The population structure in the Baltic herring reflects natural selection and local adaptation

How species time reproduction and adapt to environmental conditions are key topics in ecology and evolutionary biology. Here, we conducted a high-resolution population genetic analysis of Baltic herring, a subspecies of Atlantic herring (<i>Clupea harengus</i>). Genotypes at >4,500 SNPs were generated from >4,500 spawning individuals, sampled from 150 locations spanning Swedish's eastern coast. Abiotic factors-week of spawning, latitude, temperature, salinity-were used to assess how genetic variation is shaped by temporal, spatial, and environmental gradients. Our results reaffirm strong genetic differentiation between spring- and autumn-spawning ecotypes, despite hybridization suggesting ongoing gene flow between the two ecotypes. We document significant substructuring within the spring-spawning ecotype, delineating three main, previously unidentified, genetic clusters underpinned by adaptative genetic variation associated with latitude, salinity, temperature, and spawning time. Complementary linkage disequilibrium (LD) partitioning showed that adaptive loci-especially those in inversion regions-exhibit strong elevated among-population LD, consistent with divergence maintained by local selection despite ongoing gene flow. Clinal variation in allele frequencies indicated regionally distinct selection pressures, including shifts in allele frequencies at two major supergenes (inversions) and at a suite of genes correlated with abiotic factors. Importantly, rare genetic outlier populations are identified within each geographic region which further illustrates the unexpected fine-grained population structure of Baltic herring and implies a strong homing behavior in this abundant marine fish. Overall, this study demonstrates the capacity for targeted population genetic studies to detect adaptive variation in natural populations, the outcomes of which have direct implications for sustainable fisheries and biodiversity management.