Genome-wide local ancestries discriminate homoploid hybrid speciation from secondary introgression in the red wolf (Canidae: Canis rufus)

Abstract

Hybridization is well recognized as a driver of speciation, yet it often remains difficult to parse phylogenomically in that post-speciation gene flow frequently supersedes an ancestral signal. Here we examined how interactions between recombination and gene flow shaped the phylogenomic landscape of red wolf to create non-random retention of introgressed ancestry. Our re-analyses of genomic data recapitulate fossil evidence by demonstrating red wolf was indeed extant and isolated prior to more recent admixture with other North American canids. Its more ancient divergence, now sequestered within low-recombinant regions on the X-chromosome (i.e., chromosomal refugia), is effectively masked by multiple, successive waves of secondary introgression that now dominate its autosomal ancestry. These interpretations are congruent with more theoretical explanations that describe the manner by which introgression can be localized within the genome through recombination and selection. They also tacitly support the large-X effect, i.e., the manner by which loci that contribute to reproductive isolation can be enriched on the X-chromosome. By contrast, similar, high recombinant regions were also found as enriched within very shallow gene trees, thus reflecting post-speciation gene flow and a compression of divergence estimates to 1/20th of that found in recombination cold spots. Our results effectively reconcile conflicting hypotheses regarding the impact of hybridization on evolution of North American canids and support an emerging framework within which the analysis of a phylogenomic landscape structured by recombination can be used to successfully address the macroevolutionary implications of hybridization.