Inversion of dominance-diversity relationships along a latitudinal stress gradient

Background/Question/Methods   Knowing how stochastic and deterministic processes interactively structure biotic communities may improve ecosystem management, but a synthetic empirical understanding of how such interactions operate in different environmental contexts is lacking. We investigated whether native herbaceous oak savanna communities were limited by seed dispersal, and whether the importance of dispersal limitations depended upon the extent to which communities were dominated by exotic perennial grasses. A 5-year field experiment was conducted, comprised of seeded and not-seeded subplots nested within whole plots exposed to different disturbance treatments (including no disturbance and different combinations of mowing, burning, and herbicide application) designed to differentially reduce exotic perennial grass cover. Richness and cover of seeded species were evaluated as functions of exotic perennial grass cover, and relationships observed in seeded and not-seeded subplots were contrasted to detect how degree of dominance influenced strength of seed limitation. To determine how relationships between dominance and establishment diversity depended upon environmental context, our experiment was replicated at 10 sites spanning a 500 km latitudinal gradient. Principle Components Analysis (PCA) was employed to reduce complexity of several collinear stress/resource variables coincident with this latitudinal gradient (e.g. soil moisture, fertility). The first PCA eigenvector (termed here the Site Stress Index) was incorporated as a covariate in analysis of dominance-diversity relationships.

Results/Conclusions   Dominance by exotic perennial grasses influenced the extent to which native communities were seed-limited, but the nature of this influence varied strongly with the Site Stress Index. At moderately-stressed sites, reduction of dominance had no influence on the degree to which richness and cover of seeded species increased with seed addition. In contrast, at low-stress sites dominance-reduction enhanced the degree to which richness and cover of seeded species increased with seed addition, while at high-stress sites dominance-reduction weakened species responses to seed addition. Dominance thus appears to have divergent roles in structuring communities depending on the level of stress in the regional environment. Under low-stress conditions dominant species can suppress the degree to which communities are limited by dispersal, presumably through effects of competition. However, under high-stress conditions dominant species can increase community sensitivity to immigration, presumably by ameliorating local stress levels and facilitating establishment of arriving propagules. These results suggest that management of dominant species should account for both negative and positive effects of dominance on diversity that may result from stress-mediated switching in the direction of interspecific interactions.