Friedland et al.: Post-smolt growth, maturation, and survival of two stocks of Atlantic salmon 



661 



illustrates the influence of smolt-size and condition 

 and stocking circumstances on the return rates and 

 yields of salmon stocks (Ward et al., 1989; Hvidsten 

 and Johnsen, 1993; Farmer, 1994; Lundqvist et al., 

 1994). We lack smolt size information on the fish that 

 did not survive; they may have had a different fresh- 

 water zone length-size frequency than that observed 

 for the survivors. Smolt age composition was similar 

 each year and appears unrelated to the differences 

 between stocks. However, the trend in smolt age com- 

 position is not similar to the time-series trends in 

 survival and maturity for the two stocks and should 

 be considered along with other factors when evalu- 

 ating long-term changes in stock performance 

 (Friedland et al., 1993). 



Summer circuli spacing and post-smolt growth- 

 zone length data suggest that the Penobscot fish grew 

 faster than the Connecticut fish as post-smolts. These 

 differences may have influenced survivorship and 

 maturation. These data suggest that fish from the 

 Connecticut stock grew more slowly during the post- 

 smolt year than did Penobscot fish, matured as 1SW 

 less frequently, and were more vulnerable to mortal- 

 ity. It is generally accepted that larger individuals 

 are less vulnerable to predation (Peterson and 

 Wroblewski, 1984; McGurk, 1986; Anderson, 1988; 

 Miller et al., 1988; Pepin, 1991; L'Abee-Lund et al., 

 1993). Therefore, smaller, slower-growing Connecti- 

 cut salmon may be vulnerable to a wider range of 

 potential sources of mortality. For example, under 

 these conditions, we would predict that Connecticut 

 post-smolts would be vulnerable to predators for a 

 longer period of the post-smolt year than would 

 Penobscot post-smolts. Likewise, smaller post-smolts 

 may not effectively compete with other predators for 

 prey or may experience ontogenetic mismatches with 

 prey resources that are also growing or that may be 

 transient in post-smolt habitats (Brodeur, 1991; 

 Healey, 1991; Levings, 1994). 



Sea-age at maturation is partly a growth-related 

 phenomenon associated with the seasonal accumu- 

 lation of lipid stores (Rowe et al., 1991; Thorpe, 1994). 

 In general, maturation at a certain age has been as- 

 sociated with individual growth rate and other causal 

 effects (Aim, 1959; Svedang, 1991; Thorpe, 1994); 

 however, the effect of sea growth on maturation in 

 salmon has not always been obvious (Power, 1986; 

 Randall et al., 1986; Myers and Hutchings, 1987). 

 Using sea ranching and cage culture experiments 

 with the same genetic stock of salmon, Saunders et 

 al. ( 1983 ) reported evidence supporting a hypothesis 

 that first seawinter temperature minima are criti- 

 cal determinants of maturation in salmon. This work 

 was further supported by Herbinger and Newkirk 

 (1987) who described a relationship between 1SW 



maturation and favorable (or perhaps minimum) 

 winter growth. However, the specificity of seasonal 

 growth effects can be challenged by other experimen- 

 tal evidence that shows that spring growth can in- 

 fluence 1SW maturation (Thorpe et al., 1990). In an 

 analysis of the Penobscot stock, Friedland and Haas 

 (1996) showed that maturation fraction varies with 

 summer growth rate of the cohort as indicated by 

 circuli spacing indices for 2SW returns. This finding 

 is further supported by our comparison of Penobscot 

 and Connecticut fish which shows that the stock with 

 the greater summer growth had produced a higher 

 percentage of mature 1SW fish. 



How an environmentally driven maturation 

 mechanism would optimize age at maturation is not 

 clear. For salmon stocks with complex maturation 

 age structures, early maturing fish (1SW) are pre- 

 dominantly males; egg-producing females more fre- 

 quently mature at a later age when their egg pro- 

 duction is maximized. Therefore, a shift in age at 

 maturity allows a brood class to receive genes from 

 more than one spawning cohort without significant 

 loss of egg production. This plasticity in spawning 

 age ensures that genes move within the population 

 while remaining robust to environmental effects on 

 the deposition of female gametes, which are limiting 

 (Stearns and Crandall, 1984; Stearns, 1992). How- 

 ever, Atlantic salmon exhibit a wide range of matu- 

 ration age structures (Power, 1981; Saunders, 1981; 

 Saunders and Schom, 1985) suggesting within-popu- 

 lation heterozygosity may also be maintained by pro- 

 tracted freshwater residency that allows many brood 

 years to contribute to a smolt run in a given year. 



Genetic influences have also been shown to affect 

 stock-specific patterns of age at maturity (Saunders 

 et al., 1983; Thorpe et al., 1983). However, genetic 

 factors are unlikely to explain the differences be- 

 tween the Penobscot and Connecticut stocks because 

 the Connecticut stock is derived predominantly from 

 the Penobscot gene pool (Rideout and Stolte, 1988). 

 When the Connecticut River broodstock was devel- 

 oped, gametes from Canadian and U.S. origin 

 ( Penobscot River broodstock ) donor stocks were used. 

 However, the crosses with Canadian genetic sources, 

 as demonstrated with tagging, produced virtually no 

 progeny; therefore it can be concluded that the Con- 

 necticut broodstock is principally derived from the 

 Penobscot River broodstock. 



The systematic differences in growth, survival, and 

 maturation between these two stocks may be related 

 to their post-smolt migrations. When salmon first 

 enter the marine environment, they move by active 

 and passive mechanisms (Jonsson et al., 1993). Be- 

 cause of the differences in the timing of the smolt 

 migration, the starting point of the post-smolt feed- 



