534 
Fishery Bulletin 99(4) 
Table 4 
Observed numbers of each mtDNA haplotype, haplotype diversity, and nucleotide diversity (substitutions per 1000 bp) within col- 
lections of North Pacific coho salmon screened for variable sites detected in a preliminary survey (Table 3)(Nei and Tajima, 1983; 
Nei, 1987). Standard errors are in parentheses. Homogeneity of haplotype frequencies ( P MC < 10 -4 ) was tested by using Monte-Carlo 
simulation based on 10,000 resampling iterations to estimate probability (Roff and Bentzen, 1989). 
Collection 
n 
Haplotype 
Haplotype 
diversity 
Nucleotide 
diversity 
A 
B 
C 
D 
E 
F 
G 
H 
I 
J 
K 
L 
M 
N 
0 
p 
Hugh Smith River 
20 
6 
0 
12 
0 
2 
0 
0 
0 
0 
0 
0 
0 
0 
0 
0 
0 
0.57 
3.64 
Fish Creek (Taku River) 
20 
11 
0 
3 
2 
0 
2 
0 
0 
0 
1 
0 
1 
0 
0 
0 
0 
0.68 
5.68 
Berners River 
20 
5 
0 
11 
0 
0 
0 
0 
0 
1 
2 
0 
0 
0 
0 
0 
1 
0.64 
2.40 
Indian River 
20 
19 
0 
1 
0 
0 
0 
0 
0 
0 
0 
0 
0 
0 
0 
0 
0 
0.10 
0.22 
Ford Arm River 
20 
10 
6 
1 
0 
2 
0 
1 
0 
0 
0 
0 
0 
0 
0 
0 
0 
0.68 
4.51 
Crooked Creek 
20 
9 
0 
10 
0 
1 
0 
0 
0 
0 
0 
0 
0 
0 
0 
0 
0 
0.57 
2.50 
Little Susitna River 
20 
4 
0 
15 
0 
0 
0 
0 
0 
1 
0 
0 
0 
0 
0 
0 
0 
0.42 
1.13 
Buskin River 
20 
5 
0 
12 
0 
3 
0 
0 
0 
0 
0 
0 
0 
0 
0 
0 
0 
0.58 
4.69 
Karluk River 
20 
1 
0 
9 
0 
5 
0 
0 
0 
0 
0 
1 
0 
0 
4 
0 
0 
0.73 
9.07 
Eek River 
20 
1 
0 
7 
0 
9 
0 
0 
3 
0 
0 
0 
0 
0 
0 
0 
0 
0.68 
8.38 
Kanektok River 
20 
5 
0 
8 
0 
5 
0 
0 
1 
0 
0 
0 
0 
1 
0 
0 
0 
0.75 
7.91 
Delta Clearwater River 
21 
0 
0 
0 
0 
0 
0 
0 
20 
0 
0 
0 
0 
0 
0 
1 
0 
0.1 
0.2 
Kamchatka River 
17 
9 
0 
7 
0 
1 
0 
0 
0 
0 
0 
0 
0 
0 
0 
0 
0 
0.58 
2.7 
Total 
258 
85 
6 
96 
2 
28 
2 
1 
24 
2 
3 
1 
1 
1 
4 
1 
1 
0.734 
6.73 
(±0.016) 
(±0.00) 
Average 
0.54 
4.08 
(±0.06) 
(±0.83) 
lections spanned the southern portion of the geographic 
range from southern Southeast Alaska to the Kamchatka 
Peninsula. Within that set of collections, the two coastal 
Southeast Alaskan collections (Ford Arm and Indian Riv- 
er) appeared to form a weak cluster, but there was no obvi- 
ous structure among the remaining collections. 
We conducted AMOVA analyses reflecting a geograph- 
ical hierarchy: Southeast Alaska, Southcentral Alaska, 
western and interior Alaska (Bering Sea), and Asia to 
examine the geographic basis of variation. Although the 
Karluk River collection resembled Bering Sea collections 
more than other northern Gulf of Alaska collections, we in- 
cluded it with the Southcentral Alaska group to maintain 
the geographic basis of the analysis. Analyzing the data 
based on haplotype frequencies (analogous to allelic dif- 
ferences in analysis of variance described by Weir [1996]) 
revealed highly significant divergence among collections 
((P ST =0.291, P MC <0.0001), most of which is attributable 
to average divergence among drainages within a region 
( d> gc =0.227, P MC <0.0001 ), rather than differences between 
regions (<f> rr =0.083, P MC =0.094). Incorporating relation- 
ships between the haplotypes into the analysis increased 
the proportion of the total divergence observed among 
drainages ( & ST = 0.449, P MC <0.0001) and among drainages 
within regions (d> sc =0.273 , P MC <0.0001). The estimate of 
the proportion of divergence among regions also increased 
(0 ct =O.‘242, P mc = 0.083). Estimates of long-term gene flow 
[AI e( pm=(d>Yy _1_ 1^2] from <P SC and @ CT based on haplotype 
relationships were about 1 female per generation between 
collections within regions and about two females per gen- 
eration between regions. Such estimates assume that an 
equilibrium between gene flow and random drift exists. 
The nested clade analysis (Fig. 3, Table 5) collapses the 
“gene tree” from the periphery and analyzes each subclade 
for significantly small or large geographical distributions 
of the components as compared with the subclade as a 
whole. The first two levels of nesting are 1-step clades (nest- 
ing 0-step haplotypes) and 2-step clades (nesting 1-step 
clades). The significance in their geographic distributions 
were consistent with restricted gene flow with isolation by 
distance for haplotypes of the A-D (plus I, J, and P) sub- 
clade and past fragmentation for haplotypes of the E-H 
(less G plus M, N, and O) subclade. The 3-step clade (nested 
2-step clades) that includes all of the E-H (plus L, M, N, 
and O) haplotypes also indicates past fragmentation. The 
most interior level of nesting, which contrasts the A group 
of haplotypes with the E group, assuming that the A group 
is interior (ancestral) based on the interpretations of Cas- 
telloe and Templeton (1994), is consistent with contiguous 
range expansion. We reanalyzed the data without the Del- 
ta Clearwater population, the Kamchatka population, and 
without either. These two geographically distinct popula- 
tions did not alter the overall interpretation, but the Delta 
Clearwater population was responsible for significance of 
clade 2-4 and the Kamchatka population was responsible 
for the significance of clade 2-2. 
