Mitochondrial Analysis of the Most Basal Canid Reveals Deep Divergence between Eastern and Western North American Gray Foxes (Urocyon spp.) and Ancient Roots in Pleistocene California

Chromosome-level assembly of the gray fox (Urocyon cinereoargenteus) confirms the basal loss of PRDM9 in Canidae

Reference genome assemblies have been created from multiple lineages within the Canidae family; however, despite its phylogenetic relevance as a basal genus within the clade, there is currently no reference genome for the gray fox (Urocyon cinereoargenteus). Here, we present a chromosome-level assembly for the gray fox (U. cinereoargenteus), which represents the most contiguous, non-domestic canid reference genome available to date, with 90% of the genome contained in just 34 scaffolds and a contig N50 and scaffold N50 of 59.4 and 72.9 Megabases, respectively. Repeat analyses identified an increased number of simple repeats relative to other canids. Based on mitochondrial DNA, our Vermont sample clusters with other gray fox samples from the northeastern United States and contains slightly lower levels of heterozygosity than gray foxes on the west coast of California. This new assembly lays the groundwork for future studies to describe past and present population dynamics, including the delineation of evolutionarily significant units of management relevance. Importantly, the phylogenetic position of Urocyon allows us to verify the loss of PRDM9 functionality in the basal canid lineage, confirming that pseudogenization occurred at least 10 million years ago.

Can demographic histories explain long-term isolation and recent pulses of asymmetric gene flow between highly divergent grey fox lineages?

Secondary contact zones between deeply divergent, yet interfertile, lineages provide windows into the speciation process. North American grey foxes (Urocyon cinereoargenteus) are divided into western and eastern lineages that diverged approximately 1 million years ago. These ancient lineages currently hybridize in a relatively narrow zone of contact in the southern Great Plains, a pattern more commonly observed in smaller-bodied taxa, which suggests relatively recent contact after a long period of allopatry. Based on local ancestry inference with whole-genome sequencing (n = 43), we identified two distinct Holocene pulses of admixture. The older pulse (500-3500 YBP) reflected unidirectional gene flow from east to west, whereas the more recent pulse (70-200 YBP) of admixture was bi-directional. Augmented with genotyping-by-sequencing data from 216 additional foxes, demographic analyses indicated that the eastern lineage declined precipitously after divergence, remaining small throughout most of the late Pleistocene, and expanding only during the Holocene. Genetic diversity in the eastern lineage was highest in the southeast and lowest near the contact zone, consistent with a westward expansion. Concordantly, distribution modelling indicated that during their isolation, the most suitable habitat occurred far east of today's contact zone or west of the Great Plains. Thus, long-term isolation was likely caused by the small, distant location of the eastern refugium, with recent contact reflecting a large increase in suitable habitat and corresponding demographic expansion from the eastern refugium. Ultimately, long-term isolation in grey foxes may reflect their specialized bio-climatic niche. This system presents an opportunity for future investigation of potential pre- and post-zygotic isolating mechanisms.

Population Genetics of California Gray Foxes Clarify Origins of the Island Fox

We used mitochondrial sequences and nuclear microsatellites to investigate population structure of gray foxes (Urocyon cinereoargenteus) and the evolutionary origins of the endemic island fox (Urocyon littoralis), which first appeared in the northern Channel Islands <13,000 years ago and in the southern Channel Islands <6000 years ago. It is unclear whether island foxes evolved directly from mainland gray foxes transported to the islands one or more times or from a now-extinct mainland population, already diverged from the gray fox. Our 345 mitochondrial sequences, combined with previous data, confirmed island foxes to be monophyletic, tracing to a most recent common ancestor approximately 85,000 years ago. Our rooted nuclear DNA tree additionally indicated genome-wide monophyly of island foxes relative to western gray foxes, although we detected admixture in northern island foxes from adjacent mainland gray foxes, consistent with some historical gene flow. Southern California gray foxes also bore a genetic signature of admixture and connectivity to a desert population, consistent with partial replacement by a late-Holocene range expansion. Using our outgroup analysis to root previous nuclear sequence-based trees indicated reciprocal monophyly of northern versus southern island foxes. Results were most consistent with island fox origins through multiple introductions from a now-extirpated mainland population.

Genomic Assessment of Cancer Susceptibility in the Threatened Catalina Island Fox (Urocyon littoralis catalinae)

Small effective population sizes raise the probability of extinction by increasing the frequency of potentially deleterious alleles and reducing fitness. However, the extent to which cancers play a role in the fitness reduction of genetically depauperate wildlife populations is unknown. Santa Catalina island foxes (Urocyon littoralis catalinae) sampled in 2007-2008 have a high prevalence of ceruminous gland tumors, which was not detected in the population prior to a recent bottleneck caused by a canine distemper epidemic. The disease appears to be associated with inflammation from chronic ear mite (Otodectes) infections and secondary elevated levels of Staphyloccus pseudointermedius bacterial infections. However, no other environmental factors to date have been found to be associated with elevated cancer risk in this population. Here, we used whole genome sequencing of the case and control individuals from two islands to identify candidate loci associated with cancer based on genetic divergence, nucleotide diversity, allele frequency spectrum, and runs of homozygosity. We identified several candidate loci based on genomic signatures and putative gene functions, suggesting that cancer susceptibility in this population may be polygenic. Due to the efforts of a recovery program and weak fitness effects of late-onset disease, the population size has increased, which may allow selection to be more effective in removing these presumably slightly deleterious alleles. Long-term monitoring of the disease alleles, as well as overall genetic diversity, will provide crucial information for the long-term persistence of this threatened population.

Role of microhabitat and temporal activity in facilitating coexistence of endemic carnivores on the California Channel Islands

Most extinctions worldwide have occurred on islands, primarily due to interactions with exotic species, but rarely due to interactions among endemic species. This potential exists on two of the California Channel Islands, Santa Cruz and Santa Rosa, as endemic island spotted skunks (Spilogale gracilis amphiala) appear to have rapidly declined to rarity, possibly due to interference competition with endemic island foxes (Urocyon littoralis). Niche partitioning is expected in response to interference competition; hence, it is surprising that skunks and foxes show extensive overlap in macrohabitat use and circadian activity. However, the role of microhabitat associations and fine-scale temporal activity in facilitating coexistence of the two species has not been explored. We evaluated microhabitat associations of island spotted skunks and island foxes across both islands using data from live traps and wildlife cameras collected during 2015–2017, and we analyzed fine-scale temporal activity using camera data from 2016 and 2017. On both islands, skunks had a positive response to ground-level cover provided by rugged topography or woody vegetation such as shrubs or trees, whereas foxes had a weak or negative response, suggesting partitioning of these microhabitat characteristics. Additionally, on both islands the peak in timing of skunk activity offset the peak for foxes, which implies that skunks use fine-scale adjustments in activity to avoid foxes. Past grazing by exotic herbivores likely reduced habitat refuges for skunks; however, as vegetation on both islands recovers, regrowth of shrubs and trees may provide cover that will improve prospects for coexistence of island spotted skunks and island foxes.

Developmental features of the canid proximocaudal femur

We continued direct morphological studies of the canid coxofemoral joint, considering early-life spatial relationships around the locus of the proximocaudal joint capsule insertion. Our primary goal was to elucidate the postnatal developmental gross anatomy of the proximocaudal femur, among juveniles across Canidae. From an original database of 267 independent (museum) specimens from 11 canid taxa and 1 hybrid taxon, we identified 29 ancient or modern candidate juvenile specimens (nine taxa and one hybrid taxon). Based on optimal ability to recognize landmarks, the best photographic data were categorized into five groups of four each (n = 20). The data groups approximated early juvenile, early-mid juvenile, mid-juvenile, mid-late juvenile; and young adult stages. In this descriptive photographic essay, we demonstrate the developmental spatial proximity among (a) the dorsal meeting of the respective lateral and medial extensions from the growth centers of the femoral head and greater trochanter; (b) the caudodorsal aspect of the coxofemoral joint capsule attachment; (c) a segment of the proximocaudal femoral shaft physis; and (d) an eventual associated mineralized prominence. The latter occurs frequently but not universally, suggesting natural population variability across taxa. Across taxa and juvenile age categories, the morphology thus supports developmental conservation among ancient and modern Canidae. The biomechanical and biological cause-effect implications are not yet clear. For zoological purposes, we apply the term postdevelopmental mineralized prominence to the residual caudolateral surface feature. We extend the original anatomical work of Morgan in zoological and phylogenic arenas, using direct observation of cleared skeletal specimens.

The nature of coxofemoral joint pathology across family Canidae

We evaluated coxofemoral joints from museum specimens of: Vulpes lagopus; Vulpes vulpes; Vulpes velox; Nyctereutes procyonoides; Urocyon cinereoargenteus; Aenocyon [Canis] dirus; Canis latrans; Canis lupus lupus; Canis lupus familiaris; C. l. familiaris × latrans; and Canis dingo. Acetabular components included: fossa; articular surface; medial and lateral articular margins; and periarticular surfaces. Acetabular components variably revealed: osteophyte-like features; varying appearance of articular margin rims (especially contour changes); rough bone surfaces (especially fossa and articular surface); and surface wear. Proximal femoral components included: articular surface; articular margin; periarticular surfaces; and joint capsule attachment. Femoral components variably revealed: rough bone surface; bone loss; articular margin osteophyte-like features; caudal post-developmental mineralized prominence; and enthesophytes along the joint capsule attachment. Non-metric multidimensional scaling was used to analyze right-left asymmetric relationships between observed traits, across taxa. Significantly different acetabular trait asymmetry involved only C. latrans-C. l. familiaris; V. vulpes-N. procyonoides, and U. cinereoargenteus-N. procyonoides. There were no significant lateralized differences in proximal femoral traits involving modern canids, ancient and modern C. l. familiaris, or modern vulpines. Thus, the observations were strongly bilateral. We hypothesized high similarity of traits across taxa. The data confirm the hypothesis and strongly suggest broad and deep morphological and mechanistic conservation that almost certainly pre-existed (at least) all modern canids. Further zoological studies are needed to evaluate phylogenic implications in greater detail.

Mitochondrial Genomes of the United States Distribution of Gray Fox (Urocyon cinereoargenteus) Reveal a Major Phylogeographic Break at the Great Plains Suture Zone

We examined phylogeographic structure in gray fox (Urocyon cinereoargenteus) across the United States to identify the location of secondary contact zone(s) between eastern and western lineages and investigate the possibility of additional cryptic intraspecific divergences. We generated and analyzed complete mitochondrial genome sequence data from 75 samples and partial control region mitochondrial DNA sequences from 378 samples to investigate levels of genetic diversity and structure through population- and individual-based analyses including estimates of divergence (FST and SAMOVA), median joining networks, and phylogenies. We used complete mitochondrial genomes to infer phylogenetic relationships and date divergence times of major lineages of Urocyon in the United States. Despite broad-scale sampling, we did not recover additional major lineages of Urocyon within the United States, but identified a deep east-west split (∼0.8 million years) with secondary contact at the Great Plains Suture Zone and confirmed the Channel Island fox (Urocyon littoralis) is nested within U. cinereoargenteus. Genetic diversity declined at northern latitudes in the eastern United States, a pattern concordant with post-glacial recolonization and range expansion. Beyond the east-west divergence, morphologically-based subspecies did not form monophyletic groups, though unique haplotypes were often geographically limited. Gray foxes in the United States displayed a deep, cryptic divergence suggesting taxonomic revision is needed. Secondary contact at a common phylogeographic break, the Great Plains Suture Zone, where environmental variables show a sharp cline, suggests ongoing evolutionary processes may reinforce this divergence. Follow-up study with nuclear markers should investigate whether hybridization is occurring along the suture zone and characterize contemporary population structure to help identify conservation units. Comparative work on other wide-ranging carnivores in the region should test whether similar evolutionary patterns and processes are occurring.