Mitochondrial genomes reveal the pattern and timing of marten (Martes), wolverine (Gulo), and fisher (Pekania) diversification

https://doi.org/10.1016/j.ympev.2014.08.002Get rights and content

Highlights

  • Pekania constitutes a taxon that is phylogenetically apart and well differentiated from Martes.

  • Genus Martes containing Charronia is monophyletic to the exclusion of Gulo.

  • Subgenus Martes phylogeny is: foina (americana (melampus (zibellina, martes))).

  • Fossils older than the Tortonian–Messinian transition do not represent Martes.

Abstract

Despite recent advances in understanding the pattern and timescale of evolutionary diversification in the marten, wolverine, fisher, and tayra subfamily Guloninae (Mustelidae, Carnivora), several important issues still remain contentious. Among these are the phylogenetic position of Gulo relative to the subgenera of Martes (Martes and Charronia), the phylogenetic relationships within the subgenus Martes, and the timing of gulonine divergences. To elucidate these issues we explored nucleotide variation in 11 whole mitochondrial genomes (mitogenomes) from eight gulonine species and two outgroup meline species. Parsimony, maximum likelihood, and Bayesian phylogenetic analyses yielded fully resolved and identical patterns of relationships with high support for all divergences. The generic status of Pekania (P. pennanti), the monophyly of the genus Martes containing M. flavigula (subgenus Charronia) to the exclusion of the genus Gulo (G. gulo), and the M. foina (M. americana (M. melampus (M. zibellina, M. martes))) phylogeny of the subgenus Martes were strongly supported. Dating analyses (BEAST) using a set of five newly applied fossil calibrations provided divergence times considerably younger than previous multigene mitochondrial estimates, but similar to multigene nuclear and nuclear–mitochondrial estimates. The 95% confidence (highest posterior density) intervals of our divergence times fell within those inferred from nuclear and nuclear–mitochondrial sequence data, and were markedly narrower than in earlier studies (whether nuclear, mitochondrial, or combined). Notably, and contrary to long-held beliefs, our findings indicate that fossils older than the Tortonian–Messinian transition (late Late Miocene) do not represent Martes, excluding from this genus its putative members from the Early, Middle, and early Late Miocene. This study demonstrates the high informativeness of the mitogenome for phylogenetic inference and divergence time estimation within Guloninae, and suggests that mitogenomes can be highly informative also for other clades at similar levels of evolutionary divergence.

Introduction

The clade containing the martens (genus Martes, including M. foina, M. martes, M. zibellina, M. melampus, M. americana, M. flavigula, and M. gwatkinsii), the wolverine (Gulo gulo), the fisher (Pekania pennanti), and the tayra (Eira barbara) has recently been ranked as a subfamily and referred to as either Martinae (e.g., Fulton and Strobeck, 2006, Koepfli et al., 2008, Yu et al., 2011) or Guloninae (e.g., Sato et al., 2009, Sato et al., 2012, Wolsan and Sato, 2010, Wolsan, 2013). Here we use Guloninae Gray, 1825 because it has priority over Martinae Wagner, 1841 under the International Code of Zoological Nomenclature (Sato et al., 2009). It should be noted that the name Guloninae has indeed been referred recently to two different clades: (1) the crown clade of Martes, Gulo, Pekania, and Eira, and (2) the total clade of these four genera composed of the crown clade and its extinct paraphyletic stem group. Here we apply Guloninae to the crown clade, which follows the phylogenetic definition of this name as proposed in Wolsan and Sato, 2007, Wolsan and Sato, 2010.

Mainly due to ecomorphological differences, Martes and Gulo were long regarded as distantly related. Although regularly classified in the same carnivoran family (Mustelidae), these genera were often assigned to different subfamilies or tribes (e.g., Anderson, 1989, McKenna and Bell, 1997, Ginsburg, 1999, Pavlinov, 2003). Some morphological and ethological observations (Pocock, 1920, Krott, 1959, Rabeder, 1976, Sotnikova, 1982) and the results of early molecular studies (Dragoo and Honeycutt, 1997, Koepfli and Wayne, 1998, Hosoda et al., 2000, Stone and Cook, 2002) suggested a closer affinity between both genera, but it is only in recent years that evidence pointing to the monophyly of these mustelids has been presented (Koepfli and Wayne, 2003, Koepfli et al., 2008, Wolsan and Sato, 2010, Yu et al., 2011, Sato et al., 2012). This evidence also indicated that the fisher (which was formerly commonly treated as congeneric with martens and accordingly referred to as Martes pennanti) is outside the Martes and Gulo clade. The fisher has therefore been reclassified in its own genus (Wolsan and Sato, 2005, Wolsan and Sato, 2007, Koepfli et al., 2008, Sato et al., 2012).

Molecular investigations within the last 20 years (Carr and Hicks, 1997, Dragoo and Honeycutt, 1997, Hosoda et al., 1997, Hosoda et al., 2000, Koepfli and Wayne, 1998, Koepfli and Wayne, 2003, Stone and Cook, 2002, Sato et al., 2003, Sato et al., 2009, Sato et al., 2012, Koepfli et al., 2008, Wolsan and Sato, 2010, Yu et al., 2011; and others) have considerably extended and refined knowledge on the pattern and timescale of gulonine phylogeny. Yet there are important issues that still remain to be elucidated, such as (1) the relative phylogenetic placement of Pekania and Eira, (2) the position of Gulo with respect to the subgenera Martes (containing M. foina [type species], M. martes, M. zibellina, M. melampus, and M. americana) and Charronia (containing M. flavigula [type species] and M. gwatkinsii), (3) the relationships within the subgenus Martes, and (4) the timing of gulonine divergences.

Various molecular markers offer specific advantages for evolutionary studies. The characteristics of mitochondrial DNA (mtDNA) such as relatively high nucleotide substitution rates, lack of recombination, uniparental inheritance, short coalescence time, and high copy number made mtDNA markers the choice of molecular evolutionary studies in the past three decades. Although these markers can be less accurate than nuclear DNA (nDNA) markers at deep levels of evolutionary divergence because of the higher substitution rate and hence faster accumulation of homoplasies over time, mtDNA markers can be more informative and effective than nDNA markers at shallow levels of divergence because the higher rates of sequence evolution in mtDNA enable rapid accumulation of informative character changes useful for phylogeny reconstruction and divergence time estimation. As mtDNA has a single gene history, whole-genome mtDNA sequence data can be used in a concatenated format. The analysis of the whole or majority of mitochondrial genome (mitogenome) data is more likely to provide accurate results than analyses of one or a few mtDNA markers because mitogenomes contain increased phylogenetic signal, which reduces the effect of homoplasy, and because they include increased number of characters, which decrease stochastic errors. With the recent rapid growth in the taxonomic coverage of complete mtDNA sequences, mitogenome analyses have become increasingly popular. These analyses have been shown to be useful for phylogenetic reconstruction and divergence time estimation across a wide range of vertebrate taxa from fishes (e.g., Inoue et al., 2010) through amphibians (e.g., San Mauro et al., 2014) to birds (e.g., Pacheco et al., 2011) and mammals (e.g., Pozzi et al., 2014).

In this study we used comparative nucleotide sequence information from whole mitogenomes to resolve the phylogenetic relationships between Gulo and the subgenera of Martes and among all extant species of the subgenus Martes, as well as to re-estimate Martes, Gulo, and Pekania divergence times.

Section snippets

Sequence data sets

Complete mitogenome sequences were gathered for eight species of Guloninae and two species of Melinae (Table 1). All extant gulonine species except Eira barbara and Martes gwatkinsii were sampled. An entire mitogenome from Martes martes was used for the first time. The melines were employed as an outgroup in all analyses. Choice of this outgroup was based on the results of recent molecular studies, which have recovered Melinae among close relatives of Guloninae (Sato et al., 2012 and references

Homoplasy levels

Given that the lesser the slope and R2 of the linear regression of patristic versus p distances the more homoplastic the data set (Table 2), 2CD was the least homoplastic partition; 3CD, CR, and, in a lesser degree, tRNA were more homoplastic than other partitions; and rRNA and 1CD were intermediate. MTG, MTGexclCR, and PCG showed similar levels of homoplasy, with MTG being the least homoplastic, PCG the most, and MTGexclCR intermediate.

Phylogeny reconstruction

All phylogenetic analyses on MTG, MTGexclCR, and PCG

Monophyly of the genus Martes

Until very recently, the genus Martes was commonly circumscribed to contain three subgenera: Martes, Charronia, and Pekania. This classification derived from morphological observations (Anderson, 1970) and has been questioned by molecular investigations, which have, with growing evidence, suggested the paraphyly of this circumscription relative to Gulo (Hosoda et al., 2000, Stone and Cook, 2002, Koepfli and Wayne, 2003, Sato et al., 2003, Flynn et al., 2005; and others). The identity of Pekania

Conclusions

The generic status of Pekania and the monophyly of Martes composed of the subgenera Martes and Charronia are strongly supported in this study, which presents a compelling resolution for the interrelationships of these taxa and Gulo. A long-awaited conclusive solution to the riddle of phylogenetic relations among M. martes, M. zibellina, M. melampus, and M. americana is also provided.

The accuracy and precision of previous gulonine divergence time estimates derived from multigene nDNA and

Author contributions

B.L. and M.W. conceived and designed the study; B.L., M.W., D.W., and Z.Z. collected the data; B.L., M.W., W.Z., and Y.X. analyzed the data; M.W. drafted the manuscript; and M.W., B.L., and Y.X. contributed to the final manuscript.

Acknowledgments

We thank three anonymous reviewers for their critical reading of the manuscript and constructive comments and suggestions. This study was supported by the National Natural Science Foundation of China (grant 31370392), the Special Fund of Forestry Industrial Research for Public Welfare of China (grant 201304809), and the National Science Centre, Poland (grant 2012/06/M/NZ8/00189).

References (124)

  • L. Pozzi et al.

    Primate phylogenetic relationships and divergence dates inferred from complete mitochondrial genomes

    Mol. Phylogenet. Evol.

    (2014)
  • D. San Mauro et al.

    Life-history evolution and mitogenomic phylogeny of caecilian amphibians

    Mol. Phylogenet. Evol.

    (2014)
  • J.J. Sato et al.

    Deciphering and dating the red panda’s ancestry and early adaptive radiation of Musteloidea

    Mol. Phylogenet. Evol.

    (2009)
  • J.J. Sato et al.

    Evolutionary and biogeographic history of weasel-like carnivorans (Musteloidea)

    Mol. Phylogenet. Evol.

    (2012)
  • K.D. Stone et al.

    Molecular evolution of Holarctic martens (genus Martes, Mammalia: Carnivora: Mustelidae)

    Mol. Phylogenet. Evol.

    (2002)
  • W.G. Thompson et al.

    A radiometric calibration of the SPECMAP timescale

    Quat. Sci. Rev.

    (2006)
  • E. Anderson

    Quaternary evolution of the genus Martes (Carnivora, Mustelidae)

    Acta Zool. Fenn.

    (1970)
  • E. Anderson

    The phylogeny of mustelids and the systematics of ferrets

  • E. Anderson

    Evolution, prehistoric distribution, and systematics of Martes

  • E. Anderson

    The oldest known Pleistocene marten

    Martes Work. Group Newsl.

    (1997)
  • E. Anderson

    The Carnivora from Porcupine Cave

  • Anonymous

    Marte fossile

    Écho Monde Savant

    (1839)
  • A.D. Barnosky

    A summary of fossilized species in Porcupine Cave

  • A.D. Barnosky et al.

    Age and correlation of key fossil sites in Porcupine Cave

  • G.F. Baryshnikov

    Pleistocene Mustelidae (Carnivora) from Paleolithic site in Kudaro caves in the Caucasus

    Russ. J. Theriol.

    (2010)
  • G.F. Baryshnikov et al.

    Sredneplejstocenovye hiŝnye mlekopitaûŝie (Carnivora, Mammalia) Srednej Azii

    Tr. Zool. Inst. RAN

    (1994)
  • J.A. Baskin

    Mustelidae

  • J.A. Baskin

    Carnivora from the late Miocene Love Bone Bed of Florida

    Bull. Fla. Mus. Nat. Hist.

    (2005)
  • P.R. Bjork

    The Carnivora of the Hagerman local fauna (late Pliocene) of southwestern Idaho

    Trans. Am. Philos. Soc. New Ser.

    (1970)
  • G. Brunner

    Eine präglaciale Fauna aus dem Windloch bei Sackdilling (Oberpfalz)

    Neues Jahrb. Miner. Geol. Paläontol. Abh. B

    (1933)
  • S.M. Carr et al.

    Are there two species of marten in North America? Genetic and evolutionary relationships within Martes

  • T. Czyżewska

    Natural endocranial casts of the Mustelinae from Węże I near Działoszyn (Poland)

    Acta Zool. Crac.

    (1981)
  • R. Dehm

    Altpleistocäne Säuger von Schernfeld bei Eichstätt in Bayern

    Mitt. Bayer. Staatssamml. Paläontol. Hist. Geol.

    (1962)
  • J.W. Dragoo et al.

    Systematics of mustelid-like carnivores

    J. Mammal.

    (1997)
  • A.J. Drummond et al.

    BEAST: Bayesian evolutionary analysis by sampling trees

    BMC Evol. Biol.

    (2007)
  • A.J. Drummond et al.

    Relaxed phylogenetics and dating with confidence

    PLoS Biol.

    (2006)
  • J.J. Flynn et al.

    Molecular phylogeny of the Carnivora (Mammalia): assessing the impact of increased sampling on resolving enigmatic relationships

    Syst. Biol.

    (2005)
  • M. Fourment et al.

    PATRISTIC: a program for calculating patristic distances and graphically comparing the components of genetic change

    BMC Evol. Biol.

    (2006)
  • S.J. Friedmann et al.

    Magnetostratigraphic constraints on the age of Pleistocene fossiliferous strata in Porcupine Cave’s DMNH Velvet Room excavation

  • L. Ginsburg

    Order Carnivora

  • J. Głazek et al.

    On the stratigraphic position of Węże I locality (Middle Poland)

  • R.W. Graham et al.

    Late Quaternary distribution of Martes in North America

  • J.E. Gray

    An outline of an attempt at the disposition of Mammalia into tribes and families, with a list of the genera apparently appertaining to each tribe

    Ann. Philos. (New Ser.)

    (1825)
  • F. Heller

    Eine Forest-Bed-Fauna aus der Sackdillinger Höhle (Oberpfalz)

    Neues Jahrb. Miner. Geol. Paläontol. Abh. B

    (1930)
  • F. Heller

    Ein Nachtrag zur Forest-Bed-Fauna aus der Sackdillinger Höhle (Oberpfalz)

    Centralbl. Miner. Geol. Paläontol. B

    (1933)
  • T. Hosoda et al.

    Phylogenetic relationships within Martes based on nuclear ribosomal DNA and mitochondrial DNA

  • T. Hosoda et al.

    Evolutionary trends of the mitochondrial lineage differentiation in species of genera Martes and Mustela

    Genes Genet. Syst.

    (2000)
  • T. Hosoda et al.

    Phylogenetic history of mustelid fauna in Taiwan inferred from mitochondrial genetic loci

    Can. J. Zool.

    (2011)
  • S.S. Hughes

    Synthesis of Martes evolutionary history

  • J.G. Inoue et al.

    Evolutionary origin and phylogeny of the modern holocephalans (Chondrichthyes: Chimaeriformes): a mitogenomic perspective

    Mol. Biol. Evol.

    (2010)
  • Cited by (31)

    • Wolverines (Gulo gulo) in a changing landscape and warming climate: A decadal synthesis of global conservation ecology research

      2022, Global Ecology and Conservation
      Citation Excerpt :

      Extant wolverines spans the Holarctic but Gulo’s evolutionary origin is debated. Mitogenomic analysis suggests Gulo separated from close relatives Pekania and Martes in the late Miocene, 8MYA (Li et al., 2014). However, a recent USA fossil predates other Gulo records by a million years (Samuels et al., 2018) suggesting Gulo arose in the Nearctic and spread to the Palearctic during the Pliocene era.

    • Evolutionary progression of mitochondrial gene rearrangements and phylogenetic relationships in Strigidae (Strigiformes)

      2018, Gene
      Citation Excerpt :

      The variances were estimated with Bootstrap method (500 replications). The inferring phylogenetic relationships can be affected by mtDNA datasets and partitioning strategies in vertebrates (Duchene et al., 2011; Li et al., 2014; Powell et al., 2013; Meiklejohn et al., 2014; Urantowka et al., 2017). For ardeid birds, Zhou et al. (2014) performed phylogenetic analysis using nucleotide sequences of the two rRNAs and 12 PCGs (excluding the stop codon) encoded on heavy strand (H-strand), while Haring et al. (2001) conducted phylogenetic analysis using the nucleotide sequences of PCGs alone among the raptors.

    • Beyond the raccoon roundworm: The natural history of non-raccoon Baylisascaris species in the New World

      2017, International Journal for Parasitology: Parasites and Wildlife
      Citation Excerpt :

      When larvae migrated to the lungs, very few were retained, and even fewer were found in the small intestine after migration was completed (Babero, 1960b). Baylisascaris devosi infects North American mustelids belonging to the clade Guloninae (martens, fishers and wolverines), which are all carnivorous mammals that inhabit forests in the northern United States and Canada (Ruggiero et al., 1994; Li et al., 2014) (Table 5). B. devosi was formally described by Sprent (1952a) showing that parasites from fisher (Martes pennant) and marten (Martes americana) were distinct from B. columnaris from striped skunks.

    • The complete mitochondrial genome of the Tibetan fox (Vulpes ferrilata) and implications for the phylogeny of Canidae

      2016, Comptes Rendus - Biologies
      Citation Excerpt :

      mtDNA typically comprises a small closed circular DNA strand that generally encodes 37 genes: 13 protein-coding genes (PCGs), 2 rRNAs, and 22 tRNAs [6–8]. Compared with nuclear genes, mtDNA is characteristic of maternal inheritance, a high substitution rate, an absence of recombination, and a high copy number, which make it more effective in analyses [7,9,10]. In addition, the proteins involved in oxidative phosphorylation have provided insight into the molecular basis of adaptation to extreme environments [8,11–13].

    View all citing articles on Scopus
    View full text