Species Competition and Predation - Ecological Systems - page 236

without the presence of other stabilizing mechanisms, because alone it would not add
the sort of structure to feedback loops able to separate many species [ 16 , 35 ]. Indeed,
the patterning of feedback loops would be quite limited. Lotka-Volterra theory in this
case predicts that at most two species can coexist stably when species in the focal guild
do not partition resources, and there is only one predator present, regardless of how the
predator distributes its attacks [ 41 ].
Fundamentally, development of the keystone species concept has viewed preda-
tion and competition as very different kinds of interactions, and has not focused
adequately on the true requirement that density dependence needs to be concentrated
more strongly within species than between species to promote stable coexistence
[ 16 ]. Simply reducing the magnitude of competition uniformly does not achieve this,
and equalizing fitnesses alone does not achieve this in the absence of a separate and
strong stabilizing mechanism [ 16 ]. Certainly it is conceivable that a keystone preda-
tor might focus its attacks in such a way that fitness is approximately equalized across
species, which would achieve approximate neutrality, but there is no mechanism that
makes this likely. Trade-offs between susceptibility to predation and competitive
ability would work in the right direction [ 40 ], but there is no reason for these trade-
offs to be sufficiently precise to yield approximate neutrality, which nevertheless
would not stabilize diversity, just slow its loss. Though keystone predation can indeed
be shown to have strong effects in some communities [ 16 ], the stabilizingmechanisms
needed to make it effective have neither been identified nor sought.
The theory discussed above identifies predator partitioning as an important way in
which predation can promote diversity. Indeed, predators and other natural enemies
can have very strong effects on their prey populations, and it is not uncommon for
predators to be relatively specialized [ 11 ]. In fact, if the predators of a guild partition
their resources (the prey populations in the guild in question), it follows reciprocally
that the prey are partitioning predators. Thus, predator partitioning should have about
the same prevalence in nature as resource partitioning. If both predator partitioning
and resource partitioning are present for any given guild ( Fig. 13.2a ), the reasonable
expectation is that the strength of coexistence would be stronger than if only one of
these were partitioned ( Fig. 13.2c and d ). Indeed, the evidence from models points in
this direction [ 17 ]. However, this outcome applies when predators that partition their
prey are substituted for predators that do not partition their prey. A situation of great
importance in nature is removal, addition, or restoration of predators, as occurs with
human activities, although human activities might sometimes substitute one kind of
predator for another [ 42 - 44 ]. In this case, adding predators that partition the focal
guild to a comparable extent to the partitioning of resources in the focal guild would
not greatly change the strength of coexistence. The reason is that there would be
little change in the ratio of interspecific to intraspecific density dependence. However,
adding predators that do not partition would increase the ratio of interspecific to
intraspecific density dependence, undermining resource partitioning and therefore
undermining coexistence, as noted above.
These effects of adding or removing predators, or changing their properties, can
be summarized in terms of changing values of r as the scenario changes. Note that r
is the niche overlap measure, but is also a symmetric measure of the ratio of