Interannual variability can be large for survival rate or body size of adults in stocks of Pacific salmon (Oncorhynchus spp.). Within-stock or between-stock density-dependent processes usually explain little of this variation, suggesting that environmental processes are important (Peterman 1987). Our purpose was to identify patterns of covariation across years and among salmon stocks in components of recruitment of fishable biomass, such as survival rate and body size of adults. Such patterns can help characterize the spatial and temporal scale over which environmental processes influence recruitment.

We explored patterns of covariation in two components of recruitment, an index of survival rate and an index of body size of adults, across 21 sockeye salmon (O. nerka) stocks from British Columbia (BC) (20 stocks from the Fraser River plus Skeena River, for brood years 1948-86) and 9 sockeye salmon stocks from Bristol Bay, Alaska (brood years 1954-83). Stocks from these two regions overlap for much of their marine life history. For each stock, an index of survival rate was calculated to account for lognormal error structure and correct for possible within-stock density-dependent effects. This index was the brood-year residuals from a Ricker stock-recruitment model fit by linear regression of log_e(recruits per spawner) vs. spawners. The index of body size was average length of female spawners by age.

Correlation coefficients were calculated for pairwise comparisons among the 21 BC stocks (210 possible comparisons), among the 9 Bristol Bay stocks (36 possible comparisons), and between the BC and Bristol Bay stocks (189 possible comparisons) for a) the survival index calculated for sub-2 returns (e.g., age 4₂ and 5₂ fish) by brood year and b) length data for age 4₂ female spawners by brood year. In addition, to determine which lifestage, freshwater or marine, may be the largest contributor to interannual variability, within-stock comparisons for Bristol Bay stocks were made by correlating the survival rate of sub-2 and sub-3 returns with a) the same brood year and b) the same ocean-entry year (OEY). For body size, these same within-stock correlations were done in addition to correlations between sub-2 and sub-3 fish with the same return year.

There were large positive correlations in survival rate index among the Bristol Bay stocks and, to a lesser extent, among the BC stocks, with a tendency toward negative correlations between Bristol Bay and BC stocks (Fig. 1). This suggests that within each region (i.e., Bristol Bay or BC) the interannual variability in survival rate of sockeye stocks is influenced by common environmental processes, but that these processes are distinct for the stocks of each region. For eight of the nine Bristol Bay stocks, the survival index of sub-2 adults was more highly correlated (average r = 0.47) with the survival index of sub-3 adults of the same stock that had the same ocean-entry year versus adults that had the same brood year (average r = 0.25). This indicates that the late freshwater or early marine lifestage (shared by age groups with the same OEY) may be more important than the early freshwater lifestage (shared by age groups with the same brood year) for determining variability in survival rate.

Body size tended to be positively correlated among BC stocks and Bristol Bay stocks, but there was no consistent correlation between Bristol Bay and BC stocks (Fig. 2). As with the survival index, these results suggest that environmental processes influence the interannual variability in body size of adult sockeye within each region, but not between regions. For the Bristol Bay stocks, within-stock correlations of lengths of female spawners were all positive and large between age groups that shared the same return year and OEY (e.g., age 4₂ vs. 5₃ and 5₂ vs. 6₃, average r = 0.66). However, for age groups that shared the same brood year and OEY but not the same return year, the correlations were considerably weaker (e.g., 4₂ vs. 5₂, and 5₃ vs. 6₃, average r = 0.19). This suggests that the final year of ocean residence (shared by age groups with the same return year) is the most critical period for determining interannual variability in age-specific size of adults.

Patterns of covariation can also be examined by fitting different types of models that include environmental effects to data for the residuals in log_e(recruits/spawner) (details given in Adkison et al. in press). For Bristol Bay sockeye salmon, the best-fit model was a one-time shift in parameters of the Ricker stock-recruitment curve, coinciding with the rapid change in the mid-1970s in intensity of the Aleutian low-pressure weather systems and associated wind-driven processes (Trenberth and Hurrell 1994, Hare and Francis 1995). On average across the nine Bristol Bay sockeye salmon stocks, the Ricker 'a' parameter (an index of productivity) increased threefold between the early and late 1970s, whereas the Ricker 'b' parameter did not change appreciably or consistently among stocks. This increased productivity is consistent with Brodeur and Ware's (1992) finding that zooplankton abundance increased between those two periods in the Gulf of Alaska.

These results suggest that large-scale environmental processes can strongly influence survival and growth rates of sockeye salmon. Analyses of historical data as well as forecasts by management agencies and the fishing industry must therefore take the spatial and temporal scales of those environmental processes into account.

Funding was provided by grants to R. M. Peterman from the Ocean Production Enhancement Network, one of Canada's Networks of Centres of Excellence, and the Natural Sciences and Engineering Research Council of Canada.

Adkison, M. D., R. M. Peterman, M. F. Lapointe, D. M. Gillis, and J. Korman. In press. Alternative models of climatic effects on sockeye salmon (Oncorhynchus nerka) production in Bristol Bay, Alaska, and the Fraser River, British Columbia. Fish. Oceanogr. 5.

Brodeur, R. D., and D. M. Ware. 1992. Interannual and interdecadal changes in zooplankton biomass in the subarctic Pacific Ocean. Fish. Oceanogr. 1:32-38.

Hare, S. R., and R. C. Francis. 1995. Climate change and salmon production in the Northeast Pacific Ocean. In R. J. Beamish (editor), Climate change and northern fish populations. Can. Spec. Publ. Fish. Aquat. Sci. 121:357-372.

Peterman, R. M. 1987. Review of the components of recruitment of Pacific salmon. In M. Dadswell, R. Klauda, C. Moffitt, R. Saunders, R. Rulifson, and J. E. Cooper (editors), Common strategies of anadromous and catadromous fishes, p. 417-429. Am. Fish. Soc. Symp. 1.

Trenberth, K. E., and J. W. Hurrell. 1994. Decadal atmosphere-ocean variations in the Pacific. Climate Dynamics 9:303-319.

Table of Contents