Print PageDisturbance experiments typically focus on changes in average values of diversity or abundance as the main response variables. The predictability of the response, defined as the amount of variation across replicates of the same treatment, is often overlooked despite its importance as a consequence of disturbance. Using 72 manipulative studies that included a control and a disturbed treatment we performed a meta-analysis on among-replicate variability in changes in abundance (n=162) and diversity (n=73) following five different types of disturbances including species removals, habitat loss and fragmentation, species additions/invasions, eutrophication, and climate warming, and further compared among-replicate variability for experiments conducted in terrestrial versus aquatic systems and between producers and consumers.
Results/Conclusions
Our results show that response predictability of disturbed treatments is not significantly different from controls across all disturbance types. Instead, the effects of disturbance on response predictability are highly contingent on the type of disturbance, with biotic disturbances leading to greater changes in response predictability than abiotic disturbances. Response predictability also differed according to habitat type and trophic position with stronger effects observed in terrestrial systems than in aquatic systems and stronger effects for producers than consumers. The greatest decreases in response predictability were observed in terrestrial systems following species removals and in aquatic systems following nutrient addition showing that removing species, particularly producers, results in significant divergence in community structure in terrestrial systems while nutrient addition is the most destabilizing disturbance in aquatic systems. In contrast, response predictability in diversity increased following species invasions in aquatic systems showing a widespread homogenizing effect of species invasions on community structure in aquatic systems.
