The genomic architecture of local adaptation in two connected populations of three-spined stickleback

Populations often adapt to their local environments despite the homogenizing effects of gene flow, but the genomic mechanisms enabling this process remain unclear. Theory predicts that adaptive divergence under high connectivity is favored when beneficial alleles cluster in regions of reduced recombination, a pattern that can be reinforced by structural variants (SVs). We investigated this in three-spined sticklebacks (Gasterosteus aculeatus) from the St. Lawrence Estuary, where distinct freshwater and marine ecotypes meet and interbreed along a short ecological gradient. Using long- and short-read whole-genome sequencing, we mapped fine-scale recombination landscapes, catalogued SVs, and examined their relationship with adaptive genomic regions. Recombination landscapes differed between populations, with population-specific shifts in recombination rate estimated by an LD-based method. Putatively adaptive regions were not confined to low-recombination regions, yet SVs (inversions, insertions, and deletions) frequently coincided with local recombination suppression and elevated differentiation, suggesting they may contribute to local adaptation. Differentiated regions also overlapped disproportionately with previously-identified regions involved in repeated local adaptation across the species range, which tended to be strongly enriched on chromosomes IV, VII and XXI. These repeated regions were associated with lower recombination rates, suggesting that recombination suppression may contribute to their reuse across populations. As found in stickleback populations from other regions, the St. Lawrence populations exhibit elements suggestive of concentrated architectures clustered in a few genomic regions, along with relatively diffuse patterns of highly differentiated regions distributed genome-wide, across a wide range of recombination rates. These results highlight the intertwined roles of recombination variation and structural variation in shaping evolutionary trajectories in connected populations.