The genetic origins of species boundaries at subtropical and temperate ecoregions in the North American racers (Coluber constrictor)

Abstract

Phylogeographically structured lineages are a common outcome of range-wide population genetic studies. In the southeastern United States, disconnection between populations found at the intersection of the southeastern coastal plains of peninsular Florida and the southeastern plains of the adjacent continent is readily apparent among many plants and animals. However, the timing and maintenance of species boundaries between these distinctly different subtropical and temperate regions remains unknown for all organisms studied there. Using genome-scale data, we examine the timing of origins, gene flow, and the movement of genes under selection in unique ecoregions within the North American racers (Coluber constrictor). Isolation-migration models along with tests of genome-wide selection, locus-environment associations, and spatial and genomic clines demonstrate that two unrecognized species are present and are in contact at the boundary of these two ecoregions. We show that selection at several loci associated with unique environments have maintained species boundaries despite constant levels of gene flow between these lineages over thousands of generations. This research provides a new avenue of research to examine speciation processes in poorly studied biodiversity hotspots.

Fig. 1. Spatial admixture results for the two lineages of Coluber constrictor in the eastern United States.

A Ancestral coefficients over space using TESS3r. B Interpolation of admixture across space showing the location of the hybrid zone. C Spatial cline between the two lineages using admixture estimates.

Fig. 2. Estimates of groups using self organizing maps (SOM) machine learning methods.

A Relative distance to the closest unit over each iteration. B Support for number of clusters. C Data layer weights. D SOM neighbor distances and clusters. E Map showing the location of individuals and admixture. F admixture estimates across all individuals.

Fig. 3. Geographic and environmental association with genetic data.

A Genetic distance by spatial distance highlighting the presence of two groups. B GDM spline model for significant variables. C GDM variable importance over all climate variables with inset showing a photo of Coluber constrictor from the East lineage taken by F. T. Burbrink.

Fig. 4. Historical demographic estimates using GADMA.

A The timing of origin of the Florida, East, and Central-West group, migration rates, and population size changes through time for all SNPs. B The timing of origin of the Florida, East, and Central-West group, migration rates, and population size changes through time for SNPs not showing significant selection.

Fig. 5. Detecting significant loci isolating the two lineages of Coluber constrictor.

A Spatial cline width vs distance from the cline center (km) and significant loci circled in red. B Significant genome clines (color) against background non-significant clines (gray). C Venn diagrams showing overlap among significant loci for genome and spatial clines, genome scans, and group prediction via DAPC. D SqCL loci from Coluber constrictor mapped to the Ahaetulla prasina genome (ASM2864084v1) with inset showing the location of significantly selected loci on chromosomes 4 (Z sex chromosome) and 7.