HARVARD UNIVERSITY

Ernst Mayr Library

of the Museum of

Comparative Zoology

ASIATIC

HERPETOLOGICAL

RESEARCH

VOLUME 3

1990

Asiatic Herpetological Research

Editor

ERMI ZHAO

Chengdu Institute of Biology, Acadcmia Sinica, Chengdu, Sichuan, China

Associate Editors

KELLAR AUTUMN THEODORE j, PAPENFUSS

Museum of Vertebrate Zoology, University of Museum of Vertebrate Zoology, University of

California, Berkeley, California, USA California, Berkeley, California, USA

J. ROBERT MACEY

Museum of Vertebrate Zoology, University of

California, Berkeley, California, USA

Editorial Board

Kraig Adi.fr

Cornell University, Ithaca, New York, USA

Meiiii a Huang

Zhcjiang Medical University, Hangzhou, Zhejiang,

China

Natalia B. Ananjeva

Zoological Institute, Leningrad, USSR

Leo Borkin

Zoological Institute, Leningrad, USSR

Bihui Chen

Anhui Normal University, Wuhu, Anhui, China

Ylanchong Chen

Shanghai Institute of Biochemistry, Shanghai,

China

ILLYA DAREVSKY

Zoological Institute, Leningrad, USSR

INDRANEIL DAS

Madras Crocodile Bank, Vadanemmeli Perur, Madras,

India

CarlGans

University of Michigan, Ann Arbor, Michigan, USA

David M. Green

McGill University, Montreal, Quebec, Canada

Robert F. Inger

Field Museum, Chicago, Illinois, USA

KlJANGYANG LUE

National Taiwan Normal University, Taipei, Taiwan,

China

Hidetoshi Ota

Department of Biology, University of the Ryukyus,

Nishihara, Okinawa, Japan

Anming Tan

University of California, Berkeley, California, USA

William E. Duellman

University of Kansas, Lawrence, Kansas, USA

Hajime Fukada

Scnnyuji Sannaicho, Higashiyamaku, Kyoto, Japan

Datong Yang

Kunming Institute of Zoology, Kunming, Yunnan,

China

Asiatic Herpetological Research is published by the Asiatic Herpetological Research Society (AHRS) and the

Chinese Society for the Study of Amphibians and Reptiles (CSSAR) at the Museum of Vertebrate Zoology, University

of California. The editors encourage authors from all countries to submit articles concerning but not limited to Asian

herpctology.

Authors should consult Guidelines for Manuscript Preparation and Submission at the end of this issue.

All correspondence outside of the People's Republic of China and requests for subscription should be directed to AHR,

Museum of Vertebrate Zoology, University of California, Berkeley, California, USA 94720. All correspondence within

the People's Republic of China should be directed to Ermi Zhao, Editor, Chengdu Institute of Biology, P.O. Box 416,

Acadcmia Sinica, Chengdu, Sichuan Province, People's Republic of China.

Subscription and membership are $20 per year ($25 outside of the USA and China, $40 for libraries). Make

checks or money orders payable in US currency to AHRS. If you do not have access to US currency, please notify us,

and we will make other arrangements.

Asiatic Herpetological Research Volume 3 succeeds Chinese Herpetological Research Volume 2, which

was published at the Museum of Vertebrate Zoology, 1988-1989 as the journal for the Chinese Society for the Study of

Amphibians and Reptiles. Volume 2 succeeded Chinese Herpetological Research 1987, published for the

Chengdu Institute of Biology by the Chongqing Branch Scientific and Technological Literature Press, Chongqing,

Sichuan, China. Acta Herpetologica Sinica ceased publication February 1988, with Vol. 6, No. 2.

April 1990

Asiatic Herpetological Research

Vol. 3, pp. 1-36

Three Species in the Vipera kaznakowi Complex (Eurosiberian Group) in

the Caucasus: Their Present Distribution, Possible Genesis, and

Phylogeny*

NIKOLAI L. ORLOV1 AND BORIS S. TUNIYEV2

^Zoological Institute. USSR Academy of Sciences, Leningrad, USSR

2Causasian State Biosphere Reserve, Sochi, USSR

Abstract.-Thiee valid species: Vipera ka2nakowi Nikolsky, V. dinniki Nikolsky and V. darevskii

Vedmedeja, Orlov, and Tuniyev of the "Vipera kaznakowi " complex belonging to the Eurosiberian viper

group are recognized. The distributions of the closely related species, V. kaznakowi and V. dinniki are

defined. The habitat of V. kaznakowi is present along the Black Sea coast, from Khopa Village (Turkey)

in the south, to Maikop (USSR) in the north. Vipera kaznakowi is associated with montane areas from

sea level up to 1000 m. The range of V. dinniki covers the northern and southern slopes of the Great

(Main) Caucasus Ridge, ranging from the Fisht-Oshten Massive in the west to Shkhara Mountain in the

east. Vipera dinniki tends to be restricted to alpine and subalpine zones at elevations from 1 500 m to

3000 m. Vipera darevskii occurs in the southeastern part of the Dzhavachet Mountain Ridge, Armenia,

near the border with Georgia. The history of studies on the vipers of the "Vipera kaznakowi " complex is

summarized. The possible genesis of the present distributions of these vipers and their phylogenetic

relationships are discussed. Comparative morphological and ecological characters of the three species are

listed.

Key Words: Reptilia, Serpentes, Viperidae, Vipera, USSR, Caucasus, systematics, ecology.

Introduction

The study of the following viper species

of the Caucasus, Vipera kaznakowi

Nikolsky 1909, V. dinniki (Nikolsky

1913), and V. darevskii Vedmederja,

Orlov and Tuniyev 1986 and the history of

their distributions is of great interest for

understanding the phylogenetic links of the

Eurosiberian shield-headed vipers from the

Caucasus (subgenus Pelias, Merrem 1808)

and the formation of the present snake

fauna in the west Caucasus Isthmus.

Ever since Nikolsky (1909) described

the Caucasus Viper, Vipera kaznakowi, the

understanding of the taxonomic position of

this species has constantly varied. Also,

ideas regarding the habitats of those forms

referred to this species have varied (Orlov

and Tuniyev 1986). We recognize at least

three closely related species within the V.

f This publication combines material previously

published in Russian by Orlov and Tuniyev (1986)

and Vedmederja et al. (1986) with additional

information.

kaznakowi complex. All of them occur

primarily in the western part of the

Caucasus Isthmus. The eastern boundary

of the species range needs to be defined.

The Caucasus vipers (Fig. 1) have been

recorded from the following localities:

Western Dagestan in the vicinity of

Khasav-Yurt (Krasovsky 1929, 1932).

Along the slope of Makh-khokh

Mountain from the highlands of Ingushetia

(Chernov 1929).

On Legli Mountain in the Gukasyansky

region, Armenia (Darevsky 1956).

In the vicinity of Ushguli at the foothills

of Shkhara Mountain in Svanetia, Georgia

(Muskhelishvili 1959).

In Borzhomi Canyon, Georgia

(Bakradze 1969, 1975; Bannikov et al.

1977).

In Lagodekhi, Georgia (Zoological

Institute, the USSR Academy of Sciences,

Nos. 8389 and 13769).

1990 by Asiatic Herpetological Research

Vol. 3, p. 2

Asiatic Herpetological Research

April 1990

FIG. 1. The distributions of the vipers from the Vipera kaznakowi complex and the Vipera ursini

complex in the Caucasus. 1. V. kaznakowi, 2. V. dinniki, 3. V. darevskii, 4. V. ursini renardi

(northern population) and V. ursini eriwanensis (southern population). Type localities: A -V.

kaznakowi; B, B' -V. dinniki; C-V. darevskii. Localities: 1 -Makh-Khokh Mountain; 2 -the

settlement of Khasav-Yurt; 3 -Lagodekhi; 4 -Benis-Kheri Canyon; 5 -the settlement of Khopa, Turkey;

6 -Mikhailovsky Pass; 7 -the settlement of Ubinskaya; 8 -the town of Maikop; 9 -the mouth of the

Urushten River; 10 -the Fisht-Oshtenovsky Massive; 11 -Shkhara Mountain. Extreme points of the

distributions are marked.

All three species, Vipera kaznakowi, V.

dinniki and V. darevskii, differ in

morphology and ecology (Vedmederja

1984; Orlov and Tuniyev 1986).

History of the study of the Vipera

kaznakowi complex.

The taxonomic problems and

interrelationships of the viper forms of the

Eurosiberian group in the Caucasus has

caused contradictions and has confused

zoologists for nearly a hundred years. The

fragmented highland landscapes within the

overall range of the Caucasus vipers

isolates even neighboring populations.

Hence, there is an accumulation of unique

characteristics in populations. Very

complex situations emerge in sympatric

areas of closely related species of" this

complex. This may be connected with

natural hybridization. Occasionally the

identification of some individuals from

intergrading populations is difficult. This

particularly concerns Vipera dinniki in the

areas of its interactions with V. kaznakowi,

and additionally with the vipers of the V.

ursini complex (Bonaparte). The high

April 1990

Asiatic Herpetological Research

Vol. 3, p. 3

FIG. 2. Type localities of the vipers of the Vipera kaznakowi complex proposed over the history of the

study. A -V. kaznakowi Nikolsky, described in 1909, collected from Tsebelda, the vicinity of Sukhumi,

Abkhazia, Georgia. B, B' -V. Berus dinniki Nikolsky, described in 1913, from the upper reaches of the

Malaya Laba River, the Caucasus Reserve, Krasnodarsky Territory and Svanetia, Georgia (respectively).

C -V. tigrina Tzarewsky, described in 1916, taken from the Northern Caucasus. D-V. berus ornata

Basoglu, described in 1947 from the settlement of Khopa, Artviisky Vilayet, Turkey. E -V. darevskii

Vedmederja, Orlov, and Tuniyev, 1986, from Legli Mountain, the Mokrye Mountains, Gukasyansky

Region, Armenia. F-V. kaznakowi orientalis Vedmederja, 1984 from the eastern part of the Main

Caucasus Ridge.

degree of phenotypic polymorphism in

vipers from the V. kaznakowi complex

also creates additional problems in the

evaluation of their taxonomic position.

Due to the complexity in identification of

these vipers and the absence of definite

localities for specimens an analysis of old

literature gives only an approximate idea

regarding the relationships of the Vipera

kaznakowi complex.

Rossikow (1890) mentioned "multi-

colored" vipers in the Northern Caucasus

defined as Vipera berus. Boettger (1893)

regarded the dispersal of two closely related

viper species, V. berus and V. renardi

from the Caucasus as a phenomenon that

deserves attention. Nikolsky (1905)

considered V. berus to be associated with

Transcaucasia whereas V. renardi inhabits

steppe areas of the Precaucasus and

mountains of the Northern Caucasus.

Leister (1908a) in his sketch on the

geographic distribution of V. renardi and

V. berus within the Caucasus wrote that V.

Vol. 3, p. 4

Asiatic Herpetological Research

April 1990

renardi is associated with steppe areas

which separate the Caucasus from the

habitat of V. berus, and also the mountains

of the Northern Caucasus. He also

confirmed Boettger's and Nikolsky's

opinions that V. berus also appears in

Transcaucasia. In the northern

Transcaucasia there is an isolated

population. According to Boettger (1899),

the individuals of V. berus which are

preserved at the Caucasus Museum (now

the Museum of Georgia) were collected

from Suani, Tiflis, Avar, Kodzhori,

Khasav-Yurt, and Kazikoparan, whereas

those from the Senckenburg Museum

(Frankfurt-am-Main, West Germany) were

taken from Sukhumi, Georgia.

Leister (1908a,b,c) gives a list of Vipera

berus specimens in the collection of the

Zoological Museum of the Academy of

Sciences (now Zoological Institute of the

USSR Academy of Sciences) taken from

the Caucasus: 1) from Tiflis and

Lagodekhi regions, Georgia and 2) from

Yelenovka near the Gochka Lake (now the

Sevan Lake). Leister then states that he

found V. renardi on the bank of the

Gochka Lake at the same locality where

Kessler and Brandt collected V. berus (this

specimen, ZIN 5478, collected by Brandt

we determined to be V. ursini ). On the

basis of the findings of Kessler and Brandt

and his own findings, Leister comes to the

conclusion that both species, V. berus and

V. renardi, live sympatrically. Concerning

the specimens collected by Kessler and

Brandt, Nikolsky (1905) writes that their

rostral touches only one apical scale like in

V. renardi. In all other characters they look

more like V. berus. This led Leister

(1908a) to consider that a certain

intermediate species is possible — an

ancestral form of V. berus and V. renardi.

Nikolsky (1913) referring to the vipers

from Transcaucasia listed by Kessler,

Brandt and Leister nevertheless concluded

that they are V. renardi. Leister (1908a,b),

having analyzed the collections, concludes

that the territory of the Northern Caucasus

and Transcaucasia are inhabited by V.

berus and V. renardi. He thought that V.

berus and V. renardi live sympatrically in

the Northern Caucasus (for instance, the

vicinity of Grozny is a sympatric area).

Since Nikolsky (1909) described Vipera

kaznakowi, a number of synonyms have

been proposed for this species as new

forms have been described and new

combinations proposed. A number of

forms, primarily from the Northern

Caucasus and from the eastern range of V.

kaznakowi have confused taxonomists as

to their relationship to V. kaznakowi.

These forms were placed in different

combinations within V. berus and V.

ursini (Nikolsky 1913, 1916; Basoglu

1947; Knoepfler and Sochurek 1955;

Kramer 1961). Nikolsky (1913) assigned

the common Caucasus viper to a new

subspecies, V. berus dinniki. He defined

the viper's range as the northern and

southern slopes of the Caucasus Ridge

from the Malaya (Small) Laba River

headwaters up to Elbrus Mountain. Morits

(1916) also recorded V. berus in the

Northern Caucasus. Tsarevsky (1916)

described a new form V. tigrina that was

closely related to V. renardi and V.

kaznakowi. Unfortunately he cited the

species locality solely as "the Northern

Caucasus."

Nikolsky (1909, 1911, 1913, 1916), the

author of Vipera kaznakowi and V. berus

dinniki descriptions, studied the

relationships of the species V. kaznakowi,

V. berus, and V. renardi (= V. ursini

renardi ) and noted the difficulties in

identification of these forms. Nikolsky

(1913) described V. berus dinniki from a

single specimen collected by N. Y. Dinnik

at the upper reaches of the Malaya Laba

River, and from three specimens found by

Shelkovnikov in Svanetia, Georgia. In his

next monograph, Nikolsky (1916)

discussed both forms: Coluber berus

dinniki and Coluber kaznakowi. The

range of the former was defined as the

Caucasus Mountains on both sides of the

Great Caucasus Ridge. The upper reaches

of the Bolshaya (Big) Laba River was

designated as a sympatric zone for V. berus

dinniki and V. renardi. In his opinion, V.

renardi was closely associated with the

Caucasus Black Sea coast and the northern

slope of the Caucasus Ridge.

April 1990

Asiatic Herpetological Research

Vol. 3, p. 5

Dinnik (1926) noted that Viper a berus

and V. kaznakowi occurred in the Northern

Caucasus. Krasovsky (1929) indicated that

V. berus inhabited Khasav-Yurt and

Rutulsky Kanton, Dagestan. Chernov

(1929) recorded Coluber berus dinniki in

the highlands of Ingushetia on the southern

slope of Makh-khokh Mountain. The

specimens were collected by D. Krasovsky

in 1926 (Fig. 1). Krasovsky (1933)

included both V. kaznakowi and V. berus

in the fauna of the Caucasus State Reserve.

Bartenev and Reznikova (1935) concluded

that V. kaznakowi and V. berus dinniki

were distributed in the western Caucasus,

whereas in the alpine, V. ursini renardi

also occurred. Rostombekov (1939)

recorded V. kaznakowi as part of the fauna

of Abkhazia. A new form, V. berus

ornata, was described from northeast

Anatolia (Basoglu 1947). Later on Mertens

(1952a) synonymized it with V .

kaznakowi. Terentyev and Chernov

(1949) regarded V. tigrina Tzarevsky and

V. berus dinniki Nikolsky as junior

synonyms of V. kaznakowi Nikolsky

which is, in their opinion, related to V.

ursini renardi.

Darevsky (1956) presented a new

combination of names for the viper from

the Gukasyansky region, Armenia: Vipera

ursini renardi. Fyodorov (1956) recorded

V. berus in the forests of the premontane

area of the Stavropolsky Territory and V.

kaznakowi in the subalpine belt. In his

survey of the snake fauna of Abkhazia,

Georgia, Milyanovsky (1957) mentions V.

kaznakowi. Muskhelishvili (1959) records

V. kaznakowi on Mount Shkhara from the

vicinity of Ushguli, Svanetia, Georgia.

Kramer (1961) regards V. tigrina a junior

synonym of V. kaznakowi, whereas V.

berus dinniki should be synonymized with

V. ursini renardi. Bakradze (1969) found a

female V. kaznakowi in the town of Banis-

Khevi, near Borzhomi, Georgia. On the

basis of this finding, he suggested that the

species habitat covered the entire Adzharo-

Imeretinsky Ridge and some of the

Trialetsky Ridge.

In the field guide of Bannikov et al.

(1977) the names Vipera berus dinniki, V.

tigrina and V. berus ornata are mentioned

as junior synonyms of V. kaznakowi.

Vedmederja (1977) recorded V. kaznakowi

in Adgaria, Georgia. In his opinion

(Vedmederja 1984) V. kaznakowi is a

polytypic species comprising four

subspecific forms. Tertyshnikov (1977),

in defining ecological and zoogeographic

subdivisions of the herpetofauna of the

Northern Caucasus, noted that V.

kaznakowi tends to be restricted to western

and southwestern montane regions. The

most recent taxonomic works dealing with

these vipers state that V. berus occurs

neither in the Caucasus nor in the

Precaucasus. Hence, V. kaznakowi is the

sole valid name with regard to shield-

headed vipers from the western Caucasus

Isthmus and northeast Anatolia. The

taxonomic status of the eastern populations

from Dagestan and Armenia was not

discussed (Terentyev and Chernov 1949;

Mertens 1952a; Mertens and Wermuth

1960; Klemmer 1963; Bannikov et al.

1977; Baran 1977; Harding and Welch

1980).

The study of old collections and

literature shows that the majority of

researchers did differentiate the vipers from

the Vipera ursini complex and the V.

kaznakowi complex. The major

difficulties in defining the systematic

position and taxonomic status concerned

primarily the vipers from the eastern range

of the V. kaznakowi complex. References

that V. berus occurs in the Northern

Caucasus and Transcaucasia more often

than not concern snakes from the V.

kaznakowi complex rather than those from

the V. ursini complex.

An analysis of literature data and

morphometric characteristics of 141 viper

specimens allowed Vedmederja et al.

(1986) to designate three species within the

Vipera kaznakowi complex:

1. V. kaznakowi Nikolsky, 1909

2. V. dinniki Nikolsky, 1913

3. V. darevskii Vedmederja, Orlov and

Tuniyev, 1986

Vol. 3, p. 6

Asiatic Herpetological Research

April 1990

Fig. 2 represents type localities of the

forms proposed for the Vipera kaznakowi

complex over the history of its study.

Systematic Accounts

Vipera kaznakowi Nikolsky, 1909

(Fig. 3, 4, 14a, Plate 1)

Chronology of species description

Vipera renardi (Christoph) - Silantyev,

1903, 30:37 (part).

Vipera kaznakowi Nikolsky, 1909:174;

Nikolsky, 1910:81, table 1.

Vipera kaznakowi - Nikolsky,

1913:179-181; colored plate III (=Fig. 3,

this paper).

Coluber kaznakowi - Nikolsky,

1916:244-247.

Vipera berus ornata Basoglu, 1947:182-

190.

Vipera ursini kaznakowi - Knoepfler and

Sochurek, 1955:185-188.

Vipera kaznakowi - Terentyev and

Chernov, 1949:270-277 (Map 15);

Bannikov et al., 1977:323-324 (map 133;

colored plate 31, 4).

Vipera kaznakowi kaznakowi -

Vedmederja, 1984:8.

The English common name is the Caucasus

Viper, or Kaznakow's Viper.

Lectotype1: No. 4408, an adult female.

Collected by Y. V. Voronov from

Tsebelda, the vicinity of Sukhumi,

Abkhasia, the Caucasus. It is stored in the

Caucasian Museum (now the Museum of

Georgia, Tbilisi).

+ Kramer (1961) was mistaken regarding this

specimen as the holotype, because Nikolsky (1909)

had not singled out a holotype from the five

specimens from which he described the species.

Diagnosis: A large snake for the

Eurosiberian group. Total length reaches

650-700 mm. Dorsally the head is covered

with large scales. In size and shape they

differ considerably from body scales.

Nostril is cut through either in the middle or

slightly closer to the lower edge of the

nasal. Upper-lateral edge of snout is

pointed. Rostral normally reaches two

apical scales on upper snout. Upper edge

of the nasalorostral scale is slightly curved

at obtuse angles. Scales of anterior frontal

have weak longitudinal keels. Head is

broad and normally black dorsally. Head is

separated from body by a sharp nuchal

collar.

Description: The ratio of body length to

tail length is 5.7 to 6.4 in males and 7.5 to

10.9 in females. Unlike other species

assigned to the Vipera kaznakowi

complex, red and yellow colors prevail in

V. kaznakowi. Melanistic specimens are

common. However, unlike complete

melanistic individuals of V. dinniki, those

of V. kaznakowi preserve yellow or red

color on either upper or lower labials.

Vipera kaznakowi is either black or dark

brown striped dorsally and laterally. Often

stripes merge so that only red or yellow

spots remain between them. Ventrum is

black. Head is very broad, impressed

dorsally. This fact is emphasized by a

slightly pointed upper edge of snout.

Cheeks are greatly swollen. Head is well

separated from body by a thin nuchal

collar. Comparative data on scalation and

size characters are listed in Tables 1-4.

Variability and comparative remarks

V. kaznakowi (Fig. 4, Plate 1) is

undoubtedly more closely related to V.

berus and V. ursini. Supposedly, when

interacting with V. ursini renardi and V.

ursini eriwanensis it forms two species of

hybrid genesis: V. dinniki and V.

darevskii. From all listed forms it differs

by: 1) the extraordinarily great width of the

triangular head, 2) the head width of

specimens of this species equals the

distance between tip of snout and posterior

angle of mouth slot, 3) the "cheeks" are

greatly swollen, hence a broad furrow

April 1990

Asiatic Herpetological Research

Vol. 3, p. 7

Nikolski. Nova species viperae.

H3B. K»bk. My3. T. V.

A. 3. E. Tom J. I.

II KiiiniiKiir C, TI- E.

Vipera kaznakovi sp n

FIG. 3. Type specimen of Vipera kaznakowi Nikolsky, 1913 (reproduced from Plate III, originally

printed in color, Nikolsky 1913).

Vol. 3, p. 8

Asiatic Herpetological Research

April 1990

TABLE 1 . Comparative scale characters of the three viper species of the "Vipera kaznakowi " complex.

TABLE 2. Sexual dimorphism of body length and number of subcaudals in the three viper species of the

"Vipera kaznakowi " complex.

to* 9

4.5

5.6

2.82

< 0.001

< 0.001

<0.05

TABLE 3. A comparison of the number of vertebrae carrying ribs in Shield Headed .Vipers of the USSR.

X-ray analysis of the vipers indicates that the number of rib-carrying vertebrae is stable, except for Vipera

dinniki, (indicated by **) in which the number varied from 128 to 140.

Specific and subspecific

viper forms

Vipera berus

Vipera kaznakowi

Vipera darevskii

Vipera dinniki

Vipera ursini renardi

Vipera u. eriwanensis

Vipera dinniki

appears between eye and temple. The

furrow goes up to upper head, parallel to

upper edge of parietals. Dorsally the head

is either impressed or flat. The viper also

differs from closely related species in body

proportions: 1) it is relatively thicker, more

massive; 2) the head is conspicuously

bigger. These habitus characters

differentiate all age groups of V. kaznakowi

ranging from juveniles to mature

specimens. Comparative data on

pholidosis, size, and number of vertebrae

April 1990

Asiatic Herpetological Research

Vol. 3, p. 9

TABLE 4. Comparative morphology of vipers in the "Vipera kaznakowi " complex.

Vipera kaznakowi, 64 specimens

• SVL 8 8 ^475 mm

SVL V 9 y< 600 mm

• SVL 86/ CI. x 5.6-6.4

SVL 9 9/ CL as 7.5-10.9

• Head is either impressed or flat

dorsally

• Edge of snout is slightly rounded

V. dinniki, 68 specimens

• SVL 6 6 ^ 412 mm

SVL 9 9 ^486 mm

• SVL 8 8/ CL a 5.9-7.4

SVL 9 9/ CL as 7.9-13 5

• Head is either flat or slightly

protuberant dorsally

• Edge of snout is rounded

• Rostral is broad. Usually touches <

two apical scales (in 91% specimens)

rarely one apical (in 9%)

• The width of the frontal is equal to i

1.21-1.72 of its length

• The distance between anterior edge i

of the frontal and the rostral is equal

to 0.75-1.05 of the frontal edge

• Frontal is either smaller than (

parietals or equals them

• Large lower oculars are separated i

from frontal either by one row (in

87.0%) or rarely by two rows of

small scales (in 12.4%)

• The upper pre -ocular does not touch (

the nasal in 96.1%. In 3.9% it does

• Nostril is either cut through in the i

middle of nasal or is slightly shifted

downwards

• Nasal does not touch rostral <

• Body scales with expressed keels; <

some scale rows that reach ventrals

have no keels

Rostral is narrow reaching either one

(in 48.6%) or two (in 51.4%) apicals

The width of the frontal is equal to

1.13-1.83 of its length

The distance between anterior edge of

the frontal and the rostral is equal to

0.77-1.35 of the frontal edge

Frontal is either smaller than parietals

or equals them

Large lower oculars are separated from

frontal either by one row (in 58.5%) or

two rows (in 41.5%)

The upper pre-ocular does not touch

the nasal in 86.6%. In other cases it

reaches the nasal

Nostril is cut through in the center or

nasal. It may be shifted downwards

exceedingly rarely

Nasal does not touch rostral

Body scales have expressed keels. 76%

of scales that reach ventrals have no

keels, 18.4% have slightly expressed

keels, 5.6% have well-expressed keels

V. darevskii, 9 specimens

• SVL 8 8 ? 258 mm

SVL 9 9?* 421 mm

SVL 88 / CL a 6.0-6.5

SVL 9 9 / CL a8.4-9.3

• Head is either flat or slightly

protuberant dorsally

• Edges of snout are slightly pointed

laterally. Anterior edge is a little

rounded

• Rostral is more narrow and touches

one (in three specimens) or two (in six

specimens) apical scales

• The width of the frontal is equal to

1.48-1.71 of its length

• The distance between anterior edge of

the frontal and the rostral is equal to

0.67-1.07 of the frontal edge

• Frontal is larger than parietals

• Large lower oculars are separated from

frontal only by a single row of small

scales (in all nine specimens)

• The upper pre-ocular does not touch

the nasal in two specimens. In seven

specimens it does

• Nostril is cut through in the lower part

of the nasal

• Nasal does not touch rostral

• Body scales have expressed keels.

Some scale rows that reach ventrals

have no keels

supporting ribs are given in Tables 1-4.

Coloration. Vipera kaznakowi are

incredibly diverse in color and

extraordinarily bright among shield-headed

vipers. Red hues prevail in the color

pattern. Newborns are similarly bright in

color, primarily red brown, unlike the gray

young of V. dinniki. The typical intensive

red color appears in V. kaznakowi after

they have shed twice. Complete melanists

are frequent in populations. Often the black

dorsal stripe merges with lateral stripes so

that either red or yellow spots arranged in

two rows along the dorsum remain. The

dorsal stripe can be either zig-zag shaped or

in the shape of an even broad line.

Ventrum is black. Head pattern of adult

specimens normally blends with the dorsal

stripe. In immature specimens, the head

pattern may be separated from the dorsal

stripe by a light interval that disappears

after maturation.

Vol. 3, p. 10

Asiatic Herpetological Research

April 1990

FIG. 4. Vipera kaznakowi (ZIN 1 1529) from the isolated forest of Babuk-Aul.

PLATE 1 . Vipera kaznakowi (ZIN 1 1529) from the isolated forest of Babuk-Aul.

PLATE 2. One pattern variation of Vipera dinniki from Krasnaya Polyana (ZIN 12153), a male.

PLATE 3. Vipera ursini eriwanensis from the valley of Kassach, the foothills of the Ara-Fler Mountains.

Orlov and Tuniyev

Asiatic Herpetological Research

Plate 1

%

Muy^ \JU

Vipera kaznakowi

Orlov and Tuniyev

Asiatic Herpetological Research

Plate 2

Vipera dinniki

\

au^ if\\

Orlov and Tuniyev

Asiatic Herpetological Research

Plate 3

Vipera ursini eriwanensis

1

2Uu# W

April 1990

Asiatic Herpetological Research

Vol. 3, p. 11

Sexual dimorphism. Maximum body

length is longer in females (up to 600 mm)

than in males (does not exceed 457 mm).

Males have longer tails (Table 2).

Accordingly, the number of ventrals is

greater in females, whereas the number of

subcaudals is greater in males. Males are

more slender. Sexual dimorphism in

coloration is feebly marked. Melanistic

individuals are more frequent among

females.

Age variability. Vipera kaznakowi are

born with the typical adult color and

pattern. However, in newborns these

characters are less pronounced than in

adults. Newborn snakes may be either

pinkish or reddish. According to our

observations on birth and development of

snakes from the Sochi-Khosta populations,

the coloration becomes stronger with each

subsequent shedding. Maximal color

intensity is achieved by the season after the

first hibernation.

Melanistic specimens are born with the

typical species pattern, but their coloration

is darker. The coloration becomes darker

during subsequent sheddings and elements

of the pattern merge.

New born litters are homogeneously

colored. On maturing, the coloration

becomes diverse. Phenotypic

polymorphism in mature vipers of a litter is

great. Either partial or nearly complete

melanism is observed in all vipers from this

population. This is typical for a number of

other animals from humid subtropic areas

adjacent to the Black Sea coast (Bartenev

and Reznikova 1935). Specimens showing

maximal melanism always preserve

elements of orange and red on the throat

scales and chin shields, the rostral, upper

labials and subcaudals. Melanistic

specimens of V. dinniki can have

completely black coloration by maturity.

Geographic range and ecology. The

species ranges along the Black Sea coast

from the town of Khopa, Turkey and

Suramsky Pass in the east, then throughout

Kolchida (Colchis) up to Mikhailovsky

Pass in the west. It is then found up to the

northern slope of the Main Caucasus

Ridge. Here Vipera kaznakowi occurs

along the foothills from the settlement of

Ubinskaya in the west to the town of

Maikop, USSR in the north and the mouth

of the Urushten River in the east (Fig. 1

from Vedmederja et. al. 1986). The

species generally occurs up to an elevation

of 800 m. Along the river valleys of the

Black Sea coast, it may occur up to an

elevation of 1000 m or even higher.

Vipera kaznakowi is a forest dwelling

species. It occurs on montane wooded

slopes, in the bottoms of humid canyons,

and in meadows adjacent to forests. It is

recorded in Quercetum azaleorum and

Quercetum coggygriosecornosum oak

groves, in mixed subtropic forests with

evergreen subforests of Quercus

hartwissiana, Quercus iberica, Alnus

barbata, Fagus orientalis, Taxus baccata,

Laurocerasus officinalis, Buxux colchicus,

Castanetum colchicum, Fagetum nudum,

Salictum fontenale, and Alnetum

strut hiopteridor sum. In addition, this

species is found in polydominant forests in

river terraces and large overgrown

outcroppings. At the upper limits of its

elevational range, the species reaches

coniferous forests. It is recorded within the

ecotone of Fageto-Abieta athyridosa-

maxtoherbosa, but the viper never

penetrates deep into coniferous forests.

Vipera kaznakowi is also present within

transformed areas such as meadows formed

after forests are cut, fruit orchards, kitchen

gardens, vineyards, and dilapidated parks

(Red Data Book of the USSR 1978, 1984).

As a rule, vipers occur within sites where

the density of lizards is high.

Typical biotypes are as follows: small

meadows and other illuminated spots in

forests with an exposure providing high

solar radiation in conditions of humid

subtropical climate of the Black Sea coast

of the Caucasus. The sites are located in

the vicinity of standing water, where rocks

exit that are suitable for hibernation. The

climate of the Caucasus coast is greatly

dependent on topographic conditions of the

area that forms a narrow line between the

Caucasus Ridge and the Black Sea. It is

Vol. 3, p. 12

Asiatic Herpetological Research

April 1990

situated from northwest to southeast along

the coast. The wall of the Ridge starts from

Anapa. Near Novorossiisk it reaches an

elevation of up to 600 m. Near Tuapse, the

wall exceeds 1000 m above sea level. At

the latitude of Sochi, it is 3000 m high.

The Great Caucasus Ridge presents a

barrier for cold northeastern winds. This

barrier separates the warm humid coast

from the comparatively cold continental

Prekuban area (Korostelyov 1933). In the

east the Adzharo-Imeretinsky Ridge

separates the humid subtropical Kolchida

(Colchis) from the arid regions of the

Eastern Precaucasia. The climatic zone of

the Black Sea coast of the Caucasus

appears to be extremely favorable for

Vipera kaznakowi. The northwestern most

localities for this species (Mikhailovsky

Pass and Stanitsa Ubinskaya) coincide with

Pontic and northwestern floristic regions.

In relief, climate, and vegetation the

northwestern region is a continuation of the

southern coast of the Crimea.

The Pontic region is marked by 1)

subtropical vegetation in the foothills, 2)

great temperature stability, and 3) high

humidity (Korostelyov 1933). The limited

dispersal of the species along the northern

slope of the Big Caucasus is in the region

of the "Kolchidian (Colchis) Gates." Due

to the lowering of the western portion of

the Main Caucasus Ridge, interchange of

flora and fauna of Kolchida (Colchis) and

the Prekuban area occurred in the past and

presently occurs. Warm humid air from the

Black Sea breaks through in this part of the

Ridge. On the northern slope of the Big

Caucasus, it creates a small refugium on a

Kolchidian (Colchis) type associated with

Castanea satyra, Buxux colchicus, Ostrya

carpinifolia, Corylus colurna, and others

(Galushko 1974; Kharadze 1974;

Kholyavko et al. 1978). Of the

herpetofauna of Kolkhida B ufo

verucossimus, Triturus vittatus ophryticus,

Pelodytes caucasicus, and Lacerta derjugini

occur along with V. kaznakowi. Along the

entire area of of the Black Sea coast of the

Caucasus, the viper is rare. In a number of

spots it has disappeared. In some

situations, only single specimens of V.

kaznakowi might be encountered.

The highest density of Vipera kaznakowi

that have been observed are in rocky

outcroppings within the forest belt in the

mountains of the Caucasus Reserve. In

beech forests along a road in the valley of

the Achipse River, we have recorded three

specimens per kilometer of road. Single

specimens have been recorded in the

vicinity of Khosta, Babuk-Aul, Guzeripl,

and Kisha Kordons. The density of the

viper is also low in the southern habitats in

Adzharia and Lazistan, Georgia (Basoglu

1947; Vedmederja 1977; Basoglu and

Baran 1977).

Anthropogenic factors are responsible

for declining numbers and populations of

Vipera kaznakowi. There are pressures due

to recreational use of health resorts along

the Black Sea coast, ploughing of sites

near the foothills, and to a smaller extent,

hay-mowing (Red Data Book of the USSR

1978, 1984). Within some resort areas V.

kaznakowi is completely extinct.

Diet. Vipera kaznakowi feeds on various

animals. Different populations show

specific feeding patterns with regard to prey

available. According to data recorded in

captivity, individual preference is observed.

When stimulated to regurgitate, the

following food species were recorded in the

field: Apodemus sylvaticus, a forest

mouse, A. agrarius, a field mouse,

Microtus majori, a juvenile specimen, M.

gud, Sorex raddei, Lacerta saxicola, L.

derjugini, L. praticola and L. agilis. In the

collection of the Zoological Museum of the

Moscow State University a viper specimen

from Bebysyry Lake is preserved. In its

stomach a juvenile grass snake (Natrix

natrix ) was found. Immature individuals

feed on juvenile lizards of the above

mentioned species, and to a lesser extent,

on Orthoptera.

In captivity adult vipers readily take any

small rodents, fledgling sparrows, lizards,

and pieces of chicken. Young vipers

usually start on small lizards and crickets

(Grillus bimaculatus and Achaeta

domestica ). After some months they

usually take newborn mice. After the

viper's bite a prey normally dies in 5-7

April 1990

Asiatic Herpetological Research

Vol. 3, p. 13

minutes. The snake never persecutes its

prey. It finds the prey some time later

using its olfactory organs. Swallowing

ranges from one minute to 3.5 hours

depending on prey size and the state of a

snake. Complete digestion in the wild

takes up to five days. In captivity digestion

may take 30 to 40 hours at stable

temperature. Optimal day temperature is

26-30°C, and night temperature 18°C

during activity period.

Shedding. Mature vipers normally shed 2

to 3 times during the activity period.

General shedding is observed in June.

New born vipers shed in the first hours

after birth. Before entering hibernation,

they shed again.

Reproduction. Mating is recorded from late

March to April. Birth occurs in late

August. Females give birth to 3 to 5

young. The observed time of birth lasts for

about two hours with intervals of 20 to 40

minutes (Zinyakova and Trofimov 1977).

In captivity, the majority of females give

birth at night, between 2400 and 0600 hrs.

Some individuals are born in transparent

capsules, which the neonates leave in the

first minutes after birth. Females reproduce

annually. Gravid females continue to take

food right up to birth.

Development of young. Neonate Vipera

kaznakowi have a mean body length of

144.75 mm, tail length of 14.0 mm. Mean

body weight is 4.1 g (n=8). After first

shedding on the second day after birth,

newborns start actively feeding on insects

or small lizards. After birth the activity

period is 1.5 to 2.5 months, whereas

newborns of V. dinniki never take food

and almost immediately enter hibernation,

during which snakes increase in length

from 10 to 20 mm. Vipera kaznakowi lose

0.3 g of initial weight during the first

month after birth. In the second month

they restore their initial weight and then

increase it approximately 1 g before

entering hibernation. One year old

specimens have a body length of 200 mm

and a tail length of 24 mm. Vipers reach

sexual maturity by the third year at a SVL

of 350-400 mm.

Territoriality. Like other vipers, Vipera

kaznakowi is conservative in territory use.

The same individuals can be encountered in

the same places during different seasons.

Vipera kaznakowi utilizes considerably

larger individual ranges than V. dinniki.

Seasonal and daily activity. Along the

Black Sea coast of the Caucausus, Vipera

kaznakowi emerge after hibernation in

March at an altitude of 600 to 800 m. On

the northern slope of the Big Caucasus the

vipers appear in the second half of April to

early May when the mean day temperature

is 13°to 16°C. In the foothills the vipers

enter hibernation at the beginning of

November at an altitude of up to 600 m. In

the upper elevational limits of its range V.

kaznakowi hibernate in late September to

early October. New born vipers are more

active than those of other age groups.

Two sharply marked peaks of daily

activity can be observed in Vipera

kaznakowi. In the morning the period of

daily activity ranges from 0730 to 1130

hrs, and in the evening from 1630 to 1830

hrs. At those times the soil temperature

does not exceed 30-32°C in the sites

inhabited by V. kaznakowi.

Sympatric species. Along with Vipera

kaznakowi the following species of reptiles

occur: Lacerta derjugini, L. saxicola, L.

praticola, L. agilis, Anguis fragilis,

Pseudopus apodus, Coluber najadum, C.

jugularis, Elaphe longissima, Natrix natrix,

N. tessellata, and Coronella austriaca.

Narrow sympatric areas appear with

Testudo graeca, Lacerta trilineata, L.

caucasica, L. rudis, L. parvula, L. mixta,

Elaphe hohenackeri, Vipera ursini, and V.

dinniki.

Vipera dinniki Nikolsky, 1913

(Fig. 5, 6, 7, 12, 14ab,16, Plate 2)

Species description in chronology

Vipera berus - Boettger in Radde,

1899:286; Nikolsky, 1905:304 (ad

Caucasus).

Pelias chersea - Menetries, 1 832:73 (part).

Vol. 3, p. 14

Asiatic Herpetological Research

April 1990

Vipera xanthina -Dinnik, 1902:34.

Vipera renardi - Silantyer, 1903, 30:37

(part).

Vipera berus dinniki Nikolsky, 1913:176-

179.

Coluber berus dinniki - Nikolsky,

1916:240-244.

Vipera tigrina Tzarevsky, 1916:32-37.

Vipera ursini renardi - Kramer, 1961:715.

Vipera ursini kaznakowi - Knoepfler and

Sochurek, 1955:185-188.

Vipera kaznakowi - Terentyev and

Chernov, 1949:270-271 (map 5);

Bannikov et al., 1977:323-324 (map 133,

colored plate 31,4).

Vipera kaznakowi dinniki - Vedmederja,

1984:8.

Vipera kaznakowi orientalis - Vedmederja,

1984:9, nomen nudum.

The English common name is Dinnik's

Viper or the Caucusus Subalpine Viper.

Lectotype: No. 26044, an adult female

collected by N. Y. Dinnik (Fig. 5) from

the upper reaches of the Malaya (Small)

Laba River, Northern Caucasus and

Svanetia, Georgia (Fig. 2: B, B'). The

specimen is preserved at the Museum of

Natural History, Kharkov State University,

Ukraine.

Diagnosis: Total length reaches 500 to

550 mm. Dorsally the head is covered with

large scales. Nostril is cut through in the

center of the nasal. Upperlateral snout edge

is rounded and slightly pointed. Rostral

touches either one or two apical scales on

upper head. Three to four scale rows with

no keels are between the rostral and frontal.

Head is not broad. Nuchal collar is not

expressed.

Description: In males body length is not

greater than 412 mm; in females it does not

FIG. 5. Lectotype of Vipera dinniki (The

Museum of Natural History of Kharkov State

University 26044).

exceed 486 mm. The ratio of body length

to tail length is 5.9 to 7.4 in males; 7.8 to

13.5 in females. General coloration is not

as bright as in Vipera kaznakowi.

However, specimens with bright yellow

and orange elements can be observed.

Normally V. dinniki (Fig. 6,.Plate 2)

have light brown, grey, silver greyish or

green greyish color which never occurs in

V. kaznakowi. Some specimens have a

dark even dorsal stripe along the center of

the body. The latter substitutes for the zig-

zag shaped stripe typical for the majority of

plate-headed vipers. Ventrum is either dark

and light spotted or light grey and dark

speckled. Neonates of V. dinniki

occasionally don't have the red color of the

body typical for V. kaznakowi. They can

be bom grey brown. The head is relatively

narrower than that of V. kaznakowi. The

nuchal collar is not expressed, unlike in V.

kaznakowi. Upper edge of snout is either

rounded or slightly pointed. Head can be

either protuberant or flat dorsally, but never

impressed like that of V. kaznakowi. Body

is thinner and more delicate. Comparative

data on pholidosis and size characters are

listed in Tables 1-4.

Remarks and variability.

Vipera dinniki greatly resembles

morphologically V. kaznakowi, V. ursini

renardi and V. berus. In size V. dinniki is

smaller than V. kaznakowi and larger than

V. ursini renardi (Table 2). The head

normally is slightly protuberant, seldom flat

and never as broad as that of V .

April 1990

Asiatic Herpetological Research

Vol. 3, p. 15

FIG. 6. A Female Vipera dinniki from the upper part of the Laba River (ZDN 17281).

kaznakowi. Hence, the nuchal collar is

hardly marked. In body proportions it

primarily resembles V. berus. Normally

the parietals are shorter than the frontal.

Three to four lower labials reach the lower

jaw scale.

Color. Normally Vipera dinniki is not as

brightly colored as V. kaznakowi.

However specimens with either bright

yellow or orange elements occur. Often V.

dinniki have greyish, silver greyish or

green greyish color. This is never recorded

in V. kaznakowi. For all shield-headed

vipers a zig-zag shaped dorsal stripe is a

common pattern element. In V. dinniki the

stripe is often an even broad dark line

which runs along the dorsum. This is

occasionally seen in V. kaznakowi, where

it is usually represented by a row of oblique

Vol. 3, p. 16

Asiatic Herpetological Research

April 1990

FIG. 7. A male Vipera dinniki from the vicinity of Krasnaya Polyana (Red Meadow), [ZIN 12153].

diametrical spots. The dorsal stripe is

separated from the dark sides by lighter

lateral stripes. The ventrum is either dark

with light spots or light grey. The number

of melanistic specimens in populations

ranges from 20 to 25%. Complete

melanistic individuals of V. dinniki do not

have a single light spot in color, unlike

melanists of V. kaznakowi.

Sexual dimorphism. Maximum body

length is greater in females (up to 486 mm),

and smaller in males (up to 412 mm).

Males have longer tails, characteristically

April 1990

Asiatic Herpetological Research

Vol. 3, p. 17

FIG. 8. A Vipera dinniki from Lagodekhi, an eastern population (ZIN 13769).

thicker at the tail base (Table 2). The

number of ventral scales is greater in

females. The number of subcaudals is

greater in males. Males are more delicately

built than females. Sexual dimorphism in

color is hardly expressed. Coloration in

males is generally brighter and more

contrasting.

Age variability. Neonate vipers are

patterned like adult individuals. However,

the general color is normally grey, unlike

bright red neonates of Vipera kaznakowi.

Only after the third shedding, faint

Vol. 3, p. 18

Asiatic Herpetological Research

April 1990

coloration typical for the species (yellow,

reddish, greenish) emerges in V. dinniki.

Color becomes stronger with each

subsequent shedding. Maximum color

strength is reached by maturity. Melanistic

specimens are born with a specific color.

They gradually darken with each shedding.

By the third year, they acquire a black

velvet color.

Geographical range and ecology. The

species ranges from the Fisht-Oshtenovsky

Mountain Range in the west up to Mount

Shkhara in the east. The eastern limits of

this species distribution has not been

worked out and additional eastern localities

are probable. The southern distributional

limit runs along the dark coniferous

highland ecological belt on the southern

slope of the Main Caucasus and South

Frontal Ridges. The northern limit goes

from Mount Shkhara to the west along the

crest of the Main Caucasus Ridge up to the

Bolshaya (Big) Laba River head where it

passes on to a northern macroslope. In the

north it occurs on the Peredovoy (Frontal)

Ridge and reaches the Fisht-Oshtenovsky

Mountain Massive (Orlov and Tuniyev

1986). Research during 1987-1989

showed that Vipera dinniki occurs further

to the east than previously known (Fig. 9).

The problem of subspecific status of the

eastern populations and their interaction

with the vipers of the V. ursini complex

will be addressed in a future paper.

The elevational distribution is generally

between 1500 and 3000 m. Occasionally

the viper may descend slightly lower.

Vipera dinniki is a subalpine montane-

meadow species. It tends to be restricted to

the upper forest belt, subalpine and alpine

meadows, rocky outcroppings and montane

moraines.

Vipera dinniki can be observed in

vegetation associations of Betuletum

calamagrostidosum, Pinetum mystillosum

subalpinum, Fageto Betuleto-Sorbetum

altherbosa-subalpinum, also in Aceretum

trantaltherbosum subalpinum. It is

associated with rock outcrops interspaced

with shrubs of Rhodoretum caucasicus

subalpinus, subalpine highland herbs and

rock debris. It occurs widely in moraines

overgrown with moss, lichens and

Thymus. The biotype of V. Kaznakowi is

always located in the vicinity of water.

Hibernation sites are also located in the

immediate vicinity of summer biotypes. All

types of rock outcrops are inhabited.

Vipera dinniki can be found in limestone,

slate, and crystalline outcrops.

Climatic conditions within the range of

Vipera dinniki are much more severe than

those of V. kaznakowi. However, the

viper's distribution in severe highlands

coincides with the mildest climatic places in

this severe area. Vipera dinniki commonly

occurs on slopes with a southern or

southeastern exposure. For instance, on

Mount Aibga, 23 km away from the sea, at

an altitude in excess of 1800 m, the mean

temperature in January is -0.1°C

(Korostelyov 1933; Shkadova 1979).

Mean temperature in July is only 13.7°C on

Mount Achisko at an altitude of 1750 m (37

km away from the sea). However, great

temperature drops never occur in this area,

even in winter. As regards to soil freezing,

minimum temperature in January is -7°to -

8°C (Selyaninov 1933). Due to solar

radiation on slopes with southern and

southeastern exposures, the vipers manage

to maintain high body temperatures during

activity periods.

Vipera kaznakowi reaches the edge, but

does not penetrate deep into the dark

coniferous belt. Vipera dinniki, at its lower

limits, reaches the edge but never penetrates

the dark coniferous belt. Thus, the dark

coniferous belt is a barrier between them.

Occasionally, in spots where this belt is

either absent or is fragmented, for instance

along river valleys, both species occur

sympatrically, forming a narrow line with

intergrading characters. The valley of the

Mzymta River may serve as an example of

a site where the two species come in

contact. The limited distribution of the

species on the northern slope of the Main

Caucasus Ridge might be connected with

the increase of the mountain's aridity and

severe climate towards the east.

Xerophytization is observed from west to

east in the subsequent change of beech

April 1990

Asiatic Herpetological Research

Vol. 3, p. 19

|*| V 'ipera darevskii

J Vipera dinnik

1 Vipera kaznakowi

| Vipera ursini eriwanensis

J Vipera ursini renardi

FIG. 9. Distribution map of shield-headed vipers in the Caucasus.

forests and Abies forests to pine forests,

and further east, to steppe areas

(Adamyants 1971, Kharadze 1974,

Lavrenko 1980; Agakhanyants 1981). This

change is noted in amphibians and reptiles.

Along with V. dinniki, Lacerta derjugini

and Pelodytes caucasicus fall out and

more xeric species such as Bufo viridis,

Lacerta agilis, and Vipera ursini renardi

dominate.

Population density is different in various

parts of the habitat. It is maximal on rock

outcrops and moraines in the subalpine belt

of the Caucasus Reserve. In July, in

subalpine meadows of the Gertsen Ridge

and along the valley of the Bezymyanka

(Without name) River, 1700 to 1900 m, we

have recorded as many as 5 to 7 vipers per

1 km of a route. At the same time, on an

area of 500 m2, we have recorded up to 6

specimens. This is at an elevation of 1800

m on the Aishkha Ridge in a subalpine

meadow of Aceretum trantaltherbosum

subalpinum. In late June to early July

along the Moloshnaya (Milk) and

Sumasshedshya (Crazy) rivers of the

Mzymta River basin, up to 4 specimens per

100 to 500 m of a route were recorded. In

August, on a bank of the high altitude

Kardavych Lake, we recorded up to 8

specimens per 300 m of a route. On the

Aspidny Ridge, 2000 m above sea level,

we recorded 5 to 6 specimens per 300 to

400 m of a route.

According to Bozhansky (1979), the

density in the subalpine belt is 2 to 6

specimens per hectare. Occasionally,

seasonal accumulations of up to 30-40

Vol. 3, p. 20

Asiatic Herpetological Research

April 1990

individuals per hectare are possible. In

other regions the population densities are

much lower. Degradation of localities is

due to intensive cattle grazing in subalpine

meadows. Overall numbers of both V.

kaznakowi and V. dinniki are estimated at

some dozens of thousands (Bannikov and

Makeyev 1976). Apparently the figure may

primarily concern V. dinniki as the number

of V. kaznakowi is rather small.

Diet. Adult Vipera dinniki eat fewer types

of food items than V. kaznakowi. For

instance, in the highlands there are viper

populations that prey either on lizards or on

small mammals as no other food items are

available. When field collected vipers were

stimulated to regurgitate, the following prey

were observed: Apodemus sylvaticus,

Microtus majori, Sicista caucasica,

fledglings of ground nesting birds such as

Anthus spinoletta, and Lacerta caucasica.

Immature specimens feed on Orthoptera

and small lizards. In captivity V. dinniki

takes similar food items as V. kaznakowi.

It is often observed that a bitten mouse or a

field-vole makes a few desparate leaps,

falls into a deep crack between rocks and

dies there. The viper skillfully catches its

prey between its teeth, drags it out onto a

level spot, then drops it, examines it from

all sides and having found the prey's head,

swallows it. This behavior is typical for

other rock dwelling vipers from the

Caucasus, such as V. raddei, and V.

ammodytes.

Shedding. Overall, shedding is observed

in June and in late August to early

September. Neonate vipers shed during the

first hours after birth. Some days later they

enter hibernation.

Reproduction. Copulation occurs in late

April to May (Bozhansky 1984). Birth of

neonates on the northern slope of the Main

Caucasus Ridge occurs in August. On the

southern slope it occurs throughout

September. Bozhansky (1983) found two

groups of specimens during the summer.

One group containing males and females

which do not reproduce that year, and the

other group containing gravid females.

Bozhansky suggests that in montane

conditions, female Vipera dinniki have a

reproductive cycle of many years.

Development of the young. In late July to

early August viper embryos reach 70 mm in

length (Orlova 1973). Mean body length of

the neonates is 131.0 mm, tail length is

14.8 mm, mean weight is 3.1 g (N=28).

In the highlands almost right after birth the

vipers enter hibernation. Unlike newly

born Vipera kaznakowi, they do not feed

until the next season. By the third year the

vipers become mature.

Territoriality. Gravid females tend to move

on small areas ranging from 1-4 to 51 m2

Their individual places to a great extent (up

to 98%) overlap. Within their sites the

vipers actively utilize only 2 or 3 places

where they may be encountered at different

times and under various weather

conditions. During the morning the vipers

can be found in places with a western

exposure. Normally, this place is on a rock

surface shaded by a bush. The vipers

utilize direct sun for a short time, leaving a

part of the body under the sun and a part

under a half shaded area. From 1 100-1200

the vipers retreat to burrows and normally

appear after 1500 hours. In gloomy

weather they lie flat on a rock during the

entire period of activity. This increases the

surface of the body contiguity with the

rock. On rare hot days, after midday the

vipers never reappear on the surface.

Males and non-reproducing females emerge

much more seldom, normally after they

have taken food. Bozhansky (1984)

observed some specimens using the same

territories for three seasons.

Insolation is of great importance in the

viper's life. Gravid females take food

rarely and two months before giving birth,

they stop eating. Territory utilization is

entirely dependent on optimal insolation

patterns. Absence of aggressiveness

allows all adult females in a territorial group

to utilize the warmest microhabitats

(Bozhansky 1983, 1984). These data were

collected by Bozhansky during three

summers at an elevation of 2000 m along

the border between the forest and subalpine

belts on Aishkha Ridge, the right ridge of

April 1990

Asiatic Herpetological Research

Vol. 3, p. 21

the Mzymta River valley, the western

Caucasus. It is amazing that Vipera

kaznakowi from the low altitudes of the

coast have considerably larger individual

sites. The females do not have a multi-year

cycle and never stop taking food while

gravid. The adaptations in the highland V.

dinniki are very important in relatively low

temperatures, rainy periods, and when it

suddenly becomes cold.

Seasonal dynamics in populations.

Biological monitoring in highland

populations from the Caucasus Reserve

testify to their climax state (the collections

of the Zoological Museum of the Moscow

State University and the Zoological

Institute, the USSR Academy of Sciences

were adopted as a zero counting point).

Seasonal dynamics within climax

populations is monotypical. At the

beginning of spring, males emerge first.

Then the percentage of female occurrences

gradually increases. By early-mid June the

sex ratio is 1:1. In mid-late summer males

are seen on the surface less frequently than

females. During the period of pre-

hibernation at the end of summer, males

become more common again. From late

August to mid September, 80% of the

snakes observed are gravid females and

new born individuals. The later can be

observed on the surface even when adults

have entered hibernation. During this

period, the location of new bom individuals

does not necessarily correlate with rock

outcrops which are used as winter shelters.

The newborn snakes migrate much more

than adult snakes.

Seasonal and daily activity. In the spring

vipers emerge from mid April to May when

mean day temperature on the surface

reaches 11°C. Seasonal activity is

dependent on weather conditions. In the

highlands of the Caucasus Reserve the first

snow usually falls in the second half of

September and snow is present until May.

Snow cover is 7-8 m thick. Hence, Vipera

dinniki inhabiting the subalpine belt enter

hibernation in the second half of

September. At elevations of 1800 to

2400 m morning activity is hardly

expressed, whereas evening activity is

shifted from 1700 to 2000 hrs. In gloomy

weather snakes are active throughout

daylight at temperatures higher than 10°C.

At 8°C vipers are not seen on the surface.

Gravid females are seen on the surface even

in drizzling rain. Even if the temperature is

10°C, body temperature in vipers is 30°C,

and cloacal temperature is 26-28°C due to

accumulation of warmth from solar

radiation. This is an important thermal

adaptation of a number of highland reptiles.

Sympatric species. Throughout nearly the

entire range Vipera dinniki is sympatric

with Lacerta caucasica, L. saxicola, Anguis

fragilis, and Coronella austrica.

Vipera darevskii Vedmederja, Orlov,

andTuniyev, 1986

(Fig. 10, 11, 12, 13)

Chronology of species description

Vipera kaznakowi dinniki - Darevsky,

1956:128.

Vipera kaznakowi darevskii - Vedmederja,

1984:8, nomen nudum.

The English common name is Darevsky's

Viper.

Holotype: ZIN 19934, an adult female

from Legli Mountain, the Mokrye (Wet)

Mountains, Gukasynsky region, Armenia.

The specimen was collected in June, 1980

by I. S. Darevsky. The specimen is

preserved at the Zoological Institute, the

USSR Academy of Sciences, Leningrad

(Fig. 10).

Paratypes: ZIN 16546 a and b. The

specimens were collected May 28, 1954;

ZIN 17545, the specimen was collected

August 6, 1955; ZIN 19935, the specimen

was collected June, 1980 by I. S. Darevsky

(Fig. 11).

Holotype description: Body length is

421 mm, head included. Tail length is 46

mm. A female. Head is slightly impressed

dorsally. Lateral snout edges are slightly

pointed. Anterior edge of snout is slightly

rounded. Rostral is narrow. Frontal is

Vol. 3, p. 22

Asiatic Herpetological Research

April 1990

FIG. 10. Holotype of Vipera darevskii (ZIN

19934). a. dorsal view, b. ventral view, c. head.

narrow. Its length is equivalent to 1.66

times its width. Parietals are slightly longer

than the frontal. The frontal is separated

from supraoculars, which protrude over

lateral edge of the snout, by a row of three

scales. Prefrontal is triangular, three times

shorter than the frontal, and is separated

from the rostral by 2 scale rows. Nostril is

cut through in the lower part of the nasal.

The latter is separated from the rostral by a

broad scale. Upper labials and lower

labials are both 9 to the right and 8 to the

left. There are 5 rows of throat scales.

Around the center of the body there are 21

scale rows with strongly expressed keels,

except for two scale rows on both sides

adjacent to the ventrals, which are smooth.

The number of ventrals is 138. There are

25 pairs of subcaudals.

The color is yellowish grey. A zig-zag

shaped brown stripe runs along the

dorsum. At the center of the body its width

is nearly 8 mm. A row of hardly

conspicuous spots is present laterally. The

spots merge into a light brown stripe.

Dorsally, the head has light yellowish spots

along the edges of the frontal, parietals and

lower oculars with yellowish temporals.

Ventrum is blackish marked by light

contours of ventrals (Fig. 10).

Paratypes: Morphological characters of 8

paratypes are listed in the tables 1-4.

General color resembles that of the

holotype, except for 2 specimens. In the

latter the dorsal stripe is interrupted in the

anterior part of the body (Fig. 11).

Diagnosis: This viper is not large. SVL

reaches 460 to 489 mm. Head is slightly

impressed dorsally, covered by big scales.

Lateral edges of snout are slightly pointed.

Anterior snout edge is rounded. Nostril is

cut through in lower part of the nasal.

Head is narrow, hardly separated from

body.

Remarks and variability.

Morphologically, Vipera darevskii

occupies an intermediate position between

V. kaznakowi and V. ursini eriwanensis.

To be more precise, morphologically V.

darevskii occupies a middle position

between the two species of the "Vipera

kaznakowi " complex (V. kaznakowi and

V. dinniki ), on the one hand, and the

steppe vipers from the "V. ursini "

complex, on the other hand. This viper is

considerably smaller in body size than V.

kaznakowi. The head is narrower and the

nuchal collar is less expressed. It differs

April 1990

Asiatic Herpetological Research

Vol. 3, p. 23

mrf ■■-■ •**<*£*♦>«

^T1 Jr.- •>"♦.:..'.- .Ti>v» Y»£*^a

■ *. .

FIG. 11. The paratypes of Vipera darevskii (ZIN 19935). a., b. dorsal view, c, d. ventral view.

from V. ursini eriwanensis by greater head

height and much less pointed upper anterior

snout edge. Pholidosis data are listed in

Tables 1-3.

Coloration is yellowish or yellowish

grey, which never occurs in Vipera ursini.

Along the dorsum a contrasting zig-zag

shaped brown stripe is present. The

ventrum is dark grey, speckled black and

white. Of the small number of known

specimens, no melanistic individuals have

been found. The coloration of V. darevskii

is evidently more stable than that of the

polymorphic and motley colored V.

kaznakowi and V. dinniki (Fig. 13).

Yellow-greyish hues are prevalent. The

pattern is more homogeneous. The neck

transition is hardly expressed, like in V.

dinniki. It differs from V. ursini

eriwanensis by 1) a relatively high head, 2)

yellowish general coloration, 3) clear

contrasted pattern, and 4) special

pholidosis.

Sexual dimorphism. Maximum body size

is greater in females than males (Table 2).

Males have longer tails. Number of ventral

scales is greater in females (Fig. 14).

Number of subcaudal scales is greater in

males. Sexual dimorphism in color is not

recorded.

Vol. 3, p. 24

Asiatic Herpetological Research

April 1990

<fc* ' ■ k

FIG. 12. Vipera darevskii in the field.

Age variability. Newly born vipers are

uniquely colored.

Geographic range and ecology. The

species range shows relict characters. It

covers the southeast part of the submeridian

Dzhavakhetsky Ridge within the territory of

Armenia (here this ridge has the name

Mokrye (Wet) Mountains) and evidently

adjacent regions of Georgia (Fig. 1: 9).

Vertical distribution lies within a narrow

interval at elevations from 2600 (rarely

2500) to 3000 m on Mount Legli, Georgia.

It is a montane meadow subalpine species

inhabiting detritus slopes with great

amounts of big black outcrops of volcanic

rocks at angles of 35°-45° (Darevsky pers.

comm.). At recorded elevations, according

to Tumadzhanov (1980), subalpine

meadows (so called Transcaucasian oats

outcrops) are widely represented. It is a

leading formation on steep slopes with all

exposures which form the first stages of

invasion of bare rock outcrops.

Characteristic components of these

meadows are as follows: Festuca

woronowii, F. ovina, Bromopsis variegata,

B. villosula, and Carex tristis. Lower

along the slopes at elevations from 2300 to

2400 m the formations of transition

montane-meadow steppe are replaced by

montane Festuca valesidea and Stipa

(feather grass). From elevations of 2200 m

down to 1000 m Festuca valesiaca steppes

occur (Lavrenko 1980). Here in the

montane-steppe highland ecological belt,

Vipera ursini eriwanensis occurs in contact

with V. darevskii in the transition zone of

the meadow steppes.

The numbers of Vipera darevskii are not

great. To date it is known only from the

type locality cited as "Mount Legli", where

it occurs within a narrow band in subalpine

meadows. The species biology is nearly

unknown.

April 1990

Asiatic Herpetological Research

Vol. 3, p. 25

FIG. 13. Polymorphism in Vipera dinniki from the population that occurs in the valley of the upper part

of the Mzymta River.

Discussion

Phytogeny and the history of present viper

distributions in the Vipera kaznakowi

complex in the Caucasus Isthmus.

The head of the vipers assigned to

Vipera berus, V. ursini, and the V.

kaznakowi complex (Fig. 15) is covered

with rectangular scales (the rostral, nasals,

nasal-rostrals, upper oculars, parietals, and

the frontal), unlike in V. lebetina, V.

xanthina, V. raddei, and V. persica. In the

latter species, small scales on the upper

head are typical.

Marx and Rabb (1965) consider these

characters important in the assessment of

phylogenetic relationships of the vipers.

On the basis of body vertebrae close

relationships and isolation of Vipera berus,

V. ursini and V. kaznakowi within the

genus Vipera have been proposed

(Chkhikvadze and Zerova 1983). They

restore the prior generic name of Pelias to

the vipers. Reuss (1935) noted the

isolation of the vipers that have large

regular head plates, on the basis of skull

kinesis and analysis of the head muscles.

Without going into extensive detail on the

taxonomic position of these vipers with

large regular head scales from the Euro-

Siberian group, we shall simply refer to

them as shield-headed vipers. These vipers

show great morphological resemblance.

They are share a common color pattern,

behavior, and reproductive biology.

We consider Vipera berus and V .

kaznakowi to be close relatives. These

vipers probably diverged from a mesophilic

form in the Miocene. Northern populations

tend to be restricted to humid biotypes up to

tundra areas, whereas for southern

populations, humid subtropical forests are

inhabited. In Kramer's (1961) book

concerning the taxonomy of V. ursini and

V. kaznakowi, in the chapter on phylogeny

and history of the vipers distribution, he

regards V. berus and V. ursini as a united

circle of races, a rassenkreis. Such facts

that V. ursini renardi was subspecifically

included in V. berus also emphasizes close

relationships of shield-headed vipers.

Vol. 3, p. 26

Asiatic Herpetological Research

April 1990

FIG. 14. A female Vipera darevskii from the vicinity of Gukasyansky Region, Armenia (ZIN 19934).

Basoglu (1947) recognized the following

varieties within V. berus: berus, renardi,

ornata. His view point was criticized and

the V. ursini renardi was restored

(Terentyev and Chernov 1949; Mertens

1952 a, b). Kramer (1961) noted great

morphological resemblance in V. berus

and V. kaznakowi. He considered them

April 1990

Asiatic Herpetological Research

Vol. 3, p. 27

FIG. 15. The shield-headed vipers from the Caucasus, a. Vipera kaznakowi, b. V. dinniki, c. V. ursini

renardi, d. V. dinniki (Lagodekhi, Georgia. ZIN 8389).

phylogenetically close species and felt that

montane populations of the vipers played a

big role in the formation and isolation of the

forms. In his opinion, when it became cold

they were forced out of the mountains onto

the plains by glaciers. When the climate

became warm again, they retreated into the

mountains. In the mountains it was easy to

choose temperature and humidity optima,

whereas on the plains this was not

possible. Kramer felt that in some places

on the plains where refugia might be

present, the climate remained more or less

stable. For the vipers in the V. ursini

complex, Kramer suggests first separation

in the Miocene and Pliocene and later

independent development of the eastern and

western forms. The complex formation of

the climate and relief in the Caucasus

coincides with regular regressions and

transgressions of the sea. Breaks in links

between the faunas of the Caucasus and the

European platform, the Balkans and Turkey

stipulate emergence and isolation of original

viper forms, beginning with the Miocene

and Pliocene. At this time the formation of

the Caucasus as a montane country

occurred (Bogachev 1938). Schwarz

(1936) suggests that the viper fauna of the

Mediterranean islands are Miocene relicts.

Present ranges of shield-headed vipers in

Vol. 3, p. 28

Asiatic Herpetological Research

April 1990

the Caucasus have precise altitudinal and

ecological limits to their distributions.

They are supported by natural historic

events. Overlap of ranges occurs in

comparatively small areas. The

"kaznakowi "like vipers evidently invaded

the Caucasus during the Miocene from the

south, when the Caucasus island was

connected with Middle Asian land. The

southern invasion of the Caucasus in upper

Miocene by different mammal species was

observed by Vereschagin (1958); that of

lizards from the Podarchis-Archaeolacerta

groups was noted by Darevsky (1967).

Fossils of Vipera have been known in

Europe since the Miocene (Tatarinov

1964). Of the European Miocene fossils,

Provipera boettgeri Kinkelin 1892, was

described. Of those from lower Pliocene,

V. gedulyi Bolkay 1913, was described by

Marx and Rabb (1965). This was at about

the time the Caucasus Mountains were

formed (Bogachev 1938). A warm

subtropical climate which stimulated rich

development of mesophilic vegetation, and

contributed to the formation and wide

distribution of warmth and mesophilic

species like V. kaznakowi in the Caucasus,

including the Adzharo-Imeretinsky Ridge

where even until the Pliocene, bay trees,

rubber plants, araucarias and palms were

preserved (Vereschagin 1958). Different

mammal fossils: Middle Asian and

Caucasus hamsters, Mesocricetus,

Prometheomys, Sorex, and Talpa

(Vereschagin 1958) and also those of

insects: Orthoptera, Hemiptera, Blattoidea,

Coleoptera (Rodendorf 1939) testify to the

presence of good foodstuff for the shield-

headed vipers in the Miocene.

The end of the Tertiary period was

marked by a decrease in tectonics and with

the emergence of a broad link between the

Caucasus and the Balkans via the Crimea

(Vereschagin 1958). Steppes arose in

northern areas adjacent to the Black Sea

coast (Pidoplichko 1954; Scherbak 1966).

During this period an arid adapted

"urisini "-like viper associated with steppe

areas penetrated into western Precaucasia

from the east. It might have already

separated from Vipera ursini renardi and

have become widely distributed throughout

the plains area along the northern slope in

the Big Caucasus.

The middle-upper Pliocene, when the

ridges of both the Big and Small Caucasus

were subjected to considerable glaciation,

should be considered the beginning of the

initial break in the range of Vipera

kaznakowi (Markov et al. 1965). The

Kolchida (Colchis) had become a basic

nucleus in V. kaznakowi dispersal. There,

even in the epochs when it became severely

cold in the Pleistocene, a relatively warm-

loving vegetation of a Caucasus type was

preserved (Vereschagin 1958). Along with

Kolchida (Colchis) much smaller refugia

were sporadically preserved along the

Black Sea coast up to the town of

Lazorevskoye. The same is true for the

northern slope of the Main Caucasus Ridge

between the Pshekha and Malaya rivers.

This is supported by the present

distribution of the Tertiary vegetation of the

Kolchida (Colchis) type in the western

Caucasus (Adamyants 1971; Kharadze

1974; Pechorin and Lozovoy 1980;

Kholyavko et al. 1978; Tuniyev 1990, this

volume).

It is in the narrow humid canyons with a

relatively regular thermal regime that Vipera

kaznakowi is preserved. Also, isolated

populations might be preserved in midlands

where the refugia with Kolchida (Colchis)

vegetation are known in regions of the

Fisht-Oshterovsky Massive, the Lago-Naki

Plateau, and even in the Central Caucasus

(Kharadze 1974; Kholyavko et al. 1978).

Small refugia might also be preserved along

the southern slope of the eastern Big

Caucasus. Similar refugia for

Archaeolacerta were recorded by Darevsky

(1967). Most highland populations of V.

kaznakowi undoubtedly went extinct

during the glacial period. Those that were

preserved in refugia have been

accumulating unique characters. Data

obtained by Takhtadzhan (1946) and

Maruashvili (1956) reveal that in glacial

epochs, mean annual temperature never

dropped lower than 1.5-2.0°C, and the

amount of precipitation was never less than

1500-2000 mm. These data support the

proposed preservation of relictual V .

April 1990

Asiatic Herpetological Research

Vol. 3, p. 29

kaznakowi populations in the mountains.

Darevsky (1967) regards this argument

as proof that due to the development of

montane glaciers, a fundamental

reconstruction of all animals and plants in

these areas occured from premontane and

montane refugia. Lizards for example,

might have been preserved in the

Gagrinsky and Bzybsky ridges facing the

sea, along with other regions. The end of

the Pleistocene was marked by the

development of steppe landscapes on the

Zakubauskaya sloping plain along with

already formed landscapes and a

characteristic biological community of the

present type along the Black Sea coast of

the Caucasus (Vereschagin 1958). This

situation contributed to a much wider

dispersal of Vipera ursini renardi and

further preservation of V. kaznakowi.

During the xerothermal epoch in the

Holocene, intensive glacial melting and the

upwards elevational shift of vegetation belts

along the mountain slopes occurred

throughout the area. Within a considerable

area of the Caucasus, steppe landscapes

became prevalent. This allowed the steppe

animals, including the steppe viper, Vipera

ursini renardi, to ascend into the mountains

(Vereschagin 1958). At that time an initial

contact of V. ursini renardi with montane

relict populations of V. kaznakowi might

have occurred in the midmountains of the

Western Caucasus. This region is

presently inhabited by V. dinniki which

shows mixed morphologies of both V.

kaznakowi and V. ursini renardi. The

approximate ways of possible interaction

and characters are shown in Fig. 16.

Darevsky (1967) writes about the form-

building role of hybridization in the

Caucasus. He observes the genesis of new

forms in the lizard group Archaeolacerta.

Hybridogenity in Vipera dinniki is

problematic, whereas introgressing

hybridization in the zones of contact

between V. kaznakowi and V. dinniki is

presently not doubted, age interbreedings

apparently occurring in midmontane

populations of V. dinniki. For example, in

one of these populations associated with the

Mzymta River head, all the initial types of

vipers can be found (Fig. 15). In

premontane populations of V. kaznakowi

we have not found specimens which have

hybrid characters. Generally all

premontane populations of V. kaznakowi

show great homogeneity. In the eastern

range of V. dinniki the species is presently

disjunct geographically from the closely

related V. ursini renardi, but rare

individuals with an intermediate

morphology between these species are

present.

Mayr (1968), Borkin and Darevsky

(1980) and Solbrig and Solbrig (1982)

report on various types of hybridization.

For instance, it is recorded that hybrids

possessing vital capacity occur between

sympatric species. Some of these hybrids

are capable of recurrent crosses with one or

both parental forms. The fate of both

species between which hybridization occurs

is partially dependent on 1) the intensity of

hybridization occurring between them and

2) the level of genetic and ecological

sterility of hybrids (Solbrig and Solbrig

1982).

In most cases the hybrids themselves do

not have greater evolutionary significance.

Rather it is the products of recurrent

interbreeding between the hybrids and the

parental forms, i.e. the process of

introgressive hybridization, that may form

new taxa; i.e. the blending of genomes of

ancestral taxons, or "borrowing" of parts

from other genomes by means of

introgression of genes as a variant of

hybridogenic species formation (Mayr

1968; Borkin and Darevsky 1980).

After the xerothermal epoch, the climate

again became more humid. This fact

contributed to the restoration of the former

limits of the forest belt (Vereschagin 1958).

Throughout the subalpine belt along the

southern slope of the Main Caucasus Ridge

within the area from the Central Caucasus

up to the Fisht-Oshtenovsky Massive in the

west, subalpine meadows and montane

curved forests are widely developed

(Galushko 1974; Dolukhanov 1974;

Kharadze 1974; Kholyavko et al. 1978).

Vol. 3, p. 30

Asiatic Herpetological Research

April 1990

V. ursini eriwanensis

V. darevskii

I

FIG. 16. Model of the proposed genesis of the vipers from the Vipera kaznakowi complex.

April 1990

Asiatic Herpetological Research

Vol. 3, p. 31

In the western part, this vegetation is

present along the northern slope. To the

east, it gradually changes to steppe

(Lavrenko 1980). Definite limits of the

subalpine belt in the area influenced by the

warm Black Sea mark the present

distributional limit of Vipera dinniki.

The final establishment of the present

climate stabilized the range of Vipera

kaznakowi and allowed the joining of

scattered populations from the refugia

within the given region. The favorable

conditions along the Black Sea coast made

it possible for V. kaznakowi to disperse

along warm river valleys up to

midmountains and for V. dinniki to

disperse along the northern slope of the Big

Caucasus to the east up to the Belshaya

Laba River head. Where V. kaznakowi

and V. dinniki distributions meet, and

those of V. dinniki and V. ursini renardi,

zones of secondary hybridization emerged.

In these zones the following is observed:

intergradation of characters, a high degree

of polymorphism, and the "plucking" of

specimens from the "initial" type. In these

populations the specimens are hard to

define.

Phenotypical diversity is great with

regard to the Mzymta River valley

(Fig. 15). Non-identified individuals

having characters of both Vipera ursini

renardi and V. kaznakowi are known from

a number of localities in the complex

eastern range of these vipers. We think that

the history of V. darevskii is also

connected with the invasion of V .

kaznakowi from the south during the

Miocene and its further wide dispersal

throughout the Caucasus. Vipera

kaznakowi apparently used to inhabit the

majority of the Maly (Small) Caucasus

area, including the Adzharo-Imeretinsky,

Meskhetsky, and Dzhavakhetsky ridges.

The fact that this range did exist in the past

is demonstrated by a relict population of V.

kaznakowi that occurs in Baniskhevsky

Canyon, Georgia and by the proposed

distribution of the vipers along the

Trialetsky Ridge (Bakradze 1969). The

latter might be a link between the areas

presently inhabited by the closely related

species, V. kaznakowi and V. darevskii.

In middle Pliocene, Vipera kaznakowi

apparently invaded the Dzhavakhetsky

Ridge (the Mokrye [Wet] Mountains). At

that time the hot, arid climate of the region

was replaced by a more humid, cool

climate. Later on, however, beginning

with the middle Pleistocene, processes of

aridization and glaciation caused the final

degradation of forests (Agakhanyants

1981). By that time the diverged vipers

were restricted to the highest places in the

mountains where maximum humidity could

be found. At the same time the newly

formed steppe areas were actively invaded

by V. ursini eriwanensis (Reuss 1935)

from the Anatolyskoye Plateau. Later on

the viper became widely distributed

throughout the entire montane-steppe belt in

the Small Caucasus and the Armyanskoye

Plateau. At that site one more source of

hybridization, including further form-

building, evidently existed. The genesis of

V. darevskii can be explained by the

hybridization of V. ursini eriwanensis and

V. kaznakowi that split from its major

range in the vicinity of the Mokrye

Mountains (Plate 3).

Apparently, introgression of genes from

Vipera ursini eriwanensis which essentially

surrounded a small Pleistocene population

of V. kaznakowi was very substantial.

Morphological propinquity of V. darevskii

to V. ursini eriwanensis and V. kaznakowi

support its possible hybrid genesis (Borkin

and Darevsky 1980). It goes without

saying that further ecological, ethological,

cytological and biochemical research of

these vipers is needed to prove it.

Similarly, Agakhanyants (1981) explains

the floristic richness of the region by a

complex interaction of invaders from the

south and north in the highlands of the

Small Caucasus. Further aridization of the

climate caused the mesophilic form of V.

darevskii to ascend into the mountains.

Ecological disconnection in the present

high-altitude belts caused complete

separation of V. darevskii and V. ursini

eriwanensis.

The presence of "kaznakowi " -like

Vol. 3, p. 32

Asiatic Herpetological Research

April 1990

vipers in Lagodekhi and other regions of

the Big Caucasus may be accounted for

solely by changes in climate in refugia that

preserved populations and further

divergence in isolated montane canyons. A

number of isolated populations might by

modified due to hybridization with Vipera

ursini renardi in a manner similar to the

above listed scheme.

During the period of xerophytization in

the Holocene, breaks within the range of

Vipera kaznakowi occurred. At that time

isolated populations existed over a long

period. A number of populations in

hydrophilous epochs to follow, would have

been capable of connecting many times.

This is supported by the restoration of the

former forest belt after Holocene

xerophytization (Galushko 1974; Kharadze

1974). Pleistocene glaciation obviously

also influenced the emergence of isolated

populations and changes in the

distributional limits (Markov et al. 1965).

Acknowledgments

The assistance of many people

contributed to the completion of this study.

We would like to extend our special

recognition to II ya S. Darevsky and Natalia

B. Ananjeva from the Leningrad Zoological

Institute, USSR Academy of Sciences who

aided us in the collection of data, analysis

of the material, and preparation of this

manuscript We are grateful to Valentina F.

Orlova from the Zoological Museum, the

Moscow State University and Valery I.

Vedmederja from the Museum of Natural

History, the University of Kharkav,

Ukraine for providing an opportunity to

work with systematic museum collections.

Illustrative assistance was provided by

Rostislav A. Danov.

Literature Cited

ADAMYANTS, G. I. 1971. [On Castanetum of the

Caucasus]. Reports of the Sochi Department of

the USSR Geographic Society, Science

Publishing, Leningrad. Issue 2:398-404. (In

Russian).

AGAKHANYANTS.O. Y. 1981. [Arid mountains

of the USSR]. Mysl Publishing, Moscow.

270 pp. (In Russian).

ANONYMOUS. 1978. [Red Data Book of the

USSR]. Lesnaya Promyshlennost Publishing,

Moscow. (In Russian).

ANONYMOUS. 1984. [Red Data Book of the

USSR]. Lesnaya Promyshlennost Publishing,

Moscow. (In Russian).

BAKRADZE, M. A. 1969. [New data on the range

of the Caucasus viper {Vipera kaznakowi

Nikolsky) in Georgia]. Academic Press,

Tbilisi, Georgia 56(2):467-468. (In Russian).

BAKRADZE, M. A. 1975. [Distribution of Vipera

laurenti in south Georgia]. Bulletin of the State

Museum of Georgia, Tbilisi 28a:41 1-412. (In

Russian).

BANNIKOV, A. G. AND V. M. MAKEYEV. 1976.

[Conservation of venomous snakes of the

USSR]. Priroda Publishing, Moscow 5:71-83.

(In Russian).

BANNIKOV, A. G., I. S. DAREVESKY, V. G.

ISCHENKO, A. K. RUSTAMOV, AND N. N.

SCHERBAK. 1977. [Field guide of the USSR

amphibians and reptiles]. Prosveschenye

Publishing, Moscow. 414 pp. (In Russian).

BARAN, J. 1977. Tiirkiye Gilanlamin Taxonomik

Revizyonu ve cografi Dagilislari. TBAK,

Ankara. TBAK ser. 9, n. 309.

BARTENEV, A. N. AND M. A. REZNIKOVA.

1935. [Papers on the fauna of amphibians and

reptiles from the western and central parts of the

State Caucasus Reserve]. Bulletin of the

Museum of Georgia, Tbilisi 8:11-41. (In

Russian).

BASOGLU, M. 1947. On some varieties of Vipera

berus from the extreme north-eastern Anatolia.

Rev. Fac. Sci. Univ. Istanbul. (B), 12/1:182-

190.

BASOGLU, M. AND J. BARAN. 1977. Tiirkiye

siiriingleri Kism. I. Gilandar. The Reptiles of

Turkey. Part II. The snakes. Bomova-Izmir Ege

Universitesi Matbaasi. 205 pp.

BOETTGER, O. 1893. Katalog der Reptilien-

sammlung im Museum der Senckenbergischen

Naturforschenden Gesellschaft in Frankfurt am

Main. Frankffurt a. M. Teil 1-2. 160 pp.

April 1990

Asiatic Herpetological Research

Vol. 3, p. 33

BOETTGER, O. 1899 Reptillia et Batrachia. Pp.

274-294. In Radde, Die sammlungen der

Kaukasischen Museums, I, Zoologie, Tiflis.

Caucasus Department of the Russian

Geographic Society, Vladikavkaz 32(5):9-10.

(In Russian).

BOGACHEV, V. V. 1938. [On the zoogeography

of the Caucasus and its geological history].

Bulletin of the Azerbajzhan Affiliation of the

USSR Academy of Sciences, Baku no. 4:29-47.

(In Russian).

BORKIN, L. Y. AND I. S. DAREVSKY. 1980.

[Hybridogenic species formation in vertebrates].

Journal of General Biology, Moscow 41 (4):485-

505. (In Russian).

BOZHANSKY, A. T. 1979. [Distribution and daily

activity of the Caucasus viper on southern

slopes of the Big Caucasus]. Pp. 15-80. In

Materials of the Second All Union Conference

of Young Scientists on Conservation of Fauna

and Flora, preserved at VNIIITEISH. Moscow.

N. 100. (In Russian).

BOZHANSKY, A. T. 1983. [Territory relationships

in groupings of the Caucasus viper]. Pp. 90-

91. In Behavior of Animals in Communities.

Materials of the Third All Union Conference on

Animal Behavior, Moscow, Vol. 11. (In

Russian).

BOZHANSKY, A. T. 1984. [The Caucasus viper].

Priroda, Moscow 41(6):53-55. (In Russian).

CHERNOV, S. A. 1929. [Materials on the fauna

of Amphibia and Reptilia of highland

Ingushetia]. Bulletin of the Ingushsky Research

Institute of Natural Study, Vladikavkaz 2:93-

110. (In Russian).

CHKHIKVADZE, V. M. AND G. A. ZEROVA.

1983. [Morphology of body vertebrae and

systematic position of some viper snakes].

Reports of the Academy of Sciences of Georgia,

Tbilisi 112(3):653-656. (In Russian).

DAREVSKY, I. S. 1956. [A new species of the

venomous snake, Vipera Icaznakowi Nikolsky,

for the fauna of Armenia]. Bulletin of the

Academy of Sciences of Armenia, Yerevan

9(12): 127- 130. (In Russian).

DAREVSKY, I. S. 1967. [Rock lizards from the

Caucasus (Lacerta saxicola )]. Science

Publishing House, Leningrad. 214 pp. (In

Russian).

DINNIK, N., Y. 1902. [The upper reaches of the

Malaya Laba and Mzymta rivers]. The West

DINNIK, N. Y. 1926. [Snakes of the Northern

Caucasus]. Scientific Notes of the Northern

Caucasus Institute of Natural Study,

Vladikavkaz 1:11-17. (In Russian).

DOLUKHANOV, A. G. 1974. [Subalpine

landscapes of the Caucasus as shelters for relict

fauna elements]. Problems of Botanies,

Leningrad 12:27-34. (In Russian).

FYODOROV, S. M. 1956. [Reptiles, amphibians,

Cyclostomata, and fishes of Stavropolsky

Territory]. Pp. 205-220. In Materials of the

Study of Stavropolsky Territory, Stavropol,

Issue 8. (In Russian).

GALUSHKO, A. I. 1974. [Relicts of the highland

flora of the western part of the Central

Caucasus]. Pp. 19-26. In Problems of

Botanies, Vol. 12, Flora of Highlands and its

Exploration. Science Publishing Leningrad.

(In Russian).

HARDING, K. A. AND K. R. G. WELCH. 1980.

Venomous snakes of the world. A checklist.

Pergamon Press, Riverside, New Jersey. 188

pp.

KHARADZE, A. L. 1974. [On some floristic

groups of endemics of the Big Caucasus]. Pp.

70-75. In Problems of Botanies, Vol. 12,

Flora of Highlands and its Exploration, Science

Publishing, Leningrad. (In Russian).

KHOLYAVKO, V. S., D. A. GOLBA-

MIKHAYLENKO AND Y. S. KHOLYARKO.

1978. [Atlas of arborious races of the

Caucasus]. Lesnaya Promyshlennost

Publishing, Moscow. 215 pp. (In Russian).

KLEMMER, D. 1963. Liste der rezenten

Giftschlangen. Elapidae, Hydrophiidae,

Viperidae und Crotalidae. Pp. 255-464. In

Behringwerk-Mitteil. Die Giftschlangen der

Erde. Marburg/Lahn.

KNOEPFLER P. S. AND E. SOCHUREK. 1955.

Neues iiber Rassen der Wiesenotter.

Burgenlandische Heimatbiatter 4: 1 85-1 88.

KOROSTELYOV, N. A. 1933. [Climatic sketch of

the Black Sea coast of the Caucasus]. Pp. 143-

168. In The Black Sea Coast of the Caucasus.

In Standardization and Rationalization,

Vol. 3, p. 34

Asiatic Herpetological Research

April 1990

Moscow-Leningrad. (In Russian).

KRAMER E. 1961. Variation,

Sexualdimorphismus, Wachstum und

Taxonomie von Vipera ursini (Bonaparte,

1835) und Vipera kaznakowi Nikolsky 1909.

Revue Suisse de Zoologie 61:627-725.

KRASOVSKY, D. 1929. [Materials on the fauna of

Reptilia et Amphibia of Khasav-Yurt region of

Dagestan]. Bulletin of the Gorsky Pedagogical

Institute, Makhachkala 6:175-200. (In

Russian).

KRASOVSKY, D. 1932. [Materials on the fauna of

terrestrial vertebrates of Rutulsky Kanton of

Dagestan]. Bulletin of the Second Northern

Caucasus Pedagogical Institute 10:89-95. (In

Russian).

LAVRENKO, Y. M. 1980. [Caucasus montane and

meadow steppes and stepped meadows]. Pp.

269-272. In Vegetation of the European part of

the USSR, Science Publishing, Leningrad. (In

Russian).

LEISTER, A. F. 1908a. [On the geographic range

of V. renardi Christoph and Vipera berus L.

within the Caucasus]. Works of the Kharkov

Society of Nature Researches at the Royal

University of Kharkov, Kharkov 42:3-7. (In

Russian).

LEISTER, A. F. 1908b. [New data on the

herpetology of the Terskaya region]. Bulletin of

the Caucasus Museum, Tiflis, Georgia 1-

4(4):209-217. (In Russian).

LEISTER, A. F. 1908c. [On the fauna of the

Erivan region. Notes on findings of the steppe

viper Vipera renardi Christoph in

Transcaucasis]. Pp. 223-226. In Press of the

Erivan Regional Statistics Committee, Erivan,

Armenia. (In Russian).

MARKOV, K. K., G. I. LAZUKOV, AND V. A.

NIKOLA YEV. 1965. [The Caucasus]. Pp. 306-

321. In Quaternary Period, v. 1, Territory of

the USSR, part 2. Area of Ancient Glaciation

of High Mountains of the USSR South, chapter

2, Moscow University Press, Moscow. (In

Russian).

MARUASHVILI, L. I. 1956. [The reason to review

the present day concepts about

paleogeographical conditions of the glacier

period in the Caucasus]. Press of the Institute

of Geography named after Vakhushti, Tbilisi,

Georgia. 113 pp. (In Russian).

MARX H. AND G. RABB. 1965. Relationships

and zoogeography of the Viperine snakes

(Family Viperidae). Chicago Natural History

Museum 44(21): 161-206.

MAYR, E. 1968. [Zoological species and

evolution]. Moscow 597 pp. (In Russian).

MENfiTRlfiS, E. 1832. Catalogue raisonnddes

objects zoologiques de Caucase. St.

Petersbourg, Academy Imperial, des sciences.

272 pp.

MERTENS, R. 1952a. Amphibien und Reptilien

aus der Tiirkei. Rev. Fac. Sci. Istanbul (B),

17:41-75.

MERTENS, R. 1952b. Nachtrag zu "Amphibien

und Reptilien aus der Tiirkei". Rev. Fac. Sci.

Univ. Istanbul (B) 17:353-355.

MERTENS R. AND H. WERMUTH. 1960. Die

Amphibien und Reptilien Europas. Dritte

Liste, nach dem Stand vom 1 Januar 1960,

Frankfurt a. Main. 264 pp.

MILYANOVSKY, Y. S. 1957. [On the snake fauna

of Abkhazia]. Works of the State Museum of

Abkhazia, Sukhumi, Tbilisi 2:199-203. (In

Russian).

MORITS, L. D. 1916. [On the snakes of the

Northern Caucasus]. Wildlife Hobbyist,

Petrograd, Russia 1-2:1-21. (In Russian).

MUSKHELISHVILI, T. A. 1959. [On the

herpetofauna of Svanetia]. Report of the

Academy of Sciences of Georgia, Tbilisi

22(6):729-831. (In Russian).

NIKOLSKY, A. M. 1905. [Reptiles and

amphibians of the Russian Empire

(Herpetologica Rossica)]. Press of the Western

Empire Academy of Sciences, Ser 8, 17(1): 1-

517. (In Russian).

NIKOLSKY, A. M. 1909. [A new viper species

from the Caucasus, Vipera kaznakowi.]

Bulletin of the Caucasus Museum, Tiflis Nos

3,4:173-174. (In Russian).

NIKOLSKY, A. M. 1911. [Reptiles and

amphibians of the Caucasus]. The Caucasus

Museum Press, Tiflis, Georgia 10(1): 1-18. (In

Russian.)

April 1990

Asiatic Herpetological Research

Vol. 3, p. 35

NIKOLSKY, A. M. 1913. [Reptiles and

amphibians of the Caucasus]. The Caucasus

Museum Press, Tiflis, Georgia 272 pp. (In

Russian).

NIKOLSKY, A. M. 1916. [Reptiles. The fauna of

Russia and adjacent countries, II, Ophidia].

Academic Press, Petrograd. 350 pp. (In

Russian).

ORLOV, N. L. AND B. S. TUNIYEV. 1986.

[Present ranges, possible ways of their

formation and phylogeny of the three species of

vipers from the Eurosiberic group (V .

kaznakowi complex) in the Caucasus].

Proceedings of the Zoological Institute, USSR

Academy of Sciences, Leningrad 157:104-135.

(In Russian).

ORLOVA, V. F. 1973. [Herpetofauna of the

northern area of the State Caucasus Reserve].

Zoological News, Kiev, Ukraine (2):61-65. (In

Russian).

PECHORIN, A. I. AND S. P. LOZOVOY. 1980.

[Natural monuments of the Krasnodar Territory].

Krasnodar Press, Krasnodar (2): 1-138. (In

Russian).

PIDOPLICHKO, I. G. 1954. [On the glacier

period]. The Ukraine Academy of Sciences

Press, Kiev, Ukraine. 139 pp. (In Russian).

REUSS.T. 1935. Erlauterungen zur Typenfrage in

der Zoologie an Hand von Bildem europaischer

Giftschlangen. Nach. Bl. Aquar. u. Terr. Ver.,

Berlin 16:215-218.

RODENDORF, B. B. 1939. [On the Miocene fauna

of insects in the vicinity of Voroshilovsk].

Priroda Publishing, Moscow, N. 12. (In

Russian).

ROSSIKOW, K. N. 1890. [In the mountains of the

northwestern Caucasus. A trip to Zaakdan and

the Bolshaya Laba River head for the purpose of

zoogeographic research]. Bulletin of the

Imperial Russian Geographic Society. 1:1-198.

(In Russian).

ROSTOMBEKOV, V. N. 1939. [On the

herpetofauna of Abkhazia]. Bulletin of the

Abkhazian Affiliation of the Academy of

Sciences, Sukhumi, Georgia. 1:117-217. (In

Russian).

SCHERBAK, N. N. 1966.

reptiles of the Crimea].

[Amphibians and

Naukora Dumka

Publishing House, Kiev, Ukraine. 239 pp. (In

Russian).

SCHWARZ, E. 1936. Untersuchungen tiber die

Systematik und Verbreitung der europaischen

und mediterranien Ottern. Behringwerk Mitteil.,

Marburg 7:159-162.

SELYANINOV, G. T. 1933. [Sketch on the

climate of Krasnaya Polyana (Red Meadow).]

Pp. 169-201. In The Black Sea Coast of the

Caucasus. Moscow-Leningrad. (In Russian).

SHKADOVA, A. K. 1979. [Temperature regime of

terrains on the territory of the USSR].

Gidrometeoizdat Publishing, Leningrad. 240

pp. (In Russian).

SILANTYEV, A. 1903. [Materials on the

herpetology of the Black Sea region]. Annual

Journal of the Zoological Museum, Academy of

Sciences, St. Petersbourg, Russia 8:1-10. (In

Russian).

SOLBRIG, O., AND D. SOLBRIG. 1982. [Biology

and evolution of populations]. Mir Publishing,

Moscow. 488 pp. (In Russian).

TAKHTADZHAN, A. L. 1946. [On the history of

the vegetation developments in Armenia].

Works of the Botanic Institute, Armenian

Academy of Sciences, Yerevan, Armenia. 4:51-

107. (In Russian).

TATARINOV, L. P. 1964. [Order Ophidia,

snakes]. Pp. 484-493. In Fundamentals of

Paleontology. Reptiles, Amphibians and Birds.

Science Publishing, Moscow. (In Russian).

TERENTYEV, P. V. AND S. A. CHERNOV. 1949.

[Guide to reptiles and amphibians]. Soviet

Sciences Publishing, Moscow. 315 pp. (In

Russian)

TERTYSHNIKOV, M. F. 1977 [Ecological and

zoogeographic characteristics of the herpetofauna

of the Northern Caucasus]. Fauna of

StavTopolsky Territory, Stavropol, Issue 11:3-

25. (In Russian).

TSAREVSKY, S. F. 1916. [New forms of vipers

from within the Russian Empire]. Pp. 32-38 In

Annual Journal of the Zoological Museum of

the Imperial Academy of Sciences, St.

Petersburg. (In Russian).

TUMADZHANOV, 1. 1. 1980. [Caucasus meadows

and curved forests]. Pp. 198-202. In

Vol. 3, p. 36

Asiatic Herpetological Research

April 1990

Vegetation of the European Part of the USSR,

Science Publishing, Leningrad. (In Russian).

TUNTYEV, B.S. 1990. On the independence of the

Colchis center of amphibian and reptile

speciation. Asiatic Herpetological Research

3:67-84

VEDMEDERJA, V. I. 1977. [New data on the

herpetofauna of Adzharia]. Pp. 54-55. In

Questions of Herpetology. Abstracts of the

Report at Fourth All Union Conference of

Herpetology, Science Press, Leningrad. (In

Russian).

VEDMEDERJA, V. I. 1984. [Range, variability and

peculiarities of ecology of the Caucasus viper].

Pp. 8-9. In Species and its Productivity in the

Habitat. Part V. Questions of Herpetology,

Sverdlovsk. (In Russian).

VEDMEDERJA, V. I., N. L. ORLOV AND B. S.

TUNIYEV. 1986. [On the taxonomy of three

viper species of the Vipera kaznakowi

complex]. Proceedings of the Zoological

Institute, Leningrad 157:55-61. (In Russian).

VERESCHAGIN, N. K. 1958. [Genesis of the

terrain fauna of the Caucasus Isthmus]. Pp.

506-514. In Fauna of the USSR, Vol. 5.

Montane Areas of the European Part of the

USSR, Academic Press, Moscow. (In

Russian).

ZINYAKOVA, M. P. AND A. G. TROFIMOV.

1977. [On the biology of reproduction of the

Caucasus viper, Vipera kaznakowi Nikolsky].

P. 95. In Questions of Herpetology. Abstracts

of the Report at the Fourth All Union

Conference of Herpetology, Science Publishing,

Leningrad. (In Russian).

April 1990

Asiatic Herpetological Research

Vol. 3, pp. 37-45

Studies on Hynobiid Salamanders, With Description of a New Genus

KRAIG ADLER1 AND ERMI ZHAO2

^Cornell University, Section of Neurobiology and Behavior, Mudd Hall, Ithaca, New York 14853, USA

2Chengdu Institute of Biology, P.O. Box 416, Academia Sinica, Chengdu, Sichuan, China

Abstract. -The types of Hynobius chinensis Giinther, 1889, were reexamined and are redescribed; the

known range in China is mapped. Hynobius yiwuensis Cai, 1985, is relegated to the synonymy of

chinensis. Hynobius retardatus Dunn, 1923, of Japan, differs markedly from all other Hynobius and is

here placed in a new genus (Satobius ). The new genus is characterized and compared to the other eight

genera of hynobiids.

Key words: Amphibia, Caudata, salamanders, Hynobiidae, Hynobius, Satobius, China, Japan.

Introduction

As part of our work on a handbook of

Chinese amphibians and reptiles, we have

had to re-study certain Chinese hynobiid

salamanders described in the literature and

consider their relationships to extralimital

species, especially those native to Japan.

This paper is a result of these current

investigations and concerns the identity and

distribution of Hynobius chinensis

Giinther, 1889, the taxonomic status of H.

yiwuensis Cai, 1985, and the proper

generic status of a Japanese species, H.

retardatus Dunn, 1923.

Results and Discussion

1. Identity and Distribution of

Hynobius chinensis Giinther, 1889

In 1889, Albert Giinther described

Hynobius chinensis from two specimens

collected in Hubei Province. Previously,

Hynobius species had been known only

from Japan and Korea, thus Gunther's

records from central China were quite

isolated from the ranges of the then-known

species. Since 1889, however, no other

specimens of Hynobius have been reported

from Hubei, despite much collecting there,

leading Zhao and Hu (1983, English

translation 1988) to suggest that either the

locality data for Gunther's specimens are

wrong or that chinensis is not, in fact, a

species of Hynobius.

We have examined Gunther's two

syntypes (British Museum [Nat. Hist.]

numbers 1946.9.6.54 and .55, formerly

catalogued as 1889.6.25-26). Both are

females, as judged from the external

appearance of their cloacae (the specimens

are too hardened to examine internally).

Based on our examination of the types,

H. chinensis is a true Hynobius as defined

by Zhao and Hu (1984, English translation

1988). Since Gunther's (1889) original

description is brief, we provide the

following redescription: Head large, its

length from snout to gular fold longer than

its width (15.2 x 10.6 mm; 12.8 x 9.5

mm); tip of snout rounded. Eyes are

dorsolateral in position, slighdy protruded;

diameter of eye shorter than the distance

from its anterior corner to the tip of snout;

pupil rounded. Nostril between eye and tip

of snout, and slightly closer to the latter;

distance between nostrils slightly larger or

equal to the distance between eyes. A "V"-

shaped bulge on top of head. No labial

fold. An indistinct gular fold. Angle of

jaw just behind the posterior corner of eye.

Both maxilla and mandible with tiny teeth.

Tongue elliptical, large, almost covering the

entire floor of mouth. Series of vomerine

teeth " 1/ "-shaped, the outer branch

comprising 6-9 and inner branch 11-15

teeth; the angle formed by outer and inner

branches just beyond the anterior margin of

choanae; inner branches much longer than

the outer ones and extending backwards to

the level of the middle of eye ball; the

1990 by Asiatic Herpetological Research

Vol. 3, p. 38

Asiatic Herpetological Research

April 1990

FIG. 1. Map of central China illustrating all known localities for Hynobius chinensis. The inset shows

the location of the more detailed map. Locality records in Zhejiang and Fujian provinces are noted by solid

circles; the star-shaped symbol in Hubei Province is the type locality (Yichang). Other localities mentioned

in the text are noted by hollow circles; some major cities are added for reference. The river mapped in

Sichuan Province (upstream from Chongqing) that is continuous with the Chang Jiang (Yangtze River) is

known variously as the Dajin, Dadu, and Min.

posterior ends of two inner branches close

but do not meet at midline. Body short and

stout; limbs well developed and tips of

digits meet when limbs adpressed. Costal

grooves 11, very prominent and meeting on

ventral midline. Fingers four, 2-3-4-1 in

order of length, the first finger almost equal

in length to the fourth. Toes five, 3-4-2-5-

1 in order of length. Digits flattened, free;

without palmar and tarsal tubercles; no

cornified covering on palms, tarsa, fingers,

and toes. Tail length shorter than snout-

vent length; tail compressed, but cylindrical

at the base and pointed at the end, without

crest on ventral edge and only slightly so

on its dorsal side. Skin smooth. The

dimensions cannot be re-measured due to

the specimens' hardened condition; only

head length and width can be given

(above). Further details about the types are

given in Zhao and Adler (1989). Cai et al.

(1985) have described the embryonic

development and larval features of

chinensis.

Lectotype: We hereby designate BMNH

1946.9.6.54 as the lectotype of H .

chinensis.

Type Locality: Gunther (1889) stated

that "two specimens were collected by Mr.

Pratt at Ichang [=Yichang]." (Here and

below, modern spellings of place names,

where different, are given in brackets. All

localities are mapped in Fig. 1.) A. E. Pratt

explored China during 1887-1890 and

summarized his experiences in a book

(1892). As Gunther noted in the appendix

to Pratt's book, herpetological specimens

were collected at various localities in Hubei

and Sichuan provinces, and he repeated the

"Ichang" locality for the two Hynobius

specimens. Pope and Boring (1940),

however, stated that "Mell ('29) claims that

Pratt's records from 'Ichang' were

collected in the mountains south of the

Yangtze River near Changyang."

Curiously, no such paper by Mell is listed

in the bibliography of Pope and Boring's

monograph, nor can we find such a

statement in any of Mell's publications, so

we are unable to verify the basis of Pope

and Boring's information attributed to Mell.

In his book, Pratt (1892) mentioned

finding salamanders only twice (pages 179

and 224) but both localities are in Sichuan:

a lake above Ta-tsien-lu [=Kangding], at

April 1990

Asiatic Herpetological Research

Vol. 3, p. 39

14,070 feet elevation, and at Kia-ting-fu

[=Leshan], both of which are localities far

to the west of Yichang (see Fig. 1). Four

other new species of reptiles and

amphibians were described by Gunther in

his 1889 paper, all having "Ichang" as then-

type locality, and for these taxa the locality

has never been questioned, to our

knowledge. Thus, we are inclined to

accept the type locality as stated by

Gunther, until new evidence comes to

hand, although this locality is about 700-

900 km west of the other known localities

for this species (see Fig. 1 and below).

Distribution: Subsequent to Giinther's

report, several other specimens of H .

chinensis have been reported. Here we

summarize all known records (see Fig. 1):

Fujian: Kuatun [=Guadun, 27° 42' N

117° 50' E, a town at Mt. Wuyi],

Ch'ungan Hsien [=Chong'an County, 27°

46' N 118° 01'E] (Pope, 1931).

Hubei: Ichang [=Yichang, 30° 42' N 111°

17' E] (Gunther, 1889).

Zhejiang: Dachen, 29° 28' N 120° 06' E,

140 m, and Chalin, (both in Yiwu County,

29° 18' N 120° 04' E), Zhenhai, 29°

57' N 121° 42' E, and Xiaoshan, 30°

10' N 120° 15' E, (Cai, 1985; types of

H. yiwuensis); Wenling, 28° 22' N 121°

22' E, 1500 ft., (Boring and Chang, 1933;

Chang, 1933).

Boring and Chang's specimens from

Zhejiang Province have been examined and

their identifications verified. Cai's species,

H . yiwuensis, is a synonym of H .

chinensis (see section 2, below). Thus,

chinensis is the only Hynobius found in

mainland China, except for two species in

the extreme northeast; three other species

are endemic to Taiwan.

2. Taxonomic Status of

Hynobius yiwuensis Cai, 1985

Cai (1985) described Hynobius

yiwuensis as new, based upon a series of

adults (16 males, 11 females), juveniles,

larvae, and eggs from Zhejiang Province.

Unfortunately, he was unable to compare

his specimens directly with the types of H.

chinensis. Based on our comparison with

those types, however, we believe that the

two taxa are synonymous. We base our

conclusion primarily on the six features

used by Cai to diagnose his new form.

Vomerine Teeth. Cai described the length

of the inner branch of vomerine teeth in

yiwuensis as longer than that of chinensis,

yet our examination reveals no significant

difference. Moreover, the shape of the

vomerine teeth series is very similar and the

numbers of teeth in both the inner and outer

branches are within the same ranges in the

two forms.

Head and Body Proportions. Cai stated

that in yiwuensis the head is much longer

than broad, in contrast to Giinther's

measurements for one specimen in which

the length was only slightly greater (11 x

10 mm). Our re-measurement of Giinther's

specimens, taking snout to gular fold as

head length, shows that in both types of

chinensis the head is much longer than

broad (see measurements in section 1,

above). Cai also claimed that chinensis and

yiwuensis differ somewhat in body

proportions, the latter being less stout. Our

comparisons failed to notice significant

differences, when relative sizes are taken

into account.

Adpressed Limbs. In both types of

chinensis, the tips of the digits touch when

the limbs are adpressed, but Cai's

diagnosis states that in yiwuensis the tips

usually do not touch. However, in Cai's

specimens there is variation in this character

and in about one third of them the tips of

the digits do meet

Costal Grooves. In his diagnosis, Cai

stated that yiwuensis has 10 costal grooves,

in contrast to chinensis which has 1 1 . We

have confirmed the presence of 1 1 in the

types of chinensis. However, in Cai's type

series there are a few individuals having 1 1

costal grooves, as Cai himself even noted

(1985, p. 110). In a series of ten

specimens of chinensis from Wenling,

about 160 km southeast of Cai's type

Vol. 3, p. 40

Asiatic Herpetological Research

April 1990

TABLE 1 . Comparison of the genera of hynobiid salamanders. See text for further details.

Bairachuperus* Hynobius Ltua

Onychodactylus Pachy hynobius* Pachypalaminus* Ranodon* Salamandrella Saiobtus

'Batrachuperus includes Paradaclylodon, Pachyhynobius includes Xenobius. and Randon includes Pseudohynobius. According to some authorities Pachypataminus is a synonym of

Hynobms (Nishio el al.. 1987).

locality, the number of costal grooves

varied from 10 to 12 (Chang, 1933).

Tail Proportions. Cai noted that the tail of

yiwuensis is compressed, with fin folds

especially distinct in the males, whereas

Gunther stated that chinensis was without a

tail crest. Gunther did not mention the

sexes of his two specimens, which are both

females (see section 1), and this was not

known to Cai. Thus, we believe that the

differences in the tail may be due to sexual

dimorphism and, in addition, perhaps also

to further elaboration of the crests during

the breeding season.

Color Pattern. Giinther's description was

based on specimens that had been

preserved for some time. Moreover, Cai

noted that the color pattern changes during

the breeding season, when it becomes

lighter and mostly green in color.

Nevertheless, our comparisons of the two

forms do not reveal any striking

differences.

Further details of our comparison

between the types of chinensis and Cai's

specimens of yiwuensis are given in Zhao

and Adler (1989). In summary, we believe

that all of these specimens represent a

single taxon which, because of the priority

of Giinther's name, must be designated

Hynobius chinensis.

3. Generic Status of

Hynobius retardatus Dunn, 1923

The identity of this Japanese species was

first recognized by E. R. Dunn, who

briefly characterized it (Dunn, 1923a) and

soon described it in more detail (Dunn,

1923b). The animal itself had been known

to Japanese biologists since at least 1907

under the names Hynobius fuscus, H.

lichenatus, and H. nigrescens. These

names, however, are now known to be

properly applied to species found elsewhere

April 1990

Asiatic Herpetological Research

Vol. 3, p. 41

in Japan (fuscus is a synonym of

nigrescens). Dunn (1923a) noted that H.

retardatus was a "well-marked species"

and, as knowledge of this animal increased

in succeeding years, its distinctiveness

from other Hynobius and hynobiids

generally became more apparent.

In 1932, Makino reported that H.

retardatus has a diploid somatic

chromosome number of 40, but since the

comparable numbers for most other

hynobiids were not known at that time, the

full significance of this very low number

was not recognized. In fact, retardatus has

by far the lowest number of chromosomes

in the family Hynobiidae (Table 1). Only

in 1943, with Sato's magnificent review of

Japanese salamanders, in which he

included a special comparative study of

their chromosomes, could the matter be

properly evaluated, and Sato himself

(1943, p. 489) suggested that retardatus

might be worthy of generic rank.

Unfortunately, Sato's premature death in

August 1945, during the atomic bombing

of Hiroshima (for a biography of Sato, see

Adler, 1989), prevented him from pursuing

this matter. In subsequent years, additional

data on anatomy, karyology, and biology

have accumulated which further support the

separation of retardatus into a new genus,

which we name:

Satobius, new genus

Type Species: Hynobius retardatus

Dunn, 1923a.

Content: A single species.

Diagnosis: A genus of hynobiid

salamanders (family Hynobiidae)

characterized by very long limbs and toes

(tips of digits of adpressed limbs overlap

+ 1 to +4 intercostal spaces in adults); a

very long tail (in adults, 100 to 118% of

head and body length combined); a long

neck and small head; no premaxillary

fontanelle or basi branchial radii; two short

series of vomerine teeth arranged in

transverse arcs between the internal nares;

vomer sutured to anterior end of

parasphenoid; lungs present; diploid (2«)

chromosome number of 40; larval duration

of one year or more (neoteny sometimes

occurs); and both terrestrial and aquatic

habits in adults during non-breeding

season.

These characteristics are discussed

below and, for easy access, are tabulated

for each of the nine genera of hynobiid

salamanders (Table 1).

Costal Grooves. These vertical grooves on

the side of the body correspond to the

position of ribs and, thus, to trunk

vertebrae; generally, the number of costal

grooves that can be counted is one less than

the number of trunk vertebrae.

The typical number of costal grooves in

5. retardatus is 11. In the Japanese species

of Hynobius the modal numbers of grooves

range from 11 to 13 (Misawa, 1989) and

some mainland species have as many as 14

(Dunn, 1923b). For species in other

hynobiid genera, the modal numbers of

costal grooves range from 10 to 14.

Adpressed Limbs. As a relative measure of

limb length, the minimum distance between

the tips of the digits is determined with the

limbs adpressed along the sides of the

body. Distances are then measured in

intercostal spaces, the fleshy folds between

adjacent costal grooves.

In metamorphosed adults of 5 .

retardatus the intercostal distance between

the digits of adpressed limbs is +1 to +4;

that is, because the limbs and toes are very

long, the digits actually overlap by one to

four intercostal spaces. Proportionately,

these are the longest limbs found in any

member of the entire family. In species of

Hynobius, this measurement ranges from 5

to +3; among hynobiids other than

Hynobius and Satobius, the longest limbs

are found in Pachypalaminus (-2 to 0), but

the toes are shorter than those of Satobius.

Tail. The tail of 5. retardatus, as measured

from posterior angle of the vent, is longer

than the combined measurements of head

plus body length. In adults, tail length

varies from 100 to 118% of head-body

Vol. 3, p. 42

Asiatic Herpetological Research

April 1990

length, whereas in all other hynobiids it is

shorter except in Ranodon (about 100-

120% of head-body length) and in

Onychodactylus (about 100-115%). The

tail lengths of Hynobius species are

significantly shorter than the head-body

length (60-80%); only in the adults of H.

nigrescens does the tail occasionally equal

head-body in length. Apparently in all

hynobiids, tail length relative to head-body

increases with overall size, so the numbers

given here are all taken, for comparative

purposes, from large adults.

Vomer and Vomerine Teeth. The paired

vomer bones (or prevomers for those who

deny homology to the vomer of mammals)

of the palate bear teeth near their posterior

edges. The relationship between the

vomers and the parasphenoids lying

posterior to them varies among hynobiid

genera (Table 1). In S. retardatus, the

posterior edge of the vomer is sutured to

the parasphenoid and overlaps very little, if

at all in some specimens, onto the palatal

surface of the parasphenoid (Inukai, 1932;

Sato, 1943). The two series of vomerine

teeth extend broadly between the internal

nares in two slight arcs which nearly meet

at the midline; the overall length of this

patch of teeth along the midline is about

30% of the width of the series.

The general pattern of these teeth is like

those in the genera Onychodactylus and

Ranodon, and quite unlike that in Hynobius

where the vomers (and the vomerine teeth

on their posterior surface) extend onto the

parasphenoids to a degree varying from

species to species. In some Hynobius this

overlap is small (e.g., H. leechii and

lichenatus) but in most it extends onto the

palatal surface for from one-third to as

much as one-half the length of the

parasphenoids (e.g., H. formosanus and

sonani) (Sato, 1943). Since the vomerine

teeth are located on this edge of the vomer,

in these Hynobius the vomerine series has a

pattern wholly unlike that in retardatus,

beginning at the nares and extending far

posteriorly on the palate in the shape of a

lyre (see Table 1). In H. formosanus, for

example, the length of the vomerine teeth

along the midline of the palate is fully twice

the total width of the two series (versus

30% in 5. retardatus).

Chromosomes. Makino (1932) was the

first to report that S. retardatus has a

diploid chromosome number of 40, as

confirmed by others (Azumi and Sasaki,

1971; Morescalchi, 1975). No other

hynobiid is known to have fewer than 56.

In Japanese species of Hynobius, the so-

called pond-type species are 2n=5 6

(retardatus is a pond breeder) and the

mountain brook types are 2/i=58 (except H.

okiensis where 2n=56); H. kimurai is

ordinarily 2n=56, except in one population

(2n=60) which may represent a separate

species (Morescalchi, 1975; Morescalchi et

al., 1979; Ikebe et al., 1989). The Korean

H. leechii is 2n=56 (Makino, 1934), but

the chromosome number is not known for

any of the Chinese species of Hynobius.

Recent studies by Japanese and Italian

workers show that the karyology of S.

retardatus is even more different from that

of Hynobius species than the low number

of chromosomes would suggest. The

combined lengths of the chromosomes in

the genomes of Hynobius species and in

retardatus are nearly equal at the same

degree of condensation, and the amount of

nuclear DNA is also approximately equal

(Morescalchi, 1975). Despite these

similarities, there are important differences

that set retardatus apart from species of

Hynobius. Kuro-o et al. (1987, as

modified in 1989) were able to compare

chromosome pairs in four Hynobius (3

Japanese and 1 Korean species) and

retardatus, using the R-banding technique

(RBG method), allowing them to identify

18 of 28 pairs in Hynobius and 16 of 20 in

retardatus. Based on this analysis,

chromosome pairs 2 and 8 of retardatus are

not at all represented in the genomes of

these Hynobius species, whereas pairs 2,

12, 20, and 22, found in all four

Hynobius, were lacking altogether in that

of retardatus; other chromosome pairs were

completely (11 pairs) or partially (3 pairs)

homeologous. To summarize, among these

four Hynobius species homeologies

totalled about 90%, but this value fell to

65% when retardatus was included. DNA

April 1990

Asiatic Herpetological Research

Vol. 3, p. 43

analysis was also performed by these

authors, using highly-repetitive DNAs as

probes on Southern blot hybridization,

which showed that retardatus was

distinctively different from the five

Hynobius species tested (4 Japanese and 1

Korean).

Breeding Biology. The reproductive

biology of 5. retardatus has been studied in

detail (Sasaki, 1924; Makino, 1933; Sato,

1989 and references cited therein). The

larval period is normally less than a year,

but at high elevations can take more than

one year. Neotenous individuals are well

known (Sasaki, 1924). Adults breed in

ponds and are terrestrial during non-

reproductive periods. However, unlike

species of Hynobius which remain on land

while not breeding, retardatus often visits

the water during non-breeding periods

(Sasaki, 1924).

Distribution. Satobius retardatus is found

only in Hokkaido, the northernmost of the

main Japanese islands. No Hynobius are

known from Hokkaido and, among

hynobiids, only Salamandrella keyserlingii

is found there.

Relationships. Historically, 5. retardatus

has been said to be most closely related to

H. nigrescens of neighboring Honshu

Island, Japan (Sato, 1943). Indeed, the

two species are superficially similar in that,

as adults, both are blackish in color with

little pattern and both possess relatively

long tails. However, on closer

examination, these two species are

fundamentally different. H. nigrescens is a

large-headed and short-necked species, as

are all Hynobius, when compared with S.

retardatus. The skull of nigrescens, in

particular the vomer-parasphenoid

relationship and the pattern of vomerine

teeth, are typical of Hynobius and unlike

the situation in Satobius. Furthermore, the

number of chromosomes and details of

chromosome structure, based on both C-

banding and R-banding methods, are

distinctly different. In short, we believe

that the similarities between nigrescens and

retardatus are due to convergence and do

not reflect close phylogenetic relationship.

Several of the diagnostic characteristics

of S. retardatus are similar to those of

members of the R a no don group of

hynobiids, which includes Batrachuperus,

Liua, Onychodactylus, and Ranodon (Zhao

and Hu, 1984). The suturing of the vomer

and parasphenoid, the pattern of vomerine

teeth, and the relative lengths of limbs and

tail in retardatus are similar to the condition

in one or more members of the Ranodon

group, but in most other respects Satobius

is clearly a member of the Hynobius group

of genera: Hynobius, Pachypalaminus

(synonymized with Hynobius by Nishio et

al., 1987) and Salamandrella (for

discussion of these two groups of genera,

see Zhao and Hu, 1984). Within this

group, the closest living relative of

retardatus may be H. leechii of Korea and

northeastern China, according to their

chromosome structure; indeed, as Ikebe et

al. (1989) point out, based on C-banding

patterns, retardatus is more similar to the

northernmost populations of leechii rather

than those in the southern part of the

Korean peninsula.

We suggest that the ancestral stock

leading to S. retardatus was derived from a

Hynobius-likc ancestor and arrived in

Japan from the mainland very early, before

the formation of the Tsugaru Strait which

later isolated Hokkaido from the southern

Japanese islands. Satobius differentiated in

Japan but was later excluded from the more

southern Japanese islands by a later

invasion of hynobiids from the mainland.

Thus isolated by the Tsugaru barrier,

Satobius further differentiated, yet retained

some of its primitive characters, including a

few today found only in the Ranodon

group.

The long isolation from Hynobius has

led also to the chromosomal differentiation

described earlier (although retaining some

similarities to H. leechii on the mainland).

The great reduction in chromosome number

in retardatus may be due in part to fusions,

since the total composite lengths and DNA

content are approximately the same as for

the Hynobius genome (Kuro-o et al.,

1989). However, in view of the many

differences based on the C-banding and R-

Vol. 3, p. 44

Asiatic Herpetological Research

April 1990

banding studies mentioned above, the

karyotype of retardatus cannot be explained

by a simple mechanism of fusions alone.

Etymology. We take great pleasure in

naming this new genus for Ikio Sato (1902-

1945), who first recognized its

distinctiveness from Hynobius. The name

Satobius is derived in part from the stem of

Hynobius (hynis [or hynnis], Greek for

plowshare, the cutting part of a plow, and

bios, life).

Acknowledgments

Our research has been supported by the

USA National Academy of Sciences

(Committee on Scholarly Exchange with

the People's Republic of China). We thank

Barry Clarke (British Museum [Nat. Hist.])

and Robert C. Drewes and Jens V. Vindum

(California Academy of Sciences) for the

loan of specimens. We thank Satoshi

Amagai (Cornell University) for some

translations of articles in Japanese.

Literature Cited

ADLER, K. (ed.) 1989. Contributions to the

History of Herpetology. Society for the Study

of Amphibians and Reptiles, Contributions to

Herpetology 5:1-202.

AZUMI, J. AND M. SASAKI. 1971. Karyotypes of

Hynobius retardatus Dunn and Hynobius

nigrescens Stejneger. Chromosome Information

Service 12:31-32.

BORING, A. M. AND T.-K. CHANG. 1933. The

distribution of the Amphibia of Chekiang

province. Peking Natural History Bulletin

8:63-74.

CAI, C.-M. 1985. [A survey of tailed amphibians

of Zhejiang, with description of a new species

of Hynobius]. Acta Herpetologica Sinica

4:109-114. (In Chinese).

CAI, M.-Z., J. ZHANG, AND D.-J. LIN. 1985.

[Preliminary observation on the embryonic

development of Hynobius chinensis Guenther].

Acta Herpetologica Sinica 4:177-180. (In

Chinese).

CHANG, T.-K. 1933. Two new amphibian records

from Chekiang. Peking Natural History

Bulletin 8:75-80.

DUNN, E. R. 1923a. New species of Hynobius

from Japan. Proceedings of the California

Academy of Sciences, ser. 4, 12:27-29.

DUNN, E. R. 1923b. The salamanders of the

family Hynobiidae. Proceedings of the

American Academy of Arts and Sciences

58:445-523.

GONTHER, A. 1889. Third contribution to our

knowledge of reptiles and fishes from the Upper

Yangtsze-Kiang. Annals and Magazine of

Natural History, ser. 6, 4:218-229.

IKEBE, C, M. KURO-O, T. OHHASHI, H.

YAMADA, K.-I. AOKI, AND S.-I. KOHNO.

1989. Evolution of chromosome 10 in ten

pond-type Hynobius from Korea and Japan, with

comments on phylogenetic relationships. Pp.

85-90. In M. Matsui et al. (eds.), Current

Herpetology in East Asia. Herpetological

Society of Japan, Kyoto.

INUKAI, T. 1932. Urodelenarten aus Nordjapan

mit besonderer Beriicksichtigung der

Morphologie des Schadels. Journal of the

Faculty of Science, Hokkaido Imperial

University ser. 6, 1:191-217.

KURO-O, M., C. IKEBE, AND S. KOHNO. 1987.

Cytogenetic studies of Hynobiidae (Urodela).

VI. R-banding patterns in five pond-type

Hynobius from Korea and Japan. Cytogenetics

and Cell Genetics 44:69-75.

KURO-O, M., C. IKEBE, R. KATAKURA, Y.

IZUMISAWA, AND S.-I. KOHNO. 1989.

Analyses of phylogenetic relationships in the

genus Hynobius by means of chromosome

banding and Southern blot hybridization. Pp.

91-94. In M. Matsui et al. (eds.), Current

Herpetology in East Asia. Herpetological

Society of Japan, Kyoto.

MAKINO, S. 1932. The chromosome number in

some salamanders from northern Japan. Journal

of the Faculty of Science, Hokkaido Imperial

University ser. 6 (Zool.) 2:97-108.

MAKINO, S. 1933. [Several ecological

observations on Hynobius retardatus Dunn].

Shokubutsu oyobi Dobutsu (Botany and

Zoology), Tokyo 1:1 136-1 142. (In Japanese).

MAKINO, S. 1934. The chromosomes of

Hynobius leechii and H. nebulosus.

April 1990

Asiatic Herpetological Research

Vol. 3, p. 45

Transactions of the Sapporo Natural History

Society 13:351-354.

MIS AW A, Y. 1989. The method of counting

costal grooves. Pp. 129-134. In M. Matsui et

al. (eds.), Current Herpetology in East Asia.

Herpetogical Society of Japan, Kyoto.

MORESCALCHI, A. 1975. Chromosome

evolution in the caudate Amphibia. Pp. 339-

387. In T. Dobzhansky et al. (eds.),

Evolutionary Biology, vol. 8. Plenum Press,

New York.

MORESCALCHI, A., G. ODIERNA, AND E. OLMO.

1979. Karyology of the primitive salamanders,

family Hynobiidae. Experientia 35:1434-1436.

NISHIO, K., M. MATSUI, AND M. TASUMI.

1987. The lacrimal bone in salamanders of the

genera Hynobius and Pachypalaminus: a

reexamination of its taxonomic significance.

Monitore Zoologico Italiano, new ser. 21:307-

315.

POPE, C. H. 1931. Notes on amphibians from

Fukien, Hainan, and other parts of China.

Bulletin of the American Museum of Natural

History 61:397-611.

POPE, C. H. AND A. M. BORING. 1940. A

survey of Chinese Amphibia. Peking Natural

History Bulletin 15:13-86.

PRATT, A. E. 1892. To the Snows of Tibet

Through China. Longmans, Green, and Co.,

London. 268 pp.

SASAKI, M. 1924. On a Japanese salamander, in

Lake Kuttarush, which propagates like the

axolotl. Journal of the College of Agriculture,

Hokkaido Imperial University 15:1-36

("Hynobius lichenatus" = Satobius retardatus).

SATO, I. 1943. [A Monograph of the Tailed

Batrachians of Japan]. Nippon Shuppan-sha,

Osaka, 520, 7 pages, 31 color plates. (In

Japanese).

SATO, T. 1989. Breeding environment and

spawning of a salamander, Hynobius retardatus,

at the foot of Hidaka Mountains, Hokkaido,

Japan. Pp. 292-304. In M. Matsui et al. (eds.),

Current Herpetology in East Asia.

Herpetological Society of Japan, Kyoto.

ZHAO, E.-M. AND Q.-X. HU. 1983. [Taxonomy

and evolution of Hynobiidae in western China,

with description of a new genus.] Acta Herpetol.

Sinica 2(2):29-35 (In Chinese). (English

translation published in Zhao et al., 1988,

Studies on Chinese Salamanders. Society for

the Study of Amphibians and Reptiles,

Contributions to Herpetology 4: ix, 1-67).

ZHAO, E.-M. AND Q.-X. HU. 1984. [Studies on

Chinese Tailed Amphibians.] Sichuan Sci.

Tech. Publ. House, Chengdu, (3), 68 pages (In

Chinese). (English translation published in

Zhao et al., 1988, Studies on Chinese

Salamanders. Society for the Study of

Amphibians and Reptiles, Contributions to

Herpetology 4: ix, 1-67).

ZHAO, E.-M. AND K. ADLER. 1989. [Hynobius

chinensis, a re-description on its 100th

anniversary]. Sichuan Journal of Zoology

8(2): 18-20, 1 plate. (In Chinese).

April 1990

Asiatic Herpetological Research

Vol.3, pp. 46-51

Relationships Between Serum T4, T3, Cortisol and the Metabolism of

Chemical Energy Sources in the Cobra During Pre-hibernation, Hibernation

and Post-hibernation

RUIMIN WU1 AND JlE HUANG1

^Department of Biology, Fujian Medical College, Fujian, China

Abstract. -This study analyzed the variation of the serum thyroxine (T4), triiodothyronine (T3), and

Cortisol in the cobra (Naja naja Linnaeus) in pre-hibernation, hibernation, and post-hibernation. The

variations were compared with the changes of indexes of cobra metabolism in those three periods, such as

oxygen consumption, serum glucose, serum triglyceride, glycogen content of liver, and triglyceride content

of fat bodies. From pre-hibernation to hibernation, the change of metabolic indexes of the cobra indicated

that the metabolic rate tends to decline because of the very low levels of the three serum hormones in pre-

hibernation and falling temperature. Low temperature during hibernation limited the activities of the three

serum hormones which were rising at higher levels during hibernation. From hibernation to post

hibernation, the rising temperature allowed the three serum hormones to increase markedly, stimulating

metabolism gradually. Therefore the metabolic rate of the cobra during post-hibernation tended to rise

again. The metabolic indexes of the cobra in post-hibernation showed a significant increase in metabolic

rate, which had close relation with the high levels of the three serum hormones. The indexes of

metabolism of the cobra indicate that it is hepatic glycogen and not fat that hibernating cobras use as their

main energy source.

Key Words: Reptilia, Serpentes, Elapidae, Naja naja, cobra, thyroxine, triiodothyronine, Cortisol,

metabolism, energy metabolites, chemical energy sources.

Introduction

Hibernation is an adaptive strategy of

ectotherms for keeping out of the cold

during winter. The metabolic rate of the

hibernating reptiles is apparently different

from that of active ones, as a result of

physiological adaptation. The endocrine

system and its close relation to metabolism

may play an important role in physiological

adaptation. Maher (1965) pointed out that

the maximum metabolic response in lizards

to thyroxine occurred at 3°C; Wilhoft

(1966) found that injections of thyroxine in

lizards increased their metabolism.

Musucchia (1984) suggested that the

injection of Cortisol into hamsters during

hibernation or hypothermia could achieve

the same result as the injection of glucose,

which might prolong the survival of

hamsters under the same conditions. We

consider that the metabolism of energy

metabolites in hibernating snakes must have

a close relation to the control of serum

thyroxine (T4), triiodothyronine (T3), and

Cortisol. This study may provide a way to

find the mechanism of snake hibernation

and guidance for maintaining snakes.

Methods

This study was carried out from

September 1987 to May 1988. Cobras

Naja naja (Linnaeus) used as experimental

animals were captured at Changlou, Fujian

Province, China. We studied serum T4, T3,

and Cortisol, and related indexes of

metabolism in cobras during pre-

hibernation, hibernation, and post-

hibernation.

Cobra groups: 12 adult cobras (6 males, 6

females) were captured in May 1987,

weighing 176-273 g. These cobras had

been fed in our college snake garden for 4

months before the experiment. On

September 9, 1987, 6 cobras were taken

out of the snake garden and tested as a pre-

hibernation group (Pre G). On October 2,

1987, the remaining 6 cobras were

transferred from the snake garden into a

dark cement pool with a perforated metal

window above. The temperature in the

pool was similar to the atmosphere. On

1990 by Asiatic Herpetological Research

April 1990

Asiatic Herpetological Research

Vol. 3, p. 47

February 10, 1988, 3 cobras were taken

out from the pool and tested as a

hibernation group (H G), and on May 7,

the last three cobras were tested as a post-

hibernation group (Post G).

Determination of oxygen consumption:

The oxygen consumption of each group

was determined using the method of Dong

et al. (1986) corrected by the authors.

Heart beats and respiratory state: Heart rate

and respiratory state of each group were

measured by routine methods.

Analysis of three hormones: Blood was

drawn from the posterior vena cava.

Serum T4, T3, and Cortisol of each group

were analyzed by radioimmunoassay by

means of kits produced by the Institute of

Shanghai Biologicals. Each sample was

analyzed twice.

Analysis of serum energy metabolites:

Serum glucose and serum triglycerides of

each group of snakes were analyzed by

clinical chemistry methods, and each

sample was analyzed twice.

Analysis of hepatic glycogen: Fresh snake

liver of each group was weighed and cut

into pieces. After having been in boiling

water for 5 minutes, the liver pieces were

homogenized. The homogenate was in

boiling water for another twenty minutes

and filtered immediately. The filtered

homogenate was mixed with 95% alcohol -

A. R. (mean analysis reagent) to twice the

volume and kept at room temperature for

ten minutes before it was centrifuged for

twenty minutes at 3000 rpm. The

supernate was drawn out, the sediment

(glycogen) soluble in hot water was

estimated by the following calculation:

Hepatic glycogen (g/Kg(BW)) = total glycogen

content in liver/body weight (BW)

Analysis of fat body triglycerides: Fresh

snake fat bodies of each group was

weighed and a small weighed part was

homogenized in a fixed volume of n-

heptane (C7H16). The fat tissue was

extracted in this way three times.

Triglycerides of fat bodies extracted in n-

heptane were analyzed by clinical chemical

methods.

Triglycerides of fat body(g/Kg(BW)) = triglyceride

content in fat body/body weight (BW)

Results

1. Variations of the three serum hormones

during Pre G, HG and Post G

The levels of serum T4, T3, and Cortisol

of Pre G were the lowest in the three cobra

groups. The level of serum Cortisol of HG

was markedly lower than that of post G.

The levels of T4 and T3 of HG were not

distincdy different from those of Post G

(Table 1).

2. Variations of the oxygen consumption

and energy metabolites during Pre G, HG

and Post G

The oxygen consumption of the HG

group was significantly the lowest of the

three cobra groups. The oxygen

consumption of Pre G was not significantly

different from that of Post G.

The contents of serum glucose of both

TABLE 1. Experimental results of the three

serum hormones and t tests.

* Too small to measure. **Significant at 0.05.

Vol. 3, p. 48

Asiatic Herpetological Research

April 1990

TABLE 2. Experimental results of oxygen consumption by energy substances, and t tests.

BW = Body Weight. **P < 0.05

Pre G and Post G were apparently higher

than that of HG. But the content of serum

glucose of Pre G was not significantly

different from that of Post G.

The hepatic glycogen content of the

cobras appeared to decline during

hibernation. The hepatic glycogen content

of HG was markedly less than that of Pre

G, and the hepatic glycogen of Post G was

less than that of HG.

Similar to the variation of serum

glucose, the content of serum triglycerides

of Pre G was higher than that of HG, but

there were not significant differences

between the serum triglycerides of post G

and that of HG, and between that of Pre G

and that of Post G. The triglyceride

contents of fat bodies in the three cobra

groups did not have any statistical

differences from one another (Table 2).

Discussion

1. Variations of the oxygen consumption

and energy metabolites in the cobras during

the three periods

The contents of serum glucose and

triglycerides of Pre G were relatively high

due to the cobras active intake before

entering hibernation, which was

advantageous for the storage of energy

metabolites for overwintering. The oxygen

consumption was large during this period,

but the contents of glucose and triglycerides

of HG were apparently at low level.

Oxygen consumption decreased markedly

in this period. The hepatic glycogen

content during hibernation became 99%

less than during pre-hibernation. The

triglyceride content of fat bodies of the HG

group seemed to rise a little, perhaps due to

some triglyceride synthesis at the beginning

of hibernation.

These results indicate that the metabolic

rate of cobras during hibernation remains at

a low level and oxygen consumption,

serum glucose, and serum triglyceride

content fell markedly, and that hepatic

glycogen is used as the main energy source

instead of fat body triglycerides. In post-

hibernation, the serum glucose content rose

significantly and the content of hepatic

glycogen decreased by 92 percent less than

that of HG, showing that cobras had used

hepatic glycogen to the greatest degree for

evoking their activities. This variation

provided a good situation for cobras to

come out from hibernation, and at the same

time, oxygen consumption increased

gready. The rising serum triglycerides of

post G also meant that the cobras' fatty

metabolism began to become active.

April 1990

Asiatic Herpetological Research

Vol. 3, p. 49

Compared with that of pre G, the

oxygen consumption of post G was

relatively large, but the serum glucose and

serum triglyceride content of post G were

relatively low. The reasons may be: 1) that

the pre G cobras were not fasting so that

their contents of serum glucose and

triglycerides were relatively high, 2) that

the post G cobras had so little hepatic

glycogen as to be unable to raise their

serum glucose and triglyceride contents as

high as those of the Pre G group, and 3)

that the rising temperature and the high

serum hormone level after hibernation

accelerated the organs and tissues to take in

glucose and triglycerides from the blood.

Among the three reasons the latter was the

most important because of the apparent rise

in oxygen consumption. This amount

indicated that the metabolism in the organs

and tissues had been enhanced markedly.

2. Variation of the three serum hormones

Serum T3 of HG was 56.6% higher than

that of Pre G, and serum T3 of Post G

51.7% higher than that of HG. Similar to

serum T3, serum T4 of Pre G was too low

to be tested, but T3 of HG rose

significantly, and T3 of Post G continued to

rise 43.2% more than that of HG. As a

result, serum T3 and T4 were increasing

steadily during hibernation. This partem of

serum T4 was similar to the results reported

by Nauleau, et al. (1987).

A comparison between the variations of

serum T3 and serum T4 can provide some

important information about the secretive

state of the thyroid gland. Before

hibernation, serum T4 was very low, and

serum T3 was higher, being in a dominant

position, but in hibernation, serum T4 was

twice as great as serum T3, though both of

them had increased. These variations were

caused by the increased activity of the

thyroid gland during hibernation. Turner

and Bagnara (1976) suggested that the

activity of the thyroid gland of ectotherms

was low both in summer and during pre-

hibernation. It was enhanced during

hibernation, reaching a peak during post-

hibernation. The thyroid gland mainly

secretes T4, which is the precursor of T3.

The secreted T4 is converted into T3 in the

blood or in the tissues and organs. The

function of stimulating metabolism by T3 is

at least three times stronger than that by T4.

The activity of the cobra thyroid gland was

at a low level in pre-hibernation. The

synthesized and secreted T4 was very low

and a portion of T4 was converted into T3

thus the concentration of serum T4 was at a

low level and the level of T3 was relatively

high. After the cobra entered hibernation,

the secretive function of the thyroid gland

began to become active and produce more

and more T4, therefore serum T4 dominated

at a higher level though both T4 and T3

increased in the blood. The variation of

serum T4 and T3 in post G was the same as

in HG. The very high levels of serum T4

and T3 in post G were of great advantage to

the cobra enhancing its metabolic rate for

arising from hibernation. Besides these,

the results in Table 1 and Table 2 also

indicate that the levels of serum T4, and T3,

of HG were higher than that of Pre G, but

the amount of oxygen consumption of HG

was less than that of Pre G. This seemed

to be self-contradictory because both T4,

and T3, were able to stimulate metabolism.

These phenomena resulted in the falling

temperature, inhibiting the function of

hormones. Wilhoft (1966) pointed out that

the function of T4 stimulating metabolism

appeared to be inactive in many reptiles at

low body temperature, and to be active only

at higher temperature. Stimulation of

metabolism was carried out in such a way

that T4 and T3 were capable of inducing the

synthesis of aerobic metabolic enzymes.

The low levels of serum T4 and T3 in pre-

hibernation indicated that the metabolic rate

of cobras in this period tends to fall.

The metabolic enzymes which had been

synthesized were still in an active state

because of the higher temperature, thus the

pre G cobras had a high oxygen

consumption. The metabolic rate of HG,

though serum T4 and T3 levels had risen,

was at a low level due to the low

temperature during hibernation which

inhibited the function of T3 (T4) and the

activity of metabolic enzymes synthesized

before. The cobras of HG had a small

oxygen consumption. The very high levels

Vol. 3, p. 50

Asiatic Herpetological Research

April 1990

of serum T4 and T3 gradually performed the

function of stimulating metabolism in the

phase at the end of hibernation as the

temperature was rising.

Serum Cortisol was increasing

continuously during hibernation. Serum

Cortisol of HG was 2.8 times higher than

that of pre G, and that of post G which

continued to rise by 3.6 times more than

that of HG. Musacchia (1984) reported

that the survival of animals during

hibernation or hypothermia had a close

relation to the level of serum Cortisol in

their bodies. An injection of Cortisol into

hamsters during hibernation or hypothermia

could yield the same result as an injection

of glucose, and might increase the survival

of hamsters under the same conditions. He

also reported that glucocorticosteroids

played an important role in animals arising

from hibernation or hypothermia. Some

clues perhaps could be found for reptiles

from the experimental results on hamsters,

although they were very different in other

respects.

The results in Table 2 indicated that the

hepatic glycogen content of HG was very

low. It was difficult for cobras to continue

hibernating while only using so little an

amount of hepatic glycogen as a energy

source. They must receive energy

metabolites in other ways.

Gluconeogenesis may be an important

process in which Cortisol has a close

relation. The surprarenal cortex of cobras

during hibernation became active gradually

and the level of serum Cortisol became

higher and higher, which was related to

accelerating gluconeogenesis and to the

cobra's arising from hibernation. The

variation of Cortisol showed that fat

metabolism became active after cobras were

aroused out of hibernation.

3. Relationship between the three serum

hormones and energy metabolites

In pre-hibernation, the function of T4 and

T3 in inducing synthesis of aerobic

metabolic enzymes was weak because of

the low levels of serum T4 and T3. The

metabolism of cobras in this period was

under the control of the enzymes which had

been synthesized before. The metabolic

rate of cobras tended to decrease from pre-

hibernation to hibernation as the

temperature was falling and as the amount

of enzyme reduced (synthesizing less and

degrading). During hibernation, when the

temperature was low, cobras had a low

metabolism under the control of the low-

temperature isozymes, though the levels of

serum hormones were at relatively higher.

The low serum glucose and triglyceride

contents implied that the decreased

metabolism of cobras in hibernation, and

the reduction of glycogen in the liver

indicated the metabolic rate of cobras at a

certain level. In post-hibernation, high

temperature provided a favorable factor for

hormones to perform their functions, and

T3 (T4) induced synthesis of enzymes and

Cortisol accelerated the gluconeogenesis and

fat catabolism. Serum glucose was

increasing markedly though the hepatic

glycogen was low, and quite a lot of

glucose was produced from

gluconeogenesis. Serum T4 and T3 play an

important role in accelerating somatic cells

to take in energy metabolites and in raising

metabolism.

In short, during hibernation the cobra

nearly used up all the hepatic glycogen but

consumed little fat body triglyceride. Body

weight tended to be lost significantly during

the period after the snake was aroused out

of hibernation. Perhaps fat bodies were

used up because there was little hepatic

glycogen. The high level of Cortisol

indicated the tendency for fat catabolism.

4. One putative pattern

From the above results, we consider that

perhaps the cobra has formed a seasonal

regulation of endocrine and that hibernation

is controlled by this mechanism. When the

season for hibernation comes, the contents

of serum T4, T3, and Cortisol decrease so

that the metabolic system of all organs and

tissues in cobras is controlled effectively.

The metabolic rate of cobras tends to go

down. With temperature falling and the

three serum hormones reducing, the

remaining metabolic enzymes in cobras

April 1990

Asiatic Herpetological Research

Vol. 3, p. 51

become inactive gradually (content less and

activity low) and cobras go into hibernation

keeping their metabolic rate at a fairly low

level under the control of the low

temperature isozymes, and after physical

regulation. This suggests that the external

factor for cobra hibernation is the fall of

temperature, while the internal factor is the

drop of the three serum hormone levels. In

hibernation, the contents of serum T4, T3,

and Cortisol increases gradually, but the

levels of the three serum hormones are not

high enough to accelerate metabolism of the

organs and tissues, and the low temperature

inhibits the activities of the three serum

hormones. When the levels of serum T4,

T3, and Cortisol are high enough, the

metabolism of the organs and tissues in

cobras tends to go up. With the

temperature rising, the three serum

hormones induce the synthesis of aerobic

metabolic enzymes in the organs and

tissues and accelerate the metabolism of

energy metabolites. The metabolic rate

rises and the cobra recovers from

hibernation. This suggests that the external

factor for the cobra to come out of

hibernation is the rise of temperature, while

the internal factor is the rise of the three

serum hormone levels.

Literature Cited

DONG, Y. J. ZHAO, J. YANG, Y. CUI, G. LI, AND

B. CHEN. 1986. Studies on physiological

ecology of Rana spinosa during hybernation

and activation. Acta Herpetologica Sinica 1986,

5(4):24 1-245.

MAHER, M. T. 1965. The role of the thyroid

gland in the oxygen consumption of lizards.

General Comparative Endocrinology 5(3):320-

325.

MUSACCIA, X. J. 1984. Comparative

physiological and biochemical aspects of

hypothermia as a model for hibernation.

Cryobiology 21(6):583-592.

NAULLEAU, G., F. FLEURY, AND J. BOISSIN.

1987. Annual cycles in plasma testosterone and

thyroxine in male Aspic Vipera aspis L.

(Reptilia, Viperidae), in relation to the sexual

cycle and hibernation. General Comparative

Endocrinology 65(2):254-263.

TURNER, G. AND T. J. BAGNARA. 1976. General

endocrinology, sixth ed. W.R. Co.

Philadelphia.

WILHOFT, D. C. 1966. The metabolic response to

thyroxine of lizards maintained in a thermal

gradient. General Comparative Endocrinology

7(3):445-451.

5. Conclusion

Serum T4, T3, and Cortisol at low levels

were involved in the regulation of

depressing the metabolism of cobras before

hibernation. The low temperature during

hibernation inhibited activity of the three

serum hormones, and the metabolism was

at a low level. With the rising temperature,

the levels of serum T4, T3, and Cortisol at

very high levels accelerated the metabolic

rate of cobras which resulted in arousal

from hibernation. Cobras in hibernation

used hepatic glycogen as their main energy

source instead of fat body triglycerides.

I April 1990

Asiatic Herpetological Research

Vol. 3, pp. 52-53

Intergradation Between Melanochelys trijuga trijuga and M. t. coronata

(Testudines: Emydidae: Batagurinae)

INDRANEIL DAS1 AND PETER C. H. PRITCHARD2

1 Animal Ecology Research Group, Department of Zoology, Oxford University, Oxford 0X1 3PS, England

^Florida Audubon Society, 1101 Audubon Way, Maitland, Florida 32751, USA

Key words: Reptilia, Testudines, Emydidae, Melanochelys, India, distribution, intergrade.

The Indian Black Turtle or pe amai,

Melanochelys trijuga, is one of the most

abundant chelonians in the Indian

subcontinent, with a distribution extending

from Sri Lanka (Deraniyagala 1939),

through India and Burma, to western

Thailand (Wirot 1979), although apparently

excluding Bangladesh (Das, 1989).

Nevertheless, details of the distribution of

the seven described subspecies remain

obscure. The range maps provided by

Smith (1931), Das (1985), and Tikader and

Sharma (1985) indicate rather clearly

allopatric distributions for the Indian

subspecies, but the overall range of the

species is based on extremely few and

widely-separated locality points (except for

Sri Lanka and Kerala), as is evident in the

map provided by Iverson (1986).

Deraniyagala (1939) recognized two

subspecies in Sri Lanka, but the geographic

separation, if any, between these two forms

was not made clear.

Of the various subspecies, the most

distinctive is probably Melanochelys trijuga

coronata, whose distribution is restricted to

the state of Kerala in southwestern India

(the distribution map provided by Tikader

and Sharma (1985) involves a transposition

of the range of coronata and trijuga ). M.

t. coronata has a striking head pattern, with

a broad, black diamond-shaped marking on

the crown of the head, contrasting with the

white to yellow temporal region. The head

pattern of the other subspecies consists at

most of small, yellow to pink spots and

reticulations that may disappear with age.

The shell as a whole is usually unrelieved

black, in contrast to other subspecies in

which at least a lighter plastral rim is

evident (although Deraniyagala (1939)

reported completely black specimens of M.

t . trijuga from Kalpitiya, Sri Lanka and

we have seen increased pigmentation with

age in the north Indian subspecies,

indopeninsularis, in which adult animals

may lose the lighter plastral rim). The

distribution of M. t. coronata is generally

indicated as widely separated from that of

M. t. trijuga by the Western Ghats, but the

southward penetration of the forma typica

into the hiatus between these two

subspecies is not represented on existing

range maps, and the subspecific

relationship between these two forms has

been assumed rather than demonstrated.

We here report upon four specimens that

bridge the geographic map between M. t.

trijuga and M. t. coronata, and three that

show characters intergradient between

them.

An adult male (CL 23.9 cm; CW

16.4 cm) was collected by the two authors

from a dry stream bed in Chichli, Indira

Ghandi (formerly Annamalai) Wildlife

Sanctuary, Coimbatore District, Tamil

Nadu, on March 27, 1989 (Fig. 1). The

right anterior margin of the shell showed

signs of an old injury, with several

peripheral bones lost. The head pattern of

the specimen included the diamond- shaped

marking typical of Melanochelys trijuga

coronata. The yellow head reticulations

and the yellow plastral margin is suggestive

of M. t. trijuga, and the size of the

specimen is greater than that of any

recorded specimen of M. t. coronata

(maximum 17.5 cm (Smith 1931) or

18 cm (Tikader and Sharma 1985) of

20.8 cm (Das 1985)). The specimen was

retained alive and housed at the Madras

Crocodile Bank.

1990 bv Asiatic ileroetoloeical Research

April 1990

Asiatic Herpetological Research

Vol. 3, p. 53

FIG. 1. Intergrade between Melanochelys trijuga

coronata and M. t. trijuga.

A second specimen — a fragmentary

shell only — was also found in the Indira

Ghandi Wildlife Sanctuary, in a tribal

settlement. The nuchal area and anterior

peripherals are missing. Shell width is

13.7 cm, again indicating a specimen larger

than is typical of Melanochelys trijuga

coronata. Ankylosis of the shell, as

illustrated for Sri Lanka specimens by

Deraniyagala (1939), is essentially

complete, except for faint indications of the

peripheral sutures. The specimen is

registered as PCHP 2803 in the private

collection of the second author.

In addition to these, two intergradient

specimens were found in the collection of

the Southern Regional Station of the

Zoological Survey of India (Madras), both

collected by M. Vasant and party. An

8.1 cm juvenile (Lot no. 10) was collected

at Kombiar Charagam, Kalakkadu Wildlife

Sanctuary, Turunelveli District, Tamil

Nadu (altitude 210 m), on October 13,

1987. A 20.6 cm adult male (Lot no. 6)

was collected at Nambiar, Nambi Kovil

Road, also within Kalakkadu Wildlife

Sanctuary (altitude 140 m), on January 11,

1987.

Melanochelys trijuga is basically a pond

turtle, and the Western Ghats appear to

separate the essentially lowland ranges of

M. t. coronata and M. t. trijuga

Nevertheless, the new localities all fall

within upland areas actually in the Western

Ghats. It appears that the hills may form a

sufficient barrier to allow the evolution of

different subspecies east and west of the

Ghats, but the ability of M. trijuga to exist

far from water suggests that individuals

wandering into these hills may provide the

stock for an intergradient population. It

would be worthwhile to search for a

lowland intergradient population in the

coastal area at the northern extreme of the

distribution of M. t. coronata.

Other cases exist of congeneric

batagurine species of a similar ecological

role differing primarily in head pattern, as

is the case with some of the species of

Rhinoclemmys in the Neotropics. One is

tempted to interpret the different markings

as a means of avoiding cross-matings in

situations of sympatry. But, the overall

allopatry of the species of Rhinoclemmys

in northern South America, for example,

suggests that this interpretation may be an

incomplete one, as does the discovery of

intergradation between adjacent populations

of Melanochelys that differ primarily in

cephalic pigmentation.

Literature Cited

DAS, I. 1985. Indian turtles. A field guide.

World Wildlife Fund India, Eastern Region,

Calcutta.

DAS, I. 1989. Report on a survey of freshwater

turtles and land tortoises in Bangladesh. Report

to Fauna and Flora Preservation Society, July

1989. Mimeo9pp.

DERANIYAGALA, P. E. P. 1939. The tetrapod

reptiles of Ceylon. Colombo Museum,

Colombo. 412 pp.

IVERSON, J. B. 1986. A checklist with

distribution maps of the turtles of the world.

Paust Printing, Richmond, Indiana. 283 pp.

SMITH, M. A. 1931. The fauna of British India,

including Ceylon and Burma. Reptilia and

Amphibia, Vol. 1. Loricata, Testudines.

Taylor and Francis, London. 185 pp.

TIKADER, B. K. AND R. C. SHARMA. 1985.

Handbook: Indian Testudines. Zoological

Survey of India, Calcutta.

WIROT, N. 1979. The turtles of Thailand.

Mitbhadung Press, Bankok.

[April 1990

Asiatic Herpetological Research

Vol. 3, pp. 54l9]

Thermal Sensitivity of Sprinting and Clinging Performance in the Tokay

Gecko (Gekko gecko)

JONATHAN B. LOSOS1

1 Museum of Vertebrate Zoology and Department of Integrative Biology,

University of California, Berkeley, CA 94720 USA

Abstract. -The thermal sensitivity of sprinting and clinging ability was measured in tokay geckos

(Gekko gecko). Sprinting performance was maximal at high (35-4 1°C) temperatures, as is the case for

other nocturnal lizards, but the optimal temperature for clinging was considerably lower (approximately

17°C). These different optima could be adaptive if maximal sprinting and clinging capabilities are needed at

different temperatures. Alternatively, they could result from constraints on adaptive evolution.

Key Words: Reptilia, Sauria, Gekkonidae, Gecko gecko, thermal sensitivity, optimal performance.

Introduction

Most physiological processes are

temperature-dependent. Ectothermic

animals, which do not maintain a constant

body temperature, are thus subject to

fluctuation in the rate at which they can

perform many vital tasks. By regulating

body temperature behaviorally, however,

many reptiles can maximize performance

capability (Huey 1982a). In some cases,

maximal performance temperature might

differ for different tasks. For example, a

wealth of behavioral data indicates that

many lizards and snakes increase their body

temperature after feeding, which suggests

that digestion has a higher optimal

temperature than other activities (reviewed

in Huey 1982a).

Early work centered on the thermal

dependence of sub-organismal traits (e.g.,

enzyme activity, muscle contractile speed

[Huey and Stevenson 1979]). In many

cases, however, the effect of temperature

on the functional capacities (e.g., sprint

speed, critical thermal maximum

temperature) cannot be predicted by study

at sub-organismal levels (e.g., Licht 1967;

Marsh and Bennett 1985, 1986).

Consequently, recent studies have focused

on whole-organism performance. The

thermal dependence of sprinting ability has

been studied in great detail. Maximum

sprinting speed and/or endurance in many

species occurs at the temperature they most

frequently experience in nature (Bennett

1980); however, nocturnal lizards appear

exceptional (Huey and Bennett 1987; Huey

etal. 1989). The thermal sensitivity of few

other whole-organism locomotor

performance measures has been determined

for lizards. Here I report the thermal

sensitivity of tokay geckos (Gekko gecko)

[Fig. 1] for two ecologically relevant tasks,

maximum sprinting and clinging capability.

I ask whether maximum performance

capability occurs at the relatively low

temperatures most commonly experienced

by geckos (Huey et al. 1989) and whether

the optimal temperature is the same for the

two performance measures.

Methods

Tokays are relatively large geckos found

on trees and walls throughout southeastern

Asia (Smith 1935). Lizards were captured

on Phuket Island, Phuket Province,

Thailand, and transported to the University

of California, Berkeley in late September

1987. They were maintained in 30 x 17 x 9

cm plastic shoeboxes at ambient

temperatures and provided with water and

crickets ad libitum. An ontogenetic series

was used in this study (11 individuals;

snout-vent length: 85-180 mm; mass: 10-

140 g).

All trials were conducted in a walk-in

environmental chamber in the Museum of

Vertebrate Zoology, University of

1990 by Asiatic Herpetological Research

April 1990

Asiatic Herpetological Research

Vol. 3, p. 55

FIG. 1 . Gekko gecko from Phuket Province, Thailand.

California, Berkeley, within a month of

capture. Lizards were placed in the chamber

at least one hour prior to performance

measurement. Humidity, which could not

be regulated, was determined on several

occasions using a Bacharach sling

psychrometer.

Clinging capability was measured by

placing lizards on a plexiglass plate and, at

a gradual and steady rate, lifting the end of

the plate so that the lizard was clinging with

its head directed down. The angle at which

the lizard fell from the plate was recorded

(protocol modified from Emerson and Diehl

1980; Alberch 1981). Lizards that jumped

from the plate were not included in the

analysis. As lizards began to slide, they

usually attempted to maintain their grip by

moving and re-applying their toe pads

regardless of temperature. There was no

evidence that temperature affected the

lizards' efforts to prevent sliding and falling

from the experimental plate.

One trial per lizard was conducted per

temperature. Performance at nine

temperatures (12, 16, 17, 22, 24, 31 [three

times], 34, 35, 41; the order of

temperatures is presented in figure 2) was

measured over a six-day period. On some

days, two trials were conducted.

Sprint capability was measured by

placing lizards at the end of a 2.25 m

trackway covered with a rough rubber

surface and inducing them to run by

repeated taps to the tail (protocol following

Huey 1982b; Huey et al. 1989; Garland

1985). As the lizard ran, it interrupted light

beams stationed every 0.25 m. The time

elapsed during each interval was computed

by a Compaq personal computer, the fastest

single interval during four trials, conducted

at hourly intervals, was considered the

maximum speed for that lizard at that

Vol. 3, p. 56

Asiatic Herpetological Research

April 1990

155

20 30

Temperature

1.50

.25

1.00

25 35

Temperature

FIG. 2. Clinging ability (mean + 1 s.e.) at

different temperatures. Clinging ability is

measured as the angle of the plate at which a lizard

lost its grip and fell. The points are numbered by

the order in which the trials were conducted.

temperature. Lizards that did not sprint at

maximal capability on any of the trials at a

given temperature, as judged by their gait,

were not included in the calculations for

that temperature. Lizards were tested at

five temperatures (30, 36, 19, 26, 41°C, in

that order), one temperature per day, over

an eight-day period. Trials were never held

on three consecutive days. A second trial at

30°C was held at the conclusion of the

study to determine whether a performance

decline had occurred. Animals whose

performance decreased > 30 % were

excluded from the analysis.

Results

Clinging performance is temperature-

dependent, with a peak at 17°C (Fig. 2). A

non-linear equation (In [clinging ability] =

0.99 + 2.69 * In [temp] - 0.45 *

{In [temp]}2; F28 = 5.08, P < 0.05)

better fits the data than a linear regression

(Fli9 = 2.77, P>0.10).

Although between-day variation exists in

clinging ability at a given temperature, no

general pattern of increased or decreased

performance over the duration of the study

exists. For example, three trials were

conducted at 31 °C. The second trial had

the lowest mean, whereas the last trial had

the highest mean. Three other sets of trials

were conducted at approximately the same

temperature (16-17°C, 22-24°C, 34-35°C).

In two cases, performance ability decreased

FIG. 3. Sprint speed (mean ± 1 s.e.) at different

temperatures. The numbers indicate which

represents the first and the second 30° trials.

in the second trial, but in the third case it

increased.

Sprinting capability is also temperature-

dependent (Fig. 3). Maximal performance

ability occurs at 36-4 1°C, but interpretation

of the results is difficult because

performance ability declined over the

course of the experiment, as evidenced by

the difference in sprinting ability in the two

sets of trials at 30°C. Despite this decline,

several results are clear from inspection of

Fig. 3: 1. performance at 36°C (tested 8

July) is slightly higher than at 30 (7 July);

performance at 40°C (tested 13 July, after

all trials except the second set at 30) is

nearly as high or higher than all other

temperatures; 3. performance at 30°C (16

July) is greater than performance at 26°C

(11 July), which, in turn, is greater than

performance at 19°C (10 July).

Consequently, even if performance steadily

decreased over time, it is reasonable to

conclude that maximal sprint performance

occurs around 36-4 1°C

Discussion

Sprinting and clinging are ecologically

relevant performance measures for geckos,

but their optimal performance temperatures

differ greatly for tokays. Sprint

performance is greatest at relatively high

temperatures, as is the case for a number of

other nocturnal gecko species (Huey et al.

1989) and the nocturnal skink

April 1990

Asiatic Herpetological Research

Vol. 3, p. 57

Eremiascincus fasciolatus (Huey and

Bennett 1987). Clinging capability, the

thermal dependence of which has never

previously been investigated, is maximal at

considerably lower (approx. 17°C)

temperatures.

It is difficult to envision how such

different optima could evolve adaptively.

Most lizards sprint maximally at

temperatures close to those they normally

experience (Huey 1982a; Huey et al.

1989). The low optimal temperature for

clinging matches the field temperatures of

many active nocturnal geckos (Huey et al.

1989). However, many nocturnal geckos

bask and/or are active to some extent during

the day (Bustard 1967, 1968; Werner and

Whitaker 1978; Nagy and Knight 1989).

Consequently, the high optimal temperature

for sprinting seen in many nocturnal lizards

might represent adaptation for diurnal

capability (Huey and Bennett 1987; Huey et

al. 1989). Nonethless, it seems implausible

that one aspect of locomotion, sprint

performance, should be selected at high

temperatures, whereas another important

component of effective movement,

clinging, should be favored at considerably

lower temperatures.

As an alternative explanation, the

differences in thermal optima might result

from differences in evolutionary lability of

the two performance capabilities and thus

represent constraints on adaptive evolution

in either sprinting or clinging ability. To

understand why these performance abilities

are affected differently by temperature, a

better understanding is needed of their

underlying physical and physiological

bases (e.g., Garland 1984, 1985; Marsh

and Bennett 1985, 1986). In many species

of lizards, sprint performance is maximal at

temperatures close to the critical thermal

maximum (Huey and Bennett 1987; Huey

et al. 1989). The cause for this linkage is

unclear. However, if geckos must be able

to survive diurnal temperatures (either

because they intentionally maintain high

temperatures to maximize other processes,

or because environmental conditions

preclude the maintenance of lower

temperatures), then they would have to

evolve high critical thermal maxima. The

high sprint performance maximum might be

a correlated effect of this physiological

adaptation to high temperatures and not be

adaptive per se (Huey et al. 1989).

Clinging capability depends upon both

physical and physiological processes.

Geckos cling to smooth surfaces by dry

adhesion. The subdigital lamellar pads of

geckos are covered with millions of

microscopic setal hairs. When the pads are

adpressed to a surface, these hairs form

intermolecular bonds with molecules on the

surface of the substrate (Hiller 1975). If

the surface energy (a measure of the

number of free electrons on the surface of

the substrate) is relatively high, then

enough bonds can form to support the

lizard. Because these bonds result from the

activity of electrons, the forces theoretically

should be temperature-independent over the

range of temperatures in this study.

However, geckos have an elaborate

muscular and vascular system for the

adpression and removal of their toe pads

(Russell 1975); the thermal dependence of

these muscles has not been investigated.

The poor clinging performance of geckos at

10-12°C is clearly the result of

physiological incapacitation. At that

temperature, geckos were generally

inactive, moved slowly and infrequently,

and even rarely bit or barked when

handled. In contrast to trials at higher

temperatures, the lizards did not attempt to

adjust their pads or posture when the plate

was tilted and quickly lost their hold. More

research is required to determine whether

the performance decline at temperatures

above 17-1 8°C results from decreased

capability of the muscles and enzymes

involved in clinging.

One possible confounding effect in the

clinging experiments is the variation in

absolute humidity. Moisture might

decrease the formation of high-energy

intermolecular bonds. Absolute humidity

in the environmental chamber was

measured at several temperatures and

increased from 8.5 Barrs at 11.9°C to 25.1

Barrs at 34.3°C (relative humidity,

Vol. 3, p. 58

Asiatic Herpetological Research

April 1990

however, was greatest at intermediate

temperatures). Consequently, the greater

absolute humidity at higher temperatures in

the environmental chamber might have

caused a decrease in clinging ability.

Further research is needed to investigate to

what extent humidity affects clinging.

Although different processes may often

be maximized at different temperatures,

rarely is the difference as great as is

observed between sprinting and clinging in

tokay geckos (Huey 1982a). One would

not expect these geckos to need maximum

ability at these aspects of locomotion at

such different temperatures, but neither

would one expect the physiological

sensitivity of different systems to be so

different. Interestingly, the optimal

temperature for hearing sensitivity in the

tokay gecko is intermediate between the

sprinting and clinging optima (Werner

1976). Further research is required to

understand the processes shaping and

constraining performance evolution.

Acknowledgments

I thank T. Papenfuss and R. Macey for

providing the geckos, R. Jones for

assistance with equipment, R. Huey for

advice on racing geckos, J. Herron for

providing me with a pre-publication

manuscript, and A. Bauer, H. Greene and

R. Huey for constructive comments on a

previous version of this paper.

Literature Cited

ALBERCH, P. 1981. Convergence and parallelism

in foot morphology in the neotropical

salamander genus Bolitoglossa. I. Function.

Evolution 35:84-100.

BENNETT, A. F. 1980. The thermal dependence of

lizard behavior. Animal Behavior 28:752-762.

BUSTARD, H. R. 1967. Activity cycle and

thermoregulation in the Australian gecko,

Gehyra variegata. Copeia 1967:753-758.

BUSTARD, H. R. 1968. The ecology of the

Australian gecko Heteronotia binoei in northern

New South Wales. Journal of Zoology, London

156:483-497.

EMERSON, S. B., AND D. DIEHL. 1980. Toe pad

morphology and mechanisms of sticking in

frogs. Biological Journal of the Linnaen

Society 13:199-216.

GARLAND, T., JR. 1984. Physiological correlates

of locomotory performance in a lizard: an

allometric approach. American Journal of

Physiology 247:R806-R815.

GARLAND, T., JR. 1985. Ontogenetic and

individual variation in size, shape and speed in

the Australian agamid lizard Amphibolurus

nuchalis. Journal of Zoology, London (A)

207:425-439.

HILLER, U. 1975. Comparative studies on the

functional morphology of two gekkonid lizards.

Journal of the Bombay Natural History Society

73:278-282.

HUEY, R. B. 1982a. Temperature, physiology,

and the ecology of reptiles. Pp. 25-91. In C.

Gans and F. H. Pough (Eds.), Biology of the

Reptilia, Vol. 12. Academic Press, New York,

USA.

HUEY, R. B. 1982b. Phylogenetic and

ontogenetic determinants of sprint performance

in some diurnal Kalahari lizards. Koedoe 25:43-

48.

HUEY, R. B., AND A. F. BENNETT. 1987.

Phylogenetic studies of coadaptation: preferred

temperatures versus optimal performance

temperatures of lizards. Evolution 41:1098-

1115.

HUEY, R. B., P. H. NIEWIAROWSKI, J.

KAUFMANN, AND J. C. HERRON. 1989.

Thermal biology of nocturnal ectotherms: is

sprint performance of geckos maximal at low

body temperatures? Physiological Zoology

62:488-504.

HUEY, R. B., AND R. D. STEVENSON. 1979.

Integrating thermal physiology and ecology of

ectotherms: a discussion of approaches.

American Zoologist 19:357-366.

LICHT, P. 1967. Thermal adaptation in the

enzymes of lizards in relation to preferred body

temperature. Pp. 131-145. In C. L. Prosser

(Ed.), Mechanisms of Temperature Adaptation.

American Association for the Advancement of

Science, Washington, D. C.

MARSH, R. L., AND A. F. BENNETT. 1985.

April 1990

Asiatic Herpetological Research

Vol. 3, p. 59

Thermal dependence of isotonic contractile

properties of skeletal muscle and sprint

performance of the lizard Dipsosaurus dorsalis.

Journal of Comparative Physiology B 155:541-

551.

MARSH, R. L., AND A. F. BENNETT. 1986.

Thermal dependence of contractile properties of

skeletal muscle from the lizard Sceloporus

occidentalis with comments on methods for

fitting and comparing force-velocity curves.

Journal of Experimental Biology 126:63-77.

NAGY, K. A., AND M. H. KNIGHT. 1989.

Comparative field energetics of a Kalahari skink

(Mabuya striata) and gecko (Pachydactylus

bibroni). Copeia 1989:13-17.

RUSSELL, A. P. 1975. A contribution to the

functional analysis of the foot of the Tokay,

Gekko gecko (Reptilia: Gekkonidae). Journal of

Zoology, London 176:437-476.

SMITH, M. A. 1935. The Fauna of British India,

Reptilia and Amphibia. Vol. 2 — Sauria.

Taylor and Francis, Ltd., London. 440 pp.

WERNER, Y. L. 1976. Optimal temperatures for

inner-ear performance in gekkonid lizards.

Journal of Experimental Zoology 195:319-352.

WERNER, Y. L., AND A. H. WHITAKER. 1978.

Observations and comments on the body

temperatures of some New Zealand reptiles.

New Zealand Journal of Zoology 5:375-393.

I April 1990

Asiatic Herpetological Research

Vol. 3, pp. 60-63

Mating Call Structures of the Chinese Frog, Rana nigromaculata

(Amphibia, Anura, Ranidae)

YONG MU1 AND ERMI ZHAO1

^Chengdu Institute of Biology. P.O. Box 416, Academia Sinica, Chengdu, China

Abstract. -Mating calls of the Chinese Pond Frog, Rana nigromaculata at three localities were recorded

and compared. The calls of R. nigromaculata are short in duration and consist of a few notes. Each note

has several distinct pulses. The call structures of northern and central China are different from those of

Sichuan, hence we tentatively regard them as different subspecies.

Key words: Amphibia, Anura, Ranidae, Rana nigromaculata, mating calls.

Introduction

Rana nigromaculata occurs widely in the

lower Amur and Ussuri river valleys

(USSR), Korea, Japan, and throughout

northeastern and midwestern China. In

China it occurs in all but Xinjiang, Tibet

and Guangxi autonomous regions, and

Yunnan, Taiwan, and Hainan provinces.

As far as we know, no paper has been

published in China analyzing the mating

calls of Chinese anurans.

The mating call of frogs is a useful clue

in revealing systematic and evolutionary

relationships. Vocalizations are species-

specific in anurans, and serve as isolating

mechanisms. Therefore, some taxonomic

revisions have been based primarily on call

differences [see Kuramato (1977) for

review].

Methods

Calls of Rana nigromaculata were

recorded in 1988 at three localities. Several

calls of a single male were recorded on 19

March, in the suburbs of Hongya County,

about 100 km southwest of Chengdu,

Sichuan Province. The frog was calling

from a stone near water. The water

temperature was 19°C.

A number of calls from three males were

recorded at East Lake Park, Wuchang,

Wuhan, Hubei Province on 17 June. The

frogs called while standing in shallow

water shaded by grass. The water

temperature was 22°C.

The calls of 10 males were recorded at

Wofoshi, Xiangshan Park, Beijing, on 28

May. Male frogs were found calling in

surroundings similar to those of frogs in

Hubei Province. The water temperature

was 24°C.

Temperature records were always taken

in the vicinity of calling males. Calls were

recorded with a Sony TCS-370 cassette

recorder and a Feidec TSM-91 microphone.

Sonograms were prepared with Kay 7800

and Kay 7900 sonographs. For the

analysis of mating calls, standard (2.56

sec) sonograms with a 150 Hz bandwidth

filter were used.

When a call was composed of several

groups of pulses, each group was termed a

note. The main acoustic parameters

measured were: duration of call, number of

notes and pulses, pulse repetition rate

(number of pulses per second),

fundamental frequency, and dominant

frequency.

Results and Discussion

The vocalizations of Rana nigromaculata

consist of 4-10 notes, and each note has 3-

7 distinct pulses. The higher frequency

parts (above 2 kHz) are usually absent in

the first few notes (1-3). Time intervals

between pulses tend to become longer at the

end of the call. In all localities, the frog's

frequency band ranges from 0 to 8 kHz.

© 1990 by Asiatic Herpetological Research

April 1990

Asiatic Herpetological Research

Vol. 3, p. 61

TABLE 1 . Analysis of the calls of Rana nigromaculata. The dominant frequency and pulse repetition rate

of R. nigromaculata from Hongya, Sichuan are about half of those recorded at the other localities. Note the

divergence of the Sichuan call structure from northern (Beijing) and central (Wuhan) China.

Their fundamental frequencies are

approximately identical (0.5 kHz). Since

fundamental frequencies are dependent on

the oscillations resulting from air passing

over the vocal cord, causing it to vibrate at

a frequency which is a function of the mass

and tension of the cord (Duellman and

Trueb 1986), it is likely that the frogs have

a common vocal structure.

As well as multi-note calls, Rana

nigromaculata has a single-note call.

Frogs in northern and central China

(Beijing and Wuhan) emit this sound

repeatedly and usually in groups. Each

group is composed of 5-10 single-note

calls. In Hongya, the single-note call could

rarely be heard. Because R. nigromaculata

in Sichuan is peripheral in distribution, we

believe that the call structure in northern

and central China represent the basic

structure, and that the Sichuan structure

was derived from it.

Table 1 and Figure 1 clearly demonstrate

the divergence of call structure between

northern and central China, and Sichuan.

In comparing the three main acoustic

parameters, call duration, pulse repetition

rate, and mean dominant frequency, a

general pattern is evident (Zweifel 1959).

Pulse repetition rate and dominant

frequency correlate positively with

temperature, but duration of call correlates

negatively with temperature. However, we

feel that this is not sufficient to explain the

differences between the populations. At a

temperature only 3 and 5°C lower than in

Wuhan and Beijing, respectively, the

dominant frequency and pulse repetition

rate of Sichuan Rana nigromaculata are

about half of those found in the former

localities. The length of calls of Sichuan

frogs is approximately equal to the

maximum call length of frogs from the

Wuhan and Beijing localities, and their

vocal sacs are of the same size. In

addition, they have very similar

fundamental frequencies. We conclude that

they have nearly identical sound-producing

Vol. 3, p. 62

Asiatic Herpetological Research

April 1990

TYPE B/65 SONA&Ram * KAY

TYPE B/65 SONAGBAM e KAY ELEMETRICS CO PINE BROOK. N.J.

fc fa t,- it

8

1-

p p

I f I f M

I 1 I I i t'

TYPE B/65 SONAGRAM * KAY ELEMETRICS CO PINE BROOK N.J.

JTT

W'.W-%

;h i i i

1 :

Hi'

'

in w it ftt i

i

"V I' Hil

HM 1H

MUM

f

in

r1

i

FIG. 1. Sonagrams of the calls of R ana nigromaculata in China, a. Seven note call from Beijing on

May 27, 1988. The water temperature was 24°C. b. One 6 note call and two 1 note calls from Wuchang,

Hubei Province, June 17, 1988. The water temperature was 22°C. c. Two 5 note calls from Hongya,

Sichuan Province, March 19, 1988. The water temperature was 19°C.

April 1990

Asiatic Herpetological Research

Vol. 3, p. 63

structures, but different controlling

mechanisms.

In acknowledging that our analysis of

frog calls may not be exhaustive, we can

only tentatively suggest that the Sichuan

population of Rana nigromaculata is a

separate subspecies from the Wuhan and

Beijing populations.

Acknowledgments

We would like to thank Theodore J.

Papenfuss (Museum of Vertebrate

Zoology, University of California,

Berkeley) for donating the sonograph

paper, and Mitsuru Kuramoto of the

Fukuoka University of Education, Japan.

Literature Cited

DUELLMAN, W. E. AND L. TRUEB. 1986.

Biology of Amphibians. McGraw-Hill, New

York. 670 pp.

KURAMOTO, M. 1977. Mating call structures of

the Japanese pond frogs, Rana nigromaculata

and Rana brevipoda (Amphibia, Anura,

Ranidae). Journal of Herpetology 11(3):249-

254.

ZWEIFEL, R. G. 1959. Effect of temperature on

call of the frog, Bombina variegata. Copeia

1959(4):322-327.

April 1990

Asiatic Herpetological Research

Vol. 3, pp. 64-66 1

Four Remarkable Reptiles from South China Sea Islands,

Hong Kong Territory

JAMES LAZELL1 AND WENHUA LU1

lThe Conservation Agency, Rhode Island 02835, USA

Key Words: Reptilia, Squamata, Hong Kong, distribution.

FIG. 1. Hong Kong and nearby islands. 1. Hei Ling Chau. 2. Shek Kwu Chau. 3. High West

Pokfulam, Hong Kong Island.

Introduction

There are at least one hundred vegetated

islands, presumably supporting amphibians

and/or reptiles, within the Hong Kong

Territory. This is estimated to be about

three percent of the total number of

continental shelf islands in the South China

Sea between Taiwan and Hainan Dao.

Nevertheless, the number of endemic and

isolated species documented from the

territory in the comprehensive work of

Karsen, Lau, and Bogadek (1986) is high:

four endemics and at least fifteen widely

disjunct populations. Here we add two

species not previously recorded (both

widely disjunct) and new records of species

poorly documented before (Fig. 1).

Voucher specimens are in the Museum of

Comparative Zoology (MCZ), Harvard

University.

Dibamus cf. bourreti

(White-tailed Two-footed Lizard)

A single specimen, MCZ 172041, was

April 1990

Asiatic Herpetological Research

Vol. 3, p. 65

collected by Anthony Bogadek, 1 April

1987, on Hei Ling Chau ca. 10 km

southwest of Victoria, Hong Kong — a

range extension northeast for the genus,

and family Dibamidae of ca. 800 km

(Lazell 1988).

The specimen, a 177 mm SVL male,

was sent to Allen Greer, Australian

Museum, Sydney, for identification. He

believed the specimen was "probably"

Dibamus bourreti (in litt., 9 May 1988).

However, we note the following sharp

distinctions from D. bourreti as diagnosed

and described by Greer (1985: 148): the

rostral suture is not complete, not present

from lip to nostril but only posterior to the

nostril. There is a prominent labial suture.

There are no preanal or tibial pores.

Dibamis bourreti is diagnosed as having a

complete rostral suture, no labial suture,

and four preanal pores on each side, even

in a female — the highest count in Dibamus.

This specimen, MCZ 172041, has 23

scale rows at midbody, six scales fronting

the hindlimb where Greer (1985:120)

shows D. novaeguinae having four, and

six rows of preanal scales where Greer

(1985:120) shows D. novaeguinae having

three. The hindlimb is 2.7% of SVL. The

tail is 22.6% of SVL. In life this specimen

was lilac of lavender-gray shading to buff

on the head and chalk-white on the tail. It

is much paler and less contrastingly marked

than the Guangxi D. bourreti illustrated by

Tian and Jiang (1986). See Figure 2.

Typhlops albiceps

This tiny snake, apparently rare

throughout its range, was known from

Hong Kong Island only on the basis of two

specimens collected in 1959 and 1966

(Karlsen et al. 1986), now in the British

Museum (Natural History): BMNH

1954.1.13.4 and BMNH 1983.946,

respectively. It was rediscovered 27 May

1988 at High West, Pokfulam, Hong Kong

Island by Sandra Brown (Macklin 1988).

The species appears reasonably common at

this site. Voucher specimens are at the St.

Louis School, West Point, Hong Kong

(curated by Anthony Bogadek) and at MCZ

FIG. 2. Dibamus sp. from Hei Ling Chau. The

rostral, R, lacks a suture from lip to nostril. A

labial suture, L, is present.

(MCZ 173290).

Lance (1976) did not recognize

"Typhlina" or "Ramphotyphops" and

neither do we. The putative distinction

from Typhops is entirely in male genetalia

(McDowell 1974). Because females are

indistinguishable, recognition of

"Typhlina" or "Ramphotyphops" directly

violates the principles of systematic

zoology (Mayr et al. 1953) and the

principles of animal taxonomy (Simpson

1961).

This is the only species of Typhlops in

China with 18 scale rows. It is very

slender, the head spatulate. MCZ 173290

is 155.5 mm SVL with a 2.5 mm tail. The

dorsum is wood brown shading to buff on

the head. The chin is near-white; this color

extends posteriorly for 10 scales onto the

throat The tail tip is near- white; this color

extends anteroventrally on all 10 subcaudal

scales and 4 rows anterior to the vent.

Dendrelaphis pictus

(Painted Bronze Back Snake)

This elegant snake is known from

Guangdong, Guangxi, and Yunnan (Hu et

al. 1980), and from Hong Kong on the

basis of a single report (Wall 1903). One

of us (Lazell) has searched both BMNH

and MCZ for this specimen with no

success. Peaker (1987) criticized Karsen et

al. (1986) for including this species in the

Hong Kong fauna. Deletion would have

Vol. 3, p. 66

Asiatic Herpetological Research

April 1990

TABLE 1. Some scale counts for Chinese

Ahaetulla prasina.

been premature. Between 1971 and 1984

Dr. Barry Hollinrake, resident of Shek

Kwu Chau, a small island 1.3 km south of

Lanuau, amassed a collection of 35 snakes

on the island of 9 species, now housed at

MCZ (Boalch 1988). Among these is a

beautiful female Dendrelaphis pictus with

15 dorsal scale rows, 171 ventrals, and 131

subcaudals: MCZ 173278.

The ventral and subcaudal counts of the

Shek Kwu Chau specimen are significantly

lower than those given by Hu et al. (1980:

39): 184 ventrals (females 186-193) and

141-169 subcaudals (females 141-169).

Ahaetulla prasina (Jade Vine Snake)

Two specimens of this spectacular

snake, never previously recorded in Hong

Kong Territory, are in the Hollinrake

collection from Shek Kwu Chau: male,

MCZ 173303, and female, MCZ 173304.

Their scale counts are much higher than

those given by Pope (1935) of Hu et al.

(1980) for this widespread South China

species: Table 1.

Literature Cited

BOALCH, K. 1988. Snakes alive —a specialist's

delight. South China Post 154(1 13):3.

GREER, A. 1985. The relationships of the lizard

genera Anelytropsis and Dibamus. Journal of

Herpetology 19:116-156.

HU, B., M. HUANG, Z. XIE, E. ZHAO, Y. JIANG,

Z YU, AND J. MA. 1980. Illustrated snakes of

China. Science and Technology Press,

Shanghai.

KARSEN, S. M. LAU, AND A. BOGADEK. 1986

Hong Kong amphibians and reptiles. Urban

Council, Hong Kong. 136 pp.

LANCE, V. A. 1976. The land vertebrates of

Hong Kong, Pp. 6-22. In B. Lofts (ed). The

fauna of Hong Kong. Royal Asiatic Society,

Hong Kong.

LAZELL, J. 1988. A leg up —by 800 kilometers.

Assoc. Pacific Systematists Newsletter 5:2.

MACKLIN, S. 1988.

biologists bewildered.

Post 154(147):!.

Reclusive reptile has

South China Morning

MAYR, E., E. G. LINSLEY, AND R. L. USFNGER.

1953. Methods and principles of systematic

zoology. McGraw-Hill, New York. 336 pp.

MCDOWELL, S. 1974. A catalogue of the snakes

of New Guinea and the Solomons, with special

reference to those in the Bernice P. Bishop

Museum 1. Scolecophidia. Journal of

Herpetology 8:1-57.

PEAKER, M. 1987. Review of Karsen et al.

(1986). The Herptile 12(l):36-37.

POPE, C. H. 1935. The reptiles of China.

Natural History of Central Asia 10, American

Museum of Natural History, New York. 604

pp.

SIMPSON, G. G. 1961. Principles of animal

taxonomy. Columbia University Press, New

York. 247 pp.

TIAN, W. AND Y. JIANG. 1986. Chinese

amphibian and reptile identification manual.

Science Press, Beijing. 364 pp.

WALL, F. 1903. A prodromus of the snakes

hitherto recorded from China, Japan, and the

Loo Choo Islands; with some notes.

Proceedings of the Zoological Society of

London 84-102.

April 1990

Asiatic Herpetological Research

Vol. 3, pp. 67-84

On the Independence of the Colchis Center of Amphibian and Reptile

Speciation

BORIS S. TUNIYEV1

^Causasian State Biosphere Reserve, Sochi, USSR

Abstract. -The Colchis region of Western Transcaucasia is characterized by a rather uniform thermal

regimen, corresponding to a subtropical climate. The Colchis forests contain an extraordinary abundance

and diversity of tree, shrub, and vine species. The herpetofauna of the Colchis forests is surprisingly poor,

despite its uniqueness.

Key Words: Amphibia, Reptilia, USSR, Caucasus, biogeography.

Introduction

The herpetofauna of Western

Transcaucasia is not homogeneous, due to

the different age and genesis of the species

distributions. Along with autochtonous

and endemic forms, one can find species

whose main areas of distribution are in the

European part of the USSR and in the

Eastern Mediterranean. At the same time, a

number of species which have main

distributional centers in the Colchis occur

beyond the bounds of Western

Transcaucasia, in other parts of the

Caucasian Isthmus. For these reasons, it is

necessary to define the Colchis

herpetofauna and to determine its place in

the fauna of reptiles and amphibians of the

Caucasian Isthmus as a whole.

Research on this issue started with the

works of Nordmann (1840), Derjugin

(1899), Silantyev (1903), Brauner (1905),

and Nesterov (1911). However, the first

well-grounded definition of the fauna in

question from a zoogeographical point of

view was presented in the works of

Satunin (1912). Satunin wrote in 1910,

"So far I cannot say much about the genesis

of the fauna of this region called Western

Transcaucasia. This country with its

evergreen plants and scanty fauna

resembles a piece of the Mediterranean in

the narrow sense of the word. True, here

are endemic species and forms, but not a

single genus of vertebrate is unrepresented

in the countries of the Mediterranean.

Often they are inhabited by the same

species. The question is whether this fauna

has appeared from the west or is it the

remainder of the fauna that has populated

densely the shores of the Black Sea at one

time. It is impossible to answer these

questions at the present level of our

knowledge. But even now I can definitely

say that this fauna by its origin, has nothing

in common with the faunas in other regions

of the Caucasus."

In 1912, Satunin divided the Caucasian

Isthmus into five subregions and 11

districts, including the Colchis in the West-

Transcaucasian district of the Littoral

subregion.

Among other merits of this work by

Satunin, one cannot but mention the fact

that for the first time, he defined in an exact

way the Colchis region proper. He defined

the northern border as the spurs of the Main

Caucasian Range up to the basin of the

Tuapse River, the southern border as the

Pontic Range, and the eastern border as the

Arsijanskij Range. The valley of the Rioni

River and the adjacent southern slopes of

the Main Range were defined as the central

part of the region. Satunin emphasized the

depauperate herpetofauna of this region on

one hand, and the presence of endemic

species such as Vipera kaznakowi and

Bufo verrucosissimus on the other hand.

Nikolsky (1911) assigned the entire

Caucasus, excluding eastern Precaucasia,

© 1990 by Asiatic Herpetological Research

Vol. 3, p. 68

Asiatic Herpetological Research

April 1990

to the Mediterranean. However, he could

not differentiate the forest and the alpine

belts of the Greater Caucasus, because of

the absence of data.

Results and Discussion

Investigations of the last decades made it

possible to add the majority of the species

of the Colchis herpetofauna to an overall

picture of Colchis faunal distributions

(Turov 1928; Bartenev and Reznikova

1935; Khozatsky 1941; Milyanovskiy

1957; Bannikov et al. 1977;

Negmedzyanov and Bakradze 1977; Orlova

1973, 1978a, 1978b; Golubev 1980, 1985;

Tuniyev 1983, 1985). In addition there has

been a revision of the taxonomic status of

such forms as Vipera kaznakowi

(Vedmederja et al. 1986; Orlov and

Tuniyev 1986a, 1990 this volume), Lacerta

agilis (Peters 1960), L. derjugini

(Bartenev and Reznikova 1931; Orlova

1978a ; Bischoff 1982, 1984), L. saxicola

(Darevsky 1967; Darevsky and Vedmederja

1977), Anguis fragilis colchicus (Lukina

1965; Scherbak and Scherban 1980), and

others.

Accumulation of this information along

with works on fossil amphibians and

reptiles of the Caucasus (Vekua et al. 1979;

Chkhikvadze 1981, 1983, 1984; Bakradze

and Chkhikvadze 1977; Zerova and

Chkhikvadze 1984; Yefimov and

Chkhikvadze 1987) have made it possible

to revise the zoogeography of the region.

Darevsky (1957) singled out seven

different groups of species and subspecies

of the herpetofauna in the Caucasus, based

on their origin. Among the species

representatives of the region of interest to

us, it is necessary to pay attention to

Lacerta strigata (Asia Minor species),

Emys orbicularis, Anguis fragilis,

Coronella austriaca, Elaphe quatuorlineates

sauromates, Natrix natrix (European boreal

species), Testudo graeca, Natrix tessellata

(Mediterranean species), Pseudopus

apodus, Coluber najadum (east-

Mediterranean species), and Lacerta

saxicola, L. praticola, L. derjugini, L.

media (autochthonous species). The

Colchis, however, was not distinguished as

an independent center of speciation in this

work.

Scherbak (1981) included the Colchis in

the Caucasian Region of the Mediterranean

Province. He suggested that the typical

species of the region were Mertensiella

caucasica, Pelodytes caucasicus, Lacerta

saxicola-complex and others. However,

the Colchis proper was again not

distinguished as an independent center of

herpetofaunal formation.

For the analysis of the herpetofauna of

the Colchis proper, it is necessary to

exactly define the term "the Colchis

phytolandscapes", and to decide what types

of vegetation are universally recognized as

"Colchis types". Albov (1885) was the

first to clearly depict plant landscapes of

the Colchis. He singled out a region,

unique for Russia, of mountain limestone

flora which had been developing mainly

autochthonously in a large refugium with

numerous endemic and relict species and

even genera. Kolakovskiy (1980) regarded

the Colchis flora as basically forest and

alpine-meadow, and suggested that its main

phytolandscapes had existed since old times

with changes only in the composition of

their edificators, except for the extinct

formation of evergreen subtropical forests

in the lower mountain belt. The tertiary-

relict character of the forest mesophile flora

and vegetation is fully revealed here due to

slight changes in this region's climatic

conditions (Kuznetsov 1891). The most

characteristic features of the tertiary-relict

Colchis forest are: extraordinary

abundance and diversity of tree and shrub

species, impossibility of singling out the

dominant species (which is also

characteristic of tropical forest with extreme

density of trees), abundance of vines and

epiphytes, and almost total absence of grass

cover. All these attributes make the Colchis

forest similar in many aspects to a tropical

rain forest (Pavlov 1984). According to

Sinskaya (1933), the Colchis forest

vegetation underwent three main stages of

development: the tropical forest; the forest

of Colchis type, but rich and covering a

wider area; and last, the modern Colchis

April 1990

Asiatic Herpetological Research

Vol. 3, p. 69

forest.

The Colchis type of vegetation includes a

number of phytocenoses differing in

structure, composition, and ecological

peculiarities: it may be mixed

(polydominant), or may be presented by

cenosis of one or two species, but the

common and obligatory attribute of

phytocenosis of the Colchis type is an

abundance of tertiary relicts. The area with

Colchis type vegetation is characterized by

a rather monotonous thermal regimen,

corresponding to a subtropical climate, but

with highly diverse soils (Gulisashvili et al.

1975).

The herpetofauna of the Colchis forests

is surprisingly poor, despite its uniqueness.

The species composition is different in the

southeastern and northwestern parts of the

Colchis compared to the other portions of

its territory. Species such as Mertensiella

caucasica, Lacerta clarkorum, L. parvula,

and L. mixta, whose distributions are

connected with forests growing on acid

soils above volcanic rocks, are found on

the western slopes of the Adzharo-

Imeretinsky, Shavshetsky and Lazistansky

(Pontic) mountain ranges. Similarly,

floristic endemics of this part of the Colchis

{Rhododendron ungernii, Osmanthus

decorus, Betula medwedewii, Epigaea

gaultheriodes, and others), are Adzharo-

Lazistan endemics, sometimes with slight

radiations to the adjoining regions, but are

not Colchis endemics in the broad sense of

the word. The same applies to Lacerta

saxicola darevskii and L. saxicola brauneri

which are widespread in the northwestern

part of the Colchis, but are absent in the

central and southeastern Colchis. These

animals, by analogy with floristic endemics

{Allium candolleanum, Campanula

mirabilis, C. bzybica, C. calcarea, C.

jadvigae, Genista abchasica, Gentiana

paradoxa, Omphalodes kusnetzovii, and

others) are northern-Colchis endemics. For

example, of the 450 endemic Colchis

species of flora, 83 (25%) are endemics of

the northern Colchis (Adzinba 1980).

Triturus vittatus ophryticus, T. vulgaris

lantzi, Bufo verrucosissimus, Pelodytes

caucasicus, Lacerta derjugini, L. agilis

grusinica, Natrix megalocephala, and

Vipera kaznakowi are Colchis endemics in

the broad sense of the word.

In addition to the Colchis endemics,

there are three more ecological-geographical

groups of amphibians and reptiles in the

region. They have similar ecological

characteristics (habitat first of all), and

overlapping geographic distributions.

1. The East-Mediterranean group

consists of Triturus cristatus karelini,

Testudo graeca nikolskii (Fig. 1 ), Lacerta

media, L. praticola pontica, L. strigata,

Pseudopus apodus tracius, Natrix

tessellata, and Coluber najadum. This

group's distribution includes either the

Balkans and the Caucasus or the Balkans,

Crimea, and the Caucasus. According to

ecological characteristics, these are

xeromesophiles or hemixerophiles whose

spreading is related to dry foothills of

Western Transcaucasia up to 200-300 m

above sea level with an annual sum of

temperatures exceeding 5000°C. Thus,

Testudo graeca, Pseudopus apodus, and

Coluber najadum occur in the Colchis on a

narrow seaside strip of land with enclaves

of Mediterranean vegetation from Tuapse to

Pitsunda-Sukhumi. A local population of

L. strigata occurs in the Pitsunda region,

and a local population of L. media occurs

in the environs of Pitsunda and Salme. The

majority of localities of L. praticola and

Natrix tessellata in the Colchis are in the

seaside hills up to 400 m above sea level.

It is only along the valleys of large rivers

like the Shakhe River, the Mzymta River,

the Bzyb River, and others that N .

tessellata penetrates into the Colchis up to

600 m above sea level. Thus the majority

of these species are found either in places

with vegetation of the Mediterranean type,

or in places where the initial Colchis

vegetation has been reduced to zero and the

landscapes resemble the Mediterranean

ones by their thermo-biotopic conditions

(substitute shibliaks and tomillares,

Pitsunda pine groves, foothill post-forest

glades, and landplots with Erica tetraliz,

Juniperus oxicedrus, Pinus pityusa, and

Arbutus andrachne).

Vol. 3, p. 70

Asiatic Herpetological Research

April 1990

FIG. 1. Testudo graeca nikolskii is a Mediterranean species, and in the Colchis it is found only in the

seaside strip of land with enclaves of Mediterranean vegetation.

2. The Caucasian group includes Hyla

arborea schelkownikowi (Fig. 2), Rana

macrocnemis, Lacerta caucasica alpina, L.

rudis, and Vipera dinniki. Distributions of

these species in the Caucasian Isthmus are

broader than those of the Colchis group.

At the same time, the majority of them are

Colchis autochthons. These species are

mesophiles and occur in mesophillous

forest and mountain meadow formations.

This group seems to be of Colchis origin,

retaining close connections with the main

center of the formation. Broader ecological

tolerance in comparison with typical

Colchis species makes it impossible to

include them within the Colchis group.

3. The European group consists of Bufo

viridis, Rana ridibunda, Emys orbicularis,

Anguis fragilis, Natrix natrix, Coronella

austriaca, and Coluber jugularis caspius.

The composition of this group is not

homogeneous. It includes both species

typical for the steppe areas {Bufo viridis

and Coluber jugularis ) and those that are

widely distributed in Europe (all the rest).

Only Anguis fragilis and Coronella

austriaca are widely distributed in the

Colchis. This makes it difficult to

definitely consider them late migrants to the

Caucasus. Other species either occur in

several spots along the Colchis seashore

(B. viridis and Natrix natrix ) or populate a

narrow strip of land along the sea together

with the Mediterranean species (C.

jugularis ) or a somewhat wider strip

(Emys orbicularis and Rana ridibunda ).

Despite the possibility of finding these

species in the typical Colchis forest

formations, the majority of them are still

attributed to the Mediterranean type of

vegetation.

April 1990

Asiatic Herpetological Research

Vol. 3, p. 71

FIG. 2. Hyla arborea schelkownikowi seems to be of a Colchis origin, but because of its broad

ecological tolerance, it is found in a large part of the Caucasian Isthmus.

Let us consider the dispersal and

distribution of the representatives of the

Colchis group in detail. Triturus vittatus

occurs in the territory from the seashore to

the subalpine meadows in all the forest

types. In the place called "the Colchis

Gates" (lowering of the Main Caucasian

Range between Mt. Fisht and Mt.

Chugush) the species crosses over to the

northern slope of the Western Calucasus,

reaching the environs of Goriachij Klutch

and Krasnodar in the northwest and the

basin of the Laba River in the northeast. In

the eastern part of the area, the species

crosses the Adzharo-Imeretinskij mountain

range and reaches the outskirts of Tbilissi-

Oni. It occurs in the Lagodekhi region as

an isolate. Outside the boundaries of the

Colchis this species occurs either in the

Colchis type forests or in their derivatives.

Triturus vulgaris lantzi (Fig. 3) occurs

in the same places in the Colchis as T.

vittatus ophryticus does. Very often both

species are symbiotopic (Tuniyev and

Beregovaya 1986). On the northern slope

of the Western Caucasus, its home range is

wider than that of the previous species, but

it is only through the mesophillous forests

and subalpine meadows that it penetrates

into the Eastern Transcaucasia up to the

Trialet Ridge. The isolated population in

Talysh occurs in the Hirkan forests which

are ecologically and genetically close to the

Colchis forests.

Bufo verrucosissimus (Fig. 4) occurs in

all parts of the Colchis from the sea shore

up to the subalpine forests. On the

northern slope of the Western Caucasus, it

is found on the territory from the environs

Vol. 3, p. 72

Asiatic Herpetological Research

April 1990

FIG. 3. Triturus vulgaris lantzi. This subspecies is widespread throughout the Colchis.

of Krasnodar in the west to the Psebaj

settlement and the Shakhgirej Canyon in the

east, where it is found in the derivatives of

the Colchis forests from 400 m to 1000 m

above sea level. In the Eastern Caucasus it

is found in the Borzhomy Gorge, the

Lagodekhi-Zakataly, and the Talysh region,

where its distribution is limited to the

mesophilous forests, which are abundant in

the Colchis and Hirkan floral elements.

The distribution of Pelodytes caucasicus

(Fig. 5) in the Colchis is more restricted. It

does not occur in the coastal belt and oak-

forests. It is found both in the mesophilous

beech, chestnut, and fir-tree forests, and in

mixed broad-leaved forests with an

evergreen understory. On the northern

slope of the Western Caucasus, its

distribution coincides with that of B.

verrucosissimus, but unlike the latter, it is

not found in deforested places. Its

southeastern distributional limits also

coincide with that of Bufo verrucosissimus.

Pelodytes caucasicus does not occur east

of the Trialet Ridge. There is an isolated

population in the mesophilous forests in the

Lagodekhi-Zakataly region.

Lacerta derjugini has a distribution in

the Colchis similar to P. caucasicus. It

reaches the sub-alpine belt. On the

northern slope of the Western Caucasus, it

is found in the Colchis forest derivatives

from the Belaja River to the Small Laba

River (the Shakhgirej Gorge). The species

penetrates through the Eastern

Transcaucasia up to the Trialet Ridge.

Separate populations occur in north-eastern

Georgia up to Lagodekhi-Zakataly.

Lacerta agilis grusinica (Fig. 6) is

known to occur only on the territory of the

Colchis and the adjoining sea coast up to

April 1990

Asiatic Herpetological Research

Vol. 3, p. 73

FIG. 4. Bu/o verrucosissimus. This Colchis endemic occurs in the four Colchis refugia.

Novorossijsk. Its vertical distributional

limit is 700 m above sea level, though local

populations can be found in the sub-alpine

belt (Mt. Aishkho and Mt. Uglovoj).

The distribution of Natrix megalocephala

is similar to that of many Colchis species.

It is found from the environs of Tuapse and

Gorjachij Kljutch in the west, to the region

between the Belaja Laba River and the

Small Laba River in the north, eastwards

through the whole territory of the Colchis

up to the Borzhomy Gorge and separately

in the Lagodekhi-Zakataly region (Orlov

and Tuniyev 1986b). In the Colchis it

reaches the sub-alpine belt. In the rest of

the area it does not exceed 1000 m above

sea level.

Elaphe longissima (Fig. 7) occurs in the

region of Novorossijsk and throughout

Western Transcaucasia, except for the mid-

mountains and highlands. There are

isolated populations in the Borzhomy

Gorge, the Lagodekhi-Zakataly region and

the Belaja River basin on the northern slope

of the Western Caucasus.

Vipera kaznakowi (Fig. 8) occurs

throughout the territory of the Colchis up to

1000 m above sea level. On the northern

slope of the Western Caucasus it occurs

between the Belaja River and the Small

Laba River. Separate populations are

known in the Borzhomy Gorge and the

Lagodekhi region.

The dispersal and distribution of Lacerta

saxicola darevskii (Fig. 9), L. s. brauneri,

L. mixta, L. parvula and Mertensiella

caucasica have been analyzed above.

A comparison of the endemic Colchis

species, whose distributions are associated

Vol. 3, p. 74

Asiatic Herpetological Research

April 1990

FIG. 5. Pelodytes caucasicus occurs in all four Colchis refugia.

with forest and meadow formations of the

Colchis type, makes it possible to identify

three more regions in the Caucasian

Isthmus, besides the Western Caucasus

(the Colchis proper), in which the Colchis

herpetofauna occurs. The three regions are:

the Bjelo-Labinskij region on the northern

slope of the Western Caucasus, the

Kakhetinskij (Lagodekhi-Zakataly) region

on the southern slope of the Eastern

Caucasus, and the Borzhomskij region in

Eastern Transcaucasia (Fig. 10). The

comparative composition of the

herpetofauna of these regions is shown in

Table 1.

It is evident from Table 1, that the most

significant differences are found between

the Colchis and the Kahetinskij regions and

the least significant differences are between

the Colchis and the Borzhomskij and the

Belo-Labinskij regions. Taking into

account the above mentioned peculiarities

of distribution of herpetofauna within the

Colchis refugium itself, the differences

become even less significant. In this case,

we deal with three regions smaller in space,

and a wealth of species of the Colchis

herpetofauna that occur in refugia and have

survived off the main territory of the

Colchis.

The Belo-Labinskij region is only

conventionally separated from the Colchis

by the crest of the Main Caucasian

Mountain Range. All the northern Colchis

species, except L. a. grusinica, occur on

the northern slope of the Western Caucasus

in the Belaja and the Small Laba river

drainages. It should be stressed that this

unity is based on the fact that the

characteristic Colchis elements of flora and

vegetation cross over the Main Caucasian

Range in the place known as "the Colchis

Gates" to its northern slope. The basins of

April 1990

Asiatic Herpetological Research

Vol. 3, p. 75

FIG. 6. Lacerta agilis grusinica is found throughout the main Colchis refugium.

TABLE 1 . Distribution of the endemic Colchis herpetofauna in the main refugia of the Caucasian

Isthmus.

Species

R

Colchis

E G

Bjelo-

Labinskij

I O N

Kakhetinskij Borzhomskij

Triturus vittatus

T. vulgaris lantzi

Mertensiella caucasica

Bufo verrucosissimus

Pelodytes caucasicus

Lacerta derjugini

L. agilis grusinica

L. saxicola darevskii

L. s. brauneri

L. mixta

L. parvula

L. clarkorum

Natrix megalocephala

Elaphe longissima

Vipera kaznakowi

+ + + +

+ + - +

+ - - +

+ + + +

+ + + +

+ + + +

+ - - -

+ + - -

+ + - -

+ - - +

+ - - +

+ - - -

+ + + +

+ + + +

+ + + +

Total:

15

10

11

Vol. 3, p. 76

Asiatic Herpetological Research

April 1990

FIG. 7. Elaphe longissima. The Colchis refugia populations are disjunct from the main distribution in

Europe.

the Belaja, Tsitse and Laba rivers abound in

those elements. To the north of these

watersheds their distributions are

continuous up to the Skalistij (Rocky)

limestone ridge. Maleyev (1939) has noted

that a part of the Maikop district abounds in

Colchis elements and is inseparable from

the Colchis according to the character of its

flora and vegetation.

The Borzhomskij region is also

conventionally separated from the Colchis

by the Adzhara-Imeretinskij Mountain

Ridge. The flora and vegetation of the

Baniskhevskoje Gorge, the Likanskoje

Gorge, and the upper belt of Mt. Lomis-

Mta, as well as the environs of Bakuriani,

hardly differ from those of the Colchis.

April 1990

Asiatic Herpetological Research

Vol. 3, p. 77

FIG. 8. Vipera kaznakowi occurs throughout the territory of the Colchis up to 1000 m above sea level.

In the isolated eastern Kakhetinskij

region a considerable number of the ancient

Tertiary vegetation representatives survived

due to the warm and humid climate

(Gulisashvili et al. 1975).

Modern distributions of eco-geographic

groups of amphibians and reptiles

distinguished by our scientists have distinct

altitudinal-ecological limits owing to natural

and historical reasons.

Migration of ancestoral species of the

Colchis and the Caucasian groups from the

south apparently took place in the Miocene

when the Caucasian island joined vast

territories of Asia Minor. Colonization of

the Caucasus from the south by different

species of mammals during the early

Miocene has been studied by Vereschagin

(1958), and that of lizards of the Podarchis-

Archaeolacerta group by Darevsky (1967).

The early Miocene was about the time of

the formation of the Caucasian Mountains

(Bogachev 1938).

The warm subtropical climate and

vegetation in the Caucasus favored the

evolution and dispersal of heat and

mesophilic forms (Triturus vittatus

ophryticus, T. vulgaris lantzi, Pelodytes

caucasicus, Bufo verrucosissimus, Lacerta

saxicola darevskii, L. s. brauneri, L.

derjugini, Elaphe longissima, Natrix

megalocephala, and Vipera kaznakowi ), as

well as species with a broader ecological

tolerance (Rana macrocnemis, Hyla arborea

schelkownikowi, Lacerta rudis, and L.

agilis grusinica ).

Fossil remains of mammals

(Mesocricetus, Prometheomys, Sorex,

Talpa ) [Vereschagin 1958] and insects

(Orthoptera, Hemiptera, Blattoidea, and

Vol. 3, p. 78

Asiatic Herpetological Research

April 1990

FIG. 9. Lacerta saxicola darevskii. This lizard is a Colchis endemic whose distribution is restricted to the

northwestern part of the Colchis.

Coleoptera) [Rodendorf 1939] suggest a

good food supply for amphibians and

reptiles during the Miocene. It was also in

the Miocene that the majority of these

species reached the eastern-most parts of

the Greater Caucasian Range along its

southern slopes and penetrated from there

into the Talysh across the so-called

"Karabakhi Bridge". Safarov (1966), and

other scientists, have studied the former

direct relations between the Colchis and the

Hirkan floras. Even at the present time, the

floristic composition of the Kakhetinskij

region and of the Karabakh has many

common features with that of the Colchis

and the Talysh forests (Arushanyan 1973;

Sokolov 1977; Takhtadzhan 1978;

Gadzhiyev et al. 1985).

The end of the Tertiary period was

characterized by damping of tectonics due

to the broad correlation of the Caucasus and

the Balkans (Vereschagin 1958) and the

formation of the steppe landscapes along

the northern Black Sea coast (Pidoplichko

1954; Scherbak 1966). During that period,

such South-European species as Rana

ridibunda, Bufo viridis, Emys orbicularis,

Anguis fragilis, Coluber jugularis, and

Coronella austriaca seem to have

penetrated to the Precaucasia from the west.

At the same time, such species as Testudo

graeca, Pseudopus apodus, Triturus

cristatus karelini, Lacerta praticola pontica,

L. media, Coluber najadum, and Natrix

tessellata got into the Colchis from the

west along the Black Sea coast.

Early and Middle Pliocene should be

considered the beginning of initial

fragmentation of the Colchis faunal areas

when the Greater and the Lesser Caucasian

April 1990

Asiatic Herpetological Research

Vol. 3, p. 79

FIG. 10. The main refugia of the endemic Colchis herpetofauna.

Mountain ranges underwent substantial

glaciation (Grozdetskiy 1954; Markov et al.

1965). The main center of dispersal of

these species was the Colchis, where

relatively heat-loving vegetation of the

Caucasian type survived even during the

periods of the extreme Pleistocene cooling

(Vereschagin 1958; Adamyants 1971).

Along with the Colchis, smaller refugia

sporadically survived on the territory of the

Caucasus Black Sea coast, and also on the

northern slope of the Main Caucasian

Range between the Pshekha River and the

Small Laba River. The present distribution

of the Tertiary vegetation of the Colchis

type in the western Caucasus testifies to

this (Kharadze 1974; Pechorin and

Lozovoy 1980; Kholyavko et al. 1978;

Adamyants 1971; Koval and Litvinskaya

1986).

It was in the narrow humid gorges with

a relatively constant thermal regimen that

the representatives of the Colchis group

remained intact. At the same time,

independent populations might also have

been preserved in mid-mountain areas

where refugia of the Colchis vegetation

exist in the vicinities of the Fisht-Oshtenskij

Mountain Massive, the Lagonaki Plateau

and even in the Central Caucasus

(Kholyavko et al. 1978; Kharadze 1974).

Small refugia seem to have also remained

on the southern slope of the eastern pan of

the Greater Caucasus and in the Kuru River

Gorge. It is indisputable that the majority

of the mountainous populations of Colchis

species perished during the Pleistocene and

Vol. 3, p. 80

Asiatic Herpetological Research

April 1990

that the ones that survived in refugia have

been accumulating unique characteristics

which led to different geographical forms

(subspecies) on different slopes of the Main

and the Adzharo-Imeretinskij mountain

ranges. The data presented by Takhtadzhan

(1946) and Maruashvili (1956) support the

hypothesis concerning the preservation of

relict Colchis species in the mountains.

According to their data, the average annual

temperature during the glacial periods

decreased not more than 1.5-2.0°C, while

precipitation amounted to not less than

1500-2000 mm.

Darevsky (1967) considers that this

argument supports the possibility of foothill

refugia of reptiles existing on the sea facing

slopes of the Gagrinskij and the Bzhybskij

mountain ranges, and in other regions,

despite radical reorganization of the

distributions of all the species of plants and

animals in connection with glaciation.

During the interglacial and especially the

postglacial periods, reconstruction of all

vegetation belts took place (Vereschagin

1958). This favored the isolation of the

species of the Colchis and the Caucasian

groups in the above-mentioned refugia, but

favored wider dispersal of the European

and the Mediterranean groups in

Transcaucasia. In the northwestern part of

the Caucasus Black Sea coast, mesophilic

vegetation gave way to xerophytic

vegetation of the Mediterranean type. Plant

formations of this type with the prevalence

of Juniperetum, Querceto, Pinetum

carpinulosum, Pinetum fruticosum and

shibliaks are characteristic of the Anpa-

Gelendzhik region. Enclaves of

Mediterranean vegetation remained still

further to the south, up to Pitsunda

(Takhtadzhan 1978; Kolakovskiy 1961).

When the xerothermic period ended, the

climate became more humid again. This

favored the reestablishment of the former

borders of the forest belt (Vereschagin

1958). Subalpine meadows and elfin

woodlands were expanding throughout the

whole subalpine belt of the southern slope

of the Main Caucasian Mountain Range

from the Central Caucasus to the Fisht-

Oshtenskij Mountain Massive (Kholyavko

et al. 1978; Kharadze 1974; Dolukhanov

1974; Galushko 1974). On the northern

slope such vegetation is present in the

western parts and changes its character to

the east, transforming into a steppe type

(Lavrenko 1980). Modern areas of Vipera

dinniki and Lacerta caucasica alpina have

fixed distributional limits in the subalpine

belt influenced by the warm Black Sea.

Final settling of the present-day climate

favored fixation of the Colchis species

distributions, with their distinct populations

exceeding the bounds of the refugia. This

was coupled by simultaneous depression

and reduction of the European and

especially of the Mediterranean species

distributions.

Concluding this review of the Colchis

herpetofauna and their main refugia, it is

necessary to enumerate the most important

characteristics. The Colchis species are

characterized by antiquity (conservation

since the Tertiary period), Autochthonity,

depression — for some species {Vipera

kaznakowi; Lacerta clarcorum, L. a.

grusinica), existence of the northern-

Colchis limestone, and the southern-

Colchis volcanic centers of formation of

narrow-endemic forms. Reptiles have a

common tendency toward melanism, while

amphibians approach their low temperature

thresholds. These are the adaptive features

acquired during the glacial period. As a

rule, the modem distribution of the Colchis

species does not exceed the bounds of the

Colchis vegetation refugia or their

derivatives.

The maximal vertical distribution is in

the center of the Colchis up to 1800 m

above sea level, while in the other portions

of the refugia it does not exceed 1000 m

above sea level as a rule. The existence of

four refugia of the Colchis herpetofauna in

the Caucasian Isthmus is determined by

natural factors of high order- these are the

areas with slightly changed climatic

conditions characterized by modern

crossing of the January -3°C isotherm and

800 mm isohyet.

April 1990

Asiatic Herpetological Research

Vol. 3, p. 81

Literature Cited

ADAMYANTS, G. I.. 1971. [On Castanetum of

the Caucasus]. A Report of the Sochi

Department at the USSR Geographical Society.

Issue 2. Leningrad. (In Russian).

ADZINBA, Z. I. 1980. [The plants of Abkhazia

from limited area]. Proceedings of the Botanical

Gardens. Sukhumi, Tbilisi, Georgia Issue

26:69-82. (In Russian).

ALBOV, N. M. 1885. [Papers on the flora of

Kolkhida]. Proceedings of the Tiflis Botanical

Gardens. Tiflis, Georgia Issue 1:1-288. (In

Russian).

ARUSHANYAN, R. I. 1973. [The relicts of

Nagomy Karabakh and their new findings].

Botanical Journal Leningrad 58(5):700-706. (In

Russian).

BAKRADZE, M. A. AND V. M. CHKHIVADZE

1977. [Fossils of the tortoises from Testudo

genus in the USSR]. Prosveschenye

Publishing House, Moscow. 369pp. (In

Russian).

BANNIKOV, A. G., I. S. DAREVESKY, V. G.

ISCHENKO, A. K. RUSTAMOV, AND N. N.

SCHERBAK. 1977. [Field guide of the USSR

amphibians and reptiles]. Prosveschenye

Publishing House, Moscow. 369pp. (In

Russian).

BARTENEV, A. AND M. REZNIKOVA. 1931.

NEE LAC ERT A -Formen (Reptilia) aus dem

Kaukasichen Staatsnaturschtzgebiete

(Westkaukasus). Zoologischer Anzeiger 96(9-

10):268-271.

BARTENEV, A. AND M. REZNIKOVA. 1935.

[Papers on the fauna of amphibians and reptiles

from the western and central parts of the State

Caucasus Reserve]. Bulletin of the Museum of

Georgia 8:1 1-41. (In Russian).

BISCHOFF, W. 1982. Zur Kenntnis der

innerartlichen Gliederung der Artwiner Eicechse,

Lacerta derjugini Nikolskij, 1898. Zool. Abh.

Bd. 38, Nr. 1 S. 1-51.

BISCHOFF, W. 1984. Bemerkungen zur

innerartlichen Gliederung und zur Verbreitung

der Artwiner Eidechse (Lacerta derjugini

Nikolskij, 1989) an den Sudhangen des GroBen

Kaukasus. Salamandra, Bd. 20, Nr. 2/3, S.

101-111.

BOGACHEV, V. V. 1938. [On the zoogeography

of the Caucasus and its geological history].

Materials of the Azerbajzhan Affiliation. The

USSR Academy of Sciences, Moscow. 4:29-

47. (In Russian).

BRAUNER.A. 1905. [Preliminary data on reptiles

and amphibians from the Crimea Kubanskaya

region, Volynskaya and Varshavskaya

gubernias]. Notes of the Novorossiskoye

Society of Researchers. Odessa, Russia. 28:1-

88. (In Russian).

CHKHIKVADZE, V. M. 1981. [Survey on the data

of amphibian and reptile neogene fossils from

the northern area adjacent to the Black Sea

coast]. Science Publishing, Leningrad, pp. 151-

152. (In Russian.)

CHKHIKVADZE, V. M. 1983. [Fossil turtles of the

Caucasus and northern territory adjacent to the

Black Sea coast]. Metsniyereba Press, Tbilisi,

Georgia 48 pp. (In Russian).

CHKHIKVADZE, V. M. 1984. [Survey on the

fossils of tailed and tailless amphibians of the

USSR]. Academic Press, Georgian Academy of

Sciences, Tbilisi 10(1):5-13. (In Russian).

DAREVSKY, I. S. 1957. [The fauna of the

Armenian reptiles and their zoogeographical

analysis]. Ph. D. Thesis. Yerevan, Armenia.

28 pp. (In Russian).

DAREVSKIY I. S. 1967. [Rock lizards from the

Caucasus (Lacerta saxicola).} Science

Publishing House, Leningrad. 210 pp. (In

Russian).

DAREVSKY, I. S. AND V. I. VEDMEDERJA. 1977.

[A new species form of the Lacerta saxicola

group from southeastern Turkey and adjacent

areas of Adzharia]. Proceedings of the

Zoological Institute, USSR Academy of

Sciences, Leningrad 74:50-54. (In Russian).

DERJUGIN, K. M. 1899. [Report on a field trip

and zoological research in Charivski territory,

southwestern Transcaucasia and the vicinity of

Trapezond]. Proceedings of Saint Petersburg

Society of Researchers, St. Petersburg 30(2): 1-

65. (In Russian).

DOLUKHANOV, A. G. 1974. [Subalpic landscapes

of the Caucasus as shelters for relict fauna

elements]. Problems of Botanies, Leningrad

12:27-34. (In Russian).

Vol. 3, p. 82

Asiatic Herpetological Research

April 1990

GADZHIYEV, V. D., A. M. ALEKPEROV, M. R.

EFENDIJEV, AND R. K. MUSTAFAJEVA. 1985.

[The Zakatalski Reserve]. Agroprom Press,

Moscow. 184 pp. (In Russian).

GALUSHKO, A. I. 1974. [Relicts of the highland

flora of the western part of the Central

Caucasus]. Pp. 19-26. In Problems of

Botanies, Vol. 12, Flora of Highlands and its

Exploration. Science Publishing Leningrad.

(In Russian).

GOLUBEV, N. S. 1980. [On the habitat of

Pelodytes caucasicus (Amphibia, Pelobatidae)].

Zoological News, Kiev, Ukraine (3):52-55. (In

Russian).

KOVAL, I. P. AND S. A. LITVINSKAYA. 1986.

[Rare vegetation communities on the Krasnodar

territory]. Rostov State University Press,

Rostov-upon-Don, USSR, part 3, 336 pp. (In

Russian).

KUZNETSOV, N. I. 1891. [Elements of the

Mediterranean area in Western Transcaucasia].

Notes of the Russian Geographical Society, St.

Petersberg 23(3): 1-190. (In Russian).

LAVRENKO, Y. M. 1980. [Caucasus montane and

meadow steppes and stepped meadows]. Pp.

269-272. In Vegetation of the European part of

the USSR, Science Publishing, Leningrad. (In

Russian).

GOLUBEV, N. S. 1985. [Caucasus land toad,

Pelodytes caucasicus Boulenger (distribution,

morphology and ecology)]. Ph.D. Thesis.

Leningrad. 25 pp. (In Russian).

GROZDETSKIY, N. A. 1954. [Physical geography

of the Caucasus]. Moscow State University

Press, Moscow 2(1): 1-208. (In Russian).

GROZDETSKIY, N. A. 1958. [Physical geography

of the Caucasus]. Moscow State University

Press, Moscow 2:(2)l-264. (In Russian).

GUUSASHVILI, V. Z., L. B. MAHATADZE, AND

L. I. PRILIPKO. 1975. [Vegetation of the

Caucasus]. Science Publishing House,

Moscow. 223 pp. (In Russian.)

KHARADZE, A. L. 1974. [On some floragenetic

endemic groups in the Big Caucasus].

Problems of Botanies, Science Publishing

House, Leningrad 12:70-76. (In Russian).

KHOLYAVKO, V. S., D. A. GLOBA-

MIKHAILENKO, AND E. S. KHOLYAVKO

1978. [The tree atlas of the Caucasus].

Moscow. 215 pp. (In Russian).

KHOZATSKY, L. I. 1941. [Sea storms and some

land living animals]. Wildlife, Moscow 1:112-

113. (In Russian).

KOLAKOVSKIY, A. A. 1961. [Vegetation of

Kolkhida]. Materials on the USSR fauna and

flora. MOIP Press, Moscow, Issue 10. 460 pp.

(in Russian).

KOLAKOVSKIY, A. A. 1980. [The flora of

Abkhazia]. Metsniyereba Publishing, Tbilisi,

Georgia 1:7-18. (In Russian).

LUKINA, G. P. 1965. [On the taxonomy of the

legless lizard]. Herpetology, Tashkent,

Uzbekistan, USSR, pp. 6-10. (In Russian).

MALEYEV, V. P. 1941. [Tertiary relicts in the

flora of the Western Caucasus and major stages

of the Quaternary history of its USSR flora].

The USSR Academy of Sciences Press,

Moscow, pp. 61-144. (In Russian.)

MARKOV, K. K., G. I. LAZUKOV, AND V. A.

NIKOLAYEV. 1965. [The Caucasus]. Pp. 306-

321. In Quaternary Period, v. 1, Territory of

the USSR, part 2. Area of Ancient Glaciation

of High Mountains of the USSR South, chapter

2, Moscow University Press, Moscow. (In

Russian).

MARUASHVILI, L. I. 1956. [The reason to review

the present day concepts about

paleogeographical conditions of the glacier

period in the Caucasus]. Press of the Institute

of Geography named after Vakhushti, Tibilisi,

Georgia. 113 pp. (In Russian.)

MILYANORSKIY, Y. S. 1957. [On the snake

fauna of Abkhazia]. Proceedings of the

Abkhazski State Museum, Georgia 2:199-203.

(In Russian).

NEGMEDZYANOV, V. A. AND M. A. BAKRADZE.

1977. [On the herpetology of Kolkhida]. Pp.

153-155. In Questions of Herpetology,

Leningrad. (In Russian).

NESTEROV, P. V. 1911. [Materials on the

herpetology of the southwestern Transcaucasia

(Chokhotski Territory)]. Annual Journal of the

Zoological Museum, Academy of Sciences, St.

Petersburg 10(1): 1-18. (In Russian).

April 1990

Asiatic Herpetological Research

Vol. 3, p. 83

NIKOLSKY, A. M. 1911. [Reptiles and

amphibians of the Caucasus]. The Caucasus

Museum Press, Tiflis, Georgia 10(1): 1-18. (In

Russian.)

NORDMANN, A. 1840. Cataloque raisonne des

mammiferes de la faune Pontique. Voyage dans

la Russia meridionale et la Crimec, par la

Hongric, la Valachie et la Moldavic, vol. 3, s.

535-551.

ORLOV, N. L. AND B. S. TUN1YEV. 1986a.

[Present ranges, possible ways of their

formation and phylogeny of the three species of

vipers from the Eurosiberic gTOup (V .

kaznakowi complex) in the Caucasus].

Proceedings of the Zoological Institute, USSR

Academy of Sciences, Leningrad 157:104-135.

(In Russian).

ORLOV, N. L. AND B. S. TUNIYEV. 1986b. [A

new species of the grass snake Natrix

megalocephala sp. nov from the Caucasus]. Pp.

116-130. In Proceedings of the Zoological

Institute, Leningrad, Vol. 12. Leningrad. (In

Russian).

ORLOV, N. L. AND B. S. TUNIYEV. 1990. Three

species in the Vipera kaznakowi complex

(Eurosiberian Group) in the Caucasus: their

present distribution, possible genesis, and

phylogeny. Asiatic Herpetological Research

3:1-36.

ORLOVA, V. F. 1973. [Herpetofauna of the

northern area of the State Caucasus Reserve].

Zoological News, Kiev, Ukraine 2:61-65. (In

Russian).

ORLOVA, V. F. 1978a. [Geographical variation

and lektotype description of Lacerta derjugini

Nikolski]. Pp. 188-203. In Birds and

Reptiles. Moscow State University Press,

Moscow. (In Russian).

ORLOVA, V. F. 1978b. [Geographical

distribution and intraspecific variation of Lacerta

praticola from the Caucasus. Pp. 204-215. In

Birds and Reptiles. Moscow State University

Press, Moscow. (In Russian).

PAVLOV, N. B. 1984. Botanic geography of the

USSR, Moscow, pp.13-17. (In Russian).

PESHORIN, A. I. AND S. P. LOZOVOY. 1980.

[Natural monuments of the Krasnodar Territory].

Krasnodar Press, Krasnodar. 138 pp. (In

Russian).

PETERS, G. 1960. Die Grusinische Zauneidechse

Lacerta agilis grusinica nomen novum. Zool.

Anz. Bd. 165, Nr. 7/8, S. 113-126.

PIDOPLICHKO, I. G. 1954. [On the glacier

period]. The Ukraine Academy of Sciences

Press, Kiev, Ukraine. 264 pp. (In Russian).

RODENDORF, B. B. 1939. [On the Miocene fauna

of insects in the vicinity of Voroshilovsk].

Priroda Publishing, Moscow, N. 12. (In

Russian).

SAFAROV, I. V. 1966. [Dzelkva, its biological

and ecological peculiarities and industrial

significance]. Academic News, Azerbajzhan

1:15-17. (In Russian).

SATUNIN, K. A. 1912. [On zoogeographical

circles of the Caucasus Territory]. The

Caucasus Museum News, Tiflis, Georgia

7:(1)1-106. On Russian).

SCHERBAK, N. N. 1966. [Amphibians and

reptiles of the Crimea]. Naukova Dumka

Publishing House, Kiev, Ukraine. 239 pp. (In

Russian).

SCHERBAK, N. N. 1981. [Fundamentals of

herpetogeographic regional divisions of the

USSR territory]. Pp. 157-158 In Questions

of Herpetology, Leningrad. (In Russian).

SCHERBAK, N. N. AND SCHERBAN. 1980.

[Amphibians and reptiles of the Ukrainian

Karpat Mountains]. Naukova Dumka

Publishing House, Kiev, Ukraine. 253 pp. (In

Russian).

SILANTYEV, A. 1903. [Materials on the

herpetology of the Black Sea region]. Annual

journal of the Zoological Museum, Academy of

Sciences, St. Petersbourg 8:1-10. (In Russian).

SINSKAYA, Y. N. 1933. [Major evolutional

characters of the forest vegetation in the

Caucasus with regard to the specific history].

Botanical loumal, Leningrad 18(5-6): 117- 130.

(In Russian).

SOKOLOV, S. Y. 1977. [The habitats of trees and

bushes of the USSR]. Science Publishing

House, Leningrad 1:1-163. (In Russian).

TAKHTADZHAN. 1946. [On the history of the

vegetation developments in Armenia].

Proceedings of the Botanic Institute, Armenian

Academy of Sciences, Yerevan, Armenia 4:51-

Vol. 3, p. 84

Asiatic Herpetological Research

April 1990

107. (In Russian).

TAKHTADZHAN. 1978. [Floristic areas of land].

Science Publishing, Leningrad. 219 pp. (In

Russian).

VEDMEDERJA, V. I., N. L. ORLOV AND B. S.

TUNIYEV. 1986. [On the taxonomy of three

viper species of the Vipera kaznakowi

complex]. Proceedings of the Zoological

Institute, Leningrad 157:55-61. (In Russian).

TUNIYEV, B. S. 1983. [The herpetofauna of the

southern part of the Caucasus State Reserve].

Pp. 84-94. In Conservation of the Relict Flora

and Fauna in the Northwestern Caucasus.

Leningrad. (In Russian).

TUNIYEV, B. S. 1985. [Rare and endangered

amphibians and reptiles of the Caucasus

Reserve]. Pp. 104-116. In Ecological

Researches in the Caucasus Biospheric Reserve.

Rostov State University Press, Rostov-upon-

Don. (In Russian).

TUNIYEV, B. S. AND S. Y. BEREGOVAYA 1986.

[Sympatric amphibians of yew-box groves].

Proceedings of the Zoological Institute, USSR

Academy of Sciences, Leningrad 157:136-151.

(In Russian).

VEKUA, A. K., K. S. KALANDADZE, AND V. M.

CHIKVADZE. 1979. [New paleontological

findings from Belaya (White) Cave, Georgia,

USSR]. Reports of the Georgian Academy of

Sciences, Tbilisi, Georgia. 96(3):745-748. (In

Russian).

VERESCHAGIN, N. K. 1958. [Genesis of the

terrain fauna of the Caucasus Isthmus]. Pp.

506-514. In Fauna of the USSR, Vol. 5.

Montane Areas of the European Part of the

USSR, Academic Press, Moscow. (In

Russian).

YEFIMOV, M. B., AND V. M. CHKHIVADZE.

1987. [Survey on crocodile fossils in the

USSR]. Georgian Academy of Science News,

Tbilisi, Georgia 13(3):200-206. (In Russian).

TUROV, S. S. 1928. [Materials on the fauna of

the State Caucasus Reserve]. Works of the

North Caucasus Research Institutions

Association, Rostov-upon-Don 2:1-15. (In

Russian).

ZEROVA, G. A. AND V. M. CHKHIVADZE. 1984.

[Survey on cenozoic lizards and snakes from the

USSR]. Georgian Academy of Sciences News,

Tbilisi, Georgia 10(5):319-325. (In Russian).

April 1990

Asiatic Herpetological Research

Vol. 3, pp. 85-100

Studies of the Early Embryonic Development of Rana rugulosa Wiegmann

jionghuaPan1 and Danyu Liang1

^Department of Biology, South China Normal University, Guangzhou, Guandong, China

Abstract. -This paper deals with the early embryonic developmental stages of Rana rugulosa

Wiegmann, and with methods of artificial fertilization and experimentally accelerated development.

Developmental stages are distinguished by morphological changes and obvious physiological features. At a

temperature of 25.1 to 27°C it takes 70 hr 20 min to complete development from fertilized eggs to tadpoles

with opercular folds. This course of development is divided into 25 stages, which are standardized with

normal table equivalents.

Key Words: Amphibia, Anura, Ranidae, Rana rugulosa, embryonic development, artificial fertilization,

accelerated development, fertilized eggs, normal table.

Introduction

Rana rugulosa Wiegmann, whose

popular name is the Field Chicken in

China, is strong and big of body. It is

famous for its delicacy of meat and the

price is dear. Domestic and foreign

markets are in great need of it. The growth

and development of these frogs are

relatively quick. Hatchlings can become

sexually mature in one year. The average

body length is about 1 10 mm. The weight

is about 0.25 kg. Consequently, Rana

rugulosa farming has been a growing

enterprise. More papers deal with the

embryonic development of R. guentheri

Boulonger and R. catesbeiana. Currently

there are few domestic and foreign studies

on the embryonic development of R .

rugulosa. In order to produce reference

material for the captive breeding of this

frog, we studied its early embryonic

development, using artificial fertilization

and higher temperatures to accelerate

development experimentally, during the

months of April and June, 1987 and 1988.

Methods

We made 1 1 observations on the

embryonic development of Rana rugulosa.

Six of the 11 were artificially fertilized.

The parent frogs were bought in

Guangzhou. The description of the

features of embryonic development was

mainly based on the six artificially fertilized

eggs. The water temperature of the six

developing embryos was 25.1 to 27°C; the

pH was 6.5-6.8. Several minutes after

oviposition and ejaculation or artificial

fertilization, we put the oocyte into the

laboratory. Then we used a microscope to

observe the course of development and to

measure the size and shape. The

developmental stages were defined as

beginning when half of the embryos in a

sample showed all of the distinguishing

characters of that stage.

For every developmental stage we took

10 to 20 embryos and fixed them in 5%

formalin and Bouin's fluid. This material

was used in drawings, tissue-slices, and

photomicrographs. The figures shown

here were taken from observations of living

material.

Results

The 25 stages of the early embryonic of

Rana tigrina are as follows (time in

parentheses is hours and minutes after

fertilization):

I) Oocyte stage. Unfertilized eggs are

spheres 1.4-1.7 mm in diameter (Fig. 26).

(00:25-00:30): The egg membrane absorbs

water and expands. The eggs reach 2.6-

4.0 mm in perivitellic space. (00:50): The

pigment crown of the animal pole extends

downward upon the gray crescent

(Fig. la, lb, lc, Id, le, 26).

© 1990 by Asiatic Herpetological Research

Vol. 3, p. 86

Asiatic Herpetological Research

April 1990

2) 2 cell stage. (01:12): The first

longitudinal cleavage progresses and a

cleavage furrow in the animal pole appears,

forming 2 equal hemispheres (Fig. 2a, 2b,

27).

3) 4 cell stage. (01:33): A cleavage

furrow in the animal pole appears,

perpendicular to the first cleavage furrow.

It progresses to the vegetal pole and forms

four equal sections (Fig. 3a, 3b, 28).

4) 8 cell stage. (01:42): The third

cleavage furrow appears parallel to the

equator, approximately bisecting the animal

hemisphere, forming four small animal

cells and four large vegetal cells (Fig. 4a,

4b, 29).

5) 16 cell stage. (01:48): Observed from

the animal pole, the fourth cleavage forms

eight blastomeres in two circular or

elliptical tiers (Fig. 5a, 5b, 30).

6) 32 cell stage. (02:05): The fifth

cleavage is horizontal. The animal pole has

eight small cells; the vegetal pole has 8

large cells (Fig. 6a, 6b, 31).

7) Early blastula stage. (02:23): The

embryo is at the large cell blastula stage,

but the cells are still clearly distinguished.

The blastocoel begins to appear in the

middle part of the embryo near the animal

pole (Fig. 7a, 7b, 32).

8) Mid-blastula stage. (02:50): There are

many small blastomeres. The blastocoel

continues to amplify (Fig. 8a, 8b, 33).

9. Late blastula stage. (03:18): The

blastomeres are the color of a red bayberry.

The cell boundary becomes indistinct and

the blastocoel expands to its maximum

(Fig. 9a, 9b, 34)

10) Early gastrula stage. (05:56): The

pigmented crown epiboly of the animal cap

occurs, extending to cover over 75% of the

embryo. The dorsal lip forms. Involution

of cells begins (Fig. 10a, 10b, 35).

11) Mid-gastrula stage. (07:00): The

dorsal lip continues to amplify. The cells

on the side of the lip are drawn into the

future archenteron and a semicircle appears

(Fig. 11a, lib, 36).

12) Late gastrula stage. (08:05): The

blastopore shrinks and closes gradually.

Eventually, the yolk plug is enclosed in the

embryo (Fig. 12a, 12b, 12c, 12d, 37)

13) Neural plate stage. (10:30): The

blastopore becomes a fissure, and the

embryo begins to elongate along the

longitudinal axis, becoming pear-shaped

(Fig. 13a,.13b, 38).

14) Neural fold stage stage. (11 :50): A

fold appears on both sides of the neural

plate. The neural groove forms the median

sinus of the embryo. The front of the plate

is bigger, and the neural fold gradually

approaches the dorsal axis from both sides.

The embryo elongates to 1.9 mm (Fig. 14a,

14b, 39).

15) Cilial movement stage. (12:40): The

neural folds are joining. The embryo

rotates within the vitelline membrane (Fig.

15a, 15b, 40).

16) Nerve tube stage. (13:45): The neural

tube has formed and the gill plate and the

cement gland can be seen (Fig. 16a, 16b,

41).

17) Tail bud stage. (14:55): Two outer

gill buds protrude on the side of each gill

plate. The total length is 3.0-3.6 mm, and

the length of the tail bud is 1/10 to 1/7 of

this (Fig. 17a, 17b, 17c, 42).

18) Muscle effect stage. (15:55):

Muscular response begins in most

individuals. The olfactory organ appears,

the cement gland is complete. Total length

is 3.3-3.8 mm., and the length of the tail

bud is W of this (Fig. 18a, 18b, 18c, 43).

19) Hatching stage. (16:45): The embryo

hatches from the egg membrane and two

outer gill budlets with 3-5 branches

protrude obviously (Fig. 19a, 19b, 19c,

19d, 44).

20) Heart beat stage. (29:25): The heart

April 1990

Asiatic Herpetological Research

Vol. 3, p. 87

begins to move, the eyes protrude , the pair

of otoliths can be seen, and the 3rd external

gill matures (Fig. 20a, 20b, 20c, 45).

21) Open mouth stage. (31:10): The

membrane covering the mouth splits to

show the mouth cavity. The alimentary

canal is complete and body segments are

obvious. The inverted "V" shape

myomeres appear on the side of the

embryo. Body length is 5.8-6.2 mm. Tail

fin length is half of the body length (Fig.

21a, 21b, 21c, 46).

22) Tail fin blood circulatory stage.

(32:00): The cement gland begins to

degrade, circulation begins in the tail bud,

and tadpoles are able to swim in a straight

line (Fig. 22a, 22b, 22c, 47).

23) Gill opercular fold stage. (35:55):

The opercular fold appears in the base of

the external gill and the intestines form a

bow (Fig. 23a, 23b, 23c, 48).

24) Right side operculum closed stage.

(55:30): The opercular fold stretches

toward the right side, the right external gill

is covered and forms the right internal gill.

The left external gill is exposed. The

intestines have 2-3 twists. Total length is

6-9 mm (Fig. 24a, 24b, 24c, 49).

25) Operculum completion stage. (70:20):

The external gill in entirely enclosed in the

gill cavity, the posterior of which has an

exhalent pore to the left side. The yolk has

almost been absorbed, and the tadpoles

begin feeding. Total length is 6.8-10.5 mm

(Fig. 25a, 25b, 25c).

Table 1 shows N & F normal table

equivalents (Nieuwkoop and Faber 1967)

of our results.

Discussion

1. The circumstances of the early

embryonic development of Rana tigrina, R.

catesbeiana, R. Umnochiris, and R.

nigromaculata basically identical, but R.

rugulosa has some minor differences in

developmental timing.

TABLE 1. Normal table equivalents

(Nieuwkoop and Faber 1967) of the

developmental stages of Rana rugulosa as listed

in this paper.

2. During early embryonic

development, we raised Rana rugulosa in

water of about 25.1 to 27.1°C. It takes

only 70 hr, 20 min to develop from oocyte

to gill cover stage. Under the same

conditions, R. catesbeiana is 60 to 70 hr

slower, and R. Umnochiris is 80 hr

slower. Speed of development is closely

related to temperature. At 26 to 28.5°C this

developmental period in R. rugulosa takes

only 64 hr, 30 min.

Vol. 3, p. 88

Asiatic Herpetological Research

April 1990

3. Fifty-five minutes after fertilization,

the grey crescent of Rana rugulosa

appears. This is identical to the timing of

R. catesbeiana, but different from that of R.

nigromaculata, in which the grey crescent

can barely be seen, or is not visible.

4. The development of Rana rugulosa

from the eight cell stage to the 16 cell stage

takes six min, while the same period takes

59 min in R. catesbeiana.

5. In Rana rugulosa, external gill

develops 3-5 branches during the period of

hatching, while in R. catesbeiana, the

external gill develops 2-3 branches during

the tail bud circulatory stage. The third

external gill appears sooner in R. rugulosa

than in R. catesbeiana or R. guentheri.

6. Eggs laid by Rana rugulosa adhere

in a pile or on the surface of the water. A

reason for low hatching success is that

oxygen levels can be low in the center of

the egg mass. According to our

observations, eggs at the edge of a mass are

easy to hatch. If we separate the masses

early enough, adequate dispersion can be

achieved. However, the operation must be

a careful one, or the egg membranes will be

broken and the proper development of the

embryo can be affected, causing deformity

and death.

References

CAI, M. 1980. [Hasten parturition and artificial

fertilize method of frogs]. Chinese Journal of

Zoology, Beijing. 2:49-50. (In Chinese).

CHENGDU INSTITUTE OF BIOLOGY 1977. [Key

to Amphibians of China]. Science Press,

Beijing, China. 93 pp. (In Chinese).

GU, S. AND J. LE. 1986. [Animal Embryology].

People's Education Press, Beijing, China. (In

Chinese).

LIU, C. AND S. HU. 1961. [The Tailless

Salientia in China]. Science Press, Beijing,

China. 364 pp. (In Chinese).

NIEUWKOOP, P. D. AND J. FABER. 1967.

Normal table oiXenopus laevis. North Holland

Publishing Co. Amsterdam.

SHANDONG FISHERIES INSTITUTE. 1966. [The

embryonic development of Rana catesbeiana.]

Chinese Journal of Zoology 1966(4): 182-185.

(In Chinese).

ZHANG, J., J. LIU, AND M. CAI. 1985. [A study

on the seasonal variation in the ovary and the

reproductive frequency in Rana limnochiris].

Acta Herpetologica Sinica 1985, 4(4):276-281.

(In Chinese).

April 1990

Asiatic Herpetological Research

Vol. 3, p. 89

Fig. 1 A

I m m

H Fig. 1 B

Fig. ID I m m Fig. 1 E

» 1

Fig. 2 A

Fig. 2 B

Fig. 3 A

Fig. 3 B

FIG. la, lb, lc, Id, le, 2a, 2b, 3a, 3b. Embryonic development of R ana rugulosa. See text for

details.

Vol. 3, p. 90

Asiatic Herpetological Research

April 1990

Fig. 4 A

Fig. 4 B

Fig. 5 A

l 1

Fig. 5 B

Fig. 6 A

Fig. 6 B

Fig. 7 A

Fig. 7 B

Fig. 8 A

FIG. 4a, 4b, 5a, 5b, 6a, 6b, 7a, 7b, 8a. Embryonic development of Rana rugulosa. See text for

details.

April 1990

Asiatic Herpetological Research

Vol. 3, p. 91

Fig. 8 B

Fig. 9 A

Fig. 9 B

I m m

Fig. 10 A

Fig. 10 B

Fig. 1 1 A

Fig. 1 1 B

Fig. 12 A

Fig. 12 B

FIG. 8b, 9a, 9b, 10a, 10b, 11a, lib, 12a, 12b. Embryonic development of Rana rugulosa. See

text for details.

Vol. 3, p. 92

Asiatic Herpetological Research

April 1990

Fig. 12 C

Fig. 12 D

I m m

H 1

Fig. 13 A

Fig. 13 B

Fig. 14 A

Fig. 14 B

FIG. 12c, 12d, 13a, 13b, 14a, 14b. Embryonic development of Rana rugulosa. See text for details.

April 1990

Asiatic Herpetological Research

Vol. 3, p. 93

Fig. 15 A

I m m

I 1

Fig. 15 B

Fig. 16 A

Fig. 16 B

Fig. 17 A

Fig. 17 B

I m m

i <

Fig. 17 C

FIG. 15a, 15b, 16a, 16b, 17a, 17b, 17c. Embryonic development of Rana rugulosa. See text for

details.

Vol. 3, p. 94

Asiatic Herpetological Research

April 1990

Fig. 18 B

Fig. 18 C

2mm

Fig. 19 A i 1 Fig. 19 B

Fig. 19 C

m m

Fig. 19 D

Fig. 20 A

Fig. 20 B

Fig. 20 C

FIG. 18a, 18b, 18c, 19a, 19b, 19c, 19d, 20a, 20b, 20c. Embryonic development of Rana

rugulosa. See text for details.

April 1990

Asiatic Herpetological Research

Vol. 3, p. 95

Fig. 21 B

Fig. 21 A

m m

Fig. 21 C

Fig. 22 A

Fig. 22 B

Fig. 22 C

Fig. 23 B

Fig. 23 C

FIG. 21a, 21b, 21c, 22a, 22b, 22c, 23a, 23b, 23c. Embryonic development of Rana rugulosa.

See text for details.

Vol. 3, p. 96

Asiatic Herpetological Research

April 1990

Fig. 24 A

Fig. 24 B

2mm

i «

Fig. 24 C

Fig. 25 A

Fig. 25 B

Fig. 25 C

FIG. 24a, 24b, 24c, 25a, 25b, 25c. Embryonic development of R ana rugulosa. See text for details.

April 1990

Asiatic Herpetological Research

Vol. 3, p. 97

FIG. 26 (X30)

FIG. 27 (X30)

FIG. 28 (X30)

FIG. 29 (X30)

FIG. 30 (X30)

FIG. 31 (X30)

FIG. 26, 27, 28, 29, 30, 31. Embryonic development of Rana rugulosa. See text for details.

Vol. 3, p. 98

Asiatic Herpetological Research

April 1990

FIG. 32 (X30)

FIG. 33 (X30)

FIG. 34 (X30)

FIG. 35 (X30)

FIG. 36 (X30)

FIG. 37 (X30)

FIG. 32, 33, 34, 35, 36, 37. Embryonic development of Rana rugulosa. See text for details.

April 1990

Asiatic Herpetological Research

Vol. 3, p. 99

FIG. 38 (X30)

FIG. 39 (X30)

FIG. 40 (X30)

FIG. 41 (X30)

FIG. 42 (X30)

FIG. 43 (X20)

FIG. 38, 39, 40, 41, 42, 43. Embryonic development oiRana rugulosa. See text for details.

Vol. 3, p. 100 Asiatic Herpetological Research

April 1990

FIG. 44 (X20)

FIG. 45 (X15)

FIG. 46 (X15)

FIG. 47 (X15)

FIG. 48 (X10)

FIG. 49 (X10)

FIG. 44, 45, 46, 47, 48, 49. Embryonic development oiRana rugulosa. See text for details.

April 1990

Asiatic Herpetological Research

Vol. 3, pp. 101-103

The Validity of Elaphe perlacea, a Rare Endemic Snake from

Sichuan Province, China

ermi Zhao1

^Chengdu Institute of Biology, P.O. Box 416, Academia Sinica, Chengdu, Sichuan, China

Key Words: Reptilia, Serpentes, Colubridae, Elaphe perlacea, China, Sichuan, endemic, taxonomic

validity.

Introduction

Recently Schulz (1989) reported on the

validity of the specific status of Elaphe

perlacea Stejneger, 1929. Schulz (1989)

concluded, "Since data appear to be chiefly

within the range of E. mandarina, E.

perlacea is supposed to be placed in a

subspecies rank of E. mandarina pointing

out that further investigations may reveal

that it is even a variety only." Elaphe

perlacea is an endemic to Sichuan Province,

China. For fifty years after the description

of Elaphe perlacea by Stejneger (1929) no

specimens were found. Many

herpetologists including those of China

doubted its validity.

In the last ten years three additional

specimens have been found in Sichuan

Province (Table 1, 2). Two of these

specimens (80-1, 87-2, and 88-3) were

examined by me. All three specimens have

a dorsal scale formula of 19-19-17, 7 upper

labials (the third and fourth entering the

eye), one preocular, two post oculars, 1+2

temporals, and a divided anal. Both the

scale counts and the patterns are similar to

those of the type. The new material differs

from that of the type in having: 1) Four

lower labials in contact with the anterior

chin shield; 2) Dorsal scales of the male

are smooth, only the posterior 2 to 3 middle

rows are slightly keeled; 3) slight

modification of the dorsal head pattern

(Fig. 1).

Elaphe perlacea differs from E.

mandarina (Cantor) in many ways. E.

mandarina has: 1) 23 scale rows on the

neck and mid-body, 19 or 21 before the

vent; 2) Two anterior temporal scales

(occasionally one); 3) A much different

dorsal pattern. Thus, Elaphe perlacea is a

valid species.

On a field trip during May of 1989 I

went to Hailuo Gou one of the new

localities. On the way we passed through a

small town, Moxi. In the market were

several Elaphe perlacea skins. This

suggests that this snake is common in this

area.

The type locality of Elaphe perlacea is

Yachow Prefecture (Ya'an Prefecture),

Sichuan Province, China. This is an

imprecise locality since a prefecture is quite

large. The two new localities are not in this

prefecture but are close to it. The new

TABLE 1. Elaphe perlacea specimens found in Sichuan recenUy.

Sex, Number Locality

Male

Male, 87-2

Female, 88-3

Female, 80-1

Hailuo Gou, Luding

Hailuo Gou, Luding

Hailuo Gou, Luding

Wolong, Wenchuan

© 1990 by Asiatic Herpetological Research

Vol. 3, p. 102

Asiatic Herpetological Research

April 1990

TABLE 2. Scale counts of Elaphe perlacea Stejneger, 1929.

a b c

FIG. 1. Back of head of Elaphe perlacea, showing the main pattern, a. Male, type specimen, b. Female

from Wolong. c. Male from Hailuo Gou.

FIG. 2. Map of Sichuan showing approximate position of localities.

April 1990 Chinese Herpetological Research Vol. 3, p. 103

localities are: Elev. 2000 and 2500 m, Literature Cited

Hailuo Gou, Luding County, Garze Zang

Autonomous Prefecture, and Elev. 2000 m, SCHULZ, K.-D. 1989. Die hinterasiatischen

Wolong, Wenchuan County, Aba (Ngawa) Kletternattern der Gattung Elaphe Teil XVI

Zang Autonomous Prefecture, Sichuan Elaphe perlacea Stejneger, 1929 Sauria, Berlin-

Province, China (Fig. 2). As presently W., 11(2):15-16.

understood Elaphe perlacea is endemic to

Sichuan Province, China, and only occurs Stejneger, L. 1929. A new snake from China,

in the foot hills of the Himalayan Plateau Proceedings of the Biological Society of

directly west of the Sichuan Basin. Wash.ngton 42:129-130.

[April 1990

Asiatic Herpetological Research

Vol.3, pp. 104-115

Stellio sacra (Smith 1935)

a Distinct Species of Asiatic Rock Agamid

from Tibet*

NATALIA B. ANANJEVA1, GUNTHER PETERS2, J. ROBERT MACEY3

AND THEODORE J. PAPENFUSS3

^Zoological Institute, USSR Academy of Sciences, Leningrad, USSR

^Museum fur Naturkunde der Humboldt-Universitdt, Invalidenstrasse 43, Berlin, DDR

^Museum of Vertebrate Zoology, University of California, Berkeley, California, USA

Abstract. -An examination of the type series of Agama himalayana sacra Smith (1935) and new

material collected in Tibet in 1988 has shown that this form should be considered a distinct species. In

following the recent revision of the genus Agama (Moody 1980), Stellio sacra is included in the genus

with all other Asiatic rock agamids.

Key Words: Reptilia, Sauria, Agamidae, Stellio sacra, Tibet, systematics, distribution.

Introduction

Four agamid lizards were collected at the

beginning of the 20th Century near Lhasa,

Tibet. They were described in a short

account by Smith (1935) as a subspecies of

the Himalayan rock agamid Agama

himalayana sacra.

The examination of these type specimens

and new material collected in Lhasa and the

vicinity of Lhasa, Tibet in 1988 by the third

and fourth authors, led us to the conclusion

that it is not a subspecies of Stellio

himalayanus. Furthermore it is not a

member of the Stellio himalayanus species

complex which includes Stellio

badakhshanus, Stellio chernovi, Stellio

himalayanus, and Stellio stoliczkanus. The

results of the examination of the agamids

from Lhasa, Tibet confirm the opinion of

Ananjeva et al. (1981) that Stellio sacra

should be elevated to full specific status.

According to the present revision of the

genus Agama (Wermuth 1967) into Agama,

Stellio (Moody 1980; Sokolovsky 1975,

1977) Trapelus, Psuedotrapelus, and

Xenagama (Moody 1980) the lizards

examined from Tibet should have the

generic name, Stellio.

* Sino-Soviet- American Arid Asian Desert Regions

Research Paper no. 5.

Methods

To the best of our knowledge all

specimens housed outside the Peoples

Republic of China were examined. Of the

four specimens in the type series, one

lectotype and two paralectotypes are housed

at the British Museum (Natural History)

[BMNH], and one paralectotype is housed

at the Indian Museum, Calcutta (Zoological

Survey of India [ZSI]). Thirteen additional

specimens were collected during the 1988

joint Chengdu Institute of Biology -

University of California Expedition and are

housed at the California Academy of

Sciences (CAS).

Stellio sacra

(Smith 1935), new combination

Known Material

Material Examined:

Lectotype: BMNH 1946.8.28.57

(formerly 1904.12.28.1) [Fig. 1.]

Locality: Near Lhasa, Tibet.

Paralectotypes: BMNH 1946.8. 28.58,

BMNH 1946.8.28.59 [Fig. 2], and ZSI

15740 [Fig. 3]. Locality: near Lhasa,

Tibet.

CAS 170545. Locality: Elev. 3740 m,

Sera Monastery, Lhasa (29° 39' N 91°

fr\ 1 fifin u.,

April 1990

Asiatic Herpetological Research

Vol. 3, p. 105

FIG. 1. Lectotype of Stellio sacra BMNH 1946.8.28.57 (formerly 1904.12.28.1).

Sera Monastery, Lhasa (29° 39' N 91°

06' E), Lhasa Municipality, Xizang (Tibet)

Autonomous Region, China. Collected by:

T. J. Papenfuss and R. Macey. Date: 24

Sept., 1988.

CAS 170546-53. Locality: Elev.

3700 m, at base of mountains approx. 3

km WNW (airline) of the Potala Palace,

Lhasa (29° 39' N 91° 06' E), Lhasa

Municipality, Xizang (Tibet) Autonomous

Region, China. Collected by: CAS

170546-49 R. Macey and T. J. Papenfuss,

and CAS 170550-53 T. J. Papenfuss and

R. Macey. Date: 25 Sept., 1988.

CAS 170554-57. Locality: Elev. 3990

m, 52.4 km south of Yangbajan (30° 13' N

90° 25' E), also at km 1900.8 from

Xining, on the Xining-Golmud-Lhasa Rd.,

Lhasa Municipality, Xizang (Tibet)

Autonomous Region, China. Collected by:

CAS 170554-55 R. Macey and T. J.

Papenfuss, CAS 170556-57 T. J.

Papenfuss and R. Macey. Date: 27 Sept.,

1988. (See Table 1).

Other Material.

Note that the following material is

reported on in Hu et al. (1987) but no

numbers or reference to a museum

collection are mentioned. It is probable that

all or most of these specimens are housed at

the Chengdu Institute of Biology.

Three males, 3 females, and 3 juveniles.

Locality: Bomi (29° 50' N 95° 45' E),

Qamdo Prefecture, Xizang (Tibet)

Autonomous Region, China.

Six males, and 12 females. Locality:

Lhasa (29° 39' N 91° 06' E), Lhasa

Municipality, Xizang (Tibet) Autonomous

Region, China.

Vol. 3, p. 106

Asiatic Herpetological Research

April 1990

FIG. 2. Paralectotypes oiStellio sacra BMNH 1946.8. 28.58 and BMNH 1946.8.28.59.

Two males, 2 females, and 1 juvenile.

Locality: Nyingchi (29° 32' N 94°

25' E), Lhasa Municipality, Xizang (Tibet)

Autonomous Region, China.

Distribution

Stellio sacra as presently understood, is

restricted to the river drainage of the

Yarlung Zangbo in the Lhasa Valley,

Xizang (Tibet) Autonomous Region, China

(Fig. 4). Only four localities are known,

all between 3000 and 4000 m.

Populations occurring in the Kunlun

Mountains of southern Xinjiang Uygur

Autonomous Region, China were

previously assigned to Stellio himalayanus

himalayanus, however their present

taxonomic position is uncertain.

Diagnosis: Rock agamid with flattened

head and body which is typical for this

lizard group. They are comparatively large

lizards with a snout-vent length of 120-

150 mm and a tail length of 180-240 mm

(Table 1).

Gular Sac seems to be developed to a

greater degree than in other Stellio. Body

scales are small and granular. The scales

are not well differentiated. There is a very

slight but noticeable nuchal crest on the

head. It begins from the middle of the

occiput and continues as a poorly

differentiated vertebral stripe. The

longitudinal rows of enlarged and feebly

keeled scales on the vertebral region are

arranged parallel to each other. There are

neither groups of enlarged scales nor

separate enlarged scales on the dorsal lateral

regions.

The males have a large patch of callous

scales on the belly. The annuli and

segmentation of the scales on the basal

quarter of the tail are not prominent. On the

April 1990

Asiatic Herpetological Research

Vol. 3, p. 107

\

/

FIG. 3. Paralectotype of Stellio sacra ZSI 15740.

lateral surface of the tail there are three to

four annuli in each segment.

There is a small granular dark pattern on

the back. The center of the back tends to

have more black and toward the sides a

dark golden brown dominates. The

separate elements of this pattern are

connected to heavily marked diffuse

transverse stripes. The narrow stripes form

two rows of the dark colored scales that

continue from the neck to the tail. Overall

the lizard is darkly colored but there are a

few randomly scattered yellow blotches on

the back (Fig. 5). Juveniles are lighter in

color tending more toward a dark golden

brown with darker speckling all over the

back. The dark golden brown forms bands

across the back which are offset at the

spine.

Comparative Description

Stellio sacra differs from Stellio

himalayanus and related forms in all the

diagnosis characters. Stellio sacra is only

similar in body size to Stellio stoliczkanus

among the species examined. Other species

are notably smaller (Table 1; Anderson and

Leviton 1969; Ananjeva et al. 1981).

The body proportion data concerning

Stellio sacra and the Stellio himalayanus

species complex indicates small differences

Vol. 3, p. 108

Asiatic Herpetological Research

April 1990

TABLE 1 . Characters of Stellio sacra specimens used for analysis.

(Table 2). Stellio sacra has a shorter tail

than Stellio badakhshanus, Stellio

chernovi, and Stellio himalayanus.

One of the paralectotypes, a female

(BMNH 1946.8.28.58), has a regenerated

tail. The length of the regenerated tail after

autotomy is about 80 mm which is rather

great. The scales of the tail are mucronate,

and the regular annular arrangement in the

regenerated portion of the tail is disturbed.

Stellio sacra has a more similar head

height index to Stellio erythrogaster and

Stellio lehmanni than to Stellio caucasius

and Stellio himalayanus (Ananjeva 1981);

i.e. the head is not so flat (depressed).

It is necessary to also remark about the

differences in the structure of the digits in

Stellio sacra and that of the specimens

examined of Stellio caucasius, Stellio

lehmanni, and Stellio stoliczkanus of the

same size. The digits, and especially the

terminal phalanges, of Stellio sacra

specimens are not round in their section as

in the above group of species, but instead

are compressed laterally.

The distinctive character of Stellio sacra

pholidosis is its comparative homogeneity.

The scales on the back, sides and belly are

small. It has a considerably large scale

count around the mid-body of 239 scales.

The mid-body scale count of Stellio

badakhshanus, Stellio chernovi, Stellio

himalayanus, and Stellio stoliczkanus all

do not exceed 180 scales. Such a large

mid-body scale count is only observed in

Stellio nuristanicus, which has from 230-

248 mid-body scales. During this phase of

our study two of the type specimens

(Anderson and Leviton 1969) were

examined.

The longitudinal rows of the enlarged

poorly keeled scales are not arranged as the

obvious "dorsal stripe," as typically

observed in many other species of Stellio.

The dorsal scales gradually decrease in size

and degree of keeling, from the center of

the back toward the small dorso-ventral

scales. The ventral scales are smaller than

the dorsal scales. The posterior margins of

these scales slightly overlap the scales of

the following row.

April 1990

Asiatic Herpetological Research

Vol. 3, p. 109

FIG. 4. Distribution of Stellio himalayanus and Stellio sacra. Question mark indicates Xinjiang

Autonomous Region population of questionable taxonomic status.

The ventral head scales are also small

and not regular. The size of these scales

decreases from the nasal end toward the

gular fold. On the sides of the head and on

the neck are situated small groups of very

strongly keeled spinose scales.

Stellio sacra has nostrils shaped

longitudinally oval, similar to Stellio

chernovi and Stellio himalayanus. Where

as Stellio agrorensis, Stellio melanurus,

Stellio nuristanicus and Stellio tuberculatus

have round nostrils. Stellio sacra has a

round tympanum.

In Stellio sacra, the scales of the upper

surface of the fore and hind limbs are

comparatively small, but their size is similar

to the largest dorsal scales. These scales

are strongly keeled. Single slightly

enlarged scales are distinguishable from the

surrounding small scales. The scales on

the base of the tail are keeled and

mucronate, but to a smaller degree than for

example in Stellio himalayanus.

The results of the comparative study of

Asiatic rock agamids enables the typical

distinguishing characters of Stellio sacra to

be identified:

1 . The small size of the scales.

2. The presence of a short nuchal crest

made up of from two to three scale rows of

slightly enlarged, narrow, keeled, dark

colored scales. Such a nuchal crest is

absent in Stellio chernovi, Stellio

himalayanus, Stellio stoliczkanus and other

species. A nuchal crest is noted however in

Stellio melanurus, Stellio nuptus, Stellio

Stellio, and Stellio tuberculatus.

Vol. 3, p. 110

Asiatic Herpetological Research

April 1990

,- «

FIG. 5. Stellio sacra from elevation 3990 m, 52.4 km south of Yangbajan (30° 13' N 90° 25' E),

Xizang (Tibet) Autonomous Region, China.

3. The ends of the occipital (nuchal) and

neck scales are directed backwards but not

foreword; the parietal scales are oriented in

a laterocaudal direction.

4. The rows of the enlarged dorsal

scales are parallel but do not meet as in

species of the "Stellio himalayanus "

complex.

TABLE 2. Comparative data on the relative proportions of the tail, limbs, and head of rock agamids of the

genus Stellio (5. sacra and agamids of the 5. himalayanus complex).

April 1990

Asiatic Herpetological Research

Vol.3, p. Ill

FIG. 6. Habitat of Stellio sacra, elevation 3990 m, 52.4 km south of Yangbajan (30° 13' N 90° 25' E),

Xizang (Tibet) Autonomous Region, China.

5. In the region over the shoulders

there are two arched diffuse rows of

slightly enlarged mucronate (spinose)

scales.

6. The absence of enlarged scales or

regions of enlarged scales on the back of

the body.

7. The upper head scales distributed

between the nasal scales have a stretched

shape. The length of these scales is twice

the width as also in Stellio agrorensis,

Stellio nuristanicus, and Stellio tuberculatus

where as the scales of Stellio chernovi and

Stellio himalayanus have a roundish,

polygonal shape.

8. Only a single row of small scales is

noted between the suboculars and the

supralabials. Other species have two to

five scale rows.

9. The scales of the back, shoulders,

thighs and some parietal scales have an

unusual microstructure. The margins of the

scales are jagged resembling that of fringes.

This character was also observed in Stellio

annectens, Stellio erythrogaster, Stellio

melanurus, and Stellio nuptus.

10. There is a patch of callous

(granular) scales in the middle of the belly,

which is large in males.

1 1 . The males have a large patch of anal

pores and from six to seven rows of callous

scales before the cloaca.

Coloration

The dorsal coloration of the type

Vol. 3, p. 112

Asiatic Herpetological Research

April 1990

specimens consists of diffuse components,

each scale is either totally dark or light in

coloration. The parietal region and along

the middle of the back (two scale rows of

absolutely dark scales) are more dark in

coloration. From the middle of the back to

the sides there are light brown transverse

stripes. This coloration is poorly

developed in some specimens.

The subadult specimen has a similar

small speckled pattern and in addition has

alternating light and dark transverse stripes

coming from the vertebral ridge toward the

sides of the body. This pattern is similar to

that of juvenile specimens of Stellio

tuberculatus.

As far as we can tell from the available

specimens of Stellio sacra, no contrasting

pattern and coloration on the surface of the

gular region, neck and nuchal region are

noted. Such a contrasting pattern and

coloration are typical for specimens of

Stellio himalayanus and related species.

Natural History

Stellio sacra are common in the rocky

hills surrounding the Lhasa Valley. They

were seen only on slopes covered with

large boulders (Fig. 6). Often a single

adult male occupied a pile of boulders in

association with several females and

juveniles. Stellio sacra is an agile, alert

species that is difficult to approach closely.

The only other reptile that occurs on the

rocky slopes is the gecko, Cyrtodactylus

tibetanus. Another agamid,

Phrynocephalus theobaldi is abundant on

the sandy soil at the base of the rocky hills.

Discussion

The large review of the rich Indian fauna

that Smith (1935) dealt with explains why

the Sacred Agamid, Stellio sacra, was

originally described as a subspecies of the

Himalayan agamid, Stellio himalayanus.

The study presented here concludes that

Stellio sacra is not related to the Stellio

himalayanus complex. On the other hand

the above descriptions of the morphological

characters do not allow an interpretation of

the direct relation of Stellio sacra to such

species as Stellio agrorensis, Stellio

melanurus, Stellio nuptus, and Stellio

tuberculatus. The characters which enable

us to bring together Stellio sacra and these

species should be considered as

plesiomorphic similarities. These

characters are long limbs, the presence of a

small gular sac and nuchal crest, the

polyannular structure of the caudal

segments, the juvenile color pattern, and so

on. This assumption about a plesiomorphic

condition is based on the idea that the

oligomerization (decreasing of number of

elements) of homologic organs (here

pholidosis elements) is one of the main

directions of agamid evolution (Dogel

1954).

Characters, such as the orientation of the

scale axes in nuchal and neck regions and

in parietal scales is also of great interest. It

is suggested (Peters unpublished data) that

the more common condition of the scale,

i.e. the orientation of the point backwards,

is typical not only for most of the species in

the genus Stellio (including Stellio sacra ),

but also for the majority of agamids and

lizards as a whole. Hence it is a

plesiomorphic condition. The contrary

direction, where the point of the scale is

forwards, is observed in Stellio melanurus,

Stellio nuptus and a number of African

species. In accordance with Peter's point

of view, this should be considered as an

apomorphic condition. However there is a

problem in interpreting such a character as

the presence of the scales with the jagged

margins. This scale structure is found in

species with both types of scale orientation,

including Stellio sacra as well as Stellio

erythrogaster, Stellio melanurus, Stellio

nuptus and the African species, Stellio

annectens. The interpretation of such

characters makes all possible explanations

of the relationships of these species

controversial.

These facts and the inability at this time

to determine even a single obvious

synapomorphy for all Asiatic rock agamids

of the genus Stellio (with or without Stellio

melanurus and Stellio nuptus ) does not

allow an acceptable phylogenetic

relationship to be developed. This problem

April 1990

Asiatic Herpetological Research

Vol. 3, p. 113

TABLE 3. The distribution of the species in the genus Stellio.

Species

Distribution

Stellio adramitana (Anderson 18%)

Stellio agrorensis (Stoliczka 1872)

Stellio annectens (Blanford 1870)

Stellio atricollis (Smith 1849)

Stellio badakhshanus (Anderson and Leviton 1969)

Stellio caucasius (Eichwald 1831)

Stellio chernovi (Ananjeva, Peters,

and Rzepakovsky 1981)

Stellio cyanogaster (Ruppell 1835)

Stellio erythrogaster (Nikolsky 1896)

Stellio himalayanus (Steindachner 1869)

Stellio lehmanni (Nikolsky 1896)

Stellio melanurus (Blyth 1854)

Stellio microlepis (Blanford 1874)

Stellio nuptus (De Filippi 1843)

Stellio nuristanicus (Anderson and Leviton 1969)

Stellio phillipsii (Boulenger 1895)

Stellio sacra (Smith 1935)

Stellio stellio (Linnaeus 1758)

Stellio stoliczkanus (Blanford 1875)

Stellio trachypleurus (Peters 1982)

Stellio tuberculatum (Hardwicke and Gray 1827)

Stellio yemenensis (Klauzewitz 1954)

Stellio zonorus (Boulenger 1895)

Arabia

Afghanistan, Pakistan, India

Africa (Somalia, Ethiopia)

Eastern and southern Africa

Afghanistan

Caucasus, Tadjikistan, Turkmania, Turkey,

Iran, Afghanistan, Pakistan

Tadjikistan, Turkmania, Uzbekistan

Iraq,

Somalia, Ethiopia

Iran, Turkmania

Tadjikistan, Uzbekistan, Afghanistan, Pakistan,

India

Tadjikistan, Turkmania, Uzbekistan, Afghanistan

Iran, Pakistan

Iran

Iraq, Iran, Afghanistan, Pakistan

Afghanistan

Ethiopia

Tibet

Greece, southwest Asia, northern Egypt

Mongolia, China

Ethiopia

India, Nepal, Afghanistan, Pakistan

Arabia

Somalia

may be explained not only by the small

sample size of Stellio sacra and the fact that

many other forms are poorly studied but

also by the obvious existence of parallel

trends in different developmental lines

within this lizard group.

In order to arrive at a more trust worthy

hypothesis of the relationships of rock

agamids, it is necessary to carry out

biochemical investigations. The first

preliminary results of such investigations

are that of Joger and Arano (1987) and

Ananjeva and Sokolova (in prep.). This

type of data will be highly interesting to

compare with the results of comparative

morphological studies.

The distributional patterns of Asiatic

rock agamids of the genus Stellio (Table 3)

and their chorological isolation seems to

support the idea that Stellio is a

monophyletic group. Rock agamids of the

genus Stellio are distributed from Greece

and the Nile River Delta in the west to the

Gobi Altai of southern Mongolia, and the

western deserts of China in the east. In the

western part of this region, Greece to the

Nile River Delta, Stellio stellio is found.

The southern portion of this region from

southwestern Iran to Pakistan Stellio

melanurus and Stellio nuptus occur.

These species are probably of African

origin (Peters unpublished). In southern

Iran and Afghanistan along with Stellio

melanurus and Stellio nuptus also occur

Stellio agrorensis, Stellio badakhshanus,

Stellio caucasius, Stellio erythrogaster,

Stellio himalayanus, Stellio microlepis,

and Stellio nuristanicus.

The Asiatic rock agamids of the genus

Stellio sensu stricto are absent from the

mountainous regions of western Indostan

Vol. 3, p. 114

Asiatic Herpetological Research

April 1990

and the eastern portion of the Arabian

Peninsula. They are also not found across

the Red Sea in Ethiopia and Somalia.

These areas are poorly studied

herpetologically and it is possible that they

do in fact occur in these regions.

The distribution of Asiatic rock agamids

seems to be a single unit. Most problematic

is the origin of such species as Stellio

melanurus, Stellio nuptus, and Stellio

stellio, which may be of African origin as it

was already mentioned above.

It is possible that this assumption will be

corroborated during further research in

biochemical phylogeny. The preliminary

results of Joger and Arano (1987) on

Stellio stellio is a first step. The origin of

the genus Stellio is interesting and the data

about the early divergence of Stellio in

Stellio sensu stricto (Asiatic species) and

the Stellio atricollis species group from

Africa and southern Arabia is useful (Joger

and Arano 1987). It is possible that Stellio

is a paraphyletic group of species. The

Arabian - African species group is Stellio

adramitana, Stellio annectens, Stellio

atricollis, Stellio cyanogaster, Stellio

phillipsi, Stellio trachypleurus, Stellio

yemenensis, and Stellio zonurus (Table

3). These species are similar to the Asiatic

species Stellio melanurus and Stellio

nuptus in a number of characters. This

allows an assumption that they are related.

Acknowledgments

We are most greatful for the loan of type

material of Stellio sacra from Dr. A. C. G.

Grandison, Dr. E. H. Arnold and Dr. A. F.

Stimson of the British Museum of Natural

History, London, Great Britain, and Dr.

Shara of the Indian Museum, Calcutta,

India.

Literature Cited

ANANJEVA, N. B. 1981. [The peculiarities of the

skull structures, dental system, and hyoid

apparatus of the lizards of Agama genus of the

USSR fauna]. Proceedings of the Zoological

Institute of the USSR Academy of Sciences,

Leningrad 101:3-20. (In Russian).

ANANJEVA, N. B., G. PETERS, AND V. T.

RZEPAKOWSKY. 1981. [New species of rock

agamid from Tadjikistan Agama chernovi sp.

nov.]. Proceedings of the Zoological Institute of

the USSR Academy of Sciences, Leningrad

101:23-27. (In Russian).

ANDERSON, S. T. AND A. E. LEVITON. 1969.

Amphibians and reptiles collected by the Street

Expedition to Afghanistan, 1965. Proceedings of

the California Academy of Sciences, series 4.

37(2):25-56.

DOGEL, V. A. 1954. [Oligomerization of the

homological organs as one of the main

tendencies of animal evolution]. Leningrad

State University. 368 pp. (In Russian).

HU, S., E. ZHAO, Y. JIANG, L. FEI, C. YE, Q.

HU, Q. HUANG, Y. HUANG, AND W. TIAN.

(incorrectly stated in English as Hu, S, L. Fei,

Q. Hu, Q. Huang, Y. Huang, Y. Jiang, W.

Tian, C. Ye, E. Zhao). 1987. Amphibia-

Reptilia of Xizang. The series of the scientific

expedition to the Qinghai-Xizang Plateau. The

comprehensive scientific expedition to Qinghai-

Xizang Plateau, Academia Sinica. Chengdu

Institute of Biology, Academi (sic.) Sinica. 153

pp. (If Chinese).

JOGER, U. AND B. ARANO. 1987. Biochemical

phylogeny of the Agama genus group. Pp. 215-

218. In Van Gelder, Strijbosch.and Bergers

(eds.). Proceedings of the 4th Ordinary General

Meeting of the Society of European

Herpetologists, Nijmegen.

MOODY, S. M. 1980. Phylogenetic and historical

biogeographical relationships of the genera in the

Agamidae (Reptilia: Lacertilia). Ph.D. Thesis.

University of Michigan. 373 pp.

PETERS, C. 1982. Eine neue Wirtelschwanzagame

aus Ostafrika (Agamidae: Agama ).

Mitteilungen aus der Zoologishen Sammlung des

Museums fur Naturkunde in Berlin 58(2):265-

268.

SMITH, M. A. 1935. The fauna of British India,

including Ceylon and Burma. Reptilia and

Amphibia. Vol. II. Sauria. Taylor and Francis,

London. 440 pp.

SOKOLOVSKY, V. V. 1975. [Comparative

karyological study of the lizards from the family

Agamidae. II. Karyotypes of five species of the

genus Agama ]. Cytologica 16(l):91-93. (In

Russian).

April 1990 Asiatic Herpetological Research Vol. 3, p. 115

SOKOLOVSKY, V. V. 1977. [Taxonomic

relationships in the family Agamidae according

to karyological data]. Soviet Herpetological

Conference, Leningrad. [English abstract, in

Russian].

WERMUTH, H. 1967. Liste der rezenten

Amphibien und Reptilien: Agamidae. Das

Tierreich 86, Berlin. 127 pp.

I April 1990"

Asiatic Herpetological Research

Vol. 3, pp. 116-119|

Isolation and Amino Acid Sequence of a New Dodecapeptide from the Skin

of Oreolalax pingii +

YIQUAN TANG1, SHENGHAI TlAN1, SHIXIANG WU1, JlACHENG HUA1, XlNQUAN Jl1,

GUANFU WU2, ERMI ZHAO2, AND GANG ZOU1

^Shanghai Institute of Materia Medica, Academia Sinica, Shanghai, China

2Chengdu Institute of Biology, P.O. Box 416, Academia Sinica, Chengdu, China

Abstract. -A novel dodecapeptide has been isolated by alumina column chromatography and HPLC from

methanol extracts of the skin of the Chinese frog Oreolalax pingii. The sequence of the peptide is: Gly-

Leu-Val-Ser-Asp-Leu-Met-Tyr-Gly-Ile-Gly-Leu-NH2 Th's peptide differs from all the other amphibian skin

peptides and should be regarded as a member of a new peptide family.

Key Words: Amphibia, Anura, Pelobatidae, Oreolalax pingii, China, biochemistry, peptides.

Introduction

Many active peptides have been

discovered from amphibian skin during the

past 20 years or more. Amphibian skin

peptides have proved to be of considerable

value not only in pharmacology, but also in

taxonomic and evolutionary domains

(Erspamer 1984; Cei 1985; Lazarus et al.

1985). In order to discover new active

peptides, we have carried out research on

amphibian skin peptides from Chinese

frogs since 1983 (Hua et al. 1985; Tang et

al. 1985). This paper concerns another

novel dodecapeptide obtained from the skin

of Oreolalax pingii.

Methods

The materials and experimental

procedures were previously reported (Tang

et al. 1985), with the following brief

mentions and additions.

Six hundred specimens of Oreolalax

pingii were collected in May, 1983 from the

Daliangshan of Sichuan Province, China.

The fresh skins (400 g) were removed and

extracted with methanol. The methanol

extracts were evaporated until dry. The

residue was dissolved in 95% ethanol and

the solution distributed on the alumina

* This publication combines material previously

published in Chinese by Tang et al (1985) with

additional information.

columns. The column was eluted with

ethanol-water mixtures of descending

concentrations of ethanol (also see

Montecucchi et al. 1981)

HPLC was performed on a Waters

HPLC system. Details of individual

chromatographic procedure are shown in

the figures. Amino acid analysis of

peptides after hydrolysis in HC1 were

carried out on a LKB 4400 amino acid

analyzer. Sequence analysis of the peptides

were performed by manual DABITC/PITC

procedure (Chang 1983). The complete

sequence analysis of the dodecapeptide was

carried out on an Applied Biosystems

Model 470A gas phase sequencer.

Enzymatic digestions of the peptide with

a-Chymotrypsin, carboxypeptidase A (CP-

A) and Y (CP-Y) were also as before (Tang

et al. 1985). Bioassays of each isolated

product were tested on the longitudinal

muscle myenteric plexus preparation of the

guinea pig ileum (GPI),

Results

The water portion eluted from alumina

columns was lyophilized to give 214 mg of

residue. The residue was separated by

HPLC semi-preparatively as in Fig. 1.

Perks 32 and 33 each were single peak by

verifying on HPLC (analytical uBondapak

Cj8 column, the elution conditions were the

same as in Fig. 1), respectively. The

ir»r»r\ i_.

April 1990

Asiatic Herpetological Research

Vol. 3, p. 117

100

o

CM

C\J

<

40 60

time(min)

FIG. 1. RP-HPLC of the water eluate from the alumina columns. Column: uBondapak C\% 7.8X300

mm. Mobile phase: A=0.05% CF3COOH B=60% CH3CN in A. Concave gradient elution from 0-85%B

(dotted line), 90 min, at 1.0 ml/min. Detected at UV 220 nm, 0.4 aufs.

Gly-Leu-Val-Ser-Asp-Leu-Met-Tyr-Gly-Ile-Gly-Leu-NH2

■GH-1-

A h

■CH-2-

FlG. 2. Profile of sequence anlaysis of the dodecapeptide. (-*): DABITC/PITC method. (-»): gas

phase sequencing. («-): CP-Y digestion. CH: a-chymotryptic peptides.

difference of retention time of the two

peaks was 0.8 min. The peak 32 acted as

the representative of the dodecapeptide for

further studies and the peak 33 was also

investigated simultaneously.

Amino acid composition of the peak 32

was Asp (1), Ser (l),Gly (3), Val (1), Met

(1), He (1), Leu (3), Tyr (1). Amino acid

sequence of the peak 32 was determined by

the DABITC/PITC method and gas phase

sequencing. The former method proceeded

to the tenth step, but the latter to the

penultimate residue (Fig. 2). For C-

terminal residue analysis, CP-A and CP-Y

digestions of the peak 32 were carried out,

and did not release any amino acids by the

former. This indicated a blocked C-

terminus; upon latter, however, Leu and

Gly were obtained. Thus, we deduced that

the C-terminal structure of the peak 32 is

Leu-NH2. Digestion of the peptide by a-

chymotrypsin provided further structural

confirmation. The fragment peptides, CH-

1 and CH-2 were separated on HPLC as

depicted in Fig. 3. Amino acid

compositions of the two fragments are in

accordance with their sequences (see

Fig. 2), respectively.

From the above results the complete

amino acid sequence of the peak 32 was

Vol. 3, p. 118

Asiatic Herpetological Research

April 1990

0.2-r

100

10

20

30

time(min)

40

50

60

FIG. 3. RP-HPLC of a-chymotryptic digests of the dodecapeptide. Column: uBondapak Cis 3.9X300

mm. Mobile phase: A=0.1% CF3COOH B=60% CH3CN in A. Linear gradient elution from 0-60%B ,

40 min, at 0.6 ml/min. Detected at UV 220 nm, 0.2 aufs. Peak (*) is the unreacted dodecapeptide.

established as in Fig. 2.

The peak 33 had the same properties

with the peak 32 in the amino acid

compositions and sequence analysis,

respectively. We do not know the

structural differences between the two

peaks.

Based on the amino acid analysis, the

yield of the pure dodecapeptide (including

peak 32 and 33 in Fig. 1) was estimated to

be at least 1.0 nmol starting from 1.0 g of

fresh skin, according to that the rate of

recovery of all above isolation steps was

10%. The dodecapeptide was inactive in

GPI test, and its activity awaits to be

established by assay methods other than

those used in the present screening.

Discussion

The dodecapeptide described above is a

completely novel peptide. The peptide may

be a member of a new peptide family, and it

should be regarded as the biochemical

characteristic of Oreolalax pingii in

taxonomy. Furthermore, the C-terminal

portion of the dodecapeptide has some

homologies to the mammalian Leu-

enkephalin (Hughes et al. 1975) as shown

below:

The dodecapeptide: Gly-Leu-Val-Ser-Asp-

Leu-Met-Try-Gly-Ile-Gly-Leu-NH2

Leu-enkephalin: Try-Gly-Gly-Phe-Leu

In amphibian skin peptide research, it

April 1990

Asiatic Herpetological Research

Vol. 3, p. 119

was observed that one peptide can be

separated by HPLC into two components—

the peak splitting, such as that discovered

in HPLC separation of PGL-a (Andreu et

al. 1985) and ranamargatin (Tang et al.

1988). The peak 32 and 33 (Fig. 1) both

have the same amino acid composition and

sequence. Therefore, they may be

produced by one peptide in HPLC

separation. The reason of the peak splitting

mentioned above, however, is presently not

known. To our knowledge, there are no

similar findings besides the amphibian skin

peptides. Hence, it is necessary to

understand if the peak splitting is unique to

the amphibian skin peptides.

Acknowledgments

We thank Mr. C. Chen for amino acid

analysis.

Literature Cited

ANDREU, D., H. ASCHAUER, G. KREIL, AND R.

B. MERRIFIELD 1985. Solid-phase synthesis

of PYLa and isolation of its natural counterpart,

PGLa[PYLa-(4-24) peptide amide] from skin

secretion of Xenopus laevis. European Journal

of Biochemistry 149:531-535.

CEI, J.M. 1985. Taxonomic and evolutionary

significance of peptides in amphibian skin.

Peptides 6(Suppl. 3): 13-16.

CHANG, J. 1983. Manual micro-sequence analysis

of polypeptides using dimethylaminoazobenzene

isothiocyanate. Methods in Enzymology

91:455-466.

ERSPAMER, V. 1984. Half a century of

comparative researchs on biogenic amines and

active peptides in amphibian skin and molluscan

tissues. Comparative Biochemistry and

Physiology 79C: 1-7.

HUA, J., S. WU, Y. TANG, W. ZHANG, AND G.

ZOU 1985 [Isolation and characterization of

bradykinin and its two fragments from the skin

of the Chinese frog Rana nigromaculata ]. Acta

Biochemica Biophysica Sinica 17:171-173. (In

Chinese).

HUGHES, J., T. W. SMITH, H. W. KOSTERLITZ,

L. A. FOTHERGILL, B. A. MORGAN, AND H.

R. MORRIS. 1975. Identification of two

related pentapeptides from the brain with potent

opiate agonist activity. Nature (London)

258:577-579.

LAZARUS, L.H., W. E. WILSON, G. GAUDINO, B.

J. IRONS, AND A. GUGLIETTA 1985.

Evolutionary relationships between non-

mammalian and mammalian peptides. Peptides

6(Suppl. 3):295-307.

MONTECUCCHI, P.C., R. DE CASTIGLIONE, S.

PIANI, L. GOZZIN, AND V. ERSPAMER 1981.

Amino acid composition and sequence of

dermorphin, a novel opiate-like peptide from the

skin of Phyllomedusa sauvagei. International

Journal of Peptide Protein Research 17:275-284.

TANG, Y., S. TIAN, S. WU, J. HUA, G. HU, X.

JI, G. ZOU, G. WU, AND E. ZHAO. 1985.

[Separation and structure of a novel hexapeptide

obtained from the skin of Oreolalax pingii].

Acta Herpetologica Sinica 1985, 4(2):99-102.

(In Chinese).

TANG, Y., S. TIAN, S. WU, J. HUA, G. WU, E.

ZHAO, Y. LU, Y. ZHU, AND G. ZOU. 1988.

[Isolation and structure of ranamargarin, a new

tachykinin from the skin of Chinese frog Rana

margaratae ]. Scienua Sinica (B) 9:967-974. (In

Chinese).

I April 1990

Asiatic Herpetological Research

Vol.3, pp. 120-122

The Validity of Sacalia quadriocellata

JINZHONG FU1 AND ERMI ZHAO2

^Zoological Institute, Academia Sinica, Beijing, China

^Chengdu Institute of Biology, Academia Sinica, P. 0. Box 416, Chengdu, Sichuan, China

Key Words: Reptilia, Testudines, Emydidae, Sacalia quadriocellata, China, taxonomic validity.

Introduction

Sacalia quadriocellata was first

described by Siebenzock in 1903 as

Clemmys bealei quadriocellata. Pope

(1935) described Clemmys quadriocellata

from Hainan, and compared it with

Clemmys bealei. But most people except

Pope thought that S. quadriocellata was a

synonym of S. bealei. Sachsse (1984a)

held that it was the female of S. bealei.

Methods

Specimens of Sacalia quadriocellata

from Hainan and Guangxi provinces, along

with specimens of S. bealei from Hainan,

Fujian and Anhui provinces were

examined. Data were taken on external,

skull, and shell suture characters. Sexual

differences were also examined.

Results

External differences

1) The dorsal surface of the head is

uniform olive or chocolate brown in Sacalia

quadriocellata , but it is vermiculated with

black in S. bealei.

2) Sacalia quadriocellata have two ocelli

on each side of the dorsal surface of the

head. The ocelli always have distinct

boundaries and there is one black spot

within each ocellus (Fig. 1). In contrast,

S. bealei have one or two ocelli on each

side of the dorsal surface of the head

(Fig. 1). If two ocelli are present, they

may not have clear boundaries, but tend to

run together. There are from one to three

black spots within each ocellus.

3) The anterior margin of the carapace has

many little black or chocolate brown

speckles in Sacalia bealei . There are few

or none in S. quadriocella .

Differences in skull characters

1) The length from anterior of prefrontal to

posterior of basi occipital to supra occipital

is 2.59-2.62 (N = 3) in S. bealei ; but

2.92-3.39 (N = 3) in S. quadriocellata..

2) The quadrate is not in contact with the

opisthotic in Sacalia bealei, but in S.

quadriocellata the posterior part of the

quadrate tends to contact the opisthotic.

abed

FIG. 1. Ocelli of Sacalia quadriocellata and S. bealei. A: Male 5. quadriocellata. B: Female S.

quadriocellata. C: S. bealei with a pair of ocelli. C: S. bealei with two pairs of ocelli.

(Pi ioon v,„

April 1990

Asiatic Herpetological Research

Vol. 3, p. 121

TABLE 1 . Analysis of shell suture data.

Fig. 2 for details of measurement.

See

FIG. 2. Means of measurement of shell sutures.

3) The posterior process of the jugal turns

up in S. bealei, but turns down in S.

quadriocellata.

4) The shortest distance between the orbital

is located at the anterior prefrontal in

Sacalia bealei. It is located at the joint of

the frontal and prefrontal in S .

quadriocellata.

Shell suture differences

Thirteen shell suture characters in S.

bealei (N = 27) and S. quadriocellata

(N = 26) were analyzed (Fig. 2). Seven

characters (PA/PL, PB/PL, PE/PL, IH/PL,

IAB/PL, IAN/PL, BL/PL) are different

from each other (t, p< 0.05), [Table 1].

Sexual dimorphism

Sexual dimorphism was evident in

Sacalia quadriocellata. In life, males have

very distinct orange-red speckles near the

neck and limbs, whereas females have not.

The carapace is narrower at the anterior end

than the posterior end in females, but both

ends are nearly equal in males. The most

interesting character is ocelli coloration.

They tend to be greyish with a white ring

surrounding the two ocelli on each side in

males (Fig. 1), but tend to be yellow

without a white ring in females. The skulls

of the males were somewhat flat, narrow

and long. The skulls of the females were

somewhat convex, wide and short.

Discussion

Since Sacalia bealei and S .

quadriocellata have distinguishable

characters, we consider 5. quadriocellata

to be a valid name. Scalia quadriocellata

occurs in China on Hainan Island and

Guangxi Province. It is also found in

central Annam. The specimens reported on

by Zheng and Ding (1965) from Fujian

Province and by Zhong (1981) from

Jiangxi Province are misidentified S.

bealei. The later species ranges over most

of southern China, including Guizhou,

Anhui, Jiangxi, Fujian, Guangdong

provinces, Hainan Island, and Hong Kong.

Vol. 3, p. 122

Asiatic Herpetological Research

April 1990

Literature Cited

Salamandra 11:20-26.

FANG P. W. 1934. Notes on some Chelonians of

China. Sinensia 4(7): 145- 199.

POPE, C. H. 1935. The reptiles of China.

Natural History of Central Asia, Vol. 10.

American Museum of Natural History, New

York. 604 pp.

SACHSSE W. 1984. Chinemys reevesii var.

unicolor und Clemmys bealei var.

quadriocellata — Auspragungen von Sexual

dormorphismus der beiden, Nominat formen.

SCHMIDT K. P. 1927. The reptiles of Hainan.

Bulletin American Museum of Natural History

65:395-465.

ZHENG J. AND H. DING. 1965. [A preliminary

survey of the turtles and tortoises from Fukien].

Journal of Fukien Teachers College 1:163-193.

(In Chinese).

ZHONGC. 1981. [Two new records of reptiles of

Jiangxi Province]. Acta Herpetologica Sinica

1981,5(15):95-98. (In Chinese).