Research ArticleEVOLUTIONARY GENETICS

Whole-genome sequence analysis shows that two endemic species of North American wolf are admixtures of the coyote and gray wolf

See allHide authors and affiliations

Science Advances  27 Jul 2016:
Vol. 2, no. 7, e1501714
DOI: 10.1126/sciadv.1501714
  • Fig. 1 Admixture proportions, hypothesized branching patterns, and the geographic distribution of Canis in North America.

    Top: Previously proposed phylogenetic relationships among Canis lineages, with gray lines indicating putative admixture events (5). Bottom: Geographic distributions of Canis in North America. Sample locations are indicated by dots and abbreviations are described in Table 1. Ancestry proportions from vonHoldt et al. (5) are indicated (proportion gray wolf/proportion coyote; see also new values in Table 3). IRNP, Isle Royale National Park; Ma, million years ago.

  • Fig. 2 PCA of 5.4 million unphased SNPs and 23 Canis genomes.

    The dashed line contains genomes that are considered admixed.

  • Fig. 3 Estimates of ancestry proportions using the Embedded Image statistic.

    Sequences grouped as coyotes are from Alabama, California, Illinois, Ohio, and Florida. (*Individual labeled as wolf but is likely to derive from a coyote; see discussion in the text.)

  • Fig. 4 Demographic history inferred using G-PhoCS.

    A schematic depiction of the population phylogeny assumed in the analysis. The phylogeny was augmented with migration bands from all canids to the red wolf and the Great Lakes region wolf. G-PhoCS infers significant rates of gene flow primarily from the gray wolf and the coyote to the red wolf and the Great Lakes region wolf (shaded box). Ninety-five percent Bayesian credible intervals are shown for the total rates transformed into proportions between 0 and 100% (see Materials and Methods). Similarly high rates were also inferred when assuming three alternative topologies for the population phylogeny (fig. S4).

  • Table 1 Samples, origin, and genome code used in the article; average genome coverage; and ancestry proportions.

    If a population resides within the gray wolf and coyote hybrid zone, the location is indicated with “HZ”; reference populations are indicated by “REF.” When previously sequenced, the appropriate citation is provided. EuGW, Eurasian gray wolf; NAGW, North American gray wolf; RW, red wolf; GLW, Great Lakes region wolf; COY, coyote; DOG, dog; JACK, golden jackal.

    Genome (reference)Species, common nameLocationGroup codeAverage coverage
    Alaska (26)C. lupus, gray wolfAlaska, USANAGW6.37
    AlabamaC. latrans, coyoteAlabama, USA (REF)COY5.10
    Algonquin1C. lupus, eastern wolf*Algonquin Provincial Park, Canada (HZ)GLW5.60
    Algonquin2C. lupus, eastern wolf*Algonquin Provincial Park, Canada (HZ)GLW4.48
    Boxer (26)C. familiaris, dog (boxer)Domestic dogDOG22.31
    Basenji (63)C. familiaris, dog (Basenji)Domestic dogDOG21.93
    California (26)C. latrans, coyoteCalifornia, USA (REF)COY24.25
    FloridaC. latrans, coyoteFlorida, USACOY7.10
    Minnesota (26)C. lupus, Great Lakes wolf*Minnesota, USA (HZ)GLW24.90
    GShep (63)C. familiaris, dog (German shepherd)Domestic dogDOG23.86
    IllinoisC. latrans, coyoteIllinois, USACOY6.21
    India (26)C. lupus, gray wolfIndia (REF)EuGW15.88
    Iran (54)C. lupus, gray wolfIran (REF)EuGW28.70
    IRNPC. lupus, Great Lakes wolf*IRNP, USA (HZ)GLW4.62
    Kenya (26)C. aureus, golden jackalKenya, AfricaJACK25.74
    Mexican (26)C. lupus baileyi, Mexican wolfNew Mexico, USANAGW23.88
    Mongolia (54)C. lupus, gray wolfMongolia (REF)EuGW25.65
    OhioC. latrans, coyoteOhio, USACOY5.92
    Qinghai (54)C. lupus, gray wolfChina (REF)EuGW27.03
    QuebecCoyC. latrans, coyoteQuebec, Canada (REF)COY6.41
    QuebecWolfC. lupus, gray wolf*Quebec, Canada (HZ)GLW3.88
    Redwolf1 (54)C. rufus, red wolfCaptive breeding program, USA (HZ)RW28.30
    Redwolf2C. rufus, red wolfCaptive breeding program, USA (HZ)RW5.57
    Redwolf3C. rufus, red wolfCaptive breeding program, USA (HZ)RW6.29
    WisconsinC. lupus, Great Lakes wolf*Wisconsin, USA(HZ)GLW7.49
    Yellowstone1C. lupus, gray wolfYellowstone National Park, USANAGW12.90
    Yellowstone2 (26)C. lupus, gray wolfYellowstone National Park, USANAGW26.39
    Yellowstone3 (26)C. lupus, gray wolfYellowstone National Park, USA (REF)NAGW24.32

    *Species designation uncertain: the Great Lakes region may contain a wolf-like endemic North American species, C. lycaon, only a mixture of gray wolf and coyote genes, or mixtures of all three taxa.

    †Because of their parent-offspring relationship (Yellowstone1 and Yellowstone2 are the parents of Yellowstone3), we excluded these two individuals.

    • Table 2 Pairwise FST estimates between canid lineages.

      Abbreviations are found in Table 1.

      NAGWCoyote (REF)Red wolfGLW
      EuGW0.0990.1600.1880.076
      NAGW0.1530.1770.057
      Coyote (REF)0.1080.045
      Red wolf0.085
    • Table 3 Estimated fraction of wolf-like versus coyote-like alleles at the 16,184 fixed differences between wolves and putatively unadmixed coyotes.

      See Table 1 for sample abbreviations.

      SampleNcoyoteNheterozygousNwolf% Wolf alleles% Wolf ancestry
      Florida7464189711423.60
      Illinois6192577762317.60
      Ohio7173300610323.40
      Redwolf12,4654657899829.79.3
      Redwolf22,1063136486436.420.3
      Redwolf32,2113983563335.518.8
      Alaska8,5442510106480.993.9
      Yellowstone110,4002654153680.493.1
      Yellowstone211,1503052187378.960.6
      Mexican11,639111433047685.8
      Algonquin12,5165508302147.739
      Algonquin22,3934855187852.847.4
      Minnesota6,714616531746161
      Isle Royale4,3722257198763.865.6
      QuebecWolf131807157621.30
      Wisconsin6,4085501220764.967.4
      Boxer12,0011026221982.195.9
      GShep12,6841317211582.897
      Basenji13,102832217283.998.8
    • Table 4 Fraction of unique (non-EuGW + COY-ref) alleles in the complete sequence data and in sequences downsampled to six times.
      SampleNew SNP%New SNP% (downsampled)Coverage
      Basenji*6.576.4221.93
      GShep6.822.5124.25
      Redwolf18.784.6228.30
      Minnesota7.133.9624.90
      Redwolf23.473.895.57
      Redwolf34.284.736.29
      Algonquin23.594.084.48
      Algonquin13.193.555.60
      IRNP2.092.244.62
      QuebecWolf2.413.223.88
      Wisconsin4.374.647.49
      FLcoy4.185.193.88
      ILcoy4.575.86.21
      OHcoy3.564.45.92

      *Genome was not downsampled.

      Supplementary Materials

      • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/2/7/e1501714/DC1

        Supplementary Results

        table S1. The D statistic test for admixture in coyotes and North American hybrids. Results are calculated separately for the X chromosome and the autosomes.

        table S2. Formula estimates for wolf and coyote ancestry proportions in each sample across the 38 autosomes.

        table S3. Estimates of total migration rates and migration proportions.

        table S4. Fraction of nonwolf alleles in the full-sequence data and in sequences downsampled to six times (in parenthesis).

        fig. S1. Profile likelihood curve for T, the time of split between gray wolves and coyotes under a simple isolation model (blue).

        fig. S2. PCA of downsampled genomes.

        fig. S3. Results of all possible D statistic tests for Eurasian wolf introgression into North American canids, following D(P1, P2, Eurasian wolf, fox), where P1 and P2 are all possible combinations of North American canids.

        fig. S4. Estimates of total migration rates under four different models assumed in four separate G-PhoCS runs.

        fig. S5. Estimates of population divergence times and effective population sizes in the four separate G-PhoCS runs.

        fig. S6. Estimates of divergence time obtained in eight separate runs of G-PhoCS.

      • Supplementary Materials

        This PDF file includes:

        • Supplementary Results
        • Legend for table S1
        • table S2. ˆf estimates for wolf and coyote ancestry proportions in each sample across the 38 autosomes.
        • table S3. Estimates of total migration rates and migration proportions.
        • table S4. Fraction of nonwolf alleles in the full-sequence data and in sequences downsampled to six times (in parenthesis).
        • fig. S1. Profile likelihood curve for T, the time of split between gray wolves and coyotes under a simple isolation model (blue).
        • fig. S2. PCA of downsampled genomes.
        • fig. S3. Results of all possible D statistic tests for Eurasian wolf introgression into North American canids, following D(P1, P2, Eurasian wolf, fox), where P1 and P2 are all possible combinations of North American canids.
        • fig. S4. Estimates of total migration rates under four different models assumed in four separate G-PhoCS runs.
        • fig. S5. Estimates of population divergence times and effective population sizes in the four separate G-PhoCS runs.
        • fig. S6. Estimates of divergence time obtained in eight separate runs of G-PhoCS.

        Download PDF

        Other Supplementary Material for this manuscript includes the following:

        • table S1 (Microsoft Excel format). The D statistic test for admixture in coyotes and North American hybrids.

        Files in this Data Supplement: