Volume 1, Number 1, January 1994 Russian Journal of Herpetology

TOBUSHING COMPANY 1 rOLIUM Folium Publishing Co., Moscow, Russia Russian Journal of Herpetology Vol 1, No. 1, 1994, pp. 42 — 52

SOME ASPECTS OF HISTORICAL BIOGEOGRAPHY OF ASIAN ROCK AGAMIDS

N. B. Anan’eva1 and B. S. Tuniev2

Oi igiiial article submitted October 21,1993

The genesis of areals of Asian rock agamids belonging to the genus Laudakia Gray, 1845 is considered. The hypothesis about two routes of radiation of these lizards endemic for arid mountain regions of Palearctic is offered. The touthefn route from the hypothetical center of origin in the area of mountain systems of Hindukush and Himalayas passed in the conditions of southern subtropics with well developed belt of sclerophilous forests. This evolutional line with “caucasia” complex remains the relic areas in forest regions. Northern group of species with “himalayam” complex had its speciation in northern subtropic area with more severe climatic situation determined by strong continental and arid conditions. Altitudinal and biotopic differences of Laudakia species under their sympatiic distribution are specially discussed.

Key words: Reptilia, Sauna Agamidae, Laudakia, Palearctic Asia, systematics, biogeography, distribu­ tion, evolution

The Asian rock agamids of the genus Laudakia phylogenetic scheme is impeded by the difficulty ob­ Gray, 1845 includes 16 species distributed in moun­ taining material for genetic analysis for several groups tain rock landscapes of arid zone from Greece and of species and by the obvious existence of parallel Nile river delta on the west through Middle East and trends in different developmental lines within Central Asia to Gobi Altai on the north-east and Laudakia genus. Bramaputra river on the east. The mountain ring­ We collected data on geographic and biotopic dis­ tailed agamas belonging to this genus have been con­ tribution of rock agamas for their discussion in con­ sidered for a long time as a part of the genus Agama text of notions on the climate and genesis of land­ Daudin, 1802 (Boulenger 1885; Wermuth 1967). Af­ scapes and vegetation. A comparison of recent are- ter Moody’s revision (1980) they were referred to a alographic patterns of the studied species with known distinct genus Stellio Laurenti, 1768. The subsequent data un paleogeography and paleobotany of the region study of the whole complex of mountain ring-tailed of their distribution may be used as additional method agamas showed that it is a paraphyletic group (Joger of research for understanding the trends of reptiles ra­ and Arano 1987; Ananjeva et al. 1990; Joger 1991). diation (Ananjeva and Tuni^ev 1992). It might be es- Taking into consideration these data and some nomen­ pecially fruitful for Laudakia because of the unique clature remarks some authors (Leviton et al. 1992) of­ chorological isolation of this genus (Anan’eva and Pe­ fer to use the generic name Laudakia Gray, 1845 for ters 1990). Asian rock agamids and for Afro-Arabian group of We gathered data on geographic and biotopic dis­ species - -th e generic nameAcanthocercus Fitzinger, tribution of Laudakia caucasia, L. chemovi, L. ery- 1843. throgastra, L. himalayana, L. lehmanni, L. stolicz- The analysis of phylogenetic relations of Lau­ kana in 1975 — 1992 during field work in Armenia, dakia species was accomplished on the basis of both Georgia, Dagestan, Turkmenistan, Tadzhikistan, morphological and biochemical data (Joger and Arano Uzbekistan and Mongolia. We also used all the avail­ 1987; Ananjeva et al. 1990; Anan’eva and Sokolova able published information on the histoiy of landscape 1990; Joger 1991). These results are partly contradic­ formation typical for Laudakia areal and biotopical tory and can be considered only as preliminary ideas. distribution of these lizards in Iran Afghanistan, Paki­ They can not definitely determine the relations within stan, Turkey, Greece, Israel, and China. this groups of lizards. The construction of more clear Asian rock agamids endemic for Palearctic have the distribution looking like a integral unit (Fig. 1, Ta­ Zoological Institute, Russian Academy of Sciences, St. Peters­ ble 1). Within their diversity we can distinguish at burg, Russia. I least two complexes of species with a number of mor­ ' Caucasian State Biosphere Reserve, Sochi, Russia phological and ecological differences: “caucasia”

О 1993 Folium Publishing Company Some Aspects of Historical Biogeography of Asian Rock Agamids 43 complex and “himalayana” complex. The first one in­ cludes L. caucasia, L. microlepis, L. nuristanica, Black L. tuberculata, and probably L. sacra, L. lehmanni, Caspian Sea and L. erythrogastra. “Himalayana” complex con­ sists of related species L. himalayana, L. chernovi, L. badakhshana, L. stoliczkana (Peters 1971; Anan’e­ va etal. 1981; Anan’eva and Peters 1990). It is possi­ ble, that L. pakistanica that is most closely related to Arabian Sea L. agrorensis and L. melanura (Baig 1989) can be at­ tributed to this complex. We realize that it is a specu­ lative scheme, and probably future genetic examina­ F1S. I. General distribution of Asiatic rock agamids of Laudakia tion will permit to distinguish more groups and com­ genus in Palearctic. plexes of species within Laudakia. It was ilreadv noted (Anan’eva and Peters 1990), that long limbs, presence of a small gular sac and Despite the absence of clear phylogenetic hy­ nuchal crest, the poly annular structure of the caudal pothesis for Laudakia all the herpetologists who stud­ segments, juvenile color patterns should be consid­ ied their taxonomy and ecology are unanimous in the opinion, that speciation of these lizards has been con­ ered as plesiomorphic characters. But the interpreta­ nected in the past and at present with Hindukush, tion of separate morphological characters (including Pamir and Himalayas mountains (Peters 1971; Anan- scalation) as indication of species relationships is un­ jeva et al. 1981; Anan’eva and Peters 1990; Baig acceptable, and we use here such notions on plesio- 1992). Strong heterogeneity of dissected mountain re­ morphic characters of scalation only in context of are- lief probably promoted the isolation of populations, an alographic patterns of Laudakia. As it was noted ecological differentiation of these lizards is now nor­ above, the parallel trends occur in different develop­ mally expressed in different altitudinal distribution. In mental lines within Laudakia genus, expressed for ex­ this context it seems to be useful to discuss our hy­ ample in parallel changing of the number of annuli in pothesis for geographic distribution of the two com­ caudal segments in L. stellio, L. stoliczkana, and plexes of Asian mountain agamas and formation of L. caucasia (Peters 1971; Anan’eva and Ataev 1984). their contemporary ranges.

TABLE 1. The distribution of the species of Laudakia genus. Species Distribution

Laudakia agrorensis (Stoliczka 1872) Afghanistan, Pakistan, India Laudakia badakhshana (Anderson and Leviton 1969) Afghanistan Laudakia caucasia (Eichwald 1831) Armenia, Georgia, Azerbajan, Tajikistan, Turkmenistan, Turkey, Iraq, Iran, Afghanistan, Pakistan Laudakia chernovi (Ananjeva et al. 1981) Tajikistan, Turkmenistan, Uzbekistan Laudakia erythrogastra (Nikolskyi 1896) Iran, Turkmenistan, Afghanistan Laudakia himalayana (Steindachner 1869) Tajikistan, Uzbekistan, Kirgizstan, Afghanistan, Pakistan, India Laudahu lehmanni (Nikolskyi 1896) Tajikistan, Uzbekistan, Kirgizstan, Turkmenistan, Afghanistan Laudakia melanura (Blyth 18S4) Iran, Pakistan Laudakia microlepis (Blanford 1874) Iran, Pakistan, Afghanistan Laudakia nupta (De Filippi 1843) Iraq, Iran, Afghanistan, Pakistan Laudakia nuristanica (Anderson and Leviton 1969) Afghanistan, Pakistan Laudakia sacra (Smith 1935) Tibet (China) Laudakia stellio (Linnaeus 1758) Greece, Southwest Asia, Northern Egypt Laudakia stoliczkana (Blanford 1875) Mongolia, China Laudakia tuberculata (Hardwicke and Gray 1827) India, Nepal, Afghanistan, Pakistan Laudakia pakistanica (Baig 1989) Pakistan 44 N. В. Anan’eva and В. S. liiniev

We have veiy - ant;' paleontological data that there arose acceptable arid areas as distinct from luxu­ cannot elucidate the problems of origin and ancient riant tropical vegetation of the Himalayas proper in diversity of agamid lizards of Laudakia genus appropriate for colonization by these arid lizards. (Moody 1980; Estes 1983) especially because we Ouickly rising and reaching already cold desert condi­ have no reliable arguments for allocation of fossil re­ tions Tibet could not serve as a conductor for the ra­ cords to Agama, Trapelus, Laudakia or Ananthocer- diation of Asian mountain agamas which retain even cus. now the main recently the chorological patterns in The analvsis of present diversity of Asian moun­ subtropical area of Asia, Europe, and northern Africa tain ring-tailed agamas permits to assume the area of (L. stellio). recent Hindukush and Himalayas mountain systems We shall consider two complexes of species of to be the center of Laudakia origin. Their speciation Laudakia genus in the framework of this hypothesis. was influenced by aridization of the southern part of Our observations are based mainly on the field Asian continent and Alpic orogenesis. The allozyme work results in the areas of distribution of L. caucasia, based phylogenetic hypothesis for agamid genus L. himalayana, L. chernovi, L. lehmanni, L. erythro- Phrynocephalus (Macey et al. 1991, 1992) argues for gastra, L. stoliczkana. These observations and eco­ ancient divergence of agamids of Afro-Asian radia­ logical data from literature show that spccies of “cau­ tion group including Asian .nountain again,.1 casia" complex retain (to a certain extent) the ten­ (Laudakia). It is possible to think that the first stages dency to a semi arboreal mode of life. Unlike “hima- of this radiation took place in Miocene on the present layana” complex species these lizards are closely as­ territory of Afghanistan, Pakistan, and Tadzhikistan. sociated in their distribution with xerophilous forests The patterns of current distribution of Laudakia were and their derivatives. Thus L. caucasia in Caucasus is formed as a result of a number of geological events in rather common in the afforested gorges along Kura Middle East and Central Asia The established fact river (Borzomsky gorge and others), in the oak forests concerning the continuous subsistence of mountain of northern Armenia (Sevkar), in J uni pe re turn, Pis- belt of arid climate with corresponding subtropical tacio-Juniperetum light forests of eastern Georgia vegetation beginning with the Cretaceous on the terri­ (Vashlovani) and southern and eastern Armenia tory of ancient Mediterranean geosynclinal is of spe­ (Chosrov, Megri district), etc. L. caucasia is mostly cial importance (Kolakovskii 1974a 1974b). Theo­ common in shibliak belt in western Kopet-Dagh, it retically it could serve as a basis for the wide radiation can reach also Junipereturn and Aceretum forests. of the ancestor of mountain agamas. L. en’throgastra can occupy a considerably wide We suggest h^re the following hypothesis for the spectrum of biotops, but reaches its highest density of radiation of Asian mountain agamas that could occur populations in the gorges with Pistacia vera. by two ways. The first route probably passed from the L. lehmanni is associated in the most of localities area of the supposed center of origin to the regions of with “blackwood” forests of Platanus orientalis, Eastern Pami, from where further radiation round Juglans regia, and others. It is necessaiy to note that Takla-Makan along the southern periphery of Altai the aiboreal patterns of behavior for L. caucasia in and Tien Shan was possible. The radiation of Asian Caucasus and Kopet-Dagh and fo ri, erythrogastra in mountain agamas could go both in the eastern and Badchyz are observed considerably often But in com­ north-eastern directions to Mongolian Altai and in parison with these species L. lehmanni uses trees and south-western direction along the northern foot-hills bushes permanently and it can be reffered to both of Central Asian mountains. The use of natural corri­ petrophilous and semiaiboreal species. All the adult dors between Altai and Saur-Tar bagatai (Saissan de­ and juvenile specimens of L. lehmanni except for pression area), “Dzhungar Gates” between Taibagatai gravid females during and just before egg-laying were and Dzhunggar Alatau (AJakol depression), ?nd Ili observed only on the trees (Gissar and Peter the Great river valley between Dzhungar Alatau and Tien Shan mountain ranges). They can climb up to the height of seems to be very probable for this radiation. The fossil 8 — 15 m on the trees, where they find places for records of agamids were found in Zaisan depression basking, rest and feeding. One male, three-four fe­ and Balkhash lake area (Chkhikvadze et al. 1983). males, and young specimens were observed in one ar­ The second route of radiation of the Asian moun­ boreal microgroup. tain agamas probably passed round Transhimalayas Habitats of L. tuberculata and L. nuristanica are mountains. Under the conditions of “rainy shade” also correlated with typical forest biotopes. Massinger Some Aspects of Historical Biogeography of Asian Rock Agamids 45

(cit. after Anderson and Leviton 1969) notes conifer­ typical at elevation below 2000 m, wormwood ous and evergreen oak forests and moist habitats steppes are typical at elevations of 2000 — 2700 m among the main biotopes for L. nuristanica. Baig and arid Junipereto — Querceto — Pinetum light for­ (1988) also gives a similar information about the habi­ ests are typical above 2700 m (Menitskii 1984). tat of L. nuristanica in Pakistan (Ziarat, District The most eastern distribution has L. stoliczkana. Chitral of N. W.F.P.): “wet, temperate, mountain, co­ Its preferred biotopes are gorges and slopes of the niferous forest along a water courses”, p. 200. Accord­ mountains with screes, crevices and thorny bushes of ing to Menitskii (1984) the dense oak forests of ever­ Caragana (Borkin et aJ 1990). L. sacra according to green Quercus baloot and thick underwood of Frax- recent knowledge is restricted in its distribution by the inus xanthoxyhides, Lonicera nummularifolia, Co- river drainage of the Yarlung Zangbo in Lhasa valley, toneaster insignis are typical for the narrow mountain Tibet (Anaiyevaetal. 1990). They were observed only valleys of mountain rivers at the elevations of 1200 — on slopes covered with large boulders at elevations 1700 m in Nuristan In the higher areas forests are re­ from 3000 to 4000 m. This area referred to Tibet Aus­ placed by coniferous forests of Cedrus deodara, Pinus tralis district of Provincia Tibetica of Central Asian gerardiana, Abies webbiana, Picea morinda. In Pak- subregion of Mediterranean region (Grubov, 1959) is tia (Kunar province, Gardez region) Quercus baloot the warmest region of Tibet. This is a considerably oak forests are replaced at an elevation above 2200 m arid region in spite of 500 mm of average annual pre­ by forests of Cedrus deodara and more mesophilous cipitation The precipitation is brought by the summer oak forests of Quercus dilitata and also Acer turke- monsoon that enters to the Himalayas from the south stanicum, Celtis caucasica, Juglans regia, Pyrus pa- and breaks through the valley of Zangpo (Brachmapu- shia, etc. tra) river from the east. Precipitation is of a stormy The areaL. tuberculata is defined by more luxuri­ nature. The average annual temperature of Lhasa is ant subtropical and temperate forests. Thus in the for­ + 9°C at an altitude of 3650 m. The average winter ests widely distributed from Penjab to Western Nepal temperature is slightly below 0°C and the average evergreen oak Quercus incana dominates with its temperature of the warmest month is only about usual satellites Rhododendron arboreum, Mallotus + 15°C. The climatic situation is characterized also by philippiensis, Woodfordia fruticosa, Rhamnus tri- an absolute maximum of temperature + 30°C and an quetta, and Ilex dipyrena. absolute minimum of temperature - 17°C. Open Unlike “caucasia” complex species of “himalay­ v nodland of xerophilous bushes with dominating ana” complex inhabit open areas without forests, as a Sophora vicifolia, Berberis, Lonicera spinosa, Cera- rule, with semidesert vegetation, or herbaceous asso­ tostigma griffitii, Buddleia tibetica and parcipitation ciations on rocks and screes. L. chernovi living in re- of Rosa sericea, Cotoneaster acutifolia, Dasiphora fug

fauna. At the same time the northern route of radiation We can suppose that it was the time of the widest dis­ passed in the area of northern subtropics of Kun Lun tribution of polymoiphic species L. caucasia. Accord- and Tien Shan with more severe arid an continental i.:g to the general idea of “forest” development of climatic conditions. Probably the speciadon of Asian “caucasia” group the populations from the north of mountain agamas could be divided into two trends Suleiman mountains, (the southern slope of Hin­ correlated with environmental conditions. The south­ dukush) are likely to be the most ancient. This area ern group of the diverse “caucasia1’ complex retained presents also the limit of the Asian summer monsoon the relict areas in the forest belt of western Himalayas influence and development of evergreen oak forests. and Hindukush, and the northern group of species of The taxonomic status of these populations and L. cau­ the “himalayana” complex developed in the condi­ casia from southern Tadjikistan and regions adjacent tions of increasing aridizatioa to Afghanistan (where L. reticulata (Agama reticulata The active adaptive radiation could occur prob­ Chernov, 1935) was descovered) and Southern Af­ ably on the border of Miocene and Pliocene, and later ghanistan — Western Pakistan (between Kandagar in Pliocene. Pliocene was the time of raising of moun­ and Quetta) is of special interest for following re­ tains and orogenetic movements of “block-type”, i.e., search. The origin of L. caucasia distributed from different parts of mountains raised irregularly Turkmen — Khorassan mountains to Transcaucasus (Gvozdetskii and Mikhailov 1987). Ecological condi­ seams to be more recent Probably in Pliocene L. cau­ tions had different trends of changing during this oro­ casia rached the northern slope of Great Caucasus. genesis. Species of “caucasia” complex were pre­ The South-Eastern part of Iranian high plateau might served in the least modified forest conditions of sub­ be considered as the place of formation of L. mi- tropical climate and phytolandscapes. Species of ''hi­ crolepis in this period. The problem of relationships malayana” complex distributed in Pamir and Hin­ of this species (or subspecies L. c. microlepis) needs dukush probably extended their areas under increas­ further examination (Anan’eva and Ataev 1984). Fur­ ing aridization further to the south right up to the inte­ ther dispersion of the species of L. caucasia complex rior regions of Kashmir. The numerous disjunctions of both to the west and to the east was possibly stopped formerly continuous areas of species belonging to by climatic situation and probably by comoetition both complexes took place probably also in Pliocene. with other related species: L. nupta in the southern re­ The rapid raising of Tibet determined the conditions gions of Iranian high plateau and L. melanura on the of a cold desert that could not be suitable for the radia­ east. L. caucasia occupied at that time the widest area tion of mountain agamas. An exception in this area is including not only the majority of the Eastern Kopet- the warmer Lhasa valley and maybe some refugia of Dagh, all the Central and the Western Kopet-Dagh, Kun Lun that have different microclimadc conditions. but also reached Lesser and Bigger Balkhan moun­ It seems that the glaciation and rapid raising of moun­ tains on the North-West of Turkmenistan along the tains made impossible the survival of these lizards in canyons of small rivers from Kopet-Dagh to the coast the Eastern Pamir and in most high parts of the Central of the ancient Amu-Daiya river flowing at that time Tien Shan. into Caspian Sea. It reached also the regions of upland The vertical segregation of the species might have of Kjujuki and Meshed-Messerian inclined plain on been important for evolution of “himalayana” com­ the south-west along the incised valley of Atrek river. plex in Pliocene. L. chemovi was probably common It’s interesting to note, that now a similar dispersion in the belt of deciduous and shibliak vegetation, and of L. caucasia along the steep banks of rivers to the L. himalayana was distributed in open meadow and plains in the Eastern Transcaucasus and in the South- rock biotopes above the forest belt. The mountain rais­ Western Kopet-Dagh in Sumbar river valley (Makeev ing in Central Asia separated the species of “himalay­ 1985) is observed. ana" complex to a number of populations that must We would like to discuss the possible ways of the have been important for the speciation of L. sto- formation of L. erythrogastra areal. If we share the liczkana and probably of another species. view that L. erythrogastra and L. melanura are related Pliocene was also an important period for the spe­ species it is possible to propose the way of penetration cies of “caucasia” complex. At that time Iranian pla­ of their ancestors from the south of Iranian high pla­ teau and other high plateaus of South-Western Asia teau. This way could pass along the depressions of that arised at the site of the Sarmat Sea became a new Mash-kel, Sabari, Dagi-Tundi, Namaksar to the basin center of speciation in xerophilous fauna and flora. of Geri-Rud (Tejen) river. L. erythrogastra dispersed Some Aspects of Historical Biogeography of Asian Rock Agamids 47 to the east and to the west from this area to Meshed, disappeared. As a result three localities of L. chernovi on the south to the regions of Kopet-Dagh, Badkhyz, were preserve: 1. Tupolang river valley on the south­ Karabil and to Central Hindukush along its northern ern slope of Gissar mountain range (Uzbekistan) at an foot-hills. It inhabited the biotopes that were not occu­ elevation of 800 — 1000 m; 2. Surkhu ridge (Tadjik­ pied by L. caucasia, a species with similar ecological istan) — elevation of 900 — 1400 m, and 3. on the requirements. In rare cases of sympatiy of L. caucasia crest of the southern part of Kugi-Tang mountain and L. erythrogastra they have altitudinal differences: ridge (the border between Turkmenistan and Uzbekis­ biotopes more xerophilous and thermophilous L. ery­ tan) — elevation of 2400 — 3000 m (Ataev 1985; throgastra are characterized everywhere by lower Khabibulov 1990). It is important to note the increase hypsometric indexes. This is also typical for L. nupta, of hypsometric marks of biotops of L. chernovi in the L. melanura, L. stellio. On the northern slope of East­ southern direction from Tupolang to Kugi-Tang. It is ern Kopet-Dagh L. erythrogastra inhabits the eleva­ a result of both Holocene altitudinal displacement of tions of 600 — 900 m and L. caucasia inhabits the vegetation upwards along the slopes and of the general biotopes at the elevations 800 — 1500 in According recent alterations of temperature and solar radiation in to Anderson’s (1968) data, on the southern slope of the southern direction. This regularity is observed for Eastern Kopet-Dagh the elevations of 800 — 1600 m most mountain Asian agamas with relatively exten­ are typical for L. erythrogastra and 960 — 2575 m are sive areals. The altitudinal displacements of biotopes typical fori, caucasia. In the area of sympatiy in Hin- of L. erythrogastra from the northern slope of Eastern dukush L. erythrogastra is distributed at an elevation Kopet-Dagh to Hindukush was already noted. The al­ of 915 — 2440 m (Anderson and Leviton 1969) and titudinal differences in biotops was noted also for L. caucasia has a higher altitudinal distribution — L. caucasia: elevation of 600 m in the north of its 2440 — 2745 (Clark et al. 1969; Anderson and Levi­ range (Great Caucasus, Dagestan) to 1000 m on the ton 1969; Leviton and Anderson 1970). L. erythrogas­ north of Lesser Caucasus (Georgia, gorge in the mid­ tra penetrated into the northern slope of Kopet-Dagh dle stream of Kura river); 900 — 2000 m in Armenian (Tuniev et al. 1991) probably in Pliocene by the Tejen highland; 950 — 2575 m in Kopet-Dagh (Ataev 1985; river valley. A distinct subspecies, L. e. nurgeldievi is Anderson 1968) and 2745 m in South-Eastern limit of known from some localities between the Chaacha and the L. caucasia distribution in Hindukush (Leviton Layensuv rivers (Tuniyev et al. in press). L. erythro­ and Anderson 1970). gastra is petrophilous lizard, but it inhabits areas with L. himalayana inhabits the elevations of 2400 — burrows of rodents normally used as shelters. That 2600 m in the Big offspur of Gissar range, the eleva­ was also noted for L. stellio (Clark and Clark 1973). tion of 3200 m in Western Pamir (Vanch river valley Quartemary period was characterized over the near glacier of Russian Geographical Society) and the whole Euroasiatic continent by the global alteration of elevations above 3500 m in Southern Pamir. climate and landscapes connected with Pleistocene Pleistocene fluctuation of the Caspian Sea modi­ glaciations on the north of the continent and its moun­ fied the primary distribution of L. caucasia along tain regions and periods of hot arid climate during in­ Western and Eastern seashores. The maximal trans­ terglacial periods and Postglacial epoch It was Pleis­ gression of the Caspian Sea flooded Kura-Araks river tocene that probably determined the current patterns lowland and the littoral part of Dagestan in the Cauca­ of distribution of Asian mountain agamas. sian part of the areal of L. caucasia. As a result some It seems possible to assume the trends of specia- populations of this species were isolated. In Central tion in “himalayana” complex during this time. The Asian part of its range the flow of Amu-Darya river area o f L. himalavana desintegrated into a number was turned to the north to Aral Sea. Accelerated proc­ of populations isolated by high mountain ranges ess of allegation divided the continuous areal of (5000 — 7000 m elevation) of Gissaro-Darvaz, West­ L. caucasia into several isolated populations inhabit­ ern Pamir, etc. This Pleistocene isolation seems to be ing Krasnovodsk plateau, Bigger Balkhan, Lesser a definite factor of speciation for such species as Balkhan and Western Kopet-Dagh. Khvalyn trans­ L. pakistanica or L. badakhsharn gression of Caspian Sea flooded most of the territory In Pleistocene the refugia of subtropical microcli­ of Transcaspian plain including the western part of mate were reduced to small territories on the western Meshed sands and residual — mountain Kyuyuki. Its outskirts of Pamir-Altai mountains; in all the other overthrust sheet was later covered additionally by the parts of this mountain system the subtropics totally drift of deposits of the Atrek river (Durdyev 1969). 48 N. В. Anan’eva and В. S. Ttiniev

Only the eastern part of Meshed sands remained, as an teau. Their absence is the general regularity for all in­ island not flooded during all the period. L. caucasia terior plateaus in the southern part of Asia For exam­ triannulata was speciated just in this unique land­ ple, they are absent in the lesser territories of Ana­ scape. Meshed sands wi*h the capacity o f25 — 350 m tolian and Kars plateaus although their northern and (Durdyev 1969) extends for 50 km from north to eastern outlying districts are inhabited by L caucasia, south and for 20 — 25 km from west to east (Kalenov and L. stellio lives on their southern outlying districts. 1973). The development of the ravine system in this It is also possibly can be explained by the screening area is of special interest This unique phenomenon is effect of the marginal mountain ranges along all the absent in all other sandy deserts of Turkmenistan On asiatic mountain belt of the Alpic folding. the east of Meshed massif a stratum of non stratifi- The survival of L. sacra during Pleistocene can be cated Meshed formation was distinguished. Its sands explained also by the climatic peculiarities of the have a more prominent thickness in comparison with Lhasa valley. The average temperature in this area is the sands of diagonal stratification of Meshed forma­ even more high than in the adjacent subtropical region tion They are characterized by more intensive devel­ of Sichuan at the same latitude. Being measured at the opment of ravine system and by the absence of bush same elevation the average July temperature of Lhasa vegetatioa is by 10°C and the average January temperature is by According to Kalenov (1973) Meshed sands are 14°C exceed the average temperatures in Chunzin city distributed

Kugitang-Tau mountain range (Eastern Turkmenis­ derson and Laviton 1969) on the southern slope of tan). The vertical zoning of this mountain range is Hindukush L. erythrogastra inhabits the belt up to an clearly expressed (Ataev 1985; Khabibulov 1990). elevation of 2440 m, and L. caucasia and L. badakh- L. lehmanni was found here on the elevations of shana are distributed at higher elevations up to 800 — 1200 m, the upper mark of its distribution is 2600 — 2650 m. 2600 m. This semiaiboreal lizard quite often pene­ The case of sympatric distribution of related spe­ trates to the plain along separate rocky chains. L. cher­ cies from the same complex: L. chernovi and L. hima­ novi in Kugitanp-Tau inhabits exceptionally the layana is of special interest. This information was re­ ridge of the mountain range in the elevation 2700 — ceived from L. chernovi specimen 11421 stored in 2800 m. It is proved to be a typical petrobiont, prefer­ Dresden museum fur Naturkunde. This specimen was ring bare rocks without vegetation on the edge of east­ collected by German Alpinist G.SchOne in 1974 on ern scarp. It was observed on the highest middle part Matcha river (upper part of Zeravshan river) near of the mountain range, which is the highest, along the Khudgif village. This collector noted also sympatric upper cornices of the gorge walls near their sources. L. himalayana in the same place. The examination of Khabibulov (1990) determined it as one of the most this specimen and also of another specimen 8513 stenotopic species of lizards in this region. Altitudinal (Dresden museum fur naturkunde) collected on preference of high elevation is typical for L. chernovi Kshtut river, Gissar mountain range, elevation of from Kugitang-Tau (Ataev 1985; Khabibulov 1990) 2900 — 3500 m, show that these L. chernovi speci­ as for the agamas from Baisum-Tau in Uzbekistan. As mens do not demonstrate such strong differences in it was noted above, the populations from Nuiek and scalation with L. himalayana as L. chernovi from Tupolang areas inhabit lower vertical belts. It is ex­ terra typica (Anan’eva et al. 1981). This fact of sym­ plained by the landscape and climatic differences in patric record of L. chernovi and L. himalayana seems these pans of areal. The reasons of this ecological dis­ to be an additional argument for their specilic distinc- placement were considered above. The altitudinal dis­ tivity. placement is usual for Laudakia species with exten­ Another couple of sympatric species: L. caucasia sive areals. It relates also to L. stellio that shows ver­ and L. erythrogastra presumably belongs to one “cau­ tical differences in Turkey (1200 m), Cyprus casia” complex. The data on altitudinal preferences of (1600 m), Lebanon (2000 m). It was found at the Her- these species were already considered above. In rare mon foothills at an elevation of 1650 (Werner and cases of their sympatry (for example in С hare him? n Avital 1980). gorge in the Eastern Kopet-Dagh) these lizards have With regard to L. chernovi Khabibulov (1990) biotopical differences: L. caucasia inhabits steep supposed that as far as the relative humidity of the air slopes of canyons and L. erythrogastra occupies the on the external mountain ranges of this mountain sys­ gently sloping of hills in the gate of gorge and above tem in the area of the ridge is reduced in comparison the rocks. The conclusion about the possibility of with the middle mountains and is close in humidity to sympatric occurance of no more than two species in the low-mountain regions of interior mountain ranges the same altitudinal belt seems to be possible. Nor­ it can determine strong differences of the altitudinal mally they belong to different complexes and have mi­ preferences of lizards from the populations of Kugi­ crobiotopic differences. The sympatric species of the tang-Tau and Baisun-Tau (high elevations) and south­ same complex usually have not only biotopical but ern Gissar and Suikh-Ku (low elevations). Humidity also altitudinal differences. really can be important as a limiting factor for the dis­ The problem of sympatiy of related species from tribution of arid reptiles. It is especially significant for the same complex of Laudakia is correlated with ideas L. chernovi preferring low indexes of the moistering of hybridization and hybridogenous origin of some coefficient typical for the regions of interior mountain forms of mountain agamas. Panov et al. (1987) ex­ ranges of Pamir. pressed interesting hypothesis for hybridogenous ori­ Very few data on sympatric distribution of gin of the western group of populations of L. caucasia Laudakia species and on their altitudinal and biotopi- in Turkmenistan They supposed that this area may be cal preferences are available. According to Anderson considered as the zone of secondary intergradation. and Leviton’s (1969) data L. lehmanni and L. hima­ Two hypothesis are discussed in this context: a) these layana are sympatric at an elevation of 457 m near populations are a result of L. caucasia and L. c. trian- Mazar-i-Sharif in Hindukush. In Paghman area (An­ nulata hybridization or b) these populations and also Some Aspects of Historical Biogeography of Asian Rock Agamids 51

“triannulata” form are a result of hybridization of Ananjeva N. R and Tiiniyev (1992), “Historical biogeogra­ L. caucasia and L. microlepis. Zykova and Panov phy of the Phrynocephalus species in the USSR,” Asiatic Herpetol. Res., 4,76 - - 98. (1990) were also the authors of a hypothesis for hybri- Anderson S. C. (1968), “Zoogeographic analysis of the liz­ dogenous nature of some individuals in sympatric area ards fauna of Iran,” in: The Cambridge History o f Iran. I. of L. caucasia and L. erythrogastra. They suggested The Land of Iran, London, pp. 305 — 3T1. ’ that the type specimen of Laudakia (Agama) erythro­ Anderson S. C. and Leviton A E. (1969), “Amphibians and gastra pallida Nikolsky, 1897 considered should be as reptiles collected by the Street expedition to Afghanistan,” hybrid from sympatric zone. These ideas about hybri- Proc. Calif. Acad. Sci., 37(2), 25 — 56. dogenous origin of some forms of Laudakia may be Ataev Ch. (1985), Reptiles o f the Mountains o f Turkmenistan useful also in elucidation of some debatable problems [in Russian], Ilym, Ashkhabad. of taxonomy in “himalayana” complex. This consents Baig K. J. (1988), “New record of Agama nuristanica the form Laudakio (Stellio) bochariensis Nikolsky, (Sauria, Agamidae) from Pakistan,” Biology, 34(1), 1899 that was reduced to synonyms of L. himalayana 199 — 201. by Chernov (1959). This form has a number of char­ Baig K. J. (1989), “A new species of Agama (Sauria, Agami­ dae) from Northern Pakistan,” Bull. KitakyushuMus. Nat. acters intermediate fori, himalayana and L chernovi H ist,9 ,117— 122. that can be treated as an argument for hybridogenous Baig К J. (1992), Systematic Studies o f the Stellio-Gmup of origin of some forms of Laudakia. Agama (Sauria, Agamidae), Ph. D. Thesis Quaid-i-Azam The hypothesis for Laudakia radiation and their University, Islamabad. arealographic patterns can be modified and corrected Beutler A (1981), “Agama stellio ( Linn. 1758), Hardun,” after new records and results of using of new methods in: W. Bohme (ed.), Handbuch der Reptilien und Am- in taxonomy. One might expect new records and de­ phibien Europas, Akad. Verlagsges, Wiesbaden, scriptions of new forms from unexplored high moun­ pp. 161 — 177. tain areas of Afghanistan, Iran, Western China, and Borkin L. Ya., Monkhbayar Kh., Orlov N. L., Semenov D. Northern India. The proposed shame of radiation of V., and Terbish Kh. (1990), “Distribution of reptiles in Mongolia,” Trudy Zool. Inst. Akad. Nauk SSSR, Lenin­ mountain Asian agamas is typical for speciation in grad, 207,22— 138. mountain regions. Such type of speciations resulting Boulenger G. A. (1885), Catalogue o f the Lizards in the Brit­ in polytypic genus (or supergenus) was described be­ ish Museum (Nat. Hist.). I. Gekkonidae, Eublepharidae, fore for Archaeolacerta from Caucasus and Asia Mi­ Uroplatidae, Pygopodidae, Agamidae, Taylor & Francis, nor (Darevskii 1967) and Euro-Siberian viperid London. snakes of Vipera kaznakowi complex from Great and Chernov S. A. (1959), “Reptilia. The Fauna of Tajik SSR,” Small Caucasus (Orlov and Tuniev 1986). Trudy Mad. Nauk Tadzhikskoi SSk, Stalinabad, 42. Chkhikvadze V. N., Shammakov S. М., and Zerova G. A. 1983, “On the the history of the formation of the Sqamata REFERENCES fauna (Squamata) оt Middle Asia and Kazakhstan,” Izv. Anan’eva N. &, Peters G., and Rzepakovskii V. Г. (1981), Acad. Sci. Turkm. SSR, Ser. Biol., 2, 3 — 8. “New species of the mountain agamas from Tajikistan, Clark R J. and Clark E. D. 1973, “Report on a collection Agama chernovi sp. nov.,” Trudy Zool. Inst. Akad. Nauk of amphibians and reptiles from Turkey,” Occas. pap. SSSR, Leningrad, 101,23 — 26. California Acad. Sci., 104, 1 — 62. Anan’eva N. & and Ataev Ch. (1984), “Stellio caucasius Clark R. J., Clark E. D., Anderson S. C., and Leviton A. E. triannulatus ssp. nov. — a new subspecies of the Cauca­ (1969), “Report on a collection of amphibians and reptiles sian agama from South-Western Turkmenia,” Trudy Zool. from Afghanistan,” Proc. California Acad. Sci., 4th ser., Inst. Akad. Nauk SSSR, Leningrad, 124,4 — 11. 36(10), 279 — 315. Anan’eva N. R and Peters G. (1990), “Stellio sacer (Smith Darevskii L S. (1967), Rock Lizards of the Caucasus (Lacerta 1935) — a distinct species of the Asiatic rock agamas,” saxicola) [in Russian], Nauka, Leningrad. Trudy Zool. Inst. Akad. Nauk SSSR, Leningrad, Durdyev Ch. (1969), “Structural analysis and history of the 207,3 — 11. relief development of the Southern part of Western Turk­ Ananjeva N. R , Peters G., Matey J. R , and Papen- menistan lowland,” Abstr. Ph. D. Thesis [in Russian], fuss T. J. (1990), “Stellio sacer (Smith 193 5)— a distinct Ashkhabad. species of Asiatic rock agamid from Tibet,” Asiatic Her- Estes R (1983), Handbuch der Paleoherpetologie. Teil petol. Res., 3,104 — 115. Ю/A. Sauria terrestria. Amphisbaenia, Gustav Fischer Anan’eva N. R and Sokolova Т. M. (1990), “The position Verlag, Stuttgart — New York. of the genus Phrynocephalus Каир 1825 in agamids sys­ Grubov V. J. (1959), Botanical-geographical regions o f tem,” Trudy Zool. Inst. Akad. NaukSSSR, Leningrad, 207, Central Asia [in Russian], Izd. Lenigradskogo Univ.,Len­ 12— 21 ingrad. 52 N. В. Anan’eva and В. S. Ttiniev

Gvozdetilul N. A. and Mikhailov N. J. (1987), Physical Makeev V. N. (1985), The Populations o f Amphibians and Geography o f the USSR. Asian part [in Russian], Reptiles in South-Western Kopet-Dagh. Ecological Pecu­ \fysshaya Shkola, Moscow. liarities o f the Protection o f Animal World [in Russian], JogerU. (1991), “A molecular phytogeny of agamid lizards,” Moscow, pp. 30 — 34. Copeia, 3,616 — 622. Menitsldi Yu. L. (1984), The Oaks o f Asia [in Russian], Joger U. and Arano E (1987), “Biochemical phylogeny of Nauka, Leningrad. the Agama genus group,” in: Van Gelder, Strijbosch, and Monin A S. (1977), The History of the Earth [in Russian], Berger (eds.) Proc. 4th Otd. Gen. Meet, Soc. Europ. Nauka, Leningrad. Herp., Nijmegen, 215 — 218. Moody S. M (1980), “Phylogenetic and historical bio- Kalenov G. S. (1973), “Essay of the vegetation of Meshed geographical relationships of the genera in the Agamidae sandy massive,” Botan. Zh., 58(5), 658 — 665. (Reptilia, Lacertidae),” Ph. D. Thesis, University of Khabibulov M R. (1990), “Reptiles of Kugitang-Tau (East­ Michigan. ern Turkmenistan). Fauna, ecology, conservi 5 on,” Abstr. Ph. D. Thesis [in Russian], Leningrad. Orlov N. L. and TUniev R S. (1986), “Present ranges, pos­ Kolakovskii A. A (1974a), “Data on paleogeography of sible ways of their formation and phylogeny of the three Caucasus in association with the formation of flora,” species of vipers from the Eurosiberian group (V. kaznak- Trudy Sukhumsk Bat. Soda, 20,115 — 131. owi complex) in the Caucasus,” Trudy Zool. Inst Akad. Kolakovsky A. A. (1974b), “Vertical belts of Kolchis area in NaukSSSR, Leningrad, 157,134 — 135. Tertiary,” TrudyLesn. Inst, Tbilisi, 21,98 — 115. Panov E. N.. Zykova L. Yu., Gamer M E., and Va­ Lcviton A. E. and Anderson & C. (1970). “The amphibians sil’ev V. J. (1987), “The intergradation zone for different and reptiles of Afghanistan. A checklist and key to the forms of the Stellio caucasius complex in the South-’West­ herpetofauna,” Proc. Calif. Acad. Sci., Vol. 38, No. 10, pp. ern Turkmenia,” Zool. Zh, 66(3), 402 — 411. 163 — 206. Peten G. (1971), “Die Wirbelschwanze Zentralasiens Leviton A E., Anderson & C., Adler K, and Minton & A (Agamidae, Agama),” Mitt. Zool. Mus., Berlin, 47(2) (1992), Handbook to Middle East Amphibians and Rep­ 357 — 381. tiles. Soc. Study Amphibians and Reptiles. Tbniev R S., Ataev Ch, Shammakov S. (1991), “Stellio Le Kai Fu (1957), ‘The scheme of physic-geographical zon­ erythrogasternurgeldievi ssp. nov. (Agamidae, sauna)— ing of China,” in: E. N. Muizayev (ed.) Physical-Geo­ new subspecies from Eastern Kopet-Dagh,” Izv. Akad. graphical Zoning o f China [in Russian], Moscow, Nauk Turion. SSR, 6, 50 — 60. pp. 11 — 44. Macey J. R , Ananjeva N. &, Zhao E., and Papenfuss T. J. Wermuth H. (1967). Agamidae. Das Tierreich. Lisle darre- (1991), “A phylogeny for the agamid genus Phryno- centen Amphibien und Reptilien, Lief, Berlin. cephalus and its biogeographical implications,” in: Abstr Werner Y. L. and Avital E. (1980), “The herpetofauna of VI Ord. Meet. EHS, p 59. Mr. Hermon and its altitudinal distribution,” Israel J. Macey J. R , Ananjeva N. R, Zhao E., and Papenfuss T. J. Zool.,29,192— 193. (1992), “A phylogenetic hypothesis for the arid Asian Zykova L. Yu. and Panov E. N. (1990), “On possible hy­ agamid lizards genus Phrynacephalus,” in: Abstr. Asian bridization between the agamas Stellio caucasius and S. Herp. Meet Huangshan, Annul, China, pp. 59 — 60. erythrogaster,”Zool. Zh, 69(7), 103 — 106.