Two tests of enemy release of commonly co-occurring bunchgrasses native in Europe and introduced in the United States

The popularly cited enemy release hypothesis, which states that non-native species are released from population control by their enemies, has not been adequately tested in plants. Many empirical studies have compared damage to native versus non-native invaders only in the invaded range, which can lead to erroneous conclusions regarding enemy release. Biogeographical studies that have compared natural enemies in native and introduced ranges have typically focused on a small area of the plants' distributions in each range, only one plant species, and/or only one guild of natural enemies. To test enemy release, we first surveyed both pathogens and herbivores in multiple populations in both the native and naturalized ranges of three commonly co-occurring perennial bunchgrasses introduced to the United States from Europe. We then compared our field results to the number of fungal pathogens that have been documented on each species from published host-pathogen data compilations. Consistent with enemy release, our field survey showed less herbivory and denser populations in the naturalized range, but there was no evidence of release from pathogens. In contrast, the published host-pathogen data compilations produced evidence of enemy release from pathogens. The difference in results produced by the two approaches highlights the need for multiple approaches to testing mechanisms of invasions by introduced species, which can enable well supported theory to inform sound management practices.