FISHERY BULLETIN: VOL. 86, NO. 4 



ocean growth, particularly in the year of maturity. 



Body shape of Yukon River and British Colum- 

 bia chum salmon was different, with Yukon River 

 chum salmon having a shorter head, thinner caudal 

 peduncle, and smaller fins than British Columbia 

 chum salmon. This more fusiform body shape is pre- 

 sumably an adaptation to the long migration in fresh 

 water, as selection should result in a body shape that 

 would minimize energy consumption during migra- 

 tion. Morphological differentiation with respect to 

 distance of upstream migration has been reported 

 to occur in other Oncorhynchus species (Eniutina 

 1954; Taylor and McPhail 1985), as well as with 

 respect to river size (Hjort and Schreck 1982; 

 Beacham and Murray 1987). 



The mean fecundity of 2,325 eggs per female for 

 Yukon River fall chum salmon reported in our study 

 is similar to other results. Elson (1975) reported 

 mean fecundities of 2,360 eggs and 2,513 eggs per 

 female for Porcupine River chum salmon sampled 

 in 1971 and 1973, respectively. Raymond (1981) 

 reported mean fecundities of Tanana River chum 

 salmon of 2,355 eggs in 1977 and 2,762 eggs in 1978. 

 Fecundities of Yukon River fall chum salmon are 

 less than those reported for many chum salmon 

 stocks in British Columbia (Beacham 1982), and also 

 less than fall chum salmon in the Amur River (3,200 

 to 4,300 eggs) (Smirnov 1975). Mean egg size of 

 Yukon River fall chum salmon is also less than that 

 of Amur River chum salmon (180 to 300 mg, 6.7 to 

 9.0 mm diameter) (Smirnov 1975). 



The different fecundities and ages at maturity of 

 Yukon River and British Columbia chum salmon 

 present an interesting contrast in life history char- 

 acters. Yukon River fall chum salmon mature at an 

 average age of 0.28 years older than British Colum- 

 bia salmon, which means that they incur an addi- 

 tional 5% mortality if the instantaneous mortality 

 rate during the last year of life for chum salmon is 

 0.013 per month (Ricker 1976). The lower fecundity 

 and older age at maturity of Yukon River salmon 

 indicate that they are not as productive as chum 

 salmon in British Columbia or that mean survival 

 rates of the two groups are not equivalent. Egg-to- 

 fry survival rates for Yukon River chum salmon 

 have been reported as a mean of about 2.5% (Buklis 

 and Barton 1984), whereas those for British Colum- 

 bia chum salmon average about 10% (Bakkala 1970; 

 Beacham and Starr 1982). If Yukon River chum 

 salmon are as productive as those in British Colum- 

 bia, then ocean survival rates of Yukon River chum 

 salmon must be higher than those of British Colum- 

 bia chum salmon. 



When incubated under the same water tempera- 

 tures, Yukon River chum salmon alevins hatch and 

 the fry emerge sooner than most chum salmon 

 populations in British Columbia (Beacham and Mur- 

 ray 1987). The faster development rates presumably 

 occur as a response to lower water temperatures 

 during the winter in the Yukon River tributaries 

 than in rivers in British Columbia. Yukon River 

 chum salmon alevins and fry are shorter and lighter 

 than those from British Columbia (Beacham and 

 Murray 1987), presumably reflective of smaller ini- 

 tial egg size of Yukon River chum salmon. At incu- 

 bation temperatures of 4°, 8°, and 12 °C, maximum 

 alevin and fry size for Yukon River chum salmon 

 was observed at 4°C, but for British Columbia 

 stocks, maximum alevin and fry size was generally 

 observed at 8°C. These results suggest that Yukon 

 River chum salmon are better adapted for develop- 

 ment under low water temperatures than are British 

 Columbia chum salmon. 



Yukon River chum salmon are generally distinc- 

 tive in electrophoretic characteristics from chum 

 salmon in Cook Inlet in Alaska (Okazaki 1981) and 

 British Columbia (Okazaki 1981; Beacham et al. 

 1985, 1987). For example, the allelic frequency of 

 Idh-3^^ is 0.28 in Queen Charlotte Islands popula- 

 tions and 0.17 in populations in northern British 

 Columbia (Beacham et al. 1987), but this allele was 

 not detected in our study of Yukon River chum 

 salmon. Heterozygosity of Yukon River chum 

 salmon was lower than that observed for British 

 Columbia salmon (the same loci were included in the 

 analysis) (Beacham et al. 1987). Kijima and Fujio 

 (1984) reported that average heterozygosity is 

 related to effective population size in Japanese chum 

 salmon populations, with more abundant populations 

 having increased genetic variance. Abundance of the 

 Yukon River populations examined in our study is 

 unknown, but the catch data suggest that the in- 

 dividual Yukon River populations may not be as 

 abundant as major chum salmon populations in 

 British Columbia. 



Allelic frequencies for most salmon populations 

 are reported to show little annual variation (Grant 

 et al. 1980; Utter et al. 1980; Beacham et al. 1985, 

 1987), allowing for pooling of samples from a par- 

 ticular population over several years. It should thus 

 not be necessary to conduct annual sampling in 

 order to characterize the populations contributing 

 to fisheries. Stock identification based on stable 

 traits, such as allelic frequencies, reduces annual 

 sampling costs for the baseline stocks. This differs 

 from scale analysis, in which variation in the char- 



672 



