|Document Type:||Contract Report|
|Title:||Genetic stock identification study|
|Author/Editor:||George B. Milner, David J. Teel, Fred M. Utter|
|Publisher:||National Marine Fisheries Service|
|Contracting Agency:||Bonneville Power Administration. Portland, Oregon|
The contributions of wild and hatchery stocks to the May 1982 troll fishery of chinook salmon off the Washington coast were estimated by the genetic stock identification (GSI) method. A description of the GSI method was presented including the model and statistical procedure for obtaining maximum likelihood estimates of stock contributions. A final data base for analysis of this fishery (derived by identifying genetic variants expressed in liver and heart tissues, and eliminating populations estimated to be non-contributors) included 14 polymorphic loci from 49 stocks and 17 stock groups from California through southern British Columbia. GSI estimates from approximately 2,000 mixed fishery samples indicated that about 90% of the catch originated from the Columbia River including a 76.5% total contribution returning to the stock group of the Lower Columbia River and Bonneville Pool areas. Separate analyses of catches from the northern and southern areas indicated a much higher northern area representation of Canadian (increasing from 1 to 22%) and Puget Sound (increasing from 1 to 9%) stock groups, and a substantially reduced contribution of the Lower Columbia River and Bonneville Pool stock group (decreasing to 45.6% from 78.9%). Accuracy of GSI estimates compare favorably with average estimates of previous years based on such information as abundances in terminal areas and analyses of scale, fin clipping, and coded–wire tag data; the data base was therefore regarded as adequate for analysis of this fishery. Possible improvements in the procedure included more efficient sampling by collecting tissues at landing sites rather than on shipboard and more efficient analyses through a continual search to increase the number of polymorphic loci.