Beacham et al.: Population Identification of Oncorhynchus Ishawytscha by variation at microsatellite loci 



249 



Table 4 



Estimated percentage composition of three simulated mixtures of Eraser River chinook salmon incorporating variation at 13 micro- 

 satellite loci and estimated with a 50-population baseline. Each mixture of 150 fish was generated 500 times with replacement, and 

 stock compositions of the mixtures were estimated by resampling each baseline population with replacement to obtain a new distri- 

 bution of allele frequencies. "Regional sum" is the regional sum of all populations in the region. The appearance of each population in 

 each of the simulated mixtures is indicated in parentheses after the population name. Standard deviations are given in parentheses. 



Population and sum for region 



Birkenhead (spring) 

 Harrison (fall) 

 Chilliwack(fall)' 



Sum for Lower Eraser region 

 Westroad (spring) 

 Bridge (summer) 

 Cottonwood (spring) 

 Elkin (spring) 

 L.Chilcotin (spring) 

 Quesnel (summer, fall) 

 Stuart-Nechako (summer, fall) 



Sum for Mid Eraser region 

 Slim (spring) 

 Swift (summer) 

 Bowron (spring) 

 Willow (spring) 

 Holmes (spring, summer) 

 MacGregor (spring) 



Sum for Upper Eraser region 

 Clearwater (summer) 

 Finn (summer) 

 Louis 



Mahood (fall) 

 Raft (spring, summer) 



Sum for North Thompson region 

 Eagle (summer, fall) 

 L Shuswap (spring, summer) 

 M Shuswap (spring, summer) 

 South Thompson (summer, fall) 



Sum for South Thompson region 



Sum for Lower Thompson region 



' White-fleshed population. 



estimation of regional stock compositions should be avail- 

 able when the genetic data outlined in our study is applied 

 to actual mixed-fishery samples. 



Identification of individuals 



Individuals were classified with respect to origin for 50 

 populations in the Fraser River drainage. Success rate of 

 classification of individuals varied considerably among 

 populations and to some extent was reflective of sample 

 size of individual populations. For example, success rate of 

 identication of individual Goat Creek and Horsey River chi- 

 nook salmon was about 5%, but approximately only 20 fish 



from each population were included in the analysis used 

 to characterize the populations (Table 5). Success rate of 

 classification of other upper Fraser River populations was 

 higher, but more fish were available to quantify the varia- 

 tion in the populations. Success rate was partially attrib- 

 utable to sample size but was also markedly influenced by 

 genetic differentiation in the population. For example, all 

 individual Birkenhead River fish were correctly assigned 

 to the population of origin in a 50-population baseline — 

 indicative of the genetic distinctiveness of this population. 

 Overall, highest rates of classification to individual popu- 

 lations were observed for lower Fraser River and lower 

 and North Thompson River populations, with an average 



