253
possibly are not suffi cient to provide a positive balance between
survival and mortality. Theoretically, the consequences of a
decrease in the number of offspring might be compensated by
embryonic incubation, enabling development inside the
maternal organism or in specialized brooding structures. In
both cases, the time spent in the water column, the most
hazardous for the young organism, is drastically shortened.
The suggested connection between incubation and a
decrease in the number of young corresponds to the conclusion made by Smith and Fretwell ( 1974 ) – the more energy
(including parental care) is allocated to an offspring, the better are its chances for survival (also discussed in Emlen
1973 ; Strathmann 1978b ; Poulin and Féral 1996 ). Moreover,
according to Picken ( 1980 ), a free larval stage is absent in
some species with protected development, since they produce relatively few ova, and incubation ensures a high level
of offspring survival. A similar correlation between larger
size and smaller number of eggs and parental care was
described in fi shes (discussed in Balon 1991 ).
As for invertebrates, most ophiuroids that brood their
young are characterized by reduced fecundity (Byrne 1991a ).
In this group embryonic incubation supposedly evolved in
relation to the acquisition of larger eggs, non-feeding larvae
and smaller adult size (Byrne 1991a ; Byrne et al. 2008 ).
Considering bivalve mollusks, Sellmer ( 1967 ) wrote that
incubation is an evolutionary adaptation to having a reduced
number of young, which in turn is connected to the small
parental size (see also Mackie 1984 ). Zarenkov ( 1982 ) noted
that the decrease in body size characteristic of the crustacean
subclass Copepoda is associated with reduced productivity,
such species tending to evolve parental care. Incubation and
short-lived larvae are characteristic of many colonial epibiotic invertebrates. Notably in ascidians, almost all colonial
species (with smaller zooids) have parental care, whereas
solitary species do not (Strathmann and Strathmann 1982 ;
Strathmann 1990 ). Strathmann ( 1978b , 1990 , 1986 ) explained
the association of brooding with small adult size by the
necessity of a normal oxygen supply to the embryos (see also
Strathmann et al. 1984 ). According to this relationship, since
fecundity increases disproportionally with surface area as
adult size increases, larger animals are less capable of
successfully brooding their offspring.
On the whole, internal incubation, often associated with
viviparity, is usual in groups of small-sized invertebrates
(Levin and Bridges 1995 ; see also the review of hypotheses
in Ghiselin 1987 ). Taking into account the microscopic size
of bryozoans and the relatively small number of oocytes (and
even smaller number of brooded larvae) formed by zooids in
brooding species, the evolution of embryonic incubation
appears to have been an extremely important, possibly crucial, event in bryozoan evolution, allowing them to compensate for the reduction in the numbers of offspring during the
transition from planktotrophy to lecithotrophy.
Finally, when analyzing the hypothesis that is in question
here, Strathmann and Strathmann ( 1982 ) asked why brooding is also not so typical of large animals. Indeed, numerous
small eggs and planktotrophic larvae usually develop in
larger broadcasting species whereas smaller species are normally brooders producing relatively large eggs and nonfeeding larvae (reviewed in Olive 1985 ). Following the idea
of Chia ( 1974 ), I suggest that non-brooding Echinodermata
with non-feeding larvae compensate for the decrease in the
number of oocytes accompanying the evolution of lecithotrophy by the larger size of the maternal individual and thus
by the greater number of gametes it produces. In other words,
for relatively large animals, with their numerous eggs, the
decrease in the number of oocytes associated with an increase
in size is not so risky as it is for smaller animals.
A shift in oogenesis (reduction in egg number and increase
in their size) and parental care can apparently evolve in the
cheilostomes sequentially, with a short time lag. One can
argue that oogenesis becomes modifi ed fi rst, with the
decrease in the number of offspring caused by it compensated soon after by the origin of brooding. Wray ( 1995a ) also
suggested that lecithotrophy preceded the origin of brooding.
Besides, the above-described independent multiple origin of
brood chambers within Cheilostomata (see Chap. 2 ) indicates that the non-feeding larva and thus the new mode of
oogenesis also evolved several times. In my opinion, brooding originated in cheilostomes every time oogenesis was
altered within the broadcasting basal clades. If macrolecithal
oogenesis, and thus non-feeding larvae, evolved only once in
the evolutionary history of cheilostomes, then it is indeed
puzzling why a single extant species combining broadcasting and lecithotrophy has not survived (see below for detailed
analysis).
3.4.1.1 Multiple Origins of Lecithotrophy
in Cheilostomata
Judging from the patchy distribution of planktotrophy in the
phylogenetic scheme of the bryozoan order Ctenostomata
(Todd 2000 ), brooding and lecithotrophy may have originated at least fi ve times in this group. I suggest the same
happened in the order Cheilostomata, which acquired brooding (and thus lost planktotrophy) independently on several
occasions and at different geological times. Apart from
perhaps the superfamily Aeteoidea (see Jebram 1992 ),
brooding cheilostomes root their ancestry in the suborder
Malacostegina, which comprises only broadcasters and thus
lacks both lecithotrophy and parental care. In general, the
presence of lecithotrophic larvae in Bryozoa is always associated with embryonic incubation, which, pari passu , never
goes hand in hand with a feeding larva. This means that, if
bryozoans with a planktotrophic larva and no parental care
were the ancestors of the clades with independently acquired
embryonic incubation, the endotrophic larva originated as
3.4 Causes, Stages and Consequences of Transition to Endotrophy in Cheilostomata and Ctenostomata
possibly are not suffi cient to provide a positive balance between
survival and mortality. Theoretically, the consequences of a
decrease in the number of offspring might be compensated by
embryonic incubation, enabling development inside the
maternal organism or in specialized brooding structures. In
both cases, the time spent in the water column, the most
hazardous for the young organism, is drastically shortened.
The suggested connection between incubation and a
decrease in the number of young corresponds to the conclusion made by Smith and Fretwell ( 1974 ) – the more energy
(including parental care) is allocated to an offspring, the better are its chances for survival (also discussed in Emlen
1973 ; Strathmann 1978b ; Poulin and Féral 1996 ). Moreover,
according to Picken ( 1980 ), a free larval stage is absent in
some species with protected development, since they produce relatively few ova, and incubation ensures a high level
of offspring survival. A similar correlation between larger
size and smaller number of eggs and parental care was
described in fi shes (discussed in Balon 1991 ).
As for invertebrates, most ophiuroids that brood their
young are characterized by reduced fecundity (Byrne 1991a ).
In this group embryonic incubation supposedly evolved in
relation to the acquisition of larger eggs, non-feeding larvae
and smaller adult size (Byrne 1991a ; Byrne et al. 2008 ).
Considering bivalve mollusks, Sellmer ( 1967 ) wrote that
incubation is an evolutionary adaptation to having a reduced
number of young, which in turn is connected to the small
parental size (see also Mackie 1984 ). Zarenkov ( 1982 ) noted
that the decrease in body size characteristic of the crustacean
subclass Copepoda is associated with reduced productivity,
such species tending to evolve parental care. Incubation and
short-lived larvae are characteristic of many colonial epibiotic invertebrates. Notably in ascidians, almost all colonial
species (with smaller zooids) have parental care, whereas
solitary species do not (Strathmann and Strathmann 1982 ;
Strathmann 1990 ). Strathmann ( 1978b , 1990 , 1986 ) explained
the association of brooding with small adult size by the
necessity of a normal oxygen supply to the embryos (see also
Strathmann et al. 1984 ). According to this relationship, since
fecundity increases disproportionally with surface area as
adult size increases, larger animals are less capable of
successfully brooding their offspring.
On the whole, internal incubation, often associated with
viviparity, is usual in groups of small-sized invertebrates
(Levin and Bridges 1995 ; see also the review of hypotheses
in Ghiselin 1987 ). Taking into account the microscopic size
of bryozoans and the relatively small number of oocytes (and
even smaller number of brooded larvae) formed by zooids in
brooding species, the evolution of embryonic incubation
appears to have been an extremely important, possibly crucial, event in bryozoan evolution, allowing them to compensate for the reduction in the numbers of offspring during the
transition from planktotrophy to lecithotrophy.
Finally, when analyzing the hypothesis that is in question
here, Strathmann and Strathmann ( 1982 ) asked why brooding is also not so typical of large animals. Indeed, numerous
small eggs and planktotrophic larvae usually develop in
larger broadcasting species whereas smaller species are normally brooders producing relatively large eggs and nonfeeding larvae (reviewed in Olive 1985 ). Following the idea
of Chia ( 1974 ), I suggest that non-brooding Echinodermata
with non-feeding larvae compensate for the decrease in the
number of oocytes accompanying the evolution of lecithotrophy by the larger size of the maternal individual and thus
by the greater number of gametes it produces. In other words,
for relatively large animals, with their numerous eggs, the
decrease in the number of oocytes associated with an increase
in size is not so risky as it is for smaller animals.
A shift in oogenesis (reduction in egg number and increase
in their size) and parental care can apparently evolve in the
cheilostomes sequentially, with a short time lag. One can
argue that oogenesis becomes modifi ed fi rst, with the
decrease in the number of offspring caused by it compensated soon after by the origin of brooding. Wray ( 1995a ) also
suggested that lecithotrophy preceded the origin of brooding.
Besides, the above-described independent multiple origin of
brood chambers within Cheilostomata (see Chap. 2 ) indicates that the non-feeding larva and thus the new mode of
oogenesis also evolved several times. In my opinion, brooding originated in cheilostomes every time oogenesis was
altered within the broadcasting basal clades. If macrolecithal
oogenesis, and thus non-feeding larvae, evolved only once in
the evolutionary history of cheilostomes, then it is indeed
puzzling why a single extant species combining broadcasting and lecithotrophy has not survived (see below for detailed
analysis).
3.4.1.1 Multiple Origins of Lecithotrophy
in Cheilostomata
Judging from the patchy distribution of planktotrophy in the
phylogenetic scheme of the bryozoan order Ctenostomata
(Todd 2000 ), brooding and lecithotrophy may have originated at least fi ve times in this group. I suggest the same
happened in the order Cheilostomata, which acquired brooding (and thus lost planktotrophy) independently on several
occasions and at different geological times. Apart from
perhaps the superfamily Aeteoidea (see Jebram 1992 ),
brooding cheilostomes root their ancestry in the suborder
Malacostegina, which comprises only broadcasters and thus
lacks both lecithotrophy and parental care. In general, the
presence of lecithotrophic larvae in Bryozoa is always associated with embryonic incubation, which, pari passu , never
goes hand in hand with a feeding larva. This means that, if
bryozoans with a planktotrophic larva and no parental care
were the ancestors of the clades with independently acquired
embryonic incubation, the endotrophic larva originated as
3.4 Causes, Stages and Consequences of Transition to Endotrophy in Cheilostomata and Ctenostomata