Ecosystem change and decadal variation in stock–recruitment relationships of Lake Erie yellow perch (Perca flavescens)

Abstract

Abstract Fish stock–recruitment relationships (SRRs) may vary in response to ecosystem change, increasing uncertainty for fisheries management. We defined three periods between 1975 and 2015 over which Lake Erie, a Laurentian Great Lake, underwent significant ecosystem changes: before zebra mussel (Dreissena polymorpha) establishment, after zebra mussel establishment and before re-eutrophication, and after re-eutrophication. To examine the extent to which SRRs of Lake Erie yellow perch (Perca flavescens) also varied over these periods, we compared the performance of Baseline (constant recruitment), Ricker (constant SRR), Periodic Ricker (different SRRs among three periods) and Random-walk Ricker (annually varying SRRs) models fitted to data for yellow perch stocks corresponding to three lake basins. Periodic and Random-walk Ricker models performed better for stocks in the western and eastern basins, but the Baseline model performed best in the central basin. Annual variation in the SRRs coincided with the timing of zebra mussel establishment and re-eutrophication in the shallower western basin, but not in the deeper eastern basin, where quagga mussels (Dreissena bugensis) established later and conditions are less eutrophic. These results underscore that temporally and spatially varying SRRs associated with ecosystem change should be taken into account in models of fish population dynamics.