005; UFBA 1903 - POR, UFBA 1904 - POR (1247.083 S - 3806.640 W), 26 m depth; UFBA 1905 - POR, UFBA 2036-2038 - POR (1247.970 S - 3807.355 W), 28 m depth, vii.2005. Diagnosis: Acanthotetilla with encrusting to endolithic habit, no anatriaenes, two categories of protriaenes, and two categories of acanthoxeas (the larger and always present with

I: 966-1157.3-1372 28.8-38.4-54 II: 308-684.0-1050 3.6-3.6-3.6 (clads) I: 28.8-38.4-54 5.4-6.6-7.2 II: 18-40.3-57.6 1.8-3.4-5.4

I: 238-297.1-378 28-28.0-28 II: 100.0 x 20 192-216-237 4.8-9.2-12 212-278.4-322 4-8.0-9 211-225.8-244 16-19.5-23

742-995.1-1232 5.4-13.0-21

A. walteri sp. nov.

I: 688-939.3-1193 1.2-5.6-9 II: 120-172.4-218 2.4-3.2-4.8 582-829.0-1018 2.4

(clads) 19.4-22.838.8 1.2 2.8-2.4

A. rocasensis sp. nov.

A. celebensis de Voogd and van Soest, 2007

1763-1923.0-2123 20-31.0-33

1977 22 (clads) 45-52.0-60

length not reported 5-5.5-6 (clads) 42-53.0-64 4-5.0-6 228-281.1-371 24-29.3-35 (clads) 50-70 1260-1450.0-1600 9-9.5-10 (clads) 55-70.0-80 9 325-372.3-414 40-46.4-60 Anatriaenes rhabdome Anatriaenes cladome Acanthoxea 3000

1260-1377.0-1540 4-5.3-9

(clads) 41-63.0-81 3-4.1-7

A. gorgonosclera van Soest, 1977

770-1216.0-1600 3-13.2-17

I: 1400-1516.0-1680 34-37.8-47 II: 740-1138.0-1260 6-10.9-14 360-907.0-1880 1.5-2.5-4

(clads) 28-38.0-50 1.5-3.0-4

A. seychellensis Thomas, 1973

2200-2960.0-3800 14-25.1-30

(clads) 50-68.0-95 9

2000-2500-3000 9

A. enigmatica Lvi, 1964

3100-3812.0-4400 24-29.9-35

1000-1920.0-2520 6-10.0-14

(clads) 30-38.0-56 6-8.0-10

A. hemisphaerica Burton, 1959

Sigmaspires

Protriaenes rhabdome

Protriaenes cladome

Species

Oxeas

9-11.0-13

8-9.6-11

8-10.3 -12

9-13.1-16

I: 300-405.0-442 20-25.0-33 II: 199-257.0-284 10-15.0-17 8-10.0-12

(clad) 40-45.0-50 34

1248 7

10-11.6-15

7-10.2-18

513

Fig. 5: Acanthotetilla walteri sp. nov. preserved specimen (UFBA 1902 - POR). Holotype. Scale bar = 1cm.

mean length ca. 300 m, and the smaller and very rare, ca. 100 m long). Description: The largest specimen was 47.5 mm in maximum diameter. The mean larger diameter observed in the entire population was only 20 mm and the mean smaller diameter was 13 mm. Specimens were encrusting (up to 3.5 mm thick) over and underneath small agglutinated carbonate/limestone gravel, and appear to be endolithic too. Consistency slightly compressible. Surface irregular and texture smooth to slightly hispid due to spicules projecting approximately 0.5 mm beyond the surface. All the samples were studied in the preserved state, presenting an external color ranging from dark to light gray, slightly lighter in the interior. Porocalices (0.2 - 2.5 mm) distributed at random.

Fig. 7: Acanthotetilla walteri sp. nov. SEM of the microscleres of the holotype: A. Acanthoxea II; B. Sigmaspire; C. Acanthoxea I. Scale bars = 40m (A), 1m (B), 100m (C).

Skeleton: The skeletal architecture is radial, composed of bundles of oxeas and protriaenes I and II, crossing the choanosome and piercing the surface. The subectosomal skeleton is reinforced by abundant acanthoxeas and sigmaspires scattered between the bundles. Microscleres are considerably less common in the choanosome, in which

Fig. 6: Acanthotetilla walteri sp. nov. A. Schematic representation of the skeletal architecture; BG. Schematic representation of the spicules: protriaene II (B), protriaene I (C), prodiaene (D), oxea (E), regular acanthoxea and young form (F), sigmaspire (G). Scale bars = I - 200 m (A); II 100 m (B-F); III - 20 m (G).

514

Table 2: Measurements, in m (means underlined), of spicules of the holotype and two paratypes of Acanthotetilla walteri sp. nov. Specimens Oxeas Acanthoxea Protriaenes rhabdome Protriaenes clads Sigmaspires Holotype UFBA 1902-POR 770-977.2-1162 x 5.4-13.4-18 I: 266-299.6-336 x 28-28.0-8 II: 100.0 x 20.0 I: 518-673.4-1008 x 3.6-3.6-3.6 II: 966-1085.0-1218 x 7.2-7.2-7.2 I: 25.2-40.6-54 x 1.8-3.3 4 II: 32.4-41.4-54 x 6.5-6.6-7.2 7.2-10.0-14.4 Paratype UFBA 1894-POR 910-997.6-1232 x 10.8-13.9-18 I: 238-281.4-336 x 28-28.0-28 II: not found I: 308-653.8-1050 x 3.6-3.6-3.6 II: 1232 1302.0-1372 x 7.2-7.2-7.2 I: 18-37.4-57.6 x 1.8-3.3-3.6 II: 28.8-32.4-36 x 5.8-5.9-6.1 7.2-11.3-18 Paratype UFBA 1896-POR 742-910.5-1078 x 7.2-11.6-21 I: 252-310.3-378 x 28-28.0-28 II: not found I: 462 -724.7-1008 x 3.6-3.6-3.6 II: not found I: 25.2-42.8-57.6 x 3.2-3.6-5.4 II: not found 7.2-9.4-10.8

developping stages of the acanthoxeas can be seen, as well as abundant canals of the aquiferous system. Acanthoxeas II were observed in very small numbers, only by electron microscopy. Spicules (measurements in m): Megascleres: Protriaenes I, sometimes prodiaenes, rhabdome thicker immediately under the cladome, tapering gradually until the distal end, length 308 - 684.0 - 1050 and thickness 3.6 - 3.6 - 3.6; clads forming approximately a 90 angle with the rhabdome, length of clads 18 - 40.3 - 57.6 and thickness 1.8 - 3.4 - 5.4 (n = 90). Protriaenes II, rare, rhabdome longer and stouter than observed in the other category, length 966 - 1157.3 1372 and thickness 28.8 - 38.4 - 54 (n = 6); clads shorter and stouter than observed in the other category, length 28.8 - 38.4 - 54 and thickness 5.4 - 6.6 - 7.2. Oxeas, abundant, often straight with sharp ends, length 742 - 995.1 - 1232 and thickness 5.4 - 13.0 - 21.6 (n = 90). Acanthoxeas I, large, slightly curved at central region with sharp ends, occasionally forked in one of these, length 238 - 297.1 - 378 and thickness 28 - 28.0 - 28 (n = 90); spines sharp, frequently over 5 m high, robust and pointed to the center of the spicule. Growth forms are smooth and/or bear thin, straight thorns with an expanded central region. Acanthoxeas II, very rare, smaller, with sharp terminations, length 100 and thickness 20 (n = 2); spines very robust, straight, apparently flattened conical and densely packed, frequently nearly 10 m high. Microscleres: Sigmaspires, length 7.2 - 10.2 - 18 (n = 90). Ecology and distribution: Acanthotetilla walteri sp. nov. was found in seven out of twenty-four sampling stations. Five of these were on sand/gravel under the influence of treated organic effluents (Fig. 1B - I), and the remaining two on gravel under the influence of acidic chemical effluents (Fig. 1B II). Most of specimens, eighteen, were collected right next to the discharge of the outfall with acidic effluents. This is in marked contrast to the majority of the sponges communities collected in Bahias northern littoral in this project, which seemed to be more abundant near the organic effluents (PesoAguiar unpublished). The species is so far known only from its type locality in the northern sector of Bahia states littoral, between 22 and 35 m depth. Etymology: The species is named after Walter Andrade, who has been responsible for the collections in the above mentioned integrated biomonitoring programme since the beginning of the 1990s.

Remarks: Acanthotetilla walteri sp. nov. differs from all other species of Acanthotetilla by the frequent presence of two categories of protriaenes associated to a lack of anatriaenes (Table 2). Its larger protriaenes are notably stouter than any previously reported ones, with rhabdomes reaching as much as 50 m in diameter, as opposed to no more than 15 m in other Acanthotetilla spp. The second category of acanthoxeas was observed to be very rare in the holotype, and was not found at all in any of the paratypes. This spicules exceeding rarity makes it a likely aberrant form, rather than a regular member of the new species spicule set. The Bahian species differs further from A. gorgonosclera by its growth form, the new species being encrusting and not possessing the irregular tubes described by van Soest (1977); and by the shape of the first category of acanthoxeas, which are curved in the central region and have a larger number of spines than reported for A. gorgonosclera (almost 18 whorls). Acanthotetilla walteri sp. nov. differs additionally from A. seychellensis (Thomas, 1973) by the lack of a second category of oxeas, and by the morphology of the acanthoxeas, which are thinner and with larger spaces between groups of whorls of spines in the latter. The new species differs from A. enigmatica also by the much smaller length of the oxeas (995 m in A. walteri sp. nov. x 2960 m in A. enigmatica); and by the morphology of the acanthoxeas, which in A. enigmatica exhibit very spaced spines as opposed to a rather dense arrangement in A. walteri sp. nov. The distinction from A. hemisphaerica lies in the length of the oxeas as well (995 m in A. walteri sp. nov. x 3812 m in A. hemisphaerica), and in the morphology of the acanthoxeas, stouter in A. hemisphaerica. Supplementary distinctive features between A. walteri sp. nov. and A. celebensis are the latters possession of much larger and stouter oxeas and second category of acanthoxeas. Finally, A. walteri sp. nov. differs from the other new species described here, A. rocasensis sp. nov., by the lack of a second category of oxeas and by its considerably larger acanthoxeas I. The new species is thus considered well distinguished from the remaining known species of Acanthotetilla.

Discussion
Acanthotetilla had a western Indian Ocean/Caribbean distribution classified as discontinuous by van Soest (1994). The recent description of a species from Indonesia (de Voogd and van Soest 2007) and the present finding of

515

two southwestern Atlantic species shuffles the scenario by expanding the genus distribution easterly at the same time that the number of western hemisphere species is increased. It is tempting to consider the Indian Ocean and Atlantic subgroups as likely monophyletic, which could be mirrored in the classification by establishment of two subgenera. Biogeographic criteria need not be the sole ones, as all the Atlantic species agglutinate debris and are possibly endolithic, while none of those in the Indian Ocean were reported as likely so (e.g. van Soest 1977, de Voogd and van Soest 2007). On the other hand, the presence of anatriaenes cuts right through these putative clades, by uniting the Atlantic A. gorgonosclera and the Indian Ocean A. hemisphaerica, A. enigmatica and A. celebensis; or, through the lack of triaenes, both new Atlantic species and the western Indian Ocean A. seychellensis. An explanation for the known distribution of Acanthotetilla is still difficult. The observation that the genus appears to be absent from most of the tropical/sub-tropical South Atlantic favors a scenario with relatively recent invasion of northeastern Brazilian waters from the north (hypothesis 1), rather than from the Indian Ocean, around Cape of Good Hope (hypothesis 2). In hypothesis 1, either the Caribbean or the Indian Ocean species should be basal in the Acanthotetilla clade, with southwestern Atlantic species being derived in the tree, as a consequence of their later divergence. This relatively recent north-south vicariance would account for the apparent lack of Acanthotetilla in most of the South Atlantic. If hypothesis 2 is the correct one instead, invasion of the Atlantic could have occurred much earlier, before the establishment of a connection between the North Atlantic/ western Tethys and the South Atlantic Ocean. In this scenario, species inhabiting the present days sub-tropical/temperate South Atlantic, formerly a typically tropical sea, would have become extinct after the onset of the circum-Antarctic current, which is responsible for the low temperatures of the Falklands/Malvinas and Benguela currents. This lowered temperature could explain the large distribution gap in this area. The new species reported here would also be more derived than their Indian Ocean congeners, but Acanthotetilla gorgonosclera, from the Caribbean, would be closer to the northeastern Brazilian species, and possibly more derived than all.

de Desenvolvimento Cientfico e Tecnolgico (CNPq), Fundao de Amparo Pesquisa do Estado do Rio de Janeiro (FAPERJ), Fundao de Amparo Pesquisa do Estado da Bahia (FAPESB) and Universidade Federal do Rio de Janeiro are deeply thanked for the provision of (travel) grants and/or fellowships.

References
Burton M (1959) Sponges. John Murray Exped 1933 - 1934, Sci Rep 10(5): 151-218 de Voogd NJ, van Soest RWM (2007) Acanthotetilla celebensis sp.nov., a new species from North Sulawesi, Indonesia (Porifera: Demospongiae: Spirophorida: Tetillidae). Zootaxa 1397: 25-28 Hajdu E (1994) A phylogenetic interpretation of hamacanthids (Demospongiae, Porifera), with the redescription of Hamacantha popana. J Zool 232(1): 61-77 Hajdu E, Muricy G, Berlinck RGS, Freitas, JC (1996) Marine poriferan diversity in Brazil. Through knowledge to management. In: Bicudo CEM, Menezes NA (Org). Biodiversity in Brazil. A first approach. CNPq, Brasil. pp 157-172 Kikuchi RKP, Leo ZMAN (1997) Rocas (southeastern Equatorial Atlantic): an atoll built primarily by coralline algae. Proc 8th Int Coral Reef Symp, Balboa 1: 731-736 Lana PC, Camargo MG, Brogim RA, Isaac VJ (1996) O bentos da costa brasileira: avaliao crtica e levantamento bibliogrfico (1858 - 1996). FEMAR, Rio de Janeiro Lvi C (1964) Spongiaires du Canal de Mozambique. Bull Mus Nat Hist nat Paris (2) 36(3): 384-395 Moraes F, Vilanova EP, Muricy G (2003) Distribuio das esponjas (Porifera) na Reserva Biolgica do Atol das Rocas, nordeste do Brasil. Arq Museu Nac 61: 13-22 Moraes F, Ventura M, Klautau M, Hajdu E, Muricy G (2006) Biodiversidade de esponjas das ilhas ocenicas brasileiras. In: Alves RJV, Castro JW de A (eds). Ilhas ocenicas brasileiras, da pesquisa ao manejo. Ministrio do Meio Ambiente, Braslia. pp. 147-177 Muricy G, Santos CP, Batista D, Lopes DA, Pagnoncelli D, Monteiro LC, Oliveira MV, Moreira MCF, Carvalho MS, Melo M, Klautau M, Dominguez PR, Costa RN, Silvano RG, Schwientek S, Ribeiro SM, Pinheiro US, Hajdu E (2006) Filo Porifera. In: Lavrado HP, Ignacio BL (eds). Biodiversidade bentnica da regio central da zona econmica exclusiva brasileira. Museu Nacional, Rio de Janeiro. pp. 109-145 van Soest RWM (1977) Revision of the megacanthoxea-bearing Tetillids (Porifera, Spirophorida with a description of a new species. Stud Fauna Curacao Caribb Isl 53(172): 1-14 van Soest RWM (1994) Demosponge distribution pattems. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 213-223 van Soest RWM, Rtzler K (2002) Family Tetillidae Sollas, 1886. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges. Kluwer Academic/Plenum Publishers, New York. pp. 85-98 Thomas PA (1973) Marine Demospongiae of Mah Island in the Seychelles Bank (Indian Ocean). Ann Kon Mus Midd - Afrika Tervuren (Zool Wetensch) 203: 1-96

Acknowledgements
Maurizlia Brito, Head of Reserva Biolgica do Atol das Rocas (ReBIO Atol das Rocas) and Instituto Brasileiro do Meio Ambiente e dos Recursos Naturais Renovveis (IBAMA), are greatly thanked for the provision of access to and facilities at das Rocas Atoll, as well as for granting the collecting permits for EH and MVO 2002 and 2003 field trips. Fernando Moraes and Ulisses dos Santos Pinheiro helped with field work at das Rocas Atoll. Mrcia Attias and Nomia Rodrigues (Instituto de Biofsica Carlos Chagas Filho/ UFRJ) are thanked for the provision of access to SEM (JEOL 5310) and technical support in operating the equipment. Conselho Nacional

Porifera research: Biodiversity, innovation and sustainaBility - 2007

517

A new species of Cladorhiza (Porifera: Cladorhizidae) from S. California (USA)

Henry M. Reiswig(1,2*), Welton L. Lee(3)
Department of Biology, University of Victoria, P.O. Box 3020 Stn CSC, Victoria, British Columbia, V8W 3N5 Canada. hmreiswig@shaw.ca (2) Natural History Section, Royal British Columbia Museum, P.O. Box 9815, Stn. Prov. Govt., Victoria, British Columbia, V8W 9W2 Canada (3) 6600 Mokelumne Ave., Oakland, CA, 94605 USA. fiddlesponge@att.net
(1)

Abstract: A new, very large species of Cladorhiza, collected from 1,442 m depth on the San Juan Seamount, S. California, by the ROV Tiburon, is strikingly bilateral in symmetry and feather-like in form. The specimen, 382 mm in total length, consists of a narrow stalk attached to hard substrate by a small disc, and an elongate spatulate body. The main body, triangular in section, bears a continuous fringe of about 400, 21-mm-long marginal filaments and a series of 13 fleshy lobes projecting from the midline of the frontal surface. Major biologic processes are regionally separated. Male reproduction (as spermatic cysts) is restricted to the tissues of the frontal surface of the main body, including the frontal lobes. Female reproduction, as oocyte production, embryo development and larval maturation, occurs exclusively in two abfrontal surface tissue bands in the cushions between the keel and the more fleshy main body. Prey, exclusively small crustaceans, are captured and digested only on/in marginal filaments. The elegant bilateral symmetry attained by this species attests to the continuing experimentation with development patterns within Porifera. Keywords: California, carnivorous, Cladorhiza, new species, Porifera

Introduction
Cladorhizid sponges are a widely recognized group of about 100 species of highly specialized small, thin, branching, often symmetrical, deep-water forms that all probably capture small crustaceans as prey. Hajdu and Vacelet (2002) assigned authority for the family Cladorhizidae to Dendy (1922) because he erected the group name Cladorhizeae, assigning only his new species, Amphilectus unguiculatus, to it; Dendy gave no diagnosis of the group. De Laubenfels (1936) provided the first summary of the distinctive morphological and habitat characters shared by members of the three genera, Cladorhiza Sars, 1872, Asbestopluma Topsent, 1901, and Chondrocladia Thomson, 1873, that consitute the modern scope of family Cladorhizidae. Since the unique cladorhizid feeding habit remained only vaguely suspected (Sars 1872), interest in the group remained fairly minor until the spectacular revelation by Vacelet and Boury-Esnault (1995) that a member of the family (later named Asbestopluma hypogea, by Vacelet and Boury-Esnault 1996), inhabiting a shallow-water (22 m) Mediterranean cave near Marseille, France, entirely lacked choanocyte chambers and fed instead by passively capturing and digesting small crustaceans. This publication stimulated general interest in this highly specialized family of carnivorous sponges and incited greater attempts to collect these largely deep-water species. New information has been especially sought to help reveal the present distribution of the feeding habit and evidence for

the environmental factors which facilitated evolution of this unexpected feeding method within Porifera. In response to our wide-ranging request for new specimens to aid our compilation of the marine sponges of California (Lee et al. 2007), L. Lundsten (Monterey Bay Research Institute = MBARI) presented us in May, 2005, with a recent submersible-collected, feather-form sponge for identification. That specimen has turned out to be a new species of Cladorhiza which is unusual in both its large size and exquisite bilateral symmetry. Here we describe the taxonomic characters distinguishing this new species, its regional functional specialization, and evidence of its carnivory, the most compelling available so far for a deepwater member of the Cladorhizidae.

Materials and methods

One specimen of the new species was collected from the San Juan Seamount, S. California (Fig. 1) by Dave Clague using MBARIs ROV Tiburon supported by RV Western Flyer and fixed in 70% ethanol. The specimen was broken and folded during collection. Additional occurrences of the species were extracted from dive records. For histological analysis by light microscopy, tissue blocks from various body regions were desilicified in 4% hydrofluoric acid in water for 8 h, dehydrated, embedded in paraffin, and sectioned at 12 m. Tissues were stained with eriochrome cyanin either en block before dehydration or after mounting

518

sections on slides. For surface tissue examination by scanning electron microscopy (SEM), tissue blocks were dehydrated, critical-point dryed (CPD) via carbon dioxide, attached to pegs with epoxy, sputter coated with gold-palladium, and viewed in an Hitachi S-3500 SEM at the University of Victoria. Marginal filaments were desilicified, stained with rose bengal, embedded in paraffin, sectioned longitudinally to expose prey cysts, deparaffinized, and handled as surface tissues above. Spicule preparations were made by digesting tissue fragments from various body regions in hot nitric acid. Cleaned spicules were isolated by either accumulating the spicules on filters, 0.22 m pore size nitrocellulose filters for light microscopy and ion-etched 0.2 m pore size polycarbonate membrane filters for SEM, or rinsing the spicules by settling/decantation with water, with final transfer to cover glass on SEM pegs by pipette or forceps. Spicule preparations were viewed in either the SEM noted above or a Leo 1950 VP SEM with a standard secondary electron detector (HV=20 kV and beam current = 80A) at the California Academy of Sciences (CAS). Measurements of slide preparations were made either directly with an ocular micrometer or indirectly by a computer-digitizer coupled to light microscopes by drawing tube (camera lucida). Data are presented as mean st. dev. (range, number of measurements).

Systematics
Phylum Porifera Grant, 1836 Class Demospongiae Sollas, 1885 Order Poecilosclerida Topsent, 1928 Family Cladorhizidae Dendy, 1922

Fig. 1: Location of Cladorhiza pteron sp. nov., holotype collection and additional sighting of the species, San Juan Seamount, 160 km SSW of Pt. Conception, California. Fig. 2-4: In situ photos of additional Cladorhiza pteron sp. nov., on San Juan Seamount, reproduced with permission from Monterey Bay Aquarium Research Institute. Fig. 5: Cladorhiza pteron sp. nov. holotype in frontal (left) and abfrontal (right) views. The body is twisted at a midbody discontinuity caused by folding during collection. Arrows indicate the levels of insertion of frontal lobes. Fig. 6: Frontal lobe from midbody of Cladorhiza pteron sp. nov. before sectioning. The entire lobe was desilicified and sectioned - see Figs 17A and B. Fig. 7A-B: Long styles and enlargement of the two ends from Cladorhiza pteron sp. nov. (SEM). Microscleres have been retained in Fig. 7A for size comparison. Fig. 8A-B: Short style and enlargement of the rounded head from Cladorhiza pteron sp. nov. (SEM). Fig. 9A-C: Anchorate/unguiferate anisochelae and enlargement of tips from Cladorhiza pteron sp. nov. (SEM). Fig. 10: Surface of a marginal filament of Cladorhiza pteron sp. nov. with anchorate/unguiferate anisochelae anchored in surface tissues by their narrow ends with anchorate end exposed for prey capture (SEM). Fig. 11: Sigmancistra of Cladorhiza pteron sp. nov. from the frontal surface (SEM). The equality of ends is obscured by 90o twisting of the spicule.

Genus Cladorhiza Sars, 1872

Cladorhiza pteron sp. nov. Figs. 1-21. Type material: Holotype: California Academy of Sciences, Invertebrate Zoology 173204, coll. D. Clague, 2 May 2004, San Juan Seamount, S. California, 33.1323oN, 120.9043oW (within Eclusive Economic Zone of USA), 1446m, ROV Tiburon, dive T664 from RV Western Flyer (previously MBARI specimen T664-A16). Additional material: The species was reported from 17 locations over a total range of 96 km of San Juan and Rodriguez Seamounts on Tiburon dives 629 and 664 (Table 1) and photographed in situ at most sites (Figs. 2-4). Diagnosis: Vertically elongate, bilaterally symmetrical body borne on a thin stalk attached to hard substrate by basal plate; spatulate body, attached abfrontally to edge of extended stalk, bears a complete fringe of marginal filaments with upper apical growth point; without branching. Male reproductive lobes are spaced along midline of the frontal body surface. Megascleres as two size classes of styles; microscleres as anchorate/ unguiferate anisochelae and sigmancistras. Description: The holotype (Fig. 5), the only specimen collected, is 382 mm in overall length, consisting of an elongate spatulate and feather-like body, 235 mm long attached onto

a dense, rigid, 8.1 mm thick stalk 165 mm in length, ending in a small basal disc 22 mm in diameter. The stalk, attached to the abfrontal surface of the body, extends the length of the body providing its primary support. The body, triangular in section, with a slightly concave frontal surface, is 16.3 mm wide and 216 mm long, bearing a complete fringe of ca. 396 marginal filaments 0.69 0.12 (0.5-1.0, n = 26) mm in thickness and 21.4 5.8 (7.1-32.8, n = 31) mm in length. With filaments included, the main body is effectively 68.5 mm in greatest width. Ten additional filaments occurring in a single transverse series across the frontal body surface appear to be an irregularity in their distribution. Clumping of filaments in groups in the fixed holotype appears to be an artefact of tissue fusion occurring while it was captive in the ROV holding tank after capture; it is not seen in specimens photographed in situ. The number of marginal filaments in a smaller uncollected but photographed specimen with only 6 frontal lobes was ca. 192 (Fig. 4). Eleven cylindrical or flattened-palmate frontal lobes (Fig. 6), 12.0 2.8 (8.6-19.0, n = 11) mm long by 3-5 mm wide by ca. 1 mm thick, occur in midline of the frontal body surface, spaced 17.9 10 (4.1-42, n = 10) mm apart; two additional developing frontal lobes are small knobs near the upper growth end. Spicules, megascleres: Styles of two size groups. Long styles (Figs 7A-B); length 2569 660 (1042-3636, n = 108) m, width 51.1 10.2 (24-73, n = 108) m. These are the most abundant styles. They are mostly fusiform with usually the stylote end somewhat restricted as noted above, and the point, gently rounded, not sharp. In some cases smaller constrictions may occur on the opposite (oxeote) tip. In such cases both ends are usually gently rounded. Rarely the spicules ap-

519

pear as typically stylote. Length/width ratios are highly variable. These spicules dominate all central areas of the sponge, from the core of the stalk and the axis to the core of all of the lateral filaments. Short styles (Fig. 8A-B); length,788 235 (370-1333, n = 102) m, width 37.7 8.5 (24-55, n = 102) m. Short styles range from stylote with rounded head and

gently rounded to pointed tip, to fusiform with either one end or both constricted so as to form a somewhat narrow, almost mucronate-like handle. In the latter case it is often difficult to say if these are fusiform styles or anisoxeas. These spicules are rare in the general body structure but dominate in the base or holdfast area.

520

Styles intermediate in size also occurs. These, however, are so rare (only two were found in spicule preparations) that we hesitate to designate them as a distinct size class. They measured 1539 x 18 m and 2460 x 15 m and are distinguished by their extreme thinness as compared to all other megascleres. These were only found between the larger styles, which support the central core of the stalk. The central core is made up of extremely tightly packed styles of the largest size class. The intermediate size styles were found in areas where the core changes thickness. While we suspect that many more of these intermediate sized spicules may occur in the core, especially in curved areas, this was impossible to verify since the core spicules are so tightly bound together

that efforts to separate the individual spicules only damages and breaks spicules. We suspect that these thinner spicules fill the spaces that the larger spicules naturally form when the core becomes curved and the alignment of the larger styles is disrupted by the curvature. Microscleres: two types, anchorate/unguiferate anisochelae and sigmancistras. Anchorate/unguiferate anisochelae (Figs 9A-C), length 28.4 1.5 (24-33, n = 100) m, have the shaft gently curved. The anchorate head has three prominent lanceolate alae with their sides parallel on the upper two thirds and then tapered to a sharp point. The fimbriae are prominent, with the somewhat extended upper half gently curved. The unguiferous foot has three short, claw-like and pointed

521

Fig. 12: Diagram of positional tissue specialization of Cladorhiza pteron sp. nov. in cross-section (above), frontal and abfrontal renderings, showing regions of male gamete production, female gamete production and feeding. Fig. 13: Part of the right female tissue band lying between the marginal filaments (above) and keel (below) of Cladorhiza pteron sp. nov.; the late embryos and larvae are visible beneath the thin transparent surface tissues. Fig. 14: Oocyte in the process of feeding from surrounding trophocytes in the female tissue band of Cladorhiza pteron sp. nov. (LM). Fig. 15: Moderately developed embryo lying within a pinacocytelined follicle in the female tissue band of Cladorhiza pteron sp. nov. (LM). Fig. 16: Mature larvae, posterior end up, in a more expanded developmental follicle in the female tissue band of Cladorhiza pteron sp. nov. (LM). Fig. 17: Stained cross-section of the frontal lobe of Fig. 6 in entirety (A, LM montage) and in magnified view of near surface tissues (B) of Cladorhiza pteron sp. nov. Sperm follicles occur in two to three layers with dark-staining basal structure more evident as maturing follicles near the lobe surface. Fig. 18: Mature sperm follicle from the frontal body surface of Cladorhiza pteron sp. nov. (LM). Fig. 19: An intact, recently captured and encysted prey copepod within a marginal filament of Cladorhiza pteron sp. nov. (SEM). Fig. 20: An encysted prey thoroughly invaded by parenchymal cells of the sponge within a marginal filament of Cladorhiza pteron sp. nov. (SEM). Fig. 21: Completely cleaned cuticular remains of a digested prey copepod within a marginal filament cyst of Cladorhiza pteron sp. nov. (SEM).

Table 1: Location of Cladorhiza pteron on Rodriguez (dive 629) and San Juan (dive 664) Seamounts, S. California from ROV Tiburon dive records. Dive 629 629 629 629 629 629 629 629 629 629 629 629 629 664 664 664 664 Latitude 33.95391 33.95505 33.955154 33.955227 33.96699 33.967518 33.968365 33.968723 33.969788 33.970573 33.970577 33.97089 33.97093 33.13229 33.132286 33.1323 33.132286 Longitude -121.145134 -121.145294 -121.14513 -121.14521 -121.13368 -121.13392 -121.13361 -121.133705 -121.1335 -121.133446 -121.13336 -121.1332 -121.1328 -120.90432 -120.904305 -120.90428 -120.90431 Depth(m) 1762.3 1749.9 1711.9 1705.4 1785.8 1760.6 1750.6 1741.4 1718.8 1707.1 1700.9 1689.5 1667.1 1443.3 1443.5 1443.5 1443.5

alae. These microscleres are most abundant on the marginal filaments (Fig. 10) where they occur in massive numbers between the ectosomal membrane and the central core of the filaments. They are likewise abundant throughout the sponge on all surfaces with the exception of the base and the very lowest portion of the peduncle. Sigmancistras (Fig. 11), length 36.7 2.0 (32-42, n = 100) m, are contort with both ends ending in an acerate tip (sharp spine) the axis of which swells abruptly just behind the point and bends. The spicule is widest at this point, appearing to gradually decrease in width toward the spicule center due to rotation of the flattened axis. The two ends of this spicule type are mirror images of one another. These spicules are found predominantly on peels of surface tissues on the frontal face, the surface bearing the blunt lobes. They are abundant here and may on occasion be found elsewhere, but not reliably. Skeletal organization: the sponge is attached to the substrate via an expanded base which is formed of massive numbers of short styles closely embedded in a spongin matrix to form an exceedingly strong basal mass. These generally merge with the long styles to form an equally thick and strong core to the peduncle and body axis. In all cases, spicules of all widths fill spaces where the mass increases or decreases in width. In the base and lower axis, the parenchymal tissue is thin, thus the surface is close to the mass of styles. In the lower half of the sponge, some anisochelae may occur along with a very few sigmancistras. In the upper body, near the base of filaments, the number of microscleres borne by the surface and the thicker parenchymal tissues increases. The filaments themselves are each cored by a thin cylindrical mass of long styles which arise as lateral branches from the axis core of megascleres. The filaments have a distinct, thickened surface membrane (dermis) within which can be seen multitudes of anisochelae oriented with their larger anchorate end exposed. Thickness of the parenchymal tissue mass, between the anisochela-bearing dermis and style-bearing core varies from about 169-181 m at its widest point to almost nothing near the filament tips. Functional body regionation: The major physiological processes of male reproduction, female reproduction and feeding are restricted to very specific, non-overlapping body regions in this bilaterally symmetrical species (Fig. 12). Male gamete production takes place only on the frontal surface, concentrated in the frontal lobes. Female gamete production and embryonic development takes place in two strips of abfrontal tissues. Food capture and processing takes place only on the marginal filaments. Egg and larval production: small, young, unfed oocytes before vitellogenesis, large vitellogenous, but uncleaved oocytes, early embryos, and late larvae, occur along two abfrontal tissue bands lying in the cushion between the extended stalk and the marginal filaments (Fig. 13). The distal five centimeters lack female reproductive elements. Young oocytes are ovoid, average 42 x 72 m diameters, with large 16 x 20 m nucleus and a single 5.9 m diameter nucleolus (Fig. 14). Moderately developed embryos, 567 x 892 m in diameters, are enclosed in a cell follicle (Fig. 15). Late larvae with approximately 10,000 cells are parenchymellae (Fig. 16), typical of those known in Poecilosclerida generally. They have dimensions

522

722 x 778 m, and lie loosely in a thin-walled, but expanded follicle 777 x 1407 m in total dimensions. The anterior and lateral surfaces of the larvae are multistratified, flagellated, and 67 m thick while the posterior surface is thinner, 33 m thick, with unflagellated, and columnar epidermis. Spicules were not present in the brooded larvae examined. Extrapolation of measured density of embryos and larvae indicate the entire holotype has a total of 482 of these stages. Sperm Production: sperm follicles occur in a nearly continuous layer just below the dermis of the entire frontal body surface excluding the 20 mm apex growth region, but are two or three layers deep on all surfaces of the frontal lobes (Figs 17A, B). The follicles are ovoid in transverse section and larger on the frontal lobes, length 201 47 (134-318, n = 28) m, thickness 141 20 (93-219, n = 28) m (vs. means of 110 x 83 m under the frontal surface. All stages of sperm maturation are present; development is synchronous within follicles but asynchronous between follicles. Many, if not all follicles, have a dark-staining, structureless lens-shaped structure on the proximal side, larger and more dense in follicles in later maturation stages and in contact with the dermis (Figs 17B, 18). Extrapolation of follicle density measured at the two locations resulted in an estimate of over 36,000 sperm follicles on the entire holotype, 88% of them produced on the frontal lobes. Prey capture and digestion: Choanocyte chambers and an aquiferous canal system are absent in the holotype. Small copepod crustaceans are attached to the surfaces of only the marginal filaments, about one copepod every five or six filaments. Some of these are partly enclosed in a pocket of the surrounding surface tissues and small swellings about the size of the copepods are clearly evident on the filaments. Serially sectioned filaments contain copepods and their remains in fully enclosed cysts within the fairly spacious but dense parenchymal layer, between the dermis and the axial spicule bundle. All stages of prey digestion are encountered, from intact and apparently freshly enclosed stages (Fig. 19), stages infiltrated by massive numbers of sponge parenchymal cells (Fig. 20), and completely cleaned and empty prey cuticles ready for elimination (Fig. 21). This set of conditions and arrangement of microscleres on filament surfaces is consistent with the conclusion that the copepods are prey, snared on projecting microscleres, overgrown by parenchymal tissues and digested in prey cysts in the parenchyme, as has been well documented only for Asbestopluma hypogea by Vacelet and Boury-Esnault (1995) and Vacelet and Duport (2004). The number and size of copepods in three completely serial-sectioned, mid-body filaments amounted to a mean of 19 copepods per filament with a size range of 0.39 0.34 (0.17-1.47) mm in total length. This is extrapolated to provide an estimate of 7,000 copepod prey being processed in cysts in the entire holotype at time of collection. Etymology: The name, pteron, is derived from the Greek word, for feather, reflective of the body form of the new species.

Discussion
In recently describing a new cladorhizid, Cladorhiza corona, from the Aleutian Islands, Lehnert et al. (2005) produced a valuable table listing the known Cladorhiza species and their characters. The 28 presently known species can be effectively compared to the new form, C. pteron sp. nov., on the basis of three principal parameters: body shape, symmetry, and spicule complement Body shape can be quite variable within Cladorhiza, however it is possible to recognize a few general categories and compare these to C. pteron sp. nov. with its stalked, unbranched, pinnate and bilateral body form. The first category includes species with a branched form, these having either small branches coalescing with the stalk and/or short branchlets to more robust, highly branched forms, almost tree-like in appearance. Filaments, if present, can be either very small or large and obvious (individual branches may appear pinnate). This group includes C. abyssicola Sars, 1872, C. corticocancellata Carter, 1876, C. gelida Lundbeck, 1905, C. methanophila Vacelet and Boury-Esnault, 2002, C. oxeata Lundbeck, 1905, and C. thomsoni Topsent, 1909. They clearly differ from the unbranched C. pteron sp. nov. A second shape category includes forms without branching of the main stem: Cladorhiza arctica Burton, 1946 (clavate body), C. grimaldi Topsent, 1909, C. fristedti (Lambe, 1900), C. rectangularis Ridley and Dendy, 1886, C. schistochela Lvi, 1993, C. septemdentalis Koltun, 1970 (cylindrical body), C. ephyrula Lvi, 1964 (discoid or vase-shape body), C. segonzaci Vacelet, 2006 (rarely partly pinnate). These clearly differ from the flattened pinnate bilateral form of C. pteron sp. nov. A third common body form within the genus is stalked with some kind of terminal swelling which may have numerous extending filaments. These include Cladorhiza bathycrinoides Koltun, 1955, C. corona Lehnert et al., 2005, C. inversa Ridley and Dendy, 1886, C. longipinna Ridley and Dendy, 1886, C. minuta (Lambe, 1900), C. mirabile (Ridley and Dendy, 1886), C. moruliformis Ridley and Dendy, 1886, C. nematomorpha Lvi, 1964, and C. similis Ridley and Dendy, 1886. No terminal swelling occurs in C. pteron sp. nov. Although three species are listed in Lehnert et al. (2005) as pinnate in form, review of original figures and descriptions show these to have cylindric body forms. They are not remotely pinnate in the sense shown by C. pteron sp. nov. A few species fall outside of these general categories: C. linearis Ridley and Dendy, 1886, has a slender axis with lateral tufts of spicules, C. mani Koltun, 1964, is a short stalked form, C. depressa Kieschnick, 1896, is laterally flattened, C. flosabyssi Topsent, 1909 has a long thin stalk and flower-like tentacles, C. microchela Lvi,1964, is a small thin filament, C. tenuisigma Lundbeck, 1905, has a main trunk with rhizoids and long tentacles, and C. tridentata Ridley and Dendy, 1886, is a small hemispheric dome. None of these come close to the form seen in C. pteron sp. nov. While C. pteron sp. nov. shares its general stalked, pinnate form with a few other species, it diverges from these dramatically by its elongate, trowel-shaped body, with the continuation of the stalk forming a flattened keel along

523

the abfrontal surface. The frontal surface is fringed with numerous filaments and a series of lobes projecting from the midline. With this body form, which is distinctly bilateral in symmetry, C. pteron sp. nov. is to our knowledge the only known species in this genus with this set of characteristics. Virtually all other cladorhizids are either asymmetric or show radial symmetry. Cladorhiza pteron sp. nov. has a spicule complement which includes large styles, short styles-anisoxeas and sigmancistras. Within the genus Cladorhiza, the spicule complement is generally similar. Megascleres always include one or more of the following: styles-subtylostyles, tylostyles, oxeas or acanthoxeas. Acanthoxeas are rare, occurring in only one species, C. arctica, where they are accompanied by styles-strongyles. Several cladorhizids differ by having tylostyles. These include C. flosabyssi, C. fristedti, C. inversa, C. longipinna, C. mani, C. minuta and C. tridentata. In addition a host of other cladorhizids, unlike C. pteron sp. nov., have true sigmas: C. abyssicola, C. bathycrinoides, C. corticocancellata, C. depressa, C. ephyrula, C. flosabyssi, C. gelida, C. grimaldi, C. linearis, C. mani, C. moruliformis, and C. segonzaci. Sigmancistras, like those present in C. pteron sp. nov., are reported from only three other species, C. abyssicola, which is a branched cladorhizid, C. corona and C. segonzaci. C. septemdentalis is unique by virtue of its septemdentate anisochelae. Cladorhiza pteron sp. nov., which is clearly unique among all known cladorhizid species, shares several additional characteristics with the recently reported C. corona. Both are attached directly to hard substrate and have short, thick styles-anisoxeas only in the basal plate. These were suggested to be related to attachment to a solid substrate (Lehnert et al., 2005). Likewise, both species have equal-ended sigmancistras which are not common in cladorhizids. The two species, however, differ significantly in body form and symmetry. Cladorhiza corona has a stalked form with a basal plate and two sets of distal appendages, the basal one radiating in a full circle and the distal forming a quarter circle of triangularshaped structures oriented in a plane almost perpendicular to the basal appendages, the lower portion distinctly radial in symmetry, while the upper appendage is bilateral. One can envision this to be a variant or precursor to the strict bilateral symmetry of C. pteron sp. nov.

funding was provided by a Discovery Grant from the Natural Sciences and Engineering Research Council of Canada to HMR. Finally we extend our thanks to Dr. Bob Van Syoc who made possible the extensive use of the facilities at the California Academy of Sciences, Mr. Scott Serata who aided in the use of the Academies SEM and to former librarian Ms. Marion Taylor who was so helpful in obtaining much of the more obscure literature.

References
de Laubenfels MW (1936) A discussion of the sponge fauna of the Dry Tortugas in particular and the and the West Indies in general, with material for a revision of the families and orders of the Porifera. Carnagie Institution of Washington (Tortugas Laboratory, Paper No. 467) 30: 1-225 Dendy A (1922) Report on the Sigmatotetraxonida collected by H.M.S. Sealark in the Indian Ocean. Reports of the Percy Sladen Trust Expedition to the Indian Ocean in 1905, vol. 7. Trans Linn Soc London, Ser. 2, 18: 1-164 Hajdu E, Vacelet J (2002) Family Cladorhizidae Dendy, 1922. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges, vol. 1. Kluwer Academic/Plenum Publishers, New York. pp. 636-641 Lee WL, Elvin DW, Reiswig HM (2007) The sponges of California, a guide and key to the marine sponges of California. Monterey Bay Sanctuary Foundation, Monterey, California Lehnert H, Watling L, Stone R (2005) Cladorhiza corona sp. nov. (Porifera: Demospongiae: Cladorhizidae) from the Aleutian Islands (Alaska). J Mar Biol Assoc UK 85: 1359-1366 Sars GO (1872) Spongiae. In: Kongelige Norske Universitat (ed), On some remarkable forms of animal life from the great depths off the Norwegian coast. I. Partly from posthumous manuscripts of the late professor Dr. Michael Sars. Brgger & Christie, Christiania, Norway. pp. 62-82 Vacelet J (2006) New carnivorous sponges (Porifera, Poecilosclerida) collected from manned submersibles in the deep Pacific. Zool J Linn Soc Lond 148: 553-584 Vacelet J, Boury-Esnault N (1995) Carnivorous sponges. Nature 373(6512): 333-335 Vacelet J, Boury-Esnault N (1996) A new species of carnivorous sponge (Demospongiae: Cladorhizidae) from a Mediterranean cave. In: Willenz P (ed). Recent Advances in Sponge Biodiversity. Inventory and Documentation. Bulletin de lInstitut royal des Sciences naturelles de Belgique. Biologie, 66(Suppl.): 109-115 Vacelet J, Duport E (2004) Prey capture and digestion in the carnivorous sponge Asbestopluma hypogea (Porifera: Demospongiae). Zoomorphology 123: 179-190

Acknowledgements
We thank Lonny Lundston (MBARI) for providing access to the specimen, in situ digital photos and data from dive records. Partial

Porifera research: Biodiversity, innovation and sustainaBility - 2007

525

Sponges of genus Myxilla Schmidt, 1862, collected in Antarctic waters by Spanish Antarctic expeditions
Pilar Ros(1), Javier Cristobo(2*, 3)
Departamento de Bioloxa Animal. Facultade de Bioloxa. Universidade de Santiago de Compostela, Spain. pilar.rios.lopez@gmail.com (2) Ministerio de Educacin y Ciencia, Instituto Espaol de Oceanografa. Centro Oceanogrfico de Gijn, C/ Prncipe de Asturias 70 bis, 33212 Gijn, Asturias, Spain. cristobo@gi.ieo.es (3) Departamento de Zoologa y Antropologa Fsica. Universidad de Alcal. Alcal de Henares, Madrid, Spain
(1)

Abstract: Between 1994-2003, four different Spanish expeditions were undertaken at the Antarctic Peninsula and Bellingshausen Sea (Antarctica). Many sponge species were collected during these expeditions, six of which belong to the genus Myxilla Schmidt, 1862, redescribed and illustrated here, including reaffirming the validity of Myxilla (Burtonanchora) magna characterised by the presence of styles in the polygonal choanosmal skeleton, tylotes with spined ends in tangentially arranged in ectosome, anchorate chelae and two size classes of sigmas in a dense ectosomal layer. Keywords: Antarctica, Myxilla, Poecilosclerida, Porifera, systematics

Introduction
During the Spanish Antarctic expeditions Bentart 94, Bentart 95, Bentart 03 and Gebrap 96 large collections of Antarctic sponges were made (Ros et al. 2004, Ros and Cristobo 2006). Amongst there were 38 samples of the family Myxillidae belonging to six species of Myxilla. Van Soest (2002), in the revision of this family, recognizes four subgenera. Species belonging to three of them are described in this paper: M. (Burtonanchora) de Laubenfels, 1936, M. (Ectyomyxilla) Hentschel, 1914 and M. (Myxilla) Schmidt, 1862. From these and other Antarctic collections, species of the subgenus Myxilla seem to be rare in the Southern Ocean, with species of subgenera Burtonancora and Ectyomyxilla in abundance at both shallow and deep water environments (van Soest 2002). Two species are redescribed here for the first time since their original descriptions: M. (B.) pistillaris Topsent, 1916 described from Peter I Island and M. (E.) hentscheli Burton, 1929 from Trinity Island. Five of the six species are reported here with extended bathymetric distributions: M. (M.) elongata Topsent, 1916 (46-672 m), M. (B.) magna Topsent, 1916 (92-233 m), M. (B.) pistillaris (86-92 m), M. (B.) lissostyla Burton, 1938, (15-1400 m) and M. (E.) hentscheli (214-385 m). As part of a larger project on Antarctic sponges we redescribe these species here, including reinstating the validity of M. (B.) magna previously synonymised with another species, and redefining characteristics of this species. Characters determined to be diagnostic for Burtonancora are the presence of styles in the

polygonal choanosmal skeleton, tylotes with spined ends tangentially arranged in the ectosome, anchorate chelae and two size classes of sigmas in a dense ectosomal layer. Sponges were collected by either a rock dredge or Agassiz trawl from the Antartic peninsula. Spicules of these species are illustrated for the first time by Scanning Electron Microscopy (SEM).

Material and methods

Specimens were collected during the Spanish expeditions Bentart 94, Bentart 95, Bentart 03 and Gebrap 96 (Fig. 1; Table 1), using rock dredge, scuba diving and Agassiz trawl. Specimens were preserved in 70% ethanol. Spicules were prepared for SEM as follows. The organic matter was digested with nitric acid taken to boiling point, following the methods of Rtzler (1978) and Cristobo et al. (1993), and spicules were examined using a Leica S440 Scanning Electron Microscope. The data for spicule sizes are based on 25 measurements for each spicule category, comprising minimum, maximum and average lengths in micrometers (m); width of chelae was measured in lateral view. Skeletal architecture was studied by light microscopy. The classification system adopted in this work is that proposed by van Soest (2002). Stations mentioned in the text are placed in these Antarctic areas

526

Fig. 1: Maps of Bentart and Gebrap expeditions with the locations of the sampling stations where Myxilla species were found.

527

Table 1: List of Antarctic stations with sponges of genus Myxilla. Expedition Bentart 94 Bentart 94 Bentart 94 Bentart 95 Bentart 95 Bentart 95 Gebrap 96 Station 11 33 100 23 24 24 6 Species Myxilla (Burtonanchora) lissostyla Burton, 1938 Myxilla (Burtonanchora) lissostyla Burton, 1938 Myxilla (Burtonanchora) lissostyla Burton, 1938 Myxilla (Myxilla) elongata Topsent, 1916 Myxilla (Burtonanchora) magna Topsent, 1916 Myxilla (Burtonanchora) magna Topsent, 1916 Myxilla (Ectyomyxilla) hentscheli Burton, 1929 Myxilla (Myxilla) elongata Topsent, 1916 Long W Lat S 603321 624003 602600 624130 604202 623933 635708 605944 635832 605236 635828 605159 595219 595125 624848 625016 585901 590028 624212 624103 590956 590851 624127 624158 902115 685005 785716 680424 632517 650116 630045 645407 Samples 4 1 2 1 2 1 1 1 Depth (m) 30 15 24 141 233 214 647-672 Sampler Rock dredge Diving Rock dredge Agassiz trawl Agassiz trawl Rock dredge Rock dredge

Gebrap 96

Myxilla (Burtonanchora) asigmata Topsent, 1901

1192-1379

Rock dredge

Gebrap 96

Myxilla (Burtonanchora) lissostyla Burton, 1938

1182-1400

Rock dredge

Bentart 03 Bentart 03 Bentart 03 Bentart 03

8 19 20 21

Myxilla (Burtonanchora) pistillaris Topsent, 1916 Myxilla (Burtonanchora) lissostyla Burton, 1938 Myxilla (Burtonanchora) lissostyla Burton, 1938 Myxilla (Myxilla) elongata Topsent, 1916 Myxilla (Burtonanchora) magna Topsent, 1916

1 15 1 3 3

86 517 46 104

Agassiz trawl Agassiz trawl Agassiz trawl Agassiz trawl

Systematic description
Class Demospongiae Sollas, 1885 Order Poecilosclerida Topsent, 1928 Suborder Myxillina Hajdu, van Soest and Hooper, 1994 Family Myxillidae Topsent, 1928 Genus Myxilla Schmidt, 1862 Subgenus Myxilla Schmidt, 1862 Myxilla (Myxilla) elongata Topsent, 1916 Material examined: Five specimens from the following stations: Stations 23 (Bentart 95, Trinity Island, 141 m), 6 (Gebrap 96, Bransfield Strait, 647-672 m) and 20 (Bentart 03, Gerlache, 46m). Depth range: 46-672 m. Additional material examined: Holotype. Microscopic slides of ND.T. 671 and D.T. 672 (Myxilla elongata), Museum National dHistoire Naturelle (MNHN), Paris.

Description: The species varies in growth form being either massive-erect or tubular, with dimensions of 12 x 8 x 4 cm. The surface is microlobate and uneven. Oscules lying on the sides of the sponge are larger, up to 3 mm, than those apically located, of 1 mm in diameter. The consistency of the sponge is soft and easy to break. One of the fragments was mucilagenous when collected. External colour yellow in life; in preservative (alcohol) it is beige (Fig.2). Skeleton: The choanosome is reticulate, partly unispicular or paucispicular principal tracts of acanthostyles, interconnected by secondary tracts of one or two acanthostyles. Numerous sigmas are present within the choanosome. The ectosomal tylotornotes are arranged in a palisade. Two size classes of isochelae are present in the ectosomal layer (Fig.2B). Spicules (Figs. 2, 3, 4): Megascleres - Acanthostyles straight or slightly curved with a few spines concentrated near each end. Size: 405520 (469.55) x 12.5-22.5 (17.11) m. Tylotornotes mainly straight. One end rounded with some spines and the other

528

Fig. 2: Myxilla (Myxilla) elongata Topsent, 1916. A. Habitus. B. Skeleton. C. Spicules. a. Acanthostyle. b. Tylotornote. c. Anchorate chelae I. d. Anchorate chelae II. e. Sigma.

with a mucron, and short and strong spines. Size: 235-320 (280.30) x 6.25-12.5 (9.39) m. Microscleres - Anchorate chelae I with three free teeth with rounded ends. The shaft is thinner in the centre. Size: 47.5-70 (57.58) x 12.5-20 (16.32) m. Anchorate chelae II with three long oval teeth. Size: 20-32.5 (26.95) x 3.75-6.25 (5.06) m. Sigmas C shape or sometimes S shape. Size: 45-65 (54.10) m. Distribution (Fig. 18): Antarctic and Subantarctic: Peltier Channel (Topsent 1917); McMurdo Sound (Burton 1929); South Georgia, Shag Rock (Burton 1932); Weddell Sea (Barthel et al. 1990); Terra Nova Bay-Ross Sea (Pansini et

al. 1994); Trinity Island, Bransfield Strait, Gerlache Strait (present study). Remarks: In the original description (Topsent 1916), Myxilla elongata is characterised by possessing slightly curved and fusiform tylotornotes or ectosomal subtylotes. Short spines are spread along the shaft, and these subtylotes are swollen at one end and slender at the other (with a mucron). The length of these spicules varied between 250-300 m with a thickness of 10 m. The choanosomal acanthostyles are slightly curved with one end either rounded or subtylote and the other with sharply tapering points. The acanthostyles are also moderately spined (460-470 x 17-24 m). Abundant anchorate chelae (28-33 m) and sigmas (50-60 m) (Fig. 4) are also present (Topsent 1917). The recently collected

529

Fig. 3: Myxilla (Myxilla) elongata Topsent, 1916. A. Acanthostyle. B. Tylotornote. C-D. Ends of Acanthostyle. E-F. Two bases of Tylotornote. G. Anchorate chelae. H. Detail of extremity of anchorate chelae I. I-J. Sigmas. K-M. Anchorate chelae II.

material all share the same morphological and spicular characteristics with the holotype, except for one specimen collected during the Bentart 03 expedition also having microxeas (size 25-85 m), although we conclude that there are foreign. Subgenus Burtonanchora de Laubenfels, 1936 Myxilla (Burtonanchora) asigmata Topsent, 1901 Material examined: Station 8 (Gebrap 96, Bransfield Strait). Depth: 1192-1379 m. 1 specimen.

Additional material examined: Holotype. Slides MNHN Paris ND.T. 1608 (Myxilla spongiosa var. asigmata). Description: The sponge is spherical in shape, 39 mm high and 28 mm in diameter. The surface is very perforate and reddish due to protruding fibres that are scattered over the surface and also incorporates sand grains. The texture is finely hispid due to the protruding spicules. Three big hollow oscules, 3 to 5 mm in diameter, are present in the upper surface of the sponge, with drainage canals radiating away from the oscules and forming stellate grooves. The sponge is firm and not easilty breakable. Exterior colour in life unknown; in preservative it is beige (Fig.5).

530

Fig. 4: Type of Myxilla (Myxilla) elongata Topsent, 1916. A. Slides. B. Spicules. C. Tylotornotes. D. Sigmas and anchorate chelae. E. End of acanthostyle. F. Tylotornote and anchorate chelae. G. Anchorate chelae I and II. H. Diatom.

Table 2: Spicule micrometries of Myxilla (Burtonanchora) asigmata Topsent, 1901. Reference Topsent (1901) Topsent (1908) Topsent (1913) Hentschel (1914) Burton (1932) Koltun (1964) Pansini et al. (1994) Exp. Gebrap 96 Styles Megascleres (m) Tylotes 380 x 7-8 285-320 x 10 375-400 x 10 285-380 180 x 6 118-400 x 7-10 120-400 295-382.5 x 5-10 Microscleres(m) Anchorate chelae (n teeth) 60-70 (3) 50-60 (?) 63-75 (3-5) 40-70 (?) 30-63 (?) 30-75 (3) 30-75 (3) 37.5-65 x 11.25-20 (5)

715-775 x 20 495-650 x 20-27 870-900 x 26-28 495-816 300 x 9 300-900 x 9-45 300-900 x 30 650-755 x 30-40

531

Fig. 5: Myxilla (Burtonanchora) asigmata Topsent, 1901. A. Habitus. B. Skeleton. C. Spicules. a. Style. b. Tylote. c. Anchorate chelae.

Skeleton: The choanosomal skeleton is reticulate, with multispicular principal tracts (160-390 m) compared of styles, lined with spongin fibres and interconnected by small secondary tracts of 3 to 5 spicules. Principal tracts diverge towards the ectosome and protrude beyond the surface. Tornotes are very scarce, and where present are found in groups of 10 to 12 spicules without a clear disposition. Microscleres are present in both the choanosomal tracts and spongin fibres (Fig.5B). Spicules (Figs. 5, 6): Megascleres - Styles thick slightly curved with tapering sharp points. Thick axial canal. Size: 650-755 (721) x 30-40 (36.7)

m. Tylotes straight with slightly acanthose base and smooth shaft. Size: 295-382.5 (339.85) x 5-10 (7.75) m. Microscleres - Anchorate chelae with five spatuliferous teeth. Three to five free teeth and two lateral teeth are fused to the shaft. Two incipient lateral alae. The shaft is thinner in the middle. Size: 37.5-65 (44.45) x 11.25-20 (13.45) m. Distribution (Fig. 18): Antarctic and Subantarctic: Bellingshausen Sea (Topsent 1901); Booth-Wandel Island (Topsent 1908); Weddell Sea (Topsent 1913); Gauss Station (Hentschel 1914); South Georgia (Burton 1932); Wilkes Land and South Shetland Islands (Koltun 1964); MacRobertson Land and Enderby Land (Koltun 1976); Weddell Sea (Barthel

532

Fig. 6: Myxilla (Burtonanchora) asigmata Topsent, 1901. A. Style. B. Tylote. C-D. Base and end of style. E-F. Ends of tylote. G-I. Anchorate chelae.

et al. 1990); Ross Sea (Pansini et al. 1994); Lazarev Sea (Gutt and Koltun 1995); Bransfield Strait (present study). Remarks: Myxilla (Burtonanchora) asigmata was first described by Topsent (1901) and characterised as follows: Styles and tylotes present as megascleres. The styles are smooth, and the tylotes are microspined at the top of the heads. Microscleres are tridentate chelae (Topsent 1901). This author, however, described the species as a variety of Lissodendoryx spongiosa (Ridley and Dendy, 1886), but Burton (1932) considered Myxilla asigmata was a valid species. Although Topsent was the first to indicate the variability in the chelae teeth structure (up to 5 teeth found), he described the species as having only three teeth on the

chelae. The specimen described here from Bransfield Strait shows similar morphological characteristics to the species described from Bellingshausen and Weddell Sea by Topsent (1901, 1913). The variations in the size of spicules are shown in Table 2. Van Soest (2002) defined Myxilla as having anchorate chelae possessing only three teeth, whereas the variations observed in the number of teeth in the chelae of Myxilla asigmata contrast to this definition, and as such suggests the species may be more appropriately included in Stelodoryx (Myxillidae). This latter genus possesses styles and tornotes as megascleres, a reticulate choanosomal skeleton and an average number of five teeth on each claw of the chelae.

533

Fig. 7: Myxilla (Burtonanchora) magna Topsent, 1916. A. Habitus. B. Skeleton. C. Spicules. a. Style. b. Tylote. c. Achorate chelae I. d. Anchorate chelae II. e. Sigma I. f. Sigma II.

Nevertheless, this species is provisionally retained in Myxilla, and it appears to us that Myxillas definition should extend to those species possessing chelae with three as well as more teeth. Myxilla (Burtonanchora) magna Topsent, 1916 Material examined: Stations 23 (141 m), 24 (233 m) (Bentart 95, Trinity Island), and 21 (104 m) (Bentart 03, Paradise Bay). Depth: 104-233 m. 6 specimens. Additional material examined: Holotype. Slides MNHN Paris ND.T. 674 (Myxilla magna).

Description: The sponge is massive, ovate or conical in shape, with dimensions of up to 13 x 7 x 9 cm (biggest specimen). The surface is uneven, microlobular, crumbly, prominently sculptured, with ridges of 4-5 mm in length and hispid. The consistency is firm, rough to the touch, easy to tear, mucous. No differentiation between ectosome and choanosome. Oscules on the apex between the projections, 6-7 mm in diameter; some specimens have subectosomal drainage canals. Their colour in life is yellow, and beige or brownish in formalin. It may give the alcohol a slightly yellowish tint (Fig.7).

Table 3: Spicule micrometries of M. mollis Ridley and Dendy, 1886; M. spongiosa Ridley and Dendy, 1886 and M. magna Topsent, 1916. Megascleres (m) Styles 700 x 20 Tridentate chelae 50 Sigmas 63 x 4.5 Microscleres (m)

534

Reference

Species

Ridley and Dendy (1887)

M. spongiosa

Cuartas (1992) Isochelas II 20-23

M. spongiosa

Desqueyroux-Fandez M. spongiosa (remeasured) and van Soest (1996) BMNH 1887:5:2:93 Ridley and Dendy M. mollis (1887) Burton (1934) M. mollis Anchorate chelae II 16-27

Tylotes (spined at the end) 400 x 10 Tornotes 120 x 9-11 Anisotylotes 250-300 x 8-10 Smooth tylotes 220 x 6

Sigmas II 20-31

Koltun (1964) Anchorate chelae 37-78 Anchorate chelae I 41-51 Anchorate chelae 20

M. mollis

Unguiferous chela 30-69 Isochelae I 47-55 Tridentate chelae 40 Isochelae 16-60 Anchorate chelae I 43-80 Sigmas 20-60 Anchorate chelae II 19-23 Sigmas I 38-57 Sigmas I 52 x 4

Sigmas 30-40 Sigmas I 45-79 Sigmas 63 Sigmas 20-64 Sigmas 20-100

Desqueyroux (1975)

M. mollis

Boury-Esnault and van Beveren (1982)

M. mollis

Sigmas II 25-32 Sigmas II 16 x 1

Desqueyroux-Fandez M. mollis (1989)

Cuartas (1992)

M. mollis

Mothes and Lerner (1995) Styles 421-486 x 8 Styles 414-526 x 10-13 Styles 500-570 x 27-29 Styles 480-690 x 20-40 Styles 440-590 x 12-30

M. mollis

Styles 225-230 x 10-12 Styles 539-617 x 15-20 Subtylostyles/styles 420 x 10 Styles 720 x 5 Styles/subtylostyles (spined at the end) 420-728 x 10-37 Styles/subtylostyles (smooth or spined at the end) 400-730 Styles (smooth or spined) 492-614 x 16-22 Styles (spined at the end) 368-500 x 22-30 Styles 220-225 x 9-11 Styles 468-621 x 20-26 Tridentate chela 40-60 Tridentate chelae 18-44 Isochelae I 32-40 Isochelae I 34-49 Anchorate chelae I 73-80 Anchorate chelae I 42-85x 8-32 Tylotornotes 245-310 x 5-10 Anchorate chelae I 35-50 x 10-15 Sigmas 30-60 Sigmas 36-75 Sigmas I 40-49 Sigmas I 40-79 Sigmas I 140-220

Desqueyroux-Fandez M. mollis (remeasured) and van Soest (1996) BMNH 1887:5:2:112 Desqueyroux-Fandez M.(Myxilla) mollis and van Soest (1996) Topsent (1917) M. magna

Sigmas II 16-18 Sigmas II 20-32 Sigmas II 40-70 Sigmas I 50-182 x 2-10 Sigmas I 58-80 x 2-3 Sigmas II 30-47 x 2 Sigmas II 25-55

Exp. Bentart 95

M. magna

Tornotes (spined at the end) 220-400 x 6-10 Tornotes (spined at the end) 300-400 Tylotes (spined at the end) 252-339 x 6-9 Tornotes tylotes (spined at the end) 250 x 10 Tornotes 160 x 7 Subtylostyles (spined at the end) 294-370 x 8-11 Anysotylotes 227-283 x 4-8 Anisotylotes 221-280 x 7-10 Tylotornotes (spined at the end) 280-300 x 10 Tylotornotes 275-330 x 7-12

Exp. Bentart 03

M. magna

Isochelas II 24-28 Isochelas II 16-22 Anchorate chelae II 23-27 Anchorate chelae II 20-28 x 2-7 Anchorate chelae II 18-30 x 3-6

535

Fig. 8: Myxilla (Burtonanchora) magna Topsent, 1916. A. Style. B-C. Detail of extremities of style. D. Tylote. E-F. Ends of tylote. G-I. Anchorate chelae I. J-L. Anchorate chelae II. M. Sigma I. N. Sigma II.

Skeleton: The choanosomal skeleton is composed of parallel tracts of 5-6 styles interconnected with secondary tracts of one or two spicules, perpendicular or oblique near the surface of the sponge. The tracts can reach the surface and support the conules. Inside, the choanosome is reticulate with an irregular polygonal network of spicules. Anchorate chelae of both classes are rare here. The ectosomal skeleton is composed of peripheral choanosomal fibres tangential to surface; with a dense layer of sigmas of two size classes and some anchorate chelae. The ectosomal tylotes are arranged paratangentialy or in a paucispicular palisade (Fig. 7B).

Spicules (Figs. 7, 8, 9): Megascleres - Choanosomal styles straight or slightly curved in distal third of shaft, or with two curvatures in the shaft. Size: 440-690 (565.33) x 12.5-40 (24.93) m. Ectosomal tylotes straight with a slight swelling on each end, with some short spines or sometimes mucronate. Size: 245-330 (292.08) x 5-12.5 (8.64) m. Microscleres - Anchorate chelae I with the shaft slightly curved and three separated and spatulated teeth in each end. The shaft is thinner in the middle. Size: 35-85 (61) x 8.75-32.5 (20.78) m. Anchorate chelae II have three, occasionally four

536

Fig. 9: Myxilla (Burtonanchora) magna Topsent, 1916. A. Slides of type. B-C. Skeleton. D. Spicules. E. Tylotes and sigmas I. F. Microscleres. G. Spicules. H. Sigmas I and II. I-K. Anchorate chelae I and II.

teeth and two alae. Short shaft slightly curved. Size: 18.7530 (22.69) x 2.5-7.5 (5.08) m. Sigmas I are big C and S shaped with tapering sharp points. Size: 50-182.5 (112.11) x 2.5-10 (5.75) m. Sigmas II are smaller C shaped and some of them S shaped. Tapering or abrupt points. Size: 25-75 (40.94) x 2.5 m. Distribution (Fig. 18): Antarctic: Peltier Channel (Topsent, 1917); Trinity Island, Paradise Bay (present study). Remarks: Myxilla (Burtonanchora) magna is defined as possessing ectosomal tylotes (Topsent 1916: 7) with different ends, being either smooth with a terminal spine

like a mucron, or possessing groups of spines at each end. The megascleres are smooth styles, and the microscleres are anchorate chelae and sigmas in two size classes each (Fig. 9). The validity of this species has, however, been questioned since it is often described or referred to as a synonym of other Burtonanchora species. Boury-Esnault and van Beveren (1982) described two species of Myxilla from the Kerguelen Islands: M. basimucronata and M. mollis. In their paper the synonymies proposed by Burton (1932) were refuted and Myxilla magna considered a valid species. We concur with this opinion and add an additional character to clearly identify this species (Table 3). Within the ectosomal

537

Fig. 10: Myxilla (Burtonanchora) pistillaris Topsent, 1916. A. Habitus. B. Skeleton. C. Spicules. a. Style. b. Tylote. c. Raphide. d. Anchorate chelae.

skeleton this species possesses a dense layer of sigmas and anchorate chelae and the choanosomal skeleton is composed of a network of parallel tracts of styles interconnected by secondary tracts (Figs. 7B, 9B). Myxilla (Burtonanchora) pistillaris Topsent, 1916 Material examined: Station 8 (Bentart 03, Peter I Island). Depth: 8 m. 1 specimen. Additional material examined: Holotype, Slides MNHN Paris ND.T. 673 (Myxilla pistillaris) P.q.P. n 55.

Description: The sponge is thickly encrusting to massive in shape, 2 cm in diameter and 4 mm thick, fixed to brachiopods. The surface is hispid and irregular. The consistency is soft, easy to tear. Their colour is grey or greenish in vivo and white, beige or brownish in ethanol due to an accumulation of sand grains in the surface (Fig. 10). Skeleton: The choanosomal skeleton is plumoreticulate with ascending irregular tracts of styles that protrude through the surface of the sponge, interconnected by single spicules. Diverging fibres of spicule tracts occur near the surface. The ectosomal skeleton is composed of tylotes forming a loose palisade with anchorate chelae in a separate layer. Raphides are not abundant but can be found in this area (Fig. 10B).

538

Fig. 11: Myxilla (Burtonanchora) pistillaris Topsent, 1916. A. Style. B-C. Base and end of style. D. Tylote. E-F. Ends of tylote. G-I. Anchorate chelaes. J. Raphide.

Spicules (Figs. 10, 11, 12): Megascleres - Styles greatly curved, sometimes with double curvature; tapering to sharp points. Axial canal very thin. Size: 590-800 (717.57) x 12.5-22.5 (18.54) m. Tylotes slightly fusiform with both ends swollen with small spines. Size: 265-327.5 (295.08) x 10-12.5 (10.83) m. Microscleres - Anchorate chelae with the shaft slightly curved and three spatulate teeth in each end. Size: 27.5-46.25 (37.95) x 11.25-17.5 (13.8) m. Raphides straight or slightly curved at centre. Size: 60-135 (90.08) m.

Distribution (Fig. 18): Antarctic: Peltier Channel (Topsent, 1917); Peter I Island (present study). Remarks: Myxilla (Burtonanchora) pistillaris was first described by Topsent (1916). It was characterised by the possession of curved, smooth (with exception of heads), fusiform ectosomal tornotes (300 x 10 m) and smooth curved styles (480-500 m) with short points. Microscleres are anchorate chelae (37-73 m) with 3 teeth and very thin scarce raphides in trichodragmata are also present (90 m) (Topsent 1916) (Fig. 12). The specimen collected from Peter I Island differs slightly in the size of the anchorate chelae

539

Fig. 12: Myxilla (Burtonanchora) pistillaris Topsent, 1916. A. Slides of type. B. Skeleton. C. Styles. D. Tornote. E. Tylote and anchorate chelae. F. Microscleres. G. Anchorate chelae and raphide.

compared to Topsent (1916), but the shape and form are similar. Apart from this, the ectosomal spicules have slight swelling at each end, while Topsent described tornotes with different ends (one narrowed and the other straight). The presence of raphides differentiates this species from others in the subgenus. Myxilla (Burtonanchora) lissostyla Burton, 1938 Material examined: Stations 11 (30 m), 33 (15 m), 100 (24 m) (Bentart 94, Livingston Island), 9 (1182-1400)

(Gebrap 96, Bransfield Strait) and 8 (86 m), 19 (517 m) (Bentart 03, Peter I Island, Marguerite Bay). Depth range: 15-1400 m. 21 specimens. Additional material examined: Holotype, slides MNHN Paris ND.T. 1608 (Myxilla asigmata) and slides of the specimen BMNH 1935:10:26:29a Holotype (Myxilla lissostyla). Description: The sponge is massive in form, 8 x 6 x 2.5 cm (biggest specimen). The surface is smooth or rough and contains small sand grains. It has a cavernous appearance. The consistency is firm but easy to tear. The oscules are circular

540

Fig. 13: Myxilla (Burtonanchora) lissostyla Burton, 1938. A-B. Habitus. C. Skeleton. D. Spicules. a. Styles. b. Tylote. c. Anchorate chelaes.

and flat, 1-2 mm in diameter and uniformly distributed. Some specimens have shallow exhalant canals. Colour in life is dark orange, brown or grey; in ethanol it is beige or brown staining the ethanol a slightly yellowish tint. Embryos were noted in some specimens among bundles of styles (Fig.13). Skeleton: The choanosomal skeleton has primary spicule tracts of styles (approximately eight spicules wide). The tracts are interconnected by secondary tracts of styles at right angles. Anchorate chelae are abundant throughout the mesohyl, particularly around the inhalant and exhalant canals. The ectosomal skeleton is arranged with spined tylotes projecting through the pinacoderm in a palisade or brushes of several spicules (Fig. 13B).

Spicules (Figs. 13, 14, 15): Megascleres - Styles I, thick, short and slightly curved, sometimes with different terminations from points to rounded. Size: 350-980 (572.71) x 12.5-40 (21.86) m. Its possible to also find thinner and curved styles, probably young spicules. Size: 205-580 (435.46) x 1.25-12.5 (5.81) m. The ectosomal tylotes are straight, smooth and with acanthose heads. Size: 205-380 (276.27) x 3.75-15 (9.93) m. We have also observed tylotes with smooth tyles. Size: 190-415 (271) x 3.75-10 (6.72) m. Microscleres - Mature and developing anchorate chelae with a pronounced curvature and 3 short, sometimes bifid pointed

541

Fig. 14: Myxilla (Burtonanchora) lissostyla Burton, 1938. A-B. Detail of extremities of style. CD. Ends of tylote. E-M. Anchorate chelaes.

Table 4: Spicule micrometries of Myxilla (Burtonanchora) lissostyla Burton, 1938. Reference Burton (1938) Remeasurements (25) of type Desqueyroux (1975) Exp. Bentart 94 Exp. Gebrap 96 Exp. Bentart 03 Styles Megascleres (m) Tylotes Microscleres (m) Anchorate chelae 110 112-140 x 17.5-35 30-75 22.5-80 x 1.5-17.5 60-130 x 20-40 45-55 x 10-20

800 x 35 770-890 x 27.5-37.5 600 x 24 205-640 x 1.25 x 27.5 830-980 x 20-40 610-770 x 15-20

350 x 10 315-365 x 5-12.5 290 x 7 190-415 x 3.75-15 290-380 x 5-10 280-350 x 7.5-15

542

Fig. 15: Myxilla (Burtonanchora) lissostyla Burton, 1938. Slides of type. A-B. Skeleton. C. Styles and anchorate chelae. D-E. Detail of extremities of style. F-G. Ends of tylote. H-K. Anchorate chelaes.

teeth at each end. Size: 22.5-130 (56.09) x 1.25-40 (10.3) m. Distribution (Fig. 18): Antarctic: Hunters Station 9 (Burton 1938); Brabant Island (Desqueyroux 1975); Weddell Sea (Bartel et al. 1990); Livingston Island, Bransfield Strait, Peter I Island, Marguerite Bay (present study). Remarks: There are several references to this species in the literature (Burton 1938, Koltun 1964, Barthel et al. 1990, Desqueyroux 1975), but the lack of details about spicules or geographic distribution prevented a direct detailed comparison and therefore a clear identification of our specimens as M.

lissostyla. An examination of the type species, however, confirmed their conspecificity with M. lissostyla. The spicule dimensions are somewhat smaller than those of the holotype, also previously noted by Desqueyroux (1975). Our present collection extends the bathymetric distribution of this species from 15 to 1400 m. The original description characterized this species as possessing styles, tylotes (tornotes, strongylotes sensu Burton 1938: 12) being terminally spined and chelae (Fig. 15). Burton (1938) considered M. pistillaris Topsent, 1916 and M. novaezealandiae Dendy, 1924 as closely related taxa to Myxilla lissostyla, but differing in form and size of

543

Fig. 16: Myxilla (Ectyomyxilla) hentscheli Burton, 1929. A. Habitus. B. Skeleton. C. Spicules. a. Styles. b. Tornote. c. Acanthostyle. d. Anchorate chelae I. e. Anchorate chelae II. f. Sigma I. g. Sigma II.

chelae. Abundant material was collected during the Antarctic expedition for comparative purposes and it was found that although the spicule morphologies were similar in all specimens, spicule size varied. Specimens from Gebrap 96 expedition have spicule dimensions that are more similar to that of the holotype (Table 4). Subgenus Ectyomyxilla Hentschel, 1914 Myxilla (Ectyomyxilla) hentscheli Burton, 1929 Material examined: Station 24 (Bentart 95, Trinity Island). Depth: 214 m. 1 specimen.

Description: The sponge is massive and ficiform (fragmented specimen), 4 x 3.5 cm (Fig. 16A), easy to tear. No differentiation between ectosome and choanosome. The surface is slightly rough to the touch. It has a single oscule 6 mm in diameter on apex of the sponge. Colour is beige in formalin (Fig. 16). Skeleton: The choanosomal skeleton is reticulate, paucispicular. Ascending tracts are composed of columns of styles interconnected by thinner oblique bundles of the same styles. Microscleres are abundant, in particular in the endopinacoderm. The ectosomal skeleton is formed by the ends of the tracts with styles protruding in bouquets of the choanosomal tracts. Tornotes are tangential or paratangential to surface. Acanthostyles are very rare, seen only in spicule

544

Fig. 17: Myxilla (Ectyomyxilla) hentscheli Burton, 1929. A. Style. B-C. Base and end of style. DE. Ends of Tornote. F. Tornote. G. Acanthostyle. H-I. Detail of extremities of acanthostyle. JK. Anchorate chelae I. L-M. Anchorate chelae II. N-O. Sigma I. P. Sigma II.

preparation but not in skeletal sections. Spherulous cells are abundant (Fig. 16B). Spicules (Figs. 16, 17): Megascleres - Styles slightly curved in distal third of the shaft, with tapering sharp points. In some of them we have observed a reduced number of spines at the base. Axial canal very thin. Size: 560-615 (589) x 10-20 (16.8) m. Acanthostyles short and thin, with spines evenly dispersed over shaft and base, except on point. Rare. Size: 97.5-125 (114) x 5-10 (7.5) m. Anisotornotes smooth, straight or slightly curved, isodiametric along the shaft with unequal swelling at each end (anisotornotes - tylote or subtylote on

one side and mucronate or lanceolated on the other). Size: 252.5-335 (298) x 7.5-10 (9.5) m. Microscleres - Anchorate chelae I with three spatulate teeth on each end, slightly curved towards the shaft. Size: 48.7552.5 (51.25) x 11.25-15 (13) m. Anchorate chelae II with three spatulate teeth recurved towards the shaft. Teeth from both extremities are separated by a small gap at middle length of the chelae. Size: 22.5-30 (26) x 5-10 (6) m. Sigmas I are C shaped. Tapering points with opposed orientation. Size: 55-82.5 (73.5) x 2.5-5 (4.45) m. Sigmas II are C and S shaped with tapering sharp points. Size: 35-50 (43) x 1.252.5 (2.25) m.

545

Figure 18: Map of localities and areas where Myxilla species cited in this work were found.

Distribution (Fig. 18): Antarctic: Gauss Station (Hentschel, 1914); Trinity Island (present study). Remarks: Hentschel (1914) identified this species as Myxilla spongiosa Ridley and Dendy, 1886, highlighting the presence of small choanosomal acanthostyles, styles, tornotes and two size categories of spatuliferous anchorate chelae and sigmas. The acanthostyles are scarce and scattered within the choanosome. Burton (1929) concluded this was a valid species and renamed this population Ectyomyxilla hentscheli. The spicule characteristics of the present specimen are similar to that of the hoolotype. This species is recorded for the first time outside the type locality, extending both its geographic and bathymetric distributions.

(Australia) for thoughtful review of the manuscript. This work is part of the projects: Spanish Ministry of Science and Technology (CICYT ANT93-0996; CICYT ANT94-1161/E; CICYT ANT951011; CICYT ANT97-2097-E; CICYT ANT96-2440-E; MCYT REN2001-1074/ANT; MCYT REN2003-01881/ANT and MEC CGL2004-21066-E).

References
Barthel D, Tendal O, Panzer K (1990) Ecology and taxonomy of sponges in the eastern Weddell Sea shelf and slope communities. Rep Polar Res 68: 120-131 Boury-Esnault N, van Beveren M (1982) Les dmosponges du plateau continental de Kerguelen-Heard. Comit National franais des recherches antarctiques 52: 1-175 Burton M (1929) Porifera. Part II. Antarctic Sponges. British Antarctic Terra Nova Expedition 1910. Natural History Report. Zoology 6: 393-458 Burton M (1932) Sponges. Discovery Reports 6: 237-392 Burton M (1934) Sponges. Further zoological results of the Swedish Antarctic expedition 1901-1903 III: 1-58 Burton M (1938) Non calcareous sponges. Scientific Reports Australasian Antarctic Expedition, Serie C (Zoology and Botany) 9: 5-22

Acknowledgements
We thank colleagues for provided the material studied here particularly Dr. Ana Ramos of the Spanish Institute of Oceanography, leader of Bentarts Projects and Professor Francisco Ramil of Vigo University who collected the material of Gebrap 96 Expedition. We are particularly grateful to Professor Claude Lvi of the Museum National dHistoire Naturelle of Paris for allowing access to type collections and for his many positive comments and friendship, and to Professor John Hooper from the Queensland Museum

546

Cristobo FJ, Urgorri V, Solrzano MR, Ros P (1993) Mtodos de recogida, estudio y conservacin de las colecciones de porferos. In: Palacios F, Martnez C, Thomas B (eds). International Symposium and First World Congress on Preservation and Conservation of Natural History Collections, Madrid 2. Direccin General de Bellas Artes y Archivos. Ministerio de Cultura, Madrid. pp. 277-287 Cuartas E (1992) Poriferos de la provincia biogeogrfica argentina. III. Poecilosclerida (Demospongiae), del litoral marplatense. PHYSIS Seccin A 47(113): 73-88 Desqueyroux R (1975) Esponjas (Porifera) de la regin antrtica chilena. Cah Biol Mar 16: 47-82 Desqueyroux-Fandez R (1989) Demospongiae (Porifera) del litoral chileno antrtico. Serie Cientfica Instituto Antrtico Chileno 39: 97-158 Desqueyroux-Fandez R, van Soest RWM (1996) A review of Iophonidae, Myxillidae and Tedaniidae occurring in the South East Pacific (Porifera: Poecilosclerida). Rev Suisse Zool 103(1): 3-79 Gutt J, Koltun VM (1995) Sponges of the Lazarev and Weddell Sea, Antarctica. Explanations for their patchy occurrence. Antarctic Sci 7(3): 227-234 Hentschel E (1914) Monaxone Kieselschwmme und Hornschwmme der deutschen Sdpolar Expedition. 1901-1903. Deutschen Sdpolar Expedition Zoology 7: 37-141 Hooper JNA, van Soest RWM (eds) (2002) Systema Porifera: a guide to the classification of sponges. Kluwer Academic/Plenum Publishers, New York Koltun VM (1964) Sponges of the Antarctic. I. Tetraxonida and Cornacuspongidae. Academy of Sciences of U.S.S.R. Zoological Institute. Explorations of the Fauna of the Seas. Biological results of the Soviet Antarctic Expedition (1955-1958) 2: 1-116 Koltun VM (1976) Porifera. Part I: Antarctic sponges. British Australian New Zealand Antarctic Research Expedition 19291931 Reports Series B (Zoology and Botany) 9(4): 147-198 Mothes B, Lerner C (1995). Ectyonancora ruthae sp. n. (Myxillidae) e outras esponjas detectadas na 1 expedio antrtica brasileira (Porifera; Hexactinellida e Demospongiae). Biocincias 3(1): 155-171

Pansini M, Calcinai B, Cattaneo-Vietti R, Sar M (1994) Demosponges from Terra Nova Bay (Ross Sea, Antarctica): 1987/88 and 1989/89. Programma Nazionale di Ricerche in Antartide. Expeditions. National Scientific Commission for Antarctica Oceanographic Campaign 1987/88 Oceanographic Campaign 1989/90 3: 67-100 Ridley SO, Dendy A (1886) Preliminary report on the Monaxonida collected by H.M.S. Challenger. Part II. Ann Mag Nat Hist 18: 470-496 Ridley SO, Dendy A (1887) Report on the Monaxonida collected by H.M.S. Challenger. Rep Sci Res Voy H.M.S. Challenger, 20: 1-275 Ros P, Cristobo FJ (2006) A new species of Biemna (Porifera: Poecilosclerida) from Antarctica: Biemna strongylota. J Mar Biol Assoc UK 86: 949-955 Ros P, Cristobo FJ, Urgorri V (2004) Poecilosclerida (Porifera, Demospongiae) collected by the Spanish Antarctic expedition BENTART-94. Cah Biol Mar 45: 97-119 Rtzler K (1978) Sponges on coral reefs. In: Stoddart DR, Johanness RE (eds). Coral reefs: research methods. Unesco, Paris. pp. 81120 Topsent E (1901) Spongiaires. Rsultats du Voyage du S.Y. Belgica (1897-1899) sous le commandement de A. de Gerlache de Gomery. Expdition antarctique belge, Zoologie 4: 1-54 Topsent E (1908) Spongiaires. Expdition antarctique franaise (1903-1905) commande par le Dr. Jean Charcot (Paris) 4: 1-37 Topsent E (1913) Spongiaires de la expdition antarctique nationale cossaise. Trans Royal Soc Edinburgh 49(3): 579-673 Topsent E (1916) Diagnoses dponges recueillies dans lAntarctique par le Pourquoi-pas?. Bull Mus Hist Nat Paris 3: 163-172 Topsent E (1917) Spongiaires. In: Joubin L (ed.). Deuxime Expdition Antarctique Franaise (19081910) commande par le Dr. Jean Charcot (Paris) 4: 1-88 van Soest RWM (2002) Family Myxillidae Dendy, 1922. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges. Kluwert Academic/Plenum Publishers, New York. pp 602-620

Porifera research: Biodiversity, innovation and sustainaBility - 2007

547

A new species of Cinachyra (Demospongiae: Tetillidae) collected by Project REVIZEE off Esprito Santo State, SE Brazil
Pablo R.D. Rodriguez, Guilherme Muricy(*)
Laboratrio de Porifera, Departamento de Invertebrados, Museu Nacional, Universidade Federal do Rio de Janeiro. Quinta da Boa Vista s/n, 20940-040, Rio de Janeiro, RJ, Brasil. muricy@acd.ufrj.br Abstract: A new species of Cinachyra Sollas (Porifera, Spirophorida) was dredged from 500 m depth off Espirito Santo State, SE Brazil, by Project REVIZEE Central SCORE. Cinachyra helena sp. nov. is characterized by the presence of a single category each, of protriaenes, anatriaenes, anisoactinal choanosomal oxeas, isoactinal cortical oxeas, and sigmaspires. The cortex is formed by small isoactinal oxeas arranged obliquely to the surface, and an anchoring basal spicule mass is absent. This is the first valid record of Cinachyra from the Atlantic; all similar species lacking a cortex should be transferred to Cinachyrella or other genera of Tetillidae. At the present state of knowledge, there are four valid species in the genus Cinachyra: C. barbata Sollas, C. crustata (Wilson), C. uteoides Dendy, and C. helena sp. nov. Keywords: Cinachyra helena, new species, Porifera, Project REVIZEE, Southwestern Atlantic, taxonomy

Introduction
The genus Cinachyra Sollas, 1886, of the demosponge family Tetillidae Sollas, 1886, comprises globular sponges with a radial skeleton of oxeas and triaenes, sigmaspires, a cortex reinforced by auxiliary oxeas, and flask-shaped porocalyces (van Soest and Rtzler 2002). There is considerable confusion in the literature distinguishing between the closely related genera Cinachyra and Cinachyrella Wilson, 1925, because many authors disregarded the presence or absence of a distinctive cortex, which is now considered the most important character to distinguish between the two genera (Rtzler 1987, van Soest and Rtzler 2002). Rtzler (1987) and Rtzler and Smith (1992) revised the Western Atlantic species of Cinachyra, and concluded that they all belong instead to Cinachyrella, in which the cortex is absent. In the most recent revision of the family Tetillidae, van Soest and Rtzler (2002) recognized only a single valid species as belonging with certainty to the genus Cinachyra, viz., C. barbata Sollas, 1886. So far, twelve valid species of tetillid sponges have been described from Brazil: Acanthotetilla rocasensis Peixinho et al., 2007; Acanthotetilla walteri Peixinho et al., 2007; Cinachyrella alloclada (Uliczka, 1929), Cinachyrella apion (Uliczka, 1929), Cinachyrella kuekenthali (Uliczka, 1929), Craniella carteri Sollas, 1886, Craniella corticata (BouryEsnault, 1973), Craniella quirimure Peixinho, Cosme and Hajdu, 2005, Tetilla euplocamus Schmidt, 1868, and Tetilla radiata Selenka, 1879 (Schmidt 1868, Selenka 1879, Sollas 1886, 1888, Fischel-Johnson 1971, Boury-Esnault 1973, Rtzler and Smith 1992, Lazoski et al. 1999, Santos and Hajdu 2003; Peixinho et al. 2005; Muricy and Hajdu 2006).

Craniella cranium Mller, 1776 was quoted from Fernando de Noronha Archipelago (Carter 1890); however, Carters description was insufficient, and this record was considered questionable and unrecognisable (Hechtel 1976, Moraes et al. 2006). Cinachyra rhizophyta Uliczka 1929, recorded from Cear State (Fischel-Johnson 1971) was put in synonym with Cinacyhrella apion (cf. Rtzler and Smith 1992). Unfortunately, many Brazilian records of these species, particularly of Cinachyrella spp., did not include descriptions (Mello Leito et al. 1961, Hechtel 1976, Collette and Rtzler 1977, Atta et al. 1989, Muricy et al. 1991, 1993, 2006, Muricy and Moraes 1998, Lobo-Hajdu et al. 1999, Santos et al. 1999, 2002, Muricy and Silva 1999, Moraes et al. 2003, 2006, Santos et al. 2004, Hajdu et al. 2004). Two of such more recent records are of undescribed species of Cinachyra (Cinachyra sp.; Hajdu et al. 2004, Muricy et al. 2006). Since the genus Cinachyra was until recently considered to have a cold-temperate to polar distribution (Kerguelen Islands, Patagonia and Antarctica; van Soest and Rtzler 2002), we decided to investigate in more detail the two records of Cinachyra from the subtropical SW Atlantic. Both species turned out to be new, but only one of them really belongs to Cinachyra. In this study, we describe the new species of Cinachyra collected off Espirito Santo State, SE Brazil, by Project REVIZEE Central SCORE (Muricy et al. 2006), and discuss the identity of the record of Cinachyra sp. from So Paulo State, Southern Brazil (2420527S 4346759W), collected by Project REVIZEE South SCORE (Hajdu et al. 2004). The Project REVIZEE (in portuguese, Programa de Avaliao do Potencial Sustentvel dos REcursos VIvos da Zona Econmica Exclusiva, or Program of Evaluation of the

548

Sustainable Resources of the Economic Exclusive Zone) is an initiative of the Brazilian Government to comply with the United Nations Convention on the Law of the Sea (UNCLS). The Brazilian continental shelf and slope (down to 2,076 m depth) were divided in four sectors (called SCORES: North, Northeast, Central, and South), in which extensive surveys were done to estimate the diversity and abundance of planktonic, nectonic and benthic organisms and their sustainable exploitation potential (eg, Amaral and RossiWongtschowski 2004, Costa et al. 2005, Lavrado and Ignacio 2006).

Cinachyra helena sp. nov. (Figs. 23, Tab. 1) Synonyms: Cinachyra sp., Muricy et al. 2006: 115 Diagnosis: Cinachyra with a single category each of protriaenes, anatriaenes, anisoactinal choanosomal oxeas, isoactinal cortical oxeas, and sigmaspires. Cortex formed by small isoactinal oxeas arranged obliquely to the surface. Anchoring spicule tufts absent. Material examined: Holotype: MNRJ 3635C, Esprito Santo State, Brazil; dredging, project REVIZEE Central SCORE II, station 20C-deep, RV Astro Garoupa, 22/XI/1997, 1917`S 3757`W, 500 m depth. Paratype: MNRJ 3658B, same locality, date and collector. Material examined for comparison: Cinachyra sp. sensu Hajdu et al. (2004): MNRJ 2814 A-H, So Paulo state, Brazil; dredging, project REVIZEE South SCORE, station 6659, RV Prof. W. Besnard, 09/I/1998, 2420S 4346W, 505 m depth. Description: External form hemispherical to subspherical (Fig. 2). Size up to 2.5 cm in diameter by 1.5 cm high. Color in vivo unknown; beige after fixation in alcohol 70%. The holotype has irregular porocalyces, up to 3 x 1 mm wide by 2.5 mm deep (Fig. 2A). A thin, whitish cortex is clearly visible in sectioned specimens (Fig. 2B). The paratype was fragmented, and the part of its surface that is left has no porocalyces (Fig. 2C). Oscules are not visibe in preserved specimens, probably contracted. Surface even, strongly hispid, with adhering sand grains. The sponge is fixed by a broad attachment base, without anchoring basal spicule mass (Fig. 2D). Consistency firm, slightly compressible, inelastic. Skeleton: Choanosomal skeleton with radial tracts of anisoactinal coanossomal oxeas, anatriaenes and protriaenes, which project through the cortex and the surface of the sponge (Fig. 3A). The cladomes of both pro- and anatriaenes are directed outwards. Tracts are wider (580-733 m thick) in the ectosome than in the choanosome (76-202 m thick), and are 50530 m apart. The cortex is composed of two layers (Fig. 3A): the internal layer (505-884 m thick) is formed by small isoactinal oxeas arranged obliquely to the surface; the outer layer (253-1137 m thick) is formed only by sigmaspires and the extremities of the radial tracts of megascleres. Sigmaspires are dispersed randomly both in the cortex and in the choanosome. Spicules: Choanosomal anisoactinal oxeas, slightly curved at the thinner extremity: 2075-3305-4300 x 40-43-50 m (Fig. 3B). Cortical isoactinal oxeas short, robust, fusiform: 425650-850 x 26-34-43 m (Fig. 3C). Anatriaenes rare, varying from very thin, reduced forms to long and robust spicules, with clads short and slightly curved, forming an almost straight angle with the rhabdome: clads 19-72-116 x 2-18-31 m; rhabdome 1310-3421-8381 x 5-24-34 m (Fig. 3D, E). Protriaenes abundant, occasionally irregular, with clads long and straight and rhabdome long, ending abruptly: clads 130214-275 x 12-16-21 m, rhabdome 1500-3258-6250 x 17-2334 m (Figs. 3F, G). Sigmaspires C- or S-shaped, with spines

Material and methods

The area sampled by Project REVIZEE Central SCORE ranges from Salvador (11S), Bahia state, to Cabo de So Tom (22S), in Northern Rio de Janeiro state, including the islands and seamounts of the Vitria-Trindade chain (Fig. 1). The continental shelf breaks at approximately 70 m depth, and the seafloor is dominated by carbonatic sediments, corals and calcareous algae. The slope is mostly covered by foraminiferan deposits. Sponges have the greatest biomass among benthic organims in this area (Lavrado 2006). A complete description of the environment and benthic communities of the Central SCORE region is given by Lavrado and Ignacio (2006). Collections were done by dredging and bottom trawling on board of the RV Astro Garoupa, between 20 and 2,076 m depth, from 19/X/1997 a 24/XI/2003. Sponges were fixed in ethanol 70% or formalin 4%, and deposited in the Porifera collection of Museu Nacional (Rio de Janeiro, Brazil). Photographs of both preserved specimens and of the skeleton by light microscopy were taken with a Nikon Coolpix digital camera. Spicule slides were prepared by dissociation of a small fragment of sponge in boiling nitric acid. Sponge fragments were dehydrated in an alcohol series (50-100%) with a final xylene step and included in paraffin. Transverse sections of the skeleton were mounted on microscope slides for identification. Twenty spicules of each kind were measured per specimen. Size ranges and means (underlined) are given in the text.

Systematics
Phylum Porifera Grant, 1836 Class Demospongiae Sollas, 1885 Subclass Tetractinomorpha Lvi, 1953 Order Spirophorida Bergquist and Hogg, 1969 Family Tetillidae Sollas, 1886

Genus Cinachyra Sollas, 1886

Definition: Tetillidae with cortex reinforced by auxiliary oxeas, with flask-shaped porocalyces (van Soest and Rtzler 2002).

549

Fig. 1: Area studied by the Project REVIZEE, Central SCORE, between Salvador and Cabo de So Tom, Brazil, with the location of the collecting site of Cinachyra helena sp n. (BA, Bahia State; ES, Esprito Santo State; RJ, Rio de Janeiro State).

large, recurved, and more abundant close to the tips, which are enlarged, irregular: 10-14-21 m (Fig. 3H). Ecology: The paratype (MNRJ 3658B) was growing over a foliaceous sponge, Phakellia sp. The two specimens were partially covered by sand grains. Geographical and bathymetric distribution: Off Espirito Santo State, Brazil (1917`14``S 3757`13``W), 500 m depth (Fig. 1). Etymology: This species was named in honour of Dr. Helena Passeri Lavrado, in recognition of her successful efforts to coordinate the Benthic Ecology Group of Project REVIZEE Central SCORE.

Discussion
Only two specimens of the new species were found; of these, only the holotype had porocalyces (Fig. 2A). This is an important issue, since the presence of porocalyces is the main character that separates Cinachyra from Craniella Schmidt, 1870 (van Soest and Rtzler 2002). The paratype was fragmented, and most of its surface was lost (Fig. 2C). Since the spicules and all other characters of the two

specimens were nearly identical in shape and size (Table 1), we assumed that the porocalyces were present in the lost part of the paratypes surface. The genus Cinachyra is currently considered to contain a single species, Cinachyra barbata Sollas, 1886, and most other species previously described as Cinachyra were transferred to Cinachyrella (van Soest and Rtzler 2002). However, a literature survey indicates that at least one other species described as Cinachyra (C. uteoides Dendy, 1924) and another described as Tetilla (Tetilla (Cinachyrella) crustata Wilson, 1925) do have both porocalyces and a cortex made up of small oxeas, and should therefore be considered as valid species of Cinachyra. Bergquist (1968) expressed doubts whether C. novae-zealandiae Brondsted, 1924 has true porocalyces, since Brondsteds (1924) description is not clear about the nature of its aquiferous openings. It is clear however from the original description that C. novaezealandiae has a cortex of small curved oxeas forming a palisade. We agree with Bergquist (1968) that a revision of the holotype is needed to clarify the best taxonomic placement of C. novae-zealandiae, if in Cinachyra or in Craniella (but not in Tetilla, as suggested by Bergquist, loc. cit.). For the time being, our interpretation is that the aquiferous openings

550

Fig. 2: Cinachyra helena sp. nov. A. upper view of holotype (MNRJ 3635C) showing the irregular porocalyces in the surface (arrows). B. side view of holotype, showing the radial arrangement of the skeleton and the lighter cortex (arrow). C. upper view of paratype (MNRJ 3658B), in which most of the surface was lost. D. basal view of paratype, showing the broad, flattened attachement surface (scale bars = 1 cm).

Table 1: Measurements (min-med-max length / min-med-max width in m) of the spicules of Cinachyra helena sp. nov. (n=20). Spicules Choanosomal oxeas Cortical oxeas Protriaene clads Protriaene rhabdome Anatriene clads Anatriaene rhabdome Sigmaspires Holotype MNRJ 3635C 3050-3635-4300 / 34-41-50 425-734-850 / 26-34-43 130-198-240 / 15-17-20 2150-3615-6250 / 19-23-26 38-75-97 / 10-20-29 2425-4315-8381 / 14-26-34 12-14-22 / 0.5-2 Paratype MNRJ 3658B 2075-2976-4050 / 40-44-50 450-566-775 / 26-34-41 175-231-275 / 14-17-22 1500-2901-5000 / 17-22-34 19-70-117 / 2-17-31 1310-2527-6383 / 5-21-44 10-14-18 / 0.5-2

of C. novae-zealandiae are not porocalyces, and therefore we suggest to classify it as Craniella novae-zealandiae. A detailed taxonomic revision of the family Tetillidae, including observation of type specimens of as many species as possible, is clearly needed for a better understanding of the scope of the genus Cinachyra. Cinachyra barbata presents two categories of protriaenes, one of anatriaenes, two of isoactinal oxeas, and one of sigmaspires. The anatriaenes only occur in the basal spicule mass. Cinachyra helena sp. nov. differs from Cinachyra barbata by the possession of only one category of protriaenes, choanosomal anatriaenes, and the absence of a basal mass. Furthermore, its choanosomal anisoactinal oxeas and cortical isoactinal oxeas are smaller than those of Cinachyra barbata, which has choanosomal isoactinal oxeas of approximately 8000 x 70 m and cortical isoactinal oxeas around 900 x 36 m (Sollas 1888).

Cinachyra helena sp. nov. is similar to C. uteoides in the absence of a basal (anchoring) spicule mass: both species are more or less flattened basally, with a broad attachment base. Cinachyra uteoides differs from the new species however by the presence of protriaenes in two size classes, the isoactinal choanosomal oxeas, the larger cortical oxeas (2600 x 80 m), and the smaller rhabdome of the anatriaenes (2400-2900 x 12-16 m) (Dendy 1924, Bergquist 1968). Cinachyra crustata (Wilson 1925) differs from C. helena sp. nov. in many characters, such as the flattened body shape, presence of root-like spicule tracts (= anchoring basal spicule mass), larger porocalyces, thinner cortex with tangentiallydisposed rather than obliquely-disposed oxeas, and larger, isoactinal choanosomal oxeas (7000-9000 m). Although it is uncertain whether it really has porocalyces (cf. Bergquist 1968), Craniella novae-zealandiae Brondsted 1924 shares with the new species the globular shape without spicule basal mass and a strongly hispid surface. They differ,

551

Fig. 3: Skeletal characters of Cinachyra helena sp. nov. A. transverse section of the skeleton showing the cortex of smaller oxeas and the radial megasclere tracts. B. choanosomal, anisoactinal oxea. C. cortical, isoactinal oxea. D. anatriaene. E. cladome of anatriaene. F. protriaene. G. cladome of protriaene (SEM) H. sigmaspire (SEM).

however, in that the cortical oxeas of C. novae-zealandiae are curved, its choanosomal oxeas are isoactinal, all megascleres are smaller, and it apparently has no sigmaspires (Brondsted 1924). We reexamined the specimens of Cinachyra sp. recorded from Southern Brazil (Hajdu et al. 2004). They are flattened, discoidal, with a fringe of projecting spicules.

Both porocalyces and a spiculated cortex are absent, and therefore they would be better classified in Tetilla Schulze 1868, or maybe in a new genus defined by a discoidal shape, rather than in Cinachyra. Cinachyra rhizophyta, recorded from Cear state, NE Brazil (Fischel-Johnson 1971), was synonymized with Cinachyrella apion by Rtzler and Smith (1992). However, the description of C. rhizophyta from

552

Cear does not mention the two categories of protriaenes or the presence of raphides, characters typical of C. apion. We consider the record of Cinachyra rhizophyta sensu FischelJohnson (1971) as a synonym of Cinachyrella alloclada, rather than of C. apion. This, however, does not change the list of tetillid species from Brazil, which now includes eleven valid species: Acanthotetilla rocasensis, Acanthotetilla walteri, Cinachyra helena sp. nov., Cinachyrella alloclada, Cinachyrella apion, Cinachyrella kuekenthali, Craniella carteri, Craniella corticata, Craniella quirimure, Tetilla euplocamus, and Tetilla radiata. The genus Cinachyra appears to be more frequent in deep than in shallow water, with records of C. barbata from 45549 m depth (Sollas 1886, 1888), of C. uteoides from 55182 m (Dendy 1924, Bergquist 1968), and of the new species from 500 m depth. Cinachyra helena sp. nov. is the first valid record of the genus from the Brazilian coast and the Atlantic Ocean. The distribution of the genus Cinachyra now includes Kerguelen Island, Patagonia and Antarctica (C. barbata), New Zealand (C. uteoides), Philippines (C. crustata) and SE Brazil (C. helena sp. nov.).

Acknowledgements
We are grateful to Conselho Nacional de Desenvolvimento Cientfico e Tecnolgico (CNPq), Fundao Carlos Chagas Filho de Apoio Pesquisa do Estado do Rio de Janeiro (FAPERJ), Secretaria da Comisso Interministerial para os Recursos do Mar (SECIRM), and CENPES-PETROBRAS for financial and logistic support. We also thank Dr. Marcia Attias and Nomia Rodrigues (Laboratrio de Ultraestrutura Celular Herta Meyer, Instituto de Biofsica Carlos Chagas Filho, UFRJ) for their help in the use of SEM. We are particularly grateful to Dr. Helena P. Lavrado for inviting us to identify the sponges of Project REVIZEE Central SCORE.

References
Amaral ACZ, Rossi-Wongtschowski C (eds). (2004) Biodiversidade bentnica da regio sudeste-sul do Brasil plataforma externa e talude superior. Srie Documentos REVIZEE Score Sul. Instituto Oceanogrfico USP, So Paulo Atta AM, Barral M, Peixinho S, Souza-Atta MLB (1989) Isolation and functional characterization of a mitogenic lectin from the marine sponge Cinachyrella alloclada. Braz J Med Biol Res 22: 379-385 Bergquist PR (1968) The marine fauna of New Zealand: Porifera, Demospongiae, Part I. (Tetractinomorpha and Lithistida). New Zeal Oceanogr Inst Mem 37(188): 1-105 Bergquist PR, Hogg JJ (1969) Free aminoacid patterns in Demospongiae: a biochemical approach to sponge classification. Cah Biol Mar 10: 205-220 Boury-Esnault N (1973) Campagne de la Calypso au large des ctes Atlantiques de lAmrique du Sud (1961-1962) 29. Spongiaires. Rs Sci Camp Calypso 10: 263-295 Brondsted HV (1924) Papers from Dr. Th. Mortensens Pacific Expedition 1914-16. XXIII. Sponges from New Zealand. Part I. Vidensk Meddel Dansk Naturh Foren Kjoben 77: 435-483 Carter HJ (1890) Porifera. Notes on the zoology of Fernando de Noronha. J Linn Soc 20: 564-569

Collette B, Rtzler K (1977) Reef fishes over sponge bottoms off the mouth of the Amazon River. Proc 3rd Int Coral Reef Symp, Miami 1: 305-310 Costa PAS, Martins AS, Olavo G (eds). Pesca e potenciais de explorao de recursos vivos na regio central da Zona Econmica Exclusiva brasileira. Museu Nacional, srie Livros 13, Rio de Janeiro Dendy A (1924) Porifera. Part I. Non-Antarctic sponges. British Antarctic (Terra Nova) Expedition, 1910 (Zoology). Nat Hist Rep 6(3): 269-392 Fischel-Johnson M (1971) Some marine sponges of northeast Brazil. Arq Cinc Mar 11(2): 103-116 Grant RE (1836) Animal Kingdom. In: Todd RB (ed). The cyclopaedia of anatomy and physiology. Vol. 1. Sherwood, Gilbert, and Piper, London. pp. 107-118 Hajdu E, Santos CP, Lopes DA, Oliveira MV, Moreira MCF, Carvalho MS, Klautau M (2004) Filo Porifera. In: Amaral CZA, Rossi-Wongtschowski CLB (Eds) Biodiversidade bentnica da regio sudeste-sul do Brasil plataforma externa e talude superior. Srie Documentos REVIZEE Score Sul. Instituto Oceanogrfico USP, So Paulo. pp. 49-56 Hechtel GJ (1976) Zoogeography of Brazilian marine Demospongiae. In: Harrison FW, Cowden RR (eds). Aspects of sponge biology. Academic Press, New York & London. pp. 237-260 Lavrado HP (2006) Captulo 1 Caracterizao do ambiente e da comunidade bentnica. In: Lavrado HP, Igncio BL (eds). Biodiversidade bentnica da regio central da Zona Econmica Exclusiva brasileira. Museu Nacional, Srie Livros 18, Rio de Janeiro. pp. 19-64 Lavrado HP, Igncio BL (2006) Biodiversidade bentnica da regio central da Zona Econmica Exclusiva brasileira. Museu Nacional, Srie Livros 18, Rio de Janeiro Lazoski C, Peixinho S, Russo CAM, Sol-Cava AM (1999) Genetic confirmation of the specific status of two sponges of the genus Cinachyrella (Porfera: Demospongiae: Spirophorida) in the southwestern Atlantic. Mem Qld Mus 44: 299-305 Lvi C (1953) Sur une nouvelle classification des Dmosponges. Compt Rend Hebdom San Acad Sci Paris 236(8): 853-855 Lbo-Hajdu G, Mansure JJ, Salgado A, Hajdu E, Muricy G, Albano RM (1999) Random amplified polymorphic DNA (RAPD) analysis can reveal intraspecific evolutionary patterns in Porifera. Mem Qld Mus: 317-328. Mello-Leito A, Pgo AF, Lopes WM (1961) Porferos assinalados no Brasil. Avul Centr Est Zool Univ Brasil 10: 1-29 Moraes FC, Ventura M, Klautau M, Hajdu E, Muricy G (2006) Biodiversidade de esponjas das ilhas ocenicas brasileiras. In: Alves RJV, Castro, JWA (eds). Ilhas Ocenicas Brasileiras da pesquisa ao manejo. Ministrio do Meio Ambiente, Brasilia. pp. 147-178 Moraes FC, Vilanova EP, Muricy G (2003) Distribuio das esponjas (Porifera) na Reserva Biolgica do Atol das Rocas, Nordeste do Brasil. Arq Mus Nac 61(1): 13-22 Muricy G, Moraes FC (1998) Marine sponges of Pernambuco state, NE Brazil. Rev Bras Oceanogr 46(2): 213-217 Muricy G, Silva OC (1999) Esponjas marinhas do Estado do Rio de Janeiro: um recurso renovvel inexplorado. In: Silva SHG, Lavrado HP (eds). Ecologia dos Ambientes Costeiros do Estado do Rio de Janeiro. Srie Oecologia Brasiliensis 7: 155-178

553

Muricy G, Hajdu E (2006) Porifera Brasilis. Guia de identificao das esponjas marinhas mais comuns do sudeste do Brasil. Museu Nacional, Srie Livros 17, Rio de Janeiro Muricy G, Hajdu E, Custdio M, Klautau M, Russo C, Peixinho S (1991) Sponge distribution at Arraial do Cabo, SE Brazil. In: Magoon OT, Converse H, Tippie V, Tobin LT, Clark D (eds). Coastal Zone 91. Proc 7th Symp Coast Ocean Manag. ASCE Publications, Long Beach. pp. 1183-1196 Muricy G, Hajdu E, Araujo FV, Hagler AN (1993) Antimicrobial activity of Southwestern Atlantic shallow-water marine sponges (Porifera). Sci Mar 57(4): 427-432 Muricy G, Santos CP, Batista D, Lopes DA, Pagnoncelli D, Monteiro LC, Oliveira MV, Moreira MCF, Carvalho MS, Melo M, Klautau M, Rodriguez PRD, Costa RN, Silvano RG, Schientek S, Ribeiro SM, Pinheiro US, Hajdu E (2006) Captulo 3. Filo Porifera. In: Lavrado HP, Igncio BL (2006) Biodiversidade bentnica da regio central da Zona Econmica Exclusiva brasileira. Museu Nacional, Srie Livros 18, Rio de Janeiro. pp. 109-145 Peixinho S, Cosme B, Hajdu E (2005) Craniella quirimure sp. nov. from the mangroves of Bahia (Brazil) (Tetillidae, Spirophorida, Demospongiae). Zootaxa 1036: 31-42 Peixinho S, Fernandez J, Oliveira MV, Cares S, Hajdu E (2007) Description of two new species of Acanthotetilla Burton, 1959 from NE Brazil, Southwestern Atlantic (Tetillidae, Spirophorida, Demospongiae). In: Custdio MR, Lbo-Hajdu G, Hajdu E, Muricy G (eds). Porifera research: biodiversity, innovation and sustainability. Srie Livros 28. Museu Nacional, Rio de Janeiro. pp. 509-515 Rtzler K (1987) Tetillidae (Spirophorida, Porifera): a taxonomic reevaluation. In: Vacelet J, Boury-Esnault N (eds). Taxonomy of Porifera from the N.E. Atlantic and Mediterranean Sea. NATO ASI Series G 13. Springer-Verlag, London. pp.187-203 Rtzler K, Smith KP (1992) Guide to western Atlantic species of Cinachyrella (Porifera: Tetillidae). Proc Biol Soc Wash 105: 148164 Santos JP, Mothes B, Tenrio DO, Cantarelli J (1999) Porifera (Demospongiae, Calcarea) entre os estados do Cear e Pernambuco, Brasil. Taxonomia e distribuio. Trab Oceanogr UFPE 27(2): 4960

Santos JP, Cantarelli J, Tenrio DO (2002) Porifera do Estado de Pernambuco, Brasil. In: Tabarelli M, Silva JMC (eds). Diagnstico da biodiversidade de Pernambuco 2. Editora Massangana, Recife. pp. 385-404 Santos CP, Hajdu E (2003) Redescription of Tetilla radiata Selenka from the Southwestern Atlantic (Porifera, Spirophorida, Tetillidae) and designation of its neotype. Rev Bras Zool 20(4): 637-642 Santos CP, Hajdu E, Muricy G (2004) An identification system for common Demospongiae of the So Sebastio Channel Area, SW Atlantic, developed with the Linnaeus II software. Boll Mus Ist Biol Univ Gen 68: 587-591 Schmidt O (1868) Die Spongien der kste von Algier. Mit nachtrgen zu den Spongien des Adriatischen Meeres (Drittes Supplement). Wilhelm Englemann, Leipzig Schmidt O (1870) Grundzge einer Spongien fauna des Atlantischen Gebietes. Wilhelm Engelmann, Leipzig Selenka E (1879) Ueber einem Kieselschwamm von achtstrahligem Bau, und ber Entwicklung der Schwammknospen. Zeitsch Wiss Zool: 472-475 Sollas WJ (1885) A classification of the sponges. Ann Mag Nat Hist 5(16): 95-395 Sollas WJ (1886) Preliminary account of the tetracinellid sponges dredged by H.M.S Challenger 1872-76. Part I. The Choristida. Sci Proc Royal Dublin Soc 5: 177-199 Sollas WJ (1888) Report on the Tetractinellida collected by H.M.S. Challenger during the years 1873-1876. Rep Sci Res Voy H.M.S Challenger, Zool 25: 1-458 Uliczka E (1929) Die tetraxone Schwmme Westindiens (auf Grund der Ergebnisse der reise Kkenthal-Hartmeyer). Zool Jahrb, Suppl 16: 35-62 van Soest RWM, Rtzler K (2002) Family Tetillidae Sollas, 1886. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges. Kluwer Academic/Plenum Publishers, New York. pp. 85-98 Wilson HV (1925) Siliceous and horny sponges collected by the US Fisheries Steamer Albatross during the Philippine Expedition, 1907-10. Bull US Nat Mus 100(2): 273-532

Porifera research: Biodiversity, innovation and sustainaBility - 2007

555

Field preservation and optimization of a DNA extraction method for Porifera

Adriana Salgado(1), Thomz Vieiralves(1), Flvia R.M. Lamaro(1), Leonardo L.M. Assumpo(1), Dbora Gomes(2), Lia Jascone(2), Ana Luiza Valado(2), Rodolpho M. Albano(3), Gisele Lbo-Hajdu(1*)
Departamento de Biologia Celular e Gentica/DBCG, Instituto de Biologia Roberto Alcantara Gomes/IBRAG, Universidade do Estado do Rio de Janeiro/ UERJ, Rua So Francisco Xavier, 524 PHLC, sala 205, Maracan, 20550013, Rio de Janeiro, RJ, Brazil (2) Curso de Cincias Biolgicas, Disciplina de Gentica Bsica, DBCG, IBRAG, UERJ (3) Departamento de Bioqumica/DBq, IBRAG, UERJ, Av. 28 de Setembro, 87 fundos, PAPC 4o andar, 20551-013, Universidade do Estado do Rio de Janeiro, Rio de Janeiro, RJ, Brazil. lobohadju@oi.com.br
(1)

Abstract: The small number of molecular studies on lower invertebrates may be due to a limited availability of fresh or properly preserved biological material. Specimens collected and preserved in different fixatives can influence the quality of the extracted DNA. Variables such as the type of fixative, time of storage, and extraction protocol are critical for obtaining DNA in sufficient quantity and of good quality. This work evaluates the efficiency of six different field fixatives and the most effective DNA extraction protocol for marine sponges. Sponges were collected and preserved in one of the following: 1) 96% ethanol, 2) 70% ethanol, 3) dry-ice, 4) air-dried, 5) lyses buffer with guanidine hydrochloride (LBWGH), or 6) silica gel. Genomic DNA was extracted by one of four different protocols: lyses buffer with proteinase K, cetyl trimethyl ammonium bromide (CTAB), guanidine hydrochloride or DNAzol. The quality of the DNA obtained was determined with scores of the DNA degradation level. Our results showed that high molecular weight DNA was seen with all six fixatives albeit with a great variation in DNA quality. Based on gel analysis, the most effective preservation methods, both in quality and quantity, were dry ice, silica gel and LBWGH. Regarding the DNA extraction procedures, CTAB, LBWGH and the DNAzol methods produced high quality genomic DNA. However, considering the cost-benefit ratio of the methods for the processing of a large number of samples, a short term preservation in combination with extraction with LBWGH is the best protocol among the techniques tested here. Keywords: field preservation, DNA extraction, method optimization, Porifera

Introduction
The constant need to unravel the phylogenetic relationships of various organisms and the popularization of molecular methods transformed museum collections in valuable sources of DNA. Biologists have been extracting DNA from specimens deposited in collections for decades and the protocols to use this material have been continuously improving (Arrighi et al. 1968, Pbo 1989, Post et al. 1993, Thomas 1994, Reiss et al. 1995, Dilon et al. 1996, Hammond et al. 1996, Shedlock et al. 1997, Kalmr et al. 2000, Berntson and France 2001, Rohland et al. 2004, Chakraborty et al. 2006). However, the integrity of the extracted DNA will vary according to the organism, preservation conditions, time of storage and DNA extraction method. Few reports comparing preservation methods have been published for marine invertebrates (France and Kocher 1996, Chase et al. 1998, Dawson et al. 1998, Berntson and France 2001, Crabbe 2003), and, very recently, one study has addressed members of Porifera (Ferrara et al. 2006). Overcoming the preservation step, the next decision is to choose a reliable DNA extraction method. However, as

observed for several other marine invertebrates, the quality of the DNA extracted from sponges is rarely suitable for PCR reactions. One possible explanation can be the presence of acidic polysaccharides in several marine sponges, which could inhibit PCR amplification (Demeke and Adams 1992). Furthermore, it was previously reported that the yield of DNA extraction varies considerably among taxa and should not be extrapolated from one organism to another (Dillon et al. 1996, Dawson et al. 1998, Mtambo et al. 2006). Several nucleic acid extraction procedures are currently available. They go from a a simple salting out method to the use of complex buffers with high salt, detergents and reagents which cleave disulfide bridges (Miller et al. 1988, Boom et al. 1990, Seutin et al. 1991, Hong et al. 1997, Dawson et al. 1998, Berntson and France 2001, Mtambo et al. 2006, Ferrara et al. 2006). The aim of this work was to assess simple protocols for the preservation and DNA extraction from marine sponges. For this, different protocols were compared to evaluate the quality and quantity of the extracted genomic DNA and its suitability for PCR amplification.

556

Material and methods Collection, preservation and storage

Specimens of Hymeniacidon heliophila Parker, 1910 and Paraleucilla magna Klautau, Monteiro and Borojevic, 2004 (Calcarea) were collected at Praia Vermelha beach, Rio de Janeiro (2300S, 4312W), while Amphimedon viridis Duchassaing and Michelloti, 1864 and Aplysina fulva (Nardo, 1834) were collected at Joo Fernandinho beach, Bzios, Rio de Janeiro (2244S, 4154W). After careful dissection to remove macroscopic symbionts and substrate debris, all individuals were divided in six fragments, each one preserved differently: 1) 96% ethanol, 2) 70% ethanol, 3) frozen in dryice (solid CO2), 4) air-dried, 5) lyses buffer with guanidine hydrochloride (LBWGH - 4 M guanidine hydrochloride, 50 mM Tris-HCl pH 8.0, 0.05 M EDTA, 0.5% sodium-Nlauroylsarcosine, 1% -mercaptoethanol), and 6) silica gel. Three replicates of all samples were preserved in each fixative for seven to sixty days at room temperature. The exception were the air-dried samples, which were kept after one day in the oven at 60oC and the fragments frozen in dryice, which were stored in -80oC ultra freezer. In order to compare the preservation protocols, DNA was extracted from these samples by the same method (LBWGH, see below).

Fig. 1: Degradation scores for analysis of DNA quality, modified from Amos and Hoelzel (1991). 1 = not degraded high weight DNA, 2 = high weight DNA with little degradation, 3 = high weight DNA with degradation, 4 = DNA with high degradation, 5 = completely degraded DNA, and 6 = no DNA.

DNA Extraction
Hymeniacidon heliophila was collected, divided in four fragments of same size and wet weight, and frozen in dry-ice immediately after removal from sea water. Afterwards, the samples were submitted to four methods of genomic DNA extraction (described in the appendix): - Method 1, Lyses buffer with proteinase K (LBWPK); - Method 2, Lyses buffer with cetyl trimethyl ammonium bromide (CTAB); - Method 3, Lyses buffer with guanidine hydrochloride (LBWGH); - Method 4, DNAzol.

PCR Amplification test

Two pairs of primers (Lobo-Hajdu et al. 2004): 18S Forward /5-TCATTTAGAGGAAGTAAAAGTCG-3, 5,8S Reverse /5-GCGTTCAAAGACTCGATGATTC-3, and 5,8S Forward /5-GAATCATCGAGTCTTTGAGC C-3, 28S Reverse /5-GTTAGTT TCTTTTCCTCCGCTT-3 were used to test the amplification of the two internal transcribed spacers (ITS-1 and ITS-2) of nuclear ribosomal RNA (rRNA). Each 30 l PCR amplification reaction mixture contained 10 ng of genomic DNA, reaction buffer (10 mM KCl, 20 mM Tris-HCl pH 8.8, 10 mM (NH4)2SO4, 0.1% Triton-X-100, 100 mg/ml gelatin), 3 mM MgSO4, 200 M dNTPs, 80 ng of each primer and 1 unit of DNA polymerase (Platinum Taq, Invitrogen, SP, Brazil). PCR amplification was carried out in a DNA thermal cycler (M.J. Research PTC-100). An initial denaturation step of 5 min at 96C was followed by 35 cycles of 30 s at 94C, 45 s at 52C and 1 min at 72C, with an additional final step of 5 min at 72C for final expansion. The amplified bands were separated by electrophoresis on 2% agarose gels in 0.5X TBE. The size of the amplified fragments was estimated by comparison with standard DNA ladders.

Determination of quality and quantity of DNA

The DNA concentration was estimated in 0.8% agarose gels run in TBE 0.5X (50 mM Tris Base, 50 mM boric acid, 1 mM EDTA) by comparison with solutions of known concentrations of lambda bacteriophage genomic DNA (10 ng/l, 20 ng/l, 50 ng/l, 100 ng/l), and visualized under UV light after staining with 0.6 g/ml ethidium bromide. To standardize the quantification, each sample applied on a gel had its volume adjusted according to the initial wet weight. In order to evaluate the quality of the genomic DNA, the scores given in Amos and Hoelzel (1991) were adopted, with some modifications. For each sample, a score from 0 to 5 was given, depending on the degree of DNA degradation. A diagrammatic representation of how those scores were correlated to DNA degradation when visualized in agarose gels is presented in Fig. 1.

Results and discussion

All of the six fixation procedures tested allowed for some extracted DNA. However, depending on the fixative used a great variation on DNA amount and quality was observed (Fig. 2). LBWGH, freezing and silica gel dried samples obtained the best scores (scores 1 and 2). Even though air-dried samples resulted in good quality DNA (score 3), the yield was lower than with the above mentioned methods. The worst quality results were obtained for 70% ethanol with a score

557

Fig. 2: DNA quality state determination and scores from samples of Hymeniacidon heliophila preserved in six different fixatives and extracted by the LBWGH method. Concentration markers are 10 ng and 50 ng lambda DNA. D = air dried, E70 = 70% ethanol, E96 = 96% ethanol, F = frozen in dry-ice, L = LBWGH solution and S = silica gel.

Fig. 3: Agarose gel (2%) electrophoresis showing ITS-1 PCR amplification from one H. heliophila individual preserved in six different fixatives. M = 100bp ladder molecular weight marker. D = air dried, E70 = 70% ethanol, E96 = 96% ethanol, F = frozen in dry-ice, L = LBWGH solution and S = silica gel.

of 5. Despite the variation in DNA quantity and quality, all extraction procedures produced DNA that rendered single PCR products for both ITS-1 and ITS-2. Figure 3 shows the 380 bp fragment resultant of ITS-1 PCR amplification from H. heliophila. To test long-term preservation, each sample, sorted by the fixation procedures used was submitted to the LBWGH DNA extraction method after seven, 15, 30 and 60 days of storage. The quality of the DNA extracted from the seven and 15 days samples is shown in Fig. 4. It can be clearly seen that after a week all yields decrease, which was reflected on the quality scores. Nevertheless, even after 60 days of storage at room temperature, enough DNA to be amplified could be extracted (data not shown). The best quality DNA was obtained preserving sponges in dry-ice followed by freezing at -80oC. LBWGH and silica gel were also good fixatives. The other tested fixatives yielded DNA with less quality according to the following order, from best to worst: air-dried, 96% ethanol and 70% ethanol for H. heliophila. Other tested sponges gave different results. For example, A. viridis showed excellent quality DNA for air-dried preserved samples. One possible justification for these discrepancies is the different composition of sponge tissues and the amount of polysaccharides they possess. Amphimedon viridis can be squeezed until almost all water is removed, which facilitates the denaturing of nucleases (DNase and RNAse). So, for A. viridis, samples preserved in a dry state, yielded DNA as good as those kept frozen or in LBWGH solution. Two other sponges tested, P. magna and A. fulva, were well preserved in LBWGH, frozen and in 96% ethanol. They differ from A. viridis and H. heliophila regarding the use of 70% ethanol, silica gel and air for fixation. Paraleucilla magna would allow high quality DNA extractions from these three fixatives which imply in quick removal of water, while A. fulva provided low quality, if any, genomic DNA. Again, these two sponges presented different consistencies, being

P. magna more friable and A. fulva more meaty (firm when pressed). These results demonstrated that LBWGH solution yielded reasonable DNA, both in quality and in quantity. This solution acts denaturing nucleases because it contains a chaotropic agent (guanidine hydrochloride), EDTA, which absorbs Ca++ and Mg++ essential ions for some nucleases, and Nlaurylsarcosine, a detergent responsible by the disintegration of cellular membranes, allowing the release of nuclear DNA. Although the LBWGH solution produced better results than 96% ethanol, it did not preserve the shape and integrity of the organism. Guanidine hydrochloride cannot be considered as a permanent fixative for biological material, but it is perfectly viable as a temporary transport solution of specimens for DNA assays from the field to the laboratory. If kept frozen (-20o C), samples in LBWGH can render DNA suitable for PCR amplification up to 10 years after being collected (data not shown). Additional problems related to shipping preserved biological specimens were introduced by new rules imposed by the International Air Transport Association (IATA). In a query launched at the Porifera Mailbase (see Archives of Porifera at http://www.jiscmail.ac.uk/lists/porifera.html) by Dr John Hooper, Queensland Museum, on January 24th 2005, a consensus list of recommended preservative methods for sponges was announced. Three methods for preserving and shipping overseas sponge samples for DNA studies were: 1) small fragments mummified in high analytical grade silica gel, 2) samples immersed in DMSO buffer (20% DMSO, 250 mM EDTA, NaCl to saturation, pH 8.0) (adapted from Seutin et al. 1991), and 3) freeze-dried sponges. More recently, the subscribers of the Porifera Mailbase once more inquired about sponge preservation for genetic work (Dr. Claire Goodwin, Ulster Museum, on May 25th 2006). From that debate additional suggestions were made: preserving small pieces in RNALater (Ambion), 95% ethanol, 70% ethanol (1:10 weight sponge to ethanol volume), isopropanol,

558

Fig. 4: DNA quality state determination and scores from samples of H. heliophila preserved in six different fixatives for seven and fifteen days. Concentration markers are 10, 20, 50 and 100 ng, respectively). D = air dried, E70 = 70% ethanol, E96 = 96% ethanol, F = frozen in dry-ice, L = LBWGH solution and S = silica gel.

lyses buffer with high concentration of a chaotropic agent, such as guanidine hydrochloride (GuHCl), and FTA Classic Cards (Whatman BioScience, Fast Technology for Analysis of nucleic acids) (Crabbe 2003). Considering altogether, for small to medium projects, it is viable to collect samples of sponges in LBWGH solution. Large marine faunistic surveys will demand a less expensive and more easily accessible fixative. In that case, 96% ethanol is recommended to preserve Porifera samples, which keep the morphology of the sample while yielding an acceptable amount of good quality DNA. One point to be highlighted is the importance of taking only a small sample when preserving in silica gel and LBWGH. The same probably stands true for some species preserved in 96% or 70% ethanol (Seutin et al. 1991, Reiss et al. 1995). Dessauer et al. (1996) suggested that samples should be cut in fragments not bigger than 1 mm3, although Dawson et al. (1998) had success with 0.2 cm3 finely chopped tissues. This feature is important because the essential step in alcohol and silica gel preservation is the fast elimination of water out of the tissues in order to limit hydrolytic cleavage of the nucleic acids. Thus, the sponge should be cleaned and the sea water drained just after collection and separated in two subsamples. A bigger piece, with visible representative anatomic features, should be placed in 96% ethanol. Even drained, the sponge will keep some sea water inside that will lower the ethanol concentration. After some days, the ethanol should be replaced. A second, smaller piece from the clean inner part of the sponge should be cut in small fragments and placed in LBWGH, silica gel, RNALater or FTA Cards. A comparison between the four extraction methods applied to fresh material is shown in Fig. 5. Notably, LBWGH and LBWPK extraction methods if combined (Fig. 5, LPK) yield genomic DNA with improved quality. DNAzol rendered genomic DNA with higher quality score than the other techniques. This kit, and others such as Trizol and RNALater, has guanidine hydrochloride (GuHCl) or

thiocyanate (GuSCN) on its composition (Boom et al. 1990) and therefore resembles the LBWGH method. The CTAB buffer works well for fresh (Fig. 5, CT) and ethanol preserved samples, although resulted in an smaller amount of DNA. The combination of CTAB with proteinase K reduces so much the amount of DNA obtained that it is not visible in agarose gels (Fig. 5, CTK). Finally, in Fig. 6, a comparison of preservation and extraction methods is shown. Sponge samples preserved and extracted in/with LBWGH buffer resulted in perfect DNA for PCR amplification. A proposal taken from this work is the use of the LBWGH buffer for the short-time fixation of sponge samples at room temperature, followed by the LBWGH plus proteinase K DNA extraction. Being a simple solution, like RNALater, the LBWGH should not be a problem to exchange loans by conventional mail. The use of screw capped 2 mL vials is recommended, always using three times more LBWGH buffer than sponge fragments. For long-time fixation in LBWGH buffer, samples should be kept at -20o.C.

Appendix
Method 1, Lyses Buffer with Proteinase K (LBWPK): Homogenization in 1:3 (weight:volume) solution of SET buffer (0.15 M NaCl, 0.05 M Tris/HCl pH 8.0, 10 mM EDTA, 0.4% SDS) plus 20 g/L proteinase K. The suspension was incubated at 55C for 1 hr and centrifuged at 3000 g/10 min. The supernatant was extracted once with an equal volume of phenol:chloroform:isoamyl alcohol (25:24:1) and twice with one volume of chloroform. Precipitate DNA from the homogenate by the addition of 2 volumes of ethanol plus 1/10 volume of 3 M sodium acetate pH 5.2, followed by centrifugation at 10000 g/10 min at 4C. The pellet was washed in 70% ethanol, air dried, dissolved in sterile water plus 20 g/ml RNAse A (GIBCO BRL), and incubated for 1 h at 37C.

559

Fig. 5: DNA quantification and scores from fresh collected samples showing the extraction products from two different individuals of H. heliophila with four different methods and combinations of them. Concentration markers are 10, 20, 50 and 100 ng, respectively). L = LBWGH solution, LPK = LBWGH plus proteinase K, Dz = DNAzol, CT = CTAB solution, CK = CTAB plus proteinase K, PK = proteinase K.

Method 2, Lyses Buffer with Cetyl Trimethyl Ammonium Bromide (CTAB): Homogenization of sponge fragments in 1:3 (weight:volume) solution of 2% CTAB in 100 mM Tris-HCl pH 8.0, 20 mM EDTA, 1.4 M NaCl plus 1 L mercaptoetanol and 10 g/L proteinase K. The suspension was incubated at 50C for at least 1 hr and centrifuged at 3000 g/10 min. The supernatant was extracted twice with one volume of chloroform: isoamyl alcohol (24:1). Precipitate DNA from the homogenate by the addition of 0.8 volumes of isopropanol plus 1/10 volume of 3 M sodium acetate pH 5.2, followed by centrifugation at 10000 g/10 min at 4C. The pellet was washed in 70% ethanol, air dried, dissolved in sterile water plus 20 g/ml RNAse A (GIBCO BRL), and incubated for 1 h at 37C. Method 3, Lyses Buffer with Guanidine Hydrochloride (LBWGH): Specimens were ground with a rod in a mortar with 1:5 (weight:volume) solution of 4 M guanidine hidrochloride, 50 mM Tris-HCl pH 8.0, 0.05 M EDTA, 0.5% sodium-Nlauroylsarcosine and 1% -mercaptoethanol. The suspension was incubated at 50C for 1 hr and centrifuged at 3000 g/10 min. The supernatant was extracted with an equal volume of phenol:chloroform:isoamyl alcohol (25:24:1) and nucleic acids were precipitated with 2 volumes of ethanol. The pellet was washed in 70% ethanol and air dried. The dried pellet was dissolved in sterile water plus 20 g/ml RNAse A (GIBCO BRL) and incubated at 37C for 2 h. Method 4, DNAzol: Homogenization of 25-50 mg fragments in 1 ml of DNAzol in a hand held homogenizer by applying as few strokes as possible. The homogenate was centrifuged at 3000 g/10 min. The supernatant was transferred to a fresh tube and the DNA precipitated by the addition of 0.5 ml of 100% ethanol per 1 ml of DNAzol used for the isolation. The DNA precipitate was removed by spooling with a pipette tip or by centrifugation at 10000 g/10 min at 4C. DNA pellet was washed twice with 75% ethanol and dissolved in water.

Fig. 6: DNA quantification of total genomic DNA extracted by the LBWGH method. Molecular weight marker are from top to bottom: 23100, 9400, 6600, 4400, 2300, 2000 bp. E96 = sponge specimens preserved at museum collections for long term in ethanol 96%, and L = sponge specimens preserved directly in LBWGH.

Acknowledgements
D. M. Braga de Mello is thanked for technical assistance. This work was carried out with financial assistance from Programa Procincia, Sub-reitoria de Ps-graduao e Pesquisa (SR2-UERJ), Fundao Carlos Chagas Filho de Amparo Pesquisa do Estado do Rio de

560

Janeiro (FAPERJ) and Conselho Nacional de Desenvolvimento Cientfico e Tecnolgico (CNPq).

References
Amos B, Hoezel R (1991) Long-term preservation of whale skin for DNA analysis. Rep Int Whal Comm Special Issue 13: 99-103 Arrighi FE, Bergendahl J, Mandel M (1968) Isolation and characterization of DNA from fixed cells and tissues. Exp Cell Res 50: 47-53 Berntson EA, France SC (2001) Generating DNA sequence information from museum collections of octocoral specimens (Phylum Cnidaria: Class Anthozoa). Bull Biol Soc Wash 10: 119129 Boom R, Sol CJA, Salimans MMM, Jansen CL, Wertheim-van Dillen PME, van der Noordaa J (1990) Rapid and simple method for purification of nucleic acids. J Clin Microbiol 28(3): 495-503 Chakraborty A, Sakai M, Iwatsuki Y (2006) Museum fish specimens and molecular taxonomy: A comparative study on DNA extraction protocols and preservation techniques. J Appl Ichthyol 22(2): 160166 Chase MR, Etter RJ, Rex MA, Quattro JM (1998) Extraction and amplification of mitochondrial DNA from formalin-fixed deep-sea mollusks. BioTechniques 24(2): 243-247 Chomczynski P, Mackey K, Drews R, Wilfinger W (1997). DNAzol: a reagent for the rapid isolation of genomic DNA. BioTechniques 22: 550-553 Crabbe MJ (2003) A novel method for the transport and analysis of genetic material from polyps and zooxanthellae of scleractinian corals. J Biochem Biophys Meth 57: 171-176 Damato ME, Corach D (1996). Genetic diversity of populations of the freshwater shrimp Macrobrachium borelli (Cardidea, Palaemonidae) evaluated by RAPD analysis. J Crust Biol 16: 650655 Dawson MN, Raskoff KA, Jacobs DK (1998). Field preservation of marine invertebrates tissue for DNA analyses. Mol Mar Biol Biotechnol 7(2): 145-152 Demeke T, Adams RP (1992) The effects of plant polysaccharides and buffer additives on PCR. BioTechniques 12: 332-334 Dessauer HC, Cole CJ, Hafner MS (1996) Collection and storage of tissues. In: Hillis DM, Moritz C, Mable BK (eds). Molecular systematics, 2nd ed. Sinauer Associates, Sunderland. pp. 29-47 Dillon N, Austin AD, Bartowsky E (1996) Comparison of preservation techniques for DNA extraction from hymenopterous insects. Insect Mol Biol 5(1): 21-24 Ferrara GB, Murgia B, Parodi AM, Valisano L, Cerrano C, Palmisano G, Bavestrello G, Sara M (2006) The assessment of DNA from

marine organisms via a modified salting-out protocol. Cell Mol Biol Lett 11(2): 155-160 France SC, Kocher TD (1996) DNA sequencing of formalin-fixed crustaceans from archival research collections. Mol Mar Biol Biotechnol 5(4): 304-313 Hammond JBW, Spanswick G, Mawn JA (1996) Extraction of DNA from preserved animal specimens for use in Randomly Amplified Polymorphic DNA analysis. Anal Biochem 240: 300-302 Hong Y-K, Sohn CH, Lee KW, Kim HG (1997) Nucleic acid extraction from seaweed tissues for polymerase chain reaction. J Mar Biotechnol 5: 95-99 Kalmr T, Bachrati CZ, Marcsik A, Rask I (2000) A simple and efficient method for PCR amplifiable DNA extraction from ancient bones. Nucleic Acids Res 28(12): e67 Lobo-Hajdu G, Guimares ACR, Salgado A, Lamaro FRM, Vieiralves T, Mansure JJ, Albano RM (2004) Intragenomic, intraand interspecific variation in the rDNA ITS of Porifera revealed by PCR-Single-Strand Conformation Polymorphism (PCR-SSCP). In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Boll Mus Ist Biol Univ Genova 68: 413-423 Miller SA, Dykes DD, Polesky HF (1988) A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res 16(3): 1215 Mtambo J, van Bortel W, Madder M, Roelants P, Backeljau T (2006) Comparison of preservation methods of Rhipicephalus appendiculatus (Acari: Ixodidae) for reliable DNA amplification by PCR. Exp Appl Acarol 38: 189-199 Pbo S (1989) Ancient DNA: extraction, characterization, molecular cloning, and enzymatic amplification. Proc Natl Acad Sci USA 86(6): 1939-1943 Post RJ, Flook PK, Millest AL (1993) Methods for the preservation of insects for DNA studies. Biochem Syst Ecol 21(1): 85-92. Reiss RA, Schwert DP, Ashworth AC (1995) Field preservation of Coleoptera for molecular genetic analyses. Environm Entomol 24(3): 716-719 Rohland N, Siedel H, Hofreiter M (2004) Nondestructive DNA extraction method for mitochondrial DNA analyses of museum specimens. BioTechniques 36(5): 814-816, 818-821 Seutin G, White BN, Boag PT (1991) Preservation of avian blood and tissue samples for DNA analyses. Can J Zool 69: 82-90 Shedlock AM, Haygood MG, Pietsch TW, Bentzen P (1997) Enhanced DNA extraction and PCR amplification of mitochondrial genes from formalin-fixed museum specimens. BioTechniques 22(3): 394-396, 398, 400 Thomas RH (1994) Analysis of DNA from natural history museum collections. EXS 69: 311-321

Porifera research: Biodiversity, innovation and sustainaBility - 2007

561

Morphological and molecular analyses of microorganisms in Caribbean reef adult sponges and in corresponding reproductive material
Susanne Schmitt(1), Markus Wehrl(1), Niels Lindquist(2), Jeremy B. Weisz(2, ), Ute Hentschel(1*)
University of Wuerzburg, Research Center for Infectious Diseases, Roentgenring 11, D-97070 Wuerzburg, Germany. susanne.schmitt@mail.uni-wuerzburg.de, markus.wehrl@mail.uni-wuerzburg.de, ute.hentschel@mail.uni-wuerzburg.de (2) University of North Carolina at Chapel Hill, Institute of Marine Sciences, 3431 Arendell Street, Morehead City, NC 28557, USA. nielsl@email.unc.edu
(1)

Abstract: Caribbean reef sponges were surveyed for the presence of microorganisms in the mesohyl tissue of adult sponges and the respective reproductive material (embryos, larvae). A clear correlation was found in that high microbial abundance (HMA) sponges always contained microorganisms in their reproductive stages. In contrast, low microbial abundance (LMA) sponges did not contain microorganisms in their reproductive stages. Based on these data, Ircinia felix Duchassaing and Michelotti, 1864 was chosen as a model organism for the molecular analysis of microorganisms within the adult sponge and its larvae and juveniles. Denaturing gradient gel electrophoresis (DGGE) of eubacterial 16S rDNA sequences revealed similar banding patterns for the adult individual and its reproductive stages. However, resolution of the DGGE gel was found to be limited. Selected DGGE bands (n=21) were excised and sequenced. The majority of sequences were most similar to sequences obtained from other HMA sponges indicating the presence of members of the previously identified, sponge-specific community in the adult sponge and its reproductive stages. Keywords: larvae, microbial diversity, Porifera, reproductive stages, sponge

Introduction
Sponges are filter-feeders that pump large volumes of seawater through their aquiferous system and take up food particles like organic particles and microorganisms by phagocytosis (Brusca and Brusca 1990). In addition to these food bacteria, many so called bacteriosponges (Reiswig 1981) can permanently harbor large amounts of extracellular microorganisms in their mesohyl that make up 40-60% of the sponge biomass and exceed seawater concentrations by 2-4 orders of magnitude (Friedrich et al. 2001, Webster and Hill 2001). These sponge-associated microorganisms are morphologically diverse and often show unusual membrane structures like additional sheaths, slime layers or putative nuclear membranes (Vacelet and Donadey 1977, Wilkinson 1978, Fuerst et al. 1998, Friedrich et al. 1999, Fieseler et al. 2004). 16S rRNA gene based studies revealed phylogenetically complex, sponge-specific microbial consortia that are present in different sponges and that are remarkably different from seawater bacterioplankton, both in terms of concentration and diversity (Hentschel et al. 2003, 2006, Taylor et al. 2007). In total, representatives of the eubacterial phyla Proteobacteria
() J.B. Weisz present address: Old Dominion University Norfolk, Department of Biological Science, VA 23529, USA. jweisz@odu.edu

(Alpha-, Gamma- and Deltaproteobacteria), Acidobacteria, Actinobacteria, Bacteroidetes, Chloroflexi, Cyanobacteria, Nitrospira and Gemmatimonadetes (Friedrich et al. 1999, Hentschel et al. 2002, Taylor et al. 2005, Schmitt et al. 2007) and of the candidate phylum Poribacteria (Fieseler et al. 2004) as well as of the archaeal phylum Crenarchaeota (Preston et al. 1996, Margot et al. 2002) were detected as specific members of the sponge microbiota. Moreover, it could be shown that the sponge-specific microbial consortia are stable over time and space (Hentschel et al. 2002). The microbial community profiles of Cymbastela concentrica von Lendenfeld, 1887 were stable over large distances in temperate waters but differed between temperate and tropical seas (Taylor et al. 2005). The microbial consortia of verongid sponges were also stable after experimental perturbation (Friedrich et al. 2001, Thoms et al. 2003); however, microbial community variability was observed after copper exposure in Rhopaloeides odorabile Thompson, Murphy, Bergquist and Evans, 1987 (Webster et al. 2001). Vertical transmission has been proposed as a potential mechanism for the establishment and maintenance of specific sponge-microbe-associations. Microorganisms have already been detected by electron microscopy in oocytes of several oviparous sponges (e.g. Aplysina spp. (Gallissian and Vacelet 1976), Stelletta grubii Schmidt, 1862 (Sciscioli et al. 1991), Geodia cydonium Jameson, 1811 (Sciscioli et al. 1994), Chondrilla spp. sponges (Maldonado et al. 2005)) and

562

in embryos and larvae of several viviparous sponges (e.g. Spongia spp. and Hippospongia spp. (Kaye 1991, Kaye and Reiswig 1991), Chondrosia reniformis Nardo, 1847 (Lvi and Lvi 1976), Astrosclera willeyana Lister 1900 (Wrheide 1998)). Usher et al. (2001) and Ereskovsky et al. (2005) reported in detail on the incorporation and transmission of Cyanobacteria in Chondrilla australiensis Carter, 1873 and of a spiral bacterium in Halisarca dujardini Johnston, 1842, respectively. Furthermore, Enticknap et al. (2006) succeeded in cultivating several alphaproteobacterial strains from larvae of the sponge Mycale laxissima Duchassaing and Michelotti, 1864. These bacteria were closely related to each other and also to the strain NW001 isolated from the sponge R. odorabile (Webster and Hill 2001). To obtain a better understanding of the establishment and maintenance of this unique association between sponges and complex microbial consortia we performed a general electronmicroscopical survey for the presence of microorganisms in adult sponges as well as in the respective reproductive stages. In total, ten Caribbean species representing five orders and two different modes of reproduction (ovipary, vivipary) were included. Based on these results, Ircinia felix Duchassaing and Michelotti, 1864 was chosen as a model system for the detailed molecular study of vertical transmission. I. felix is a viviparous species with internal fertilization. Embryos are brooded in the sponge mesohyl and free swimming larvae spend a few hours in the water column before they settle on a suitable substrate and metamorphose into juveniles. We performed settlement experiments on reefs offshore Florida and applied DGGE analysis and subsequent sequencing of excised bands to characterize and compare the associated microbiota of three different developmental stages (adult, larvae, and juveniles).

Denaturing Gradient Gel Electrophoresis (DGGE)

Larvae settlement experiments were performed with the sponge I. felix as described by Schmitt et al. (2007). Briefly, larvae released by one adult individual were caught using the methodology of Lindquist et al. (1997), transferred into sealed plastic containers (~ 75 ml volume) and returned to the reef where they were cable tied to racks at 9 m depth. After settlement, pieces of the Nylon with settled juvenile sponges as well as pieces of Nylon without sponge tissue (control) were collected. DGGE was performed with larvae, juveniles and the respective adult sample. DNA of the adult sponge was extracted using the Fast DNA Spin Kit for Soil (Q-Biogene, Heidelberg, Germany) whereas DNA of three pooled larvae, three pooled juveniles and the control was extracted by heating (100C) the samples for 10 min. Following PCR amplification with universal, eubacterial 16S rDNA-targeted primers 341f with GC clamp and 907r (Muyzer et al. 1998), two independent PCR reactions of each sample (adult, larvae, juveniles, control) were run on a 10% (w/v) polyacrylamide gel in 1x TAE using a 090% denaturing gradient; 100% denaturant corresponded to 7 M urea and 40% (v/v) formamide. Electrophoresis was performed for 6 h at 150 V and 60C. Gels were stained for 30 min in SYBR Gold (Molecular Probes) and scanned on a Typhoon 8600 scanner (Amersham Biosciences). DGGE banding pattern similarities were determined by cluster analysis using the software Quantity One (Bio-Rad, Mnchen, Germany). Selected bands were excised with an EtOH sterilized scalpel and incubated in 25l H2Odd overnight at 4C. 4 l of eluted DNA was subsequently used for reamplification with primers 341f and 907r. PCR products were ligated into the pGEM-T-easy vector (Promega) and transformed by electroporation into competent E. coli XL 1-Blue cells. Plasmid DNA of up to three different clones per excised band was isolated by standard miniprep procedures and the correct insert size was verified by using agarose gel electrophoresis following restriction digestion. Sequencing was performed on an ABI 377XL automated sequencer (Applied Biosystems). 16S rRNA gene sequences were deposited in the EMBL/GenBank/DDBJ database under accession numbers DQ661773-DQ661787 and EU095956EU095961.

Materials and methods Transmission Electron Microscopy (TEM)

Adult and reproductive material of ten sponge species was collected by SCUBA diving offshore off Key Largo, Florida in June 2002 and June and August 2004 using the NOAAs National Undersea Research Center (NURC) facilities and vessels (Table 1). The adult sponge samples were cut into small pieces of about 1mm3. All samples (adult, larvae) were preserved in 2.5% glutaraldehyde/H2Odd and washed five times in cacodylate buffer (50 mM, pH 7.2), fixed in 2% osmium tetroxide for 90 min, washed again five times in H2Odd and incubated overnight in 0.5% uranyl acetate. After dehydration in an ethanol series (30, 50, 70, 90, 96, and three times in 100% for 30 min, respectively), samples were incubated three times for 30 min in 1x propylene oxide, overnight in 1:1 (v/v) propylene oxide/Epon 812 (Serva), two times for 2 h in Epon 812 and finally embedded in Epon 812 for 48 hours at 60C. The samples were sectioned with an ultramicrotom (OM U3, C. Reichert, Austria) and pieces of three different individuals of each sponge developmental stage (except for A. coralliphagum embryos where only one sample was available) were examined by transmission electron microscopy (Zeiss EM 10, Zeiss, Germany).

Results Transmission Electron Microscopy

High microbial abundance sponges In adult samples of Agelas wiedenmayeri Alcolado, 1984, Aka coralliphagum Ruetzler, 1971, Ectyoplasia ferox Duchassaing and Michelotti, 1864, I. felix and Smenospongia aurea Hyatt, 1875 large numbers of extracellular microorganisms were scattered throughout the sponge mesohyl (Table 1, Fig. 1A, C, E, G, I). These species are therefore regarded as high microbial abundance (HMA) sponges. The microorganisms showed a high variety of morphotypes, such as rods, cocci and other, irregular forms. Many microorganisms possessed additional membrane structures similar to those that were described previously from other HMA sponges (Vacelet and Donadey 1977, Wilkinson

563

Table 1: Detection of microorganisms in adult sponges and reproductive stages using electron microscopy. Detection of microorganisms by TEM Species High microbial abundance sponges Agelas wiedenmayeri Alcolado, 1984 Aka coralliphagum Ruetzler, 1971 Ectyoplasia ferox Duchassaing and Michelotti, 1864 Ircinia felix Duchassaing and Michelotti, 1864 Smenospongia aurea Hyatt, 1875 Low microbial abundance sponges Callyspongia vaginalis Lamarck, 1814 Mycale laxissima Duchassaing and Michelotti, 1864 Niphates digitalis Lamarck, 1814 Tedania ignis Duchassaing and Michelotti, 1864 Ulosa ruetzleri Wiedenmayer, 1977
1

Order Agelasida Haplosclerida Poecilosclerida Dictyoceratida Verongida Haplosclerida Poecilosclerida Haplosclerida Poecilosclerida Poecilosclerida

Adult + + + + + (+)1

Reproductive stages + + + + +

Low microbial abundance and diversity in adult mesohyl

1978, Friedrich et al. 1999). Morphotype C is characterized by several additional sheaths, type D by a copious, irregular slime layer, and type E by a putative nuclear membrane. Cyanobacteria could be identified by their typical thylacoid membranes and were particularly dominant in the I. felix mesohyl (Fig. 1G). Some loosely scattered sponge cells were also present in the mesohyl. Most of these cells were amoeboid-like and contained large nuclei and often vesicles and phagosomes showing their phagocytotic activity. A layer of pinacocytes and/or choanocytes always separated the mesohyl of these sponges from seawater. Aka coralliphagum, I. felix and S. aurea have a viviparous mode of reproduction and release free swimming parenchymella-type larvae into the water column. Many microorganisms were predominantly located in the central region of the larvae (Fig. 1D, H, J). These morphologically diverse microorganisms were extracellular and similar in shape to the microorganisms present in the respective adult tissues including the morphotypes C, D and E. Few amoeboidlike sponge cells, that contain large amounts of lipids, were also present in the center of the larvae. Agelas wiedenmayeri and E. ferox have an oviparous mode of reproduction. They release oocytes or zygotes, which are embedded in a gelatinous sheath. These early reproductive stages were densely filled with lipids and electron-dense vesicles (Fig. 1B, F). Microorganisms that resembled the adult microbial community were predominantly found in the outer regions of the reproductive stages of A. wiedenmayeri and E. ferox. (Fig. 1B, F). Low microbial abundance sponges EM inspection of Callyspongia vaginalis Lamarck, 1814, M. laxissima, Niphates digitalis Lamarck, 1814, Tedania ignis Duchassaing and Michelotti, 1864 and Ulosa ruetzleri Wiedenmayer, 1977 adult samples revealed a low abundance and diversity of microorganisms (Table 1, Fig. 2C) or the complete absence in the mesohyl matrix (Table 1, Fig. 2A, E, G, I) and are therefore classified as low microbial abundance (LMA) sponges. The mesohyl contained few sponge cells

that were embedded in a voluminous extracellular matrix. All investigated species are viviparous. The larvae contained many morphological structures that could not be identified unambiguously. Noticable are high numbers and sometimes very large vesicles and lipids. However, no microorganisms could be detected in any of these larvae in this study (Fig. 2B, D, F, H, J).

Denaturing Gradient Gel Electrophoresis (DGGE)

Figure 3A represents the bacterial profiles of I. felix adult, larvae and juvenile as well as the control (piece of Nylon without sponge tissue). The DGGE banding patterns of each of two adult, larvae, and juvenile PCR reactions differed in only one, four, and two band positions, respectively, indicating that a PCR bias is negligible. The number of bands in I. felix adult was higher than in the larvae (adult n =20.5; larvae n=16), but the DGGE banding patterns appeared highly similar. Overall, adult and larvae samples shared more than 70% of all bands (Fig. 3B). The juvenile sample differed from the adult and larvae samples in that it had generally less bands (n=13) and shared only 53% of all bands with adult and larvae (Fig. 3B). The cluster analysis placed the juvenile sample next to the control (number of bands: n=17), albeit with only 54% similarity (Fig. 3B).

16S rDNA sequence analysis

Twenty four bands were excised from the I. felix DGGE gel (Fig. 3A). After removal of 5 sequences as chimaeras (sequences of bands 4, 14, 15, 16, 21), a total of 21 16S rRNA gene sequences were obtained: 9 from adult, 3 from larvae, 8 from juvenile, and 1 additional sequence from the control (Table 2). Three clones of DGGE band 18 revealed different sequences whereas two clones of DGGE bands 1 and 17 each revealed identical sequences. The overall diversity was high with representatives of four different bacterial phyla (Acidobacteria, Chloroflexi, Gemmatimonadetes, and Proteobacteria (Alpha-, Gamma-, Deltaproteobacteria)). In the adult sample, all ten sequences

564

Fig. 1: Transmission electron microscopy of the HMA sponges A. wiedenmayeri adult (A) and embryo (B), A. coralliphagum adult (C) and larva (D), E. ferox adult (E) and embryo (F), I. felix adult (G) and larva (H) and S. aurea adult (I) and larvae (J). Lines indicate microorganisms. Cy: cyanobacteria, ECM: extracellular matrix, L: lipids, MO: microorganisms, Ph: phagosome, SC: sponge cell. Scale bar: 1 m (A, C, D, G, H, I, J), 2 m (B, E, F).

were most similar to sequences derived from other sponges: Aplysina aerophoba Schmidt, 1862, Aplysina cavernicola Vacelet, 1959, Theonella swinhoei Gray, 1868, Agelas dilatata Duchassaing and Michelotti, 1864 and Plakortis sp. In the larvae, two sequences were most similar to a 16S rRNA gene sequence from A. cavernicola, whereas clone B13-1 was related to a seawater clone. The juvenile sample contained four 16S rRNA gene sequences most similar to sequences derived from the sponges A. aerophoba and A. dilatata, two sequences most similar to Alcanivorax sp., and two sequences most similar to a cold seep and a hot spring clone, respectively. The 16S rRNA gene sequence obtained from the control was related to a marine Pseudoalteromonas sp. sequence (Table 2).

Discussion
The EM survey for the presence of microorganisms in the sponge mesohyl yielded two different groups of sponges. One group contained large numbers of morphologically diverse microorganisms whereas the mesohyl of the second group was almost devoid of microorganisms. These data expand early observations on patterns of microbial abundances in sponges by Vacelet (1975) and Vacelet and Donadey (1977). Whenever microorganisms were present in the adult sponge, microorganisms were also contained in the respective reproductive stages (Table 1, Fig. 1). Whenever microorganisms were present in low numbers or absent in the adult sample, microorganisms were also missing in the reproductive stages (Table 1, Fig. 2). This correlation suggests that HMA sponges transfer microorganisms vertically through their reproductive stages. Morphotypes C, D and E that were found to be abundant and consistently associated with other sponges (Vacelet and Donadey 1977, Wilkinson 1978, Friedrich et al. 1999) could also frequently be detected in adult mesohyl

565

Fig. 2: Transmission electron microscopy of the LMA sponges C. vaginalis adult (A) and larva (B), M. laxissima adult (C) and larva (D), N. digitalis adult (E) and larva (F), T. ignis adult (G) and larva (H) and U. ruetzleri adult (I) and larva (J). ECM: extracellular matrix, L: lipids, SC: sponge cell. Scale bar: 3 m (J), 4 m (F), 5 m (A, B, C, D, E, G, H), 8 m (I).

and reproductive material of the sponges of this study. Moreover, these types were also present in juvenile sponges of I. felix (data not shown, Schmitt et al. 2007). This indicates that several similar microbial morphotypes are vertically transmitted in different sponges. Apparently, vertical transmission is common and widespread among HMA sponges. The mode of reproduction seems not to be a determining factor for vertical transmission as both oviparous (A. wiedenmayeri, E. ferox) and viviparous (A. coralliphagum, I. felix, S. aurea) species belong to the HMA sponge group (Table 1). This is consistent with previous electron microscopy studies that also documented the presence of microorganisms in oocytes of oviparous species (Gallissian and Vacelet 1976, Usher et al. 2001) and in oocytes and embryos of viviparous species (Kaye 1991, Ereskovsky et al. 2005). This further supports the general character of the microbial transfer through larvae in sponges. Based on these microscopic results the HMA sponge I. felix was chosen for a molecular comparison of the bacterial profile of the adult sponge and its developmental stages (larvae, juveniles) using DGGE. Ideally, DGGE bands represent single 16S rDNA sequence fragments that are separated by their GC content on an increasing denaturing gradient gel. The number of bands and the banding pattern correspond to the microbial numbers and diversity of a certain sample. In previous studies DGGE was found to be useful to describe the total microbial profile of sponges as well as the profile of specific microbial groups (Diaz et al. 2004, Taylor et al. 2005, Wehrl et al. 2007). In this study, some bands represented single 16S rDNA sequence fragments (e.g. bands 1 and 17 each revealed two identical sequences) whereas other bands represented more than one sequence (e.g. band 18 revealed three different sequences). Therefore, the total microbial diversity in each sponge developmental stage is probably higher than indicated by the number of bands per sample. The banding patterns of the I. felix adult sponge and its larvae and juveniles appeared

566

Fig. 3: A. 16S rDNA-DGGE gel of I. felix adult, larvae and juvenile samples as well as a Nylon-control sample. Two independent PCR reactions were run for each sample. Arrows mark excised bands. B. Cluster analysis of the DGGE gel showing percentage similarity of banding patterns.

Table 2: 16S rDNA sequence analysis of bands excised from the DGGE gel of Ircinia felix. Clone B1-1 B2-1 B3-1 B5-1 B6-1 B7-2 B8-1 B9-1 B10-1 B11-1 B12-1 B13-1 B17-1 B18-1 B18-3 B18-4 B19-2 B20-1 B22-1 B23-1 B24-1 Closest sequence match in GenBank sponge clone PAUC43f (AF186415) sponge DGGE Band 6 (AY180081) sponge clone TK19 (AJ347028) sponge clone TK13 (AJ347034) sponge clone AD040 (EF076132) sponge clone PK035 (EF076097) sponge clone PK035 (EF076097) sponge clone TK16 (AJ347035) sponge clone AD015 (EF076136) sponge DGGE Band 6 (AY180081) sponge DGGE Band 6 (AY180081) seawater clone (AY592226) sponge clone TK34 (AJ347030) cold seep clone (AB015247) sponge clone TK97 (AJ347054) sponge clone TK34 (AJ347030) marine Alcanivorax sp. (AY726812) hot spring clone pItb-vmat-60 (AB294961) sponge clone AD015 (EF076136) Alcanivorax sp. Mho1 (AB053124) Pseudoalteromonas sp. (AM111085) Similarity (%) 99 97 96 98 96 95 99 97 93 97 97 96 98 94 96 98 89 95 92 99 96 Length (bp) 585/589 487/499 567/589 577/588 566/589 554/579 582/587 548/564 524/561 487/499 486/499 569/589 548/559 530/560 529/547 551/559 502/561 567/594 522/562 586/587 472/489 Taxonomic affiliation Gemmatimonadetes Acidobacteria Gemmatimonadetes Deltaproteobacteria Deltaproteobacteria Gammaproteobacteria Gammaproteobacteria Chloroflexi Alphaproteobacteria Acidobacteria Acidobacteria Acidobacteria Alphaproteobacteria Alphaproteobacteria Alphaproteobacteria Alphaproteobacteria Alphaproteobacteria Gammaproteobacteria Alphaproteobacteria Gammaproteobacteria Gammaproteobacteria

control

juvenile

larvae

adult

567

similar (Fig. 3A, B). However, sequencing of excised bands that showed the same migration distance revealed only once the same sequence (bands 2 and 12). Overall, there appears to be little overlap among the sequences obtained from the adult individual, its larvae and the juvenile sponges. This might also be the result of a lack of resolution of the DGGE gel. The total phylogenetic diversity of I. felix is high and encompasses at least four bacterial phyla. Interestingly, most of the sequences (15 out of 20) obtained from the three sponge developmental stages show highest homology to sequences derived from other HMA sponges whereas the sequence from the control is related to the seawater bacterium Pseudoalteromonas sp. (Table 2). Apparently, the sponge-specific microbial consortium is present in the I. felix adult sponge as well as in its reproductive stages although the phylogenetic diversity seems reduced in the latter one. However, this might be due to the smaller number of excised and sequenced bands in larvae and juvenile samples compared to the adult sponge. Similarly, the presence of the spongespecific microbial community was previously documented in embryos of the HMA sponge Corticium sp. (Sharp et al. 2007). Furthermore, three selected phylotypes were consistently found in adult sponges and throughout the embryonic development indicating vertical transmission of these microbes. In a recent study, a large set of sequences including 15 sequences of this study were used to compare the microbial diversity of I. felix adult sponges and reproductive stages (larvae and juveniles) (Schmitt et al. 2007). Phylogenetic tree construction revealed vertical transmission clusters (IF-clusters) that contained sequences of both adult sponges and reproductive material. Therefore, this study in conjunction with the larger sequence dataset (Schmitt et al. 2007) clearly showed an overlap among the microbial communities of I. felix adult sponges and reproductive stages suggesting vertical transmission of the sponge specific microbial community in I. felix. In summary, the TEM survey revealed that the Caribbean sponges A. wiedenmayeri, A. coralliphagum, E. ferox, I. felix and S. aurea are associated with large amounts of microorganisms and that these microorganisms are most likely transferred vertically via the sponge reproductive stages. Other sponges that coexist in the same habitat (C. vaginalis, M. laxissima, N. digitalis, T. ignis and U. ruetzleri) contain few or no microorganisms in the adult mesohyl and the corresponding larvae. DGGE sequence analysis of adult, larvae and juvenile samples of I. felix revealed that representatives of the previously described sponge specific microbial consortium (Hentschel et al. 2002) are present in I. felix and its reproductive stages. Vertical transmission might be important to establish and maintain the phylogenetically complex yet highly sponge-specific microbial community in many other marine HMA sponges.

suggestions on the manuscript. This research was supported by Deutsche Forschungsgemeinschaft grant HE3299/1-1 and 1-2 to U.H.

References
Brusca RC, Brusca GJ (1990) Invertebrates. Sinauer, Sunderland. pp. 181-210 Diaz CM, Akob D, Cary CS (2004) Denaturing gradient gel electrophoresis of nitrifying microbes associated with tropical sponges. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Boll Mus Ist Biol Univ Genova 68: 279-289 Enticknap JJ, Kelly M, Peraud O, Hill RT (2006) Characterization of a culturable alphaproteobacterial symbiont common to many marine sponges and evidence for vertical transmission via sponge larvae. Appl Environ Microb 72: 3724-3732 Ereskovsky AV, Gonobobleva E, Vishnyakov A (2005) Morphological evidence for vertical transmission of symbiotic bacteria in the viviparous sponge Halisarca dujardini Johnston (Porifera, Demospongiae, Halisarcida). Mar Biol 146: 869-875 Fieseler L, Horn M, Wagner M, Hentschel U (2004) Discovery of the novel candidate phylum Poribacteria in marine sponges. Appl Environ Microb 70: 3724-3732 Friedrich AB, Fischer I, Proksch P, Hacker J, Hentschel U (2001) Temporal variation of the microbial community associated with the Mediterranean sponge Aplysina aerophoba. FEMS Microbiol Ecol 38: 105-113 Friedrich AB, Merkert H, Fendert T, Hacker J, Proksch P, Hentschel U (1999) Microbial diversity in the marine sponge Aplysina cavernicola (formerly Verongia cavernicola) analyzed by fluorescence in situ hybridisation (FISH). Mar Biol 134: 461-470 Fuerst JA, Webb RI, Garson MJ, Hardy L, Reiswig HM (1998) Membrane-bounded nucleoids in microbial symbionts of marine sponges. FEMS Microbiol Lett 166: 29-34 Gallissian MF, Vacelet J (1976) Ultrastructure de quelques stades de l`ovognse des spongiaires du genre Verongia (Dictyoceratida). Ann Sci Nat Zool Biol Anim 18: 381-404 Hentschel U, Fieseler L, Wehrl M, Gernert C, Steinert M, Horn M, Hacker J (2003) Microbial diversity of marine sponges. In: Mller WEG (ed). Molecular marine biology of sponges. Springer-Verlag, Heidelberg. pp. 59-88 Hentschel U, Hopke J, Horn M, Friedrich AB, Wagner M, Hacker J, Moore BS (2002) Molecular evidence for a uniform microbial community in sponges from different oceans. Appl Environ Microb 68: 4431-4440 Hentschel U, Usher KM, Taylor MW (2006) Marine sponges as microbial fermenters. FEMS Microbiol Ecol 55: 167-177 Kaye H (1991) Sexual reproduction in four Caribbean commercial sponges. II. Oogenesis and transfer of bacterial symbionts. Invertebr Reprod Dev 19: 13-24 Kaye HR, Reiswig HM (1991) Sexual reproduction in four Caribbean commercial sponges III. Larval behaviour, settlement and metamorphosis. Invertebr Reprod Dev 19: 25-35 Lvi C, Lvi P (1976) Embryognse de Chondrosia reniformis (Nardo), dmosponge ovipare, et transmission des bactries symbiotiques. Ann Sci Nat Zool 18: 367-380 Lindquist N, Bolser R, Laing K (1997) Timing of larval release by two Caribbean demosponges. Mar Ecol Progr Ser 155: 309-313

Acknowledgments
We gratefully acknowledge the staff of the University of North Carolina at Wilmingtons NOAA National Undersea Research Center at Key Largo, FL for their exceptional assistance during the field work and we thank Martin Meinhold, Hilde Angermeier and Roswitha Schiller (University of Wuerzburg, Germany) for interesting discussions and three anonymous reviewers for helpful

568

Maldonado M, Cortadellas N, Trillas MI, Rtzler K (2005) Endosymbiotic yeast maternally transmitted in a marine sponge. Biol Bull 209: 94-106 Margot H, Acebal C, Toril E, Amils R, Fernandez Puentes J (2002) Consistent association of crenarchaeal archaea with sponges of the genus Axinella. Mar Biol 140: 739-745 Muyzer G, Brinkhoff T, Nbel U, Santegoeds C, Schfer H, Wawer C (1998) Denaturing gradient gel electrophoresis (DGGE) in microbial ecology. In: Akkermans ADL, Elsas JD van, Bruijn FJ de (eds). Molecular microbial ecology, manual 3.4.4. Kluwer, Dordrecht. pp. 1-27 Preston CM, Wu KY, Molinski TF, DeLong EF (1996) A psychrophilic crenarchaeon inhabits a marine sponge: Cenarchaeum symbiosum gen. nov., sp. nov. Proc Natl Acad Sci USA 93: 6241-6246 Reiswig HM (1981) Partial carbon and energy budgets of the bacteriosponge Verongia fistularis (Porifera: Demospongiae) in Barbados. PSZN Mar Ecol 2: 273-293 Sciscioli M, Lepore E, Gherardi M, Scalera-Liaci L (1994) Transfer of symbiotic bacteria in the mature oocyte of Geodia sydonium (Porifera, Demospongiae): an ultrastructural study. Cah Biol Mar 35: 471-478 Sciscioli M, Scalera-Liaci L, Lepore E, Gherardi M, Simpson TL (1991) Ultrastructural study of the mature egg of the marine sponge Stelletta grubii (Porifera, Demospongiae). Mol Reprod Dev 28: 346-350 Schmitt S, Weisz JB, Lindquist N, Hentschel U (2007) Vertical transmission of a phylogenetically complex microbial consortium in the viviparous sponge Ircinia felix. Appl Environ Microb 73: 2067-2078 Sharp KH, Eam B, Faulkner DJ, Haygood MG (2007) Vertical transmission of diverse microbes in the tropical sponge Corticium sp. Appl Environ Microbiol 73: 622-629 Taylor MW, Radax R, Steger D, Wagner M (2007) Sponge-associated microorganisms: evolution, ecology, and biotechnological potential. Microbiol Mol Biol Rev 71: 295-347

Taylor MW, Schupp PJ, de Nys R, Kjelleberg S, Steinberg PD (2005) Biogeography of bacteria associated with the marine sponge Cymbastela concentrica. Environ Microbiol 7: 419-433 Thoms C, Horn M, Wagner M, Hentschel U, Proksch P (2003) Monitoring microbial diversity and natural products profiles of the sponge Aplysina cavernicola following transplantation. Mar Biol 142: 685-692 Usher KM, Kuo J, Fromont J, Sutton DC (2001) Vertical transmission of cyanobacterial symbionts in the marine sponge Chondrilla australiensis (Demospongiae). Hydrobiologia 461: 15-23 Vacelet J (1975) tude en microscopie lectronique de lassociation entre bactries et spongiaires du genre Verongia (Dictyoceratida). J Microsc Biol Cell 23: 271- 288 Vacelet J, Donadey C (1977) Electron microscope study of the association between some sponges and bacteria. J Exp Mar Biol Ecol 30: 301-314 Webster N, Hill RT (2001) The culturable microbial community of the Great Barrier Reef sponge Rhopaloeides odorabile is dominated by an -proteobacterium. Mar Biol 138: 843-851 Webster NS, Webb RI, Ridd MJ, Hill RT, Negri AP (2001) The effects of copper on the microbial community of a coral reef sponge. Environ Microbiol 31: 19-31 Wehrl M, Steinert M, and Hentschel U (2007) Bacterial uptake by the marine sponge Aplysina aerophoba. Microb Ecol 53: 355-365 Wilkinson CR (1978) Microbial associations in sponges. III. Ultrastructure of the in situ associations in coral reef sponges. Mar Biol 49: 169-176 Wrheide G (1998) The reef cave dwelling ultraconservative coralline demosponge Astrosclera willeyana Lister, 1900 from the Indo-Pacific - micromorphology, ultrastructure, biocalcification, isotope record, taxonomy, biogeography, phylogeny. Facies 38: 1-88

Porifera research: Biodiversity, innovation and sustainaBility - 2007

569

Why bioeroding sponges may be better hosts for symbiotic dinoflagellates than many corals
Christine H.L. Schnberg(1*), Ryota Suwa(2)
Centre for Marine Studies, The University of Queensland, St. Lucia, QLD 4072, Australia; present address: Carl von Ossietzky University Oldenburg, Faculty 5, Institute of Biology and Environmental Sciences, Department of Zoomorphology and Systematics, 26129 Oldenburg, Germany, ph +441-9783611, fax +441-9783250. christine.schoenberg@uni-oldenburg.de (2) University of the Ryukyus, Department of Biology, Faculty of Science, Nishihara, Okinawa 903-0213, Japan. k048565@eve.u-ryukyu.ac.jp
(1)

Abstract: Bioeroding sponges in symbiosis with dinoflagellates of the genus Symbiodinium (zooxanthellae) appear to be more bleaching-resistant than other coral reef organisms. We propose that stress tolerance in invertebrate-dinoflagellate symbiosis is strongly related to protection provided by the host. We investigated whether sponge zooxanthellae can be redistributed depending on external factors such as light, which can cause stress. Our observations on daily colour change in the endolithic demosponge Cliona orientalis Thiele, 1900 indicated that the intracellular symbionts are moved by the sponge. We propose that this mechanism can effectively be used to prevent bleaching. Applying different approaches we clearly showed that zooxanthellae were redistributed. Following a diel rhythm, zooxanthellae are most concentrated and closest to the sponge surface during the day and widely distributed and drawn into the sponge tissue during the night. We suggest that this behaviour (1) helps to maximise photosynthetic yield during daylight, (2) minimises loss of symbionts to night- and twilight-active grazers and, (3) can be used during periods of stress, moving the symbionts further into the protective sponge tissue and into the substrate. Accidental light stress in the holding tank caused a different symbiont distribution at the end of the study than at the beginning, with a more even distribution and an increased concentration of zooxanthellae in about -2 to -4 mm tissue depth, partly overriding diel patterns. Due to this distribution, a dilution effect and removal of the symbionts from the source of stress was achieved. We conclude that hosts can provide significant protection to their microbial symbionts if they are able to move them, which strongly increases bleaching-resistance. Keywords: bleaching tolerance, Cliona orientalis, diel rhythm, redistribution, Symbiodinium

Introduction
Tropical coral reefs are experiencing growing levels of environmental pressure. Probably the most devastating but not the only threat is the increase of sea surface temperatures related to climate change (e.g. Hoegh-Guldberg 1999, Reaser et al. 2000). Stress impairs the photosynthesis of symbiotic dinoflagellates in marine organisms and triggers the loss of the symbionts and their pigments, an event known as bleaching. In the last two decades, environmental stress has repeatedly led to severe bleaching events and to mass mortalities of reef organisms, occasionally resulting in recognisable phase shifts affecting coral dominance (e.g. Loya et al. 2001). Bioeroding sponges that can colonize reef corals appeared immune to events that led to coral bleaching (Vicente 1990), even though some of the more successful sponge species are known to contain the same genus of symbiotic dinoflagellates found in corals (Symbiodinium spp.; Vacelet 1981, Rtzler 1990, Schnberg and Loh 2005). The putative hardiness of bioeroding sponges would have two important implications: (1) Studying bioeroding sponges, we may gain a better understanding of mechanisms supporting stress resistance

in coral reef organisms. (2) We may have to expect a phase shift of the reef community from coral-dominated to spongedominated, and thus an increase in bioerosion. There is circumstantial evidence that bioeroding sponges will survive presently occurring negative conditions better than some corals. Symbiotic bioeroding sponges were observed to appear unchanged when other reef organisms bleached during heating events (Vicente 1990, Schnberg and Wilkinson 2001, Mrquez et al. 2006). Increased illumination resulted in bleaching in Cliona varians. However, the sponge eventually revovered, re-acquiring symbionts horizontally (Hill and Wilcox 1998), suggesting that it may be able to survive an asymbiotic phase relying on heterotrophy. If this is the case, sponges may survive and recover where many corals would die, eventually leading to changes in the community. Bioeroding sponges are known to benefit from deteriorating environmental conditions such as eutrophication (Rose and Risk 1985, Holmes 1997, Holmes 2000, Holmes et al. 2000). Aggressive, encrusting bioeroding sponges in the Caribbean and on the Australian Great Barrier Reef have increased in abundance (Rtzler 2002, Schnberg 2001, Schnberg pers. obs. 2003-2004, Lpez-Victoria and Zea 2005). Bioeroding

570

sponge communities on severely heat-damaged reefs in Okinawa were dominated by symbiotic species (Schnberg pers. obs. 2004). We propose that the hardiness of symbiotic bioeroding sponges is strongly controlled by the host. Bioeroding sponges live endolithically in calcium carbonate substrates that provide significant shelter against strong light, which can induce bleaching in coral reef organisms (e.g. Fitt and Warner 1995, Glynn 1996, Buck et al. 2002). We further assume that the intracellular symbionts can be moved away from the source of stress if need arises. We based this theory on the following observations: (1) Colonies of the Australian bioeroding sponge Cliona orientalis are dark brown during the day, but beige during the night (Schnberg 2000). (2) After sampling sponge tissue by coring, dark rims appear at the cut edges of C. orientalis, presumably formed by higher concentrations of symbionts assisting in the healing process (Schnberg 2000, 2006). (3) When trying to artificially bleach C. orientalis, it may undergo a colour change, but when sponge grafts are cut in half, dense dark accumulations of symbionts can be found in the centres of the grafts (Schnberg pers. obs. 2003). As the phenomenon of symbiont transport may play a central role in the hardiness of photosynthetic bioeroding sponges, we conducted a study testing the following hypotheses: (1) The diel colour change observed in the field putatively is caused by differing symbiont concentrations in the uppermost layer of the sponge. The sponge symbionts are thought to be redistributed within the sponges in a diel rhythm. If this cannot be shown, the sponge may release its symbionts overnight and re-acquire them in the morning. The first cause for the diel colour change was regarded to be the more likely. (2) When exposed to stress factors that influence the holosymbionts physiology, the sponge is expected to redistribute the zooxanthellae and transport them away from the immediate action of the stress.

concentration of photosynthetic pigments throughout the sponge overnight.

Sampling, sample preparation and experimental design

Live sponge tissue was cored from three different colonies on horizontal surfaces in August 2005 with an air-driven underwater drill from infested massive Porites colonies in 12 m depth off Heron Island, Capricorn Bunker Group, Great Barrier Reef, Australia (Coral Garden, Fig. 2). Sponge cores were transported back to the Heron Island Research Station (HIRS) in seawater and transferred to a 100 l holding tank in flow-through, natural seawater conditions. The tank was covered with a double layer of shade cloth to reduce the ambient light to levels more similar to the sample depth (sample site: 180-300 mol photons m-2 s-1, tank without shade cloth in full sun: 2500-2800 mol photons m-2 s-1; tank with shade cloth: 340-350 mol photons m-2 s-1). Light measurements were conducted with light loggers at the sample site and with a LI-COR light meter and an air-water hybrid sensor at the aquarium. For each measurement series on the sponges one core per sampled colony was removed from the tank haphazardly, resulting in a sample size of N = 3 per observation. At HIRS we took photographs, visually inspected the cores with a colour chart (home-made chart that was later compared to CoralWatch coral health chart, see references) and monitored their photosynthetic properties with pulse amplitude modulated fluorometry (see below). For three days, data were collected three times during daylight (morning, noon and late afternoon) and once in the dark. On the fourth day, the amount of remaining cores allowed only three observations in total, i.e. during the morning, afternoon and during night. Cores were then frozen and stored at -20C until further examinations as described below.

Material and methods The studied sponge, Cliona orientalis Thiele, 1900
The study was carried out on the bioeroding sponge Cliona orientalis Thiele, 1900, which harbours G-type Symbiodinium (Schnberg and Loh 2005). We used the encrusting growth form that has a coherent tissue layer on the substrate surface and evenly penetrates the substrate, which made observations easier than in earlier growth stages. Symbiotic dinoflagellates in healthy C. orientalis are distributed inhomogeneously (Schnberg 2000). The vast majority of the symbionts is located in a dense, dark brown, superficial layer of about 1 mm thickness (Fig. 1). The sponge extends to about 1.3 to 1.6 cm depth in the coral skeleton, but underneath the brown surface layer its tissue looks ochre yellow, suggesting a much lower density of symbiont cells compared to the surface layer (Fig. 1). If the sponge were to expel symbionts overnight and take them up again in the morning, the brown layer would vanish and pigment concentrations would be significantly lower in the sponge tissue by night compared to by day. If the symbionts were redistributed we would expect to find either a change of width of the brown surface layer of symbionts, a change in its location relative to the surface or a homogeneous

Measurements and observations

Photosynthetic properties of each series of three live sponge cores were examined with pulse amplitude modulated (PAM) fluorometry employing a Maxi imaging PAM fluorometer (Walz, Effeltrich, Germany; Schreiber et al. 2003, Consalvey et al. 2005 and references therein). In contrast to earlier PAM fluorometers, an imaging fluorometer allows two-dimensional visualisation of chlorophyll a fluorescence and can be used for mapping spatial heterogeneity of photosynthetic parameters (see e.g. Ralph et al. 2005, who used this tool for investigations of spatial patterns of coral zooxanthella photosynthesis). In the present publication we follow concepts and terminology as given by Consalvey et al. (2005). Present photosynthetic parameters measured with PAM do not have units. Cores were taken from the holding tank, and photosynthetic properties were directly measured from the top (the original surface) and in vertical crossection after splitting the cores in half with hammer and chisel. Measurements during the day were thus not dark-adapted, and our photosynthetic parameters are effective, rather than absolute values. To visualise the spatial distribution of active chlorophyll a (chl a) in the cores, we obtained images of the minimum lightadapted fluorescence yield F0, and took comparative, true

571

Fig. 1: Live Cliona orientalis cores as seen from the surfaces (left) and in crossection (right) at different daytimes. Core diameters are 3.5 cm. Enlarged core in crossection, far lower right: Symbiotic dinoflagellates are mainly concentrated in the uppermost 1-1.5 mm of the sponge (brown band), the rest of the sponge tissue appears ochreyellow. White parts of the cores are areas of coral skeleton not invaded by the sponge. Other enlargements of crossections show that the dark band of symbionts was wider during night (N) than during day (D). Core surfaces were darker and browner during the day than when observed after dark. The surface views of cores in the fourth row clearly show the higher accumulation of zooxanthellae at the cut rims of the cores (asterisk). During the end of the second day (row 8) and during the third day (rows 9-12), cores look paler and more yellow, which was an effect of light stress. On some cores scars can be seen that resulted from gas bubbles in the tissue (crosses). Digital photographs from the fourth day were regrettably lost during a hard drive failure in the field.

colour photographs with a digital camera using the flash (Cybershot DSC-P100, Sony, Japan). We also measured the effective photosystem II quantum yield or light utilization efficiency (Fq/Fm) that represents unbleached photosynthetic units available for photosynthesis and is sensitive to stress. Whereas F0 was used as a proxy for chlorophyll or symbiont distribution (see Consalvey et al. 2005 and references therein), Fq/Fm was employed to monitor the health and function of the sponge symbionts. Using a Maxi imaging PAM allowed two-dimensional integration of these values after designating areas of interest (AOI). An AOI from a core top was a circle of constant area placed centrally to avoid parts that may have

been damaged from the coring (example on Fig. 3). AOIs on the split cores in crossection were five consecutive slim belts of about 2 mm height x a third of the width of a given core, aligned with the edge of the upper core surface (example on Fig. 3). AOIs of surface and crossection measurements differed in area, and resulting values cannot necessarily be compared to each other. Furthermore, the sponge has a spicule palisade in the surface tissue, in which the vertical, opaline spicules possibly act like optical fibres allowing the light to penetrate deeper into the tissue (similar mechanism was observed for Tethya seychellensis: Gaino and Sar 1994). Spicules within the body of C. orientalis are unordered. In statistical analyses

572

Fig. 2: Sample location Coral Gardens, south side of Heron Island, Capricorn Bunker Group, Great Barrier Reef, Queensland (QLD), Australia.

we thus only used PAM data obtained from sponge cores in crossection that were directly comparable. We manually measured the width of the superficial layer of symbionts in crossection with a ruler at different times of the day and estimated chl a concentrations in different tissue layers of the sponge cores. Using a razor blade and a ruler we carefully shaved off three consecutive layers of sponge material aiming at taking 1.5 mm thick layers each from the top down (ca. 100 mg each sample). Removed layers in daylight-samples were (1) a dark brown sample from the surface, including the dense band of symbionts, (2) a brown to dark yellow sample from the layer immediately below the surface, and (3) an ochre-yellow sample from the layer below that. The material was weighed (Sartorius analytical scale), and chl a was extracted from each tissue sample in 1.4 ml laboratory-grade methanol for 2h in dark (Wellburn 1994). After centrifugation absorption per extracted sample was quantified at a Shimadzu spectrophotometer and used to calculate chl a concentrations per g sample. As the used tissue samples were composite material comprised of sponge tissue, sponge spicules and coral skeleton, resulting chl a concentrations represent relative rather than absolute values. Chl a concentrations were used to confirm fluorescence data, i.e. to obtain additional information on the symbiont distribution. After testing the assumptions for statistical approaches, data were analysed with ANOVA, the used factors being daytime and tissue depth. As towards the end of the experiment light levels of the holding tanks accidentally rose above ambient levels at the sample site, the influence of possible stress from increased illumination was investigated as well. However, as we had no clear evidence when this situation started to act and photosynthetic efficiency remained comparatively stable throughout the experiment, we were unable to separate our

data set into unstressed data and stressed data for statistical comparison, and observations remained qualitative.

Results Colour change

Using different tools, we were able to provide clear evidence that dinoflagellate symbionts in our samples of Cliona orientalis were redistributed in a diel cycle and that this cycle can be affected by stress. Sponge core upper surfaces were visibly paler during the night than during the day (Fig. 1, Table 1). Cores did not become as pale as colonies previously observed during night dives in the field (Schnberg 2000), but the colour change was still quite evident. Daylight colours of sponge core surfaces ranged from dark chocolate brown to brown, night colours from hazelnut brown to dark beige (see colour numbers of the CoralWatch coral health chart as given in Table 1). From visual inspection only we also had the impression that overnight and with increasing length of the experiment, the dark superficial band of zooxanthellae widened or moved deeper into the sponge cores. Towards the end of the experiment, sponge cores became affected by stronger light despite of the shade cloth. They developed a bright yellow pigment that was already noted during earlier stress experiments (Schnberg and Walpersdorf unpubl. data 2003), and the cores contained gas bubbles in the superficial tissue layer. On the fourth day the yellow pigment faded again, and the bubbles left scars that were obviously healing.

Changes in fluorescence
Chl a fluorescence imaging supported and confirmed the above results. F0 values, our first proxy for zooxanthella distribution, followed a diel rhythm and was influenced by

573

Fig. 3: Chlorophyll a fluorescence images of live Cliona orientalis cores as seen from the surfaces and in crossection at different daytimes. Core diameters are 3.5 cm. Minimum light-adapted fluorescence yield F0 was used as a proxy for the distribution of chlorophyll a. High values of F0 are bright green, low values orange and red. Lack of fluorescence or chlorohyll is displayed in black. On the first core in surface view we displayed a circular area of interest (AOI) to indicate how fluorescence was measured from the surface. On enlargement on the lower right side the sampling procedure in crossection is demonstrated (AOI 1-5). Other enlargements show that the chlorophyll distribution differed between day (D) and night (N) samples. In the latter samples, brighter colours for F0 indicated that zooxanthellae lay deeper in the sponge cores than during the day (arrows).

increased illumination (Fig. 3 and Fig. 4: white bars, Table 1). F0 significantly differed with tissue layer, but differences with daytime were only significant at the 92% level, or at the 95% level if data from crossectional observations were included (Table 2). However, trends were nevertheless quite obvious. Daytime surface values ranged between 0.2 and close to 0.3, whereas night values always remained below 0.2 (Fig. 4: light grey parts of graphs). In crossection, and

in the unstressed daylight cores of the first two days of the experiment, F0 continuously decreased from the surface into the sponge, with typical values being between 0.15 and close to 0.2 in the uppermost 2 mm, and approximately 0.04 in the innermost layer, near the clean coral skeleton. During night and with increasing stress, F0 was lowered in the uppermost surface layer, and the second crossectional layer in -2 to -4 mm often displayed another peak (Fig. 3, Fig. 4, Table 1). On the

574

Table 1: Summary of results on diel and stress-related symbiont redistribution in the bioeroding sponge Cliona orientalis. Values are means of three replicates and two measurements (except for noon values of stressed cores, they are based on three replicates only). F0 was used as a proxy for zooxanthella distribution, Fq/Fm for photosynthetic health (higher values) the amount of stress acting on the symbionts (lower values). No stress (first 2 days of experiment) Morning brown (RHS200C) beige to pale brown with yellow brown (RHS199C- tinge (RHS165A N199A) with RHS12A) brown with yellow tinge (RHS199A with RHS12A) Noon Afternoon Night Morning Noon Afternoon Stress (last two days of experiment) Night beige with yellow tinge (RHSN199A with RHS12A)

Observation

Surface colour (according to the mini colour chart of the Royal Horticultural Society, London) F0 continuously decreases with tissue depth. In surface layers lowest values occur during the night surf. = 0.26 to 2 = 0.18 to 4 = 0.15 to 6 = 0.09 to 8 = 0.06 to 10 = 0.05 surf. = 0.23 to 2 = 0.19 to 4 = 0.16 to 6 = 0.11 to 8 = 0.07 to 10 = 0.05 surf. = 0.25 to 2 = 0.15 to 4 = 0.13 to 6 = 0.09 to 8 = 0.06 to 10 = 0.05 surf. = 0.17 to 2 = 0.11 to 4 = 0.13 to 6 = 0.09 to 8 = 0.06 to 10 = 0.05 surf. = 0.19 to 2 = 0.13 to 4 = 0.16 to 6 = 0.11 to 8 = 0.08 to 10 = 0.06 surf. = 0.20 to 2 = 0.12 to 4 = 0.15 to 6 = 0.11 to 8 = 0.08 to 10 = 0.07

F0

F0 decreases with tissue depth, but has another peak in the layer between 2 and 4 mm. Variability between tissue layers and daytime decreases with increasing stress. surf. = 0.19 to 2 = 0.14 to 4 = 0.14 to 6 = 0.11 to 8 = 0.08 to 10 = 0.06 surf. = 0.13 to 2 = 0.15 to 4 = 0.15 to 6 = 0.10 to 8 = 0.07 to 10 = 0.05

F0 from the surface and in 5 consecutive tissue layers of 2 mm each

Fq/Fm surf. = 0.57 to 2 = 0.51 to 4 = 0.53 to 6 = 0.50 to 8 = 0.47 to 10 = 0.42 mean: 1.9 mm SD: 0.8 120-130 40-85 10-30 50-70 10-20 130-140 mean: 2.3 mm SD: 0.6 mean: 3.2 mm SD: 0.8 100-120 70-95 20-30 surf. = 0.51 to 2 = 0.51 to 4 = 0.56 to 6 = 0.57 to 8 = 0.53 to 10 = 0.42 surf. = 0.61 to 2 = 0.49 to 4 = 0.58 to 6 = 0.55 to 8 = 0.49 to 10 = 043 surf. = 0.61 to 2 = 0.44 to 4 = 0.55 to 6 = 0.57 to 8 = 0.53 to 10 = 0.46

Fq/Fm does not change with daytime but differs in different tissue layers. surf. = 0.57 to 2 = 0.46 to 4 = 0.57 to 6 = 0.57 to 8 = 0.57 to 10 = 0.51 surf. = 0.61 to 2 = 0.43 to 4 = 0.51 to 6 = 0.51 to 8 = 0.50 to 10 = 0.52 surf. = 0.64 to 2 = 0.53 to 4 = 0.57 to 6 = 0.56 to 8 = 0.54 to 10 = 0.48 increasing by 0.5-1.5 mm 60-90 70-90 30-60 100 85 30 50-60 60-80 30-80 60-90 50-70 70-85 surf. = 0.58 to 2 = 0.44 to 4 = 0.55 to 6 = 0.57 to 8 = 0.54 to 10 = 0.47

Fq/Fm from the surface and in 5 consecutive tissue layers of 2 mm each

Zooxanthella band width (mm) 100-110 70-110 15-25

mean: 4.0 mm SD: 0.4

g chl a mg1 in uppermost 1.5 mm

g chl a mg1 in -1.5 to -3 mm

g chl a mg1 in-3 to -4.5 mm

575

Table 2: Statistical results of testing properties of Cliona orientalis zooxanthellae against daytime (one-way ANOVA ) or daytime and tissue depth (two-way ANOVA). For values obtained with pulse amplitude modulated fluorometry (F0 and Fq/Fm) only data from the crossectional surface were included in the analysis. Source F0 Daytime Tissue layer df Sum of squares Mean square F P Posthoc Tukey Kramer

3 4

0.003 0.114 0.001 0.021 0.003 0.107 0.015 0.127 14.964 47.347 180.366 155977.671 5789.480 78713.572

0.001 0.029 <0.001 <0.001 0.001 0.027 0.001 0.002 4.988 1.155 60.122 77988.836 964.913 639.948

2.484 74.168 0.303

0.0703 0.0001 0.9862

No significant difference at the 95% level, but significant when surface values were included: night differed against morning and noon Only the deepest vs. the second deepest and uppermost vs. the second layer did not differ

Daytime * tissue layer 12 Residual 55 Fq/Fm Daytime Tissue layer 3 4

0.394 11.590 0.560

0.7575 0.0001 0.8638

No significant difference Significant differences: uppermost vs. second and third layer, the deepest (5th) layer vs. the second, third and fourth layer

Daytime * tissue layer 12 Residual 55 Zooxanthella band width Daytime 3 Residual Chl a concentration Daytime Tissue layer 41 3 2

4.319

0.0098

Zooxanthella band width significantly differed for noon vs. morning

0.094 121.868 1.508

0.9633 0.0001 0.1810

No significant difference Chl a concentration significantly differed in each tissue layer compared to each other tissue layer

Daytime * tissue layer 6 Residual 123

fourth day of the experiment, F0 measured from the top was markedly reduced compared to earlier measurements (Fig. 3, Fig. 4) and suggested a decrease in zooxanthella density rather than degeneration of chlorophyll, because effective photosynthetic efficiency remained strong. When tentatively declaring the last two days of the experiment as influenced by increased illumination and entering the factor light into statistical analysis, there was a significant interaction factor between tissue layer and light, supporting the above observations and indicating that the increased illumination had an influence on zooxanthella distribution. Effective photosynthetic efficiency Fq/Fm varied between 0.5 and 0.6+. It did not display a diel rhythm, but significantly differed between tissue layers (Fig. 4: black bars, Table 1 and 2). Highest values were measured from the top of the cores, or in the second layer in crossection (2-4 mm into the cores). During night and with increasing stress, the highest photosynthetic efficiency was detected further inside the

cores and was often associated with the third layer into the cores (at 4-6 mm tissue depth, Table 1). With increasing light levels in the holding tank, Fq/Fm values as a whole slightly dropped and data variability increased (end of second day, third day, Table 1), but this was not a significant effect, and light and tissue layer did not interact in a tentative analysis. During the fourth day of the experiment, Fq/Fm recovered and reached maximum values. The holding conditions were not changed, but the weather on the fourth day was variable and the sky partly overcast.

Changes in symbiont band width and chlorophyll a concentration

Measuring the width of the dark brown surface layer of the highest zooxanthella concentration and quantifying the relative chl a concentration of the topmost three tissue layers of C. orientalis again supported above results of

576

Fig. 4: Minimum light-adapted or effective fluorescence yield (F0, a proxy for the distribution of chlorophyll a, white bars) and light utilisation efficiency (Fq/Fm, a proxy for photosynthetic health of the symbionts, black bars) of cores taken from Cliona orientalis. The first pair of bars of all graphs (on light grey area) represents measurements taken from the upper surfaces of the cores, the other 5 pairs of bars per graph (on white area) are values measured in crossection. Crossectional values were taken from the surface down in five layers: in 0 to -2, -2 to -4, -4 to -6, -6 to -8 and 8 to -10 mm tissue depth of the cores. Data sets of different daytimes are distributed vertically (morning, noon, afternoon, night) and consecutive days horizontally. Graphs with a dark grey background represent values taken during the night. On day four only 3 data sets are available. Error bars are standard deviations (N = 3).

577

Fig. 5: Width of the dark brown symbiont layer observed in crossection at or near the surface of Cliona orientalis. A diel rhythm is expressed in the repeated widening (night and morning) and narrowing (afternoon) of the symbiont bands. The trendline indicates an overall increase of symbiont layer width with stress. Grey bars mark dark phases at night. Error bars are standard deviations (N = 3).

symbiont redistribution. Zooxanthella layer width followed a distinct and significant diel rhythm (Fig. 5, Table 1 and 2), with the narrowest extent of about 2 mm at noon and in the afternoons. Layer width increased overnight to a width of 2-4 mm, further widening in the morning with 3.5-4.5 mm, and then again distinctly concentrating and narrowing to regain the condensed daylight layer width. Morning values differed significantly from noon values (Table 2). Over the course of the experiment and with increasing light levels in the holding tank, the layer of symbionts widened increasingly. As could be expected from the colouring of sponge samples and from fluorescence measurements, chl a concentrations per tissue layer, i.e. densities of zooxanthellae, significantly decreased with increasing depth into the sponge (Fig. 6, Table 1 and 2). Under unstressed conditions during the first two days of the experiment, the distribution of the symbionts could be roughly described as being 5x as concentrated in the dark brown surface layer (Fig. 6, black circles) compared to the ochre yellow parts of the sponge (> 3 mm into the sponge tissue). In a distance of > 1.5 to < 3 mm from the surface (second layer sampled; Fig. 6, grey circles) the symbiont density was already reduced to 3/5 of the surface concentration. In the uppermost surface layer a diel rhythm could be discerned, with chl a concentrations rising over the day, displaying the highest values in the afternoon and dropping again over night and towards the morning. This effect was not statistically significant (Table 2), but it was inversely proportional to band-width measurements. Trends in chl a concentrations were not quite as clear in the two lower sample layers, maybe due to difficulties in keeping the sampling procedure exactly the same between the samples. However, chl a concentrations in the two lower layers were roughly contrasting the developments of the superficial layer, with the third sample layer showing the least pronounced changes (Fig. 6). During the last two days of the experiment the effects of light stress became apparent in a very different symbiont distribution pattern than during the first two days: Mean chl a concentrations of all three sampled layers were the same, and variability of the concentrations per sample increased drastically. This indicated that zooxanthellae were now much more evenly distributed in the upper 4.5 mm of the

sponge cores, which is congruent with the observation that the symbiont band width increased with increasing exposure to higher light levels. All results are summarised in Table 1.

Discussion
Redistribution of dinoflagellate symbionts in the bioeroding sponge Cliona orientalis could be shown by using different proxies: direct observations of colour change and band width of symbiont distribution together with chlorophyll a concentrations obtained with pulse amplitude modulated fluorescence and spectroscopy. Symbionts were in densest distribution in the sponge surface layer, but were wider distributed and deeper in the sponge during night compared to daytime observations. The possible effect of stress caused by increased illumination reduced the amplitude of the rhythmic diel effects and resulted in a more even distribution of symbionts or their distribution deeper into the sponge. A reaction to a possibly harmful situation was evident by the above observations, but could not be shown by a significant reduction of effective light utilisation efficiency. Symbiont redistribution may have beneficial effects for the sponge and the symbionts and occurs for different reasons. Observed diel changes are most likely related to optimising the photosynthetic harvest during the day and reducing risks during the night. At noon and during the second half of the day the symbionts were densely crowded at the surface of the sponge, receiving high levels of light. It has been shown that C. orientalis and other, closely related bioeroding sponges benefit from their symbionts, putatively by receiving nutrition in form of zooxanthella photosynthates (Rosell and Uriz 1992, Hill 1996, Schnberg 2006). Therefore not only the symbionts profit from an effective light-harvest, but also the sponge. Overnight, the band of zooxanthellae is slightly retracted into the sponge and the symbionts are more evenly distributed over deeper tissue areas in the sponge. As there is no need to keep them near the surface in the dark, this reduces the risk of losing part of the symbiont population to nightactive grazers. Similar mechanisms have been described from the zooxanthellate holotrich ciliate Meristentor dinoferus that

578

Fig. 6: Relative chlorophyll a concentrations, i.e. symbiont densities, in the three top tissue layers of Cliona orientalis (uppermost layer = 0-1.5 mm, black circles and continuous line; second layer = 1.5-3 mm, grey circles and dashed line; third layer = 3-4.5 mm, white circles and dotted line). Chlorophyll a concentrations decreased with tissue depth, and highest surface concentrations were measured during the afternoons. Due to stress the chlorophyll concentration evened out between the tissue layers. Grey bars mark dark phases at night. Error bars are standard deviations (N = 3).

also withdraws its symbionts over night into less exposed body parts, in this case into the stalk (Lobban et al. 2002). It is unknown whether one or both symbiont partners control the symbiont redistribution. As the symbionts in bioeroding sponges are intracellular (e.g. Vacelet 1981, Rtzler 1990), the movement is most likely carried out by active sponge cell migration, but we do not yet know how it occurs and the exact factors that affect this behaviour. The sponge may receive chemical cues from the zooxanthellae, either by the formation of oxygen radicals or by other substances released by the symbionts. Extreme photosynthesis levels during our experiment putatively led to the formation of gas bubbles in the sponge tissue, i.e to gas embolism, and later to temporary scars. This negative effect can be reduced by moving the zooxanthellae into deeper tissue layers. Another stress-related, but yet mostly unknown reaction is the production of a yellow pigment in the sponge cores. This colour has previously been observed in the field in patchy-pale colonies of C. orientalis, presumably during reproduction or heat stress during summer (Schnberg unpubl. data 1997) and in the tank during stress experiments (Schnberg pers. obs. 2003). Nothing is known about this pigment, except that it is produced during times of stress. The pigment could be a carotenoid helping to quench surplus light energy (Consalvey et al. 2005) or may provide a sunscreen-effect similar to that of mycosporine-like amino acids (MAAs) known from other reef organisms (e.g. Benaszak 1995, 1998, Dunlap 1998). Even though we were unable to monitor the effects of light stress by a significant reduction of light utilisazion efficiency, we took the occurrence of the yellow pigment, the gas bubbles in the tissue and the redistribution of the symbionts as stress indicators. Stress-related redistribution of symbionts partly overrode the clear diel rhythm and resulted in the following patterns: symbionts were more evenly distributed, they were moved further away from the surface, i.e. from the source of stress, and could eventually be accumulated at the greatest distance of the source of stress, e.g. the bottom of the sponge core or in its centre (Schnberg pers. obs. 2003). Redistribution may result in the reduction of the stress level on the symbiont population. By drawing symbionts deeper into the body,

C. orientalis provides significant shading that effectively protects the zooxanthellae and reduces their photosynthetic activity. This process is particularly effective in bioeroding sponges, where the substrate provides extra shading for the endolithic tissue (see also Schnberg 2006). A more even symbiont distribution additionally results in the dilution of the toxic effect of surplus oxygen that can cause bleaching in other reef organisms. By using these strategies, possibly in combination with the production of the yellow pigment, C. orientalis reduced the stress acting on the zooxanthellae so well that we could not detect stress by using pulse amplitude modulated fluorometry, and the photosynthetic efficiency did not change. Many corals cannot use the strategy of redistributing their zooxanthellae, as they only form a near one-dimensional tissue layer that coats their skeletons. However, some corals such as massive Porites do not have consecutive horizontal skeletal barriers (coenostea) and can stretch their tissue a few cm into their own skeletons to reduce and evade external stress (e.g. Brown et al. 1991). Lower mortality rates and sublethal damage in massive Porites (Marshall and Baird 2000, Baird and Marshall 2002, Baker et al. 2004) and symbiotic bioeroding sponges (Vicente 1990, Schnberg and Wilkinson 2001, Mrquez et al. 2006) compared to other coral reef organisms suggest that the ability of moving zooxanthellae into deeper regions is a very important mechanism for survival. C. orientalis thus provides a superior protection simply by the combination of the above mechanisms and its endolithic life style. Symbiont stress tolerance adds to these effects: C. orientalis contains G-clade Symbiodinium (Schnberg and Loh 2005) that are comparatively hardy with respect to heat (Schnberg et al. in press). At community level, from the greater resistance to bleaching of symbiotic bioeroding sponges such as C. orientalis we may expect increasing abundances of these sponges and ultimately extending rates of bioerosion on coral reefs (Schnberg 2001, Rtzler 2002, Schnberg pers. obs. 2003-2004, Lpez-Victoria and Zea 2005). In the present times of expanding coral reef demise, this adds yet another threat to structural and organismic diversity of this unique environment.

579

Acknowledgements
The study was conducted at the Centre for Marine Studies at the University of Queensland, Australia. C. S. received a Feodor Lynen Fellowship by the Alexander von Humboldt Foundation in Bonn, Germany, funds from the Australian Research Council to O. Hoegh-Guldberg and from the University of Queensland Chancellors Fund. We thank M. Hidaka for always supporting R. Suwa and for letting him come to Australia for the present study. We are indebted to excellent assistance by all station staff at Heron Island Marine Station. S. Harii assisted during the sampling dive, and A. Jones participated in laboratory analyses for this study as work experience. F. Reichel and J. Kolbowski from Walz provided technical support. We appreciate comments of A. Lawton and W. Loh prior to manuscript submission.

References
Baird AH, Marshall PA (2002) Mortality, growth and reproduction in scleractinian corals following bleaching on the Great Barrier Reef. Mar Ecol Prog Ser 237: 133-141 Baker AC, Starger JC, McClanahan TR, Glynn PW (2004) Corals adaptive response to climate change. Shifting to new algae may safeguard devastated reefs from extinction. Nature 430: 741 Benaszak AT (1995) Effects of ultraviolet (UV) on marine microalgalinvertebrate symbioses. 2. The synthesis of mycosporine-like amino acids in response to exposure to UV in Anthopleura elegantissima and Cassiopeia xamachana. J Exp Mar Biol Ecol 194: 233-250 Benaszak AT (1998) Relationship between ultraviolet (UV) radiation and mycosporine-like amino acids (MAAs) in marine organisms. Bull Mar Sci 63: 617-628 Brown BE, Tudhope AW, Le Tissier MDA, Scoffin TP (1991) A novel mechanism for iron incorporation into coral skeletons. Coral Reefs 10: 211-215 Buck BH, Rosenthal H, Saint-Paul U (2002) Effect of increased irradiance and thermal stress on the symbiosis of Symbiodinium microadriaticum and Tridacna gigas. Aquat Living Res 15: 107117 Consalvey M, Perkins RG, Paterson DM, Underwood GJC (2005) PAM fluorescence: a beginners guide for benthic diatomists. Diatom Research 20: 1-22 Dunlap WC (1998) Ultraviolet radiation-adsorbing mycosporinelike amino acids in coral reef organisms: a biochemical and environmental perspective. J Phycol 34: 418-430 Fitt WK, Warner ME (1995) Bleaching patterns in four species of Caribbean reef corals. Biol Bull 189: 298-307 Gaino E, Sar M (1994) Siliceous spicules of Tethya seychellensis (Porifera) support the growth of a green alga a possible light conducting system. Mar Ecol Prog Ser 108: 147-151 Glynn PW (1996) Coral reef bleaching: facts, hypotheses and implications. Global Change Biol 2: 495-509 Hill MS (1996) Symbiotic zooxanthellae enhance boring and growth rates of the tropical sponge Anthosigmella varians forma varians. Mar Biol 125: 649-654 Hill MS, Wilcox T (1998) Unusual mode of symbiont repopulation after bleaching in Anthosigmella varians: acquisition of different zooxanthellae strains. Symbiosis 25: 279-289 Hoegh-Guldberg O (1999) Climate change, coral bleaching and the future of the worlds coral reefs. Mar Freshwater Res 50: 839-866

Holmes KE (1997) Eutrophication and its effect on bioeroding sponge communities. Proc 8th Int Coral Reef Symp, Balboa 2: 1411-1416 Holmes KE (2000) Effects of eutrophication on bioeroding sponge communities with the description of new West Indian sponges, Cliona spp. (Porifera: Hadromerida: Clionidae). Invertebr Biol 119: 125-138 Holmes KE, Edinger EN, Hariyadi, Limmon GV, Risk MJ (2000) Bioerosion of live massive corals and branching coral rubble on Indonesian coral reefs. Mar Poll Bull 40: 606-617 Lobban CS, Schefter M, Simpson AGB, Pochon X, Pawlowski J, Foissner W (2002) Meristentor dinoferus n. gen., n. sp., a giant heterotrich ciliate (Spirotrichea: Heterotrichida) with zooxanthellae, from coral reefs on Guam, Mariana Islands. Mar Biol 140: 411-423 Lpez-Victoria M, Zea S (2005) Current trends in space occupation by encrusting excavating sponges on Colombian coral reefs. Mar Ecol 26: 33-41 Loya Y, Sakai K, Yamazato K, Nakano Y, Sambali H, van Woesik, R (2001) Coral bleaching: the winners and the losers. Ecology Letters 4: 122-131 Mrquez JC, Zea S, Lpez-Victoria M (2006) Is competition for space between the encrusting excavating sponge Cliona tenuis and corals influenced by higher-than-normal temperatures? Bol Invest Mar Cost 35: 256-265 Marshall PA, Baird AH (2000) Bleaching of corals on the Great Barrier Reef: differential susceptibilities among taxa. Coral Reefs 19: 155-163 Ralph PJ, Schreiber U, Gademann R, Khl M, Larkum AWD (2005) Coral phytobiology studied with a new imaging pulse amplitude modulated fluorometer. J Phycol 41: 335-342 Reaser JK, Pomerance, R, Thomas PO (2000) Coral bleaching and global climate change: scientific findings and policy recommendations. Conserv Biol 14: 1500-1511 Rose CS, Risk MJ (1985) Increase in Cliona delitrix infestation of Montastrea cavernosa heads on an organically polluted portion of the Grand Cayman fringing reef. PSZN Mar Ecol 6: 345-363 Rosell D, Uriz, MJ (1992) Do associated zooxanthellae and the nature of the substratum affect survival, attachment and growth of Cliona viridis (Porifera: Hadromerida)? An experimental approach. Mar Biol 114: 503-507 Royal Horticultural Society (2007) RHS gardening for all. http:// www.rhs.org.uk, or mailorder@rhs.org.uk Rtzler K (1990) Associations between Caribbean sponges and photosynthetic organisms. In: Rtzler K (ed). New perspectives in sponge biology. Smithsonian Institution Press, Washington DC. pp. 455-466 Rtzler K (2002) Impact of crustose clionid sponges on Caribbean reef corals. Acta Geol Hispanica 37: 61-72 Schnberg CHL (2000) Bioeroding sponges common to the central Great Barrier Reef: descriptions of three new species, two new records, and additions to two previously described species. Senckenbergiana marit 30: 161-221 Schnberg CHL (2001) Small-scale distribution of Great Barrier Reef bioeroding sponges in shallow water. Ophelia 55: 39-54 Schnberg CHL (2006) Growth and erosion of the zooxanthellate Australian bioeroding sponge Cliona orientalis are enhanced in light. Proc 10th Int Coral Reef Symp, Okinawa. 1: 168-174

580

Schnberg CHL, Loh WK (2005) Molecular identity of the unique symbiotic dinoflagellates found in the bioeroding demosponge Cliona orientalis. Mar Ecol Prog Ser 299: 157-166 Schnberg CHL, Suwa R, Hidaka M (in press) Sponge and coral zooxanthellae in heat and light preliminary results of photochemical efficiency monitored with pulse amplitude modulated fluorometry. Mar Ecol Schnberg CHL, Wilkinson CR (2001) Induced colonization of Great Barrier Reef corals by a clionid bioeroding sponge. Coral Reefs 20: 69-76 Schreiber U, Walz H, Kolbowski J (2003) Propagation of spatial variations of chlorophyll fluorescence parameters in dandelion leaves induced by spot laser heating. PAM News 1: 1-18

Thiele J (1900) Kieselschwmme von Ternate 1. Abh Senckenb naturf Ges 25: 19-80 Vacelet J (1981) Algal-sponge symbioses in the coral reefs of New Caledonia: a morphological study. Proc 4th Int Coral Reef Symp. Manila 2: 713-719 Vicente VP (1990) Response of sponges with autotrophic endosymbionts during the coral-bleaching episode in Puerto Rico. Coral Reefs 8: 199-202 Wellburn AR (1994) The spectral determination of chlorophylls a and b, as well as total carotenoids, using various solvents with spectrophotometers of different resolution. J Plant Physiol 144: 307-313

Porifera research: Biodiversity, innovation and sustainaBility - 2007

581

Silicateins, silicase and spicule-associated proteins: synthesis of demosponge silica skeleton and nanobiotechnological applications
Heinz C. Schrder(1*), Anatoli Krasko(1), David Brandt(1), Matthias Wiens(1), Muhammad Nawaz Tahir(2), Wolfgang Tremel(2), Werner E.G. Mller(1)
Institut fr Physiologische Chemie, Abteilung Angewandte Molekularbiologie, Universitt, Duesbergweg 6, D-55099 Mainz, Germany. hschroed@uni-mainz.de, krasko@uni-mainz.de, brandtd@uni-mainz.de, wiens@uni-mainz.de, wmueller@uni-mainz.de (2) Institut fr Anorganische Chemie und Analytische Chemie, Universitt, Duesbergweg 10-14, D-55099 Mainz, Germany. tahir@uni-mainz.de, tremel@uni-mainz.de
(1)

Abstract: The major skeletal elements of the Demospongiae and Hexactinellida are spicules formed from amorphous silica (biosilica). Demosponges are unique in their ability to synthesize their silica skeleton enzymatically. In the past few years, we have cloned several isoforms of the silica-forming enzyme, silicatein, from both marine sponges (example: Suberites domuncula) and freshwater sponges (example: Lubomirskia baikalensis). The silicateins are very similar to cathepsins, a family of cysteine proteases which do not precipitate silica. In the silicatein sequence, the cysteine residue of the catalytic triad of these proteases is replaced by a serine residue which is essential for the catalytic mechanism of the enzyme. In addition, a hydroxy amino acid (serine) cluster is present in the molecule. Silicatein undergoes posttranslational modifications of the protein, e.g. phosphorylation, as revealed by 2-dimensional gel electrophoresis. Using primmorphs (a special form of 3-dimensional cell aggregates) we established that spicule formation begins intracellularly and is completed extracellularly. Immunoblotting and immunogold labeling experiments revealed that silicatein is not only present in the axial filament but also at the surface of the spicules. Applying the technique of differential display of transcripts, we identified a further enzyme involved in silica metabolism, the silicase, which is able to depolymerize amorphous silica. The silicase shares highest similarity to the carbonic anhydrases, a family of zinc metal enzymes. The recombinant sponge enzymes, silicatein and silicase allow the synthesis/degradation of silica under ambient conditions that do not damage biomolecules. They are key enzymes for a variety of potential applications in (nano)biotechnology and medicine, including (i) surface modification of biomaterials, (ii) encapsulation of biomolecules and (iii) biofabrication of nanostructure materials for opto- and micro-electronics. Keywords: biosilica, nanobiotechnology, silicase, silicatein, spicules

Introduction
The biomineral constituting the spicules of the inorganic skeleton of two classes of sponges (Porifera), the Demospongiae and the Hexactinellida, is amorphous silica, while the third class of sponges, the Calcarea, has a skeleton made of calcium carbonate. The siliceous spicules are the main stabilizing inorganic elements in the body of demosponges and hexactinellid sponges. They have a characteristic, speciesspecific morphology. The marine demosponge Suberites domuncula, which we use as a model to study spicule formation (reviewed in: Mller et al. 2006b), has a skeleton which is composed of only two types of megascleres, monactinal tylostyles (main fraction) and diactinal oxeas (smaller fraction). The spicules reach lengths of up to 450 m and diameters of 5 to 7 m (Fig. 1A). The tylostyles have one pointed end and one swollen knob (Fig. 1B), while the two ends of the oxeas are pointed. In their centre, the spicules have a 0.3 to 1.6 m wide axial

canal (Fig. 2A), around which the silica is deposited under formation of concentric ~0.3 to 1 m thick layers. Growth of spicules through apposition of lamellar silica layers is particularly prominent in hexactinellid sponges (Fig. 2B). The inorganic phase of the siliceous spicules contains a relatively high content of water (6-13%) which is largely present in a mobile form, as revealed by high resolution magnetic resonance microimaging studies (Mller et al. 2006c). The formation of the spicules is a very rapid process and can be completed within a few days; for example, the spicules of the freshwater sponge Ephydatia fluviatilis (length: 100 - 300 m) are formed within 40 h (Weissenfels 1989). The synthesis of the spicules occurs in specialized cells, the sclerocytes (Simpson 1984). To study spicule formation we use a 3-dimensional cell culture system that has been established for S. domuncula (Custdio et al. 1998, Mller et al. 1999) and many other sponges including freshwater sponges (Mller et al. 2007a).

582

Fig. 1: A. Spicules, oxeae and tylostyles from the demosponge Suberites domuncula. SEM analysis. B. Higher magnification. Diameter of spicules 5 to 7 m.

Fig. 2: Cross-sections through spicules from (A) the demosponge Suberites domuncula (TEM analysis) and (B) the hexactinellid sponge Hyalonema sieboldi (SEM analysis). ac, axial canal; af, axial filament. Bar: 2 m (A) and 80 m (B).

This primmorph system allows the investigation of the different phases of spicule formation (Mller et al. 2005b). Primmorphs are a special type of cell aggregates which are formed from dissociated, sponge single cells under suitable conditions. The cells of primmorphs have a high proliferation and differentiation capacity (Zhang et al. 2003). Applying the primmorph system we could establish that the initial steps of spicule formation occur intracellularly in the sclerocytes (Mller et al. 2005b, 2006b).

Synthesis of silica: silicatein

The discovery of the key enzyme involved in spiculogenesis was a major breakthrough in the understanding of spicule formation. The axial canal of the spicules harbors an organic filament, the axial filament (Uriz et al. 2000). Morse and his coworkers (Shimizu et al. 1998, Cha et al. 1999) discovered that this axial filament is composed of a cathepsin Lrelated enzyme, which they termed silicatein. These results (reviewed in: Weaver and Morse 2003) and later studies using the enzymatically active recombinant silicatein (Tahir et al. 2004, Schrder et al. 2006, reviewed in: Mller et al. 2007c) demonstrated that silica formation in sponges is an enzymatic process, in contrast to silica formation in diatoms which is mediated by polyamines (Krger et al. 2000) and/ or polycationic peptides (silaffins) modified at their lysine residues by methylpropylamine units (Krger et al. 1999, 2001). Silicatein catalyzes the synthesis of silica from monomeric silica precursors such as the synthetic silicon alkoxide, tetraethoxysilane (TEOS) which is the preferred substrate under experimental conditions. The first silicatein cDNA has been cloned from the marine demosponge Tethya aurantium (Cha et al. 1999, Mller et al. 2004b); two isoforms of silicateins (silicatein- and silicatein-) have been characterized. In subsequent years the genes/cDNAs encoding silicatein- and silicatein- have also been cloned from other sponges, including the marine sponges, S. domuncula (Krasko et al. 2000, 2002, Schrder et al. 2004b, 2005b, Mller et al. 2006b) and Petrosia ficiformis (Pozzolini et al. 2004), and the freshwater sponges E. fluviatilis (Funayama et al. 2005) and Lubomirskia baikalensis (Kaluzhnaya et al. 2005). In the latter sponge, four isoforms of silicatein- (silicatein a1 to a4) have

been identified (Kaluzhnaya et al. 2005, Wiens et al. 2006). These four silicateins are phylogenetically closely related (see below), suggesting their emergence by gene duplication (Wiens et al. 2006). Analysis of the silicatein- and genes isolated from S. domuncula revealed that they consist of six exons (Mller et al. 2003a, Schrder et al. 2005b). Silicateins can be purified from isolated axial filaments. The common procedure involves the treatment of the isolated spicules with hydrofluoric acid (HF). Thereby the axial filaments are freed from the surrounding silica, and can be subsequently visualized by staining with Coomassie brilliant blue (Mller et al. 2006c). The time course of this process can be followed if the dye is present in the HF solution (Fig. 3). As outlined before silicateins are related to the cathepsin family of proteases; the highest similarity is shared with cathepsin L (Shimizu et al. 1998, Krasko et al. 2000, Mller et al. 2003b). Cathepsins do not precipitate silica. The alignment of the deduced amino acid sequence of silicatein with the silicatein- sequence and the closely related

Fig. 3. Dissolution of spicules from the freshwater sponge Lubomirskia baikalensis with hydrofluoric acid in time lapse images. After 10 min, the silica is completely dissolved leaving behind only the organic axial filament, which was stained simultaneously with the liberation using Coomassie brilliant blue.

583

cathepsin L sequence from S. domuncula is shown in Fig. 4. Prominent sites within the silicatein sequence are the catalytic triad (silicatein-: Ser aa138, His aa277, and Asn aa279) and the hydroxy amino acid (serine) cluster (aa267 to aa274; Shimizu et al. 1998, Krasko et al. 2000). The serine cluster is thought to act as a template for biosilica deposition.

Silicatein: post-translational processing

Protein chemical studies as well as analyses of the silicatein genes/cDNAs revealed that silicateins are synthesized as proenzymes which undergo two processing steps to form the mature, active enzymes (Mller et al. 2003a, 2005b). The predicted cleavage site of the S. domuncula silicatein proenzyme is at aa112/aa113 (GlnAsp; based on comparisons with cathepsins; see Fig. 4). In axial filaments only the mature 23-kDa form of silicatein - without the signal peptide and the propeptide - exists (Mller et al. 2005b). In addition, silicatein undergoes further post-translational modification steps (Mller et al. 2005b). Five phosphoisoforms have been identified (Mller et al. 2005b). The sizes of these five phospho-isoforms of silicatein are compatible with the predicted size of the mature form of the protein. Interestingly, in 2-dimensional gels each of these five silicatein phospho-isoforms appears as two or three spots (Mller et al. 2007c). Analysis of the spots by electrospray ionization-mass spectrometry revealed that silicatein also underlies posttranslational modification by de-hydroxylation from tyrosine to phenylalanine (Mller et al. 2007c).

modeling; Fig. 5A), this serine residue, Ser aa138, as well as His aa277,and Asn aa279 are part of the active site pocket of the enzyme; the serine residues of the hydroxyl amino acid cluster are present at the surface of the molecule. Silicatein catalyzes the formation of amorphous silica from organosilicon compounds; TEOS is the most commonly used substrate. The proposed mechanism of the reaction comprises two steps (Cha et al. 1999); step 1 is the (ratelimiting) hydrolysis of the alkoxide substrate (Fig. 6) and step 2 is the subsequent polymerization of the resulting silanol compounds under formation of amorphous silica. Silicatein is also able to promote the condensation of organically modified polysiloxanes (silicones; Cha et al. 1999; to be published). The hydroxyl group of the serine residue and the imidazole group of the histidine residue in the active site of the enzyme are the crucial moieties involved in the catalytic mechanism of the silicatein molecule (Zhou et al. 1999). The nucleophilicity of the hydroxyl group at the serine residue is thought to be increased by formation of a hydrogen bond to the imidazole nitrogen, facilitating the nucleophilic attack of the hydroxyl group on the silicon of the alkoxide substrate (Fig. 6; Cha et al. 1999, Zhou et al. 1999). The transitory covalent linkage between the enzyme and the substrate is then hydrolyzed by water. The reactive silanol molecules generated by hydrolysis subsequently undergo condensation reactions.

Silicatein: phylogenetic analysis

The result of a phylogenetic analysis of the hitherto characterized silicateins from the marine sponges T. aurantium and S. domuncula, as well as the freshwater sponges L. baikalensis, Spongilla lacustris and E. fluviatilis with the cathepsins L from these demosponges and the hexactinellid sponge Aphrocallistes vastus and with the papain cysteine peptidase from the plant Arabidopsis thaliana as outgroup is shown in Fig. 7. The calculated, slanted tree shows that the cathepsin L sequences form the basic branches from which the silicatein sequences originate. This result suggests that the silicateins derive from a common ancestor of the cathepsin L

Silicatein: catalytic mechanism

The catalytic center of the silicateins differs from the catalytic center of cathepsins in the presence of a serine residue instead of a cysteine residue present in the protease molecules (Fig. 4; Krasko et al. 1997); this serine residue is thought to be essential for the catalytic mechanism of silicatein (Shimizu et al. 1998, Krasko et al. 2000). In the presumptive 3D structure of silicatein (obtained by computer

Fig. 4: Silica-synthesizing enzyme: silicatein. Alignment of the deduced polypeptide sequences of silicatein- und silicatein- from S. domuncula (SILCAa_SUBDO and SILCAb_SUBDO) with cathepsin L from S. domuncula (CATL_SUBDO). Residues conserved (similar or related with respect to their physico-chemical properties) in all sequences are shown in white on black and those in at least two sequences in black on gray. The characteristic sites in the sequences are marked (); the catalytic triad amino acids, Ser in silicateins [Cys in cathepsins] and His and Asn, and the processing sites for the conversion of the proenzyme to the mature enzyme (PRO [ ]), the signal peptide (SIGNAL { }), and the serine cluster [ Ser ].

584

Fig. 5: 3-Dimensional modeling of silicatein- (A) and silicase (B). The published X-ray structures of cathepsin S and carbonic anhydrase were used as a reference for modeling the silicatein- and silicase structure, respectively. The active center of silicatein, comprising the amino acids Ser, His and Asn (yellow) and the localization of the Ser residues (green) of the serine cluster are marked. The silicase model shows the three histidine residues (red), which bind the zinc ion, in the catalytic center of the enzyme. Secondary structure elements are marked in red (-helices) or blue (-strands).

Fig. 6: Silicatein. Proposed mechanism of action (modified after Cha et al. 1999). The rate-limiting step (hydrolysis of the alkoxide substate) is shown.

sequences from the marine hexactinellid sponge A. vastus and the marine demosponges (here: S. domuncula). In addition, this tree indicates that the silicateins from the cosmopolitan species S. lacustris and E. fluviatilis form the basal branch from which the silicateins of the Baikalian endemic sponge L. baikalensis emerge. The result of this analysis also supports the view that the freshwater sponges evolved later than the marine sponges (see also Mller et al. 2006a, 2006d).

Degradation of silica: silicase

Besides the silica-anabolic enzymes, the silicateins, another enzyme, termed silicase, has been identified in the marine sponge S. domuncula. Silicase is able to depolymerize amorphous silica (Schrder et al. 2003). The expression of the gene encoding this silica-catabolic enzyme is strongly upregulated in response to higher concentrations of silicon, like the expression of silicatein (Krasko et al. 2000) and collagen (Krasko et al. 2000, Schrder et al. 2000). The silicase cDNA has been identified in primmorphs from S. domuncula, applying the technique of differential display of transcripts. The deduced polypeptide is closely related to the carbonic anhydrases. Most of the amino acids which are characteristic for the eukaryotic-type carbonic anhydrase signature are also present in the sponge silicase (Fig. 8; Schrder et al. 2003). Carbonic anhydrases are a family of zinc metal enzymes (Sly and Hu 1995). The three conserved histidine residues which are characteristic for these enzymes are also found in the deduced sponge protein at aa181, aa183 and aa206 (Fig. 8). These histidine residues bind a zinc ion. The structure of the S. domuncula silicase obtained by computer modeling is shown in Fig. 5B. The proposed mode of action of the silicase (depolymerisation of amorphous silica) is shown in Fig. 9. It is assumed that the reaction mechanism of the sponge enzyme is analogous to that of other zinc-dependent enzymes involved in ester hydrolysis. The zinc ion (= Lewis acid) interacts
Fig. 7: Phylogenetic relationship of the silicateins. Four deduced silicatein sequences of the isoform silicatein- (-1, -2, -3 and -4) from L. baikalensis (SILICAa1_LUBAI; SILICAa2_LUBAI; SILICAa3_LUBAI; SILICAa4_LUBAI) and the two cathepsin L sequences (CATL1_LUBAI; CATL2_LUBAI) were aligned with silicatein- from S. domuncula (SILICAa_SUBDO) from T. aurantium (SILICAa_TETHYA) and with the -isoenzymes from S. domuncula (SILICAb_SUBDO) and T. aurantium (SILICAb_ TETYHA), as well as with the cathepsin L sequences from S. domuncula (CATL_SUBDO), G. cydonium (CATL_GEOCY) and Aphrocallistes vastus (CATL_APHRVAS) and the related papain-like cysteine peptidase XBCP3 from Arabidopsis thaliana (PAPAIN_ARATH) [outgroup]. Additionally the deduced silicateins from the cosmopolitan freshwater sponges Ephydatia fluviatilis (SILCA1_EPHYDAT and SILCA2_EPHYDAT) and Spongilla lacustris (SILCA_SPONGILLA) are included in this analysis. The numbers at the nodes are an indication of the level of confidence for the branches as determined by bootstrap analysis (1000 bootstrap replicates; modified after Mller et al. 2007a).

with water (= Lewis base). The hydroxide ion formed by splitting of the water molecule is bound to the zinc ion. This hydroxide ion then performs a nucleophilic attack at one of the silicon atoms of the polymeric silicate. In the next step the zinc-complex binds to the silicon under cleavage of the

585

Fig. 8: Silicase. Alignment of the silicase from S. domuncula (SIA_SUBDO) with the human carbonic anhydrase II (CAH2_HUMAN). The carbonic anhydrase domain is framed ( e-CAdom ). The similar amino acid residues in both sequences are shown in white on black. The three zinc-binding histidine residues () and the characteristic amino acids forming the eukaryotic-type carbonic anhydrase signature (, found in both sequences; , present only in the carbonic anhydrases but not in the silicase) are indicated; +, residues forming the active-site hydrogen network.

oxygen bond in the polymeric silicate. The transiently formed zinc-bound silicate is then hydrolyzed by water, resulting in the release of the silicic acid product and regeneration of the zinc-bound hydroxide. Based on its ability to dissolve or to etch silica substrates, the silicase is of interest for a wide range of applications in nanobiotechnology (Mller et al. 2005c); studies to use this enzyme in soft lithography (Pisignano et al. 2005) are going on.

Silicatein-associated proteins
Analysis of the protein composition of native axial filaments revealed a series of further silicatein-associated proteins. To isolate these proteins, spicules were subjected to a novel extraction procedure. Instead of dissolving the spicules with HF, purified spicules were pulverized and extracted with a lysis-buffer (Tris-buffered saline, pH 7.5) containing of 4 M urea 1 mM EDTA, 1% Nonidet-P40, and a protease inhibitor cocktail (Schrder et al. 2006). The dominant protein in extracts from S. domuncula was the 24-kDa silicatein(s). The second major band found by SDS PAGE analysis of extracts from spicules corresponded to a Mr of 35 kDa; this protein was identified as galectin-2 (see below). In addition, a strong protein band of a size of ~250 kDa could be identified, which represents collagen. Finally, a 14-kDa protein was identified in the spicules, which displays sequence similarity to selenoprotein M (see below; Mller et al. 2005a). Galectin-2. Applying the methods of differential display of transcripts the cDNA coding for a galectin-2 has been identified in S. domuncula (Schrder et al. 2006). Galectin-2 has two galactose-binding sites (Fig. 10). This lectin forms aggregates in the presence of Ca2+ ions, to which silicatein molecules bind (Schrder et al. 2006). This interaction is most likely mediated by a highly hydrophobic stretch which is present at the C-terminus of galectin-2. Competition experiments revealed that the interaction between galectin-2 and silicatein- is abolished in the presence of a synthetic oligopeptide corresponding to the hydrophobic C-terminal stretch of galectin-2 (Schrder et al. 2006). Selenoprotein M. The deduced protein sequence of selenoprotein M (14 kDa) from S. domuncula shows the characteristic features known from other metazoan selenoproteins

Fig. 9: Silicase. Proposed mechanism of action (modified after Schrder et al. 2003).

(Fig. 11). The expression of this protein was found to be upregulated following exposure of primmorphs to selenium, as demonstrated by differential display (Mller et al. 2005a). Silicatein-associated protein. The expression of this 26-kDa protein is also induced by selenium but, unlike selenoprotein M, this polypeptide represents a new (sponge-specific) protein (Mller et al. 2005a). The secondary structure of the deduced polypeptide sequence of silicatein-associated protein shows a high structural regularity. The protein comprises six repeats of 20 amino acids and 10 highly similar hydrophobic regions consisting of ~9 amino acids. These hydrophobic regions comprise charged amino acids (glutamine) at their Nterminus, and proline and the hydroxy amino acids serine and threonine at the C-terminal ends (Fig. 12). At present the function of the selenium-dependent silicateinassociated proteins is not well understood. Antibodies against recombinant selenoprotein M and silicatein-associated protein were found to stain both the axial filament as well as

586

Fig. 10: Galectin-2. The two S. domuncula galectins [GALEC1_SUBDO and GALEC2_SUBDO (in white letters on black)] were aligned with the G. cydonium galectin-1 (GALEC1_GEOCY) and the human galectin-8 isoform b (GALEC8b_HUMAN). In addition, the two segments within GALEC2_SUBDO (in white letters on gray) and GALEC8b_HUMAN, spanning the galactose-binding domains have been included. In S. domuncula galectin-2, the parts ranged from aa1 to aa117 and from aa118 to aa293 (designated GALEC2a_SUBDO and GALEC2b_SUBDO) and in the human galectin-8b from aa1 to aa150 and from aa151 to aa316 (GALEC8ba_HUMAN and GALEC8bb_ HUMAN). These domains are marked Gal-binding-1 and Gal-binding-2. In addition, the N-terminus of the mature galectin-2 (+::::) and the hydrophobic terminus (hydrophobic) at the C-terminus of this protein are indicated. Amino acids that are similar among all sequences are in reversed type, and those conserved in at least two sequences are shaded.

Fig. 11: Selenoprotein M. From the S. domuncula nucleotides sequence (SDSelM), selenoprotein M (SelM_SUBDO) is predicted and aligned with the human selenoprotein M precursor (SelM_HUMAN); the human sequence was shortened between amino acids 2035 and 119124, indicated by square brackets [ ]). Residues conserved (similar or related with respect to their physicochemical properties) in the two sequences are shown in white on black. The TGA triplet that encodes selenocysteine (U) is underlined.

587

Fig. 12: Spicule-associated protein. The deduced protein SPIaP_SUBDO was analyzed for the predicted secondary structure (Mller et al. 2005a); the helical conformation (X), the extended conformation (), the turn (>) and the coil conformation (w) are indicated. The six highly similar segments of 20 amino acids are marked in white on black or are underlined. In addition, the 10 hydrophobic regions, present in the six 20-amino-acid blocks, are indicated and numbered (#).

the surface of the spicules (Mller et al. 2005a). We assume that these proteins are active in the structural/morphological organization of the silicatein-mediated synthesis of biosilica. Based on the complexity of the process of biosilica formation it is likely that besides silicatein, spicule formation requires additional (regulatory) proteins. Electron microscopical studies including immunogold labeling experiments presented first insights in the possible function of galectin-2 in spicule formation. The results indicate that galectin-2, together with collagen, provides the structural matrix for the assembly of silicatein molecules (Schrder et al. 2006; see below). Based on these findings novel strategies for the design of artificial spicule-like elements based on silicatein-mediated catalysis of metal oxide formation have been developed for potential applications in nanobiotechnology (to be published).

Spicule formation: morphological aspects

Primmorphs (S. domuncula) were used as a model to investigate spicule formation. Electron microscopical studies demonstrated that the synthesis of the spicules and the formation of the first silica layer around the axial filament start within the sclerocytes (Fig. 13A-C). The small (up to 10 m long) spicules are then extruded by the cells. In the extracellular space, the spicules grow up to 450 m in length with a diameter of 5 m. In the initial stage, the 1.6-m wide axial canal is primarily filled with the axial filament and additional membrane structures (Fig. 13E). In the final stage, it is almost completely filled with the axial filament, which displays the characteristic triangular form (Fig. 13F). Immunofluorescence studies revealed that silicatein exists both in the axial canal and on the surface of the spicules (Mller et al. 2005b, 2006b, Belikov et al. 2005). To further understand the synthesis of spicules, immunogold labeling/ TEM experiments were performed (Mller et al. 2005b). Using (gold-labeled) antibodies against silicatein an intense labeling of both the outer surface of the (mature) spicules and the inner surface toward the axial canal was found (Fig. 13I). In addition, many clusters of gold particles were seen in the axial filament which is mainly composed of silicatein

(Fig. 13J). Even more interesting were the results obtained by fine structure analysis of immature, growing spicules in the extracellular space (Fig. 13G and H). At first concentric rings which are 0.2 to 0.5 m apart are seen around the forming spicules. Subsequently, the inner rings fuse and electron dense clods become visible. Later, during maturation, the number of the concentric rings increases from two to ten rings with a total diameter 4-6 m. These results demonstrate that the spicules grow through apposition of lamellar silica layers. To get insights in the organizing principles that determine the specific morphology of the spicules and their arrangement within the sponge body, studies using specific antibodies and high resolution SEM analyses were performed. The results revealed that the spicules in demosponges (S. domuncula) are surrounded by collagen fibers forming an ordered network (Schrder et al. 2006, Eckert et al. 2006). We conclude that biosilica deposition catalyzed by silicatein is matrix-guided by (i) galectin (formation of aggregates and complexes with silicatein) and (ii) collagen (determination of spicule shape). Based on our studies, the process of spicule formation can be divided into three phases: Phase 1: Intracellular phase (sclerocytes). Silicatein is synthesized as a pro-enzyme (36.3 kDa) and processed via the 34.7-kDa form to the 23-kDa mature enzyme. In addition, the protein undergoes post-translational modification (phosphorylation and possibly further modifications). These modifications may occur during the transport of the protein through the endoplasmic reticulum and the Golgi complex. Finally it is transported into vesicles where it forms the axial filaments. Thereafter the first layer of silica is formed. Finally the spicules are released into the extracellular space. Phase 2: Extracellular phase (appositional growth). In the extracellular space the spicules grow in length and diameter by appositional growth. The silica deposition occurs in two directions; (i) centrifugal - from the axial canal to the surface - and (ii) centripedal - from the mesohyl to the surface of the spicule. The immunogold electron microscopical analysis showed that (i) silicatein is present also in the extracellular space and (ii) the silicatein molecules are arranged along strings, which are organized in parallel to the surfaces of the spicules (Schrder et al. 2006). In the presence of Ca2+, silica-

588

Fig. 13: Formation of spicules in primmorphs from S. domuncula. A. Section through a primmorph, showing the formation of an axial filament (af), a process that proceeds intracellularly; TEM analysis. B. The 15 m large sclerocytes produce one to three of up to 6 m long spicules (sp). C. At higher magnification, 15nm round fibrils (fi) adjacent to the axial filament (af) become visible. D F. Maturation of the axial filament in spicule from primmorphs; TEM analysis. D. Initial stage of spicule (sp) growth showing the electro-dense homogenous axial filament (af). E. At a later stage the spicules are filled with membrane structures and a small axial filament. F. Finally, the triangular axial filament appears as a homogenous core that does not completely fill the canal. G and H. Immunogold electron microscopy of cross sections through growing spicules in primmorphs. Distribution was detected by the secondary antibody labeled with nanogold. The band-like concentric rings are regularly arranged around the surface of the spicule/axial canal. G. The inner rings fuse and electron dense linear clods are formed. H. Finally, the silica layer grows and the number of rings increases. I. At higher magnification, the association with the inner and outer surface of the spicule (open arrowheads) is notable. J. Strong immunogold staining of the axial filament in the axial canal (ac).

tein associates with galectin-containing strings allowing the appositional growth of the spicules. Phase 3: Extracellular phase (shaping). The galectin-containing strings formed in phase 2 are organized by collagen fibers to net-like structures. These collagen fibers control the spatial arrangement of the silicatein/galectin-2 complexes (formation of concentric rings around the axis of a growing spicule). Hence, they provide the organized platform for the morphogenesis of the spicules. The prerequisite for spicule formation is the availability of sufficient amounts of silicic acid. Sponges live in an aqueous environment which is undersaturated with respect to silicic acid. Therefore they must accumulate silicic acid from the surrounding water. The uptake of silicic acid in sponge cells is an energy consuming process (Perovi-Ottstadt et al. 2005) which is mediated by a specific silicon transporter (Schrder et al. 2004a). In addition, silicon uptake by sponges may be facilitated by their high filtration rate (Vogel 1977).

Application of biosilica enzymes: silica nano(bio)technology

The silica-anabolic and silica-catabolic enzymes, silicatein (Mller et al. 2004a, 2007b) and silicase (Mller et al. 2005c), are of extreme interest for a variety of applications in nanobiotechnology. Silica-based materials are widely used in industry and medicine (reviewed in: Iler 1979). They are present in high-tech products such as microelectronics, optoelectronics, (bio)sensors and catalysts. Silica is also used for fabrication of glasses, ceramics, paints, adhesives, and separation materials, or as insulator in semiconductor devices. The technical (chemical) production of silica requires high temperature and pressure or extremes of pH, while silica formation in sponges occurs under mild, ambient conditions. Moreover, siliceous sponges are able to produce the building blocks of their skeletons with high fidelity and in

large copy number. Therefore, the mechanism(s) underlying biosilica formation in these organisms are of high interest for the design, in particular on the nano-scale, of novel biosilicas to be used in nano(bio)technology. Tremel and coworkers could show that the recombinant silicatein retains its biocatalytic activity after immobilization of the protein onto gold surfaces (Tahir et al. 2004) and selfassembled polymer layers (Tahir et al. 2005). The results revealed the formation of interconnected silica nanospheres with a diameter about 70-300 nm. Recombinant His-tagged silicatein was immobilised onto a gold surface which had been functionalized with nitrilotriacetic acid (NTA) alkanethiol

589

(Tahir et al. 2004). The recombinant protein bound through its His-tag affinity sequence to the chelator NTA alkanethiol through Ni2+ complex formation. Also the method of formation of self-assembled monolayers (SAM) was applied to bind silicatein to metal or metal oxide surfaces. The surfaces had been activated by cysteamine to allow the binding of a reactive ester polymer which specifically reacts with primary amines. This polymer then allowed the binding of -terminated NTA amine molecules. Finally, His-tagged silicatein was bound to the NTA anchor via Ni2+ complexation. Surprisingly, the immobilised silicatein was also able to catalyze, at neutral pH and room temperature, the synthesis of titania (TiO2) and zirconia (ZrO2), using monomeric alkoxide precursors (Tahir et al. 2005). The principle of construction of this selfassembled polymer layer resembles that of the concentric rings formed around the growing spicules, consisting of the silicatein/galectin-2 complex-containing strings and the surrounding collagen fibers. Thus nature can be used as a model for the design of new nano-scale objects for application in nanobiotechnology. Silica is a constituent of bioactive glasses. Such glasses are used as scaffolds in tissue engineering. We showed that mineralization of bone-forming cells (SaOS-2 cells) increases if the cells are grown on silicatein (biosilica)-modified culture plates (Schrder et al. 2005a). Biocatalytically formed silica may also be applied for preparation of coatings of metal implants used in surgery. The biosilica layer is thought to increase the biocompatibility of the implant. In addition, bioactive substances could be physically (by entrapment) or chemically (after surface functionalization) attached to the (otherwise relatively inert) biosilica-modified metal surface. Such materials are of extreme interest, e.g., in bone replacement and dentistry. Besides its hydrolytic activity, silicatein has a second, technologically relevant activity (Tahir et al. 2006); it catalyzes the reductive formation of colloidal gold nanoparticles which further aggregate to form gold nanocrystals. Tahir et al. (2006) reported a procedure to immobilize His-tagged silicatein onto TiO2 nanowires. The immobilized silicatein was then used to produce Au nanoparticles at the surface of the nanowires. Surface functionalization was achieved using a multifunctional polymeric ligand which contained (i) dopamine residues to attach the polymer onto the surface of the nanowire and (ii) a NTA linker to immobilize the Histagged silicatein (Fig. 14A). Immobilization of silicatein on the surface of the functionalized nanowire was demonstrated by immunofluorescence microscopy using antibodies against the bound protein (Fig. 14B). The morphology of the synthesized nanocomposite material consisting of TiO2 nanowires decorated with Au nanocrystals was studied by electron microscopy. The SEM image of an Au-decorated TiO2 nanowire is shown in Fig. 14C. The formed Au nanocrystals showed a triangular morphology (Fig. 14D). Our results show that the recombinant silicatein (and also the recombinant silicase) are of potential importance for a series of applications in nano(bio)technology. These applications comprise the use of the enzymes for preparation of surface coatings of metals, metal oxides, glasses and other materials (in particular biomaterials), the synthesis of nano-containers and nano-devices (e.g., nano-sieves) made of amorphous silica, and

Fig. 14: Formation of TiO2 nanowire/Au nanoparticle composites. A. Schematic presentation of the structure of the TiO2/Au nanocomposite. The TiO2 nanowire was functionalized using a multifunctional polymer ligand (red) which was bound to the nanowire by complexation through its catechol groups. His-tag silicatein molecules (brown/green) were attached to the NTA terminated side chains of the polymer by Ni2+ complexation. Au nanocrystals (yellow) were formed by reduction of tetrachloroauric acid by the sulfhydryl groups of immobilized silicatein. They are chemically bound to the amino groups at the outer surface of the protein. B. Immunodetection of silicatein molecules bound to the TiO2 nanowire. C. TEM image of the TiO2 nanowire decorated with Au nanocrystals. D. Magnified view of Au nanocrystals on the surface of the TiO2/Au nanocomposite.

the encapsulation of bioactive molecules (drugs, hormones, etc) allowing the controlled release of these substances (drug delivery). Silicatein-mediated biosilicification may also be an innovative approach in lithography and for the production of microelectronics (Mller et al. 2005c, Pisignano et al. 2005). This new technology allows, for the first time, to link via an enzyme (silicatein) - two worlds, the inorganic and the organic world, opening the development of new fabrication procedures previously thought to be impossible. Moreover, biosilica turned out to be of potential interest in (nano)photonics. Silica spicules have been demonstrated

590

Fig. 15: Sponge spicules as optical fibers. A. Experimental scheme. A stalk spicule (Sp) of the hexactinellid sponge Hyalonema sieboldi was immobilized within a canula (Ca) and illuminated with white light source (WLS) focused through a biconvex lens (L) onto one end of the spicule. The emitted light was recorded by a optical spectrum analyzer (A). B. Changes of the color to red during the course through the fiber (from left to right). C. Output end of the sponge fiber after illumination. The central core of the fiber is brighter due to the higher reflection caused by the organic axial filament (af).

to function as optical glass fibers with unique properties (Cattaneo-Vietti et al. 1996, Aizenberg et al. 2004, Mller et al. 2006e). Fig. 15 shows a spicule of the hexactinellid sponge Hyalonema sieboldi free-spaced coupled with a white light (laser) source. The stalk spicules of this sponge are composed of approximately 40 siliceous layers around the central axial filament (thickness, 1 m) and can reach 30 cm in length and a diameter of 300 m. Only light with wavelengths between 615 nm and 1310 nm can pass through the spicule fiber, while wavelengths <615 nm and >1310 nm are filtered out (Mller et al. 2006e). Therefore, these spicules act as sharp high- and low pass filters.

biosilica glass occurs under mild, physiological conditions (low temperature and pressure, near-neutral pH), whereas physical-chemical methods require the application of high temperatures and pressures, and the use of caustic chemicals. This is particularly advantageous for the fabrication of silica (and other metal oxide) coatings on organic (bio)material surfaces.

Acknowledgements
This work was supported by grants from the European Commission, the Deutsche Forschungsgemeinschaft, the Bundesministerium fr Bildung und Forschung Germany (project: Center of Excellence BIOTECmarin) and the International Human Frontier Science Program.

Conclusion
In summary, the sponge enzymes involved in biosilica metabolism, silicatein and silicase (protected by patents; silicatein: EP 1320624 and US 7,169,589 B2; silicase: DE 10246186), turned out to be of high interest and potential importance for a variety of medical and technical applications. Biosilica is a key material in nano(bio)technology. This technology, based on the application of enzymatically formed silica, is bioinspired, i.e. nature is used as a model to synthesize new materials based on amorphous silica (and other metal oxides) in the nano-scale. The unique advantage of this new technology compared to conventional procedures is the fact that silicatein-mediated (enzymatic) production of

References
Aizenberg J, Sundar V, Yablon AD, Weaver JC, Chen G (2004) Biological glass fibers: correlation between optical and structural properties. Proc Natl Acad Sci USA 101: 3358-3363 Belikov SI, Kaluzhnaya OV, Schrder HC, Krasko A, Mller IM, Mller WEG (2005) Expression of silicatein in spicules from the Baikalian sponge Lubomirskia baicalensis. Cell Biol Int 29: 943951

591

Bhattacharyya P, Vulcani BE (1980) Sodium-dependent silicate transport in the apochlorotic marine diatom Nitzschia alba. Proc Natl Acad Sci USA 77: 6386-6390 Cattaneo-Vietti R, Bavestrello G, Cerrano C, Sar A, Benatti U, Giovine M, Gaino E (1996) Optical fibres in an Antarctic sponge. Nature 383: 397-398 Cha JN, Shimizu K, Zhou Y, Christianssen SC, Chmelka BF, Stucky GD, Morse DE (1999) Silicatein filaments and subunits from a marine sponge direct the polymerization of silica and silicones in vitro. Proc Natl Acad Sci USA 96: 361-365 Custdio MR, Prokic I, Steffen R, Koziol C, Borojevic R, Brmmer F, Nickel M, Mller WEG (1998) Primmorphs generated from dissociated cells of the sponge Suberites domuncula: a model system for studies of cell proliferation and cell death. Mech Ageing Develop 105: 45-59 Eckert C, Schrder HC, Brandt D, Perovic-Ottstadt S, Mller WEG (2006) A histochemical and electron microscopic analysis of the spiculogenesis in the demosponge Suberites domuncula. J Histochem Cytochem 54: 1031-1040 Funayama N, Nakatsukasa M, Kuraku S, Takechi K, Dohi M, Iwabe N, Miyata T and Agata K (2005). Isolation of Ef silicatein and Ef lectin as molecular markers for sclerocytes and cells involved in innate immunity in the freshwater sponge Ephydatia fluviatilis. Zool Sci 22: 1113-1122 Hildebrandt M, Wetherbee R (2003) Components and control of silicification in diatoms. Prog Mol Subcell Biol 33: 11-57 Iler RK (1979) The chemistry of silica. John Wiley & Sons, New York Kaluzhnaya OV, Belikov SI, Schrder HC, Wiens M, Giovine M, Krasko A, Mller IM, Mller WEG (2005) Dynamics of skeleton formation in the Lake Baikal sponge Lubomirskia baicalensis. Part II. Molecular biological studies. Naturwissenschaften 92: 134-138 Krasko A, Gamulin V, Seack J, Steffen R, Schrder HC, Mller WEG (1997) Cathepsin, a major protease of the marine sponge Geodia cydonium: purification of the enzyme and molecular cloning of cDNA. Molec Mar Biol Biotechnol 6: 296-307 Krasko A, Lorenz B, Batel R, Schrder HC, Mller IM, Mller WEG (2000) Expression of silicatein and collagen genes in the marine sponge Suberites domuncula is controlled by silicate and myotrophin. Eur J Biochem 267: 4878-4887 Krasko A, Schrder HC, Batel R, Grebenjuk VA, Steffen R, Mller IM, Mller WEG (2002) Iron induces proliferation and morphogenesis in primmorphs from the marine sponge Suberites domuncula. DNA Cell Biol 21: 67-80 Krger N, Deutzmann R, Sumper M (1999) Polycationic peptides from diatom biosilica that direct silica nanosphere formation. Science 286: 1129-1132 Krger N, Deutzmann R, Bergsdorf C, Sumper M (2000) Speciesspecific polyamines from diatoms control silica morphology. Proc Natl Acad Sci USA 97: 14133-14138 Krger N, Deutzmann R, Sumper M (2001) Silica-precipitating peptides from diatoms, the chemical structure of silaffin-1a from Cylindotheca fusiformis. J Biol Chem 276: 26066-26070 Mller WEG, Wiens M, Batel R, Steffen R, Borojevic R, Custdio RM (1999) Establishment of a primary cell culture from a sponge: primmorphs from Suberites domuncula. Mar Ecol Progr Ser 178: 205-219 Mller WEG, Krasko A, Le Pennec G, Schrder HC (2003a) Biochemistry and cell biology of silica formation in sponges. Microsc Res Tech 62: 368-377

Mller WEG, Krasko A, Le Pennec G, Steffen R, Wiens M, Ammar MSA, Mller IM, Schrder HC (2003b) Molecular mechanism of spicule formation in the demosponge Suberites domuncula: silicatein - collagen - myotrophin. Prog Mol Subcell Biol 33: 195221 Mller WEG, Schrder HC, Lorenz B, Krasko A (2004a) European Patent No. EP1320624 Mller WEG, Wiens M, Adell T, Gamulin V, Schrder HC, Mller IM (2004b) Bauplan of Urmetazoa: basis of genetic complexity of Metazoa. Int Rev Cytol 235: 53-92 Mller WEG, Borejko A, Brandt D, Osinga R, Ushijima H, Hamer B, Krasko A, Xupeng C, Mller IM, Schrder HC (2005a) Selenium affects biosilica formation in the demosponge Suberites domuncula: Effect on gene expression and spicule formation. FEBS J 272: 3838-3852 Mller WEG, Rothenberger M, Boreiko A, Tremel W, Reiber A, Schrder HC (2005b) Formation of siliceous spicules in the marine demosponge Suberites domuncula. Cell Tissue Res 321: 285-297 Mller WEG, Schrder HC, Krasko A (2005c) German Patent No. DE10246186 Mller WEG, Belikov SI, Schrder HC (2006a) Science First Hand 3: 26-35 Mller WEG, Belikov SI, Tremel W, Perry CC, Gieskes WWC, Boreiko A, Schrder HC (2006b) Siliceous spicules in marine demosponges (example Suberites domuncula). Micron 37: 107120 Mller WEG, Kaluzhnaya OV, Belikov SI, Rothenberger M, Schrder HC, Reiber A, Kaandorp JA, Manz B, Mietchen D, Volke F (2006c) Magnetic resonance imaging of the siliceous skeleton of the demosponge Lubomirskia baicalensis. J Struct Biol 153: 3141. Mller WEG, Schrder HC, Wrede P, Kaluzhnaya OV, Belikov SI (2006d) Speciation of sponges in Baikal-Tuva region: an outline. J Zool Syst Evol Res 44: 105-117 Mller WEG, Wendt K, Geppert C, Wiens M, Reiber A, Schrder HC (2006e) Novel photoreception system in sponges? Unique transmission properties of the stalk spicules from the hexactinellid Hyalonema sieboldi. Biosens Bioelectron 21: 1149-1155 Mller WEG, Belikov SI, Kaluzhnaya OV, Perovi-Ottstadt S, Fattorusso E, Ushijima H, Krasko A, Schrder HC (2007a) Cold stress defense in the freshwater sponge Lubomirskia baicalensis: Role of okadaic acid produced by symbiotic dinoflagellates. FEBS J 274: 23-36 Mller WEG, Schrder HC, Lorenz B, Krasko A (2007b) United States Patent No. US 7,169,589 B2 Mller WEG, Wang X, Belikov SI, Tremel W, Schlomacher U, Natoli A, Brandt D, Boreiko A, Tahir MN, Mller IM, Schrder HC (2007c) Formation of siliceous spicules in demosponges: example Suberites domuncula. In: Buerlein E (ed). Handbook of Biomineralization, vol. 1. Biological aspects and structure formation. Wiley-VCH, Weinheim, pp 59-82 Perovi-Ottstadt S, Wiens M, Schrder HC, Batel R, Giovine M, Krasko A, Mller IM, Mller WEG (2005) Arginine kinase in the demosponge Suberites domuncula: regulation of its expression and catalytic activity by silicic acid. J Exp Biol 208: 637-646 Pisignano D, Maruccio G, Mele E, Persano L, Di Benedetto F, Cingolani R (2005) Polymer nanofibers by soft lithography. Appl Physics Lett 87: 123109 Pozzolini M, Sturla L, Cerrano C, Bavestrello G, Camardella L, Parodi AM, Raheli F, Benatti U, Mller WEG, Giovine M (2004)

592

Molecular cloning of silicatein gene from the marine sponge Petrosia ficiformis (Porifera, Demospongiae) and development of primmorphs as a model for biosilicification studies. Mar Biotechnol 6: 594-603 Schrder HC, Krasko A, Batel R, Skorokhod A, Pahler S, Kruse M, Mller IM, Mller WEG (2000) Stimulation of protein (collagen) synthesis in sponge cells by a cardiac myotrophin-related molecule from Suberites domuncula. FASEB J 14: 2022-2031 Schrder HC, Krasko A, Le Pennec G, Adell T, Wiens M, Hassanein H, Mller IM, Mller WEG (2003) Silicase, an enzyme which degrades biogenous amorphous silica: contribution to the metabolism of silica deposition in the demosponge Suberites domuncula. Prog Mol Subcell Biol 33: 250-268 Schrder HC, Perovi-Ottstadt S, Rothenberger M, Wiens M, Schwertner H, Batel R, Korzhev M, Mller IM, Mller WEG (2004a) Silica transport in the demosponge Suberites domuncula: fluorescence emission analysis using the PDMPO probe and cloning of a potential transporter. Biochem J 381: 665-673 Schrder HC, Perovi-Ottstadt S, Wiens M, Batel R, Mller IM, Mller WEG (2004b) Differentiation capacity of the epithelial cells in the sponge Suberites domuncula. Cell Tissue Res 316: 271280 Schrder HC, Boreiko O, Krasko A, Reiber A, Schwertner H, Mller WEG (2005a) Mineralisation of SaOS-2 cells on enzymatically (silicatein) modified bioactive osteoblast-stimulating surfaces. J Biomed Mater Res Part B: Appl Biomater 75B: 387-392 Schrder HC, Perovi-Ottstadt S, Grebenjuk VA, Engel S, Mller IM, Mller WEG (2005b) Biosilica formation in spicules of the sponge Suberites domuncula: Synchronous expression of a gene cluster. Genomics 85: 666-678 Schrder HC, Boreiko A, Korzhev M, Tahir MN, Tremel W, Eckert C, Ushijima H, Mller IM, Mller WEG (2006) Co-expression and functional interaction of silicatein with galectin: matrixguided formation of siliceous spicules in the marine demosponge Suberites domuncula. J Biol Chem 281: 12001-12009 Shimizu K, Cha J, Stucky GD, Morse DE (1998) Silicatein alpha: cathepsin L-like protein in sponge biosilica. Proc Natl Acad Sci USA 95: 6234-6238 Simpson TL (1984) The cell biology of sponges. Springer-Verlag, New York

Sly WS, Hu PY (1995) Human carbonic anhydrases and carbonic anhydrase deficiencies. Annu Rev Biochem 64: 375-401 Tahir MN, Thato P, Mller WEG, Schrder HC, Janshoff A, Zhang J, Huth J, Tremel W (2004) Monitoring the formation of biosilica catalysed by histidine-tagged silicatein. Chem Commun 2004: 2848-2849 Tahir MN, Thato P, Mller WEG, Schrder HC, Borejko A, Faiss S, Janshoff A, Huth J, Tremel W (2005) Formation of layered titania and zirconia catalysed by surface-bound silicatein. Chem Commun 28: 5533-5535 Tahir MN, Eberhardt M, Therese HA, Kolb U, Theato P, Mller WEG, Schrder HC, Tremel W (2006) From single molecules to nanoscopically structured functional materials: Au nanocrystal growth on TiO2 nanowires controlled by surface bound silicatein. Angew Chem Int Ed 45: 4803-4809 Uriz MJ, Turon X, Becerro MA (2000). Silica deposition in Demosponges: spiculogenesis in Crambe crambe. Cell Tissue Res 301: 299-309 Vogel S (1977) Current-induced flow through living sponges in nature. Proc Natl Acad Sci USA 74: 2069-2071 Weaver J, Morse DE (2003) Molecular biology of demosponge axial filaments and their roles in biosilification. Micr Res Techn 62: 356367 Weissenfels N (1989) Biologie und mikroskopische anatomie der Swasserschwmme (Spongillidae). Gustav Fischer Verlag, Stuttgart Wiens M, Belikov SI, Kaluzhnaya OV, Krasko A, Schrder HC, Perovic-Ottstadt S, Mller WEG (2006) Molecular control of serial module formation along the apical-basal axis in the sponge Lubomirskia baicalensis: silicateins, mannose-binding lectin and mago nashi. Dev Genes Evol 216: 229-242 Zhang XY, Cao XP, Zhang W, Yu XJ, Jin MF (2003). Primmorphs from archaeocytes-dominant cell population of the sponge Hymeniacidon perleve: Improved cell proliferation and spiculogenesis. Biotechnol Bioeng 84: 583-590 Zhou Y, Shimizu K, Cha JN, Stucky GD, Morse DE (1999) Efficient catalysis of polysiloxane synthesis by silicatein requires specific hydroxyl and imidazole functionalities. Angew Chemie Int Ed 38: 780-782

Porifera research: Biodiversity, innovation and sustainaBility - 2007

593

Redescription of the Brazilian endemic sponge Geodia glariosa (Demospongiae: Geodiidae), with new records on its geographic and bathymetric distribution
Carla M.M. da Silva(1), Meiryelen V. da Silva(2), Bruno Cosme(3)
Laboratrio de Biologia de Porifera, Departamento de Zoologia, Instituto de Biologia, Universidade Federal da Bahia, Rua Baro de Geremoabo, s/n, CEP 40170-290, Ondina, Salvador, BA, Brasil. carlamms@ufba.br (2) Laboratrio de Gentica Marinha, Departamento de Biologia Celular e Gentica, IBRAG, Universidade do Estado do Rio de Janeiro, Rio de Janeiro, RJ (3) Laboratrio de Porifera, Departamento de Invertebrados, Museu Nacional, Universidade Federal do Rio de Janeiro, RJ, Brasil
(1)

Abstract: Redescription of Geodia glariosa (Sollas), an endemic species of shallow waters on the Brazilian coast, is another step of a broad ongoing study on the Western Atlantic species of Geodia, which aims to upgrade the level of their descriptions of inner and outer morphology to elucidate intra- and interspecific variations. This article offers a complete redescription of the species via comparison of syntypes with samples collected on the coastal region of north-eastern, south-eastern and southern Brazil. The species is recorded for the first time for the coasts of the States of Pernambuco, Espirito Santo and So Paulo. New occurrences for the State of Bahia are also included. Two additional categories of spicules, rare choanosomal styles and plagiotriaenes, are recorded. A comparison with other Brazilian records is given. This paper expands the species bathymetry to the intertidal region (Praia da Pituba and Ilha do Medo, Bahia) and extends its northern limit to Recife, Pernambuco State coast (Praia da Boa Viagem, 080805S). Keywords: taxonomy, distribution, Astrophorida, Geodia, SEM analysis

Introduction
The Brazilian coasts poriferan fauna is still largely unknown, specially when compared to those of the Caribbean and Indo-Pacific regions. Foreign researchers and scientific expeditions, like those of the Challenger (Poljaeff 1884, Sollas 1886, 1888, Ridley and Dendy 1887) and the Calypso (Boury-Esnault 1973) made significant contributions to broaden this knowledge, by dredging our continental shelf, being responsible for about 73% of the records of poriferans belonging to Class Demospongiae known until the present time (Muricy and Hajdu 2006). The increasing efforts by Brazilian researchers in the last couple decades resulted in the recording of about 320 species, particularly for the Northeast, Southeast and South regions of Brazil. In Geodia, four new taxa were described (Hajdu et al. 1992, Silva and Mothes 2000), totalling nine species known from the Brazilian coast, five of which are provisionally endemic: Geodia glariosa (Sollas, 1886), G. tylastra BouryEsnault, 1973, G. splendida Silva and Mothes, 2000, G. riograndensis and G. australis Silva and Mothes, 2000, and four were also recorded for the Caribbean: G. gibberosa (Lamarck, 1815), G. neptuni (Sollas, 1888), G. papyracea Hechtel, 1983 and G. corticostylifera Hajdu et al., 1992.

Redescription of Geodia glariosa is part of a broad ongoing study (Silva and Mothes 2000, Silva 2002, Silva et al. 2004), which aims to describe in the same level of detail the Western Atlantic and Eastern Pacific species belonging to that genus, in regard to external and internal morphological characters (spicules and skeleton), to elucidate intra- and interspecific variations. This species was recorded for the Brazilian coast by Sollas (1886) as Cydonium glariosus based on two juvenile specimens (according to Volkmer-Ribeiro and Mothes-deMoraes, 1975). They were both collected between 13 and 46 m deep along the coast of the State of Bahia, without any accurate record of site of occurrence. In his subsequent work (Sollas 1888), the author recorded that this species of sponge had the peculiar trait of incorporating grains of sand, which are placed on the cortical region as an intermediary crust between the underliyng sterraster layer and the outer strongylaster layer. Topsent (1892) recorded Cydonium glareosum [nomen imperfectum (nomen correctum: Cydonium glariosus)] for the Gulf of Gascogne (Spain, Eastern Atlantic), from a depth of 248 m, without any description of the spicules or skeleton. Mello-Leito et al. (1961), in his inventory of

594

poriferans recorded for Brazil, quotes Sollas record (1886, 1888) as Cydonium gloriosus [nomen imperfectum (nomen correctum: Cydonium glariosus)]. The redefinition of Geodia and inclusion of genera Cydonium Fleming, 1828 and Pyxitis Schmidt, 1870 as its junior synonyms (von Lendenfeld 1903) resulted in ammending Sollas species name to Geodia glariosa, a name already included in a series of comparative studies of the Brazilian coasts sponge fauna (Hechtel 1976, Hajdu 1994, Silva and Mothes 2000). Volkmer-Ribeiro and Mothes-de-Moraes (1975) recorded G. glariosa (as Cydonium glariosus) for the coast of Santa Catarina, in depths lower than 13m, highlighting this sponges ability to incorporate dead diatom shells instead of grains of sand as seen in the specimens from Bahia (Sollas 1886, 1888), probably due to the shortage of sand grains on the intertidal rocky-shores where specimens occurred. This work aims to provide a complete redescription of Geodia glariosa based on comparison of the syntypes described by Sollas (1886, 1888) and the specimens recorded by Volkmer-Ribeiro and Mothes-de-Moraes (1975) with nine samples collected in different environments in north-eastern and south-eastern Brazil, rocky-shores and mangroves (intertidal) and rocky substrate or gravel bottom (sublitoral), and varying depths, 0-20m. Two additional categories of choanosomal megascleres, morphological variations on the ectosomal structure and new occurrences for the States of Pernambuco, Espirito Santo and So Paulo are recorded. The species bathymetry is increased to the intertidal region (Praia da Pituba and Ilha do Medo, Bahia) and its northern distribution limit is expanded until the coast of Pernambuco, Recife (Praia da Boa Viagem, 080805S).

Skeletal slides and dissociated spicules mounts follow Hajdu (1994). Scanning electron micrographs were taken according to Mothes and Silva (2002). Measures for single specimens are given in Table 1 (oxeas measures are length / width, triaenes measures are shaft length / shaft width / cladome length / clade length / clade width, and microscleres measures are total length / center diameter / rays length / rays width). The range of spicule sizes for the species as a whole are given in the text (megascleres measures are length / width and microcleres measures are total length). All measures are in m. Abbreviations used in the text: BMNH The Natural History Museum, London, England; MCNPOR Porifera Collection, Museu de Cincias Naturais, Fundao Zoobotnica do Rio Grande do Sul, Porto Alegre, Brazil; MNRJ Museu Nacional, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil; UFBA-POR Porifera Collection, Museu de Zoologia, Universidade Federal da Bahia, Salvador, BA, Brazil; UFRJPOR Porifera Collection, Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil; MZUSP Museu de Zoologia, Universidade de So Paulo, So Paulo, Brazil.

Results
Order Astrophorida Sollas, 1888 Family Geodiidae Gray, 1867 Genus Geodia Lamarck, 1815 Diagnosis: See Uriz (2002: 134) Geodia glariosa (Sollas, 1886) (Figs 1- 5, Table 1) Cydonium glariosus Sollas, 1886: 196 (Type-locality: off Bahia State coast); 1888: 223. Cydonium gloriosus; MelloLeito et al., 1961: 18. Cydonium glariosus; Volkmer-Ribeiro and Mothes-de-Moraes, 1975: 7. Geodia glariosa; Hajdu et al., 1992: 212; Silva and Mothes, 2000: 31. Material examined: Brazil: MZUSP unregistered (slides MCNPOR 3647): Pernambuco, Recife, Praia da Boa Viagem, M. Vannucci coll., 06/VI/1955 (det. de Laubenfels, 1956, as G. gibberosa); MCNPOR 2143: Praia da Boa Viagem, 080805S, 345519W, A. A. Lise coll., 04/XI/1974, 7 m depth; B. Mothes det, 07/I/1991; Syntype BMNH 1889.1.1.85, 1889.1.1.86 (slides MCNPOR 3634): Bahia, H.M.S. Challenger coll., IX/1873, 7-25 fathoms (12,8-45,7 m); UFBA-POR 1308: Bahia, Salvador, Praia da Pituba, 130100S, 382800W, A.V. Madeira coll., 5/VI/1993, intertidal (0-1 m); UFBA-POR 1235: Bahia, Ilha do Medo, 1253S, 3842W, F. Kelmo coll., 28/X/1992, intertidal (0-1 m); MNRJ 4322: Bahia, Una, Canavieiras, Station #5 shallow, 153408S, 384981W, Astro Garoupa coll. / Program REVIZEE CENTRAL V, 01/VII/2001, 20 m; UFRJPOR 1393 (MCNPOR 4036): Bahia, Porto Seguro, 1626S, 3905W, C.E.Z. coll., 27/VIII/1980; UFRJPOR 1364 (MCNPOR 4029): Bahia, northern reefs of Cumuruxatiba, 171197S, 391874W, M. N. Prado coll.,

Material and methods

Samples were collected from several coastal Brazilian states, between 0 and 50 m depth, by hand (intertidal), SCUBA diving, or dredging undertaken by expeditions such as those of CEZ - Zoological Studies Commission/UFRJ (Brazil) and H.M.S. Challenger (England). Collection sites of Geodia glariosa are shown in Fig. 1: #1. Praia da Boa Viagem, Recife, Pernambuco (080805S, 345519W); #2. Praia da Pituba, Salvador, Bahia (130100S, 382800 W); #3. Ilha do Medo, Baa de Todos os Santos, Bahia (1253S, 3842W); #4. Canavieiras, Una, Bahia, (153408S, 384981W); #5. Porto Seguro, Bahia (1626S, 3905W); #6. Northern reefs of Cumuruxatiba, Bahia (171197S, 391874W); #7. Santa Cruz, Esprito Santo (195713S, 400920W); #8. So Sebastio, So Paulo (234536S, 452435W); #9. off Imbituba coast, Santa Catarina (281603S, 484721W). Stations 1 to 8 are all first records for the Brazilian coast. The sponges are stored in the Marine Porifera Collection, Museu de Cincias Naturais, Fundao Zoobotnica do Rio Grande do Sul, Porto Alegre, Brazil; Porifera Laboratory of Museu Nacional/Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil and Porifera Collection of Museu de Zoologia, Universidade Federal da Bahia, Salvador, Brazil. The map (Fig. 1) was obtained from the webpage Online Map Creation (www.aquarius.geomar.de/omc).

595

Fig. 1: Map of South American coast, showing the collection sites of Geodia glariosa (Sollas, 1886) along the northeastern and southeastern Brazilian coast. Stations 1-8 = new locality records for the Brazilian coast; 9 = Volkmer-Ribeiro and Mothes-deMoraes (1975).

Fig. 2: Specimens of Geodia glariosa: A, B. MCNPOR 175, specimen from Santa Catarina State coast showing its association with Balanus sp. (Cirripedia); C. MCNPOR 3647, specimen from Pernambuco State coast; D. UFRJPOR 1393, specimen from Porto Seguro, Bahia State coast.

596

Table 1: Comparative data on spicular micrometries of syntype and additional material of Geodia glariosa (Sollas. 1888). Triaenes measures are shaft length/shaft width/cladome length/clade length/clade width. Microscleres measures are total diameter/center diameter/rays length/ rays width. Means are underlined. All measures are in m (n=50). n.r. = not referred; n.o. = not observed.
Material Syntype BMNH 1889.1.1.86 Syntype BMNH 1889.1.1.86 (remeasured) Not available MCNPOR 175 Reference Sollas (1886, 1888) (present article) Oxea I 1856.0/ 26 954.5-1372-1794/ 9.5-17.7-23.8 Oxea II 350.0-400/ 15.8 266-370.7-522.5/ 4.8-13.6-19 Orthotriaene 2856.0/ 51.6 1587-2021.6-2461/ 23-33.2-42.6/ 247-445.1-646/ 127.6-231.9-??/ 16.1-21.1-25.3 n.r. 1012-1525-2059/ n.r./ 143-267-430/ n.r. 1610-1825.4-2231/ 23-31.1-43.7/ 289.8-398.5-494/ 142.6-195.9-236.9/ 13.8-17.1-25.3 1610-1683-2553/ 24.2-32.7-39.1/ 304-395.7-456/ 170.2-209-241.5/ 13.8-17.5-23 n.r. 1118-1808.9-2340/ 30-42.9-60/ 240-397.1-500/ 109.6-179.7-239.4/ 23.9-36.1-53.2 1300-1736.4-2200/ 30-50.9-70/ 380-433.3-520/ 150-205.5-270/ 30-42.7-60 1780-1902-2100/ 40.45-50/ 320- 370- 400/ 160-184-200/ 30-37-50 1350-1781.3-2120/ 25-43.1-60/ 220-448.8-1000/ 110-240-550/ 20-31.3-40 1100-1545-2100/ 30-43.8-60/ 620-767.5-1000/ 250-375-550/ 15-22.5-30 1580.5-2180.7-2800/ 28.2-42.7-56.0/ 146.5-308/ 22.9-56 420-1747-2800/ 7-32.5-56/ 28-155.5-280/ 5.6-28.4-56 1404.8-1624.2-1944.4/ 20.4-27.2-30.5/ 285-345.6-447.9/ 91.6-181.7-223.9/ 10.2-19.4-20.4 Plagiotriaene n.r. 529-1360.8-2024/ 9.2-20.2-27.6/ 95-274.7-389.5/ 52.9-136-202.4/ 6.9-11.8-13.8 n.r. n.r.

Topsent (1892) Volkmer-Ribeiro and Mothes-deMoraes (1975) (present article)

n.r. 912-1200-1634/ 12-19-26 805-1103.1-1357/ 9.5-15.6-23.8

n.r. 370-437-500/ 10-19-28 276-345.7-437/ 9.5-13-19

MCNPOR 175 (remeasured)

MCNPOR 176

(present article)

503.5-862-1187.5/ 4.6-14-23

218.5-300.8-370.5/ 6.9-12.5-18.4

MZUSP w/n MZUSP w/n (MCNPOR 3647)

de Laubenfels (1956)

n.r. 780-1176.2-1460/ 10-23.5-31.9

n.r. 300-443.4-580/ 10.6-19-31.9

828-1246.1-1541/ 16.1-23.3-29.9/ 180.5-287.3-475/ 82.8-146.2-230/ 9.2-11.5-16.1 483-997.1-1495.0/ 11.5-19-27.6/ 85.5-242.9-427.5/ 48.3-127.3-213.9/ 4.6-9.9-14.9 n.r. n.o./ 21.3-21.3-1.3/ 111.7-111.7-111.7/ 58.5-58.5-58.5/ 18.6-18.6-18.6 735-1232.5-1515/ 12.2-25.5-43.7/ 325-526.9-790/ 43.7-64.5-85.1/ 10.9-26.7-36.5 750-798-900/ 27.9-30.7-34/ 325-381-490/ 158-189.5-211.4/ 21.9-30.1-42.5 1030-1375-5-1680/ 13.4-23.1-35.2/ 240-495.5-680/ 26.7-57.7-87.5/ 10.9-17.6-21.9 630-938.2-1590/ 12-26.5-41.3/ 4.6-25.8-46.7/ 27-15.1-28.7/ 1-2.5-3.9 420/ 7/ 28/ 5.6 780/ 18.5/ 38.2/ 17.5 / 4.5 1404.8-1624.2-1944.4/ 20.4-27.2-30.5/ 285-345.6-447.9/ 91.6-181.7-223.9/ 10.2-19.4-20.4

UFRJPOR 1393 (MCNPOR 4036)

(present article)

1044.2-1471-2240/ 20-29.9-37.2

65.6-96.1-33.7/ 3.6-4.9-7.3

UFRJPOR 1364 (MCNPOR 4029)

(present article)

1460-1658-1900/ 18.6-26.6-31.9

65.6- 95.5-137.7/ 2.4-4.8-133.7

UFRJPOR 189 (MCNPOR 4032)

(present article)

990-1278-1600/ 18.6-20.2-21.3

63.2-97.2-123.9/ 0.7-1.5-2.2

UFRJPOR 113 (MCNPOR 4030)

(present article)

1220-1464-1820/ 16-24.2-31.9

21.9-29.6-48.6/ 1-1.9-3.9

UFBA-POR 1308

(present article)

590-1161.5-1750/ 8-19.8-31 728-938.9-1134/ 8-14-21 590.4-897.4-1445.6/ 10.2-11.8-20.4

238-390.8-518/ 1.5-13.4-21 238-348.1-448/ 1.5-12.5-14 274.9-378.7-468.3/ 4.1-8.4-13.2

UFBA-POR 1235

(present article)

MNRJ 4322

(present article)

597

Table 1 (cont.) Protriaene 5355.0/ 29 1035-1924.3-3473/ 6.9-13.1-23/ 114-226.4-351.5/ 23-58.4-85.1/ 4.6-9.4-16.1 n.r. 4243-4871-6086/ n.r./ 36-47-57/ n.r. 1035-2349.8-3726/ 6.9-12.8-18.4/ 142.5-191.1-256.5/ 29.9-46.3-69/ 4.6-6.8-11.5 943-2714-4301/ 9.2-14-16.1/ 123.5-191-247/ 34.5-52.7-87.4/ 4.6-7.7-9.2 n.r. n.o./ 13.3-16.1-26.6/ 42.6-57.9-79.8/ 26.6-40.7-53.2/ 8-11.1-18.6 3725-3931.3-4250/ 20-23.8-30/ 50-65-80/ 30-32.5-40/ 10 1750-2487.5-3000/ 18-20.8-25/ 40-91.3-125/ 30-35-40/ 10-11.3-15 2000-2460-2750/ 20-22-25/ 40-52-70/ 30-36-50/ 10-11-15 1500-2150-3000/ 20-28-40/ 50-58-70/ 20-34-40/ 10-13-15 1834 - >2000 mm/ 11.2-17.87-28/ 14-22.81-42/ 7-9.89-14 >2000/ 11-14.1-21/ 14-14.7-28/ 7-11.9-14 1476.1-4021.1/ 10.1/ 20.4-40.7/ 5.1-7.1/ 30.5-50.9 Anatriaene 4641.0/ 11.8 897-2698.1-4094/ 6.9-12.8-20.7/ 50.6-79.6-117.3/ 23-35-55.2/ 9.2-10.5-13.8 n.r. 3782-4527-5319/ n.r./ 53-86-112/ n.r. / n.r. 1357-2909.5-4117/ 6.9-11.6-16.1/ 34.5-71.3-105.8/ 20.7-40-59.8/ 6.9-12.6-18.4 1610-2282.8-2737/ 6.9-9.2-11.5/ 43.7-51.2-59.8/ 32.2-41.4-46/ 11.5-12.7-16.1 n.r. -/ 21.3-21.3-21.3/ 93.1-99.8-106.4/ 53.2-53.2-53.2/ 18.6-20-21.3 1625-2475-2975/ 18-21.5-24/ 40-67-87/ 20-35.3-46/ 10-11.3-15 1675-2487.5-2950/ 19-21.8-26/ 90-108.8-130/ 48-68.3-88/ 10-12.5-15 1800-2512.5-2900/ 18-20.5-24/ 40-66.5-86/ 26-38.3-52/ 10-13.5-16 1675-2456.3-2975/ 18-20.5-25/ 43-68-86/ 28-37.5-49/ 10-13-15 >2000 length/ n.o./ n.o./ n.o. > 2000/ 14-22.7-28/ 14-28-42/ 11-14.6-17 n.o. Sterraster 51.6-58 46-51.8-57.5 Oxyaster 16.0-19/ n.r. / 8.0 9.2-15.7-25.3/ 6-12 Spheroxyaster 16.0/ 8.0 n.o. Strongylaster 10.0 n.o.

n.r. 42-59-67

n.r. 20-22-26

n.r. 13-15-16

n.r. 6-7-10

47.5-55.6-66.5

9.2-17.2-25.3/ 6-12

n.o.

n.o.

23-44.4-55.2

9.2-17.4-25.3/ 6-10

n.o.

n.o.

n.r. 39.9-56.9-69.2

n.r. 14-23-32.4/ 4.3-10.2-16.2/ 3.2-4.1-5.4 14-23.9-31.4/ 4.3-9.7-13/ 3-4.3-5.6 15-24.8-34.2/ 4.9-9.7-13.2/ 3.4-4.5-5.8 16.2-25.9-33.5/ 5.4-10.7-14/ 3.2-4.4-6.5 16-24.4-35.2/ 5.1-9.5-14/ 3.2-4.4-5.6 7-15.2-22

n.r. 10.5-13.0-16.7/ 5.2-6.5-7.9

n.r. 5.5-8.9-12

41.3-59.7-85.1

9.5-12.5-16/ 5.9-7.2-8.6/ 1.2-1.6-2.4 9.8-12.6-15/ 5.8-6.9-8/ 1.3-1.6-2 9.7-12.5-16/ 5.5-7-8/ 1.4-1.7-2 9.7-12.6-16/ 5.6-6.9-8/ 1.2-1.5-2 5-9.3-14

5.6-8.5-11

48.6-60-68

5.8-8.4-10

49-55-65.6

5.7-8-9.5

55.9-64-72.9

5.8-8.3-11

40-47.6-54

5-7.6-10.0

54-64-75

7-20.9-30

5-11-14

6.5-10.2-15.0

30.5-47.1-58.4

35.6-40.6-45.7/ 7.6-9.7-10.2/ 12.7-16.5-20.3

10-12.5-16/ 5-7.5-10/ 1.5-1.8-2.0

8-12.5-16

598

Fig. 3: Skeletal arrangement of Geodia glariosa (MCNPOR 3647). A. View of the ectosome and choanosome; B. Detailed view of the threelayered ectosome, showing the ectocortex with beams of oxeas, the intermediate crust of grains of sand and the thick layer of sterrasters.

18/I/1980; UFRJPOR 189 (MCNPOR 4032): Esprito Santo, Santa Cruz, 195713S, 400920W, C.E.Z. coll., VII/1970; UFRJPOR 113 (MCNPOR 4030): So Paulo, So Sebastio, Ara, 234536S, 452435W, H. R. Costa coll., VII/1964; MCNPOR 175 (7 specimens), MCNPOR 176: Santa Catarina, 281603S, 484721W, B. V. Neto coll., 13/VI/1971, <13 m depth. Redescription: Massive, globose or hemispherical sponges (Fig. 2C, D and A, respectively) of light purple to black color, about 6.3 to 10.6 cm long, 6.0 to 8.8 cm width and 5.8 to 8.8 cm thick. Hard to firm consistency. Slightly rugose and hispid surface with rounded projections, partially covered by algae or bivalves shells (Mollusca) and/or cirripedians (Crustacea). Oscules and pores not observed. [Oscules not distinguishable from the pores, which have the usual sieve-like arrangement (Sollas 1888)]. Skeleton: The ectocortex, which lies beneath the exopinacoderm and the cortical strongylasters, is crowded with many foreign materials (sand grains, rock fragments and minute shells), and traversed by radiating brushes of short oxeas; and by the distal ends of the radiating choanosomal megascleres, particularly the ortho- and protriaenes, the cladomes of which frequently lie in the cortex, but sometimes pierce the surface. Preserved material has beige color and compressible consistency. This species is usually associated with barnacles (Balanus sp.) (Fig. 2A, D) or bivalves. The species has a three-layered ectosome measuring 800 to 1500 m thick (Fig. 3A, B), composed almost completely by the collagenous region, strongly yellow or yellowish-brown pigmented. The first layer containing bundles of small fanshaped oxeas is about 250-500 thickness and is superposed by a thin layer of strongylasters. The brushes of radiating cortical oxeas pass between some adjacent grains of sand, several brushes surrounding each individual grain as though keeping

599

it in place. The upper ends of some oxeas pierce the surface. The second layer is a crust of grains of sand of about 300 to 500 m, embedded among the cortical oxeas or disposed just bellow them. The grains are angular or subangular and irregular in shape, disposed in a layer considerably variable in thickness (160-400 m). Just below the ectocortex, there are 7 to 10 layers of sterrasters (Fig. 3B) irregularly scattered (about 300-500 m thick). Towards the innner portion of the sponge skeleton, there is a collagenous layer with few or no spicules about 47.5 m thick that completes the cortex on the inner side. Protriaenes and anatriaenes pierce the ectosome. Subectosomal region with aligned triaenes, mainly ortho-, but plagiotriaenes also occur, like pillars of ectosome. This region shows many canals and radial architecture. Outer portion of the choanosome exhibiting a radial arrangement. Inner region disorganized, composed of dense bundles with oxeas and triaenes. Oxyasters and spheroxyasteres disposed around the canals and throughout the choanosome.

Spicules
Megascleres Oxeas I (Fig. 4A-F): Choanosomal. Slightly curved to straight, stout, fusiform, with stepped (Fig. 4F) or gradually pointed (Fig. 4D) ends, sometimes abruptly pointed (Fig. 4E). Variable to styles (Fig. 4C) (590-1315.2-2240 / 4.6-19.7-37.2 m). Oxeas II (Fig. 4G-I): Cortical. Straight, slight or markedly curved, fusiform, with abruptly or gradually sharpening ends. Some spicules modified to strongyles (Fig. 4I) (200-366.1580 / 4.1-18-31.9 m). Orthotriaenes (Fig. 4J): Thick, conical rhabome, with gradually pointed end. Cladi are first curved upwards and then slightly downwards (790-1781.7-2856 / 14-38.8-60 m). Plagiotriaenes (Fig. 4K): Thin, straight or slightly sinuous rhabdome with gradually pointed end. Cladome with gradually sharpening cladi, usually curved upwards at the basal portion, and straight or curved downwards at the distal end (420-1134.1-2024 / 7-24.1-43.7 m). Anatriaenes (Fig. 4L, M): Long, thin and straight rhabdome, with thin, sinuous or slightly curved pointed ends. Cladi conical, slightly curved downwards, with gradually sharpening tips (897-2987.5-5319 m). Protriaenes (Fig. 4N-P): Long and slender rhabdome with stout, round, or thin, sinuous pointed ends. Cladome with strongyliform cladi strongly projected upwards (943-32616086 / 6.9-28.4-50.9 m). Microscleres Sterrasters (Fig. 5A-C): Spherical, with conical microspines. Young spicules provided of mucronate or strongyloid tips and with ends which are irregular or star-like (23-55.4-85 m). Oxyasters (Fig. 5D): Choanosomal. Small centrum; 6 to 10 long, thin and conical rays, with blunt or gradually pointed ends (7-26.2-45.7 m). Spheroxyasters (Fig. 5E, F): Cortical. Large centrum (about half of the spicule diameter), with 12 to 15 fusiform, blunt

Fig. 4: Megascleres of Geodia glariosa (Sollas, 1886). A, B. Oxea I; C. Styloid (Oxea II); D-F. Oxea I details of the tips; G-I. Oxea II; J. Orthotriaene; K. Plagiotriaene; L, M. Anatriaene; N-P. Protriaene.

or gradually pointed rays, provided with conical microspines curved upwards at the distal half. These spicules are scattered along the inner, fibrous portion of the cortex and surrounding it (5-10.9-16.7 m). Strongylasters (Fig. 5G): Somal. Small centrum, 8 to 12 cylindrical rays, blunt or truncated, rarely conical or pointed, with conical or blunt microspines all along their length (59.2-15 m). Ecology: The sponges were collected from rocky shores and mangroves (intertidal) and rocky substrate or gravel bottom (sublitoral). The specimens from Praia da Boa Viagem (Pernambuco), Praia da Pituba and Ilha do Medo (Bahia) are always associated with algae, particularly UFBA-POR 1235, which has 40% of its surface covered. The specimens from Bahia (off the coast, syntype), Porto Seguro, So Paulo and Santa Catarina are partially covered by barnacles (Balanus

600

Fig. 5: Microscleres of Geodia glariosa. A. Young sterraster; B. Adult sterraster; C. Sterraster surface with hilum; D. Oxyaster; E, F. Spheroxyaster; G. Strongylaster.

sp.) or bivalve mollusks. This species usually absorbs shells and/or rock fragments, as well as other kinds of available materials. Geographic distribution: South-west Atlantic. Endemic from Brazil. Pernambuco State: Praia da Boa Viagem, Recife, (first record); Bahia State: off the coast (Sollas, 1888); Salvador, Praia da Pituba and Ilha do Medo (first record); Una, Canavieiras (first record); Porto Seguro (first record); northern reefs of Cumuruxatiba (first record); Esprito Santo State: Santa Cruz (first record); So Paulo State: So Sebastio, Ara (first record); Santa Catarina State: Imbituba (Volkmer-Ribeiro and Mothes-de-Moraes, 1975). Bathymetric distribution: From intertidal region (0-1 m), Salvador, Bahia, Brazil (present paper) to 45.7 m, off Bahia State coast (Sollas, 1886).

Discussion
Geodia glariosa is the only species from the Brazilian coast to have an ectocortex composed of robust, gradually thinning oxeas opening in minute fans on the surface of the sponge, above and visibly distinct from the sterraster crust, and the only one of its genus (according to Sollas 1886, 1888) to incorporate sand or biodebris sediment (such as dead diatom shells) aggregated as a variably thick crust in the lower region of the ectosome, on the base of the bundles of cortical oxeas, or even between them, when such sediments are sparse. Those peculiarities, together with occurrence of robust, strongly

curved cortical oxeas differentiate G. glariosa from the other eight species of the genus recorded for the Brazilian coast and confirm its status as a valid species. Geodia glariosa is equally close to Geodia gibberosa (widely recorded for the Westerns Atlantic, Caribbean and Brazil) and G. megastrella, recorded for Eastern Atlantic, along the coast of Portugal (Carter 1876) and Barbados (Caribbean) by van Soest and Stentoft (1988). It differs from the former by presenting gradually pointed, robust cortical oxeas instead of thin strongyloxeas, besides more robust triaenes and bigger aster size; regarding G. megastrella, it differs in presenting protrieanes instead of promesotriaenes, cortical oxeas instead of strongyloxeas and smaller aster sizes. The species ability to incorporate sediments and biodebris material into its skeleton was widely discussed by Sollas (1886, 1888), who included considerations on how these materials are captured by the sponge and the path they follow to be introduced into the median part of the ectosome, from its entrance through the pores. In our comparative study, we detected considerable variation on the thickness of the layer of inclusions. On the specimen from Santa Catarinas coast, which grew on rocky shores of the shallow infralitoral, sand is replaced with dead diatoms shells, confirming VolkmerRibeiro and Mothes-de-Moraes (1975) observations. On specimens from the coast of Espirito Santo and So Paulo, the sand layer is thinner (340 m), and on specimens from Bahia (Porto Seguro), sand grains are sparsely distributed

601

between bundles of cortical oxeas, and do not constitute a distinct layer in the middle of the ectosome. Of the sponges collected on the mesolitoral, on rocky shores, sample UFBAPOR 1308 presented an inclusion layer twice as thick (1200 m) as that seen on specimen UFBA-POR 1235 (650 m). In this study, two additional categories of spicules are recorded besides those cited in the literature. Plagiotriaenes shorter and thinner than ortotriaenes (420-2024/7-43.7 versus 790-2856/14-60, respectively) were seen on the syntypes and all other examined specimens, and were easy to distinguish by having clades visibly, though discreetly, turned upwards (far from the spicule axis), with clades considerably shorter than those of ortotriaenes. Choanosomal styles were also seen in the specimen from Pernambuco (MCNPOR 3647) and in one from the the coast of Bahia (Porto Seguro, UFRJPOR1393). These spicules occur rarely and seem to represent variations of choanossomal oxeas. In the intertidal samples from Salvador (Bahia), spicular set varies in at least one triaene category according to spicular density and choanosome thickness: UFBA 1308-POR has rare anatriaenes and UFBA 1235-POR shows rare protriaenes. These spicules were seen on all other specimens examined in this study. Regarding the spicules, some variations are important records on the present redescription: in specimen MCNPOR 176, from the coast of Santa Catarina, all spicule categories are indeed smaller and thinner than the standard found for the species (see Table I), which confirms that it is a young specimen, as recorded by Volkmer-Ribeiro and Mothes-deMoraes (1975). Previous records of G. glariosa were restricted to the coast of Bahia (Sollas 1886) and Santa Catarina (VolkmerRibeiro and Mothes-de-Moraes 1975). This former gap on the species distribution is almost completely filled here with samples collected on the south-eastern and north-eastern coasts, on the States of Pernambuco (Recife), Bahia (Porto Seguro, Canavieiras, Cumuruxatiba, Praia da Pituba and Ilha do Medo), Esprito Santo (Santa Cruz) and So Paulo (So Sebastio, Ara). The specimen identified by de Laubenfels (1956) as Geodia gibberosa for Praia da Boa Viagem (Recife, Pernambuco State) is ascribed here to Geodia glariosa, expanding the species northern distribution limit. The species bathymetry is also expanded to the intertidal region, Praia da Pituba, Salvador, Bahia. The minute, light brown-colored specimen collected on the Coast of Spain, Gulf of Gascogne, at a depth of 248m, and identified by Topsent (1892) as Cydonium glariosus, is not considered co-specific with Geodia glariosa due to its disjoint geographical and bathymetrical distribution (the specimen was collected approximately at 44N), but definitive evaluation of this record was not possible due to unsuccessful search for the aforementioned sample, which was not found on the Collections of the Natural History Museum of Paris, where the slides of specimens collected on the expeditions of Prince Albert of Monaco are deposited (Rob van Soest, pers. comm.). In our opinion, such sample may possibly be attributed to Geodia megastrella, which was already cited for deep waters on the coast of Portugal (684 m) and for Barbados, Caribbean (153 m), and which also possesses three

categories of microscleres, of a shape quite similar to those found on G. glariosa. Both species differ though as regards their megascleres, in that G. megastrella has promesotriaenes instead of the protriaenes found in G. glariosa. In this way, we conclude that G. glariosa is an endemic species of shallow waters on the Brazilian coast.

Acknowledgements
Dr. Beatriz Mothes (MCN/FZB), Dr. Eduardo Hajdu and Dr. Guilherme Muricy (MNRJ) for loan of samples or microscopical slides; Dr. Cla Lerner (MCN/FZB) for the photographs of the specimens from Bahia (Cumuruxatiba) and Santa Catarina States; MCN technicians for the SEM photographs; FAPESP, FAPESB and CNPq for financial support.

References
Boury-Esnault N (1973) Campagne de la Calypso au large des ctes atlantiques de lAmrique du Sud (1961-1962). I, 29. Spongiaires. Rs Sci Camp Calypso 10: 263-295 Carter HJ (1876) Descriptions and figures of deep-sea sponges and their spicules, from the Atlantic Ocean, dredged up on board H.M.S. Porcupine, chiefly in 1869 (concluded). Ann Mag Nat Hist 18(105): 226-240; (106): 307-324; (107): 388-410; (108): 458-479 de Laubenfels MW (1956) Preliminary discussion of the sponges of Brazil. Contrib Avulsas Inst Oceanogr So Paulo, Oceanogr Biol 1: 1-4 Hajdu ECM (1994) A phylogenetic interpretation of hamacanthids (Demospongiae, Porifera) with a redescription of Hamacantha popana (de Laubenfels, 1935). J Zool 232: 61-77 Hajdu ECM, Muricy GR, Custdio M, Russo C, Peixinho S (1992) Geodia corticostylifera (Demospongiae, Porifera) new Astrophorid from the brazilian coast (Southwestern atlantic). Bull Mar Sci 51(2): 204-217 Hechtel GJ (1976) Zoogeography of Brazilian marine Demospongiae. In: Harrison FW, Cowden RR (ed). Aspects of sponge biology. Academic Press, New York. pp. 237-259 Mello-Leito A, Pego AF, Lopes WM (1961) Porferos assinalados no Brasil. Av Cent Est Zool Univ Brasil 10: 1-29 Mothes B, Silva CMM (2002) Stelleta ruetzleri sp. nov., a new anchorinid from the Southwestern Atlantic (Porifera, Demospongiae). Sci Mar 66(1): 69-75 Muricy G, Hajdu E (2006) Porifera brasilis: guia de identificao das esponjas marinhas mais comuns do Sudeste do Brasil. Srie Livros 17. Museu Nacional, Rio de Janeiro Poljaeff N (1884) Report on the Keratosa collected by H.M.S. Challenger during the years 1873-1876. Rep Sci Res Voy H.M.S. Challenger, Zool 11: 1-88 Ridley SO, Dendy A (1887) Report on the Monaxonida collected by the H.M.S. Challenger during the years 1873-1876. Rep Sci Res Voy H.M.S. Challenger, Zool 20: 170-186 Silva CMM (2002) Reviso das espcies de Geodia Lamarck, 1815 (Porifera, Astrophorida, Geodiidae) do Atlntico Ocidental e Pacfico Oriental. PhD Thesis, Universidade de So Paulo, So Paulo Silva CMM, Mothes B (2000) Three new species of Geodia Lamarck, 1815 (Porifera, Geodiidae) from the Brazilian coast, Southwestern Atlantic. Rev Suisse Zool 107(1): 31-48

602

Silva CM, Mothes B, Lyrio-Oliveira I (2004) Redescription of Geodia papyracea (Hechtel, 1965) with new records along the Northeastern and Southeastern Brazilian coast. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Boll Mus Ist Biol Univ Genova 68: 605-612 Sollas WJ (1886) Preliminary account of the tetractinellid sponges dredged by H.M.S. Challenger, 1872-1876. Part I. The Choristida. Sci Proc R Dublin Soc 5: 177-199 Sollas WJ (1888) Report on the Tetractinellida collected by H.M.S. Challenger during the years 1873-1876. Rep Sci Res Voy H.M.S. Challenger, Zool 25(63): 1-458 Topsent E (1892) Contribution ltude des spongiaires de lAtlantique Nord. Rsult Camp Sci Albert I Monaco 2: 1-165

Uriz MJ (2002) Family Geodiidae Gray, 1867. In: Hooper JNA, van Soest RMW (eds). Systema Porifera: a guide to the classification of sponges, vol. 1. Kluwer Academic/Plenum Publishers, New York. pp. 134-140 van Soest RWM, Stentoft N (1988) Barbados deep water sponges. Stud Fauna Curacao Caribb Isl 70: 1-175 Volkmer-Ribeiro C, Mothes-de-Moraes B (1975) Esponjas tetraxonidas do litoral sul-brasileiro. I - Redescrio de Cydonium glariosus Sollas, 1886 e Erylus formosus Sollas, 1886. Iheringia sr Zool 47: 3-22 von Lendenfeld R (1903) Tetraxonia. In: Schuzle FE (ed). Das Tierreich, Lief. 19. Friedlnder, Berlin. pp. 1-168

Porifera research: Biodiversity, innovation and sustainaBility - 2007

603

The perils and merits (or the Good, the Bad and the Ugly) of DNA barcoding of sponges a controversial discussion
Antonio M. Sol-Cava(1*), Gert Wrheide(2)
Molecular Biodiversity Laboratory - Genetics Department Instituto de Biologia Universidade Federal do Rio de Janeiro. Bloco A CCS Ilha do Fundo. 21941-490 Rio de Janeiro, RJ, Brazil. sole@biologia.ufrj.br (2) Abteilung Geobiologie, Geowissenschaftliches Zentrum, Georg-August Universitt Gttingen, Goldschmidtstr. 3, D37077 Gttingen, Germany. gert.woerheide@geo.uni-goettingen.de
(1)

Abstract: DNA barcodes are defined as signature sequences used to identify unknown specimens. They have been proposed as a means to quickly solve the taxonomical impediment, through an organised and highly structured effort, using the same part of the mitochondrial sub-unit I of the Cytochrome c oxidase gene to identify all species of the planet. There has been much debate about the uses and misuses of DNA for taxonomy, and the radical proposition of DNA barcodes has heated the debate. In this paper we present two contrasting views of how DNA barcodes may or may not help sponge taxonomy. Keywords: Barcodes, critical analysis, cytochrome oxidase, molecular systematics, Porifera

Introduction
a classification founded on any single character, however important that may be, has always failed Charles Darwin, 1859

The use of genetic markers to detect cryptic species and formulate phylogenetic hypotheses has revolutionised systematics and taxonomy in the last thirty years. From the early studies with allozymes (reviewed in Thorpe and Sol-Cava 1994) to the recent analyses of DNA sequences (reviewed in Avise 2004), molecular systematics have mostly corroborated classic taxonomy. However, the use of molecular markers has also challenged many long-held beliefs in taxonomy, such as the cosmopolitanism of many marine invertebrate species (Klautau et al. 1999, Knowlton 2000), the closer relationship of Nematoda to Arthropoda (forming the Ecdysozoa Aguinaldo et al. 1997) than to other worms (Halanych 2004, Mallatt and Giribet 2006) or the recent hypothesis about the phylogenetic position of Placozoa (Dellaporta et al. 2006). Over 20 years have passed since the first paper on molecular systematics of sponges was published (Sol-Cava and Thorpe 1986), and much progress has been made in technical and analytical approaches, which led to amazing discoveries, like the close affinity between some chondrosids and aplysinids (Borchiellini et al. 2004, Nichols 2005), the polyphyletism of Axinella (whose species seem to be scattered among different orders; Borchiellini et al. 2004) and the deconstruction of the Ceractinomorpha and Tetractinomorpha sub-classes of Demospongiae (Borchiellini et al. 2001, Boury-Esnault 2006). Clearly, taxonomy and

systematics have benefited immensely from these new approaches, and will continue to do so. The continuous advances in DNA sequencing technology have recently led to the proposition of using short (about 650 bp) mitochondrial DNA sequences (more specifically, of the Cytochrome Oxidase c subunit I gene, CO1 or cox1) to identify all living species (Hebert et al. 2003a). Those sequences would ideally function as species-specific signature sequences (so-called DNA barcodes), which would allow quick (in a few minutes) and unambiguous identification of any organism straight in the field. The proposed development and use of very small and cheap hand-held CO1 sequencers would obviate the need for field guides or taxonomists to identify samples at some point in the future (Hebert et al. 2003a). Obviously, such ambitious claims have achieved much attention and publicity (and some funding), to what its proponents would like to turn into something like the Human Genome project. The essence of the Consortium Barcodes of Life (CBoL) initiative has produced a heated debate. In this paper we discuss the merits (good) and perils (bad and ugly) aspects of DNA barcodes applied to sponge taxonomy. Rather than produce positive or negative conclusions about the utility of DNA barcodes for sponges, we expect to foment the discussion, and help the readers see both sides of this contentious issue. Since this paper reports on the debate between the two authors during the Buzios Sponge Symposium, each author will present his point of view on the subject. AMSC will present his personal view on the bad and the ugly aspects of DNA barcodes in general as well as a more specific view for sponges, and GW will present why he believes DNA barcodes (or better a DNAassisted taxonomy) might aid sponge taxonomists in species

604

description and discovery. But first we must define what DNA barcodes are and what they are not.

What are DNA barcodes?

A DNA barcode is a DNA signature sequence that allows the identification of a specimen to a known species. The stated goal of the Consortium for the Barcodes of Life (CBoL) is that anyone, anywhere, anytime be able to identify quickly and accurately the species of a specimen whatever its condition (http://phe.rockefeller.edu/barcode/). To be useful, that sequence must be short, ubiquitous, and easily amplifiable and sequenceable using universal primers. It must be conserved enough to allow the identification of higher taxonomical ranks, like phyla down to genera, variable enough to distinguish even highly similar species, but not too variable, so that levels of intraspecific variability do not add up too much noise to species identification. The choice of a mitochondrial gene as the barcoding marker was based on three facts: 1) mitochondrial genes are relatively easy to amplify by PCR, because mitochondria are abundant in the cells, making DNA extraction straightforward even from degraded samples; 2) the mitochondrial genome is haploid, allowing for direct sequencing of PCR products without phasing alleles by e.g. cloning or SSCP as usually is the case for nuclear, diploid, genes and 3) levels of recombination on the mitochondrial genome are very low, reducing problems of paralogy. The mitochondrial gene chosen was the Cytochrome Oxidase c subunit I, whose choice was based primarily on the availability of a large number of sequences in GenBank, the existence of universal primers that allow amplification of this fragment from most phyla (Folmer et al. 1994), and the claim that different taxonomic levels could be resolved with the marker in most organisms (Hebert et al. 2003b).

that the methodology must be robust enough to be used by laypersons, under varied circumstances. Consequently, many problems commonly faced and properly handled by scientists working with molecular systematics, like contamination, paralogy and identification errors can become important sources of error. The problems summarised below have all been handled appropriately by molecular systematists through careful, case-by-case, analysis of the data. Therefore, for molecular systematics, those problems are just a source of noise/homoplasy, which can be made explicit and be solved through the critical analysis of the data. They become important and very bad, however, when the middle-man (the biologist) is removed from the process, and the communication is made directly by the DNA sequencer and the databanks, as envisaged by CBoL.

Contamination
Many organisms live in intimate associations with other species. In those cases, contamination can be an important problem for sequencing using universal (i.e. not phylum specific) primers. Sponges harbour enormous amounts of other organisms, and direct sequencing of PCR products will often produce misleading results (see Erpenbeck et al. 2002 for a good analysis of this problem in sponges). This can be circumvented by scientists with a careful analysis of the produced sequences using phylogenetic methods, but it may be an important problem if a barcoder is to be trusted by people in the field to identify single specimens (Hurst and Jiggins 2005).

Paralogy
One of the advantages of using CO1 was that recombination was rare, and the haploid nature of mtDNA made it easy to assume homology of the analysed sequences. However, copies of parts of the mitochondrial genome are often found in the nuclear genome (Mourier et al. 2001). In most cases, direct sequencing of PCR products of mitochondrial genes will produce the true mitochondrial sequence. However, sometimes the nuclear, pseudo-mitochondrial, copies will be preferentially amplified and sequenced instead of the true mtDNA (Williams and Knowlton 2001, Thalmann et al. 2004). In those cases, the produced sequences will be paralogous to those present in the databases. If their transfer to the nucleus is old enough, those sequences will have diverged over the threshold of 2.5% divergence used by the CBoL to exclude identification to known species, and the result will be the wrong identification of the samples. Another source of paralogy is incomplete lineage sorting (Wahlberg et al. 2003). Using a single gene sequence to identify species will miss much of their evolutionary history, and taxa recently diverged may all too easily be overlooked (Choat 2006).

What DNA barcodes are not

There is much confusion between the concepts of DNA barcodes (a Technique) and Molecular Systematics (a Science). In order to be really useful for quick identification of specimens in the field, as said before, DNA barcode systems must be ubiquitous and use universal primers (precluding the need for prior identification of the phylum or class of the organism by the user). Therefore, as stated by the Barcodes of Life (BoL) consortium (http://barcoding. si.edu/), DNA barcoding is a technique to identify specimens. Nothing else. DNA barcoding is not phylogenetic analysis. If the system requires a critical phylogenetic analysis to work, it is not useful for barcoding purposes. DNA barcoding is not molecular systematics (Moritz and Cicero 2004). It cannot rely on extensive geographic sampling to decide the taxonomic status of each unknown specimen.

Perils and pitfalls: the Bad about DNA barcoding

The aim of the Barcodes of Life project is to develop a methodology that will allow the quick, on the fly, identification of specimens in the field, using portable sequencing devices (barcoders) connected by satellite to large databanks (http:// www.dnabarcoding.ca/barcode_initiative.php). This means

Horizontal gene transfer and introgression

In plants and protists, horizontal gene transfer can be a relatively common phenomenon (Bergthorsson et al. 2003), while in animals it is considered to be rare (Kurland et al. 2003). We do not know how common horizontal gene transfer may be in sponges, but there are some evidences indicating

605

that this may have happened in Tetilla (Rot et al. 2006). Another source of polyphyletism on mtDNA is introgression, a phenomenon that is not uncommon in animals (Moritz 1987, Quesada et al. 1995). When those processes occur, an immediate result is that mitochondrial gene trees will not be adequate representations of the species phylogenies. Nuclear DNA determines about 100% of the phenotype of each organism, the way it looks and adapts to the environment. An individual from one species that has a mitochondrial DNA from another will still belong to the former species, but it would be wrongly identified, if we used the CO1 sequence as the sole parameter to identify it, as the one whence its mtDNA came from.

Identification errors in the database

Any database is only as good as the data put in it. GenBank is riddled with errors, which are often dismissed by many authors using their data for their own research. These errors include sequencing errors (Karlin et al. 2001, Foster 2003) but, more importantly, identification errors. For example, an ad hoc identification analysis of fungi species whose sequences had been deposited in GenBank revealed that over 20% of them had been wrongly identified (Bridge et al. 2003). This problem has been efficiently handled by CBoL through the establishment of quality standards for the submitted sequences, and the requirement of voucher museum specimens for each sequence entered into their database. However, the sheer volume of specimens deposited into the museii will inevitably mean that most specimens will not have their taxonomic identification verified after they have received their first name. Once the name has been tied to the sequence in the database, the error may be perpetuated in subsequent identifications, leading to a cascade of taxonomic errors. Most results presented by barcoding advocates are of groups with well resolved taxonomy, where the system is more likely to work well. However, a recent, large-scale (over 2,000 individuals belonging to 263 taxa) evaluation of the barcoding approach to a marine invertebrate group (Gastropoda) found that when no representatives were present in the database (simulating what would happen when using barcodes to unveil new species), barcodes failed to identify species recognised by taxonomy over 20% of the time (Meyer and Paulay 2005). Errors included the lumping of different species as single entities, and considering conspecific specimens as belonging to different species (Meyer and Paulay 2005). This lack of correlation between identification by barcodes and by conventional taxonomy may, in fact, indicate that conventional taxonomy is wrong, and that levels of paraphyly and polyphyly observed all resulted from oversplitting and overlumping real biological species (Funk and Omland 2003, Meyer and Paulay 2005). However, even if 100% of the mismatches between species identified by taxonomists and by barcodes were due to taxonomical errors, this would still be a major drawback to the BoL initiative, since it would mean that even the initial database, built on species identified by experts in the field, would be liable to be wrong. Consequently, unless only holotypes were used for building the database, sequences could not be reliably attached to species names. For example,

we know, now, that Chondrilla nucula, formerly considered to be a cosmopolitan species is, in fact, a species complex (Klautau et al. 1999, Usher et al. 2004). If that information was not available, during the building of the CBoL database of known sponge species what sequence would definitely represent C. nucula would depend on where the sample had been collected. It could be argued that, since all Chondrilla specimens from the Mediterranean analysed to date formed a monophyletic, low divergent cluster, the sequence from a Mediterranean specimen (the type locality of C. nucula) would adequately represent that species. But what would happen, then, with the cosmopolitan Oscarella lobularis, that aggregates two sibling species (Boury-Esnault et al. 1992, Loukaci et al. 2004) within the Mediterranean, where it was originally described?

Reification of species
One of the things that made the CBoL so attractive was their clear aim and the promise of unambiguously identifying, in a short time, all species of the planet. To identify a specimen to a species, it is important, above all, to know what a species is. There is an enormous ongoing debate about what a species may be, with over 22 species definitions used by different authors (Mallet 1996). CBoL does not try to define what a species is. They follow the pragmatic approach of verifying how much divergence in CO1 sequence exists between species acknowledged as different by taxonomists, and use the average divergence as a rule of thumb threshold above which specimens are considered to belong to different species. The currently accepted threshold for the CBoL consortium is a p distance of 0.025 (Hebert et al. 2003b). This means that, if the sequence of an unknown specimen is less than 2.5% divergent from a sequence present in the database, it will be identified as belonging to that species. A species, then, is reified by barcoders as a group of organisms that is over 2.5% different from any other groups. It is as simple as that. No doubts, no grey zones. That is the advantage of relying on a single character to identify species. However attractive that can be to ecologists, pharmaceutical companies or other users of taxonomic identifications, it is at least nave, and at worst very dangerous. Anyone with some experience in taxonomy knows how this simplistic approach to identification is prone to error and can seriously go wrong. The definition of the threshold value above which sequences are considered to belong to different species is also very important: setting a high threshold value means that false positives (= incorrectly deciding that a given sequence belongs to a different species from that in the database, which would correspond to a type I statistical error) will be more rare, but it will also mean that many different species will be considered as belonging to the same genetic species (= false negatives, which would correspond to type II statistical errors). Conversely, setting a low threshold value will increase the number of species likely to be detected, but it will also mean giving species status to what may be simply intraspecific varieties (see e.g. Bradley and Baker 2001). This question was addressed by the analysis of a large dataset by Meyer and Paulay (2005). They found that, using a carefully built phylogeny for all the 263 evolutionary significant

606

units (ESUs) sampled (through the use of molecular, morphological, ecological and reproductive data), if they chose a threshold value of 2% they would have between 11% and 20% (depending on number of individuals sampled per ESU) false positives (oversplitting), and 8% false negatives (overlumping). Increasing the threshold value to 3% would diminish the number of false positives considerably, to 2% to 3%, but it would also increase the proportion of false negatives to 16%. For the gastropods studied, Meyer and Paulay found that the threshold value that would produce the smallest number of false positives and false negatives would be 2.6%. However, even at that best threshold level total error rates were still as high as 17% (Meyer and Paulay 2005). A big problem with having to decide on threshold distance values as a basis for taking taxonomic decisions is that they are reductionist and bound to lead to artificial taxonomic entities. Taxonomy took a long time to incorporate the conceptual and analytical advances of cladistics and evolutionary biology. It would be sad, now, to return to phenetic, distance-based approaches (de Queiroz and Good 1997), abandoning critical character-based thinking. Furthermore, even if we were to accept the overall idea of a distance-based taxonomy, we would have to deal with the probably insurmountable problem of the inexistence of a precise evolutionary clock. Evolutionary rates can vary enormously not only between different genes or taxa, but also between different parts of the same genes (Stevens and Schofield 2003). There are analytical ways to deal with this problem (Aris-Brosou and Yang 2002, Thorne and Kishino 2005) but, again, they depend on a case-by-case analysis incompatible with the idea of automatic identification. In sponges, there are few works using CO1 sequences for species-level taxonomy (Schroder et al. 2003, Duran et al. 2004, Nichols and Barnes 2005, Wrheide 2006), but it appears that the barcoding region of CO1 may be too conserved in sponges (Wrheide et al. 2004, Erpenbeck et al. 2006b). For example, several species of Chondrilla that could be identified through allozymes, ribosomal sequences and conventional taxonomy would all be clustered into a single species if we used a 2% CO1 divergence threshold to separate them (Zilberberg, personal communication). It is clear that at this point in time, we do not have sufficient amount of data at hand to decide on any threshold, should there be a universal one for sponges.

that scientific knowledge can be crystallised. But they also include serious political questions, like the brain-drain of young students and scientists from taxonomic work into the band-wagon of methodologically easy, well funded, highly publishable but scientifically empty barcode programs.

Brain-drain from classical systematics

The recognition of Science as a major source of National wealth resulted in increased levels of funding and a strong sustained growth of Graduate programs and Research Institutes. This has led to deep changes in the way scientists are funded and evaluated, with much weight being put into publication and impact factors, and public accountability of the work done. Those are welcomed changes, since Science largely relies on public funding, and it is natural that scientists should be evaluated in relation to the way they perform their work. However, because the evaluation system is still being constructed, there are large distortions, which favour scientists working in the more fashionable areas of Genetics and Biotechnology, in detriment of more slow producing fields like Zoology and Ecology. A consequence of this distortion has been a brain-drain, with graduate students and young scientists migrating from the slower to faster publishing fields. A program that puts even more emphasis on DNA for systematics will only make matters worse (Ebach and Holdrege 2005), to a point where we may irreversibly lose expertise, as the best sponge taxonomists will fail to train students interested on identifying and describing sponge species before they retire from their field. This problem is particularly serious because taxonomy expertise takes years to build.

Reductionism and pragmatism:

Only through the ignorance of arrogance could one fail to learn the lessons of several centuries of comparative morphology. Single-character systems rarely work for even one truly diverse clade and never work for all clades Will et al. 2005

The Ugly
Your work, Sir, is both new and good, but whats new is not good and whats good is not new Samuel Johnson, XVIII century (cited by Will et al. 2005)

Bad as they may currently be, the technical problems of molecular barcodes may eventually be circumvented through technological developments and rigorous methodological approaches. However, there are more serious, deeper philosophical and political problems with the idea of molecular barcodes, particularly in relation to their ultimate end of identifying all species of the planet. The ugly aspects of the BoL initiative are related to philosophical issues, like the return to a 19th century typological thinking and the idea

The barcoding of life project is not scientific. It has been successful in capturing the attention from the media and from some funding agencies because it makes huge promises and downplays the enormous difficulties associated with taxonomy. The CBoL site justifies the development and application of Barcodes saying that, in 250 years of existence, Zoologists have only described about 15% of animal diversity (www.dnabarcoding.ca/rationale.php). However, it is possible that taxonomic work has not been slow because zoologists havent worked hard enough or because they lacked technology. Progress may have been slow because systematics is a complex science. If taxonomists were willing to make the same conceptual compromises as the CBoL proponents, by oversimplifying the complex task of delimiting biological species, they would have finished the description of biodiversity very quickly (and just as wrongly as CBoL would). If we are willing to accept that a species can be defined based on 650 bp of a mitochondrial gene (this represents less than 0.00000001% of the total genome

607

of species that have had their genome completely sequenced thus far), then we could, for example accept that sponge species could be described solely based on spicule types. We could even envisage a Spiculometer (fig 1), which could do an image recognition of spicule slides and, based on the different spicule combinations, give us a quick, reproducible and precise identification. The fact that that identification would be wrong (grouping, for example, most haliclonids as a single species), would be secondary to our objective of naming all of sponge biodiversity. By making false promises (like barcoding the whole biodiversity of the planet in ten years with 1 billion dollars; Hebert et al. 2003a) with very competent public relations and lobbying activists, CBoL has quickly attracted the attention of the media, which always welcome golden pill solutions to the problems of society. Because evolutionary rates are not the same for the same genes across the taxonomical landscape, the relationships obtained from DNA sequence comparisons reflect only indirectly the evolutionary history of each group. Only through the sampling of several, different characters (including morphology, ecology and genetic data) can we begin to understand the limits between evolutionary lineages and, based on those, take informed decisions about species borders. A good example of how the use of multiple datasets has helped understanding taxonomic relationships in sponges can be found in Erpenbeck et al. (2006a). The taxonomical impediment is real, but cutting corners in identifying species may only make matters worse, generating confusion and deviating resources from proper species descriptions into just discovering possible new biological entities. The rate at which systematists are describing new species is not limited by the number of new things to describe. The shelves in taxonomists laboratories are already full of specimens waiting to be analysed, and the real limiting factor has been, and still is, the access to collections, bibliography and qualified personnel. In other words, taxonomists are already overburdened with new species to describe, and the bottleneck of species description may be made worse by the large amounts of putative new species found by barcoding. This unavoidable crisis may have a positive result to systematics, through the final realization that conventional taxonomy was, after all, what really needed support. However, this crisis may also have a different, less bright outcome to Systematics. Faced with huge numbers of species waiting to be named, it may become too tempting to simply replace formal taxonomic descriptions with some barcode species name (Baker and Bradley 2006) or molecular Operational taxonomy unit (Blaxter 2004) that will link specimens to gene sequences without further studies. We will have, then, a name (or a code) and a sequence, but will that be useful at all for biology? What is the difference between a conventional label in a collection jar, with data on time and local of collection and an arbitrary voucher number, and a similarly arbitrary number, linked to a DNA sequence? Without formal study by taxonomists how will those newly found species serve the biological community or society?

Crystalisation of knowledge
Moreover, the generation of COI profiles will provide a partial solution to the problem of the thinning ranks of morphological taxonomists by enabling a crystallization of their knowledge before they leave the field Hebert et al. 2003a

Barcodes for identifying things can be very good and useful, but only AFTER the taxonomic work has been done properly. For example, barcoding birds or whales can be very useful to control the illegal traffic of endangered species. However, the main argument used by the BoL initiative to justify its very large budget was the zoological impediment, which means that their ultimate promise is to identify the 85% species that have not been described by taxonomists. Barcoding things is essentially typological and, as Paul Hebert correctly puts it, could be a way to crystallise knowledge. It is true that stability in names is something important, but it should not be made arbitrarily (Knapp et al. 2004). There is a huge gap in taxonomy to be filled, and crystallising the current knowledge in a rapidly changing field, like sponge taxonomy, is bound to be a step backwards.

Merits and opportunities: the Good about DNA barcoding (or better a DNA-assisted taxonomy)
Because what keeps on moving, is eternal (Nam quod semper movetur, aeternum est) M.T. Cicero (106-43 BCE): De Re Publica VI (27) (Scipios Dream)

DNA barcoding provides exciting new means for quick species identification and discovery. The use of DNA signature sequences (aka DNA barcodes) in sponge taxonomy, supplementing conventional morphological characters, will revolutionize future ways in which we conduct taxonomic research to define and describe species. The fascinating idea of a universal DNA barcode for all organisms and a hand-held DNA barcode scanner, similar to the tricorder in the now famous science fiction series Star Trek (www.startrek.com), that enables identification of any life form on our planet on the fly, might sound a bit too ambitious at present, but technological advances might enable such a system at some stage in the not too distant future. However, even nowadays, scientific research around DNA barcodes will provide multiple exciting opportunities for sponge research, e.g. to increase our knowledge and understanding about principles of molecular evolution, speciation processes, community ecology and species delimitation. A DNA sequence-assisted taxonomic system for sponges, providing the means to quickly and unequivocally identify taxa, will significantly ease the workload of taxonomic service provided by the few experts in the field to pharmaceutical and ecological researchers, among others, who need to identify the taxa they encounter in their surveys or that show promising biochemical activities. DNA barcoding approaches will open up a new dimension and quality in biodiversity research and will become of vital importance for the survival and acknowledgement of sponge taxonomy and increase its

608

Fig. 1: The Spiculometer. A parody on how morphological sponge taxonomy could join the fast lane.

reputation over the coming decades. It would be a serious disadvantage to disregard the opportunities that molecular (DNA barcoding) approaches bring to the field. We, as the community of scientists working on sponges, need to capitalize on (and not ignore) the new potential of scientific and financial opportunities and resources that the DNA barcoding movement creates and use them to our advantage, before others, who do not have the necessary taxonomic experience, do it. DNA barcoding resources will be vital to actually get the work done when attempting to identify taxa in large collections that exist in various museums around the world in a reasonable timeframe (i.e. before retirement and with a respectable publication list) otherwise we will never create interest among young scientists to endeavour in sponge taxonomic research. A good example is the large collection of the Great Barrier Reef Seabed Biodiversity mapping project (www.reef.crc.org.au/resprogram/programC/seabed/index. htm), coordinated by the Australian Institute of Marine Science, which is attempting to document the sessile epibenthic fauna in the inter-reefal areas of the GBR. Thousands of samples have been collected, but without additional funding from DNA barcoding initiatives (or pharmaceutical companies

for that matter), taxonomic work on such large collections will only proceed very, very slowly. Another yet unexplored aspect is the identification of the vast diversity of cryptic and/ or small encrusting sponges (e.g. Richter et al. 2001), which then can be identified from tiny biopsies. This will open up a whole new dimension of sponge biodiversity, pivotal e.g. for our understanding of nutrient cycling and bentho-pelagic coupling in coral reefs (Lesser 2006). Once funding for DNA barcoding is obtained, those new resources can, should and will be utilized to create also new positions for conventional taxonomic work and train a new generation of multidisciplinary taxonomists ready for the challenges of an integrative taxonomy of the 21st century. Those new resources (monetary and human) will also create exciting new opportunities for international collaborations (see for example the Sponge Barcoding Project, introduced in this volume by Wrheide et al.) to tackle the many methodological and intellectual challenges that lie ahead. DNA barcoding of sponges will also change the societys appreciation of the taxonomic work done in dusty natural history museums and turn that into a picture of modern science that is methodologically up-to-date and ready for

609

future challenges. With world-wide declining funding for a not-so- terribly-sexy science like taxonomy, new resources from DNA barcoding might be pivotal for the survival of conventional taxonomy and will also enable research in natural history museums that goes beyond barcoding, i.e. do molecular systematics, phylogeography and molecular evolutionary research, to better understand the processes that shaped present-day biodiversity. However, there are certainly some aspects of DNA barcoding that need careful consideration. First of all, especially in marine organisms harbouring numerous microbial and/or metazoan commensals or symbionts, contamination is definitely an issue. Designing sponge-specific primers for DNA-taxonomy markers should circumvent this issue, however, sequences obtained will have to be verified by phylogenetic tests in any case (this should be the usual procedure in any lab anyway). Paralogy, horizontal gene transfer and introgression on the other hand, can and will only be detected by phylogenetic tests once sufficient comparative data is accumulated and we have to start doing so otherwise we will never get a deeper understanding of those issues. It is also clear that one mitochondrial marker will not be sufficient to establish a DNA taxonomic system for sponges that will aid species description and discovery; we will have to include at least one nuclear marker an approach discussed in Wrheide et al. (2007). This can and will only be done together with taxonomic experts. Identification errors inherently occur in databases, but can be minimized by cross-verification by those taxonomic experts, an approach advocated in the Sponge Barcoding Project (SBP) (www.spongebarcoding. org; see also Wrheide et al. 2007). A philosophical (and practical) problem certainly is the definition of what a (sponge) species actually is. (Sponge) taxonomists still mostly use fixed diagnostic characters (e.g. spicules and architecture) derived from comparative morphology to diagnose and separate species, not necessarily adhering to the biological species concept or any other than a typological one. While this has served reasonably well to catalogue diversity and is practical, it remains contentious whether it reflects the real biological diversity of sponges, considering that so-called cosmopolitan sponge species, often only possessing a small number of morphological characters, are most likely a set of sibling (cryptic) species with different and divergent evolutionary histories, as uncovered by numerous genetic studies (e.g. Klautau et al. 1999). Existing morphological alpha-taxonomy of sponges is a rather artificial system solely based on morphological differences without considering evolutionary history and/or reproductive isolation. Furthermore, those morphological characters (spicules) used to define species differences have been shown to potentially vary with environmental conditions, i.e. the silica content of seawater has the potential to modulate the phenotypic expression of various spicule types (Maldonado et al. 1999). Quite disturbing. The time has come to seriously consider additional characters, like DNA signature sequences, to corroborate taxonomic hypotheses. The argument that species identities will be reduced to single characters (a gene fragment) is not valid. Foremost, in a DNA sequence (how ever long it might be) each nucleotide position represents a separate character

with four character states, in a protein sequence each amino acid represents one character, each with 20 character states (see textbooks like Page and Holmes 1998). So in e.g. the standard barcoding marker COI we have about 650 characters, some of them diagnostic, so it should be possible to quantify differences among species recognized by conventional taxonomy based on DNA sequences, preferably a combination of one mitochondrial and one nuclear marker. Inherent difficulties with species level analysis of DNA signature sequence are widely appreciated (e.g. Hickerson et al. 2006) and recent novel analytical approaches begin to tackle those problems at least in terrestrial organisms (Pons et al. 2006). Also the argument that a species can not be defined based on a single gene sequence is debatable, as it has been shown numerously that putative barrier genes exist (most of them found in model species such as Drosophila spp.) that are associated with reproductive incompatibilities (see recent review by Noor and Feder 2006). Even if this makes only 0.00000001% of one species genome, it certainly can make a difference. Further, the phenotype, solely recognized in conventional taxonomy, certainly is a reflection of the genotype, but only a reflection of a small fraction of this genotype. However, in sponge taxonomy we are at the very beginning of establishing a system of DNA taxonomy and DNA barcodes that could aid in future species discovery and description, and currently we do not have data to decide on thresholds of genetic distances for species delimitation nature certainly is not black and white in this regard and will not provide us with a simple solution, but we can only learn, develop and advance by gathering additional DNA sequence data and develop and apply novel analytical approaches to solve old problems where conventional taxonomy is at its limits. The DNA barcoding approach provides now novel ways to obtain funding to be able to do so and to rejuvenate taxonomy, and increase public awareness and appreciation of its new relevance. The brain-drain from classical systematics can not be overcome by disregarding technological and analytical novelties and advances. It would be similar to argue against using email just because the guys in the post office might lose their jobs. The not-negligible brain-drain from classical taxonomy has deeper roots in the practice of how science is currently conducted and evaluated, as outlined above, and it would be nave to believe that condemning DNA taxonomy/ barcoding would solve this problem. Instead, we should take the opportunity and promote a taxonomic system that has a solid base both in conventional comparative morphology and DNA sequence analysis. We should not replace formal taxonomic (morphological) descriptions with DNA sequences (or worst barcodes), but try to unify both into one integrated system and use the potentially new funding sources from barcoding initiatives to support such a system.
Where we go from there is a choice I leave to you (Neo, 1999 The Matrix)

610

The gold nugget of DNA barcoding

The new methodological approaches, intellectual challenges and potential funding resources provided by the DNA barcoding movement now puts us, as scientificallyconscious researchers, in the unique position to actually use to our advantage the gold nugget dangling in front of us. We will be the ones who steer future sponge taxonomy in the right direction, get it out of the dust and make it ready for the multiple challenges of the 21st century. DNA barcoding will enable creation of new exciting positions for a new generation of integrative taxonomists, enable novel research directions and research collaborations. In a world of dwindling resources for taxonomy those new opportunities can not be dismissed and might prove pivotal for the future survival and appreciation of sponge taxonomy.

informed, careful way, never losing sight of conventional taxonomy based on morphological characters.

Acknowledgements
We thank the organizers and the participants of the 7th Sponge Symposium for hosting the discussion on this important topic. AMSC was financed by the Brazilian (CNPq) and Rio de Janeiro (FAPERJ) Research Agencies. AMSC thanks Jean Vacelet for competently modelling the spiculometer photo, and the team from the Molecular Biodiversity Laboratory (LBDM) for suggestions and long debates. GW acknowledges funding from the German Research Foundation (DFG), from the sponsors of the symposium to fund his participation, and Dirk Erpenbeck as well as his group from the Molecular Marine Biodiversity Lab (MMBL) of the Dept. of Geobiology for fruitful discussions, Dirk Erpenbeck for the Cicero and Catherine Vogler for the Neo citation.

Concluding remarks
Molecular systematics is clearly a mature, growing science. It helps conventional taxonomy because it adds a new dimension to the analysis of species and their phylogenetic relationships. However, DNA barcodes are not synonymous with molecular systematics. Their sole stated aim is to provide a quick means to identify specimens to known species, in a similar way that supermarket barcodes serve for the identification of goods to the supermarket database. For that single end molecular barcodes are, indeed, very useful, and the error rates associated with the boundary between intraspecific and interspecific levels of CO1 sequence divergence (estimated at around 5% for the comparison of unknown samples with reference sequences of the species in the database; Meyer and Paulay 2005) are quite acceptable and possibly similar to error rates associated with the use of field guides and taxonomic keys. Problems with DNA barcodes become more prominent when they are used to identify new species, particularly in groups where conventional taxonomy is still far from being complete, as in the case of sponges. DNA barcodes can be seen as a welcome additional source of funds for museii and zoology departments, but they may also result in a brain-drain of young scientists away from conventional taxonomy which, ultimately, is what needs more support, since it represents the bottleneck in the description of the Worlds biodiversity. We believe that the only way the Barcodes of Life consortium will achieve its objective is through concurrent support of conventional taxonomy, and we propose that sponge barcoding projects should have that aim clearly stated, objectively allocating about 20% of all obtained resources specifically to the work of species descriptions by conventional taxonomists. Ultimately, embracing or not the DNA barcodes program will be a personal decision, for which matters beyond scientific criteria may be important. With time, it will become clear if the promises of the Consortium Barcodes of Life will be fulfilled (original claim by Paul Hebert: all species barcoded until 2010 with a budget of about 1 billion dollars. Hebert et al. 2003a; revised targets: 2020 deadline and under 2 billion dollars budget. Paul Hebert, in Whitfield 2003). In any case, we believe that it is important that the choice be made in an

References
Aguinaldo AMA, Turbeville JM, Linford LS, Rivera MC, Garey JR, Raff RA, Lake JA (1997) Evidence for a clade of nematodes, arthropods and other moulting animals. Nature 387: 489-493 Aris-Brosou S, Yang Z (2002) Effects of models of rate evolution on estimation of divergence dates with special reference to the Metazoan 18S ribosomal RNA phylogeny. Syst Biol 51: 703-714 Avise JC (2004) Molecular markers, natural history and evolution. Chapman & Hall, London Baker RJ, Bradley RD (2006) Speciation in mammals and the genetic species concept. J Mammal 87: 643-662 Bergthorsson U, Adams KL, Thomason B, Palmer JD (2003) Widespread horizontal transfer of mitochondrial genes in flowering plants. Nature 424: 197-201 Blaxter ML (2004) The promise of a DNA taxonomy. Phil Trans Roy Soc Lond B 359: 669-679 Borchiellini C, Chombard C, Manuel M, Alivon E, Vacelet J, Boury-Esnault N (2004) Molecular phylogeny of Demospongiae: implications for classification and scenarios of character evolution. Mol Phylogenet Evol 32: 823-837 Borchiellini C, Manuel M, Alivon E, Boury-Esnault N, Vacelet J, Le Parco Y (2001) Sponge paraphyly and the origin of Metazoa. J Evol Biol 14: 171-179 Boury-Esnault N (2006) Systematics and evolution of Demospongiae. Can J Zool 84: 205-224 Boury-Esnault N, Sol-Cava AM, Thorpe JP (1992) Genetic and cytological divergence between colour morphs of the Mediterranean sponge Oscarella lobularis Schmidt (Porifera, Demospongiae, Oscarellidae). J Nat Hist 26: 271-284 Bradley RD, Baker RJ (2001) A test of the genetic species concept. cytochrome b sequences and mammals. J Mammal 82: 960-973 Bridge PD, Roberts PJ, Spooner BM, Panchal G (2003) On the unreliability of published DNA sequences. New Phytologist 160: 43-48 Choat JH (2006) Phylogeography and reef fishes: bringing ecology back into the argument. J Biogeogr 33: 967-968 de Queiroz K, Good DA (1997) Phenetic clustering in biology: A critique. Q Rev Biol 72: 3-30 Dellaporta SL, Xu A, Sagasser S, Jakob W, Moreno MA, Buss LW, Schierwater B (2006) Mitochondrial genome of Trichoplax adhaerens supports Placozoa as the basal lower metazoan phylum. Proc Natl Acad Sci USA 103(23): 8751-8756

611

Duran S, Pascual M, Turon X (2004) Low levels of genetic variation in mtDNA sequences over the western Mediterranean and Atlantic range of the sponge Crambe crambe (Poecilosclerida). Mar Biol 144: 31-35 Ebach MC, Holdrege C (2005) More taxonomy, not DNA barcoding. BioScience 55: 822-823 Erpenbeck D, Breeuwer JAJ, Parra-Velandia FJ, van Soest RWM (2006a) Speculation with spiculation? Three independent gene fragments and biochemical characters versus morphology in demosponge higher classification. Mol Phylogenet Evol 38: 293305 Erpenbeck D, Breeuwer JAJ, Velde HCvd, van Soest RWM (2002) Unravelling host and symbiont phylogenies of halichondrid sponges (Demospongiae, Porifera) using a mitochondrial marker. Mar Biol 141: 377-386 Erpenbeck D, Hooper JNA, Wrheide G (2006b) CO1 phylogenies in diploblasts and the Barcoding of Life; are we sequencing a suboptimal partition? Mol Ecol Notes 6: 550-553 Folmer O, Black M, Hoeh W, Lutz R, Vrijenhoek R (1994) DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol Mar Biol Biotechnol 3: 294-299 Foster P (2003) To Err is Human. Ann Hum Genet 67: 2-4 Funk DJ, Omland KE (2003) Species-level paraphyly and polyphyly: Frequency, causes, and consequences, with insights from animal mitochondrial DNA. Annu Rev Ecol Evol Syst 34: 397-423 Halanych KM (2004) The new view of animal phylogeny. Annu Rev Ecol Evol Syst 35: 229-256 Hebert PDN, Cywinska A, Ball SL, de Waard JR (2003a) Biological identifications through DNA barcodes. Proc R Soc Lond B 270: 313-321 Hebert PDN, Ratnasingham S, deWaard JR (2003b) Barcoding animal life: cytochrome c oxidase subunit 1 divergences among closely related species. Proc R Soc Lond B 270: 96-99 Hickerson M, Meyer C, Moritz C (2006) DNA Barcoding will often fail to discover new animal species over broad parameter space. Syst Biol 55: 729-739 Hurst GDD, Jiggins FM (2005) Problems with mitochondrial DNA as a marker in population, phylogeographic and phylogenetic studies: the effects of inherited symbionts. Proc R Soc Lond B 272: 525-1534 Karlin S, Bergman A, Gentles AJ (2001) Genomics - Annotation of the Drosophila genome. Nature 411: 259-260 Klautau M, Russo CAM, Lazoski C, Boury-Esnault N, Thorpe JP, Sol-Cava AM (1999) Does cosmopolitanism result from overconservative systematics? A case study using the marine sponge Chondrilla nucula. Evolution 53: 1414-1422 Knapp S, Lamas G, Lughadha EN, Novarino G (2004) Stability or stasis in the names of organisms: the evolving codes of nomenclature. Phil Trans R Soc Lond B 359: 611-622 Knowlton N (2000) Molecular genetic analyses of species boundaries in the sea. In: Sol-Cava AM, Russo CAM, Thorpe JP (eds). Marine Genetics. Kluwer Academic Publishers, Amsterdam, pp. 73-90 Kurland CG, Canback B, Berg OG (2003) Horizontal gene transfer: A critical view. Proc Natl Acad. Sci USA 100(17): 9658-9662 Lesser M, P. (2006) Benthic-pelagic coupling on coral reefs: Feeding and growth of Caribbean sponges. J Exp Mar Biol Ecol 328: 277288

Loukaci A, Muricy G, Brouard JP, Guyot M, Vacelet J, BouryEsnault N (2004) Chemical divergence between two sibling species of Oscarella (Porifera) from the Mediterranean Sea. Biochem Systemat Ecol 32: 893-899 Maldonado M, Carmona MC, Uriz MJ, Cruzado A (1999) Decline in Mesozoic reef-building sponges explained by silicon limitation. Nature 785-788 Mallatt J, Giribet G (2006) Further use of nearly complete, 28S and 18S rRNA genes to classify Ecdysozoa: 37 more arthropods and a kinorhynch. Mol Phylogenet Evol 40: 772-794 Mallet J (1996) The genetics of biological diversity: from varieties to species. In: Gaston KJ (ed). Biodiversity: a Biology of Numbers and Divergence. Blackwell Science Inc., Malden, MA, USA, pp. 13-53 Meyer CP, Paulay G (2005) DNA Barcoding: Error rates based on comprehensive sampling. PLoS Biology 3: e422 Moritz C, Cicero C (2004) DNA Barcoding: Promise and Pitfalls. PLoS Biology 2: 279-354 Moritz C, Dowling, T.E. and Brownn, W.M. (1987) Evolution of animal mtDNA: relevance for population biology and systematics. Annu Rev Ecol Syst 18: 269-292 Mourier T, Hansen AJ, Willerslev E, Arctander P (2001) The Human Genome Project reveals a continuous transfer of large mitochondrial fragments to the nucleus. Mol Biol Evol 18: 1833-1837 Nichols SA (2005) An evaluation of support for order-level monophyly and interrelationships within the class Demospongiae using partial data from the large subunit rDNA and cytochrome oxidase subunit I. Mol Phylogenet Evol 34: 81-96 Nichols SA, Barnes PAG (2005) A molecular phylogeny and historical biogeography of the marine sponge genus Placospongia (Phylum Porifera) indicate low dispersal capabilities and widespread crypsis. J Exp Mar Biol Ecol 323: 1-15 Noor MAF, Feder JL (2006) Speciation genetics: evolving approaches. Nat Rev Genet 7: 851-861 Page RDM, Holmes EC (1998) Molecular evolution: a phylogenetic approach. Blackwell Science, Oxford Pons J, Barraclough T, Gomez-Zurita J, Cardoso A, Duran D, Hazell S, Kamoun S, Sumlin W, Vogler A (2006) Sequence-Based species delimitation for the DNA taxonomy of undescribed insects. Syst Biol 55: 595-609 Quesada H, Wenne R, Skibinski DOF (1995) Differential introgression of mitochondrial DNA across species boundaries within the marine mussel genus Mytilus. Proc R Soc Lond B 262: 51-56 Richter C, Wunsch M, Rasheed M, Kotter I, Badran MI (2001) Endoscopic exploration of Red Sea coral reefs reveals dense populations of cavity-dwelling sponges. Nature 413: 726-730 Rot C, Goldfarb I, Ilan M, Huchon D (2006) Putative cross-kingdom horizontal gene transfer in sponge (Porifera) mitochondria. BMC Evol Biol 6: 71 Schrder HC, Efremova SM, Itskovich VB, Belikov S, Masuda Y, Krasko A, Mller IM, Mller WEG, Medina M, Weil E, Szmant AM (2003) Molecular phylogeny of the freshwater sponges in Lake Baikal. J Zool Systemat Evol Res 41: 80-86 Sol-Cava AM, Thorpe JP (1986) Genetic differentiation between morphotypes of the marine sponge Suberites ficus (Demospongiae: Hadromerida). Mar Biol 93: 247-253 Stevens JR, Schofield CJ (2003) Phylogenetics and sequence analysis - some problems for the unwary. Trends Parasitol 19: 582-588

612

Thalmann O, Hebler J, Poinar HN, Paabo S, Vigilant L (2004) Unreliable mtDNA data due to nuclear insertions: a cautionary tale from analysis of humans and other great apes. Mol Ecol 13: 321335 Thorne JL, Kishino H (2005) Estimation of divergence times from molecular sequence data. In: Nielsen R (ed). Statistical methods in molecular evolution. Springer Verlag, London, pp. 235-256 Thorpe JP, Sol-Cava AM (1994) The use of allozyme electrophoresis in invertebrate systematics. Zoolog Scripta 23: 3-18 Usher KM, Sutton DC, Toze S, Kuo J, Fromont J (2004) Biogeography and phylogeny of Chondrilla species (Demospongiae) in Australia. Mar Ecol Progr Ser 270: 117-127 Wahlberg N, Oliveira R, Scott JA (2003) Phylogenetic relationships of Phyciodes butterfly species (Lepidoptera: Nymphalidae): complex mtDNA variation and species delimitations. Syst Entomol 28: 257-273 Whitfield J (2003) DNA barcodes catalogue animals. News@nature DOI:10.1038/news030512-7

Will KW, Mishler BD, Wheeler QD (2005) The perils of DNA barcoding and the need for integrative taxonomy. Syst Biol 54: 844-851 Williams SL, Knowlton N (2001) Mitochondrial pseudogenes are pervasive and often insidious in the snapping shrimp genus Alpheus. Mol Biol Evol 18(8): 1484-1493 Wrheide G (2006) Low variation in partial cytochrome oxidase subunit I (COI) mitochondrial sequences in the coralline demosponge Astrosclera willeyana across the Indo-Pacific. Mar Biol 148: 907-912 Wrheide G, Erpenbeck D, Menke C (2007) The Sponge Barcoding Project: aiding in the identification and description of poriferan taxa. In: Custdio MR, Lbo-Hajdu G, Hajdu E, Muricy G (eds). Porifera research: biodiversity, innovation and sustainability. Srie Livros 28. Museu Nacional, Rio de Janeiro. pp. 123-128 Wrheide G, Fromont J, Sol-Cava AM (2004) Population genetics and phylogeography of sponges - a workshop synthesis. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Boll Mus Ist Biol Univ Genova 68: 683-688

Porifera research: Biodiversity, innovation and sustainaBility - 2007

613

Sexual reproduction of Geodia barretti Bowerbank, 1858 (Porifera, Astrophorida) in two Scandinavian fjords
Frank Spetland(1), Hans Tore Rapp(1*), Friederike Hoffmann(2), Ole Secher Tendal(3)
Department of Biology, University of Bergen, Bergen High-Technology Centre, PO Box 7800, N-5020 Bergen, Norway. frank@naturexpo.no, hans.rapp@bio.uib.no (2) Max Planck Institute for Marine Microbiology, Celsiusstr. 1. D 28359 Bremen, Germany. fhoffman@mpi-bremen.de (3) Zoological Museum, SNM, University of Copenhagen, Universitetsparken 15, DK - 2100 Copenhagen , Denmark. ostendal@snm.ku.dk
(1)

Abstract: The gametogenesis in the common cold-water sponge Geodia barretti is described from two Scandinavian fjords through a year cycle: in Korsfjorden, western Norway, and in Kosterfjorden on the Swedish west-coast. The reproductive cycle is annual for both populations, with one or two periods of gamete release per year. Individuals within the same local population reproduce simultaneously within a restricted period of time. Geodia barretti is a dioecious and oviparous sponge, with oocytes (up to 100 m in diameter) and spermatic cysts (up to 125 m in diameter) organised in clusters within the mesohyl. The sponge has asynchronous spermatogenesis and synchronous oogenesis. Asexual reproduction has not been observed. The onset of the reproduction coincides with the phytoplankton blooms in both fjords: Gametes are released in early summer, just after the phytoplankton spring bloom is over. In Kosterfjorden, however, an additional release of gametes occurs in October, just after the autumn phytoplankton bloom. In both fjord systems reproduction of G. barretti thus matches the peaks in sedimentation of organic matter that follow after phytoplankton blooms. Keywords: Geodiidae, reproduction, spermatogenesis, oogenesis, sedimentation

Introduction
There are relatively few studies on the reproduction within the Astrophorida. All species belonging to this order are thought to be oviparous and dioecious (Scalera and Sciscioli 1970, Bergquist 1978, Fell 1983). The oocytes of the astrophorid sponges are small and the reproduction is expected to run through an annual cycle (Simpson 1968, Scalera and Sciscioli 1970, Hoffmann et al. 2003). The oogenesis is assumed to last more than two months, and to be initiated before the spermatogenesis (Fell 1974, Bergquist 1978, Kaye 1990, Kaye and Reiswig 1991). The sponges are expected to have a spawning phase where each population shed their gametes simultaneously. A recent study on the shallow-water Mediterranean species Geodia cydonium (Jameson, 1811) confirmed the oviparity and gonochorism of this species (Mercurio et al. 2007). Sponges belonging to the Astrophorida form mass occurrences on the outer shelf and upper slope over large areas of the NE Atlantic. Geodia barretti Bowerbank, 1858 is the dominating sponge species on such grounds along the Norwegian coast, around the Faroe Islands, in the Barents Sea, and in The Denmark Strait (Klitgaard and Tendal 2004). Reproduction and recruitment details of sponges on the grounds are largely unknown, a fact that is true for many sponge groups (Leys and Ereskovsky 2006).

Despite the abundance of Geodia barretti Bowerbank, 1858 in the Northeast Atlantic, the knowledge on the reproductive biology of the species is very limited. It is suggested that G. barretti is oviparous, dioecious and reproduces in the spring (material from Korsfjorden) (Hoffmann et al. 2003). Methods for cultivating G. barretti have recently been developed (Hoffmann et al. 2003), and this has set a new focus on the reproductive biology of this species. In order to goal-direct further work on cultivation of G. barretti, it is necessary to investigate the natural reproductive cycle, and to find out if any periodicity in the reproduction can be detected, possibly influenced by hydrographical conditions. Such a study combined with an estimate of the reproductive output may aid in developing a model for the population dynamics of the species, and perhaps can explain how mass occurrences of this and related species are built up and sustained. Furthermore, this kind of information is essential to develop successful methods in sponge aquaculture for biotechnological production of secondary metabolites. This study investigates the sexual reproduction and reproductive cycle of two populations of Geodia barretti in the NE Atlantic and its relation to hydrography and vertical carbon flux.

614

Material and methods

The sampling took place in the Korsfjord (6010N, 510E) outside Bergen, Norway (Fig. 1), using a triangular dredge between 60 and 250 m depth, and in the Kosterfjord (585N, 1106E) outside Strmstad, Sweden (Fig. 2) at depths between 80 and 200 m. The samples in the Korsfjord were taken between December 2002 and June 2004, while the samples from the Kosterfjord were taken in 1974, 1975 and 1976. In the Kosterfjord, the samples came from three localities; Kruggl, Ulvillarna and Bjrns Rev. See Table 1 for overview of samples. Several subsamples were taken from all Norwegian individuals for histological examination. The Swedish samples were already subsamples from the original sponge. The subsamples were relatively large (> 10 cm3), from the cortex to the centre of the animals. The tissue sampled in Norway was fixed for 24h in a mixture of 0.03% glutaraldehyde and 4% formaldehyde in seawater and subsequently dehydrated and stored in 70% ethanol. The Swedish material was fixed in Bouins solution and stored in 70% ethanol. Subsamples were embedded in paraffin and Technovit 7100 for microtome sectioning. Before embedding and sectioning (2 m) the samples, the spicules were removed. This was done using 5% hydrofluoric acid (HF) in phosphate buffer saline (PBS) for 1h. The sections were stained in two different ways, some with GIEMSA and others with Toluidine blue. The samples were studied in a regular light microscope (Zeiss Axioplan microscope).

The origin of these cells could not be observed. The early spermatogonia gather to form a cluster (Fig. 3B) and are soon surrounded by a single layer of one or more cells (Fig. 3C). These cells are surrounded by the sponge associated bacteria, but these seem to take no part in the spermatogenesis. A cavity is formed and the spermatogonia start their development (Fig. 3D). At this point, the size of the early spermatic cyst is ~20-30 m, and the spermatogonia are about 5 m across. The spermatic cysts appear in clusters in the mesohyl, although, solitary cysts do occur. When the cyst is complete, the differentiation of the developing sperm cells continues. When all sperm cells have completed their differentiation, the cysts have grown to become up to 120 m across, and contain numerous mature spermatozoa (Fig. 3E and 3F). The development from spermatogonia to spermatozoa within the cysts lasted for approximately 2.5 months. Using only a light microscope it was not possible to observe details of spermatogonial stage. However, it was observed that sperm development within one specific spermatic cyst is asynchronous, with some cells being in division, while others were in different phases of the meiosis. The developing sperm cells appear to be scattered randomly within the cyst; their single flagellum pointing in any direction.

Oogenesis
The oogenesis is largely synchronous within a specimen and between different individuals in the population. All the oocytes have approximately the same size at the same time. The oogenesis starts with the transformation of an unknown cell type into an early version of the previtellogenetic oocyte (Fig. 4A). In the beginning of the oogenesis, the oocyte is about 15-20 m in diameter, amoeboid, with pseudopodia (true for the entire process of oogenesis) and probably still mobile. At this point in the process, the nucleus makes up

Results Spermatogenesis
Spermatogenesis begins in the mesohyl with some cells aggregating at a seemingly random location (Fig. 3A).

Fig. 1: Map of the Korsfjord. The sill is marked in the lower left corner of the image. The sampling area (arrows) is situated along the steep vertical rocky walls of the SE side of the fjord (from Wassmann 1991).

615

Fig. 2: Map of the Kosterfjord (modified from Bmstedt 2000). A. Bjrns Rev. B. Kruggl and C. Ulvillarna.

Table 1: Overview of the sampled material. KF=Korsfjorden, KsF-BjR= Bjrns Rev in Kosterfjorden, KsF-Kr=Kruggl in Kosterfjorden, KsF-Ulv= Ulvillarna in Kosterfjorden. Locality KF KF KF KF KF KsF - Bj.R KsF - Bj.R KsF - Bj.R KsF- Kr KsF- Kr KsF- Kr KsF - Ulv KsF - Ulv KsF - Ulv Sampling date 21.05.04 12.02.03 11.04.03 21.05.03 16.06.03 23.03.77 17.06.76 18.10.76 07.05.75 25.05.76 16.02.76 06.05.75 08.06.76 17.11.75 Specimens collected 10 12 15 14 11 7 11 15 12 4 10 3 8 7 Reproductive specimens 2 2 4 1 6 2 5 7 1 1 2 1 2 1 Male : Female ratio 0;2 0:2 1:1 0:1 1:1 1:1 4:1 4:3 0:1 0:1 1:1 0:1 0:2 0:1 Depth 60-250 m 60-250 m 60-250 m 60-250 m 60-250 m ~200 m ~150 m ~150 m ~80 m ~100 m ~100 m ~140 m ~110 m ~120 m

about 80% of the cell volume (Fig. 4B and C), and the quantity of cytoplasm is modest. The early oocytes are not surrounded by any layer of supportive cells, and they seem to migrate to the vicinity of an excurrent canal before the oogenesis is completed. Now the oocyte becomes surrounded by a layer of what may be fibre bundles of collagen type (Fig. 4D and 4E). The origin of the fibre bundles is not known. Before onset of the vitellogenesis, the diameter of the oocyte is ~15-20 m. The vitellogenetic phase was dominated by the accumulation of yolk, assumed to be obtained from the mesohyl via endocytosis due to the lack of nurse cells. During this phase, the diameter of the cell increases from ~15-20 m to 90-100 m, and the fibril layer becomes thicker (Fig. 4E, 4F and 4G). After the vitellogenetic phase the oocyte reaches the mature stage (Fig. 4H). It is then 90-100 m in diameter and has a nucleolated nucleus. In addition to the nucleolus, there is a brighter sphere in the nucleus (Fig. 4D). The fibril layer has

reached 12 m across and the nucleus diameter has reached 30 m. The inclusions that may be seen in the nucleus of a maturing oocyte of G. barretti are assumed to be condensed chromatin (Fig. 4E and 4H).

The reproductive cycle

In the Korsfjord the gametogenesis starts in February/ March and ends with spawning sometime in late May or beginning of June. In the Kosterfjord two reproductive periods were observed within these three geographically very close localities. The two periods of gametogenesis run from February and July with spawning in May/June and October respectively. The spawning phase at Kruggl is in May/June, and at Ulvillarna and Bjrns Rev in October and June (Fig. 5). The hydrographical conditions in the area show cyclic changes through vertical mixing and inflow of Atlantic

616

Fig. 3: Spermatogenesis in Geodia barretti. A, B. Earliest observed stage of spermatogenesis (arrows). c = water canal. C, D. The spermatic cyst is almost complete. The arrows in D indicate the asynchronous development within the same cyst. CWI = cells with inclusions, Sp = Spermatic cyst. E. The surrounding cell (Sc) is prominent, and the space within the cyst (Cs) is obvious. F. The most developed spermatic cyst observed. Note that some spermatocytes are still in division (D). All sections except C, were embedded in Technovit and stained with Toluidine blue. C was embedded in LR-White and stained with GIEMSA.

water through a year cycle (Matthews and Sands 1973), but no direct correlation between hydrography and timing of gametogenesis or spawning was observed. In G. barretti from Korsfjorden, no gametes have been found from July until February (Hoffmann et al. 2003), and

hence, no storing of gametes was observed. No hermaphrodites were found. Not all specimens reproduce every year. Of the 125 specimens examined only 36 were reproductively active. Out of these 15 were males and 21 females. The overall

617

reproductive M:F fraction hence becomes 1:1.4, while the overall reproductive: non-reproductive fraction becomes 1:3.5. For Korsfjorden, the M:F fraction is 1:2.1 and for Kosterfjorden 1:1.2. The time required for differentiation of spermatozoa is shorter than for oocytes, and spermatogenesis starts later in the year than the oogenesis. The oogenesis lasts about four months, while the spermatogenesis lasts for about 2.5 months. The oocyte probably leaves the sponge through an excurrent canal during the same time as the spermatozoans. The density of oocytes in the tissue varies, as for the spermatic cysts, but at some places it reaches 12 oocytes within an area of 5 mm2 mesohyl. The reproductive phase is the same each year for the local population, but whether or not the single individual reproduces annually or every other year, is not yet known.

Discussion
As suggested for all Astrophorida, Geodia barretti is oviparous and dioecious (Scalera and Sciscioli 1970), and as it has a relatively low density of oocytes when compared to other species, it is assumed to behave as a K-strategist (Ereskovsky 2000).

Spermatogenesis
The arrangement and process of spermatogenesis in G. barretti does not differ in any great extent from closely related species. The development of spermatozoa is asynchronous, and takes place within cysts scattered around the mesohyl in clusters (Simpson 1968, Scalera and Sciscioli 1970, Fell 1974, Bergquist 1978, Gaino et al. 1986, Wolfgang 1989, TanakaIchihara and Watanabe 1990, Ilan 1995, Mercurio et al. 2007). The origin of spermatic cysts from choanocyte chambers is widely accepted, and has been suggested in many studies (Simpson 1968, Scalera and Sciscioli 1970, Fell 1974). However, it seems that the origin of the spermatocysts in Geodia barretti is not from choanocyte chambers as in its close relative Erylus discophorus (Scalera and Sciscioli 1970) as most of the spermatic cysts are much larger than any choanocyte chamber and no fusion of spermatic cysts or choanocyte chambers was observed. Synchronous release of sperm has been reported earlier (Reiswig 1970). However, considering the asynchronous development of spermatozoa it is likely that sperm release may happen several times during the reproductive period of G. barretti.

mesohyl, a pattern also known from other sponges (Bergquist 1978, Ayling 1980). Oocytes are arranged in clusters scattered seemingly without any special order. Oogenesis in the early stages proceeds without the apposition of follicle cells to enclose the irregular oocyte. The oocytes of G. barretti are surrounded by a layer of fibrils or fibre bundles in the same manner as in e.g. Geodia conchilega (Scalera and Sciscioli 1970) and G. cydonium (Sciscioli et al. 1994). After the reproductive phase there are no traces of the fibril layer in the mesohyl of G. barretti. This suggests that the fibrils are released from the sponge together with the oocytes. Growing oocytes have irregular surfaces because of the presence of numerous pseudopodia (Fig. 4D) protruding into the mesohyl, as also observed in other species (Lepore et al. 2000, Mercurio et al. 2007). The mature oocyte has a more rounded shape, as it has withdrawn the pseudopodia from the mesohyl (Fig. 4H). Lack of nurse cells has previously been reported in E. discophorus, G. cydonium, G. conchilega and Stelletta grubii Schmidt, 1862 (Scalera and Sciscioli 1970), and suggests that yolk is accumulated by active uptake from the mesohyl (Gaino et al. 1986, Sciscioli et al. 2002). No incorporation of bacteria in the oocyte was observed in G. barretti although this has been reported for other species of Geodia (Lepore et al. 2000). However, bacteria surrounding the oocyte in the previtellogenic stage was observed. The single nucleolus in the nucleus of maturing oocytes of G. barretti is sometimes accompanied by a brighter sphere of unknown origin.

The reproductive fraction of the specimens

The reproductive vs. the non-reproductive fraction in G. barretti was 1:3.5. This is relatively high for the nonreproductive part. It is however important to evaluate how many specimens and of which sizes it is necessary to sample to find the true fraction (Fell 1983). In a recent study of the shallow-water species G. cydonium all examined individuals (n = 10) were sexually active (Mercurio et al. 2007). Many sponges have a seasonal, repetitive reproductive period, with gametogenesis occurring in asynchronous cycles once a year (Scalera and Sciscioli 1970, Diaz 1973, Fell 1983, Kaye 1990). Geodia barretti has a synchronous oogenesis associated to one or two annual cycles respective to the geographical location. It is known that some sponges do not reproduce every year (Ayling 1980), and that a single sponge specimen may wait more than one year before it reproduces again. Such a strategy may explain the low number of reproductive individuals in the examined material from the two populations of G. barretti. It may also be that some of the smallest specimens of G. barretti had not reached reproductive age. The sex ratio in G. barretti is more equal between the sexes than in most other sponges (M:F = 1:1.4). Most often the male fraction tends to be lower than the female, such as in Ancorina alata (Dendy, 1924) M:F = 1:5.3, Polymastia hirsuta = 1:1.99, Aaptos aaptos = 1:8.3 (Ayling 1980), Halichondria panicea = 1:2 (Witte and Barthel 1994), with some exceptions such as Halichondria panicea = 10:1-7 (Witte et al. 1994) and

Oogenesis
In most sponge species, the oocytes are scattered throughout the mesohyl (Fell 1983), and this is also the case in G. barretti. In G. barretti, both single oocytes and aggregates of oocytes may be found. The germinal cells are not known, but for related species like E. discophorus the oocyte originates from cells of amoeboid type (Scalera and Sciscioli 1970). Except from the outer zone, down to approximately 1 cm below the cortex, gametogenesis takes place throughout the entire

618

619

Fig. 5: Annual reproductive cycle of G. barretti compared to sedimentation in Fana Fjord (close to Korsfjord, see Fig. 1) at 60 m and 90 m depths, representing typical sedimentation events in a fjord system. Spawning in Korsfjord and at two sites in Kosterfjord correlates well with the sedimentation peak just after the phytoplankton spring bloom. The second spawning event in the Ulvillarna/Bjrns Rev population corresponds to the sedimentation peak after the autumn bloom. Figure modified from Wassmann (1991).

Cinachyra tarentina Pulitzer-Finali, 1983 = 7:1 (Lepore et al. 2000). In other studies the males were even absent (Corriero et al. 1996, Corriero et al. 1998). No hermaphrodites were found in this study, and in general hermaphroditic individuals are very rare within the Astrophorida (Scalera and Sciscioli 1970). Our findings contrast an earlier report that G. barretti could reproduce asexually (Burton 1947), as no indications of this were observed. Burtons study was most probably performed on Geodia mesotriaena (Hentschel, 1829) or Isops phlegraei pyriformis (Vosmaer, 1885) (Klitgaard and Tendal 2004).

What triggers reproduction?

Temperature seems to have no direct influence on the reproductive cycle of Geodia barretti as the specimens from Bjrns Rev and Ulvillarna in the Kosterfjord show two reproductive periods within the year (Fig. 5). The populations of G. barretti in the Korsfjord and the Kosterfjord showed different reproductive peaks. Both fjords are sill fjords, and have an annual renewal of Atlantic water, changing the
Fig. 4: Oogenesis in Geodia barretti. A. The earliest recognizable oocyte (Ooc) in a previtellogenetic state, surrounded by bacteria (Bact). B. The beginning of the vitellogenetic phase showing that the nucleolus is well developed already, and that the oocyte may still be mobile due to the lack of a fibril layer and the presence of pseudopodia (Pp). Nu = nucleolus. C, D. Accumulation of yolk in the oocytes. The nucleus in D contains a bright sphere (X) in addition to the nucleolus (Nu) which is of unknown origin, task and importance. CWI = Cells With Inclusions/ Spherulous cell. E. growth of the surrounding fibril layer (Fl) and the uptake of yolk from the mesohyl (Cyt). F, G. An almost full-grown oocyte. Some cytoplasmic bridges (Cb) are still present and yolk (Y) is still being transported into the oocyte. H illustrates the mature oocyte in comparison to a spherulous cell (CWI), and shows that the nucleus contains chromatin (Chr).

temperature, salinity and oxygen content in their deepest parts. The fjords are also similar in the sense that their hydrography varies little in the deeper water layers. No direct relation seems to exist between the reproductive period and seasonal variation in water temperature or salinity. Studies in which the temperature has effect on the reproductive cycle are numerous (e.g. Fell 1976, Witte et al. 1994, Ereskovsky 2000), however studies reporting that temperature has no effect on the reproductive cycle are scarce (Corriero et al. 1998). No positive correlation between the salinity and the reproductive cycle was observed. Studies show that there is no effect on the reproductive cycle of Halichondria panicea due to increase or decrease in salinity (Witte et al. 1994, Corriero et al. 1998). Studies on Thenea abyssorum Koltun, 1959, Trichostemma sol (Sars, 1872) and Tentorium semisuberites (Schmidt, 1870), state that the sexual reproduction cycle in deep sea sponges is probably triggered by sedimentation and advection of particulate organic carbon (Witte et al. 1994). This may also be the case for G. barretti as the onset of the reproduction coincides with the phytoplankton blooms in both fjords. Normally there is a quite massive spring bloom from March to April and a smaller autumn bloom in August/September in both fjords, both followed by an increased sedimentation (Wassmann 1991). The gametes in the Korsfjorden population are fully mature in early summer and released in late May or the beginning of June, just after the phytoplankton spring bloom is over, and the sedimentation of dissolved and particulate organic matter is at its highest (Wassmann 1991). In Kosterfjorden release of gametes seems to occur at the same time after the spring bloom; however, in this fjord an additional release of gametes occurred in October, just after the autumn phytoplankton bloom has ended (Fig. 5). The reproductive cycle of G. barretti is then influenced by the spring bloom in Korsfjorden and both the spring and autumn bloom in Kosterfjorden. Most likely the local population of

620

sponges releases their eggs and sperm within a short period of time (presumably simultaneously) into the surrounding water masses. This study has shown that Geodia barretti is a dioecious and oviparous sponge, with oocytes and spermatic cysts organised in clusters within the mesohyl. The sponge has asynchronous spermatogenesis and synchronous oogenesis. The reproductive cycle is annual with one or two periods of gamete release per year. The onset of the reproduction coincides with the phytoplankton blooms and the release of gametes is likely to occur when sedimentation of particulate organic matter is at its highest after the phytoplankton blooms.

Acknowledgements
This work has been supported by the University of Bergen and The Norwegian Research Council (grants to Hans Tore Rapp), The European Commission through contract number HPRI-CT-199900056 (Friederike Hoffmann and Hans Tore Rapp) and The European Commission through Copenhagen Biosystematics Centre (COBICE) (grant to Frank Spetland). The crew of RV Hans Brattstrm is thanked for assistance during field work. Thanks are due to Joachim Reitner and Wolfgang Drse (University of Gttingen) for help and support during a research visit in Gttingen. We are grateful to the anonymous referees for their helpful suggestions which improved the content of the paper.

References
Ayling AL (1980) Patterns of sexuality, asexual reproduction and recruitment in some subtidal marine Demospongiae. Biol Bull 158: 271-282 Bergquist P (1978) Sponges. Hutchinson & Co, London Burton M (1947) Non-sexual reproduction in sponges, with special reference to a collection of young Geodia. Proc Linn Soc London 160: 163-178 Bmstedt U (2000) Life cycle, seasonal vertical distribution and feeding of Calanus finmarchicus in Skagerrak coastal water. Mar Biol 137: 279-289 Corriero G, Sar M, Vaccaro P (1996) Sexual and asexual reproduction in two species of Tethya (Porifera: Demospongiae) from a Mediterranean coastal lagoon. Mar Biol 126: 175-181 Corriero G, Scalera LL, Nonnis MC, Gaino E (1998) Reproductive strategies of Mycale contarenii (Porifera: Demospongiae). Mar Biol 131: 319-327 Diaz JP (1973) Cycle sexuel de deux demosponges de letang de Thau: Suberites massa Nardo et Hymeniacidon sanguinea Bowerbank. Bull Soc Zool France 98: 145-156 Ereskovsky AV (2000) Reproduction cycles and strategies of the cold-water sponges Halisarca dujardini (Demospongiae, Halisarcida), Myxilla incrustans and Iophon piceus (Demospongiae, Poecilosclerida) from the White Sea. Biol Bull 198: 77-87 Fell PE (1974) Porifera. In: Giese AC, Pearse JS (eds). Reproduction of marine invertebrates 1. Academic Press, New York. pp 51-132 Fell PE (1976) The reproduction of Haliclona loosanoffi and its apparent relationship to water temperature. Biol Bull 150: 200210 Fell PE (1983) Oogenesis, Oviposition and Oosorption. In: Adiyodi KG, Adiyodi RG (eds). Reproductive biology of invertebrates. Volume 1. John Wiley & Sons, Chichester. pp. 1-29

Gaino E, Burlando B, Buffa P, Sar, M (1986) Ultrastructural study of spermatogenesis in Oscarella lobularis (Porifera, Demospongiae). Int J Inverteb Reprod Dev 10: 297-305 Hoffmann F, Rapp HT, Zller T, Reitner J (2003). Growth and regeneration in cultivated fragments of the boreal deep water sponge Geodia barretti Bowerbank, 1858 (Geodiidae, Tetractinellida, Desmospongiae). J Biotechnol 100: 109-118 Ilan M (1995) Reproductive biology, taxonomy and aspects of chemical ecology of Latrunculiidae (Porifera). Biol Bull 188: 306312 Kaye HR (1990) Sexual reproduction in four Caribbean commercial sponges. II. Oogenesis and transfer of bacterial symbionts. Invertebr Reprod Dev 19(1): 13-24 Kaye HR, Reiswig HM (1991) Sexual reproduction in four Caribbean commercial sponges. I. Reproductive cycles and spermatogenesis. Inverteb Reprod Dev 19(1): 1-11 Klitgaard AB, Tendal OS (2004) Distribution and species composition of mass occurrences of large-sized sponges in the northeast Atlantic. Progr Oceanogr 61: 57-98 Lepore E, Sciscioli M, Scalera LL, Santarelli G (2000) Sexual reproduction of Cinachyra tarentina (Porifera, Demospongiae). Ital J Zool 67: 153-158 Leys S, Ereskovsky AV (2006) Embryogenesis and larval differentiation in sponges. Can J Zool 84: 262-287 Matthews JBL, Sands NJ (1973) Ecological studies on the deepwater community of Korsfjorden, Western Norway. Sarsia 52: 2952 Mercurio M, Corriero G, Gaino E (2007) A 3-year investigation of sexual reproduction in Geodia cydonium (Jameson, 1811) (Porifera, Demospongiae) from a semi-enclosed Mediterranean bay. Mar Biol 151: 1491-1500 Reiswig HM (1970) Porifera. Sudden sperm release by tropical Demospongiae. Science 170: 538-539 Scalera LL, Sciscioli M (1970) The sexual cycle of Erylus discophorus (Schmidt) (Porifera, Tetractinellida). Riv Biol 63: 265-270 Sciscioli M, Lepore E, Gherardi M, Scalera LL (1994) Transfer of symbiotic bacteria in the mature oocyte of Geodia cydonium (Porifera, Demospongiae): an ultrastructural study. Cah Biol Mar 35: 471-478 Sciscioli M, Lepore E, Mastrodonato M, Scalera LL, Gaino E (2002) Ultrastructural study of the mature oocyte of Tethya aurantium (Porifera, Demospongiae). Cah Biol Mar 43: 1-7 Simpson TL (1968) The cell biology of sponges. Springer-Verlag, New York Tanaka-Ichihara K, Watanabe Y (1990) Gametogenetic cycle in Halichondria okadai. In: Rtzler K (ed). New perspectives in sponge biology. Smithsonian Institution Press, Washington DC. pp 170-174 Wassmann P (1991) Dynamics of primary production and sedimentation in shallow fjords and polls of western Norway. Oceanography. Mar Biol Ann Rev 29: 87-157 Witte U, Barthel D (1994) Reproductive cycle and oogenesis of Halichondria panicea (Pallas) in Kiel Bight. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp 297-305 Witte U, Barthel D, Tendal O (1994) The reproductive cycle of the sponge Halichondria panicea (Pallas, 1766) and its relationships to temperature and salinity. J Exp Mar Biol Ecol 183: 41-52

Porifera research: Biodiversity, innovation and sustainaBility - 2007

621

Phylogenetic relationships among the filamentous cyanobacterial symbionts of Caribbean sponges and a comparison of photosynthetic production between sponges hosting filamentous and unicellular cyanobacteria
Robert W. Thacker(1*), Maria Cristina Diaz(2), Klaus Rtzler(3), Patrick M. Erwin(1), Steven J.A. Kimble(1), Melissa J. Pierce(1), Sandra L. Dillard(1)
Department of Biology, University of Alabama at Birmingham, Birmingham, AL 35294-1170, USA. thacker@uab.edu, erwin@uab.edu, sjkimble@uab.edu, mslissa@uab.edu, leedill@uab.edu (2) Museo Marino de Margarita, Blvd. El Paseo, Boca del Ro, Margarita, Edo. Nueva Esparta, Venezuela. crisdiaz@ix.netcom.com (3) National Museum of Natural History, Smithsonian Institution, Washington, D.C. 20560-0163, USA. ruetzler.klaus@nmnh.si.edu
(1)

Abstract: We investigated the filamentous cyanobacteria associated with two newly described species from the Caribbean coast of Panam, Haliclona walentinae and Xestospongia bocatorensis. In addition to sequencing cyanobacterial 16S ribosomal RNA genes from Hyrtios violaceus, H. walentinae and X. bocatorensis, we measured the chlorophyll a content of H. walentinae and X. bocatorensis as an index of symbiont abundance. The photosynthetic and respiration rates of these two associations were compared to those of two sympatric sponges that host unicellular cyanobacteria, Aplysina fulva and Neopetrosia subtriangularis. A phylogeny of 16S ribosomal RNA genes reveals that the symbionts of H. violaceus, H. walentinae and X. bocatorensis are part of the O. spongeliae clade and that each sponge hosts a unique ribotype of this cyanobacterium. H. walentinae yielded the highest chlorophyll a concentrations, while X. bocatorensis, A. fulva, and N. subtriangularis were not significantly different. All sponges measured showed gross productivity to respiration (P:R) ratios greater than 1.5, indicating that cyanobacterial photosynthesis can compensate for host sponge respiration, and that all four species can be considered phototrophic. X. bocatorensis yielded the highest P:R ratios, while those of H. walentinae, A. fulva, and N. subtriangularis were not significantly different. Specialized associations with filamentous cyanobacteria may provide a valuable source of carbon to host sponges. These associations occur over a broader phylogenetic range of hosts than previously described, including representatives of the orders Dictyoceratida and Haplosclerida. Keywords: cyanobacteria, molecular systematics, photosynthesis, phylogeny, symbioses

Introduction
Many marine sponges host diverse communities of extracellular symbiotic bacteria (Wilkinson et al. 1981, Hentschel et al. 2006). These symbionts may contribute to the host sponges metabolism by providing nutrients such as fixed carbon or nitrogen (Wilkinson and Fay 1979, Rai 1990, Sar et al. 1998). In particular, symbiotic cyanobacteria may provide their hosts with the products of photosynthesis. Cyanobacteria constitute nearly 50% of the biomass of some sponges, with up to 50% of the hosts energy budget and 80% of the hosts carbon budget derived from symbiont photosynthesis (Wilkinson 1983, Cheshire et al. 1997). Wilkinson (1987) suggested that these phototrophic symbioses were more prevalent in the Indo-Pacific region

than in the Caribbean, concluding that Caribbean species were rarely phototrophic. Two groups of cyanobacteria are most commonly reported as associates of marine sponges: unicellular cyanobacteria currently classified as Candidatus Synechococcus spongiarum and filamentous cyanobacteria classified as Oscillatoria spongeliae (Steindler et al. 2005, Diaz et al. 2007). Unicellular, S. spongiarum-like symbionts were previously described as Aphanocapsa feldmani by investigators using electron microscopy to identify sponge symbionts (Rtzler 1990, Usher et al. 2006). Subsequent investigations based on molecular phylogenetic techniques have placed these symbionts into the genus Synechococcus (Usher et al. 2004, Thacker 2005). Although Usher et al. (2001) provided evidence for vertical transmission of these

622

symbionts, molecular phylogenetic analyses based on 16S ribosomal RNA (rRNA) gene sequences suggest there is no specialization of these symbionts for particular host species (Steindler et al. 2005, Thacker 2005). Common Caribbean hosts of S. spongiarum include Aplysina fulva (Pallas, 1766) and Neopetrosia subtriangularis (Duchassaing, 1850). Filamentous Oscillatoria spongeliae have been previously reported from a variety of Indo-Pacific sponges based on both morphological (Rtzler 1990, Diaz 1996) and molecular (Thacker and Starnes 2003, Thacker 2005, Ridley et al. 2005) evidence. These hosts include members of the dictyoceratid genera Dysidea, Lamellodysidea, Lendenfeldia, and Phyllospongia. O. spongeliae filaments are extracellular, approximately 10 m wide, and range in length from 5 to 50 cells (Hinde et al. 1994). Previous research has documented the evolutionary specialization of these symbionts for particular host sponges. Molecular phylogenies of the IndoPacific hosts and their symbionts reveal that each sponge species hosts a unique clade of O. spongeliae, and suggest that cospeciation occurs between hosts and symbionts (Thacker and Starnes 2003, Ridley et al. 2005). Filamentous cyanobacteria have previously been reported as symbionts from only two Caribbean sponges: Hyrtios violaceus (Duchassaing and Michelotti, 1864) (formerly Oligoceras hemorrhages de Laubenfels, 1936), which is common in the Bahamas and Belize (Wiedenmayer 1977, Rtzler 1990), and an undescribed species of Niphates, also from the Bahamas and Belize (Diaz 1996). From the Bocas del Toro region of Panam, Diaz et al. (2007) have described two additional Caribbean sponges that host filamentous cyanobacteria: Haliclona walentinae and Xestospongia bocatorensis. The objectives of our study were to PCRamplify and sequence cyanobacterial 16S rRNA genes from three of these sponges: H. violaceus, H. walentinae, and X. bocatorensis. We compared these sequences to known sequences from O. spongeliae to examine their host specificity and phylogenetic relationships. For the two species from Panam, we measured chlorophyll a concentrations to estimate the abundance of these cyanobacteria in their host sponges and measured photosynthetic and respiration rates to determine whether these sponges are phototrophic. We tested the hypotheses that sponges hosting filamentous cyanobacterial symbionts (1) contain higher concentrations of chlorophyll a and (2) have higher ratios of gross photosynthetic production to respiration than sympatric sponges hosting unicellular cyanobacterial symbionts.

Methods
Specimens of Hyrtios violaceus were collected from shallow reefs at Twin Cays, Belize, near the Smithsonian Institutions research station at Carrie Bow Cay, Belize. Specimens of Haliclona walentinae, Xestospongia bocatorensis, Aplysina fulva and Neopetrosia subtriangularis were collected from shallow reefs, between 2 and 5 m depth, near the Smithsonian Tropical Research Institutes Bocas Research Station, Bocas del Toro, Panam. Sponges were preserved in 95% ethanol and RNAlater (Ambion, Inc.). Genomic DNA extractions were prepared from preserved sponges (2 H. walentinae individuals, 4 X. bocatorensis

individuals, and 3 H. violaceus individuals) using the Wizard Genomic DNA Purification Kit, following the manufacturers protocol (Promega Corporation, Madison, WI). Nearly complete bacterial 16S rRNA genes were amplified from these extracts using universal bacterial primers (MartinezMurcia et al. 1995). PCR products were inserted into plasmid vectors using the pGEM T-Easy Vector System (Promega). Representative clones from each individual sponge were screened for the presence of cyanobacterial 16S rRNA genes using cyanobacteria-specific PCR primers (Nbel et al. 1997). Three positive clones from each individual sponge were sequenced at the UAB CFAR DNA Sequencing Core Facility, using the plasmids sequencing primers and internal sequencing primers (Nbel et al. 1997). Sequences were assembled and aligned using Sequencher 4.1 (GeneCodes, Ann Arbor, MI) and Se-Al (Rambaut, University of Oxford). For each individual sponge, a single consensus sequence was constructed from the three sequenced clones, since in all cases these clones showed less than 1% sequence divergence within an individual host. Several reference sequences from GenBank were included in the 1,416 bp alignment, including Oscillatoria spongeliae from other marine sponges and two outgroup sequences, O. cf. corallinae and Arthrospira sp. PCC 7345 (Fig. 1). Phylogenetic analyses included a distance-based neighborjoining approach conducted with MEGA version 3.1 (Kumar et al. 2004) using the Kimura 2-parameter model of nucleotide substitutions with 500 bootstrap replicates. We also performed a maximum likelihood (ML) phylogenetic analysis using GARLI version 0.951 (Zwickl 2006; download available at: http://www.bio.utexas.edu/faculty/antisense/garli/Garli. html), including 100 bootstrap replicates. The hierarchical Akaike information criterion (AIC) implemented by Modeltest 3.7 (Posada and Crandall 1998) was used to select the best model of DNA substitution, the general time reversible model with proportion of invariable sites and a gamma distribution of variable substitution rates among variable sites (GTR+I+G). For Bayesian phylogenetic analyses, MrBayes 3.1.2 (Ronquist and Huelsenbeck 2003) was used to calculate the posterior probabilities of branch nodes, implementing the GTR+I+G likelihood model. The Monte Carlo Markov chain length was set at 500,000 generations with sampling every 100th generation and a burn-in value of 1250 cycles. After 250,000 generations, the average standard deviation of split frequencies reached less than 0.01. For H. walentinae and X. bocatorensis, chlorophyll a concentrations and photosynthetic production were compared to two sympatric sponges, A. fulva and N. subtriangularis. Microscopic examinations of all four sponges did not reveal the presence of photosynthetic eukaryotes; thus, chlorophyll a concentrations are directly correlated with the abundance of cyanobacteria within a sponge, as reported by other investigators (Wilkinson 1983, Rai 1990). For each specimen, approximately 0.25 g (wet mass) of finely chopped sponge ectosome was placed in 10 ml of a 90% acetone:water mixture and held overnight at 4C. Each sample was briefly spun in a centrifuge to remove suspended solids, after which the supernatant was transferred to a cuvette and absorbance measured at 750, 664, 647, and 630 nm in a spectrophotometer. Chlorophyll a concentrations were calculated based on

623

Fig. 1: Neighbor-joining phylogeny of 16S ribosomal RNA gene sequences amplified from Oscillatoria spongeliae symbionts of marine sponges. Numbers at each node represent percentage bootstrap support from 500 replicates of a neighbor-joining analysis, percentage bootstrap support from 100 replicates of a maximum-likelihood analysis, and percentage Bayesian posterior probabilities, respectively. Asterisks indicate less than 50% support. Branch tips are labeled with the identity of the host sponge and GenBank accession number, while shaded bars indicate the genus, family, and order of the host sponges. Scale bar indicates number of substitutions per site based on the Kimura two-parameter model of nucleotide substitutions.

equations provided by Parsons et al. (1984) and standardized to sponge mass. Differences in chlorophyll a concentrations among sponge species and symbiont type (unicellular vs. filamentous) were compared using a nested analysis of variance (ANOVA), with species nested within symbiont type. Post hoc comparisons among species were conducted using Fishers least significant difference test. Photosynthetic and respiration rates were measured for fragments of sponges that were incubated in 0.5 l bottles filled with filtered seawater. Each fragment was incubated sequentially in a light bottle, which allowed approximately 75% of ambient light (an average of 1000 mol quanta / m2 / s during this experiment) to reach the sponge, and a dark bottle, which allowed no light to reach the sponge.

Immediately prior to each incubation, the initial oxygen concentration in each bottle was measured using a YSI Model 85 oxygen meter. After 1 hour of incubation in a water bath, which maintained a temperature of 30C (reflecting the ambient water temperature), final oxygen concentrations were measured. Respiration was calculated as the change in oxygen concentrations in the dark bottles, standardized by sponge wet mass, while net photosynthesis was calculated from the change in oxygen concentrations in the light bottles, standardized by sponge wet mass. Gross photosynthesis was calculated as the difference between respiration and net photosynthesis. The gross production to respiration ratio (P: R; Wilkinson 1983) was calculated as gross photosynthesis divided by respiration. Differences in P:R ratios among sponge

624

species and symbiont type (unicellular vs. filamentous) were compared using a nested analysis of variance (ANOVA), with species nested within symbiont type. Post hoc comparisons among species were conducted using Fishers least significant difference test.

Results
The cyanobacterial symbionts of Hyrtios violaceus, Haliclona walentinae, and Xestospongia bocatorensis yielded 16S rRNA gene sequences that shared 96.7% to 99.4% identity with other sequences obtained from Oscillatoria spongeliae. Sequences have been deposited in GenBank under accession numbers EF537054 to EF537062. Each sponge species hosts a unique and well-supported monophyletic clade of O. spongeliae (Fig. 1), clearly illustrating the high degree of host-specificity observed for this symbiont. The H. violaceus symbiont is most similar to that of Lendenfeldia chondrodes, and is part of a monophyletic clade that includes the symbionts of Lamellodysidea chlorea and Lamellodysidea herbacea. The symbionts hosted by H. walentinae and X. bocatorensis were very similar, showing only 0.78% sequence divergence. These symbionts formed a monophyletic clade with those of Dysidea granulosa and Phyllospongia papyracea. Chlorophyll a concentrations did not vary significantly among sponge species nested within symbiont types (F = 3.405, df = 2, P = 0.059), but did vary significantly between symbiont types (F = 5.548, df = 1, P = 0.032). However, post hoc tests revealed that this pattern was driven by a single species. H. walentinae yielded significantly higher chlorophyll a concentrations than the other three sponges, creating the significant difference between symbiont types (Fig. 2). Gross photosynthetic production to respiration (P:R; Fig. 3) ratios varied significantly among sponge species nested within symbiont types (F = 7.314, df = 2, P = 0.007) and between sponges hosting filamentous and unicellular cyanobacteria (F = 18.508, df = 1, P = 0.001). Post hoc tests revealed that this pattern was also driven by a single species, since X. bocatorensis yielded the highest P:R ratio, while those of the other three sponges were not significantly different. All of these sponges showed P:R ratios greater than 1.0, indicating that cyanobacterial photosynthesis can compensate for sponge respiration. In addition, Wilkinson (1987) defined phototrophic sponges as having a P:R ratio greater than 1.5; all four of these Caribbean sponges can be considered phototrophic under this definition.

Fig. 2: Chlorophyll a concentrations (mean SE) measured in Caribbean sponges that host filamentous cyanobacterial symbionts (Xestospongia bocatorensis and Haliclona walentinae; open bars) and unicellular symbionts (Aplysina fulva and Neopetrosia subtriangularis; shaded bars). Five individuals were sampled per species; different letters above bars indicate significantly different means.

Fig. 3: Gross photosynthetic production to respiration ratios (mean SE) measured in Caribbean sponges that host filamentous cyanobacterial symbionts (Xestospongia bocatorensis and Haliclona walentinae; open bars) and unicellular symbionts (Aplysina fulva and Neopetrosia subtriangularis; shaded bars). Numbers in parentheses indicate the number of individuals sampled; different letters above bars indicate significantly different means.

Discussion
Microscopic examination of the Caribbean sponges Hyrtios violaceus, Haliclona walentinae, and Xestospongia bocatorensis revealed that all three of these species host abundant filamentous cyanobacterial symbionts (Rtzler 1990, Diaz et al. 2007). These symbionts are morphologically similar to the symbionts classified as Oscillatoria spongeliae that are hosted by Indo-Pacific sponges in the genera Dysidea, Lamellodysidea, Lendenfeldia, and Phyllospongia (Thacker and Starnes 2003, Ridley et al. 2005). The molecular phylogenetic analyses described in this study confirm that the cyanobacterial symbionts of H. violaceus, H. walentinae, and

X. bocatorensis are members of the Oscillatoria spongeliae clade. Moreover, each of these sponge species hosts a unique subclade or ribotype of the cyanobacterium, as has been demonstrated for the Indo-Pacific hosts (Thacker and Starnes 2003, Ridley et al. 2005). Previous molecular phylogenies of the Indo-Pacific hosts and their symbionts supported the hypothesis of cospeciation between sponges and O. spongeliae (Thacker and Starnes 2003, Ridley et al. 2005). Ridley et al. (2005) found qualitative evidence of only one host-switching event among the IndoPacific hosts; however, this single event was not supported by statistical analyses. The addition of the Caribbean taxa to the phylogenetic tree of O. spongeliae raises questions about the

625

hypothesis of cospeciation. If a thorectid ancestor was initially colonized by O. spongeliae, two independent colonizations of Dysideidae and two independent colonizations of two haplosclerid taxa are needed to generate the observed phylogenetic pattern (Fig. 1). In addition the symbionts of the Caribbean sponges are derived from two different lineages of O. spongeliae. Data on the molecular systematics of the Caribbean host sponges are clearly needed to quantify the contribution of sponge hosts to these patterns. However, the current data set provides qualitative support for a hypothesis of independent colonization of these hosts by the symbionts, and suggests that cospeciation events might only occur within genera. Cyanobacterial abundance within a host sponge is directly correlated with chlorophyll a concentrations (Wilkinson 1983, Rai 1990). Since photoacclimation can also influence chlorophyll a concentrations within cyanobacterial cells (MacIntyre et al. 2002), all sponges used in this study were collected from similar light environments between 2 and 5 m depth. Chlorophyll a concentrations varied significantly among H. walentinae, X. bocatorensis, and two sympatric sponges that host the unicellular cyanobacterial symbionts classified as Candidatus Synechococcus spongiarum (Usher et al. 2004, Steindler et al. 2005), Aplysina fulva and Neopetrosia subtriangularis. H. walentinae yielded a higher concentration of chlorophyll a than the other three sponges, indicating that it may host a higher density of symbiotic cyanobacteria. On average, sponges hosting filamentous cyanobacteria contained a higher concentration of chlorophyll a than sponges hosting unicellular cyanobacteria. However, since our comparisons only included two species hosting each symbiont type, and since this pattern was driven by the extremely high concentration of chlorophyll a in a single species, this hypothesis clearly needs to be tested with a wider range of host species. The observed variation in chlorophyll a concentrations within and among host species and between symbiont types may reflect variation in the abundance of cyanobacteria, which may reflect variation in the costs and benefits associated with these symbioses (Thacker 2005). Our current data can also be compared to chlorophyll a concentrations measured in other sponges (e.g., Wilkinson 1983, Thacker 2005); however, there is a large degree of variability in chlorophyll extraction and analysis methods among studies. For example, in Palau, Lamellodysidea chlorea (which hosts O. spongeliae) contained 685 84 (mean SE) g chlorophyll a / g sponge, while Neopetrosia exigua (Kirkpatrick, 1900) (which hosts S. spongiarum) contained 610 40 g chlorophyll a / g sponge (Thacker 2005). These values are greater than twice the average of the Caribbean sponges sampled in this study, indicating a large amount of variation that could be due not only to differences in sample processing (the Palauan sponges were freeze-dried prior to chlorophyll a extraction) but also to biogeographic factors that influence chlorophyll a concentrations and/or cyanobacterial abundance. These sources of variation can be larger than the variation observed between particular hosts and symbionts within a single study. Previous experiments used shading to manipulate the interactions between L. chlorea and O. spongeliae and between N. exigua and S. spongiarum (Thacker 2005). L. chlorea

was dependent on photosynthesis for survival and growth, rapidly dying when shaded, while N. exigua did not suffer any reduction in survival or growth when shaded, indicating that filamentous symbionts may make larger contributions to their hosts than unicellular symbionts. Based on these results, we hypothesized that sponges hosting filamentous symbionts would have higher gross production to respiration (P:R) ratios than those hosting unicellular symbionts. On average, sponges hosting filamentous symbionts did display higher P:R ratios than those hosting unicellular symbionts; however, this pattern was driven by a single species, X. bocatorensis. Thus, this hypothesis should also be tested with a larger selection of host species. The large variability in P:R ratios, also observed by Wilkinson (1983), may be biologically significant, potentially illustrating the changing balance of costs and benefits involved in these symbioses. All four sponges showed that cyanobacterial photosynthesis can more than compensate for sponge respiration. Furthermore, according to Wilkinsons (1987) definition, all four of these species can be considered phototrophic, suggesting that more phototrophic sponges exist in the Caribbean than previously acknowledged. Rtzler (unpublished) measured P:R ratios for H. violaceus in Belize, and found values of 4.6 (full sunlight), 3.4 (lightly overcast), and 2.2 (overcast), concluding that this species is also phototrophic. The phylogenetic data obtained in this study confirms that the filamentous cyanobacteria found within three species of Caribbean sponges are members of the Oscillatoria spongeliae clade, with each sponge hosting a unique subclade or ribotype of symbiont. The current phylogeny emphasizes multiple colonizations or host-switches over cospeciation; however, there is a clear need to obtain the host phylogeny to quantify these patterns. In general, filamentous symbionts may provide a larger contribution to sponge metabolism than unicellular symbionts (Thacker 2005); however, since we found very similar P:R ratios for A. fulva and H. walentinae, such generalizations may ignore the substantial variation observed among host sponges. This variability may reflect a dynamic relationship between sponge hosts and their cyanobacterial symbionts, which may be strongly influenced by the type of symbiont, the host sponge species, and environmental conditions.

Acknowledgements
Logistical support for this project was generously provided by Rachel Collin and the Smithsonian Tropical Research Institutes Bocas Research Station, as well as by the Smithsonian Institutions Caribbean Coral Reef Ecosystems Program (CCRE Contribution no. 774). Portions of this project were completed during the 2005 Taxonomy and Ecology of Caribbean Sponges course held at the Bocas Research Station. Maria Salazar and the UAB CFAR DNA Sequencing Core Facility provided DNA sequencing support; this work has been facilitated by the infrastructure and resources provided by the NIH CFAR Core Grant P30 AI27767. This material is based upon work supported by the National Science Foundation under Grant No. 0209329 awarded to RWT.

626

References
Cheshire AC, Wilkinson CR, Seddon S, Westphalen G (1997) Bathymetric and seasonal changes in photosynthesis and respiration of the phototrophic sponge Phyllospongia lamellosa in comparison with respiration by the heterotrophic sponge Ianthella basta on Davies Reef, Great Barrier Reef. Mar Freshwater Res 48: 589-599 Diaz MC, Thacker RW, Rtzler K, Piantoni C (2007) Two new haplosclerid sponges from Caribbean Panama with symbiotic filamentous cyanobacteria, and an overview of spongecyanobacteria associations. In: Custdio MR, Lbo-Hajdu G, Hajdu E, Muricy G (eds). Porifera research: biodiversity, innovation and sustainability. Srie Livros 28. Museu Nacional, Rio de Janeiro. pp. 31-39 Diaz MC (1996) Molecular and ecological studies of spongemicrobial associations. Ph.D. Thesis, University of California, Santa Cruz. Hentschel U, Usher KM, Taylor MW (2006) Marine sponges as microbial fermenters. FEMS Microbiol Ecol 55: 167-177 Hinde RF, Pironet F, Borowitzka MA (1994) Isolation of Oscillatoria spongeliae, the filamentous cyanobacterial symbiont of the marine sponge Dysidea herbacea. Mar Biol 119: 99-104 Kumar S, Tamura K, Nei M (2004) MEGA3: Integrated software for molecular evolutionary genetics analysis and sequence alignment. Brief Bioinfor 5: 150-163 MacIntyre HL, Kana TM, Anning T, Geider RJ (2002) Photoacclimation of photosynthesis irradiance response curves and photosynthetic pigments in microalgae and cyanobacteria. J. Phycol. 38: 17-38 Martnez-Murcia AJ, Acinas SG, Rodriguez-Valera F (1995) Evaluation of prokaryotic diversity by restrictase digestion of 16S rDNA directly amplified from hypersaline environments. FEMS Microbiol Ecol 17: 247-256 Nbel U, Garcia-Pichel F, Muyzer G (1997) PCR primers to amplify 16S rRNA genes from cyanobacteria. Appl Environ Microbiol 63: 3327-3332 Parsons TR, Maita Y, Lalli C (1984) A manual of chemical and biological methods for seawater analysis. Pergamon Press, New York Posada D, Crandall KA (1998) Modeltest: testing the model of DNA substitution. Bioinformatics 14: 817-818 Rai AN (1990) CRC Handbook of symbiotic cyanobacteria. CRC Press, Boca Raton Ridley CP, Bergquist PR, Harper MK, Faulkner DJ, Hooper JN, Haygood MG (2005) Speciation and biosynthetic variation in

four dictyoceratid sponges and their cyanobacterial symbiont, Oscillatoria spongeliae. Chem Biol 12: 397-406 Ronquist F, Huelsenbeck JP (2003) MrBayes3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19: 1572-1574 Rtzler K (1990) Associations between Caribbean sponges and photosynthetic organisms. In: Rtzler K (ed). New Perspectives in Sponge Biology. Smithsonian Institution Press, Washington, DC. pp. 455-466 Sar M, Bavestrello G, Cattaneo-Vietti R, Cerrano C (1998) Endosymbiosis in sponges: Relevance for epigenesis and evolution. Symbiosis 25: 5770 Steindler L, Huchon D, Avni A, Ilan M (2005) 16S rRNA phylogeny of sponge-associated cyanobacteria. Appl Environ Microbiol 71(7): 4127-4131 Thacker RW (2005) Impacts of shading on sponge-cyanobacteria symbioses: a comparison between host-specific and generalist associations. Int Comp Biol 45: 369-376 Thacker RW, Starnes S (2003) Host specificity of the symbiotic cyanobacterium Oscillatoria spongeliae in marine sponges, Dysidea spp. Mar Biol 142: 643-648 Usher KM, Kuo J, Fromont J, Sutton DC (2001) Vertical transmission of cyanobacterial symbionts in the marine sponge Chondrilla australiensis (Demospongiae). Hydrobiologia 461: 1523. Usher KM, Toze S, Fromont J, Kuo J, Sutton DC (2004) A new species of cyanobacterial symbiont from the marine sponge Chondrilla nucula. Symbiosis 36: 183-192 Usher KM, Kuo J, Fromont J, Toze S, Sutton DC (2006) Comparative morphology of five species of symbiotic and non-symbiotic coccoid cyanobacteria. European J Phycol 41(2): 179-188 Wiedenmayer F (1977) Shallow-water sponges of the western Bahamas. Experientia Suppl 28: 1-287 Wilkinson CR (1983) Net primary productivity in coral reef sponges. Science 219: 410-412 Wilkinson CR (1987) Interocean differences in size and nutrition of coral reef sponge populations. Science 236: 1654-1657 Wilkinson CR, Fay P (1979) Nitrogen fixation in coral reef sponges with symbiotic cyanobacteria. Nature 279: 527-529 Wilkinson CR, Nowak M, Austin B, Colwell RR (1981) Specificity of bacterial symbionts in Mediterranean and Great Barrier Reef sponges. Microbial Ecol 7: 13-21 Zwickl DJ (2006) Genetic algorithm approaches for the phylogenetic analysis of large biological sequence datasets under the maximum likelihood criterion. Ph.D. dissertation, The University of Texas at Austin

Porifera research: Biodiversity, innovation and sustainaBility - 2007

627

Chemical defense strategies in sponges: a review

Carsten Thoms, Peter J. Schupp
University of Guam Marine Laboratory, UOG Station, Mangilao, Guam 96923, USA. cthoms@guam.uog.edu, pschupp@guam.uog.edu Abstract: Sponges, as well as other sessile marine invertebrates, share numerous ecological features with plants and have evolved similar strategies to defend themselves against threats from their biotic environment. Chemical defense plays a preeminent role in this context. General concepts from plant chemical ecology applied to sponges have revealed interesting parallels in regard to resource investment, defense compound allocation, and synergism between chemical and structural defenses. However, these concepts often cannot be generalized in sponges since also numerous contradictory examples exist. Sponges frequently use their compounds as multi-purpose tools, with concurrent activity against various threats (e.g., predation, pathogens, and biofouling). Apparently, a latitudinal gradient in chemical deterrence, as it was previously shown for marine algae, is lacking in sponges. Recently, facultative defenses (i.e., activated and inducible defenses; immune reactions) have received increased attention in sponge chemical ecology. Although there are examples that clearly demonstrate that these strategies exist in sponges, their number is still very low. One focus of this review is laid on the discussion of the various difficulties inherent to experiments in regards to facultative defenses in sponges that may explain why, as yet, only few studies have found compelling evidence for their existence. Keywords: activated defense, chemical ecology, growth-differentiation balance hypothesis, inducible defense, optimal defense theory

Introduction
The ecological factors structuring sponge communities resemble in many respects that of plants. Sponges are often abundant and apparent in the habitat they grow in (Bergquist 1978, McClintock et al. 2005), lack behavioral defenses, and in many cases they are autotrophic (due to photosynthetic symbionts) (Arillo et al. 1993, Usher et al. 2001, Hentschel et al. 2006). Similar to plants, they are most of the times nonfatally grazed by predators. It is, therefore, not surprising that sponges also have evolved defense strategies similar to those known from terrestrial and aquatic plants. Already in the early 1950s it was discovered that sponges yield secondary metabolites with pronounced bioactivity (Bergmann and Feeney 1950). Since then, more than 5000 compounds have been isolated from sponges (Blunt and Munro 2003). Numerous ecological studies have shown that they often serve defensive purposes to protect the sponges from threats such as predator attacks, microbial infections, biofouling, and overgrowth by other sessile organisms (reviewed in McClintock and Baker 2001, Paul and Puglisi 2004, Paul et al. 2006). However, there is also a large number of sponge secondary metabolites with no apparent ecological function. One theory addressing this issue suggests that some secondary metabolites simply do not have any ecological function, but rather represent evolutionary baggage (Jones and Firn 1991, McClintock and Baker 1998). One the other hand, one has to acknowledge that marine natural products research has been and still is driven by pharmacological screening programs, which aim at the discovery of new chemical structures with

pharmacological activity rather than investigating possible ecological functions of these compounds. The results of recent ecological studies indicate that in addition to the simple storage of chemical weapons in their tissues sponges have evolved mechanisms to increase the efficiency of their chemical defense, to save metabolic energy invested in their defense, and to protect themselves from cell damage caused by their own bioactive defense compounds. In the following, we review studies that describe such chemical defense strategies in sponges.

Constitutive defense Optimal defense theory and growth-differentiation balance hypothesis in sponges
The first reports on predation on sponges appeared in the 1960s (Bakus 1966, Randall and Hartman 1968). Nowadays, it is generally accepted that predation has a major impact on sponge ecology, and that sponge populations can be significantly reduced by predators if they are not sufficiently protected (e.g., Hill 1998, Pawlik 1998, Hill et al. 2005, Wulff 2006). Chemical defense undoubtedly ranges among the most important anti-predator strategies of sponges (e.g., Pawlik et al. 1995, Uriz et al. 1996, Wright et al. 1997). This holds especially true in habitats with high predation pressure, such as tropical coral reefs (Pawlik 1998), where highly mobile predators such as fishes or turtles can quickly remove substantial biomass, compared to more urchin dominated grazing in temperate regions, or starfish dominated grazing

628

in polar regions (Wright et al. 1997, McClintock and Baker 2001, Davis et al. 2003). For plants, there are two prominent hypotheses to explain spatial and temporal variation in defense expression, the growth-differentiation balance hypothesis (GDBH) and the optimal defense theory (ODT). The GDBH assumes that a balance must be maintained between resources invested in growth and in differentiation (which includes the production of defense compounds) (Stamp 2004, Barto and Cippolini 2005). A key premise of this hypothesis is that defense is costly. Due to the high structural complexity of defense compounds in sponges it can be assumed that in many cases their biosynthesis is, indeed, metabolically expensive (Paul 1992, Pawlik 1993). Whereas few studies have assessed the metabolic costs of sponge chemical defense so far, several authors have investigated the interrelations between sponge growth and the investment in chemical defense. Turon et al. (1998) reported seasonal patterns in growth rate and toxicity level of the Mediterranean sponge Crambe crambe and observed a significant negative correlation between these parameters. Moreover, they found that C. crambe growing in shaded areas had lower growth rates but invested more resources in chemical defense than individuals growing in well-illuminated habitats (Turon et al. 1998). Wulff (2005) reported a positive correlation between growth rates and palatability to fish predators of twelve reef and mangrove sponge species. Walters and Pawlik (2005) found that sponge species with a pronounced chemical protection had slower wound-healing rates than chemically unprotected species. The authors ascribed this to a trade-off between investment of resources in chemical defense and tissue growth (Walters and Pawlik 2005). However, it has to be noted that growth rates and regeneration capabilities in sponges may not necessarily be equated but can differ substantially in the same species (Reiswig 1973, Ayling 1983). This, in turn, raises interesting questions about differences in resource allocation to each of these processes in relation to investment in anti-predator defenses. The optimal defense theory (ODT) postulates that defenses are primarily allocated to plant parts of high fitness value (e.g., reproductive tissues) and / or that have a higher risk of predation (Rhoades 1976). By restricting metabolite allocation to these areas rather than distributing them over the entire plant, biosynthesis costs may be lowered. Numerous, but not all studies on terrestrial plants support the ODT (see Baldwin and Ohnmeiss 1994, Zangerl and Rutledge 1996, Heil et al. 2002 in support of the ODT, and Zangerl 1986, Zangerl and Nitao 1998 contradicting the theory). Similarly, many but not all studies on marine organisms, such as algae (Cronin and Hay 1996, van Alstyne et al. 1999, Pavia et al. 2002, Toth et al. 2005), sea fans (Dube et al. 2002), mollusks (Avila and Paul 1997, Thoms et al. 2006a) and brachiopods (Mahon et al. 2003) are conform with this hypothesis. This also holds true for sponges, where no clear pattern regarding defense compound allocation has emanated so far. The Micronesian sponge Oceanapia sp. is an example that supports predictions of the ODT, as the sponge allocates the highest concentrations of the pyridoacridine alkaloids kuanoniamine C and D in tissue parts that are most apparent to predators and that most likely play a role in reproduction

(Eder et al. 1998, Schupp et al. 1999). Schupp et al. (1999) demonstrated in a series of field and laboratory experiments using different predators that both alkaloids were deterrent at natural concentrations towards generalist reef fish and the spongivorous angelfish Pomacanthus imperator. In detailed field experiments using the two major predators Becerro et al. (1998) found intracolonial variation of crude organic extracts containing the sesterterpenes scalaradial and desacetylscalaradial in the tropical sponge Cacospongia sp. The concentrations were highest at the sponge tips and in the ectosome. However, when tested against fish predators, even the lowest concentration of the extract found in the sponge tissue was already effective. The specialized nudibranch Glossodoris pallida, on the other hand, preferred pieces of Cacospongia base over tips, thereby selecting the chemically less defended sponge parts (Becerro et al. 1998). Latrunculia apicalis, a spherically shaped Antarctic sponge, is protected against the keystone spongivorous sea star Perknaster fuscus by the sequestration of discorhabdin G. Consistent with the ODT the concentration of this alkaloid decreases rapidly from the surface tissue of the sponge towards its core (Furrow et al. 2003). In the tropical sponge Ectyoplasia ferox the concentrations of defensive triterpene glycosides were found to be approximately twice as high in the outer 2 mm layer than in the deeper tissue layers of this sponge (Kubanek et al. 2002), a finding that again supports the ODT. However, the same study reported that in the sponge Erylus formosus concentrations of the defensive triterpene glycoside formoside were only about one-third as high in the outer 1 mm layer of the sponge as in its more interior layers (Kubanek et al. 2002). Swaeringen and Pawlik (1998), studying chemical gradients in sponge tissue as well as differences in antifeeding properties against predators in the field, found no evidence that deterrent compounds were concentrated in the surface tissues of the sponge Chondrilla nucula collected from the Bahamas and Florida. Becerro et al. (1995) found no differences in toxicity between the periphery and the central parts of the Mediterranean sponge Crambe crambe. However, in this study toxicity was only evaluated by a Microtox bioassay and no feeding experiments with predators in the field were conducted. Burns et al. (2003) reported no difference in deterrence towards the wrasse Thalassoma klunzingeri and the sea urchin Diadema setosum when they compared extracts of ectosome and endosome layers of six sponges from the Red Sea. Furrow et al. (2003) offer a possible explanation for the discrepancy between studies comparing inner and outer tissue layers of sponges in regard to the ODT: they suggest that sequestration of anti-predatory metabolites primarily to the outermost layers in Antarctic sponges such as L. apicalis could be highly adaptive because of the ubiquity of sea star sponge predators in Antarctic marine benthic environments. Other than fish, whose bites penetrate well below the sponge surface, sea stars feed on sponges by extrusion of the cardiac stomach for external digestion. This feeding behavior could be a particularly strong selective force for surface sequestration of chemical defenses (Furrow et al. 2003). Thus, differential distribution of defensive secondary metabolites to outer compared to inner layers may reflect the feeding behavior of the predominant predators in the respective habitat (i.e., surface feeding affecting only the ectosome versus biting

629

larger pieces including both ecto- and endosomal layers), and may, therefore, not always be present. Moreover, the often amorphous morphology and anatomy of sponges as well as their extraordinary ability to rapidly regenerate lost tissue after wounding may complicate the assignment of high fitness value to distinct parts of their body. This may further explain why defensive metabolite allocation in accordance with the ODT is less apparent in sponges than in other sessile organisms. It is conceivable that the efficiency of chemical defense can also be optimized by utilizing the same compound for different ecological purposes. Biosynthesis costs may be saved, if instead of producing several compounds for multiple purposes, only one metabolite is sequestered that is active against a variety of target organisms and other threats. However, Schmitt et al. (1995) pointed out that multiple uses of defensive compounds could limit adaptive changes following the evolution of resistance to these compounds by the affected organisms (e.g., predators). The first study to assess this topic in sponges was conducted by Thompson et al. (1985). They tested 28 compounds isolated from eight sponge species for a broad range of bioactivities including antimicrobial properties, inhibition of larval settlement, fish toxicity, inhibition of sexual reproduction, and antipredator activity. Most of the compounds tested showed activity in at least one assay, but usually they were active in several of these tests. Bobzin and Faulkner (1992) tested the metabolites manool and cholesterol endoperoxide isolated from the Bahamian sponge Aplysilla glacialis for their feeding deterrent and antifouling properties. Whereas the compounds significantly deterred feeding by fish, they actually increased the rate of fouling. Becerro et al. (1997) tested three fractions of different polarity of crude extracts from the sponges Crambe crambe and Hemimycale columella for their inhibitory activity against cell division, photosynthesis, and settlement of organisms growing in the same habitat. They found that several compounds in these fractions displayed multiple activities and concluded that secondary metabolites may be multi-purpose tools. Thacker et al. (1998) reported that 7-deacetoxyolepupuane, a secondary metabolite isolated from Dysidea sp., caused necrosis in the competing sponge Cacospongia sp., and additionally showed feeding-deterrent activity against fish. Newbold et al. (1999) observed that certain sponge crude extracts with anti-feeding activity against fishes (Pawlik et al. 1995) at the same time inhibited growth of marine bacteria. In a follow-up study, several of these extracts were also tested for activity against bacterial attachment (Kelly et al. 2003). Seven compounds from different sponges were isolated and identified that proved to be active in both deterring predators and inhibiting bacterial attachment (Kelly et al. 2003). Kubanek et al. (2002) reported multiple defensive roles for triterpene glycosides isolated from Erylus formosus and Ectyoplasia ferox, two Caribbean sponges belonging to different taxonomic orders. Formoside and other triterpene glycosides from Erylus formosus concurrently deterred predators, inhibited microbial attachment and prevented fouling by invertebrates and algae, whereas triterpene glycosides from Ectyoplasia ferox had both antipredatory and allelopathic activities (Kubanek et al. 2002).

For terrestrial plants it has been reported that concentrations of chemical defenses are significantly higher in species growing in tropical than in temperate forests (Levin and York 1978, Coley and Aide 1991). This has been interpreted as an evolutionary response to greater herbivory in the tropics (Coley and Aide 1991). Similar observations have been made for marine algae: tropical algae yield higher numbers and more deterrent secondary metabolites (Faulkner 1984, Hay and Fenical 1988, Hay 1996, Bolser and Hay 1996) and, thus, seem to be better defended than temperate species. Again, this has been attributed to the higher number of herbivorous fish on tropical compared to temperate reefs (Bolser and Hay 1996, Meekan and Choat 1997). Since sessile invertebrates in tropical coral reefs do, indeed, suffer greater predation pressure than in any other marine environment (Vermeij 1978, Carpenter 1997), it seemed not surprising that Bakus and Green (1974) found an inverse relationship between latitude and ichthyotoxicity in sponges. However, several subsequent studies did not find support for this latitudinal gradient theory (e.g., McCaffrey and Endean 1985, McClintock 1987, van de Vyver et al. 1990). This motivated Becerro et al. (2003) to test this theory by directly comparing chemical defenses from tropical and temperate sponges (collected from Guam and the Mediterranean Spanish coast respectively). Contrary to their predictions, they found the chemical defenses of tropical and temperate sponges to be equally effective against both sympatric (i.e., co-occurring with the sponges) and allopatric (i.e., not sharing habitat with the prey sponges) predatory fish. However, the authors point out that their results may be due to a response of the sponges and their predators to specific traits of the areas they investigated. They advise to be cautious with generalizing their results until they are confirmed by studies in other geographic areas (Becerro et al. 2003).

Interactions of chemical and structural features in sponge defense

It has been well documented that structural features in plants can also act as a defense against predators (e.g., McNaughton et al. 1985, Pennings and Paul 1992). In sponges inorganic spicules can amount for up to 75% of the total dry mass (Rtzler and Macintyre 1978) and are often arranged with their sharp end towards or protruding the sponge surface (Uriz et al. 2003). Thus, it was hypothesized already early on that these skeletal components of sponges provide antipredator defense, too (Randall and Hartman 1968, Sar and Vacelet 1973). The most likely mechanism of action for sponge spicules is abrasion or injury of feeding structures (e.g., mouth parts, lining of the digestive system), as has been observed in the gut of the hawksbill turtle (Meylan 1988). Nevertheless, several studies on the antipredatory properties of sponge spicules found results contrary to this assumption and concluded that sponges may have evolved spicules solely for structural purposes (Chanas and Pawlik 1995, 1996, Waddell and Pawlik 2000). However, an additive or even synergistic feeding deterrent effect between sponge spicules and secondary metabolites is conceivable if it is assumed that spicules act as an abrasive while passing through the gut of a potential predator. This way, they may facilitate or enhance the action of defense compounds (Hill et al. 2005).

630

Similar synergistic effects have been observed in plants (e.g., Pennings 1996). In recent studies, Hill et al. (2005) as well as Jones et al. (2005) analyzed interactions between sponge spicules and secondary metabolites in the context of predator deterrence. Both studies found examples of synergistic or additive effects in 1 of 4 and 4 of 8 tested sponge species, respectively. However, both studies came to the conclusion that synergism between structural and chemical defense cannot be considered the general rule in sponges. Moreover, differential results of various studies dealing with the effectiveness of structural defenses in sponges suggest that the results cannot be extended to all predators and different predator species can be affected differently by sponge spicules in their diets (Paul and Puglisi 2004).

Inducible defense and immune reactions

Inducible defenses were defined by Harvell (1990) as responses activated through a previous encounter with a consumer or competitor that confer some degree of resistance to subsequent attacks. Inducible defenses are most common if levels of disturbing impacts are unpredictable and display a high spatial or temporal variability (Harvell 1990, Zangerl and Rutledge 1996, Toth and Pavia 2007). Under these circumstances they can be more economical and effective against herbivores than constitutive defenses (Karban et al. 1997, Heil 2002, Toth and Pavia 2007). In contrast to plants, sponge chemical defense has largely been considered static. Only in recent years a small number of studies has looked into facultative, inducible defense mechanisms in this phylum. Thacker et al. (1998) investigated changes in the chemical profiles of the Indopacific sponges Dysidea sp. and Cacospongia sp. in the process of the former overgrowing the latter. They found no changes in the chemistry of Dysidea sp., but observed an increase in quantity of organic extract in portions of Cacospongia sp. that were covered by agar strips containing Dysidea crude extract, suggesting an induced defense against overgrowth. Richelle-Maurer et al. (2003) detected a sharp increase in the concentrations of the alkaloids oroidin and sceptrin in the Caribbean sponge Agelas conifera after experimental simulation of predator bites. Both compounds deterred feeding when tested at near natural concentrations against the predatory reef fish Stegastes partitus (Richelle-Maurer et al. 2003). A mixture of the compounds also proved to be active against the coral Madracis mirabilis, a potential competitor for space. Addition of the two compounds to ambient seawater at 0.0125% of the natural sponge concentration resulted in closure and retraction of the coral polyps. However, forced confrontation of A. conifera with the corals did not yield measurable changes in oroidin and sceptrin concentrations in the sponge tissue (Richelle-Maurer et al. 2003). An explanation for the low number of studies on inducible defenses in sponges may be that changes in chemical profiles most likely are a function of numerous biotic and abiotic factors influencing secondary metabolite biosynthesis (Thompson et al. 1987, Becerro et al. 1995, Turon et al. 1996). Moreover, unfavorable influences from the environment can do both increase defense compound metabolism (as a defensive response to the influencing factor) or decrease investment in

the secondary metabolite production, if energy is preferentially invested in cell repair (Agell et al. 2001, Walters and Pawlik 2005). As induced reactions in the chemical profile often are observable only after days or even weeks following the inducing event (Taylor et al. 2002, Richelle-Maurer 2003), this severely complicates interpretations on interrelations between observed secondary metabolite changes and assumed inducing factors. Mller and coworkers approached this problem by investigating adaptive antibacterial responses in sponges at the genetic level (Mller and Mller 2003). They found various immune reactions, primarily in Suberites domuncula, and described the signal transduction pathways as well as the defensive agents involved. The sponge responded to treatment with the bacterial endotoxin lipopolysaccharide (LPS) (Mller et al. 2004) with increased biosynthesis of two alkyl-lipid derivatives, 1-O-hexadecyl-sn-glycero-3-phosphocholine and 1-O-octadecyl-sn-glycero-3-phosphocholine. Both compounds showed pronounced activity in an antibacterial assay. In order to prove that the compounds were indeed produced by S. domuncula, a key enzyme of their biosynthetic pathway was cloned from the sponge. In a subsequent study, Wiens et al. (2005) discovered a receptor for LPS at the surface of cells from S. domuncula. They identified a signal transduction pathway that is induced upon elevated LPS levels and resulted in the enhanced expression of a perforin-like protein primarily at the sponge surface. The protein eliminates Gram-negative bacteria, whereas it is inactive against Gram-positive species. Based on these findings the authors concluded that the sponge S. domuncula possesses an innate immune system against Gramnegative bacteria (Wiens et al. 2005). Thakur et al. (2005) were able to show that S. domuncula also exhibits immune reactions against Gram-positive bacteria. The sponge reacts to exposure to peptidoglycan the characteristic cell wall component of Gram-positive bacteria with activation of endocytosis and release of lysozyme. Activation of endocytosis was determined by differential expression of an adaptor gene (AdaPTin-1) isolated from the sponge that encodes for a putative protein involved in endosome formation (Thakur et al. 2005). The release of lysozyme results in digestion and, thus, in elimination of the bacteria. Immunofluorescence studies with antibodies raised against lysozyme revealed that this immune reaction is targeted exclusively against extracellular bacteria in the sponge mesohyl and not against potentially symbiotic bacteria located in sponge bacteriocytes (Thakur et al. 2005). These examples clearly demonstrate that sponges, indeed, have inducible defenses and immune reactions. The application of biomolecular techniques to analyze the responses of sponges toward predator or pathogen attacks on the gene expression level may help to unravel mechanisms that otherwise are concealed by the complexity of factors influencing the sponge secondary chemistry.

Activated defense
Rapid wound-induced conversions of stored precursors to potent defensive compounds have been referred to with various terms, including short-term inducible defense

631

(STID) (Haukioja 1980, Clausen et al. 1989), dynamic defense (Reichardt et al. 1990) and induced direct defense (van Hulten et al. 2006). Paul and van Alstyne (1992), when reporting the first defense of this type in the marine environment, termed this process activated defense to clearly distinguish it from the predator-induced biosynthesis of defensive metabolites (see inducible defense). By converting inactive or less active precursors to defense metabolites with pronounced activity only upon tissue damage and locally restricted to the wounded tissue area, the risk of autotoxicity caused by the defensive conversion products can be alleviated (Saunders et al. 1977, Frehner and Conn 1987, Poulton 1988). Typically, the rapid activated defense reactions are catalyzed by enzymes that upon disruption of tissue compartments get into contact with the precursors and facilitate the conversion reactions (Matile 1984, Wittstock and Gershenzon 2002). Activated chemical defenses are widespread in terrestrial vascular plants. The most prominent example is the conversion of cyanogenic glycosides to HCN (e.g., Jones 1988, Seigler 1991, Wajant and Effenberger 1996, Gleadow and Woodrow 2002, and references cited therein). Numerous analogous defense mechanisms in the terrestrial environment involve other molecules such as glucosinolates, phenolic glycosides, and sesquiterpenes (Sterner et al. 1985, Clausen et al. 1989, Stoewsand 1995, Fahey et al. 2001). In aquatic habitats, activated defenses so far have predominantly been found in plants [see reviews by Paul and Puglisi (2004) and Pohnert (2004)]. Reported examples include numerous macroalgal species (e.g., Paul and van Alstyne 1992, Cetrulo and Hay 2000, Jung et al. 2002, van Alstyne and Houser 2003) as well as planktonic diatoms and dinoflagellates (Pohnert 2005, Strom et al. 2003). In contrast, there are very few reports on analogous defense mechanisms in sessile marine invertebrates. It is yet unresolved whether this is due to a limited distribution of this strategy among invertebrates or rather reflects the fact that the majority of ecological studies so far have focused on constitutive defenses. When discussing the possibility of activated defenses in sponges, it is interesting to note that secondary metabolites in sponges are often stored in specialized cells (e.g., spherulous cells, choanocytes), which may provide the necessary compartments to separate precursors from converting enzymes a prerequisite for activated defense reactions (e.g., Thompson et al. 1983, Turon et al. 2000, Richelle-Maurer et al. 2003). To our knowledge, to date only two examples of activated defenses in sessile marine invertebrates have been reported one occurring in the sponge genus Aplysina (Teeyapant and Proksch 1993, Weiss et al. 1996, Ebel et al. 1997, Thoms et al. 2004, 2006b), the other in the marine hydroid species Tridentata marginata (Lindquist 2002). Ettinger-Epstein et al. (2007) recently observed a deacetylation of acetylated sesterterpenes in Luffariella variabilis to the corresponding alcohols when they thawed frozen tissue of the sponge. Since the compounds were stable when isolated from the tissue, the authors proposed that the conversion to the alcohols may be enzyme-mediated. Further, they speculated about a role of this reaction as an activated defense, but have not tested this hypothesis yet (Ettinger-Epstein et al. 2007). Recently, we

found another example of an activated defense in the sponge Aplysinella rhax, which we presented at the 7th International Sponge Symposium (Thoms and Schupp 2006). Despite these reported examples, the existence of activated defenses in sponges has been a controversially discussed topic (see Puyana et al. 2003 and Thoms et al. 2006b). Here, we highlight the various aspects that need to be considered when examining chemical profiles of sponges for changes that may be interpreted as activated defenses. Moreover, we point out various methodological constraints (some being more and some less sponge-specific) that complicate interpretations in this context and, therefore, resulted in this controversy.

Sample handling and preservation

Due to the circumstances involved in marine sampling (e.g., wave action and currents during sampling, transport and storage during dive and on the boat, etc.) sponge samples are at high risk of unintentional damage before they are finally preserved. Thus, to gain insight into the chemistry of an intact sponge, it is necessary to minimize damage as well as transportation times (e.g., by using sturdy sample containers instead of plastic bags; by on-board preservation of the samples, etc.). Further, the method of preservation can considerably impact the intactness of the analyzed chemical profile. Freezing wet sponge tissue results in the formation of intracellular ice crystals that may cause decompartmentalization by disrupting cellular membranes (Hllgren and quist 1990). Upon thawing, enzymes may be reactivated and catalyze conversion reactions (Gahan 1981, Ettinger-Eppstein et al. 2007). Interestingly, a similar effect can be caused by extraction or preservation of wet sponge tissue with organic solvents (Teeyapant and Proksch 1993, Thoms and Schupp unpublished). This, at first, may seem surprising, since enzymes usually are considered sensitive to contact with organic solvents. However, sustained catalytic activity of enzymes in aqueous solvents, as it may occur when wet sponge tissue gets gradually soaked during the extraction or preservation procedure, is a known phenomenon (see Klibanov 2001 and references cited therein). If the disintegrating effect of organic solvents on biomembranes (Jones 1989, Weber and deBont 1996) causes decompartmentalization within the sponge tissue, contact between the active enzymes and the precursors may be facilitated and the conversion reactions can take place. To ensure enzyme inactivation in the sponge tissue, samples should, therefore, be processed by flashfreezing and subsequent lyophilization. The existence of activated chemical defenses in sponges is a rather recent concept and, thus, earlier studies did not necessarily have possible enzymatic reactions in sponge tissue in mind. This may explain why several compounds that originally were considered constitutive in sponges (Fattorusso et al. 1970, Kernan et al. 1987, Shin et al. 2000) later revealed to be conversion products (Thoms et al. 2006b, EttingerEpstein et al. 2007, Thoms and Schupp unpublished).

Natural variability of the sponge chemistry

Many sponge species display pronounced variability in their chemical profiles. Not only do individuals of the same species

632

show considerable quantitative and qualitative differences in their secondary metabolite chemistry, but even within single individuals vast divergences are observed (Schupp et al. 1999, Furrow et al. 2003, Thoms et al. 2006b). This variability can do both conceal activated defense reactions as well as falsely hint to them. To avoid misinterpretations, each study on activated defenses should be preceded by a survey on the chemistry of intact individuals under various natural conditions. Interestingly, all activated defenses in sessile marine invertebrates discovered so far involve components that are easily detectable by HPLC-UV and are present in the organisms tissues in extraordinarily high concentrations (Lindquist 2002, Thoms et al. 2006b, Thoms and Schupp unpublished). However, secondary metabolites can possess pronounced activity and mediate ecological interactions even at minute concentrations (Paul and Puglisi 2004, Paul et al. 2006). Moreover, due to specific chemical characteristics (e.g., lack of chromophores) compounds involved in defense reactions may not be readily detectable by standard chemoanalytical techniques. Therefore, changes in the chemical profiles may not always become apparent, even if they have major ecological effects. Pronounced natural variability also entails difficulties for data analysis of wounding experiments. Substantial fluctuations of compound concentrations may impede validation of observed wound-activated changes. It may be reasonable to analyze shifts in relative compound proportions rather than measuring changes in their absolute concentrations if the relative pattern of the sponges chemistry turns out to be more uniform. Further, if there is more than one assumed precursor and/or product, pooling their respective concentrations can help to identify wound-activated changes. High intensities of wounding even if ecologically irrelevant can help to initially observe wound-activated chemical reactions. By gradually decreasing wounding intensity in a series of samples and analyzing the resulting chemical profiles, a causal link between wounding and the reactions in the chemical profiles can be investigated.

conversion event over time. Thus, only start and end points of the conversion reactions can be appropriately analyzed in field experiments. Moreover, it is difficult to determine the effect of defined wounding intensities in field experiments, since unintentional damage in the course of sample handling are likely to occur. Due to these difficulties, laboratory experiments conducted in seawater tanks with carefully handled, entire sponge individuals may be a reasonable alternative that allows for more controllable conditions. By ensuring the healthy condition of the sponges (e.g., by using individuals without any signs of damage and with open oscules indicating metabolic activity) and by comparing their chemical profiles to those of intact sponges in the field, bias caused by the laboratory conditions can be minimized.

Determining ecological relevance and target organisms

Determining wound-activated reactions in chemical profiles of sponges is a matter of careful investigation and use of appropriate analytical techniques (allowing for compound identification, description of reaction kinetics, and determination of enzymatic catalysis). Considerably more ambiguous are interpretations on the ecological relevance of such reactions. To provide evidence for an activated defense, the conversion of less active precursors into defensive agents with higher activity has to be shown (Paul and van Alstyne 1992). Thus, precursors and products have to be compared in bioassays in their respective naturally occurring concentrations. Since the product concentrations are a function of the wounding caused to the sponge tissue, an ecologically relevant manner of wounding needs to be applied. Often tissue grinding has been employed to elicit wound-activated reactions and the compound concentrations thereof have been used in bioassays to assess the defensive function of the conversion products (e.g., Paul and van Alstyne 1992, Cetrulo and Hay 2000, Jung and Pohnert 2001). Yet, if large tissue pieces are bitten off from the sponge and immediately swallowed by the predator, measurements in ground tissue are likely to overestimate the naturally formed concentrations. Puyana et al. (2003) chose stabbing of sponge tissue with a scalpel as an alternative to grinding. This likely resulted in the disruption of tissue compartments at the surface of the scalpel cuts, but left the tissue underneath unaffected. Subsequently, the extracts from entire sponge pieces bearing the scalpel cuts were analyzed (Puyana et al. 2003). However, if the ratio of damaged to undamaged tissue in the sampled sponge pieces is low, concentrations of conversion products might become too low for detection. In our recent study on an activated defense in the sponge Aplysinella rhax we picked tissue pieces with forceps in order to mimic predator bites and elicit conversion reactions (Thoms and Schupp unpublished). Compound concentrations were analyzed in the picked tissue pieces. At best, these approaches will imitate feeding behavior of one predator type, only. The actual predator may bite off larger or smaller pieces, cause less or more tissue squeezing, or may abrade the surface layers instead (Toth and Pavia 2007). Thus, a reliable comparison of the bioactivity before and after

Field experiments versus laboratory experiments

Generally, field experiments are clearly to be favored over laboratory experiments when investigating ecological phenomena. However, investigations on activated defenses in sponges in the field entail several experimental constraints. If wounding is caused to sponges in situ, i.e., in their natural habitat, usually a time-consuming sampling procedure has to follow (underwater bagging, transportation to the surface, etc.) before the samples can be preserved and enzymatic reactions can be stopped. Underwater handling of the samples entails the risk of elution of the conversion products from the sponge tissue, especially if hydrophilic compounds are formed. This can be minimized if after wounding samples are immediately sealed underwater in small containers and if the ambient seawater in these containers is analyzed as well. Since wound-activated reactions often occur within seconds (Paul and van Alstyne 1992, Pohnert 2000, Jung and Pohnert 2001, Thoms et al. 2006b, Thoms and Schupp unpublished), a prolonged sampling procedure precludes monitoring the

633

wound-activated conversion is only possible if the actual predator is known and its feeding behavior can appropriately be mimicked. But even if this is feasible, effects on compound concentrations occurring after wounding, such as dilution by seawater or leakage from the sponge tissue, may impede their proper assessment. Water solubility of the defensive metabolites also poses a challenge to the design of bioassays testing their anti-predator effect. Compound loss from the experimental food needs to be minimized in order to keep the assay conditions constant over the experiment course. However, under natural conditions the defensive agents may be exuded from the tissue directly into the predators mouth when the tissue gets squeezed and cells disrupt (Thoms and Schupp unpublished) an effect that can hardly be imitated with food designed to retain the compounds efficiently. Further, it has to be taken into account that predators may learn to link sensing the precursors with the formation of repellent conversion products (see Chivers and Smith (1998), Rochette et al. (1998), and Larson and McCormick (2005) on learning and recognition of chemical cues in potential sponge predators). Thus, a comparative bioassay may not necessarily reveal any difference between the compounds if the predator stops feeding already upon contact with the non-repellent precursors. The above considerations are based on the assumption that the target organism of an activated defense is a predator and, by this, also represents the elicitor of the defense reactions. Providing evidence for the ecological relevance of an observed wound-activated chemical reaction gets even more intricate if the eliciting organisms and the target organisms are not identical. This, for example, is the case in sponges of the genus Aplysina (Thoms et al. 2004, 2006b). Here, the conversion precursors possess a pronounced repellent effect against potential fish predators (Thoms et al. 2004). If despite this chemical protection the sponge gets wounded, these compounds are enzymatically converted into agents with considerably enhanced antimicrobial properties, presumably providing a barrier against microbial pathogens and protecting the wounded sponge tissue against infection (Thoms et al. 2006b). To reveal such functions of activated defense reactions, the search for potential targets has to be based broadly, including both macro- and microorganisms. Moreover, it has to be taken into account that defensive compounds may be active at various scales (e.g., in quantities high enough to overcome dilution effects en route to a predators olfactory organs, or at only locally arising concentrations that form a barrier against microbes). Taken together, various methodological constraints as well as inherent limitations on the interpretability of the results considerably complicate the accumulation of evidences for activated defense mechanisms in sponges. This, on the one hand, may explain the low number of reports in this context with respect to sponges, but also with respect to sessile marine invertebrates in general. On the other hand, this is an exciting challenge for future studies aiming to shed light on the question whether in this group of animals activated defenses are, indeed, isolated phenomena, or may represent a common but as yet largely overlooked strategy.

Conclusions
While considerable work has been done on sponge chemical ecology over the last decade and we are seeing some trends emerging from the multitude of studies, it has to be acknowledged that often ecological concepts can not be generalized. For instance, there are numerous examples that support the optimal defense theory and the growthdifferentiation balance hypothesis but almost as many contradicting them. Many tropical sponge species have a stronger protection against predators than their temperate relatives, however, a general proof for the latitudinal gradient theory failed as a whole, the defensive chemistry of tropical sponges is apparently not more repellent than that of temperate species. Sponges do make use of synergisms between structural and chemical defenses but not all sponge species do and this strategy is not equally effective against all types of predators. It is obvious that single concepts are unlikely to be valid for all the numerous sponge species in the multitude of habitats they live in. However, from the studies reviewed it becomes apparent that in many cases evidences for these concepts may be obscured by methodological constraints as well as by the complexity of parameters affecting sponges and their secondary metabolisms. Secondary metabolite profiles of sponges often are characterized by pronounced variability. In fact, sponges have been described as dynamic multicellular systems that undergo constant changes in adaptation to altering external factors (Gaino and Magnino 1999). This versatility undoubtedly complicates seeing clear patterns in sponge traits. Moreover, there is evidence that microbial symbionts often contribute substantially to both nutrition and secondary metabolite biosynthesis of sponges (Taylor et al. 2007). Still very little is known about these interactions in most sponge species, which severely complicates answering such basic questions as to whether chemical defense is costly for a sponge. Further complexity is added by sponges employing chemical defenses simultaneously against various threats on various sizes of scale (e.g., against predators, competitors, biofouling, pathogens), and using metabolites in multiple ways, being concurrently active against several of these threats (multi-purpose tools). The resulting interferences may obscure links between single effects and single causes and, this way, complicate discerning clear defense concepts. This is similar for facultative defenses. While parallels between sponge and plant ecology make it rational to search for such defenses in sponges (i.e., for activated and inducible defenses), so far only few examples have been identified. Here as well, experimental constraints and interfering parameters complicate investigations on effects and causes. Hence, the question whether facultative defenses in sponges are isolated phenomena or common but as yet largely overlooked strategies remains to be resolved. In many cases it will be inevitable to evaluate the ecological concepts and defense strategies at the species or even at the individual level to be able to contemplate all the factors that potentially impact their outcome. To break down the complexity of parameters, investigation of certain processes in artificial systems may be necessary. Biomolecular approaches similar to those currently employed to elucidate

634

innate immune reactions in sponges on the genetic level may, in the future, prove useful also to study other aspects of sponge chemical ecology. However, to draw legitimate conclusions on ecological interrelations, findings from in vitro experiments should always be validated in natural systems.

Acknowledgements
We thank Claudia Kohlert-Schupp and several colleagues for interesting discussions and three anonymous reviewers for their comments that helped to improve this review. Carsten Thoms gratefully acknowledges a Feodor Lynen Fellowship by the Alexander von Humboldt-Foundation, Bonn, Germany. Peter Schupp acknowledges funding by NIH MBRS SCORE S06-GM44796 and NIH SCORE S06-GM044796-16A1. This is contribution number 608 of the University of Guam Marine Laboratory.

References
Agell G, Uriz MJ, Cebrian E, Marti R (2001) Does stress protein induction by copper modify natural toxicity in sponges? Environ Toxicol Chem 20: 2588-2593 Arillo A, Bavestrello G, Burlando B, Sar M (1993) Metabolic integration between symbiotic cyanobacteria and sponges: a possible mechanism. Mar Biol 117: 159-162 Avila C, Paul VJ (1997) Chemical ecology of the nudibranch Glossodoris pallida: Is the location of diet-derived metabolites important for defense? Mar Ecol Prog Ser 150: 171-180 Ayling AL (1983) Growth and regeneration rates in thinly encrusting Demospongiae from temperate waters. Biol Bull 165: 343-352 Bakus GJ (1966) Some relationships of fishes to benthic organisms on coral reefs. Nature 210: 280-284 Bakus GJ, Green G (1974) Toxicity in sponges and holothurians: a geographic pattern. Science 185: 951-953 Baldwin IT, Ohnmeiss TE (1994) Coordination of photosynthesis and alkaloidal responses to damage in uninducible and inducible Nicotiana sylvestris. Ecology 75: 1003-1014 Barto EK, Cipollini D (2005) Testing the optimal defense theory and the growth-differentiation balance hypothesis in Arabidopsis thaliana. Oecologia 146: 169-178 Becerro MA, Paul VJ, Starmer J (1998) Intracolonial variation in chemical defenses of the sponge Cacospongia sp. and its consequences on generalist fish predators and the specialist nudibranch predator Glossodoris pallida. Mar Ecol Prog Ser 168: 187-196 Becerro MA, Thacker RW, Turon X, Uriz MJ, Paul VJ (2003) Biogeography of sponge chemical ecology: comparisons of tropical and temperate defenses. Oecologia 135: 91-101 Becerro MA, Turon X, Uriz MJ (1995) Natural variation of toxicity in encrusting sponge Crambe crambe (Schmidt) in relation to size and environment. J Chem Ecol 21: 1931-1946 Becerro MA, Turon X, Uriz MJ (1997) Multiple functions for secondary metabolites in encrusting marine invertebrates. J Chem Ecol 23: 1527-1547 Bergmann W, Feeney RJ (1950) The isolation of a new thymine pentoside from sponges. J Am Chem Soc 72: 2809-2810 Bergquist PR (1978) Sponges. University of California Press, Berkeley Blunt J, Munro M (2003) MarinLit. A database of the literature on marine natural products prepared and maintained by the

Marine Chemistry Group. Department of Chemistry, University of Canterbury, Canterbury Bobzin SC, Faulkner DJ (1992) Chemistry and chemical ecology of the Bahamian sponge Aplysilla glacialis. J Chem Ecol 18: 309332 Bolser RC, Hay ME (1996) Are tropical plants better defended? Palatability and defenses of temperate vs. tropical seaweeds. Ecology 77: 2269-2286 Burns E, Ifrach I, Carmeli S, Pawlik JR, Ilan M (2003) Comparison of anti-predatory defenses of Red Sea and Caribbean sponges. I. Chemical defense. Mar Ecol Prog Ser 252: 105-114 Carpenter RC (1997) Invertebrate predators and grazers. In: Birkeland C (ed). Life and death of coral reefs. Chapman and Hall, New York. pp. 198-229 Cetrulo GL, Hay ME (2000) Activated chemical defenses in tropical versus temperate seaweeds. Mar Ecol Prog Ser 207: 243-253 Chanas B, Pawlik JR (1995) Defenses of Caribbean sponges against predatory reef fish. 2. Spicules, tissue toughness, and nutritional quality. Mar Ecol Prog Ser 127: 195-211 Chanas B, Pawlik JR (1996) Does the skeleton of a sponge provide a defense against predatory reef fish? Oecologia 107: 225-231 Chivers DP, Smith RJF (1998) Chemical alarm signalling in aquatic predator-prey systems: a review and prospectus. Ecoscience 5: 338-352 Clausen TP, Reichardt PB, Bryant JP, Werner RA, Post K, Frisby K (1989) Chemical model for short-term induction in quaking aspen (Populus tremuloides) foliage against herbivores. J Chem Ecol 15: 2335-2346 Coley PD, Aide TM (1991) Comparison of herbivory and plant defenses in temperate and tropical broad-leaved forests. In: Price PW, Lewinsohn TM, Fernandes GW, Benson WW (eds). Plant-animal interactions: evolutionary ecology in tropical and temperate regions. Wiley, New York. pp. 25-49 Cronin G, Hay ME (1996) Within plant variation in seaweed palatability and chemical defenses: optimal defense theory versus the growth differentiation balance hypothesis. Oecologia 105: 361-368 Davis AR, Fyfe SK, Turon X, Uriz MJ (2003) Size matters sometimes: wall height and the structure of subtidal benthic invertebrate assemblages in south-eastern Australia and Mediterranean Spain. J Biogeogr 30: 1797-1807 Dube D, Kim K, Alker AP, Harvell CD (2002) Size structure and geographic variation in chemical resistance of sea fan corals Gorgonia ventalina to a fungal pathogen. Mar Ecol Prog Ser 231: 139-150 Ebel R, Brenzinger M, Kunze A, Gross HJ, Proksch P (1997) Wound activation of protoxins in marine sponge Aplysina aerophoba. J Chem Ecol 23: 1451-1462 Eder C, Schupp P, Proksch P, Wray V, Steube K, Muller CE, Frobenius W, Herderich M, van Soest RWM (1998) Bioactive pyridoacridine alkaloids from the Micronesian sponge Oceanapia sp. J Nat Prod 61: 301-305 Ettinger-Epstein P, Motti CA, de Nys R, Wright AD, Battershill CN, Tapiolas DM (2007) Acetylated sesterterpenes from the Great Barrier Reef sponge Luffariella variabilis. J Nat Prod 70: 648651 Fahey JW, Zalcmann AT, Talalay P (2001) The chemical diversity and distribution of glucosinolates and isothiocyanates among plants. Phytochemistry 56: 5-51

635

Fattorusso E, Minale L, Sodano G (1970) Aeroplysinin-1, a new bromo-compound from Aplysina aerophoba. J Chem Soc D-Chem Commun 12: 751-753 Faulkner DJ (1984) Marine natural products: metabolites of marine algae and herbivorous marine mollusks. Nat Prod Rep 1: 251-280 Frehner M, Conn EE (1987) The linamarin beta-glucosidase in Costa Rica wild bean (Phaseolus lunatus L.) is apoplastic. Plant Physiol 84: 1296-1300 Furrow FB, Amsler CD, McClintock JB, Baker BJ (2003) Surface sequestration of chemical feeding deterrents in the Antarctic sponge Latrunculia apicalis as an optimal defense against sea star spongivory. Mar Biol 143: 443-449 Gahan PB (1981) Cell senescence and death in plants. In: Bowen ID, Lockshin RA (eds). Cell death in biology and pathology. Chapman & Hall, London. pp. 145-169 Gaino E, Magnino G (1999) Dissociated cells of the calcareous sponge Clathrina: a model for investigating cell adhesion and cell motility in vitro. Microsc Res Tech 44: 279-292 Gleadow RM, Woodrow IE (2002) Constraints on effectiveness of cyanogenic glycosides in herbivore defense. J Chem Ecol 28: 1301-1313 Hllgren J, quist G (1990) Adaptations to low temperatures. In: Alscher RG, Cumming JR (eds). Stress responses in plants: adaptation and acclimation mechanisms. Wiley-Liss, New York. pp. 265-293 Harvell CD (1990) The ecology and evolution of inducible defenses. Q Rev Biol 65: 323-340 Haukioja E (1980) On the role of plant defenses in the fluctuation of herbivore populations. Oikos 35: 202-213 Hay ME (1996) Marine chemical ecology: whats known and whats next? J Exp Mar Biol Ecol 200: 103-134 Hay ME, Fenical W (1988) Marine plant - herbivore interactions: the ecology of chemical defense. Annu Rev Ecol Syst 19: 111-145 Heil M (2002) Ecological costs of induced resistance. Curr Opin Plant Biol 5: 345-350 Hentschel U, Usher KM, Taylor MW (2006) Marine sponges as microbial fermenters. FEMS Microbiol Ecol 55: 167-177 Hill MS (1998) Spongivory on Caribbean reefs releases corals from competition with sponges. Oecologia 117: 143-150 Hill MS, Lopez NA, Young KA (2005) Anti-predator defenses in western North Atlantic sponges with evidence of enhanced defense through interactions between spicules and chemicals. Mar Ecol Prog Ser 291: 93-102 Jones AC, Blum JE, Pawlik JR (2005) Testing for defensive synergy in Caribbean sponges: Bad taste or glass spicules? J Exp Mar Biol Ecol 322: 67-81 Jones CG, Firn RD (1991) On the evolution of plant secondary chemical diversity. Phil Trans R Soc Lond B Biol Sci 333: 273280 Jones DA (1988) Cyanogenesis in animal-plant interactions. In: Evered D, Harnett S (eds). Cyanide compounds in biology. Wiley, Chichester. pp. 151-170 Jones RP (1989) Biological principles of the effects of ethanol: a review. Enzyme Microb Technol 11: 130-152 Jung V, Pohnert G (2001) Rapid wound-activated transformation of the green algal defensive metabolite caulerpenyne. Tetrahedron 57: 7169-7172 Jung V, Thibaut T, Meinesz A, Pohnert G (2002) Comparison of the wound-activated transformation of caulerpenyne by invasive and

noninvasive Caulerpa species of the Mediterranean. J Chem Ecol 28: 2091-2105 Karban R, Agrawal AA, Mangel M (1997) The benefits of induced defenses against herbivores. Ecology 78: 1351-1355 Kelly SR, Jensen PR, Henkel TP, Fenical W, Pawlik JR (2003) Effects of Caribbean sponge extracts on bacterial attachment. Aquat Microb Ecol 31: 175-182 Kernan MR, Faulkner DJ, Jacobs RS (1987) The luffariellins, novel antiinflammatory sesterterpenes of chemotaxonomic importance from the marine sponge Luffariella variabilis. J Org Chem 52: 3081-3083 Klibanov AM (2001) Improving enzymes by using them in organic solvents. Nature 409: 241-246 Kubanek J, Whalen KE, Engel S, Kelly SR, Henkel TP, Fenical W, Pawlik JR (2002) Multiple defensive roles for triterpene glycosides from two Caribbean sponges. Oecologia 131: 125-136 Larson JK, McCormick MI (2005) The role of chemical alarm signals in facilitating learned recognition of novel chemical cues in a coral reef fish. Anim Behav 69: 51-57 Levin DA, York BM (1978) The toxicity of plant alkaloids: an ecographic perspective. Biochem Syst Ecol 6: 61-76 Lindquist N (2002) Tridentatols D-H, nematocyst metabolites and precursors of the activated chemical defense in the marine hydroid Tridentata marginata (Kirchenpauer 1864). J Nat Prod 65: 681684 Mahon AR, Amsler CD, McClintock JB, Amsler MO, Baker BJ (2003) Tissue-specific palatability and chemical defenses against macropredators and pathogens in the common articulate brachiopod Liothyrella uva from the Antarctic Peninsula. J Exp Mar Biol Ecol 290: 197-210 Matile P (1984) The toxic compartment of plant cells. Naturwissenschaften 71: 18-24 McCaffrey EJ, Endean R (1985) Antimicrobial activity of tropical and subtropical sponges. Mar Biol 89: 1-8 McClintock JB (1987) Investigation of the relationship between invertebrate predation and biochemical composition, energy content, spicule armament and toxicity of benthic sponges at McMurdo Sound, Antarctica. Mar Biol 94: 479-487 McClintock JB, Amsler CD, Baker BJ, van Soest RWM (2005) Ecology of Antarctic marine sponges: an overview. Integr Comp Biol 45: 359-368 McClintock JB, Baker BJ (1998) Chemical ecology in Antarctic Seas: chemical interactions can lead to unusual arrangements between species. Am Sci 86: 254-264 McClintock JB, Baker BJ (2001) Marine chemical ecology. CRC, Boca Raton McNaughton SJ, Tarrants JL, McNaughton MM (1985) Silica as a defense against herbivory and a growth promotor in African grasses. Ecology 66: 528-535 Meekan MG, Choat JH (1997) Latitudinal variation in abundance of herbivorous fishes: a comparison of temperate and tropical reefs. Mar Biol 128: 373-383 Meylan A (1988) Spongivory in hawksbill turtles: a diet of glass. Science 239: 393-395 Mller WEG, Klemt M, Thakur NL, Schrder HC, Aiello A, DEsposito M, Menna M, Fattorusso E (2004) Molecular/chemical ecology in sponges: evidence for an adaptive antibacterial response in Suberites domuncula. Mar Biol 144: 19-29

636

Mller WEG, Mller IM (2003) Origin of the metazoan immune system: identification of the molecules and their functions in sponges. Integr Comp Biol 43: 281-292 Newbold RW, Jensen PR, Fenical W, Pawlik JR (1999) Antimicrobial activity of Caribbean sponge extracts. Aquat Microb Ecol 19: 279284 Paul VJ (1992) Chemical defense of benthic marine invertebrates. In: Paul VJ (ed). Ecological roles of marine natural products. Cornell University Press (Comstock), Ithaca. pp. 164-188 Paul VJ, van Alstyne KL (1992) Activation of chemical defenses in the tropical green algae Halimeda spp. J Exp Mar Biol Ecol 160: 191-203 Paul VJ, Puglisi MP (2004) Chemical mediation of interactions among marine organisms. Nat Prod Rep 21: 189-209 Paul VJ, Puglisi MP, Ritson-Williams R (2006) Marine chemical ecology. Nat Prod Rep 23: 153-180 Pavia H, Toth GB, Aberg P (2002) Optimal defense theory: elasticity analysis as a tool to predict intraplant variation in defenses. Ecology 83: 891-897 Pawlik JR (1993) Marine invertebrate chemical defenses. Chem Rev 93: 1911-1922 Pawlik JR (1998) Coral reef sponges: do predatory fishes affect their distribution? Limnol Oceanogr 43: 1396-1399 Pawlik JR, Chanas B, Toonen RJ, Fenical W (1995) Defenses of Caribbean sponges against predatory reef fish. I. Chemical deterrency. Mar Ecol Prog Ser 127: 183-194 Pennings SC (1996) Testing for synergisms between chemical and mineral defenses - a comment. Ecology 77: 1948-1950 Pennings SC, Paul VJ (1992) Effect of plant toughness, calcification, and chemistry on herbivory by Dolabella auricularia. Ecology 73: 1606-1619 Pohnert G (2000) Wound-activated chemical defense in unicellular planktonic algae. Angew Chem Int Ed 39: 4352-4254 Pohnert G (2004) Chemical defense strategies of marine organisms. In: Schulz S (ed). Topics in current chemistry, volume 239: the chemistry of pheromones and other semiochemicals I. Springer, Berlin, Heidelberg. pp. 179-219 Pohnert G (2005) Diatom/copepod interactions in plankton: The indirect chemical defense of unicellular algae. Chem Biochem 6: 946-959 Poulton JE (1988) Localization and catabolism of cyanogenic glycosides. In: Evered D, Harnett S (eds). Cyanide compounds in biology. John Wiley & Sons, Chichester. pp. 67-91 Puyana M, Fenical W, Pawlik JR (2003) Are there activated chemical defenses in sponges of the genus Aplysina from the Caribbean? Mar Ecol Prog Ser 246: 127-135 Randall JE, Hartman WD (1968) Sponge-feeding fishes of the West Indies. Mar Biol 1: 216-225 Reichardt PB, Bryant JP, Mattes BR, Clausen TP, Chapin FS, Meyer M (1990) Winter chemical defense of Alaskan balsam poplar against snowshoe hares. J Chem Ecol 16: 1941-1959 Reiswig HM (1973) Population dynamics of three Jamaican Demospongiae. Bull Mar Sci 23: 191-226 Rhoades DF, Cates RG (1976) Toward a general theory of plant antiherbivore chemistry. Recent Adv Phytochem 10: 168-213 Richelle-Maurer E, de Kluijver MJ, Feio S, Gaudencio S, Gaspar H, Gomez R, Tavares R, van de Vyver G, van Soest RWM (2003) Localization and ecological significance of oroidin and sceptrin in the Caribbean sponge Agelas conifera. Biochem Syst Ecol 31: 1073-1091

Rochette R, Arsenault DJ, Justome B, Himmelman JH (1998) Chemically-mediated predator-recognition learning in a marine gastropod. Ecoscience 5: 353-360 Rtzler K, Macintyre I (1978) Siliceous sponge spicules in coral reef sediments. Mar Biol 49: 147-159 Sar M, Vacelet J (1973) cologie des dmosponges. In: Grass P (ed). Spongiaires. Trait de zoologie, vol. 3. Masson, Paris. pp. 462-576 Saunders JA, Conn EE, Lin CH, Stocking CR (1977) Subcellular localization of the cyanogenic glycoside of Sorghum by autoradiography. Plant Physiol 59: 647-652 Schmitt TM, Hay ME, Lindquist N (1995) Constraints on chemically mediated coevolution multiple functions for seaweed secondary metabolites. Ecology 76: 107-123 Schupp P, Eder C, Paul V, Proksch P (1999) Distribution of secondary metabolites in the sponge Oceanapia sp. and its ecological implications. Mar Biol 135: 573-580 Seigler DS (1991) Cyanide and cyanogenic glycosides. In: Rosenthal GA, Berenbaum MR (eds). Herbivores: their interactions with secondary plant metabolites. Academic Press, San Diego. pp. 3578 Shin J, Lee HS, Seo Y, Rho JR, Cho KW, Paul VJ (2000) New bromotyrosine metabolites from the sponge Aplysinella rhax. Tetrahedron 56: 9071-9077 Stamp N (2004) Can the growth-differentiation balance hypothesis be tested rigorously? Oikos 107: 439-448 Sterner O, Bergman R, Kihlberg J, Wickberg B (1985) The sesquiterpenes of Lactarius vellereus and their role in a proposed chemical defense system. J Nat Prod 48: 279-288 Stoewsand GS (1995) Bioactive organosulfur phytochemicals in Brassica oleracea vegetables a review. Food Chem Toxicol 33: 537-543 Strom S, Wolfe G, Slajer A, Lambert S, Clough J (2003) Chemical defense in the microplankton II: Inhibition of protist feeding by beta-dimethylsulfoniopropionate (DMSP). Limnol Oceanogr 48: 230-237 Swearingen DC, Pawlik JR (1998) Variability in the chemical defense of the sponge Chondrilla nucula against predatory reef fishes. Mar Biol 131: 619-627 Taylor RB, Sotka E, Hay ME (2002) Tissue-specific induction of herbivore resistance: seaweed response to amphipod grazing. Oecologia 132: 68-76 Taylor MW, Radax R, Steger D, Wagner M (2007) Sponge-associated microorganisms: evolution, ecology, and biotechnological potential. Microbiol Mol Biol Rev 71: 295-347 Teeyapant R, Proksch P (1993) Biotransformation of brominated compounds in the marine sponge Verongia aerophoba. Evidence for an induced chemical defense? Naturwissenschaften 80: 369370 Thacker RW, Becerro MA, Lumbang WA, Paul VJ (1998) Allelopathic interactions between sponges on a tropical reef. Ecology 79: 1740-1750 Thakur NL, Perovic-Ottstadt S, Batel R, Korzhev M, Diehl-Seifert B, Mller IM, Mller WEG (2005) Innate immune defense of the sponge Suberites domuncula against gram-positive bacteria: induction of lysozyme and AdaPTin. Mar Biol 146: 271-282 Thompson JE, Barrow KD, Faulkner DJ (1983) Localization of two brominated metabolites, aerothionin and homoaerothionin, in spherulous cells of the marine sponge Aplysina fistularis. Acta Zool Stockh 64: 199-210

637

Thompson JE, Murphy PT, Bergquist PR, Evans EA (1987) Environmentally induced variation in diterpene composition of the marine sponge Rhopaloeides odorabile. Biochem Syst Ecol 15: 595-606 Thompson JE, Walker ND, Faulkner DJ (1985) Screening and bioassays for biologically active substances from forty marine sponge species from San Diego, California, USA. Mar Biol 88: 11-21 Thoms C, Ebel R, Proksch P (2006a) Sequestration and possible role of dietary alkaloids in the sponge-feeding mollusk Tylodina perversa. In: Cimino G, Gavagnin M (eds). Molluscs. Progress in molecular and subcellular biology, vol. 43. Springer, Berlin. pp. 261-275 Thoms C, Ebel R, Proksch P (2006b) Activated chemical defense in Aplysina sponges revisited. J Chem Ecol 32: 97-123 Thoms C, Schupp PJ (2006) Activated chemical defense in Aplysinella rhax. In: Custdio MR, Lbo-Hajdu G, Hajdu E, Muricy M (eds). 7th International Sponge Symposium - Book of Abstracts (Armao dos Bzios. Brazil). Museu Nacional, Srie Livros, vol. 16. pp. 143 Thoms C, Wolff M, Padmakumar K, Ebel R, Proksch P (2004) Chemical defense of Mediterranean sponges Aplysina cavernicola and Aplysina aerophoba. Z Naturforsch 59c: 113-122 Toth GB, Langhamer O, Pavia H (2005) Inducible and constitutive defenses of valuable seaweed tissues: consequences for herbivore fitness. Ecology 86: 612-618 Toth GB, Pavia H (2007) Induced herbivore resistance in seaweeds: a meta-analysis. J Ecol 95: 425-434 Turon X, Becerro MA, Uriz MJ (1996) Seasonal patterns of toxicity in benthic invertebrates: the encrusting sponge Crambe crambe (Poecilosclerida). Oikos 75: 33-40 Turon X, Becerro MA, Uriz MJ (2000) Distribution of brominated compounds within the sponge Aplysina aerophoba: coupling of X-ray microanalysis with cryofixation techniques. Cell Tissue Res 301: 311-322 Turon X, Tarjuelo I, Uriz MJ (1998) Growth dynamics and mortality of the encrusting sponge Crambe crambe (Poecilosclerida) in contrasting habitats: correlation with population structure and investment in defence. Funct Ecol 12: 631-639 Uriz MJ, Turon X, Becerro MA, Agell G (2003) Siliceous spicules and skeleton frameworks in sponges: origin, diversity, ultrastructural patterns, and biological functions. Microsc Res Tech 62: 279-299 Uriz MJ, Turon X, Becerro MA, Galera J (1996) Feeding deterrence in sponges. The role of toxicity, physical defenses, energetic contents, and life-history stage. J Exp Mar Biol Ecol 205: 187204 Usher KM, Kuo J, Fromont J, Sutton DC (2001) Vertical transmission of cyanobacterial symbionts in the marine sponge Chondrilla australiensis (Demospongiae). Hydrobiologia 461: 15-23 van Alstyne KL, Houser LT (2003) Dimethylsulfide release during macroinvertebrate grazing and its role as an activated chemical defense. Mar Ecol Prog Ser 250: 175-181

van Alstyne KL, McCarthy JJ, Hustead CL, Kearns LJ (1999) Phlorotannin allocation among tissues of northeastern pacific kelps and rockweeds. J Phycol 35: 483-492 van de Vyver G, Huysecom J, Braekman JC, Daloze D (1990) Screening and bioassays for toxic substances in sponges from western Mediterranean Sea and North Brittany. Vie Milieu 40: 285-292 van Hulten M, Pelser M, van Loon LC, Pieterse CMJ, Ton J (2006) Costs and benefits of priming for defense in Arabidopsis. Proc Natl Acad Sci USA 103: 5602-5607 Vermeij GJ (1978) Biogeography and adaption patterns of marine life. Harvard University Press, Cambridge Waddell B, Pawlik JR (2000) Defenses of Caribbean sponges against invertebrate predators. I. Assays with hermit crabs. Mar Ecol Prog Ser 125: 125-132 Wajant H, Effenberger F (1996) Hydroxynitrile lyases of higher plants. Biol Chem 377: 611-617 Walters KD, Pawlik JR (2005) Is there a trade-off between woundhealing and chemical defenses among Caribbean reef sponges? Integr Comp Biol 45: 352-358 Weber FJ, deBont JAM (1996) Adaptation mechanisms of microorganisms to the toxic effects of organic solvents on membranes. Biochim Biophys Acta-Rev Biomembr 1286: 225-245 Weiss B, Ebel R, Elbrchter M, Kirchner M, Proksch P (1996) Defense metabolites from the marine sponge Verongia aerophoba. Biochem Syst Ecol 24: 1-12 Wiens M, Korzhev M, Krasko A, Thakur NL, Perovic-Ottstadt S, Breter HJ, Ushijima H, Diehl-Seifert R, Mller IM, Mller WEG (2005) Innate immune defense of the sponge Suberites domuncula against bacteria involves a MyD88-dependent signaling pathway Induction of a perforin-like molecule. J Biol Chem 280: 2794927959 Wittstock U, Gershenzon J (2002) Constitutive plant toxins and their role in defense against herbivores and pathogens. Curr Opin Plant Biol 5: 300-307 Wright JT, Benkendorff K, Davis AR (1997) Habitat associated differences in temperate sponge assemblages: the importance of chemical defence. J Exp Mar Biol Ecol 213: 199-213 Wulff JL (2005) Trade-offs in resistance to competitors and predators, and their effects on the diversity of tropical marine sponges. J Anim Ecol 74: 313-321 Wulff JL (2006) Ecological interactions of marine sponges. Can J Zool 84: 146-166 Zangerl AR (1986) Leaf value and optimal defense in Pastinaca sativa L. (Umbelliferae). Am Midl Nat 116: 432-436 Zangerl AR, Nitao JK (1998) Optimal defense, kin conflict and the distribution of furanocoumarins among offspring of wild parsnip. Evolut Ecol 12: 443-457 Zangerl AR, Rutledge CE (1996) The probability of attack and patterns of constitutive and induced defense: a test of optimal defense theory. Am Nat 147: 599-608

Porifera research: Biodiversity, innovation and sustainaBility - 2007

639

Influence of temperature on primmorph production in Petrosia ficiformis (Porifera: Demospongiae)

Laura Valisano(1), Attilio Arillo(1), Giorgio Bavestrello(2), Marco Giovine(3), Carlo Cerrano(1*)
Dipartimento per lo Studio del Territorio e delle sue Risorse, Universit di Genova, Corso Europa 26, 16132 Genova, Italy. Tel: 010/3538563. Fax: 010/3538220. valisano@dipteris.unige.it, arillo@unige.it, cerrano@dipteris.unige.it (2) Dipartimento di Scienze del Mare, Universit Politecnica delle Marche, Via Brecce Bianche, 60131, Ancona, Italy. g.bavestrello@univpm.it (3) Dipartimento di Biologia, Universit di Genova, Viale Benedetto XV, 16132 Genova, Italy. mgiovine@unige.it
(1)

Abstract: This study focused on the influence of temperature on primmorph formation for the Mediterranean demosponge Petrosia ficiformis. In particular two temperatures, 12C and 24C were tested. They represent respectively the winter minimum and the summer maximum of the annual range of the natural environment of the species. At 12C the primmorphs showed a high level of survival and growth reaching about 0.1 mm2 in 4 weeks. On the contrary at 24C although a high number of primmorphs was initially formed, this number drastically decreased, while the size of aggregates remained small, indicating a high mortality rate. The reared primmorphs produced during the experimental time up to 20 spicules each, about 100 m long, demonstrating to be a powerful model for in vitro studies on spiculogenesis in siliceous sponges. Keywords: Cell culture, Primmorphs, spiculogenesis, temperature, Demospongiae

Introduction
In the last years, the discovery of a great number of new bioactive compounds (Mller et al. 2000) produced by marine sponges has spurred the studies on sponge cell culture in order to obtain large amount of biomass. However, a limiting step for a development of biological researches on sponges is the availability of experimental models suitable for in vitro studies. The results of the various attempts to obtain large-scale production of sponge cell lines are still debated (Pomponi and Willoughby 1994, Klautau et al. 1994, Ilan et al. 1996, Mller et al. 1996) but the most promising results have been obtained with the development of a 3D cell aggregate named primmorphs (Custdio et al. 1998). Dissociated cells are in fact able to re-aggregate, in presence of Ca2+ and Mg2+, thanks to their property of self/non self recognition (Gaino et al. 1999), forming round shaped three dimensional structures, which represent the intermediate stage between a diamorphic multicellular aggregate characterized by a continuous pinacoderm (Borojevic et al. 1967) and the initial developmental stage of a complete pumping specimen (Custdio et al. 1998). While sponge cells turn into telomerase negative state when maintained into single cell suspension and die very quickly due to a process of apoptosis (Pomponi and Willoughby 1994, Koziol et al. 1998, Wegner et al. 1998), in primmorphs cells seem to maintain high level of telomerase (Koziol et al. 1998), suggesting the possibility to obtain the complete reorganization of the entire organism after dissociation (Wilson 1907, Diaz 1979).

Up today, primmorphs have been obtained from 25 sponge species (Table 1) with wide differences in the number, size and growth dynamic of the aggregates (Valisano et al. 2006a). Moreover studies on Petrosia ficiformis (Poiret, 1798) suggested that primmorphs production is strongly influenced by the seasonal cycle of the sponge (Valisano et al. 2006b). Few species are better investigated. Primmorphs from Dysidea avara and Suberites domuncula were mostly used as models for biological studies (e.g. biosilification in primmorphs of S. domuncula) (Mller et al. 2005) and for the in vitro production of bioactive compounds (e.g. avarol from D. avara) (Mller et al. 2000). Primmorphs from Hymeniacidon perleve and Petrosia ficiformis were used to understand respectively the role of archaeocytes (Zhang et al. 2003, Sun et al. 2006) and the influence of different factors such as seasonality or silica availability (Valisano et al. 2006b, 2007) in primmorphs formation. It is known that thermal oscillations can affect sponge biology at different level, modulating biomineralization processes (Simpson 1984), inducing the loosing of autotrophic symbionts (Cerrano et al. 2001), producing variations in the oxygen consumption and pumping rate (Zocchi et al. 2003), and increasing the intracellular Ca2+ mobilisation via cADPR until cell death (Zocchi et al. 2001). The aim of this work is to study the effect of culture temperature on primmorph formation and spicule production, using primmorphs of P. ficiformis, a species with a simple spicular complement composed only by monoaxonic spicules.

640

Table 1: Review of the largest size reached by primmorphs formed by species already analysed in literature (* refers to the general size range reported for the group of species analysed in the paper). Species Cliona celata Agelas oroides Axinyssa aurantiaca Axinella polypoides Geodia cydonium Haliclona oculata Halichondria panicea Pseudosuberites aff. andrewsi Stylissa massa Axinella damicornis Haliclona fulva Spirastrella cunctatrix Acanthella acuta Phorbas fictitius Hemimycale columella Pleraplysilla spinifera Batzella inops Suberites domuncula Dysidea avara Stylotella agminata Petrosia ficiformis Axinella verrucosa Hymeniacidon perleve Ircinia muscarum Xestospongia muta Largest size 0.004 mm2 0.041 mm2 0.074 mm2 0.2 - 2 mm * 0.2 - 2 mm * 0.2 - 2 mm * 0.2 - 2 mm * 0.2 - 2 mm * 0.2 - 2 mm * 0.562 mm2 0.562 mm2 0.585 mm2 0.785 mm2 1.045 mm2 1.676 mm2 1.85 mm2 1.858 mm2 1-2 mm 1-5 mm 1.50 m 2.457 mm2 no data no data no data no data Reference Valisano et al. 2006 Valisano et al. 2006 Valisano et al. 2006 Sipkema et al. 2003 Sipkema et al. 2003 Sipkema et al. 2003 Sipkema et al. 2003 Sipkema et al. 2003 Sipkema et al. 2003 Valisano et al. 2006 Valisano et al. 2006 Valisano et al. 2006 Valisano et al. 2006 Valisano et al. 2006 Valisano et al. 2006 Valisano et al. 2006 Valisano et al. 2006 Custdio et al. 1998 Muller et al. 2000 Zhang et al. 2003 Valisano et al. 2006 Nickel et al. 2003 Zhang et al. 2003 De Rosa et al. 2001 Richelle-Maurer et al. 2003

11.23 x 104 cells/ml-1 (means SD), on an oscillating table to avoid cells to attach on the bottom of plates and every three days one third of seawater was replaced with new filtered sea water. Twelve replicates were put at the temperature of 12C, the temperature we used in previous reports (Valisano et al. 2006a, 2006b) for primmorphs production, while twelve replicates were put at 24C. The two temperatures of 12C and 24C represent respectively the winter minimum and the summer maximum of temperature of sea water in the Northern Mediterranean Sea (average values). For three weeks the size of primmorphs was recorded every three days from six replicates maintained at the two temperatures. Measures were taken at a stereomicroscope, and according to Sipkema et al. (2003) areas of primmorphs were calculated assuming that primmorphs are perfect spheres. From the other six replicates at both 12C and 24C, primmorphs were picked for spicules analysis. With this purpose three primmorphs were picked up from different samples for each experimental set, for a total of 54 primmorphs analysed and 336 spicules measured. Primmorphs were observed on glass slides at the optic microscope and spicules were identified thanks to their bright light reflection properties (Schreder 2005). The number, the length and the thickness of spicules within primmorphs were measured every three days for three weeks.

Results
The culture temperature strongly affected both the rate of survival and the final size of primmorphs of P. ficiformis. Dissociated cells maintained at 12C formed in three days numerous small primmorphs that in the following weeks aggregated reaching in four weeks a size of about 0.1 mm2 (Fig. 1). On the contrary at 24C, after three days the number of aggregates formed is comparable, but this number drastically decreased in the first week until about 10 primmorphs of reduced size (0.04 mm2), thus demonstrating a high mortality rate (Fig. 1). Moreover the temperature of 24C stimulated the proliferation of ciliates, not evidenced in the set maintained at 12C. In both the experimental sets the spicule formation was evident about one week after the start of the experiment, when 1-5 thin spicules were detected in each primmorph. These values remained almost constant for about three weeks when the spicule number in both the experimental sets drastically increased reaching about 20 spicules per primmorph (Fig. 2). The trend of spicule length agreed with that of their number in the aggregate: the spicules remained about 40 m long for about three weeks and after quickly increased their size resulting, at the end of the experiment, about 100 m length (Fig. 3). The spicule width increased until about 2 m in the first two weeks and remained almost constant until the end of the experiment (Fig. 4). No differences between the two cultures temperature were recorded both in number and size of spicules produced.

Materials and methods

Specimens of Petrosia ficiformis were collected on the rocky cliff of the Marine Protected Area of the Portofino Promontory (Ligurian Sea, Italy) between 15-20 m depth during July 2005. In this month during a previous study a high number of primmorphs, high rates of survival and aggregation were recorded (Valisano et al. 2006b). Sponges were immediately carried to the laboratory and maintained in aquaria at 12C and salinity at 38. The day after sampling, sponges were processed to dissociate cells according to Mller et al. (1999). Sponge samples of 4 to 5 cm3, continuously submersed in sea-water, were cut in small pieces and transferred into 50 ml conical tubes filled with CMFSW-EDTA. After gentle shaking for 20 minutes, the solution was discarded and new CMSFWEDTA was added. After continuous shaking for 40 minutes the supernatant was collected and filtered through a 40 m mesh nylon net and the process repeated once. Samples were centrifuged (1600 rpm, 458 g for 5 minutes) and washed twice in CMFSW. Cells in final pellets were re-suspended in natural filtered sea-water and dissociated cells were put into tissue culture plastic plates, with inoculum cell density of 360

Discussion
This study was focused on the influence of temperature of culture on primmorphs formation in the Mediterranean

641

Fig. 2: Number of spicules formed within primmorphs at 12 and 24C (means SE).

Fig. 1: Size and number of primmorphs formed at 12 and 24C (means SE).

Petrosia ficiformis. In particular we chose two temperatures that are the winter minimum and the summer maximum of the annual range in the environment of the species. Our results showed an abundant primmorphs formation and a high rate of survival in the experimental set maintained at 12C. In this condition we have obtained a dynamic of aggregation completely overlapped to that already observed for specimens collected in July (Valisano et al. 2006b). On the contrary at 24C the rate of survival was low and the maximal size reached by primmorphs very small, in agreement with the situation recorded culturing primmorphs of sponges collected in September (Valisano et al. 2006b). Similar results were observed in Suberites domuncula, whose cells showed an evident and rapid decrease of viability at 28C while temperatures lower than 22C minimize the death rates. The low death rate recorded at 12-16C correspond with the temperature S. domuncula experienced in its habitat (Sipkema et al. 2003). It is likely that the optimal water temperature for primmorph production is related to the ecological requirement of different species. Higher temperatures (25-30C) resulted to positively affect the formation of primmorphs for the Chinese sponge Stylotella agminata even if the fungal contamination is higher at this temperature (Zhang et al. 2003). On the contrary in

Fig. 3: Length of spicules formed within primmorphs at 12 and 24C (means SE).

Fig. 4: Thickness of spicules formed within primmorphs at 12 and 24C (means SE).

642

the same species no primmorphs formation was recorded at 15C. While many authors suggest that without supplementing silica to culture medium, no spicules formation occurs (Krasko et al. 2000), we demonstrated that primmorphs of P. ficiformis are able to produce spicules already in the first phases of aggregation using the low silica amount available in the medium. During the period of rearing the number and size of the newly formed spicules increased reaching in four weeks a length comparable with that of the species in natural conditions (Bavestrello et al. 1994) while the width remained very thin probably due to insufficient silica availability in the culture medium. It is generally assumed that spiculogenesis is a two-steps process with the increasing in length due to the growth of the proteinaceous filament and the increasing in width affected by silica concentration and water temperature (Uriz 2006, Mller et al. 2006). This two-steps process provides the explanation for the first increase in length of spicules, due to the formation of their axial filament, and the consequent increase of width of spicules, due to the apposition on it of silica particles, as described in literature. In the marine sponge Microciona prolifera, lower temperatures stimulate the formation of wider spicules, suggesting a more efficient up-take and transport of silica (Simpson 1978). Fry (1970) demonstrated that Ophlitaspongia seriata spicules show different size frequency distributions in different populations from the Northern France and Welsh coasts. In agreement with these evidences Bavestrello et al. (1993) put in evidence a strong influence of depth on the spicule size of P. ficiformis; these changes can be considered related to the thermal gradient of the column water during the year. Nevertheless other studies indicate that in Halichondria panicea the silica uptake is not influenced by temperature, but is related to dissolved silica concentration, while temperature is responsible of the level of polymerization (Frlich and Barthel 1997). Pozzolini et al. (2004) demonstrated primmorphs of P. ficiformis express the silicatein gene, and here we quantify spicules production, highlighting the level of functionality reached by these aggregates and confirming once again they could be a powerful model for in vitro studies of spiculogenesis in siliceous sponges.

References
Bavestrello G, Bonito M, Sar M (1993) Influence of depth on the size of sponge spicules. Sci Mar 57: 415-420 Bavestrello G, Pansini M, Sar M (1994) The variability and taxonomic status of different Petrosia-like sponges in the Mediterranean Sea. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 83-92 Borojevic R, Fry WG, Jones WC, Lvi C, Rasmont R, Sar M, Vacelet J (1967) Mise au point actuelle de la terminologie des ponges (A reassessment of the terminology for sponges). Bull Mus Natl Hist Nat 39: 1224-1235 Cerrano C, Magnino G, Sar A, Bavestrello G, Gaino E (2001) Necrosis in a population of Petrosia ficiformis (Porifera,

Demospongiae) in relations with environmental stress. Ital J Zool 68: 131-136 Custdio MR, Prokic I, Steffen R, Koziol C, Borojevic R, Brmmer F, Nickel M, Mller WEG (1998) Primmorphs generated from dissociated cells from the sponge Suberites domuncula: a model system for studies of cell proliferation and cell death. Mech Ageing Dev 105: 45-59 De Rosa S, De Caro S, Tommonaro G, Slantchev K, Stefanov K, Popov S (2001) Development in a primary cell culture of the marine sponge Ircinia muscarum and analysis of the polar compounds. Mar Biotechnol 3: 281-286 Diaz J (1979) Variations, differentiations et functions des categories cellulaires de la Demosponge deau saumaitre, Suberites massa Nardo, au cours du cycle biologique annuel et dans des conditions experimentales. Thse Univ Sci Tech Languedoc, Montpellier Frlich H, Barthel D (1997) Silica uptake of the marine sponge Halichondria panicea in Kiel Bight. Mar Biol 128: 115-125 Fry WG (1970) The sponge as a population: a biometric approach. In: Fry WG (ed). The Biology of the Porifera. Academic Press, London. pp. 135-161 Gaino E, Bavestrello G, Magnino G (1999) Self/non-self recognition in sponges. Ital J Zool 66: 299-316 Ilan M, Contini H, Carmeli S, Rinkevich B (1996) Progress towards cell cultures from a marine sponge that produces bioactive compounds. J Mar Biotechnol 4: 145-149 Klautau M, Custdio MR, Borojevic R (1994) In vitro culture of primary cell lines from marine sponges. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 401406 Koziol C, Borojevic R, Steffen R, Mller WEG (1998) Sponges (Porifera) model systems to study the shift from immortal to senescent somatic cells: the telomerase activity in somatic cells. Mech Ageing Dev 100: 107-120 Krasko A, Batel R, Schreder HC, Mller IM, Mller WEG (2000) Expression of silicatein collagen genes in the marine sponge Suberites domuncula is controlled by silicate and myotrophin. Eur J Biochem 167: 4878-4887 Mller WEG, Steffen R, Rinkevich B, Matranga V, Kurelec B (1996) The multixenobiotic resistance mechanism in the marine sponge Suberites domuncula: its potential applicability for the evaluation of environmental pollution by toxic compounds. Mar Biol 125: 165-170 Mller WEG, Wiens M, Batel R, Steffen R, Schreder HC, Borojevic R, Custdio MR (1999) Establishment of a primary cell culture from a sponge: primmorphs from Suberites domuncula. Mar Ecol Prog Ser 178: 205-219 Mller WEG, Bohm M, Batel R, De Rosa S, Tommonaro G, Mller IM, Schreder HC (2000) Application of cell culture for the production of bioactive compounds from sponges: synthesis of avarol from Dysidea avara. J Nat Prod 63: 1077-1081 Mller WEG, Rothenberger M, Boreiko A, Tremel W, Reiber A, Schreder H (2005) Formation of siliceous spicules in the marine demosponge Suberites domuncula. Cell Tissue Res 321: 285-297 Mller WEG, Belikov S I, Tremel W, Perry C, Gieskes WWC, Boreiko A, Schreder H (2006) Siliceous spicules in marine demosponges (example Suberites domuncula). Micron 37: 107120 Pomponi SA, Willoughby R (1994) Sponge cell culture for production of bioactive metabolites. In: van Soest RWM, van Kempen TMG,

643

Braekman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 395-400 Pozzolini M, Sturla L, Cerrano C, Bavestrello G, Camardella L, Parodi AM, Raheli F, Benfatti U, Mller WEG, Giovine M (2004) Molecular cloning of silicatein gene from the sponge Petrosia fciformis (Porifera, demospongia) and development of primmorphs as a model for biosilification. Mar Biotechnol 6: 594-603 Richelle-Maurer E, Gomez R, Braekman JC, van de Vyver G, van Soest RWM, Devijver C (2003) Primary cultures from the marine sponge Xestospongia muta (Petrosiidae, Haplosclerida). J Biotechnol 100: 169-176 Schreder HC, Perovic-Ostadt S, Grebenjuk VA, Engel S, Mller IM, Mller WEG (2005) Biosilica formation in spicules of the sponge Suberites domuncula: synchronous expression of a gene cluster. Genomics 85: 666-678 Simpson TL (1978) The biology of the sponge Microciona prolifera (Ellis and Solander). III. Spicule secretion and the effect of temperature on spicule size. J Exp Mar Biol Ecol 35: 31-42 Simpson TL (1984) The cell biology of sponges. Springer-Verlag, New York Sipkema D, Van Wielink R, Van Lammeren AAM, Tramper J, Osinga R, Wijffels RH (2003) Primmorphs from seven marine sponges: formation and structure. J Biotechnol 100: 127-139 Sun LM, Song YF, Qu Y, Yu XJ, Zhang W (2006) Purification and in vitro cultivation of the archaeocytes (stem cells) of the marine sponge Hymeniacidon perleve (Demospongiae). Cell Tissue Res 328: 223-237 Uriz JM (2006) Mineral skeletogenesis in sponges. Can J Zool 84: 322-356 Valisano L, Bavestrello G, Giovine M, Cerrano C (2006 a) Primmorphs formation dynamics: a screening among Mediterranean sponges. Mar Biol 149: 1037-1046

Valisano L, Bavestrello G, Giovine M, Arillo A, Cerrano C (2006b) Seasonal production of primmorphs from the marine sponge Petrosia ficiformis (Poiret, 1789) and new culturing approaches. J Exp Mar Biol Ecol 337: 171-177 Valisano L, Bavestrello G, Giovine M, Arillo A, Cerrano C (2007) Effect of iron and dissolved silica on primmorphs of Petrosia ficiformis (Poiret, 1789) Chem Ecol 23: 233-241 Wegner C, Steffen R, Koziol C, Batel R, Lacorn M, Steinbart H, Simat T, Mller WEG (1998) Apoptosis in marine sponges: a biomarker for environmental stress (cadmium and bacteria). Mar Biol 131: 411-421 Wilson HV (1907) On some phenomena of coalescense and regeneration in sponges. J Exp Zool 5: 245-258 Zhang W, Zhang X, Cao X, Xu J, Zhao Q, Yu X, Jin M, Deng M (2003) Optimizing the formation of in vitro sponge primmorphs from the Chinese sponge Stylotella agminata (Ridley). J Biotechnol 100: 161-168 Zhang XY, Cao XP, Zhang W, Yu XJ and Jin MF (2003) Primmorphs from archaeocytes-dominant population of the sponge Hymeniacidon perleve: Improved cell proliferation and spiculogenesis. Biotechnol Bioeng 84(5): 583-590 Zocchi E, Carpaneto A, Cerrano C, Bavestrello G, Giovine M, Bruzzone S, Guida L, Luisa F, Usai C (2001). The temperaturesignaling cascade in sponges involves a heat-gated cation channel, abscisic acid and cyclic ADP-ribose. Proc Natl Acad Sci 98(26): 14859-14864 Zocchi E, Basile G, Cerrano C, Bavestrello G, Giovine M, Buzzone S, Usai C (2003) ABA and cADPR-mediated effects on respiration and filtration downstream of the temperature-signaling cascade in sponges. J Cell Sci 116: 629-636

Porifera research: Biodiversity, innovation and sustainaBility - 2007

645

Mass occurrence of Rossella nodastrella Topsent on bathyal coral reefs of Rockall Bank, W of Ireland (Lyssacinosida, Hexactinellida)
Rob W.M. van Soest(1*), Fleur C. van Duyl(2), Connie Maier(2), Marc S.S. Lavaleye(2), Elly J. Beglinger(1), Konstantin R. Tabachnick(3)
Zoological Museum of the University of Amsterdam, P.O. Box 94766, 1090 GT Amsterdam, the Netherlands. soest@science.uva.nl, beglinger@science.uva.nl (2) Royal Netherlands Institute for Sea Research, P.O. Box 17, 1790 AB Den Burg, the Netherlands. duyl@nioz.nl, maier@nioz.nl, lava@nioz.nl (3) Department of Bottom Fauna, Institute of Oceanology, Academy of Sciences of Russia, Nahimovsky 36, Moscow, Russia. tabachnick@mail.ru
(1)

Abstract: We report on the mass occurrence of a hexactinellid species, Rossella nodastrella Topsent, 1915 in coldwater coral reefs W of Ireland. The species was until now known only from the holotype, a single small specimen collected off the Azores. In recent boxcore sampling at a reef mound situated in 580 m water at 55.4N 15.7W along the south-eastern slope of Rockall Bank numerous specimens of this species were collected, showing a range of morphologies from 1 cm high, spiny urn-shaped individuals to megabenthic cup-shaped forms of 30-40 cm high and wide. We provide an extensive redescription of the species including SEM images of all the microscleres. Underwater video transects showed this species to be densely concentrated, up to approx. 6 specimens per m2, over a distance of more than 1 km. Presence of individuals was negatively correlated with live coral cover. This dense concentration was a local phenomenon, because the species was virtually absent in nearby similar habitats. Keywords: bathyal, coral reefs, Hexactinellida, mass occurrence, North Atlantic

Introduction
Diversity of Hexactinellida in the North East Atlantic is limited to approx. 45 species (van Soest et al. 2006; http:// www.marinespecies.org/porifera/), but this is compensated by reported mass occurrences of several of these species. This is especially documented for Pheronema carpenteri (Thompson, 1869) (see Rice et al. 1990, Barthel et al. 1996) and Schaudinnia rosea (Fristedt, 1887) (cf. Klitgard and Tendal 2004). These mass occurrences were reported from areas where also bathyal coral reefs have been sighted, but the general impression conveyed by various studies is that they are indeed neighbouring these reefs but usually downslope from them, for reasons as yet unexplained (Rice et al. 1990). In recent cruises with the Dutch research vessel Pelagia in waters west of Ireland (Moundforce 2004, BIOSYS 2005) large build ups were encountered of an irregularly cup-shaped species initially thought to be Asconema aff. setubalense Kent, 1870 or Asconema aff. foliata Fristedt, 1887 (see van Soest and Lavaleye 2005). However, subsequent studies including exchange of specimens between us, made it clear that these large cups conformed in their skeletal characters to a forgotten species described by Topsent (1915) from a single small urn-shaped specimen collected in bathyal depths off the Azores as Rossella nodastrella. The species can now be

more completely described from a large series of specimens connecting tiny urn-shaped specimens to large cup-shaped specimens. It is the purpose of this paper to provide a redescription and quantitative information on its occurrence in the area W of Ireland.

Material and methods

Specimens were collected mainly by boxcoring. We used a boxcore of 50 cm diameter capable of bringing up 2000 cm2 of ocean bottom, including reef corals. Additional samples were obtained in a few trawl attempts, which skirted a local reef. In situ observations of larger Rossella nodastrella individuals were made from Hopper camera transects. Images were obtained by an analogue video camera hung into a frame with two strong light strobes, which was lowered to just above the bottom and moved along it at a speed of 2 miles per hour for one hour. Position of the camera was monitored from the surface, thus the image field width varied over transects between approx. 1.2 m and several meters, depending of the vertical movement of the ship relative to the bottom. Due to this, densities of observed sponges could only be approximated.

646

Position and depth of all collecting attempts that contained Rossella nodastrella specimens are given in Table 1. Specimens were preserved in 96% alcohol, with an occasional specimen kept dry. All specimens are incorporated in the collections of the Zoological Museum of the University of Amsterdam (ZMA). Microscopic sections were made tangential and perpendicular to the surface to study in situ arrangement of the spicules using light microscopy. Spicule mounts for SEM and light microscopy were made by boiling fragments in concentrated HNO3.

Table 1: Rossella nodastrella specimens collected during Moundforce 2004 and BIOSYS/HERMES 2005 at the SE slopes of Rockall Bank. All specimens have been incorporated in the collections of the Zoologisch Museum of the University of Amsterdam. Sample BX31-02 BX31-A BX37-01 BX10/5 BX10/6 BX11/2 DR18/2 DR31/10 BX37/3 BX46/1 BX63/1 BX71/1 BX71/2 BX88/1 BX89/1 BX89/3 BX90/1 BX93/1 BX96/1 BX96/3 BX96/8 BX97/1 DR111/1 BX114/2 BX114/3 BX153/4 BX154/2 BX157/3 BX158/1 BX159/1 BX161/2 BX173/1 Date 2004/9/1 2004/9/1 2004/9/2 2005/6/25 2005/6/25 2005/6/25 2005/6/26 2005/6/27 2005/6/28 2005/6/29 2005/7/1 2005/7/3 2005/7/3 2005/7/6 2005/7/6 2005/7/6 2005/7/6 2005/7/6 2005/7/6 2005/7/6 2005/7/6 2005/7/6 2005/7/8 2005/7/9 2005/7/9 2005/7/11 2005/7/11 2005/7/11 2005/7/11 2005/7/11 2005/7/11 2005/7/12 Depth 560 560 557 602 602 584 672 844 757 580 581 586 586 586 586 586 590 590 577 577 577 581 524 644 644 573 572 588 583 586 585 629 560 560 557 602 602 584 675 857 757 580 581 586 586 586 586 586 590 590 577 577 577 581 524 644 644 573 572 588 583 586 585 629 N 55.49395 55.49395 55.49642 55.49978 55.49978 55.49918 55.50347 55.51888 55.44457 55.49940 55.49932 55.50073 55.50073 55.50115 55.50123 55.50123 55.50093 55.50078 55.50112 55.50112 55.50112 55.50135 55.49365 55.44280 55.44280 55.49072 55.49110 55.50103 55.50098 55.50093 55.50105 55.44463 W -15.80390 -15.80390 -15.80215 -15.79815 -15.79815 -15.79795 -15.80405 -15.80765 -16.07537 -15.79832 -15.79832 -15.78927 -15.78927 -15.78788 -15.78798 -15.78798 -15.78795 -15.78770 -15.78868 -15.78868 -15.78868 -15.78848 -15.80088 -16.09748 -16.09748 -15.80132 -15.80110 -15.78830 -15.78857 -15.78840 -15.78842 -16.07178

Results Systematic description

Class Hexactinellida Schmidt, 1870 Subclass Hexasterophora Schulze, 1886 Order Lyssacinosida Zittel, 1877 Family Rossellidae Schulze, 1885 Subfamily Rossellinae Schulze, 1885 Genus Rossella Carter, 1872 Rossella nodastrella Topsent, 1915 Figs. 1A-N, 2A-I, 3A-H Rossella nodastrella Topsent, 1915: 1, figs 1-5; Topsent, 1928: 76, pl. III fig. 22, pl. IV fig. 3. Asconema aff. foliata; van Soest and Lavaleye, 2005: figs 2A-B. Material examined: All samples listed in Table 1 were examined. Additionally: BIOSYS/HERMES Hopper Camera Transect Station 109 (08-07-2005).

Description
Shape and size: Smaller specimens, from approx. 1 cm (Fig. 1C) to approx. 6 cm (Fig. 1K) are urn-shaped to tubular, with a spiny (Figs 1C, H, J-L, N) or occasionally smooth surface (Fig. 1M). They are thin-walled and have a large atrial cavity ending in a conspicuous oscule with fringeless rim, occupying approx. one third of the diameter of the sponge (Figs 1C, H, N). Occasionally, two oscules are found (Fig. 1K). Rarely, specimens were observed which appeared to be flabellate, i.e. their atrial cavity was exposed over the entire length of the individual (Fig. 1J); possibly these were damaged-and-repaired specimens. When reaching sizes over 6 cm in length, shape tends to alter into the Asconema type, i.e. tubular-trumpetshaped, with a predominantly smooth surface, widely flaring vent and recurved rims (Figs 1B, E-G, I). With increasing length, also the diameter increases to wider-than-high cups (Figs 1E, G, I). Shape in larger specimens, which may grow to reach sizes of 30-40 cm high and in diameter, may be highly irregular and is often based on a prostrate anchoring plate or lobe (Figs 1B, F). Individuals may show more than one vent and look distinctly mushroom-like (Fig. 1A). Detailed examination of a large number of adjacent specimens

Fig. 1: Habits and growth forms of Rossella nodastrella Topsent, 1915 individuals observed and collected at Rockall Bank. A. In situ image grabbed from a Moundforce 2003 Hopper Camera video, depth approx. 580 m (image size 1.5 x 1.2 m). B. In situ specimens (left and center) collected in Boxcore M2004-BX31 (diameter of boxcore 50 cm). C. Ditto, showing small urnshaped spiny individuals (largest approx. 2 cm high). D. A large cup-shaped specimen approx. 25 cm in diameter from boxcore BIOSYS-BX93. E. Cup-shaped specimen approx. 20 cm diameter from BIOSYS-BX71. F. Trumpet-shaped specimen 13 cm high and approx. 9 cm diameter, from M2004-BX31. G. Smaller cupshaped specimen approx. 10 cm diameter from BIOSYS-BX46. H. Three attached spined-tubular specimens, 5-6 cm high, from BIOSYS-BX71. I. Large cup-shaped specimen of approx. 25 cm diameter from BIOSYS-BX71. J. Specimen with exposed atrial tube, approx. 6 cm high, from BIOSYS-BX100. K. Larger tubular specimen approx. 7 cm high from BIOSYS-BX161. L. Tubular specimen of 4.5 cm high, from BIOSYS-BX10. M. Squat smooth small specimens 2 cm diameter, from BIOSYS-BX100. N. Larger (4.5 cm high) and smaller (1.5 cm high) spined-tubular specimens from BIOSYS-BX10.

647

648

observed in a Hopper Camera transect of Haas Mound suggests that many of the smaller or prostrate specimens are fragmented off nearby larger individuals, indicating clonal processes might be common. Colour: Greyish white. Consistency: Of smaller specimens fragile but keeping their shape when lifted out of the water; larger individuals collapse when taken out of the water and tear very easily. Their consistency is best described as similar to a wet towel or a thick wad of wet paper. Skeleton (Figs 2A, B): The dermal skeleton is built almost exclusively of spined stauractines (Fig. 2B) with rare spined pentactins mixed in. This beautiful network is carried by larger smooth hypodermal pentactines, with their long ray directed inward into the parenchyme. The parenchymal spicules making up most of the skeleton of the thin walls of the specimens are long smooth centrotylote diactines. Shorter diactines protrude up to approx. 1 mm from the dermal surface causing the spined surface which characterizes most of the younger/smaller specimens. The atrial skeleton consists entirely of spined hexactines. No special anchoring spicules are present. Microscleres are predominantly hemioxyhexasters, with the other hexaster-types only common in the smaller specimens, becoming rare in large specimens; the latter are positioned most commonly subatrially. Macrodiscasters/discasters and calycocomes are located mostly in the vicinity of atrial surface while microdiscohexasters are located close to the dermal surface. This is very unusual for Rossellidae, as in all other genera of this family the situation is opposite. Spicules: Megascleres (Figs. 2C-I) 1. Dermalia are stauractines (Fig. 2D), entirely spined, with blunt ending rays, each ray 50-113.4-147 x 4-5.6-9 m, and some pentactines (Fig. 2E), entirely spined, rays 117-150 x 6 m. Stauractines may have a short rudimental tubercle. 2. Hypodermal pentactines (Figs. 2G,H), orthotropal, smooth, except for rugose apices; tangential rays 285-393.6-520 x 1318.0-21 m, proximal rays 480-634.1-755 x 17-19.2-22 m. 3. Atrial hexactines (Fig. 2C), entirely spined, rays 108-133.7155 x 6-6.3-7 m. The proximal rays are usually slightly longer and more spined than the tangential and distal rays (see Table 2). 4. Short diactines (Fig. 2F), centrotylote, apically spined and somewhat swollen, 366-739.4-1520 x 6-7.4-9 m. 5. Long diactines (Fig. 2A), centrotylote, smooth, 2,2603,773.3-6,300 x 12-31.7-70 m; rare truncated diactines may have a swollen spined end (Fig. 2I). Microscleres (Fig. 3A-G) 1. Oxyoidal microscleres: oxyhexasters, hemioxyhexasters (Fig. 3B) and oxyhexactines. The hexasters invariably have two secondary rays (not three as Topsent described for the type), rays rugose or finely spined; rarely rays have a clawlike termination (Fig. 3C) and these may be considered hemionychexasters; malformed smaller thick-centred forms are not uncommon (Fig. 3D). Diameter: 48-67.9-93, with prima-

ry ray length: 2-4 m, secondary ray length: 23-30.4-38 m, diameter of primary rosette: 7-19 m. 2. Macrodiscasters (Fig. 3E), with smooth centre and approx. 20 spined rays, ends provided with toothed discs. Diameter: 56-95.9-120 m, ray length 28-43.6-54 m, diameter of primary rosette 10-30 m 3. Calycocomes (Figs. 3A, F, G), with smooth primary rays, crowned with 6-12 densely spined secondary rays each ending with (larger, Fig. 3G) or without (smaller, Fig. 3F) toothed discs. Diameter: 48-83.7-126 m, primary ray length: 3-6 m, secondary ray length: 18-31.2-39 m, diameter of primary rosette: 6-30 m. 4. Microdiscohexaster (Figs. 3A, H), with smooth primary rays, crowned with 12-20 sparingly spined small thin rays each ending in toothed discs (barely visible in light microscopy, so measurements were done in SEM preparations). Diameter: 14-32 m, primary ray length: 2 m, secondary ray length: 4-6 m, diameter of primary rosette: 8-12 m. Distribution and ecology: The type specimen was collected on August 18, 1911, at station 3140 of the cruises of the Prince of Monaco, close to Sao Miguel, Azores, 3738N 2601W, at a depth of 1378 m; it was fixed on the dead skeleton of another hexactinellid, Hertwigia falcifera Schmidt, 1880. The Rockall Bank specimens were collected on or at the fringe of reef mounds found at the SE slopes of Rockall Bank, 55.455.5N 15.6-16.1W, at depths of 524-857 m. Most often they were growing on dead coral branches adjacent to or in the midst of live corals of both species Lophelia pertusa (L., 1758) and Madrepora oculata L., 1758. Remarks: The smaller specimens reported here resemble Topsents (1915) drawing of the type and all spicule types reported by Topsent were found in our specimens, although presence of microscleres varied considerably among individuals. Nevertheless, some clear discrepancies in spicule sizes were found with data reported by Topsent from the type specimen (see Table 2): - proximal rays of the hypodermal pentactines are only 755 m in maximum length, compared to 2,000 m in the type - short diactines are up to 2-3 x larger in the type - long diactines are 2 x larger in the type - hexasters have two secondary rays, not three - the five hexaster types are all distinctly smaller than in the type. In spite of these differences, we refrain here from erecting a new (sub-)species, as the variation over geographic distance is not properly known. Topsents material consisted only of a single individual. In the absence of measurements of further specimens from other localities, the observed differences are here explained as individual variation caused by geographic separation. Our Rockall Bank individuals all originate from a small reef mound area of approx. 15 km2 in size, and it is conceivable that many were propagated by fragmentation (see below), which would explain the narrow range of variation of the spicule sizes (Table 2). The different branching condition of the hexasters should be verified in the type, because the drawing of Topsent shows only four in stead of six primary

649

Fig. 2: Rossella nodastrella Topsent, 1915, skeleton and megascleres. A. Overview of dermal stauractine network supported by parenchymal oxea bundles. B. Detail of stauractine network. C. Atrial hexactine. D. Dermal stauractine. E. Rare dermal pentactine. F. Short diactine with central knobs. G. Hypodermal pentactine from the side. H. Ditto from the front. I. Iare club-ended long diactine.

rays and the number of secondary rays visible in the drawing exactly matches the twelve secondary rays in the Rockall Bank hexaster. In the 1928 repetition of the description of the type, Topsent admits that there are also hexasters with two secondary rays, although the majority were still considered having three secondary rays. It is possible that Topsents drawing is not accurate and in fact shows a hexaster with two secondary rays. Should future material obtained from the Azores and elsewhere show consistent differences in spicule measurements with our material, then recognition of the Rockall Bank population at the subspecific level is probably warranted. Until then, we maintain the name Rossella nodastrella for it. A second North Atlantic species of Rossella, Rossella mortenseni Burton, 1928 was reassigned to Mellonympha by Koltun, 1967, as a junior synonym of M. velata (Thomson), but is considered a valid species of Mellonympha by Tabachnick (2002) (here confirmed), among other things because the type specimen lacks calycocomes, which are characteristic for Rossella. No other Rossella species are known to occur in

the North Atlantic; the genus has a predominantly AntarcticSouthern Ocean distribution.

In situ observations of the Rockall Bank populations

Of the 107 samples taken in the Rockall Bank area during Moundforce 2004 and BIOSYS/HERMES 2005 19 contained one or more specimens of Rossella nodastrella. The samples were taken at two locations approx. 15 km apart, dubbed HAAS and CLAN Mounds. Of 44 stations made at CLAN Mound only two contained a small individual each, whereas at 64 stations made at HAAS Mound 43 individuals were obtained divided over 17 stations. The dominant occurrence at HAAS mound was confirmed by in situ observations made from Hopper video camera transects. One particular transect at HAAS mound (Station 109) going uphill from approx. 5529.572 N/ 1547.213 W, depth 728 m, to approx. 5529.739 N / 1548.149 W, depth 529 m, showed extremely high densities of Rossella nodastrella in the upper half. Between 552 and 529 m, over a

650

Fig. 3: Rossella nodastrella Topsent, 1915, microscleres. A. overview of various hexasters. B. dominant oxyhexaster. C. rare claw-ending hemionychexaster. D. rare thick-centred, malformed oxyhexaster. E. macrodiscaster. F. calycocome with rounded ray apices. G. calycocome with disc-ended rays. H. microdiscohexaster.

Table 2: Spicule dimensions (m) reported by Topsent (1915, 1928) from the Azores type specimen and the Rockall Bank specimens. Azores Stauractines, rays Dermal pentactines, rays Hypodermal pentactines, tangential Hypodermal pentactines, proximal Hexactines, rays Short diactines Long diactines Oxyhexasters Macrodiscasters Calycocomes Microdiscohexasters 160 x 8 present 270 x 9-27 up to 2,000 200 x 11 900-4,000 x 4-20 12,000 x 100 100-120 170 175 27-37 Rockall Bank 50-147 x 4-9 117-150 x 6 285-520 x 13-21 480-755 x 17-22 108-150 x 6-7 366-1,520 x 6-9 2,260-6,300 x 12-70 53-78 84-120 48-93 16-32

distance of approx. 1350 m a total of 1387 individuals could be counted. Since image field width varied considerably, only an approximation of the observed bottom surface area can be given: we estimate this to be around 2100 m2, thus an average of 0.66 individuals per m2 were present over the second half of the transect. Locally, densities were as high as 6 large individuals per m2. The percentage living corals and the presence of Rossella nodastrella were negatively correlated: averaging 19% live coral in the presence of Rossella nodastrella vs. 42% live coral cover in its absence (see Fig. 4). The irregular shapes of the observed sponges and the frequent occurrence of larger individuals surrounded by a number of smaller individuals gives the strong impression that processes of propagation by fragmentation and regeneration of partly dead or damaged individuals could be a part of the life strategy this species. Crabs (Paramola cuvieri (Risso, 1816)) were observed carrying fragments of Rossella cups around as camouflage, a further indication that fragments may easily become isolated from the parent individuals. This

651

Fig. 4: Quantitative presence of Rossella nodastrella Topsent, 1915 individuals and % cover of live corals observed during 1 hour with a Hopper camera in a transect (stat. B2005/109) at Haas mound, going uphill from approx. 5529.572 N/ 1547.213 W, depth 728 m, to approx. 5529.739 N / 1548.149 W, depth 529 m, total distance of transect approx. 3200m. Rossella nodastrella individuals were counted in 10 second intervals and averaged for 5 min intervals. Likewise, % live coral cover was estimated in 10 seconds intervals and averaged for 5 min intervals.

would also explain why the species has such an extremely patchy distribution. However, the capability for regeneration of fragments has never been demonstrated in hexactinellid sponges in general, and in rossellid sponges in particular, so the remarks made here remain hypothetical. Other species recognizable in the video transect, Mellonympha velata (Thomson, 1873) (153 individuals) and Geodia macandrewi Bowerbank, 1858 (13 individuals) had a much lower density and were more evenly spread over the transect.

Acknowledgements
The material of this research was collected with grants from the EUROMARGIN Programme of the European Science Foundation (Moundforce 2004 Project, 813.03.006/855.01.040), the Netherlands Organisation for Scientific Research (BIOSYS project 814.01.005/835.20.024), and the EU HERMES Project (contract no. GOCE-CT-2005-511234). The following colleagues and shipmates provided assistance: Dr Gerard Duineveld (NIOZ), Dr Henk de Haas (NIOZ); Mr Arthur Palacs (International University, Bremen, Germany), and the captain and the crew of RV Pelagia (Royal Netherlands Institute for Sea Research).

References
Barthel D, Tendal OS, Thiel H (1996) A wandering population of the hexactinellid sponge Pheronema carpenteri on the continental slope off Morocco, Northwest Africa. PSZN Mar Ecol 17: 603616 Carter HJ (1872) On two new sponges from the Antarctic Sea, and on a new species of Tethya from Shetland; together with observations on the reproduction of sponges commencing from zygosis of the sponge-animal. Ann Mag Nat Hist (4) 9(54): 409-435

Bowerbank JS (1858) On the anatomy and physiology of the Spongiadae. Part I. On the spicula. Phil Trans Roy Soc 148(2): 279-332 Burton M (1928) Hexactinellida. Danish Ingolf Exped 6(4): 1-18 Fristedt K (1887) Sponges from the Atlantic and Arctic Oceans and the Behring Sea. Vega-Exped Vetensk Iakt (Nordenskild) 4: 401471 Kent WS (1870) On the Hexactinellidae or hexaradiate spiculed silicious sponges taken in the Norma Expedition off the coast of Spain and Portugal. With description of new species, and revision of the order. Month Microsc J 4: 241-252 Klitgaard AB, Tendal OS (2004) Distribution and species composition of mass occurrences of large-sized sponges in the northeast Atlantic. Progr Oceanogr 61(1): 57-98 Koltun VM (1967) Glass, or Hexactinellid sponges of the Northern and Far-Eastern Seas of the USSR (Class Hyalospongiae). [In Russian]. Opred faune SSR izd Zool muz Akad nauk 94: 1-124 Rice AL, Thurston MH, New AL (1990) Dense aggregations of a hexactinellid sponge, Pheronema carpenteri, in the Porcupine Seabight (northeast Atlantic Ocean), and possible causes. Progr Oceanogr 24: 179-196 Schmidt OS (1870) Grundzge einer Spongien-Fauna des Atlantischen Gebietes. Wilhelm Engelmann, Leipzig Schmidt OS (1880) Die Spongien des Meerbusen von Mexico (Und des caraibischen Meeres). Abtheilung II. Hexactinelliden. Heft II. In: Reports on the dredging under the supervision of Alexander Agassiz, in the Gulf of Mexico, by the USCSSBlake. Gustav Fischer, Jena. pp. 1-32 Schulze FE (1885) The Hexactinellida. In: Tizard TH, Moseley HM, Buchanan JY, Murray J (eds). Rep Sci Res Voy H.M.S. Challenger, 18731876. Narrative 1(1): 437-451 Schulze FE (1886) ber den Bau und das System der Hexactinelliden. Abhandl Knigl Akad Wiss Berlin (Phys-Mathem Classe) 1886: 197

652

Schulze FE (1887) Report on the Hexactinellida collected by H.M.S. Challenger during the years 18731876. Rep Sci Res Voy H.M.S. Challenger, Zool 21: 1-514 Tabachnick KR (2002) Family Rossellidae Schulze, 1885. In: Hooper JNA, van Soest RWM (eds). Systema Porifera: a guide to the classification of sponges. Kluwer Academic/Plenum Publishers, New York. pp. 1442-1508 Thomson CW (1869) On Holtenia, a genus of vitreous sponges. Proc Roy Soc London 18: 32-35 Thomson CW (1873) The depths of the sea. Macmillan and Co, London Topsent E (1915) Une Rossella des Aores (Rossella nodastrella n.sp.). Bull Inst ocanogr Monaco 303: 16

Topsent E (1928) Spongiaires de lAtlantique et de la Mditerrane provenant des croisires du Prince Albert Ier de Monaco. Rs Camp sci Prince Albert I Monaco 74: 1-376 van Soest RWM, Boury-Esnault N, Janussen D, Hooper JNA (2006) The world list of extant Porifera. http://www.marinespecies.org/ porifera/ (accessed on May 15, 2006) van Soest RWM, Lavaleye MSS (2005) Diversity and abundance of sponges in bathyal coral reefs of Rockall Bank, NE Atlantic, from boxcore samples. Mar Biol Res 1: 338-349 Zittel KA (1877) Studien ber fossile Spongien. 1: Hexactinellidae. Abhandl Mathem-Phys Classe Knigl-Bayer Akad Wiss 13 (1): 163

Porifera research: Biodiversity, innovation and sustainaBility - 2007

653

A novel biochemical method to distinguish cryptic species of Chondrilla (Chondrosida, Demospongiae) based on its sulfated polysaccharides
Eduardo Vilanova(1*), Carla Zilberberg(2), Michele Kochem(1), Mrcio R. Custdio(3), Paulo A.S. Mouro(1)
Laboratrio de Tecido Conjuntivo Hospital Universitrio Clementino Fraga Filho and Instituto de Bioqumica Mdica/ UFRJ. Av Brigadeiro Trompowsky s/n, sala 4a01, Rio de Janeiro (RJ), Brasil, CEP 21941-590. evilanova@hucff.ufrj.br (2) Dept. Biologia Celular e Gentica UERJ. Rua So Francisco Xavier 524, PHLC, sala 205, Rio de Janeiro (RJ), Brasil, CEP 20550-013. carlazilber@yahoo.com.br (3) Departamento de Fisiologia Instituto de Biocincias/USP. Rua do Mato n.321, Sala 300, So Paulo (SP), Brasil, CEP 05508-900. mcust@usp.br
(1)

Abstract: Sulfated polysaccharides from marine sponges are highly complex molecules with species specific composition. We now propose a novel biochemical method to distinguish cryptic species of Chondrilla, built on the analysis of the sulfated polysaccharides content. The major difference between the sulfated polysaccharides from Chondrilla australiensis and Chondrilla nucula is their sulfate content, which was enough to give different electrophoretic motilities on agarose gel. Additionally, the sulfated polysaccharides from the cryptic species C. nucula, Chondrilla sp. B, Chondrilla sp. E and Chondrilla sp. F also showed distinct electrophoretic motilities on agarose gel. This method allowed the distinction of two sympatric cryptic species of Chondrilla (sp. E and sp. F) found through allozymes by the presence of a diagnostic locus. Analysis of the sulfated polysaccharides has advantages over allozymes or DNA since it can be applied to specimens fixed either in ethanol, formaldehyde, frozen or dried. Keywords: marine sponges, molecular systematics, glyconectins

Introduction
The cellular adhesion and recognition of marine sponges (Porifera) is mediated by proteoglycan-like molecules, also called aggregation factors (AFs), spongicans or glyconectins (e.g. Fernndez-Busquets and Burger 2003, Guerardel et al. 2004, Misevic et al. 2004). These proteoglycan-like molecules are composed of a protein core attached to several sulfated polysaccharide units (Humphreys et al. 1977, Jarchow et al. 2000). The sulfated polysaccharide units of glyconectins are responsible for the cell-cell recognition and adhesion in sponges (Bucior and Burger 2004). The interaction between the sulfated polysaccharides of adjacent sponge cells is calcium dependent and a highly species specific event (Spillmann and Burger 1996, Bucior and Burger 2004, Misevic et al. 2004). The species specific interaction of the sulfated polysaccharides from glyconectins was demonstrated by the selective and homophilic aggregation of beads coated with sulfated polysaccharides from different sponges (Popescu and Misevic 1997, Misevic et al. 2004). Another evidence for the species specificity of sulfated polysaccharides from Porifera species is their chemical and structural diversity (Zierer and Mouro 2000, Guerardel et al. 2004). These sulfated polysaccharides are highly complex and all the

species previously studied showed polymers with different structures and/or sugar and sulfate content (Table 1). The taxonomy of sponges is an unsolved problem due to the low numbers of usable morphological characters to discriminate species (Sol-Cava et al. 1991, Sol-Cava 1994). Due to the lack of consistent morphological traits, many species of sponges are considered cosmopolitan (SolCava et al. 1991). However, recent studies using molecular markers, such as allozymes and DNA sequences, showed that many species considered cosmopolitan were actually complexes of cryptic species (e.g., Sol-Cava and Thorpe 1986, Boury-Esnault et al. 1992, Muricy et al. 1996, Klautau et al.1999, Lazoski et al. 2001, Loukaci et al. 2004, Usher et al. 2004). Although analysis of allozymes seems to have enough resolution to distinguish cryptic species, the use of this methodology is impounded due to the need of fresh or frozen samples (Wrheide et al. 2004). In addition, comparisons among allozymes and other currently available molecular markers in detecting cryptic species of sponges have yield conflicting results (Zilberberg 2006). Therefore, there is a great need to find novel markers for the detection of cryptic species of sponges.

654

Table 1: Chemical differences among the sulfated polysaccharides from marine sponges. Species Aplysina fulva Chondrilla nucula Cliona celata Dysidea robusta Halichondria panicea Hymeniacidon heliophila Microciona prolifera Myxilla rosacea Ophlithaspongia tenius Suberites ficus Sulfated polysaccharide HexUA, Glu, (sulfated) HexUA, Ara, Gal, Fuc, (sulfated) Sulfated HexNac, Ara, Fuc HexUA, Ara, Gal, Fuc 4-O-sulfated Gal Py(4,6), Fuc, GlcNac N-sulfated HexUA, Gal, Fuc, (sulfated) Gal, Fuc, Gal Py (4,6), GlcNac N-sulfated Glc 4,6-disulfated, Fuc 2,4-disulfated HexUA, Glc, GlcNac N-sulfated HexUA, GlcNac, Fuc, Man, Gal (sulfated) Reference Zierer and Mouro 2000 Zierer and Mouro 2000 Guerardel et al. 2004 Zierer and Mouro 2000 Guerardel et al. 2004 Zierer and Mouro 2000 Guerardel et al. 2004 Cimino et al. 2001 Parrish et al. 1991 Bucior and Burger 2004

Chondrilla (Demospongiae: Chondrillidae) is a good model for the detection of cryptic species due to the large number of cryptic species that have been found along the Atlantic and Pacific Oceans (Klautau et al. 1999, Usher et al. 2004, Zilberberg 2006, Zilberberg et al. 2006). Usher et al. (2004) found, through DNA sequence analyses, two cryptic species of Chondrilla australiensis along the western coast of Australia. Similarly, along the Atlantic Ocean there are about six to eight cryptic species of Chondrilla nucula, which have been found through allozymes or DNA sequence analyses (Klautau et al. 1999, Zilberberg 2006, Zilberberg et al. 2006). Based on the species specificity of the sulfated polysaccharides from sponges, we propose, here, a novel biochemical method to distinguish cryptic species within the genus Chondrilla through agarose gel electrophoresis of its sulfated polysaccharides. The chemical composition of the sulfated polysaccharides from the species C. nucula and C. australiensis were analyzed to evaluate the differences between sulfated polysaccharides from congeneric species. We also tested the efficiency of this methodology in detecting cryptic species of Chondrilla found through allozymes (Klautau et al. 1999, Zilberberg et al. 2006). Additionally, we tested the ability to analyze specimens fixed in different media, including formaldehyde, a preservative that makes the study of allozymes and DNA sequences unfeasible.

Brazil) were collected and one was fixed in 70% ethanol, one in 4% formaldehyde, one dried and one was frozen.

Extraction of the sulfated polysaccharides

Each specimen was cut into small pieces (1 mm3), washed with 70% ethanol, immersed tree times in acetone and dried at a 60C oven. Sulfated polysaccharides were extracted from the dried tissues (100 mg from C. australiensis and C. nucula, and 30 mg from all the other specimens) by extensive papain digestion, and the extracts were partially purified by cethylpyridinium and ethanol precipitations using the same methodology described for other invertebrate tissues (Vieira et al. 1991). Approximately 3 mg (dry weight) of crude extract was obtained from C. australiensis and C. nucula and 1 mg from the other species.

Purification of the sulfated polysaccharides

The crude extracts of sulfated polysaccharides (3 mg of each specimen) were applied to a DEAE cellulose column, equilibrated with 5 mM sodium acetate (pH 5.0) with 10 mM EDTA (ethylenediaminetetraacetic acid). The polysaccharides were eluted from the column using a linear gradient of 0-3 M NaCl, at a flow rate of 0.5 ml/min. Fractions of 0.5 ml were collected and checked by metachromatic assay using 1,9-dimethylmethylene blue (Farndale et al. 1986), and by measuring conductivity. The fractions containing sulfated polysaccharides were pooled, dialyzed against distilled water and lyophilized.

Material and methods Sponge samples

To analyze the sulfated polysaccharides from cryptic species of Chondrilla, two specimens from each of five species were used. These species were: Chondrilla australiensis (Melbourne, Australia); C. nucula (Marseille, France); two cryptic species found in sympatry in the Bahamas (Lee Stocking Island), named Chondrilla sp. E and sp. F (Zilberberg et al. 2006); and Chondrilla sp. B (Klautau et al. 1999), one individual from Ubatuba (So Paulo, Brazil) and another from Bzios (Rio de Janeiro, Brazil). All these specimens were fixed in 70% ethanol. Additionally, to compare the efficiency of this methodology using different fixatives, four specimens of Chondrilla sp. B from Arraial do Cabo (Rio de Janeiro,

Agarose gel electrophoresis

The crude extracts and purified sulfated polysaccharides were analyzed by agarose gel electrophoresis. The sulfated polysaccharides (15 g) were applied to a 0.5% agarose gel and run for 1 h at 110 V in a 0.05 M 1,3-diaminopropaneacetate buffer (pH 9.0). The sulfated polysaccharides in the gel were fixed with 0.1% N-cetyl-N,N,Ntrimethylammonium bromide solution. After 12 h, the gel was dried and stained with 0.1% toluidine blue in 0.1:5:5 acetic acid:ethanol:water.

655

Polyacrylamide gel electrophoresis

The molecular masses of the sulfated polysaccharides were estimated by polyacrylamide gel electrophoresis. Sulfated polysaccharides (15 g) were applied to a 6% 1 mm thick polyacrylamide gel slab in 0.02 M sodium barbital (pH 8.6). After electrophoresis (100 V for 30 min), the gel was stained with 0.1% toluidine blue in 1% acetic acid and then washed for about 4 h in 1% acetic acid. The molecular mass markers were the low-molecular-mass dextran sulfate (8 kDa), chondroitin 4-sulfate from shark cartilage (40 kDa) and high-molecular-mass dextran sulfate (500 kDa).

Chemical analysis
Total hexose and uronic acid were estimated by the phenolH2SO4 reaction (Dubois et al. 1956) and carbazole reaction (Dische 1947), respectively. After acid hydrolysis of the polysaccharides (6.0 M trifluoroacetic acid for 5 h at 100C), sulfate was measured by the BaCl2gelatin method (Saito et al. 1958). The presence of different neutral sugars was estimated by paper chromatography in 3:2:1 n-butanol pyridinewater for 48 h (Kircher 1954).

Results
Fractionation of the sulfated polysaccharides from C. australiensis resulted in a single and sharp peak, eluted with 0.5 M NaCl (Fig. 1A). C. nucula showed a single sulfated polysaccharide fraction too, but eluted with a higher NaCl concentration of 1M (Fig. 1B). These results indicate the presence of a single and homogeneous population of sulfated polysaccharides in the two species. The presence of a single population of sulfated polysaccharides and the purity of the fractions were confirmed by agarose gel electrophoresis. The sulfated polysaccharides from two specimens of either C. australiensis or C. nucula showed narrow bands with similar electrophoretical motility (Fig. 2). However, the electrophoretical motility of the sulfated polysaccharides from C. australiensis and C. nucula differs significantly (Fig. 2), which indicates that C. australiensis and C. nucula have distinct sulfated polysaccharides. Polyacrylamide gel electrophoresis showed that sulfated polysaccharides from C. australiensis and C. nucula have high molecular weights (approximately 500 kDa; Fig. 3). The differences between the sulfated polysaccharides from C. australiensis and C. nucula were evaluated by their sugar compositions, as well as, hexuronic acid and sulfate contents. The sulfated polysaccharides from C. australiensis and C. nucula showed the same sugar composition and similar hexuronic acid content (Table 2). However, the sulfate content of the sulfated polysaccharides from C. nucula was approximately 50% higher than that from C. australiensis (Table 2). Therefore, the major difference between the sulfated polysaccharides from C. australiensis and C. nucula is their sulfate content. The crude extract of sulfated polysaccharides from two specimens of C. nucula from France, two specimens of Chondrilla sp. B from Brazilian coast and two specimens of Chondrilla sp. E and two of Chondrilla sp. F from Bahamas were applied to an agarose gel electrophoresis (Fig. 4).

Fig. 1: Purification of the sulfated polysaccharides. The crude extracts of sulfated polysaccharides from C. australiensis (A) and C. nucula (B) (~3mg each) were purified by ion exchange chromatography (DEAEcellulose-FPLC). The samples were eluted by a linear gradient of 03 M NaCl. The fractions were checked by its metachromatic property () and NaCl concentration (). The fractions indicated by the horizontal bar corresponding to the purified sponge sulfated polysaccharides.

The sulfated polysaccharides from these cryptic species of Chondrilla showed distinct electrophoretical motilities. These differences in the electrophoretical motility show that the agarose gel electrophoresis of crude extract of sulfated polysaccharides has a good resolution to distinguish cryptic species of Chondrilla. In order to evaluate the effectiveness of the method for samples of sponges fixed in different media, we extracted sulfated polysaccharides from specimens of Chondrilla sp. B fixed either in 70% ethanol, 4% formaldehyde, frozen at -20oC or dried at 60oC. The electrophoretic motility of the sulfated polysaccharide was the same, independent of the method used to fix the sponge (Fig. 5).

Discussion
In the present study we show a new, simple and efficient method to distinct cryptic species within Chondrilla. We also demonstrate that this methodology can be performed with specimens fixed in formaldehyde, a preservative that

656

Fig. 2: Agarose gel electrophoresis of the purified sulfated polysaccharides from C. australiensis and C. nucula (~15 g of each).

Fig. 3: Polyacrylamide gel electrophoresis of the purified sulfated polysaccharides from C. australiensis and C. nucula (~15 g of each). The molecular mass markers were high-molecular-mass dextran sulfate (Dex500, 500 kDa), chondroitin 4-sulfate from whale cartilage (C-4-S, 40 kDa), and low-molecular-mass dextran sulfate (Dex8, 8 kDa).

Table 2: Chemical composition of the sulfated polysaccharides from C. nucula and C. australiensis. Species C. australiensis C. nucula
a

Sugar compositiona Ara, Gal, Fuc and HexUA

Total hexoseb,d 2.19 2.44

Total sulfateb,d 2.16 3.73

Hexuronic acidc,d 0.12 0.10

Sulfate/ total hexosec 0.99 1.53

The sugar composition was determined by paper chromatography of hydrolyzed sulfated polysaccharides. b nMoles/ml. c Molar ratio. d Total hexose, total sulfate and hexuronic acid were measured by phenol-sulfuric acid, BaCl2-gelatin and carbazole methods, respectvely.

impounds the use of allozymes or even DNA sequencing analyses. The sulfated polysaccharides from C. australiensis and C. nucula showed high molecular weight (~500 kDa), the same sugar composition (hexuronic acid, Ara, Fuc and Gal) and similar hexuronic acid content (12% and 10%, respectively). The only chemical difference between the sulfated polysaccharides from these species was the sulfate: total sugar molar ratio (1:1 and 1.5:1, respectively). Zierer and Mouro (2000) reported the chemical characterization of C. nucula from Arraial do Cabo, Brazil. This species is actually a cryptic species of C. nucula temporarily named Chondrilla sp. B (Klautau et al. 1999). The sulfated polysaccharide from Chondrilla sp. B contains hexuronic acid, Ara, Fuc and Gal, the hexuronic acid accounts for 25% of the total sugar and the sulfate:total sugar molar ratio is 2.5:1 (Zierer and

Mouro 2000). Therefore, the differences detected among the sulfated polysaccharides from C. australiensis, C. nucula and Chondrilla sp. B were mostly related to their sulfate and hexuronic acid content. This confirms the species specific composition of the sulfated polysaccharides from sponges, even among species of the same genus. Differences among sulfated polysaccharides from congeneric species have also been observed in the -L-fucans isolated from the jelly coat of eggs of four sea urchin species within the genus Strongylocentrotus (Alves et al. 1998). The differences among these fucans are only in the sulfate pattern and the position of the glycosidic bonds (Alves et al. 1998, Villela-Silva et al. 1999, 2001). These structural differences of sulfated polysaccharides are sufficient to avoid interspecific fertilization among these congeneric species (Biermann et al. 2004). The sulfated polysaccharides from the three species of

657

Fig. 4: Agarose gel electrophoresis of the crude extracts of sulfated polysaccharides from Chondrilla nucula, Chondrilla sp. B, and the two sympatric species Chondrilla sp. E and Chondrilla sp. F (15 g of each).

Chondrilla also showed differences in their sulfation pattern. However, we still need a structural characterization of these sulfated polysaccharides, such as the position of glycosidic bonds and sulfation sites to determine all their differences. The electrophoretic motility of sulfated polysaccharides in agarose gel is mostly determined by their interaction with 1,3-diaminopropane, which depends on the structure and sulfation pattern of the sulfated polysaccharide (Dietrich and Dietrich 1972, 1976). This methodology had enough resolution to separate sulfated polysaccharides with small structural differences. For instance, it can separate the glycosaminoglicans dermatan sulfate and condroitin-4sulfate, which differs exclusively on the type of hexuronic acid in the chains (glycuronic acid in chondroitin-4-sulfate and iduronic acid in dermatan sulfate) (Dietrich and Dietrich 1972, 1976). Therefore, agarose gel electrophoresis in 1,3diaminopropane buffer can be used to distinguish sulfated polysaccharides with small structural differences. Chondrilla nucula was once considered as cosmopolitan. However, a study using allozymes electrophoresis showed that C. nucula was in fact a complex of cryptic species (Klautau et al. 1999). In the present study, four cryptic species of Chondrilla detected through allozymes by the presence of at least one diagnostic locus (Zilberberg 2006, Zilberberg et al. 2006) were analyzed through their sulfated polysaccharides. The four species were separated by their sulfated polysaccharides, including the two sympatric and cryptic Bahamian species (named sp. E and sp. F; Zilberberg et al. 2006). This result demonstrates the good resolution of the technique to separate cryptic species of Chondrilla. Additionally sulfated polysaccharides analyses by agarose gel electrophoresis showed some advantages in relation to allozymes or DNA analyses. Allozyme electrophoresis techniques require fresh or frozen samples (Wrheide et al. 2004). Therefore, DNA sequencing analyses were advantageous over allozymes by the ability to work with

Fig. 5: Agarose gel electrophoresis of the crude extracts of sulfated polysaccharides from specimens of Chondrilla sp. B fixed in ethanol 70%, formaldehyde 4%, dried and frozen (~15 g of each).

dried or ethanol preserved specimens. However, DNA sequencing analyses are unfeasible with tissues preserved in formaldehyde, and most of the earlier preserved museum specimens used this fixative. Thus, sulfated polysaccharides

658

analyses have some advantages over allozymes and DNA, since it is a quick (~ 5 days) and very low cost technique (~ US$ 2.00 / sample), it requires a very small sample (30mg of sponge tissue), and most importantly, it can be performed in formaldehyde preserved specimens. We can conclude that the analysis of the sulfated polysaccharides by agarose gel electrophoresis is a promissory biochemical technique to distinguish cryptic species within Porifera. However, further analyses with a larger sample size, a higher number of cryptic species of Chondrilla and other sponge taxa must be performed to establish the general effectiveness and robustness of this technique.

Acknowledgements
We thank A.M. Sol-Cava and M. Maldonado for the collection of Chondrilla sp. E and Chondrilla sp. F from the Bahamas; K. Usher for the collection of C. australiensis, N. Boury-Esnault for the collection of C. nucula and F. Cavalcanti for the collection of Chondrilla sp. B. To Conselho Nacional de Desenvolvimento Cientfico e Tecnolgico (CNPq: FNDCT, PADCT, and PRONEX), Financiadora de Estudos e Projetos (FINEP), Fundao de Amparo Pesquisa do Estado do Rio de Janeiro (FAPERJ) for financial support.

References
Alves AP, Mulloy B, Moy GW, Vacquier VD, Mouro PAS (1998) Females of the sea urchin Strongylocentrotus purpuratus differ in the structures of their egg jelly sulfated fucans. Glycobiology 8: 939-394 Biermann CH, Marks JA, Vilela-Silva ACES, Castro MO, Mouro (2004) Carbohydrate-based species recognition in sea urchin fertilization: another avenue for speciation? Evol Dev 6: 353-361 Boury-Esnault N, Sol-Cava AM, Thorpe JP (1992) Genetic and cytological divergence between colour morphs of the Mediterranean sponge Oscarella lobularis Schmidt (Porifera, Demospongiae, Oscarellidae). J Nat Hist. 26: 271-284 Bucior I, Burger MM (2004) Carbohydrate-carbohydrate interaction as a major force initiating cell-cell recognition. Glycoconjugate J 21: 111-123 Cimino P, Bifulco G, Casapullo A, Gmez-Paloma L, Riccio R (2001) Isolation and NMR characterization of rosacelose, a novel sulfated polysaccharide from the sponge Myxilla rosacea. Carbohydr Res 334: 39-47 Dietrich CP, Dietrich SMC (1972) Simple micro method for identification of heparin and other acidic mucopolysaccharides from mammalian tissues. Anal Biochem 46: 209-215 Dietrich CP, Dietrich SMC (1976) Electrophoretical behaviour of acidic mucopolysaccharides in diamine buffers. Anal Biochem 70: 645-657 Dische Z (1974) A new specific color reaction of hexuronic acids. J Biol Chem 167: 189-198 Dubois M, Guilles KA, Hamilton JK, Rebels PA, Smith F (1956) Colorimetric method for determination of sugars and substances. Anal Chem 28: 350-356 Farndale RW, Buttle DJ, Barret AJ (1986) Improved quantification and discrimination of sulphated glycosaminoglycans by use of dimethylmethylene blue. Biochim Biophys Acta 883: 173-177

Fernandez-Busquets X, Burger MM (2003) Circular proteoglycans from sponges: first members of spongican family. Cell Mol Life Sci 60: 88-112 Guerardel Y, Czeszak X, Sumanovski LT, Karamanos Y, Popescu O, Strecker G, Misevic GN (2004) Molecular fingerprinting of carbohydrate structure phenotypes of three porifera proteoglycanlike glyconectins. J Biol Chem 279: 15591-15603 Humphreys S, Humphreys T, Sano J (1977) Organization and polysaccharides of sponge aggregation factor. J Supramol Struct 7: 339-351 Jarchow J, Fritz J, Anselmetti D, Calabro A, Hascall VC, Gerosa D (2000) Supramolecular structure of a new family of circular proteoglycans mediating cell adhesion in sponges. J Struct Biol 132: 95-105 Kircher HW (1960) Gas-liquid chromatography of methylated sugars. Anal Chem 32: 1103-1106 Klautau M, Russo CA, Lazoski C, Boury-Esnault N, Thorpe JP, Sol-Cava AM (1999) Does cosmopolitanism results from overconservative systematics? A case study using the marine sponge Chondrilla nucula. Evolution 53: 1414-1422 Lazoski C, Sol-Cava A.M, Boury-Esnault N, Klautau M, Russo CAM (2001) Cryptic speciation in a high gene flow scenario in the oviparous marine sponge Chondrosia reniformis. Mar Biol 139: 421-429 Loukaci A, Muricy G, Brouard JP, Guyot M, Vacelet J, BouryEsnault N (2004) Chemical divergence between two sibling species of Oscarella (Porifera) from the Mediterranean Sea. Biochem System Ecol 32: 893-899 Misevic GN, Guerardel Y, Sumanovski LT, Slomianny MC, Demarty M, Ripoll C, Karamanos Y, Mes E, Popescu O, Strecker G. (2004) Molecular recognition between glyconectins as an adhesion self-assembly pathway to multicellularity. J Biol Chem 279: 15579-15590 Muricy G, Sol-Cava AM, Thorpe JP, Boury-Esnault N (1996) Genetic evidence for the extensive cryptic speciation in the subtidal sponge Plakina trilopha (Porifera: Demospongiae: Homoscleromorpha) from the Western Mediterranean. Mar Ecol Prog Ser 138: 181-187 Parrish CR, Jakobsen KB, Coombe DR, Bacic A (1991) Isolation and characterization of cell adhesions molecules from the marine sponge Ophlitaspongia tenius. Biochim Biophys Acta 1073: 56-64 Popescu O, Misevic GN (1997) Self-recognition by proteoglycans. Nature 386: 231-232 Saito H, Yamagata T, Suzuki S (1968) Enzymatic methods for the determination of small quantities of isomeric chondroitin sulfate. J Biol Chem 243: 1542-1563 Sol-Cava AM (1994) Evolutionary genetics of marine sponges. In: van Soest RWM, van Kempen TMG, Braekman JC (eds). Sponges in time and space: biology, chemistry, paleontology. Balkema, Rotterdam. pp. 55-63 Sol-Cava AM, Klautau M, Boury-Esnault N, Borojevic R, Thorpe JP (1991). Genetic evidence for cryptic speciation in allopatric populations of two cosmopolitan species of the calcareous sponge genus Clathrina. Mar Biol 111: 381-186 Sol-Cava AM, Thorpe JP (1986) Genetic differentiation between morphotypes of the marine sponge Suberites ficus (Demospongiae: Hadromerida). Mar Biol 93: 247-253 Spillmann D, Burger MM (1996) Carbohydrate-carbohydrate interactions in adhesion. J Cell Biochem 61: 562-568

659

Usher KM, Sutton DC, Toze S, Kuo J, Fromont J (2004) Biogeography and phylogeny of Chondrilla species (Demospongiae) in Australia. Mar Ecol Prog Ser 270: 117-127 Vieira RP, Mulloy B, Mouro PAS (1991) Structure of a fucosebranched chondroitin sulfate from sea cucumber. Evidence for the presence of 3-O-sulfo-beta-D-glucuronosyl residues. J Biol Chem 266: 13530-13536 Vilela-Silva ACES, Alves AP, Valente AP, Vacquier VD, Mouro PAS (1999) Structure of the sulfated alpha-1-fucan from the egg jelly coat of the sea urchin Strongylocentrotus franciscanus: patterns of preferential 2-O and 4-O-sulfation determine sperm cell recognition. Glycobiology 9: 927-933 Vilela-Silva ACES, Castro MO, Valente AP, Biermann CH, Mouro PAS (2002) Sulfated fucans from the egg jellies of the closely related sea urchins Strongylocentrotus droebachiensis and Strongylocentrotus pallidus ensure species-specific fertilization. J Biol Chem 277: 379-387

Wrheide G, Fromont J, Sol-Cava AM (2004). Population genetics and phylogeography of sponges - a workshop synthesis. In: Pansini M, Pronzato R, Bavestrello G, Manconi R (eds). Sponge science in the new millennium. Boll Mus Ist Biol Univ Genova 68: 683-688 Zierer MS, Mouro PA (2000) A wide diversity of sulfated polysaccharides are synthesized by different species of marine sponges. Carbohydr Res 328: 209-216 Zilberberg C (2006) Reproduo clonal e sistemtica molecular de esponjas (Porifera). PhD Thesis, Universidade Federal do Rio de Janeiro, Rio de Janeiro Zilberberg C, Sol-Cava AM, Klautau M (2006) The extent of asexual reproduction in sponges of the genus Chondrilla (Demospongiae: Chondrosida) from the Caribbean and the Brazilian coasts. J Exp Mar Biol Ecol 336: 221-220

661

Authors Index
A
Albano, Rodolpho M. ...................................................... 555 Alcolado, Pedro M. .............................................................. 3 Alivon, Eliane .................................................................. 383 Almeida, Marise....................................................... 427, 491 Andra, Brbara R. .......................................................... 131 Arillo, Attilio .................................................................... 639 Assumpo, Leonardo L.M.............................................. 555 Austin, William C. ........................................................... 139 vila, Enrique .................................................................. 147 Azzini, Francesca ..................................................... 157, 203 Correia, Monica Dorigo ................................................... 233 Cosme, Bruno .......................................................... 275, 593 Costa-Lotufo, Letcia V. ................................................... 313 Coutinho, Cristiano C. ..................................................... 281 Cristobo, Javier ................................................................ 525 Cruz-Barraza, Jos Antonio ............................................. 147 Custdio, Mrcio R. ......................................................... 653

D
de Voogd, Nicole J. .......................................................... 173 Diaz, Maria Cristina ................................................... 31, 621 Dillard, Sandra L.............................................................. 621 Duckworth, Alan R. ......................................................... 297

B
Ballet, Pascal ...................................................................... 79 Barrie, J. Vaughn .............................................................. 139 Batista, Daniela ................................................................ 131 Bavestrello, Giorgio ......................... 157, 203, 239, 503, 639 Bayer, Kristina ................................................................. 165 Becking, Leontine E......................................................... 173 Beglinger, Elly J. .............................................................. 645 Belikov, Sergey ................................................................ 383 Belikov, Sergey I. ............................................................. 179 Beresi, Matilde Sylvia.........................................................11 Bertolino, Marco .............................................................. 189 Bzac, Chantal ................................................................... 23 Bonasoro, Francesco ........................................................ 503 Borchiellini, Carole .......................................................... 383 Boury-Esnault, Nicole................................................ 23, 383 Brandt, David ................................................................... 581 Bremec, Claudia............................................................... 189 Brmmer, Franz ....................................................... 361, 373 Buckeridge, John.............................................................. 393 Buell, Nicole .................................................................... 419

E
Eckert, Rafael ................................................................... 477 Efremova, Sofia................................................................ 383 Efremova, Sofia M. .......................................................... 483 Ehrlich, Hermann ............................................................. 303 Ellwood, Michael ............................................................. 393 Ereskovsky, Alexander V. .......................................... 41, 327 Erpenbeck, Dirk ............................................................... 123 Erwin, Patrick M. ............................................................. 621 Esteves, Ana ..................................................................... 491 Evans-Illidge, Elizabeth ................................................... 297

F
Fassini, Dario ................................................................... 503 Fernandez, Jlio ............................................................... 509 Ferreira, Elthon G. ........................................................... 313 Freeman, Christopher J. ................................................... 319 Fusetani, Nobuhiro........................................................... 173

C
Cares, Simone ................................................................. 509 Calcinai, Barbara...................................... 157, 189, 203, 239 Calderon, Emiliano Nicolas ..............................................211 Camillo, Cristina Gioia Di ............................................... 239 Campos, Maurcio .................................................... 219, 477 Carballo, Jos Lus ........................................................... 147 Carnevali, Daniela Candia ............................................... 503 Carraro, Joo Lus .................................................... 219, 477 Cedro, Victor Ribeiro ....................................................... 233 Cerrano, Carlo .................................. 157, 189, 203, 239, 639 Chaves-Fonnegra, Andia .................................................. 247 Chiappone, Mark.............................................................. 255 Conway, Kim W. .............................................................. 139 Cook, Steve de C.............................................................. 265

G
Gaggero, Laura ................................................................ 203 Gerasimova, Elena I. ........................................................ 327 Giovine, Marco ................................................................ 639 Gleason, Daniel F. ............................................................ 319 Gochfeld, Deborah J. ....................................................... 335 Gomes, Dbora ................................................................ 555 Gonobobleva, Elisaveta L. ............................................... 345 Guerardel, Yann ................................................................. 79

H
Hajdu, Eduardo ................................ 233, 313, 353, 449, 509 Hammel, Jrg U. .............................................................. 373 Harvey, Alan W. ............................................................... 319 Heim, Isabel ............................................................. 361, 373

662

Hentschel, Ute .......................................................... 165, 561 Hill, April ......................................................................... 419 Hill, Malcolm ................................................................... 419 Hoffmann, Friederike ............................................... 379, 613 Humanes, Madalena................................................. 427, 491

I
Itskovich, Valeria ............................................................. 383

J
Jascone, Lia ...................................................................... 555 Jimenez, Paula C. ............................................................. 313

K
Kaluzhnaya, Oksana V. .................................................... 179 Karamanos, Yannis............................................................. 79 Kelly, Michelle ................................................................. 393 Kelve, Merike .................................................................. 405 Kimble, Steven J.A. ......................................................... 621 Klages, Michael ............................................................... 379 Klautau, Michelle............................................................. 413 Kochem, Michele ............................................................. 653 Kotterman, Michiel .......................................................... 497 Krasko, Anatoli ................................................................ 581 Krautter, Manfred............................................................. 139 Kumeiko, Vadim V. .......................................................... 483 Kuusksalu, Anne .............................................................. 405

McLean, Elizabeth L. ....................................................... 443 Menke, Christian .............................................................. 123 Menshenina, Larisa L....................................................... 449 Metsis, Madis ................................................................... 405 Miller, Steven L. .............................................................. 255 Misevic, Gradimir N. ......................................................... 79 Misevic, Nikola .................................................................. 79 Monteiro, Leandro C........................................................ 413 Moraes, Fernando ............................................................ 467 Moraes, Manoel O. de ...................................................... 313 Mothes, Beatriz ........................................................ 219, 477 Mouro, Paulo A.S. .......................................................... 653 Mukhina, Yulia I. ............................................................. 483 Mller, Isabel M. ........................................................ 89, 179 Mller, Werner E.G. ........................................... 89, 179, 581 Muricy, Guilherme ........................................... 131, 467, 547

N
Nakao, Yoichi ................................................................... 173 Nickel, Michael ........................................................ 361, 373 Nicolai, Marisa H. ............................................................ 491 Norris, Jonathan ................................................................. 79 Norris, Vic .......................................................................... 79

O
Oliveira, Mara V.............................................................. 509 Oliveira-Silva, Patricia..................................................... 439 Osinga, Ronald................................................................. 497

L
Lamaro, Flvia R.M. ...................................................... 555 Lanna, Emilio................................................................... 413 Lavaleye, Marc S.S. ......................................................... 645 Lee, Welton L. .................................................................. 517 Lemoine, Nathan .............................................................. 419 Lerner, Cla .............................................................. 219, 477 Leys, Sally P. ...................................................................... 53 Lindquist, Niels ................................................................ 561 Lbo-Hajdu, Gisele .......................................................... 555 Lopes, Daniela A. ..................................................... 353, 449

P
Paiva, Paulo Csar de........................................................211 Pansini, Maurizio ............................................................. 157 Parma, Lorenzo ................................................................ 503 Peixinho, Solange .................................................... 275, 509 Perez, Thierry ................................................................... 383 Pessoa, Cludia ................................................................ 313 Piantoni, Carla.................................................................... 31 Pierce, Melissa J............................................................... 621 Podgornaya, Olga I. ......................................................... 483 Popescu, Octavian .............................................................. 79 Portela, Tiago A. .............................................................. 313 Pronzato, Roberto .............................................................. 61

M
Maes, Emmanuel................................................................ 79 Maia, Guilherme de Azevedo .......................................... 281 Maier, Connie................................................................... 645 Maldonado, Manuel ......................................................... 477 Manconi, Renata ................................................................ 61 Marques, Daniela ............................................................. 427 Mrquez, Diana M. .......................................................... 433 Martnez, Alejandro ......................................................... 433 Masuda, Yoshiki ............................................................... 383 Matsunaga, Shigeki .......................................................... 173 Mazzoli-Dias, Mirna ........................................................ 439 McFall, Greg .................................................................... 319

R
Rapp, Hans Tore ............................................................... 613 Reintamm, Tnu............................................................... 405 Reiswig, Henry M. ........................................................... 517 Ribeiro, Suzi M. ............................................................... 439 Ros, Pilar......................................................................... 525 Ripoll, Camille ................................................................... 79 Robledo, Sara M. ............................................................. 433 Rodriguez, Pablo R.D. ..................................................... 547 Ry, Hans ......................................................................... 379

663

Rutten, Leanne M............................................................. 255 Rtzler, Klaus............................................................. 31, 621 Ruzicka, Rob .................................................................... 319

Tubbs, Lincoln ................................................................. 393

V
Vacelet, Jean ..................................................................... 107 Valado, Ana Luiza .......................................................... 555 Valisano, Laura ........................................................ 239, 639 van Duyl, Fleur C............................................................. 645 van Soest, Rob W.M......................................... 173, 319, 645 Veitenheimer-Mendes, Inga Ludmila............................... 219 Vieiralves, Thomz .......................................................... 555 Vilanova, Eduardo............................................................ 653 Volkmer-Ribeiro, Ceclia ..................................................117

S
Sachs, Oliver .................................................................... 379 Salgado, Adriana .............................................................. 555 Sampaio, Cludio L.S. ..................................................... 131 Sauter, Eberhard ............................................................... 379 Schejter, Laura ................................................................. 189 Schlder, Carmen ............................................................. 335 Schmitt, Susanne ...................................................... 165, 561 Schnberg, Christine H.L................................................. 569 Schrder, Heinz C. ................................................... 179, 581 Schupp, Peter J. ................................................................ 627 Silva, Carla M.M. da ........................................................ 593 Silva, Meiryelen V. da ...................................................... 593 Silveira, Edilberto R......................................................... 313 Sol-Cava, Antonio M. .................................................... 603 Sovierzosky, Hilda Helena ............................................... 233 Spetland, Frank ................................................................ 613 Strecker, Gerard ................................................................. 79 Sumanovski, Lazar T.......................................................... 79 Suwa, Ryota ..................................................................... 569 Swanson, Dione W. .......................................................... 255

W
Wehrl, Markus.................................................................. 561 Weisz, Jeremy B............................................................... 561 Wiens, Matthias ............................................................... 581 Wilke, Diego V. ................................................................ 313 Wilkie, Iain C. .................................................................. 503 Wolff, Carsten .................................................................. 297 Worch, Hartmut................................................................ 303 Wrheide, Gert ......................................................... 123, 603

X
Xavier, Joana .................................................................... 427

T
Tabachnick, Konstantin R. ....................................... 449, 645 Tahir, Muhammad Nawaz ................................................ 581 Tendal, Ole Secher ........................................................... 613 Thacker, Robert W. ............................................ 31, 335, 621 Thoms, Carsten ................................................................ 627 Tremel, Wolfgang............................................................. 581

Y
Yoshioka, Paul M. ............................................................ 443

Z
Zea, Sven.......................................................................... 247 Zilberberg, Carla .......................................................211, 653

665

Subject Index
#
12S 361, 362, 364, 365, 367, 369 16S 165, 166, 168, 361, 362, 364, 365, 367, 369, 431, 561-566, 621-624 18S 90, 91, 181, 361, 383-386, 388, 389, 556 3-alkylpyridine 173-175, 177 Alignment 95, 166, 182, 183, 364, 365, 367, 374-376, 384, 385, 406, 407, 409, 582, 583, 585, 622 Allelochemical 247, 252, 253, 443, 446, 447 Allografts 96, 245 Alloimmune response 95 Alphaproteobacteria 562, 566 Ammonia 165-168, 337, 340 Ammonia-oxidizing bacteria 165, 168 Amphiblastula 41, 48, 49, 100, 285, 413, 414, 417, 419 Amphidiscosida 357 Amphimedon 4, 33, 42, 159, 160, 173-176, 233, 234, 236, 313317, 337, 338, 341, 444-446, 556, 557 Amphimedon aff. complanata 233, 236 Amphimedon compressa 176, 313-317, 337, 338, 341, 444, 445, 446 Amphimedon erina 176, 445, 446 Amphimedon paraviridis 176 Amphimedon queenslandica 42, 45, 47, 173-175 Amphimedon viridis 4, 176, 233, 234, 236, 341, 444, 556, 557 Amphitoxin 173-177 Anaderma 395, 397 Anaderma rancureli 397 Anatomy 23, 26, 27, 284, 629 Ancestor 41, 44, 53, 55, 89, 90, 91, 94, 95, 102, 114, 179, 180, 181, 185, 281, 282, 286-288, 384, 385, 390, 583, 625 Anchialine lakes 157, 162 Ancorinidae 32, 275, 356, 386 Anheteromeyenia 63, 70, 118, 119 Antarctica 14, 199, 219, 220-224, 226, 227, 229, 230, 400, 525, 535, 547, 552 Antero-posterior axis 50, 89, 100, 283, 289, 291 Anthosigmella varians 234 Anthropogenic 256, 259, 261, 335, 340-342, 427, 431 Antibodies 82, 83, 483-485, 488, 585, 587, 588, 589, 630 Antileishmanial activity 433-436 Antimitotic activity 313, 316 Antiproliferative activity 313, 315, 316 Aphanocapsa feldmani 621 Aphrocallistes 94, 139, 140-144, 181, 183, 304, 306, 354, 357, 358, 583, 584 Aphrocallistes beatrix 354, 357, 358 Aphrocallistes vastus 94, 139, 140-144, 181, 183, 304, 306, 583, 584 Aphrocallistidae 357 Apical-basal axis 48, 50, 102, 103 Aplysilla 33, 99, 323, 324, 384, 386, 388, 629 Aplysilla glacialis 629 Aplysilla longispina 323, 324 Aplysilla sulfurea 99, 384, 386, 388 Aplysillidae 33, 384, 388 Aplysina 3-5, 33, 34, 56, 159, 165, 166, 168, 273, 303, 314-316, 323, 324, 335-338, 340, 341, 354, 357, 358, 361, 362, 364, 365, 367, 368, 369, 384, 385, 387, 389, 419, 444, 493, 556, 561, 564, 621, 622, 624, 625, 631, 633, 654 Aplysina aerophoba 33, 165-168, 361, 362, 364, 365, 367-369, 419, 564, Aplysina archeri 33, 354, 357, 358, 362, 364, 365, 367, 368 Aplysina aurea 273 Aplysina cauliformis 3-5, 34, 336, 338, 340, 341, 354, 357, 358, 362, 364, 365, 367, 368 Aplysina cavernicola 361, 362, 364, 365, 367-369, 419, 564

A
Aaptos 159, 160, 162, 354, 356, 358, 617 Aaptos aaptos 617 Aaptos pernucleata 159, 160, 162 Abyssocladia 107-110, 112, 114 Abyssocladia agglutinans 108 Abyssocladia dominalba 112 Abyssocladia huitzilopochtli 108 Abyssocladia naudur 110 Acalle 66, 118 Acanthanchora 133 Acanthella 159, 160, 497, 498, 500, 640 Acanthella acuta 497, 498, 500, 640 Acanthella hispida 159, 160 Acanthotetilla 509-515, 547, 552 Acanthotetilla gorgonosclera 515 Acanthotetilla rocasensis sp. nov. 509-512, 547, 552 Acanthotetilla walteri sp. nov. 509, 510, 512-514, 547, 552 Acarnidae 220, 394 Acetylcholinesterase 427, 428, 430 Aciculites 394, 396, 397, 398 Aciculites manawatawhi 396-398 Aciculites orientalis 397 Aciculites oxytylota 397 Aciculites papillata 397 Aciculites pulchra 394, 396-398 Aciculites sulcus 396-398 Acidobacteria 165, 561, 563, 566 Actin 45, 56, 57, 286, 316, 408, 410, 489 Activated defense 627, 630-633 Africa 277, 393 Agelas 233, 234, 236, 274, 314-316, 354, 356, 358, 562-564, 630, 640 Agelas clathrodes 314-316, 354, 356, 358 Agelas dispar 233, 234, 236, 314-316, 354, 356, 358 Agelas oroides 640 Agelas schmidti 354, 356, 358 Agelas wiedenmayeri 562-565, 567 Agelasida 236, 314, 356, 386, 388, 563 Agelasidae 314, 356 Aggregation factor 45, 80, 90, 91, 102, 653 Aiolochroia 314-316, 323, 324, 338, 354, 357, 358, 364, 367 Aiolochroia crassa 314-316, 323, 324, 338, 354, 357, 358, 364, 367 Aka 159, 160, 252, 253, 255, 562, 563 Aka coralliphaga 253, 562, 563 Aka mucosa 159, 160 Alectona 354, 356, 358 Alectona mesatlantica 354, 356, 358 Alectonidae 32, 356

666

Aplysina fistularis 4, 34, 314-316, 341, 362, 364, 365, 367, 368, 384, 385, 387, 389 Aplysina fulva 556, 557, 621, 622, 624, 625, 640, 654 Aplysina gerardogreeni 34, 354, 357, 358 Aplysina insularis 362, 364, 365, 367, 368 Aplysina lactuca 314, 316 34, 338, 354, 357, 358 Aplysina lacunosa Aplysina muricyana 314-316 Aplysina solangeae 314-316 Aplysinidae 33, 34, 38, 273, 274, 314-317, 385, 387, 389 Aplysinopsis 269, 271, 272 Apoptosis 93, 97, 102, 408, 639 Aquaculture 157, 297, 301, 498, 613 Aquarium 419, 422, 484, 485, 497, 498, 501, 503, 518, 570, 640 Aquiferous system 54-56, 99, 102, 107-110, 112-114, 162, 165, 239, 240, 242, 245, 283, 285, 286, 290, 416, 497, 499, 514 Archaea 170, 379, 381, 561 Archaeocyte 27, 29, 95-99, 165, 287, 290, 291, 639 Arctic 51, 327, 332, 333, 379, 380, 651 Argentina 11-14, 19-21, 63-66, 118, 189, 190, 195, 199, 223, 227 133, 219, 220, 222, 444 Artemisina 219, 220, 222 Artemisina apollinis Artemisina melana 444 Asbestopluma 107-110, 112, 113, 517, 522 Asbestopluma agglutinans 110 107-109, 112, 113, 517, 522 Asbestopluma hypogea Asbestopluma occidentalis 109 Asbestopluma stylivarians 112 Asexual reproduction 44, 203, 283, 417, 503, 613 Associations 14, 17, 19, 31, 37, 147, 148, 151, 152, 153, 154, 419, 439, 443-446, 604, 621 Astrophorida 32, 236, 275, 314, 356, 385, 386, 388, 389, 393, 467, 471, 477, 478, 593, 594, 613, 617, 619 362, 367, 386, 562 Astrosclera 362, 367, 386, 562 Astrosclera willeyana Atlantic Ocean 107, 190, 361, 431, 449, 464, 552, 654 Aulospongus 321, 323, 324, 354, 356, 358 321, 323, 324 Aulospongus pearsi 321, 323, 324 Aulospongus samariensis Australia 32, 63-66, 72, 73, 118, 123, 124, 153, 162, 175, 176, 270, 297, 298, 301, 569, 570, 572, 608, 654 Autografts 96 Awhiowhio 393-398 Awhiowhio osheai 395-398 Awhiowhio sepulchrum 395-398 Awhiowhio unda 396-398 Axial filament 27, 181-184, 303, 305, 581-583, 585, 587, 588, 590, 642 Axinella 95, 242, 321-324, 354, 356, 358, 386, 603, 640 Axinella bookhouti 321, 323, 324 386 Axinella corrugata 386, 640 Axinella damicornis Axinella polypoides 386, 640 Axinella pomponiae 322-324 Axinella verrucosa 95, 640 Axinella waltonsmithi 321-324 Axinellidae 32, 314, 356, 386 Axinyssa 32, 321-324, 493, 640 321-324 Axinyssa ambrosia Axinyssa aplysinoides 32 493, 640 Axinyssa aurantiaca 367, 386 Axos 367, 386 Axos cliftoni Azores 451, 473, 645, 648-650 Azoriciidae 396

B
Bacteria 5, 56, 62, 95, 96, 109, 147, 162, 165, 168, 170, 177, 310, 337, 341, 342, 379, 381, 420-422, 491, 497, 561, 562, 614, 617, 619, 621, 629, 630 Bahamas 32, 38, 135, 176, 278, 336, 341, 362, 364, 622, 628, 654, 655 Baikalospongia 65, 180-182, 384, 385, 387-389 Baikalospongia bacillifera 65, 180-182, 384, 385, 387 Baikalospongia fungiformis 180, 384, 387 Baikalospongia intermedia 65, 181, 182, 384, 385, 387 Baikalospongia recta 181, 182, 387 Balliviaspongia 67 Balliviaspongia wirrmanni 67 Barcoding 123-127, 603-610 Basal apparatus 345, 346, 348-351, 489 Basal lamina 56, 285-287, 291 Bath sponge 297, 298, 300, 301, 309 Bathymetric distribution 61, 219, 230, 357, 540, 545, 549, 593, 600 Batzella 32, 174, 176, 640 Batzella inops 640 Batzella melanos 32 Bauplan 61, 89, 90, 180, 284 Behaviour 42, 44, 206, 239, 244, 304, 483, 485, 489, 503, 505, 569, 578 Belize 32, 38, 64, 136, 240, 622, 625 Benthic community 189, 419, 327, 548 Bergquistia 270 Betaproteobacteria 165, 166, 168 Biemna 159, 160, 162, 233, 234, 236, 240 Biemna fortis 240 Biemna megalosigma 159, 160, 162 Biemna microacanthosigma 233, 234, 236 Bilateral symmetry 517, 523 Bioactive compounds 152, 154, 173, 174, 303, 310, 313, 433, 589, 627, 639 Biocomposites 303, 304, 308, 309 Biodiversity 23, 61, 62, 108, 109, 117-119, 123-125, 158, 199, 313, 335, 339, 357, 384, 390, 443, 446, 447, 509, 606-610 Bioerosion 131, 252, 255, 259, 261, 262, 569, 578 Biogeography 29, 61-63, 162, 319, 367, 467 Bioindicator 3, 5, 335, 341, 342 Biomarker 341, 379, 427, 430, 431 Biosilicification 303, 589 Biostratigraphy 11 Blastomeres 98, 282, 284, 331, 345, 346, 417 Bleaching 260, 262, 420, 569, 570, 578 Body plan 15, 44, 53, 56, 89, 90, 93, 100, 101, 281-283, 483 Boring pattern 203, 204, 206, 207 Boring sponge 4, 153, 154, 157, 160, 203, 207, 247, 256, 262 Brazil 4, 62-67, 90, 118, 119, 131-136, 176, 211, 212, 229, 233, 234, 236, 275, 277, 278, 313, 314, 316, 317, 340, 341, 353, 354, 356-358, 413, 415, 417, 439, 440, 449, 457, 460, 465, 467, 468, 471, 473, 477, 478, 481, 509, 510, 512, 515, 547-549, 551, 552, 556, 593-595, 600, 601, 654-656 Bromoperoxidase 491-493, 495 Bubaridae 356 Bubaris 159, 354, 356, 358

C
Cacospongia 233, 265, 266, 272, 274, 354, 357, 358, 628-630 Cacospongia levis 233, 354-358 Cacospongia mollior 265, 266

667

Cacospongia scalaris 265 Calcarea 11-20, 41, 42, 44, 46, 48-50, 89, 90, 93-95, 124, 303, 308, 309, 349, 356, 357, 413, 417, 494, 556, 557, 581 Calcaronea 41, 42, 44, 46, 48-50, 284, 413, 416, 417 Calciblastula 48, 49, 284 Calcinea 41, 48-50, 284, 413, 417 Calcium carbonate 90, 94, 179, 203, 247, 255, 262, 303, 309, 570, 581 Callipelta 394, 396-398 Callipelta punctata 396-398 32, 95, 159, 160, 173-176, 314-316, 321, 323, Callyspongia 324, 385, 386, 444-446, 563 Callyspongia fallax 321, 323, 324 Callyspongia plicifera 444 Callyspongia vaginalis 314-316, 385, 386, 444-446, 563, 565, 567 Callyspongiidae 31, 32, 173-176, 314, 384, 386 Calmodulin 94 Calthropellidae 366 Calycosoma 449-451, 463-465 Calycosoma validum 449-451, 464, 465 Calyx 33, 383, 385, 387 Calyx podatypa 33, 385, 387 Cambrian 11-14, 16, 20, 94, 190 Canada 63, 64, 66, 118, 374, 460 Candidaspongia 273 Carbohydrates 79, 80, 82, 83 Caribbean 5, 7, 31, 36-38, 73, 131, 134-136, 147, 152, 153, 166, 176, 212, 234, 240, 247, 248, 255, 256, 261, 262, 278, 297, 316, 319, 320, 323, 335, 336, 339, 340, 361, 367, 384, 434, 443, 446, 465, 467, 473, 497, 509, 514, 515, 561, 562, 567, 569, 593, 600, 601, 621, 622, 624, 625, 629, 630 Carnivorous feeding 108, 109, 112, 113 Carnivorous sponge 107-110, 112, 114, 517 Carteriospongia 33, 266, 273 Carteriospongia foliascens 33 Cathepsin 103, 179, 181-186, 309, 581-584 cDNA 96-99, 101, 181-183, 185, 405-411, 582-585 Cell adhesion 42, 79, 82-87, 90, 91, 93, 102 Cell culture 56, 89, 94, 97, 100, 310, 434, 581, 639 Cell differentiation 102, 281, 284, 287, 291, 345, 349 Cell recognition 79, 80, 82, 84-86, 91, 92, 653 Cell-cell interactions 45, 47, 247 Ceractinomorpha 41, 411, 603 Chalinidae 31, 33, 34, 36, 133, 147, 153, 174-176, 225, 357, 384386 147, 148, 151, 152, 233, 236, 321, 323, 324, 493 Chalinula Chalinula molitba 233, 236, 321, 323, 324, 493 Chalinula nematifera 147-149, 151-154 Characella 354, 356, 358, 477, 478, 480-482 354, 356, 358, 477, 480, 481 Characella aspera Characella capitolii sp. nov. 477, 478, 480, 481 Characella pachastrelloides 482 Characella sollasi 482 Checklist 63, 157, 353, 396, 397 Chemical defense 627-631, 633 Chemical ecology 627, 633, 634 Chemotaxonomy 173, 177 Chile 118, 190, 195, 199, 222, 223, 481 Chitin 303-311 Chlorophyll 341, 570, 571, 573, 575-578, 621-625 Choanocyte 26, 41, 89, 97, 107, 165, 282, 283, 285, 286, 416, 483-485, 488, 489, 563, 631 Choanocyte chamber 26, 38, 56, 67, 89, 100, 102, 103, 107, 108, 112, 265, 269, 270, 271, 273, 283, 285, 414, 483, 485, 489, 517, 522, 617

Chondrilla 4, 32, 133, 159, 233, 234, 236, 323, 324, 335, 337, 338, 340, 341, 444, 445, 503, 561, 605, 606, 628, 653-658 Chondrilla australiensis 32, 34, 159, 160, 562, 653-656 Chondrilla nucula 4, 32, 133, 233, 234, 236, 323, 324, 335, 337, 338, 340, 341, 444, 445, 503, 605, 628, 653-657 Chondrocladia 107-110, 112, 114, 517 Chondrocladia gigantea 108, 112 Chondrocladia lampadiglobus 109 Chondrosia 44, 46, 49, 165, 166, 233, 234, 236, 239, 242, 244, 245, 310, 322-324, 367, 405, 407, 409, 411, 497-500, 503, 562 Chondrosia collectrix 243, 244, 246, 322-324 44, 46, 49, 165, 166, 168, 242, 244, 310, Chondrosia reniformis 323, 324, 405-411, 497-500, 503, 504, 506, 507 Cinachyra 354, 356, 358, 444, 547-552, 619 Cinachyra barbata 547, 549, 550, 552 Cinachyra crustata 547, 549, 550, 552 Cinachyra helena sp. nov. 547-552 Cinachyra novae-zealandiae 549 Cinachyra rhizophyta 547, 551, 552 Cinachyra tarentina 629 Cinachyra uteoides 547, 549, 550, 552 Cinachyrella 34, 159, 160, 233, 234, 236, 322-324, 338, 354, 356-358, 387, 547, 549, 551, 552 Cinachyrella alloclada 233, 234, 236, 322-324, 338, 354, 356, 358, 547, 552 Cinachyrella apion 233, 234, 236, 354, 356, 358, 387, 547, 551, 552 Cinachyrella australiensis 34, 159, 160 354, 356-358, 547, 552 Cinachyrella kuekenthali Cinctoblastula 47-49, 285 Ciocalypta 321, 323, 324, 493 Ciocalypta gibbsi 321, 323, 324 493 Ciocalypta penicillus Circumtropical 68, 71 Citronia 273 Cladocroce 159, 160 517, 518, 521-523 Cladorhiza Cladorhiza abyssicola 108, 522, 523 Cladorhiza arctica 522, 523 Cladorhiza bathycrinoides 522, 523 Cladorhiza corona 522, 523 Cladorhiza corticocancellata 522, 523 Cladorhiza depressa 522, 523 Cladorhiza ephyrula 522, 523 522, 523 Cladorhiza flosabyssi Cladorhiza fristedti 522, 523 Cladorhiza gelida 522, 523 Cladorhiza grimaldi 522, 523 522, 523 Cladorhiza inversa Cladorhiza linearis 522, 523 Cladorhiza longipinna 522, 523 Cladorhiza mani 522, 523 Cladorhiza methanophila 109, 112, 522 Cladorhiza microchela 522 Cladorhiza minuta 522, 523 Cladorhiza mirabile 522 522, 523 Cladorhiza moruliformis 522 Cladorhiza nematomorpha Cladorhiza oxeata 522 Cladorhiza pteron sp. nov. 518, 521-523 Cladorhiza rectangularis 522 Cladorhiza schistochela 522 110, 522, 523 Cladorhiza segonzaci Cladorhiza septemdentalis 522, 523 Cladorhiza similis 522 Cladorhiza tenuisigma 522

668

Cladorhiza thomsoni 522 Cladorhiza tridentata 522, 523 Cladorhizidae 107-110, 112, 517, 518 Clathria 3-5, 32, 133, 159, 160, 321, 323, 324, 354, 356, 358 Clathria (Clathria) carteri 321, 323, 324 Clathria (Clathria) prolifera 5, 321, 323, 324 323, 324 Clathria (Thalysias) schoenus Clathria venosa 3-5 34, 321-324, 413, 415, 416, 493, 494 Clathrina Clathrina canariensis 321, 323, 324 Clathrina cerebrum 413, 415, 417, 493 Clathrina contorta 493 Clathrina coriacea 322-324, 413, 415, 417 Clathrinida 32, 34 Clathrinidae 34 Cleavage 41, 48-50, 56, 102, 186, 282-285, 290, 314, 316, 317, 329, 330, 345, 346, 349, 558, 583, 584 Cliona 3-5, 7, 20, 79, 80, 85, 159, 160, 162, 203-207, 210, 233, 234, 236, 239-241, 244, 247-253, 255-259, 261, 321, 323, 324, 338, 354, 356, 358, 427, 444, 493-495, 569-576, 578, 640, 654 203-207 Cliona albimarginata Cliona ameghinoi 20 7, 338 Cliona aprica Cliona aurivilli 159, 160 4, 5, 323, 324, 444 Cliona caribbaea Cliona celata 79, 80, 84-87, 159, 160, 162, 233, 234, 236, 321, 323, 324, 354, 356, 358, 427-431, 493, 495, 940, 654 3, 4, 247-253, 255-262, 338, 444 Cliona delitrix 20 Cliona entrerriana Cliona lampa 210, 261 Cliona nigricans 240, 241, 244 Cliona orientalis 159, 160, 210, 569, 570-578 Cliona varians 3, 4, 233, 234, 236, 444, 569 Cliona vermifera 253 Cliona vesparia 3, 4 427, 493, 494 Cliona viridis Clionaidae 32, 203, 247, 356, 386, 427 Cliothosa hancocki 159, 160 Cloning 90, 91, 166, 364, 604 Coelocarteria singaporense 32 Coelosphaera 321, 323, 324 Coelosphaeridae 133, 356 COI 361, 362, 364, 365, 367-369, 373-375, 607, 609 Coliform 337, 338, 340-342 Collagen 25, 27, 29, 56, 92, 99, 102, 135, 239, 244, 245, 265, 269, 270-272, 285, 286, 303, 304, 306, 309, 310, 405, 407-410, 422, 483, 488, 500, 507, 584, 585, 587-589, 598, 599, 615 Collospongia 271, 272 Colombia 4, 5, 247, 248, 340, 341, 433, 434 Colonization 61, 62, 119, 247, 255-262, 440, 625 Commensalism 62, 131, 147, 439, 440 Community structure 3, 4, 256, 335-337, 341, 342, 419 Competition 3, 131, 151-154, 162, 211, 212, 215, 216, 233, 240, 247, 322, 585 Confocal microscopy 486, 488 Conservation 62, 97, 132, 158, 184, 286, 289, 290, 291, 373, 410, 467 Continental shelf 189, 199, 255, 316, 357, 393, 394, 440, 473, 477, 480, 509, 548, 593 Coral reef 3-5, 7, 8, 31, 123, 131, 132, 136, 152, 153, 203, 204, 233, 234, 236, 237, 247, 255-262, 335-337, 340, 342, 443, 446, 447, 569, 570, 578, 608, 627, 629, 645 Corallistes 354, 356, 358, 386, 394, 395, 397, Corallistes australis 397 Corallistes multituberculatus 397 354, 356, 358 Corallistes typus

Corallistes undulatus 397 Corallistidae 356, 386, 395-397 Cortispongilla 65 Cortispongilla barroisi 65 Corvoheteromeyenia 63, 118, 119 Corvoheteromeyenia australis 63, 119 Corvoheteromeyenia heterosclera 63, 119 Corvomeyenia 66, 118, 385, 387, 388 66 Corvomeyenia carolinensis Corvomeyenia epilithosa 66 Corvomeyenia everetti 66 Corvomeyenia thumi 66 Corvospongilla 61, 63, 72, 74, 119 Corvospongilla becki 63 Corvospongilla bhavnagarensis 63 Corvospongilla boehmii 63 Corvospongilla burmanica 63 Corvospongilla caunteri 63 Corvospongilla lapidosa 63 Corvospongilla loricata 63 Corvospongilla mesopotamica 63, 64 Corvospongilla micramphidiscoides 63 Corvospongilla novaeterrae 63 Corvospongilla scabrispiculis 63 Corvospongilla seckti 63 Corvospongilla sodenia 63 Corvospongilla thysi 63 Corvospongilla ultima 63 Corvospongilla victoriae 63 Corvospongilla volkmeri 63, 119 Corvospongilla zambesiana 63 33, 269-271, 297-301, 323, 324 Coscinoderma 323, 324 Coscinoderma lanuga Cosmopolitan 14, 61, 63, 68, 72, 123, 147, 162, 179, 180, 181, 183, 384, 427, 430, 584, 605, 609, 653, 657 Costifer wilsoni 396, 397, 398 COXI 383-386, 388-390 Crambe 32, 56, 317, 362, 367, 394, 493, 628, 629 56, 317, 362, 367, 628, 629 Crambe crambe 99, 190, 191, 354, 356, 358, 547, 549, 550, 552 Craniella Craniella carteri 547, 552 Craniella corticata 547, 552 Craniella cranium 547 Craniella leptoderma 190, 191 Craniella novae-zealandiae 549-551 Craniella quirimure 547, 552 Craniella schmidtii 99 354, 356, 358 Crella (Yvesia) Crella elegans 387 Crellidae 356, 387 Crellomyxilla chilensis 354, 356, 358 Cretaceous 11, 19, 62, 118 Cribrochalina 33 Cribrochalina dura 33 Cribrochalina vasculum 33 Croatia 63, 361, 362, 364, 367, 368, 374, 498, 501 Crude extract 492, 493, 630, 654, 655 Cryptic species 158, 320, 603, 609, 653-658 Cryptobiosis 61, 62, 67, 68 Cuba 3-8, 63, 278, 340, 341 Cultivation 405, 407, 484, 497, 498, 500, 501, 613 Cyamon 133 Cyanobacteria 147, 154, 165, 336, 561-564, 621, 622, 624, 625 Cymbastela 32, 341, 419, 561 Cymbastela concentrica 341, 419, 561

669

Cytochrome oxidase 124, 180, 361, 362, 373-375, 383, 411, 603, 604 Cytology 23, 26, 28, 29, 284, 483 Cytoplasm 26-29, 144, 288, 329, 345, 346, 348, 408, 414, 416, 422, 483, 485, 486, 488, 489, 615, 619 Cytoplasmic streaming 144 Cytoskeleton 45, 50, 56, 57, 101, 286, 305, 410, 484 Cytotoxicity 313, 314, 316, 317, 434, 436

D
Dactylocalycidae 357 Dactylocalyx 354, 357, 358 Dactylocalyx pumiceus 354, 357, 358 Dactylospongia 33, 272 Dactylospongia elegans 33 Darwinella 33, 211-216 Darwinellidae 33, 386 Deep sea 107-109, 112-114, 124, 144, 240, 305, 353, 354, 357379, 393, 399, 439, 619 Deltaproteobacteria 165, 561, 563, 566 Demospongiae 16, 19, 23, 29, 34, 41, 42, 44, 46, 48, 50, 61, 89, 90, 93-96, 124, 147, 154, 163, 165, 179, 181, 190, 219, 233, 247, 275, 282-284, 286, 303, 327, 345, 346, 348, 351, 356-358, 379, 383, 385, 393, 394, 396-398, 405, 433, 467, 471, 477, 478, 484, 486, 494, 509, 510, 518, 526, 547, 548, 581, 593, 639, 653, 654 Dendroceratida 32, 44, 46, 211, 384, 386, 388, 444 Depth range 27, 153, 321, 328, 333, 353, 357, 398, 400 Desmacella 354, 356, 358 Desmacella aff. pumilio 354, 356, 358 Desmacella annexa 354, 356, 358 Desmacellidae 32, 356 Desmanthidae 314, 356 Desmapsamma anchorata 133, 278, 323, 324, 443-447 Desmoxyidae 356 Developmental biology 41, 42, 44, 281 Developmental genes 41, 43, 47, 281 DGGE 419, 421, 422, 561-563, 565-567 Diacarnus spinipoculum 387 Diagoniella 11-14, 16 Dictyoceratida 32, 38, 44, 46, 47, 49, 236, 265, 269, 270-273, 297, 314, 357, 433, 563, 621 Dictyonella 32, 314-316, 386 Dictyonella funicularis 32 Dictyonella incisa 386 Dictyonellidae 32, 314, 386 Didiscus 386 Diel rhythm 569, 570, 572, 575, 577, 578 Digestion 80, 96, 305, 306, 308, 522, 562, 628, 630, 654 Dinoflagellates 147, 181, 569, 570, 571, 631 Diploblastic 55, 100, 124 Discodermia 32, 394, 396-399 Discodermia dissoluta 32 Discodermia proliferans 396-399 Discodermia sinuosa 394 Disease 119, 256, 260, 262, 297, 335, 336, 341, 342 Dispersion 503 Disphaerula 47-49, 282, 283, 346 Disturbance 6, 153, 255, 335 DNA barcoding 123, 124, 603, 604, 607-610 DNA degradation 555, 556 DNA extraction 166, 362, 374, 375, 384, 555-558, 604, 622 DNA sequencing 124, 374, 406, 409, 603, 622, 656, 657 DNA taxonomy 123, 609 Dosilia 63

Dosilia brouni 63 Dosilia palmeri 63 Dosilia plumosa 63 Dosilia pydanieli 63 Dosilia radiospiculata 63 Dragmacidon 233, 234, 236, 314, 316, 323, 324, 338, 354, 356, 358 Dragmacidon reticulatum 233, 234, 236, 314, 316, 323, 324, 338 Dredging 108, 139, 353, 473, 548, 593 Drulia 66, 70, 119 Drulia browni 66, 70, 119 Drulia conifera 66 Drulia cristata 66 Drulia ctenosclera 66 Drulia uruguayensis 66 Duosclera mackayi 63 Dysidea 160, 162, 165, 166, 233, 236, 242, 269, 273, 323, 324, 443-446, 493, 497, 499, 622, 624, 629, 630, 639, 640, 654 Dysidea avara 94, 165, 166, 168, 242, 497-499, 639, 640 Dysidea cinerea 158, 159, 160 Dysidea etheria 233, 236, 242, 269 Dysidea fragilis 159, 160, 162, 323, 324, 493 Dysidea granulosa 33, 624 Dysidea janiae 147, 443-446 Dysidea robusta 654

E
East Atlantic 369, 593, 600 Echinodictyum 159, 233, 234, 236 Echinodictyum asperum 159 Echinodictyum dendroides 233, 234, 236 Echinometra lucunter 211-215 Echinospongilla 66, 384, 385, 387-389 Echinospongilla brichardi 66, 384, 385, 387 Ecionemia 386 Ecology 6, 36, 61, 86, 107, 109, 123, 160, 278, 390, 439, 471, 503, 511, 514, 549, 599, 606, 607, 627, 633, 648 Ectyoplasia ferox 314, 315, 338, 444, 562, 563, 628, 629 EEZ (Economic Exclusive Zone) 353, 394 Electrophoresis 80, 81, 364, 374, 406, 419, 421, 422, 491, 556, 557, 561-563, 581, 654-658 Embryo 43-47, 97-99, 102, 112, 113, 144, 283, 288, 290, 314, 316, 317, 329-332, 345, 346, 348, 349, 414, 417, 517, 521, 522, 538, 561, 562, 564, 565, 567 Embryogenesis 43, 47, 50, 53, 56, 98-101, 103, 281-285, 327330, 332, 346, 348 Embryology 41, 42, 282, 291, 345, 417 Embryonic development 41-47, 50, 98, 345, 346, 348, 483, 521, 567 Endemic 17, 30, 61, 72, 73, 179, 180, 181, 182, 187, 234, 316, 321, 323, 324, 357, 384, 385, 394-398, 401, 467, 473, 584, 593, 600, 601 Endobiont 379 Environmental stress 157, 162, 316, 335, 341, 427, 569 Enzyme 82, 93, 96, 97, 103, 166, 179, 181, 183-185, 303, 308, 309, 405, 427, 428, 431, 491, 492, 494, 495, 497, 581-583, 585, 587-590, 630, 631 Ephydatia 54, 55, 61, 63, 68, 95, 97, 119, 179, 181, 384, 385, 387, 388, 414, 497, 500, 581, 584 Ephydatia cooperensis 384, 387 Ephydatia facunda 63 Ephydatia fluviatilis 61-63, 68, 97, 101, 179, 181, 183, 384, 387, 414, 497, 500, 581-584

670

Ephydatia fortis 63 Ephydatia japonica 63 Ephydatia meyeni 63 Ephydatia millsii 63 Ephydatia muelleri 54, 55, 57, 63, 95, 385, 387 Ephydatia ramsayi 63 Ephydatia robusta 63 Ephydatia syriaca 63 Epibiont 148, 150, 240 Epidioxysterol 433, 435, 436 Epithelium 45, 53, 56, 95, 99, 101, 102, 247, 281-287, 291, 345, 346, 349, 598 Eroding sponge 203 Erosion pattern 203, 205, 206, 207, 208 Erosion rate 203, 204, 207, 210, 261 Erylus 354, 356-358, 467, 468, 471-473, 491-493, 495, 617, 628, 629 Erylus alleni 467, 473 Erylus clavatus 473 Erylus corneus 467, 471-474 354, 356-358, 467, 472, 473 Erylus diminutus Erylus discophorus 491-495, 617 Erylus expletus 473 Erylus formosus 467, 471-474, 628, 629 Erylus granularis 473 Erylus latens sp. nov. 467, 468, 471-474 Erylus ministrongilus 473 Erylus pappilatus 473 354, 356, 358, 467, 472-474 Erylus soesti Erylus toxiformis 467, 472-474 Erylus transiens 467, 468, 472-474 Erylus trisphaera 473 Esperiopsidae 107, 109, 110, 114, 356 Esperiopsis 107, 109, 110, 112-114, 354, 356-358 354, 356-358 Esperiopsis bathyalis Esperiopsis desmophora 109, 112, 114 Esperiopsis symmetrica 109 Esperiopsis villosa 109, 110 Etching 203, 205, 206, 208, 247, 251, 252, 304, 306 Euchelipluma 107-110, 112-114 Euchelipluma arbuscula 109, 112 Euchelipluma elongata 109 Euchelipluma pristina 109, 110, 112 Eunapius 63, 71, 72, 385, 387, 388 Eunapius aetheriae 63 Eunapius ambiguus 63 Eunapius calcuttanus 63 Eunapius carteri 63, 71 Eunapius conifer 63 Eunapius crassissimus 63 Eunapius fragilis 63, 387 Eunapius geei 63 Eunapius geminus 63 Eunapius michaelseni 63 Eunapius nitens 63, 71 Eunapius potamolepis 63 Eunapius ryuensis 63 Eunapius sinensis 63 Eunapius subterraneus 63 Eunapius tinei 63 102, 303, 304, 306-308, 354, 357, 358 Euplectella Euplectella aspergillum 102, 303, 304, 306-308, 309 354, 357, 358 Euplectella suberea Euplectellidae 305, 357 Eurypon 159, 160, 387 Eurypon clavatum 387

Euryspongia 273 Evenness 335, 337, 339, 340 Evolution 41, 44, 48, 50, 53, 55, 56, 61, 79, 80, 82, 86, 87, 89, 90, 92, 94, 102, 107, 109, 110, 112, 113, 117, 123, 179, 180, 185, 281, 283, 286-289, 291, 308, 309, 369, 410, 439, 517, 607, 629 Excavation 204, 207, 208, 247, 248, 249, 250, 253 Excretion 165-168 Explants 297-301, 497, 498, 499 Extracellular matrix 45, 56, 82, 90, 92, 285, 288, 406, 410, 504, 507, 563-565 Extraction method 555, 557

F
Farming 297, 299, 300, 301 Farrea occa 139, 303, 304, 306-308 Fascaplysinopsis 272 Fasciospongia 272 Fauna 313, 316, 317, 319, 320, 321, 324, 340, 353, 357, 390, 393, 394, 396, 420, 439, 440, 467, 477, 593, 608 Fecal coliform 337, 338, 340-342 Feeding 53, 56, 102, 107, 109, 112, 131, 132, 134-136, 153, 162, 180, 301, 322, 405, 484, 497-500, 517, 521, 628, 629, 630, 632, 633 Fenestraspongia 271, 272 Ferric iron 99, 497-499, 501 Filter-feeding 89, 107, 109, 114, 283, 301, 335, 427 Fjord 139, 142-144, 498, 501, 613, 614, 619 Flagella 26, 39, 41, 283-286, 345, 346, 348, 350, 483-486, 488, 489, 614 Flagellated cells 488, 489 Foraminifera 199, 439, 440, 441 Forcepia 354, 356, 358 Fossil 11, 12, 14, 20, 62, 63, 91, 93, 94, 117, 118, 165, 179, 180, 384, 394, 395, 399 France 41, 166, 340, 362, 364, 367, 368, 517, 642, 654, 655 Freshwater sponge 19, 20, 54, 56, 61-63, 67, 69, 70, 72-75, 95, 97, 101, 103, 117-120, 179-183, 383, 384, 388, 390, 497, 500, 581-584 Function 54, 56, 79, 82, 85, 89, 94, 100, 101, 109, 114, 125, 165, 170, 179, 258, 259, 288, 291, 345, 408, 421, 429, 443, 480, 486, 491, 571, 585, 587, 590, 603, 627, 630, 632

G
Galectin-2 585-589 Gametes 54, 328, 329, 414, 416, 613, 616, 619, 620 Gametogenesis 327-330, 332, 333, 413, 417, 613, 615-617 Gammaproteobacteria 166, 566 Gap junctions 54, 286 Gastraea 41, 284, 413 Gastrophanella 354, 356, 358 Gastrulation 41, 43, 44, 53, 56, 100, 102, 281-291 Gemmules 19, 54, 61, 62, 67-72, 74, 97-99, 112, 119, 283, 385, 388, 390 GenBank 124-126, 166, 364, 367, 374, 383, 384, 386, 562, 566, 604, 605, 622-624 Gene 44, 47, 90-92, 95, 97-103, 114, 124, 165, 166, 168, 179184, 244, 281, 282, 288-291, 367, 374, 383-385, 406, 408, 409, 499, 561- 564, 582, 584, 603-605, 607, 609, 622-624, 630, 642 Geodia 32, 45, 89, 90, 147, 148, 150, 152, 154, 233, 234, 236, 313-316, 321, 323, 324, 354, 356, 358, 386, 405, 561, 593-596, 598-601, 613, 616, 617, 619, 620, 640, 651 Geodia australis 354, 356, 358, 593

671

Geodia barretti 613, 615-617, 619, 620 Geodia corticostylifera 233-234, 313-316, 593 Geodia cydonium 45, 89-97, 152, 183, 386, 405-407, 409, 561, 584, 586, 613, 617, 640 Geodia gibberosa 152, 321, 323, 324, 593, 594, 600, 601 Geodia glariosa 593-596, 598-601 Geodia media 147, 148, 150, 152, 386 Geodia neptuni 32, 233, 354, 356, 358, 386, 593 Geodia papyracea 32, 234, 236, 386, 593 Geodia riograndensis 354, 356, 358, 593 Geodia splendida 354, 356, 358, 593 Geodia tylastra 593 Geodiidae 314, 356, 386, 467, 471, 593, 594, 613 Geographic distribution 16, 61, 62, 69, 72, 73, 277, 278, 396398, 540, 600 Geographic range 67, 68, 71, 72 Geological provinces 11, 12 Germ layers 41, 43, 56, 97, 281-283, 285-288, 291, 484, 489 Global warming 185, 419, 422 Glutamate 54 Glyconectin 79-86 Glyconectins 80, 82-87, 653 Glycoprotein 99, 491, 493 Golgi complex 345, 348-349, 587 Gondwanian origin 68 Gonochorism 327, 613 Gorgonian 420, 443-446 Gorgonians 4, 5, 7, 135-136, 261, 321, 443-447 Grantia 354, 356, 358, 413, 416-417 Grantia compressa 413, 416-417 Great Barrier Reef 32, 153, 298, 301, 569-570, 572, 608 Growth 31, 36, 56, 61, 67, 68, 92, 93, 95, 102, 103, 132, 139-142, 144, 150, 152, 153, 157, 160, 162, 180-182, 205, 216, 224, 240, 244, 245, 247-253, 255, 259, 281, 287, 297-301, 310, 313-317, 319, 320-322, 324, 329, 331, 341, 405, 434, 443, 446, 447, 485, 497-500, 511, 514, 518, 522, 527, 570, 581, 587, 588, 606, 619, 625, 627-629, 639, 642, 646 Growth-differentiation balance hypothesis 627, 628, 633 Guam 32, 629 Guancha 413, 415, 416 Guancha blanca 413, 415, 417 Guitarra 109, 189, 190, 199, 200 Guitarra dendyi 189, 190, 199, 200 Gulf stream 320

H
Habitat 109, 117, 119, 123, 124, 135, 139, 144, 165, 212, 233, 234, 255-262, 305, 319-321, 324, 361, 379, 420, 443, 446, 477, 491, 517, 567, 627-629, 632, 641 Hadromerida 23, 29, 32, 38, 49, 50, 181, 190, 234, 236, 247, 283, 356, 379, 384-386, 388, 389, 405 Halichondria 4, 23, 32, 79, 80, 85, 95, 144, 159, 174, 176, 212, 321, 323, 324, 327-332, 354, 356, 358, 385, 386, 419-421, 484, 493, 497, 617, 619, 640, 642, 654 Halichondria bowerbanki 321, 323, 324, 330, 332, 333, 419421, 424 Halichondria melanodocia 386 Halichondria moorei 484 Halichondria panicea 79, 80, 84-87, 95, 212, 327-330, 332, 333, 385, 386, 493, 617, 619, 640, 642, 654 Halichondria sitiens 327-333 Halichondrida 32, 44, 46, 123, 174, 236, 314, 356, 385, 386, 388, 389, 393 Halichondriidae 133, 176, 314, 327, 356, 386

Haliclona 31, 33, 34, 36-38, 48, 67, 68, 147-150, 152-154, 158162, 173-176, 219, 225-227, 229, 230, 233, 234, 236, 283, 290, 338, 354, 357, 358, 385-387, 444, 478, 484, 489, 493, 495, 621, 622, 624, 625, 640 158-160, 162, 219, 225, 226, 354, 357, 358 Haliclona (Gellius) Haliclona (Gellius) caerulea 149 Haliclona (Gellius) cymaeformis 158-160, 162 Haliclona (Gellius) rudis 219, 225-226 Haliclona (Halichoclona) 338, 354, 357, 358 Haliclona (Halichoclona) vansoesti 338 Haliclona (Haliclona) 158-160 Haliclona (Reniera) 33, 159-160, 493, 495 Haliclona (Reniera) cinerea 493, 495 Haliclona (Rhizoniera) 176, 219, 226, 227, 230, 338 Haliclona (Rhizoniera) curacaoensis 233, 236, 338 Haliclona (Rhizoniera) dancoi 219, 226, 227, 230 Haliclona (Soestella) 31, 34, 36-38 Haliclona (Soestella) caerulea 36, 147, 149, 151-154 Haliclona (Soestella) melana 36, 233, 234, 236 Haliclona (Soestella) walentinae sp. nov. 31, 34, 36- 38, 621, 622, 624, 625 Haliclona amphioxa 386 Haliclona aquaeductus 385, 386 Haliclona fulva 640 Haliclona implexiformis 154, 444 Haliclona manglaris 233, 234, 236 Haliclona mediterranea 385, 386 Haliclona oculata 385, 387, 640 Haliclona sonorensis 147, 148, 150, 152, 153 Haliclonissa verrucosa 219, 227, 228, 230 Halicometes minuta 355, 356, 358 Halisarca 42, 44, 46-49, 333, 338, 345, 346, 348, 350, 414, 483, 484, 562 Halisarca caerulea 338 Halisarca dujardini 2, 44, 46-48, 333, 345, 346, 348, 350, 351, 414, 483, 484, 488, 489, 562 Halisarcida 42, 44, 47-50, 282, 283, 345, 346, 348, 444 Halitoxin 175, 176 Haloperoxidase 491-495 Hamacantha microxifera 355-358 Hamacanthidae 356 Haplosclerida 31, 32, 34, 36, 38, 44, 46, 47, 49, 61, 62, 67, 133, 147, 153, 154, 160, 173, 174, 225, 234, 236, 314, 357, 383-386, 388, 389, 416, 563, 621 Hard-bottom 239, 242, 245, 255-262, 319-321 Heat shock proteins 94, 408, 410 Hemiasterella 386 Hemiasterellidae 386 Hemimycale columella 629, 640 Hemolytic activity 313, 315, 317 Hemolytic assay 315 Herbivory 131, 211, 629 Herengeria 395-399 Herengeria auriculata 395-399 Herengeria vasiformis 395-398 Hermaphroditism 327, 330 Heteromeyenia 64, 119 Heteromeyenia baileyi 64 Heteromeyenia horsti 64 Heteromeyenia insignis 64, 119 Heteromeyenia latitenta 64 Heteromeyenia stepanowii 64 Heteromeyenia tentasperma 64 Heteromeyenia tubisperma 64 Heterorotula 64 Heterorotula capewelli 64

672

Heterorotula contraversa 64 Heterorotula fistula 64 Heterorotula kakauensis 64 Heterorotula multidentata 64 Heterorotula multiformis 64 Heterorotula nigra 64 Hexactinellida 17, 19, 20, 41, 42, 44, 46, 48, 49, 89, 90, 93-95, 112, 124, 139, 165, 179, 282, 283, 303-306, 308, 357, 379, 393, 449, 463, 581, 645, 646, 651 Hexactinosida 16, 18, 20, 357, 464 Higginsia strigilata 321, 323, 324 Hippospongia 265-267, 271, 484, 562 Histology 55, 94, 108, 248 Holacanthus ciliaris 131-136 Holacanthus tricolor 131-136 Homeobox 89, 93, 95, 99-102, 281, 289-291 Homeostasis 102, 281 Homoclerophorida 32 Homophymia 394, 396-398 Homophymia pollubrum 397 Homophymia stipitata 394, 396-398 Homosclerophorida 38, 356 Horizontal gene transfer 604, 609 Housekeeping proteins 94 Houssayella 66, 119 Houssayella iguazuensis 66, 119 Hyalonema 90, 304, 306, 310, 355, 357, 358, 582, 590 Hyalonema schmidti 355, 358 Hyalonema sieboldi 304, 306, 582, 590 Hyalonematidae 305, 357 Hyattella 159, 271, 314-316, 355, 357, 358 Hyattella intestinalis 159, 314- 316 Hybridization 98-100, 406, 407, 409 Hydrothermal sites 107, 109 Hymedesmia 355, 356, 358, 493 Hymedesmiidae 356, 387 Hymeniacidon 335, 337, 338, 340, 341, 556, 639, 640, 654 Hymeniacidon heliophila 556-559, 654 Hymeniacidon perleve 639, 640 Hyrtios 33, 38, 272, 323, 324, 338, 621, 622, 624 Hyrtios proteus 338 Hyrtios violaceus 33, 38, 323, 324, 621, 622, 624, 625

Iophon terranovae 219-221, 230 3, 4, 233, 234, 236, 337, 338, 341, 387, Iotrochota birotulata 443-446, 495 Iotrochotidae 387 Ircinia 33, 47, 159, 160, 233, 234, 236, 239, 242, 266-271, 314316, 322-324, 338, 355, 357, 358, 433-436, 444-446, 493, 561563, 566, 640 Ircinia aucklandensis 268, 269 Ircinia campana 33, 322-324, 338, 433-436 Ircinia echinata 159, 160 Ircinia fasciculata 435 Ircinia felix 33, 242, 322-324, 3338, 561-567 267 Ircinia irregularis Ircinia muscarum 640 47 Ircinia oros Ircinia ramosa 33 242 Ircinia retidermata Ircinia strobilina 233, 234, 236, 314-316, 338, 355, 357, 358, 444-446 268 Ircinia subaspera Ircinia variabilis 33, 242 Irciniidae 266, 268-270, 314, 357, 433 Ireland 645 Iron 95, 99, 206, 408, 497-501 Isabella 395, 397 Isabella mirabilis 397 219, 223, 225 Isodictya erinacea Isoraphiniidae 396 Italy 219, 362, 364, 368, 415, 640 ITS 124, 361, 362, 365, 367, 369, 556, 557

J
Jania adherens 147, 149-153 Japan 32, 63-65, 174-176, 374, 384, 571 32, 274, 355, 356, 358 Jaspis Jaspis stellifera 32 Jereicopsis graphidophora 395, 397

K
Kinases 91, 94, 96, 410

I
IATA 373, 375, 557 Identification 8, 62, 91, 97, 117-119, 123, 124, 128, 132, 168, 176, 219, 289, 306, 308, 309, 320, 336, 384, 407, 471, 489, 492, 517, 542, 548, 603-610, 632 Igernella notabilis 321, 323, 324 Immunofluorescence 184, 484, 485, 587, 589, 630 Indicator species 458 Indonesia 63-66, 118, 157, 162, 174, 176, 203, 240, 509, 514 Inducible defense 627, 630 Inland water 62 Innate immunity 79, 96, 405 Integrin 45, 47, 91, 92 Interactions 3, 8, 45, 47, 50, 82-86, 90, 100, 147, 149, 151-153, 165, 212, 215, 216, 233, 247, 285, 288, 289, 308, 439, 443, 446, 447, 507, 625, 629, 630, 632, 633 Intercellular adhesion 45 Intracellular calcium 54, 100 Introgression 604, 605, 609 Iophon 219-221, 230, 333, 414 Iophon piceus 333, 414

L
Lake Baikal 102, 179-182, 185, 187, 384, 388, 390 Lamellodysidea 33, 273, 622, 624, 625 Lamellodysidea chlorea 33, 624, 625 Lamellodysidea herbacea 33, 624 Laminins 50 Lanuginellinae 449, 451, 463, 464 Larispongia magdalenae 12, 13, 17 Larva 46-48, 50, 56, 99, 103, 112, 251, 252, 282-285, 287, 288, 345, 346, 350, 414, 416, 483, 485, 486, 488, 489, 564, 565 Larval settlement 139, 152, 251, 285, 399, 483, 629 Latrunculia 32, 355, 356, 358, 628 Latrunculiidae 356 Leiodermatium 32, 395-398 Leiodermatium dampieri 396-398 Leiodermatium linea 396-398 Leiosella 271 Leishmania 317, 433-436 Lendenfeldia 33, 273, 622, 624

673

Lendenfeldia dendyi 33 Lendenfeldia frondosa 33 Lepidothenea incrustans 394-398 Leucandra 321, 323, 324, 416 Leucetta 34, 323, 324 Leucetta imberbis 323, 324 Leucettida 32 Leucosolenia 54, 355, 356, 358, 416, 417 Leucosolenida 356 Leucosoleniidae 356 Life cycle 48, 61, 67, 68, 74, 413, 488 Life history 154, 173, 327, 330, 413 Limestone 13, 157, 203-206, 208, 256, 319, 321, 399, 513 Lipopolysaccharide (LPS) 95, 96, 409, 630 Lissodendoryx 133, 323, 324, 338, 355, 356, 358, 532 Lissodendoryx (Anomodoryx) sigmata 323, 324 Lissodendoryx (Lissodendoryx) colombiensis 338 Lithistid 11, 14-16, 385, 387, 388, 393-397, 399-402 Lithistida 32, 38, 174, 356, 396 Living fossils 93, 94, 179 Lophocalyx 355, 357, 358, 449-465, Lophocalyx atlantiensis sp. nov. 450, 458-460, 463, 464 Lophocalyx biogasi sp. nov. 450, 452, 454, 455, 463 Lophocalyx brasiliensis sp. nov. 450, 457-459, 463, 464 Lophocalyx oregoni sp. nov. 450-453, 455, 463, 464 Lophocalyx pseudovalida sp. nov. 450, 460-463 Lophocalyx reiswigi sp. nov. 461 Lubomirskia 65, 102, 179-181, 385, 388-390, 581, 582 Lubomirskia baikalensis 65, 102, 179-187, 385, 387-390, 581, 582 Lubomirskiidae 61, 62, 65, 67, 68, 179, 181, 182, 383-385, 387390 Luffariella 266, 268, 271, 272, 631 Lyidium 394 Lyssacinosida 16, 18, 20, 357, 645, 646

M
Macandrewia spinifoliata 394, 396, 397, 398 Macandrewiidae 396 Makedia 67 Malawispongia 65 Malawispongia echinoides 65 Malawispongiidae 61, 62, 65, 67, 68, 76, 180, 383, 385, 390 Mangrove habitats 31, 340 Marine sediments 11 Mechanical properties 303, 309, 503, 504, 507 Mediterranean 23, 27, 29, 30, 32, 38, 75, 95, 109, 147, 165-167, 240, 266, 297, 300, 317, 361, 362, 367, 369, 384, 405, 407, 419, 420, 427, 497, 498, 500, 501, 517, 523, 605, 613, 628, 629, 639, 640 Mellonympha 464, 649, 651 Mellonympha velata 649, 651 Mesenchyma 44, 56, 98, 99, 281, 283, 285-291, 310 Mesoderm 55, 56, 58, 100, 281, 282, 285-291 Mesohyl 26, 27, 54-56, 89, 165, 167, 168, 242, 244, 265-272, 281, 285, 286, 329, 331, 414, 416, 422, 503, 507, 561-564, 567, 587, 613-615, 617, 619, 630 Mesohyl creep 503 Metamorphosis 41, 44, 47, 282-285, 290, 413, 483-485, 488, 489 Metania 66, 70, 118, 119 Metania fittkaui 66 Metania pottsi 66 Metania reticulata 66, 70

Metania rhodesiana 66 Metania spinata 66 Metania vesparia 66 Metania vesparioides 66 Metania. godeauxi 66 Metania. kiliani 66, 119 Metania. ovogemata 66 Metania. subtilis 66 Metaniidae 61, 62, 66-71, 383-385, 387, 388, 390 Metazoa 41, 42, 44, 45, 47, 48, 50, 53, 55, 56, 82, 89-91, 93-95, 97, 98, 100-102, 107, 123, 165, 179-181, 183-185, 242, 281-289, 291, 305, 308, 309, 327, 345, 439, 484, 585, 609 Metschnikowia 66 Metschnikowia tuberculata 66 Metschnikowiidae 61, 62, 66-68, 180, 383, 385 Microbes 95, 96, 177, 341, 379, 420, 567, 633 Microbial consortia 165, 561, 562 Microbial diversity 170, 420, 561, 565, 567 4, 45, 79, 80, 85, 90, 387, 484, 642, 654 Microciona 4 Microciona microchela Microciona prolifera 45, 79, 80, 84-87, 90, 387, 484, 642, 654 Microcionidae 32, 133, 220, 356, 387 Microelectrodes 379 Microenvironment 287, 379 Microorganisms 97, 147, 165, 168, 498, 561-565, 567 Microscleroderma 394, 396, 397 Microscleroderma herdmani 397 Microscleroderma novaezealandiae 397, 540 Microscleroderma stonae 397 Microxina benedeni 219, 228, 229 Microxina phakelloides 219, 229 , 230 Mineralogy 203, 210, 239, 244 Mitochondrial 124, 180, 181, 361, 603, 604, 605, 606, 609 Mitochondrial DNA 369, 603, 605 Mitogen-activated protein (MAP) 96 Molecular evolution 92, 607, 609 Molecular markers 97, 98, 361, 362, 364, 365, 367, 370, 411, 484, 591, 603, 653 Molecular phylogeny 50, 89, 114, 180, 383 Molecular recognition 79 133, 313-317, 338, 443-446 Monanchora Monanchora arbuscula 313-317, 338, 446 Monanchora barbadensis 443, 444 Monanchora unguifera 444-446 Monophyly 90, 91, 93, 114, 178, 180, 281, 383-385, 389 Morphogenesis 41-46, 48, 50, 80, 92, 93, 100, 179, 284, 483, 493, 498, 588 Morphological descriptions 609 Morphology 5, 23, 31, 34, 36, 37, 61, 103, 108-110, 112, 118, 123, 124, 139, 150, 162, 163, 165, 203, 212, 240, 253, 255, 265, 284, 286, 287, 291, 319, 330, 361, 368, 390, 395, 413, 414, 446, 468, 471, 483, 484, 486, 489, 504, 512, 514, 542, 558, 581, 587, 589, 593, 606, 607, 609, 629 mtDNA 124, 374, 604, 605 Multicellularity 79, 80, 82, 87, 91, 94, 281, 284, 287 Multi-dimensional scaling 337, 339 Multilayer embryos 44 Mutualism 147, 151, 153, 439 Mycale 4, 32, 131, 133, 153, 158-160, 219, 222, 224, 233, 234, 236, 321, 323, 324, 338, 341, 355-358, 387, 439, 440, 441, 444446, 484, 562, 563 Mycale (Mycale) crassissima 159, 160 Mycale (Oxymycale) acerata 219, 222, 224 Mycale beatrizae 355-358 Mycale carmigropila 444-446 Mycale diversisigmata 233, 234, 236

674

Mycale fibrexilis 321, 323, 324, 387 Mycale hentscheli 32 Mycale laevis 153, 338, 341, 444-446 Mycale laxissima 133, 338, 444, 562, 563, 565, 567 Mycale microsigmatosa 4, 338, 341, 439, 440, 446 Mycale philippensis 158, 159, 160 Mycalidae 32, 113, 133, 222, 356, 387 Myotrophin 93, 95, 99 Myxilla 190, 191, 193, 219, 221, 223, 230, 333, 355, 356, 358, 414, 493, 494, 525-542, 544, 545, 654 Myxilla (Burtonanchora) asigmata 527, 529, 530-532, 538 Myxilla (Burtonanchora) lissostyla 525, 527, 538, 540-542 Myxilla (Burtonanchora) magna 525, 527, 531, 533-536 Myxilla (Burtonanchora) pistillaris 525, 527, 535, 537, 539, 540 Myxilla (Ectyomyxilla) hentscheli 525, 527, 542, 544 Myxilla (Ectyomyxilla) mariana 219, 221, 223, 230 Myxilla (Ectyomyxilla) tenuissima 355, 356, 358 Myxilla (Myxilla) elongata 525-530 Myxilla (Myxilla) mollis 190, 191, 193, 534, 535 Myxilla (Myxilla) rosacea 493, 494, 654 Myxillidae 191, 221, 356, 525, 526, 531, 545

N
NaCl 315, 373-376, 492, 557-559, 654, 655 Nanobiotechnology 581, 585, 587-589 Narrabeena 272 Natural products 152, 173, 177, 313, 316, 317, 335, 627 Neamphius huxleyi 32 Necrosis 629 Neighbour-Joining 95, 182, 365, 367, 369, 388, 389 Neoaulaxinia 393-399 Neoaulaxinia clavata 396-398 Neoaulaxinia persicum 396-399 Neoaulaxinia zingiberadix 396-398 Neofibularia 31, 32, 36, 337, 338, 341, 355, 356, 358, 444 Neofibularia irata 32 Neofibularia nolitangere 36, 337, 338, 341 Neopelta 396-398 Neopelta plinthosellina 397 Neopelta pulvinus 396-398 Neopeltidae 396 Neopetrosia 33, 337, 338, 340, 341, 621, 622, 624, 625 Neopetrosia carbonaria 338 Neopetrosia exigua 33, 625 Neopetrosia subtriangularis 33, 337, 338, 340, 341, 621, 622, 624, 625 Neophrissospongia 395, 397 Neophrissospongia microstylifer 397 Neoschrammeniella 393, 395-398 Neoschrammeniella castrum 397 Neoschrammeniella fulvodesmus 396-398 Neoschrammeniella moreti 397 Neoschrammeniella norfolkii 397 Neosiphonia 393-398 Neosiphonia motukawanui 396-398 Neosiphonia superstes 394-398 Nervous system 53, 99, 291, 431 Netherlands 124, 173, 174, 319, 330, 332, 427, 450, 497, 498, 501, 645 New records 159, 162, 189, 190, 219, 237, 319, 321, 324, 593 New species 15, 23, 30, 31, 34, 37, 38, 73, 107, 108, 110, 117, 118, 120, 189, 190, 219, 234, 265, 274, 275, 277, 279, 393, 394,

395, 449, 452, 455, 460, 463, 464, 467, 468, 473, 477, 480, 481, 509, 512, 514, 515, 517, 522, 547, 549, 550, 552, 605, 607, 610 New Zealand 63, 64, 108, 110, 112, 265-270, 339, 394-402, 552 Niphates 33, 38, 233, 236, 314, 316, 323, 324, 335, 337, 338, 340, 355, 357, 358, 385, 387, 444-446, 563, 622 Niphates caycedoi 338, 444, 445 Niphates digitalis 385, 387, 563, 565, 567 Niphates erecta 233, 236, 323, 324, 335, 337, 338, 340, 444-446 Niphatidae 31, 33, 38, 133, 173-176, 227, 314, 357, 384, 387 Nitrate 165-168, 170, 337, 340, 493, 498 Nitrification 165, 166, 167, 170 Nitrite 165, 166, 168, 337, 338, 340 Noggin 98, 99, 101, 102 Non-synonymous substitutions 91, 92 North Vietnam 157, 162 Northeastern Atlantic 427, 491 Northern blot 99, 100, 406, 407, 409 Norway 62, 124, 449, 613, 614 Nuclear membrane 348, 349, 561, 563 Nucleus 26-29, 41, 95, 100, 288, 329, 330, 331, 345, 346, 348350, 413, 414, 416, 521, 604, 614, 615, 617, 619 Nudospongilla 64 Nudospongilla coggini 64 Nudospongilla cunningtoni 64 Nudospongilla ehraiensis 64 Nudospongilla moorei 64 Nudospongilla vasta 64 Nudospongilla yunnanensis 64 Nutrient 123, 154, 165, 203, 248, 255, 259, 261, 262, 335-337, 339-342, 381, 402, 415, 419, 443, 499, 608, 621

O
Oamaru 394, 399 Oceanapia 33, 239, 240, 241, 244, 338, 355, 357, 358, 383, 387, 628 Oceanapia ambionensis 33 Oceanapia nodosa 338 Oceanic islands 61, 63, 233, 236, 467, 468, 473 Ochridaspongia 65 Ochridaspongia interlithonis 65 Ochridaspongia rotunda 65 Ohridospongilla 67 Ohridospongilla stankovici 67 Oncosclera 66, 118, 119 Oncosclera atrata 66 Oncosclera diahoti 66 Oncosclera gilsoni 66 Oncosclera intermedia 66 Oncosclera jewelli 66, 119 Oncosclera navicella 66 Oncosclera petricola 66 Oncosclera ponsi 66 Oncosclera rousseletii 66 Oncosclera schubarti 66 Oncosclera schubotzi 66 Oncosclera spinifera 66 Oncosclera stolonifera 66 Oncosclera tonolli 66 Ontogeny 41, 55, 56, 245 Oocytes 26, 97, 98, 328-332, 413, 414, 416, 521, 563, 565, 613615, 617, 619, 620 Oogenesis 613-615, 617, 619, 620 Ophlithaspongia tenius 654 Optimal defense theory 627, 628, 633

675

Ordovician 11-20, 109, 114 Organelles 345, 346, 349 Organic pollution 4, 5, 248, 340 Oscillatoria spongeliae 31, 36-38, 621-625 Oscula 34, 48, 54, 100, 102, 108, 140, 148, 150, 240, 241, 247, 249, 273, 414, 473, 497 Oscules 29, 31, 34, 36, 100, 102, 148, 182, 190, 193, 199, 220, 221, 225-229, 247, 249, 250, 271, 275, 278, 471, 497, 499, 500, 510, 526, 527, 532, 538, 548, 598, 632, 646 Osculum 18, 48, 54, 55, 107, 141, 144, 285, 380, 451, 458, 485, 486 Ostia 54, 107, 148, 242, 478 Overgrowth 31, 174, 211, 443, 444, 446, 447, 627, 630 Ovoviviparous 56, 327-330 Oxygen 96, 102, 142, 247, 379, 380, 381, 408, 578, 585, 619, 623, 639

P
Pachastrella 355, 356, 358, 477, 481 Pachastrella monilifera 355, 356, 358, 477 Pachastrellidae 133, 356, 477, 478, 482 Pachastrissa 355, 356, 358 Pachataxa 477 Pachataxa lutea 477 Pachydictyum 66 Pachydictyum globosum 66 Pachypellina 355, 357, 358 Pachyrotula 64 Pachyrotula raceki 64 Pacific Ocean 63, 147, 162, 304, 361, 402, 654 Palaeospongilla chubutensis 11, 18, 19, 62 Palaeospongillidae 19, 384 Palau 32, 157, 162, 625 Panama 31, 34, 37, 65, 135, 153, 335, 336, 338-340, 621, 622 Papua New Guinea 32 Parabiosis 96 Paraleucilla 413-417, 556, 557 Paraleucilla magna 413-417, 556, 557 Parasites 95, 108, 132, 313, 434 Parasitism 147, 151, 439 Parenchymella 41, 45, 47-50, 61, 67, 100, 282, 283, 285, 329, 483, 485, 521, 563 Parsimony 166, 168, 384, 385, 388, 389 PCR 99, 165, 166, 289, 290, 362, 373-376, 384, 406, 420-422, 424, 555-558, 562, 563, 566, 604, 622 Pectispongilla 64, 73 Pectispongilla aurea 64 Pectispongilla botryoides 64 Pectispongilla stellifera 64 Pectispongilla subspinosa 64 Pedra da Risca do Meio Marine State Park 313, 316, 317 Pellina semitubulosa 33 Penares schulzei 32 Penares sollasi 159, 160 Percent cover 211, 213-215, 341 Pericharax heteroraphis 34 Petromica 314, 315, 355, 356, 358 Petromica ciocalyptoides 314, 315 Petrosaspongia 271, 272 Petrosia 33, 159, 338, 383, 582, 639-641 Petrosia ficiformis 33, 582, 639-642 Petrosia pellasarca 33, 338 Petrosiidae 31, 33, 36, 174-176, 383-385, 387 Petrosina 67, 383, 385, 388, 389

Phakellia 355, 356, 358, 549 355, 356, 358 Phakellia connexiva Phelloderma 107, 108, 110 233, 305, 355, 357, 358, 645 Pheronema Pheronema carpenteri 233, 355, 357, 358, 645 Pheronematidae 305, 357 Phloeodictyidae 33, 357, 385, 387 Phorbas 32, 321, 323, 324, 387, 493, 640 321, 323, 324 Phorbas amaranthus Phorbas fictitius 493, 640 Phorbas tenacior 387 Phoriospongiidae 387 Photosynthesis 569-571, 578, 621, 623-625, 629 Phototaxis 483-486 Phyllospongia 33, 273, 622, 624 Phyllospongia alcicornis 33 Phyllospongia foliacens 33 Phyllospongia papyracea 33, 624 Phylogenetic 31, 37, 38, 41, 42, 48, 50, 89, 90, 91, 93-95, 100, 102, 110, 165, 166, 168, 170, 173, 179, 181-183, 185, 186, 282, 289, 346, 364, 367, 368, 383-385, 388-390, 555, 561, 567, 582584, 603, 604, 609, 610, 621, 622, 624, 625 Phylogeny 41, 48, 62, 89, 114, 117, 168, 173, 180, 181, 362, 383385, 388, 605, 621, 623, 625 Phymatellidae 395-397 Physiology 56, 162, 341, 483, 497, 499, 500, 503, 570 Phytoplankton 401, 613, 619, 620 Pinacocytes 95, 99, 239, 244, 285, 286, 483, 485, 488, 489, 563 Pinacoderm 56, 57, 89, 98, 100, 101, 103, 240, 242, 244, 271, 284-286, 420, 483, 485, 488, 489, 539, 639 159, 160, 252, 386 Pione Pione carpenteri 159, 160 Pione lampa 252 386 Pione velans Placinolopha mirabilis 32 233, 234, 236, 238, 286 Placospongia Placospongia aff. melobesioides 233, 234, 236 238 Placospongia intermedia Placospongiidae 386 Plakina 355, 356, 358 Plakinastrella 133, 355, 356, 358 Plakinidae 32, 38, 356 Plakortis 337, 338, 341, 344, 564 337, 338, 341 Plakortis angulospiculatus 338 Plakortis halichondrioides Plasma membrane 81, 91, 92, 346 Pleraplysilla 47, 273, 640 Pleraplysilla spinifera 47, 640 Pleroma 394-399 Pleroma aotea 394-399 Pleroma menoui 394, 396-398 Pleroma turbinatum 394, 395-398 Pleromidae 396, 397 Pocillopora 151-154 Pocillopora capitata 149 Pocillopora damicornis 149 Pocillopora meandrina 149 Pocillopora verrucosa 149 Podospongiidae 387 Poecillastra 355, 356, 358, 477, 478, 480-482 355, 356, 358, 477, 481, 482 Poecillastra sollasi Poecilosclerida 32, 44, 46, 49, 107, 109, 110, 112-114, 163, 174, 191, 220, 234, 236, 314, 356, 358, 384, 385, 387-389, 518, 521, 525, 526, 563 Polarity 45, 56, 89, 93, 100, 101, 239, 244, 283, 284, 346, 485, 629

676

Polluted sites 03-05, 341 Pollution 3-8, 248, 259, 261, 335, 336, 339-342, 427 Polymastia 23, 24, 27-29, 49, 163, 283, 341, 355, 356, 358, 617 Polymastia arctica 27-29 Polymastia corticata 355, 356, 358 Polymastia grimaldi 27-29 Polymastia harmelini sp. nov. 23, 24, 27-30 Polymastia janeirensis 29, 30 Polymastia mamillaris 29 Polymastia penicillus 28-30 Polymastia robusta 29, 30, 49 Polymastiidae 23, 24, 356 Polysaccharides 81, 82, 87, 304, 555, 557, 653-658 Pomacanthus 131-136, 628 Pomacanthus arcuatus 131-135 Pomacanthus paru 131-135 Poribacteria 165, 561 Portugal 364, 368, 427, 491, 492, 600, 601 Potamolepidae 61, 62, 66-68, 70, 71, 118, 180, 383, 384, 385, 387, 388, 389 Potamolepis 66 Potamolepis belingana 66 Potamolepis chartaria 66 Potamolepis leubnitziae 66 Potamolepis marshalli 66 Potamolepis micropora 66 Potamolepis pechuelii 66 Potamolepis weltneri 66 Potamophloios 67, 70 Predation 131, 132, 136, 152, 153, 154, 211, 215, 300, 322, 467, 627, 628, 629 Preservation 11, 14, 19, 108, 114, 119, 148, 285, 291, 373, 374, 376, 431, 510, 555-558, 631 Prey capture 107, 112, 522 Primmorph 89, 93-97, 99-101, 288, 497-499, 581, 582, 584, 585, 587, 588, 639-642 Propagule 67, 69, 199, 503, 505, 507 Prosuberites 386 Prosuberites laughlini 386 Protachileum kayseri 12, 14, 15, 17 Proteobacteria 165, 561, 563 Proteoglycans 79 Protospongia 11-14, 16 Protosuberites 159, 160, 386 Psammocinia 33, 266-270 Psammocinia beresfordae 269 Psammocinia halmiformis 268-270 Pseudaxinella 32, 234, 386 Pseudaxinella reticulata 234, 386 Pseudaxinella tubulosa 32 Pseudaxinyssa 32 Pseudoceratina 364, 365 Pseudosuberites 162, 189-191, 199, 498, 640 Pseudosuberites andrewsi 498, 640 Pseudosuberites cf. antarcticus 189-191, 199 Ptilocaulis 323, 324, 386 Ptilocaulis gracilis 386 Ptilocaulis walpersi 323, 324 Puerto Rico 32, 135, 443, 447 Pumping activity 167, 380, 381, 497, 499

Racekiela pictovensis 64 Racekiela ryderi 64 Racekiela sheilae 64, 119 Radiation 18, 61, 152-154, 162, 180, 204, 361, 368 Radiospongilla 64, 389 Radiospongilla amazonensis 64 Radiospongilla cantonensis 64 Radiospongilla cerebellata 64 Radiospongilla cinerea 64 Radiospongilla crateriformis 64 Radiospongilla hemephydatia 64 Radiospongilla hispidula 64 Radiospongilla hozawai 64 Radiospongilla indica 64 Radiospongilla multispinifera 64 Radiospongilla philippinensis 64 Radiospongilla sansibarica 64 Radiospongilla sceptroides 64 Radiospongilla sendai 64 Radiospongilla sinoica 64 Radiospongilla streptasteriformis 64 Rarefaction 337, 339 Raspaciona 133, 355, 356, 358 Raspailia 321, 323, 324, 355-358 Raspailia (Parasyringella) 355, 356, 358 Raspailia phakellina 358 Raspaillidae 387 rDNA 94, 124, 181, 361, 362, 367, 388, 411, 420, 421, 561-563, 565, 566 Reactive oxygen species (ROS) 96 Red Sea 32, 42, 147, 162, 176, 623, 628 Reidispongia 394-398 Reidispongia coerulea 394-398 Reproduction 42, 44, 67, 68, 107, 109, 112, 114, 131, 132, 136, 203, 233, 283, 327, 332, 341, 413-416, 439, 503, 517, 521, 562, 563, 565, 578, 613, 619, 620, 628, 629 Reproductive period 332, 413-415, 417, 615, 617, 619 Respiration 53, 142, 381, 621-625 REVIZEE 353-355, 357, 449, 460, 477, 478, 547-549, 594 Rezinkovia 65, 389, 390 Rezinkovia echinata 65 Rhabderemia 32, 355-358 Rhabderemia besnardi 355-358 Rhabderemia itajai 355-358 Rhabderemia sorokinae 32 Rhabderemia uruguaiensis 355, 356, 358 Rhabderemiidae 356 Rhizaxinella 386 Rhopaloeides 271, 298, 561 RNA 96, 281, 290, 362, 383, 405-409, 556, 621, 622 Rootlets 345, 346, 348-351 Rossella 142, 451, 645, 646, 649-651 Rossella nodastrella 645, 646, 649-651 Rossellidae 357, 449, 451, 646, 648 ROV 109, 379, 380, 517, 518, 521 rRNA 90, 91, 165, 166, 168, 281, 367, 383-386, 388, 406, 556, 561-564, 621, 622, 623, 624 Russia 41, 62-64, 179, 196, 200, 327, 345, 383, 384, 449, 483, 645

R
Racekiela 64, 118, 119 Racekiela biceps 64

S
Saccospongia baccata 109, 114 Sandstone 12, 14, 19, 319, 321, 510 Sanidastra 64

677

Sanidastra yokotonensis 64 270, 493 Sarcotragus Sarcotragus spinosulus 493 Saturnospongilla 64, 69, 70, 118 Saturnospongilla carvalhoi 64, 118 Scalarispongia 265, 266, 269, 272 Scalarispongia scalaris 266 Scanning Electron Microscopy 26, 34, 37, 48, 61, 148-152, 182, 184, 185, 189, 219, 234, 304-307, 421, 477, 478, 480, 481, 509, 511, 513, 518, 521, 523, 525, 545, 551, 582, 587, 589, 593, 645, 646, 648 355, 356, 358 Sceptrella 645 Schaudinnia Schaudinnia rosea 645 Scleritoderma 394, 395-398 Scleritoderma camusi 397 Scleritoderma flabelliformis 395-398 Scleritodermiidae 396 Scopalina 3, 4, 233, 234, 236, 323, 324, 341, 386, 444-446 3, 4, 233, 234, 236, 323, 324, 341, 386, 444Scopalina ruetzleri 446 Screening 98, 100, 313, 316, 317, 405-407, 409, 495, 627 SCUBA 23, 34, 148, 158, 166, 257, 314, 328, 362, 374, 384, 428, 471, 473, 498, 504, 562, 594 Sea of Cortes 147, 148, 152 Sea urchin 82, 95, 131, 206, 211-216, 252, 255, 256, 313-317, 628, 656 Secondary metabolite 123, 131, 173, 216, 240, 273, 303, 361, 362, 419, 613, 627, 628-630, 631-633 Sediment incorporation 239, 242 Sedimentation 3-5, 12, 14, 90, 139, 144, 153, 154, 162, 180, 233, 242, 261, 322, 335, 336, 440, 507, 613, 619, 620 Sediments 3, 8, 11, 12, 14, 19, 20, 24, 62, 119, 203, 207, 239-242, 244, 245, 339, 379, 399, 428, 440, 444, 548, 600 Selectivity 239, 245, 260 Selenoprotein 585, 586 Self-recognition 79, 86, 281, 287 Semitaspongia 269, 270, 272 Semitaspongia incompta 270 Sequencing 80, 81, 90, 91, 94, 124, 126, 166, 180, 289, 362, 364, 368, 369, 374, 375, 384, 406, 407, 409, 411, 562, 567, 603-605, 621, 622, 656, 657 Settlement 3, 103, 139, 152, 153, 160, 162, 211, 251, 252, 284, 285, 336, 397, 399, 401, 483, 485, 486, 488, 562, 567, 629 Sewage pollution 4 Sexual reproduction 42, 44, 67, 68, 203, 283, 503, 613, 619, 629 Shipping 373, 376, 557 Sibling 124, 136, 147, 153, 165, 609 Signal transduction molecule 91, 101 Signaling molecule 50, 54, 100 Silica 90, 93-95, 144, 174, 179, 182-184, 186, 242, 303-311, 373, 393, 400-402, 433, 434, 555-558, 581-584, 587-590, 609, 639, 642 Silicase 581, 584, 585, 588-590 Silicatein 90, 93, 99, 103, 179-187, 303, 309, 497, 581-585, 587590, 642 Silicates 144, 179, 244 Silicon 99, 180, 185, 304, 307, 311, 402, 500, 501, 582-584, 588 Siphonidiidae 32, 356 Siphonochalina 32 Skeleton 11, 15, 23-25, 34, 36, 53, 56, 72, 89, 92-94, 102, 109, 112, 132, 142-144, 149, 150, 160, 182, 185, 190, 191, 193-199, 206, 220-230, 239, 245, 247-253, 255, 259-262, 265, 266, 270275, 278, 283, 284, 294, 297, 303-311, 327, 336, 383, 384, 393,

414, 446, 451, 452, 455, 457, 463, 464, 471, 478, 484, 510, 513, 525, 527, 528, 531, 532, 535, 536, 538, 539, 542, 543, 545, 547, 548, 550, 551, 570-573, 578, 581, 588, 593, 598-600, 648, 649 272-274, 321, 323, 324, 562, 563 Smenospongia Smenospongia aurea 562-565, 567 321, 323, 324 Smenospongia cerebriformis Smothering 139, 242, 446 Soft bottom 239-242, 244, 379, 440 South Amrica 12, 14, 63, 64, 117-120, 190, 192, 222, 223, 227, 229, 230, 258, 400, 595 South Atlantic 108, 195, 219, 221, 223, 224, 227, 229, 231, 319, 440, 509, 515 South Atlantic Bight 319 Southwestern Atlantic 131, 233, 275, 278, 439, 509, 515, 547 Sp. nov. 24, 31, 34, 36-38, 189, 190, 193-195, 197, 199, 275279, 321, 323, 324, 354-358, 450-452, 454-464, 471-473, 477, 478, 509-514, 518, 521-523, 547, 548, 550, 551, 552 Spain 124, 364, 367, 368, 428, 477, 492, 501, 525, 593, 601 Species composition 157, 162, 341, 393 Species discrimination 361, 362, 367-369 Species diversity 6, 340, 341 Species richness 5, 7, 62, 67, 71-73, 133, 199, 255, 422, 424, 440 Species-area curve 335, 337, 339 Species-specific 42, 45, 79, 82, 84-86, 90, 92, 443, 446, 447, 603 Spermatocysts 112, 328-331, 617 Spermatogenesis 42, 112, 328-332, 613, 614, 616, 617, 620 Sphaerotylus 355, 356, 358 Spheciospongia 32, 158-160, 240, 321, 323, 324, 386 Spheciospongia cuspidifera 240 Spheciospongia florida 32 Spheciospongia tentorioides 158-160 Spheciospongia vesparium 321, 323, 324, 386 Spiculation 61 Spiculogenesis 499, 582, 639, 642 Spinospongilla 66 Spinospongilla polli 66 Spirastrella 32, 131, 133, 159, 160, 233, 234, 236, 323, 324, 338, 640 233, 234, 236, 323, 324 Spirastrella coccinea Spirastrella cunctatrix 159, 160, 640 Spirastrella decumbens 159, 160 233, 236, 338 Spirastrella hartmani 323, 324 Spirastrella mollis Spirastrellidae 32, 386 Spirophorida 32, 48, 190, 236, 283, 356, 385, 387-389, 509, 510, 547, 548 Sponge abundance 5, 234 Sponge Barcoding Database (SBD) 123-125 Sponge Barcoding Project (SBP) 123, 124, 609 Sponge communities 3, 4, 7, 319, 570, 627 Sponge development 41-44, 47, 50, 56, 281, 284, 285, 290, 291, 562, 565, 567 Sponge distribution 3, 199, 212, 216, 261 Sponge diversity 7, 316, 335, 336, 339, 341 Sponge fauna 11, 12, 14, 15, 19, 20, 23, 62, 73, 117-119, 124, 157, 160, 162, 163, 180, 189, 219, 231, 233, 254, 236, 313, 319, 321, 324, 340, 353, 357, 390, 394, 467, 594 Sponge larvae 44, 45, 282, 285, 308, 351, 483, 486, 488 Sponge-alga association 147, 149, 153 Sponge-coral association 147, 149, 152 Sponge-cyanobacteria association 31 Spongia 23, 33, 159, 160, 265-268, 271, 274, 310, 323, 324, 338, 341, 428, 430, 431, 484, 493 323, 324 Spongia graminea

678

Spongia irregularis 159, 160, 167 Spongia officinalis 167, 341, 428, 430, 431 338 Spongia pertusa 323, 324 Spongia tubulifera Spongiidae 33, 38, 270, 271, 314, 357 Spongilla 62, 64, 71-73, 180, 181, 385, 387, 484, 583, 584 Spongilla alba 64 Spongilla aspinosa 64 Spongilla cenota 64 Spongilla chaohuensis 64 Spongilla fluviatilis 62 Spongilla heterosclerifera 64 Spongilla inarmata 64 Spongilla jiujiangensis 64 Spongilla lacustris 64 Spongilla mucronata 64 Spongilla palustris 62 Spongilla patagonica 62 Spongilla permixta 65 Spongilla prespensis 65, 72, 73 Spongilla shikaribensis 65 Spongilla spoliata 65 Spongilla stankovici 65, 72, 73 Spongilla wagneri 65 Spongillidae 11, 20, 42, 61-63, 67-73, 179-181, 383-385, 387390, 501 Spongillina 61, 62, 67-69, 73, 180, 383-385, 388-390 Spongiomorph structure 281 Spongivory 131, 134, 135 Spongosorites 355, 356, 358, 386 386 Spongosorites genitrix SST 419, 420, 422 Stelletta 32, 159, 275-279, 323, 324, 355, 356, 358, 493, 561, 617 Stelletta anancora 275, 277, 278 Stelletta anasteria 275, 278 Stelletta aruensis 159, 160 Stelletta beae 275, 278 277, 323, 324 Stelletta carolinensis Stelletta clavosa 32 Stelletta gigas 275, 277, 278 Stelletta grubii 493, 561, 617 Stelletta hajdui 275, 278 Stelletta kallitetilla 32, 275-279 Stelletta pudica 32, 277 Stelletta purpurea 275 Stelletta ruetzleri 275, 278 Stelletta soteropolitana sp. nov. 275-279 Stelodoryx 189, 190, 193, 194, 531 Stelodoryx argentinae sp. nov. 189, 190, 193, 194 Stelodoryx cribrigera 193 Stem cell 97-100, 281, 283, 285, 287, 290, 291, 310 Stem cell marker gene 89, 99 Sterol 383, 394, 408, 433-436 Sterrastrolepis 67, 118, 119 Sterrastrolepis brasiliensis 67, 118, 119 Stratospongilla 65, 69, 70 Stratospongilla africana 65 Stratospongilla akanensis 65 Stratospongilla bombayensis 65, 70 Stratospongilla clementis 65 Stratospongilla gravelyi 65 Stratospongilla indica 65 Stratospongilla lanei 65 Stratospongilla penney 65 Stratospongilla sumatrana 65

Strepsichordaia 273 Strongylacidon 133, 387 Strongylacidon bermudae 387 Strongylophora 383 Stylissa massa 640 Stylotella agminata 640, 641 Suberea 33 Suberea azteca 33 Suberea mollis 33 Suberites 4, 45, 89, 90, 159, 160, 162, 179, 181, 355, 356, 358, 386, 405, 408, 409, 493, 497, 581, 582, 630, 639-641, 654 Suberites aurantiaca 4 Suberites caminatus 355, 356, 358 Suberites carnosus 493 Suberites domuncula 45, 89-103, 179, 181-184, 186, 386, 405410, 497, 581-588, 630, 639-641 Suberites ficus 386 Suberitidae 133, 181, 190, 356, 386 Submarine canyon 109, 189, 190, 199 Submersible 107-109, 517 Sulawesi 32, 64, 66, 68, 174, 203, 240, 509 Sulfated polysaccharides 653-658 Survival 61, 95, 135, 136, 139, 152, 153, 239, 297-301, 305, 443, 491, 578, 607, 609, 610, 625, 639, 640, 641 Svenzea zeai 32, 338 Swartschewskia 65, 181, 384, 385, 387-390 Swartschewskia papyracea 65 Sweden 614 Swimming 47, 61, 112, 282, 283, 285, 287, 320, 345, 348, 350, 483-486, 488, 562, 563 Sycettida 32 Sycettidae 34 Sycon 34, 94, 413, 417 Sycon ciliatum 413, 417 Sycon raphanus 94, 95 Symbiosis 38, 62, 109, 131, 152, 162, 203, 439, 569 Symmetry 48, 93, 517, 522, 523 Sympagella 449, 464 Sympatric species 136, 242, 327, 657 Synechococcus spongiarum 32, 37, 621, 625

T
Taonura 272 Taxonomy 8, 23, 73, 118, 119, 123, 124, 162, 174, 196, 219, 233, 265, 274, 324, 327, 353, 375, 383, 393, 449, 467, 477, 509, 547, 593, 603, 605-610, 625, 653 Tectitethya 3, 4, 239, 240-242, 244 Tectitethya crypta 3, 4, 240-242, 244 Tedania 131, 133, 136, 159, 160, 189, 190, 193, 195-199, 233, 234, 236, 338, 341, 355, 357, 358, 387, 563 Tedania (Tedania) brevispiculata 159, 160 Tedania (Tedania) ignis 131, 133, 136, 234, 236, 338, 341, 387, 563, 565, 567 Tedania (Tedaniopsis) charcoti 190, 193, 195, 196, 199 Tedania (Tedaniopsis) massa 190, 195 Tedania (Tedaniopsis) sarai sp. nov. 189, 190, 195, 197, 199 Tedania (Tedaniopsis) vanhoffeni 355, 357, 358 Tedania (Trachytedania) mucosa 190, 198, 199 Tedaniidae 193, 357, 387 Temperature 6, 54, 84, 86, 157, 160, 162, 179, 180, 203, 256, 304, 305, 308, 310, 313, 314, 316, 328, 332, 333, 336, 346, 361, 373-375, 379, 393, 397-400, 415, 419-422, 424, 433, 434, 484, 492, 495, 497, 498, 503, 505, 507, 511, 515, 556-558, 588, 589, 590, 619, 623, 639-642

679

Tentorium semisuberites 380, 381, 619 Tertiary 11, 13, 18-20 32, 38, 49, 54, 133, 158-160, 162, 165, 166, 168, 181, Tethya 183, 233, 234, 236, 321, 323, 324, 367-369, 373, 374, 386, 571, 582, 584 374, 386 Tethya actinia Tethya aurantium 49, 181, 183, 373-376, 582, 583, 584 386 Tethya californiana Tethya seychellensis 158-160, 162, 571 Tethyidae 32, 38, 356, 386 Tetilla 34, 48, 49, 282, 387, 547, 549, 551, 552, 605 Tetilla (Cinachyrella) crustata 549 Tetilla arb 34 Tetilla euplocamus 547, 552 Tetilla japonica 387 Tetilla radiata 547, 552 Tetillidae 34, 133, 190, 283, 356, 387, 509, 510, 547, 548, 550 Tetractinomorpha 41, 405, 411, 548, 603 Tetraspanin 100, 101 233, 355, 356, 358, 619 Thenea Thenea abyssorum 619 Thenea fenestrata 233, 355, 356, 358 32, 174, 394, 564 Theonella Theonella conica 32 Theonella swinhoei 32, 174, 564 Theonellidae 32, 38, 396 Thorecta 267, 269, 271, 272 271, 272 Thorectandra Thorectaxia 272, 273 Thorectidae 33, 38, 268-271, 273, 274, 357 355, 356, 358 Thrombus 133, 355, 356, 358 Timea Timeidae 356 Tonkin Gulf 157, 162 159, 314-316, 355, 356, 358 Topsentia Totipotent cells 67-69, 98 Transcription factor 45, 47 Transmission electron microscopy 23, 305, 345, 346, 562, 564, 565 Trichimella 48, 49, 112, 283 619 Trichostemma Trichostemma sol 619 Tridentata marginata 631 Triploblasts 53, 55, 100, 291 Trochospongilla 19, 20, 65, 385, 387 Trochospongilla horrida 65, 385 Trochospongilla pennsylvanica 65, 387 Tubulin 94, 484, 485

Ulosa ruetzleri 341, 563, 565 Ultrastructure 345, 346 65, 71, 72 Umborotula Urmetazoa 89, 90, 91, 93, 95, 102, 185, 281, 283, 288, 291 67, 70, 118, 119 Uruguaya Uruguayella 65 USA 63-66, 124, 314, 315, 319, 419, 484, 489, 501, 517, 518

V
Venezuela 63-67, 118, 124, 262 Verongida 32, 38, 273, 311, 314, 357, 384, 385, 387, 389, 563 34, 303, 338, 444 Verongula Verongula gigantea 34, 303 34, 338 Verongula reiswigi Verongula rigida 34, 338, 444 Vetulina 385, 387, 394 Vetulina oamaruensis 394 Vetulina stalactites 385, 387 Vetulinidae 387 Video 54, 645, 646, 649, 651 Viviparous 284, 562, 563, 565 355, 356, 358, 477 Vulcanella

W
Western Atlantic 31, 361, 367-600 Western Pacific 147 White Sea 48, 327, 332, 333, 346, 483, 484, 489

X
Xestospongia 31, 33, 34, 36, 37, 38, 159, 160, 176, 338, 383, 385, 387, 444-446, 621, 622, 624, 640 Xestospongia bocatorensis sp. nov. 31, 34, 36-38, 621, 622, 624, 625 Xestospongia cf. testudinaria 159, 160 Xestospongia muta 33, 37, 385, 387, 640 33, 37, 338, 444, 445 Xestospongia proxima Xestospongia rosariensis 33, 37, 338 Xestospongia testudinaria 33, 159, 160 Xestospongia wiedenmayeri 33, 36

Z
Zambia 63, 64, 384 Zanzibar 32, 64 Zooxanthellae 203, 204, 240, 569, 570-573, 575, 577, 578 Zygochlamys patagonica 189

U
Ulosa 341, 484, 563

680

681

682

Participants List
A
Adams, Charles Alcolado, Pedro (20) Almeida, Maria Marise Andersen, Raymond J. (147) Andrea, Barbara Rustum (41) Assumpo, Leonardo Lus Marques (186) Austin, Catherine Austin, William C. (79) Azevedo, Fernanda Correia (71) Cosme, Bruno (85) Coutinho, Cristiano Carvalho (204) Coutinho, Maria Alice de Almeida Cunha, Andreia Brazo A.C. da Custdio, Mrcio Reis (47)

D
Davis, Andy (21) Desqueyroux-Fandez, Ruth (3) Diaz, Maria Cristina (1) Dicks, Emilie Fleur Dohrmann, Martin Dresch, Roger Remy Duckworth, Alan (194) Duyl, Fleur C. van (200)

B
Bakran-Petricioli, Tatjana (99) Bannister, Raymond John Barnes, Peter Brendan (26) Barreto, Maria do Carmo R.L. Felgueiras Bastos, Murillo Moreira (135) Batista, Daniela (154) Batista, Twiggy Cristina Alves Bayer, Kristina (180) Becking, Leontine E. (193) Bell, James J. Beresi, Matilde Sylvia (102) Berlinck, Roberto Gomes de Souza (203) Bert, Theresa (216) Bessa, Julia Manuela Marques Blanquer, Andrea Borchiellini, Carole Borojevic, Radovan Boury-Esnault, Nicole (227) Buckeridge, John St James Stewart (191)

E
Efremova, Sofia (42) Ehrlich, Hermann (139) Ellwood, Michael (197) Ereskovsky, Alexander V. (63) Erpenbeck, Dirk (89) Erwin, Patrick M. Esteves, Ana Isabel dos Santos Esteves, Eduardo Leal (151) Ettinger-Epstein, Piers (190)

F
Fernandez, Diana Margarita Marquez (72) Fernandez, Jlio Csar Cruz (211) Fernandez, Maria Patricia Curbelo (62) Ferreira, Elthon Gois Ferreira, Patrcia Fernandes Ferretti, Cristina (205) Freeman, Chris Frota, Mario Luiz Conte da

C
Cabral, Sergio de Andrade Calcinai, Barbara (24) Calderon, Emiliano N. (155) Campos, Maurcio Alves de (174) Caralt, Sonia de (168) Carballo, Jos Luis (142) Crdenas, Paco (146) Carvalho, Mariana de Souza (136) Cavalcanti, Bruno Colho Cavalcanti, Fernanda Fernandes (30) Cavalcanti, Guarani de Hollanda Cebrian, Emma (165) Cedro, Victor Ribeiro (188) Cerrano, Carlo (50) Cerrano, Giovanni (23) Chaves-Fonnegra, Andia (59) Chiappone, Mark Coles, Steve L. (58) Cook, Steve Correia, Monica Dorigo (6) Cortes, Javier Galiana (7)

G
Gamulin, Vera Gaspar, Helena Gil, Ana Riesgo (109) Giovine, Marco Gochfeld, Deborah (117) Goeij, Jasper Merijn de (125) Gmez, Adriana Alvizu Gonobobleva, Elisaveta L. (10) Goodwin, Claire (208) Gugel, Jochen

H
Hajdu, Eduardo (202) Hajdu, Erik (222)

683

Hajdu, Karina (218) Harper, Mary Kay (159) Hausmann, Rudolf (103) Haygood, Margo Heim, Isabel (111) Heim, Mark Hill, Malcolm (164) Hoffmann, Friederike (115) Hooper, John N.A. (54) Hoshino, Sayumi (8) Humanes, Madalena (68)

M
Magalhes, Alexandre de Oliveira Maia, Guilherme de Azevedo (48) Maldonado, Manuel (137) Manconi, Renata Manuel, Michal (131) Masuda, Yoshiki (5) McCormack, Grace (212) Mclean, Elizabeth Layli (44) Mendes, Adriana Maria Salgado (183) Monteiro, Leandro de Campos (158) Moraes, Fernando Coreixas de (2) Moraes, Francisco Leo Pardo (226) Moreira, Ana Paula Barbosa (38) Moreno, Srgio Taboada Mota, Sula Salani (98) Mukhina, Yulia Mller, Isabel M. Mller, Werner E.G. (49) Muricy, Guilherme Ramos da Silva (126)

I
Ireland, Chris (143) Ise, Yuji (94) Ivanov, Marija (83)

J
Janussen, Dorte (161) Jimenez, Paula Christine

N K
Kaandorp, Jaap A. (52) Kanagasabhapathy, Manmadhan Kauffman, Anne Kathryn (25) Kay-Nishiyama, Cynthia Kelly, Michelle (220) Kelve, Merike (46) Kijjoa, Anake (182) Kim, Hyung June (160) Klautau, Michelle Kljajic, Zoran (112) Klppel, Anne (178) Knowlton, Ann L. (97) Koopmans, Marieke (122) Kuusksalu, Anne Nicolai, Marisa Helena Fonseca Nishiyama, Gregory

O
Oliveira, Amanda Borges Martins de (184) Oliveira, Brbara Rodrigues de Oliveira, Mara Ventura de (195) Oliveira, Marcos Paulo Carvalho de (150) Osinga, Ronald (56)

P
Pansini, Cecilia Carpi Pansini, Maurizio Parma, Lorenzo (15) Parolin, Mauro Paula, Andr Figueira de Paula, Thiago Silva de (119) Pauls, Sheila M. Peixinho, Solange Perdomo, Viviane (162) Pereira, Fabio Renato Prez, Thierry Petricioli, Donat (100) Pfannkuchen, Martin (78) Pick, Kerstin Picton, Bernard (123) Pinheiro, Ulisses dos Santos (22) Pinho, Paulo Miguel Martins de (113) Pisera, Andrzej (224) Pohler, Susanne M.L. Pomponi, Shirley A. (40) Pozzolini, Marina (214)

L
Lanna, Emlio (166) Laport, Marinella Silva (39) Lavrov, Dennis V. (33) Lee, Arlene V.H. (171) Lee, Kyung Jin (148) Lee, Welton L. (172) Lejeusne, Christophe Leys, Sally (57) Liberatore, Don Lima, Gabriela Menezes do Amaral (185) Lira, Simone Possedente de Lbo-Hajdu, Gisele (221) Lopes, Daniela de Almeida (132) Lopez, Jose (170) Lpez, Pilar Ros (217) Lopp, Annika

684

R
Raleigh, Jean (206) Rangel, Marisa Rapp, Hans Tore (80) Redmond, Niamh (209) Reintamm, Tnu (73) Reis, Estfane Cardinot (149) Reiswig, Henry M. (101) Reveillaud, Julie Ribeiro, Suzi Meneses (138) Ribeiro, Venina Pires (34) Roberts, Daniel (13) Rodrguez, Javier Cristobo (219) Rodriguez, Pablo Rodrigues Dominguez (156) Rosa, Salvatore De (96) Rossi, Andr Linhares (201) Rua, Cintia Paula Jandre (31) Rhle, Sebastian (128) Russo, Claudia Augusta de Moraes (32) Rtzler, Klaus (144) Ryan, Molly K. (145)

Sym, Christiane de Azevedo (157)

T
Tabachnik, Konstantin (134) Tavares, Maria da Conceio Marques Thacker, Robert W. Thomas, Olivier P. Thoms, Carsten (177) Tokina, Daria B. (64) Tompkins, Gabrielle (17) Trindade, Amaro (152) Turon, Xavier (163)

U
Uriz, Maria-J. (108) Usher, Kayley M. (19)

V
Vacelet, Jean (223) Vaillancourt, Yvonne R. Valderrama, Diego Fernando (14) Valisano, Laura Vzquez, Adriana M. Prez Velandia, Fernando Jose Parra (127) Vilanova, Eduardo Prata Voigt, Matthias (173) Voigt, Oliver (90) Volkmer-Ribeiro, Cecilia (9) Voogd, Nicole J. de

S
Samaai, Toufiek (93) Santos, George Joaquim Garcia Santos, Josivete Pinheiro dos (153) Santos, Loyana Docio Schlacher-Hoenlinger, Monika Schlappy, Marie-Lise (11) Schmitt, Susanne (181) Schoenberg, Christine Hanna Lydia (67) Schrder, Heinz C. (4) Schupp, Peter (175) Selvin, Joseph Sidri, Marzia Silva, Carla Maria Menegola da (36) Silva, Gustavo Bastos da (140) Silva, Mara Flvia Lima da Silva, Meiryelen Vieira da (225) Silva, Teresinha Andrea da Sim, Chung Ja Sipkema, Detmer Slattery, Marc Soest, Rob van (76) Sol-Cava, Antnio Mateo (104) Stafford, David F. Stevely, John M. (213) Swain, Timothy D. (86) Sweat, Donald (215)

W
Whalan, Steve Wiens, Matthias Philipp (45) Wilke, Diego Veras Willenz, Philippe (133) Wolff, Carsten Wrheide, Gert (129) Wulff, Janie L. (124)

X
Xavier, Joana Rita Bogalho Teixeira

Z
Zea, Sven (141) Zhang, Wei (169) Zilberberg, Carla (196)