774 



Fishery Bulletin 90(4). 1992 



complete details of the individual studies. (Bartley et 

 al. 1992, Waples et al. 1991). 



Variability within populations 



The levels of genetic variation within populations were 

 evaluated using only the loci found to be polymorphic. 

 Because this restriction does not represent a random 

 sample of gene loci, values reported here are applicable 

 only for comparisons among populations included in 

 this study or with other studies using the same set of 

 loci. Indices of genetic variability were consistently 

 slightly higher in the Snake River samples; the average 

 number of alleles per locus was 1.63 vs. 1.51 for the 

 Klamath River, and the average heterozygosity was 

 0.079 vs. 0.065 (0.05>p>0.01 in both instances, based 

 on Mann- Whitney tests). Heterozygosities ranged from 

 0.058 to 0.090 in the Snake River populations and were 

 less uniform in the Klamath River groups, where both 

 the lowest (0.039 in Shasta River) and the highest 

 (0.126 in Omagar Creek) values were found. Neither 

 of these latter two populations were represented in the 

 initial study of the Klamath River by Utter et al. (1989). 

 The actual heterozygosity values reported here are 

 higher than those reported by Utter et al. (1989), 

 primarily because a number of new, very polymorphic 

 systems are included in the more recent analyses. 

 Nevertheless, Utter et al. (1989) also found a slightly 

 higher average heterozygosity in Snake River spring- 

 run and summer-nm chinook salmon (0.035-0.045) than 

 in those from the Klamath River (0.027-0.032). Based 

 on the new data, Waples et al. (1991) concluded that, 

 in comparison with other Columbia River populations. 

 Snake River spring-run and summer-run chinook 

 salmon have somewhat reduced levels of genetic vari- 

 ability, but that the difference is apparently not as large 

 as suggested by earlier studies (Utter et al. 1989, 

 Winans 1989). 



Vanability between regions 



Allele frequency distributions differed substantially 

 between the two regions at a number of gene loci. 

 Although three or more alleles were found at some of 

 these loci, most of the important differences were 

 reflected in differing frequencies of the common allele 

 (Table 3). Particularly large differences were found at 

 mMDH-2* and sMEP-1* (Fig. 2); for these loci, the 

 range of allele frequencies was nonoverlapping be- 

 tween regions, with substantially higher frequencies 

 of the common (i.e., iOO*) allele found in the Klamath 

 River samples at both loci. 



Genetic differences between the two regions based 

 on data for all 30 loci are summarized in a phenogram 

 resulting from clustering of pairwise genetic distances 



(Fig. 3). The Snake and Klamath River populations are 

 separated by a mean genetic distance of 0.014, whereas 

 the within-river separations average 0.004 and 0.007, 



