Genome Study Maps Polygenic Coastal Adaptation in Song Sparrows

A range-wide genome study of the threatened Atlantic song sparrow finds that coastal adaptation is polygenic rather than the work of one major gene. Whole-genome resequencing of 53 birds revealed candidate loci under divergent selection in both isolated barrier-island populations and coastal mainland populations that exchange genes with inland song sparrows. The most promising signals occurred in genes connected with salt and water balance, kidney function, bill development and melanin production, although the researchers stress that these are candidates rather than experimentally proven causes of adaptation.

The work focused on two closely related song sparrow subspecies that meet across the Atlantic coast of North America. The specialist Atlantic song sparrow, Melospiza melodia atlantica, breeds in salt marshes and coastal dunes from the Outer Banks of North Carolina to Cape Cod, Massachusetts. Its inland relative, the eastern song sparrow, Melospiza melodia melodia, occupies shrubby habitats east of the Rocky Mountains and can also breed in coastal areas north of the Atlantic subspecies’ range. The coastal birds face heat, limited freshwater, periodic tidal flooding and high environmental salinity, and they possess several conspicuous adaptations, including larger bills, grayer plumage and heavier dark streaking.

Those differences have long been considered potentially adaptive, but their genetic basis was unresolved. Atlantic song sparrows and nearby inland populations show high levels of gene flow, while their morphological differences become more continuous near the coast. That combination suggests adaptation involving many genes rather than a single locus of large effect. It also creates a useful test of how divergent selection can preserve biologically important traits in populations that continue to interbreed. The researchers sampled 11 sites across Maine, New York, New Jersey, Delaware, Maryland, Virginia and North Carolina during the 2021 and 2022 breeding seasons, ultimately retaining 20 barrier-island birds, 15 coastal-mainland birds and 18 inland birds for filtered analyses.

To characterize the genome, the team generated a new chromosome-level song sparrow reference assembly in collaboration with the Vertebrate Genomes Project. Long-read sequencing and Hi-C data produced an assembly of 1.55 billion base pairs, organized into 353 scaffolds, including 31 complete autosomal chromosomes and the Z and W sex chromosomes. It contained 20,371 annotated genes and was larger and more contiguous than recent reference assemblies for other song sparrow subspecies. Resequenced genomes averaged 14.6-fold coverage and 2.5% missing genotypes. After filtering, 11,496,474 single-nucleotide polymorphisms remained, of which 11,493,090 were represented in a statistically phased dataset covering 96.4% of the genome.

Population structure was subtle but measurable. The first principal component separated inland birds from coastal birds, with southern barrier-island sparrows at the opposite extreme and mainland coastal birds between them. A second component distinguished Pea Island, North Carolina, from the other sampled islands. Discriminant analyses also separated the two subspecies, three broad geographic groups and individual barrier islands. In contrast, the inland birds showed no geographic substructure across a 1,500-kilometre north–south sampling span, a pattern the authors associated with the eastern subspecies’s high dispersal ability and unpredictable migration. The result indicates that Atlantic song sparrow diversity is structured by geography and isolation even though the overall genetic differentiation between subspecies remains low.

Across the genome, mean fixation index, FST, between inland birds and barrier-island birds was only 0.027. A few 10-kilobase windows reached much higher values, but many contained few variable sites and did not show consistently elevated individual markers. Absolute divergence, dxy, was also low and tightly correlated with nucleotide diversity rather than FST, a pattern consistent with recent separation and substantial gene flow. The researchers therefore used a cross-population extended haplotype homozygosity analysis, xpEHH, to identify genomic regions where one population carries unusually long, homogeneous haplotypes. False-discovery-rate correction and a threshold of −log(p) greater than 10 identified discrete outliers, including 11 regions shared by the inland–barrier-island and inland–coastal-mainland comparisons.

Most of the shared signals were consistent with positive selection in the Atlantic song sparrow rather than the inland subspecies. Three of the 11 regions occurred near putative centromeres, and one included a linked group of three genes, but most candidate loci were scattered through the genome rather than concentrated in large inversion-like blocks. That arrangement matters because the song sparrow genome has unusually low linkage disequilibrium: average r-squared fell below 0.10 within 310 base pairs and reached 0.03 at 300 kilobases. High recombination can help separate beneficial alleles and remove deleterious ones, potentially allowing strong selection to maintain a polygenic coastal phenotype without a single “supergene” or extensive reproductive isolation. Nevertheless, the authors caution that the system may not be in stable migration–selection balance.

Several outlier genes offer plausible physiological routes to coastal survival. ANO3 belongs to a family of chloride transmembrane transporters, while S100A16 and TRIP13 have documented associations with the kidney. A predicted gene orthologous to CYP11A1, a mitochondrial cholesterol side-chain cleavage enzyme involved in producing steroid hormones such as aldosterone, was also flagged, linking the genomic pattern to salt and water balance. TRIP13 and PCSK2 appeared in comparable coastal–inland comparisons in swamp sparrows. PCSK2 cleaves hormone precursors, including pro-vasopressin, and has also been associated with the pathway producing α-melanocyte-stimulating hormone, making it relevant to pigmentation as well as physiology. COL15A1, found near another selection signal, has been proposed as an influence on body-size variation in other song sparrow subspecies.

These results identify a genetically distributed basis for the Atlantic song sparrow’s salt-marsh syndrome, but they do not establish that any named gene directly causes its distinctive bill, plumage or salt tolerance. The analysis associates selected genomic regions with candidate genes, and functional experiments would be needed to connect those variants to traits and survival. The comparison also covers only 53 birds from part of the subspecies’ eastern range, and some coastal sites may contain admixed individuals. Demographic models cited by the authors suggest secondary contact may follow earlier allopatry, raising the possibility that gene flow could eventually erode divergence if selection is not strong enough. Nevertheless, the study shows that Atlantic song sparrows retain coastally associated traits in the face of gene flow through a broad, recombinantly dispersed set of loci, while genomic data reveal unexpectedly fine population structure along their threatened coastal range.

Subject of Research: Genomic basis of coastal adaptation in the Atlantic song sparrow

Article Title: Local Adaptation With Gene Flow Has a Polygenic Basis in a Threatened Coastal Songbird

Article References: Clark, J. D., Lim, H. C., Olsen, B. J., & Kovach, A. I. (2026). Local Adaptation With Gene Flow Has a Polygenic Basis in a Threatened Coastal Songbird. Ecology and Evolution, 16(10), Article e74506. https://doi.org/10.1002/ece3.74506

Image Credits: AI Generated

DOI: 10.1002/ece3.74506

Keywords: Atlantic song sparrow, local adaptation, gene flow, polygenic adaptation, whole-genome sequencing, divergent selection, coastal birds, osmoregulation, population genomics, salt marsh adaptation, Local, Adaptation