Nelson et al.: Population structure of Oncorhynchus tshawytscha of the Fraser River 
95 
al., 1989; Wilmot et al., 1992), of California and Oregon 
(Bartley et al. 1992), and in British Columbia (Teel et al. 
2000). These studies suggest that chinook salmon popula- 
tions were genetically heterogeneous and that populations 
could be placed into genetically defined groups correspond- 
ing to geographic regions. In some of the earlier studies 
there was little genetic distinction between geographical- 
ly separate groups because the allozyme markers showed 
low polymorphism. However, using 25 polymorphic allo- 
zymes, Teel et al. (2000) detected strong population diver- 
gence within the Fraser River and among major rivers in 
British Columbia. 
DNA markers can be more polymorphic than allozyme 
markers and thus may be more sensitive to population 
structure; with higher levels of polymorphism, there is an 
increased likelihood for populations to contain unique al- 
leles or to have frequency differences in alleles that are 
shared among populations. Among the DNA-based mark- 
ers, mitochondrial DNA has been used to examine genetic 
structure in chinook salmon populations of the West Coast 
of North America (Wilson et al., 1987; Cronin et al., 1993). 
These studies suggest that there is structuring among 
West Coast chinook salmon populations. However, the low 
resolution of this method limits its utility. Minisatellite 
DNA has been used to study Canadian chinook salmon 
populations. Beacham et al. (1996) found that chinook 
salmon formed two major regional groups in British Co- 
lumbia: a southern group consisting of populations of the 
Fraser River, Vancouver Island, and the southern main- 
land; and a northern group consisting of populations of 
the Skeena River, the Yukon River, and the northern main- 
land. However, owing to technical complexity, the tech- 
nique is unsuitable for studies involving large numbers of 
individuals. 
Microsatellite DNA loci are highly polymorphic and 
technically easy to use (Nelson et al., 1998) and provide 
powerful markers for elucidating population structure. 
Microsatellite loci have provided information regarding 
population divergence in chinook salmon (Banks et al., 
1996) and other salmonids (Angers et al., 1995; Scribner 
et al., 1996; Nelson et al., 1998; Small et al., 1998a, 
1998b). In our study we exploited the ease of analysis 
and the highly polymorphic nature of microsatellite DNA 
loci to study population structure of chinook salmon. We 
surveyed variation at three microsatellite loci within 20 
Fraser River chinook salmon populations and examined 
temporal stability of microsatellite allele frequencies. We 
used this information to hypothesize the genetic structure 
of chinook salmon populations within the Fraser River 
watershed. 
Materials and methods 
DNA extraction 
Liver or scale samples were analyzed from 2612 individual 
chinook salmon from 20 populations of the Fraser and 
Thompson River watersheds (Fig. 1.). Sample sizes ranged 
from 30 to 347 fish (Table 1). Liver samples were obtained 
from spawning wild adults. Hatchery adults were sources 
for the Chehalis-red and Chilliwack-red samples. The 
nomenclature “-red” refers to the red flesh color of the 
fish in the population. DNA was extracted from liver and 
scales archived on scale cards according to the methods of 
Nelson et al. ( 1998). Liver samples taken prior to 1994 were 
subjected to DNA extractions as described in Beacham et 
al. (1996). Each 25 pL of polymerase chain reaction (PCR) 
required either 100 ng of genomic DNA, 0.1 to 1 pL of liver 
extracts, or 5 to 10 pL of scale extract. 
PCR amplification 
The loci amplified in this work were OfslOO (Nelson et 
al., 1998), OislOl (Small et al., 1998a) and Ofsl02 (Nelson 
and Beacham, 1999). PCR amplification was carried out 
in 96-well microtiter plates with a MJ PTC- 100 thermal 
cycler (MJ research, Watertown, MA). 25-pL PCR reactions 
contained 10 pmol (0.4 pM) of each primer, 80 pM of each 
nucleotide, 20 mM tris-pH 8.8, 2 mM MgS0 4 , 10 mM KC1, 
0.1% triton x-100, 10 mM (NH 4 )S0 4 , and 0.1 mg/mL of 
bovine serum albumin. Primer set OtslOO required a 10% 
final volume of glycerol in the PCR. PCR temperature 
cycles were preceded by a denaturation incubation of 3 
min at 94°C; samples then were held at 80°C while 1 unit 
of DNA polymerase was added. PCR cycle parameters and 
primer sequences for each locus are presented in Table 2. 
Three pLs of lOx loading dye (50 mM EDTA pH 8.0, 30% 
glycerol, 0.25%' bromphenol blue) were added to each reac- 
tion and ten pLs of this solution was loaded on each gel 
lane for electrophoresis. 
Gel electrophoresis 
Microsatellite alleles were size-fractionated on nondena- 
turing polyacrylamide gels 17 cm wide by 14.5 cm long. 
Gels consisted of a 19:1 ratio of acrylamide to bis-acryl- 
amide. Gel contained 2x TAE buffer (Maniatis et al., 1982) 
as did the gel box reservoirs. Electrophoretic conditions 
are described in Table 2. Each gel included three 20 base- 
pair (bp) marker lanes (GenSura Labs Inc., Del Mar, CA) 
to create a molecular size grid for sizing amplified micro- 
satellites, 24 population samples, and one “standard fish” 
to estimate the precision of allele sizes (Table 3). Standard 
deviations were calculated for alleles from two different 
standard fish for each primer set. Gels were stained with 
0.5 pg/mL of ethidium bromide in water and visualized 
with ultraviolet light (Fig. 2). 
Digital images of gels were obtained as described in 
Nelson et al. (1998). Individual alleles were identified by 
using the procedure outlined in Small et al. (1998a). A 
four-bp bin was used for all Ots 101 alleles. A four-bp bin 
was used to define smaller alleles of OtslOO and five- to 
eight-bp bins were used for larger alleles. A four-bp bin 
was used for the smaller alleles of Ots 102 and five- to six- 
bp bins were used for larger alleles. These bin sizes (see 
Table 1 for bin designations) were four or more standard 
deviations wide according to estimates derived from the 
standard fish. Bins are referred to as “alleles” throughout 
the text. 
