Gene J Sandone’s research while affiliated with Alaska Department of Fish and Game and other places

ABSTRACT A modified form of catch-age analysis, which used limited observations of absolute abundance as auxiliary information, was applied to Prince William Sound herring (Clupea harengus pctlktsi). Catch-at-age data were from the 1973-88 purse seine and gill net sac roe fisheries in Prince William Sound. Estimates of spawning biomass from spawn deposition surveys conducted in 1984 and 1988 were used as auxiliary information. The catch-age analysis model estimated exploitation rates for each gear by year, gear selectivity coefficients by age and gear, partial recruitment coefficients for age-3 and age-4 herring, initial age-3 cohort sizes in each year, and the size of cohorts of all ages in 1973. From these estimates and the results of age-weight sampling, estimates of population size and spawning biomass were constructed. Spawning biomass estimates averaged almost double the peak biomass estimates from aerial surveys. Exploitation rates ranged from 25%-35% from 1972 to 1977 and averaged 12% from 1979 to 1988. Maxi- mum gear selectivity for purse seine gear was reached about age 6, whereas gill net gear had a much more restricted dome shaped selectivity with a maximum at age 8. Over the 1973-88 period, 78% of the age-3 herring were estimated to be recruited and present on the fishing and spawning grounds. Almost all age-4 herring were estimated to be fully recruited. Based upon the range of literature estimates of natural mortality, a survival rate of 64% was used for the catch-age analysis estimates. Lower survival rates changed the biomass estimates very little but dramatically increased initial cohort sizes. Higher survival rates had the opposite effect. Based upon recruit year class strengths observed from 1973 through 1988, a strong year class of at least 300 million age-3 recruits would be expected about every 4 years on the average. No overall relationship between spawners and recruits was evident, although a positive density dependent trend was evident for recent data (1979-85 year classes). We suggest that the catch-age analysis should be updated each year and that the analysis should incorporate the latest year's stock assessment information. In addition to generating biomass estimates for the current year, this process would update the estimates of biomass for prior years. The model's estimates of partial recruitment coeffl- cients would then be used to project the age 4 to age 9+ biomass that should return the following year. The model's estimates of average age-3 recruitment over time could be used to set a range of possible age-3 recruitment scenarios for the upcoming year.