658 



Fishery Bulletin 94(4). 1996 



growth zone length was taken as the distance from 

 the last circulus in the freshwater zone to the outer 

 circulus of the third circuli pair of the winter zone 

 measured along the 360° axis of the scale (Fig. 2). 

 This distance reflects the average spacing of circuli 

 and the number of circuli deposited during the post- 

 smolt period. 



Comparison of post-smolt growth 



Circuli spacing patterns and lengths of the post-smolt 

 growth zone for both the Penobscot and Connecticut 

 stocks were compared by analysis of variance. Cir- 

 culi spacings for the spring, summer, and winter 

 zones were compared by means of two-way analysis 

 of variance with year and stock as factors. The growth 

 zone of an individual fish was represented by the 

 mean circuli spacing of the growth-zone circuli pairs 

 identified in the pattern of mean spacings for all fish 

 of that smolt year and stock. In all cases, the inter- 

 action terms for these ANOVA's were significant and 

 thus the two-way ANOVA results are not reported. 

 Instead, circuli spacings of both stocks were com- 

 pared with one-way analysis of variance for each 

 smolt year. The length of the post-smolt growth zone 

 for both stocks was also compared for each year with 

 a series of one-way ANOVA's. 



The annual variation in circuli spacing growth- 

 zone indices was compared by using Pearson prod- 

 uct-moment correlation. The correlation coefficients 

 were computed by using the seasonal growth-index 

 means for the two stocks. 



Results 



Return rate and cohort age at maturity 



Return rates for both 1SW and 2SW salmon 

 were higher in the Penobscot River than in the 

 Connecticut River. The return rate for 1SW 

 salmon averaged 0.97%r and ranged from 0.36 

 to 2.20%f for the Penobscot stock, whereas for 

 the Connecticut stock the return rate averaged 

 0.03% c and ranged from 0.0 to 0.127« (Table 1). 

 For 2SW salmon, the return rate for the 

 Penobscot stock averaged 4.89%c and ranged 

 from 1.98 to 12.69%f , whereas in the Connecti- 

 cut the return rate averaged 1.40%c and ranged 

 from 0.28 to 5.11%r. Return rates were signifi- 

 cantly different between stocks in a paired t- 

 test (P<0.01 for both 1SW and 2SW rates). De- 

 spite the differences in the level of 2SW return 

 rates between the two stocks, the pattern of 

 annual variation in 2SW return rates of the two 

 stocks was highly correlated (r=0.88, P<0.01). 

 Peaks in return rate that occurred with the 1978 

 and 1983 smolt classes are seen in both time 

 series (Fig. 4A). The relation between the two 

 time series is seen clearly in the normalized or Z- 

 transformed data (Fig. 4B). The transform is simply 

 the observation minus the time-series mean divided 

 by the time-series standard deviation. 



The 1SW fraction for the Penobscot, which aver- 

 aged 11% during the study period, was an order of 

 magnitude higher than the fraction observed for the 

 Connecticut River (Table 1). The difference between 

 stocks was highly significant (paired £-test, P<0.01). 



Smolt size and age 



Smolt sizes of 2SW returns, as indicated by the 

 lengths of the freshwater zone, showed no system- 

 atic differences between stocks. Mean freshwater 

 zone lengths ranged from 1.000 to 1.243 mm for both 

 stocks (Fig. 5A). Significant differences between 

 freshwater zone lengths were found for 6 of the 14 

 smolt classes, but of those 6 years, there were equal 

 numbers of years with greater means in the 

 Penobscot and Connecticut stocks (Table 3). 



Circuli spacing patterns and post-smolt 

 growth 



Considerable variation in circuli spacing was evident 

 between stocks and among years. Annual mean cir- 

 culi spacing indices had an approximate range of 0.04 

 to 0.070 mm (Fig. 5, B-D). Spring and summer spac- 

 ing indices averaged greater than 0.060 mm for both 



