252 
Fishery Bulletin 106(3) 
Table 6 
Estimated contribution (%) of individual populations and regional estimates of chum chum salmon ( Oncorhynchus keta) in Russia 
in simulated mixed-population samples. Each mixture of 150 fish was generated 100 times from baseline allele frequencies, and 
stock compositions of the mixt ures were estimated with parametric resampling of each of the 34 baseline populations to obtain a 
new distribution of allele frequencies on each iteration. Actual= 100% accurate. Standard deviations are shown in parentheses. 
Fourteen microsatellites were used to estimate stock compositions of the simulated mixtures. 
Populations 
Actual % 
Estimated % 
Populations 
Actual % 
Estimated % 
Mixture 1 
Mixture 3 
Amur 
30 
30.5 (4.0) 
Plotnikova 
20 
14.0(3.3) 
Avakumovka 
10 
8.7 (2.0) 
Kol 
10 
6.7 (2.5) 
Ryazanovka 
10 
9.1 (2.4) 
Utka 
10 
5.5 (1.9) 
Naiba 
20 
21.0(3.3) 
Kamchatka 
10 
8.2 (2.4) 
Udarnitsa 
10 
7.6 (2.2) 
Ossora 
10 
10.6(2.8) 
Tauy 
10 
7.1 (2.5) 
Olutorsky Bay 
10 
6.0 (2.6) 
Ola 
10 
8.9 (2.5) 
Anadyr 
20 
17.9(3.4) 
Regions 
Kanchalan 
10 
12.3 (3.1) 
Amur 
30 
30.5(4.0) 
Regions 
Primorye 
20 
17.9 (3.2) 
West Kamchatka 
40 
40.8(4.7) 
Sakhalin Island 
30 
28.7 (3.8) 
East Kamchatka 
30 
25.4(3.9) 
Magadan 
20 
16.5(3.6) 
Northeast Russia 
30 
30.2(4.1) 
West Kamchatka 
0 
4.0 (1.9) 
Magadan 
0 
1.4 (1.2) 
East Kamchatka 
0 
1.6 (1.2) 
Northern Sea of Okhotsk 
0 
1.1 (0.8) 
Mixture 2 
Mixture 4 
Naiba 
20 
19.0(3.2) 
Avakumovka 
10 
8.2 (2.3) 
Magadan 
10 
7.8 (2.6) 
Naiba 
10 
10.1 (2.3) 
Ola 
10 
9.7 (2.8) 
Tym 
10 
8.1 (2.5) 
Okhota 
10 
9.0 (2.5) 
Ola 
15 
13.0(2.8) 
Oklan 
10 
7.8 (2.5) 
Okhota 
10 
8.7 (2.6) 
Bolshaya 
20 
16.5(3.3) 
Hairusova 
15 
13.7(3.1) 
Hairusova 
10 
10.0(3.0) 
Utka 
10 
5.5 (2.2) 
Kikchik 
10 
7.5 (2.6) 
Kamchatka 
10 
8.1 (2.5) 
Regions 
Kanchalan 
10 
11.6(2.6) 
Sakhalin Island 
20 
19.2(3.3) 
Regions 
Magadan 
30 
27.1 (4.0) 
Primorye 
10 
8.312.3) 
Northern Sea of Okhotsk 
10 
8.0 (2.6) 
Sakhalin Island 
20 
18.3(3.1) 
West Kamchatka 
40 
42.1 (4.7) 
Magadan 
25 
22.7 (4.0) 
East Kamchatka 
0 
2.9 (1.6) 
West Kamchatka 
25 
26.6(3.8) 
East Kamchatka 
10 
10.9(2.9) 
Northeast Russia 
10 
12.2(2.7) 
and thus the estimation of allele frequencies would be 
subject to sampling error, the clustering of these popula- 
tions was well supported by our bootstrap calculations 
(100%). Regional clustering of samples or populations 
is typically observed in chum salmon (Beacham et al., 
1987; Winans et ah, 1994), and thus it is unlikely that 
close genetic relationships among these populations from 
Primorye were inferred incorrectly. 
If populations spawn in remote areas, opportunities 
to collect samples may be limited. In our study, all sam- 
ples that were available for a specific sampling site or 
population were combined in order to estimate genetic 
differentiation among populations. Annual variation 
in allele frequencies within a population is typically 
less than the geographic and population differences ob- 
served; therefore pooling annual samples over time is a 
reasonable approach to estimate population-level allele 
frequencies. Relative annual stability of microsatellite 
allele frequencies is a general feature of microsatel- 
lite loci in salmonids (Tessier and Bernatchez, 1999; 
Beacham et ah, 2006). 
The population structure of chum salmon in Russia 
has been investigated previously. For example, Winans 
et al. (1994) after examining 35 allozyme loci, indicated 
that there were four groups of Russian chum salmon 
populations, and that one group generally comprised 
populations from west Kamchatka, a second group com- 
prised populations from Magadan, the Sea of Okhotsk, 
and east Kamchatka, the third group was solely from 
east Kamchatka, and a fourth group comprised the 
