Ecosystem stability in space: Theoretical framework and dynamical models

Background/Question/Methods

Understanding the mechanisms governing ecosystem stability is essential for predicting the consequences of global environmental changes on ecosystem sustainability. The past two decades have seen great progress in understanding the mechanisms of ecosystem stability in local ecological systems. There is, however, an urgent need to extend existing knowledge to larger spatial scales to match the scale of management and conservation. Recent theoretical efforts have extended research to metacommunities and metaecosystems, which designate sets of communities and ecosystems, respectively, that are spatially distinct but connected by dispersal or movement of abiotic entities. Based on metacommunity models, we develop a general theoretical framework to study the stability and variability of ecosystems at multiple scales. We also develop dynamical models to examine how local (e.g. interspecific competition & species correlations) and spatial (e.g. dispersal & environmental correlations) factors affect ecosystem stability at different scales.

Results/Conclusions

We proposes the concepts of alpha, beta, and gamma variability. Gamma variability at regional (metacommunity) scale can be partitioned into local alpha variability and spatial beta variability, either multiplicatively or additively. This partitioning framework is analogous to that underlying the concepts of alpha, beta, and gamma diversity. We show that, on average, variability decreases from local to regional scales, which creates a negative variability-area relationship. We also show that diversity can provide insurance effects at the various levels of variability, thus generating alpha, beta, and gamma diversity-stability relationships. As a consequence, the loss of biodiversity and habitat impairs ecosystem stability at the regional scale.

In our dynamical models, although local interspecific competition and dispersal can affect the variability measured at local population and community level, neither of them has effects on gamma variability measured at the metacommunities level. Dispersal can reduce alpha variability, but it increases spatial synchrony simultaneously; these two effects cancel out at the metacommunity level. Besides, local species correlations and spatial environmental correlations both increase gamma variability. Overall, our results enables a synthetic understanding of ecosystem stability at multiple scales and has practical implications to predict the consequence of biodiversity and habitat changes.