Farley et al.: Early marine growth in relation to marine-stage survival rates for Oncorhynchus nerka 



127 



coastal domain along the Alaska Peninsula. During 

 2002. 75% of the total catch of juvenile sockeye salmon 

 occurred north of 57°N, with 50% of the total catch 

 occurring north of 58°N within the shallow stratified 

 waters near the northern coastal domain. 



Average fork length of juvenile sockeye salmon was 

 significantly smaller during 2001 than during 2000 and 

 2002 for growth rates greater than 0.3 mm/day (^-test; 

 P<0.01) and not significantly different from 2000 (Mest; 

 P=0.05) for growth rates equal to mm/day (Table 3). 

 The rank order of juvenile sockeye salmon fork lengths 

 was the same for all growth rates, and the largest fish 

 taken in 2002 and the smallest, in 2001. For all three 

 growth rates, average fork length was significantly 

 larger during 2002 than during 2000 and 2001 (^test; 

 P<0.01). 



The marine-stage survival rate and abundance in 

 the indices mirrored the observed variation in fish fork 

 length; they were highest during 2002 and at or near 

 their lowest during 2001 (Table 4). In addition, the 

 nearshore distribution of juvenile sockeye salmon (2001; 

 Fig. 3) appeared to coincide with lower indices of abun- 

 dance and marine-stage survival rate, whereas fish 

 distributed in the northern area of our survey (2002; 

 Fig. 3) exhibited higher marine-stage survival rate and 

 abundance. 



Discussion 



Our study indicates that the size of Bristol Bay sockeye 

 salmon after their first year at sea is not directly related 

 to their survival, when size is measured from growth 

 rings on the scales of adults returning to the Egegik 

 and Kvichak rivers. Analyses of the MSWl growth index 

 indicated that most of the time series had a constant 

 mean and variance. Similar studies where adult scales 



Table 3 



Average fork length of juvenile sockeye salmon (Onco- 

 rhynchus nerka) collected along the eastern Bering Sea 

 during 2000, 2001. and 2002. Daily growth rate ( mm ) was 

 assumed to be 0, 0.30, and 1.7. Statistics include sample 

 size (n), average fork lengths, and standard deviation 

 (SDl of the original length data. 



2000 

 2001 

 2002 



834 



802 



1936 



174.77 

 171.91 

 197.96 



176.53 

 170.61 

 197.94 



184.76 19.99 

 164.53 35.55 

 197.83 34.74 



from Atlantic (Salmo salar L.; Crozier and Kennedy, 

 19991, coho (O. kisutch; Briscoe, 2004), and chum (O. 

 keta\ Helle, 1979) salmon were used to measure growth 

 (size) of salmon during their first year at sea revealed 

 that the survival rate of a cohort was statistically unre- 

 lated to variation in growth (size) of the salmon. The 

 relative uniformity in the size of salmon after their first 

 year at sea and the lack of a relationship between size 

 and survival rate is contrary to the prevailing paradigm 

 that the size achieved by fish after their first summer 

 at sea is important to survival (Beamish and Mahn- 

 ken, 2001). However, these results do not necessarily 

 invalidate this paradigm; the adult scale samples avail- 

 able for analysis may reflect only those juvenile salmon 

 that had attained sufficient size in order to survive to 

 adulthood, and not those that died at sea (Crozier and 

 Kennedy, 1999). 



In support of this possibilty, when we directly mea- 

 sured the fork length of juvenile sockeye salmon (Tables 

 3 and 4) during late summer and early fall surveys 

 along the eastern Bering Sea shelf (2000-02), smaller 

 fish had lower indices of marine-stage survival rate. 

 This result is consistent with that from other studies of 

 teleost fish, where larger individuals gained a survival 

 advantage over smaller conspecifics during the juvenile 

 life-history stage (Parker, 1968; Healey, 1982; Holtby et 



