Species Competition and Predation - Ecological Systems - page 235

a cause of high mortality, and therefore not distinguished from harsh or stressful
conditions in its effects [ 16 ]. However, as remarked above, such conditions need
have no effect on species coexistence because they need not be associated with
density-dependent effects, although they would alter species average fitness ratios if
they were species-specific in effect.
The feedback loops expected with trophic interactions mean that predation
would likely act in a density-dependent way. However, it is also possible that its
density dependence might be weak relative to other sources and function on
different spatial and temporal scales. For instance, predators that range over
a much larger area than the focal guild might not be very responsive to the changes
in the density of the focal guild. Predators might also live much longer and so
change in density much more slowly than their prey in the focal guild, and they
might also depend more heavily on species outside the focal guild and so not be so
responsive to the focal guild even though they inflict mortality on them. Thus, it is
not unreasonable to entertain the hypothesis that in some situations predation acts in
an approximately density-independent way.
Density-independent predation would reduce the fitnesses k of the species in the
focal guild. From the two-species coexistence condition, it is clear that it would
only affect coexistence if the ratio k 1 / k 2 were affected, or in the multispecies case,
if the ratio of any species' fitness to the guild average fitness were affected [ 15 ]. If
fitnesses are unequal in the absence of predation, the same reduction for each fitness
has the effect of making the fitness ratio deviate further from the value 1, which is
more likely to lead to exclusion than coexistence. However, a trade-off that led to a
disproportionately large reduction in fitness for the species with larger fitness
(“selective predation on the competitive dominant”) [ 16 , 36 ] would make the fitness
ratios more equal provided predation was not too strong. The requirement that the
predation be not too strong leads to an intermediate predation prediction: predation
promotes diversity for intermediate intensities of predation [ 16 , 36 ].
Empirical studies have noted strong effects of predators in some guilds [ 37 ]. Of
particular note are those consisting of sessile intertidal organisms, where the predator
or herbivore is necessary for the maintenance of diversity [ 36 , 38 ]. These predators
are often termed keystone predators due to their essential role [ 38 , 39 ]. Discussions of
keystone predators often emphasize selective predation without regard for a density-
dependent role of predation [ 37 ]. However, density-independent selective predation
is purely equalizing and would be incapable of stabilizing coexistence in the absence
of a separate stabilizing mechanism, such as resource partitioning [ 16 ].
It is reasonable to expect that a keystone predator would respond at least to some
extent in a density-dependent way to its prey. A density-dependent but nonselective
predator, i.e., one that does not affect the fitness ratio, would in fact have the effect of
undermining competition-based coexistence, and thereby undermine coexistence
overall. This would occur because such a predator adds equally to intraspecific and
interspecific density dependence, diluting the effect of resource partitioning and
increasing the value of r , weakening coexistence. A density-dependent but selective
predator need not have negative effects on coexistence provided it is selective on the
competitive dominant [ 40 ], but it would still be limited in its effects on coexistence