FISHERY BULLETIN: VOL. 80, NO. 4 



years. Multiple regression was used to describe 

 the effect of pink salmon on chum salmon fry sur- 

 vivals by: 



£ = a L + h{ D) + c 

 S P 



(10) 



where R 



S 



P 

 D 



total returns from a given brood 

 year 



spawning escapement of chum 

 salmon that produced a given 

 brood year 



spawning excapement of pink salm- 

 on in same brood year 



median time of downstream pink 

 salmon fry migration minus 

 median for chum salmon fry in 

 days 



regression constant. 



The analysis yielded: 

 Variable Coefficient 



SE 



CHUM ESCAPEMENT / PINK ESCAPEMENT 



The regression was significant (F = 9.60, df = 2, 

 and 3, P<0.05) and accounted for 85% of the vari- 

 ation in R/S in odd brood years (R = 0.92). The 

 chum salmon R/S ratio increases the earlier that 

 chum salmon fry migrate downstream relative 

 to pink salmon fry (Table 6). If chum salmon es- 

 capement is 10% of that of pink salmon, then the 

 time of the median downstream migration of 

 chum salmon fry must be at least 9 d earlier 

 than that of pink salmon fry if the chum salmon 

 spawners are just to replace themselves (R/S = 

 1.0) (Fig. 10). However, if chum salmon escape- 



Table 6.— Dates when 50% of the fry were estimated to 

 have migrated downstream and estimated spawning es- 

 capements for odd-numbered brood years, 1965-77. 



Figure 10.— Contour plot for returns/spawners for odd-num- 

 bered brood years of Fraser River chum in relation to timing of 

 pink and chum fry migrations and relative spawning escape- 

 ments of pink and chum. Contour lines of R/S = 1.0, 2.0, 3.0, and 

 4.0 are plotted. 



ment is 30% of the pink salmon escapement and 

 chum salmon fry still have a 9-d advantage over 

 pink salmon fry, then the R/S ratio for chum 

 salmon will be between 3.0 and 4.0, as happened 

 for the 1969 brood year (Fig. 10). The chum 

 salmon R/S ratio will still be between 3.0 and 4.0 

 if the chum salmon escapement is 15% of the pink 

 salmon escapement, provided the chum salmon 

 fry have at least a 19-d advantage over the pink 

 salmon fry, similar to the 1965 brood year (Fig. 

 10). 



DISCUSSION 



Mean lengths at maturation for 4-yr-old chum 

 salmon in the present study were found to be sig- 

 nificantly correlated with oceanic water temper- 

 atures in the penultimate growing season but not 

 in the final one. Helle (1979) found that oceanic 



822 



