U.S. Dept. of Commerce / NOAA / OAR / PMEL / Publications
NOAA's National Marine Fisheries Service (NMFS) advises the North Pacific Fisheries Management Council on the status of pollock stocks in the Gulf of Alaska and Bering Sea. The process includes a stock projection model, initialized with results from a model of stock assessment. The latter model employs commercial catch statistics and operational survey information to produce estimates of stock abundance. Given these results, the stock projection model forecasts future abundance as a function of different harvest and recruitment scenarios. The stock assessment model requires information independent of the fishery to calibrate estimates of absolute population. Prior to FOCI guidance, these were limited to the hydroacoustic and bottom-trawl surveys. For the last 2 years, estimates of spawning biomass for the Gulf of Alaska derived from the Annual Egg Production Method [Picquelle and Megrey, 1993] have provided a third source of information.
The effects of management decisions on the Shelikof Strait pollock population are examined using alternative harvest and recruitment scenarios in the stock projection model. Beginning in 1992, FOCI has analyzed biological and physical time series to estimate recruitment qualitatively. This prediction significantly simplifies the stock projection analysis by limiting the number of viable recruitment scenarios. The forecast of year-class strength uses hydroacoustic survey results of spawning aggregations, ichthyoplankton surveys of eggs and larvae, estimates of spawning biomass/recruitment from the annual stock assessment [Hollowed et al., 1993], and a suite of biological and physical factors. Previous modeling of pollock recruitment [Megrey et al., 1995] did not address the autocorrelated nature of the recruitment data. In 1993, the recruitment modeling was augmented with the development of a transfer function model. This model uses the same physical environmental data as before, but accounts for the autocorrelation of recruitment. In addition, it directly predicts recruitment. In 1993 it was used to generate recruitment scenario candidates for the stock projection model. FOCI's prediction made in 1992 was for weak 1989-90 year classes, a weak to average 1991 year class, and a strong 1992 year class. Recent observations of recruitment to the 1989 and 1990 year classes are consistent with this prediction.
Acknowledgments. We thank the numerous scientists who planned, conducted and published their research, and those who have provided information not yet published. In particular, Phyllis Stabeno, Ron Reed, and Jeff Napp for their continuous support and critical comments on this manuscript, and Allen Macklin, FOCI Program Coordinator. We also thank all the technical staff who assisted with the data acquisition, preparation, and analysis. Continued superlative support has been provided by the NOAA Ship Miller Freeman, and on specific cruises by the NOAA Ships Discoverer and Surveyor. A portion of this research and analysis was funded by Minerals Management Service, Interagency Agreement #14-35-0001-14165. The majority of research was funded by Fisheries Oceanography Coordinated Investigations and NOAA's Coastal Ocean Program. Wopila Tunkashila, Maki Unci and Grandmother Ocean. This is FOCI contribution #212 and Pacific Marine Environmental Laboratory contribution #1535.
Return to previous section or go to next section