INCREASED SPATIAL VARIANCE ACCOMPANIES REORGANIZATION OF TWO CONTINENTAL SHELF ECOSYSTEMS

Phase transitions between alternate stable states in marine ecosystems lead to disruptive changes in ecosystem services, especially fisheries productivity. We used trawl survey data spanning phase transitions in the North Pacific (Gulf of Alaska) and the North Atlantic (Scotian Shelf) to test for in...

Full description

Bibliographic Details
Main Authors: Litzow, Michael A., J. Daniel Urban, Laurel, Benjamin J.
Format: Article in Journal/Newspaper
Language:unknown
Published: Figshare 2016
Subjects:
Online Access:https://dx.doi.org/10.6084/m9.figshare.c.3294032
https://figshare.com/collections/INCREASED_SPATIAL_VARIANCE_ACCOMPANIES_REORGANIZATION_OF_TWO_CONTINENTAL_SHELF_ECOSYSTEMS/3294032
Description
Summary:Phase transitions between alternate stable states in marine ecosystems lead to disruptive changes in ecosystem services, especially fisheries productivity. We used trawl survey data spanning phase transitions in the North Pacific (Gulf of Alaska) and the North Atlantic (Scotian Shelf) to test for increases in ecosystem variability that might provide early warning of such transitions. In both time series, elevated spatial variability in a measure of community composition (ratio of cod [Gadus sp.] abundance to prey abundance) accompanied transitions between ecosystem states, and variability was negatively correlated with distance from the ecosystem transition point. In the Gulf of Alaska, where the phase transition was apparently the result of a sudden perturbation (climate regime shift), variance increased one year before the transition in mean state occurred. On the Scotian Shelf, where ecosystem reorganization was the result of persistent overfishing, a significant increase in variance occurred three years before the transition in mean state was detected. However, we could not reject the alternate explanation that increased variance may also have simply been inherent to the final stable state in that ecosystem. Increased variance has been previously observed around transition points in models, but rarely in real ecosystems, and our results demonstrate the possible management value in tracking the variance of key parameters in exploited ecosystems.