255
concerning the origin of parental care and the evolutionary
enlargement of oocytes. Some authors think that the former
preceded the latter (Shine 1978 , 1989 ) and others, that the
reverse is true (Nussbaum 1985 , 1987 ; Summers et al. 2006 ).
Thus, theoretically the simplest variants of brooding might
have evolved in species with planktotrophic larvae. For
instance, some Streblospio polychaetes (Levin 1984 ; Levin
and Bridges 1995 ; Pernet and McArthur 2006 ) and the
kamptozoan Loxosomella elegans (Nielsen 1998 ) brood
small eggs that develop into planktotrophic larvae. Also,
some phoronid species brood their embryos within the tentacle crown for several days, after which they leave the parent organism to develop into planktotrophic actinotrocha
larvae (Silén 1954 ; Emig 1982 , 1983 ; Zimmer 1991 ). In this
phylum the smallest eggs are produced by non-brooding species, however.
An interesting example in this respect is provided by the
reproductive pattern in Tendra zostericola , monotypic for
the genus (Tendridae) ( Electra pontica Gruncharova, 1980
is apparently synonymous). In the Black Sea this species
co- exists with Electra repiachowi , which produces
cyphonautes larvae. In contrast, T . zostericola (morphologically very close to Electra ) produces ciliated coronate larvae whose early development occurs in the space between
the membranous frontal wall of the zooid and overarching
protective mural spines of the acanthostegal brood chamber
(Ostroumoff 1886a , b ; Braiko 1967 ; see also Sect. 2.3.5 ).
During reproduction, small oocytes (70 μm in diameter)
ovulate and accumulate in the coelom (4–10 in number; see
Nordmann 1839 ; Paltschikowa-Ostroumowa 1926 ; Braiko
1967 ; pers. obs.) of the maternal autozooid (Repiachoff
1875 ). After that they are transferred via the intertentacular
organ from the zooid cavity into the brood chamber. It is
unknown whether the ovary continues to form oocytes after
oviposition. Each zygote develops into a ciliated larva with
a non-functioning gut (Repiachoff 1875 , 1878 ; Ostroumoff
1886b ). According to Braiko ( 1967 ), the larvae of T . zostericola develop in the brood chamber in less than 10 h, whereas
Paltschikova- Ostroumowa ( 1926 ) wrote that they leave the
brood chamber after 2 days, meaning in both cases that larval development is almost as fast as in the cyphonautes
embryo before it starts feeding. As with most bryozoan
endotrophic larvae, those of T . zostericola swim from 6–8
(warm water) to 24 h (cold water) before settlement (Braiko
1967 ). It should be also stressed that embryonic development is successfully accomplished outside the brood chamber in experiments (Braiko 1967 ).
These data provide evidence that the reproductive mode
of Tendra recapitulates an early stage in the evolution of
reproductive pattern II, showing a number of transitional
traits between broadcasters with planktotrophic larvae and
brooders with lecithotrophic larvae. On the one hand, Tendra
is morphologically very close to Electra , producing similar
number of small eggs of similar size. On the other hand,
these eggs possess enough yolk for larval development without feeding. Tendra also has a primitive and independently
evolved brood chamber in which several embryos are
brooded simultaneously, developing to non-feeding larvae
with a rudimentary gut. Embryogenesis occurs in the water
entering the brood cavity in this species, which can be also
be adduced as a primitive trait since embryos of advanced
cheilostome brooders die when removed from ovicells to sea
water (see Silén 1945 ).
Structural and reproductive similarities between Tendra
and malacostegans may also demonstrate the possible mode
of transition to brooding in species with planktotrophic larvae (similar to what occurs in phoronids and some polychaetes, see above). For instance, instead of spawned zygotes
exiting into the water, the polypide might allow them to exit
onto the spine-fl anked frontal surface of the distal zooid,
which further transformed into an acanthostegal brood chamber protecting the eggs from predators and/or silting. In contrast to extant brooders with endotrophic larva, ancient
brooders may have produced planktotrophic larvae in which
only the early stages of embryonic development took place
in such primitive brood chambers. The early stages of planktotrophic larval development have high development rates in
malacostegans. For instance, in Conopeum seurati , embryos
begin to move inside the fertilization envelope as early as 8 h
after spawning, leaving the envelope after 9 h. Early embryos
of Einhornia crustulenta (Electridae) complete this stage in
12 h. The gut becomes visible in the cyphonautes of the former species 32 h after the start of development (Cook 1962 ).
In Membranipora serrilamella the embryo begins to swim
slowly, while still enclosed within the fertilization envelope,
less than 24 h after spawning (using groups of cilia that protrude through openings in the envelope). It starts to feed
2 days after spawning (Mawatari 1975 ). So, early cyphonautes
larvae could leave the brood chamber (had they originated
before endotrophy) very early, for instance, within the fi rst
2 days, as do the larvae of Tendra . In this case the transition
to a new mode of oogenesis and hence to an lecithotrophic
larva could occur in the future, after the evolution of brooding, which would have increased survival of the still rather
numerous offspring.
Nevertheless, there are no living bryozoans with brood
chambers and planktotrophic larvae. If we consider that
brooding might have compensated for a reduction in the
number of offspring during the transition to lecithotrophy,
the lack of brooders with cyphonautes larvae may shed
light on the question of what came fi rst: lecithotrophy or
parental care? Species of Conopeum have fewer oocytes than
other malacostegans, and, possibly, relatively short-lived
cyphonautes larvae (Cook 1962 ; Dudley 1973 ). This perhaps
indicates an evolutionary trend towards a change in oogenesis mode accompanied by a reduction of the larval feeding
3.4 Causes, Stages and Consequences of Transition to Endotrophy in Cheilostomata and Ctenostomata
concerning the origin of parental care and the evolutionary
enlargement of oocytes. Some authors think that the former
preceded the latter (Shine 1978 , 1989 ) and others, that the
reverse is true (Nussbaum 1985 , 1987 ; Summers et al. 2006 ).
Thus, theoretically the simplest variants of brooding might
have evolved in species with planktotrophic larvae. For
instance, some Streblospio polychaetes (Levin 1984 ; Levin
and Bridges 1995 ; Pernet and McArthur 2006 ) and the
kamptozoan Loxosomella elegans (Nielsen 1998 ) brood
small eggs that develop into planktotrophic larvae. Also,
some phoronid species brood their embryos within the tentacle crown for several days, after which they leave the parent organism to develop into planktotrophic actinotrocha
larvae (Silén 1954 ; Emig 1982 , 1983 ; Zimmer 1991 ). In this
phylum the smallest eggs are produced by non-brooding species, however.
An interesting example in this respect is provided by the
reproductive pattern in Tendra zostericola , monotypic for
the genus (Tendridae) ( Electra pontica Gruncharova, 1980
is apparently synonymous). In the Black Sea this species
co- exists with Electra repiachowi , which produces
cyphonautes larvae. In contrast, T . zostericola (morphologically very close to Electra ) produces ciliated coronate larvae whose early development occurs in the space between
the membranous frontal wall of the zooid and overarching
protective mural spines of the acanthostegal brood chamber
(Ostroumoff 1886a , b ; Braiko 1967 ; see also Sect. 2.3.5 ).
During reproduction, small oocytes (70 μm in diameter)
ovulate and accumulate in the coelom (4–10 in number; see
Nordmann 1839 ; Paltschikowa-Ostroumowa 1926 ; Braiko
1967 ; pers. obs.) of the maternal autozooid (Repiachoff
1875 ). After that they are transferred via the intertentacular
organ from the zooid cavity into the brood chamber. It is
unknown whether the ovary continues to form oocytes after
oviposition. Each zygote develops into a ciliated larva with
a non-functioning gut (Repiachoff 1875 , 1878 ; Ostroumoff
1886b ). According to Braiko ( 1967 ), the larvae of T . zostericola develop in the brood chamber in less than 10 h, whereas
Paltschikova- Ostroumowa ( 1926 ) wrote that they leave the
brood chamber after 2 days, meaning in both cases that larval development is almost as fast as in the cyphonautes
embryo before it starts feeding. As with most bryozoan
endotrophic larvae, those of T . zostericola swim from 6–8
(warm water) to 24 h (cold water) before settlement (Braiko
1967 ). It should be also stressed that embryonic development is successfully accomplished outside the brood chamber in experiments (Braiko 1967 ).
These data provide evidence that the reproductive mode
of Tendra recapitulates an early stage in the evolution of
reproductive pattern II, showing a number of transitional
traits between broadcasters with planktotrophic larvae and
brooders with lecithotrophic larvae. On the one hand, Tendra
is morphologically very close to Electra , producing similar
number of small eggs of similar size. On the other hand,
these eggs possess enough yolk for larval development without feeding. Tendra also has a primitive and independently
evolved brood chamber in which several embryos are
brooded simultaneously, developing to non-feeding larvae
with a rudimentary gut. Embryogenesis occurs in the water
entering the brood cavity in this species, which can be also
be adduced as a primitive trait since embryos of advanced
cheilostome brooders die when removed from ovicells to sea
water (see Silén 1945 ).
Structural and reproductive similarities between Tendra
and malacostegans may also demonstrate the possible mode
of transition to brooding in species with planktotrophic larvae (similar to what occurs in phoronids and some polychaetes, see above). For instance, instead of spawned zygotes
exiting into the water, the polypide might allow them to exit
onto the spine-fl anked frontal surface of the distal zooid,
which further transformed into an acanthostegal brood chamber protecting the eggs from predators and/or silting. In contrast to extant brooders with endotrophic larva, ancient
brooders may have produced planktotrophic larvae in which
only the early stages of embryonic development took place
in such primitive brood chambers. The early stages of planktotrophic larval development have high development rates in
malacostegans. For instance, in Conopeum seurati , embryos
begin to move inside the fertilization envelope as early as 8 h
after spawning, leaving the envelope after 9 h. Early embryos
of Einhornia crustulenta (Electridae) complete this stage in
12 h. The gut becomes visible in the cyphonautes of the former species 32 h after the start of development (Cook 1962 ).
In Membranipora serrilamella the embryo begins to swim
slowly, while still enclosed within the fertilization envelope,
less than 24 h after spawning (using groups of cilia that protrude through openings in the envelope). It starts to feed
2 days after spawning (Mawatari 1975 ). So, early cyphonautes
larvae could leave the brood chamber (had they originated
before endotrophy) very early, for instance, within the fi rst
2 days, as do the larvae of Tendra . In this case the transition
to a new mode of oogenesis and hence to an lecithotrophic
larva could occur in the future, after the evolution of brooding, which would have increased survival of the still rather
numerous offspring.
Nevertheless, there are no living bryozoans with brood
chambers and planktotrophic larvae. If we consider that
brooding might have compensated for a reduction in the
number of offspring during the transition to lecithotrophy,
the lack of brooders with cyphonautes larvae may shed
light on the question of what came fi rst: lecithotrophy or
parental care? Species of Conopeum have fewer oocytes than
other malacostegans, and, possibly, relatively short-lived
cyphonautes larvae (Cook 1962 ; Dudley 1973 ). This perhaps
indicates an evolutionary trend towards a change in oogenesis mode accompanied by a reduction of the larval feeding
3.4 Causes, Stages and Consequences of Transition to Endotrophy in Cheilostomata and Ctenostomata