122 



Fishery Bulletin 105(1) 



years their migration is farther offshore (Farley et al., 

 2005). We also compared the size of juvenile Bristol Bay 

 sockeye salmon collected during late summer and early 

 fall (2000-02) trawl surveys along the eastern Bering 

 Sea shelf with indices of their abundance, marine stage 

 survival rate after our survey, and returns per spawner 

 from these cohorts. Interannual differences in the size 

 and growth rates of juvenile sockeye salmon were also 

 compared to their early marine distribution and ocean 

 conditions. The specific objectives of this study were to 

 determine whether larger, presumably faster growing, 

 juvenile sockeye salmon in fact had higher survival 

 rates than smaller, presumably slower growing indi- 

 viduals, and what aspects of the marine environment 

 might influence these growth rates. 



Materials and methods 



Data 



Our research focused on Bristol Bay sockeye salmon 

 because this region has the largest returns and com- 

 mercial harvest of sockeye salmon in the world. Scales 

 from adult sockeye salmon, their fork lengths, and 

 data on brood-year return per spawner for the Egegik 

 and Kvichak Rivers in Bristol Bay, Alaska, as well as 

 annual totals of the number of adult Bristol Bay sock- 

 eye salmon returns and spawners, were obtained from 

 the Alaska Department of Fish and Game (ADF&G). 

 Salmon scales are collected annually by ADF&G to 

 estimate the age composition of adult sockeye salmon 

 for fishery management. Age was designated by the 

 European notation, i.e., 0.6, where a = the number of 

 winters spent in freshwater prior to going to sea and b = 

 the number of winters spent in the ocean (Koo, 1962a). 

 Salmon scale collections and brood year return per 

 spawner data were available for the dominant freshwater 

 and ocean age groups of sockeye salmon sampled in the 

 Kvichak River (brood year returns for ages 1.2, 1.3, 2.2, 

 2.3; 1965-97) and the Egegik River (brood year returns 

 for ages 1.3, 2.2, 2.3; 1965-97). 



Scales were selected for measurement by following the 

 procedures described in Ruggerone et al. (2005). Brief- 

 ly, scales were selected when our age determination 

 matched that previously made by ADF&G, the shape 

 of the scale indicated that the scale was from the "pre- 

 ferred area" (below the dorsal fin and above the lateral 

 line — see Koo, 1962b I, and the circuli and annuli were 

 clearly defined and not affected by scale regeneration or 

 significant resorption along the measurement axis. The 

 number of scale samples for each river system and age 

 group are provided in Table 1. 



Scales from adult sockeye salmon were digitized fol- 

 lowing procedures described by Hagen et al.' and Rug- 



1 Hagen, P. T., D. S. Oxman, and B. A. Agler 2001. Devel- 

 oping and deploying a high resolution imaging approach for 

 scale analysis. Doc. 567, p. 11. North Pacific Anadromous 

 Fish Commission, 889 Pender Street, Vancouver, Canada. 



gerone et al. (2005). The scale measurement axis was 

 determined by a perpendicular line drawn from a line 

 intersecting each end of the first saltwater annulus. 

 Distance (mm) between the focus and the outer edge of 

 the scale was designated as the total scale length. The 

 relationship between total scale length and adult fork 

 length was linear for both the Egegik River (F-test, 

 P<0.001; r- = 0.41) and Kvichak River (F-test, P<0.001; 

 r^ = 0.36) sockeye salmon samples. For an index of total 

 growth through the first year at sea, we measured the 

 distance from the focus to the outer edge of the first 

 saltwater growth zone for each fish. A time series of 

 annual means of the individual growth during the first 

 year at sea (MSWl^ ^ ,) estimated for each adult fresh- 

 water age group (o represents 1 or 2) within a river 

 system (i represents Egegik, Kvichak) was used as an 

 index of size that sockeye salmon would have attained 

 after their first year it) at sea. 



The total number of fish caught and the fork lengths 

 (mm) of juvenile sockeye salmon within each trawl haul 

 were recorded during the Bering-Aleutian Salmon In- 

 ternational Survey (BASIS) research cruises along the 

 eastern Bering Sea shelf during fall (August-September) 

 2000 to 2002 (Fig. 1). The surveys were conducted over 

 a broad area of the shelf and over major oceanograph- 

 ic domains (coastal and middle domains; Kinder and 

 Schumacher, 1981) along the eastern Bering Sea shelf. 

 In addition, the surveys were designed to sample the en- 

 tire population of juvenile sockeye salmon from Bristol 

 Bay lake systems to reduce the chance of sample vari- 

 ability that could affect one's ability to interpret results 

 from these samples. Recent descriptions of juvenile 

 salmon migration pathways along the eastern Bering 

 Sea shelf (Farley et al., 2005) and genetic stock com- 

 position indicate that juvenile sockeye salmon collected 

 during the surveys were primarily from Bristol Bay. 



Fish were collected by using a mid water rope trawl 

 (see Farley et al., 2005 for description) rigged to sample 

 the top 15 m of the water column. We attempted to 

 collect scales from juvenile sockeye salmon during all 

 three years of the survey; however, sample sizes of 

 scales (from the preferred location on fish) were too 

 small for statistical analyses because of descaling of 

 the juvenile salmon by our mid-water rope trawl. Data 



