Journal of Mammalogy, 95(3):455–466, 2014

Molecular phylogeny of East and Southeast Asian fossorial moles (Lipotyphla, )

AKIO SHINOHARA,* SHIN-ICHIRO KAWADA,NGUYEN TRUONG SON,CHIHIRO KOSHIMOTO,HIDEKI ENDO,DANG NGOC CAN, AND HITOSHI SUZUKI Division of Bio-resources, Department of Biotechnology, Frontier Science Research Center, University of Miyazaki, Downloaded from https://academic.oup.com/jmammal/article-abstract/95/3/455/876160 by guest on 13 September 2019 Kihara 5200, Kiyotake, Miyazaki 889-1692, Japan (AS, CK) Department of Zoology, National Museum of Nature and Science, 4-1-1, Amakubo, Tsukuba, Ibaraki 305-0005, Japan (SK) Department of Vertebrate Zoology, Institute of Ecology and Biological Resources (IEBR), Vietnam Academy of Science and Technology, 18 Hoang Quoc Viet Road, Hanoi, Vietnam (NTS, DNC) The University Museum, The University of Tokyo, Hongo 7-3-1, Bunkyo-ku, Tokyo 113-0033, Japan (HE) Graduate School of Environmental Science, Hokkaido University, Kita-ku, Sapporo 060-0810, Japan (HS) * Correspondent: [email protected]

The diversity of fossorial moles in East and Southeast Asia is contained in the 2 species-rich genera (8 species) and (8 or more species), and the 3 monospecific genera Scapanulus, Scaptochirus, and Parascaptor. To better understand the evolution and biogeography of these fossorial moles, we conducted molecular phylogenetic analyses using mitochondrial cytochrome-b (Cytb; 1,140 base pairs [bp]) and 12S rRNA (approximately 830 bp) and nuclear recombination activating gene 1 (Rag1; 1,010 bp) gene sequences from 5 species of Euroscaptor,6ofMogera, and the single species of Scaptochirus. Phylogenetic estimates revealed 5 distinct lineages of East and Southeast Asian fossorial moles: Mogera, Scaptochirus, Euroscaptor mizura, E. parvidens, and E. malayana–E. klossi–E. longirostris. Our results support the monophyly of Mogera but not Euroscaptor, indicating a need for taxonomic revision of the latter genus. We hypothesize that Mogera originated in the central portion of its range and then dispersed to peripheral islands, such as Taiwan and the Japanese Islands. The fragmented distribution of Southeast Asian Euroscaptor presumably arose from habitat competition (invasion) from Mogera species, long-range dispersal, vicariance events, or a combination of these, explaining the high species richness of fossorial moles in this region.

Key words: Euroscaptor, , Lipotyphla, Mogera, molecular phylogeny, phylogeography, Scaptochirus, , Talpidae

Ó 2014 American Society of Mammalogists DOI: 10.1644/13-MAMM-A-135

The diversity of fossorial moles in East and Southeast Asia (Allen 1938; Ziegler 1971; Motokawa 2004; Kawada and is contained in the 2 species-rich genera Mogera (8 species) Yokohata 2009; see also Sa´nchez-Villagra et al. 2006). and Euroscaptor (8 or more species), and the 3 monospecific The genus Mogera is widely distributed in East Asia (eastern genera Scapanulus, Scaptochirus, and Parascaptor (Hutterer China, Taiwan, Korea, Japan, and Primorye) and Southeast 2005; Kawada et al. 2008, 2012). Four of these genera Asia (southeastern China to northern Vietnam). Of the 8 (Mogera, Euroscaptor, Scaptochirus, and Parascaptor) belong currently recognized species, the large Japanese (M. wogura) and La Touche’s mole (M. latouchei) have wide to the tribe (Hutterer 2005), and have been predom- distributions in East Asia and Southeast Asia, respectively inantly classified by their dental formula: 44 teeth in (Allen 1938). The remaining 6 species are each confined to Euroscaptor (i 3/3, c 1/1, p 4/4, m 3/3, total 44), 42 in rather small geographic areas. The insular mole (M. insularis) Mogera (i 3/2, c 1/1, p 4/4, m 3/3, total 42), 42 in Parascaptor and Kano’s mole (M. kanoana) are endemic to Taiwan. The (i 3/3, c 1/1, p 3/4, m 3/3, total 42), and 40 in Scaptochirus (i 3/ 3, c 1/1, p 3/3, m 3/3, total 40). Because of different combinations of tooth loss, it has been suggested that the dental formula observed in Euroscaptor is the ancestral condition www.mammalogy.org 455 456 JOURNAL OF MAMMALOGY Vol. 95, No. 3 remaining 4 species are endemic to Japan. The lesser Japanese surveys of Southeast Asian fossorial mole populations are mole (M. imaizumii) occurs in eastern Japan (Abe 1974, 2005; needed to elucidate the evolutionary history of fossorial taxa in Ohdachi et al. 2009). The Sado mole (M. tokudae) and the Asia. Etigo mole (M. etigo) are confined to Sado Island and the In our previous studies, we examined the phylogenetic Echigo Plain, respectively. The Senkaku mole (M. uchidai)is relationships of 21 talpid species, including 8 Asian fossorial restricted to Uotsuri Island in the Senkaku Islands (Abe 2005; mole species, and showed that the Asian fossorial moles Ohdachi et al. 2009). formed a monophyletic group to the exclusion of the European Phylogenetic analyses of the genus Mogera are available for species group (Shinohara et al. 2003, 2004a, 2004b, the East Asian taxa. In particular, the Japanese species, except 2008; Kawada et al. 2007; see also Cabria et al. 2006; for M. uchidai, have been subjected to analyses using Colangelo et al. 2010; Crumpton and Thompson 2013). The Downloaded from https://academic.oup.com/jmammal/article-abstract/95/3/455/876160 by guest on 13 September 2019 mitochondrial DNA (mtDNA) sequences (Okamoto 1999; phylogenetic relationships both among and within the genera Tsuchiya et al. 2000) and nuclear gene sequences (Shinohara et examined, however, have not yet been completely resolved, al. 2004a, 2005; Kirihara et al. 2013). In addition, phylogeo- partly due to ambiguous relationships (Shinohara et al. 2008). graphic analyses of M. imaizumii (Iwasa et al. 2006), M. To obtain finer resolution of Asian talpine phylogeny, we insularis, M. kanoana (Kawada et al. 2007), and M. wogura conducted molecular phylogenetic analyses focusing on the from the Korean Peninsula and adjacent regions (M. w. Southeast Asian fossorial moles. coreana and M. w. robusta—Koh et al. 2012) also have been documented using mtDNA sequences. These recent molecular phylogenetic and phylogeographic studies suggested that the MATERIALS AND METHODS species richness of the modern Mogera lineage is the result of a DNA extraction and sequencing.—The species examined in combination of multiple migration events from mainland Asia this study are listed in Table 1 and mapped in Fig. 1. Details of to the peripheral islands and vicariance events (Tsuchiya et al. collecting localities and species identification of the 2000; Shinohara et al. 2004a, 2005; Iwasa et al. 2006; Kawada Vietnamese and Thai specimens were given in Kawada et al. et al. 2007; Kirihara et al. 2013); however, none of these (2009) and Kawada et al. (2006), respectively. The specimens studies included the Southeast Asian member of the genus, M. of Taiwanese moles used in this study were the same as in latouchei. Kawada et al. (2007). La Touche’s mole (M. latouchei)is The genus Euroscaptor is distributed mainly in Southeast sometimes considered a synonym of the insular mole (M. Asia, with at least 8 species described. One species, the insularis), but recent studies (Kawada et al. 2007, 2009) based Japanese mountain mole (E. mizura), occurs on the Japanese on morphological and karyological characteristics have Islands, whereas all others occur in mainland Asia and indicated that M. latouchei is a distinct species. In this study Southeast Asia (Hutterer 2005). The 7 Southeast Asian species we follow Kawada et al. (2007, 2009). The nomenclature and are the Kloss’s mole (E. klossi) from Thailand; the greater classification follows Ohdachi et al. (2009) for Japanese moles Chinese mole (E. grandis) from Mt. Emei, China; the long- and Hutterer (2005) for other species. We used the species nosed mole (E. longirostris) from southern China to northern name E. malayana for the Malaysian mole; the history of the Vietnam; the small-toothed mole (E. parvidens) from central species name for this mole was reported in Kawada et al. Vietnam; the Malaysian mole (E. malayana) from peninsular (2003, 2008). Malaysia; the Himalayan mole (E. micrura) from the Genomic DNA from preserved liver, spleen, or muscle, or a Himalayas; and a recently described species from northern combination of these, was extracted by proteinase-K digestion Vietnam, E. subanura (Hutterer 2005; Can et al. 2008; Kawada and phenol–chloroform–isoamyl alcohol extraction procedures, et al. 2008, 2012). Their habitats are extremely fragmented and followed by ethanol precipitation to purify extracted DNA. The usually restricted to highlands, although the ranges of these polymerase chain reaction method for amplification of Southeast Asian species are not well documented. To date, complete mitochondrial cytochrome-b (Cytb; 1,140 base pairs genetic studies to reconstruct the evolutionary history and [bp]), partial mitochondrial 12S rRNA (12S; approximately explain the current species diversity of Euroscaptor have not 900 bp), and partial nuclear recombination activating gene 1 been conducted. Motokawa (2004) examined the skulls of 2 (Rag1; 1,010 bp) genes followed our previous study (Shino- Euroscaptor species and reported that the genus is not hara et al. 2004a). The Cytb gene was first amplified using the monophyletic. By using molecular genetic markers, Shinohara universal primer pair L-14724 and H-15915 (Irwin et al. 1991) et al. (2004b, 2008) reported the phylogenetic position of the and then nested amplifications were carried out using 2 primer Malaysian mole (reported as E. micrura, but specified as E. pairs: L-14724 (Suzuki et al. 1997) and SNH-570, and H- malayana in Kawada et al. [2008]) and the short-faced mole 15916 (Suzuki et al. 2000) and SNL-491. Amplification of 12S (Scaptochirus moschatus), which is widely distributed in was performed using the universal primer pair L-613 (Mindell China; this analysis also supported a paraphyletic origin of et al. 1991) and H-1478 (Kocher et al. 1989), with nested Euroscaptor. This finding was recently supported by Crumpton amplifications carried out using 2 primer pair sets: R-L613 and and Thompson (2013). Because no extensive molecular U-H1066 (Suzuki et al. 1997), and R-L 946 (Shinohara et al. phylogenetic analyses have yet been performed with samples 2004a) and U-H1478 (Suzuki et al. 1997). The Rag1 gene was covering the broad geographic range of Euroscaptor, intensive amplified using the primer pair Rag1-F1851 (Sato et al. 2004) June 2014 SHINOHARA ET AL.—MOLECULAR PHYLOGENY OF ASIAN MOLES 457

TABLE 1.—Samples used in this study. Cytb ¼ cytochrome-b; 12S ¼ 12S RNA; Rag1 ¼ recombination activating gene 1.

GeneBank accession no. Cytb 12S Rag1 Species Collecting locality (and map code) Specimen identification Code (1,140bp) (~850 bp) (1,010 bp) Euroscaptor longirostris Tam Dao (elevation 930 m), Vinh Phuc, Vietnam (1) IEBR-M-574/SIK0775 EloTD1 AB823108 AB823143 AB823180 (n ¼ 9) IEBR-M-575/SIK0776 EloTD2 AB823109 AB823144 AB823181 IEBR-M-576/SIK0777 EloTD3 AB823110 AB823145 AB823182 SAPA (elevation 2,000 m), Lao Cai, Vietnam (2) IEBR-M-1066/SIK0820 EloSP1 AB823111 AB823146 AB823183

IEBR-M-1067/SIK0821 EloSP2 AB823112 AB823147 AB823184 Downloaded from https://academic.oup.com/jmammal/article-abstract/95/3/455/876160 by guest on 13 September 2019 IEBR-M-1068/SIK0822 EloSP3 AB823113 AB823148 AB823185 Nguyen Binh, Cao Bang, Vietnam (3) IEBR-M-1467/SIK0865 EloNB1 AB823114 AB823149 AB823186 IEBR-M-1468/SIK0866 EloNB2 AB823115 AB823150 AB823187 IEBR-M-1469/SIK0867 EloNB3 AB823116 AB823151 AB823188 Euroscaptor parvidens Chu Yang Sin National Park, Dak Lak, Vietnam (4) IEBR-M-1346/SIK0852 EpaCY1 AB823117 AB823152 AB823189 (n ¼ 6) IEBR-M-1347/SIK0853 EpaCY2 AB823118 AB823153 AB823190 IEBR-M-1348/SIK0854 EpaCY3 AB823119 AB823154 AB823191 Dong Giang, Quang Nam, Vietnam (5) IEBR-M-1352/SIK0858 EpaDG1 AB823120 AB823155 AB823192 IEBR-M-1353/SIK0859 EpaDG2 AB823121 AB823156 AB823193 IEBR-M-1354/SIK0860 EpaDG3 AB823122 AB823157 AB823194 Euroscaptor klossi Mae Sa Long, Chiang Rai, Thailand (6) SIK0673 Ekl1 AB823106 AB823141 AB823178 (n ¼ 2) SIK0674 Ekl2 AB823107 AB823142 AB823179 Euroscaptor malayana Cameron Highlands, Pahang, Malaysia (7a) SIK0550 Ema1 AB185151b AB185153b AB185155b (n ¼ 2) SIK0557 Ema2 AB185152b AB185154b AB185156b Euroscaptor mizura Ashiu, Miyama, Nantan, Kyoto, Japan (9) KT3032/HS624 Emi1 AB037604c AB106233d AB176543d (n ¼ 6) Kita-azumi, Nagano, Japan (10) KT4091 Emi2 AB823103 AB823138 AB823175 Towadako, Towada, Aomori, Japan (11) SAS3 Emi3 AB076828e AB823136 AB823173 Mt. Iwaki, Iwaki, Aomori, Japan (12) SO2001/5/26/2 Emi4 DQ630413f AB823137 AB823174 Mt. Fuji, Fujinomiya, Shizuoka, Japan (13) Miyake1 Emi5 AB823104 AB823139 AB823176 Miyake2 Emi6 AB823105 AB823140 AB823177 Mogera latouchei SAPA (elevation 1,400 m), Lao Cai, Vietnam (2) SAS148 MlaSP1 AB823091 AB823124 AB823161 (n ¼ 13) SAS149 MlaSP2 AB823092 AB823125 AB823162 SAS150 MlaSP3 AB823093 AB823126 AB823163 IEBR-M-1069/SIK0823 MlaSP4 AB823090 AB823123 AB823160 IEBR-M-1070/SIK0824 MlaSP5 AB823094 AB823127 AB823164 IEBR-M-1071/SIK0825 MlaSP6 AB823095 AB823128 AB823165 Nguyen Binh, Cao Bang, Vietnam (3) IEBR-M-1472/SIK0870 MlaNB1 AB823096 AB823129 AB823166 IEBR-M-1473/SIK0871 MlaNB2 AB823097 AB823130 AB823167 IEBR-M-1474/SIK0872 MlaNB3 AB823098 AB823131 AB823168 IEBR-M-1475/SIK0873 MlaNB4 AB823099 AB823132 AB823169 IEBR-M-1476/SIK0874 MlaNB5 AB823100 AB823133 AB823170 Co Ma, Thuan Chau, Son La, Vietnam (8) IEBR-M-1860/SIK0886 MlaCM1 AB823101 AB823134 AB823171 IEBR-M-1861/SIK0887 MlaCM2 AB823102 AB823135 AB823172 Mogera wogura Mishima, Shizuoka, Japan (14c) KT3201 — AB037623c AB106237d AB106244d Mogera imaizumii Niigata, Niigata, Japan (15c) KT2795 — AB037609c AB106236d AB106242d Mogera tokudae Sado Island, Niigata, Japan (16c) KT2672 — AB037607c AB106235d AB106243d Mogera insularis Pingtung, Pingdong, Taiwan (17g) NSMT-M34015/SIK0587 — AB181616g AB181640g AB823158 Mogera kanoana Kenting National Park, Pingtung, Taiwan (18g) NMNS009312/SIK0583 — AB181624g AB181648g AB823159 Scaptochirus moschatus Western Jilin, China (19h) MH5404 — AB306502i AB306503i AB353298i Talpa altaica Cytb, Rag1: Novosibirisk, Russiac Cytb, Rag1: MH6804 — AB037602c AY012100j AB176542d Talpa europaea Cytb, Rag1: Aarhus, Denmarke Cytb, Rag1: EM-1 — AB076829e Y19192k AB106246d 12S: Dalby, Swedenk Urotrichus talpoides Cytb, 12S: Mt. Tsurugi, Tokushima, Japan Cytb, 12S: HEG85-97 — AB076833e AB106239d AB106245d Rag1: Mt. Gomadan, Wakayama, Japan Rag1: SAS1 gracilis Mt. Yulong, Lijiang, Yunnan, China KIZ211023 — AB076700d AB106231d AB106240d

a See also Kawada et al. (2003) for collecting locality and Kawada et al. (2008) for species description. b Shinohara et al. (2004b). c Tsuchiya et al. (2000). d Shinohara et al. (2004a). e Shinohara et al. (2003). f Dubey et al. (2007). g Kawada et al. (2007). h Oda et al. (1992). i Shinohara et al. (2008). j Murphy et al. (2001). k Mouchaty et al. (2000). 458 JOURNAL OF MAMMALOGY Vol. 95, No. 3 Downloaded from https://academic.oup.com/jmammal/article-abstract/95/3/455/876160 by guest on 13 September 2019

FIG.1.—Collection localities of the Asian fossorial mole samples used in this study. Detailed information on the samples is provided in Table 1. June 2014 SHINOHARA ET AL.—MOLECULAR PHYLOGENY OF ASIAN MOLES 459 and Rag1-R2951 (¼ Rag1-R2864—Teeling et al. 2000). The version 3.2.1, the optimum substitution models for each polymerase chain reaction product was then reamplified using partition were selected by Kakusan4 (Tanabe 2011) based on the following primer sets: Rag1-F1851 and Rag1-R2486 (Sato the Bayesian information criterion. Two independent runs of 4 et al. 2004), and Rag1-F2401 (Shinohara et al. 2004a) and Markov chains were conducted for 5 million generations in the Rag1-R2951. Primer sequences are listed in Supporting Bayesian analyses. We sampled 1 tree every 100 generations Information S1 (DOI: 10.1644/13-MAMM-A-135.S1). All of and calculated a consensus topology for 375,000 trees after the polymerase chain reaction products were purified using a discarding the first 25% of trees. Parameter estimates and Qiaquick PCR purification kit (Qiagen, Tokyo, Japan), primed convergence were checked using Tracer version 1.5 (Rambaut using the Big-Dye terminator cycle sequencing kit version 3.1 et al. 2013). (Applied Biosystems, Foster City, California), and both strands Testing the monophyly of the genus Euroscaptor.—Because Downloaded from https://academic.oup.com/jmammal/article-abstract/95/3/455/876160 by guest on 13 September 2019 were directly sequenced (model 3130 or 3730; Applied the members of the genus Euroscaptor used in this study did Biosystems). Compiling of sequences from both strands was not form a monophyletic clade in our phylogenetic analyses, performed using DNASIS Mac version 3.1 (Hitachi Software we conducted hypothesis testing using both nonpartitioned ML Engineering, Tokyo, Japan). The obtained sequences were and Bayesian criteria. First, we constructed phylogenetic trees deposited in GenBank under the accession numbers using both the nonpartitioned ML and Bayesian methods AB823090–AB823194. Additional nucleotide sequences (Ta- described above, but enforcing the monophyly of the genus ble 1) were retrieved from DNA databases (DDBJ/GenBank/ Euroscaptor. The Kishino–Hasegawa test (Kishino and EMBL) and included in our phylogenetic analyses. Hasegawa 1989), the Shimodaira–Hasegawa test (Shimodaira Phylogenetic inference.—We constructed phylogenetic trees and Hasegawa 1999), and the approximately unbiased test of using maximum-likelihood (ML) and Bayesian methods. Shimodaira (2002) were performed using CONSEL version Because the sequence lengths of the 12S products varied 0.20 (Shimodaira and Hasegawa 2001) for comparisons of the among the species, they were aligned using MUSCLE multiple ML trees. Of these tests, the Kishino–Hasegawa test is known sequence alignment algorithm (Edgar 2004) implemented in to be biased when few trees were used for comparison MEGA version 5.2 (Tamura et al. 2011). We could not find (Shimodaira and Hasegawa 1999; Goldman et al. 2000; see any insertions or deletions, or both, in the Cytb and Rag1 gene also Shimodaira 2002; Strimmer and Rambaut 2002; Shi et al. sequences and aligned them manually in codon view using 2005). To reduce the bias in the Kishino–Hasegawa test, we MEGA version 5.2. The final alignment lengths of the Cytb, prepared a set of candidate trees that enforced the monophyly 12S, and Rag1 gene sequences were 1,140, 872, and 1,010 bp, of Euroscaptor using the bootstrap analyses (100 replicates) respectively, and are available as NEXUS files from the online with PAUP and reexecuted the Kishino–Hasegawa test. We database TreeBASE under the submission identification 15035. performed these alternative-topology tests using the Pairwise genetic distances of the Cytb gene sequences among nonpartitioned data set, because CONSEL does not assume populations were estimated using the Kimura-2 parameter the site likelihoods from a partitioned model (see Zwickl model (Kimura 1980) implemented in MEGA version 5.2, and 2011). are expressed as mean 6 SD. For Bayesian-based comparisons, we employed the Bayes For the ML analysis, we calculated trees using both factor (Kass and Raftery 1995). A Bayes factor is calculated as partitioned and nonpartitioned analyses. Optimal parameters a difference of harmonic mean estimates of the log-likelihood for the nonpartitioned data set were obtained using jModeltest scores between nonconstrained and constrained trees imple- 2 (Darriba et al. 2012) based on the Akaike information mented in MrBayes version 3.2.1 (Ronquist et al. 2012). A log criterion (AIC). The nonpartitioned ML trees were inferred by difference in the range of 3–5 is typically considered strong heuristic search (addition sequence ‘‘as-is’’ with tree-bisection- evidence, whereas a value above 5 is considered very strong reconnection branch swapping). The statistical confidences of evidence (Kass and Raftery 1995) in favor of the topology (see the clades were evaluated by 100 bootstrap replications. These the manual of MrBayes 3.2 [Ronquist et al. 2011]). phylogenetic analyses were carried out using the software Divergence time estimation.—The divergence time at each package PAUP* 4.0b10 (Swofford 2003). For the partitioned node was estimated using BEAST version 1.7.5 (Drummond et ML analysis, we defined 12S rRNA and each codon position of al. 2012). Although the fossil record of Asian fossorial moles is both Cytb and Rag1 as separate data blocks. The best-fit not well documented, we set 2 calibration points. First, the partitioning scheme and substitution models were selected by earliest record of Talpini in East Asia is recovered from early Partitionfinder version 1.1.1 (Lanfear et al. 2012) under the Miocene (Burdigalian stage: 20.5–16.4 million years ago AIC. The partitioned ML analysis was carried out using Garli [mya]) of Olkhon, Irkutsk region, Russia (Fortelius 2013). 2.0 (Zwickl 2006) with 5 replicated searches. The statistical Therefore, we treated the calibration as a lognormal confidences of the clades also were evaluated by 100 bootstrap distribution, set the earliest possible age to 16.4 mya and the replications using Garli 2.0. older 95% to 20.5 mya (offset ¼ 16.4 mya, X¯ ¼ 1.3 mya, and The Bayesian analysis was conducted using MrBayes SD ¼ 1.0 mya). Second, the oldest known fossil of version 3.2.1 (Ronquist et al. 2012). For the analysis, the data Scaptochirus was recovered from the Gaozhuang Formation set was divided into 7 partitions as suggested by Partition- (4.3–4.0 mya) of Yushe Basin, Shanxi Province, China (Flynn finder. To implement the best substitution models for MrBayes and Wu 1994). This record was used to calibrate the split 460 JOURNAL OF MAMMALOGY Vol. 95, No. 3 between Scaptochirus and its sister taxon as a uniform 83–86% bootstrap support; Fig. 2). In addition, the Kishino– distribution (lower boundary ¼ 4.3 mya). Hasegawa, Shimodaira–Hasegawa, and approximately The data set was divided into 7 partitions and the best-fit unbiased tests significantly (P , 0.05) rejected monophyly partitioning scheme and substitution models for each partition of Euroscaptor in paired tests (Table 2). When compared with were selected by Partitionfinder version 1.1.1 (Lanfear et al. multiple trees obtained by bootstrap analyses (n ¼ 101, because 2012) under the Bayesian information criterion. We employed 1 of the 100 replicates found 2 trees) in the Kishino–Hasegawa relaxed uncorrelated lognormal clock models, and adopted test, the monophyly of Euroscaptor is mostly rejected (Table them independently for mitochondrial and nuclear genes. We 2). Furthermore, comparison of Bayes factors also indicated used a random starting tree, Yule process tree prior, and the that a monophyletic Euroscaptor was not supported; the program’s default prior distributions of model parameters. The difference of harmonic mean log-likelihood scores between the Downloaded from https://academic.oup.com/jmammal/article-abstract/95/3/455/876160 by guest on 13 September 2019 analysis was run for 20 million generations and sampled every best and constrained trees (logB10) is 14.39, indicating very 1,000 generations. strong evidence (see Kass and Raftery 1995). The remaining 3 species of Euroscaptor (E. malayana, E. klossi,andE. longirostris) formed a clade with strong support from all ML RESULTS and Bayesian analyses, but there was no clear support for For the nonpartitioned ML analysis, the GTR (Tavare´ 1986) relationships among the 3 species. þIþG was selected as the best substitution model for the In contrast to Euroscaptor, the monophyly of Mogera was concatenated data set. For the partitioned ML analysis, the 7 strongly supported (1.00 posterior probability; 98–100% ML partitioning was selected as the best scheme with following bootstrap; Fig. 2). The trees constructed with concatenated models: SYM (Zharkikh 1994) þG; TIM (Posada and Crandall gene sequences recovered a weekly supported clade compris- 2001) þI; GTRþIþG; GTRþIþG; GTR; K81uf (Posada and ing the 3 Japanese species (M. wogura, M. imaizumii and M. Crandall 2001) þI; and TIMef (Posada and Crandall 2001) þG tokudae) to the exclusion of those from Taiwan (M. insularis model for the 1st, 2nd, and 3rd codon positions of Cytb, 12S, and M. kanoana) and Southeast Asia (M. latouchei; 0.85 and the 1st, 2nd, and 3rd codon positions of Rag1, respectively. posterior probability; 62–69% bootstrap support; Fig. 2). For the Bayesian analysis, SYM (Zharkikh 1994) þG; HKY85 However, the sister relationship between the Taiwanese and (Hasegawa et al. 1985) þGþI; HKY85þG; GTRþG; JC69 Southeast Asian species was robust (1.00 posterior probability; (Jukes and Cantor 1969); F81 (Felsenstein 1981) þI; and K80 89–95% bootstrap support; Fig. 2). The 13 specimens of M. (Kimura 1980) þG were selected as the best substitution latouchei were grouped according to the 3 collecting localities, models for the 1st, 2nd, and 3rd codon positions of Cytb, 12S, but genetic diversities between each population were not high and the 1st, 2nd, and 3rd codon positions of Rag1, respectively. (Kimura-2 parameter distances in Cytb: from Lao Cai versus The final average SD of split frequencies of the Bayesian Cao Bang provinces: 1.30% 6 0.11%, from Lao Cai versus analysis was 0.002631, and all average effective sample sizes Son La provinces: 1.32% 6 0.03%, and from Cao Bang versus were . 200. Son La provinces: 1.56% 6 0.08%). Incongruence among the nonpartitioned ML, partitioned Each of the species currently assigned to Euroscaptor was ML, and Bayesian trees was found only between the topologies recovered as monophyletic, even if the genus as a whole was within a monophyletic E. mizura (see Fig. 2). This incongru- not. Individuals of E. mizura were collected from 5 localities in ence is due to the position of Emi1, which was collected in Japan, and exhibited a relatively high degree of genetic Kyoto, Japan. Neither of these alternative relationships divergence given their geographic distribution. For example, received strong support; therefore, we show only the non- the genetic divergence among individuals of E. mizura partitioned ML tree with the bootstrap support values for (Kimura-2 parameter distances in Cytb: 5.95% 6 2.63%, nonpartitioned and partitioned analyses of the ML and the ranging from 0.18% to 8.04%) is comparable to the posterior probabilities of the Bayesian analysis on each node interspecies genetic diversities among E. malayana, E. klossi, (Fig. 2). and E. longirostris (Kimura-2 parameter distances in Cytb: Phylogenetic relationships.—The phylogenetic trees 8.35% 6 0.71%, ranging from 7.00% to 9.35%). strongly supported monophyly of the Asian fossorial mole The remaining Euroscaptor specimens collected from lineages relative to Talpa. Within the Asian taxa, we recovered Southeast Asia were divided into 2 groups: one included E. Mogera as a monophyletic group (1.00 posterior probability; parvidens from central Vietnam and the other comprised E. 98–100% ML bootstrap; Fig. 2), but Euroscaptor was malayana, E. klossi,andE. longirostris from Malaysia, paraphyletic. The Japanese species E. mizura was recovered Thailand, and northern Vietnam, respectively (Fig. 1). as the sister taxon to the other Asian fossorial moles, with Interestingly, monophyly of E. longirostris was not strongly Scaptochirus forming part of a clade including the remaining supported in our analyses (0.93 posterior probability; 56–75% species of Euroscaptor. Support for the sister-taxon ML bootstrap; Fig. 2), with 2 clades with strong supports relationship between Scaptochirus and E. parvidens was not recovered: the first from Vinh Phuc (EloTD1-3) and Cao Bang concrete (0.97 posterior probability, but only 61–70% provinces (EloNB1-3), Vietnam, and the second from Lao Cai bootstrap support; Fig. 2), whereas support for the position Province, Vietnam (EloSP1-3). The genetic distances between of E. mizura was more reliable (1.00 posterior probability but these 2 clades are substantial (Kimura-2 parameter distances in June 2014 SHINOHARA ET AL.—MOLECULAR PHYLOGENY OF ASIAN MOLES 461 Downloaded from https://academic.oup.com/jmammal/article-abstract/95/3/455/876160 by guest on 13 September 2019

FIG.2.—Maximum-likelihood (ML) tree constructed using cytochrome-b (Cytb; 1,140 bp), 12S RNA (12S; approximately 830 bp), and recombination activating gene 1 (Rag1; 1,010 bp) gene sequences without partitioning. The partitioned ML and Bayesian analyses had an almost identical tree topology, except within Euroscaptor mizura (*). Bootstrap values in nonpartitioned ML, partitioned ML analyses, and posterior probabilities in Bayesian analysis are shown on each node, respectively. Branch lengths are given as the ML distance with the selected model. The numbers in the parentheses following taxon labels show the map code in Fig. 1. Each terminal taxon name is identified by a code as reported in Table 1. 462 JOURNAL OF MAMMALOGY Vol. 95, No. 3

TABLE 2.—Tests for the monophyly of the genus Euroscaptor in maximum-likelihood analysis. * P , 0.05.

P-value Maximum likelihood–based lnL Kishino–Hasegawa test Shimodaira–Hasegawa test Approximately unbiased test Paired comparison Original tree (Fig. 2) 14,931.69751 — — -— Constrained best tree 14,948.47768 0.049* 0.049* 0.038* Multiple comparison Constrained bootstrap trees (n ¼ 91) 14,949.20522–15,001.86694 , 0.05* — — Constrained bootstrap trees (n ¼ 8) 14,948.47768–14,954.42633 0.051–0.057 — — Downloaded from https://academic.oup.com/jmammal/article-abstract/95/3/455/876160 by guest on 13 September 2019 Constrained bootstrap trees (n ¼ 2) 14,951.94884–14,952.24992 0.061 and 0.064 — —

Cytb: 6.08% 6 0.10%), and are comparable to the genetic and Southeast Asia, based on molecular phylogenetic analysis distance between the 2 collecting localities of E. parvidens using both mitochondrial and nuclear gene sequences from the (from Dak Lak Province: EpaCY1-3, and from Quang Nam 3 Asian genera Euroscaptor (5 species), Mogera (6 species), Province: EpaDG1-3; Kimura-2 parameter distances in Cytb: and Scaptochirus (monotypic) and the European genus Talpa 6.55% 6 0.20%). (2 species). The results reveal previously unknown phases of Divergence time estimation.—Our BEAST analysis lineage differentiation, at the genus and species levels, among suggested that the European clade Talpa diverged from Asian the Asian taxa (Fig. 2). fossorial lineages 20.9 mya (95% confidence interval ¼ 18.0– Lineage divergence of the East and Southeast Asian 25.2 mya), with the ancestral Asian fossorial mole lineage fossorial moles.—The monophyly of Mogera was strongly diverging into E. mizura (node B in Fig. 2), Mogera (node C), E. supported by our phylogenetic analyses, with our divergence klossi–E. malayana–E. longirostris (node D), and E. parvidens time analysis suggesting this clade emerged around 10.0 mya and S. moschatus (node F) between 10.5 and 19.7 mya (Table 3; (8.0–12.6 mya [Table 3]). At least 4 additional genetically Supporting Information S2, DOI: 10.1644/13-MAMM-A-135. distinct East and Southeast Asian lineages were recovered in S2). The interspecies diversification of Mogera (node G) started our analyses: Scaptochirus, E. mizura, E. parvidens, and E. around 10.0 mya (8.0–12.6 mya). The diversification of E. malayana–E. klossi–E. longirostris (Fig. 2; Supporting klossi–E. malayana–E. longirostris (node K) was estimated at Information S2). These 5 lineages diverged between 10.5 and 5.3 mya (3.8–7.9 mya), whereas local populations of E. mizura 19.7 mya (Table 3; Supporting Information S2), which is (node M: 4.1 mya, 2.9–5.7 mya) and E. parvidens (node O: 3.2 consistent with fossil records: fossorial moles were widely mya, 2.0–4.6 mya) diverged more recently. distributed in the Northern Hemisphere in the middle Miocene (McKenna and Bell 1997; Ziegler 2003). During this period, the global climate dramatically changed to cool and dry after DISCUSSION the end of a warm and humid period called the middle Miocene In this study, we aimed to provide deeper insights into the climatic optimum (17–15 mya—Zachos et al. 2001). In the species diversity of fossorial moles in the tribe Talpini in East Asian region, the Indian and East Asian monsoons began and

TABLE 3.—Estimated divergence times based on concatinated data sets using BEAST. Node codes correspond to Fig. 2. Time-scaled tree is provided in Supporting Information S2.

Divergence time (mya) Node Lineage X¯ 95% highest posterior density A Eurasian moles (tribe Talpini) 20.9 18.0–25.2 B Asian moles (Euroscaptor, Scaptochirus, and Mogera) 17.1 16.4–19.7 C Asian moles (except E. mizura) 16.1 14.0–18.9 D E. malayana þ E. klossi þ E. longirostris þ E. parvidens þ S. moschatus 14.8 12.0–17.6 E Talpa 13.5 9.5–18.0 F E. parvidens þ S. moschatus 13.4 10.5–16.6 G Mogera 10.0 8.0–12.6 H Japanese moles (M. tokudae þ M. wogura þ M. imaizumii) 8.6 6.3–10.9 I Taiwanese moles (M. insularis þ M. kanoana) þ M. latouchei 8.0 6.1–10.4 J M. wogura þ M. imaizumii 5.7 3.8–7.9 K E. malayana þ E. klossi þ E. longirostris 5.3 3.8–7.0 L Taiwanese moles (M. insularis þ M. kanoana) 4.5 3.0–6.2 M E. mizura 4.1 2.9–5.7 N E. longirostris 3.7 2.5–5.2 O E. parvidens 3.2 2.0–4.6 P M. latouchei 0.8 0.5–1.2 June 2014 SHINOHARA ET AL.—MOLECULAR PHYLOGENY OF ASIAN MOLES 463 developed in association with the uplift of the Himalayan– mole) and M. wogura (Japanese ‘‘western’’ mole), to the Tibetan plateau (Amano and Taira 1992; An et al. 2001); exclusion of M. tokudae (endemic to Sado Island), also was furthermore, middle-latitude aridity appeared (Fortelius et al. supported by our results. Tsuchiya (1990) proposed the 2002; Liu et al. 2009) at this time. Such drastic climate changes hypothesis that M. tokudae (including M. etigo), M. likely shaped the biodiversity of the Asian fossorial mole imaizumii, and M. wogura differentiated on the continent and lineages. entered Japan from the Korean Peninsula in this order during Revisiting the genus Euroscaptor.—Our phylogenetic trees separate Pleistocene glacial periods based on combined did not support the monophyly of Euroscaptor (Fig. 2), and our evidence from biogeography, morphology, karyology, and hypothesis testing also rejected it (Table 2). The isozyme analyses (see Shinohara et al. 2005). This hypothesis nonmonophyletic status of Euroscaptor is due to 2 factors: was well supported in recent molecular phylogenetic studies Downloaded from https://academic.oup.com/jmammal/article-abstract/95/3/455/876160 by guest on 13 September 2019 the early split of E. mizura from the other Asian fossorial mole (Tsuchiya et al. 2000; Shinohara et al. 2004a, 2005; Kirihara et taxa and interposition of S. moschatus. Five genera of fossorial al. 2013). Our results provide further support for this moles (Mogera, Euroscaptor, Scaptochirus, Parascaptor, and hypothesis, in that an identical topology was recovered Scapanulus [Hutterer 2005]) are distributed in East and following the addition of the Taiwanese and Vietnamese Southeast Asia. Interestingly, Scapanulus is classified specimens. together with 3 North American fossorial mole genera In Southeast Asia, M. latouchei is distributed rather widely (Scalopus, Parascalops, and ) in the tribe Scalopini, (from southeastern China to northern Vietnam, but scattered) whereas the remaining 4 genera belong to tribe Talpini. These compared to M. kanoana and M. insularis. The of latter genera are primarily classified by dental formula because this former species is controversial; it is usually treated as a they exhibit different combinations of tooth loss (Allen 1938; subspecies of M. insularis (e.g., Hutterer 2005). Conversely, Ziegler 1971); the dental formula observed in Euroscaptor is Kawada et al. (2007, 2009) reported that M. latouchei should hypothesized to be the ancestral condition (Ziegler 1971; be recognized as a separate species based on morphological Motokawa 2004; Kawada and Yokohata 2009). and karyological characteristics. Our molecular phylogenetic According to the molecular phylogeny inferred from the data support this latter conclusion (Fig. 2). Although we were present study, the dental formula specific to Euroscaptor thus able to survey only 3 localities from northern Vietnam, we did appears to be plesiomorphic. As first suggested by Motokawa not detect any marked genetic diversity among specimens of (2004), we propose that the taxonomic status of the genus this species (Fig. 2). However, Kawada et al. (2009) reported Euroscaptor should be reconsidered. The genus Euroscaptor that Vietnamese M. latouchei was morphologically different was originally applied to E. klossi from Thailand (type species from Chinese M. latouchei, suggesting the existence of a Talpa klossi Thomas, 1929) by Miller (1940). In our cryptic species in the continental southern Mogera lineage. In phylogenetic analyses, specimens of E. klossi collected from addition, Kawada et al. (2007) reported that the 2 Taiwanese Thailand fit within a clade comprising E. klossi, E. malayana, species showed different ranges of intraspecific genetic and E. longirostris, although the relationships among these 3 diversity, with much greater diversity observed within M. species was not fully resolved (Fig. 2). Nonetheless, E. mizura kanoana than within M. insularis, and concluded that M. is not a true member of the genus Euroscaptor as evidenced by insularis expanded rapidly in Taiwan over a short period. the early split of this lineage from the Asian fossorial mole Together with our findings, these observations provide us with lineages recovered in our tree (Fig. 2). Surprisingly, our a hypothesis of multiple dispersal events in the Mogera linage. phylogenetic trees further indicated that E. mizura has a high Evolution of Southeast Asian Euroscaptor.—In the intraspecific genetic diversity (Fig. 2). The habitat of E. mizura Southeast Asian Euroscaptor moles, 2 lineages were is in the high mountain zones and its distribution is thus highly recovered: E. parvidens (central Vietnam) and E. malayana fragmented (Abe 2005; Kawada and Yokohata 2009). (Malaysia)–E. klossi (Thailand)–E. longirostris (northern Therefore, the large intraspecific genetic diversity observed Vietnam [Fig. 2]). Within these lineages, 6 conspecific local in this lineage may correlate with a long history of isolation populations were recognized: E. malayana, E. klossi, and E. between populations. We propose that E. mizura was widely longirostris from Vinh Phuc and Cao Bang provinces; E. distributed over the Japanese islands in the past, but became longirostris from Lao Cai Province; E. parvidens from Dak displaced to higher elevations following subsequent invasions Lak Province; and E. parvidens from Quang Nam Province. To by larger Mogera species. Further phylogeographic study may date, the phylogenetic relationships among these southeastern help resolve this issue. Euroscaptor species have remained unknown. In this study, we Evolution and biogeography of Mogera.—Our phylogenetic provide molecular evidence that each species forms a trees clearly supported the monophyly of Mogera, as suggested monophyletic clade with high support values, except for E. by others (Shinohara et al. 2003, 2004a, 2004b, 2008; Kawada longirostris (Fig. 2). The known distribution of E. longirostris et al. 2007; Crumpton and Thompson 2013). Among the 6 is northern Vietnam and southwestern China, and the species of Mogera, our phylogenetic trees strongly supported morphology of this species is highly variable (Kawada et al. the monophyly of the Taiwanese and Vietnamese species (M. 2009). In addition, Kawada et al. (2009) pointed out that the 2 kanoana, M. insularis, and M. latouchei, respectively [Fig. 2]). populations of E. longirostris (from Vinh Phuc and Cao Bang A close relationship between M. imaizumii (Japanese ‘‘eastern’’ provinces, and from Lao Cai Province) are morphologically 464 JOURNAL OF MAMMALOGY Vol. 95, No. 3

different in terms of coat color, tail shape, and tail length. ABE, H. 2005. Insectivora. Pp. 1–24 þ 154–158 in A guide to the These morphological groupings clearly correlate with our of Japan Revised edition (H. Abe, ed.). Revised ed. Tokai molecular phylogenetic results (Fig. 2). Therefore, the University Press, Kanagawa, Japan. taxonomy of E. longirostris remains a matter of debate. ALLEN, G. M. 1938. The mammals of China and Mongolia. American The divergence times of the 6 conspecific local populations Museum of Natural History, New York 1:1–620 þ 9. AMANO,K.,AND A. TAIRA. 1992. Two-phase uplift of Higher of southeastern Euroscaptor were estimated to have taken Himalayas since 17 Ma. Geology 20:391–394. place between 2.0 and 7.0 mya (Table 3). It is not yet certain AN, Z., J. E. KUTZBACH,W.L.PRELL, AND S. C. PORTER. 2001. how these divergences were generated because our sampling is Evolution of Asian monsoons and phased uplift of the Himalaya– very limited in light of the wide distributions of these moles: Tibetan plateau since late Miocene times. Nature 411:62–66. Downloaded from https://academic.oup.com/jmammal/article-abstract/95/3/455/876160 by guest on 13 September 2019 from Peninsular Malaysia in the south, to southern China in the CABRIA, M. T., J. RUBINES,B.GO´ MEZ-MOLINER, AND R. ZARDOYA. 2006. north, and to the Himalayan region in the west (Hutterer 2005). On the phylogenetic position of a rare Ibelian endemic , the In addition, their habitats are highly fragmented, with accurate (Galemys pyrenaicus). Gene 375:1–13. distribution maps still lacking. However, observed intraspecific CAN, D. N., ET AL. 2008. Checklist of wild mammal species of genetic divergences among several southeastern Euroscaptor Vietnam. Primate Research Institute, Inuyama, Japan. taxa may reflect histories of long isolation. For instance, we COLANGELO, P., ET AL. 2010. Molecular systematics and evolutionary collected E. longirostris and M. latouchei from different biogeography of the genus Talpa (Soricomorpha: Talpidae). altitudes of the same mountain districts in Vietnam (SAPA, Molecular Phylogenetics and Evolution 55:372–380. Lao Cai Province, and Nguyen Binh, Cao Bang Province CRUMPTON, N., AND R. S. THOMPSON. 2013. The holes of moles: osteological correlates of the trigeminal nerve in Talpidae. Journal [Table 1; see also Kawada et al. 2009]) and detected much of Mammalian Evolution 20:213–225. larger genetic distances in E. longirostris (Kimura-2 parameter DARRIBA, D., G. L. TABOADA,R.DOALLO, AND D. POSADA. 2012. distances in Cytb: 6.09% 6 0.13%) than M. latouchei (1.30% jModeltest 2: more models, new heuristics and parallel computing. 6 0.11%). Such differences of genetic distance and habitat Nature Methods 9:772. altitude between 2 species may suggest a recent invasion of M. DRUMMOND, A. J., M. A. SUCHARD,D.XIE, AND A. RAMBAUT. 2012. latouchei into the habitat of E. longirostris. Finally, it is of note Bayesian phylogenetics with BEAUti and the BEAST 1.7. that E. parvidens formed a clade with Scaptochirus, though Molecular Biology and Evolution 29:1969–1973. support values are not high. To better understand the DUBEY, S., N. SALAMIN,S.D.OHDACHI,P.BARRIERE` , AND P. VOGEL. evolutionary histories of these and other Southeast Asian 2007. Molecular phylogenetics of (Mammalia: Soricidae) species in the Talpini, finer taxon sampling together with reveal timing of transcontinental colonizations. Molecular Phylo- accurate distribution maps—as are currently available for genetics and Evolution 44:126–137. Japanese fossorial species (Abe 2005; Ohdachi et al. 2009)— EDGAR, R. C. 2004. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Research 32:1792– are essential. 1797. FELSENSTEIN, J. 1981. Evolutionary trees from DNA sequencies—a ACKNOWLEDGMENTS maximum-likelihood approach. Journal of Molecular Evolution 17:368–376. This study was conducted using the results of our continuous field FLYNN, J. F., AND Y. W. WU. 1994. Two new shrews from the Pliocene survey in Southeast Asia, cited in the literature. Such field surveys of Yushe Basin, Shanxi Province, China. Vertebrata PalAsiatica could not have been done without the dedicated support of many 32:73–86. anonymous local people. We sincerely appreciate their help, kindness, FORTELIUS, M. (coordinator). 2013. New and Old Worlds database of and friendship. We also thank K. Tsuchiya, M. A. Iwasa, S. D. fossil mammals (NOW). University of Helsinki, Helsinki, Finland. Ohdachi, and T. Miyake for providing tissues of E. mizura, and H. Kai http://www.helsinki.fi/science/now/. Accessed 31 August 2013. and K. L. Campbell for help with revising the manuscript. This study FORTELIUS, M., ET AL. 2002. Fossil mammals resolve regional patterns was supported in part by MEXT/JSPS KAKENHI grants 15770060 (AS), 21370033 (HE), and 19405010 (HS) and by the Vietnam of Eurasian climate change over 20 million years. Evolutionary National Foundation for Science and Technology Development Ecology Research 4:1005–1016. (NAFOSTED) grant 106.12–2011.06. GOLDMAN, N., J. P. ANDERSON, AND A. G. RODRIGO. 2000. Likelihood- based tests of topologies in phylogenetics. Systematic Biology 49:652–670. SUPPORTING INFORMATION HASEGAWA, M., H. KISHINO, AND T. YANO. 1985. Dating of the human– ape splitting by a molecular clock of mitochondrial DNA. Journal SUPPORTING INFORMATION S1.—Primer sequences used in this study. Found at DOI: 10.1644/13-MAMM-A-135.S1 of Molecular Evolution 22:160–174. HUTTERER, R. 2005. Order Soricomorpha. Pp. 220–311 in Mammal SUPPORTING INFORMATION S2.—Divergence time estimates using BEAST. species of the world: a taxonomic and geographic reference (D. E. Found at DOI: 10.1644/13-MAMM-A-135.S2 Wilson and D. M. Reeder, eds.). 3rd ed. Johns Hopkins University Press, Baltimore, Maryland. IRWIN, D. M., T. D. KOCHER, AND A. C. WILSON. 1991. Evolution of the LITERATURE CITED cytochrome b gene of mammals. Journal of Molecular Evolution ABE, H. 1974. Change of the boundary-line of two moles’ 32:128–144. distributions in a period of 14 years. Journal of the Mammalogical IWASA, M. A., C. KAWAKUBO,K.TSUCHIYA, AND H. SUZUKI. 2006. Society of Japan 6:13–23 (in Japanese with English summary). Intraspecific differentiation in the lesser in eastern June 2014 SHINOHARA ET AL.—MOLECULAR PHYLOGENY OF ASIAN MOLES 465

Honshu, Japan, indicated by nuclear and mitochondrial gene MCKENNA, M. C., AND S. K. BELL. 1997. Classification of mammals analyses. Zoological Science 23:955–961. above the species level. Columbia University Press, New York. JUKES, T. H., AND C. R. CANTOR. 1969. Evolution of protein molecules. MILLER, G. S., JR. 1940. Notes on some moles from Southeastern Asia. Pp. 21–132 in Mammalian protein metabolism (H. N. Munro, ed.). Journal of Mammalogy 21:442–444. Academic Press, New York. MINDELL, D. P., C. W. DICK, AND R. J. BAKER. 1991. Phylogenetic KASS, R. E., AND A. E. RAFTERY. 1995. Bayes factors. Journal of relationships among megabats, microbats, and primates. Proceed- American Statistical Association 90:773–795. ings of the National Academy of Science of the USA 88:10322– KAWADA, S., S. KOBAYASHI,H.ENDO,W.RERKAMNUAYCHOKE, AND S. 10326. ODA. 2006. Karyological study on Kloss’s mole Euroscaptor klossi MOTOKAWA, M. 2004. Phylogenetic relationships within the family Talpidae (Mammalia: Insectivora). Journal of Zoology (London) (Insectivora, Talpidae) collected in Chiang Rai Province, Thailand. Downloaded from https://academic.oup.com/jmammal/article-abstract/95/3/455/876160 by guest on 13 September 2019 Mammal Study 31:105–109. 263:147–157. KAWADA, S., A. SHINOHARA,S.KOBAYASHI,M.HARADA,S.ODA, AND L. MOUCHATY, S. K., A. GULLBERG,A.JANKE, AND U. ARNASON. 2000. The K. LIN. 2007. Revision of the mole genus Mogera (Mammalia: phylogenetic position of the Talpidae within based on Lipotyphla: Talpidae) from Taiwan. Systematics and Biodiversity analysis of complete mitochondrial sequences. Molecular Biology 5:223–240. and Evolution 17:60–67. KAWADA, S., A. SHINOHARA,M.YASUDA,S.ODA, AND B. L. LIM. 2003. MURPHY, W. J., E. EIZIRIK,W.E.JOHNSON,Y.P.ZHANG,O.A.RYDER, The mole of Peninsular Malaysia: notes on its identification and AND S. J. O’BRIEN. 2001. Molecular phylogenetics and the origins of ecology. Mammal Study 28:73–77. placental mammals. Nature 409:614–618. ODA, S., ET AL. 1992. Survey and capture of small wild mammals in KAWADA, S., N. T. SON, AND D. N. CAN. 2009. Moles (Insectivora, Talpidae, ) of Vietnam. Bulletin of the National Museum the central and west regions of Jilin Province, China. Annals of the of Nature and Science 35:89–101. Research Institute of Environmental Medicine, Nagoya University 43:244–247 (in Japanese). KAWADA, S., N. T. SON, AND D. N. CAN. 2012. A new species of mole of the genus Euroscaptor (Soricomorpha, Talpidae) from northern OHDACHI, S. D., Y. ISHIBASHI,M.A.IWASA, AND T. SAITOH (eds.). 2009. Vietnam. Journal of Mammalogy 93:893–850. The wild mammals of Japan. Shoukadoh Book Sellers and Mammalogical Society of Japan, Kyoto, Japan. KAWADA, S., M. YASUDA,A.SHINOHARA, AND B. L. LIM. 2008. OKAMOTO, M. 1999. Phylogeny of Japanese moles inferred from Redescription of the Malaysian mole as to be a true species, mitochondrial COI gene sequences. Pp. 21–27 in Recent advances Euroscaptor malayana (Insectivora, Talpidae). Memoirs of the in the biology of Japanese Insectivora (Y. Yokohata and S. National Museum of Nature and Science 45:65–74. Nakamura, eds.). Hiba Society of Natural History, Shobara, Japan. KAWADA, S., AND Y. YOKOHATA. 2009. Euroscaptor mizura (Gunther,¨ POSADA,D.,AND K. A. CRANDALL. 2001. Selecting models of 1880). Pp. 32–33 in The wild mammals of Japan (S. D. Ohdachi, Y. nucleotide substitution: an application to human immunodeficiency Ishibashi, M. A. Iwasa, and T. Saitoh, eds.). Shoukadoh Book virus1 (HIV-1). Molecular Biology and Evolution 18:897–906. Sellers and Mammalogical Society of Japan, Kyoto, Japan. RAMBAUT, A., M. A. SUCHARD,D.XIE, AND A. J. DRUMMOND. 2013. KIMURA, M. 1980. A simple method for estimating evolutionary rates Tracer v1.5. http://tree.bio.ed.ac.uk/software/tracer/. Accessed 24 of base substitutions through comparative studies of nucleotide July 2012. sequences. Journal of Molecular Evolution 16:111–120. RONQUIST, F., J. HUELSENBECK, AND M. TESLENKO. 2011. Draft MrBayes KIRIHARA,T.,A.SHINOHARA,K.TSUCHIYA,M.HARADA,A.P. version 3.2 manual: tutorials and model summaries. http://mrbayes. KRYUKOV, AND H. SUZUKI. 2013. Spatial and temporal aspects of sourceforge.net/manual.php. Accessed 24 July 2012. Mogera species occurrence in the Japanese Islands inferred from RONQUIST,F.,ET AL. 2012. MrBayes 3.2: efficient Bayesian mitochondrial and nuclear gene sequences. Zoological Science phylogenetic inference and model choice across a large model 30:267–281. space. Systematic Biology 61:539–542. ISHINO AND ASEGAWA K , H., M. H . 1989. Evaluation of the maximum SA´ NCHEZ-VILLAGRA, R. M., I. HOROVITZ, AND M. MOTOKAWA. 2006. A likelihood estimate of the evolutionary tree topologies from DNA comprehensive morphology analysis of talpid moles (Mammalia) sequence data, and the branching order in Hominoidea. Journal of phylogenetic relationships. Cladistics 22:59–88. Molecular Evolution 29:170–179. SATO, J. J., T. HOSODA,M.WOLSAN, AND H. SUZUKI. 2004. Molecular KOCHER, T. D., ET AL. 1989. Dynamics of mitochondrial DNA phylogeny of arctoids (Mammalia: Carnivora) with emphasis on evolution in : amplification and sequencing with conserved phylogenetic and taxonomic positions of the ferret-badgers and primers. Proceedings of the National Academy of Science of the skunks. Zoological Science 21:111–118. USA 86:6196–6200. SHI, X., H. GU,E.SUSKO, AND C. FIELD. 2005. The comparison of the KOH, H. S., ET AL. 2012. Lack of genetic divergence between Mogera confidence regions in phylogeny. Molecular Biology and Evolution wogura coreana from Korea and M. w. robusta from northeastern 22:2285–2296. China and adjacent Russia (Soricomorpha: Mammalia), reexamined SHIMODAIRA, H. 2002. An approximately unbiased test of phylogenetic from 12S rRNA and cytochrome b sequences. Cells and tree selection. Systematic Biology 51:492–508. Systems 16:408–414. SHIMODAIRA, H., AND M. HASEGAWA. 1999. Multiple comparisons of LANFEAR, R., B. CALCOTT,S.Y.HO, AND S. GUINDON.2012. log-likelihoods with applications to phylogenetic inference. Molec- Partitionfinder: combined selection of partitioning schemes and ular Biology and Evolution 16:1114–1116. substitution models for phylogenetic analyses. Molecular Biology SHIMODAIRA, H., AND M. HASEGAWA. 2001. CONSEL: for assessing the and Evolution 29:1695–1701. confidence of phylogenetic tree selection. Bioinformatics 17:1246– LIU, L., J. T. ERONEN, AND M. FORTELIUS. 2009. Significant mid-latitude 1247. aridity in the middle Miocene of East Asia. Palaeogeography, SHINOHARA, A., K. L. CAMPBELL, AND H. SUZUKI. 2003. Molecular Palaeoclimatology, Palaeoecology 279:201–206. phylogenetic relationships of moles, -moles and desmans 466 JOURNAL OF MAMMALOGY Vol. 95, No. 3

from the New and Old Worlds. Molecular Phylogenetics and combined molecular phylogenetic analyses of multilocus sequence Evolution 26:247–258. data. Molecular Ecology Resources 11:914–921. SHINOHARA, A., K. L. CAMPBELL, AND H. SUZUKI. 2005. An evolutionary TAVARE´ , S. 1986. Some probabilistic and statistical problems in the view on the Japanese talpids based on nucleotide sequences. analysis of DNA sequences. Pp. 57–86 in Some mathematical Mammal Study 30:S19–S24. questions in biology—DNA sequence analysis (R. M. Miura, ed.). SHINOHARA, A., ET AL. 2004a. Evolution and biogeography of talpid American Mathematical Society, Providence, Rhode Island. moles from continental East Asia and the Japanese Islands inferred TEELING, E. C., O. MADSEN,R.A.VAN DEN BUSSCHE,W.W.DE JONG, from mitochondrial and nuclear gene sequences. Zoological Science M. J. STANHOPE, AND M. S. SPRINGER. 2000. Microbat paraphyly and 21:1177–1185. the convergent evolution of a key innovation in Old World SHINOHARA, A., S. KAWADA,M.HARADA,K.KOYASU,S.ODA, AND H. rhinolophoid microbats. Proceedings of the National Academy of Downloaded from https://academic.oup.com/jmammal/article-abstract/95/3/455/876160 by guest on 13 September 2019 SUZUKI. 2008. Phylogenetic relationships of the short-faced mole, Sciences of the USA 99:1431–1436. Scaptochirus mochatus (Mammalia: ), among Eurasian THOMAS, O. 1929. A new mole from western Siam. Annals and fossorial moles, as inferred from mitochondrial and nuclear gene sequences. Mammal Study 33:77–82. Magazine of Natural History, Series 10, 3:206–207. TSUCHIYA, K. 1990. Phylogeny and evolution of the Japanese moles. SHINOHARA, A., S. KAWADA,M.YASUDA, AND B. L. LIM. 2004b. Phylogenetic position of the Malaysian mole, Euroscaptor micrura Collecting and Breeding 52:378–380 (in Japanese). (Mammalia: Eulipotyphla), inferred from three gene sequences. TSUCHIYA, K., ET AL. 2000. Molecular phylogeny of East Asian moles Mammal Study 29:185–189. inferred from the sequence variation of the mitochondrial STRIMMER, K., AND A. RAMBAUT. 2002. Inferring confidence sets of cytochrome b gene. Genes & Genetic Systems 75:17–24. possibly misspecified gene trees. Proceedings of the Royal Society ZACHOS, J., M. PAGANI,L.SLOAN,E.THOMAS, AND K. BILLUPS. 2001. of London, B. Biological Sciences 269:137–142. Trends, rhythms, and aberrations in global climate 65 Ma to SUZUKI, H., S. MINATO,S.SAKURAI,K.TSUCHIYA, AND I. M. FOKIN. present. Science 292:686–693. 1997. Phylogenetic position and geographic differentiation of the ZHARKIKH, A. 1994. Estimation of evolutionary distances between Japanese dormouse, Glirulus japonicus, revealed by variations nucleotide sequences. Journal of Molecular Evolution 39:315–329. among rDNA, mtDNA and the Sry gene. Zoological Science ZIEGLER, A. C. 1971. Dental homologies and possible relationships of 14:167–173. Recent Talpidae. Journal of Mammalogy 52:50–68. SUZUKI, H., K. TSUCHIYA, AND N. TAKEZAKI. 2000. A molecular ZIEGLER, R. 2003. Moles (Talpidae) from the late middle Miocene of phylogenetic framework for the Ryukyu endemic rodents Tokudaia south Germany. Acta Palaeontologica Polonica 48:617–648. osimensis and Diplothrix legata (Rodentia, Mammalia). Molecular ZWICKL, D. J. 2006. Genetic algorithm approaches for the phylogenetic Phylogenetics and Evolution 15:15–24. analysis of large biological sequence datasets under the maximum SWOFFORD, D. L. 2003. PAUP*: phylogenetic analysis using likelihood criterion. Ph.D. dissertation, University of Texas at parsimony (*and other methods). Version 4. Sinauer Associates, Inc., Publishers, Sunderland, Massachusetts. Austin, Austin. ZWICKL, D. J. 2011. Garli configuration settings. https://www.nescent. TAMURA, K., D. PETERSON,N.PETERSON,G.STECHER,M.NEI, AND S. KUMAR. 2011. MEGA5: molecular evolutionary genetics analysis org/wg_garli/GARLI_Configuration_Settings. Accessed 30 August using maximum likelihood, evolutionary distance, and maximum 2013. parsimony methods. Molecular Biology and Evolution 28:2731– 2739. Submitted 31 May 2013. Accepted 3 February 2014. TANABE, A. S. 2011. Kakusan4 and Aminosan: two programs for comparing nonpartitioned, proportional and separate models for Associate Editor was Sharon A. Jansa.