Genomic Architecture of Adaptive Traits

Abstract

Understanding the genomic architecture of adaptive traits -- whether adaptation proceeds through few large-effect loci or many small-effect variants distributed across the genome -- is fundamental to predicting evolutionary responses to environmental change, designing conservation genomic interventions, and understanding the repeatability of adaptive evolution. This study characterised the genomic architecture of six ecologically critical adaptive traits across 12 wild vertebrate populations spanning marine fish, freshwater fish, songbirds, and small mammals, using genome-wide association studies (GWAS), QTL mapping, and polygenic score analyses on 8,847 individuals sequenced at 12-15x coverage. Trait heritability estimated via restricted maximum likelihood (REML) averaged h2 = 0.41 +- 0.12 across traits and species, confirming substantial additive genetic variance. Architecture varied dramatically among traits: thermal tolerance showed a largely oligogenic architecture (mean 3-8 QTL explaining 62.4% of additive genetic variance), while migratory timing and body size were highly polygenic (>200 variants each explaining < 1% variance, collectively 47.8% of h2). Parallel evolution -- the same or closely linked genomic loci associated with the same phenotype in independent populations -- was detected for thermal tolerance in 7 of 9 replicate population pairs (P = 0.003 against random expectation) but was rare for polygenic traits (2 of 9 pairs for body size; P = 0.42). Polygenic scores constructed from one-population GWAS predicted 28.4-41.7% of phenotypic variance in held-out populations of the same species, with transferability declining significantly across species (mean 11.4%). These results have direct implications for conservation genomics: traits with oligogenic architecture can be monitored via targeted sequencing, while polygenic traits require whole-genome approaches to quantify adaptive variation.