Host plant species identity is known to influence herbivores’ likelihood of mortality from parasitoids. Possible mechanisms responsible for host plant effects on herbivore-parasitoid interactions include differences between plant species in the emission of volatile organic compounds that attract parasitoids to herbivore infested plants, differences between plant species in the apparency of herbivores, differences between plant species in the ability of herbivores to resist parasitoid attack, and differences in herbivore densities across plant species.  The purpose of this study is to separate the effects of trait-mediated and density-mediated interactions by testing whether host plant species identities explain variation in the outcome of caterpillar-parasitoid interactions while accounting for variation in caterpillar densities. We used an observational experiment to do this, collecting caterpillars from eight tree species in forests of central Connecticut for three weeks each June from 2004 to 2007.  Caterpillar collection was performed with a quantitative sampling protocol by sampling whole tree branches, counting the number of leaves on the branches, and measuring leaf areas for each host plant species, allowing calculations of caterpillar densities. Caterpillars were reared in lab on foliage from the same host plant species on which they were collected, and instances of parasitism by wasps and flies were recorded.  

Results/Conclusions

Host plant identity affected the frequency of mortality from parasitoids for six, abundant, generalist, caterpillar species grouped together in analyses conducted at two temporal scales. At both scales, we observed differences between parasitism of caterpillars on Quercus rubra, which had high parasitism frequencies, and those on Quercus alba, which had low parasitism frequencies.  Analyses conducted for tachinid flies and wasps separately demonstrated that the host plant effect, and the Quercus effect in particular, was driven by tachinid fly parasitism. Another analysis, comparing parasitism frequencies between caterpillars on these two Quercus species only, and grouping only those caterpillars common to both host plant species, demonstrated a host plant effect for wasps only. Like the previous analyses, Quercus rubra had high parasitism frequencies while Quercus alba had low parasitism frequencies. These results indicate that host plant identities can affect herbivores’ likelihood of mortality from parasitoids, and that this effect exists outside effects due to variation in herbivore density across host plants.