FOOD INPUTS, UTILIZATION, CARBON FLOW AND ENERGETICS
361
of food-chain links has often reflected the intensity of
study of a system (Hall and Rafaelli, 1991). The claim
that longer chains are less stable than shorter ones,
which would encourage severe population fluctations,
has held sway for for nearly two decades in the
ecological literature as a result of models, dating from
May (1973), demonstrating the destabilizing effect of
complexity in food webs. The recognition that in nature
complexity is of common occurrence has encouraged
new work and modelling, which has succeeded in
demonstrating the stabilizing effect of complexity as
deduced by intuition, in terms of an interwoven matrix
of interaction holding the community together. This
does not employ an approach with nonlinear terms in
predator–prey relationships. Predators are constrained
so they cannot maintain high feeding efficiency on
many different prey organisms at the same time, thus
relaxing pressure on low-density resources and so
preventing their extinction. These many weak links,
with few strong interactions, in the system connecting
species in the community dampen oscillations between
consumers and resources, and thus tend to stabilize,
rather than destabilize, the dynamics of the component
species (McCann et al., 1998; Polis, 1998).
Although it has been thought that food-web relationships among individual species in the deep sea are
still too poorly understood to allow construction of
the benthic food web for any single deep-sea station
(Sibuet, 1992), there is steady movement towards the
realization of this goal. Because of high local species
richness, a high degree of taxonomic resolution in
food-web description is difficult, but it seems unlikely
that the number of food-web links approaches those
found in the tropical rain forest, even if similar
sorts of interaction occur. However, one interesting
consequence of the patchy system caused by spatially
uneven detrital deposition (which may also help explain
high species richness – see Rice and Lambshead, 1994)
is that food chains will be shorter in small habitat
patches (Pimm et al., 1991).
Stable-isotope analysis has been applied in order to
track sources in marine food webs involving deposit
feeders (see overview of subject in Michener and
Schell, 1994). Use of such natural tracers involves a
number of assumptions. For example, measurements of
d
13 C may be biased by lipid contents, with differing
rates of
13 C enrichment occurring depending on relative
amount of carbohydrate, proteins and lipids present
in the tissue. It is claimed that
15 N ratios are more
conservative among biochemical fractions, so that
d
13 N is a more reliable tool in tracking trophic
linkages (Hobson and Welch, 1992). However, stableisotope studies conducted in tandem with analysis
of stomach contents seem to provide most certain
data. Iken et al. (2001), using d
13 N and analysis
of gut contents, concluded that, although sedimented
particulate matter seems to provide the major food
source for the benthic community at the BENGAL
site, food-web structure is not simple. Their analysis
of 90 species shows that 35 macro- and megafauna
are deposit feeders, with 17 suspension feeders, and
29 predators/scavengers. However, d
13 N values overlap
between feeding types and cover a large range within
them, indicating a considerable overlap in food sources,
and probably high competition for food. Invertebrate
predators, such as the polychaete Nicon sp. and the
asteroid Dytaster grandis showed the highest ratios
(i.e., greatest depletion). Interestingly, d
13 N values for
the latter species are greater than those for highly
motile predator/scavengers, such as amphipods, decapod crustaceans and benthopelagic fish (see pp. 354–
359).
Modelling food webs, energy flow and carbon
dynamics on the deep-sea floor
Perhaps one of the best-developed models of food web
structure and energy flow has been put together by fisheries biologists for the Grand Banks off Newfoundland.
This model aims at a better understanding of species
interactions for the purposes of stock management of
the top predators, such as cod (Gadus morhua). A study
of a subset of seven consumer species used augmentation or depletion of single species in order to test
effects on the resilience of the community, and found
highly indeterminate outcomes (Gomes and Haedrich,
1992). No data of remotely comparable resolution yet
exist for the deep sea, where incorporation of additional
complexity in order to approximate biological realism
will be even more necessary than for the Grand Banks.
Therefore attempts to model trophic relationships in the
deep sea will face much greater problems.
For this reason the approach up to now has been
to categorize the entities of the deep-sea ecosystem
in terms of size and functional groups, which have
been sufficient as descriptive compartments in models
of early diagenesis (e.g., Soetaert et al., 1996). Such
models are typically constrained by particle-flux data
from sediment traps, sediment porosity profiles and
estimated sedimentation rates and bioturbation using
361
of food-chain links has often reflected the intensity of
study of a system (Hall and Rafaelli, 1991). The claim
that longer chains are less stable than shorter ones,
which would encourage severe population fluctations,
has held sway for for nearly two decades in the
ecological literature as a result of models, dating from
May (1973), demonstrating the destabilizing effect of
complexity in food webs. The recognition that in nature
complexity is of common occurrence has encouraged
new work and modelling, which has succeeded in
demonstrating the stabilizing effect of complexity as
deduced by intuition, in terms of an interwoven matrix
of interaction holding the community together. This
does not employ an approach with nonlinear terms in
predator–prey relationships. Predators are constrained
so they cannot maintain high feeding efficiency on
many different prey organisms at the same time, thus
relaxing pressure on low-density resources and so
preventing their extinction. These many weak links,
with few strong interactions, in the system connecting
species in the community dampen oscillations between
consumers and resources, and thus tend to stabilize,
rather than destabilize, the dynamics of the component
species (McCann et al., 1998; Polis, 1998).
Although it has been thought that food-web relationships among individual species in the deep sea are
still too poorly understood to allow construction of
the benthic food web for any single deep-sea station
(Sibuet, 1992), there is steady movement towards the
realization of this goal. Because of high local species
richness, a high degree of taxonomic resolution in
food-web description is difficult, but it seems unlikely
that the number of food-web links approaches those
found in the tropical rain forest, even if similar
sorts of interaction occur. However, one interesting
consequence of the patchy system caused by spatially
uneven detrital deposition (which may also help explain
high species richness – see Rice and Lambshead, 1994)
is that food chains will be shorter in small habitat
patches (Pimm et al., 1991).
Stable-isotope analysis has been applied in order to
track sources in marine food webs involving deposit
feeders (see overview of subject in Michener and
Schell, 1994). Use of such natural tracers involves a
number of assumptions. For example, measurements of
d
13 C may be biased by lipid contents, with differing
rates of
13 C enrichment occurring depending on relative
amount of carbohydrate, proteins and lipids present
in the tissue. It is claimed that
15 N ratios are more
conservative among biochemical fractions, so that
d
13 N is a more reliable tool in tracking trophic
linkages (Hobson and Welch, 1992). However, stableisotope studies conducted in tandem with analysis
of stomach contents seem to provide most certain
data. Iken et al. (2001), using d
13 N and analysis
of gut contents, concluded that, although sedimented
particulate matter seems to provide the major food
source for the benthic community at the BENGAL
site, food-web structure is not simple. Their analysis
of 90 species shows that 35 macro- and megafauna
are deposit feeders, with 17 suspension feeders, and
29 predators/scavengers. However, d
13 N values overlap
between feeding types and cover a large range within
them, indicating a considerable overlap in food sources,
and probably high competition for food. Invertebrate
predators, such as the polychaete Nicon sp. and the
asteroid Dytaster grandis showed the highest ratios
(i.e., greatest depletion). Interestingly, d
13 N values for
the latter species are greater than those for highly
motile predator/scavengers, such as amphipods, decapod crustaceans and benthopelagic fish (see pp. 354–
359).
Modelling food webs, energy flow and carbon
dynamics on the deep-sea floor
Perhaps one of the best-developed models of food web
structure and energy flow has been put together by fisheries biologists for the Grand Banks off Newfoundland.
This model aims at a better understanding of species
interactions for the purposes of stock management of
the top predators, such as cod (Gadus morhua). A study
of a subset of seven consumer species used augmentation or depletion of single species in order to test
effects on the resilience of the community, and found
highly indeterminate outcomes (Gomes and Haedrich,
1992). No data of remotely comparable resolution yet
exist for the deep sea, where incorporation of additional
complexity in order to approximate biological realism
will be even more necessary than for the Grand Banks.
Therefore attempts to model trophic relationships in the
deep sea will face much greater problems.
For this reason the approach up to now has been
to categorize the entities of the deep-sea ecosystem
in terms of size and functional groups, which have
been sufficient as descriptive compartments in models
of early diagenesis (e.g., Soetaert et al., 1996). Such
models are typically constrained by particle-flux data
from sediment traps, sediment porosity profiles and
estimated sedimentation rates and bioturbation using