Quantifying predation on folivorous insect larvae: the perspective of life-history evolution

Assumptions about mortality rates form a cornerstone for models of life-history evolution. When seeking adaptive explanations for body sizes, the size dependence of predation risk is of particular interest. Here, we review published studies that provide (1) estimates of the daily predation rates experienced by insect larvae feeding on tree leaves or (2) evidence concerning the relationship between predation risk and larval size. Larvae were found to experience an average avian predation rate of 3.1% per day and an average arthropod predation rate of 10.5% per day. In some studies, mortality rates were systematically dependent on parameters of the larvae (e. g. coloration) or of the environment (host plant, season), but not to the extent that would render generalizations meaningless. Nevertheless, mortality rates varied considerably more for arthropod than avian predators, making an estimate of avian predation rate more reliable for use in quantitative models. Moreover, birds tend to be a more important predator guild exploiting the larger larval stages, as indicated by the predominantly positive size dependence of avian predation risk. By contrast, predation by arthropods was generally negatively size dependent. Based on the available data, we estimate that avian predation rates increase approximately 3.6-fold, while arthropod predation rates decrease approximately 4.9-fold, in response to a 2-fold increase in the linear size of prey. A modelling exercise showed that realistic mortality rates - if assumed to be independent of size - cannot serve as a basis for adaptive explanations for observed body sizes. However, by assuming a positive size dependence of mortality risk within the limits observed for bird predation, it is possible to explain a wide range of body sizes within an optimality framework. (C) 2011 The Linnean Society of London, Biological Journal of the Linnean Society, 2011, 104, 1-18.