The Complete Sequence of the Mitochondrial Genome of the Crustacean

Penaeus monodon: Are Malacostracan Crustaceans More Closely Related to

Insects than to Branchiopods?
Kate Wilson, Valma Cahill, Elizabeth Ballment, and John Benzie
Australian Institute of Marine Science, Townsville, Queensland, Australia

The complete sequence of the mitochondrial genome of the giant tiger prawn, Penaeus monodon (Arthropoda,
Crustacea, Malacostraca), is presented. The gene content and gene order are identical to those observed in Dro-
sophila yakuba. The overall AT composition is lower than that observed in the known insect mitochondrial genomes,
but higher than that observed in the other two crustaceans for which complete mitochondrial sequence is available.
Analysis of the effect of nucleotide bias on codon composition across the Arthropoda reveals a trend with the
crustaceans represented showing the lowest proportion of AT-rich codons in mitochondrial protein genes. Phylo-
genetic analysis among arthropods using concatenated protein-coding sequences provides further support for the
possibility that Crustacea are paraphyletic. Furthermore, in contrast to data from the nuclear gene EF1␣, the first
complete sequence of a malacostracan mitochondrial genome supports the possibility that Malacostraca are more
closely related to Insecta than to Branchiopoda.

Introduction
Mitochondrial genome sequence and structure is
widely used to provide information on phylogenetic re-
lationships and on the genetic structure of populations
and patterns of gene flow. This information may derive
from studies of gene order, the sequences of individual
genes, restriction fragment length polymorphism
(RFLP) analysis of mtDNA, or the sequences of com-
plete genomes.
The giant tiger prawn, Penaeus monodon, is the
major species of shrimp farmed in aquaculture world-
wide. The Penaeidae fall within the order Decapoda
(class Malacostraca), which contains most commercially
important crustacean species. Analyses of penaeid mi-
tochondrial genetics have been carried out using RFLP
analysis and single-gene sequence analysis (Palumbi and
Benzie 1991; Bouchon, Souty-Grosset, and Raimond
1994; Klinbunga et al. 1999). However, there is no com-
plete mitochondrial genome sequence available for any
decapod, or even for any malacostracan crustacean. To
date, the only complete crustacean mitochondrial ge-
nome sequences available are those of two branchio-
pods: the brine shrimp Artemia franciscana (Branchio-
poda, Anostraca) (Valverde et al. 1994) and the water
flea Daphnia pulex (Branchiopoda, Cladocera) (Crease
1999). Thus, there is a dearth of mitochondrial sequence
data available for phylogenetic comparisons within the
Crustacea and related groups.
Recently, partial sequence (8,754 bp) was pub-
lished for the southern pink shrimp, Penaeus notialis
(Garcı́a-Machado et al. 1999). This sequence encom-
passed seven protein- coding genes, and the aligned
amino acid sequences of this species, Artemia, and four
insects were used to make inferences about phylogeny
Key words: Malacostraca, complete mitochondrial genome se-
quence, arthropod phylogeny, crustacean/insect relationships, penaeid
shrimp.
Address for correspondence and reprints: Kate Wilson, Australian
Institute of Marine Science, PMB 3, Townsville MC, Queensland 4810,
Australia. E-mail: k.wilson@aims.gov.au.
Mol. Biol. Evol. 17(6):863–874. 2000
䉷 2000 by the Society for Molecular Biology and Evolution. ISSN: 0737-4038
within the Arthropoda. The resulting analysis supported
a paraphyletic Crustacea, with Penaeus being more
closely related to insects than to the branchiopod crus-
tacean, Artemia. However, this analysis was based on
only partial mitochondrial sequence from P. notialis and
a single branchiopod, A. franciscana, as the complete
mitochondrial sequence of D. pulex was not then avail-
able for inclusion in the analysis. Moreover, the mtDNA
of Artemia appears to be evolving at an accelerated rate
(Crease 1999; see fig. 5), which could affect molecular
phylogenetic inference (Hillis, Moritz, and Mable 1996).
In this paper, we present the first complete mito-
chondrial sequence for a malacostracan, P. monodon.
We discuss the composition of the P. monodon mito-
chondrial genome and analyze the effect of nucleotide
bias on codon composition across the Arthropoda. Fi-
nally, we describe the results of phylogenetic analyses
among members of the Arthropoda for which complete
mitochondrial sequences are available, paying particular
attention to the relationships of the different crustacean
groups and Insecta.
Materials and Methods
Determination of Sequence
Mitochondrial DNA was extracted from pleopod
muscle from a single hatchery-grown P. monodon spec-
imen derived from parents collected near Cairns, North
Queensland, Australia, according to Dowling, Moritz,
and Palmer (1990), using 10 mM EDTA in the homog-
enization buffer. This mtDNA was cut with XbaI to pro-
duce three fragments (2.1, 3.8, and 10.0 kb), which were
cloned into pUC19. The resulting three plasmids,
AIMS-P.mon74, AIMS-P.mon75, and AIMS-P.mon76,
were sequenced using fluorescent cycle-sequencing kits
from the Applied Biosystems division of Perkin-Elmer.
Initial sequencing was carried out using vector primers,
and subsequent sequencing was performed by ‘‘primer
walking’’; i.e., primers were synthesized corresponding
to previously obtained P. monodon mitochondrial se-
quence. Sequence was obtained for both strands for the
complete genome. To determine the relative orientation

863
864 Wilson et al.
of the three clones and to ensure that no sequence was
missing due to a failure to clone very small XbaI frag-
ments, primers that faced ‘‘out’’ of the clones were de-
signed and used to amplify junction fragments from in-
tact mitochondrial DNA by the polymerase chain reac-
tion (PCR). In each case, single bands of the predicted
size were obtained. These three PCR products were also
sequenced using ABI reagents.
Analysis of Codon Bias
Potential amino acid bias due to underlying nucle-
otide bias was analyzed using the ‘‘square plot’’ analysis
of Foster, Jermiin, and Hickey (1997). In brief, this an-
alyzes the amino acid composition by looking at the
proportion of amino acids with codons that have (1) A
or T in both the first and the second positions (2) A or
T in either the first or the second position, and (3) only
G or C in these two positions. The amino acid leucine
was excluded from this analysis, as it can be encoded
by codons from two different classes (TTR and CTR)
and hence can accommodate changes in underlying base
composition without a change in encoded amino acids.
The data are presented here as a histogram rather than
as square plots to show underlying trends across the
Arthropoda.
Phylogenetic Analyses
Protein-coding sequences were individually aligned
using Pileup (Genetics Computer Group 1994) with the
gap creation penalty set to 5.0 and the gap extension
penalty set to 0.3. Sequences were from 10 of the 11
arthropods for which complete mitochondrial genome
sequences were available in GenBank: among the in-
sects (Hexapoda), the dipteran flies Drosophila yakuba
(X03240, Hexapoda, Insecta, Diptera), Drosophila me-
lanogaster (U37541, Hexapoda, Insecta, Diptera), Cer-
atitis capitata (CCA242872, Hexapoda, Insecta, Dip-
tera), Anopheles quadrimaculatus (L04272, Hexapoda,
Insecta, Diptera), and Anopheles gambiae (L20934,
Hexapoda, Insecta, Diptera) and the orthopteran Locusta
migratoria (X80245, Hexapoda, Insecta, Orthoptera);
the branchiopod crustaceans A. franciscana (X69067,
Crustacea, Branchiopoda, Anostraca) and Daphnia pu-
lex (AF117817, Crustacea, Branchiopoda, Cladocera);
and from the Chelicerata, the prostriate tick Ixodes hex-
agonus (AF081828, Chelicerata, Arachnida, Acari) and
the metastriate tick Rhipicephalus sanguineus
(AF081829, Chelicerata, Arachnida, Acari). The hon-
eybee, Apis mellifera (L06178, Hexapoda, Insecta, Hy-
menoptera), was not included in the analysis due to the
extreme nucleotide bias which confounds phylogenetic
inference based on both nucleotide and protein sequence
analyses (Galtier and Gouy 1995; Foster, Jermiin, and
Hickey 1997). The earthworm Lumbricus terrestris
(U24570, Annelida, Oligochaeta) was used as an out-
group. Internal sections of nad2 and nad4 were excluded
from the analysis due to severe ambiguities in align-
ment. In addition, all sequences were trimmed at the N-
and C-terminal ends to give an exact and unambiguous
alignment. The alignments were then concatenated to
give a single multiple sequence alignment containing
3,327 characters (2,335 variable and 1,687 parsimony-
informative) for phylogenetic analysis. The alignments
generated are available from the corresponding author
on request.
The aligned sequences were subjected to distance
analysis using MEGA 1.01 (Kumar, Tamura, and Nei
1993) and PHYLIP 3.572c (Felsenstein 1993), parsi-
mony analysis using PAUP 3.1.1 (Swofford 1993), and
maximum-likelihood analysis using MOLPHY 2.3b
(Adachi and Hasegawa 1996b). Distances between taxa
were estimated in the program MEGA using the number
of differences, p-distance, the Poisson correction, and
gamma probabilities with gamma set to 1 or 2. Trees
were then constructed using neighbor joining. Bootstrap
probabilities were estimated with 1,000 replications us-
ing pairwise deletion for gaps and missing data. Dis-
tance analyses were also carried out with the PHYLIP
package using the Dayhoff PAM matrix for amino acid
substitutions and neighbor-joining tree building. Boot-
strap probabilities were estimated in PHYLIP with 100
replicates. For parsimony analysis, the heuristic search
option in PAUP 3.1.1 was used. A simple addition se-
quence was used and tree bisection-reconnection branch
swapping was performed. Bootstrap values were esti-
mated using 1,000 replications. For both distance and
parsimony analyses, L. terrestris was defined as the out-
group. For maximum-likelihood analysis, the mtREV-24
model of amino acid substitution in mitochondrial pro-
tein-coding genes was used (Adachi and Hasegawa
1996a), and the data were analyzed using both the star
decomposition method and local rearrangement options.
For the latter, the 200 best trees using stepwise addition
were first generated, and these tree topologies were then
subjected to local rearrangements to determine the max-
imum-likelihood tree.
Results
Genome Size
Three plasmids were constructed from the three
major XbaI fragments of the P. monodon mitochondrial
genome. As the total size of these fragments was esti-
mated to be approximately 16 kb, it was assumed that
this comprised the majority of the mitochondrial ge-
nome. PCR analysis of intact mtDNA using primers that
spanned the XbaI sites at the plasmid ends revealed that
one XbaI site had been missed in the cloning strategy.
This was because there were two XbaI sites which were
almost adjacent within a region which was subsequently
identified as containing the trnG gene. These sites are
separated by only 6 bp and thus add 12 bp to the total
sequence length obtained from the plasmids, making a
total genome size of 15,984 bp. This sequence has been
deposited in GenBank (accession number AF217843).
Gene Content and Order
The gene content was identified primarily by align-
ment with the known genes from D. yakuba, D. melan-
ogaster, and A. gambiae. In addition, tRNA genes and
their anticodons were detected using the program t-

FIG. 1.—Gene order in the Penaeus monodon mitochondrial genome. tRNA genes are represented by the single-letter code for the cognate
amino acid. For Leu and Ser, for which there are two tRNAs, the codon recognized has been specified. The numbers are the numbers of bases
between genes, with negative numbers representing probable overlaps between genes (e.g., atp8 and atp6 are separated by ⫺7 bp, ATGTTAA).
RNAscan (Lowe and Eddy 1997). Like all of the other
arthropod mtDNAs, the P. monodon mitochondrial ge-
nome contains genes for 13 proteins, 22 tRNAs, and two
rRNAs. In addition, there is a 991-bp-long A⫹T-rich
region that does not contain any known genes. As the
origin of replication of the Drosophila mtDNA maps in
the corresponding region in the Drosophila mtDNA ge-
nome, this region is presumed to be the equivalent of
the vertebrate control region (Wolstenholme 1992). The
order and arrangement of the genes in P. monodon is
illustrated in figure 1. It is identical to that observed in
the insects D. melanogaster, D. yakuba, and C. capitata
and in the branchiopod D. pulex.
The genes are densely packed, as in other mito-
chondrial genomes. Excluding the 991-bp control re-
gion, there are a total of only 139 bp found between the
different genes in the P. monodon mitochondrial ge-
nome. There are three clear cases of overlapping genes,
which all occur between adjacent genes encoded on the
same strand. There is a 5-bp overlap (TCTAA) between
the last codon and the termination codon of cox1 and
trnL(taa) and a 7-bp overlap (ATGATAA) between the
termination of atp8 and the initiation of atp6 which is
seen in most arthropod mitochondrial genomes (the hon-
eybee, A. mellifera, has an even greater overlap, 19 bp);
likewise, the end of nad4L and nad4 overlap by 7 bp
(ATGTTAA) on the minority coding strand. In addition,
there are a further three cases in which an overlap could
occur if the complete stop codon present in the mtDNA
sequence is used. These are a 2-bp overlap (AA) be-
tween the potential termination codon of nad2 (TAA)
and the start of trnW, a 1-bp overlap (A) between the
termination of nad6 (TAA) and the initiation of cob, and
a 2- bp overlap (AG) between the termination of cob
(TAG) and the start of trnS(uga). However, it is also
possible that RNA processing clips the genes apart up-
stream of these potential overlaps and that polyadenyl-
ation completes the stop codon on the mRNA (Anderson
et al. 1981).
Complete mtDNA Sequence of the Shrimp P. monodon 865
Coding Sequences
The P. monodon mitochondrial genome appears to
use the same genetic code as Drosophila. It is certainly
the case that UGA specifies an amino acid rather than
a stop codon, since otherwise the majority of proteins
would show premature termination. It is also reasonable
to presume that the amino acids specified by each codon
family are the same as in Drosophila, as use of this code
gives rise to the expected open reading frames with a
high degree of similarity to the same proteins identified
from Drosophila and other arthropods. However, this
could be formally proven only by purification and se-
quencing of P. monodon mitochondrial proteins.
The location and putative initiation and termination
codons of the protein-coding sequences are shown in
table 1. Twelve out of the 13 protein coding genes ap-
pear to initiate with the codon ATN. However, as is the
case with several other arthropod species, the initiation
codon of cox1 is unclear. In P. monodon, it appears to
start with ACG based on alignment with other species.
In Drosophila, Locusta, and Daphnia, there is a tetra-
nucleotide sequence (ATAA or ATTA) immediately pre-
ceding the apparent first codon, and it has been sug-
gested that this tetranucleotide could serve as the initi-
ation codon (Clary and Wolstenholme 1983; de Bruijn
1983). However, no such sequence is present in P. mon-
odon, and we are not aware of any model indicating how
a tetranucleotide might function in this capacity. The
absence of such a tetranucleotide has also been observed
in other arthropods, e.g., A. gambiae (Beard, Mills, and
Collins 1993), leading to the conclusion that each of the
disparate array of first codons in aligned cox1 genes
most likely does serve as the initiation codon. An alter-
native model to explain these atypical start codons is
that they are converted to more conventional start co-
dons in the mRNA; for example, in the tomato, the cox1
gene starts with an ACG codon, but this is converted to
an AUG in the mRNA by RNA editing (Kadowaki et
866 Wilson et al.

Table 1
Protein-Coding Genes in the Penaeus monodon Mitochondrial Genome
Majority (J) or
Minority (N)
Gene Start Position End Positiona Start Codon End Codon Amino Acids Strand
nad2. . . . . . . . . 252 1250 ATT TAA 333 J
cox1 . . . . . . . . . 1465 3000 ACGb TAA 512 J
cox2 . . . . . . . . . 3071 3757 ATG T 229 J
atp8 . . . . . . . . . 3900 4055 ATT TAA 52 J
atp6 . . . . . . . . . 4052 4723 ATG TAA 224 J
cox3 . . . . . . . . . 4737 5525 ATG T 263 J
nad3. . . . . . . . . 5594 5944 ATG T 117 J
nad5. . . . . . . . . 8095 6374 ATA T or TA 574 N
nad4. . . . . . . . . 9512 8175 ATG TAA 446 N
nad4L . . . . . . . 9805 9509 ATG TAA 99 N
nad6. . . . . . . . . 9942 10460 ATT TAA 173 J
cob . . . . . . . . . . 10463 11596 ATG TAG 378 J
nada1. . . . . . . . 12627 11692 ATA TAA 312 N
a Designated as the last base of the codon encoding the final amino acid, i.e., excluding partial or complete stop codons.
b See discussion in the text and figure 2 regarding the initiation codon of cox1.

al. 1995). The presumed initiation context of cox1 genes
from arthropods, for which there is complete mitochon-
drial sequence, plus that of the penaeid shrimp P. no-
tialis, for which there is partial mitochondrial sequence
(GenBank accession numbers X84350–X84357), is
shown in figure 2.
Nine of 13 genes have a complete termination co-
don, either TAA (eight genes) or TAG (cytochrome b
only). The other genes all have incomplete termination
codons, T, or possibly TA for nad5 (table 1). These are
believed to be converted to complete TAA codons by
the addition of A residues during RNA processing (An-
derson et al. 1981).
tRNAs
The predicted structures of the P. monodon tRNAs
are shown in figure 3. All tRNAs have the typical clo-
verleaf structures of other mitochondrial tRNAs except
for trnS(ucg), which recognizes the codon AGN. This
latter tRNA lacks the dihydrouridine loop, as observed
for this tRNA in all other arthropod mitochondrial ge-
nomes (Wolstenholme 1992).
The anticodons are identical to those observed in
D. yakuba, except for the trnK anticodon, which is UUU
in P. monodon, in contrast to CUU in D. yakuba. The
trnK anticodon is also UUU in the related shrimp P.
notialis and in the honeybee A. mellifera (Crozier and
Crozier 1993), but it is CUU in all other arthropods
known to date. In fact, UUU is the anticodon which
would be predicted for lysine because, in most mtDNAs,
two-codon families which end in a purine have a U in
the wobble position of the anticodon. In the case of D.
yakuba and other arthropods, the CUU anticodon for
lysine (codons AAA and AAG) requires C-A pairing
(Clary and Wolstenholme 1985).
Base Composition
The A⫹T content of the P. monodon genome is
70.6% (A, 5,636; C, 2,669; G, 2,028; T, 5,651 on the
strand that contains the majority of open reading
frames). The 991-bp control region has an AT compo-
sition of 81.5%. These values are lower than those for
the published insect genomes (75.3%–84.9% for total
A⫹T content and 86.0%–96.0% for the AT-rich control
region), but higher than those for the crustaceans A.
franciscana (64.5% total AT content, 68.0% AT in the
control region) and D. pulex (62.3% total AT, 67.1% AT
in the control region). Among the chelicerates, the

FIG. 2.—Initiation context for cox1 genes from a variety of arthropods. The first five to seven codons are shown in uppercase letters. Where
the first codon is not a typical start codon for arthropod mitochondrial DNA (ATN), the upstream context is shown in lowercase letters. In the
case of the drosophilid flies, the preceding tetranucleotide is postulated to act as the initiation ‘‘codon.’’ However, while this theory might also
be applicable to the locust and to Daphnia, the upstream sequence for the Anopheles species, Ceratitis capitata, and Penaeus monodon do not
conform to this pattern. The presumed initial amino acids are shown on the right-hand side of the figure.
 Complete mtDNA Sequence of the Shrimp P. monodon 867

FIG. 3.—The inferred secondary structure of the 22 tRNAs in the Penaeus monodon mitochondrial genome. The tRNAs for Leu and Ser
are identified in the figure by the codons recognized, rather than by the anticodon present in the tRNA itself.

horseshoe crab is 67.7% A⫹T in the 8,459 bp sequenced
(Staton, Daehler, and Brown 1997), and the ticks Ixodes
and Rhipicephalus are 72.6% and 77.9% AT-rich,
respectively.
There are a total of 3,716 codons in all 13 protein-
coding genes, excluding stop codons. The overall AT
composition of protein-coding regions is 69.3% but at
third positions it is 83.7%. When the nucleotide com-
position of the third positions of fourfold-degenerate co-
dons is examined, it is 84.9% AT (see table 2). This
contrasts with much lower values of 68.7% in Artemia
and 62.9% in Daphnia. The corresponding value for the
partial sequence available for P. notialis (for the protein-
coding genes atp6, atp8, cox1, cox2, cox3, nad2, nad3,
and nad5) is 74.6%. Table 2 illustrates that these values
for the two penaeid shrimp, representing the class Ma-
lacostraca, are intermediate between those for the bran-
chiopod crustaceans and the insects. The percentage of
A⫹T in the third positions of fourfold-degenerate co-
dons for the two tick species is similar to that for the
penaeids.
The codon usage in P. monodon is shown in table
3. A feature of most arthropod genomes sequenced to
date is that the bias toward the nucleotides A and T also
868 Wilson et al.

leads to a bias in the amino acids used. This can be

Daphnia
analyzed by comparing the proportions of amino acids

pulex
17.3
26.7
36.2
19.8
62.9
37.1
with A or T versus C or G at the first and second codon
positions. Penaeus monodon appears to share the bias
of all arthropod genomes toward amino acids encoded

franciscana
by AT-rich codons (fig. 4). In P. monodon, the bias to-

Artemia

14.1
30.9
37.8
17.1
68.7
31.2
ward AT-rich codons appears to be very similar to that
observed in Artemia and Daphnia and is slightly less
than that observed in the insects. The gradation in the
proportion of codons with A or T in both of the first
Penaeus
natialis
10.6 two positions which is observed in the different taxo-
32.4
42.2
14.8
74.6
25.4
nomic groups indicates a lack of evolutionary station-
arity; i.e., there is heterogeneity in the evolutionary pro-
cess leading to differing average codon and, hence, ami-
monodon
Penaeus

no acid compositions.
6.3
38.5
46.4
8.8
84.9
15.1

Phylogenetic Analysis
quadrimaculatus

The P. monodon mitochondrial sequence represents

Aopheles

the 12th complete arthropod mitochondrial genome se-

6.0
46.5
43.0
4.5
89.5
10.5

quenced to date; there are now complete mitochondrial

sequences available for seven insects, two chelicerates,
and three crustaceans. Hence, it was of interest to see
whether mitochondrial sequence comparisons could
Aopheles
gambiae

yield any evidence on phylogenetic relationships be-

3.2
47.8
46.0
2.9
93.8
6.1

tween the different subphyla represented, particularly on

crustacean/insect relationships. As the phylogenetic
comparisons being made are between widely divergent
Ceratitis

phylogenetic groups, most sequence signal will be ob-

capitata
3.1
44.7
48.8
3.4
93.5
6.5

scured by saturation. This was indeed found to be the

case when nucleotide sequences of individual genes
(small- and large-subunit rRNA genes and cox1) were
Base Composition at the Third Codon Positions of Fourfold-Degenerate Codons

melanogaster
Drosophila Drosophila

used in phylogenetic analyses: the phylogenetic trees

produced from these data were largely unresolved
3.3
45.7
48.8
2.2
94.5
5.5

(nodes supported by bootstrap values of ⬍70% were

considered insufficiently supported; Hillis and Bull
1993). However, by aligning the amino acid sequences
yakuba

of all mitochondrially encoded proteins and using the

3.5
46.2
48.3
2.0
94.5
5.5

concatenated set of aligned sequences in phylogenetic

analysis, more robust phylogenetic trees were obtained.
The phylogenetic analysis used data from 11 of the
migratoria

12 complete arthropod mitochondrial genomes available

Locusta

2.3
54.4
40.2
3.1
94.6
5.4

(10 from the database plus P. monodon). Sequence from

the honeybee A. mellifera was omitted as the extreme
AT bias in this species makes phylogenetic analyses
even of protein-coding regions unreliable (Foster, Jer-
mellifera

miin, and Hickey 1997), as evidenced by the fact that

Apis

0.8
62.7
34.1
2.4
96.8
3.2

A. mellifera clustered together with the ticks with a

bootstrap value of 99–100 when included in a similar
phylogenetic analysis (Black and Roehrdanz 1998). The
sanguineus hexagonus
Rhipicephalus Ixodes

earthworm L. terrestris (Boore and Brown 1995) was

10.1
38.5
36.6
14.9
75.1
25.0

used as the outgroup.

Three different tree structures resulted from appli-
cation of different phylogenetic algorithms (fig. 5A–C).
In all cases, the two chelicerate species form a clade
3.5
48.7
40.2
7.5
88.9
11.0

which is quite distinct from a Hexapoda/Crustacea

clade. In the first tree structure (fig. 5A), Artemia and
Daphnia form a clade separate from Hexapoda (repre-
% AT . . . . .
% GC . . . .
G ........
A ........
T ........
C ........
Table 2

sented by the class Insecta) and P. monodon; in the sec-

%

ond (fig. 5B), all Crustacea and Hexapoda form a single

clade, with Penaeus being the crustacean most closely
 Complete mtDNA Sequence of the Shrimp P. monodon 869

Table 3
Codon Usage in 13 Protein-Coding Genes of Penaeus monodon
Phe . . . . . . . .TTT 234 Ser . . . . . . . .TCT 109 Tyr . . . . . . . .TAT 113 Cys . . . . . . . .TGT 32
TTC 71 TCC 26 TAC 30 TGC 9
Leu . . . . . . . .TTA 340 TCA 89 End. . . . . . . .TAA 8 Trp . . . . . . . .TGA 88
TTG 40 TCG 13 TAG 1 TGG 14
Leu . . . . . . . .CTT 91 Pro . . . . . . . .CCT 89 His . . . . . . . .CAT 58 Arg . . . . . . . .CGT 22
CTC 21 CCC 13 CAC 27 CGC 2
CTA 74 CCA 36 Gln . . . . . . . .CAA 69 CGA 34
CTG 19 CCG 6 CAG 7 CGG 3
Ile . . . . . . . . .ATTa 250 Thr . . . . . . . .ACT 92 Asn. . . . . . . .AAT 106 Ser . . . . . . . .AGT 43
ATC 48 ACC 19 AAC 35 AGC 8
Met. . . . . . . .ATA 192 ACA 68 Lys . . . . . . . .AAA 70 AGA 71
ATG 26 ACG 16 AAG 10 AGG 0
Val . . . . . . . .GTT 100 Ala . . . . . . . .GCT 129 Asp. . . . . . . .GAT 58 Gly . . . . . . . .GGT 109
GTC 20 GCC 35 GAC 17 GGC 5
GTA 131 GCA 60 Glu . . . . . . . .GAA 65 GGA 87
GTG 11 GCG 6 GAG 17 GGG 33
a Three of the ATT codons are likely to specify methionine (Fearnley and Walker 1987), as they appear to act as the initiation codons of nad2, atp8, and nad6.

related to the Hexapoda and Artemia being the most
distantly related. In the third tree, Penaeus and Hexap-
oda form one clade, and the relationship of Daphnia and
Artemia to each other and to this clade remains unre-
solved. The most striking feature held in common by all
three trees is the paraphyly of the Crustacea, with Pen-
aeus being more closely related to the insects than to
the branchiopods. This relationship of Penaeus to In-
secta is supported by bootstrap values of ⬎90% with all
tree-building methods.
Discussion
The length of the P. monodon mitochondrial ge-
nome falls within the range observed for most animal
mitochondrial genomes, including those of the other ar-
thropods for which complete sequences are available:
FIG. 4.—Amino acid content of protein-coding genes in arthropod
mitochondrial genomes is affected by codon bias. The histogram rep-
resents the proportion of amino acids for codons that have (1) A or T
in both the first and the second positions (AT/AT), (2) A or T in either
the first or the second position (AT/CG or CG/AT), or (3) only G or
C in these two positions (CG/CG). Leucine and termination codons
are excluded. All arthropod taxa for which complete mitochondrial
genome sequences are available are included.
the insects D. yakuba (16,019 bp) (Clary and Wolsten-
holme 1985), D. melanogaster (19,517 bp) (Lewis, Farr,
and Kaguni 1995), A. gambiae (15,363 bp) (Beard,
Mills, and Collins 1993), A. quadrimaculatus (15,455
bp) (Cockburn, Mitchell, and Seawright 1990), C. cap-
itata (15,980 bp) (Spanos et al. 2000), A. mellifera
(16,343 bp) (Crozier and Crozier 1993), and L. migra-
toria (15,722 bp) (Flook, Rowell, and Gellissen 1995),
the crustaceans A. franciscana (15,822 bp) (Valverde et
al. 1994) and D. pulex (15,333 bp) (Crease 1999), and
the chelicerates I. hexagonus (14,539 bp) and R. san-
guineus (14,710 bp) (Black and Roehrdanz 1998).
The gene order in the P. monodon mitochondrial
genome is identical to that of the three dipteran species
D. yakuba, D. melanogaster, and C. capitata, as well as
to the branchiopod D. pulex. All other arthropod
mtDNAs for which complete mitochondrial sequences
are available show distinct differences in the placement
of certain genes. The two Anopheles species differ in
the placement of two tRNAs, Locusta in the placement
of one tRNA, Apis in the placement of eight tRNAs,
Artemia in the placement of the block of trnI to trnQ,
and Ixodes in the placement of one tRNA, and Rhipi-
cephalus shows major rearrangements. In the latter spe-
cies, the block of genes from nad1 to trnQ appears to
have transposed, and there are two further tRNA
rearrangements.
Boore et al. (1995), and Boore and Brown (1998),
and Boore, Lavrov, and Brown (1998) argue that rear-
rangements of metazoan mtDNA are highly unlikely to
be duplicated by convergences and hence can be used
to infer phylogenetic lineages. These authors therefore
conclude that the Crustacea and Insecta form a sister
group, with the Chelicerata and Myriapoda occupying
separate arthropod lineages (Boore, Lavrov, and Brown
1998). The arrangement of genes in the P. monodon
genome is in accord with this hypothesis. This arrange-
ment is identical to that observed in the crustacean D.
pulex and appears to be also conserved in the partially
characterized genomes of the malacostracan crustaceans
Homarus americanus (Boore et al. 1995) and P. notialis
870 Wilson et al.
FIG. 5.—Phylogenetic trees based on analysis of concatenated,
aligned protein-coding sequences from 13 mitochondrial coding genes.
For all parsimony and distance trees, only nodes with ⬎70% bootstrap
support have been retained (Hillis and Bull 1993). All trees are rooted
with L. terristris as outgroup. A, Tree derived from maximum-likeli-
hood analysis; the same tree was obtained using both star decompo-
sition and local rearrangement options. Bootstrap probabilities shown
are local bootstrap probabilities for the subsequent node estimated by
the RELL method with 1,000 replications (Adachi and Hasegawa
1996b). The total branch length of the tree is 349.73, and the log
likelihood of the data given this tree is ⫺48,236.81. B, Tree obtained
using parsimony analysis. A single tree of 8,282 steps was found with
a consistency index excluding uninformative characters of 0.761, a
homoplasy index excluding uninformative characters of 0.239, and a
retention index of 0.561. Similar trees were obtained from the neigh-
bor-joining algorithm in the program MEGA using distance matrices
generated using either number of differences or p-distances. C, Tree
obtained using distance analysis, as implemented in the PHYLIP pro-
gram. Bootstrap values (100 replicates) are shown. Similar trees were
obtained from the neighbor-joining algorithm in the program MEGA
using distance matrices generated using either the Poisson correction
or the gamma parameter (␥ ⫽ 1 or ␥ ⫽ 2).
(Garcı́a-Machado et al. 1999). Taken together, these data
provide further support for the proposed relationship be-
tween Crustacea and Insecta (Boore, Lavrov, and Brown
1998), and the conclusion that the drosophilid mito-
chondrial gene order represents the ancestral crustacean/
insect mitochondrial gene order (Crease 1999).
Phylogenetic relationships among the major arthro-
pod groups were also investigated by sequence compar-
isons using concatenated protein-coding sequences. The
resulting trees all gave strong support for several con-
clusions regarding arthropod phylogeny, although it
should be noted that these trees are necessarily based on
the restricted set of arthropod taxa for which complete
mitochondrial sequences are currently available. In par-
ticular, the available sequences only represent three of
the four extant arthropod subphyla, with no complete
mitochondrial sequence from the Myriapoda being
available in the database.
The first conclusion is that the Chelicerata form a
separate clade from Crustacea/Hexapoda. This appears
to be widely supported by molecular evidence on gene
sequence (Friedrich and Tautz 1995; Black and Roehr-
danz 1998), on mitochondrial gene order (Boore, Lav-
rov, and Brown 1998), and on the molecular basis of
development (Averof and Akam 1995). It is also a view
held by most morphological taxonomists (e.g., Snod-
grass 1938; Kukalová-Peck 1992; Wheeler, Cartwright,
and Hayashi 1993), although there have been proposals
that have placed crustaceans closer to chelicerates
(Schram and Emerson 1991; Briggs, Fortey, and Willis
1992).
The second conclusion is that Crustacea are a par-
aphyletic group. This contradicts many phylogenetic
trees proposed on the basis of morphological data
(Snodgrass 1938; Wheeler, Cartwright, and Hayashi
1993). However, the concept of crustacean paraphyly
has been proposed previously based on both morpho-
logical and molecular characters (Briggs and Fortey
1989; Averof and Akam 1995), although the details of
the paraphyly may vary (see discussion below). The
conclusion that Crustacea is paraphyletic with respect to
Hexapoda was also drawn by Garcı́a-Machado et al.
(1999) in their analysis of partial mitochondrial se-
quence of P. notialis.
The third conclusion relates to the relationship of
the two classes of crustaceans that are now represented
by complete mitochondrial genome sequences, the Ma-
lacostraca and the Branchiopoda, to each other and to
the Hexapoda, represented here by the class Insecta. A
corollary of several theories supporting the idea that the
Crustacea are paraphyletic is that Hexapoda derive from
a crustacean ancestor. If this is the case, then it is an
inescapable conclusion that some Crustacea will be
more closely related to Hexapoda than to other crusta-
cean groups. This theory is consistent with the trees gen-
erated in the present analysis, which indicate that P.
monodon is related to the Hexapoda (Insecta), with
bootstrap support ranging from 92% to 100%. The exact
relationship of the two branchiopods both to each other
and to the Penaeus/insect clade is unclear, but in all
cases our analysis indicates that the branchiopods are
less closely related to insects than is the malacostracan
P. monodon. This result was also obtained in the anal-
ysis of Garcı́a-Machado et al. (1999), although their
analysis included only one branchiopod, A. franciscana,
which appears to be subject to accelerated evolution
(Crease 1999; see long branch lengths for Artemia in
fig. 5).
This conclusion on the relative relationships of Ma-
lacostraca/Branchiopoda/Insecta contradicts the results
of two other molecular studies that included represen-

tatives of both crustacean classes. In the first, a com-
bined analysis of the 18S and 28S nuclear ribosomal
RNA genes gave strong support for a single crustacean/
hexapod clade and weak support for a possible closer
relationship of the branchiopod Artemia salina to Hex-
apoda than the malacostracan Procambarus clarkii
(Friedrich and Tautz 1995). In a more wide-ranging
analysis using protein-coding sequence from the nuclear
gene encoding EF1␣, Regier and Schultz (1997, 1998)
propose that Crustacea is polyphyletic, with Branchio-
poda forming a single clade with Hexapoda and Cheli-
cerata being more closely related to this clade than
Malacostraca.
The reasons for the disparity in phylogenetic trees
obtained with the analysis of different molecules are un-
clear. One possibility is that the heterogeneity in the
evolutionary processes occurring in the different arthro-
pod lineages leads to a misleading result. This was cer-
tainly the case in the results of Black and Roehrdanz
(1998), who carried out a similar analysis using concat-
enated protein sequences which placed A. mellifera in
the same clade as the hard ticks with 99% bootstrap
support. The reason for the latter result is the extreme
AT bias in the Apis mitochondrial genome, which has a
significant influence on the resulting amino acid com-
position of the encoded proteins (see fig. 4). However,
as figure 4 illustrates, although this bias is evident in
the P. monodon mitochondrial genome, it is not as ex-
treme as in the case of Apis or, indeed, any of the other
insects. In fact, the degree to which amino acid com-
position is affected by codon bias appears to be no dif-
ferent for P. monodon than for A. franciscana, and only
slightly higher than for D. pulex. Hence, it is unlikely
to affect the relative positions of these three species on
a phylogenetic tree with respect to each other or to In-
secta. In addition, the proposed phylogenetic tree struc-
ture is also strongly supported by maximum-likelihood
analysis. This method, although still susceptible to pro-
ducing erroneous phylogenetic trees in cases of extreme
bias in nucleotide or amino acid composition, is far less
sensitive to small compositional differences than are dis-
tance and parsimony analyses (Galtier and Gouy 1995).
A more likely explanation for the different trees is
that each of these molecular phylogenies is based on
analysis of only one or two genes or, in the case of the
present paper, a set of genetically linked genes. There is
always a danger that such analyses may give misleading
results due to factors such as insufficient informative
characters or convergent evolution.
A more far-fetched, but not impossible, scenario is
that some horizontal transfer of either nuclear or mito-
chondrial sequences has occurred from insects to crus-
taceans or vice versa. As increasing amounts of data
emerge from complete genome sequences, evidence
suggests that the evolutionary origins of different groups
of genes may differ. For example, Rivera et al. (1998)
recently demonstrated that some groups of genes in eu-
karyotes are derived from archaebacteria, whereas other
groups appear to derive primarily from proteobacteria,
leading to the conclusion that there may have been con-
Complete mtDNA Sequence of the Shrimp P. monodon 871
siderable horizontal gene transfer in the evolutionary
process.
If the phylogeny suggested by this analysis of mi-
tochondrial sequences is correct, this would have strong
implications for the polarity of evolution of morpholog-
ical characters used in the taxonomy of Crustacea and
Hexapoda. In particular, it would imply that characters
which are used to distinguish crustaceans from hexapods
have been secondarily lost in hexapods, rather than be-
ing autapomorphies of Crustacea. However, as the Crus-
tacea are notoriously diverse in body form (Schram
1986) and the characters that serve to uniquely define
them are few, it is possible to develop plausible scenar-
ios by which this may have happened. In fact, evolu-
tionary scenarios which would conform with the pro-
posed phylogeny have been already developed on both
morphological and molecular-developmental grounds
(Averof and Akam 1995; Telford and Thomas 1995).
For example, one character which has been used to
define Crustacea is the structure of the head. It has been
described as having five segments, two preoral segments
bearing the first and second antennae and three postoral
segments bearing, respectively, the mandibles and the
first and second maxillae (Cisne 1982; Schram 1986).
Insects, in contrast, have been described as having six
segments in the head, three preoral and three postoral
(Lawrence, Nielsen, and Mackerras 1991). However, the
nature of the most anterior portion of the crustacean
head was never clear (see Averof and Akam 1995), and
recent work indicates that both crustaceans and insects
have six cephalic segments, as revealed by looking at
the expression of genes that specify segmental identity
(Popadic et al. 1998; Scholtz 1998). Another commonly
cited character defining Crustacea is the presence of a
second antenna on the second appendage-bearing seg-
ment of the head. However, it is easy to envisage sec-
ondary loss of such a structure in Insecta, and, indeed,
some insects develop small appendages on this segment
during the embryonic stage which could be vestiges on
an earlier adult appendage (Tamarelle 1984).
A third set of characters often deemed to be unique
to Crustacea relate to embryology and larval develop-
ment, namely, the nature of the embryonic fate map and
the formation of a nauplius larva (larvae with three sets
of appendages used for propulsion, corresponding to the
antennules, antennae, and mandibles) or passage through
an egg-nauplius phase (Anderson 1982; Schram 1986).
However, more recent studies have contested Anderson’s
(1982) claim that crustacean egg cleavage is spiral
(Hertzler and Clark 1992), and the wisdom of using a
character as variable as fate maps in phylogenetic anal-
ysis has also been questioned (Averof and Akam 1995).
Moreover, as larval development has clearly undergone
considerable diversification within Hexapoda, ranging
from ametabolous insects that show no marked meta-
morphosis at all, to hemimetabolous insects with a grad-
ual larval-to-adult transition, to the holometabolous in-
sects which show complete metamorphosis from larval
to adult forms (Chapman 1991), it is certainly possible
to envisage that the nauplius larval stage has been sec-
ondarily lost.
872 Wilson et al.
Other features which in the past have been used to
argue that Crustacea and Hexapoda occupy distinct evo-
lutionary lineages, namely, the origin of the mandibles
(gnathobasic vs. whole-limb) and the nature of the limbs
(uniramous vs. bi- or polyramous) have already effec-
tively been discounted as being only apparent, rather
than real, differences by a number of authors (Kukalová-
Peck 1992; Averof and Akam 1995; Telford and Thom-
as 1995; Scholtz, Mittmann, and Gerberding 1998).
The proposed phylogeny also has implications for
the relative affinities of different crustacean classes to
each other and to Insecta. The divergence between the
branchiopod clade and the malacostracan/insect clade is
consistent with the fossil record. Although their exact
affinities may be disputed, fossils with substantial sim-
ilarities to both Branchiopoda and Malacostraca are
known from the Cambrian (Dahl 1983; Briggs 1983),
and these classes were certainly established as indepen-
dent lineages well before the appearance of the first fos-
sil insects in the Devonian (Kukalová-Peck 1991).
Moreover, there are certain features held in common be-
tween Malacostracan crustaceans and insects that are not
shared with ‘‘simpler crustaceans,’’ most notably the de-
tailed structures of the compound eye and of the seg-
mental nervous systems (Averof and Akam 1995; Tel-
ford and Thomas 1995; Osorio, Averof, and Bacon
1995). The former has been alluded to as ‘‘one of the
most disconcerting problems of arthropod phylogeny’’
by Tiegs and Manton (1958), who argued that Arthrop-
oda were polyphyletic and hence that the malacostracan
and insect eyes had evolved by convergence.
A question that the present paper cannot address is
the relative position of the myriapods on the proposed
phylogenetic tree, as no complete myriapod mitochon-
drial gene sequences were publicly available at the time
of writing. Traditionally, it has been assumed that Hex-
apoda evolved from a myriapodlike ancestor and that
Crustacea formed a sister group to this ‘‘Atelocerate
clade’’ (Kristensen 1991). However, almost all molecu-
lar analyses indicate that Crustacea, rather than Myri-
apoda, are sister to Insecta (Wheeler, Cartwright, and
Hayashi 1993; Friedrich and Tautz 1995; Boore, Lavrov,
and Brown 1998), and arguments have also been put
forward on morphological grounds, taking into account
key features of Myriapoda, Hexapoda, and Crustacea,
that make this scenario plausible (Averof and Akam
1995; Osorio, Averof, and Bacon 1995; Telford and
Thomas 1995).
One conclusion that is clear from the present paper
and the recent completion of the D. pulex mitochondrial
genome sequence (Crease 1999) is that while identity of
mitochondrial gene order may be a strong indicator of
phylogenetic relatedness due to the low probability of
convergence (Boore and Brown 1998), differences in
mitochondrial gene order do not necessarily reflect ma-
jor phylogenetic differences. This is illustrated in the
Arthropoda by the fact that two widely divergent crus-
taceans and three dipteran flies all share the same mi-
tochondrial gene order, while those of other dipterans
and hexapods differ in the placement of one or more
tRNAs. Most strikingly, analysis of the mitochondrial
genome of two hard ticks, which lie within the same
chelicerate family, has revealed major mitochondrial re-
arrangements in four out of five subfamilies within that
group (Black and Roehrdanz 1998; Campbell and Bark-
er 1998). Thus, the various tRNA rearrangements pre-
sent in the mitochondrial genomes of Anopheles, Lo-
custa, Apis, and Artemia all appear to be characters that
were derived after the initial radiation of hexapods and
crustaceans.
In summary, the complete sequence of the mala-
costracan crustacean P. monodon gives further weight
to the conclusion that the mtDNA gene order observed
in D. yakuba is the ancestral crustacean/hexapod ar-
rangement. Phylogenetic analysis comparing available
arthropod mitochondrial genomes indicates a sister
group relationship between Malacostraca and Insecta
and hence throws further fat on the controversial fire of
arthropod phylogeny. However, a more in-depth under-
standing of arthropod phylogeny will require the com-
pletion both of the sequence of a greater variety of ar-
thropod mitochondrial genomes, most notably with the
inclusion of some myriapod genomes, as well as a better
understanding of the processes of mitochondrial
evolution.
Acknowledgments
The authors thank Lars Jermiin for his assistance
with the analysis of codon bias in arthropod genomes
and with the maximum- likelihood analysis, and Ger-
hard Scholtz for discussions of crustacean phylogeny.
We also thank two anonymous reviewers for their help-
ful comments on the manuscript. This is contribution
number 994 of the Australian Institute of Marine
Science.
LITERATURE CITED
ADACHI, J., and M. HASEGAWA. 1996a. Model of amino acid
substitution in proteins encoded by mitochondrial DNA. J.
Mol. Evol. 42:459–468.
———. 1996b. MOLPHY version 2.3: programs for molecular
phylogenetics based on maximum likelihood. Compu. Sci.
Monogr. 28:1–150.
ANDERSON, D. T. 1982. Embryology. Pp. 1–41 in L. G. ABELE,
ed. Embryology, morphology and genetics. Academic
Press, New York.
ANDERSON, S., A. T. BANKIER, B. G. BARRELL et al. (11 co-
authors). 1981. Sequence and organization of the human
mitochondrial genome. Nature 290:457–470.
AVEROF, M., and M. AKAM. 1995. Insect-crustacean relation-
ships: insights from comparative developmental and molec-
ular studies. Philos. Trans. R. Soc. Lond. B Biol. Sci. 347:
293–303.
BEARD, C. B., H. MILLS, and F. H. COLLINS. 1993. The mito-
chondrial genome of the mosquito Anopheles gambiae:
DNA sequence, genome organization, and comparisons
with mitochondrial sequences of other insects. Insect Mol.
Biol. 2:103–124.
BLACK, W., and R. ROEHRDANZ. 1998. Mitochondrial gene or-
der is not conserved in arthropods: prostriate and metas-
triate tick mitochondrial genomes. Mol. Biol. Evol. 15:
1772–1785.

BOORE, J. L., and W. M. BROWN. 1995. Complete DNA se-
quence of the mitochondrial DNA of the annelid worm
Lumbricus terrestris. Genetics 141:305–319.
———. 1998. Big trees from little genomes: mitochondrial
gene order as a phylogenetic tool. Curr. Opin. Genet. Dev.
8:668–674.
BOORE, J. L., T. M. COLLINS, D. STANTON, L. L. DAEHLER,
and W. M. BROWN. 1995. Deducing the pattern of arthropod
phylogeny from mitochondrial DNA rearrangements. Na-
ture 376:163–165.
BOORE, J. L., D. V. LAVROV, and W. M. BROWN. 1998. Gene
translocation links insects and crustaceans. Nature 392:667–
668.
BOUCHON, D., C. SOUTY-GROSSET, and R. RAIMOND. 1994.
Mitochondrial DNA variation and markers of species iden-
tity in two penaeid shrimp species: Penaeus monodon Fa-
bricius and P. japonicus Bate. Aquaculture 127:131–144.
BRIGGS, D. E. G. 1983. Affinities and early evolution of the
Crustacea: the evidence of the Cambrian. Pp. 1–22 in F. R.
SCHRAM, ed. Crustacean phylogeny. A. A. Balkema, Rot-
terdam, the Netherlands.
BRIGGS, D. E. G., and R. A. FORTEY. 1989. The early radiation
and relationships of the major arthropod groups. Science
246:241–243.
BRIGGS, D. E. G., R. A. FORTEY, and M. A. WILLIS. 1992.
Morphological disparity in the Cambrian. Science 256:
1670–1673.
CAMPBELL, N. J. H., and S. C. BARKER. 1998. An unprece-
dented major rearrangement in an arthropod mitochondrial
genome. Mol. Biol. Evol. 15:1786–1787.
CHAPMAN, R. F. 1991. General anatomy and function. Pp. 33–
67 in C.S.I.R.O. Division of Entomology, ed. The insects
of Australia. A textbook for students and research workers.
Melbourne University Press, Melbourne, Australia.
CISNE, J. L. 1982. Origin of the Crustacea. Pp. 65–92 in L. G.
ABELE, ed. The biology of Crustacea: systematics, the fossil
record and biogeography. Academic Press, New York.
CLARY, D. O., and D. R. WOLSTENHOLME. 1983. Genes for
cytochrome c oxidase subunit I, URF2, and three tRNAs in
Drosophila mitochondrial DNA. Nucleic Acids Res. 11:
6859–6872.
———. 1985. The mitochondrial molecule of Drosophila yak-
uba: nucleotide sequence, gene organisation, and genetic
code. J. Mol. Evol. 22:252–271.
COCKBURN, A. F., S. E. MITCHELL, and J. A. SEAWRIGHT. 1990.
Cloning of the mitochondrial genome of Anopheles quad-
rimaculatus. Arch. Insect Biochem. Physiol. 14:31–36.
CREASE, T. J. 1999. The complete sequence of the mitochon-
drial genome of Daphnia pulex (Cladocera: Crustacea).
Gene 233:89–99.
CROZIER, R. H., and Y. C. CROZIER. 1993. The mitochondrial
genome of the honeybee Apis mellifera: complete sequence
and genome organization. Genetics 133:97–117.
DAHL, E. 1983. Malacostracan phylogeny and evolution. Pp.
189–212 in F. R. SCHRAM, ed. Crustacean phylogeny. A. A.
Balkema, Rotterdam, the Netherlands.
DE BRUIJN, M. H. L. 1983. Drosophila melanogaster mito-
chondrial DNA, a novel organization and genetic code. Na-
ture 304:234–241.
DOWLING, T. E., C. E. MORITZ, and J. D. PALMER. 1990. Nu-
cleic acids II: restriction site analysis. Pp. 250–317 in D.
M. HILLIS, C. MORITZ, and J. D. PALMER, eds. Molecular
systematics. Sinauer, Sunderland, Mass.
FEARNLEY, I. M., and J. E. WALKER. 1987. Initiation codons
in mammalian mitochondria; differences in genetic code in
the organelle. Biochemistry 26:8247–8251.
Complete mtDNA Sequence of the Shrimp P. monodon 873
FELSENSTEIN, J. 1993. PHYLIP (phylogeny inference package).
Version 3.57c. Distributed by the author, Department of Ge-
netics, University of Washington, Seattle.
FLOOK, P., C. ROWELL, and G. GELLISSEN. 1995. The sequence,
organization and evolution of the Locusta migratoria mi-
tochondrial genome. J. Mol. Evol. 41:928–941.
FOSTER, P. G., L. S. JERMIIN, and D. A. HICKEY. 1997. Nucle-
otide composition bias affects amino acid content in pro-
teins coded by animal mitochondria. J. Mol. Evol. 44:282–
288.
FRIEDRICH, M., and D. TAUTZ. 1995. Ribosomal DNA phylog-
eny of the major extant arthropod classes and the evolution
of myriapods. Nature 376:165–167.
GALTIER, N., and M. GOUY. 1995. Inferring phylogenies from
DNA sequences of unequal base compositions. Proc. Natl.
Acad. Sci. USA 92:11317–11321.
GARCÍA-MACHADO, E., M. PEMPERA, N. DENNEBOUY, M. OLI-
VA-SUAREZ, J.-C. MOUNOLOU, and M. MONNEROT. 1999.
Mitochondrial genes collectively suggest the paraphyly of
Crustacea with respect to Insecta. J. Mol. Evol. 49:142–149.
GENETICS COMPUTER GROUP. 1994. Program manual for the
Wisconsin package. Version 8. Genetics Computer Group,
Madison, Wis.
HERTZLER, P. L., and W. H. CLARK. 1992. Cleavage and gas-
trulation in the shrimp Sicyonia ingentis: invagination is
accompanied by oriented cell division. Development 116:
127–140.
HILLIS, D. M., and J. J. BULL. 1993. An empirical test of boot-
strapping as a method for assessing confidence in phylo-
genetic analysis. Syst. Biol. 42:182–192.
HILLIS, D. M., C. MORITZ, and B. K. MABLE. 1996. Molecular
systematics. Sinauer, Sunderland, Mass.
KADOWAKI, K.-I., K. OZAWA, S. KAZAMA, N. KUBO, and T.
AKIHAMA. 1995. Creation of an initiation codon by RNA
editing in the cox1 transcript from tomato mitochondria.
Curr. Genet. 28:415–422.
KLINBUNGA, S., D. J. PENMAN, B. J. MCANDREW, and A. TAS-
SANAKAJON. 1999. Mitochondrial DNA diversity in three
populations of the giant tiger shrimp, Penaeus monodon.
Mar. Biotechnol. 1:113–121.
KRISTENSEN, N. P. 1991. Phylogeny of extant hexapods. Pp.
125–140 in C. Division of Entomology, ed. The insects of
Australia. A textbook for students and research workers.
Melbourne University Press, Melbourne, Australia.
KUKALOVÁ-PECK, J. 1991. Fossil history and the evolution of
hexapod structures. Pp. 141–179 in C.S.I.R.O. Division of
Entomology, ed. The insects of Australia. A textbook for
students and research workers. Melbourne University Press,
Melbourne, Australia.
———. 1992. The ‘Uniramia’ do not exist: the ground plan
of the Pterygota as revealed by Permian Diaphanopterodea
from Russia (Insecta: Paleodictyopteroidea). Can. J. Zool.
70:236–255.
KUMAR, S., K. TAMURA, and M. NEI. 1993. MEGA: molecular
evolutionary genetics analysis. Version 1.0. The Pennsyl-
vania State University, Philadelphia.
LAWRENCE, J. F., E. S. NIELSEN, and I. M. MACKERRAS. 1991.
Skeletal Anatomy. Pp. 3–32 in C. Division of Entomology,
ed. The insects of Australia. A textbook for students and
research workers. Melbourne University Press, Melbourne,
Australia.
LEWIS, D., C. FARR, and L. KAGUNI. 1995. Drosophila melan-
ogaster mitochondrial DNA: completion of the nucleotide
sequence and evolutionary comparisons. Insect Mol. Biol.
4:263–278.
874 Wilson et al.
LOWE, T., and S. EDDY. 1997. tRNAscan-SE: a program for
improved detection of transfer RNA genes in genomic se-
quence. Nucleic Acids Res. 25:955–964.
OSORIO, D., M. AVEROF, and J. P. BACON. 1995. Arthropod
evolution: great brains, beautiful bodies. Trends Ecol. Evol.
10:449–454.
PALUMBI, S. R., and J. A. H. BENZIE. 1991. Large mitochon-
drial DNA differences between morphologically similar
Penaeid shrimp. Mol. Mar. Biol. Biotechnol. 1:27–34.
POPADIC, A., A. ABZHANOV, D. ROUSCH, and T. KAUFMAN.
1998. Understanding the genetic basis of morphological
evolution: the role of homeotic genes in the diversification
of the arthropod bauplan. Int. J. Dev. Biol. 42:453–461.
REGIER, J. C., and J. W. SHULTZ. 1997. Molecular phylogeny
of the major arthropod groups indicates polyphyly of crus-
taceans and a new hypothesis for the origin of hexapods.
Mol. Biol. Evol. 14:902–903.
———. 1998. Molecular phylogeny of arthropods and the sig-
nificance of the Cambrian ‘‘explosion’’ for molecular sys-
tematics. Am. Zool. 38:918–928.
RIVERA, M. C., R. JAIN, J. E. MOORE, and J. A. LAKE. 1998.
Genomic evidence for two functionally distinct gene clas-
ses. Proc. Natl. Acad. Sci. USA 95:6239–6244.
SCHOLTZ, G. 1998. Cleavage, germ band formation and head
segmentation: the ground pattern of the Euarthropoda. Pp.
317–332 in R. A. FORTEY and R. H. THOMAS, eds. Arthro-
pod phylogeny. Chapman and Hall, London.
SCHOLTZ, G., B. MITTMANN, and M. GERBERDING. 1998. The
pattern of Distal-less expression in the mouthparts of crus-
taceans, myriapods and insects: new evidence for a gnatho-
basic mandible and the common origin of Mandibulata. Int.
J. Dev. Biol. 42:801–810.
SCHRAM, F. R. 1986. Crustacea. Oxford University Press, New
York and Oxford.
SCHRAM, F., and M. J. EMERSON. 1991. Arthropod pattern the-
ory: a new approach to arthropod phylogeny. Mem. Qld.
Mus. 31:1–18.
SNODGRASS, R. 1938. Evolution of the Annelida, Onychophora
and Arthropoda. Smithson. Misc. Coll. 97:1–159.
SPANOS, L., G. KOUTROUMBAS, M. KOTSYFAKIS, and C. LOUIS.
2000. The mitochondrial genome of the Mediterranean
fruitfly, Ceratitis capitata. Insect Mol. Biol. 9:139–144.
STATON, J. L., L. L. DAEHLER, and W. M. BROWN. 1997. Mi-
tochondrial gene arrangement of the horseshoe crab Limulus
polyphemus L.: conservation of major features among ar-
thropod classes. Mol. Biol. Evol. 14:867–874.
SWOFFORD, D. L. 1993. PAUP: phylogenetic analysis using
parsimony. Version 3.1.1. Illinois Natural History Survey,
Champaign.
TAMARELLE, M. 1984. Transient rudiments of second antennae
on the ‘‘intercalary’’ segment of embryos of Anurida mar-
itima Guer. (Collembola: Arthropleona) and Hyphantria cu-
nea Drury (Lepidoptera: Arctiidae). Int. J. Insect Morphol.
Embryol. 13:331–336.
TELFORD, M. J., and R. H. THOMAS. 1995. Demise of the Ateo-
cerata? Nature 376:123–124.
TIEGS, O. W., and S. M. MANTON. 1958. The evolution of the
Arthropoda. Biol. Rev. 33:255–337.
VALVERDE, J. R., B. BATUECAS, C. MORATILLA, R. MARCO,
and R. GARESSE. 1994. The complete mitochondrial DNA
sequence of the crustacean Artemia franciscana. J. Mol.
Evol. 39:400–408.
WHEELER, W., P. CARTWRIGHT, and C. HAYASHI. 1993. Ar-
thropod phylogeny: a combined approach. Cladistics 9:1–
39.
WOLSTENHOLME, D. R. 1992. Animal mitochondrial DNA:
structure and evolution. Int. Rev. Cytol. 141:173–216.
B. FRANZ LANG, reviewing editor
Accepted February 1, 2000