Abstract
Enemy release often is invoked to explain the success of invasive plants: an invader benefits from reduced attack as it escapes specialized enemies through the invasion process. Enemy release typically is thought of as occurring at large geographic scales, but local-scale interactions may also be important for invader establishment and success. Furthermore, most tests of local enemy release have been conducted over a single year even though release may be a transient phenomenon, especially at small scales. In this study, we used a multi-year field experiment to investigate whether locally isolated populations of the noxious non-native weed Cirsium arvense benefit from reduced levels of aboveground damage, and whether any initial advantage is lost over subsequent growing seasons. Populations of C. arvense were grown in plots at set distances from established source populations for 4 years. In the first year of the experiment, folivory significantly declined with host isolation, but damage from specialist stem gallers and seed predators did not. However, in subsequent years of observation, folivores began colonizing isolated C. arvense populations while stem gallers exhibited very slow colonization of more isolated plots; seed predation showed no pattern with distance in any year. Local enemy escape did not result in increased plant performance, which instead negatively correlated with degree of isolation. Nonetheless, our results stress the importance of multi-year observations in tests of enemy release, since the herbivory patterns initially observed often changed within subsequent years depending on the dispersal ability and biology of the causal organism involved.