HARVARD UNIVERSITY Ernst Mayr Library of the Museum of Comparative Zoology ASIATIC HERPETOLOGICAL RESEARCH VOLUME 3 1990 Asiatic Herpetological Research Editor ERMI ZHAO Chengdu Institute of Biology, Acadcmia Sinica, Chengdu, Sichuan, China Associate Editors KELLAR AUTUMN THEODORE j, PAPENFUSS Museum of Vertebrate Zoology, University of Museum of Vertebrate Zoology, University of California, Berkeley, California, USA California, Berkeley, California, USA J. ROBERT MACEY Museum of Vertebrate Zoology, University of California, Berkeley, California, USA Editorial Board Kraig Adi.fr Cornell University, Ithaca, New York, USA Meiiii a Huang Zhcjiang Medical University, Hangzhou, Zhejiang, China Natalia B. Ananjeva Zoological Institute, Leningrad, USSR Leo Borkin Zoological Institute, Leningrad, USSR Bihui Chen Anhui Normal University, Wuhu, Anhui, China Ylanchong Chen Shanghai Institute of Biochemistry, Shanghai, China ILLYA DAREVSKY Zoological Institute, Leningrad, USSR INDRANEIL DAS Madras Crocodile Bank, Vadanemmeli Perur, Madras, India CarlGans University of Michigan, Ann Arbor, Michigan, USA David M. Green McGill University, Montreal, Quebec, Canada Robert F. Inger Field Museum, Chicago, Illinois, USA KlJANGYANG LUE National Taiwan Normal University, Taipei, Taiwan, China Hidetoshi Ota Department of Biology, University of the Ryukyus, Nishihara, Okinawa, Japan Anming Tan University of California, Berkeley, California, USA William E. Duellman University of Kansas, Lawrence, Kansas, USA Hajime Fukada Scnnyuji Sannaicho, Higashiyamaku, Kyoto, Japan Datong Yang Kunming Institute of Zoology, Kunming, Yunnan, China Asiatic Herpetological Research is published by the Asiatic Herpetological Research Society (AHRS) and the Chinese Society for the Study of Amphibians and Reptiles (CSSAR) at the Museum of Vertebrate Zoology, University of California. The editors encourage authors from all countries to submit articles concerning but not limited to Asian herpctology. Authors should consult Guidelines for Manuscript Preparation and Submission at the end of this issue. All correspondence outside of the People's Republic of China and requests for subscription should be directed to AHR, Museum of Vertebrate Zoology, University of California, Berkeley, California, USA 94720. All correspondence within the People's Republic of China should be directed to Ermi Zhao, Editor, Chengdu Institute of Biology, P.O. Box 416, Acadcmia Sinica, Chengdu, Sichuan Province, People's Republic of China. Subscription and membership are $20 per year ($25 outside of the USA and China, $40 for libraries). Make checks or money orders payable in US currency to AHRS. If you do not have access to US currency, please notify us, and we will make other arrangements. Asiatic Herpetological Research Volume 3 succeeds Chinese Herpetological Research Volume 2, which was published at the Museum of Vertebrate Zoology, 1988-1989 as the journal for the Chinese Society for the Study of Amphibians and Reptiles. Volume 2 succeeded Chinese Herpetological Research 1987, published for the Chengdu Institute of Biology by the Chongqing Branch Scientific and Technological Literature Press, Chongqing, Sichuan, China. Acta Herpetologica Sinica ceased publication February 1988, with Vol. 6, No. 2. April 1990 Asiatic Herpetological Research Vol. 3, pp. 1-36 Three Species in the Vipera kaznakowi Complex (Eurosiberian Group) in the Caucasus: Their Present Distribution, Possible Genesis, and Phylogeny* NIKOLAI L. ORLOV1 AND BORIS S. TUNIYEV2 ^Zoological Institute. USSR Academy of Sciences, Leningrad, USSR 2Causasian State Biosphere Reserve, Sochi, USSR Abstract.-Thiee valid species: Vipera ka2nakowi Nikolsky, V. dinniki Nikolsky and V. darevskii Vedmedeja, Orlov, and Tuniyev of the "Vipera kaznakowi " complex belonging to the Eurosiberian viper group are recognized. The distributions of the closely related species, V. kaznakowi and V. dinniki are defined. The habitat of V. kaznakowi is present along the Black Sea coast, from Khopa Village (Turkey) in the south, to Maikop (USSR) in the north. Vipera kaznakowi is associated with montane areas from sea level up to 1000 m. The range of V. dinniki covers the northern and southern slopes of the Great (Main) Caucasus Ridge, ranging from the Fisht-Oshten Massive in the west to Shkhara Mountain in the east. Vipera dinniki tends to be restricted to alpine and subalpine zones at elevations from 1 500 m to 3000 m. Vipera darevskii occurs in the southeastern part of the Dzhavachet Mountain Ridge, Armenia, near the border with Georgia. The history of studies on the vipers of the "Vipera kaznakowi " complex is summarized. The possible genesis of the present distributions of these vipers and their phylogenetic relationships are discussed. Comparative morphological and ecological characters of the three species are listed. Key Words: Reptilia, Serpentes, Viperidae, Vipera, USSR, Caucasus, systematics, ecology. Introduction The study of the following viper species of the Caucasus, Vipera kaznakowi Nikolsky 1909, V. dinniki (Nikolsky 1913), and V. darevskii Vedmederja, Orlov and Tuniyev 1986 and the history of their distributions is of great interest for understanding the phylogenetic links of the Eurosiberian shield-headed vipers from the Caucasus (subgenus Pelias, Merrem 1808) and the formation of the present snake fauna in the west Caucasus Isthmus. Ever since Nikolsky (1909) described the Caucasus Viper, Vipera kaznakowi, the understanding of the taxonomic position of this species has constantly varied. Also, ideas regarding the habitats of those forms referred to this species have varied (Orlov and Tuniyev 1986). We recognize at least three closely related species within the V. f This publication combines material previously published in Russian by Orlov and Tuniyev (1986) and Vedmederja et al. (1986) with additional information. kaznakowi complex. All of them occur primarily in the western part of the Caucasus Isthmus. The eastern boundary of the species range needs to be defined. The Caucasus vipers (Fig. 1) have been recorded from the following localities: Western Dagestan in the vicinity of Khasav-Yurt (Krasovsky 1929, 1932). Along the slope of Makh-khokh Mountain from the highlands of Ingushetia (Chernov 1929). On Legli Mountain in the Gukasyansky region, Armenia (Darevsky 1956). In the vicinity of Ushguli at the foothills of Shkhara Mountain in Svanetia, Georgia (Muskhelishvili 1959). In Borzhomi Canyon, Georgia (Bakradze 1969, 1975; Bannikov et al. 1977). In Lagodekhi, Georgia (Zoological Institute, the USSR Academy of Sciences, Nos. 8389 and 13769). 1990 by Asiatic Herpetological Research Vol. 3, p. 2 Asiatic Herpetological Research April 1990 FIG. 1. The distributions of the vipers from the Vipera kaznakowi complex and the Vipera ursini complex in the Caucasus. 1. V. kaznakowi, 2. V. dinniki, 3. V. darevskii, 4. V. ursini renardi (northern population) and V. ursini eriwanensis (southern population). Type localities: A -V. kaznakowi; B, B' -V. dinniki; C-V. darevskii. Localities: 1 -Makh-Khokh Mountain; 2 -the settlement of Khasav-Yurt; 3 -Lagodekhi; 4 -Benis-Kheri Canyon; 5 -the settlement of Khopa, Turkey; 6 -Mikhailovsky Pass; 7 -the settlement of Ubinskaya; 8 -the town of Maikop; 9 -the mouth of the Urushten River; 10 -the Fisht-Oshtenovsky Massive; 11 -Shkhara Mountain. Extreme points of the distributions are marked. All three species, Vipera kaznakowi, V. dinniki and V. darevskii, differ in morphology and ecology (Vedmederja 1984; Orlov and Tuniyev 1986). History of the study of the Vipera kaznakowi complex. The taxonomic problems and interrelationships of the viper forms of the Eurosiberian group in the Caucasus has caused contradictions and has confused zoologists for nearly a hundred years. The fragmented highland landscapes within the overall range of the Caucasus vipers isolates even neighboring populations. Hence, there is an accumulation of unique characteristics in populations. Very complex situations emerge in sympatric areas of closely related species of" this complex. This may be connected with natural hybridization. Occasionally the identification of some individuals from intergrading populations is difficult. This particularly concerns Vipera dinniki in the areas of its interactions with V. kaznakowi, and additionally with the vipers of the V. ursini complex (Bonaparte). The high April 1990 Asiatic Herpetological Research Vol. 3, p. 3 FIG. 2. Type localities of the vipers of the Vipera kaznakowi complex proposed over the history of the study. A -V. kaznakowi Nikolsky, described in 1909, collected from Tsebelda, the vicinity of Sukhumi, Abkhazia, Georgia. B, B' -V. Berus dinniki Nikolsky, described in 1913, from the upper reaches of the Malaya Laba River, the Caucasus Reserve, Krasnodarsky Territory and Svanetia, Georgia (respectively). C -V. tigrina Tzarewsky, described in 1916, taken from the Northern Caucasus. D-V. berus ornata Basoglu, described in 1947 from the settlement of Khopa, Artviisky Vilayet, Turkey. E -V. darevskii Vedmederja, Orlov, and Tuniyev, 1986, from Legli Mountain, the Mokrye Mountains, Gukasyansky Region, Armenia. F-V. kaznakowi orientalis Vedmederja, 1984 from the eastern part of the Main Caucasus Ridge. degree of phenotypic polymorphism in vipers from the V. kaznakowi complex also creates additional problems in the evaluation of their taxonomic position. Due to the complexity in identification of these vipers and the absence of definite localities for specimens an analysis of old literature gives only an approximate idea regarding the relationships of the Vipera kaznakowi complex. Rossikow (1890) mentioned "multi- colored" vipers in the Northern Caucasus defined as Vipera berus. Boettger (1893) regarded the dispersal of two closely related viper species, V. berus and V. renardi from the Caucasus as a phenomenon that deserves attention. Nikolsky (1905) considered V. berus to be associated with Transcaucasia whereas V. renardi inhabits steppe areas of the Precaucasus and mountains of the Northern Caucasus. Leister (1908a) in his sketch on the geographic distribution of V. renardi and V. berus within the Caucasus wrote that V. Vol. 3, p. 4 Asiatic Herpetological Research April 1990 renardi is associated with steppe areas which separate the Caucasus from the habitat of V. berus, and also the mountains of the Northern Caucasus. He also confirmed Boettger's and Nikolsky's opinions that V. berus also appears in Transcaucasia. In the northern Transcaucasia there is an isolated population. According to Boettger (1899), the individuals of V. berus which are preserved at the Caucasus Museum (now the Museum of Georgia) were collected from Suani, Tiflis, Avar, Kodzhori, Khasav-Yurt, and Kazikoparan, whereas those from the Senckenburg Museum (Frankfurt-am-Main, West Germany) were taken from Sukhumi, Georgia. Leister (1908a,b,c) gives a list of Vipera berus specimens in the collection of the Zoological Museum of the Academy of Sciences (now Zoological Institute of the USSR Academy of Sciences) taken from the Caucasus: 1) from Tiflis and Lagodekhi regions, Georgia and 2) from Yelenovka near the Gochka Lake (now the Sevan Lake). Leister then states that he found V. renardi on the bank of the Gochka Lake at the same locality where Kessler and Brandt collected V. berus (this specimen, ZIN 5478, collected by Brandt we determined to be V. ursini ). On the basis of the findings of Kessler and Brandt and his own findings, Leister comes to the conclusion that both species, V. berus and V. renardi, live sympatrically. Concerning the specimens collected by Kessler and Brandt, Nikolsky (1905) writes that their rostral touches only one apical scale like in V. renardi. In all other characters they look more like V. berus. This led Leister (1908a) to consider that a certain intermediate species is possible — an ancestral form of V. berus and V. renardi. Nikolsky (1913) referring to the vipers from Transcaucasia listed by Kessler, Brandt and Leister nevertheless concluded that they are V. renardi. Leister (1908a,b), having analyzed the collections, concludes that the territory of the Northern Caucasus and Transcaucasia are inhabited by V. berus and V. renardi. He thought that V. berus and V. renardi live sympatrically in the Northern Caucasus (for instance, the vicinity of Grozny is a sympatric area). Since Nikolsky (1909) described Vipera kaznakowi, a number of synonyms have been proposed for this species as new forms have been described and new combinations proposed. A number of forms, primarily from the Northern Caucasus and from the eastern range of V. kaznakowi have confused taxonomists as to their relationship to V. kaznakowi. These forms were placed in different combinations within V. berus and V. ursini (Nikolsky 1913, 1916; Basoglu 1947; Knoepfler and Sochurek 1955; Kramer 1961). Nikolsky (1913) assigned the common Caucasus viper to a new subspecies, V. berus dinniki. He defined the viper's range as the northern and southern slopes of the Caucasus Ridge from the Malaya (Small) Laba River headwaters up to Elbrus Mountain. Morits (1916) also recorded V. berus in the Northern Caucasus. Tsarevsky (1916) described a new form V. tigrina that was closely related to V. renardi and V. kaznakowi. Unfortunately he cited the species locality solely as "the Northern Caucasus." Nikolsky (1909, 1911, 1913, 1916), the author of Vipera kaznakowi and V. berus dinniki descriptions, studied the relationships of the species V. kaznakowi, V. berus, and V. renardi (= V. ursini renardi ) and noted the difficulties in identification of these forms. Nikolsky (1913) described V. berus dinniki from a single specimen collected by N. Y. Dinnik at the upper reaches of the Malaya Laba River, and from three specimens found by Shelkovnikov in Svanetia, Georgia. In his next monograph, Nikolsky (1916) discussed both forms: Coluber berus dinniki and Coluber kaznakowi. The range of the former was defined as the Caucasus Mountains on both sides of the Great Caucasus Ridge. The upper reaches of the Bolshaya (Big) Laba River was designated as a sympatric zone for V. berus dinniki and V. renardi. In his opinion, V. renardi was closely associated with the Caucasus Black Sea coast and the northern slope of the Caucasus Ridge. April 1990 Asiatic Herpetological Research Vol. 3, p. 5 Dinnik (1926) noted that Viper a berus and V. kaznakowi occurred in the Northern Caucasus. Krasovsky (1929) indicated that V. berus inhabited Khasav-Yurt and Rutulsky Kanton, Dagestan. Chernov (1929) recorded Coluber berus dinniki in the highlands of Ingushetia on the southern slope of Makh-khokh Mountain. The specimens were collected by D. Krasovsky in 1926 (Fig. 1). Krasovsky (1933) included both V. kaznakowi and V. berus in the fauna of the Caucasus State Reserve. Bartenev and Reznikova (1935) concluded that V. kaznakowi and V. berus dinniki were distributed in the western Caucasus, whereas in the alpine, V. ursini renardi also occurred. Rostombekov (1939) recorded V. kaznakowi as part of the fauna of Abkhazia. A new form, V. berus ornata, was described from northeast Anatolia (Basoglu 1947). Later on Mertens (1952a) synonymized it with V . kaznakowi. Terentyev and Chernov (1949) regarded V. tigrina Tzarevsky and V. berus dinniki Nikolsky as junior synonyms of V. kaznakowi Nikolsky which is, in their opinion, related to V. ursini renardi. Darevsky (1956) presented a new combination of names for the viper from the Gukasyansky region, Armenia: Vipera ursini renardi. Fyodorov (1956) recorded V. berus in the forests of the premontane area of the Stavropolsky Territory and V. kaznakowi in the subalpine belt. In his survey of the snake fauna of Abkhazia, Georgia, Milyanovsky (1957) mentions V. kaznakowi. Muskhelishvili (1959) records V. kaznakowi on Mount Shkhara from the vicinity of Ushguli, Svanetia, Georgia. Kramer (1961) regards V. tigrina a junior synonym of V. kaznakowi, whereas V. berus dinniki should be synonymized with V. ursini renardi. Bakradze (1969) found a female V. kaznakowi in the town of Banis- Khevi, near Borzhomi, Georgia. On the basis of this finding, he suggested that the species habitat covered the entire Adzharo- Imeretinsky Ridge and some of the Trialetsky Ridge. In the field guide of Bannikov et al. (1977) the names Vipera berus dinniki, V. tigrina and V. berus ornata are mentioned as junior synonyms of V. kaznakowi. Vedmederja (1977) recorded V. kaznakowi in Adgaria, Georgia. In his opinion (Vedmederja 1984) V. kaznakowi is a polytypic species comprising four subspecific forms. Tertyshnikov (1977), in defining ecological and zoogeographic subdivisions of the herpetofauna of the Northern Caucasus, noted that V. kaznakowi tends to be restricted to western and southwestern montane regions. The most recent taxonomic works dealing with these vipers state that V. berus occurs neither in the Caucasus nor in the Precaucasus. Hence, V. kaznakowi is the sole valid name with regard to shield- headed vipers from the western Caucasus Isthmus and northeast Anatolia. The taxonomic status of the eastern populations from Dagestan and Armenia was not discussed (Terentyev and Chernov 1949; Mertens 1952a; Mertens and Wermuth 1960; Klemmer 1963; Bannikov et al. 1977; Baran 1977; Harding and Welch 1980). The study of old collections and literature shows that the majority of researchers did differentiate the vipers from the Vipera ursini complex and the V. kaznakowi complex. The major difficulties in defining the systematic position and taxonomic status concerned primarily the vipers from the eastern range of the V. kaznakowi complex. References that V. berus occurs in the Northern Caucasus and Transcaucasia more often than not concern snakes from the V. kaznakowi complex rather than those from the V. ursini complex. An analysis of literature data and morphometric characteristics of 141 viper specimens allowed Vedmederja et al. (1986) to designate three species within the Vipera kaznakowi complex: 1. V. kaznakowi Nikolsky, 1909 2. V. dinniki Nikolsky, 1913 3. V. darevskii Vedmederja, Orlov and Tuniyev, 1986 Vol. 3, p. 6 Asiatic Herpetological Research April 1990 Fig. 2 represents type localities of the forms proposed for the Vipera kaznakowi complex over the history of its study. Systematic Accounts Vipera kaznakowi Nikolsky, 1909 (Fig. 3, 4, 14a, Plate 1) Chronology of species description Vipera renardi (Christoph) - Silantyev, 1903, 30:37 (part). Vipera kaznakowi Nikolsky, 1909:174; Nikolsky, 1910:81, table 1. Vipera kaznakowi - Nikolsky, 1913:179-181; colored plate III (=Fig. 3, this paper). Coluber kaznakowi - Nikolsky, 1916:244-247. Vipera berus ornata Basoglu, 1947:182- 190. Vipera ursini kaznakowi - Knoepfler and Sochurek, 1955:185-188. Vipera kaznakowi - Terentyev and Chernov, 1949:270-277 (Map 15); Bannikov et al., 1977:323-324 (map 133; colored plate 31, 4). Vipera kaznakowi kaznakowi - Vedmederja, 1984:8. The English common name is the Caucasus Viper, or Kaznakow's Viper. Lectotype1: No. 4408, an adult female. Collected by Y. V. Voronov from Tsebelda, the vicinity of Sukhumi, Abkhasia, the Caucasus. It is stored in the Caucasian Museum (now the Museum of Georgia, Tbilisi). + Kramer (1961) was mistaken regarding this specimen as the holotype, because Nikolsky (1909) had not singled out a holotype from the five specimens from which he described the species. Diagnosis: A large snake for the Eurosiberian group. Total length reaches 650-700 mm. Dorsally the head is covered with large scales. In size and shape they differ considerably from body scales. Nostril is cut through either in the middle or slightly closer to the lower edge of the nasal. Upper-lateral edge of snout is pointed. Rostral normally reaches two apical scales on upper snout. Upper edge of the nasalorostral scale is slightly curved at obtuse angles. Scales of anterior frontal have weak longitudinal keels. Head is broad and normally black dorsally. Head is separated from body by a sharp nuchal collar. Description: The ratio of body length to tail length is 5.7 to 6.4 in males and 7.5 to 10.9 in females. Unlike other species assigned to the Vipera kaznakowi complex, red and yellow colors prevail in V. kaznakowi. Melanistic specimens are common. However, unlike complete melanistic individuals of V. dinniki, those of V. kaznakowi preserve yellow or red color on either upper or lower labials. Vipera kaznakowi is either black or dark brown striped dorsally and laterally. Often stripes merge so that only red or yellow spots remain between them. Ventrum is black. Head is very broad, impressed dorsally. This fact is emphasized by a slightly pointed upper edge of snout. Cheeks are greatly swollen. Head is well separated from body by a thin nuchal collar. Comparative data on scalation and size characters are listed in Tables 1-4. Variability and comparative remarks V. kaznakowi (Fig. 4, Plate 1) is undoubtedly more closely related to V. berus and V. ursini. Supposedly, when interacting with V. ursini renardi and V. ursini eriwanensis it forms two species of hybrid genesis: V. dinniki and V. darevskii. From all listed forms it differs by: 1) the extraordinarily great width of the triangular head, 2) the head width of specimens of this species equals the distance between tip of snout and posterior angle of mouth slot, 3) the "cheeks" are greatly swollen, hence a broad furrow April 1990 Asiatic Herpetological Research Vol. 3, p. 7 Nikolski. Nova species viperae. H3B. K»bk. My3. T. V. A. 3. E. Tom J. I. II KiiiniiKiir C, TI- E. Vipera kaznakovi sp n FIG. 3. Type specimen of Vipera kaznakowi Nikolsky, 1913 (reproduced from Plate III, originally printed in color, Nikolsky 1913). Vol. 3, p. 8 Asiatic Herpetological Research April 1990 TABLE 1 . Comparative scale characters of the three viper species of the "Vipera kaznakowi " complex. Species Vipera kaznakowi Vipera d inniki Vipera dare vskii Characters n i Tiin-max x±se n min-max x±se n min-max x±se 1. Number of scales 64 8-12 10.03±0.1 65 9-12 9.610.1 9 8-9 8.410.2 2. round the eye Number of upper 64 8-10 9.0±0.1 68 8-11 10.311.3 9 9-10 9.210.1 3. labials Number of lower 64 8-12 10.910.05 68 8-12 10310.1 . 9 9-10 9.510.2 4. 5. labials Number of ventrals Number of scale rows around body 64 64 130-143 18-21 135310.8 20.010.1 68 68 126-141 21-23 133.911.8 21.110.15 9 9 134-140 19-21 135.510.8 20.610.2 TABLE 2. Sexual dimorphism of body length and number of subcaudals in the three viper species of the "Vipera kaznakowi " complex. to* 9 Characters 66 99 n min-max x±se n min-max x±se Vipera kaznakowi: Body length Number of subcaudals 23 23 358-475 31-40 415.1+7.5 34.811.5 16 20 375-600 22-32 504.11118.6 25.410.9 Vipera dinniki: Body length Number of subcaudals 29 29 259-112 31-37 33118.76 32.210.6 20 20 321-486 18-30 441.8+11.4 23.7+0.8 Vipera darevskii: Body length Number of subcaudals 3 3 236-258 29-35 249.315.53 32.311.4 5 5 233^21 25-30 331.8128.2 2710.79 4.5 5.6 2.82 < 0.001 < 0.001 <0.05 TABLE 3. A comparison of the number of vertebrae carrying ribs in Shield Headed .Vipers of the USSR. X-ray analysis of the vipers indicates that the number of rib-carrying vertebrae is stable, except for Vipera dinniki, (indicated by **) in which the number varied from 128 to 140. Specific and subspecific viper forms Vipera berus Vipera kaznakowi Vipera darevskii Vipera dinniki Vipera ursini renardi Vipera u. eriwanensis Vipera dinniki Number of specimens Ni jmber of rib bearing analyzed vertebrae 10 142 8 135 9 138 25 138" 10 148 10 140 2 132 and 135 appears between eye and temple. The furrow goes up to upper head, parallel to upper edge of parietals. Dorsally the head is either impressed or flat. The viper also differs from closely related species in body proportions: 1) it is relatively thicker, more massive; 2) the head is conspicuously bigger. These habitus characters differentiate all age groups of V. kaznakowi ranging from juveniles to mature specimens. Comparative data on pholidosis, size, and number of vertebrae April 1990 Asiatic Herpetological Research Vol. 3, p. 9 TABLE 4. Comparative morphology of vipers in the "Vipera kaznakowi " complex. Vipera kaznakowi, 64 specimens • SVL 8 8 ^475 mm SVL V 9 y< 600 mm • SVL 86/ CI. x 5.6-6.4 SVL 9 9/ CL as 7.5-10.9 • Head is either impressed or flat dorsally • Edge of snout is slightly rounded V. dinniki, 68 specimens • SVL 6 6 ^ 412 mm SVL 9 9 ^486 mm • SVL 8 8/ CL a 5.9-7.4 SVL 9 9/ CL as 7.9-13 5 • Head is either flat or slightly protuberant dorsally • Edge of snout is rounded • Rostral is broad. Usually touches < two apical scales (in 91% specimens) rarely one apical (in 9%) • The width of the frontal is equal to i 1.21-1.72 of its length • The distance between anterior edge i of the frontal and the rostral is equal to 0.75-1.05 of the frontal edge • Frontal is either smaller than ( parietals or equals them • Large lower oculars are separated i from frontal either by one row (in 87.0%) or rarely by two rows of small scales (in 12.4%) • The upper pre -ocular does not touch ( the nasal in 96.1%. In 3.9% it does • Nostril is either cut through in the i middle of nasal or is slightly shifted downwards • Nasal does not touch rostral < • Body scales with expressed keels; < some scale rows that reach ventrals have no keels Rostral is narrow reaching either one (in 48.6%) or two (in 51.4%) apicals The width of the frontal is equal to 1.13-1.83 of its length The distance between anterior edge of the frontal and the rostral is equal to 0.77-1.35 of the frontal edge Frontal is either smaller than parietals or equals them Large lower oculars are separated from frontal either by one row (in 58.5%) or two rows (in 41.5%) The upper pre-ocular does not touch the nasal in 86.6%. In other cases it reaches the nasal Nostril is cut through in the center or nasal. It may be shifted downwards exceedingly rarely Nasal does not touch rostral Body scales have expressed keels. 76% of scales that reach ventrals have no keels, 18.4% have slightly expressed keels, 5.6% have well-expressed keels V. darevskii, 9 specimens • SVL 8 8 ? 258 mm SVL 9 9?* 421 mm SVL 88 / CL a 6.0-6.5 SVL 9 9 / CL a8.4-9.3 • Head is either flat or slightly protuberant dorsally • Edges of snout are slightly pointed laterally. Anterior edge is a little rounded • Rostral is more narrow and touches one (in three specimens) or two (in six specimens) apical scales • The width of the frontal is equal to 1.48-1.71 of its length • The distance between anterior edge of the frontal and the rostral is equal to 0.67-1.07 of the frontal edge • Frontal is larger than parietals • Large lower oculars are separated from frontal only by a single row of small scales (in all nine specimens) • The upper pre-ocular does not touch the nasal in two specimens. In seven specimens it does • Nostril is cut through in the lower part of the nasal • Nasal does not touch rostral • Body scales have expressed keels. Some scale rows that reach ventrals have no keels supporting ribs are given in Tables 1-4. Coloration. Vipera kaznakowi are incredibly diverse in color and extraordinarily bright among shield-headed vipers. Red hues prevail in the color pattern. Newborns are similarly bright in color, primarily red brown, unlike the gray young of V. dinniki. The typical intensive red color appears in V. kaznakowi after they have shed twice. Complete melanists are frequent in populations. Often the black dorsal stripe merges with lateral stripes so that either red or yellow spots arranged in two rows along the dorsum remain. The dorsal stripe can be either zig-zag shaped or in the shape of an even broad line. Ventrum is black. Head pattern of adult specimens normally blends with the dorsal stripe. In immature specimens, the head pattern may be separated from the dorsal stripe by a light interval that disappears after maturation. Vol. 3, p. 10 Asiatic Herpetological Research April 1990 FIG. 4. Vipera kaznakowi (ZIN 1 1529) from the isolated forest of Babuk-Aul. PLATE 1 . Vipera kaznakowi (ZIN 1 1529) from the isolated forest of Babuk-Aul. PLATE 2. One pattern variation of Vipera dinniki from Krasnaya Polyana (ZIN 12153), a male. PLATE 3. Vipera ursini eriwanensis from the valley of Kassach, the foothills of the Ara-Fler Mountains. Orlov and Tuniyev Asiatic Herpetological Research Plate 1 % Muy^ \JU Vipera kaznakowi Orlov and Tuniyev Asiatic Herpetological Research Plate 2 Vipera dinniki \ au^ if\\ Orlov and Tuniyev Asiatic Herpetological Research Plate 3 Vipera ursini eriwanensis 1 2Uu# W April 1990 Asiatic Herpetological Research Vol. 3, p. 11 Sexual dimorphism. Maximum body length is longer in females (up to 600 mm) than in males (does not exceed 457 mm). Males have longer tails (Table 2). Accordingly, the number of ventrals is greater in females, whereas the number of subcaudals is greater in males. Males are more slender. Sexual dimorphism in coloration is feebly marked. Melanistic individuals are more frequent among females. Age variability. Vipera kaznakowi are born with the typical adult color and pattern. However, in newborns these characters are less pronounced than in adults. Newborn snakes may be either pinkish or reddish. According to our observations on birth and development of snakes from the Sochi-Khosta populations, the coloration becomes stronger with each subsequent shedding. Maximal color intensity is achieved by the season after the first hibernation. Melanistic specimens are born with the typical species pattern, but their coloration is darker. The coloration becomes darker during subsequent sheddings and elements of the pattern merge. New born litters are homogeneously colored. On maturing, the coloration becomes diverse. Phenotypic polymorphism in mature vipers of a litter is great. Either partial or nearly complete melanism is observed in all vipers from this population. This is typical for a number of other animals from humid subtropic areas adjacent to the Black Sea coast (Bartenev and Reznikova 1935). Specimens showing maximal melanism always preserve elements of orange and red on the throat scales and chin shields, the rostral, upper labials and subcaudals. Melanistic specimens of V. dinniki can have completely black coloration by maturity. Geographic range and ecology. The species ranges along the Black Sea coast from the town of Khopa, Turkey and Suramsky Pass in the east, then throughout Kolchida (Colchis) up to Mikhailovsky Pass in the west. It is then found up to the northern slope of the Main Caucasus Ridge. Here Vipera kaznakowi occurs along the foothills from the settlement of Ubinskaya in the west to the town of Maikop, USSR in the north and the mouth of the Urushten River in the east (Fig. 1 from Vedmederja et. al. 1986). The species generally occurs up to an elevation of 800 m. Along the river valleys of the Black Sea coast, it may occur up to an elevation of 1000 m or even higher. Vipera kaznakowi is a forest dwelling species. It occurs on montane wooded slopes, in the bottoms of humid canyons, and in meadows adjacent to forests. It is recorded in Quercetum azaleorum and Quercetum coggygriosecornosum oak groves, in mixed subtropic forests with evergreen subforests of Quercus hartwissiana, Quercus iberica, Alnus barbata, Fagus orientalis, Taxus baccata, Laurocerasus officinalis, Buxux colchicus, Castanetum colchicum, Fagetum nudum, Salictum fontenale, and Alnetum strut hiopteridor sum. In addition, this species is found in polydominant forests in river terraces and large overgrown outcroppings. At the upper limits of its elevational range, the species reaches coniferous forests. It is recorded within the ecotone of Fageto-Abieta athyridosa- maxtoherbosa, but the viper never penetrates deep into coniferous forests. Vipera kaznakowi is also present within transformed areas such as meadows formed after forests are cut, fruit orchards, kitchen gardens, vineyards, and dilapidated parks (Red Data Book of the USSR 1978, 1984). As a rule, vipers occur within sites where the density of lizards is high. Typical biotypes are as follows: small meadows and other illuminated spots in forests with an exposure providing high solar radiation in conditions of humid subtropical climate of the Black Sea coast of the Caucasus. The sites are located in the vicinity of standing water, where rocks exit that are suitable for hibernation. The climate of the Caucasus coast is greatly dependent on topographic conditions of the area that forms a narrow line between the Caucasus Ridge and the Black Sea. It is Vol. 3, p. 12 Asiatic Herpetological Research April 1990 situated from northwest to southeast along the coast. The wall of the Ridge starts from Anapa. Near Novorossiisk it reaches an elevation of up to 600 m. Near Tuapse, the wall exceeds 1000 m above sea level. At the latitude of Sochi, it is 3000 m high. The Great Caucasus Ridge presents a barrier for cold northeastern winds. This barrier separates the warm humid coast from the comparatively cold continental Prekuban area (Korostelyov 1933). In the east the Adzharo-Imeretinsky Ridge separates the humid subtropical Kolchida (Colchis) from the arid regions of the Eastern Precaucasia. The climatic zone of the Black Sea coast of the Caucasus appears to be extremely favorable for Vipera kaznakowi. The northwestern most localities for this species (Mikhailovsky Pass and Stanitsa Ubinskaya) coincide with Pontic and northwestern floristic regions. In relief, climate, and vegetation the northwestern region is a continuation of the southern coast of the Crimea. The Pontic region is marked by 1) subtropical vegetation in the foothills, 2) great temperature stability, and 3) high humidity (Korostelyov 1933). The limited dispersal of the species along the northern slope of the Big Caucasus is in the region of the "Kolchidian (Colchis) Gates." Due to the lowering of the western portion of the Main Caucasus Ridge, interchange of flora and fauna of Kolchida (Colchis) and the Prekuban area occurred in the past and presently occurs. Warm humid air from the Black Sea breaks through in this part of the Ridge. On the northern slope of the Big Caucasus, it creates a small refugium on a Kolchidian (Colchis) type associated with Castanea satyra, Buxux colchicus, Ostrya carpinifolia, Corylus colurna, and others (Galushko 1974; Kharadze 1974; Kholyavko et al. 1978). Of the herpetofauna of Kolkhida B ufo verucossimus, Triturus vittatus ophryticus, Pelodytes caucasicus, and Lacerta derjugini occur along with V. kaznakowi. Along the entire area of of the Black Sea coast of the Caucasus, the viper is rare. In a number of spots it has disappeared. In some situations, only single specimens of V. kaznakowi might be encountered. The highest density of Vipera kaznakowi that have been observed are in rocky outcroppings within the forest belt in the mountains of the Caucasus Reserve. In beech forests along a road in the valley of the Achipse River, we have recorded three specimens per kilometer of road. Single specimens have been recorded in the vicinity of Khosta, Babuk-Aul, Guzeripl, and Kisha Kordons. The density of the viper is also low in the southern habitats in Adzharia and Lazistan, Georgia (Basoglu 1947; Vedmederja 1977; Basoglu and Baran 1977). Anthropogenic factors are responsible for declining numbers and populations of Vipera kaznakowi. There are pressures due to recreational use of health resorts along the Black Sea coast, ploughing of sites near the foothills, and to a smaller extent, hay-mowing (Red Data Book of the USSR 1978, 1984). Within some resort areas V. kaznakowi is completely extinct. Diet. Vipera kaznakowi feeds on various animals. Different populations show specific feeding patterns with regard to prey available. According to data recorded in captivity, individual preference is observed. When stimulated to regurgitate, the following food species were recorded in the field: Apodemus sylvaticus, a forest mouse, A. agrarius, a field mouse, Microtus majori, a juvenile specimen, M. gud, Sorex raddei, Lacerta saxicola, L. derjugini, L. praticola and L. agilis. In the collection of the Zoological Museum of the Moscow State University a viper specimen from Bebysyry Lake is preserved. In its stomach a juvenile grass snake (Natrix natrix ) was found. Immature individuals feed on juvenile lizards of the above mentioned species, and to a lesser extent, on Orthoptera. In captivity adult vipers readily take any small rodents, fledgling sparrows, lizards, and pieces of chicken. Young vipers usually start on small lizards and crickets (Grillus bimaculatus and Achaeta domestica ). After some months they usually take newborn mice. After the viper's bite a prey normally dies in 5-7 April 1990 Asiatic Herpetological Research Vol. 3, p. 13 minutes. The snake never persecutes its prey. It finds the prey some time later using its olfactory organs. Swallowing ranges from one minute to 3.5 hours depending on prey size and the state of a snake. Complete digestion in the wild takes up to five days. In captivity digestion may take 30 to 40 hours at stable temperature. Optimal day temperature is 26-30°C, and night temperature 18°C during activity period. Shedding. Mature vipers normally shed 2 to 3 times during the activity period. General shedding is observed in June. New born vipers shed in the first hours after birth. Before entering hibernation, they shed again. Reproduction. Mating is recorded from late March to April. Birth occurs in late August. Females give birth to 3 to 5 young. The observed time of birth lasts for about two hours with intervals of 20 to 40 minutes (Zinyakova and Trofimov 1977). In captivity, the majority of females give birth at night, between 2400 and 0600 hrs. Some individuals are born in transparent capsules, which the neonates leave in the first minutes after birth. Females reproduce annually. Gravid females continue to take food right up to birth. Development of young. Neonate Vipera kaznakowi have a mean body length of 144.75 mm, tail length of 14.0 mm. Mean body weight is 4.1 g (n=8). After first shedding on the second day after birth, newborns start actively feeding on insects or small lizards. After birth the activity period is 1.5 to 2.5 months, whereas newborns of V. dinniki never take food and almost immediately enter hibernation, during which snakes increase in length from 10 to 20 mm. Vipera kaznakowi lose 0.3 g of initial weight during the first month after birth. In the second month they restore their initial weight and then increase it approximately 1 g before entering hibernation. One year old specimens have a body length of 200 mm and a tail length of 24 mm. Vipers reach sexual maturity by the third year at a SVL of 350-400 mm. Territoriality. Like other vipers, Vipera kaznakowi is conservative in territory use. The same individuals can be encountered in the same places during different seasons. Vipera kaznakowi utilizes considerably larger individual ranges than V. dinniki. Seasonal and daily activity. Along the Black Sea coast of the Caucausus, Vipera kaznakowi emerge after hibernation in March at an altitude of 600 to 800 m. On the northern slope of the Big Caucasus the vipers appear in the second half of April to early May when the mean day temperature is 13°to 16°C. In the foothills the vipers enter hibernation at the beginning of November at an altitude of up to 600 m. In the upper elevational limits of its range V. kaznakowi hibernate in late September to early October. New born vipers are more active than those of other age groups. Two sharply marked peaks of daily activity can be observed in Vipera kaznakowi. In the morning the period of daily activity ranges from 0730 to 1130 hrs, and in the evening from 1630 to 1830 hrs. At those times the soil temperature does not exceed 30-32°C in the sites inhabited by V. kaznakowi. Sympatric species. Along with Vipera kaznakowi the following species of reptiles occur: Lacerta derjugini, L. saxicola, L. praticola, L. agilis, Anguis fragilis, Pseudopus apodus, Coluber najadum, C. jugularis, Elaphe longissima, Natrix natrix, N. tessellata, and Coronella austriaca. Narrow sympatric areas appear with Testudo graeca, Lacerta trilineata, L. caucasica, L. rudis, L. parvula, L. mixta, Elaphe hohenackeri, Vipera ursini, and V. dinniki. Vipera dinniki Nikolsky, 1913 (Fig. 5, 6, 7, 12, 14ab,16, Plate 2) Species description in chronology Vipera berus - Boettger in Radde, 1899:286; Nikolsky, 1905:304 (ad Caucasus). Pelias chersea - Menetries, 1 832:73 (part). Vol. 3, p. 14 Asiatic Herpetological Research April 1990 Vipera xanthina -Dinnik, 1902:34. Vipera renardi - Silantyer, 1903, 30:37 (part). Vipera berus dinniki Nikolsky, 1913:176- 179. Coluber berus dinniki - Nikolsky, 1916:240-244. Vipera tigrina Tzarevsky, 1916:32-37. Vipera ursini renardi - Kramer, 1961:715. Vipera ursini kaznakowi - Knoepfler and Sochurek, 1955:185-188. Vipera kaznakowi - Terentyev and Chernov, 1949:270-271 (map 5); Bannikov et al., 1977:323-324 (map 133, colored plate 31,4). Vipera kaznakowi dinniki - Vedmederja, 1984:8. Vipera kaznakowi orientalis - Vedmederja, 1984:9, nomen nudum. The English common name is Dinnik's Viper or the Caucusus Subalpine Viper. Lectotype: No. 26044, an adult female collected by N. Y. Dinnik (Fig. 5) from the upper reaches of the Malaya (Small) Laba River, Northern Caucasus and Svanetia, Georgia (Fig. 2: B, B'). The specimen is preserved at the Museum of Natural History, Kharkov State University, Ukraine. Diagnosis: Total length reaches 500 to 550 mm. Dorsally the head is covered with large scales. Nostril is cut through in the center of the nasal. Upperlateral snout edge is rounded and slightly pointed. Rostral touches either one or two apical scales on upper head. Three to four scale rows with no keels are between the rostral and frontal. Head is not broad. Nuchal collar is not expressed. Description: In males body length is not greater than 412 mm; in females it does not FIG. 5. Lectotype of Vipera dinniki (The Museum of Natural History of Kharkov State University 26044). exceed 486 mm. The ratio of body length to tail length is 5.9 to 7.4 in males; 7.8 to 13.5 in females. General coloration is not as bright as in Vipera kaznakowi. However, specimens with bright yellow and orange elements can be observed. Normally V. dinniki (Fig. 6,.Plate 2) have light brown, grey, silver greyish or green greyish color which never occurs in V. kaznakowi. Some specimens have a dark even dorsal stripe along the center of the body. The latter substitutes for the zig- zag shaped stripe typical for the majority of plate-headed vipers. Ventrum is either dark and light spotted or light grey and dark speckled. Neonates of V. dinniki occasionally don't have the red color of the body typical for V. kaznakowi. They can be bom grey brown. The head is relatively narrower than that of V. kaznakowi. The nuchal collar is not expressed, unlike in V. kaznakowi. Upper edge of snout is either rounded or slightly pointed. Head can be either protuberant or flat dorsally, but never impressed like that of V. kaznakowi. Body is thinner and more delicate. Comparative data on pholidosis and size characters are listed in Tables 1-4. Remarks and variability. Vipera dinniki greatly resembles morphologically V. kaznakowi, V. ursini renardi and V. berus. In size V. dinniki is smaller than V. kaznakowi and larger than V. ursini renardi (Table 2). The head normally is slightly protuberant, seldom flat and never as broad as that of V . April 1990 Asiatic Herpetological Research Vol. 3, p. 15 FIG. 6. A Female Vipera dinniki from the upper part of the Laba River (ZDN 17281). kaznakowi. Hence, the nuchal collar is hardly marked. In body proportions it primarily resembles V. berus. Normally the parietals are shorter than the frontal. Three to four lower labials reach the lower jaw scale. Color. Normally Vipera dinniki is not as brightly colored as V. kaznakowi. However specimens with either bright yellow or orange elements occur. Often V. dinniki have greyish, silver greyish or green greyish color. This is never recorded in V. kaznakowi. For all shield-headed vipers a zig-zag shaped dorsal stripe is a common pattern element. In V. dinniki the stripe is often an even broad dark line which runs along the dorsum. This is occasionally seen in V. kaznakowi, where it is usually represented by a row of oblique Vol. 3, p. 16 Asiatic Herpetological Research April 1990 FIG. 7. A male Vipera dinniki from the vicinity of Krasnaya Polyana (Red Meadow), [ZIN 12153]. diametrical spots. The dorsal stripe is separated from the dark sides by lighter lateral stripes. The ventrum is either dark with light spots or light grey. The number of melanistic specimens in populations ranges from 20 to 25%. Complete melanistic individuals of V. dinniki do not have a single light spot in color, unlike melanists of V. kaznakowi. Sexual dimorphism. Maximum body length is greater in females (up to 486 mm), and smaller in males (up to 412 mm). Males have longer tails, characteristically April 1990 Asiatic Herpetological Research Vol. 3, p. 17 FIG. 8. A Vipera dinniki from Lagodekhi, an eastern population (ZIN 13769). thicker at the tail base (Table 2). The number of ventral scales is greater in females. The number of subcaudals is greater in males. Males are more delicately built than females. Sexual dimorphism in color is hardly expressed. Coloration in males is generally brighter and more contrasting. Age variability. Neonate vipers are patterned like adult individuals. However, the general color is normally grey, unlike bright red neonates of Vipera kaznakowi. Only after the third shedding, faint Vol. 3, p. 18 Asiatic Herpetological Research April 1990 coloration typical for the species (yellow, reddish, greenish) emerges in V. dinniki. Color becomes stronger with each subsequent shedding. Maximum color strength is reached by maturity. Melanistic specimens are born with a specific color. They gradually darken with each shedding. By the third year, they acquire a black velvet color. Geographical range and ecology. The species ranges from the Fisht-Oshtenovsky Mountain Range in the west up to Mount Shkhara in the east. The eastern limits of this species distribution has not been worked out and additional eastern localities are probable. The southern distributional limit runs along the dark coniferous highland ecological belt on the southern slope of the Main Caucasus and South Frontal Ridges. The northern limit goes from Mount Shkhara to the west along the crest of the Main Caucasus Ridge up to the Bolshaya (Big) Laba River head where it passes on to a northern macroslope. In the north it occurs on the Peredovoy (Frontal) Ridge and reaches the Fisht-Oshtenovsky Mountain Massive (Orlov and Tuniyev 1986). Research during 1987-1989 showed that Vipera dinniki occurs further to the east than previously known (Fig. 9). The problem of subspecific status of the eastern populations and their interaction with the vipers of the V. ursini complex will be addressed in a future paper. The elevational distribution is generally between 1500 and 3000 m. Occasionally the viper may descend slightly lower. Vipera dinniki is a subalpine montane- meadow species. It tends to be restricted to the upper forest belt, subalpine and alpine meadows, rocky outcroppings and montane moraines. Vipera dinniki can be observed in vegetation associations of Betuletum calamagrostidosum, Pinetum mystillosum subalpinum, Fageto Betuleto-Sorbetum altherbosa-subalpinum, also in Aceretum trantaltherbosum subalpinum. It is associated with rock outcrops interspaced with shrubs of Rhodoretum caucasicus subalpinus, subalpine highland herbs and rock debris. It occurs widely in moraines overgrown with moss, lichens and Thymus. The biotype of V. Kaznakowi is always located in the vicinity of water. Hibernation sites are also located in the immediate vicinity of summer biotypes. All types of rock outcrops are inhabited. Vipera dinniki can be found in limestone, slate, and crystalline outcrops. Climatic conditions within the range of Vipera dinniki are much more severe than those of V. kaznakowi. However, the viper's distribution in severe highlands coincides with the mildest climatic places in this severe area. Vipera dinniki commonly occurs on slopes with a southern or southeastern exposure. For instance, on Mount Aibga, 23 km away from the sea, at an altitude in excess of 1800 m, the mean temperature in January is -0.1°C (Korostelyov 1933; Shkadova 1979). Mean temperature in July is only 13.7°C on Mount Achisko at an altitude of 1750 m (37 km away from the sea). However, great temperature drops never occur in this area, even in winter. As regards to soil freezing, minimum temperature in January is -7°to - 8°C (Selyaninov 1933). Due to solar radiation on slopes with southern and southeastern exposures, the vipers manage to maintain high body temperatures during activity periods. Vipera kaznakowi reaches the edge, but does not penetrate deep into the dark coniferous belt. Vipera dinniki, at its lower limits, reaches the edge but never penetrates the dark coniferous belt. Thus, the dark coniferous belt is a barrier between them. Occasionally, in spots where this belt is either absent or is fragmented, for instance along river valleys, both species occur sympatrically, forming a narrow line with intergrading characters. The valley of the Mzymta River may serve as an example of a site where the two species come in contact. The limited distribution of the species on the northern slope of the Main Caucasus Ridge might be connected with the increase of the mountain's aridity and severe climate towards the east. Xerophytization is observed from west to east in the subsequent change of beech April 1990 Asiatic Herpetological Research Vol. 3, p. 19 |*| V 'ipera darevskii J Vipera dinnik 1 Vipera kaznakowi | Vipera ursini eriwanensis J Vipera ursini renardi FIG. 9. Distribution map of shield-headed vipers in the Caucasus. forests and Abies forests to pine forests, and further east, to steppe areas (Adamyants 1971, Kharadze 1974, Lavrenko 1980; Agakhanyants 1981). This change is noted in amphibians and reptiles. Along with V. dinniki, Lacerta derjugini and Pelodytes caucasicus fall out and more xeric species such as Bufo viridis, Lacerta agilis, and Vipera ursini renardi dominate. Population density is different in various parts of the habitat. It is maximal on rock outcrops and moraines in the subalpine belt of the Caucasus Reserve. In July, in subalpine meadows of the Gertsen Ridge and along the valley of the Bezymyanka (Without name) River, 1700 to 1900 m, we have recorded as many as 5 to 7 vipers per 1 km of a route. At the same time, on an area of 500 m2, we have recorded up to 6 specimens. This is at an elevation of 1800 m on the Aishkha Ridge in a subalpine meadow of Aceretum trantaltherbosum subalpinum. In late June to early July along the Moloshnaya (Milk) and Sumasshedshya (Crazy) rivers of the Mzymta River basin, up to 4 specimens per 100 to 500 m of a route were recorded. In August, on a bank of the high altitude Kardavych Lake, we recorded up to 8 specimens per 300 m of a route. On the Aspidny Ridge, 2000 m above sea level, we recorded 5 to 6 specimens per 300 to 400 m of a route. According to Bozhansky (1979), the density in the subalpine belt is 2 to 6 specimens per hectare. Occasionally, seasonal accumulations of up to 30-40 Vol. 3, p. 20 Asiatic Herpetological Research April 1990 individuals per hectare are possible. In other regions the population densities are much lower. Degradation of localities is due to intensive cattle grazing in subalpine meadows. Overall numbers of both V. kaznakowi and V. dinniki are estimated at some dozens of thousands (Bannikov and Makeyev 1976). Apparently the figure may primarily concern V. dinniki as the number of V. kaznakowi is rather small. Diet. Adult Vipera dinniki eat fewer types of food items than V. kaznakowi. For instance, in the highlands there are viper populations that prey either on lizards or on small mammals as no other food items are available. When field collected vipers were stimulated to regurgitate, the following prey were observed: Apodemus sylvaticus, Microtus majori, Sicista caucasica, fledglings of ground nesting birds such as Anthus spinoletta, and Lacerta caucasica. Immature specimens feed on Orthoptera and small lizards. In captivity V. dinniki takes similar food items as V. kaznakowi. It is often observed that a bitten mouse or a field-vole makes a few desparate leaps, falls into a deep crack between rocks and dies there. The viper skillfully catches its prey between its teeth, drags it out onto a level spot, then drops it, examines it from all sides and having found the prey's head, swallows it. This behavior is typical for other rock dwelling vipers from the Caucasus, such as V. raddei, and V. ammodytes. Shedding. Overall, shedding is observed in June and in late August to early September. Neonate vipers shed during the first hours after birth. Some days later they enter hibernation. Reproduction. Copulation occurs in late April to May (Bozhansky 1984). Birth of neonates on the northern slope of the Main Caucasus Ridge occurs in August. On the southern slope it occurs throughout September. Bozhansky (1983) found two groups of specimens during the summer. One group containing males and females which do not reproduce that year, and the other group containing gravid females. Bozhansky suggests that in montane conditions, female Vipera dinniki have a reproductive cycle of many years. Development of the young. In late July to early August viper embryos reach 70 mm in length (Orlova 1973). Mean body length of the neonates is 131.0 mm, tail length is 14.8 mm, mean weight is 3.1 g (N=28). In the highlands almost right after birth the vipers enter hibernation. Unlike newly born Vipera kaznakowi, they do not feed until the next season. By the third year the vipers become mature. Territoriality. Gravid females tend to move on small areas ranging from 1-4 to 51 m2 Their individual places to a great extent (up to 98%) overlap. Within their sites the vipers actively utilize only 2 or 3 places where they may be encountered at different times and under various weather conditions. During the morning the vipers can be found in places with a western exposure. Normally, this place is on a rock surface shaded by a bush. The vipers utilize direct sun for a short time, leaving a part of the body under the sun and a part under a half shaded area. From 1 100-1200 the vipers retreat to burrows and normally appear after 1500 hours. In gloomy weather they lie flat on a rock during the entire period of activity. This increases the surface of the body contiguity with the rock. On rare hot days, after midday the vipers never reappear on the surface. Males and non-reproducing females emerge much more seldom, normally after they have taken food. Bozhansky (1984) observed some specimens using the same territories for three seasons. Insolation is of great importance in the viper's life. Gravid females take food rarely and two months before giving birth, they stop eating. Territory utilization is entirely dependent on optimal insolation patterns. Absence of aggressiveness allows all adult females in a territorial group to utilize the warmest microhabitats (Bozhansky 1983, 1984). These data were collected by Bozhansky during three summers at an elevation of 2000 m along the border between the forest and subalpine belts on Aishkha Ridge, the right ridge of April 1990 Asiatic Herpetological Research Vol. 3, p. 21 the Mzymta River valley, the western Caucasus. It is amazing that Vipera kaznakowi from the low altitudes of the coast have considerably larger individual sites. The females do not have a multi-year cycle and never stop taking food while gravid. The adaptations in the highland V. dinniki are very important in relatively low temperatures, rainy periods, and when it suddenly becomes cold. Seasonal dynamics in populations. Biological monitoring in highland populations from the Caucasus Reserve testify to their climax state (the collections of the Zoological Museum of the Moscow State University and the Zoological Institute, the USSR Academy of Sciences were adopted as a zero counting point). Seasonal dynamics within climax populations is monotypical. At the beginning of spring, males emerge first. Then the percentage of female occurrences gradually increases. By early-mid June the sex ratio is 1:1. In mid-late summer males are seen on the surface less frequently than females. During the period of pre- hibernation at the end of summer, males become more common again. From late August to mid September, 80% of the snakes observed are gravid females and new born individuals. The later can be observed on the surface even when adults have entered hibernation. During this period, the location of new bom individuals does not necessarily correlate with rock outcrops which are used as winter shelters. The newborn snakes migrate much more than adult snakes. Seasonal and daily activity. In the spring vipers emerge from mid April to May when mean day temperature on the surface reaches 11°C. Seasonal activity is dependent on weather conditions. In the highlands of the Caucasus Reserve the first snow usually falls in the second half of September and snow is present until May. Snow cover is 7-8 m thick. Hence, Vipera dinniki inhabiting the subalpine belt enter hibernation in the second half of September. At elevations of 1800 to 2400 m morning activity is hardly expressed, whereas evening activity is shifted from 1700 to 2000 hrs. In gloomy weather snakes are active throughout daylight at temperatures higher than 10°C. At 8°C vipers are not seen on the surface. Gravid females are seen on the surface even in drizzling rain. Even if the temperature is 10°C, body temperature in vipers is 30°C, and cloacal temperature is 26-28°C due to accumulation of warmth from solar radiation. This is an important thermal adaptation of a number of highland reptiles. Sympatric species. Throughout nearly the entire range Vipera dinniki is sympatric with Lacerta caucasica, L. saxicola, Anguis fragilis, and Coronella austrica. Vipera darevskii Vedmederja, Orlov, andTuniyev, 1986 (Fig. 10, 11, 12, 13) Chronology of species description Vipera kaznakowi dinniki - Darevsky, 1956:128. Vipera kaznakowi darevskii - Vedmederja, 1984:8, nomen nudum. The English common name is Darevsky's Viper. Holotype: ZIN 19934, an adult female from Legli Mountain, the Mokrye (Wet) Mountains, Gukasynsky region, Armenia. The specimen was collected in June, 1980 by I. S. Darevsky. The specimen is preserved at the Zoological Institute, the USSR Academy of Sciences, Leningrad (Fig. 10). Paratypes: ZIN 16546 a and b. The specimens were collected May 28, 1954; ZIN 17545, the specimen was collected August 6, 1955; ZIN 19935, the specimen was collected June, 1980 by I. S. Darevsky (Fig. 11). Holotype description: Body length is 421 mm, head included. Tail length is 46 mm. A female. Head is slightly impressed dorsally. Lateral snout edges are slightly pointed. Anterior edge of snout is slightly rounded. Rostral is narrow. Frontal is Vol. 3, p. 22 Asiatic Herpetological Research April 1990 FIG. 10. Holotype of Vipera darevskii (ZIN 19934). a. dorsal view, b. ventral view, c. head. narrow. Its length is equivalent to 1.66 times its width. Parietals are slightly longer than the frontal. The frontal is separated from supraoculars, which protrude over lateral edge of the snout, by a row of three scales. Prefrontal is triangular, three times shorter than the frontal, and is separated from the rostral by 2 scale rows. Nostril is cut through in the lower part of the nasal. The latter is separated from the rostral by a broad scale. Upper labials and lower labials are both 9 to the right and 8 to the left. There are 5 rows of throat scales. Around the center of the body there are 21 scale rows with strongly expressed keels, except for two scale rows on both sides adjacent to the ventrals, which are smooth. The number of ventrals is 138. There are 25 pairs of subcaudals. The color is yellowish grey. A zig-zag shaped brown stripe runs along the dorsum. At the center of the body its width is nearly 8 mm. A row of hardly conspicuous spots is present laterally. The spots merge into a light brown stripe. Dorsally, the head has light yellowish spots along the edges of the frontal, parietals and lower oculars with yellowish temporals. Ventrum is blackish marked by light contours of ventrals (Fig. 10). Paratypes: Morphological characters of 8 paratypes are listed in the tables 1-4. General color resembles that of the holotype, except for 2 specimens. In the latter the dorsal stripe is interrupted in the anterior part of the body (Fig. 11). Diagnosis: This viper is not large. SVL reaches 460 to 489 mm. Head is slightly impressed dorsally, covered by big scales. Lateral edges of snout are slightly pointed. Anterior snout edge is rounded. Nostril is cut through in lower part of the nasal. Head is narrow, hardly separated from body. Remarks and variability. Morphologically, Vipera darevskii occupies an intermediate position between V. kaznakowi and V. ursini eriwanensis. To be more precise, morphologically V. darevskii occupies a middle position between the two species of the "Vipera kaznakowi " complex (V. kaznakowi and V. dinniki ), on the one hand, and the steppe vipers from the "V. ursini " complex, on the other hand. This viper is considerably smaller in body size than V. kaznakowi. The head is narrower and the nuchal collar is less expressed. It differs April 1990 Asiatic Herpetological Research Vol. 3, p. 23 mrf ■■-■ •**<*£*♦>« ^T1 Jr.- •>"♦.:..'.- .Ti>v» Y»£*^a ■ *. . FIG. 11. The paratypes of Vipera darevskii (ZIN 19935). a., b. dorsal view, c, d. ventral view. from V. ursini eriwanensis by greater head height and much less pointed upper anterior snout edge. Pholidosis data are listed in Tables 1-3. Coloration is yellowish or yellowish grey, which never occurs in Vipera ursini. Along the dorsum a contrasting zig-zag shaped brown stripe is present. The ventrum is dark grey, speckled black and white. Of the small number of known specimens, no melanistic individuals have been found. The coloration of V. darevskii is evidently more stable than that of the polymorphic and motley colored V. kaznakowi and V. dinniki (Fig. 13). Yellow-greyish hues are prevalent. The pattern is more homogeneous. The neck transition is hardly expressed, like in V. dinniki. It differs from V. ursini eriwanensis by 1) a relatively high head, 2) yellowish general coloration, 3) clear contrasted pattern, and 4) special pholidosis. Sexual dimorphism. Maximum body size is greater in females than males (Table 2). Males have longer tails. Number of ventral scales is greater in females (Fig. 14). Number of subcaudal scales is greater in males. Sexual dimorphism in color is not recorded. Vol. 3, p. 24 Asiatic Herpetological Research April 1990 h+b or th+b th+b Uh+b t>h+b Preraaxillary fontanelle present absent present present absent absent present absent absent Basibranchial radii present absent present present present absent present absent absent Vomerine teeth (o=posilion of internal oahs) 0/~- ""N.0 or O'l po o/"^\o 0^-* ^0 O^-s ^v° 0/"\ p 0 O^v ^o o ^ /» 0 °"V/-° Vomer vs. parasphenoid sutured broadly overlaps sutured sutured overlaps overlaps sutured overlaps sutured, may overlap tip Lungs reduced present reduced absent reduced present reduced present present Chromosome number (2n) 62, 68 56, 58, 60 64 78 64 58 66 62 40 Larval duration 2-3 years 1 year 2-3 years 2-3 years ? 1 year 2-3 years 1 year 1 year or more Habitat in non- breeding period water land water land water land land land water or land 'Batrachuperus includes Paradaclylodon, Pachyhynobius includes Xenobius. and Randon includes Pseudohynobius. According to some authorities Pachypataminus is a synonym of Hynobms (Nishio el al.. 1987). locality, the number of costal grooves varied from 10 to 12 (Chang, 1933). Tail Proportions. Cai noted that the tail of yiwuensis is compressed, with fin folds especially distinct in the males, whereas Gunther stated that chinensis was without a tail crest. Gunther did not mention the sexes of his two specimens, which are both females (see section 1), and this was not known to Cai. Thus, we believe that the differences in the tail may be due to sexual dimorphism and, in addition, perhaps also to further elaboration of the crests during the breeding season. Color Pattern. Giinther's description was based on specimens that had been preserved for some time. Moreover, Cai noted that the color pattern changes during the breeding season, when it becomes lighter and mostly green in color. Nevertheless, our comparisons of the two forms do not reveal any striking differences. Further details of our comparison between the types of chinensis and Cai's specimens of yiwuensis are given in Zhao and Adler (1989). In summary, we believe that all of these specimens represent a single taxon which, because of the priority of Giinther's name, must be designated Hynobius chinensis. 3. Generic Status of Hynobius retardatus Dunn, 1923 The identity of this Japanese species was first recognized by E. R. Dunn, who briefly characterized it (Dunn, 1923a) and soon described it in more detail (Dunn, 1923b). The animal itself had been known to Japanese biologists since at least 1907 under the names Hynobius fuscus, H. lichenatus, and H. nigrescens. These names, however, are now known to be properly applied to species found elsewhere April 1990 Asiatic Herpetological Research Vol. 3, p. 41 in Japan (fuscus is a synonym of nigrescens). Dunn (1923a) noted that H. retardatus was a "well-marked species" and, as knowledge of this animal increased in succeeding years, its distinctiveness from other Hynobius and hynobiids generally became more apparent. In 1932, Makino reported that H. retardatus has a diploid somatic chromosome number of 40, but since the comparable numbers for most other hynobiids were not known at that time, the full significance of this very low number was not recognized. In fact, retardatus has by far the lowest number of chromosomes in the family Hynobiidae (Table 1). Only in 1943, with Sato's magnificent review of Japanese salamanders, in which he included a special comparative study of their chromosomes, could the matter be properly evaluated, and Sato himself (1943, p. 489) suggested that retardatus might be worthy of generic rank. Unfortunately, Sato's premature death in August 1945, during the atomic bombing of Hiroshima (for a biography of Sato, see Adler, 1989), prevented him from pursuing this matter. In subsequent years, additional data on anatomy, karyology, and biology have accumulated which further support the separation of retardatus into a new genus, which we name: Satobius, new genus Type Species: Hynobius retardatus Dunn, 1923a. Content: A single species. Diagnosis: A genus of hynobiid salamanders (family Hynobiidae) characterized by very long limbs and toes (tips of digits of adpressed limbs overlap + 1 to +4 intercostal spaces in adults); a very long tail (in adults, 100 to 118% of head and body length combined); a long neck and small head; no premaxillary fontanelle or basi branchial radii; two short series of vomerine teeth arranged in transverse arcs between the internal nares; vomer sutured to anterior end of parasphenoid; lungs present; diploid (2«) chromosome number of 40; larval duration of one year or more (neoteny sometimes occurs); and both terrestrial and aquatic habits in adults during non-breeding season. These characteristics are discussed below and, for easy access, are tabulated for each of the nine genera of hynobiid salamanders (Table 1). Costal Grooves. These vertical grooves on the side of the body correspond to the position of ribs and, thus, to trunk vertebrae; generally, the number of costal grooves that can be counted is one less than the number of trunk vertebrae. The typical number of costal grooves in 5. retardatus is 11. In the Japanese species of Hynobius the modal numbers of grooves range from 11 to 13 (Misawa, 1989) and some mainland species have as many as 14 (Dunn, 1923b). For species in other hynobiid genera, the modal numbers of costal grooves range from 10 to 14. Adpressed Limbs. As a relative measure of limb length, the minimum distance between the tips of the digits is determined with the limbs adpressed along the sides of the body. Distances are then measured in intercostal spaces, the fleshy folds between adjacent costal grooves. In metamorphosed adults of 5 . retardatus the intercostal distance between the digits of adpressed limbs is +1 to +4; that is, because the limbs and toes are very long, the digits actually overlap by one to four intercostal spaces. Proportionately, these are the longest limbs found in any member of the entire family. In species of Hynobius, this measurement ranges from 5 to +3; among hynobiids other than Hynobius and Satobius, the longest limbs are found in Pachypalaminus (-2 to 0), but the toes are shorter than those of Satobius. Tail. The tail of 5. retardatus, as measured from posterior angle of the vent, is longer than the combined measurements of head plus body length. In adults, tail length varies from 100 to 118% of head-body Vol. 3, p. 42 Asiatic Herpetological Research April 1990 length, whereas in all other hynobiids it is shorter except in Ranodon (about 100- 120% of head-body length) and in Onychodactylus (about 100-115%). The tail lengths of Hynobius species are significantly shorter than the head-body length (60-80%); only in the adults of H. nigrescens does the tail occasionally equal head-body in length. Apparently in all hynobiids, tail length relative to head-body increases with overall size, so the numbers given here are all taken, for comparative purposes, from large adults. Vomer and Vomerine Teeth. The paired vomer bones (or prevomers for those who deny homology to the vomer of mammals) of the palate bear teeth near their posterior edges. The relationship between the vomers and the parasphenoids lying posterior to them varies among hynobiid genera (Table 1). In S. retardatus, the posterior edge of the vomer is sutured to the parasphenoid and overlaps very little, if at all in some specimens, onto the palatal surface of the parasphenoid (Inukai, 1932; Sato, 1943). The two series of vomerine teeth extend broadly between the internal nares in two slight arcs which nearly meet at the midline; the overall length of this patch of teeth along the midline is about 30% of the width of the series. The general pattern of these teeth is like those in the genera Onychodactylus and Ranodon, and quite unlike that in Hynobius where the vomers (and the vomerine teeth on their posterior surface) extend onto the parasphenoids to a degree varying from species to species. In some Hynobius this overlap is small (e.g., H. leechii and lichenatus) but in most it extends onto the palatal surface for from one-third to as much as one-half the length of the parasphenoids (e.g., H. formosanus and sonani) (Sato, 1943). Since the vomerine teeth are located on this edge of the vomer, in these Hynobius the vomerine series has a pattern wholly unlike that in retardatus, beginning at the nares and extending far posteriorly on the palate in the shape of a lyre (see Table 1). In H. formosanus, for example, the length of the vomerine teeth along the midline of the palate is fully twice the total width of the two series (versus 30% in 5. retardatus). Chromosomes. Makino (1932) was the first to report that S. retardatus has a diploid chromosome number of 40, as confirmed by others (Azumi and Sasaki, 1971; Morescalchi, 1975). No other hynobiid is known to have fewer than 56. In Japanese species of Hynobius, the so- called pond-type species are 2n=5 6 (retardatus is a pond breeder) and the mountain brook types are 2/i=58 (except H. okiensis where 2n=56); H. kimurai is ordinarily 2n=56, except in one population (2n=60) which may represent a separate species (Morescalchi, 1975; Morescalchi et al., 1979; Ikebe et al., 1989). The Korean H. leechii is 2n=56 (Makino, 1934), but the chromosome number is not known for any of the Chinese species of Hynobius. Recent studies by Japanese and Italian workers show that the karyology of S. retardatus is even more different from that of Hynobius species than the low number of chromosomes would suggest. The combined lengths of the chromosomes in the genomes of Hynobius species and in retardatus are nearly equal at the same degree of condensation, and the amount of nuclear DNA is also approximately equal (Morescalchi, 1975). Despite these similarities, there are important differences that set retardatus apart from species of Hynobius. Kuro-o et al. (1987, as modified in 1989) were able to compare chromosome pairs in four Hynobius (3 Japanese and 1 Korean species) and retardatus, using the R-banding technique (RBG method), allowing them to identify 18 of 28 pairs in Hynobius and 16 of 20 in retardatus. Based on this analysis, chromosome pairs 2 and 8 of retardatus are not at all represented in the genomes of these Hynobius species, whereas pairs 2, 12, 20, and 22, found in all four Hynobius, were lacking altogether in that of retardatus; other chromosome pairs were completely (11 pairs) or partially (3 pairs) homeologous. To summarize, among these four Hynobius species homeologies totalled about 90%, but this value fell to 65% when retardatus was included. DNA April 1990 Asiatic Herpetological Research Vol. 3, p. 43 analysis was also performed by these authors, using highly-repetitive DNAs as probes on Southern blot hybridization, which showed that retardatus was distinctively different from the five Hynobius species tested (4 Japanese and 1 Korean). Breeding Biology. The reproductive biology of 5. retardatus has been studied in detail (Sasaki, 1924; Makino, 1933; Sato, 1989 and references cited therein). The larval period is normally less than a year, but at high elevations can take more than one year. Neotenous individuals are well known (Sasaki, 1924). Adults breed in ponds and are terrestrial during non- reproductive periods. However, unlike species of Hynobius which remain on land while not breeding, retardatus often visits the water during non-breeding periods (Sasaki, 1924). Distribution. Satobius retardatus is found only in Hokkaido, the northernmost of the main Japanese islands. No Hynobius are known from Hokkaido and, among hynobiids, only Salamandrella keyserlingii is found there. Relationships. Historically, 5. retardatus has been said to be most closely related to H. nigrescens of neighboring Honshu Island, Japan (Sato, 1943). Indeed, the two species are superficially similar in that, as adults, both are blackish in color with little pattern and both possess relatively long tails. However, on closer examination, these two species are fundamentally different. H. nigrescens is a large-headed and short-necked species, as are all Hynobius, when compared with S. retardatus. The skull of nigrescens, in particular the vomer-parasphenoid relationship and the pattern of vomerine teeth, are typical of Hynobius and unlike the situation in Satobius. Furthermore, the number of chromosomes and details of chromosome structure, based on both C- banding and R-banding methods, are distinctly different. In short, we believe that the similarities between nigrescens and retardatus are due to convergence and do not reflect close phylogenetic relationship. Several of the diagnostic characteristics of S. retardatus are similar to those of members of the R a no don group of hynobiids, which includes Batrachuperus, Liua, Onychodactylus, and Ranodon (Zhao and Hu, 1984). The suturing of the vomer and parasphenoid, the pattern of vomerine teeth, and the relative lengths of limbs and tail in retardatus are similar to the condition in one or more members of the Ranodon group, but in most other respects Satobius is clearly a member of the Hynobius group of genera: Hynobius, Pachypalaminus (synonymized with Hynobius by Nishio et al., 1987) and Salamandrella (for discussion of these two groups of genera, see Zhao and Hu, 1984). Within this group, the closest living relative of retardatus may be H. leechii of Korea and northeastern China, according to their chromosome structure; indeed, as Ikebe et al. (1989) point out, based on C-banding patterns, retardatus is more similar to the northernmost populations of leechii rather than those in the southern part of the Korean peninsula. We suggest that the ancestral stock leading to S. retardatus was derived from a Hynobius-likc ancestor and arrived in Japan from the mainland very early, before the formation of the Tsugaru Strait which later isolated Hokkaido from the southern Japanese islands. Satobius differentiated in Japan but was later excluded from the more southern Japanese islands by a later invasion of hynobiids from the mainland. Thus isolated by the Tsugaru barrier, Satobius further differentiated, yet retained some of its primitive characters, including a few today found only in the Ranodon group. The long isolation from Hynobius has led also to the chromosomal differentiation described earlier (although retaining some similarities to H. leechii on the mainland). The great reduction in chromosome number in retardatus may be due in part to fusions, since the total composite lengths and DNA content are approximately the same as for the Hynobius genome (Kuro-o et al., 1989). However, in view of the many differences based on the C-banding and R- Vol. 3, p. 44 Asiatic Herpetological Research April 1990 banding studies mentioned above, the karyotype of retardatus cannot be explained by a simple mechanism of fusions alone. Etymology. We take great pleasure in naming this new genus for Ikio Sato (1902- 1945), who first recognized its distinctiveness from Hynobius. The name Satobius is derived in part from the stem of Hynobius (hynis [or hynnis], Greek for plowshare, the cutting part of a plow, and bios, life). Acknowledgments Our research has been supported by the USA National Academy of Sciences (Committee on Scholarly Exchange with the People's Republic of China). We thank Barry Clarke (British Museum [Nat. Hist.]) and Robert C. Drewes and Jens V. Vindum (California Academy of Sciences) for the loan of specimens. We thank Satoshi Amagai (Cornell University) for some translations of articles in Japanese. Literature Cited ADLER, K. (ed.) 1989. Contributions to the History of Herpetology. Society for the Study of Amphibians and Reptiles, Contributions to Herpetology 5:1-202. AZUMI, J. AND M. SASAKI. 1971. Karyotypes of Hynobius retardatus Dunn and Hynobius nigrescens Stejneger. Chromosome Information Service 12:31-32. BORING, A. M. AND T.-K. CHANG. 1933. The distribution of the Amphibia of Chekiang province. Peking Natural History Bulletin 8:63-74. CAI, C.-M. 1985. [A survey of tailed amphibians of Zhejiang, with description of a new species of Hynobius]. Acta Herpetologica Sinica 4:109-114. (In Chinese). CAI, M.-Z., J. ZHANG, AND D.-J. LIN. 1985. [Preliminary observation on the embryonic development of Hynobius chinensis Guenther]. Acta Herpetologica Sinica 4:177-180. (In Chinese). CHANG, T.-K. 1933. Two new amphibian records from Chekiang. Peking Natural History Bulletin 8:75-80. DUNN, E. R. 1923a. New species of Hynobius from Japan. Proceedings of the California Academy of Sciences, ser. 4, 12:27-29. DUNN, E. R. 1923b. The salamanders of the family Hynobiidae. Proceedings of the American Academy of Arts and Sciences 58:445-523. GONTHER, A. 1889. Third contribution to our knowledge of reptiles and fishes from the Upper Yangtsze-Kiang. Annals and Magazine of Natural History, ser. 6, 4:218-229. IKEBE, C, M. KURO-O, T. OHHASHI, H. YAMADA, K.-I. AOKI, AND S.-I. KOHNO. 1989. Evolution of chromosome 10 in ten pond-type Hynobius from Korea and Japan, with comments on phylogenetic relationships. Pp. 85-90. In M. Matsui et al. (eds.), Current Herpetology in East Asia. Herpetological Society of Japan, Kyoto. INUKAI, T. 1932. Urodelenarten aus Nordjapan mit besonderer Beriicksichtigung der Morphologie des Schadels. Journal of the Faculty of Science, Hokkaido Imperial University ser. 6, 1:191-217. KURO-O, M., C. IKEBE, AND S. KOHNO. 1987. Cytogenetic studies of Hynobiidae (Urodela). VI. R-banding patterns in five pond-type Hynobius from Korea and Japan. Cytogenetics and Cell Genetics 44:69-75. KURO-O, M., C. IKEBE, R. KATAKURA, Y. IZUMISAWA, AND S.-I. KOHNO. 1989. Analyses of phylogenetic relationships in the genus Hynobius by means of chromosome banding and Southern blot hybridization. Pp. 91-94. In M. Matsui et al. (eds.), Current Herpetology in East Asia. Herpetological Society of Japan, Kyoto. MAKINO, S. 1932. The chromosome number in some salamanders from northern Japan. Journal of the Faculty of Science, Hokkaido Imperial University ser. 6 (Zool.) 2:97-108. MAKINO, S. 1933. [Several ecological observations on Hynobius retardatus Dunn]. Shokubutsu oyobi Dobutsu (Botany and Zoology), Tokyo 1:1 136-1 142. (In Japanese). MAKINO, S. 1934. The chromosomes of Hynobius leechii and H. nebulosus. April 1990 Asiatic Herpetological Research Vol. 3, p. 45 Transactions of the Sapporo Natural History Society 13:351-354. MIS AW A, Y. 1989. The method of counting costal grooves. Pp. 129-134. In M. Matsui et al. (eds.), Current Herpetology in East Asia. Herpetogical Society of Japan, Kyoto. MORESCALCHI, A. 1975. Chromosome evolution in the caudate Amphibia. Pp. 339- 387. In T. Dobzhansky et al. (eds.), Evolutionary Biology, vol. 8. Plenum Press, New York. MORESCALCHI, A., G. ODIERNA, AND E. OLMO. 1979. Karyology of the primitive salamanders, family Hynobiidae. Experientia 35:1434-1436. NISHIO, K., M. MATSUI, AND M. TASUMI. 1987. The lacrimal bone in salamanders of the genera Hynobius and Pachypalaminus: a reexamination of its taxonomic significance. Monitore Zoologico Italiano, new ser. 21:307- 315. POPE, C. H. 1931. Notes on amphibians from Fukien, Hainan, and other parts of China. Bulletin of the American Museum of Natural History 61:397-611. POPE, C. H. AND A. M. BORING. 1940. A survey of Chinese Amphibia. Peking Natural History Bulletin 15:13-86. PRATT, A. E. 1892. To the Snows of Tibet Through China. Longmans, Green, and Co., London. 268 pp. SASAKI, M. 1924. On a Japanese salamander, in Lake Kuttarush, which propagates like the axolotl. Journal of the College of Agriculture, Hokkaido Imperial University 15:1-36 ("Hynobius lichenatus" = Satobius retardatus). SATO, I. 1943. [A Monograph of the Tailed Batrachians of Japan]. Nippon Shuppan-sha, Osaka, 520, 7 pages, 31 color plates. (In Japanese). SATO, T. 1989. Breeding environment and spawning of a salamander, Hynobius retardatus, at the foot of Hidaka Mountains, Hokkaido, Japan. Pp. 292-304. In M. Matsui et al. (eds.), Current Herpetology in East Asia. Herpetological Society of Japan, Kyoto. ZHAO, E.-M. AND Q.-X. HU. 1983. [Taxonomy and evolution of Hynobiidae in western China, with description of a new genus.] Acta Herpetol. Sinica 2(2):29-35 (In Chinese). (English translation published in Zhao et al., 1988, Studies on Chinese Salamanders. Society for the Study of Amphibians and Reptiles, Contributions to Herpetology 4: ix, 1-67). ZHAO, E.-M. AND Q.-X. HU. 1984. [Studies on Chinese Tailed Amphibians.] Sichuan Sci. Tech. Publ. House, Chengdu, (3), 68 pages (In Chinese). (English translation published in Zhao et al., 1988, Studies on Chinese Salamanders. Society for the Study of Amphibians and Reptiles, Contributions to Herpetology 4: ix, 1-67). ZHAO, E.-M. AND K. ADLER. 1989. [Hynobius chinensis, a re-description on its 100th anniversary]. Sichuan Journal of Zoology 8(2): 18-20, 1 plate. (In Chinese). April 1990 Asiatic Herpetological Research Vol.3, pp. 46-51 Relationships Between Serum T4, T3, Cortisol and the Metabolism of Chemical Energy Sources in the Cobra During Pre-hibernation, Hibernation and Post-hibernation RUIMIN WU1 AND JlE HUANG1 ^Department of Biology, Fujian Medical College, Fujian, China Abstract. -This study analyzed the variation of the serum thyroxine (T4), triiodothyronine (T3), and Cortisol in the cobra (Naja naja Linnaeus) in pre-hibernation, hibernation, and post-hibernation. The variations were compared with the changes of indexes of cobra metabolism in those three periods, such as oxygen consumption, serum glucose, serum triglyceride, glycogen content of liver, and triglyceride content of fat bodies. From pre-hibernation to hibernation, the change of metabolic indexes of the cobra indicated that the metabolic rate tends to decline because of the very low levels of the three serum hormones in pre- hibernation and falling temperature. Low temperature during hibernation limited the activities of the three serum hormones which were rising at higher levels during hibernation. From hibernation to post hibernation, the rising temperature allowed the three serum hormones to increase markedly, stimulating metabolism gradually. Therefore the metabolic rate of the cobra during post-hibernation tended to rise again. The metabolic indexes of the cobra in post-hibernation showed a significant increase in metabolic rate, which had close relation with the high levels of the three serum hormones. The indexes of metabolism of the cobra indicate that it is hepatic glycogen and not fat that hibernating cobras use as their main energy source. Key Words: Reptilia, Serpentes, Elapidae, Naja naja, cobra, thyroxine, triiodothyronine, Cortisol, metabolism, energy metabolites, chemical energy sources. Introduction Hibernation is an adaptive strategy of ectotherms for keeping out of the cold during winter. The metabolic rate of the hibernating reptiles is apparently different from that of active ones, as a result of physiological adaptation. The endocrine system and its close relation to metabolism may play an important role in physiological adaptation. Maher (1965) pointed out that the maximum metabolic response in lizards to thyroxine occurred at 3°C; Wilhoft (1966) found that injections of thyroxine in lizards increased their metabolism. Musucchia (1984) suggested that the injection of Cortisol into hamsters during hibernation or hypothermia could achieve the same result as the injection of glucose, which might prolong the survival of hamsters under the same conditions. We consider that the metabolism of energy metabolites in hibernating snakes must have a close relation to the control of serum thyroxine (T4), triiodothyronine (T3), and Cortisol. This study may provide a way to find the mechanism of snake hibernation and guidance for maintaining snakes. Methods This study was carried out from September 1987 to May 1988. Cobras Naja naja (Linnaeus) used as experimental animals were captured at Changlou, Fujian Province, China. We studied serum T4, T3, and Cortisol, and related indexes of metabolism in cobras during pre- hibernation, hibernation, and post- hibernation. Cobra groups: 12 adult cobras (6 males, 6 females) were captured in May 1987, weighing 176-273 g. These cobras had been fed in our college snake garden for 4 months before the experiment. On September 9, 1987, 6 cobras were taken out of the snake garden and tested as a pre- hibernation group (Pre G). On October 2, 1987, the remaining 6 cobras were transferred from the snake garden into a dark cement pool with a perforated metal window above. The temperature in the pool was similar to the atmosphere. On 1990 by Asiatic Herpetological Research April 1990 Asiatic Herpetological Research Vol. 3, p. 47 February 10, 1988, 3 cobras were taken out from the pool and tested as a hibernation group (H G), and on May 7, the last three cobras were tested as a post- hibernation group (Post G). Determination of oxygen consumption: The oxygen consumption of each group was determined using the method of Dong et al. (1986) corrected by the authors. Heart beats and respiratory state: Heart rate and respiratory state of each group were measured by routine methods. Analysis of three hormones: Blood was drawn from the posterior vena cava. Serum T4, T3, and Cortisol of each group were analyzed by radioimmunoassay by means of kits produced by the Institute of Shanghai Biologicals. Each sample was analyzed twice. Analysis of serum energy metabolites: Serum glucose and serum triglycerides of each group of snakes were analyzed by clinical chemistry methods, and each sample was analyzed twice. Analysis of hepatic glycogen: Fresh snake liver of each group was weighed and cut into pieces. After having been in boiling water for 5 minutes, the liver pieces were homogenized. The homogenate was in boiling water for another twenty minutes and filtered immediately. The filtered homogenate was mixed with 95% alcohol - A. R. (mean analysis reagent) to twice the volume and kept at room temperature for ten minutes before it was centrifuged for twenty minutes at 3000 rpm. The supernate was drawn out, the sediment (glycogen) soluble in hot water was estimated by the following calculation: Hepatic glycogen (g/Kg(BW)) = total glycogen content in liver/body weight (BW) Analysis of fat body triglycerides: Fresh snake fat bodies of each group was weighed and a small weighed part was homogenized in a fixed volume of n- heptane (C7H16). The fat tissue was extracted in this way three times. Triglycerides of fat bodies extracted in n- heptane were analyzed by clinical chemical methods. Triglycerides of fat body(g/Kg(BW)) = triglyceride content in fat body/body weight (BW) Results 1. Variations of the three serum hormones during Pre G, HG and Post G The levels of serum T4, T3, and Cortisol of Pre G were the lowest in the three cobra groups. The level of serum Cortisol of HG was markedly lower than that of post G. The levels of T4 and T3 of HG were not distincdy different from those of Post G (Table 1). 2. Variations of the oxygen consumption and energy metabolites during Pre G, HG and Post G The oxygen consumption of the HG group was significantly the lowest of the three cobra groups. The oxygen consumption of Pre G was not significantly different from that of Post G. The contents of serum glucose of both TABLE 1. Experimental results of the three serum hormones and t tests. Groups Nor N' Serum T3 (ng/ml) Serum T4 (ng/ml) Serum cortiso 1 (ng/ml) Pre-hibemation (A, 22.9°C) 6 3 3 7 4 7 0.2303 0.3607 0.5467 3.967* 1.767 4.703* * 0* 0.700 1.233 4.399* * 1.153 4.792* * 369.0 1509 6953 12.35* 2.808* * 5.168* * Hibernation (B, 13°C) Post-hibernation (C. 24.9°C) tAB value *BC value tAC value * Too small to measure. **Significant at 0.05. Vol. 3, p. 48 Asiatic Herpetological Research April 1990 TABLE 2. Experimental results of oxygen consumption by energy substances, and t tests. Groups Nor NT Oxygen consumption (ml 02/hr • kg(BW*)) Serum glucose (mg%) Serum triglyceride (mg%) Hepatic glycogen (g/kg(BW)) Fat body triglycerides (g/kg(BW)) Pre hibernation 6 127.6 193.9 102.9 47.99 10.29 (A, 22.9°C) Hibernation 3 54.9 50.40 18.73 0.3825 16.04 (B, 13°C) Post-hibernation 3 788.6 143.5 24.4 0.03084 17.61 (C, 24.9°Q *AB 7 9.731** 7.742** 6.542** 2.680** 1.108 tBC 4 2.791** 5.473** 0.679 2.985** 0.1615 tAC 7 2.316 2.220 2.165 2.699** 1.736 BW = Body Weight. **P < 0.05 Pre G and Post G were apparently higher than that of HG. But the content of serum glucose of Pre G was not significantly different from that of Post G. The hepatic glycogen content of the cobras appeared to decline during hibernation. The hepatic glycogen content of HG was markedly less than that of Pre G, and the hepatic glycogen of Post G was less than that of HG. Similar to the variation of serum glucose, the content of serum triglycerides of Pre G was higher than that of HG, but there were not significant differences between the serum triglycerides of post G and that of HG, and between that of Pre G and that of Post G. The triglyceride contents of fat bodies in the three cobra groups did not have any statistical differences from one another (Table 2). Discussion 1. Variations of the oxygen consumption and energy metabolites in the cobras during the three periods The contents of serum glucose and triglycerides of Pre G were relatively high due to the cobras active intake before entering hibernation, which was advantageous for the storage of energy metabolites for overwintering. The oxygen consumption was large during this period, but the contents of glucose and triglycerides of HG were apparently at low level. Oxygen consumption decreased markedly in this period. The hepatic glycogen content during hibernation became 99% less than during pre-hibernation. The triglyceride content of fat bodies of the HG group seemed to rise a little, perhaps due to some triglyceride synthesis at the beginning of hibernation. These results indicate that the metabolic rate of cobras during hibernation remains at a low level and oxygen consumption, serum glucose, and serum triglyceride content fell markedly, and that hepatic glycogen is used as the main energy source instead of fat body triglycerides. In post- hibernation, the serum glucose content rose significantly and the content of hepatic glycogen decreased by 92 percent less than that of HG, showing that cobras had used hepatic glycogen to the greatest degree for evoking their activities. This variation provided a good situation for cobras to come out from hibernation, and at the same time, oxygen consumption increased gready. The rising serum triglycerides of post G also meant that the cobras' fatty metabolism began to become active. April 1990 Asiatic Herpetological Research Vol. 3, p. 49 Compared with that of pre G, the oxygen consumption of post G was relatively large, but the serum glucose and serum triglyceride content of post G were relatively low. The reasons may be: 1) that the pre G cobras were not fasting so that their contents of serum glucose and triglycerides were relatively high, 2) that the post G cobras had so little hepatic glycogen as to be unable to raise their serum glucose and triglyceride contents as high as those of the Pre G group, and 3) that the rising temperature and the high serum hormone level after hibernation accelerated the organs and tissues to take in glucose and triglycerides from the blood. Among the three reasons the latter was the most important because of the apparent rise in oxygen consumption. This amount indicated that the metabolism in the organs and tissues had been enhanced markedly. 2. Variation of the three serum hormones Serum T3 of HG was 56.6% higher than that of Pre G, and serum T3 of Post G 51.7% higher than that of HG. Similar to serum T3, serum T4 of Pre G was too low to be tested, but T3 of HG rose significantly, and T3 of Post G continued to rise 43.2% more than that of HG. As a result, serum T3 and T4 were increasing steadily during hibernation. This partem of serum T4 was similar to the results reported by Nauleau, et al. (1987). A comparison between the variations of serum T3 and serum T4 can provide some important information about the secretive state of the thyroid gland. Before hibernation, serum T4 was very low, and serum T3 was higher, being in a dominant position, but in hibernation, serum T4 was twice as great as serum T3, though both of them had increased. These variations were caused by the increased activity of the thyroid gland during hibernation. Turner and Bagnara (1976) suggested that the activity of the thyroid gland of ectotherms was low both in summer and during pre- hibernation. It was enhanced during hibernation, reaching a peak during post- hibernation. The thyroid gland mainly secretes T4, which is the precursor of T3. The secreted T4 is converted into T3 in the blood or in the tissues and organs. The function of stimulating metabolism by T3 is at least three times stronger than that by T4. The activity of the cobra thyroid gland was at a low level in pre-hibernation. The synthesized and secreted T4 was very low and a portion of T4 was converted into T3 thus the concentration of serum T4 was at a low level and the level of T3 was relatively high. After the cobra entered hibernation, the secretive function of the thyroid gland began to become active and produce more and more T4, therefore serum T4 dominated at a higher level though both T4 and T3 increased in the blood. The variation of serum T4 and T3 in post G was the same as in HG. The very high levels of serum T4 and T3 in post G were of great advantage to the cobra enhancing its metabolic rate for arising from hibernation. Besides these, the results in Table 1 and Table 2 also indicate that the levels of serum T4, and T3, of HG were higher than that of Pre G, but the amount of oxygen consumption of HG was less than that of Pre G. This seemed to be self-contradictory because both T4, and T3, were able to stimulate metabolism. These phenomena resulted in the falling temperature, inhibiting the function of hormones. Wilhoft (1966) pointed out that the function of T4 stimulating metabolism appeared to be inactive in many reptiles at low body temperature, and to be active only at higher temperature. Stimulation of metabolism was carried out in such a way that T4 and T3 were capable of inducing the synthesis of aerobic metabolic enzymes. The low levels of serum T4 and T3 in pre- hibernation indicated that the metabolic rate of cobras in this period tends to fall. The metabolic enzymes which had been synthesized were still in an active state because of the higher temperature, thus the pre G cobras had a high oxygen consumption. The metabolic rate of HG, though serum T4 and T3 levels had risen, was at a low level due to the low temperature during hibernation which inhibited the function of T3 (T4) and the activity of metabolic enzymes synthesized before. The cobras of HG had a small oxygen consumption. The very high levels Vol. 3, p. 50 Asiatic Herpetological Research April 1990 of serum T4 and T3 gradually performed the function of stimulating metabolism in the phase at the end of hibernation as the temperature was rising. Serum Cortisol was increasing continuously during hibernation. Serum Cortisol of HG was 2.8 times higher than that of pre G, and that of post G which continued to rise by 3.6 times more than that of HG. Musacchia (1984) reported that the survival of animals during hibernation or hypothermia had a close relation to the level of serum Cortisol in their bodies. An injection of Cortisol into hamsters during hibernation or hypothermia could yield the same result as an injection of glucose, and might increase the survival of hamsters under the same conditions. He also reported that glucocorticosteroids played an important role in animals arising from hibernation or hypothermia. Some clues perhaps could be found for reptiles from the experimental results on hamsters, although they were very different in other respects. The results in Table 2 indicated that the hepatic glycogen content of HG was very low. It was difficult for cobras to continue hibernating while only using so little an amount of hepatic glycogen as a energy source. They must receive energy metabolites in other ways. Gluconeogenesis may be an important process in which Cortisol has a close relation. The surprarenal cortex of cobras during hibernation became active gradually and the level of serum Cortisol became higher and higher, which was related to accelerating gluconeogenesis and to the cobra's arising from hibernation. The variation of Cortisol showed that fat metabolism became active after cobras were aroused out of hibernation. 3. Relationship between the three serum hormones and energy metabolites In pre-hibernation, the function of T4 and T3 in inducing synthesis of aerobic metabolic enzymes was weak because of the low levels of serum T4 and T3. The metabolism of cobras in this period was under the control of the enzymes which had been synthesized before. The metabolic rate of cobras tended to decrease from pre- hibernation to hibernation as the temperature was falling and as the amount of enzyme reduced (synthesizing less and degrading). During hibernation, when the temperature was low, cobras had a low metabolism under the control of the low- temperature isozymes, though the levels of serum hormones were at relatively higher. The low serum glucose and triglyceride contents implied that the decreased metabolism of cobras in hibernation, and the reduction of glycogen in the liver indicated the metabolic rate of cobras at a certain level. In post-hibernation, high temperature provided a favorable factor for hormones to perform their functions, and T3 (T4) induced synthesis of enzymes and Cortisol accelerated the gluconeogenesis and fat catabolism. Serum glucose was increasing markedly though the hepatic glycogen was low, and quite a lot of glucose was produced from gluconeogenesis. Serum T4 and T3 play an important role in accelerating somatic cells to take in energy metabolites and in raising metabolism. In short, during hibernation the cobra nearly used up all the hepatic glycogen but consumed little fat body triglyceride. Body weight tended to be lost significantly during the period after the snake was aroused out of hibernation. Perhaps fat bodies were used up because there was little hepatic glycogen. The high level of Cortisol indicated the tendency for fat catabolism. 4. One putative pattern From the above results, we consider that perhaps the cobra has formed a seasonal regulation of endocrine and that hibernation is controlled by this mechanism. When the season for hibernation comes, the contents of serum T4, T3, and Cortisol decrease so that the metabolic system of all organs and tissues in cobras is controlled effectively. The metabolic rate of cobras tends to go down. With temperature falling and the three serum hormones reducing, the remaining metabolic enzymes in cobras April 1990 Asiatic Herpetological Research Vol. 3, p. 51 become inactive gradually (content less and activity low) and cobras go into hibernation keeping their metabolic rate at a fairly low level under the control of the low temperature isozymes, and after physical regulation. This suggests that the external factor for cobra hibernation is the fall of temperature, while the internal factor is the drop of the three serum hormone levels. In hibernation, the contents of serum T4, T3, and Cortisol increases gradually, but the levels of the three serum hormones are not high enough to accelerate metabolism of the organs and tissues, and the low temperature inhibits the activities of the three serum hormones. When the levels of serum T4, T3, and Cortisol are high enough, the metabolism of the organs and tissues in cobras tends to go up. With the temperature rising, the three serum hormones induce the synthesis of aerobic metabolic enzymes in the organs and tissues and accelerate the metabolism of energy metabolites. The metabolic rate rises and the cobra recovers from hibernation. This suggests that the external factor for the cobra to come out of hibernation is the rise of temperature, while the internal factor is the rise of the three serum hormone levels. Literature Cited DONG, Y. J. ZHAO, J. YANG, Y. CUI, G. LI, AND B. CHEN. 1986. Studies on physiological ecology of Rana spinosa during hybernation and activation. Acta Herpetologica Sinica 1986, 5(4):24 1-245. MAHER, M. T. 1965. The role of the thyroid gland in the oxygen consumption of lizards. General Comparative Endocrinology 5(3):320- 325. MUSACCIA, X. J. 1984. Comparative physiological and biochemical aspects of hypothermia as a model for hibernation. Cryobiology 21(6):583-592. NAULLEAU, G., F. FLEURY, AND J. BOISSIN. 1987. Annual cycles in plasma testosterone and thyroxine in male Aspic Vipera aspis L. (Reptilia, Viperidae), in relation to the sexual cycle and hibernation. General Comparative Endocrinology 65(2):254-263. TURNER, G. AND T. J. BAGNARA. 1976. General endocrinology, sixth ed. W.R. Co. Philadelphia. WILHOFT, D. C. 1966. The metabolic response to thyroxine of lizards maintained in a thermal gradient. General Comparative Endocrinology 7(3):445-451. 5. Conclusion Serum T4, T3, and Cortisol at low levels were involved in the regulation of depressing the metabolism of cobras before hibernation. The low temperature during hibernation inhibited activity of the three serum hormones, and the metabolism was at a low level. With the rising temperature, the levels of serum T4, T3, and Cortisol at very high levels accelerated the metabolic rate of cobras which resulted in arousal from hibernation. Cobras in hibernation used hepatic glycogen as their main energy source instead of fat body triglycerides. I April 1990 Asiatic Herpetological Research Vol. 3, pp. 52-53 Intergradation Between Melanochelys trijuga trijuga and M. t. coronata (Testudines: Emydidae: Batagurinae) INDRANEIL DAS1 AND PETER C. H. PRITCHARD2 1 Animal Ecology Research Group, Department of Zoology, Oxford University, Oxford 0X1 3PS, England ^Florida Audubon Society, 1101 Audubon Way, Maitland, Florida 32751, USA Key words: Reptilia, Testudines, Emydidae, Melanochelys, India, distribution, intergrade. The Indian Black Turtle or pe amai, Melanochelys trijuga, is one of the most abundant chelonians in the Indian subcontinent, with a distribution extending from Sri Lanka (Deraniyagala 1939), through India and Burma, to western Thailand (Wirot 1979), although apparently excluding Bangladesh (Das, 1989). Nevertheless, details of the distribution of the seven described subspecies remain obscure. The range maps provided by Smith (1931), Das (1985), and Tikader and Sharma (1985) indicate rather clearly allopatric distributions for the Indian subspecies, but the overall range of the species is based on extremely few and widely-separated locality points (except for Sri Lanka and Kerala), as is evident in the map provided by Iverson (1986). Deraniyagala (1939) recognized two subspecies in Sri Lanka, but the geographic separation, if any, between these two forms was not made clear. Of the various subspecies, the most distinctive is probably Melanochelys trijuga coronata, whose distribution is restricted to the state of Kerala in southwestern India (the distribution map provided by Tikader and Sharma (1985) involves a transposition of the range of coronata and trijuga ). M. t. coronata has a striking head pattern, with a broad, black diamond-shaped marking on the crown of the head, contrasting with the white to yellow temporal region. The head pattern of the other subspecies consists at most of small, yellow to pink spots and reticulations that may disappear with age. The shell as a whole is usually unrelieved black, in contrast to other subspecies in which at least a lighter plastral rim is evident (although Deraniyagala (1939) reported completely black specimens of M. t . trijuga from Kalpitiya, Sri Lanka and we have seen increased pigmentation with age in the north Indian subspecies, indopeninsularis, in which adult animals may lose the lighter plastral rim). The distribution of M. t. coronata is generally indicated as widely separated from that of M. t. trijuga by the Western Ghats, but the southward penetration of the forma typica into the hiatus between these two subspecies is not represented on existing range maps, and the subspecific relationship between these two forms has been assumed rather than demonstrated. We here report upon four specimens that bridge the geographic map between M. t. trijuga and M. t. coronata, and three that show characters intergradient between them. An adult male (CL 23.9 cm; CW 16.4 cm) was collected by the two authors from a dry stream bed in Chichli, Indira Ghandi (formerly Annamalai) Wildlife Sanctuary, Coimbatore District, Tamil Nadu, on March 27, 1989 (Fig. 1). The right anterior margin of the shell showed signs of an old injury, with several peripheral bones lost. The head pattern of the specimen included the diamond- shaped marking typical of Melanochelys trijuga coronata. The yellow head reticulations and the yellow plastral margin is suggestive of M. t. trijuga, and the size of the specimen is greater than that of any recorded specimen of M. t. coronata (maximum 17.5 cm (Smith 1931) or 18 cm (Tikader and Sharma 1985) of 20.8 cm (Das 1985)). The specimen was retained alive and housed at the Madras Crocodile Bank. 1990 bv Asiatic ileroetoloeical Research April 1990 Asiatic Herpetological Research Vol. 3, p. 53 FIG. 1. Intergrade between Melanochelys trijuga coronata and M. t. trijuga. A second specimen — a fragmentary shell only — was also found in the Indira Ghandi Wildlife Sanctuary, in a tribal settlement. The nuchal area and anterior peripherals are missing. Shell width is 13.7 cm, again indicating a specimen larger than is typical of Melanochelys trijuga coronata. Ankylosis of the shell, as illustrated for Sri Lanka specimens by Deraniyagala (1939), is essentially complete, except for faint indications of the peripheral sutures. The specimen is registered as PCHP 2803 in the private collection of the second author. In addition to these, two intergradient specimens were found in the collection of the Southern Regional Station of the Zoological Survey of India (Madras), both collected by M. Vasant and party. An 8.1 cm juvenile (Lot no. 10) was collected at Kombiar Charagam, Kalakkadu Wildlife Sanctuary, Turunelveli District, Tamil Nadu (altitude 210 m), on October 13, 1987. A 20.6 cm adult male (Lot no. 6) was collected at Nambiar, Nambi Kovil Road, also within Kalakkadu Wildlife Sanctuary (altitude 140 m), on January 11, 1987. Melanochelys trijuga is basically a pond turtle, and the Western Ghats appear to separate the essentially lowland ranges of M. t. coronata and M. t. trijuga Nevertheless, the new localities all fall within upland areas actually in the Western Ghats. It appears that the hills may form a sufficient barrier to allow the evolution of different subspecies east and west of the Ghats, but the ability of M. trijuga to exist far from water suggests that individuals wandering into these hills may provide the stock for an intergradient population. It would be worthwhile to search for a lowland intergradient population in the coastal area at the northern extreme of the distribution of M. t. coronata. Other cases exist of congeneric batagurine species of a similar ecological role differing primarily in head pattern, as is the case with some of the species of Rhinoclemmys in the Neotropics. One is tempted to interpret the different markings as a means of avoiding cross-matings in situations of sympatry. But, the overall allopatry of the species of Rhinoclemmys in northern South America, for example, suggests that this interpretation may be an incomplete one, as does the discovery of intergradation between adjacent populations of Melanochelys that differ primarily in cephalic pigmentation. Literature Cited DAS, I. 1985. Indian turtles. A field guide. World Wildlife Fund India, Eastern Region, Calcutta. DAS, I. 1989. Report on a survey of freshwater turtles and land tortoises in Bangladesh. Report to Fauna and Flora Preservation Society, July 1989. Mimeo9pp. DERANIYAGALA, P. E. P. 1939. The tetrapod reptiles of Ceylon. Colombo Museum, Colombo. 412 pp. IVERSON, J. B. 1986. A checklist with distribution maps of the turtles of the world. Paust Printing, Richmond, Indiana. 283 pp. SMITH, M. A. 1931. The fauna of British India, including Ceylon and Burma. Reptilia and Amphibia, Vol. 1. Loricata, Testudines. Taylor and Francis, London. 185 pp. TIKADER, B. K. AND R. C. SHARMA. 1985. Handbook: Indian Testudines. Zoological Survey of India, Calcutta. WIROT, N. 1979. The turtles of Thailand. Mitbhadung Press, Bankok. [April 1990 Asiatic Herpetological Research Vol. 3, pp. 54l9] Thermal Sensitivity of Sprinting and Clinging Performance in the Tokay Gecko (Gekko gecko) JONATHAN B. LOSOS1 1 Museum of Vertebrate Zoology and Department of Integrative Biology, University of California, Berkeley, CA 94720 USA Abstract. -The thermal sensitivity of sprinting and clinging ability was measured in tokay geckos (Gekko gecko). Sprinting performance was maximal at high (35-4 1°C) temperatures, as is the case for other nocturnal lizards, but the optimal temperature for clinging was considerably lower (approximately 17°C). These different optima could be adaptive if maximal sprinting and clinging capabilities are needed at different temperatures. Alternatively, they could result from constraints on adaptive evolution. Key Words: Reptilia, Sauria, Gekkonidae, Gecko gecko, thermal sensitivity, optimal performance. Introduction Most physiological processes are temperature-dependent. Ectothermic animals, which do not maintain a constant body temperature, are thus subject to fluctuation in the rate at which they can perform many vital tasks. By regulating body temperature behaviorally, however, many reptiles can maximize performance capability (Huey 1982a). In some cases, maximal performance temperature might differ for different tasks. For example, a wealth of behavioral data indicates that many lizards and snakes increase their body temperature after feeding, which suggests that digestion has a higher optimal temperature than other activities (reviewed in Huey 1982a). Early work centered on the thermal dependence of sub-organismal traits (e.g., enzyme activity, muscle contractile speed [Huey and Stevenson 1979]). In many cases, however, the effect of temperature on the functional capacities (e.g., sprint speed, critical thermal maximum temperature) cannot be predicted by study at sub-organismal levels (e.g., Licht 1967; Marsh and Bennett 1985, 1986). Consequently, recent studies have focused on whole-organism performance. The thermal dependence of sprinting ability has been studied in great detail. Maximum sprinting speed and/or endurance in many species occurs at the temperature they most frequently experience in nature (Bennett 1980); however, nocturnal lizards appear exceptional (Huey and Bennett 1987; Huey etal. 1989). The thermal sensitivity of few other whole-organism locomotor performance measures has been determined for lizards. Here I report the thermal sensitivity of tokay geckos (Gekko gecko) [Fig. 1] for two ecologically relevant tasks, maximum sprinting and clinging capability. I ask whether maximum performance capability occurs at the relatively low temperatures most commonly experienced by geckos (Huey et al. 1989) and whether the optimal temperature is the same for the two performance measures. Methods Tokays are relatively large geckos found on trees and walls throughout southeastern Asia (Smith 1935). Lizards were captured on Phuket Island, Phuket Province, Thailand, and transported to the University of California, Berkeley in late September 1987. They were maintained in 30 x 17 x 9 cm plastic shoeboxes at ambient temperatures and provided with water and crickets ad libitum. An ontogenetic series was used in this study (11 individuals; snout-vent length: 85-180 mm; mass: 10- 140 g). All trials were conducted in a walk-in environmental chamber in the Museum of Vertebrate Zoology, University of 1990 by Asiatic Herpetological Research April 1990 Asiatic Herpetological Research Vol. 3, p. 55 FIG. 1 . Gekko gecko from Phuket Province, Thailand. California, Berkeley, within a month of capture. Lizards were placed in the chamber at least one hour prior to performance measurement. Humidity, which could not be regulated, was determined on several occasions using a Bacharach sling psychrometer. Clinging capability was measured by placing lizards on a plexiglass plate and, at a gradual and steady rate, lifting the end of the plate so that the lizard was clinging with its head directed down. The angle at which the lizard fell from the plate was recorded (protocol modified from Emerson and Diehl 1980; Alberch 1981). Lizards that jumped from the plate were not included in the analysis. As lizards began to slide, they usually attempted to maintain their grip by moving and re-applying their toe pads regardless of temperature. There was no evidence that temperature affected the lizards' efforts to prevent sliding and falling from the experimental plate. One trial per lizard was conducted per temperature. Performance at nine temperatures (12, 16, 17, 22, 24, 31 [three times], 34, 35, 41; the order of temperatures is presented in figure 2) was measured over a six-day period. On some days, two trials were conducted. Sprint capability was measured by placing lizards at the end of a 2.25 m trackway covered with a rough rubber surface and inducing them to run by repeated taps to the tail (protocol following Huey 1982b; Huey et al. 1989; Garland 1985). As the lizard ran, it interrupted light beams stationed every 0.25 m. The time elapsed during each interval was computed by a Compaq personal computer, the fastest single interval during four trials, conducted at hourly intervals, was considered the maximum speed for that lizard at that Vol. 3, p. 56 Asiatic Herpetological Research April 1990 155 20 30 Temperature 1.50 .25 1.00 , 1 • 1 • t y \ 2 • 1 V 25 35 Temperature FIG. 2. Clinging ability (mean + 1 s.e.) at different temperatures. Clinging ability is measured as the angle of the plate at which a lizard lost its grip and fell. The points are numbered by the order in which the trials were conducted. temperature. Lizards that did not sprint at maximal capability on any of the trials at a given temperature, as judged by their gait, were not included in the calculations for that temperature. Lizards were tested at five temperatures (30, 36, 19, 26, 41°C, in that order), one temperature per day, over an eight-day period. Trials were never held on three consecutive days. A second trial at 30°C was held at the conclusion of the study to determine whether a performance decline had occurred. Animals whose performance decreased > 30 % were excluded from the analysis. Results Clinging performance is temperature- dependent, with a peak at 17°C (Fig. 2). A non-linear equation (In [clinging ability] = 0.99 + 2.69 * In [temp] - 0.45 * {In [temp]}2; F28 = 5.08, P < 0.05) better fits the data than a linear regression (Fli9 = 2.77, P>0.10). Although between-day variation exists in clinging ability at a given temperature, no general pattern of increased or decreased performance over the duration of the study exists. For example, three trials were conducted at 31 °C. The second trial had the lowest mean, whereas the last trial had the highest mean. Three other sets of trials were conducted at approximately the same temperature (16-17°C, 22-24°C, 34-35°C). In two cases, performance ability decreased FIG. 3. Sprint speed (mean ± 1 s.e.) at different temperatures. The numbers indicate which represents the first and the second 30° trials. in the second trial, but in the third case it increased. Sprinting capability is also temperature- dependent (Fig. 3). Maximal performance ability occurs at 36-4 1°C, but interpretation of the results is difficult because performance ability declined over the course of the experiment, as evidenced by the difference in sprinting ability in the two sets of trials at 30°C. Despite this decline, several results are clear from inspection of Fig. 3: 1. performance at 36°C (tested 8 July) is slightly higher than at 30 (7 July); performance at 40°C (tested 13 July, after all trials except the second set at 30) is nearly as high or higher than all other temperatures; 3. performance at 30°C (16 July) is greater than performance at 26°C (11 July), which, in turn, is greater than performance at 19°C (10 July). Consequently, even if performance steadily decreased over time, it is reasonable to conclude that maximal sprint performance occurs around 36-4 1°C Discussion Sprinting and clinging are ecologically relevant performance measures for geckos, but their optimal performance temperatures differ greatly for tokays. Sprint performance is greatest at relatively high temperatures, as is the case for a number of other nocturnal gecko species (Huey et al. 1989) and the nocturnal skink April 1990 Asiatic Herpetological Research Vol. 3, p. 57 Eremiascincus fasciolatus (Huey and Bennett 1987). Clinging capability, the thermal dependence of which has never previously been investigated, is maximal at considerably lower (approx. 17°C) temperatures. It is difficult to envision how such different optima could evolve adaptively. Most lizards sprint maximally at temperatures close to those they normally experience (Huey 1982a; Huey et al. 1989). The low optimal temperature for clinging matches the field temperatures of many active nocturnal geckos (Huey et al. 1989). However, many nocturnal geckos bask and/or are active to some extent during the day (Bustard 1967, 1968; Werner and Whitaker 1978; Nagy and Knight 1989). Consequently, the high optimal temperature for sprinting seen in many nocturnal lizards might represent adaptation for diurnal capability (Huey and Bennett 1987; Huey et al. 1989). Nonethless, it seems implausible that one aspect of locomotion, sprint performance, should be selected at high temperatures, whereas another important component of effective movement, clinging, should be favored at considerably lower temperatures. As an alternative explanation, the differences in thermal optima might result from differences in evolutionary lability of the two performance capabilities and thus represent constraints on adaptive evolution in either sprinting or clinging ability. To understand why these performance abilities are affected differently by temperature, a better understanding is needed of their underlying physical and physiological bases (e.g., Garland 1984, 1985; Marsh and Bennett 1985, 1986). In many species of lizards, sprint performance is maximal at temperatures close to the critical thermal maximum (Huey and Bennett 1987; Huey et al. 1989). The cause for this linkage is unclear. However, if geckos must be able to survive diurnal temperatures (either because they intentionally maintain high temperatures to maximize other processes, or because environmental conditions preclude the maintenance of lower temperatures), then they would have to evolve high critical thermal maxima. The high sprint performance maximum might be a correlated effect of this physiological adaptation to high temperatures and not be adaptive per se (Huey et al. 1989). Clinging capability depends upon both physical and physiological processes. Geckos cling to smooth surfaces by dry adhesion. The subdigital lamellar pads of geckos are covered with millions of microscopic setal hairs. When the pads are adpressed to a surface, these hairs form intermolecular bonds with molecules on the surface of the substrate (Hiller 1975). If the surface energy (a measure of the number of free electrons on the surface of the substrate) is relatively high, then enough bonds can form to support the lizard. Because these bonds result from the activity of electrons, the forces theoretically should be temperature-independent over the range of temperatures in this study. However, geckos have an elaborate muscular and vascular system for the adpression and removal of their toe pads (Russell 1975); the thermal dependence of these muscles has not been investigated. The poor clinging performance of geckos at 10-12°C is clearly the result of physiological incapacitation. At that temperature, geckos were generally inactive, moved slowly and infrequently, and even rarely bit or barked when handled. In contrast to trials at higher temperatures, the lizards did not attempt to adjust their pads or posture when the plate was tilted and quickly lost their hold. More research is required to determine whether the performance decline at temperatures above 17-1 8°C results from decreased capability of the muscles and enzymes involved in clinging. One possible confounding effect in the clinging experiments is the variation in absolute humidity. Moisture might decrease the formation of high-energy intermolecular bonds. Absolute humidity in the environmental chamber was measured at several temperatures and increased from 8.5 Barrs at 11.9°C to 25.1 Barrs at 34.3°C (relative humidity, Vol. 3, p. 58 Asiatic Herpetological Research April 1990 however, was greatest at intermediate temperatures). Consequently, the greater absolute humidity at higher temperatures in the environmental chamber might have caused a decrease in clinging ability. Further research is needed to investigate to what extent humidity affects clinging. Although different processes may often be maximized at different temperatures, rarely is the difference as great as is observed between sprinting and clinging in tokay geckos (Huey 1982a). One would not expect these geckos to need maximum ability at these aspects of locomotion at such different temperatures, but neither would one expect the physiological sensitivity of different systems to be so different. Interestingly, the optimal temperature for hearing sensitivity in the tokay gecko is intermediate between the sprinting and clinging optima (Werner 1976). Further research is required to understand the processes shaping and constraining performance evolution. Acknowledgments I thank T. Papenfuss and R. Macey for providing the geckos, R. Jones for assistance with equipment, R. Huey for advice on racing geckos, J. Herron for providing me with a pre-publication manuscript, and A. Bauer, H. Greene and R. Huey for constructive comments on a previous version of this paper. Literature Cited ALBERCH, P. 1981. Convergence and parallelism in foot morphology in the neotropical salamander genus Bolitoglossa. I. Function. Evolution 35:84-100. BENNETT, A. F. 1980. The thermal dependence of lizard behavior. Animal Behavior 28:752-762. BUSTARD, H. R. 1967. Activity cycle and thermoregulation in the Australian gecko, Gehyra variegata. Copeia 1967:753-758. BUSTARD, H. R. 1968. The ecology of the Australian gecko Heteronotia binoei in northern New South Wales. Journal of Zoology, London 156:483-497. EMERSON, S. B., AND D. DIEHL. 1980. Toe pad morphology and mechanisms of sticking in frogs. Biological Journal of the Linnaen Society 13:199-216. GARLAND, T., JR. 1984. Physiological correlates of locomotory performance in a lizard: an allometric approach. American Journal of Physiology 247:R806-R815. GARLAND, T., JR. 1985. Ontogenetic and individual variation in size, shape and speed in the Australian agamid lizard Amphibolurus nuchalis. Journal of Zoology, London (A) 207:425-439. HILLER, U. 1975. Comparative studies on the functional morphology of two gekkonid lizards. Journal of the Bombay Natural History Society 73:278-282. HUEY, R. B. 1982a. Temperature, physiology, and the ecology of reptiles. Pp. 25-91. In C. Gans and F. H. Pough (Eds.), Biology of the Reptilia, Vol. 12. Academic Press, New York, USA. HUEY, R. B. 1982b. Phylogenetic and ontogenetic determinants of sprint performance in some diurnal Kalahari lizards. Koedoe 25:43- 48. HUEY, R. B., AND A. F. BENNETT. 1987. Phylogenetic studies of coadaptation: preferred temperatures versus optimal performance temperatures of lizards. Evolution 41:1098- 1115. HUEY, R. B., P. H. NIEWIAROWSKI, J. KAUFMANN, AND J. C. HERRON. 1989. Thermal biology of nocturnal ectotherms: is sprint performance of geckos maximal at low body temperatures? Physiological Zoology 62:488-504. HUEY, R. B., AND R. D. STEVENSON. 1979. Integrating thermal physiology and ecology of ectotherms: a discussion of approaches. American Zoologist 19:357-366. LICHT, P. 1967. Thermal adaptation in the enzymes of lizards in relation to preferred body temperature. Pp. 131-145. In C. L. Prosser (Ed.), Mechanisms of Temperature Adaptation. American Association for the Advancement of Science, Washington, D. C. MARSH, R. L., AND A. F. BENNETT. 1985. April 1990 Asiatic Herpetological Research Vol. 3, p. 59 Thermal dependence of isotonic contractile properties of skeletal muscle and sprint performance of the lizard Dipsosaurus dorsalis. Journal of Comparative Physiology B 155:541- 551. MARSH, R. L., AND A. F. BENNETT. 1986. Thermal dependence of contractile properties of skeletal muscle from the lizard Sceloporus occidentalis with comments on methods for fitting and comparing force-velocity curves. Journal of Experimental Biology 126:63-77. NAGY, K. A., AND M. H. KNIGHT. 1989. Comparative field energetics of a Kalahari skink (Mabuya striata) and gecko (Pachydactylus bibroni). Copeia 1989:13-17. RUSSELL, A. P. 1975. A contribution to the functional analysis of the foot of the Tokay, Gekko gecko (Reptilia: Gekkonidae). Journal of Zoology, London 176:437-476. SMITH, M. A. 1935. The Fauna of British India, Reptilia and Amphibia. Vol. 2 — Sauria. Taylor and Francis, Ltd., London. 440 pp. WERNER, Y. L. 1976. Optimal temperatures for inner-ear performance in gekkonid lizards. Journal of Experimental Zoology 195:319-352. WERNER, Y. L., AND A. H. WHITAKER. 1978. Observations and comments on the body temperatures of some New Zealand reptiles. New Zealand Journal of Zoology 5:375-393. I April 1990 Asiatic Herpetological Research Vol. 3, pp. 60-63 Mating Call Structures of the Chinese Frog, Rana nigromaculata (Amphibia, Anura, Ranidae) YONG MU1 AND ERMI ZHAO1 ^Chengdu Institute of Biology. P.O. Box 416, Academia Sinica, Chengdu, China Abstract. -Mating calls of the Chinese Pond Frog, Rana nigromaculata at three localities were recorded and compared. The calls of R. nigromaculata are short in duration and consist of a few notes. Each note has several distinct pulses. The call structures of northern and central China are different from those of Sichuan, hence we tentatively regard them as different subspecies. Key words: Amphibia, Anura, Ranidae, Rana nigromaculata, mating calls. Introduction Rana nigromaculata occurs widely in the lower Amur and Ussuri river valleys (USSR), Korea, Japan, and throughout northeastern and midwestern China. In China it occurs in all but Xinjiang, Tibet and Guangxi autonomous regions, and Yunnan, Taiwan, and Hainan provinces. As far as we know, no paper has been published in China analyzing the mating calls of Chinese anurans. The mating call of frogs is a useful clue in revealing systematic and evolutionary relationships. Vocalizations are species- specific in anurans, and serve as isolating mechanisms. Therefore, some taxonomic revisions have been based primarily on call differences [see Kuramato (1977) for review]. Methods Calls of Rana nigromaculata were recorded in 1988 at three localities. Several calls of a single male were recorded on 19 March, in the suburbs of Hongya County, about 100 km southwest of Chengdu, Sichuan Province. The frog was calling from a stone near water. The water temperature was 19°C. A number of calls from three males were recorded at East Lake Park, Wuchang, Wuhan, Hubei Province on 17 June. The frogs called while standing in shallow water shaded by grass. The water temperature was 22°C. The calls of 10 males were recorded at Wofoshi, Xiangshan Park, Beijing, on 28 May. Male frogs were found calling in surroundings similar to those of frogs in Hubei Province. The water temperature was 24°C. Temperature records were always taken in the vicinity of calling males. Calls were recorded with a Sony TCS-370 cassette recorder and a Feidec TSM-91 microphone. Sonograms were prepared with Kay 7800 and Kay 7900 sonographs. For the analysis of mating calls, standard (2.56 sec) sonograms with a 150 Hz bandwidth filter were used. When a call was composed of several groups of pulses, each group was termed a note. The main acoustic parameters measured were: duration of call, number of notes and pulses, pulse repetition rate (number of pulses per second), fundamental frequency, and dominant frequency. Results and Discussion The vocalizations of Rana nigromaculata consist of 4-10 notes, and each note has 3- 7 distinct pulses. The higher frequency parts (above 2 kHz) are usually absent in the first few notes (1-3). Time intervals between pulses tend to become longer at the end of the call. In all localities, the frog's frequency band ranges from 0 to 8 kHz. © 1990 by Asiatic Herpetological Research April 1990 Asiatic Herpetological Research Vol. 3, p. 61 TABLE 1 . Analysis of the calls of Rana nigromaculata. The dominant frequency and pulse repetition rate of R. nigromaculata from Hongya, Sichuan are about half of those recorded at the other localities. Note the divergence of the Sichuan call structure from northern (Beijing) and central (Wuhan) China. Number of Pulse Number of Fundamental Dominant Notes Call Duration Repetition Rate Calls Frequency Frequency Beijing 10 668.6 89.7 1 9 529.9±77.1 84.7110.8 3 0.5310.02 2.3910.20 8 450.5±26.4 80.6127.4 12 7 396.4±21.0 109.8144.0 7 6 344.7±33.1 104.7122.1 6 5 324.2±17.9 84.5126.2 3 4 255.0135.8 43.2111.2 5 1 50.8112.1 154.8145.9 5 Wuhan 8 467.3 69.3 1 0.5110.09 2.3210.17 7 452.6 86.2 1 6 426.6+53.6 77.718.7 3 5 319.7 90.7 1 1 50.8117.5 120.8134.1 10 Hongya 5 637.6166.9 45.417.7 7 0.471.0.06 1.4210.10 4 576.5157.0 49.815.5 7 1 79.9113.1 125.412.1 2 Their fundamental frequencies are approximately identical (0.5 kHz). Since fundamental frequencies are dependent on the oscillations resulting from air passing over the vocal cord, causing it to vibrate at a frequency which is a function of the mass and tension of the cord (Duellman and Trueb 1986), it is likely that the frogs have a common vocal structure. As well as multi-note calls, Rana nigromaculata has a single-note call. Frogs in northern and central China (Beijing and Wuhan) emit this sound repeatedly and usually in groups. Each group is composed of 5-10 single-note calls. In Hongya, the single-note call could rarely be heard. Because R. nigromaculata in Sichuan is peripheral in distribution, we believe that the call structure in northern and central China represent the basic structure, and that the Sichuan structure was derived from it. Table 1 and Figure 1 clearly demonstrate the divergence of call structure between northern and central China, and Sichuan. In comparing the three main acoustic parameters, call duration, pulse repetition rate, and mean dominant frequency, a general pattern is evident (Zweifel 1959). Pulse repetition rate and dominant frequency correlate positively with temperature, but duration of call correlates negatively with temperature. However, we feel that this is not sufficient to explain the differences between the populations. At a temperature only 3 and 5°C lower than in Wuhan and Beijing, respectively, the dominant frequency and pulse repetition rate of Sichuan Rana nigromaculata are about half of those found in the former localities. The length of calls of Sichuan frogs is approximately equal to the maximum call length of frogs from the Wuhan and Beijing localities, and their vocal sacs are of the same size. In addition, they have very similar fundamental frequencies. We conclude that they have nearly identical sound-producing Vol. 3, p. 62 Asiatic Herpetological Research April 1990 TYPE B/65 SONA&Ram * KAY TYPE B/65 SONAGBAM e KAY ELEMETRICS CO PINE BROOK. N.J. fc fa t,- it 8 1- p p I f I f M I 1 I I i t' TYPE B/65 SONAGRAM * KAY ELEMETRICS CO PINE BROOK N.J. JTT W'.W-% ;h i i i 1 : Hi' ' in w it ftt i i "V I' Hil HM 1H MUM f in r1 i FIG. 1. Sonagrams of the calls of R ana nigromaculata in China, a. Seven note call from Beijing on May 27, 1988. The water temperature was 24°C. b. One 6 note call and two 1 note calls from Wuchang, Hubei Province, June 17, 1988. The water temperature was 22°C. c. Two 5 note calls from Hongya, Sichuan Province, March 19, 1988. The water temperature was 19°C. April 1990 Asiatic Herpetological Research Vol. 3, p. 63 structures, but different controlling mechanisms. In acknowledging that our analysis of frog calls may not be exhaustive, we can only tentatively suggest that the Sichuan population of Rana nigromaculata is a separate subspecies from the Wuhan and Beijing populations. Acknowledgments We would like to thank Theodore J. Papenfuss (Museum of Vertebrate Zoology, University of California, Berkeley) for donating the sonograph paper, and Mitsuru Kuramoto of the Fukuoka University of Education, Japan. Literature Cited DUELLMAN, W. E. AND L. TRUEB. 1986. Biology of Amphibians. McGraw-Hill, New York. 670 pp. KURAMOTO, M. 1977. Mating call structures of the Japanese pond frogs, Rana nigromaculata and Rana brevipoda (Amphibia, Anura, Ranidae). Journal of Herpetology 11(3):249- 254. ZWEIFEL, R. G. 1959. Effect of temperature on call of the frog, Bombina variegata. Copeia 1959(4):322-327. April 1990 Asiatic Herpetological Research Vol. 3, pp. 64-66 1 Four Remarkable Reptiles from South China Sea Islands, Hong Kong Territory JAMES LAZELL1 AND WENHUA LU1 lThe Conservation Agency, Rhode Island 02835, USA Key Words: Reptilia, Squamata, Hong Kong, distribution. FIG. 1. Hong Kong and nearby islands. 1. Hei Ling Chau. 2. Shek Kwu Chau. 3. High West Pokfulam, Hong Kong Island. Introduction There are at least one hundred vegetated islands, presumably supporting amphibians and/or reptiles, within the Hong Kong Territory. This is estimated to be about three percent of the total number of continental shelf islands in the South China Sea between Taiwan and Hainan Dao. Nevertheless, the number of endemic and isolated species documented from the territory in the comprehensive work of Karsen, Lau, and Bogadek (1986) is high: four endemics and at least fifteen widely disjunct populations. Here we add two species not previously recorded (both widely disjunct) and new records of species poorly documented before (Fig. 1). Voucher specimens are in the Museum of Comparative Zoology (MCZ), Harvard University. Dibamus cf. bourreti (White-tailed Two-footed Lizard) A single specimen, MCZ 172041, was April 1990 Asiatic Herpetological Research Vol. 3, p. 65 collected by Anthony Bogadek, 1 April 1987, on Hei Ling Chau ca. 10 km southwest of Victoria, Hong Kong — a range extension northeast for the genus, and family Dibamidae of ca. 800 km (Lazell 1988). The specimen, a 177 mm SVL male, was sent to Allen Greer, Australian Museum, Sydney, for identification. He believed the specimen was "probably" Dibamus bourreti (in litt., 9 May 1988). However, we note the following sharp distinctions from D. bourreti as diagnosed and described by Greer (1985: 148): the rostral suture is not complete, not present from lip to nostril but only posterior to the nostril. There is a prominent labial suture. There are no preanal or tibial pores. Dibamis bourreti is diagnosed as having a complete rostral suture, no labial suture, and four preanal pores on each side, even in a female — the highest count in Dibamus. This specimen, MCZ 172041, has 23 scale rows at midbody, six scales fronting the hindlimb where Greer (1985:120) shows D. novaeguinae having four, and six rows of preanal scales where Greer (1985:120) shows D. novaeguinae having three. The hindlimb is 2.7% of SVL. The tail is 22.6% of SVL. In life this specimen was lilac of lavender-gray shading to buff on the head and chalk-white on the tail. It is much paler and less contrastingly marked than the Guangxi D. bourreti illustrated by Tian and Jiang (1986). See Figure 2. Typhlops albiceps This tiny snake, apparently rare throughout its range, was known from Hong Kong Island only on the basis of two specimens collected in 1959 and 1966 (Karlsen et al. 1986), now in the British Museum (Natural History): BMNH 1954.1.13.4 and BMNH 1983.946, respectively. It was rediscovered 27 May 1988 at High West, Pokfulam, Hong Kong Island by Sandra Brown (Macklin 1988). The species appears reasonably common at this site. Voucher specimens are at the St. Louis School, West Point, Hong Kong (curated by Anthony Bogadek) and at MCZ FIG. 2. Dibamus sp. from Hei Ling Chau. The rostral, R, lacks a suture from lip to nostril. A labial suture, L, is present. (MCZ 173290). Lance (1976) did not recognize "Typhlina" or "Ramphotyphops" and neither do we. The putative distinction from Typhops is entirely in male genetalia (McDowell 1974). Because females are indistinguishable, recognition of "Typhlina" or "Ramphotyphops" directly violates the principles of systematic zoology (Mayr et al. 1953) and the principles of animal taxonomy (Simpson 1961). This is the only species of Typhlops in China with 18 scale rows. It is very slender, the head spatulate. MCZ 173290 is 155.5 mm SVL with a 2.5 mm tail. The dorsum is wood brown shading to buff on the head. The chin is near-white; this color extends posteriorly for 10 scales onto the throat The tail tip is near- white; this color extends anteroventrally on all 10 subcaudal scales and 4 rows anterior to the vent. Dendrelaphis pictus (Painted Bronze Back Snake) This elegant snake is known from Guangdong, Guangxi, and Yunnan (Hu et al. 1980), and from Hong Kong on the basis of a single report (Wall 1903). One of us (Lazell) has searched both BMNH and MCZ for this specimen with no success. Peaker (1987) criticized Karsen et al. (1986) for including this species in the Hong Kong fauna. Deletion would have Vol. 3, p. 66 Asiatic Herpetological Research April 1990 TABLE 1. Some scale counts for Chinese Ahaetulla prasina. Mainland Shek Kwu Chau Ventrals, males 186-209 217 Ventrals, females 191-204 224 Subcaudals, males 155-177 - Subcaudals, females 150-169 184 been premature. Between 1971 and 1984 Dr. Barry Hollinrake, resident of Shek Kwu Chau, a small island 1.3 km south of Lanuau, amassed a collection of 35 snakes on the island of 9 species, now housed at MCZ (Boalch 1988). Among these is a beautiful female Dendrelaphis pictus with 15 dorsal scale rows, 171 ventrals, and 131 subcaudals: MCZ 173278. The ventral and subcaudal counts of the Shek Kwu Chau specimen are significantly lower than those given by Hu et al. (1980: 39): 184 ventrals (females 186-193) and 141-169 subcaudals (females 141-169). Ahaetulla prasina (Jade Vine Snake) Two specimens of this spectacular snake, never previously recorded in Hong Kong Territory, are in the Hollinrake collection from Shek Kwu Chau: male, MCZ 173303, and female, MCZ 173304. Their scale counts are much higher than those given by Pope (1935) of Hu et al. (1980) for this widespread South China species: Table 1. Literature Cited BOALCH, K. 1988. Snakes alive —a specialist's delight. South China Post 154(1 13):3. GREER, A. 1985. The relationships of the lizard genera Anelytropsis and Dibamus. Journal of Herpetology 19:116-156. HU, B., M. HUANG, Z. XIE, E. ZHAO, Y. JIANG, Z YU, AND J. MA. 1980. Illustrated snakes of China. Science and Technology Press, Shanghai. KARSEN, S. M. LAU, AND A. BOGADEK. 1986 Hong Kong amphibians and reptiles. Urban Council, Hong Kong. 136 pp. LANCE, V. A. 1976. The land vertebrates of Hong Kong, Pp. 6-22. In B. Lofts (ed). The fauna of Hong Kong. Royal Asiatic Society, Hong Kong. LAZELL, J. 1988. A leg up —by 800 kilometers. Assoc. Pacific Systematists Newsletter 5:2. MACKLIN, S. 1988. biologists bewildered. Post 154(147):!. Reclusive reptile has South China Morning MAYR, E., E. G. LINSLEY, AND R. L. USFNGER. 1953. Methods and principles of systematic zoology. McGraw-Hill, New York. 336 pp. MCDOWELL, S. 1974. A catalogue of the snakes of New Guinea and the Solomons, with special reference to those in the Bernice P. Bishop Museum 1. Scolecophidia. Journal of Herpetology 8:1-57. PEAKER, M. 1987. Review of Karsen et al. (1986). The Herptile 12(l):36-37. POPE, C. H. 1935. The reptiles of China. Natural History of Central Asia 10, American Museum of Natural History, New York. 604 pp. SIMPSON, G. G. 1961. Principles of animal taxonomy. Columbia University Press, New York. 247 pp. TIAN, W. AND Y. JIANG. 1986. Chinese amphibian and reptile identification manual. Science Press, Beijing. 364 pp. WALL, F. 1903. A prodromus of the snakes hitherto recorded from China, Japan, and the Loo Choo Islands; with some notes. Proceedings of the Zoological Society of London 84-102. April 1990 Asiatic Herpetological Research Vol. 3, pp. 67-84 On the Independence of the Colchis Center of Amphibian and Reptile Speciation BORIS S. TUNIYEV1 ^Causasian State Biosphere Reserve, Sochi, USSR Abstract. -The Colchis region of Western Transcaucasia is characterized by a rather uniform thermal regimen, corresponding to a subtropical climate. The Colchis forests contain an extraordinary abundance and diversity of tree, shrub, and vine species. The herpetofauna of the Colchis forests is surprisingly poor, despite its uniqueness. Key Words: Amphibia, Reptilia, USSR, Caucasus, biogeography. Introduction The herpetofauna of Western Transcaucasia is not homogeneous, due to the different age and genesis of the species distributions. Along with autochtonous and endemic forms, one can find species whose main areas of distribution are in the European part of the USSR and in the Eastern Mediterranean. At the same time, a number of species which have main distributional centers in the Colchis occur beyond the bounds of Western Transcaucasia, in other parts of the Caucasian Isthmus. For these reasons, it is necessary to define the Colchis herpetofauna and to determine its place in the fauna of reptiles and amphibians of the Caucasian Isthmus as a whole. Research on this issue started with the works of Nordmann (1840), Derjugin (1899), Silantyev (1903), Brauner (1905), and Nesterov (1911). However, the first well-grounded definition of the fauna in question from a zoogeographical point of view was presented in the works of Satunin (1912). Satunin wrote in 1910, "So far I cannot say much about the genesis of the fauna of this region called Western Transcaucasia. This country with its evergreen plants and scanty fauna resembles a piece of the Mediterranean in the narrow sense of the word. True, here are endemic species and forms, but not a single genus of vertebrate is unrepresented in the countries of the Mediterranean. Often they are inhabited by the same species. The question is whether this fauna has appeared from the west or is it the remainder of the fauna that has populated densely the shores of the Black Sea at one time. It is impossible to answer these questions at the present level of our knowledge. But even now I can definitely say that this fauna by its origin, has nothing in common with the faunas in other regions of the Caucasus." In 1912, Satunin divided the Caucasian Isthmus into five subregions and 11 districts, including the Colchis in the West- Transcaucasian district of the Littoral subregion. Among other merits of this work by Satunin, one cannot but mention the fact that for the first time, he defined in an exact way the Colchis region proper. He defined the northern border as the spurs of the Main Caucasian Range up to the basin of the Tuapse River, the southern border as the Pontic Range, and the eastern border as the Arsijanskij Range. The valley of the Rioni River and the adjacent southern slopes of the Main Range were defined as the central part of the region. Satunin emphasized the depauperate herpetofauna of this region on one hand, and the presence of endemic species such as Vipera kaznakowi and Bufo verrucosissimus on the other hand. Nikolsky (1911) assigned the entire Caucasus, excluding eastern Precaucasia, © 1990 by Asiatic Herpetological Research Vol. 3, p. 68 Asiatic Herpetological Research April 1990 to the Mediterranean. However, he could not differentiate the forest and the alpine belts of the Greater Caucasus, because of the absence of data. Results and Discussion Investigations of the last decades made it possible to add the majority of the species of the Colchis herpetofauna to an overall picture of Colchis faunal distributions (Turov 1928; Bartenev and Reznikova 1935; Khozatsky 1941; Milyanovskiy 1957; Bannikov et al. 1977; Negmedzyanov and Bakradze 1977; Orlova 1973, 1978a, 1978b; Golubev 1980, 1985; Tuniyev 1983, 1985). In addition there has been a revision of the taxonomic status of such forms as Vipera kaznakowi (Vedmederja et al. 1986; Orlov and Tuniyev 1986a, 1990 this volume), Lacerta agilis (Peters 1960), L. derjugini (Bartenev and Reznikova 1931; Orlova 1978a ; Bischoff 1982, 1984), L. saxicola (Darevsky 1967; Darevsky and Vedmederja 1977), Anguis fragilis colchicus (Lukina 1965; Scherbak and Scherban 1980), and others. Accumulation of this information along with works on fossil amphibians and reptiles of the Caucasus (Vekua et al. 1979; Chkhikvadze 1981, 1983, 1984; Bakradze and Chkhikvadze 1977; Zerova and Chkhikvadze 1984; Yefimov and Chkhikvadze 1987) have made it possible to revise the zoogeography of the region. Darevsky (1957) singled out seven different groups of species and subspecies of the herpetofauna in the Caucasus, based on their origin. Among the species representatives of the region of interest to us, it is necessary to pay attention to Lacerta strigata (Asia Minor species), Emys orbicularis, Anguis fragilis, Coronella austriaca, Elaphe quatuorlineates sauromates, Natrix natrix (European boreal species), Testudo graeca, Natrix tessellata (Mediterranean species), Pseudopus apodus, Coluber najadum (east- Mediterranean species), and Lacerta saxicola, L. praticola, L. derjugini, L. media (autochthonous species). The Colchis, however, was not distinguished as an independent center of speciation in this work. Scherbak (1981) included the Colchis in the Caucasian Region of the Mediterranean Province. He suggested that the typical species of the region were Mertensiella caucasica, Pelodytes caucasicus, Lacerta saxicola-complex and others. However, the Colchis proper was again not distinguished as an independent center of herpetofaunal formation. For the analysis of the herpetofauna of the Colchis proper, it is necessary to exactly define the term "the Colchis phytolandscapes", and to decide what types of vegetation are universally recognized as "Colchis types". Albov (1885) was the first to clearly depict plant landscapes of the Colchis. He singled out a region, unique for Russia, of mountain limestone flora which had been developing mainly autochthonously in a large refugium with numerous endemic and relict species and even genera. Kolakovskiy (1980) regarded the Colchis flora as basically forest and alpine-meadow, and suggested that its main phytolandscapes had existed since old times with changes only in the composition of their edificators, except for the extinct formation of evergreen subtropical forests in the lower mountain belt. The tertiary- relict character of the forest mesophile flora and vegetation is fully revealed here due to slight changes in this region's climatic conditions (Kuznetsov 1891). The most characteristic features of the tertiary-relict Colchis forest are: extraordinary abundance and diversity of tree and shrub species, impossibility of singling out the dominant species (which is also characteristic of tropical forest with extreme density of trees), abundance of vines and epiphytes, and almost total absence of grass cover. All these attributes make the Colchis forest similar in many aspects to a tropical rain forest (Pavlov 1984). According to Sinskaya (1933), the Colchis forest vegetation underwent three main stages of development: the tropical forest; the forest of Colchis type, but rich and covering a wider area; and last, the modern Colchis April 1990 Asiatic Herpetological Research Vol. 3, p. 69 forest. The Colchis type of vegetation includes a number of phytocenoses differing in structure, composition, and ecological peculiarities: it may be mixed (polydominant), or may be presented by cenosis of one or two species, but the common and obligatory attribute of phytocenosis of the Colchis type is an abundance of tertiary relicts. The area with Colchis type vegetation is characterized by a rather monotonous thermal regimen, corresponding to a subtropical climate, but with highly diverse soils (Gulisashvili et al. 1975). The herpetofauna of the Colchis forests is surprisingly poor, despite its uniqueness. The species composition is different in the southeastern and northwestern parts of the Colchis compared to the other portions of its territory. Species such as Mertensiella caucasica, Lacerta clarkorum, L. parvula, and L. mixta, whose distributions are connected with forests growing on acid soils above volcanic rocks, are found on the western slopes of the Adzharo- Imeretinsky, Shavshetsky and Lazistansky (Pontic) mountain ranges. Similarly, floristic endemics of this part of the Colchis {Rhododendron ungernii, Osmanthus decorus, Betula medwedewii, Epigaea gaultheriodes, and others), are Adzharo- Lazistan endemics, sometimes with slight radiations to the adjoining regions, but are not Colchis endemics in the broad sense of the word. The same applies to Lacerta saxicola darevskii and L. saxicola brauneri which are widespread in the northwestern part of the Colchis, but are absent in the central and southeastern Colchis. These animals, by analogy with floristic endemics {Allium candolleanum, Campanula mirabilis, C. bzybica, C. calcarea, C. jadvigae, Genista abchasica, Gentiana paradoxa, Omphalodes kusnetzovii, and others) are northern-Colchis endemics. For example, of the 450 endemic Colchis species of flora, 83 (25%) are endemics of the northern Colchis (Adzinba 1980). Triturus vittatus ophryticus, T. vulgaris lantzi, Bufo verrucosissimus, Pelodytes caucasicus, Lacerta derjugini, L. agilis grusinica, Natrix megalocephala, and Vipera kaznakowi are Colchis endemics in the broad sense of the word. In addition to the Colchis endemics, there are three more ecological-geographical groups of amphibians and reptiles in the region. They have similar ecological characteristics (habitat first of all), and overlapping geographic distributions. 1. The East-Mediterranean group consists of Triturus cristatus karelini, Testudo graeca nikolskii (Fig. 1 ), Lacerta media, L. praticola pontica, L. strigata, Pseudopus apodus tracius, Natrix tessellata, and Coluber najadum. This group's distribution includes either the Balkans and the Caucasus or the Balkans, Crimea, and the Caucasus. According to ecological characteristics, these are xeromesophiles or hemixerophiles whose spreading is related to dry foothills of Western Transcaucasia up to 200-300 m above sea level with an annual sum of temperatures exceeding 5000°C. Thus, Testudo graeca, Pseudopus apodus, and Coluber najadum occur in the Colchis on a narrow seaside strip of land with enclaves of Mediterranean vegetation from Tuapse to Pitsunda-Sukhumi. A local population of L. strigata occurs in the Pitsunda region, and a local population of L. media occurs in the environs of Pitsunda and Salme. The majority of localities of L. praticola and Natrix tessellata in the Colchis are in the seaside hills up to 400 m above sea level. It is only along the valleys of large rivers like the Shakhe River, the Mzymta River, the Bzyb River, and others that N . tessellata penetrates into the Colchis up to 600 m above sea level. Thus the majority of these species are found either in places with vegetation of the Mediterranean type, or in places where the initial Colchis vegetation has been reduced to zero and the landscapes resemble the Mediterranean ones by their thermo-biotopic conditions (substitute shibliaks and tomillares, Pitsunda pine groves, foothill post-forest glades, and landplots with Erica tetraliz, Juniperus oxicedrus, Pinus pityusa, and Arbutus andrachne). Vol. 3, p. 70 Asiatic Herpetological Research April 1990 FIG. 1. Testudo graeca nikolskii is a Mediterranean species, and in the Colchis it is found only in the seaside strip of land with enclaves of Mediterranean vegetation. 2. The Caucasian group includes Hyla arborea schelkownikowi (Fig. 2), Rana macrocnemis, Lacerta caucasica alpina, L. rudis, and Vipera dinniki. Distributions of these species in the Caucasian Isthmus are broader than those of the Colchis group. At the same time, the majority of them are Colchis autochthons. These species are mesophiles and occur in mesophillous forest and mountain meadow formations. This group seems to be of Colchis origin, retaining close connections with the main center of the formation. Broader ecological tolerance in comparison with typical Colchis species makes it impossible to include them within the Colchis group. 3. The European group consists of Bufo viridis, Rana ridibunda, Emys orbicularis, Anguis fragilis, Natrix natrix, Coronella austriaca, and Coluber jugularis caspius. The composition of this group is not homogeneous. It includes both species typical for the steppe areas {Bufo viridis and Coluber jugularis ) and those that are widely distributed in Europe (all the rest). Only Anguis fragilis and Coronella austriaca are widely distributed in the Colchis. This makes it difficult to definitely consider them late migrants to the Caucasus. Other species either occur in several spots along the Colchis seashore (B. viridis and Natrix natrix ) or populate a narrow strip of land along the sea together with the Mediterranean species (C. jugularis ) or a somewhat wider strip (Emys orbicularis and Rana ridibunda ). Despite the possibility of finding these species in the typical Colchis forest formations, the majority of them are still attributed to the Mediterranean type of vegetation. April 1990 Asiatic Herpetological Research Vol. 3, p. 71 FIG. 2. Hyla arborea schelkownikowi seems to be of a Colchis origin, but because of its broad ecological tolerance, it is found in a large part of the Caucasian Isthmus. Let us consider the dispersal and distribution of the representatives of the Colchis group in detail. Triturus vittatus occurs in the territory from the seashore to the subalpine meadows in all the forest types. In the place called "the Colchis Gates" (lowering of the Main Caucasian Range between Mt. Fisht and Mt. Chugush) the species crosses over to the northern slope of the Western Calucasus, reaching the environs of Goriachij Klutch and Krasnodar in the northwest and the basin of the Laba River in the northeast. In the eastern part of the area, the species crosses the Adzharo-Imeretinskij mountain range and reaches the outskirts of Tbilissi- Oni. It occurs in the Lagodekhi region as an isolate. Outside the boundaries of the Colchis this species occurs either in the Colchis type forests or in their derivatives. Triturus vulgaris lantzi (Fig. 3) occurs in the same places in the Colchis as T. vittatus ophryticus does. Very often both species are symbiotopic (Tuniyev and Beregovaya 1986). On the northern slope of the Western Caucasus, its home range is wider than that of the previous species, but it is only through the mesophillous forests and subalpine meadows that it penetrates into the Eastern Transcaucasia up to the Trialet Ridge. The isolated population in Talysh occurs in the Hirkan forests which are ecologically and genetically close to the Colchis forests. Bufo verrucosissimus (Fig. 4) occurs in all parts of the Colchis from the sea shore up to the subalpine forests. On the northern slope of the Western Caucasus, it is found on the territory from the environs Vol. 3, p. 72 Asiatic Herpetological Research April 1990 FIG. 3. Triturus vulgaris lantzi. This subspecies is widespread throughout the Colchis. of Krasnodar in the west to the Psebaj settlement and the Shakhgirej Canyon in the east, where it is found in the derivatives of the Colchis forests from 400 m to 1000 m above sea level. In the Eastern Caucasus it is found in the Borzhomy Gorge, the Lagodekhi-Zakataly, and the Talysh region, where its distribution is limited to the mesophilous forests, which are abundant in the Colchis and Hirkan floral elements. The distribution of Pelodytes caucasicus (Fig. 5) in the Colchis is more restricted. It does not occur in the coastal belt and oak- forests. It is found both in the mesophilous beech, chestnut, and fir-tree forests, and in mixed broad-leaved forests with an evergreen understory. On the northern slope of the Western Caucasus, its distribution coincides with that of B. verrucosissimus, but unlike the latter, it is not found in deforested places. Its southeastern distributional limits also coincide with that of Bufo verrucosissimus. Pelodytes caucasicus does not occur east of the Trialet Ridge. There is an isolated population in the mesophilous forests in the Lagodekhi-Zakataly region. Lacerta derjugini has a distribution in the Colchis similar to P. caucasicus. It reaches the sub-alpine belt. On the northern slope of the Western Caucasus, it is found in the Colchis forest derivatives from the Belaja River to the Small Laba River (the Shakhgirej Gorge). The species penetrates through the Eastern Transcaucasia up to the Trialet Ridge. Separate populations occur in north-eastern Georgia up to Lagodekhi-Zakataly. Lacerta agilis grusinica (Fig. 6) is known to occur only on the territory of the Colchis and the adjoining sea coast up to April 1990 Asiatic Herpetological Research Vol. 3, p. 73 •^ •* |^^w — * — i 1^^ * 1 - f&jj& J ■si 9b. v^ 5* *3| I f' • ■ ■ k J 1 <■ '% if? * HI* 1 "... FIG. 4. Bu/o verrucosissimus. This Colchis endemic occurs in the four Colchis refugia. Novorossijsk. Its vertical distributional limit is 700 m above sea level, though local populations can be found in the sub-alpine belt (Mt. Aishkho and Mt. Uglovoj). The distribution of Natrix megalocephala is similar to that of many Colchis species. It is found from the environs of Tuapse and Gorjachij Kljutch in the west, to the region between the Belaja Laba River and the Small Laba River in the north, eastwards through the whole territory of the Colchis up to the Borzhomy Gorge and separately in the Lagodekhi-Zakataly region (Orlov and Tuniyev 1986b). In the Colchis it reaches the sub-alpine belt. In the rest of the area it does not exceed 1000 m above sea level. Elaphe longissima (Fig. 7) occurs in the region of Novorossijsk and throughout Western Transcaucasia, except for the mid- mountains and highlands. There are isolated populations in the Borzhomy Gorge, the Lagodekhi-Zakataly region and the Belaja River basin on the northern slope of the Western Caucasus. Vipera kaznakowi (Fig. 8) occurs throughout the territory of the Colchis up to 1000 m above sea level. On the northern slope of the Western Caucasus it occurs between the Belaja River and the Small Laba River. Separate populations are known in the Borzhomy Gorge and the Lagodekhi region. The dispersal and distribution of Lacerta saxicola darevskii (Fig. 9), L. s. brauneri, L. mixta, L. parvula and Mertensiella caucasica have been analyzed above. A comparison of the endemic Colchis species, whose distributions are associated Vol. 3, p. 74 Asiatic Herpetological Research April 1990 FIG. 5. Pelodytes caucasicus occurs in all four Colchis refugia. with forest and meadow formations of the Colchis type, makes it possible to identify three more regions in the Caucasian Isthmus, besides the Western Caucasus (the Colchis proper), in which the Colchis herpetofauna occurs. The three regions are: the Bjelo-Labinskij region on the northern slope of the Western Caucasus, the Kakhetinskij (Lagodekhi-Zakataly) region on the southern slope of the Eastern Caucasus, and the Borzhomskij region in Eastern Transcaucasia (Fig. 10). The comparative composition of the herpetofauna of these regions is shown in Table 1. It is evident from Table 1, that the most significant differences are found between the Colchis and the Kahetinskij regions and the least significant differences are between the Colchis and the Borzhomskij and the Belo-Labinskij regions. Taking into account the above mentioned peculiarities of distribution of herpetofauna within the Colchis refugium itself, the differences become even less significant. In this case, we deal with three regions smaller in space, and a wealth of species of the Colchis herpetofauna that occur in refugia and have survived off the main territory of the Colchis. The Belo-Labinskij region is only conventionally separated from the Colchis by the crest of the Main Caucasian Mountain Range. All the northern Colchis species, except L. a. grusinica, occur on the northern slope of the Western Caucasus in the Belaja and the Small Laba river drainages. It should be stressed that this unity is based on the fact that the characteristic Colchis elements of flora and vegetation cross over the Main Caucasian Range in the place known as "the Colchis Gates" to its northern slope. The basins of April 1990 Asiatic Herpetological Research Vol. 3, p. 75 / FIG. 6. Lacerta agilis grusinica is found throughout the main Colchis refugium. TABLE 1 . Distribution of the endemic Colchis herpetofauna in the main refugia of the Caucasian Isthmus. Species R Colchis E G Bjelo- Labinskij I O N Kakhetinskij Borzhomskij Triturus vittatus T. vulgaris lantzi Mertensiella caucasica Bufo verrucosissimus Pelodytes caucasicus Lacerta derjugini L. agilis grusinica L. saxicola darevskii L. s. brauneri L. mixta L. parvula L. clarkorum Natrix megalocephala Elaphe longissima Vipera kaznakowi + + + + + + - + + - - + + + + + + + + + + + + + + - - - + + - - + + - - + - - + + - - + + - - - + + + + + + + + + + + + Total: 15 10 11 Vol. 3, p. 76 Asiatic Herpetological Research April 1990 FIG. 7. Elaphe longissima. The Colchis refugia populations are disjunct from the main distribution in Europe. the Belaja, Tsitse and Laba rivers abound in those elements. To the north of these watersheds their distributions are continuous up to the Skalistij (Rocky) limestone ridge. Maleyev (1939) has noted that a part of the Maikop district abounds in Colchis elements and is inseparable from the Colchis according to the character of its flora and vegetation. The Borzhomskij region is also conventionally separated from the Colchis by the Adzhara-Imeretinskij Mountain Ridge. The flora and vegetation of the Baniskhevskoje Gorge, the Likanskoje Gorge, and the upper belt of Mt. Lomis- Mta, as well as the environs of Bakuriani, hardly differ from those of the Colchis. April 1990 Asiatic Herpetological Research Vol. 3, p. 77 FIG. 8. Vipera kaznakowi occurs throughout the territory of the Colchis up to 1000 m above sea level. In the isolated eastern Kakhetinskij region a considerable number of the ancient Tertiary vegetation representatives survived due to the warm and humid climate (Gulisashvili et al. 1975). Modern distributions of eco-geographic groups of amphibians and reptiles distinguished by our scientists have distinct altitudinal-ecological limits owing to natural and historical reasons. Migration of ancestoral species of the Colchis and the Caucasian groups from the south apparently took place in the Miocene when the Caucasian island joined vast territories of Asia Minor. Colonization of the Caucasus from the south by different species of mammals during the early Miocene has been studied by Vereschagin (1958), and that of lizards of the Podarchis- Archaeolacerta group by Darevsky (1967). The early Miocene was about the time of the formation of the Caucasian Mountains (Bogachev 1938). The warm subtropical climate and vegetation in the Caucasus favored the evolution and dispersal of heat and mesophilic forms (Triturus vittatus ophryticus, T. vulgaris lantzi, Pelodytes caucasicus, Bufo verrucosissimus, Lacerta saxicola darevskii, L. s. brauneri, L. derjugini, Elaphe longissima, Natrix megalocephala, and Vipera kaznakowi ), as well as species with a broader ecological tolerance (Rana macrocnemis, Hyla arborea schelkownikowi, Lacerta rudis, and L. agilis grusinica ). Fossil remains of mammals (Mesocricetus, Prometheomys, Sorex, Talpa ) [Vereschagin 1958] and insects (Orthoptera, Hemiptera, Blattoidea, and Vol. 3, p. 78 Asiatic Herpetological Research April 1990 FIG. 9. Lacerta saxicola darevskii. This lizard is a Colchis endemic whose distribution is restricted to the northwestern part of the Colchis. Coleoptera) [Rodendorf 1939] suggest a good food supply for amphibians and reptiles during the Miocene. It was also in the Miocene that the majority of these species reached the eastern-most parts of the Greater Caucasian Range along its southern slopes and penetrated from there into the Talysh across the so-called "Karabakhi Bridge". Safarov (1966), and other scientists, have studied the former direct relations between the Colchis and the Hirkan floras. Even at the present time, the floristic composition of the Kakhetinskij region and of the Karabakh has many common features with that of the Colchis and the Talysh forests (Arushanyan 1973; Sokolov 1977; Takhtadzhan 1978; Gadzhiyev et al. 1985). The end of the Tertiary period was characterized by damping of tectonics due to the broad correlation of the Caucasus and the Balkans (Vereschagin 1958) and the formation of the steppe landscapes along the northern Black Sea coast (Pidoplichko 1954; Scherbak 1966). During that period, such South-European species as Rana ridibunda, Bufo viridis, Emys orbicularis, Anguis fragilis, Coluber jugularis, and Coronella austriaca seem to have penetrated to the Precaucasia from the west. At the same time, such species as Testudo graeca, Pseudopus apodus, Triturus cristatus karelini, Lacerta praticola pontica, L. media, Coluber najadum, and Natrix tessellata got into the Colchis from the west along the Black Sea coast. Early and Middle Pliocene should be considered the beginning of initial fragmentation of the Colchis faunal areas when the Greater and the Lesser Caucasian April 1990 Asiatic Herpetological Research Vol. 3, p. 79 FIG. 10. The main refugia of the endemic Colchis herpetofauna. Mountain ranges underwent substantial glaciation (Grozdetskiy 1954; Markov et al. 1965). The main center of dispersal of these species was the Colchis, where relatively heat-loving vegetation of the Caucasian type survived even during the periods of the extreme Pleistocene cooling (Vereschagin 1958; Adamyants 1971). Along with the Colchis, smaller refugia sporadically survived on the territory of the Caucasus Black Sea coast, and also on the northern slope of the Main Caucasian Range between the Pshekha River and the Small Laba River. The present distribution of the Tertiary vegetation of the Colchis type in the western Caucasus testifies to this (Kharadze 1974; Pechorin and Lozovoy 1980; Kholyavko et al. 1978; Adamyants 1971; Koval and Litvinskaya 1986). It was in the narrow humid gorges with a relatively constant thermal regimen that the representatives of the Colchis group remained intact. At the same time, independent populations might also have been preserved in mid-mountain areas where refugia of the Colchis vegetation exist in the vicinities of the Fisht-Oshtenskij Mountain Massive, the Lagonaki Plateau and even in the Central Caucasus (Kholyavko et al. 1978; Kharadze 1974). Small refugia seem to have also remained on the southern slope of the eastern pan of the Greater Caucasus and in the Kuru River Gorge. It is indisputable that the majority of the mountainous populations of Colchis species perished during the Pleistocene and Vol. 3, p. 80 Asiatic Herpetological Research April 1990 that the ones that survived in refugia have been accumulating unique characteristics which led to different geographical forms (subspecies) on different slopes of the Main and the Adzharo-Imeretinskij mountain ranges. The data presented by Takhtadzhan (1946) and Maruashvili (1956) support the hypothesis concerning the preservation of relict Colchis species in the mountains. According to their data, the average annual temperature during the glacial periods decreased not more than 1.5-2.0°C, while precipitation amounted to not less than 1500-2000 mm. Darevsky (1967) considers that this argument supports the possibility of foothill refugia of reptiles existing on the sea facing slopes of the Gagrinskij and the Bzhybskij mountain ranges, and in other regions, despite radical reorganization of the distributions of all the species of plants and animals in connection with glaciation. During the interglacial and especially the postglacial periods, reconstruction of all vegetation belts took place (Vereschagin 1958). This favored the isolation of the species of the Colchis and the Caucasian groups in the above-mentioned refugia, but favored wider dispersal of the European and the Mediterranean groups in Transcaucasia. In the northwestern part of the Caucasus Black Sea coast, mesophilic vegetation gave way to xerophytic vegetation of the Mediterranean type. Plant formations of this type with the prevalence of Juniperetum, Querceto, Pinetum carpinulosum, Pinetum fruticosum and shibliaks are characteristic of the Anpa- Gelendzhik region. Enclaves of Mediterranean vegetation remained still further to the south, up to Pitsunda (Takhtadzhan 1978; Kolakovskiy 1961). When the xerothermic period ended, the climate became more humid again. This favored the reestablishment of the former borders of the forest belt (Vereschagin 1958). Subalpine meadows and elfin woodlands were expanding throughout the whole subalpine belt of the southern slope of the Main Caucasian Mountain Range from the Central Caucasus to the Fisht- Oshtenskij Mountain Massive (Kholyavko et al. 1978; Kharadze 1974; Dolukhanov 1974; Galushko 1974). On the northern slope such vegetation is present in the western parts and changes its character to the east, transforming into a steppe type (Lavrenko 1980). Modern areas of Vipera dinniki and Lacerta caucasica alpina have fixed distributional limits in the subalpine belt influenced by the warm Black Sea. Final settling of the present-day climate favored fixation of the Colchis species distributions, with their distinct populations exceeding the bounds of the refugia. This was coupled by simultaneous depression and reduction of the European and especially of the Mediterranean species distributions. Concluding this review of the Colchis herpetofauna and their main refugia, it is necessary to enumerate the most important characteristics. The Colchis species are characterized by antiquity (conservation since the Tertiary period), Autochthonity, depression — for some species {Vipera kaznakowi; Lacerta clarcorum, L. a. grusinica), existence of the northern- Colchis limestone, and the southern- Colchis volcanic centers of formation of narrow-endemic forms. Reptiles have a common tendency toward melanism, while amphibians approach their low temperature thresholds. These are the adaptive features acquired during the glacial period. As a rule, the modem distribution of the Colchis species does not exceed the bounds of the Colchis vegetation refugia or their derivatives. The maximal vertical distribution is in the center of the Colchis up to 1800 m above sea level, while in the other portions of the refugia it does not exceed 1000 m above sea level as a rule. The existence of four refugia of the Colchis herpetofauna in the Caucasian Isthmus is determined by natural factors of high order- these are the areas with slightly changed climatic conditions characterized by modern crossing of the January -3°C isotherm and 800 mm isohyet. April 1990 Asiatic Herpetological Research Vol. 3, p. 81 Literature Cited ADAMYANTS, G. I.. 1971. [On Castanetum of the Caucasus]. A Report of the Sochi Department at the USSR Geographical Society. Issue 2. Leningrad. (In Russian). ADZINBA, Z. I. 1980. [The plants of Abkhazia from limited area]. Proceedings of the Botanical Gardens. Sukhumi, Tbilisi, Georgia Issue 26:69-82. (In Russian). ALBOV, N. M. 1885. [Papers on the flora of Kolkhida]. Proceedings of the Tiflis Botanical Gardens. Tiflis, Georgia Issue 1:1-288. (In Russian). ARUSHANYAN, R. I. 1973. [The relicts of Nagomy Karabakh and their new findings]. Botanical Journal Leningrad 58(5):700-706. (In Russian). BAKRADZE, M. A. AND V. M. CHKHIVADZE 1977. [Fossils of the tortoises from Testudo genus in the USSR]. Prosveschenye Publishing House, Moscow. 369pp. (In Russian). BANNIKOV, A. G., I. S. DAREVESKY, V. G. ISCHENKO, A. K. RUSTAMOV, AND N. N. SCHERBAK. 1977. [Field guide of the USSR amphibians and reptiles]. Prosveschenye Publishing House, Moscow. 369pp. (In Russian). BARTENEV, A. AND M. REZNIKOVA. 1931. NEE LAC ERT A -Formen (Reptilia) aus dem Kaukasichen Staatsnaturschtzgebiete (Westkaukasus). Zoologischer Anzeiger 96(9- 10):268-271. BARTENEV, A. AND M. REZNIKOVA. 1935. [Papers on the fauna of amphibians and reptiles from the western and central parts of the State Caucasus Reserve]. Bulletin of the Museum of Georgia 8:1 1-41. (In Russian). BISCHOFF, W. 1982. Zur Kenntnis der innerartlichen Gliederung der Artwiner Eicechse, Lacerta derjugini Nikolskij, 1898. Zool. Abh. Bd. 38, Nr. 1 S. 1-51. BISCHOFF, W. 1984. Bemerkungen zur innerartlichen Gliederung und zur Verbreitung der Artwiner Eidechse (Lacerta derjugini Nikolskij, 1989) an den Sudhangen des GroBen Kaukasus. Salamandra, Bd. 20, Nr. 2/3, S. 101-111. BOGACHEV, V. V. 1938. [On the zoogeography of the Caucasus and its geological history]. Materials of the Azerbajzhan Affiliation. The USSR Academy of Sciences, Moscow. 4:29- 47. (In Russian). BRAUNER.A. 1905. [Preliminary data on reptiles and amphibians from the Crimea Kubanskaya region, Volynskaya and Varshavskaya gubernias]. Notes of the Novorossiskoye Society of Researchers. Odessa, Russia. 28:1- 88. (In Russian). CHKHIKVADZE, V. M. 1981. [Survey on the data of amphibian and reptile neogene fossils from the northern area adjacent to the Black Sea coast]. Science Publishing, Leningrad, pp. 151- 152. (In Russian.) CHKHIKVADZE, V. M. 1983. [Fossil turtles of the Caucasus and northern territory adjacent to the Black Sea coast]. Metsniyereba Press, Tbilisi, Georgia 48 pp. (In Russian). CHKHIKVADZE, V. M. 1984. [Survey on the fossils of tailed and tailless amphibians of the USSR]. Academic Press, Georgian Academy of Sciences, Tbilisi 10(1):5-13. (In Russian). DAREVSKY, I. S. 1957. [The fauna of the Armenian reptiles and their zoogeographical analysis]. Ph. D. Thesis. Yerevan, Armenia. 28 pp. (In Russian). DAREVSKIY I. S. 1967. [Rock lizards from the Caucasus (Lacerta saxicola).} Science Publishing House, Leningrad. 210 pp. (In Russian). DAREVSKY, I. S. AND V. I. VEDMEDERJA. 1977. [A new species form of the Lacerta saxicola group from southeastern Turkey and adjacent areas of Adzharia]. Proceedings of the Zoological Institute, USSR Academy of Sciences, Leningrad 74:50-54. (In Russian). DERJUGIN, K. M. 1899. [Report on a field trip and zoological research in Charivski territory, southwestern Transcaucasia and the vicinity of Trapezond]. Proceedings of Saint Petersburg Society of Researchers, St. Petersburg 30(2): 1- 65. (In Russian). DOLUKHANOV, A. G. 1974. [Subalpic landscapes of the Caucasus as shelters for relict fauna elements]. Problems of Botanies, Leningrad 12:27-34. (In Russian). Vol. 3, p. 82 Asiatic Herpetological Research April 1990 GADZHIYEV, V. D., A. M. ALEKPEROV, M. R. EFENDIJEV, AND R. K. MUSTAFAJEVA. 1985. [The Zakatalski Reserve]. Agroprom Press, Moscow. 184 pp. (In Russian). GALUSHKO, A. I. 1974. [Relicts of the highland flora of the western part of the Central Caucasus]. Pp. 19-26. In Problems of Botanies, Vol. 12, Flora of Highlands and its Exploration. Science Publishing Leningrad. (In Russian). GOLUBEV, N. S. 1980. [On the habitat of Pelodytes caucasicus (Amphibia, Pelobatidae)]. Zoological News, Kiev, Ukraine (3):52-55. (In Russian). KOVAL, I. P. AND S. A. LITVINSKAYA. 1986. [Rare vegetation communities on the Krasnodar territory]. Rostov State University Press, Rostov-upon-Don, USSR, part 3, 336 pp. (In Russian). KUZNETSOV, N. I. 1891. [Elements of the Mediterranean area in Western Transcaucasia]. Notes of the Russian Geographical Society, St. Petersberg 23(3): 1-190. (In Russian). LAVRENKO, Y. M. 1980. [Caucasus montane and meadow steppes and stepped meadows]. Pp. 269-272. In Vegetation of the European part of the USSR, Science Publishing, Leningrad. (In Russian). GOLUBEV, N. S. 1985. [Caucasus land toad, Pelodytes caucasicus Boulenger (distribution, morphology and ecology)]. Ph.D. Thesis. Leningrad. 25 pp. (In Russian). GROZDETSKIY, N. A. 1954. [Physical geography of the Caucasus]. Moscow State University Press, Moscow 2(1): 1-208. (In Russian). GROZDETSKIY, N. A. 1958. [Physical geography of the Caucasus]. Moscow State University Press, Moscow 2:(2)l-264. (In Russian). GUUSASHVILI, V. Z., L. B. MAHATADZE, AND L. I. PRILIPKO. 1975. [Vegetation of the Caucasus]. Science Publishing House, Moscow. 223 pp. (In Russian.) KHARADZE, A. L. 1974. [On some floragenetic endemic groups in the Big Caucasus]. Problems of Botanies, Science Publishing House, Leningrad 12:70-76. (In Russian). KHOLYAVKO, V. S., D. A. GLOBA- MIKHAILENKO, AND E. S. KHOLYAVKO 1978. [The tree atlas of the Caucasus]. Moscow. 215 pp. (In Russian). KHOZATSKY, L. I. 1941. [Sea storms and some land living animals]. Wildlife, Moscow 1:112- 113. (In Russian). KOLAKOVSKIY, A. A. 1961. [Vegetation of Kolkhida]. Materials on the USSR fauna and flora. MOIP Press, Moscow, Issue 10. 460 pp. (in Russian). KOLAKOVSKIY, A. A. 1980. [The flora of Abkhazia]. Metsniyereba Publishing, Tbilisi, Georgia 1:7-18. (In Russian). LUKINA, G. P. 1965. [On the taxonomy of the legless lizard]. Herpetology, Tashkent, Uzbekistan, USSR, pp. 6-10. (In Russian). MALEYEV, V. P. 1941. [Tertiary relicts in the flora of the Western Caucasus and major stages of the Quaternary history of its USSR flora]. The USSR Academy of Sciences Press, Moscow, pp. 61-144. (In Russian.) MARKOV, K. K., G. I. LAZUKOV, AND V. A. NIKOLAYEV. 1965. [The Caucasus]. Pp. 306- 321. In Quaternary Period, v. 1, Territory of the USSR, part 2. Area of Ancient Glaciation of High Mountains of the USSR South, chapter 2, Moscow University Press, Moscow. (In Russian). MARUASHVILI, L. I. 1956. [The reason to review the present day concepts about paleogeographical conditions of the glacier period in the Caucasus]. Press of the Institute of Geography named after Vakhushti, Tibilisi, Georgia. 113 pp. (In Russian.) MILYANORSKIY, Y. S. 1957. [On the snake fauna of Abkhazia]. Proceedings of the Abkhazski State Museum, Georgia 2:199-203. (In Russian). NEGMEDZYANOV, V. A. AND M. A. BAKRADZE. 1977. [On the herpetology of Kolkhida]. Pp. 153-155. In Questions of Herpetology, Leningrad. (In Russian). NESTEROV, P. V. 1911. [Materials on the herpetology of the southwestern Transcaucasia (Chokhotski Territory)]. Annual Journal of the Zoological Museum, Academy of Sciences, St. Petersburg 10(1): 1-18. (In Russian). April 1990 Asiatic Herpetological Research Vol. 3, p. 83 NIKOLSKY, A. M. 1911. [Reptiles and amphibians of the Caucasus]. The Caucasus Museum Press, Tiflis, Georgia 10(1): 1-18. (In Russian.) NORDMANN, A. 1840. Cataloque raisonne des mammiferes de la faune Pontique. Voyage dans la Russia meridionale et la Crimec, par la Hongric, la Valachie et la Moldavic, vol. 3, s. 535-551. ORLOV, N. L. AND B. S. TUN1YEV. 1986a. [Present ranges, possible ways of their formation and phylogeny of the three species of vipers from the Eurosiberic gTOup (V . kaznakowi complex) in the Caucasus]. Proceedings of the Zoological Institute, USSR Academy of Sciences, Leningrad 157:104-135. (In Russian). ORLOV, N. L. AND B. S. TUNIYEV. 1986b. [A new species of the grass snake Natrix megalocephala sp. nov from the Caucasus]. Pp. 116-130. In Proceedings of the Zoological Institute, Leningrad, Vol. 12. Leningrad. (In Russian). ORLOV, N. L. AND B. S. TUNIYEV. 1990. Three species in the Vipera kaznakowi complex (Eurosiberian Group) in the Caucasus: their present distribution, possible genesis, and phylogeny. Asiatic Herpetological Research 3:1-36. ORLOVA, V. F. 1973. [Herpetofauna of the northern area of the State Caucasus Reserve]. Zoological News, Kiev, Ukraine 2:61-65. (In Russian). ORLOVA, V. F. 1978a. [Geographical variation and lektotype description of Lacerta derjugini Nikolski]. Pp. 188-203. In Birds and Reptiles. Moscow State University Press, Moscow. (In Russian). ORLOVA, V. F. 1978b. [Geographical distribution and intraspecific variation of Lacerta praticola from the Caucasus. Pp. 204-215. In Birds and Reptiles. Moscow State University Press, Moscow. (In Russian). PAVLOV, N. B. 1984. Botanic geography of the USSR, Moscow, pp.13-17. (In Russian). PESHORIN, A. I. AND S. P. LOZOVOY. 1980. [Natural monuments of the Krasnodar Territory]. Krasnodar Press, Krasnodar. 138 pp. (In Russian). PETERS, G. 1960. Die Grusinische Zauneidechse Lacerta agilis grusinica nomen novum. Zool. Anz. Bd. 165, Nr. 7/8, S. 113-126. PIDOPLICHKO, I. G. 1954. [On the glacier period]. The Ukraine Academy of Sciences Press, Kiev, Ukraine. 264 pp. (In Russian). RODENDORF, B. B. 1939. [On the Miocene fauna of insects in the vicinity of Voroshilovsk]. Priroda Publishing, Moscow, N. 12. (In Russian). SAFAROV, I. V. 1966. [Dzelkva, its biological and ecological peculiarities and industrial significance]. Academic News, Azerbajzhan 1:15-17. (In Russian). SATUNIN, K. A. 1912. [On zoogeographical circles of the Caucasus Territory]. The Caucasus Museum News, Tiflis, Georgia 7:(1)1-106. On Russian). SCHERBAK, N. N. 1966. [Amphibians and reptiles of the Crimea]. Naukova Dumka Publishing House, Kiev, Ukraine. 239 pp. (In Russian). SCHERBAK, N. N. 1981. [Fundamentals of herpetogeographic regional divisions of the USSR territory]. Pp. 157-158 In Questions of Herpetology, Leningrad. (In Russian). SCHERBAK, N. N. AND SCHERBAN. 1980. [Amphibians and reptiles of the Ukrainian Karpat Mountains]. Naukova Dumka Publishing House, Kiev, Ukraine. 253 pp. (In Russian). SILANTYEV, A. 1903. [Materials on the herpetology of the Black Sea region]. Annual journal of the Zoological Museum, Academy of Sciences, St. Petersbourg 8:1-10. (In Russian). SINSKAYA, Y. N. 1933. [Major evolutional characters of the forest vegetation in the Caucasus with regard to the specific history]. Botanical loumal, Leningrad 18(5-6): 117- 130. (In Russian). SOKOLOV, S. Y. 1977. [The habitats of trees and bushes of the USSR]. Science Publishing House, Leningrad 1:1-163. (In Russian). TAKHTADZHAN. 1946. [On the history of the vegetation developments in Armenia]. Proceedings of the Botanic Institute, Armenian Academy of Sciences, Yerevan, Armenia 4:51- Vol. 3, p. 84 Asiatic Herpetological Research April 1990 107. (In Russian). TAKHTADZHAN. 1978. [Floristic areas of land]. Science Publishing, Leningrad. 219 pp. (In Russian). VEDMEDERJA, V. I., N. L. ORLOV AND B. S. TUNIYEV. 1986. [On the taxonomy of three viper species of the Vipera kaznakowi complex]. Proceedings of the Zoological Institute, Leningrad 157:55-61. (In Russian). TUNIYEV, B. S. 1983. [The herpetofauna of the southern part of the Caucasus State Reserve]. Pp. 84-94. In Conservation of the Relict Flora and Fauna in the Northwestern Caucasus. Leningrad. (In Russian). TUNIYEV, B. S. 1985. [Rare and endangered amphibians and reptiles of the Caucasus Reserve]. Pp. 104-116. In Ecological Researches in the Caucasus Biospheric Reserve. Rostov State University Press, Rostov-upon- Don. (In Russian). TUNIYEV, B. S. AND S. Y. BEREGOVAYA 1986. [Sympatric amphibians of yew-box groves]. Proceedings of the Zoological Institute, USSR Academy of Sciences, Leningrad 157:136-151. (In Russian). VEKUA, A. K., K. S. KALANDADZE, AND V. M. CHIKVADZE. 1979. [New paleontological findings from Belaya (White) Cave, Georgia, USSR]. Reports of the Georgian Academy of Sciences, Tbilisi, Georgia. 96(3):745-748. (In Russian). VERESCHAGIN, N. K. 1958. [Genesis of the terrain fauna of the Caucasus Isthmus]. Pp. 506-514. In Fauna of the USSR, Vol. 5. Montane Areas of the European Part of the USSR, Academic Press, Moscow. (In Russian). YEFIMOV, M. B., AND V. M. CHKHIVADZE. 1987. [Survey on crocodile fossils in the USSR]. Georgian Academy of Science News, Tbilisi, Georgia 13(3):200-206. (In Russian). TUROV, S. S. 1928. [Materials on the fauna of the State Caucasus Reserve]. Works of the North Caucasus Research Institutions Association, Rostov-upon-Don 2:1-15. (In Russian). ZEROVA, G. A. AND V. M. CHKHIVADZE. 1984. [Survey on cenozoic lizards and snakes from the USSR]. Georgian Academy of Sciences News, Tbilisi, Georgia 10(5):319-325. (In Russian). April 1990 Asiatic Herpetological Research Vol. 3, pp. 85-100 Studies of the Early Embryonic Development of Rana rugulosa Wiegmann jionghuaPan1 and Danyu Liang1 ^Department of Biology, South China Normal University, Guangzhou, Guandong, China Abstract. -This paper deals with the early embryonic developmental stages of Rana rugulosa Wiegmann, and with methods of artificial fertilization and experimentally accelerated development. Developmental stages are distinguished by morphological changes and obvious physiological features. At a temperature of 25.1 to 27°C it takes 70 hr 20 min to complete development from fertilized eggs to tadpoles with opercular folds. This course of development is divided into 25 stages, which are standardized with normal table equivalents. Key Words: Amphibia, Anura, Ranidae, Rana rugulosa, embryonic development, artificial fertilization, accelerated development, fertilized eggs, normal table. Introduction Rana rugulosa Wiegmann, whose popular name is the Field Chicken in China, is strong and big of body. It is famous for its delicacy of meat and the price is dear. Domestic and foreign markets are in great need of it. The growth and development of these frogs are relatively quick. Hatchlings can become sexually mature in one year. The average body length is about 1 10 mm. The weight is about 0.25 kg. Consequently, Rana rugulosa farming has been a growing enterprise. More papers deal with the embryonic development of R. guentheri Boulonger and R. catesbeiana. Currently there are few domestic and foreign studies on the embryonic development of R . rugulosa. In order to produce reference material for the captive breeding of this frog, we studied its early embryonic development, using artificial fertilization and higher temperatures to accelerate development experimentally, during the months of April and June, 1987 and 1988. Methods We made 1 1 observations on the embryonic development of Rana rugulosa. Six of the 11 were artificially fertilized. The parent frogs were bought in Guangzhou. The description of the features of embryonic development was mainly based on the six artificially fertilized eggs. The water temperature of the six developing embryos was 25.1 to 27°C; the pH was 6.5-6.8. Several minutes after oviposition and ejaculation or artificial fertilization, we put the oocyte into the laboratory. Then we used a microscope to observe the course of development and to measure the size and shape. The developmental stages were defined as beginning when half of the embryos in a sample showed all of the distinguishing characters of that stage. For every developmental stage we took 10 to 20 embryos and fixed them in 5% formalin and Bouin's fluid. This material was used in drawings, tissue-slices, and photomicrographs. The figures shown here were taken from observations of living material. Results The 25 stages of the early embryonic of Rana tigrina are as follows (time in parentheses is hours and minutes after fertilization): I) Oocyte stage. Unfertilized eggs are spheres 1.4-1.7 mm in diameter (Fig. 26). (00:25-00:30): The egg membrane absorbs water and expands. The eggs reach 2.6- 4.0 mm in perivitellic space. (00:50): The pigment crown of the animal pole extends downward upon the gray crescent (Fig. la, lb, lc, Id, le, 26). © 1990 by Asiatic Herpetological Research Vol. 3, p. 86 Asiatic Herpetological Research April 1990 2) 2 cell stage. (01:12): The first longitudinal cleavage progresses and a cleavage furrow in the animal pole appears, forming 2 equal hemispheres (Fig. 2a, 2b, 27). 3) 4 cell stage. (01:33): A cleavage furrow in the animal pole appears, perpendicular to the first cleavage furrow. It progresses to the vegetal pole and forms four equal sections (Fig. 3a, 3b, 28). 4) 8 cell stage. (01:42): The third cleavage furrow appears parallel to the equator, approximately bisecting the animal hemisphere, forming four small animal cells and four large vegetal cells (Fig. 4a, 4b, 29). 5) 16 cell stage. (01:48): Observed from the animal pole, the fourth cleavage forms eight blastomeres in two circular or elliptical tiers (Fig. 5a, 5b, 30). 6) 32 cell stage. (02:05): The fifth cleavage is horizontal. The animal pole has eight small cells; the vegetal pole has 8 large cells (Fig. 6a, 6b, 31). 7) Early blastula stage. (02:23): The embryo is at the large cell blastula stage, but the cells are still clearly distinguished. The blastocoel begins to appear in the middle part of the embryo near the animal pole (Fig. 7a, 7b, 32). 8) Mid-blastula stage. (02:50): There are many small blastomeres. The blastocoel continues to amplify (Fig. 8a, 8b, 33). 9. Late blastula stage. (03:18): The blastomeres are the color of a red bayberry. The cell boundary becomes indistinct and the blastocoel expands to its maximum (Fig. 9a, 9b, 34) 10) Early gastrula stage. (05:56): The pigmented crown epiboly of the animal cap occurs, extending to cover over 75% of the embryo. The dorsal lip forms. Involution of cells begins (Fig. 10a, 10b, 35). 11) Mid-gastrula stage. (07:00): The dorsal lip continues to amplify. The cells on the side of the lip are drawn into the future archenteron and a semicircle appears (Fig. 11a, lib, 36). 12) Late gastrula stage. (08:05): The blastopore shrinks and closes gradually. Eventually, the yolk plug is enclosed in the embryo (Fig. 12a, 12b, 12c, 12d, 37) 13) Neural plate stage. (10:30): The blastopore becomes a fissure, and the embryo begins to elongate along the longitudinal axis, becoming pear-shaped (Fig. 13a,.13b, 38). 14) Neural fold stage stage. (11 :50): A fold appears on both sides of the neural plate. The neural groove forms the median sinus of the embryo. The front of the plate is bigger, and the neural fold gradually approaches the dorsal axis from both sides. The embryo elongates to 1.9 mm (Fig. 14a, 14b, 39). 15) Cilial movement stage. (12:40): The neural folds are joining. The embryo rotates within the vitelline membrane (Fig. 15a, 15b, 40). 16) Nerve tube stage. (13:45): The neural tube has formed and the gill plate and the cement gland can be seen (Fig. 16a, 16b, 41). 17) Tail bud stage. (14:55): Two outer gill buds protrude on the side of each gill plate. The total length is 3.0-3.6 mm, and the length of the tail bud is 1/10 to 1/7 of this (Fig. 17a, 17b, 17c, 42). 18) Muscle effect stage. (15:55): Muscular response begins in most individuals. The olfactory organ appears, the cement gland is complete. Total length is 3.3-3.8 mm., and the length of the tail bud is W of this (Fig. 18a, 18b, 18c, 43). 19) Hatching stage. (16:45): The embryo hatches from the egg membrane and two outer gill budlets with 3-5 branches protrude obviously (Fig. 19a, 19b, 19c, 19d, 44). 20) Heart beat stage. (29:25): The heart April 1990 Asiatic Herpetological Research Vol. 3, p. 87 begins to move, the eyes protrude , the pair of otoliths can be seen, and the 3rd external gill matures (Fig. 20a, 20b, 20c, 45). 21) Open mouth stage. (31:10): The membrane covering the mouth splits to show the mouth cavity. The alimentary canal is complete and body segments are obvious. The inverted "V" shape myomeres appear on the side of the embryo. Body length is 5.8-6.2 mm. Tail fin length is half of the body length (Fig. 21a, 21b, 21c, 46). 22) Tail fin blood circulatory stage. (32:00): The cement gland begins to degrade, circulation begins in the tail bud, and tadpoles are able to swim in a straight line (Fig. 22a, 22b, 22c, 47). 23) Gill opercular fold stage. (35:55): The opercular fold appears in the base of the external gill and the intestines form a bow (Fig. 23a, 23b, 23c, 48). 24) Right side operculum closed stage. (55:30): The opercular fold stretches toward the right side, the right external gill is covered and forms the right internal gill. The left external gill is exposed. The intestines have 2-3 twists. Total length is 6-9 mm (Fig. 24a, 24b, 24c, 49). 25) Operculum completion stage. (70:20): The external gill in entirely enclosed in the gill cavity, the posterior of which has an exhalent pore to the left side. The yolk has almost been absorbed, and the tadpoles begin feeding. Total length is 6.8-10.5 mm (Fig. 25a, 25b, 25c). Table 1 shows N & F normal table equivalents (Nieuwkoop and Faber 1967) of our results. Discussion 1. The circumstances of the early embryonic development of Rana tigrina, R. catesbeiana, R. Umnochiris, and R. nigromaculata basically identical, but R. rugulosa has some minor differences in developmental timing. TABLE 1. Normal table equivalents (Nieuwkoop and Faber 1967) of the developmental stages of Rana rugulosa as listed in this paper. N & F stages Pan & Liang Stages 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 10% 10 10V2 11 11 12A UV2 12B 12 12C 12V2 12D 13 13 15 14 17 15 21-22 16 24-25 17 29/30 18 33/34 19A 37/38 19D 39 20 40 21 41 22 42 23 45 24 47 25 2. During early embryonic development, we raised Rana rugulosa in water of about 25.1 to 27.1°C. It takes only 70 hr, 20 min to develop from oocyte to gill cover stage. Under the same conditions, R. catesbeiana is 60 to 70 hr slower, and R. Umnochiris is 80 hr slower. Speed of development is closely related to temperature. At 26 to 28.5°C this developmental period in R. rugulosa takes only 64 hr, 30 min. Vol. 3, p. 88 Asiatic Herpetological Research April 1990 3. Fifty-five minutes after fertilization, the grey crescent of Rana rugulosa appears. This is identical to the timing of R. catesbeiana, but different from that of R. nigromaculata, in which the grey crescent can barely be seen, or is not visible. 4. The development of Rana rugulosa from the eight cell stage to the 16 cell stage takes six min, while the same period takes 59 min in R. catesbeiana. 5. In Rana rugulosa, external gill develops 3-5 branches during the period of hatching, while in R. catesbeiana, the external gill develops 2-3 branches during the tail bud circulatory stage. The third external gill appears sooner in R. rugulosa than in R. catesbeiana or R. guentheri. 6. Eggs laid by Rana rugulosa adhere in a pile or on the surface of the water. A reason for low hatching success is that oxygen levels can be low in the center of the egg mass. According to our observations, eggs at the edge of a mass are easy to hatch. If we separate the masses early enough, adequate dispersion can be achieved. However, the operation must be a careful one, or the egg membranes will be broken and the proper development of the embryo can be affected, causing deformity and death. References CAI, M. 1980. [Hasten parturition and artificial fertilize method of frogs]. Chinese Journal of Zoology, Beijing. 2:49-50. (In Chinese). CHENGDU INSTITUTE OF BIOLOGY 1977. [Key to Amphibians of China]. Science Press, Beijing, China. 93 pp. (In Chinese). GU, S. AND J. LE. 1986. [Animal Embryology]. People's Education Press, Beijing, China. (In Chinese). LIU, C. AND S. HU. 1961. [The Tailless Salientia in China]. Science Press, Beijing, China. 364 pp. (In Chinese). NIEUWKOOP, P. D. AND J. FABER. 1967. Normal table oiXenopus laevis. North Holland Publishing Co. Amsterdam. SHANDONG FISHERIES INSTITUTE. 1966. [The embryonic development of Rana catesbeiana.] Chinese Journal of Zoology 1966(4): 182-185. (In Chinese). ZHANG, J., J. LIU, AND M. CAI. 1985. [A study on the seasonal variation in the ovary and the reproductive frequency in Rana limnochiris]. Acta Herpetologica Sinica 1985, 4(4):276-281. (In Chinese). April 1990 Asiatic Herpetological Research Vol. 3, p. 89 Fig. 1 A I m m H Fig. 1 B Fig. ID I m m Fig. 1 E » 1 Fig. 2 A Fig. 2 B Fig. 3 A Fig. 3 B FIG. la, lb, lc, Id, le, 2a, 2b, 3a, 3b. Embryonic development of R ana rugulosa. See text for details. Vol. 3, p. 90 Asiatic Herpetological Research April 1990 Fig. 4 A Fig. 4 B Fig. 5 A l 1 Fig. 5 B Fig. 6 A Fig. 6 B Fig. 7 A Fig. 7 B Fig. 8 A FIG. 4a, 4b, 5a, 5b, 6a, 6b, 7a, 7b, 8a. Embryonic development of Rana rugulosa. See text for details. April 1990 Asiatic Herpetological Research Vol. 3, p. 91 Fig. 8 B Fig. 9 A Fig. 9 B I m m Fig. 10 A Fig. 10 B Fig. 1 1 A Fig. 1 1 B Fig. 12 A Fig. 12 B FIG. 8b, 9a, 9b, 10a, 10b, 11a, lib, 12a, 12b. Embryonic development of Rana rugulosa. See text for details. Vol. 3, p. 92 Asiatic Herpetological Research April 1990 Fig. 12 C Fig. 12 D I m m H 1 Fig. 13 A Fig. 13 B Fig. 14 A Fig. 14 B FIG. 12c, 12d, 13a, 13b, 14a, 14b. Embryonic development of Rana rugulosa. See text for details. April 1990 Asiatic Herpetological Research Vol. 3, p. 93 Fig. 15 A I m m I 1 Fig. 15 B Fig. 16 A Fig. 16 B Fig. 17 A Fig. 17 B I m m i < Fig. 17 C FIG. 15a, 15b, 16a, 16b, 17a, 17b, 17c. Embryonic development of Rana rugulosa. See text for details. Vol. 3, p. 94 Asiatic Herpetological Research April 1990 Fig. 18 B Fig. 18 C 2mm Fig. 19 A i 1 Fig. 19 B Fig. 19 C m m Fig. 19 D Fig. 20 A Fig. 20 B Fig. 20 C FIG. 18a, 18b, 18c, 19a, 19b, 19c, 19d, 20a, 20b, 20c. Embryonic development of Rana rugulosa. See text for details. April 1990 Asiatic Herpetological Research Vol. 3, p. 95 Fig. 21 B Fig. 21 A m m Fig. 21 C Fig. 22 A Fig. 22 B Fig. 22 C Fig. 23 B Fig. 23 C FIG. 21a, 21b, 21c, 22a, 22b, 22c, 23a, 23b, 23c. Embryonic development of Rana rugulosa. See text for details. Vol. 3, p. 96 Asiatic Herpetological Research April 1990 Fig. 24 A Fig. 24 B 2mm i « Fig. 24 C Fig. 25 A Fig. 25 B Fig. 25 C FIG. 24a, 24b, 24c, 25a, 25b, 25c. Embryonic development of R ana rugulosa. See text for details. April 1990 Asiatic Herpetological Research Vol. 3, p. 97 FIG. 26 (X30) FIG. 27 (X30) FIG. 28 (X30) FIG. 29 (X30) FIG. 30 (X30) FIG. 31 (X30) FIG. 26, 27, 28, 29, 30, 31. Embryonic development of Rana rugulosa. See text for details. Vol. 3, p. 98 Asiatic Herpetological Research April 1990 FIG. 32 (X30) FIG. 33 (X30) FIG. 34 (X30) FIG. 35 (X30) FIG. 36 (X30) FIG. 37 (X30) FIG. 32, 33, 34, 35, 36, 37. Embryonic development of Rana rugulosa. See text for details. April 1990 Asiatic Herpetological Research Vol. 3, p. 99 FIG. 38 (X30) FIG. 39 (X30) FIG. 40 (X30) FIG. 41 (X30) FIG. 42 (X30) FIG. 43 (X20) FIG. 38, 39, 40, 41, 42, 43. Embryonic development oiRana rugulosa. See text for details. Vol. 3, p. 100 Asiatic Herpetological Research April 1990 FIG. 44 (X20) FIG. 45 (X15) FIG. 46 (X15) FIG. 47 (X15) FIG. 48 (X10) FIG. 49 (X10) FIG. 44, 45, 46, 47, 48, 49. Embryonic development oiRana rugulosa. See text for details. April 1990 Asiatic Herpetological Research Vol. 3, pp. 101-103 The Validity of Elaphe perlacea, a Rare Endemic Snake from Sichuan Province, China ermi Zhao1 ^Chengdu Institute of Biology, P.O. Box 416, Academia Sinica, Chengdu, Sichuan, China Key Words: Reptilia, Serpentes, Colubridae, Elaphe perlacea, China, Sichuan, endemic, taxonomic validity. Introduction Recently Schulz (1989) reported on the validity of the specific status of Elaphe perlacea Stejneger, 1929. Schulz (1989) concluded, "Since data appear to be chiefly within the range of E. mandarina, E. perlacea is supposed to be placed in a subspecies rank of E. mandarina pointing out that further investigations may reveal that it is even a variety only." Elaphe perlacea is an endemic to Sichuan Province, China. For fifty years after the description of Elaphe perlacea by Stejneger (1929) no specimens were found. Many herpetologists including those of China doubted its validity. In the last ten years three additional specimens have been found in Sichuan Province (Table 1, 2). Two of these specimens (80-1, 87-2, and 88-3) were examined by me. All three specimens have a dorsal scale formula of 19-19-17, 7 upper labials (the third and fourth entering the eye), one preocular, two post oculars, 1+2 temporals, and a divided anal. Both the scale counts and the patterns are similar to those of the type. The new material differs from that of the type in having: 1) Four lower labials in contact with the anterior chin shield; 2) Dorsal scales of the male are smooth, only the posterior 2 to 3 middle rows are slightly keeled; 3) slight modification of the dorsal head pattern (Fig. 1). Elaphe perlacea differs from E. mandarina (Cantor) in many ways. E. mandarina has: 1) 23 scale rows on the neck and mid-body, 19 or 21 before the vent; 2) Two anterior temporal scales (occasionally one); 3) A much different dorsal pattern. Thus, Elaphe perlacea is a valid species. On a field trip during May of 1989 I went to Hailuo Gou one of the new localities. On the way we passed through a small town, Moxi. In the market were several Elaphe perlacea skins. This suggests that this snake is common in this area. The type locality of Elaphe perlacea is Yachow Prefecture (Ya'an Prefecture), Sichuan Province, China. This is an imprecise locality since a prefecture is quite large. The two new localities are not in this prefecture but are close to it. The new TABLE 1. Elaphe perlacea specimens found in Sichuan recenUy. Sex, Number Locality Altitude (m) Date Snout-vent (mm) Tail (mr 2500 22 Oct 1987 - - 2500 22 Oct 1988 884 146 2000 5 Jun 1988 1055 189 2000 Apr 1980 865 194 Male Male, 87-2 Female, 88-3 Female, 80-1 Hailuo Gou, Luding Hailuo Gou, Luding Hailuo Gou, Luding Wolong, Wenchuan © 1990 by Asiatic Herpetological Research Vol. 3, p. 102 Asiatic Herpetological Research April 1990 TABLE 2. Scale counts of Elaphe perlacea Stejneger, 1929. Sex Locality Dorsal Ventral Anal Subcaudal Upper Lower Pre& Temp labial labial post ocular oral M Yachow, the type specimen 19 19- 17 229 2 69/69+1 2-2-3 (3) 1-2 1+2 M Hailuo Gou, Luding 19 19- 17 231 2 62/62+1 2-2-3 9(4) 1-2 1+2 F Wolong, Wenchuan 19 19- 17 224 2 67/67+1 2-2-3 8/9(4) 1-2 1+2 a b c FIG. 1. Back of head of Elaphe perlacea, showing the main pattern, a. Male, type specimen, b. Female from Wolong. c. Male from Hailuo Gou. FIG. 2. Map of Sichuan showing approximate position of localities. April 1990 Chinese Herpetological Research Vol. 3, p. 103 localities are: Elev. 2000 and 2500 m, Literature Cited Hailuo Gou, Luding County, Garze Zang Autonomous Prefecture, and Elev. 2000 m, SCHULZ, K.-D. 1989. Die hinterasiatischen Wolong, Wenchuan County, Aba (Ngawa) Kletternattern der Gattung Elaphe Teil XVI Zang Autonomous Prefecture, Sichuan Elaphe perlacea Stejneger, 1929 Sauria, Berlin- Province, China (Fig. 2). As presently W., 11(2):15-16. understood Elaphe perlacea is endemic to Sichuan Province, China, and only occurs Stejneger, L. 1929. A new snake from China, in the foot hills of the Himalayan Plateau Proceedings of the Biological Society of directly west of the Sichuan Basin. Wash.ngton 42:129-130. [April 1990 Asiatic Herpetological Research Vol.3, pp. 104-115 Stellio sacra (Smith 1935) a Distinct Species of Asiatic Rock Agamid from Tibet* NATALIA B. ANANJEVA1, GUNTHER PETERS2, J. ROBERT MACEY3 AND THEODORE J. PAPENFUSS3 ^Zoological Institute, USSR Academy of Sciences, Leningrad, USSR ^Museum fur Naturkunde der Humboldt-Universitdt, Invalidenstrasse 43, Berlin, DDR ^Museum of Vertebrate Zoology, University of California, Berkeley, California, USA Abstract. -An examination of the type series of Agama himalayana sacra Smith (1935) and new material collected in Tibet in 1988 has shown that this form should be considered a distinct species. In following the recent revision of the genus Agama (Moody 1980), Stellio sacra is included in the genus with all other Asiatic rock agamids. Key Words: Reptilia, Sauria, Agamidae, Stellio sacra, Tibet, systematics, distribution. Introduction Four agamid lizards were collected at the beginning of the 20th Century near Lhasa, Tibet. They were described in a short account by Smith (1935) as a subspecies of the Himalayan rock agamid Agama himalayana sacra. The examination of these type specimens and new material collected in Lhasa and the vicinity of Lhasa, Tibet in 1988 by the third and fourth authors, led us to the conclusion that it is not a subspecies of Stellio himalayanus. Furthermore it is not a member of the Stellio himalayanus species complex which includes Stellio badakhshanus, Stellio chernovi, Stellio himalayanus, and Stellio stoliczkanus. The results of the examination of the agamids from Lhasa, Tibet confirm the opinion of Ananjeva et al. (1981) that Stellio sacra should be elevated to full specific status. According to the present revision of the genus Agama (Wermuth 1967) into Agama, Stellio (Moody 1980; Sokolovsky 1975, 1977) Trapelus, Psuedotrapelus, and Xenagama (Moody 1980) the lizards examined from Tibet should have the generic name, Stellio. * Sino-Soviet- American Arid Asian Desert Regions Research Paper no. 5. Methods To the best of our knowledge all specimens housed outside the Peoples Republic of China were examined. Of the four specimens in the type series, one lectotype and two paralectotypes are housed at the British Museum (Natural History) [BMNH], and one paralectotype is housed at the Indian Museum, Calcutta (Zoological Survey of India [ZSI]). Thirteen additional specimens were collected during the 1988 joint Chengdu Institute of Biology - University of California Expedition and are housed at the California Academy of Sciences (CAS). Stellio sacra (Smith 1935), new combination Known Material Material Examined: Lectotype: BMNH 1946.8.28.57 (formerly 1904.12.28.1) [Fig. 1.] Locality: Near Lhasa, Tibet. Paralectotypes: BMNH 1946.8. 28.58, BMNH 1946.8.28.59 [Fig. 2], and ZSI 15740 [Fig. 3]. Locality: near Lhasa, Tibet. CAS 170545. Locality: Elev. 3740 m, Sera Monastery, Lhasa (29° 39' N 91° fr\ 1 fifin u., April 1990 Asiatic Herpetological Research Vol. 3, p. 105 FIG. 1. Lectotype of Stellio sacra BMNH 1946.8.28.57 (formerly 1904.12.28.1). Sera Monastery, Lhasa (29° 39' N 91° 06' E), Lhasa Municipality, Xizang (Tibet) Autonomous Region, China. Collected by: T. J. Papenfuss and R. Macey. Date: 24 Sept., 1988. CAS 170546-53. Locality: Elev. 3700 m, at base of mountains approx. 3 km WNW (airline) of the Potala Palace, Lhasa (29° 39' N 91° 06' E), Lhasa Municipality, Xizang (Tibet) Autonomous Region, China. Collected by: CAS 170546-49 R. Macey and T. J. Papenfuss, and CAS 170550-53 T. J. Papenfuss and R. Macey. Date: 25 Sept., 1988. CAS 170554-57. Locality: Elev. 3990 m, 52.4 km south of Yangbajan (30° 13' N 90° 25' E), also at km 1900.8 from Xining, on the Xining-Golmud-Lhasa Rd., Lhasa Municipality, Xizang (Tibet) Autonomous Region, China. Collected by: CAS 170554-55 R. Macey and T. J. Papenfuss, CAS 170556-57 T. J. Papenfuss and R. Macey. Date: 27 Sept., 1988. (See Table 1). Other Material. Note that the following material is reported on in Hu et al. (1987) but no numbers or reference to a museum collection are mentioned. It is probable that all or most of these specimens are housed at the Chengdu Institute of Biology. Three males, 3 females, and 3 juveniles. Locality: Bomi (29° 50' N 95° 45' E), Qamdo Prefecture, Xizang (Tibet) Autonomous Region, China. Six males, and 12 females. Locality: Lhasa (29° 39' N 91° 06' E), Lhasa Municipality, Xizang (Tibet) Autonomous Region, China. Vol. 3, p. 106 Asiatic Herpetological Research April 1990 FIG. 2. Paralectotypes oiStellio sacra BMNH 1946.8. 28.58 and BMNH 1946.8.28.59. Two males, 2 females, and 1 juvenile. Locality: Nyingchi (29° 32' N 94° 25' E), Lhasa Municipality, Xizang (Tibet) Autonomous Region, China. Distribution Stellio sacra as presently understood, is restricted to the river drainage of the Yarlung Zangbo in the Lhasa Valley, Xizang (Tibet) Autonomous Region, China (Fig. 4). Only four localities are known, all between 3000 and 4000 m. Populations occurring in the Kunlun Mountains of southern Xinjiang Uygur Autonomous Region, China were previously assigned to Stellio himalayanus himalayanus, however their present taxonomic position is uncertain. Diagnosis: Rock agamid with flattened head and body which is typical for this lizard group. They are comparatively large lizards with a snout-vent length of 120- 150 mm and a tail length of 180-240 mm (Table 1). Gular Sac seems to be developed to a greater degree than in other Stellio. Body scales are small and granular. The scales are not well differentiated. There is a very slight but noticeable nuchal crest on the head. It begins from the middle of the occiput and continues as a poorly differentiated vertebral stripe. The longitudinal rows of enlarged and feebly keeled scales on the vertebral region are arranged parallel to each other. There are neither groups of enlarged scales nor separate enlarged scales on the dorsal lateral regions. The males have a large patch of callous scales on the belly. The annuli and segmentation of the scales on the basal quarter of the tail are not prominent. On the April 1990 Asiatic Herpetological Research Vol. 3, p. 107 \ / FIG. 3. Paralectotype of Stellio sacra ZSI 15740. lateral surface of the tail there are three to four annuli in each segment. There is a small granular dark pattern on the back. The center of the back tends to have more black and toward the sides a dark golden brown dominates. The separate elements of this pattern are connected to heavily marked diffuse transverse stripes. The narrow stripes form two rows of the dark colored scales that continue from the neck to the tail. Overall the lizard is darkly colored but there are a few randomly scattered yellow blotches on the back (Fig. 5). Juveniles are lighter in color tending more toward a dark golden brown with darker speckling all over the back. The dark golden brown forms bands across the back which are offset at the spine. Comparative Description Stellio sacra differs from Stellio himalayanus and related forms in all the diagnosis characters. Stellio sacra is only similar in body size to Stellio stoliczkanus among the species examined. Other species are notably smaller (Table 1; Anderson and Leviton 1969; Ananjeva et al. 1981). The body proportion data concerning Stellio sacra and the Stellio himalayanus species complex indicates small differences Vol. 3, p. 108 Asiatic Herpetological Research April 1990 TABLE 1 . Characters of Stellio sacra specimens used for analysis. Rostrum to Hind limb front Number of Sex/age Body length Tail length length margin of tympanum scales at midbody Type specimens BMNH 1946.8.28.57 M 136 215 98.5 30.5 232 BMNH 1946.8.28.58 F 130 autotomized 92.0 27.6 230 BMNH 1946.8.28.59 subadult 74 146 48.5 17.0 225 ZSI 15740 F 124 182 83.5 26.6 - CAS specimens 170555 M 147 246 92.2 33.6 249 170554 M 145 233 98.0 31.2 240 170556 M 143 autotomized 89.0 30.3 250 170546 F 114 188 74.0 24.3 230 170547 F 100 185 67.2 21.9 245 170548 F 100 autotomized 66.8 21.0 247 170545 juv. 67.0 128 45.0 14.8 225 170549 juv. 72.5 142 52.0 16.1 256 170550 juv. 71.0 137 48.0 15.7 246 170551 juv. 75.5 155 54.0 16.5 250 170552 juv. 75.0 136 52.0 16.2 234 170553 juv. 66.3 132 48.0 15.4 248 170557 juv. 82.5 146 61.0 18.0 223 (Table 2). Stellio sacra has a shorter tail than Stellio badakhshanus, Stellio chernovi, and Stellio himalayanus. One of the paralectotypes, a female (BMNH 1946.8.28.58), has a regenerated tail. The length of the regenerated tail after autotomy is about 80 mm which is rather great. The scales of the tail are mucronate, and the regular annular arrangement in the regenerated portion of the tail is disturbed. Stellio sacra has a more similar head height index to Stellio erythrogaster and Stellio lehmanni than to Stellio caucasius and Stellio himalayanus (Ananjeva 1981); i.e. the head is not so flat (depressed). It is necessary to also remark about the differences in the structure of the digits in Stellio sacra and that of the specimens examined of Stellio caucasius, Stellio lehmanni, and Stellio stoliczkanus of the same size. The digits, and especially the terminal phalanges, of Stellio sacra specimens are not round in their section as in the above group of species, but instead are compressed laterally. The distinctive character of Stellio sacra pholidosis is its comparative homogeneity. The scales on the back, sides and belly are small. It has a considerably large scale count around the mid-body of 239 scales. The mid-body scale count of Stellio badakhshanus, Stellio chernovi, Stellio himalayanus, and Stellio stoliczkanus all do not exceed 180 scales. Such a large mid-body scale count is only observed in Stellio nuristanicus, which has from 230- 248 mid-body scales. During this phase of our study two of the type specimens (Anderson and Leviton 1969) were examined. The longitudinal rows of the enlarged poorly keeled scales are not arranged as the obvious "dorsal stripe," as typically observed in many other species of Stellio. The dorsal scales gradually decrease in size and degree of keeling, from the center of the back toward the small dorso-ventral scales. The ventral scales are smaller than the dorsal scales. The posterior margins of these scales slightly overlap the scales of the following row. April 1990 Asiatic Herpetological Research Vol. 3, p. 109 FIG. 4. Distribution of Stellio himalayanus and Stellio sacra. Question mark indicates Xinjiang Autonomous Region population of questionable taxonomic status. The ventral head scales are also small and not regular. The size of these scales decreases from the nasal end toward the gular fold. On the sides of the head and on the neck are situated small groups of very strongly keeled spinose scales. Stellio sacra has nostrils shaped longitudinally oval, similar to Stellio chernovi and Stellio himalayanus. Where as Stellio agrorensis, Stellio melanurus, Stellio nuristanicus and Stellio tuberculatus have round nostrils. Stellio sacra has a round tympanum. In Stellio sacra, the scales of the upper surface of the fore and hind limbs are comparatively small, but their size is similar to the largest dorsal scales. These scales are strongly keeled. Single slightly enlarged scales are distinguishable from the surrounding small scales. The scales on the base of the tail are keeled and mucronate, but to a smaller degree than for example in Stellio himalayanus. The results of the comparative study of Asiatic rock agamids enables the typical distinguishing characters of Stellio sacra to be identified: 1 . The small size of the scales. 2. The presence of a short nuchal crest made up of from two to three scale rows of slightly enlarged, narrow, keeled, dark colored scales. Such a nuchal crest is absent in Stellio chernovi, Stellio himalayanus, Stellio stoliczkanus and other species. A nuchal crest is noted however in Stellio melanurus, Stellio nuptus, Stellio Stellio, and Stellio tuberculatus. Vol. 3, p. 110 Asiatic Herpetological Research April 1990 ,- « FIG. 5. Stellio sacra from elevation 3990 m, 52.4 km south of Yangbajan (30° 13' N 90° 25' E), Xizang (Tibet) Autonomous Region, China. 3. The ends of the occipital (nuchal) and neck scales are directed backwards but not foreword; the parietal scales are oriented in a laterocaudal direction. 4. The rows of the enlarged dorsal scales are parallel but do not meet as in species of the "Stellio himalayanus " complex. TABLE 2. Comparative data on the relative proportions of the tail, limbs, and head of rock agamids of the genus Stellio (5. sacra and agamids of the 5. himalayanus complex). Species tail length/ hind limb length/ head length/ n body length body length (%) body length (%) Male Female Male Female Male Female Male Female Stellio sacra 5 4 1.61 1.58 72.2 69.0 27.0 25.1 Stellio stoliczkanus 35 14 1.65 1.48 66.3 63.7 25.9 25.1 Stellio himalayanus 11 15 1.88 1.75 67.9 66.7 26.3 25.7 Stellio badakhshanus 3 1 1.72 1.67 76.0 70.0 27.3 26.2 Stellio chernovi 18 8 1.85 1.77 77.2 74.0 27.6 26.3 April 1990 Asiatic Herpetological Research Vol.3, p. Ill FIG. 6. Habitat of Stellio sacra, elevation 3990 m, 52.4 km south of Yangbajan (30° 13' N 90° 25' E), Xizang (Tibet) Autonomous Region, China. 5. In the region over the shoulders there are two arched diffuse rows of slightly enlarged mucronate (spinose) scales. 6. The absence of enlarged scales or regions of enlarged scales on the back of the body. 7. The upper head scales distributed between the nasal scales have a stretched shape. The length of these scales is twice the width as also in Stellio agrorensis, Stellio nuristanicus, and Stellio tuberculatus where as the scales of Stellio chernovi and Stellio himalayanus have a roundish, polygonal shape. 8. Only a single row of small scales is noted between the suboculars and the supralabials. Other species have two to five scale rows. 9. The scales of the back, shoulders, thighs and some parietal scales have an unusual microstructure. The margins of the scales are jagged resembling that of fringes. This character was also observed in Stellio annectens, Stellio erythrogaster, Stellio melanurus, and Stellio nuptus. 10. There is a patch of callous (granular) scales in the middle of the belly, which is large in males. 1 1 . The males have a large patch of anal pores and from six to seven rows of callous scales before the cloaca. Coloration The dorsal coloration of the type Vol. 3, p. 112 Asiatic Herpetological Research April 1990 specimens consists of diffuse components, each scale is either totally dark or light in coloration. The parietal region and along the middle of the back (two scale rows of absolutely dark scales) are more dark in coloration. From the middle of the back to the sides there are light brown transverse stripes. This coloration is poorly developed in some specimens. The subadult specimen has a similar small speckled pattern and in addition has alternating light and dark transverse stripes coming from the vertebral ridge toward the sides of the body. This pattern is similar to that of juvenile specimens of Stellio tuberculatus. As far as we can tell from the available specimens of Stellio sacra, no contrasting pattern and coloration on the surface of the gular region, neck and nuchal region are noted. Such a contrasting pattern and coloration are typical for specimens of Stellio himalayanus and related species. Natural History Stellio sacra are common in the rocky hills surrounding the Lhasa Valley. They were seen only on slopes covered with large boulders (Fig. 6). Often a single adult male occupied a pile of boulders in association with several females and juveniles. Stellio sacra is an agile, alert species that is difficult to approach closely. The only other reptile that occurs on the rocky slopes is the gecko, Cyrtodactylus tibetanus. Another agamid, Phrynocephalus theobaldi is abundant on the sandy soil at the base of the rocky hills. Discussion The large review of the rich Indian fauna that Smith (1935) dealt with explains why the Sacred Agamid, Stellio sacra, was originally described as a subspecies of the Himalayan agamid, Stellio himalayanus. The study presented here concludes that Stellio sacra is not related to the Stellio himalayanus complex. On the other hand the above descriptions of the morphological characters do not allow an interpretation of the direct relation of Stellio sacra to such species as Stellio agrorensis, Stellio melanurus, Stellio nuptus, and Stellio tuberculatus. The characters which enable us to bring together Stellio sacra and these species should be considered as plesiomorphic similarities. These characters are long limbs, the presence of a small gular sac and nuchal crest, the polyannular structure of the caudal segments, the juvenile color pattern, and so on. This assumption about a plesiomorphic condition is based on the idea that the oligomerization (decreasing of number of elements) of homologic organs (here pholidosis elements) is one of the main directions of agamid evolution (Dogel 1954). Characters, such as the orientation of the scale axes in nuchal and neck regions and in parietal scales is also of great interest. It is suggested (Peters unpublished data) that the more common condition of the scale, i.e. the orientation of the point backwards, is typical not only for most of the species in the genus Stellio (including Stellio sacra ), but also for the majority of agamids and lizards as a whole. Hence it is a plesiomorphic condition. The contrary direction, where the point of the scale is forwards, is observed in Stellio melanurus, Stellio nuptus and a number of African species. In accordance with Peter's point of view, this should be considered as an apomorphic condition. However there is a problem in interpreting such a character as the presence of the scales with the jagged margins. This scale structure is found in species with both types of scale orientation, including Stellio sacra as well as Stellio erythrogaster, Stellio melanurus, Stellio nuptus and the African species, Stellio annectens. The interpretation of such characters makes all possible explanations of the relationships of these species controversial. These facts and the inability at this time to determine even a single obvious synapomorphy for all Asiatic rock agamids of the genus Stellio (with or without Stellio melanurus and Stellio nuptus ) does not allow an acceptable phylogenetic relationship to be developed. This problem April 1990 Asiatic Herpetological Research Vol. 3, p. 113 TABLE 3. The distribution of the species in the genus Stellio. Species Distribution Stellio adramitana (Anderson 18%) Stellio agrorensis (Stoliczka 1872) Stellio annectens (Blanford 1870) Stellio atricollis (Smith 1849) Stellio badakhshanus (Anderson and Leviton 1969) Stellio caucasius (Eichwald 1831) Stellio chernovi (Ananjeva, Peters, and Rzepakovsky 1981) Stellio cyanogaster (Ruppell 1835) Stellio erythrogaster (Nikolsky 1896) Stellio himalayanus (Steindachner 1869) Stellio lehmanni (Nikolsky 1896) Stellio melanurus (Blyth 1854) Stellio microlepis (Blanford 1874) Stellio nuptus (De Filippi 1843) Stellio nuristanicus (Anderson and Leviton 1969) Stellio phillipsii (Boulenger 1895) Stellio sacra (Smith 1935) Stellio stellio (Linnaeus 1758) Stellio stoliczkanus (Blanford 1875) Stellio trachypleurus (Peters 1982) Stellio tuberculatum (Hardwicke and Gray 1827) Stellio yemenensis (Klauzewitz 1954) Stellio zonorus (Boulenger 1895) Arabia Afghanistan, Pakistan, India Africa (Somalia, Ethiopia) Eastern and southern Africa Afghanistan Caucasus, Tadjikistan, Turkmania, Turkey, Iran, Afghanistan, Pakistan Tadjikistan, Turkmania, Uzbekistan Iraq, Somalia, Ethiopia Iran, Turkmania Tadjikistan, Uzbekistan, Afghanistan, Pakistan, India Tadjikistan, Turkmania, Uzbekistan, Afghanistan Iran, Pakistan Iran Iraq, Iran, Afghanistan, Pakistan Afghanistan Ethiopia Tibet Greece, southwest Asia, northern Egypt Mongolia, China Ethiopia India, Nepal, Afghanistan, Pakistan Arabia Somalia may be explained not only by the small sample size of Stellio sacra and the fact that many other forms are poorly studied but also by the obvious existence of parallel trends in different developmental lines within this lizard group. In order to arrive at a more trust worthy hypothesis of the relationships of rock agamids, it is necessary to carry out biochemical investigations. The first preliminary results of such investigations are that of Joger and Arano (1987) and Ananjeva and Sokolova (in prep.). This type of data will be highly interesting to compare with the results of comparative morphological studies. The distributional patterns of Asiatic rock agamids of the genus Stellio (Table 3) and their chorological isolation seems to support the idea that Stellio is a monophyletic group. Rock agamids of the genus Stellio are distributed from Greece and the Nile River Delta in the west to the Gobi Altai of southern Mongolia, and the western deserts of China in the east. In the western part of this region, Greece to the Nile River Delta, Stellio stellio is found. The southern portion of this region from southwestern Iran to Pakistan Stellio melanurus and Stellio nuptus occur. These species are probably of African origin (Peters unpublished). In southern Iran and Afghanistan along with Stellio melanurus and Stellio nuptus also occur Stellio agrorensis, Stellio badakhshanus, Stellio caucasius, Stellio erythrogaster, Stellio himalayanus, Stellio microlepis, and Stellio nuristanicus. The Asiatic rock agamids of the genus Stellio sensu stricto are absent from the mountainous regions of western Indostan Vol. 3, p. 114 Asiatic Herpetological Research April 1990 and the eastern portion of the Arabian Peninsula. They are also not found across the Red Sea in Ethiopia and Somalia. These areas are poorly studied herpetologically and it is possible that they do in fact occur in these regions. The distribution of Asiatic rock agamids seems to be a single unit. Most problematic is the origin of such species as Stellio melanurus, Stellio nuptus, and Stellio stellio, which may be of African origin as it was already mentioned above. It is possible that this assumption will be corroborated during further research in biochemical phylogeny. The preliminary results of Joger and Arano (1987) on Stellio stellio is a first step. The origin of the genus Stellio is interesting and the data about the early divergence of Stellio in Stellio sensu stricto (Asiatic species) and the Stellio atricollis species group from Africa and southern Arabia is useful (Joger and Arano 1987). It is possible that Stellio is a paraphyletic group of species. The Arabian - African species group is Stellio adramitana, Stellio annectens, Stellio atricollis, Stellio cyanogaster, Stellio phillipsi, Stellio trachypleurus, Stellio yemenensis, and Stellio zonurus (Table 3). These species are similar to the Asiatic species Stellio melanurus and Stellio nuptus in a number of characters. This allows an assumption that they are related. Acknowledgments We are most greatful for the loan of type material of Stellio sacra from Dr. A. C. G. Grandison, Dr. E. H. Arnold and Dr. A. F. Stimson of the British Museum of Natural History, London, Great Britain, and Dr. Shara of the Indian Museum, Calcutta, India. Literature Cited ANANJEVA, N. B. 1981. [The peculiarities of the skull structures, dental system, and hyoid apparatus of the lizards of Agama genus of the USSR fauna]. Proceedings of the Zoological Institute of the USSR Academy of Sciences, Leningrad 101:3-20. (In Russian). ANANJEVA, N. B., G. PETERS, AND V. T. RZEPAKOWSKY. 1981. [New species of rock agamid from Tadjikistan Agama chernovi sp. nov.]. Proceedings of the Zoological Institute of the USSR Academy of Sciences, Leningrad 101:23-27. (In Russian). ANDERSON, S. T. AND A. E. LEVITON. 1969. Amphibians and reptiles collected by the Street Expedition to Afghanistan, 1965. Proceedings of the California Academy of Sciences, series 4. 37(2):25-56. DOGEL, V. A. 1954. [Oligomerization of the homological organs as one of the main tendencies of animal evolution]. Leningrad State University. 368 pp. (In Russian). HU, S., E. ZHAO, Y. JIANG, L. FEI, C. YE, Q. HU, Q. HUANG, Y. HUANG, AND W. TIAN. (incorrectly stated in English as Hu, S, L. Fei, Q. Hu, Q. Huang, Y. Huang, Y. Jiang, W. Tian, C. Ye, E. Zhao). 1987. Amphibia- Reptilia of Xizang. The series of the scientific expedition to the Qinghai-Xizang Plateau. The comprehensive scientific expedition to Qinghai- Xizang Plateau, Academia Sinica. Chengdu Institute of Biology, Academi (sic.) Sinica. 153 pp. (If Chinese). JOGER, U. AND B. ARANO. 1987. Biochemical phylogeny of the Agama genus group. Pp. 215- 218. In Van Gelder, Strijbosch.and Bergers (eds.). Proceedings of the 4th Ordinary General Meeting of the Society of European Herpetologists, Nijmegen. MOODY, S. M. 1980. Phylogenetic and historical biogeographical relationships of the genera in the Agamidae (Reptilia: Lacertilia). Ph.D. Thesis. University of Michigan. 373 pp. PETERS, C. 1982. Eine neue Wirtelschwanzagame aus Ostafrika (Agamidae: Agama ). Mitteilungen aus der Zoologishen Sammlung des Museums fur Naturkunde in Berlin 58(2):265- 268. SMITH, M. A. 1935. The fauna of British India, including Ceylon and Burma. Reptilia and Amphibia. Vol. II. Sauria. Taylor and Francis, London. 440 pp. SOKOLOVSKY, V. V. 1975. [Comparative karyological study of the lizards from the family Agamidae. II. Karyotypes of five species of the genus Agama ]. Cytologica 16(l):91-93. (In Russian). April 1990 Asiatic Herpetological Research Vol. 3, p. 115 SOKOLOVSKY, V. V. 1977. [Taxonomic relationships in the family Agamidae according to karyological data]. Soviet Herpetological Conference, Leningrad. [English abstract, in Russian]. WERMUTH, H. 1967. Liste der rezenten Amphibien und Reptilien: Agamidae. Das Tierreich 86, Berlin. 127 pp. I April 1990" Asiatic Herpetological Research Vol. 3, pp. 116-119| Isolation and Amino Acid Sequence of a New Dodecapeptide from the Skin of Oreolalax pingii + YIQUAN TANG1, SHENGHAI TlAN1, SHIXIANG WU1, JlACHENG HUA1, XlNQUAN Jl1, GUANFU WU2, ERMI ZHAO2, AND GANG ZOU1 ^Shanghai Institute of Materia Medica, Academia Sinica, Shanghai, China 2Chengdu Institute of Biology, P.O. Box 416, Academia Sinica, Chengdu, China Abstract. -A novel dodecapeptide has been isolated by alumina column chromatography and HPLC from methanol extracts of the skin of the Chinese frog Oreolalax pingii. The sequence of the peptide is: Gly- Leu-Val-Ser-Asp-Leu-Met-Tyr-Gly-Ile-Gly-Leu-NH2 Th's peptide differs from all the other amphibian skin peptides and should be regarded as a member of a new peptide family. Key Words: Amphibia, Anura, Pelobatidae, Oreolalax pingii, China, biochemistry, peptides. Introduction Many active peptides have been discovered from amphibian skin during the past 20 years or more. Amphibian skin peptides have proved to be of considerable value not only in pharmacology, but also in taxonomic and evolutionary domains (Erspamer 1984; Cei 1985; Lazarus et al. 1985). In order to discover new active peptides, we have carried out research on amphibian skin peptides from Chinese frogs since 1983 (Hua et al. 1985; Tang et al. 1985). This paper concerns another novel dodecapeptide obtained from the skin of Oreolalax pingii. Methods The materials and experimental procedures were previously reported (Tang et al. 1985), with the following brief mentions and additions. Six hundred specimens of Oreolalax pingii were collected in May, 1983 from the Daliangshan of Sichuan Province, China. The fresh skins (400 g) were removed and extracted with methanol. The methanol extracts were evaporated until dry. The residue was dissolved in 95% ethanol and the solution distributed on the alumina * This publication combines material previously published in Chinese by Tang et al (1985) with additional information. columns. The column was eluted with ethanol-water mixtures of descending concentrations of ethanol (also see Montecucchi et al. 1981) HPLC was performed on a Waters HPLC system. Details of individual chromatographic procedure are shown in the figures. Amino acid analysis of peptides after hydrolysis in HC1 were carried out on a LKB 4400 amino acid analyzer. Sequence analysis of the peptides were performed by manual DABITC/PITC procedure (Chang 1983). The complete sequence analysis of the dodecapeptide was carried out on an Applied Biosystems Model 470A gas phase sequencer. Enzymatic digestions of the peptide with a-Chymotrypsin, carboxypeptidase A (CP- A) and Y (CP-Y) were also as before (Tang et al. 1985). Bioassays of each isolated product were tested on the longitudinal muscle myenteric plexus preparation of the guinea pig ileum (GPI), Results The water portion eluted from alumina columns was lyophilized to give 214 mg of residue. The residue was separated by HPLC semi-preparatively as in Fig. 1. Perks 32 and 33 each were single peak by verifying on HPLC (analytical uBondapak Cj8 column, the elution conditions were the same as in Fig. 1), respectively. The ir»r»r\ i_. April 1990 Asiatic Herpetological Research Vol. 3, p. 117 100 o CM C\J < 40 60 time(min) FIG. 1. RP-HPLC of the water eluate from the alumina columns. Column: uBondapak C\% 7.8X300 mm. Mobile phase: A=0.05% CF3COOH B=60% CH3CN in A. Concave gradient elution from 0-85%B (dotted line), 90 min, at 1.0 ml/min. Detected at UV 220 nm, 0.4 aufs. Gly-Leu-Val-Ser-Asp-Leu-Met-Tyr-Gly-Ile-Gly-Leu-NH2 ■GH-1- A h ■CH-2- FlG. 2. Profile of sequence anlaysis of the dodecapeptide. (-*): DABITC/PITC method. (-»): gas phase sequencing. («-): CP-Y digestion. CH: a-chymotryptic peptides. difference of retention time of the two peaks was 0.8 min. The peak 32 acted as the representative of the dodecapeptide for further studies and the peak 33 was also investigated simultaneously. Amino acid composition of the peak 32 was Asp (1), Ser (l),Gly (3), Val (1), Met (1), He (1), Leu (3), Tyr (1). Amino acid sequence of the peak 32 was determined by the DABITC/PITC method and gas phase sequencing. The former method proceeded to the tenth step, but the latter to the penultimate residue (Fig. 2). For C- terminal residue analysis, CP-A and CP-Y digestions of the peak 32 were carried out, and did not release any amino acids by the former. This indicated a blocked C- terminus; upon latter, however, Leu and Gly were obtained. Thus, we deduced that the C-terminal structure of the peak 32 is Leu-NH2. Digestion of the peptide by a- chymotrypsin provided further structural confirmation. The fragment peptides, CH- 1 and CH-2 were separated on HPLC as depicted in Fig. 3. Amino acid compositions of the two fragments are in accordance with their sequences (see Fig. 2), respectively. From the above results the complete amino acid sequence of the peak 32 was Vol. 3, p. 118 Asiatic Herpetological Research April 1990 0.2-r 100 10 20 30 time(min) 40 50 60 FIG. 3. RP-HPLC of a-chymotryptic digests of the dodecapeptide. Column: uBondapak Cis 3.9X300 mm. Mobile phase: A=0.1% CF3COOH B=60% CH3CN in A. Linear gradient elution from 0-60%B , 40 min, at 0.6 ml/min. Detected at UV 220 nm, 0.2 aufs. Peak (*) is the unreacted dodecapeptide. established as in Fig. 2. The peak 33 had the same properties with the peak 32 in the amino acid compositions and sequence analysis, respectively. We do not know the structural differences between the two peaks. Based on the amino acid analysis, the yield of the pure dodecapeptide (including peak 32 and 33 in Fig. 1) was estimated to be at least 1.0 nmol starting from 1.0 g of fresh skin, according to that the rate of recovery of all above isolation steps was 10%. The dodecapeptide was inactive in GPI test, and its activity awaits to be established by assay methods other than those used in the present screening. Discussion The dodecapeptide described above is a completely novel peptide. The peptide may be a member of a new peptide family, and it should be regarded as the biochemical characteristic of Oreolalax pingii in taxonomy. Furthermore, the C-terminal portion of the dodecapeptide has some homologies to the mammalian Leu- enkephalin (Hughes et al. 1975) as shown below: The dodecapeptide: Gly-Leu-Val-Ser-Asp- Leu-Met-Try-Gly-Ile-Gly-Leu-NH2 Leu-enkephalin: Try-Gly-Gly-Phe-Leu In amphibian skin peptide research, it April 1990 Asiatic Herpetological Research Vol. 3, p. 119 was observed that one peptide can be separated by HPLC into two components— the peak splitting, such as that discovered in HPLC separation of PGL-a (Andreu et al. 1985) and ranamargatin (Tang et al. 1988). The peak 32 and 33 (Fig. 1) both have the same amino acid composition and sequence. Therefore, they may be produced by one peptide in HPLC separation. The reason of the peak splitting mentioned above, however, is presently not known. To our knowledge, there are no similar findings besides the amphibian skin peptides. Hence, it is necessary to understand if the peak splitting is unique to the amphibian skin peptides. Acknowledgments We thank Mr. C. Chen for amino acid analysis. Literature Cited ANDREU, D., H. ASCHAUER, G. KREIL, AND R. B. MERRIFIELD 1985. Solid-phase synthesis of PYLa and isolation of its natural counterpart, PGLa[PYLa-(4-24) peptide amide] from skin secretion of Xenopus laevis. European Journal of Biochemistry 149:531-535. CEI, J.M. 1985. Taxonomic and evolutionary significance of peptides in amphibian skin. Peptides 6(Suppl. 3): 13-16. CHANG, J. 1983. Manual micro-sequence analysis of polypeptides using dimethylaminoazobenzene isothiocyanate. Methods in Enzymology 91:455-466. ERSPAMER, V. 1984. Half a century of comparative researchs on biogenic amines and active peptides in amphibian skin and molluscan tissues. Comparative Biochemistry and Physiology 79C: 1-7. HUA, J., S. WU, Y. TANG, W. ZHANG, AND G. ZOU 1985 [Isolation and characterization of bradykinin and its two fragments from the skin of the Chinese frog Rana nigromaculata ]. Acta Biochemica Biophysica Sinica 17:171-173. (In Chinese). HUGHES, J., T. W. SMITH, H. W. KOSTERLITZ, L. A. FOTHERGILL, B. A. MORGAN, AND H. R. MORRIS. 1975. Identification of two related pentapeptides from the brain with potent opiate agonist activity. Nature (London) 258:577-579. LAZARUS, L.H., W. E. WILSON, G. GAUDINO, B. J. IRONS, AND A. GUGLIETTA 1985. Evolutionary relationships between non- mammalian and mammalian peptides. Peptides 6(Suppl. 3):295-307. MONTECUCCHI, P.C., R. DE CASTIGLIONE, S. PIANI, L. GOZZIN, AND V. ERSPAMER 1981. Amino acid composition and sequence of dermorphin, a novel opiate-like peptide from the skin of Phyllomedusa sauvagei. International Journal of Peptide Protein Research 17:275-284. TANG, Y., S. TIAN, S. WU, J. HUA, G. HU, X. JI, G. ZOU, G. WU, AND E. ZHAO. 1985. [Separation and structure of a novel hexapeptide obtained from the skin of Oreolalax pingii]. Acta Herpetologica Sinica 1985, 4(2):99-102. (In Chinese). TANG, Y., S. TIAN, S. WU, J. HUA, G. WU, E. ZHAO, Y. LU, Y. ZHU, AND G. ZOU. 1988. [Isolation and structure of ranamargarin, a new tachykinin from the skin of Chinese frog Rana margaratae ]. Scienua Sinica (B) 9:967-974. (In Chinese). I April 1990 Asiatic Herpetological Research Vol.3, pp. 120-122 The Validity of Sacalia quadriocellata JINZHONG FU1 AND ERMI ZHAO2 ^Zoological Institute, Academia Sinica, Beijing, China ^Chengdu Institute of Biology, Academia Sinica, P. 0. Box 416, Chengdu, Sichuan, China Key Words: Reptilia, Testudines, Emydidae, Sacalia quadriocellata, China, taxonomic validity. Introduction Sacalia quadriocellata was first described by Siebenzock in 1903 as Clemmys bealei quadriocellata. Pope (1935) described Clemmys quadriocellata from Hainan, and compared it with Clemmys bealei. But most people except Pope thought that S. quadriocellata was a synonym of S. bealei. Sachsse (1984a) held that it was the female of S. bealei. Methods Specimens of Sacalia quadriocellata from Hainan and Guangxi provinces, along with specimens of S. bealei from Hainan, Fujian and Anhui provinces were examined. Data were taken on external, skull, and shell suture characters. Sexual differences were also examined. Results External differences 1) The dorsal surface of the head is uniform olive or chocolate brown in Sacalia quadriocellata , but it is vermiculated with black in S. bealei. 2) Sacalia quadriocellata have two ocelli on each side of the dorsal surface of the head. The ocelli always have distinct boundaries and there is one black spot within each ocellus (Fig. 1). In contrast, S. bealei have one or two ocelli on each side of the dorsal surface of the head (Fig. 1). If two ocelli are present, they may not have clear boundaries, but tend to run together. There are from one to three black spots within each ocellus. 3) The anterior margin of the carapace has many little black or chocolate brown speckles in Sacalia bealei . There are few or none in S. quadriocella . Differences in skull characters 1) The length from anterior of prefrontal to posterior of basi occipital to supra occipital is 2.59-2.62 (N = 3) in S. bealei ; but 2.92-3.39 (N = 3) in S. quadriocellata.. 2) The quadrate is not in contact with the opisthotic in Sacalia bealei, but in S. quadriocellata the posterior part of the quadrate tends to contact the opisthotic. abed FIG. 1. Ocelli of Sacalia quadriocellata and S. bealei. A: Male 5. quadriocellata. B: Female S. quadriocellata. C: S. bealei with a pair of ocelli. C: S. bealei with two pairs of ocelli. (Pi ioon v,„ April 1990 Asiatic Herpetological Research Vol. 3, p. 121 TABLE 1 . Analysis of shell suture data. Fig. 2 for details of measurement. See Sacalia Sacalia Measure bealei quadriocellata (N=27) (N=26) mean ± SE mean ± SE PA/PL 0.259+0.0180 0.249±0.0137 PB/PL 0.424±0.0240 0.417±0.0207 PC/PL 0.620±0.0280 0.616±0.0337 PD/PL 0.501±0.0198 0.520±0.0226 PE/PL 0.377±0.0140 0.367±0.0140 IG/PL 0.114±0.0140 0.10910.0117 IH/PL 0.105±0.0160 0.096±0.0150 IP/PL 0.171±0.0240 0.169±0.0172 IAB/PL 0.230±0.0177 0.248±0.0218 IF/PL 0.164±0.0140 0.157±0.0154 IAN/PL 0.194±0.0163 0.18010.0168 BL/PL 0.359±0.0200 0.37010.0163 FL/PL 0.387±10.050 0.38310.0150 FIG. 2. Means of measurement of shell sutures. 3) The posterior process of the jugal turns up in S. bealei, but turns down in S. quadriocellata. 4) The shortest distance between the orbital is located at the anterior prefrontal in Sacalia bealei. It is located at the joint of the frontal and prefrontal in S . quadriocellata. Shell suture differences Thirteen shell suture characters in S. bealei (N = 27) and S. quadriocellata (N = 26) were analyzed (Fig. 2). Seven characters (PA/PL, PB/PL, PE/PL, IH/PL, IAB/PL, IAN/PL, BL/PL) are different from each other (t, p< 0.05), [Table 1]. Sexual dimorphism Sexual dimorphism was evident in Sacalia quadriocellata. In life, males have very distinct orange-red speckles near the neck and limbs, whereas females have not. The carapace is narrower at the anterior end than the posterior end in females, but both ends are nearly equal in males. The most interesting character is ocelli coloration. They tend to be greyish with a white ring surrounding the two ocelli on each side in males (Fig. 1), but tend to be yellow without a white ring in females. The skulls of the males were somewhat flat, narrow and long. The skulls of the females were somewhat convex, wide and short. Discussion Since Sacalia bealei and S . quadriocellata have distinguishable characters, we consider 5. quadriocellata to be a valid name. Scalia quadriocellata occurs in China on Hainan Island and Guangxi Province. It is also found in central Annam. The specimens reported on by Zheng and Ding (1965) from Fujian Province and by Zhong (1981) from Jiangxi Province are misidentified S. bealei. The later species ranges over most of southern China, including Guizhou, Anhui, Jiangxi, Fujian, Guangdong provinces, Hainan Island, and Hong Kong. Vol. 3, p. 122 Asiatic Herpetological Research April 1990 Literature Cited Salamandra 11:20-26. FANG P. W. 1934. Notes on some Chelonians of China. Sinensia 4(7): 145- 199. POPE, C. H. 1935. The reptiles of China. Natural History of Central Asia, Vol. 10. American Museum of Natural History, New York. 604 pp. SACHSSE W. 1984. Chinemys reevesii var. unicolor und Clemmys bealei var. quadriocellata — Auspragungen von Sexual dormorphismus der beiden, Nominat formen. SCHMIDT K. P. 1927. The reptiles of Hainan. Bulletin American Museum of Natural History 65:395-465. ZHENG J. AND H. DING. 1965. [A preliminary survey of the turtles and tortoises from Fukien]. Journal of Fukien Teachers College 1:163-193. (In Chinese). ZHONGC. 1981. [Two new records of reptiles of Jiangxi Province]. Acta Herpetologica Sinica 1981,5(15):95-98. (In Chinese). April 1990 Asiatic Herpetological Research Vol. 3, pp. 123-i: Interim Report on the Freshwater Turtle Trade in Bangladesh S. M. A. RASHID1 AND IAN. R. SWINGLAND1 ^Durrell Institute of Conservation and Ecology, Rutherford College. University of Kent, Canterbury, Kent CT2 7NX, United Kingdom Key Words: Reptilia, Testudines, Bangladesh, conservation, commercial trade. Introduction Trade in the freshwater turtle species of Bangladesh has been occurring for a very long time. Exploitation of this natural resourse was limited prior to 1980, but during the past decade there has been a rapid and drastic increase both in terms of commercial exploitation and volume of trade. At present freshwater turtles are captured everywhere in the country. Since there are potential buyers, who have emerged due to the increase in turtle trade, the hunters prefer to sell their catches to those buyers rather than selling it in the small country markets. Turtle meat has served as a source of protein to most of the ethnic groups and non-moslems in Bangladesh. The turtle meat and eggs are mostly consumed by the Hindus and to some extent by the Christians, Buddhists and other minorities. Recent observations indicate that the rate of turtle meat consumption has accelerated due to the high price of other meat sources, making them unaffordable to most people, as well as the scarcity of fish and meat products. This trend has already proved to be a threat to the chelonian population. Many of the freshwater turtles are becoming rarer. Personal observations and interviews with the local people and hunters (turtle catchers) have confirmed it. Lack of turtle trade regulation is also one of the reasons for the increase in the magnitude of turtle trade. Though on paper the Forest Department, "godfather" of all the wildlife in Bangladesh, looks after it practically there is no one to execute the regulation. There is no regulation to control the hunting and capturing of the turtles. As a result, freshwater turtles are exported all round the year though the bulk fluctuates with the season. Some of the freshwater turtles are included in CITES I, but trade is still continuing. The Bangladesh Wildlife (Preservation Amendment) Act, 1973 does not include any turtle species in its schedules and as such gives free access to the exporters. Moreover the government also charges a nominal amount of duty (Taka 5.00 per maund) on the export. Study Period This is a preliminary report on the freshwater turtle trade in Bangladesh, covering the period from May 1989 to mid- August, 1989. The work is still in progress and the final report will be submitted after the completion of the study. Objectives The primary objectives of this study are to identify/determine: 1. Species involved in trade. 2. Volume of trade. 3. Proportion of each species exported. 4. Methods of collecting. 5. Major collecting sites, habitat preferences of the species concerned. 6. Sex ratio of the species exported. 7. Status of the species in the wild involved in trade. 8. Methods of transportation, packing, stocking. 9. Mortality rate during transportation. Previous Information The chelonian fauna of Bangladesh are © 1990 by Asiatic Herpetological Research Vol. 3, p. 124 Asiatic Herpetological Research April 1990 Year FIG. 1. Monetary value in Bangladesh Taka of turtle export from 1972 through 1987. The current exchange rate is approximately $1 US = Taka 30. Figures are as reported by the Bangladesh Export Promotion Bureau. not well documented. We have to rely mostly on Smith (1931), and it is necessary to revise and update the information for most of the chelonians. Later works include that of Ahmed (1958), Shaft and Quddus (1977), Husain (1979), Khan (1982,1985) and Fugler (1984). All these publications throw some light on the chelonians with some indication of the species being exploited for trade either locally or for export. Fugler (1984) put forward further information on the extent of trade and the species involved. But since he worked for a very short time much information is lacking. A recent study by Hosain (1989 unpublished) gives some information on the food and feeding habits of some freshwater turtles in Bangladesh. Information on turtle export is also on file at the Export Promotion Bureau (EPB). They only give the value of the turtles exported (Fig. 1). Neither the quantity nor the species are known. Methods Based on earlier information regarding the location of some of the turtle export centers, those centers were visited and the owners were briefed about our intention to collect information on the turtle trade. These centers were located at Baidyar Bazar, Sonargaon; Narayanganj (BIWTA Ghat, Jam Tala, Panchaputi); and Dhaka (Mirpur Section 1 and 10) [Fig. 2]. After seeking information about the export schedule, members of the study group were present physically at the centers observing the packing methods, taking measurements of the turtles exported, identifying and sexing them and also estimating the proportion of each species exported in the consignment. The group members had to face a lot of non- cooperative attitudes from the traders. But patience and interest proved worthwhile and when the traders understood that we are not doing any harm to their trade, they gradually came forward. At present the first author and two research assistants are working on this turtle trade project. The two export centers located within the metropolis Dhaka, Mirpur Sec. 1 and Sec. 10, are being visited weekly. Turtles are being exported every week and those are being monitored. The other centers are being visited once in a fortnight but it is not possible to check the turtle specimens there. The reason for this is that those centers are located far away and that in all the cases the turtles are packed for export after midnight so that they can be transported to the airport by dawn when most of the airlines are operating to the Far East. Most of these turtles are destined for Japan, Hong Kong, Singapore, Thailand and Malaysia where they are mostly used for food. Results Zia International Airport, Dhaka is the only shipping port. The turtles are exported live, packed in bamboo-woven wicker baskets. Two to three individuals weighing about 10 - 12 kg are packed in a single basket but when the specimen is large, it is packed singly. Prior to packing the turtles are washed and cleaned of any foreign material. The sturdy cord which is April 1990 Asiatic Herpetological Research Vol. 3, p. 125 BANGLADESH International Boundary District Boundary Subdivisional Boundary FIG. 2. Regional map of Bangladesh. Vol. 3, p. 126 Asiatic Herpetological Research April 1990 used to tie the fore and hind limbs of the turtle on each side together (to restrict its movement) is cut off and then the individuals are placed in the baskets to be packed. The lid is placed over the basket and tied with aluminium wires. The packed baskets are marked with the trademark of the exporter and then transported to the airport. Turtles under 1 kg in weight are not exported. While weighing, packing the exporters do not treat the different species separately. During this period (summer), the volume of trade is much less. Because of monsoon rain the water level rises and it becomes difficult to catch turtles. But according to some of the turtle hunters the catch is greater because of increased rate of movement of the turtles and the hunters can also maneuver and place traps and other collecting gears in suitable localities. The peak time for exporting the turtles is winter (late October- February). At that time the number of individuals as well as the number of species is higher compared to the summer catches. Information is being collected about the frequency of catches of the different species in different seasons and also in different habitats. During this interim period it has been observed that 1 metric ton of live freshwater turtles are exported every week. The volume is much less than during the peak time, in winter, when the volume rises to 5 - 6 metric tons per week. Presently turtles are being shipped once a week but during winter they are shipped almost everyday. The species exported during this period were Aspideretes hurum, and/4. gangeticus. Aspideretes hurum comprises the major portion of the bulk followed by A. gangeticus. Lissemys punctata is not being exported but smuggled to neighboring India with whom Bangladesh shares most of the boundaries. The sex ratio of the exported turtles have been found to be 45.71% males and 54.29% females irrespective of the species. The average weight of the exported turtles was 7.72 kg ranging from 1 kg to 33.20 kg. From one of the export centers in Dhaka 4017 kg of live turtles were exported from June 10th to August 21st . The ratio of the turtles, in terms of number of individuals were A. hurum 11.43% and A . gangeticus- 28.57% and in terms of weight was A. hurum 83.62%; A. gangeticus- 16.48%. Assuming that turtle demand for export is the same and that the number of people going for it is also the same, the decline in the export figures during 1987- 88 (Fig. 1) can be related to the decline of the turtle population in the wild. There are occasional reports of Kachuga tecta hatchlings being exported to Japan for pet trade. About 200 hatchlings were exported in 1988, and 40 kilograms of live Kachuga tecta hatchlings and juveniles were exported to Singapore last year. In June 1989, there was a consignment of 40 kilos of live Geochlemys hamiltoni shipped to Singapore. The prices of the turtles also vary considerably. The collectors sell A . hurum, A. gangeticus and Chitra indica at the rate of Taka 1200 - 1300 per maund (1 maund = 33 kg. approx.) to the middleman (locally known as maha ian or bePari) who in turn sells it at the rate of Taka 1400 - 1500 to the suppliers, who feed the exporters. The rate of the final exchange between the supplier and the exporter is not known. Lissemys punctata is bought at the rate of Taka 500 - 600 per maund from the collectors and sold by the middleman to the supplier at the rate of Taka 700 per maund (1 US $ = Taka 30). Last year at one time the prices went up to Taka 1 100 - 1200 per maund. The meat of the turtles which die off during transportation is consumed locally and is sold at the rate of Tk.50 - 55 per kilogram and that of L. punctata Tk. 15 per kg. Kachuga tecta is being sold in the country markets at the rate of Tk. 10 - Tk. 15 per kilo. Earlier the money was channeled down by the exporter to the supplier, who gave it to the middleman and finally it reached the collectors. The collectors were committed to the middleman to supply turtles at a rate determined by him. This trend has changed a lot now. The middleman invests his own money and the collectors negotiate to fix the April 1990 Asiatic Herpetological Research Vol. 3, p. 127 rate. This is because of the low rate of catches. The collectors complain of the non-availability of the turtles and the middleman has to comply with it. Recently the working group has been able to contact the airport officials in Dhaka by whom the exact weight of the turtles shipped is recorded. In most of the cases the weight figures measured in front of the group members at the turtle export centers are not the same as the airport figures. The airlines carrying it and the final destination of the shipment will be known. Very often this information is concealed by the exporters. Also the business organizations exporting turtles will be known. So far some of the exporters have been contacted. The Forest Department personnel have also been contacted and liaison is being maintained with them. Based on the information from the turtle suppliers about the collecting sites, the senior author and research assistants have travelled extensively to some areas in south Bangladesh to see the techniques used to collect chelonians and the habitat from where they are collected. Some of the major supply and collecting areas have been identified. Interviews have been taken of the traders and hunters and collectors. The areas visited thus far are Maijdi, Begumganj, Chandraganj, Lakhipur, Dalalbazar, Char, Alexandar, Ramgoti, Comilla, Laksham, Chandpur, Hajiganj, Sahatali, and Chittagong. Discussion Species Involved in Local and International Trade The freshwater turtles found in Bangladesh belong to the families Trionychidae and Emydidae (Testudines, Reptilia). Family Trionychidae Subfamily Cyclanorbinae 1. Lissemys punctata andersoni (Lacepede 1788), Flapshell Turtle. Subfamily Trionychinae 2. Aspideretes gangeticus (Meylan 1987), Softshell turtle 3. Aspideretes hurum (Meylan 1987), Peacock Softshell Turtle. Family Emydidae Subfamily Batagurinae 4. Geoclemmys hamiltoni (Gray 1831), Spotted Pond Turtle. 5. Morenia petersi (Anderson 1879), Eyed Turtle. 6. Hardella thurji (Gray 1831), Crowned River Turtle. 7. Kachuga tecta (Gray 1831), Roofed Turtle. 8. Kachuga tentoriaflaviv enter (Gray 1834), Tent Turtle. Classification follows Obst (1988). The generic name of Trionyx has been replaced by Aspideretes and the common English names are used following Stubbs (1989). There are reports of some of the other species being involved in the trade, like Chitra indica, Kachuga dhongoka and some others. But during the interim study period, no specimens of these species were observed. This gives rise to a great concern for particularly C. indica. In the past this species has been heavily exploited and the hunters and exporters believe that there is a serious decline in its population. Pelochelys bibroni was included in the chelonian list by Shafi and Quddus (1977), Husain (1979), and Khan (1982, 1985) but no precise localities were mentioned. Fugler (1984) was also not certain whether this species is included in trade or not. All the species mentioned above are consumed locally, mostly by the non- moslems. The rate of consumption has increased to a considerable extent, which needs to be monitored. Species Involved in International Trade 1. Lissemys punctata andersoni. 2. Aspideretes gangeticus. 3. Aspideretes hurum. Vol. 3, p. 128 Asiatic Herpetological Research April 1990 4. Kachuga tecta. 5. Hardella thurji. 6. Geoclemmys hamiltoni. Because of the wide distribution and availability of Lissemys punctata, it is consumed on most occasions. Apart from this, a good lot is also smuggled to India. The present investigator detected one smuggling route in eastern Bangladesh. There are reports of some more well established routes in the northwest and southwest Bangladesh. Kachuga tecta is also consumed quite often. It is estimated that most of the ethnic groups in the northeast, northwest and southeast consume at least either one Kachuga tecta or Lissemys punctata per week per household. Acknowledgments The authors are grateful to "Care for the Wild", Horsham, U.K., for sponsoring this project. Thanks are also due to Messers Md. Ghulam Mustafa, Rasheduzzaman Ahmed and Najmul Hasan for their tireless assistance in carrying out the study. Cooperation extended by Mr. Hashem, owner of the Mirpur, Sec. 10, Turtle Supply Center deserves mention. Last but not least, we thank Mr. Bill Jordan for his whole hearted cooperation from the very beginning, and Mr. Clifford Warwick for his suggestions. Literature Cited AHMAD, N. 1958. On edible turtles and tortoises of East Pakistan. Directory of Fisheries, East Pakistan. 18 pp. FUGLER, C. M. 1984. The commercially exploited Chelonia of Bangladesh: taxonomy, ecology, reproductive biology and ontogeny. Fisheries Information Bulletin 2(l):l-52. HOSAIN, L. 1989. Ecology of freshwater turtles of Bangladesh. Masters thesis. Department of Zoology, University of Dhaka. Unpublished. HUSAIN, K. Z. 1979. Bangladesher bonyajontu- swampad o tar songrakhshan. Bangla Academy Bignan Patrika 5(3):29-31 (In Bangla). KHAN, M. A. R. 1982. Chelonians of Bangladesh and their conservation. Journal of the Bombay Natural History Society 79:1 10-1 16. KHAN, M. A. R. 1985. Bangladesher bonya prani. Vol. I. Amphibian & Reptiles. Bangla Academy, Dhaka. 169 pp. (In Bangla). OBST, F. J. 1988. Turtles, tortoises and terrapins. St. Martin's Press, New York. 230 pp. SHAFI, M. AND M. M. A. QUDDUS. 1979. Bangladesher mothshya swampada. Bangla Academy Bignan Potrika 3(2): 14-36 (In Bangla). SMITH, M. 1931. The fauna of British India, including Ceylon and Burma. Reptilia and Amphibia. Vol. 1. Loricata and Testudines. Taylor and Francis, London. 185 pp. STUBBS, D. 1989. Tortoises and freshwater turtles. An action plan for their conservation. IUCN/SSC Tortoise and Freshwater Turtle Specialist Group. IUCN, Gland, Switzerland. 47 pp. I April 1990 Asiatic Herpetological Research Vol. 3, pp. 129-136 The Past and Present Situation of the Chinese Alligator Bihui Chen1 ^Department of Biology, Anhui Normal University, Wuhu, Anhui, China Key Words: Reptilia, Crocodilia, Alligatoridae, Alligator sinensis, China, endangered species, history, conservation. Introduction The Chinese Alligator is one of 21 species of existing crocodilians in the world today. Their numbers have dwindled and their distribution is so narrow that it has aroused concern among experts, conservationists, and amateurs in wildlife at home and abroad. It is reported in foreign countries that the Chinese Alligator is an extinct animal in the wilderness. It has been placed on the list of endangered species by the International Union for the Conservation of Nature and Natural Resources. How did the Chinese Alligator live? How does it live now? This is a problem of worldwide attention. Changes in the geographic distribution of the Chinese alligator The fossils of the Chinese Alligator have been found in Taian and Yanzhou in Shandong Province, Maqian in Shanghai, Yuyao in Zhejiang Province, and Hexian in Anhui Province. In addition, some unidentified fossil alligators have been discovered on the south border of the Junggar Basin in Xinjiang Uygur Autonomous Region; Jiulengshan, Douan in Guangxi Zhuang Autonomous Region; Nanjing in Jiangsu Province; and Danxian in Hainan Province. These fossils reveal that the range of the Chinese Alligator extended from Taian, Shandong Province, to Yuyao, Zhejing Province during the Neolithic age of the Recent epoch. If the unidentified alligator fossils are included, the range during the late Eocene and the beginning of the Oligocene would have extended from Shanghai and Yuyao, Zhejiang Province in the east, as far west as Douan, Guangxi Zhuang Autonomous Region to the border of Hainan Island in the south, and to the Junggar Basin, Xinjiang Uygur Autonomous Region in the north. According to ancient records which can be traced back to 3000 before present, the habitat of the Chinese Alligator was limited to the extensive lake and marshland of the middle-lower Yangtse River, along the banks of the Yangtse River from Shanghai to Jiangling City in Hubei Province, around Dongting Lake in southern Hubei and northern Hunan Provinces, including the extensive river network between the two provinces, and the Shaoxing, south of Hangzhou Bay in Zhejiang Province, approximately 28.5° - 32.5° N, 116.6° - 121.9° E. This range probable continued until the mid- 19th century. However, in many regions, the Chinese Alligator became extinct. Reliable records show that they became extinct in the south of Hangzhou Bay in 1201 AD. A great number were killed around Nanjing in the 1870's. The bank of the Yangtse River frequently collapsed, and flooding occurred. It was thought that the alligators picked holes along the bank and caused the disasters. Thus, people killed a great deal of alligators, and drove them to extinction there. The investigation in the 1950's revealed that the range of the Chinese Alligator stretched west from Pengze in Jiangxi Province, east to the west bank of Tai Lake, the north border of the Yangtse River, and south at the foot of Mt. Huang. To the north of Mt. Dongtianmu, 30.0° - 31.6° N, 118° - 120°E, the range has dwindled. Further supplementary investigations beginning in 1976 proved 1990 by Asiatic Herpetological Research Vol. 3, p. 130 Asiatic Herpetological Research April 1990 that the range became smaller and was limited to ponds of a hilly region to the north of Mt. Huang, below 200 meters in elevation. There were a few individuals that extended northward along the riverside plain, and even as far as the Yangtse River, approx. 30.6° - 31.6° N, 118.0° - 119.6°E. Within this range, they are sparse. The present habitat of the species is mainly located in some villages in Xuancheng, Nanling, Jingxian, Wuhu, Langqi, and Guangde Counties in Anhui Province, and is also situated around a few villages in Anji and Changxing counties in Zhejiang Province, adjacent to the Anhui border. In 1983 the Chinese Alligator's Natural Refuge organized the research workers of several counties to make a survey and statistically estimated that the number of alligators was about 500. Among the 200 animals captured in part by the investigation, and in part by the Research Center of Chinese Alligator Reproduction, only 4.6% were immature, and 95.4% were older than 10 years of age. Thus, the age pyramid was inverted. Other recent investigations on alligator eggs found that the number of eggs has gradually declined in recent years. The Research Center of Chinese Alligator Reproduction obtained 270 eggs in 1982, 278 eggs in 1983, 154 eggs in 1984, and 85 eggs in 1985. The eggs have not only declined in quantity. They also rarely hatched normally. Thus, it is clear that the wild Chinese Alligator population is declining. Reasons for the dwindling of the range and number of the Chinese Alligator Climatic change Fossil alligators were distributed in the Junggar Basin, Xinjiang Uygur Autonomous Region during the late Eocene to the Oligocene. In the Recent epoch, their range extended to the Yellow River. However, in our country it is recorded with reliable authority that there is no trace of them in these regions. This fact indicates that in some regions, the Chinese Alligator has long been extinct. Climatic variations are an important factor, and it is known that the worldwide climate became cold during ice-ages in the Quaternary period. According to information provided by Jiacheng Zhang, the greater ice-age of the Quaternary period in our country may be divided into six sub-ice-ages and six inter- ice-ages. The yearly average temperature of the inter-ice-ages was 3° - 6°C higher than that of the present, but the average yearly temperature of the sub-ice-ages was 6° - 12°C lower than that of the present. According to records in literature, in 903 BC and 897 BC, the Han River froze twice. In 366 AD, continuous freezing prevailed for three years on the surface of Bohai Bay from Changli to Yingkou. At that time, vehicles, horses, and troops 3000 to 40000 strong could pass over the ice surface. Carriages could pass over Taihu Lake when it was frozen in 1111 AD. The Chinese Alligator is an animal that is adapted to warm weather and not to cold. Late hibernation is an important period, then the reproductive organs develop. At low temperatures, reproduction cannot proceed normally. The hatching stage is about one month, and requires a temperature of about 30°C. If it is lower than 28°C, the young can hardly hatch. Thus it is certain that it is difficult for the Chinese Alligator to proliferate in chilly regions. They can only occupy the area south of the Yangtse River because there is a cool climate to the north. Habitat destruction, the most important factor The Chinese Alligator likes to live in water habitats such as ditches, ponds, reservoirs, etc. These habitats accumulate water year-round. These provide a mild and damp climate. Grass and trees grow luxuriantly, and numerous species of animals are common. This environment enables the Chinese Alligator to not only procure food, build holes, and mate in the water, but also to build nests and produce offspring on land. In the course of thousands of years, it was here that people reclaimed wasteland and built water conservation projects. Artificially cultivated plants were grown instead of natural vegetation. The alligator's holes April 1990 Asiatic Herpetological Research Vol. 3, p. 131 and nest-sites were destroyed extensively. All of this resulted in reduction of range and population size. We have investigated the habitat of the Chinese Alligator in the villages of Wanchun and Yitai, and the town of Qingshuihe on the outskirts of Wuhu. At present, there is level and open terrain with a vast extent of farmland, numerous villages, and well-developed roadways. The Shuiyang River lies to the east, flowing westward where it meets the Yangtse River. The Chinese Alligator is extinct in this region, which was altered radically about 80 years ago. Beach used to stretch for tens of miles along both sides of the Qingyi and Shuiyang Rivers. When the tide ebbed, reeds and other plants were exposed, providing habitat for many varieties of animals. When the tide was at flood, alligators were common on remote beaches in sparsely-inhabited areas. From the late 16th to the early 20th centuries, people from the north of the Yangtse River moved into this area, and began to cultivate an increasing amount of beach land. The Wanchun Embankment was built in 1904. When a dam was being built, alligators were common around the flooded plain. When a dam was completed, they still inhabited ponds, pools, and ditches within the dam. As the alligators dug holes and built dens, they often destroyed the dam, flooding seedlings, and endangering fish and ducks. Peasants made a point of hunting Chinese Alligators whenever they were discovered, and the number of individuals living in the embankment has declined. However, on the plain that had once been flooded, where the Shuiyang and the western Qingyi rivers meet, a great number of Chinese Alligators survived. In 1927, people began to build a dam at this spot. The Yitai Embankment was completed in 1931. Farmland and villages replaced the beach. Within the memory of the former generation, "Chinese Alligators could be heard roaring everywhere in summer -as much as frogs are heard croaking- echoing the whole neighborhood". It was here in 1935 that Z. D. Xiao conducted a study and described the state of the Chinese Alligator. There were a number of the animals in this region then. C. G. Zhu (1951-1956) made an investigation in the same region and only discovered alligators in Wuhu. This region was a desolate beach then. In 1954 I captured one alligator and made it onto a specimen. It has been preserved in the Department of Biology, Anhui Normal University. In 1959, the Wanchun floodgate was built to irrigate farmland. In the meantime, people built a complete set of engineering equipment and an irrigation canal. All Chinese Alligators were dug out and killed. The area became farmland, and the habitat of the Chinese Alligator was destroyed. A large-scale drive to eradicate the blood fluke was launched in 1958. A large amount of sodium pentachlorophenate was applied to the river basin. The last alligator at Yitai Village was poisoned. During the twenty-odd years that followed, Chinese Alligators and their sign were not found. The Qingshuihe River is only an example. Other locations met the same fate but in various degrees. People have multiplied considerably in the present range of the Chinese Alligator. They weeded all corners to get brushwood burnt. When alligators began to build nests and hatch eggs, sufficient weeds were needed. The weeds grew less and less, and in 1984 only one alligator deprived of a nest was captured by a peasant. Due to the fact that vegetation was destroyed, the areas of water in reserve had sharply dropped, and the area of wasteland to reclaim had greatly increased. The ponds and ditches that have never dried in history show frequent droughts at present. Moreover, because the forests have disappeared, soil and water cannot be conserved. There are rivers that flow into the Yangtse River (e.g. the Qingyi, Shuiyang, and Zhang rivers) whose beds have risen 1-2 meters in height, and have frequently flooded. Flood and drought have rendered the alligator's livelihood exceedingly difficult. When habitats incurred drought and flood, the alligator had to move away from the dry land or the submerged holes and look for habitat elsewhere. They were often captured or killed during the process of moving. Vol. 3, p. 132 Asiatic Herpetological Research April 1990 Excessive, indiscriminate capture or slaughter Chinese Alligators are often killed because they consume fish, ducklings, and small geese, and damage dams by digging holes. Thus they damage things that are of immediate benefit to people. Another significant factor is that the alligator supplies edible meat, useful skin, and medicinal materials. So, excessive and indiscriminate capture and slaughter has not been rare in the past, and remains common in the present day. For example, Guoxianjiayou, a book written during the reign of Jiajing in the Ming Dynasty, records: The bank of the upper Yangtse River near Nanjing often collapsed during the early years of the Ming Dynasty. It was due to this reason that the Chinese Alligators had herein drilled holes. It was reported that there was an old fisherman who had once said: 'Roast dogs were used as bait, put a hook; by raising the bait, you trapped alligators'. The above record was not mentioned in the history of the middle Ming Dynasty, which verified the fact that after excessive harvesting during the early years of the Ming Dynasty, Chinese Alligators living in the Nanjing area had become very scarce. The Classical Chinese Materia Medica, written by S. Z. Li, recorded the medicinal value of Chinese Alligators, and their meat was served as favorite dishes at wedding feasts. All these serve to verify that catching alligators was in great vogue at that time. After going through the great disaster of catching and killing during the Ming Dynasty, the alligator's population was reduced, and its range dwindled rapidly. During the Qing Dynasty, the government began to divide land into regions to provide cattle ranches for the Mandarin. For example, Wanchun Dan in Wuhu district was an area of land reserved for cattle ranches. People were forbidden to plough or sow in such land so that wild grass could grow densely. Such reservations gave alligators fine conditions for prolific reproduction. But the custom of killing alligators continued.. Peasants not only caught alligators by using fish hooks, but developed a new device of string bows set around the burrows. As soon as the alligators emerged, they were trapped. Application of large amounts of chemical fertilizers and insecticides in the field Alligators feed on snails, mother-of- pearl, shrimp, aquatic insects, fish, frogs, turtles, birds, and small mammals. Among these, aquatic animals are their chief food. Since a great quantity of chemical fertilizers and insecticides have been applied to farmland in recent years, the growth of the fish, shrimp, and mothers-of-pearl is affected when the polluted water flows into ponds. So, the decrease of natural food has greatly influenced the reproductive success and survival capacity of alligators. It is common for young alligators to die during the winter season due to insufficiency of food. Protection and captive breeding of the Chinese Alligator In order to ensure the survival of the Chinese Alligator, the Forestry Department of the China National Government, and the Anhui Provincial Government have taken two measures. 1. Conservation The first measure (in 1980) was to set up natural conservation effort for the species. The main goal was the protection and enlargement of the existing Chinese Alligator population and the protection of their habitat. Thus, a unit of leadership was established in the project which covers five counties with their respective species- protection work units. Reported here is the conservation effort: In order to observe and investigate the location and population size of the species, the conservation workers in each county employ identical approaches. They inquire with the local folks, explore caves, and calculate statistics using headlamps. Every possible water habitat is examined. Since alligators have stationary territories, observation in the field is convenient. A April 1990 Asiatic Herpetological Research Vol. 3, p. 133 stock of information resulted from the project, and provided valuable data for the development of the species. Through legislative measures, the National Government has ruled that the Chinese Alligator is a first-class rare animal to be conserved and protected. Hunting is not allowed. According to Item 1 30 of the National Offense Law, those who break hunting regulations by hunting in a restricted region, during a restricted period, by forbidden means, or harm rare animals, will be sentenced to two years in prison, or be subject to a heavy penalty. These regulations are strictly enforced by the police, who exercise their mission to keep the dignity of the law. It is openly advocated that the Chinese Alligator is one of China's rare and valuable animals, and belongs to our national resources. Local folks are enlightened and therefore change their view of the Chinese Alligator as a harmful species. Folks are also made to recognize the great significance of the Chinese Alligator in academic research and its considerable economic benefits. Thus, everyone begins to be concerned about the species, and protects it. The provincial and local governments also post notices to inform the public that protection of the Chinese Alligator is rewarded, while the killing of the species is punished. A full consideration of spatial and temporal factors strengthens the effectiveness of the conservation effort. Because the alligators presently have a fragmented distribution, their conservation is negatively affected by overcrowded human habitats with their large stretches of farmland and weaving highways. Considering these factors, many conservations stations are set up in the conservation area. If alligators are spotted in their caves, a conservation station is immediately set up with someone (usually a local farmer) who is responsible for the station and is paid from a special fund from the National Forestry Department. As the conservation station, no one is allowed to use chemicals, to cut grass, or to ravage the alligator nests and eggs. Consequently, the alligators in the conservation area enjoy proper protection. In some areas, alligators obtain the necessary conditions for reproduction, such as in Xuancheng, Nanling, and Jingxian counties, where in the past two years, alligators have been seen nesting, laying eggs, and hatching young. The alligator population in these areas is recovering and increasing in size. 2. Research In order to achieve the rapid and effective recovery of the Chinese Alligator population, the second measure taken by the Forestry Department and the Anhui Provincial Government was to set up the Anhui Research Center of Chinese Alligator Reproduction in close cooperation with the Department of Biology, Anhui Normal University. Here, a thorough study of the ecology of the species has been made. This has resulted in a systematic theory to be applied to captive breeding and artificial reproduction. Currently, the artificial incubation of Chinese Alligator eggs, and the raising of the young alligators has had much success. Details are reported as follows: Pen construction. A pen for alligators should be located in quiet marshy areas with appropriate temperatures and rich sources of food. After the location in decided, pens must be constructed with walls of brick, stone and cement. The height of the walls should be over two meters. When the rainy season comes, the walls may by surrounded by standing water, so the foundation should by 1.5 m below ground level, so as to keep the alligators from escaping by digging. The pen pond should be planted with trees and bushes, making it possible for the alligators to nest under the cover of vegetation. The land should be scattered with wild seeds to provide the necessary grass for nesting. The depth of the pond water should be kept at 0.5 m or more. Within the pond, small islands are built, where the alligators can enjoy sunshine. Captive breeding. Captive breeding efforts Vol. 3, p. 134 Asiatic Herpetological Research April 1990 must take into account such biological factors as sexual maturity, courtship, copulation, nesting, and hatching. At different stages of growth, alligators have different requirements as to the quality and quantity of food. At sexual maturity, alligators need many kinds of substances, so food should be prepared with variety. During courtship and copulation, when they are stimulated by sexual hormones, both males and females are very active, and will often fight for a sexual partner. Males and females should be grouped on a ratio of one male with three to five females. The water depth should meet the requirements of activity and copulation. For constructing nests, large quantities of grass should be provided in the reproduction areas, so as to provide sufficient nesting materials for the gravid females, and to prevent fighting, harmful interactions, and decrease in egg deposition. During the entire period of reproduction, female alligators must be kept in extreme quiet in order for them to perform. As the alligator's appetite is slightly decreased by the process of reproduction, feeding should be monitored. After egg deposition, alligators begin to eat more, and the usual level of feeding is resumed. Overfeeding is to be avoided, lest the alligators gain too much fat, which affects their health and lowers reproductive productivity. During the winter, more care should be given, and all alligators who have not entered their caves before hibernation should be captured and sent to artificial hibernation chambers. Egg deposition of captive alligators. Given the above requirements, captive alligators can lay eggs normally. The adult captive alligators we bred laid 264 eggs in 1983, 503 eggs in 1984, 809 eggs in 1985, 801 eggs in 1986, 1045 eggs in 1987, and 1219 eggs in 1988. Artificial incubation of alligator eggs Construction of incubation chambers The incubation of alligator eggs requires high temperature and humidity. In constructing incubation chambers, enough emphasis should be placed on temperature and means of raising temperature. The surface of the chamber walls should be waterproof. The walls and floor should allow for cleaning readily. The incubation chambers at he Chinese Alligator Research Center have a double glass roof, and a double glass wall that faces south. Temperature is controlled by the temperature-controller. The incubation chambers should not be too spacious, so as to save energy and allow for easy control. Equipment for incubation Make a round egg collector and a square incubator which can let water go through at the bottom, to keep the eggs from being suffocated by the standing water at the bottom. Plants which can help stabilize temperature and humidity, and allow for ventilation, best serve as the filling of the incubator. The research center finds that moss is a very good filling material. The search is continuing for better fillings. The chamber temperature is generally controlled and monitored by means of electricity. Techniques of incubation 1) Immediate collection of alligator eggs is required, as too many alligators in the reproduction area can possibly get the eggs crushed if not collected in time. 2) Egg collectors are used in collecting eggs. Quick action is required, and attention is to be given to the natural order of the eggs. The alligator eggs are stacked in baskets according to their natural order, and are sent to the incubation chambers immediately. Try to avoid shaking as much as possible in carrying the eggs. 3) Incubators and incubation chambers should be kept clean and be adequately sterilized to avoid bacterial or viral infections. 4) Alligator eggs are often in multilayer order when discovered in the wild. Artificial incubation requires monolayer arrangement. April 1990 Asiatic Herpetological Research Vol. 3, p. 135 5) Incubation temperature should be maintained at about 31°C. Humidity for the first four weeks is 95%. Use 85-95% humidity for the rest of the incubation period. Raising young alligators The mortality of young alligators is high in the first year, thus the key to breeding is the successful raising these juveniles. Construction of cages Cages for young alligators have requirements identical to those of the incubation chambers, as well as ventilation, sunlight, and adequate water and drainage. Methods of raising hatchlings The time between pipping and breakdown of the shell varies from a little over an hour, to two to three days. Hatchlings slow in liberating themselves are often poor in health, so they should be taken good care of. The hatchling has a crack 1.3-1.5 cm long in the vent in which some unconsumed yolk remains. This continues to provide sustenance for the hatchling, which needs no food during its early infancy. When to feed the hatchlings is still being studied. The newly-hatched alligators can be kept in the cage with enough area of water and dry sand and stone for the alligator to move about freely. Water supply and sanitary conditions are a significant factor in survival rate. Consequently, daily cleaning is required to remove food remains and excrement. When the hatchlings need to be fed, some Oryzias latipes can be thrown into the cage alive. Hatchlings in dry areas can be fed with fish meat in a small saucer, or with artificially-prepared food. Hatchlings are especially greedy; however, too much food can give rise to gout. If hatchlings are found in low spirits, tired of eating, or suffering from paralysis of the legs (a symptom of gout), feeding should be stopped immediately. Soon after the hatchlings liberate themselves from the shell, the weather gets cool, which could reduce their appetite. Now, the temperature should be raised. The temperature in the cage must be maintained at approximately 31°C. The hatchlings will recover their appetite and will gain weight rapidly. In fine weather when the temperature rises to its normal level, hatchlings should be given more of a chance to enjoy sunshine. Individual differences in growth rate will be evident by now, and the young alligators should be grouped accordingly. In the area around the cages it is desirable to keep quiet, cut down human interference, reduce the possibility of infection by pathogens, and to reduce the possibility of" predation on the alligators by mice. When the young alligators grow as weighty as 40 g, they can be sent into hibernation. Two or three days before hibernation, feeding is stopped so as to avoid diseases and ailment caused by food left in the digestive canal when temperature is lowered. It is desirable to reduce the temperature gradually until all food is completely digested. Temperature should be lowered by 2-3°C each time. Below 20°C, a larger decrease is allowable. 10°C is the proper hibernation temperature. Hibernation chambers can be set up underground, at a temperature of 10-12°C, and at 80% humidity. Late in hibernation, bowls of fresh water are supplied for the young alligators. To conclude, our breeding techniques generally agree with the natural growth of young alligators in the wild. Consequently, the population rises year by year. The number of surviving captive- bred Chinese Alligators is as follows: 66 in 1982, 77 in 1983, 117 in 1984, 300 in 1985, 450 in 1986, and 975 in 1987. Acknowledgments I thank Prof. Ermi Zhao for his kind help in correcting this paper. References CAO, K. 1984. [On the geographical and historical origins and the reasons for the decline of Alligator sinensis in China]. Acta Herpetologica Sinica 1984, 3(3):73-76. (In Vol. 3, p. 136 Asiatic Herpetological Research April 1990 Chinese). CHEN, B., S. WANG, AND B. WANG. 1984. [The rarely seen reptile animal — Chinese Alligator]. Anhui Science Press, Anhui, China. 120 pp. (In Chinese). CHEN, B., Z. HUA, AND B. LI. 1985. [Chinese Alligator]. Anhui Science Press, Anhui, China, pp. 186-239. (In Chinese). CHU, C. 1954. Chinese Alligator. Biological Bulletin 9: 9-11. WEN, H. AND Z. HUANG. 1981. [A discussion about the changes of the geographical distribution of the Chinese Alligators]. Acta Xiangtan University 1: 112-122. (In Chinese). XU, Q. AND C. HUANG. 1984. [Some problems in the evolution and distribution of alligators]. Vertebrate Paleontology 22(l):49-53. (In Chinese). ZHANG, J. AND Z. LIN. 1985. The [Chinese climate]. Shanghai Science and Technique Press, pp. 506-522. (In Chinese). CHU, C. 1957. [Observation of the life history of Chinese Alligator]. Acta Zoologica Sinica 9(2): 129-144. (In Chinese) ZHANG, M. AND Z. HUANG. 1978. (Treatise on the study of Reptilia). Journal of Haerbin Teachers Collage, 2. (In Chinese). HUANG, W., D. FANG, AND Y. YU. 1982. [Preliminary study on the fossil hominid skull and fauna of Hexian, Anhui]. Palaontologia Sinica, series C, 20(3):248-256. (In Chinese). ZHOU, B. 1982. [Skeletal remains of the Yangtse Alligator, discovered at the Wangyin Neolithic Site, Yanzhou, Shandong]. Acta Archeologica Sinica 2:251-262. (In Chinese) I April 1990 Asiatic Herpetological Research Vol. 3, pp. 137-140| Guidelines for Manuscript Preparation and Submission Summary Manuscripts must: 1) be written in English. 2) be of letter quality (typewritten on bond paper). 3) include camera ready figures (if any). 4) include complete and accurate literature citations. 5) include complete and accurate localities with latitude and longitude. 6) include a camera ready map illustrating regions discussed (when applicable). Manuscripts failing to meet these criteria will be returned for correction. Purpose and Content Asiatic Herpetological Research publishes articles concerning but not limited to Asian herpetology. The editors encourage publications from all countries in an attempt to create an open forum for the discussion of Asian herpetological research. Articles should be in standard scientific format and style. The following sections should be included: Title. The title should reflect the general content of the article in as few words as possible. A poor title may cause readers to read no further. Names and Addresses. The names and addresses of all authors must be complete enough to allow postal correspondence. Abstract. The abstract should briefly summarize the nature of the research, its results, and the main conclusions. Abstracts should be less than 300 words. Key Words. Key words provide an index for the filing of articles. Key words provide the following information (when applicable): 1) Taxonomy (e.g. Reptilia, Squamata, Gekkonidae, Gekko gecko ). 2) Geography (e.g. China, Thailand). 3) Subject (e.g. taxonomic validity, ecology, biogeography). The order of taxonomy, geography, and subject should be observed. Text. Manuscripts must be in English and spelling must be correct and consistent. Use Webster's New International Dictionary for reference. For clarity, use active voice whenever possible. For example, the following sentences in active voice are preferable to those in passive voice. Active voice: "Lizards were extremely common on the site." and "The three snakes examined were female." Passive voice: "Lizards were observed to be extremely common on the site." and "Three snakes were examined and were found to be female." Abbreviation. Do not abbreviate unless the full phrase has already appeared. Scientific names may be abbreviated only if they have appeared fully in the same paragraph. Never begin a sentence with an abbreviation of a scientific name. © 1990 by Asiatic Herpetological Research Vol. 3, p. 138 Asiatic Herpetological Research April 1990 Standard Format Manuscripts following standard format should include introduction, methods, results, and discussion sections. While other formats are acceptable, the editors encourage the use of standard format. Introduction. The introduction typically states the significance of the topic and reviews prior research. Methods. This section should clearly state where, when, and how research was carried out. Include sample sizes. Protocols designed by other investigators must be properly cited. Research materials and their manufacturers should be listed. The reader must be able to replicate the methods of the author(s). Results. This section states the results and their significance to the investigation. Figures and tables may be used to clarify, but not to replace, results statements in the text. Statistics should be used when applicable. Large amounts of data should be avoided, or included as an appendix at the end of the article. Discussion. The discussion is a synthesis of the introduction and the results. No new information should be discussed unless it was presented in the results section. New findings should be discussed in relation to prior research. The author(s) should feel free to present several possible interpretations of the results. The editors particularly encourage suggestions of future research in Asian herpetology. Section Headings Articles will be published using three section heading styles. Level 1: text is bold and centered Level 2: text is italic and centered. Level 3: text is italic and at the beginning of the first paragraph. Authors should take this style into account when writing manuscripts. Statistics Statistics must be accompanied by sample sizes, significance levels, and the names of any tests. Investigators should pay careful attention to independence and applicability of tests, and randomness of samples. One of the most frequent examples of nonindependence is the use of multiple, paired t-tests instead of analysis of variance (ANOVA). In general, multiple tests on the same data set are not valid. Descriptive statistics are in many cases more appropriate than inferential statistics. References Accurate and standard references are a crucial part of any article. This is especially important when dealing with publications from many different countries. The reader must be able to precisely identify any literature cited. References in the text must be checked for consistency with references in the literature cited section. All references cited in the text must be in the literature cited section. The literature cited section may not contain any references not mentioned in the text. Articles containing inaccurate or inconsistent literature citations will be returned for correction. References In Text 1) References to articles by one or two authors must include both surnames in the order they appear in the original publication. References to articles by more than two authors must include the first author's surname, followed by "et al." 2) The year of article follows the authors, separated only by a space. 3) References with the same author and year are distinguished by the lower case characters a, b, c, ... April 1990 Asiatic Herpetological Research Vol. 3, p. 139 4) References cited in text are listed in alphabetical order by first author. For example, "My results also incorporate literature records (Marx et al. 1982; Marx and Rabb 1972; Mertens 1930; Pope 1929; Wall 1909, 1910a, 1910b, 1910c). References In Literature Cited 1) References must include all authors, in the order that they appear in the original publication; "et al." is never used in a literature cited section. 2) The first author is listed surname first, initial(s) last. All other authors are listed initial(s) first, surname last. 3) References with the same author and year are distinguished by the lower case characters, "a, b, c, ..." 4) References cited are listed in alphabetical order by first author. 5) Names of journals are not abbreviated. See below for examples: Journal article. Dial, B. E. 1987. Energetics and performance during nest emergence and the hatchling frenzy in loggerhead sea turtles (Caretta caretta ). Herpetologica43(3):307-315. Journal article from a journal that uses year instead of volume. Gatten, R. E. Jr. 1974. Effect of nutritional state on the preferred body temperatures of turtles. Copeia 1974(4):912-917. Journal article, title translated, article not in English. Ananjeva, N. B. 1986. [On the validity of Megalochilus mystaceus (Pallas, 1776)]. Proceedings of the Zoological Institute, Leningrad 157:4-13. (In Russian). Note that for Acta Herpetologica Sinica, the year must precede the volume number. This is to distinguish between the old and new series, and between 1987, Vol. 6 numbers 1-4 and 1988, Vol. 6 numbers 1-2. Cai, M., J. Zhang, and D. Lin. 1985. [Preliminary observation on the embryonic development of Hynobius chinensis Guenther]. Acta Herpetologica Sinica 1985, 4(2): 177-180. (In Chinese). Book. Pratt, A. E. 1892. To the snows of Tibet through China. Longmans, Green, and Co., London. 268 pp. Article in book. Huey, R. B. 1982. Temperature, physiology, and the ecology of reptiles. Pp. 25-91. In C. Gans and F. H. Pough (eds.), Biology of the Reptilia, Vol. 12, Physiological Ecology. Academic Press, New York. Government publication. United States Environmental Data Service. 1968. Climatic Atlas of the United States. Environmental Data Service, Washington, D. C. Abstract of oral presentation Arnold, S. J. 1982. Are scale counts used in snake systematics heritable? SSAR/HL Annual Meeting. Raleigh, North Carolina. [Abstr]. Thesis or dissertation. Moody, S. 1980. Phylogenetic and historical biogeographical relationships of the genera in the Agamidae (Reptilia: Lacertilia). Ph.D. Thesis. University of Michigan. 373 pp. Anonymous, undated. Anonymous. Undated. Turpan brochure. Promotion Department of the National Tourism Administration of the People's Republic of China, China Travel and Tourism Press, Turpan, Vol. 3, p. 140 Asiatic Herpetological Research April 1990 Xinjiang Uygur Autonomous Region, China. Figures and Tables Figures and tables should be referenced in order in the text Each table should be typewritten, double spaced on a separate sheet. See below for instructions for Figures. Plates All figure plates submitted must be of publication quality, and should ideally be camera ready. All text in figures must be of typeset quality. Times Roman typeface is preferred. If typeset quality lettering is not possible for the author(s), Asiatic Herpetological Research will accept figure plates without lettering. The following instructions must be followed precisely: 1) Carefully label figures in pencil, on the back, or attach a photocopy or an additional sheet with instructions. 2) Do not submit figures with poor type or handwriting on the face of the figure. Substandard figures will be returned for correction. In order to avoid wasted effort, please follow the above instructions carefully. Figure Legends Figure legends should be typed on a separate sheet. Legends should explain the figure without reference to the text. A figure and legend should make sense if separated from the rest of the article. For example: FIG. 2. Lateral view of live Psammodynastes pulverulentus holding a prey lizard (Anolis carolinensis ). Note buccal tissue surrounding the enlarged anterior maxillary and dentary teeth of the snake. Copyright Asiatic Herpetological Research reserves the copyrights to all material published therein, except that excluded by permission of the editors. Any material under a prior copyright submitted to Asiatic Herpetological Research must be accompanied by the written consent of the copyright holder. Submission of Manuscripts Authors should submit letter quality, double spaced, single-sided manuscripts both in English and in the original language on 21.5 x 28 cm (8.5 x 11 inch) white bond paper. If possible, include a computer diskette containing the manuscript. Macintosh diskettes with Macwrite, Write Now, Microsoft Word, or text files, or MS/PC DOS diskettes with Word Perfect, Wordstar, Microsoft Word, RTF, or ASCII files are preferable. Computerized manuscripts should include italic, bold, and centered text only. Additional formatting is not necessary or desirable. Manuscripts will be reviewed. The editors will attempt to choose reviewers whose research knowledge most closely matches the content of the manuscript. Asiatic Herpetological Research requests $25 US per printed page from authors with funds available. Please indicate if funds are available. C.C. Liu 1900-1976 This volume of Asiatic Herpetological Research is dedicated to the memory of Cheng-chao Liu. This year is the 90th anniversary of his birth. C. C. Liu, China's most eminent herpetologist, was the author of 55 papers and two books on herpetological subjects. Liu died in Chengdu, Sichuan, China on 9 April 1976. (ISSN 1051-; CONTENTS ORLOV, NIKOLAI L. AND BORIS S. TUNIYEV. Three Species in the Vipera kaznakowi Complex (Eurosiberian Group) in the Caucasus: Their Present Distribution, Possible Genesis, and Phylogeny 1 ADLER, KRAIG AND ERMI ZHAO. Studies on Hynobiid Salamanders, With Description of a New Genus 37 WU, RUIMIN AND JIE HUANG. Relationships Between Serum T4, T3, Cortisol and the Metabolism of Chemical Energy Sources in the Cobra During Pre-hibernation, Hibernation and Post-hibernation 46 DAS, INDRANEIL AND PETER C. H. PRITCHARD. Intergradation Between Melanochelys trijuga trijuga and M . t. coronata (Testudines: Emydidae: Batagurinae) 52 LOSOS, JONATHAN B. Thermal Sensitivity of Sprinting and Clinging Performance in the Tokay Gecko {Gekko gecko ) 54 Mu, YONG AND ERMI ZHAO. Mating Call Structures of the Chinese Frog, Rana nigromaculata (Amphibia, Anura, Ranidae) 60 LAZELL, JAMES AND WENHUA LU. Four Remarkable Reptiles from South China Sea Islands, Hong Kong Territory 64 TUNIYEV, BORIS S. On the Independence of the Colchis Center of Amphibian and Reptile Speciation 67 PAN, JIONGHUA AND DANYU LIANG. Studies of the Early Embryonic Development of Rana rugulosa Wiegmann 85 ZHAO, ERMI. The Validity of E lap he perlacea, a Rare Endemic Snake from Sichuan Province, China 101 ananjeva, Natalia b., Gunther peters, J. Robert Macey, and Theodore J. PAPENFUSS. Stellio sacra (Smith 1935) - a Distinct Species of Asiatic Rock Agamid from Tibet 104 Tang, Yiquan, Shengkai Tian, Shixiang Wu, Jiacheng Hua, Xinquan Ji, Guanfu WU, ERMI ZHAO, AND GANG ZOU. Isolation and Amino Acid Sequence of a New Dodecapeptide from the Skin of Oreolalax pingii 116 Fu, JlNZHONG AND ERMI ZHAO. The Validity of Sacalia quadriocellata 120 RASHID, S. M. A. AND IAN R. SWINGLAND. Interim Report on the Freshwater Turtle Trade in Bangladesh 123 CHEN, BlHUl. The Past and Present Situation of the Chinese Alligator 129 Guidelines for Manuscript preparation and Submission 137 Harvard MCZ Library 3 2044 066 300 419