Genetically distinct breeding populations
Sequencing efforts resulted in sequences from 330 individuals with a mean coverage of 1.6X (range: 0.02X – 5.2X; Table S1). For the 182 breeding samples, the mean coverage was 1.7X (range: 0.6X – 5.2X), while the 148 nonbreeding samples had a mean coverage of 1.5X (range: 0.02 – 3.4X). Down sampling individuals above 2X coverage to 2X coverage, resulted in an overall mean coverage of 1.4X (1.5X for breeding samples, 1.3X for nonbreeding samples). Our SNP filtering produced genotype likelihood data for 4,722,390 variants. We removed 13 individuals from subsequent analyses due to high relatedness by removing a single individual from each related pair. Principal components analysis with the breeding samples revealed five genetic clusters that aligned with geography: Western Boreal (Alaska to Saskatchewan), Basin Rockies (South Dakota and Montana), Southern Temperate (from Missouri, east to Maryland, south to Louisiana), Northern Temperate (from Minnesota, east to Quebec, south to Pennsylvania), and Maritime Provinces (New Brunswick and Newfoundland; Figure S1). Admixture results for five groups revealed a similar delineation of individuals as in the principal components analysis (Figure 1, Figure S1). Pairwise FST values among these breeding populations had a mean of 0.009 and ranged from the weakest differentiation (FST = 0.004) between the Northern Temperate and Southern Temperate groups and the strongest (FST = 0.018) between Maritime Provinces and Basin Rockies (Table S2). Based on our comprehensive sampling of the core regions of the American Redstart breeding range and the population structure results, we expect there to be no unsampled genetically distinct breeding populations. For example, while we did not sample individuals from Alaska, we expect Alaska breeding birds to be a part of the Western Boreal breeding population.