660 



Fishery Bulletin 94(4), 1996 



Cumulative growth during the post-smolt year 

 appeared greater in the Penobscot stock. The post- 

 smolt growth zone circuli counts were significantly 

 higher in the Penobscot stock (P<0.01). Post-smolt 

 growth zone lengths ranged from 0.913 to 1.577 mm 

 for both Penobscot and Connecticut origin fish (Fig. 

 5E). Significant differences between Penobscot and 

 Connecticut spring growth-zone means were found 

 for all 14 smolt years, and of these, 12 were years in 

 which the Penobscot stock had a greater mean than 

 the Connecticut stock (Table 3). 



Annual variation in circuli spacing was similar for 

 two stocks for one of the three post-smolt growth 

 zones. The correlation coefficient for winter spacing 

 indices was significant, whereas the correlation co- 

 efficient for spring and summer indices was not sig- 

 nificant (Table 4). 



Discussion 



The Penobscot and Connecticut stocks exhibited simi- 

 lar interannual patterns of 2SW survivorship despite 

 significant differences between the two stocks in 

 magnitude of survival rate and maturation fraction. 

 Therefore, the sets of conditions affecting the two 



Table 4 



Correlation coefficients (r) between Penobscot and Con- 

 necticut river circuli spacing seasonal growth indices. 

 Bolding indicates significance at P<0.05, n=14. 



Seasonal growth index 



P-level 



Spring 



Summer 



Winter 



-0.263 



-0.018 



0.535 



0.364 

 0.950 

 0.049 



stocks during the post-smolt year must to some de- 

 gree intersect to account for the similarity in the 

 survival-rate time series. At the same time, however, 

 there must also be significant differences in these 

 condition sets to account for the significantly lower 

 survival and maturation fraction found for the Con- 

 necticut stock. We have found it instructive to con- 

 sider those factors that appear to be similar and those 

 that appear to be different for the two stocks. 



The scales from Penobscot and Connecticut fish 

 returns had similar freshwater zone lengths, circuli 

 spacing indices in the spring and winter, and a cor- 

 related pattern of annual variability between win- 

 ter growth indices. These similarities should be in- 

 terpreted in respect to the correlation in annual 2S W 

 survivorship. The survivorship of North American 

 multiseawinter stocks has been related to ocean con- 

 ditions in the Labrador Sea during winter (Friedland 

 et al., 1993). It has been hypothesized that stocks 

 emanating from various rivers in North America in- 

 termix and are acted upon by a common set of sur- 

 vival conditions during the winter season. The co- 

 herence of the survival time series for the Penobscot 

 and Connecticut stocks, as well as for other North 

 American stocks, is considered evidence that stocks 

 converge on the same overwintering ground. Scale 

 analysis shows that annual variation in growth in- 

 dices during winter were correlated, which would be 

 consistent with the assertion that these fish were in 

 the same location during that season. The lack of 

 correlation between spring and summer indices sug- 

 gests that the fish from the respective stocks were 

 acted upon by differing conditions, possibly while 

 they resided in different areas. These data suggest 

 that fish from the Penobscot River are not fully mixed 

 with fish from the Connecticut River until the win- 

 ter season, thus supporting the survival hypothesis 

 for all North American stocks in general. 



Though smolt sizes, as inferred by the freshwater 

 zone lengths of the scales, were similar for the two 

 stocks, we can not conclusively eliminate freshwa- 

 ter effects from contributing to observed differences 

 in survival. There is a large body of evidence that 



