843 
Abstract .——■We used genetic varia- 
tion at three microsatellite DNA loci to 
describe population structure in 34 coho 
salmon populations from British Co- 
lumbia and to perform stock composi- 
tion analysis on simulated mixed-stock 
fishery samples. Each microsatellite 
locus was highly polymorphic, with 31 
alleles at Ots 101, 20 alleles at Ots 3, and 
38 alleles at Ofsl03. Average observed 
heterozygosities were 86.3%, 73.3% and 
74.9%, respectively. Analysis of genetic 
distances revealed three relatively ho- 
mogeneous, geographically based 
groups of coho salmon populations in 
the following regions: the upper Skeena 
and Nass River watersheds, the lower 
Fraser River drainage, and the upper 
Fraser River-Thompson River water- 
sheds. Coastal populations from the 
mainland of British Columbia, Van- 
couver Island, and the Queen Charlotte 
Islands formed a more heterogeneous 
regional stock grouping. Significantly 
different allele frequencies were ob- 
served among populations within re- 
gions, and allele frequencies were gen- 
erally temporally stable in multiyear 
samples. Phylogenetic lineages within 
British Columbia coho salmon likely 
reflect geographic patterns of recoloni- 
zation from at least three separate gla- 
cial refugia after the last ice age. Local 
spawning populations within regions 
may form metapopulations, but current 
levels of gene flow among subpopula- 
tions are apparently insufficient to pre- 
vent differentiation at neutral genetic 
loci. Maximum-likelihood estimates of 
stock composition were accurate and 
precise, indicating great potential for 
management of coho salmon at the level 
of metapopulations or “evolutionarily 
significant units” in domestic and in- 
ternational mixed-stock fisheries. Indi- 
vidual fish were identified to stock by 
using a discriminant analysis with a 
high degree of accuracy in a few re- 
gions, but more generally with approxi- 
mately 50% success. 
Manuscript accepted 26 January 1998. 
Fish. Bull.: 96:843-858 ( 1998). 
Population structure and stocl< 
identification of British Columbia coho 
salmon, Oncorhynchus kisutch , 
based on microsatellste DNA variation 
Maureen P. Small 
Ruth E. Withler 
Terry D. Beacham 
Pacific Biological Station, Department of Fisheries and Oceans 
Nanaimo, British Columbia V9R 5K6, Canada 
E-mail address (for R. E, Withler, contact author):withlerr@dfo-mpo.gc.ca 
Of the five commercially important 
Pacific salmon (Oncorhynchus) spe- 
cies exploited in the commercial, 
recreational and aboriginal fisher- 
ies of British Columbia, coho salmon 
is the one that has shown the largest 
decline in abundance in both his- 
torical (the past 100 years) and re- 
cent (the past 40 years) times 
(Northcote and Atagi, 1997). Coho 
salmon abundance has also declined 
dramatically in Washington State, 
Oregon, and California, to the ex- 
tent that a number of coho salmon 
populations, including all of those 
from the upper Columbia River sys- 
tem, are considered extinct and oth- 
ers are considered endangered 
(Nehlsen et ah, 1991; Weitkamp et 
al., 1995). The list of factors that 
may have contributed to declining 
coho salmon abundance is long, and 
includes overfishing, loss of fresh- 
water habitat, misguided enhance- 
ment efforts, limited marine carry- 
ing capacity, and climate changes 
resulting in altered oceanic condi- 
tions (Fraser et ah, 1982; Beamish 
and Bouillon, 1993; Walters, 1993; 
Reisenbichler, 1997). 
Because coho salmon tend to 
spawn in small streams and rivers 
that are especially vulnerable to 
degradation through human activi- 
ties, freshwater habitat degradation 
has been considered a major factor 
in declining coho abundance (Fraser 
et ah, 1982). Similarly, coho salmon 
commonly are believed to be over- 
exploited because of the mixed- 
stock nature of most coho salmon 
fisheries and the frequent bycatch 
of coho salmon in fisheries directed 
toward other species (Fraser et ah, 
1982). Nevertheless, the consistent 
long-term decline in coho abun- 
dance, which has occurred over a 
broad geographic range in habitats 
both degraded and pristine and un- 
der a range of exploitation levels 
(Weitkamp et ah, 1995; Northcote 
and Atagi, 1997), indicates that 
changes occurring in the marine 
environment are also having a ma- 
jor negative influence on coho 
salmon populations. This seems es- 
pecially plausible because in some 
regions the recent decline in coho 
abundance coincides with a stable 
number or increase in abundance of 
other heavily freshwater-dependent 
salmonids, such as sockeye salmon 
and stream-type chinook salmon 
(Northcote and Atagi, 1997). Thus, 
coho salmon populations may be 
less amenable to rebuilding through 
local or regional habitat improvement 
and harvest management than popu- 
lations of other species. 
The increasing conservation con- 
cerns for coho salmon have led to 
substantial efforts to quantify 
