A General Model for Individual Host Selection

Abstract

We formalise some ideas on host choice by individual insects, into a model which allows an animal to change diet during a lifetime. We assume that factors affecting acceptance thresholds through insect condition are not host-specific. Major predictions are that: individual insects may change diets, but should never become monophagous on plants which have previously been rejected; current eggload may affect diet width in ovipositing females; conditioning of host selection behaviour may be quite complex, including cross-conditioning to different hosts; genetic variance for host choice may be maintained in populations; genetic correlations between use of different novel hosts, and between host choice and fecundity, should be significant and positive. Evidence for these and corollary predictions appears to corroborate the model. In addition, we offer explanations for two previously unexplained phenomena: asymmetrical cross-conditioning; and genetic variance for use of hosts which have not been previously experienced. Such genetic variance allows host shifts to evolve rapidly. Finally, we argue that these ideas and results indicate that chemical similarity between hosts is sufficient but not necessary for host shifts to occur.