Beacham: The use of DNA variation for stock identification of Oncorhynchus keta 



623 



no marked distinction among Fraser River stocks 

 (Beacham et al., 1987). Minisatellite DNA variation 

 may provide the means to discriminate among Fraser 

 River stocks if necessary in management applica- 

 tions. In the Yukon River drainage, summer-run 

 stocks were well differentiated from fall-run stocks 

 by analysis of variation at protein-coding loci ( Wilmot 

 et al., 1992, 1994). However, in the fall-run stocks, 

 which spawn in rivers in both Alaska and the Yukon 

 Territory, some stocks spawning in Alaskan rivers 

 (Sheenjek and Chandalar rivers) were more similar 

 to stocks spawning in the Yukon Territory rivers in 

 Canada than they were to other Alaskan fall-run 

 stocks. The addition of minisatellite DNA variation 

 to techniques of stock discrimination may provide 

 more accurate estimates of stock composition in 

 mixed stocks (e.g. Canadian and American stocks). 

 Compared with some other methods of stock iden- 

 tification that are based upon biological tags, genetic- 



based methods of stock identication generally have 

 the advantage of temporal stability of the charac- 

 ters used. For example, variation at protein-coding 

 loci has usually been found to be temporally stable 

 (Beacham et al., 1985, 1987; Kondzela et al., 1994; 

 Wilmot et al., 1994), and similar results have been 

 reported for minisatellite loci (Taylor et al., 1994; 

 Beacham et al., 1996). The two-year temporal sta- 

 bility of allelic frequencies at the Ssa-A33 and Ssa- 

 A34 loci observed in the current study suggests that 

 annual sampling of contributing baseline stocks 

 would not be required in order to estimate stock com- 

 position of appropriate mixtures. Because the popu- 

 lations sampled in this study tended to be major ones 

 (escapement of British Columbia populations is usu- 

 ally >10,000 fish), the ratio of sample size to effective 

 population size was less than 0.1, and the probability 

 of a significant test to detect temporal variation is simi- 

 lar to the presumed level of significance ( Waples, 1989). 



