The effects of plant community structure on natural enemies (diseases and herbivores) and the role that they play in invasions by exotic species have long been studied. There is broad support for the reduction of enemy attack as community diversity increases. With respect to invasions by exotic species, two roles of enemy attack have received attention. First, according to the enemy release hypothesis, introduced species flourish because they leave their natural enemies behind in their native range and are no longer regulated by them. Alternatively enemy spillover may occur, in which the natural enemies of introduced species invade a new region with their host, and attack other species in the community. We examined the effects of diversity, exotic species dominance, dominance of the host species, and other factors on foliar damage by herbivores and fungal pathogens to six native plant species in a seasonal wetland prairie restoration experiment in Eugene, Oregon. The species we surveyed are Agrostis exarata, Deschampsia cesptiosa, Madia glomerata, Prunella vulgaris, Grindelia integrifolia and Epilobium densiflorum. Combined these species comprise 80% of community plant cover, and our survey spanned a gradient of diversity and relative cover of native and exotic species.

Results/Conclusions We found more combined pathogen and herbivore damage on the most common exotic grass Lolium multiflorum than on either of the two common native perennial grasses A. exarata or D. cespitosa, indicating that L. multiflorum harbors more natural enemies than the two native grasses. Multiple regression analysis using Akaike’s information criterion on 14 factors revealed that different factors influenced herbivores vs. pathogens, and the importance of individual factors varied among the six native host species. E. densiflorum dominance was strongly positively associated with herbivory on itself and G. integrifolia. Exotic species dominance was positively correlated with herbivory on M. glomerata and G. integrifolia, but negatively correlated with herbivory on A. exarata and P. vulgaris. L. multiflorum dominance was positively associated with pathogens on the two native grasses and G. integrifolia, but negatively associated with pathogens on P. vulgaris. Total cover was strongly positively associated with pathogens on M. glomerata and E. densiflorum. Diversity had a relatively small effect, but was negatively correlated with pathogen damage on four of the six species. Our results indicate that enemy spillover from exotics to natives occurs in some cases for both herbivores and pathogens, and community diversity has a small but consistent effect on pathogens across the community.