Host-Associated Differentiation in Goldenrod Insects

Accounting for the diversity of life on Earth is among the most fundamental problems facing ecologists and evolutionary biologists. A special puzzle is posed by the astounding diversity of plant-feeding (phytophagous) insects and their insect parasitoids, which together account for more than half of the world’s described animal taxa. Sister-group comparisons imply that phytophagy can drive rapid diversification, and the profusion of large genera and sibling species among phytophagous insects suggests that their rapid diversification continues today.

An important candidate mechanism for the diversity of phytophagous insects is frequent, rapid speciation via host race-formation. While a few cases of host-associated differentiation are well documented, it remains unknown whether, among phytophagous insects, this process is rare and special or frequent and predictable. A deeper understanding of the evolutionary importance of differentiation following host shifts will require perspective from two complementary approaches. The first is the study of multiple host shifts within a single clade (e.g., Rhagoletis). The second is the study of parallel host shifts by multiple evolutionarily independent insect lineages on the same host plant pair.

Collaborators Dr. Steve Heard (U. of New Brunswick), Dr. John Stireman (Wright State U.), and I are testing the hypothesis that reproductive isolation and genetic divergence can evolve as a consequence of divergent selection on traits between differing host environments. Little is known of the frequency among insect lineages of differentiation and eventual speciation via host race formation. We do not know the likelihood of host race formation given the opportunity (i.e., given one insect species attacking two hosts, either in allopatry or sympatry) or how that likelihood might depend on the ecology of the insect or the hosts.

In contrast to the traditional approach of focusing on a particular clade (such as the Rhagoletis pomonella group), we are assessing the frequency of host race formation by examining genetic divergence in the diverse community of herbivores that feed on two closely related and broadly sympatric species of goldenrods (Solidago altissima and S. gigantea). Steve and I pioneered this latter approach, showing that the goldenrod elliptical-ball moth Gnorimoschema gallaesolidaginis (Gelichidae) and the tephritid fly Eurosta solidaginis (investigated by Warren Abrahamson and colleagues) together represent the first known case of parallel host-associated differentiation (30).

Integrating ecological, population genetic, and phylogenetic approaches, John has been the driving force in more recently expanding our work to eight species of goldenrod herbivores. In a recent paper in Evolution (44), John presents mtDNA sequence, allozyme, and ecological data that, in conjunction with data on two herbivores from Warren Abrahamson’s group, provides unequivocal evidence that host shifts are common (at least 4 of 10 taxa), phylogenetically diverse (Coleoptera, Diptera, and Lepidoptera), and in each case morphologically and taxonomically cryptic. Further, in a subsequent paper in the Proceedings of the Royal Society (45) John has shown that host-associated differentiation cascades across trophic levels, extending from the herbivores to their wasp parasitoids (3 of 4 taxa). These results make perhaps strongest case to date for the importance of ecological selection and host race formation as general mechanisms of insect diversification.

Along with previous behavioral and ecological studies of some of the taxa we studied, our work provides novel evidence that host-related divergence and subsequent speciation has likely been a significant mode of speciation in phytophagous insects and their insect parasitoids. In turn, this argues for an important role for ecological (and possibly sympatric) mechanisms of speciation in the diversification of two of the earth’s most speciose groups of terrestrial organisms. A full evaluation of the importance of host-race formation in the diversification of these and other diverse parasitic groups will require continued detailed behavioral, ecological, and genetic study of specific interactions along with broader comparative analyses such as ours in additional systems. Our tests of the effects of enemy-free space (48) and of the local and geographical distribution of ploidy variation (Halverson et al. in revision) and its effects on patterns of insect attack (Halverson et al. in review) are significant steps in this direction. So too is the transfer of knowledge and techniques gained from our study of goldenrod parasites to the study of host-shifting in the the non-pollinating wasps of figs (Bernhard et al. in revision), a phylogentically independent and ecologically very different system.