Trends in marine survival of Atlantic salmon populations in eastern Canada

Sebastián A Pardo,Dustin Raab,J Brian Dempson,Jeffrey A Hutchings,Ross A Jones,Geir H Bolstad,Julien April,Shijie Zhou

doi:10.1093/icesjms/fsab118

Abstract

Abstract Declines in wild Atlantic salmon (Salmo salar) abundance throughout the north Atlantic are primarily attributed to decreases in survival at sea. However, comparing trends in marine survival among populations is challenging as data on both migrating smolts and returning adults are sparse and models are difficult to parameterize due to their varied life histories. We fit a hierarchical Bayesian maturity schedule model to data from seven populations in eastern Canada to estimate numbers of out-migrating smolts, survival in the first and second year at sea, and the proportion returning after 1 year. Trends in survival at sea were not consistent among populations; we observe positive, negative, and no correlations in these, suggesting that large-scale patterns of changes in marine survival are not necessarily representative for individual populations. Variation in return abundances was mostly explained by marine survival in the first winter at sea in all but one population. However, variation in the other components were not negligible and their relative importance differed among populations. If salmon populations do not respond in a uniform manner to changing environmental conditions throughout their range, future research initiatives should explore why.

Highlights

Reductions in fishing mortality, albeit necessary, are not always sufficient to facilitate population recovery
These priors, when converted to yearly survival, are roughly constrained between 0 and 0.2 for S1 and between 0.2 and 0.5 for S2 (Figure 2a). They are informative only to the extent that they limit the estimates of marine survival to what we considered to be biologically realistic based on previous studies on return rates of 1SWdominated populations in Newfoundland and Labrador (NL), as well as the few studies that have explored survival in the second winter at sea of Atlantic salmon in eastern Canada
We explored the effect of using weaker priors for Z1, Z2, and proportion returning as grilse (Pg), and found that the posteriors of Z1 were not sensitive to the choice of prior, while the posteriors Z2 and Pg became more uncertain with weaker priors while maintaining the same temporal trends observed when using stronger priors

Summary

Introduction

Reductions in fishing mortality, albeit necessary, are not always sufficient to facilitate population recovery. Gradual efforts to close commercial Atlantic salmon (Salmo salar) fisheries in eastern Canada culminated in full moratoria in all regions, beginning in the Maritime provinces (1984) and following in Newfoundland (1992), Labrador (1998), and Québec (QC) (2000). Since these closures, many populations have not increased as expected (Dempson et al, 2004; ICES, 2019); some have been assessed as threatened or endangered by the Committee on the Status of Endangered Wildlife in Canada (COSEWIC, 2010), Canada’s science advisory body (to the national government) on. While the mechanisms behind population declines are not fully understood, the potential drivers of these are many (see Cairns, 2001, for a detailed discussion of possible causes), including but not limited to: fishing mortality (Dempson et al, 2004); damming of waterways and changes in the freshwater habitat (Dunfield, 1985; Clarke et al, 2014); acidification ( in Nova Scotia’s [NS] Southern Uplands, see Gibson et al, 2009); predation by seals and birds (Cairns and Reddin, 2000); negative effects of interbreeding or interactions with escaped farmed salmon (Keyser et al, 2018); and climate-driven changes in survival and productivity (Mills et al, 2013)

Methods

Results

Conclusion