FISHERY BULLETIN: VOL. 80, NO. 4 



1970-79 indicated that the proportion of age-5 

 fish in the run decreased from October through 

 December, whereas the proportion of age-3 fish 

 increased (Table 1). Age-4 chum salmon com- 

 prised about 74% of the run in each month, while 

 age-3 comprised 24% of the run in December but 

 only 14% in October. 



Sex ratios were variable among ages, with 

 more male chum salmon returning at age 3 (X 2 = 

 23.7, df = 1, P<0.01) and age 5 (X 2 = 10.7, 

 P<0.01 ) than did females (Table 2). More female 

 chum salmon returned at age 4 than did males 

 (X 2 = 70.6, P<0.01). Since the mean age of chum 

 salmon in the run decreased through time, and 

 sex ratios vary with age, the sex ratio of the run 

 may also vary through time. However, an appli- 

 cation of the sex ratios by age in Table 2 to the age 

 compositions in Table 1 shows that sex ratios of 

 the total run were nearly the same each month 

 (52.9% females in October, 53.1% in November, 

 and 52.9% in December). 



Table 1.— Percentage age composition by month (Octo- 

 ber-December) of chum salmon sampled in the Fraser 

 River test fishery, 1970-79. Sample sizes are in paren- 

 theses. 



Age 



October 



November 



December 



Table 2.— Sex ratios (males:females) of age-3, 

 -4, and -5 chum salmon sampled in the Fraser 

 River test fishery, 1960-79. 



Age: 3 4 5 



Ratio 

 Sample size 



1.19 

 3,172 



0.79 

 5.378 



1.30 

 626 



Fecundity 



Fecundity was determined from samples of 

 chum salmon taken in 1966, 1968, and 1978 with 

 females ranging in postorbital-hypural length 

 from 50 to 66 cm. Females sampled in 1978 were 

 generally larger than those in 1966 and 1968 

 (Table 3). A two-way analysis of covariance with 

 sampling year and age as factors and length as a 

 co-variate (co-variate was accounted for before 

 factors were tested) indicated that there was no 

 significant difference in fecundity (F) among 

 years (F= 0.92, df = 2 and 228, P>0.05) or among 



ages (F = 2.66, P>0.05). The mean fecundity for 

 all samples was 3,250 eggs/female (Table 3). 



The relationship between fecundity and length 

 of 234 females sampled was described by: 



loge F = 2.8659 + 0.8193 log e L (r = 0.29) (1) 



where F = fecundity and L = postorbital-hypural 

 length in millimeters. The regression was signifi- 

 cant (F = 20.1, P<0.01), but accounted for about 

 only 9% of the variability in fecundity (Fig. 2). 

 Most of the females sampled were between 54 

 and 63 cm in length, and with high variability in 

 fecundity within this short length range, little of 

 the variability in fecundity could be accounted 

 for by the regression. 



Table 3. — Number of females sampled, mean length (cm), 

 mean age (yr), and mean fecundity of chum salmon from the 

 Fraser River. SE = standard error of mean. 



Year: 



1966 



1968 



1978 



Total 



Fry Migrations and Survival 



The major portion of the downstream migra- 

 tion of chum salmon fry passed the Mission City 

 sampling site from mid-March to the end of 

 April. Usually 50% of the fry had passed the sam- 

 pling site between April 13th and the 23d (Fig. 

 3), although 50% of the fry have passed the sam- 

 pling site as early as April 3 (1976 brood year) 

 and as late as May 3 ( 1971 brood year). About 80% 

 of the chum salmon fry migrate downstream at a 

 length of <40 mm. However, the proportion of 

 fry >40 mm long during the downstream migra- 

 tion tends to increase with time. By the second 

 week of May 1978, 20% of the fry were >43 mm 

 long and weighed 1.0 g, which suggests a period 

 of freshwater rearing for these fry. 



Linear regression was used to determine the 

 relationship of air temperature (T), determined 

 as the mean of monthly air temperatures at the 

 Abbotsford airport from December through 

 February, to timing of the fry migration (F), 



816 



