North Pacific Anadromous Fish Commission Technical Report No. 7: 56–58, 2007 56 David A. Beauchamp 1 , Alison D. Cross 2 , and Jamal H. Moss 3 1 U.S. Geological Survey, University of Washington, School of Aquatic and Fisheries Sciences, Box 355020, Seattle, WA 98105-5020, USA 2 University of Washington, Box 355020, Seattle, WA 98105-5020, USA 3 Alaska Fisheries Science Center, Auke Bay Laboratory, NMFS, NOAA, 11305 Glacier Highway, Juneau, Alaska 99801-8626, USA Keywords: Growth, survival, pink salmon, Gulf of Alaska, Prince William Sound All correspondence should be addressed to D.A. Beauchamp. e-mail: davebea@u.washington.edu Inter-Annual Patterns in Stage-Specific Feeding, Growth, and Survival of Juvenile Pink Salmon in the Gulf of Alaska Evidence is mounting for the positive relationship between marine survival of salmon and larger juvenile size or higher growth rates, but little is known about the precise timing of critical life stages or the mechanisms underlying size-selective marine mortality. The temporal dynamics and variety of factors that potentially affect distribution, growth, and survival can confound determinations of the processes that most influence performance during critical life stages in the ocean. Hatcheries in Prince William Sound (PWS) release approximately 600 million otolith- marked juvenile pink salmon annually, and these hatchery fish represent the largest biomass of juvenile salmon in the eastern coastal Gulf of Alaska (GOA). Pink salmon are considered the primary zooplankton consumers in this region. Juvenile pink salmon in PWS and the coastal GOA exhibited a three-fold difference in marine survival during the GLOBEC sampling program in 2001–2004. For otolith-marked hatchery pink salmon released in PWS Sound during 2001–2004, we investigated the relationship of growth and size-selective mortality during and after the first growing season. Using a bioenergetics model, we examined monthly and inter-annual patterns in spatial distribution, thermal experience, growth, and diet composition to relate feeding and growth performance to inter- annual variability in stage-specific mortality and total ocean survival. During 2001–2004, monthly sampling cruises were conducted from July through September or October at three sites in PWS and six or more sites along the Seward Line (GAK 1-6) in the coastal Gulf of Alaska (GOA). At each station, fish were sampled with a surface trawl and zooplankton were collected with a 500 micron Tucker trawl towed at the surface or 0–10 m vertical tows with a 333 micron bongo net; vertical temperature and salinity profiles were recorded with CTD casts. Fish were identified, counted, measured, weighed, and frozen for additional processing. In the lab, fish were thawed, reweighed and measured, and scales, otoliths, and stomach contents were removed. Energy densities were measured for a subset of fish and some zooplankton using a bomb calorimeter. Energy densities for the remainder of the prey were taken from literature values. Stage-specific, size-selective mortality was examined following Moss et al. (2005) by comparing back- calculated size and growth (based on the radius of specific circuli on scales) at the same life stages for both juveniles sampled during the first summer in PWS and GOA, and for surviving adults from the same hatchery groups that were harvested in cost recovery fisheries close to each hatchery. A bioenergetics model simulated monthly consumption rates and growth efficiency for hatchery cohorts of juvenile pink salmon, from the time of their release into PWS in May, through September each year during 2001– 2004. Inputs for these simulations tracked the modal distribution, diet composition, and thermal experience of identifiable hatchery cohorts within and among different specific water masses (PWS; the Alaska Coastal Current- ACC; and the shelf transition zone-TRANS). Monthly spatial distribution patterns were assigned to the water mass(es) with CPUE ≥ 30% of the monthly total catch rate. For these water masses, back-calculated lengths were converted to weights and were combined with monthly diet composition and thermal experience to estimate feeding rates using a bioenergetics model (Beauchamp et al. in press). Feeding rates were reported both as the biomass of prey consumed over prescribed periods and as proportions of the maximum consumption rate expected for a given size of consumer and thermal regime. Feeding rates and growth efficiency were used to infer inter-annual differences in prey availability. As indicated by larger average scale radius at the same circuli, juvenile pink salmon were generally smaller during years when the ocean survival was low (3% smolt-adult) for cohorts entering the ocean in 2001 and 2003, and were larger when the ocean survival was higher (8–9%) in 2002 and 2004 (Fig. 1). Juveniles that survived to adulthood were significantly larger than the average juveniles from the same hatchery group in situ at the same life stage, and the size of the survivors tended to diverge from average juveniles around August (circuli 8-13; Fig. 1). Since the average size of the juveniles at large in GOA had not converged with the size of the surviving adults by the © 2007 North Pacific Anadromous Fish Commission