Quantifying the effects of intraspecific variation on predator feeding rates through nonlinear averaging

Abstract

Handling Editor: Shawn Leroux Abstract 1. Theory suggests that intraspecific trait variation will alter species interaction strengths through nonlinear averaging when interaction strengths are nonlinear functions of individuals traits. This effect is expected to be widespread, yet what factors mediate its magnitude in nature and hence its potential effects on ecosystems and communities are unclear. 2. We sought to quantify how nonlinear predator functional responses, variation in prey densities and counteracting variation in attack rates and handling times among predator individuals of similar body size alter their populationlevel feeding rates through nonlinear averaging in a natural system, and to determine the processes influencing the net magnitude of this effect. 3. We used a field caging experiment in the rocky intertidal of Oregon, USA to quantify attackrate variation and feeding rates of the whelk Nucella ostrina on its barnacle and mussel prey. We also used empirically parameterized simulations to examine the effects of handlingtime variation among individuals on populationlevel feeding rates. 4. Within cages, individual attackrate variation reduced populationlevel whelk feeding rates. However, the magnitude of this reduction differed among prey species and cages depending on cagespecific magnitudes of attackrate variation and functionalresponse nonlinearity. The inferred effects of handlingtime variation among individuals were of smaller magnitude than those of attackrate variation, yet counteracted them to cause a net weakening of the effect of individual attackrate variation on populationlevel feeding rates. Across cages, attackrate and preydensity variation had nonadditive effects that produced greater feedingrate reductions at the experiment scale relative to the cage scale. 5. Our results indicate that the effects of trait variation via nonlinear averaging depend critically on the features of systems that determine the magnitudes of nonlinearities and trait variation. Because of counteracting trait variation, nonlinearaveraging effects may be quite complex, involving both the variances and covariances of all traits and environmental variables influencing the ecological process of interest.