The open-type parking lot Kalavan-2 was found in the area of the lake. Sevan (Armenia) in 2005 at an altitude of about 1,600 m above sea level in the mountains that rise above the northern shore of the lake. The monument, whose sediments contain Paleolithic industry and faunal remains, marks an important stage in the prehistoric past of Armenia. Excavations of Kalavan-2 revealed the stratigraphic sequence of several phases of settlement, which date back to the Middle Paleolithic era. The radiocarbon date of 34,200 ± 360 BP, established from a fragment of tooth enamel of a large bovine animal, determines the age of the Mousterian level 7. The Cfa confirms the determination of these complexes as belonging to the final phase of the Middle Paleolithic, as well as the importance of the site for studying the last episodes of the existence of Neanderthals in the South Caucasus.

Key words: mustye, Middle Paleolithic finale, Lesser Caucasus, Armenia.

Introduction

Distribution of prehistoric human populations in Armenia, which occupies the southernmost part of the Caucasus Isthmus (Figure 1).

It is caused by paleogeographic events in the region. The Caucasus Isthmus is divided by the Greater Caucasus Mountain range. The latter creates a climatic and cultural barrier between the northern plains of the European continent and the southern elevations

1. Location of the Kalavan-2 site (a) and the main sources of obsidian in the Lesser Caucasus (b).

Transcaucasia, bordering the Middle East. The landscapes of Armenia are formed by the mountains of the Lesser Caucasus and high volcanic plateaus, dissected by river valleys and lake basins, which create ecological niches suitable for human habitation. However, this area is located at an average altitude of approx. 1 700 - 1 800 m above sea level, which could have been a factor determining the settlement and depopulation phases.

As a result of a recent archaeological survey conducted in Northwestern Armenia on the forested northern slopes of the Areguni Mountains (Barepat River Valley), which rise above the northern shore of Lake Baikal. Sevan, an open parking lot Kalavan-2 was discovered. This article presents preliminary results of the first excavation seasons (in 2006 and 2007, the sequence of deposits belonging to the final phase of the Middle Paleolithic was identified here [Colonge et al., 2006, 2007]).

Parking characteristics

The Kalavan-2 site is located at an altitude of 1,630 m above sea level on the left bank of the Barepat River, which belongs to one of the main hydrological systems of the Areguni Mountains. This river of the 3rd - 4th order follows the main fault line, which has a north-western strike and separates Mesozoic volcanic and sedimentary rocks (sandstones, porphyritic rocks, andesites, conglomerates, tuff breccias, etc.). In this area, at least five levels of terraces, level or tiered, can be distinguished, which emphasizes the extent of the area. changes in the landscape of northern Armenia in the Quaternary period. The Kalavan-2 site is located on the third Upper Pleistocene alluvial terrace at the confluence of the Barepat River with a small tributary of the 2nd order. The terrace is located at an altitude of 30 m above the water's edge and is subject to changes due to the tectonic activity of the area and the slope processes caused by it, as well as the arrival of stone blocks from the rocks above the eastern bank of the river.

The main excavation site (excavation 2) is located on the longitudinal axis of a terraced ledge, next to its northern tip, in a place where part of the rock base is exposed by erosion from under the cover deposits (Fig. 2). This part of the monument was inhabited in the late Bronze and Iron Ages: there is a burial ground, some burials of which were subjected to predatory attacks excavations. The flat surface of this area gradually slides to the northeast as a result of the movement of soil masses.

To clarify the stratigraphy and spatial position of the layers, the area of excavation 2 was increased to 7 m ² and penetrated to a depth of 99-115 cm (layer 11 was reached); in addition, in sq.L22, deposits were opened to a depth of 380 cm (layer 20 was reached). At the same time, excavation 1 in the east and excavation 3 in the west were laid on opposite sides of the slope. Chronostratigraphic correlation of various excavated sites of the monument is planned to be carried out in the next field season based on new dates and the results of analysis of artifacts.

To date, 20 main stratigraphic layers have been identified at the Kalavan-2 site, which represent individual episodes of sedimentation (Figure 3) and are sources of paleoclimatic and morphogenetic information (Colonge, Nahapetyan, Monchot, 2007; Olivier, Nahapetyan, 2008).

Fig. 2. Topographic map of the area where the Kalavan-2 parking lot is located.

The stratigraphic section of the monument indicates a variable nature of sedimentation, including alluvial manifestations, products of erosive terragenic slope processes and poor soil formation. Judging by the nature of gravel and pebbles, which correspond to a low-power flow, and the geometry of layers showing a pronounced drop, the alluvial component of sediments was formed by the right tributary of the Barepat.

Layers 19-16 form the main body of the alluvial terrace. There is an alternation of river sediments of various hydrodynamic types - pebbles, sand, silt.

Layers 15-11 correspond to a short alluvial-nival sequence of sedimentation characterized by angular gravel and sandy-silty lenses. These sediments may be the result of cyclical seasonal deposits in cool, humid climates.

This is followed by a bundle of layers 10-6, which reflects the mixed slope and aquatic genesis of sediments, which probably corresponded to relatively short-term temperate climatic conditions with poorly developed soil formation. Gravel destruction and sediment carbonation have been established in these layers, which indicates a significant post-depositional evolution (diagenesis) of sediments.

Layers 5-3 contain numerous detritus remnants (the surface relief of rock and gravel varies from slightly rounded to slightly angular), which are in contact with each other by erosion. Their accumulation in the layer was caused by a sharp change in the nature of sedimentation - erosion-slope processes were replaced by alluvial ones. This phenomenon may have been caused by unstable climatic conditions (for example, changes in the distribution mode or intensity of precipitation).

Layer 2 is the uppermost of the well-preserved Pleistocene levels. It corresponds to silty-sandy deposits, in which fine, slightly rounded gravel is occasionally found. This layer may represent the last episode of the return of the alluvial phase of sedimentation, which occurred within shallow meandering channels, in a cool and humid climate

3. Stratigraphic section of excavation 2. Drawing by S. Nahapegyan, D. Kolonzha.

and relative morphogenetic stability (no traces of slope erosion have been traced).

The last layer 1 is identified on the basis of a relatively young humus horizon, which reflects the conditions of modern soil formation. The upper part of the sediment section (layers 2 - 13) reaches a maximum thickness of 1.8 m. It is characterized by carbonatization along the secondary internal cavities of sediments and frost-breaking cracks and wedges that fix the conditions of cold climatic periods. Some of these cracks literally tear a pack of layers 2-7 into separate blocks, which are vertically displaced from a few centimeters to a decimeter.

Due to the movement of the ground, a stepped structure of layers was formed, falling in the direction of the northeast. The section shows that these cracks developed obliquely in layers 2 - 7, whereas in layers 8 - 13 they are laid subvertically. In the light of a preliminary sedimentological analysis of the entire sediment thickness, these structural features, in particular, individual ice lenses that once existed, show that the "frost wave" that reached layer 17 marks a local or even regional permafrost zone during a period whose chronological framework is still unknown. The ground movement was caused by the return of a humid temperate climate and is probably also associated with the lowering of the Barepat level during the main phase of the river's incision, which caused a destabilization of the slopes.

Eight archaeological layers with varying numbers of finds were discovered. Approximately 3/4 of the finds are artefacts and about 1/4 are faunal remains. The archaeological materials of layer 1 were not taken into account, as it contained elements introduced by slope processes, as well as artifacts from a nearby Iron Age burial ground.

The complex is characterized by the predominance of unidirectional primary splitting technology, a large specific weight of retouched tools, including scrapers of various types and various spiky points, as well as the use of tronking-faceting techniques in finishing tools (Yeritsyan, 1970; Beliaeva and Lioubine, 1998). For the middle horizon of the monument (layer 4), there is a radiocarbon date of about 49 Ka BP (Cohen and Stepanchuk, 1999). The industrial complex of the Yerevan-1 cave shows considerable affinity with the collection that was formed during excavations of the surface under a destroyed rock canopy in the Angegakot region of Southeastern Armenia [Liagre et al., 2006], as well as with assemblages from Taglar and Zar caves in Azerbaijan [Jafarov, 1983; Mansurov, 1990]. All these industries are very similar to the Zagros-Tavrian Mousterian materials (Golovanova and Doronichev, 2003; Fourloubey et al., 2003; Adler and Tushabramishvili, 2004; Bar Yosef, Belfer-Cohen, Adler, 2006).

The stone industries found in the Lusakerg-1 and -2 caves, located near the obsidian outcrops on Gutansar, are based on the use of this rock. In the Lusakert-1 cave, thanks to excavations carried out by B. Yeritsyan in 1970-1975 (Yeritsyan, 1975) and S. Furlube in 1999, six archaeological levels were identified on the preserved part of the terrace (the rock base was not reached). They show the alternation of different facies of the Middle Paleolithic: layers F-H (corresponding to level 6) represent a recurrent unidirectional Levallois technology focused on obtaining thin and elongated chips; layers D and C2 (levels 5-4) represent a fairly close facies, but with numerous toothed tools (most likely caused by mechanical damage in the layer); layer C1 (level 3) - again a plate-shaped faceted moustache with numerous points; layer B (level 2) - a return to a toothed moustache (with unintended toothiness). However, all Lusakert-1 complexes are characterized by the predominance of the recurrent unidirectional Levallois splitting technology and the use of the tronking-faceting technique.E. features inherent in the Zagros-Tavrian moustache (Fourloubey, 1999).

An AMS date of 26,920 ± 220 BP (GRA 14949/Ly1006) was obtained from a horse tooth found in layer C2 (level 4) of the Lusakert-1 cave (Fourloubey et al., 2003). But it can be wrong because of young carbon pollution.

The excavation carried out in Lusakert-2 cave in 1999 by D. Kolonjem partially confirmed the results of B. Yeritsyan's excavations: three levels of archaeological material were revealed (the rock base was not reached). They contained a complex of artefacts of the Levallois appearance, made mainly by recurrent unidirectional splitting. Scrapers, Mousterian points and "Yerevan" type points (retouched triangular points with a refined base) determine the composition of the gun set. The appearance of this industry is comparable to the Zagros-Tavria Mousterian complex [Ibid.].

The Middle Paleolithic of the neighboring regions of Western Georgia is characterized by significant facies variability. N. Tushabramishvili [1984], V. P. Lubin [1989], and M. Nioradze [1990] identified five local cultural variants, which, as these researchers assumed, represented various Middle Paleolithic cultural groups that simultaneously inhabited this territory (Adler and Tushabramishvili, 2004).

This grouping of industries was based mainly on technological and typological criteria. However, as shown by D. S. Adler and N. Tushabramishvili [Ibid.], the variability of the Middle Paleolithic regions of Georgia is more likely due to changes in environmental conditions, rather than cultural factors. Among the Middle Paleolithic variants of different ages, the earliest (Jruchula-Kudaro) may be related to the early Levantine Mustier (Ibid.; Bar-Yosef and Kuhn, 1999; Meignen, 2000; Meignen and Tushabramishvili, 2006), while the most recent variant (Ortwala-klde; Sakazhiya; Sagvarjile, layer V) shows similarities with Zagros-Tavrian moustache (Adler and Tushabramishvili, 2004; Adler et al., 2006).

For the entire territory of the South Caucasus, the question of the correlation of complexes from the above-mentioned sites of the end of the Middle Paleolithic and monuments of the early Upper Paleolithic period is highly debatable. According to some experts, the upper Mousterian layers at some sites of the southern slope of the Greater Caucasus (Sakazhiya, Ortvala-kld, Sagvarjile, layer V, etc.), according to data obtained during excavations in the early 1970s, include many Upper Paleolithic tools and can be considered as "transitional" to the Upper Paleolithic [Tushabramishvili, 1994; Nioradze and Otte, 2000; Golovanova and Doronichev, 2003]. Elements of these "transitional" periods can be traced even after the Denekamp interstadial, i.e. after 28,000 AD. Hence, the Caucasus could have been a place where Middle Paleolithic traditions were preserved and co-existed for a long time with Upper Paleolithic ones. The point of view of other researchers is based on the results of recent excavations in the Ortvala-klde cave (Adler, 2002; Adler and Tushabramishvili, 2004; Bar-Yosef and Belfer-Cohen, Adler, 2006): there is no evidence of the transition from the final Middle Paleolithic to the early Upper stage. On the contrary, there was a distinct archaeological, stratigraphic, settlement, and time gap between the final Middle Paleolithic (layer 5) and the early Upper Paleolithic (layer 4) associated with the initial stage of the Denekamp interstadial.

Archaeological materials of the site

In general, the good preservation of stone artefacts, although varying depending on the layer and type of raw material, makes it possible to clearly identify the material from which they were made (Table 1). Many objects from the upper part of the site sediments (layers 1 - 13) are covered with a carbonate crust. Carbonation was the result of water seeping through the pores of loose sediments, cracks and frost-breaking wedges.

The variety of raw materials for making artifacts found during the excavations of the Kalavan-2 site is due to the geological complexity of the Lesser Caucasus, which is located at the intersection of two sedimentary complexes - volcanic and sedimentary. Three types of raw stone materials were found at the site. The first two are local varieties. More than 1/3 (34 %) of the products are made of sedimentary rocks (silicified sandstone, corneas, flint or jasper). These rocks of various colors are represented in the alluvium of the rivers flowing near the monument. Local varieties of raw materials include volcanic rocks (basalt). They were brought to the site from ancient igneous geological formations of the mountain range that separates the lake. Sevan from the Getik river basin, into which Barepat flows. 2% of the artifacts are made from volcanic rocks.

The highest proportion (64 %) of products made from raw materials of the third type (volcanic glass, or obsidian, usually black, less often brown, red or colorless). Obsidian is not known in the Barepat basin and, therefore, is an exotic raw material for this region. The nearest obsidian springs are located on the western and eastern banks of Sevan, approximately 80-90 km in a straight line from the site (Badalyan, Chataigner, Kohl, 2004). Since Armenia is a country with a mountainous landscape, such distance indicators can hardly give an idea of the real distance. To determine how the distance factor might have affected the use of obsidian [Barge and Chataigner, 2004], we tried to use the capabilities of GIS (geographic information systems). Such spatial analysis features (included in ESRI's ArcGIS software) as" distance estimation in terms of labor costs "and" easiest path " allowed us to calculate the time required to cover the distance between points on the map and determine the path that requires minimal effort and time, taking into account the distance and steepness of mountain slopes. According to this study, the distance from the Kalavan-2 site to the nearest obsidian springs is 24-28 hours of continuous walking, or 3 to 4 days of travel (Chataigner and Barge, 2008).

Industries of the lower levels (layers 9-11). Isolated artifacts found in the lower levels suggest the presence of Mousterian rocks in these layers. A lamellar chip and three fragments of flakes were found in layer 11, including a convex chip on the proximal fragment of a thick flake (Figs. 4, 17). Layer 9 contains 8 finds, including a plate cleavage prepared using the recurrent convergent-unidirectional Levallois method. This is the only element that finds analogs in the overlying layer 7.

Table 1. Artefacts and bone remains from different layers of the site, copies.

Layer 7 industry. Most numerous in the Kalavan-2 parking lot. The collection contains 214 artifacts, of which 130 are made of obsidian. The industry has a very high proportion of small artefacts (61.7 % of items less than 20 mm long), mostly made of obsidian. Objects made of local pale purple or green flint are most often more than 20 mm long. The high degree of depletion of obsidian artifacts probably reflects their longer exploitation (Adler and Tushabramishvili, 2004). The largest cleavages were the result of applying only recurrent convergent and unidirectional splitting methods. Production was mainly aimed at obtaining lamellar chips (five whole and five fragmented); Levallois points are few (3 specimens), even fewer oval-shaped flakes (2 specimens). Chips of the latter type are most likely plates or points, the cleavage of which was unsuccessful due to insufficient convexity of the working front of the nucleus.

The industry of layer 7 is exclusively convergent and unipolar Levallois, with a set of tools, half of which are scrapers and retouched Levallois points (Figs. 4, 10-16). This allows you to be quite confident

Figure 4. Stone tools from layers 6 (1 - 9), 1 (10 - 16) and 11 (17, 18) of excavation 2. Drawings by Zh. Devilde.

It should be attributed to the Moutier type, which is represented in the lower layers (layers F - H) in the Lusakert-1 cave in Armenia. The only difference is that layer 7 of the Kalavan-2 monument does not contain any Yerevan-type pinnacles.

For layer 6, there are 41 artifacts, including 33 obsidian items. Three splitting methods can be distinguished. The first is the method of discoidal splitting; the result of its application is medium - sized flakes (30 - 50 mm long). The chips had a wide impact area, convergent cutting edges, one of which is sometimes occupied by the rim. Some of the chips are typical Pseudolevallois points. The results of the second method, which is aimed at removing plates, are only one whole object and one fragment. The convergent unidirectional Levallois method of removing plate chips is more widely used, probably similar to the one whose products are common in the underlying layer 7. However, the poor surface preservation of these artifacts suggests that they can be moved from other layers. There are 12 guns, all retouched. Among them: four retouched obsidian points, similar in size (length 27-37 mm, width 17-23 mm, thickness 6-9 mm); a flint tip, larger than the objects of the previous series (dimensions 55x19x10 mm); two end scrapers; an atypical chisel; a scraper with cool elements edge retouching, pièce esquillée; two indeterminate fragments (Fig. 4, 1-9).

The predominant role of unidirectional nucleation preparation technology, a moderate proportion of plates, a high specific weight of chips with convergent edges (retouched spiky points, convergent scrapers) , and the manifestation of tronking techniques are characteristic of the Zagros-Tavria moustache. This mousterian variant is widespread from Western Iran (Kunji Caves, Varvasi) (Dibble and Holdaway, 1990, 1993; Baumler and Speth, 1993) to Central Anatolia (Karain Complex I) (Otte et al., 1995), where its age is determined within the oxygen-isotope stage 3 (60,000 years). - 25,000 BC). As noted by D. S. Adler (Adler, 2002), hominids of the late Middle Paleolithic stage were part of a large prehistoric social system bounded by the Caucasus Range to the north and the Zagros and Taurus Mountains to the south.

Faunal remains found at the site

During the excavation of the Kalavan-2 site, 129 bones were found. The proportion of identified bone remains by species and / or relation to the skeleton is low (14 %). Most of the bones were found in Mousterian layer 7 (Table 2). The bones are very poorly preserved, usually porous and calcified. Their outer surface is usually destroyed, which often makes their interpretation impossible; fragments of spongy bones are rare. All materials bear traces of the influence of natural factors, for example, the enamel of teeth as a result of repeated freezing and thawing is strongly modified, cracked [Todisco and Monchot, 2008].

Bison cf. Bos primigenius. In 2006, a fragment of the upper jaw was found in layer 7 with two preserved right molars (M2 and M3), which belonged to a young adult large individual from the Bovidae family (Figs. In 2007, almost the entire occipital and parietal parts of the skull with preserved left and right horn openings were found in the same layer. The finds were in poor condition. The front part of the skull was crushed under the weight of the ground, scattered fragments of bones of the upper jaw, upper teeth (right M1, right P4, left M2, left MZ) and other cranial parts were found nearby, for example, the pyramid of the temporal bone. Due to the very poor preservation of the bones, reliable measurements are not available. Measurements of the teeth were also not made due to the destruction and fragmented state of the enamel. It can be assumed that the structure and shape of the horn openings are more characteristic of the bison than of the bison. There are very few detectable postcranial skeletal remains; fragments of the diaphysis of the long bones (humerus, tibia, and metacarpal bones) were assigned to this large polorogid, taking into account the significant thickness of the tubular bones.

Various morphometric features that can be used to distinguish representatives of the genera Bos and Bison are described in the literature (Slott-Moller, 1990; McCuaig Balkwill and Cumbaa, 1992; Bragal, 1984, 1993). However, the developed criteria can only be used if there is a sufficiently large sample. The upper molars of the animal from the Kalavan-2 site have a significant amount of dental cement, which covers the columns between the dental lobes. This makes it possible to assign the upper molars to the genus Bos (Fig. 5, 1). The additional column (entostille) of bison is shorter and hidden between the lobes located at the neck of the tooth. The upper left M1 has a distinct central island, which is more common in members of the genus Bos. All of the above features are more characteristic of bison than of bison. However, due to significant difficulties in conducting reliable species attribution, we are forced to limit ourselves to interpreting this sample as cf. Bos primigenius.

Bone remains of members of the genus Bison are rare in Crimean sites, but they are dominant in North Caucasian sites (Baryshnikov, 1999). They are also represented in the South Caucasus (Western Georgia).

Bone remains of various animal species presented at the site Table 2.*

* Fragment categories are determined by the thickness of the cortical bone layer.

in a context dating from the end of the Middle Paleolithic to the beginning of the Upper Paleolithic (the Ortwala-klde and Dzudzuan cave sites) [Adler et al., 2006; Bar-Oz et al., 2008]. Remains of the aurochs, an open-landscape animal, have been found in the sediments of Paleolithic sites in Western Georgia. Bison remains found in the epipaleolitic layers of Dzudzuana (layer 1B, ca. 13 000 - 11 000 (Bar-Oz et al., 2008) and in the Middle Paleolithic layers of the Ortval-klde (Adler and Tushabramishvili, 2004; Bar-Oz and Adler, 2005).

Wild goat / ibex Capra sp. Several bone remains of a mountain goat were found in layer 7 of the Kalavan-2 monument. We were able to identify the proximal limb of the second phalanx (L = 22.6; B = 26.5) and the distal limb of the first phalanx (L = 21.0; B = 17.6). These two elements were found at the anatomical junction, and their morphological and biometric characteristics indicate that they belong to the genus Capra (Boessneck, Muller, and Teichert,1964;Prammel and Frisch, 1986; Clutton-Brock et al., 1990; Fernandez, 2001). The third element is the whole sesamoid bone. Due to the lack of horn openings, it is impossible to determine in more detail the species belonging of an animal from the genus Capra found at the parking lot. Wild goats live in rocky, open landscapes and are resistant to harsh climatic conditions. Goats (Ovis and Ovis/Capra) served as prey for hunters of the Kalavan-1 Late Paleolithic site located nearby (Liagre and Balasescu, 2007). In the neighboring regions of Western Georgia at the Ortvala-klde parking lotCapra caucasica was the main prey (more than 90 % of all faunal remains); its remains are represented in both the Final Middle and Upper Paleolithic layers (Adler et al., 2006). In the Kozarnika Cave in Northern Bulgaria, this species meets-

5. Faunal remains from excavation 2. 1-5 - right upper molars of a large bovid (Bos sp.) (1-lingual norm; 3-vestibular; 4 - mesial norm M2; 2-vestibular norm M3; 5-vestibular norm M1, M2, M3); 6-8 - lower molar Equus caballus (6-lingual norm; 7, 8 - occlusal); 9, 10-fragment of the metacarpal diaphysis of a large bovid (9 - rear view; 10-front view). Photos by E. Monchot and M. Coutureau.

It has been observed since the Middle Pleistocene deposits (Fernandez and Cregut-Bonnoure, 2007).

Equus caballus horse. Only one lower molar (M1/M2) of a representative of the species was found in the sediments of the site in the contact zone between layers 6 and 7 (Figs. The tooth, whose position cannot be accurately determined (probably M2, because the developed mesial part of the hypoconulid and a strong bend of the crown can be traced), is assigned to Equus caballus by morphological features. Tooth dimensions (mm): occlusal length 26.3; occlusal width 16.2; height 52.0; postflexid length 10.3; "double node" length 14.0 (according to [Eisenmann, 1981]). The wild horse is an animal most characteristic of open landscapes, and its remains have never been found in large numbers in the Paleolithic sites of the Caucasus. Equus caballus is mentioned in descriptions of Taglar Cave and Dashsalakhli monuments in Azerbaijan, Yerevan-1 and Lusakert-1 in Armenia (Lubin, 1989), Dzudzuana Cave in Georgia (Bar-Oz et al., 2008), Ilskaya-1 and -2 sites, and Barakaevskaya Cave in the North Caucasus (Baryshnikov and Hoffecker, 2008). 1994; Doronichev, 2000], but it was not found in the Ortvala-kld cave in Georgia (Bar-Oz and Adler, 2005).

Red deer Cervus elaphus. The presence of red deer can be justified by the presence of a fragment of the mediolateral metacarpal diaphysis and a fragment of the dorsal metatarsal diaphysis (layer 7). According to their characteristics, these fragments are close to the Cerviclae family; the thickness of the cortical bone excludes their belonging to roe deer, in the latter it is much smaller. Red deer are very often listed in the fauna lists of Middle and Upper Paleolithic sites in the Caucasus, Middle and Middle East. It is an inhabitant of open forest spaces and is resistant to cold climatic conditions.

Taphonomy

Several types of surface modification have been identified on the bones, which reflect the complex taphonomic history of the site. First, most of the objects showed traces of the plant root system (Figs. 5, 9). The interweaving of these traces shows that the bones, before being buried by loose sediments, were for a long time in the zone of active vegetation of plants (Mottershead et al., 2003).

The bone material of the site was exposed to anthropogenic impact, which is confirmed, in particular, by the presence of traces of stone tool cutting on a small fragment of bone. Seven bones, including the tibial diaphysis of a large bovid, show signs of heat exposure. Some fragments of the bones of a large bovid preserved traces of a spiral fracture, which could have been formed as a result of fragmentation by a transverse impact (Figs. 5, 10), but these signs should be interpreted with caution as the results of human activity, given the origin of the objects (Morlan, 1983).

It is important to note the small number of so-called short bones (bones, wrists, tarsal, ankles, kneecaps and phalanges) and the absence of isolated teeth. The very high representativeness of the tibial and humeral diaphyses allows their determination to be made confidently (Morlan, 1994). Badgley (Badgley, 1986) drew attention to the fact that the frequency of occurrence of certain skeletal bones is closely related to the nature of sedimentation at a particular site. For example, a large number of vertebrae and phalanges are found in river estuaries, while teeth are mostly found in riverbed and alluvial deposits on plains.

Radiocarbon dating

Radiocarbon dating was performed for four samples from excavation 2: two samples were obtained in the Poznań Laboratory (Poland), and two more were obtained in the CAIS laboratory (Center for Applied Isotope Research) at the University of Georgia (USA). The sequence of dates obtained raises many questions (Table 3). For the UGAMS-2295 sample (layer 19), a date of 42,040 ± 400 BP (~43,500 ± 800 cal. bp) was determined, which is at the limit of the method's capabilities, and therefore this age should be considered as the minimum. The age difference between two samples taken from Bovid tooth fragments (NC-2006 and NC-2007, layer 7), possibly belonging to the same animal and analyzed in the same laboratory, can be explained by contamination of the sample. Definitions for the Poz-22181 sample due to its too young age should not be taken into account, and layer 7 can be dated to 34,200 ± 360 BP (~37,700 ± 880 cal. bp). The radiocarbon date for a horse tooth found at the boundary between layers 6 and 7 refers to the age of these deposits to the maximum of the last glaciation within 20 000 - 16 000 L. N. (oxygen-isotope stage 2). This date also looks too young, which is a consequence of possible contamination of the sample and its stratigraphic location.

Analysis of the results of radiometric dating of the Late Middle Paleolithic and Early Upper Paleolithic layers from the Georgian Ortvala-klde cave (Adler et al., 2008) suggests a long extension of the Upper Paleolithic age.-

Table 3. Key radiocarbon dates for parking

* Calibration using the IntCal 09 calibration curve [Reimer et al., 2009] and the OxCal 4.1 computer program.

** Calibration using the CalPal2007_Hulu calibration curve and CalPal software [Weninger and Joris, 2008].

habitation of Neanderthals in this region. About 38 000 - 34 000 l. n. (40 000 - 37 000 cal. l. n), their relatively rapid disappearance and settlement of the territory by human populations of the modern anthropological type took place. The study of objects in the adjacent territories of the South and North Caucasus revealed some discrepancy in the AMS dates. In particular, the age of the final Middle Paleolithic deposits in the Mezmay cave was determined to be close to 32,000 BP (35,000 cal. bp). The AMS date for the Kalavan-2 site is close to the last result. However, additional radiocarbon and EPR dating is required to provide greater confidence in the reliability of this chronological definition.

Conclusion

Some provisions justified above allow us to make a preliminary assessment of the sediments of the Kalavan-2 site. We consider this site promising for understanding the cultural characteristics of the ancient population not only in Armenia, but also in the Lesser Caucasus as a whole. The industry of layer 6, which reflects the predominance of tronking techniques, can be associated with the Zagros-Tavria mousse. The materials of layer 6 are an example of the classical Mousterian of Armenia with Levallois splitting, aimed at obtaining Levallois plate blanks and Levallois points. This industry and the faunal complex, which includes bison, wild goat, and red deer, belong to the oxygen-isotope stage 3, taking into account the radiocarbon date of 34,200 ± 360 BP.Special attention should be paid to the presence of several Mousterian artifacts in the lower layers 9 and 11. Sedimentation conditions, when erosion horizons with fine-grained sediments were replaced by larger detrital formations coming from the slopes, make it problematic to more accurately determine the archaeological material.

The site of Kalavan-2, located at an altitude of 1,600 m above sea level in the Barepata Valley, was most likely chosen by a group of hunter-gatherers because of the abundance of raw stone in the riverbed. In addition, the proposed paths of seasonal migration of ungulates moving in the spring to the upper summer pastures and descending in the autumn to reproduce and feed their young are established here [Adler et al., 2006]. On the opposite bank of the Barepat, a few hundred meters from the Kalavan-2 site, the Kalavan-1 hunting camp site was discovered and excavated, dating back to the end of the Upper Paleolithic; radiocarbon dates are available for it: 14,070 ± 60 and 13,750 ± 60 BP. the genus Capra (Liagre and Balasescu, 2007). It seems that the area where the camps were located allowed Mousterian and Upper Paleolithic hunters to plan and carry out hunting trips.

For a better understanding of the structure and post-depositional changes in the cultural deposits of the Kalavan-2 site, the results of all previous studies should be supplemented with data from systematic exploration of the Barepata Valley, geoarchaeological and geomorphological surveys. If new archaeological complexes are discovered, new radiometric dates and cultural (typological and technological) characteristics will be required. All this will help to reconstruct the natural environment and behavioral and lifestyle features of hominid groups that lived in the Upper Pleistocene in Armenia. Finally, since the rock base at Kalavan-2 has not yet been reached, the possibility of detecting anthropogenic activity in the lower, 20th layer may lead to the discovery of older than the Middle Paleolithic archaeological horizons of this site.

Acknowledgements

Excavations at the Kalavan-2 site were carried out by the Armenian-French Mission, the Institute of Archeology and Ethnography of the Academy of Sciences of the Republic of Armenia (Director P. Avetisyan; head of excavations from the Armenian side B. Gasparyan), the archaeological mission "Caucasus" under the Ministry of Foreign Affairs of France (Director K. A. Abramovich). The head of excavations on the French side, D. Colonge). We thank the Gfeller Foundation (USA) for their financial support for the project, as well as colleagues Liore Kolske Horwitz, John D. Speth, Marcel Ott, and Damien Fleasa for their helpful comments in preparing this text.

List of literature

Jafarov A. K. Mousterian culture of Azerbaijan (based on the materials of the Tatar Cave). Baku: Elm Publ., 1983, 97 p.

Yeritsyan B. G. Yerevan cave site and its place among the oldest monuments of the Caucasus: abstract of the dissertation of the Candidate of Historical Sciences. - M., 1970. - 32 p.

Yeritsyan B. G. New Lower Paleolithic cave site Lusakert-1 (Armenia) / / KSIA. -1975. - Issue 141. - pp. 42-50.

Lyubin V. P. Paleolith of the Caucasus / / Paleolith of the World: Paleolith of the Caucasus and Northern Asia. - L.: Nauka, 1989. - pp. 9-144.

Mansurov M. M. Novye paleoliticheskie pamyatniki v srednego basseine r. Kura [New Paleolithic monuments in the middle basin of the Kura River]. Tbilisi: Metsniereba Publ., 1990, pp. 40-42.

Nioradze, M., Cave Paleolithic sites of the Tskalitsela River Gorge, Paleolit Kavkaza i sopredel'nykh territorii. Tbilisi: Metsniereba Publ., 1990, pp. 55-60.

Tushabramishvili D. Paleolith of Georgia. State Museum of Georgia. - 1984. - Vol. 37B. - pp. 5-27.

Tushabramishvili N. The Middle Paleolithic of Western Georgia and the stages of its transition to the Upper Paleolithic (based on the materials of Ortvala-klde): abstract of the dissertation of the Candidate of Historical Sciences. - Tbilisi, 1994. - 31 p.

Adler D.S. Tate Middle Palaeolithic Patterns of Lithic Reduction, Mobility, and Land Use in the Southern Caucasus: Unpublished Ph.D. thesis. - Harvard, 2002. - 488 p.

Adler D.S., Tushabramishvili N. Middle Palaeolithic patterns of settlement and subsistence in the Southern Caucasus // Settlement Dynamics of the Middle Palaeolithic and Middle Stone Age / eds. N. Conard, A. Kandel. - Tubingen: Kerns Verlag, 2004. - P. 91 - 132.

Adler D.S., Bar-Oz G., Belfer-Cohen A., Bar-Yosef O. Ahead of the game - Middle and Upper Palaeolithic hunting behaviors in the Southern Caucasus // Current Anthropology. - 2006. - Vol. 47. - P. 89 - 118.

Adler D.S., Bar-Yosef O., Belfer-Cohen A., Tushabramishvili N., Boaretto E., Mercier N., Valladas H., Rink W.J. Dating the demise: Neandertal extinction and the establishment of modern humans in the southern Caucasus // J. of Human Evolution. - 2008. - Vol. 55(5). - P. 817 - 833.

Badalyan R., Chataigner C., Kohl R. Trans-Caucasian Obsidian: The Exploitation of the Sources and Their Distribution // A view from the Highlands: Trans-Caucasus, Eastern Anatolia and Northwestern Iran / ed. A. Sagona. - Eeuwen: Peteers Press, 2004. - P. 437 - 465.

Badgley C. Taphonomy of mammalian fossil remains from Siwalik rocks of Pakistan // Paleobiology. - 1986. - Vol. 12. - P. 119 - 142.

Barge O., Chataigner C. Un SIG pour l'analyse des approvisionnements: l'exemple de l'obsidienne en Armenie // Revue d'Archeometrie. - 2004. - Vol. 28. - С. 25 - 33.

Bar-Oz G., Adler D.S. Taphonomic history of the Mddle and Upper Palaeolithic assemblage from Ortvale Klde, Georgian Republic // J. of Taphonomy. - 2005. - Vol. 3. - From 185-211.

Bar-Oz G., Belfer-Cohen A., Meshveliani T., Djakeli N., Bar-Yosef O. Taphonomy and zooarchaeology of the Upper Palaeolithic cave of Dzudzuana, Republic of Georgia // International J. of Osteoarchaeology. - 2008. - Vol. 18. - P. 131 - 151.

Bar-Yosef O., Kuhn S. The big deal about blades: Eaminar technologies and human evolution // American Anthropologist. - 1999. - Vol. 101. - P. 322 - 338.

Bar-Yosef O., Belfer-Cohen A., Adler D.S. The implications of the Mddle-Upper Palaeolithic Chronological Boundary in the Caucasus to Eurasian Prehistory // Anthropologic. - 2006. - Vol. 46. - P. 49 - 60.

Baryshnikov G. Bison from early Paleolithic sites in the Caucasus and Crimea // Ге bison: gibier et moy en de subsistance des hommes du Paleolithique aux Paleoindiens des grandes plaines: Actes du colloque international - Antibes, 1999. - P. 319 - 331.

Baryshnikov G., Hoffecker J. Mousterian Hunters of the NW Caucasus: Preliminary Results of Recent Investigations // J. of Field Archaeology. - 1994. - Vol. 21. - P. 1 - 14.

Baumler M.F., Speth J.D. A Mddle Paleolithic assemblage from Kunj i cave, Iran // The Paleolithic Prehistory of the Zagros-Taurus / eds. D.I Olszewski, H.L. Dibble. - Philadelphia: University of Pennsylvania, the University Museum, 1993. - P. 1 - 73. - (University Museum Monograph 83; University Museum Symposium Series 5).

Beliaeva E.V., Lioubine V.R. The Caucasus-Levant-Zagros: Possible Relations in the Middle Palaeolithic // Prehistoire d'Anatolie: Genese de deux mondes (Anatolian Prehistory at the Crossroads of two worlds) / ed. M. Otte. - Liege: Universite de Liege, 1998. - P. 39 - 55. - (ERAIII. 85; vol. I).

Boessneck J., Muller H.H., Teichert M. Osteologische Unterscheidungsmerkmale zwischen Schaf (Ovis aries Linne) und Ziege (Capra hircus Einne) // Kuhn-Archiv. - 1964. - Vol. 78. - P. 1 - 129.

Brugal J.-R. Le Bos primigenius Boj. 1827 du pleistocene moyen des grottes de Lunel-Viel (Herault) // Bull, du Musee d'Anthropologie prehistorique de Monaco. - 1984. - Vol. 28. - P. 7 - 62.

Brugal J.-R. La faune des grands mammiferes de l'abri des Canalettes, materiel 1980 - 1986 // L'abri des Canalettes / ed. L. Meignen. - P.: CNRS (Monographie du CRA), 1993. - P. 89 - 137.

Chataigner C., Barge O. Quantitative approach to the diffusion of obsidian in the ancient northern Near East // Layers of Perception. Proceedings of the 35th International Conference on Computer Applications and Quantitative Methods in Archaeology (CAA) / eds. A. Posluschny, K. Lambers, I. Herzog. - Bonn: Dr. Rudolf Habelt GmbH Verlad, 2008. - 375 p.

Clutton-Brock J., Dennis-Brian K., Armitage P., Jewel P. Osteology of the Soay sheep // Bull, of the British Museum. Zoology Series 1. - 1990. - Vol. 56. - P. 1 - 56.

Cohen V., Stepanchuk V. Late Middle and Early Upper Palaeolithic evidence from the east European plain and Caucasus: an new look at variability, interactions and transitions // J. of World Prehistory. - 1999. - Vol. 13. - P. 265 - 319.

Colonge D., Gasparyan B., Ghukasyan R., Nahapetyan S. Le gisement de Kalavan-2 // Mission Caucase, Rapport Scientifique 2006 / ed. C. Chataigner-P: Ministeredes Affaires Etrangeres, 2006. - P. 14 - 19.

Colonge D., Nahapetyan S., Monchot H. Le gisement de Kalavan-2 // Caucase, Rapport Scientifique 2007 / ed. C. Chataigner. - P.: Ministere des Affaires Etrangeres, 2007. - P. 15 - 29.

Dibble H.L., Holdaway S.J. Le Paleolithique moyen de l'abri sous roche de Warwasi et ses relations avec le Mousterien du Zagros et du Levant // L'Anthropologiero -1990. - Vol. 94. - P. 619 - 641.

Dibble H.L., Holdaway S.J. The Middle Palaeolithic Industries of Warwasi // The Paleolithic Prehistory of the Zagros-Taurus / ed. D. Olszewski, H. Dibble. - P.: University of Pennsylvania, the University Museum, 1993. - P. 75 - 99. - (University Museum Monograph 83; University Museum Symposium Series 5).

Doronichev V.B. Lower Palaeolithic occupation of the Northern Caucasus // Early Humans at the Gate of Europe: Proceedings of the first international symposium, Dmanisi, Tbilisi (Georgia), September 1998. / ed. D. Lordkipanidze, O. Bar-Yosef, M. Otte. - Tbilisi, 2000. - P. 67 - 77.

Eisenmann V. Etudes des dents jugales inferieures des Equus (Mammalia, Perissodactyla) actuels et fossiles // Palaeovertebrata. - 1981. - Vol. 10. - P. 127 - 226.

Fernandez H. Osteologie comparee des petits ruminants eurasiatiques sauvages et domestiques (genre Rupicapra, Ovis, Capra et Capreolus). Diagnose differentielle du squelette appendiculaire: These de doctorat, Biologie. - Geneve, 2001. - 465 p.

Fernandez P., Cregut-Bonnoure E. Les Caprinae (Rupicaprini, Ovibovini, Ovini et Caprini) de la sequence pleistocene de Kozarnika (Bulgarie du Nord): morphometrie, biochronologie et implications phylogeniques // Revue de Paleobiologie. - 2007. - Vol. 26(2). - P. 425 - 503.

Fourloubey C. Recherches sur le Paleolithique // Mission Caucase, Rapport Scientifique 1999 / ed. C. Chataigner. - P.: Ministere des Affaires Etrangeres, 1999. - P. 7 - 23.

Fourloubey C., Beauval C., Colonge D., Liagre J., Olivier V., Chataigner C. Le Paleolithique en Armenie: etat des connaissances acquises et donnees recentes // Paleorient. - 2003. - Vol. 29. - P. 5 - 18.

Golovanova L.V., Doronichev V.B. The Middle Palaeolithic of the Caucasus // J. of World Prehistory. - 2003. - Vol. 17. - P. 71 - 138.

Liagre J., Balasescu A. Kalavan-1, une (ou plusieurs) occupations du paleolithique superieur final // Mission Caucase, Rapport Scientifique 2007 / ed. C. Chataigner. - P.: Ministere des Affaires Etrangeres, 2007. - P. 30 - 40.

Liagre J., Gasparyan B., Olivier V., Nahapetyan S. The site of Angeghakot 1 (Republic of Armenia) and the identification of the Mousterian cultural facies of "Yerevan points" type in the Southern Caucasus // Paleorient. - 2006. - Vol. 32. - P. 5 - 18.

McCuaig Balkwill D., Cumbaa S.L. A guide to identification of postcranial bones of Bos taurus and Bison bison. - Ottawa: Musee canadien de la nature, 1992. - 277 p. - (Syllogeus; vol. 71).

Meignen L. Early Middle Palaeolithic Blade Technology // Southwestern Asia / ed. D. Tushabramishvili. - 2000. - Vol. 19. - P. 158 - 168.

Meignen L., Tushabramishvili N. Paleolithique moyen laminaire sur les flancs sud du Caucase : productions lithiques et fonctionnement du site de Djruchula (Georgie) // Paleorient. - 2006. - Vol. 32. - P. 81 - 104.

Morlan R.E. Spiral fractures on limb bones: which ones are artificial? // Carnivores, humans scavengers and predators: a question of bone technology / eds. G.M Lemoine, A.S. MacEachern. - Calgary: University of Calgary Archaeological Association, 1983. - P. 241 - 269.

Morlan R.E. Bison bone fragmentation and survivorship: a comparative method // J. of Archaeological Science. - 1994. - Vol. 21. - P. 797 - 807.

Mottershead D.N., Baily B., Collier P., Inkpen R.J. Identification and quantification of weathering by plant roots // Building and Environment. - 2003. - Vol. 38. - P. 1235 - 1241.

Nioradze M., Otte M. Paleolithique superieur de Georgie // L'Anthropologic. - 2000. - Vol. 104. - P. 265 - 300.

Olivier V., Nahapetyan S. Analyse geomorphologique des sites de Kalavan-1 et Kalavan-2 // Mission Caucase, Rapport Scientifique 2008 / ed. C. Chataigner. - P.: Ministere des Affaires Etrangeres, 2008. - P. 4 - 8.

Otte M., Yalcinkaya I., Taskiran H., Kozlowski J.K, Bar-Yosef O., Noiret P. The Anatolian Middle Paleolithic: New Research at Karain Cave // J. of Anthropological Research. - 1995. - Vol. 51. - P. 287 - 299.

Prummel W., Frisch H.J. A guide for the distinctions of the species, sex and body size in bones of sheep and goat // J. of Archaeological Science. - 1986. - Vol. 13. - P. 567 - 577.

Reimer P.J., Baillie M.G.L., Bard E., Bayliss A., Beck J.W., Blackwell P.G., Bronk Ramsey C., Buck C.E., Burr G.S., Edwards R.L., Friedrich M., Grootes P.M., Guilderson T.P., Hajdas I., Heaton T.J., Hogg A.G., Hughen K.A., Kaiser K.F., Kromer B., McCormac G., Manning S.W., Reimer R.W., Richards D.A., Southon J.R., Talamo S., Turney C.S.M., van der Plicht J., Weyhenmeyer C.E. Intcal09 and Marine09 radiocarbon age calibration curves, 0 - 50,000 years cal BP // Radiocarbon. - 2009. - N51(4). - P. 1111 - 1150.

Slott-Moller R. La faune // Les chasseurs d'Aurochs de la Borde. Un site du paleolithique moyen (Livernon, Lot) / eds. J. Jaubert, M. Lorblanchet. - 1990. - Vol. 27. - P. 33 - 68.

Todisco D., Monchot H Bone weathering in a periglacial context: the Tayara site (KbFk-7), Qikirtaq island, Nunavik (Canada) //Arctic. - 2008. - Vol. 61. - P. 87 - 101.

Weninger B., Jöris O. A ¹⁴C calibration curve for the last 60 ka: the Greenland-Hulu U/Th timescale and its impact on understanding the Middle to Upper Palaeolithic transition in Western Eurasia // J. of Human Evolution. - 2008. - Vol. 55. - P. 772 - 781.

The article was submitted to the Editorial Board on 10.02.10.

Permanent link to this publication: