REPRODUCTION, DEVELOPMENT AND LIFE-HISTORY TRAITS
415
larvae or direct development (reviewed by Jablonski
and Lutz, 1983). Planktotrophic development requires
that larvae find food before they exhaust the energy
and nutrients sequestered by the mother in the egg,
so most planktotrophic animals depend either directly
or indirectly upon phytoplankton. Those species that
do not produce larvae, including peracarid crustaceans,
must disperse as adults or juveniles.
It is generally supposed that dispersal should be
less crucial for the maintenance of genetic diversity
in stable habitats than in habitats where there is
a high likelihood of environmental change and a
commensurate requirement for adaptive change in
the genome. However, Hamilton and May (1977)
have disputed this point with a model showing the
population-wide importance of successfully competing
for distant sites. Long-distance dispersal should not
be advantageous where densities remain too low to
foster competition among siblings or between siblings
and offspring. Disadvantages of larval dispersal include
high planktonic mortality and the improbability of
locating a site as good as that of one’s parents.
Predictions
One might predict that long-distance dispersal of
larvae would not be favored by natural selection on
the abyssal plain because great expanses of uniform
substratum would make colonization of new habitats
unnecessary. Moreover, in habitats like seamounts
where a small area of suitable substratum is surrounded
by extensive areas of marginal or unsuitable habitat,
selection should favor only limited dispersal. Large
eggs, parental protection and lecithotrophic development should be found in those deep-sea habitats
where dispersal is not critical to the survival of the
species. However, in ephemeral deep-sea habitats such
as hydrothermal vents, the temporal variability of the
system might compensate for the danger of gamete loss
to unsuitable habitat, favoring the evolution of larval
dispersal mechanisms.
Because the density of living phytoplankton declines
rapidly below the deep chlorophyll maximum at
about 200 m, and they are virtually absent below the
permanent thermocline, it has been predicted that
planktotrophy should not be a viable strategy for deepsea animals (Thorson, 1946, 1950). Thorson argued
that it would be impossible for a microscopic larva
to migrate all the way to the euphotic zone without
feeding, and Scheltema (1994) provided calculations in
support of this contention.
Etter and Caswell (1994) used a cellular automaton
model to explore the advantages of dispersal in habitats
with various levels of disturbance in a competitive
system. The model predicts that long-distance dispersal
would only be advantageous at certain disturbance frequencies, and that the range of these frequencies would
be influenced by reproductive output, competition and
the intensity of disturbance. If it is assumed that the
frequency of disturbance decreases as a function of
depth down the slope, then the greatest advantages to
dispersal may be expected at some intermediate depth,
which would correspond to an intermediate level of
disturbance.
Evidence
As indicated in Tables 12.1 to 12.6, planktotrophic
larvae are produced by animals in many phyla living
at slope depths and planktotrophy is common in
some groups (notably prosobranch gastropods) even at
abyssal depths. The mechanisms of nutrition for these
larvae remain generally unknown. They may use nutritional resources available at depth (bacteria, dissolved
organic matter, detrital seston) or they may migrate
to a depth where sufficient food is available. Both
possibilities have been discussed in the literature and
explored experimentally. Young et al. (1997) presented
a mathematical model incorporating swimming speeds,
respiration rates and plankton concentrations to predict
the likelihood that larvae could reach the euphotic
zone by migrating from bathyal depths. When run
with reasonable parameter estimates taken from deepwater echinoderm larvae where data were available
and shallow-water larvae when the former were not
available, the model showed that migration could
be possible for many species if they were respiring
at low rates and swimming at high (but realistic)
speeds. However, larvae of some of these species
cannot tolerate the temperatures or pressures found at
or near the level where phytoplankton are abundant
(Young and Cameron, 1989; Young and Tyler, 1993;
Young et al., 1996a, 1997; Tyler and Dixon, 2000).
Preliminary data (W. Jaeckle, A. Pile, I. Bosch and
C. Young, unpublished) show that larvae of bathyal
echinoderms can absorb dissolved organic matter and
remove heterotrophic bacteria from the water column
at relatively high rates. These various pieces of indirect
evidence indicate that planktotrophic development may
be a viable developmental strategy in the deep sea, even
for species that cannot migrate to the euphotic zone.
Compelling evidence that some larvae migrate
415
larvae or direct development (reviewed by Jablonski
and Lutz, 1983). Planktotrophic development requires
that larvae find food before they exhaust the energy
and nutrients sequestered by the mother in the egg,
so most planktotrophic animals depend either directly
or indirectly upon phytoplankton. Those species that
do not produce larvae, including peracarid crustaceans,
must disperse as adults or juveniles.
It is generally supposed that dispersal should be
less crucial for the maintenance of genetic diversity
in stable habitats than in habitats where there is
a high likelihood of environmental change and a
commensurate requirement for adaptive change in
the genome. However, Hamilton and May (1977)
have disputed this point with a model showing the
population-wide importance of successfully competing
for distant sites. Long-distance dispersal should not
be advantageous where densities remain too low to
foster competition among siblings or between siblings
and offspring. Disadvantages of larval dispersal include
high planktonic mortality and the improbability of
locating a site as good as that of one’s parents.
Predictions
One might predict that long-distance dispersal of
larvae would not be favored by natural selection on
the abyssal plain because great expanses of uniform
substratum would make colonization of new habitats
unnecessary. Moreover, in habitats like seamounts
where a small area of suitable substratum is surrounded
by extensive areas of marginal or unsuitable habitat,
selection should favor only limited dispersal. Large
eggs, parental protection and lecithotrophic development should be found in those deep-sea habitats
where dispersal is not critical to the survival of the
species. However, in ephemeral deep-sea habitats such
as hydrothermal vents, the temporal variability of the
system might compensate for the danger of gamete loss
to unsuitable habitat, favoring the evolution of larval
dispersal mechanisms.
Because the density of living phytoplankton declines
rapidly below the deep chlorophyll maximum at
about 200 m, and they are virtually absent below the
permanent thermocline, it has been predicted that
planktotrophy should not be a viable strategy for deepsea animals (Thorson, 1946, 1950). Thorson argued
that it would be impossible for a microscopic larva
to migrate all the way to the euphotic zone without
feeding, and Scheltema (1994) provided calculations in
support of this contention.
Etter and Caswell (1994) used a cellular automaton
model to explore the advantages of dispersal in habitats
with various levels of disturbance in a competitive
system. The model predicts that long-distance dispersal
would only be advantageous at certain disturbance frequencies, and that the range of these frequencies would
be influenced by reproductive output, competition and
the intensity of disturbance. If it is assumed that the
frequency of disturbance decreases as a function of
depth down the slope, then the greatest advantages to
dispersal may be expected at some intermediate depth,
which would correspond to an intermediate level of
disturbance.
Evidence
As indicated in Tables 12.1 to 12.6, planktotrophic
larvae are produced by animals in many phyla living
at slope depths and planktotrophy is common in
some groups (notably prosobranch gastropods) even at
abyssal depths. The mechanisms of nutrition for these
larvae remain generally unknown. They may use nutritional resources available at depth (bacteria, dissolved
organic matter, detrital seston) or they may migrate
to a depth where sufficient food is available. Both
possibilities have been discussed in the literature and
explored experimentally. Young et al. (1997) presented
a mathematical model incorporating swimming speeds,
respiration rates and plankton concentrations to predict
the likelihood that larvae could reach the euphotic
zone by migrating from bathyal depths. When run
with reasonable parameter estimates taken from deepwater echinoderm larvae where data were available
and shallow-water larvae when the former were not
available, the model showed that migration could
be possible for many species if they were respiring
at low rates and swimming at high (but realistic)
speeds. However, larvae of some of these species
cannot tolerate the temperatures or pressures found at
or near the level where phytoplankton are abundant
(Young and Cameron, 1989; Young and Tyler, 1993;
Young et al., 1996a, 1997; Tyler and Dixon, 2000).
Preliminary data (W. Jaeckle, A. Pile, I. Bosch and
C. Young, unpublished) show that larvae of bathyal
echinoderms can absorb dissolved organic matter and
remove heterotrophic bacteria from the water column
at relatively high rates. These various pieces of indirect
evidence indicate that planktotrophic development may
be a viable developmental strategy in the deep sea, even
for species that cannot migrate to the euphotic zone.
Compelling evidence that some larvae migrate