Evolutionary History of the Odd-Nosed Monkeys and the Phylogenetic Position of the Newly Described Myanmar Snub-Nosed Monkey Rhinopithecus strykeri

Odd-nosed monkeys represent one of the two major groups of Asian colobines. Our knowledge about this primate group is still limited as it is highlighted by the recent discovery of a new species in Northern Myanmar. Although a common origin of the group is now widely accepted, the phylogenetic relationships among its genera and species, and the biogeographic processes leading to their current distribution are largely unknown. To address these issues, we have analyzed complete mitochondrial genomes and 12 nuclear loci, including one X chromosomal, six Y chromosomal and five autosomal loci, from all ten odd-nosed monkey species. The gene tree topologies and divergence age estimates derived from different markers were highly similar, but differed in placing various species or haplogroups within the genera Rhinopithecus and Pygathrix. Based on our data, Rhinopithecus represent the most basal lineage, and Nasalis and Simias form closely related sister taxa, suggesting a Northern origin of odd-nosed monkeys and a later invasion into Indochina and Sundaland. According to our divergence age estimates, the lineages leading to the genera Rhinopithecus, Pygathrix and Nasalis+Simias originated in the late Miocene, while differentiation events within these genera and also the split between Nasalis and Simias occurred in the Pleistocene. Observed gene tree discordances between mitochondrial and nuclear datasets, and paraphylies in the mitochondrial dataset for some species of the genera Rhinopithecus and Pygathrix suggest secondary gene flow after the taxa initially diverged. Most likely such events were triggered by dramatic changes in geology and climate within the region. Overall, our study provides the most comprehensive view on odd-nosed monkey evolution and emphasizes that data from differentially inherited markers are crucial to better understand evolutionary relationships and to trace secondary gene flow.


Introduction
The course of mammalian evolution in the Tertiary and Quaternary has been affected profoundly by changes in continental configuration, mountain building, and associated changes in patterns of oceanic and atmospheric circulation. Nowhere is this better illustrated than in eastern Eurasia during the late Tertiary (23.0-2.588 million years), where the raising of the Himalayas and adjacent Tibetan Plateau altered the biogeographic landscape by creating new physical and climatic barriers to gene flow. Environmental changes in the region of the Hengduan Mountains (including today's Three Parallel Rivers National Park and World Heritage Site, Figure 1) in South-western China had pronounced effects on biotas, creating opportunities for the preservation of paleoendemics and for the creation of neoendemics in sheltered, trenchant intermontane valleys [1,2]. Tectonic activities, changes in global temperature, and the inauguration of Pleistocene glacial and interglacial fluctuations led to further cycles of dissection and coalescence of landscape and biotas. Heightened environmental seasonality and changes in sea level affecting the inundation and exposure of low-lying continental shelf environments, such as large parts of South-East Asia (e.g., Sundaland), created dynamic landscapes, which witnessed both the extinction and origination of numerous mammalian species. The primate order was no exception with radiations in gibbons, macaques, langurs and oddnosed monkeys as well as extinctions, e.g. in the Pongidae (Gigantopithecus and Pongo on the Asian mainland) [3,4]. In this geographical and historical context odd-nosed monkeys are of particular interest, because they occupy a large latitudinal and altitudinal range including their presumed centre of origin (Hengduan Mountains, Figure 1) [5].
Odd-nosed monkeys are enigmatic and rare, and they show remarkable anatomical and behavioral adaptations to a range of habitats unusual for primates, such as mangrove swamps, and temperate and high altitude forests [5,6]. The group consists of four genera, snub-nosed monkeys (Rhinopithecus), douc langurs (Pygathrix), proboscis monkeys (Nasalis) and pig-tailed monkeys (Simias). Three of the five snub-nosed monkey species, R. roxellana, R. brelichi and R. bieti are endemic to China, while a fourth species, R. avunculus is restricted to the North of Vietnam. A fifth species, R. strykeri was recently described from Myanmar (Burma) [7], and has now also been confirmed in Nujiang prefecture, China, in contiguous forests with Myanmar. Douc langurs, represented by the three species P. nemaeus, P. cinerea and P. nigripes, are distributed through parts of Vietnam, Laos and Cambodia, east of the Mekong River. Both Nasalis and Simias are monotypic. N. larvatus is a pure Bornean species, whereas S. concolor is endemic to the Mentawai Islands, west off Sumatra. All species are endangered or critically endangered [7,8]. R. avunculus and R. strykeri are the most threatened species with populations sizes of ,250 and 260-330 individuals, respectively, followed by P. cinerea (550-700 individuals) and R. brelichi (,750 individuals). R. bieti has less than 2,000 individuals, and R. roxellana and S. concolor have ,15,000 and 6,700-17,300 individuals. For P. nemaeus, P. nigripes and N. larvatus no reliable estimates are available [7,8].
Within genera, a clearly resolved branching pattern was obtained for Pygathrix [17][18][19], but relationships among the Rhinopithecus species are still disputed. Jablonski and Peng [9] found that R. roxellana groups together with the R. bieti+R. brelichi clade and that R. avunculus forms a sister species to all other snubnosed monkeys. Molecular studies based only short fragments of the mitochondrial genome suggested an unresolved polytomy [18][19][20][21], while in a recent study by Li et al. [15] using complete mitochondrial genome data, R. avunculus appears as sister lineage to a R. bieti+R. roxellana clade. Unfortunately, all previous molecular studies lack a complete taxon sampling. In particular, R. strykeri was not studied so far, but it can be assumed that this species is closely related to R. bieti [7].
Although our understanding of the phylogenetic relationships within the odd-nosed monkey group became clearer in recent years, several questions remain. All molecular studies so far used mainly only short mitochondrial fragments and/or did not include all species. To obtain a more complete picture about the evolutionary history of the odd-nosed monkeys, we have analyzed complete mitochondrial genomes (mtDNA) and 12 nuclear loci (nucDNA, five autosomal, six Y chromosomal, one X chromosomal) from all ten odd-nosed monkey species. Moreover, we included data from the two known major haplogroups of R. bieti [20,22,23] and P. cinerea (Roos unpublished). As a result of our study, we present here the most complete and updated molecular phylogeny of odd-nosed monkeys and discuss their phylogeographic implications.

Mitochondrial phylogeny
The mtDNA alignment including a total of 24 primate sequences had a length of 16,920 bp (Table S1). After excluding indels and poorly aligned positions, 15,617 bp remained in the mtDNA1 dataset (for further details see Materials and Methods). Another alignment including only the 12 protein-coding genes on the heavy strand (mtDNA2) consisted of 10,851 bp. Phylogenetic relationships as obtained from maximum-likelihood (ML) and Bayesian reconstructions resulted in identical and mainly significantly supported branching patterns ( Figure 2, Figure 3A). Asian colobines initially split into the three lineages/clades Trachypithecu-s+Presbytis, Semnopithecus and odd-nosed monkeys. While reconstructions based on the mtDNA1 dataset allowed no resolution of this trichotomy, the mtDNA2 dataset suggested a sister grouping of Semnopithecus with odd-nosed monkeys. Among odd-nosed monkeys, Rhinopithecus separated 7.28 million years ago (Ma) (for 95% highest posterior probabilities see Table 1), before also Pygathrix diverged 6.63 Ma from the Nasalis+Simias clade. Within Rhinopithecus, R. avunculus formed a sister lineage to the remaining species, which further divided into a clade consisting of R. roxellana and R. brelichi, and a clade with R. bieti and R. strykeri. Interestingly, R. bieti appeared paraphyletic with haplogroup B being more closely related to R. strykeri than to the con-specific haplogroup A. Within Pygathrix, P. nigripes separated first and haplogroup B of P. cinerea formed a sister lineage to a clade consisting of the P. cinerea haplogroup A and P. nemaeus. Initial splits in Rhinopithecus and Pygathrix, and the divergence of Nasalis and Simias occurred 2.50-1.75 Ma. Haplogroup A of R. bieti separated from haplogroup B and R. strykeri 0.60 Ma, while the latter two diverged 0.24 Ma. Haplogroup B of P. cinerea split from haplogroup A and P. nemaeus 0.69 Ma, and the latter two separated 0.23 Ma. According to alternative tree topology tests, alternative phylogenetic relationships among odd-nosed monkey species as suggested by nucDNA ( Figure 3B) were rejected (P,0.001).

Nuclear phylogeny
For phylogenetic analysis, nuclear sequence data were combined, because single loci provided only limited resolution in an earlier study [14] and partition homogeneity tests revealed no significant difference in their evolutionary history (Y chromosomal loci combined: P = 0.945; autosomal loci combined: P = 0.066; all nuclear loci combined: P = 0.200). The concatenated nuclear alignment had an original length of 15,101 bp (Table S1), but was reduced to 13,102 bp after indels and poorly aligned positions were removed. Phylogenetic relationships among Asian colobines were mainly significantly supported ( Figure 3B), but partially disagreed with the mitochondrial topology. Notably, phylogenetic relationships as suggested by mtDNA ( Figure 3A) were rejected (P = 0.002) by alternative tree topology tests. Differences to the mitochondrial phylogeny include the basal position of Presbytis among Asian colobines, the sister grouping of Semnopithecus and Trachypithecus, and species/haplogroup relationships within Rhinopithecus and Pygathrix. In Rhinopithecus, an initial split occurred between the R. roxellana+R. brelichi clade and the remaining taxa, including R. avunculus. Among the latter, R. avunculus diverged first and R. strykeri is basal to a monophyletic R. bieti clade. Within Pygathrix, P. nigripes represented the first split and P. nemaeus formed a sister lineage to the monophyletic P. cinerea clade. Calculated divergence ages for odd-nosed monkey genera and species were similar to the mtDNA estimates (Table 1). Accordingly, Rhinopithecus, Pygathrix and the Nasalis+Simias clade separated 6.77-5.99 Ma, and the latter split 1.12 Ma. Initial differentiation of Rhinopithecus (R. roxellana+R. brelichi -other Rhinopithecus spp., R. roxellana2R. brelichi) and Pygathrix (P. nigripes -other Pygathrix spp.) occurred on a similar time scale (1.69-1.31 Ma). Further splitting events in both genera took place 0.73-0.40 Ma (R. avunculus2R. bieti+R. strykeri, R. bieti2R. strykeri) and 0.66 Ma (P. cinerea2P. nemaeus), respectively.

Discussion
As in earlier molecular studies including extended sequence data [11][12][13][14][15][16], the monophyly of the odd-nosed monkeys is strongly supported. Further, our data clearly suggests a basal position of Rhinopithecus and a sister grouping of Pygathrix with the Nasalis+Simias clade, a pattern also supported by two SINE integrations [14], and earlier nuclear [14,16] and mitochondrial genome data [14]. Moreover, our study provides information concerning the branching pattern within the two polytypic genera Rhinopithecus and Pygathrix. In all gene trees, R. roxellana and R. brelichi form a monophyletic clade and P. nigripes is basal within Pygathrix. For all other Rhinopithecus and Pygathrix taxa, incongruent tree topologies were obtained. In Rhinopithecus, R. avunculus appears to be basal in the mtDNA tree, but in the nucDNA tree, the species is connected with the R. bieti+R. strykeri clade. The mtDNA tree further suggests paraphyly of R. bieti, but monophyly in the nucDNA tree. Similarly, within Pygathrix, P. cinerea is paraphyletic in the mtDNA tree, but monophyletic in the nucDNA tree.
Various reasons for incongruent gene tree topologies are possible [24,25]. Hybridization (i.e., secondary gene flow) is one reason that becomes more and more accepted and recent studies show that hybridization occurs more frequently than previously thought in a broad range of species. Also for primates, hybridization was confirmed for almost all major radiations [26,27]. In many cases, phylogenetic incongruences are caused by sex-biased introgessive hybridization which can occur if a secondary contact between two populations arises, migration is either male-or female-biased, and hybridization is unidirectional [28][29][30].
For odd-nosed monkeys, secondary gene flow seems to be also the most likely explanation for the observed incongruent gene tree topologies, because geographically close-by taxa cluster together. However, low resolution of phylogenetic relationships among species due to limited informative sites and incomplete lineage sorting might have an effect as well, in particular for nucDNA.
Assuming secondary gene flow as main reason for the observed gene tree discordances among odd-nosed monkey species and by incorporating estimated divergence ages, R. avunculus diverged first in Rhinopithecus, but came into secondary contact with the progenitor of R. bieti and R. strykeri. The progenitor of the latter two originally split into three major lineages (R. bieti haplogroup A, R. bieti haplogroup B, R. strykeri), with the R. bieti haplogroup B being closely related to R. strykeri. Interestingly, the R. bieti haplogroup B occurs more in the southern part of the species' range [22] and thus geographically closest to R. strykeri. Both R. bieti haplogroups came into contact again and homogenized their nuclear gene pools [31]. Today, both mitochondrial haplogroups of R. bieti intergrade, although there is still a trend of a northern and southern clade [22]. In Pygathrix, originally four major lineages emerged (P. nigripes, P. cinerea haplogroup A, P. cinerea haplogroup B, P. nemaeus), but secondary exchange of nuclear genes occurred between both haplogroups of P. cinerea and of mitochondria between P. cinerea haplogroup A and P. nemaeus. Both mitochondrial haplogroups of P. cinerea are today still geographically separated (Roos unpublished).
By combining the available information on phylogeny and pastgeology and -climate in Asia, we develop the following dispersal scenario for odd-nosed monkeys ( Figure 4). The origin of Asian colobine monkeys and also of the odd-nosed monkeys might have been the region of South-western China and the Hengduan Mountains in the border region of today's Myanmar, India and China [5,14,32]. The ancestor of the Asian colobines, including the odd-nosed monkeys, was probably a species of Mesopithecus, the colobine genus which was widespread in woodland and forested environments of eastern and southern Europe and western and southern Asia during the latest Miocene and earliest Pliocene [5,33,34]. The recent discovery of fossil colobine remains belonging to Mesopithecus in North-eastern Yunnan Province, near Zhaotong, supports this interpretation [35]. In the region, all the larger Southeast Asian rivers (Mekong, Salween, Yangtze) rise, which are all dispersal barriers for arboreal primates [36] at least since the early Miocene [37]. Habitat fragmentation became more pronounced in this region in the late Miocene as the result of regional tectonic uplift and subsidence related to the Himalayan orogeny [38]. After the separation of the langur progenitor from the odd-nosed monkey ancestor, members of the odd-nosed monkey group successively migrated from China to the South and expanded their range into Indochina and Sundaland in the latest Miocene. The migration into Sundaland by a progenitor of Nasalis+Simias was probably via land bridges connecting the mainland with Sundaland islands during periods of lowered sea levels [39]. Finally, Nasalis on Borneo and Simias on the Mentawai Islands diverged in the early Pleistocene. Due to the relatively prominent dating discrepancy, gene flow between both genera after the initial separation might have occurred, especially considering that migration was repeatedly possible via land bridge connections during the Pleistocene [39]. Simias or at least a common ancestor of Simias and Nasalis was most likely temporarily also present on Sumatra, but became extinct there of unknown reasons. Interestingly, the Mentawai archipelago shows a high rate of faunal and floral endemism and might have acted as refuge, as e.g. proposed for macaques [40]. In light of the prominent morphological differences between Nasalis and Simias the relatively late divergence is surprising and suggests that autapomorphic traits can emerge in a short time period.
Differentiation in Rhinopithecus started in the early Pleistocene, most likely triggered by reduction and fragmentation of suitable habitat. Rhinopithecus originally occupied a wide distribution, which became fragmented as the result of reduction of forest habitats resulting from climate/geological changes and changes in the course of rivers. Environmental changes occurring in the early Pleistocene appear to have produced recurrent episodes of habitat fragmentation and coalescence, allowing secondary contact and gene exchange. R. avunculus split off first, followed by the split between the northern species (R. roxellana, R. brelichi) and ''Himalaya'' species (R. bieti, R. strykeri). Shortly afterwards, R. roxellana and R. brelichi diverged and remained separated from each other. R. avunculus came into secondary contact with the progenitor of R. bieti and R. strykeri until the early middle Pleistocene. Subsequently, the ancestral R. bieti+R. strykeri population divided into two (R. bieti haplogroup A, R. bieti haplogroup B+R. strykeri) and later on into three populations (R. bieti haplogroup A, R. bieti haplogroup B, R. strykeri). R. strykeri remained afterwards separated Odd-Nosed Monkey Phylogeny PLoS ONE | www.plosone.org from R bieti, while representatives of both R. bieti haplogroups came into secondary contact which lasts until today (as far as possible due to anthropogenic habitat fragmentation) [31]. The fossil record of Rhinopithecus broadly supports this interpretation. The genus enjoyed a wide distribution in central, eastern, and southern China in the Plio-Pleistocene [3,41,42], and even included a large-bodied extinct species, R. lantianensis, north of the Yellow River [43]. The extent of species differentiation in the Pleistocene is hard to determine from the fossil record, because most Rhinopithecus fossils are represented only by teeth and jaw fragments, and these exhibit only little diagnostic, species-specific morphological traits.
For Pygathrix, our genetic data suggest differentiation events on a similar time scale as in snub-nosed monkeys (see above), limestone langurs [16] and crested gibbons [44]. Thus, the same environmental changes as mentioned above might have influenced not only snub-nosed monkey differentiation, but also differentiation processes in douc langurs and other primates in the region. After the split of P. nigripes, the ancestor of P. nemaeus and P. cinerea successively diverged in the middle Pleistocene, leading first to the separation of the southern P. cinerea haplogroup (haplogroup B) and later on also to the divergence of the northern P. cinerea haplogroup (haplogroup A) and P. nemaeus. Subsequently, both populations of P. cinerea and P. nemaeus came into secondary contact and exchanged again genetic material.
By analyzing sequence data from complete mitochondrial genomes and 12 nuclear loci from all ten odd-nosed monkey species and major haplogroups, the present study provides new and comprehensive insight into the evolutionary and biogeographic history of this enigmatic primate group. Our study supports the hypothesis that odd-nosed monkeys originated on the Asian mainland and migrated into Indochina and Sundaland during the late Miocene. Differentiation into species started in the early Pleistocene, but secondary gene flow among various lineages occurred until the middle Pleistocene. The odd-nosed monkeys as a taxonomic group provide another example that our knowledge of primate diversity is still incomplete (two taxa have been newly described within the last 15 years) and that our understanding of diversification processes is limited. In particular, the impact of secondary gene flow and hybridization seems to be underestimated. The analysis of additional nuclear loci and a population genetic approach including more samples from each taxon will shed further light on the underlying speciation processes in this primate group.

Ethics statement
Our work was conducted according to relevant German and international guidelines, including countries where we obtained and analyzed samples. Fecal samples from captive animals held at the breeding station of Fanjingshan National Nature Reserve (FNNR), the Kunming Institute of Zoology (KIZ), Beijing Zoo and the Endangered Primate Rescue Center (EPRC), and permissions to use them for our study were provided by Yeqin Yang, director of FNNR, Yaping Zhang, president of KIZ, Jinyuan Zhang, director of Beijing Zoo and Tilo Nadler, director of the EPRC. Fecal samples were collected by institutional staff in a non-invasive way without disturbing, threatening or harming the animals during daily enclosure cleaning. Dried skin material from R. strykeri was obtained from a museum specimen described earlier [7] and collection adhered to the legal requirements of Myanmar. Collection of samples adhered to the American Society of Primatologists (ASP) Principles for the Ethical Treatment of Non Human Primates (see www.asp.org/society/resolutions/ EthicalTreatmentOfNonHumanPrimates.cfm).

Sample collection
We obtained fecal samples from captive but wild-born R. roxellana, R. brelichi, R. bieti, P. cinerea and P. nigripes from the breeding station of FNNR, KIZ and Beijing Zoo, all China, and the EPRC, Vietnam (Table S2). From R. bieti and P. cinerea, representatives of both major mitochondrial haplogroups (A, B) ( [20,22,23], Roos unpublished) were sampled. Fresh fecal samples were preserved in 90% ethanol and stored at room temperature before further processing. From R. strykeri, we obtained a dried skin fragment from the holotype (AIMZ 15504) housed at the Anthropological Institute and Museum of the University of Zurich (AIMZ), Switzerland [7]. This specimen was collected on March 7, 2010 from two hunters from Pade village (Kachin State, Myanmar; 26.424861uN, 98.312371uE) who had caught the monkey one or two days before [7]. Orthologous sequences of the remaining species (R. avunculus, P. nemaeus, N. larvatus, S. concolor) were recently published by our group [14].

Laboratory methods
DNA from fecal and tissue material was extracted using the QIAamp DNA Stool Mini and DNeasy Blood & Tissue kits from Qiagen following recommendations of the supplier. Complete mitochondrial genomes and nuclear loci were amplified and sequenced using methods outlined in Roos et al. [14]. Due to degradation of DNA extracted from the fecal and dried skin material, mitochondrial genomes were amplified via five overlapping fragments each with a size of ,5,000 bp. Amplification of nuclear loci was conducted in fragments of less than 1,000 bp [14]. Nuclear markers included five autosomal loci (intron 3 of the serum albumin gene, ALB3; intron 3 of the interstitial retinol-binding protein, IRBP3; intron 1 of the transition protein 2, TNP2; intron 1 of the transthyretin gene, TTR1; intron 11 of the von Willebrand Factor, VWF11), a fragment of the X chromosomal Xq13.3 region and six Y chromosomal loci (intron 5 of the Dead Box gene, DBY5; introns 7 and 11 of the SMC mouse homologue, SMCY7, SMCY11; the SRY gene, SRY; intron 18 of the ubiquitous motif gene, UTY18; last intron of the Zinc finger gene, ZFYLI). Sequences were assembled with GeneiousPro 5.4 [45] and mitochondrial genomes were annotated with the online program DOGMA [46]. Sequences were deposited at GenBank (for accession numbers see Table S2).
For phylogenetic analysis, nuclear sequence data were combined, because single loci provided only limited resolution in an earlier study [14] and partition homogeneity tests in PAUP 4.0b10 [50] with 1,000 replications revealed no significant difference in their evolutionary history (see Results). In contrast, mitochondrial and nuclear sequence data were not combined, because partition homogeneity tests suggested that both datasets track different evolutionary histories (P = 0.001, P = 0.002) and because of known incongruent phylogenetic positions of the three langur genera Semnopithecus, Trachypithecus and Presbytis among Asian colobines in mitochondrial and nuclear phylogenies [14].
Phylogenetic trees were constructed with ML and Bayesian algorithms, using the programs GARLI 2.0 [51] and MrBayes 3.1.2 [52,53]. For all reconstructions, the optimal nucleotide substitution model for each locus was chosen using the Bayesian information criterion (BIC) as implemented in jModeltest 0.1 [54] ( Table S1). For phylogenetic analysis, the datasets were whenever appropriate partitioned treating each locus separately and each with its own substitution model. The solely protein-coding alignment of the mitochondrial genome (mtDNA2) was further partitioned into codon positions. In GARLI, only the model specification settings were adjusted, while all other settings were left at their default value. Relative support of internal nodes was assessed by bootstrap analyses with 1,000 replications and ML majority-rule consensus trees were calculated in PAUP. For Bayesian analyses, we used four independent Markov Chain Monte Carlo (MCMC) runs with the default temperature of 0.2. Four repetitions were run for 10 million generations with tree and parameter sampling occurring every 100 generations. Acceptance rates were in the optimal range of 10-70%. The first 25% of samples were discarded as burn-in, leaving 75,001 trees per run. The adequacy of this burn-in and convergence of all parameters was assessed by examining the uncorrected potential scale reduction factor (PSRF) [55] as calculated by MrBayes, which should approach 1 as runs converge and by visual inspection of the trace of the parameters across generations using the software TRACER 1.5 [56]. AWTY [57] was used to check whether posterior clade probabilities were also converging. Posterior probabilities for each split and a phylogram with mean branch lengths were calculated from the posterior density of trees. Alternative phylogenetic relationships among odd-nosed monkey species were tested with the Kishino-Hasegawa test [58] with full optimization and 1,000 bootstrap replications in PAUP.
To estimate divergence ages from the nuclear and mtDNA2 datasets, we applied a Bayesian MCMC method, which employs a relaxed molecular clock approach [59] as implemented in BEAST 1.6.1 [60]. Therefore, we assumed a relaxed lognormal model of lineage variation and a Birth-Death Process prior for branching rates.
As calibrations we used the fossil-based divergence between Homo and Pan, which has been dated at 6-7 Ma [61][62][63], the separation of Pongo from the Homo+Pan lineage ,14 Ma [64], the split between Theropithecus and Papio ,4 Ma [65,66], the split between M. sylvanus and M. mulatta ,5.5 Ma [67] and the divergence of hominoids and cercopithecoids 24-29 Ma [68,69]. Instead of hardbounded calibration points, we used the published dates as a normal distribution prior for the respective node. For the Homo2Pan divergence, this translates into a normal distribution with a mean of 6.5 Ma and a standard deviation (SD) of 0.5 Ma, for the separation of Pongo from the Homo+Pan clade into a mean of 14.0 Ma and a SD of 1.0 Ma, for the Theropithecus-Papio split into a mean of 4.0 Ma and a SD of 0.5 Ma, for the split between M. sylvanus and M. mulatta into a mean of 5.5 Ma and a SD of 0.5 Ma, and for the hominoid -cercopithecoid divergence into a mean of 26.5 Ma and a SD of 2.5 Ma.
In BEAST, four replicates were run for 10 million generations with tree and parameter sampling occurring every 100 generations. The adequacy of a 10% burn-in and convergence of all parameters were assessed by visual inspection of the trace of the parameters across generations using TRACER. Subsequently, the sampling distributions were combined (25% burn-in) using the software LogCombiner 1.6.1 and a consensus chronogram with node height distribution was generated and visualized with TreeAnnotator 1.6.1 and FigTree 1.3.1 [70].

Supporting Information
Table S1 Locus-specific information including alignment length, number of variable and parsimony-informative sites, and selected substitution models. (DOC)