Adaptive and neutral genetic variation in spring- and fall-spawning herring (Clupea harengus L.) in the Northwest Atlantic

Abstract

Understanding the factors influencing the spatial distribution of intraspecific diversity is both a fundamental goal in evolutionary biology and necessary for the design of robust management strategies for harvested species. Intraspecific diversity can now be assessed using novel genomic techniques that enable the high-density screening of neutral and adaptive genetic variation. In this thesis, I used whole-genome resequencing of pooled DNA of Atlantic herring (Clupea harengus) to (i) identify patterns, genomic regions, evolutionary processes and environmental variables involved in the origin and maintenance of population divergence and local adaptation in the northwest Atlantic, and (ii) develop a diagnostic genetic tool that can inform conservation and fisheries management. I found fine-scale population structure in herring at putatively adaptive loci, despite low differentiation at neutrally evolving loci. Populations were distinguishable by spawning time and along a latitudinal cline defined by winter sea-surface temperature. Divergent selection leading to adaptation to seasonal reproduction and spawning at different latitudes is likely maintaining molecular divergence patterns for these traits in this environment. Each pattern was underpinned by thousands of outlier SNPs distributed in specific genomic regions spanning several candidate genes, some with a known role in the timing of reproduction (i.e. TSHR). Many spawning time-associated SNPs were shared with populations across the ocean, suggesting such adaptation predates the last glacial maximum. Finally, I developed and evaluated the efficacy of two cost-effective SNP-panels diagnostic of spawning season and latitudinal origin. Individual genotypes at these loci confirmed temporal stability of genetic differences among northern populations and between reproductive strategies despite their mixing outside of the spawning season, suggesting spawning time and site fidelity in ecological time scales. Admixture between reproductive and latitudinal components is unrestricted, and the proportion of admixed individuals vary across aggregations. Some individuals with intermediate admixture levels spawned in either season, suggesting that spawning time is not completely fixed. The analysis of mixture samples revealed the dynamic composition of aggregations outside of the breeding season and demonstrated the utility of the SNP-panels for mixed stock assessment. Altogether, these results contribute to the hypothesis that selection influences spatial distributions of genetic variation, highlighting the need to manage or conserve ecologically important adaptive variation in nature.