Integrating chemical plant trait- and ecological-based approaches to better understand differences in insect herbivory between cultivated and natural systems

Abstract

Cultivated plants are often much more susceptible to insect herbivores than wild-type plants. In addressing this observation, much attention has focused on a trait-based approach, and especially on how artificial selection via domestication has modified morphological and chemical traits, in particular levels of defensive secondary metabolites. However, larger scale ecological processes, such as the spatial distribution and diversity of species in a plant community, also determine how insects locate and exploit their food plants, and these differ profoundly between natural and agricultural ecosystems. In this paper we discuss these two approaches to better understand differences in levels of insect herbivory between agricultural and natural ecosystems. We argue that studies investigating the effects of secondary metabolites on insect herbivory are compromised by the methodological approach that is often used. Insect feeding assays testing the effect of reduced concentrations of secondary metabolites in domesticated plants rely on testing a limited subset of insect species, usually those that can easily be reared in the laboratory and often are agricultural pest species. The responses of these insects do not reflect the full range of responses of the species present in the plant’s natural habitat. This may explain why reduced levels of secondary metabolites in crop plants may only partially explain increased susceptibility to herbivory. Hypotheses explaining larger scale patterns of insect herbivore abundance are often based on studies in agricultural settings. In our opinion, developing broad ecological hypotheses based on studies in agricultural systems do not necessarily apply to natural systems and vice versa. To fully understand how susceptibility or resistance to insect herbivory is affected by plant traits and habitat heterogeneity, these have to be studied together in both natural and agricultural settings.