Can Rising Variance Predict Sudden Shifts in Populations and Ecosystems? A Test Using Alaskan Crustacean Data

Marine ecosystems may respond to external forcing with abrupt reorganizations that are economically and socially devastating to fishing communities.  There is currently no method allowing early warning of these reorganization events.  However, recent developments in ecosystem modeling suggest a novel approach to this problem: using the variance of key parameters to monitor ecosystem status, rather than the mean.  This “variance tracking” approach has been validated by empirical observations in the Gulf of Alaska and North Atlantic, suggesting that it may have real world management utility.  We are conducting an extensive evaluation of the variance tracking method, using commercially important Alaskan crustacean populations as a model system for developing parameter variance as an indicator of resilience in both ecosystems and individual fisheries.  Using retrospective analysis of extensive fisheries-derived and fisheries-independent data sets, we are testing key parameters for increases in spatial variance prior to the collapse of crustacean fisheries associated with ecological reorganization in Alaskan waters during the 1970s and 1980s.  We are also testing for “false positives” in situations where fisheries did not collapse, and we are comparing the ability of fisheries-derived and fisheries-independent parameters to forecast historical fisheries collapses.  Our study may contribute to an eventual tool for early warning of ecological reorganizations, and could also suggest an approach (preventing variance increases) for maintaining ecosystem resilience.