LAUSR

Abstract Declines in North American freshwater biodiversity can be largely attributed to anthropogenic disturbances that reduce habitat quantity, quality, and connectivity. In the Laurentian Great Lakes, wetlands have been significantly lost as a result of agricultural land-use practice, yet high species richness and several at-risk fishes persist. These patterns lead to the hypothesis that Great Lakes wetland fishes are in extinction debt, meaning there is a time lag between habitat loss and local extinction. To test this hypothesis, we sampled current species richness of wetland-specialist fishes in protected and unprotected wetlands in the Lake Erie basin. We predicted species richness in unprotected wetlands using parameter estimates (slope and intercept) from protected wetlands. Extinction debt was measured as a positive difference between predicted and observed species richness. We determined how much area would be required to support current species richness and compared this to minimum-area requirements for eight at-risk fishes. Species richness was significantly higher than predicted in unprotected wetlands, indicating that fishes in them are in extinction debt. This result indicates that there is time available to restore disturbed wetlands and prevent local extinction. We identified 17 wetlands of high priority for restoration (>10 species bound for extinction) and determined that the restoration of 178 km2 of wetland habitat would reduce the risk of future biodiversity loss. This is the first direct study of extinction debt in freshwater fishes. Additional extinction-debt assessments in aquatic systems are needed to extend the ecological theory and fundamental application of extinction debt beyond terrestrial systems.