Explaining the differential abundance of species is a fundamental goal of community ecology. The theory of metabolic ecology relates organism abundance to mass using a simple power equation. Although this relationship holds well at large spatial and temporal scales, it is not as well-supported in local samples because ecological factors have a larger influence on species density than the inherent limits to energy use set by allometric scaling. In this study, we hypothesized that we could improve the understanding of the mechanisms driving the relative abundance of insect herbivores on a shared host plant, a long-standing goal in plant-herbivore ecology, by incorporating the influence of local intrinsic and extrinsic ecological factors, e.g., individual growth rate, survival, reproduction, and degree of polyphagy. Using an assemblage of forest caterpillars found co-occurring on a single host plant species (box elder [Acer negundo L.]), we tested whether species relative abundance patterns were best explained by metabolic theory, or alternatively, by models incorporating life history traits of the caterpillars calculated in association with the host plant in both field and lab settings.

Results/Conclusions Using a generalized least squares regression method incorporating phylogenetic relatedness of the herbivores, we found no scaling relationship between herbivore abundance and mass. However, we found a strong positive relationship between abundance and intrinsic population growth rate as calculated from traits measured in association with the host plant. Survivorship (including parasitism rates on the host plant in the field) and larval development on the host plant independently drove herbivore population growth estimates, and species mass was not correlated with these influential traits. Species mass did influence mean species fecundity, but this had a small overall effect on population growth estimates and abundance. Diet breadth showed a non-linear relationship with abundance, with the four most abundant species including two extremely polyphagous species and two specialists. Overall we found that the local abundance of a suite of herbivores on a given host plant is best explained by quantifying intrinsic population parameters of the herbivore specific to that host. Although data-intensive, this hybrid approach may prove fruitful for addressing long-standing questions of community assembly, and suggests limitations to metabolic theory at scales where ecological processes dominate.