TESTING PREDATOR-PREY THEORY BY STUDYING FLUCTUATING POPULATIONS OF SMALL MAMMALS

Fluctuating populations of small mammals provide an excellent opportunity to study the functional and numerical responses of predators because of the wide range in prey density that occurs. I reinterpret data from six studies that have examined the role of predation in the population dynamics of votes in California, southern Sweden and western Finland, of snowshoe hares in northern Canada, and of house mice and rabbits in Australia. Most studies have measured functional responses by relying on changes in diet as reflected by seat or stomach contents. These methods are probably biased toward showing predator satiation. Contrary to previous conclusions I find that there is little evidence for non-linear (Type III) functional-response curves or predator satiation at high prey densities. Recent studies indicate that the functional and numerical responses of predators can be rapid and strong enough to initiate cyclic declines, dampen fluctuations, or even cause stable numbers. The exception to this appears to be the irruptions of mice and rabbits in Australia. I propose a general explanation for the role of predation whereby the effect of predation is largely dependent on the entire prey community. When potentially cyclic prey are a small component of the overall prey biomass, generalist predators are able to prevent fluctuations by strong functional or numerical responses. As the prey community becomes dominated by a few species that fluctuate, limit cycles predominate. Limit cycles turn into irruptive population dynamics when seasonal prey reproduction is eliminated because of extended periods of vegetation growth (vegetation flushes following drought). In the future we must test assumptions underlying the way we study predation by telemetric monitoring of prey mortality and by experimentally manipulating predation.