Author's personal copy
Molecular Phylogenetics and Evolution 54 (2010) 726–745
Contents lists available at ScienceDirect
Molecular Phylogenetics and Evolution
journal homepage: www.elsevier.com/locate/ympev
Dogs, cats, and kin: A molecular species-level phylogeny of Carnivora
ˇ Ingi Agnarsson a,c,*, Matjaz Kuntner c, Laura J. May-Collado a,b
Biology Department, University of Puerto Rico—Rio Piedras (UPR-RP), San Juan, PR 00931, Puerto Rico Department of Environmental Science and Policy, George Mason University, MSN 5F2, 4400 University Drive, Fairfax, VA 22030, USA c Institute of Biology, Scientiﬁc Research Centre of the Slovenian Academy of Sciences and Arts, Novi trg 2, P.O. Box 306, SI-1001 Ljubljana, Slovenia
a r t i c l e
i n f o
a b s t r a c t
Phylogenies underpin comparative biology as high-utility tools to test evolutionary and biogeographic hypotheses, inform on conservation strategies, and reveal the age and evolutionary histories of traits and lineages. As tools, most powerful are those phylogenies that contain all, or nearly all, of the taxa of a given group. Despite their obvious utility, such phylogenies, other than summary ‘supertrees’, are currently lacking for most mammalian orders, including the order Carnivora. Carnivora consists of about 270 extant species including most of the world’s large terrestrial predators (e.g., the big cats, wolves, bears), as well as many of man’s favorite wild (panda, cheetah, tiger) and domesticated animals (dog, cat). Distributed globally, carnivores are highly diverse ecologically, having occupied all major habitat types on the planet and being diverse in traits such as sociality, communication, body/brain size, and foraging ecology. Thus, numerous studies continue to address comparative questions within the order, highlighting the need for a detailed species-level phylogeny. Here we present a phylogeny of Carnivora that increases taxon sampling density from 28% in the most detailed primary-data study to date, to 82% containing 243 taxa (222 extant species, 17 subspecies). In addition to extant species, we sampled four extinct species: American cheetah, saber-toothed cat, cave bear and the giant short-faced bear. Bayesian analysis of cytochrome b sequences data-mined from GenBank results in a phylogenetic hypothesis that is largely congruent with prior studies based on fewer taxa but more characters. We ﬁnd support for the monophyly of Carnivora, its major division into Caniformia and Feliformia, and for all but one family within the order. The only exception is the placement of the kinkajou outside Procyonidae, however, prior studies have already cast doubt on its family placement. In contrast, at the subfamily and genus level, our results indicate numerous problems with current classiﬁcation. Our results also propose new, controversial hypotheses, such as the possible placement of the red panda (Ailuridae) sister to canids (Canidae). Our results conﬁrm previous ﬁndings suggesting that the dog was domesticated from the Eurasian wolf (Canis lupus lupus) and are congruent with the Near East domestication of the cat. In sum, this study presents the most detailed species-level phylogeny of Carnivora to date and a much needed tool for comparative studies of carnivoran species. To demonstrate one such use, we perform a phylogenetic analysis of evolutionary distinctiveness (EDGE), which can be used to help establish conservation priorities. According with those criteria, and under one of the many possible sets of parameters, the highest priority Carnivora species for conservation of evolutionary diversity include: monk seals, giant and red panda, giant otter, otter civet, Owston’s palm civet, sea otter, Liberian mongoose, spectacled bear, walrus, binturong, and the fossa. Published by Elsevier Inc.
Article history: Received 10 February 2009 Revised 18 October 2009 Accepted 29 October 2009 Available online 10 November 2009 Keywords: Adding taxa Ailuridae Bayesian Canidae Caniformia Cytochrome b Extinct species Eupleridae Felidae Feliformia Herpestidae Hyaenidae Mammalia Mephitidae Mustelidae Nandiniidae Odobenidae Otariidae Perissodactyla Phocidae Pholidota Pinnipedia Prionodontidae Procyonidae Taxon sampling Ursidae Viverridae
1. Introduction Detailed species-level phylogenies represent the sine qua non for hypothesis testing in comparative biology (Felsenstein, 1985; Harvey and Pagel, 1991). Particularly valuable are phylogenies that sample taxa densely, are well resolved, and contain estimates of
* Corresponding author. Address: Biology Department, University of Puerto Rico—Rio Piedras (UPR-RP), San Juan, PR 00931, Puerto Rico. E-mail address: email@example.com (I. Agnarsson). URL: http://www.theridiidae.com (I. Agnarsson). 1055-7903/$ - see front matter Published by Elsevier Inc. doi:10.1016/j.ympev.2009.10.033
the branching history of the tree, such as branch lengths, in addition to revealing the interrelationships of species. Such phylogenies are high-utility tools that can be used to test evolutionary hypotheses, reveal coevolutionary histories of multiple traits, estimate ages of lineages and their biogeographical histories, and serve as tools for conservation, to name but a few of their uses. Despite their obvious utility, and the availability of data on GenBank, the most species-rich primary-data phylogenies available for most higher-level taxa (such as orders) within mammals contain only a small portion of the known extant species. This has sparked the reconstruction of consensus phylogenies using supertree methods
Author's personal copy
I. Agnarsson et al. / Molecular Phylogenetics and Evolution 54 (2010) 726–745
(Bininda-Emonds, 2005; Cardillo et al., 2004; Jones et al., 2002; Price et al., 2005) that stitch together many smaller phylogenies and/or taxonomies to provide a best-guess hypothesis for higherlevel taxa. As useful as supertrees have proved to be (BinindaEmonds, 2005; Bininda-Emonds et al., 2003) they are, however, constrained by the available input data, and thus the resulting trees often have relatively low resolution, with a high number of species placed without quantitative phylogenetic estimates (i.e., derived directly from taxonomy), and typically lacking accurate estimates of branch lengths. Therefore, it remains an urgent priority to provide more detailed primary phylogenies for mammals, and other organisms. Here we present a near species-complete phylogeny of Carnivora that increases taxon sampling density from 28% in the most detailed primary-data study to date (Flynn et al., 2005) to 82%. The order Carnivora consists of about 270 extant species that include many of the world’s top terrestrial predators (e.g., cheetah, lion, leopard, tiger, hyena, wolf), man’s favorite pets (dog, cat), and iconic wild animals (e.g., panda, polar bear, brown bear). Carnivores are highly diverse ecologically, having occupied all continents and all major habitat types on the planet and being diverse in traits such as sociality, communication, body/brain size, feeding ecology, etc. Thus, studies have addressed comparative questions within the order, such as the evolution of sociality (Dalerum, 2007), brain size (Finarelli, 2008; Finarelli and Flynn, 2007; Wroe and Milne, 2007), metabolic rate (Munoz-Garcia and Williams, 2005), and potential coevolution of such traits. Furthermore, the timing and the history of domestication of the cat and, especially the dog (ﬁrst domesticated animal (Leonard et al., 2002)), have played a major role in the foundation of evolutionary theory (Darwin, 1859) as well as in understanding the origin and cultural transition of modern man (Savolainen et al., 2002). Clearly, a detailed species-level phylogeny will provide a tool that can further advance many types of studies. Carnivoran species, both as top predators and as highly diverse, charismatic, and ‘popular’ animals also play an important role in conservation. As is true of mammalian species worldwide, many carnivores are experiencing population decline, some are currently at the brink of extinction (e.g., Falkland island fox, Caribbean monk seal, sea mink, Japanese sea lion) while many more are listed as in critical danger (e.g., Mediterranean monk seal, red wolf, Darwin’s fox, Malabar civet, Iberian lynx, and the California channel island fox) (http://www.iucnredlist.org/). Certainly, maintaining biodiversity at the top trophic level is an important task, and understanding the phylogenetic relationships among, and within, species can aid in conservation decisions where the aim is to maximize representation of not only species, but of phylogenetic diversity (Faith, 1992, 2007; Redding and Mooers, 2006; Isaac et al., 2007; Steel et al., 2007; Helmus et al., 2007; Graham and Fine, 2008). Carnivora has been divided into two superfamilies, Caniformia (dog-like carnivores) and Feliformia (cat-like carnivores). Caniformia is, traditionally, further divided into two infraorders Canoidea (or Cynoidea) that consist of Canidae (dogs and relatives) and Arctoidea which includes Ursidae (bears), Procyonidae (raccoons, coatis, and relatives), Mephitidae (skunks), Mustelidae (weasels, otters, and relatives), Otariidae (fur seals), Odobenidae (walrus), and Phocidae (‘true’ seals). Feliformia consists of the families Felidae (cats), Herspestidae (mongooses), Hyaenidae (hyenas), Priodontidae (linsangs), Nandiniidae (African palm civet), and Viverridae (genets, civets, and binturong). Recent phylogenetic studies, however, often differ in the interrelationships among families so that suprafamiliar classiﬁcation remains unstable both within Caniformia and Feliformia (see Section 4). As data continue to accumulate from various sources (morphology, nuclear and mitochondrial DNA, behavior, fossils, etc.) certain
high-level clades tend to be conﬂicted, while others are repeatedly supported by independent datasets. The latter form the backbone of current knowledge, clades that have withstood many tests and can essentially be considered as ‘known’ or as benchmark clades by which novel phylogenetic results can be judged (Agnarsson and May-Collado, 2008; May-Collado and Agnarsson, 2006). Here, we use the single gene for which data are available for majority of mammals, cytochrome b, to provide a near complete species-level phylogeny of Carnivora. The record of several strongly supported high-level clades from prior studies serves to assure us that the results ‘make sense’—a single-gene phylogeny that fails to recover many ‘known’ clades cannot be expected to represent a good working phylogenetic hypothesis at any level (May-Collado and Agnarsson, 2006; May-Collado et al., 2007; Agnarsson and May-Collado, 2008). In contrast, a phylogeny that captures known clades may also advance our understanding of relationships among species and clades that remain in dispute, or that have not yet been tested. The main conﬂicts in Carnivora phylogeny surround the interrelationships among families and some lower taxonomic groups, both within Caniformia and Feliformia. For instance, the placements of Canidae, Felidae, Ursidae, and of the red panda, have been continuously disputed with each new phylogenetic study. A series of studies, using a variety of data, have been designed in an attempt to clear these higher-level conﬂicts, or to address more detailed species-level relationships among a subclade of Carnivora (Arnason et al., 2007; Barycka, 2007; Gaubert and Begg, 2007; Higdon et al., 2007, 2008; Koepﬂi et al., 2007, 2008; Krause et al., 2008; Pages et al., 2008; Patou et al., 2008). These prior studies have included a fraction of the species diversity within Carnivora, with the largest analysis to date including only about 28% of the known species diversity (76 species) (Flynn et al., 2005). Considering the known importance of dense taxon sampling for phylogenetic accuracy (Agnarsson and May-Collado, 2008; Heath et al., 2008a,b; Hedtke et al., 2006) it is clear that a more complete taxon sampling is desirable. In this study we propose a phylogenetic hypothesis at the species-level for Carnivora including 222 extant species (out of 270 spp.) and several subspecies, along with four extinct species. This represents about 82% of the extant carnivoran species described to date. The phylogeny is based on a Bayesian analysis of cytb sequences, a marker that has proved to be reliable for mammals (Agnarsson and May-Collado, 2008). We evaluate our results based on the recovery of well established ‘benchmark clades’ and discuss the implications and potential uses of the new phylogeny for classiﬁcation, conservation, and comparative biology.
2. Materials and methods Cytochrome b sequences were downloaded from GenBank for 243 carnivoran taxa (see Table 1 for accession numbers), including 222 extant species, 17 subspecies, and four extinct species: two cats (a saber-toothed cat, Homotherium serum, and American cheetah Miracinonyx trumani), and two bears (cave bear, Ursus spelaeus and giant short-faced bear, Arctodus simus). As outgroups we selected 13 species of Perissodactyla and both extant species of Pholidota (pangolins), the putative sister group of Carnivora (Arnason et al., 2008; Nishihara et al., 2006). Because missing data can cause problems in phylogenetic reconstruction, we created two data matrices; one in which all taxa with available cytb sequences were included (full dataset = 258 taxa) and another where taxa with less than 50% of the full cytb sequence length were excluded (pruned dataset = 202 taxa). The sequences were aligned in Mesquite, a trivial task given that it is a protein-coding gene with no implied gaps. The appropriate model for the Bayesian analysis
/ Molecular Phylogenetics and Evolution 54 (2010) 726–745
Table 1 GenBank accession numbers for taxa included in the full and pruned matrix. schoutedeni Genetta tigrina Genetta thierryi Genetta victoriae Genetta rubiginosa Genetta servalina Genetta poensis Genetta pardina Genetta maculata Genetta johnstoni Genetta felina Genetta genetta Genetta cristata Genetta bourloni Genetta angolensis Cynogale bennettii Civettictis civetta Chrotogale owstoni Arctogalidia trivirgata Arctictis binturong Poiana richardsonii Paradoxurus jerdoni Paradoxurus hermaphroditus Paguma larvata Hemigalus derbyanus Viverricula indica Prionodon linsang Prionodon pardicolor Viverra zibetha Viverra tangalunga Viverra megaspila Family Hyaenidae Hyaena hyaena Crocuta crocuta Parahyaena brunnea Proteles cristatus Family Eupleridae Cryptoprocta ferox Fossa fossana Galidia elegans Galidictis fasciata Mungotictis decemlineata Salanoia concolor Family Herpestidae Helogale hirtula Herpestes urva Helogale parvula GenBankaccession number Full matrix(n = 274) Pruned matrix(n = 202)
Y07726 DQ318533 EU107377 AJ245723 EU257216 DQ470804 NC_001640 DQ464015 DQ223534 EU030681 DQ464013 AJ245725 AY152859 EU057637 NC_004027
Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â
Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â
AF522350 AY241900 AY241877 DQ395206 AY241894 AY751476 AY170115 AY241897 AY397707 AY241912 AY241892 AF125150 AY397700 AY241888 AY241898 DQ395207 DQ683992 AY170105 AF125144 AF125140 AY048793 AY241891 DQ683994 AF511056 AB303952 AF125143 AY241890 AF125141 AF511048 AF511047 AF511045 AF511046 AF511063 AY928676 AY928677 AY048792 AY928681 AF511062 AY170099 AY170100 AY170094 AY187007 AF522335 DQ519074 AF522333
Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â
Â Â Â
Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â
Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â
.Author's personal copy
I. Extinct species are marked with a cross. Agnarsson et al. GenBankscientiﬁc name Outgroups Perissodactyla Ceratotherium simum Coelodonta antiquitatis Diceros bicornis Dicerorhinus sumatrensis Equus asinus Equus burchellii Equus cab Equus hemionus Equus przewalskii Equus neogeus Equus hydruntinus Rhinoceros sondaicus Hippidion saldiasi Pholidota Manis pentadactyla Manis tetradactyla Carnivora Feliformia Family Nandiniidae Nandinia binotata Family Viverridae Genetta sp.
Full matrix(n = 274) Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â
Pruned matrix(n = 202) Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â
(1996) (1996) (1996) (1996) (1996)
AM711897 AF028135 AY333749 AM711902 NC_002008 NC_010340 L29415 U47038 AF028143 NC_008093 AY291432 AY291431 AY291433
Â Â Â Â Â Â Â Â Â Â Â Â Â Â
Â Â Â
Â Â Â Â (continued on next page)
. EF587169 EF587026 EF587025 Masuda et al. / Molecular Phylogenetics and Evolution 54 (2010) 726–745 Table 1 (continued) GenBankscientiﬁc name Herpestes auropunctatus Herpestes edwardsii Herpestes ichneumon Herpestes naso Herpestes javanicus Galerella sanguinea Galerella pulverulenta Cynictis penicillata Crossarchus obscurus Crossarchus alexandri Bdeogale nigripes Bdeogale crassicauda Atilax paludinosus Suricata suricatta Rhynchogale melleri Mungos mungo Liberiictis kuhni Ichneumia albicauda Family Felidae Acinonyx jubatus Catopuma temminckii Felis margarita Felis domesticus (silvestris catus) Felis chaus Felis nigripes Felis silvestris silvestris Felis silvestris ornata Felis silvestris caffra Felis lybica Lynx canadensis Lynx rufus Lynx pardinus Lynx lynx Leopardus tigrinus Leopardus pardalis Leopardus wiedi Homotherium serum Herpailurus yaguarondi Panthera leo Panthera tigris tigris Panthera tigris sumatrae Panthera tigris corbetti Panthera tigris altaica Panthera pardus Panthera onca Neofelis nebulosa Neofelis diardi Miracinonyx trumani Prionailurus bengalensis Prionailurus viverrinus Pardofelis marmorata Uncia (Panthera) uncia Puma concolor Otocolobus manul Oncifelis geoffroyi Oncifelis guigna Ictailurus planiceps Lynchailurus colocolo Caniformia Family Ailuridae Ailurus fulgens Family Canidae Atelocynus microtis Canis lupus pallipes Canis lupus lupus Canis lupus familiaris Canis lupus chanco Canis simensis Canis rufus Canis mesomelas elongae Canis latrans Canis indica Canis himalayensis Canis aureus GenBankaccession number X94926 AY170107 EF689052 AF522339 DQ519072 AF522331 AF522330 AF511060 AF522327 AF522326 AY950655 AY950654 AF522325 AY170111 AF522344 AF522347 AF522342 AF511058 NC_005212 AY773082 EF587036 X82296 EU362125 Masuda et al. AY928671 AY499331 EF689048 D28902 AY886752 DQ469951 Masuda et al. Agnarsson et al. Masuda et al. Masuda et al. Masuda et al. DQ097176 AY886751 AF384815 AF053018 AF053048 AF053050 AF053039 EF551002 EF437582 DQ257669 EF437579 DQ097175 AB210233 AB210239 AY499335 NC_010638 AY598487 Masuda et al. Masuda et al.Author's personal copy
/ Molecular Phylogenetics and Evolution 54 (2010) 726–745
GenBankaccession number AF028136 AF028139 AF028137 AF028154 AF028153 AF028152 AF028151 AF028148 AF028150 AF280070 AF028147 AF028145 AF028146 AB292765 AF042174 AF028156 AF028155 FM177762 U23554 U23562 U18899 L21881 NC_009492 AF007931 AF007925 AF007908 AF007915 AF007934 AF007911 AF264047 NC_003428 U18872 AY522430 EF587265 AY522429 NC_009331 AB020910 AJ428576 X82293 U18456 AM181021 AM181018 AF380893 X82292 AF380900 AY377329 AM422150 AM181017 AM181019 AY377328 AM181020 NC_004030 AM181016 AM181036 NC_008429 AM181029 AM181033 NC_001325 NC_008432 AM181031 NC_008425 NC_001602 AM181027 AM181028 AY377322 AM181022 AY377327 AM181023 AY424646
Full matrix(n = 274) Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â
Pruned matrix(n = 202)
Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â
Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â
.Author's personal copy
730 Table 1 (continued) GenBankscientiﬁc name Canis adustus Chrysocyon brachyurus Cuon alpinus Pseudalopex sechurae Pseudalopex gymnocercus Pseudalopex griseus Pseudalopex culpaeus Pseudalopex vetulus Otocyon megalotis Nyctereutes procyonoides Lycaon pictus Dusicyon thous Vulpes zerda Vulpes vulpes Vulpes macrotis Licalopex fulvipes Urocyon cinereoargenteus Speothos venaticus Family Ursidae Arctodus simus Tremarctos ornatus Melursus ursinus Helarctos malayanus Ursus americanus Ailuropoda melanoleuca Ursus americanus vancouveri Ursus americanus kermodei Ursus americanus cinnamomum Ursus americanus carlottae Ursus americanus americanus Ursus americanus altifrontalis Ursus spelaeus Ursus maritimus Ursus arctos Ursus thibetanus ussuricus Ursus thibetanus thibetanus Ursus thibetanus mupinensis Ursus thibetanus formosanus Selenarctos thibetanus Family Odobenidae Odobenus rosmarus rosmarus Family Otariidae Arctocephalus forsteri Arctocephalus tropicalis Arctocephalus townsendi Arctocephalus pusillus Arctocephalus philippii Arctocephalus gazella Arctocephalus galapagoensis Arctocephalus australis Zalophus wollebaeki Zalophus californianus Phocarctos hookeri Otaria byronia Neophoca cinerea Eumetopias jubatus Callorhinus ursinus Family Phocidae Phoca hispida Phoca groenlandica Phoca fasciata Phoca caspica Phoca vitulina Phoca sibirica Phoca largha Hydrurga leptonyx Halichoerus grypus Erignathus barbatus Cystophora cristata Ommatophoca rossii Monachus schauinslandi Monachus monachus Mirounga leonina Mirounga angustirostris
I. Agnarsson et al.
sampling the Markov chain every 1000 generations.000 generations (repeated twice). 2004) with a parsimony tree chosen as the basis for Modeltest. 1994). The Markov chain Monte Carlo search for each matrix was run with four chains for 20.000. / Molecular Phylogenetics and Evolution 54 (2010) 726–745 Table 1 (continued) GenBankscientiﬁc name Lobodon carcinophaga Leptonychotes weddellii Family Mephitidae Mephitis macroura Mephitis mephitis Spilogale gracilis Spilogale putorius Family Procyonidae Bassaricyon alleni Bassaricyon gabbii Bassariscus astutus Bassariscus sumichrasti Nasua narica Nasua nasua Potos ﬂavus Procyon cancrivorus Procyon lotor Family Mustelidae Vormela peregusna Taxidea taxus Pteronura brasiliensis Poecilogale albinucha Neovison vison Mustela strigidorsa Mustela sibirica Mustela putorius Mustela nudipes Mustela nivalis Mustela nigripes Mustela lutreola Mustela itatsi Mustela frenata Mustela eversmannii Mustela erminea Mustela altaica Melogale moschata Mellivora capensis Meles meles anakuma Martes zibellina Martes pennanti Martes melampus Martes foina Martes ﬂavigula Martes caurina Martes americana Lutrogale perspicillata Lutra sumatrana Lutra maculicollis Lutra lutra Lutra canadensis Lontra provocax Lontra longicaudis Lontra felina Lontra canadensis Galictis cuja Galictis vittata Ictonyx striatus Ictonyx libyca Gulo gulo Enhydra lutris Eira barbara Arctonyx collaris Aonyx capensis Amblonyx cinereus GenBankaccession number AM181024 AM181025 DQ471833 X94927 DQ471834 AM711898 DQ660299 DQ660300 AF498159 DQ660301 DQ660302 DQ660303 DQ660304 DQ660305 NC_009126 EF581360 AF057132 AF057126 EF472349 EF689076 EF987748 EF987747 EF987746 EF987745 EF987744 EF987743 EF987742 AB026104 AF498153 EF987741 AF457442 AB026100 AF498158 EF987755 NC_009677 EF987753 AF448248 EF987752 EF987750 EF987749 L77954 AF268274 EF472348 EF472347 AF057125 AF057124 AF418979 DQ341273 AF057123 AF057122 AF057121 EF987754 AF498155 AF498156 EF987739 NC_009685 AB291077 AF498154 AF498157 AF057118 AF057119 Full matrix(n = 274) Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â
Pruned matrix(n = 202) Â Â
Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â
Â Â Â Â Â Â Â Â Â Â Â Â Â
was selected with Modeltest (Posada and Crandall.2 (Huelsenbeck and Ronchist. second. Bayesian analysis was performed using MrBayes V3.
.000 generations were discarded as ‘‘burnin”.000.. 2001). The best model was GTR+C+I (Rodríguez et al.Author's personal copy
I.1. 1990. and the sample points of the ﬁrst 5. using the AIC criterion (Posada and Buckley. Yang. and third codon positions. 1998. Agnarsson et al. 2001) with settings as in May-Collado and Agnarsson (2006) and
Agnarsson and May-Collado (2008) with separate model estimation for ﬁrst.
1. Steel et al. The likelihood analysis (see Supplementary Fig. 2007). 2) recovers the ‘traditional’ Arctoidea. The best tree (lowest log likelihood) from these 200 analyses (see Supplementary Fig.g. I.
Maximum likelihood analyses were conducted in the program Garli (Zwickl. in turn. and support for it. fossa. We illustrate one example of such an analysis in Fig. 3. that
Fig. Summary of relationships among carnivoran families. 2 and Table 2). red panda). Genetta felina (Viverridae.. Isaac et al. Parsimony analyses were also conducted (see May-Collado and Agnarsson. parsimony results are in greater conﬂict with the available external evidence (morphology. although overall the results are quite similar. kinkajou (R. We focus our discussion on the Bayesian results as they simultaneously provide an estimate of the phylogeny. 1). e. the kinkajou which here. estimated their extinction risk as in between the ‘least concern’ and ‘near threatened’ categories. walrus (J. for analysis parameters using parsimony). 2007) and (2) the expected terminal branch length of species. due to the placement of the red panda (Ailurus fulgens. 8. To allow the inclusion of ‘data deﬁcient’ species we. arbitrarily. (2008) results of EDGE and HEDGE analyses are sensitive to how IUCN categories are translated to extinction probabilities. Agnarsson et al. and note that while results from other types of analyses differed in many details. Feliformia) with bears (Ursidae. We note that the ‘pessimistic’ transformation gives relatively great weight to phylogenetic distinctiveness as all taxa are considered at some considerable risk of extinction. 8 ranked relatively high in every analysis.org/copyleft/fdl. taking into account the extinction probability of all species. Caniformia The Bayesian results suggest a sister relationship between the red panda (Ailurus fulgens) and Canidae and. hence potentially a result of long branch attraction. May-Collado and Agnarsson (2006) showed that. mitogenomic studies. the Bayesian result supports a slightly modiﬁed Arctoidea and Musteloidea. in terms of the recovery of ‘known’ clades (benchmark clades) parsimony performed relatively poorly compared to Bayesian analyses of cytb sequences within Cetartiodactyla. However. wolf. 1. Mustelidae. (2008). meerkat (Fir0002/Flagstaffotos). we do not discuss the parsimony analysis further. The following obtained with permission from ASM mammal image library: palm civet (K. see http://www. for our purposes of showing one potential use of this phylogeny. using the GTR+C+I model and 200 search replications. / Molecular Phylogenetics and Evolution 54 (2010) 726–745
We use the Tuatara module (Maddison and Mooers. Hence. we will conﬁne most of our discussion to the Bayesian analysis including all taxa. is sister to a clade containing Procyonidae. The parsimony results differ further. We calculated (1) evolutionary distinctiveness (ED). a measure of a species terminal branch length and its share in ancestral branches.. 2006. and Mephitidae (Fig. Results 3. by uniting the red panda and kinkajou. 2006). differing only in terms of support values. Kuntner (tiger. the top priority species for conservation seen in Fig. Ailuridae) sister to Canidae (Fig. 1). However. Photographs by M. needed as input values in Tuatara. The placement of the red panda has not been stable across prior analyses and must be considered unresolved. instead of nesting within Procyonidae. we only illustrate one example of analyzing the data.D. 3. The following photographs are licensed under GFDL. 1). weighted by ‘global extinction’ (GE) risk (=EDGE. 2007) in the evolutionary analysis packet Mesquite (Maddison and Maddison.) and recover many fewer benchmark clades than Bayesian analyses (see Supplementary Fig. genet (Guérin Nicolas). In the parsimony analysis the two long branches. but probably conservatively. and for all but one of the currently recognized families within both suborders. However. hyena. panda. 1. The phylogeny recovers the monophyly of all benchmark clades (Figs. we discuss the likelihood results when there are major differences with the Bayesian analysis. but not Musteloidea. which were typically higher than in the full analysis. as it places the red panda sister to a clade containing all Caniformia minus Canidae. its subdivision into Caniformia and Feliformia. Agnarsson (otter. placing one cat (Homotherium serum) with viverrids. As emphasized by Mooers et al.. coati.Author's personal copy
I. 1) broadly agrees with the Bayesian results. using the ‘pessimistic’ approach of Mooers et al. conﬁrming our previous ﬁndings. Haweeli). Kutunidisz). Lord). sea lion). Benchmark clades The Bayesian analysis of the pruned dataset gave essentially identical result.html. etc. Photos of elephant seals (Jan Roletto) and skunks from public domain. We calculate extinction risk by translating IUCN threat categories to extinction probabilities.gnu. Caniformia).2. and placing skunks (Mephitidae) within raccoons (Procyonidae). nuclear DNA. Our Bayesian and likelihood results are incongruent with current classiﬁcation at the family level only in the placement of a single species. linsang (Alessio Marrucci). These unlikely relationships in most cases involve taxa with notably long branches. red panda and kinkajou (Potos ﬂavus) attract. again weighted by extinction risk (HEDGE. Given these issues. We ﬁnd support for Carnivora. 2009) to assess conservation priorities for carnivoran species.D.
59 0.00 1.00
0.00 1.00 1.00 1.14 Prionailurus viverrinus Otocolobus manul 0.00 1.00 1.98 Mustela eversmannii 0.92 Viverra megaspila Viverra zibetha Poiana richardsonii Genetta thierryi 0.62 0.84 0.73 0.00 1.00 Cystophora cristata 0.86 0.00 1.99 0.78 0.58 0.00 0.80
Manis pentadactyla Manis tetradactyla 0.99 0.00 1.69 0.90 0.58 0.41 1.59 0.00 1.59 0.)
this group is sister to the remaining caniforms.19 0.00 1. / Molecular Phylogenetics and Evolution 54 (2010) 726–745
Fig.00 1.99 0.00 1.59 0.61 1.00 1.00 1.43 0.00 1.52 0.00 Spilogale gracilis Spilogale putorius 1.28 0.25 Ursus americanus altifrontalis 0.75 0.94 0.590.00 0. and Otariidae—fur seals and sea lions).42 1.54 0.55 Martes pennanti 0.30
0.35 0.00 1.00 1.00 1.99 0.00 Mustela strigidorsa Mustela erminea 1.20
0.00 1.00 1.69 0.96 0. 6 and 7).77 0.00 1.24 0.96 0.44 Monachus schauinslandi 0.00 1.00 Herpestes auropunctatus 1.00 Nasua narica 1.35 0.99 0.90 0.00 1.00 1.00 1.63 0.00 1.00 1.64 0.99 0.26 Mustela frenata 0.00 Paradoxurus hermaphroditus Paradoxurus jerdoni Viverricula indica Civettictis civetta 1.99 0.00 1.00 1.00 0.99 0.97 0.62 Leopardus wiedi 1.50 0.31 0.84 0.99 1.00 1.00 1. thus suggesting a slightly modiﬁed Arctoidea (Figs.00 1. The gray fox (Urocyon cinereoargentus) was sister to the
remaining extant canids.00 Ommatophoca rossii 0.13 0.79 0.83 Monachus monachus 0.44 0.57 0.71 0.00 1.00 Hippidion saldiasi Equus caballus Equus przewalskii 1.86 0.00 1.80 0.92 0.63 0.81 0.00 Chrotogale owstoni Hemigalus derbyanus Arctictis binturong Paguma larvata 1.00 1.00 1.98 Canis lupus pallipes 1.00 0.47 0.99 Ursus americanus americanus 0.92 0.00 Bassaricyon gabbii 1. These results support the monophyly of all carnivoran families as currently circumscribed.00 1.00 Lutra canadensis 1. within which two main lineages were identiﬁed.59 0.92 Lynchailurus colocolo 0.00 1.98 0.98 0.00 1.00 1.00 1.00 1.54 0.00 0.90 0.50 1.43 0.00 1.90 0.00 Mephitis mephitis 1.00 1.00 1.14 0.90 0.59 Ursus thibetanus thibetanus 0.50 Acinonyx jubatus 0. the likelihood analysis recovered the traditional Arctoidea (Supplementary Fig.00 1.00 Martes melampus 1.96 Panthera tigris altaica Panthera tigris tigris 0.00 1.00 1.39 Ursus spelaeus 1.99 Phoca largha 1.99 0.39
0.00 Canis latrans Canis rufus Canis aureus 0.00 1.43 1.
.98 0.69 0.33 0.00 1. the other containing jackals.99 0. They further support all benchmark clades.00 1.71 0.59 0.00 1.80
0.00 1.00 0.94 Bassariscus astutus 0.00 Nasua nasua 1.96 0.00 1.63 0.00 1.78 Galerella pulverulenta 0.00 1.55 0.22 0.30 0.82 0.24 0.43 0.30 0.30
1.00 1.00 1.94 Canis lupus chanco 0.69 0.62 0.00 Proteles cristatus 0.39 0.00 1.00 Galictis cuja Galictis vittata 1.69 0.00 1.00 1.00 1.00 1.51 0.82 0.25 0.19 Vormela peregusna 0.00 1.00 1.59 0.00 0.27 Ursus americanus kermodei Ursus americanus vancouveri Odobenus rosmarus rosmarus 1.00 Ursus arctos 0.54 0.74 0.22 Oncifelis guigna 0.00 1.00 1.98 0.00 1.86
1.41 0.30 0.44 0.00 1. The genus Canis is also paraphyletic with the blackbacked jackal (Canis mesomelas) and the Dhole or Asian wild dog (Canis alpinus.00 1.58 Genetta johnstoni 0.96 0.75 0.27 0.45 0.47
0.52 0.90 0.49 0.79 0.00 1.27 0.00 1.98 0.59 Genetta victoriae 0.97
1.00 1.00 0.50 0.82 Halichoerus grypus 0.00 1.00 1.00 1.00 1.26 0.00 1.00 1. previously Coun) placing outside the genus.00 0.00 1.00 1.99 0.90 0.62 0.00 Genetta rubiginosa 1.59 0.00 1.00 0. crab-eating dog (Cerdocyon).00 1.26 0.00 1.00 1.59 0.00 Canis lupus familiaris Canis lupus lupus Ailuropoda melanoleuca
1.33 0.82 0.59
0.00 1.00 0.00
0.39 1.74 1.83 Vulpes zerda 0.00 1.81 0.00 1.99 0.84 0.64 0.99 Mustela nivalis Mustela itatsi 1.61 0.00 1.41 Speothos venaticus 0.97 Aonyx capensis 0.00 Bassaricyon alleni 1.00 Arctonyx collaris 1.55 0.59 0.64 0.59 0.49 Mustela lutreola Mustela putorius
0.00 Prionodon linsang Prionodon pardicolor 1.Author's personal copy
I.88 0.86 0.54 0.00 Galidictis fasciata 1.59 0.22 0.59
0.00 0.43 Ursus thibetanus 1.00 0.86 0.00 1.00 1.50 0.14 0.00 Ceratotherium simum Diceros bicornis 1.98 0.00 1.71 Eira barbara 0.00 1.49 0.00 1.84 0.00 1.94 0.88 1.99 Lycalopex sechurae 0.97 0.39 0. although with a modiﬁed Arctoidea and Musteloidea (asterisked) due to the placement of the red panda as sister to Canidae.98 0.00 1.00 1.00 Phoca vitulina 0.74 0.63 Ursus thibetanus formosanus 0.95 0.59 0.00 Genetta tigrina 1.79 0.98 0. adustus does not nest with Canis in the likelihood analysis.58 0.00 1.00 1.00 1.00 1.64 0.00 1.95 Martes foina 0.98 0.97 0.59 0.67 0.50 0.99 0.81 Gulo gulo 1.00 Rhinoceros sondaicus Rhinoceros unicornis 1. the reader is referred to the web version of this article.79 1.00 Neofelis diardi Neofelis nebulosa 1.00 1.80 0.00 1.00 1.00 1.99 0.79 0.80 1.59 0.00 1.48 0.28 Phoca caspica 0.00 1.58 0.00 1.00 1.79 0.00 1.00 Arctocephalus philippii 1.64 0.35 0.00 Lontra felina 1.98 0.00 Lontra canadensis 0.00 1.79 Neophoca cinerea 1. and foxes (Vulpes).55 0.44 Vulpes macrotis 0.00 1.00 1.52 0.49 0.71 0.22 Catopuma temminckii Pardofelis marmorata 0.58 0.00 1.98 0. Phocidae—seals.57 0.69 0.00 1.88 Nyctereutes procyonoides 0.59 0.98 0.99 0.95 0.99 0.56 0. in Canis).80 0.00 Viverra tangalunga 0.51 0.39 0.78 0.59 0.97 0.58 Felis silvestris catus Felis silvestris ornata Ailurus fulgens Urocyon cinereoargenteus Otocyon megalotis 0.00 1.00 1.98 0.10
1.00 Bdeogale crassicauda 0.52 0.43 0.64 Hyaena hyaena Parahyaena brunnea Fossa fossana Cryptoprocta ferox 0.31 0.80 0.00 1.00 1.81 Genetta cristata 0.50 0.00 0.00 0.63 0.89 0.52 0.94 0.00 0.73 Ictonyx striatus 0.00 1.75 0.00 1.90 0.00 1.95 1.00 1.61 1. However.00 1.83 0.00 1.97 0.47 0.52 0.81 0.79 0.90 Miracinonyx trumani Puma concolor 0.94 0.63 Canis simensis 0. short-eared dog (Atelocynus).86 0.00 1.75 Ursus americanus cinnamomum 0.00 1.78 Bdeogale nigripes Rhynchogale melleri 0. (For interpretation of the references to color in this ﬁgure legend.97 Lycalopex griseus 0.41 0.45 0.00 1.78 0.27 0.98 Felis margarita 0.80 0.71 0.84 0.81 0.00 1.24 0.50 0.24 0.99 0.98 Poecilogale albinucha 0.00 Herpestes javanicus 0.95 0. Within Pinnipedia the following familial relationships (Phocidae(Odobenidae + Otariidae) was strongly supported.94 Panthera uncia 0.43 Genetta pardina 0.88 Panthera pardus 0.90
0.44 0.62 0.96 0.63 0.00 1.00 0.99 0.78 0.00 1.52 0.87 0.46 1.00 1.99 0.84 0.00 1.00 1.99 0.96 0.97 0.00 1.00 1.80 0.80 0.95 0.13 0.92 0.59
0.55 0.00 1.44 0.99 0. African wild dog (Lycaon).99 0.00 1.00 Crossarchus alexandri 0.59 0.46 0.83 0.00 Ursus thibetanus mupinensis 0.48 Herpestes urva 0.99 0.54 0.88 0.00 Arctocephalus gazella 0.26 0.00 1.63 0.00 1.86 0.00 1.63 0.00 1.98 0.44 0. The genus Vulpes appears to be paraphyletic. Canidae The monophyly of Canidae was well supported in each analysis (Fig.00 Lutra lutra 1.00 1. bush dog (Speothos).59 0.88 Chrysocyon brachyurus 0.94 1. The domesticated dog (Canis lupus familiaris) is sister to the Eurasian wolf (Canis lupus lupus).31 1.30 0.30 0.88 0.41 0.00 1.44 0.73 0.42 0.13 0.53 0. Agnarsson et al.40
1.00 1.96 Oncifelis geoffroyi 0. racoon dog (Nyctereutes).00 1.27 Lynx rufus 1.59 0.41 Ichneumia albicauda 0.00 Hydrurga leptonyx Leptonychotes weddellii Erignathus barbatus 0.64 0.81 0.00 1.00 1.00 1.95 0.00 Arctocephalus forsteri 1.54 0.83 0. except in the placement of Potos outside Procyonidae.00 1.84 Neovison vison 1.00 1.98 0. This Ursidae/Pinnipedia clade was sister to a well supported modiﬁed Musteloidea containing Potos ﬂavus as sister to the clade (Mustelidea(Mephitidae + Procyonidae)).73 0.00 1.79 0.98 0.48 0.00 1.00 Coelodonta antiquitatis Dicerorhinus sumatrensis Equus asinus Equus burchellii 1.43 Felis nigripes 0.00 1.90 Genetta bourloni Genetta sp schoutedeni 1.97 0.00 1.98 0.47 0.86 0.84 0.74 1.74 0.00 1.77 0.00 Lutra sumatrana 0.82 Crossarchus obscurus 1.00 Mustela nudipes 1.63 0.00 Tapirus indicus Tapirus terrestris 1.31 0.86 0.87 0.44 0.00 1.00 1.92 0.00 1.78 0.90 0.00 1.88 0.79 0.00 1.54 0.00 1.00 Eumetopias jubatus 1.84 0.00 1.94 0.59 0.80 Felis chaus 0.99 0.41 Panthera tigris corbetti 1.00 0.89 0.00 1.88 0.90 0.00 1.54 0.81 0.00
0.33 0.00 Procyon cancrivorus Procyon lotor Taxidea taxus 1.00 Zalophus wollebaeki Otaria byronia 0.98
0.51 0.00 Helogale hirtula Helogale parvula Homotherium serum Herpailurus yaguarondi 0.00 1.69 0.84 0.98 0.00 1.92 0.14 0.1.10
0.54 0.00 1.94 0.58 0.69 0.89 Ursus americanus carlottae 0.94 Mustela altaica 0.67 0.00 Mirounga angustirostris Mirounga leonina 1.78 0.00 1.00 1.43 0.00 1.52
0.00 1. maned wolf (Chrysocyon).2.53 0.00 0.00 1.89 0.00 1.00 Mephitis macroura 1.50 0.59
0.46 0.59 Ursus maritimus 0.00 Lontra longicaudis 1.00 1.00 1.00 Liberiictis kuhni Mungos mungo 1.94 0.78 0.59 Melursus ursinus 0.64 1.69 0.90 0.39 0.48 0. 2.00 Atelocynus microtis 0.28 0.79 0.80 0.00 Zalophus californianus 1.44 0.63 1. 3.24 1.98 0.00 1.49 0.00 1.98 0.00 1.00 1.96 0.65 0.90 0.00 Amblonyx cinereus Lutrogale perspicillata Ictonyx libyca 1.58 0.57 Arctocephalus australis 1.56 Galerella sanguinea 0.98 0.00 1.00 1.00 0.14 0.99
Canis adustus 0.50 0.00 1.42 0.00 1.25 0.22 0.00 1.00 0.50 0.64 0.22 0.77 1.96 Nandinia binotata
0.90 0.43 0.63 0.88 0. wolves.00 Genetta poensis 0.55 0.54 0.53 Felis silvestris caffra 0.00 1.79
Arctodus simus Tremarctos ornatus 0.64 0.35 0.00 1.99 0.43 0.86 0.35 0.00 1. 6).53 Phoca sibirica Potos flavus 1.64 0. one containing bat-eared fox (Otocyon).00 1.00 1.00 Dusicyon thous 0.84 0.00 1.75 Ursus thibetanus ussuricus 0.00 Galidia elegans 1.92 0.54 1.71 0.00 Callorhinus ursinus 1.00 1.98 0.92 Lontra provocax Enhydra lutris 0.00 Arctocephalus townsendi 1.00 1.00 1.78 Atilax paludinosus 0.87 0.98 Martes zibellina 0.00 1.60
0.00 1.71 Arctocephalus tropicalis 1.00 1.41 0.98 0.94 0.97 0.31 Felis silvestris lybica 0.51 Euarctos americanus 0.56 0.27 0.74 0.73 Leopardus tigrinus 0.99 0.26 0.00 0.00 1.73 0.86 Lobodon carcinophaga 0.50
0.51 0.99 0.43 Helarctos malayanus 0.22 1.78 0.96 0.00 Ictailurus planiceps 1. colored by families.99 Vulpes vulpes Canis mesomelas elongae Cuon alpinus 0.00 1.44 0.00
0.00 Meles meles anakuma Mellivora capensis 0.00 1.95 0.00 Martes americana Martes caurina Melogale moschata Pteronura brasiliensis 1.00 0.00 1.00 1.00 1.00 Canis himalayensis 0.83 0.98 Herpestes ichneumon 0.55 0.48 1.00 1.00 1.99 Phoca fasciata Phoca groenlandica 1.98 0.00 1.82 0.59 0.98 0.97 Lutra maculicollis 0.00 1.54 0.54 Arctocephalus pusillus 0.00 1.00 1.41 0.52 0.97 0.59 0.00 Mustela sibirica 0.00 1.56 Genetta angolensis 0.0
0.92 0.97 1.80 0.90 0.63 0.00 1.70
0.82 0.00 1.99 0.95 Lynx pardinus 0.19 1.44 1.00 1.94 0.88 1.00 1.00 Herpestes edwardsii 1.95 Panthera tigris sumatrae 0.63 0.86 0.43 0.90 0.00 0.75 1.00 1.63 Genetta maculata 1.00 1.00 1.00 1.41 0.53 0.53 Phoca hispida 0.59 0.84 1.96 0.52 0.84 Mustela nigripes 0.00 Bassariscus sumichrasti 0. and ‘dogs’ (in part.33 Phocarctos hookeri 0.00 1.65 0.49 0.99 0.00 1.00 1.45 Panthera onca 0.75 0.69 0.00 Panthera leo 0.64 0.39 0.92 0.64 1.56 0.31 1.59 0.00 1.00 Mungotictis decemlineata Salanoia concolor 0.80 0.84 0.71 0. and the remaining ‘foxes’ (Pseudalopex).94 0.44 0.99 0.00 1.82 0.41 1.00 1.00 1.00 1. The monophyly of the remaining Canis species also is very poorly supported (39% posterior probability) and C.00 1.47 0.97 0.00 1.69 Lycalopex culpaeus Lycalopex gymnocercus
Hyaenidae Eupleridae Herpestidae
1.00 1.00 1.00 1.46 0.20
1.56 0.00 1.22 0.00 0.57 0.00 1.00 0.54 0.99 Lycalopex vetulus 0.00 0. Branches are proportional to lengths.69 Prionailurus bengalensis 0.89 0.99 0.00 1.00 1.97 0.00 1.59
0.00 0.86 Lynx canadensis Lynx lynx 0.00 1. Results of Bayesian analysis of the full matrix.67 0.00 Crocuta crocuta 1.90 0.49
1. 2).88 0.79 0.88 0.97 0.58 Genetta felina Genetta genetta 0.59 Genetta servalina 0.90 0.00 0.57 0.65 Canis indica 0.99 0.97 0.24 0.55 Herpestes naso 0.44 Cynictis penicillata 1.58 0.00 0.78 0. Our results also place bears (Ursidae) sister to Pinnipedia (Odobenidae—walrus.98 0.69 0. it contains Nyctereutes procyonoides.00
Arctogalidia trivirgata Cynogale bennettii 1.99 Leopardus pardalis 0.86 Martes flavigula 1.00 1.97 0.46 1.95 0.00 Arctocephalus galapagoensis 1.63 Suricata suricatta 1.69 0.75 0.00 1.78 0.33 0.59 0.59 0.
The domestic cat (F. The genus Arctocephalus is furthermore paraphyletic. Our results indicate that three of these subfamilies are not monophyletic (Fig. Otaria. Lynx spp.or polyphyletic. Current classiﬁcation subdivides Mustelidae into four subfamilies. Neofelis. and Pteronura). the hypothesized ancestor of the domestic cat. Within Hyaenidae. Ictonyx. linsang) and spotted (P. Otariidae. silvestris catus) is always sister to the Asiatic wildcat (F. spelaeus) is sister to a clade containing the polar bear (U. excluding Potos) 100 100 100 93 100 100 65 100
third subfamily. Melinae (true badgers: Arctonyx. Monophyletic but relatively weakly supported are Prionailurus and Felis. 3. Felidae Within monophyletic Felidae.2. the extinct saber-toothed (or Scimitar-toothed) cat (Homotherium serum) was sister to all remaining cats (Figs. within Phocidae. Paradoxurus). the clade Catopuma + Pardofelis. with relationships among the remainder of cat-like carnivorans as: (Viverridae(Felidae(Prionodontidae(Hyaenidae(Eupleridae + Herespestidae))))). while the genera Lynx and Leopardus are paraphyletic. and Gulo). uncia. sister to Procyon plus Bassariscus. Poiana). Ursidae Our results support a monophyletic Ursidae with the panda (Ailuropoda melanoleuca) as sister to the remaining bears (Fig. for example the placement of jaguarondi. Viverrinae (Viverricula. and Halichoerus) was strongly supported.
3. and each pinniped family was strongly supported (Fig. However. the brown bear (U. and Hemigalus). Vormela. Ommatophoca. and the sun bear (Helarctos malayanus) nest within it. the small-toothed palm civet (Arctogalidia trivirgata. Prionodontidae and Hyaenidae Prionodontidae (Fig. Erignathus. the genera Lutra. Hydrurga.Author's personal copy
I. Agnarsson et al. thus suggesting quite different placement of Taxideninae (Supplementary Fig. 6). but not that of Procyonidae and the subfamily Potosinae.3. and the pair of brown hyenas (Hyaena hyaena and Parahyaena brunnea).3. 2006). Our ﬁndings refute the monophyly of the subfamily Procyoninae (Procyon. Mydaus. 3–5).3. 6). 3 and Supplementary Fig. The Odobenidae is monotypic (walrus) and is sister to Otariidae. as well as those within Felis. Musteloidea—Mephitidae.3. Mellivora). P. 3). In addition. Our results dispute the monophyly of the genus Ursus as the sloth bear (Melursus ursinos). excluding Melogale. In the Bayesian analysis the monotypic Nandiniidae (African palm civet. Noephoca. were supported. Galictis. Mellivorinae (honey-badger. 1996. 4 and 5) in all analyses. Nandinia binotata) was sister to the remaining extant Feliformia. Genettinae. Mustelinae. and Bassariscus). The extinct Cave bear (U. due to the placement of Potos ﬂavus sister to all remaining musteloideans. The two putative subfamilies. All results dispute the hypothesized sister group relationship of the jaguarondi (Herpailurus yagouarundi) and the puma (Puma concolor) (Masuda et al. Paradoxurinae (Arctictis. Enhydra. Prionodontidae. and Melogale). Procyonidae.. as indicated in braces. Our results support the monophyly of Viverrinae. 2). The
. the clade Lynchailurus + (Leopardus + Oncifelis). particularly in the placement of African palm civet. 3. pusillus outside it. Octolobus. and Taxideninae (Taxidea). 3. Feliformia Our results support the monophyly of all seven families within Feliformia (Figs. Mustelinae (Eira. and both Mephitidae and Mustelidae was strongly supported (Fig. Lutrinae (otters: Aonyx. Poecilictis.4. 3. Ictonyx. the sister pair spotted hyena (Crocuta crocuta) and aardwolf (Proteles cristatus). 2). due to the placement of Melogale sister to Lutrinae and Mustelinae. Mustela.2. Arctocephalinae (Callorhinus and Arctocephalus) and Otariinae (Eumetopias. and the extinct giant short-faced bear (Arctodus simus) then group strongly. See text for details. *refers to support for slightly modiﬁed benchmark clades. The big cat clade contains Panthera and Neofelis species and is strongly supported. Benchmark clades Carnivora Caniformia Feliformia Arctoidea Musteloidea Mustelidae Canidae Mephitidae Procyonidae Pinnipedia Otariidae Phocidae Ursidae Felidae Herpestidae Eupleridae Hyaenidae Full matrix 100 59 59 49* (excluding Ailuridae) 100* (excluding Ailuridae) 100 100 100 NO (100. The subdivision of Otariidae into fur seals and sea lions is refuted. and these two in turn sister to the African wildcat F.2. and Phocarctos) were also refuted. instead strongly supporting a clade containing Bassaricyon and Nasua. the monophyly of both subfamilies Monachinae (southern seals and monk seals: Monachus.2. The majority of clades within Felidae are poorly supported. Lobodon. The likelihood results differ mainly in the placement of the root of Mustelidae. 2000. Viverridae Within Viverridae four subfamilies have been proposed. Our results place the monotypic Taxideninae (Taxidea taxus) sister to the remaining Mustelidae. and Viverridae (Supplementary Fig. to the exclusion of all remaining bears. Lyncodon. and Martes were para.1. and Panthera incl. Meles. Nasua.3. In contrast. Martes. seemingly due to the majority of felids having available only very short cytb sequences available. Johnson et al. Mirounga. / Molecular Phylogenetics and Evolution 54 (2010) 726–745
Table 2 Recovery and support of benchmark clades within Carnivora (numbers are posterior probability values).. only so by the weakly supported phylogenetic position of A. currently in Paradoxurinae) was sister to hemigalins and paradoxurins together. and Leptonychotes) and Phocinae (northern seals: Phoca.2. and Hemigalinae (Fig. 2). silvestris ornata). 3) consists of only two species: banded (P. Pinnipedia—Odobenidae. arctos). Spectacled bear (Tremarctos omalus). and the interrelationships of Panthera uncia. Viverra). 3. containing two species each. and Hemigalinae (Chrotogale. due to the placement of Pteronura brasiliensis (giant otter) also sister to both Lutrinae and Mustelinae. 2). However. 7). but the Bayesian and likelihood analyses differ in the placement of several genera on the felid tree (see Supplementary Fig. The interrelationships of these subfamilies was generally weakly supported and differed between the Bayesian and likelihood analyses (Fig. Phocidae The monophyly of Pinnipedia. Mustelidae The monophyly of Musteloidea. maritimus) and its sister species. Cystophora. 7): (1) Melinae. the interrelationships of families were generally poorly supported and the likelihood analysis differed. pardicolor) linsangs. however. Mustela.. onca and pardus. excluding Potos) 100 100 100 59 90 100 54 100 Pruned matrix 100 93 93 49* (excluding Ailuridae) 100* (excluding Ailuridae) 100 100 100 NO (100. Mattern and McLennan. 3. Other generally well supported clades are Oncifelis. Poecilogale. Genettinae (Genetta. (2) Lutrinae. silvestris lybica. Lutra. two lineages. Paguma. thus rendering Paradoxurinae paraphyletic. Civettictis. Cynogale.3. was polyphyletic with its various genera scattered. Zalophus.
naso) was placed sister to the white-tailed mongoose (Ichneumia).Author's personal copy
I.3.00 0.63 1. Conservation priorities Our preliminary analysis of conservation priorities identiﬁes a number of candidate species that.55 0.3. Numbers are posterior probability values.00 1.81 0. 8).00 0.00 1. Agnarsson et al. Herpestidae In general.00 1.78 0.54 1.00
0. and the long-
.35 0.00 0. the relationships within Herpestidae are rather weakly supported (Fig.59 0.30 0.00 1.52 1.00 0. Ow-
nosed mongoose (H. Mungotictis.50 1.78
1.59 0. The Egyptian mongoose (Herpestes ichneumon) nested with Galerella (slender mongooses).64
1. but place Crytoprocta sister to it so that Euplerinae is paraphyletic (Fig. Although the ranking of species strongly depends on the parameters of choice (Mooers et al.98 0.00 0.41 0.00 1.63
1.00 0.99 1. Galidictis.4.00 0.78 0.59 0.58 0. Our results support Galidiinae.44 1. Details of relationships among feliformian carnivores.59 0.00
Nandinia binotata Arctogalidia trivirgata Cynogale bennettii Chrotogale owstoni Hemigalus derbyanus Arctictis binturong Paguma larvata Paradoxurus hermaphroditus Paradoxurus jerdoni Viverricula indica Civettictis civetta Viverra tangalunga Viverra megaspila Viverra zibetha Poiana richardsonii Genetta thierryi Genetta victoriae Genetta johnstoni Genetta cristata Genetta servalina Genetta felina Genetta genetta Genetta rubiginosa Genetta maculata Genetta tigrina Genetta angolensis Genetta pardina Genetta poensis Genetta bourloni Genetta sp schoutedeni Prionodon linsang Prionodon pardicolor Crocuta crocuta Proteles cristatus Hyaena hyaena Parahyaena brunnea Fossa fossana Cryptoprocta ferox Galidia elegans Galidictis fasciata Mungotictis decemlineata Salanoia concolor Atilax paludinosus Herpestes naso Herpestes urva Herpestes edwardsii Herpestes auropunctatus Herpestes javanicus Ichneumia albicauda Herpestes ichneumon Galerella pulverulenta Galerella sanguinea Cynictis penicillata Bdeogale crassicauda Bdeogale nigripes Rhynchogale melleri Suricata suricatta Liberiictis kuhni Mungos mungo Crossarchus alexandri Crossarchus obscurus Helogale hirtula Helogale parvula
Fig.00 1. 2008).48 1. two subfamilies have been proposed: Galidiinae (Galidia.00
0. 3).59 0.42 1. panda.4.74 1.00
1. Eupleridae Within Eupleridae.43 0. high ranking species for conservation priorities in most analysis include: monk seals. 3. / Molecular Phylogenetics and Evolution 54 (2010) 726–745
0. and Salanoia) and Euplerinae (Crytoprocta and Fossa).56 0. 3).00
0.97 1.00 0. 3.5.90 0.00 0. red panda.00 0.00
0.74 0. Our ﬁndings indicate that the genus Herpestes is polyphyletic.
3. otter civet.00 0. may deserve special attention and conservation effort (Fig.90 1.00 0.77
1.86 0.97 1.58 0. based on evolutionary distinctiveness weighted by extinction risk..
1. and a guide to carnivoran classiﬁcation. Kuntner (cheetah. Drawing of Homotherium by Alexis Vlachos. 5.. Thus. In addition. giant otter. While we anticipate that future studies using more character data will likely refute some of our ﬁndings. 2007). Photographs by M. banded palm civet.. with the most robustly supported higher-level clades that prior studies tend to agree upon (e. Given the high information content in cytb and ease of ampliﬁcation. / Molecular Phylogenetics and Evolution 54 (2010) 726–745
Fig. and marbled polecat. Siberian tiger (MJCdetroit). 4. based on sequences of the cytochrome b mitochondrial gene. jaguarondi (Bodlina).S. We proposed a phylogenetic hypothesis placing 222 out of the 270 extant species of Carnivora. Finarelli and Flynn.html. 4. reproduced from Lyras and Van Der Geer (2007) with permission. Numbers are posterior probability values with a second number representing support value in the pruned analysis. Clearly. Geoffroy’s cat (Daf-de). fossa. Kutunidisz). I. 1. Relationships among species of Felidae. jaguar). when higher than in the full analysis. Agnarsson et al. The following obtained with permission from ASM mammal image library: pampas cat and oncilla (K.g. although the Bayesian analysis recovered a slightly modi-
ﬁed Arctoidea due to a novel placement of the red panda as sister to canids (Figs. 4. in particular the detailed patterns of relationships among carnivoran species represent a useful working hypothesis.0.g.org/copyleft/fdl. Indochinese tiger (Kabir Bakie). in addition to 17 subspecies and four extinct species. it seems reasonable to assume that the novel results presented here. 4.Author's personal copy
I. 2007) and thus represents a useful tool for comparative studies. spectacled bear. Discussion Despite their obvious utility. the signal present in cytb agrees well. the current phylogeny overall agrees with results that have been most stable in previous studies based on various data (e.
ston’s palm civet. obtaining complete cytb sequences for all carnivorans is an obvious priority. snow leopard. see http://www. 2. clouded leopard. Liberian mongoose. species-complete phylogenies are lacking for most higher-level taxa (such as orders). and 6). snow leopard). Finarelli and Flynn. Acosta-Jamett). our results support the sister relationship between Carnivora and Pholidota. Bengal tiger (Paul Mannix). overall. Our analysis of a pruned matrix where taxa with short sequences were removed indicates that low support for placement of species is often a result of missing data (short sequences). blackfooted cat.gnu. Fish and Wildlife). Photograph of kodkod from Arkive (G. and all but one (Bayesian) benchmark clades gives credibility to our results in general. Photographs of Bornean clouded leopard (Canorus). see also Fig. The following photographs are licensed under GFDL. the sea otter. Benchmark clades We found strong support for the monophyly of all benchmark clades. 1 and 2). Within Caniformia relationships among families have been unstable in prior studies (Flynn and
. Agnarsson (leopard. licensed under the Creative Commons Attribution ShareAlike 3. Sumatran tiger (Monika Betley). The recovery of all (likelihood). Caniformia The monophyly of Caniformia is uncontroversial and is here supported (Figs. lion.2. regardless of the organism. binturong. Photograph of mountain lion from public domain (U.
. / Molecular Phylogenetics and Evolution 54 (2010) 726–745
Fig. 1994).g. 2004. WesleyHunt and Flynn. Delisle and Strobeck. 1993. morphological data alone. 2005. The following images are licensed under the Creative Commons Attribution ShareAlike 2. Flynn et al. 1996a. Flynn et al. Creative Commons Attribution 2. 2007). Pallas’ cat (Scottmliddell)... Berta and Ray. Flynn and Nedbal. Dragoo and Honeycutt. 2004..Author's personal copy
I. Bininda-Emonds et al. Flynn and Nedbal. Delisle and Strobeck. 2006.. 2005. 1998.php?title=Image:Prionailurus_planiceps. 1998). 1996). Creative Commons Attribution 3.g.org/wiki/ File:Prionailurus_viverrinus.. Goswami.. 1998.. 1990. Davis et al. Fish and Wildlife Service). Goswami. morphology and molecular data (e. 1995.5–3. our study is the ﬁrst to strongly support this clade. Photographs of Canadian lynx (Erwin and Peggy Bauer. Yu et al. The photograph of jungle cat is not copyrighted (GorillazFanAdam). Lento et al.. Prior studies place Ailuridae within Arctoidea.
2004).wikipilipinas. (2007) using complete mitogenomic data highlights ‘a gradually established molecular understanding’ referring to the sister relationship between Musteloidea and Pinnipedia (to the exclusion of Ursidae). ﬁshing cat (http://en. Our results. 6).g.0: ocelot (Ana Cotta).lynxexsitu. Flynn et al. continued from Fig.. these studies have only provided weak support for this clade.S. in contrast. 2006. Ledje and Arnason.g. Fish and Wildlife Service) and sand cat (DocTaxon) are from public domain. as most morphological studies have suggested (e. Fig. This arrangement has been weakly supported by combined morphological and mtDNA data (Vrana et al.wikipedia. Numbers are posterior probability values with a second number representing support value in the pruned analysis. the placement of Ailuridae has varied considerably in previous studies (e. and mtDNA alone (e. Relationships among species of Felidae. Koepﬂi et al.g. Nevertheless.b.. Arnason et al.org/ index. however. 1996). 2005). mtDNA alone (e.. Asian golden cat (OpenCage). Our Bayesian results challenge for the ﬁrst time the monophyly of Arctoidea placing the red panda (Ailuridae) sister to canids (Fig. and was method dependent in this study. Flynn and Nedbal. Thus we may still regard its phylogenetic afﬁnities as unsettled. 5. bobcat (U.g.jpg). (e. Photographs by M. Finarelli and Flynn. African wildcat (Sonelle). 6). 1999). This relationship has also been supported by other lines of evidence including nuDNA alone (e.0: Iberian lynx (Programa de Conservación Ex-situ del Lince Ibérico www. GNU Free Documentation Licence: ﬂat headed cat (http://en.0 Unported License: Eurasian lynx (Michael Gäbler).JPG#ﬁle). 2006. strongly support the more traditional sister relationship between Ursidae and Pinnipedia (to the exclusion of Musteloidea.. 1998. 2006). 2000). Sato et al.
Nedbal.. 2005. Bininda-Emonds and Russell.
. 1997. 2000.g. 2005. 4. Yu et al..g. Wolsan. black-footed cat (Zbyszko). domestic cat). Agnarsson et al.... Kuntner (leopard cat. Asiatic wildcat. and combined nuclear and mitochondrial DNA (e. U. Zhang and Ryder.. 2000. margay (Ltshears).es). However. when higher than in the full analysis.S. a result that may well be attributable to the dense taxon sampling.
44 0. and mtDNA
data for 23 species (Zrzavy and Ricankova.99 0.28
4..64 0. Numbers are posterior probability values.99 0.54 0.97 0.1.00
1. 7.00 1.00 0.59 0.41 0. cytogenetic.33 0.31 0.00 0.69 0.59 0.99 0.69 0.00 1.00 0. / Molecular Phylogenetics and Evolution 54 (2010) 726–745
0.00 0. Several other studies based on a combination of morphological.99
Ailurus fulgens Urocyon cinereoargenteus Otocyon megalotis Vulpes zerda Nyctereutes procyonoides Vulpes macrotis Vulpes vulpes Canis mesomelas elongae Cuon alpinus Chrysocyon brachyurus Speothos venaticus Atelocynus microtis Dusicyon thous Lycalopex sechurae Lycalopex vetulus Lycalopex griseus Lycalopex culpaeus Lycalopex gymnocercus Canis adustus Canis simensis Canis latrans Canis rufus Canis aureus Canis himalayensis Canis lupus chanco Canis indica Canis lupus pallipes Canis lupus familiaris Canis lupus lupus Ailuropoda melanoleuca Arctodus simus Tremarctos ornatus Ursus spelaeus Ursus arctos Ursus maritimus Melursus ursinus Helarctos malayanus Ursus thibetanus Ursus thibetanus formosanus Ursus thibetanus ussuricus Ursus thibetanus mupinensis Ursus thibetanus thibetanus Ursus americanus americanus Ursus americanus Ursus americanus cinnamomum Ursus americanus carlottae Ursus americanus altifrontalis Ursus americanus kermodei Ursus americanus vancouveri Odobenus rosmarus rosmarus Callorhinus ursinus Eumetopias jubatus Zalophus californianus Zalophus wollebaeki Otaria byronia Arctocephalus pusillus Phocarctos hookeri Neophoca cinerea Arctocephalus forsteri Arctocephalus australis Arctocephalus galapagoensis Arctocephalus gazella Arctocephalus tropicalis Arctocephalus philippii Arctocephalus townsendi Mirounga angustirostris Mirounga leonina Monachus monachus Monachus schauinslandi Lobodon carcinophaga Ommatophoca rossii Hydrurga leptonyx Leptonychotes weddellii Erignathus barbatus Phoca fasciata Phoca groenlandica Cystophora cristata Phoca largha Phoca vitulina Halichoerus grypus Phoca caspica Phoca hispida Phoca sibirica
0. and combined molecular and morphological data (Wayne et al.25
0.00 0.98 1.00
0.79 0. Details of relationships among caniformian carnivores.00 0.51 1. 2005).65 0. behavioral.64 1.59 0.89 0. Canidae Our study places the gray fox (Urocyon cinereoargentus) sister to other extant canids (Fig.88 0.00 1.47 0. 2004).84 1.83
1.46 1. 6.69 0. Agnarsson et al.00 1.00 0.57 1.00 0.00
1.39 0.82 0.99 1. nuclear DNA (Bardeleben et al..00 1.71 1.54 0.94
1. see also Fig.84 1.99 1.59 0.75 0. 1997) support the same relationship.00 0.00 0.83 0. The gray fox thus represents an old lineage that has persisted but not
0.43 0.63 0.27
0.59 0.39 0.00
0.Author's personal copy
Higdon et al. 2007. (ringed-tailes) and another subclade consist of Nasua spp. Phocidae Several studies have supported the monophyly of Pinnipedia. Higdon et al. Our results suggest two alternative
. americanus. we support Krause et al.. but differed in its exact phylogenetic position (e. and we must consider the placement of Potos an open question.. we found a clade containing Otaria (South American sea lion). found support for the monophyly of Lutrinae. Mustelidae)).3. 2006. In our study. contrasting the results by Pages et al. forsteri (New Zealand
fur seal). thibetanus and U. subfamilies within it are not. Instead we ﬁnd a sister relationship between Mephitidae and Procyonidae. and Arnason et al..4. Mustelidae Within Musteloidea our results differ in relationships among major groups from the recent study of Arnason et al. In addition. Arctocephalus (fur seals) is again refuted due to the phylogenetic position of the South African (or Cape) fur seal (A. 2006 [nuDNA]. (2003. Higdon et al.. it is placed in a grade of sea lions within which the remainder of Arctocephalus are monophyletic. australis (South American fur seal). A.. while Sato et al... found support for the monophyly of Melinae. (2008) we found strong support for the following relationship: Ursus spelaeus (cave bear) sister to Ursus arctos (brown bear) + U. Sato et al. or Coun alpinus) sister to the remaining canids. 4. 2006. and Artocephalus (fur seals) (see also Higdon et al. Dragoo and Honeycutt.. Dragoo and Honeycutt (1997) based on 12S and 16S ribosomal RNA (rRNA) genes.. pusillus). Flynn et al. the majority of them based on molecular evidence (e. 2005) and supertree analysis (Bininda-Emonds et al. 2000 [morphology and molecular data] Sato et al. Procyonidae. Wynen et al. (sea lions) (e. Most studies have placed it within Arctoidea. The relationships found in our study within Pinnipedia (Fig.2. Arnason et al. 2008.g.Author's personal copy
I. pusillus and A. Flynn et al. gazella (Antarctic fur seal) also represents an expected biogeographical pattern. more curious is the sister relationship between A. mtDNA]. malayanus. Agnarsson et al. (raccoons) sister to Bassariscus spp. Pages et al. 2004. as the monotypic sloth bear (Melursus ursinus) and sun bear (Helarctos malayanus) nest within it. Ursidae Within Ursidae. (2008). This is a somewhat surprising result. as also found by Krause et al. Musteloidea—Mephitidae.2. Fulton and Strobeck. 2007).g.. ursinus and U. and A. which has been referred to as ‘enigmatic’ (Bininda-Emonds et al. The spectacled bear (Tremarctos ornatus) and the extinct giant short-faced bear (Arctodus simus). 1997. tropicalis (Subantarctic fur seal) and A. Like in these prior studies... Like Krause et al. Further. crab-eating dog) and Atelocynus (short-eared dog) and Pseudalopex (foxes. Mustelidae)). In contrast to our ﬁndings. 2004) using nuclear genes also found Mustelinae to be paraphyletic (see also Stone and Cook. twonsendi (Guadalupe fur seal) suggesting a long distance dispersal to the Juan Fernandez islands.. 4. 2004. Delisle and Strobeck (2005) suggested a sister relationship between the red panda and Mephitidae. 2004. In agreement with prior studies our results support a sister relationship between Eumetopias (Steller sea lion) and Zalophus spp. the monophyly of the Otariidae subfamilies. In concordance with previous molecular studies.. Wynen et al. With the exception of the suggested position of Ailurus. the subantarctic fur seal. (2004) using cytb and repetitive ﬂaking regions. The phylogenetic position of the red panda (Ailurus fulgens) The phylogenetic position of the red panda has been one of major controversies in the history of Carnivora phylogenetics. then branch off next. Koepﬂi and Wayne (1998) similarly suggested paraphyletic Lutrinae. Wynen et al. (2001) using partial cytb and control region sequences found some support for the relationship between A. and that Potos ﬂavus is sister to all other musteloideans (Fig. and arboreal habits. 4. 3) based on a combined analysis of 11 gene segments supported the monophyly of Procyonidae. Our results strongly place Potos ﬂavus (kinkajou) as sister to all musteloideans thus rendering Procyonidae paraphyletic. Slattery and O’Brien (1995) suggested (based on very few species) that the red panda diverged early within the lineage of Procyonidae. Vila et al. Fulton and Strobeck. thus agreeing with previous ﬁndings that the dog was domesticated from it (e. Similar to Wayne et al. Our results refute the monophyly of the genus Ursus.. our study concords with other molecular studies (Krause et al.g. Although in our study the monophyly of Mustelidae is supported. who share with the panda conspicuous black patches around the eyes.2. Davis et al. (olingos). Delisle and Strobeck. our ﬁndings refute the monophyly of Canis with the black-backed jackal (Canis mesomelas) and dhole or Asian wild dog (Canis. (2007) which found support for the following relationship (Ailurus(Mephitidae(Procyonidae). 1997). 2008). (2008) using nuclear genes. 2005 [nuDNA. 2006. Higdon et al. This is another example of a difference between a taxon rich versus character rich studies (Agnarsson and May-Collado. and Mustelinae. 2006. is refuted. philippii (Juan Fernandez fur seal) and A. the giant panda (Ailuropoda melanoleuca) is sister to the remaining ursids (Fig. 2001. 1983). 2001. 4. In addition. given that a recent study by Fulton and Strobeck (2007) based on nuclear and mitochondrial DNA and a study by Koepﬂi et al. which must be related to a denser taxon sampling in our study.b) supporting a sister relationship between U. Neophoca (Australian sea lion).. King. 2005 [mtDNA].. respectively). in these studies Potos was sister to the remaining Procyonidae.g. see Fig. 6) agree with previous studies supporting the monophyly of each family. (2007) found support for the position of the red panda as sister to (Mephitidae(Procyonidae.. Otariidae. (2006) and Fulton and Strobeck (2006) proposed a clade containing the red panda. The relationships within this second lineage include strong supported southern hemisphere clade containing A. and was considered ‘highly autapomorphic’.. suggesting the colonization of Galapagos from Southern America. the musteloidean family relationships that we support here have previously been supported by mitochondrial and nuclear data (Flynn et al.g. Krause et al.. 1999. vegetarian diet. galapagoensis (Galapagos fur seal).. In other aspects our results contrast the cytb study of Koepﬂi and Wayne (1998). where a subclade contains both Procyon spp.5. despite the ecological success of this species judged by its current abundance. 2007).. 2007).. 2006 [nuDNA]. (1997) we found support for a clade containing Cerdocyon (=Dusicyon. 6). 2007). the relationships between procyonid species resemble those by Fulton and Strobeck (2007) and Koepﬂi et al. Our study provides strong support for the placement of Callorhinus ursinus (northern fur seal) sister to the rest of otariids (Fig. previously Lycalopex). 2008. Arnason et al. Most studies placed the red panda in Musteloidea. 2008). 2002). initially proposed based on vague morphological data (Repenning et al. (coatis) sister to Bassaricyon spp. 2007 [complete mitogenomic data]). Sato et al. / Molecular Phylogenetics and Evolution 54 (2010) 726–745
diversiﬁed compared to its sister lineage. The sister relationship of A.2. Our ﬁndings place the domestic dog sister to the Eurasian wolf (Canis lupus lupus). (2007. Other recent studies have also suggested that the giant panda branched off prior to the major diversiﬁcation of Ursidae (Yu et al. tropicalis. as well as placing seals (Phocidae) sister to Otaroidea containing eared seals and the walrus (Odobenidae + Otariidae) (e.2. However.. Therefore. 2007. Procyonidae. 2007).. 1971. Talbot and Shields. Lutrinae... 1996a. and Mustelidae. and Marmi et al. maritimus (polar bear). Davis et al.. Pinnipedia—Odobenidae.. Otherwise. Arnason et al. however. however. 1999). 6). 7). (2008) in transferring these species to Ursus (as U. Phocarctos (New Zealand sea lion). (2007)..
see also Fig.00
1.26 0.79 1.00 0.94 0.80 1. the position of the
1. 6.92 1.00 0.87 0.00 1.98 0.84 0.19 1. which challenges the monophyly of Arctoidea.00 1.00 0.00
0.81 1.00 1.00
0.95 0. 7.98 0. / Molecular Phylogenetics and Evolution 54 (2010) 726–745
Potos flavus Mephitis macroura Mephitis mephitis Spilogale gracilis Spilogale putorius Bassaricyon alleni Bassaricyon gabbii Nasua narica Nasua nasua Bassariscus sumichrasti Bassariscus astutus Procyon cancrivorus Procyon lotor Taxidea taxus Arctonyx collaris Meles meles Mellivora capensis Gulo gulo Eira barbara Martes pennanti Martes flavigula Martes melampus Martes zibellina Martes foina Martes americana Martes caurina Melogale moschata Pteronura brasiliensis Lontra canadensis Lutra canadensis Lontra longicaudis Lontra felina Lontra provocax Enhydra lutris Lutra maculicollis Lutra lutra Lutra sumatrana Aonyx capensis Amblonyx cinereus Lutrogale perspicillata Ictonyx libyca Poecilogale albinucha Ictonyx striatus Vormela peregusna Galictis cuja Galictis vittata Mustela frenata Neovison vison Mustela nudipes Mustela strigidorsa Mustela erminea Mustela altaica Mustela nivalis Mustela itatsi Mustela sibirica Mustela nigripes Mustela eversmannii Mustela lutreola Mustela putorius
Fig. Agnarsson et al.00 1. the Bayesian analysis supports a sister relationship between the red panda and Canidae with relatively strong support.00
0.97 1.00 0.
and novel placements.00 1.94 1.00 1. The likeli-
hood analysis places the red panda sister to all Caniformia minus Canidae.00 1.84 1. Details of relationships among caniformian carnivores. Numbers are posterior probability values.00
1.52 0. Essentially the difference between the two methods is a matter of root placement within Caniformia.99 1.00 0.00 0.00 1.00 0.00
1.98 0.00 0.00 1.73 0.71
0.44 0.97 0.99 1.Author's personal copy
are Homotherium. The ‘‘big cat clade”. Gaubert and Cordeiro-Estrela. Wesley-Hunt and Flynn. such results strongly imply independent evolution towards big size in cats (contra BinindaEmonds et al. 2008). This result is consistent with the monotypic subfamily Arctogalidiinae proposed by Pocock (1933). it is not currently under threat (IUCN. the short terminal branches and 100% support for the clade Catopuma + Pardofelis suggest these are synonymous. The genera for which monophyly cannot be tested due to monotypy/taxon sample. prior studies have also supported the following relationship (Hyaenidae(Eupleridae + Herespestidae)) (e. Herpailurus. 2006).. 2001) or almost exclusively osteological characters (Christiansen.3.. Flynn et al. poor character to taxon ratio (Christiansen. we focus our discussion on clades that are better supported and thus more likely to withstand the addition of data. Gaubert and Cordeiro-Estrela. Bininda-Emonds et al. they are incongruent with previously proposed topologies within Panthera (Mattern and McLennan. Christiansen. e. lion. with the enigmatic cat form the Central Asian mountain ranges. as do our ﬁndings. 2006. 2009). Either ﬁnding is surprising given prior work. (2008).. Apart from Arctogalidia. our results are nearly identical to those of Patou et al. Our results dispute the sister group relationship of the jaguarondi (Herpailurus yagouarundi) and the puma (Puma concolor) (Masuda et al. 4). Neofelis..1. gathering full cytb sequences from cat species can provide the most rapid means to advance understanding of Felidae phylogenetics. 3). highlighting that some of the interrelationships among Feliformia families are here poorly supported. (2004) using morphological and molecular data. and the jaguar (P. Our results group a Neotropical big cat. but this is not surprising as prior studies also differed greatly among themselves. All our analyses agree that the latter groups with relatively distal Panthera species. Nevertheless. 2008. and differ among the analyses. We have previously detected such incongruence between taxon sparse versus character sparse phylogenies (Agnarsson and May-Collado. 2008). We highlight this family here to draw attention to the lack of consensus regarding the phylogenetic relationships within this important and high-proﬁle group of mammals. Miracionyx. all other three subfamilies (Viverrinae. Felidae The monophyly of Felidae receives strong support (Fig. but also the non-roaring snow leopard (Panthera uncia = Uncia uncia). 2001). (2006). and thus support the monophyly of Pardofelis s. 2008).. Johnson et al.g. many clades within it are poorly supported. Agnarsson et al. but the genera Lynx and Leopardus are paraphyletic in our tree. Our Bayesian results support several studies. in contrast. The paraphyly of Paradoxurinae is due to the placement of Arctogalidia invirgata sister to Hemigalinae and the rest of Paradoxurinae. the jaguar. Both Bayesian and likelihood here agree that it is not a true civet (Viverridae). average nodal support within Felidae is generally lower in the full dataset than in the pruned dataset. Lynchailurus. the snow leopard (Panthera uncia). (2008).3. Genettinae. e. Puma. 2005). Relationships within Feliformia Our results support the monophyly of all families within Feliformia (Figs. The main strength of our study is a threefold increase in taxon sample compared to the largest prior study (Yu and Zhang. This clade is nested within an Old World big cat group making it difﬁcult to
. by Gaubert and Cordeiro-Estrela (2006). relatively weak support for Prionailurus and Felis. In the likelihood analysis. Our ﬁndings.. The genus Panthera includes the four roaring cats (tiger. and Pantera. this species represents an ancient lineage and thus much unique evolutionary history. 2006). Our results agree with recent molecular phylogenies (mtDNA and nuDNA) that have found support for a sister relationship between Viverridae and the remaining feliformians (e. but underlines the uncertainty surrounding the phylogenetic afﬁnities of this enigmatic animal.2. Acionyx. 1994. 2008). 2005. Ictailurus. Viverridae With the exception of Paradoxurinae. These include low taxon sampling (Bininda-Emonds et al. Given the low support for many clades. or the derivation of the phylogeny from a single. 2006): further studies including both many taxa and many characters are needed to resolve them. 1997.. tigris) form a group that is sister to all other Panthera species. 2000.. data source. Finarelli and Flynn. Johnson et al.g. 2005. (2008) placed Arctogalidia sister to the remaining Paradoxurinae (hence a monophyletic Paradoxurinae) based on mtDNA and nuDNA. Finarelli and Flynn. as discussed above. However. 2006). 2001. which suggested the African palm civet Nandinia binotata (Nandiniidae) as sister to a clade containing the remaining extant feliformians (Flynn and Nedbal. leopard.. The tiger subspecies (P. as previously found. 2007). The relatively big cats form two distantly related clades. 1–5) (Barycka. is well supported (Fig.g. Gathering more phylogenetic data from a broad taxonomic sample of felids is thus an urgent priority. 2007. There is thus little consensus regarding Felidae phylogenetics and a need for more comprehensive studies. 2001). pardus). 2008).3. 2001.. Our results place Prionodontidae (linsangs) as sister to the above clade. suggesting Uncia is a junior synonym of Panthera (as in Johnson et al. 2005. 4.. Pardofelis. Flynn et al. 2005. Yu and Zhang..g.. 2000.. 4).g. the snow leopard (Bayesian analysis) or the leopard (likelihood). these interrelationships are poorly supported and quite different in the likelihood analysis (Supplementary Fig.. 4. Further. Our results differ from most previously proposed felid phylogenies. 2005. and the jaguar. poor resolution (Yu and Zhang. therefore. Nandinia instead is sister to cats. Peters and Hast. 2007). 2008). Patou et al. and a much denser outgroup sampling than prior studies (e.. Some prior studies have found support for the division of the cats into sister clades of relatively big cats and small bodied cats (Bininda-Emonds et al. these in turn separating the lion (P. Christiansen. and Hemigalinae) were monophyletic (Fig. However. although the exact placement of the jaguarondi remains ambiguous. and very similar to those of Gaubert et al. Flynn et al. Mattern and McLennan. Fortunately. rel-
atively narrow. and Gaubert and Begg (2007) using molecular data. May-Collado and Agnarsson. refuted this. 2). although the latter three form a grade rather than a clade in the likelihood tree. 1998. the phylogenies of Yu and Zhang (2005) and Johnson et al. In the short term. best explain the inability of the snow leopard to roar as secondary loss. Christiansen. 2006). The low support seems in large part to be due to missing data—the high number of felids with short sequences. and Otocolobus. However. 2005). Catopuma.Author's personal copy
I. Christiansen. and Patou et al. 4.. Other studies place the snow leopard as sister to other species of Panthera (Johnson and O’Brien. 1996. There is good support for monophyly of Oncifelis. the puma (Puma concolor) and the extinct American cheetah (Miracinonyx trumani). However. Arctogalidia is morphologically peculiar and its placement requires further scrutiny. 2000. / Molecular Phylogenetics and Evolution 54 (2010) 726–745
red panda in the phylogenetic network is the same in the two analyses. containing Panthera and Neofelis species. 2002). implying the single origin of roaring within Panthera. 2006. Neofelis + Panthera and the cheetah (Acionyx jubatus). prior studies have some shortcomings. While these ﬁndings are generally well supported. however.l. animal chemical signals (BinindaEmonds et al.. Similar to our study. leo) from the clade including the leopard (P. Weissengruber et al. Gaubert and Cordeiro-Estrela. Mattern and McLennan. onca). In terms of genetic diversity.
the clouded leopard (Neofelis nebulosa) and the Bornean clouded leopard (Neofelis diardi). see text. all sometimes considered to be subspecies of the Eurasian wildcat. based on mitochondrial DNA data. suggesting diversiﬁcation within America. to the Americas. and Herpestidae are all strongly supported here (Fig. Species such as the walrus (currently listed as data deﬁcient. relatives.2
Black footed cat Marbled polecat Snow leopard
Mediterranean monk seal Spectacled bear
Sea otter Banded palm civet
Walrus Clouded leopard
0. which dates further back than the archeological evidence for cat domestication from ancient Egypt 5700 years ago (Linseele et al. placed the domestic cat as sister to the European wildcat (F. They concluded that cat domestication from as few as ﬁve founders took place in the Near East more than 9000 years ago. The neighbor joining analysis with mitochondrial data of Driscoll et al. 2004). in closely grouping the two subspecies F. 2000). 8. s. Eupleridae. Hyaenidae.2 0. Based both on Driscoll et al. group with the African cheetah. ornata with the domestic cat.)
. our results weakly place the domestic cat as sister to the Asiatic wildcat (F. ornata. EDGE and HEDGE. s. nigripes) and the other groups the remaining Felis species. (2006) in strong
0. 2007). / Molecular Phylogenetics and Evolution 54 (2010) 726–745
interpret biogeographically.3. which expanded their range from Africa to include Europe and the Americas during the Pleistocene (Yamaguchi et al. All analyses recovered two distinct clades within the Old World clade Felis. a recent study by Driscoll et al. Johnson et al. yellow indicating vulnerable. and dark red critically endangered. reﬂecting their evolutionary distinctiveness. (2007) tested the origin of cat domestication by genotyping 979 individuals belonging to all F. for example. orange indicating endangered. with dark blue reﬂecting least concern. The analysis of Johnson et al. However. under the ‘pessimistic’ transformation of IUCN categories to extinction risk. light blue indicating near threatened..3.0 0. but recently as least concern) and ﬂat headed cat (near threatened) rank high. Prionodontidae. less threatened. Pink indicate ‘data deﬁcient’ species (see text). One groups the fairly heavy bodied jungle cat (Felis chaus) with the small bodied black-footed cat (F.. support for this relationship is Bayesian only and is too weak to allow preliminary speculations on the running and climbing abilities of the American cheetah. It is notable that a number of endangered and critically endangered species rank lower than many species that are not currently seriously threatened. s. silvestris subspecies from the entire range for 36 short tandem repeats (STR) of DNA and by analyzing two mitochondrial DNA genes. (For interpretation of the references to color in this ﬁgure legend. These. (2006). However. s. while Eupleridae is weakly supported. 4. and Herpestidae The monophyly of Prionodontidae. A scattergram of two conservation priority metrics. The interrelationships of species within Hyaenidae differ slightly from those suggested by Koepﬂi et al. (2007).1 0. 3). lybica from the Near East. s. the fastest land animal and an unusual cat considering its diminished claw retraction abilities (Russell and Bryant. loosing these endangered species would loose relatively little unique evolutionary diversity. and our study there seems to remain a possibility that the domestic cat was bred from both F. here represented by Homotherium. A similar scenario has been suggested for lions. but their phylogenetic analysis grouped the domestic cat (European and Asian lineages being monophyletic) unequivocally with F. see Supplementary Table 1). F. silvestris silvestris).3
Expected Terminal Branch Length [HEDGE]
Fig. Colors of dots refer to IUCN conservation status.Author's personal copy
I. or perhaps only from the latter. Agnarsson et al. respectively. 2001).. lybica and F. This supports Christiansen’s (2006) claims of convergent evolution in ‘‘saber-tooth” characteristics in the clouded leopard with those in the extinct saber-toothed cats. ornata. lybica and F. the reader is referred to the web version of this article. 2006) for vicariance to have played a major role (Mattern and McLennan. silvestris ornata). during glaciations.
Our Bayesian results agree with the ﬁndings of Driscoll et al. The clade Neofelis + Panthera groups the biggest cats (Panthera) with two medium sized Asian species. Our ﬁndings thus suggest a long-range dispersal of the jaguar ancestor. species [EDGE]
0. Species ranking high in both analyses are highlighted (for details. Hyaenidae. The modern cat radiation is much too recent (<11 mio years ago.0 0. (2007) grouped the domestic cat with both F. lybica and F. while some endangered species have closely genetically similar. However.1
Flat headed cat
Canis simensis MustelaProcyon lotor lutreola
0. Hence. silvestris.3
Panda Hawaiian monk seal Giant otter Otter civet Owston's palm civet
Evolutionary Distinctiveness. s. and the likelihood result placed such clade as sister to the remaining Felis. The sister relationship of the puma and the extinct American cheetah is well supported. s. in turn.
. 1859. N.. 192–201. R.. 5. 90–97. D. the hypothesis we propose here should. Purvis. Mitogenomic relationships of placental mammals and molecular estimates of their divergences. Mol. Strobeck.. 426–443. 361–373... K.. 2008). but contrast considerable to morphological based phylogenies (see Gaubert et al. how one estimates extinction risk of species strongly impacts EDGE and HEDGE analyses (Mooers et al.. The phylogeny of Cetartiodactyla: the importance of dense taxon sampling..E. Petrov. Janke... and their kin. Moore.L. Faith. Honeycutt. A. E.A. Phylogenet.E. our analysis underlines the importance of many ongoing conservation efforts for species such as monk seals.P. Science 317. snow leopard. 1186–1198. Bardeleben. Linn. cats. Berta. 2006. A species-level phylogenetic supertree of marsupials. O.. 2007) and HEDGE (Steel et al. 143–175. etc. C.. Strobeck.1016/j.. Arnason. Bininda-Emonds..R.. our results are in general similar to Veron et al.. M. 2008. 1991). Pontier.. E. Thus we do not intend this preliminary analysis as a guideline to. Hupe. Building large trees by combining phylogenetic information: a complete phylogeny of the extant Carnivora (Mammalia). 273.. 48. London. we conclude by providing a preliminary phylogenetic assessment of conservation priorities within Carnivora (Fig. S. 41. A. Cardillo.. A. instead of grouping the hyenas together. Adegoke.L. I.R. O.. Jones. Biol. Driscoll. e. Zool.. Evol. phylogenetic understanding of carnivores seems strong. Mol. and the placement of carnivoran species in families. Systematics of mustelid-like carnivores. Cladistics 8. 977–992.P. Phylogenet.. A. C. 141–157. Biol. A. economic value. 2007... A. point to several problematic issues with taxonomy at lower levels. S. Evolution and systematics of the feliform Carnivora..4. D.A.. and the collection of other types of data where a broad taxon sampling is better emphasized than in studies hitherto. This implies that the insectivorous habits of the aardwolf evolved within this clade of otherwise hunters/scavengers. Macdonald. our study. 41.. The EDGE (Isaac et al. including ecological role of species.. Supertrees are a necessary not-so-evil: a comment on Gatesy et al.A... it does provide information beyond mere extinction risk of species. 52. and thus a good candidate for rapid improvement in phylogenetic knowledge of large clades. The Near Eastern origin of cat domestication. 2007) metrics consider both evolutionary distinctiveness (i.A. Phylogenetic reconstruction of carnivore social organizations. Mitogenomic analyses of caniform relationships. Bonn. J. a Slovenian Research Agency research fellowship ARRS Z1-97990618-07 to I. U.P. Bininda-Emonds. Syst. J. Gene 421.. In general. D. Within mongooses (Herpestidae).
. M. 1990. missing data. 267. Mol. charisma. O. Johnson. Lehman. A phylogeny of the Caniformia (order Carnivora) based on 12 complete protein-coding mitochondrial genes. P. 2005). Habib. Part B. A. Wayne. in the online version. our study highlights the need for additional sampling of full-length cytb se-
quences as a rapid means of estimating phylogenetic relationships.P. 1997. Janke. 2005.B. suggesting serious taxonomic problems within the family as also found by Veron et al.. John Murray. and the remarkable promise of cytochrome b to provide reliable species-level phylogenies.. E.J. 2003. 724–729. cytb is an extraordinary phylogenetic marker when taxa are densely sampled.. Gullberg.. R.ympev. P. Stirling. Arnason. Phylogenet. Evol. Appendix A. Gullberg. J. Bininda-Emonds. J.. 257–282. 37. A. Systematics and conservation—on predicting the feature diversity of subsets of taxa... Ray. Evol.. Boakes. 45... 363–377.. W. D. Harley. I. overall. Mol. but also highlights species which have received relatively less conservation attention such as the Liberian mongoose. O. As our results largely agree with the best supported clades that prior studies agree on (benchmark clades). J.e. C. Kullberg. and the inﬂuence of fossil taxa and missing characters..H. Paleontol. and giant and sea otters..S... 815–831. Pinniped phylogeny and a new hypothesis for their origin and dispersal.R. Kullberg. Delisle. 72. Darwin. F. However. 2007.. O. Christiansen. Phylogenet. Delisle. On the Origin of Species by Means of Natural Selection. L. M. 72. 345–354. in preparation). N. J. C. 2004. Within Eupleridae the subfamily Galidiinae (Galidia. Bininda-Emonds.E..A. J. Conclusions We provide a novel species based phylogeny of Carnivora containing about 82% of the known taxonomic diversity of the dogs. Delibes. 1992. A phylogeny of the extant Phocidae inferred from complete mitochondrial DNA coding regions. 2008. Davis. Evol. Purvis... and Salanoia) was supported. at doi:10.L. O’Brien. Yamaguchi. I.. Based on this. while Euplerinae (Crytoprocta and Fossa) was not.Author's personal copy
I. Biol...10. Mamm. and some unresolved issues in terms of placement of some families. May-Collado. M. Cardillo. 863–874.. Biol. Cladistics 24. and the highly evolutionary distinct walrus.M. Phylogenet.R. Supertree construction in the genomic age. Gittleman. M. Vert. (2004).P. Evol. R. and thus should be considered as important for conservation of evolutionary diversity. the genus Herpestes is polyphyletic. Skeletal morphology and locomotor capabilities of the archaic pinniped Enaliarctos mealsi. A.P. 4. 2007.. M.K. Notably. Phylogenet.C. Dalerum. Soc. how to allocate funding to conservation effort in Carnivora—a more detailed analysis will be presented elsewhere (May-Collado and Agnarsson.P... Our results are similar to those of Gaubert and Cordeiro-Estrela (2006) based on molecular data. A. A. to assess how conservation efforts might proﬁtably be prioritized (Vane-Wright et al. Kitchener.A. Bininda-Emonds.. A.g. M. C. Mol. Phylogeny of the great cats (Felidae: Pantherinae). U. Morphol.. 2006. J.. Bininda-Emonds. Vainola.. Mol. Decker-Flum. C. M.L. 78. Furthermore.. many other factors play a role when conservation priorities are assessed. 2005. Price. Grenyer. The utility of chemical signals as phylogenetic characters: an example from the Felidae. 519–523. J... I. Methods Enzymol. 2005. giant and red panda. how much unique evolutionary history the species represents) as well as extinction risk. K. Siniff. Gittleman. Mammal. 10. 2001... Menotti-Raymond. What to protect? Clearly. Janke. Galidictis. or the Preservation of Favoured Races in the Struggle for Life. 2008. A. In terms of the basic subdivision of Carnivora into cat-like (Feliformia) and dog-like (Caniformia) animals. 37. J.W. / Molecular Phylogenetics and Evolution 54 (2010) 726–745
support for a sister species relationship between the aardwolf (Proteles cristatus) and the spotted hyena (Crocuta crocuta). Zool... However. species-level. A morphological perspective on the phylogenetic relationships of the extant phocid seals (Mammalia: Carnivora: Phocidae). as have others before. Roca. Arnason.. 2004.2009. 33. O. U. 964–985.033.R. R.H. the exact placement of Fossa is very weakly supported... Monogr. In: Molecular Evolution: Producing the Biochemical Data. Supplementary data Supplementary data associated with this article can be found.L. to what species contain much unique evolutionary history. Our analyses highlight species that may deserve special conservation attention if we are to maximally preserve evolutionary diversity on our planet. E. A molecular phylogeny of the Canidae based on six nuclear loci.L. Rev.. 1996.W. 37–51. R. The primary purpose of the study is to provide a tool for comparative. 1–256..P. 2007. be a useful tool. Sabertooth characters in the clouded leopard (Neofelis nebulosa Grifﬁths 1821). 8). However. the binturong. D. 395. E. Agnarsson et al. E.R. Barycka. Christiansen. Russell. As we have previously found for Cetartiodactyla. Harley. Gullberg. Acknowledgments Funding for this work came from the University of Puerto Rico. comparative studies of carnivores that supplements existing supertrees. 745–757... Dragoo. 74. 11–31. Mungotictis.. Kullberg. (2004) based on complete cytb sequences and eco-ethological characters. Zool. Evol. Gittleman... Geffen. Purvis. and the Integrated Bioscience Program at University of Akron. 264.. References
Agnarsson. 1–15. Conservation priorities As one potential application of phylogenetic results such as ours. 1999.. J.R. C.
2002.. P. Pecon-Slattery. Soibelzon. The Comparative Method in Evolutionary Biology.H. Hughes. J.K. Slater.. Mol. Higdon. In: Carroll. 2007. I. Ronchist. Mol. Lopez-Giraldez. J.. Basal metabolic rate in carnivores is associated with diet after controlling for phylogeny.. T. J. Van Valkenburgh. Baillie. T. S. 41. Phylogenetic diversity and conservation. Ho... 135–144. Wayne. Obrien. Viverridae) within feliformians: implications for faunal exchanges between Asia and Africa. Mosaics of convergences and noise in morphological phylogenies: what’s in a viverrid-like carnivoran? Syst. Spriggs..J. 8. Taxon sampling affects inferences of macroevolutionary processes from phylogenetic trees. W. Mol. Arnason. O.. Hillis.. Redding.P. 2007.. J.M.. J. Collen. D. Zwickl.M.H. S..L.. O.. 17. Phylogenet.. Converting endangered species categories to probabilities of extinction for phylogenetic conservation prioritization. 2008a. Johnson... tempo and biogeographic history of a mammalian adaptive radiation. J. D. T. P. 42.J.W.. C. O.. Pegasoferae. A complete phylogeny of the whales. Contributions to the systematics of the southern fur seals. 243– 264. Wayne. Theor. morphology and the Great American Interchange. J. 232–253. J. Suzuki. Hasegawa. 38. 1991. 2008. Crandall.. Rabeder.C.O. 6.. Hickson. Cladistics 16. Lopez. D. M. 1996. B. Sato. Evidence from nuclear DNA sequences sheds light on the
Faith. Mattern.H. C.A.. The general stochastic model of nucleotide substitution. J. Phylogenet. 1990. Carnivora. M. Price. Kobayashi. BMC Evol. 1171–1177. Washington.. IUCN. 43.. Ledje. Beck. Kolokotronis. Hubbs. S. 2001. Yamaguchi. Leonard.. 1996b.. American Geophysical Union. J. 254.A. Beck..P.
. Teeling. 169. J. 485–501. 20. 67–76. Fine. Maddison.A. J. Turvey. 54. Morphological integration in the carnivoran skull. Y. Phylogenet.. Mol. Biol. 1758–1772. A. D. C. 2009. 11. New York.. 44. K.. USA 103. Bland.A. Bininda-Emonds. A.. Nocon. 7. Ledje. Nedbal. J. PLoS ONE 2 (3)... Evol.Author's personal copy
I. / Molecular Phylogenetics and Evolution 54 (2010) 726–745 Krause.. Patou. R. 1998.. 2007. Nedbal.. Evol. R.iucnredlist. Maddison. 53. S. Sato.J.D. Mammalia).. Munoz-Garcia. Resolution of phylogenetic conﬂict in large data sets by increased taxon sampling..... M.. Soc. 2001.H. Am..P...L.M. Science 298. Gaubert. Hendrickx. M. Penny. C. Veron.. W. 2004. 54.. D. Phylogenetic beta diversity: linking ecological and evolutionary processes across space in time. Begg. J. Rev. 1933. 2008. Biochem. Antarctic Pinnipedia. NY. T. 6. Acad. Nishihara. L.. A. J. Phylogenetic relationships and divergence times among mustelids (Mammalia: Carnivora) based on nucleotide sequences of the nuclear interphotoreceptor retinoid binding protein and mitochondrial cytochrome b genes.B. E. 1971.W. J. Hast. Zool.M. J. Molecular phylogeny of the Arctoidea (Carnivora): effect of missing data on supertree and supermatrix analyses of multiple gene data sets.). M. D.R.A. C. Isaac. 2006.J.. 44. S. Zool. 136. A phylogenetic supertree of the bats (Mammalia: Chiroptera). Strobeck. 28–52. T. BMC Biol..C.K. 2007.. Genets (Carnivora. D. A...P. Mammals on the EDGE: conservation priorities based on threat and phylogeny.M. Wayne. J.. Masuda. The rarer genera of oriental Viverridae.A. Paleobiology 34. S. 2007. 190–199. 9929–9934.P. Nedbal.org>. 47. laryngeal anatomy. J. 1998.. Mol.. Linn. 2007. Slattery.R..A. 2006. Zool. T.org/>.J. New York.D.. Available from: <http://mesquiteproject. 33..P. Hillis.F.B... P.R.. King. Agnarsson. Tuatara: conservation priority in a phylogenetic context. 1998.Ø. an unexpected mammalian clade revealed by tracking ancient retroposon insertions.I. I. Oxford University Press. 11–24. P. G.A. 142. Bray.S..0. Okada.J.. S. Finarelli. 401– 416. Agnarsson et al.. O. A.. Evol.P.N. 2008. Cordeiro-Estrela.O... A. Wayne.... J.M. O’Brien. M.E. Unger. 2006.. Begg. Evol.. Zubaid. Hosoda. P. 2000. J...P.. H..J.. 8. 80. Hiyama. Ancient DNA evidence for Old World origin of New World dogs.. Gaubert.. Phylogeny of the Procyonidae (Mammalia: Carnivora): molecules. Lucherini. K. 2006. C. 1670–1678. Flynn. Pocock... Begg. Multigene phylogeny of the Mustelidae: resolving relationships. Evol. e3700. N.. Koepﬂi. Hsu. U. Incorporating evolutionary measures into conservation prioritization.. Phylogenetic reconstruction of the felidae using 16s rRNA and NADH-5 mitochondrial genes. K. Huelsenbeck. C. Calvignac. Saugetierkd. Fulton.. Peters.. Mooers.. R. 1–15. Biol. A. Natl. Evol. Evol. Deere. 1983. 2005.F. Biol. Mol. Biol. e296.. M. The Late Pliocene vertebrate fauna of Vatera (Lesvos Island. Honeycutt. 2008. A. A. Goswami. Chambers.C.J. 1613– 1616. MarÍn. Z. Wheeler. 246. Sci. 2008.. O’Brien. Available from: <http://mesquiteproject..L.6. A. Wozencraft. version 2. Biol.. Zool.. 220. J. version 1. 2007. Evol. Pagel. 55.. J. Sato.).. Syst. Testing hypotheses of the evolution of encephalization in the Canidae (Carnivora. V..J. 522–529. K.. Fernandes. Finarelli. Felidae). Zool.. Antunes. J. H. N. Buckley. Vila.J.R. 1997.L. Mitochondrial genomes reveal an explosive radiation of extinct and extant bears near the Miocene–Pliocene boundary.. Greece). G.. 1–52. A. J. R. W. Flynn. T.. 1076–1095. D. T. Strobeck. Finarelli.A. 1985. G. Y. Eizirik. J. T. Phylogeny of the Carnivora (Mammalia): congruence vs incompatibility among multiple data sets. 46..P. Wolsan. Proc. S. Modeltest: testing the model of DNA substitution.J..E.... Eizirik. Syst. 2005. May-Collado. J. Cooper.. R. K. Sci. Syst. E. Re-assessed molecular phylogeny and evolutionary scenario within genets (Carnivora.Ø.A. W. Van Neer. E68–E83. Malaspinas. H.. Oliver.. Grassman.. J. J. Biol. Koepﬂi. Flynn. 351–365. 9.. Heath. 445–473. 266–278. 1994.S. Seals of the World. 481–499. Oxford University Press. A.. E. Biol..E. L. Sci.R. Higdon. Phylogenetic measures of biodiversity. D. F. Maddison. 793–808. Hanni. Pages. Purvis. Evolution 60.. Phylogenet. Phylogenetic systematics and tempo of evolution of the Viverrinae (Mammalia.J. G. Mol.. M. Evol. Dragoo. D. Sci.. J. Phylogenetic relationships of the Asian palm civets (Hemigalinae and Paradoxurinae. Biol.. Guillen.A. In: Burt. Hillis.. S.. S. K. 2005. M. dolphins and even-toed hoofed mammals (Cetartiodactyla)..P. D. Hedtke. Mol. Evol. Phylogenet. Zehr. Conserv. 580–601.. Peterson. R. Felsenstein. Williams.. 1933. Bryant.T. Heath... T.. Briggs. University of Queensland Press. Viverridae.. J. Molecular systematics of the Hyaenidae: relationships of a relictual lineage resolved by a molecular supermatrix. Syst. M... Nat. Biol. Jones..M. N. G... Phylogenet. Bininda-Emonds. 589–610.P. 920–927. Linseele. 47.. Koepﬂi. Phylogenet. Phylogenetic review of tonal sound production in whales in relation to sociality.A. M. W. J. Posada. Lett. J. Evol. Veron. Wartzok. 57. O’Brien.J. A. Hofreiter. 20. B. Jenks.. Yamamo. Lento. C. S.A. BMC Evol. Mol. Soc.D. J..A. Biol. 87–104. Novel phylogeny of the raccoon family (Procyonidae: Carnivora) based on nuclear and mitochondrial DNA evidence... Townsend. St... H.M. Physiol. Bioinformatics 17. Phylogeny and divergence of the pinnipeds (Carnivora: Mammalia) assessed using a multigene dataset. A.M. M. S. Proc. 43.J. 125. Yamaguchi. 603– 620. Hosoda.. Science 311.. Evolution 61. 2004. 817–818. Gaubert. S. A. J. Hedtke. P.R. Mesquite: a modular system for evolutionary analysis. J.. Ives. Phylogeny and speciation of felids.. S. Arnason. 169–183. 35–45. Combined analysis of fourteen nuclear genes reﬁnes the Ursidae phylogeny. G. The evolution of encephalization in caniform carnivorans. G. M. Gittleman. 1274. Evol. 38. 865–894.J... 2006. 2007. A.V. L.A. Paris. Biol. BMC Evol. C.P. 2006. Lond. M.. Johnson. Mol. W. 2008b. Mol. Paabo. Phylogeny. 1996a. Wolsan. 2007. May-Collado. Fulton. MRBAYES: Bayesian inference of phylogenetic trees.E. L. (Ed.R. E.. and vocalization in felids (Mammalia.E.K. Suzuki.. M.E...P. D. Slatkin.. Helmus. 2006. Agnarsson. S. <http://www.V. Bininda-Emonds.. 2005. Evol.J. Lucia. Faith. Eizirik. C.. Phylogenetic relationships of otters (Carnivora: Mustelidae) based on mitochondrial cytochrome b sequences. Cytochrome b and Bayesian inference of whale phylogeny. Kim.. 344–354. Simmons. Ferguson. P. Van Der Geer. Maldonado. Rovie-Ryan.M.. pp. 2008. Biol. 2006.. G. Phylogenet..A. H. K. with particular reference to the Juan Fernandez and Guadelupe species. Scr. T. R. Genettinae). R.K. 73–77. Claw retraction and protraction in the Carnivora: the cheetah (Acinonyx jubatus) as an atypical felid. Bioinformatics 14. 2003. 2006. J. W. 160–166... 165–181. Posada.. 2007. J.. T. Phylogenetic analyses of complete cytochrome b sequences of the order Carnivora with particular emphasis on the Caniformia. Phylogenet. Viverridae) in Africa: an evolutionary synthesis based on cytochrome b sequences and morphological characters.K. Harvey.. C... Rev. U. K... N. Whence the red panda? Mol.. J. 2004. G. Bruford.E.. 1995. Mol. P. Gompper. Wayne. 50. 2008..R.. S98–S116. Crandall. Carnivora. PLoS ONE 3 (11). Mol. 2004.. Conservation Biology: Evolution in Action.E. M. evolutionary history and taxonomy of the Mustelidae based on sequences of the cytochrome b gene and a complex repetitive ﬂanking region. Hyoid structure. R... 21. Fox. Medina. Mooers.. BMC Evol. Use of spectral-analysis to test hypotheses on the origin of pinnipeds.A. Archaeol. 81. P.P. C.. W. 414– 426.. Ecol. Evol.B.K. Evol.. (Eds. Phylogenet. K.org/packages/ tuatara>. 2006. Lyras. 78. Zool.. 59. 2081–2090. J. Gaubert.... Waterman.. 77.. F.. S. Stiller.. Minami.. Repenning..Y. C.. D. C.R. Mol.. 969–1035. 43. Evol. 34.. 641– 649. D. J. H. 1265–1277. 12. 7.L. A. 2007. Suzuki.. McLennan. Taxon sampling and the accuracy of phylogenetic analyses. Evidence for early cat taming in Egypt. Marmi. Flynn. S. Veron. G. Zahirpour... 2007.K. M. Mol.. 317–337. Murphy. A. 2000. Russell.. Evol. R. Biol. Syst. Phylogenet.. Selecting the best-ﬁt model of nucleotide substitution.. Phylogenies and the comparative method. Yuhki.P... S. W.J.P.M. Cranium 24.M.. 216. 2002. 73–83. Jennings.L. 2009. Model selection and model averaging in phylogenetics: advantages of Akaike information criterion and Bayesian approaches over likelihood ratio tests. Biol.. R. DC. 754–755... Rodríguez..J. Bininda-Emonds. A. Graham. Zool.A. The Late Miocene radiation of modern Felidae: a genetic assessment. Ferguson.H.W...W. W..P. Veron. Posada. Dear.. Molecular phylogeny of the Carnivora (Mammalia): assessing the impact of increased sampling on resolving enigmatic relationships.H.. M. Carnivora). Matheus.E. Valadez.. R... MacLarnon... Domingo-Roura. Williams..P.. Viverridae. M. S. 883–892.. C. K. R. Molecular phylogeny of mitochondrial cytochrome b and 125 rRNA sequences in the Felidae: ocelot and domestic cat lineages. Linden. 223– 259. M....W. X.. Koepﬂi.E... R. L.E. 239–257. C. Correlation between the radiation event of weasels and Martens (Mustelidae: Carnivora) inferred from multiple nuclear sequences. Klein. NY. 2008. Cordeiro-Estrela.W. Am. S. Biol. P. 2001.A. Wolsan. Debruyne. Maddison. Syst. Phylogeny and divergence of the pinnipeds (Carnivora: Mammalia) assessed using a multigene dataset. S.. Biol.. 41. Mooers. Phylogenet. 1039–1056. M. Phylogenetic relationships within caniform carnivores based on the analyses of the mitochondrial 12S rRNA gene. C. Hosoda. Nat.
2007.. Milne... Gallo. D.. Mol. S. Williams.E. 2005. Conserv.D. 2001. Acad...T. Phylogenet. G. Insley. 33. 329–342. C. C. R. Molecular phylogeny of the red panda (Ailurus fulgens)..C.. G.. Slade. 483–495. 311–333. 306–314..C. Evol.. Wolsan. 270–284.P. A. What to protect? Systematics and the agony of choice. Peters. Weissengruber. Powell. Zhang.. Wayne. Leitner. 1991. Phylogenet. Y..G. Evol. S. G. catus).. 567–575. Science 298. 2004. Y.L.. Zool. 1251–1260.. Scr. 46.. 2006. W.. Fitch. G. Mooers. Evol...
.. Palaeontol. N. J.. 35..D. 235. Phylogenetic relationships within the eared seals (Otariidae: Carnivora): implications for the historical biogeography of the family. Science 276. 39. Phylogenet.. Lundeberg. Marshall. Taylor. Molecular systematics and origin of sociality in mongooses (Herpestidae. / Molecular Phylogenetics and Evolution 54 (2010) 726–745 phylogenetic relationships of pinnipedia: single origin with afﬁnity to Musteloidea. jaguar (Panthera
onca). Shields. 2002.F.I. Lundeberg.. 622–653. D. Anat. Carnivora).R. N. 38. and assessment of the position of ‘Miacoidea’ relative to Carnivora. D. J. Ph.O. K. Rice. Syst. Biol.. Mitochondrial–DNA sequence evolution in the Arctoidea..D. S. (Online) 3. Phylogenet. P. Zrzavy. Milinkovitch. Bickham. 201. Evol. Mol. 47–58.. M. J.P.. M. Savolainen..E. M.J. Hedging one’s bets: quantifying a taxon’s expected contribution to future phylogenetic diversity. O. Cooper.A. Maximum likelihood phylogenetic estimation from DNA sequences with variable rates over sites: approximate methods. 9557–9561. G. Carnivora).. P. Evolution 61.D. 237–244.E.R. Phylogenet... Yang.. Zool. Amorim.H. Evol. Zwickl. A. Carnivora): relative reliability and utility of morphological and molecular datasets. Evol.E. 1996. Crandall. J.A. L.. 3. Maldonado. R. I... 5. Koepﬂi. Yu.P. K. 477–494. Wynen. Yamaguchi.. Sci. Forstenpointner. 24. Mammalia: Carnivora: Mustelidae).W. Y. 86. Gaubert. O.. 2004. 1996b.. 5. J... Evolution of the mane and group-living in the lion (Panthera leo): a review. Phylogenetic relationships within mammalian order Carnivora indicated by sequences of two nuclear DNA genes.W.. Mol.L.. E. Mol. 1–28. Phylogenetic studies of pantherine cats (Felidae) based on multiple genes. Zhang... White. Ryder. 33. Hyoid apparatus and pharynx in the lion (Panthera leo).. Phylogenet. Zool. Veron. L. Slattery. Q. Z. J. 345–384. 2004. Molecular systematics of the Canidae. Girman. G. 1687–1689. R. Dunham. 1996a. Natl. Sci. 125–146.. Higher level relationships of the arctoid Carnivora based on sequence data and ‘‘total evidence”.W. Phylogenet. S. Werdelin. J. J. K.... Wayne. Majluf. 1995. Adams. Zhang. Humphries. Ryder. Hered. Stone... 2007. A... J. Vane-Wright.J. tiger (Panthera tigris).M. Cook. 1997.K. 2004. T. Mol. The University of Texas at Austin. P. Francis. 2002. Biol... Hoelzel. G. Mol. Bioinform. R. Mol. Savolainen.R..J.J.. M. Phylogenet. Agnarsson et al. Molecular evolution of Holarctic martens (genus Martes. A. J.. A.P. Flynn. 2002. Mol.. J. Phylogeny of the Carnivora: basal relationships among the Carnivoramorphans. J. Dissertation. L. Multiple and ancient origins of the domestic dog. 2005.W.. Proc. 23.J. 1994.A. Lau. Genetic evidence for an East Asian origin of domestic dogs. 1994.. Genetic algorithm approaches for the phylogenetic analysis of large biological sequence datasets under the maximum likelihood criterion.. 195–209. P. Wheeler. Geffen. Wroe. 169–179.....P. Honeycutt. 30. S. Evol.F. Zhang. Phylogeny and classiﬁcation of early European Mustelida (Mammalia. Luo. 263. Li.A. Talbot... Shields.P. Wesley-Hunt. J.L... Vila. cheetah (Acinonyx jubatus) and domestic cat (Felis silvestris f... Macdonald. M. S. V. P. Evol.K. Evol. 694–705. J. Mimoto.J. 1610–1613.Author's personal copy
I. 3. J.P. 21. USA 90. C. with novel application of nuclear beta-ﬁbrinogen intron 7 to carnivores. J.R. L. Mol.. R. 582–598. Evol..P. O’Brien. 413–422. Yu. Colyn. 55. P. 1993. L. A phylogeny of the bears (Ursidae) inferred from complete sequences of three mitochondrial genes. Ricankova. J..B. W. R. Phylogeny of recent Canidae (Mammalia. Syst. Goldsworthy. Steel. Talbot. J. Kubber-Heiss. A. P. Convergence and remarkably consistent constraint in the evolution of carnivore skull shape. Acta Theriol. Vrana.... Y. 1997. Phylogeography of brown bears (Ursus arctos) of Alaska and paraphyly within the Ursidae.