On the role of host-trait variation in insect diversification

Abstract

Ecology’s driving role in niche expansion and subsequent diversification has gained appreciation in recent years, but the specific mechanisms driving such processes remain poorly understood. For the herbivorous insects, diversification may be facilitated by their ability to radiate via host shifts (=host-associated differentiation, HAD). The Gape-and-Pinch (GAP) model (Heard 2012) hypothesizes that HAD may be facilitated by insects’ selective patterns of host use, both within and between host species, which may follow a predictable sequence during divergence. I provide the first tests of the GAP model and investigate a recent possible case of HAD. First, I test hypotheses derived from the GAP model, specifically that host trait variation plays a role in the divergence of specialist herbivorous insects. I quantify trait variation among phenotypes of Solidago altissima and S. gigantea and compare patterns of attacked versus available phenotypes. For specialists, results are generally consistent with GAP predictions: use of available host phenotypes depends on the stage of HAD, and there may be a link between traits diagnosing alternative host species and divergent selection acting on host-associated populations. Second, I ask whether the goldenrod ball-gall fly Eurosta solidaginis, which has evolved several host-associated races, has initiated HAD on S. rugosa. Mitochondrial DNA sequence data indicate that flies on S. altissima and S. rugosa are likely one oligophagous host race, but fine-scale interpretation of fly genotypes across the landscape suggest that S. rugosa-flies may be in the early stages of divergence. Finally, I ask whether a ‘key’ host trait may have driven divergence in one case of HAD and initiated another. For E. solidaginis, selective use of pubescent S. altissima may have promoted divergence from S. gigantea-flies and facilitated acceptance of S. rugosa, but detectability of the pattern varies across the landscape. Identifying such key traits could provide new insights into the mechanisms of HAD and ecological speciation. My results suggest that divergent use of plant phenotypes plays a role in insect diversification, and that early divergence may occur at local scales. More broadly, my work indicates adaptation to within-niche variation may facilitate colonization of novel niches and ultimately drive ecological speciation.