Origin and Identity of Fejervarya (Anura: Dicroglossidae) on Guam

Origin and Identity of Fejervarya (Anura: Dicroglossidae) on Guam By Elijah Wostl* Eric N. Smith and Robert N. Reed Abstract We used morphological and molecular data to infer the identity and origin of frogs in the genus Fejervarya that have been introduced to the island of Guam. Mensural and meristic data were collected from 96 specimens from throughout their range on the island and a principal component analysis (PCA) was used to investigate the distribution of these data in morphological space. We also amplified a fragment of the 16S rrna mitochondrial gene from twelve of these specimens and compared it to 63 published sequences of Fejervarya and the morphologically similar Zakerana. All examined Fejervarya from Guam are morphologically indistinguishable and share an identical haplotype. The molecular data identify them as Fejervarya cancrivora with a haplotype identical to F. cancrivora from Taiwan. *Corresponding Author E-mail: ewostl@uta.edu Introduction Pacific Science, vol. 70, no. 2 December 16, 2015 (Early view)

Guam is an oceanic island separated from the nearest large land mass by approximately 1800 km and consequently lacks native amphibians. However, it has acquired an anuran fauna through the introduction of non-native species. Currently eight species have been identified as established on the island: Litoria fallax (Peters 1880), Rhinella marina (Linnaeus 1758), Hylarana guentheri (Boulenger 1882), Microhyla pulchra ( Hallowell 1861), Polypedates braueri (Vogt 1911), Eleutherodactylus planirostris (Cope 1862), Fejervarya cancrivora (Gravenhorst 1829) and Fejervarya cf. limnocharis (Gravenhorst 1829) (Christy et al. 2007, Zug 2013). Two of these species have been present for several decades. Rhinella marina was brought to the island in 1937 (in Eldridge 1988) and Litoria fallax has been present since 1968 (Falanruw 1976). The remaining six species were first detected between 2003 and 2005 (Christy et al. 2007). Since initial detection, all species have greatly expanded their range on Guam and appear to be thriving. Christy et al. (2007) reported the presence of both F. cancrivora and F. cf. limnocharis on Guam. However, these species are poorly delimited and each has been found to be composed of several morphologically similar or cryptic species (Chen et al. 2005, Djong et al. 2007, Djong et al. 2011, Hasan et al. 2012, Islam et al. 2008, Kotaki et al. 2008, Kotaki et al. 2010, Kuromoto et al. 2007, Kurniawan et al. 2012, Sumida et al. 2007, Toda et al. 1998, and Veith et al. 2001). In this paper, we use morphology and a fragment of the 16S rrna mitochondrial gene to infer the number, identity, and origin of the Fejervarya species present on Guam. Knowledge of the specific identity and origin of an invasive species is critical for developing interdiction measures, sheds light on the pathways utilized by invasive species, and can highlight gaps in the standing procedures designed to prevent the introduction of non-native species. 2

Materials and Methods Taxa Sampling Between 18 June 2011 and 04 February 2012, we collected 96 specimens of Fejervarya (57 males, 39 females) at 12 localities across their distribution on Guam. Specimens were collected, sacrificed by immersion in a chlorobutanol solution, photographed post mortem (venter and dorsum) and fixed using 10 percent formalin (3.7% formaldehyde). Once fixed, specimens were transferred to 70% ethanol for preservation. Prior to fixation and preservation, a piece of liver was stored in 100 percent ethanol for future molecular work. All specimens and tissue samples were deposited at the Smithsonian National Museum of Natural History (USNM 580816-580911). Morphology We collected mensural data on 15 morphological traits from adult specimens: snout vent length (SVL), head length (HL), mid-head width (MHW), posterior head width (PHW), inter-narial distance (IND), eye nares (EN), eye snout (ES), tympanum eye (TE), tympanum width (TW), forelimb length (FL), thigh length (TL), tibia length (TiL), tarsus length (TarL), hind-foot length (HFL), and inner metatarsal tubercle length (IMT). We also collected meristic data on an additional five traits: the extent of webbing on the hind feet (following Savage and Heyer 1975 as modified in Myers and Duellman 1982 and coding to the nearest ½ phalanx), the total number of vomerine teeth, the presence or absence of a fringe of skin along the outer metatarsal and fifth toe, the presence or absence of a dorsal stripe, and the throat pattern of males. We used the total number of vomerine teeth because of intra-specimen variation between left and right sides. Sex and maturity were determined by making an incision through the abdominal wall and examining the gonads and state of the seminal duct or ovaries. Females were considered mature if they contained eggs or if the oviducts were enlarged. Males were 3

considered mature if the seminal duct was enlarged and convoluted. See Table 1 for a definition of the morphological traits. <<Table 1 near here>> As an initial step to assess the number of Fejervarya species on Guam, we performed a principal component analysis (PCA) on specimens with a full morphological data set (n = 95), reasoning that if multiple species of Fejervarya are present on Guam, then each may form a distinct cluster in morphological space. The mensural data were log 10 transformed and regressed against the log 10 transformed SVL. This procedure reduces the influence on body size on morphometric traits (Reist 1986). The resulting residuals were used along with vomerine tooth count in the PCA. Males and females were analyzed separately to avoid the effects of sexual dimorphism. In order to evaluate the number of principal components with statisctically significant eigenvalues (alpha=0.05) we used an eigenvalue Monte Carlo simulation (parallel analysis) as outlined in O Connor (2000). The number of vomerine teeth encompassed a broad range. Therefore, we tested this character for normality and tested the smallest and largest counts as outliers. The throat pattern of males and state of the metatarsal fringe were invariant and were not included in morphological analyses. Webbing formulas have long been used to aid in the identification of anurans; however, they do not lend themselves to statistical evaluation. We compared the webbing formulas derived from the Guam specimens to those for F. cancrivora and F. limnocharis in the literature (Dubois and Ohler 2000, Inger 1966, Manthey and Grossmann 1997). We mapped the presence or absence of a dorsal stripe onto the PCA plot to visualize its distribution in morphological space. This character has been used to distinguish F. limnocharis from F. cancrivora (Inger and Steubing 2005), and may have led to the allocation of Guam specimens to F. limnocharis. 4

Molecular Analysis We used the morphological data to identify a subset of individuals from each sex for use in the molecular analysis. This subset included individuals from the extremes of each axis for each principal component, individuals from the center of each morphological cluster, and individuals that showed divergent variation for non-mensural characters (i.e foot webbing). Twenty-seven of the 95 individuals used in the morphometric analysis were used in the molecular analysis, including the two specimens tentatively identified as F. limnocharis. DNA was extracted from liver tissue using a serapure magnetic bead suspension (Rohland and Reich 2012) by transferring a small piece of the original tissue sample to 1000 µl of cell lysis buffer and agitating it at 57 with 20µl of Proteinase K until it was completely digested. Fifty microliters of the resulting solution was combined with 50µl of water in a microcentrifuge tube and mixed with 180 µl of serapure (1.8:1 serapure:sample). The remaining steps of DNA extraction followed the procedures for cleaning PCR products with AMPure magnetic beads (Agencourt, Bioscience, Beverly, Massachusetts, USA). A fragment of the mitochondrial 16S rrna gene was amplified via polymerase chain reaction (PCR) using the forward primer 16S AR (5 -CGC CTG TTT ATC AAA AAC AT-3 ) and the reverse primer 16S BR (5 -CCG CTC TGA ACT CAG ATC ACG-3 ). The thermal cycling profile consisted of an initial denaturation at 95 C followed by five cycles of a 30 s, 95 C denaturation, a 30 s 50 C annealing and a 2 min 72 C extension followed by 40 similar cycles with the annealing temperature raised to 54 C and then a final 10 min extension at 72 C. Amplification products were cleaned using serapure magnetic beads following the same procedure as the extraction. Cycle sequencing reactions (in both primer directions) and Sanger sequencing were performed using standard protocols associated with BigDye terminator 5

chemistry (Applied Biosystems, Foster City, California, USA) at the UTA genomics core facility (Arlington, Texas, USA; gcf.uta.edu). Alignment and Phylogenetic Analysis Raw sequence chromatograms were assembled and edited using the program Sequencher 4.8 (Gene Codes Corporation Ann Arbor, MI). The resulting sequences were submitted to Genbank under accession numbers KR816718 KR816729 and KT972720 KT972734. To identify the species present, and their origin, we compared the sequences from Guam specimens to sequences for representatives of all Fejervarya species and the morphologically similar Zakerana available on Genbank. We aligned sequences using Clustal W in MEGA 6.0 (Tamura et al. 2013) and constructed a phylogenetic tree from uncorrected pairwise distances using the unweighted pairwise method with arithmetic mean (UPGMA) as implemented in MEGA 6.0. Once we identified the species most closely related to our specimens from Guam, we downloaded all available sequences for that species, aligned them with our sequences, and constructed a second distance tree. In most instances, pairwise distance is inappropriate for phylogenetic inference due to the inability to account for multiple mutations at a site or variable rates of molecular evolution between branches. However, we are interested in the origin of an introduced population that was established within the last 15 years. A direct comparison of raw pairwise distances is the most appropriate method to infer this relationship. Results Morphological data All examined specimens possessed a fringe of skin along the fifth toe and metatarsal and the throat pattern of all reproductively active males consisted of two dark lateral blotches separated medially by the mottled grey ground color of the throat. Other characters were 6

variable. The SVL of adults ranged from 48 77 mm in males and 58 103 mm in females. Several male (USNM 580879, SVL 67.58 mm; USNM 580839, SVL 59.33 mm), and female (USNM 580858, SVL 58.11 mm; USNM 580859, SVL 60.53; USNM 580876, SVL 65.16; USNM 580910, SVL 61.0 mm) specimens fall within this range yet did not exhibit evidence of sexual maturity. We ran the PCA both with and without these specimens. The results were largely the same and so we present the PCA in which they were included. One of these specimens, a male (USNM 580879), was also included in the molecular analysis. The webbing formula of the hind foot was variable and can be summarized as I(0 0.5) (1)II(0 1) (0 1)III(0 1) (0.5 2)IV(0 2) (0 0.5)V and is closer to that reported for F. cancrivora than F. limnocharis (Dubois and Ohler 2000, Inger 1966, Manthey and Grossmann 1997). The total number of vomerine teeth varied considerably, ranging from eight to 24. There was no significant difference between males and females (p = 0.075, df = 92), and when combined the values were normally distributed (Shapiro-Wilk test, p = 0.067, df =94). No outliers were detected using the outlier labeling rule following Hoaglin and Iglewicz (1987). Among males, the first three principal components had an eigenvalues >1.0 and accounted for 50.4%, 10.0%, 8.6% of the observed variation respectively. Only the first component had a significant eigenvalue (respective eigenvalues for the first three components 7.5, 1.5 and 1.3). Similarly, in females the first three components also had eigenvalues >1.0 which respectively account for 50.4%, 11.1% and 8.3% of the observed variation. Like males only the first component had a significant eigenvalue (7.6, 1.7, and 1.3 respectively). For each sex, scatter plots of the individual loading scores along the axis of each components with eigenvalues >1.0 failed to reveal distinct morphological clusters (Figure 1). In both males and females the mensural variables all strongly loaded onto the first principal component and constitute a latent size variable. The second component largely reflects variation in the number of 7

vomerine teeth. Table 2 presents the loading scores of each variable on each component. The presence or absence of a dorsal stripe was distributed throughout the morphological cluster revealed by the PCA. <<Figure 1 near here>>, <<Table 2 near here>> Molecular data All sequenced specimens of Fejervarya from Guam shared an identical haplotype, suggesting a single species and origin. For this reason, we included just two of the sequences in further analyses. Compared to other Fejervarya species for which sequences are available, the specimens from Guam are most similar to F. cancrivora. Comparison of the sequences from Guam specimens to all available sequences of F. cancrivora revealed an identical haplotype with specimens from southern Taiwan (Figure 2). Three sequences identified as F. cancrivora on Genbank (accession numbers EU 435283, EU 365386, and EU 365385), all from Taiwan, were similar to our F. limnocharis outgroup and divergent from other F. cancrivora. These sequences probably represent misidentifications and were excluded from the second comparison. <<Figure 2 near here>> Discussion All Fejervarya from Guam, while phenotypically variable, are morphologically indistinguishable. Those examined at the molecular level share an identical haplotype that is identical to F. cancrivora from Taiwan and similar to that of specimens from Java, the type locality for the species. Unfortunately, the Taiwan and Java material are not linked to curated specimens, raising the possibility that the associated identifications are incorrect. Given the shallow divergence of the clade that contains these sequences and number of other samples within this clade identified as F. cancrivora, we feel that misidentification is unlikely in this case. 8

Of note is that two other species of frog now occurring on Guam (Hylarana guentheri and Polypedates braueri) are also native to Taiwan, and a third (Microhyla pulchra) is found on adjacent mainland China. Additionally, we found no evidence of the occurrence of specimens belonging to the F. limnocharis complex on the island. While we are confident of our identification of the species of Fejervarya on Guam and its origin, the poor quality of the data associated with the Genbank sequences represents an unanticipated complication that prevented absolute surety. Of the 63 Fejervarya and Zakerana sequences downloaded, only 14 are associated with a specimen. We must rely on the identification of the other 49 specimens provided by the authors. As illustrated by the three putative F. cancrivora that were excluded from the second analysis because of their molecular affinity with F. limnocharis, we know that these identifications can be unreliable. Additionally, 15 Genbank sequences are not associated with a locality, either because they are part of unpublished work or because the authors did not include locality data in the publication. Our experience on Guam causes us to infer that this species is a very efficient colonizer. This frog has spread rapidly across the island, achieves high local population density, thrives in degraded and anthropogenically altered habitats, is tolerant of a broad range of environmental conditions (we found individuals on karst islets separated from the mainland by several meters of ocean) and is a voracious and indiscriminant predator. Prey items observed by the first author on Guam included other Fejervarya, the blindsnake Ramphotyphlops braminus, the skink Carlia ailanpalai, and a large centipede (Scolopendra sp.). The large and growing population of Fejervarya cancrivora on Guam, the cargo and transportation hub of the West Pacific, greatly increases the risk of this species being transported to, and becoming established on, other islands in the region. 9

Acknowledgments We thank G. R. Zug for providing the impetus for this project and his continual support and encouragement. We would also like to thank S. Gotte and A. Wynn from the National Museum of Natural History for curating and loaning the specimens and tissues used in this study. For their valuable insight and suggestions, we thank J. M. Meik, K. J. Shaney, W. Schargel, and D. Sanchez. For her work in the field and continuous support, EW would like to thank R. Wostl. All specimens used in this work were collected and handled under an IACUC issued by the USGS Fort Collins Science Center. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government. 10

Table 1 Description of Morphological Traits. Character Snout-Vent Length (SVL) Head Length (HL) Head Width Mid (MHW) Head Width Posterior (PHW) Inter-narial Distance (IND) Eye to Nares (EN) Eye to Snout (ES) Tympanum to Eye (TE) Tympanic Width (TW) Forelimb Length (FL) Thigh Length (TL) Tibia Length (TiL) Tarsus Length (TarL) Hind foot Length (HFL) Inner Metatarsal Tubercle Length (IMT) Description tip of snout to cloaca with specimen flat Strait line distance from tip of snout to posterior margin of mandible transverse distance between outer edge of lips at posterior margin of the eyes width of head at posterior margin of the tympanum Distance between medial margin of the nares Anterior margin of the orbit to posterior margin of the nares Tip of snout to anterior corner of orbit Anterior edge of the tympanum to the posterior corner of the eye Horizontal width of tympanic membrane Straight line distance from bent elbow to base of outer palmer tubercle with forearm parallel to body Straight-line distance from vent to distal margin of flexed knee Straight-line distance from outer margin of bent knee to outer margin of flexed tibio-tarsal articulation Straight line distance from margin of flexed tibio-tarsal articulation to base of inner metatarsal tubercle Straight line distance from base of inner metatarsal tubercle to tip of fourth toe Length of inner metatarsal tubercle 11

Table 2 loading scores of each variable on each principal component with an eigenvalue greater than one.. Males Females Character PC 1 PC 2 PC 3 PC 1 PC 2 PC 3 Vomerine Teeth 0.17 0.57 0.59 0.20 0.80 0.35 Head Length 0.75-0.38-0.08 0.84 0.04-0.11 Mid Head Width 0.75-0.45-0.07 0.87 0.18-0.30 Posterior Head Width 0.78-0.24-0.02 0.86 0.07-0.31 Inter-narial Distance 0.69-0.2 0.46 0.70 0.31 0.42 Eye Nares 0.84-0.23-0.05 0.85 0.19 0.05 Eye Snout 0.7-0.24-0.17 0.84-0.20-0.14 Tympanum Eye 0.83 0.26 0.09 0.81-0.03 0.25 Tympanum Width 0.57-0.38 0.49 0.44 0.18 0.44 Forearm Length 0.77 0.33 0.25 0.66-0.19-0.16 Thigh Length 0.61 0.36-0.08 0.72 0.18-0.17 Crus Length 0.81 0.1-0.21 0.80-0.23-0.18 Tarsus Length 0.76 0.35-0.05 0.82-0.05-0.03 Hindfoot Length 0.63 0.28-0.52-0.38-0.64 0.33 Inner metatarsal tubercle 0.73 0.07-0.06 0.39-0.51 0.53 12

Figure 1. Plots of the first two principal components of male (A) and female (B) Fejervarya sp. from Guam. Filled symbols represent individuals used in the molecular analysis. Squares indicate the presence of a dorsal stripe. 13

Figure 2: Pairwise distance tree of Fejervarya from Guam and all available sequences of Fejervarya cancrivora. Note the shared haplotype between Guam and southern Taiwan. Sequences are identified by Genbank accession number when vouchered specimens were not available. Branch tips marked with an asterisk represent a haplotype belonging to several individuals. The distribution of the haplotype is indicated in the label. 14

Literature Cited Boulenger, G. A. 1882. Catalogue of the Batrachia Salientia s. Ecaudata in the Collection of the British Museum. Second Edition. London: Taylor and Francis. Chen, L., R. W. Murphy, A. Lathrop, A. Ngo, N. L. Orlov, C. T. Ho, and I. L. M. Somorjai. 2005. Taxonomic chaos in Asian ranid frogs. An initial phylogenetic resolution. Herpetol. J.15:231 243. Christy, M. T., C. S. Clark, D. E. I. Gee, D. Vice, D. S. Vice, M. P. Warner, C. L. Tyrrell, G. H. Rodda, and J. A. Savidge. 2007. Recent records of alien anurans on the Pacific island of Guam. Pac. Sci. 61:469 484. Cope, E. D. 1862. On some new and little known American Anura. Proc. Acad. of Nat. Sci. Philadelphia 14: 151 159 Djong, T. H., M. M. Islam, M. Nishioka, M. Matsui, H. Ota, M. Kuramota, M. M. R. Khan, M. S. Alam, D. S. Anslem, W. Khonsue, and M. Sumida. 2007. Genetic relationships and reproductive-isolation mechanisms among the Fejervarya limnocharis complex from Indonesia (Java) and other Asian countries. Zool. Sci. 24:360 375. Djong, T. H., M. Matsui, M. Kuramoto, M. Nishioka, and M. Sumida. 2011. A new species of the Fejervarya limnocharis complex from Japan (Anura, Dicroglossidae). Zool. Sci. 28:922 929. 15

Dubois, A. and A. Ohler. 2000. Systematics of Fejervarya limnocharis (Gravenhorst, 1829) (Amphibia, Anura, Ranidae) and related species. 1. Nomenclatural status and type-specimens of the nominal species Rana limnocharis Gravenhorst, 1829. Alytes 18:15 50 Eldredge, L. G. 1988. Case studies of the impacts of introduced animal species on renewable resources in the U.S.-Affiliated Pacific Islands. Pp. 118-146 in B. D. Smith, ed. Topic Reviews in Insular Development and Management in the Pacific U. S.-Affiliated Islands. University of Guam Marine Laboratory. Falanruw, M. C. 1976. Life on Guam: Savanna, Old Fields, Roadsides. Department of Education, Guam. 79 pp. Gravenhorst, J. L. C. 1829. Deliciae Musei Zoologici Vratislaviensis. Fasciculus primus. Chelonios et Batrachia. Leipzig: Leopold Voss Hasan, M., M. M. Islam, M. M. R. Khan, M. S. Alam, A. Kurabayashi, T. Igawa, M. Kuramota, and M. Sumida. 2012. Cryptic anuran biodiversity in Bangladesh revealed by mitochondrial 16S rrna gene sequences. Zool. Sci. 29:162 172. Hallowell, E. 1861 "1860". Report upon the Reptilia of the North Pacific Exploring Expedition, under command of Capt. John Rogers, U.S. N. Proceedings of the Academy of Natural Sciences of Philadelphia 12: 480 510 16

Hoaglin, D.C. and B. Iglewicz. 1987. Fine-tuning some resistant rules for outlier labeling. J. Am. Stat. Assoc. 82: 1147 1149. Inger, R.F. 1966. The systematics and zoogeography of the amphibia of Borneo. Fieldiana: Zoology 52:1 402 Inger, R. F., and R. B. Stuebing. 2005. A field guide to the frogs of Borneo. Natural History Publications, Kota Kinabalu. 424 pp. Islam, M. M., N. Kurose, M. M. R. Khan, T. Nishizawa, M. Kuramoto, M. S. Alam, M. Hasan, N. Kurniawan, M. Nishioka, and M. Sumida. 2008. Genetic divergence and reproductive isolation in the genus Fejervarya (Amphibia: Anura) from Bangladesh inferred from morphological observations, crossing experiments, and molecular analysis. Zool. Sci. 25:1084 1105. Kotaki, M., A. Kurabayashi, M. Matsui, W. Khonsue, T. H. Djong, M. Tandon, and M. Sumida. 2008. Genetic divergences and phylogenetic relationships among the Fejervarya limnocharis complex in Thailand and neighboring countries revealed by mitochondrial and nuclear genes. Zool. Sci. 25:381 390. Kotaki, M., A. Kurabayashi, M. Matsui, M. Kuramota, T. H. Djong, and M. Sumida. 2010. Molecular phylogeny of the diversified frogs of the genus Fejervarya (Anura: Dicroglossidae). Zool. Sci. 27:386 395. 17

Kuraishi, N., M. Matsui, H. Ota, and S.-L. Chen. 2011. Specific separation of Polypedates braueri (Vogt, 1911) from P. megacephalus (Hallowell, 1861)(Amphibia: Anura: Rhacophoridae). Zootaxa 2744:53 61. Kuramota, M., S. H. Joshy, A. Kurabayashi, and M. Sumida. 2007. The genus Fejervarya (Anura: Ranidae) in Central Western Ghats, India, with description of four new cryptic species. Curr. Herpetol. 26:81 105. Kurniawan, N., M. M. Islam, T. H. Djong, T. Igawa, M. B. Daicus, H. S. Yong, R. Wanichanon, M. M. R. Khan, D. T. Iskandar, M. Nishioka, and M. Sumida. 2010. Genetic divergence and evolutionary relationship in Fejervarya cancrivora from Indonesia and other Asian countries inferred from allozyme and mtdna sequence analysis. Zool. Sci. 27:222 233. Linnaeus, C. 1758. Systema Naturae per Regna Tria Naturae, Secundum Classes, Ordines, Genera, Species, cum Characteribus, Differentiis, Synonymis, Locis. 10th Edition. Volume 1. Stockholm, Sweden: L. Salvii. Manthey, U., and W. Grossmann. (1997). Amphibien & Reptilien Sudostasiens. Berlin: Natur und Tier-Verlag. 512 pp. Myers, C. W., and W. E. Duellman. 1982. A new species of Hyla from Cerro Colorado, and other tree frog records and geographical notes from Western Panama. Am. Mus. Novit. 2752:1 32. 18

O Conner. B. P. 2000. SPSS and SAS programs for determining the number of components using parallel analysis and Velicer's MAP test. Behav.Res.Meth Ins C 32, 396-402. Peters, W. C. H. 1880. Über neue oder weniger bekannte Amphibien des Berliner Zoologischen Museums (Leposoma dispar, Monopeltis (Practogonus) jugularis, Typhlops depressus, Leptocalamus trilineatus, Xenodon punctatus, Elapomorphus erythronotus, Hylomantis fallax). Monatsberichte der Königlichen Preussische Akademie des Wissenschaften zu Berlin 1880: 217 224 Reist, J.D. 1986. An empirical evaluation of coefficients used in residual and allometric adjustment of size covariation. Can. J. Zool. 64:1363 1368. Rohland, N., and D. Reich. 2012. Cost-effective, high-throughput DNA sequencing libraries for multiplexed target capture. Genome Res. 22:939 946. Savage, J. M., and W. R. Heyer. 1967. Variation and distribution in the tree-frog genus Phyllomedusa in Costa Rica, Central America. Beitrage zur Neotropischen Fauna 5:111 131. Sumida, M., M. Kotaki, M. M. Islam, T. H. Djong, T. Igawa, Y. Kondo, D. S. Anslem, W. Khonsue, and M. Nishioka. 2007. Evolutionary relationships and reproductive isolating mechanisms in the rice frog (Fejervarya limnocharis) species complex from Sri Lanka, Thailand, Taiwan and Japan, inferred from mtdna gene sequences, allozymes, and crossing experiments. Zool.Sci. 26:547 562. 19

Tamura, K., G. Stecher, D. Peterson, A. Filipski, and S. Kumar. 2013. MEGA6: Molecular Evolutionary Genetic Analysis version 6.0. Mol. Biol. and Evol. 30:2725 2729. Toda, M., M. Matsui, M. Nishida, and H. Ota. 1998. Genetic divergence among southeast and east Asian populations of Rana limnocharis (Amphibia: Anura), with special reference to sympatric cryptic species in Java. Zool. Sci. 15:607 613. Veith, M., J. Kosuch, A. Ohler, and A. Dubois. 2001. Systematics of Fejervarya limnocharis (Gravenhorst, 1829)(Amphibia, Anura, Ranidae) and related species 2. morphological and molecular variation in frogs from the Greater Sunda Islands (Sumatra, Java, Borneo) with the definition of two species. Alytes 19:5 28. Vogt, T. 1911. Beitrag zur Amphibien-fauna der Insel Formosa. Sitzungsberichte der Gesellschaft Naturforschender Freunde zu Berlin 1911: 179 184 Zug, G. R. 2013. Amphibians and reptiles of the Pacific islands: A comprehensive guide. University of California Press, Berkeley and Los Angeles. 20