Bottom-Up Effects of Habitat, Host Plant and Herbivore Quality on Parasitoid Host Use

Bottom-up (resource mediated) effects have been shown to strongly influence higher trophic levels, with the potential to drive trophic cascades. Parasitoids, or insects that must complete larval development within or on the body of another arthropod host, are important members of the third trophic level. Parasitoids inflict substantial mortality on herbivores in both natural and agricultural systems and have been extensively studied in theoretical ecology. However, few studies have taken an empirical approach to understanding the bottom-up factors that mediate host use by both parasitoid wasps and flies in the field. The objective of this dissertation was to determine how forest light environment, intra- and interspecific host plant quality, host caterpillar size and host caterpillar nutritional quality affect parasitoid host use. My study system included several species of `slug' caterpillars (Limacodidae), their shared host plant species, and the community of wasp and fly parasitoids that attacks these caterpillars.;I found that parasitoids primarily partition their shared hosts based on size, but not by host species or phenology. In general, wasps were reared from caterpillars collected in early instars, while flies were reared from caterpillars collected in late instars. Furthermore, for at least one species of solitary wasp, Triraphis discoideus (Braconidae), host size influenced sex allocation of offspring by ovipositing females. Host plant species of differing quality indirectly affected the size attained by the tachinid fly parasitoid, Austrophorocera sp., through its direct effects on the size and performance of the caterpillar host. The host plant species that resulted in the highest caterpillar host performance in the absence of enemies also yielded the largest parasitoid flies, which suggests that host plant quality can cascade up to influence the third trophic level.;In a field experiment that manipulated leaf phenotype (sun or shade white oak leaves) and forest light environment (sunny light gaps or shaded forest understory microhabitats) in a two-way factorial design, I found that light environment strongly affected the likelihood of parasitism, while the effects of leaf phenotype were more subtle. Euclea delphinii larvae were 4.6 - 11.5 times more likely to be parasitized in sunny light gap habitats than in shaded understory habitats. This pattern was consistent for both tachinid fly and wasp parasitoids across two separate experiments. The performance of two species of caterpillars (E. delphinii and Acharia stimulea) was maximized in the shade habitat + sun leaf treatment, a habitat/leaf type combination that occurs infrequently in nature. These results suggest that 1) the direct effects of light environment on the incidence of parasitism supersede any indirect effects resulting from altered leaf quality and 2) an inherent ecological tradeoff exists for herbivores choosing between sunny (high leaf quality, physiologically harsh environment, high parasitism risk) and shaded (reduced leaf quality, less harsh environment, reduced parasitism risk) habitats.;I conclude that bottom-up factors, especially host caterpillar size and forest light environment, strongly affect parasitoid host use, and that wasp and fly parasitoid guilds may be under selection to partition hosts based on these factors.