The major goal of this project was to document the effects of herbivore ecology and developmental stage on enemy-induced mortality. Insect herbivores are killed by a taxonomically and ecologically diverse collection of natural enemies that includes vertebrate and invertebrate predators, insect parasitoids, nematodes, fungi, protists, bacteria and viruses. However, not all herbivores suffer enemy attacks to the same degree. Mortality due to enemies can be severe enough to cause extinctions of local populations, but in some cases it can be trivial. Given their variable impact the ecological and biological factors that influence enemies need to be identified in order to predict when and where they are likely to be important.

In a previous study, we found that enemy attack was the most frequent cause of death for immature herbivores and was more frequent than all other factors combined for late immature stages(late instar larvae and pupae). However, the different types of enemies were not distinguished in this analysis. Instead, they were pooled into a single category, which was compared against a variety of other mortality sources (competition, plant factors, weather, developmental failure and miscellaneous/unknown). This method obscured any differences in the way parasitoids, predators and pathogens might respond to the ecological settings of herbivore populations.

In this project, we distinguished mortality caused by three basic types of enemies: predators, parasitoids, and pathogens. We then tested for differences associated with herbivore developmental stage and four ecological characteristics of herbivores: invasion status (native vs. exotic), feeding biology (whether they are exophytic and feed externally on plant tissues, or endophytic and feed internally on plant tissues), cultivation (natural/semi-natural vs. cultivated) and latitudinal zone(tropical/subtropical vs. temperate).

Total enemy-indiced mortality was highest in the late developmental stages, and overall, parasitoids kill more herbivores than do either predators or pathogens. Among the ecological variables, both feeding biology and latitude showed significant enemy effects in at least one late developmental stage, whereas neither cultivation status nor invasion status was associated with enemy-induced mortality at any stage.

Bonferroni adjustment of probabilities for multiple comparisons resulted in few significant interactions between enemy type and the ecological variables. However, raw probabilities and comparisons across herbivore immature stages suggest several patterns that deserve attention in future studies:

These results identify several general patterns in insect demographics that should be useful for hypothesis testing.