Taxonomic Revision and Evolutionary Phylogeography of Dusky Langur (Trachypithecus obscurus) in Peninsular Malaysia

© 2020 Academia Sinica, Taiwan Open Access Taxonomic Revision and Evolutionary Phylogeography of Dusky Langur (Trachypithecus obscurus) in Peninsular Malaysia Nor Rahman Aifat1, Muhammad Abu Bakar Abdul-Latiff1,2, Christian Roos3, and Badrul Munir MdZain1,* 1Department of Biological Sciences and Biotechnology, Faculty of Science and Technology, Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia. *Correspondence: E-mail: [email protected] (Md-Zain) E-mail: [email protected] (Aifat) 2Centre of Research for Sustainable Uses of Natural Resources (CoR-SUNR), Faculty of Applied Sciences and Technology, Universiti Tun Hussein Onn Malaysia, Pagoh Campus, 84000 Muar, Johor, Malaysia. E-mail: [email protected] (Abdul-Latiff) 3Primate Genetics and Gene Bank of Primates, German Primate Center, Kellnerweg 4, 37077 Goettingen, Germany. E-mail: [email protected] (Roos) Received 29 May 2020 / Accepted 4 October 2020 / Published 30 November 2020 Communicated by Benny K.K. Chan Dusky langur, Trachypithecus obscurus, inhabits tropical rainforests in Peninsular Malaysia, Thailand, and Myanmar. Morphologically, five subspecies are distributed in Peninsular Malaysia, but few studies have used genetic data to verify the classification. It is difficult to differentiate subspecies based on morphological characteristics, so this study used molecular data to differentiate subspecies of T. obscurus. The issue was addressed by analyzing 723 and 649 base pairs of the mitochondrial D-loop region and COI, respectively. DNA amplifications were performed using species-specific primer toward 35 individuals representing different populations. Phylogenetic analyses showed that two main clades representing populations in southern and northern Peninsular Malaysia. The results demonstrate that subspecies of T. obscurus in Peninsular Malaysia does not support classification based on the morphology that recognizes five subspecies. Previous study based on morphology that classified the subspecies on Perhentian Island, Terengganu, as T. obscurus styx is not recognized in this study. This subspecies happened to merge with the population in northern Peninsular Malaysia. Trachypithecus o. styx probably inhabited the southern peninsula and, due to the terminal Pleistocene sea level rise, spread to the east coast but could not spread farther because the subspecies was situated on offshore islands during the period. This assumption was supported by the molecular clock, which showed that subspecies on Perhentian Island spread after the Perlis population (T. obscurus flavicauda). Key words: Trachypithecus obscurus, Dusky leaf monkey, Dusky langur, Colobinae, Primate phylogeography. BACKGROUND Trachypithecus obscurus, also known as dusky leaf monkey or spectacled langur (Fig. 1), is a langur that is distributed in Myanmar, Thailand, and Peninsular Malaysia (Fig. 2), including surrounding islands (Groves 2001; Md-Zain and Ch`ng 2011). Primatologists do not agree on the total number of subspecies in this langur species. Groves (2001) suggests that there are seven subspecies of T. obscurus, which are T. obscurus obscurus (Peninsular Malaysia to northern Perlis), T. obscurus halonifer (restricted to Penang Island), T. Citation: Aifat NR, Abdul-Latiff MAB, Roos C, Md-Zain BM. 2020. Taxonomic revision and evolutionary phylogeography of dusky langur (Trachypithecus obscurus) in Peninsular Malaysia. Zool Stud 59:64. doi:10.6620/ZS.2020.59-64. Zoological Studies 59:64 (2020) doi:10.6620/ZS.2020.59-64 1 © 2020 Academia Sinica, Taiwan obscurus carbo (Langkawi Island), T. obscurus styx (restricted to Perhentian Island), T. obscurus flavicauda (northern Perlis), T. obscurus sanctorum (Myanmar), and T. obscurus seimundi (Phangan Island, Thailand). Meanwhile, Brandon-Jones et al. (2004) suggested 11 subspecies, seven of which are similar to those of Groves (2001) with four additional subspecies: T. obscurus corax, T. obscurus phayrei, T. obscurus shanicus, and T. obscurus smithi. Roos et al. (2014), the latest study to classify T. obscurus subspecies, identified seven subspecies based on Groves (2001). Trachypithecus o. phayrei, suggested by Brandon-Jones et al. (2004), has been upgraded to species level as T. phayrei. This subspecies was upgraded to the species level based on the fur coloration of the tail, leg, and pelage being lighter than that of the obscurus group (Blyth 1847; Groves 2001; Roos et al. 2014). The biogeography of T. obscurus is poorly defined because the geography and environment of its habitats change regularly. During the Pliocene and Pleistocene, drastic climate change had influenced the species distribution in certain tropical forest habitats (Bird et al. 2005; Meijaard 2003; Morley 2000). The Asian colobine’s ancestor most likely invaded Southeast Asia via an emerging land bridge connecting Africa and the Arabian Peninsula during the late Miocene (Stewart and Disotell 1998; Whybrow 1992). Habitat isolation that happened around 3.17 ± 0.31 million years ago (MYA) caused other species and subspecies of Trachypithecus to emerge (Karanth et al. 2008). Subspecies in Thailand are classified as T. o. seimundi, and T. o. sanctorum was classified as a subspecies in Myanmar. Trachypithecus obscurus in Peninsular Malaysia was classified into five subspecies: T. o. obscurus, T. o. halonifer, T. o. carbo, T. o. styx, and T. o. flavicauda (Brandon-Jones et al. 2004; Groves 2001; Roos et al. 2014). However, there are discrepancies around the distributions of these subspecies. Groves (2001) proposed that T. o. halonifer is endemic to Penang Island, but BrandonJones et al. (2004) expanded the distribution of the subspecies to Langkawi Island because they considered the subspecies’ pelage colour to be nearly the same as that of the population in Langkawi. Meanwhile, Groves (2001) argued that the population in Langkawi has darker hair dorsal to that of T. o. halonifer in Penang. Therefore, he proposed a subspecies in Langkawi: T. o. carbo. These subspecies are said to have been distributed at Terutau, Langkawi Island, and Dayang Bunting Island as well as the west coast of Peninsular Malaysia in both the Thai-Malaysia boundary areas (Groves 2001). Morphological characters in taxonomic studies were used to hypothesize T. obscurus subspecies classification, but no previous study has used genetics to classify subspecies. Page et al. (1999) studied the phylogenetic relationships of 16 species in family Cercopithecidae using genetic samples of T. obscurus and other Old World monkeys. In order to prove that Presbytis and Trachypithecus are indeed two separate genera, Md-Zain et al. (2005 2008) used T. obscurus as a representative of Trachypithecus. Md-Zain et al. (2008) sequenced two loci in the Y chromosome— testis-specific protein and sex-determining region— to explain the separation. They precisely separated Fig. 1. Spectacled langur owing to its interesting looks. Photographed by Ridwan Rahman. page 2 of 11 Zoological Studies 59:64 (2020) © 2020 Academia Sinica, Taiwan Trachypithecus and Presbytis supported by genetic data, behavior, morphology, and ecology (Abdul-Latiff et al. 2019; Md-Zain et al. 2010; Ruslin et al. 2019; Vun et al. 2011). Raaum et al. (2005) were the first to use complete mtDNA sequences from four main tribes in Cercopithecoid, and T. obscurus was used to represent Asian Colobinae. Osterholz et al. (2008) used T. obscurus to represent Trachypithecus to study the phylogenetic position of the langur from Semnopithecus and Trachypithecus among Asian colobines. Genetic relationships of langur species of Trachypithecus and Presbytis were determined by Chaveerach et al. (2007) using amplified fragment length polymorphism markers. This study used blood samples from five species of T. obscurus from Southern Thailand. Perelman et al. (2011) presented a molecular phylogeny of living primates comprising approximately 90% of all living genera, including Trachypithecus, but T. obscurus was highlighted to represent Trachypithecus in this study. Groves (2001) and Brandon-Jones et al. (2004) classified T. obscurus in Malaysia. However, from their classification, there is a conflict for certain subspecies. Based on the findings from previous research, using T. obscurus as the sole answer to the taxonomic position of T. obscurus in Peninsular Malaysia does not accurately estimate its position. Thus, we aimed to determine the molecular systematics and biogeographical history of T. obscurus in Peninsular Malaysia using D-loop and COI genes of mtDNA sequences. Based on the analysis of the samples derived predominantly from wild living animals of known location, our results enabled us to a) provide the phylogeny for T. obscurus, b) estimate divergence times between lineages, and c) provide a reliable taxonomic classification for T. obscurus. This is the first study to test classifications of subspecies T. obscurus using genetic tools. MATERIALS AND METHODS DNA Extraction, Polymerase Chain Reaction (PCR), and Sequencing A total of 35 genetic samples of T. obscurus were used in this study (Table 1). All 35 samples were collected across mainland Peninsular Malaysia including surrounding islands such as Perhentian Island, Langkawi Island, and Penang Island (Fig. 2). Fecal samples were collected fresh in the field and were kept in alcohol 97% to preserve the DNA, whereas tissue samples were collected from roadkill accidents. DNA was extracted from 0.5–1.0 g of fecal samples using the innuPREP Stool DNA Kit (Analytik Jena, Germany) and 0.04 g of tissue samples using the innuPREP DNA Mini Kit (Analytik Jena, Germany) following the manufacturer’s protocol (Aifat et al. 2016a). About 800 bp of mtDNA D-loop region and COI were amplified through a polymerase chain reaction (PCR) using a Mastercycler Nexus (Eppendorf North America, Inc.). PCR reactions were generated using MyTaq Red Mix (Bioline). The primers used in this study were designed to be species specific to maximize the extent of both region sequences and to avoid amplifying Numts. The primers for the D-loop region were designed by Abdul-Latiff et al. (2017), and those for COI by the authors in this study (Table 2). The parameters for the PCR reaction were as follows: initial denaturation for 3 min at 95°C; followed by 30 cycles of denaturation for 15 s at 95°C, annealing for 30 s at 52°C (D-loop) and 49°C (COI), extension for 10 s at 72°C; and a final extension stage for 10 min at 72°C (Aifat et al. 2016b). A double PURE kit was used to purify the PCR products, and the samples were sent to Apical Scientific Sdn. Bhd. (Malaysia) for sequencing. Sequence and Phylogenetic Analyses The sequences from both regions were obtained and edited using Bioedit Sequence Alignment Editor. All T. obscurus DNA sequences were deposited into GenBank under accession numbers MT193169– MT193202. The edited sequences were validated using sequence similarity searches (GenBank BLASTn) to ensure that the targeted species and loci were obtained. The sequences were treated separately to detect any incongruence and each individual sequence was evaluated. The datasets were then combined, and phylogenetic analysis was constructed. All 35 sequences of T. obscurus and one sequence representing Trachypithecus cristatus (NC023971: Jiang and Huang 2016), one sequence of T. auratus (KY117596: MohdSalleh et al. 2017), three sequences of T. obscurus (AY863425, NC006900: Raaum et al. 2005; MF872234: Roos et al. 2019) from GenBank were aligned using the MEGA 7 ClustalW multiple alignment algorithm (Kumar et al. 2016). The sequence information from GenBank was unknown as samples were obtained from captivity from Wuppertal Zoo, Germany and Houston Zoological Gardens, USA. The aligned sequences were analyzed at three levels: sequence analysis, phylogenetic analysis, and haplotype analysis. The analyses were performed using PAUP 4.0b10 (Swofford 2002) and DnaSP 4.0 (Rozas et al. 2003). Sequence analysis was done to reveal genetic distances and single-nucleotide polymorphisms (Md-Zain et al. 2018). Phylogenetic analysis was used to construct three types of phylogenetic trees using three different page 3 of 11Zoological Studies 59:64 (2020) © 2020 Academia Sinica, Taiwan softwares: distance based (Neighbor-Joining, NJ) using MEGA 7 (Kumar et al. 2016), character based (Maximum Parsimony, MP) using PAUP 4.0b10 (Swofford 2002), and Bayesian inference (BI) using MrBayes 3.1 (Huelsenbeck and Ronquist 2001). NJ tree reconstruction was tested with bootstrap values of 1,000 using the Kimura two-parameter model. For the MP trees, tree bisection reconnection algorithms were used. The divergence times of T. obscurus were estimated using BEAST v 1.10.4 and BEAUti v 1.10.4 (Drummond et al. 2012). As calibration point, T. cristatus was used as a second calibration point according to Perelman et al. (2011), whereas T. auratus was used as a fossil record around 1.9 ± 0.05 MYA (Jablonski and Tyler 1999). RESULTS Amplification products 723and 649-bp long were obtained for D-loop and COI gene fragments, respectively. The combined analysis, about 1372 bp, was used for phylogenetic and haplotype analyses. Of about 49 variable sites from sequence analyses, 35 were parsimony informative. Analysis of the pairwise sequences using the Kimura two-parameter for the northern population of T. obscurus in Langkawi IslandKedah, Perlis, and Penang Island was 0.011 to 0.015. The Perhentian Island-Terengganu population showed a low genetic distance value for northern population (0.011–0.013) (Table 3). The eastern population of T. obscurus in Selangor, Johor, and Melaka showed genetic distances of 0.007–0.016. The NJ, MP, BI, and molecular clock phylogenetic Table 1. Details on the samples used in this study No. Sample Code Locality Type of sample 1 ALTOK408 Langkawi Island Fecal 2 ALTOK409 Langkawi Island Fecal 3 ALTOK410 Langkawi Island Fecal 4 ALTOK411 Langkawi Island Fecal 5 ALTOK412 Langkawi Island Fecal 6 ALTOK416 Langkawi Island Fecal 7 ALTOK417 Langkawi Island Tissue 8 BM19 Kedah Tissue 9 MMTOK595 Kedah Fecal 10 ALTOP512 Penang Island Fecal 11 ALTOP514 Penang Island Fecal 12 ALTOP522 Penang Island Fecal 13 ALTOP509 Penang Island Fecal 14 ALTOP508 Penang Island Fecal 15 ARTOT525 Perhentian Island Fecal 16 ARTOT527 Perhentian Island Fecal 17 ARTOT607 Terengganu Fecal 18 ALTOR497 Perlis Fecal 19 ALTOR498 Perlis Fecal 20 ALTOR499 Perlis Fecal 21 FRTOB09 Selangor Tissue 22 FRTOB442 Selangor Fecal 23 FRTOB443 Selangor Fecal 24 FRTOB444 Selangor Fecal 25 FRTOB446 Selangor Fecal 26 FRTOB447 Selangor Fecal 27 FRTOB448 Selangor Fecal 28 ALTOJ544 Johor Fecal 29 ALTOJ545 Johor Fecal 30 ALTOJ546 Johor Fecal 31 ALTOJ547 Johor Fecal 32 ARTOM611 Melaka Fecal 33 ARTOM612 Melaka Fecal 34 ARTOM615 Melaka Fecal 35 BM17 Melaka Tissue page 4 of 11Zoological Studies 59:64 (2020) © 2020 Academia Sinica, Taiwan trees exhibited identical topological formations and were summarized in a single phylogenetic tree (Fig. 3). All the phylogenetic trees showed absolute division between T. obscurus from GenBank and other obscurus individuals supported with 0.99 (BI), 99% (NJ) and 100% (MP) bootstrap values. Excluding obscurus samples from GenBank, T. obscurus diverged to form two separate major clades supported by 0.8 (BI), 88% (NJ) and 90% (MP) bootstrap values. The first major clade comprised all samples from southern populations of T. obscurus—Selangor, Johor, and Malacca. The second major clade comprised samples from the northeast population. Sampel ARTOT607 from Terengganu was the earliest to diverge from the clade, a split that was supported by 0.6 (BI), 70% (NJ) and 77% (MP) bootstrap values. The molecular divergence phylogenetic tree was constructed using the uncorrelated lognormal relaxed-clock model. This was done to estimate the Fig. 2. Map of T. obscurus distribution in Southeast Asia and the sampling locations in Peninsular Malaysia. N Table 2. Sequence primers that were designed to be species specific to the target locus Primer name Primer sequence (5’-3’) Region Annealing temperature (°C) LATIFF1978_F AAGTGCTTAACCGTCCATAG D-loop 52.0 LATIFF1909_R AAGGCAAATTATGTGGGAGT AIFAT0326_F AACAGATCGTAATCTCAACA COI 49.0 AIFAT0109_R GTTCATCAAATGTGTGGTAG Table 3. Average pairwise genetic distances among populations of T. osbcurus [1] [2] [3] [4] [5] [6] [7] [1] Langkawi Island-Kedah - [2] Perlis 0.014 - [3] Penang Island 0.011 0.015 - [4] Perhentian Island-Terengganu 0.012 0.013 0.011 - [5] Selangor 0.019 0.025 0.022 0.020 - [6] Johor 0.018 0.025 0.021 0.019 0.007 - [7] Melaka 0.022 0.026 0.024 0.025 0.016 0.013 - page 5 of 11Zoological Studies 59:64 (2020) © 2020 Academia Sinica, Taiwan substitution rate for all nodes in the tree and establish the divergence dates of the T. obscurus population in Peninsular Malaysia. The analysis indicated that the T. obscurus group diverged from T. cristatus around 2.12 MYA. Molecular clock separation between the southern and north-east populations of T. obscurus occurred around 0.15 MYA. Fig. 3. Summary of Neighbor-Joining, Maximum Parsimony, Bayesian Inference, and Molecular clock phylogenetic tree of T. obscurus. Values above the branch represent posterior probability for BI/bootstrap values for NJ/bootstrap values for MP. Values on nodes represent divergence times of the clade. Table 4. Haplotype analysis with segregation sites for T. obscurus Hap Segregation sites [ [ [ #Hap_1 #Hap_2 #Hap_3 #Hap_4 #Hap_5 #Hap_6 #Hap_7 #Hap_8 #Hap_9 #Hap_10 #Hap_11 #Hap_12 #Hap_13 #Hap_14 #Hap_15 #Hap_16 1111 1111111111 1122223333 555555667] 1222333355 6666880000 1113445788 8901284555 134478552] 0236036956 0129041789 1563123801 7051329267 996834132] TTTCGGTATC ACTGGATGTT CTACGATTAT TATGACCTAA GCATTATCG ....A..... .T.....AC. ..G.....G. ..C...T.G. ......... ....A..... ....A..AC. .....G.... CG....T... .......T. ....A..... .T..A.CA.. ...T...... ........G. .T.C..... ....A..... .T..A.CA.. ...T...... ........G. .TGC..... ....A....T ....AG.A.. .......... ..C....C.. ......... .C..A..... ....A.CA.. .......... ..C..A.... .T...G... ....A..... ....A.CA.. .......... ..C....... .T...G... ..CTA..G.. ..C....A.. .......... ......T... ......... ....A..... ..C....A.C T...A.CC.. ....G.T..G ......C.. ..C.A..G.. G.CA...A.C ......CC.. ...A..T... ......... ....AAC... G.CA..CA.. ....A.CC.C ...A..T..G .T..C.... ..C.A..... G.CA...A.C ....A.CC.. ...A..T... ......... ....AAC.CT GTCA...A.C .C..A.C... ......T.G. A........ ....AAC.CT GTCA...A.C .C..A.C... ......T.G. ........A C...A..... G.CA...A.C ....A.CC.. .........G ......... page 6 of 11Zoological Studies 59:64 (2020) © 2020 Academia Sinica, Taiwan Haplotype analysis and Minimum Spinning Network (MSN) Data were collected for a total of 16 T. obscurus haplotypes (Table 4), with a haplotype diversity of 0.9234. Based on the analysis, the populations in Langkawi-Kedah, Perhentian-Terengganu, Selangor, and Melaka were recorded to have four haplotypes: Hap 1–3, Hap 7–9, Hap 10–12, and Hap 14–16. The population in Perlis was recorded to have two haplotypes: Hap 4 and Hap 5. Meanwhile, both of populations in Penang Island and Johor were recorded to have one haplotype each: Hap 6 and Hap 13, respectively. Based on the haplotype analysis, a minimum spanning network (MSN) was used to picture the relationships among haplotypes (Fig. 4). The populations on Perhentian Island-Terengganu were found to differ at 14 mutational sites, even though both populations were in the same state. The population on Terengganu (Hap 9) was closer to the population in Langkawi Island, with seven mutational sites. Meanwhile, the population on Perhentian Island (Hap 7 and Hap 8) was closest to the population in Perlis, with six mutational sites in the same sequence. DISCUSSION Taxonomy status of Trachypithecus obscurus in Peninsular Malaysia The results from this study support the classification in Chasen (1940), which found that T. o. obscurus is distributed in southern Peninsular Malaysia. According to Brandon-Jones et al. (2004), Groves (2001), and Roos et al. (2014), T. o. obscurus is found all over Peninsular Malaysia, except along the north coast or south of Perlis. The genetic distance analysis supported this by its results that the southern and northeastern populations had higher genetic distances. The tree topologies indicated a separation between the south and north populations, and T. o. obscurus is found in the southern Peninsular Malaysia. The classification distributions of T. o. carbo by Groves (2001) and Roos et al. (2014) are different from that of Brandon-Jones et al. (2004), which can be seen in this study. Pocock (1935) was the first to recognize T. o. carbo on Langkawi Island based on morphology— fur color; this subspecies has a darker pelage than T. o. halonifer on Penang Island (Fig. 5). Meanwhile, Brandon-Jones et al. (2004) argued that a darker pelage Fig. 4. MSN network for both regions for populations of T. obscurus. The size of the circle is directly proportional to haplotype frequency. Red nodes represent the median vector, which is often hypothesized to be the ancestral sequence. page 7 of 11Zoological Studies 59:64 (2020) © 2020 Academia Sinica, Taiwan than T. o. halonifer keeps T. o. carbo from being discriminated, but this subspecies is distributed on Terutau Island in Thailand and not on Langkawi Island. Groves (2001) stated that subspecies of T. obscurus in Terutau Island, Thailand, has a darker coloration than the subspecies on Langkawi Island and Dayang Bunting Lake as it is likely the result of the fur coloration becoming darker along a gradient from south to north. This study provides the classification based on genetic data, and the results are found to agree with those of Fig. 5. Morphology of T. obscurus in Peninsular Malaysia. A, Langkawi Island (T. o. carbo), B, Penang Island (T. o. halonifer), C, Perlis (T. o. flavicauda), D, Selangor (T. o. obscurus), E, Perhentian Island (T. o. styx). Photographed by Md-Zain, Aifat Rahman and Abdul-Latiff. A. C. E. B. D. page 8 of 11Zoological Studies 59:64 (2020) © 2020 Academia Sinica, Taiwan Groves (2001) and Roos et al. (2014). In this study, the population of T. obscurus from Penang Island formed its own clade and did not shared the same haplotype with the population on Langkawi Island. Thus, this study rejected the classification of subspecies by Brandon-Jones et al. (2004), who stated that subspecies in Penang and Langkawi Island are T. o. halonifer. The results from this study supports the morphological classifications by Groves (2001) and Roos et al. (2014). Chasen (1935 1940) was the earliest to identify subspecies T. obscurus on Langkawi Island as T. o. carbo and stated that T. o. halonifer is endemic to Penang Island. This statement is supported by Groves (2001) and Roos et al. (2014). The population of T. obscurus on Perhentian Island was represented on an unstable clade in the phylogeny. In the NJ phylogenetic trees for both regions, this population formed a paraphyletic clade with the population in Perlis and Penang Island (63% bootstrap value). Meanwhile, in the MP and BI phylogenetic trees of both regions, the population on Perhentian Island formed monophyletic and polyphyletic clades with the populations in Perlis. Subspecies of T. obscurus on Perhentian Island has been classified as T. o. styx (Kloss 1911). Pocock (1935) did not differentiate T. o. styx and T. o. carbo (Langkawi Island population) based on fur coloration. However, both of these subspecies have opposite geographical areas and are separated by T. o. obscurus and T. o. halonifer. Groves (2001) states that T. o. styx may have had a darker pelage and inhabited the northern Peninsular Malaysia (T. o. flavicauda) (Fig. 5) until the late Pleistocene. During this time, sea level rose to the north of Peninsular Malaysia and caused the species to migrate to the south of Peninsular Malaysia. At that time, most of Peninsular Malaysia was underwater, and the species could not migrate or inhabit the area because it was an isolated island (Groves 2001). Biogeography of T. obscurus in Peninsular Malaysia Roos et al. (2011) suggested that the primitive ancestor of the Asian Colobinae migrated to Asia via a land route connecting Africa and the Arabian Peninsula during the late Miocene (Stewart and Disotell 1998; Whybrow 1992). This is supported by the discovery of Colobinae primate fossils dated to the last days of the Miocene, which were found in the Tibetan plateau via the northern road of the Himalayas (Davies and Oates 1994). The Hengduan Mountain, which is now a border area between Myanmar and China, is said to be the hotspot of this genetic diversity (Peng et al. 1993; Thinh et al. 2010). The rise in wide river waters in Southeast Asia such as the Mekong River, Salween River, and Yangtze River caused Trachypithecus to migrate south of the Asia mainland, which includes Thailand and Peninsular Malaysia (Hallet and Mollnar 2001). According to Abdul-Latiff (2015), species diversification from this genus formed four groups of species: those in South Asia, center of Southeast Asia, southern Southeast Asia, and northern Southeast Asia. This situation illustrates that the ancestors of extant species are likely to survive in these four separate groups. The cool climate during the Pliocene led to declines in the area of evergreen forest (Batchelor 1979; Morley 2000) and caused these four groups to separate (Wang 1994). In this study, we proposed that T. obscurus invaded Peninsular Malaysia around 0.25 MYA. This process caused the species to separate into north-east and south populations. Separation of T. obscurus occurred in the mainland before the species were distributed in the archipelago area around Peninsular Malaysia. Trachypithecus o. styx diverged after separation from the Perlis population T. o. flavicauda. The molecular clock results showed that the hypothesis of Groves (2001) is acceptable. CONCLUSIONS Classification of the Trachypithecus obscurus populations in Peninsular Malaysia showed that the north-eastern and southern populations are distinct. Based on the studies conducted, genetic data show that the subspecies classification in Peninsular Malaysia is not fully resolved for the north-eastern populations, whereas for southern population, the classification of T. o. obscurus is acceptable. Classification of subspecies T. o. styx in Perhentian Island that is based on morphology is not relevant as this subspecies is nested with populations in the north, such as Perlis and Langkawi Island. The population on Perhentian Island was a continuation of the northern population before separation occurred because of geographical factors. The molecular clock phylogenetic tree for both sequences was effective in classifying subspecies and evolutionary histories. The results of this study should be considered as preliminary study for the taxonomic revision of subspecies T. obscurus using a molecular approach. In the future, additional samples of T. obscurus from other states—especially Terengganu, Pahang and Kelantan—should be used to confirm the subspecies for T. obscurus populations in the east. From the conservation standpoint, genetic information helps the Department of Wildlife and National Parks (DWNP) manage dusky leaf monkey populations via translocation. This leaf monkey is believed to have a page 9 of 11Zoological Studies 59:64 (2020)