POPULATION-DYNAMICS AND MANAGEMENT OF WALLEYE POLLOCK (THERAGRA-CHALCOGRAMMA) IN THE GULF-OF-ALASKA, 1976-1986

Information from resource assessment surveys and commercial fisheries statistics was used to investigate the population dynamics of Gulf of Alaska walleye pollock (Theragra chalcogramma) from 1976 to 1986. Stock assessment analyses were used to estimate historical stock size, rates of fishing and numbers of recruits. Three different age-structured stock assessment models were fitted to two data sets: a nonlinear log catch model calibrated with fishery-independent population biomass estimates, a nonlinear catch/effort model, and a linear model based on catch per unit of effort. Variances of parameter estimates were obtained when possible. Results indicate that (1) a shift in selectivity-at-age occurred between the 1976-1981 and 1982-1986 periods; (2) population biomass was 1 000 000 t in 1976, peaked at 2 500 000 t in 1981-1982 and dropped to 400 000 t in 1986; and (3) annual changes in biomass were influenced mainly by five consecutive strong year classes (1975-1979) moving into the fishery followed by four weak year classes (1980-1983). The decline in abundance is supported by independent biomass estimates from hydroacoustic/midwater trawl surveys. Results from different catch-at-age models compare favorably with respect to estimated trends in abundance. Differences in scale are attributed to the data set, the choice of the model, and the way the models were constrained. Estimates of instantaneous natural mortality (M year-1) ranged from 0.35 to 0.50 from stock assessment methods and from 0.23 to 0.44 from empirical and approximate methods. Catchability estimates ranged from 0.014 (M = 0.42) to 0.0191 (M = 0.35). It also appears that catchability varies interannually and may be density dependent. A Ricker spawner-recruit model explained 83.5% of the variation in recruit abundance, indicating that density-dependent processes may be important in pollock population dynamics.