Topological structure of behaviorally-mediated indirect effects: consequences for large food webs
Adaptive behaviors are widely recognized as ubiquitous in ecological communities. However, our understanding of how specific behaviors affect the stability of large ecological networks remains poor. Recent work analyzing randomly-generated community matrices has shown that community stability may be determined in part by the relative frequencies of different classes of pairwise direct effects (namely mutualism, competition, and predation). We adopt this approach to explore the local stability consequences of three different adaptive behaviors: diet choice, predator-specific defenses, and adaptive adjustment of general foraging effort by species on the second trophic level. We determine the locations and signs of additional community matrix entries consistent with each behavior. These are added separately to randomly-generated community matrices consistent with food webs without adaptive behavior, and the eigenvalues of each replicate matrix are compared to determine how each behavior affects community stability.
Diet choice and predator-specific defenses each imply the addition of pairwise positive:positive interactions (between species which share at least one predator in the former case, or species which share at least one prey in the latter). Both of these inclusions had negative effects on community stability (i.e. they increased the real parts of the leading eigenvalues of the community matrix). This is consistent with prior findings that mutualistic interactions tended to be destabilizing. Specific defenses were less destabilizing than diet choice, and these results held if the behaviors were only displayed by species on the second trophic level. Adaptive adjustment of general foraging effort implies negative:negative interactions between species which share an adaptive forager as either a prey item or a predator, and negative:positive interactions between a species which preys upon an adaptive forager and one upon which that adaptive forager preys. Inclusion of these terms increased community stability. Our findings emphasize the importance of adaptive behavior to the stability of food webs and extend fruitful community matrix analyses previously applied to networks with direct interactions only. This approach complements work which simulates systems' dynamics by highlighting the potential significance of topological patterns in behaviorally-mediated indirect interactions.