Heterogeneous habitats are predicted to have higher species diversity, suggesting they may be more susceptible to exotic invasions. However, native species may take advantage of habitat heterogeneity to coexist with invaders. Examining spatial and temporal variability in the composition of invaded plant communities can provide insight into how native and exotic species utilize environmental heterogeneity and how invasion and coexistence processes operate within these communities. Because spatial and temporal heterogeneity in environmental conditions occur simultaneously, with varying effects at different spatio-temporal scales, examining the relative roles of each on plant community structure is critical to understanding species coexistence in and informing management of these communities. Furthermore, understanding contingency of community responses at various spatial and temporal scales is necessary to justify extrapolation of results of local research or management activities to other locations or years. We used a large-scale spatially nested, 5-year plant community data set in the California grasslands to investigate spatio-temporal variation in community structure. Our goal was to determine how space and time at several scales contributed to variation in plant communities and which species were the strongest drivers of these patterns.

Results/Conclusions

Communities were tightly grouped by site in a 3-dimensional ordination produced by nonmetric multidimensional scaling, with two exotic annual grasses appearing to drive the strong latitudinal gradient in community structure. Blocks nested within sites were grouped more loosely. Ordinations at finer spatial scales suggested that different species were driving patterns in community structure at these scales. Temporal variability in communities tended to be greater than spatial variability at the finest scales. Temporal variability also increased with increasing time lags. Interannual variability in communities differed among sites and, in one case, among blocks within a site. However, spatial variables explained little of the variation in interannual variability, suggesting that yearly variation in species abundances depends primarily on temporally-varying environmental factors. Overall, these results suggest that patterns in community structure and the species driving those patterns vary across spatial and temporal scales, thus findings from community ecology research or management actions may be contingent on scale and location at which they are conducted. Furthermore, these patterns suggest that species with shared provenance and life history may significantly differ ecologically, thus extrapolating environmental responses and roles of individual species to groups based on these characteristics must be done with caution.