Toward Climate-Resilient Aquaculture: Breeding Fish with Genomics for Thermal Tolerance

Rising temperatures challenge aquaculture, but the genetic basis of thermal resilience may provide a route to adaptation. In this study, we synthesize the current evidence on heritability, genomic architecture, genotype-by-environment interaction (G×E), and genomic prediction for thermal tolerance in cultured fishes while distinguishing acute thermal limits from chronic heat-survival and thermal-performance traits. We conducted a structured synthesis of primary quantitative-genetic, genome-wide association studies (GWASs), QTL (quantitative trait locus), reaction-norm, and genomic-prediction studies, prioritizing recent evidence and verifying phenotype definitions and reported estimates against original publications. Evidence is strongest in Atlantic salmon (Salmo salar), rainbow trout (Oncorhynchus mykiss), and olive flounder (Paralichthys olivaceus). Atlantic salmon showed high heritability for CTmax (h²=0.47) and ITmax (h²=0.40), but these traits aregenetically uncorrelated, demonstrating that acute and chronic tolerance should not be treated as equivalent. A separate Atlantic salmon genomic study reports moderate genomic heritability (h²=0.208) and a polygenic architecture, with no significant QTL. In rainbow trout, acute hyperthermia resistance was heritable (h²=0.29±0.05) and highly polygenic, with six QTLs explaining less than 5% of genetic variance. In olive flounder, GWASs identified 204 significant SNPs across eight chromosomes; 35 common significant SNPs were confirmed in an independent population. Bayesian B and random-forest models showed strong prediction performance, with Bayesian B providing superior reliability for genomic breeding values. Reaction-norm analysis in Atlantic salmon further demonstrates substantial G×E across summer heat loads and better statistical properties of genomic than pedigree-based breeding values. Collectively, the evidence supports polygenic inheritance and selection potential, but indicates that prediction reliability across thermal environments remains a key translational gap for climate-resilient breeding in aquaculture fishes.