Full text of DOI:10.1577/1548-8659(2000)129<0194:GPSAOO>2.0.CO;2

This article was downloaded by: [University of Exeter]
On: 20 July 2015, At: 13:15
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: 5 Howick Place,
London, SW1P 1WG
Transactions of the American Fisheries Society
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/utaf20
Genetic Population Structure and Origin of Life History
Types in Chinook Salmon in British Columbia, Canada
David J. Teel a , George B. Milner a , Gary A. Winans a & W. Stewart Grant a
a Conservation Biology Division, Northwest Fisheries Science Center , 2725 Montlake
Boulevard East, Seattle, Washington, 98112-2097, USA
Published online: 09 Jan 2011.
To cite this article: David J. Teel , George B. Milner , Gary A. Winans & W. Stewart Grant (2000) Genetic Population Structure
and Origin of Life History Types in Chinook Salmon in British Columbia, Canada, Transactions of the American Fisheries
Society, 129:1, 194-209, DOI: 10.1577/1548-8659(2000)129<0194:GPSAOO>2.0.CO;2
To link to this article: http://dx.doi.org/10.1577/1548-8659(2000)129<0194:GPSAOO>2.0.CO;2
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions

Transactions of the American Fisheries Society 129:194–209, 2000
American Fisheries Society 2000
Genetic Population Structure and Origin of Life History Types
in Chinook Salmon in British Columbia, Canada
D
AVID J. TEEL,* GEORGE B. MILNER, GARY A. WINANS
W. STEWART RANT
Conservation Biology Division, Northwest Fisheries Science Center,
2725 Montlake Boulevard East, Seattle, Washington 98112-2097, USA
, AND
G
Abstract.—We used protein electrophoresis to examine genetic population structure and origin
of life history types of chinook salmon Oncorhynchus tshawytscha in British Columbia, Canada.
Among 31 allozyme loci resolved in 91 samples from 63 populations of chinook salmon in rivers
and hatcheries throughout British Columbia, population heterozygosities averaged 0.084 (range
0.048–0.108) and were typical of values for populations in other regions. A hierarchical gene
diversity analysis indicated that 91.3% of the total allele-frequency diversity was attributable to
within-population variability; the remaining 8.7% was attributable to geographic variability among
populations, which was partitioned into among-river (3.3%), among-area (3.5%), and among-region
(1.9%) components. Two major groups of populations appeared in the principal components anal-
ysis and in cluster analysis of genetic distances. A coastal group included populations in four
subgroups: Central coast, Georgia Strait, lower Fraser River, and west Vancouver Island. An inland
group included six subgroups: Nass River, Skeena River, north Thompson River, upper and mid-
Fraser River, south Thompson River, and lower Thompson River. The geographic extents of the
inland and coastal groups largely coincided with the geographic distributions of stream- and ocean-
type juvenile forms and may reflect postglacial colonization by two ancestral lineages that survived
in Pleistocene refugia. The presence of genetically undifferentiated stream-type fish in coastal
streams populated by ocean-type fish may reflect either postglacial life history differentiation from
ancestral ocean-type fish or life history flexibility of ocean-type fish.
Genetic studies of chinook salmon Oncorhyn- ment, analyses of these variables may not provide
chus tshawytscha in British Columbia have been
of particular interest to evolutionary biologists, be-
cause the diversity of life history patterns in this
region provides an opportunity for understanding
the origin of within-species variability. Various
life history forms of chinook salmon spawn in
about 260 rivers that are scattered throughout the
province at locations ranging from just above the
tidal limit to nearly 1,000 km inland (Aro and
Shepard 1967). These forms are characterized by
differences in the timing of juvenile outmigration
from freshwater, in ocean distribution of maturing
fish, in the timing of adult return to freshwater, in
spawning timing, and in the geographic distribu-
tions of spawning populations (Fraser et al. 1982;
Healey 1991). One of these traits, adult return tim-
ing, has been used to classify populations as man-
agement units, particularly in Washington, Ore-
gon, Idaho, and California (e.g., WDF et al. 1993)
and to interpret the results of genetic studies (e.g.,
Utter et al. 1989; Nielsen et al. 1994; Banks et al.
1996). Although adult run timing and other vari-
ables may provide useful categories for manage-
sufficient information to understand the evolution
of major lineages within chinook salmon.
Some authors have postulated that the length of
juvenile freshwater residence, along with ocean
distribution, and the geography of spawning pop-
ulations delineate two evolutionary lineages or
races along the west coast of North America, lin-
eages that arose as a consequence of geographic
isolation during Pleistocene glaciation (e.g., Heal-
ey 1991). An ocean-type lineage, consisting of fish
that migrate to estuaries in their first year, contrasts
with a stream-type lineage of fish that resides in
freshwater for at least 1 year before migrating to
marine waters. An extended estuarine residence is
associated with ocean-type juveniles, and long-
distance ocean migrations are associated with
stream-type juveniles. The delineation between
life history types, however, is not always clear.
Individuals in some populations, particularly those
in coastal areas, can show ocean- or stream-type
life history traits (Healey 1991), which suggests
mixing between evolutionary lineages, or pheno-
typic plasticity. Inheritance studies have shown
that in at least some populations, early and late
juvenile migration timings and other traits asso-
ciated with ocean- and stream-type fish have a ge-
* Corresponding author: david.teel@noaa.gov
Received December 22, 1997; accepted April 8, 1999
194

CHINOOK SALMON POPULATION STRUCTURE
195
netic as well as an environmental basis (Taylor nets or seines and were collected from several sites
1988; Clarke et al. 1992, 1994). along a stream in order to avoid taking offspring
Life history variability within a species may re- from only a few matings. Wild adults were sam-
flect adaptive convergence in similar habitats in pled on or near spawning grounds, typically over
different drainages. Taylor (1988, 1990b) has sug- a period of 1 or 2 d, near the peak of the spawning
gested that much of the life history variability in run. Many samples were collected in conjunction
chinook salmon has resulted from adaptation to with government or community enhancement proj-
local environments. In a study of genetic vari- ects, in which wild adult fish were captured on or
ability in Alaskan and southern populations of chi- near a spawning ground, transported to a hatchery,
nook salmon, Gharrett et al. (1987) suggested that and spawned artificially. Progeny from these mat-
divergence in the timing of juvenile outmigration ings were reared until their release near the site of
appeared to reflect natural selection after postgla- the adult capture. Many of our samples from these
cial colonization, at least for populations in south- projects were juvenile fish. Most of the broods we
eastern Alaska. However, in a study of chinook sampled were the progeny of 100 to 300 adult fish.
salmon in the Columbia River Basin, Waples et al. Several samples were taken from broods that were
(1991) found that genetic relationships were con- the progeny of 60 to 100 spawners (Table 1). In a
cordant with juvenile migration timing and that few cases, in which efforts to supplement natural
significant genetic subdivision occurred within spawning have continued for several years, the
these two major life history groups.
captured adults may have included both naturally
In the present study, we surveyed allozyme vari- reared and hatchery-reared fish. Specimens were
ation in 63 river and hatchery populations of chi- collected and stored as described by Aebersold et
nook salmon in British Columbia from 1981 to al. (1987), and specimens were shipped to the
1991. These genetic data have been used in pre- Northwest Fisheries Science Center for processing
vious studies to estimate the contributions of ge- at the Manchester Marine Experimental Station in
netically distinct population groups to numerous Manchester, Washington.
commercial and recreational marine fisheries (e.g.,
Electrophoresis.—Sample preparation and hor-
Milner et al. 1985; Utter et al. 1987b; Shaklee et izontal starch gel protein electrophoresis followed
al. 1990b) and to study the migratory timing of the method of Aebersold et al. (1987). The gel
adult chinook salmon caught in research fisheries buffers and tissues used to resolve proteins are
in the lower Fraser (Waples et al. 1990) and lower given in Waples et al. (1993). The guidelines out-
Skeena rivers (G. A. Winans and D. J. Teel, North- lined by Utter et al. (1987a) were used to infer
west Fisheries Science Center, unpublished data). genotypes from banding patterns on electropho-
In the analysis presented here, we estimated the retic gels. Enzyme systems and locus designations
distribution of genetic diversity among popula- for polymorphic loci are presented in Table 2. No-
tions at three hierarchical levels: rivers, areas, and menclature for locus names and alleles followed
regions. We then inferred genetic relationships Shaklee et al. (1990a).
among populations of chinook salmon using clus-
Data analysis.—Genotypic frequencies of poly-
ter analyses, and we used these results to search morphic loci for each sample were examined for
for patterns of divergence that might be concordant departures from expected Hardy-Weinberg pro-
with the distributions of ocean and stream life his- portions using chi-square goodness-of-fit tests,
tory types.
with Levene’s (1949) correction for small sample
sizes. A locus was considered polymorphic if more
than one allele was observed in at least one sample.
Methods
Samples.—A total of 91 samples were collected When a locus had more than two alleles, so that
from 63 British Columbia rivers and hatcheries the expected frequency of occurrence of rare ge-
from 1981 to 1991. Forty localities were sampled notypes was less than 1.0, we pooled less frequent
in a single year, 18 locations over 2 years, and 5 alleles and calculated frequencies for three classes
locations over 3 years (Figure 1; Table 1). Spawn- of genotypes: homozygotes for the most frequent
ing adults were sampled at hatcheries, and juve- allele, heterozygotes for the most frequent allele
niles were collected from rearing ponds. Samples and one other allele, and all other genotypes. This
of hatchery broods included the progeny of at least pooling reduced the number of tests in which the
100 adult fish, except at Birkenhead Hatchery, test statistic might not follow the chi-square dis-
where about 60 fish had been spawned. Wild ju- tribution (Sokal and Rohlf 1981). Hardy-Weinberg
venile fish were sampled in natal rivers with fyke tests were performed on isoloci (comigrating pro-

196
TEEL ET AL.
F
IGURE 1.—Locations of 63 chinook salmon populations in British Columbia. Numbers correspond to population
names in Table 1.
tein products of duplicated loci) using the methods large to be attributable to sampling error and ge-
of Waples (1988).
netic drift.’’
We estimated allelic frequencies from genotypic
We used five methods to examine levels and pat-
frequencies. Allelic frequencies for isoloci were terns of genetic variation within and between 63
calculated as mean frequencies over both loci and populations using 31 loci (sAAT-1,2 ; sAAT-3
were treated as a single tetrasomic locus in our sAAT-4 ; ADA-1 ; ADA-2 ; sAH ; FDHG ; GPI-
analyses. In this form, the frequencies of isoloci ; GPI-B1 ; GR ; mIDHP-1 ; mIDHP-2 ; s-
(sAAT-1,2 and sMDH-B1,2 ) can be compared to IDHP-1 ; sIDHP-2 ; LDH-B1 ; LDH-B2 ; LDH-
those reported by Utter et al. (1989) for chinook ; sMDH-A1,2 ; sMDH-B1,2 ; MPI ; PEPA
salmon populations that ranged from California PEPB-1 ; PEPD-2 ; PEPLT ; PGDH ; PGK-2
through British Columbia. Allelic frequencies of sSOD-1 ; TPI-1 ; TPI-2 ; TPI-3 ; and TPI-4
*
*;
*
*
*
*
*
A*
*
*
*
*
*
*
*
*
*
*
C*
*
*
*
*
*
*
;
;
*
*
*
*
*
*
*
*
).
samples taken in different years from the same First, average heterozygosity per locus (isoloci ex-
location were combined. Waples (1990) recom- cluded) for each population was calculated using
mended combining temporal samples to estimate an unbiased estimator of expected heterozygosity
population allelic frequencies ‘‘unless there is rea- (Nei 1978). Second, the proportion of P0.95 loci was
son to believe that temporal differences are too computed for each population. A locus was con-

CHINOOK SALMON POPULATION STRUCTURE
197
sidered a P0.95 locus if the frequency of the most
The percentage of P0.95 loci and average hetero-
common allele was less than or equal to 0.95. Third, zygosity per locus for each population over 31 loci
gene diversity analysis (Chakraborty et al. 1982) are presented in Table 1. The percentage of P0.95
was used to examine genetic variation at three geo- loci ranged from 12.9% in Puntledge River to
graphic levels: populations (a putative breeding 35.5% in Kitimat River, with a mean for all pop-
population in a river or hatchery) within areas ulations of 25.0%. Mean percentage of P0.95 loci
(GSA); areas (groups of nearby populations) within for areas ranged from 20.6% in the south Thomp-
regions (GAR); and regions (coastal versus inland son River to 31.2% in the Central coast. Average
populations) (GRT) (Table 1). Fourth, a principal heterozygosity ranged from 0.048 in Birkenhead
components analysis (PCA) was applied to a cor- River to 0.108 in Harrison River, with a mean for
relation matrix based on alleles occurring at fre- all populations of 0.084. The average heterozy-
quencies of greater than 0.05 in one or more pop- gosity for areas ranged from 0.058 in the Nass
ulations. The number of alleles (k) analyzed at a River to 0.094 in the lower Thompson River. The
locus was k minus 1; therefore, we arbitrarily gene diversity analysis of the 63 samples resulted
dropped the most common allele at each locus and in a total gene diversity of 0.085 (HT) and an av-
used only less frequent alleles, for a total of 26 erage sample diversity of 0.077 (HS), indicating
alleles from 21 loci. Fifth, chord distances (Cavalli- that 91.3% of the total genetic diversity was at-
Sforza and Edwards 1967) were computed between tributable to within-sample variability and that
all pairs of populations, and relationships among 8.7% was attributable to variability among sam-
populations were depicted in a dendrogram, using ples (Table 3). About 3.3% of the total gene di-
the unweighted pair-group method with arithmetic versity was due to variability among populations
averaging algorithm (Sneath and Sokal 1963). An within areas, 3.5% of the total was due to vari-
a priori grouping of populations by area, based on ability among areas within regions, and 1.9% was
geography (12 areas total; Table 1), was used to due to variability between populations in inland
calculate a mean between-population within-area and coastal regions.
distance for each area. Chord distances were cal-
culated with the BIOSYS program (Swofford and of populations. For example, sAAT-3
Some alleles were confined to particular groups
90 occurred
*
Selander 1981), and the cluster and PCA analyses in all populations from Vancouver Island, Georgia
were made using the NTSYS program (Rohlf 1993). Strait, and the Central coast, but it was not found
in Fraser or Nass river populations (Figure 2).
sAAT-4*63 occurred at high frequencies in the Fra-
Results
Of the 31 loci screened, 25 were polymorphic, ser and Skeena river populations, with frequencies
and 21 of these were at the P0.95 level of poly- greater than 0.40, but it was absent in populations
morphism in at least one population (Table 2). Al- on west Vancouver Island and was found in only one
lelic frequencies for polymorphic loci are available Georgia Strait population (Big Qualicum Hatch-
from FishBase (www.fishbase.org). Eleven of the ery).
25 polymorphic loci had three alleles per locus,
and 14 had two alleles per locus. Three loci (sAH
MPI , and PGK-2
tions examined. Six loci (GPI-B1
mIDHP-2 , sMDH-A1,2 , PGDH
were monomorphic in all populations.
We tested observed genotypic proportions at and GR
polymorphic loci (up to 25) in 91 samples for de- ings for PEPA
partures from expected Hardy-Weinberg propor- A scatterplot of scores for PC1 and PC2 (Figure
tions. Significant (P 0.05) deviations were de- 3) showed a clear geographic pattern. Populations
The first two eigenvectors of the PCA explained
*
,
about 31.6% of the total variation (Table 4; Figure
*
*
) were variable in all popula- 3). Large negative loadings on PC1 for sAAT-4
, mIDHP-1 and GR 110 contrasted with large positive load-
, and TPI-2 ings for sAAT-3 90 and PEPB-1 130 on PC1.
Large negative loadings on PC2 for PGK-2 100
85 contrasted with the large positive load-
90 on PC2.
*63
*
*
,
*
*
*
*
*
)
*
*
*
*
*
Ͻ
tected in 51 of 1,267 tests (4.0%). This number of from the Central coast, west Vancouver Island,
significant tests is close to the number expected Georgia Strait, and the lower Fraser River had pos-
by chance for this rejection level. There were no itive PC1 scores, whereas inland populations from
consistent trends in the distribution of significant the Nass and Skeena rivers and from the Fraser
tests by sample, sample type (wild versus hatchery, River above the Fraser Canyon had negative PC1
juvenile versus adult), or locus. Therefore, we ac- scores. PC2 separated populations in north and
cepted the null hypothesis that the samples were central British Columbia from coastal populations
drawn from disscrete panmictic populations.
to the south. Populations in the Nass and Skeena

198
TEEL ET AL.
TABLE 1.—Sample information and indices of genetic variability for chinook salmon in British Columbia. Map codes
refer to Figure 1. Sources are hatchery (H) and wild (W). Life-history type is stream type (S), ocean type (O), and
intermediate (I). Age is adult fish (A) and juvenile fish (J). Brood year of juvenile fish is 1 year prior to year sampled.
Indices of variability are based on data for 31 loci except for heterozygosity, which excluded 3 isoloci. Indices of
genetic variability are %P
ϭ
percentage of P
loci, H
ϭ
heterozygosity, and D
ϭ
average chord distance from
0.95
0.95
other populations in the area.
Life
Number
Map
code
Area (region)
and locality
Sample
source
history
type
of
fish
Year
sampled
a
Age
A, A
%P
H
D
0.95
Nass River (inland)
1
2
Cranberry
Damdochax
Area mean
Lower Skeena River (inland)
Kitsumkalum
Cedar
Area mean
Upper Skeena River (inland)
Kitwanga
W
W
S
S
93
75
88, 89
88
22.6
19.4
21.0
0.064
0.051
0.058
0.059
0.059
0.059
A
b
3
4
W
W
S
S
A, A, A
A
281
100
88, 89, 91
91
32.3
32.3
32.3
0.082
0.102
0.092
0.076
0.076
0.076
b
5
6
7
8
9
W
W
S
S
S
S
S
S
S
A
111
192
100
80
80
113
100
91
22.6
19.4
22.6
25.8
29.0
25.8
25.8
24.4
0.076
0.085
0.065
0.080
0.078
0.064
0.072
0.075
0.061
0.085
0.056
0.061
0.052
0.068
0.049
0.062
b
c
Bulkley
Morice
Kispiox
Fort Babine
Babine
J, A
89, 91
91
W
W
A
b
c
J
89
90
H
W
J
b
10
11
A, A
A
82, 88
91
Bear
W
Area mean
Central Coast (coastal)
Kitimat
Atnarko
Wannock
Area mean
b
12
13
14
W
W
W
I
I
O
J, A
J, A, A
A, A
190
329
180
85, 88
85, 90, 91
88, 91
35.5
29.0
29.0
31.2
0.085
0.077
0.084
0.081
0.073
0.065
0.087
0.075
b
b
West Vancouver Island (coastal)
Sucwoa
15
16
17
18
19
W
H
W
H
O
O
O
O
O
J
J
J
40
100
40
200
100
85
85
85
81, 85
85
25.8
22.6
29.0
22.6
19.4
23.9
0.078
0.075
0.079
0.086
0.065
0.076
0.053
0.050
0.050
0.059
0.051
0.052
Conuma
Conuma
Robertson
Nitinat
A, J
J
W
Area mean
Georgia Strait (coastal)
Quinsam
20
21
22
23
24
H
H
H
H
O
O
O
O
O
A, J
J
J, J
J
194
100
185
100
150
81, 85
85
81, 85
85
25.8
12.9
25.8
19.4
25.8
21.9
0.093
0.071
0.098
0.085
0.087
0.086
0.089
0.090
0.089
0.080
0.097
0.089
Puntledge
Big Qualicum
Nanaimo
Squamish
Area mean
b
W
J
85
Lower Fraser River (coastal)
Harrison
Birkenhead
Area mean
b
25
26
W
H
O
S
J
150
180
85
85, 87
29.0
16.1
22.6
0.108
0.048
0.078
0.125
0.125
0.125
c
c
J, J
Lower Thompson River (inland)
Spius
b
27
28
29
30
31
W
W
W
W
S
S
S
S
S
J
J
78
196
162
80
87
87
82, 87
87
87
22.6
22.6
22.6
32.3
22.6
24.5
0.090
0.082
0.088
0.095
0.092
0.094
0.058
0.056
0.054
0.060
0.060
0.058
b
Nicola
Coldwater
Bonaparte
Deadman
b
A, J
J
J
b
b
W
120
Area mean
South Thompson River (inland)
Adams
32
33
34
35
36
W
W
W
W
I
A
J, J, J
J, J, J
J
J
80
260
260
80
87
29.0
29.0
25.8
19.4
25.8
25.8
0.102
0.101
0.090
0.088
0.093
0.091
0.065
0.064
0.054
0.055
0.057
0.059
b
Salmon
Eagle
Lower Shuswap
Middle Shuswap
Area mean
S
S
I
85, 87, 88
85, 87, 88
87
b
b
b
W
I
120
87

CHINOOK SALMON POPULATION STRUCTURE
199
TABLE 1.—Continued.
Life
Number
of
fish
Map
code
Area (region)
and locality
Sample
source
history
Year
sampled
a
type
Age
A, J
%P
H
D
0.95
North Thompson River (inland)
Clearwater
b
37
38
39
40
41
42
W
W
W
W
W
W
S
S
S
S
S
S
145
80
80
100
160
200
82, 85
87
87
85
87
25.8
0.087
0.103
0.086
0.090
0.075
0.099
0.089
0.062
0.063
0.062
0.074
0.087
0.070
0.070
b
White Horse Bluff
Horseshoe
Raft
J
J
J
J
J
32.3
25.8
19.4
19.4
29.0
25.3
b
b
c
b
Finn
b
North Thompson
Area mean
87
Mid Fraser River (inland)
Chilcotin
b
43
44
45
46
47
48
49
50
51
W
W
W
W
W
W
W
W
S
S
S
S
S
S
S
S
S
J
80
169
340
160
180
220
240
80
88
29.0
32.3
29.0
29.0
25.8
25.8
29.0
25.8
29.0
28.3
0.088
0.088
0.092
0.096
0.089
0.083
0.084
0.090
0.091
0.091
0.055
0.063
0.054
0.067
0.054
0.065
0.050
0.056
0.063
0.059
b
Chilko
Quesnel
Lower Cariboo
Upepr Cariboo
Cottonwood
Blackwater
Nazko
Baezaeko
Area mean
A, J
J, J, J
c
J
82, 87
85, 87, 88
87
85, 87
85, 87
85, 87
87
b
b
b
c
c
J , J
b
J, J
J, J
J
b
b
b
c
W
J , J
220
85, 87
Upper Fraser River (inland)
Willow
52
53
54
55
56
57
58
59
60
61
62
63
W
W
W
W
W
S
S
S
S
S
S
S
S
S
S
S
S
J, J
J
J
J
J
J, J
J
J
J
J
252
71
85, 87
87
85
85
85
85, 87
85
87
87
87
22.6
29.0
29.0
25.8
22.6
29.0
19.4
25.8
29.0
29.0
29.0
29.0
26.6
0.083
0.094
0.083
0.079
0.078
0.087
0.072
0.085
0.091
0.075
0.087
0.093
0.090
0.050
0.045
0.048
0.048
0.054
0.042
0.055
0.050
0.055
0.048
0.059
0.072
0.052
Salmon
Seebach—McGregor
James—McGregor
Fontoniko—McGregor
Bowron
Indianpoint—Bowron
Slim
Walker—Torpy
Morkill
Horsey
100
100
100
270
100
80
80
80
80
80
b
W
W
W
W
W
W
W
J
A
87
87
Swift
Area mean
a
Data from Taylor (1989) and personal communications from fishery biologists of the Canadian Department of Fisheries and Oceans.
Sample taken in conjunction with enhancement project.
Brood sampled was the progeny of 60 to 100 adult fish.
b
c
rivers and in rivers along the Central coast had between the 63 populations (Figure 4) showed sim-
high PC2 scores. Populations in west Vancouver ilar relationships. Two large clusters were joined
Island and along Georgia Strait had lower PC2 at a genetic distance of 0.126. The first group con-
scores. Within the Fraser River, PC2 also separated tained populations north of the Wannock River of
inland and northernmost populations from lower the Central coast and Fraser River populations
river and southern populations. Populations in the above Fraser Canyon. The other cluster contained
mid- and upper Fraser River had higher PC2 scores the populations from west Vancouver Island and
than did populations in Harrison, south Thompson, the Georgia Strait areas and two lower Fraser River
and lower Thompson rivers. Most populations in populations.
the north Thompson River had intermediate PC2
scores. One geographical exception was the Bir- (D
Each of these two clusters contained subgroups
0.074–0.083 between subgroups) that in-
ϭ
kenhead Hatchery population in the lower Fraser cluded populations in the same drainage or area.
River, which had a relatively high PC2 score. The However, a few genetic outliers also appeared. For
scatterplot showed that in addition to these broad example, Cedar River and Bulkley River popula-
patterns of PC1 and PC2 scores, most populations tions did not cluster closely with others from the
clustered with other populations from the same Skeena River, and the Finn Creek population was
major drainage or area. In many cases, these clus- not part of the north Thompson River cluster. Pop-
ters were distinct from one another.
ulations in the Birkenhead Hatchery and Harrison
Unweighted pair-group method with arithmetic River of the lower Fraser River clustered relatively
averaging algorithm analysis of chord distances distantly with the southern coastal populations.

200
TEEL ET AL.
TABLE 2.—Enzymes and study results for 25 polymorphic loci in samples of British Columbia chinook salmon.
Enzyme
Number of
populations
polymorphic
Range of
common allele
frequency
Commission
a
Enzyme name
number
Locus
Aspartate amino transferase
2.6.1.1
3.5.4.4
sAAT-1,2*
sAAT-3*
sAAT-4*
ADA-1*
ADA-2*
sAH*
FDHG*
GPI-A*
GR*
sIDHP-1*
sIDHP-2*
LDH-B1*
LDH-B2*
LDH-C*
sMDH-B1,2*
MPI*
30
19
55
54
7
63
14
52
58
37
38
8
1.000–0.930
1.000–0.779
1.000–0.415
1.000–0.680
1.000–0.957
0.994–0.528
1.000–0.942
1.000–0.800
1.000–0.622
1.000–0.918
1.000–0.854
1.000–0.933
1.000–0.925
1.000–0.958
1.000–0.847
0.894–0.341
1.000–0.887
1.000–0.560
1.000–0.992
1.000–0.740
0.650–0.006
1.000–0.650
1.000–0.988
1.000–0.878
1.000–0.929
Adenosine deaminase
Aconitate hydratase
4.2.1.3
1.2.1.1
5.3.1.9
1.6.4.2
1.1.1.42
Formaldehyde dehydrogenase (glutathione)
Glucose-6-phosphate isomerase
Glutathione reductase
Isocitrate dehydrogenase
Lactate dehydrogenase
1.1.1.27
9
14
51
63
42
53
8
49
63
60
3
Malate dehydrogenase
Mannose-6-phosphate isomerase
Dipeptidase
Tripeptide aminopeptidase
Proline dipeptidase
Leucyl-L-tyrosine peptidase
Phosphoglycerate kinase
Superoxide dismutase
Triose-phosphate isomerase
1.1.1.37
5.3.1.8
3.4.–.–
3.4.–.–
3.4.13.9
3.4.–.–
2.7.2.3
1.15.1.1
5.3.1.1
PEPA*
PEPB-1*
PEPD-2*
PEPLT*
PGK-2*
sSOD-1*
TPI-1*
TPI-3*
TPI-4*
10
24
a
IUBMBNC (1992).
The largest average chord distances (Table 1) were
found between two populations in the lower Fraser
River (0.125). Other areas had much smaller av-
erage distances, which ranged from 0.052 for west
Vancouver Island and upper Fraser River to 0.089
for Georgia Strait.
Discussion
Genetic Variation within Populations
The levels of genetic diversity in chinook salm-
on in British Columbia do not indicate that pop-
ulations have experienced strong reductions in
T
ABLE 3.—Components of genetic diversity in chinook salmon for 13 highly polymorphic loci. H
ϭ
total expected
relative differentiation among populations within
relative differentiation between regions (coastal
T
heterozygosity; H
areas; G
ϭ
average within-sample heterozygosity; G
ϭ
S
SA
ϭ
relative differentiation among areas within regions; G
ϭ
AR
RT
vs. inland populations); (1
Ϫ
G
ST
) ϭ within-population variation.
Locus
H
T
H
S
G
SA
G
AR
G
RT
(1
Ϫ
G
ST
)
sAAT-3*
sAAT-4*
ADA-1*
sAH*
0.053
0.335
0.145
0.224
0.262
0.052
0.011
0.441
0.144
0.099
0.402
0.186
0.013
0.085
0.047
0.285
0.134
0.206
0.240
0.049
0.011
0.417
0.121
0.093
0.356
0.177
0.012
0.077
0.036
0.045
0.046
0.039
0.027
0.024
0.014
0.032
0.040
0.048
0.022
0.036
0.016
0.033
0.035
0.032
0.018
0.015
0.045
0.031
0.040
0.020
0.054
0.015
0.091
0.009
0.042
0.035
0.034
0.072
0.013
0.026
0.014
0.003
0.002
0.001
0.069
0.000
0.000
0.006
0.004
0.019
0.895
0.851
0.923
0.919
0.915
0.943
0.944
0.946
0.837
0.936
0.887
0.949
0.938
0.913
GR*
sIDHP-2*
LDH-B2*
MPI*
PEPB-1*
PEPLT*
PGK-2*
sSOD-1*
TPI-3*
Mean(25 loci)

CHINOOK SALMON POPULATION STRUCTURE
201
size, nor do they indicate that chinook salmon have because these studies included samples from larger
experienced founder effects. Chinook salmon in
British Columbia contained intermediate levels of
within-population genetic variation in comparison
with populations in other regions. We found 21
P0.95 loci in British Columbia chinook salmon that
were similar to values for chinook salmon in Cal-
ifornia and Oregon (34 of 78 loci; Bartley et al.
1992), in the Snake River (24 of 63 loci; Waples
et al. 1993), and in the upper Columbia River (18
of 32 loci; Utter et al. 1995).
Estimates of heterozygosity for British Colum-
bia populations of chinook salmon appear to be
intermediate to those for populations in other ar-
eas. Our estimates of population heterozygosity
averaged 0.084 (range: 0.048–0.108; 31 loci) and
were similar to those of Utter et al. (1989), who
found a mean heterozygosity for 10 British Co-
lumbia chinook salmon populations of 0.085 (25
loci). In contrast, populations of chinook salmon
along the Oregon coast had a mean population
heterozygosity of 0.122 (Utter et al. 1989). How-
ever, populations located toward the edge of the
range of this species tend to have smaller hetero-
zygosities. For example, heterozygosities for pop-
ulations in the Snake River averaged 0.040, and
those in the Klamath River averaged 0.029 (Utter
et al. 1989). The various levels of genetic diversity
in chinook salmon populations may reflect popu-
lation size or population stability. If so, chinook
populations in British Columbia appear to have
had population histories that were free of severe
population bottlenecks.
areas.
Slightly less than one-third of the total among-
population gene diversity occurred within areas,
the smallest unit of the geographic hierarchy that
we examined. In general, populations within areas
appeared to be genetically distinct from, although
clearly related to, one another. In some cases, the
areas we defined were quite large, and isolation
by distance may account for genetic differences
among populations. However, in other cases, ge-
netic distances were pronounced among popula-
tions that spawned in different locations within the
same drainage. Differences in adult migration and
spawn timing, in addition to spatial separation,
provided a likely mechanism to limit gene flow
between these populations. For example, the Kit-
sumkalum River and Cedar River populations in
the lower Skeena River do not cluster closely to-
gether, although both rivers are part of the Kit-
sumkalum River subbasin. The two populations
have quite different timings for both adult migra-
tion and spawning. The Cedar River has an adult
spring-run population with individuals that spawn
largely in July and August. Fish return to the Kit-
sumkalum River in summer and spawn mostly in
September. In the lower Fraser River, the two pop-
ulations we sampled are genetically distinct and
are also characterized by very different adult life
histories (Fraser et al. 1982). Fish return to the
Harrison River in fall and spawn in October and
November, later than other Fraser River popula-
tions. In contrast, the Birkenhead Hatchery pop-
ulation originated from spring-run fish that are
among the earliest migrating and spawning chi-
nook salmon in the Fraser River.
Over greater geographic distances, the pattern
of genetic differentiation among populations is
largely congruent with geographic relationships.
In some cases, groups of populations can be char-
acterized by frequencies of individual alleles. Re-
sults of PCA and cluster analyses that use multi-
locus information further depict genetic relation-
ships that have geographic coherence. Distinct
groups of populations can be identified in the Nass
River, the Skeena River, the Central coast, the west
coast of Vancouver Island, the east coast of Van-
couver Island and the southern mainland, the lower
Thompson River, the south Thompson River, the
north Thompson River, and the mid- and upper
Fraser River. In many cases, genetic groups are
also identifiable on a larger geographic scale. For
Variability among Populations
One goal of this study was to examine allele-
frequency variability among populations of chi-
nook salmon in British Columbia. We found strong
differences among populations at all levels of the
geographic hierarchy that we examined. The pro-
portion of genetic variability among populations
(8.7%) is typical for groups of populations in other
areas. Smaller among-population diversities were
found by Gharrett et al. (1987) for Alaskan chi-
nook salmon (5.9%) and by Waples et al. (1993)
for Snake River populations (3.4%). However, the
samples from these studies encompassed smaller
geographic ranges and contained fewer genetically
differentiated life history types than do samples in
the present study. High levels of among-popula-
tion diversity were reported by Bartley et al.
(1992) for California and Oregon populations
(10.6%) and by Utter et al. (1989) for populations
from the Pacific Northwest (12.3%), most likely example, the Thompson River genetic groups clus-

202
TEEL ET AL.
F
IGURE 2.—Frequencies for six alleles in 63 chinook salmon populations in British Columbia. Samples are
arranged in the order listed in Table 1. Areas are Nass River (N), lower Skeena River (LS), upper Skeena River
(US), Central coast (CC), west Vancouver Island (WV), Georgia Strait (GS), lower Fraser River (LF), lower
Thompson River (LT), south Thompson River (ST), north Thompson River (NT), mid-Fraser River (MF), and upper
Fraser River (UF).

CHINOOK SALMON POPULATION STRUCTURE
203
F
IGURE 2.—Continued.
ter together and are most genetically similar to of British Columbia chinook salmon. The first
populations from the mid-Fraser River.
group contained populations in the Skeena and
Nass rivers, along with those in the Fraser River
above the Fraser Canyon. These rivers transverse
Variability among Regions
On a broader level, the PCA and dendrogram the Coast Mountains and drain inland regions. A
revealed two large genetic groups of populations second genetic group contained populations in riv-

204
TEEL ET AL.
T
ABLE 4.—Factor scores of principal component anal-
The separation of British Columbia chinook
ysis (PCA).
salmon into coastal and inland genetic groups is
congruent with groupings based on chinook salm-
on life history forms. In coastal rivers as far north
as the Nass River, populations are predominately
ocean type (Taylor 1989, 1990a). Fish in these
populations migrate to sea primarily as subyear-
lings, they tend to spend most of their ocean life
in coastal waters, and they return to freshwater
shortly prior to spawning, usually in the fall or
summer (Healey 1991). In the upstream drainages
of the Fraser, Skeena, and Nass rivers, populations
are predominately stream type, and fish migrate
downstream primarily as yearlings, move exten-
sively offshore, and return to freshwater several
months prior to spawning, usually in the spring or
summer (Healey 1983; Taylor 1989, 1990a).
The geographic separation between ocean- and
stream-type spawning populations is not complete,
however, and several British Columbia coastal
populations consist of stream-type individuals
(Taylor 1989). Two of the populations we studied
in mainland rivers having headwaters in the Coast
Mountains are stream type. One was in the Squa-
mish River of the lower Strait of Georgia. The
other was the Birkenhead Hatchery in the lower
Fraser River, which uses stream-type fish from the
Birkenhead River for broodstock. Both popula-
tions clustered with the coastal ocean-type popu-
lations and were genetically dissimilar to inland
stream-type populations.
Several British Columbia populations also con-
tain substantial proportions of both ocean-type and
stream-type chinook salmon, particularly in the
south Thompson River and along the Central coast
(Taylor 1989). Several populations in these areas
were genetically intermediate between the coastal
and inland populations. For example, in the PCA
plot (Figure 3), populations in the Adams and
Shuswap rivers, which are tributaries of the south
Thompson River, and in the Atnarko and Kitimat
rivers (from the Central coast) were positioned
near the convergence of the coastal and inland ge-
netic groups. These populations have frequencies
Allele
Factor 1
Factor 2
sAAT-1,2*85
sAAT-3*90
sAAT-4*63
ADA-1*83
sAH*86
FDHG*143
GPI-A*105
GPI-A*93
0.58
0.70
0.35
0.08
0.14
0.15
Ϫ
0.76
0.58
0.56
0.44
0.48
0.17
0.07
0.68
0.10
0.48
0.11
0.09
0.16
0.50
0.36
0.03
0.11
0.66
0.00
0.05
0.48
0.43
0.44
0.19
4.74
Ϫ
Ϫ
0.39
0.59
0.35
0.04
0.52
0.23
0.26
0.28
0.15
0.54
0.22
0.36
0.24
0.43
0.67
0.47
0.16
0.62
0.14
0.06
0.53
0.31
3.49
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
GR*85
GR*110
Ϫ
Ϫ
sIDHP-1*74
sIDHP-2*127
sIDHP-2*50
LDH-B1*48
LDH-B2*71
sMDH-B1,2*121
sMDH-B1,2*70
MPI*109
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
PEPA*90
PEPB-1*130
PEPLT*110
PGK-2*100
sSOD-1*Ϫ260
sSOD-1*580
TPI-3*96
TPI-4*104
Eigenvalue
% Variation explained
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
Ϫ
18.2
13.4
ers draining coastal regions and in the western
slopes of the Coast Mountains.
Differentiation between these two genetic groups
is particularly pronounced, over a relatively small
geographic distance, within the Fraser River. The
genetic similarity between lower Fraser River pop-
ulations and coastal fish and the divergence from
those fish above the Fraser Canyon have also been
observed in steelhead Oncorhynchus mykiss (Par-
kinson 1984), coho salmon Oncorhynchus kisutch
(Wehrhahn and Powell 1987; Small et al. 1998),
and sockeye salmon Oncorhynchus nerka (Wood
et al. 1994).
Our allozyme data agree with the genetic group-
ings of populations that were based on a study of
minisatellite DNA in 30 stocks of British Colum-
bia chinook salmon (Beacham et al. 1996). That
study delineated a southern group composed of
southern mainland, Vancouver Island, and Fraser
River populations and a northern group containing
populations from the northern mainland, the Skee-
na River, and the Yukon River. Although Beacham
et al. (1996) placed inland Fraser River popula-
tions in the southern group, the authors presented
results from a PCA in which populations from the
mid- and upper Fraser River clustered more closely
to Skeena River populations.
of the sAAT4
*63 and PEPB1*130 alleles that are
intermediate between frequencies in southern
coastal populations and those in inland popula-
tions. Although the PC1 values show the genetic
affinities between the Atnarko and Kitimat popu-
lations and the southern coastal populations, the
PC2 values and the dendrogram (Figure 4) indicate
genetic similarities to populations in the Skeena
and Nass rivers.

CHINOOK SALMON POPULATION STRUCTURE
205
F
IGURE 3.—Scatterplot of scores along the first two principal component (PC) axes for 63 chinook salmon
populations in British Columbia. Location numbers are given in Table 1 and Figure 1. Populations within an area
are identified with polygons where possible.
Zoogeography of Life History Types
British Columbia chinook salmon. The genetic
similarity among geographically isolated stream-
type populations in the upper Fraser, Skeena, and
Nass rivers suggests that these chinook salmon
descended from a common ancestor. Such a pattern
would not arise from convergent selection for a
particular life history type. Previous genetic data
indicate that the upper Fraser River may have been
colonized by chinook salmon from the upper Co-
lumbia and Snake rivers (Utter et al. 1989).
Stream-type populations from these rivers possess
behaviors—spring or summer adult return-timing
and juvenile residence in freshwater for at least a
year—that would be needed to successfully col-
onize the upstream areas of large British Columbia
rivers.
The genetic study of Utter et al. (1989), how-
ever, included only a single sample north of Van-
couver Island. Genetic analyses of populations in
Alaska and in Asia, north of the glaciation, where
stream-type juveniles predominate, may be partic-
ularly important. Recent allozyme studies of chi-
nook salmon in Alaska and Asia (P. Crane, Alaska
Department of Fish and Game, and C. Guthrie,
The dichotomous distributions of ocean- and
stream-type chinook salmon life history forms
most likely reflect postglacial dispersals and col-
onizations (Healey 1983, 1991; Gharrett et al.
1987; Taylor 1990a). Most of British Columbia
was covered with ice during the last glacial max-
imum and has been repopulated by chinook salmon
within the last 13,000 years (McPhail and Lindsey
1986). Potential sources for recolonization include
chinook salmon that survived in the Bering refug-
ium to the north and in the Pacific and Columbia–
Snake River refugia to the south (McPhail and
Lindsey 1970; Gharrett et al. 1987). Ocean- and
stream-type chinook salmon may have survived
the glaciation in different refugia, and following
retreat of the ice sheet, the two forms may have
expanded into deglaciated areas in Washington,
British Columbia, and Alaska.
The pattern of allozyme variability in British
Columbia largely supports this hypothesis. If we
can assume near-neutrality of the majority of allo-
zyme alleles (Kimura 1968), our results imply
long-standing genetic isolation between inland
stream-type and coastal ocean-type populations of National Marine Fisheries Service, personal com-

206
TEEL ET AL.
F
IGURE 4.—Dendrogram of Cavalli-Sforza and Edwards (1967) chord distances (31 loci) among 63 chinook
salmon populations in British Columbia (cophenetic correlation 0.848). Areas are Nass River (N), lower Skeena
ϭ
River (LS), upper Skeena River (US), Central coast (CC), west Vancouver Island (WV), Georgia Strait (GS), lower
Fraser River (LF), lower Thompson River (LT), south Thompson River (ST), north Thompson River (NT), mid-
Fraser River (MF), and upper Fraser River (UF).

CHINOOK SALMON POPULATION STRUCTURE
207
munications) will make these comparisons possi- trait, the response to photoperiod change, appears
ble and may provide more insight into the origins to be under the control of a single genetic locus
of stream-type populations in inland British Co- (Clarke et al. 1992). Clarke et al. (1992) suggested
lumbia.
that disruptive selection at this locus in ocean-type
Genetic data indicate that southern British Co- populations might have repeatedly produced
lumbia coastal rivers were colonized by chinook stream-type responses to photoperiod. The close
salmon from the Pacific refuge (Utter et al. 1989). genetic affinity between stream- and ocean-type
If fish surviving in coastal rivers south of the ice forms in coastal areas in our study and the occur-
sheet had chiefly ocean-type life histories, as is rence of populations with a mix of individuals with
presently the case, a northward expansion along different life history types support this hypothesis.
the coast would explain the distribution of ocean- Therefore, various life history patterns in chinook
type populations in southern British Columbia. salmon appear to have arisen on long and short
The transition from ocean-type populations to time scales.
stream-type populations at 55ЊN (Figure 1), with-
Acknowledgments
out a corresponding change in environmental con-
ditions, supports a zoogeographic explanation for
the appearance of different life history types in
British Columbia (Healey 1983; Taylor 1990a).
We thank Ken Pitre, Brian Riddell, Neil Schu-
bert, and Barb Snyder of the Canadian Department
of Fisheries and Oceans and Marc Miller and Sew-
all Young of the Washington Department of Fish
and Wildlife, who coordinated sample collections.
Paul Aebersold and Don Van Doornik of the U.S.
National Marine Fisheries Service supervised
much of the laboratory analysis. Jim Peacock and
Kathleen Neely drafted the figures for this article.
Tony Gharrett, Anne Marshall, and Brian Riddell
generously provided valuable reviews of an earlier
manuscript.
Life History Variability within Regions
The patterns of life history variability in some
areas conflict with a strictly zoogeographic expla-
nation for the origin of ocean- and stream-type
forms. Healey (1991) thought these exceptions
represented a serious inconsistency with the zoo-
geographic model. In British Columbia, some
stream-type populations occur in areas that are oth-
erwise inhabited by ocean-type populations, and
many populations contain some individuals of both
life history types. Populations with intermediate
genetic and life history patterns in the south
Thompson River and along the Central coast may
be explained by mixing or hybridization of inland
stream-type and coastal ocean-type lineages. How-
ever, other stream-type populations in southern
British Columbia coastal rivers were genetically
most closely related to ocean-type populations and
appear to have been derived from the same an-
cestral ocean-type populations, which implies a
postcolonization divergence in life history types.
Several studies have shown that life history pat-
terns in chinook salmon can change rapidly. For
example, ocean-type chinook salmon introduced
in New Zealand have developed some stream-type
behaviors over a period of about 30 generations
(Quinn and Unwin 1993; Quinn et al. 1996). Also,
in the Great Lakes, where fall-spawning chinook
salmon were introduced, spring-spawning behav-
ior has appeared in only a few generations (Kwain
and Thomas 1984). Breeding experiments have
shown that several behavioral and life history
characteristics are controlled by the interaction be-
tween genetic and environmental variability (Tay-
lor 1988, 1990b; Clarke et al. 1992, 1994). One
References
Aebersold, P. B., G. A. Winans, D. J. Teel, G. B. Milner,
and F. M. Utter. 1987. Manual for starch gel elec-
trophoresis: a method for the detection of genetic
variation. NOAA Technical Report NMFS 61.
Aro, K. V., and M. P. Shepard. 1967. Salmon of the
North Pacific Ocean. Part IV. Spawning populations
of North Pacific salmon. 5. Pacific salmon in Can-
ada. International North Pacific Fisheries Commis-
sion Bulletin 23:225–327.
Banks, M. A., B. A. Baldwin, and D. Hedgecock. 1996.
Research on chinook salmon (Oncorhynchus tshaw-
ytscha) stock structure using microsatellite DNA.
Bulletin of the National Research Institute of Aqua-
culture 2(Supplement):5–9.
Bartley, D., B. Bentley, J. Brodziak, R. Gomulkiewicz,
M. Mangel, and G. A. E. Gall. 1992. Geographic
variation in population genetic structure of chinook
salmon from California and Oregon. U.S. National
Marine Fisheries Service Fishery Bulletin 90:77–
100. [Authorship amended per errata, Fishery Bul-
letin 90(3):iii.]
Beacham, T. D., R. E. Withler, and T. A. Stevens. 1996.
Stock identification of chinook salmon (Oncorhyn-
chus tshawytscha) using minisatellite DNA varia-
tion. Canadian Journal of Fisheries and Aquatic Sci-
ences 53:380–394.
Cavalli-Sforza, L. L., and A. W. F. Edwards. 1967. Phy-
logenetic analysis: models and estimation proce-
dures. Evolution 21:550–570.

208
TEEL ET AL.
Chakraborty, R., M. Hagg, N. Ryman, and G. Sta¨hl.
1982. Hierarchical gene diversity analysis and its
application to brown trout population data. Here-
ditas 97:17–21.
Mitochondrial DNA polymorphism in unique runs
of chinook salmon (Oncorhynchus tshawytscha)
from the Sacramento–San Joaquin river basin. Con-
servation Biology 8:882–884.
Clarke, W. C., R. E. Withler, and J. E. Shelbourn. 1992. Parkinson, E. A. 1984. Genetic variation in populations
Genetic control of juvenile life history pattern in
chinook salmon (Oncorhynchus tshawytscha). Ca-
nadian Journal of Fisheries Aquatic Sciences 49:
2300–2306.
of steelhead trout (Salmo gairdneri) in British Co-
lumbia. Canadian Journal of Fisheries and Aquatic
Sciences 41:1412–1420.
Quinn, T. P., and M. J. Unwin. 1993. Variation in life
history patterns among New Zealand chinook salm-
on (Oncorhynchus tshawytscha) populations. Ca-
nadian Journal of Fisheries and Aquatic Sciences
50:1414–1421.
Clarke, W. C., R. E. Withler, and J. E. Shelbourn. 1994.
Inheritance of smolting phenotypes in backcrosses
of hybrid stream-type
ϫ ocean-type chinook salm-
on. Estuaries 17:13–25.
Fraser, F. J., P. J. Starr, and A. Y. Fedorenko. 1982. A Quinn, T. P., and six coauthors. 1996. Origin and genetic
review of the chinook and coho salmon of the Fraser
River. Canadian Technical Report of Fisheries and
Aquatic Sciences 1126.
structure of chinook salmon, Oncorhynchus tshaw-
ytscha, transplanted from California to New Zea-
land: allozyme and mtDNA evidence. U.S. National
Marine Fisheries Service Fishery Bulletin 94:506–
521.
Gharrett, A. J., S. M. Shirley, and G. R. Tromble. 1987.
Genetic relationships among populations of Alas-
kan chinook salmon (Oncorhynchus tshawytscha). Rohlf, F. J. 1993. NTSYS-pc. Numerical taxonomy and
Canadian Journal of Fisheries and Aquatic Sciences
44:765–774.
multivariate analysis system, version 1.80. Exeter
Software, Setauket, New York.
Healey, M. C. 1983. Coastwide distribution and ocean Shaklee, J. B., F. W. Allendorf, D. C. Morizot, and G.
migration patterns of stream- and ocean-type chi-
nook salmon (Oncorhynchus tshawytscha). Cana-
dian Field-Naturalist 97:427–433.
S. Whitt. 1990a. Gene nomenclature for protein-
coding loci in fish. Transactions of the American
Fisheries Society 119:2–15.
Healey, M. C. 1991. Life history of chinook salmon Shaklee, J. B., C. Busack, A. Marshall, M. Miller, and
(Oncorhynchus tshawytscha). Pages 311–393 in C.
Groot and L. Margolis, editors. Pacific salmon life
histories. University of British Columbia Press,
Vancouver.
S. R. Phelps. 1990b. The electrophoretic analysis
of mixed-stock fisheries of Pacific salmon. Pages
235–265 in Z. I. Ogita and C. L. Markert, editors.
Isozymes: structure, function, and use in biology
and medicine. Progress in clinical and biological
research, volume 344. Wiley-Liss, New York.
IUBMBNC (International Union of Biochemistry and
Molecular Biology, Nomenclature Committee).
1992. Enzyme nomenclature, 1992. Academic Small, M. P., R. E. Withler, and T. D. Beacham. 1998.
Press, San Diego, California.
Kimura, M. 1968. Evolutionary rate at the molecular
level. Nature (London) 217:624–626.
Kwain, W., and E. Thomas. 1984. The first evidence of
spring spawning by chinook salmon in Lake Su-
Population structure and stock identification of Brit-
ish Columbia coho salmon, Oncorhynchus kisutch,
based on microsatellite DNA variation. U.S. Na-
tional Marine Fisheries Service Fishery Bulletin 96:
843–858.
perior. North American Journal of Fisheries Man- Sneath, P. H. A., and R. R. Sokal. 1963. Numerical
agement 4:227–228. taxonomy. W. H. Freeman, San Francisco.
Levene, H. 1949. On a matching problem arising in Sokal, R. R., and F. J. Rohlf. 1981. Biometry, 2nd edi-
genetics. Annals of Mathematics and Statistics 20:
91–94.
McPhail, J. D., and C. C. Lindsey. 1970. Freshwater
fishes of northwestern Canada and Alaska. Fisheries
Research Board of Canada Bulletin 173.
tion. W. H. Freeman, San Francisco.
Swofford, D. L., and R. B. Selander. 1981. BIOSYS-1:
a FORTRAN program for the comprehensive anal-
ysis of electrophoretic data in population genetics
and systematics. Journal of Heredity 72:281–283.
McPhail, J. D., and C. C. Lindsey. 1986. Zoogeography Taylor, E. B. 1988. Adaptive variation in rheotactic and
of the freshwater fishes of Cascadia (the Columbia
system and rivers north to the Stikine). Pages 615–
637 in C. H. Hocutt and E. O. Wiley, editors. The
zoogeography of North American freshwater fishes.
John Wiley and Sons, New York.
agonistic behavior in newly emerged fry of chinook
salmon, Oncorhynchus tshawytscha, from ocean-
and stream-type populations. Canadian Journal of
Fisheries and Aquatic Sciences 45:237–243.
Taylor, E. B. 1989. Adaptive diversification of juvenile
life histories in the chinook salmon, Oncorhynchus
tshawytscha, (Walbaum). Doctoral dissertation.
University of British Columbia, Vancouver.
Milner, G. B., D. J. Teel, F. M. Utter, and G. A. Winans.
1985. A genetic method of stock identification in
mixed populations of Pacific salmon, Oncorhynchus
spp. U.S. National Marine Fisheries Service Marine Taylor, E. B. 1990a. Environmental correlates of life-
Fisheries Review 47:1–8.
history variation in juvenile chinook salmon, On-
corhynchus tshawytscha (Walbaum). Journal of Fish
Biology 37:1–17.
Nei, M. 1978. Estimation of average heterozygosity and
genetic distance from a small number of individuals.
Genetics 89:583–590.
Taylor, E. B. 1990b. Phenotypic correlates of life-
history variation in juvenile chinook salmon, On-
Nielsen, J. L., D. Tupper, and W. K. Thomas. 1994.

CHINOOK SALMON POPULATION STRUCTURE
209
corhynchus tshawytscha. Journal of Animal Ecology
59:455–468.
stock fishery analysis. Canadian Journal of Fish-
eries and Aquatic Sciences 47:968–976.
Waples, R. S., and six coauthors. 1993. A genetic mon-
itoring and evaluation program for supplemented
populations of salmon and steelhead in the Snake
River basin. Annual Report to the Bonneville Power
Administration, Portland, Oregon.
Utter, F., P. Aebersold, and G. Winans. 1987a. Inter-
preting genetic variation detected by electrophore-
sis. Pages 21–45 in N. Ryman and F. Utter, editors.
Population genetics and fishery management. Uni-
versity of Washington Press, Seattle.
Waples, R. S., D. J. Teel, and P. B. Aebersold. 1991. A
genetic monitoring and evaluation program for sup-
plemented populations of salmon and steelhead in
the Snake River basin. Annual Report to the Bon-
neville Power Administration, Portland, Oregon.
Waples, R. S., G. A. Winans, F. M. Utter, and C. Mahn-
ken. 1990. Genetic approaches to the management
of Pacific salmon. Fisheries 15(5):19–25.
Utter, F. M., D. W. Chapman, and A. R. Marshall. 1995.
Genetic population structure and history of chinook
salmon of the upper Columbia River. Pages 149–
165 in J. L. Nielsen and D. A. Powers, editors.
Evolution and the aquatic ecosystem: defining
unique units in population conservation. American
Fisheries Society, Symposium 17, Bethesda, Mary-
land.
WDF (Washington Department of Fisheries), WDW
(Washington Department of Wildlife), and WWTIT
(Western Washington Treaty Indian Tribes). 1993.
1992 Washington State salmon and steelhead stock
inventory (SASSI). Washington Department of Fish
and Wildlife, Olympia.
Wehrhahn, C. F., and R. Powell. 1987. Electrophoretic
variation, regional differences, and gene flow in the
coho salmon (Oncorhynchus kisutch) of southern
British Columbia. Canadian Journal of Fisheries
and Aquatic Sciences 44:822–831.
Wood, C. C., B. E. Riddell, D. T. Rutherford, and R. E.
Withler. 1994. Biochemical genetic survey of sock-
eye salmon (Oncorhynchus nerka) in Canada. Ca-
nadian Journal of Fisheries and Aquatic Sciences
51(Supplement 1):114–131.
Utter, F., G. Milner, G. Sta¨hl, and D. Teel. 1989. Genetic
population structure of chinook salmon, Oncorhyn-
chus tshawytscha, in the Pacific Northwest. U.S. Na-
tional Marine Fisheries Service Fishery Bulletin 87:
239–264.
Utter, F., D. Teel, G. Milner, and D. McIsaac. 1987b.
Genetic estimates of stock compositions of 1983
chinook salmon, Oncorhynchus tshawytscha, har-
vests off the Washington coast and the Columbia
River. U.S. National Marine Fisheries Service Fish-
ery Bulletin 85:13–23.
Waples, R. S. 1988. Estimation of allele frequencies at
isoloci. Genetics 118:371–384.
Waples, R. S. 1990. Temporal changes of allele fre-
quency in Pacific salmon: implications for mixed-