We hypothesized that native and invasive insects might differ in their impacts on native host plants. The impact of herbivory on plants may be manifested in different ways, depending on which part of the plant’s life history is affected, as well as such factors as density-dependence. Our study system consists of the invasive South American cactus moth, Cactoblastis cactorum (Berg) (Lepidoptera: Pyralidae) and three native specialist insects, all attacking populations of native Opuntia in Florida. We surveyed prickly pear cacti (O. humifusa and O. stricta) for three years across six sites in the Florida Panhandle. During each visit, we recorded plant size and fruit number, as well as presence/absence data for insect herbivore populations. We used linear mixed models, incorporating site and plant identity as random effects and presence/absence of each insect, initial plant size, and plant species as fixed effects, to predict the relative growth and relative fruit production rates of Opuntia. We used model comparison with Akaike’s Information Criterion (AIC) to determine the best set of explanatory variables.
Results/Conclusions
The best models predicting a plant’s relative growth rate included plant species identity and the number of times that C. vittiger and C. cactorum were observed. In contrast, we found that the best model predicting a plant’s relative fruit production rate includes the number of times each of the four insects (Dactylopius sp., Chelinidea vittiger, Melitara prodenialis, and C. cactorum) was observed, as well as initial plant size and plant species identity. This suggests that insect herbivory differentially impacts vegetative and reproductive parts of the plants. All four insect species feed almost exclusively on the vegetative tissue of the plant. However, only C. cactorum and C. vittiger were found to impact the relative growth rate of the plant. Our results show that all herbivores impact cactus reproductive output but the invasive species in particular affects vegetative growth. These results can be used to design manipulative experiments, and at face value have implications in projecting the potential long-term consequences of C. cactorum herbivory for these native cacti.