Biotic interactions can structure communities, drive succession, and account for patterns of biodiversity, yet we currently know little about how multiple biotic factors interact to influence traits in ecological and evolutionary time. Moreover, variation in biotic interactions among native and introduced plant ranges may account for the spectacular success of some invasive species. For example, in novel environments plants may experience ecological release from coevolved specialist enemies and may in turn evolve allocation away from costly defense towards competitive ability. This assumes that plants face a fundamental tradeoff between growth and defense, but we predict that this tradeoff depends on the biotic environment and specificity of plant responses to herbivores. I examined variation in growth and herbivore defense allocation in a model woody invader, using native and invasive populations of Triadica sebifera that differed in historic herbivore pressure. Greenhouse studies demonstrated that populations from the native range with a history of intense herbivory invested more in indirect herbivore resistance (extrafloral nectar production) than invasive populations. Additionally, the induction of indirect defense depended on herbivore feeding mode, suggesting tradeoffs among defenses against multiple herbivores. Further, I showed that even similarly feeding generalist herbivores can induce specific defense responses of plants, by demonstrating a loss of specificity in invasive populations compared to native populations, despite higher tolerance of herbivory overall. Together, these patterns suggest that selection for competitive ability may result in correlated selection for increased tolerance but decreased resistance of herbivory and reduced specificity of defense responses. I tested mechanisms causing this pattern of reallocation by examining the relative contributions of selection from herbivores and competitors in shaping herbivore defense in a different model plant species, common mustard. I used an experimental evolution approach to generate lines that differed in historical herbivore identity and competitive intensity. I detected that both herbivore and competitive history influence plant resistance to contemporary herbivores, and revealed a strong role for competitive history in contemporary plant defense. Understanding how variation in biotic interactions contributes to plant success is one way to predict the traits likely to evolve within a given selective environment and their ecological effects.