Partitioning species contributions to ecological stability in disturbed communities

Ecosystems worldwide are experiencing a range of natural and anthropogenic disturbances, many of which are intensifying as global change accelerates. Ecological responses to those disturbances are determined by both the vulnerabilities of species and their interspecific interactions. Understanding how individual species contribute to the (in-)stability of an aggregated community property, or function, is fundamental to ecological management and conservation. Here, we present a framework to identify species contributions to stability based on their absolute and relative responses to disturbances. Using simulations, we show that these two dimensions enable identification of (de-)stabilizing species and reveal that competitive dominance determines the magnitude of both absolute and relative contributions to stability. Applying our framework to empirical data from a multi-site mesocosm experiment showed that species contributions varied among treatments, sites, and seasons. Despite this dependency on both biotic and abiotic contexts, species contributions were generally constrained by their relative dominance in undisturbed conditions. Rare species contributed positively to stability, while dominant species contributed negatively, indicating compensatory dynamics. Our framework offers an important step toward a more mechanistic understanding of ecological stability based on species performance.