c° HARVARD UNIVERSITY \>V Ernst Mayr Library of th« t Museum of Comparative Zoo logy ASIATIC HEMATOLOGICAL Li ARVARD VERSITY RESEARCH . *5 ■ft * • VOLUME 5 1993 Asiatic Herpetological Research Editor Ermi Zhao Chengdu Institute of Biology, Academia Sinica, Chengdu, Sichuan, China Associate Editors J. Robert Macey Department of Biology, Washington University, St. Louis, Missouri, USA Theodore J. Papenfuss Museum of Vertebrate Zoology, University of California, Berkeley, California, USA Editorial Board KRAIG ADLER Cornell University, Ithaca, New York, USA NATALIA B. ANANJEVA Zoological Institute, St. Petersburg, Russia STEVEN C. ANDERSON University of the Pacific, Stockton, California, USA AARON BAUER Villanova University, Villanova, Pennsylvania, USA LEO BORKIN Zoological Institute, St. Petersburg, Russia BIHUI CHEN Anhui Normal University, Wuhu, Anhui, China ILYA DAREVSKY Zoological Institute, St. Petersburg, Russia INDRANEIL DAS Madras Crocodile Bank, Vadanemmeli Perur, Madras, India WILLIAM E. DUELLMAN University of Kansas, Lawrence, Kansas, USA HAJIME FUKADA Sennyuji Sannaicho, Higashiyamaku, Kyoto, Japan CARL GANS University of Michigan, Ann Arbor, Michigan, USA HUI-QING GU Hangzhou Teacher's College, Hangzhou, Zhejiang, China ROBERT F. INGER Field Museum, Chicago, Illinois, USA MAHMOUD LATIFI Institut d'Etat des serums et vaccins Razi, Teheran, Iran KUANGYANG LUE National Taiwan Normal University, Taipei, Taiwan, China RONALD MARLOW University of Nevada, Las Vegas, Nevada, USA ROBERT W. MURPHY Royal Ontario Museum, Toronto, Ontario, Canada GOREN NILSON University of Goteborg, Goteborg, Sweden HIDETOSHI OTA Department of Biology, University of the Ryukyus, Nishihara, Okinawa, Japan JIONG-HUA PAN South China Normal University, Guangzhou, Guangdong, China YUN-XU TONG Lanzhou University, Lanzhou, Gansu, China KE-MING XU Liaoning Normal University, Dalian, Liaoning, China YU-HUA YANG Sichuan University, Chengdu, Sichuan, China KEN-TANG ZHAO Suzhou Railway Teacher's College, Suzhou, Jiangsu, 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 herpetology. Authors should consult Guidelines for Manuscript Preparation and Submission at the end of this issue. All correspondence outside of the of China and requests for subscription should be sent to AHR, Museum of Vertebrate Zoology, University of California, Berkeley, California, USA 94720. All correspondence within China should be sent to Ermi Zhao, Editor, Chengdu Institute of Biology, P.O. Box 416,Chengdu, Sichuan Province, China. Subscription and membership are $25 per year ($45 for libraries). Postage outside of the USA and China, please add $5 per issue for surface mail or $10 per issue for air mail. 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 5 succeeds Volume 4 published in 1992, Volume 3 published in 1990 and 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 in June, 1988. Cover: Eublepharus turkmenicus from vicinity of Temen Spring, 2.5 km west of Danata (39° 07' N 55° 08' E), Krasnovodsk Region, Turkmenistan. Photo by J. Robert Macey. December 1993 Asiatic Herpetological Research Vol. 5, pp. l-To] i U Coluber atayevi Sp. Nov. (Ophidia, Colubridae) from the Kopet-Dag Mountains of Turkmenistan BORIS S. TUNIYEV1 AND SAHAT M. SHAMMAKOV2 Caucasian Slate Biosphere Reserve, Sochi, Russia ^Institute of Zoology, Turkmenistan Academy of Sciences, Azadi Street 6, 744000 Ashgabat, Turkmenistan Abstract. -An examination snakes, formerly considered to be Coluber najadum, from the Kopet-Dag Mountains of Turkmenistan, leads us to believe that this population represents an undescribed species. We here describe this population as Coluber atayevi. Key words: Reptilia, Ophidia, Colubridae, Coluber, Turkmenistan, systematics. Introduction During field work in the Kopet-Dag Mountains, Turkmenistan, we observed and analyzed many individuals of Coluber najadum under natural conditions. Comparison of these individuals with Caucasian material led to the conclusion that the Kopet-Dag snakes belong to a previously unrecognized species. This conclusion is supported by significant morphological divergence from the other representatives of the najadum-rubriceps complex. Methods We analyzed 10 specimens of Coluber najadum (Eichwald) from various areas of the Caucasian Isthmus and 10 specimens from the Kopet-Dag belonging to the new taxon. Morphometric data were compared with the available information in the literature about Coluber najadum in its natural habitat throughout its range (Terentjev and Chernov, 1949; Bannikov et al., 1977) and for separate regions (Ananjeva and Orlov, 1977; Muzskheleshvili, 1970). The following features and indices have been used: 1) L. = length of body, mm; 2) L. cd. = length of tail, mm; 3) Sq.= number of scales around body; 4) Ventr.= number of ventral scutes; 5) S. cd.= number of subcaudal scutes; 6) Lab.= number of upper labials; 7) Sublab.= number and size of sublabials; 8) form of mandibular scutes; 9) shape of the head; 10) distribution of scutes on throat; 11) L/L. cd.= body length/tail length; 12) Pr. oc.= number of preorbital scales; 13) Post. oc.= number of postorbital scales; 14) Temp.= number of temporal scales; 15) A.= form of anal scutes. For numerical features and indices, we have calculated the mean (x), mean error (m), mean square deviation (S2), using the formulas for small samples (Lakin, 1980) When describing the biotopes, we determined the plant species according to Nitikin and Geldykhanov (1988); the general vegetation type follows Korovin (1934), with some corrections. History of the Study of ''Coluber najadum'" in the Kopet-Dag Zamenis dahli Fitzinger was first mentioned from the environs of Sulukli Spring and the Kuchan road by Varentzov (1894). Nikolskij (1905, 1916) observed this species in the vicinity of Ashkhabad. On the basis of these records, Coluber najadum (Eichwald) was included in the list of the Turkmenistan herpetofauna (Chernov, 1934; Terentjev and Chernov, 1949; Bogdanov, 1962). All previous collections from the territory of Turkmenistan and the neighboring parts of Iran have been analyzed by Ananjeva and Orlov (1977). They added the localities of © 1993 by Asiatic Herpetological Research Vol. 5, p. 2 Asiatic Herpetological Research December 1993 TABLE 1 . Morphometry of Coluber atayevi paratypes in the collection of the Caucasian Reserve, Sochi, Russia (see text for abbreviations). No. L. L.cd. Sq. L./L.cd Ventr. S.cd. Lab. Preoc. Postoc. Temp. Sublab. 421 533 201 19 2.65 208 102 8/8 1/1 2/2 2+2/2+1 10/11 422 455 143 19 3.18 214 90 8/8 1/1 2/2 2+1/2+2 10/10 423 381 141 19 2.70 210 106 8/8 1/1 2/2 2+1/2+3 10/10 424 392 122 19 3.21 190 90 8/8 1/1 2/2 2+1/1+2 9/9 425 246 78 19 3.15 207 105 8/8 1/1 2/2 2+1/2+3 10/10 426 242 73 19 3.31 209 102 8/8 1/1 2/2 2+1/2+1 10/10 427 238 74 19 3.22 197 96 8/8 1/1 2/2 2+2/2+1 9/9 428 242 73 19 3.31 204 101 8/8 1/1 2/2 2+3/1+3 10/10 429 250 79 19 3.16 203 100 8/8 1/1 2/2 2+1/2+3 10/10 Firuza settlement and, tentatively, Dzhebel Station to the distribution of Coluber najadum. The last locality was doubtful (Shcherbak and Golubev, 1981; Ataev 1985). Rustamov and Shammakov (1979) and Shcherbak and Golubev (1981) mentioned it from Dushak Mountain. Recent localities include the Babazon region in the Kopet-Dag Reserve (Shcherbak, et al., 1986), Saivan and Imarat villages and Kara-Kala settlement (Ataev, et al., 1991). During almost a century, less than 20 representatives of this taxon have been recorded, half of them in recent years (Ataev etal., 1991). The small amount of preserved material, part of which had been lost (Ananjeva and Orlov, 1977), dissociation of time of collection and place of storage of specimens, led to the opinion that Coluber najadum was the taxon distributed in the Kopet-Dag. This form was included in the nominate form, and had never even been considered as a separate subspecies (Bannikov et al., 1977). It is interesting to note that Bannikov et al. (1977) included C. n. rubriceps Mertens (now recognized as a distinct species [Engelman et al., 1986; Rehak, 1986; Ananjeva et al., 1988]) in the synonymy of Coluber najadum. Coluber atayevi Shammakov, sp. nov. Tuniyev and Zamenis dahli: Varentzov, 1894:27; Nikolskij, 1905:233; 1916:92. Coluber najadum: Chernov, 1934:273; Terentjev and Chernov, 1949:240-241 (part); Bogdanov, 1962:167; Bannikov et al., 1977:262-263 (part); Ananjeva and Orlov, 1977:14-16; Rustamov and Shammakov, 1979:144; Shcherbak and Golubev, 1981:70-72; Ataev, 1985:242- 243; Shcherbak et al., 1986:98-100; Latifi, 1991:102-103. We name the new species in honor of the famous Turkmen herpetologist, Chary Ataevich Ataev, who studies reptiles of the mountains of Turkmenistan. Holotype: Collection of the Caucasian Reserve, Sochi, Russia, No. 420, adult male, environs of Saivan Village, Saivan- Nokhur Plateau, western Kopet-Dag, Bakharden Region, Turkmenistan, 12 May 1991, collected by B. S. Tuniyev (Fig. 1). Paratypes: Twenty two specimens. Collection of the Caucasian Reserve, Sochi, Nos. 421-429, 4 adults and 5 juveniles, same data as holotype, collected by C. A. Ataev and B. S. Tuniyev; California Academy of Sciences (CAS) Nos. 182948-182950, same locality as holotype, May 1990; CAS 185185- 185194, 5 adults and 13 juveniles, Elev. 1200-1300 m, 38° 30 N, 56° 47' E, 2 km SE (airline) of Saivan, Ashgabad Region Turkmenistan, 21 May 1992, collected by B. S.. Tuniyev, S. M. Shammakov, N. B. Ananjeva, T. J. Papenfuss and R. Macey (Plate 1). Tuniyev and Shammakow Asiatic Herpetological Research Adult Coluber atayevi. Type locality of Coiluber atayevi, environs of Saivan Village, Saivan-Nokhur Plateau, western Kopet-Dag, Bakharden Region, Turkmenistan. December 1993 Asiatic Herpetological Research Vol. 5, p. 3 FIG. 1 . Holotype of Coluber atayevi sp. nov. (Collection of the Caucasian Reserve no. 420). Description of holotype: Snout- vent length 500 mm, tail 172 mm; head length 19.6 mm, head width 8.4 mm, head height 6.2 mm. Head smoothly rounded, narrow, covered with large regular scutes; 8 upper labials, 5th upper labial touches lower postorbital and large lower temporal with its extended upper posterior side; 9 lower labials, 6th largest, last two pairs almost covered by upper labials; a single preorbital on either side, 2 small scutes below; posterior pair of upper temporals slightly larger than anterior one; seen from above, rostral extends slightly between internasals. Narrow genial scutes in contact along mental groove, no space between posterior genials. Nineteen scale rows at midbody; 206 abdominal scutes; 97 pairs of subcaudals; anal divided. Scales bordering abdominal scutes of same size as other lateral scales; body scales smooth, rhombic; ridge on lateral aspect of abdominal scutes indistinct, almost absent. Coloration in preservative: Dorsum gray, venter grayish; 5 large dark ocelli bordered with light circles on sides of neck; lateral row of small black dots ending abruptly on anterior third of body. Eye outlined with white lines extending anteriorly and posteriorly; lower white stripe extends over upper labials; a narrow black streak running posteriorly and down from eye, situated on 5th upper labial and slightly touching 6th. Description of paratypes: Counts and measurements of the paratypes in the collection of the Caucasian Reserve, Sochi, Russia are given in Table 1 . Diagnosis: Comparatively small snake (Fig. 2), smaller than C. n. najadum, C. n. dahli, and C. r. rubriceps in dimensions. It is comparable in size with the European subspecies, C. rubriceps thracius. In contrast to C. najadum, whose tail constitutes 1/3 of its total length, C. atayevi has a comparatively short tail, Vol. 5, p. 4 Asiatic Herpetological Research December 1993 -•iSfe, - ' FIG. 2. Representatives of the " najadum-rubriceps" -complex. Left- Largest specimen of Coluber atayevi sp. nov. (Collection of the Caucasian Reserve no. 421). Right- Medium-sized specimen of Coluber najadum (Sochi environs, Maly Akhun, Collection of the Caucasian Reserve no. 94). approximately 1/4 of total length, also characteristic of C. rubriceps. Habitus and elements of coloration of C. atayevi are intermediate between C . najadum and C. rubriceps; its narrow, sharp, and flat head with the rostrum beveled downward is closer to the C. rubriceps head shape than to that of C. najadum, with its comparatively wide, rounded, and high head, where the upper and lower surfaces of head are parallel. Color pattern of C. atayevi resembles that of C. najadum, but brown colors prevail instead of olive-green ones. Lateral abdominal ridges are practically absent in C. atayevi, in contrast to the above- mentioned species, and consequently the body is round in cross-section and not rectangular as in C. najadum and C . rubriceps. Genial scutes of C. atayevi contact one another along the mental groove, rarely having a few isolated granules between the posterior pair, whereas between the widely separated posterior genials of C. najadum there are always 2-4 rows of well- developed scales (Fig. 3). Posterior upper labials of C. atayevi are weakly December 1993 Asiatic Herpetological Research Vol. 5, p. 5 FIG. 3. Distribution of head scales on representatives of the Coluber najadum-Coluber rubriceps-complex: a, b, c- a young specimen of C. najadum; d, e, f- a young specimen of C. atayevi sp. nov. distinguished from the throat scales, whereas all upper labials of C. najadum are strongly pronounced. Geographic distribution: The range of Coluber atayevi includes the western and central Kopet-Dag, from the surroundings of the Kara-Kala settlement in the west to the Sulukli Spring in the east. This is an upland species, associated with such vegetation types as "prashiblyak," "broad- leaved forest" (Kamelin, 1970) and "phrygana," and in the western part of the uplands, where these plant associations occur at lower elevations, individuals of C. atayevi are found at elevations of 400-1600 meters (Shcherbak et al., 1991). At the eastern end of its range (Dushak Mountain), the snake has been found at 2000 meters elevation (Shcherbak et al., 1986). Biotopes: According to our observations, Coluber atayevi is found on the highest parts of the Saivano-Nokhur Plateau and on the crests of mountains at 900-1400 meters elevation. The most typical biotopes of the species are ecotones of mesophilous derivatives of deciduous forest and meadow-steppe coenosis along the edges of small ravines having deposits of limestone and argillaceous slates. Indicators of the deciduous forest are isolated old trees of Oriental plane (Platanus orientalis) and English walnut (Juglans regia). Forest plots of "prashiblyak" are typified by Aceretum fruticans and Acer turcomanicum, a subdominant role played by Crataegus turcomanica, Lonicera floribunda, Prunus cerasifera, Cotoneaster nummularioides, Cotoneaster ovatus, and Rubus anatolicus. In the herb layer, there are such species as Alliaria petiolata, Lamium album, Geranium pussilum, Arum juquemontii, and Allium paradoxum. Rocky-shrubby vegetation is usual for the ecotone of forest ravines on the rocky and scree slopes, with isolated trees of Celtis caucasica and many different shrubs and semi-shrubs: Colutea gracillis, Ephedra equisitina, Thelycrania meyeri, Rhamnus coriacea, Hymenocrater bituminosus, and Artemisia turcomanica. Representatives of Coluber atayevi are found in Cousinia smirnovii associations and Astragalus piletoclados groups (phrygana type) in the immediate vicinity of the forest and shrubby communities with steppe wedges, mainly in overgrazed places. The region is comparatively well- watered, because almost every ravine has a spring or stream. The vegetation of the eastern border of the distribution of C. atayevi is described by Korovin (1934): "Dushak mountain is an isolated massif, formed by light limestone. Its steep slopes serve as a home for typical mountainous xerophytes. Here we find juniper, both isolated trees and groups of them. A number of semishrubs and xerophytic herbs form flora of these mountains." Later, the author mentions the domination of shrubs of Astragalus piletoclados, as well as groups of different species of Acantholimon and here very common gray cushion-like groups of Onobrychis comma. Korovin concludes Vol. 5, p. 6 Asiatic Herpetological Research December 1993 that the cushion-like xerophytes of the Kopet-Dag are better developed on the tops of mountains (about 2000 meters elevation). This is higher than the steppe zone, so phrygana belongs to high-altitude vegetation. To our regret, absence of data about the biotopes of species from the other places gives us no opportunity to characterize the Cenozoic ties of the species throughout the whole area. Population density: Coluber atayevi is the most numerous snake species on the Saivano-Nokhur Plateau. Six specimens were found during a three-hour excursion in the vicinity of Saivan village in May, 1990; 12 specimens were noticed during the same period of time in May, 1991. The largest number of snakes (as many as 5 specimens per 300 meters) was among the shrubs of rocky-scree plots of ravines. Isolated specimens of snakes were met in stoneless places. The fact that this species is rather common for the western Kopet- Dag is proved by the data of Ataev et al. (1991). All other authors mention only isolated findings, reckoning it among the rarer species of the Kopet-Dag. Apparently, the sporadic distribution of the species and the considerable altitudes at which its habitat occurs, are the reasons why it is rarely met. It is not excluded that its population density in the eastern part of its range is significantly lower than in its western part. Seasonal and daily activity: Presumably, the species' activity begins in the middle of April, considering temperature conditions of this mountain zone (Babaev, et al., 1982) and preferable temperatures of daily activity, noted in May, 1990-1991. Coluber atayevi is a diurnal species with two-peak activity in May; the morning peak (9:00-11:00) is strongly pronounced and the evening one is feebly marked. The snakes are active in sunny, windless weather. We have never seen them when it rains, there are strong winds, or heavy overcast. In bad weather the snakes are absent not only on the surface, but from under the plates of slate, where they are usually met in sunny weather. Diet: Lizards in the habitat of Coluber atayevi are Ablepharus pannonicus, Stellio caucasicus, and Pseudopus apodus. Cyrtopodion caspius and Eremias strauchi are common, though not so numerous, while Mabuya aurata and Eumeces taeniolatus are rare. Considering the small size of the head and body of Coluber atayevi, the bulk of its diet must be formed only of Stellio caucasicus, young specimens of Eremias strauchi, and possibly Cyrtopodion caspius. Evidently, C. atayevi is a saurophage exclusively, since the common and abundant Microtus socialis and Saxetania cultricolis (a micromammal and an orthopteran insect, respectively) are too large to be eaten by this snake. Syntopical species of snakes: Subdominants of Coluber atayevi are C. nummifer, C. ravergieri and Vipera lebetina; Typhlops vermicularis is common; Agkistrodon halys caucasicus and Natrix tesselata are rare. At the borders of the species' biotopes Eirenis meda, Psammophis lineolatum, Naja oxiana, and Eryx miliaris are met. These species are more characteristic for smaller hypsometric marks. Discussion Coluber atayevi is most closely related to C. najadum and C. rubriceps, possessing features of both species; in habitus (head shape, in particular) it resembles C . rubriceps, but it is similar to C. najadum in color pattern. We should note that these features are characteristic for juvenile specimens as well as adults; in other words, we cannot distinguish ancestral features that would allow us to unite C. atayevi with either of the species mentioned (Table 2, Fig. 3). Judging from the contemporary distribution of the three species, we propose that the center of the complex is the Eastern Mediterranean region, so-called dry land of Asia Minor or Balkan-Caucasian December 1993 Asiatic Herpetological Research Vol. 5, p. 7 „ „ in in O & 5? §■ ri * 5 £^ ^ £^ J^ ri ri NO X. + _ ■; ". 7 oo" -^ oo — . ri + in w~« mi 0 rl ri r^ yr. no — no r- , OJ + + + + + ri r4 + m en n — rj — r\ - rT n «2 11 r+i - rJ ri rJ — — + + + + + + n — — — n ri >*> ^ „ -- X) 3 r- c O on lA-i • ■ • O — m Cr, 0 ^ no __; ON ri S 0^ ^ "(-in 1 C/3 c r-* # # e? t? 8 ON w w a ~2 ^^ t£ ^ ^ &■ in in 00 m ri mi xi 03 c oo — oo -^ in »n r-" ri c~ JC ! f*" r* *o •n fi ~ (*-. — ri ri 00 ^^ *-" w ' ■"- ' -^06 -r ri -^ >n ■ -J OO — • 00 — 1 oo — 00 r~ 00 on r-- 00 00 OO 1 — OO ON oc O on 00 r- 00 00 00 -- r- oo •d C z OO m r*"i <"*l — in z »n -r rl u +1 c ri in > r 8 o nO z J". ON 0 ri rj ri r. ri n ri 1 g in *n d- ~ ON ON r^ ri ON ON 1 i ON t/3 ™ 00 ™ ~™ — ' ON -^ — rj ON ^ -3 oo ON p. " f m O nO ir ~ ■ r nO r t oo O rn ri g ri ri ri ri T ri ri f 1 OO rJ X ^ ' J + oo O M3 • ■ ■ 3- c 0 O 0 C rJ ri NO in - 1 -r 2S in 0 0 J r-i +1 ' r-i rsi 2 2 in +1 NO ' ■ ' • J p i in in s £ r— ON *~* r-i vC ON CI ri § 01 o r-- 00 -r 00 r4 — oo o 2 o 5 u. OC r-- in +1 Tf vO 00 > O E oj -c U o 00 ® S r- r — ON OO OO t> On — ON ON ^^ > u i .c | y- X? ^ <* 'o ; i > z> CD 5 x: 0 5G- ft. «o ^ u c c/3 0 5 c 5 CO 3 %1 -a | 1^ T3 0 I ~ s s c Ira 6^ 1 s - 3 1 3 %> ^J s° 3 t. ) ■5 1 ,0 1 -1 s J t 'a1 '51 -C5 '5- £ u 1800 m. Localities of specimens morphologically intermediate between this race and Calotes v. versicolor arc indicated as half-circles.. radiating circumorbital bars are almost always evident, which also fade with age. Taxonomioc Considerations locality Pondicherry, India. Calotes versicolor (Daudin), Jerdon 1853:470. This report recognizes a subspecies if 75 percent of the available individuals from a geographic area can be correctly assigned to that provenance on the basis of one or more characters. Analysis of the data suggests that Calotes versicolor is divisible into at least two subspecies. Geographic discontinuities in the diagnostic characters are the basis for the generalized racial distributions shown in Figure 9. Calotes versicolor versicolor (Daudin) Agama versicolor Daudin 1802:395. Type locality "India". Agama tiedmanni Kuhl 1820:109. Type ? Calotes viridis Gray 1846:648. Type locality Madras. (Type specimen lost). A subspecies of C. versicolor distributed from Sri Lanka north through most of peninsular India and Pakistan, west to the Kabul Valley in southeastern Afghanistan, northeast to Hainan Island, China and southeast to Sumatra, Indonesia; replaced in the northern mountains of Pakistan and adjacent Afghanistan and India by C. v. nigrigularis (nov. ssp, described below). Other undescribed races probably replace this plains form in northeastern India, and lowland areas of Myanmar, Thailand, Malaysia and Sumatra. Vol. 5, p. 22 Asiatic Herpetological Research December 1993 - - ' ■ ■ v^v --r ■ - v~\ , , FIG. 10. Chin and throat color pattern in adult Calotes v. versicolor. A, FMNH/UF 7051 1, adult male, Karachi, Karachi Dist., Sindh Prov., Pakistan. B, FMNH/UF 19952, adult male, 8 mi. W. Madras, Tamil Nadu State, India. C. v. vesicolor has the following suite of characters which distinguishes it from C. v. nigrigulariss: larger adult size, mean angle of dorsal scale rows 90 - 105 , mean number two distinct postocular stripes (absent in largest males, Fig. 10A), dorsal body pattern usually indistinct, tending to uniform tan during most of the year, becoming pink to reddish in males during the breeding season; 5-7 crossbars may be present (particularly in juveniles and adult females), each 1-2 scales long at the vertebral line; gulars either uniformly light- colored or marked with narrow, diagonal, faint, dusky or sometimes black stripes; scales of the throat and pre-shoulder areas vary from the same color as the gulars to having dark brown or black bases. Large males from southern India often have a ventrally located black, partial collar (Fig. 11); the belly is always uniformly yellowish- to dirty-white. Several months before the breeding season, the color and pattern of adult males change. At this time the lateral and dorsal surfaces of the head, neck, and shoulders, and the sides of the body all become suffused with yellow, pink, orange, or even dull red (depending on geographic location, and age of the individual). The throat and chest change to orange or red with black mottling (seasonal adult color changes in Indian subcontinent populations © 1993 by Asiatic Herpetological Research also described by Murray 1886, Smith 1935, and Minton 1966); the tail and limbs become black. Holotype — Presumably in the Paris Museum, but now lost. The type locality had been simply stated as "India", but this was later restricted to Pondicherry, India by Kuhl (1820). Exemplary Material Examined. — The specimen materials examined by us that best fit the type description are listed below. We do not include any specimens here from the eastern Himalaya Mountains and from West Bengal eastward, as we believe that those populations will eventually be recognized as representing one or more races distinct from the nominate form and that one described below: FMNH/UF 19886, 19955-9, Kanheri Caves, nr. Borivli, Maharashtra State, India; FMNH/UF 70535-7, Khadiji Falls, Dadu Dist., Sindh Prov., Pakistan; FMNH/UF 19949-53, 8 mi. W. Madras, Tamil Nadu State, India;FMNH/UF 79087, 79099, Sujabad, Deri Ghazi Khan Dist., Punjab Prov., Pakistan; FMNH/UF 78932, Sonmiani, Las Bela Dist., Baluchistan Prov., Pakistan; AMNH 39377-8, 5 mi. E. Kalka and AMNH 39382, nr. Kalka, Amballa Dist., Punjab State, India; CAS 94337-8, 3 mi. SE Sirohi, Rajasthan State, India; MCZ 55502-3, Baroda, Gujarat December 1993 Asiatic Herpetological Research Vol. 5, p. 23 B FIG. 11. Chin and throat color pattern in Calotes v. nigrigularis. A, FMNH/UF 79470, adult male, nr. Chergal, Manshera Dist., NWFP, Pakistan. B, FMNH/UF, juvenile male, Miandam, Swat Dist., NWFP, Pakistan. C, FMNH/UF 70503, adult female, Abbottabad, Abbottabad Dist., NWFP, Pakistan. State, India; FMNH/UF 19884-5, New Delhi, India; FMNH/UF 78420, Multan, Multan Dist., Punjab Prov., Pakistan; ZSI 20798, 20800, Pali, Pali Dist., Rajasthan State, India; ZSI 1383, 13486, Ajmer, Rajasthan State, India; ZSI 20796, Jodhpur, Rajasthan State, India; BNHS 325-6, Wanothi, Kutch Dist., Gujarat State, India; BNHS 318 Bhavnagar, Rajkot Dist., Gujarat State, India; SMF 70074 Amritsar, Punjab State, India; SMF 61925, Bangalore, Karnataka State, India; SMF 55444, Meerut, Uttar Pradesh State, India; UMMZ 172083-90, Bhubaneswar, Orissa State, India; and BMNH 1923-3-445, Mirpur Sakro, Thatta Dist., Sindh State, Pakistan. Scute lation Characteristics. — Overall ranges and means for the scale characters of the samples of Calotes v. versicolor examined are as follows (Details regarding geographic variation in these parameters are found in the diagnosis above, in Figures 5- 8, and in the text discussions regarding them): Upper head scales unequal, smooth to feebly keeled; two well -separated spines on each side of the back of the head above the tympanum; canthus and superciliary ridge sharp; 11-13 (mean 12.0) infra- and 11-15 (mean 12.3) supralabials; dorsal scales large, distinctly keeled, all pointing backwards and upwards, larger than the ventral scales, which are always strongly keeled and mucronate, in 35-52 scale rows at midbody (mean 44.5); subdigital laminae of 4th toe 20 - 27 (mean 23.1); gular scales behind mental to middle of eye 7 to 15 (mean 11.2). SVL and Color Variation. — Juveniles have a dorsal color partem like that of adult females, except that the ground color is grayish, rather than the usual brownish, Vol. 5, p. 24 Asiatic Herpetological Research December 1993 and the dorso-lateral stripes are usually dirty white, rather than yellow. The chin and throat are the same color and pattern as in adult females, except that the lateral black diagonal stripes are usually better defined (Fig. 11). Nomenclature. — The type localities of both C. vultuosa Harlan (1825) and C. gigas Blyth (1853) (synonyms of Calotes versicolor) are Calcutta, West Bengal, India. Our studies show that C. versicolor from West Bengal exhibit a high level of character variation. They are excluded from our synonomy of C. v. versicolor on the basis that we cannot confidently place them in any named valid race at the present time, as they are intermediate between surrounding populations in many respects. Murray (1886) reported Calotes viridis Gray (1846) from Upper Sindh, Baluchistan, Punjab, southern India and the Deccan Plateau (this reference not included in C. versicolor synonomy given by Smith 1935). Murray's identification of some material from southern Pakistan as C. viridis is clearly incorrect, for no specimens referable to this name have been found there by any of the several thorough herpetologists who have worked extensively in Sindh Province since that time. The species has been considered a probable synonym of C. versicolor; the species type locality is Madras; the type specimen is lost. Because the original type locality of Calotes versicolor was imprecise, ("India"), Kuhl (1820) re-designated it as Pondicherry, India. It then follows that the peninsular Indian and Indus Valley (sensu latu) populations are to be given the name Calotes versicolor versicolor. In addition, we note an apparently distinct population of C. versicolor which occurs from Thailand, Myanmar, Assam, Sikkim, Darjeeling, and Nepal. However, we believe that taxonomic recognition of this population is not currently warranted until sample sizes are increased and fresher material becomes available for study. Specimens from this area differ from C. v. versicolor in having more mucronate dorsal scales, a lower number of scales under the fourth toe, a higher number of gular scales, wider body crossbands, in having many adults and subadults (in addition to in juveniles) with dusky longitudinal stripes on the belly, and the adults having a smaller SVL. Specimens we recognize as intermediate between these populations and those typical of Calotes v. versicolor occur in parts of Nepal and West Bengal. Geographic and Vertical Range. — Calotes v versicolor does not occur above 2000 m in the Indian Ghats (this study). Populations from the Himalaya Mountains (racially not yet defined), from Garwhal, India east through Sikkim and Bhutan are found to 2500 m elevation (this study), but to only 1030 m in Indochina (Smith 1935). Calotes v. versicolor (sensu latu) is distributed from the drier, more open forests of Sumatra and the Malay Peninsula north to near Hong Kong and Hainan, west through the mainland to southeastern Afghanistan, and eastern Iran, including the Andaman Islands and Sri Lanka. Additional study will undoubtedly lead to the recognition of additional races in the eastern parts of the range as here defined. If so, the nominate form C. v. versicolor will undoubtedly become restricted to those populations living in the lowlands of the Indo-Pakistan subcontinent. Calotes versicolor nigrigularis ssp. nov. Holotype.— FMNH/UF 79470 (Figs. 11, 12), adult male, on shrub on rocky hillside, Shargal, 20 km S Balakot, Manshera Dist., Northwest Frontier Province (lat. 34.3° N, long. 73.4° E), Pakistan. Pakistan Museum Natural Science field crew, June 15, 1990. Paratypes (N 16, all from Pakistan). — AZAD KASHMIR PROVINCE: FMNH/UF 79049, Gulpur; Kotli Dist., FMNH/UF 79396, 81165, Red Fort, Muzaffarabad, Muzaffarabad Dist.; FMNH/UF 79472, Chalpani, Muzaffarabad Dist.; FMNH/UF 79494, Panyola, Poonch Dist.; FMNH/UF 79495, December 1993 Asiatic Herpetological Research Vol. 5, p. 25 g£^K£ tk* U. n B FIG. 12. Side views of head of adult Calotes versicolor. A, C. v. nigrigularis, FMNH/U 79470, adult male, Chergal, Manshera Dist., NWFP, Pakistan.. B, Calotes v. versicolor, FMNH/UF 78926, adult male, Mach, Quetta Dist., Baluchistan, Pakistan. C, C. v. versicolor, FMNH/UF 70516, adult male, Karachi, Karachi Dist., Sindh Prov., Pakistan. Seri, Muzaffarabad Dist.; FMNH/UF 79601, Chela, Muzaffarabad Dist.; NORTHWEST FRONTIER PROVINCE: FMNH/UF 78944, Charsadda, Mandan Dist.; FMNH/UF 79229, Bahrain, Swat Dist.; FMNH/UF 79326, Miandam, Swat Dist.; FMNH/UF 79471 Khakai, Manshera Dist.; FMNH/UF 81133, 11 km W. Hungu, Togh Serai, Kohat Dist.; FMNH/UF 82243, Temargarh, Dir Dist.; FMNH/UF 70503, Abbottabad, Abbottabad Dist.; FMNH/UF 79462 Dharial, 4 km SW Balakot. PUNJAB PROVINCE: FMNH/UF 81136, Company Bagh, N. Tret, Rawalpindi Dist.; FMNH/UF 82242, 1.9 km SE Kohala, Rawalpindi Dist. Other exemplary material (all juveniles, or in poor condition): AZAD KASHMIR: FMNH/UF 82244, Pottri, nr. Bhimber, Mirpur Dist.; PUNJAB PROVINCE: ZSD 1231, Ghora Gali, Rawalpindi Dist.; FMNH/UF 82656, 1 1 km S Kohala, Rawalpindi Dist.; NORTHWEST FRONTIER PROV.: FMNH/UF 79138, Miandam, Swat Dist.; BNHS 313, Drosh, Chitral Dist.; FMNH/UF 82821, 1.1 km SW Garh Habibullah, Manshera Dist.; FMNH/UF 82822, 0.6 km SW Garh Habibullah, Manshera Dist.; BNHS 341, Parachinar, Kuram Dist.; FMNH/UF 81218, 13 km NE Abbottabad, Manshera Dist.; FMNH/UF 81093, 2 km W Hungu, Kohat Dist.; and FMNH/UF 82077, 5.8 km NW Khaki, Manshera Dist. Diagnosis. — Conspecific with Calotes versicolor on the basis of it's short head, the scales on the sides of the body pointing upwards and backwards, and that it lacks a fold or pit in front of the shoulder. It differs from the nominate race in having more strongly keeled (and usually more mucronate) body scales, more transverse scale rows at midbody, generally more Vol. 5, p. 26 Asiatic Herpetological Research December 1993 median gular scales from the tip of the jaw behind the mental to a level perpendicular to the middle of the eyes, fewer enlarged vertebral scales composing the in the nucho-dorsal crest, and in the adult state it lacks dark postocular stripes. During the breeding season the skin over the posterior part of the lower jaw in adult males (only) is jet black, except for a longitudinal median ventral band, which varies with season from pink to scarlet. The most vivid red color is found in the largest males during July and August. Each black gular patch extends (during the breeding season) posteriorly along the side of the neck, thence dorso- posteriorly at an upward angle to the vertebral line, including the entire shoulder region (Figs. 10, 12A). At the same time the entire head and dorsal neck surface are pinkish-red. In some individuals, both the red and black pattern may disappear at death. Adult males of this race lack the greatly swollen jaw muscle mass of the nominate form, resulting in a head that in top view has more parallel posterior borders behind the eyes than that of C. v. versicolor (where these edges are clearly divergent). In many individuals the toes are shorter than in those of the nominate population and the brachium is usually as long as the antibrachium; in C. v. versicolor the antibrachium is often shorter. Description of the Holotype. — Length of the head 1.48 times its width; snout broad, a little longer than the orbit; top of head from side slightly convex, slightly concave from the front; upper head scales unequal, smooth, to faintly keeled or tuberculate; canthus rostralis and supraciliary ridge sharp; two thin, spinous scales above the tympanum, the anterior one smallest, separated from the tympanum by about 5 scale rows; 13 supra- and infralabials; body somewhat compressed laterally; dorsal scales medium in size, distinctly keeled, most being mucronate, pointing backwards and upwards, larger than the ventrals, which are more strongly keeled and mucronate; 51 scales round the middle of the body. No gular pouch; gular scales like those of the ventrals, but larger. Nuchal and dorsal crests developed, composed anteriorly of lanceolate spines, gradually decreasing in size to the base of the tail. Limbs moderate; fingers 3 and 4 almost equal in length; toe 4 longer than 3. Tail rounded, covered with more or less equal- sized, strongly keeled, mucronate scales. The measurements (in mm) are as follows: total length 339; SVL 94; tail length 245; body length (axilla-groin) 47; greatest head length (snout tip to posterior extent of lower jaw) 34; greatest head width (across most posterior part of lower jaw) 23; greatest head height (just behind posterior edge of eye) 18.5; height of ear opening 3.1; length of brachium (axilla to elbow) 15; length of antibrachium (elbow to wrist) 13.6; posterior limb when extending anteriorly nearly reaches posterior edge of eye. The dorsal ground color is more or less grayish-tan (pinkish-tan in life) over the posterior 2/3 of the body. From the level of the posteriorly extended elbows to and onto the base of the tail. The body is very indistinctly marked with 3 slightly darker cross bars. Anteriorly it is almost completely black, being lightest along the vertebral line. Ventrally it is dark gray from near a line connecting the anterior edges of the shoulder posteriorly to just before the insertion of the hind limbs, where the color changes abruptly to grayish cream. The hind limbs and tail base are more or less uniform above and below, matching the colors of adjacent body surfaces. On the dorsal caudal surface, from about the level of the posteriorly adpressed knee to slightly beyond the claw tips of the hind foot, faint darker cross bars can be discerned, fading posteriorly as the tail becomes suffused with dark gray from its middle to the tip. The front limbs from shoulder to claw tips are uniform grayish black. The sides of the neck are black, continuing anteriorly to the black color of the limbs and the sides of the body. The most intense black on the entire individual occurs from the anterior lateral surface of the neck anteriorly onto the jaws and gular region. This black jaw marking is distinctly set off from the lighter color of December 1993 Asiatic Herpetological Research Vol. 5, p. 27 FIG. 13. General body color pattern of Cables v. nigrigularis; A, adult female; B, adult male in breeding coloration. the head and midgular areas, forming a more or less arrow-shaped black patch on each side (Fig. 10, 12A). Between the two black jaw patches is a dark pink (scarlet in life) median longitudinal band, beginning in the postmental region and extending onto the gular fold. The top of the head is medium gray, slightly mottled with grayish-tan. Laterally and posteriorly this fades into a pinkish-tan which covers all of the temporal areas and extends posteriorly in a V-shaped mark to above the shoulders. The eyelids above and below are light gray, with a nearly black spot in the anterior corner and a larger one in the posterior comer. Both the supra- and infralabials are light grayish-tan with faint grayish stripes radiating from the orbit. The lightest part of the body is in the area of the mental and surrounding shields. The most striking part of the entire color scheme is the black and scarlet gular pattern. Sex and Color Variation.. — Adult females lack distinct metachroic color changes (Fig. 13). The dominant dorsal color is gray to grayish-brown, usually with a narrow white to yellow dorsolateral stripe (sometimes represented by a series of dashes) from the neck to above the hind limb insertion. All these markings are variable in intensity and completeness. Along the vertebral area between the light stripes are 5 to 6 darker brown to black blotches or crossbands, 4 to 5 scales long, which in the largest females fade into the ground color. The ground color of the gular region is white or pale gray to pink (latter during the breeding season only). There are no large jet-black gular patches as found in adult males, though the base of the scales in this area may be dark gray (Fig. 11). Frequently the gular area is also marked with 5 to 7 more or less distinct black lines or dashes running postero- medially from the infralabials toward the midline. The reddish throat of adult males is first evident in individuals about 50 mm SVL (FMNH/UF 82656), i.e., at the end of the first year of life. The ventral surface of neonates of both sexes (mean ca 37 mm SVL), through nearly the entire first year is uniform dirty white. The smallest male with well defined black jowls with a red median area has a SVL of 76 mm (FMNH/UF 82623). Distribution. — This subspecies is restricted to the foothills and outliers of the Himalaya Mountains, from the Jhellum and Neelam River Valleys of Azad Kashmir, Pakistan, west to the Hindu Kush Mountains and foothills bordering the Kabul River Valley in southeastern Afghanistan, south to include the Safed Koh Range on the Pakistan-Afghanistan border (Fig. 1). It may extend further south to Waziristan or even Quetta, but this will only be proven with fresh material (see below). Within this area the race is apparently restricted to subtropical chir pine (Pinus roxburghii) and oak (Quercus Vol. 5, p. 28 Asiatic Herpetological Research December 1993 incana) forests, which are found at elevations between 1000 and 3000 m, depending on exposure and slope conditions. Individual character states of the plains race C. v. versicolor extend into the foothills along several of the larger rivers. Such changes coincide with the general floral change from the plains into the foothills. The Indus Valley (which further specimens may prove completely divides C. v. nigrigularis into isolated eastern and western populations) is an example (see remarks of intergradation under C. v. versicolor). Such intergrade populations nearly bisect the mountain range of C. v. nigrigularis in the Kabul River Valley near Peshawar (specimens in ZSDP, uncat.). We have not yet found any intermediate populations in the valleys of the Neelam or Jhellum Rivers. However, intergrades do occur near the foot of the Himalayan front range (Taxila and Islamabad, PMNH uncat.). The fact that specimens somewhat intermediate between the two races have been found as far south as Khuzdar, Baluchistan (BMNH H 1964-276-8) suggests that C. v. nigrigularis may eventually be found throughout the Quetta area and the Central Brahui Range as well, though no fresh specimens are available for study at this time. The only other intensive study of geographic variation of character states in Pakistan with which these results can be compared is our earlier study of Echis carinatus (Auffenberg and Rehman 1991). Like Calotes versicolor, this species is found over virtually all of Pakistan except the higher mountains. Calotes versicolor does not, however, occur in the sandy deserts of northwest Pakistan. In the Echis study we analyzed 12 characters. While each of these demonstrate a unique pattern of geographic variation, several features common to most of them stand out. These comprise what we believe to be five major centers of adaptive speciation in Echis carinatus - Transcaspia, Iranian Plateau, Astola Island, Indo-Gangetic Plain, Himalayan foothills, and the Cholistan-Thar Desert. Of these, the first three are essentially extralimital from the standpoint of the current study. The remaining three areas (Himalayan foothills, Indo-Gangetic Plain and Cholistan-Thar Desert) are also recognizable on the basis of distinctive character states, or combination thereof, in the characteristics of Calotes versicolor populations studied in this report. Thus, the Himalaya foothills populations are distinguishable from those of the Indo- Gangetic Plain and Cholistan-Thar Desert on the basis of several significantly different scale characters as well as a strikingly different metachroic color and pattern change in the adult males during the breeding season. Likewise, the Desert populations are different at statistically significant levels from those of all the surrounding Indo-Gangetic Plains populations in regard to certain scale characters. Though the the recommended nomenclatorial designation for these Echis and Calotes populations is different in each case, the correspondence of similar geographic patterns of variation is certainly important from the standpoints of both zoogeography and speciation in the subcontinent. We have not found any evidence for the curious mosaic of mean character states found in the Indus Delta region, as we did for Echis carinatus. The reason may be related to the fact that Calotes versicolor is often found in riverain forests, so that river and channel changes may be less important as an isolating mechanism in this species than in Echis carinatus. Acknowledgments We particularly thank the United States Fish and Wildlife Service (Washington), the Deutscher Akademischer Austauschdienst (Bonn, Germany), and the Office of Sponsored Research , University of Florida for providing funds to conduct this study. To all curators and collection managers of the institutions listed above, we extend our sincere thanks for the many ways in which they have contributed to the success of this project. Finally we wish to December 1993 Asiatic Herpetological Research Vol. 5, p. 29 acknowledge the support offered by our respective institutions. Literature Cited AUFFENBERG, W. AND H. REHMAN. 1991. Studies on Pakistan reptiles. Pt. 1. The genus Echis (Viperidae). Bulletin of the Florida Museum, Natural History 35(5):263- 314. BOULENGER, G. A. 1885. Catalog of the Lizards in the British Museum of Natural History. British Museum, London. 497 pp. KUHL, H. 1820. Beitrage zur Zoologei und vergleichenden Anatomic Frankfurt a. Mein. 1 14 pp. MURRAY, J. A. 1886. The Reptiles of Sind; A Systematic Account. Richardson and Co, London. 92 pp. SMITH, M. A. 1935. Fauna of British India, Vol. 2, Reptilia and Amphibia. Taylor and Francis, London. 440 pp. TIWARI, M. AND AUROFILIO (sic). 1990. Biology of the Indian garden lizard, Calotes versicolor (Daudin). Part I: Morphometries. Hamadryad 15(l):30-33. APPENDIX 1 Localities (to district only) from which specimens were examined, the museum collections in which they are found, and the number studied (in parentheses). AFGHANISTAN: BMNH (3); Jalalabad CAS (3). BANGLADESH: Chittagong MCZ (1). INDIA: Assam State: Chabus AMNH (1), Goalpara Dist., Raimona FMNH (3); Behar State: Benares BMNH (8), Patna BMNH (1); Gujarat State: Baroda MCZ (2); Bhaunagar BNHS (1), Hingolgadh BNHS (2) Rajkot BNHS (1), Kutch BNHS (2); Himachal Pradesh State: Amballa MCZ (2); Kulu Valley MCZ (2); Jammu-Kashmir State: Jammu BMNH (1); Karnataka State: Bangalore SMF (1); Maharashtra State: AMNH (1), Bombay FMNH/UF (7); Orissa State: Bhubaneswar UMMZ (13); Punjab State: BMNH (1), Amritsar SMF (1), Amballa Dist. nr. Kalka AMNH (3); Rajasthan State: Ajmer ZSI (2), Bikaner ZSD (1), Pali ZSI (1), Jodhpur BNHS (1) CAS (2), SMF (1), ZSI (3) Ml Abu/Abu Rd. CAS (1), AMNH (1); Nagaur ZSI (2), Jaipur ZSI (4); Tamil Nadu State: Madurai FMNH/UF (7); Uttar Pradesh State: nr, Chalu BMNH (1), Delhi FMNH/UF (2), Kanpur AMNH (1); Pitharagah (Kumaon) BMNH (1), Meerut SMF (1), Mussoorie ZSI (1); West Bengal State: Calcutta UMMZ (1), FMNH (2), MCZ (6), Darjeeling MCZ (1), Kalimpong Dist. Tarkhala MCZ (1). HONGKONG: BMNH (3). MY ANMAR ("Burma"): Rangoon FMNH/UF (27); Arakan FMNH/UF (4); Mandalay FMNH/UF (2); "at Chinese border" BMNH (1). MALAYSIA: Penang State: Penang FMNH/UF (8). NEPAL: BMNH (10), Katmandu SMF (1), Swayabonath SMF (1), Lapha Kamali Valley BMNH (1), Rasna Dist. BMNH (1), Maewa-Khola BMNH(12). PAKISTAN: Azad Kashmir Prov.: Muzzafarabad Dist. SMF (1), FMNH/UF (5); Kotli Dist. FMNH/UF (2); Poonch Dist. FMNH/UF (1); Mirpur Dist. FMNH/UF (2); Baluchistan Prov.: Kalat Dist., AMNH (1), BMNH (2), ZSI (1); Las Bela Dist. AMNH (4), FMNH/UF (2); Panjgur Dist., MCZ (1); Quetta Dist. FMNH/UF (1); Khuzdar Dist. BMNH (3); Waziristan Dist. BNHS (1); Northwest Frontier: Chitral Dist. BNHS (1); Dir Dist. FMNH/UF (1); Abbottabad Dist. ZSD (1), FMNH/UF (1); Kohat Dist. FMNH/UF (1); Kuram Dist. BNHS (1); Manshera Dist. ZSD (2), FMNH/UF (7);Peshawar Dist. BNHS (1), ZSDP (3), Swat Dist. FMNH/UF 6); Punjab Prov.: Dera Ghazi Khan Dist. FMNH/UF (2); Lahore Dist. ZSI (1); Kohat Dist. FMNH/UF (1); Multan Dist. ZSDM (4), FMNH/UF (1); Chakwal Dist. ZSI (2); Bahawalpur Dist. ZSDM (3); Rawalpindi Dist. ZSD (1), CAS (1), FMNH/UF (3); Sindh Prov.: FMNH (1); Badin Dist. ZSD (1); Dadu Dist. AMNH (1), ZSD (1), FMNH/UF (4); Hyderabad Dist. AMNH (7), BMNH (1), ZSD (1); Karachi Dist. AMNH (4), BMNH (4), CAS (7), FMNH (4), UMMZ (4), SDSNH (1), FMNH/UF (91), ZSD (30), ZSI (2); Thatta Dist. AMNH (7), UMMZ (1), ZSD (6), FMNH/UF (2); Thar Parkar Dist. BNHS (1), ZSD (2). REPUBLIC OF THE MALDIVE ISLANDS: Addu Atoll BMNH (2), Baras Isl. BMNH (1); Hulalay Isl. Bmnh (3); RAF BAse BMNH (1). Vol. 5, p. 30 Asiatic Herpetological Research December 1993 SIKKIM: BMNH (1); Mangpu FMNH (34), Teesta Valley MCZ (2). THAILAND: Bangkok Prov.: FMNH/UF (2); Chiang Mae Prov.: FMNH/UF (5); Mae Hong Son Prov.: FMNH/UF (1); Udon Thani Prov.: FMNH/UF (1); Yala Prov.: FMNH/UF (1); Phattabung Prov.: FMNH/UF (1). I December 1993 Asiatic Herpetological Research Vol. 5, pp. 3 1-44 J Holocene anurans from Caucasus ZBYNEK rocek Department of Paleontology, Geological Institute, Czechoslovak Academy of Sciences, Rozvojova 135, CS-165 00 Prague, Czech Republic Abstract. -Holocene deposits of the Kudaro I Cave from the vicinity of Ertso Lake (South Ossetia, NW Caucasus) yielded, among others, rather numerous disarticulated anuran bones. Examination of this sample revealed that majority of this material belongs to the genus Bufo and to the family Ranidae. This generally corresponds to the composition of the contemporary anuran fauna of that region. Key Words: Anura, Holocene, Caucasus, osteology. Introduction The material described in the present paper was recovered from the deposits of the cave Kudaro I. The cave is on the slope of Mt. Chasavalskaya (1600 m altitude), valley of Dzhordzhori River, in the vicinity of Ertso Lake, about 90 km NE from the town Kutaisi, South Ossetia, near Rachinsky ridge in the NW Caucasus (approx. 42° N, 43° W; see also Lyubin 1980a). The deposits are of the Holocene age (Lyubin 1980b). First description of the anuran and reptile material from this cave was published by Darevsky (1980). His taxonomic assignments generally agree with those in the present paper. The material consists of isolated bones; anuran bones described in this paper are deposited in the Zoological Institute of the USSR Academy of Sciences, St. Petersburg, under collection numbers ZIL/EL/1 - ZIL/EL/185. Anatomical terminology mostly follows that of Bolkay (1919) and Gaupp (1896). It should be noted that most of elements in the sample are postcranial bones and only few cranial bones are present. Fragmentary material bearing no diagnostic characters was excluded from the account below. Systematic Paleontology Bufo sp. Material: Left scapula, ZIL/EL/7 (Fig. 7B). Probably also ZIL/EL/1 40 (Fig. 7 A). Description: The margin of the cavitas glenoidalis is elevated. Although both the pars acromialis and proc. glenoidalis are broken off it is obvious that there was a deep incision between them. There is a moderately prominent and rather pointed outgrowth on the margo anterior. Both scapulae are comparatively big elements corresponding by their size to the below described humeri and ilia. Material: Humeri ZIL/EL/21 (Fig. 2B), ZIL/EL/25, ZIL/EL/55, ZIL/EL/79, ZIL/EL/1 12 (Fig. 2A), ZIL/EL/132 (Fig. 2C), ZIL/EL/139, ZIL/EL/153, ZIL/EL/173. These specimens (except for ZIL/EL/153 that includes also the most distal section of the crista ventralis) are preserved only as distal parts of the humerus. All of them are characteristic by conspicuous degree of development of the epicondylus medialis, the distal margin of which extends almost to the level of the distal margin of the caput humeri. Hence, there is a distinct notch between the both structures that can continue onto the dorsal surface of the distal section of the bone. On the ventral surface of the medial epicondylus one can discern an indistinct ridge running onto its distal surface. The lateral epicondylus is developed in lesser degree, extending laterally from the outline marked by the crista lateralis in some specimens (see Fig. 1993 by Asiatic Herpetological Research Vol. 5, p. 32 Asiatic Herpetological Research December 1993 2A, B). The lateral surface of the caput humeri is worn away in larger specimens, so the ball is not complete. The whole distal end of the bone is bent ventrally; this is correspondingly reflected on its dorsal surface which is convex along its longitudinal axis. Some variation may be observed concerning the extent of the medial and lateral cristae which might be, however, assigned to secondary sex differences. This might be suggested also by small specimen ZIL/EL/55 which may represent an immature individual, and in which both cristae are lacking. On the other hand, all specimens have their crista medialis directed laterally, so its ventral surface meets the medial surface of the diaphysis almost perpendicularly. In ZIL/EL/25 and ZIL/EL/132 the margins of both cristae are rather undulated and thickened. Discussion: These humeri (preserved only as distal sections) are morphologically closest to those of Bufo. In large with the determination of size-corresponding ilia. However, some (esp. smaller) humeri may fall into the variation range of the ranids, but the latter assignment lacks reliable foundation if only distal part of the bone is at the disposal. Discoglossids may be excluded because their medial and lateral cristae are confluent with the diaphysis, with no distinct border. The only fossil anuran that is of similar size as ZIL/EL/1 12 is Latonia seyfredi v. Mayer (= Discoglossus giganteus Wettstein- Westerheimb). However, its morphology and stratigraphic range are different (see e.g. Mlynarski, 1976, pi. 1/4). Material: Radioulnae ZIL/EL/43, ZIL/EL/62 (Fig. 8 A), ZIL/EL/1 06 (Fig. 8B), ZIL/EL/1 16, ZIL/EL/120 (Fig. 8C), ZIL/EL/1 62. Description: The margin of the olecranon rimming the articular cavity is formed either by calcified cartilage or ossified tissue lacking periostal bone. The border between the smooth periostal bone and rough surface rimming (and also covering) the articular cavity is distinctive. It seems that this most proximal part of the olecranon may be abraded in large specimens (e.g., ZIL/EL/62). On the inner surface of the bone, close to the point where the margins of the articulation cavity of the olecranon and capitulum meet with one another, is a small but deep pit. Similar pit is lacking or not so deep in ranids, but regularly present in Bufo. It serves as a muscle insertion area and in some specimens may be doubled. The posterior margin of the bone (i.e., of its ulnar part) bears an indistinct crista in some specimens. Material: Ilia ZIL/EL/41, ZIL/EL/66, ZIL/EL/75, ZIL/EL/94 (Fig. 4A), ZIL/EL/97, ZIL/EL/99 (Fig. 4B), ZIL/EL/1 24, ZIL/EL/1 29. Description: The ala ossis ilii, if compared with the posterior part of the bone, are stout (ZIL/EL/94) or rather slender (ZIL/EL/99). Their dorsal margin is rounded, only in the mid-part it becomes an edge distinctly bent medially along its whole extent. In its anterior part, the ala is compressed dorsoventrally, ellipsoid in cross-section. On the medial surface of the ala, approximately at the level of the highest point of their arch, there is an orifice of the narrow horizontal canal coming onto the bone surface from the posterior. There is certain variation in the location and morphology of this canal - it may continue as a groove for a short distance anteriorwards, and the orifice may be located on the bottom of a horizontal depression developed below the above- mentioned edge. The depression may terminate anteriorly on the dorsal surface of the ala or, in some specimens (esp. smaller ones), the orifice is located posterior to the depression. The torus superior is extensive, with two to three tubercles continuing (except the most anterior one) as a short and low ridges laterally. The anterior-most tubercle continues as a short rounded ridge anteroventrally, onto the medial surface of the bone. ZIL/EL/1 24 (and some other ones) is much smaller but except for size its morphology corresponds in all principal features to that described above. December 1993 Asiatic Herpetological Research Vol. 5, p. 33 «&& CM EM FIG.1. A- Rana sp. (ZIL/EL/148). Left humerus in ventral, dorsal and lateral views (from the left to right). B-fianasp. (ZIL/EL/151). Left humenis in ventral, dorsal and lateral view (from the left to right). C- Rana sp. (ZIL/EL/74). Ventral (left) and dorsal (right) view of the proximal section of theright humerus (drawing reversed for comparison). Bar equals 1mm. Abbreviations: C. A. - crista adventiva; C.L. - crista lateralis; C. M. - crista medialis; C. V. - crista ventralis; E. M. - epicondylus medialis; F. D. - fossula dividens; P. L. -processus lingualis. Discussion: The shape of the ilium corresponds to that in contemporary Bufo bufo and B. viridis. The only difference is that in both latter forms the longitudinal depression on the medial surface of the ala is developed in much lesser degree (due to lesser extent of the edge). It should also be noted that the orifice of the mentioned canal on the medial surface of the bone displays certain variation in contemporary forms (the orifice may be doubled, and the differences in this respect may be found also between the left and right ilium of a single individual). The same seems to hold for fossil material. Size differences may be ascribed either to interspecific variation or to differences between both sexes (the latter may reach quite a high degree in some contemporary representatives of the genus Bufo. Vol. 5, p. 34 Asiatic Herpetological Research December 1993 B D FIG. 2. A- Bufo sp. Right humerus (ZIL/EL/1 12). B- Bufo sp. Right humerus (ZIL/EL/21). C- Bufo sp. Left humerus, drawing reversed for comparison (ZIL/EL/1 32). D- Ranidae indet. Right humerus (ZIL/EL/150). E- Ranidae indet. Right humerus (ZIL/EL/1 10). Bar is 1 mm. Rana sp. Material: Humeri ZIL/EL/56, ZIL/EL/74 (Fig. 1C), ZIL/EL/83, ZIL/EL/128, ZIL/EL/1 48 (Fig. 1A), ZIL/EL/1 51 (Fig. IB), ZIL/EL/1 79. Description: The crista ventralis and crista ad ventiva delimit a wide, shallow and rather longitudinal depression for muscle insertion. The latter crista may be developed to various degree, whereas the crista ventralis is well developed in nearly all individuals, with a distinctive lingual process (only in ZIL/EL/128 this process is poorly developed, and the crista ventralis continues distally as a gradually lowering ridge). The crista ventralis has a hollow inside its free margin; consequently, it is thinner along its attachment to the diaphysis. This is manifested externally by grooves along the insertion of the crista, on :ither side. The crista lateralis and medialis are directed dorsally, forming thus a longitudinal groove on the dorsal surface of the bone. The proc. lingularis and the outgrowth produced by the crista adventiva may form together a roof over the fossula dividens; this nearly results in a canal. The caput humeri is well prominent ventrally (clearly seen in lateral aspect). Although the distal part of the bone is straight, the crista medialis and lateralis make it seemingly "S" shaped. Lateral epicondylus is entirely absent. Other features may be seen in Fig. IB. Some variation may be observed, mainly in the degree of development of the lingual process and in the extent of the medial and lateral cristae, as well as of the crista adventiva. Anatomical notes: As may be deduced from the condition in Rana esculenta (Gaupp, 1896) the depression between the crista ventralis and adventiva could serve as an area of insertion for three heads of the December 1993 Asiatic Herpetological Research Vol. 5, p. 35 FIG. 3. A- Ranidae indet. Right ilium in lateral view (ZIL/EL/87). B- Ranidae indet. Left ilium in lateral view (ZIL/EL/1). C- Ranidae indet. Lett ilium in lateral view (ZIL/EL/14). D- Anura indet. Left ilium in lateral view (ZIL/EL/3). B, C, and D reversed for comparison. Bar equals 1mm. Abbreviations: A. - acetabulum; A. O. I. - ala ossis ilii; C. O. I. - crista ossis ilii; P. A. - pars ascendens; P. C. - parscylindriformis; P. D. - pars descendens; T. S.- tuber superius. m. pectoralis (portio epicoracoidea, sternalis and abdominalis), whereas the proximal part of the crista itself (its edge) could serve for attachment of two heads of the m. deltoideus (pars clavicularis and scapularis). The third head of the deltoideus muscle (pars episternalis) is fixed to the ventral ridge of the medial epicondylus. The fossula dividens probably served for tendon of the m. coracoradialis. The groove between both the crista medialis and lateralis served no doubt for insertion of the caput profundum of the m. anconeus. The medial crista serves in anurans for attachment of the m. flexor carpi radialis and the lateral crista for the m. extensor carpi radialis. The former has its special function in amplexus. Consequently, the crista medialis is usually better developed in males, and the degree of its development is considered secondary sex character. Taxonomic notes: Humeri recalling those described above may be found in some individuals of contemporary "brown" frogs, i.e. of Rana temporaria R. arvalis, R. dalmatina, R. latastei, and R . macrocnemis. I found morphology closely resembling that in ZIL/EL/1 51 (Fig. IB) in contemporary Rana arvalis (DP FNSP 5830) from S Bohemia (Czechoslovakia), in R. arvalis wolterstorfii (DP FNSP 6264) from Soroksar (Hungary), and in R. latastei (DP FNSP 6419) from Como (Italy). In other individuals, the cristae and the lingual process were developed to various degree, similar to the condition in the described fossil material. In all cases these humeri belonged to males. Hence, it Vol. 5, p. 36 Asiatic Herpetological Research December 1993 FIG. 4. A- Bufo sp. Right ilium in lateral view (ZIL/EL/94). B-Bufo sp. Left ilium in lateral view (ZIL/EL/99). C- Ranidae indet. Left ilium in lateral view (ZIL/EL/68). D- Ranidae indet. Right ilium in lateral view (ZIL/EL/1 11). B and C reversed for comparison. Bar equals 1 mm. may be concluded that the described characters on the humerus may be ascribed to sexual dimorphism, but they are not present in all males. In any case, relations to contemporary "brown" ranids seems to be beyond any doubt. It is quite possible that the above described humeri assigned to Rana sp. and humeri (and other elements) described below as Ranidae indet. might belong to a single form. Material: (Fig. 7E). Ranidae indet. Right coracoid, ZIL/EL/54 Description: The intumescentia glenoidalis is circular, with distinct but small fovea acetabuli where the ligament of the humerus is inserted. The fovea is surrounded by marginal part for the cartilago paraglenoidalis that connects this bone with the proc. glenoidalis scapulae. The pars epicoracoidealis is nearly regularly arch-like, exceeding by its antero-posterior diameter the lateral part of the bone. Material: Humeri ZIL/EL/1 10 (Fig. 2E), ZIL/EL/1 50 (Fig. 2D), ZIL/EL/1 68. Description: The crista ventralis humeri in ZIL/EL/1 10 (and in ZIL/EL/1 68 that is similar) is prominent, gradually lowering distally. Part of its margin is tongue-like bent medially. Only within the proximal section of the crista there is a groove along the medial surface of its basis. The crista medialis is well developed, but only in the distal third of the bone. The lateral crista is developed in lesser degree. The medial December 1993 Asiatic Herpetological Research Vol. 5, p. 37 FIG. 5. Ranidae indet. Vertebrae in ventral view. A- Sacral vertebra (ZIL/EL/10). B- V6 (ZIL/EL/70). C- V2 (ZIL/EL/8). Bar equals 1 mm. Fig.6 Ranidae indet. Praesacral vertebrae in dorsal (above) and ventral (below) views. A- V5 (ZIL/EL/17). B- V2 (ZIL/EL/125). Bar equals 1 mm. Vol. 5, p. 38 Asiatic Herpetological Research December 1993 epicondylus is well developed, the lateral one is rudimentary. On the medial surface of the diaphysis there is low but discernible crista. Specimen ZIL/EL/150 has its ventral crista well developed, with rounded and almost straight margin. It is thin along its attachment to the diaphysis. This, together with the fact that the diaphysis is oval in cross-section, caused that there is a groove along the basis of the crista, especially well developed on its medial surface. Neither medial nor lateral crista are developed in this specimen, and the lateral epicondylus is absent, too. In spite of these differences, both humeri may be assigned to the Ranidae. Material: Ilia ZIL/EL/1 (Fig. 3B), ZIL/EL/2, ZIL/EL/14 (Fig. 3C), ZIL/EL/18, ZIL/EL/19, ZIL/EL/28, ZIL/EL/29, ZIL/EL/49, ZIL/EL/68 (Fig. 4C), ZIL/EL/87 (Fig. 3 A), ZIL/EL/1 11 (Fig. 4D), ZIL/EL/135, ZIL/EL/164, ZIL/EL/177. Specimens ZIL/EL/2, ZIL/EL/19, ZIL/EL/29, ZIL/EL/135 and ZIL/EL/177 are similar to ZIL/EL/1 1 1 (Fig. 4D) in that the tuber superius is continuous with the crista ilii. Description: The crista ossis ilii (vexillum of some authors) and the pars cylindriformis can be well distinguished in medial view, whereas only with some difficulties in lateral view. The crista does not reach up to the anterior end of the ala. It is directed dorsally in its section adjacent to the tuber superius, and bent dorsomedially in its anterior portion. Consequently, it forms wide groove on the medial surface of the ala, roofed dorsally by the edge of the crista, and ventrally delimited by the pars cylindriformis. The tuber superius is prominent above the level of the crista (but not in ZIL/EL/1 11), however, its margin corresponds by its thickness to the edge of the latter. Only posteriorly the tuber has a rounded margin, declined rather laterally. In specimen ZIL/EL/68 the tuber is prominent but not extensive; it has conspicuous ridge running down obliquely from its top. There is a distinct depression between the tuber and pars cylindriformis. The depression extends to the dorsal margin of the ala, separating thus the tuber from the crista. The acetabulum is with acute and prominent margins, considerably lifted above the pars descendens, but rather slanting down towards the pars ascendens. However, even here the margin of the acetabulum is represented by a distinct ridge. There is a considerable notch between the tuber superius and the dorsal margin of the pars ascendens. The dorsal margin of the pars ascendens continues anteroventrally onto the medial surface of the bone as a rounded ridge that disappears after a short distance. ZIL/EL/1 (Fig. 3B) is essentially the same but the crista is much lower than is the dorsoventral diameter of the pars cylindriformis, in whole its extent anterior to the tuber superius. Anteriorwards it is even getting lower, so its transition into the dorsal margin of the pars cylindriformis is indistinct. The tuber is prominent and extensive anteroposteriorly. Anatomical notes: The lateral surface of the crista ossis ilii is an insertion area for the m. iliacus externus, the other end of which is fixed by a tendon to the proximal part of the femur (Gaupp, 1896, figs 104, 105). The inner surface of the crista is occupied by the m. coccygeo-iliacus that runs to the urostyle. The iliacus externus muscle is, together with the iliacus internus, one of the most robust pelvic muscles in ranids and perhaps it plays an important role in jumping, despite of the fact that its tendon is fixed close to the proximal end of the femur. The tuber superius serves for attachment of the m. glutaeus magnus, m. ilio-fibularis, and m. ilio-femoralis. The first is the most robust muscle of the dorsal side of the thigh, and together with other heads of the m. triceps femoris it stretches the knee joint. All muscles that are inserted on the tuber superius are important for locomotion. Discussion: Contemporary European ranids mostly have the torus superior ilii continuous with the crista ilii, regardless if they belong to the esculenta or temporaria complexes. However, certain variation exists in this respect, so one can find individuals with prominent torus also in those forms in which it is continuous with December 1993 Asiatic Herpetological Research Vol. 5, p. 39 B FIG. 6. Ranidae indet. Prae sacral vertebrae in dorsal (above) and ventral (below) views. (ZIL/EL/17). B- V2(ZIL/EL125). Bar equals 1 mm. A- V5 the crista in most individuals. This is why more precise assignment is difficult. Material: Praesacral vertebrae V2 - ZIL/EL/8 (Fig. 5C) and ZIL/EL/125 (Fig. 6B); V3 - ZIL/EL/9; V5 - ZIL/EL/17 (Fig. 6A); V6 - ZIL/EL/70 (Fig. 5B). Description: The centra are procoelous, dorsoventrally compressed. In V2, the diapophyses are distinctly inclined anteriorly and slightly also ventrally; they are oval in cross-section due to moderate dorsoventral compression. ZIL/EL/125 is similar in its preserved characters but differs in having the postzygapophyses more robust, and the posterior convexity of the centrum more prominent (see Fig. 6B). Besides that, the anterior-posterior distance between the prae- and postzygapophyses is greater than in ZIL/EL/8 because the former processes extend anteriorly beyond the level of the centrum. The neural arches of ZIL/EL/125 (they are not preserved in ZIL/EL/8) produce distinct proc. spinosus which is, as usually in V2 of ranids, directed posteriorly. Its dorsal surface is flat, only anteriorly there is a narrow and low ridge paralleled by a depression on either side. V3 is represented by ZIL/EL/9 which is preserved only as fragment lacking the centrum, but its diapophysis with some rugosity in the middle of its length, as well as an extent of its neural canal and shape of its praezygapophysis, suggest its relations to the ranids. ZIL/EL/17 is V5; it has its proc. spinosus directed dorsally (again, as usual in ranids). Perhaps it might be associated with ZIL/EL/70 (see below), judging by the shape of the centrum in ventral view (also in this specimen the posterior convexity is divided by a slot, though visible only in posterior aspect). Peculiar feature of this specimen is the ventral edge of its anterior concavity which runs out anteriorly in the mid-line (see Fig. 6A). It is difficult to say whether this is of some taxonomic Vol. 5, p. 40 Asiatic Herpetological Research December 1993 FIG. 7. A- cf. Bufo sp., left scapula in lateral view (ZIL/EL/140). B- Bufo sp., left scapula in lateral view (ZIL/EL/7). C- Anura indet., left praearticular in dorsal view (ZIL/EL/12). D- Anura indet., parasphenoid (ZIL/EL/185). E-Ranidae indet., right coracoid in ventral view (ZIL/EL/54). F- Anura indet., right pterygoid (ZIL/EL/141). Abbreviationsx. gl. - cavitas glenoidalis; i. gl. - intumescentiaglenoidalis; m.a. - margo anterior; p. a. pars acromialis; p. gl. - proc. glenoidalis. Bar equals 1 mm. importance. V6 is represented by ZIL/EL/70. Its diapophyses are rounded in cross-section, and are of the same diameter both proximally and distally. They are horizontal, not inclined dorsally. The posterior convexity is clearly divided vertically by a slot which is better developed than in ZIL/EL/17. Signs of such slots may be observed in corresponding vertebrae of some individuals of the contemporary Ranidae (e.g., Rana esculenta). All the described praesacral vertebrae have in common a distinct indentation along the posterior edge December 1993 Asiatic Herpetological Research Vol. 5, p. 41 of the neural arches; this is interrupted only in the mid-line where a distinct ridge runs down from the proc. spinosus. All the above features suggest that vertebrae of the Ranidae should be concerned. Bufonids are excluded mainly because of the morphology of their proc. spinosus and because their neural canal is less spacious. Material: (Fig. 5A). Sacral vertebra ZIL/EL/10 Description: The centrum is dorsoventrally compressed, bicondylar posteriorly, both condyli being separated by a comparatively wide notch. The anterior side of the centrum is not preserved but comparison with contemporary ranids suggests that there could be a mineralized intervertebral disc. The diapophyses are inclined dorsally, and are distinctly compressed dorsoventrally. The articulation surface of the praezygapophyses is, in correspondence with the inclination of the diapophyses, facing dorsomedially. Anura indet. Material: Parasphenoid, ZIL/EL/185 (Fig. 7D). Description: The shape and proportions of the bone may be seen from Fig. 7D. Among the characters that should be mentioned are the proc. posterior which is well developed, narrow proximal parts (compared with the distal sections) of the lateral processes and of the pars medialis, and distinct indentations on the transition between the pars medialis and lateral processes caused by a low ridge on either side; similar ridge continues on both sides from the lateral edge of the proc. posterior onto the surface of the proc. lateralis where it disappears. Material: Left praearticular, ZIL/EL/12 (Fig. 7C). Description: The proc. coronoideus is well developed, nearly vertical in position. The sulcus pro cart. Meckeli is, especially in the posterior moiety of the bone, only moderately developed. These characters do not allow precise assignment. Material: Right pterygoid, ZIL/EL/141 (Fig. 7F). Description: The inner surface (margo orbitalis) of the ramus maxillaris does not bear any crista and the ramus itself is almost straight. The distal section of the ramus posterior is broken off, so the real proportions of the bone are difficult to reconstruct. Material: Ilium, ZIL/EL/3 (Fig. 3D). Description: Although this ilium is preserved only as a small section, important diagnostic characters are preserved. The crista ossis ilii is well developed, and may be distinguished both in medial and lateral view. In contrast to ranids, the torus superior is not developed, and the anterior margin of the pars descendens is directed posteroventrally instead of ventrally or even anteroventrally (see Bohme 1977, fig. 9). Material: Praesacral vertebra (most probably V5 or V6), ZIL/EL/1 1. Description: Only the centrum and bases of the left transverse process incl. adjacent praezygapophysis are preserved. However, one can conclude after the declination of the proximal part of the transverse process that V5 or V6 should be concerned. The centrum is dorsoventrally compressed and procoelous, though its posterior side is also slightly concave. As its surface does not display spongious structure (indicating a crack) it can be admitted that there could be a free intervertebral disc that in living animal adhered the posterior end of the centrum. The ventral surface of the centrum is almost at the same level as the proximal section of the transverse processes, and the centrum itself is short antero-posteriorly. The praezygapophysis is comparatively small and located far laterally (its distance from the lateral edge of the proximal concavity of the centrum is about half the diameter of Vol. 5, p. 42 Asiatic Herpetological Research December 1993 -OL D Fig. 8. Radioulnae in medial view. A- Bufo sp. (ZIL/EL/62). B-Bufo sp. (ZIL/EL/106). C- Bufo sp. (ZIL/EL/120). D- Anura indet. (ZIL/EL/36). E- Anura indet. (ZIL/EL/46). F- Anura indet. (ZIL/EL/72). D-F reversed for comparison. Arrows indecate border between periost and that part of the bone without periostal surface. Abbreviations: CAP. - capitulumradioulnae; OL. - olecranon; cr. - crista on the ulnar margin. Line equals 1 mm. this concavity). The neural arches are comparatively thin, and the neural canal was obviously quite spacious. Material: Urostyle, ZIL/EL/88. Description: This element fits morphologically into the variation range of contemporary Ranidae. Both in Ranidae and Bufonidae this range is rather broad which precludes precise assignment of the specimen. Material: Radioulnae ZIL/EL/36 (Fig. 8D), ZIL/EL/46 (Fig. 8E), ZIL/EL/72 (Fig. 8F), ZIL/EL/73, ZIL/EL/1 14, ZIL/EL/152. Description: These radioulnae are medium to small sized (compared with those identified as Bufo). A conspicuous character is that most of them are laterally compressed in their columnar section. This results in that their anterior and posterior margins run out in a distinct ridge. The smallest specimen (ZIL/EL/72), however, has its margins rounded. These radioulnae might be ascribed to the Ranidae, however, lack of diagnostic characters of these elements casts some doubts on this assignment. Accompanying Vertebrate Fauna in the Sample From Kudaro I Cave, Tsepkin (1980) gave a list of accompanying fishes, Darevsky (1980) mentioned one lizard genus (Lacerta sp.), Burchak-Abramovich (1980) gave a list of birds, Gadzhiev (1980) bats, Gromov & Fokanov (1980) December 1993 Asiatic Herpetological Research Vol. 5, p. 43 rodents, and Vereshchagin & Baryshnikov (1980) large mammals. In the sample that was placed at my disposal there were, besides frogs, also two different forms of birds, and following mammals (determination by Dr. Ivan Horacek, Department of Zoology, Charles University, Prague): Talpa cf. caeca Prometheomys schaposchnikovi, Arvicola cf. terresths, Microtus cf. gud, Microtus ("Pitymys") cf. majori, and cf. Lupus. References BANNIKOV, A. G., I. S. DAREVSKY, V. G. ISHCHENKO, A. K. RUSTAMOV, AND N. N. SHCHERBAK. 1977. [Key to determination of amphibians and reptiles of the fauna of the USSR]. Prosveshchenie, Moscow. (In Russian). BOHME, G. 1977. Zur Bestimmung quarterer Anuren Europas an Hand von Skelettelementen. Wiss. Zeitschr. Humb. Univ. Berlin, math.- nat., 26:283-300. Conclusions Determination of the material revealed that its substantial part belongs to the genus Bufo and to the family Ranidae. Minor part (represented by fragmentary or less numerous elements) could be determined only as Anura indet. Precise determination was impossible because of supposed individual and developmental variation. Nevertheless, results of this determination show that generic composition of the Holocene anuran fauna in this region was basically the same as contemporary one. The genus Bufo in the corresponding altitudes of Caucasus is nowadays represented by. Bufo verrucosissimus, and B. viridis, genus Rana by R. macrocnemis, and possibly also by R. ridibun da(Bannikov et al., 1977; Kuznetsov, 1974; Tuniyev, 1990). Besides, there occurs sporadically also Pelodytes caucasicus in South Ossetia, however, presence of this genus in the fossil material could not be confirmed. Acknowledgments I am grateful to Professor I. S. Darevsky (Zoological Institute, St. Petersberg) for the loan of the fossil material for study, and to Dr. I. Horacek (Department of Zoology, Charles University, Prague) for the determination of accompanying micromammalian fauna. Thanks are due also to Dr. B. Sanchiz (Museum of Natural History, Madrid) for his valuable suggestions. BOLKAY, S. 1919. Osnove uporedne osteologije anurskih batrahija. Glasnik Zemaljskog muzeja Bosni i Hercegovini, (1919):277-357. BURCHAK-ABRAMOVICH, N. I. 1980. [Remains of birds from the Kudaro I Cave]. Pp. 98-110. In Ivanova, I. K. and A. G. Tchemyakovsky (eds) Kudarskye peshchernye paleoliticheskie stoyanki v Yugo-Ossetii. Nauka, Moscow. (In Russian). DAREVSKY, I. S. 1980. [Amphibians and reptiles from Kudaro I cave]. Pp. 125-127. In Ivanova, I. K. and A. G. Tchernyakhovsky (eds) Kudarskye peshchernye paleoliticheskie stoyanki v Yugo- Osetii. Nauka, Moscow. (In Russian). GADZHIEV, D. V. 1980. [Remains of bats (Chiroptera) from the Kudaro I Cave]. Pp. 11- 124. In Ivanova, I. K. and A. G. Tchernyakhovsky (eds) Kudarskye peshchernye paleoliticheskie stoyanki v Yugo-Ossetii. Nauka, Moscow. (In Russian). GAUPP, E. 1896. Anatomie des Frosches. Lehre vom Skelet und vom Muskelsystem. Fridrich Vieweg und Sohn, Braunschweig. GROMOV, I. M. AND V. A. FOKANOV. 1980. [On remains of Late- Pleistocene rodents from the Kudaro I Cave].Pp. 79-89. In Ivanova, I. K. and A. G. Tchernyakhovsky (eds) Kudarskye peshchernye paleoliticheskie stoyanki v Yugo- Ossetii. Nauka, Moscow. (In Russian). KUZNETSOV, B. A. 1974. [Key to determination of vertebrate animals of the fauna of the USSR. I. Cyclostomata, fishes, amphibians and reptiles]. Prosveshchenie, Moscow. (In Russian). LYUBIN, V. P. 1980a [Geographical position of cave settlements of Yugo-Ossetia]. Pp. 6-12. Vol. 5, p. 44 Asiatic Herpetological Research December 1993 In Ivanova, I. K. and A. G. Tchemyakhovsky (eds) Kudarskye peshchernye paleoliticheskie stoyanki v Yugo-Ossetii. Nauka, Moscow. (In Russian). LYUBIN, V. P. 1980b. [Geological-stratigraphic conditions of paleolithic deposition in eastern gallery of the Kudaro I Cave]. Pp. 13-32. In Ivanova, I. K. and A. G. Tchemyakhovsky (eds) Kudarskye peshchemye paleoliticheskie stoyanki v Yugo- Ossetii. Nauka, Moscow. (In Russian). MLYNARSKI.M. 1976. Discoglossus giganteusWellslein- Westerheimb, 1955 (Discoglossidae, Anura) from the Miocene of Przeworno in Silesia (Poland). Acta Zool. Cracov. 21:1.12. TSEPKIN, E. A. 1980. [Remains of fishes from the Kudaro I Cave]. Pp. 90-97 In Ivanova, I. K. and A. G. Tchemyakhovsky (eds) Kudarskye peschemye paleoliticheskie stoyanki v Yugo- Ossetii. Nauka, Moscow. (In Russian). TUNIYEV, B. S. 1990. On the independence of the Colchis Center of amphibian and reptile speciation. Asiatic Herpetological Research 3:67-84. VERESHCHAGIN, N. K. AND G. F. BARYSHNIKOV. 1980. [Remains of mammals in eastern gallery of the Kudaro I Cave (excavations made by V. P. Lyubin in 1957- 1958)]. Pp. 51-62 In Ivanova, I. K. and A. G. Tchemyakhovsky (eds) Kudarskye peschemye paleoliticheskie stoyanki v Yugo-Ossetii, pp. 51-62. (In Russian). I December 1993 Asiatic Herpetological Research Vol. 5, pp. 45-50 Karyotype, C-Band and Ag-Nors Study of Three Stink Frogs Gang Wei1, ning Xu1, dejun Li1, Guanfu wu2 and xiquan Song3 'Department of Biology, Zunyi Medical College, 563003 Guizhou, China 2Chengdu Institute of Biology, Academia Sinica, 610015 Sichuan, China ■* Department of Biology, Zunyi Teachers College, 563002 Guizhou. China Abstract.-The karyotypes.C-bands and Ag-NORs of Rana kuangwuensis, R.andersonii and R.margaratae were analyzed. Intra- and interspecific chromosome variations, including centromeric type and C-banding patterns, were detected. It was assumed that the Guizhou Plateau was the distributional center of the origi- nal place of the group. Key Words: Amphibia, Ranidae, Rana kuangwuensis, Rana andersonii, Rana margaratae, China, kary- otype, C-band, Ag-NORs. U«M«K"*» lYltfXXftlAAAA •t}f!Mtt(,t,«tt|,|'*M FIG. 1. a: Karyotype of Rana andersonii. b: showing C-bands. c: showing Ag-NORs. Introduction The group of stink frogs which have a special stink smell from the skin consists of nine species, i.e. Rana andersonii, R. an- lungensis, R. grahami, R. kuangwuensis, R. lungshengensis, R. margaratae, R. schmackeri, R. tiannanensis and R. wuchuanensis. They are considered to be phylogenetically close because of similar morphological characters in adults and tad- poles. Among them, the karyotype and C- bands of R. grahami from Kunming, Yun- nan has been studied by Li (1982). In ad- dition, the karyotype, C-bands and Ag- NORs of R. margaratae from Emei Moun- tain, Sichuan have been analyzed by Wang (1983) and Wu (1990). In the present pa- per, the karyotypes, C-bands and Ag- NORs of R. kuangwuensis, R. andersonii and R. margaratae were analyzed. Methods Two females and one male R. andersonii were captured at Qianxi, Guizhou Province, (27°20' N,106°16' E). One fe- male and three male R. kuangwuensis were captured at Nanjiang, Sichuan Province (32°30' N, 106°40' E) and one female and 1993 by Asiatic Herpetological Research Vol. 5, p. 46 Asiatic Herpetological Research December 1993 ^{((IK11'1"1""1" at XilYA if H h }] ,| M .1 •» I! f. v r vtffVWM* ik mMm^mw^m A A RG.2. a: Karyotype of Rana kuangwuensis. b: showing C-bands. c: showing Ag-NORs. one male R. margaratae from Zunyi (27°40' N, 106°50' E) were captured. Karyotypes, C-bands and Ag-NORs preparations were made after Wei et al. (1990). Results Figures 1, 2, and 3 depict the karyotypes, C-bands and Ag-NORs of/?. andersonii, R. kuangwuensis and R. margaratae. For the measurment of the karyotypes see table 1 . The diploid number of the species are all the same, 2n=26, comprising two groups. The large chromosome group includes chromosome Nos. 1-5, with a relative length (R.L.) larger than 9%. With regard to the arm ratio (A.R.), chromosome Nos. 1 and 5 are metacentric in all three species. No. 2 is metacentric in R. margaratae and R. andersonii, but submetacentric in R. kuangwuensis. No.3 is submetacentric in R. margaratae and R. kuangwuensis , but metacentric in R. andersonii. No.4 is sub- metacentric in R. andersonii and R. mar- garatae, but metacentric in R. kuangwuen- sis. The small chromosome group comprises chromosome Nos. 6-13, with a R.L. less than 7%. Nos.6, 8, 10, 12 and 13 are metacentric, No.7 is submetacentric in all the three species. Nos. 9 and 1 1 are sub- metacentric in R. kuangwuensis and R. an- dersonii but metacentric in R. margaratae. Secondary constrictions are observed in the long arms of No. 10 of R. margaratae (only one homologous) and R. andersonii but not observed in R. kuangwuensis. Treatment of the chromosome of the three species according to the C-banding method shows that each species has a centromeric C-band on each chromosome. For interstitial C-band, it is quite different among the species. There is only an interstitial C-band in lOq in R. andersonii. And there is an interstitial C-band in 2p (stained weakly), 3q and 4q (only one homologous) in R. margaratae. But there are much more interstitial C-bands in R. kuangwuensis than in the other two species. R. margaratae and R. andersonii have not any telomeric C-band. But R. kuangwuensis has some telomeric C- bands. December 1993 Asiatic Herpetological Research Vol. 3, p. 47 Knuunxu" lIAXtlAXfti |^ KK]J ]i ,! iiwii it u •« ii |5 Hi M'* II » *i I FIG. 3. a: Karyotype of Rana margaratae. b: showing C-bands. c: showing Ag-NORs. Specific staining of the NORs with silver (Ag) confirms that the regions of NORs in all the three species are the same, in the long arms of chromosome No. 10. But the relation between the NORs and the con- strictions is quite different. The regions of the one pair of NORs in R. andersonii cor- respond to the regions of the secondary constritions. R. margaratae has only one NORs in the chromosome where the sec- ondary constriction locates, and no NORs is observed in the other homologous which has no secondary constriction. While for R. kuangwuensis, no secondary constric- tion is observed in the regions of the NORs. Discussion Comparing the karyotypes between the three species, we detect some interspecific variations. The secondary constriction is not detected in R. kuangwuensis. The chromosomes consist of 8 metacentric and 5 submetacentric pairs in R. kuangwuensis and R. margaratae, while 9 metacentric and 4 submetacentric in R. andersonii. C-banding patterns of all 13 pairs of R. margaratae are similar to those of R. ander- sonii, except for the variant bands on chromosomes 2, 3, 4, and 10. While C- banding pattern of R. kuangwuensis is quite different from those of the other two species, for R. kuangwuensis has more in- terstitial C-bands and telomeric C-bands as well. And the relation between the NORs and the secondary constrictions is quite dif- ferent among the three species. This study also indicates that intraspecific chromosome variations exist in R. margaratae from different distributional areas. We observe 8 metacentric and 5 submetacentric pairs in the present study, as opposed to 12 metacentric and 1 submetacentric pairs from Emei Mountain, Sichuan (Wang et al., 1983) and 9 metacentric and 4 submetacentric pairs also from Emei Mountain (Wu 1990). The C-banding pattern of chromosomes 1-13 of R. margaratae from Zunyi is com- pared with those from Emei Mountain. There is a centromeric C-band in each chromosome and a terminal C-band at each chromosome terminus, and an interstitial C- band in the long arm of No. 3 from Emei Vol. 5 p. 48 Asiatic Herpetological Research December 1993 TABLE 1 . Karyotypic data for Rana margaratae, R. andersonii, and R. kuangwuensis. Chromosome Nos. R. margaratae R. andersonii R. kuanguensis 1 R. L. 14.64±0.83 A. R. 1.26+0.31 R. L. 14.3010.91 A. R. 1.2610.33 R. L. 15.7010.91 A. R. 1.4610.26 2 R. L. 12.38±0.84 A. R. 1.48+0.38 R. L. 11.9010.70 A. R. 1.4310.35 R. L. 13.1110.32 A. R. 1.5410.41 3 R. L. 11.73±0.86 A. R. 1.72±0.49 R. L. 11.2010.68 A. R. 1.5710.28 R. L. 11.9310.94 A. R. 2.0110.39 4 R. L. 11.39+0.77 A. R. 1.76+0.39 R. L. 10.2110.49 A. R. 1.7110.41 R. L. 11.4210.92 A. R. 1.2910.39 5 R. L. 10.2010.84 A. R. 1.39+0.45 R. L. 9.5810.34 A. R. 1.3310.20 R. L. 10.2510.84 A. R. 1.4110.37 6 R. L. 6.52±0.45 A. R. 1.21±0.35 R. L. 6.8210.48 A. R. 1.3210.29 R. L. 6.5910.65 A. R. 1.2010.16 7 R. L. 5.6510.37 A. R. 2.5310.37 R. L. 6.0910.61 A. R. 2.1410.30 R. L. 5.9210.41 A. R. 2.5410.39 8 R. L. 5.4410.48 A. R. 1.3910.38 R. L. 6.0110.60 A. R. 1.3810.46 R. L. 5.4010.35 A. R. 1.2410.37 9 R. L. 5.2710.42 A. R. 1.9210.34 R. L. 5.7310.57 A. R. 2.0610.39 R. L. 5.1310.27 A. R. 2.0110.36 10 R. L. 4.7310.43 A. R. 1.3510.26 R. L. 5.3010.53 A. R. 1.2710.25 R. L. 4.8710.32 A. R. 1.5510.43 11 R. L. 4.3910.38 A. R. 1.3210.36 R. L. 4.9210.32 A. R. 1.5710.42 R. L. 4.3410.36 A. R. 1.9210.37 12 R. L. 4.2810.56 A. R. 1.2710.29 R. L. 4.5610.41 A. R. 1.5310.37 R. L. 4.2110.32 A. R. 1.4510.37 13 R. L. 3.9810.34 A. R. 1.4710.39 R. L. 4.1210.39 A. R. 1.6610.33 R. L. 3.5510.40 A. R. 1.6610.35 Mountain (Wang et al , 1983). Besides those above, there is an interstitial C-band in the aero long arm of No.7, and even het- erogeneity observed in No.9. There is an interstitial C-band in the middle of the long arms of both homologues of No.9 in fe- male, while only one homologous of No.9 is observed having an interstitial C-band, the other homologue has not an interstitial C-band in the middle of the long arm, but has an interstitial C-band near the terminus of the long arm (Wu 1990). We also ob- served indeed C-band heterogeneity of chromosome No.9 from Emei Mountain. Yet in our present study, we do not detect C-band heterogeneity of chromosome No.9 from Zunyi. And we detect other interstitial C-band (2p, 4q) but no telomeric C-band has been observed. In the early stage of the karyotypic evolution, a karyotype had generally more metacentric chromosomes. With the development, the karyotype differenciated in the direction of having more submetacentric or telocentric chromosomes (Li, 1985). Generally speaking, karyotypes with more telomeric and less interstitial C-band are more original. Between the two distributional areas of R. margaratae, the specimens from Emei Mountain has 12 (Wang 1983) or 10 (Wu, 1990) metacentric, and has more telomeric and less interstitial C-bands, and that from Zunyi has 8 metacentric and has less telomeric and more interstitial C-bands. On the view of point above, R. margaratae from Emei Mountain is more original than that from Guizhou. The stink frog group is composed of 9 species. The distributions of them are as follows: R. andersonii: upper Burma to Yunnan, Guizhou, Guangxi, Hainan December 1993 Asiatic Herpetological Research Vol. 5, p. 49 R. anlungensis: Guizhou (Anlung County) R. grahami: Sichuan, Guizhou, Yunnan R. kuangwuensis: Sichuan (Nanjiang County) R. lungshengensis: Guizhou, Guangxi, Hunan R. margaratae: Gansu, Sichuan, Guizhou R. schmackeri: Henan, Gansu, Sichuan, Guizhou, Hubei, Anhui, Jiangsu, Zhejiang, Jiangxi,Hubei, Guangdong R. tiannanensis: Yunnan, Hainan R. wuchuanensis: Guizhou (Wuchuan, Libo) From the discription above, it could be found that the distributional areas of some species are very limited, only one or two counties. So they are very rare and pre- cious wildlife. And it could also be found that the stink frog group is distributed mainly in the south of China, and most of them (7 species) are found on the Guizhou Plateau. So, the Guizhou Plateau might be the distributional center of the group. There were another two species of stink frogs. Their karyotypes and C-banding patterns were published. They are R. gra- hami and R. schmackeri. Both the species have 10 metacentric and 3 submetacentric pairs in their karyotypes, and both species have one telomeric C-band, but the former has 5 and the latter has 4 interstitial C- bands. Among the 5 species published their karyotypes and C-banding patterns, R. margaratae from Emei Mountain has most metacentric pairs and most telomeric C- bands. Although it does not have less in- terstitial C-bands, it could still be consid- ered as the most original in the viewpoint of cytogenetics. Considering that R. mar- garatae from Guizhou is more evoluted than that from Emei Mountain, it might be as- sumed that the stink frog group originated in Emei Mountain and its adjacent plateau. Acknowledgments We are most grateful to Ms. Wei Wu for providing us with a pre-publication manuscript and Ms. Xiaomong Zheng for help in collecting specimens and experimental work. This research was supported by the Guizhou Provincial Committee of Science and Technology. Literature Cited FROST, D. R. 1985. Amphibian species of the world. Pp. 480-520. The Association of Sys- tematics, Lawrence, Kansas. IKEBE, C, A. KEN-ICHI and K. SEI-ICHI. 1987. Karyotype analysis of two Japanese salamanders, Hynobius nebulosus (Schlegel) and Hynobius dunni Tago, by means of C-banding. Zoological Science 4:621-626. LI, GUOZHENG. 1985. Chromosome and its research method. Science Press, Beijing Pp.l- 261. (In Chinese). LI, S., Y. WANG, C, LI, R. WANG and G. LIU. 1982. An investigation for the karyotypic and C-banding pattern on the two anuran Amphibia. Acta Genetica Sinica (6):473-478. (In Chinese). SCHMID, M. 1978. Chromosome banding in Amphibia. Constitutive heterochromatin and nucleolus organizer regions in Ranidae, Microhylidae and Rhacophorodae. Chromosoma 68:131-148. WANG, Z., X, WANG, and W. CHEN. 1983. A comparative study on constitutive hete- rochromatin and nucleolus organizing regions (NORs) of three species of the genus Rana. Acta Herpetologica Sinica 2(4): 1-6. (In Chinese). WEI, G., F. CHEN, and N. XU. 1990. An investigation of the karyotypic C-banding and Ag-NORs pattern on Rana chensinensis. Hereditas (Beijing) 12(l):24-26. (In Chinese). WEI, G., N. XU, and D. LI. 1990. An investigation of the karyotype, C-banding, Ag- NORs pattern on Rana schmackeri. Cytogenetics of Amphibians and Reptiles, Advances in Life Sciences. 147-152. Birkhauser Vol. 5, p. 50 Asiatic Herpetological Research December 1993 Verlag (publisher) 270 pp. and F. H. Pough (Eds.), Biology of the Rep- tilia, Vol. 12, Physiological Ecology. Huey, R. B. 1982. Temperature, physiology, and Academic Press, New York, the ecology of reptiles. Pp. 25-91. In C. Gans I December 1993 Asiatic Herpetological Research Vol. 5, pp. 5 1-58 1 The Variegated Toad Agama in Djungar Gate (Eastern Kazakstan) with Notes on Certain Systematic Problems of Phrynocephalus versicolor Str. (Reptilia: Agamidae) M. L. GOLUBEV Institute of Zoology. Academy of Sciences, Kiev, Ukraine Abstract.-The distribution of Phrynocephalus versicolor in Djungar (=Junggar) Gate (Eastern Kazakstan) was investigated. The characteristics of these lizards are: absence of red axillary spots and the presence of red-orange subcaudal coloration. The taxonomic status of this population and of subspecies of P. versicolor is discussed. The variegated toad agama is presumed to be a "composed" species. A^y Words: Reptilia, Sauria, Agamidae, Phrynocephalus versicolor, Kazakstan, China, Djungar Gate, systematics, distribution. 75 sa-oo' B2"30' ^KA 1 . k o 1 tJii:^ £Bjm I \ ill '2 h ■ I ft n * ihh o IWH CHINA KAZAKHSTAN Djungar Gate ^J> ruih bfi^^3' U r u w q i FIG. 1. The distribution of Phrynocephalus versicolor hispida (I), P. gutlatus salenskyi (II), P. g. alpherakii (III), and P. v. doriai (IV) in Eastern Kazakhstan and Chinese Djungaria. Introduction There are few reliable literature citations concerning the distribution of the variegated toad agama, P. versicolor Str. from the eastern part of the Balkhash-Alkol Depression. Only Paraskiv (1956) reported finding the toad agama "near Lake Zhalanashkol". He also indicated that the distribution of this lizard is a natural continuation of the the Djungar Gate in the Alakol Depression, "somewhere along the northern Lake Alakol shore" (Paraskiv, 1956). According to Kubykin (1975), one specimen of this lizard was captured by him on Sredniy Island (Lake Alakol). However, this specimen is not mentioned in the collection list of the Institute of Zoology of the Kazakh Academy of Sciences (Brushko and Kubykin, 1988). Semenov (1986) andSemenov et al. (1987) refered to toad agamas collected "from Alakol Hollow" in the Zoological Museum of Moscow State University. All other references to the distribution of this lizard cite the above references. © 1993 by Asiatic Herpetological Research Vol. 5, p. 52 Asiatic Herpetological Research December 1993 FIG. 2. Habitat of Phrynocephalus versicolor in Djungar Gale near Druzhba Railway Station: crushed- stony and gravel semidesert covered with boyalych (Sal sola orbuscula). Methods We studied the distribution of the variegated toad agama in June 1991 during investigations of the Djungar Gate territory (Fig. 1). Results and Discussion The Djungar Gate is a relatively narrow pass between the Balkhash-Alakol Depression and Chinese Djungaria. This pass is oriented from the northwest to the southeast and rising in elevation towards the southeast. The pass enters China near the Druzhba ("Friendship") railway station along the Lankol Valley. The Djungar Gate Valley is formed by the broad alluvial plain and gentle foothills of Maily Ridge and the Djavlau Mountains to the northeast and the more abrupt upthrust of the Djungar Alatau to the southwest. The surface of the valley alluvial plains varied with the degree of slope and ranged from boulders, rubble. gravels and fine gravels on a loess base with the finer sorted materials deposited farther from the mountains slopes. Southeast of Lake Zhalanashkol the valley floor is a broad alkaline plain (20-25 km.) with subsurface water. The dominate plant, Salsola orbuscula, (common Russian name = boyalych, also known in the United States as Russian thistle) is found on the lower slopes and alluvial plains of the valley. It is more widely distributed on the northeastern slopes but it is sometimes replaced by saxaul (Haloxylon sp). Wormwood (Artemisia sp.) is dominate among the grasses and nearly the only plant on flats without shrubs. From the Lankol Valley the toad agamas are distributed along both the northeastern and southwestern slopes above the Djungar Gate. Along the southwestern alluvial plain the toad agama is distributed 15-20 km December 1993 Asiatic Herpetological Research Vol. 5, p. 53 from the Druzhba station to the northwest. The lizard occurs along the foothills of the Maily Ridge and the Djavlau Mountains for 55-60 km where the northern limit of its distribution coincides with the border of the boyalych dominate, gently sloping alluvial plain composed of rock rubble and gravels. Further north, on the steeper slopes of the alluvial plain composed of larger rock rubble and dominated by wormwood, the sunwatcher (P. helioscopus) is found. It is possible that P. versicolor occurs much further to the northeast into the Alakol Depression along the foothills, however field work in this area is difficult because of the presence of military installations along the border. A small isolated area with conditions which would make good habitat for this species is found along the railway tracks between the Zhalanashkol Station and the 19th Station near the mouth of the Irgaty River. Toad agamas are found under single bushes in small groups composed of 1-2 males and 2-5 females of different ages. In addition, groups of up to 10 subadults were observed. The density of the lizards is variable, with higher densities in gravelly areas with boyalych (Fig. 2) as well as in areas of colonies of the great gerbil (Rhombomys opimus ) that have excavated through the darker colored gravels and rock rubble and where the lighter loes makes up the predominant coloration of the surface. It is interesting to note that the lizards inhabiting the gravel plains in the Djungar Gate have retained the sand burrowing behavior, involving rapid lateral movements of the body, observed in populations inhabiting sandy areas. Pregnant P. versicolor as well as females of other Phrynocephalus species assume the "copulation avoidance" posture when pursued by males (Polynova, 1982; 1989; Rogovin, 1991; and our observations of P. strauchi in the Fergan Valley). To assume the "copulation avoidance" posture the female turns onto her back as the male approaches and maintains this position while he is nearby. Currently P. versicolor is considered to be a polymorphic species and it is interesting to determine the subspecific position of the form inhabiting the Djungar Gate. For a long time it was assumed that in eastern Kazakhstan this toad agama was found in three isolated populations: the Zaissan Depression, Alakol Depression and the Hi River Depression (Paraskiv, 1956; Bannikov et al., 1977). Peters (1984) considered the Zaissan Depression and Hi River Depression lizards to be two seperate species: P. salenskyi Bedr. and P. alpheraki Bedr. Three years later a new subspecies, P. versicolor paraskiwi (Semonov et al., 1987), was described from the Hi River Depression. These authors speculated that the two Chinese Djungar Depression subspecies, P. v. hispida Bedr. and P. v. doriai Bedr. were conspecifics. However, becuase of a shortage of material, they did not determine the taxonomic status of the Alakol Depression variegated toad agama. Soon after the most significant attempt to analyze the intraspecific variaton of P. versicolor was undertaken (Semenov and Shenbrot, 1989). Semenov and Shenbrot (1989) examined 675 specimens: 580 from Mongolia and Tuva, 65 from the Hi River Depression, 19 from Chinese Kuldja (now Yining, Xinjiang, China), 11 from the Alakol Depression, but no specimens from the remainder of the range of this species in China. The authors, using discriminant analysis techniques, felt their material was adequate to discuss all known subspecies of the variegated toad agama. Semenov and Shenbrot (1989) indicated that P. v. paraskiwi was detached from the main group. Also, P. v. doriai from Kuldja and P. v. kulagini from western Mongolia were resurrected. The Alakol variegated toad agama was singularly attributed to P. v. doriai. These authors were unable to distinguish the lizards from Mongolian Djungaria from "typical" P. v. hispida, however they did not indicate which P. v. hispida they considered Vol. 5, p. 54 Asiatic Herpetological Research December 1993 TABLE 1 . Differences between Djungar subspecies of PhrynocepMlus versicolor (after Bedriaga, 1909). Characters P. v. hispida P. v. doriai Bodv lenuth (I,.) 122 mm 133 mm Tail length (L. cd.) Medium (male), short (female) Long Supraocular scales Slightly smaller than surrounding Distinctly smaller Head scales Laree Small No. of scales across top of head 21-26(23-24) 25-29 (rarely 23) No. of scales along top of head 10-13 12-15 Thieh scales Smooth Smooth or keeled Dorsal coloration Gray, olive, light brown, gray- brown Dark gray, grey-brown, dirty red Dorsal bands Can be distinguished on shoulders and hind part of dorsum Distinct to absent Ventral coloration White, throat and chest slighty pigmented Commonly dark Axillary spots Tracks of yellowish or pink- yellow spots may be present 9 typical. They rejected their original view on the close relationship of the Djungar forms. Phrynocephalus v. hispida was recognized as identical to the nominative form. Discriminat analysis also has shown that the presence of red axillary spots are useful characteristics for seperating the closely related pairs of subspecies, P. v. versicolor - P. v. kulagini and P. v. paraskiwi - P. v. doriai. The remaining characteristics were found not to be useful for this purpose. This was already noted in earlier research (Nikolsky, 1915; Leroy, 1940; Terentjev and Chernov, 1949). However, a number of questions remain unanswered. Has it been demonstrated that axillary spots are absent in P. v. doriai and in the Alakol toad agama? Is it apropriate to include in the nominative subspecies, charaterised by the presence of axillary spots, the form P. v. hispida in which the axillary spots may be present or absent (Bedriaga, 1909)? If so, then what are the reasons for seperating into distinct taxa P. v. kulagini , which lacks axillary spots and P. v. paraskiwi which has the axillary spots? The latter form should be excluded from further discussion since it has been shown (Golubev, 1989) that it was erroneously described and should be attributed to P. gutatus alpherakii. Discriminant analysis did not clarify the relationship between P. v. doriai and P. v. hispida. It is clear (Table 1) that such characteristics as the relative size of the head and supraocular scales can not be used unless they are standardized. Dorsal and ventral coloration vary widely and are effected in life by such physiological considerations as body temperature and ambient light and in preserved specimens by the manner of preservation. Body length to tail length, when expressed in ratios (Bedriaga, 1909) and repeated measurements of type specimens from the Zoological Institute of the Russian Academy of Sciences, St. Petersberg (ZIN) did not confirm the differences noted by Bedriaga (P. v. hispida ZIN 6637 females: 0.78-0.81; males 0.66-0.68; P. v. doriai ZIN 5549, 8160 females 0.71-0.74; males 0.64-0.72). The ratios could be confirmed by using liner dimension L. and L. cd. but this was not done. Two other characteristics, number of scales along and across the top of the head are known to vary widely among populations. Only presence or absence of axillary spots remains as a useful character for seperating subspecies. However, we have no information concerning this character in P. v. doriai. Bedriaga (1909) used material that had been in preservative for more than December 1993 Asiatic Herpetological Research Vol. 5, p. 55 FIG. 3. Dorsal view of Phrynocephalus versicolor from Djungar Gate (Djungar Railway Station). 10 years and these spots might have disappeared during this time. Also there are no detailed data on the distribution of this character in lizards from eastern Djungaria. It is known only that such spots are present in the Mongolian part of the range of the toad agama (Semenov and Shenbrot, 1989). However, the nearly isolated Mongolian Djungaria (Barun- Khuray Depression) differs from Chinese Djungaria in several geogrphic parameters such as altitude. The type specimens of P. v. hispida are dated 1879 (ZIN 6637 and 6638) and 1880 (ZIN 6639). Nikolay M. Przewalsky's Third Central Asian Expedition (First Tibetan Expedition) took place at this time (Dubrovin, 1980). Przewalsky left the city of Zaissan on March 21 (April 4 by the modern calendar) and reached Ulungur Lake (Ulungur Hu) on March 31 (April 12 by the modern calendar). He followed the Urungu River (Ulungur He) and its tributary the Bulugun to the Barun-Khuray Depression and crossed it from north to south. He then crossed the Baytik Shan Ridge and the plains of south eastern Djungaria. On May 18 (May 30 by the modern calendar) he reached Barkul and did not return to Djungaria during this expedition (Przewalsky, 1883). In Przewalsky's journal the toad agamas {Phrynocephalus sp.) are casually mentioned for the middle and lower reaches of the river. Thus, it is possible to draw two conclusions: (1) the P. v. hispida type specimens may have been collected in different parts of Djungaria and thus include different forms, (2) the date of collection for the ZIN 6639 sample is incorrect. There is an overlap between both subspecies of toad agamas from the Djungar Gate region in body proportions (females 0.6-1.06; males 0.64-0.75), number of scales across (19-29; not counting the supraocular scales) and along (8-14) the top of the head. The scales on the thigh are smooth and dorsal and ventral coloration are highly variable (Figs. 3, 4, 5). In the material we collected, a slight shift in these character's values toward Vol. 5, p. 56 Asiatic Herpetological Research December 1993 FIG. 4. Ventral view of Phrynocephalus versicolor from Djungar Gale (Djungar Railway Station). hispida can be noted. It is important to note that the subcaudal surface in living lizards of both sexes is a red-orange color, that with age looses its lustre and disappears (Fig. 4). We also discovered remnants of the red-orange coloration in one of 13 specimens from "Alakol Depression" (Zoological Museum of Moscow State University, MSU R7779). Recently we examined specimens from China with the same subcaudal coloration in the California Academy of Sciences collected in the central (northeast of the city of Karamay) and southeastern (east from the city of Urumqi) Djungaria. In all specimens examined there are dark transverse bars on the ventral tail surface and this agrees with Bedriaga (1909) but not with Semenov (1986). From the above data, it follows that variegated toad agamas inhabiting the Alakol Depression and Djungar Gate to southern Djungaria (>500 km) differ from all other forms of P. versicolor in the absence of axillary spots and brightly colored subcaudal surface. Does this indicate the existence of a new subspecies? There are also red tailed toad agamas in northern Djungaria and the Zaissan Depression variously described as P. guttatus, P. salenskyi and P. versicolor. In the Zaissan Depression this lizard is mostly sand-dwelling and similar in habits to P. guttatus and P. frontalis (Golubev, 1989), while in northern Djungaria and in some places in the Zaisson Depression they are found in more stabilized soils. From the Alakol Depression and the Djungar Gate, where this species is found, there are over 300 km of continuous habitat without noticeable barriers into central Djungaria where P. versicolor is found. This may represent a cline with a gradual transition from one form to the other. In September 1991 repeated copulations between a male P. salenskyi (Zaissan Depression) and a female P. versicolor (Djungar Gate) were observed in the Kiev Zoo terrarium. If precopulation barriers exist, they apparently can be broken in terrarium conditions. Both Przewalsky (1883) and Potanin (1948), when traveling December 1993 Asiatic Herpetological Research Vol. 5, p. 57 along the lower reaches of the Urungu River and the southern shore of Ulungur Lake (type locality of P. salenskyi ), noted the variable coloration of toad agamas. We discovered fragments of a light longitudinal caudal stripe (a characteristic of P. g. salenskyi) in some specimens of P. versicolor from Djungar Gate (Fig. 5). Red axillary spots are present not only in the Alashan variegated toad agama (species type locality) but in lizards inhabiting the area south of Beishan Ridge in the Gashun Goby. However, here P. versicolor are connected by coloration and pattern transitions with P. axillaris Blanf. Thus, the question of the taxonomic status of the variegated toad agamas from Djungaria and Alakol Depression should again be considered open as does the question of the position of P. v. doriai. The Kuldja and western Djungarian populations are seperated by the Tianshan Mountains. There are reasons to believe that the Kuldja P. v. doriai is actually an ecological race of the Hi P. g. alpherakii while the Ebinurian P. v. doriai is assignable to the acutirostris group (which also might be no more than one of the color variants of the axillaris- guttatus complex. In summary, it appears that only two subspecies of the variegated toad agama can be recognized. P. v. kulagini inhabits southern Tuva and northwestern Mongolia and forms a narrow zone of intergradation with the nominative subspecies P. v. versicolor. However, there are doubts that the axillary red spots constitute a characteristic which allows one to delineate populations specifically on the level of geographaic race, i.e. subspecies. It is possible that P. versicolor consists of isolated, genetically differentiated color morphs associated with stabilized soils. Taxonomic seperation of these variants should occur only after a detailed study of the entire group. Bedriaga (1909) recognized five subspecies and considered P. versicolor to be a species composed of many highly variable populations. Bedriaga expressed concern that his taxonomic arragement of these subspecies did not represent a natural assemblage. FIG. 5. A specimen of Phrynocephalus versicolor from Djungar Gate (Djungar Railway Stauon)with light longitudinal caudal stripe. Further he stated "[only when we are more familiar with the fauna of Central Asia, will it be clear whether I have exaggerated distinguishing characters]". Leroy (1940) proposed that the species P. versicolor be abolished. Leroy's point of view may be closest to the truth. Acknowledgments The author is most grateful for the loan Vol. 5, p. 58 Asiatic Herpetological Research December 1993 of Phrynocephalus collections from M. E. A. Dunajev and Dr. V. F. Orlova of the Zoological Museum of the Moscow State University, Russia; Mrs. L. Jogansen and Prof. I. S. Darevsky of the Zoological Institute of the Russian Academy of Sciences, St. Petersberg; and Mr. J. V. Vindum and Dr. A. E. Leviton of the California Academy of Sciences, San Francisco, California, U.S.A. Also, I would like to thank Mr. V. B. Zaykovsky, Mr. G. Makhnin, and Mr. G. I. Zveryansky of the Railway Antiplague Service, Alma-Ata, Kazakhstan for the friendly help during the 1991 field season. Literature Cited NIKOLSKY, A. M. 1915. [Fauna of Russia and adjacent countries. Reptiles. Vol. I. Chelonia and Sauria]. Petrograd, Imperial Academy of Sciences. 532 pp. (In Russian). PARASKIV, K. P. 1956. [Reptiles of Kazakhstan]. Kazakh Academy of Sciences Press, Alma-Ata. 228 pp. (In Russian). PETERS, G. 1984. Die Krotenkopfagamen Zentralasiens (Agamidae: Phrynocephalus). Mitteilungen aus dem Zoologischen Museum in Berlin. Akademie Verlag, Berlin 60(1): 23-67. POLYNOVA, G. V. 1982. [Demonstrative behavior of Phrynocephalus mystaceus]. Zoological Journal, Moscow 61(5):734-741. (In Russian). BANNIKOV, A. G., I. S. DAREVSKY, V. G. ISCHENKO, A. K. RUSTAMOV, AND N. N. SCHERBAK. 1977. [Field guide of the USSR amphibians and reptiles]. Moscow, Prosveschenje Publishing House, Moscow. 369 pp. (In Russian). BEDRIAGA, Y. 1909. Amphibian und Reptilien. Pp. 73-502. In Wissenschaftliche Resultate der Reisen N. M. Prezewalskijs durch Zentralasien. Zoologische Teil. Band 3. Part 1. Lacertilia. St. Petersburg. (In Russian/German). Brushko, Z. K. and R. A. Kubykin. 1988. [Catalogue of the herpetological collection of the Institute of Zoology of the Kazakh Academy of Sciences]. Kazakh Academy of Sciences Press, Alma-Ata. 42 pp. (In Russian). DUBROVIN, N. F. 1980. [Nikolay Mikhailovich Przewalsky. Biographical essay]. St. Petersburg, Military Printing House. 602 pp. (In Russian). GOLUBEV, M. L. 1989. [Phrynocephalus guttatus (Gmel.) or Ph. versicolor Str. (Reptilia, Agamidae): which Phrynocephalus species occurs in Kazakhstan?]. Zoological News, Kiev 5:38-46. (In Russian). KUBYKIN, R. A. 1975. [Ecological-faunistical survey of reptiles of the Lake Alakol islands]. Transaction of the Kazakh Academy of Sciences. Biological Series 3:10-16. (In Russian). LEROY, P. 1940. Les Phrynocephales de Mongolie et du N.-W. Chinois. Peking Natural M, i,,,, n,,n..i.., i/i/")\.iir» iah History Bulletin 14(2): 139-146. POLYNOVA, G. V. 1989. [The new data on the functional role of the posture "copulatory avoidance" in the genus Phrynocephalus]. Pp. 200-210. In Questions in herpetology. Abstracts of the Report at the Seventh All Union Conference of Herpetology, Naukova Dumka, Kiev. (In Russian). POTANIN, G. N. 1948. [Field trips to Mongolia]. Geographgiz Publishing, Moscow. 480 pp. (In Russian). PRZEWALSKY, N. M. 1883. [The trip from Zaissan diroughout Hami to Tibet and the upper reaches of the Yellow River]. Balashev's Press, St. Petersburg. 473 pp. (In Russian). ROGOVIN, D. V. 1991. [Social behavior of Phrynocephalus helioscopus and Ph. reticulatus (Reptilia, Agamidae) and their relationships in joint settlements]. Zoological Journal, Moscow 7(X3):61-72. (In Russian). SEMENOV, D. V. 1986. [Materialen zur Variabitat und innerartlichen systematik der Buntkrotenkopfagame (Phrynocephalus versicolor Sir.) in der Mongolei], Pp. 157-173. In Herpetologissche Untersuchungen in der Mongolischen volksrepublik. Moscow, Academy Press. (In Russian; German summary). SEMENOV, D. V., Z. K. BRUSHKO, R. A. KUBYKIN, G. I. SHENBROT. 1987. [Taxonomic position and protected status of the Round-headed Lizard (Reptilia, Agamidae) in the territory of the USSR]. Zoological Journal, Moscow 68(12):79-87. (In Russian) December 1993 Asiatic Herpetological Research Vol. 5, pp. 59-64 Allozyme Variation and Genetic Relationships within the Phrynocephalus guttatus Species Group (Sauria: Agamidae) in the Former USSR. Sergei mezhziierin and Michael l. Golubev Institute of Zoology, Academy of Sciences, Kiev, Ukraine Abstract. -An electrophoretic analysis of several populations of Phrynocephalus guttatus s. lato. shows that there are two groups with a remarkable level of genetic differentiation. There is an eastern Palearctic P. versicolor from southern Mongolia, and a western Palearctic P. guttatus s. str. which includes: g. guttatus, g. kushackewitschii g. alpherakii, g. salenskyi, g. moltschanovii, guttatus ssp. from northern Turkmenia and versicolor hispida from Djungar Gate. There are no objective criteria for subspecific separation by biochemical genetic markers. Key words: Reptilia, Sauria, Agamidae, Phrynocephalus, electrophoresis, systematics. 40 FIG. 1. Distribution of the Phrynocephalus guttatus species group in the former USSR. la- P. g. guttatus; lb- P. g. moltschanovi; II- P. g. kushackewitschii; III- P. g. alpherakii; IV- P. g. salenskyi; V- P. versicolor hispida; VI-P. guttatus ssp. The numbering of the populations is as given in Table 1. Introduction The agamid genus Phrynocephalus includes some polytypic species groups. One of the most complicated species complexes is Phrynocephalus guttatus s. lato. Representatives of this species group are widely distributed in Middle and Central Asia from the northern Caucasus to China. The systematics of this species group is highly controversial and needs revision. There are some alternative viewpoints on the status and systematic relationships of its representatives. The classical viewpoint of Terentjev and Chernov (1949) recognized only two species: P. versicolor and P. guttatus (P. g. guttatus and P. g. kushackewitschii). A new concept was developed during the last decade by Semenov and Shenbrot (Semenov and Shenbrot, 1982; Shenbrot and Semenov, 1987; Semenov et. al., 1987). According to this concept, the guttatus-group consists of four species: P. guttatus (Gmel), P. moltschanovi Nik., P. melanurus Eichw. (=P. salenskyi Bedr.) and P. versicolor Str. The last form includes the nominal subspecies (China: Alashan to Djungaria), P. v. kulagini (Tuva, Russia; western Mongolia) and the western Palearctic subspecies, P. v. paraskii Semenov, Brushko, Kubykin et Shenbrot. Golubev (1989) lowered the status of "salenskyi" to subspecific level, united P. v. paraskii with P. guttatus alpherakii Bedr. and included "moltschanovi" only as a color variation of P. g. guttatus. © 1993 by Asiatic Herpetological Research Vol. 3, p. 60 Asiatic Herpetological Research December 1993 TABLE 1 . Localities, sample sizes and taxa of Phrynocephalus guttatus S. lato. populations collected and investigated in this study. N Taxa Locality No. 1 P. g. guttatus NORTHERN TRANSCAUCASUS REGION Daghestan: Tersky Sands near Chervlenny Buruny Russia: Stavropol Dist., Tersky Sands: Roshchino Chencheno-Ingushety: near Starogladkovsky N. Daghestan: sands on the right bank of Kuma River 2 3 3 1 2 P. g. moltschanovi N. Kysylkum in Karakalpakia: Bellau Mount. N. Kysylkum in Karakalpakia: Kostruba Well 4 5 3 P. guttatus ssp. N.Turkmenia: Kazakhlyshor near Kumsebshen Well 3 4 P. g. salenskyi E. KAZAKHSTAN: Zaissan Depression: Left bank fo Black Irtysh near Karatal Irtysh Sands near Chingildy Left bank of Bukhtarma Reservior: Kuludjunsky Sands 22 18 18 5 P. g. kushackewitschii E. KAZAKHSTAN: Taldy-Kurghan Dist. (Alakol Basin): Near Andreevka (left bank of Chyndjaly River 23 6 P. g. kushackewitschii E. KAZAKHSTAN: Taldy Kurghan Dist. (Balkhash Basin): NW bank of Kapchagay Reservior SWof Balkhash: Ortadyressin 10 2 7 P. g. alpherakii (+P. v. paraskii E. KAZAKHSTAN: Taldy-Kurghan Dist. (right bank of Illi ); Panfilov Region: 25 km. from Aidarly Kerbulak Region: Ayakkalkan Alma-Ata Dist.: near Chundja (left bank of Illi) 2 7 2 8 P. versicolor hispida E.Kazakhstan: DjungarGate 7 9 P. v versicolor Mongolia: South Gobi Aimag: Dalanzadagad 5 10 P. strauchi Fergan Valley: left bank of Kajrakkum Reservior near Kyly 20 11 P. helioscopus saidalievi S. part of Fergan Valley near Kim 7 A high level of morphological variation on the one hand, and caryological conservatism on the other, doesn't allow one to decide problems of systematic status and specific identity of representatives of P. guttatus s. lato. Therefore, in order to decide controversial systematic problems of this agamid group, we used biochemical genetic markers. Methods Electrophoretic analysis was carried out on geographic forms of P. guttatus s. lato from different geographic regions (Fig. 1, Table 1). The geographic form from northern Turkmenia was excluded from P. g. guttatus on the basis of the red spots on the arm pits, a very rare characteristic in P. guttatus. We also studied two well differentiated species, P. strauchi Nik. and P. helioscopus saidalievi Sattorov, both from the Fergan Valley, as an external control for genetic differentiation. Each adult specimen was processed in the laboratory for blood and muscle samples and immediately studied by standard vertical acrylamide electrophoresis. Homogenates obtained from muscle, crushed in distilled water with 5 per cent sucrose, were processed for the following enzymes and proteins (Table 2). Isozymes were numbered in order of decreasing mobility from the most anodal one. Allozymes were designated numerically according to their mobility, relative to the most frequent allele (100), faster mobility (>100), slower mobility (<100). The genetic divergence between populations and divergence time were estimated with indices of standard genetic distances by formulas proposed by Nei (1975). December 1993 Asiatic Herpetological Research Vol. 3, p. 61 TABLE 2. Enzymes studied and electroph oretic conditions employed. Enzyme or protein Locus abrev. EC no. Tissue Buffer Aspartate aminotransferase s-Aat 2.6.1.1 Muscle TEB Glycerol-3-phospate dehydrogenase G-3-pdh 1.1.1.18 Muscle TEB Isocitrate dehydrogenase s-Idh 1.1.1.42 Muscle TEB Lactate dehydrogenase Ldh-A 1.1.1.27 Muscle TEB Lactate dehydrogenase Ldh-B 1.1.1.27 Muscle TEB Malate dehydrogenase s-Mdh 1.1.1.37 Muscle TEB Malic enzyme s-Me 1.1.1.40 Muscle TEB Superoxide dismutase s-Spd 1.15.1.1 Muscle TEB 6-phosphogluconate dehydrogenase 6-pgdh 1.1.1.44 Muscle TEB Phosphoglucomutase Pgm 2.7.5.1 Muscle TG Esterase Es-D 3.1.1.1 Muscle TEB Esterase Es-2 3.1.1.1 Muscle TG Esterase Es-3 3.1.1.1 Hemolizate TG Hemoglogin Hb - Hemolizate TG Albumin Alb - Muscle TG Structural muscle proteins Pt-1. 2. 3 - Muscle TG Note: TEB- Tris-EDTA-NA2 boric acid Ph 8.5 (Peacock et al., 1965). TG- Tris-glycin, disc- electrophoresis (Davis, 1964) Results Allozyme variation. — Four of 18 loci analyzed (Es-3, Pt-2, Pt-3, IdhS) were monomorphic and fixed for the same allele in all populations and species considered. Fourteen loci were polymorphic within or between population and their allelic frequencies are given in Table 3. Expected genotypes distributions were in equilibrium according to Hardy-Weinberg formula in the investigated populations at all loci observed. Exceptions were obtained only in "salenskyi" and "kushackewitschii" at the Ldh-B locus (Table 4). In our opinion, absence of heterozygote genotypes can be explained by introgression of Ldh-B 90 from "salenskyi" to "kushackewitschii" or vice versa. Levels of genetic variation are given in Table 3 (Mean proportion of heterozygosity observed (H obs.) and expected (H exp.)}. H obs. ranged from 0 in "salenskyi" to 0.05 in "kushackewitschii" with a mean of 0.04. This meaning of heterozygosity is near the level usual for Reptilia (Nevo, 1984). Genetic divergence. — Only two loci (Me-S, Mdh-S) display fixation or predominance of alternative alleles between P. versicolor (Mongolia) and the P. guttatus group. All representatives of Western palearctic P. guttatus have common gene pools of the loci considered. Phrynocephalus. v. hispida is an exception and has alternative allelic fixation at Es-2 (Table 3). Genetic distances among P. guttatus forms are rather low and range from 0 to 0.05 i.e. on interspecific level of differentiation. There are four loci which display alternative fixation between P. strauchi and P. guttatus. It shows a clear intraspecific level of genetic divergence. The largest genetic distance is between P. h. saidalievi and P. guttatus species group (D=0.832). Differences between these species include 8 loci which display alternative allelic fixation (Pgm, Mdh-S; Aat-S, Me-S, Alb; Pt. 1; Ldh-B, Es-2). From the allelic frequencies at 18 loci tested, we calculated Nei's genetic distance and constructed a matrix of genetic distances (Table. 5). A UPGMA phenogram was calculated on the basis of this matrix. This reflects the relationships Vol. 3. p. 62 Asiatic Herpetological Research December 1993 TABLE 3. Allelic frequencies. Locus Allele 1 2 3 4 5 6 7 8 9 10 11 s-Aal 90 95 100 105 1.00 1.00 1.00 0.98 0.02 0.12 0.88 1.00 1.00 1.00 1.00 0.80 0.20 0.97 0.03 G-3-pdh 95 100 1.00 0.25 0.75 Ldh-A 90 100 1.00 0.25 0.75 1.00 0.08 0.92 0.92 0.08 0.81 0.19 0.39 0.61 1.00 0.67 1.00 1.00 Ldh-B 90 100 1.00 1.00 1.00 1.00 1.00 0.97 0.03 1.00 1.00 1.00 1.00 1.00 s-Mdh -100 80 90 100 1.00 1.00 1.00 0.02 0.98 1.00 1.00 1.00 1.00 0.93 0.17 1.00 1.00 s-Me -100 90 91 98 100 102 105 108 1.00 1.00 1.00 0.20 0.53 0.14 0.11 0.83 0.13 0.04 1.00 1.00 1.00 0.50 0.40 0.10 1.00 1.00 Pgm 95 100 102 105 1.00 1.00 1.00 0.97 0.03 1.00 1.00 1.00 0.17 0.83 1.00 1.00 6-pgdh Es-2 80 85 88 90 98 100 103 108 110 112 1.00 1.00 1.00 0.03 0.01 0.96 0.04 0.96 0.08 0.04 1.00 0.88 1.00 0.08 0.32 0.60 0.05 0.87 0.08 94 0.17 96 0.12 0.25 0.50 0.04 0.05 0.55 0.57 98 1.00 0.88 0.75 0.50 0.92 0.95 0.35 0.50 0.37 100 0.04 0.33 0.07 103 105 0.05 0.21 110 0.05 0.70 115 0.09 0.03 0.94 0.03 s-Sod 100 110 1.00 1.00 1.00 1.00 1.00 0.91 0.09 1.00 1.00 1.00 1.00 1.00 Es-D 95 98 100 105 0.95 0.05 1.00 1.00 0.10 0.86 0.04 1.00 1.00 0.23 0.77 1.00 1.00 1.00 1.00 1.00 Hb 99 100 102 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Alb 98 100 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 Pl-1 a b 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 H obs. % 0.6 3.0 2.2 6.3 4.9 3.4 6.9 4.0 6.6 5.6 1.0 H exp, % 0.6 3.2 2.1 8.0 5.0 4.3 7.8 3.8 13.7 5.8 1.0 Under the electrophoretic condition used, the following loci are monomorphic: s-Idh, Es-3, Pt-2, Pt-3. See table 1 for population numbers. between the P. guttatus s. lato representatives and the other two species that we studied (Fig. 2). Discussion Two distinctive gene pools. differentiated from one another only by two diagnostic loci, were found between representatives of P. guttatus s. lato and P. versicolor (southern Mongolia). The genetic differentiation corresponds to the division of eastern Palearctic (P. versicolor from southern Mongolia) and Western December 1993 Asiatic Herpetological Research Vol. 3, p. 63 TABLE 4. Distribution of genotypes at the Ldh-A locus in populations of different geographiac forms of toad agamas of the Phrynocephalus guttatus group. Form Locality Genotypes distribution salensky Zaissan 0 E 42 36.8 0 6 10.5 0.75 x2 =13.44** d. f. =1 kushackevitschi Andreevka 0 E 2 0.31 0 10 3.2 8.5 x2 =4.05* d. f. =1 kushackevitschi Kapchagay O E 1 0.6 moltschanovi Beltau O E 7 6.75 Note: O- observed distribution; E- expected distribution; *p < 0.05; **p < 0.001 TABLE 5. Matrix of genetic distances (D, Nei, 1975) among the taxa P. guttatus s. lato, P. strauchi, and P. helioscopus saidalievi. 1 2 3 4 5 6 7 8 9 10 11 1 X 0.004 0.003 0.043 0.053 0.038 0.033 0.046 0.151 0.257 0.809 2 x 0.002 0.041 0.028 0.012 0.016 0.042 0.128 0.281 0.820 3 x 0.032 0.055 0.041 0.019 0.035 0.146 0.269 0.795 4 X 0.078 0.078 0.040 0.049 0.132 0.234 0.878 5 X 0.003 0.043 0.099 0.117 0.344 0.920 6 X 0.032 0.084 0.120 0.326 0.896 7 X 0.043 0.133 0.280 0.845 8 X 0.165 0.303 0.784 9 X 0.388 0.749 10 x 0.659 11 X Palearctic forms (P. guttatus) which diverged around 500,000 years ago (Late Pleistocene). P. guttatus consists of conspecific forms, diverse morphologically, but conservative on the molecular level. In this species, the more differentiated form is P. v. hispida. This is supported by the fixation of Es-2 (94) which is absolutely absent in P. v. versicolor. The level of genetic differentiation of P. guttatus s. lato from P. strauchi and P. helioscopus saidalievi is higher and corresponds to good species which diverged about 1,500,000-2,000,000 years ago, i.e. in Middle or Early Pleistocene. On the basis of the data we obtained, our main conclusion is that P. guttatus s. lato. consists of two groups with a remarkable level of genetic differentiation. There is an eastern Palearctic P. versicolor from southern Mongolia, and a western Palearctic P. guttatus s. str.* which includes: g. guttatus, g. kushackewitschii g. alpherakii, g. salenskyi, g. moltschanovii, guttatus ssp. from northern Turkmenia and versicolor hispida from Djungar Gate. There are no objective criteria for subspecific separation by biochemical genetic markers. *This abbreviation, which as generally known means "sensu stricto", was erroneously deciphered as "s. Strauch" (Mezhzherin and Golubev, 1992). This somewhat distorted the intended meaning. Literature Cited DAVIS, B. J. 1964. Disc-electrophoresis. 2. Method and application to human serum Vol. 3. p. 64 Asiatic Herpetological Research December 1993 -/h -//- ■ P. g guttatus *• P. g. moltschanovi V-P.g.spp. 1 — P. v. paraskivi P. g. salenskyi P. v. hispida P. g. kushackewitschii - I P. g. kushackewitschii - II P. v. versicolor P. strauchi P. helioscopus saidalievi -//- 0.8 0.7 0.6 0.3 0.2 D ( Nei, 1975) 0.1 0 FIG. 2. UPGMA Phenogram of relationships among Phrynocephalus guttatus s. lato. proteins. Annals of the New York Academy of Sciences 121:404-408. GOLUBEV, M. L. 1989. [Phrynocephalus guttatus (Gmel. ) or Ph. versicolor Str. ( Reptilia, Agamidae): which Phrynocephalus species occurs in Kazakhstan?]. Zoological News, Kiev (5):38-46. (In Russian). MEZHZHERIN, S. V. AND M. L. GOLUBEV. 1992. Allozyme variation and genetic relationships among Phrynocephalus guttatus species group (Agamidae) of the former USSR fauna. 1st Asian Herpetological Meeting, 15-20 July, 1992, Huangshan, China. Abstract Book:52. NEVO E., A. BEILES AND R. BEN-SHLOMO. 1984. The evolutionary significance of genetic diversity: ecological, demographic and life history correlates. Lecture Notes in Biomathematics 53:13-213. PEACOCK, A. C, S. L. BUNTING AND K. G. QUINN. 1965. Serum protein electrophoresis in acrylamide gel: patterns from normal human subjects. Science 147:1451-1455. SEMENOV, D. V. AND G. I. SHENBROT. 1982. [On species independence of Phrynocephalus moltschanovi (Reptilia, Agamidae)]. Zoological Journal, Moscow 61(8):1 194-1204. (In Russian). SEMENOV, D. V. , Z. K. BRUSHKO, R. A. KUBYKIN AND G. I. SHENBROT. 1987. [Taxonomic position and protective status of the Round-headed Lizard (Reptilia, Agamidae) in the territory of the USSR]. Zoological Journal, Moscow 68(1 2):79-87. (In Russian) . SHENBROT, G. I. AND D. V. SEMENOV. 1987. [Present distribution and taxonomy of Phrynocephalus guttatus (Reptilia, Agamidae)]. Zoological Journal, Moscow 66(2):259-272. (In Russian). TERENTYEV, P. V. AND S. A. CHERNOV. 1949. [Guide to reptiles and amphibians], Soviet Sciences Publishing, Moscow. 315 pp. (In Russian). December 1993 Asiatic Herpetological Research Vol. 5 pp. 65-73 1 Geographic Variation and Diversity in Three Species of Phrynocephalus the Tengger Desert, Western China YlTEZHAO WANG1 AND HUIZHAO WANG2 'Chengdu Institute of Biology, Accidentia Sinica, Chengdu, Sichuan, China ^ Chengdu Library. Academia Sinica, Chengdu, Sichuan, China Abstract. -Univariate and multivariate statistical analysis of 5 merisuc characters, 4 metric characters and 8 ratio characters recorded for 9 samples of lizards were used to assess non-geographic variation and geographic variation in the southeast area of the Tengger Desert. A cluster analysis indicated that there were three major groups which represented Phrynocephalus versicolor, P. pnewalskii, and P. frontalis, based on morphological characters, respectively. The cluster analysis and canonical analysis showed that among samples, phenetic similarity was not always predicted by geographic proximity. Dispersal and divergence of these species of Phrynocephalus with the relationships of paleogeography and paleoclimatology are discussed. It is evident that the Yellow River fails to cause geographic isolation. Key words: Reptilia, Sauria, Agamidae, Phrynocephalus, China, biogeography. Introduction Strauch (1876) described four new species of Phrynocephalus from the Altan Desert (including the Tengger Desert) and the Mu Us Desert (Ordos). They were P. pnewalskii, P. versicolor, P. affinis, and P. frontalis. He pointed out that P. affinis was very similar to P. pnewalskii. Pope (1935), with out any discussion, considered P. affinis to be synonymous with P. pnewalskii. Leroy (1939), who studied the geographic variation and distribution of P. pnewalskii, P. frontalis, and P. vlangalii, failed to recognize P. frontalis and P. versicolor, which were quite different on morphological characters. Therefore, his distribution of P. frontalis included P. versicolor . Zhao (1979) deduced that the distribution of Phrynocephalus in China almost reached the shore of the Bohai Sea and in the Altan Desert (including the Tengger Desert) no distribution for P. frontalis was mentioned. In this paper we discuss (1) the geographic variations and distributions of P. pnewalskii, P. frontalis, and P. versicolor, and (2) the dispersal and divergence tracts of the three species in the Tengger Desert. Materials and Methods Study Areas. — The Tengger Desert is situated in western Nei Mongol Autonomous Region, China. The east edge reaches the Helan Mountains, and the southern edge, the Yellow River. In the west and north, the study area connects with the Badain Jaran Desert and the Ulan Behou Desert, respectively. The climate in these areas is arid-continental. The topography is moving sand dunes and gobi. In this area the vegetation is scarce. Ammopiptanthus sp., Potaninia sp., Teraena sp., and Caragana sp. are present. Populations of P. pnewalskii, P. frontalis, and P. versicolor were sampled from nine localities in the southeast area of the Tengger Desert from the south and north shore of the Yellow River, through southeast Tengger Desert to Hala Woo Valley of the Helan Mountains, and in the west at Xial Hong Shan (Fig. 1 and Table 1). Methods. — Five meristic characters, four metric characters and eight ratio characters were chosen for study. These characters are shown in Table 2. Statistical analysis. — For analysis of geographic variation, lizards were grouped into nine samples (Fig. 1 and Table 1) within which gene flow was assumed to occur freely. Sexual variation in metric characters was examined in sample 1 (representing P. pnewalskii), and samples 2 and 3 (representing P. frontalis). We © 1993 hy Asiatic Herpetological Research Vol. 3, p. 66 Asiatic Herpetological Research December 1993 TABLE 1 . Locality data for Phrynocephalus from the Tengger Desert, western China used in this study. Population Locality Males Females 1 Shapotou North shore of the Yellow River (37°30'N 104°58'E) 26 14 2 Shapoltou South shore of the Yellow River 31 12 3 7.2 km SW of Alxa Zuogi (38°50'N 105°32'E) 21 1') 4 2.4 km north of Mujing/.i (3X°20'N 105°42'E) 9 15 5 1.9 km SW of AlxaZouqi (38°50'N 105°32'E) 10 9 6 1.7 km SWof Luanjing (37°58'N 105°32'E) 6 4 7 47.3 km SW of Luanjing 5 2 8 9.3 km east of Xiao Hong Shan (37°31'N 104°27'E) 5 1 9 Mala Woo Vallcv 1 Man Mountains (3X Sl'N 1(>SC,34T-;) 5 3 100 km l TENGGER DESERT FIG. 1. Locations in the Tengger Desert for the nine samples of Phrynocephalus used in the study of geographic variation (see Table 1 for exact localities). determined that individuals of the three species are reproductively mature at an SVL of 45 mm or larger. The ANOVA test was used to test for differences between the sexes in samples 1, 2, and 3. Standard univariate statistics were calculated for these three samples and to examine geographic variation in single meristic characters for each sample. For all December 1993 Asiatic Herpetological Research Vol. 3, p. 67 TABLE 2. Characters used in analysis of variation in three species of Phrynocephalus. A. Meristic Characters TABLE 3. Sexual variation in three samples of Phrynocephalus (sample number as in Fig. 1 and Table 1. Levels of significance are indicated. (*=p<0.05: **=p<0.01) 1. NSPL No. of supralabials 2. NIFL No. of infralabials 3. NSDT No. of subdigital lamellae on longest toe 4. NSNT Number of scales between nostrils 5. NDVT No. of ventral dark bands on tail B. Metric Characters 6. SVL Snout-vent length 7. TL Tail Length 8. LFL Foreleg length (including fingers) 9. LHL Hindleg length (including longest toe) C. Ratio characters 10. SVL/TL 11. SVL/LFL 12. SVL/LHh 13. TL/LFL 14 TL/LHL 15 LFL/LHL 16 HL/HW* 17 LNE/LN** * Head length (rostral up to and including parietales)/Head width (front of ears). ** Length from nostril to eye/Length between nostrils. samples, meristic characters were calculated and standard values from the matrix of intersample phenotypic distances were clustered with the unweighted furthest- neighbor method using arithmetic average. To overcome some disadvantages of the clustering, the multivariate analysis of variance (MANOVA) and canonical analysis were used to provide weighted combinations of characters to analyze variations of the nine samples. A set of canonical was calculated and the mean values of each sample were plotted on the first two axes. Additionally, the relative contribution of each character in Table 4 to each of these two axes was calculated, which collectively accounted for over 86% of the total variation. All calculations were performed on an IBM-PC/XT computer at the Chengdu Library, Academia Sinica by means of FORTRAN Program II. Character n M (SD) n M (SD) ANOV A Males Females Population 1 SVL 26 61.51 (1.42) 14 58.65 (2.88) 4.11** TL 26 86.20 (3.50) 14 75.72 (3.54) 1.02 LFL 26 31.35 (1.10) 14 29.47 (1.17 1.13 LHL 26 51.10 (2.16) 14 47.17 (1.82) 1.41 Population 2 SVL 31 51.51 (1.45) 12 52.08 (1.64 1.28 TL 31 72.55 (2.67 12 65.39 (2.00) 1.78 LFL 31 26.09 (1.42) 12 25.36 (1.01) 1.98 LHL 31 43.05 (2.83) 12 40.16 (1.69 2.80** Population 3 SVL 21 58.07 (1.59) 19 54.97 (2.37) 2.22* TL 21 77.85 (3.23) 19 67.86 (3.59 1.24 LFL 21 29.51 (1.28) 19 27.25 (0.69) 3.44** LHL 21 47.68 (1.79 19 42.52 (0.73) 6.01** Results Sexual variation (non-geographic variation). — Results for intersex comparisons (adults only) in three samples are summarized in Table 3. In ANOVA the sexes from sample 1 differed significantly in SVL. In sample 2, the metric character LHL showed significant sexual dimorphism . In sample 3, three metric characters, SVL, LFL, and LHL showed significant sexual dimorphism. Multivariate analysis. — Fig. 3 presents a distance phenogram, based on 5 meristic characters, clustering 9 OTUs which correspond to the group sample localities (Fig. 1 and Table 1) used in this study. Vol. 3, p. 68 Asiatic Herpetological Research December 1993 Vec tor II 2.5 2 1.5 *9 1 3 - * 0.5 * V e c t o r 1 n 0.5 '"I — i '"■£■"1 8 : ,i n|iMi|i ni|M ii|i in | * 2 7 -1 4 6 * : * 5 1.5 -2 2.5 -5-4-3-2-1012345 FIG. 2. Projections on tlie first two canonical vectors of centroids representing the nine samples of Phrynocephalus. The number of centroid corresponds to the localities in Fig. 1. The cophenetic correlation of the phenogram with the distance matrix was 0.912. The first major dichotomy separates sample 9 from all others, and the second major dichotomy groups samples 1 and 7. The third major dichotomy groups samples 2, 3, 4, 5, 6, and 8. Most of the clusters group samples that do not have any geographic affinities or relationships. For example, samples 3 and 9 in fairly close geographic proximity (about 10 km separate each other) were separated by different dichotomies in the phenogram, conversely, samples 2, 8, and 3 clustered with one another, despite being separated by over 150 km. Univariate analysis (geographic- variation). — Standard statistics are presented for two meristic characters. Fig. 4 and Fig. 5 depicts geographic variation in NSPL and NIFL counts for nine samples (males only). The mean NSPL counts of samples 1 and 7 were 17.50 and 16.00, respectively. For samples 2, 3, 4, 5, 6, and 8, the mean NSPL counts were between 13.00 and 14.90, and for sample 9 was 15.00 (Fig. 4). Fig. 5 shows NIFL counts for all nine samples. Samples 1 and 7 show higher NIFL counts, 17.10 and 5- 4" 453826179 Sample FIG. 3. Distance phenogram resulting from the cluster analysis with five meristic characters of nine samples (each considered an OTU) of Phrynocephalus. The cophenetic correlation coefficient was 0.912. 16.00 respectively, than those of samples 2 (15.00), 3 (14.80), 4 (14.10), 5 (14.00), 6 (13.20), and 8 (14.00), as well as 9 (14.00). In the MANOVA, the overall tests of the hypothesis of no effect due to geography were rejected (P < 0.001) for Wilks' Criterion, and Roy's Maximum Root Criterion. The first two canonical vectors extracted from the variance-covariance matrix accounted for 77.58 % and 8.67 % of the total variation. The samples are plotted along these two vectors in Fig. 2 and Table 4 summarizes the percent influence of each character to each of the two vectors. Three major groups were discernible in Fig. 2, the most strongly differentiated of which separated primarily along vector I and to some extant along vector II. The second group in clustering (samples 1 and 7) was separated along vector I, which also separated by high NSPL and NIFL counts in Figs. 4 and 5, and form a distinct phenetic cluster in Fig. 3. The first group December 1993 Asiatic Herpetological Research Vol. 3, p. 69 TABLE 4. Variable coefficients for canonical variates I and II and estimated % influence of each vector for nine samples of Phrynocephalus. Vector I (77.58%) Vector II (8.67%) Character Variable coefficients % influence Variable coefficients % influence NSPL 0.332 8.040 0.715 4.430 NIFL NSDT 0.355 0.871 9.780 58.862 0.715 -0.450 4.430 1.763 NSNT NDVT 0.035 -0.093 0.097 0.007 -0.245 -0.036 0.011 0.001 SVLTTL SVL/LFL SVL/LHL LNE/LN LFL/LHL -0.008 0.026 -0.004 0.032 -0.001 0.000 0.001 0.000 0.001 0.000 -0.004 -0.043 0.005 -0.010 0.001 0.000 0.000 0.000 0.000 0.000 HL/HW TL/LFL TL/LHL 0.003 0.007 0.023 0.000 0.000 0.000 0.009 -0.003 0.031 0.000 0.000 0.000 TABLE 5. Characters for three species of population-groups of Phrynocephalus. Character SVL TL LFL LHL NSNT NSPL NIFL NDVT NSDT P. frontalis Range (Mean) P. przewalskii Range (Mean) P. versicolor Range (Mean) 45.8-57.2(51.5) 50.7-71.4(61.6) 46.1-55.6(49.5) 58.4-82.1 (72.4) 76.1-101.5(86.3) 53.9-69.4 (60.8) 23.3-29.5 (26.1) 27.4-35.9 (31.4) 22.7-26.9 (24.4) 37.1-48.7(43.1) 43.7-59.6(51.1) 36.5-43.4 (39.1) 2-4(3.2) 2-4 (2.9) 2-4 (2.0) 11-17 (14.9 15-20(17.6) 13-17(15.0) 12-17(15.1) 14-20(17.1) 12-16(14.0) 2-5 (2.8) 1-4(2.2) 0-3 (2.0) Color on ventral surface of tail tip Color on armpits 24-30 (26.4) 27-34 (29.9) 21-26(22.0) dark dark dark grey-white grey-white grey-white Color on back of body No. of teeth on maxillary No. of teeth on dentary 2-5 pairs of black spots irregular spots longitudinal line 2-4 transverse black- reddish bands 11 11 10 11 12 11 Ishium Cartilage Cartilage Posterior 1/2 is cartilage Meckel's cartilage Covered by splenial Covered by splenial Native not covered by splenial in clustering (sample 9) was separated along vector II and vector I. The third group, represented by samples 2, 3, 4, 5, 6, and 8 in clustering, was plotted along the centroid in Fig. 2. Characters having the greatest influence to vector I and vector II are NSDT, NSPL, and NIFL, respectively. Only males were used in univariate and multivariate analysis for geographic variation. Vol. 3, p. 70 Asiatic Herpetological Research December 1993 -'.1- on .2 20- Xi ea "m Q. = 15- 00 10 i i i i i i i i i 12 3 4 5 6 7 8 9 Localily FIG. 4. Diagram depicting geographic variation among nine samples of Phrynocephalus in supralabial number. Verticle line represents sample mean; open and closed bars represent range and one standard deviation, respectively. Discussion According to the morphological characters, populations of all nine samples can be divided into three major groups as shown in Fig. 2, 3, and Table 5. The first group (sample 9), a second group (samples 1 and 7), and a third group (samples 2, 3, 4, 5, 6, and 8). These are very similar to the descriptions by Strauch (1876) of P. versicolor, P. przewalskii and P. frontalis, respectively. It is apparent that P. frontalis is distributed in the Tengger Desert. This was not mentioned by authors who had studied the distribution of P. frontalis before. Strauch (1876) had pointed out that the color patterns of Phrynocephalus were very variable. In our study, we found that the color on the armpits of the group of P. frontalis and P. versicolor in the Tengger Desert was grayish-white. This was different from P. frontalis in the Mu Us Desert which has reddish on the armpits, and P. versicolor with reddish-blue on the armpits at Anxi and Wuwui in Gansu Province. In 142 specimens of P. frontalis in the Tengger Desert, only 4 specimens had reddish on the armpits. Therefore, we suggest that the color on armpits could not be a character for taxonomy in the genus Phrynocephalus. In the group of P. frontalis, sample 2 is !s 2(H ca = 15i 10 * 1 1 1 8 ill Locality FIG. 5. Diagram depicting geographic variation among nine samples of Phrynocephalus in infralabial number. Verticle line represents sample mean; open and closed bars represent range and one standard deviation, respectively. located on the south shore of the Yellow River, but samples 3, 4, 5, 6, and 8 are located on the north side of the Yellow River. They are very similar in morphological characters and they belong to the same species, P. frontalis. It is evident that the Yellow River does not function to geographically isolate P. frontalis. According to information on paleogeography and paleoclimatology, in Late Tertiary the Tengger Desert began to form . The area was very dry and cool. The Yellow River began to form in the Pleistocene with the uplift of the Qinghai- Xizang Plateau (Li, 1984). This means that the Tengger Desert was forming earlier than the Yellow River. Additionally, the Yellow River in this area had shifted its route several times from west to east during the past. We suggest the Phrynocephalus might have invaded and dispersed into the Tengger Desert before the formation of the Yellow River and that the river fails to be a geographic barrier for these lizards. Although the group of P. przewalskii located on the north side of the Yellow River is the nearest neighbor of the group of P. frontalis, they are quite different from each other on morphological characters. It is believed that this is the result of interspecific isolation. According to the study of morphological and skeletal characters of P. versicolor, P. przewalskii and P. frontalis, P. versicolor was quite different from the other two December 1993 Asiatic Herpetological Research Vol. 3, p. 71 60° 80° 100° 120° r- i v e, Sochi, Russia Abstract. -The Yew-box Grove of the Caucasian State Biosphere Reserve is home to seven species of amphibian. These species occur in a wide range of aquatic environments. The species composition, physical characteristics and history of each aquatic site was evaluated. The reproductive biology and food habits of each species was studied. These amphibians divide their niches on daily activity, seasonal activity, breeding site, microhabitat and food habits. The highest amphibian diversity and species overlap occurs in the most stable aquatic environments. Key Words: Amphibia, Russia, Caucasus, ecology. Introduction It is important when studying the influence of environmental factors on life history characteristics to distinguish those factors that are significant and those that are part of the "neutral background" (Monchadsky, 1958). The study of environmental influences is accomplished on sympatric species, usually closely related species (Orr and Maple, 1978; Ananjeva, 1981) at the population level (Pianka, 1973; Schoener, 1974; Schoener, 1977; Lyapkov and Severtsev, 1981). This is the study of the ecological niche (Pianka, et al., 1979. In the former Soviet Union one of the areas of highest amphibian diversity is found in the western Caucuses. The Yew- box Grove of the Caucasian State Biosphere Reserve is inhabited by eight amphibian species: Triturus vulgaris lantzi, T. cristatus karelini, T. vittatus ophryticus, P elodyte s caucasicus, Bufo verrucosissimus, Hyla arborea schelkownikowi, Rana ridibunda, and R. macrocnemis. Methods Field studies were conducted from 1980- 1982 in the Yew-box Grove (approximate area 302 ha) in the Caucasian State Biosphere Reserve and on adjacent land. Transect routes and study sites were selected on the basis of local forest typography (Gulisashvili, et al., 1975). Observations on these study sites along the FIG. 1. Study sites of sympatric amphibians in the Yew-box Grove, Caucasian State Biosphere Reserve. 1- Spring 118; 2- Opolznevaya Ravine: 3- Labirintovaya Ravine; 4- Glubokaya Ravine; 5- Khosta River; 6- Samshit Pond; 7- Pond on Malaya Khosta River. transects were made throughout the year (Fig. 1). The intensity of calls was recorded in spring (2-3 times per week), summer (1-2 times per month) and fall (2-3 times per week). Over 200 individuals were recorded. © 1993 by Asiatic Herpetological Research December 1993 Asiatic Herpetological Research Vol. 5 p. 75 iy y yi yn yui ix x xi xn Note : - river Khosta - - spring 118 - ravine Labirintovaya - ravine Glubokaya FIG. 2. Mean monthly temperatures of perennial bodies of water in the Yew-box Grove (1 980-1982). 2C 10 0 91 95 i n II iy y yi yn yn ix x xi xn ioo humidity FIG. 3. Climatram of the Yew-box Grove (1982). The location, weather conditions, air and body temperature, and behavior of each specimen was recorded. During 1982, detailed microclimate records were made in at the depth of 5 cm at the depth of 1C cm at the depth of 15 cm ■ Bt the depth of c'O cm FIG. 4. Winter soil temperature in the Yew-Box Grove (1982-1983). the Yew-box Grove (Fig. 2). Air temperature and humidity above the surface were recorded weekly by a thermograph (M-16AN) and a hydrograph (M-16AN) placed in a meteorological kiosk (Fig. 3). Soil temperatures were recorded at 5, 10, 15 and 20 cm depths (Fig. 4). Water samples were taken periodically for chemical analysis by established methods (Anonymous, 1978). Ambient light was measured with a light meter and converted into percent relative illumination. Climagrams were made (Formozov, 1934). Species composition of each biotope was calculated using conventional methods (Kashkarov, 1927; Dinesman and Kaletskaya, 1952). Food habits were studied using non-lethal methods (Verzhutsky and Zhuravlev, 1977). Results Description of Study Sites Study Site 1. Spring 118. This site contains a small perennial stream running Vol. 5 p. 76 Asiatic Herpetological Research December 1993 TABLE 1 . Some hydrochemical indices of the study sites. Sample site No. of samples pll Mean ions, total mg/1 total hardness mg-equiv/1 ammonium nitrogen nitrites nitrates mineral phosphorous mg/1 Glubokaya 15 7.36-8.30 8.02 196.0-292.1 223.7 2.42-3.60 2.78 0.00-0.04 0.02 0.00-0.026 0..009 0.16-0.37 0.24 0.003-0.008 0.003 Labirintovaya 10 7.30-8.45 7.87 182.8-380.1 868.7 2.28-3.67 3.08 0.00-0.60 0.19 0.00-0.020 0.004 0.03-0.55 0.18 0.003-0.029 0.009 Opolznevaya 10 8.00-8.27 8.17 254.3-350.0 301.3 3.04-3.71 3.42 0.01-0.11 0.05 0.00-0.034 0.020 0.08-1.41 0.45 0.003-0.009 0.006 Spring 118 10 7.54-8.50 8.12 347.1-472.9 413.4 3.91-5.61 4.85 0.01-0.45 0.12 0.001-0.07 0.019 0.97-2.67 1.78 0.013-0.164 0.089 Khosta River 10 7.70-8.50 8.12 208.2-315.9 252.3 2.35-3.39 2.98 0.00-2.32 0.22 0.00-0.016 0.006 0.02-1.44 0.37 0.003-0.022 0.008 through a sub-tropical, mixed broad-leafed forest (Fagus orientalis, Taxus baccata, Carpinus betulus) with an evergreen understory (Buxus colchicus, Ilex colchicus, Laurocerasus officinalis) and lianas (Hedera colchica, Smilax excelssior). Relative illumination is 1-2%. The stream flows over a bed composed of clay and sandstone. The stream flow derives from runoff and a sub-surface aquifer. The water chemistry of the spring water was hydrocarbonic-calcic with moderate mineralization, and moderately hard (Table 1). Hydrogen ion concentration is neutral to slightly basic, pH ranges from 6.89- 8.50. The concentration of nitrogen and nitrates (0.97-2.67 mg/1) is higher than in other waterways. This is a result of the subterranean flow. Ammonium concentration is normally low and increases during flash floods (up to 0.45 mg/1). Nitrites are also found in low concentrations except during flash floods (up to 0.08 mg/1). Phosphorus concentration is considerably higher than in other waterways (up to 0.16 mg/1). Rana macrocnemis, R. ridibunda, Bufo verrucosissimus, Hyla arborea schelkownikowi , Pelodytes caucasicus, and Triturus vittatus ophryticus are found at this study site and the latter two species breed there (Fig. 5, Table 2). Study Site 2. Opolznevaya Ravine. A small intermittent waterway flows through an eroded ravine through carbonic rock and clay. The vegetational community is analogous to Study Site 1, but box yews and beeches are dominate. Relative illumination is 2%. Stream flow is derived from runoff and aquifer. The water chemistry of the spring water was hydrocarbonic-calcic with moderate mineralization, and soft. Hydrogen ion concentration is neutral to slightly basic, pH ranges from 8.0-8.27. No amphibians were observed at this study site. Study Site 3 . Labirintovaya Ravine. A small intermittent, vernal-autumnal stream flows through an eroded, karst gorge with steep walls in a box yew forest. Relative illumination is 2%. Stream flow is derived from runoff and subsurface flow. During low waters periods the stream falls into a number of stagnant pools. The water chemistry of the spring water was hydrocarbonic-calcic with moderate mineralization, and soft. Hydrogen ion concentration is neutral to slightly basic, pH ranges from 7.3-8.45. Water content of nitrogen compounds is low but increases during flash floods. R. ridibunda is found here at this study site and P. caucasicus reproduces here. Study Site 4. Glubokaya Ravine. This site is a small pond in a limestone gorge with steep walls and surrounded by a box yew forest. Relative illumination is 2-3%. The pond is fed by a small, relatively constant spring flowing from Karst. December 1993 Asiatic Herpetological Research Vol. 5 p. 77 FIG. 5. Distribution of sympatric amphibians at Spring 118 in the Yew Box Grove. 1- Triturus vittatus ophryticus; 2- Pelodytes caucasicus; 3- Rana ridibunda; 4- Hyla arborea schelkownikowi; 5- Rana macrocnemis; 6- Bufo verrucosissimus. Spring temperature ranges from 11-16° C. The water chemistry of the spring water was hydrocarbonic-calcic with moderate mineralization, and moderately hard (2.42 - 3.6 mg-eq/1). Ammonium concentration is low (maximum = 0.04 mg/1) and in winter drops to 0. Nitrates are highest during summer low water period (up to 0.37 mg/1). Phosphorus concentration is very low (>0.01 mg/1). Bufo verrucosissimus, R. macrocnemis, P. caucasicus, R. ridibunda and T. v. ophryticus are found living at this study site and the last three species breed here. Study Site 5. Khosta River. This site is a small mountain stream 21 km long that drains a watershed of approximately 96 km2. The stream flows through a canyon formed in Cretaceous limestone at an average of 5 m-Vsec. Water flow is derived from runoff and springs arising in karst rock formations. Vegetation is a broad- leafed, subtropical Colchis type forest. Relative illumination is as high as 100%. The water chemistry of this stream is hydrocarbonic-calcic with moderate mineralization and basic pH (7.7-8.5). Ammonium concentration is not high and varies from 0 to 0.07 mg/1, but during floods it can reach 2.32 mg/1. This stream is subject to occasional flooding. Concentration of nitrates reaches a maximum during floods of 1.44 mg/1. Dissolved oxygen is 10-15mg/l and carbonic acid is 10mg/l. The Nizhe-Khostinsky Spring, flowing out of a karst formation, contributes 1-1.5 m-Vsec of flow at 11-13° C to the Khosta River. In low water periods the Khosta River above the spring nearly dries up and its temperature ranges from 0.6 (the river freezes) - 26° C. During low waters on the Khosta River the Nizhe-Khostinsky Spring provides a relatively stable flow and temperature regime. This spring serves as a barrier for dispersal of amphibians at this study site. Rana ridibunda and Hyla arborea schelkownikowi live and reproduce above the spring. Pelodytes caucasicus and Bufo verrucoisissimus live and breed below the spring. Study Site 6. Samshit Pond. This site is a small pond with flowing water in a stand of hornbeam {Carpinus betulus) in the broad-leafed forest. Relative Vol. 5 p. 78 Asiatic Herpetological Research December 1993 TABLE 2. Diversity of amphibian species at the study sites. Species Study Sites No. 1 No. 2 No. 3 No. 4 No. 5 No. 6 No.7 Triturus vulgaris X X Triturus vittatus X X X X Bufo verrucosissimus O 0 X 0 Pelodytes caucasicus X X X X Hyla arborea 0 o X Rana ridibunda o o X X 0 Rana nwcrocnemis 0 0 0 0 X X NOTE: X- reproduce, O- inhabit illumination is 100%. The pond flow is feed by run off and flow from the aquifer. The water chemistry of this pond is hydrocarbonic-calcic with moderate mineralization, moderately hard and alkaline. Hyla arborea schelkownikowi, Triturus vulgaris lantzi, Triturus vittatus ophryticus, Rana ridibunda, and R. macrocnemis, are found living at this study site and the last three species breed here. Study Site 7. Pond on the Malaya Khosta River. This site is a small stagnant pond located in the flood plain forest. The pond is filled by runoff and flow from the aquifer. Relative illumination is 50%. Bufo verrucosissimus, T. vulgaris lantzi, T v. ophryticus ,H. a. schelkownikowi and R. macrocnemis, are found living at this study site and the last two species breed here. Species Accounts Triturus cristatus karelini. This newt is an extremely rare and declining species along the Caucasian Black Sea coast. It was observed only once, at Study Site 1, in the box yew forest. Triturus vulgaris lantzi. This species exclusively inhabits stagnant ponds and ponds with flowing water, in well illuminated stands of hornbeam in broad- leafed forests, and adjacent areas. These newts are found in the ponds beginning in early March. Reproductive activity begins when water temperature reaches 10° C (usually from mid-March to early April. Females lay their eggs in shallow, thoroughly warmed waters at a depth of 5 cm and remain in the pond until the end of June. The newts over-winter in forest leaf litter and underground (Table 3). Triturus vittatus ophryticus. This species is found in both well illuminated broad-leafed flood plain forests and thick yew-box groves. It appears in bodies of water from the end of November through January. Reproduction occurs from January until the middle of April at water temperatures of 7-9° C. Females lay their eggs at depths of 5-10 cm. The adults stay in the water until the end of May. The newts over-winter in forest leaf litter. Bufo verrucosissimus. This toad is found throughout the Yew-box Grove with the exception of the steeper parts of the Khosta Canyon. Reproduction takes place in well illuminated running water in the Khosta River from February until May at water temperatures from 9.5-16° C. Eggs are deposited at a depth of 20-70 cm in strings through vegetation and other underwater objects. These toad over- winter in forest leaf litter beginning in December. Pelodytes caucasicus. This species inhabits back-water vegetation communities with flowing water. Reproduction lasts from the end of May until the end of October at water temperatures of 13-16° C. Females lay their eggs at a depth of 10-20 cm. These anurans over-winter in forest leaf litter. December 1993 Asiatic Herpetological Research Vol. 5 p. 79 I— U o. Rana macrocnemis X X X X X X X X X XXX X X X Rana ridibunda XXX X X XXX X XXX X X X Hyla arborea XX X X X X X X X Pelodytes caucasicus XXX X XXX X X X X X Bufo verrucosissimus X X X X X X X X X X X X X X X Triturus vittatus XXX XXX X X X X X X X Triturus vulgaris X X X X X X X X Z O P D CQ on 3 at summer biotopes: flood plain forest box tree forest hornbeam forest ash and oak forest yew and cherry laurel forest agrocenocis at spawning site: river pools (rain) lotic streams semi-lotic streams ponds at hibernation sites: stagnant, semi-lotic body of water soil and litter Breeding periods: January February March April May June-October Depth of egg-laying from water surface: 0-5 cm 5-10 cm 10-20 cm 20-50 cm over 50 cm Preferable water temp, for breeding: 4-5 degrees Celcius 5-7 degrees Celcius 7-9 degrees Celcius 9-18 degrees Celcius Vol. 5 p. 80 Asiatic Herpetological Research December 1993 TABLE 4. Comparison of development periods of Triturus vittatus ophryticus and Pelodytes caucasicus. MONTHS: SPECIES: xn I n ru IV V VI VU Vffl IX X XI Triturus vittatus (ad.) X X X X X X Triturus vittatus (larvae) *™8™ X X X X X X Pelodytes caucasicus (ad.) :;::::x:;::;:::::::: X X X X X X Pelodytes caucasicus (larvae) X X X X X X X X X X X X * Note: Shaded area represents the periods of mutual number limitation Hyla arborea schelkownikowi. This species is found in open, well illuminated ecotones throughout the grove. Reproduction take place from March until October in warm (>11" C), stagnant waters. Eggs are laid at a depth of 10-12 cm. These frogs over-winter in the forest leaf litter. Rana ridibunda. This frog is numerous in Khostinsky Canyon in open, well illuminated areas in the water-box yew forest ecotone. Reproduction last from January until March at water temperatures from 5-9° C. Frogs over-winter at the bottom of stagnant bodies of water. Rana macrocnemis. This frog is found in low numbers in all areas of the grove except in the rocks. Reproduction takes place from February to March in warm (4- 9° C), shallow water. Eggs are laid at 0-5 cm depth. During cool winters this frog over-winters in the forest leaf litter and during warm winter it remains abroad. Discussion During the summer all of the species of amphibians in the Caucasian Biosphere Reserve are broadly sympatric. However, during reproduction and winter retreat there is habitat segregation (Table 3). The highest level of overlap occurs in the summer in the flood plain forest where all of the above are found. The lowest level of overlap occurs in the box yew-cherry laurel stands with only B. verrucosissimus, and R. macrocnemis present. The box yew- cherry laurel stands are the most ancient forest type preserved in the Yew-box Grove. This ancient forest is dominated by box yew trees ranging from 500 to 2000 years old and has been virtually unchanged in appearance during that time. It is interesting that this ancient forest is inhabited by the indigenous Caucasian species B. verrucosissimus, and R . macrocnemis. The hornbeam tree is a pioneering species that invades disturbed areas such as those that have burned or been logged. It also grows in barren areas. The diversity of amphibian species in the hornbeam forest is caused by a number of factors: 1) secondary character of hornbeam forest; 2) relatively higher illumination (compared to the box yew forest; 3) presence of suitable water conditions for reproduction. When the hornbeam trees are young and the habitat is still open, such species as R. ridibunda, T. v. lantzi, and H. a. schelkownikowi are found. Later, when the trees become mature and the forest more closed and less illuminated, B . verrucosissimus, R. macrocnemis and P. caucasicus become established. This environment supports the highest level of sympatry of amphibian species in the Yew- box Grove. In box yew stands T. v. lantzi and H. a. schelkownikowi, the most illumination tolerant species, are not represented. These two species appear in December 1993 Asiatic Herpetological Research Vol. 5 p. 81 TABLE 5. Daily activity of amphibians of the Yew-box Grove. TIME OF ACTIVITY: SPECIES: 45 67 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 1 2 3 Triturus vittatus X X X X X X X X X XXX Triturus vulgaris X X X X X X X X X X XXX Pelodytes caucasicus X X X X X X X Bufo verrucosissimus X X X X X Hyla arborea XXX X X X X X X X X X X X X X XXX Rana ridibunda X X X X X X X X X X X X X X X X X Rana macroenemis X X X X X X X X the ecotone adjoining the southwest boundary of the grove (Table 3). It is interesting to note that B . verrucosissimus and R. macroenemis have very specific breeding requirements in terms of the aquatic environment required, but they occur over a large area during the terrestrial stages of their life history. They can be termed stenotopic or very restricted in terms of the reproductive requirements (B. verrucosissimus lays its eggs in rivers and R. macroenemis in small pools) and eurytopic in terms of their general distribution (Dazho, 1975). These two species are autochthones or indigenous to the broken country Caucasian region and reproduce in rapid mountain stream and ephemeral pools. The majority of lakes in this region are of recent origin and formed by glaciation, karst or from landslides. As this range of aquatic environments became available at the end of the last glacial period, amphibians successfully colonized those environments which met their reproductive requirements (Monchadsky, 1958). Bufo verrucosissimus lays its eggs in strings and wraps them around aquatic vegetation and other anchored objects in the water. This allows the toad to lay its eggs in fast flowing mountains streams which are generally unsuitable for other species. It lays its eggs at depths of 20 to 70 cm thus providing some protection from flash Hoods though many eggs perish in such floods. Pelodytes caucasicus and R. ridibunda are found sympatric with B. verrucosissimus. Rana ridibunda lays its eggs in the shallow, slow-moving sections of the river or in pools formed by floods. Reproduction in R. ridibunda is limited to the short period of winter low water and lasts from the end of January to the end of March. Pelodytes caucasicus is isolated from other breeding anurans temporally. It attaches its eggs to thin roots (2-10 cm) beginning in mid-June, when other species have finished breeding. Typically breeding sites are in the backwaters of streams, under vegetation canopies where temperatures are moderate. In small streams R. ridibunda and T. v. ophryticus are sympatric with P. caucasicus . At Glubokaya Ravine, Study Site 4, there is a high level of sympatry among the amphibian species. There is significant temporal segregation. Adult T. v. ophryticus remain in the water from the end of November to the end of May. Larval Vol. 5 p. 82 Asiatic Herpetological Research December 1993 TABLE 6. Faeces compostion of sympatric amphibian species of the Yew-box Grove SPECIES: Triturus Triturus Bufo Pelodytes Rana Rana cristatus vittatus vermcosissimus caucasicus ridibunda macrocnemis COLEOPTERA: Curculionidae X X X X Carabidae X X X X X Cerambycidae X X Chrysomelidae X X X Coccinelidae X X Scarabaeidae X Lucanidae X Silphidae X Elateridae X non det. X X DIPTERA: Tipulidae X Muscidae X non det. X X HYMENOPTERA: Apidae X Vespidae X X X Formicidae X HEMIPTERA: Pentatomidae X X X Pyrrhocoridae X X ANNELIDES: Hirudinea X Oligochaeta X ARTROPODA: Lepidoptera X Scorpiones X Amphipoda X Isopoda X X X MOLLUSCA: Pisidium X Oxychilus X development occurs from March until August (Table 4). Pelodytes caucasicus can be heard calling from May until the middle of October. The adults of T. v. ophryticus and P. caucasicus prey upon each others larval stages. When post metamorphic T. v. ophryticus begin leaving the water in late July they are prey upon by adult P. caucasicus . In December when adult newts enter the water to breed they capture the smaller sizes of P. caucasicus tadpoles. During these periods of intense competition and predation both species adopt several strategies for preying upon the other for avoiding predation (Smith 1981). December 1993 Asiatic Herpetological Research Vol. 5 p. 83 TABLE 7. Size limits of feeding objects of sypatric amphibians in the Yew-box Grove SIZE OF FEEDING OBJECTS ( in mm) SPECIES: up to 3 3-3. 5 3. 5-4 4-4. 5 4. 5-5 5-5. 5 5. 5-14. 5 14 5-19 19-36 Triturus vulgaris X X Triturus vittatus X X Bufo verrucosissimus x X X X X X X X Pelodytes caucasicus x x X X X X X Rana macrocnemis X X X X X Rana ridibimda X X X The highest population densities of T. v. ophryticus, T. v. lantzi and T. c. karelini occur in small forest lakes. Much lower population densities are found in mountain streams. Recently formed lakes are apparently the most suitable habitat for these species of newts. As lakes mature, sediments accumulate and they become less suitable habitat and populations decline and are preserved at low levels in nearby streams. The highest level of sympatry occurs in lakes during breeding season (Table 3). There is, however, very little competition because of temporal and microhabitat segregation for egg deposition and deposition sites. Lakes and deep pits (Study Site 7) are breeding sites for H. a. schelkownikowi where it is spatially segregated from other amphibians but overlaps temporally with P. caucasicus (Table 3) and activity patterns (Table 5). In the Yew-box Grove P. caucasicus and H. a. schelkownikowi are allopatric at breeding sites. On the Caucasian Black Sea Coast they are sympatric at breeding sites. The interrelationships of these populations in the zones of sympatry have not been studied. On the Caucasian Black Sea Coast winters are mild with abundant precipitation. Most amphibians remain active and abroad through the winter. Though, on the occasional cold days they become torpid. The exceptions are adult P. caucasicus and H. a. schelkownikowi. During some cold winters when night temperatures fall to -10 to -12° C all amphibian species enter hibernation. Bufo verrucosissimus, P. caucasicus, H. a. schelkownikowi. and T. v. lantzi pass hibernation hidden in the soil and leaf litter. Rana macrocnemis hibernates in the soil, leaf litter and in the water. Winter soil temperatures at 20 cm depth ranges from 3.5 - 7.2" C. In the spring the water warms more quickly than the soil and this may explain why the those amphibians hibernating in the water breed earlier than those hibernating in the soil. The mechanism of niche isolation of symbiotopic species includes layering, or the formation of adaptive groups (Dinesman 1948a) and differences in daily and seasonal activity. For example, at Study Site 1, T. v. ophryticus occupies the deepest aquatic level, R. ridibunda and P. caucasicus are found in the lower intermediate levels, R. macrocnemis and B. verrucosissimus occupy the higher intermediate levels and H. a. schelkownikowi is found at the shallowest level and onto land (Fig. 4). Some of the pattern of species distribution can be explained by varying tolerance of desiccation among the different species (Dinesman, 1948b). Those species that are found at the same level are active at different times of the day (Table 5). The amphibians of the Yew-box Grove can be divided into three groups based on their food habits: 1) feeding on hydrobionts (newts), 2) feeding on arboreal invertebrates (tree frogs), 3) feeding on Vol. 5 p. 84 Asiatic Herpetological Research December 1993 terrestrial invertebrates (all other species). Rana ridibunda feeds on both aquatic and terrestrial invertebrates (Table 6, 7). In a given area species diversity is dependent upon niche separation (Pianka, 1981). The seven amphibian species studied here are characterized by biotopic (including breeding site choice, hibernation sites and summer activity ranges), seasonal activity period, daily activity period and food habits isolation. In the Yew-box Grove the highest level of species diversity is achieved in the most stable aquatic environments. Water temperature during the breeding season, level of illumination and water chemistry appear to be an important characteristics. Acknowledgments We wish to express our gratitude to N. B. Ananjeva for valuable consultations and to I. V. Marchukaitis for preparing the illustrations. Literature Cited ANANJEVA, N. B. 1981. [On the studies of sympatric species using reptiles as an example].. Nauka Publishing House, Leningrad. Pp. 15- 26. (In Russian). ANONYMOUS. 1978. [Unified methods of water analysis in the USSRL Leningrad, Gidrometizdat Publishing, Leningrad. (In Russian). DAZHO, P. 1975. [Foundations of ecology! Progress Publishing House, Moscow. 408 pp. (In Russian). DINESMAN, L. G. 1948a. [On the question of ecological differentiation of amphibian species]. Bull. MSNT 3(6):47-50. (In Russian). DINESMAN, L. G. 1948b. [Adaptation of amphibians to varying humidity]. Zoological Journal 27(3):231-240. (In Russian). DINESMAN, L. G. AND M. L. KALETSKAYA. 1952. [Methods of amphibian and reptile counts! Pp. 329-341. In Methods of counting and geographical distribution of terrestrial vertebrates. Moscow. FORMOZOV, A. N. 1934. [Special type of climagrams for ecological studies]. Scientific Transactions of Moscow University 2:271-274. (In Russian). GULISASHVILI, V. Z., L. B. MAHATADZE, AND L. I. PRILIPKO. 1975. [Vegetation of the Caucasus]. Nauka Publishing House, Moscow. 227 pp. (In Russian). KASHKAROV, D. N. 1927. [Methods of qualitative studies of vertebrate and analysis]. Tashkent. 23 pp. (In Russian). LYAPKOV, S. M. AND A. S. SEVERTSEV. 1981. [Mechanism of co-existence of two species of Far-Eastern Anura]. Zoological Journal 3:398- 408. (In Russian). MONCHADSKY, A. S. 1958. [On classification of factors of environment]. Zoological Journal 37(5):680-691. (In Russian). ORR, L. P. AND W. T. MAPLE. 1978. Competition avoidance mechanisms in salamander larvae of the genus Desmognathus. Copeia 1978(4):679-685. PIANKA, E. R. 1973. The structure of lizard communities. Annual Review of Ecology and Systematics 4:53-74. PIANKA, E. R. 1981 (Evolutionary ecology! Mir Publishing House, Moscow. (In Russian). PIANKA, E. R., R. B. HUEY, AND L. R. LAWLOR. 1979. Niche segregation in desert lizards. Pp. 67-115. In Analysis of ecological systems. Columbus, Ohio State University Press. SCHOENER, T. W. 1974. Resource partitioning in ecological communities. Science 185:27-38. SCHOENER, T. W. 1977. Competition and the niche. Pp. 35-136. In C. Cans and D. W. Tinkle (eds.). Biology of the Reptilia, Vo. 7, Ecology and Behaviour. Academic Press, New York. SMITH, J. M. 1981. [Evolution of behaviour]. Mir Publishing House, Moscow. (In Russian). VERZHUTSKY, B. N. AND B. E. ZHURAVLEV. 1977. [Sparing method of studies of reptile trophic spectrum]. Pp. 58-59 In The Problems of Herpetology. Fourth USSR Herpetological Conference. Science Press, Leningrad. (Abstr.XIn Russian). December 1993 Asiatic Herpetological Research Vol. 5, pp. 85-95 1 Notes on a Collection of Squamate Reptiles from Eastern Mindanao, Philippine Islands Part 1: Lacertilia Brian E. Smith Museum of Natural History. Louisiana Stale University, Baton Rouge, Louisiana 70803-3216. Present Address: Department of Biology. University of Texas at Arlington Box 19498 Arlington, Texas 76019-0498 Abstract. -In 1982, I spent six months collecting the herpetofauna of several areas in eastern Mindanao, Philippine Islands. I present species accounts for all lizards collected at these sites, and draw conclusions about the biodiversity of second-growth and primary forest habitats. I found second-growth habitats to be depauparate compared to primary forest habitats. Species which may depend on primary forest habitat and also species apparently restricted to such habitats are detailed. The importance of primary forest reserves, selective logging, and a mosaic of successional habitats within primary forest reserves is discussed. Key words: Reptilia, Squamata, Lacertilia, Philippines, taxonomy, ecology, biodiversity Introduction Little is known of the ecology, distribution, and life history of Philippine lizards. Taylor (1922a, b, c; 1923; 1925) investigated the systematics and zoogeography of the Philippine herpetofauna and provided some ecological information. Brown and Alcala (1961, 1964, 1978, 1980) undertook ecological and systematic studies on some islands, Alcala (1986) recently reviewed the herpetofauna, and Auffenberg (1988) has detailed the ecology of the Gray's monitor, Varanus olivaceus. However, many islands remain poorly known. This and a subsequent paper provide information on the natural history of squamate reptiles collected in eastern Mindanao, Philippine Islands, during a collecting trip to the area made in 1982. I also comment on squamate assemblages occurring in primary and second-growth dipterocarp forest in an effort to pinpoint species of special concern should primary forest continue to be lost to logging and agricultural practices. Site Description Vegetation. — The vegetation of the Philippines was described by Brown (1919) and Dickerson (1928) and a detailed study of southeast Asian rain forests was published by Richards (1952). Trees of the family Dipterocarpaceae dominate the rain forests of the Philippines. For the purposes of my study, primary dipterocarp forest is defined as dipterocarp forest that has apparently never been logged. Early second-growth dipterocarp forest is defined as dipterocarp forest selectively logged one or two years previously. It has abundant ground cover of rattan, herbaceous vines, shrubs, and lianas. A few small (<2 m) trees were present, and the ground was littered with fallen logs. Selective logging was practiced at my study sites, and some small dipterocarps were left standing, though not enough to cast appreciable shade. Late second-growth dipterocarp forest is defined as forest selectively logged three or more years ago. It usually had a closed canopy composed of young fast-growing trees (mostly Trema species) about 10-15 m tall. Small dipterocarps were present, and the herbaceous understory was extremely dense. The vegetation of my study areas is discussed in more detail in Smith (1985). Site 1. — This site was located in the coastal mountains (Diuata Range) of east- central Mindanao, 55 km south, 20 km west of Bislig Bay and 10 km southeast of Mt. Agtuuganon (Fig. 1). Areas from 500- 800m elevation were sampled. Vegetation of this area was early second-growth and primary dipterocarp forest. Slopes were generally very steep (Fig. 2, 3, 4). This site was sampled from April 6 to August 15, 1982. Site 2. — This site was also in the Diuata Range, 33 km south and 7 km west of Bislig Bay (Fig. 1). Vegetation at this site consisted entirely of late second-growth © 1993 by Asiatic Herpetological Research Vol. 5 p. 86 Asiatic Herpetological Research December 1993 M-f 12^° 125° East Longitude 2# FIGURE 1. Map of Mindanao, Philippine Islands, showing locations of collecting sites 1 and 2 (triangle) and site 3 (circle). dipterocarp forest. Slopes were moderately inclined to steep. The site was 400 m in elevation, and was sampled from August 21 to September 4, 1982. Site 3. — I also visited Mount Talomo, a 2693 m peak located 30 km west of Davao City in the Mount Apo Range of southern Mindanao (Fig. 1). I sampled areas around the Philippine Eagle Captive Breeding Project (PECBP) field station (about 1000 m elevation) from September 8-13, 1982. Workers at the field station made incidental collections at this site from April to September. According to residents, logging in this area was discontinued sometime in the mid- to late-1960's. The area was selectively logged and appears to be more similar to typical pristine forest than many other logged areas I visited. Human disturbance is considerable on the slopes of Mount Talomo. Farms extend up the slopes from Davao Gulf to about 900 m elevation. Coconuts, bananas, pineapples, coffee, and various fruits and vegetables are grown. Areas sampled range from 600- 1050 m. Slopes are gently to steeply inclined. Climate. — Rainfall was heavy at all sites. At sites 1 and 2, there is no marked dry season. The wetter season usually occurs from November to March, with the heaviest rains in December and January (Census Office of the Philippine Islands 1920; Dickerson, 1928; Willmott et al, 1981). Annual rainfall in Surigao (the nearest weather station to sites 1 and 2) is 3647 mm, with 2360 mm falling from November to March, and 1 191 mm of this amount in December and January alone (Willmott et al., 1981). Surigao is in the lowlands, and rainfall at my collecting sites may have been higher. Site 3 also does not have a marked wet or dry season (Census Office of the Philippine Islands, 1920; Dickerson, 1928). Annual rainfall in Davao (at the base of the Mount Apo range) is 1942 mm (Willmott et al., 1981). This figure undoubtably increases with elevation. During some times of the year, the PECBP station may be shrouded in clouds for weeks at a time. Temperature was relatively constant at all sites. At site 1 it usually ranged from 20-25 C under the canopy, and from 20-30 C in the open. The minimum temperature recorded was 18 C, the maximum was 34 C. Temperatures at site 2 were similar. Temperature readings were not available from site 3, but it was slightly cooler because of the higher elevation. Humidity at all sites varied from 79-100%. Occasionally typhoons strike Mindanao, but they generally lack the severity of those hitting more northerly islands (Census Office of the Philippine Islands, 1920, Dickerson, 1928). In March of 1982 a typhoon struck Bislig Bay near sites 1 and 2. The only effect at the study sites was moderate, steady rain for several days. Methods Specimens were collected using drift fences (Gibbons and Semlitsch, 1981) and by hand. Drift fences were 0.5 m high and 18 m long and constructed as detailed in Gibbons and Semlitsch (1981). Pit-cans and funnel traps were placed at either end of the fence. Pit-cans seemed effective at capturing all terrestrial lizards except Varanus spp. Funnel traps were primarily useful in capturing snakes. Drift fences appeared to adequately sample reptiles moving on the soil surface. To sample December 1993 Asiatic Herpetological Research Vol. 5 p. 87 FIGURE 2. Primary forest at the edge of a road cut at site 1. The ridge top is about 900m elevation. arboreal fauna, I examined epiphytic ferns and trees felled during logging. These efforts were largely unsuccessful and arboreal species are under-represented in the collection. Data on macrohabitat (primary, early, or late second-growth dipterocarp forest), microhabitat (fossorial, terrestrial, or arboreal), elevation, date, and time of day were recorded for each specimen collected. Standard measurements including snout-vent (SVL), tail (TL), and total length (TTL) were taken in the field on freshly killed specimens. General weather conditions were noted daily. Specimens were dissected in the laboratory to determine sex, stomach contents, and reproductive condition of females. Food contents were identified usually to family for the arthropod prey of these lizards. Standard scale counts were also taken, but no deviation from previously published data was noted, and scale count data are not reported herein. Species Accounts Family Gekkonidae Cyrtodactylus agusanensis: — Specimens of this species were collected in all habitats at sites 1 and 2. This species' morphology suggests nocturnal habits (vertically elliptical pupils), but half the specimens were captured in drift fences during the day. Specimens captured at night were taken on bushes and logs, 1-3 m above ground. My observations do not agree with Alcala (1986), who states that this species is found in swamps and along rivers. My specimens were all taken far from such habitats. I also did not find this species to be particularly rare, as did Alcala Vol. 5 p. Asiatic Herpetological Research December 1993 (1986). Females taken June 14 and 23 had one large egg (15.7-18.2 mm long) in each oviduct. Small males were captured April 2 and 9 and a small female July 31. One juvenile was captured at site 2 on August 26. Stomach contents included insects of several families and a shed skin, prohahly of C. agusanensis. Specimens examined: LSUMZ 41601- 41609, 41640. Family Agamidae From their morphology, all agamids collected on the study sites appear to be arboreal, although most specimens, except Draco species, were caught on or near the ground. Calotes cristatellus: — A female was collected at about 550 m in early second- growth forest at site 1. A male specimen lacks additional data. This species is rare, transient in the habitats sampled, or mostly arboreal and hence under-represented in the collection. It has morphology indicative of a highly arboreal lifestyle. The female, collected on July 1, had one large egg (38.4 and 35.2 mm) in each oviduct. Stomach contents included lepidopteran larvae, unidentified insects, and a snail. Specimens examined: LSUMZ 41737, 41738. Draco mindanensis: — Taylor (1922a) collected only two specimens of this species. They were taken at 1 100 ft. at the base of Malindang Mountain, northwestern Mindanao. Inger (1983) examined taxonomic characters in nine specimens but gave no ecological data for them. My specimens were taken at 650 m at site 1 in primary forest. The species may be confined to primary forest. Draco mindanensis is diurnal and arboreal. A female contained two oviducal eggs 17.5 and 18.4 mm long. The date of capture of this specimen is unknown. Stomach contents consisted of several families of insects. This species is apparently not an ant-feeding specialist like its congener, D. volans (see below). D. mindanensis is reported from Catagan and Malindang Mountain in northwestern Mindanao (Taylor, 1922a) and the Diuata Mountains in the province of Davao del Norte, east- central Mindanao (this study). Specimens examined: LSUMZ 41678- 41680. Draco volans. — This species is very common in early second-growth forest and is probably the most conspicuous lizard species in this or any other habitat. Individuals are commonly seen running along branches, displaying, and gliding. They are diurnal and exclusively arboreal. This species was never seen in primary forest. A female containing two eggs (14.7 and 13.9 mm long) was collected on July 20. Taylor (1922a) stated, and my data confirm, that this species feeds exclusively on ants. Specimens examined: LSUMZ 41741- 41748. Gonyocephalus semperi. — Although arboreal by morphology, five specimens were captured on the ground in drift fences. One was caught by hand 1 m above the ground on a large tree. This species is diurnal, as is its' congener G. godeffroyi in the Solomon Islands (McCoy 1980). Four were captured in primary forest, two in late second-growth forest. Due to the species' arboreal habits, it is highly likely that G. semperi does not occur in highly disturbed areas largely lacking trees. McCoy (1980) found that G. godeffroyi also avoids open areas in the Solomons. One G. semperi exhibited aggression and grunted when handled. The single female captured June 16 contained three developing eggs (5.8, 6.6, and 7.3 mm in length). Stomachs examined contained the remains of insects of the families Chilopoda, Coleoptera, Hymenoptera, Orthoptera, and larval Lepidoptera. Specimens examined: LSUMZ 41730- 41735. Hydrosaurus pustulosus. — This is a juvenile specimen collected by a native on December 1993 Asiatic Herpetological Research Vol. 5 p. 89 Figure 3. Early second-growth forest at site 1. This site had been selectively logged two or three years prior to this photograph. Trees in the middle backgound were deliberately left standing during the selective logging procedure. June 19. Its' stomach was empty. It is said by Alcala (1986) to be omnivorous, which generally agrees with observations of captive specimens at the Dallas Zoo. Auffenberg (1988) states that adult H. pustulosus are entirely folivorous in the wild. Captive specimens usually lay 6-8 eggs measuring roughly 50 mm in length about once a year (Mitchell 1985). This species is said to be common in the Philippines near unpolluted mountain streams (Alcala 1986). It has also been observed around coastal fishing villages, utilizing as vertical perches the stilts or piers that support homes over water (L. A. Mitchell, personal observation). Specimen examined: LSUMZ 41739. Family Scincidae In the Philippines, skinks far exceed the other lizard families in number of species, abundance, and probably in the variety of niches they occupy. They are often the most abundant lizards in all habitats sampled, with the exception of early second-growth forest, where Draco volans is more abundant. The leaf litter herpetofauna is dominated by species of the family Scincidae. Brachymeles gracilis hilong: — Specimens were captured only in late second-growth forest (site 2, 400 m) and at Mount Talomo (site 3). Brown and Alcala (1980) stated that this fossorial species is found under leaves, duff, rotting logs, and Vol. 5 p. 90 Asiatic Herpetological Research December 1993 Figure 4. Small permanent stream in primary forest at site 1. Small streams such as this one were common in areas of primary forest at all sites. in loose soil, usually only in primary forest from 50-1000 m elevation. This species is apparently rare or absent in early second- growth forest. Stomach contents indicate that this species is a generalized insectivore, however, part of a skink tail (Mabuya or Sphenomorphus species) was found in one stomach. Specimens examined: 41725. LSUMZ 41719- Brachymeles schadenbergi oriental is. — This fossorial species was often caught in drift fences after long, steady rains. It is found under logs and in leaf litter and loose soil in primary and second-growth forests at elevations from 50- 10(H) m (Brown and Alcala, 1980). A female collected July 5 had three developing embryos 18.1, 15.9, and 15.0 mm in her oviduct. Brown and Alcala (1980) stated that this subspecies is ovoviviparous and usually has 2 or 3 young. This species is also a generalized insectivore. In addition, a lizard tail (probably Brachymeles species) was found in one specimen's stomach. Specimens examined: LSUMZ 41726- 41729. Lamprolepis smaragdina philippinica. — This was a very common arboreal species in cultivated areas and villages, and a single specimen was taken in a coconut plantation on Mount Talomo between 700 and 800 m elevation. In contrast to Brown and Alcala (1980) and Alcala (1986), I never observed this species in primary or second-growth forests. My study sites did not encompass December 1993 Asiatic Herpetological Research Vol. 5 p. 91 agricultural areas or villages. Specimen examined: LSUMZ 4 1 648. Lipinia semperi. — Taylor (1922a) noted that this species is commonly found in old tree stumps and hollow trees. Alcala (1986) states that it is arboreal and rare. My specimen was taken in daylight from an epiphytic fern 4 m above ground in primary forest at 800 m elevation. Specimen examined: LSUMZ 41740. Mabuya multicarinata multicarinata. — This is a very active and abundant skink that favors open areas. It is diurnal and terrestrial. All but two specimens were caught in early second-growth forest. One specimen was taken in late second-growth forest. An additional specimen was observed in primary forest, but this was within 100 m of a logging road and it may have been transient. This species seems to be a lizard of open areas, and it may have originally occupied natural treefall gaps in the forest. With extensive clearing of areas for logging and cultivation, M. m. multicarinata has probably increased substantially in numbers. Of eleven females collected throughout this study, only two failed to contain well-developed eggs. The rest had 2 (6 specimens) or 3 (3 specimens) large oviducal eggs 11.0-15.5 mm long. Juveniles 29-41 mm SVL were captured on June 24 and August 1,4, 12, and 26. Another juvenile was taken from the stomach of a snake (Cyclocorus nuchalis taylori) that was collected July 1. Stomachs examined contained insects of many taxa, and I consider this species a generalized insectivore. Specimens examined: LSUMZ 41610- 41639. Sphenomorphus Species Lizards of the genus Sphenomorphus dominate the leaf litter herpetofauna of the primary forest in the areas I investigated, and they are sometimes conspicuous in secondary growth as well. Sphenomorphus exceeds all other genera at my sites in number of species and individuals. Its' species are mostly diurnal and terrestrial. Sphenomorphus acutus: — Brown and Alcala (1980) state that this species is strictly arboreal, which may account for the paucity of specimens collected. However, three specimens were collected in drift fences during daylight, so they are at least occasionally found on the ground. This agrees with Alcala's (1986) observations. Stomach contents included arachnids and orthopterans. Specimens examined: LSUMZ 41713- 41715. Sphenomorphus coxi coxi. — This was the most common ground-dwelling lizard in the primary forest and is the most conspicuous member of the leaf-litter herpetofauna in this habitat. It is fairly common in second-growth forest as well, but it is not seen in the open as often as Mabuya multicarinata multicarinata. Although unquantified, I believe that these observations represent a real difference, and that M. m. multicarinata replaces 5. c. coxi as the dominant skink in second- growth habitats. Gravid females were collected April 27 (1 egg, 5.2 mm long), July 3 (2 eggs, 8.5 and 9.2 mm), and August 5 (2 eggs, 14.7 and 15.2 mm). One juvenile was collected May 2 1 and two on May 24. Four others were collected August 9, 13 (2), and 28. These juveniles measured 34-45 mm SVL. One small specimen identified as a male measured 45 mm SVL. These data suggest two hatching seasons during the six months of this study, one in May and the other in August. Apparently, reproductive maturity is reached at approximately 45 mm SVL. Stomachs examined contained insects of many taxa. I consider this species to be a generalized insectivore. Specimens examined: LSUMZ 41649- 41677. Sphenomorphus decipiens. — Although Alcala (1986) states that this species is rare, I found it to be a fairly common diurnal Vol. 5 p. 92 Asiatic Herpetological Research December 1993 member of the leaf litter herpetofauna in primary forest. It was never found in second-growth forest. A juvenile (25 mm SVL) was collected June 23. Stomachs contained larval lepidopterans and other insects. Specimens examined: LSUMZ 41704-41711. Sphenomorphus fasciatus. — Brown and Alcala (1980) noted that this species was a common terrestrial skink. During my study, I collected only one specimen at site 1. Since my capture techniques seemed especially efficient in sampling the terrestrial herpetofauna, I conclude that S. fasciatus was rare at my collecting sites. The stomach of this specimen contained insects of several taxa. were either captured in drift fence traps or found under cover. At night, individuals were found actively foraging. This species is apparently secretive during daylight. Specimens frequently grunted during capture and attempted to bite. All specimens were caught in primary forest except one caught in late second-growth forest. This species is apparently absent from highly disturbed areas. Gravid females were caught March 16 (5 eggs, 3.5-6.3 mm) and June 14 (5 eggs, 12.3- 13.9 mm). Specimens examined: LSUMZ 41641- 41647. Discussion Specimen examined: LSUMZ 41716. Sphenomorphus steerei. — This species is common in primary forest, and is occasionally found in second-growth forest. In my study, it was more common above 650 m elevation, although Brown and Alcala (1980) stated that it occurs from sea level to 2000 m. Stomachs contained insects of several different taxa. Specimens examined: LSUMZ 41697- 41703,41712. Sphenomorphus variegatus. — This species is a common diurnal leaf-litter lizard of the primary forest. It was never found in early second-growth forest, but two specimens were taken in late second- growth forest. S. variegatus appeared to be completely absent from highly disturbed areas. Females taken April 1 1 and May 24 each contained two eggs (11.0 and 10.3 mm; and 6.5 and 6.1 mm, respectively). Juveniles (SVL 24-39 mm) were captured May 23, June 29, August 9, and August 26. Stomachs contained insects of many different taxa. Specimens examined: LSUMZ 41681- 41696. Tropidophorus partelloi. — This species was found to be active during the day and night. Specimens collected during the day The continuing destruction of tropical rain forest worldwide makes it imperative that species of special concern are identified in these habitats. In the section below, I will attempt to pinpoint species which could be adversely affected by continuing deforestation, where my data are adequate to do so. I make special reference to the lizards of the forest floor, since I feel that these lizards were the most accurately sampled species with the techniques that I used. These species are all skinks: Mabuya multicarinata multicarinata, Sphenomorphus coxi coxi, S. decipiens, S. steerei, S. variegatus, and Tropidophorus partelloi. Sphenomorphus fasciatus is not considered, since only one specimen of this species was collected. Primary Forest This was the richest habitat sampled, containing all six of the skinks mentioned above. Of these six, five seem to be regular inhabitants of the primary forest. In addition, there were two common arboreal lizards: The gecko Cyrtodactylus agusanensis and the agamid Gonyocephalus semperi. The most commonly collected terrestrial lizards listed in decreasing order of abundance were: Sphenomorphus coxi coxi, S. variegatus, S. steerei, S. decipiens, and Tropidophorus partelloi. Species which were sparsely collected in primary forest include the December 1993 Asiatic Herpetological Research Vol. 5 p. 93 arboreal lizards Calotes cristate/Ins, Draco mindanensis, Lipinia semperi, and Sphenomorphus acutus, and the terrestrial lizard Mabuya multicarinata multicarinata. I do not consider M. m. multicarinata to be a regular inhabitant of primary forest. Early Second-Growth Forest This was the most depauperate habitat sampled. Two arboreal lizards were common: Cyrtodactylus agusanensis and Draco volans, the latter species being the most common lizard observed in any habitat. The common terrestrial lizards were: Sphenomorphus coxi coxi and the extremely abundant Mabuya multicarinata multicarinata. This habitat contained only three of the six terrestrial skinks of interest. The skink Sphenomorphus steerei was only represented by two individuals collected in this habitat, and I consider it to be rare in early second-growth habitats. M. m. multicarinata was virtually absent in primary forest, but it was found to be twice as common in second-growth habitats as the next most common lizard, S. c. coxi. I spent approximately two man-months apiece working in primary forest and early second-growth forest at site 1. I have assumed that the time spent in these two habitat types was equal and have calculated the Brillouin diversity index (Krebs 1989) for each habitat using the absolute number of specimens collected in each habitat type of the six terrestrial skinks of interest noted above. Indices of 0.610 and 0.309 are calculated for primary forest and early second-growth forest habitat types, respectively. Although not an exact measurement, this rough estimate of diversity points out the major difference in biodiversity between these two habitat types as regards terrestrial skinks. Late Second-Growth Forest I spent very little time in this habitat type, yet the species collected here provide valuable insights into possible lizard population successional patterns. The arboreal agamid Gonyocephalus semperi, absent in early second-growth forest, was found in late second-growth forest. whereas Draco volans, common in early second-growth forest, was absent from late second-growth forest. The terrestrial skinks Sphenomorphus variegatus and Tropidophorus partelloi, both absent from early second-growth forest and common in primary forest, were collected in late second-growth forest. Species of Special Concern It is clear that less complex habitats will support a less complex community of plants and animals, by definition. Although my data are sparse, it is obvious that the biodiversity of terrestrial skinks (and other lizards) is far different after logging operations alter the forest structure. The clearing of primary forest creates habitat for open habitat specialists such as Mabuya multicarinata multicarinata and Draco volans, while decreasing habitat for the species which seem to be confined to the primary forest. Especially notable is die lack of Sphenomorphus variegatus, which is common in primary forest but absent in early second-growth forest. Other species which may be adversely affected by logging activities include the skinks Tropidophorus partelloi, Sphenomorphus steerei, and 5. decipiens; and the agamids Draco mindanensis and Gonyocephalus semperi. Except for D. mindanensis, all these species were collected in late second- growth and primary forest. D. mindanensis was only found in primary forest. Sphenomorphus coxi coxi was the only lizard commonly found in all habitats, although it was less frequently seen in the open where it was syntopic with Mabuya multicarinata multicarinata. Conclusions The collection of primary forest species in areas of late second-growth forest points towards the possibility of a sustained harvest of the primary forest for lumber. However, there is absolutely no data on the periodicity of such a harvest, nor any precise information on changes in biodiversity through a successional series. It is imperative that primary forest be conserved as a refuge for certain species Vol. 5 p. 94 Asiatic Herpetological Research December 1993 which may only occur there, such as Draco mindanensis. Possibly, a system of reserves with rotating areas of selective logging may help to conserve Philippine lizard species. However, at the present time research towards such an end is entirely lacking, and the conversion of large parts of the Philippines towards a much impoverished lizard fauna continues unabated. Acknowledgments This paper is part of a thesis submitted to the Louisiana State University Graduate School in partial fulfillment of the requirements for the degree of Master of Science. Dr. D. A. Rossman supported and guided all phases of this work. Dr. J. V. Remsen, Dr. J. W. Fleeger, and Dr. W. J. Harman also assisted in innumerable ways. Dr. R. S. Kennedy of the Philippine Eagle Conservation Project initially suggested this study and was instrumental in obtaining funding as were Dr. J. P. CTNeill and Dr. D. A. Rossman. The Philippine Paper Company provided the study site, storage space, and transportation. A collecting permit was graciously provided by the government of the Philippine Islands. Mario Caleda, Hector Miranda, Kevin Moore, Bill Wischusen, Mark Witmer, and especially Susing Babao, all of the Philippine Eagle Conservation Project, assisted in obtaining specimens. Robin Lawson, Bill Sanderson, and Randy Vaeth assisted in identification of specimens and food remains. Jim Murphy and staff at the Dallas Zoo Herpetology Department provided an intellectually stimulating environment in which to finish final editing of this paper. Ardell Mitchell shared his observations on both captive and wild- caught Hydrosaurus pustulosus. Jim Murphy, Dr. E. D. Brodie, Jr., and Dr. J. A. Campbell provided helpful comments on the final draft. Literature Cited ALCALA, A. C. 1986. Guide to Philippine Flora and Fauna, Vol.X: Amphibians and Reptiles. Natural Resources Management Center, Ministry of Natural Resources and University of the Philippines, Manila. 195pp. AUFFENBERG, W. 1988. Gray's Monitor Lizard. University of Florida Press, Gainesville, Florida. 419pp. BROWN, W. C., AND A. C. ALCALA. 1961. Populations of amphibians and reptiles in the submontane and montane forests of Cuernos de Negros, Philippine Islands. Ecology 42:628- 636. BROWN, W. C., AND A. C. ALCALA. 1964. Relationship of the herpetofauna of the non- dipterocarp communities to that of the dipterocarp forest of southern Negros Island, Philippines. Senckenbergiana Biologia 45:591- 611. BROWN, W. C., AND A. C. ALCALA. 1978. Philippine Lizards of the Family Gekkonidae. Silliman University Natural Sciences Monograph Series Number 1. Silliman University, Dumaguete City, Philippines. 146pp. BROWN, W. C, AND A. C. ALCALA. 1980. Philippine Lizards of the Family Scincidae. Silliman University Natural Sciences Monograph Series Number 2. Silliman University, Dumaguete City, Philippines. 264pp. BROWN, W. H. 1919. Vegetation of Philippine Mountains. Manila Bureau of Printing, Manila. 434pp. CENSUS OFFICE OF THE PHILIPPINE ISLANDS. 1920. Census of the Philippine Islands, 1918. Vol. I, Part 1. Geography, History, and Climatology. Manila Bureau of Printing, Manila. 630pp. DICKERSON, R. E., ED. 1928. Distribution of Life in the Philippines. Monograph of the Bureau of Science, Manila, Number 21. 322pp. GIBBONS, J. W., AND R. D. SEMLITSCH. 1981. Terrestrial drift fences with pitfall traps: An effective technique for quantitative sampling of animal populations. Brimleyana7:l-16. INGER, R. F. 1983. Morphological and ecological variation in the flying lizards (genus Draco). Fieldiana: Zoology, New Series 18:1-35. KREBS, C. J. 1989. Ecological Methodology. December 1993 Asiatic Herpetological Research Vol. 5 p. 95 Harper & Row, New York. 654pp. MCCOY, M. 1980. Reptiles of the Solomon Islands. Wau Ecology Institute, Handbook Number 7. Wau, Papua New Guinea. 80pp. MITCHELL, L. A. 1985. Comments on the maintenance and reproduction of Hydro saurus pustulosus at the Dallas Zoo. Pp. 185-186. In S. McKeown, F. Caporaso, and K. H. Peterson (eds.). Proceedings of the Ninth International Herpetological Symposium on Captive Propagation and Husbandry. International Herpetological Symposium, San Diego, California. 265 pp. TAYLOR, E. H. 1922A. The Lizards of the Philippine Islands. Manila Bureau of Printing, Manila. 269pp. TAYLOR, E. H. 1922B. Additions to the herpetological fauna of the Philippine Islands, I. Philippine Journal of Science 21:161-206. TAYLOR, E. H. 1922C. Additions to the herpetological fauna of the Philippine Islands, II. Philippine Journal of Science 21:257-303. TAYLOR, E. H. 1923. Additions to the herpetological fauna of the Philippine Islands, III. Philippine Journal of Science 22:515-555. RICHARDS, P. W. 1952. The Tropical Rain Forest: An Ecological Study, tlniversity Press, Cambridge. 450 pp. TAYLOR, E. H. 1925. Additions to the herpetological fauna of the Philippines, IV. Philippine Journal of Science 26:97-111. SMITH, B. E. 1985. A systematic survey of the squamate reptiles of eastern Mindanao, Philippine Islands, with notes on ecomorphology. M. S. Thesis, Louisiana State University. 78pp. WILLMOTT, C. J., J. R. MATHER, AND C. M. ROWE. 1981. Average monthly and annual surface air temperature and precipitation data for the world. Part 1. The eastern hemisphere. Publications in Climatology 34:1-395. Vol. 5, pp. 96-102 Asiatic Herpetological Research December 1993 Notes on a Collection of Squamate Reptiles from Eastern Mindanao, Philippine Islands Part 2: Serpentes Brian E. Smith Museum of Natural History. Louisiana State University, Baton Rouge. Louisiana 70803-3216. Present Address: Department of Biology. University of Texas at Arlington. Uta Box 19498, Arlington. Texas 76019-0498 Abstract. -A systematic collection of the herpetofauna occurring in the Diuata Range of eastern Mindanao, Philippine Islands, with incidental collections from the Mount Apo Range, was made from April-September 1982. This paper completes taxonomic and ecological notes that were taken on the squamates of this region. Although conclusions about the effect of habitat alteration on snake populations are necessarily tentative due to sampling difficulties, some comments on apparent and potential shifts in population size and habitat use of the more common snake species are made. Key words: Reptilia, Squamata, Serpentes, Philippines, taxonomy, ecology Introduction Except for reviews by Taylor (1922) and Leviton (1959) and biogeographic works by Brown and Alcala (1970), Leviton (1963a, 1970) and Wiister and Thorpe (1989, 1990, 1991a), very little is known about the ecology and distribution of Philippine snakes. This paper describes ecological and distributional data taken on the snakes of an area in eastern Mindanao, Philippine Islands, that I visited from April- September 1982. An earlier paper on the lizards of this area describes the sites and methodologies I used during this study. Scalation data taken on snakes included dorsal scale rows at ten ventral scutes posterior to the head, mid-body, and ten ventral scutes anterior to the anal plate; ventral scute counts; and subcaudal scale counts. This data is given only when it adds to data previously published. Species Accounts Family Colubridae Ahaetulla prasina preocularis. — Two specimens were taken, one lacking specific data that was probably collected at or above the upper elevational limit of 800 m given by Leviton (1967). This species has a morphology typical of arboreal snakes, but is frequently taken on the ground, where it may descend to forage on Mabuya skinks (Leviton 1967). If Mabuya is a favored food item, then it is highly likely that A. prasina favors open areas, as does Mabuya. A female was taken on April 6 on a road in early second-growth habitat near site 1. Five oviducal eggs (22.8-31.1 mm in length) were found in this specimen. Specimens examined: LSUMZ 41804- 41805. Boiga cynodon. — This snake is arboreal by morphology, but two specimens were captured in drift fences and one found dead on a road at site 1 . They were all collected or killed at night, and were found in both early second-growth and primary forest habitats. Prior studies give no elevational information (Alcala, 1986; Leviton, 1968a; Smith, 1943; Taylor, 1922); the present specimens were found at 450-650 m. This species is aggressive and has very large palatine teeth which are quite effective in defense. Taylor (1922) stated that this species is rare in the Philippines, but it was the most commonly captured Boiga species at my study sites. As Leviton (1968a) noted, B. cynodon has an arboreal morphology and diet (birds and bird eggs), but all specimens were captured on the ground. Specimens examined: LSUMZ 41814- 41816. Boiga dendrophila latifasciata. — This specimen was taken on a road at night at about 500 m elevation in early second- growth forest. Previous reports indicate 1993 by Asiatic Herpetological Research December 1993 Asiatic Herpetological Research Vol. 5 p. 97 that this species is a common inhabitant of lowland swampy areas (Taylor, 1922; 1965; Tweedie, 1983), but the present specimen was found in hilly country far from such habitat. The stomach of this specimen contained bird feathers. Previous studies found this species to also eat bats and lizards (Alcala, 1986; Leviton, 1968a; Taylor, 1922). Specimen examined: LSUMZ 41812. Calamaria gervaisi. — This burrowing species was common in primary forest, but was rarely found in early second-growth forest. One specimen was collected 20 cm underground when digging a pit-can hole. Four adults were taken in a single pit-can during four days in April. The significance of this aggregration is unknown. A female collected April 10 contained one egg 3.4 mm in length. Segmented worms were found in the stomachs of two specimens. Specimens examined: LSUMZ 41769- 41778. Cyclocorus nuchalis taylori:. — One specimen was found in leaf litter in the primary forest, the other was caught in a drift fence in early second-growth forest. One specimen had eaten a juvenile Mabuya multicarinata. Leviton (1965) found specimens of the skink genera Mabuya and Sphenomorphus in stomachs of the C. nuchalis he examined. Specimens examined: LSUMZ 41796- 41797. Dendrelaphis caudolineatus terrificus. — Leviton (1968b) gave an elevational range of 0-35 m for this species, but I collected specimens of this snake from 100-1000 m. Also, I did not note any particular association with water, as did Alcala (1986). Although thought to be primarily arboreal (Leviton, 1968b), all the specimens I collected were taken on the ground. One specimen had eaten a terrestrial skink, Sphenomorphus coxi. A specimen examined by Leviton (1968b) also contained a terrestrial skink, Mabuya species. One specimen that I collected was taken in grassland in an extremely large ca. 100 square km) clear-cut area. This species is arboreal in its morphology, but these data indicate common use of the terrestrial microhabitat. Specimens examined: LSUMZ 41798- 41800. Dryphiops philippina. — This specimen was collected by a native near Mount Talomo and is the first specimen of this species collected on Mindanao. The ventral scute count of this specimen is 172, increasing the known variation of 177-188 reported by Leviton (1964a). Previous specimens were taken near sea level (Leviton, 1964a), but this specimen may have been collected as high as ca. 1000 m. This species was previously known only from Luzon, Negros, and Sibuyan (Leviton 1964a). Subsequent investigations should discover it on other large islands in the Philippine archipelago. Specimen examined: LSUMZ 41790. Elaphe erythrura erythrura. — This species is a common ground-dwelling diurnal inhabitant of all forest situations. It feeds on lizards, birds, and mammals (Alcala, 1986; Leviton, 1977; Taylor, 1922). My specimens commonly had mammals or mammal remains in their stomachs, and I consider this species to be a typical, heavy-bodied, mammal-eating constrictor that opportunistically takes other prey. It is also a common food item of the Philippine Eagle (Pithecophaga jefferyi) and the Philippine Serpent-Eagle (Spilornis holospilus). Alcala (1986) reports an altitudinal range to 500 m, but I took specimens up to 1065 m. I collected young snakes August 1 and September 8 (two snakes). These specimens measured 465, 383, and 409 SVL, respectively. The two largest of this group were identified as young males. Leviton (1977) stated that year-old young probably measure 400 mm SVL. Specimens examined: LSUMZ 41807- 41811. Vol. 5 p. 98 Asiatic Herpetological Research December 1993 Oligodon maculatus. — This is only the fifth known specimen of this species. Taylor (1922) took two specimens beneath sod and trash piles at Bunawan, Agusan, in the Agusan Valley of Mindanao. A third specimen was taken in northern Surigao Province (Taylor, 1925). A fourth was collected on Mount Todaya in the Mount Apo Range (Leviton, 1962). My specimen was taken during daylight in the forest floor litter on Mount Talomo, Mount Apo Range, at about 1000 m elevation. Alcala (1986) reports two specimens from 400 and 850 m elevation, but gives no further information. Including my data, an altitudinal range of 400-1000 m is indicated. This species is known only from eastern Mindanao. Leviton (1962) stated that this species has 17 dorsal scale rows throughout; my specimen reduces to 15 in the posterior third of the body. Leviton also noted that the loreal may be present or absent; in my specimen it is absent. This specimen also differs from the ones analyzed by Leviton in having one instead of two preoculars. My specimen also has fewer scutes than those examined by Leviton, increasing the known variability in ventral scute counts to 156-164. The specimen I collected is male. The stomach was empty. Specimen examined: LSUMZ 41806. Oxyrhabdium modestum. — T h e specimen taken was a gravid female with eight eggs measuring 18.9-26.1 mm in length. This specimen was collected on a road at about 400 m elevation in late second-growth forest. There is no further data to add to that given by Leviton (1964b). Specimen examined: LSUMZ 41803. Psammodynastes pulverulentus. — This is a common and aggressive rear-fanged colubrid of the forest floor litter. It is diurnal and was taken in all habitat types. Juveniles 190 and 197 mm SVL were collected May 3 and May 23 at site 1. I found insects, a lizard tail, and a snake {Calamaria gervaisi) in the stomachs of the specimens I examined. This species is primarily known as a lizard feeder, but frogs and snakes are also taken (Greene, 1989; Leviton, 1983). Greene (1989) did not report insects as a food item in the specimens he examined. Specimens examined: LSUMZ 41779- 41789. Rhabdophis auriculata auriculata. — This species is a common diurnal member of the leaf-litter herpetofauna in all habitat types investigated. I did not find it to be associated with water, as reported by Alcala (1986). These are small inoffensive snakes which usually attempt to conceal themselves rather than flee or bite when captured. I collected gravid females June 17 (three eggs, 9.9-10.5 mm in length) and 30 (one egg, 9.0 mm), and July 8 (three eggs, 5.3-7.6 mm) and 23 (two eggs, 9.0- 9.2 mm). Juveniles 135-235 mm SVL were collected April 7, 11, and 12, May 18 and 23, June 14, and July 8. The gradually increasing SVL of these specimens indicate that they could be the members of a single cohort. Leviton (1970) indicated two hatching seasons in the Mount Apo Range, during June- July and October-November. My observations within the Diuata Range indicate that the hatching season here could occur as early as March, but data are scanty. Specimens had eaten frogs and frog eggs, as also reported by Leviton (1970). Specimens examined: LSUMZ 41749- 41768. Stegonotus muelleri. — This is the eighth known specimen of this species, and only the second taken on Mindanao (Leviton 1959). Virtually no ecological data are available. This specimen was found dead on a logging road in early second-growth forest at about 590 m elevation. Leviton (1959) found three adult Rana limnocharis in the stomach of a large adult; the present specimen's stomach was empty. The specimen I collected is male. Its ventral scute count is higher than that given for males by Leviton (1959), thereby increasing the range of this measurement to 217-236. December 1993 Asiatic Herpetological Research Vol. 5 p. 99 Specimen examined: LSUMZ 41 802. Tropidonophis dendrophiops dendrophiops. — Specimens were collected during the day in all habitats, but always near swift-flowing streams. The altitudinal range given by Alcala (1986) extends to 700 m; one individual I collected was taken at 900 m. Juveniles were collected April 18 (161 mm SVL), May 21 (242 mm SVL), and June 24 (308 mm SVL) at site 1. These could be members of a single cohort born in March or April. One specimen was collected in a pit-can eating a frog. Taylor (1922) also reported frogs as common food items of this species. Alcala (1986) refers to this species as Natrix dendrophiops. Malnate and Underwood (1988) have recently assigned this species to the genus Tropidonophis Specimens examined: LSUMZ 41791- 41795. Family Elapidae Maticora intestinalis philippina. — One specimen was taken in a pit-can at site 2 in late second-growth forest at 400 m elevation. This snake is thought to be rare in the Philippines (Alcala, 1986; Taylor, 1922). Leviton (1963b) gives no ecological information, and Taylor (1922) describes an apparent anti-predator display during which specimens exhibit aimless thrashing motions, similar to displays described for other New and Old World coral snakes (Greene, 1973). Despite Alcala's (1986) statement that this species is found in arboreal ferns as well as under rotten logs, I consider this species to be a typical semi-fossorial coral snake. This agrees with Taylor's (1922) observations. There are no data on food habits of this species; my specimen's stomach was empty. Specimen examined: LSUMZ 41817. Naja samarensis. — This is an alert diurnal species. It is quite common around habitations and early second-growth habitats. Contrary to Alcala (1986) and Taylor (1922), this species was never found in primary forest, and residents told me that they never saw individuals of this species in the forest. It is likely that the conversion of much of the Philippines into altered habitat has resulted in an increased abundance of this highly venomous snake. I saw or collected this species from sea level to 1000 m elevation. Despite its highly toxic venom (Minton 1967), the local residents believed that this snake brought luck, and individuals found under and around houses were invariably left alive. I actively sought out reports of envenomation resulting from bites of N. samarensis, but received no such reports, and was told by locals that bites of this species were seldom problematical. This is in contrast to conclusions reached by Reyes and Lamana (1955). This is a spitting cobra, and there is at least one report of an accurate strike in the eyes resulting in a great deal of pain but no serious after- effects (Van Wallach, personal communication). I handled many specimens, but never saw one exude any appreciable quantity of venom. This species is known to eat frogs, snakes (Calamaria gervaisi), and rodents (Gressitt, 1937; Leviton, 1964c; Taylor, 1922). A specimen I examined contained a small Bufo marinus (SVL 95mm), and Van Wallach reports (personal communication) only B. marinus in the stomachs of N. samarensis he collected in rice paddies on Mindanao near the city of Surigao. N. samarensis itself is eaten by the Philippine Eagle {Pirhecophagas jefferyi) and the Philippine Serpent-Eagle {Spilornis holospilus) Two juveniles were taken with an adult male from a hole in the ground at 1000 m on Mount Talomo. These specimens were brought to me by a local resident. It was not possible to ascertain whether there was any significance to this aggregation, but other cobras are known to guard both eggs and young (Campbell and Quinn, 1975; Tryon, 1979; Tweedie, 1983). Wiister and Thorpe (1991b) have recently elevated Naja naja samarensis to full species status, and I use this new species designation in this paper. Specimens examined: LSUMZ 41819- 41824. Vol. 5 p. 100 Asiatic Herpetological Research December 1993 Discussion Unlike lizards, snakes are rarely taken in such numbers that it is possible to gauge their relative abundance in any given habitat. Therefore, ecological conclusions based on studies such as this one are few and tentative. Most of these conclusions have already been reached in individual species accounts. The pace of destruction of primary forest in the Philippines has created a great deal of second-growth habitats of various types. At least one snake, Naja samarensis, has probably benefitted from this trend. Taylor (1922) observed this species in primary forest, but neither I nor any other biologists that I worked with, nor any local residents, ever told me of seeing this species in the primary forest. I consider it a common snake of second-growth and agricultural areas. Other species, such as Elaphe erythrura erythrura and Rhabdophis auriculata auriculata appear to be common in all habitats, although they have not undergone an apparent shift in habitat preference similar to N. samarensis. The only other very abundant snake, Calamaria gervaisi, may be adversely affected by deforestation, in that it was common in primary forest but not in second-growth habitats. Logging causes both soil compaction from the movement of heavy equipment and soil drying from increased insolation. It seems likely that these changes would adversely affect a burrowing animal such as C. gervaisi. Snake ecology suffers from the typical problems inherent to studies of higher level predators. Since they are frequently at the top of the food chain, snakes never seem to be very abundant. In addition, they are secretive by nature, making collection and observation difficult. If we are to understand the effect of habitat loss on snake populations, it will be necessary to use novel approaches which are uniquely designed to overcome these specific problems. Acknowledgments This is a portion of a thesis submitted to the graduate school of Louisiana State University in partial fulfillment of requirements for the Master of Science degree. I thank my major advisor. Dr. D. A. Rossman, and my committee members, Drs. J. V. Remsen, J. W. Fleeger, and W. J. Harman for their comments on my thesis. Drs. R. S. Kennedy and J. P. O'Neill helped to arrange funding for my trip. Members of the Philippine Eagle Conservation Project and local residents assisted in the collections. The Dallas Zoo provided an intellectually stimulating environment in which to finish final editing of this paper. Comments by Jim Murphy, E. D. Brodie, Jr., and J. A. Campbell improved the final draft. Literature Cited ALCALA, A. C. 1986. Guide to Philippine Flora and Fauna, Vol. X: Amphibians and Reptiles. Natural Resources Management Center, Ministry of Natural Resources and University of the Philippines, Manila. 195pp. BROWN, W. C, AND A. C. ALCALA. 1970. The zoogeography of the herpetofauna of the Philippine Islands, a fringing archipelago. Proceedings of the California Academy of Sciences, Fourth Series 38:105-130. Campbell, j. a., and h. r. quinn. 1975. Reproduction in a pair of Asiatic cobras, Naja naja (Serpentes, Elapidae). Journal of Herpetology 9:229-233. GREENE, H. W. 1973. Defensive tail display by snakes and amphisbaenians. Journal of Herpetology 7:143-161. GREENE, H. W. 1989. Defensive behavior and feeding biology of the Asian mock viper, Psammodynastes pulverulentus (Colubridae), a specialized predator on scincid lizards. Chinese Herpetological Research 2:21-32. GRESSITT, J. L. 1937. Note on a Philippine cobra. Copeia 1937:73. LEVITON, A. E. 1959. Systematics and zoogeography of Philippinesnakes. Ph. D. Dissertation, Stanford University. 865pp. December 1993 Asiatic Herpetological Research Vol. 5 p. 101 LEVITON, A. E. 1962. Contribution to a review of Philippine snakes, I: The snakes of the genus Oligodon. Philippine Journal of Science 91:459-484. LEVITON, A. E. 1977. Contributions to a review of Philippine snakes, XIII: The snakes of the genus Elaphe. Philippine Journal of Science 106:99-119. LEVITON, A. E. 1963a. Remarks on the zoogeography of Philippine terrestrial snakes. Proceedings of the California Academy of Sciences, Fourth Series 31:369-416. LEVITON, A. E. 1963b. Contributions to a review of Philippine snakes, III: The genera Maticora and Calliophis. Philippine Journal of Science 92:523-550. LEVITON, A. E. 1964a. Contributions to a review of Philippine snakes, IV: The genera Chrysopelea and Dryophiops. Philippine Journal of Science 93:131-145. LEVITON, A. E. 1964b. Contributions to a review of Philippine snakes, VI: The snakes of the genus Oxyrhabdium. Philippine Journal of Science 93:407-422. LEVITON, A. E. 1964c. Contributions to a review of Philippine snakes, VII: The snakes of the genera Naja and Ophiophagus. Philippine Journal of Science 93:531-550. LEVITON, A. E. 1965. Contributions to a review of Philippine snakes, IX: The snakes of the genus Cyclocorus. Philippine Journal of Science 94:519-533. LEVITON, A. E. 1967. Contributions to a review of Philippine snakes, X: The snakes of the genus Ahaetulla. Philippine Journal of Science 96:73-90. LEVITON, A. E. 1968a. Contributions to a review of Philippine snakes, XI: The snakes of the genus Boiga. Philippine Journal of Science 97:291-314. LEVITON, A. E. 1968b. Contributions to a review of Philippine snakes, XII: The Philippine snakes of the genus Dendrelaphis (Serpentes: Colubridae). Philippine Journal of Science 97:371-396. LEVITON, A. E. 1970. Description of a new subspecies of Rhabdophis auriculata in the Philippines, with comments on the zoogeography of Mindanao Island. Proceedings of the California Academy of Sciences, Fourth Series 38:347-362. LEVITON, A. E. 1983. Contributions to a review of Philippine snakes, XIV: The snakes of the genera Xenopeltis, Zaocys, Psammodynastes, and Myersophis. Philippine Journal of Science 112:195-223. MALNATE, E. V., AND G. UNDERWOOD. 1988. Australasian natricine snakes of the genus Tropidonophis. Proceedings of the Academy of Natural Sciences of Philadelphia 140:59-201. MINTON, S. A., JR. 1967. Paraspecific protection by elapid and sea snake antivenins. Toxicon 5:47-55. REYES, A. C, AND C. LAMANA. 1955. Snakebite mortality in the Philippines. Philippine Journal of Science 84:189-194. SMITH, M. A. 1943. The Fauna of British India: Reptilia and Amphibia. Volume III: Serpentes. Taylor and Francis, London. 583pp. TAYLOR, E. H. 1922. The Snakes of the Philippine Islands. Manila Bureau of Printing, Manila. 312pp. TAYLOR, E. H. 1925. Additions to the herpetological fauna of the Philippines, IV. Philippine Journal of Science 26:97-111. TAYLOR, E. H. 1965. The serpents of Thailand and adjacent waters. University of Kansas Science Bulletin 45:609-1096. TRYON, B. W. 1979. Reproduction in captive forest cobras, Naja melanoleuca (Serpentes, Elapidae). Journal of Herpetology 13:499-504. TWEEDIE, M. W. F. 1983. The Snakes of Malaya. Singapore National Printers, Ltd. Singapore. 167pp. WUSTER, W„ AND R. S. THORPE. 1989. Population affinities of the asiatic cobra (Naja naja) species complex in south-east Asia: Reliability and random resampling. Biological Journal of the Linnean Society 36:391-409. 424pp. WUSTER, W., AND R. S. THORPE. 1990. Systematics and biogeography of the Asiatic Vol. 5 p. 102 Asiatic Herpetological Research December 1993 cobra (Naja naja) species complex in the Philippine Islands. Pp. 333-344. In G. Peters and R. Hutterer (eds.), Vertebrates in the Tropics. Museum Alexander Koenig, Bonn. WiJSTER, W., AND R. S. THORPE. 1991b. Systematics of Asiatic cobras. SSAR/HL Annual Meeting. University Park, Pennsylvania. [Abstr]. WUSTER, W., AND R. S. THORPE. 1991a. Asiatic cobras: Systematics and snakebite. Experientia 47:205-209. December 1993 Asiatic Herpetological Research Vol. 5, pp. 103-104 First Records for Ophisaurus hard and Python molurus bivittatus from Jiangxi Province, China Changfu zhong Department of Biology. Jiangxi Medical College, Nanchang, Jiangxi. China. Abstract.- Ophisaurus harti and Python molurus bivittatus are reported for the first time from Jiangxi Province China. The measurements, characteristics, and distributions of these two species and subspecies are described in detail. Key Words: Reptilia, Lacertilia, Anguidae, Ophisaurus harti, Serpentes, Boidae, Python molurus bivittatus, China, distribution Ophisaurus harti Boulenger (Fig. 1) On October 22, 1980 a specimen of O. harti was caught by Weitao Ji and Songlin Cheng of the Wuyi Shan Natural Reserve at an altitude of 900 meters on Mt. Wuyi, Yanshan County, Jiangxi Province, China. The specimen is kept in the Department of Biology, Jiangxi Medical College, Nanchang. Measurements of specimen, in mm. Specimen Number 600 -t- > Sex male Head width 21 Snout to ear 28 Body width 19 Total length 279 Tail* length 166 *Regenerated Characteristics. — Body cylindrical and no vestiges of limbs externally. Head with large symmetrical shields; two shields in a line between the nasal and the azygous prefrontal; ear-opening minute, smaller than the nostril. Dorsal scales keeled, in 16 longitudinal rows and 99 transverse series (counted in the length of the lateral fold); ventrals smooth, in 10 longitudinal series. The tail is long and fragile, and regenerates quickly. The regenerated tail is shorter than the original one, and the regenerated scales are smaller than the original ones. Brown above, with 21 transverse blue marking; under parts whitish. Distribution. — Vietnam; China: Sichuan, Yunnan, Guizhou, Anhui, Jiangsu, Zhejiang, Jiangxi (Mt. Wuyi in FIG. 1. Ophisaurus harti. Jiangsu, Zhejiang, Jiangxi (Mt. Wuyi in Yanshan County), Hunan, Fujian, Taiwan, Guangxi. Python molurus bivittatus Schlegel (Fig. 2) A piece of skin of P. molurus bivittatus was collected in 1965 from the people of Mt. Daji in Quannan County by the author, and one living specimen was caught in a mountain stream in the countryside of the city of Longnan County by a fisherman with a fishing net. The specimen was purchased by Chunhuo Teng of Nanchang People's Park in 1979. Measurements of specimen, in mm. Specimen Number 1001 Sex unknown Head width 55 Body width 120 Total length 2936 Tail length 34 © 1993 by Asiatic Herpetological Research Vol. 5 p. 104 Asiatic Herpetological Research December 1993 FIG. 2. Python molurus bivittatus. Characteristics. — Size large, with vestiges of hind limbs externally. Head distinct from neck, with large symmetrical shields; rostral with a deep pit on either side; two internasals ; two pairs of prefrontals, the anterior pair is longer than the posterior one; frontal a little larger than the supraocular, divided longitudinally; parietal, loreal and temporal regions covered with irregular scales; supralabials 13, the first two deeply pitted, 6th and 7th separated from the eye by suboculars. Scales smooth, in 50 rows on neck, 75 rows on midbody and 38 rows before the vent; anal entire. Tail rather short. Ventrals and subcaudals were not counted. Light yellowish above, with a dorsal series of large, more or less subquadrangular dark gray, black-edged spots; flanks with smaller, rounded or irregularly-shaped spots of the same color. A lance-shaped mark on the top of the head and the neck; yellowish below, with a border of dark spots on the outermost row of the scales; tail below marbled with yellow and black. Distribution.— Asia, Indo- Australian Region; China: Yunnan, Jiangxi (Dajishan in Quannan County, Longnan County), Fujian, Guangdong, Hainan, Guangxi. References GRESSIT, J. L. 1941. Amphibians and reptiles from southeastern China. Philippine Science Journal 75(l):29-58. POPE, C. H. 1935. The Reptiles of China. Vol. 10 of Natural History of Cenual Asia , New York. 604 pp. 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. SMITH, M. A. 1943. The fauna of British India, including Ceylon, and Burma. Reptilia and Amphibia. Vol. Ill, Serpentes. Taylor and Francis, London. 583 pp. TIAN, W. S. and Y. M. JIANG (eds.). 1986. [Handbook for identification of Chinese amphibians and reptiles]. Science Press, Beijing. 164pp. (In Chinese). ZHONG, C. 1984. [A new record of Jiangxi Province]. Acta Herpetologica Sinica 3(2):20. (In Chinese, with English abstract). ZHONG, C. F. 1986. [Preliminary of survey of reptiles in the Jinggangshan Natural Reserve]. Jiangxi University Journal (Natural Science) 10(2):71-74. (In Chinese, with English abstract). ZHONG, C. F. 1990. [Survey of reptiles in Ruijin County, Jiangxi Province]. Pp. 236-238 In Ermi Zhao (ed.) From Water Onto Land. China Forestry Press, Beijing. (In Chinese, with English abstract). ZHONG, C. F. and G. F. WU. 1981. [A preliminary list and its geographical distribution of reptiles of Jiangxi Province]. Acta Herpetologica Sinica (old series) 5(16):99-110. (In Chinese, with English abstract). December 1993 Asiatic Herpetological Research Vol. 5, pp. 105-108 Karyotype Information on some Toad Agamas of the Phrynocephalus guttatus Species Group (Sauria, Agamidae) of the former USSR. VALANTINA V. MANILO AND MICHAEL L. GOLUBEV Institute of Zoology, Academy of Sciences, Kiev, Ukraine Abstract. -Karyotypes of several toad agamas of the Phrynocephalus guttatus species group (sensu law) were investigated in specimens from a variety of localities of the former USSR. Differences in the diakinetic stage of meiosis have been observed, permitting distinctions among three groups of species. The forms guttatus, moltschanovi, kushackewitschii, and alpherakii comprise Group I; P. guttatus salenskyi represents the second group; and P. versicolor hispida represents Group III. Key words: Reptilia, Sauria, Agamidae, Phrynocephalus guttatus, Kazakhstan, Middle Asia, Precaucasus, karyology. 40 . FIG. 1. Scheme of distribution of forms of P. guttatus species group of the former USSR fauna: la- P. g. guttatus; lb- P. g. moltschanovi; II- P. g. kushackewitschii; III- P. g. alpherakii; IV- P. g. salenskyi; V- P. versicolor hispida; VI- P. guttatus spp. (the numbering of populations is in accordance with the data in table 1). Introduction The first and only extensive karyological investigation of the agamid lizard genus Phrynocephalus Kaup is the work of Sokolovsky (Sokolovsky, 1974; 1977). Karyotype characteristics permitted the recognition of five groups. The "guttatus" group included two species, P. guttatus (Gmel.) and P. versicolor Str. These species have a diploid number of 46, all chromosomes are telocentric. The karyotypes could be divided into 12 pairs of macrochromosomes and 1 1 pairs of microchromosomes. Approximately 50% of the metaphase plates in P. guttatus contained satellite chromosomes on the first pair of chromosomes, but these were never observed in the P. versicolor karyotype. The specimens examined came from Daghestan (P. guttatus) and Central Gobi, Mongolia (P. versicolor) and were believed to represent the nominative forms of both species. The systematics of the P. guttatus group based on external morphological characteristics is extremely complicated and remains unclear. At various times the forms alpherakii Bedr., moltschanovi Nik., kushackewitschii Bedr., salenskyi Bedr., etc. have either been included in P. guttatus, sensu stricto or treated as related species. The forms bogdanowi Bedr., hispida Bedr., and paraskiwi Semenov et al. have been assigned to P. versicolor. (Bedriaga, 1909; Nikolsky, 1915; Terentjev and Chernov, 1949; Peters, © 1993 by Asiatic Herpetological Research Vol. 5 p. 106 Asiatic Herpetological Research December 1993 m \ i • w • i H HI lif) fl« AH Hi M fin no no ~» >• -- -- -* st V i ^ *% \v - ■ FIG. 2 The karyogramme of Phrynocephalus guttatus salenskyi. 1984; Semenov et al. 1987). Golubev (1989) suggested that P. guttatus and P. versicolor from Kazakhstan are conspecific. Karyotype details of the forms listed above have never before been examined. The purpose of this study is to determine whether karyotype information will aid in our understanding of the systematics and evolution of Phrynocephalus. Methods Between 1989-1991 we collected specimens of nearly all listed forms of both species of Phrynocephalus inhabiting the territory of the former USSR with the exception of P. v. bogdanowi from the extreme south of Tuva (Central Asia) and P. guttatus ssp. from Turkmenistan (Fig. 1 and Table 1). Chromosome samples were prepared from cellular suspension of bone marrow, blood, and testis. We used a smear method and a method known as "digging out" in conformity with procedures described by Ford and Hamerton (1956) and McGregor and Varley (1986) as partially modified by Manilo (1986). Chromosomal staining was FIG 3. Diakinetic stage of meiosis of five forms of P. guttatus species group, s. lato: I- all elements are ring- or stick-shaped; (guttatus, moltschanovi, kushackewitschii, alpherakii); II- one or two elements are cross-shaped (salenskyi); III- two and more (up to four) elements are cross-shaped (hispida). performed by Giemsa stain (2% solution) in 0.01 M sodium-phosphate buffer (pH 6.8) for 20-30 minutes. After washing in distilled water, the preparations were passed through alcohols and xylols (orthoxylol) and subsequently embedded in Canadian Balsam. In excess of 30 metaphase plates from each form were investigated using a Biolam 1-212 microscope. Metaphase plates of spermatogonial division, spermatocyte I (diakinesis) and spermatocyte II (metaphase II) bivalents were investigated in testis preparations. Chromosome morphology is described according to the classification proposed by Levari etal. (1964). Results and Discussion Our data support the findings of Sokolovsky (1974, 1977). The diploid number is uniformly 46 and the Fundamental Number (NF) is 46. In several forms {guttatus, moltschanovi, kushackewitschii) we noted satellite chromosomes on several plates; whereas in other forms (alpherakii, salenskyi) we saw no evidence of satellites. The revelation of this structure largely depends on the degree of spiralization. It is possible that satellites will be found in the latter forms with further investigation and more extensive material. December 1993 Asiatic Herpetological Research Vol. 5 p. 107 TABLE 1 . Localities, sample sizes, and taxa of Phrynocephalus guttatus s. lato populations collected and investigated in this study (numbering of populations is given in accordance with the data shown in Fig. 1. No Taxa Locality Sample size, sex 1 P. g. guttatus N. Transcaucasus: N. Daghestan: sands on right bank of Kuma River 2 male 2 P. g. moltschanovi N. Kysylkum in Karakalpakia: Kostruba Well 2 male 3 P. g. alpheraku E. Kazakhstan: Alma Ata District; Near Karakuldek 3 male 4 5 P. g. kushackewitschii P. g. kushackewitschii E. Kazakhstan: Taldy-Kurghan District: NW bank of Kapchagay Reservior Near Andreevka (left bank of Chyndjaly River) 3 male; 2 female 9 male 6 P. versicolor hispida E. Kazakhstan: Djungar Gate 4 male 7 8 P. g. salenskyi P. g. salenskyi E. Kazakhstan: Zaissan Depression: Left bank of Bukhtarma Reservior: Kkuludjunsky Sands Left bank of Black Irtysh near Karatal 3 male 3 male Sokolovsky described all chromosomes as telocentric. We cannot confirm this with confidence. Second arms are clearly visible on metaphasic plates with premetaphasic (elongated) chromosomes on several pairs of large elements. Such chromosomes could be aero- or even subtelocentric. The karyotype of salenskyi is an example (Fig. 2). This characteristic is not peculiar to any one form or group of forms of the guttatus group and cannot be used to distinguish a subordinate group. We also observed distinct peculiarities of chromosome morphology in meiosis in the diakinetic stage. The chromosome bivalents of the various taxa differ in the number of ring-shaped and cross-shaped bivalents. Based on this difference in pairing, it is possible that the taxa of the toad agamas of the "guttatus" group might be grouped in the following way; I: guttatus*, moltschanovi, kushackewitschii, alpherakii all diakinetic bivalents are ring or stick-shaped. II: salenskyi one or two of the elements are cross-shaped (Fig. 3), the remainder as in Group I. Ill: hispida from two to four cross-shaped elements (Fig. 3), the remainder as in Group I. This grouping by cross-shaped elements in diakinesis is a continuum. In this system, P. v. hispida is closer to P. g. salenskyi from the Zaissan Depression [sometimes attributed to P. versicolor (Paraskiv, 1953; Bannikov et al., 1977)] than to other forms of P. guttatus sensu stricto. However, it may be important that the toad agama from the Zaissan Depression occupies an intermediate position between Groups I and III. It is interesting to note the absence of chromosomal differences in the I-st. group, while its members, as mentioned above, are attributed by a number of authors to different species. The data may be interpreted to suggest uniformity in the species of the guttatus group from Kazakhstan, Middle Asia, and the Precaucasus (Golubev, 1989) as well as a close relationship between salenskyi from Zaissan Depression and hispida from Djungar Gate and northern Djungaria (Golubev, 1992). * On several testis preparations of the nominative form we observed a picture similar to that of the preparations in Group II. Literature Cited BANNIKOV, A. G., I. S. DAREVSKY, V. G. ISCHENKO, A. K. RUSTAMOV AND N. N. Vol. 5 p. 108 Asiatic Herpetological Research December 1993 SHCHERBAK. 1977. [Field guide of the USSR amphibians and reptiles]. Prosveschenje Publishing House, Moscow. 369 pp. (in Russian). BEDRIAGA, YA. 1909. Amphibien und Reptilien. Pp. 73-502. In: Wissenschaftliche Resultate der Reisen N. M. Przewalskijs durch Zentralasien. Zoologische Teil. Band 3. Part 1. Lacertilia. Sankt-Petersbourgh. (In Russian/German). FORD, C. E., AND J. L. HAMERTON. 1956. A colchicine hipotonic citrate squash sequence for mammal's chromosomes. Staining Technology 31:247-251. GOLUBEV, M. L. 1989. [Phrynocephalus guttatus (Gmel.) or P. versicolor Str. (Reptilia, Agamidae): which Phrynocephalus species occurs in Kazakhstan?]. Zoological News, Kiev (5):38-46. (In Russian). GOLUBEV, M. L. 1992. [Variegated toad agama Phrynocephalus versicolor (Reptilia: Agamidae) of the Djungar Gate (Eastern Kazakhstan) with notes on systematics of the species]. Zoological News, Kiev no. 2:31-38. (In Russian). LEVAN, A., K. FREDGA, AND A. A. SANDBERG. 1964. Nomenclature for centromeric position on chromosomes. Hereditas 52:201-220. MACGREGOR, H. C., AND J. M. VARLEY. 1986. [Working with animal chromosomes]. Mir Publishing House, Moscow. 272 pp. (In Russian). MANILO, V. V. 1986. [Karyotypes of gecko genera Alsophylax and Crossobamon]. Zoological News, Kiev (5):46-54. (In Russian). NIKOLSKY, A. M. 1915. [Fauna of Russia and adjacent countries. Reptiles. Vol.1. Chelonia and Sauria]. Imperial Academy of Sciences, Petrograd. 532 pp. (In Russian). PARASKIV, K. P. 1956. [Reptiles of Kazakhstan]. Kazakh Academy of Sciences Press, Alma Ata. 228 pp. (In Russian). PETERS, G. 1984. Die krotenkopfagamen Zentralasiens (Agamidae: Phrynocephalus). Mitteilungen aus dem Zoologischen Museum in Berlin. Akademie Verlag, Berlin 60(l):23-67. SEMENOV, D. V., Z. K. BRUSHKO, R. A. KUBYKIN, AND G. I. SHENBROT. 1987. [Taxonomic position and protective status of the round-headed lizard (Reptilia, Agamidae) in the territory of the USSR], Zoological Journal, Moscow 68(12);79-87. (In Russian). SOKOLOVSKY, V. V. 1974. [A comparative karyological study of lizards of the family Agamidae. 1. Chromosome complements of 8 species of the genus Phrynocephalus (Reptilia, Agamidae)]. Cytology, Moscow 16(7):920- 925. (In Russian). SOLOLOVSKY, V. V. 1977. [Systematic relations in the family Agamidae based on karyological data]. P. 195 In: Questions of herpetology abstracts of the report at the Fourth All Union Conference of Herpetology, Nauka, Leningrad. (In Russian). TERENTYEV, P. V. AND S. A. CHERNOV. 1949. [Guide to reptiles and amphibians]. Soviet Sciences Publishing, Moscow. 315 pp. (In Russian). December 1993 Asiatic Herpetological Research Vol. 5, pp. 109-111 A Karyosystematic Study of the Plate Tailed Geckos of the Genus Teratoscincus (Sauria, Gekkonidae) VALENTINA v. manilo The Schmalhausen Institute of Zoology. Academy of Science of the Ukraine, Kiev, Ukraine Abstract. -Karyotypes of two subspecies of Teratoscincus scincus are described (T. s. scincus and T. s. rustamowi ). Both have 2n=36, and 46 arms in the karyotype (N. F. =46). A minor difference in centromere position in two of the smaller chromosome pairs was noted, and may characterize the respective subspecies sampled, or may reflect intra-population variation or even error in preparation. The karyotypes differ from an earlier published description for T. scincus, in which de Smet (1981) reported 2n=34, N. F. =42. Whether this represents intraspecific variation, error in preparation or interpretation, or a suggestion that the name T. scincus is being applied to more than a single species will only be resolved with further systematic study of this gekkonid lizard. Key words: Reptilia, Sauria, Gekkonidae, Teratoscincus scincus, Kazakhstan, Tadjikistan, karyology. TABLE 1. Karyotypic data for Teratoscincus scincus. Legends: M- macrochromosome, m- microchromosome, v- metacentric, sT- subtelocentric, a (A)- acrocentric, NF- basic number. Species, subspecies Chromosomal formula 2n NF Author T. scincus T. s. scincus T. s. rustamowi 4sT+4v+26A 24M(6sT+ 1 8 A)+ 1 2m(4v+8a) 24M(6sT+ 1 8 A+ 1 2m(4v+8a) 34 42 de Smet 36 46 our data 36 46 our data Introduction The Central Asian Gekkonid genus Teratoscincus is comprised of four recognized species (Szczerhak and Golubev, 1986). Karyotype data are available only for the species Teratoscincus scincus (de Smet, 1981). T. scincus is presently divided into three subspecies: T s. scincus; T. s. rustamowi; and T s. keyserlingii. This paper provides karyotypic descriptions of two races of T. scincus. Methods A total of seven females and four males representing four populations from Turkmenistan (20 km north of Bami station; 50 km north of Bakhardok; 45 km north of Ashkabad; and the vicinity of Gyaurs) were studied. Also, a Kazakhstan population (the Chimkent Region, Syutkent Settlement) was sampled, as were three additional males of T. s. rustamowi from Tadjikistan (Leninabad Region, in the vicinity of Yakkatarak Settlement). Chromosomal samples were prepared from cellular suspensions of bone marrow, blood and testis by the smear method and by using the method of "digging out" as described in Ford and Hamerton (1956) and McGregor and Varley (1986), as modified in part by Manilo (1986). Cellular mitotic activity was increased by injections of phytohemagglutinine solution (0.02 ml/g body mass) and chorionic gonadotrophin (50 units/g body mass). Chromosome preparations were stained with Giemsa (2% solution) in a 0.01 M sodium-phosphate buffer (pH 6.8) for 20- 30 minutes. After washing in distilled water, the preparations were passed through alcohols and xylols (ortho-xylol) and subsequently embedded in Canadian balsam. An NU-2 microscope with a 100x10 magnification was used for microscopy and photomicrography. Chromosomes are described using the centromeric position to define morphology following the © 1993 by Asiatic Herpetological Research Vol. 5 p. 110 Asiatic Herpetological Research December 1993 AO »fl />fl °* ,Ml "M ri^ IS ^1 *• •• •• ,0 < « a • An oi *• •*_ *» "^ A #j ~ a «* •»• »— ~ ~ AA *»*v **• *-«• - - ** - |! i i < D FIG. 1. Teratoscincus scincus scincus. a- mitotic metaphase of a dividing cell of bone marrow; b- bivalents of diakinesis; c, d- karyotype of female and male, respectively; e- idiogram of the karyotype. classification suggested by Levan et al. (1964). Karyotype descriptions Teratoscincus scincus scincus (Schlegel, 1858). Type locality: The Hi River in "Turkestan" The diploid chromosome consists of 36 chromosomes. The karyotype is provisionally divisible into 24 macrochromosomes (M) and 12 microchromosomes (m), but there is no sharp demarcation between M and m. Chromosomes decrease in size gradually from largest (pair 1) to smallest (pair 18). Chromosome pairs 4, 7, and 9 appear subtelocentric; pairs 14 and 15 are metacentric; and the remaining pairs are FIG. 2. Teratoscincus scincus rustamowi. a- mitotic metaphase of a dividing blood cell; b, c- male karyotype; d- idiogram of the karyotype. acrocentric. The chromosomal formula could be stated: 2n=24 M (6 sT+18A) + 12m (4v+8a) = 36. The "fundamental number" (N. F.) is 46. Sex chromosomes are not evident. Male and female karyotypes do not appear to differ in chromosome number or morphology. (Fig. 1). In male meiosis, the number of bivalents at diakinesis is 18. The bivalents which correspond to macrochromosomes have a ring-like shape; the smaller bivalents (microchromosomes) have a rod-like shape (Fig. lb). Teratoscincus scincus rustamowi (Szczerbak, 1979) Type locality: Fergan Valley in the sands in the vicinity of Cokand and Kairakkum. As in T. s. scincus the diploid number is December 1993 Asiatic Herpetological Research Vol. 5 p. Ill 36 with a somewhat arbitrary division between 24 macrochromosomes and 12 microchromosomes; no obvious sex chromosome heteromorphism; and chromosome pairs 4, 7, and 9 are subtelocentric. One possible difference noted between the subspecies, however, is that pairs 13 and 15 (instead of 14 and 15) appear to be metacentric. Meiotic material was not studied (Fig. 2). Comparative analysis of karyotypes of the genus The karyotype of T. scincus was first described by de Smet (1981). He reported 2n=34 with the following formula: 4sT+4V+26A. The total number of arms (Fundamental Number, or N. F.) was 42. Subsequently Manilo and Pisanets (1984) obtained a different result (2n=36). This stimulated us to re-examine the group in a more detailed manner (Table 1). Our conclusion is that two of the three subspecies have very similar karyotypes (the karyotype of T. s. keyserlingii remains unknown). The present studies with relatively small sample sizes, do not permit us to determine whether the difference we describe between the two subspecies represents individual variation or a real difference. Because de Smet did not indicate the locality for his specimens, we cannot determine whether his description of 2n=34 and our description of 2n=36 represent a difference in diploid number among populations; variation within populations; or problems in preparation and description. Clearly this group of geckos should be studied more extensively. Literature Cited DE SMET, W. H. O. 1981. Description of the orsein stained karyotypes of 136 lizard species (Lacertilia, Reptilia) belonging to the families Teiidae, Scincidae, Lacertidae, Cordylidae and Varanidae (Austarchoglossa. Acta Zoologica et pathologica, Antverpiensia 76:407-420. FORD, C. E. AND J. L. HAMERTON. 1956. A colchicine, hypotonic citrate squash sequence for mammalian chromosomes. Stain Technology 31:247-251. LEV AN, A., K. FREDGA, AND A. A. SANDBERG. 1964. Nomenclature for centromeric position on chromosomes. Hereditas 52:201-202. MACGREGOR, H. C. AND J. M. VARLEY. 1986. Working with animal chromosomes. Mir Publishing House, Moscow. 272 pp. (In Russian). MANILO, V. V. 1986. Karyotypes of geckos of the genera Alsophylax and Crossobamon. Vestnik Zoologii 5:46-54. (In Russian). MANILO, V. V. AND YE. M. PISANETS. 1984. Karyotype of the plate-tailed gecko (Teratoscincus scincus) from the Turkmenia territory. Vestnik Zoologii 5:83-84. (In Russian). SZCZERBAK, N. N. 1979. A new subspecies of plate-tailed gecko (Teratoscincus scincus rustamowi ssp. n., Sauria, Reptilia) from Uzbekistan and systematics of the species. Protection of Turkmenistan Nature 1979:129- 138. SZCZERBAK, N. N. AND M. L. GOLUBEV. 1986. Geckos of the USSR fauna and of adjacent countries. Naukova dumka, Kiev. 231pp. (In Russian) Vol. 5, pp. 112-116 Asiatic Herpetological Research December 1993 Resting Metabolic Rate in Three Age-groups of Alligator sinensis Pei-chao Wang and Jiang-hua Zhang Department of Biology, East China Normal University, Shanghai 200062, People's Republic of China Abstract. -The resting metabolic rate in three age-groups of Alligator sinensis is influenced the temperature, season, body weight and other factors. Among these factors, the effect of body weight, temperature, season, which are greater than that of others. The relationship between the body weight and the resting metabolism is not consistent with the "third-quarter" surface area law. In the equation M=aW^, the value of "a" ranges from 0.009 to 0.028, and the value of "b" ranges from 0.522 to 0.591, so that b is approximately equal to 2/3 in empirical equation. Key Words: Reptilia, Crocodilia, Alligatoridae, Alligator sinensis, China, metabolism. Introduction Alligator sinensis is a kind of special and precious reptile. It is classed as a first grade protected form of wildlife in China. Much research on the adaptability of Alligator sinensis have been done in the fields of morphology, distribution, reproduction and so on (Chen, 1985; Chen and Wang, 1984). Zhang (1986; 1989) reported information about infant Alligator sinensis whose weights ranged from 35 to 50 grams. Because energy metabolism is a very important criterion for the adaptability of Alligator sinensis, the purpose of our work was to explore the regularity of daily gain in total energy from diet, with relation to digestibility and energy allocation and to explore the resting metabolic rate of Alligator sinensis. This paper deals with the resting metabolic rate of three age groups in Alligator sinensis. Materials and Methods The Alligator sinensis that we used were provided by the Shanghai Zoo. The total number of tested animals was 13. Five of them were bom in 1980, and their average body weight was 2.84 ± 0.053 (M±SD) kg at the beginning of the experiment in 1986 and 3.126±0.053 kg at the end of 1987. Four of them were born in 1981, and their average body weight was 1.5410.59 kg at the beginning of the experiment in 1986 and 2.13910.857 kg at the end of the experiment in 1987. Four of them were born 1982, and their average body weight A 30 A A A 25 - 1 20 - Ai> 15 O ** 0 A A *x A %° ■5* 10 • • • * * W * * x X 5 O - '••'• .. • 0 12 3 k 5 6 Dody weitht(Ke) FIG. 1. Relationship between RMR and body weight in Alligator sinensis. Open circle- May (20°C); solid triangle- July (28°C); x- September (24°C); *- October (20°C); solid circle- January (12'C). was 0.90010.284 kg at the beginning of the experiment and 1.23110.536 kg at the end of the experiment in 1987. The experiment began in May 1986 and ended in January 1987. These animals were reared in the three ponds according their age differences. These ponds were simple artificial environments and the size of each pond was 3x4 square meters. © 1993 by Asiatic Herpetological Research December 1993 Asiatic Herpetological Research Vol. 5 p. 113 Table 1 . The linear regression equations and the approximate surface area equations of Alligator sinensis. Months Ta=C Linear repression equations Approximate surface area equations May 20° log Y=log- 1.9 11 +0.568 logX M=0.012W°-568 May 25° log Y=log- 1.789+0.549 logX M=0.016W°-549 July 25° log Y=log- 1.584+0.579 logX M=0.026W°-579 July 28° log Y=log- 1.549+0.571 logX M=0.028W°-571 Sep. 24° logY=log- 1.750+0.542 logX M=0.018W°-542 Sep. 25° log Y=log- 1.730+0.550 logX M=0.019W°-550 Oct. 20 log Y=log- 1.882+0.591 logX M=0.013W°-S91 Oct. 25° log Y=log- 1.774+0.525 logX M=0.017W°-525 Jan. 12° logY=log-2.050+0.544 logX M=0.009W°-544 TABLE 2. T-test of regressional coefficient. Months May July September October January T=C t-value 20° 25" 25" 28" 24° 25° 20° 25° 12° 1201 1203 1072 637 302 1819 1043 1272 4599 Note: all t values are over t o.ooidf6=5.96 and t 0.ooidf5=6.86. In this paper the measure of the resting metabolic rate is in ml CVkg-'hr1 or ml 02/W° 56hr-!. The closed-system respirator meter of Wang et al. (1980) was used to measure the oxygen consumption of Alligator sinensis under two different temperatures. The first, 25°C, is the contrast temperature that comes from the adaptive temperature of Alligator mississippiensis reported by Coulson and Coulson (1986). The other is the seasonal temperature that is derived from the average temperature of each month in the last five years in Shanghai (Table 1). The ingestive food behavior in the animals was fasted to avoid its effect on metabolism during the measuring of oxygen consumption. Results and Discussions The relationship between the body weight and the resting metabolic rate (RMR) The relationship between the body weight and the resting metabolic rate of Alligator sinensis is summarized in Fig. 1. The resting metabolic rate to unit weight declines with the raise of individual weight in each month or under each temperature. That is, there is a negative correlation between the body weight and the weight- specific resting metabolic rate of Alligator sinensis that corresponds with the surface area law. The further analysis of the correlation between the weight and the Vol. 5 p. 1 14 Asiatic Herpetologica / Research December 1993 TABLE 3. A comparison of RMR in three age groups of Alligutoi ■ sinensis from 1986-1987. Age groups RMR May 20°C 1986 July 1986 25°C 25°C 28°C Sept. 24°C 1986 25°C Oct. 20°C 1986 25°C Jan. 1987 12°C 10.28 16.19 17.54 10.51 11.10 8.80 10.31 1.07 1.15 1.24 0.71 0.79 0.62 0.75 100 100 100 100 100 100 100 mlO^g"1!!-1 Mean 8.52 S. E. 0.77 A group % 100 Born 1980 ml 02/W°-56rr l Mean 13.26 16.39 27.98 28.87 17.44 18.40 14.12 16.84 S. E. 0.39 0.46 1.85 0.66 0.32 0.37 0.34 0.91 % 100 100 100 100 100 100 100 100 5.44 0.36 100 8.76 0.09 100 mlO^g-1!!"1 Mean 10.98 13.89 21.08 21.85 13.09 13.84 9.87 12.19 6.49 S. E. 0.64 0.72 2.35 2.76 1.33 1.45 0.79 1.51 0.66 B group % 128.9 135.1 130.2 124.6 124.5 124.7 100.8 118.2 119.3 Born 1981 ml O2/^-56!!'1 Mean 12.61 16.55 26.41 27.23 17.73 18.72 13.53 16.55 8.63 S. E. 0.53 0.35 0.42 0.52 0.13 0.23 0.22 0.62 0.67 % 95.1 101.0 94.5 94.3 101.7 101.7 95.8 98.3 98.5 mlO^g"1!!-1 Mean 13.87 18.62 25.08 27.42 15.18 16.52 11.61 14.52 8.47 S. E. 1.16 1.42 1.84 2.33 0.76 1.46 0.73 1.14 0.79 C group % 162.8 181.1 154.9 156.3 144.4 148.8 131.9 140.8 155.7 Born 1982 ml O^0-56^1 Mean 12.28 15.96 26.79 29.19 17.00 18.31 12.87 15.88 8.87 S. E. 0.50 0.49 1.44 0.80 0.61 0.04 0.76 0.14 0.20 % 92.6 97.4 95.8 101.1 97.5 99.5 91.2 94.3 101.3 resting metabolic rate of Alligator sinensis begin by converting or "transforming" observed values to their logarithms and the linear regression equation and the approximate surface area equation (Table 1) which are formed on the base of their logarithms according to the methods of Avery (1979). In equation M=aW"b from Table 1, the value "a" ranges from 0.009 to 0.028, the value "b" ranges from 0.522 to 0.591. The significance of coefficients on the linear regression equations in Table 1 are also examined through the T-test (T=b/sb) and the results are shown in the Table 2. All values of "t" are larger than to.ooidf5=6.86 and to.ooidf5=5.96 (Table 2), so the values of p are less than 0.001. We may be to deduce that the body weight has a great effect on the metabolic rate, and has a similar effect in other crocodilians (Coulson and Hernandez, 1983). December 1993 Asiatic Herpetological Research Vol. 5 p. 115 30 25 4= s 15 7 Months FIG. 2. Seasonal influence on the RMR of Alligator sinensis maintained at 25°C. Solid circle- 1980 age group; x- 1981 age group; open circle- 1982 age group. Effects of temperature and season Table 3 shows that the resting metabolic rate of Alligator sinensis affected by the temperature and season. In the cool season or under the low temperature, the resting metabolic rate is at a lower level, and vice versa. This is similar to that of the other reptiles (Coulson and Coulson, 1986; Huggins et al., 1971; Wang and Lu, 1986; Wang and Xu, 1987; Wang et al., 1983, 1988). This metabolic character is due to the result of acclimatization of seasonal temperature rhythm in evolution of animals. We further analyze the relationship between RMR (resting metabolic rate) and temperature as well as season. We used 0.56 power of body weight to adjust all values of observing, so that the effect of body weight in RMR is eliminated. The results are summarized in Table 3. The values of ml O2/W0-56hr1 from Table 3 show the temperature and season effect on RMR. The levels of RMR are higher under high temperatures than low, and it is a similar state that there is a higher level of RMR during hot seasons than during cool seasons. From Fig. 2, is is further shown that the RMR changes with seasons under the same temperature of 25°C. In July, the RMR is the highest level; in September, the RMR becomes lower and it is continues to fall in October. This suggests that the energy consumption is relevant to the seasonal change which also corresponds with the rules of the energy consumption and requirements of Alligator sinensis. Among the growth months of Alligator sinensis as in July, there is a high water temperature, and there is intensive metabolism and rapid growth in Alligator sinensis. The data below support this statement. The group born in 1980 ingested daily 172.02±77.65 (M±SD) g fresh fish in June, 221.35±35.58 g fresh fish in July, 54.73±41.14 g fresh fish in September. The daily body weight growth of the group was also rapid, such as 10.3 g in June, 44.6 g in July and 15.8 g in September. In May, Alligator sinensis had just awaked from hibernation, when the RMR was lower. In October, Alligator sinensis will stop the feeding when the RMR declined to a low level, and there were some changes in their physiological attributes for the coming hibernation stage. It is indicated that Alligator sinensis has a series of adaptive strategies for the seasonal changes. Tlie relationship between age and RMR The relationships between age and RMR in three age groups are shown in Table 3. There are two kinds of data on the RMR. The first RMR in Table 3 is influenced by body weight, and does not eliminate the effect of body weight. It is expressed in ml 02/kg1hr1. The second RMR eliminates the effect of body weight by expressing RMR in ml O2/W056hr1. A comparison on the level of both kinds of RMR in three age groups is based on 100 in RMR of age groups in 1980. The results of comparison suggest that the first kind of RMR falls as the age of groups increases. The second kind of RMR slightly falls in younger groups, except in a few months. Acknowledgments This study was supported by the Scientific Fund of the State Educational Committee, China, East China Normal Vol. 5 p. 116 Asiatic Herpetological Research December 1993 University and Shanghai Zoo. We thank Professor Ruyong Sun, Prof. Wenji Huang, Prof. Ermi Zhao for helpful discusssions and encouragement. We would also like to thank Dr. Ted Joanen, Dr. E. Norber Smith, and Dr. Shoji Tokunaga for providing some literature. Literature Cited AVERY, R. A. 1979. Lizards-- A study in thermoregulation. Printed by Lidio Ltd., East Rilbride Scotland. CHEN, B. H. AND C. L. WANG. 1984. Artificial reproduction of Alligator sinensis. Acta Herpetologica Sinica (new ser.) 3(2):49-54. (In Chinese, with English abstract). CHEN, B. H. 1985. Relationship between the metabolic rate and seasonal changes in activity of Alligator sinensis. Acta Herpetologica Sinica (new ser.) 4(3):173-176. (In Chinese, with English abstract). COULSON, R. A. AND T. D. COULSON. 1986. Effect of temperature on the rates of digestion, amino acid absorption and assimilation in the Alligator. Comparative Biochemistry and Physiology 83A(l):585-588. COULSON, R. A. AND T. HERNANDEZ. 1983. Alligator metabolism. Comparative Biochemistry and Physiology 74B(1): 1-182. HUGGINS, S. E., H. E. HOFF AND M. E. VALENTINUZZI. 1971. Oxygen consumption of small caimans under basal conditions. Physiological Zoology 44:40-47. WANG, P. C, S. ZHAO, H. J. LU, L. B. ZHU, AND Z. X. CHI. 1980. A simple closed-system respirometer for measuring the oxygen consumption of the terrestrial vertebrate. Journal of East China Normal University (Natural Science) 2:126-131. (In Chinese, with English abstract). WANG, P. C, K. C. QIAN, H. J. LU, L. B. ZHU, AND S. ZHAO. 1983. Studies on physiological ecology of Pallas' pit-viper. Acta Herpetologica Sinica (new ser.) 2(1): 19-32. (In Chinese, with English abstract). WANG, P. C. AND H. J. LU 1986. Heat metabolism and Uiermoregulation of Cope's rat snake. Acta Herpetologica Sinica (new ser.) 5(1): 10- 16. (In Chinese,with English abstract). WANG, P. C, AND H. F. XU. 1987. Influence of ambient temperature on body temperature and heat energy metabolism of Takydromus septentrionalis. Acta Herpetologica Sinica (new ser.) 6(2):10-15. (In Chinese, with English abstract). WANG, P. C, H. F. XU, W. MA, AND X. Jl. 1988. The influence of ambient temperature on the body temperature and energy metabolism in Chinemys reevesii. Acta Herpetologica Sinica (new ser.) 7(2):122-127. (In Chinese, with English abstract). December 1993 Asiatic Herpetological Research Vol. 5, pp. 117-126 Effects of Chinese Snake Venoms on Blood Coagulation, Purified Coagulation Factors and Synthetic Chromogenic Substrates YUN ZHANG1-2, YULIANG XlONG2 AND CASSIAN BON1 'Unite des Venins, Unite associee Instilul Pasteur/INSERM 285, 25, rue du Dr. Roiix, 75724 Paris Cedex 15 France * Kunming Institute of Zoology, Academia Sinica, Kunming 650107, Yunnan, China Abstract. -We examined the action of venoms from common Chinese Crotalidae and Elapidae snakes on blood coagulation mechanisms. Procoagulant effects were observed with venoms from Agkistrodon acutus, Trimeresurus stejnegeri, Ophiophagus luinnah and Bungarus fasciatus, the latter two only in the presence of Ca2+. After treatment with a serine protease inhibitor (phenylmethanesulfonyl fluoride, PMSF), Agkistrodon acutus venom lost its ability to clot purified fibrinogen but retained its capacity to clot human plasma in the absence of Ca2+. An anticoagulant action was obtained with venoms from Trimeresurus mucrosquamatus, Agkistrodon halys and Naja naja atra. This action was abolished after treatment with a specific inhibitor of PLA2 activity (p-bromophenacyl bromide, BrPBr), revealing a procoagulant action with high concentrations of treated venoms in the cases of Trimeresurus mucrosquamatus and Agkistrodon halys. The effects of these venoms on hemostasis have been further characterized by measuring their phospholipase A2 activity, their ability to hydrolyze synthetic chromogenic substrates and to activate purified blood coagulation factors (prodirombin, factor X, protein C and plasminogen). These venoms showed an amidolytic activity which was mainly due to serine proteases (90 to 95% of inhibition with PMSF). Combining the observations obtained with human plasma and purified blood coagulation factors, we concluded Uiat: i) six of the eight tested Chinese venoms {i.e.: Ophiophagus hannah, Bungarus fasciatus, Agkistrodon acutus, Trimeresurus mucrosquamatus, Trimeresurus stejnegeri and Naja naja atra) contain components which activate factor X in a Ca2+-dependent manner; ii) three venoms (Agkistrodon acutus, Agkistrodon halys and Trimeresurus stejnegeri) contain prothrombin activators; iii) Ophiophagus hannah venom has a weak protein C activating activity; and iv) Trimeresurus stejnegeri venom possesses plasminogen activating activity. In addition, several of these venoms have previously been shown to contain thrombin-like and fibrinogenolytic enzymes, anticoagulant phospholipases A2 (PLA2s) and/or non enzymatic anticoagulant components. Key Words: Snake venoms, blood coagulation, purified blood coagulation factors, chromogenic substrates. Introduction Snake venoms are known to be a rich source of hydrolytic enzymes, mainly proteases and phospholipases A 2 and of non-enzymatic proteins, which induce disorders of blood coagulation, hemorrhage and shock (Pirkle and Markland, 1988; Ouyang and Teng, 1972; Teng and Seegers, 1981). Many proteases acting on different steps of the blood coagulation cascade have been purified from snake venoms. They cleave blood coagulation factors, either in a specific or in a non- specific manner, and cause acceleration or retardation of blood coagulation (Pirkle and Markland, 1988). Some of these proteases, such as thrombin-like enzymes (Stocker and Meier, 1988) or protein C activators (Kisiel et al., 1987), are serine proteases which may be rapidly and irreversibly inactivated by alkylation with PMSF. Other procoagulant or anticoagulant proteases, like factor X or prothrombin activators from Bothrops atrox venom (Hofmann and Bon, 1987a; 1987b) or from Echis carinatus venom (Morita and Iwanaga, 1978) are insensitive to PMSF and have been postulated to be metalloenzymes. PLA2s have also been recognized for their anticoagulant activity, which has been attributed to their ability to antagonize the procoagulant action of negatively charged phospholipids (Ouyang et al., 1978). In order to better understand the pathophysiological action of snake venoms on haemostasis, and to examine the potential use of their procoagulant or anticoagulant components as pharmacological tools, we examined the © 1993 by Asiatic Herpetological Research Vol. 5 p. 118 Asiatic Herpetological Research December 1993 effects of various snake venoms on blood coagulation mechanisms in vitro, using human plasma, purified blood coagulation factors (fibrinogen, prothrombin, factor X, protein C and plasminogen), and synthetic chromogenic substrates. We examined in detail venoms from the eight most common venomous snakes in China; four belonging to the Elapidae family {Ophiophagus hannah, Naja naja atra, Bungarus fasciatus and Bungarus multicinctus) and the other four to the Crotalidae family (Trimeresurus mucrosquamatus, Trimeresurus stejnegeri, Agkistrodon halys and Agkistrodon acutus). Methods Venoms were supplied by the Kunming Institute of Zoology (Academia Sinica, China). The venoms were collected from snakes living in the southern provinces of China and stored desiccated. They were dissolved in 50 mM Tris-HCl buffer, pH 7.8, at a concentration of 1 mg-ml"1 and were used immediately. Bovine factor X, human prothrombin and human Glu-plasminogen were obtained from Sigma (St. Louis, MO, USA). Human protein C was obtained from Diagnostica Stago (Asnieres, France). Human fibrinogen (grade L) from Kabi Vitrum (Stockholm, Sweden) was treated with diisopropylfluorophosphate according to the instructions of the manufacturer, in order to irreversibly inactivate traces of thrombin or other blood coagulation factors. Platelet-poor human plasma was the supernatant of human blood mixed with 1/10 volume of 3.8% sodium citrate and centrifuged at 3000 rpm for 15 min. Pools of normal citrated plasma obtained from 5- 10 healthy donors were stored at -20°C. Chromogenic substrates H-D-Phe-Pip- Arg-pNA (S-2238), H-D-Val-Leu-Lys- pNA (S-2251), H-D-Val-Leu-Arg-pNA (S- 2266), H-D-Pro-Phe-Arg-pNA (S-2302) and Bz-Ile-Glu-Gly-Arg-pNA (S-2222) were obtained from Kabi Vitrum ^Stockholm, Sweden) and the chromogenic substrate H-D-Lys(Cbo)-Pro-Arg-pNA 365-25) was from Diagnostica Stago (Asnieres, France). Phenylmethanesulfonyl fluoride (PMSF) and /?-bromophenacyl bromide (BrPBr) were purchased from Sigma (St. Louis, MO, USA). All other reagents were of the highest purity available. Chemical modifications Inactivation of serine proteases by PMSF was performed in 50 mM Tris-HCl, pH 7.8. Venom samples (2 mg-ml"1) were incubated at 37°C for two hours with 5 mM PMSF (stock solution: 0.1 M in dimethylsulfoxid). Inactivation of PLA2s was carried out in the same buffer by incubating the venom (1 mg-ml"1) at 37°C for one hour with 2 mM BrPBr (stock solution: 0.1 M in acetone). Treated venom samples were then dialyzed for 4 to 8 hours against large volumes of the same buffer. Chromogenic assays Amidolytic activity was measured with a Kontron spectrophotometer in 1 cm path- length plastic cuvettes. Assays were performed in 500 ml of 50 mM Tris-HCl, pH 7.8, containing the appropriate chromogenic substrate (0.2 mM). The reactions was initiated by addition of the sample to be tested (5 mg-ml"1 to 100 mg-ml"1, final concentrations) and the formation of p-nitroanilide was monitored at 405 nm. The amount of substrate hydrolyzed was calculated using a molar extinction coefficient of 10,000 M^-cm"1 for free p-nitroanilide. Determination ofPLA2 activity PLA2 activity was determined by the titrimetric method described by Desnuelle et al. (1955), according to the procedure of Radvanyi and Bon (1982). Effects of the venoms on blood coagulation Citrated platelet-poor human plasma (200 ml) was incubated at 37°C for 1 min, then a 20 ml aliquot of diluted venom samples was added and clotting time was recorded. In some cases, 5 ml of CaCl2 December 1993 Asiatic Herpetological Research Vol. 5 p. 119 6- 5"" 4 - B. multicinctus i i i r" 0-1 1 1 1— — r .001.01 .1 1 10 100 .001 .01 / - B. fasciatus 6- 5-i 4- 3- 1 \- • \ \ 2- W" 1 - o- T I 10 100 .001 .01 10 100 .001 .01 10 100 Venom concentration ((ig/ml) Venom concentration (|!g/ml) FIGURE 1. Effects of Elapidcie snake venoms on blood coagulation. Citrated platelet-poor human plasma (200 ml) was incubated at 37°C for 1 min; dilutions of the sample to be tested (20 ml) were then added simultaneously with 5 ml of 0.45 M CaCl2 (10 mM final concentration); native venom ( 9), PMSF- treated venom (A) orp-bromophenacyl bromide-treated venom ( I). Values are the mean of triplicates. Vol. 5 p. 120 Asiatic Herpetological Research December 1993 TABLE 1. Phospholipase A2 activity of the venoms from the common Chinese venomous snakes (unmolmiir'mg'1). native venom treated venoms Agkistrodon acutus Agkistrodon halys Trimeresurus mucrosquamatiis Trimeresurus stejnegeri Ophiophagus harmah Bungarus fascia tus Bungarus multicinctus Naja naja 37±2 <1.0 220±9 <1.0 48±3 <1.0 50±6 <1.0 95±8 <1.0 160±11 <1.0 180+12 <1.0 35±4 <1.() The phospholipase A2 activity of the venoms were determined as described by Radvanyi and Bon (1982) using egg lecithin solubilized by sodium cholate. Venoms were treated by p-bromophenacyl bromide as indicated in the method then dialyzed to remove the excess of reagent. Each value is the mean of three determinations ± standard deviation. (10 mM final concentration) were added simultaneously with the venom samples. Thromhin-like activity was determined by measuring the clotting time of purified human fibrinogen (0.5%) in 50 mM Tris- HCl, pH 7.8, containing 0.1 M NaCl. Fibrinogen (200 ml) was incubated for 2 min at 37 °C before addition of 20 ml of diluted venom samples. Activation of blood coagulation factors by snake venoms Activation of prothrombin and factor X was performed as described by Hofmann and Bon (1987a; 1987b). Briefly, purified human prothrombin (50 mg-ml"1) was incubated at 37°C in 50 mM Tris-HCl, pH 7.8, containing 0.1 M NaCl and different concentrations of the samples to be tested. Aliquots (50 ml) were removed at various times and their amidolytic activity was tested in 500 ml of the same buffer containing S-2238 (0.2 mM). Purified bovine factor X (25 mg-ml"1) was incubated at 37°C in 50 mM Tris-HCl, pH 7.8, containing 0.1 M NaCl, 10 mM CaCl, and different concentrations of the samples to be tested. Aliquots (50 ml) were removed at various times and their amidolytic activity was immediately assayed in 500 ml of the same buffer containing S-2222 (0.2 mM). Protein C activation was assayed according to the method of Orthner et al. (1988), with minor modifications. Human protein C (5 mg-ml"1) was incubated at 37°C in 50 mM Tris-HCl, pH 7.8, containing 1 mg-ml"1 polyethylene glycol, 5 mM EDTA and dilutions of the samples to be tested. At various times, aliquots (50 ml) were taken to measure the amidolytic activity of activated protein C, in 500 ml of the same buffer containing CBS65-25 (0.3 mM). Plasminogen activation assay Human Glu-plasminogen (100 mg-ml"1) was incubated at 37°C in 200 ml of 50 mM Tris-HCl, pH 7.8, containing 0.1 M NaCl, 0.01% Tween-80, and different concentrations of the samples to be tested. Aliquots (50 ml) were taken at various times and assayed for plasmin activity. They were introduced into a plastic cuvette containing 450 ml of the same buffer December 1993 Asiatic Herpetological Research Vol. 5 p. 121 TABLE 2. Amidolytic activity (nmol-mirr'-mg"1) of venoms from Chinese snakes. Venom Sul bstrate S-2238 S-2251 S-2222 S-2302 S-2266 CB S65-25 A. acutus Native 230 <1 20 110 50 550 Treated <1 <1 <1 <1 <1 <1 A. halys Native 30 50 70 360 290 260 Treated <1 <1 <1 10 40 30 T. mucrosquamatus Native 2200 110 20 1500 2900 1600 Treated 40 <1 <1 130 60 70 T. stejnegeri Native 800 120 30 2400 950 1400 Treated 10 <1 <1 240 10 10 0. hannah Native 10 <1 10 60 50 70 Treated <1 <1 <1 <1 <1 <1 B. fasciatus Native <1 <1 <1 <1 <1 <1 B. multicinctus Native <1 <1 <1 <1 <1 <1 N. naja atra Native <1 <1 <1 <1 <1 <1 The amidolytic activities of each venom were determined as described in Methods, with the indicated substrates. Venoms were treated by PMSF as indicated in Methods, then dialyzed to remove the excess of reagent. Indicated values are the means of three determinations (standard errors were less than 10%). supplemented with S-2251 (0.3 mM) and the formation of p-nitroanilide was monitored at 405 nm. Results Procoagulant and anticoagulant properties of the venoms The procoagulant and anticoagulant actions of the venoms from the eight most common venomous snakes in China (Ophiophagus hannah, Naja naja atra, Bungarus fasciatus, Bungarus multicinctus, Trimeresurus stejnegeri, Trimeresurus mucrosquamatus, Agkistrodon halys and Agkistrodon acutus) were examined with human plasma in the presence and in the absence of calcium ions. Each venom was tested in its native form, after treatment with a specific and irreversible inhibitor of serine proteases (PMSF), or after treatment with a specific and irreversible inhibitor of PLA2s (BrPBr) (Volwerk et al., 1974). As indicated in Figure 1, Bungarus multicinctus venom did not modify blood coagulation in vitro. Venoms from Bungarus fasciatus and from Ophiophagus hannah showed a procoagulant action, dependent on the presence of Ca2+. They were unable to clot purified fibrinogen (result not shown), indicating the absence of thrombin-like enzymes. Their procoagulant effect might therefore result either from an ability to convert prothrombin into thrombin in a calcium- dependent manner, or more probably, from a direct or indirect activation of factor X into factor Xa. The fact that a treatment of these venoms with PMSF significantly but not completely reduced their procoagulant action (Figure 1) suggests that this effect is due, at least in part, to serine protease(s). The venom from Naja naja atra was characterized by a pronounced anticoagulant action (Figure 1). It did not prevent clotting of purified human Vol. 5 p. 122 Asiatic Herpetological Research December 1993 fibrinogen in the presence of thrombin (result not shown), indicating that the anticoagulant action is not due to fibrinogenolysis. The anticoagulant effect of Naja naja atra venom was completely and irreversibly prevented by treatment with BrPBr (Figure 1) which inhibited PLA2 activity of the venom (Table 1). This activity is therefore due to one or several anticoagulant PLA2s, as reported in the case of many other snake venoms. The effects of venoms from Chinese Crotalidae snakes on blood coagulation mechanisms (Figure 2) appeared much more complex than those observed with Elapidae venoms (Figure 1). The venoms from Agkistrodon acutus and from Trimeresurus stejnegeri were characterized by a procoagulant action, observed both in the presence and in the absence of Ca2+ (Figure 2). These venoms also clotted purified fibrinogen in a calcium- independent manner (result not shown), indicating that they contain potent thrombin-like enzymes, in agreement with the observations reported by Ouyang e t al. (1971) and by Liu and Xiong (1990). Treatment of Agkistrodon acutus venom with PMSF completely abolished its ability to clot purified fibrinogen (result not shown) and strongly reduced its procoagulant action (Figure 2), as expected since this effect is mainly due to thrombin- like serine proteases. However, PMSF- treated Agkistrodon acutus venom showed a complex action, an anticoagulant activity at 10 mg-ml"1, and a clotting activity at higher concentrations (Figure 2), suggesting that it contains other procoagulant components, in addition to thrombin-like enzymes. The anticoagulant effect of PMSF-treated Agkistrodon acutus venom is consistent with the presence of an anticoagulant protein of 26 kD, which is devoid of enzymatic activity and which prevents the formation of thrombin by binding to the prothrombin activation complex (Teng and Seegers, 1981). In addition to this non-enzymatic component, Agkistrodon acutus venom might contain anticoagulant PLA2s, since treatment with BrPBr (Table 1) significantly increased its procoagulant effect (Figure 2). Agkistrodon halys and Trimeresurus mucrosquamatus venoms were characterized by an anticoagulant action which was completely abolished after treatment with BrPBr, but not with PMSF (Figure 2). This indicates that they contain potent anticoagulant PLA2s, as previously reported for Trimeresurus mucrosquamatus (Ouyang et al., 1978) and for Agkistrodon halys (Chen et al., 1987). In fact, treatment of Agkistrodon halys and Trimeresurus mucrosquamatus venoms with BrPBr suppressed the anticoagulant activity of these venoms and revealed a weak procoagulant activity (Figure 2), indicating the presence of procoagulant components. Further, this procoagulant action was observed only in the presence of Ca2+, suggesting that the procoagulant components are able to convert prothrombin into thrombin or to activate factor X. High concentrations (100 mg-ml" ') of Agkistrodon halys venom were able to clot purified fibrinogen (result not shown). This may be explained by the presence of a thrombin-like enzyme, which has been purified (Guan et al., 1988) but the level of this enzyme in the venom is low and its activity is weak. Action of the venoms on purified blood coagulation factors We determined me amidolytic activity of the venoms on a number of chromogenic substrates, classically used for assaying blood coagulation factors: thrombin (S- 2238), factor Xa (S-2222), activated protein C (CBS65-25), kallikrein (S-2302 and S-2266) and plasmin (S-2251). Assays were performed with native, PMSF-treated and BrPBr-treated venoms (Table 2). The venoms from Elapidae snakes did not present detectable activities, with the exception of the venom from Ophiophagus hannah which hydrolysed several substrates (Table 2). The venoms from Crotalidae snakes exhibited significant activities towards most chromogenic substrates, but with important species differences: the venoms from Trimeresurus snakes presented much higher activities than those of Agkistrodon snakes and the venom from Agkistrodon December 1993 Asiatic Herpetological Research Vol. 5 p. 123 c E E c o o .001 .01 1 00 .001 .01 1 00 10 i .001 .01 100 ,001 .01 1 00 Venom concentration ()jg/ml) Venom concentration ((ig/ml) Figure 2. Effects of Crotalidae snake venoms on blood coagulation. Citrated platelet-poor human plasma (200 ml) was incubated at 37°C for 1 min then the sample to be tested (20 ml) was added simultaneously with calcium (5 ml of 0.45 M CaCl2', 10 mM final concentration; closed symbols) or without calcium (open symbols and dashed lines). Native venom (9; O); PMSF-treated venom (A; /\or p-bromophenacyl bromide-treated venom ( I: D). The values are the mean of triplicates, standard errors being 10% of the values. Vol. 5 p. 124 Asiatic Herpetological Research December 1993 TABLE 3: Activation of factor X, prothrombin, protein C and plasminogen by the venoms from common Chinese snakes. Activation activity (Arb. Unit) i Factor X Prothrombin Protein C Plasminogen A. acutus 0.67 1.3 0 0 PMSF treated 0.59 1.0 ND ND A. halys 0 3.0 0 0 PMSF treated ND 2.9 ND ND T. mucrosquamatus 0.66 0 0 0 PMSF treated 0.58 ND ND ND T. stejnegeri 0.27 0.16 0 0.42 PMSF treated ND ND ND 0 Ophiophagus hannah 9.7 0 0.44 0 PMSF treated 4.2 ND ND ND B. fasciatus 13.3 0 0 0 PMSF treated 5.8 ND ND ND B. multicinctus 0 0 0 0 PMSF treated ND ND ND ND N. naja 0.22 0 0 0 PMSF treated ND ND ND ND The activation of the indicated blood coagulation factors was determined as described in Materials and Methods by measuring the amidolytic activity of the factors after activation. Each value is the mean of at least three independent determination, standard errors being less than ±10%. ND means not determined. The results were expressed as (lie A O.D./min at 405 nm acquired in analysed solution divide incubation time and divide venom concentration in activation solution. acutus was 10 times more active on substrate S-2238 than that from Agkistrodon halys; in contrast substrate S- 2251 was hydrolysed by the venom of Agkistrodon halys but not by the venom of Agkistrodon acutus. These results are in agreement with the general concept that the proteolytic activities of the venoms from Elapidae snakes are much lower than those of Crotalidae snakes. The procoagulant action of the various venompurified bovine factor X, human prothrombin, protein C and plasminogen, and measuring their amidolytic activity after activation (Table 3). Except for Agkistrodon halys and Bun gar us multicinctus, all venoms were able to activate factor X, those from Ophiophagus hannah and Bungarus fasciatus being far more active than the others. Furthermore the ability of Bungarus fasciatus venom to activate factor X was markedly reduced after treatment with PMSF (Table 3), suggesting that the venom components responsible for this activity may be serine proteases. It should however be noticed that neither Ophiophagus hannah nor Bungarus fasciatus venoms showed any activity on prothrombin or plasminogen, emphasizing their rather specific action on factor X. On the other hand, three venoms (Agkistrodon acutus, Agkistrodon halys and Trimeresurus stejnegeri) among the eight which have been tested, possessed componentss was further examined in detail, using which converted prothrombin into thrombin. In all cases, this activity was insensitive to PMSF, suggesting that it is not due to serine proteases. Table 3 also shows that Ophiophagus hannah venom was able to activate protein C with a low activity compared to that observed in the venom of Agkistrodon contortrix contortrix (Kisiel et al., 1987). Interestingly, the venom from Trimeresurus stejnegeri was characterized by the capacity to activate plasminogen in vitro (Table 3). This action December 1993 Asiatic Herpetological Research Vol. 5 p. 125 appeared to be due to (a) serine protease(s), since it was completely abolished by PMSF. We observed no correlation between the amidolytic activity of snake venoms, measured with the chromogenic thrombin substrate (S-2238) and their thrombin-like activity as determined by their ability to clot fibrinogen (result not shown). Similarly, there was no correlation between activation of prothrombin (Table 3) and the amidolytic activity measured with factor Xa substrate S-2222 (Table 2). These results emphasize the differences which exist between the substrate specificity of human coagulation factors and snake venom activators, and the existence in snake venoms of proteases which are able to hydrolyse chromogenic substrates without possessing the capacity to activate the corresponding blood coagulation factors. Discussion In the present study, we found that, except for Bungarus multicinctus, the venoms from common Chinese venomous snakes (Ophiophagus hannah, Naja naja atra, Bungarus fasciatus, Trimeresurus stejnegeri, Trimeresurus mucrosquamatus, Agkistrodon acutus and Agkistrodon halys) possessed procoagulant and/or anticoagulant activities. An in vitro analysis of these venoms indicated that their action on blood coagulation results from the combined effects of several procoagulant and anticoagulant components. In particular, comparing the effects of native venom with those of venom in which PLA2 activity has been blocked revealed the presence of anticoagulant PLA2s in Agkistrodon halys venom, similar to that descrided by Chen et al. (1987), which masked the effect of procoagulant component(s). We also showed that the procoagulant action of Agkistrodon acutus venoms does not result only from the previously described thrombin-like enzyme (Ouyang et al., 1971), but also from at least two other components, a PMSF-insensitive prothrombin activator and a calcium- dependent factor X activator. This illustrates the complexity of action of the venoms on blood coagulation mechanisms. We demonstrated the presence of prothrombin activators in Agkistrodon acutus, Agkistrodon halys and Trimeresurus stejnegeri venoms, as well as of factor X activators in Agkistrodon acutus, Trimeresurus mucrosquamatus, Trimeresurus stejnegeri, Ophiophagus hannah, Bungarus fasciatus and Naja naja atra venoms. These studies further indicated that prothrombin and factor X activators from Crotalidae (Agkistrodon acutus, Agkistrodon halys, Trimeresurus mucrosquamatus and Trimeresurus stejnegeri) venoms are PMSF-insensitive, and Ca2+-dependent in the case of factor X activators. These activators might be similar to those of Bothops atrox and Vipera russelli venoms (Kisiel et al., 1976; Hoffman and Bon, 1987a; 1987b). Interestingly, factor X activators found in Elapidae venoms (Ophiophagus hannah and Bungarus fasciatus) were inactivated after u-eatment with PMSF, suggesting that they might be serine proteases. We also showed the existence of a protein C activator in the venom of Ophiophagus hannah, although its activity was low compared to that found in the venom of Agkistrodon contortrix contortrix (Kisiel et al., 1987; Orthneret al., 1988). These studies also revealed the first evidence of the existence of a plasminogen activator. Trimeresurus stejnegeri contains a PMSF-sensitive plasminogen activator. Its biochemical structure and mechanism of action are currently under investigation. Acknowledgments This research was supported in part by funds from the Direction des Recherches et Etudes Techniques. Zhang Yun is the recipient of a fellowship from the Fondation pour la Recherche Medicale. References CHEN, Y. C, J. M. MARAGANORE, I. REARDON, AND R. L. HEINRIKSON. 1987. Characterization of the structure and function of Vol. 5 p. 126 Asiatic Herpetological Research December 1993 three phospholipases A2 from the venom of Agkistrodon halys pallas.. Toxicon 25:401- 409. DESNUELLE, P., J. M. CONSTANTIN. AND J. BALDY. 1955. Potentiometric assay of pancreatic lipase activity. Bulletin tie la Societd de Chimie Biologique 37:285-290. GUAN, L. F., X. ZHANG, AND C. W. CHI. 1988. Differences in fibrin polymerization and fibrinopeptide A and B release induced by human thrombin and by the thrombin-like enzyme from the venom of Agkistrodon halys pallas. Pp. 93- 106. In: Pirkle, H. and F. S.' Markland (eds.), Hemostasis and Animal Venoms. Marcel Dekker, New York. HOFMANN, H., AND C. BON. 1987a. Blood coagulation induced by the venom of Bothrops atrox. I. Identification, purification and properties of a prothrombin activator. Biochemistry 26:772-780. HOFMANN, H., AND C. BON. 1987b. Blood coagulation induced by die venom of Bothrops atrox. II. Identification, purification and properties of two factor X activators. Biochemistry 26:780-787. Journal of Biochemistry (Japan) 83:559-570. (In Japanese). ORTHNER, C. L., P. BHATTACHARYA, AND D. K. STRICKLAND. 1988. Protein C activator from the venom of Agkistrodon contort rix contortrix. Biochemistry 27:2558-2564. OUYANG, C, J. S. HONG, AND C. M. TENG. 1971. Purification and properties of the thrombin-like principle of Agkistrodon acutus venom and its comparison with bovine thrombin. Thrombosis et Diathesis Haemorrhagia 26:224-234. OUYANG, C, AND C. T. TENG. 1972. Purification and properties of the anticoagulant principle of Agkistrodon acutus venom. Biochimica Biophysica Acta 278:155-162. OUYANG, C, C. M. TENG, Y. C. CHEN, AND S. C.LIN. 1978. Purification and characterization of the anticoagulant principle of Trimeresurus mucrosquamatus venom. Biochimica Biophysica Acta 541:394-407. PIRKLE, H., AND F. S. MARKLAND. 1988. Hemostasis and Animal Venoms. Marcel Dekker, New York. 486 pp. KISIEL, W., M. A. HERMODSON, AND E. W. DAVIE. 1976. Factor X activating enzyme from Russell's viper venom: isolation and characterization. Biochemistry 15:4901-4906. KISIEL, W., S. KONDO, K. J. SMITH, B. A. MCMULLEN, AND L. F. SMITH. 1987. Characterization of a protein C activator from Agkistrodon contortrix contortrix venom. Journal of Biological Chemistry 262:12607- 12613. RADVANYI, F. R., AND C. BON. 1982. Catalytic activity and reactivity with /?-bromophenacyl bromide of the phospholipase subunit of crotoxin. Journal of Biological Chemistry 257:12616-12623. STOCKER, K. F., AND J. MEIER. 1988. Thrombin-like snake venom enzymes. Pp. 67- 84. In: Pirkle, H. and F. S. Markland (eds), Hemostasis and Animal Venoms. Marcel Dekker, New York. LIU N. K., AND Y. L. XIONG. 1990. Purification and characterization of thrombin-like enzymes from the venom of Trimeresurus stejnegeri. Zoological Research (China) 11:234-240. (In Chinese). MORITA, T., AND S. IWANAGA. 1978. Purification and properties of die prothrombin activator from the venom of Echis carinatus. TENG, C. M., AND W. H. SEEGERS. 1981. Agkistrodon acutus snake venom inhibits prothrombinase complex formation. Thrombosis Research 23:255-263. VOLWERK, J. J., W. A. PIETERSON, AND G. H. DE HAAS. 1974. Histidine at the active site of phospholipase. Biochemistry 13:1446-1454. December 1993 Asiatic Herpetological Research Vol. 5, pp. 127-136J Herpetogeographical Map of Turkmenistan SAHAT SCHAMMAKOV, CHARI. ATAEV, AND ELDAR. A. RUSTAMOV Institute of Zoology, Turkmenistan Academy of Sciences, Azadi Street 6, 744000 Ashgabad, Turkmenistan Abstract. -The herpetological map presented in this paper shows the distribution and abundance of the reptiles of Turkmenistan. The country is divided into 17 complexes and the 84 species and subspecies found in Turkmenistan are listed as occurring in mountains, plains, or both. Key words: Reptilia, Turkmenistan, biogeography, distribution. Introduction In the mid-1960's biogeography entered a new state of development with the practice of ecosystems mapping (Chyel'tsov-Bebutov, 1963, 1964, 1970, 1976). We do not here discuss the principles of the preparation and classification of geographical maps which depict animal population areas. We can only note that they make-up the series of sections included in many integrated regional atlases. Special surveys (Chel'tsov and Chibisova, 1976; Chel'tsov-Bebutov et al., 1972) have dealt with them as well. Nevertheless, the above mentioned maps were prepared for birds and mammals. Until now there have been no geographical maps (as geographical science visualized these) which present the quantitative proportions of reptiles in the total animal kingdom of any region (Chel'tsov-Bebutov and Chibisova, 1976). The three authors of this article (Ataev, Rustamov, and Shammakov, 1989) created a color version of the Herpetogeographical Map of Turkmenistan in 1989. It was presented in 1989 at the All-Union Seminar dealing with the animal kingdom registration and cadastre (in Ufa), the Zoological Section of the Moscow Naturalists Society (in Moscow) and the Vll-th All-Union Herpetological Conference (in Kiev). This article presents a black and white version of the map, giving no consideration to color qualitative background, on the scale 1:2,000,000, to be included into the Turkmen SSR Geographical Atlas (Fig. 1). Field data, gathered throughout the whole Turkmenistan during 1960-1985 (Schammakov, 1981; Ataev, 1985), served as the main sources for this map. Other data were obtained from literature (Rustamov, 1966, 1981; Ataev, 1975; Rustamov and Schammakov, 1982; Ataev, Rustamov and Schammakov, 1985; Rustamov and Shcherbak, 1986; Makeev et al., 1988). Topographic maps on the scale of 1:1,000,000 and 1:500,000 were used as the cartographic basis. The taxonomic generalization level of the topological contours shown on the map were dependent on both its scale and an analysis of data gathered by Ataev and Schammakov, unfortunately, apart, not in assemblage, with the zoogeographical survey of the country by Rustamov. A further point: the whole complex of a habitat and the animal population, which it supports, was taken as a unit undergone to zoogeographical mapping (Chel'tsov- Bebutov, 1963, 1964, 1976). We tried to single out the larger habitats at a level of an ecosystem (landscape or land system, according to Christian, 1975), not of the land unit, which is in close correlation with both the chosen scale and the content of the rest of the maps belonging to the Nature Division of the Atlas. The map scale provoked the necessity to single out such complexes of the reptiles population territorial aggregation, which should be grouped into a definite unity with regard to both common conditions of the habitats (the integral components of which are those aggregations) and the dominant species prevailing in number. A total of 17 complexes as such were revealed. Thus, © 1993 by Asiatic Herpetological Research Vol. 5, p. 128 Asiatic Herpetological Research December 1993 the map was build up on consideration of the habitats of reptiles and their species composition and density. Any territorial differentiation not proved by distinctions in the reptile population was not, as a rule, taken into account. The reptile fauna of the Turkmenistan (Table 1) includes 78 species (84 subspecies) which belong to 2 orders and 14 families. The fauna consists of 3 species (3 subspecies) of tortoises, 47 (51) lizards, and 28 (30) snakes. The information on the reptiles species and population quantities distributed through every complex is placed in a special table that is not given in the atlas, as well as the Table 1, because of the lack of space. One needs this table because the map contours contain no concrete figures on the general density and species number of reptiles. The reptile populations are characterized only according to their appropriate abundance levels. This is quite enough for examining the general content of the map. Nevertheless, we provide herpetologists using this map with more concrete figures (Table 2). Reptile distributions, their abundance, and correlation are dependent upon habitats diversity as well as the fauna richness and specific ecologico- geographical peculiarities (Rustamov, 1966, 1981; Ataev, 1975; Rustamov and Schammakov, 1982). This, in turn, forms the physiognomy of the 17 territorial herpetological complexes. To optimize the reptiles population characteristics, the map legend was made up of 2 parts: the table (placed at the Supplement) and the text. In addition, the insets give information on the fauna composition and contain the out-scale signs characterizing the loci of the habitats. The tables series are arranged according to the principle that permitted us to depict the territorial structure of herpetological complexes, although the map scale and content give no possibility to illustrate the morphological specification of the habitats occupied by these complexes. For example, the table-legend horizontal columns present the main groups of the territorial herpetological complexes revealed on the basis of common ecosystems availability within the compared habitats. Those (groups) are: plain-desert (4 habitats), flood-plain valley (7), piedmont semi-desert (3), and mountain-arid (3). The vertical columns present the territorial units obtained as a result of geographical regionalization that, in our case, merely ground the boundaries of the herpetocomplexes. Such units of the regionalization scheme (zoogeographical regions) within Turkmenistan include: 1 area, 1 sub-area, 3 provinces, 4 districts, 6 regions and 10 sections (Rustamov and Scherbak, 1986). The text of the legend gives the reptile population characteristics for every habitat gone into either complex. In front of the latter's name there is a circle under the correspondent number, the color map has qualitative background representing the complex. The latter's name is followed by the species number and the animals total density index (individuals per ha). The text of the legend is reduced in this article as the abundance indices are brought out in the special table (see Table 2). Further reptile population characteristics for every complex are presented with species numeration of a fixed sequence: first species which use large areas are listed, then the stenotypic ones, which are confined to individual, smaller habitats within a contour. For example, clay surface, solonchaks, construction sites, etc., which are evidently differentiated due to their decreased sizes. The species names are arranged one after another according to decreasing population number within the habitat, of which a brief description is given immediately prior to the species enumeration (see the text of the legend). The dominant species are followed by (1), the codominant ones by (2), and the minor species by (3). The dominant species are defined by us as those whose number is over 10 per hectare, codominant species from two to nine per hectare, and minor species only one per hectare. Thus, the map shows the herpetological territorial complexes differentiated December 1993 Asiatic Herpetological Research Vol. 5, p. 129 according to their species composition, total abundance and dominance levels (with regard to the species number) as well as principle features of the territory's morphology and its ecosystems structure, including the pattern of soils and vegetation cover. Mapping had proved to be the most effective means to manifest and analyze the reptiles population richness throughout the country. The present map can serve as the data source to evaluate the actual situation with the Turkmenistan reptile resources, or to elaborate the practical measures on resource use and conservation. The map can be a help to anybody who will create new, more detailed, large-scaled herpetological maps of either Turkmenistan or any other country. TABLE 1. Reptiles of Turkmenistan. Su- USSR Red Data Book; T- Turkman SSR Red Data Book. Mountains Plains Mountains & Plains Order Testudines Emys orbicularis (Linnaeus, 1758) Mauremys caspica (Gmelin, 1774) Agrionemys horsfieldi (Gray, 1844) Order Squamata Suborder Sauria + + + Phrynocephalus helioscopus helioscopusPaWas, 1771 P. interscapularis Lichtenstein, 1858 P. maculatus Anderson, 1872 (Su, T) P. mystaceus mystaceus Pallas, 1776 P. raddei raddei Boettger, 1888 P. r. boettgeri Bedriaga, 1905 P. reticulatus reticulatus Eichwald, 1831 P. r. bannikovi Darevsky, Rustamov et Schammakov, 1976 P. rossikowi rossikowi Nikolsky, 1899(Su, T) P. r. schammakowi Szczerbak et Golubev 1979, (Su, T) Stellio caucasius caucasius + Stellio c. triannulatus Ananjeva et Ataev, 1984 S. chernovi (Ananjeva, Peters et Rzepakovsky, 1981) + S. erythrogaster Nikolsky, 1896 + S. lehmanni Strauch, 1896 + Trapelus sanguinolentus aralensis (Lichtenstein, 1823) Pseitdopus apodus apodus Pallas, 1775 + Alsophylax laevis Nikolsky, 1907 (Su, T) A. loricatus szczerbaki Golubev et Sattorov, 1979 (Su, T) A. pipiens (Pallas, 1814) (T) Bunopus tuberculatus Blanford, 1874 (Su, T) + Crossobamon eversnumni (Wiegmann, 1834) Cyrtopodion caspius caspius Eichwald, 1831 C. fedtschenkoi (Strauch, 1 887) + C. longipes microlepis Lantz, 1918 (Su, T) + C. russowi (Strauch, 1887) C. spinicauda (Strauch, 1887) (Su, T) + C. turcmenicus (Szczerbak, 1978) (Su, T) + Eublepharis turcmenicus Darevsky, 1977 (Su, T) + Teratoscincus scincus scincus Schlegel, 1858 + + + + + + + + + + + + Vol. 5, p. 130 Asiatic Herpetological Research December 1993 Eremias arguta uzbekistanica Chernov, 1934 (T) - + - E. grammica (Lichtenstein, 1823) - + - E. intermedia (Strauch, 1876) - + - K lineolata (Nikolsky, 1896) - + E. nigrocellata Nikolsky, 1896 (T) - + E. persica Blanford, 1874 - - + E. regeli Bedriaga, 1905 (T) - - + E. scripta scripta Strauch, 1867 - + - E. strauch i kopetdaghica Szczerhak, 197 1 + - E. velox velox Pallas, 1771 - - + Lacerta raddei raddei Boettger, 1892 (T) + L. strigata Eichwald, 1 83 1 - - + Mesalina guttulata wotsonana Stoliczka, 1872 - + - Ablepharus deserti Strauch, 1868 - + A. pannonicus (Lichtenstein, 1 823) - - + Chalcides ocellatas ocellatus Forskal, 1775 (Su, T) + - - Eumeces schneideri princeps Eichwald, 1839 - - + E. taeniolatus taeniolatus Blyth, 1854 - - + Mabuya aurata septemtaeniata Reuss, 1 834 + Ophiomorus chernovi Anderson et Leviton 1966 (Su, T) + Varanus griseus caspius Eichwald, 1831 (Su, T) - - + Suhorder Serpentes Eryx elegans (Gray, 1849) (Su, T) + E. miliaris miliaris Pallas, 1773 - + - E. tataricus specious Tsarevsky, 1915 (T) + - Boiga trigonatum melanocephalia Annandale, 1904 (Su, T) - - + Coluber caspius Gmelin, 1789 (T) - - + C. karelini karelini Brandt, 1838 - + C. najadum najadum Eichwald, 1831 (T) + - C. ravergieri Menetries, 1 832 - - + C rhodorhachis rhodorhachis (Jan, 1865) - - + C. r. ladacensis (Anderson, 1871) - - + Eirenis medus (Gernov, 1949) + Elaphe dione (Pallas, 1773) - + E. quatuorlineata sauromates Pallas, 1814 (T) - + - Lycodon striatus bicolor Nikolsky, 1903 (Su, T) + - Lythorhynchus ridgewayi Boulenger, 1887 (Su, T) - - + Natrix natrix persa Pallas, 1814 - + - N. tessellata (Laurenti, 1768) - - + Oligodon taeniolatus (Jordan, 1853) (Su, T) + Psammophis lineolatum (Brandt, 1838) - + P. schokari schokari Forskal, 1775 + Pseudocyclophis persicus persicus Anderson, 1872 + Ptyas mucosus nigric ans Cernov, 1949 (Su, T) - - + Spalerosophis diadema schiraziana Jan, 1865 - - + Telescopus rhynopoma (Blanford, 1874) (Su, T) + - Agkistrodon halys caraganus Eichwald, 1 83 1 (T) - + A. h. caucasicus Nikolsky, 1916 (T) + Naja oxiana (Eichwald, 1831) (Su, T) - - + Typhlops vennicularis Menem, 1820 + - Echis multisquamatus Cherlin, 1981 - - + Wiper a lebetina turanica Cernov, 1940 - - + December 1993 Asiatic Herpetological Research Vol. 5, p. 131 TABLE 2. Abundance and proportions of ecologico-systematic groups within the territorial complexes of Turkmenistan. 1*- species number; 2*- individuals per hectare. Systematic groups and abundance Testudines Sauria Scrpentes Total Complexes % % 1 % 1. South-Ustjurt 1 3.7 2.4 17 141.4 92.8 9 7.3 4.8 27 151.5 2. Caspian 1 3.6 3.3 16 95.0 84.8 13 13.2 11.8 30 111.8 3. Karakum 1 3.7 2.2 23 155.0 89.5 10 14.4 8.3 34 173.1 4. Sundukli 1 3.8 3.2 17 106.9 90.6 7 7.3 6.2 25 118.0 5. Sarykamysh 1 3.7 3.1 14 104.1 88.0 7 10.5 8.9 22 120.3 6. Uzboi 2 7.8 6.9 16 92.7 81.9 7 8.0 11.2 26 107.5 7. Atrek-Sumbar 3 7.9 8.4 12 65.5 69.6 10 20.5 22.0 25 93.9 8. Tedzhen - Haushan 1 3.4 3.2 16 99.6 84.6 12 14.4 12.2 29 117.4 9. Murgab 1 3.5 2.2 16 153.0 88.3 12 14.5 9.5 29 171.0 10. Amu-Darya 1 3.3 2.0 19 172.7 90.5 13 14.2 7.5 33 190.2 11. Kopetdag piedmont-anthropogenic 1 3.9 7.4 6 35.5 72.6 8 10.5 20.0 15 49.9 12. Kopetdag piedmont 1 3.6 2.3 19 148.2 93.1 10 7.3 4.6 30 159.1 13. Kugitang piedmont 1 3.8 3.8 16 86.7 88.8 8 7.3 7.4 25 97.8 14. Badghyz-Karabil 1 11.5 27.2 18 27.4 59.2 12 3.8 13.6 31 42.7 15. Balkhan 1 1.2 4.4 6 21.2 89.0 4 1.5 6.6 11 23.9 16. Kopetdag mountain 3 9.1 11.3 16 58.6 81.5 18 5.2 7.2 37 71.9 17. Kugitang mountain 1 1.3 4.8 9 22.6 83.4 10 3.2 11.8 20 27.1 (•) Alaophylax laevia (\ Alaophylax laricatua ^^ Bunopua tuberculatum ^-p. Cyrtopodion longapee UphiOinorua chernovi iremiaa regeli ,r).arguta K.nigrocellata ; £ryx tataricua tAj Chalcidea ocellatua I J Lacerta raddei V Cyrtopodion turcmenicua 30 Lacerta atrigata L^ Coluber najadutn Phrynocephalua roaaicorf'i Te lea c opus rhynopo.na total reptile density (sped esAa 1/ / X <50 \/'//X50-'00 \fv^W0-!50\ tortoises 150-200 preval species /r^Sp-prevajent ^^^ < species 0> secondary -pedes mountain sped es plain species is ESS pn«kfF 1 if *rdp FIG. 1. Herpetological map of Turkmenistan. The species composition of various habitats within each of the 17 complexes is listed below. We define dominant species (1) as those that number over 10 per hectare, co-dominant species (2) as those that number from 2-9 per hectare, and minor species (3) as those that number one or less per hectare. Vol. 5, p. 132 Asiatic Herpetological Research December 1993 1. South-Ustjurt Complex Various types of northwestern Turkmenistan deserts — Cyrtopodion caspius (1), Eremias intermedia (2), Trapelus sanguinolenta (2), Agrionemys horsfieldi (2), Eryx miliaris (2), Psammophis lineolatum (2), Coluber karelini (3), Spalerosophis diadema (3). Clay, crushed-stone and solonchak habitats — Phrynocephalus helioscopus (2). Sandy and clay — Varanus griseus (3), Naja oxiana (3), Boiga trigonatum (3), Agkistrodon halys (3). Clay — Cyrtopodion russowi (2). Sandy — Phrynocephalus inter scapularis (1), Teratoscincus scincus (1), Crossobamon eversmanni (2), Eremias grammica (2), Eremias scripta (2), Phrynocephalus mystaceus (2). 2. Caspian Complex Various types of eastern Caspian deserts — Cyrtopodion caspius(2), Eryx miliaris (2), Coluber karelini (3), Eremias intermedia (2), Eremias velox (2), Trapelus sanguinolentus (2), Agrionemys horsfieldi (2), £c/;/.v multisquamatus (2), Psammophis lineolatum (2), £c/!/5 multisquamatus (2), Sandy, clay and solonchak habitats — Eremias lineolata (2). Clay, crushed stone and solonchak — Phrynocephalus helioscopus (2). Sandy, clay and construction sites — Eumeces schneideri (2), Coluber rhodorhachis (3), Varanus griseus (3), Mi/a oxiana (3), Boiga trigonatum (3). Sandy and clay — Agkistrodon halys (3). Sandy and crushed stone — Phrynocephalus reticulums (1). Sandy and on construction sites — Coluber ravergieri (3). Clay and on construction sites — Mabuya aurata (2). Sandy — Phrynocephalus interscapularis (1), Teratoscincus scincus (1), Crossobamon eversmanni (3), Eremias grammica (3), Eremias scripta (3), Phrynocephalus mystaceus (3). Clay — Elaphe quatuorlineata (3). Solonchak — Lythorhynchus ridgewayi (3). By water bodies — Natnx tessellata (3). i. Karakum Complex Various types of Karakum deserts — Cyrtopodion russowi (2), Cyrtopodion caspius (2), Agrionemys horsfieldi (2),Eryx miliaris (2),Coluber karelini (3), Eremias grammica (2),Trapelus sanguinolentus (2), Psammophis lineolatum (2), Eremias velox (2), Echis multisquamatus (2), Ecto multisquamatus (2), Spalerosophis diadema (3). Sandy, clay and solonchak habitats — Phrynocephalus raddei (2), Eremias grammica (2). Clay, crushed stone and solonchak — Phrynocephalus helioscopus (2). Sandy, clay and on construction sites — Eumeces schneideri (2), Coluber rhodorhachis (3), Varanus griseus (3), Naja oxiana (3), Bo/^a trigonatum (3). Sandy, less common crushed stone — Phrynocephalus reticulums (1). Sandy, clay, less common crushed stone — Mesalina guttulata (2). Sandy and on construction sites — Coluber ravergieri (3). Clay and crushed stone — Lythorhynchus ridgewayi (3). Clay and on construction sites — Mabuya aurata (2). Sandy — Phiynocephalus interscapularis (I), Teratoscincus scincus (1), Crossobamon eversmanni (2), Eremias grammica (2), Eremias scripta (2), Phrynocephalus mystaceus (2), Vipera lebetina (2). 4. Sundukli Complex Various types of Sundukli massif deserts — Cyrtopodion caspius (I), Eremias intermedia (2),Trapelus sanguinolentus (2),Agrionemys horsfieldi (2),Eryx miliaris (2), Psammophis lineolatum (2),Coluber karelini (3), Spalerosophis diadem (3). Sandy, clay and solonchak habitats — Phrynocephalus raddei (2), Eremias lineolata (2),Eremias velox (2),Cyrtopodion russowi (2),Echis multisquamatus (2). Clay, crushed-stone and solonchak) — Phiynocephalus helioscopus (2). Sandy and clay — Varanus griseus (?>),Naja oxiana (3), Boiga trigonatumO). Sandy — Phrynocephalus interscapularis (1), Teratoscincus scincus (1), Crossobamon eversmanni (2), Eremias grammica (2), Eremias scripta (2), Phrynocephalus mystaceus (2). Crushed stone — Cyrtopodion fedtschenkoi (2). 5. Sarykamysh Complex Various habitats of the Sarykamysh Depression — Coluber karelini (3), Eremias velox (2), Trapelus sanguinolentus (2), Agrionemys horsfieldi (2), Cyrtopodion caspius (2), Psammophis lineolatum (2), Eryx miliaris (2), Spalerosophis diadema (3), Varanus griseusO). Sandy, clay and solonchak habitats — Eremias grammica (2). Sandy, clay, solonchak and on construction sites — Cyrtopodion russowi (2). Clay, crushed- stone and solonchak — Phiynocephalus helioscopus (2). Sandy, clay and solonchak — Eremias lineolata (2). Sandy, clay and on irrigated lands) — Agkistrodon halys (3). Sandy — Phrynocephalus interscapularis (1), Teratoscincus scincus (1), Crossobamon eversmanni (2), Eremias grammica (2), Phrynocephalus mystaceus December 1993 Asiatic Herpetological Research Vol. 5, p. 133 (2). Solonchak habitats, along collectors and canals, in settlements — Elaphe dione (3). On irrigated lands and water bodies; — Natrix tessellata (2). 6. Uzboi Complex Various habitats of Western Uzboi Valley — Coluber karelini (3), Cyrtopodion caspius (2), Agrionemys horsfieldi (2), Trapelus sanguinolentus (2), Eremias velox (2), Echis multisquamatus (2), Eryx miliahs (2), Psammophis lineolatum (2), Spalerosophis diadema (3), Coluber rhodorhachis (3), Varanus griseus (3), Naja oxianaO). Sandy, clay, solonchak and on construction sites — Cyrtopodion russowi (2). Sandy, clay, solonchak and crushed-stone — Eremias intermedia (2). Flood-plain, clay and on construction sites — Mabuya aurata (2). Clay, crushed-stone and solonchak — Phrynocephalus helioscopus (2). Sandy, clay and solonchak — Phrynocephalus raddei (2), Eremias lineolata (2). Flood plain — Emys orbicularis (2). Sandy — Phrynocephalus interscapularis (1), Teratoscincus scincus (1), Crossobamon eversmanni (2), Eremias scripta (2), Eremias grammica (2), Phrynocephalus mystaceus (2). 7. Atrek-Sumbar Complex Various habitats of Atrek and Lower Sumbar valleys — Trapelus sanguinolentus (2), Eremias velox (2) Agrionemys horsfieldi (2), Echis multisquamatus (2), Eumeces schneideri (2), Cyrtopodion caspius (2), Coluber karelini (3), Elaphe dione (3), Varanus griseus (3),Psammophis lineolatum (2), Spalerosophis diadema (3), Boiga trigonatum (3). Flood plains and irrigated lands — Natrix natrix (2), Natrix tessellata (2), Ablepharus pannonicus (2), Coluber caspius (3). Sandy, clay, crushed stone, and solonchak habitats — Eremias intermedia (2), Eryx miliahs (2). Clay, crushed-stone and solonchak — Phrynocephalus raddei (2). Clay, crushed-stone and solonchak — Phrynocephalus helioscopus (2). By water bodies — Emys orbicularis (2), Mauremys caspica (2). Sandy — Teratoscincus scincus (1), Crossobamon eversmanni (2). 8, Tedzhen-Hauzkhan Complex Various habitats of Tedzhen Valley and Hauzkhan Massif — Eremias velox (2), Natrix tessellata (2), Trapelus sanguinolentus (2), Agrionemys horsfieldi (2), Echis multisquamatus (2), Eumeces schneideri (2), Mabuya aurata (2), Coluber karelini 0),Coluber rhodorhachis (3), Naja oxiana (3), Varanus griseus (3), Coluber ravergieri (3), Cyrtopodion caspius (2), Psammophis lineolatum (2), Boiga trigonatum (3), Vipera lebetina (2), Eumeces taeniolatus (3), Spalerosophis diadema (3), Eryx miliahs (2), Eremias intermedia (2). Sandy, clay and solonchak habitats — Phrynocephalus raddei (2), Eremias lineolata (2). Clay, crushed-stone and solonchak— Phrynocephalus helioscopus (2). Sandy and clay— Mesalina guttulata (3). Sandy— Phrynocephalus interscapularis(l), Phrynocephalus mystaceus (2), Eremias grammica (2). Clay — Lytorhynchus hdgewayiO). 9. Murgab Complex Various habitats of the Murgab Valley— Eremias velox (2), Trapelus sanguinolentus (2), Agrionemys horsfieldi (2), Cyrtopodion caspius (2), Echis multisquamatus (2), Eumeces schneideri (2), Psammophis lineolatum (2), Vipera lebetina (2), Mabuya aurata (2), Coluber karelini (3), Coluber rhodorhachis (3), Varanus griseus (3), Naja oxiana (3), Spalerosophis diadema (3), ftya5 mucosus (3), flo/ga trigonatum (3). On flood-plains and irrigated lands— Ablepharus deserti (1), Ablepharus pannonicus (2), Atom* tessellata (2), Eumeces taeniolatus (3). Sandy, clay, crushed-stone and solonchak habitats— Eremias intermedia (2), Eryx miliahs (2). Sandy, clay, solonchak habitats and on construction sites— Cyrtopodion russowi (2). Sandy, clay and solonchak habitats— Phrynocephalus raddei (2), Eremias lineolata (2). Sandy— Phrynocephalus interscapularis (1), Teratoscincus scincus (I), Crossobamon eversmanni (2). Clay— Lythorhynchus hdgewayi (3). /0. A/nu Darya Complex Various habitats of the Amu Darya Valley— Ablepharus deserti (1), Eremias velox (2), Natrix tessellata (2), Trapelus sanguinolentus (2), Agnortemy.? horsfieldi (2), Cyrtopodion caspius (2), £cto multisquamatus (2), Psammophis lineolatum (2), V/pera lebetina (2), Eumeces schneideri (2), Mabuya aurata (2), Coluber karelini (3), Spalerosophis diadema (3), Varanus griseus (3), My'a ox/ana (3), Coluber ravergieh,0), Elaphe dione (3), flo/ga trigonatum (3) Agkistrodon halys (3), Eumeces taeniolatus (3). Sandy, clay, crushed-stone and solonchak habitats— £ryx miliahs (2), Eremias grammica (2). Clay, crushed stone and solonchak— Vol. 5, p. 134 Asiatic Herpetological Research December 1993 Phrynocephalus helioscopus (2). Sandy, clay and crushed stone — Phrynocephalus raddei(2), Eremias lineolata(2). Sandy and crushed stone — Phrynocephalus reticulatus(l). Clay and crushed stone — Lythorhynchus ridgewayi (3). Construction sites — Cyrtopodion fedtschenkoi (2). 77. Kopetdag Piedmont Anthropogenic Complex Various habitats of Kopetdag piedmont oases — Mabuya aurata (1), Cyrtopodion caspius (1), Eremias velox (2), NatrLx tessellata (2), Trapelus sanguinolentus (2), Agrionemys horsfiekii (2), Coluber rhodorhachis (3), Coluber ravergieri.O), Naja oxiana (3). In flood-plains of shallow rivers and on construction sites — Eryx miliaris (2), Echis multisquamatus (2), Spalerosophis diadema (3). On flood plains — Eremias lineolata (2), Lythorhynchus ridgewayi (3). 12. Kopetdag Piedmont Complex Various habitats of the Kopetdag piedmont plain — Cyrtopodion caspius (1), Phrynocephalus raddei (2), Phrynocephalus helioscopus (2), Eremias intermedia (2), Eremias lineolata (2), Eremias velox (2), Cyrtopodion russowi (2), Trapelus sanguinolentus (2), Agrionemys horsfieldi (2), Eryx miliaris (2), Echis multisquamatus (2), Psammophis lineolatum (2), Coluber karelini (3), Spalerosophis diadema (3), Varanus griseus (3), Naja oxiana (3). Clay and crushed-stone habitats — Mabuya aurata (1), Lythorhynchus ridgewayi (3). Sandy and clay — Mesalina guttulata (2), Boiga trigonalum (3). Clay — Pseudocyclophis persicus (3), Eirenis medus (3). Sandy — Phrynocephalus interscapularis (1), Teratoscincus scincus (1), Crossobamon eversmanni (2), Eremias grammica (2), Eremias scripta (2), Phrynocephalus mystaceus (2). 7j. Kugitang Piedmont Complex Various habitats of Kugitang piedmont plain — Cyrtopodion caspius (1), Phrynocephalus raddei (2), Eremias intermedia (2), Phrynocephalus helioscopus (2), Eremias lineolata (2), Eremias velox (2), Trapelus sanguinolentus (2), Agrionemys horsfieldi (2), £nu miliaris (2), £c/n.t multisquamatus (2), Psammophis lineolatum (2), Coluber karelini (3), Spalerosophis diadema (3). Clay and crushed-stone habitats — Cyrtopodion fedtschenkoi (2), Lythorhynchus ridgewayi (3). Sandy and clay — Varamw griseus (3), TVa/a oxiana (3). Boiga trigonalum (3). Sandy — Phrynocephalus interscapularis (1), Phrynocephalus mystaceus (2) Eremias grammica (2), Eremias scripta (2). 74. Badghyz-Karabil Complex Various habitats of the Badghyz and Karabil hills — Agrionemys horsfieldi (1), Mabuya aurata (2), Cyrtopodion caspius (2), Trapelus sanguinolentus (2), Eremias velox (2), Ablepharus pannonicus (2), ,Sre//io erythrogaster (2), Eumeces taeniolatus (2), Eumeces schneideri (2), Pseudophus apodus (3), Vipera lebetina (3), /Va/a oxiana (3), Psammophis lineolatum (3), Varanus griseus (3), Spalerosophis diadema (3), Coluber ravergieri (3), Coluber rhodorhachis (3). On slopes covered by stones and mud-streams — Stellio caucasius (2), Typhlops vermicularis (3). On bare rocks-outcrops — Lycodon striatus (3). On stone surfaces — Pseudocyclophis persicus (3). On slopes covered by loess and stones — Psammophis schokari (3). Food-plains and mud-streams — Eremias persica (2). Mud-streams — Oligodon taeniolatus (2). Flood plains — Natrix tessellata (3). 75. Balkhan Complex Various habitats of the Great and Small Balkhan Mountains — Cyrtopodion caspius (1), Stellio caucasius (2), Trapelus sanguinolentus (2), Eremias velox (2), Agrionemys horsfieldi (2), Coluber rhodorhachis (3), iVa/fl oxiana (3), Ablepharus pannonicus (3). Piedmonts and inter-ridge hills — Va/xmi/.? griseus (3). Mud streams, undulated surfaces, piedmonts and inter-ridge hills — Psammophis lineolatum (3). Inter-ridge hills, piedmonts and construction sites — Spalerosophis diadema (3). 76. Kopetdag Mountain Complex Various habitats of the Kopetdag Mountains — Stellio caucasius (2), Ablepharus pannonicus (2), Agr/onmy.s horsfieldi (2), Coluber rhodorhachis (3), Vipera lebetina (3), /Va/a oxiana (3), Cyrtopodion caspius (1), Trapelus sanguinolentus (2), Coluber ravergieri (3), Pseudopus apodus (3), Mabuya aurata (1), Eremias strauchi (2), Eremias velox (2), Typhlops vermicularis (2), Eumeces schneideri (2), Eumeces December 1993 Asiatic Herpetological Research Vol. 5, p. 135 COMPLEXES OF THE MODERN REPTILE POPULATION \ zoogeogra- \ phical \ regions main groups \ of the ter- \ ritorial her- \ petologioal \ ARID MEDITERRANEAN-ASIATIC SUB-AREA pro- vinces TURAN PLAIN-DESERT TRANSITORY IRAN-AFGHAN MOUN- TAIN- ASIA dist- ricts K A R A K U M KYZYLKUM IRAN-AFGHAN UPLAND CENTRA! ASIA MOUTAII regi- ons Ustjurt by-Cas- pian K a r a k u in Central- Kyzylkum North- Iran- Afghan upland Gissar Alaj complexes \ sec- \ti- \ons Ustjurt crushed stone Kara- -kum sand Kara- kum takyr Atrek tugai Ainu Darya tugai Kyzyl- kum sand Kyzyl- kxim crushed' stone takyr Badghyz Karabil pied- mont -Khora- san-Ko- petdag moun- tain Gissar Turke- stan pied- mont [ U3 PLAIN DESERT _ zi an U±J FLOOD-PLAIN VALLEY U3 q t£2 m R5-| PIEDMONT SEMI-DESERT UL2 nn £TJ MOUNTAIN ARID un PT| un FIG. 2. Supplement to herpetological map of Turkmenistan. taeniolatus (2), Cyrtopodion spintccuida (3), Pseudocyclophis persicus (3). Steppe-like, stone and inter-ridge hills — Coluber caspius (3). Stone inter-ridge hills and piedmonts — Lycodon striatus (3), Psammophis schokari (3). Steppe-like inter-ridge hills and piedmonts — Eirenis niedus (3). Inter-ridge hills, piedmonts, and on construction sites — Psammophis lineolatum (3), Spalerosophis diadema (3). Stone and inter-ridge hills — Oligodon taeniolatus (3), Eublepharis turcmenicus (3). Piedmonts — Varanus gnseus (3). Along shallow rivers and other water bodies — Natri.x tessellata (2). Juniper stands — Eryx elegans (3), Agkistrodon halys (3). 17. Kugitang Mountain Complex Various habitats of the Kugitang Mountains — Cyrtopodion fedtschenkoi (1), Agrionemys horsfieldi (2), Vipera lebetina (3), Coluber ravergieri (3), Stellio lehmanni (2), Eremias velox (2), Naja oxiana (3), Ablepharus pannonicus (2), Coluber rhodorhachis (3), Eumeces schneideri (2). Inter-ridge hills, piedmonts, and on construction sites — Trapelus sanguinolenlus (2). Piedmonts, stone habitats, and on construction sites — Cyrtopodion caspius (2). Stone habitats and juniper stands — Stellio chernovi (3). Piedmonts, stone and inter-ridge hills — Spalerosophis diadema (3). Inter-ridge hills and stone surfaces — Pseudopus apodus (3), Typhlops vermicularis (3). Stone surfaces and piedmonts — Psammophis lineolatum (3). Piedmonts — Lytorhynchus ridgewayi (3). Along shallow rivers and other water bodies — Nairix tessellata (2). Literature Cited House, Ashkhabad. 343pp. ATAEV, CH. A. 1975. Geograficheskii obzor herpetofauni gor Turkemenistana. (Geographical overview on the Turkmenistan mountains herpetofauna). Izvestiya Akad. nauk TSSR, ser. biol. nauk, N 3, pp.75-80. ATAEV, CH. A. 1985 gor Turkmenistana. Presmykayushchiesya (Reptiles of the Turkmenistan mountains). Ylym Publishing ATAEV, CH. A, A. K. RUSTAMOV, AND S. M. SCHAMMAKOV. 1985. Presmykayushchiesya (Reptiles). Pp. 208-270. In Krasnaya kniga Turkmenskoi SSR (Turkmen SSR Red Data Book). Turkmenistan Publishing House, Ashkhabad. ATAEV, CH. A., E. A. RUSTAMOV, AND S. M. SCHAMMAKOV. 1989. Opit Vol. 5, p. 136 Asiatic Herpetological Research December 1993 gerpetogeograficheskogo kartografirovaniya Turkmenskoi SSR (Experience of herpetogeographical mapping of Turkmen SSR). Pp. 14-15. In Voprosy gerpetologii (Problems of herpetology - 7th All-Union Herpetological Conference). Naukova dumka Publishing House, Kiev. CHEL'TSOV-BEBUTOV, A. M. 1963. Voprosi melkomasshtabnogo zoogeograficheskogo kartografirovaniya na primere karty Kustanaiskoi oblasti (On small- scale zoogeographical mapping with special reference to Kustanai region). Pp. 124-126. In Tezisi dokladov po voprosam zoologicheskoi kartografii (Theses of reports read at the seminar on problems of zoological mapping). Moscow. Chel'tsov-Bebutov, A. M. 1964. Nekotorie voprosi zoogeograficheskogo kartografirovaniya na primere karti Kustanaiskoi oblasti (Some problems of zoogeographical mapping with special reference to Kustanai region). Pp. 3-24. In Biogeograficheskie ocherki Kustanaiskoi oblasti (Biogeographical essays of Kustanai region). Moscow. Chel'tsov-Bebutov, A. M. 1970. Zoogeograficheskoye kartografirovanie i landshaftovedeniye (Zoogeographical mapping and land system discipline). Pp. 49-94. In Landshafmyi sbornik (Articles on land systems). Moscow. Chel'tsov-Bubutov, A. M. 1976. Zoogeograficheskoye kartografirovanie: osnovnie printsipi i polozheniya (Zoogeographical mapping: main principles and theses). Vestnik MGU, series geographical 2:50-56. Chel'tsov-Bebutov, A. M., A. K. Danilenko, and V. V. Chibisova. 1972. Karti zhivotnogo mira v Sovetskih regional'nih atlasah (Animal kingdom maps in Soviet Regional Atlases). In Metody sozdaniya kompleksnih regional'nih atlasov SSSR. Kati prirodi. (Methods of creation of Regional Integrated Atlases of USSR. The Environmental maps). Moscow. Chel'tsov-Bebutov, A.M., and V. V. Chibisova. 1976. Zhivotnii mir i yego resursi (Animal kingdom and its resources). Pp. 326-341. In Kompleksnie regional'nie atlasi (Integrated Regional Atlases). Moscow. Christian, C. S. 1975. The concept of land units and land systems. Humid Tropics 20:74-81 . Makeev, V. M., A. T. Bozhanskii, S. V. Kudryavtsev, V. I. Frolov, and Yu. D. Khomustenko. 1988. Nekotorie resul'tati gerpetologicheskogo obsledovaniya Vostochnoi Turkmenii (Some results of herpetological examination of the Eastern Turkmenistan). Pp. 127-143. In Redkie i maloizuchennie zhivotnie Turkmenistana (Rare and poor-studied animals of the Turkmenistan). Ylym Publishing House, Ashkhabad. Rustamov, A. K. 1966. Kratki obzor gerpetofauni Turkmenii i eyo zoogeograficheskie osobennosti (Brief overview of the Turkmenistan herpetofauna and its zoogeographical peculiarities). Pp. 158-168. In Pozvonochnie zhivotnie Srednei Asii (Middle Asian vertebrates). Tashkent. Rustamov, A.,K. 1981. Opit otsenki vidovogo yendimizma gerpetofauni Irana, Afganistana i Srednei Azii (Experimental estimation of specific endemism of herpetofauna of Iran, Afganistan and Middle Asia). Pp. 118-119. In Voprosi gerpetologii (Problems of herpetology), Leningrad. Rustamov, A. K., and S. M. Schammakov. 1982. On the herpetofauna of Turkmenistan. Vertebrata Hungarica, Budapest 21:215-226. Rustamov, A. K., and N. N. Shcherbak. 1986. Cierpetogeograficheskoye raionirovaniye Srednei Azii (Herpetogeographical regionalization of Middle Asia). Izvestiya Akad. nauk SSSR, ser. biol., pp. 13-20. Schammakov, S. M. 1981. Premykayushchiesya ravninnogo Turkmenistana (Reptiles of the flat Turkmenistan). Ylym Publishing House, Ashkhabad. 310 pp. December 1993 Asiatic Herpetological Research Vol. 5, pp. 137-142 A Karyosystematic Study of the Genus Bombina from China (Amphibia: Discoglossidae) WAN-ZHAO LIU AND DA-TONG YANG Department of Vertebrate Zoology. Kunming Institute of Zoology, Academia Sinica, Kunming. Yunnan. China Abstract. -Chromosome number, morphology and positions of Ag-NORs were determined for four Chinese species of Bombina. Chromosome numbers are: B. orientalis (2n=24, NF=48), B. maxima (2n=28, NF=56), B. microdeladigitora (2n=28, NF=56). The Ag-NORs of B. orientalis are located on the long arm of the 7th chromosome pair, where as those of the latter three are on the short arm of the 1 1th pair. The subdivision of Bombina into two subgenera is supported by the karyology. A close relation between Bombina and Discoglossus is suggested. Key words: Anura, Discoglossidae, Bombina; karyotypes, Ag-NORs, China. TABLE 1. Species, localities, and number of individuals used in this karyological study. Species Locality No. of Individuals B. (G.) fortinuptialis B. (G) maxima B. (G) microdeladigitora B. (B.) orientalis Jinxiu, Guangxi Dayao, Yunnan Jingdon, Yunnan Qindao, Shandong 5 males 1 female 1 males 5 females 5 males 2 females 3 males 1 female Introduction There are five genera in the family Discoglossidae. The systematics of this family have long been under discussion. The systematic position of the genus Bombina within this family is the most problematical. A variety of studies dealing with this genus have been presented during the past years (summarized by Lang, 1988; 1989a; 1989b). However, the systematics of Bombina is still quite confusing. The relationships within Bombina have not been fully worked out. Only six species belong to the genus Bombina. All are distributed in Eurasia. Karyological data are known for B. bombina (2n=24, NF=48), B. variegata (2n=24, NF=48) [Morescalchi, 1973], B. orientalis (2n=24, NF=48) [Jiang et al., 1984], B. maxima (2n=28, NF=56) [Zhao, 1986, no photographs presented]. No chromosome banding data are available for Bombina. Careful morphological analysis and banding of the chromosomes may yield useful information not only on the phylogeny of the genus itself, but also on the relationships between Bombina and other genera of the family Discoglossidae. The purpose of this study is to analyze the karyotypes and Ag-NORs of four species in the genus from China Liu and Hu, 1961). The results, when compared with known karyological data of B. bombina, B. variegata and other genera of Discoglossidae, should be helpful in understanding the taxonomy and phylogeny of Bombina. Materials and Methods The specimens used in this investigation are listed in Table 1. The toads were collected by the authors at time of the year when both sexes are active for mating. The specimens were kept at room temperature (15-20°C) until the time of investigation. To block mitosis at metaphase, we used a freshly prepared colchicine solution of 0.05%, and injected intraperitonealy l/20ml of this solution per gram of body weight. The animals were sacrificed 20-24 hrs later, the spleen and small intestine were © 1993 by Asiatic Herpetological Research Vol. 5 p. 138 Asiatic Herpetological Research December 1993 TABLE 2. Observational results of the diploid chromosome number of four species of Bombina from China. Species # of cells Number of diploid chromosomes observed 22 23 24 25 26 27 28 29 B. (G.)fortinuplialis percentage (%) 113 1 12 99 1 0.88 10.6 87.6 0.88 B. (G.) maxima percentage (%) 198 3 1.5 3 1.5 14 7 176 88.9 B. (G.) microdeladigitora percentage (%) 238 2 4 22 209 1 0.8 1.7 9.2 87.8 0.4 B. (G.) orientalis percentage (%) 116 2 8 104 1.7 6.9 89.5 2 1.7 removed, and the intestine was opened with a pair of fine scissors to expose the inner epithelial surface. The exposed epithelial surface was washed with a 0.64% NaCl solution for several minutes in order to remove all mucous and debris. The tissue was cut into small pieces and placed in a Petri-dish. We added 8-10 ml of 0.4% KC1 into the dish, suspended vigorously with a Pasteur pipette. The hypotonic treatment lasted 30- 40 minutes. It was centrifuged at 1000 rpm for five minutes, and then the hypotonic solution was removed. The tissue was fixed with 8- 10ml of freshly prepared solution of 3:l(v/v) absolute methanol- glacial acetic acid for three times, with a total time of 60 minutes. The samples will keep indefinitely in the fixative if stored at 1-4°C. We prepared slides by transferring 3-4 pieces of the fixed tissue onto a dry, warm (about 50°C) slide. We then added 5-10 drops of 60% acetic acid and siphoned the solution up and down until the solution evaporated completely. Slides were stained in 10% Giemsa (pH 6.8) for 10 minutes. Staining of nucleolus organizer regions (NORs) followed the methods of Howell et al. (1980). A total of 675 mitotic chromosome spreads were observed. Ten selected metaphase plates for each species were photographed, enlarged, and measured. The chromosome nomenclature used is mat suggested by Levan et al. (1964). For the convenience of comparison, the chromosomes are defined as being large (A group), medium (B group) and small (C group) according to their relative lengths. Large chromosomes have a value of 100 units or more, small chromosomes have a value of 40 or fewer units. Chromosomes whose length falls between 40-99 units are considered to be medium. Results The observed diploid chromosome numbers are presented in Table 2. Measurements of metaphase chromosomes of four Chinese species of Bombina are shown in Table 3. It is obvious that, the karyotypes of maxima, microdeladigitora and fortimiptialis are equal to each other. They have 2n=28, NF=56, composed of 6 pairs of large homologous, one pair of medium- small chromosomes and seven pairs of small homologous; all the chromosomes are metacentric (m), except for 6th, 7th, and 9th pair, which are submetacentric (sm). A weak secondary constriction is observed on the short arm of 1 1th pair. It appears in about 10% of the cases. The Ag-NORs are observed on short arm of 11th pair and December 1993 Asiatic Herpetological Research Vol. 5 p. 139 TABLE 3. Measurements of metapha.se chromosomes of four Chinese B. fortinuptialis species of Bombina Mean ± SE B. maxima Group No. Relative length ratio type Group No. Relative length ratio type 1 160.0+4.22 1.1810.06 m 1 165.215.30 1.2310.10 m 2 147.0±3.72 1.3610.03 m 2 144.4+4.03 1.4310.07 m A 3 129.915.36 1.4810.05 m A 3 127.916.21 1.4710.07 m (1-6) 4 119.8±2.10 1.4710.06 m (1-6) 4 122.814.15 1.5810.11 m 5 109.9±5.20 1.2110.08 m 5 108.116.52 1.3810.11 m 6 107.3+3.94 1.8610.13 sm 6 105.814.30 1.7110.06 sm B(7) 7 40.511.05 2.5110.11 sm B(7) 7 40.913.25 2.6910.23 sm 8 33.5±0.97 1.2810.02 m 8 33.412.51 1.5610.09 m 9 28.4±0.69 2.1910.09 sm 9 30.111.75 2.3310.12 sm 10 27.4±1.27 1.5010.03 m 10 28.111.02 1.3710.06 m C 11 26.4±0.78 1.4810.05 m C 11 26.711.33 1.4210.08 m (8-14) 12 25.4+0.76 1.6210.08 m (8-12 12 24.511.06 1.5510.03 m 13 23.1 ±0.41 1 .4410.07 m 13 22.811.10 1.4410.02 m 14 21.4±0.71 1.2210.04 m 14 20.311.04 1.30+0.03 m B. microdehuligi, . . to* . % ; !! if If 11 if ii :#» •»••»»- • < II J5ji_ ;; ;; ji :•* n a tt ** •• «« »» > /• FIG. 1. Karyotypes and Ag NO3 stained karyotypes of Bombina from China. Arrows show Ag-NORs. A: B. fortinuuptialis; B: B. maxima; C: S. microdeladigitora; D: £. orientalis. are presented in Fig. 1. Discussion Now, karyotypes are known for all the recognized species of Bombina. We compare the karyotypes and Ag-NORs of them in Table 4. All the chromosomes of Bombina have median or submedian centromeres. In the discoglossids, Alytes are rich in acrocentrics, and with some microchromosomes (2n=38, NF=64-72), Discoglossus have 2n=28, NF=54, with one pair of telocentrics (Morescalchi, 1973). So, from the karyololgical point of view, Bombina is the most highly differentiated. Within Bombina, two different kinds of karyotypes exist. The differences between the two are mainly as follows: 1). The morphology of 7th pair are quite different. The 7th pair of maxima, microdeladigitora and fortinuptialis are medium-small and s, where as those of bombina, variegata and orientalis are medium-large, m, with a clear secondary constriction on the long arm. 2). The number of smaller homologues is different. The former three have 7 pairs of small homologues, but the latter three have only 5 small homologues. Tian and Hu (1985) suggested a subdivision of Bombina into two subgenera, the subgenus Bombina containing the Palaearctic bombina, variegata and orientalis, and the Oriental Glandula containing maxima, microdeladigitora and fortinuptialis. Our December 1993 Asiatic Herpetological Research Vol. 5 p. 141 TABLE 4. Comparison of karyotypes and Ag-NORs in Bombina. Species NF Chromosome formula Secondary centremere Locality of A-NORs Data B. (G.) fort in uptialis 28 56 22m+6sm No. 11 No. 11 Present study B. (G.) maxima 28 56 22m+6sm No. 11 No. 11 Present study B. (G. ) microdeladigitora 28 56 22m+6sm No. 11 No. 11 Present study B. (G.) orienlalis 24 48 22m+2sm Nos. 7,8 No. 7 Present study B. (G. ) bombina 24 48 unknown Nos. 7,8 unknown Morescalchi 1973 B. (G.) variegata 24 48 unknown Nos. 7,8 unknown Morescalchi 1973 karyological evidence support this subdivision. In B. maxima, microdeladigitora and fortirtuptialis, the karyotypes are practically equal to each other which indicates that these three species may have diverged recently. In the group consisting of B. bombina, variegata and orientalis, bombina and variegata are equal to one another (Morescalchi, 1973), but orientalis has some differences from them. The 8th and 12th pairs are m in orientalis, but st in bombina and variegata. Thus the two European species are more closely related, which is congruent with immunological evidence of Maxon (1979) and Maxson and Szymura(1984). Morescalchi et al. (1977) could not resolve relationships within Discoglossidae because the karyotypes of Discoglossus, Alytes and Bombina are so different from each other. Our investigation indicated that Discoglossus and the Glandnla-group of Bombina have the same diploid chromosome number (Lanza et al., 1975; 1976). We suggest that those two genera may be related. The secondary constriction is the only "marker" currently available for analysis in most anuran karyosystematics studies. However, in the present study, we found that the secondary constriction is abrupt and the position of it is quite variable. The Ag-NORs are stable and clear and may be a more useful tool than the place of the secondary constriction in some cases. The mechanisms of karyotype evolution of Bombina may take place through unequal translocation. This is still an open question. A further chromosome banding study is necessary. References JIANG, S., WEN, C, SHEN, C. and MEN, Y. 1984. Preliminary observations on karyotypes of Bombina orientalis. Acta Herpetologica Sinica 3(l):25-27. (In Chinese). LANG, M. 1988. Notes on the genus Bombina Oken (Anura: Bombinatoridae). I. Recognized species, distribution, characteristics and use in laboratory. British Herpetological Society Bulletin No. 26:6-13. LANG, M. 1989a. Notes on the genus Bombina Oken (Anura: Bombinatoridae). II. Life history aspect. British Herpetological Society Bulletin No. 27:13-17. LANG, M. 1989b. Notes on the genus Bombina Oken (Anura: Bombinatoridae). III. Anatomy, systematics, hybridization, fossil record and biogeography. British Herpetological Society Bulletin No. 28:43-49. LANZA, B„ J. M. CEI AND E. G. CRESPO. 1975. Immunological evidence for the specific status of Discoglossus pictus Otth. 1837 and D. sardus Tschudi 1837, with notes on the families Discoglossidae Giinther, 1858 and Bombinidae Fitzinger 1826 (Amphibia: Salientia). Monitore Zoologico Italiano (N.S.) 10:153-162. Vol. 5 p. 142 Asiatic Herpetological Research December 1993 LANZA, B., J. M. CEI and E. G. CRESPO. 1976. Further immunological evidence for die validity of die family Bombinidae (Amphibia: Salientia). Monitore Zoologico Italaliano (N.S.) 10:311-314. LEVAN, A., K. FREDGA, AND A. A. SANDBERG. 1964. Nomenclature for centromeric position on chromosomes. Hereditas 52:201-220. LIU, C. C. and S. Q. HU. 1961. Tail-less amphibians of China. Science Press, Beijing. 1-358 pp. MAXSON, L. R. and J. M. SZYMURA. 1984. Relationships among Discoglossid frogs: An albumin perspective. Amphibia-Reptilia 5:245- 252. MAXSON, L. R. 1979. Quantitative immunological studies of the albumins of several species of fire-bellied toads, genus Bombina.. Comparative Biochemestry and Physiology 63B:517-519. MICHALOWSKI, J. 1961 Studies on species characters in Bombina variegata (L.) and B. bombina (L.): I. Applying the L:T indicator to the classifying purposes. Acta Zoologica Cracoviensia 6(3):51-59. MORESCALCHI, A. 1973. Amphibia. Pp. 233- 348 In A. B. chiarelli and C. Capanna (eds). Cytotaxonomy and vertebrate evolution. Academic Press, New York. MORESCALCHI, A. 1977. Phylogenetic species of karyological evidence. Pp. 149-167. In Hecht, M. K., P. C. Gody and B. M. Hecht (eds.), Major patterns in vertebrate evolution. Plenum Press, New York. MORESCALCHI, A. , E. OLMO and V. STINGO. 1977. Trends of karyological evolution in pelobatid frogs. Experientia 3:1577-1578. TIAN, W. S. AND Q. X. HU. 1985. Taxonomical studies on the primajtive anurans of the Hengduan Mountains, with descriptions of a new subfamily and subdivision of Bombina. Acta Herpetologica Sinica4(3):219-224. ZHAO Y. F. 1986. Studies of the karyotype of Bombina maxima. Acta Herpetologica Sinica 5(3):227-228. (In Chinese). December 1993 Asiatic Herpetological Research Vol. 5, pp. 143-146 A Study on the Purification and Pharmacological Properties of Two Neurotoxins from the Venom of the King Cobra (Ophiophagus Hannah) JIANX1NG SONG, YULIANG XlONG, WANYU WANG, AND XlAOCHUN PU Toxinologica! Laboratory, Kunming Institute of Zoology. Academia Sinica, Kunming, Yunnan, China Abstract- Using Sephadex G-50, CM-Sephadex C-25 and CM-cellulose 52 columns, two neurotoxins of Ophiophagus hannah were purified to be homogeneous on acidic PAGE which contain 53 and 73 amino acid residues respectively. The two neurotoxins were used as a substitute for morphine. Using a morphine withdrawal jumping model, the results demonstrated that the effects of the two neurotoxins when administered by injecting are very significant (P<0.01), and when administered orally are significant (P<0.05). Key words: King Cobra, Ophiophagus hannah, neurotoxins, morphine addiction, naloxone jumping model. Introduction The King Cobra (Ophiophagus hannah) is the largest poisonous snake in the world (Tu, 1977). It is extensively distributed in southern China, India, Thailand and other Asian countries. Joubert (1973) reported on the purification and sequence determination of two toxins from the venom of King Cobras grown in Thailand. Shun (1981) reported on the purification of four postsynaptically acting toxins from the venom of the King Cobra in Guangxi, China. Our research showed that among four toxins, there is a postsynaptically acting toxin containing only 63 amino acid residues. We determined the complete amino acid sequence of neurotoxin, which contain 73 amino acid residues. This neurotoxin's sequence is analogous to that of the neurotoxins determined by Joubert (1973), but its C-terminal four amino acid residues are very similar to that of bungartoxin in hydrophobicity (Lin and Wang, 1984). Now, the sequence of the neurotoxin containing 53 amino acid residues is in the process of being determined. Xiong and Wang (1987) reported the clinical observation results of using the neurotoxin from cobra venom to achieve better analgesic effect on moiphine addicted patients. The possibility to use the neurotoxins from snake venom as a substitute for morphine was also first reported by Xiong (1990). The mechanism has been discussed in the other papers. In this paper, the research concentrated on the purifying of two neurotoxins from the venom of King Cobras from Guangxi, China and using them as a substitute for morphine during tests on the mice-jumping model. Methods Venom of King Cobras was purchased from Guangxi Province. Male and female mice were provided by the feed lot in our institute. Naloxone and morphine were purchased from Qinghai Medical Factory. CM-Sephadex C-25, Sephadex G-50, and CM-Cellulose 52 were provided by our pharmacy. Other chemicals of A. R. grade were produced in China. 1. The purification of two neurotoxins from the venom of King Cobras: The dry venom powder was dissolved in the buffer solution (pH 5.8, 0.05 M HAc-NaAc buffer), then the solution was centrifuged to discard the insoluble material. The supernatant was loaded on a Sephadex G- 50 column (25 x 200 cm). The same buffer was used to elute the column. The fractions which contained only low molecular weight components were collected (The process was directed by acidic PAGE). The collected material was desalted and concentrated. It was then loaded on a CM- Sephadex C-25 column (4x80 cm). The column was eluted first by the equilibrium buffer, then was eluted by the buffer 1993 by Asiatic Herpetological Research Vol. 5 p. 144 Asiatic Herpetological Research December 1993 {0 ;in 1 Neurotoxin 2 Amino Acid Min. Residue Residue Number Min. Residue Residue Number Numbers 100 Residues Numbers Number Lye 6 6.81 4 7.55 His 0 0 1 1.89 Arg 4 5.26 1 1.89 Asp 8 10.53 6 11.33 Thr 9 11.84 2 3.77 Ser 3 3.95 5 9.43 Glu 5 6.58 2 3.77 Pro 7 9.21 4 7.55 Gly 4 5.26 4 7.55 Ala 4 5.26 3 5.66 Val 6 7.89 2 3.77 Met 1 1.32 0 0 He 3 3.95 3 5.66 Leu 1 1.32 3 5.66 Try 2 2.63 0 0 Phe 2 2.63 4 7.55 1/2 Cys 8 10.53 7 13.21 Try 3 3.95 2 3.77 Total Residue 76 53 HAc-Ac buffer), loaded on the Sephadex G-50 column, then eluted with the same buffer. The fractions containing neurotoxins were put together, m desalted, concentrated, then loaded on the CM- sephadex C-25 column, which was eluted first with equilibrium buffer, then with different NaCI gradient. After the CM- Sephadex C-25 column, the two neurotoxins were further purified on a CM- Cellulose column. The crude venom was isolated first on a Sephadex G-50 column, then CM- Sephadex C-25, and finally, a CM- Cellulose column. Figure 1 and 2 shows the acidic PAGE of different fractions. The amino acid sequence of neurotoxin 1 was determined by the Immobilized Phase Edman Method as: 1 Thr • Lys • Cys • Tyr • Val • Thr • Pro • Asp • Val • Lys • Ser • Glu • Thr • Cys • Pro • Ala • Gly • Gin • Asp • Leu • Cys • Tyr • Thr • Glu • Thr • Trp • Cys • Val • Ala • Trp • Cys • Thr • Val • Arg • Gal • Lys • Arg • Val • Ser • Leu • Thr • Cys • Aal • Ala • He • Cys • Pro • He • Val • Pro • Pro • Lys • Val • Ser • He • Lys • Cys • Cys • Ser • Thr • Asp • Aal • Cys • Gly • Pro • Phe Asn • Val • Arg Pro • Thr • Trp • Pro The toxicity of neurotoxins 1 and 2 was determined by the following method: The mice, with body weight of 18-20 g, were divided at random into groups of 5 mice. The different doses of the neurotoxins were given to the groups S. V. The death numbers of mice within 24 hours were recorded and the LD 50 was determined by modified a Ginsberg method. For neurotoxin I the LD 50=0.21±0.013. For neurotoxin 2 the LD 50=0.24±0.009 For the morphine withdrawal jumping model, the effects of purified neurotoxins on mice jumping numbers and the results of statistical analysis are shown in Table 2. The results demonstrated that the two neurotoxins display very significant effects (P<0.01) as a substitute for morphine, by injecting S. V. and significant effects (P<0.05) when administered orally. Discussion The origination and evolutionary relationships of neurotoxins and Vol. 5 p. 146 Asiatic Herpetological Research December 1993 TABLE 2. The effects of the two neurotoxins on mice jumping test. Administered S. V. Administered Orally Groups Dose No. of jumps Statistics Dose No. of jumps Statistics Saline 0.1 ml 44.4 Neurotoxin 1 0. lug/20 g 11.3 p<0.01 Neurotoxin 2 0.1 ug/20 g 19.3 p<0.01 0.1 ml 44.4 1.0 u,g/20g 18.2 p<0.05 1.0 ug/20 g 16.8 p<0.05 phospholipase A2 from snake venom are still disputed problems. Some experts thought that the original molecule is the postsynaptical toxins containing only 60-61 amino acid residues. By increasing the cycles in the molecules, the short-chain neurotoxins evolved into long-chain neurotoxin containing 7374 amino acid residues, then evolved into phospholipase A2. On the other hand, some experts thought Phospholipase A2 was the original molecule which diverged into neurotoxins (including postsynaptical and presynaptical toxin), proteinase inhibitor and a new neurotoxin (dendrotoxin). In King Cobra venom, both short and long chain neurotoxins were isolated. Especially, a postsynaptical toxin containing 53 amino acid residues was purified. So this neurotoxin is very important to understand the evolutionary relationship of the neurotoxin. Now, we are focused on the determination of its sequences and its conformational properties in the solution by means of 2D- NMR method (mainly by different correlated spectrum and NOSY spectrum). By using the withdrawal jumping model, two neurotoxins display significant effects on substituting for morphine. The clinical observation in Kunming also demonstrated remarkable effects. These facts suggest that these two neurotoxins from the King Cobra venom may be a better medicine to be used as a substitute for morphine. Now, experiments have been accomplished to understand its mechanism. References GU, H. Y. 1965. Pharmacology (Vol. 1) edited by C. S. Zheng. People Health Publishing House. 393 pp. (In Chinese) JOUBERT, E. J. 1973. The amino acid sequences of two toxins from Opliiophagus hannah venom. Biochemica et Biophysica 317:85-93. LIN, N. Q. AND W. Y. WANG. 1984. [The amino acid sequence of the neurotoxin from the venom of King Cobra in Guangxi Province, China]. Acta Biochemica et Biophysica Sinica 16(6):592-596. (In Chinese). SHUN, X. 1981. [The Study on the purification of four neurotoxins from the venom of king cobra (Ophiophagus hannah)]. Zoological Research 2(l):260-270. (In Chinese). TU, A. T. 1977. Venoms: chemistry and molecular biology. New York. XIONG, Y. L. 1990. The study on given up addiction by using snake venom. ABSTRACTS of Second International Congress of Ethnobiology. XIONG, Y. L. AND W. Y. WANG. 1987. [The exploiting on snakes resources.] Symposium on Exploiting Yunnan Biological Resource. Yunnan People's Publishing House. (In Chinese). I December 1993 Asiatic Herpetological Research Vol. 5, pp. 147-165 The Ecology of the Caucasian Salamander (Mertensiella caucasica Waga) in a Local Population DAVID N. TARKHNISHVILI1 AND IRINA. A. SERBINOVA2 'Institute of Zoology, Georgian Academy of Sciences, Tbilisi, Georgia ^Moscow Zoo, 123242 Moscow, Russia Abstract. -The different aspects of ecology of Mertensiella caucasica Waga, 1876 were investigated in a local population from Borjomi Canyon (central Georgia) for five years (1985-1990). The aspects of the species' life cycle were more precisely determined. The main fecundity is about 16.9 eggs per female. There are about 2 years in a period from egg deposition (June to first half of July) to the end of metamorphosis in nature. Animals have spent most of the time in shelters after metamorphosis. They appear on the ground surface at night during the breeding period. Commonly the adults don't retreat to a great distance from population localities. Localities are situated in comparatively small plots (100-300m) along the streams. Estimation of adult animal number showed that the population consists of 1189 specimens (1989). Annual adult survival is higher than known values of most amphibians (approaches 0.77). Larval survival is 0.27-0.32 in the second year of life. The characteristics of demography (especially, low renewal rates) and spatial restriction in localities depends mostly on subUe constitution of the species (which is a result of allometric growth specifics). The small recent geographical range of M. caucasica is explained as a result of morphological and ecological peculiarities. General morphological constitution limits adaptive possibilities of any particular representative of die European salamander tribe. This is an explanation of quite high ecological similarity of M. caucasica and Chioglossa lusitanica. Key Words: Amphibia, Caudata, Salamandridae, Mertensiella caucasica, Caucasus Mountains, Georgia, population ecology. Introduction Natural populations are the single way of species existence. Autoecological research doesn't allow a complete understanding of the life of a species in nature. That is why there must be information of life cycles, geographical range, population size, number dynamics, etc. On the other hand it is hard to explain ecological aspects of the species existence without any information of their habitat preferences, feeding habits, breeding sites, etc. By analyzing connections between species population ecology and autecology, as well as morphology and geographical distribution, the most complete notion can be formed. Investigations on some amphibian species biology have allowed scientists to elaborate complex works connected with different aspects of their life history. A wonderful example is Bell's works on the Smooth Newt (Bell and Lawton, 1975; Bell, 1977). There aren't many data of regular stationary investigations about the ecology of the rare or narrow-ranged species. Long term research of such species enlarges the knowledge of the biology of wide taxonomic groups. Moreover, these investigations may be useful to find out ways of rare species preservation. A local population of the endemic salamander, (Mertensiella caucasica), from the western Caucasus of Georgia has been investigated for a five year period (1985- 1990). This work gives additional information about the life history of this species. The geographical distribution of the Caucasian Salamander was mainly established in the beginning of this century. Information was summarized by Nikolsky (1913). Later investigations commonly took place in earlier reported localities or in adjacent areas. Some new localities for salamanders were found by Bakradze and Tartarashvili (pers. comm.). The real geographic range of M. caucasica was established. The Caucasian Salamander is © 1993 by Asiatic Herpetological Research Vol. 5 p. 148 Asiatic Herpetological Research December 1993 FIG. 1. Distribution of Mertensiella caucasica. distributed in external spurs of the Trialetian Mountain Range. Probably it is the result of historical changes in the Kura River bed (Fig. 1). Populations are mainly distributed in the forest belt, but in some places they can be found close to subalpine meadows. Humidity in the species' locations reaches 1000 mm or more per year (another narrow-ranged representative of the salamander tribe, Chioglossa lusitanica, has similar requirements of humidity). In the most dry part of the range of M. caucasica, the eastern one, salamanders live only in coniferous forest. When humidity reaches 1200 mm/year in the middle area, they can also be found in subalpine meadows. Salamanders are distributed in deciduous forest only close to the Black Sea coast, where humidity is very high (2000-2400 mm/year; Fig. 1). The high dependence of the animal on humidity does not itself limit the species distribution, but determines sensitivity of specimens to other environmental factors. It is very interesting that the rheophilous species Ranodon sibiricus, more restricted to water habitats than M. caucasica, is geographical limited by coniferous forests like M. caucasica in eastern localities (Paraskiv, 1953). Local populations, distributed along tributaries of the Chorokh and Kura rivers (in upper flow), are formed by salamanders within its area. Width of streams in salamander plots is not more that 1-1.5 m in spring and because of stepped disposition of streams, they run slowly in some places. There are many slowly draining pools about 20-30 cm in depth with a lot of shelters. The bottoms of streams and pools are covered with stones, and there is a lot of non- decayed organic matter. Stepped disposition of streams is formed by stoned conglomerations and fallen logs. Apparently, mountain ranges between stream canyons don't allow wide salamander migration and local populations are comparatively isolated. There is no evidence that direct migrations of animals occurs during their life cycle. Individuals are found a maximum of 200-300 m distance from streams. December 1993 Asiatic Herpetological Research Vol. 5 p. 149 yyj — stream bed and branches -*S — fallen trees and piles of stones O — large pools where larvae were found \ — steep banks where shelters are located FIG. 2. Schematic diagram of the study site for Mertensiella caucaska. Tfie Study Area The studied population inhabits coniferous forest ecosystems along the second range tributary of the Kura River in Borjomi Canyon (eastern part of the species' range) (Fig. 1). The plant association is formed by Taxus baccate, Picea orientalis and deciduous spots. The size of the inhabited location is a bit more than 200 m, and it is situated between 1000 and 1300 m altitude, about 2 km from the stream mouth. Slopes are precipitous, built by corrosion of underground tree roots or relatively gradual, partially covered by pteridium, Matteuccia struthiopteris, from the adjoining stream banks. There are some stone conglomerations and fallen trees in the study area, shown on the map (Fig. 2). Air temperature is close to stream water temperature (13-15°C in summer) in shelters formed by stones and logs. Dynamics of air temperature in Borjomi Canyon in May to July, 1989 is shown in Fig. 3. Quiet pools and shelters are relative rare, slopes are steeper and stream flow is faster at upper and lower localities. Density of salamanders here falls rapidly as well as away from the stream banks in these places. Methods The main quantitative data were obtained during excursions with a lantern after sunset along the study area. The location of each adult animal was mapped, substrate type and distance from stream bank (more or less than 50 cm) was recorded. Adult animals were marked individually by toe- clipping. Combinations from clipped digits in hind-limbs (not more than 2 in 1 foot) responds to individual number of animals from 1 to 99. Zero-1 clipped digit in the front leg mean number of hundred. Marks of salamanders recaptured in the next year were renewed. Data of capture-recapture were statistically counted as in Kaughley (1977). Substrates of animals caught were subdivided in 6 types: shallow water; sand and pebbles above water shore; wet stones; wet ground; moss or lichens; dry ground and stones. These types were ranked according to their humidity. Basic investigations were conducted on 8-10 and 21-23 June, 1986, 24-28 June, and 5-7. August, 1987, 3-5 and 21-24 July, 1988, 16 June- 12 July, 1989, 2-9 July, 1990. We had 337 contacts with males and 202 with females (including specimens found two or more time). Recording of larvae was conducted during night excursions. We caught females from nature in the reproductive period and obtained eggs using a hormonal stimulation method (Gontcharov et al., 1989) to study some ecological and morphological features of early development. Eggs were incubated in Weiss bowls in dechloronated water at a temperature of 14°C as well as in aquaria at Vol. 5 p. 150 Asiatic Herpetological Research December 1993 Date FIG. 3. Air temperature at the Mertensiella caucasica study site during the period of reproductive activity. Stippled bars represent periods of rainfall. 40 -i i- 30- X> S 20- 3 Z 10- Males Females 12 3 4 YV~ i-2 Substrate 5 6 FIG. 4. Location of salamanders by substrate type. 1- shallow water; 2- wet sand and pebbles; 3- wet stones; 4- moist ground; 5- moss or lichens; 6- dry stones. Solid bars represent males. Stippled bars represent females. a temperature varying from 5-22°C. Before completion of metamorphosis, larvae were kept in 20 liter aquaria, where water was changed every third day. Food consisted of crustaceans (Daphnia, Cyclops), Tubifex and Chironomid larvae. Morphological studies were conducted on larvae, juvenile and adult animals with a binocular magnifies and calipers. Their snout-vent length (L), head length (Lc), and tail length (Led) were measured with a precision of 0.1 mm. The coloration patterns of some animals was also recorded. Results The niche of larvae and adult specimens. Salamanders don't have an even distribution within the study site. Preference to every substrate depends on the amount of moisture of each particular substrate. Frequency of captures decreases with distance from water or potential shelters. The number of animals captured out of shelters depends on time and season. Most adult specimens were recorded close to the stream (less than 50 cm from the water shore): 60±4% of males, 62±5% of females. In contrast to data on the ecologically similar species, Chioglossa lusitanica (Arntzen, 1981), there is no difference between male and female attachment to water in M. caucasica. Animals commonly may be found on wet sand or stones at the water shore, and they avoid dry soil and stones. The distribution of substrate type of captured animals for the December 1993 Asiatic Herpetological Research Vol. 5 p. 151 30- O— 1986 »-- 1987 A— 1988 a r"1 Males 20- • 1989 \ : i o 10- A w i i i 0 >J ° • "tX) ^ ^-^o. 3 z 20 10- Females ^KS^Ll ••*-*.. tC» 10 20 30 1 10 Date FIG. 5. The seasonal dynamics of salamander occurrence. period from 1988-1989 is presented in Fig. 4. It is obvious that females avoid open rocky plots, fallen leaves or moss cover. According to our observations, adult salamanders spend only a small part of their life on the surface of the ground. Even during the active period, only a small part of the population leaves their shelters at night. Apparently, salamanders spent the rest of the time in shelters, where they live and feed. Seasonal dynamics and diurnal activity are reported on a number of captures during excursions in a 4 year period (Fig. 5). The number of animals found above ground decreased in July. The decrease in capture was especially sharp in the 5-15 July period. Even after rains, individuals could hardly be found in the second half of July (Fig. 5). Earlier seasonal activity of males is noted for other flowing-water (Arntzen, 1981) and standing-water (Golubev, 1981; Beneski et al., 1986) tailed amphibians. According to phenological data the large number of salamanders found in June is connected with their reproductive activity. The beginning of the reproductive period is determined by air and water temperature. The active season begins when minimal night temperature is about 15°C (Fig. 3). After mating and egg deposition, activity decreases. Apparently, the decreasing number of specimens found in July does not depend on regular migrations. Observations of adjacent parts of the stream 100 E 3 z 50 Females 23 24 1 22 23 24 Hour of occurence FIG. 6. Number of salamanders observed by hour of day. did not have any result in late July and August. The decrease in animal numbers either depends on the dispersion of individuals in the forest (or along stream banks, as proposed by Arntzen, 1981, for C. lusitanica), or more probably, they are in shelters most of the time because of greater abundance of their food, such as Gammaridae and Lumbricidae, there. The main nocturnal period of activity of M. caucasica is between 2200 and 0100. No active salamander has been found before 2130, and active animals have been rather rare before 2230. The number of active animals decreased after 0130-0200. The peak of activity was observed at about 2300 (Fig. 6). Other nocturnal tailed amphibians also have a short active period and the time of activity is species specific (Semlitch and Peachmann, 1985). In the summer, Vol. 5 p. 152 Asiatic Herpetological Research December 1993 D T3 6- y 5- U 5 4 3 1 A - 2 FIG. 7. Consecutive stages of courtship and amplexux in Mertensiella caucasica. Clutch Number FIG. 8. Inter and intra-clutch variability in fertilized eggs of Mertensiella caucasica. salamander larvae can be found in some parts of the stream bed in small pools with slowly drying water. Outside of the shelters, they are mainly in shallow water with a depth of less than 5 cm. Generally, the number of larvae doesn't exceed 10 individuals (maximum 14) in a pool. Comparatively large premetamorphosed larvae can be found in the stream. They move actively along the stream. Like adults, larvae spend most of the time in shelters. The first individuals can be seen in open water in late May. Larvae leave their shelters at twilight, when absolute sunlight is about 10 Lk. There are no nitrates in the water composition in salamander breeding and developing sites. The pH reaches 7.8-8.3 and the hardness of the water is 0.6-2.8 mg/equivalent 1. Juveniles rarely leave their shelters during the period from the end of metamorphosis until first breeding. The diet of larvae, juveniles, and adults reflects their biotopical preferences and does not show specialization in any invertebrate group (Kuzmin, 1992). Both terrestrial and aquatic organisms are found in the diet of adult salamanders (Ekvtimishvili, 1948; Kuzmin, 1992). Only terrestrial organisms are found in the diet of juveniles (Kuzmin, 1992). Space used by salamanders. The requirements of salamanders to environmental characters limit the space they use even within a local habitat. Thus, salamander distribution isn't homogeneous December 1993 Asiatic Herpetological Research Vol. 5 p. 153 in a given locality. Obviously, animals prefer places with plenty of shelters. The mean number of salamanders captured during night excursions per 10 m along the stream bank in 1986, 1987, and 1989 was 9.2 ±1.86 males and 5.26+1.13 females. According to the "mean crowding" index of Lloyd (1967) (m=m+2^ -1) the mean value is respectively 21.8™ and 12.78. The highest density was observed in places where there were logs and wooden blocks, combined with stone conglomerations, and a lot of small pools and shelters under tree roots. The total number of adults recorded for the 1986-1990 period, including recaptures was 455. The place where each specimen was found in 1989 is noted in Fig. 2. This data could give a notion of real space distribution of animals. Note that larvae can be found outside of the local population habitat significantly more often than adults. This is connected with the fact that some of the larvae leave the shelters and continue development in the lower parts of the stream (see below). Apparently, this process does not disturb normal metamorphosis and juvenile animals return to the population locality. Life Cycle. The salamander breeding period in investigated habitats occurs from the second half of June until early July. Most of the females found in July are ready for egg deposition. The large oocytes can be observed through the transparent ventral skin. There are well distinguished mating corns at the adult male shoulders. Amplexing animals were found on the ground close to shelters. Cyren (191 1) and Obst and Rotter (1962) described normal sexual behavior of salamanders in water in natural and laboratory conditions. We observed normal sexual behavior twice: on 28 June, 1988 and 4 July, 1990. In the first case it took place about 2 m from the stream bank in a conglomeration of tree roots. In the second case it happened close to water, at the entrance of a rock chink. We don't exclude the possibility of normal copulation in water. For example, mating of C. lusitanica may take place both in streams and on the shore. The consequent states of courtship and amplexus are shown in Fig. 7, b-f. According to our observations, sexual behavior of M. caucasica is similar to that of Salamandra salarnandra (Joly, 1966). The corn on the dorsal side of male tails has no special role in courtship and amplexus. We have also observed an attempt of copulation (Fig. 7h) on 28 June, 1989 and 2 males in an amplexus pose on 8 June, 1986. Apparently, we have observed, in the latter case, rival combats, described for S. salamandra by Kiistle (1986), but in this work the behavioral display is not the same. Copulated females have a slightly opened cloaca. There are more than three days between copulation and egg deposition. Each female deposits from 1 1 to 24 eggs (N=9, M=16.9, a=3.9). Inter- and intra-clutch variability of fertilized eggs sizes is shown in Fig. 8. Darevsky and Polozhikhina (1966) found that the sizes of 90 eggs found in nature ranged from 5.0- 5.6 mm. Females deposit separate eggs, sticking them to the substrate in shaded places. Activity of animals gradually decreases after the completion of the reproductive period. Egg development takes 45 days until hatching in aquaria, where the average temperature is 14.8°C. When the temperature changes from 6° to 26° (M=16.5, o=4.7) development is extended to about 48-5 1 days. We can expect similar developmental rates in nature, when the temperature of the water is about 14-15° in July to August. The hatching of most part of the generation takes place not earlier than late August. Larvae found in June can be divided into 3 groups on the basis of snout- vent length: the I group- L=14.6-19.5 mm (M=106-180 mg); the II group- L=23.7- 27.5 mm (M=330-664 mg); the III group- L=29.0-35.4 mm (M=632-1400 mg). By virtue of larval size distribution, Freytag (1954) as well as Koroljov (1986) concluded a 3 year period of larval development in the Caucasian Salamander. Kuzmin (1992) established no annual ring in the I and II larvae group hip bones and only one annual ring in the III group larvae Vol. 5 p. 154 Asiatic Herpetological Research December 1993 YEAR 4 YEAR active HHH hibernation loo! egg deposition YEAR 2 | embryonal development j^s hatching ci=i^ metamorphosis FIG. 9. The reproductive cycle of Mertensiella caucasica. A- underground parts of the stream bed; B- open water; C- surface and temporary shelters on die ground; D- constant shelters on the ground. hip bones. On the basis of this information Kuzmin (1992) supposed that larvae of I and II size groups had most probably hatched in the given year. Nevertheless, analysis of time of reproduction and embryonic development during the year opposes Kuzmin's opinion. ApparenUy, salamander larvae remain in shelters after hatching and go out when water temperature increases al least to 13°C (at the beginning of the next summer). Lack of annual ring on hip bones can be explained by incomplete development of hind legs just after the hatch. Larvae growth is delayed by autumn temperature decreasing. Thus, animals of the I and II size groups have a previous year hatch and represent a single generation. Size differences within a generation are formed by prolonged breeding time (not only between breeding locations [Mertens, 1968], but within populations, too) and/or by variation of individual growth rates. Larvae have developed during the warm period of the second year. After hibernation, they have a metamorphosis in July-August of the 3rd year. Their development from fertilized egg to completed metamorphosis takes about two years in nature. Salamanders have a concealed life during the period after metamorphosis and before maturity. According to Kuzmin (1992), 3- December 1993 Asiatic Herpetological Research Vol. 5 p. 155 5 annual rings can be observed in hip bones of adults. Therefore, salamanders can first breed in the 3rd year after metamorphosis. The total life cycle of the Caucasian Salamander from egg to egg is about 4 years (Fig. 9). Growth and Development Embryogenesis of M. caucasica is similar to other large-size egg amphibian development (Fig. 10 a-c). Analyzing experimental observations, hatching takes place when total length is 17-20 mm (L=10.5-11.5 mm). When hatching starts, a larvae has well developed external gills and a tail fin. Sometimes the rudiments of 3 toes can be distinguished on the hind legs. Pigmentation is formed by a couple of faded pigmented stripes. There are rare individual melanophores on the surface of the stripes. A line of small circular non- pigmented patches lays along each stripe (Fig. 10 d). The gut is filled with yolk. The total length of the smallest larva caught in nature was at least 25 mm (commonly, L=15 mm, minimum = 14.6 mm). Larvae found in nature already have no yolk in the gut. Their external gills are smaller than those of animals that have not yet hatched. There are 4-5 toes on the hind legs and more pigment cells (Fig. 10 e). In laboratory conditions, when temperature is 14.8°C, yolk disappears from the frontal part of the gut 16 days after hatching at a length of 13.8114 mm. The first larvae with a snout-vent length of 15- 16 mm can be found in streams in early June, but most of them appear in July. The small larvae, which have over wintered, appear in stream pools in small, probably sibling groups. The largest group (8 larvae) was found on July 5, 1988. Individual sizes in that group provides some information about intra-clutch larval size variation: when L=17.35±0.48, min=16.0, max = 19.2, coefficient of variation approaches 7.8, L total=29.25±0.60, min=27, max=31.3, CV=5.8%. Summer growth of first hibernated animals (in 1985) is shown in the histograms of June and August larval size distribution (Fig. 11). Mean total FIG. 10. Embryo and larval development of Mertensiella caucasica. length increases 6.27 mm for 70 days and the specific growth approaches 0.24%/day. At the same time homogeneity of generation increases: CV=23% in June and becomes 9.9% in August. This process is probably caused by more rapid growth and/or comparatively high mortality of the small larvae. Morphological changes, connected with metamorphosis (i.e. yellow coloration of unpigmented spots, decreasing of gill size, reduction of tail fin- Fig. 10 f) began in animals with at least 30 mm snout-vent length. They approach that size in the 3rd year of larval development. Comparing sizes of the 2nd and 3rd year larvae, the specific body growth rate is 0.12%/day in the period between August and June of the next year. Commonly, metamorphosis takes place at a snout- vent length of 30-35 Vol. 5 p. 156 Asiatic Herpetological Research December 1993 10 1 5- June Males ru o-1 10 kLi xi —I T" 30 40 50 60 70 £ s Z August r^M ■■ ~i 1 1 1 1 t~ 30 40 50 1 1 1 60 70 L (mm) FIG. 11. Size distribution of Mertensiella caucasica larvae at the study site in 1985. mm. Maximal larval size approaches 35.4 mm (L total=70.3 mm) in the population studied. Snout-vent length of metamorphosed animals varies within the limits of 32.9-42.4 mm (body mass, M=694-1592 mg, N=ll). At the same time 2 larvae, with L=43.8 and 44.6 mm, are in a collection from the surroundings of Batumi (Kuzmin, pers. comm.). Animals of all three size groups can be found in streams even in May (Korolyov, 1986). Peculiarities of natural growth of M. caucasica are similar to C. lusitanica. Although this species passes metamorphosis at smaller sizes (i.e. L=24- 25 mm) their linear growth for two summer months approaches 0.29%/day and 0. 10%/day for the rest of the year (Arntzen, 1981) and it is very similar to the analogous index of M. caucasica. The slow growth of salamander larvae is mainly the result of low water temperature in streams. The specific total length growth rate of a single animal (from 27.9 to 56.7 mm) was 0.54%/day, and total length increased from 37.0 to 59.2 mm was 0.42%/day under laboratory conditions, at 23-25°. The snout-vent length of adults varies insignificantly. Data on animals measured X> £2 N=10 x = 71.1 6 = 8.4 SE = 2.65 Females 50 I H 60 L (mm) 1 1 m i 70 80 FIG. 12. Size distribution of adult Mertensiella caucasica in the study area. L- snout-vent length. in the studied population are shown in Fig. 12. There aren't considerable intersexual differences in sizes and general body proportions, but apparently females can begin breeding at a smaller body size. Other authors (Cyren, 1911; Knoblauch, 1905; Nesterov, 1911) reported that mean L in males approached 68.9 mm (N=7) and females, 63.5 mm (N=ll). The animal body length of an outlying population (Goderdzi Mountain Pass) varies between 68-77 mm in males and 56-73 mm in females. However, our population does not show any specifics in adult animal size distribution. We will briefly discuss morphological changes in the period from die beginning of active feeding to the end of metamorphosis. When larvae begin to feed, melanophores gradually disperse from the lateral sides, filling the ventral surface of the larval body. Even the size II larval group have only a narrow non-pigmented stripe remaining on the ventral side. All lower surface is filled by pigmented cells and non-pigmented patches remain only on the lateral sides of the size III larval group. These patches are used as a substrate of xanthophores and iridiophors, forming yellow spots later on (Tarkhnishvili and Tartarashvili, 1987). The intensity of basic coloration is correlated with the size of the animal that has already started metamorphosis. The animals with a large size at the beginning of metamorphosis have a dark-brown (not as dark as in spotted salamander) coloration with bright and comparatively large yellow spots. Smaller size larvae do not have such December 1993 Asiatic Herpetological Research Vol. 5 p. 157 an intense basic coloration and spot pattern is more or less reduced (spots are smaller and/or poorly expressed). The intensity of salamander pigmentation, like other tailed amphibians, may vary depending on the light intensity at the larval location (Fernandez and Collins, 1988). Ground coloration of adults varies from reddish-brown (similar to Chioglossa lusitanica or some M. luschani subspecies (Winter et al., 1987) to dark brown. The spotted pattern may be expressed in a different degree to full reduction (especially in light colored specimens) (Fig. 13). Poorly pigmented animals with comparatively reduced spots (described by Tartarashvili and Bakradze (1989) as the subspecies M. c. djanashvilii- Fig. 13 a) predominate in some populations from the surroundings of Batumi, at the Black Sea coast. Nevertheless, dark colored animals, with well developed spots (Fig. 13 e, f) predominate in the population from the subalpic zone (Mountain Pass Goderdzi, in Bakradze's collection). Salamanders with an intermediate intensity of coloration are more abundant in our studied population, but there are some specimens with less or more reduced spot pattern. There is also a female, colored as the form described by Tartarashvili and Bakradze. Probable, the specific coloration of adults is connected with the character of larval development, which depends on the special climatic conditions of each habitat. That is why many light-colored animals occur in the warm sea cost habitat and dark colored ones are found at high altitudes. Populations from Borjomi Canyon are in an intermediate place. Or course, we don't exclude the possibility of inheritable fixing of one or another coloration type in different populations. The Caucasian salamander is included with the Luschan Salamander, M. luschani, in the same genus because of the tail corn, the secondary sexual character of males (Ozeti, 1967). This character appears in males with a length of at least 130 mm and it seems to be of no functional importance as some investigators have proposed, for example Cyren, (191 1). FIG. 13. Color variation in Menensiella caucasica. The main changes of general body proportion occur during ontogenetic development. First, relative tail length increases after hatch. Mertens (1968) reported this for larvae with a total length of more than 45 mm. Based on our data, comparatively rapid tail growth begins at the earliest stages of development and extends to the adult stage. On the other hand, comparative length of head decreases (Fig. 14). The changes of general proportions have different intensity in different developmental stages. Allometrical dependence of head and body length on the total length of the I and II size group larvae is described by equations: Lcd=0.35L126 Lv=1.71L°-26 The coefficients of allometric equations for the III size group larvae is different: Lcd=0.08L17' Lc=1.5L°-52 Vol. 5 p. 158 Asiatic Herpetological Research December 1993 2.0 1.5 Led 10 0.5 0J ■ Spotted Salamander D Caucasian Salamander 10 _L Lc Stage of Ontogenesis Fig. 14. Changes in body length proportions in Mertensiella caucasica (open square) and Salamandra salamandra (solid square) during ontogenesis. 1- larvae with total length less than 35 mm; 2- larvae with total length greater than 35 mm; 3- yearlings; 4- juveniles; 5- adults, solid line- L/Lccj; broken line- 1/LC. The coefficients of static allometrical equations for recently metamorphosed animals are: Lcd=0.23L142 Lc=1.04L062 Hence, the most rapid comparative increasing of tail length is during late larval development. The comparative decreasing of head is the most rapid in the I and II size groups, but later this process isn't so clear. Nevertheless, it takes part in metamorphosis. Plenty of eco-morphological features separate M. caucasica from the other representatives of European salamander tribe, depended on subtilization (Ozeti, 1967). The latter is a base determined ecological similarity between M. caucasica and C. lusitanica (Borja-Sanchiz and Mlinarsky, 1979). Perhaps this is a reason of similar breeding ways of these species differing from other European salamanders. The changes of general proportion of M. caucasica and S. salamandra, which are shown in Fig. 14, allow a comparison of these species. The general trends of body proportion changes are common in the two species (as in most Tetrapoda): the comparative length of head (L/Lc) decreases and that of tail (L/Lcd) increases. But in these tendencies, both are quite rapid in M. caucasica. In S. salamandra the changes are gradual and moreover, tail length growth is poorly distinguished (Fig. 14). Population Number Dynamics and Regulation. The analysis of our 1986-1990 capture- recapture data allow us to study population number, number dynamics and demographic peculiarities. We obtained quite full information in 1989. In that year the captured animal number, in relation to real population number, was comparatively high. The general picture of results is shown in Table 1. We had 532 contacts with animals. We met an animal twice in the same night only on 7 occasions. They were on the surface, essentially not moving. In all cases animals were captured again in the same plot, not more than an hour later. The 50 animals of 68 recaptured December 1993 Asiatic Herpetological Research Vol. 5 p. 159 TABLE 1. Mark-recapiure results for Mertensiella caucasica in 1989. Date i nj Ri Ma les 6-16 1 5 5 i 6-17 2 11 11 0 II 6-22 3 27 26 0 1 III 6-25 4 15 7 0 1 1 rv 6-26 5 26 29 0 2 5 0 V x'j 6-28 6 28 28 2 2 1 2 2 VI 6-29 7 13 13 1 0 1 0 0 0 vn 7-1 8 18 18 0 1 2 0 1 1 0 vra 7-2 9 8 8 0 0 2 0 1 1 0 0 rx 7-4 10 7 7 1 1 1 0 0 0 0 0 1 X 7-10 11 4 4 0 0 0 0 1 0 0 0 0 1 XI 7-12 12 2 2 0 0 1 0 1 0 0 0 0 0 0 n 4 8 14 2 6 2 0 0 1 1 0 Nj 579 619 349 681 104 44 SEj 151 376 135 474 97 25 Pi,i+1 1 1 1 0.22 (P6.9) 0.45 Ai,i+1 70 (A3.5) Dale i 'M Ri Females 6-16 1 I 6-17 2 0 II 6-22 3 0 0 HI 6-25 4 0 1 0 rv 6-26 5 0 0 0 0 V x'i 6-28 6 1 1 0 1 0 VI 6-29 7 0 0 0 0 1 0 vn 7-1 8 0 1 0 0 2 0 0 VIII 7-2 9 0 0 0 0 0 1 0 0 rx 7-4 10 1 0 0 0 1 2 0 0 0 X 7-10 11 0 0 0 0 0 0 1 0 0 0 XI 7-12 12 0 0 0 0 2 0 0 0 0 0 0 n 2 3 0 1 6 3 1 0 0 0 0 Ni 78 340 497 SEj 66 200 476 97 25 Pi.i+1 1 1 Ai,i+1 70 265 157 (A3.5) (A4.6) Note: i, j- the number of census; nj- the number of specimens examined in i-tli census; Rj- the number of recaptured and escaped specimens; X4_ [jj- the number of animals recaptured in the 4-th census that had been marked during the 3-rd census: rj- the total number of recaptured specimens that had been marked in the i-th census; Nj- the estimate of population size in the moment i, using the lolly-Seber method; SEj- standard error (lolly-Seber); Pj^ j- the probability of the individual remaining in the active part of the population between the i-th and j-th census; Aj_ j- the number of specimens supplemented the active part of the population between the i-th and j-th census. in a year of marking were caught in the same or adjacent plot (the place of 28 captured animals was not recorded). On the basis of these data, we conclude that these salamanders have a low moving ability and a short period of nocturnal activity. Only an approximate representation of activity dynamics can be given by captured animal number, though in some cases, data of captured animals are Vol. 5 p. 160 Asiatic Herpetological Research December 1993 used as an index of number (Bozhanski and Semenov, 1982). We used the Jolly-Seber method (see Caughley, 1977) to estimate the real number of the local population and its dynamics during the breeding period in 1989. The total number of breeding animals was 1187 and the percentage of males was 58±0.001. The greatest number of active animals was in the end of June. Females appeared a bit later than males. Unfortunately, the data of 1986-1988 don't allow us to correctly estimate the number of salamanders in those years. It varied from 10-20 to 460 individuals, when the errors exceeded mean values, i.e. significantly lower than real quantity. The highest value (460) was recorded when the Schumacher method was used on 1990 data. The values recorded for the 1986- 1988 data didn't exceed 200. Thus, the annual number value significantly increases when the research period is prolonged and captured specimen number increases. That is because only a small part of the population left their shelters, even in the highest activity period in late June to early July. The estimate of the C. lusitanica population (Arntzen, 1981) is different because of the permanent migration of part of the population. The number of two local populations of this species is 1236 and 1324 respectively. This is similar to our information about M. caucasica, moreover, the study sites have a size similar to ours. Usually, the number of animals in the widely distributed in Europe S. salamandra populations can exceed some thousand specimens (Klewen, 1986). Their populations are spread over several hectares, and animals can be found far from the breeding sites. The capture of salamanders marked in previous years gives some information about mortality rates of adults. Ninety eight males and 45 females were marked in June, 1989. There were 38.5% (OM=7.8%) recaptured males of the total of 39 found in July 1989, and 28.9% (OM=7.3%) recaptured females of the 38 total found. The 31.8% (OM=5.7%) of males and 17% (OM=5.9%) of females captured in July 1990 were marked in June 1989. Hence, survival rates of the period from July, 1989 to July, 1990 is Pm=31.8/38. 5=0.83 for males and Pm= 17/28.9=0.59 for females. The part of all marked adults was 33.8±9.3% in July, 1989 and 26.2±4.2% in July, 1990. Hence, the annual survival approached 0.77. It should be noted that only 8 individuals (9.0±2.0%) of all 89 marked in 1986-1988 were found again in 1989. This is quite a high number because 89 individuals aren't more that 10% of the adult population. Unfortunately, there is a lack of information about renewal rates of tailed amphibian populations. Ignoring age structure, the annual mean survival of Ambystoma maculatum approaches 0.72 for males and 0.60 for females (the mean male number is 641 and capturing of males is a bit more often (Husting, 1965). These data are quite similar to ours. On the other hand, annual survival of the Smooth Newt is only 0.45 for males and 0.55 for females in England (Bell, 1977). There are considerable low survival rates of Notophtalmus viridescens and some Anura when higher mortality of males is observed (Ischenko, 1989). Klewen's (1986) quantitative data for S. salamandra show that annual survival varies between 0.55- 0.81 (mean 0.66 in four years of investigation. A higher female survival was recorded for the genus Desmognathus (Husting, 1965). The Caucasian Salamander has a comparatively low population renewal rate, when mortality is low. Perhaps, this kind of population dynamics is typical for populations with low total number and high male survival. Organ (see Husting, 1965) mentioned that a higher male survival was a result of significant energy expenses of females during breeding. We can only indirectly estimate salamander mortality before mating. On the basis of adult female number and mean fecundity, we estimate that there were 8000-9000 eggs deposited in the study site. December 1993 Asiatic Herpetological Research Vol. 5 p. 161 TABLE 2. Data on mark-recapture of Meriensiella caucasica from 1986 to 1988. 1986 1987 1988 Sex Data 6-7 6-8 6-9 6-10 1 .-21 6-22 6-2 6-24 6-2: 7-3 7-4 7-5 7-22 ni 9 5 14 7 6 4 4 49 13 8 6 27 6 7 2 1 16 95 f Ri 9 2 8 7 5 2 3 36 10 8 4 23 5 4 1 1 11 70 mi 0 1 2 0 0 1 1 5 0 0 0 0 0 1 0 0 1 6 ni 2 5 5 2 2 8 2 28 2 0 5 7 6 8 1 0 15 48 m Ri 2 4 4 1 1 4 2 18 0 0 0 0 0 1 1 1 3 20 mi 0 1 0 0 0 1 0 3 0 0 0 0 0 0 0 0 0 3 Note: nj- number of individuals examined in the i-th census; Rj- number of recaptured and escaped individuals; m;- total number of individuals captured in the i-th census that were marked in the same year (but on another day). Nevertheless, significantly few larvae could be found in the stream when salamander density was the highest. Korolyov (1986) found only 90 larvae in 1984 in the whole area of the stream described here. Bozhansky and Semenov (1982) counted 1-33 larvae in every 700 m of flow in August. We found 1 16 larvae in the study site in June, 1985, of which 91 had already gone through hibernation. In the beginning of July, 1990, 74 larvae of the I and II size groups (hibernated once) were counted in the same plot. The number of second year larvae is quite constant in different years. Although it was considerably lower that the total number of eggs deposited, it did not have an influence on the real larval mortality rates in the first year. More probably, larvae were carried by flow along the stream and were distributed more uniformly than adults and eggs. On the base of the 2nd and 3rd year larvae proportion at the site, we were able to judge larval mortality from egg deposition to the second year. The twice hibernated were 27.8% (1985) to 31.9% (1990) of the total number of once hibernated specimens. Apparently these values express the real survival rates of a year (within twelve months). The stable larval density and ratio of the second and third year animals at the locality is a result of the stability of the adult salamander population number, and moreover, conversely to stagnant water amphibians, quite constant developmental conditions. The later is a reason of Caucasian Salamander number dynamics peculiarities. The basic reasons of natural salamander mortality are not completely clear. Perhaps, egg and larval mortality caused by ponds drying up is not as important for M. caucasica as it is for many amphibians (for example, Ambystoma maculatum, Albersetal., 1987). There are practically no predatory insect larvae dangerous to salamander larvae in the stream. Young trout (Salmotrutta labrax), which are possible predators, are also very rare. Perhaps the main reasons of larval mortality are over wintering and larval diversion into the stream flow. Grass Snakes (Natrix natrix) could of course cause great damage during metamorphosis. There were from one to five just metamorphosed specimens in the stomachs of the five Grass Snakes captured in the salamander locality. We did not find any adult salamanders in snake stomachs. It is unlikely the low vulnerability of adults is connected with autotomy ability or coloration. The juvenile animals do not have the same coloration as adults and they are able to autotomise also (Golubev, 1981). The reason is rather adult animal size and antipredator behavior of this species described by Brodie et al. (1984). Discussion According to the view of adaptionists, the peculiarities of the morphology of the Caucasian Salamander mainly are a result of general body constitution. The vulnerability of this species results in the high requirements to environmental Vol. 5 p. 162 Asiatic Herpetological Research December 1993 condition, especially temperature and humidity. The comparative large body surface reduces homeostatic ability. Another main feature of M. caucasica as well as of C. lusitanica delimiting these species from all other European salamanders is breeding by egg deposition, when fecundity is comparatively low. The biogeographical and ecological characteristics of M. caucasica can be explained by consequences of its morphological type. On the basis of paleontological date, Mertensiella aff. caucasica was distributed sympatrically with the Spotted Salamander in an area extending to central Europe in the Pliocene. The reduction of the range of Mertensiella was the result of the last glacial periods (Borja-Sanchiz and Mlinarsky, 1979). Nevertheless, the present range of S. salamandra is quite wide (Thorn, 1968), while the range of M. caucasica, like other representatives of the tribe, M. luschani and C. lusitanica, is comparatively narrow and in areas with mild climate. It is unlikely that the Spotted Salamander can affect the geographic range of these species. Although members of the genus Mertensiella are allopatric to S. salamandra, C. lusitanica has a wide sympatric zone with this species (Bas Lopez, 1984). The absence of S. salamandra in the Caucasus, including the Great Caucasus, obviously depends on historical reasons. The geographic range of subtle species is limited most of all by climatic factors. Wolterstorff et al. (1936) mentioned that the range of M. caucasica has not changed considerably since the Eocene. Perhaps the low homeostatic ability of adults limits their migratory possibilities. As we already noted, the captures of animals far from their population locality were very rare. The salamanders don't penetrate the comparatively distant mountain systems like the Great Caucasus. They also don't occur in comparatively dry localities along the Trialeti Mountain Range in the East (Fig. 1) where there is no relief limit. This is one of the reasons for the restricted salamander distribution. On the basis of different research (Obst and Rotter, 1962; Tartarashvili, pers. comm.) we conclude that the area of salamander localities of high altitude and on the Black Sea coast isn't larger than ours. Chioglossa lusitanica localities have a similar distribution (Arntzen, 1981). On the other hand, a small population area might depend on attachment to the breeding sites (stream plots suitable for egg deposition and larval development). Spotted salamander populations are always distributed in significantly wider areas (Klewen, 1985). As a result of the small area of the localities and the lack of breeding sites, the population number is limited at a comparatively low level, about 1000 individuals, when the sex ratio is close to equal. This amount is enough to maintain the populations demographically and genetically (Lande and Barrowclough, 1989). The potential population growth rate is also limited by comparatively low fecundity. Nevertheless, the breeding sites are used rather efficiently. Temporary ponds vulnerable to periodic natural disturbances aren't used for egg deposition. Thus, the population renewal possibilities of M. caucasica are different from stagnant- water amphibians. A significant part of the latter species are not able to breed efficiently because many of the breeding ponds within localities are destroyed during egg and larval development. Apparently the egg deposition of M. caucasica is done only in places suitable for further development. This type of reproduction is correlated with the high stability of population number though the resilience to habitat transformation is low. This described model of population dynamics practically excludes the number of outbursts caused by climatic perturbations which could stimulate considerable migrations. Since, settling of investigated species is determined by the low tolerance of adults, the small area of population localities and breeding sites, and the low fecundity. Hence the main reason for the narrow geographical range of M. caucasica are its morphological features and the stable type of population cycle. December 1993 Asiatic Herpetological Research Vol. 5 p. 163 In analyzing morphological reasons of Caucasian Salamander ecological specifics in comparison to related species, we can not accept as a main character only body proportions. Ecological particularities of all the European salamander tribe mainly depend on comparatively large egg formation (and probably skin structure). These characters limit European salamanders to breed only in flowing water in comparatively wet places. Although the European salamander adaptive type allows comparatively wide interspecies variability in some features, for example coloration patterns and degree of subtilisation, the general morphological constitution restricts adaptive ability of particular representatives of the group. In a sense M. caucasica is a morpho-ecological equivalent of C. lusitanica. These species have similar life cycle, population spatial structure and number dynamics, climatic and biotopic preferences, etc. The central adaptive possibility and the feature determined place in the group can be distinguished among a lot of morphological characteristics. There are some other features which separate these species and reveal the independent origin of both of them. This is, for example, tail corn in males. Nevertheless, this structure does not take place among main ecological features of species and reflects only the complexity of the phylogenetical ways. Color patterns have rarely been used in phylogenetical speculations, but this characteristic is a favorable object of adaptationists. The presence of light- colored specimens in some M. caucasica populations, their predominance in other populations of this species and in M. luschani and, finally, fully reduced of spotted specimens in C. lusitanica are not connected with variability of the plant cover in localities and do not affect their ecological preferences. We can't speculate about the adaptive meaning of coloration in this case. Coloration is closely related to the climate type of localities. Apparently, we could consider this characteristic as a fixed non-adaptive reaction to temperature and humidity changes. In conclusion, we would like to give an opinion on an interesting detail connected with the distribution of M. caucasica. Three anuran species, the Colchic Toad {Bufo verrucosissimus), the Caucasian Parsley frog (Pelodytes caucasicus), and the Asia Minor Frog (Rana macrocnemis) live sympatrically with the Caucasian Salamander. Rana macrocnemis is distributed all over the Caucasus. Bufo verrucosissimus and Pelodytes caucasicus like M. caucasica do not penetrate the eastern part of the Trialeti Mountains because of lack of humidity. However, they are distributed in some locations in the Great Caucasus. The northern parts of the Trialetian and Adjaro-Imeretian mountains are more that 50 km from the southern parts of the Great Caucasus in Central Georgia. There are no suitable localities for forest amphibians between these mountain systems. The comparatively small transitional zone could have been crossed many times by B. verrucosissimus and P. caucasicus after the Great Caucasus system was formed. If we take into consideration the comparatively high fecundity (about 500 eggs per year for P. caucasicus and 10,000 eggs per year for B . verrucosissimus) and the temporal variability of the breeding sites, a few climatically favorable seasons could cause a great increase in population number and as a final result, a massive migration. According to the peculiarities of M. caucasica population dynamics, we can not expect any similar process. Hence, the lack of M. caucasica in the Great Caucasus has historical rather than autecological reasons. Acknowledgments The authors wish to express their gratitude to Dr. S. L. Kuzmin for his great help in providing material and literature. We also thank Mr. R. V. Tartarashvili for interesting information and material. Dr. J. Ilieva helped us greatly to prepare the English text of this article. Literature Cited ALBERS, P. H. AND R. M. PROUTY 1987. Survival of spotted salamanders eggs in Vol. 5 p. 164 Asiatic Herpetological Research December 1993 temporary woodland ponds of coastal Maryland. Environmental Pollution 46(1):45-61. ARNTZEN, J. W. 1981. Ecological observation on Chioglossa lusitanica (Caucata, Salamandridae). Amphia Reptilia 3/4:187-203. BAS LOPEZ, S. 1984. Biogeographia de los amfibies y reptiles de Galicia, un ensago de sintesis. Amphibia Reptilia 5:289-310. BELL, G. A. C. 1977. The life of the smooth newt (Triturus vulgaris) after metamorphosis. Ecological Monographs 47(3):279-299. BELL, G. A. C. AND J. H. LAWTON. 1975. The ecology of the eggs and larvae of the smooth newt, Triturus vulgaris (Linn.). Journal of Animal Ecology 44(2):393-424. BENESKI, J. T., E. J. ZABISKO, AND J. H. LARSEN. 1986. Demography and migratory patterns of the eastern long-toed salamander, Ambystoma macrodactylum columbianum. Copeia 1986(2):398-408. ' BOZHANSKI, A. T. AND D. V. SEMONOV. 1982. [The data on the biology of the Caucasian salamander (Mertensiella caucasica)]. Zoologichesky Zhurnal 61(8): 1 1 18- 1 129. (In Russian). BORJA-SANCHIZ, F. DE AND M. MLYNARSKI. 1979. Pliocene salamandids (Amphibia, Caudata) from Poland. Acta Zoologica Cracov 24(4): 175-188. FERNANDEZ, P. J. AND J. P. COLLINS. 1988. Effect of environment and ontogeny on color pattern variation in Arizona tiger salamanders (Ambystoma tigrinum nebulosum Hallowell). Copeia 1988(4):928-938. FREYTAG, G. E. 1954. Der Kaukasus- Salamander, ein seltener Terrarien pflegling. Die Aquarien und Terrarien Jahrbuch:l 15-1 19. GOLUBEV, N. S. 1981. [The peculiarities of the tail autotomy and regeneration of the Caucasian salamander]. Voprosy gerpetologii. Nauka, Leningrad 5:40. (In Russian). GONTCHAROV, B. F., O. I. SHUBRAVY, I. A. SERBINOVA, AND V. K. UTESHEV. 1989. The USSR program for breeding amphibians, including rare and endangered species. International Zoo Yearbook 28:10-21. HUSTING, E. L. 1965. Survival and breeding structure in a population of Ambystoma maculatum. Copeia 1965(3):352-362. ISCHENKO, V. G. 1989. [The reproductive tactics and demography of amphibian population]. Itogi nauki i tekhniki. Zoologiya pozvonochnykh. V. 17. Problemy populyatsionnoy ekologii zemnovodnykh i presmykayushchikhsya. Moscow, VINITI:5-51. (In Russian). JOLY, J. 1966. Sur l'ethologie sexuelle de Salamandra salamandra (L.). Z. Tierpsychol. 23(l):8-27. BRODIE, E. D., R. A. NUSSBAUM, AND M. Dl GIOVANI. 1984. Antipredator adaptations of Asian salamanders (Salamandridae). Herpetologica 40(l):56-68. CYREN, O. 1911. Beitrage zur Kenntnis der Kaukasichen Feuer-salamanders, Salamandra caucasica (Waga), seiner Lebensweis und Fortpflanzung. Bericht der Senckenbergischen Naturforschended Gesellschaft, Frankfurt a/M 42:175-189. DAREVSKY, I. S. AND V. F. POLOZHIKHINA. 1966. [On the reproductive biology of the Caucasian salamander- Mertensiella caucasica (Waga)]. Zoologichesky Zhurnal 45(3):465- 466. (In Russian). EKVTIMISHVILI, Z. 1948. [Feeding of the Caucasian salamander]. Trudy Zoologitcheskogo Instituta AN GSSR 8:239-245. (In Georgian). KASTLE, W. 1986. Rival combaLs in Salamandra salamandra. Pp. 525-528 In Z. Rocek (ed.), Studies in Herpetology. Prague. KAUGHLEY, G. 1977. Analysis of vertebrate populations. A. Wiley- Interscience Publications., London, New York, Sydney, Toronto. 361 pp. KLEWEN, R. 1986. Population ecology of Salamandra salamandra terrestris in an isolated habitat. Pp. 395-398 In Z. Rocek (ed.), Studies in Herpetology. Prague. KOROLJOV, A. V. 1986. Some data on the larvae of Mertensiella caucasica. Pp. 281-283 In Z. Rocek (ed.), Studies in Herpetology. Prague. KUZMIN, S. L. 1992. Feeding ecology of the Caucasian Salamander (Mertensiella caucasica) with comments on life history. Asiatic December 1993 Asiatic Herpetological Research Vol. 5 p. 165 Herpetological Research 4:123-131. LANDE, R. AND G. F. BARROWCLOUGH. 1989. The effective numbeity of population and genetic variability and their application in the population monitoring. Pp. 117-157 In M. E. Sould (ed.) Viable populations for conservation. Mir, Moscow. (Russian translation). OZETI, N. 1967. The morphology of the salamander Mertensiella luschani (Steindachner) and the relationships of Mertensiella and Salamandra. Copeia 1967(2):287-298. PARASKIV, K. P. 1953. [Siberian salamander (Ranodon sibiricus)]. Izvestiya AN KazSSR, Biologicheskaya seriya 1:47-56. (In Russian). KNOBLAUCH, A. 1905. Der Kaukasische Feuersalamander, Salamandra caucasica (Waga). Berichte der Senckenbergischen Naturforschenden Gesellschaft in Frankfurt am Main 36:89-110. LLOYD, M. 1967. Mean crowding. Journal of Animal Ecology 36(1): 1-30. MERTENS, R. 1968. Bemerkungen zur "Normalentwicklung" des Kjaukasus- Salamanders. Salamandra 4:44^45. NESTEROV, P. V. 1911. Salamandra caucasica Waga. Izvestiya Kavkazskogo Muzuya 5:319- 327. (In Russian). NIKOLSKY, A. M. 1913. Herpetologia Caucasica. Kantselariya Namestnika E. I. V. na Kavkaze, Tiflis. 272 pp. (In Russian). OBST, F. J. AND J. ROTTER. 1962. Notizen zu Mertensiella caucasica (Waga) 1876. Aquarien und Terrarien Zeitschrift 15(2):50-52; (3):84-86. SEMLITCH, R. D. AND H. K. PEACHMANN. 1985. Diel pattern of migratory activity for several species of pond-breeding salamanders. Copeia 1985(1):86-91. TARKHNISH VILI, D. N. AND R. V. TARTARASHVILI. 1987. [On the development of spotted pigmentation in some Salamandridae]. Zoologichesky Zhurnal 66(9):1328-1338. (In Russian). TARTARASHVILI, R. V. AND M. L. BAKRADZE. 1989. [The new subspecies of the Caucasian Salamander]. Soobshcheniya AN GSSR 133(1):177-179. (In Russian). THORN, R. 1968. Les salamandres d'Europe, d'Asie et d'Afrique du Nord. Editions Paul Lechevarier, Paris. 376 pp. WOLTERSTORFF, W., L. A. LANTZ, AND W. HERRE. 1936. Ein Import des kaukasischen Salamanders (Mertenjsiella caucasica). Zoologischer Anzeiger 116(1/2):1-13. December l'W Asiatic Herpetological Research Vol. 3, pp. 166-169 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 die general content of the article in as few words as possible. A poor tide 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. 1993 by Asiatic Herpetological Research December 1993 Asiatic Herpetological Research Vol. 3. p. 167 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, ..." Vol. 3, p. 168 Asiatic Herpetological Research December 1993 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 ). Herpetologica 43(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. Phylogeneuc 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, Xinjiang Uygur Autonomous Region, China. December 1993 Asiatic Herpetological Research Vol. 3, p. 169 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 1 1 inch) white bond paper. If possible, include a computer diskette containing the manuscript. Macintosh diskettes with Mac write. 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. ISSN 1051-3825 CONTENTS TUNIYEV, BORIS S., AND SAHAT M. SHAMMAKOV. Coluber atayevi Sp. Nov. (Ophidia, Colubridae) from the Kopet-Dag Mountains of Turkmenistan 1 CHOU, WEN-HAO. On the Status of Rhacophorus prasinatus Mou, Risch, and Lue (Anura: Rhacophoridae) 11 AUFFENBERG, WALTER, AND HAFIZUR REHMAN. Studies on Pakistan Reptiles. Pt. 3. Calotes versicolor 14 ROCEK, ZBYNEK. Holocene anurans from Caucasus 31 Wei, Gang, Ning xu, dejun Li, Guanfu wu and xiquan Song. Karyotype, C- Band and Ag-Nors Study of Three Stink Frogs 45 GOLUBEV, M. L. The Variegated Toad Agama in Djungar Gate (Eastern Kazakstan) with Notes on Certain Systematic Problems of Phrynocephalus versicolor Str. (Reptilia: Agamidae) 51 Mezhzherin, Sergei and Michael L. Golubev. Allozyme Variation and Genetic Relationships within the Phrynocephalus guttatus Species Group (Sauria: Agamidae) in the Former USSR .' 59 WANG YUEZHAO AND HurZHAO WANG. Geographic Variation and Diversity in Three Species of Phrynocephalus in the Tengger Desert, Western China 65 TUNIYEV, BORIS. S. AND SVETLANA. YU. BEREGOVAYA. Sympatric Amphibians of the Yew-box Grove, Caucasian State Biosphere Reserve, Sochi, Russia 74 SMITH, BRIAN E. Notes on a Collection of Squamate Reptiles from Eastern Mindanao, Philippine Islands Part 1 : Lacertilia 85 SMITH, BRIAN E. Notes on a Collection of Squamate Reptiles from Eastern Mindanao, Philippine Islands Part 2: Serpentes 96 ZHONG, CHANGFU. First Records for Ophisaurus harti and Python molurus bivittatus from Jiangxi Province, China 103 MANILO, VALANTINA V. AND MICHAEL L. GOLUBEV. Karyotype Information on some Toad Agamas of the Phrynocephalus guttatus Species Group (Sauria, Agamidae) of the former USSR ' 105 MANILO, VALENTINA V. A Karyosystematic Study of the Plate Tailed Geckos of the Genus Teratoscincus (Sauria, Gekkonidae) 109 WANG, PEI-CHAO AND JlANG-HUA ZHANG. Resting Metabolic Rate in Three Age-groups of Alligator sinensis 112 ZHANG, YUN, YULIANG XlONG AND CASSIAN BON. Effects of Chinese Snake Venoms on Blood Coaguladon, Purified Coagulation Factors and Synthetic Chromogenic Substrates 117 SCHAMMAKOV, SAHAT, CHARI. ATAEV, AND ELDAR. A. RlJSTAMOV. Herpetogeographical Map of Turkmenistan 1 27 LIU, WAN-ZHAO, AND DA-TONG YANG. A Karyosystematic Study of the Genus Bombina from China (Amphibia: Discoglossidae) 137 SONG, JIANXING, YULIANG XlONG, WANYU WANG, AND XlAOCHUN PIT. A Study on the Purification and Pharmacological Properties of Two Neurotoxins from the Venom of the King Cobra (Ophiophagus hannah) 143 TARKHNISHVILI, DAVID N., AND IRINA. A. SERBINOVA. The Ecology of the Caucasian Salamander {Mertensiella caucasica Waga) in a Local Population 1 47 GUIDELINES FOR M ANUSCRIPT PREPARATION AND Si IBMISSION 1 66 Harvard MCZ Library 3 2044 066 300 419