Fishing genomes: empowering and innovating Mediterranean fisheries with a genomic toolbox

Conserving biodiversity means protecting not only species and habitats but also genetic diversity, which underpins fitness, resilience, and adaptive capacity. In marine systems, dispersal can blur population boundaries and fishery-dependent indicators can be sparse or delayed; genomic data provide direct, heritable evidence that complements biomass and catch statistics. Stock assessment often assumes a single “unit stock” (a discrete, self-recruiting population with homogeneous vital rates); when this is violated, management can be biased. By quantifying both neutral and functional components of variation in an environmental/oceanographic context, genomics can refine stock boundaries, reveal environment-associated divergence, and monitor erosion of diversity under exploitation and climate change. This thesis operationalizes that framework for Mediterranean fisheries through three case studies. First, using chromosome-scale reference genomes to anchor legacy reduced-representation data, red mullet (Mullus barbatus) shows high basin-wide connectivity consistent with larval dispersal, yet localized signals consistent with selection, indicating that strong gene flow can coexist with spatially structured adaptation. Second, a seascape-genomics analysis of European hake (Merluccius merluccius) integrates genotype–environment associations with geography: neutral loci resolve broad structure, while candidate adaptive loci show additional divergence aligned with temperature–salinity regimes, supporting environment-aware, cross-jurisdictional delineations. Third, high-coverage ancient and modern genomes of Atlantic bluefin tuna (Thunnus thynnus) reveal a subtle but consistent genome-wide erosion of diversity associated with twentieth-century declines, detectable earlier than fishery-dependent baselines. Beyond the neutral signal, higher-impact variants show stronger depletion; conditioning on pre-bottleneck frequency indicates a major role for frequency-dependent drift, but higher-impact classes remain more depleted even at matched starting frequency, consistent with an added contribution of purifying/background selection through the contraction. Together, these contributions translate genomic signals into operational tools such as stock structure, genotype-environment divergence, and temporal indicators of evolutionary capacity, that can be integrated into assessment workflows to support stock identification, movement/productivity parameterization, and rebuilding that safeguards adaptive potential.