244 



Fishery Bulletin 101(2) 



September and October (DFO'^). However, the adequacy of 

 managing Fraser River chinook salmon based upon run tim- 

 ing is currently under review (Candy et al.'), and changes 

 in management objectives are likely. During their migra- 

 tion through the lower river, chinook salmon are exploited 

 by several distinct fisheries, among them commercial net 

 fisheries, and the recreational fishery. 



Conservation concerns for specific populations requires 

 more information on timing of returns beyond the present 

 designations of "spring," "summer," and "fall." Timing of 

 some specific populations through the lower river can be 

 inferred from their arrival on the spawning grounds. For 

 example, chinook from the Birkenhead River (about 300 

 km from the Fraser River mouth) have historically sup- 

 ported a local recreational fishery on that river in April 

 and early May (Fraser et al., 1982), indicating a very early 

 timing of passage through the lower Fraser River. Coded- 

 wire tagging (CWT) has been conducted on some enhanced 

 populations, but these populations have been limited in 

 scope given the size of the drainage, and coded-wire tag 

 returns within the Fraser River have provided limited 

 information on the timing of population returns. 



Stock or population identification of chinook salmon 

 migrating through the lower river is a continuing issue of 

 management concern, and until recently there has been no 

 effective way to provide estimates of population composi- 

 tion in the detail required by fishery managers. Although 

 allozyme-based methods of stock identication have proven 

 useful in estimation of chinook salmon stock composition 

 in mixed-stock fisheries (Shaklee et al., 1999), the level of 

 population discrimination available in the Fraser River was 

 not sufficient for population-specific applications. Chinook 

 salmon returning to spawn in specific rivers were consid- 

 ered as populations in our analysis, whereas stocks were a 

 collection of populations from a particular geographic area 

 or management unit. Minisatellite variation has been very 

 effective in discriminating among individual Fraser River 

 populations (Beacham et al., 1996), but we considered this 

 technology not practical because of the cost and the time 

 required for laboratory analysis. Any new technology em- 

 ployed had to meet these criteria: rapidity of analysis, mod- 

 erate cost, and ability to discriminate among populations. 

 Because variation in microsatellite loci meets these criteria 

 and has been applied to other species requiring discrimina- 

 tion among salmonid populations within watersheds (Small 

 et al., 1998; Beacham and Wood, 1999; Beacham et al., 2000), 



' DFO (Department of Fisheries and Oceans). 1995. Fraser 

 River chinook salmon. Fraser River Action Plan, Fishery 

 Management Group, Vancouver, British Columbia, Canada, 24 

 p. (Available from Fi.sheries and Oceans, 5.55 West Hasting St., 

 Suite 1220, Vancouver, Firitish Columbia, Canada V6B 5G3. 



•^ DFO(DepartmcntofTisheriesandOcoansl. 1999. FraserRiver 

 chinook salmon DFO Science Stock Status Report D6-11, 7 p. 

 [Available from Fisheries and Oceans, 555 West Hasting St., 

 Suite 1220, Vancouver, British Columbia. Canada V6B 5G3.J 



' Candy, J R , .J R. Irvine, C K. Farken. S. L. Lemkc, R. E. Bailey, 

 M. Wetklo, and K. .lohsen. 2002. A discussion paper on pos- 

 sible new stock groupings (conservation units) for Fraser River 

 chinook salmon. Can .Stock Assessment Secretariat. Res. Doc. 

 2002/85, 57 p. lAvailablc Irom hUp://www.(if'o-mpo.gc.ca/Csas/ 

 English/Index_e.htm.| 



and has been shown to be useful in stock discrimination in 

 chinook salmon (Banks et al., 2000 ), we chose to survey vari- 

 ation at microsatellite loci for Fraser River chinook salmon 

 (Beacham et al., 2003) and evaluate and apply the variation 

 to practical problems of stock identification. 



In the current study, we surveyed variation at 13 mi- 

 crosatellite loci and evaluated the utility of using micro- 

 satellite variation for stock identification of Fraser River 

 chinook salmon. These procedures were accomplished by 

 analysis of simulated mixtures and application to a sample 

 of chinook salmon that had previously been marked with 

 coded-wire tags. We evaluate the accuracy of identifying 

 individuals to population and region of origin. We also es- 

 timate stock compositions from fisheries in the lower river 

 to determine the timing and relative abundance of specific 

 populations through the fishery. 



Materials and methods 



Collection of baseline DNA samples and 

 laboratory analysis 



Genomic DNA was extracted from either liver, scales, oper- 

 culum punches, or fin clips from chinook salmon sampled 

 between 1987 and 1998 by using the phenol-chloroform pro- 

 tocol of Miller et al. (1996) (early extractions) or the chelex 

 resin protocol of Small et al. ( 1998) (later extractions) (Table 

 1, Fig. 1). Samples were derived from adults in all areas 

 except the McGregor River, where because of the difficulty 

 of obtaining adults, juveniles were sampled. For the survey 

 of baseline populations, PCR products at six microsatel- 

 lite loci— O/s;00. OtslOl. Otsl02. Otsl04. Otsl07 (prim- 

 ers outlined by Nelson and Beacham, 1999) and Ssal97 

 (O'Reilly et al., 1996) — were initially size-fractionated on 

 nondenaturing polyacrylamide gels by staining with 0.5 

 mg/mL ethidium bromide in water and illuminating with 

 ultraviolet light. Nelson et al. ( 1998) have provided a more 

 complete description of gel electrophoretic conditions. Three 

 20-bp marker lanes were run on each gel, and the size of 

 the amplified alleles was determined from the molecular 

 size grid created with the 20-bp markers. Beacham and 

 Wood (1999) have provided a more complete description of 

 the methods used to identify alleles with this technology. 

 With the acquistion of an automated sequencer (the ABI 

 377) in our laboratory, PCR products at seven additional 

 loci— 0^o2, Ogo4 (Olsen et al., 1998), Oke4 (Buchholz et 

 al., 1999), Omy325 (O'Connell et al., 1997), OkilOO (K. M. 

 Miller, unpubl. data), and Ots2. Ots9 (Banks et al.. 1999)— 

 were size-fractionated on denaturing polyacrylamide gels 

 and allele sizes were determined with Genescan software 

 (PE Brosystems, Foster City, CA). The six loci previously 

 analyzed on nondenaturing polyacrylamide gels stained 

 with ethidium bromide were subsequently analyzed on the 

 automated sequencer when it became available. 



Collection of fishery samples 



In 199.5, samples were collected from a daily gillnet test 

 fishery at Albion in the lower Fraser River in southern 



