PARADOXICAL AVOIDANCE OF ENRICHED HABITATS: HAVE WE FAILED TO APPRECIATE OMNIVORES?

Abstract

Omnivores have not figured prominently in our understanding of food webs and prey dynamics even though they can have substantial direct and indirect effects on the structure of ecological communities and on the dynamics of interacting species. The important role of omnivores is implicated by the paradoxical results of a food-supplementation experiment. The experiment was designed to test theories that predict how habitat change affects the distribution of habitat-selecting species. According to theories of habitat selection, a quantitative change in habitat (as caused by supplemental food) should increase consumer population size and alter habitat selectivity. Related theories of patch use predict that consumers should increase their use of enriched patches. A two-year experiment on two species of small mammals in Canada's boreal forest failed to alter population densities of red-backed voles, but did cause a dramatic shift in vole habitat use. Rather than increasing their use of feeding stations as predicted by classical theory, voles avoided them. Deer mice did not respond to the experimental treatments. The paradoxical results occurred because omnivorous black bears altered prey behavior by increasing predation risk at feeding stations. Revised theory confirms the indirect omnivore effect, and demonstrates that the behavioral paradox is far more likely for omnivores than for other types of predators. The behavioral paradox of enrichment highlights not only important new, and potentially stabilizing, roles for omnivores, but also the pervasive influences of behavior and habitat selection on population dynamics and regulation.