Zavreliella inawaheia Sasa, Kitami & Suzuki,

© 2017 Academia Sinica, Taiwan Open Access Species Identification of Shed Snake Skins in Taiwan and Adjacent Islands Tein-Shun Tsai1,* and Jean-Jay Mao2 1Department of Biological Science and Technology, National Pingtung University of Science and Technology 1 Shuefu Road, Neipu, Pingtung 912, Taiwan 2Department of Forestry and Natural Resources, National Ilan University No.1, Sec. 1, Shennong Rd., Yilan City, Yilan County 260, Taiwan. E-mail: [email protected] (Received 28 August 2017; Accepted 25 November 2017; Published 19 December 2017; Communicated by Jian-Nan Liu) Tein-Shun Tsai and Jean-Jay Mao (2017) Shed snake skins have many applications for humans and other animals, and can provide much useful information to a field survey. When properly prepared and identified, a shed snake skin can be used as an important voucher; the morphological descriptions of the shed skins may be critical for taxonomic research, as well as studies of snake ecology and conservation. However, few convenient/ expeditious methods or techniques to identify shed snake skins in specific areas have been developed. In this study, we collected and examined a total of 1,260 shed skin samples - including 322 samples from neonates/ juveniles and 938 from subadults/adults - from 53 snake species in Taiwan and adjacent islands, and developed the first guide to identify them. To the naked eye or from scanned images, the sheds of almost all species could be identified if most of the shed was collected. The key features that aided in identification included the patterns on the sheds and scale morphology. Ontogenetic differences and intraspecific variation in the patterns of sheds were evident in some snake species, and the proportion of young snakes with patterned shed skins was larger than that of adults. The retention of markings on the ventral side of the body (especially the ventral head) during sloughing was much lower than that on the dorsal side. We hope that this pioneering work will not only encourage other researchers to develop similar keys for their country, but also promote local schools, organizations, and citizen scientists to conduct snake inventories. Key words: Ecdysis, Snake monitoring technique, Ontogenetic differences, Intraspecific variations, Guide and key. *Correspondence: E-mail: [email protected]; [email protected] BACKGROUND Molting (shedding, ecdysis) in reptiles results from cyclical changes in the underlying skin structure. Snakes periodically molt their outermost layer of epidermis, permitting the replacement of skin that has been abraded or damaged, the disposal of parasites, and growth (Greene 1997). The process of ecdysis is therefore viewed to signify a renewing of vital forces or as a sign of immortality (Crump 2015). Shed snake skins have several practical applications. First, in folk and traditional medicine, snake sheds have been used for treatment of ailments like glaucoma, eczema, hemorrhoids, wound healing, psoriasis, and parturition problems (Lev 2003; Mukherjee et al. 2013). Second, shed skins can serve as a tissue - collected without harming the animal - for studies of contaminant exposure (Kaur 1988; Hopkins et al. 2001; Jones et al. 2005) or DNA (Bricker et al. 1996; Clark 1998). DNA in shed skins is stable for at least one month as long as the shed is kept dry (Fetzner 1999). Third, ethical considerations make using human skin a major problem in transdermal research (Ngawhirunpat et al. 2006), but shed snake skin may offer a model membrane because Zoological Studies 56: 38 (2017) doi:10.6620/ZS.2017.56-38 1 © 2017 Academia Sinica, Taiwan of its partial similarities to human skin (Itoh et al. 1990; Priprem et al. 2007 2008; KumpugdeeVollrath et al. 2013; Torri et al. 2014). Shed skins can be used in exhibits and employed in educational programs at places such as snake farms, museums, or nature centers, where they may engender less fear than a live snake. Snake sheds can also be very aesthetically pleasing; many people have taken to creating beautiful snake shed jewelry (https://is.gd/ZBeJwB). In addition, skin lipid pheromones have been shown to play important roles in snake behavior. Shed skin releases pheromones into the environment and informs conspecifics about the reproductive status of the newly shed snake (Kubie et al. 1978). Other animals, such as ground squirrels, exploit the scent of rattlesnake (Crotalus sp.) sheds to reduce the risk of rattlesnake predation (Clucas et al. 2008a b), and some birds use snake sheds in their nests, which may function to decrease predation by mammalian predators (Medlin and Risch 2006) or reveal female parental quality (Trnka and Prokop 2011). Wildlife detector dogs can also be trained to find hidden snake sheds during surveys of wild snakes (Stevenson et al. 2010). In addition, a shed snake skin can provide much useful information to a field survey. When properly prepared and identified, a shed snake skin can be used as a voucher, documenting the presence of a species without sacrificing or even necessarily finding an individual (Gray 2009 2012). In some species, if the shed skin is complete, the sex of the snake that left it may be inferred by counting the ventral and/or subcaudal scales. The location where a shed skin is found can provide insight regarding species habitat preferences (Gray 2009 2012). During ecdysis, a snake’s stratum corneum is stretched, resulting in a shed skin that is usually slightly longer (on average 11% increase reported in Gray 2009) than the actual snake. Using a regression equation, the actual snout-vent length (SVL) can be estimated from the length of a shed skin (Gray 2009). The relative size, shape, and arrangement of scales from sheds, as well as the sex and size measurements of the actual snake, could potentially be used to identify individuals and aid in investigations of site fidelity during ecdysis (Gray 2008). Studies on the global decline of snakes and other reptiles (Gibbons et al. 2000; Reading et al. 2010) have revealed that it is necessary to monitor wild populations in the long term, which must be aided by the establishment of standard methods and techniques (McDiarmid et al. 2012). In snake monitoring or surveying, snake sheds are normally ignored or discarded, but they can be a useful voucher for a specimen if the methods or technique to identify them are well developed. Because keratin is slow to break down, sheds may persist in an arid environment for well over a month, assuming they are not consumed by fungi or invertebrates. Marks made by scale clippings are observable in cast skins, so sheds can also be of value in mark-recapture studies (Gray 2002). Researchers could also sequence DNA from shed skins to use in population genetics studies. Several publications or guides address identifying actual snakes, yet these same works may not be useful in identifying sheds because many rely on coloration, which sheds lack. Few convenient/fast methods or techniques to identify species of snake sheds in specific areas, however, have been developed, except in Canada (Gray 2012) and some areas in the United States (ex. Gray 2002 2015). In this study, we collected and examined shed skins from 53 species of snakes found in Taiwan and adjacent islands in order to develop the first guide and key to identify them; we hope that this will facilitate related studies on the biology and conservation of snakes. MATERIALS AND METHODS We collected shed skin samples of snakes from Taiwan (for most species), Green Island (for Laticauda spp. and Oligodon formosanus), Orchid Island (for Elaphe carinata yonaguniensis, Laticauda spp., O. formosanus, and Psammodynastes pulverulentus papenfussi), Kinmen (for Python bivittatus bivittatus), and Mazu (for O. chinensis). Sheds were collected in the field or obtained from individual snakes in captivity. The study was approved (IACUC approval number: NPUST-104-029) by the Animal Care and Use Committee of the National Pingtung University of Science and Technology (NPUST) and the captive animals were mainly raised in the Reptile and Amphibian Facility at NPUST, which has been accredited by the International Association for the Assessment and Accreditation of Laboratory Animal Care (AAALAC). The sheds were processed and prepared before identification. The sheds were first immersed in clean water and lightly squeezed to soften them. They were then cut using dissecting scissors underneath the water from the angle of the mouth to the center of the tenth ventral; once this point was reached the page 2 of 23Zoological Studies 56: 38 (2017) © 2017 Academia Sinica, Taiwan cutting continued down the middle of the ventrals in the direction of the tail, until the tenth ventral anterior to the anal plate. The sheds were then cut at an angle towards the boundary line between the ventrals and the dorsals, and cutting continued along this line towards the tail tip. During the above procedure, any debris on the sheds was removed by hand or with forceps, and the sheds were simultaneously and carefully spread out. After that, the sheds were carefully spread out under the water onto mosquito nets, and the nets with sheds were lifted out of the water and placed on shelves to dry. After the sheds dried, they were separated carefully from the net, and scanned (Epson Perfection V37, Seiko Epson Corp., Japan) to obtain images for identification. The images were displayed using the auto-contrast function in Adobe Photoshop 7.0.1. Finally, the individually prepared shed skin samples were sealed in transparent bags, which were put into A4 size ring binders and stored in dry boxes. The shed samples were deposited as voucher specimens in 1) Department of Biological Science and Technology, NPUST; and 2) Department of Forestry and Natural Resources, National Ilan University. The scientific names of snakes in this study followed the Reptile Database (http://www.reptile-database.org), except for the undescribed endemic species (“Paiwan Keelback”; Hebius sp.) found in southern Taiwan and the Formosan Tiger Snake (Rhabdophis formosanus; Takeuchi et al. 2012 2014). The validity of subspecies in Trimeresurus stejnegeri (Guo et al. 2016), Hebius sauteri, and Sinonatrix percarinata (Zhao 2006) was questionable, and subspecies names for these species were not used in this study. RESULTS A total of 1,260 shed samples (one sample per individual, including 322 samples from neonates/juveniles and 938 samples from subadults/adults) from 53 species of snakes (including one undescribed species, Hebius sp.) were collected and examined in this study. Two samples of O. chinensis were collected from Hangzhou, China. Sheds were collected in the field (10%) or obtained from individual snakes in captivity (90%). Representative images of dorsal head, anterior body (about two head lengths posterior to the head), mid-body, and posterior body (about two head lengths anterior to the vent) of shed skins from each species are shown in figures 1-8. At least one complete set of shed images for each species was available, except for O. chinensis, which lacked the head region in all four shed samples. The ranges in the number of dorsal scale rows, ventrals, subcaudals, and other scale characters for each species were mainly summarized from previous reports (Maki 1931; Pope 1935; Wang and Wang 1956; Kuntz 1963; Ota and Toyama 1989; Tu et al. 1990; Zhao et al. 1998; Ota et al. 1999; Zhao 2006; You et al. 2013 2015). The sample sizes of sheds for each snake species are listed in table 1, where the summarized ranges of dorsal scale row, numbers of ventrals and subcaudals are also shown. The catalogue numbers of vouchers are listed in appendix table A1. We used “type I” and “type II” to designate different pattern forms (defined below) when interindividual or ontogenetic variations in shed skins existed within a species. The key features (except for those summarized above) of a complete shed skin for each species are described as follows. Achalinus formosanus formosanus Whole body without markings (Fig. 1A); dorsal scales keeled (except for the outermost row); anal scale entire and subcaudals not divided; the parietal contacted posterolaterally by an enlarged paraparietal scale; the last supralabial very slender. Achalinus niger Whole body without markings (Fig. 1B); dorsal scales at least on anterior part of body smooth; anal scale entire and subcaudals not divided; the parietal contacted posterolaterally by an enlarged paraparietal scale; the last supralabial very slender. Amphiesma stolatum Two light-colored stripes extending along the dorsal body (or at least the posterior part of body) and occasionally crossed by blackish bars anteriorly (Fig. 1C); dorsal body scales keeled (except for the outermost row); anal scale divided. Boiga kraepelini Dorsal body with (Fig. 1D) or without (Fig. 1E) cross bands; body scales smooth; dorsolateral body scales strongly oblique, at least on anterior part of body; temporals small and similar to nuchal page 3 of 23Zoological Studies 56: 38 (2017) © 2017 Academia Sinica, Taiwan Table 1. The sample sizes of neonate/juvenile (N/J) and adult (A) sheds, ranges of number of dorsal scale rows (at anterior-middle-posterior body), ventrals, and subcaudals, and geographic distribution for 53 snake species. See the text for detailed descriptions of the classification of body patterns Species Body patterns N/J ANo. of dorsal scale rowsaNo. of ventralsa Pairs of subcaudalsa Occurrence in Taiwan or adjacent islandsb Achalinus formosanus formosanus Not-patterned 0 1 27(29)-27(25)-25 158-184 61-83 (single row) Mountain areas of Taiwan, at altitudes of 1000-2000 m Achalinus niger Not-patterned 0 4 25-25-25 169-185 51-72 (single row) Mountain areas of Taiwan, at altitudes of 1000-2000 m Amphiesma stolatum Patterned 2 27 19-19-17 142-165 41-87 Throughout Taiwan, Matsu, and Orchid Is., up to 500 m altitude Boiga kraepelini Patterned 0 6 23(21,25)-21(19,23)-17(15) 212-250 115-158 Throughout Taiwan, up to 1500 m altitude Not-patterned 0 4 Bungarus multicinctus multicinctus Patterned 13 11 15(16,17)-15(16)-15 198-250 26-65 (single row) Throughout Taiwan, Penghu, Kinmen, Matsu, Xiaoliuqiu, and Gueishan Is., up to 1000 m Not-patterned 051 Calamaria pavimentata pavimentata Not-patterned 0 2 13-13-13 167-192 13-23 Throughout Taiwan and Orchid Is., up to 1500 m altitude Cyclophiops major Patterned 14 0 15-15-15 155-189 61-97 Throughout Taiwan, up to 1000 m altitude Not-patterned 0 18 Daboia siamensis Patterned 4 57 29(27,31)-29(27,31,33)-21(23) 151-169 40-54 Southern and south-eastern Taiwan, up to 500 m altitude Deinagkistrodon acutus Patterned 1 8 21(22,23)-21(23)-17(18,19) 152-176 41-63 Throughout Taiwan, at altitudes of 1501500 m Elaphe carinata yonaguniensis Patterned-type I 12 0 23(21,25)-23(21,25)-19(17) 209-230 78-99 Throughout Taiwan and adjacent islands, up to 2000 m altitude Patterned-type II 1 7 Elaphe taeniurus friesi Patterned-type I 90 17 25(23)-23(21,25)-19(17) 243-262 101-122 Throughout Taiwan, up to 2000 m altitude Patterned-type II 0 102 Euprepiophis mandarinus Patterned-type I 011 23(21,25)-23(21)-19(17,21) 181-238 49-76 Northern, central and eastern Taiwan, at altitudes of 1000-2500 m Patterned-type II 0 3 Gonyosoma frenatum Patterned 3 0 19(21)-19(17)-15(13) 200-227 108-151 Southern and southeastern Taiwan, up to 1500 meters altitude Not-patterned 0 2 Hebius miyajimae Patterned 0 2 19-19-17 141-152 87-92 Mainly in northern Taiwan, at altitudes of 500-1600 m Hebius sauteri Not-patterned 1 4 17-17-17 120-147 65-92 Throughout Taiwan, up to 1500 m altitude Hebius sp. (“Paiwan Keelback”) Not-patterned 1 0 19-19-17(18) 156 95 Limited areas of southern or eastern Taiwan, between 500 and 1300 meters altitude Hypsiscopus plumbea Not-patterned 3 8 19-19-17(15) 122-136 23-47 Western and northern Taiwan, including Gueishan Is., up to 500 m altitude Indotyphlops braminus Not-patterned 0 8 20-20-20 300-303 8-14 Throughout Taiwan and adjacent islands, up to 500 m altitude Laticauda colubrina Patterned 0 4 25-23,25-21 225-245 33-44 Orchid Is., Green Is., eastern or southern coast of Taiwan Laticauda laticaudata Patterned 0 8 19-19(21)-17 215-252 33-47 Orchid Is., Green Is., eastern or southern coast of Taiwan Laticauda semifasciata Patterned 0 10 23-23-19(21) 178-203 32-43 Orchid Is., Green Is., eastern or southern coast of Taiwan Lycodon rufozonatus rufozonatus Patterned 512 17(19,21)-17(19)-15(17) 184-225 45-95 Throughout Taiwan and Matsu, up to 2000 m altitude Not-patterned 0 34 Lycodon ruhstrati ruhstrati Patterned 38 7 17(19)-17-15 193-233 64-116 Throughout Taiwan, up to 1500 m altitude Not-patterned 0 7 Macropisthodon rudis rudis Patterned 3 3 23-23(21,25)-19 123-156 37-65 Throughout Taiwan, at altitudes of 5001500 m Not-patterned 0 2 Myrrophis chinensis Type I 2 0 23(25)-23-21(17,19) 131-155 35-52 Limited areas of Taipei, Taoyuan, Nantou, Kenting, and Kinmen, at altitudes up to 500 m Type II 0 9 Naja atra Type I 54 0 23(21,25,27)-21(19,20)-15(14,13) 158-185 38-53 Throughout Taiwan and Matsu, up to 1000 m altitude Type II 0 138 page 4 of 23Zoological Studies 56: 38 (2017) © 2017 Academia Sinica, Taiwan Species Body patterns N/J ANo. of dorsal scale rowsaNo. of ventralsa Pairs of subcaudalsa Occurrence in Taiwan or adjacent islandsb Oligodon chinensis Patterned-type I 2 0 17-17-15 158-206 40-73 Matsu Patterned-type II 0 2 Oligodon formosanus Patterned-type I 7 0 19-19(17)-17(15) 154-189 39-60 Throughout Taiwan and adjacent islands, up to 1000 m altitude Patterned-type II 2 19 Oligodon ornatus Patterned 0 3 15-15-15 156-182 27-44 Northern and central Taiwan, at altitudes of 500-1500 m Oreocryptophis porphyraceus kawakamii Patterned-type I 2 0 19(18,17)-19(17)-17(15) 194-214 47-75 Throughout Taiwan, up to 2200 m altitude Patterned-type II 0 7 Ovophis makazayazaya Patterned 0 7 25~31-25~29-21 144-155 39-54 Northern and western Taiwan, at altitudes of 500-2200 m Not-patterned 0 3 Pareas atayal Not-patterned 011 15-15-15 174-188 71-79 Northern Taiwan, up to 1500 m altitudes Pareas formosensis Not-patterned 0 10 15-15-15 170-180 69-82 Throughout Taiwan except for the northeastern tip of the island, up to 2000 m altitudes Pareas komaii Not-patterned 0 7 15-15-15 162-182 60-76 Central, southern, and eastern Taiwan, up to 2000 m altitudes Plagiopholis styani Patterned 0 5 15-15-15 102-126 21-32 Very limited areas around Taipei and the Northern Cross-Island Highway, at altitudes of 500-1100 m Protobothrops mucrosquamatus Patterned 3 37 25-29-25(21,27,29)-17-21 194-233 70-108 Throughout Taiwan, up to 1500 m altitude Psammodynastes pulverulentus papenfussi Patterned 0 11 17-17-15(13) 161-177 58-79 Throughout Taiwan and Orchid Is., up to 1500 m altitude Pseudoxenodon stejnegeri stejnegeri Patterned 0 2 19(17,21)-17-15(14) 150-162 42-66 Throughout Taiwan, at altitudes of 8002500 m Not-patterned 0 3 Ptyas dhumnades Patterned-type I 2 0 16-16(14)-14 186-217 91-188 Throughout Taiwan, up to 2000 m altitude Patterned-type II 0 3 Ptyas korros Patterned 2 0 15-15(13,14)-11 156-184 100-154 Throughout Taiwan, up to 1000 m altitude Not-patterned 0 2 Ptyas mucosa Patterned-type I 25 019(17,21)-17(16)-14(13,15) 170-213 94-143 Throughout Taiwan, Kinmen, and Matsu, up to 1500 m altitude Patterned-type II 0 43 Python bivittatus bivittatus Patterned 1 18 53-64-64-72-40-44 255-263 63-71 Kinmen Rhabdophis formosanus Patterned 1 15 19-19-17(15) 161-171 74-89 Mountain areas of Taiwan, at altitudes of 1500-3000 m Rhabdophis swinhonis Patterned 1 20 15(17)-15-15 124-165 49-74 Throughout Taiwan, up to 1800 m altitude Sibynophis chinensis chinensis Patterned 0 4 17-17-17 164-187 171-208 Throughout Taiwan and Gueishan Is., up to 1500 m altitude Sinomicrurus hatori Patterned 0 3 13-13-13 (occasionally 15) 223-245 30-34 Northern and eastern Taiwan, up to 2200 m altitude Sinomicrurus macclellandi swinhoei Patterned 0 1 13-13-13 207-240 32-41 Northern and western Taiwan, at altitudes up to 1000 m Sinomicrurus sauteri Patterned 0 2 13-13-13 (occasionally 15) 234-268 27-36 South and central Taiwan, up to 2500 m altitude Sinonatrix annularis Patterned 10 0 19-19-17 136-167 47-74 Very limited areas of Taipei, and Taoyuan, at altitudes up to 1000 m Not-patterned 0 9 Sinonatrix percarinata Patterned 4 3 19-19-17 131-160 44-87 Throughout Taiwan, at altitudes up to 1500 m Not-patterned 0 2 Trimeresurus gracilis Patterned 7 33 25~27-21~19-15~17 144-149 43-53 Central mountains in Taiwan, at elevations of 2000-3500 m Not-patterned 05 Trimeresurus stejnegeri Not-patterned 452 21(22~25)-21(23)-15 154-178 43-80 Throughout Taiwan, Green Is. and Orchid Is., up to 2000 m altitude Xenochrophis piscator Not-patterned 2 9 19-19(17)-17(15) 121-152 42-90 Throughout Taiwan and Kinmen, up to 500 m altitude aThe numbers of dorsal scale rows, ventrals, and subcaudals are mainly summarized from previous reports (see the text). The data of Hebius sp. are measured from a fluid-preserved specimen (voucher number: NIU-HE-001). (Maki 1931; Pope 1935; Wang and Wang 1956; Kuntz 1963; Ota and Toyama 1989; Tu et al. 1990; Zhao et al. 1998; Ota et al. 1999; Zhao 2006; You et al. 2013, 2015). bThe geographical distributions of snakes in Taiwan and adjacent islands are mainly according to Shang et al. (2009). Table 1. (continued) page 5 of 23Zoological Studies 56: 38 (2017) © 2017 Academia Sinica, Taiwan Fig. 1. Representative images of the dorsal head, anterior body (at a position about two head lengths posterior to the head), middle body, and posterior body (at a position about two head lengths anterior to the vent) of potential types of shed skins for the snake species: (A) Achalinus formosanus formosanus (not patterned; adult); (B) A. niger (not patterned; adult); (C) Amphiesma stolatum (patterned; adult); (D) Boiga kraepelini (patterned; adult); (E) B. kraepelini (not patterned; adult); (F) Bungarus multicinctus multicinctus (patterned; juvenile); (G) B. m. multicinctus (not-patterned but tinted with gray; adult); (H) B. m. multicinctus (not patterned; adult); (I) Calamaria pavimentata pavimentata (not patterned; adult); (J) Cyclophiops major (patterned; neonate). Scale bar = 0.5 cm. (A) (C) (E) (G) (I) (B) (D) (F) (H) (J) page 6 of 23Zoological Studies 56: 38 (2017) © 2017 Academia Sinica, Taiwan scales; anal scale divided. Bungarus multicinctus multicinctus Dorsal body of neonates/juveniles and 18% of adults with cross bands (Fig. 1F); body of other adults without obvious markings (Figs. 1G, H), but the scales around the mid-dorsal line on 51% of these adults grayish (Fig. 1G) or grayish green; some sheds with grayish green or even light orange ventrals; body scales smooth; vertebral scales enlarged and hexagonal; anal scale entire and subcaudals not divided. Calamaria pavimentata pavimentata Whole body without markings (or with very faint, light stripes along the dorsal flanks) (Fig. 1I); body scales smooth; parietals in contact with supralabials; internasals absent; anal scale entire. Cyclophiops major Dorsal body with scattered, dark small spots in neonates (Fig. 1J) and without markings (Fig. 2A) in juveniles and adults; dorsal scales of at least anterior part of body smooth; anal scale divided. Daboia siamensis Dorsal body with blotches or designs edged by black and/or light lines (Fig. 2B); dorsal body scales strongly keeled (except for the outermost row); top of head with three large, prominent, dark patches forming a light “Y” between them and covered only with small scales (except for supraoculars or nasals); anal scale entire. Deinagkistrodon acutus Dorsal body with a series of large, triangular designs laterally (Fig. 2C); upper head brownish, with light-colored sides; dorsal body scales strongly keeled (except for the outermost rows); anal scale entire. Elaphe carinata yonaguniensis Nape or anterior body of neonates/juveniles with pairs of dark marks or cross bands (Figs. 2D, E; type I); posterior body of neonates/juveniles sometimes with light brown longitudinal stripes extending onto tail (Figs. 2D, E); the posterior half of the body of adults with longitudinal intermittent lines of darkened scale keels (Fig. 2F; type II); sides of ventrals in adults and some juveniles darkened; dorsal body scales strongly keeled (except for the outermost row); anal scale divided. Elaphe taeniurus friesi Anterior dorsal body with 2 or 4 rows of dark blotches or bands in neonates/juveniles and subadults (Fig. 2G; type I), but with a faint median band in adults (Fig. 2H; type II); the designs turning into two lateral series of dark bands on the posterior body, most obvious on tail; head usually with a stripe extending posteriorly from the eye; dorsal body scales feebly keeled except those on the flanks; anal scale divided. Euprepiophis mandarinus Dorsal head with 2-3 broad, black crossbands; dorsal body with a series of light-centered saddles (Fig. 2I; type I) or a series of curved and paired spots (Fig. 2J; type II); sides of some ventrals usually darkened; body scales smooth (or occasionally feebly keeled); anal scale divided. Gonyosoma frenatum In neonates/juveniles, top of head with dark stripes, upper body may have crossbars (Fig. 3A); in adults, body without markings (Fig. 3B) and head occasionally with a stripe mainly extending posteriorly from the eye; dorsal body scales smooth or feebly keeled; prefrontals in contact with supralabials; anal scale divided. Hebius miyajimae Nape and dorsal body with a pair of lightcolored longitudinal stripes (Fig. 3C); mid-dorsal and/or dorsolateral body scales keeled; anal scale divided. Hebius sauteri Body generally without markings (Fig. 3D); nape occasionally with a light line extending upward and posteriorly from corner of mouth; middorsal and dorsolateral body scales keeled; anal scale divided. Hebius sp. Dorsal body without obvious markings page 7 of 23Zoological Studies 56: 38 (2017) © 2017 Academia Sinica, Taiwan Fig. 2. Representative images of the dorsal head, anterior body, middle body, and posterior body of potential types of shed skins for the snake species: (A) Cyclophiops major (not patterned; adult); (B) Daboia siamensis (patterned; adult); (C) Deinagkistrodon acutus (patterned; adult); (D) Elaphe carinata yonaguniensis (patterned-type I; neonate); (E) E. c. yonaguniensis (patterned-type I; juvenile); (F) E. c. yonaguniensis (patterned-type II; adult); (G) E. taeniurus friesi (patterned-type I; juvenile); (H) E. t. friesi (patterned-type II; adult); (I) Euprepiophis mandarinus (patterned-type I; adult); (J) E. mandarinus (patterned-type II; adult). Scale bar = 0.5 cm. (A) (C) (E) (G) (I) (B) (D) (F) (H) (J) page 8 of 23Zoological Studies 56: 38 (2017) © 2017 Academia Sinica, Taiwan (Fig. 3E) or occasionally with light cross lines anteriorly; nape with a light line extending upward and posteriorly from corner of mouth; mid-dorsal and dorsolateral body scales keeled; anal scale divided. Hypsiscopus plumbea Whole body without markings (Fig. 3F); body scales smooth; head with only one internasal scale; anal scale divided. Indotyphlops braminus Body very small (diameter less than 5 mm) and without designs; tail quite short, ending in a spine; scales all around body similar in size and smooth (Fig. 3G). Laticauda colubrina Body with bands that are narrower than interspaces (Fig. 3H); tail laterally compressed; body scales smooth; upper head distinctly marked; head with 3 prefrontals and no loreals, while the 2nd chin shields not in contact with the infralabials; anal scale divided. Laticauda laticaudata Body with bands of similar width to interspaces (Fig. 3I); tail laterally compressed; body scales smooth; upper head distinctly marked; head with 2 prefrontals and no loreals, while the 2nd chin shields not in contact with the infralabials; anal scale divided. Laticauda semifasciata Body with bands that are broadest at vertebral line, tapering off laterally (Fig. 3J); tail laterally compressed; body scales smooth; ventral scales on posterior half to two-thirds of body keeled; ventral scales on poseterior one-third of body notched in central region of posterior margin; upper head distinctly marked; head with 3 prefrontals and no loreals, while the 2nd chin shields not in contact with the infralabials; anal scale divided. Lycodon rufozonatus rufozonatus Back of upper head may have a lightcolored “Λ” configuration; dorsal body of neonates, juveniles, and 26% of adults with dark or gray cross bands (Fig. 4A), while that of other adults without markings (Fig. 4B); body scales smooth; anal scale entire. Lycodon ruhstrati ruhstrati Back of upper head with or without a light band; dorsal body of neonates, juveniles, and 50% of adults with dark or grey cross bands (Fig. 4C), while that of other adults without markings (Fig. 4D); usually greater than 6 rows of body scales (on the flanks) smooth and others feebly keeled; anal scale entire. Macropisthodon rudis rudis Dorsal body of neonates/juveniles and 60% of adults with blotches or markings, variable in size, shape, and intensity of pigmentation (Fig. 4E), while that of other adults without obvious markings (Fig. 4F); upper head light brown in neonates/ juveniles or dark dirty brown in adults, may have a stripe extending posteriorly from the eye; sides of upper head light-colored; sides of ventrals of adults may be darkened; all dorsal body scales strongly keeled, keels darkened in adults; temporals keeled and spectacle separated from supralabials by suboculars; anal scale divided. Myrrophis chinensis Nape with a vertebral stripe; in neonates/ juveniles, upper body may have scattered dark spots, lateral body with an obvious longitudinal band of light color, involving first to fourth scale rows (Fig. 4G; type I); in adults, upper body may have scattered dark spots (Fig. 4H; type II) or a faint longitudinal band of light color on the lateral body (Fig. 4I; type II); body scales smooth; head with only one internasal scale; anal scale divided. Naja atra Upper body without markings (Fig. 4J) or with irregular or scattered crosslines of light color; nape may have a spectacle-like mark (Figs. 5A, B); upper body of 15% of neonates/juveniles and 96% of adults darkened; ventrals without dark color in neonates/juveniles (Fig. 4J and Fig. 5A; type I), but generally darkened and mottled with white in adults (Fig. 5B; type II); body scales smooth and dorsolateral scales strongly oblique; vertebral scales much smaller than lateral scales; loreals absent and temporals larger than nuchal scales; page 9 of 23Zoological Studies 56: 38 (2017) © 2017 Academia Sinica, Taiwan Fig. 7. Representative images of the dorsal head, anterior body, middle body, and posterior body of potential types of shed skins for the snake species: (A) Ptyas dhumnades (patterned-type II; adult); (B) P. korros (patterned; juvenile; note the faint, light scales arranged in a transverse row on the anterior body); (C) P. korros (not patterned; adult); (D) P. mucosa (patterned-type I; juvenile); (E) P. mucosa (patterned-type II; adult); (F) Python bivittatus bivittatus (patterned; adult); (G) Rhabdophis formosanus (patterned; adult); (H) R. swinhonis (patterned; adult); (I) Sibynophis chinensis chinensis (patterned; adult); (J) Sinomicrurus hatori (patterned; adult). Scale bar = 0.5 cm. (A) (C) (E) (G) (I) (B) (D) (F) (H) (J) page 16 of 23Zoological Studies 56: 38 (2017) © 2017 Academia Sinica, Taiwan behind eyes; loreal scales absent, prefrontals not in contact with supralabials; body scales smooth; anal scale divided. Sinonatrix annularis Dorsal body with crossbands (Fig. 8C) or without markings (Fig. 8D); dorsal body scales feebly keeled in neonates or keeled in juveniles and adults, except those on the flanks (outer 1-2 rows of scales); internasals triangular and pointed anteriorly; usually one labial in contact with the eye; anal scale divided. Sinonatrix percarinata Dorsal body with crossbands (Fig. 8E) or without markings (Fig. 8F); dorsal body scales feebly keeled in neonates or keeled in juveniles and adults, except those on the flanks (outer 1-2 rows of scales); internasals triangular and pointed anteriorly; usually two labials in contact with the eye; anal scale divided. Trimeresurus gracilis Upper body of neonates/juveniles and 87% of adults brownish with a mid-dorsal series of dark blotches with lateral blotches on each side tending to be in line with the mid-dorsal one (Fig. 8G), while that of other adults without obvious markings (Fig. 8H); upper head of those with patterned body generally bearing a dark, thick band extending from eye to corner of mouth; upper dorsal body scales (feebly) keeled, except those on the flanks; upper head covered only with small scales (except for supraoculars) and with fewer than 9 scales located between supraoculars; anal scale entire. Trimeresurus stejnegeri Whole body light-colored and without designs; upper dorsal body scales keeled; upper head covered only with small scales (except for supraoculars) and with 9-15 scales located between supraoculars (Fig. 8I); anal scale entire. Xenochrophis piscator Whole body without obvious markings; dorsal body scales feebly keeled, except those on the flanks (outer 3 rows of scales); internasals triangular and pointed anteriorly (Fig. 8J); usually two labials in contact with the eye; anal scale divided. In addition, for convenient and expeditious identification of complete shed skins of the 53 snake species in Taiwan and adjacent islands, we provide a key in appendix I. DISCUSSION The identification of shed snake skins can augment data collected for species inventories (Gray 2002). Investigations of shed skin patterns could enhance our understanding of the reproductive, behavioral, and physiological ecology of snakes, and help structure conservation strategies (Lillywhite and Sheehy 2016). Both accurate and precise morphological descriptions of shed snake skins are critical not only for taxonomic research, but also for forensic efforts, customs inspections, and the conservation of species (Baker 2006). Although the exact rate of decay for snake sheds in the wild is almost unknown, snake sheds either are quickly consumed by fungi/ animals or become fragmented and faded. In a series of field experiments, Gray (2005) found that 10 cm sections of shed skins lasted an average of 7 days when covered by objects and 11.5 days in the open; one shed skin section lasted up to 46 days. When a shed is incomplete or faded, it will have fewer characters identifiable to the naked eye or from scanned images and the key for shed snake skins in this study may not be applicable. Therefore, to accurately identify snake sheds using the key in this study, one should endeavor to collect complete shed samples as much as possible and soon after ecdysis. The sheds of the 53 species of snakes in this study can be identified using the guide or the key, although it was not easy to differentiate between H. sauteri and Hebius sp., S. hatori and S. sauteri, or even S. percarinata and X. piscator. The differences between S. hatori and S. sauteri are minor even on the actual skins and their external morphology is similar to S. japonicus boettgeri, which is distributed on several islands of the Ryukyu Archipelago, Japan (Ota et al. 1999; Mochida et al. 2015; Kaito et al. 2017); the body coloration of this species probably works as crypsis through background matching or disruptive camouflage rather than aposematism (Mochida et al. 2015). The differences between shed skins of Sinomicrurus snakes in Taiwan and on the Ryukyu Islands could be the presence or absence of a transverse light band on the head, the number of longitudinal stripes on the dorsal body, or the page 17 of 23Zoological Studies 56: 38 (2017) © 2017 Academia Sinica, Taiwan Fig. 8. Representative images of the dorsal head, anterior body, middle body, and posterior body of potential types of shed skins for the snake species: (A) S. macclellandi swinhoei (patterned; adult); (B) S. sauteri (patterned; adult); (C) Sinonatrix annularis (patterned; neonate); (D) S. annularis (not patterned; adult); (E) S. percarinata (patterned; juvenile); (F) S. percarinata (not patterned; adult); (G) Trimeresurus gracilis (patterned; adult); (H) T. gracilis (not patterned; adult); (I) T. stejnegeri (not patterned; adult); (J) Xenochrophis piscator (not patterned; adult). Scale bar = 0.5 cm. (A) (C) (E) (G) (I) (B) (D) (F) (H) (J) page 18 of 23Zoological Studies 56: 38 (2017) © 2017 Academia Sinica, Taiwan number and extent of transverse bands on the dorsal body. In addition, apical pits and an apical notch on a scale (Chiasson 1981; Chiasson and Lowe 1989; Gray 2002 2012 2015) could partly be examined from the scanned images, but examining such characters in neonates or some species (e.g., L. r. rufozonatus, L. r. ruhstrati, or P. mucosa) might be difficult and require the use of a microscope. When shed skins are incomplete and/or their markings have faded, species identification may instead depend on examination of micro-structure morphology on the surface of scales using light microscopy or scanning electron microscopy (Tsai et al. unpublished data). It has been shown that the microdermatoglyphic patterns of the dorsal scales are species-specific (e.g., Price 1982; Wang and Zhou 1998) and the scales of the original animals and sloughs of the same species showed identical microdermatoglyphic characters (Wang and Zhou 1998). Nevertheless, scanning electron microscopy may be expensive and of limited availability for some identifiers. In this study we develop a more convenient identification method, requiring only the naked eye or scanned images, which could easily be used in a snake inventory by a local school or other organization or by citizen scientists (Todd et al. 2017). Ontogenetic differences and intraspecific variation in shed patterns were evident in about 23 snake species in this study. Gray (2002) mentioned that pattern is usually less visible on the sheds of younger snakes. We found that the markings on the shed skins of young snakes tended to be lighter than those on the shed skins of patterned adult individuals of the same species, such as D. siamensis, P. mucrosquamatus, P. b. bivittatus, and R. swinhonis. However, we also found that the proportion of young snakes with patterned shed skins was larger than that of adults, corresponding to ontogenetic loss of pattern in some individuals of species such as B. m. multicinctus, L. r. rufozonatus, L. r. ruhstrati, M. r. rudis, M. chinensis, S. annularis, S. percarinata, and T. gracilis. Ontogenetic differences in the distribution of melanocytes (Krey and Farajallah 2013), the quantity of melanin deposition in the skin tissues, and the growth and development of skin are likely underyling causes of the above trends. In addition, although we did not collect the sheds from young individuals of 22 snake species, this should not reduce the applicability of the key developed in this study. The patterns on the sheds of most young snakes could be inferred when conspecific adults have patterned sheds, because most ontogenetic shifts in snake patterns primarily involve a loss in pattern clarity. Most of the above 22 snake species (except Achalinus spp., I. braminus, C. p. pavimentata, and Pareas spp.) had patterned sheds as adults and should also have patterned sheds in young snakes, while the shed skins from young snakes of Achalinus spp. and I. braminus should not be patterned because these snakes have no patterns on their actual body. Moreover, we did not find distinct geographical variation in the patterns visible on shed skins. It has been shown that O. formosanus from different regions of Taiwan, Green Is. or Orchid Is. may have polymorphic markings or colorations on their body (https://goo.gl/WjsW2H), but a similar striped pattern was visible on the dorsal body of sheds of this species from above sites, with only occasional variation in the markings on the head (Figs. 5E, F). The distinct sexual dimorphism of T. stejnegeri’s actual skin color (Tsai and Tu 1998; Zhao 2006) was not visible in their sheds. Many markings on the actual skins were not visible on the shed skins. In appendix table A2, we summarize the visibility of markings on the actual and shed skin of all 53 species of snakes in this study, separated by body region. We found that most of these snake species have markings on the dorsal side of their actual body (81% of species with markings on the head, 91% on the body, and 91% on the tail); in contrast, markings were visible on the shed skins of only 55% (dorsal head), 75% (dorsal body), and 68% (dorsal tail). Fewer species have markings on ventral side of the actual body (43% on the ventral head, 75% on the ventral body, and 77% on the ventral tail), and even fewer have markings on the ventral side of the shed skins (6% ventral head; 17% ventral body; 19% ventral tail). Thus, 67% of snake species with markings on the actual skin of the dorsal head, 83% with markings on the dorsal body, and 75% with markings on the dorsal tail also had some markings on their shed skins, whereas only 13% (ventral head), 23% (ventral body), and 24% (ventral tail) of snake species with markings on the ventral side also had some markings on the sheds. The degree of retention for the markings on the ventral side of the body (especially on the ventral side of the head) during sloughing was much lower than that on the dorsal side. Pattern retention in shed snake skins is due to deposition of melanin from epidermal melanophores. It can be hypothesized that there is a lack or reduction in epidermal melanophores in the ventral epidermis. The causes of these differences and their potential page 19 of 23Zoological Studies 56: 38 (2017) © 2017 Academia Sinica, Taiwan biological significance, if any, remain to be investigated in future studies. The sheds of some species bear distinct traits that are useful for species identification. For example, 1) the sheds of many viperids (such as D. siamensis, D. acutus, P. mucrosquamatus, and T. gracilis) and sea snakes (such as L. colubrina and L. laticaudata) had deeply darkened, apparent, and unfaded patterns even on adults; 2) the sheds of some species (such as E. t. friesi, O. p. kawakamii, P. p. papenfussi, P. b. bivittatus, S. c. chinensis, and Sinomicrurus spp.) had distinctive markings on the head; 3) the sheds of some species (such as E. mandarinus, O. chinensis, and R. formosanus) had unique markings on the body; 4) the sheds of some species (such as B. kraepelini, B. m. multicinctus, M. r. rudis, N. atra, and Pareas spp.) had special scutellation on the head or body; 5) the sheds of some species (such as E. c. yonaguniensis, E. mandarinus, M. r. rudis, N. atra, P. s. stejnegeri, Ptyas spp., and R. formosanus) sometimes had darkened markings on the sides of ventrals, which could be a good character for species identification. In addition, the geographic location and habitat where sheds are found in the wild could also be used as auxiliary information for species identification (Gray 2012 2015). Mainly according to Shang et al. (2009), we listed the geographical distribution of the 53 species in Taiwan and the adjacent islands in table 1 to help make shed identification faster and more accurate. Also, ventral scales and smooth (or feebly keeled) dorsal scales on the shed skins of most species in this study may produce iridescence when observed at a certain angle through the light, except for those of I. braminus, Laticauda spp., and/or P. b. bivittatus. Therefore, checking for the presence of iridescence (see also Monroe and Monroe 1967; Verveen and Rouwkema 2007) on shed skins might also be considered an applicable method for species identification. CONCLUSIONS A shed snake skin found during a field survey can provide much useful information. Investigations on the use of shed skins for species identification can advance our understanding of the geographic distribution, activity patterns, and reproductive, behavioral, and physiological ecology of snakes, with important implications for conservation. In this study, we present a pioneering guide to the identification of local snake sheds in Taiwan, which may also be useful in other parts of Asia. The sheds of almost all species can be identified to the naked eye or from scanned images if a complete shed is collected. The key features used include the patterns on sheds and scale morphology. Ontogenetic differences and intraspecific variation in the patterns on the sheds were evident in some snake species, and the proportion of young snakes with patterned shed skins was higher than that of adults. The degree of retention for the markings on the ventral side of body (especially the ventral head) during sloughing was much lower than that on the dorsal side. We hope this pioneer work will not only encourage other researchers to develop similar keys for their country, but also promote local schools, organizations, and citizen scientists to conduct snake inventories. Acknowledgments: We thank Bo-Shen Wang, Shih-Hao Wang, Wei-Chieh Hsu, Chao-An Tu, Zuo-Ming Cai, En-Dao Liang, Pi-Hang Chen, ChinHsiu Lin, Kai-Xiang Zhang, Chun-Yu Chen, KaiCing Liou, Qing-Guo Ji, Miaoju Lin, Ya-Chi Yang, Zi-You Fan, Yi-Jie Lee, Meng-Qun Zhou, Chih-Chi Liu, Ruo-Cian Hu, and other colleagues and friends for their great help and assistance in collecting samples and preparing specimens. We thank Chung-Wei You and Dr. Juan Lei for offering shed samples of O. chinensis collected from Mazu and Hangzhou, respectively. We thank Shih-Ping Chou and Dr. Si-Min Lin for offering some shed samples of P. b. bivittatus collected from Kinmen. We also thank Dr. Ming-Chung Tu for assistance with the study. Special thanks go to Dr. Andrew M. Durso for reviewing a previous manuscript of this paper. This work was funded by the National Science Council (NSC 100-2621-B-020-001), Council of Agriculture, and Kaohsiung City Government. Authors’ contributions: TST and JJM collected samples. TST designed the study, wrote the manuscript, prepared the tables and figures, and provided the identification key. Both authors participated in revising the manuscript and approved the final manuscript. Competing interests: The authors have no competing interests to declare. Availability of data and materials: The supporting data will be provided by the corresponding author on request. Consent for publication: Not applicable. page 20 of 23Zoological Studies 56: 38 (2017) © 2017 Academia Sinica, Taiwan Ethics approval consent to participate: Not applicable. REFERENCES Baker BW. 2006. Forensic implications of dorsal scale row counts on puff-faced water snakes (Colubridae: Homalopsinae: Homalopsis buccata). Herp Rev 37:171173. Bricker J, Bushar LM, Reinert HK, Gelbert L. 1996. Purification of high quality DNA from shed skin. Herp Rev 27:133-134. Chiasson RB. 1981. The apical pits of Agkistrodon (Reptilia: Serpentes). J Arizona-Nevada Acad Sci 16:69-73. Chiasson RB, Lowe CH. 1989. Ultrastructural scale patterns in Nerodia and Thamnophis. J Herpetol 23:109-118. Clark A. 1998. Reptile sheds yield high quality DNA. Herp Rev 29:16-18. Clucas B, Owings DH, Rowe MP. 2008a. Donning your enemy’s cloak: Ground squirrels exploit rattlesnake scent to reduce predation risk. Proc R Soc Lond B Biol Sci 275:847-852. Clucas B, Rowe MP, Owings DH, Arrowood PC. 2008b. Snake scent application in ground squirrels, Spermophilus spp.: A novel form of antipredator behaviour? Anim Behav 75:299-307. Crump M. 2015. Eye of Newt and Toe of Frog, Adder’s Fork and Lizard’s Leg. University of Chicago Press, Chicago, Illinois, USA. Fetzner JW Jr. 1999. Extracting high-quality DNA from shed reptile skins: A simplified method. BioTechniques 26:10521054. Gibbons JW, Scott DE, Ryan TJ, Buhlmann KA, Tuberville TD, Metts BS, Greene JL, Mills T, Leiden Y, Poppy S, Winne CT. 2000. The global decline of reptiles, déjà vu amphibians. Bioscience 50:653-666. Gray BS. 2002. A key to the shed skins of northeastern snakes. Bull Chicago Herp Soc 37:121-128. Gray BS. 2005. The Serpent’s Cast: A Guide to the Identification of Shed Skins from Snakes of the Northeast and Mid-Atlantic States. The Center for North American Herpetology Monograph Series Number 1, USA. Gray BS. 2008. A note on site fidelity for ecdysis in the northern brown snake, Storeria dekayi dekayi. Bull Chicago Herp Soc 43:1-3. Gray BS. 2009. Estimating snout to vent length from data acquired from the shed skins of the northern brown snake, Storeria dekayi dekayi. J Kansas Herpetol 32:17-19. Gray BS. 2012. Guide to the Identification of the Shed Skins of the Snakes of Canada, USA. Gray BS. 2015. Shed Snakeskin Identification: A Guide to Snakeskins Found in Pennsylvania. Pennsylvania Amphibian & Reptile Survey, USA. Greene HW. 1997. Snakes: the Evolution of Mystery in Nature. University of California Press, California, USA. Guo P, Liu Q, Chen X, Nguyen T. 2016. Complex longitudinal diversification across South China and Vietnam in Stejneger’s pit viper, Viridovipera stejnegeri (Schmidt, 1925) (Reptilia: Serpentes: Viperidae). Mol Eco 25:29202936. Hopkins WA, Roe JH, Snodgrass JW, Jackson BP, Kling DE, Rowe CL, Congdon JD. 2001. Nondestructive indices of trace element exposure in squamate reptiles. Environ Pollut 115:1-7. Itoh T, Xia J, Magavi R, Nishihata T, Rytting JH. 1990. Use of shed snake skin as a model membrane for in vitro percutaneous penetration studies: comparison with human skin. Pharm Res 7:1042-1047. Jones DE, Gogal RM Jr, Nader PB, Holladay SD. 2005. Organochlorine detection in the shed skins of snakes. Ecotoxicol Environ Saf 60:282-287. Kaur S. 1988. Lead in the scales of cobras and wall lizards from rural and urban areas of Punjab, India. Sci Total Environ 77:289-290. Kaito T, Ota H, Toda M. 2017. The evolutionary history and taxonomic reevaluation of the Japanese coral snake, Sinomicrurus japonicus (Serpentes, Elapidae), endemic to the Ryukyu Archipelago, Japan, by use of molecular and morphological analyses. J Zool Syst Evol Res 55:156166. Krey K, Farajallah A. 2013. Skin histology and microtopography of Papuan white snake (Micropechis ikaheka) in relation to their zoogeographical distribution. HAYATI J Biosci 20:7-14. Kubie JL, Cohen J, Halpern M. 1978. Shedding enhances the sexual attractiveness of oestradiol treated garter snakes and their untreated penmates. Ani Behav 26:562-570. Kumpugdee-Vollrath M, Subongkot T, Ngawhirunpat T. 2013. Model membrane from shed snake skins. Int J Med Heal Biomed Bioeng Pharm Eng 7:669-676. Kuntz RE. 1963. Snakes of Taiwan. Quart J Taiwan Mus 16:180. Lev E. 2003. Traditional healing with animals (zootherapy): Medieval to present-day Levantine practice. J Ethnopharmacol 85:107-118. Lillywhite HB, Sheehy CM III. 2016. Synchrony of ecdysis in snakes. Herpetol Conserv Biol 11:286-292. Maki M. 1931. A Monograph of the Snakes of Japan. Dai-ichi Shobo, Publisher. Tokyo, Japan. McDiarmid RW, Foster MS Guyer C, Gibbons JW, Chernoff N. 2012. Reptile Biodiversity: Standard Methods for Inventory and Monitoring. University of California Press, Berkeley, California, USA. Medlin EC, Risch TS. 2006. An experimental test of snake skin use to deter nest predation. Condor 108:963-965. Mochida K, Zhang W-Y, Toda M. 2015. The function of body coloration of the hai coral snake Sinomicrurus japonicus boettgeri. Zool Stud 54:33. doi:10.1186/s40555-015-01102. Monroe EA, Monroe SE. 1967. Origins of iridescent colors on the indigo snake. Science 159:97-98. Mukherjee S, Dasgupta SC, Gomes A. 2013. Effect of Naja naja Laurenti shed skin extract on estrous cycle, hormone - cytokine profiles, histopathology of ovary and uterus of Swiss albino mice. Indian J Exp Biol 51:235-240. Ngawhirunpat T, Panomsuk S, Opanasopit P, Rojanarata T, Hatanaka T. 2006. Comparison of the percutaneous absorption of hydrophilic and lipophilic compounds in shed snake skin and human skin. Pharmazie 61:331-335. Ota H, Ito A, Lin J-T. 1999. Systematic review of the elapid snakes allied to Hemibungarus japonicus (Gunther, 1868) in the East Asian islands, with description of a new subspecies from the central Ryukyus. J Herpetol 33:675687. Ota H, Toyama M. 1989. Taxonomic re-definition of Achalinus formosanus Boulenger (Xenoderminae: Colubridae: Ophidia), with description of a new subspecies. Copeia 1989:597-602. page 21 of 23Zoological Studies 56: 38 (2017) © 2017 Academia Sinica, Taiwan Pope CH. 1935. The Reptiles of China: Turtles, Crocodilians. Snakes, Lizards. Natural History of Central Asia, Vol. 10. American Museum of Natural History, New York, USA. Price RM. 1982. Dorsal snake scale microdermatoglyphics: ecological indicator or taxonomic tool? J Herpetol 16:294306. Priprem A, Khamlert C, Pongjanyakul T, Radapong S, Rittirod T, Chitropas P. 2008. Comparative permeation studies between scale region of shed snake skin and human skin in vitro. Am J Agri Biol Sci 3:444-450. Priprem A, Pongjanyakul T, Khamlert C, Chitropas P, Kanla P, Sripanidkulchai K. 2007. Shed skin of Ophiophagus hannah: Structural topography and in vitro permeation of nicotine and phenol. Am J Anim Vet Sci 2:84-88. Reading CJ, Luiselli LM, Akani GC, Bonnet X, Amori G, Ballouard JM, Filippi E, Naulleau G, Pearson D, Rugiero L. 2010. Are snake populations in widespread decline? Biol Lett 6:777-780. Shang G, Li PX, Yang YJ. 2009. Guide to the Amphibians and Reptiles of Taiwan. OWL Publishing House, Taipei, Taiwan. (in Chinese) Stevenson DJ, Ravenscroft KR, Zappalorti RT, Ravenscroft MD, Weigley SW, Jenkins CL. 2010. Using a wildlife detector dog for locating eastern indigo snakes (Drymarchon couperi). Herp Rev 41:437-442. Takeuchi H, Ota H, Oh H-S, Hikida T. 2012. Extensive genetic divergence in the East Asian natricine snake, Rhabdophis tigrinus (Serpentes: Colubridae), with special reference to prominent geographical differentiation of the mitochondrial cytochrome b gene in Japanese populations. Biol J Linn Soc 105:395-408. Takeuchi H, Zhu G-X, Ding L, Tang Y, Ota H, Mori A, Oh H-S, Hikida T. 2014. Taxonomic validity and phylogeography of the East Eurasian natricine snake, Rhabdophis lateralis (Berthold, 1859) (Serpentes: Colubridae), as inferred from mitochondrial DNA sequence data. Curr Herpetol 33:148153. Todd BD, Nowakowski AJ, Rose JP, Price SJ. 2017. Species traits explaining sensitivity of snakes to human land use estimated from citizen science data. Biol Cons 206:31-36. Torri C, Mangoni A, Teta R, Fattorusso E, Alibardi L, Fermani S, Bonacini I, Gazzano M, Burghammer M, Fabbri D, Falini G. 2014. Skin lipid structure controls water permeability in snake molts. J Struct Biol 185:99-106. Trnka A, Prokop P. 2011. The use and function of snake skins in the nests of great reed warblers Acrocephalus arundinaceus. Ibis 153:627-630. Tsai T-S, Tu M-C. 1998. Sexual dimorphism of Chinese green tree viper, Trimereusrus s. stejnegeri. Biol Bull Nat Taiwan Normal Univ 33:13-22. (in Chinese) Tu MC, Fong SC, Lue KY. 1990. Reproductive biology of the sea snake, Laticauda semifasciata, in Taiwan. J Herpetol 24:119-126. Verveen AA, Rouwkema J. 2007. The iridescent epidermis of Boa constrictor. Litteratura Serpentium 27:117-136. Wang CS, Wang YM. 1956. The reptiles of Taiwan. Quart J Taiwan Mus 9:1-86. Wang Y, Zhou K. 1998. A microdermatoglyphic analysis of 16 species of snakes. Chin J Appl Environ Biol 4:152-158. (in Chinese) You C-W, Li S-W, Lau A. 2013. Distribution and natural history of the ratsnakes of Taiwan: An annotated photographic review. pp. 369-384. In: Schulz KD (ed) Old World Ratsnakes-A collection of papers. Bushmaster Publications, Switzerland. You C-W, Poyarkov NA Jr, Lin S-M. 2015. Diversity of the snaileating snakes Pareas (Serpentes, Pareatidae) from Taiwan. Zool Scripta 44:349-361. Zhao E, Huang M, Zong Y et al. 1998. Fauna Sinica: Reptilia Vol. 3, Squamata, Serpentes. Science Press, Beijing, China. (in Chinese) Zhao E. 2006. Snakes of China (I, II). Anhui Science and Technology Publishing House, Anhui, China. (in Chinese). page 22 of 23Zoological Studies 56: 38 (2017) © 2017 Academia Sinica, Taiwan Appendix I. Key to the complete shed skins of 53 snake species in Taiwan and adjacent islands. (download) Appendix Table A1. The catalogue numbers of shed vouchers for 53 species of snakes. See the text for detailed descriptions on the classification of body patterns. (download) Appendix Table A2. Comparisons of the actual skin and shed skin in 53 snakes bearing banding, stripes, spots, or other markings on dorsal/ ventral head, body, and tail. Cases with markings are denoted by a check mark and those with some exceptions are denoted by a checkmark in parentheses. Note that “Y” means neonates or juveniles and “A” means adults. The counts and ratios are listed at the end of the table. (download) page 23 of 23Zoological Studies 56: 38 (2017)