Background/Question/Methods Determinants of herbivore host plant selection have been the topic of numerous ecological and evolutionary studies. Plant chemistry and natural enemies are two of the most important factors influencing host plant choice. To elucidate the mechanisms that shape host plant choice, it is important to understand the interaction between these two major factors. We examined effects of host plant chemistry on the broad generalist herbivore, Grammia incorrupta (Lepidoptera: Arctiidae), which sequesters pyrrolizidine alkaloids (PAs) from particular host plants and is attacked by a suite of parasitoids. We performed two experiments addressing the general hypothesis that host plant selection by G. incorrupta is modified by physiological conditions associated with interactions between parasitism and plant chemistry. Specifically, we asked two questions: 1) what are the effects of parasitism on PA consumption and 2) what are the effects of ingested PAs on the immune response of G. incorrupta? To address the first question, we measured the PA consumption by individuals of field collected G. incorrupta, which were presented with PA-treated filter paper, then dissected to score parasitism. For the second question, individuals were injected with silica beads as a standardized immune challenge, and their PA consumption and immune response were measured on PA and non-PA diets.

Results/Conclusions Parasitized individuals of G. incorrupta consumed significantly less of the PA filter paper compared to the non-parasitized individuals. Similarly, for the second experiment, we found that individuals injected with silica beads as an immune challenge consumed less PA filter paper than sham injected and control caterpillars. Caterpillars feeding on PA and non-PA diets showed no difference in their immune response as measured by bead melanization. In contrast to previous experiments with this system, which demonstrated increased PA-preference by parasitized individuals, we found here that PAs were relatively unacceptable to parasitized caterpillars. The reason for this discrepancy is most likely due to a difference in G. incorrupta’s defensive strategy based upon parasitoid development. At early stages of a parasitic infection, G. incorrupta avoids PAs or consumes less to maintain a healthy immune response. However, if infection progresses to later stages, G. incorrupta consumes more PAs, thereby creating a toxic environment for the parasitoid. The fact that PAs did not negatively affect G. incorrupta’s melanization suggests that there are additional variables besides PAs that determine immune function. We conclude that, when parasitized, host plant selection by G. incorrupta is flexible and may change throughout development of the parasitoid.