Plant-herbivore interactions have traditionally been framed in terms of pairwise interactions, focusing on the direct effects of plant defensive traits on herbivores, herbivores on plant performance, and the resulting co-evolution between the two. However, a bi-trophic perspective is limited in that the ecological and evolutionary outcomes of species interactions are subject to indirect effects stemming from their surrounding communities. In particular, plant-herbivore interactions are shaped by predators, who through the consumption of plant herbivores can provide an indirect defense to plants by indirectly reducing plant damage and promoting plant performance. Likewise, plants can indirectly mediate predator-herbivore interactions with potential consequences for the magnitude of trophic cascades. I use a tri-trophic approach based upon plants, herbivores, and insectivorous bird predators to understand how plant traits influences interactions between herbivores and predators, and in turn how predators indirectly contribute to the plant defense at multiple scales of biological organization. Specifically I evaluate: (1) a trade-off in direct herbivore resistance and indirect defense from birds and underlying plant traits, (2) whether herbivore predation risk is mediated by toxic host plants, and (3) the consequences of biodiversity loss on bird-herbivore interactions.

By assessing the relationships between interspecific variation in direct herbivore resistance and indirect defense from birds, I found that density-dependent foraging by bird predators shapes a trade-off in direct and indirect defense from birds. This relationship was underlain by the effects of plant structural complexity on both bird and arthropod communities (Chapter 1). Inversely, I found that toxic plants deter birds (Chapter 2). As a result, herbivores can gain enemy-free space by occurring in patches where toxic plants are more dominant. In contrast, bird diversity was positively affected by tree diversity which corresponded to stronger predation rates on herbivore models, suggesting a feedback between diversity and plant and predator trophic levels (Chapter 3). Collectively, this work highlights the importance of indirect effects across multiple trophic levels in shaping community structure and dynamics.