Optimality modelling and quantitative genetics as alternatives to study the evolution of foraging behaviours in insect herbivores.

Summary Although the evolution of large-scale dispersal has received considerable attention, we know very little about how natural selection influences foraging behaviours in herbivorous insects. Host-selection behaviours and within-habitat movements jointly determine foraging behaviours, since host selection affects the allocation of time spent on a particular host versus moving between these hosts. However, host selection is generally a labile trait, whose expression is influenced by the physiological state of the forager and hence, by characteristics of the habitat. We discuss how the quantitative genetic concepts can be used to study the evolution of such labile behaviours. Since host responses depend on the physiological state of the forager, it is argued that the state of the forager must be explicitly considered when estimating the additive genetic basis of host-selection behaviours. The lability of foraging behaviours increases the difficulty of measuring the fitness consequence of variation in the foraging phenotype in specific habitats. Therefore, it may be difficult to rely exclusively on quantitative genetic methods to test hypotheses about adaptive change in foraging behaviours across different habitats. We provide a novel approach based on optimality modelling to calculate the fitness consequence of variation in the foraging phenotype across different habitats. This method, in conjunction with quantitative genetics, can be used to test hypotheses concerning the evolution of foraging behaviours.