Spatial heterogeneity, not visitation bias, dominates variation in herbivory

Experiments in ecology can have unintended side effects. Recently, it has been suggested that the act of visiting a plant, inherent to studying herbivory, may alter plant performance and interactions. To evaluate the generality of this inference, we examined plant performance and herbivory on 14 plant species in three geographic regions. Visitation did not significantly affect any of the variables that we measured, including leaf damage, height, biomass, or survivorship, for any species. However, rates of herbivory varied significantly among sites and regions. Thus, our data do not support the generality of visitation impacting estimates of herbivory. We propose that future studies of herbivory will gain more by evaluating spatial heterogeneity in interaction outcomes than by quantifying possible experimenter-caused variation.