- Email: [email protected]
Phylogenetic relationships of megophryid frogs of the genus Leptobrachium (Amphibia, Anura) as revealed by mtDNA gene sequences
Molecular Phylogenetics and Evolution 56 (2010) 259–272
Contents lists available at ScienceDirect
Molecular Phylogenetics and Evolution journal homepage: www.elsevier.com/locate/ympev
Phylogenetic relationships of megophryid frogs of the genus Leptobrachium (Amphibia, Anura) as revealed by mtDNA gene sequences Masafumi Matsui a,*, Amir Hamidy a,b, Robert W. Murphy c,d, Wichase Khonsue e, Paul Yambun f, Tomohiko Shimada a,g, Norhayati Ahmad h,i, Daicus M. Belabut h,j, Jian-Ping Jiang k a
Graduate School of Human and Environmental Studies, Kyoto University, Kyoto 606-8501, Japan Museum Zoologicum Bogoriense, Research Center for Biology, Indonesian Institute of Sciences, Gd. Widyasatwaloka, Jl. Raya Jakarta Bogor km 46, Cibinong West Java, Indonesia Centre for Biodiversity and Conservation Biology, Royal Ontario Museum, 100 Queen’s Park, Toronto, Ont., Canada M5S 2C6 d State Key Laboratory of Genetic Resources and Evolution, Kunming Institute of Zoology, The Chinese Academy of Sciences, Kunming 650223, PR China e Department of Biology, Faculty of Science, Chulalongkorn University, Bangkok 10330, Thailand f Research and Education Division, Sabah Parks, P.O. Box 10626, Kota Kinabalu 88806, Sabah, Malaysia g Faculty of Bioenvironmental Science, Kyoto Gakuen University, Kameoka, Kyoto 621-8555, Japan h Faculty of Science and Technology, Universiti Kebangsaan Malaysia, 43600 UKM Bangi, Selangor, Malaysia i Institute for Environment and Development (LESTARI), Universiti Kebangsaan Malaysia, 43600 UKM Bangi, Selangor, Malaysia j Institute of Biological Sciences, Faculty of Science, University of Malaya, 50603 Kuala Lumpur, Malaysia k Chengdu Institute of Biology, The Chinese Academy of Sciences, Chengdu 610041, PR China b c
a r t i c l e
i n f o
Article history: Received 24 October 2009 Revised 17 February 2010 Accepted 9 March 2010 Available online 17 March 2010 Keywords: Leptobrachium Vibrissaphora Sundaland Indochina Southern China mtDNA Phylogenetics
a b s t r a c t By investigating genealogical relationships, we estimated the phylogenetic history and biogeography in the megophryid genus Leptobrachium (sensu lato, including Vibrissaphora) from southern China, Indochina, Thailand and the Sundaland. The genealogical relationships among the 30 named and unnamed taxa were estimated using 2009 bp of sequences from the mitochondrial DNA genes 12S rRNA, tRNAval, and 16S rRNA using maximum parsimony, maximum likelihood, and Bayesian inference methods. The genus Leptobrachium was a well-supported monophyletic group that contained two major clades. One clade had three subclades primarily from disjunct regions including Borneo, Peninsular Malaysia and Java, and Thailand. The Bornean subclade included one species each from the Philippines and Sumatra. The other major clade consisted of two subclades, one from Indochina and the other from southern China (Vibrissaphora). Divergence times estimated an old evolutionary history of each subclade, one that could not be explained by the geohistory of Southeast Asian major landmasses. Ó 2010 Elsevier Inc. All rights reserved.
1. Introduction The megophryid genus Leptobrachium Tschudi, 1838 (type species L. hasseltii from Java) is diagnosed by having a broad head and thin limbs (Dubois and Ohler, 1998). The genus is often considered to contain two subgenera, Vibrissaphora Liu, 1945, with adult males bearing spines on upper lip, and Leptobrachium, which is without such spines (Ohler et al., 2004). Vibrissaphora is distributed in southern China and Indochina, and it includes five to seven species (Frost, 2009; Rao and Wilkinson, 2008; Fei et al., 2009). Subgenus Leptobrachium is distributed from Indochina to Sundaland and it contains 15–17 species (Frost, 2009; Fei et al., 2009). Using DNA sequence data, both Zheng et al. (2008) and Rao and Wilkinson (2008) place Vibrissaphora within the genus Leptobrachium, and neither group of researchers recognizes subgenera. However, this * Corresponding author. Fax: +81 75 753 6846. E-mail address: [email protected] (M. Matsui). 1055-7903/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.ympev.2010.03.014
taxonomic conclusion is made without exploring the extent of genetic diversity within Sundaland and neither study includes the type species of the genus. Sundaland contains the Malay Peninsula and the Indonesian islands of Sumatra, Java, Bali, Borneo, and smaller islands west of the Makassar and Lombok straits. All of these areas are linked by the shallow-water (<200 m) Sunda Shelf, which was exposed during periods of low sea level in the Pleistocene. The eastern boundary of Sundaland is Wallace’s Line where the Indomalayan and Australasian faunas meet. Studies of some anuran lineages from Sundaland indicate that species distributions and phylogenies are strongly related to the geological history of this region (e.g., Emerson et al., 2000; Inger and Voris, 2001; Brown and Guttman, 2002; Matsui et al., 2010). Seven species of Leptobrachium are known from Sundaland and peninsular Thailand as follows: L. hasseltii Tschudi, 1838; L. montanum Fischer, 1885; L. abbotti (Cochran, 1926); L. hendricksoni Taylor, 1962; L. nigrops Berry and Hendrickson, 1963; L. gunungense
260
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
Malkmus, 1996, and L. smithi Matsui, Nabhitabhata, and Panha, 1999. Among these species, L. montanum, L. abbotti, and L. gunungense are endemic to Borneo and L. smithi is mainly distributed in Thailand and the northern part of Peninsular Malaysia, as well as through Myanmar to Assam, India (Sengupta et al., 2001). Leptobrachium hendricksoni is distributed from southern Thailand through the Malay Peninsula to western Borneo and Sumatra, and L. nigrops occurs on the Malay Peninsula, Belitung Island, and western Borneo. How and when they obtained their present-day distributions in the Sunda Islands are biogeographically interesting topics. In Leptobrachium, taxonomic problems span from the supraspecific category to the species level. The name L. hasseltii was applied to many Southeast Asian populations (Inger, 1954, 1966; Berry, 1975) until Inger et al. (1995) clarified that L. hasseltii from Borneo was not conspecific with the Javanese population. They applied the names L. montanum and L. abbotti to Bornean species. However, some populations, such as those from the Philippines, are still referred to as L. hasseltii because of the absence of a taxonomic reassessment. Iskandar (1998) and Matsui et al. (1999) suggested that true L. hasseltii was likely restricted to Java and adjacent islands, and Dubois and Ohler (1998) suggested the possible occurrence of more than one species of Leptobrachium on Java, based on a large extent of variation in female body size of Leptobrachium (sensu lato). Such taxonomic arguments still need to be clarified. These uncertainties mainly derive from the use of a few specimens from limited ranges. Intra- and interspecific variation is inadequately assessed. Some critical characteristics useful for species identification are lost upon preservation of museum vouchers. Molecular data and phylogenetic analyses can enable the identification of species and a higher level taxonomy in Leptobrachium. In this study, we use mtDNA gene sequences from various populations of Leptobrachium, especially from Sundaland, to evaluate the taxonomic status of many populations and hypothesize the evolutionary history of the species in this genus.
2. Materials and methods 2.1. Sampling design We examined a total of 80 partial DNA sequences of the mitochondrial DNA genes 12S rRNA, tRNAval, and 16S rRNA from 30 species/subspecies of the genus Leptobrachium (sensu lato) and six outgroup species (Figs. 1 and 2 and Table 1). Our own sampling of Leptobrachium consisted of 60 specimens of 21 taxa mainly obtained from Sundaland and Indochina, including several undescribed or unidentified taxa including the following: Leptobrachium sp. 1, a Sumatran species that has distinct eye color (Hamidy and Matsui, 2010); Leptobrachium sp. 2, a Philippine species often called L. hasseltii; Leptobrachium sp. 3 from Gwa, Myanmar, catalogued as L. hasseltii (Zheng et al., 2008); Leptobrachium sp. 4 from Pilok, Thailand, a dried carcass whose identification was difficult; and Leptobrachium sp. 5, which resembles Vibrissaphora but lacks male labial spines. We used Leptolalax heteropus (Boulenger, 1900) and Pelodytes punctatus (Daudin, 1802) as outgroup taxa. DNA sequences from GenBank were obtained for 11 Indochinese and Chinese species of Leptobrachium (sensu lato) and the following four outgroup species: Oreolalax rhodostigmatus Hu and Fei in Liu, Hu, and Fei, 1979 (EF397248: Fu et al., 2007), Megophrys nasuta (Schlegel, 1858) (DQ283342: Frost et al., 2006), Pelobates fuscus (Laurenti, 1768) (DQ283113: Frost et al., 2006), and Scaphiopus holbrookii (Harlan, 1835) (DQ283156: Frost et al., 2006). Leptolalax Dubois, 1980 and Oreolalax Myers and Leviton, 1962 are the sister group of Leptobrachium (Lathrop, 1997; Zheng et al., 2004, 2008; Fu
et al., 2007), whereas Pelobates Wagler, 1830 (Pelobatidae Bonaparte, 1850), Pelodytes Bonaparte, 1838 (Pelodytidae Bonaparte, 1850), and Scaphiopus Holbrook, 1836 (Scaphiopodidae Cope, 1865) form a monophyletic sister group with Megophryidae (García-París et al., 2003; Dubois, 2005). Voucher specimens/tissues are stored in ABTC (Australian Biological Tissue Collection, South Australian Museum), BOR (BORNEENSIS collection, University Malaysia Sabah), IEBR (Institute of Ecology and Biological Resources, Hanoi, Vietnam), KUHE (Kyoto University, Graduate School of Human and Environmental Studies), MDK (Department of Conservation and Ecotourism, Faculty of Forestry, Bogor Agricultural Institute), MNHN (Museum National d’Histoire Naturelle, Paris), MZB (Museum Zoologicum Bogoriense), ROM (Royal Ontario Museum), SP (Sabah Parks), UKM (University Kebangsaan Malaysia), UM (University Malaya), and UTA (Department of Biology, University Texas at Arlington). 2.2. Preparation of DNA, PCR and DNA sequencing We obtained tissues from frozen or ethanol (95–99%) preserved specimens and extracted total genomic DNA using standard Phenol–chloroform extraction procedure (Hillis et al., 1996). We homogenized tissues in 0.6 ml STE buffer containing 10 mM Tris/ HCl, pH 8.0, 100 mM NaCl and 1 mM EDTA, pH 8.0. We added Proteinase K (0.1 mg/ml) to the homogenate solutions and digested proteins for 4–12 h at 55 °C. The solution was treated with phenol and chloroform/isoamyl alcohol and DNA was precipitated with ethanol. DNA precipitates were dried and then resuspended in 0.6 ml TE (10 mM Tris/HCl, 1 mM EDTA, pH 8.0) and 1 ll was subjected to polymerase chain reaction (PCR). The PCR cycle included an initial denaturation step of 5 min at 94 °C and 33 cycles of denaturation for 30 s at 94 °C, primer annealing for 30 s at 48–50 °C, and extension for 1 min 30 s at 72 °C. Primers used in PCR are shown in Table 2. The PCR products purified using polyethylene glycol (PEG, 13%) precipitation procedures were used directly as templates for Cycle Sequencing Reactions with fluorescent-dye-labeled terminator (ABI Big Dye Terminators v.3.1 cycle sequencing kit). The sequencing reaction products were purified by ethanol precipitation following the manufacture’s protocol and then run on an ABI PRISM 3130 genetic analyzer. All samples were sequenced in both directions using the same primers as for PCR. 2.3. Phylogenetic analysis Aligned, concatenated sequences of 12S rRNA, tRNAval, and 16S rRNA yielded a total 2009 bp positions. We used Chromas Pro software (Technelysium Pty Ltd., Tewntin, Australia) to edit the sequences, and align them using the ClustalW option of Bioedit (Hall, 1999). The initial alignments were then checked by eye and adjusted slightly. Phylogenetic trees were constructed using maximum parsimony (MP), maximum likelihood (ML), and Bayesian inference (BI). MP trees, obtained using PAUP4.0b10 (Swofford, 2002), involved a heuristic search with the tree bisection recognition (TBR) branch-swapping algorithm, and 100 random additions replicates. Transitions and transversions were equally weighted, and gaps were treated as missing data. ML analysis were performed by Treefinder version June 2007 (Jobb et al., 2004), with the general time-reversible (GTR) model of DNA evolution with a gamma shape parameter (G), identified as the best-fitting model under the Akaike information criterion implemented in Kakusan 3 (Tanabe, 2007). BI and Bayesian posterior probabilities (BPP) were estimated using MrBayes 3.0b4 (Huelsenbeck and Ronquist, 2001), under the GTR model with G and proportion of invariable sites (I), selected by MrModeltest2.2 (Nylander, 2004). BI used four simultaneous Metropolis coupled Monte Carlo Markov chains for 6,000,000 generations. We sampled a tree every 100 generations
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
261
Fig. 1. Map of Southeast Asia showing sampling localities of Leptobrachium (sensu lato) and related species included in this study. Sample numbers are included in Table 1. Triangle: Subclade I; reverse triangle: Subclade II; filled square: Subclade III; circle: Subclade IV; open square: Subclade V.
and calculated a consensus topology for 30,001 trees after discarding the first 30,000 trees (burn-in = 3,000,000). Strength of nodal support in the MP analyses used non-parametric bootstrapping (MPBS; Felsenstein, 1985) with 1000 pseudoreplicates in PAUP, and 100 replicates were used for the ML tree (MLBS). A priori, we regarded tree nodes with bootstrap value 70% or greater as sufficiently resolved (Huelsenbeck and Hillis, 1993), and those between 50% and 70% as tendencies. In the BI analysis, nodes with a BPP of 95% or greater were considered significant (Leaché and Reeder, 2002).
2.4. Estimation of divergence time We estimated the divergence times using a Bayesian relaxed molecular clock calculated in BEAST (Drummond and Rambaut, 2007) using 60 million generations under a GTR+G+I substitution model and uncorrelated log-normal ‘‘relaxed” clock rate model (Drummond et al., 2006). We sampled the MCMC chain every 1000 generations, for a total of 10 million samples, and assessed convergence to the stationary distribution through inspection of the likelihood and parameter sample plots in Tracer version 1.4
262
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
Fig. 2. Maximum likelihood (ML) phylogram of 2009 bp of 12S rRNA, tRNAval and 16S rRNA mitochondrial genes for samples of Leptobrachium (sensu lato) and related species. Sample numbers are included in Table 1. Numbers above branches represent bootstrap supports for maximum parsimony (MP) /ML/ and Bayesian posterior probabilities. Asterisks indicate nodes with full bootstrap supports for ML and MP (100%) inferences and Bayesian posterior probabilities (PP = 100%).
(Rambaut and Drummond, 2007). A burn-in of the first one million samples was used. Three external calibration points provided by Roelants et al. (2007) were used to estimate dates of the cladogenic events. The
divergence time between Pelodytes and Pelobatidae + Megophryidae was assumed to be 143.1 MYBP (95% Credibility Interval [CI] of 120.8–166.1 MYBP), between Pelobates and Asian members at 114.6 MYBP (CI 93.2–135.9), and between Leptolalax and
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
263
Table 1 Sample of Leptobrachium (sensu lato) and outgroup species used for mtDNA analysis in this study together with the information on voucher, collection locality and GenBank accession numbers. UN: Unnumbered. See text for voucher abbreviations. Sample no
Species
Voucher
Locality
GeneBank
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73
L. montanum lineage 1 L. montanum lineage 1 L. montanum lineage 1 L. montanum lineage 1 L. montanum lineage 1 L. montanum lineage 1 L. montanum lineage 1 L. montanum lineage 1 L. montanum lineage 1 Leptobrachium sp. 1 Leptobrachium sp. 1 Leptobrachium sp. 1 L. gunungense L. gunungense L. abbotti L. abbotti L. abbotti L. montanum lineage 2 L. montanum lineage 2 Leptobrachium sp. 2 L. hendricksoni L. hendricksoni L. hendricksoni L. hendricksoni L. hendricksoni L. hendricksoni L. hendricksoni L. hendricksoni L. hasseltii L. hasseltii L. hasseltii L. hasseltii L. hasseltii L. hasseltii L. nigrops L. nigrops L. nigrops L. nigrops L. nigrops L. nigrops L. nigrops L. smithi L. smithi L. smithi L. smithi L. smithi L. smithi L. smithi L. smithi L. smithi Leptobrachium sp. 3 Leptobrachium sp. 4 V. l. liui V. l. yaoshanensis V. leishanensis V. jiulongshanensis L. chapaense lineage 1 L. chapaense lineage 1 L. chapaense lineage 1 L. chapaense lineage 1 V. boringii Leptobrachium sp. 5 V. echinata V. ailaonica V. promustache L. chapaense lineage 2 L. chapaense lineage 3 L. hainanense L. chapaense lineage 3 L. mouhoti L. pullum V. ngoclinhensis L. xanthospilum
KUHE UN larva KUHE 42812 BOR 22008 BOR 04B019 KUHE 42813 KUHE 42815 KUHE 17306 KUHE UN larva MDK 01 KUHE 42805 MZB Amp 15862 UTA_A53689 BOR 22956 SP 3825a BOR 12866 KUHE 39295 BOR 04B018 BOR 08481 SP 21481 ABTC 76306 KUHE 15336 KUHE 52403 UKM HC110 KUHE UN tissue MDK 10 KUHE UN tissue KUHE 15680 KUHE 52150 KUHE 42807 KUHE 42808 MZB UN tissue UTA_A53688 KUHE 42818 KUHE 42820 MZB Amp1790 MZB Amp1791 KUHE 15430 KUHE 15658 UKM HC0727 KUHE 42587 KUHE 42590 KUHE 19281 KUHE 19282 KUHE 19834 KUHE 19939 KUHE 20200 KUHE 20201 KUHE 23342 UM D0139 CAS 222215 CAS 222293 KUHE UN tissue IZCASH 30020 KUHE UN larva IZCASH 30004 IZCASH 30034 KUHE UN tissue ROM 41243 KUHE 19122 IZCASH 30048 IZCASH 30021 KUHE 34396 MNHN 1999.5657 IZCASH 30046 IZCASH 30044 AMNH 163791 ROM 32176 KUHE UNL MNHN 1997.5249 FMNH 261758 IEBR 2780 IEBR 2827 ROM 32177
Indonesia, S. Kalimantan, Banjar, Paramasan Indonesia, S. Kalimantan, Banjar, Paramasan Malaysia, Sabah, Tawau Malaysia, Sabah, Tawau Indonesia, C. Kalimantan, Lamandau, Belantikan Indonesia, C. Kalimantan, Lamandau, Belantikan Malaysia, Sarawak, Matang, Serapi Malaysia, Sarawak, Matang, Serapi Indonesia, W. Kalimantan, Betung Kerihun, Nanga Bungo Indonesia, Sumatra, Lampung, Liwa, Kubu Perahu Indonesia, Sumatra, Lampung, Liwa, Kubu Perahu Indonesia, Sumatra, Jambi, betw. Tanpa and Sungai penuh Malaysia, Sabah, Kinabalu, Mesilau Malaysia, Sabah, Kinabalu, Sungai Carson Malaysia, Sabah, Crocker, Ulu Kimanis Malaysia, Sabah, Kinabalu, Poring Malaysia, Sabah, Tawau Malaysia, Sabah, Crocker, Ulu Kimanis Malaysia, Sabah, Kinabalu, Silau-silau Philippines, Mindanao, Davao Malaysia, Penang Malaysia, Peninsula, Kelantan, Pasir Puteh Malaysia, Peninsula, Trengganu, Hulu Trengganu Malaysia, Peninsula, Trengganu, Sekayu Indonesia, Sumatra, Jambi, Bungo Indonesia, Sumatra, S. Sumatra, Lahat Malaysia, Peninsula, Selangor, Kuala Lumpur Malaysia, Peninsula, Johor, Endau Rompin, Selai Indonesia, Sumatra, Lampung, Liwa, Kubu Perahu Indonesia, Sumatra, Lampung, Liwa, Kubu Perahu Indonesia, W. Java, Gede-Pangrango National Park Indonesia, W. Java, Bogor, Cisarua Safari Park Indonesia, C. Java, Purworejo, Kaligesing Indonesia, Java, Yogyakarta, Kulon Progo, Kiskendo Indonesia, Belitung, Tanjung Pandang Indonesia, Belitung, Tanjung Pandang Malaysia, Peninsula, Selangor, Kuala Lumpur Malaysia, Peninsula, Selangor, Kuala Lumpur Malaysia, Peninsula, Pahang, Temerloh Malaysia, Sarawak, Kanowit Malaysia, Sarawak, Kanowit Thailand, Loei, Phu Luang Thailand, Loei, Phu Luang Thailand, Mae Hong Son, Phasua WF Thailand, Kanchanaburi, Pilok Thailand, Phetchaburi, Kaeng Krachan Thailand, Phetchaburi, Kaeng Krachan Thailand, Trang, Kaochong Malaysia, Perlis, Langkawi Myanmar, Kyaik Hto Myanmar, Rakhine State, Gwa Township, Rakhine Thailand, Kanchanaburi, Pilok China, Hunan, Man Shan China, Guanxi, Huapin China, Guizhou, Leigong Shan China, Zhejiang, Jiulong Shan Thailand, Doi Angkang China, Yunnan, 14.9 km SE of Simao Thailand, Chiang Mai, Doi Intanon China, Yunnan, Longling China, Sichuan, Emei Shan Laos, Xamneua, Phupan Vietnam, Lao Cai, Sa Pa China, Yunnan, Ailao Shan China, Yunnan, Dawei Shan Vietnam, Ha Giang, Vi Xuyen, Cao Bo Vietnam, Vinh Phu, Tam Dao China, Hainan, Wuzhi Shan Vietnam, Ben En Vietnam, Pichrada Vietnam, Kon Tum, Kon Plong Vietnam, Kon Tum, Ngoc Linh Mountain Vietnam, Gia Lai, Tram Lap
AB530391 AB530392 AB530393 AB530394 AB530395 AB530396 AB530397 AB530398 AB530399 AB530400 AB530401 AB530402 AB530403 AB530404 AB530405 AB530406 AB530407 AB530408 AB530409 AB530410 AB530411 AB530412 AB530413 AB530414 AB530415 AB530416 AB530417 AB530418 AB530419 AB530420 AB530421 AB530422 AB530423 AB530424 AB530425 AB530426 AB530427 AB530428 AB530429 AB530430 AB530431 AB530432 AB530433 AB530434 AB530435 AB530436 AB530437 AB530438 AB530439 EF672271 DQ283239 AB530440 EF544196 AB530441 EF544200 EF544205 AB530442 AB530443 AB530444 EF544239 EF544207 AB530445 AB530446 EF544225 EF544240 DQ283052 EF544232 AB530447 AB530448 EF672272 AB530449 AB530450 AB530451 (continued on next page)
264
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
Table 1 (continued) Sample no
Species
Voucher
Locality
GeneBank
74 75 76 77 78 79 80
L. banae Oreolalax rhodostigmatus Leptolalax heteropus Megophrys nasuta Pelobates fuscus Pelodytes punctatus Scaphiophus holbrookii
ROM 32200 CIB ZYCA746 KUHE 15487 FMNH 236525 A. Haas Collection MNHN 2000.2401 AMHH A168434
Vietnam, Gia Lai, Krong Pa China, Guizhou, Da Fang Malaysia, Peninsula, Perak, Larut Malaysia, Sabah, Crocker, Tenom Germany, Thuringia, Geroda, Triptis France, Villemoisiau USA, Florida, Alachua
AB530452 EF397248 AB530453 DQ283342 DQ283113 AB530454 DQ283156
Table 2 Primers used in this study. Target val
12S rRNA and tRNA
16 S rRNA
Primer
Sequence (50 –30 )
Reference
12Sh 12SA-L L1507 12SFLeptobrachium H1548 L1879 16sl2021 16L-1 H1923 H2315Leptobrachium 16sh2715 16H1
AAAGGTTTGGTCCTAGCCTT AAACTGGGATTAGATACCCCACTAT TACACACCGCCCGTCACCCTCTT CCGCCAAGTCCTTTGGGTTT TACCATGTTACGACTTTCCTCTTCT CGTACCTTTTGCATCATGGTC CCTACCGAGCTTAGTAATAGCTGGTT CTGACCGTGCAAAGGTAGCGTAATCACT AAGTAGCTCGCTTAGTTTCGG TCGTTGTTACTAGTYCTAACAT AAGCTCCATAGGGTCTTCTCGTC CTCCGGTCTGAACTCAGATCACGTAGG
Cannatella et al. (1998) Palumbi et al. (1991) Made in this study Modified from Goebel et al. (1999) Matsui et al. (2005a) Made in this study Tominaga et al. (2006) Hedges (1994) Made in this study Made in this study Tominaga et al. (2006) Hedges (1994)
Oreolalax + Leptobrachium 43.8 MYBP (CI 32.0–58.7). However, using external calibrations only might result in very old estimates (Matsui et al., 2010). Thus, we applied a date of 1.40 MYBP (CI 0.90–1.90) for two insular populations of L. hasseltii as estimated by Matsui et al. (2005b) for Odorrana hosii (Boulenger, 1891) from Sundaland.
eter = 0.290; nucleotide frequencies: A = 0.347, C = 0.210, G = 0.173, and T = 0.270). BI calculated average parameter estimates for nucleotide frequencies of A = 0.373, C = 0.206, G = 0.144 and T = 0.276, a gamma shape parameter 0.662, and proportion of invariable sites of 0.297. 3.2. Phylogenetic relationships
2.5. Variation in iris color Iris color has served as a valuable diagnostic morphological characteristic for species of Leptobrachium (Dubois and Ohler, 1998; Lathrop et al., 1998; Matsui et al., 1999). We accumulated data for this character from our own observations and descriptions and photographs in the literature. We hypothesized the evolution of iris color by mapping the data on the phylogenetic tree. 3. Results 3.1. Sequence and statistics Sequence statistics for the three gene fragments and for the combined alignment including all nucleotide positions are provided in Table 3. The aligned 12S rRNA, tRNAval, and 16S rRNA data set consisted of 2009 characters, in which 925 sites were variable, and 800 potentially phylogenetically informative. MP analysis yielded four most parsimonious trees of 4089 steps, a consistency index of 0.401 and retention index of 0.818. The ML analysis produced a topology with ln L 21943.875 (gamma shape paramTable 3 Alignment statistics for fragments of the 12S rRNA, tRNAval, and 16S rRNA (all nucleotide positions included); number of base bairs (bp), number of variable sites (vs), number of parsimony informative sites (pi), the transition–transversion ratio given for ingroups only (ti/tv).
12S rRNA tRNAval 16S rRNA Combined
bp
vs
pi
ti/tv
475 71 1463 2009
161 38 726 925
138 35 627 800
1.175 3.900 1.271 1.288
All analyses resulted in essentially the same topologies. They differed only in associations at poorly supported nodes. The ML tree (Fig. 2) infers the following sets of relationships: (i) Monophyly of Leptobrachium sensu lato (Leptobrachium and Vibrissaphora) with respect to Oreolalax and Leptolalax was supported in all trees (MPBS = 71%, MLBS = 78%, BPP = 98%). (ii) Leptobrachium sensu lato was divided into two basal monophyletic lineages, Clade A (all Sundaland and most Thai species; all support values = 100%) and Clade B (primarily Chinese and Indochinese species associated with Vibrissaphora: MPBS = 78%, MLBS = 93%, BPP = 100%). (iii) Clade A contained three monophyletic subclades whose relationships to each other were unresolved. Subclade I consisted of species from Borneo, Sumatra, and Mindanao, Philippines (all values 100%), Subclade II encompassed species mainly from Peninsular Malaysia and Java (MPBS = 76%, MLBS = 100%, BPP = 100%), and Subclade III contained ThaiMyanmar species (all values 100%). (iv) In Subclade I, Leptobrachium sp. 2 from Mindanao (sample number 20) was a sister lineage to the clade containing Bornean and Sumatran species. In the latter clade (all values 100%), L. montanum lineage 2 from Kinabalu (18, 19) formed the sister species to the clade of the remaining Bornean and Sumatran species (MPBS = 93%, MLBS = 97%, BPP = 100%). The latter clade contained two sublineages. First, L. montanum lineage 1 (populations 1–9 from Sabah, Sarawak, and Kalimantan, including Paramasan, the type locality of L. montanum; MPBS = 70%, MLBS = 65%, BPP = 100%) was united with Leptobrachium sp. 1 from Sumatra (10–12; all values 100%). Second, L. abbotti (15–17; all values 100%) and
265
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
(v)
(vi)
(vii)
(viii)
(ix)
L. gunungense (13, 14; all values 100%) from Sabah, formed a monophyletic clade (MPBS = 96%, MLBS = 97%, BPP = 100%). The clade containing L. montanum lineage 1 was not universally supported to be monophyletic (MPBS = 70%, MLBS = 65%, BPP = 100%). Thus, its sister species relationship with Leptobrachium sp. 1 needs to be verified. Subclade II contained three species, of which all samples clustered together with 100% support: L. hendricksoni (21–28) from Peninsular Malaysia and Sumatra, L. hasseltii (29–34) from Java and Sumatra, and L. nigrops (35–41) from Peninsular Malaysia, Belitung Island, and Sarawak. Of these, L. hendricksoni and L. hasseltii were sister species (MPBS = 78%, MLBS = 89%, BPP = 100%). Peninsular Malayan L. hendricksoni was paraphyletic, with respect to northern populations forming a sister group with Sumatran populations (MPBS = 83%, MLBS = 83%, BPP = 100%), and not with southern Malayan populations. In L. hasseltii, relationships of populations from western Java, central Java, and Sumatra were unresolved. In contrast, three genetically divergent lineages were discovered within L. nigrops. The population from Sarawak (40, 41) was the sister group of a well-supported (MPBS = 100%, MLBS = 99%, MPP = 100%) clade comprised of animals from Peninsular Malaysia (37–39) and Belitung Island (35, 36). Uncorrected p-distances were large, averaging 9.3% between Peninsular Malaysia and Belitung, 11.7% between Borneo and Belitung, and 11.2% between Borneo and Malay Peninsula. In Subclade III, populations of L. smithi (42–50) from Thailand, Langkawi Island of Malaysia, and Myanmar formed a monophyletic clade (MPBS = 100%, MLBS = 99%, BPP = 100%), but the relationships among these individuals were unresolved. Leptobrachium sp. 3 from Gwa, Myanmar (51) was the sister species to this clade and this association received full support. Together, these taxa formed the sister group of Leptobrachium sp. 4 from Pilok, Thailand (52) and with full support. Clade B was comprised of two subclades. Subclade IV primarily contained Vibrissaphora (all values 100%), and Subclade V consisted primarily of Indochinese species of Leptobrachium (all values 100%). Subclade IV consisted of three main clades. Leptobrachium chapaense (Bourret, 1937) lineage 2 (66) from Ha Gian, northern Vietnam, and it formed the sister species of the remaining species. The latter clade received full support and V. promustache Rao et al., 2006 (65) was the sister species of the other species (MPBS = 77%, MLBS = 100%, BP = 98%), in which several sublineages had unresolved relationships to each other. These sublineages included full support for five clades as follows: first, a clade containing V. echinata Dubois and Ohler, 1998 (63) from Vietnam and V. ailaonica Yang, Chen, and Ma in Yang, Ma, Chen, and Li, 1983 (64) from China; second, Leptobrachium sp. 5 (62) from Laos; third, V. boringii Liu, 1945 (61) from China; fourth, a fully supported clade containing L. chapaense lineage 1 (57–60) from Thailand and China; and fifth, a clade consisting of V. l. liui Pope, 1947 (53), V. l. yaoshanensis Liu and Hu in Hu, Tian, and Wu, 1978 (54), V. leishanensis Liu and Hu in Hu, Zhao, and Liu, 1973 (55) and V jiulongshanensis Wei and Zhao, 1981 (56) from China (MPBS = 99%, MLBS = 100%, BPP = 100%). In the last clade, two subspecies of V. liui were monophyletic (MPBS = 98%, MLBS = 100%, BP = 100%). Subclade V comprised five lineages whose relationships to one another were unresolved. Of these five, Vietnamese L. mouhoti Stuart, Sok, and Neang, 2006 (70) and L. pullum (Smith, 1921) (71) formed a fully supported clade, as did Chinese L. hainanense Ye and Fei in Ye, Fei, and Hu, 1993 (68) and Vietnamese L. chapaense lineage 3 (67, 69). The
remaining lineages consisted of the Vietnamese species L. xanthospilum Lathrop, Murphy, Orlov, and Ho, 1998 (73), L. banae Lathrop, Murphy, Orlov, and Ho, 1998 (74), and V. ngoclinhensis Orlov, 2005 (72). 3.3. Divergence times Large degrees of overlap in divergence time estimations were observed in the confidence intervals (CI). Separation of Leptobrachium sensu lato and Oreolalax was estimated to have occurred 51.8 MYBP (CI 36.2–68.5) in the early Eocene (age adopted from Holloway and Hall (1998)). In the mid Eocene (50–42 MYBP), Leptobrachium split into Clades A and B (Table 4). Divergences of Subclades I–III and IV–V occurred nearly simultaneously from the late Eocene (42–35 MYBP) to the early Oligocene (35–29 MYBP). Subclade III, from Thailand and Myanmar, separated from the common ancestor of Subclades I and II from Sundaland and Peninsular Malaysia. Soon thereafter, Subclade I, consisting species mainly from Borneo, and Subclade II, encompassing species mainly from Peninsular Malaysia and Java, split. In Clade A, speciation was initiated in the late Oligocene, 30–23 MYBP when L. nigrops separated from L. hendricksoni and L. hasseltii. It continued in the early Miocene, 23–14 MYBP, with the following events: Leptobrachium sp. 2 on Mindanao split from the common ancestor of Bornean species and Sumatran Leptobrachium sp. 1; L. hendricksoni separated from L. hasseltii; Leptobrachium sp. 4 split from the common ancestor of L. smithi and Leptobrachium sp. 3; and L. montanum lineage 2 was isolated from the other Bornean and Sumatran populations. Subsequently, in the late Miocene, 14–5 MYBP, the common ancestor of L. montanum lineage 1 and Leptobrachium sp. 1 split from that of L. abbotti and L. gunungense. In addition, L. smithi split from Leptobrachium sp. 3 from Gwa, Myanmar, L. montanum lineage 2 split from Sumatran Leptobrachium sp. 1, and L. abbotti speciated from L. gunungense. In the main Indochinese and Chinese clade, diversification occurred from the early Miocene (L. banae and other Leptobrachium from Vietnam and Hainan in Subclade V; L. chapaense lineage 2 from Ha Gian and the remaining species of Subclade IV including
Table 4 Estimated divergence times (MY) of main divergences of Leptobrachium (sensu lato). Combination
Mean
CI
Leptobrachium/Oreolalax Leptobrachium Clade A/Clade B Leptobrachium Subclade IV/Subclade V Leptobrachium Subclades I + II/Subclade III Leptobrachium Subclade I/Subcclade II L. nigrops/L. hendricksoni + L. hasseltii L. banae/other Vietnamese species Leptobrachium sp. 2 (Philippines)/Bornean and Sumatran species L. hendricksoni/L. hasseltii Leptobrachium sp. 4/other Thai-Myanmar species L. chapaense lineagae 2/other species in Subclade IV Within L. nigrops L. xanthospilum/other Vietnamese species L. montanum lineage 2/other Bornean and Sumatran species L. abbotti + L. gunungense/L. montanum lineage 1 + Leptobrachium sp. 1 Leptobrachium sp. 3/L. smithi V. l. liui/V. promustache Within L. hendricksoni L. montanum lineage 1/Leptobrachium sp. 1 L. abbotti/L. gunungense Within L. montanum lineage 1 Within L. smithi Within L. hasseltii L. hainanense/L. chapaense lineage 3 (Tam Dao)
51.8 46.4 35.8 34.0 31.0 25.5 19.6 19.0
(36.2–68.5) (31.6–61.4) (21.1–49.4) (22.5–45.7) (20.7–42.5) (16.0–34.7) (10.3–29.6) (11.3–27.4)
18.8 16.6 16.4 15.8 15.7 14.2
(11.0–27.2) (8.6–25.4) (7.8–26.3) (8.2–24.1) (7.8–24.9) (8.6–21.0)
11.2
(6.5–16.6)
9.8 9.7 7.8 6.1 5.5 5.0 4.0 2.8 2.6
(4.8–15.7) (3.6–16.7) (4.0–12.2) (3.5–9.1) (2.0–9.9) (3.0–7.5) (2.0–6.4) (1.4–4.5) (0.7–4.7)
266
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
Vibrissaphora and L. chapaense lineage 1; L. xanthospilum and, other Vietnamese and Hainan Leptobrachium and V. ngoclinhensis) to the late Miocene (Chinese V. promustache and V. liui). Intraspecific divergence was suggested to have occurred as early as the late Miocene, as exemplified by L. hendricksoni. However, most of the intraspecific divergence occurred in the Pliocene (5.0–1.8 MYBP) as expected and seen in L. montanum lineage 1, L. smithi, and L. hasseltii. Divergence in Indochina and China also occurred in the Pliocene, as exemplified in L. hainanense and L. chapaense lineage 3. An exception to the recent diversification is seen in L. nigrops where the population from Sarawak split from populations in Peninsular Malaysia and Belitung in the early Miocene, and the latter populations diverged in the middle Miocene.
Leptobrachium sp. 1 (Subclade I) had totally blue iris in adults, although it was grey in a juvenile. In Clade B, a light-colored (ranging from white through sky-blue to lime green) upper iris predominated both in Subclade IV (V. liui, V. jiulongshanensis, V. leishanensis, L. chapaense [lineages 1and 2], V. boringii, V. ailaonica, V. echinata, and L. promustache) and in Subclade V (L. chapaense lineage 3, L. hainanense, V. ngoclinhensis, L. banae, and L. xanthospilum). As in Clade A, exceptions occurred. The upper part of the iris in L. mouhoti and L. pullum (Subclade V) was yellow, scarlet, or orange red. 4. Discussion 4.1. Taxonomic considerations of Clade A
3.4. Iris color In Clade A, a totally dark black iris predominated in Subclade I (L. montanum [lineages 1, 2], L. gunungense, L. abbotti and Leptobrachium sp. 2) and Subcalde II (L. hasseltii and L. nigrops) (Table 5). However, the upper part of the iris was yellow, orange, scarlet, or red in L. hendricksoni (Subclade II) and L. smithi (Subclade III). Furthermore, in L. hendricksoni, the iris color varied from having color only on the upper part to being totally orange (in one juvenile).
Subclade I consisted of species from Borneo, Sumatra, and Mindanao. The Philippine population (sample 20), which diverged at the base of this subclade, has been treated as being L. hasseltii (Taylor, 1922; Inger, 1954, 1966; Alcala, 1986). However, L. hasseltii occurs in Subclade II (see below). Therefore, the Philippine population constituted an unnamed species, herein referred to as Leptobrachium sp. 2. Morphological studies also suggested that the Philippine population was not conspecific with L. hasseltii
Table 5 Known variatin in the iris color of Leptobrachium (sensu lato). UN: Unnumbered. See text for voucher abbreviations. Species
Iris color
Locality
Voucher [Sample No.] and/or reference
L. abbotti
Completely black
Borneo
L. gunungense L. hasseltii
Completely black Completely black
Borneo Java
L. hendricksoni
Upper half orange
Sumatra Peninsular Malaysia
L. montanum lineage 1
Completely orange Upper half red Completely black
Peninsular Malaysia ? Borneo
L. montanum lineage 2 L. nigrops
Completely black Completely black
L. smithi
Upper half scarlet Upper half orange
Borneo Peninsular Malaysia Belitung Borneo Thailand Thailand
Upper half yellow Completely blue Completely grey Completely black Upper half white Upper half pale green Upper half blue Upper half sky-blue Upper half white Upper half white Upper half light-blue Upper half white Upper half light-blue Upper half sky-blue Upper outer margin orange red Upper half? scarlet Upper half yellow Upper half white Upper half light-blue Upper half light green Upper half bluish green Upper half lime green
Peninsular Malaysia Thailand Sumatra Sumatra Philippines Vietnam Laos China Thailand Vietnam Vietnam Vietnam China China China Cambodia Vietnam Vietnam Vietnam China China China Vietnam
BOR 04B018 [17]; BOR 04B110: BOR12866 [15]; KUHE 39294; KUHE 39295 [16]; KUHE 39296; SP 21669; SP 26559; SP 26590 BOR 22959 [13]; KUHE 39377 KUHE 42818 [33]; KUHE 42820 [34]; MZB UN; KUHE 42821; MZB UN [31]; UTA A53688 [32] KUHE 42807 [29]; KUHE 42808 [30]; KUHE 42809 ABTC 4081; UKM HC 110 [23]; KUHE 15008; KUHE 15336 [21]; KUHE 15680 [27]; KUHE 15756; KUHE 52150 [28]; KUHE 52403 [22]; MDK 10 [25] KUHE 52150 (Juvenile) Manthey and Grossmann (1997) BOR 04B019 [4]; BOR 22008 [3]; BOR 04B020; BOR 04B021; MDK UN; KUHE 17306 [7]; KUHE 42811; KUHE 42812 [2]; KUHE 42813 [5]; KUHE 42814; KUHE 42815 [6]; KUHE 42816; KUHE 42817; MDK 01 [9] BOR 08481 [18]; KUHE 39204; SP 21481 [19] KUHE 15430 [37]; KUHE 15658 [38]; KUHE 15706; KUHE 35433; UKM HC 072 [39] MZB Amp1790 [35]; MZB Amp1791 [36] KUHE 081122-1; KUHE 42587 [40]; KUHE 42590 [41] KUHE 19281 [42]; KUHE 19282 [43] KUHE 19393; KUHE 19394; KUHE 19395; KUHE 19518; KUHE 19538; KUHE 19832; KU; KUHE 23318HE 19834 [44]; KUHE 19937; KUHE 19938; KUHE 19939 [45]; KUHE 20000; KUHE 20032; KUHE 23342 [48] UM D0136; UM D0139 [49] KUHE 20200 [46]; KUHE 20201 [47] MZB Amp 15862 [11]; UTA A53689 [12] KUHE 42805 [10](Juvenile) Brown and Diesmos (2009) ROM 32200 [74]; Lathrop et al. (1998), Nguyen et al. (2009) Ohler et al. (2004) Yang (1991) KUHE 19122 [59]; Dubois and Ohler (1998) Bain and Nguyen (2004) ROM 32176 [67]; Lathrop et al. (1998) Dubois and Ohler (1998) Fei et al. (2009) Fei (1999) Fei et al. (2009) Stuart et al. (2006), Nguyen et al. (2009) Smith (1921) Nguyen et al. (2009) Lathrop et al. (1998), Nguyen et al. (2009) Fei (1999), Fei et al. (2009) Fei (1999) Fei et al. (2009) Dubois and Ohler (1998)
Leptobrachium sp. 1 Leptobrachium sp. 2 L. banae L. buchardi L. chapaense lineage 1 L. chapaense lineage 2 L. chapaense lineage 3 L. L. L. L. L.
guangxiense hainanense huashen mouhoti pullum
L. xanthospilum V. ailaonica V. boringii V. echinata
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
267
Table 5 (continued) Species V. jiulongshanensis V. l. liui V. l. yaoshanensis
V. leishanensis V. ngoclinhensis V. promustache
Iris color
Locality
Voucher [Sample No.] and/or reference
Upper Upper Upper Upper Upper Upper Upper Upper Upper
Vietnam China China China China China China Vietnam China, Vietnam
Ho et al. (1999) Fei et al. (2009) Fei (1999), Fei et al. (2009) Fei (1999) Fei et al. (2009) CIB JX20071196 Fei (1999), Fei et al. (2009) Orlov (2005), Nguyen et al. (2009) Rao et al. (2006), Fei et al. (2009), Bain et al. (2009)
half blue half light greenish white half light green half light green half light greenish white half light-blue half light green half white half light-blue
(Inger, 1954; Inger et al., 1995; Dubois and Ohler, 1998). Because populations from the various islands of the Philippines might differ from the Mindanao population (Inger, 1954), further study is necessary to determine this population’s taxonomic status. Frogs morphologically identified as L. montanum are paraphyletic, and populations from the Kinabalu (19) and Crocker (18) regions of western Sabah (L. montanum lineage 2) form a lineage sister to the clade containing L. montanum from the other parts of Borneo (L. montanum lineage 1) (Fig. 1). Morphologically, L. montanum lineage 2 and L. gunungense from Kinabalu (13, 14) are difficult to distinguish, but their calls differ substantially (Malkmus, 1996; Malkmus et al., 2002; Matsui, unpublished data). Because Malkmus (1996) mainly worked on Kinabalu, he likely studied L. montanum lineage 2. Given an uncorrected p-distance of 5.7–5.8% between L. montanum lineage 2 and L. gunungense, and their different calls, these lineages undoubtedly constitute distinct species. This position is further bolstered by the genealogical relationships because L. gunungense formed a sister group not with L. montanum lineage 2 but with L. abbotti (15–17), a result not expected because L. gunungense is morphologically much more dissimilar to L. abbotti than to L. montanum lineage 2. Specimens from the type locality of L. montanum, Paramasan, southern Kalimantan (1, 2), nested in L. montanum lineage 1 and they are regarded as being true L. montanum. The lineage also contains specimens from eastern Sabah (3, 4), western Sarawak (7, 8), central Kalimantan (5, 6), and western Kalimantan (9). Further morphological and bioacoustic investigations are needed to determine taxonomic identity these populations. A sister lineage, Leptobrachium sp. 1 from Jambi (12) and Lampung (10, 11), Sumatra, surely forms an unnamed species. In these individuals, the iris is blue in color, though grayish in juveniles. This condition is very distinct among Sundaland species of Leptobrachium (Hamidy and Matsui, 2010). Within Subclade II, L. nigrops, the sister species to the monophyletic clade of L. hendricksoni and L. hasseltii, is characterized by having the smallest body size of all Leptobrachium and uniquely pointed finger and toe tips (Inger, 1966; Inger and Stuebing, 1997). Populations from Peninsular Malaysia (37–39) and Belitung Island (35, 36) are sister populations and together they form the sister lineage of the western Bornean (40, 41) population. Genetic differentiation among its allopatric populations is the greatest among all species. Populations from Peninsular Malaysia, Belitung Island, and Borneo have uncorrected p-distances ranging from 9.3% to 11.7%. Their geographical distinctiveness and level of genetic differentiation strongly suggest that each population deserves species recognition. Inger (1966) notes minor morphological differences between Peninsular Malaysian and Bornean populations. Further morphological and acoustic studies will likely verify their taxonomic distinctiveness. Leptobrachium hendricksoni is distinct in having black spots on its venter and an orange-colored upper (and sometimes whole) iris (Taylor, 1962; Inger, 1966; Matsui et al., 1999; Malkmus et al., 2002). On Peninsular Malaysia, this species is paraphyletic with respect to Sumatran populations. Interestingly, populations from the
northern part of Peninsular Malaysia (21–24) are the sister group to the Sumatran populations (25, 26) and together these form the sister group to the southern peninsular populations (27, 28). Genetic diversities in this species are much smaller (uncorrected p-distances <2.3%) than those in L. nigrops, suggesting a unique evolutionary history. In this case, taxonomic recognition of these lineages may not be warranted. Historically, the name L. hasseltii was applied to almost all populations of Leptobrachium (formerly as Megalophrys or Megophrys) ranging from southern China, Indochina, Thailand to northern India, and through Sundaland to the Philippines (e.g., Smith, 1921; Bourret, 1937; Inger, 1954, 1966; Liu and Hu, 1961; Taylor, 1962; Berry, 1975). Matsui (in Frost, 1985) split several populations as distinct species, although some authors have not recognized these changes (e.g., Alcala, 1986; Chanda, 1994; Zheng et al., 2008). The topotypic populations from Java (31–34) and the population from Lampung, southern Sumatra (29, 30) were monophyletic. Here, we have confirmed the occurrence of L. hasseltii on Sumatra and, based on specimen collection records in MZB, the species seems to be restricted to this area. Central and northern Sumatra is occupied by L. hendricksoni. The range of L. hasseltii seems limited to Java, Sumatra, Bali, and several adjacent small islands (Iskandar, 1998). The iris color of L. hasseltii is said to be scarlet (Iskandar, 1998), as cited by Matsui et al. (1999). However, all of our samples, including those used herein, and L. nigrops, have a totally black iris, unlike its sister species L. hendricksoni. Subclade III contained L. smithi from various localities of Thailand, Myanmar (Kyaik Hto) and Malaysia (Langkawi Island), Leptobrachium sp. 3 from Myanmar (Gwa), and Leptobrachium sp. 4 from Thailand (Pilok). Within L. smithi, populations from northeastern Thailand (42, 43), western Thailand (44–47), southern Thailand (48 = topotypic sample from Kaochong), including Langkawi Island (49), and Myanmar (50), formed a clade, but the relationships among them remained unresolved. Intraspecific variation in body size and iris color noted by Matsui et al. (1999) was partially confirmed by us, but these populations did not exhibit significant genetic differentiation between them. Populations from the southern and northern Isthmus Kra occurred in a well-known faunal and floral boundary in Sundaland (e.g., Tougard, 2001; De Bruyn et al., 2005). Leptobrachium sp. 3 from Gwa, Myanmar (51) was the sister lineage to L. smithi. The data used in our analysis (DQ283239) were said to be from L. hasseltii (Frost et al., 2006) yet we could not examine the voucher specimen. This sequence was more likely to have been from L. smithi rather than L. hasseltii. Large genetic uncorrected p-distance (6.0%) with L. smithi from another locality of Myanmar suggested that Leptobrachium sp. 3 represented an undescribed species. Finally, Leptobrachium sp. 4 from Pilok, western Thailand (52) occurred sympatrically with L. smithi (45) and the genetic differentiation between them was great (uncorrected p-distance = 8.1%). Inthara et al. (2005) reported that the larval denticle formula of Leptobrachium sp. from Thon Pha Phum, Kanchanaburi, which is located very close to Pilok varied from 6(2–6)/7(1–6) to 8(2–8/7(1–6) (terminology following Altig and McDiarmid, 1999). The denticle formula for L. smithi ranged
268
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
from 5(2–5)/5(1–4) to 7(2–7)/6(1–5) (Matsui et al., 1999). Unfortunately, we did not have a specimen of Leptobrachium sp. 4. Adult specimens are needed to determine this population’s taxonomic status. 4.2. Taxonomic considerations of Clade B In Subclade IV, Clade B, L. chapaense lineage 2 from Ha Giang, northern Vietnam (66) was the sister group of the other lineages. The population has been suggested to be an undescribed species (Zheng et al., 2008), and our results support this view. In the sister lineage to L. chapaense lineage 2, V. promustache (65) was the sister species to the clade encompassing five remaining lineages. These relationships are completely concordant with those of Zheng et al. (2008). One lineage contained three Chinese sublineages: first, V. jiulongshanensis from Zhejiang Province (56); second, V. leishanensis from Guizhou Province (55); and third, V. l. liui from Mangshan, Hunan Province (53) and V. l. yaoshanensis from Guangxi Province (54). Our results indicated that these four taxa were close genetically to each other (uncorrected p-distances ranged from 1.5% between V. l. liui and V. l. yaoshanensis to 3.0% between V. l. liui and V. jiulongshanensis). Zheng et al. (2008) recovered these three lineages but concluded that V. jiulongshanensis was a junior synonym of V. liui, incorrectly reflecting their phylogeny. Rao and Wilkinson (2008) also obtained the three lineages but found that V. l. liui formed a clade with V. jiulongshanensis and not with V. l. yaoshanensis. Thus, they synonymized V. jiulongshanensis into V. liui, but recognized V. yaoshanensis as a full species. We suggest that additional work is required to confirm the taxonomic status of these taxa. Rao and Wilkinson (2008) proposed the formation of a V. boringiae species group comprised of V. boringiae (=boringii) (61), V. leishanensis, V. l. liui, V. l. yaoshanensis, and V. jiulongshanensis. We did not resolve this group as a monophyletic lineage, and the association of these species was only weakly supported in Rao and Wilkinson’s trees. Our tree supported Fei et al.’s (2009) V. liui species group including V. leishanensis, V. l. liui, V. l. yaoshanensis (V. jiulongshanensis in the synonymy of V. l. liui). However, our analysis did not support their V. boringii species group as including V. boringii, V. ailaonica (64), and V. promustache. Similarly, Rao and Wilkinson’s (2008) V. ailaonica species group, including V. ailaonica, V. echinata (63) and V. ngoclinhensis (72), was not supported; our results, and those of Zheng et al. (2008), did not nest V. ngoclinhensis (Subclade V) with V. ailaonica and V. echinata (Subclade IV). Certainly, V. ailaonica and V. echinata are sister species (Dubois and Ohler, 1998; Ohler et al., 2000; Grosjean, 2001), but there remains an argument as to whether they are conspecific (Ho et al., 1999; Zheng et al., 2008) or not (Ohler et al., 2000; Grosjean, 2001; Rao and Wilkinson, 2008). The uncorrected p-distance between them was only 2.5%, the threshold of specific separation for other taxa. This low level of divergence could reflect isolation-by-distance. Leptobrachium chapaense lineage 1 from Thailand (57, 59) and Yunnan, China (58, 60) and Leptobrachium sp. 5 from Laos (62) comprise the remaining two lineages of Subclade IV. Wide-ranging L. chapaense is surely a composite of more than one species (Rao et al., 2006; Zheng et al., 2008; Rao and Wilkinson, 2008). Our L. chapaense lineage 1 is the sister lineage of the population from Sa Pa, Vietnam, the type locality of L. chapaense (Zheng et al., 2008) and likely it is conspecific. Leptobrachium chapaense clusters within Vibrissaphora and this finding conforms to the morphology of the tadpoles from Yunnan; Yang (1991) notes that Yunnan populations have a Y-shaped yellow marking on their tail, a characteristic of Vibrissaphora (Dubois and Ohler, 1998). Our single male specimen of Leptobrachium sp. 5 from Xam Neua, northeastern Laos likely represents an undescribed species.
Unfortunately, specimens of L. buchardi Ohler, Teynié, and David, 2004 from southern Laos, which has a green colored upper iris, as ascertained from one female specimen, were not available for comparison. The iris color in our male was not recorded and it is larger than the female suggesting, on one hand, the possibility of their conspecificity because of male-biased sexual size dimorphism in Vibrissaphora (Dubois and Ohler, 1998). On the other hand, given that the two specimens were collected more than 600 km in straight-line distance apart, conspecificity seems unlikely. It is also possible that our specimen is conspecific with L. huashen Fei and Ye, 2005 from Yunnan Province, China. However, L. huashen is closely related to L. chapaense (Fei and Ye, 2005; Rao and Wilkinson, 2008). Likewise, L. guangxiense Fei, Mo, Ye, and Jiang in Fei, Hu, Ye, Huang et al. (2009) from Shangsi, Guangxi Province, China was not available. However, based on the original description (Fei et al., 2009), the species seems to be a member of Subclade IV of our study. Thus, we believe it more likely than not that Leptobrachium sp. 5 represents an undescribed species. Two populations of L. chapaense lineage 3 from northern Vietnam (67, 69) and L. hainanense (68) formed a monophyletic lineage, as previously reported (Zheng et al., 2008). Lathrop et al. (1998) noted that the dorsal half of iris is white in specimens from the Tam Dao (67) population and not light-blue as in L. hainanense (Liu et al., 1973 as V. hasseltii; Matsui and Ota, 1995; Fei et al., 2009; see below). We, however, regard this difference to be within the normal range of variation, and tentatively conclude that the population from Tam Dao, as well as Ben En, should be treated as L. hainanense. The uncorrected p-distance between these two localities and L. hainanense from Hainan Island were 1.8% and 3.0%, respectively. The latter value is comparable to the distance (3.0%) found between L. abbotti and L. gunungense from Borneo. Further work is necessary to confirm taxonomic relationships and identity of these populations. Unlike the result of Zheng et al. (2008), our study resolved Vietnamese L. mouhoti (70) as the sister species of L. pullum (71) and not V. ngoclinhensis (72). As with Zheng et al.’s (2008) study, V. ngoclinhensis fell in Subclade V and this association differs from Rao and Wilkinson’s (2008) suggestion that the species is associated with Vibrissaphora (Clade IV) based on morphological data. Although L. pullum is morphologically similar to L. smithi (Matsui et al., 1999), these two species were distantly related genealogically, being placed in Subclades V and III, respectively. Finally, Vietnamese L. xanthospilum (73) and L. banae (74) represented the remaining two lineages in Subclade V and our study supported their specific status (Lathrop et al., 1998). 4.3. Taxonomic relationship of Leptobrachium and Vibrissaphora Our results for the relationships of Leptobrachium and Vibrissaphora were concordant with other recent molecular studies (Fu et al., 2007; Zheng et al., 2008; Rao and Wilkinson, 2008). As summarized by Dubois and Ohler (1998), taxonomic recognition of Leptobrachium and Vibrissaphora as genera, subgenera, or synonyms is controversial. Some authors (e.g., Fei et al., 2009) still treat them as different genera, while the others (e.g., Dubois and Ohler, 1998; Ohler et al., 2004; Delorme et al., 2006) regard them as subgenera of Leptobrachium. Zheng et al. (2008) rejected the monophyly of both subgenera and relegated Vibrissaphora to synonymy in Leptobrachium without recognizing subgenera, as originally proposed by Dubois (1983). The taxonomic conclusions of Zheng et al. (2008), now followed by Frost (2009), requires reassessment because their result seems to have been affected by inadequate sampling in Sundaland. In all of our trees, the Sundaland species plus L. smithi (Clade A, Subclades I–III) were distinctly separated from Chinese and Indochinese species (Clade B, Subclades IV and V). Clades A, B, and C in Fig. 3 of
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
269
surveyed. Although detailed morphological study is necessary, a more broadened and flattened male upper lip of Leptobrachium (Vibrissaphora) may be a synapomorphy, and a character state that separates them from Leptobrachium (Leptobrachium), as observed between L. pullum and L. smithi (Matsui et al., 1999). The presence of reversed sexual size dimorphism (Dubois and Ohler, 1998) and a Y-shaped yellow marking on the tail of tadpoles should be also reassessed using a sufficient number of samples. 4.4. Evolutionary history of Leptobrachium
Fig. 3. Iris color in frogs of the genus Leptobrachium mapped onto the maximum likelihood tree given in Fig. 2. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this paper.)
Zheng et al. (2008) corresponds to our Subclades IV, V, and I + III, respectively. They found either, Subclades V and I + III formed a monophyletic group, which was sister to Subclade IV (MP tree), or these three lineages were trichotomic (BI tree). Our results rejected both sets of relationships, and hence their conclusion. In order for the taxonomy to reflect historical relationships, we recognize Clades A and B as subgenera. Because the current megophryid taxonomy tends to recognize many genera that are neither sufficiently studied phylogenetically nor defined clearly morphologically, we refrain treating them as genera. Subgenus Leptobrachium (Clade A; type species L. hasseltii) can be characterized as the most inclusive clade containing species from the Philippines, Malaysia, Indonesia, Thailand, Myanmar, and India, but exclusive of L. chapaense lineage 1 from Thailand. Subgenus Vibrissaphora (Clade B; type species V. boringii) is the most inclusive clade that contains species from Indochina and China, and including L. (V.) chapaense lineage 1 from Thailand. Traditionally, male Vibrissaphora have been regarded as possessing nuptial spines on their upper labium during the breeding season and as having larger body sizes than females. Previously, these characteristics served to separate Vibrissaphora from Leptobrachium (see review by Dubois and Ohler (1998)). However, the character labial spines no longer defines the group. The species L. banae, L. chapaense (for three lineages), L. guangxiense, L. hainanense, L. mouhoti, L. pullum, L. xanthospilum, and Leptobrachium sp. 5 grouped in Clade B. In our new classification, these species, although lacking labial spines, should be placed in Leptobrachium (Vibrissaphora) together with V. ngoclinhensis. In this arrangement, the morphological diagnosis of Vibrissaphora is unclear and new characteristics should be
Divergence time estimations can provide valuable insights into the biogeographic history of species and regions. Often the estimated dates have extensive variance and overlap, yet they can serve as biogeographical hypotheses in the absence of detailed geological information, and can even predict unknown geological events (Lindell et al., 2006; Riddle et al., 2008). Time estimates (Table 4) suggested that Leptobrachium (sensu lato) and Oreolalax formed in the early Eocene, and subsequently diverged into two major lineages (Clades A and B) in the mid Eocene. No stratigraphic events have been associated with the origin of Clades A and B and the subsequent, nearly simultaneous divergences within them starting from the late Eocene to early Oligocene. However, the uplift of the Tibetan Plateau initiated about 40 MYBP (Zhao and Morgan, 1985; Chung et al., 1998) and this caused changes in the Mekong drainage system (Brookfield, 1998). This orogenesis is likely related to the speciation events. Estimates for the age of Sundaland Leptobrachium predate the geological separation of land masses in this region. For example, the Sunda Shelf became land-positive, forming a continuous land connection between Java and mainland Southeast Asia, including some parts of Indochina, during periods of lowered sea levels in the Pleistocene (<1.6 MYBP; Voris, 2000)). This very recent stratigraphic history is responsible for the recent cladogenic events as suggested by low levels of intraspecific genetic differentiation in Leptobrachium but it cannot account for older events. Most of the speciation is estimated to have occurred sometime between the late Oligocene to late Miocene (30.0–14.0 MYBP) in Clade A. The separation of L. nigrops from its common ancestor with L. hendricksoni and L. hasseltii likely occurred in the late Oligocene, and speciation of the latter two species in the early Miocene. These ages are comparable with those estimated for Clade B. The diversification of L. banae from other Vietnamese and Hainan Leptobrachium, and for Ha Gian L. chapaense lineage 2 from its sister taxa, likely occurred in the early Miocene. Leptobrachium sp. 2 from Mindanao, Philippines was estimated to have diverged from Bornean and Sumatran relatives at 19.0 (CI 27.4–11.3) MYBP. This estimate does not contradict the idea that Mindanao became aerial about 25–30 MYBP in the late Oligocene (Brown and Alcala, 1994), but the route of faunal invasion remains unclear (Inger, 1999). Inger and Voris (2001) showed that disjunction of the Malay Peninsula from Borneo by the South China Sea (Fig. 1) occurred once about 5 MYBP in the early Pliocene. This age corresponds to the date of intraspecific divergence estimated in L. hendricksoni for populations in the northern Malay Peninsula and Sumatra. Although specimens from Sarawak, western Borneo (Inger, 1966) were not available, separation from the peninsular population is expected at this age. ‘‘Intraspecific” divergence in L. nigrops is much older than that observed in the other species (Table 4). The mid Miocene date is comparable to or even older than that in other lineages (see below). Inger (2005) surmised that the long-term continental connection between Borneo and the Malay Peninsula (50–5 MYBP) was once interrupted in the mid Miocene, 15 MYBP. This event might have isolated Bornean L. nigrops from Malay Peninsular
270
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
and Belitung populations. Notwithstanding their morphological similarity, each of the populations currently called L. nigrops is likely a distinct species. The occurrence of L. nigrops on Belitung Island suggests the possibility of ‘‘land bridge” invasions by Leptobrachium between Mainland Asia and Borneo. Leptobrachium hendricksoni and L. nigrops are lowland species that also occur in coastal swamps. Their ecology led Inger and Voris (2001) to associate them with Pleistocene dispersal between the Malay Peninsula and Borneo, while questioning their absence on Sumatra. The presence of L. hendricksoni on Sumatra is now confirmed, and its unique ecology does not seem to be related to Pleistocene dispersal. Leptobrachium hasseltii occurs at slightly higher elevations than does L. hendricksoni. On Java and southern Sumatra, we found L. hasseltii at altitudes of 300–1500 m a.s.l. However, this altitudinal distribution may have been secondarily acquired. Within Java, most lowland forests have been logged and no longer exist. Only a few mountain forests remain today in the form of isolated habitat islands without corridors between them. This condition probably has prevented gene flow between populations from western and central Java. Diversification in L. hasseltii is estimated to have occurred in the mid Pliocene (Table 4). Available information on distributional pattern of this species is limited, but based on collection records in MZB, the westernmost distribution of L. hasseltii is the southern tip of Sumatra (Lampung), and the easternmost range includes the islands of Bali and Kangean (Iskandar, 1998), which is the southeasternmost range of the genus. The Lombok Strait, i.e., Wallace’s Line, separating Bali and Lombok, seems to have been the primary barrier preventing the invasion of Leptobrachium to Wallacea. Parapatric species of Leptobrachium occur on Borneo, where speciation occurred mostly in the Miocene, starting with the mid Miocene separation of L. montanum lineage 2 from the common ancestor of the other species. Speciation continued with the split of ancestors of L. montanum lineage 1 and Leptobrachium sp. 1, and slightly later the common ancestor of L. abbotti and L. gunungense. This was followed by speciation of L. abbotti and L. gunungense in the late Miocene. Intraspecific divergence occurred at the Miocene–Pliocene boundary as shown by L. montanum lineage 1. These very old dates of divergence in Borneo, especially on Mt. Kinabalu, are comparable to those of bufonid toads in the genus Ansonia (Matsui et al., 2010). The boundary of upland L. montanum (two lineages combined) and lowland L. abbotti occurs at elevations from 800 to 1000 m a.s.l. (Inger et al., 1995). On Mt. Kinabalu, a narrow zone of sympatry occurs between L. gunungense (1750–2200 m) and L. montanum (900–1750 m), most probably our lineage 2 (Malkmus et al., 2002; our own observation). The third species, L. abbotti, occurs at much lower elevations (500–900 m). Like the species in Borneo, Leptobrachium sp. 1 was found parapatric with L. hasseltii (Hamidy and Matsui, 2010). Interestingly, these sets of species are from different lineages. Non-sister lineages may more easily segregate their niches than the sister taxa, which might be more similar in niche requirements. Nearly simultaneous dates of divergence were estimated in the predominantly Thai clade, where splits occurred between Leptobrachium sp. 4 and the common ancestor of L. smithi and Leptobrachium sp. 3 in the early Miocene. Another speciation event occurred between L. smithi and Leptobrachium sp. 3 in the late Miocene. Subsequent intraspecific divergence in L. smithi was estimated to have begun in the early Pliocene. The Isthmus of Kra, peninsular Thailand, is usually noted as a major biogeographical transition zone (e.g., Tougard, 2001; De Bruyn et al., 2005), and it might have also affected the evolution of Leptobrachium. However, the specific date(s) of the disjunction is (are) not clear and simply estimated to have occurred as a result of a Neogene (<23 MYBP) marine transgression.
Diversification in the Indochinese–Chinese area initiated in the Miocene (Table 4). In the Indochinese clade, L. xanthospilum split from the common ancestor of other Vietnamese and Hainan Leptobrachium and V. ngoclinhensis in the early Miocene. In the Chinese clade, V. promustache and V. liui diverged in the late Miocene. Intraspecific divergence in these regions occurred mostly in the Pliocene. Our estimated date of separation between L. hainanense and Vietnamese L. chapaense lineage 3, 2.6 (CI 0.7–4.7) MYBP, closely conforms to the date of separation of these regions in the early Pleistocene, about 2 MYBP (Chang et al., 2008). In the Indochinese clade, the keratinized nuptial spines on the upper jaw of males seem to have been acquired after the middle Miocene, 15.7 (CI 7.8–24.9) MYBP by V. ngoclinhensis only. In the Chinese clade, the spines seem to have evolved in the late Miocene, 9.7 (CI 3.6–16.7) MYBP, after the basal split of L. chapaense lineage 2, which lacks spines. Subsequently, the spines were secondarily lost in the ancestor of L. chapaense (=lineage 1). As indicated by the different patterns of spine arrangement, labial spines likely have evolved independently in Chinese and Indochinese (V. ngoclinhensis: Orlov, 2005; Rao et al., 2006) clades. Each subclade is estimated to have an old evolutionary history. Divergence times are not easily explained by the geohistory of Southeast Asian major landmasses alone. In order to test hypotheses posed here, additional detailed paleogeographic evidence is necessary.
4.5. Evolution of iris color Although the data are limited, the color of the iris seems to vary ontogenetically (Table 5). The iris color of L. hendricksoni in Malay Peninsular populations varies from being totally orange to having color only on the upper part. This appears to be intraspecific variation as individuals from these populations are monophyletic and conspecific. Similarly, in Leptobrachium sp. 1, juvenile grey iris transforms to blue in adults. Further, variation in iris color ranges from being light yellow to orange in L. smithi (Matsui et al., 1999), and specimens with differently colored iris are also judged genetically to be conspecific. The color of the upper part of the iris in Indochinese–Chinese species (Clade B in this study) has been variously described as being white and sky-blue, as well as skyblue and lime, and lime and green (see Table 5). This continuous range of color could be grouped as light-colored upper iris. Although iris color varies intraspecifically, the evolution of the character can be estimated by the mapping generalized states in adults on the phylogeny (Fig. 3). Matsui et al. (1999) once surmised that the light-colored upper iris might be the primitive condition in Leptobrachium, but their hypothesis was not confirmed in this study. From the distribution of iris color on the tree (Fig. 3), a totally black iris is the plesiomorphic state within Clade A. Within Clade B, the plesiomorphic state is one of having a light color in the upper half of the iris. The yellow to scarlet color in the upper half of the iris, and sometimes also in the lower half, is regarded as being apomorphic, having evolved independently in three subclades. A totally light-colored iris also appears to be an apomorphic character in Clade A. Because no species of the sister genus Oreolalax is known to have a bicolored or totally dark iris (Liu and Hu, 1961; Zhao and Adler, 1993; Fei, 1999; Yang and Rao, 2008), the plesiomorphic state in Leptobrachium cannot be identified. Although iris color surely varies, even within a species, the range of variation seems to be limited. Thus, the character appears to be useful for species identification and in estimating phylogenetic relationships. In order to test hypotheses on the evolution of iris color posed here, data are required from taxa whose character states remain unknown.
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
Acknowledgments MM is grateful to the following for their encouragements and/or permission to conduct research and field companionship: H. Akiyama, L. Apin, K. Araya, A.-A. Hamid, T. Hikida, H. Ota, H. Kassim, J.J. Kendawong, K.B. Kueh, T. Kusano, D. Labang, M.B. Lakim, M. Maryati, the late J. Nabitabhata, K. Nishikawa, S. Panha, L.-H. Seng, A. Sudin, T. Sugahara, T. Tachi, M. Toda, and N.-S. Wong. MM is also indebted to A. Ohler and M. Delorme for exchanging tissue samples, and N. Kuraishi and N. Yoshikawa for laboratory assistance. We also thank two anonymous reviewers for improving an early version of the manuscript. The National Research Council of Thailand, the Royal Forest Department of Thailand, the Economic-Planning Unit (former Socio-Economic Research Unit) of Malaysia, the State Government of Sarawak, and Sabah Parks kindly permitted MM to conduct the project, and Chulalongkorn University, University Malaya, Universiti Malaysia Sabah, Universiti Kebangsaan Malaysia (UKM), JICA, and the Forest Department, Sarawak kindly provided all the facilities for conducting research. Field trips by MM were made possible by grants from The Monbusho International Scientific Research Program (Field Research, 01041051, 02041051, 04041068, 06041066, and 08041144), The Monbukagakusho through the Japanese Society for the Promotion of Sciences (JSPS: Field Research, 15370038, 20405013), UKM (OUPPLW-14-59/2008), and TJTTP-OECF. AH thanks M.D. Kusrini, S. Kirono, G.B. Mandigani, D. Susanto, M. Harvey, E. Smith, K.L. Sanders, T.Q. Nguyen, D.A. Anggraeni, and J. Rhamadani for providing tissue samples and to the Monbukagakusho for scholarship funding. RWM Funding for fieldwork was obtained through grants from the Natural Sciences and Engineering Research Council, Canada (Discovery Grant A3148), the Royal Ontario Museum (ROM) Foundation, and the ROM Volunteers Fieldwork Committee. All fieldwork was conducted using approved Animal Use Protocols. Fieldwork in Vietnam was assisted by N.L. Orlov, A. Lathrop, R.H. Bain, V.S. Cao, T.C. Ho and V.S. Nguyen; export permits were obtained with the assistance of V.S. Cao and the Institute of Ecology and Biological Resources, Hanoi. Specimens in Hainan Island were obtained while assisting H.-T. Shi, Hainan Normal University, and those from Yunnan were obtained with the assistance of D.-Q. Rao, Kunming Institute of Zoology and N.L. Orlov; research export permits were obtained through the Chinese Academy of Sciences.
References Alcala, A.C., 1986. Guide to Philippine Flora and Fauna 10, Amphibians and Reptiles. Xiv+195 p. Natural Resources Management Center, Ministry of Natural Resources and University of the Philippines, Quezon City. Altig, R., McDiarmid, R.W. (Eds.), 1999. Body plan: development and morphology. In: Tadpoles. University of Chicago Press, Chicago, pp. 24–51. Bain, R.H., Nguyen, T.Q., 2004. Herpetological diversity of Ha Giang province in northeastern Vietnam, with description of two new species. Am. Mus. Novit. 3452, 1–42. Bain, R.H., Nguyen, T.Q., Doan, K.V., 2009. First record of Leptobrachium promustache from Vietnam. Herpetol. Notes 2, 27–29. Berry, P.Y., 1975. The Amphibian Fauna of Peninsular Malaysia. Tropical Press, Kuala Lumpur. Bourret, R., 1937. Notes herpétologiques sur l’Indochine Française XIV–XV. Annexe Bull. Gén. Instr. Publique, Hanoi 1937, 1–80. Brookfield, M.E., 1998. The evolution of great river systems of southern Asia during the Cenozoic India–Asia collision: rivers draining southwards. Geomorphology 22, 285–312. Brown, R.M., Diesmos, A.C., 2009. HerpWatch Philippines: an online biodiversity information product for Philippine herpetological diversity. Available from:
271
Cannatella, D.C., Hillis, D.M., Chippindale, P.T., Weght, L., Rand, A.S., Ryan, M.J., 1998. Phylogeny of frogs of the Physalaemus pustulosus species group, with an examination of data incongruence. Syst. Biol. 47, 311–335. Chanda, S.K., 1994. Anuran (Amphibia) fauna of northeast India. Mem. Zool. Surv. India 18, 1–143. Chang, J., Wang, B., Zhang, Y.-Y., Liu, Y., Liang, W., Wang, J.-C., Shi, H.-T., Su, W.-B., Zhang, Z.-W., 2008. Molecular evidence for species status of the endangered Hainan peacock pheasant. Zool. Sci. 25, 30–35. Chung, S.L., Lo, C.H., Lee, T.Y., Zhang, Y.Q., Xie, Y.W., Li, X.H., Wang, K.L., Wang, P.L., 1998. Diachronous uplift of the Tibetan plateau starting 40 Myr ago. Nature 394, 769–773. De Bruyn, M., Nugroho, E., Hossain, Md.M., Wilson, J.C., Manther, P.B., 2005. Phylogeographic evidence for the existence of an ancient biogeographic barrier: Isthmus of Kra Seaway. Heredity 94, 370–378. Delorme, M., Dubois, A., Grosjean, S., Ohler, M., 2006. Une nouvelle ergotaxinomie des Megophryidae (Amphibia, Anura). Alytes 24, 6–21. Drummond, A.J., Ho, S.Y.W., Philips, M.J., Rambaut, A., 2006. Relaxed phylogenetics and dating with confidence. PLoS Biol. 4, 699–710. Drummond, A.J., Rambaut, A., 2007. BEAST, Bayesian Evolutionary Analysis Sampling Trees, version 1.4.2. Available from:
272
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
Inger, R.F., 1999. Distribution of amphibians in Southern Asia and adjacent islands. In: Duellman, W.E. (Ed.), Patterns of Distribution of Amphibians: A Global Perspective. Johns Hopkins Univ. Press, Baltimore and London, pp. 445–482. Inger, R.F., 2005. The frog fauna of Indo-Malayan region as it applies to Wallace’s line. In: Tuen, A.A., Das, I. (Eds.), Wallace in Serawak-150 years later, an international conference on biogeography and biodiversity. Institute of Biodiversity and Environmental Conservation. Universiti Malaysia Sarawak, Kota Samarahan, pp. 82–90. Inger, R.F., Stuebing, R.B., 1997. A Field Guide to the Frogs of Borneo. Nat. Hist. Publ. Sdn. Bhd, Kota Kinabalu. Inger, R.F., Voris, H., 2001. The biogeographical relations of the fogs and snakes of Sundaland. J. Biogeogr. 28, 863–891. Inger, R.F., Stuebing, R.B., Tan, F.-L., 1995. New species and new records of anurans from Borneo. Raffles Bull. Zool. 43, 115–131. Inthara, C., Lauhachinda, V., Nabhitabhata, J., Chuaynkern, Y., Kumtong, P., 2005. Mouth part structures and distribution of some tadpoles from Thailand. Thailand Nat. Hist. Mus. J. 1, 55–78. Iskandar, D.T., 1998. The Amphibians of Java and Bali. Puslitbang Biologi, LIPI, Bogor. Jobb, G., von Haeseler, A., Strimmer, K., 2004. Treefinder: a powerful graphical analysis environment for molecular phylogenetics. BMC Evol. Biol. 4, 18. Lathrop, A., 1997. Taxonomic review of the megophryid frogs (Anura: Pelobatoidea). Asiatic Herpetol. Res. 7, 68–79. Lathrop, A., Murphy, R.W., Orlov, N.L., Ho, C.T., 1998. Two new species of Leptobrachium (Anura: Megophryidae) from the central highlands of Vietnam with a redescription of Leptobrachium chapaense. Russ. J. Herpetol. 5, 51–60. Leaché, A.D., Reeder, T.W., 2002. Molecular systematics of the eastern fence lizard (Sceloporus undulatus): a comparison of parsimony, likelihood, and Bayesian approaches. Syst. Biol. 51, 44–68. Lindell, J., Ngo, A., Murphy, R.W., 2006. Deep genealogies and the mid-peninsular seaway of Baja California. J. Biogeogr. 33, 1327–1331. Liu, C., Hu, S., 1961. Tailless Batrachians of China. Science Press, Beijing. Liu, C.-C., Hu, S.-C., Fei, L., Huang, C.-C., 1973. On collections of amphibians from Hainan Island. Acta Zool. Sin. 19, 385–404. Malkmus, R., 1996. Leptobrachium gunungense sp. N. (Anura: Pelobatidae) vom Mount Kinabalu, Nord Borneo. Mitt. Zool. Mus. Berlin 72, 297–301. Malkmus, R., Manthey, U., Vogel, G., Hoffman, P., Kosuch, J., 2002. Amphibians and Reptiles of Mount Kinabalu (North Borneo). ARG Gantner Verlag Kommanditgesellschaft, Ruggell. Manthey, U., Grossmann, W., 1997. Amphibien & Reptilien Südostasiens. Natur und Tier Verlag, Münster. Matsui, M., Ota, H., 1995. On Chinese herpetology. Herpetologica 51, 234–250. Matsui, M., Nabhitabhata, J., Panha, S., 1999. On Leptobrachium from Thailand with a description of a new species (Anura: Pelobatidae). Jpn. J. Herpetol. 18, 19–29. Matsui, M., Ito, H., Shimada, T., Ota, H., Saidapur, S.K., Khonsue, W., Tanaka-Ueno, T., Wu, G., 2005a. Taxonomic relationships within the pan-oriental narrow-mouth toad Microhyla ornata as revealed by mtDNA analysis (Amphibia, Anura, Microhylidae). Zool. Sci. 22, 489–495. Matsui, M., Shimada, T., Ota, H., Tanaka-Ueno, T., 2005b. Multiple invasions of the Ryukyu Archipelago by Oriental frogs of the subgenus Odorrana with phylogenetic reassessment of the related subgenera of the genus Rana. Mol. Phylogenet. Evol. 37, 733–742. Matsui, M., Tominaga, A., Liu, W.-Z., Khonsue, W., Grismer, L.L., Diesmos, A.C., Das, I., Sudin, A., Yambun, P., Yong, H.-S., Sukumaran, J., Brown, R.M., 2010. Phylogenetic relationships of Ansonia from Southeast Asia inferred from mitochondrial DNA sequences: systematic and biogeographic implications (Anura: Bufonidae). Mol. Phylogenet. Evol. 54, 561–570. Nguyen, V.S., Ho, T.C., Nguyen, Q.T., 2009. Herpetofauna of Vietnam. Edition Chimaira, Frankfurt am Main. Nylander, J.A.A., 2004. MrModeltest v2. Program distributed by the author. Evolutionary Biology Centre, Uppsala University.
Ohler, A., Marquis, O., Swan, S., Grosjean, S., 2000. Amphibian biodiversity of Hoang Lien Nature Reserve (Lao Cai Province, northern Vietnam) with description of two new species. Herpetozoa 13, 71–87. Ohler, A., Teynie, L.A., David, P., 2004. A green-eyed Leptobrachium (Anura: Megophryidae) from southern Laos. Raffles Bull. Zool. 52, 695–700. Orlov, N.L., 2005. A new species of the genus Vibrissaphora Liu, 1945 (Anura: Megophryidae) from Mount Ngoc Linh (Kon Tum Province) and analysis of the extent of species overlap in the fauna of amphibians and reptiles of the northwest of Vietnam and central highlands. Russ. J. Herpetol. 12, 17–38. Palumbi, S.R., Martin, A., Romano, S., McMillan, W.O., Stice, L., Grabowski, G., 1991. The Simple Fool’s Guide to PCR, Version 2.0. University of Hawaii, Honolulu. Rambaut, A., Drummond, A.J., 2007. Tracer version 1.4. Available from:. Rao, D.-Q., Wilkinson, J.A., 2008. Phylogenetic relationships of the mustache toads inferred from mtDNA sequences. Mol. Phylogenet. Evol. 46, 61–73. Rao, D.-Q., Wilkinson, J.A., Zhang, M.-W., 2006. A new species of the genus Vibrissaphora (Anura: Megophryidae) from Yunnan Province, China. Herpetologica 62, 90–95. Riddle, B.R., Dawson, M.N., Hadly, E.A., Hafner, D.J., Hickerson, M.J., Mantooth, S.J., Yoder, A.D., 2008. The role of molecular genetics in sculpting the future of integrative biogeography. Prog. Phys. Geogr. 32, 173–202. Roelants, K., Gower, D.J., Wilkinson, M., Loader, S.P., Biju, S.D., Guillaume, K., Moriau, L., Bossuyet, F., 2007. Global patterns of diversification in the history of modern amphibians. Proc. Natl. Acad. Sci. USA 104, 887–892. Sengupta, S., Choudhury, N.K., Das, I., 2001. Leptobrachium smithi Matsui, Nabhitabhata and Panha, 1999 (Anura: Megophryidae), a new record for India. J. Bombay Nat. Hist. Soc. 98, 289–291. Smith, M.A., 1921. New or little-known reptiles and batrachians from southern Annam (Indo-China). Proc. Zool. Soc. Lond. 1921, 423–440. Stuart, B.L., Sok, K., Neang, T., 2006. A collection of amphibians and reptiles from hilly eastern Cambodia. Raffles Bull. Zool. 54, 129–155. Swofford, D.L., 2002. PAUP. Phylogenetic Analysis Using Parsimony (and Other Methods), Version 4. Sinauer Associates, Sunderland, MA. Tanabe, A.S., 2007. Kakusan: a computer program to automate the selection of a nucleotide substitution model and the configuration of a mixed model on multilocus data. Mol. Ecol. Notes 7, 962–964. Taylor, E.H., 1922. Additions to the herpetological fauna of the Philippine islands I. Philipp. J. Sci. 21, 161–206. Taylor, E.H., 1962. The amphibian fauna of Thailand. Univ. Kansas Sci. Bull. 63, 265– 599. Tominaga, A., Matsui, M., Nishikawa, K., Tanabe, S., 2006. Phylogenetic relationships of Hynobius naevius (Amphibia: Caudata) as revealed by mitochondrial 12S and 16S rRNA genes. Mol. Phylogenet. Evol. 38, 677–684. Tougard, C., 2001. Biogeography and migration routes of large mammal faunas in southeast Asia during the late middle Pleistocene: focus on the fossil and extant faunas from Thailand. Palaeogeogr. Palaeoclimatol. Palaeoecol. 168, 337–358. Voris, H.K., 2000. Maps of Pleistocene sea levels in Southeast Asia: shorelines, river systems and time durations. J. Biogeogr. 27, 1153–1168. Yang, D.-T. (Ed.), 1991. The Amphibian-Fauna of Yunnan. China Forestry Publ. House, Beijing. Yang, D.-T., Rao, D. (Eds.), 2008. Amphibia and Reptilia of Yunnan. Yunnan Science and Technology Press, Kunming. Zhao, E.M., Adler, K., 1993. Herpetology of China. Contr. Herpetol. 10, 1–522. Zhao, W.-L., Morgan, W.J., 1985. Uplift of Tibetan plateau. Tectonics 4, 359–369. Zheng, Y.-C., Mo, B.-H., Liu, Z.-J., Zeng, X.-M., 2004. Phylogenetic relationships of megophryid genera (Anura: Megophryidae) based on partial sequences of mitochondrial 16S rRNA gene. Zool. Res. 25, 205–213. Zheng, Y.-C., Li, S.-Q., Fu, J.-Z., 2008. A phylogenetic analysis of the frog genera Vibrissaphora and Leptobrachium, and the correlated evolution of nuptial spine and reversed sexual size dimorphism. Mol. Phylogenet. Evol. 46, 695–707.
Contents lists available at ScienceDirect
Molecular Phylogenetics and Evolution journal homepage: www.elsevier.com/locate/ympev
Phylogenetic relationships of megophryid frogs of the genus Leptobrachium (Amphibia, Anura) as revealed by mtDNA gene sequences Masafumi Matsui a,*, Amir Hamidy a,b, Robert W. Murphy c,d, Wichase Khonsue e, Paul Yambun f, Tomohiko Shimada a,g, Norhayati Ahmad h,i, Daicus M. Belabut h,j, Jian-Ping Jiang k a
Graduate School of Human and Environmental Studies, Kyoto University, Kyoto 606-8501, Japan Museum Zoologicum Bogoriense, Research Center for Biology, Indonesian Institute of Sciences, Gd. Widyasatwaloka, Jl. Raya Jakarta Bogor km 46, Cibinong West Java, Indonesia Centre for Biodiversity and Conservation Biology, Royal Ontario Museum, 100 Queen’s Park, Toronto, Ont., Canada M5S 2C6 d State Key Laboratory of Genetic Resources and Evolution, Kunming Institute of Zoology, The Chinese Academy of Sciences, Kunming 650223, PR China e Department of Biology, Faculty of Science, Chulalongkorn University, Bangkok 10330, Thailand f Research and Education Division, Sabah Parks, P.O. Box 10626, Kota Kinabalu 88806, Sabah, Malaysia g Faculty of Bioenvironmental Science, Kyoto Gakuen University, Kameoka, Kyoto 621-8555, Japan h Faculty of Science and Technology, Universiti Kebangsaan Malaysia, 43600 UKM Bangi, Selangor, Malaysia i Institute for Environment and Development (LESTARI), Universiti Kebangsaan Malaysia, 43600 UKM Bangi, Selangor, Malaysia j Institute of Biological Sciences, Faculty of Science, University of Malaya, 50603 Kuala Lumpur, Malaysia k Chengdu Institute of Biology, The Chinese Academy of Sciences, Chengdu 610041, PR China b c
a r t i c l e
i n f o
Article history: Received 24 October 2009 Revised 17 February 2010 Accepted 9 March 2010 Available online 17 March 2010 Keywords: Leptobrachium Vibrissaphora Sundaland Indochina Southern China mtDNA Phylogenetics
a b s t r a c t By investigating genealogical relationships, we estimated the phylogenetic history and biogeography in the megophryid genus Leptobrachium (sensu lato, including Vibrissaphora) from southern China, Indochina, Thailand and the Sundaland. The genealogical relationships among the 30 named and unnamed taxa were estimated using 2009 bp of sequences from the mitochondrial DNA genes 12S rRNA, tRNAval, and 16S rRNA using maximum parsimony, maximum likelihood, and Bayesian inference methods. The genus Leptobrachium was a well-supported monophyletic group that contained two major clades. One clade had three subclades primarily from disjunct regions including Borneo, Peninsular Malaysia and Java, and Thailand. The Bornean subclade included one species each from the Philippines and Sumatra. The other major clade consisted of two subclades, one from Indochina and the other from southern China (Vibrissaphora). Divergence times estimated an old evolutionary history of each subclade, one that could not be explained by the geohistory of Southeast Asian major landmasses. Ó 2010 Elsevier Inc. All rights reserved.
1. Introduction The megophryid genus Leptobrachium Tschudi, 1838 (type species L. hasseltii from Java) is diagnosed by having a broad head and thin limbs (Dubois and Ohler, 1998). The genus is often considered to contain two subgenera, Vibrissaphora Liu, 1945, with adult males bearing spines on upper lip, and Leptobrachium, which is without such spines (Ohler et al., 2004). Vibrissaphora is distributed in southern China and Indochina, and it includes five to seven species (Frost, 2009; Rao and Wilkinson, 2008; Fei et al., 2009). Subgenus Leptobrachium is distributed from Indochina to Sundaland and it contains 15–17 species (Frost, 2009; Fei et al., 2009). Using DNA sequence data, both Zheng et al. (2008) and Rao and Wilkinson (2008) place Vibrissaphora within the genus Leptobrachium, and neither group of researchers recognizes subgenera. However, this * Corresponding author. Fax: +81 75 753 6846. E-mail address: [email protected] (M. Matsui). 1055-7903/$ - see front matter Ó 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.ympev.2010.03.014
taxonomic conclusion is made without exploring the extent of genetic diversity within Sundaland and neither study includes the type species of the genus. Sundaland contains the Malay Peninsula and the Indonesian islands of Sumatra, Java, Bali, Borneo, and smaller islands west of the Makassar and Lombok straits. All of these areas are linked by the shallow-water (<200 m) Sunda Shelf, which was exposed during periods of low sea level in the Pleistocene. The eastern boundary of Sundaland is Wallace’s Line where the Indomalayan and Australasian faunas meet. Studies of some anuran lineages from Sundaland indicate that species distributions and phylogenies are strongly related to the geological history of this region (e.g., Emerson et al., 2000; Inger and Voris, 2001; Brown and Guttman, 2002; Matsui et al., 2010). Seven species of Leptobrachium are known from Sundaland and peninsular Thailand as follows: L. hasseltii Tschudi, 1838; L. montanum Fischer, 1885; L. abbotti (Cochran, 1926); L. hendricksoni Taylor, 1962; L. nigrops Berry and Hendrickson, 1963; L. gunungense
260
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
Malkmus, 1996, and L. smithi Matsui, Nabhitabhata, and Panha, 1999. Among these species, L. montanum, L. abbotti, and L. gunungense are endemic to Borneo and L. smithi is mainly distributed in Thailand and the northern part of Peninsular Malaysia, as well as through Myanmar to Assam, India (Sengupta et al., 2001). Leptobrachium hendricksoni is distributed from southern Thailand through the Malay Peninsula to western Borneo and Sumatra, and L. nigrops occurs on the Malay Peninsula, Belitung Island, and western Borneo. How and when they obtained their present-day distributions in the Sunda Islands are biogeographically interesting topics. In Leptobrachium, taxonomic problems span from the supraspecific category to the species level. The name L. hasseltii was applied to many Southeast Asian populations (Inger, 1954, 1966; Berry, 1975) until Inger et al. (1995) clarified that L. hasseltii from Borneo was not conspecific with the Javanese population. They applied the names L. montanum and L. abbotti to Bornean species. However, some populations, such as those from the Philippines, are still referred to as L. hasseltii because of the absence of a taxonomic reassessment. Iskandar (1998) and Matsui et al. (1999) suggested that true L. hasseltii was likely restricted to Java and adjacent islands, and Dubois and Ohler (1998) suggested the possible occurrence of more than one species of Leptobrachium on Java, based on a large extent of variation in female body size of Leptobrachium (sensu lato). Such taxonomic arguments still need to be clarified. These uncertainties mainly derive from the use of a few specimens from limited ranges. Intra- and interspecific variation is inadequately assessed. Some critical characteristics useful for species identification are lost upon preservation of museum vouchers. Molecular data and phylogenetic analyses can enable the identification of species and a higher level taxonomy in Leptobrachium. In this study, we use mtDNA gene sequences from various populations of Leptobrachium, especially from Sundaland, to evaluate the taxonomic status of many populations and hypothesize the evolutionary history of the species in this genus.
2. Materials and methods 2.1. Sampling design We examined a total of 80 partial DNA sequences of the mitochondrial DNA genes 12S rRNA, tRNAval, and 16S rRNA from 30 species/subspecies of the genus Leptobrachium (sensu lato) and six outgroup species (Figs. 1 and 2 and Table 1). Our own sampling of Leptobrachium consisted of 60 specimens of 21 taxa mainly obtained from Sundaland and Indochina, including several undescribed or unidentified taxa including the following: Leptobrachium sp. 1, a Sumatran species that has distinct eye color (Hamidy and Matsui, 2010); Leptobrachium sp. 2, a Philippine species often called L. hasseltii; Leptobrachium sp. 3 from Gwa, Myanmar, catalogued as L. hasseltii (Zheng et al., 2008); Leptobrachium sp. 4 from Pilok, Thailand, a dried carcass whose identification was difficult; and Leptobrachium sp. 5, which resembles Vibrissaphora but lacks male labial spines. We used Leptolalax heteropus (Boulenger, 1900) and Pelodytes punctatus (Daudin, 1802) as outgroup taxa. DNA sequences from GenBank were obtained for 11 Indochinese and Chinese species of Leptobrachium (sensu lato) and the following four outgroup species: Oreolalax rhodostigmatus Hu and Fei in Liu, Hu, and Fei, 1979 (EF397248: Fu et al., 2007), Megophrys nasuta (Schlegel, 1858) (DQ283342: Frost et al., 2006), Pelobates fuscus (Laurenti, 1768) (DQ283113: Frost et al., 2006), and Scaphiopus holbrookii (Harlan, 1835) (DQ283156: Frost et al., 2006). Leptolalax Dubois, 1980 and Oreolalax Myers and Leviton, 1962 are the sister group of Leptobrachium (Lathrop, 1997; Zheng et al., 2004, 2008; Fu
et al., 2007), whereas Pelobates Wagler, 1830 (Pelobatidae Bonaparte, 1850), Pelodytes Bonaparte, 1838 (Pelodytidae Bonaparte, 1850), and Scaphiopus Holbrook, 1836 (Scaphiopodidae Cope, 1865) form a monophyletic sister group with Megophryidae (García-París et al., 2003; Dubois, 2005). Voucher specimens/tissues are stored in ABTC (Australian Biological Tissue Collection, South Australian Museum), BOR (BORNEENSIS collection, University Malaysia Sabah), IEBR (Institute of Ecology and Biological Resources, Hanoi, Vietnam), KUHE (Kyoto University, Graduate School of Human and Environmental Studies), MDK (Department of Conservation and Ecotourism, Faculty of Forestry, Bogor Agricultural Institute), MNHN (Museum National d’Histoire Naturelle, Paris), MZB (Museum Zoologicum Bogoriense), ROM (Royal Ontario Museum), SP (Sabah Parks), UKM (University Kebangsaan Malaysia), UM (University Malaya), and UTA (Department of Biology, University Texas at Arlington). 2.2. Preparation of DNA, PCR and DNA sequencing We obtained tissues from frozen or ethanol (95–99%) preserved specimens and extracted total genomic DNA using standard Phenol–chloroform extraction procedure (Hillis et al., 1996). We homogenized tissues in 0.6 ml STE buffer containing 10 mM Tris/ HCl, pH 8.0, 100 mM NaCl and 1 mM EDTA, pH 8.0. We added Proteinase K (0.1 mg/ml) to the homogenate solutions and digested proteins for 4–12 h at 55 °C. The solution was treated with phenol and chloroform/isoamyl alcohol and DNA was precipitated with ethanol. DNA precipitates were dried and then resuspended in 0.6 ml TE (10 mM Tris/HCl, 1 mM EDTA, pH 8.0) and 1 ll was subjected to polymerase chain reaction (PCR). The PCR cycle included an initial denaturation step of 5 min at 94 °C and 33 cycles of denaturation for 30 s at 94 °C, primer annealing for 30 s at 48–50 °C, and extension for 1 min 30 s at 72 °C. Primers used in PCR are shown in Table 2. The PCR products purified using polyethylene glycol (PEG, 13%) precipitation procedures were used directly as templates for Cycle Sequencing Reactions with fluorescent-dye-labeled terminator (ABI Big Dye Terminators v.3.1 cycle sequencing kit). The sequencing reaction products were purified by ethanol precipitation following the manufacture’s protocol and then run on an ABI PRISM 3130 genetic analyzer. All samples were sequenced in both directions using the same primers as for PCR. 2.3. Phylogenetic analysis Aligned, concatenated sequences of 12S rRNA, tRNAval, and 16S rRNA yielded a total 2009 bp positions. We used Chromas Pro software (Technelysium Pty Ltd., Tewntin, Australia) to edit the sequences, and align them using the ClustalW option of Bioedit (Hall, 1999). The initial alignments were then checked by eye and adjusted slightly. Phylogenetic trees were constructed using maximum parsimony (MP), maximum likelihood (ML), and Bayesian inference (BI). MP trees, obtained using PAUP4.0b10 (Swofford, 2002), involved a heuristic search with the tree bisection recognition (TBR) branch-swapping algorithm, and 100 random additions replicates. Transitions and transversions were equally weighted, and gaps were treated as missing data. ML analysis were performed by Treefinder version June 2007 (Jobb et al., 2004), with the general time-reversible (GTR) model of DNA evolution with a gamma shape parameter (G), identified as the best-fitting model under the Akaike information criterion implemented in Kakusan 3 (Tanabe, 2007). BI and Bayesian posterior probabilities (BPP) were estimated using MrBayes 3.0b4 (Huelsenbeck and Ronquist, 2001), under the GTR model with G and proportion of invariable sites (I), selected by MrModeltest2.2 (Nylander, 2004). BI used four simultaneous Metropolis coupled Monte Carlo Markov chains for 6,000,000 generations. We sampled a tree every 100 generations
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
261
Fig. 1. Map of Southeast Asia showing sampling localities of Leptobrachium (sensu lato) and related species included in this study. Sample numbers are included in Table 1. Triangle: Subclade I; reverse triangle: Subclade II; filled square: Subclade III; circle: Subclade IV; open square: Subclade V.
and calculated a consensus topology for 30,001 trees after discarding the first 30,000 trees (burn-in = 3,000,000). Strength of nodal support in the MP analyses used non-parametric bootstrapping (MPBS; Felsenstein, 1985) with 1000 pseudoreplicates in PAUP, and 100 replicates were used for the ML tree (MLBS). A priori, we regarded tree nodes with bootstrap value 70% or greater as sufficiently resolved (Huelsenbeck and Hillis, 1993), and those between 50% and 70% as tendencies. In the BI analysis, nodes with a BPP of 95% or greater were considered significant (Leaché and Reeder, 2002).
2.4. Estimation of divergence time We estimated the divergence times using a Bayesian relaxed molecular clock calculated in BEAST (Drummond and Rambaut, 2007) using 60 million generations under a GTR+G+I substitution model and uncorrelated log-normal ‘‘relaxed” clock rate model (Drummond et al., 2006). We sampled the MCMC chain every 1000 generations, for a total of 10 million samples, and assessed convergence to the stationary distribution through inspection of the likelihood and parameter sample plots in Tracer version 1.4
262
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
Fig. 2. Maximum likelihood (ML) phylogram of 2009 bp of 12S rRNA, tRNAval and 16S rRNA mitochondrial genes for samples of Leptobrachium (sensu lato) and related species. Sample numbers are included in Table 1. Numbers above branches represent bootstrap supports for maximum parsimony (MP) /ML/ and Bayesian posterior probabilities. Asterisks indicate nodes with full bootstrap supports for ML and MP (100%) inferences and Bayesian posterior probabilities (PP = 100%).
(Rambaut and Drummond, 2007). A burn-in of the first one million samples was used. Three external calibration points provided by Roelants et al. (2007) were used to estimate dates of the cladogenic events. The
divergence time between Pelodytes and Pelobatidae + Megophryidae was assumed to be 143.1 MYBP (95% Credibility Interval [CI] of 120.8–166.1 MYBP), between Pelobates and Asian members at 114.6 MYBP (CI 93.2–135.9), and between Leptolalax and
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
263
Table 1 Sample of Leptobrachium (sensu lato) and outgroup species used for mtDNA analysis in this study together with the information on voucher, collection locality and GenBank accession numbers. UN: Unnumbered. See text for voucher abbreviations. Sample no
Species
Voucher
Locality
GeneBank
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73
L. montanum lineage 1 L. montanum lineage 1 L. montanum lineage 1 L. montanum lineage 1 L. montanum lineage 1 L. montanum lineage 1 L. montanum lineage 1 L. montanum lineage 1 L. montanum lineage 1 Leptobrachium sp. 1 Leptobrachium sp. 1 Leptobrachium sp. 1 L. gunungense L. gunungense L. abbotti L. abbotti L. abbotti L. montanum lineage 2 L. montanum lineage 2 Leptobrachium sp. 2 L. hendricksoni L. hendricksoni L. hendricksoni L. hendricksoni L. hendricksoni L. hendricksoni L. hendricksoni L. hendricksoni L. hasseltii L. hasseltii L. hasseltii L. hasseltii L. hasseltii L. hasseltii L. nigrops L. nigrops L. nigrops L. nigrops L. nigrops L. nigrops L. nigrops L. smithi L. smithi L. smithi L. smithi L. smithi L. smithi L. smithi L. smithi L. smithi Leptobrachium sp. 3 Leptobrachium sp. 4 V. l. liui V. l. yaoshanensis V. leishanensis V. jiulongshanensis L. chapaense lineage 1 L. chapaense lineage 1 L. chapaense lineage 1 L. chapaense lineage 1 V. boringii Leptobrachium sp. 5 V. echinata V. ailaonica V. promustache L. chapaense lineage 2 L. chapaense lineage 3 L. hainanense L. chapaense lineage 3 L. mouhoti L. pullum V. ngoclinhensis L. xanthospilum
KUHE UN larva KUHE 42812 BOR 22008 BOR 04B019 KUHE 42813 KUHE 42815 KUHE 17306 KUHE UN larva MDK 01 KUHE 42805 MZB Amp 15862 UTA_A53689 BOR 22956 SP 3825a BOR 12866 KUHE 39295 BOR 04B018 BOR 08481 SP 21481 ABTC 76306 KUHE 15336 KUHE 52403 UKM HC110 KUHE UN tissue MDK 10 KUHE UN tissue KUHE 15680 KUHE 52150 KUHE 42807 KUHE 42808 MZB UN tissue UTA_A53688 KUHE 42818 KUHE 42820 MZB Amp1790 MZB Amp1791 KUHE 15430 KUHE 15658 UKM HC0727 KUHE 42587 KUHE 42590 KUHE 19281 KUHE 19282 KUHE 19834 KUHE 19939 KUHE 20200 KUHE 20201 KUHE 23342 UM D0139 CAS 222215 CAS 222293 KUHE UN tissue IZCASH 30020 KUHE UN larva IZCASH 30004 IZCASH 30034 KUHE UN tissue ROM 41243 KUHE 19122 IZCASH 30048 IZCASH 30021 KUHE 34396 MNHN 1999.5657 IZCASH 30046 IZCASH 30044 AMNH 163791 ROM 32176 KUHE UNL MNHN 1997.5249 FMNH 261758 IEBR 2780 IEBR 2827 ROM 32177
Indonesia, S. Kalimantan, Banjar, Paramasan Indonesia, S. Kalimantan, Banjar, Paramasan Malaysia, Sabah, Tawau Malaysia, Sabah, Tawau Indonesia, C. Kalimantan, Lamandau, Belantikan Indonesia, C. Kalimantan, Lamandau, Belantikan Malaysia, Sarawak, Matang, Serapi Malaysia, Sarawak, Matang, Serapi Indonesia, W. Kalimantan, Betung Kerihun, Nanga Bungo Indonesia, Sumatra, Lampung, Liwa, Kubu Perahu Indonesia, Sumatra, Lampung, Liwa, Kubu Perahu Indonesia, Sumatra, Jambi, betw. Tanpa and Sungai penuh Malaysia, Sabah, Kinabalu, Mesilau Malaysia, Sabah, Kinabalu, Sungai Carson Malaysia, Sabah, Crocker, Ulu Kimanis Malaysia, Sabah, Kinabalu, Poring Malaysia, Sabah, Tawau Malaysia, Sabah, Crocker, Ulu Kimanis Malaysia, Sabah, Kinabalu, Silau-silau Philippines, Mindanao, Davao Malaysia, Penang Malaysia, Peninsula, Kelantan, Pasir Puteh Malaysia, Peninsula, Trengganu, Hulu Trengganu Malaysia, Peninsula, Trengganu, Sekayu Indonesia, Sumatra, Jambi, Bungo Indonesia, Sumatra, S. Sumatra, Lahat Malaysia, Peninsula, Selangor, Kuala Lumpur Malaysia, Peninsula, Johor, Endau Rompin, Selai Indonesia, Sumatra, Lampung, Liwa, Kubu Perahu Indonesia, Sumatra, Lampung, Liwa, Kubu Perahu Indonesia, W. Java, Gede-Pangrango National Park Indonesia, W. Java, Bogor, Cisarua Safari Park Indonesia, C. Java, Purworejo, Kaligesing Indonesia, Java, Yogyakarta, Kulon Progo, Kiskendo Indonesia, Belitung, Tanjung Pandang Indonesia, Belitung, Tanjung Pandang Malaysia, Peninsula, Selangor, Kuala Lumpur Malaysia, Peninsula, Selangor, Kuala Lumpur Malaysia, Peninsula, Pahang, Temerloh Malaysia, Sarawak, Kanowit Malaysia, Sarawak, Kanowit Thailand, Loei, Phu Luang Thailand, Loei, Phu Luang Thailand, Mae Hong Son, Phasua WF Thailand, Kanchanaburi, Pilok Thailand, Phetchaburi, Kaeng Krachan Thailand, Phetchaburi, Kaeng Krachan Thailand, Trang, Kaochong Malaysia, Perlis, Langkawi Myanmar, Kyaik Hto Myanmar, Rakhine State, Gwa Township, Rakhine Thailand, Kanchanaburi, Pilok China, Hunan, Man Shan China, Guanxi, Huapin China, Guizhou, Leigong Shan China, Zhejiang, Jiulong Shan Thailand, Doi Angkang China, Yunnan, 14.9 km SE of Simao Thailand, Chiang Mai, Doi Intanon China, Yunnan, Longling China, Sichuan, Emei Shan Laos, Xamneua, Phupan Vietnam, Lao Cai, Sa Pa China, Yunnan, Ailao Shan China, Yunnan, Dawei Shan Vietnam, Ha Giang, Vi Xuyen, Cao Bo Vietnam, Vinh Phu, Tam Dao China, Hainan, Wuzhi Shan Vietnam, Ben En Vietnam, Pichrada Vietnam, Kon Tum, Kon Plong Vietnam, Kon Tum, Ngoc Linh Mountain Vietnam, Gia Lai, Tram Lap
AB530391 AB530392 AB530393 AB530394 AB530395 AB530396 AB530397 AB530398 AB530399 AB530400 AB530401 AB530402 AB530403 AB530404 AB530405 AB530406 AB530407 AB530408 AB530409 AB530410 AB530411 AB530412 AB530413 AB530414 AB530415 AB530416 AB530417 AB530418 AB530419 AB530420 AB530421 AB530422 AB530423 AB530424 AB530425 AB530426 AB530427 AB530428 AB530429 AB530430 AB530431 AB530432 AB530433 AB530434 AB530435 AB530436 AB530437 AB530438 AB530439 EF672271 DQ283239 AB530440 EF544196 AB530441 EF544200 EF544205 AB530442 AB530443 AB530444 EF544239 EF544207 AB530445 AB530446 EF544225 EF544240 DQ283052 EF544232 AB530447 AB530448 EF672272 AB530449 AB530450 AB530451 (continued on next page)
264
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
Table 1 (continued) Sample no
Species
Voucher
Locality
GeneBank
74 75 76 77 78 79 80
L. banae Oreolalax rhodostigmatus Leptolalax heteropus Megophrys nasuta Pelobates fuscus Pelodytes punctatus Scaphiophus holbrookii
ROM 32200 CIB ZYCA746 KUHE 15487 FMNH 236525 A. Haas Collection MNHN 2000.2401 AMHH A168434
Vietnam, Gia Lai, Krong Pa China, Guizhou, Da Fang Malaysia, Peninsula, Perak, Larut Malaysia, Sabah, Crocker, Tenom Germany, Thuringia, Geroda, Triptis France, Villemoisiau USA, Florida, Alachua
AB530452 EF397248 AB530453 DQ283342 DQ283113 AB530454 DQ283156
Table 2 Primers used in this study. Target val
12S rRNA and tRNA
16 S rRNA
Primer
Sequence (50 –30 )
Reference
12Sh 12SA-L L1507 12SFLeptobrachium H1548 L1879 16sl2021 16L-1 H1923 H2315Leptobrachium 16sh2715 16H1
AAAGGTTTGGTCCTAGCCTT AAACTGGGATTAGATACCCCACTAT TACACACCGCCCGTCACCCTCTT CCGCCAAGTCCTTTGGGTTT TACCATGTTACGACTTTCCTCTTCT CGTACCTTTTGCATCATGGTC CCTACCGAGCTTAGTAATAGCTGGTT CTGACCGTGCAAAGGTAGCGTAATCACT AAGTAGCTCGCTTAGTTTCGG TCGTTGTTACTAGTYCTAACAT AAGCTCCATAGGGTCTTCTCGTC CTCCGGTCTGAACTCAGATCACGTAGG
Cannatella et al. (1998) Palumbi et al. (1991) Made in this study Modified from Goebel et al. (1999) Matsui et al. (2005a) Made in this study Tominaga et al. (2006) Hedges (1994) Made in this study Made in this study Tominaga et al. (2006) Hedges (1994)
Oreolalax + Leptobrachium 43.8 MYBP (CI 32.0–58.7). However, using external calibrations only might result in very old estimates (Matsui et al., 2010). Thus, we applied a date of 1.40 MYBP (CI 0.90–1.90) for two insular populations of L. hasseltii as estimated by Matsui et al. (2005b) for Odorrana hosii (Boulenger, 1891) from Sundaland.
eter = 0.290; nucleotide frequencies: A = 0.347, C = 0.210, G = 0.173, and T = 0.270). BI calculated average parameter estimates for nucleotide frequencies of A = 0.373, C = 0.206, G = 0.144 and T = 0.276, a gamma shape parameter 0.662, and proportion of invariable sites of 0.297. 3.2. Phylogenetic relationships
2.5. Variation in iris color Iris color has served as a valuable diagnostic morphological characteristic for species of Leptobrachium (Dubois and Ohler, 1998; Lathrop et al., 1998; Matsui et al., 1999). We accumulated data for this character from our own observations and descriptions and photographs in the literature. We hypothesized the evolution of iris color by mapping the data on the phylogenetic tree. 3. Results 3.1. Sequence and statistics Sequence statistics for the three gene fragments and for the combined alignment including all nucleotide positions are provided in Table 3. The aligned 12S rRNA, tRNAval, and 16S rRNA data set consisted of 2009 characters, in which 925 sites were variable, and 800 potentially phylogenetically informative. MP analysis yielded four most parsimonious trees of 4089 steps, a consistency index of 0.401 and retention index of 0.818. The ML analysis produced a topology with ln L 21943.875 (gamma shape paramTable 3 Alignment statistics for fragments of the 12S rRNA, tRNAval, and 16S rRNA (all nucleotide positions included); number of base bairs (bp), number of variable sites (vs), number of parsimony informative sites (pi), the transition–transversion ratio given for ingroups only (ti/tv).
12S rRNA tRNAval 16S rRNA Combined
bp
vs
pi
ti/tv
475 71 1463 2009
161 38 726 925
138 35 627 800
1.175 3.900 1.271 1.288
All analyses resulted in essentially the same topologies. They differed only in associations at poorly supported nodes. The ML tree (Fig. 2) infers the following sets of relationships: (i) Monophyly of Leptobrachium sensu lato (Leptobrachium and Vibrissaphora) with respect to Oreolalax and Leptolalax was supported in all trees (MPBS = 71%, MLBS = 78%, BPP = 98%). (ii) Leptobrachium sensu lato was divided into two basal monophyletic lineages, Clade A (all Sundaland and most Thai species; all support values = 100%) and Clade B (primarily Chinese and Indochinese species associated with Vibrissaphora: MPBS = 78%, MLBS = 93%, BPP = 100%). (iii) Clade A contained three monophyletic subclades whose relationships to each other were unresolved. Subclade I consisted of species from Borneo, Sumatra, and Mindanao, Philippines (all values 100%), Subclade II encompassed species mainly from Peninsular Malaysia and Java (MPBS = 76%, MLBS = 100%, BPP = 100%), and Subclade III contained ThaiMyanmar species (all values 100%). (iv) In Subclade I, Leptobrachium sp. 2 from Mindanao (sample number 20) was a sister lineage to the clade containing Bornean and Sumatran species. In the latter clade (all values 100%), L. montanum lineage 2 from Kinabalu (18, 19) formed the sister species to the clade of the remaining Bornean and Sumatran species (MPBS = 93%, MLBS = 97%, BPP = 100%). The latter clade contained two sublineages. First, L. montanum lineage 1 (populations 1–9 from Sabah, Sarawak, and Kalimantan, including Paramasan, the type locality of L. montanum; MPBS = 70%, MLBS = 65%, BPP = 100%) was united with Leptobrachium sp. 1 from Sumatra (10–12; all values 100%). Second, L. abbotti (15–17; all values 100%) and
265
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
(v)
(vi)
(vii)
(viii)
(ix)
L. gunungense (13, 14; all values 100%) from Sabah, formed a monophyletic clade (MPBS = 96%, MLBS = 97%, BPP = 100%). The clade containing L. montanum lineage 1 was not universally supported to be monophyletic (MPBS = 70%, MLBS = 65%, BPP = 100%). Thus, its sister species relationship with Leptobrachium sp. 1 needs to be verified. Subclade II contained three species, of which all samples clustered together with 100% support: L. hendricksoni (21–28) from Peninsular Malaysia and Sumatra, L. hasseltii (29–34) from Java and Sumatra, and L. nigrops (35–41) from Peninsular Malaysia, Belitung Island, and Sarawak. Of these, L. hendricksoni and L. hasseltii were sister species (MPBS = 78%, MLBS = 89%, BPP = 100%). Peninsular Malayan L. hendricksoni was paraphyletic, with respect to northern populations forming a sister group with Sumatran populations (MPBS = 83%, MLBS = 83%, BPP = 100%), and not with southern Malayan populations. In L. hasseltii, relationships of populations from western Java, central Java, and Sumatra were unresolved. In contrast, three genetically divergent lineages were discovered within L. nigrops. The population from Sarawak (40, 41) was the sister group of a well-supported (MPBS = 100%, MLBS = 99%, MPP = 100%) clade comprised of animals from Peninsular Malaysia (37–39) and Belitung Island (35, 36). Uncorrected p-distances were large, averaging 9.3% between Peninsular Malaysia and Belitung, 11.7% between Borneo and Belitung, and 11.2% between Borneo and Malay Peninsula. In Subclade III, populations of L. smithi (42–50) from Thailand, Langkawi Island of Malaysia, and Myanmar formed a monophyletic clade (MPBS = 100%, MLBS = 99%, BPP = 100%), but the relationships among these individuals were unresolved. Leptobrachium sp. 3 from Gwa, Myanmar (51) was the sister species to this clade and this association received full support. Together, these taxa formed the sister group of Leptobrachium sp. 4 from Pilok, Thailand (52) and with full support. Clade B was comprised of two subclades. Subclade IV primarily contained Vibrissaphora (all values 100%), and Subclade V consisted primarily of Indochinese species of Leptobrachium (all values 100%). Subclade IV consisted of three main clades. Leptobrachium chapaense (Bourret, 1937) lineage 2 (66) from Ha Gian, northern Vietnam, and it formed the sister species of the remaining species. The latter clade received full support and V. promustache Rao et al., 2006 (65) was the sister species of the other species (MPBS = 77%, MLBS = 100%, BP = 98%), in which several sublineages had unresolved relationships to each other. These sublineages included full support for five clades as follows: first, a clade containing V. echinata Dubois and Ohler, 1998 (63) from Vietnam and V. ailaonica Yang, Chen, and Ma in Yang, Ma, Chen, and Li, 1983 (64) from China; second, Leptobrachium sp. 5 (62) from Laos; third, V. boringii Liu, 1945 (61) from China; fourth, a fully supported clade containing L. chapaense lineage 1 (57–60) from Thailand and China; and fifth, a clade consisting of V. l. liui Pope, 1947 (53), V. l. yaoshanensis Liu and Hu in Hu, Tian, and Wu, 1978 (54), V. leishanensis Liu and Hu in Hu, Zhao, and Liu, 1973 (55) and V jiulongshanensis Wei and Zhao, 1981 (56) from China (MPBS = 99%, MLBS = 100%, BPP = 100%). In the last clade, two subspecies of V. liui were monophyletic (MPBS = 98%, MLBS = 100%, BP = 100%). Subclade V comprised five lineages whose relationships to one another were unresolved. Of these five, Vietnamese L. mouhoti Stuart, Sok, and Neang, 2006 (70) and L. pullum (Smith, 1921) (71) formed a fully supported clade, as did Chinese L. hainanense Ye and Fei in Ye, Fei, and Hu, 1993 (68) and Vietnamese L. chapaense lineage 3 (67, 69). The
remaining lineages consisted of the Vietnamese species L. xanthospilum Lathrop, Murphy, Orlov, and Ho, 1998 (73), L. banae Lathrop, Murphy, Orlov, and Ho, 1998 (74), and V. ngoclinhensis Orlov, 2005 (72). 3.3. Divergence times Large degrees of overlap in divergence time estimations were observed in the confidence intervals (CI). Separation of Leptobrachium sensu lato and Oreolalax was estimated to have occurred 51.8 MYBP (CI 36.2–68.5) in the early Eocene (age adopted from Holloway and Hall (1998)). In the mid Eocene (50–42 MYBP), Leptobrachium split into Clades A and B (Table 4). Divergences of Subclades I–III and IV–V occurred nearly simultaneously from the late Eocene (42–35 MYBP) to the early Oligocene (35–29 MYBP). Subclade III, from Thailand and Myanmar, separated from the common ancestor of Subclades I and II from Sundaland and Peninsular Malaysia. Soon thereafter, Subclade I, consisting species mainly from Borneo, and Subclade II, encompassing species mainly from Peninsular Malaysia and Java, split. In Clade A, speciation was initiated in the late Oligocene, 30–23 MYBP when L. nigrops separated from L. hendricksoni and L. hasseltii. It continued in the early Miocene, 23–14 MYBP, with the following events: Leptobrachium sp. 2 on Mindanao split from the common ancestor of Bornean species and Sumatran Leptobrachium sp. 1; L. hendricksoni separated from L. hasseltii; Leptobrachium sp. 4 split from the common ancestor of L. smithi and Leptobrachium sp. 3; and L. montanum lineage 2 was isolated from the other Bornean and Sumatran populations. Subsequently, in the late Miocene, 14–5 MYBP, the common ancestor of L. montanum lineage 1 and Leptobrachium sp. 1 split from that of L. abbotti and L. gunungense. In addition, L. smithi split from Leptobrachium sp. 3 from Gwa, Myanmar, L. montanum lineage 2 split from Sumatran Leptobrachium sp. 1, and L. abbotti speciated from L. gunungense. In the main Indochinese and Chinese clade, diversification occurred from the early Miocene (L. banae and other Leptobrachium from Vietnam and Hainan in Subclade V; L. chapaense lineage 2 from Ha Gian and the remaining species of Subclade IV including
Table 4 Estimated divergence times (MY) of main divergences of Leptobrachium (sensu lato). Combination
Mean
CI
Leptobrachium/Oreolalax Leptobrachium Clade A/Clade B Leptobrachium Subclade IV/Subclade V Leptobrachium Subclades I + II/Subclade III Leptobrachium Subclade I/Subcclade II L. nigrops/L. hendricksoni + L. hasseltii L. banae/other Vietnamese species Leptobrachium sp. 2 (Philippines)/Bornean and Sumatran species L. hendricksoni/L. hasseltii Leptobrachium sp. 4/other Thai-Myanmar species L. chapaense lineagae 2/other species in Subclade IV Within L. nigrops L. xanthospilum/other Vietnamese species L. montanum lineage 2/other Bornean and Sumatran species L. abbotti + L. gunungense/L. montanum lineage 1 + Leptobrachium sp. 1 Leptobrachium sp. 3/L. smithi V. l. liui/V. promustache Within L. hendricksoni L. montanum lineage 1/Leptobrachium sp. 1 L. abbotti/L. gunungense Within L. montanum lineage 1 Within L. smithi Within L. hasseltii L. hainanense/L. chapaense lineage 3 (Tam Dao)
51.8 46.4 35.8 34.0 31.0 25.5 19.6 19.0
(36.2–68.5) (31.6–61.4) (21.1–49.4) (22.5–45.7) (20.7–42.5) (16.0–34.7) (10.3–29.6) (11.3–27.4)
18.8 16.6 16.4 15.8 15.7 14.2
(11.0–27.2) (8.6–25.4) (7.8–26.3) (8.2–24.1) (7.8–24.9) (8.6–21.0)
11.2
(6.5–16.6)
9.8 9.7 7.8 6.1 5.5 5.0 4.0 2.8 2.6
(4.8–15.7) (3.6–16.7) (4.0–12.2) (3.5–9.1) (2.0–9.9) (3.0–7.5) (2.0–6.4) (1.4–4.5) (0.7–4.7)
266
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
Vibrissaphora and L. chapaense lineage 1; L. xanthospilum and, other Vietnamese and Hainan Leptobrachium and V. ngoclinhensis) to the late Miocene (Chinese V. promustache and V. liui). Intraspecific divergence was suggested to have occurred as early as the late Miocene, as exemplified by L. hendricksoni. However, most of the intraspecific divergence occurred in the Pliocene (5.0–1.8 MYBP) as expected and seen in L. montanum lineage 1, L. smithi, and L. hasseltii. Divergence in Indochina and China also occurred in the Pliocene, as exemplified in L. hainanense and L. chapaense lineage 3. An exception to the recent diversification is seen in L. nigrops where the population from Sarawak split from populations in Peninsular Malaysia and Belitung in the early Miocene, and the latter populations diverged in the middle Miocene.
Leptobrachium sp. 1 (Subclade I) had totally blue iris in adults, although it was grey in a juvenile. In Clade B, a light-colored (ranging from white through sky-blue to lime green) upper iris predominated both in Subclade IV (V. liui, V. jiulongshanensis, V. leishanensis, L. chapaense [lineages 1and 2], V. boringii, V. ailaonica, V. echinata, and L. promustache) and in Subclade V (L. chapaense lineage 3, L. hainanense, V. ngoclinhensis, L. banae, and L. xanthospilum). As in Clade A, exceptions occurred. The upper part of the iris in L. mouhoti and L. pullum (Subclade V) was yellow, scarlet, or orange red. 4. Discussion 4.1. Taxonomic considerations of Clade A
3.4. Iris color In Clade A, a totally dark black iris predominated in Subclade I (L. montanum [lineages 1, 2], L. gunungense, L. abbotti and Leptobrachium sp. 2) and Subcalde II (L. hasseltii and L. nigrops) (Table 5). However, the upper part of the iris was yellow, orange, scarlet, or red in L. hendricksoni (Subclade II) and L. smithi (Subclade III). Furthermore, in L. hendricksoni, the iris color varied from having color only on the upper part to being totally orange (in one juvenile).
Subclade I consisted of species from Borneo, Sumatra, and Mindanao. The Philippine population (sample 20), which diverged at the base of this subclade, has been treated as being L. hasseltii (Taylor, 1922; Inger, 1954, 1966; Alcala, 1986). However, L. hasseltii occurs in Subclade II (see below). Therefore, the Philippine population constituted an unnamed species, herein referred to as Leptobrachium sp. 2. Morphological studies also suggested that the Philippine population was not conspecific with L. hasseltii
Table 5 Known variatin in the iris color of Leptobrachium (sensu lato). UN: Unnumbered. See text for voucher abbreviations. Species
Iris color
Locality
Voucher [Sample No.] and/or reference
L. abbotti
Completely black
Borneo
L. gunungense L. hasseltii
Completely black Completely black
Borneo Java
L. hendricksoni
Upper half orange
Sumatra Peninsular Malaysia
L. montanum lineage 1
Completely orange Upper half red Completely black
Peninsular Malaysia ? Borneo
L. montanum lineage 2 L. nigrops
Completely black Completely black
L. smithi
Upper half scarlet Upper half orange
Borneo Peninsular Malaysia Belitung Borneo Thailand Thailand
Upper half yellow Completely blue Completely grey Completely black Upper half white Upper half pale green Upper half blue Upper half sky-blue Upper half white Upper half white Upper half light-blue Upper half white Upper half light-blue Upper half sky-blue Upper outer margin orange red Upper half? scarlet Upper half yellow Upper half white Upper half light-blue Upper half light green Upper half bluish green Upper half lime green
Peninsular Malaysia Thailand Sumatra Sumatra Philippines Vietnam Laos China Thailand Vietnam Vietnam Vietnam China China China Cambodia Vietnam Vietnam Vietnam China China China Vietnam
BOR 04B018 [17]; BOR 04B110: BOR12866 [15]; KUHE 39294; KUHE 39295 [16]; KUHE 39296; SP 21669; SP 26559; SP 26590 BOR 22959 [13]; KUHE 39377 KUHE 42818 [33]; KUHE 42820 [34]; MZB UN; KUHE 42821; MZB UN [31]; UTA A53688 [32] KUHE 42807 [29]; KUHE 42808 [30]; KUHE 42809 ABTC 4081; UKM HC 110 [23]; KUHE 15008; KUHE 15336 [21]; KUHE 15680 [27]; KUHE 15756; KUHE 52150 [28]; KUHE 52403 [22]; MDK 10 [25] KUHE 52150 (Juvenile) Manthey and Grossmann (1997) BOR 04B019 [4]; BOR 22008 [3]; BOR 04B020; BOR 04B021; MDK UN; KUHE 17306 [7]; KUHE 42811; KUHE 42812 [2]; KUHE 42813 [5]; KUHE 42814; KUHE 42815 [6]; KUHE 42816; KUHE 42817; MDK 01 [9] BOR 08481 [18]; KUHE 39204; SP 21481 [19] KUHE 15430 [37]; KUHE 15658 [38]; KUHE 15706; KUHE 35433; UKM HC 072 [39] MZB Amp1790 [35]; MZB Amp1791 [36] KUHE 081122-1; KUHE 42587 [40]; KUHE 42590 [41] KUHE 19281 [42]; KUHE 19282 [43] KUHE 19393; KUHE 19394; KUHE 19395; KUHE 19518; KUHE 19538; KUHE 19832; KU; KUHE 23318HE 19834 [44]; KUHE 19937; KUHE 19938; KUHE 19939 [45]; KUHE 20000; KUHE 20032; KUHE 23342 [48] UM D0136; UM D0139 [49] KUHE 20200 [46]; KUHE 20201 [47] MZB Amp 15862 [11]; UTA A53689 [12] KUHE 42805 [10](Juvenile) Brown and Diesmos (2009) ROM 32200 [74]; Lathrop et al. (1998), Nguyen et al. (2009) Ohler et al. (2004) Yang (1991) KUHE 19122 [59]; Dubois and Ohler (1998) Bain and Nguyen (2004) ROM 32176 [67]; Lathrop et al. (1998) Dubois and Ohler (1998) Fei et al. (2009) Fei (1999) Fei et al. (2009) Stuart et al. (2006), Nguyen et al. (2009) Smith (1921) Nguyen et al. (2009) Lathrop et al. (1998), Nguyen et al. (2009) Fei (1999), Fei et al. (2009) Fei (1999) Fei et al. (2009) Dubois and Ohler (1998)
Leptobrachium sp. 1 Leptobrachium sp. 2 L. banae L. buchardi L. chapaense lineage 1 L. chapaense lineage 2 L. chapaense lineage 3 L. L. L. L. L.
guangxiense hainanense huashen mouhoti pullum
L. xanthospilum V. ailaonica V. boringii V. echinata
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
267
Table 5 (continued) Species V. jiulongshanensis V. l. liui V. l. yaoshanensis
V. leishanensis V. ngoclinhensis V. promustache
Iris color
Locality
Voucher [Sample No.] and/or reference
Upper Upper Upper Upper Upper Upper Upper Upper Upper
Vietnam China China China China China China Vietnam China, Vietnam
Ho et al. (1999) Fei et al. (2009) Fei (1999), Fei et al. (2009) Fei (1999) Fei et al. (2009) CIB JX20071196 Fei (1999), Fei et al. (2009) Orlov (2005), Nguyen et al. (2009) Rao et al. (2006), Fei et al. (2009), Bain et al. (2009)
half blue half light greenish white half light green half light green half light greenish white half light-blue half light green half white half light-blue
(Inger, 1954; Inger et al., 1995; Dubois and Ohler, 1998). Because populations from the various islands of the Philippines might differ from the Mindanao population (Inger, 1954), further study is necessary to determine this population’s taxonomic status. Frogs morphologically identified as L. montanum are paraphyletic, and populations from the Kinabalu (19) and Crocker (18) regions of western Sabah (L. montanum lineage 2) form a lineage sister to the clade containing L. montanum from the other parts of Borneo (L. montanum lineage 1) (Fig. 1). Morphologically, L. montanum lineage 2 and L. gunungense from Kinabalu (13, 14) are difficult to distinguish, but their calls differ substantially (Malkmus, 1996; Malkmus et al., 2002; Matsui, unpublished data). Because Malkmus (1996) mainly worked on Kinabalu, he likely studied L. montanum lineage 2. Given an uncorrected p-distance of 5.7–5.8% between L. montanum lineage 2 and L. gunungense, and their different calls, these lineages undoubtedly constitute distinct species. This position is further bolstered by the genealogical relationships because L. gunungense formed a sister group not with L. montanum lineage 2 but with L. abbotti (15–17), a result not expected because L. gunungense is morphologically much more dissimilar to L. abbotti than to L. montanum lineage 2. Specimens from the type locality of L. montanum, Paramasan, southern Kalimantan (1, 2), nested in L. montanum lineage 1 and they are regarded as being true L. montanum. The lineage also contains specimens from eastern Sabah (3, 4), western Sarawak (7, 8), central Kalimantan (5, 6), and western Kalimantan (9). Further morphological and bioacoustic investigations are needed to determine taxonomic identity these populations. A sister lineage, Leptobrachium sp. 1 from Jambi (12) and Lampung (10, 11), Sumatra, surely forms an unnamed species. In these individuals, the iris is blue in color, though grayish in juveniles. This condition is very distinct among Sundaland species of Leptobrachium (Hamidy and Matsui, 2010). Within Subclade II, L. nigrops, the sister species to the monophyletic clade of L. hendricksoni and L. hasseltii, is characterized by having the smallest body size of all Leptobrachium and uniquely pointed finger and toe tips (Inger, 1966; Inger and Stuebing, 1997). Populations from Peninsular Malaysia (37–39) and Belitung Island (35, 36) are sister populations and together they form the sister lineage of the western Bornean (40, 41) population. Genetic differentiation among its allopatric populations is the greatest among all species. Populations from Peninsular Malaysia, Belitung Island, and Borneo have uncorrected p-distances ranging from 9.3% to 11.7%. Their geographical distinctiveness and level of genetic differentiation strongly suggest that each population deserves species recognition. Inger (1966) notes minor morphological differences between Peninsular Malaysian and Bornean populations. Further morphological and acoustic studies will likely verify their taxonomic distinctiveness. Leptobrachium hendricksoni is distinct in having black spots on its venter and an orange-colored upper (and sometimes whole) iris (Taylor, 1962; Inger, 1966; Matsui et al., 1999; Malkmus et al., 2002). On Peninsular Malaysia, this species is paraphyletic with respect to Sumatran populations. Interestingly, populations from the
northern part of Peninsular Malaysia (21–24) are the sister group to the Sumatran populations (25, 26) and together these form the sister group to the southern peninsular populations (27, 28). Genetic diversities in this species are much smaller (uncorrected p-distances <2.3%) than those in L. nigrops, suggesting a unique evolutionary history. In this case, taxonomic recognition of these lineages may not be warranted. Historically, the name L. hasseltii was applied to almost all populations of Leptobrachium (formerly as Megalophrys or Megophrys) ranging from southern China, Indochina, Thailand to northern India, and through Sundaland to the Philippines (e.g., Smith, 1921; Bourret, 1937; Inger, 1954, 1966; Liu and Hu, 1961; Taylor, 1962; Berry, 1975). Matsui (in Frost, 1985) split several populations as distinct species, although some authors have not recognized these changes (e.g., Alcala, 1986; Chanda, 1994; Zheng et al., 2008). The topotypic populations from Java (31–34) and the population from Lampung, southern Sumatra (29, 30) were monophyletic. Here, we have confirmed the occurrence of L. hasseltii on Sumatra and, based on specimen collection records in MZB, the species seems to be restricted to this area. Central and northern Sumatra is occupied by L. hendricksoni. The range of L. hasseltii seems limited to Java, Sumatra, Bali, and several adjacent small islands (Iskandar, 1998). The iris color of L. hasseltii is said to be scarlet (Iskandar, 1998), as cited by Matsui et al. (1999). However, all of our samples, including those used herein, and L. nigrops, have a totally black iris, unlike its sister species L. hendricksoni. Subclade III contained L. smithi from various localities of Thailand, Myanmar (Kyaik Hto) and Malaysia (Langkawi Island), Leptobrachium sp. 3 from Myanmar (Gwa), and Leptobrachium sp. 4 from Thailand (Pilok). Within L. smithi, populations from northeastern Thailand (42, 43), western Thailand (44–47), southern Thailand (48 = topotypic sample from Kaochong), including Langkawi Island (49), and Myanmar (50), formed a clade, but the relationships among them remained unresolved. Intraspecific variation in body size and iris color noted by Matsui et al. (1999) was partially confirmed by us, but these populations did not exhibit significant genetic differentiation between them. Populations from the southern and northern Isthmus Kra occurred in a well-known faunal and floral boundary in Sundaland (e.g., Tougard, 2001; De Bruyn et al., 2005). Leptobrachium sp. 3 from Gwa, Myanmar (51) was the sister lineage to L. smithi. The data used in our analysis (DQ283239) were said to be from L. hasseltii (Frost et al., 2006) yet we could not examine the voucher specimen. This sequence was more likely to have been from L. smithi rather than L. hasseltii. Large genetic uncorrected p-distance (6.0%) with L. smithi from another locality of Myanmar suggested that Leptobrachium sp. 3 represented an undescribed species. Finally, Leptobrachium sp. 4 from Pilok, western Thailand (52) occurred sympatrically with L. smithi (45) and the genetic differentiation between them was great (uncorrected p-distance = 8.1%). Inthara et al. (2005) reported that the larval denticle formula of Leptobrachium sp. from Thon Pha Phum, Kanchanaburi, which is located very close to Pilok varied from 6(2–6)/7(1–6) to 8(2–8/7(1–6) (terminology following Altig and McDiarmid, 1999). The denticle formula for L. smithi ranged
268
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
from 5(2–5)/5(1–4) to 7(2–7)/6(1–5) (Matsui et al., 1999). Unfortunately, we did not have a specimen of Leptobrachium sp. 4. Adult specimens are needed to determine this population’s taxonomic status. 4.2. Taxonomic considerations of Clade B In Subclade IV, Clade B, L. chapaense lineage 2 from Ha Giang, northern Vietnam (66) was the sister group of the other lineages. The population has been suggested to be an undescribed species (Zheng et al., 2008), and our results support this view. In the sister lineage to L. chapaense lineage 2, V. promustache (65) was the sister species to the clade encompassing five remaining lineages. These relationships are completely concordant with those of Zheng et al. (2008). One lineage contained three Chinese sublineages: first, V. jiulongshanensis from Zhejiang Province (56); second, V. leishanensis from Guizhou Province (55); and third, V. l. liui from Mangshan, Hunan Province (53) and V. l. yaoshanensis from Guangxi Province (54). Our results indicated that these four taxa were close genetically to each other (uncorrected p-distances ranged from 1.5% between V. l. liui and V. l. yaoshanensis to 3.0% between V. l. liui and V. jiulongshanensis). Zheng et al. (2008) recovered these three lineages but concluded that V. jiulongshanensis was a junior synonym of V. liui, incorrectly reflecting their phylogeny. Rao and Wilkinson (2008) also obtained the three lineages but found that V. l. liui formed a clade with V. jiulongshanensis and not with V. l. yaoshanensis. Thus, they synonymized V. jiulongshanensis into V. liui, but recognized V. yaoshanensis as a full species. We suggest that additional work is required to confirm the taxonomic status of these taxa. Rao and Wilkinson (2008) proposed the formation of a V. boringiae species group comprised of V. boringiae (=boringii) (61), V. leishanensis, V. l. liui, V. l. yaoshanensis, and V. jiulongshanensis. We did not resolve this group as a monophyletic lineage, and the association of these species was only weakly supported in Rao and Wilkinson’s trees. Our tree supported Fei et al.’s (2009) V. liui species group including V. leishanensis, V. l. liui, V. l. yaoshanensis (V. jiulongshanensis in the synonymy of V. l. liui). However, our analysis did not support their V. boringii species group as including V. boringii, V. ailaonica (64), and V. promustache. Similarly, Rao and Wilkinson’s (2008) V. ailaonica species group, including V. ailaonica, V. echinata (63) and V. ngoclinhensis (72), was not supported; our results, and those of Zheng et al. (2008), did not nest V. ngoclinhensis (Subclade V) with V. ailaonica and V. echinata (Subclade IV). Certainly, V. ailaonica and V. echinata are sister species (Dubois and Ohler, 1998; Ohler et al., 2000; Grosjean, 2001), but there remains an argument as to whether they are conspecific (Ho et al., 1999; Zheng et al., 2008) or not (Ohler et al., 2000; Grosjean, 2001; Rao and Wilkinson, 2008). The uncorrected p-distance between them was only 2.5%, the threshold of specific separation for other taxa. This low level of divergence could reflect isolation-by-distance. Leptobrachium chapaense lineage 1 from Thailand (57, 59) and Yunnan, China (58, 60) and Leptobrachium sp. 5 from Laos (62) comprise the remaining two lineages of Subclade IV. Wide-ranging L. chapaense is surely a composite of more than one species (Rao et al., 2006; Zheng et al., 2008; Rao and Wilkinson, 2008). Our L. chapaense lineage 1 is the sister lineage of the population from Sa Pa, Vietnam, the type locality of L. chapaense (Zheng et al., 2008) and likely it is conspecific. Leptobrachium chapaense clusters within Vibrissaphora and this finding conforms to the morphology of the tadpoles from Yunnan; Yang (1991) notes that Yunnan populations have a Y-shaped yellow marking on their tail, a characteristic of Vibrissaphora (Dubois and Ohler, 1998). Our single male specimen of Leptobrachium sp. 5 from Xam Neua, northeastern Laos likely represents an undescribed species.
Unfortunately, specimens of L. buchardi Ohler, Teynié, and David, 2004 from southern Laos, which has a green colored upper iris, as ascertained from one female specimen, were not available for comparison. The iris color in our male was not recorded and it is larger than the female suggesting, on one hand, the possibility of their conspecificity because of male-biased sexual size dimorphism in Vibrissaphora (Dubois and Ohler, 1998). On the other hand, given that the two specimens were collected more than 600 km in straight-line distance apart, conspecificity seems unlikely. It is also possible that our specimen is conspecific with L. huashen Fei and Ye, 2005 from Yunnan Province, China. However, L. huashen is closely related to L. chapaense (Fei and Ye, 2005; Rao and Wilkinson, 2008). Likewise, L. guangxiense Fei, Mo, Ye, and Jiang in Fei, Hu, Ye, Huang et al. (2009) from Shangsi, Guangxi Province, China was not available. However, based on the original description (Fei et al., 2009), the species seems to be a member of Subclade IV of our study. Thus, we believe it more likely than not that Leptobrachium sp. 5 represents an undescribed species. Two populations of L. chapaense lineage 3 from northern Vietnam (67, 69) and L. hainanense (68) formed a monophyletic lineage, as previously reported (Zheng et al., 2008). Lathrop et al. (1998) noted that the dorsal half of iris is white in specimens from the Tam Dao (67) population and not light-blue as in L. hainanense (Liu et al., 1973 as V. hasseltii; Matsui and Ota, 1995; Fei et al., 2009; see below). We, however, regard this difference to be within the normal range of variation, and tentatively conclude that the population from Tam Dao, as well as Ben En, should be treated as L. hainanense. The uncorrected p-distance between these two localities and L. hainanense from Hainan Island were 1.8% and 3.0%, respectively. The latter value is comparable to the distance (3.0%) found between L. abbotti and L. gunungense from Borneo. Further work is necessary to confirm taxonomic relationships and identity of these populations. Unlike the result of Zheng et al. (2008), our study resolved Vietnamese L. mouhoti (70) as the sister species of L. pullum (71) and not V. ngoclinhensis (72). As with Zheng et al.’s (2008) study, V. ngoclinhensis fell in Subclade V and this association differs from Rao and Wilkinson’s (2008) suggestion that the species is associated with Vibrissaphora (Clade IV) based on morphological data. Although L. pullum is morphologically similar to L. smithi (Matsui et al., 1999), these two species were distantly related genealogically, being placed in Subclades V and III, respectively. Finally, Vietnamese L. xanthospilum (73) and L. banae (74) represented the remaining two lineages in Subclade V and our study supported their specific status (Lathrop et al., 1998). 4.3. Taxonomic relationship of Leptobrachium and Vibrissaphora Our results for the relationships of Leptobrachium and Vibrissaphora were concordant with other recent molecular studies (Fu et al., 2007; Zheng et al., 2008; Rao and Wilkinson, 2008). As summarized by Dubois and Ohler (1998), taxonomic recognition of Leptobrachium and Vibrissaphora as genera, subgenera, or synonyms is controversial. Some authors (e.g., Fei et al., 2009) still treat them as different genera, while the others (e.g., Dubois and Ohler, 1998; Ohler et al., 2004; Delorme et al., 2006) regard them as subgenera of Leptobrachium. Zheng et al. (2008) rejected the monophyly of both subgenera and relegated Vibrissaphora to synonymy in Leptobrachium without recognizing subgenera, as originally proposed by Dubois (1983). The taxonomic conclusions of Zheng et al. (2008), now followed by Frost (2009), requires reassessment because their result seems to have been affected by inadequate sampling in Sundaland. In all of our trees, the Sundaland species plus L. smithi (Clade A, Subclades I–III) were distinctly separated from Chinese and Indochinese species (Clade B, Subclades IV and V). Clades A, B, and C in Fig. 3 of
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
269
surveyed. Although detailed morphological study is necessary, a more broadened and flattened male upper lip of Leptobrachium (Vibrissaphora) may be a synapomorphy, and a character state that separates them from Leptobrachium (Leptobrachium), as observed between L. pullum and L. smithi (Matsui et al., 1999). The presence of reversed sexual size dimorphism (Dubois and Ohler, 1998) and a Y-shaped yellow marking on the tail of tadpoles should be also reassessed using a sufficient number of samples. 4.4. Evolutionary history of Leptobrachium
Fig. 3. Iris color in frogs of the genus Leptobrachium mapped onto the maximum likelihood tree given in Fig. 2. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this paper.)
Zheng et al. (2008) corresponds to our Subclades IV, V, and I + III, respectively. They found either, Subclades V and I + III formed a monophyletic group, which was sister to Subclade IV (MP tree), or these three lineages were trichotomic (BI tree). Our results rejected both sets of relationships, and hence their conclusion. In order for the taxonomy to reflect historical relationships, we recognize Clades A and B as subgenera. Because the current megophryid taxonomy tends to recognize many genera that are neither sufficiently studied phylogenetically nor defined clearly morphologically, we refrain treating them as genera. Subgenus Leptobrachium (Clade A; type species L. hasseltii) can be characterized as the most inclusive clade containing species from the Philippines, Malaysia, Indonesia, Thailand, Myanmar, and India, but exclusive of L. chapaense lineage 1 from Thailand. Subgenus Vibrissaphora (Clade B; type species V. boringii) is the most inclusive clade that contains species from Indochina and China, and including L. (V.) chapaense lineage 1 from Thailand. Traditionally, male Vibrissaphora have been regarded as possessing nuptial spines on their upper labium during the breeding season and as having larger body sizes than females. Previously, these characteristics served to separate Vibrissaphora from Leptobrachium (see review by Dubois and Ohler (1998)). However, the character labial spines no longer defines the group. The species L. banae, L. chapaense (for three lineages), L. guangxiense, L. hainanense, L. mouhoti, L. pullum, L. xanthospilum, and Leptobrachium sp. 5 grouped in Clade B. In our new classification, these species, although lacking labial spines, should be placed in Leptobrachium (Vibrissaphora) together with V. ngoclinhensis. In this arrangement, the morphological diagnosis of Vibrissaphora is unclear and new characteristics should be
Divergence time estimations can provide valuable insights into the biogeographic history of species and regions. Often the estimated dates have extensive variance and overlap, yet they can serve as biogeographical hypotheses in the absence of detailed geological information, and can even predict unknown geological events (Lindell et al., 2006; Riddle et al., 2008). Time estimates (Table 4) suggested that Leptobrachium (sensu lato) and Oreolalax formed in the early Eocene, and subsequently diverged into two major lineages (Clades A and B) in the mid Eocene. No stratigraphic events have been associated with the origin of Clades A and B and the subsequent, nearly simultaneous divergences within them starting from the late Eocene to early Oligocene. However, the uplift of the Tibetan Plateau initiated about 40 MYBP (Zhao and Morgan, 1985; Chung et al., 1998) and this caused changes in the Mekong drainage system (Brookfield, 1998). This orogenesis is likely related to the speciation events. Estimates for the age of Sundaland Leptobrachium predate the geological separation of land masses in this region. For example, the Sunda Shelf became land-positive, forming a continuous land connection between Java and mainland Southeast Asia, including some parts of Indochina, during periods of lowered sea levels in the Pleistocene (<1.6 MYBP; Voris, 2000)). This very recent stratigraphic history is responsible for the recent cladogenic events as suggested by low levels of intraspecific genetic differentiation in Leptobrachium but it cannot account for older events. Most of the speciation is estimated to have occurred sometime between the late Oligocene to late Miocene (30.0–14.0 MYBP) in Clade A. The separation of L. nigrops from its common ancestor with L. hendricksoni and L. hasseltii likely occurred in the late Oligocene, and speciation of the latter two species in the early Miocene. These ages are comparable with those estimated for Clade B. The diversification of L. banae from other Vietnamese and Hainan Leptobrachium, and for Ha Gian L. chapaense lineage 2 from its sister taxa, likely occurred in the early Miocene. Leptobrachium sp. 2 from Mindanao, Philippines was estimated to have diverged from Bornean and Sumatran relatives at 19.0 (CI 27.4–11.3) MYBP. This estimate does not contradict the idea that Mindanao became aerial about 25–30 MYBP in the late Oligocene (Brown and Alcala, 1994), but the route of faunal invasion remains unclear (Inger, 1999). Inger and Voris (2001) showed that disjunction of the Malay Peninsula from Borneo by the South China Sea (Fig. 1) occurred once about 5 MYBP in the early Pliocene. This age corresponds to the date of intraspecific divergence estimated in L. hendricksoni for populations in the northern Malay Peninsula and Sumatra. Although specimens from Sarawak, western Borneo (Inger, 1966) were not available, separation from the peninsular population is expected at this age. ‘‘Intraspecific” divergence in L. nigrops is much older than that observed in the other species (Table 4). The mid Miocene date is comparable to or even older than that in other lineages (see below). Inger (2005) surmised that the long-term continental connection between Borneo and the Malay Peninsula (50–5 MYBP) was once interrupted in the mid Miocene, 15 MYBP. This event might have isolated Bornean L. nigrops from Malay Peninsular
270
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
and Belitung populations. Notwithstanding their morphological similarity, each of the populations currently called L. nigrops is likely a distinct species. The occurrence of L. nigrops on Belitung Island suggests the possibility of ‘‘land bridge” invasions by Leptobrachium between Mainland Asia and Borneo. Leptobrachium hendricksoni and L. nigrops are lowland species that also occur in coastal swamps. Their ecology led Inger and Voris (2001) to associate them with Pleistocene dispersal between the Malay Peninsula and Borneo, while questioning their absence on Sumatra. The presence of L. hendricksoni on Sumatra is now confirmed, and its unique ecology does not seem to be related to Pleistocene dispersal. Leptobrachium hasseltii occurs at slightly higher elevations than does L. hendricksoni. On Java and southern Sumatra, we found L. hasseltii at altitudes of 300–1500 m a.s.l. However, this altitudinal distribution may have been secondarily acquired. Within Java, most lowland forests have been logged and no longer exist. Only a few mountain forests remain today in the form of isolated habitat islands without corridors between them. This condition probably has prevented gene flow between populations from western and central Java. Diversification in L. hasseltii is estimated to have occurred in the mid Pliocene (Table 4). Available information on distributional pattern of this species is limited, but based on collection records in MZB, the westernmost distribution of L. hasseltii is the southern tip of Sumatra (Lampung), and the easternmost range includes the islands of Bali and Kangean (Iskandar, 1998), which is the southeasternmost range of the genus. The Lombok Strait, i.e., Wallace’s Line, separating Bali and Lombok, seems to have been the primary barrier preventing the invasion of Leptobrachium to Wallacea. Parapatric species of Leptobrachium occur on Borneo, where speciation occurred mostly in the Miocene, starting with the mid Miocene separation of L. montanum lineage 2 from the common ancestor of the other species. Speciation continued with the split of ancestors of L. montanum lineage 1 and Leptobrachium sp. 1, and slightly later the common ancestor of L. abbotti and L. gunungense. This was followed by speciation of L. abbotti and L. gunungense in the late Miocene. Intraspecific divergence occurred at the Miocene–Pliocene boundary as shown by L. montanum lineage 1. These very old dates of divergence in Borneo, especially on Mt. Kinabalu, are comparable to those of bufonid toads in the genus Ansonia (Matsui et al., 2010). The boundary of upland L. montanum (two lineages combined) and lowland L. abbotti occurs at elevations from 800 to 1000 m a.s.l. (Inger et al., 1995). On Mt. Kinabalu, a narrow zone of sympatry occurs between L. gunungense (1750–2200 m) and L. montanum (900–1750 m), most probably our lineage 2 (Malkmus et al., 2002; our own observation). The third species, L. abbotti, occurs at much lower elevations (500–900 m). Like the species in Borneo, Leptobrachium sp. 1 was found parapatric with L. hasseltii (Hamidy and Matsui, 2010). Interestingly, these sets of species are from different lineages. Non-sister lineages may more easily segregate their niches than the sister taxa, which might be more similar in niche requirements. Nearly simultaneous dates of divergence were estimated in the predominantly Thai clade, where splits occurred between Leptobrachium sp. 4 and the common ancestor of L. smithi and Leptobrachium sp. 3 in the early Miocene. Another speciation event occurred between L. smithi and Leptobrachium sp. 3 in the late Miocene. Subsequent intraspecific divergence in L. smithi was estimated to have begun in the early Pliocene. The Isthmus of Kra, peninsular Thailand, is usually noted as a major biogeographical transition zone (e.g., Tougard, 2001; De Bruyn et al., 2005), and it might have also affected the evolution of Leptobrachium. However, the specific date(s) of the disjunction is (are) not clear and simply estimated to have occurred as a result of a Neogene (<23 MYBP) marine transgression.
Diversification in the Indochinese–Chinese area initiated in the Miocene (Table 4). In the Indochinese clade, L. xanthospilum split from the common ancestor of other Vietnamese and Hainan Leptobrachium and V. ngoclinhensis in the early Miocene. In the Chinese clade, V. promustache and V. liui diverged in the late Miocene. Intraspecific divergence in these regions occurred mostly in the Pliocene. Our estimated date of separation between L. hainanense and Vietnamese L. chapaense lineage 3, 2.6 (CI 0.7–4.7) MYBP, closely conforms to the date of separation of these regions in the early Pleistocene, about 2 MYBP (Chang et al., 2008). In the Indochinese clade, the keratinized nuptial spines on the upper jaw of males seem to have been acquired after the middle Miocene, 15.7 (CI 7.8–24.9) MYBP by V. ngoclinhensis only. In the Chinese clade, the spines seem to have evolved in the late Miocene, 9.7 (CI 3.6–16.7) MYBP, after the basal split of L. chapaense lineage 2, which lacks spines. Subsequently, the spines were secondarily lost in the ancestor of L. chapaense (=lineage 1). As indicated by the different patterns of spine arrangement, labial spines likely have evolved independently in Chinese and Indochinese (V. ngoclinhensis: Orlov, 2005; Rao et al., 2006) clades. Each subclade is estimated to have an old evolutionary history. Divergence times are not easily explained by the geohistory of Southeast Asian major landmasses alone. In order to test hypotheses posed here, additional detailed paleogeographic evidence is necessary.
4.5. Evolution of iris color Although the data are limited, the color of the iris seems to vary ontogenetically (Table 5). The iris color of L. hendricksoni in Malay Peninsular populations varies from being totally orange to having color only on the upper part. This appears to be intraspecific variation as individuals from these populations are monophyletic and conspecific. Similarly, in Leptobrachium sp. 1, juvenile grey iris transforms to blue in adults. Further, variation in iris color ranges from being light yellow to orange in L. smithi (Matsui et al., 1999), and specimens with differently colored iris are also judged genetically to be conspecific. The color of the upper part of the iris in Indochinese–Chinese species (Clade B in this study) has been variously described as being white and sky-blue, as well as skyblue and lime, and lime and green (see Table 5). This continuous range of color could be grouped as light-colored upper iris. Although iris color varies intraspecifically, the evolution of the character can be estimated by the mapping generalized states in adults on the phylogeny (Fig. 3). Matsui et al. (1999) once surmised that the light-colored upper iris might be the primitive condition in Leptobrachium, but their hypothesis was not confirmed in this study. From the distribution of iris color on the tree (Fig. 3), a totally black iris is the plesiomorphic state within Clade A. Within Clade B, the plesiomorphic state is one of having a light color in the upper half of the iris. The yellow to scarlet color in the upper half of the iris, and sometimes also in the lower half, is regarded as being apomorphic, having evolved independently in three subclades. A totally light-colored iris also appears to be an apomorphic character in Clade A. Because no species of the sister genus Oreolalax is known to have a bicolored or totally dark iris (Liu and Hu, 1961; Zhao and Adler, 1993; Fei, 1999; Yang and Rao, 2008), the plesiomorphic state in Leptobrachium cannot be identified. Although iris color surely varies, even within a species, the range of variation seems to be limited. Thus, the character appears to be useful for species identification and in estimating phylogenetic relationships. In order to test hypotheses on the evolution of iris color posed here, data are required from taxa whose character states remain unknown.
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
Acknowledgments MM is grateful to the following for their encouragements and/or permission to conduct research and field companionship: H. Akiyama, L. Apin, K. Araya, A.-A. Hamid, T. Hikida, H. Ota, H. Kassim, J.J. Kendawong, K.B. Kueh, T. Kusano, D. Labang, M.B. Lakim, M. Maryati, the late J. Nabitabhata, K. Nishikawa, S. Panha, L.-H. Seng, A. Sudin, T. Sugahara, T. Tachi, M. Toda, and N.-S. Wong. MM is also indebted to A. Ohler and M. Delorme for exchanging tissue samples, and N. Kuraishi and N. Yoshikawa for laboratory assistance. We also thank two anonymous reviewers for improving an early version of the manuscript. The National Research Council of Thailand, the Royal Forest Department of Thailand, the Economic-Planning Unit (former Socio-Economic Research Unit) of Malaysia, the State Government of Sarawak, and Sabah Parks kindly permitted MM to conduct the project, and Chulalongkorn University, University Malaya, Universiti Malaysia Sabah, Universiti Kebangsaan Malaysia (UKM), JICA, and the Forest Department, Sarawak kindly provided all the facilities for conducting research. Field trips by MM were made possible by grants from The Monbusho International Scientific Research Program (Field Research, 01041051, 02041051, 04041068, 06041066, and 08041144), The Monbukagakusho through the Japanese Society for the Promotion of Sciences (JSPS: Field Research, 15370038, 20405013), UKM (OUPPLW-14-59/2008), and TJTTP-OECF. AH thanks M.D. Kusrini, S. Kirono, G.B. Mandigani, D. Susanto, M. Harvey, E. Smith, K.L. Sanders, T.Q. Nguyen, D.A. Anggraeni, and J. Rhamadani for providing tissue samples and to the Monbukagakusho for scholarship funding. RWM Funding for fieldwork was obtained through grants from the Natural Sciences and Engineering Research Council, Canada (Discovery Grant A3148), the Royal Ontario Museum (ROM) Foundation, and the ROM Volunteers Fieldwork Committee. All fieldwork was conducted using approved Animal Use Protocols. Fieldwork in Vietnam was assisted by N.L. Orlov, A. Lathrop, R.H. Bain, V.S. Cao, T.C. Ho and V.S. Nguyen; export permits were obtained with the assistance of V.S. Cao and the Institute of Ecology and Biological Resources, Hanoi. Specimens in Hainan Island were obtained while assisting H.-T. Shi, Hainan Normal University, and those from Yunnan were obtained with the assistance of D.-Q. Rao, Kunming Institute of Zoology and N.L. Orlov; research export permits were obtained through the Chinese Academy of Sciences.
References Alcala, A.C., 1986. Guide to Philippine Flora and Fauna 10, Amphibians and Reptiles. Xiv+195 p. Natural Resources Management Center, Ministry of Natural Resources and University of the Philippines, Quezon City. Altig, R., McDiarmid, R.W. (Eds.), 1999. Body plan: development and morphology. In: Tadpoles. University of Chicago Press, Chicago, pp. 24–51. Bain, R.H., Nguyen, T.Q., 2004. Herpetological diversity of Ha Giang province in northeastern Vietnam, with description of two new species. Am. Mus. Novit. 3452, 1–42. Bain, R.H., Nguyen, T.Q., Doan, K.V., 2009. First record of Leptobrachium promustache from Vietnam. Herpetol. Notes 2, 27–29. Berry, P.Y., 1975. The Amphibian Fauna of Peninsular Malaysia. Tropical Press, Kuala Lumpur. Bourret, R., 1937. Notes herpétologiques sur l’Indochine Française XIV–XV. Annexe Bull. Gén. Instr. Publique, Hanoi 1937, 1–80. Brookfield, M.E., 1998. The evolution of great river systems of southern Asia during the Cenozoic India–Asia collision: rivers draining southwards. Geomorphology 22, 285–312. Brown, R.M., Diesmos, A.C., 2009. HerpWatch Philippines: an online biodiversity information product for Philippine herpetological diversity. Available from:
271
Cannatella, D.C., Hillis, D.M., Chippindale, P.T., Weght, L., Rand, A.S., Ryan, M.J., 1998. Phylogeny of frogs of the Physalaemus pustulosus species group, with an examination of data incongruence. Syst. Biol. 47, 311–335. Chanda, S.K., 1994. Anuran (Amphibia) fauna of northeast India. Mem. Zool. Surv. India 18, 1–143. Chang, J., Wang, B., Zhang, Y.-Y., Liu, Y., Liang, W., Wang, J.-C., Shi, H.-T., Su, W.-B., Zhang, Z.-W., 2008. Molecular evidence for species status of the endangered Hainan peacock pheasant. Zool. Sci. 25, 30–35. Chung, S.L., Lo, C.H., Lee, T.Y., Zhang, Y.Q., Xie, Y.W., Li, X.H., Wang, K.L., Wang, P.L., 1998. Diachronous uplift of the Tibetan plateau starting 40 Myr ago. Nature 394, 769–773. De Bruyn, M., Nugroho, E., Hossain, Md.M., Wilson, J.C., Manther, P.B., 2005. Phylogeographic evidence for the existence of an ancient biogeographic barrier: Isthmus of Kra Seaway. Heredity 94, 370–378. Delorme, M., Dubois, A., Grosjean, S., Ohler, M., 2006. Une nouvelle ergotaxinomie des Megophryidae (Amphibia, Anura). Alytes 24, 6–21. Drummond, A.J., Ho, S.Y.W., Philips, M.J., Rambaut, A., 2006. Relaxed phylogenetics and dating with confidence. PLoS Biol. 4, 699–710. Drummond, A.J., Rambaut, A., 2007. BEAST, Bayesian Evolutionary Analysis Sampling Trees, version 1.4.2. Available from:
272
M. Matsui et al. / Molecular Phylogenetics and Evolution 56 (2010) 259–272
Inger, R.F., 1999. Distribution of amphibians in Southern Asia and adjacent islands. In: Duellman, W.E. (Ed.), Patterns of Distribution of Amphibians: A Global Perspective. Johns Hopkins Univ. Press, Baltimore and London, pp. 445–482. Inger, R.F., 2005. The frog fauna of Indo-Malayan region as it applies to Wallace’s line. In: Tuen, A.A., Das, I. (Eds.), Wallace in Serawak-150 years later, an international conference on biogeography and biodiversity. Institute of Biodiversity and Environmental Conservation. Universiti Malaysia Sarawak, Kota Samarahan, pp. 82–90. Inger, R.F., Stuebing, R.B., 1997. A Field Guide to the Frogs of Borneo. Nat. Hist. Publ. Sdn. Bhd, Kota Kinabalu. Inger, R.F., Voris, H., 2001. The biogeographical relations of the fogs and snakes of Sundaland. J. Biogeogr. 28, 863–891. Inger, R.F., Stuebing, R.B., Tan, F.-L., 1995. New species and new records of anurans from Borneo. Raffles Bull. Zool. 43, 115–131. Inthara, C., Lauhachinda, V., Nabhitabhata, J., Chuaynkern, Y., Kumtong, P., 2005. Mouth part structures and distribution of some tadpoles from Thailand. Thailand Nat. Hist. Mus. J. 1, 55–78. Iskandar, D.T., 1998. The Amphibians of Java and Bali. Puslitbang Biologi, LIPI, Bogor. Jobb, G., von Haeseler, A., Strimmer, K., 2004. Treefinder: a powerful graphical analysis environment for molecular phylogenetics. BMC Evol. Biol. 4, 18. Lathrop, A., 1997. Taxonomic review of the megophryid frogs (Anura: Pelobatoidea). Asiatic Herpetol. Res. 7, 68–79. Lathrop, A., Murphy, R.W., Orlov, N.L., Ho, C.T., 1998. Two new species of Leptobrachium (Anura: Megophryidae) from the central highlands of Vietnam with a redescription of Leptobrachium chapaense. Russ. J. Herpetol. 5, 51–60. Leaché, A.D., Reeder, T.W., 2002. Molecular systematics of the eastern fence lizard (Sceloporus undulatus): a comparison of parsimony, likelihood, and Bayesian approaches. Syst. Biol. 51, 44–68. Lindell, J., Ngo, A., Murphy, R.W., 2006. Deep genealogies and the mid-peninsular seaway of Baja California. J. Biogeogr. 33, 1327–1331. Liu, C., Hu, S., 1961. Tailless Batrachians of China. Science Press, Beijing. Liu, C.-C., Hu, S.-C., Fei, L., Huang, C.-C., 1973. On collections of amphibians from Hainan Island. Acta Zool. Sin. 19, 385–404. Malkmus, R., 1996. Leptobrachium gunungense sp. N. (Anura: Pelobatidae) vom Mount Kinabalu, Nord Borneo. Mitt. Zool. Mus. Berlin 72, 297–301. Malkmus, R., Manthey, U., Vogel, G., Hoffman, P., Kosuch, J., 2002. Amphibians and Reptiles of Mount Kinabalu (North Borneo). ARG Gantner Verlag Kommanditgesellschaft, Ruggell. Manthey, U., Grossmann, W., 1997. Amphibien & Reptilien Südostasiens. Natur und Tier Verlag, Münster. Matsui, M., Ota, H., 1995. On Chinese herpetology. Herpetologica 51, 234–250. Matsui, M., Nabhitabhata, J., Panha, S., 1999. On Leptobrachium from Thailand with a description of a new species (Anura: Pelobatidae). Jpn. J. Herpetol. 18, 19–29. Matsui, M., Ito, H., Shimada, T., Ota, H., Saidapur, S.K., Khonsue, W., Tanaka-Ueno, T., Wu, G., 2005a. Taxonomic relationships within the pan-oriental narrow-mouth toad Microhyla ornata as revealed by mtDNA analysis (Amphibia, Anura, Microhylidae). Zool. Sci. 22, 489–495. Matsui, M., Shimada, T., Ota, H., Tanaka-Ueno, T., 2005b. Multiple invasions of the Ryukyu Archipelago by Oriental frogs of the subgenus Odorrana with phylogenetic reassessment of the related subgenera of the genus Rana. Mol. Phylogenet. Evol. 37, 733–742. Matsui, M., Tominaga, A., Liu, W.-Z., Khonsue, W., Grismer, L.L., Diesmos, A.C., Das, I., Sudin, A., Yambun, P., Yong, H.-S., Sukumaran, J., Brown, R.M., 2010. Phylogenetic relationships of Ansonia from Southeast Asia inferred from mitochondrial DNA sequences: systematic and biogeographic implications (Anura: Bufonidae). Mol. Phylogenet. Evol. 54, 561–570. Nguyen, V.S., Ho, T.C., Nguyen, Q.T., 2009. Herpetofauna of Vietnam. Edition Chimaira, Frankfurt am Main. Nylander, J.A.A., 2004. MrModeltest v2. Program distributed by the author. Evolutionary Biology Centre, Uppsala University.
Ohler, A., Marquis, O., Swan, S., Grosjean, S., 2000. Amphibian biodiversity of Hoang Lien Nature Reserve (Lao Cai Province, northern Vietnam) with description of two new species. Herpetozoa 13, 71–87. Ohler, A., Teynie, L.A., David, P., 2004. A green-eyed Leptobrachium (Anura: Megophryidae) from southern Laos. Raffles Bull. Zool. 52, 695–700. Orlov, N.L., 2005. A new species of the genus Vibrissaphora Liu, 1945 (Anura: Megophryidae) from Mount Ngoc Linh (Kon Tum Province) and analysis of the extent of species overlap in the fauna of amphibians and reptiles of the northwest of Vietnam and central highlands. Russ. J. Herpetol. 12, 17–38. Palumbi, S.R., Martin, A., Romano, S., McMillan, W.O., Stice, L., Grabowski, G., 1991. The Simple Fool’s Guide to PCR, Version 2.0. University of Hawaii, Honolulu. Rambaut, A., Drummond, A.J., 2007. Tracer version 1.4. Available from: