A global synthesis of marine ecosystem regime shifts

Background/Question/Methods

A growing body of scientific evidence suggests that marine ecosystems around the world experience abrupt and dramatic regime shifts. These ecosystem-level shifts have important implications for ocean management as they are often unforeseen, can be rapid, large, and difficult to reverse and have direct impacts on people’s livelihoods and wellbeing. While regime shifts are increasingly documented across geographies and ecosystems, a comprehensive synthesis of where and when these shifts have occurred and their drivers and consequences in marine ecosystems has not been compiled. Building upon existing efforts (e.g., databases from the Resilience Alliance and Stockholm Resilience Centre), we have assembled a systematic global database of regime shift examples across marine ecosystems, from the intertidal to the open ocean. This database includes regime shifts that were driven both by changes in climatic or oceanic conditions as well as those driven by anthropogenic stressors, like harvest or nutrient input. Our aim is to provide a useful compendium of examples of ecosystem-level tipping points in marine systems, as well as a robust dataset that can help answer fundamental questions about the drivers, mechanisms, and consequences of these shifts.

Results/Conclusions

Through analysis of over 100 case examples we found numerous commonalities between regime shift attributes across ecosystems, geographies, and spatial and temporal scales. Regime shifts were recorded in kelp, coral, seagrass, pelagic, salt marsh, lagoon, estuarine, rocky reef, tidal flat, and oyster reef ecosystems.  Across most ecosystems the resulting shift remained stable on the temporal order of decades (20+ years). Kelp forests were a notable exception, with shifts most often remaining stable on the order of years. The majority of regimes shifts were attributed to one major stressor, with changes in harvest, climate, and eutrophication contributing in over 80% of the shifts recorded. These three major stressors were also the most likely to be cited as acting simultaneously to induce a shift. While ecosystem responses are often classified as following a “linear”, “abrupt” or “hysteretic” path, there is still a need for common, data-driven approaches to characterizing these dynamics, investigating the causal mechanisms, and rigorously testing the role of path dependency in these transitions. We hope that this effort will contribute significantly to the broader goal of helping managers recognize where ecosystem shifts have the potential to occur and how to best anticipate, avoid, or respond to these changes.