Genotype But Not Body Shape Predicts River Migration Success in Atlantic Salmon

Abstract

Migratory species typically undertake demanding long-distance journeys, across different habitat types during which they are exposed to multiple natural and anthropogenic stressors. Mortality during migration is typically high and may be human induced. Understanding individual responses to these selection pressures is rarely attempted because of the challenges of relating individual phenotypic and genetic data to migration success. Here, we show distinct single nucleotide polymorphism (SNP) sets significantly differentiated between Atlantic salmon smolts making successful migrations to sea and those that failed to migrate, in two different rivers. In contrast, morphological variation was not diagnostic of migration success. Populations from each river were genetically distinct, and while different genes were possibly implicated in migration success in each river, they related to common biological processes (e.g., osmoregulation and immune and stress response). Given that migration failure should quickly purge polymorphism at selected SNPs from a population, the question of how genetic diversity in these populations is maintained is an important one. Standing genetic variation could be maintained by different life history strategies and/or environmentally driven balancing selection. Our work highlights the importance of preserving genetic diversity to ensure evolutionary resilience at the population level and has practical implications for management.

Keywords: SNPs; animal migration; genome scan; geometric morphometrics; salmon; telemetry.

FIGURE 1

The study area (a) and genetic structuring (b, c) of Atlantic salmon populations from rivers flowing into the Moray Firth (Scotland), with samples from the River Ure (England, (a) top left panel) for comparison. Tagging locations and acoustic receivers are shown in map (a). The principal components analysis (PCA) and the ADMIXTURE analysis plots are based on 44,504 SNPs pruned for linkage disequilibrium. In the PCA scatterplot (b), dots represent individual fish, and variance (%) explained by the first and second axes are shown. Colours correspond to rivers. In the ADMIXTURE plot (K = 3; c), each fish individual is represented by a vertical bar. ‘Dev’ and ‘Fin’ are abbreviations for the Rivers Deveron and Findhorn respectively.

FIGURE 2

Fixed (blue circles) and semi (green diamonds) landmarks and linear measurements used for the GM analyses of Atlantic salmon smolts (image modified from Moccetti et al. 2023). Landmarks 14 and 15 (red stars) were used to correct for body arching. (1) Tip of snout; (2) midpoint between 1 and 3; (3) directly above middle of eye; (4) perpendicular to lateral line, projected towards 3; (5) dorsal surface posterior of cranium; (6) perpendicular to lateral line, projected towards 5; (7) anterior insertion point of dorsal fin; (8) anterior insertion point of adipose fin; (9) dorsal insertion point of caudal fin; (10) versal insertion point of caudal fin; (11) posterior midpoint of hypural plate; (12) anterior insertion point of anal fin; (13) anterior insertion point of ventral fin; (14) lateral line—perpendicular to 7; and (15) lateral line—perpendicular to 12.

FIGURE 3

The Manhattan‐style plots (a, b) show all outlier SNPs (dots) identified in bootstrap replicated datasets using ‘OutFLANK’ in each river. The outliers consistently detected in 100% of replicates and used for analysis are highlighted in green. The y‐axis shows the proportion of replicated datasets where each individual outlier SNP was identified. The x‐axis displays the position of the SNPs along the genome with chromosome numbers. (c and d) Principal components analysis scatterplots based on 70 (Oykel) and 67 (Spey) outlier SNPs between successful (blue) and unsuccessful (red) migrant Atlantic salmon smolts. Each dot represents an individual fish. Variance (%) explained by the first and second axes is also shown.

FIGURE 4

Boxplots of length (mm), weight (g) and Fulton's condition factor to compare successful and unsuccessful Atlantic salmon smolts in the Rivers Oykel and Spey.

FIGURE 5

Mean body shape projections (top) and principal components analysis scatterplots (lower panel) show an absence of shape difference between the migratory groups. Procrustes‐aligned coordinates of successful (blue) and unsuccessful (red) Atlantic salmon, where dots represent individual fish are shown below. Variance (%) explained by the first and second axes and 95% confidence ellipses are displayed. Projections show a complete overlap of the blue (successful) and red (unsuccessful) lines in both rivers despite magnifying morphological differences three times to aid visualisation.