Detecting risk of regime shifts in ecosystems

Ecosystems around the world are at riks of critical transitions due to increasing anthropogenic preasures and climate change. However, it is not clear where the risks are higher, or where ecosystems are more vulnerable. When a dynamic system is close to a threshold, it leaves a statistical signature on its time series known as critical slowing down. It takes longer to recover after a small disturbance, which translates into increases in variance, autocorrelation, and skewness or flickering. Here I measure critical slowing down on primary production proxies for marine and terrestrial ecosystems globally. Slowness is an indicator of potential instabilities and a proxy of resilience. While slowness is not a universal indicator for critical transitions, it can be used for detection of potential regime shifts.