BEACHAM and STARR: POPULATION BIOLOGY OF CHUM SALMON 



environmental factors during the final ocean 

 year strongly influenced length at maturity of 

 chum salmon from Olsen Creek, Alaska. The 

 causes of the different results of the two studies 

 are uncertain, but may in some way be related to 

 the timing of the attainment of a threshold length 

 for spawning. 



The fecundity of Fraser River chum salmon 

 (3,250 eggs/female) reported in the present study 

 was greater than that reported by Foerster and 

 Pritchard (1936) for Fraser River chum salmon 

 (2,943 eggs/ female), for Nile Creek chum salmon 

 (2,726 eggs/female) (Neave 1953), and for Hook- 

 nose Creek chum salmon (2,083-3,097 eggs/fe- 

 male) (Hunter 1959). Only a few of the mean 

 fecundities listed by Bakkala (1970) for North 

 American and Asian chum salmon were greater 

 than that of Fraser River chum salmon. How- 

 ever, size and age compositions of chum salmon 

 sampled for fecundity in the former studies were 

 unavailable, and it may be that larger sized 

 chum salmon were sampled in the Fraser than in 

 other areas because smaller females may have 

 avoided the test fishery. 



In chum salmon, males have been reported to 

 predominate in the early part of the run and fe- 

 males in the later part (Gilbert 1922; Henry 

 1954). In the present study, sex ratios as mea- 

 sured by the test fishery remained relatively con- 

 stant during the fall upriver migration. This 

 result was probably due to differences in run 

 timing of stocks in the Fraser River, so that 

 stocks in different stages of completeness of the 

 run were sampled at the same time, and thus 

 temporal shifts in sex ratios may have been ob- 

 scured. 



Freshwater and marine survival of Fraser 

 River chum salmon have varied about sixfold 

 and fivefold, respectively. Rainfall and gravel 

 permeability have been implicated in variable 

 freshwater survival elsewhere, with higher rain- 

 fall in the fall (except in flood years) and looser, 

 more permeable gravel resulting in higher sur- 

 vival (Wickett 1958). Freshwater survival of 

 chum salmon in Hooknose Creek was inversely 

 related to the total number of pink and chum salm- 

 on eggs deposited (Hunter 1959). The present 

 study indicated that freshwater survival of 

 Fraser River chum salmon was inversely related 

 to the amount of winter rainfall, and that much 

 of the variability in freshwater survival was at- 

 tributable to interactions among temperature, 

 rainfall, and egg abundance. 



Mortality among young fry has been suggested 



to be a major influence in determining the abun- 

 dance of returning adults from a brood year of 

 pink or chum salmon (Neave 1953; Hunter 1959; 

 Parker 1965). Egg-to-fry survival in Fraser 

 River chum salmon was largely dependent upon 

 physical environmental fluctuations, whereas 

 fry-to-adult survival may be dependent upon 

 chum and pink salmon abundance. The effects of 

 favorable or unfavorable environmental condi- 

 tions during incubation appear to be compen- 

 sated for by density-dependent responses of sur- 

 vival during the marine life history stage of 

 chum salmon. Density-dependent survival dur- 

 ing the marine residence period has been sug- 

 gested for several Oncorhynchus species, with 

 possible interactions within and among brood 

 years (Peterman 1978). The present study indi- 

 cated that for even brood years of Fraser River 

 chum salmon, marine survival may be inversely 

 associated with the abundance of the previous 

 brood year, which suggests that the number of 

 returns from a brood year is not determined until 

 the mixing of underyearlings and older chum 

 salmon in the ocean. Although early fry mortal- 

 ity is undoubtedly heavy in chum salmon, as it is 

 in other marine fish, and although it has been 

 suggested that the determination of year-class 

 abundance occurs soon after hatching in marine 

 fishes (Cushing and Harris 1973), some evidence 

 does suggest that year-class abundance is not de- 

 termined in the first year of life of marine fishes 

 (Ponomarenko 1973; Lett et al. 1975). 



The present study indicated that the mean age 

 of returns of a brood year increased with brood 

 year abundance. Similar observations have been 

 reported by Birman (1951) for chum salmon in 

 the Amur River in the Soviet Union and Helle 

 (1979) for chum salmon in Olsen Creek in Alaska. 

 This result implies that growth during ocean 

 residence is density-dependent, as has been re- 

 ported in other marine fishes (Sonina 1965; Palo- 

 heimo and Kohler 1968; Templeman etal. 1978). 

 Competition for food may be one of the mecha- 

 nisms of this density-dependence. The present 

 study also indicated that higher marine water 

 temperatures were accompanied by increased 

 growth rates, as indicated by annual variability 

 in size of returning 4-yr-old chum salmon. 



The present study suggests that there may be 

 competition between chum and pink salmon fry 

 in the Fraser River estuary or Strait of Georgia. 

 Phillips and Barraclough (1978) found that chum 

 salmon fry in the Strait of Georgia near the Fra- 

 ser estuary were larger in 1967 and 1969 when 



823 



