Using a Salix hybrid system, part of my research with Bob Fritz has investigated the consequences of genetic variation on the community structure of insect herbivores and the influence of plant genetics on resistance to insect herbivores (that is, the underlying genetic basis for community structure). Most of the insect herbivore species that feed on plants in this hybrid system have low motility (i.e., they are gallers, miners, or leaf-folders), making the plant on which an individual insect lives readily apparent. Thus, this system makes an excellent one to examine the community structure of insect herbivores. We created six families from each of six genetic classes (Salix eriocephala, Salix sericea, F₁ hybrids, F₂ hybrids, and backcross hybrids to each parental species) using controlled crosses in order to examine whether plant genetic variation was the basis for the community structure of these insect herbivores. Using canonical discriminant analysis, we found that plant genetics was the main structuring force underlying the willow herbivore community, with hybridization creating distinctly different communities among all six genetic classes. Moreover, line-cross analysis demonstrated that epistatic interactions were highly influential in community formation. This analysis revealed three major patterns of herbivore response, epistasis of resistance, epistasis of host recognition, and dominance of resistance. These patterns indicate that several, distinctly different herbivore species influenced community structure similarly, suggesting that community structure was more than just idiosyncratic responses of individual insect herbivores.

Using the plants mentioned above, we have also evaluated the influence of plant genetics on susceptibility to insect herbivores to determine the underlying genetic basis for community structure. To do this, we have evaluated the patterns of inheritance of susceptibility using line-cross analysis. Evidence from these analyses suggest three main points: (1) digenic epistatic interactions for plant susceptibility to herbivores were ubiquitous in this hybrid system; (2) abundance patterns were affected by resistance traits for some herbivore species, but by host recognition traits for other species; (3) herbivores that use both plant species as hosts appear to utilize differing traits for each willow species when making oviposition and feeding decisions.

Working with Bob Fritz, I have also examined the implications of plant responses to vertebrate herbivores on community structure of arthropod herbivores. Using the same willow hybrid system, the effects of the biotic environment (that is, vertebrate herbivores), as well as plant genetic variation, were incorporated into an examination of community structure. Using three genetic classes, consisting of Salix eriocephala, S. sericea, and their interspecific F₁ hybrid, we examined the influence of genetic variation and the importance of browsing damage on the community structure of the arthropod herbivores. Both the direct effects of plant genetic variation and indirect effects of browsing herbivores were important to the community structure of arthropod herbivores in this system. Because indirect interactions in this system have influential ecological consequences, they may also have important evolutionary implications. In altering the community structure of arthropod herbivores, mammalian browsers may also alter the competitive structure of those herbivores, or the interactions of herbivores with predators/parasitoids, potentially leading to further evolutionary repercussions.