Assessing the latitudinal gradient in herbivory

Abstract

Plant-herbivore interactions occur in all ecosystems and provide a major avenue for energy flow to higher trophic levels. A long-standing hypothesis to explain the latitudinal gradient in species diversity proposes that the relatively stable and frost-free climate of the tropics should lead to more intense biotic interactions in tropical compared with temperate environments, giving rise to a greater diversity of plants and herbivores. Herbivory rates have been compared across latitudes to test this biotic interactions hypothesis, with herbivory typically being measured from observable leaf damage. However, we argue that a measure of percentage leaf damage alone does not straightforwardly reflect the cost of herbivory to the plant, and on its own does not constitute an appropriate test of the biotic interactions hypothesis. For a given amount of herbivory, the impact of herbivory is dependent upon many factors, such as the construction cost of the leaf, the growth and replacement rates and leaf life span. We investigate the latitudinal gradient in herbivory by analysing a large dataset of herbivory rates for 452 tree species and separating the species into those with short and long leaf life spans. We show that annual herbivory rates tend to be greater at lower latitudes for evergreen species (which have long-lived leaves), but no trend in herbivory rate with latitude was found for species with short leaf life spans. Phylogenetic least squares regression assuming Ornstein-Uhlenbeck processes also showed a negative effect of latitude on herbivory rate for evergreen trees, but we caution that viewing herbivory as a species trait is problematic. An integrative approach that incorporates leaf life span, as well as the costs of investment in growth and potential costs of losing leaf tissue, is needed to further our understanding of the ecological and evolutionary dynamics of herbivory.