INTERANNUAL VARIABILITY OF THE EARLY-LIFE HISTORY OF WALLEYE POLLOCK NEAR SHELIKOF-STRAIT AS INFERRED FROM A SPATIALLY EXPLICIT, INDIVIDUAL-BASED MODEL

A coupled biophysical model is used to hindcast the early life history of a population of walleye pollock (Theragra chalcogramma), to assess possible physical causes of interannual Variability in recruitment. T he modelling approach combines a primitive equation, rigid-lid hydrody namic model with a probabilistic, individual-based biological model of growth, development, and mortality. Individuals are tracked through s pace using daily velocity fields generated from the hydrodynamic model , along with self-directed vertical migrations appropriate to each lif e stage in the biological model. The hydrodynamic model is driven with wind and runoff time series appropriate to each year. Biological mode l output compares favourably with observed spatial distributions for s pecific years. Lloyd's index of patchiness, calculated from model outp ut, was similar to values calculated from field data. Five noncontiguo us years were chosen for hindcasts to span a wide range of meteorologi cal conditions (winds, runoff) and recruitment success. Interannual co mparisons suggest that two years of above average recruitment (1978 an d 1988), and one year of below average recruitment (1991), experienced flow fields which carried many individuals into the Alaskan Stream. A t the same time, the vigorous flow fields generated in each of these y ears carried some individuals onto the shelf area to the south-west of the spawning site. A year with low runoff and weak winds (1989) exhib ited weak circulation, with extended retention of larvae near the spaw ning site. A year with high runoff (1987) was notable for the strength and frequency of mesoscale eddy activity. Eddies appear capable of bo th enhancing patchiness of early larvae (through retention) and dissip ating patchiness of juveniles (through mesoscale mixing), Larvae retai ned in an eddy feature exhibit a narrower range of sizes than the popu lation outside that feature.