94 
Abstract— Microsatellite DNA analy- 
sis was applied in a genetic study of 
20 chinook salmon populations from 
four regions within the Fraser River 
drainage of British Columbia, Canada. 
Twelve populations were represented 
by samples collected in different years. 
A total of 2612 fish were examined at 
three microsatellite loci. Each locus was 
highly polymorphic, with 30 alleles at 
OfslOl, 28 alleles at OtslOO, and 35 
alleles at Ofsl02. Average observed het- 
erozygosities were 86%, 88%, and 71%, 
respectively. In a dendrogram analysis 
of pairwise genetic distances, four geo- 
graphically based groups were observed 
consisting of the lower Fraser River, the 
middle Fraser River, the upper Fraser 
River, and the Thompson River. An 
analysis of molecular variance showed 
that 97.57% of the genetic variance was 
within populations and 1.80% of the 
genetic variance was partitioned among 
populations. We detected significantly 
different allele frequencies among pop- 
ulations within regional groupings and 
temporal stability in allele frequencies 
in populations for which multiple years 
of samples were analyzed. Regional 
divergence may reflect colonization pat- 
terns following the last ice age, and 
divergence among populations within 
regions may reflect local adaptation. 
The elucidation of population structure 
of chinook salmon of the Fraser River 
watershed will be useful information 
for management designed to conserve 
genetic biodiversity. 
Manuscript accepted 14 July 2000. 
Fish. Bull. 99:94-107 (2001). 
Population structure of Fraser River chinook salmon 
( Oncorhynchus tshawytscha): an analysis using 
microsatellite DNA markers 
R. John Nelson 
Maureen P. Small 
Terry D. Beacham 
K. Janine Supernault 
Pacific Biological Station 
Nanaimo, British Columbia 
V9R 5K6, Canada 
Present address (for R J Nelson): SeaStar Biotech Inc. 
32056-3749 Shelbourne St. 
Victoria, British Columbia 
V8P 5S2 Canada 
E-mail address (for R. J. Nelson) |nelson@seastarbio.com 
The Fraser River watershed produces 
greater numbers of Pacific salmon than 
any other river system in British Colum- 
bia (B.C.). Approximately 65 tributar- 
ies of the Fraser River are used as 
spawning and rearing habitat for chi- 
nook salmon (Oncorhynchus tshawyts- 
cha), and these streams produce up to 
one third of the commercial catch of 
chinook salmon from Brirish Columbia 
(Fraser et ah, 1982). Although chinook 
salmon account for only 1% to 5% of the 
total escapements of salmon within the 
watershed (Northcote and Atagi, 1997), 
these fish are an important cultural, 
sporting, and food resource. Chinook 
salmon populations in the Fraser water- 
shed have been negatively impacted 
by a variety of forces, in some cases 
reducing ( Bradford, 1994) or completely 
eliminating (Slaney et ah, 1996) local 
populations. 
Historical efforts to maintain and en- 
hance salmon runs through trans- 
plantation have had mixed results, il- 
lustrating that the characteristics of 
a population influence its ability to 
thrive in a given environment (Wood, 
1995). Also, transplantation of fish and 
hatchery production practices may al- 
ter genetic composition of wild stocks 
(Waples, 1994). In fisheries, it is impor- 
tant not to over harvest small popula- 
tions that may contain unique adap- 
tive traits. For the above reasons it is 
advantageous to understand how pop- 
ulation structures evolve in order to 
protect individual salmon runs and to 
preserve biodiversity. 
Most of the chinook salmon popula- 
tions of B.C. were founded after the 
ice of the Wisconsin glaciation retreat- 
ed approximately 10,000 to 15,000 year 
ago (McPhail and Lindsey, 1986). If chi- 
nook salmon recolonization is similar to 
that of sockeye and coho salmon (Wood 
et ah, 1994; Small et ah, 1998a), re- 
establishment of the present day B.C. 
chinook salmon populations may have 
occurred from at least two different 
sources, Beringia to the north and Cas- 
cadia to the south (Gharrett et ah, 
1987), and possibly from a refuge in 
the Queen Charlotte Islands (Warner et 
ah, 1982). The genetic character of the 
founding fish may be reflected in pres- 
ent day genetic structure, but because 
Pacific salmon return to their natal 
streams to spawn (Scheer, 1939; Quinn, 
1984), reproductive isolation can lead 
to divergence of phenotypic and geno- 
typic characters. Neutral genetic mark- 
ers can be used to measure the degree 
of reproductive isolation and potential 
for local adaptation. 
Over the years, a variety of methods 
have been used to examine population 
structure. Allozyme analysis has long 
been a mainstay in fish genetics re- 
search and has been used to determine 
population structure in chinook salm- 
on of Alaska (Gharrett et ah, 1987), 
from California to Alaska (Utter et ah, 
1989), of the Yukon River (Beacham et 
