Towards a multi-stressor theory for coral reefs in a changing world

Coral reefs are facing a constant barrage of human impacts, including eutrophication, overharvesting and climate change. While the local effects of overharvesting have been well-studied, regional nutrient loading from anthropogenic activities on land and global climate change-induced disturbances are increasing in magnitude and necessitating cross-scale multi-stressor approaches for coral reef ecology. Here, we expand on longstanding theory to develop an integrated multi-stressor framework for coral reefs. We show that: i) the geometry of a simple, empirically-motivated model suggests nutrients and harvesting can operate similarly, and synergistically, in driving shifts from coral- to algae-dominated reefs, resulting in clear context-dependent management implications; and ii) this same geometry suggests climate-driven coral mortality can drive the presence of long transients and climate-driven alternate states, even in moderately-impacted ecosystems. Reefs seemingly in a "safe space" based on individual stressors may in fact be much more susceptible to increasingly frequent storms and bleaching events in multi-stressor conditions. By integrating these findings with general ecological and theoretical concepts, we suggest that responses in benthic composition may act as "signatures of change" to multi-stressors, allowing us to develop a predictive and generalizable multi-stressor framework for coral reefs under global change. In line with this theory, we detail empirical evidence from Barbados of historical changes in reef composition and multi-stressor impacts within our framework. By bridging coral reef ecology and general ecological concepts, we can better understand ecosystem functioning and resilience in these important yet highly threatened systems. Manuscript HighlightsO_LITheoretical understanding of synergistic multi-stressor impacts on coral reefs C_LIO_LIUnexpected climate-driven alternate states, related to long transients C_LIO_LITheoretical framework predicts "signatures of change" based on dominant stressor C_LI