264
membranous sac (Cyclostomata) and the body wall
(Ctenostomata, Cheilostomata, Phylactolaemata).
The examples of Triticella fl ava (Ctenostomata), Tendra
zostericola and “ Carbasea ” indivisa (Cheilostomata), which
have small oocytes and brood several embryos simultaneously, show that the lecithotrophic larva does not require a
considerable increase in the amount of nutrients in oocytes
(as in echinoderms; see Byrne et al. 2003 ), and that the evolution of lecithotrophy is not a very diffi cult evolutionary
step (Christiansen and Fenchel 1979 ; see also below). Since
nutrient resources in the egg are limited, such a larva should
be fast-developing and short-lived. At the same time, the
number of oocytes formed in the ovary remains considerable
despite an increase in the amount of nutrients in the egg. This
combination of characters, together with the presence of
endotrophic larvae with a rudimentary gut, should be considered as transitional from plesiomorphic pattern I towards
more-derived patterns. As with pattern IV, which combines
attributes of patterns II and III, this transitional pattern combines features of patterns I and II.
Both ctenostomes and cheilostomes with reproductive
pattern II show similar “overlaps” of various kinds. The presence of numerous female gametes in the ovary, only some of
which mature and are later brooded in some way, was
described in some ctenostomes but also occurs in cheilostomes (Cribrilinidae, Margarettidae). Some cheilostomes
( Scruparia , Thalamoporella , Macropora , Monoporella )
have multiple brooding, characteristic of ctenostomes from
various families, whereas the development of embryos of
different ages in the ovicells of Thalamoporella is reminiscent of, for example, the ctenostome Nolella . In those cheilostomes that are primitive external brooders ( Aetea , Eucratea ,
Leiosalpinx ), 1–2 embryos are normally incubated at a time
(see Cook 1977b ; Eggleston 1963 ; Gordon 1986 ) and this is
combined with a rather small number of maturing oocytes
(Waters 1896 [1898] ). Reproduction in some ctenostome
species is reminiscent of this variant.
In comparing species with pattern III, the main difference
between the two bryozoan orders is in the number of oocytes
in the ovary. There are many in some ctenostomes and only a
few in cheilostomes. In consideration of the differences in
oogenesis, I suggest that pattern III evolved in these two
gymnolaemate clades on a different basis. In Cheilostomata,
placental analogues apparently fi rst evolved in species with a
few macrolecithal oocytes (pattern IV), and oocytes became
oligolecithal later (pattern III) (see Sect. 3.3 ). A similar transition from pattern II to pattern IV could have also occurred
in some ctenostomes, for instance in Flustrellidra hispida ,
which combines macrolecithal oocytes with extraembryonic
nutrition. Besides, placental analogues evolved independently in some ctenostomes with both large ( Labiostomella
gisleni , Nolela dilatata ), and relatively small ( Walkeria
uva , Zoobotryon verticillatum ) numbers of small oocytes.
I denote this variant as pattern III, based on the fact that it
combines matrotrophy, relatively small oocytes (micro- or
mesolecithal judging from published illustrations) and
noticeable embryonic enlargement. In the former case (combination of matrotrophy with many eggs in ovary), embryonic brooding evolved fi rst, but there was no reduction in egg
number (although only few of them were incubated). Further,
matrotrophy evolved but the quantity of eggs produced
remained high. In the latter case (combination of matrotrophy
with few eggs), the transition to brooding was accompanied
by a reduction in the number of oocytes. In the evolution of
this pattern, as in Walkeria uva and Zoobotryon verticillatum , transitions from patterns II to III and IV to III were both
theoretically possible in ancestors. In this connection we should
once again recall the Phylactolaemata, in which pattern III
evolved independently from Ctenostomata and in which
there are also a relatively large number of small eggs in the
ovary and sequential incubation of individual embryos.
Theoretically, an evolutionary scenario involving matrotrophy in combination with the production of numerous small
oocytes is not to be excluded for brooding cheilostomes, since
several species among them do produce numerous (although
macrolecithal) oocytes. If the oocytes of their ancestors contained fewer nutrient reserves (a transitional pattern characteristic of Tendra zostericola and “ Carbasea ” indivisa ), then
the possibility existed for matrotrophic (possibly multiple)
brooding to evolve in combination with mesolecithal oogenesis. Further changes in oogenesis (transition to fewer macrolecithal oocytes) or extraembryonic nutrition (enhanced
activity of the embryophore accompanied by a reduction in
the number of oocytes) could have resulted in patterns IV and
III, respectively. To emphasize, the above scenario is purely
speculative, since no placental cheilostomes are known to
have numerous oocytes in the ovary.
3.4.5 Environmental Factors and Radiation
of Cheilostomata in the Late Cretaceous
The upper half of the Cretaceous witnessed an explosive
radiation of Cheilostomata (Lidgard et al. 1993 ; Gordon and
Voigt 1996 ; Jablonski et al. 1997 ; Taylor 2000 ), apparently
triggered by the acquisition of a lecithotrophic larva (Taylor
1988a ; see also Taylor and Larwood 1990 ). This novelty
seems to appear several more times in cheilostome history,
being a result of changes in oogenesis and the production of
larger, more nutrient-rich eggs. The ecological factors that
infl uence, directly or indirectly, organismal development
(Schmalhausen 1949 , 1982 ; Jablonski and Lutz 1983 ;
Matsuda 1987 ; Balon 1991 ; McEdward 1995 ; Wray 1995b ),
including larval types (discussed in Wourms 1987 ), supposedly drove egg enlargement. Wray ( 1995a ) listed a number
of factors that might infl uence egg size in invertebrates with
3 Evolution of Reproductive Patterns in Cheilostomata
membranous sac (Cyclostomata) and the body wall
(Ctenostomata, Cheilostomata, Phylactolaemata).
The examples of Triticella fl ava (Ctenostomata), Tendra
zostericola and “ Carbasea ” indivisa (Cheilostomata), which
have small oocytes and brood several embryos simultaneously, show that the lecithotrophic larva does not require a
considerable increase in the amount of nutrients in oocytes
(as in echinoderms; see Byrne et al. 2003 ), and that the evolution of lecithotrophy is not a very diffi cult evolutionary
step (Christiansen and Fenchel 1979 ; see also below). Since
nutrient resources in the egg are limited, such a larva should
be fast-developing and short-lived. At the same time, the
number of oocytes formed in the ovary remains considerable
despite an increase in the amount of nutrients in the egg. This
combination of characters, together with the presence of
endotrophic larvae with a rudimentary gut, should be considered as transitional from plesiomorphic pattern I towards
more-derived patterns. As with pattern IV, which combines
attributes of patterns II and III, this transitional pattern combines features of patterns I and II.
Both ctenostomes and cheilostomes with reproductive
pattern II show similar “overlaps” of various kinds. The presence of numerous female gametes in the ovary, only some of
which mature and are later brooded in some way, was
described in some ctenostomes but also occurs in cheilostomes (Cribrilinidae, Margarettidae). Some cheilostomes
( Scruparia , Thalamoporella , Macropora , Monoporella )
have multiple brooding, characteristic of ctenostomes from
various families, whereas the development of embryos of
different ages in the ovicells of Thalamoporella is reminiscent of, for example, the ctenostome Nolella . In those cheilostomes that are primitive external brooders ( Aetea , Eucratea ,
Leiosalpinx ), 1–2 embryos are normally incubated at a time
(see Cook 1977b ; Eggleston 1963 ; Gordon 1986 ) and this is
combined with a rather small number of maturing oocytes
(Waters 1896 [1898] ). Reproduction in some ctenostome
species is reminiscent of this variant.
In comparing species with pattern III, the main difference
between the two bryozoan orders is in the number of oocytes
in the ovary. There are many in some ctenostomes and only a
few in cheilostomes. In consideration of the differences in
oogenesis, I suggest that pattern III evolved in these two
gymnolaemate clades on a different basis. In Cheilostomata,
placental analogues apparently fi rst evolved in species with a
few macrolecithal oocytes (pattern IV), and oocytes became
oligolecithal later (pattern III) (see Sect. 3.3 ). A similar transition from pattern II to pattern IV could have also occurred
in some ctenostomes, for instance in Flustrellidra hispida ,
which combines macrolecithal oocytes with extraembryonic
nutrition. Besides, placental analogues evolved independently in some ctenostomes with both large ( Labiostomella
gisleni , Nolela dilatata ), and relatively small ( Walkeria
uva , Zoobotryon verticillatum ) numbers of small oocytes.
I denote this variant as pattern III, based on the fact that it
combines matrotrophy, relatively small oocytes (micro- or
mesolecithal judging from published illustrations) and
noticeable embryonic enlargement. In the former case (combination of matrotrophy with many eggs in ovary), embryonic brooding evolved fi rst, but there was no reduction in egg
number (although only few of them were incubated). Further,
matrotrophy evolved but the quantity of eggs produced
remained high. In the latter case (combination of matrotrophy
with few eggs), the transition to brooding was accompanied
by a reduction in the number of oocytes. In the evolution of
this pattern, as in Walkeria uva and Zoobotryon verticillatum , transitions from patterns II to III and IV to III were both
theoretically possible in ancestors. In this connection we should
once again recall the Phylactolaemata, in which pattern III
evolved independently from Ctenostomata and in which
there are also a relatively large number of small eggs in the
ovary and sequential incubation of individual embryos.
Theoretically, an evolutionary scenario involving matrotrophy in combination with the production of numerous small
oocytes is not to be excluded for brooding cheilostomes, since
several species among them do produce numerous (although
macrolecithal) oocytes. If the oocytes of their ancestors contained fewer nutrient reserves (a transitional pattern characteristic of Tendra zostericola and “ Carbasea ” indivisa ), then
the possibility existed for matrotrophic (possibly multiple)
brooding to evolve in combination with mesolecithal oogenesis. Further changes in oogenesis (transition to fewer macrolecithal oocytes) or extraembryonic nutrition (enhanced
activity of the embryophore accompanied by a reduction in
the number of oocytes) could have resulted in patterns IV and
III, respectively. To emphasize, the above scenario is purely
speculative, since no placental cheilostomes are known to
have numerous oocytes in the ovary.
3.4.5 Environmental Factors and Radiation
of Cheilostomata in the Late Cretaceous
The upper half of the Cretaceous witnessed an explosive
radiation of Cheilostomata (Lidgard et al. 1993 ; Gordon and
Voigt 1996 ; Jablonski et al. 1997 ; Taylor 2000 ), apparently
triggered by the acquisition of a lecithotrophic larva (Taylor
1988a ; see also Taylor and Larwood 1990 ). This novelty
seems to appear several more times in cheilostome history,
being a result of changes in oogenesis and the production of
larger, more nutrient-rich eggs. The ecological factors that
infl uence, directly or indirectly, organismal development
(Schmalhausen 1949 , 1982 ; Jablonski and Lutz 1983 ;
Matsuda 1987 ; Balon 1991 ; McEdward 1995 ; Wray 1995b ),
including larval types (discussed in Wourms 1987 ), supposedly drove egg enlargement. Wray ( 1995a ) listed a number
of factors that might infl uence egg size in invertebrates with
3 Evolution of Reproductive Patterns in Cheilostomata