257
when the eggs leave the maternal zooid). At the same time,
this gap is rather small – several older oocytes develop and
mature almost synchronously, as in Malacostegina. Also
Stach ( 1938 ) mentioned the irregular shape of ovulated
oocytes in “ C .” indivisa , which is also known in electrid
broadcasters. We do not know whether “ C .” indivisa and
T . zostericola have nurse cells like those in the majority of
studied cheilostome brooders; if they do not, this must also
indicate that their reproductive mode is an ancient one.
Examples of plesiomorphic simultaneous brooding of
several embryos in species with the most primitive brood
chambers (external membranous sacs and acanthostegal
brood chambers) are instructive. Most of the cheilostomes
with membranous sacs, as in the genera Aetea , Eucratea and
Leiosalpinx , brood a single embryo attached to the maternal
zooid (Fig. 2.52 ). Leiosalpinx australis sometimes has two
embryos (Gordon 1986 ). Cook ( 1977b ) reported that Aetea
anguina in one of the populations studied could have up to
two embryos in the same brood sac. External membranous
sacs may have evolved independently at least three times in
cheilostomes (see Chap. 2 ), and, as in ovicell brooders, the
above examples may point to the tendency towards a gradual
reduction in the number of oocytes (probably because of
their increase in size) in the species with brood sacs.
In conclusion, it should be emphasized that the data presented in this section indicate that both lecithotrophic larvae
and parental care evolved many times in different cheilostome lineages. Brooding evolved independently at least 7–8
times. In all of these cases the ancestors appear to have been
broadcasting malacostegans with planktotrophic larvae. The
acquisition of embryonic incubation was each time accompanied (preceded or followed) by the evolution of a nonfeeding larva.
3.4.2 Fertilization and Modifi cation
of Oogenesis
In the hypothesized scenario concerning the evolution of
brooding from an antecedent broadcasting mode of sexual
reproduction, the transition to macrolecithal oogenesis was
accompanied by a shift to early fertilization, which might
have been a precondition for the origin of nurse cells. These,
in turn, could have additionally enhanced the effectiveness of
vitellogenesis (see Sect. 3.2 ).
Theoretically, the increase in the amount of nutrients contained in oocytes may have had another reason behind it. As
discussed in the review by Wourms ( 1987 ), there may be a
connection between the time of fertilization and the character of oocyte formation. In some invertebrates the fusion of
the male and the female gametes results in dramatic changes
in oogenesis. Some rotifers (for instance, Euchlanis dilatata
and Brachionus rubens ) exhibit enormous differences in the
quantity and quality of yolk in their oocytes depending on
whether or not the female has been inseminated (Gilbert
1983 , 1989 ), and fertilization may well be the reason for
these differences (Gilbert 1989 ). In other words, in these
rotifers the fertilized oocyte somehow infl uences the functioning of the vitellarium (the part of the ovary synthesizing
yolk and transporting them to the oocyte). In another
Brachionus species, B . calycifl orus , sperm are known to fuse
with early oocytes. In other words, the presence of the sperm
may determine the growth character of the female gamete, in
particular, the mode of vitellogenesis.
Although intraovarian fertilization is a generally rare phenomenon, it is obviously obligatory in all Bryozoa incubating
their offspring (in broadcasters the male and female gametes
fuse during ovulation). Thus, its role, especially its infl uence
on oogenesis in Bryozoa, may be considerable. Having in
mind the example of the rotifers, one may suggest that the
entry of sperms into the ovary and subsequent fertilization
there could additionally stimulate vitellogenesis. In brooding
cheilostomes, early fertilization ultimately resulted in the
complete dependence of oogenesis on sperm arrival. Fusion
of sperm with early oocytes became the trigger for vitellogenesis (see also Sects. 1.3.4 and 1.3.6 ). This evolutionary novelty is of paramount importance – the colony does not have to
spend energy and nutrients on vitellogenesis before receiving
the sperm, that is, before fertilization is guaranteed (Bishop
et al. 2000 ). As long as vitellogenesis has not started, the colony may invest more resources into somatic growth, for
example, and a larger colony can fi lter a greater volume of
water, increasing the chances of obtaining alien sperm.
How this trait (dependence of vitellogenesis on syngamy)
was retained during evolution is an intriguing question.
How could an external factor (entering of alien sperm) infl uence the genetic programme in such a way as to prevent vitellogenesis before syngamy? Crucial prerequisite was an ability
of the young oocyte to fuse with sperm very early (before the
onset of vitellogenesis). Further, early syngamy became an
obligatory stage for oocyte development, and indispensable
for the start of vitellogenesis. In this connection, information
concerning the time of fertilization in Tendra zostericola is
crucial. One may speculate that in this species with many
primitive characters fertilization occurs during ovulation (as it
does in malacostegans) and nurse cells are absent. If this is
true, the onset of vitellogenesis in Tendra , contrary to all other
brooding cheilostomes, does not depend on fertilization.
Intraovarian fertilization is unlikely to infl uence vitellogenesis in Ctenostomata, however, as indicated by the fact
that in the ctenostome Bowerbankia gracilis late macrolecithal oocytes are fertilized (Temkin 1996 ).
Another important consideration is that internal fertilization would seem to be a necessary prerequisite for the incubation of embryos during the evolution of invertebrates
(Ryland and Bishop 1993 ). In the view of Temkin ( 1994 ),
3.4 Causes, Stages and Consequences of Transition to Endotrophy in Cheilostomata and Ctenostomata
when the eggs leave the maternal zooid). At the same time,
this gap is rather small – several older oocytes develop and
mature almost synchronously, as in Malacostegina. Also
Stach ( 1938 ) mentioned the irregular shape of ovulated
oocytes in “ C .” indivisa , which is also known in electrid
broadcasters. We do not know whether “ C .” indivisa and
T . zostericola have nurse cells like those in the majority of
studied cheilostome brooders; if they do not, this must also
indicate that their reproductive mode is an ancient one.
Examples of plesiomorphic simultaneous brooding of
several embryos in species with the most primitive brood
chambers (external membranous sacs and acanthostegal
brood chambers) are instructive. Most of the cheilostomes
with membranous sacs, as in the genera Aetea , Eucratea and
Leiosalpinx , brood a single embryo attached to the maternal
zooid (Fig. 2.52 ). Leiosalpinx australis sometimes has two
embryos (Gordon 1986 ). Cook ( 1977b ) reported that Aetea
anguina in one of the populations studied could have up to
two embryos in the same brood sac. External membranous
sacs may have evolved independently at least three times in
cheilostomes (see Chap. 2 ), and, as in ovicell brooders, the
above examples may point to the tendency towards a gradual
reduction in the number of oocytes (probably because of
their increase in size) in the species with brood sacs.
In conclusion, it should be emphasized that the data presented in this section indicate that both lecithotrophic larvae
and parental care evolved many times in different cheilostome lineages. Brooding evolved independently at least 7–8
times. In all of these cases the ancestors appear to have been
broadcasting malacostegans with planktotrophic larvae. The
acquisition of embryonic incubation was each time accompanied (preceded or followed) by the evolution of a nonfeeding larva.
3.4.2 Fertilization and Modifi cation
of Oogenesis
In the hypothesized scenario concerning the evolution of
brooding from an antecedent broadcasting mode of sexual
reproduction, the transition to macrolecithal oogenesis was
accompanied by a shift to early fertilization, which might
have been a precondition for the origin of nurse cells. These,
in turn, could have additionally enhanced the effectiveness of
vitellogenesis (see Sect. 3.2 ).
Theoretically, the increase in the amount of nutrients contained in oocytes may have had another reason behind it. As
discussed in the review by Wourms ( 1987 ), there may be a
connection between the time of fertilization and the character of oocyte formation. In some invertebrates the fusion of
the male and the female gametes results in dramatic changes
in oogenesis. Some rotifers (for instance, Euchlanis dilatata
and Brachionus rubens ) exhibit enormous differences in the
quantity and quality of yolk in their oocytes depending on
whether or not the female has been inseminated (Gilbert
1983 , 1989 ), and fertilization may well be the reason for
these differences (Gilbert 1989 ). In other words, in these
rotifers the fertilized oocyte somehow infl uences the functioning of the vitellarium (the part of the ovary synthesizing
yolk and transporting them to the oocyte). In another
Brachionus species, B . calycifl orus , sperm are known to fuse
with early oocytes. In other words, the presence of the sperm
may determine the growth character of the female gamete, in
particular, the mode of vitellogenesis.
Although intraovarian fertilization is a generally rare phenomenon, it is obviously obligatory in all Bryozoa incubating
their offspring (in broadcasters the male and female gametes
fuse during ovulation). Thus, its role, especially its infl uence
on oogenesis in Bryozoa, may be considerable. Having in
mind the example of the rotifers, one may suggest that the
entry of sperms into the ovary and subsequent fertilization
there could additionally stimulate vitellogenesis. In brooding
cheilostomes, early fertilization ultimately resulted in the
complete dependence of oogenesis on sperm arrival. Fusion
of sperm with early oocytes became the trigger for vitellogenesis (see also Sects. 1.3.4 and 1.3.6 ). This evolutionary novelty is of paramount importance – the colony does not have to
spend energy and nutrients on vitellogenesis before receiving
the sperm, that is, before fertilization is guaranteed (Bishop
et al. 2000 ). As long as vitellogenesis has not started, the colony may invest more resources into somatic growth, for
example, and a larger colony can fi lter a greater volume of
water, increasing the chances of obtaining alien sperm.
How this trait (dependence of vitellogenesis on syngamy)
was retained during evolution is an intriguing question.
How could an external factor (entering of alien sperm) infl uence the genetic programme in such a way as to prevent vitellogenesis before syngamy? Crucial prerequisite was an ability
of the young oocyte to fuse with sperm very early (before the
onset of vitellogenesis). Further, early syngamy became an
obligatory stage for oocyte development, and indispensable
for the start of vitellogenesis. In this connection, information
concerning the time of fertilization in Tendra zostericola is
crucial. One may speculate that in this species with many
primitive characters fertilization occurs during ovulation (as it
does in malacostegans) and nurse cells are absent. If this is
true, the onset of vitellogenesis in Tendra , contrary to all other
brooding cheilostomes, does not depend on fertilization.
Intraovarian fertilization is unlikely to infl uence vitellogenesis in Ctenostomata, however, as indicated by the fact
that in the ctenostome Bowerbankia gracilis late macrolecithal oocytes are fertilized (Temkin 1996 ).
Another important consideration is that internal fertilization would seem to be a necessary prerequisite for the incubation of embryos during the evolution of invertebrates
(Ryland and Bishop 1993 ). In the view of Temkin ( 1994 ),
3.4 Causes, Stages and Consequences of Transition to Endotrophy in Cheilostomata and Ctenostomata