Are invasive species the drivers or passengers of change in degraded ecosystems?

Few invaded ecosystems are free from habitat loss and disturbance, leading to uncertainty whether dominant invasive species are driving community change or are passengers along for the environmental ride. The driver model predicts that invaded communities are highly interactive, with subordinate native species being limited or excluded by competition from the exotic dominants. The passenger model predicts that invaded communities are primarily structured by noninteractive factors (environmental change, dispersal limitation) that are less constraining on the exotics, which thus dominate. We tested these alternative hypotheses in an invaded, fragmented, and fire-suppressed oak savanna. We examined the impact of two invasive dominant perennial grasses on community structure using a reduction (mowing of aboveground biomass) and removal (weeding of above- and belowground biomass) experiment conducted at different seasons and soil depths. We examined the relative importance of competition vs. dispersal limitation with experimental seed additions. Competition by the dominants limits the abundance and reproduction of many native and exotic species based on their increased performance with removals and mowing. The treatments resulted in increased light availability and bare soil; soil moisture and N were unaffected. Although competition was limiting for some, 36 of 79 species did not respond to the treatments or declined in the absence of grass cover. Seed additions revealed that some subordinates are dispersal limited; competition alone was insufficient to explain their rarity even though it does exacerbate dispersal inefficiencies by lowering reproduction. While the net effects of the dominants were negative, their presence restricted woody plants, facilitated seedling survival with moderate disturbance (i.e., treatments applied in the fall), or was not the primary limiting factor for the occurrence of some species. Finally, the species most functionally distinct from the dominants (forbs, woody plants) responded most significantly to the treatments. This suggests that relative abundance is determined more by trade-offs relating to environmental conditions (long-term fire suppression) than to traits relating to resource capture (which should most impact functionally similar species). This points toward the passenger model as the underlying cause of exotic dominance, although their combined effects (suppressive and facilitative) on community structure are substantial.