Ecologists increasingly recognize the importance of understanding the causes and consequences of heterogeneity in ecological communities. By heterogeneity, we mean variability in an ecological factor in space and/or time. Global environmental change is a well known driver of spatial and temporal heterogeneity in ecosystems, thus understanding the relationship between spatial and temporal heterogeneity would be valuable for predicting how environmental drivers will alter the rate at which ecological communities will change over time. Theory predicts that spatial heterogeneity can drive temporal stability at many scales. Additionally, patterns have emerged from case study examples that support these predictions. Until recently, little data have been available to test if a general relationship exists between spatial and temporal heterogeneity. To address this deficiency, we empirically evaluated whether spatial heterogeneity within a community predicted temporal heterogeneity of that community from one year to the next. Based on existing theory, we predicted that spatial heterogeneity would be positively related to temporal heterogeneity both within and across sites. We tested this hypothesis by conducting a synthesis of 68 datasets spanning aquatic and terrestrial systems that included replicate measures of species abundance over space and time.
Results/Conclusions
Using mixed effects models, we found that the model incorporating spatial heterogeneity better predicted temporal heterogeneity than a null model (ΔAIC=2.9, r2=4.9, p=0.027) and that spatial heterogeneity was positively related to temporal heterogeneity. Thus communities with higher spatial heterogeneity changed more through time on average compared to communities with low spatial heterogeneity. Of the 68 communities 55 (81%) had weak to strongly positive slopes between spatial and temporal heterogeneity while the remaining communities were weak to strongly negative (19%). No experimental predictors (plot size, number of plots, spatial extent, length, and sampling interval time) had a significant influence on temporal heterogeneity. In the biotic model, spatial heterogeneity was again significant, however life span was the strongest predictor of temporal heterogeneity. Temporal heterogeneity was lower for communities with longer-lived species and higher in communities with sub-annual or annual species lifespans. Succession was positively and precipitation was negatively predictive of temporal heterogeneity in the biotic model. Based on these findings, we suggest that the positive relationship between spatial and temporal heterogeneity may be a general, albeit weak, property of many ecological communities.