Molecular evidence for the polyphyly of the genus Hemitragus (Mammalia, Bovidae)

Molecular Phylogenetics and Evolution 36 (2005) 154–168 www.elsevier.com/locate/ympev

Molecular evidence for the polyphyly of the genus Hemitragus (Mammalia, Bovidae) Anne Ropiquet, Alexandre Hassanin ¤ UMR 5202—Origine, Structure et Evolution de la Biodiversité, Département Systématique et Evolution, Muséum National d’Histoire Naturelle, Case postale No. 51, 55 rue BuVon, 75005 Paris, France Service de Systématique Moléculaire, Muséum National d’Histoire Naturelle, 43 rue Cuvier, 75005 Paris, France Received 15 October 2004; revised 20 December 2004 Available online 12 February 2005

Abstract The genus Hemitragus includes three species of tahr distributed in distant geographical areas: the Himalayan tahr, H. jemlahicus, occupies the southern Xanks of the Himalaya Mountains; the Nilgiri tahr, H. hylocrius, is endemic to southern India; and the Arabian tahr, H. jayakari, is unique to the mountains of south-east Arabia. All previous investigations were based on morphology, and these three species together have never been included in a molecular phylogenetic study. In this study, we constructed a molecular phylogeny of the tribe Caprini sensu lato to determine the taxonomic status of the three species of tahr. Phylogenetic analyses were carried out on a matrix including most extant species currently described in the tribe Caprini sensu lato, and 3165 nucleotide characters, coming from four diVerent markers, i.e., an intron of the nuclear gene coding for the protein kinase C iota, and three mitochondrial genes (subunit II of the cytochrome c oxidase, cytochrome b, and 12S rRNA). The results show that the genus Hemitragus is polyphyletic, as H. jemlahicus is associated with Capra (goats), H. hylocrius is the sister-group of Ovis (sheep), and H. jayakari is allied with Ammotragus lervia (aoudad). In the light of these unexpected results, we revaluate the validity of the morphological characters originally used for deWning the genus Hemitragus. At least, we propose a new taxonomy, where the three species of tahr are ranged into three monospeciWc genera: the genus Hemitragus is restricted to the Himalayan tahr, and two new genera are created: Arabitragus for the Arabian tahr and Nilgiritragus for the Nilgiri tahr.  2005 Elsevier Inc. All rights reserved. Keywords: Phylogeny; Taxonomy; DNA; Arabitragus; Hemitragus; Nilgiritragus; Caprini; Bovidae

1. Introduction Within the family Bovidae (Mammalia, Ruminantia), the genus Hemitragus belongs to the Caprini sensu lato, a tribe characterized by 10 additional genera (Hassanin and Douzery, 1999a): (1) Ammotragus (aoudad), (2) Budorcas (takins), (3) Capra (goats, ibexes, markhor, and turs), (4) Naemorhedus (gorals and serows), (5) Oreamnos (Rocky Mountain goat), (6) Ovibos (muskox), (7) Ovis (sheep, argalis, and mouXons), (8) Pantholops ¤

Corresponding author. Fax: +33 1 40 79 30 63. E-mail address: [email protected] (A. Hassanin).

1055-7903/$ - see front matter  2005 Elsevier Inc. All rights reserved. doi:10.1016/j.ympev.2005.01.002

(chiru), (9) Pseudois (bharals), and (10) Rupicapra (chamois and isard). The genus Hemitragus includes three species of tahr distributed in distant geographical areas (Shackleton, 1997): (1) the Himalayan tahr, H. jemlahicus (H. Smith, 1826), occupies the southern Xanks of the Himalaya Mountains from northern India east to Bhutan, as far north as Tibet; (2) the Nilgiri tahr, H. hylocrius (Ogilby, 1838), is endemic to southern India and lives along the southern parts of a range of hills known as the Western Ghats in the Indian states of Kerala and Tamil Nadu; and (3) the Arabian tahr, H. jayakari (Thomas, 1894), is unique to the mountains of south-east Arabia in Oman

A. Ropiquet, A. Hassanin / Molecular Phylogenetics and Evolution 36 (2005) 154–168

and the United Arab Emirates. Most classiWcations of mammals agree with the deWnition of three species of tahr (e.g., Grubb, 1993; Lydekker, 1913; Nowak, 1999; Shackleton and Lovari, 1997; Simpson, 1945), but the one proposed by Haltenorth (1963) considers the three types of tahr as being subspecies of H. jemlahicus. Phylogenetic aYnities between the tahrs and goats were detected from the beginning by taxonomists. In fact, the Himalayan tahr was initially included in the same genus as the goats, and named Capra jemlahica (Smith, 1826). In 1841, Hodgson created the genus Hemitragus for the Himalayan tahr and described the species H. quadrimammis as having a “general structure, odour, and horns of Capra, but having a small moist muzzle, and four teats in the females.” In 1847, Gray established a variation on the original speciWc name of the Himalayan tahr, referring to it as H. jemlaicus. The name, which is known today, H. jemlahicus, was proposed in Lydekker (1913). The Nilgiri tahr was Wrst named Kemas hylocrius by Ogilby (1838). In 1859, Blyth included the Nilgiri tahr in the genus Hemitragus, and named it H. hylocrius. However, as late as 1891, Sclater was still using the genus Capra for both Himalayan and Nilgiri species. By contrast, the Arabian tahr was described as H. jayakari by Thomas in 1894, and always belonged to the genus Hemitragus. All previous phylogenetic studies based on DNA indicated a close relationship between tahrs and goats (e.g., Gatesy et al., 1997; Groves and Shields, 1996; Hassanin and Douzery, 1999a; Hassanin et al., 1998a). Indeed, the Himalayan tahr was found allied with the domestic goat (Capra hircus) by analyzing partial sequences of 12S and 16S rRNA genes (Gatesy et al., 1997), as well as the complete cytochrome b (Cyb) gene (Groves and Shields, 1996). By integrating seven additional species of Capra in the analyses of the Cyb gene, Hassanin et al. (1998a) have conWrmed the strong aYnities between Hemitragus and Capra, but some phylogenetic inferences have suggested a possible association of the Siberian ibex (Capra sibirica) with the Himalayan tahr (H. jemlahicus), making the genus Capra paraphyletic. Relationships between the three species of tahr have been poorly studied, but all the three possible topologies can be proposed: (1) on the basis of external characters, H. jemlahicus and H. jayakari may be grouped together, as the females have four teats, whereas those of H. hylocrius and Capra have only two teats; (2) by comparing their geographic distributions, H. jemlahicus and H. hylocrius may be linked, as both are found in India; and (3) by analyzing cytogenetic data, the Arabian and Nilgiri tahrs may be associated, as they share the same diploid number of chromosomes (2n D 58), which diVers from those found for the Himalayan tahr (2n D 48), and goats (2n D 60) (Benirschke and Kumamoto, 1982). Up to now, the only species integrated in the molecular studies was H. jemlahicus (Gatesy et al., 1997; Groves and

155

Shields, 1996; Hassanin and Douzery, 1999a; Hassanin et al., 1998a), and no sequences were available for H. hylocrius and H. jayakari in the nucleotide databases. In the present study, we constructed a molecular phylogeny including all the three species of the genus Hemitragus, as well as most species currently described into the tribe Caprini sensu lato. To test whether the genus Capra is paraphyletic with respect to Hemitragus, Wve species of Capra were incorporated in the analyses: C. falconeri (markhor), C. hircus (domestic goat), C. ibex (Alpine ibex), C. nubiana (Nubian ibex), and the problematic C. sibirica (Siberian ibex). Four molecular markers were used for a taxonomic sample consisting of 30 species. Three genes of the mitochondrial genome were sequenced: the subunit II of the cytochrome c oxidase (CO2), the cytochrome b (Cyb), and the 12S rRNA gene (12S). These mitochondrial markers were chosen because they guarantee to obtain a lot of informative sites for inferring phylogenetic relationships between species of Hemitragus and Capra. Although nuclear genes are likely to be too much conserved for such recent divergences (Hassanin and Douzery, 1999b; Matthee and Davis, 2001), we also sequenced an intron of the nuclear gene coding for the protein kinase C iota (PRKCI). This nuclear marker was chosen because the PRKCI sequence of C. hircus (domestic goat) presents a deletion of four nucleotides (TYGA), which was not detected in two other caprine species, i.e., Ovibos moschatus (muskox) and Ovis aries (domestic sheep) (Matthee et al., 2001). This deletion was therefore a potential signature for diagnosing all species of the genus Capra. By analyzing these molecular data, we show that Hemitragus is polyphyletic, as the three species of tahr are associated with three diVerent genera: H. jemlahicus with Capra, H. hylocrius with Ovis, and H. jayakari with Ammotragus. We discuss the evolution of morphological characters in the light of these unexpected relationships, and we propose to place each species of tahr in its own genus for reconciling taxonomy with our phylogenetic results.

2. Materials and methods 2.1. Taxonomic sample Our taxonomic sample contains 30 taxa with 28 bovid species and two species belonging to the families Antilocapridae (Antilocapra americana) and Cervidae (Cervus elaphus) (Table 1). Within the family Bovidae, we have incorporated various members of the two major subfamilies: the subfamily Bovinae is represented by three species, including one species for each of the three tribes Bovini, Boselaphini, and Tragelaphini; and the subfamily Antilopinae sensu Kingdon (1982, 1997) is represented by 25 species, including members of Wve diVerent

156

A. Ropiquet, A. Hassanin / Molecular Phylogenetics and Evolution 36 (2005) 154–168

Table 1 Taxonomic sample used in the present study Family

Tribe

Species

Antilocapridae Antilocaprini Antilocapra americana Cervidae Cervini Cervus Bovidae

a b c d e f g h i j k l m n o p q r s t

Bovini Boselaphini Tragelaphini Antilopini

Bos taurus Boselaphus tragocamelus Tragelaphus imberbis Gazella

Aepycerotini Alcelaphini Hippotragini Caprini s.l.

Aepyceros melampus Damaliscus pygargus Hippotragus niger Ammotragus lervia Budorcas taxicolor Capra falconeri Capra hircus Capra ibex Capra nubiana Capra sibirica Hemitragus hylocrius Hemitragus jayakari Hemitragus jemlahicus Naemorhedus sumatraensis Oreamnos americanus Ovibos moschatus Ovis ammon Ovis aries Ovis aries Ovis dalli Pantholops hodgsonii Pseudois nayaur Rupicapra pyrenaica Rupicapra rupicapra

Common name

Cytochrome oxidase 2

U62571k C. unicolor U62570k Domestic Cow NC_001567d Nilgai U62566l Lesser Kudu U18815m Gazelle G. spekei U18824m Impala AY689194n Blesbok AY689195n Sable Antelope AY846771a Aoudad AY846772a Takin AY846773a Markhor AY846774a Domestic goat NC_005044g Alpine Ibex AY846775a Nubian Ibex AY846776a Asiatic Ibex AY846777a Nilgiri Tahr AY846778a Arabian Tahr AY846779a Himalayan Tahr AY846780a Sumatran Serow AY846781a Rocky Mountains Goat AY846782a Muskox AY846783a Gobi Argali AY846784a Domestic Sheep (Corse) AY846785a Domestic Sheep NC_001941i Dall’s sheep AY846786a Chiru AY846787a Bharal AY846788a Pyrenean Chamois AY846789a Alpine Chamois AY846790a

Pronghorn Deer

Cytochrome b 12S rDNA

Protein kinase C iota

AF091629c C. elaphus AJ000021o NC_001567d AJ222679p AF036279p G. gazella AF034723p AF036289p AF036287p AF036285p AF034731q AY669320f AF034736q NC_005044g AF034735q AF034740q AF034734q AY846792a AY846791a AF034733q AY669321f AF190632r AY669322f AF034727q AF034730q NC_001941i AF034728q AF034724q AF034732q AF034726q AF034725q

AF165669s C. elaphus AYH46793a AY029293t AF165725s AF165733s G. thomsonii AF165749s AF165781s AY846794a AY846795a AY846803a AY846811a AY846797a AF165797s AY846798a AY846799a AY846800a AY846808a AY846804a AY846801a AY846812a AY846814a AY846813a AY846805a AY846806a AF165789s AY846807a AY846796a AY846802a AY846810a AY846809a

M55540b C. elaphus AF091707c NC_001567d M86494e M86493e G. granti AY670652f M86496e M86499e AY670653f AY670654f AY670655f AY670656f NC_005044g AY846815a AY670657f AY670658f AY846817a AY846816a AY670659f AY670660f AY670661f AY670662f AY141135h AY670663f NC_001941i AY670664f AF400659j AY670665f AY846818a AY670666f

This study. Kraus and Miyamoto (1990). Hassanin and Douzery (1999b). Anderson et al. (1982). Allard et al. (1992). Ropiquet and Hassanin (in press). Feligini and Parma (2003). Kuznetsova and Kholodova (unpublished). Hiendleder et al. (1998). Kuznetsova and Kholodova (2002). Miyamoto et al. (1994). Janecek et al. (1996). Honeycutt et al. (1995). Hassanin and Ropiquet (2004). Randi et al. (1998). Hassanin and Douzery (1999a). Hassanin et al. (1998a). Hassanin and Douzery (2000). Matthee et al. (2001). De Donato et al. (2001).

tribes, i.e., Aepycerotini, Alcelaphini, Antilopini, Caprini sensu lato, and Hippotragini. The ingroup corresponds to the tribe Caprini sensu lato, and incorporates 21 species, with a least one species for each of the 11 genera currently recognized into this group (Hassanin and Douzery, 1999a; Hassanin et al., 1998a).

2.2. DNA extraction DNA was extracted from blood for Ammotragus lervia, Capra nubiana, C. sibirica, Damaliscus pygargus, Ovis aries, and Pseudois nayaur, from skin for Hemitragus jayakari and H. jemlahicus, from heart for Ovibos

A. Ropiquet, A. Hassanin / Molecular Phylogenetics and Evolution 36 (2005) 154–168

moschatus, from muscles for Rupicapra pyrenaica and R. rupicapra, and from cells for Capra falconeri, C. ibex, Cervus elaphus, Hippotragus niger, and Oreamnos americanus. The protocol of DNA extraction includes a digestion with hexadecyltrimethyl ammonium bromide (CTAB), followed by a deproteinisation with chloroform isoamyl alcohol (CIA) and thereafter a cold precipitation with propan-2-ol (Winnepenninckx et al., 1993). Bone samples from museum specimens were used for DNA extraction by applying the protocol detailed in Hassanin et al. (1998a). This protocol was used for the following specimens, which are preserved either in the collections of the Muséum National d’Histoire Naturelle of Paris (MNHN, Zoothèque), or at the International Foundation for the Conservation of Wildlife (IGF): Budorcas taxicolor (MNHN, No. 1902-409), H. hylocrius (MNHN, No. 1935-402), Naemorhedus sumatraensis (MNHN, No. 1993-4240), Ovis ammon (IGF, No. 1985BdC2), Ovis dalli (No. 1938-124), and Pantholops hodgsonii (No. 1993-4237). 2.3. AmpliWcation and sequencing The sequence of the subunit II of the cytochrome c oxidase gene (CO2; 582 bp, positions 7019–7600 in the mitochondrial genome of O. aries, Accession No. NC_001941) was produced from two overlapping fragments generated by polymerase chain reaction (PCR) using the two following couples of primers: U1: 5⬘-GTG AAA ATC CYG TAC AYC TCA T-3⬘ with L374 (reverse): 5⬘-CTC CTG GYT TYA RTT CTG ATG-3⬘; and U291: 5⬘-TCA ACA ACC CAT CYC TCA CAG T3⬘ with L582 (reverse): 5⬘-CCR CAR ATY TCT GAR CAT TG-3⬘. The complete sequences of the cytochrome b (Cyb; 1140 bp) and 12S rRNA (12S; 958 bp) genes were ampliWed with two sets of primers previously published (Hassanin and Douzery, 1999a; Hassanin et al., 1998a; Ropiquet and Hassanin, in 2005). The intron of the nuclear gene for the protein kinase C iota (PRKCI; 513 bp, positions 26–538 in the sequence of O. aries, Accession No. AF165789) was ampliWed with the following couple of primers: U26: 5⬘-TAT GCT AAA GTA CTG TTG GT-3⬘ and L748 (reverse): 5⬘CTG TAC CCA GTC AAT ATC CT-3⬘. Standard PCR conditions were as follows: 3 min at 94 °C; 35 cycles of denaturation/annealing/extension with 1 min at 94 °C for denaturation, 1 min at 55 °C for annealing, and 1 min at 72 °C for extension; and 7 min at 72 °C. The PCR product puriWcations were realized from Montage PCR Centrifugal Filter Devices (Millipore). The puriWed PCR products were then used as starting templates for sequencing using the CEQ2000 Dye terminator cycle Sequencing Quick start kit. The sequencing reactions were run on a Beckman CEQ2000 automatic capillary sequencer.

157

2.4. DNA alignment The nucleotide sequences were aligned manually with Se-Al v1.0al (Sequence Alignment Editor Version 1.0 alpha 1; Rambaut, 1996). The CO2 and Cyb genes were aligned using the amino-acid sequences. The 12S gene was aligned using the models of secondary structure available in the SSU rRNA database (http:// rrna.uia.ac.be/ssu/; Wuyts et al., 2002) for O. aries (Accession No. AF010406), C. hircus (M55541), Damaliscus dorcas (M86499) and Oryx gazella (M86500). By utilizing this approach, we were able to identify regions involved in helix, loop, or sites involved in non-standard pairing (any pair other than G · C, A · U, or G · U). All regions with ambiguity for the position of the gaps in 12S and PRKCI genes were excluded from the analyses to avoid erroneous hypotheses of primary homology. The gap placement was considered unambiguous when only one local sequence alignment was possible due to the conservation of both gap length and nucleotide motifs adjacent to the 5⬘ and 3⬘ boundaries of the gap. Unambiguous insertions and deletions (indels) were then coded as additional characters by using 1 and 0 symbols for insertion and deletion, respectively (SwoVord, 1993). 2.5. Phylogenetic analyses The four markers were analyzed separately and also combined to beneWt from the maximum number of molecular characters. Maximum likelihood (ML), Bayesian, and maximum parsimony (MP) methods were used for phylogenetic reconstructions. The ML analyses were carried out under PHYML (Guindon and Gascuel, 2003). MODELTEST 3.06 (Posada and Crandall, 1998) was used on the data matrix combining all the four markers for choosing the model of DNA substitution that best Wts our data. The selected likelihood model was the General Time Reversible model (Yang, 1994a) with among-site substitution rate heterogeneity described by a gamma distribution and a fraction of sites constrained to be invariable (GTR + I +
). Eight gamma categories (
8), rather than four, were used for the analyses to get a good approximation to the continuous gamma model, and therefore better manage rate heterogeneity (Yang, 1994b). Bootstrap percentages (BP) were computed as follows: (1) 1000 bootstrapped data sets were generated with the program SEQBOOT in the PHYLIP package Version 3.6b (Felsenstein, 2004); (2) the 1000 data sets were analyzed with PHYML under the GTR + I +
8 model; and (3) bootstrap percentages (BPML) were then computed using the program CONSENSE in the PHYLIP package. Bayesian analyses were performed with Mr. Bayes 3.0b4 (Huelsenbeck and Ronquist, 2001). The Bayesian approach evaluates the posterior probability (PPB) of a

158

A. Ropiquet, A. Hassanin / Molecular Phylogenetics and Evolution 36 (2005) 154–168

tree given the character matrix, i.e., the probability that the tree is correct. A partitioned Bayesian analysis was conducted to account for the combination of markers with contrasted molecular properties, i.e., mitochondrial versus nuclear and protein coding or ribosomal versus non-coding. Nine partitions were distinguished in the original data set according to the structural and functional properties of the markers: three codon-positions for CO2 and Cyb, the helix and not helix regions for 12S, and PRKCI. For each partition, MrModeltest 2.0 (Nylander, 2004) was used for choosing the model of nucleotide substitution that best Wts the data. Using the Akaike information criterion, the GTR + I +
model was selected for most partitions. In addition, the best log likelihood score was always obtained with the GTR + I +
model for each of these nine partitions. Consequently, this model was used for all nine nucleotide partitions. Unambiguous indels were analyzed as an additional partition and treated as morphological characters. All analyses were conducted with Wve independent Markov chains run for 1,000,000 Metropolis-coupled MCMC generations, with tree sampling every 100 generations, and burn-in after 1000 trees (as detected by plotting the log likelihood scores against generation number). The analyses were run twice using diVerent random starting trees to evaluate the convergence of the likelihood values and posterior clade probabilities (Huelsenbeck et al., 2002). The MP analyses were conducted in PAUP 3.1.1 (SwoVord, 1993) with diVerential weighting of the character-state transformations using the product CIex. S (CIex, consistency index excluding uninformative characters; S, slope of saturation) as detailed in Hassanin et al. (1998a,b): for each substitution-type (i.e., A–G, C–T, A–C, A–T, C–G, G–T, and indels), the amount of homoplasy was measured through the CIex, and the saturation was assessed graphically by plotting the pairwise number of observed diVerences against the corresponding pairwise number of inferred substitutions calculated by PAUP (the slope of the linear regression [S] was used to evaluate the level of saturation). The CIex and S values were calculated by distinguishing each of the three codon-positions separately for CO2 and Cyb, helix and not-helix regions for 12S, PRKCI, and indels. The MP tree was found by heuristic search using 1000 replicates of random stepwise addition. Bootstrap percentages (BPMP) were computed after 1000 replicates of the closest stepwise addition option. 2.6. Molecular dating Two diVerent approaches were used for the estimation of divergence dates on the Bayesian topology presented in Fig. 2: the relaxed molecular clock under Multidivtime, and the local molecular clock under PAML 3.14beta. Two diVerent calibration points were used for the analy-

ses: (1) in agreement with the fossil record (Vrba and Schaller, 2000), the emergence of the family Bovidae was Wxed between 18 and 20 Mya; and (2) the common ancestor between Merinos and Corsican breeds of domestic sheep was Wxed between 11,000 and 200 years before present in agreement with archaeological data concerning the origin of sheep domestication (Maijala, 1997), and the history of sheep breeds in Corsica and Sardinia (Bougler et al., 1988; Franceschi and Vallerand, 1988). The Bayesian relaxed molecular clock approach developed by Thorne et al. (1998) and Kishino et al. (2001) was performed with the software Multidivtime. The dating procedure involved two steps. First the program ESTBRANCHES estimated branch lengths and the variance–covariance matrix for each of the nine partitions: three codon positions separately for CO2 and Cyb, helix and not helix regions for 12S, and PRKCI. The F84 nucleotide substitution model was the only one implemented for DNA under ESTBRANCHES. Second, after pruning the outgroup taxa (Antilocapra), the program MULTIDIVTIME estimated the prior and posterior ages of divergence times, and their standard deviations (SD) and 95% credibility intervals (CI95%). The Markov chain was sampled 10,000 times with 100 cycles between each sample, and burn-in after 100,000 cycles. As priors, we adopted 30 Mya (SD D 15) for the expected number of time units between tip and root, and 0.089 (SD D 0.0445) substitutions per site per million years for the rate at root node, and 50 Mya for the highest possible number of time units between tip and root. These priors were determined using the strategy described in the multidivtime.readme Wle. To check for convergence of MCMC analyses, two independent runs were performed for the same data and same prior distributions, but with diVerent initial seed number. Under BASEML in the PAML 3.14beta software (Yang, 2003), the divergence times were estimated with the local clock model for combined analysis of multiplepartition data, which allows the branch group rates to vary in diVerent ways among the data partitions (Yang and Yoder, 2003). We assigned independent rates for several groups of branches on the tree, corresponding to the following clades: Bovidae, Bovinae, Antilopinae, Caprini sensu lato, (Ammotragus + H. jayakari), (Capra + H. jemlahicus), and (Ovis + H. hylocrius). The analyses were carried out using a GTR model with a gamma distribution with eight categories for each of the nine following partitions: three codon positions separately for CO2 and Cyb, helix and not helix regions for 12S, and PRKCI. Four distinct analyses were performed with four diVerent couples of calibration points: (1) 18 Mya and 11,000 years; (2) 18 Mya and 200 years; (3) 20 Mya and 11,000 years; and (4) 20 Mya and 200 years. For each node of interest, four dates were thus computed, and the values presented in Table 2 are averages calculated by using the youngest and oldest dates.

A. Ropiquet, A. Hassanin / Molecular Phylogenetics and Evolution 36 (2005) 154–168

159

Table 2 Molecular dating of the main splitting events within Bovidae Divergences between taxa

Bovinae Antilopinae sensu lato (Caprini, ƒ Hippotragini) Alcelaphini + Hippotragini Caprini sensu lato Caprini s.l. excluding Pantholops Naemorhedus + Ovibos Rupicapra Ovis H. hylocrius + Ovis H. jayakari + Ammotragus (H. jemlahicus, ƒ Ammotragus) (H. jemlahicus, ƒ Pseudois) H. jemlahicus + Capra H. jemlahicus + C. sibirica a b

Age estimates (Mya)

Epochs

PAML

Multidivtime (CredI95%)a

Maximum rangeb

16.1 § 1.6 16.3 § 1.5 13.3 § 1.4 12.1 § 1.4 10.6 § 1.3 9.2 § 1.2 4.6 § 0.7 2.1 § 0.4 3.0 § 0.5 4.5 § 0.7 6§1 7.8 § 1.1 6.4 § 1 4.5 § 0.8 4 § 0.7

14.1 § 1.4 (11.3–17) 15.7 § 1.1 (13.7–17.8) 12.5 § 1.2 (10.3–14.9) 10.9 § 1.2 (8.6–13.4) 9.9 § 1.2 (7.7–12.3) 8.6 § 1.1 (6.6–10.7) 4.1 § 0.8 (2.8–5.8) 1.5 § 0.4 (0.9–2.3) 2.4 § 0.5 (1.5–3.4) 3.3 § 0.6 (2.2–4.5) 4.7 § 0.7 (3.4–6.3) 5.8 § 0.8 (4.4–7.6) 4.8 § 0.7 (3.5–6.4) 3.4 § 0.6 (2.4–4.6) 3 § 0.5 (2–4.1)

12.7–17.7 14.6–17.8 11.3–14.7 9.7–13.5 8.7–11.9 7.5–10.4 3.3–5.3 1.1–2.5 1.9–3.4 2.7–5.2 4–7 5–8.9 4.1–7.4 2.8–5.3 2.5–4.7

Middle Miocene Middle Miocene Middle Miocene Middle/Late Miocene Late Miocene Late Miocene Early Pliocene Pleistocene Late Pliocene Pliocene Late Miocene Early Pliocene Late Miocene Late Miocene Early Pliocene Pliocene Pliocene

The 95% credibility intervals (CredI95%) for divergence ages are given in parentheses. From the two dating analyses.

3. Results 3.1. Phylogenetic analyses The four diVerent markers (CO2, Cyb, 12S, and PRKCI) were analyzed separately and also combined (3165 nucleotides and Wve unambiguous indels). In Figs. 1–3 are presented the phylogenetic trees performed with the data matrix combining all the four markers, and with ML, Bayesian, and MP methods, respectively. They appear very similar except a few topological diVerences, which are not strongly supported. We found strong evidence for the monophyly of several higher taxa, including the family Bovidae (BPML D 92; PPB D 1; BPMP D 98) and the two subfamilies Bovinae (BPML D 99; PPB D 1; BPMP D 98) and Antilopinae (BPML D 99; PPB D 1; BPMP D 99). Within the subfamily Antilopinae, the tribes Alcelaphini (represented by Damaliscus) and Hippotragini (represented by Hippotragus) were grouped together (BPML D 87; PPB D 0.96; BPMP D 93), and they were associated with the tribe Caprini sensu lato (BPML D 85; PPB D 1; BPMP D 91). The tribe Caprini s.l. was found monophyletic in the combined analyses (BPML D 100; PPB D 1; BPMP D 100), and with the three mitochondrial markers analyzed separately (see values in Figs. 1–3). Moreover, this group is characterized by three exclusive synapomorphies in the 12S alignment, including an insertion of T (at position 73 in the sequence of O. aries NC_001941) and two transitions (G ! A and T ! C, respectively at position 108 and 752). Within the tribe Caprini s.l., most basal relationships were not highly supported in terms of PPB or BP values, but Pantholops was found to diverge Wrst in the analyses combining all the four markers (BPML D 96; PPB D 1;

BPMP D 98), and in the independent analyses of CO2, 12S and PRKCI (see values in Figs. 1–3). This placement of Pantholops is also supported by the fact that all other caprines share one exclusive synapomorphy, i.e., a transversion A ! T at position 656 in the 12S. Ovibos was grouped with Naemorhedus (BPML D 100; PPB D 1; BPMP D 100), and this clade is diagnosed by numerous molecular signatures in the mitochondrial markers: one in 12S (position 665: R ! T), two in Cyb (positions 429: A ! T, and 867: A ! C), and six in CO2 (positions 114: A ! T, 127: T ! G, 157: A ! G, 201: Y ! A, 224: T ! C, and 479: R ! C). The genera Ovis and Rupicapra were, respectively, found monophyletic with high support values (BPML D 94; PPB D 1; BPMP D 100), and numerous exclusive synapomorphies (5 and 12, respectively). By contrast, the genus Hemitragus was found polyphyletic: H. hylocrius is linked to the genus Ovis (BPML D 94; PPB D 1; BPMP D 100); H. jayakari is united with the genus Ammotragus (BPML D 99; PPB D 1; BPMP D 98); and H. jemlahicus is allied with the genus Capra (BPML D 100; PPB D 1; BPMP D 98). The polyphyly of Hemitragus is supported by all the four diVerent markers, and by several molecular signatures: although H. jemlahicus and Capra do not share any exclusive synapomorphy, the clade including H. jemlahicus, Capra, and Pseudois (BPML D 97; PPB D 1; BPMP D 97) is diagnosed by a T nucleotide at position 1116 in Cyb; H. jayakari and Ammotragus share one transition G ! A in PRKCI (at position 195 in the sequence of O. aries AF165789); and the association of H. hylocrius with Ovis is supported by three exclusive synapomorphies, including one insertion of Y at position 90 in 12S, one transversion A ! C at position 876 in Cyb, and one transition T ! C at position 417 in PRKCI.

160

A. Ropiquet, A. Hassanin / Molecular Phylogenetics and Evolution 36 (2005) 154–168

Fig. 1. Maximum likelihood tree (¡ln D 20671.98) derived from the combined analysis of the four markers. The values in the black spot are the bootstrap percentages (BPML) obtained with the combination of the four markers. The other values are the BPML obtained with each of the four markers independently; from left to right: CO2, Cyb, 12S, and PRKCI. Dash indicates that the node was not retrieved with the marker, but no alternative hypothesis was supported by BPML greater than 50. Asterisk indicates that an alternative hypothesis supported by a BPML greater than 50 was obtained with the marker.

The paraphyly of the genus Capra was suggested, as H. jemlahicus was the sister-group of C. sibirica in all combined analyses (BPML D 49; PPB D 0.97; BPMP D 62). However, the genus Capra was found monophyletic in all analyses of 12S (BPML D 61; PPB D 0.77; BPMP D 49), and in the MP analysis of CO2 (BPMP D 58). The Alpine ibex (C. ibex), Nubian ibex

(C. nubiana), markhor (C. falconeri), and domestic goat (C. hircus) were robustly linked (BPML D 100; PPB D 1; BPMP D 100). This grouping is upheld by two exclusive synapomorphies in Cyb (positions 954: A ! C, and 984: A ! G). The deletion of four nucleotides previously detected in the PRKCI sequence of C. hircus (TYGA, at position 349

A. Ropiquet, A. Hassanin / Molecular Phylogenetics and Evolution 36 (2005) 154–168

161

Fig. 2. Bayesian tree derived from the combined analysis of the four markers. The values in the black spot are the posterior probabilities (PPB) obtained with the combination of the four markers. The other values are the PPB obtained with each of the four markers independently; from left to right: CO2, Cyb, 12S, and PRKCI. Dash indicates that the node was not retrieved with the marker, but no alternative hypothesis was supported by PPB greater than 0.95.

in the sequence of O. aries AF165789; Matthee et al., 2001) was retrieved not only in all species of the genus Capra, but also in H. jemlahicus, Pseudois, Ammotragus, and H. jayakari. By contrast, this deletion was not found in H. hylocrius and all other species. The clade diagnosed by the deletion of TYGA, i.e., including Capra, H. jemlahicus, Pseudois,

Ammotragus, and H. jayakari, was found in the combined analyses (BPML D 67; PPB D 0.93; BPMP D 82), in the ML analysis of Cyb (BPML D 23), and in all analyses of PRKCI (BPML D 52; PPB D 0.99; BPMP D 78). In addition, this clade is diagnosed by another molecular signature in PRKCI, corresponding to a transition A ! G at position 336.

162

A. Ropiquet, A. Hassanin / Molecular Phylogenetics and Evolution 36 (2005) 154–168

Fig. 3. Most parsimonious tree of 878,450 steps derived from the combined analysis of the four markers by using a diVerential weighting scheme based on the product of homoplasy and saturation estimators (see Section 2). The values in the black spot are the bootstrap percentages (BPMP) greater than 50 obtained with the combination of the four markers. The other values are the BPMP obtained with each of the four markers independently; from left to right: CO2, Cyb, 12S, and PRKCI. Dash indicates that the node was not retrieved with the marker, but no alternative hypothesis was supported by BPMP greater than 50. Asterisk indicates that an alternative hypothesis supported by a BPMP greater than 50 was obtained with the marker.

3.2. Divergence times Molecular divergence times estimated using Mutlidivtime and PAML methods are indicated in Table 2. The ages estimated with the two methods were similar, but those obtained with PAML were systematically older than those inferred with Multidivtime. These diVerences may be explained by the use of diVerent priors, including

the assignment of branches to rate classes with PAML, the prior for rates and for times with the Bayesian approach, and the use of diVerent models of evolution, i.e., F84 for Multidivtime, and GTR for PAML. By using the F84 model with PAML, we obtained ages slightly more recent than those obtained with the GTR model, but they were always older than those found with Multidivtime (data not shown). In addition, the use of diVer-

A. Ropiquet, A. Hassanin / Molecular Phylogenetics and Evolution 36 (2005) 154–168

ent rate classes under PAML did not produce younger date estimates (data not shown). This suggests therefore that the diVerences in age estimates between Multidivtime and PAML are mainly explained by the priors for rates and for times in Multidivtime. By considering the maximum range of age estimates deduced from the two analyses (Table 2), we found that the subfamilies Antilopinae and Bovinae diversiWed during the middle Miocene, at around 14.6–17.8 and 12.7– 17.7 Mya, respectively. The tribe Caprini s.l. occurred in the Late Miocene (8.7–11.9 Mya). Within the tribe Caprini s.l., H. jayakari diverged from Ammotragus at around the Miocene/Pliocene boundary (4–7 Mya), H. hylocrius separated from Ovis during the Pliocene (2.7– 5.2 Mya), and H. jemlahicus diverged from Capra in the Pliocene (2.5–4.7 Mya). The clade composed of Pseudois, Capra and H. jemlahicus is dated at around 4.1–7.4 Mya (Late Miocene/Early Pliocene), and it diverged from the group including Ammotragus and H. jayakari, between 5 and 8.9 Mya (Late Miocene). The genera Capra, Ovis, and Rupicapra diversiWed during the Plio-Pleistocene.

4. Discussion 4.1. Is the tribe Caprini sensu Simpson (1945) monophyletic or not? According to the classiWcation of Simpson (1945), four diVerent tribes of caprines can be deWned: (1) the tribe Caprini, including Capra, Ammotragus, Hemitragus, Ovis, and Pseudois; (2) the tribe Rupicaprini, grouping Rupicapra, Naemorhedus, and Oreamnos; (3) the tribe Ovibovini, containing Ovibos and Budorcas; and (4) the tribe Saigini, composed of Saiga and Pantholops. This classiWcation was used as a reference by many authors, but molecular investigations based on the Cyb gene have questioned its validity, as none of the four tribes were found monophyletic (Groves and Shields, 1996; Hassanin and Douzery, 1999a; Hassanin et al., 1998a). For this reason, Hassanin and Douzery (1999a) have proposed to incorporate all caprine species into a single tribe, named Caprini sensu lato. However, our present data do not agree with these previous studies concerning the association of Budorcas with Ovis. Although the position of Budorcas was not stable in our diVerent analyses, it was never found allied with Ovis (Fig. 1–3). In fact, Ropiquet and Hassanin (2005) have recently shown that Budorcas was erroneously grouped with Ovis in all preceding Cyb analyses, because all sequences of Budorcas produced by Groves and Shields (1996) were largely contaminated by DNA of Ovis. By using new sequences for Budorcas, it appears however that the tribes Ovibovini, Rupicaprini, and Saigini remain poly- or paraphyletic, as Ovibos (Ovibovini) and Naemorhedus (Rupicaprini) are robustly associated (Figs. 1–3), and as Saiga was not allied with

163

Pantholops, but with Antilopini (Ropiquet and Hassanin, 2005). In addition, the tribe Caprini sensu Simpson (1945) is not retrieved here since it is divided into two unrelated clades: the Wrst one includes Ovis and H. hylocrius, and the other one groups Ammotragus, Capra, H. jayakari, H. jemlahicus, and Pseudois (Figs. 1–3). Nevertheless, we cannot really conclude that this tribe is not monophyletic because basal relationships are not robust. Additional markers will be necessary to resolve this question. 4.2. Phylogenetic relationships between aoudad, bharals, goats, tahrs, and sheep Species of Capra (goats) can be distinguished from those of Ovis (sheep) by the presence of a strong body odor, a beard in males, a callus on the knee, and a Xat and rather long tail, naked underneath, and by the absence of preorbital glands, inguinal glands, and pedal glands in hind feet (Schaller, 1977). The species of Hemitragus (tahrs) bear a general resemblance to Capra, although they diVer in that the males have short-sized horns and lack the beard. For this reason, they have been formerly ranged in the genus Capra (e.g., Sclater, 1886), or considered as being the sister-genus of Capra (e.g., Corbet, 1978; Lydekker, 1913). By contrast, the aYnities of Ammotragus (aoudad) and Pseudois (bharals) with either Capra or Ovis have been widely discussed in the literature, as both exhibit a particular combination of goat-like and sheep-like characters. Because of their superWcial similarities to true sheep, aoudad and bharals were also named Barbary sheep and Blue sheep, respectively, and both have been formerly placed in the genus Ovis (for reviews see Gray and Simpson, 1980; Wang and HoVmann, 1987). However, more recent morphological studies have rather concluded that Ammotragus and Pseudois share close relationships with Capra (e.g., Götze, 1998; Schaller, 1977), and several authors have placed the aoudad in the genus Capra (e.g., Corbet, 1978; Van Gelder, 1977), whereas Corbet (1978) suggested that bharals “should perhaps be included” in the genus Capra. The analyses of proteins have also produced contradictory results: by analyzing high-sulphur hair proteins, Darskus and Gillespie (1971) concluded that Ammotragus and Pseudois are more similar to each other than to either Capra or Ovis; but, the comparisons of blood proteins between Capra, Ovis, and Ammotragus have rather suggested an association of Ovis with either Ammotragus or Capra (Hight and Nadler, 1976; Manwell and Baker, 1977; Schmitt, 1963). By comparing the Giemsa-band patterns of the chromosomes, Bunch and Nadler (1980) have proposed very diVerent relationships: Hemitragus was not found allied with Capra but associated with the clade uniting Ammotragus and Ovis. The impacts of both protein and cytogenetic results were however limited in

164

A. Ropiquet, A. Hassanin / Molecular Phylogenetics and Evolution 36 (2005) 154–168

scope for three major reasons: (1) the comparisons were performed by using only three to Wve genera of Caprini sensu lato (Ammotragus, Capra, Hemitragus, Ovis, and Pseudois), whereas 11 genera have been described for this tribe (Hassanin and Douzery, 1999a); (2) no outgroup taxa were included for rooting the tree(s); and (3) no explicit methodology was applied for phylogenetic inferences. By including all genera and most species described into the tribe Caprini sensu lato, and by analyzing four diVerent DNA markers, we obtained robust and reliable phylogenetic inferences. Our analyses conWrm three results previously found by analyzing either partial sequences of 12S and/or 16S rRNA genes (Gatesy et al., 1997; Ludwig and Fischer, 1998), or complete Cyb genes (Groves and Shields, 1996; Hassanin et al., 1998a): (1) the genus Ovis is monophyletic, as O. ammon, O. aries, and O. dalli are enclosed together; (2) Capra and H. jemlahicus are closely related; and (3) they share a sistergroup relationship with Pseudois. In addition, the paraphyly of Capra, which was suggested by analyzing the complete Cyb gene (Hassanin et al., 1998a), is here rediscovered by combining all four markers, as C. sibirica is grouped with H. jemlahicus rather than with other species of Capra (Figs. 1–3). But, this result is not robust (BPML D 49; PPB D 0.97; BPMP D 62), and our independent analyses reveal, Wrst, that the paraphyly of Capra is only supported by the Cyb gene, and, second, that Capra is found monophyletic with the 12S gene. In view of these conXicting results, sequences of further markers will be necessary to conclude on the systematic status of the genus Capra. All previous molecular investigations concluded to strong aYnities between Hemitragus and Capra, as the Himalayan tahr was the only species of Hemitragus included in the analyses (Gatesy et al., 1997; Groves and Shields, 1996; Hassanin and Douzery, 1999a; Hassanin et al., 1998a). By sequencing all the three species of Hemitragus, our study shows that the genus Hemitragus is polyphyletic, since the two other species of tahr are not grouped with H. jemlahicus: the Arabian tahr (H. jayakari) is allied with Ammotragus, and the Nilgiri tahr (H. hylocrius) is related to Ovis. At Wrst sight, the polyphyly of Hemitragus is very surprising, as its monophyly was never called in question in the literature. According to Lydekker (1907), the genus Hemitragus was described as follows: “the short-horned goats, as the various species of tahr may be termed, are distinguished from other goats by the absence of the beard in the bucks, and the comparative shortness of the horns, which are placed close together at the base, and do not greatly exceed the length of the head. A further distinctive feature is that the horns of females are but little smaller than those of males.” Since the beard is a synapomorphic feature of Capra, its absence in tahrs cannot be used for deWning Hemitragus. The shortness of the horns is therefore the

unique morphological character for distinguishing Hemitragus from Capra. Obviously, the usefulness of this character is expected to be doubtful for taxonomy, as the horn length has considerably varied during the evolution of bovids (Lundrigan, 1996). For instance, the diVerent species of Capra are mainly separated by diVerences in males concerning both length and shape of the horns. In this context, the polyphyly of Hemitragus is not really surprising, and it is all the more plausible as several morphological, cytogenetic, and biogeographic data argue also in its favor. First, the association of H. hylocrius with Ovis is supported by the fact that they share very similar karyotypes (Benirschke and Kumamoto, 1980). Although the number of chromosomes is variable between the diVerent species of Ovis (2n D 52– 58), Bunch et al. (2000) have inferred that their common ancestor had 58 chromosomes, i.e., exactly as in H. hylocrius. In addition, the horns in males of H. hylocrius and Ovis do not have a prominent keel in front, and are marked by deep transverse wrinkles, contrasting with the horns of the two other species of tahr. Second, the association of H. jemlahicus with C. sibirica is upheld by the fact that both are cliV dwellers living in adjacent geographic areas (Schaller, 1977). Third, the grouping of H. jayakari with Ammotragus is supported by the fact that they have in common a very similar pelage characterized by a sandy or reddish brown color, ruVs of hair on the upper forelegs and angular tufts on jaws. In addition, they have the same karyotype with 58 chromosomes (Benirschke and Kumamoto, 1982). At least, their extant geographic distributions are only separated by the Red Sea, as H. jayakari occupies the mountains of South-east Arabia, while Ammotragus is found in North Africa. Interestingly, our molecular estimations indicate that their common ancestor occurred between 4 and 7 Mya (Table 2). Since the initiation of sea Xoor spreading in the Red Sea, began at about 4–5 Mya (Ghebreab, 1998), this suggests that the Red Sea may have acted as an eYcient barrier to gene Xow between the two ancestral populations of the Arabian tahr and aoudad. 4.3. Evolution of sexual dimorphism in caprines In caprines, sexual dimorphism can be characterized by diVerences in body size and mass, in the development of horns, and in pelage color and length. DiVerences between males and females are clearly more pronounced in the two following clades: (1) the “goat” clade, which includes Capra, H. jemlahicus, Pseudois, Ammotragus, and H. jayakari; and (2) the “sheep” clade, which incorporates Ovis and H. hylocrius. These species are polygynous, i.e., males mate with multiple females, and the male–male competition resulting from this sexual behavior is though to have driven the evolution of malebiased sexual size dimorphism (Clutton-Brock, 1989; Pérez-Barberia et al., 2002). Indeed, the reproductive

A. Ropiquet, A. Hassanin / Molecular Phylogenetics and Evolution 36 (2005) 154–168

success of males is directly correlated with strength and weapon size because males Wght for gaining access to females in estrus (Clutton-Brock, 1989). In Capra, Ammotragus, Ovis, and Pseudois, horns of males are particularly developed in length and their shape is radically diVerent from those of females, and from those of males of other species. Inter-speciWc diVerences in morphology of horns may have occurred because of the evolution of divergent Wghting styles (Caro et al., 2003; Geist, 1966; Lundrigan, 1996). In addition, horns may also function as display organs for courtship (Andersson, 1994), suggesting that the diversity of horn shapes may have also resulted from variations in female mate preferences. This hypothesis is here supported by analyzing the evolution of horns and pelage in males of the “goat” clade. In contrast with goats and bharals, males of the Himalayan tahr possess short horns, which are similar in length to those of females, implying that they do not use their horns as display organs for courtship. Since H. jemlahicus is closely related to Capra, with Pseudois at the outside (Figs. 1–3), it can be inferred that horns of males were long in their common ancestor, and that they reduced in the lineage leading to H. jemlahicus. Interestingly, it seems that the reduction in horn size was balanced by the emergence of rich pelage adornments. In fact, males of the Himalayan tahr are coppery brown to black in color, and possess a huge straw-colored ruV (with hair up to 30 cm long) that covers the forequarters and a long mantle of hairs that drapes from the sides and rump (Schaller, 1977). Since males display their most splendid adornments during the rut, it seems clear that they serve for attracting females. In agreement with this view, Schaller (1977) pointed out that courtship behavior is remarkably similar in goats and sheep, but the Himalayan tahr deviates in that males tend to orient displays to the front of the female rather than to the rear and they have incorporated several new patterns into their repertoire. Similarly, horn size and pelage may have also co-evolved during the history of the “sheep” clade. In H. hylocrius, males and females have similar horn length, but their pelage are very distinct: females have a uniform grayish color, while old males are almost black, except for their grizzled white back and sides and sometimes rump. Inversely, sexual pelage dimorphism is generally less marked in the species of Ovis, whereas horns of males are clearly diVerently shaped and more developed than those of females. In the Arabian tahr, males and females do not display great diVerences in horn length and body size. In addition, males of H. jayakari are smaller than those of other species of the “goat clade” (60 cm at the shoulder versus 75–100 cm; Harrisson and Bates, 1991). As the Arabian tahr lives in rocky, desert areas, a simple explanation is that males may have adapted by reducing in both body size and horn size because of low availability of food resources. Indeed, sexual size dimorphism is expected to

165

be lower under food-limited conditions, because larger males are more likely to die than females in populations (Toïgo and Gaillard, 2003). The pelage is sexually dimorphic in H. jayakari: males have very characteristic black markings on the face, a well-deWned blackish spinal crest, and their tufts of hairs on the upper forelegs and on the jaws are denser and longer than those of females. However, sexual dimorphism in pelage appears less accentuated than in the two other species of Hemitragus. This may be explained by the phenomenon of background matching (Stoner et al., 2003), in which natural selection by predators in a desert habitat would have favored the persistence of a sandy coat color in both males and females. Our molecular estimations of divergence times (Table 2) indicate that the Pliocene epoch was an intense period of diversiWcation for sexually dimorphic characters, as several taxa exhibiting great diVerences in horn size, horn shape, body size, and pelage diverged during this period: H. hylocrius and Ovis at around 2.7–5.2 Mya in Central Asia and/or India, H. jayakari and Ammotragus at around 4–7 Mya in North Africa and/or Arabia, H. jemlahicus and the various species of Capra at around 2.8–5.3 Mya probably in Central Asia. This period was the onset of Northern Hemisphere glaciations, which resulted in a global change toward cooler, drier, and more variable climates (Peizhen et al., 2001). Since food resources availability became more seasonal during this period, these important environmental changes may have modiWed and diversiWed the feeding behavior, and pattern of aggregation of caprines, allowing the emergence of new sexual attributes involved in mating behavior. 4.4. Taxonomic conclusions Since the genus Hemitragus was found to be polyphyletic in the present study, we propose a new taxonomy where the three species of tahrs are ranged in three diVerent monospeciWc genera. HEMITRAGUS Hodgson, 1841. Calcutta J. Nat. Hist. 2, 218. TYPE SPECIES: Capra jahral Hodgson, 1833 ( D Capra jemlahica H. Smith, 1826) Hemitragus jemlahicus (H. Smith, 1826). In GriYth et al. Animal Kingdom, vol. 4, plate [1826] opp. p. 308 [1827]. TYPE LOCALITY: Nepal, Jemla Hills. DISTRIBUTION: southern Xanks of the Himalaya Mountains from northern India east to Bhutan, as far north as Tibet. COMMON NAME: Himalayan tahr ARABITRAGUS, g. nov. TYPE SPECIES: Hemitragus jayakari Thomas, 1894. Ann. Mag. Nat. Hist., Ser. 6, 13, 365.

166

A. Ropiquet, A. Hassanin / Molecular Phylogenetics and Evolution 36 (2005) 154–168

Arabitragus jayakari (Thomas, 1894) Ann. Mag. Nat. Hist., Ser. 6, 13: 365. TYPE LOCALITY: Oman, Jebel Akhdar Range, Jebel Taw. DISTRIBUTION: mountains of south-east Arabia in Oman and United Arab Emirates. COMMON NAME: Arabian tahr. NILGIRITRAGUS, g. nov. TYPE SPECIES: Kemas hylocrius (Ogilby, 1838). Proc. Zool. Soc. Lond. 1837: 81 [1838]. Nilgiritragus hylocrius (Ogilby, 1838). Proc. Zool. Soc. Lond. 1837: 81 [1838]. TYPE LOCALITY: India, Nilgiri Hills. DISTRIBUTION: southern parts of a range of hills known as the Western Ghats in the Indian states of Kerala and Tamil Nadu. COMMON NAME: Nilgiri tahr.

Acknowledgments We thank Marie-Catherine Boisselier, Céline Bonillo, Josie Lambourdière, Eric Pasquet, and Annie Tillier for laboratory facilities. We are grateful to Jacques Cuisin, Francis Renoult, Daniel Robineau, Michel Tranier, and Géraldine Véron, who provided bone fragments from specimens of the MNHN collections. We thank Céline Canler and Vitaly Volobouev for frozen cells, Jean-Luc Berthier, Jacques Rigoulet, Claire Rejaud, Gérard Dousseau, and Jean-François Marjarie for blood samples from specimens of the Ménagerie du Jardin des Plantes, Jean-Claude Thibault for O. aries blood samples, Perry S. Barboza, Kevin Budsberg and Michel Perreau for tissues of O. moschatus, Françoise Hergueta-Claro for cells from H. niger and D. pygargus, Bertrand des Clers for bone sample from Ovis ammon, Bruno GuVond for muscles from R. pyrenaica, and Sir Bani Yas, Jacob Mwanzia, Stéphane Ostrowski for skin from the Arabian tahr. We also acknowledge Marlys L. Houck for help with the bibliography, and two anonymous reviewers for their helpful comments on the Wrst version of the manuscript. References Allard, M.W., Miyamoto, M.M., Jarecki, L., Kraus, F., Tennant, M.R., 1992. DNA systematics and evolution of the artiodactyl family Bovidae. Proc. Natl. Acad. Sci. USA 89, 3972–3976. Anderson, S., de Bruijn, M.H., Coulson, A.R., Eperon, I.C., Sanger, F., Young, I.G., 1982. Complete sequence of bovine mitochondrial DNA. Conserved features of the mammalian mitochondrial genome. J. Mol. Biol. 156, 683–717. Andersson, M., 1994. Sexual Selection. Pricceton University Press, Pricceton, NJ. Benirschke, K., Kumamoto, A.T., 1980. The chromosomes of the Nilgiri tahr. Int. Zoo yearbook 20, 274–275. Benirschke, K., Kumamoto, A.T., 1982. The chromosomes of the Arabian tahr. Mamm. Chromosomes Newsletter 23, 67–68.

Blyth, E., 1859. J. Asiat. Soc. Bengal 28, 291. Bougler, J., Denis, B., Lauvergne, J.J., 1988. Le scénario du peuplement ovin français ancien. Les colloques de l’INRA 47, 277–287. Bunch, T.D., Nadler, C.F., 1980. Giemsa-band patterns of the tahr and chromosomal evolution of the tribe Caprini. J. Hered. 71, 110–116. Bunch, T.D., Wang, S., Valdez, R., HoVmann, R.S., Zhang, Y., Liu, A., Lin, S., 2000. Cytogenetics, morphology and evolution of four subspecies of the Giant Sheep argali (Ovis ammon) of Asia. Mammalia 64, 199–207. Clutton-Brock, T.H., 1989. Mammalian mating systems. Proc. R. Soc. Lond. B Biol. Sci. 236, 339–372. Caro, T.M., Graham, C.M., Stoner, C.J., Flores, M.M., 2003. Correlates of horn and antler shape in bovids and cervids. Behav. Ecol. Sociobiol. 55, 32–41. Corbet, G.B., 1978. The Mammals of the Palaearctic Region. A Taxonomic Review. Trustees of the British Museum (N.H.). The Chaucer Press, Bungay, SuVolk. Darskus, R.L., Gillespie, J.M., 1971. Breed and species diVerences in the hair proteins of four genera of Caprini. Aust. J. Biol. Sci. 24, 515–524. De Donato, M., Gallagher Jr., D.S., Davis, S.K., Stelly, D.M., Taylor, J.F., 2001. The assignment of PRKCI to bovine chromosome 1q34 ! q36 by FISH suggests a new assignment to human chromosome 3. Cytogenet. Cell Genet. 95 (1-2), 79–81. Feligini, M., Parma, P., 2003. The complete nucleotide sequence of goat (Capra hircus) mitochondrial genome. DNA Seq. 14 (3), 199–203. Felsenstein, J., 2004. PHYLIP Version 3.6b, Phylogeny Inference Package Department of Genetics, University of Washington. Franceschi, P., Vallerand, F., 1988. ProWls génétiques visibles du mouton corse. Les colloques de l’INRA. 47, 187–194. Gatesy, J., Amato, G., Vrba, E., Schaller, G., DeSalle, R., 1997. A cladistic analysis of mitochondrial ribosomal DNA from the Bovidae. Mol. Phylogenet. Evol. 7, 303–319. Geist, V., 1966. The evolution of horn-like organs. Behaviour 27, 175–214. Ghebreab, W., 1998. Tectonics of the Red Sea region reassessed. EarthScience Reviews 45, 1–44. Götze, G., 1998. Die Osteologie des Blauschafes, Pseudois nayaur (Hodgson, 1833). Documenta Naturae (Eds.). München. No. 119. Gray, G.G., Simpson, C.D., 1980. Ammotragus lervia. Mamm. Species 144, 1–7. Gray, J.E., 1847. Cat. Osteol. Brit. Mus. p. 60. Groves, P., Shields, G., 1996. Phylogenetics of the Caprinae based on cytochrome b sequence. Mol. Phylogenet. Evol. 5, 467–476. Grubb, P., 1993. Family Bovidae. In: Wilson, D.E., Reeder, D.M. (Eds.), Mammal Species of the World. A Taxonomic and Geographic Reference. Smithsonian Press, Washington, pp. 393–414. Guindon, S., Gascuel, O., 2003. A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst. Biol. 52, 696–704. Haltenorth, T., 1963. KlassiWkation des Säugetiere: Artiodactyla. Handb. D. Zoologie, Berlin 8, Lfg. 32, 1–167. Harrisson, D.L., Bates, P.J.J., 1991. The Mammals of Arabia. Publication, Harrisson Zoological Museum (Ed.). Hassanin, A., Douzery, E., 1999a. The tribal radiation of the family Bovidae (Artiodactyla) and the evolution of the mitochondrial cytochrome b gene. Mol. Phylogenet. Evol. 13, 227–243. Hassanin, A., Douzery, E., 1999b. Evolutionary aYnities of the enigmatic Saola (Pseudoryx nghetinhensis), in the context of the molecular phylogeny of Bovidae. Proc. R. Soc. Lond. B Biol. Sci. 266, 893–900. Hassanin, A., Douzery, E., 2000. Is the newly described Vietnamese bovid Pseudonovibos spiralis a chamois (genus Rupicapra)?. Naturwissenschaften. 87, 122–124. Hassanin, A., Lecointre, G., Tillier, S., 1998b. The “evolutionary signal” of homoplasy in protein-coding gene sequences and its consequences for a priori weighting in phylogeny. C R Acad. Sci. Sér. III 321, 611–620.

A. Ropiquet, A. Hassanin / Molecular Phylogenetics and Evolution 36 (2005) 154–168 Hassanin, A., Pasquet, E., Vigne, J.-D., 1998a. Molecular systematics of the subfamily Caprinae (Artiodactyla, Bovidae) as determined from cytochrome b sequences. J. Mamm. Evol. 5, 217–236. Hassanin, A., Ropiquet, A., 2004. Molecular phylogeny of the tribe Bovini (Bovidae, Bovinae) and the taxonomic status of the Kouprey, Bos sauveli Urbain 1937. Mol. Phylogenet. Evol. 33, 896– 907. Hiendleder, S., Lewalski, H., Wassmuth, R., Janke, A., 1998. The complete mitochondrial DNA sequence of the Domestic Sheep (Ovis aries) and comparison with the other major Ovine haplotype. J. Mol. Evol. 47, 441–448. Hight, M.E., Nadler, C.F., 1976. Relationships between wild sheep and goats and the aoudad (Caprini) studied by immunodiVusion. Comp. Biochem. Physiol. 54B, 265–269. Hodgson, 1833. Asiatic Researches. 18, 129. Hodgson, 1841. Catalogue of Mammals of Nepal. Calcutta J. Nat. Hist. 2, 218. Honeycutt, R.L., Nedbal, M.A., Adkins, R.M., Janecek, L.L., 1995. Mammalian mitochondrial DNA evolution: a comparison of the cytochrome b and cytochrome c oxidase II genes. J. Mol. Evol. 40, 260–272. Huelsenbeck, J.P., Larget, B., Miller, R.E., Ronquist, F., 2002. Potential applications and pitfalls of Bayesian inference of phylogeny. Syst. Biol. 51, 673–688. Huelsenbeck, J.P., Ronquist, F., 2001. MrBayes: Bayesian inference of phylogenetic trees. Bioinformatics 17, 754–755. Janecek, L.L., Honeycutt, R.L., Adkins, R.M., Davis, S.K., 1996. Mitochondrial gene sequences and the molecular systematics of the artiodactyl subfamily Bovinae. Mol. Phylogenet. Evol. 6, 107– 119. Kingdon, J., 1982. East African Mammals. An Atlas of Evolution in Africa, Vol III, Part D (Bovids). Academic Press, London. Kingdon, J., 1997. The Kingdon Field Guide to African Mammals. Academic Press, London. Kishino, H., Thorne, J.L., Bruno, W.J., 2001. Performance of a divergence time estimation method under a Probabilistic model of rate evolution. Mol. Biol. Evol. 18, 352–361. Kraus, F., Miyamoto, M.M., 1990. Rapid cladogenesis among the pecoran ruminants: evidence from mitochondrial DNA sequences. Syst. Zool. 40, 117–130. Kuznetsova, M.V., Kholodova, M.V., 2002. Molecular support for the placement of Saiga and Procapra in Antilopinae (Artiodactyla, Bovidae). J. Mammal. Evol. 9, 271–280. Ludwig, A., Fischer, S., 1998. New aspects of an old discussion—phylogenetic relationships of Ammotragus and Pseudois within the subfamily Caprinae based on comparison of the 12S rDNA sequences. J. Zool. Syst. Evol. Res. 36, 173–178. Lundrigan, B., 1996. Morphology of horns and Wghting behavior in the family Bovidae. J. Mammal. 77, 462–475. Lydekker, R., 1907. The Game Animals of India, Burma, Malaya and Tibet. Rowland ward, London. Lydekker, R., 1913. Catalogue of the ungulate mammals in the British Museum (Natural History). Trustees of the British Museum (Nat. Hist.), London. Vol. I. Artiodactyla, family Bovidae, subfamilies Bovinae to Ovibovinae (cattle, sheep, goats, chamois, serows, takin, musk-oxen, etc.). Maijala, K., 1997. Genetic aspects of domestication, common breeds and their origin. In: Piper, L., Ruvinsky, A. (Eds.), The Genetics of Sheep. Cambridge University Press, Cambridge, pp. 14–47. Manwell, C., Baker, C.M., 1977. Ammotragus lervia: Barbary sheep or barbary goat?. Comp. Biochem. Physiol. 58B, 267–271. Matthee, C.A., BurzlaV, J.D., Taylor, J.F., Davis, S.K., 2001. Mining the mammalian genome for artiodactyl systematics. Syst. Biol. 50, 367– 390. Matthee, C.A., Davis, S.K., 2001. Molecular insights into the evolution of the family Bovidae: a nuclear DNA perspective. Mol. Biol. Evol. 18, 1220–1230.

167

Miyamoto, M.M., Allard, M.W., Adkins, R.M., Janecek, L.L., Honeycutt, R.L., 1994. A congruence test of reliability using linked mitochondrial DNA sequences. Syst. Biol. 43, 236–249. Nowak, R., 1999. Walker’s Mammals of the World, sixth ed. Johns Hopkins University Press, Baltimore, Md. Nylander, J.A.A., 2004. MrModeltest 2.0. Program distributed by the author. Evolutionary Biology Centre, Uppsala University. Ogilby, W., 1838. Exhibition of the skins of two species of the genus Kemas. Proc. Zool. Soc. Lond. 5, 81. Posada, D., Crandall, K.A., 1998. MODELTEST: testing the model of DNA substitution. Bioinformatics 14, 817–818. Peizhen, Z., Molnar, P., Downs, W.R., 2001. Increased sedimentation rates and grain sizes 2–4 Myr ago due to the inXuence of climate change on erosion rates. Nature 410, 891–897. Pérez-Barberia, F.J., Gordon, I.J., Pagel, M., 2002. The origins of sexual dimorphism in body size in Ungulates. Evolution 56, 1276–1285. Rambaut, A., 1996. Se-Al: Sequence Alignment Editor. Available from: . Randi, E., Mucci, N., Pierpaoli, M., Douzery, E., 1998. New phylogenetic perspectives on the Cervidae (Artiodactyla) are provided by the mitochondrial cytochrome b gene. Proc. R. Soc. Lond. B Biol. Sci. 265, 793–801. Ropiquet, A., Hassanin, A., 2005. Molecular phylogeny of caprines (Bovidae, Antilopinae): the question of their origin and diversiWcation during the Miocene. J. Zool. Syst. Evol. Res. 43, 49–60. Schaller, G.B., 1977. Mountain Monarchs: Wild Sheep and Goats of the Himalaya. University of Chicago Press, Chicago. Schmitt, J., 1963. Ammotragus lervia Pallas, Mähnenschaf oder Mähnenziege?. Z. Säugetierk 28, 7–12. Sclater, P.L., 1886. On the various species of wild goats. Proc. Zool. Soc. Lond., 314–318. Sclater, P.L., 1891. Cat. Mamm. Ind. Mus., 146. Shackleton, D.M., 1997. Wild Sheep and Goats and their Relatives. IUCN/SSC Caprinae Specialist Group. Status Survey and Conservation Action Plan for Caprinae. IUCN, Gland, Switzerland and Cambridge, UK. Shackleton, D.M., Lovari, S., 1997. ClassiWcation adopted for the Caprinae survey. In: Wild Sheep and Goats and Their Relatives. Status Survey and Conservation Action Plan for Caprinae. Shackleton DM Eds. pp. 9–16. IUCN. Simpson, G.G., 1945. The principles of classiWcation and a classiWcation of mammals. Bull. Am. Mus. Nat. Hist. 5, 1–350. Smith, H., 1826. In GriYth’s, Cuvier Anim. Kingd. 4, 308. Stoner, C.J., Caro, T.M., Graham, C.M., 2003. Ecological and behavioral correlates of coloration in artiodactyls: systematic analyses of conventional hypotheses. Behav. Ecol. 14, 823–840. SwoVord, D.L., 1993. PAUP: “Phylogenetic Analysis Using Parsimony, Version 3.1.1,” Computer program distributed by the Illinois Natural History Survey, Champaign, IL. Thomas, O., 1894. Ann. Mag. Nat. Hist. Ser. 6 13, 365. Thorne, J.L., Kishino, H., Painter, I.S., 1998. Estimating the rate of evolution of the rate of molecular evolution. Mol. Biol. Evol. 15, 1647– 1657. Toïgo, C., Gaillard, J.M., 2003. Causes of sex-biased adult survival in ungulates: sexual size dimorphism, mating tactic or environment harshness?. IKOS 101, 376–384. Van Gelder, R.G., 1977. Mammalian hybrids and genric limits. Am. Mus. Novitates 2635, 1–25. Vrba, E.S., Schaller, G.B., 2000. Phylogeny of Bovidae based on behavior, glands, skulls, and postcrania. In: Vrba, E.S., Schaller, G.B. (Eds.), Antelopes, Deer, and Relatives. Fossil Record, Behavioral Ecology, Systematics, and Conservation. Yale University Press, New Haven, CT, pp. 203–222. Wang, X., HoVmann, R.S., 1987. Pseudois nayaur and Pseudois schaeferi. Mamm. Species 278, 1–6. Winnepenninckx, B., Backeljau, T., Dewachter, R., 1993. Extraction of high molecular weight DNA from mollusc. Trends Genet. 9, 407.

168

A. Ropiquet, A. Hassanin / Molecular Phylogenetics and Evolution 36 (2005) 154–168

Wuyts, J., Van de Peer, Y., Winkelmans, T., De Wachter, R., 2002. The European database on small subunit ribosomal RNA. Nucleic Acids Res. 30, 183–185. Yang, Z., 1994a. Estimating the pattern of nucleotide substitution. J. Mol. Evol. 39, 105–111. Yang, Z., 1994b. Maximum likelihood phylogenetic estimation from DNA sequences with variable rates over sites: approximate methods. J. Mol. Evol. 39, 306–314.

Yang, Z., 2003. Phylogenetic Analysis by Maximum Likelihood (PAML) Version 3.14. Available from: . Yang, Z., Yoder, A.D., 2003. Comparison of likelihood and Bayesian methods for estimating divergence times using multiple gene loci and calibration points, with application to a radiation of cute-looking mouse lemur species. Syst. Biol. 52, 1–12.