Host range evolution is not driven by the optimization of larval performance: the case of Lycaeides melissa (Lepidoptera: Lycaenidae) and the colonization of alfalfa

Abstract

Herbivorous insects that have recently incorporated novel hosts into their diet provide unique opportunities for understanding factors that promote or constrain the evolution of niche breadth. Lycaeides melissa has colonized both cultivated and feral alfalfa (Medicago sativa) throughout much of North America within the past 200 years. We investigated the quality of the novel host as a resource for juvenile development, and asked if the novel host is a preferred host for oviposition relative to a native host (Astragalus canadensis). Larval-performance and oviposition-preference were examined using L. melissa individuals from a population associated with both M. sativa and A. canadensis, and oviposition-preference was also examined in another population associated exclusively with M. sativa. In addition, we investigated the effects of M. sativa and A. canadensis flowers on both preference and performance. Only one of the hosts, M. sativa, has flowers that are accessible to nectaring butterflies, and we hypothesized that the presence of flowers could affect female behavior. We find that the novel host is a relatively poor larval resource: adults that were reared as larvae on M. sativa were roughly one-third the size of adults that were reared on the native host, A. canadensis. The native host, Astragalus canadensis, is the preferred host in choice experiments involving only foliage. However, when flowers were included in preference assays, the native and novel hosts received similar numbers of eggs. Thus, the presence of flowers on hosts in the field might influence the utilization of a novel and inferior larval resource. These results are consistent with a model in which host shifts are driven by adult behavior that does not directly optimize larval performance.