Mammal Study 33: 19–24 (2008) © the Mammalogical Society of

Genetic variations of the Paguma larvata, inferred from mitochondrial cytochrome b sequences

Ryuichi Masuda1,*, Yayoi Kaneko2, Boripat Siriaroonrat3, Vellayan Subramaniam4 and Masaharu Hamachi5

1 Department of Genome Dynamics, Creative Research Initiative “Sousei”, University, Sapporo 060-0810, Japan 2 Yamazaki College of Health Technology, 4-7-2 Minami-osawa, Hachiouji 192-0046, Japan 3 Research and Conservation Division, Zoological Park Organization, Dusit, Bangkok 10300, Thailand 4 Zoo Negara Malaysia, 68000 Ampang, Selangor, Malaysia 5 Fukuoka Zoo, Fukuoka 810-0037, Japan

Abstract. It is still unclear whether the masked palm civet (Paguma larvata) is native to the Japanese islands or introduced from the outside via human activities. In the present study, we sequenced the whole region (1,140 base-pairs) of the mitochondrial cytochrome b gene for 29 masked palm civets from Japan and Southeast Asia, and investigated their molecular phylogeography. Nine haplotypes were identified from the . Five halpotypes identified from 24 animals of Japan were clustered and separated from four haplotypes from five animals of Southeast Asia, showing clear differentiation between Japanese and Southeast Asian lineages. Sequence differences within Japanese haplotypes were smaller than those within Southeast Asian haplotypes and those between Japanese and Southeast Asian haplotypes. Within Japanese animals, all haplotypes found in eastern (Kanto district) were different from those of central Honshu (Chubu district). The present study highlighted the problem whether the Japanese Paguma larvata is an introduced species showing multiple original routes, or whether it is a native species genetically differentiated from Southeast Asian populations and even within Japan.

Key words: cytochrome b, genetic variation, masked palm civet, Paguma larvata.

The masked palm civet (Paguma larvata) is included in Torii (1993) reported no intraspecific variations of the family and distributed widely in East, karyotypes and morphology within Japanese individuals Southeast, and West Asia (Lekagul and McNeely 1988; of Paguma larvata, and they presumed that introduction Corbet and Hill 1992). This animal is an only viverrid of this species through multiple routes resulted in the species currently living on the main islands of Japan current irregular distribution in Japan. In addition, (Abe et al. 2005). Since 1943 when their occurrence was Harada and Torii (1993) reported the morphological first recorded from Shizuoka Prefecture (Nawa 1965), difference of the Y chromosome of the Japanese animals, Paguma larvata has been found from various localities compared with that of the Chinese animals reported by of Japan (Kuroda 1955; Imaizumi 1960; Obara 1961) Wang et al. (1984), and showed occurrence of geographic and is currently distributed in eastern Honshu and variations of the Y chromosome. Despite these previous Shikoku (Torii 1996; Abe et al. 2005). However, at studies, there has not been enough data to conclude present there are no fossil nor archaeological records of whether Paguma larvata is native to the Japanese islands this species from the Japanese islands. In order to or introduced from the outside. In order to understand uncover the origins of the Japanese populations, some the origin and status of Paguma larvata of Japan, it is researchers have investigated their ecological character- essential to further investigate their phylogeographical istics (Torii 1986; Torii and Miyake 1986). Harada and and zoogeographical characteristics, compared with

*To whom correspondence should be addressed. E-mail: [email protected] 20 Study 33 (2008) native populations living outside of Japan. In the present study, we report the findings from sequencing of the whole region (1,140 base-pairs, bp) of the mitochondrial DNA (mtDNA) cytochrome b gene for Paguma larvata from several areas of Japan and South- east Asia, and then discuss their genetic variations and the phylogenetic relationships.

Materials and methods

Samples and DNA extraction Muscle or hair samples were obtained from 24 indi- viduals collected from five localities (13 animals from , one from Saitama, four from Ibaraki, four from Gifu, and two from Aichi Prefectures) of Japan and five individuals from Southeast Asia (four from Thailand and one from the Malay Peninsula: the precise locality in these countries was unknown) (Fig. 1). Muscle tissues of Japanese animals were collected from roadkill animals or pest control, and preserved in 70–99% ethanol at room temperature. Hair roots were obtained from living ani- mals during the course of ecological survey in Japan, and Fig. 1. Sampling localities of the masked palm civet Paguma larvata in Japan and haplotypes JA1, JA2, JA3, JA4 and JA5 iden- from zoo animals of Thailand (Duzit Zoo and Khao tified in the present study. Numerals in parentheses show numbers of Khew Open Zoo) and Malaysia (Zoo Negara Malaysia), animals having those haplotypes. and preserved at room temperature. Total DNA was extracted from muscle tissues (about 2 × 2 × 2 mm) using the DNeasy Tissue kit (Qiagen) and from hair sam- (Felsenstain 1985) and for the unrooted phylogenetic tree ples (about 30 hair roots per individual) using the DNA construction of haplotypes due to the neighbor-joining Micro kit (Qiagen), both in accordance with manufac- method (Saitou and Nei 1987). A parsimony network of turer’s instructions. The DNA extracts were subjected to haplotypes were constructed using the computer program polymerase chain reaction (PCR) analysis described TCS ver. 1.21 (Clement et al. 2000). One only sequence below. (accession no. AF125151: Veron and Heard 2000) of the whole region of the Paguma larvata cytochrome b PCR, nucleotide sequencing and data analysis available in DDBJ/Genebank/EMBL databases was also The PCR amplification of the whole region (1,140 bp) included for analysis, although the nucleotide site 1,113 of the cytochrome b gene and sequencing were per- of this sequence was reported to be “Y (C or T)” and its formed using the method same as Tamada et al. (2005). sampling locality was not available in the reference. In the present study, four additional sequencing primers labeled at the 5' end with the Texas Red were newly Results and discussion designed as follows: PLCBIN-1A (5'-TTTCAGAGAC- ATGAAACATTGG-3'); PCBIN-2 (5'-TAGCAATCAT- Variations of cytochrome b sequences identified from CCCACTACTAC-3'); PLCBIN-3 (5'-TCTGACTCAG- Paguma larvata ACAAAATCCC-3'); and PLCB-R1 (5'-GGCAAATAT- For the present study, we identified five haplotypes GGGTTACTGATG-3'). Sequence alignment was done (JA1, JA2, JA3, JA4 and JA5) of the cytochrome b from using GeneWorks (Intelligenetics). Molecular evolu- Japanese animals, and four haplotypes (SE1, SE2, SE3 tionary analysis was carried out using MEGA version 3.1 and SE4) were from Southeast Asian animals (Table 1). (Kumar et al. 2004) for calculations such as nucleotide The sequence alignment showed that all of the nucleotide compositions, sequence differences, genetic distances of substitutions were transitions (Table 1), except that the Kimura’s (1980) two-parameter model, bootstrap values sequence (accession no. AF125151: Veron and Heard Masuda et al., Genetic variations of Paguma larvata 21

Table 1. Nucleotide sequences of cytochrome b of the masked palm civet Paguma larvata

Nucleotide site no. Accession Haplotype no.* 46 123 178 179 252 280 285 297 312 315 333 342 408 417 471 519 591 741 749 840 891 947 1113 JA1 GCATTTTACTGCCCACGTTTGTT AB303951 JA2 ••••••••TC•••••••••••••AB303952 JA3 ••••C••••••••••••••••••AB303953 JA4 A••••••••••••••••••••••AB303954 JA5 ••••••••••••••••A••••••AB303955 SE1 •T••••••••••••GT•••••••AB303956 SE2 •T••••••••••••GT•••••C•AB303957 SE3 •T•C•••••••T••GT••C••••AB303958 SE4 •T•••••••••••••T••••A••AB303959 AF125151•TG• •CCC• •A•AA•T•C•A• •Y Haplotypes JA1, JA2, JA3, JA4, JA5, SE1, SE2, SE3 and SE4 were identified in the present study. The sequence (accession no. AF125151; Veron and Heard 2000) has Y (C or T) at site 1113. Dots indicate identities with nucleotides of JA1. * Accession numbers will appear in the DDBJ nucleotide database.

Table 2. Pairwise comparisons of cytochrome b haplotypes of Paguma larvata

Haplotype JA1 JA2 JA3 JA4 JA5 SE1 SE2 SE3 SE4 AF125151 JA1 2111346311 JA20.18333568513 JA30.090.2622457412 JA40.090.260.182457412 JA50.090.260.180.18457412 SE1 0.26 0.44 0.35 0.35 0.35 1 3 2 10 SE2 0.35 0.53 0.44 0.44 0.44 0.09 4 3 11 SE3 0.53 0.70 0.61 0.61 0.61 0.26 0.35 5 13 SE4 0.26 0.44 0.35 0.35 0.35 0.18 0.26 0.44 10 AF125151 0.97 1.14 1.05 1.05 1.05 0.88 0.97 1.14 0.88 Percentage differences are shown below the digonal and numbers of different nucleotides are indicated above diagonal.

2000) available in the DNA databases had some trans- The sequence differences varied 0.09–0.26% (0.18% on versions to all haplotypes (JA1, JA2, JA3, JA4, JA5, average) within the Japanese haplotypes, 0.09–0.44% SE1, SE2, SE3 and SE4) identified in the present study. (0.26% on average) within the Southeast Asian haplo- All transitions among haplotypes occurred at 13 sites types and 0.26–0.70% (0.34% on average) between the (Table 1). When AF125151 was included, five transi- Japanese and Southeast Asian haplotypes (Table 2). tional and four transversional sites were observed, in These values of sequence differences within obtained addition to the 13 polymorphic sites (Table 1). The aver- haplotypes of Paguma larvata show a degree similar to age relative nucleotide frequencies among the obtained intraspecific differences of the cytochrome b gene of sequences (except AF125151) was A = 30.8%, G = other small carnivore species in Japan: less than 1.58% 12.2%, C = 26.7% and T = 30.3%. These values were in the Japanese (Martes melampus) and less than similar to those of the Owston’s palm civet (Chrotogale 0.35% in the sable (Martes zibellina) (Kurose et al. owstoni) (Veron et al. 2004) and Asiatic viverrid species 1999), and 0.49% on average in the Japanese (Veron and Heard 2000). ( meles) (Kurose et al. 2001). In an island popula- Table 2 shows that genetic variations of Japanese ani- tion (Tsushima Islands) of the domestic cat ( catus), mals are smaller than those of Southeast Asian animals. it was reported to be 0.09–0.96% (0.69% on average) 22 Mammal Study 33 (2008)

sequence AF125151 is not available in the reference (Veron and Heard 2000). The nine cytochrome b sequences obtained in the present study will appear in the DDBJ (DNA Data Bank of Japan) nucleotide database with the following acces- sion numbers: AB303951-AB303959 (Table 1).

Phylogenetic relationships among haplotypes The unrooted neighbor-joining tree (Fig. 2) and parsi- mony network (Fig. 3) show the same phylogenetic rela- tionships among haplotypes identified in the present Fig. 2. Neighbor-joining relationships among cytochrome b haplo- study: five halpotypes JA1, JA2, JA3, JA4 and JA5 iden- types (1,140 bp) of Paguma larvata. Haplotypes JA1–JA5 were iden- tified from 24 Japanese animals were phylogenetically tified from Japanese animals, and SE1, SE2, SE3 and SE4 were from clustered and separated from another cluster of four Southeast Asian animals. The sequence (accession no. AF125151; Veron and Heard 2000) was included in the analysis, while its original haplotypes SE1, SE2, SE3 and SE4 identified from locality is not available in the reference. Numbers of animals having five animals of Southeast Asia. The separation of the those haplotypes and the sampling localities are shown in parentheses. Japanese cluster was highly supported with an 86% The scale bar indicates Kimura’s (1980) two parameter distances. Numbers near internal branches show more than 50% bootstrap values bootstrap value (Fig. 2). derived from 1,000 replications. In the parsimony network (Fig. 3), the Japanese and Southeast Asian clusters were separated by one pre- sumed haplotype. Within the Japanese cluster, haplotype JA1 was positioned at the center of the networks and had the highest frequency (66.7%, 16/24 animals) among the Japanese haplotypes (Fig. 3). Haplotype JA1 was shared by 16 animals of eastern Japan (Kanto district): 12 indi- viduals of Tokyo, one of Saitama Prefecture, and three of Ibaraki Prefecture (Figs. 1 and 2). The other haplotypes JA2, JA3, JA4 and JA5 were likely derivatives from JA1 showing one or two nucleotide substitutions, and they were found from the Kanto district and central Honshu (Chubu district) (Fig. 1). Their frequencies were as fol- lows: JA2 (4.2%) from one individual of Tokyo, JA3 (4.2%) from one of Ibaraki Prefecture, JA4 (20.8%) from three of Gifu Prefecture and two of Aichi Prefectures, Fig. 3. A parsimony network among cytochrome b haplotypes of Paguma larvata identified in the present study. One bar between cir- and JA5 (4.2%) from one of Gifu Prefecture. cles means a nucleotide substitution. Open circles show haplotypes Within the Southeast Asian cluster, haplotype SE1 identified from Japan, and closed circles indicate those identified from identified from two animals (one of Thailand and one of Southeast Asia. Small open circles without haplotype numbers mean the Malay Peninsula, Fig. 2) showed the highest fre- presumed haplotypes. Numerals in parentheses show numbers of animals having those haplotypes. quency (2/5 animals) and was positioned at the center of the parsimony network (Fig. 3). The haplotypes SE2 and SE3 were likely derivatives from SE1, showing one or (Tamada et al. 2005). In addition, sequence differences three nucleotide substitutions (Fig. 3). Each of SE2, SE3 of the Owston’s palm civet (Chrotogale owstoni) of and SE4 was identified from one animal of Thailand (1/5 Vietnam were 0–2% (mean = 0.83%) and those of the animals) (Fig. 2). (Hemigalus derbyanus) of Malaysia were 0.8–1.2% (mean = 1%) (Veron et al. 2004). When Inference of origin of Japanese populations and including the sequence AF125151, it is suggested that possibility of introduction to Japan the genetic variations of Paguma larvata is larger than It is still unclear whether Paguma larvata is a native other small carnivores, but the original locality of the species to Japan or an introduced from the outside via Masuda et al., Genetic variations of Paguma larvata 23 human activities. The past records of occurrence and the Thus, the present study highlighted the two different present distribution on the Japanese islands are not geo- directions on the natural history and origins of the Japa- graphically continuous: Paguma larvata currently occurs nese populations of Paguma larvata. on Shikoku and eastern Honshu, but not on Kyushu and To further clarify molecular and morphological phylo- Hokkaido (Torii 1996; Abe et al. 2005). In addition, geography through the habitats of natural populations there are no fossil nor archaeological data regarding the provides an insight to better understanding of the ori- previous existence of the species on the Japanese islands. gin and status of the Japanese populations of Paguma From these facts, Paguma larvata is thought to have larvara. Moreover, accumulation of genetic data from been introduced from the outside of Japan (Torii 1996) the comprehensive areas contributes not only to resolu- via human activities such as the release of pets. Compar- tion of zoogeographical problems, but also to conser- ative phylogeography between the Japanese and other vation of ecosystems native to the Japanese islands. Asian populations could uncover the natural history of Paguma larvata. The mtDNA phylogenetic relationshps Acknowledgements: We thank S. Dakemoto, M. between the Japanese and Southeast Asian animals found Hasegawa (Toyohashi Museum of Natural History), O. in the present study show two possibilities on their his- Nakamura (Saitama Museum of Natural History), V. tory and origins. One is that animals of Thailand and the Arsaithamkul (Dusit Zoo), S. Kamolnorranath and W. Malay Peninsula are not direct origins of the Japanese Tunwattana (Khao Kheow Open Zoo) for supplying populations even if they are introduced animals to Japan. specimens. Many thanks go to E. Kanda (Tokyo Wild- The other is that it is still hard to deny that Paguma life Research Center) and T. Fujii (Hiroshima Prefectural larvata is a native species to Japan showing genetic Government) for their assistance of hair sampling during differentiations. Paguma larvata is distributed widely the course of ecological survey in field. We are grateful in Asia: East, Southeast, and West Asia (Lekagul and to T. Oshida (Obihiro University of Agriculture and McNeely 1988; Corbet and Hill 1992). In order to fur- Veterinary Medicine) and H. Torii (Nara Educational ther investigate the origins of the Japanese populations, University) for invaluable suggestions. We also thank S. it is necessary to clarify molecular phylogeography of Ida and A. Yoshikawa (Hokkaido University) for techni- Paguma larvata through wider habitats including cal assistance. This study was supported by Grants-in- Taiwan and the Chinese mainland. Corbet and Hill Aid for Scientific Research and by the 21st Century (1992) reported geographical variations of morphology Center of Excellence (COE) Program “Neo-Science of of this species. Precise characterization of morphology Natural History” at Hokkaido University financed from on the Japanese populations also brings invaluable infor- the Ministry of Education, Culture, Sports, Science and mation for understanding their origins. Technology, Japan. Within the Japanese populations, three haplotypes JA1, JA2 and JA3 were found from eastern Honshu References (Kanto district) and two haplotypes JA4 and JA5 were from central Honshu (Chubu district) (Fig. 1). Abe, H., Ishii, N., Itoo, T., Kaneko, K., Maeda, K., Miura, S. and Moreover, there were no haplotypes found commonly Yoneda, M. 2005. A Guide of the of Japan: Revised Edition. Tokai University Press, Kanagawa, 206 pp. between the Kanto and Chubu districts. The geographic Clement, M., Posada, D. and Crandall, K. A. 2000. TCS: A computer differences within the Japanese populations suggest that, program to estimate gene genealogies. Molecular Ecology 9: if Paguma larvata is an introduced species to Japan, the 1657–1659. origin of the Kanto population is different from that of Corbet, G. B. and Hill, J. E. 1992. The Mammals of the Indomalayan Region: A Systematic Review. Natural History Museum Publica- the Chubu district and that there were multiple routes of tions/Oxford University Press, Oxford, 488 pp. their introduction. In addition, the highest frequency Felsenstain, J. 1985. Confidence limits on phylogenies: An approach (66.7%) of JA1 and its position at the center of the parsi- using the bootstrap. Evolution 39: 783–791. Harada, M. and Torii, H. 1993. Karyological study of the masked mony networks (Fig. 3) may show founder effects of palm civet Paguma larvata in Japan (Viverridae). Journal of the introduced animals in eastern Honshu. Meanwhile, if Mammalogical Society of Japan 18: 39–42. this species is native to Japan, our findings suggest that Imaizumi, Y. 1960. Coloured Illustrations of the Mammals of Japan. the local populations have been geographically isolated Hoikusha, Tokyo, 196 pp. (in Japanese). Kimura, M. 1980. A simple method for estimating evolutionary rate within Japan, resulting in genetic differentiations similar of base substitutions through comparative studies of nucleotide to other Japanese native carnivores mentioned above. sequences. Journal of Molecular Evolution 16: 111–120. 24 Mammal Study 33 (2008)

Kumar, S., Tamura, K. and Nei, M. 2004. MEGA3: Integrated soft- domestic cats Felis catus of the Tsushima Islands, based on ware for Molecular Evolutionary Genetics Analysis and sequence mitochondrial DNA cytochrome b and control region nucleotide alignment. Briefings in Bioinformatics 5: 150–163. sequences. Zoological Science 22: 627–633. Kuroda, N. 1955. The present status of the introduced mammals in Torii, H. 1986. Food habitats of the masked palm civet, Paguma Japan. Journal of the Mammalogical Society of Japan 1: 13–18. larvata Hamilton-Smith. Journal of the Mammalogical Society Kurose, N., Masuda, R., Siriaroonrat, B. and Yoshida, M. C. 1999. of Japan 11: 39–43. Intraspecific variation of mitochondrial cytochrome b gene Torii, H. 1996. The masked palm civet Paguma larvata. In (K. Izawa, sequences of the Japanese marten Martes melampus and the sable T. Kasuya and T. Kawamichi, eds.) The Encyclopaedia of Ani- Martes zibellina (, , Mammalia) in Japan. mals in Japan. Pp. 136–137. Heibonsha, Tokyo (in Japanese). Zoological Science 16: 693–700. Torii, H. and Miyake, T. 1986. Litter size and sex ratio of the masked Kurose, N., Kaneko, Y., Abramov, A. V., Siriaroonrat, B. and palm civet, Paguma larvata, in Japan. Journal of the Mammalog- Masuda, R. 2001. Low genetic diversity in Japanese populations ical Society of Japan 11: 35–38. of the Eurasian badger Meles meles (Mustelidae, Carnivora) Veron, G. and Heard, S. 2000. Molecular systematics of the Asiatic revealed by mitochondrial cytochrome b gene sequences. Zoo- Viverridae (Carnivora) inferred from mitochondrial cytochrome logical Science 18: 1145–1151. b sequence analysis. Journal of Zoological Systematics and Lekagul, B. and McNeely, J. A. 1988. Mammals of Thailand, 2nd ed. Evolutionary Research 38: 209–217. Darnsutha Press, Bangkok, 758 pp. Veron, G., Heard Rosenthal, S., Long, B. and Roberton, S. 2004. The Nawa, A. 1965. Notes on Paguma larvata in Shizuoka Prefecture. molecular systematics and conservation of an endangered carni- Journal of the Mammalogical Society of Japan 7: 99–105 (in vore, the Owston’s palm civet Chrotogale owstoni (Thomas, Japanese). 1912) (Carnivora, Viverridae, ). Animal Conserva- Obara, I. 1961. A new locality of Paguma larvata. Journal of the tion 7: 107–112. Mammalogical Society of Japan 2: 29–30 (in Japanese). Wang, Z., Quan, G., Yie, Z. and Wang, S. 1984. Karyotypes of three Saitou, N. and Nei, M. 1987. The neighbor-joining method: A new species of Carnivora. Acta Zoologica Sinica 30: 188–195. method for reconstructing phylogenetic trees. Molecular Biology and Evolution 4: 406–425. Received 24 May 2007. Accepted 20 July 2007. Tamada, T., Kurose, N. and Masuda, R. 2005. Genetic diversity in