Landscape structure and climatic variation determine Atlantic salmon genetic connectivity in the Northwest Atlantic

The identification of landscape or climatic correlates of genetic connectivity can reveal the processes that regulate spatial diversity and inform the management and conservation of exploited or endangered species. Increasingly, the role of local adaptation in regulating spatial structure is gaining recognition, though disentangling its influence from that of dispersal and genetic drift remains challenging. We examined genetic connectivity using 15 microsatellite loci in 3546 Atlantic salmon (Salmo salar) from 47 locations in Newfoundland and Labrador. Cluster analysis revealed regional associations, but examples of unexpectedly high divergence over small spatial scales (<50 km) or low divergence over large scales (>500 km) were common with little evidence of isolation by distance (R² = 0.04). We evaluated the hypothesis that spatial structure is less influenced by straying and more by adaptation and (or) drift using sample-associated data on 22 landscape and climate variables and both stepwise multiple regression and redundancy analysis. Landscape variables explained ~40% of the variation in spatial structure with freshwater habitat area as the largest determinant. Linkages were also observed with water chemistry and winter severity, supporting hypotheses of adaptive divergence. Overall, the results suggest Atlantic salmon populations are structured by extensive genetic drift and low rates of effective straying and demonstrate how a landscape genetics approach can inform wildlife management and conservation.