Evolutionary genetics of host use in swallowtail butterflies

Abstract

PHYTOPHAGOUS insects include many of the most species-rich genera and families worldwide^1,2, and much of that diversity seems to result from speciation onto different plant species^3–7. Divergence onto different host plants is thought to involve at least two sets of loci, which may be genetically linked: loci controlling the preference of ovipositing females for a particular plant species, and loci controlling the ability of larvae or nymphs to feed on those plants^8–11. Others have argued that oviposition preference and larval performance may be pleiotropic effects of the same loci^12–15. The genetic relationship between oviposition preference and larval performance has therefore become a central problem in the developing theory of insect and plant interactions^16–23 and speciation in insects^24–27. Results from interspecific crosses between two swallowtail butterfly species that feed on different plant families indicated that oviposition preference is controlled in these insects primarily by one or more loci on the X chromosome²⁸. In a series of reciprocal interspecific crosses between these species, Papilio zelicaon and Papilo oregonius, we investigated whether larval performance on different plant species was also controlled by X-linked loci, which could allow for strong correlations between oviposition preference and larval performance. We found no X-chromosome effect for any component of larval performance. Loci from both parents influenced survivorship, and maternal effects influenced pupal mass and possibly development time. The results varied with the specific measure of performance, the host plant species used in comparing reciprocal crosses, and the sex of the larvae, all of which caution against the use of any single measure of performance in evaluating the evolutionary genetics of host shifts.