After establishing secondary contact, recently diverged populations may remain reproductively isolated or may hybridize to a varying extent depending on factors such as hybrid fitness and the strength of assortative mating. Here, we used genomic and phenotypic data from three independent contact zones between subspecies of the Variable Seedeater (Sporophila corvina) to examine how coloration and genetic divergence shape patterns of hybridization. We found that differences in plumage coloration are likely maintained by divergent selection across contact zones; however, the degree of plumage differentiation does not match overall patterns of hybridization. Across two parallel contact zones between populations with divergent phenotypes (entirely black vs. pied plumage), populations hybridized extensively across one contact zone but not the other, suggesting that plumage divergence is not sufficient to maintain reproductive isolation. Where subspecies hybridized, hybrid zones were wide and formed by later-generation hybrids, suggesting frequent reproduction and high survivorship for hybrid individuals. Moreover, contemporary gene flow has played an important role in shaping patterns of genetic structure between populations. Replicated contact zones between hybridizing taxa offer a unique opportunity to explore how different factors interact to shape patterns of hybridization. Overall, our results demonstrate that divergence in plumage coloration is important in reducing gene flow but insufficient in maintaining reproductive isolation in this clade, and that other factors such as divergence in song and time since secondary contact may also play an important role in driving patterns of reduced hybridization and gene flow.

Scorvina_SNPs_1.vcf: SNP data from Genotype-by-Sequencing approach (GBS). We called SNP variants using TASSEL version 5.2.65 following the GBSv2 SNP discovery and production pipeline. Then, we filtered genome-wide single-nucleotide polymorphisms (SNPs) using VCFTOOLS version 0.1.16. Our filtered criteria consisted of including biallelic SNPs with no indels, minimum read depth ≥ 3, minimum allele frequency ≥ 0.05, a maximum proportion of missing data ≤ 0.25, and SNPs that were ≥ 100 bp apart to reduce the effect of highly linked markers. We further filtered the data file in TASSEL, keeping SNPs with a maximum heterozygosity ≤ 0.8 to exclude potential paralogs, and removed individuals with ≥40% missing data. The resulting data set retained 255 individuals and 14,839 SNPs.

Scorvina_SNPs_2.vcf: SNP data from Genotype-by-Sequencing approach (GBS). We called SNP variants using TASSEL version 5.2.65 following the GBSv2 SNP discovery and production pipeline. Then, we filtered genome-wide single-nucleotide polymorphisms (SNPs) using VCFTOOLS version 0.1.16. Our filtered criteria consisted of including biallelic SNPs with no indels, minimum read depth of ≥ 3, maximum proportion of missing data of ≤ 0.15, and SNPs that were ≥ 1000 bp apart to obtain a complete site frequency spectrum (not truncated due to minimum allele frequency filtering) while keeping a manageable SNPs sample size, due to computer processing. This second data set retained 8,437 SNPs from 10 randomly selected genotypically pure individuals per subspecies

Scorvina_B2.txt: We measured the spectral properties from the three body patches that differ between subspecies (i.e., rump, throat, and belly), collecting five reflectance measurements (from 300 to 700 nm) per patch. We used an Ocean Optics USB2000 spectrophotometer, and a PX-2 pulsed Xenon light source (Ocean Optics Inc., Dunedin, FL, USA), with the average reading set at 50, integrating time at 16 msec, boxcar smoothing at 10, and correction for electric dark noise activated.  We then estimated mean brightness (B2) for each individual patch from the reflectance reads, using the R package PAVO2.

Scorvina_morph.txt: We measured seven morphometric traits (i.e., body mass, total culmen length, beak depth and width at the nares, right wing chord, right tarsus length, and tail length) using a caliper (accuracy ± 0.1mm) from 193 males, 125 females, and nine individuals of unknown sex.