Plant Cell, Tissue and Organ Culture 69: 215231, 2002.

2002 Kluwer Academic Publishers. Printed in the Netherlands.

215

Review

Temporary immersion systems in plant micropropagation

H. Etienne1,2, & M. Berthouly3

1 Centrede Cooperation Internationale en Recherche Agronomique pour le Developpement-Cultures Perennes
(CIRAD-CP), CIRAD, TA 80 / PS3, Boulevard de la Lironde, 34398 Montpellier Cedex 5, France; 2 Centro de
Agronoma Tropical de Investigacion e Enseanza (CATIE), CATIE, Apdo 11, 7170 Turrialba, Costa Rica; 3 Centre
de Cooperation Internationale en Recherche Agronomique pour le Developpement Amis (CIRAD-AMIS), CIRAD,
TA 40/03, Avenue Agropolis, 34398 Montpellier Cedex 5, France ( requests for offprints; Fax: +33-467 61 71 20;
E-mail: herve.etienne@cirad.fr)

Received 22 February 2001; accepted in revised form 3 December 2001

Abstract
Temporary immersion systems for plant micropropagation have been described and grouped into 4 categories
according to operation: tilting and rocker machines; complete immersion of plant material and renewal of the
nutrient medium; partial immersion and a liquid nutrient renewal mechanism; complete immersion by pneumatic
driven transfer of liquid medium and without nutrient medium renewal. The positive effects of temporary immer-
sion on micropropagation are indicated for shoot proliferation and microcuttings, microtuberization and somatic
embryogenesis. Immersion time, i.e. duration or frequency, is the most decisive parameter for system efficiency.
Optimizing the volume of nutrient medium and the volume of the culture container also substantially improves
efficacy, especially for shoot proliferation. Temporary immersion also generally improves plant material quality. It
results in increased shoot vigour and in the frequency of morphologically normal somatic embryos. Hyperhydricity,
which seriously affects cultures in liquid medium, can be eliminated with these culture systems or controlled by
adjusting the immersion times. Plant material propagated by temporary immersion can perform better during the
acclimatization phase than material obtained on semi-solid or in liquid media. Successful regeneration of plants,
after direct sowing on soil of Solanum tuberosum microtubers and Coffea arabica somatic embryos produced in
temporary immersion bioreactors, has been demonstrated. As could be expected when using liquid medium for
micropropagation, several estimations confirm large gains in efficacy from temporary immersion. The parameters
most involved in reducing production costs include: (1) the drastic reduction in work; (2) reduction in shelving area;
(3) reduction in the number of containers used; (4) better biological yields. Scaling-up somatic embryogenesis and
shoot proliferation procedures involving temporary immersion systems in order to commercialize this process are
now taking place.

Introduction tures of 46 weeks, due to exhaustion of the nutrients

Current limitations of micropropagation
Whilst offering several advantages over conventional
propagation techniques for cloning of selected plant-
ing materials, micropropagation can be an unpre-
dictable and costly production technology. Current
techniques require a large number of small containers,
semi-solid media and aseptic division of plant tissues
by hand. Plant micropropagation involves periodic
transfers of plant material to fresh media, after subcul-
in the medium and also because of continuous tissue
growth and proliferation, which is rapidly limited by
the size of the culture container (Maene and Debergh,
1985). Agar products are not inert and complicate
automation. High production costs generally limit the
commercial use of micropropagation to products with
a very high unit value, such as ornamentals, foliage
plants and selected fruit crops (Sluis and Walker,
1985; Simonton et al., 1991). Labour generally ac-
counts for 4060% of production costs. Cutting and
planting represent the most expensive part of micro-
216
propagation (Chu, 1995). Although tissue handling is
the major part of the work and the most technical, there
is also the cleaning, filling and handling of a large
number of containers (Maene and Debergh, 1985).
Other major costs result from losses during acclimat-
ization and stem and root hyperhydricity (Reuther,
1985). It has been concluded that commercial applic-
ation of micropropagation for various species would
only take place if new technologies were available to
automate procedures, and if acclimatization protocols
were improved (Kitto, 1997).
Advantages of liquid media for plant
micropropagation
Liquid media are ideal in micropropagation for re-
ducing plantlet production costs and for automation
(Debergh, 1988; AitkenChristie, 1991). Indeed, li-
quid culture systems can provide much more uniform
culturing conditions, the media can easily be renewed
without changing the container, sterilization is pos-
sible by microfiltration and container cleaning after
a culture period is much easier. In comparison with
culturing on semi-solid media, much larger containers
can be used, and transfer times can be reduced.
Plant tissues from numerous species have per-
formed better when in liquid medium rather than on
semi-solid medium. For instance, a larger number of
shoots was produced in peach (Prunus persica L.)
(Hammerschlag, 1982) and more somatic embryos
were produced in wheat (Triticum aestivum) (Jones
and Petolino, 1988) and cotton (Gossypium hirsutum)
(Gawel and Robacker, 1990). Somatic embryogenesis
in comparison with organogenesis is the least labour-
intensive bioreactor protocol (Ziv, 1995). According
to AitkenChristie and Jones (1987) a prerequisite
for automating organogenesis is a culture system in
which shoots or somatic embryos can be produced in
the same container for a long period without trans-
fer, thereby enabling regular or complete harvesting
of shoots or somatic embryos for acclimatization and
plant conversion, respectively.
Micropropagation systems with liquid medium
Bioreactors developed in the past are not suitable for
micropropagation as they were mainly developed for
bacterial culture and do not take into account the spe-
cific requirements of plant cells, such as sensitivity to
shear forces, mechanical damage and foam formation
in bubble aerated bioreactors (Teisson et al., 1999).
The advantages of in vitro culture in a liquid me-
dium are often counterbalanced by technical problems
such as asphyxia, hyperhydricity, shear forces and the
need for complex equipment. In order to avoid such
problems, other procedures have been developed that
include culture supports such as paper bridges, cellu-
lose blocks or sponges (Etienne et al., 1991; Smith and
Spomer, 1995; Wataad et al., 1997), a raft to support
plants over stationary liquid (Connor and Meredith,
1984; Hamilton et al., 1985), adding liquid medium
to established cultures on agar (Maene and Debergh,
1985) and mist bioreactors (Weathers and Giles, 1988;
Tisserat et al., 1993). Temporary immersion has also
been utilized for micropropagation, based on a prin-
ciple similar to that of mist bioreactors, preferring
temporary contact between the plants and the liquid
medium rather than permanent contact.
Temporary immersion culture systems
Harris and Mason (1983) described tilting machines
designed to achieve temporary immersion, in order
to combine aeration and the positive effects of liquid
medium culture. They noted that Stewart et al. (1952)
observed that carrot (Daucus carota L.) root explants
did not grow rapidly when immersed in a liquid me-
dium, probably due to lack of oxygen. They designed
an apparatus known as an auxophyton, which turned
the culture containers on a wheel, exposing the ex-
plants alternately to air, or immersing them in liquid.
After 20 days culture, the carrot tissue weighed 2.6
times more than tissue cultured on an agar medium.
Following Harris and Mason (1983), several semi-
automatic systems using the temporary immersion
principle have been described (Table 1). All these sys-
tems respect the conditions mentioned by Teisson et
al. (1999):
(1) avoidance of continuous immersion, which ad-
versely affects growth and morphogenesis;
(2) provision of adequate oxygen transfer;
(3) provision of sufficient mixing;
(4) limit shear levels;
(5) enable sequential medium changes and automa-
tion;
(6) reducing contamination;
(7) low cost.
The systems differ with respect to container size,
type of culture support, existence of computerized
immersion control or a simple timer, use of either
a peristaltic pump, or an air pump, or mechanical
Table 1. Summary of micropropagation processes using temporary immersion culture. For each citation, the temporary immersion system used, the gain in biological
yield compared with conventional processes, occurence of hyperhydricity and acclimatization success rate are indicated

Family Species Morphogenic Temporary Immersion Immersion times Increase in biological Hyperhydricity Successful Authors
(Common Name) pathway System useda efficacy (x fold) in acclimatization of
comparison with the plant material
conventional protocol produced in a bioreactor

Vitis vinifera Shoot proliferation Tilting and rocker 30 s every 30 s Shoot nb x 7 (/agar) Not determined Not determined Harris and Mason
(grape) machines Shoot length increased; (1983)
rooting faster and more
efficient
Potinera spp. Shoot proliferation Automated plant culture 510 min every 12 h Shoot FW x 4 (/agar) No Not determined Tisserat and
(orchid) system (APCS) Vandercook (1985)
Callistephus Shoot proliferation Automated plant culture 510 min every 12 h Shoot biomass (FW) x 1 No Yes (shoots and Tisserat and
hortensis (Aster) system (APCS) (/agar) flowers after Vandercook (1985)
culture in
automated system
were larger)
Phenix dactylifera Embryogenic Automated plant culture 5-10 min every 12 h Date palm: callus FW x No Not determined Tisserat and
(date palm) callus proliferation system (APCS) 3.2 (/agar); better quality Vandercook (1985)
Daucus carota Carrot: callus FW x 1.9
(carrot) /agar); better quality and
plant regeneration
Mitragyna inermis Shoot proliferation Automated plant culture 510 min every 12 h Shoot FW x 1.8 (/agar) No Not determined Tisserat and
(cow tree) system (APCS) Vandercook (1985)
Pinus radiata D. Shoot hedge Liquid nutrient medium 46 h every 3 Shoot FW x 1.2 and shoot Yes but reduced Yes Aitken-Christie and
Don (radiata pine) proliferation on agar with liquid days health x 2 (/agar) compared to agar Jones (1987)
replenishment system medium
Amelanchier x Shoot meristem Programmable 5 min every 30 or Shoot nb x 2.6, shoot No with 5 min Not determined Krueger et al. (1991)
grandiflora proliferation micropropagation 60 min weight x 2.1, shoot length every 1 h;
Princess Diana apparatus using cycled x 1.2 (/agar) Yes with 5 min
(serviceberry) medium (Simonton et every 30 min,
al., 1991)
Musa acuminata Shoot meristem 1-litre Bioreactor with 2 20 min every 2 h Shoot nb x 2.5 (/agar) No Not determined Alvard et al. (1993)
(banana) proliferation compartments (RITA
)
Solanum tuberosum Tuberization 10-litre Twin Jar 60 min every 6 h Tuber nb x 3-4 No Direct planting Akita and Takayama,
L. (potato) fermentors (/fermentor) without (1994)
acclimatization
Triploid banana and Somatic 1-litre Bioreactor with 2 1 min every 6 h Embryo nb x 3 (/agar) No No Escalant et al. (1994)
plantain (Musa spp.) embryogenesis compartments (RITA
)
217
 218

Table 1. Continued

Family Species Morphogenic Temporary Immersion Immersion times Increase in biological Hyperhydricity Successful Authors
(Common Name) pathway System useda efficacy (x fold) in acclimatization of
comparison with the plant material
conventional protocol produced in a bioreactor

Coffea arabica and Microcuttings 1-litre Bioreactor with 2 15 min every 6 h (C. Shoot nb x 2 (/agar) No at the Not determined Berthouly et al. (1995)
Coffea canephora compartments (RITA
) arabica) immersion times
1 min every 6 h (C. used; Yes for
canephora) longer times
Hevea brasiliensis Somatic 1-litre Bioreactor with 2 1 min every 12 h Embryo nb x 4 (/agar); No No Etienne et al. (1997b)
(rubber tree) embryogenesis compartments (RITA
) (embryo 85% cotyledonary
development); embryos (vs 26% agar)
15 min every 6 h
(germination)

Coffea arabica Somatic 1-litre Bioreactor with 1 min every 12 h Embryo nb x 2 (/agar) No No Etienne et al. (1997a)
(coffee) embryogenesis 2 compartments 90% torpedo embryos (vs
(RITA
) 30% agar)

Citrus deliciosa Somatic 1-litre Bioreactor with 2 1 min every 4 h 66% well formed No (but No Cabasson et al. (1997)
embryogenesis compartments (RITA
) cotyledonary embryos (vs systematically
0% in suspension and 60% observed with
hyperhydric cotyledonary agar medium)
embryos on solid medium)

Saccharum spp. Shoot meristem 10-litre twin flasks 2 min every 9 h Shoot nb x 4.0 to 6.0 No Yes, efficient Lorenzo et al. (1998)
(sugar cane) proliferation (BIT
) (/agar)

Ananas comosus Shoot meristem 10-litre twin flasks 2 min every 3 h Shoot nb x 3.0 (/liquid) No Yes, efficient Escalona et al. (1999)
(pineapple) proliferation (BIT
) and x 4.0 (/agar)

Coffea arabica Somatic 1-litre Bioreactor with 2 1 min every 12 h 90% well formed torpedo No Yes, efficient by Etienne-Barry et al.
(coffee) embryogenesis compartments (RITA
) (embryo embryos (vs 30% in liquid direct sowing of (1999)
development) medium) germinated
5 min every 12 h 75% embryo-to-plant somatic embryos
(germination) conversion in ex vitro
conditions

a The best immersion times are presented when several frequency/duration combinations were tested.

motion of the container to displace the liquid. Other
differences between the temporary immersion systems
include recycling or not of the medium, and separation
or incorporation of the medium tank with the culture
container. Characteristics common to these systems
include containers that are larger than conventional
culture vessels, transparent and autoclavable. These
systems are easier to use than conventional bioreact-
ors, and longer subculture periods are possible with
most of them. Temporary immersion culture systems
provide programmable partial or total contact between
the explant and the liquid medium.
The systems are represented by four different
designs (see also Figure 1).
Systems with tilting or rocker machines
Two machines were described by Harris and Ma-
son (1983). The tilting machine inclines Erlenmeyers
flasks at an angle of 30 degrees in opposite directions;
it has a capacity of 400 50-ml Erlenmeyer flasks or
320 125-ml flasks. The Rocker machine rolls 70 910-
ml wide-mouth jars lying on their sides, or tilts 120
455-ml wide-mouth jars standing upright at 3040 de-
grees every 30 sec. These machines do not include
replenishment of the liquid culture medium.
Systems with complete immersion and a liquid
medium renewal mechanism
Tisserat and Vandercook (1985) developed a large el-
evated culture chamber that was periodically drained
and then refilled with fresh medium in a sterile envir-
onment. The automated plant culture system (APCS,
Figure 1A) consists of silicone tubing, 2 impeller
pumps, 2 glass medium reservoir bottles, a 3-way
stainless steel valve, a plant culture chamber, and an
interface module containing relay boards. This system
provides a long-term method for in vitro plant culture.
Systems with partial immersion and a liquid medium
renewal mechanism
The plant tissue is always positioned on a culture sup-
port (agar medium, propylene screen, cellulose plugs).
Liquid culture medium is frequently applied, then
withdrawn into a drain-off container, so as to imbibe
the support, stabilize the composition of the culture
medium and extend the duration of subcultures, whilst
avoiding or postponing the need to change the me-
dium. Only the base of the plant material is partially
219
immersed. Descriptions of two models have been
published:
AitkenChristie and Jones (1987) and Aitken
Christie and Davies (1988) developed a semi-
automatic process in large polycarbonate containers
measuring 250 390 120 mm (Figure 1B). In
their system, Pinus spp. shoots were grown on an
agar medium, with automatic addition and withdrawal
of liquid medium by peristaltic pumps on a periodic
basis. The liquid from the fresh medium recipient
came into contact with the explants for 46 h, using
a vacuum suction system, then went to the drain-off
container. Maene and Debergh (1985) had previously
shown the positive effects of adding liquid nutrient
medium or auxins to semi-solid medium in the final
in vitro stages.
The system of Simonton et al. (1991) featured a
computer-controlled pump that intermittently applied
liquid medium to plants cultured in 7-litre vessels
(Figure 1C). Plant material rested on a perforated
polypropylene screen that was attached to the inside
of the vessel. Control capabilities included medium in-
troduction and depth regulation within four individual
culture vessels, medium cycling on an assigned sched-
ule, schedule adjustment during a culture period and
medium replacement.
Systems with complete immersion by pneumatic
driven transfer of liquid medium and without medium
replenishment
Different systems have been described since Alvard et
al. (1993), and include the most recent temporary im-
mersion systems. They are relatively simple and easy
to use. They enable contact between all parts of the
explant and the liquid medium, along with complete
renewal of the culture atmosphere by forced ventil-
ation, which propulses the liquid towards the plant
material. The plant material can be placed in the con-
tainer in bulk, removing the need to position the plant
material on a support. Such systems include pneumatic
transfer of the medium from a tank to the container
holding the plants. To avoid excess tubing, these two
volumes are preferably part of the same vessel. Over-
pressure is applied by a solenoid valve or a compressor
connected to a programmable plug. This application
determines the time and duration of floodings. As
these systems do not include a fresh medium tank,
the culture medium has to be changed at 46 week
intervals. However, replacement is rapid and there is
no need to transfer the plant material. Two variants
220

Figure 1. Diagrammatic representation of some semi-automatic temporary immersion systems: (A) APCS system with complete immersion of
plant material and renewal of the liquid culture medium (from Tisserat and Vandercook, 1985); (B and C) systems with partial immersion and
with a liquid-nutrient renewal process [(B) from AitkenChristie and Davies, 1988; (C) from Simonton et al., 1991]; (D and E) systems with
complete immersion of plant material by pneumatic driven transfer of liquid medium and without medium renewal [(D) RITA
system from
Alvard et al., 1993; (E) BIT
twin flasks system from Escalona et al.,1999].
 221

Figure 2. (A) Aspect of coffee (Coffea arabica) germinated somatic embryo production in a RITA
type temporary immersion bioreactor;
(B) View of a culture room equiped with RITA
bioreactors; (C) Direct sowing on horticultural substrate of coffee somatic embryos mass
produced in a RITA
bioreactor; (D) BIT
type twin flasks system; (E) pineapple explants proliferating in a twin flasks system; (F) pineapple
stems elongating in a BIT
bioreactor.
222
of this system have been developed and are currently
on the market: the Recipient for Automated Tempor-
ary Immersion system (RITA
) and the Twin Flasks
system (BIT
).
The RITA
system (Figures 1D and 2B; Teisson
and Alvard, 1995). The 1-litre vessel comprises two
compartments, an upper one with the plants and a
lower one with the medium. The over-pressure ap-
plied in the lower compartment pushes the medium
into the upper one. Plants are immersed as long as
overpressure is applied. During the immersion period,
air is bubbled through the medium, gently agitating the
tissues and renewing the head space atmosphere in-
side the culture vessel, with the overpressure escaping
through outlets on the top of the apparatus.
The Twin Flasks system (BIT
) (Figures 1E and
2D, E; Escalona et al., 1998). The RITA
system
was intended for mass propagation by somatic em-
bryogenesis. For organogenesis, the size of propagules
may require a larger volume and cheaper vessels. The
easiest way to carry out pneumatically-driven tem-
porary immersion is to connect two glass or plastic
flasks from 250 ml to 10 litres by tubing, and ap-
ply alternative overpressure to push the medium into
the other recipient. The RITA
vessel can easily be
adapted to this configuration. Akita and Takayama
(1994) proposed a similar system known as the sys-
tem for semi-continuous medium surface level control
culture for Solanum tuberosum L. tuberization. That
system incorporates a forced aeration in the culture
recipient, which is not found in the BIT
system.
Effect of temporary immersion on biological yield
for different types of micropropagation
When envisaging commercial use of automated tem-
porary immersion systems, it is important to take
comprehensive measurements of growth, production
and quality of the cultured material, and compare
them to those obtained with material produced with
the conventional culture systems.
Shoot proliferation and microcuttings
Temporary immersion stimulates shoot proliferation.
AitkenChristie and Jones (1987) showed that bet-
ter shoot growth could be obtained for Pinus radiata
by replenishing liquid nutrient medium in contrast to
monthly transfers on semi solid medium. This system
enabled continuous shoot growth and monthly har-
vests for 18 months, without transferring the plant
material. Shoots obtained with partial and temporary
immersion in the nutrient medium were longer and
of better quality than those obtained on semi solid
media. A complete and convincing demonstration of
the efficacy of temporary immersion was carried out
for the proliferation of Banana (Musa subgroup AAH)
shoot tip cultures. Alvard et al. (1993) showed that
applying liquid medium strongly influenced the de-
velopment and proliferation rate of micropropagated
banana explants. When four liquid medium culture
methods were compared to conventional growth on
semi solid medium, they obtained the following results
after culturing for 20 days:
(i) shoots laid in a simple liquid medium or on a
cellulose support barely proliferated or not at all;
(ii) shoots on semi solid medium, with partial im-
mersion and in a bubble-aerated medium had
multiplication rates of 2.2 to 3.1 and;
(iii) the highest proliferation rate (>5) was obtained
with explants subjected to temporary immersion.
These authors obtained their results using a RITA

bioreactor, with 20 min immersions every 2 h.
A Cuban team obtained similar results with Musa
acuminata, using the Twin Flasks system (Teisson et
al., 1999).
Serviceberry shoots (Amelanchier grandiflora
Rehd. Princess Diana) were cultured in the tempor-
ary immersion system described by Simonton et al.
(1991) and compared to those obtained on semi solid
medium or in liquid medium in baby food jars, and on
semi solid medium or in non-cycling liquid medium in
a seven-litre vessel (Krueger et al., 1991). A combina-
tion of liquid medium, a 7-litre container and intermit-
tent contact with the liquid medium gave significantly
higher proliferation rates than any other combination
tested. Compared with conventional procedures (semi
solid medium in baby food jars), cultures that grew
in intermittent contact with the culture medium gave
higher values for the number of shoots (2.6), shoot
weight (2.1), shoot length (1.2) and culture weight
(2.2).
With Saccharum spp. (sugar cane), Lorenzo et
al. (1998) demonstrated that the Twin Flasks sys-
tem for temporary immersion clearly stimulated shoot
formation and length. The multiplication rate (23.9
shoots/30 days) was 6 greater than the standard pro-
tocol (3.96 shoots/30 days; Jimnez et al., 1995).
Similar results were obtained with 3 other banana gen-
otypes. Likewise, Escalona et al. (1999) used the same
bioreactor for pineapple Ananas comosus shoot tip
culture to show that temporary immersion stimulated

the multiplication rate, along with the fresh and dry
weights, after culturing for 42 days. The multiplication
rates increased by 300% and 400% respectively, com-
pared to rates obtained with liquid or solid supports.
Up to 5000 pineapple plants can be harvested from a
single container (Figure 2E, F).
Potinera sp. (orchids) and Mitragyna inermis (cow
tree) shoot tips grown in the APCS temporary im-
mersion system (Tisserat and Vandercook, 1985) grew
more rapidly than on semi solid medium. Based on
fresh weight and volume measurements, they repor-
ted a four-fold increase for these two parameters with
orchid after 270 days, and an increase of 1.8 with
cow tree after 45 days. The same tests on Callistephus
hortensis (aster) shoot cultures did not demonstrate
any difference in plant growth and development. Nev-
ertheless, the aster plants regenerated by temporary
immersion were more vigorous in the nursery.
In the case of coffee (Coffea arabica and C. cane-
phora), multiplication by microcuttings on a semi-
solid medium is of limited value, due to the slow
growth of orthotropic shoots. The multiplication rate
is approx 6 or 7 every 3 months (Sndhal et al., 1984).
With a RITA
a similar result was obtained in just
56 weeks (Berthouly et al., 1995). Cyclically im-
mersed grapevine (Vitis vinifera L.), in side-to-side
tipping culture vessels using tilting machines with al-
ternate exposure and submergence intervals of 30 sec
or longer, produced seven times as many shoots in 90
days as explants maintained on semi solid media (Har-
ris and Stevenson, 1982; Harris and Mason, 1983).
Similar increases in shoot numbers have been obtained
with Arctostaphylos uva ursi (L.), Amelanchier alni-
folia Nutt., tobacco (Nicotiana tabacum Xanthi-nc)
and fuchsia (Fuchsia hybrida Swingtime) (Steven-
son and Harris, 1980).
Microtuberization
Plant growth and tuberization of Solanum tuberosum
L. (potato) were stimulated by temporary immersion
in a twin flask system (Akita and Takayama, 1994).
The number of tubers formed, i.e. approximatively
500960 tubers for a 10 weeks culture in a jar fer-
mentor, was greater than previous results (approx. 220
tubers for a culture [Akita and Takayama, 1993]).
Total tuber weight and homogeneity also increased.
On the other hand, there was no tuber formation under
complete or continuous immersion conditions. Teisson
and Alvard (1999) confirmed the efficiency of tempor-
ary immersion for potato microtuberization, working
223
with a double RITA
system based on the Twin Flasks
method. Three microtubers were obtained per single
node 10 weeks after inoculation. Fifty percent of the
microtubers were 0.5 g and sprouted, still in a tem-
porary immersion system. This system was very rapid
and efficient; three to four shoots developed from
a single tuber. Similar results were obtained with 3
different cultivars (Teisson and Alvard, 1999).
Somatic embryogenesis
Proliferation of embryogenic cultures
Temporary immersion culture has been used to en-
hance embryogenic culture proliferation in compar-
ison with conventional systems using semi solid me-
dium or suspensions in Erlenmeyer flasks. Tisserat
and Vandercook (1985) quantified the growth of car-
rot and Phoenix dactylifera (date palm) embryogenic
cultures in a totally automated culture system (APCS),
in which immersions of 510 min were applied every
2 h. Compared to cultures on semi solid medium, there
was a growth increase of 1.9-fold for carrot and 4-fold
for date palm. Likewise, an improvement in culture
quality was noted for the 2 species and a large quantity
of somatic embryos and plants was obtained for carrot.
Similarly, for different coffee (C. arabica) genotypes
tested, embryogenic culture growth was greater with
temporary immersion than in a stirred liquid medium
in an Erlenmeyer flask (Berthouly et al., 1995).
Embryo development
The production and quality of somatic embryos have
been improved for various species by temporary im-
mersion culture. With Citrus deliciosa, Cabasson et
al. (1997) compared the efficacy of various culture
systems on somatic embryo development. Somatic
embryos derived from suspension cultures were plated
on semi-solid medium, maintained in suspension cul-
ture or immersed temporarily. About 60% of somatic
embryos plated on semi-solid medium developed to
the cotyledonary stage, but were hyperhydric. Con-
tinuous growth in a suspension culture at 100 rpm
hindered cotyledon and protoderm formation, and so-
matic embryos were unable to develop beyond the
globular stage. Temporary immersion applied in a
RITA
bioreactor promoted somatic embryo devel-
opment, i.e. 66% of the somatic embryos produced
were cotyledonary, and were morphologically similar
to nucellar embryos. Escalant et al. (1994) showed
with several banana and plantain (Musa subgroup
AAB) cultivars that after two months of temporary
224
immersion culture in a RITA
bioreactor, three times
more somatic embryos were produced than on an agar
medium (1375 embryos vs 450). Temporary immer-
sion promoted secondary embryogenesis from the epi-
dermal cells of primary embryos. After 6 months, the
initial number of somatic embryos in the 1-liter biore-
actor had increased by 40, i.e. 6000 embryos. On
the other hand, after 2 months incubation on an agar
medium, the embryos had changed into a compact,
white callus. High conversion rates (6070%) were
obtained by transferring somatic embryos produced
with temporary immersion onto semi-solid medium.
With Hevea brasiliensis (rubber), somatic embryo
production on semi-solid medium results in low and
not particularly reproducible yields of somatic em-
bryos of poor morphological quality (Etienne et al.,
1993). Transferring embryogenic cultures to a RITA

type temporary immersion container greatly improved
the quantity and quality of somatic embryos produced,
and enabled routine production (Etienne et al., 1997b).
Somatic embryo production was three to four times
greater than on a semi-solid medium, i.e. 400 so-
matic embryos/g fresh weight embryogenic culture.
Temporary immersion also reduced the proportion
of abnormal somatic embryos by half, with an in-
crease in the germination rate. Temporary immersion
considerably increased germination (+60%) and epi-
cotyl emergence (+35%). According to Teisson et al.
(1999), approx 150 rubber somatic embryos at the
cotyledonary stage per RITA
vessel were harvested
48 weeks after transferring embryogenic cultures to
the bioreactor. However, in order to regenerate fully
developed plants, the germinated material had to be
transferred to a semi-solid medium. Plant conversion
requires plantlets should be in an upright position,
which is impossible to achieve in a RITA
vessel, as
their position changes with every flooding.
With coffee, mass somatic embryo production in
Erlenmeyer flasks and in bioreactors has been possible
for some time for C. canephora, for which the pro-
duction of several hundred thousand somatic embryos
per gram of inoculum has been reported (for review
see Berthouly and Etienne, 1999). Coffea arabica is
a more recalcitrant species. For large-scale dissemina-
tion of improved F1 hybrids of C. arabica in Central
America, reproducible production was successfully
achieved with approx. 20 clones using a RITA

type temporary immersion bioreactor (Etienne et al.,
1997a). Depending on the genotypes, yields ranging
from 15 00050 000 somatic embryos per gram cell
mass of embryogenic suspension were recorded. Such
yields are twice those obtained in Erlenmeyers flasks
under optimum conditions. The quality of C. arabica
somatic embryos produced by temporary immersion
(Figure 2A) has been very high. Whilst the propor-
tion of normal torpedo type embryos is around 30%
in a bioreactor or in Erlenmeyer flasks (Zamarripa et
al., 1991; Noriega and Sndahl, 1993), it is usually
90% with temporary immersion. This improvement
in quality is reflected in higher plant conversion rates
on semi-solid medium (8090%) and especially in
the successful regeneration of plants under ex vitro
conditions after direct sowing of somatic embryos
(EtienneBarry et al., 1999; Figure 2C). Embryo to
plant conversion rates of 7080% have been achieved
routinely under such conditions for all genotypes.
Synchronization of embryo production
With citrus, temporary immersion also improved re-
generation synchronization by inhibiting secondary
embryogenesis (Cabasson et al., 1997). Temporary
immersion also increased synchronization with rubber
and C. arabica during the embryo development and
germination phases when compared to cultures on a
semi-solid medium (Etienne et al., 1997b; Etienne
Barry et al., 1999). With C. arabica it is promoted
by high culture densities (15003000 somatic em-
bryos per 1-litre bioreactor), for which 66% of the
somatic embryos are at the same germination stage
(Figure 2A).
With Musa spp. (Escalant et al., 1994), unlike
Citrus, temporary immersion promoted embryo prolif-
eration by secondary somatic embryogenesis. With the
temporary immersion system, Musa and Coffea spp.
somatic embryos can be subcultured by transferring
part of the embryogenic culture to other bioreactors.
Culture parameters affecting the efficacy of
temporary immersion systems
The main reason for the efficacy of temporary immer-
sion systems is probably that they combine ventilation
of the plant tissues, and intermittent contact between
the entire surface of the tissue and the liquid medium.
These two characteristics are not usually combined in
other liquid culture procedures.
Immersion time
In culture systems with temporary tissue immersion,
it is clear that the immersion time is very import-
ant, since it determines nutrient uptake and control of

hyperhydricity. The immersion times used vary con-
siderably (Table 1). This is probably due to the large
variety of species, micropropagation processes and
temporary immersion systems used. Long immersion
times (1 h every 6 h) were efficient for potato tuberiza-
tion, whereas very short immersion times (1 min every
12 h) stimulated somatic embryo production with C.
arabica and rubber (Etienne et al., 1997a,b). Like-
wise, frequent immersions (30 sec every 30 sec) can be
highly efficient in tilting machines for grapevine shoot
propagation (Harris and Mason, 1983).
Krueger et al. (1991) demonstrated the importance
of immersion frequencies for the proliferation of ser-
viceberry shoots. Hyperhydricity was observed with
immersions of 5 min every 30 min, but was not seen
with immersions for 5 min every 60 min. On the other
hand, the former combination is better for the number
of shoots obtained. The authors recommended using
the former combination during an initial proliferation
cycle, and then the latter combination to maintain
shoot quality. They also observed an adaptation period
of the plant material when immersion frequency is
modified. After changing to lower immersion frequen-
cies, stress resulted in partial desiccation of the shoots,
but the material later recovered. With Pinus radiata
shoots grown on semi-solid medium, liquid nutrient
replenishment at a frequency of twice weekly more
effectively stimulated growth (FW 1.2) and shoot
quality ( 1.5) than frequencies of once every 2 or 4
weeks (AitkenChristie and Jones, 1987).
Berthouly et al. (1995) with Coffea microcuttings
showed that immersion time substantially affected the
shoot multiplication rate, as estimated by the number
of micronodes produced after 6 weeks. Indeed, immer-
sion times of 1, 5 and 15 min applied every 6 h gave
multiplication rates of 3.5, 5.4 and 8.4, respectively. In
addition, the optimum immersion time varied depend-
ing on the Coffea species. For instance, it was 15 min
every 6 h for C. arabica microcuttings and only 1 min
every 6 h for C. canephora microcuttings.
With Coffea spp., modifying the immersion time
greatly affects somatic embryo production. According
to Teisson and Alvard (1995), 15 min immersion every
6 h led to successive development and germination
of embryos, whereas for an identical culture medium,
immersions of 1 min every 24 h halted embryo de-
velopment and stimulated the production of secondary
embryos.
We recently observed that increasing the frequency
for short immersions (1 min) stimulated somatic em-
bryo formation and quality in Coffea arabica. Thus,
225
average yields of 480, 2090 and 3100 embryos were
obtained per 1-litre bioreactor, with 60, 79 and 85%
torpedo type embryos, for daily frequencies of 1, 2
and 6 immersions respectively. Hyperhydricity was
not observed. On the other hand, increasing immer-
sion times by 5 min or more led to a considerable
reduction in somatic embryo production and in their
quality, becoming more critical as the immersion fre-
quencies increased. For example, 15 min immersions
applied 2 or 6 times per day led to hyperhydricity fre-
quencies of 64 and 90%, respectively. It is likely that
each stage of somatic embryo development requires
adaptation of the immersion length and frequency to
obtain optimum results.
Volume of liquid medium
Liquid medium volume must be optimized with tem-
porary immersion systems without medium renewal,
i.e., Twin Flasks and RITA
systems or tilting and
rocker machines. Lorenzo et al. (1998) determined
an optimum volume of medium per explant (50.0 ml)
for Saccharum spp. shoot proliferation in the BIT

Twin Flasks system. An increase in multiplication
rate from 8.3 shoots/30 days to 23.9 shoots/30 days
was obtained by multiplying the volume of standard
medium by 10 from 5.0 to 50.0 ml per explant. How-
ever, the volume of medium used did not affect the
length of the shoots formed. Higher volumes proved
to be less efficient, and the authors suggested that the
cultures produced extracellular chemicals that stimu-
lated shoot formation, which would be diluted when
large volumes of medium are used. Using the same
temporary immersion system, Escalona et al. (1999)
demonstrated that an optimum medium volume exists
for pineapple shoot proliferation, which was estimated
to be 200 ml/explant. In this case, larger volumes also
led to a decreased proliferation rate.
Culture container volume
For all temporary immersion systems, the volume of
the container, and hence the head space, is much
greater than in the containers used for conventional
procedures. Moreover, containers ranging in size from
1 to 20 litres can usually be adapted to the system.
Krueger et al. (1991) demonstrated that the large size
of their culture container (7 l) had a positive effect on
micropropagation efficiency for serviceberry, notably
by preventing culture overcrowding and encouraging
shoot elongation, when compared to those obtained in
140-ml baby food jars. Monette (1983) had already
226
shown for grapevine that longer shoots could be ob-
tained in larger containers. Grapevine explant growth
was so strong in liquid medium using tilting ma-
chines, that it was necessary to use larger containers,
e.g. 910-ml square wide-mouth Mason jars, as op-
posed to 125-ml Erlenmeyer flasks to avoid crowding
of the cultures, benefitting from a larger aperture to
remove material, and larger volumes of medium to
prevent early deficiencies of certain constituents of
the medium (Harris and Mason, 1983). Using larger
containers means that larger volumes of media can be
used, which can have a positive effect on plant material
proliferation and growth.
Aeration and forced ventilation
Work by Alvard et al. (1993) on banana micro-
propagation clearly showed that a lack of air in the
liquid culture medium was a major limiting factor
for small explant growth. In their experiment, the
absence of liquid medium stirring led to explant as-
phyxia. Bubble aeration of the medium encouraged
growth, but partial immersion of the explant did
not provide sufficient aeration. Temporary immersion
clearly proved to be the most effective culture system.
However, for other systems, the positive effect of aer-
ation has not been proved. AitkenChristie and Jones
(1987) showed that aeration alone did not stimulate
shoot growth in Pinus radiata and was not a contribut-
ing factor for the increased growth found with nutrient
replenishment. Steward et al. (1952) attributed better
shoot growth in carrot to improved air supply through
alternating immersion.
The systems using pneumatic propulsion of the
nutrient medium caused forced ventilation leading to
complete renewal of the culture atmosphere at each
immersion. According to Teisson and Alvard (1995),
gas exchanges in such a system primarily occur during
immersion and are caused indirectly by movement of
the liquid and directly by the air pump. Under the most
frequently used culture conditions that corresponds to
complete renewal of the culture atmosphere after 5
min of immersion for a 1-litre bioreactor. Such forced
ventilation with air containing the gas concentrations
and relative humidity of the culture room probably has
positive effects (Krueger et al., 1991). The relative hu-
midity resulting from forced ventilation may stimulate
transpiration in the plants, which will then be more
effectively adapted to ex vitro conditions (Wardle et
al., 1983).
Effect of temporary immersion on quality
Morphological characteristics of plant material
produced in a bioreactor
Leaves originating from temporarily immersed pro-
liferating pineapple shoots were smaller than those
produced in a liquid medium (Escalona et al., 1999).
Clusters of shoots produced during the proliferation of
axillary buds in a bioreactor were almost spherical and
the shoots all formed around a central region. Some of
the shoots, which were too small for ex vitro rooting
and acclimatization, required an elongation phase in
the same bioreactor. Cow tree shoots formed in tem-
porary immersion were considerably longer and had
more foliage than those produced on semi-solid me-
dium (Tisserat and Vandercook, 1985). Serviceberry
shoots propagated by temporary immersion were also
heavier and longer than those obtained on semi-solid
medium (Krueger et al., 1991). Shoots of grapev-
ine and Amelanchier alnifolia produced in a liquid
medium on tilting or rocker machines were longer
and rooted better and more quickly than those pro-
duced on semi-solid media (Harris and Mason, 1983).
The positive effect of temporary immersion on shoot
elongation probably resulted from the larger volume
of the container (Monette, 1983; Krueger et al., 1991).
Temporary immersion had a highly positive effect
on the development of Citrus somatic embryos, lead-
ing to the development of cotyledons and protoderm,
which does not occur in suspensions (Cabasson et al.,
1997). The somatic embryos obtained were morpho-
logically identical to nucellar embryos. Likewise, in
Coffea spp., rubber and Musa, somatic embryos pro-
duced in a RITA
type bioreactor were of higher
quality than that obtained on a semi-solid medium or
in Erlenmeyer flasks (Etienne et al., 1993, 1997a, b;
Escalant et al., 1994). With rubber, the proportion of
morphologically abnormal somatic embryos was re-
duced by half with temporary immersion, compared
to semi-solid medium. Development problems due to
continuous stirring in suspensions or in conventional
bioreactors are substantially reduced in temporary im-
mersion systems. Stirring, which is a source of shear
forces that can damage the plant material, can also
have negative effects on polar distribution of growth
regulators, which is essential during early ontogen-
etic stages (Liu et al., 1993). Effective germination
in a liquid medium was successfully obtained for the
first time with C. arabica using temporary immer-
sion (EtienneBarry et al., 1999). Densities exceeding

1600 embryos per 1-litre bioreactor had a positive ef-
fect on germinated somatic embryo morphology by
stimulating elongation of the embryonic axis (+4
5 mm), an increase in fresh weight (+100%) and a
reduction in cotyledon area. These three morpholo-
gical changes were positively correlated to the rate of
conversion into plants and to plantlet growth rates.
The quality of tissues grown in a temporary immer-
sion system can change during subculture, revealing
adaptation to the system. The percentage of normal,
waxy (abundant tubular epicuticular wax) Pinus radi-
ata shoots harvested monthly increased significantly
over the culture period from 41% at the first harvest
to 93% at the eighth harvest, and remained at 97%
from the ninth to twelfth harvests (AitkenChristie and
Jones, 1987). We also observed a similar adaptation
phenomenon to changes in the immersion times with
C. arabica somatic embryo cultures.
Hyperhydricity
Micropropagation in liquid culture media increases
nutrient uptake and promotes growth, but hyperhydri-
city is common. Continuous contact of plant tissues
with the liquid medium is the source of hyperhydri-
city (Debergh et al., 1981; Ziv et al., 1983; Hussey,
1986). It is characterized by different degrees of mor-
phological and physiological disorders including a
glassy, waterlogged-tissue appearance, disorganized
growth of the shoot system, and more specifically
abnormal leaves (Ziv, 1995). Increasing aeration in
the culture recipient (Hussey, 1986) and intermittent
contact between the plant material and the liquid cul-
ture medium (AitkenChristie and Jones, 1987) can
reduce hyperhydricity. These two characteristics are
combined in most temporary immersion systems.
Alvard et al. (1993) did not report hyperhydri-
city in Musa shoots grown in a temporary immersion
system, whereas stems immersed in a liquid medium
with continuous bubble-aeration developed hyperhy-
dric outer leaf sheaths. Serviceberry shoots propagated
in a liquid medium developed severe hyperhydricity
symptoms, i.e., translucent, curled and thickened
leaves and stems (Krueger et al., 1991). Temporary
immersion of 5 min at 1 h intervals prevented hype-
rhydricity. However, these authors noted that more
frequent immersions (5 min every 30 sec) resulted in
hyperhydricity.
In the long term, the proportion of wet shoots (no
tubular epicuticular wax, small amounts of globular
epicuticular wax) in Pinus radiata was significantly
227
less in a temporary immersion culture system with
liquid medium replenishment, than on semi-solid me-
dium, and fell from 59% at the first harvest to 7%
at the eighth (AitkenChristie and Jones, 1987). Ac-
cording to these authors, this change was probably
due to the fact that monthly harvests of axillary shoots
were formed closer to the upper part of the container,
hence distal to the agar and the surface of the liquid,
where humidity was higher and where the liquid re-
mained on the surface of the needles. Coffea arabica
microcuttings showed little or no hyperhydricity with
immersion times of 15 min every 6 h, but longer im-
mersion times resulted in hyperhydricity (Berthouly
et al., 1995). On the other hand, when cultured with
the same immersion times, Coffea canephora mi-
crocuttings were hyperhydric. This species is more
susceptible to hyperhydricity, and immersion times
had to be reduced to 1 min every 6 h.
Somatic embryos of Citrus grown in a temporary
immersion system were not hyperhydric, unlike those
obtained in suspensions or semi-solid medium (Cabas-
son et al., 1997). Rubber somatic embryos produced
in a temporary immersion system are not hyperhy-
dric (Etienne et al., 1997b), although hyperhydricity
was observed in germinating material in plant conver-
sion medium. This was probably due to the excessive
immersion times used (four 15 min immersions per
day). As in the case with C. arabica, hyperhydricity
is greater in the later phases of somatic embryogenesis
(i.e. germination and plant conversion). Short immer-
sion times need to be used to avoid this problem. We
found that hyperhydricity of C. arabica somatic em-
bryos was primarily caused by the immersion times,
but not by immersion frequency if short immersions
(1 min) were used. Several 15 min immersions per
day led to embryo populations that were mostly hyper-
hydric. Using temporary immersion generally reduced
hyperhydricity.
Acclimatization of material produced in a temporary
immersion system
Losses during acclimatization can severely handicap
conventional micropropagation. Temporary immer-
sion systems have resulted in successful acclimatiza-
tion. Ex vitro acclimatization and rooting of pineapple
shoots (Escalona et al., 1999) and Saccharum spp.
shoots (Lorenzo et al., 1998) produced in a temporary
immersion system under semi-industrial conditions
has been efficient and routine. With pineapple, the
survival rate increases linearly with shoot size. Shoots
228
measuring over 6 cm can be grown directly in the
greenhouse, with 90100% successful acclimatization
(Escalona et al., 1999). Smaller shoots require a longer
culture period in the bioreactor prior to acclimatiz-
ation. Similarly, C. arabica and C. canephora mi-
crocuttings obtained by temporary immersion can be
acclimatized directly after root induction in the same
bioreactor (Berthouly et al., 1995). After planting, the
shoots and flowers of Callistephus hortensis plants
obtained from shoot tips grown in a temporary immer-
sion system were larger than those from semi-solid
medium (Tisserat and Vandercook, 1985). Shoots of
Pinus radiata harvested from a nutrient replenishment
system were effectively rooted and no decline in root-
ing ability was found during successive harvests for
18 months (AitkenChristie and Jones, 1987). Using
temporary immersion provided an unexpected bonus
during the acclimatization stage. When compared to a
conventional system on semi-solid medium, the larger
proportion of waxy shoots compared to wet shoots,
which root poorly and have low survival rates (Aitken
Christie et al., 1985), resulted in higher survival rates.
Shoots of grapevine and Amelanchier alnifolia pro-
duced in a liquid medium on tilting machines also
rooted better and more quickly than those produced
on semi-solid media (Harris and Mason, 1983).
Potato tubers propagated in a Twin Flask system
could be stored under room conditions and trans-
planted directly to soil without acclimatization (Akita
and Takayama, 1994). Temporary immersion has en-
abled mass production of germinated C. arabica so-
matic embryos capable of regenerating into plants
(7080% embryo-to-plant conversion) (EtienneBarry
et al., 1999).
Impact of temporary immersion culture on
production costs
The few estimations made so far confirm the increase
in efficiency that can be expected from using a liquid
medium culture procedure for micropropagation. For
the proliferation of Saccharum spp. shoots, Lorenzo
et al. (1998) calculated that temporary immersion
reduced costs by 46% compared with the standard
procedure on semi-solid medium. This gain primar-
ily resulted from labour reduction and in the space
required. Using temporary immersion to propagate
pineapple shoots resulted in a 100-fold increase in the
number of shoots during the four-month period fol-
lowing culture establishment (Escalona et al., 1999).
This protocol reduced production costs per pineapple
plant by 20% when compared to the conventional
method in liquid medium. According to the authors,
the key parameters for this success were: A reduction
in the number of containers and in material handling,
elimination of cutting and planting, elimination of in
vitro rooting, and reduction in contamination levels.
Long-term maintenance of Pinus radiata shoot
hedges by nutrient replenishment led to a reduction
in labour costs, facilitated the move towards automa-
tion and reduced the cost of micropropagated trees
(AitkenChristie and Jones, 1987). This was the first
report of a method of culturing shoots as hedges for a
period of up to 18 months without manual subcultur-
ing. The shoots were harvested in 600-ml glass jars at
a rate of 672 shoots/h and approximately 1100 shoots
were produced per square metre of agar surface per
month. Based on that harvesting rate, this system is
around 7 times cheaper than the normal method for
subculturing Pinus radiata shoots on semi-solid me-
dium (Smith, 1985). Culturing in liquid medium on
tilting and rocker machines, as described by Harris
and Mason (1983), led to lower costs by reducing the
volume of medium by 50% and of agar by 90% or
more, in addition to gains in shoot yield, shoot size
and better rooting.
In somatic embryogenesis, the late maturation and
germination phases are the most expensive, due to
the amount of handling required. Temporary immer-
sion bioreactors are used either for the production
and germination of somatic embryos, e.g. C. arab-
ica (EtienneBarry et al., 1999), or for germination,
e.g. Musa (Teisson et al., 1999 ). The production and
germination of C. arabica somatic embryos in a tem-
porary immersion bioreactor, combined with direct
sowing of germinated embryos under ex vitro condi-
tions reduced handling times to 13% and shelving area
requirements to 6.3% of the values obtained with con-
ventional acclimatization of plants developed on semi-
solid media (EtienneBarry et al., 1999). Moreover,
maturation and germination period was reduced by 3
months.
Commercialization of temporary immersion sys-
tems is currently underway. CIRAD has been using
mass propagation by somatic embryogenesis with the
RITA
system to disseminate clones of selected F1
C. arabica hybrids in Central America (in cooperation
with CATIE and PROMECAFE) and in Tanzania (in
cooperation with the ARTI of Lyamungu). A Cuban
team at the Centro de Bioplantas in Ciego de Avila is
also developing commercial propagation of sugarcane

and pineapple using meristem proliferation techniques
with the BIT
Twin Flasks system.
Conclusions
Temporary immersion has positive effects on all stages
of shoot proliferation and somatic embryogenesis with
many plant species. Plant growth and proliferation
rates are generally better than those obtained on semi-
solid media or in bioreactors. Regenerated plantlets
and somatic embryos are of better quality. Good res-
ults have also been obtained during acclimatization
of plant material from temporary immersion. Survival
rates and plant vigour in the nursery were gener-
ally increased with material regenerated in temporary
immersion systems. Temporary immersion combines
the advantages of solid culture media (maximum gas
exchanges) and liquid culturing (increased nutrient up-
take). Explant immersion times and frequencies are
probably the most critical parameters for developing
a micropropagation procedure. Their optimization, to-
wards lower values, results in higher biological yields
through greater control of morphogenesis, but also
through the control of hyperhydricity. This is a major
advantage over traditional bioreactors. Culture density
is a determining factor, but has not yet been examined
extensively. Immersion times can be as critical for
each culture stage as the duration of subcultures and
the chemical composition of plant growth media.
More research on the effects of temporary immer-
sion on physiology is essential to optimize culture
conditions in these simplified bioreactors. Apart from
being used for plant micropropagation, temporary im-
mersion represents a useful tool to study different
metabolic processes. For example, temporary immer-
sion systems constitute an interesting model to study
the basis of hyperhydricity. Current work with Coffea
show that immersion times directly affect hyperhy-
dricity and the water status of callus and somatic
embryos. More precisely, some water parameters, e.g.
water content and water potential, are significantly
correlated to hyperhydricity and strongly influenced
by the frequency and the time of the immersions. With
rubber, temporary immersion permits to study the in-
duction of oxidative stress on an embryogenic culture
during the immersion stage (Martre et al., 2001). An
immersion as short as 1 min is enough to cause an
increase of the superoxide dismutase activity and a
high lipid peroxidation which disappears in less than
1 h after the end of the immersed stage. The beneficial
229
effect of temporary immersion systems in reducing the
action of extracellular toxic substances or growth in-
hibitors produced by cultured material has not been
assesed. Other studies have focused on optimizing
desiccation of mature somatic embryos to establish
higher plant conversion efficiency and medium-term
storage conditions. The Rita
vessel is particularly
well configured to offer conditions for slow desicca-
tion. The relative humidity in the upper compartment
where the plant material is cultured can be controlled
by the type of saturated saline solution placed in the
lower compartment. Positive effects on plant con-
version efficiency were observed with rubber and C.
arabica somatic embryos dehydrated under such con-
ditions. Many temporary immersion systems are also
adapted to study the effect of the gaseous atmosphere
on the growth of plant material.
Since the work of Harris and Mason (1983), sev-
eral semi-automatic temporary immersion systems
have been developed and the interest in this new
culture technique has progresivelly increased (see
Table 1). From the initial complex bioreactors, the
temporary immersion systems have gradually become
simpler and can be actually assimilated to common in
vitro culture containers. The simplicity and low cost
of recently developed bioreactors are compatible with
large-scale propagation. However, it is likely that new
simplifications are possible in view to reduce more
the price of these systems. Temporary immersion has
permitted important lower labor and shelving area re-
quirements, and consecutively the production cost.
The significant reduction in the production cost of in
vitro plants from temporary immersion indicates that
somatic embryogenesis, microtuberization and shoot
proliferation production systems can be optimized for
several important species. Several pilot productions of
hundred thousands plants of C. arabica and sugarcane
were realized using different temporary immersion
systems, in view of commercial application.
Acknowledgements
Sincere thanks to Maritza Escalona for supplying
documents and photographs.
References
AitkenChristie J (1991) Automation. In: Debergh PC & Zimmer-
man RJ (eds) Micropropagation: Technology and Application
230
(pp 363388). Kluwer Academic Publishers, Dordrecht, The
Netherlands
AitkenChristie J & Davies HE (1988) Development of a semi-
automated micropropagation system. Acta Hortic. 230: 8187
AitkenChristie J, Jones C & Bond S (1985) Wet and waxy shoots
in radiata pine micropropagation. Acta Hortic. 166: 93100
AitkenChristie J & Jones C (1987) Towards automation: radiata
pine shoot hedges in vitro. Plant Cell Tiss. Org. Cult. 8: 185196
Akita M & Takayama S (1994) Stimulation of potato (Solanum
tuberosum L.) tuberization by semicontinuous liquid medium
surface level control. Plant Cell Rep. 13: 184187
Alvard D, Cte F & Teisson C (1993) Comparison of methods of
liquid medium culture for banana micropropagation. Effects of
temporary immersion of explants. Plant Cell Tiss. Org. Cult. 32:
5560
Berthouly M, Dufour M, Alvard D, Carasco C, Alemano L &
Teisson C (1995) Coffee micropropagation in a liquid medium
using the temporary immersion technique. In: ASIC Publishers
(eds) 16th International Scientific Colloquium on Coffee, Kyoto,
Japon (pp 514519), Vevey
Berthouly M & Etienne H (1999) Somatic embryogenesis of coffee.
In: Jain SM, Gupta PK & Newton RJ (eds) Somatic Embryogen-
esis in Woody Plants, Vol. 5 (pp 259288). Kluwer Academic
Publishers, Dordrecht, The Netherlands
Cabasson C, Alvard D, Dambier D, Ollitrault P & Teisson C
(1997) Improvement of Citrus somatic embryo development by
temporary immersion. Plant Cell Tiss. Org. Cult. 50: 3337
Chu I (1995) Economic analysis of automated micropropagation.
In: AitkenChristie J, Kozai T, Smith MAL (eds) Automation
and Environmental Control in Plant Tissue Culture (pp 1927).
Kluwer Academic Publischers, Dordrecht, The Netherlands
Connor AJ & Meredith CP (1984) An improved polyurethane sup-
port system for monitoring growth in plant cell cultures. Plant
Cell Tiss. Org. Cult. 3: 5968
Debergh P (1988) Improving mass propagation of in vitro plantlets.
In: Kozai T (ed) Horticulture in High Technology Era (pp 45
57). International Symposium on High Technology in Protected
Cultivation, Tokyo
Debergh P, Harbaooui Y & Lemeur R (1981) Mass propagation of
globe artichoke (Cynara scolymus) / evaluation of different hy-
potheses to overcome vitrification with special reference to water
potential. Physiol. Plant. 53: 181187
Escalant J-V, Teisson C & Cte F (1994) Amplified somatic em-
bryogenesis from male flowers of triploid banana and plantain
cultivars (Musa spp.). In Vitro Cell. Dev. Biol. 30: 181186
Escalona M, Lorenzo JC, Gonzlez B, Daquinta M, Fundora Z,
Borroto CG, Espinosa D, Arias E & Aspiolea ME (1998) New
system for in vitro propagation of pineapple (Ananas comosus L.
Merr). Pineapple News 5: 57
Escalona M, Lorenzo JC, Gonzlez B, Daquinta M, Gonzlez JL,
Desjardins Y & Borroto CG (1999) Pineapple (Ananas comosus
L. Merr) micropropagation in temporary immersion systems.
Plant Cell Rep. 18: 743748
Etienne H, Berger A & Carron MP (1991) Water status of callus
from Hevea brasiliensis during induction of somatic embryogen-
esis. Physiol. Plant. 82: 213218
Etienne H, Bertrand B, Anthony F, Cte F & Berthouly M
(1997a) Lembryogense somatique: un outil pour lamlioration
gntique du cafier. In: ASIC Publishers (eds) 17th Interna-
tional Scientific Colloquiun on Coffee, Nairobi (pp 457465),
Vevey
Etienne H, Lartaud M, MichauxFerrire N, Carron MP, Berthouly
M & Teisson C (1997b) Improvement of somatic embryogen-
esis in Hevea brasiliensis (Mll. Arg.) using the temporary
immersion technique. In Vitro Cell. Dev. Biol. 33: 8187
Etienne H, Montoro P, MichauxFerrire N & Carron MP (1993)
Effects of desiccation, medium osmolarity and abscisic acid on
the maturation of Hevea brasiliensis somatic embryos. J. Exp.
Bot. 44: 16131619
EtienneBarry D, Bertrand B, Vsquez N & Etienne H (1999) Dir-
ect sowing of Coffea arabica somatic embryos mass-produced
in a bioreactor and regeneration of plants. Plant Cell Rep. 19:
111117
Gawel NJ & Robacker CD (1990) Somatic embryogenesis in
two Gossypium hirsutum genotypes on semi-solid versus liquid
proliferation media. Plant Cell Tiss. Org. Cult. 23: 201204
Hamilton R, Pederson H & Chin CK (1985) Plant tissue culture on
membrane rafts. BioTechniques. March/April (p 96)
Hammerschlag F (1982) Factors affecting establishment and growth
of peach shoots in vitro. HortScience 17: 8586
Harris RE & Mason EB (1983) Two machines for in vitro propaga-
tion of plants in liquid media. Can. J. Plant Sci. 63: 311316
Harris RE & Stevenson JH (1982) In vitro propagation of Vitis. Vitis
21: 2232
Hussey G (1986) Problems and prospects in the in vitro propagation
of herbaceous plants. In: Withers LA & Alderson PG (eds) Plant
Tissue Culture and its Agricultural Applications (pp 6984).
Butterworths, Boston
Jimnez E , Prez J, Gil V, Herrera J, Garca Y & Alonso E (1995)
Sistema para la propagacin de la caa de azucar. In: Estrade
M, Riego E, Limonta E, Tellez P & Fuente J (eds) Avances
en Biotecnologa Moderna, Vol 3 (pp 1120). Elfos Scientiae,
Havana
Jones AM & Petolino JF (1988) Effects of support medium on
embryo and plant production from cultured anthers of soft-red
winter wheat. Plant Cell Tiss. Org. Cult. 12: 243261
Kitto SL (1997) Commercial micropropagation. HortScience. 32:
10121014
Krueger S, Robacker C & Simonton W (1991) Culture of Amel-
anchier grandiflora in a programmable micropropagation
apparatus. Plant Cell Tiss. Org. Cult. 27: 219226
Liu CM, Xu Z-H & Chua N-H (1993) Auxin polar transport is es-
sential for the establishment of bilateral symmetry during early
plant embryogenesis. The Plant Cell 5: 621630
Lorenzo JC, Gonzalez BL, Escalona M, Teisson C, Espinosa P &
Borroto C (1998) Sugarcane shoot formation in an improved
temporary immersion system. Plant Cell Tiss. Org. Cult. 54:
197200
Maene L & Debergh P (1985) Liquid medium additions to estab-
lished tissue cultures to improve elongation and rooting in vivo.
Plant Cell Tiss. Org. Cult. 5: 2333
Martre P, Lacan D, Just D & Teisson C (2001) Physiological ef-
fects of temporary immersion on Hevea brasiliensis (Mll. Arg.)
callus. Plant Cell Tiss. Org. Cult. 67: 2535
Monette PL (1983) Influence of size of culture vessel on in vitro
proliferation of grape in a liquid medium. Plant Cell Tiss. Org.
Cult. 2: 327332
Noriega C & Sndahl MR (1993). Arabica coffee micropropagation
through somatic embryogenesis via bioreactors. In: ASIC pub-
lishers (eds) 15th International Scientific Colloquium on Coffee,
Montpellier, France (pp 7381). Vevey
Reuther G (1985) Principles and application of the micropropaga-
tion of ornemental plants. In: Schfer-Menuhr A (ed) In Vitro
Techniques: Propagation and Long-Term Storage (pp 114).
Martinus Nijhoff, Dordrecht, The Netherlands

Simonton W & Robacker C (1988) Alternative system for micro-
propagation. Amer. Soc. Agr. Engineers, Technical Paper N
881028
Simonton W, Robacker C & Krueger S (1991) A programmable mi-
cropropagation apparatus using cycled medium. Plant Cell Tiss.
Org. Cult. 27: 211218
Sluis CJ & Walker KA (1985) Commercialization of plant tissue
culture propagation. Intl. Assoc. Plant Tiss. Cult. Newsl. 47: 2
12
Smith DR (1985) Pinus radiata. In: Bajaj YPS (ed) Trees. Biotech-
nology in Agriculture and Forestry, Vol 1 (pp 274291). Springer
Verlag, Berlin
Smith MAL & Spomer LA (1995) Vessels, gels, liquid media and
support systems. In: Aitken-Christie J, Kozai T & Smith MAL
(eds) Automation and Environmental Control in Plant Tissue
Culture (pp 371405). Kluwer Academic Publishers, Dordrecht,
The Netherlands
Sndhal MR, Nakamura T, MedinaFilho HP, Carvalho A, Fazuoli
LC & Costa WM (1984) Coffee. In: Ammirato PV, Evans DA,
Sharp WR & Yamada Y (eds) Handbook of Plant Cell Cul-
ture, Vol III Crop Species (pp 564590). Mac Millan Publishing
Company, New York
Stevenson JH & Harris RE (1980) In vitro plantlet formation from
shoot tip explants of Fuchsia hybrida cv. Swingtime. Can. J. Bot.
58: 21902192
Steward FC, Caplin S & Millar FK (1952) Investigations on growth
and metabolism of plant cells. I. New techniques for the invest-
igation of metabolism, nutrition and growth in undifferentiated
cells. Ann. Bot. 16: 5777
Teisson C & Alvard D (1995) A new concept of plant in vitro cul-
tivation liquid medium: Temporary immersion. In: M Terzi et
al. (eds) Current Issues in Plant Molecular and Cellular Biology
(pp 105110). Kluwer Academic Publishers, Dordrecht, The
Netherlands
Teisson C & Alvard D (1999) In vitro production of potato micro-
tubers in liquid medium using temporay immersion. Potato Res.
42: 499504
231
Teisson C, Alvard D, Lartaud M, Etienne H, Berthouly M, Escalona
M & Lorenzo JC (1999) Temporary immersion for plant tissue
culture. In: Plant Biotechnology and In vitro Biology in the 21st
Century, Proceedings of the IXth International Congress of Plant
Tissue and Cell Culture, Section H: Novel micropropagation
methods (pp 629632). Jerusalem
Tisserat B, Jones D & Galletta PD (1993) Construction and use of an
inexpensive in vitro ultrasonic misting system. HortTechnology
3: 7579
Tisserat B & Vandercook CE (1985) Development of an automated
plant culture system. Plant Cell Tiss. Org. Cult. 5: 107117
Wardle K, Dobbs EB & Short KC (1983) In vitro acclimatization
of aseptically cultured plantlets to humidity. J. Amer. Soc. Hort.
Sci. 108: 386389
Wataad A, Raghothana KG, Kochba M, Nissim A & Caba V (1997)
Micropropagation of Spathiphyllum and Syngonium is facilited
by use of interfacial membrane rafts. HortScience 32: 307308
Weathers PJ, Cheetham RD & Giles KL (1988) Dramatic in-
creases in shoot number and lengths for Musa, Cordyline and
Nephrylepsis [sic] using nutrient mists. Acta Hortic. 230: 3944
Weathers PJ & Giles KL (1988) Regeneration of plants using
nutrient mist culture. In vitro Cell. Dev. Biol. 24: 727732
Zamarripa A, Ducos JP, Tessereau H, Bollon H, Eskes AB &
Petiard V (1991) Dveloppement dun procd de multiplication
de masse en masse du cafier par embryogense somatique en
milieu liquide. In: ASIC Publishers (eds) 14th International Sci-
entific Colloquium on Coffee, San Francisco, US (pp 392402).
Vevey
Ziv M (1995) The control of bioreactor environment for plant
propagation in liquid culture. Acta Hortic. 393: 2538
Ziv M, Meir G & Halevy AH (1983) Factors influencing the produc-
tion of hardened glaucous carnation plantlets in vitro. Plant Cell
Tiss. Org. Cult. 2: 5565