Genome-Environment Associations reveal shared and unique adaptive loci across multiple pollutants and populations of the eastern mosquitofish <i>Gambusia holbrooki</i>

Abstract Rapid adaptation to human-induced stressors is commonplace in the context of global change, including during biological invasions. Identifying the genomic bases and associated biological functions underlying such adaptation is hence crucial to understand and anticipate the response of populations and species to changing conditions. In particular, the extent to which genetic responses to multiple anthropogenic stressors vary between populations in the wild has been relatively unexplored. We addressed this question by leveraging whole-genome sequence data (both PoolSeq and IndSeq) in invasive populations of Gambusia holbrooki - a widespread invasive fish species - collected from 14 locations with different multi-pollutants exposure. We sought to identify adaptive loci associated with pollution tolerance by conducting Genome-Environment Association (GEA) analyses. Additionally, we investigated the degree of adaptive loci reuse between pollutants and their combinations as well as across populations. We found strong signals of association between allele frequency changes and pollutant exposure at several genomic locations, often overlapping with genes known for their functions in detoxification and immune response. We further showed that most adaptive loci are not shared among all populations, suggesting heterogeneous genomic response to each local selection. Interestingly, multiple tests looking for footprints of selection and association to pollutants yielded consistent results. Our results shed new light on the extent of genetic convergence in the genomic bases of rapid adaptation to different cocktails of human-induced pollution, and are important to our understanding of the fate of wild populations facing increasingly complex and stressful environments.