This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright Author's personal copy Molecular Phylogenetics and Evolution 61 (2011) 854–865 Contents lists available at SciVerse ScienceDirect Molecular Phylogenetics and Evolution journal homepage: www.elsevier.com/locate/ympev Phylogeny of the Southeast Asian freshwater fish genus Pangio (Cypriniformes; Cobitidae) Jörg Bohlen a,⇑, Vendula Šlechtová a, Heok Hui Tan b, Ralf Britz c a Institute of Animal Physiology and Genetics, Laboratory of Fish Genetics, Rumburská 89, 27 721 Liběchov, Czech Republic National University of Singapore, Department of Biological Sciences, Singapore 117600, Singapore c Natural History Museum, Cromwell Road, SW7 5BD London, United Kingdom b a r t i c l e i n f o Article history: Received 15 March 2011 Revised 1 August 2011 Accepted 2 August 2011 Available online 16 August 2011 Keywords: Pangio Eel loaches Phylogeny Phylogeography Species group Sundaland a b s t r a c t The genus Pangio is one of the most species-rich of the loach family Cobitidae and widespread across South and Southeast Asia. Its species diversity has never been studied under a clear phylogenetic approach, but four ‘species-groups’ were proposed according to the most obvious morphological characters. We present here phylogenetic analyses of the genus Pangio based on sequence data of the mitochondrial cytochrome b gene, the nuclear recombination-activating gene 1 (RAG 1) and a combined dataset of 109 specimens from 18 morphologically identified species across the whole distribution area of the genus. Our data reveal the existence of three major lineages within Pangio. Two of our major lineages were congruent with formerly proposed species-groups, the remaining two species-groups together formed the third major lineage; herein we refer to the lineages as to anguillaris-group, kuhlii-oblonga group and shelfordii-group. The application of a molecular clock dated the age of the three lineages to 33–29 million years. At the species level, our data suggest about 30 distinct lineages, indicating that there is a high number of undescribed species within Pangio. The use of Pangio to address biogeographic questions is demonstrated with the example of the shelfordii-group, which is distributed across Sundaland. Ó 2011 Elsevier Inc. All rights reserved. 1. Introduction The freshwater fish family Cobitidae represents a characteristic element of the Eurasian ichthyofauna with about 19 genera in nearly all water systems from Portugal to Japan and from Siberia to Java (Bănărescu, 1990; Nelson, 2006). The greatest diversity of Cobitidae is found in East and Southeast Asia, where representatives of at least 16 of these genera occur. As recently shown, the taxa from Europe, northern and East Asia form a distinct, monophyletic group within a diverse assemblage of well-differentiated genera that are distributed in South and Southeast Asia (Šlechtová et al., 2008). Most genera of the European-East Asian lineage have already been studied from a phylogenetic and biogeographic point of view (Ludwig et al., 2001; Perdices and Doadrio, 2001; Bohlen et al., 2006a, 2007; Perdices et al., 2008) and demonstrated their suitability as models to reflect geologic events and the biogeographic history of freshwater systems in their phylogenies. In contrast, only very little is known about the intrageneric phylogeny of South and Southeast Asian cobitids, despite the fact that a sound phylogenetic hypothesis is the key for any future progress in the general understanding of the biogeography of Southeast Asia. ⇑ Corresponding author. Fax: +420 315 639 510. E-mail address: bohlen@iapg.cas.cz (J. Bohlen). 1055-7903/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.ympev.2011.08.003 The largest genus within the southern lineages of Cobitidae is Pangio Blyth, 1860 with currently 32 recognised species (Britz and Maclaine, 2007; Eschmeyer, 2010; Kottelat and Lim, 1993; Tan and Kottelat, 2009; Hadiaty and Kottelat, 2009; Britz and Kottelat, 2010). These small (<12 cm total length) freshwater fishes are commonly referred to as ‘eel loaches’ due to their very elongated body. Some eel loaches are well known and valued ornamental fish, such as the ‘kuhli loaches’, a complex of species including Pangio kuhlii. Eel loaches live in benthic substrates, typically in submerged leaf litter and dense aquatic vegetation, in slow to moderately flowing streams. Often two or three species co-occur at the same locality and up to seven species may be found within a single river basin (Ng and Tan, 1999; Kottelat and Widjanarti, 2005). At present, no detailed hypothesis has been proposed regarding the intrageneric phylogeny of the genus Pangio. Burridge (1992) recognised two ‘species groups’, viz. the ‘kuhlii group’ and the ‘shelfordii group’. In the most comprehensive work on eel loaches to date, Kottelat and Lim (1993) adopted the kuhlii- and shelfordii groups of Burridge (1992), but defined two additional species-groups: the ‘oblonga-group’ and the ‘anguillaris-group’. The anguillaris-group sensu Kottelat and Lim (1993) is defined by a vermiform body with high vertebral count (62–71) and consists of Pangio anguillaris, Pangio bitaimac, Pangio doriae, Pangio lidi, and possibly Pangio lumbriciformis and Pangio signicauda. The kuhlii-group includes all species with a dark brown or black banding pattern on yellow or Author's personal copy J. Bohlen et al. / Molecular Phylogenetics and Evolution 61 (2011) 854–865 red background including the well-known ‘kuhli-loach’, Pangio kuhlii. Species in the oblonga-group have a plain body colouration, moderate vertebral counts (45–51) and adipose keels on the caudal peduncle (Burridge, 1992; Kottelat and Lim, 1993). The species in the shelfordii-group are characterised by a pair of labial barbels (Kottelat and Lim, 1993). However, some species cannot easily be assigned to any of the species groups, since they exhibit characters of two groups, e.g. Pangio superba, which has labial barbels like the shelfordii-group plus a kind of black–red colouration that resembles species of the kuhlii-group (Y-shaped dark bands on yellow–orange ground); Pangio pulla was placed in the anguillaris-group due to its high vertebral count but also resembles the kuhlii-group in its black–red banding (Kottelat and Lim, 1993; Britz and Maclaine, 2007). Monophyly of the species groups proposed by Kottelat and Lim (1993) has so far not been tested, but a basic understanding of the species-level phylogeny of the genus is a prerequisite for any further studies on their evolution and biogeography. The genus Pangio is the most widely distributed of all southern lineages of Cobitidae (Bănărescu, 1990). Its distribution area stretches over most of Mainland SE Asia (Cambodia, Laos, Myanmar, Thailand, Vietnam), the Malay Peninsula (western Malaysia, southern Thailand, Singapore), the Sunda Islands (Borneo, Java, Sumatra) and reaches with three species into India (Menon, 1992; Kottelat and Lim, 1993; Britz and Maclaine, 2007). Due to their wide distribution and frequent occurrence, eel loaches might represent an interesting freshwater fish group for biogeographic studies. Such a model would be desirable, since there are many uncertainties in our present understanding of the biogeographic history of SE Asian freshwater fauna. For the mainland part of SE Asia from India to Vietnam, effects caused by plate tectonics were the most important geologic factor influencing the biogeography of freshwater animals during the Cenozoic (Clark et al., 2004; Brookfield, 1998). In contrast, the Malay Peninsula and Sunda Islands were tectonically more stable, but were heavily influenced by fluctuations in the sea water level (Woodruff, 2003). On the one hand, two periods of increased sea level were identified during early/ middle Miocene (24–13 MYA) and early Pliocene (5.5–4.5 MYA), where global sea level rose more than 100 m above the present level and resulted in extensive flooding of lowland areas (Woodruff, 2003). On the other hand, during the last 30 MYA and especially during Pleistocene glacial maxima, sea levels were repeatedly lower than present, sometimes up to 120 m below the present level (Voris, 2000). Nevertheless, it is still unclear, what the impact of the different sea level fluctuations has been on the biogeography of SE Asian freshwater biota. In the present study, we reconstructed for the first time a phylogenetic hypothesis for the genus Pangio using sequence data of two genes, the mitochondrial cytochrome b and the nuclear recombination activating gene (RAG-1). The reconstructed phylogeny was used to test the value of the species-group concept for the classification of eel loaches. Moreover, we analysed the data in light of species diversity in order to evaluate the present understanding about number and limits of species. Finally, by using a calibrated ultrametric tree based on our molecular analyses we evaluated the impact of past geological events on the biogeography of the genus Pangio. 2. Material and methods 2.1. Taxon sampling and identification Our study includes 109 samples representing 18 out of 32 currently recognised species of Pangio. A detailed list of samples is given in Table 1. Fishes were identified according to Kottelat and Lim (1993), Britz and Maclaine (2007), Tan and Kottelat (2009) 855 and by direct comparison with the following material: Pangio semicincta BMNH 1938.12.1.113 (holotype), BMNH 1938.12.1.114-115, BMHN 1940.2.8.2-3 (paratypes); Pangio malayana BMNH 1957.1.23.1 (holotype), BMNH 1957.1.23.2-3 (paratypes); Pangio oblonga BMNH 2001.1.15.7915-7926. Among the banded eel-loaches, the separation of Pangio semicincta from Pangio kuhlii was hampered by the lack of reliably identified material of Pangio kuhlii; we here refer to the relevant samples as Pangio kuhlii/semicincta. 2.2. Molecular techniques We used the mitochondrial cytochrome b gene (cyt b) and the nuclear recombination-activating gene 1 (RAG 1) to infer phylogenetic relationships within the genus Pangio. It has been shown that the nuclear RAG 1 is a useful marker in reconstructions of fish phylogenies on generic level as well as on the level of species-groups (e.g. Lopéz et al., 2004; Li and Ortí, 2007; Perdices et al., 2005; Rüber et al., 2004; Sullivan et al., 2006; Šlechtová et al., 2007, 2008). The mitochondrial cyt b is one of the most commonly used markers in animal phylogenies and best suited for studies on intraspecific level or among closely related species (e.g. Durand et al., 1999; Bohlen et al., 2006a, 2006b; Perdices et al., 2003; Šedivá et al., 2008). Therefore the combination of the two markers should bring good resolution in the older as well as younger evolutionary events within eel loaches. Moreover, a comparison of the maternally inherited mitochondrial with the nuclear gene can indicate hybridisation events, even if these events have occurred many generations ago (Šlechtová et al., 2008). Further, the clonally inherited cyt b gene fulfils better the requirements of the molecular clock hypothesis and has often been used to recalculate the age of evolutionary events (Doadrio and Perdices, 2005; Pérez et al., 2007; Rüber et al., 2004; Šlechtová et al., 2008). The fixation of material, DNA isolation, PCR (including primer selection), sequencing as well as alignment methods followed Šlechtová et al. (2006, 2007, 2008). For 14 specimens of Pangio, sequences of RAG 1 and cyt b have been retrieved from our recent studies (Šlechtová et al., 2007, 2008); all remaining sequences are original data. 2.3. Data analyses Chromatograms were assembled in SeqMan II (Lasergene, DNAStar). The sequences were checked for unexpected stop codons, aligned to each other using Clustal W algorhitm and later refined manually in BioEdit 7.0.5.3 (Hall, 1999). Alignments are deposited at http://purl.org/phylo/treebase/phylows/study/TB2: S11806. According to Šlechtová et al. (2008), the sister genus of Pangio is Lepidocephalichthys, therefore we used Lepidocephalichthys berdmorei as outgroup in all analyses. The separate datasets were examined for saturation by plotting the absolute number of transitions and transversions against the uncorrected p-distance for each codon position separately. The heterogeneity of base composition among taxa was checked by v2 test and the congruence of phylogenetic signals of the two datasets was assessed with partition homogeneity test (Farris et al., 1994) with 1000 bootstrap replicates in PAUP 4.0b10 (Swofford, 2002). Prior to the test, we have removed the obviously conflicting taxa. Since the test did not reveal any significant conflict (P = 0.08), both datasets could be combined into a single matrix. The phylogenies were inferred using maximum parsimony (MP), maximum likelihood (ML) and Bayesian analyses (BI). All the analyses were conducted on complete cyt b and RAG 1 matrices separately to show potential disagreement in mitochondrial versus nuclear data due to different mode of their inheritance as well as on the concatenated dataset (excluding conflicting taxa) to combine the strengths of the two genes. Author's personal copy 856 J. Bohlen et al. / Molecular Phylogenetics and Evolution 61 (2011) 854–865 Table 1 Species, individual number, geographic origin and accession numbers of the analysed specimens of Pangio. Species Individual number Geographic origin (Country, province, river system) Coordinates Biogeographic subregion Cyt b accession no. RAG-1 accession no. P. alternans A2648 Indonesia, Kalimantan Tengah, R. RunganKahayan 1°32’S 113°08’E Borneo EU670346 EU672996 P. anguillaris A0117 A1023 Ornamental fish trade Thailand, Ubon Ratchathani, R. Huai Chaera – Mainland EU670347 EU670348 EU672997 A1025 A1060 Thailand, Ubon Ratchathani, R. Huai Chaera Laos, Bolikhamsai, R. Xe Bangfai Mainland Mainland EF508575⁄ – EF056366⁄⁄ EU672998 A1061 A1064 Laos, Bolikhamsai, R. Xe Bangfai Thailand, Chiang Mai, R Mae Taeng Mainland Mainland EF508576⁄ EU670349 EF508664⁄ EU672999 A1065 A2520 Thailand, Chiang Mai, R Mae Taeng Thailand, Surat Thani, R. Tapi Mainland Malay Pen. EU670350 EU670353 EU673000 EU673003 P. cf. apoda I A1672 A1673 Myanmar, no details known Myanmar, no details known Mainland Mainland EU670354 EU670355 EU673004 EU673005 P. cf. apoda II A2264 Myanmar, Kachin, R. Tanai Mainland EU670356 EU673006 P. bitaimac A1683 Indonesia, Jambi, R. Sungei Alai Sumatra EU670351 EU673001 P. cuneovirgata P. doriae P. filinaris P. incognito P. kuhlii/semicincta I P. kuhlii/semicincta II 15°21’N 104°49’E 17°26’N 105°12’E 19°08’N 98°55’E 8°53’N 98°40’E 26°21’N 96°43’E 1°28’N 102°28’E A1684 Indonesia, Jambi, R. Sungei Alai Sumatra EU670352 EU673002 A0550 A0551 A1725 Thailand, Narathivat, River unknown Thailand, Narathivat, River unknown Malaysia, Johor, R. Labis Malay Pen. Malay Pen. Malay Pen. EF508577⁄ EF508578⁄ EU670357 EF508665⁄ EF508666⁄ – 2°25’N 103°01’E A2637 Indonesia, River unknown – EU670358 EU673007 A0115 A0116 A1583 Ornamental fish trade Ornamental fish trade Malaysia, Sarawak, R. Noren – – Borneo EU670359 EU670360 EF508580⁄ – – EF508668⁄ A1584 A1585 A1696 Malaysia, Sarawak, R. Noren Malaysia, Sarawak, R. Noren Indonesia, Jambi, R. Sungei Alai Borneo Borneo Sumatra EU670361 EU670362 EF508579⁄ EU673008 EU673009 EF508667⁄ A2650 Indonesia, Kalimantan Tengah, R. RunganKahayan Borneo EU670363 EU673010 A1438 Thailand, Narathiwat, Plu To Daeng swamp Malay Pen. EU670365 EU673012 A1439 A1440 Thailand, Narathiwat, Plu To Daeng swamp Thailand, Narathiwat, Plu To Daeng swamp Malay Pen. Malay Pen. EU670366 EU670367 EU673013 EU673014 A1592 Malaysia, Sarawak, R. Stuum Muda Borneo EU670368 EU673015 A1593 A1594 A1595 A1596 A1711 Malaysia, Malaysia, Malaysia, Malaysia, Malaysia, Borneo Borneo Borneo Borneo Borneo EU670369 EU670370 EU670371 – EU670372 EU673016 EU673017 EU673018 EU673019 EU673020 Sarawak, Sarawak, Sarawak, Sarawak, Sarawak, R. R. R. R. R. Stuum Muda Stuum Muda Stuum Muda Stuum Muda Kapuas A1712 Malaysia, Sarawak, R. Kapuas A1057 A1553 Thailand, Nakhon Sri Thammarat, R. Ai Khieo Malaysia, Johor, forest creek at Kota Tinggi A1554 A1688 Malaysia, Johor, forest creek at Kota Tinggi Malaysia, Johor, Sungai Kahang A1689 A1700 Malaysia, Johor, Sungai Kahang Indonesia, Jambi, R. Sungei Alai A1722 Malaysia, Johor, R. Labis A1723 A2638 A2639 Malaysia, Johor, R. Labis Indonesia, River unknown Indonesia, River unknown A1611 Malaysia, Sarawak, R. Stuum Muda A1645 Malaysia, Sarawak, R. Engkabang A1646 Malaysia, Sarawak, R. Engkabang 1°22’N 110°05’E 1°28’N 102°28’E 1°32’S 113°08’E 6°08’N 101°58’E 1°28’N 109°58’E 1°22’N 110°00’E Borneo EU670373 EU673021 8°33’N 99°47’E Malay Pen. – EU673022 1°52’N 103°52’E Malay Pen. EF508585⁄ EF056331⁄⁄ Malay Pen. Malay Pen. EF508584⁄ EU670374 EF508671⁄ EU673023 Malay Pen. Sumatra EU670375 EF508588⁄ EU673024 EF508674⁄ Malay Pen. EU670376 EU673025 Malay Pen. – – EU670377 EU670378 EU670379 EU673026 EU673027 EU673028 Borneo EU670380 EU673029 Borneo EF508586⁄ EF508672⁄ Borneo EF508587⁄ EF508673⁄ 2°20’N 103°37’E 1°28’N 102°28’E 2°23’N 102°55’E 1°28’N 109°58’E 1°02’N 110°44’E Author's personal copy 857 J. Bohlen et al. / Molecular Phylogenetics and Evolution 61 (2011) 854–865 Table 1 (continued) Species Individual number Geographic origin (Country, province, river system) Coordinates Biogeographic subregion Cyt b accession no. RAG-1 accession no. A1704 Malaysia, Sarawak, R. Kapuas 1°22’N 110°00’E Borneo EU670381 – P. kuhlii/semicincta III A0134 Ornamental fish trade – EU670382 – P. lumbriciformis A2606 Myanmar, Kachin, R. Tompan Chaung 25°19’N 97°10’E Mainland EU670383 EU673030 P. malayana A1685 Malaysia, Johor, Sungai Kahang 2°20’N 103°37’E Malay Pen. EU670384 EU673031 A1686 A1687 Malaysia, Johor, Sungai Kahang Malaysia, Johor, Sungai Kahang Malay Pen. Malay Pen. EU670385 EU670386 EU673033 EU673032 P cf. oblonga (malayana) A1695 Malaysia, Johor, Sungai Kahang 2°20’N 103°37’E Malay Pen. EU670410 EU673055 P. muraeniformis A1467 Malaysia, Johor, forest creek at Kota Tinggi 1°52’N 103°52’E Malay Pen. EU670387 EU673034 A1468 A1736 A1856 Malaysia, Johor, forest creek at Kota Tinggi Malaysia, Johor, forest creek at Kota Tinggi Malaysia, Johor, R. Endau Malay Pen. Malay Pen. Malay Pen. EU670388 EU670389 EU670390 EU673035 EU673036 EU673037 A2267 Thailand, Chanthaburi, R. Pong Nam Ron Mainland EU670391 – A2268 A2269 A2270 A2271 Thailand, Thailand, Thailand, Thailand, Mainland Mainland Mainland Mainland EU670392 EU670393 EU670394 EU670395 EU673038 EU673039 EU673040 EU673041 P. cf. oblonga I A0007 A0008 A2643 A2646 Ornamental Ornamental Ornamental Ornamental – – – – EU670396 EU670397 EU670398 EU670399 EU673042 EU673043 EU673044 EU673045 P. cf.oblonga II A1778 Thailand, Chanthaburi, R. Pong Nam Ron Mainland EF508582⁄ EF056346⁄⁄ A1779 A2274 Thailand, Chanthaburi, R. Pong Nam Ron Thailand, Chanthaburi, river unknown Mainland Mainland EF508581⁄ EU670400 EF508669⁄ EU673046 A0199 A0970 Ornamental fish trade Thailand, Nakhom Phanom, R. Mekong – Mainland EU670401 EU670402 – EU673047 A0971 A2293 Thailand, Nakhom Phanom, R. Mekong Thailand, Nong Khai, R. Mekong Mainland Mainland EU670403 EU670404 EU673048 EU673049 A2294 A2295 A2603 A2604 Thailand, Nong Khai, R. Mekong Thailand, Nong Khai, R. Mekong Ornamental fish trade Ornamental fish trade Mainland Mainland – – EU670405 EU670406 EU670407 EU670408 EU673050 EU673051 EU673052 EU673053 P. cf. oblonga IV A1701 Brunei, Temburong , R. Temburong 4°43’N 115°08’E Borneo EU670409 EU673054 P. pangia I A2454 A2607 Thailand, Phang Nga, R. Takua Pa Myanmar, Tenasserim, R. Dawei Malay Pen. Mainland EU670411 EU670364 EU673056 EU673011 A2608 Myanmar, Tenasserim, R. Tenasserim 8°51’N 98°20’E 14°07’N 98°14’E 11°33’N 99°05’E Mainland EU670412 EU673057 A2181 India, West Bengal, tributary of R. Tista Indian subcont EF508583⁄ EF508670⁄ A2609 Myanmar, Yangon, R. Pazundaung 26°20’N 89°30’E 17°04’N 96°32’E Mainland EU670413 EU673058 A1074 A1075 A1076 A1499 Thailand, Narathiwat, River unknown Thailand, Narathiwat, River unknown Thailand, Narathiwat, River unknown Malaysia, Johor, forest creek at Kota Tinggi Malay Malay Malay Malay EU670414 EU670415 EU670416 EU670417 EU673059 EU673060 EU673061 EU673062 A1500 Malaysia, Johor, forest creek at Kota Tinggi A1501 Malaysia, Johor, forest creek at Kota Tinggi A1697 Malaysia, Johor, Sungai Kahang P. pulla A2649 P. shelfordii A1586 P. myersi P. cf. oblonga III P. pangia II P. piperata I P. piperata II Chanthaburi, Chanthaburi, Chanthaburi, Chanthaburi, fish fish fish fish R. R. R. R. Pong Pong Pong Pong Nam Nam Nam Nam 2°26’N 103°36’E 12°55’N 102°15’E Ron Ron Ron Ron trade trade trade trade 12°55’N 102°15’E 17°18’N 104°47’E 17°57’N 103°03’E 1°52’N 103°52’E Pen. Pen. Pen. Pen. Malay Pen. EU670418 EU673063 1°52’N 103°52’E 2°20’N 103°37’E Malay Pen. EU670419 EU673064 Malay Pen. EU670420 – Indonesia, Kalimantan Tengah, R. RunganKahayan 1°32’S 113°08’E Borneo EU670421 EU673065 Malaysia, Sarawak, R. Noren 1°22’N 110°50’E Borneo EU670422 EU673066 (continued on next page) Author's personal copy 858 J. Bohlen et al. / Molecular Phylogenetics and Evolution 61 (2011) 854–865 Table 1 (continued) Species Individual number Geographic origin (Country, province, river system) A1587 A1653 Malaysia, Sarawak, R. Noren Malaysia, Sarawak, R. Engkabang A1661 Malaysia, Sarawak, R. Engkabang A1609 Malaysia, Sarawak, R. Stuum Muda A1610 A1620 Malaysia, Sarawak, R. Stuum Muda Malaysia, Sarawak, R. Sebako A1621 A1636 Malaysia, Sarawak, R. Sebako Malaysia, Sarawak, forest creek at Kuching A1637 A1651 Malaysia, Sarawak, forest creek at Kuching Malaysia, Sarawak, R. Engkabang A1652 A1654 A1655 Malaysia, Sarawak, R. Engkabang Malaysia, Sarawak, R. Engkabang Malaysia, Sarawak, R. Engkabang Pangio. sp. B A2647 Indonesia, Kalimantan Tengah, R. RunganKahayan P. superba A2636 Indonesia, River unknown Pangio. sp. A The appropriate models of nucleotide substitution for likelihoodbased analyses were determined using Modeltest 3.06 (Posada and Crandall, 1998) under the Akaike information criterion. ML analyses were performed in GARLI 0.95 with settings of GTR + I + C model without specifying the values; the values were estimated during the analyses. The statistical support of the nodes was assessed by 1000 non-parametric bootstrap resamplings and the resulting trees were used to built a 50% majority rule consensus tree in PAUP. Bayesian inference of phylogeny was performed in MrBayes 3.1 (Huelsenbeck and Ronquist, 2001). The datasets were partitioned according to codon positions and in case of combined dataset further divided into genes, summing up in total of six partitions. Analyses were set to six Markov chains running for 2,000,000 generations with default heating conditions under the GTR + I + C for each partition, starting from randomly generated trees. The trees and likelihood scores were sampled each 100 generations. After inspecting the likelihood scores of the sampled trees for stationary by plotting –lnL against generation time, the first 800 trees were discarded as ‘burnin’ and the remaining trees were used to build 50% majority rule consensus trees and to estimate the posterior probabilities. MP reconstructions were carried out in PAUP via heuristic searches with random stepwise addition of taxa and tree bisection and reconnection (TBR) branch swapping algorithm. The branch support in MP trees was assessed via 1000 non-parametric bootstrap resamplings. Since we have detected saturation in the third codon position of cyt b dataset, we performed weighted parsimony with this matrix and downweightened the transitions four times relative to transversions as corresponding to a priori estimations from the given dataset. In the combined dataset the weighting scheme followed the estimated Ti/Tv ratio of each partition. We used ML and MP bootstrap values and Bayesian posterior probabilities to assess the branch support. During the evaluation of the topology we took into consideration nodes supported by bootstrap values of 70% or greater and posterior probabilities equal or greater than 0.95. Besides, we have tested the monophyly of the formerly proposed species groups sensu Kottelat and Lim (1993) employing likelihood based Shimodaira–Hasegawa test (Shimodaira and Hasegawa, 1999) implemented in PAUP using RELL (resampling estimated log-likelihood) method with 1000 bootstrap resamplings. The constraints corresponding to the concepts of (1) the Coordinates 1°02’N 110°44’E 1°28’N 109°58’E 1°44’N 109°44’E 1°25’N 110°24’E 1°02’N 110°44’E 1°32’S 113°08’E Biogeographic subregion Cyt b accession no. RAG-1 accession no. Borneo Borneo EU670423 EU670424 EU673067 EU673068 Borneo EU670425 EU673069 Borneo EU670426 missing Borneo Borneo EU670427 EU670428 EU673070 EU673071 Borneo Borneo EU670429 EU670430 EU673072 EU673073 Borneo Borneo EU670431 EU670432 EU673074 EU673075 Borneo Borneo Borneo EU670433 EU670434 EU670435 EU673076 EU673077 EU673078 Borneo EU670436 EU673079 – EU670437 EU673080 four species groups sensu Kottelat and Lim (1993) and (2) the likelihood of the monophyly of each of the four suggested groups (kuhlii-, anguillaris-, pangio-, oblonga-group) separately were constructed in MacClade 4.0 (Maddison and Maddison, 2000). The differences in rate heterogeneity across the lineages were assessed for the cyt b dataset using the likelihood ration test (LRT) by comparing the likelihood scores of ML trees calculated with and without enforcing molecular clock in PAUP. 3. Results The aligned RAG-1 dataset contained 100 sequences of 913 basepairs (bp) length, with 27.3% of the positions variable and 20.7% of positions parsimony informative. The final alignment of the cyt b dataset included 110 sequences of 1118 bp length with 43.0% variable and 39.3% parsimony informative positions. No compositional bias was observed across the taxa. The cyt b dataset expressed saturation in the third codon position. Our phylogenetic analyses showed three major lineages within the samples of eel-loaches (ingroup) in the RAG-1 dataset as well as in the cyt b dataset and the combined dataset (Figs. 1–3). In all three analyses, the first lineage comprises all specimens of Pangio incognito, Pangio muraeniformis, Pangio piperata, Pangio shelfordii, Pangio sp. A, Pangio sp. B and Pangio superba and forms the sister-clade to the remaining eel-loaches. These remaining samples are recovered as two sister lineages: the first comprises Pangio anguillaris, Pangio bitaimac, Pangio doriae and Pangio lumbriciformis and the second all remaining samples. The monophyly of these three major lineages was supported with high statistical support in all phylogenetic analyses. The results of the SH test of the concept with four major lineages as suggested by the species-group concept resulted in a significant decrease of likelihood (D-lnL = 33.876, p SH = 0.008, p < 0.05). A detailed test of the monophyly conducted on each proposed species group separately supported the assumption of monophyly for the shelfordii-, anguillaris and kuhlii-group (all–lnL were identical to those of the unconstrained tree) except for the oblonga-group (D-lnL = 33.876, p SH = 0.008, p < 0.05). This observation indicates that the hypothesis of four major monophyletic lineages is contradicted mainly by the oblonga-group, which turned out to be paraphyletic in all phylogenetic analyses. Author's personal copy J. Bohlen et al. / Molecular Phylogenetics and Evolution 61 (2011) 854–865 859 Fig. 1. Phylogenetic relationships of eel loaches of the genus Pangio resulting from the Bayesian analysis of RAG-1 sequences. The species groups as defined by Kottelat and Lim (1993) are marked by different patterns; banded blocks mark members of the kuhlii-group, dark blocks members of the oblonga-group, light blocks members of the anguillaris-group and checked blocks members of the shelfordii-group. The likelihood ratio test of the cyt b dataset did not detect significant differences between the trees estimated with and without enforcing molecular clock (v2 = 82.076, df = 108, p = 0.9701), sug- gesting that within the studied group sequences evolved in a clock-like manner. The estimated cyt b ML tree has been converted into ultrametric tree by non-parametric rate smoothing (NPRS) Author's personal copy 860 J. Bohlen et al. / Molecular Phylogenetics and Evolution 61 (2011) 854–865 Fig. 2. Phylogenetic relationships of eel loaches of the genus Pangio resulting from the Bayesian analysis of cytochrome b sequences. method in TreeEdit 1.0 (Rambaut and Charleston, 2001. Rates of molecular diversification have been estimated several times for cobitid fishes and critically discussed by Doadrio and Perdices (2005). These authors suggested a mutation rate of 0.68% per mil- lion years for cyt b as best fitting for cobitid fishes. Applying this calibration, the divergence time between lineage I and the rest of the ingroup clade was estimated to 33 MYA and between lineage II and lineage III to 30 MYA (Fig. 4). Author's personal copy J. Bohlen et al. / Molecular Phylogenetics and Evolution 61 (2011) 854–865 861 Fig. 3. Phylogenetic relationships of eel loaches of the genus Pangio resulting from the Bayesian analysis of the combined dataset and a comparison with the species groups defined by Kottelat and Lim (1993) and the species groups as revealed from the present data. 4. Discussion 4.1. Major lineages and ‘species-groups’ among eel-loaches The presented analyses revealed the existence of several lineages within the genus Pangio. We recognise here three major lineages: Lineage I consist of Pangio incognito, Pangio muraeniformis, Pangio piperata, Pangio shelfordii, Pangio sp. A, Pangio sp. B and Pangio superba, while lineage II comprises Pangio anguillaris, Pangio bitaimac, Pangio doriae and Pangio lumbriciformis. All remaining species are grouped in lineage III. However, within lineage III, three sublineages can be identified: the first (III a) is represented by the specimens of Pangio filinaris, the second (III b) by the only sample of Pangio cf. oblonga IV and the third (III c) by all remaining sam- ples. This pattern was consistently recovered in all analyses of the single as well as the combined dataset. The species-groups proposed by Kottelat and Lim (1993) were not necessarily expected to represent monophyletic lineages, but they partly match the major lineages as revealed from the molecular genetic data. Two of their species-groups, the shelfordii- and the anguillaris-group, were identified in our analyses as monophyletic (lineage I and lineage II, respectively). Although the statistical support of the kuhlii-group was rather low in some analyses, its monophyly was not rejected. However, the oblongagroup sensu Kottelat and Lim (1993) is paraphyletic and, together with the kuhlii-group, forms lineage III in our analyses. The results of the SH-tests confirm this observation: when the hypothesis of monophyly for each of the species-group sensu Author's personal copy 862 J. Bohlen et al. / Molecular Phylogenetics and Evolution 61 (2011) 854–865 Fig. 4. Ultrametric calibrated tree of eel loaches allowing to date the cladogenetic events during the evolution of the genus. Author's personal copy J. Bohlen et al. / Molecular Phylogenetics and Evolution 61 (2011) 854–865 Kottelat and Lim (1993) is tested, the tests cannot reject monophyly of the shelfordii-, kuhlii- and the anguillaris-group, but reject monophyly of the oblonga-group. Non-monophyly of the oblonga-group is mainly due to the position of the species Pangio filinaris and Pangio cf oblonga IV, which represent two monotypic sublineages (IIIa and IIIb, respectively) basal to the remaining taxa of lineage III. The monophyly of the banded species of Pangio is statistically supported only in the RAG-1 dataset, while in the analyses of cyt b the species Pangio cuneovirgata and Pangio pulla together form a distinct group. This result may be caused by different mutation rates of the two genes; particularly the saturation in the third codon position of cyt b dataset may be responsible for the low statistical support of the kuhlii-group. Species of the kuhlii- and oblonga-groups share a similar shape of the body and similar vertebral numbers (Kottelat and Lim, 1993). The only significant difference is in the general pigmentation pattern: species of the kuhlii-group are banded while species of the oblonga-group are plain brown. The generally high degree of morphological similarity between brown and banded species is illustrated by the case of A1695. This plain brown specimen was identified as Pangio cf. oblonga based on its colour pattern, but in all genetic analysis it grouped together with the banded specimens of Pangio malayana from the same locality. These data suggest that A1695 is a specimen of Pangio malayana that did not express the banding pattern but the plain brown colour of the oblonga-group. Nearly completely brown specimens are also regularly found among specimens of Pangio kuhlii/semicincta that are imported for ornamental fish trade (Bohlen, pers. observ.). In contrast, some species from the oblonga-group are known to have black bands on the head (one on the snout and one through the eye). The phylogenetic position of the banded species within the brown species of the oblonga-group makes a clear, unequivocal separation between the kuhlii-group and the oblonga-group impossible. Consequently, we propose to recognise only three major lineages within the genus Pangio: anguillaris-group, shelfordii-group and kuhlii-oblonga group (for all species formerly belonging to the kuhlii-group and the oblonga-group). Included in the present study were specimens of Pangio lumbriciformis and Pangio pulla, two species with unusual character combinations that were not or only provisionally placed into one of the four species-groups. Kottelat and Lim (1993) stressed the contradictory morphological characters of Pangio pulla, assigned it to the anguillaris-group because of its relatively high vertebral count, but mentioned its similarity to the kuhlii-group in colour pattern. Moreover, according to Kottelat and Lim, 1993, Pangio pulla lacks pelvic fins, a character otherwise only observed in the brown species Pangio apoda and Pangio fusca (Britz and Maclaine, 2007) and in Pangio lidi (Hadiaty and Kottelat, 2009) of the anguillaris-group. Our data show a clear affiliation of Pangio pulla to the kuhlii-oblonga group, thus demonstrating that meristic characters like a high vertebral count or the absence of pelvic fins have developed at least twice independently within Pangio. Britz and Maclaine (2007) stated that Pangio lumbriciformis could not be assigned to any of the species groups, because it shares characters with the anguillaris-group and other characters with the shelfordii-group. Our data identify Pangio lumbriciformis as a member of the anguillaris-group, closely related to Pangio anguillaris, Pangio bitaimac and Pangio doriae. 4.2. Species diversity among eel-loaches and horizontal gene flow With the morphological characters used previously in taxonomic studies of the genus Pangio, the fishes in the present study could be assigned to 19 nominal species. However, our analyses revealed the existence of several additional units, indicating much higher species-level diversity among eel loaches than is presently recognised. The species 863 Pangio anguillaris, Pangio cf. apoda, Pangio pangia and Pangio piperata split into two lineages each, Pangio kuhlii/semicincta and Pangio cf. oblonga in three and four lineages, respectively. Since no material from the type localities of the last two species was available for comparison, we are unable to determine which of the lineages (if any) represent the nominal species and which lineages indicate possibly new taxa. In the case of Pangio piperata, the two lineages co-occur at the same locality and only a careful comparison with the type material may reveal which of them corresponds to the true Pangio piperata. In most cases, the new units were identified in the analysis of both single datasets as well as in the combined dataset with high statistical support, making it unlikely that they represent artefacts of the analysis. There are also cases, in which the phylogenetic position of certain specimens is inconsistent between the two datasets; these cases may indicate hybridisation events or might be a result of incomplete lineage sorting. One example regards the specimens A1688 and A1689, which were morphologically identified as Pangio semicincta and which both join the other specimens of Pangio semicincta in the cyt b dataset. In the RAG-1 dataset, however, only A1689 is found among the other samples of Pangio semicincta, but A1688 as a sister lineage to the specimens of Pangio malayana. Interestingly, these specimens of Pangio malayana were collected at the same locality as A1688 and A1689; therefore it seems most likely that a transfer of genetic information, that means a past hybridisation with subsequent backcross with the paternal (in this case maternal) species has occurred. To our knowledge, this is the first indication of a hybridisation between species of Pangio. In general, horizontal gene flow has been demonstrated in many fish taxa and its presence in co-occurring species of Pangio would not be surprising. However, the frequency of the event should be tested on more material, and its general possibility should be kept in mind in further studies on Pangio diversity. Overall, our results show that the species diversity of eel loaches in fact is much higher than previously suggested and may provide help for further detailed taxonomic studies. 4.3. Taxonomic implications The species Pangio muraeniformis has often been considered a synonym of Pangio shelfordii (Kottelat and Lim, 1993), and has only recently been recognised as a distinct species (Kottelat and Whitten, 1996); the results of our genetic analyses support the conclusion that the specimens from the Malay Peninsula represent a distinct species. At least six lineages as recovered in the present study cannot be matched with any described species. However, a morphological comparison of their voucher specimens with comparative and type material from the collection of BMNH (Bohlen and Britz, unpubl. data) showed that these fishes differ also morphologically from each other as well as from the described species and have to be considered undescribed species. In the case of two lineages (Pangio piperata and Pangio pangia), no morphological differences were observed, therefore these lineages are treated here as sublineages of the same species (Pangio piperata I and II and Pangio pangia I and II, respectively). The species referred to here as Pangio sp. V was formerly treated as Pangio shelfordii (Kottelat and Lim, 1993), but our investigations revealed genetic as well as morphologic differences (Bohlen and Britz, unpubl. data). Since the type locality of Pangio shelfordii is located in the lower basin of the Sarawak River in western Borneo, we consider our samples from the upper Sarawak basin as representing Pangio shelfordii. The analysed specimens of Pangio anguillaris form two distinct groups, which were recovered in a trichotomy with Pangio bitaimac. Four species of the anguillaris-group have been described in recent years (Britz and Maclaine, 2007; Hadiaty and Kottelat, 2009; Tan and Kottelat, 2009) and it is likely that more unidentified species are included in this complex. Since the type locality of Pangio anguillaris is in Central Borneo, but the material analysed Author's personal copy 864 J. Bohlen et al. / Molecular Phylogenetics and Evolution 61 (2011) 854–865 in the present study originated from Thailand and Laos, it is even possible that none of our lineages actually represents this species. 4.4. Dating of the cladogenetic events The ultrametric calibrated tree showed that the basal cladogenetic events that split the genus Pangio into the lineages I, II and III (nodes 1 and 2 in Fig. 4) have taken place about 29–33 MYA during early Oligocene. After a period with no further cladogenic events, the further radiation within all three lineages started more or less simultaneously 19–24 MYA during early Miocene (nodes 3–6), obviously an important time in the evolution of eel loaches. From this time onwards, the calibrated tree does not indicate any periods of increased or decreased radiation, but within all three lineages nodes appear non-synchronised. This observation can best be explained by a dominance of local factors in the evolution of eel loaches rather than largescale events triggering the evolution of these freshwater fishes. 4.5. Paleaogeological events in Sundaland reflected in the phylogeny of eel loaches It is generally accepted that palaeogeologic events are one of the major forces that have shaped the evolution of life on earth, on a local as well as on a global scale. Geologic events that lead to a separation of a formerly united habitat may lead to an interruption of the gene flow between the animals inhabiting this habitat (vicariance event), while events that join formerly separated habitats may lead to contact between isolated populations and result in hybridisation or competitive expulsion (Cox and Moore, 2010; Parenti and Ebach, 2009). One of the areas with a high number of isolating and connecting palaeogeologic events is Sundaland, which comprises the Malay Peninsula, the Sunda Islands, like Borneo, Sumatra and Java, and the shallow continental shelf between them. During periods of lowered sea level, this shelf was exposed and formed large lowland areas connecting the islands with each other and with the Malay Peninsula (Woodruff, 2003). Many rivers that nowadays drain into the Gulf of Thailand or the southern parts of the South China Sea were then united and enabled faunal exchanges of freshwater fauna between presently isolated parts of SE Asia (Voris, 2000). Although such connections were likely present repeatedly during different periods, only very few events of faunal exchanges have been demonstrated on the basis of reliable phylogenies (Dodson et al., 1995). Taking the shelfordii-group as example, we analysed the ultrametric calibrated tree for a correlation of major biogeographic events with major cladogenetic events. The analysis bases on the cytochrome b dataset because of the advantage to date events; an analysis of the RAG-1 dataset might have brought different results. The ultrametric calibrated tree shows that the basal split that separates the species Pangio incognito, Pangio piperata II, Pangio shelfordii and Pangio sp. B from the species Pangio muraeniformis, Pangio piperata I and Pangio sp. A (Fig. 4) took place during early Miocene about 19–20 MYA. Since the first group is distributed mainly on Borneo and the second mainly on the Malay Peninsula, we conclude that the observed phylogenetic pattern is the result of a vicariance event separating the distribution area of the common ancestor into a Bornean and a Peninsular subarea. A similar split was already hypothesised by Kottelat and Lim (1993) on the basis of morphological data; and our phylogenetic data provide further support for this hypothesis. The early Miocene was a period of increased global sea levels, which resulted in the flooding of large lowland areas. This flooding lasted about 11 MY (24–13 MYA) and has been identified as an important vicariance factor that separated biogeographic regions in SE Asia (Woodruff, 2003). It also efficiently separated Borneo from the Malay Penin- sula; therefore we consider this separation to be responsible for the split into the two main sublineages of the shelfordii-group. Although the two shelfordii-group sublineages in general have disjunct distribution areas, they include two exceptions to this rule: First, the species Pangio sp. A forms a monophyletic lineage with the species from the Malay Peninsula; but itself occurs on Borneo. According to our calculations, the separation between Pangio sp. A and the species from the Malay Peninsula happened around 15 MYA. During the period between 13 and 15 MYA, the sea level was repeatedly lowered and has provided the opportunity for eel-loaches to cross the land bridge. Second, the species Pangio piperata II is known only from the Malay Peninsula, but shows sister-relation to a group of Bornean species including Pangio incognito, Pangio shelfordii and Pangio sp. B. This separation between the Penninsular species and the Bornean species can be dated back to around 8 MYA, another period with low sea level during which Sundaland fell dry and rivers from western Borneo and the eastern Malay Peninsula joined and drained together to the South China Sea. In both these cases, the distribution area of the transferred population is small and located in areas close to the shortest distance between Borneo and the Malay Peninsula in the westernmost tip of Sarawak state on Borneo and in the eastern part of Johor state on the Malay Peninsula. All reconstructions of the river systems on Sundaland during the periods of low sea levels suggest that the rivers of western Sarawak and eastern Johor drained into the same major river that flew north of Borneo into the South China Sea (Inger and Voris, 2001; Bird et al., 2005). We consider two faunal exchange events the most likely explanation for the disjunct distribution of both sublineages of the shelfordii-group of eel loaches. The alternative explanation of an incomplete lineage sorting between the two sublineages during the Miocene separation period cannot explain the limited distribution of the two rare lineages on Borneo and the Malay Peninsula, respectively. It would need to assume that these areas have been isolated from the rest of Borneo and the Malay Peninsula, respectively, for at least 19 million years. Acknowledgments We would like to express our thanks to D. Bohlen, M. Kottelat, M. Kroupa, M. Lo, L. Rüber, I. Seidel, K. Udomritthiruj for their help in obtaining the samples, and to J. Maclaine and O. Crimmen for help with the type material in BMNH. The study was supported by Grant Nos. 206/05/2556 and 2006/08/0637 of the Czech Science Foundation, No. A600450508 of the Grant Agency of the Academy of Sciences of the Czech Republic, by the IRP IAPG No. AV0Z50450515, the Biodiversity Research Centre LC 06073 and the European Council founded programme ‘SYNTHESYS’. The Raffles Museum of Biodiversity Research and research grants R-154000-318-112, R-264-001-004-272 from the National University of Singapore supported THH’s field collections. References Bănărescu, P., 1990. Zoogeography of Fresh Waters, vol. 1. AULA-Verlag, Wiesbaden. Bird, M.I., Hunt, C., Taylor, D., 2005. Palaeoenvironments of insular Southeast Asia during the Last Glacial Period: a savanna corridor in Sundaland? Quat. Sci. Rev. 24, 2228–2242. Bohlen, J., Šlechtová Jr., V., Doadrio, I., Ráb, P., 2007. Low mitochondrial divergence indicates a rapid expansion across Europe in the weather loach, Misgurnus fossilis (L.). J. Fish Biol. 71 (Suppl.), 186–194. Bohlen, J., Perdices, A., Doadrio, I., Economides, P., 2006a. Vicariance, colonisation and fast local speciation in Asia Minor and the Balkan as revealed from the phylogeny of spined loaches (Osteichthyes; Cobitidae). Mol. Phylogenet. Evol. 39, 552–561. Bohlen, J., Šlechtová, V., Bogutskaya, N., Freyhof, J., 2006b. Across Siberia and over Europe: phylogenetic relationships of the freshwater fish genus Rhodeus in Europe and the phylogenetic position of R.sericeus from the River Amur. Mol. Phylogenet. Evol. 40, 856–865. Author's personal copy J. Bohlen et al. / Molecular Phylogenetics and Evolution 61 (2011) 854–865 Britz, R., Kottelat, M., 2010. Pangio longimanus, a miniature species of eel-loach from central Laos (Teleostei: Cypriniformes: Cobitidae). Ichthyological Exploration of Freshwaters 20, 317–376. Britz, R., Maclaine, J., 2007. A review of the eel-loaches, genus Pangio, from Myanmar (Teleostei: Cypriniformes: Cobitidae). Ichthyol. Explor. Freshwaters 18, 17–30. Brookfield, M.E., 1998. The evolution of the great river systems of southern Asia during the Cenozoic India-Asia collision: rivers draining southwards. Geomorphology 22, 285–312. Burridge, M., 1992. Systematics of the Acanthophthalmus kuhlii complex (Teleostei: Cobitidae), with the description of a new species from Sarawak and Brunei. Copeia 1992, 172–186. Clark, M.K., Schoenbohm, L.M., Royden, L.H., Whipple, K.X., Burchfiel, B.C., Zhang, X., Tang, W., Wang, E., Chen, L., 2004. Surface uplift, tectonics, and erosion of eastern Tibet from large scale drainage patterns. Tectonics 23, TC106. doi:10.1029/2002TC001402. Cox, C.B., Moore, P.D., 2010. Biogeography: An Ecological and Evolutionary Approach, 8th ed. John Wileys & Sons, Hoboken. Dodson, J.J., Colombani, F., Ng, P.K.L., 1995. Phylogeographic structure in mitochondrial DNA of south-east Asian freshwater fish, Hemibagrus nemurus (Siluroidae; Bagridae) and pleistocene sea-level changes on the Sunda shelf. Mol. Evol. 4, 331–346. Doadrio, I., Perdices, A., 2005. Phylogenetic relationships among the Ibero-African cobitids (Cobitis, Cobitidae) based on cytochrome b sequence data. Mol. Phylogenet. Evol. 37, 484–493. Durand, J.C., Templeton, A.R., Guinand, B., Imsiridou, A., Bouvet, Y., 1999. Nested clade and phylogeographic analyses of the chub, Leuciscus cephalus (Teleostei, Cyprinidae), in Greece. Implications for Balkan Peninsula Biogeography. Mol. Phylogenet. Evol. 13, 566–580. Eschmeyer, W., 2010. The catalog of fishes on-line. <http://www.calacademy.org>. Farris, J.S., Källerjö, M., Kluge, A.G., Butt, C., 1994. Testing significance of incongruence. Cladistics 10, 315–319. Hadiaty, R.K., Kottelat, M., 2009. Pangio lidi, a new species of loach from eastern Borneo, Indonesia. Zootaxa 2171, 65–68. Hall, T.A., 1999. BioEdit: a user-friendly biological sequence alignment editor and analysis program for Windows 95/98/NT. Nucl. Acids. Symp. Ser. 41, 95–98. Huelsenbeck, J.P., Ronquist, F.R., 2001. MRBAYES: Bayesian inference of phylogeny. Bioinformatics 17, 754–755. Inger, R.F., Voris, H.K., 2001. The biogeographic relations of the frogs and snakes of Sundaland. J. Biogeograph. 28, 863–891. Kottelat, M., Lim, K.K.P., 1993. A review of the eel-loaches of the genus Pangio (Teleostei: Cobitidae) from the Malay Peninsula, with description of six new species. Raffles Bull. Zool. 41, 203–249. Kottelat, M., Widjanarti, E., 2005. The fishes of Danau Sentarum National Park and the Kapuas lakes area, Kalimantan Barat, Indonesia. Raffles Bull. Zool. Suppl. 13, 139–173. Kottelat, M., Whitten, A.J., 1996. Freshwater Fishes of Western Indonesia and Sulawesi: Additions and Corrections, Periplus, Jakarta. Li, C., Ortí, G., 2007. Molecular phylogeny of Clupeiformes (Actinopterygii) inferred from nuclear and mitochondrial DNA sequences. Mol. Phylogenet. Evol. 44, 386–398. Lopéz, A., Chen, W.-J., Ortí, G., 2004. Esociform phylogeny. Copeia 3, 449–464. Ludwig, A., Bohlen, J., Wolter, C., Pitra, C., 2001. Phylogenetic relationships and historical biogeography of spined loaches (Cobitidae, Cobitis and Sabanejewia) as indicated by variability of mitochondrial DNA. Zool. J. Linn. Soc. 131, 381–392. Maddison, D.R., Maddison, W.P., 2000. MacClade 4: Analysis of Phylogeny and Character Evolution. Version 4.0. Sinauer Associates, Sunderland, Massachusetts. Menon, A.G.K., 1992. The Fauna of India and the Adjacent Countries. Pisces. Vol. IV. Teleostei - Cobitoidea. Part 2. Cobitidae. Zoological Survey of India, Madras. 865 Nelson, J.S., 2006. Fishes of the world, 4th ed. John Wiley and Sons, Hoboken. Ng, H.H., Tan, H.H., 1999. The fishes of the Endau Drainage, Peninsular Malaysia with descriptions of two new species of catfishes (Teleostei: Akysidae, Bagridae). Zool. Stud. 38, 350–366. Parenti, L.R., Ebach, M.C., 2009. Comparative Biogeography: Discovering and Classifying Biogeographical Patterns of a Dynamic Earth. University of California Press, Los Angeles. Pérez, G.A.C., Říčan, O., Ortí, G., Bermingham, E., Doadrio, I., Zardoya, R., 2007. Phylogeny and biogeography of 91 species of heroine cichlids (Teleostei: Cichlidae) based on sequences of the cytochrome b gene. Mol. Phylogenet. Evol. 43, 91–110. Perdices, A., Doadrio, I., 2001. The molecular systematics and biogeography of the European Cobitids based on mitochondrial DNA sequences. Mol. Phylogenet. Evol. 19, 468–478. Perdices, A., Doadrio, I., Bermingham, E., 2005. Evolutionary history of the synbranchid eels (Teleostei: Synbranchidae) in Central America and the Caribbean islands inferred from their molecular phylogeny. Mol. Phylogenet. Evol. 37, 460–473. Perdices, A., Doadrio, I., Economidis, P.S., Bohlen, J., Bãnãrescu, P., 2003. Pleistocene effects on the European freshwater fish fauna: double origin of the cobitid genus Sabanejewia in the Danube basin (Osteichthyes: Cobitidae). Mol. Phylogenet. Evol. 26, 289–299. Perdices, A., Bohlen, J., Doadrio, I., 2008. The molecular diversity of adriatic spined loaches (Teleostei, Cobitidae). Mol. Phylogenet. Evol. 46, 382–390. Posada, D., Crandall, K.A., 1998. Modeltest: testing the model of DNA substitutions. Bioinformatics 14, 817–818. Rambaut, A., Charleston, M., 2001. TreeEdit Ver 1.0. University of Oxford, Oxford. Rüber, L., Britz, R., Kullander, S.O., Zardoya, R., 2004. Evolutionary and biogeographic patterns of the Badidae (Teleostei: Perciformes) inferred from mitochondrial and nuclear DNA sequence data. Mol. Phylogenet. Evol. 32, 1010–1022. Shimodaira, H., Hasegawa, M., 1999. Multiple comparisons of Log-likelihoods with applications to phylogenetic inference. Mol. Biol. Evol. 16, 1114–1116. Sullivan, J.P., Lundberg, J.G., Hardman, M., 2006. A phylogenetic analysis of the major groups of catfishes (Teleostei: Siluriformes) using rag1 and rag2 nuclear gene sequences. Mol. Phylogenet. Evol. 41, 636–662. Šedivá, A., Janko, K., Šlechtová Jr., V., Kotlík, P., 2008. Around and across Carpathians: colonization model of the Danube basin inferred from genetic diversification of stone loach (Barbatula barbatula) populations. Mol. Ecol. 17, 1277–1292. Šlechtová, V., Bohlen, J., Freyhof, J., Ráb, P., 2006. Molecular phylogeny of the Southeast Asian freshwater fish family Botiidae (Teleostei: Cobitoidea) and the origin of polyploidy in their evolution. Mol. Phylogenet. Evol. 39, 529–541. Šlechtová Jr., V., Bohlen, J., Tan, H.H., 2007. Families of Cobitoidea (Teleostei; Cypriniformes) as revealed from nuclear genetic data and the position of the mysterious genera Barbucca, Psilorhynchus, Serpenticobitis and Vaillantella. Mol. Phylogenet. Evol. 44, 1358–1365. Šlechtová Jr., V., Bohlen, J., Perdices, A., 2008. Molecular phylogeny of the freshwater fish family Cobitidae (Cypriniformes: Teleostei): delimitation of genera, mitochondrial introgression and evolution of sexual dimorphism. Mol. Phylogenet. Evol. (Online release January 2008). Swofford, D.L., 2002. PAUP⁄: Phylogenetic analysis using parsimony (⁄ and other methods). v. 4.0. Sinauer Associates, Sunderland. Tan, H.H., Kottelat, M., 2009. The fishes of the Batang Hari drainage, Sumatra, with description of six new species. Ichthyol. Explor. Freshwaters 20, 13–69. Voris, H.K., 2000. Maps of Pleistocene sea levels in Southeast Asia: shorelines, river systems and time durations. J. Biogeogr. 27, 1153–1167. Woodruff, D.S., 2003. Neogene marine transgressions, paleogeography and biogeographic transitions on the Thai Malay Penisula. J. Biogeogr. 30, 551–567.