Abstract: The resource concentration hypothesis (RCH) predicts that herbivores will have higher population densities in areas where their host plant species are more concentrated. However, the RCH has rarely been examined outside of the context of dietary specialist insect herbivores in agroecosystems, and still results vary widely in their findings. We tested the RCH with a novel framework: we measured the poulation numerical responses of dietary specialist and generalist larval Lepidoptera (caterpillars) to their host plants in New England secondary forests, and included parameters for generalist host electivity and host density measurement scale. We found that generalist caterpillar population density correlated positively with host plant concentrations at scales < 50m, while specialists’ populations responded positively at scales < 3 m and negatively at scales between 25 and 50m. Using our models to predict omitted data, we demonstrate that taking dietary generalists’ host-specific electivities into account is necessary for detecting their response to resource concentration. These findings show the first application of the RCH to dietary generalist abundance, and has important ramifications for spatial and landscape ecology.
