FISHERY BULLETIN: VOL. 80, NO. 1 



Table 2.— Means, standard errors (in parentheses), and ranges for morphological characters of the 1977 brood 

 year hatchery samples of juvenile coho salmon and the hatchery water incubation temperatures for the first 

 month of incubation. Sample sizes were 15. The data are listed in north to south order of the sampling location. 



result of selection, several interacting selective 

 forces were involved. 



Life History Characters 



Earlier peak spawning times (Table 7) were 

 strongly associated with the northern stream 

 systems and with stream systems having large 

 estuaries (Table 6). However, the correlation of 

 peak spawning time with size of estuary may be 

 biased by the large number of samples from 

 Columbia River hatcheries: spawning times of 

 stocks from the Columbia River are earlier than 

 those of coastal stocks, and the Columbia River 

 has a large estuary. Selection for earlier spawn- 

 ing times through hatchery practices may be the 

 cause for the differences in spawning times be- 

 tween hatchery and wild stocks in the North 

 Nehalem, Trask, and Alsea Rivers. Selection for 

 earlier spawning times has been observed in a 

 steelhead trout hatchery program (Millenbach 

 1973). At hatcheries using wild stocks as sources 

 for eggs and sperm, peak spawning times were 

 similar to those of naturally spawning fish in 

 the respective stream system. 



The proportion of females (Table 7) ap- 

 peared to be higher in the southern stream sys- 



tems, suggesting that jacks were more common 

 there. The effective sex ratio, including jacks, at 

 the time of spawning should be close to 1:1 

 (Fisher 1930). If only 3-yr-old males and females 

 are counted, the proportion of females should be 

 X).50, the margin above 0.50 depending on how 

 many jacks returned in the previous year. How- 

 ever, the proportion of males was higher than 

 that of females in stocks from the Quilcene, 

 Quinault, Sandy, North Nehalem, Nehalem, 

 Trask, Salmon, Alsea, Umpqua, and Rogue 

 Rivers. Nikolskii (1969) reviewed several pos- 

 sible causes for sex ratios departing from 1:1; 

 however, the reason for the high proportion of 

 males in these stocks is not known. 



Isozyme Gene Frequencies 



Transferrin gene frequencies (Figs. 2, 3), cor- 

 related significantly with six of the stream 

 characters (Table 6). The best model from step- 

 wise multiple regression explained only 68% of 

 the variation in gene frequencies. Analysis of the 

 relationships of the "A" allele frequencies with 

 basin area (Fig. 4) and latitude (Fig. 5) explained 

 the variation more simply than did the stepwise 

 regression model. These correlations showed 



110 



