Ecological communities experience a variety of stressors ranging from very short-term to long-term in duration and over highly variable magnitudes of effect. The ability of a community to withstand stress and maintain species composition and function or to return to the initial state following a perturbation are important issues in both basic and applied ecology. A considerable amount of research has been conducted on these issues and the concept of ecological resilience. However, resilience is a complex, multidimensional property of systems resulting from the interactions of species and their environments, and due to its complexity is inherently difficult to measure. We simulate ecological communities and use approaches based on multivariate statistical gradient analysis to develop and evaluate methods to characterize community-level resilience. Using our simulated communities, we determine the ability to detect various types of changes in species composition over time and whether or not communities return to a composition similar to their initial state.
Results/Conclusions
Our results show that the use of distance-based measures from multivariate ordinations provides a simple, effective visual framework to quantify the relative resilience of ecological communities. The methods provide a means to statistically detect when meaningful changes have occurred in community composition and if/when the communities return to their initial state. Our work provides an effective empirical approach to detecting ecological change that can provide insight in academic studies, but also provides an approach to assess environmental impacts.