Mammalian Remains from the Upper Palaeolithic Site of Kamenka, Buryatia (Siberia)

Mammalian Remains from the Upper Palaeolithic Site of Kamenka, Buryatia (Siberia)

Journal of Archaeological Science (1996) 23, 35–57 Mammalian Remains from the Upper Palaeolithic Site of Kamenka, Buryatia (Siberia) Mietje Germonpré...

877KB Sizes 0 Downloads 3 Views

Journal of Archaeological Science (1996) 23, 35–57

Mammalian Remains from the Upper Palaeolithic Site of Kamenka, Buryatia (Siberia) Mietje Germonpré Department of Palaeontology, Royal Belgian Institute of Natural Sciences, Vautierstraat 29, 1040 Brussels, Belgium

Ludmila Lbova Buryat Institute of Social Sciences, Siberian Division of the Russian Academy of Sciences, Sakhyanovoistreet 8, 670042 Ulan Ude, Russia (Received 28 May 1994, revised manuscript accepted 14 November 1994) A taphonomic analysis of the mammalian remains from the early Upper Palaeolithic open-air site of Kamenka was undertaken. The site is located in the Transbaikal area near Ulan Ude, the capital of the autonomous republic of Buryatia. The faunal assemblages originate from several archaeological complexes. However, only the assemblage from the Kamenka Complex A 1993 excavation was analysed in detail. Taphonomic arguments indicate that the subsistence strategy was based primarily on two game animals: Mongolian gazelle and horse. Other species present probably belong to the group of penecontemporaneous intrusives. Kamenka functioned as a camp site that most likely was seasonally occupied during late summer/fall and/or early winter. ? 1996 Academic Press Limited Keywords: EARLY UPPER PALAEOLITHIC, SIBERIA, TRANSBAIKAL, TAPHONOMIC ANALYSIS, MONGOLIAN GAZELLE, HORSE.

Introduction he investigation of faunal assemblages from single sites can help to establish the relationship between subsistence practices, patterns of cultural development and changing palaeoclimatic and palaeoecological conditions in the Late Pleistocene of eastern Siberia as well as to help build an understanding of the nature of the Middle-to-Upper Palaeolithic transition in the region. The co-existence of Mousterian-type tools with tools that are within the range of the Upper Palaeolithic has been reported in hundreds of Siberian sites (Vasilyev, 1993). In the Transbaikal (Buryatia), the Acheulo-Mousterian was until recent times practically unknown, but some surface finds have been collected in the last decade. Early Upper Palaeolithic sites are, however, well represented (Konstantinov, Bazarova & Lbova, 1994). Due to its clear stratigraphic context and rich assemblages of faunal remains, the open-air site of Kamenka reveals much about the early Upper Palaeolithic of the area concerned. The site is located in the Transbaikal region, some 60 km southeast of Ulan Ude, the capital of the Republic of Buryatia (Figure 1), in the central part of the valley of the Brianka River (Selenga river basin). The valley, flanked by hills of 100–150 m, has a width of 3–5 km, an altitude of 580–590 m and is

T

RUSSIAN FEDERATION SIBERIA Lake Baikal

Buryatia

1.000 km Figure 1. Map of Siberia.

typical for the Transbaikal mountainous system. In the Transbaikal region, early Upper Palaeolithic sites are generally associated with mountainous slopes and the Kamenka site is located on the southeast slope of the Kamenka mountain, some 2 km from the present course of the river (Figure 2). The area is situated in the Hangay-Daurian mountain forest-steppe province, which is influenced by a highly continental climate with the maximum precipitation falling in July and August (Lavrenko & 35

0305-4403/96/010035+23 $12.00/0

? 1995 Academic Press Limited

M. Germonpre´ and L. Lbova

ka

36

Br yan

Kamenka Mountain

Kamenka

60

Novaya Bryan

0m

0m

60

600 m

m 0 1000

Lake Village Figure 2. Topographic map of the region of Kamenka.

Karamysheva, 1993). Mean annual precipitation averages 250 mm (Galazii, 1969). Winter temperatures are low with a mean for January of about "20)C, mean July temperature fluctuates around 15)C (Bugrimova, 1988). Excavations at the archaeological site of Kamenka were carried out during the years 1989–1993 by archaeological expeditions of the Buryat Institute of Social Sciences of the Siberian Division of the Russian Academy of Sciences, under the leadership of Ludmila Lbova. Some archaeological results have already been published (Lbova, 1991, 1993; Lbova & Volkov, 1993). Five archaeological complexes were recognized in different positions (Figure 3), three occur at the same stratigraphical position (Complex A 1992, Complex A 1993, Complex C 1992), two are present higher up in the sequence (Complex B 1992, Complex B 1993).

Stratigraphic Setting The Palaeolithic cultural complexes occur in polygenetic slope sediments at a depth of 7–9 m from the original surface. The stratigraphic profile can be summarized as follows (Figure 3). Layer 1 contains the modern soil. Layers 2–3 consist of grey sands with four levels of soil formation in layer 2. Layer 4 is composed of light yellow, fine grained sands with horizontal stratification. Layer 5 shows a horizontal banding of fine grained yellow, brown and dark grey sands with clay lenses; at the base of the layer ice wedges occur. Layer 6 consists of three parts: upper part (brown sand mixed with clay and small gravel), middle part (zone of calcification in brown clayey sand) and lower part

(brown sand, more clayey than in the upper and middle part, with lenses of yellow sand and dark bands of palaeosoil). Layer 7 consists of brown gravelly sands with bluish grey clay horizons and lenses of calcification. Levels 1–3 are colluvial deposits, layers 4 and 5 are fluvial, layers 6 and 7 colluvial. Layers 1–3 represent the Holocene, layers 4–7 have been assigned to the Pleistocene. We believe that layers 4 and 5 were deposited during the cold Sartan Stadial. The Sartan Stadial is part of the Zyriansk glacial (Last Glacial) and lasted from 24,000–22,000 years  to 11,000– 10,000 years  (Vorob’eva et al., 1990; Vasilyev, 1993). Based on palaeopedological features, one can assume that layers 6 and 7 were formed during the Karginsk mega-interstadial under a warmer climate. The Karginsk interstadial dates from 50,000/45,000 years  to 24,000/22,000 years  (Vorob’eva et al., 1990; Vasilyev, 1993). The Palaeolithic assemblages are included in layer 6, Complex B in the upper part and Complexes A and C in the lower part. The latter two complexes can be correlated on the basis of their stratigraphical, palaeopedological and typological characteristics. For Complex B two radiocarbon dates were obtained by the Laboratory of the Radiocarbon Method, Geological Institute of the Siberian Division of the Russian Academy of Sciences (Novosibirsk) on two bone samples: 28,815&150 years  (SD RAS-3032) and 28,060&475 years  (SD RAS-2903). The same laboratory also dated two bones of Complex A; however, these dates show a greater difference: 31,060&530 years  (SD RAS-3133) and 35,845&695 years  (SD RAS-2903) (Orlova, 1994).

Excavated Features In Complex A 1993 (picket I-12, Figure 3), the following features were excavated: a zone of artefact and bone concentration (squares C2-3/D2-3) with a diameter of 2 m, which includes a hearth; another fire place nearby (square D1-2/E1-2); a pit with a depth of 25 cm and a diameter of 50 cm containing several broken bones and stones (square E2) and a semicircular stone structure (squares G4/G5/H5). In Complex A 1992 (picket J-10, Figure 3), three zones of human activity were recognized: a concentration of bones with a few artefacts and two constructed hearths with artefacts. Complex C 1992 (picket E/F-13, Figure 3) had four aligned zones of concentration of artefacts and bones each with a diameter of about 0·8 m. In Complex B 1992 (picket E/F-13, Figure 3), the artefacts are distributed much more evenly over the surface, and only one zone of artefacts (with flakes, core nuclei and blades) could be described as a workshop. Complex B 1993 (picket I-12, Figure 3) only yielded a few artefacts.

Typology of the Artefacts The Palaeolithic Complexes A/C and B differ in morphological, typological and technological

Mammalian remains from the Upper Palaeolithic site of Kamenka, Buryatia (Siberia) 37

O N M L K IV

J

II

I

I

H G F

III

E 0m

I

D 1 C 2

B A

2 3

16 15 14 13 12 11 10

9

8

7

6

5

4

3

2

1 10 m

4 4 II 5

5 B

6 6 A

B

6

8

III 5 B 6 C

A 10

7

Figure 3. Location of the excavated areas and stratigraphic profiles I: picket H-14/15; II: Complex A and B (excavation 1993); III: Complex B and C (excavation 1992); IV: Complex A (excavation 1992); : holocene soil; : banding of palaeosoil; : zone of calcification; : sand; : clay; : gravel; : artefact.

characteristics of the artefacts. Most tools of Complexes A and C were manufactured on blades. Microwear analysis was done by P. Volkov of the Institute of Archaeology and Ethnography of the Siberian Division of the Russian Academy of Sciences (Novosibirsk) (Lbova & Volkov, 1993). The mean dimensions of the blades are 8–12 cm in length, 2 cm in

width and 0·5–1 cm in thickness. The blades were produced from prismatic cores. The industry of Complexes A and C is, in general, characterized by a faceted striking platform (index of faceting is 34). The tools (29·9%) contain a series of knives (50%) with one or two subparallel and converging working edges and predominance of dorsal design (Figure 4: (a), (d), (e),

38

M. Germonpre´ and L. Lbova (a)

(b)

(c)

(d)

(e)

A B

B (f)

(g)

(h)

(i) A

B A

A

Figure 4. Kamenka Complex A (length of scale: 3 cm). (a) Point (knife), (b) point, (c), (f), (h) A: knife; B: scraper, (d), (e), (g), knife.

(g)). The tool kit further includes scrapers (10%) with lateral and offset position of the working edge (Figure 4: (c), (f), (h)), notched and denticulated pieces (15·5%) and a few burins. Important also is the presence of several bone tools. In general, the industry of the Complexes A and C is reminiscent of a Mousterian of Levallois blade facies technological tradition. However, some technological and typological characteris-

tics such as the predominance of parallel blades, the technology of splitting, the presence of end scrapers of the Aurignac type, a few burins and the uniform bone tools suggest an attribution of these Complexes to the early Upper Palaeolithic. The Palaeolithic Complex B displays a combination of different technologies, one is reminiscent of the Levallois technique, another follows an orthogonal

Mammalian remains from the Upper Palaeolithic site of Kamenka, Buryatia (Siberia) 39 (a)

(b)

(c)

(d)

(e)

(f)

(g)

(h) (i)

Figure 5. Kamenka Complex B (length of scale: 3 cm). (a) Point (knife), (b) knife, (c), (g), (h) scraper, (d) burin, (e) scraper (Kunalei type), (f) chizel, (i) adze.

system and a third one represents incipient microblade technique. Most tools were manufactured on flakes. The mean sizes are 4–6 cm in length, 2–3·5 cm in width and 1–1·5 cm in thickness. The tool kit is dominated by a series of end scrapers (Figure 5: (c), (g), (h)) and carinate end scrapers of the ‘‘Kunaleisky’’ type (Figure 5: (e)); knives, points, chisels, burins and punches are also present (Figure 5: (a), (b), (d), (f), (i)). The typological and morphological characteristics of the industries demonstrate that two different

cultural traditions were present at the beginning of the Upper Palaeolithic in the Transbaikal region. Complexes A and C have close affinities with the Tolbaga industry (Tolbaga, Varvarina Gora, Sapun, Podzvonkaja, Kara-Bom); Complex B can be compared with the Kunalei industry (Kunalei, Priiskovoe) (Bazarov et al., 1982; Konstantinov & Konstantinov, 1991; Cauwe et al., 1993; Goebel, Derevianko & Petrin, 1993; Lbova, 1993; Tashak, 1993; Konstantinov et al., 1994).

40

M. Germonpre´ and L. Lbova Table 1. List of mammalian species from Kamenka Complex A 1993, Kamenka Complex A 1992, Kamenka Complex C 1992, Kamenka Complex B 1993 Kamenka Complex Species Mammuthus primigenius Equus caballus Equus hemionus Coelodonta antiquitatis Camelus sp. Megaloceros giganteus Bison priscus* Spiroceros kiakhtensis Procapra gutturosa Ovis ammon Bovidae Panthera leo

A 1993 NISP/MNI

A 1992 NISP/MNI

C 1992 NISP/MNI

B 1993 NISP/MNI

28/4 1/1 2/1

13/2 1/1

2/1

8/1

1/1

3/1

2/1 3/2 3/–

1/1

1/1

1/1 150/6 5/1 3/1 2/1 5/1 1/1 225/6 4/1 11/– 1/1

Total

407/18

47/10

16/5

7/4

indet

NF 1521

NF 3

NF 0

NF 8

Total

1928

50

16

15

indet: unidentified MNI: Minimal Number of Individuals NF: Number of Fragments NISP: Number of Identified Specimens *probably Bison priscus although Poëphagus baikalensis and Bos primigenius are not excluded.

Faunal Analysis Introduction Only the faunal remains of Complex A 1993 will be described in detail because both authors were present during the excavation of this complex. The sampling of Complex A 1992, Complex C 1992, Complex B 1992 and Complex B 1993 may have been conducted in a different way and the collections are very restricted. The remains of these collections will be discussed only if they provide additional information or are different from those of Complex A 1993. Analysis of the bone material includes calculation of the Number of Fragments (NF), calculation of the Minimum Number of Individuals (MNI), the description of preservation, weathering, naturally induced traces and human utilization patterns (bone breakage and cut marks). These data help to distinguish the game animals from the background fauna and they suggest implications for the manner in which the carcasses of the game were transported, butchered, processed and utilized. Table 1 gives the Number of Identified Specimens per species (NISP) and the estimated MNI per species of the different complexes. MNI were established by matching left and right elements. The elementwise NISP and MNI are indicated only for species represented with more than 25 specimens in each complex (Tables 6 & 13). The number of fragments of unidentified specimens is also presented in Table 1. Although

unidentified fragments are sometimes seen only as sources of error (Ringrose, 1993), they are considered herein to be the product of various taphonomic pathways and have therefore been included in the analysis. Species composition of the assemblage The identifiable component of the Kamenka Complex A excavation of 1993 constitutes about a fifth of the total assemblage: 407 of the 1928 bone remains were identified. All recovered bones were examined. The two best represented species, both in NISP and MNI, are the Mongolian gazelle, Procapra gutturosa and the horse, Equus caballus. The other taxa include Coelodonta antiquitatis, Camelus sp., Megaloceros giganteus, Bison priscus, Spirocerus kiakhtensis, Ovis ammon, unidentified bovids and Panthera leo (Table 1). The restricted collection of Complex A 1992 consists almost completely of identifiable bones. The taxa present are Equus caballus, Equus hemionus, Coelodonta antiquitatis, Bison priscus, Procapra gutturosa, Ovis ammon and unidentified bovids (Table 1). Only 16 identifiable bones from the Complex C 1992 excavation were curated, they are derived from Equus caballus, Coelodonta antiquitatis, Bison priscus and Procapra gutturosa (Table 1). The collection of Complex B 1993 also is very limited. Following species were identified: Mammuthus primigenius, Coelodonta antiquitatis, Bison priscus and Procapra gutturosa (Table 1).

Mammalian remains from the Upper Palaeolithic site of Kamenka, Buryatia (Siberia) 41 Table 2. Number of fragments and relative frequency of the bones per Behrensmeyer (1978) weathering stage (Kamenka Complex A 1993) Weathering stage

NF

%

Stage Stage Stage Stage

1023 772 98 15

53·6 40·5 5·1 0·8

1908

100·0

Table 3. Number of fragments and relative frequency of the bones in Behrensmeyer (1978) weathering stages 2+3 (Kamenka Complex A 1993) Species

0 1 2 3

Total

Preservation and weathering The bones of the Complex A 1993 assemblage are in an excellent state of preservation. Their colour varies from a dark greyish brown to a lighter brown, but some bones exhibit a light grey hue. Most of the bones are fresh or have only slightly weathered surfaces. Compared with the weathering stages defined by Behrensmeyer (1978), the bones belong mostly to weathering stage 0 and weathering stage 1 (Table 2). We agree with Lyman & Fox (1989) that the interpretation of the set of weathering stages of an assemblage is a complex matter, because weathering not only depends on the numbers of years since death but also on skeletal element, taxon, depositional microenvironment, accumulation history and time of exposure. However, in a very general way, some trends can be recognized. The lower frequency of bones in a later stage of weathering suggests that the bulk of the Complex A 1993 remains were only exposed for a short time on the surface or in the soil prior to deeper burial. The frequency distribution of the taxa in weathering stage 0 and 1 is comparable with the distribution of the complete assemblage. This is not surprising since altogether these stages contain 95% of the remains. The sample of weathering stages 2+3 is different, it has a much lower percentage of horse and gazelle remains, but higher frequencies of woolly rhinoceros, camel, giant deer, bison, extinct antelope, argali sheep, undetermined bovids and lion (Table 3). Both the weathering and changed frequencies point to some differences in the origin of this part of the assemblage, which will be explained later in further detail. There are no clear indications for a heavy post-depositional destruction of the bony material. Plant roots left only traces on some bones (Table 5), corrosion marks are completely absent. The bones of the small collections of Complexes A 1992 and C 1992 are relatively fresh, those of the equally small sample of Complex B 1993 are in a rather advanced stage of weathering, most of the bones belong to stage 2 or 3. Breakage Only 156 bones (8·1%) of the assemblage of Complex A 1993 are not broken, small compact bones such as

NISP

%

Equus caballus Coelodonta antiquitatis Camelus sp. Megaloceros giganteus Bison priscus Spiroceros kiakhtensis Procapra gutturosa Ovis ammon Bovidae Panthera leo

11 5 1 1 4 1 13 1 3 1

26·8 12·2 2·4 2·4 9·8 2·4 31·7 2·4 7·3 2·4

Total NISP

40

indet

NF 73

Total

113

100

Table 4. Skeletal element presentation of the Kamenka Complex A 1993 assemblage Elements

NF

%

horncore antler teeth skull mandibula rib vertebra sternum hyoid scapula humerus ulna radius MC pelvis femur patella tibia carp./tars. MT MP phalanx sesam unid. epiphysis unid. diaphysis

4 1 121 35 38 128 52 7 2 15 19 5 13 10 22 22 2 16 41 10 8 68 4 51 738

0·3 0·1 8·4 2·4 2·7 8·9 3·6 0·5 0·1 1·0 1·3 0·3 0·9 0·7 1·5 1·5 0·1 1·1 2·9 0·7 0·6 4·7 0·3 3·6 51·5

Total indet

1432 496

100·0

Total

1928

carpals/tarsals, phalanges and teeth score highest. The completeness indices for calcaneum, astragalus and centrotarsale, calculated as proposed by Marean (1991) are very high, fluctuating for gazelle between 87·5% (calcaneum) and 100% (astragalus and centrotarsale). One calcaneum and one astragalus of horse display carnivore tooth marks. More than 50% of the bones of gazelle are not broken compared with only

42

M. Germonpre´ and L. Lbova

Table 5. Frequency distribution of the bones of the Kamenka Complex A 1993 assemblage with traces Traces

NF

% Total

gnawing marks root marks burned fragments cut marks percussion marks cancellous blocks bone splinters

36 218 26 19 21 64 203

1·9 11·3 1·3 1·0 1·1 3·3 10·5

Total traces

587

Total NF

1928

60

%

40

20

0

3–5 cm 0–2 cm

6–8 cm

9–11 cm 15–17 cm 21–23 cm 12–14 cm 18–20 cm

Figure 6. Frequency distribution of the length of the bone fragments from the Kamenka Complex A 1993 assemblage.

21·4% for horse. In general, bones of smaller species tend to be more complete than those of larger animals in archaeological faunas (Klein, 1989). Complete long bones are very rare, almost 52% of the assemblage is composed of unidentified shaft fragments of long bones (Table 4). Most of the breakage patterns are undiagnostic, some 8·6% are spiral fractures and 0·6% could be described as sawtooth fractures, as defined by Shipman, Bosler and Davis (1981), indicating that these bones were broken in green condition. A few (0·5%) columnar fractures were identified. Figure 6 shows the distribution of the length of the fragments. About 79% of them are less than 6 cm long. The preponderance of small fragments is generally typical for an assemblage modified by humans. For example, at Pomongwe Cave, Zimbabwe, most of the bone flakes from human refuse have lengths of less than 5 cm (Brain, 1981). Bunn (1983) describes a bone assemblage of modern San hunter–gatherers in the Kalahari Desert, where extensive limb fragmentation occurred during marrow extraction. According to Davis & Fisher (1989), the huge number of broken bones at the 1000-year-old Los Terrace site, Montana, attest to bone breakage for marrow and grease rendering by prehistoric pronghorn hunters. Also at Denisova Cave, Altai, the high percentage of small fragments is thought to partly reflect the intensive human utilisation of the long bones (Germonpre´, 1993). It may be hypothesized that the fragmentary

condition of the bones at Kamenka Complex A 1993 reflects efforts to extract marrow and grease. Recent bone fractures were not detected on the Kamenka material. Traces The percentages of the bones with traces of the Kamenka Complex A 1993 assemblage is given in Table 5. Naturally induced Carnivore gnawing. Only a very slight percentage of the bones exhibit carnivore gnawing damage (N=26, 1·6%) indicating that the assemblage was not subjected to intensive carnivore activity. Within this context, the high occurrence of gnawing traces on bones of the Complex A 1992 sample is remarkable: 16 bones (32%) are affected. Plant root traces. Root marks are observed on some 10% of the remains of Complex A 1993. Root marking is probably not a recent phenomenon, since it occurs on deeply buried bones. The etched surfaces do not differ in colour from the unaffected surrounding bone, pointing to a root mark formation before fossilization started. Traces of human activity. Cut marks. Only a few cut marks (N=19) could be observed. The low incidence of cut marks is notable in the light of the high concentration of lithic artefacts, the associated occurrence of bone tools and the excellent preservation of the bone surfaces. Three stages of butchering can leave cut marks on bones: skinning, dismembering, and removal of meat (Davis & Fisher, 1989). The horncore of Spirocerus kiakhtensis has a series of cuts at its base (Figure 7), comparable cut marks were observed on two horncores of Procapra gutturosa. Possibly they resulted from skinning the head, perhaps to facilitate breaking the skull to reach the brain or to have easier access to the horncores which were used as raw material for bone tools. Skinning for clothing seems unlikely, at least for P. gutturosa, since cut marks are not present on the mandibles and phalanges. Transverse marks occur on the posterior surface above the distal epicondyle of a tibia of P. gutturosa (Figure 8) and on the distal epicondyle of another tibia fragment, both displaying a spiral fracture. They resemble the cut marks on the distal epiphyses of tibiae observed by Binford (1981) at Nunamiut sites. The cut marks, which he described as Td3, were made during dismembering. Almost 40% of the ribs of Mongolian gazelle have cut marks on the distal outer end, probably induced during skinning. Cut marks also occur on a cervical and a thoracal vertebra from this species. The cut marks on the thoracal vertebra correspond to the cut mark TV-2 described by Binford (1981). Filleting the tenderloin

Mammalian remains from the Upper Palaeolithic site of Kamenka, Buryatia (Siberia) 43

Figure 8. Tibia fragment of Procapra gutturosa (Mongolian gazelle) with cut marks (Kamenka Complex A 1993) (length of scale: 2 cm).

Figure 9. Cancellous block (Kamenka Complex A 1993): distal articular end of a femur of Procapra gutturosa (length of scale: 2 cm).

Figure 7. Horncore of Spirocerus kiakhtensis (extinct antelope) with cut marks (Kamenka Complex A 1993) (length of scale: 3 cm).

during primary butchery may produce this type of cut mark (Binford, 1981). Cut marks are also present on several unidentified shaft flakes and on the distal epiphysis of a tibia of horse, which according to Binford (1981) could be the product of dismembering. Percussion marks. Percussion marks occur on a small number (N=21, 1·1%) of long bones diaphyses and are almost always associated with spiral fracturing. They probably were produced when bones were fragmented in order to remove the marrow. Cancellous blocks. Some 3·3% of the bones have a conspicuous form: cubic pieces of cancellous bone shaped from epiphyses of long bones, central parts of basins or bodies of vertebrae. Most have a diameter of less than 6 cm (Figure 9). The crushed fragments display rather sharp edges and neat surface. One fragment shows even the trace of the breakage process

Figure 10. Cancellous block with trace of breakage action (Kamenka Complex A 1993) (length of scale: 3 cm).

(Figure 10). This special fragmentation can only be humanly induced and was probably inflicted to facilitate grease extraction by boiling of the cancellous portions. According to Binford (1981) cancellous

44

M. Germonpre´ and L. Lbova

portions of bones are pounded to increase the surface area of the bone exposed. They occur typically at the Nunamiut residential camps: ‘‘when women break marrow bones in residential locations, they are concerned with breaking the bone so that little compact and dense diaphyses remains attached to the articulator ends, which will later be used in the making of bone grease and bone juice.’’ Binford (1981, p. 158). Burning. Only a few bones are burned (1·3%). These are small unidentified fragments, the largest one has a length of 5 cm. It is not clear whether the burning was a result of roasting meat, intentional or accidental exposure of bones to fire. Bone splinters. Some 10% of the bones show a very special type of fragmentation, fragments in the shape of small splinters varying in length between 2–11 cm with a diameter of a few mm. They occur in the archeological horizon as clusters of 5 to 10 closely associated splinters. At the moment the significance of these splinters is not clear, but they could be byproducts of bone tool manufacturing. Bone tools and ornamented bones. Several bone tools and ornamented bone fragments occur in Complex A 1993. Most were found in the zone of concentration (squares C2-3/D2-3). They will be described in detail in a forthcoming paper. Notes on the species found The two best represented species of the Complex A 1993 assemblage, the Mongolian gazelle and horse, will be treated first. The osteometry of these two taxa is presented in Tables 7–12 and 14–19. The few measurements on the remains of the other species are mentioned in the text. All measurements are expressed in mm, unless otherwise stated. Most measurements were taken following the procedures indicated by von den Driesch (1976), most abbreviations used are also listed in that publication. Other abbreviations are: al for alveolar length, ch for crown height, cl for crown length, cw for crown width, Dob hcore for oraboral diameter of the horncore, Dt hcore for transversal diameter of the horncore, Dt burr for the transversal diameter of the burr of the antler, GB con. occ. for the greatest width of the condyli occipitales, GB for. magn. for the greatest width of the foramen magnum, H for. magn. for the height of the foramen magnum, L hcore for the length of the horncore. Following abbreviations for horse teeth were proposed by Eisenmann (1980, 1981): LP for the length of the protocone, Ip for the protocone index, LF for the length of the postflexide and IF for the postflexide index. Procapra gutturosa. Remains of the Mongolian gazelle (Procapra gutturosa) account for 55·3% of the NISP

Table 6. NISP and MNI of skeletal elements for Procapra gutturosa (Mongolian gazelle) from Kamenka Complex A 1993 Skeletal element horncore teeth M3 mand. other skull mandibula hyoid rib vertebra atlas axis other sternum scapula humerus diaphysis distal ulna proximal radius proximal diaphysis distal MC complete proximal distal pelvis femur proximal distal patella tibia proximal diaphysis distal carp./tars. calcaneum astragalus centrotarsale other MT complete proximal distal MP phalanx phalanx I phalanx II phalanx III Total

NISP

MNI

3 15

2 2 13

8 9 1 18 18

2 6 1

4 2 12 4 4 11

4 2 1 3 5

3 8 3

2 3

9

4 4 1 4

5

3 1 3 1

3 10

2 5 4 6

1 10

1 2 4 1 5

32 5 9 10 8 4

4 5 6 2

1 1 2 3 54 25 21 8 225

5 3 1 6

and 33·3% of the estimated MNI of the Complex A 1993 bone assemblage (Table 1). An inventory of the skeletal elements for gazelle is presented in Table 6. Almost all elements are represented; even a hyoid and a sternum, consisting of four associated elements, were recovered. Little variation in relative abundance of the skeletal elements exists. No important body parts are missing, suggesting that more or less whole carcasses of gazelles were brought to the site for processing and consumption. MNI estimates were calculated for a number of elements. The highest MNI estimate (6) was obtained on the mandible and the centrotarsale. At

Mammalian remains from the Upper Palaeolithic site of Kamenka, Buryatia (Siberia) 45 Table 7. Dimensions of skull fragments of Procapra gutturosa (Mongolian gazelle) from Kamenka Kamenka Complex A 1993

Cranium

GB con. occ L hcore Dt hcore Dob hcore

Kamenka Complex A 1992

N

Lim



N=1

1 1 2 2

43·2 110·4 20·7, 23·7 29·2, 30·0

22·2 29·6

24·8 30·1

Table 8. Dimensions of teeth of Procapra gutturosa (Mongolian gazelle) from Kamenka Complex A 1993

Table 9. Dimensions of skeletal elements of the forelimb of Procapra gutturosa (Mongolian gazelle) from Kamenka Complex A 1993

Teeth

Skeletal elements

N

Lim



ó

Scapula

3 3 3 3 8 8 4 3 1 4 3 3 1 4 3 1 1 1

31·2–33·8 25·3–25·4 20·5–23·2 15·1–17·5 26·3–29·6 25·6–28·8 24·0–30·3 25·3–27·2 17·1 23·6–26·2 21·9–23·4 13·3–16·4 158·1 21·4–22·9 15·7–18·0 13·4 22·1 16·2

32·3 25·3 22·2 16·0 28·5 27·2 27·4 26·5

1·1 0·0 1·2 1·1 1·0 1·1 2·4 0·9

24·9 22·5 15·3

0·9 0·6 1·4

22·1 16·9

0·6 0·9

Maxilla P2 P3 P4 M1 M2 Mandibula P2–P4 P2 P3 P4 M1 M2 M3

N

Lim

cl cw ch cl cw ch cl cw ch cl cw cl cw

1 1 1 1 1 1 1 1 1 1 1 1 1

9·6 7·8 13·1 9·3 8·8 14·8 15·9 11·4 15·3 17·9 11·1 17·6 10·1

al cl cw ch cl cw ch cl cw ch cl cw ch cl cw ch cl cw ch

1 2 2 2 2 2 2 2 2 1 2 2 1 2 1 1 l 1 2

24·0 5·6, 6·3 3·1, 3·5 5·3, 6·1 8·8, 9·2 4·4, 4·9 6·1, 8·8 9·7, 10·7 5·8, 5·8 5·7 11·4, 13·1 7·8, 8·1 4·6 13·8, 14·9 8·6 9·7 21·7 8·0 22·3, 25·1



Humerus Radius

Ulna MC

6·0 3·3 5·7 9·0 4·7 7·5 10·2 5·8 12·3 8·0 14·4

Table 10. Dimensions of skeletal elements of the hindlimb of Procapra gutturosa (Mongolian gazelle) from Kamenka Complex A 1993 Skeletal elements

N

Lim

Femur

1 3 4 4 1 2 2 1 1 3 3

48·6 19·4–21·90 22·7–25·7 17·6–20·5 158·8 19·5, 20·4 21·1, 21·6 12·4 14·1 21·8–26·8 16·2–16·7

Tibia MT

23·7

least two individuals were males, since only males of Procapra have horns. Nineteen immature ( juvenile or subadult) postcranial fragments were recovered. Deciduous teeth and teethbuds are not present. Few associations of remains definitively derived from one skeleton occur, four elements of a sternum were found together in square D2, a second and third phalanx in square A3, and two second and a third phalanx in square D2. In contrast, a left and right ulna were lying next to each other (square D2), which do not pertain to the same individual. Compared with the huge amount of broken remains in the total assemblage, the more than 40% recovery of

GLP LG BG KLC Bd BT Bp BFp KD Bd BFd BPc GL Bp Dp KD Bd Dd

Bp DC Bd Dd GL Bp Dp KD DD Bd Dd



ó

20·6 24·2 19·2

1·0 1·3 1·3

20·0 21·4 23·5 16·4

2·3 0·2

complete gazelle skeletal elements is surprising. The unbroken elements are mainly phalanges (50% of the total number of unbroken elements of this species) and carpals/tarsals (32%). The cut marks on the distal tibiae and the lack of broken carpals and tarsals, with the exception of two broken calcanea, may indicate that the wrist and ankle joints were dismembered by cutting, soon after the kill, instead of by chopping of the stiff joints at a later point in time. Cut marks occur on 5% of the bone fragments. Complete long bones are very rare, consisting only of some cannon bones. This points to the breakage of

46

M. Germonpre´ and L. Lbova

Table 11. Dimensions of tarsal bones of Procapra gutturosa (Mongolian gazelle) from Kamenka Complex A 1993 Tarsal bones Astragalus

Calcaneum Centrotarsale

GL1 GLm D1 Dm Bd GL GB

N

Lim



ó

10 10 10 1 10 2 9

27·5–30·3 25·5–28·2 15·3–17·3 15·7 15·7–18·4 58·5, 59·4 20·3–23·9

29·1 27·2 16·1

0·8 0·8 0·6

17·1 59·0 22·2

0·7 0·9

Table 12. Dimensions of phalanges of Procapra gutturosa (Mongolian gazelle) from Kamenka Complex B 1993

Complex A 1993 Phalanges

N

Lim



ó

N=1

Phalanx 1 Forelimb GL Bp Hindlimb GL Bp Phalanx 2 GL Bp Phalanx 3 DLS Ld MBS

17 16 4 4 17 18 8 2 1

35·3–42·2 10·0–11·6 32·6–35·2 9·6–11·3 19·8–23·2 9·0–10·0 21·5–22·9 19·2–20·3 5·4

38·0 10·6 33·7 10·3 21·9 9·5 22·1 19·8

1·8 0·5 1·1 0·6 0·8 0·3 0·5

37·6 10·6

the bones for marrow and grease extraction which is also indicated by the large quantity of unidentified shaft pieces. The long bones of the front limb are better represented than those of the meatier rear limb. For the front limb, the upper limb bones are somewhat more frequent than the lower ones. However, proximal epiphyses are under-represented. This is often considered to be an indication of an important amount of carnivore modification in a bone assemblage (Todd & Rapson, 1988; Binford, 1981), but humans will also seek out the most useful parts of the skeleton (Kreutzer, 1992). The high grease containing articular ends can be preferentially destroyed in an archaeological context as well. Moreover, carnivore gnawing traces are lacking on the gazelle remains. At Kamenka, many articular ends were transformed in cancellous blocks of which four (a distal epiphysis of a humerus, a pelvis fragment, a distal epiphysis of a femur and a skull fragment) were identified as Mongolian gazelle (Figure 9). The remains of the Mongolian gazelle from the sample of the Complex A 1992 excavation include a fragment of a maxillary and a skull fragment with the base of the right and the complete left horncore, both fragments are fresh. Complex C 1992 contains one scapula fragment from this bovid. A first phalanx was collected at the locus of Complex B 1993. The Mongolian gazelle lives today in the dry grassteppes and semi-deserts of the Eastern Altai,

Table 13. NISP and MNI of skeletal elements for Equus caballus (horse) from Kamenka Kamenka Complex A 1993 Skeletal elements Teeth M3 max. P2 mand. other Skull Mandibula Hyoid Rib Vertebra Sternum Scapula Humerus distal ulna Radius proximus distal MC proximal distal Pelvis Femur proximal diaf Patella Tibia proximal distal Carp./tars. calcaneum astragalus other MT split proximal diaphysis distal MP MP acc Sesam Phalanx phalanx I phalanx II phalanx III Total

NISP

MNI

82

Kamenka Complex A 1992 NISP

MNI

2 1 2 79

2 12

2 2 6

2 2

1

1 1

1 1

1 3

1 3

7

1

1

1

2 2

2 1 1

4

2

2

3 1 1 4

2 2

1 3

3

3

2

1

1 3 1 3

1 2 1 2

2

9

1 2 1 6

6

2 1 4

1 4

2 2 1 1

4 4

4 5 3 8

2 2 3 3

150

1

1 2 1 6

1 1 28

1 1 4

Mongolia and Inner Mongolia. It prefers plains and soft undulating landscapes and avoids steep slopes; furthermore it shuns regions with heavy snowfall. Its shoulderheight varies between 54 and 84 cm and its body weight between 20 and 39 kg. The meat of the Mongolian gazelle is said to be good and its winter pelt is used to make fur coats. The rut lasts from the end of November until the beginning of January; the young are born between the middle of June and the beginning of July (Heptner, Nasimovic & Bannikov, 1966). Mongolian gazelle occurs in several Late Pleistocene sites of the Transbaikal area (Imetchenov & Kalimikov, 1988). At the early Upper Palaeolithic site of Varvarina Gora, situated some 17 km southeast of Kamenka along the Brianka river, Mongolian gazelle

Mammalian remains from the Upper Palaeolithic site of Kamenka, Buryatia (Siberia) 47 Table 14. Dimensions of teeth of Equus caballus (horse) from Kamenka Complex A 1993 Teeth Maxilla P2–M3 P2–P4 M1–M3 P2 P3 P4 M1

M2

M3

Mandibula P2–M3 P2–P4 M1–M3 P2

P3

P4 M1 M2 M3

cl cl cl cl cw LP cl cw LP cl cw LP cl cw ch LP cl cw ch LP cl cw ch LP cl cl cl cl cw ch LF cl cw ch LF cl cw LF cl cw LF cl cw LF cl cw ch

N

Lim

1

74·8

1 1 1 1 1 1 1 1 1 1 1

28·2 23·5 15·1 24·0 24·6 13·3 23·1 26·4 13·2 25·3 26·7

1 1 1 1 1

13·5 26·1 27·6

1 1 1 2 2 1

165·9 84·5 78·2 31·3, 32·9 14·0, 14·9 47·1

4 4 2 4 2 2 2 1 1 1 1 1 1 1 1

24·4–31·8 12·8–17·0 12·2, 74·9 6·9–14·0 26·5, 30·1 11·5, 16·1 12·4, 12·5 24·7 15·0 8·5 25·5 14·7 10·0 31·3 15·0



Complex A 1992 ó

11·8

32·1 14·5 28·0 15·2 43·6 11·9 28·3 13·8 12·5

is the most frequent species present both in NISP (29%) and MNI (28%). Almost all skeletal elements were found. At Tolbaga however, situated some 90 km southeast of Kamenka and also dating from the early Upper Palaeolithic, Mongolian gazelle is rare (Ovodov, 1987). At the Upper Palaeolithic site of Sukhotino 4 (level 2), in the vicinity of Chita, eastern Transbaikal, Mongolian gazelle is also not well represented. Level 2 was dated at about 26,000 years  (Kasparov, 1986). Equus caballus. The systematics and nomenclature of the Pleistocene caballine horses of Eurasia is a very complicated matter (Sher, 1974; Gautier &

2·9 1·5 2·9

N

Lim



ó

1

170·4

1 1 1

77·2 37·0 25·1

1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

30·1 26·3 11·3 29·0 25·9 10·7 24·6 24·9 59·7 11·0 24·7 24·2 67·8 11·4 27·8 22·8 66·3 12·9

2 1

164·4, 174·2 101·5

169·3

3 3

32·7–37·5 12·8–15·1

34·3 13·7

2·2 1·0

2 3 3

15·1, 16·3 28·2–30·7 11·7–16·3

15·7 29·5 14·3

1·0 1·9

3 3 3 3 2 2 2 2 2 2 2 2 1

13·6–14·7 27·3–34·5 12·2–15·3 11·8–14·5 24·9, 25·6 13·3, 15·4 9·6, 10·5 26·0, 26·8 13·3, 14·5 9·5, 10·6 28·1, 29·2 10·9, 12·6 68·5

14·0 30·0 14·2 13·4 25·3 14·4 10·1 26·4 13·9 10·1 28·7 11·8

0·5 3·2 1·4 1·1

Kobusiewicz, 1992). For the moment, the caballine horse of Kamenka will be described as Equus caballus. The height at the withers could be calculated with the indices of Kiesewalter (1888) on nine complete cannon bones, these calculations resulted in a mean shoulderheight of 135 cm (range is 128 cm–142 cm). The horses of Kamenka were not large and correspond in size to the Equus caballus subspecies (small form) of the Late Pleistocene from Northern Siberia (Yedoma suite) as described by Sher (1974). Remains of horse account for 36·9% of the NISP and 33·3% of the estimated MNI of the Kamenka Complex A 1993 assemblage (Table 1). The frequency of the skeletal elements is given in Table 13. The

48

M. Germonpre´ and L. Lbova Table 15. Dimensions of isolated teeth of Equus caballus (horse) from Kamenka Complex A 1993 Isolated Teeth Maxilla P/M P3/P4

M1/M2

M3

Mandibula P2

P3/P4

M1/M2

M3

Complex C 1992

N

Lim



cl cw ch cl cw ch LP IP cl cw ch LP IP cl cw ch

2 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1

29·4, 32·0 25·6, 25·9 86·7, 97·3 29·9 26·0 84·2 11·4 38·1 27·1 24·6 65·4 13·6 50·2 34·5 23·6 72·3

30·7 25·8 92·0

cl cw ch LF IF cl cw ch LF IF cl cw ch LF IF cl cw ch

2 2 1 1 1 1 1 1 1 1 2 2 1 2 2

30·6, 36.2 13·1, 16·2 54·4 16·4 45·3 28·8 15·9 68·7 12·6 43·8 27·1, 28·0 14·9, 15·5 67·2 10·3, 10·6 37·9–38·0

33·4 14·7

N=1

Complex A 1992 N

2 2 2 2 2

27·6 15·2 10·5 38·0

distribution of the elements is more uneven than for Procapra. Not all elements occur, ribs, vertebrae, sternum and ulna are lacking. It is possible that these body parts were not brought to the site. This kind of frequency distribution is common in faunal assemblages that have been transported from a kill and initial butchering site to a residential or consumption site (Enloe, David & Hare, 1994). The absence of the axial skeleton could also be explained by differential destruction, although this seems unlikely since the preservation of the bone material is excellent. Another possibility is that the axial elements were destroyed beyond recognition by the human occupants of the site when they fragmented the bones for grease processing. The best represented element by number are the teeth, which represent more than 50% of the equid remains, followed by the mandible which provides the highest MNI estimate (6). The horse remains are much more broken than those of gazelle, only 21% is complete, consisting predominantly of phalanges (33%), carpals/tarsals (25%) and teeth (25%). This is probably due to durability and compactness of such elements as teeth, carpals/tarsals and phalanges and

Lim

25·4, 25·6, 42·7, 12·7, 50·0,

27·4 27.8 54·4 15·0 54·7



26·4 26·7 48·6 13·9 52·4

31·5 17·4 78·0 15·1 47·9 28·2 14·9 72·4 9·6 34·0 29·2 13·3 66·1

the fact that people did not exploit these elements at Kamenka. Complete long bones are absent. The frontleg bones are slightly better preserved than those of the rear leg. Proximal articular ends of the humerus are not present. The scarcity of limb bones is probably caused by the smashing up of these bones for marrow and grease extraction, resulting in a large quantity of flakes and unidentifiable articular end blocks. Three cancellous blocks were identified as derived from horse, two distal articular ends of a humerus and one distal articular end of a metapodial. Some 5% of the remains exhibit carnivore chewing. Ten isolated tooth fragments are from deciduous teeth or tooth buds, and one immature postcranial element (a femur fragment) has been found. No associations were discovered. Two metacarpal fragments, one in square H3, the other in square H4, are from the same bone. Some trends can be noted in Complex A 1992 that are undoubtedly different from the Complex A 1993 assemblage. Two associations occur. Seven articulated cervical vertebrae were found in squares B4–B5–C5, they are all unweathered. A skull fragment consisting

Mammalian remains from the Upper Palaeolithic site of Kamenka, Buryatia (Siberia) 49 Table 16. Dimensions of skeletal elements of the forelimb of Equus caballus (horse) from Kamenka Complex A 1993 Skeletal elements Scapula

LG BG BT Bd BFd GL Ll Bp Dp KD DD Bd Dd

Humerus Radius MC

N

Lim

1 1 2 1 1 2 2 3 3 1 1 2 1

62·4 50·2 76·5, 76·7 74·6 65·2 222, 230 215, 222 44·2–50·5 27·8–34·4 36·7 27·7 44·1, 48·4 35·8

Complex C 1992 x¯

Complex A 1992

N=1

N

Lim



61·0 50·4

1

48·5

2 2 2 2 1 2 1 1

205, 215 200, 210 48·7, 46·3 34·0, 30·1 32·6 25·0, 24·5 44·2 34·5

76·6 226·0 218·5 48·3 31·7 46·3

210·0 205·0 47·5 32·1 24·8

Table 17. Dimensions of skeletal elements of the hindlimb of Equus caballus (horse) from Kamenka Complex A 1993 Skeletal elements Pelvis Tibia MT

LA Bd Dd GL Ll Bp Dp KD DD Bd Dd

N

Lim

1 1 1 2

73·7 44·4 270 262, 250

1

41·4

2 1

45·3, 39·6 39·3

Complex C 1992 x¯

Complex A 1992

N

Lim



2 2

70·4, 78·5 46·9, 51·1

74·5 49·0

256

42·5

N

Lim



ó

3

62·4–65·7

63·9

1·4

4 3 3 3 3 3 3 3

245–265 245–258 46·8–51·3 41·5–46·0 31·0–32·7 30·1–30·5 44·1–47·7 37·0–37·8

255 252 49·0 43·5 31.7 30·3 46·4 37·3

7·9 5·4 1·8 1·9 0·7 0·2 1·7 0·3

Table 18. Dimensions of tarsal bones of Equus caballus (horse) from Kamenka Complex A 1993 Tarsal bones Astragalus Calcaneum

BFd LmT GL GB

N

Lim

1

52·6

2 2

107·8, 108·9 44·4, 46·1

Complex C 1992 x¯

N

108·4 45·3

2 2 2 2

mainly of the maxillary is also present in square C5; as the occipital condyle are missing, it cannot be checked whether the skull fits with the cervical column. A metatarsal and a calcaneum of the same individual were retrieved from square A4. Table 13 gives the complete list of the skeletal elements of the Complex A 1992 assemblage. Complex C 1992 contains several remains of horse: a lower jaw fragment, three fragments of jugal teeth, a shoulderblade fragment, two distal tibia fragments, two astragali, two calcanea and two first phalanxes. No horse bones were recovered from Complex B 1993. Horses prefer windswept landscapes where grasses are exposed (Guthrie & Stoker, 1990). Stomach con-

Lim 50·3, 58·5, 105·1, 45·7,

57·9 65·5 106·9 51·8

Complex A 1992 x¯

N=1

54·1 62·0 106·0 48·8

107·6 44·8

tent of the frozen horse remains from the permafrost of Siberia reveal that these animals subsisted on a diet consisting mainly of grasses, and to a lesser extent, on twigs from willow and birch. Fossils of horses are abundant all over Eurasia (Vereshchagin & Baryshnikov, 1982) and also in the Transbaikal (Imetchenov & Kalmikov, 1988). At Varvarina Gora, horse is the second best represented species both in NISP (21%) and MNI (13%), while at Tolbaga it is the second best represented species in NISP (16%) and the third in MNI (12·5%). Remains of the long bones are most frequently found (Ovodov, 1987). In level 2 of Sukhotino 4 only a few horse bones were retrieved (Kasparov, 1986).

50

M. Germonpre´ and L. Lbova

Table 19. Dimensions of phalanges of Equus caballus (horse) from Kamenka Complex A 1993 Phalanges Forelimb Phalanx 1

Phalanx 2

Hindlimb Phalanx 1

Phalanx 2

Phalanx 3

N

Lim

GL Bp Dp KD Bd GL Bp Bd

1 1 1 1 1 2 2 2

90·4 50·4 36·2 35·2 45·3 40·7, 43·5 44·1, 53·2 44·4, 48·7

GL Bp Dp KD Bd GL Bp Dp KD Bd GL GB BF

1 1 1 1 1 1 1 1 1 1 3 2 3

78·9 52·2 38·0 35·3 45·0 49·1 49·4 31·5 42·5 47·2 38·2, 44·6 49·2, 56·8 42·3, 47·8



Complex C 1992 ó

Complex A 1992

N

Lim



2 2 2 2 1

78·5, 81·9 48·3, 49·2 30·6, 32·5 31·7, 32·0 41·2

80·2 48·8 31·6 31·9

N=1

42·1 48·7 46·6 84·4 54·3 38·6 32·5 43·4 45·8 50·0 29·8 41·4 45·7 42·1 53·0 45·0

Mammuthus primigenius. The species does not occur in Complex A 1993, but was found in Complex B 1993, the find consists of a proximal femur fragment (length=25 cm, weathering stage 3) that was gnawed by carnivores. Woolly mammoth occurs in many sites of the Transbaikal (Imetchenov & Kalmikov, 1988), although it is lacking in the Upper Palaeolithic sites of Tolbaga, Varvarina Gora and Sukhotino-4 (level 2) (Kasparov, 1986; Ovodov, 1987). The woolly mammoth was adapted to the open landscapes of large river valleys and plains of northern Eurasia and North America. Its diet was based on grasses and sedges enriched with leaves and twigs of woody plants (Vereshchagin & Baryshnikov, 1982; Guthrie, 1990). Equus hemionus. Kulan was not found in Complex A 1993, but a posterior cannon bone (measurements: Bp=38·3, Dp=36·4, KD=22·7, DD=26·2) from this equid occur in Complex A 1992 (square D5). Some tooth scratches are present on the shaft and the proximal articular end. The distal articular end is missing and slight carnivore gnawing traces occur also all around the distal end surface; probably the epiphysis was not yet fused. The bone surface looks as if it came from a subadult or rather young adult. Several sites in the Transbaikal contain remains from the kulan (Imetchenov & Kalmikov, 1988). This species prefers hilly steppes and half deserts where it lives on grasses, sedges and mugwort. It can cover great distances during the summer and winter movements (Heptner et al., 1966). At Tolbaga, Varvarina

2·8 2·0

Gora and Sukhotino-4 (level 2), only a few specimens are present (Kasparov, 1986; Ovodov, 1987). Coelodonta antiquitatis. Five fragments of the woolly rhinoceros were identified in Complex A 1993: an ulna fragment (length=14 cm), the occipital parts of a skull (length=17 cm, measurement: GB Con. occ.=148·7 mm), the distal half of a humerus (length=15 cm), a thoracal vertebra fragment (length=14 cm) and a rib fragment (length=12 cm). Compared with the bulk of the assemblage, these remains are very different. Their colour is a pale grey, they are much bigger, they all belong to late weathering stages (2 and 3), and four of them show carnivore gnawing traces. This suggests a taphonomic history different from the remains already described. These bones were not buried as quickly and were more intensely gnawed by carnivores. Also, no bone smashing by prehistoric humans could be detected. These bones are possibly the remains of animals that died on the site or that were brought there by carnivores. As such, they constitute penecontemporaneous intrusives (sensu Gautier, 1987). Two complete metacarpals from the woolly rhinoceros, belonging to the same individual, were found in Complex A 1992 (squares E6 and D6). They are not broken, not weathered and without palaeoichnological marks (measurements: MCIII: GL=188, Bp=69·7, Dp=54·3, KD=51·1, Bd=62·1, Dd=51·4; MCIV: GL=150·3, Bp=50·5, Dp=47·6, KD=37·4, Bd=46·0, Dd=40·5). One fragment of an upper jugal tooth (weathering stage 1) of Coelodonta antiquitatis was

Mammalian remains from the Upper Palaeolithic site of Kamenka, Buryatia (Siberia) 51

retrieved from Complex C 1992. Complex B 1993 contains one femur fragment (weathering stage 3) and one upper jugal tooth fragment (weathering stage 3) of this rhinocerotid. The woolly rhinoceros was an inhabitant of the steppe-tundras of northern Eurasia. It is only a minor component of the faunal assemblages of the Palaeolithic sites of the Russian lowland (Vereshchagin & Baryshnikov, 1982). Fossils of the woolly rhinoceros are commonly found in the late Pleistocene deposits of the Transbaikal area (Imetchenov & Kalmikov, 1988; Kalmikov et al., 1992). The high number of rhinoceros remains collected at the Upper Palaeolithic site of Tolbaga is very exceptional (NISP: 54%, MNI: 17%). All skeletal elements are well represented (Ovodov, 1987). Some bones are from unborn and juvenile individuals (own unpublished data). At Varvarina Gora rhinoceros is less important (Ovodov, 1987). The assemblage of Sukhotino-4 (level 2) contains only a few rhinoceros bones (Kasparov, 1986). Camelus sp. A proximal half of a first phalanx was identified as from camel (measurements: Bp=48·5, Dp=33·8). It was found in Complex A 1993 (square D2), has a pale grey colour, is in a rather high weathering stage (stage 2) and shows a columnar fracture. These characteristics are comparable with those from the rhinoceros remains already described. Two other phalanges were retrieved from Complex A 1993 (square E2), a complete first phalanx (weathering stage 1, measurements: GL=93, Bp=44·6, Dp=34·1, KD=26·4, Bd=40·0) and a broken second phalanx (length=7 cm, weathering stage 1). They probably belong to the same individual. Historically, wild camels ranged in the deserts and steppes of Kazakstan, Mongolia and China. However, their numbers have steadily declined during this century (Heptner et al., 1966). Only two Pleistocene sites in the Transbaikal yielded material from Camelus knoblochi (Imetchenov & Kalmikov, 1988). Megaloceros giganteus. Only two remains from Complex A 1993 belong to the deer group: a shed antler fragment from an adult male (length 23 cm, measurements: Dt burr=87·4 mm) and the distal half of a humerus (length=24 cm; measurements: KD=50·6 mm, Bd=92·9 mm, Dd=93·2 mm). The humerus (weathering stage 1) displays a spiral fracture associated with carnivore gnawing traces. The antler (weathering stage 2) was also chewed by carnivores. The foregoing features indicate again that these finds can be placed in the same taphonomic category as those of woolly rhinoceros and camel. Megaloceros giganteus lived in the forest-steppes, steppes and steppe-tundras of Eurasia (Gromova & Baranovoi, 1981). Only two sites in the Transbaikal area yielded remains of the giant deer (Imetchenov & Kalmikov, 1988). The Transbaikal area is at the eastern limit of the distribution of the species. In

Siberia, the latest finds date from the beginning of the Upper Palaeolithic (Vereshchagin & Baryshnikov, 1984). Bison priscus/Bos primigenius/Poëphagus baikalensis. Five fragments of Complex A 1993 pertain to a large bovid of the size of Bison priscus, Bos primigenius or Poëphagus baicalensis. They are: a complete maleolar bone (weathering stage 2), a skull fragment (length= 13 cm, measurements: GB Con. occ.=120·4 mm), a mandible fragment (length=16 cm), a proximal fragment of a radius (length=8 cm) and a fragment of a second phalanx (length=5 cm, weathering stage 2). This material is too fragmented to allow a specific attribution. However, the yak and aurochs are rare in the Transbaikal region and the skull fragment points to the presence of the steppe bison Bison priscus. The radius (weathering stage 1) shows a spiral fracture. The skull fragment and lower jaw fragment belong to weathering stage 3 and were gnawed by carnivores. Thus, these specimens appear to be penecontemporaneous intrusives. The collection of Complex A 1992 is somewhat richer in large bovid bones. Eight fragments could be identified: a right lower jaw (measurements: cl M1M3=107·8, M1 cl=26·1, M1 cw=18·2, M1 ch=19·1; M2 cl=32·5, M2 cw=19·9; M3 cw=19·3); three cervical vertebrae of which two were collected in articulated position from square B8; a scapula, which could be from a steppe bison (measurements: GLP=80·3, LG=70·0, KLC=69·7); a heavily gnawed pelvis fragment (measurement: LA=82·9); a proximal half of a tiba displaying gnawing traces, and a complete calcaneum (measurements: GL=154·9, GB=50·2). Complex C 1992 contains one astragalus from Bison priscus (weathering stage 2, measurements: GLl=79·2, GLm=72·1, Bd=52·2). A large bovid is present with three fragments in the sample of the Complex B 1993 excavation: an axis fragment (weathering stage 3), a tibia fragment (weathering stage 2) and a complete radius (weathering stage 2; measurements: GL=365, Bp=110·8, Dp=58·6, KD=58·0, Bd=95·6). The latter could be ascribed to Bison priscus. It is clear that at Kamenka the large bovid, which is probably the steppe bison, belongs at least partly to the penecontemporaneous intrusives and was not an important game animal, in contrast with the bison oriented subsistence strategy at Middle Palaeolithic sites such as Mauran, France (Farizy, David & Jaubert, 1994), Wallertheim, Germany (Gaudzinski, 1995) and Il’Skaya I, Russia (Hoffecker, Baryshnikov & Potapora, 1991). Bison priscus inhabited the steppes and steppetundras of northern Eurasia, where it fed mainly on the steppe grasses. As with the bison of today, it probably did not tolerate thick snow cover (Vereshchagin & Baryshnikov, 1984). Many Pleistocene sites in the Transbaikal area contain remains of this large bovid

52

M. Germonpre´ and L. Lbova

(Imetchenov & Kalmikov, 1988). At the early Upper Palaeolithic sites of Tolbaga and Varvarina Gora, only a few remains of large bovids are present (Ovodov, 1987). At Sukhotino-4 (level 2), however, steppe bison is the most important game animal together with argali sheep and reindeer (Kasparov, 1986). Spirocerus kiakhtensis. Of this ruminant, only a horncore fragment (length=12 cm, weathering stage 2, measurements: Dob hcore=32·3 mm, Dt hcore= 25·0 mm) was recovered in Complex A 1993. The spiral winding is very clear. The horncore was detached from the skull and at the base two parallel cut marks occur (Figure 7). Although the prehistoric hunter– gatherers of Kamenka undoubtedly bagged Spirocerus kiakhtensis, this herbivore appears not to be an important game animal. The species was first described by Pavlova (1910) on the basis of cranial remains; Sokolov (1959) gives a description of some postcranial material. This extinct antelope has been found in Pleistocene deposits of the Altai, the Baikal region, Mongolia and Northern China (Gromova & Baranovoi, 1981). At the Upper Palaeolithic sites of Tolbaga, Varvarina Gora and Sukhotino-4 (level 2) it occurs only in low quantities (Kasparov, 1986; Ovodov, 1987) as is the case at Kamenka. Ovis ammon. Four bone fragments from Complex A 1993 were identified as coming from argali sheep: an occipital condylus (length=5 cm, weathering stage 0), a proximal articular end of a radius (length=4 cm, weathering stage 2, measurements: Dp=28·4 mm) showing a spiral fracture, a gnawed metacarpal shaft fragment (length=8 cm, weathering stage 1) and a proximal epiphysis fragment of a tibia (length=11 cm, weathering stage 2, measurement: Bp=93·6). Three bones were found in Complex A 1992: a complete radius (measurements: GL=299, Bp=60·8, Dp=31·6, KD=32·2, Bd=49·5, Dd=37·8) and two skull fragments (measurements: GB con. occ.=77·1, 78·9, GB for. magn.=24·6, 28·9, H for. magn.=23·7, 22·7). On the skull fragments the bases of the horncores are partly preserved. One fragment is slightly abraded, perhaps due to trampling or to carnivore activity but no unequivocal chewing traces can be seen. Ovis ammon displays a remarkable geographic variability in size and horn morphology. The largest subspecies today is Ovis ammon ammon, the Altai sheep. It can have a shoulderheight of 125 cm and its body weight can reach 200 kg. During recent times, the Altai sheep lived in the Altai, northern Mongolia and in the Transbaikal area, from where it disappeared in the 18th century. It prefers the steppes of plateaus and foothills and avoids steep slopes. Its meat is at its best during fall, before the rut, which lasts about one month between the middle of October until the middle of January. During rut, the meat of rams is said to

be unedible (Heptner et al., 1966). During the Late Pleistocene Ovis ammon was widespread in the Transbaikal area (Imetchenov & Kalmikov, 1988). At Tolbaga and Varvarina Gora it is rather well represented (Ovodov, 1987). At Sukhotino-4 (level 2), it dominates the fauna (NISP: 42%, MNI: 21%) (Kasparov, 1986). Bovidae. Eleven bone fragments from Complex A 1993 could not be identified beyond the family level. This material comprises remains of large and smaller bovids as well as remains that cannot be assigned to a distinct size category; four fragments belong to immature individuals. The sample of Complex A 1992 contains three fragments of bovid bones that also could not be attributed to a species. Panthera leo. The Kamenka Complex A 1993 assemblage contains one ulna fragment that pertains to the lion. The fragment has a length of 15 cm (measurement: BPc=48·9), it belongs to weathering stage 2, has a pale grey colour and was gnawed by carnivores. These characteristics permit its classification as a penecontemporaneous intrusive. Lion is a typical component of the mammoth fauna. This felid is, however, not frequently found in the late Pleistocene sites of the Transbaikal (Imetchenov & Kalmikov, 1988). Spatial distribution The Kamenka Complex A 1993 excavation covers 33 m2. The concentration of bones varies by square from 0 (square F3) to 369 (square D2) specimens (Figure 11). The distribution of certain features of the faunal assemblage was investigated to see whether any meaningful pattern could be detected. Some of the parameters used do not show marked changes over the excavated surface, such as the frequency of the Mongolian gazelle and horse remains and the frequency of weathering stages 0 and 1, while others show high values for distinct locations. The spatial distribution of the remains of woolly rhinoceros, camel, giant deer, bison, argali sheep and lion is shown in Figure 12. Bones of lion, woolly rhinoceros and giant deer occur only in the most eastern extension of the excavation, where the structure of large stone blocks is situated. It is thought that these remains, together possibly with the camel, argali sheep and bison material, which are partly located elsewhere, form basically an assemblage of penecontemporaneous intrusives. The main reasons for this assumption have already been given: the weathering of these remains is generally much more advanced, the fragments are bigger and carnivore gnawing marks are decidedly more frequent. The relative frequency per square metre of bones with gnawing marks was calculated for those squares that contain a minimum of 40 specimens (Figure 13).

Mammalian remains from the Upper Palaeolithic site of Kamenka, Buryatia (Siberia) 53 A

B

C

D

E

F

G

A

H

B

C

D

E

F

G

H

1

1

2

2

3

3

4

4

5

5 6

6

Coelodonta antiquitatis < 10 specimens 10–39

Megaloceros giganteus

40–69 70–99 >100 Figure 11. Spatial distribution of the faunal remains of Kamenka Complex A 1993.

Ovis ammon Camelus Bison priscus Panthera leo

Another characteristic of the bone collection shows concentrated occurrences in relation to the two hearths. These were excavated at the border of squares D1-2/E1-2 and at the border of squares C2-3/D2-3. The distribution of the burned bone fragments partially coincides with the location of the fire places (Figure 14). In the latter zone most of the bone tools and ornamented bones were found. Other characteristics show no clustering. Cancellous blocks and bone splinters occur over much of the area. Cut marks and impact points are not frequent enough to interpret their spatial distribution. As to root marks, they can be found everywhere, but their highest score is in the northwestern corner; the significance of this phenomenon remains unclear.

Discussion and Conclusion Palaeoecological setting Too few mammalian remains were collected from Complex B to arrive at a meaningful conclusion for this archaeological horizon, but the faunas of Complexes A and C suggest a dry steppe environment. The two most common mammals found are the Mongolian gazelle and horse. Other inhabitants of the steppe zone include steppe bison, kulan, rhinoceros and camel. Also, the presence of Brant’s vole (Lasiopodomys brandti) in the microfauna indicates a semi-arid steppe landscape (Khenzykhenova, pers. comm.). Real coldadapted species such as reindeer (Rangifer tarandus) are absent. The site is located at the edge of the large floodplain of the Brianka River at the foot of hilly

Figure 12. Spatial distribution of the remains of woolly rhinoceros, camel, giant deer, steppe bison and argali sheep (Kamenka Complex A 1993).

outcrops of igneous rocks. The species present avoid thick snow cover and thrive essentially on grasses. It is likely that a watering place was located nearby. During the Pleistocene a tributary of the Brianka flowed in the vicinity. This could explain the penecontemporaneous intrusives. Mammals such as rhinoceros, camels, cervids and bovids, some of them moribund, came to the area to drink and were killed or scavenged by carnivores. Carnivores were apparently also attracted by the refuse in the human camp (see gnawing traces on horse bones of Complex A 1993 and the relative high frequency of gnawing traces on bones of Complex A 1992). Chronological position Available unpublished geological and palaeopedological data suggest that the archaeological horizons were deposited during the Karginsk interstadial, which could explain the lack or under-representation of animals adapted to the extreme cold. The presence of giant deer points to an early Upper Palaeolithic age. Radiocarbon dates vary between 35,845&695 years  and 31,060&530 years  for Complex A, and between 28,815&150 years  and 28,060&475 years  for Complex B. They corroborate the attribution to the Karginsk interstadial.

54

M. Germonpre´ and L. Lbova A

B

C

D

E

F

G

H 1 2 3 4 5 6

1–5 % gnawing traces 6–10 % > 10 % Figure 13. Spatial distribution of the bone fragments with gnawing traces (Kamenka Complex A 1993).

A

B

C

D

E

F

G

H 1 2 3 4 5 6

1–5 % burned bone fragments

Only two skeletal elements (the recently fused distal humerus and the unfused first phalanges) indicate a specific period when the animals died, between midAugust and December. From this it can be assumed that Kamenka was occupied during the late summer/ autumn and/or early winter. The unfused epiphyses of the long bones and the absence of recently fused long bones do not contradict this interpretation. Late winter and spring are seasons of food stress. By spring many ungulates may become severely fat depleted, losing even much of the marrow fat (Speth, 1987). The fat reserves fluctuate not only because of climatic and nutritional variables, but also change in relation to reproductive strategies (Bunn & Ezzo, 1993). In temperate and colder environments primeaged adults store fat in the limb bones in advance of the cold season, and, in case of the males, in preparation for the rut (Stiner, 1991). According to Todd (1991), during the Late Pleistocene the period of most severe fat depletion in both sexes of bison was late winter to early spring. Presumably the highest fat content in the bones of Pleistocene Mongolian gazelle and horse was reached also during the late warm season and/or early winter. This coincides with the period of occupation of Complex A 1993. In this period, the human hunters could prepare and store grease for later consumption with relative ease. It would, moreover, seem that phalanges were practically not used for grease preparation at Kamenka. According to Speth (1987: 20) ‘‘marrow fat in the lower limbs (and possibly also in the mandible) is the last deposit to be depleted, and in a severely stressed animal may be the only remaining portion of the carcass retaining sufficient fat to be worth eating’’.

The low frequency of split phalanges at Kamenka (only 12% of the phalanges of Mongolian gazelle are broken and none of horse) may indicate that the animals were in a good condition when they were killed, making it unnecessary to exploit terminal leg elements.

6–10 % > 10 % Figure 14. Spatial distribution of the burned bone fragments (Kamenka Complex A 1993).

Seasonality Only a few indications of seasonality are available. Table 20 is based on the state of fusion of the postcranial bones of Gazella gazella, as established by Davis (1980), since comparable data on Mongolian gazelle could not be found. According to Davis (pers. comm.), the same pattern of fusion can be expected for different species of gazelle of comparable size and body weight.

Site function Several sets of cultural components were unearthed at Kamenka. Complexes A and C belong to the same stratigraphical level; Complex B is located higher in the stratigraphic sequence. The achaeological horizons are thin, but they split up locally in doublets or triplets. The archaeological features, artefacts and bones at the different loci represent the remains of repeated episodes of human occupation during which subsistencerelated activities were carried out. The presumed watering place with its high concentration of game animals probably attracted the Palaeolithic hunters, at least seasonally, and the topography and vegetation of the surrounding hills may have offered possibilities for

Mammalian remains from the Upper Palaeolithic site of Kamenka, Buryatia (Siberia) 55 Table 20. Season of death of Procapra gutturosa, based on the state of fusion of postcranial bones of Gazella gazella (Davis, 1980) (r.f.: recently fused) Procapra gutturosa* Kamenka Complex A 93 Skeletal element Distal humerus (r.f.) Phalanx I (unfused) Distal tibia (unfused) Proximal femur (unfused) Distal femur (unfused) Calcaneum (unfused) Distal MP (unfused) Proximal ulna (unfused) Proximal tibia (unfused)

NISP 1 5 2 4 1 1 3 2 1

Period of death Mid August–Beginning of September Before December Before March Before May Before May Before May Before May Before July Before July

Gazella gazella Age of fusion (months) (Davis, 1980) 22 25–8 28–10 210–16 210–18 210–16 210–16 212–18 212–18

*Season of birth: mid June–beginning of July

stalking and ambushing prey. Mongolian gazelle and horse were the primary game animals, but Spirocerus kiakhtensis was also sporadically bagged. The frequency distribution of the skeletal elements of gazelle suggests that the complete and unprocessed animals were brought to the camp. They were probably butchered soon after they died, since the carcasses were basically dismembered at the joints by cutting. The frequency distribution of the remains of horse is different. The evidence from Complex A 1993 suggests that the horses were partially butchered elsewhere. The missing elements of the axial skeleton might have been abandoned or consumed at the kill; although it is also possible that these body parts were heavily fragmented at the camp for grease processing and therefore could not be identified. At the locus of Complex A 1992, the complete articulated cervical part of the axial column was recovered. This finding could indicate that the animal was killed in the immediate vicinity and dismembered at the site. However, according to O’Connell, Hawkes & Burton Jones (1990), Hadza hunters often transport cervical vertebrae of zebra from kill site to base camp. The archaeological features (hearths, stone structure, pit, etc.), the worked bone remains, the large quantity of shaft pieces and cancellous blocks indicate that people spent some time at the site to work and process the animals and bones. The extreme fragmentation of shafts and articular ends of long bones suggests that all food resources (meat, marrow, grease, bone juice) were exploited. Subsistence strategy The fauna in the archaeological record of the Transbaikal region seems to indicate that the subsistence of the local early Upper Palaeolithic hunter– gatherers was based on a rather broad range of prey species. Although the sites of Kamenka (Complex A 1993), Tolbaga, Varvarina Gora and Podzvonkaja are roughly contemporaneous (their tool complexes are

ascribed to the Tolbaga industry), they differ in species composition. At the moment, it is impossible to know if this is related to geographical, temporal, environmental, demographical, functional or seasonal differences between the sites. At Kamenka Complex A 1993, Mongolian gazelle and horse are the main game animals which were probably hunted during the late summer/autumn and/or early winter. The other complexes at Kamenka are too restricted to allow inferences. At Tolbaga, the three main species are argali sheep, horse and, surprisingly, rhinoceros (Ovodov, 1987). At Varvarina Gora, Mongolian gazelle, horse and argali sheep are well represented (Ovodov, 1987). The scanty remains at Podzvonkaja belong to argali sheep and horse (Cauwe et al., 1993). This site is located some 200 km south of Kamenka, near the Mongolian border. The fauna of Sukhotino-4 (level 2) is definitely younger than the foregoing assemblages and is dominated by argali sheep, reindeer and steppe bison (Kasparov, 1986). The analysis of additional early Upper Palaeolithic sites in the Transbaikal is necessary to answer the question whether subsistence strategies in that region were generally based on broad ranges of prey species, but the sites investigated until now do not indicate specialized hunting of single species.

Acknowledgements The first author is grateful to the Buryat Institute of Social Sciences of the Siberian Division of the Russian Academy of Sciences for inviting her to study the mammalian remains of Kamenka, to the Royal Museums of Art and History of Brussels for inviting her to join their archaeological expedition in Buryatia in 1993, and to the Belgian Federal Office for Scientific, Technical and Cultural Affairs (contract SC-004), which supported the work in Ulan Ude and Chita in 1994.

56

M. Germonpre´ and L. Lbova

V. Kolosov and S. G. Vasilyev (Centre for the conservation of the historical-cultural heritage, Chita) are thanked for access to the faunal material of Tolbaga in their care. The discussions with S. G. Vasilyev on the site of Tolbaga were greatly appreciated. The authors would also like to thank A. Gautier, P. Van Peer and two anonymous referees for their helpful comments on the manuscript. L. Wendling checked the English. A. Wauters and W. Miseur helped with the drawings and the photographic work.

References Bazarov, D., Konstantinov, M., Imetchenov, A. Bazarova, D. & Savinova, V. (1982). The Geology and Culture of the Ancient Sites in Western Transbaikal. Novosibirsk: NAUKA (in Russian). Behrensmeyer, A. K. (1978). Taphonomic and ecologic information from bone weathering. Paleobiology 4, 150–162. Binford, L. R. (1981). Bones: Ancient Men and Modern Myths. New York, NY: Academic Press. Brain, C. K. (1981). The Hunters or the Hunted. Chicago, IL: University of Chicago Press. Bugrimova, N. P. (1988). The region of Chita. Physical map for secondary schools. Principal administration for geodesy and cartography of the soviet ministery SSSR, Moscow (in Russian). Bunn, H. T. (1983). Comparative analysis of modern bone assemblages from a San hunter–gatherer camp in the Kalahari Desert, Botswana, and from a spotted hyena den near Nairobi, Kenya. In (J. Clutton-Brock & C. Grigson, Eds) Animals and Archaeology: 1. Hunters and their Prey. Oxford: BAR International Series 163, 143–148. Bunn, H. T. & Ezzo, J. A. (1993). Hunting and scavenging by Plio-Pleistocene hominids: nutritional constraints, archaeological patterns, and behavioural implications. Journal of Archaeological Science 20, 365–398. Cauwe, N., Germonpré, M., Lbova, L. & Tashak, V. (1993). Le Paléolithique supérieur ancien de Sibérie. Les fouilles de Frans Steenhoudt and Vasily Tashak pour les Musées royaux d’Art et d’Histoire. Bulletin des Musées royaux d’Art et d’Histoire de Bruxelles 64, 129–149. Davis, S. J. M. (1980). A note on the dental and skeletal ontogeny of Gazella. Israel Journal of Zoology 29, 129–134. Davis, L. B. & Fisher, J. W. Jr. (1989). A late prehistoric model for communal utilization of pronghorn antelope in the Northwestern Plains region, North America. In (L. B. Davis & B. O. K. Reeves, Eds) Hunters of the recent past. One World Archaeology, 29, 241–276. Eisenmann, V. (1980). Les chevaux (Equus sensu lato) fossiles et actuels: crânes et dents jugales supérieures. Cahiers de Paléontologie, 3–186. Eisenmann, V. (1981). Etude des dente jugales inférieures des Equus (Mammalia, Perissodactylia) actuels et fossiles. Palaeovertebrata 10, 127–226. Enloe, J. G., David, F. & Hare, T. S. (1994). Patterns of faunal processing at Section 27 at Pincevent: the used of spatial analysis and ethnoarchaeological data in the interpretation of archaeological site structure. Journal of Anthropological Archaeology 13, 105–124. Farizy, C., David, F. & Jaubert, J. (1994). Hommes et bisons du Paléolithique Moyen à Mauran (Haute-Garonne). Gallia préhistoire, supplément, 30. Galazii, G. I. (1969). Atlas of the Baikal. Irkoetsk: Akademia Nauk SSSR, Sibirian branch, Limnological Institute (in Russian). Gaudzinski, S. (1995). Wallertheim revisited: a reanalysis of the fauna from the Middle Palaeolithic site of Wallertheim (Rheinhess/Germany). Journal of Archaeological Science 22, 51–66.

Gautier, A. (1987). Taphonomic groups: how and why? Archaeozoologia 12, 47–52. Gautier, A. & Kobusiewicz, M. (1992). Chwalim 1, a MesolithicParaneolithic site in Polish Lowlands: faunal remains. Fontes Archaeologici Posnanienses 37, 65–75. Germonpré, M. (1993). Preliminary results on the taphonomy of Denisova cave 1992 excavation. Altaica 2, 11–16. Goebel, T., Derevianko, A. P. & Petrin, V. T. (1993). Dating the Middle-to-Upper Paleolithic transition at Kara-Bom. Current Anthropology 34, 452–459. Gromova, I. M. & Baranovoi, G. I. (1981). Catalogue of the Mammals of the USSR, Pliocene-Recent. Leningrad: NUAKA (in Russian). Guthrie, R. D. (1990). Frozen Fauna of the Mammoth Steppe, the Story of Blue Babe. Chicago, IL: University of Chicago Press. Guthrie, R. D. & Stoker, S. (1990). Paleoecological significance of mummified remains of Pleistocene horses from the North Slope of the Brooks Range, Alaska. Arctic 43, 267–274. Heptner, V. G., Nasimovic, A. A. & Bannikov, A. G. (1966). Paarhufer und Unpaarhufer, In (V. G. Heptner & N. P. Naumov, Eds) Die Säugetiere der Sowjetunion. Jena: G. Fischerverlag. Hoffecker, J. F., Baryshnikov, G. & Potapova, O. (1991). Vertebrate remains from the Mousterian site of Il’skaya I (Northern Caucasus, U.S.S.R.): new analysis and interpretation. Journal of Archaeological Science 18, 113–147. Imetchenov, A. B. & Kalmikov, N. P. (1988). Fauna and Flora and the Cenozoicum of Buryatia. Ulan Ude: Akademia Nauk USSR (in Russian). Kalmikov, N. P., Piskunov, A. C. & Khandueva, V. B. (1992). Fossil Mammals from the Buryat Natural Museum. Ulan Ude: Akademia NAUK (in Russian). Kasparov, A. K. (1986). Remains of mammals from the Late Palaeolithic site Sukhotino 4 in the Transbaikalia. Proceedings of the Zoological Institute, Leningrad 149, 98–105 (in Russian). Kiesewalter, L. (1888). Skeletmessungen an Pferden als Beitrag zur theorethischen Grundlage der Beurteilungslehre des Pferdes. Leipzig (not seen, cited from Von Den Driesch, A. & Boessneck, J. 1974. Kritische Anmerkungen zur Widerristhöhenberechnung aus Längenmassen vor-und frühgeschichtlicher Tierknochen. Säugetierkundliche Mitteilungen 22, 325–348). Klein, R. G. (1989). Why does skeletal part representation differ between smaller and larger bovids at Klasies River Mouth and other archaeological sites? Journal of Archaeological Science 16, 363–381. Konstantinov, M. & Konstantinov, A. (1991). The beginning of the Late Palaeolithic in the Transbaikal. In Problems of Chronology and Periodic System of Archaeological Sites of South Siberia. Barnaul: Altai Government University, pp. 13–15 (in Russian). Konstantinov, M., Bazarova, L. & Lbova, L. (1994). Palaeolithic cultures of Transbaikalia. Abstracts ‘‘Baikal as a natural laboratory for global change’’, INTAS—Russian Academy of Sciences, Siberian branch, Irkoetsk, 2, 19. Kreutzer, L. A. (1992). Bison and deer bone mineral densities: comparisons and implications for the interpretation of archaeological faunas. Journal of Archaeological Science 19, 271–294. Lavrenko, E. M. & Karamysheva, Z. V. (1993). Steppes of the former Soviet Union and Mongolia. In (R. T. Coupland. Ed.) Natural grasslands. Eastern hemisphere and résumé. Ecosystems of the world 8B, 3–59. Lbova, L. (1991). The palaeolithic site of Kamenka-1: preliminary results of investigation. In (N. Drozdov, Ed.) The Archaeological and Ethnographical Problems of Siberia and the Far East. Krasnoyarsk: Pedagogic Institute of Krasnoyarsk, pp. 90–91 (in Russian). Lbova, L. (1993). The presentation of different cultural traditions in the Early Upper Palaeolithic in the Transbaikal region (Kamenka site). In (T. Mikhailov, Ed.) Problems of the History of Buryatia. Ulan Ude: Buryat Institute of Social Sciences, pp. 14–18 (in Russian). Lbova, L. & Volkov, P. V. (1993). The tool complex of the Palaeolithic site Kamenka-1: typology and function. In (M. V.

Mammalian remains from the Upper Palaeolithic site of Kamenka, Buryatia (Siberia) 57 Konstantinov & L. V. Lbova, Eds) Cultures and Relicts, Stone Age and Early Metal Age in the Transbaikal. Novosibirsk: Nauka, pp. 3–20 (in Russian). Lyman, R. L. & Fox, G. L. (1989). A critical evaluation of bone weathering as an indication of bone assemblage formation. Journal of Archaeological Science 16, 293–317. Marean, C. W. (1991). Measuring the post-depositional destuction of bone in archaeological assemblages. Journal of Archaeological Science 18, 677–694. O’Connell, J. F., Hawkes, K. & Burton Jones, N. (1990). Reanalysis of large mammal body part transport among the Hadza. Journal of Archaeological Science 17, 301–316. Orlova, L. (1994). About the dating of palaeolithic sites in the Transbaikal. In Cultural Traditions in Siberia and America: Ecology and Culture. Materials of the international conference, Chita (in Russian). Ovodov, N. D. (1987). Late Palaeolithic fauna at Tolbaga and Varvarina Gora in the western Transbaikal. In (I. N. Rezanov, L. D. Bazarova & E. A. Khamzina, Eds) Natural environment and ancient people of the Late Anthropogene. Ulan Ude: Akademia Nauk SSSR, geological institute pp. 122–140 (in Russian). Pavlova, M. (1910). Mammifères fossiles du Musée de Trio¨tzkossawsk-Kiakhta. Trudy, Travaux de la sous-section de Triötzkossawsk-Kiakhta, section du pays d’Amour de la Société Impériale Russe de Géographie 13, 21–59 (in Russian with French Summary). Ringrose, T. J. (1993). Bone counts and statistics: a critique. Journal of Archaeological Science 20, 121–157. Sher, A. V. (1974). Pleistocene mammals and stratigraphy of the far northeast USSR and North America. Part II: description of Pleistocene mammal remains. International Geology Review 16, 125–206. Shipman, P., Bosler, W. & Davis, K. L. (1981). Butchering of giant geladas at an Acheulian site. Current Anthropology 22, 1–10. Sokolov, I. I. (1959). On the postcranial skeleton and the outward apperance of Spirocerus kiakhtensis M. Pavlova. Vertebrata Palasiatica 2, 23–33.

Speth, J. D. (1987). Early hominid subsistence strategies in seasonal habitats. Journal of Archaeological Science 14, 13–29. Stiner, M. C. (1991). An interspecific perspective on the emergence of the modern human predatory niche. In (M. C. Stiner, Ed.) Human Predators and Prey Mortality. Boulder, CO: Westview Press, pp. 149–185. Tashak, V. (1993). Podzvonkaja-site, a site of the Early Upper Palaeolithic. In (T. Mikhailov, Ed.) Problems of the History of Buryatia. Ulan Ude: Buryat Institute of Social Sciences, pp. 18–20 (in Russian). Todd, L. C. (1991). Seasonality studies and paleoindian subsistence strategies. In (M. C. Stiner, Ed.) Human Predators and Prey Mortality. Boulder, CO: Westview Press, pp. 217–238. Todd, L. C. & Rapson, D. J. (1988). Long bone fragmentation and interpretation of faunal assemblages: approaches to comparative analysis. Journal of Archaeological Science 15, 307–325. Vasilyev, S. (1993). The Upper Palaeolithic of Northern Asia. Current Anthropology 34, 82–92. Vereshchagin, N. K. & Baryshnikov, G. F. (1982). Paleoecology of the mammoth fauna in the Eurasian arctic. In (D. M. Hopkins, J. V. Matthews, C. E. Schweger & S. B. Young, Eds) Paleoecology of Beringia. New York, NY: Academic Press, pp. 267–279. Vereshchagin, N. K. & Baryshnikov, G. F. (1984). Quaternary mammalian extinctions in Northern Eurasia. In (P. S. Martin & R. G. Klein, Eds) Quaternary Extinctions, a Prehistoric Revolution. Tucson, AZ: University Arizona Press, pp. 483–516. von den Driesch, A. (1976). A guide to the measurements of animal bones from archaeological sites. Bulletin Peabody Museum 1, 1–129. Vorob’eva, G., Medvedev, G., Aksenov, M., Savelyev, N., Gorunova, O., Semin, M. & Svinin, V. (1990). The Stratigraphy, Palaeogeography and Archaeology of Southern Central Siberia. Irkutsk: Irkutsk Government University (in Russian).