COS 58-4
Long-term patterns in species richness, diversity, and community heterogeneity: a test of the environmental heterogeneity hypothesis in the context of community assembly
Understanding what constrains the persistence of species in communities is at the heart of community assembly theory and highly relevant to restoring biodiversity. Local processes can constrain a species’ membership and relative abundance to affect richness and diversity. Following arrival, presence of a species in a community is determined by tolerance to local conditions and relative abundance by biotic interactions. By this logic, increasing environmental heterogeneity in key resources to vary the strength of abiotic filters and competitive interactions should affect the relative abundance of species at small scales and biodiversity at larger scales. In 1998, we manipulated plot-level heterogeneity by altering soil depth and soil nitrogen availability prior to sowing species in a former agricultural field. We quantified species composition in response to the plot-level heterogeneity treatments over 15 years of community assembly to test the ‘environmental heterogeneity’ hypothesis. Temporal dynamics in richness and diversity were compared between the plot-level heterogeneity treatments (maximum heterogeneity plots containing all combinations of two soil depths and three levels of soil nitrogen vs. control plots containing no manipulations) and among subplot-level nutrient and depth treatments using mixed model analyses of the split-block design with repeated measures (MIXED) and analyses of covariance (ANCOVA).
Results/Conclusions
Long-term community dynamics did not reflect short-term responses to the plot-level treatments. During the initial years of this experiment, nutrient availability was a strong determinant of community structure demonstrated by an inverse relationship between diversity and nitrogen availability. During this period, there were also limited effects of soil heterogeneity on assembling plant community structure. Following 15 years of manipulated soil heterogeneity, there is evidence for divergence in multiple aspects of plant community structure between the control and maximum heterogeneity treatments. Community heterogeneity (within-plot proportional dissimilarity) was higher in space and over time in the maximum heterogeneity plots (MIXED: P = 0.025). Richness and diversity declined over 15 years in both plot-level heterogeneity treatments, but the rate of species loss was lower the maximum heterogeneity plots (ANCOVA: P = 0.009). Subplots manipulated to enrich or reduce nitrogen exhibited lower rates of species loss than the ambient nitrogen soil (ANCOVA: P = 0.039), which explained the faster rate of decline in richness in control relative to maximum heterogeneity plots. A consistent, gradual shift in the dominant species across all treatments represents an over-riding biotic filter responsible for the low magnitude of community response to environmental heterogeneity thus far.