Sme et al.: Identification of Eleginus gracilis by means of microsatellite markers 
63 
Table 2 
Microsatellite properties of northern gadid species of the Pacific Rim and Arctic Ocean for the 9 loci designed for 
saffron cod ( Eleginus gracilis) sampled in the Chukchi Sea and Gulf of Alaska (GOA) in 2011 and 2013. Number 
of samples for each species In), the numbers of different allele observed (« a ), the range of allele sizes, the mean 
and standard error of the mean (SE) of allele sizes, expected heterozygosities ( H e ), and inbreeding coefficients (F is ) 
are given. An entry of dna means the locus did not reliably amplify. Collections were made in 2013 for navaga (E. 
nawaga), during 1997-1999 for Pacific tomcod ( Microgadus proximus ), in 2013 for Pacific cod ( Gadus macrocepha- 
lus), in 2015 for walleye pollock ( G. chalcogrammus), and in 2012 for Arctic cod (Boreogadus saida). 
Locus and species 
n 
range 
mean 
H e 
F is 
Elgr07 
Chukchi Sea E. gracilis 
30 
10 
127-175 
155.7 (1.2) 
0.867 
-0.016 
GOA E. gracilis 
41 
7 
151-179 
160.6 (0.5) 
0.683 
-0.054 
E. nawaga 
81 
14 
115-183 
133.7 (0.7) 
0.815 
-0.028 
M. proximus 
22 
1 
123 
123.0 (0.0) 
0.000 
- 
G. macrocephalus 
14 
2 
115 and 131 
128.7 (1.1) 
0.286 
-0.130 
G. chalcogrammus 
6 
2 
131 and 135 
133.3 (0.6) 
0.833 
-0.667 
B. saida 
53 
dna 
- 
- 
- 
- 
Elgrll 
Chukchi Sea E. gracilis 
30 
12 
208-272 
222.1 (1.7) 
0.833 
0.043 
GOA E. gracilis 
41 
8 
204-260 
214.0 (1.3) 
0.634 
-0.117 
E. nawaga 
81 
21 
240-336 
274.7 (1.4) 
0.877 
0.043 
M. proximus 
22 
17 
248-340 
285.8 (3.0) 
0.727 
0.230 a 
G. macrocephalus 
14 
18 
192-204 
202.9 (0.6) 
0.286 
-0.072 
G. chalcogrammus 
6 
dna 
- 
- 
- 
- 
B. saida 
53 
dna 
- 
- 
- 
- 
Elgr31 
Chukchi Sea E. gracilis 
30 
6 
191-211 
197.1 (0.7) 
0.833 
-0.103 
GOA E. gracilis 
41 
4 
191-203 
194.8 (0.5) 
0.659 
-0.015 
E. nawaga 
81 
11 
179-231 
204.4 (0.8) 
0.864 
-0.052 
M. proximus 
22 
14 
215-267 
240.5 (2) 
0.955 
-0.027 
G. macrocephalus 
14 
18 
223-299 
263.3 (3.7) 
1.000 
-0.034 
G. chalcogrammus 
6 
10 
215-267 
241.7 (4.8) 
1.000 
-0.035 
B. saida 
53 
37 
223-543 
355.6 (7.6) 
0.962 
0.005 
Elgr38 
Chukchi Sea E. gracilis 
30 
9 
112-144 
127.5 (1.1) 
0.867 
-0.026 
GOA E. gracilis 
41 
dna 
- 
- 
- 
- 
E. nawaga 
81 
dna 
- 
- 
- 
- 
M. proximus 
22 
6 
120-140 
127.9 (0.8) 
0.727 
0.068 
G. macrocephalus 
14 
6 
128-160 
141.9 (1.9) 
0.786 
0.037 
G. chalcogrammus 
6 
7 
236-276 
258.0 (4.4) 
0.833 
0.039 
B. saida 
53 
37 
252-448 
348.6 (6.3) 
0.566 
0.422 c 
Table continued 
(Hedrick, 2005), based on ratios of heterozygosities ad¬ 
justed to account for the amount of genetic variation 
observed at each locus, was estimated with the soft¬ 
ware program SMOGD, vers. 1.2.5 (Crawford, 2010). 
Estimates of chord distances (Cavalli-Sforza and Ed¬ 
wards, 1967), a geometric measure, were made with 
PHYLIP, vers. 3.6 (Felsenstein, 2005). 
Principal component analysis was used to contrast 
the genetic compositions of species groups (SYTAT, 
vers. 13 software; SYSTAT Software, Inc., San Jose, 
CA). Correlation matrix-based PCA standardizes vari¬ 
ables so that each variable has a similar scale; it was 
used to contrast the allelic compositions. Covariance 
matrix-based PCA applies the observed variances so 
that the scale of variation is included in the analysis; 
it was used to contrast allele-frequency profiles. Loci 
missing from a collection or a species did not contrib¬ 
ute to the PCA score. 
Assignment tests (GeneClass2; Piry et al., 2004) 
were used to evaluate the robustness of the differences 
among species groups. The tests removed each indi¬ 
vidual fish from the species groups before assignment. 
The criterion of Rannala and Mountain (1997) was ap¬ 
plied in all tests. 
Results 
Only genotypes from loci that could be reliably inter¬ 
preted were analyzed in for each species. Nine loci am- 
