Exotic species invasions are one of the principal threats to populations and communities of native organisms. Despite the severity of these threats, the degree of invasion varies widely among communities, due to processes occurring in the landscape surrounding invaded communities (e.g., propagule pressure), as well as intrinsic site characteristics (e.g., resource availability, community structure). However, the relative influence of these processes for controlling invasion remains poorly resolved, in part because these and other processes may represent either direct or indirect effects on invasive species. Importantly, this limits the capacity for developing management actions to promote invasion-resistant ecosystems. We used structural equation models to evaluate how external factors, specifically those influencing propagule pressure (landscape context and land-use history), and site characteristics, including resource availability (soil moisture) and community structure (richness of sown species, abundance of the community dominant Andropogon gerardii) related to the degree of invasion (richness and abundance of invasive species) in 29 tallgrass prairie restorations. We also tested whether management factors (seed mix richness and fire frequency) supported invasion resistance directly by reducing the richness or abundance of invasive species, or indirectly by modifying site characteristics.
Results/Conclusions
We found that only site characteristics, specifically sandy, dry soils and A. gerardii abundance, limited invasive richness. In contrast, invasive abundance was associated with all three classes of drivers - management, external factors, and site characteristics. Invasive species abundances were dramatically lower in sites restored from row crops (1.6% cover) than in sites restored from hay fields or old fields (23.8% cover). Seed mix richness limited invasive species abundances indirectly by increasing the richness of sown species, which in turn was generally associated with lower abundances of invasive species. The abundance of A. gerardii also generally reduced invasive abundance. In contrast, fire frequency had virtually no direct or indirect effect on invasive richness or abundance. Our results illustrate how a suite of non-independent processes, including propagule pressure, resource availability, and community structure, can together affect the degree of invasion, but that processes varied in their importance and depended on whether invasion was measured through richness or abundance. From a practical perspective, managers can resist invasion by altering the restored community through seed additions. However, land-use legacies may overwhelm the importance of restored community structure, requiring management to reduce invasive species abundance (e.g., in seed banks) prior to restoration.