Fossil, archaeological, and morphometric data suggest that indigenous red foxes in North America were derived from vicariance in two disjunct refugia during the last glaciation: one in Beringia and one in the contiguous USA. To test this hypothesis, we conducted a phylogeographical analysis of the North American red fox within its presettlement range. We sequenced portions of the mitochondrial cytochrome b (354 bp) gene and D-loop (342 bp) from 220 historical red fox specimens. Phylogenetic analysis of the cytochrome b gene produced two clades that diverged c. 400,000 years before present (bp): a Holarctic and a Nearctic clade. D-loop analyses of the Nearctic clade indicated three distinct subclades (> or = 99% Bayesian posterior probability); two that were more recently derived (rho estimate c. 20,000 bp) and were restricted to the southwestern mountains and the eastern portion of North America, and one that was older (rho estimate c. 45,000 bp) and more widespread in North America. Populations that migrated north from the southern refugium following deglaciation were derived from the colonization of North America during or prior to the Illinoian glaciation (300,000-130,000 bp), whereas populations that migrated south from the northern refugium represent a more recent colonization event during the Wisconsin glaciation (100,000-10,000 bp). Our findings indicate that Nearctic clade red foxes are phylogenetically distinct from their Holarctic counterparts, and reflect long-term isolation in two disjunct forest refugia during the Pleistocene. The montane lineage, which includes endangered populations, may be ecologically and evolutionarily distinct.

Most native red foxes (Vulpes vulpes)

Widely distributed taxa provide an opportunity to compare biogeographic responses to climatic fluctuations on multiple continents and to investigate speciation. We conducted the most geographically and genomically comprehensive study to date of the red fox (Vulpes vulpes), the world's most widely distributed wild terrestrial carnivore. Analyses of 697 bp of mitochondrial sequence in ~1000 individuals suggested an ancient Middle Eastern origin for all extant red foxes and a 400 kya (SD = 139 kya) origin of the primary North American (Nearctic) clade. Demographic analyses indicated a major expansion in Eurasia during the last glaciation (~50 kya), coinciding with a previously described secondary transfer of a single matriline (Holarctic) to North America. In contrast, North American matrilines (including the transferred portion of Holarctic clade) exhibited no signatures of expansion until the end of the Pleistocene (~12 kya). Analyses of 11 autosomal loci from a subset of foxes supported the colonization time frame suggested by mtDNA (and the fossil record) but, in contrast, reflected no detectable secondary transfer, resulting in the most fundamental genomic division of red foxes at the Bering Strait. Endemic continental Y-chromosome clades further supported this pattern. Thus, intercontinental genomic exchange was overall very limited, consistent with long-term reproductive isolation since the initial colonization of North America. Based on continental divergence times in other carnivoran species pairs, our findings support a model of peripatric speciation and are consistent with the previous classification of the North American red fox as a distinct species, V. fulva.

Introduced species can threaten native taxa in multiple ways, including competition and hybridization, which can reduce fitness, alter ecological niches or swamp native genomes. Encroachment and hybridization by introduced species also provide opportunities to study the dynamics of invasiveness and hybridization during early stages following contact. We used 33 microsatellites, 51 single nucleotide polymorphisms and a mtDNA marker to characterize the extent and spatial pattern of encroachment and hybridization between a native, endemic subspecies of red fox (Vulpes vulpes patwin) and an introduced red fox population composed of highly admixed, phylogenetically divergent stock, resulting from a century of domestication. Both nuclear and mtDNA markers indicated that hybridization was primarily restricted to a narrow zone where the two populations came into contact. Although a few introgressed genotypes were detected in the interior of the native range, we found no immigrant foxes or F(1) or F(2) hybrids there, suggesting native foxes excluded introduced individuals. We speculate that the observed interbreeding at the periphery was facilitated by low densities. In total, 98% of mtDNA haplotypes in the native range were native and 96% of the nuclear ancestry was estimated to be native. Although the introduced range had expanded fivefold over the past four decades, native and non-native haplotypes from museum samples collected in and near the native range three decades earlier showed a similar geographic distribution as today, suggesting that the native range and hybrid zone were relatively stable. We hypothesize that the monogamous mating system of red foxes and other wild canids may enhance their resistance to hybridization because of greater fitness consequences associated with mate discrimination.