The classic hypothesis that diverse native communities should resist invasions has grown into a more complex model that accounts for different processes controlling diversity at different spatial scales. Competitive exclusion may drive negative relationships between native species richness and invasibility at small scales, while resource or habitat heterogeneity emerge as dominant drivers of both native and exotic species at large scales, resulting in positive diversity-invasibility relationships. Past studies which correlated native richness with exotic richness or presence in small plots within sites have found that within-site correlations reverse from positive to negative in increasingly productive or diverse sites. Using data from the North American Breeding Bird Survey (BBS), our analysis addressed the central question: How does local community invasibility vary along gradients of native diversity and primary productivity at multiple spatial scales? We modeled the probability that a sampling location was invaded as a function of native richness separately for each BBS survey route. We then constructed and ranked multiple competing models to explain the variation in these logistic regression slopes as functions of cumulative native richness, net primary productivity (NPP), human population density and landcover-diversity aggregated within each route and across the routes within each state.
Results/Conclusions
Exotic richness of whole-route species inventories was positively correlated with native richness (r=0.21) and human population density (r=0.50), which was expected at larger spatial scales. However, the small-scale diversity-invasibility relationships summarized in the within-route regression slopes changed from positive to negative as cumulative route richness (r=-0.24) and human population density each increased (r=-0.34, df=744 for all correlations). The multiple regression analyses helped elucidate additional sources and patterns of variation in these small-scale diversity-invasibility relationships. The best supported model indicated that in states with low mean NPP (e.g. AZ, WY), slopes of within-route diversity -invasibility relationships reverse from positive to negative with increasing route native richness after accounting for human population density effects. In states with high mean NPP (e.g. FL, LA), the slopes of within-route diversity-invasibility relationships reversed from negative to positive with increasing route native richness. This result contrasts past findings that small scale diversity-invasibility relationships are either always negative or decrease monotonically along productivity gradients. While competition is an important factor affecting invasions within local communities, we find that regional productivity and diversity gradients provide a critical context for understanding the variation in local processes in bird communities.