Variable Effects of Biological and Environmental Processes on Coho Salmon Marine Survival in Southeast Alaska

Abstract

Correlation analyses, linear regression models, and multistock mixed effects models were used to examine the relationships between coho salmon Oncorhynchus kisutch marine survival and six biological and environmental covariates across 14 southeast Alaska (SEAK) stocks. A primary focus of the study was to investigate the influence of pink salmon O. gorbuscha and chum salmon O. keta fry abundances on coho salmon marine survival. The coho salmon stocks exhibited strong covariation; 88 of the 91 pairwise comparisons among the coho salmon stocks covaried positively and 54 of them were significant (P < 0.05). Only one of the covariates, the North Pacific index, which is a measure of the Aleutian low pressure zone, had consistent relationships (positive) across all 14 stocks. The other covariates, including freshwater discharge, the Pacific decadal oscillation, sea-surface temperature, and two indices of pink salmon and chum salmon fry abundances, all had inconsistent relationships with marine survival. The best-fit linear regression models varied greatly among the 14 stocks, as did the R ² values, which ranged from 0.00 to 0.54. An index representing local hatchery pink salmon and chum salmon fry abundance was the most important variable in explaining the variation in marine survival, having a stronger estimated effect on survival than an index of local wild pink salmon fry abundance. The magnitude and sign of the hatchery pink salmon and chum salmon effect varied greatly among different localities. Our results suggest that (1) SEAK coho salmon stocks are not equally influenced by the same factors and (2) there are factors that appear to affect marine survival of SEAK coho salmon stocks at varying spatial scales. This study also provides evidence that coho salmon stocks throughout SEAK experience some degree of regional concordance in the marine environment but also that local stock-specific conditions are important in fully understanding variation in marine survival.