Predation and leaf quality shape herbivore community structure on Quercus alba

Background/Question/Methods The abundance and impacts of insect herbivores are determined by both the top-down impacts of predators and the bottom-up effects of plant quality and defenses. However few researchers have examined how these forces independently and interactively determine the composition of diverse herbivore communities. Using the white oak (Quercus alba) study system, we combined a bird exclosure experiment on understory branches with quantification of leaf quality characteristics to determine their effects on herbivore community abundance and composition. We examined condensed and hydrolysable tannins, total phenolics, leaf toughness, and water content and conducted surveys of leaf-chewing herbivores during the three known distinct phonological communities (early May, early July, and late August) in this system. Herbivores were identified to the highest taxonomic resolution possible (usually species), and we recorded abundance and identities of arthropod predators. Results/Conclusions Spring total herbivore abundance was negatively correlated with leaf toughness and total phenolics, while mid- and late summer herbivores declined when birds were present. Late summer herbivores were negatively correlated with total phenolics and positively with water content. The patterns in spring were driven by free-feeding herbivores, while abundance of concealed feeders (structure-builders and leaf-miner) was not predicted by any of the variables examined. In July free-feeders responded positively to bird exclusion, leaf toughness, and water content, but were negatively correlated with arthropod predator abundance. In contrast, July concealed feeders were correlated only with total phenolics, and surprisingly their abundance increased with phenolic content. August free-feeders decreased with condensed tannins but increased with water content and arthropod predators; concealed feeders followed the same patterns while also increasing with bird exclusion and condensed tannins. Spring specialists marginally increased with bird exclusion, while generalists were stongly negatively correlated with leaf toughness and total phenolics. July specialists were positively affected by bird exclusion, toughness, condensed tannins, and arthropod predators; generalists were positively correlated with toughness and water content. August specialists were not influenced by any leaf characteristics, while generalists decreased with tannins and bird predation. Bird predation did not affect herbivore species richness in any census; spring richness declined with leaf toughness and phenolics, while mid-summer richness increased with water content and leaf toughness. While both leaf characteristics and bird predation frequently influenced herbivore community structure, their effects were additive so that bird predation never varied with leaf quality.