Detailed information on the catch, escapement, migration timing and behaviour of northern boundary sockeye salmon stocks has been obtained through national and international studies conducted over the past 20 years. The Northern Boundary Annex to the 1999 Canada/U.S. Salmon Treaty agreement requires detailed accounting of the harvests for Nass and Skeena sockeye stocks. The Northern Boundary Technical Committee was asked to provide the catch, escapement and total run size estimates for the regional sockeye stocks required to implement the new Treaty provisions. A review of the existing run reconstruction data and analyses was initiated in 2001, and lead to the preparation of a jointly approved series of run reconstructions for the 1982-2001 harvests. This report describes the revisions made to the catch and escapement data, migration parameters and assumptions used in the run reconstructions, and compares the revised results with those reported by Gazey and English (2000).

Two run reconstruction methods are used. The first assumes that the stock composition of the harvest in each fisheries is determined by the relative abundance of each stock in the fishery (i.e. the fish present in a fishery are equally vulnerable to the fishing gear). The second assumes that stock contributions to Alaskan catches are known from scale sample data analyses and relies upon the equal vulnerability assumption to estimate stock specific harvests for Canadian fisheries. Both methods use information on migration patterns that were stock-specific and fixed throughout each year, but variable across years. An initial set of migration route parameters for each stock was generated based on the 1982-83 north coast tagging studies results. Alternative sets of migration parameters were developed by adjusting the initial set of values until the estimated catches by stock and week in the Alaskan fisheries derived assuming equal vulnerability were similar to those derived using the scale stock composition estimates for these Alaskan fisheries. Five sets of migration parameter values were required to fit the 1982-2001 scale data, with one set providing the best fit for 14 of the 20 reconstruction years. The run reconstruction results derived using the second method were considered by the Canadian and Alaskan fisheries managers to be the most reasonable estimates of stock size and catch by stock for sockeye harvested in northern boundary fisheries. Consequently, the migration route parameters only affect the stock specific catch estimates for Canadian fisheries, with the vast majority of the Alaskan harvests by stock determined using the scale data or the definition of terminal fishing areas for Alaskan stocks.

The revised estimates for the Skeena stock for 1982-95 were all greater than the earlier estimates because of the combined effect of increasing the annual escapement estimates for the Skeena and decreasing the escapement estimates for Alaskan stocks. The increase in the escapement estimates for Nass sockeye resulted in higher stock size estimates for 1982-1990. The reductions in the estimates of the size of the Nass stock in 1991-1995 were the combined affect of changes to the migration routing parameters, harvest data and escapement estimates.

The 1999 Annex states that the Tree Point gillnet fishery may harvest, on average, 13.8% of the Annual Allowable Harvest (AAH) for Nass sockeye. This exploitation rate was based on earlier analyses of the 1985-1997 catch and escapement data, conducted during the negotiation of the annex. Despite the numerous and substantial changes to both BC and Alaskan catch and escapement data used in our analyses, we estimate that the 1985-1997 average exploitation rate for Nass sockeye in the Tree Point fishery was 13.9% (Table 14). Since 1997, exploitation rate for Nass sockeye in the Tree Point fishery has averaged 15.6% and ranged from harvested 20.3% in 1998 and 1999 to 10.9% in 2000.

The 1999 Annex also states that the Alaska District 104 fishery may harvest, on average, 2.45% of the combined Skeena-Nass AAH in fisheries prior to the end of July (Week 31). This percentage was based on an earlier analysis of 1985-1997 harvest of Skeena and Nass stocks in the District 104 fishery, where the percent Skeena-Nass in the District 104 fishery prior to Week 31 was applied to a fixed maximum harvest of 120,000 sockeye. We have repeated these analysis procedures using the results from the run reconstruction and derived an estimate of 2.34% for the 1985-1997 base period. The main reason for the difference between this value and that defined in the 1999 Annex is the increase in the escapement estimates for Skeena and Nass sockeye. The run reconstruction results indicate that the District 104 catch prior to Week 31 was, on average, 3.30% of the Skeena-Nass AAH during the 1985-1997 period and 2.25% of the Skeena-Nass AAH from 1999-2001. The later estimate indicates that recent harvests in District 104 have been, on average, less than the harvest share defined in the 1999 Annex and our estimate of 2.34% based on the revised catch and escapement data.

The run reconstructions reported here represent a systematic application of an analytical technique that allows the fisheries managers and analysts to combine all the available information for northern boundary fisheries into a single process to better understand the relative and potential contributions of majors stocks to each fishery. The process of organizing the data required for run reconstruction analyses reveals data gaps and assumptions critical to the interpretation of annual harvest statistics. The combined assessment of northern BC and Alaskan stocks by the Northern Boundary Technical Committee promotes an increased understanding of each others stocks and fisheries and an ongoing need to work cooperatively in the management and assessment of these valuable stocks.

This report, and other technical reports of the Pacific Salmon Commission, are available on their website at http://psc.org/publications.htm