Добавил:
Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз: Предмет: Файл:
Yepiskoposyan Azokh Cave and the Transcaucasian Corridor.pdf
Скачиваний:
14
Добавлен:
24.03.2021
Размер:
26.51 Mб
Скачать

214

M.D. Marin-Monfort et al.

demonstrated how the destruction of the organic matrix of bone (collagen) with extensive surface weathering facilitates dissolution and remineralization. Once the bones are buried these changes are more dramatic, with the original mineral of the skeleton being affected by secondary minerals of apatite (brushite, Molleson 1990), CaCO3, Fe2O3 and/or SiO3, and lling empty spaces in its molecular structure (Wyckoff 1972; Francillon-Vieillot et al. 1989).

These modications have been examined for the fossils from Units I to Vm in Azokh 1. Accumulation and preservation of bones in caves limits some types of modication but increases others; for example Smith et al. (2016) report that the bone the Azokh is poorly preserved, with no collagen preservation and extensive mineral alteration; and similarly, Bennett et al. (2016) show that no DNA could be amplied from any of the Azokh fossils. The present analysis will therefore concentrate on the surface modications found on the large mammal fossils recovered from Azokh 1. The small mammal taphonomy of the site is described by Andrews et al. (2016).

Materials and Methods

The fossil collections studied here come from the excavations and prospection work carried out in Azokh 1 Cave between 2002 and 2009. The total number of fossils studied here is 1879 specimens (Table 10.1). The fossils analyzed here all come from the back of the Azokh 1 passageway (see Murray et al. 2016). A test trench of 2 × 2 m was dug in 2003 to establish the limits between stratigraphic units, as well as to conrm lithic and fossil content richness. This test trench was made from mid Unit II to the limit of Unit III to IV. Due to the unstable sediments, the bottom of the trench did not properly reach Unit IV, which has yet to be excavated. Huseinovs excavation team indicated that Unit IV only contained microfauna (Ellobius lutescens and Microtus sociales; Markova 1982). We, however, may conrm the presence of cave bear fossils in Unit IV, for during the clearing of wall sediment from this unit, fossil teeth and bones of cave bears and other mammalian species were recovered. Finally, material recovered from the middle part of Unit V, Unit Vm, comes from an excavation area exposed by previous excavations. The lower part of Unit V, which is very thick, has not yet been reached.

Unit II is currently under excavation, with an area of 40 m2, but some of the Unit II fossils also come from the test trench described above. Unit III is restricted to the 2 × 2 m test trench. Fossils from Unit Vu come from a small excavation mainly done in the 2002 and 2008 seasons covering 8 m2. Unit Vm has

been exposed over an area of 24 m2 and has yielded both stone tools and fossil bones. Units Vm to II at Azokh 1 date from 300 to 100 ka, with Holocene periods recorded in Unit I.

Taxonomic faunal identications by Van der Made et al. (2016) may slightly differ from those identied here, because small broken fragments that could only provide a rough taxonomic identication (order family level) or only anatomical identication and animal size as well as unidentied bone splits are also included in this study.

All fossils were analyzed using a 6.3x to 50x stereoscopic light microscope (Leica MZ 7.5). A selection of these fossils (N = 22) was also analyzed using a scanning electron microscope (SEM). Two SEM microscopes were used a QUANTA 200 Environmental SEM and a FEI-INSPECT Low Vacuum SEM and both of them are housed at the Laboratory of Non-Invasive Techniques of the Museo Nacional de Ciencias Naturales of Madrid. SEM detectors were used in backscattered electron (BSE) mode combined with secondary electron (SE) emission mode, at 2030 kV, 0.60.33 Torr. Both types of microscopes enable specimens to be directly analyzed at high magnication and high resolution with no necessity for coating or any other pre-treatment. Histological analyses to establish the presence of bacterial attack or any other modication in the interior of the bone were carried out on sectioned bones (N = 53). Oxford energy dispersive spectrometry (EDS) detectors provided the chemical element composition of specic areas of interest. Some samples were analyzed by X-ray diffraction (Philips PW-1830) to obtain their mineral composition (N = 26).

Anatomical Elements and Species

Identification

Fossils were identied by Species; Body part; Segment and Portion (diaphysis, proximal and distal end; complete; lateral; body, process; arch). Dental eruption and wear, epiphyseal fusion and bone texture determined age (immature infant or juvenileor adult). We distinguished (see Supplementary Information) between number of remains (NR), number of identied specimens (NISP), minimum number of elements (MNE) and minimum number of individuals (MNI), in accordance with the criteria suggested in Lyman (1994). NR covers all recognized fragments, while NISP covers identiable specimens. MNE was calculated taking into account age, portion, and size. Calculations of survival rate (Brain 1969) or Relative Abundance (Andrews 1990) refer to the value that may be expected in the light of their MNI (%Relative Abundancei = MNE × 100/number of element i in the animal skeleton × MNI).

10 Taphonomy and Site Formation of Azokh 1

215

The skeletal proportions recovered from archaeological sites depend on the bias caused by the agents involved in the formation of the deposit (e.g., Lyman 1984). In order to check differential preservation arising from the intrinsic nature of fossil bones, we compared Relative Abundances of each bone element with its density obtained by photon densitometry (Lyman 1984). Specimens attributed to large species and sizes were compared with the values obtained for Bison (Kreutzer 1992), medium-sized species with data for reed deer (Hillson 1992) and small sizes with those of sheep (Lyman 1984). Lam et al. (1998) recorded bone mineral density values of long bones using computed tomography. The technique is more accurate than photon densitometry, but both methods show values that correlate well (R2 = 0.47 and rs = 0.68).

Correlations between density of the anatomical element (Ei) and the Relative Abundance (Ri) was statistically analyzed by means of the Spearmans correlation coefcient (r Spearman). The signicance level used was p = 0.05. Differential conservation of the anatomical element was identied by a signicant correlation. The high breakage observed at Azokh 1 results in restricted taxonomic and anatomical identications. Different anatomical categories have been distinguished taking into considerations cranial skeleton (skull, mandible and isolated teeth) axial skeleton (vertebrae, ribs), including girdles (scapulae and pelves), and major long bones (humerus, radius, ulna, femur, tibia, bula).

We also distinguished weight size groups according to the classications of Rodríguez (1997) and Blumenschine (1986), according to species present in the site (see Van der Made et al. 2016):

Large sized: >300 kg includes adults and immature individuals of Ursus spelaeus and rhinoceros (Stephanorhinus hemitoechus, S. kirchbergensis), as well as adults of Equus ferus and Equus caballus, and Bison schoetensacki.

Medium sized: 100300 kg includes Cervidae (immature

Cervus elaphus and adult Dama sp.), adult Sus scrofa, immature Equus ferus and Equus caballus, both adults and immature individuals of Equus asinus, and small bovids (Ovis, adults).

Small sized: <100 kg. Adult and immature Capreolus pygargus, Capra and Saiga. Immature Dama, Sus and

Ovis, and all carnivores identied in Azokh 1: canids (Vulpes vulpes, Canis lupus¸ Canis aureus), hyenid

(Crocuta crocuta), felids (Panthera pardus) and mustelids (Meles meles).

Carnivores are mainly restricted to Unit V (Vu and Vm), except for canids and panther that are also present in Unit II. Identications by Van der Made et al. (2016) include the fossil collections from previous excavations. These identications

include all species from Unit VI that have not been excavated by the present excavation team, except for dental remains of Dama aff. peloponesiaca found during geological sampling of this unit. The following species from Unit V also were identied from previous excavations: Meles meles, Martes cf. foina, Lynx sp., Felis chaus, Panthera pardus, Equus hydruntinus and

Megaloceros. Similarly, the following species from Unit III came from earlier excavations: Equus hydruntinus,

Stephanorhinus hemitoechus, Capreolus pygargus, Dama mesopotamica, Ovis ammon, Capra aegagrus.

Shape, Size and Fracture

Length/Width/Thickness were measured on every fossil with micrometric calipers. Lineal dimensions (mm) were classied according to Cáceres (2002) who uses the following four categories: A >2 cm, B 25 cm, C 510 cm, D >10 cm. Three orthogonal dimensions were also used to characterize the shape of the fossils (D1 length, D2 width, D3 thickness). Based on the original work of Wentworth (1919) to characterize the morphology of sedimentary particles, Frostick and Reid (1983) applied this methodology to fossils, and Blott and Pye (2008) increased the numbers of shape categories to eight. This is a bi-variant approach that relates D2/D1 in ordinates and D3/D2 in abscissas, and it shows the variety of shapes found in fossils ranging from those that are laminar/blade in form (Category 1) to those that are the most rounded/spheroid (Category 8). This approach is useful to analyse shape selection, absent, for example, if all shape categories are present, and it is usually related to hydrologically transported fossils. Similarly, Voorhies groups (Voorhies 1969) also discriminate between potentially transported fossils, this time based on the original (hydrological) shape and weight of anatomical elements which distinguishes those fossils on the basis of their potential for being transported. Other authors (Behrensmeyer 1975; Boaz and Behrensmeyer 1976; Korth 1979) repeated experiments initiated by Voorhies to obtain a more comprehensive classication.

The methodology to determine breakage patterns suggested by Bonnichsen (1979) and Bunn (1983) and completed by Villa and Mahieu (1991) was used and the following traits were recorded:

(1)Number of fractures.

(2)Fracture angle: oblique/right/mixed

(3)Fracture outline: transverse/curved-V-shaped/intermediate.

(4)Fracture edge: smooth/jagged.

(5)Shaft circumference: 1 = circumference is <1/2 of the

original; 2 = circumference is >1/2 of the original; 3 = complete.

216

M.D. Marin-Monfort et al.

(6) Shaft fragmentation: 1 = shafts <1/4 of original length;

2

= length

between

1/4

and

1/2

of

original

length

3

= length

between

1/2

and

3/4

of

original

length

4

= length >3/4 of original length (complete).

 

Villa and Mahieu (1991) compared three different French sites, Fontbrégoua a Neolithic site (4000 BC), and the collective burials of Bezouce and Sarrians (Late Neolithic, 2500 BC). At Fontbrégoua there is evidence of cannibalism and hence frequent traces of long bone breakage when the bone was still green or fresh (Villa et al. 1986). At Bezouce, the sub-fossil bones were broken by impact, and at Sarrians fossil bones were broken by sediment pressure.

Surface Modification Related to Breakage

Peeling, a term coined by White (1992), denes a roughened surface with parallel grooves or brous texture produced when fresh bone is fractured and peeled apart, similar to bending a small fresh twig with two hands. Peeling was recorded as present/absent for each fossil.

Percussion pits: These are of variable size and depth on the side of the bone opposite to where impact scars and the resulting fracture were produced. Friction of the bone against the surface on which it was resting when struck (see Blumenschine and Selvaggio 1988) produces a series of pits and scratches identied as rebound pointsby Johnson (1985), Percussion striaeby White (1992), contrecoupby Leroi-Gourhan and Brezillon (1972) or the most used term hammerstone/anvil scratchesby Turner (1983). These pits and scratches were recorded as present/absent.

Adhering akes: These are bone akes that adhere to the fracture surface of a specimen. These akes are caused

by curving incipient fracture lines, often hairline, which are sub-parallel to the fracture edge. These were recorded as present/absent.

Conchoidal percussion scars: Following Blumenschine

(1988), we distinguished between notches: arcs on the bone edge; and flakes: bone fragments splintered off by the impact.

Tool-Induced Surface Modifications

Morphology, emplacement and distribution of striations distinguish between incisions (slicing marks and sawing marks), chop marks and scraping marks. Each type of mark results from the different application of a tool on

the bone or a combination for different purposes (deeshing, dismembering or grease removal). Striation distribution (isolated marks/grouped/widespread) and orientation (oblique/transversal/longitudinal) was described for each cut mark, chop mark or scrape mark, according to the size of the mammal species. Striation length was also measured (maximum and minimum lengths when sets of cuts occurred).

Incisions: These are long striations of variable length and width characterized by a transversal V-shape section, internal microstriations and lateral roughness (Hertzian cones, Bromage and Boyde 1984). Similarly, saw marks (Noe-Nygaard 1989) are produced by a repetitive and bi-directional motion.

Scrapes: These are shallow subparallel multiple cut marks (Noe-Nygaard 1989) caused when a stone tool is dragged transversally along the length of the bone. It is traditionally assumed that this removes periosteal and grease. Some authors (Binford 1981) suggest that scraping marks are caused by the removal from the bone surface of any substance that may absorb the blow when breaking the bone to extract marrow.

Chops: These marks are the result of striking the bone surface with a sharp stone tool, leaving deep, wide and V-shaped scars. The action may be related to cutting strong muscle attachments or dismembering. They also may have internal microstriations.

Tooth Marks

Tooth marks were described and measured separately for all anatomical items following Andrews and Fernández-Jalvo (1997) and modied (written below in italic) in Fernán- dez-Jalvo and Andrews (2016):

a = Shallow pits on diaphyses of limb bones (minimum dimension);

a1 = Deep perforations on shafts of limb bones1;

b = linear marks on surface of bone (transverse measurement of grooves, minimum dimension)

b1 = linear marks on ends of bones1

b2 = linear marks on articular bone1

c = Deep perforations on articular ends of bones;

d = Deep perforations on the edges of spiral breaks;

e = Deep perforations on the edges of transverse breaks; f = Deep perforations along edges of split bones;

1These modications were not recognized in our initial classication because they were not present in the original classication by Andrews and Fernández-Jalvo 1997.

10 Taphonomy and Site Formation of Azokh 1

217

g = Multiple perforations on the bone surface made by multi-cusped teeth;

h = Deep perforations on anatomical edges;

i = Double arched puncture marks on crenulated edges.2

Punctures on spiral breaks (category d) are related to carnivore breakage, and they also depend on the size or type of the anatomical element (e.g., femur vs. radius, and at vs. long bone) and the size of the prey (large, medium, small sized animals). Breakage category emay not be produced by carnivores, but may be the result of a relatively deep puncture mark on diaphysis (category a) that facilitates post-depositional or diagenetic breakage. This has been seen at Sima de los Huesos site (Andrews and Fernández-Jalvo 1997) where single bones with two fragments having transversal breakage have a puncture mark at the break recorded on both fragments. Bones chewed by hominins have been identied at Azokh Cave (Fernández-Jalvo and Andrews 2011).

These categories have been adapted to other types of measurements of tooth marks taken by different authors investigating modern carnivore tooth marks (Selvaggio and Wilder 2001; Dominguez-Rodrigo and Piqueras 2003; Pobiner 2008; Delaney-Rivera et al. 2009) as follows:

pc: puncture marks on compact bone (category a, pits on diaphyses, and punctures on broken edges: categories d, e, f, i)

pac: puncture marks on cancellous or articular surfaces (category c, pits on epiphyses)

gc: grooves on compact bone (category b, scores on diaphyses)

gac: grooves on cancellous or articular surfaces (category b1/b2, scores on epiphyses).

We use these four categories, and sizes of tooth marks have been shown graphically in box plot diagrams. This shows the median that separates the higher half of the sample (upper quartile) from the lower half of the sample (lower quartile), as well as the range of measurements (sample minimum and sample maximum) and outliers (data exceeding the data represented in boxes). Following Andrews and Fernández-Jalvo (1997) methodology, the smaller dimension of pits and scores was always measured, which is equivalent to minor axisfor Delaney-Rivera et al. (2009), or breadthfor Domínguez-Rodrigo and Piqueras (2003) and Pobiner (2008) tooth marks produced by modern carnivores. Measurements obtained by Pinto and Andrews (2004), Pinto et al. (2005) and Rabal-Garcés et al. (2011) have been applied or adapted to the categories of Andrews and Fernández-Jalvo (1997) for cave bear fossil sites.

2This category has been proposed in Fernández-Jalvo and Andrews 2011.

Other Surface Modifications

Cracking: three different categories were distinguished under the light microscope: a01 very supercial and very thin cracking; f23: ssures, wider and deeper cracks; e45: exfoliation of cracked surface following stages described by Behrensmeyer (1975). We also distinguished cracks that show raised or warped-up ridges similar to mud-cracks that contrast with weathering cracks where the edges are just separated. These cracks differ from those produced by weathering, the edges of which are even. They were recorded as presence/absence.

Concretions: cemented sediment heavily attached to the fossil, sometimes with manganese dioxide stains.

Rodent gnawing marks, trampling marks, polishing, rounding, root-marks, and soil corrosion. Distribution of these disturbances on the fossils was described as isolated (I: a single mark), clustered (C: in patches or on a particular area on the fossil) or widespread (W: almost covering the whole fossil surface).

Sediment friction marks refer to processes that entail movement of bone against a rocky/sandy substrate or friction of rocks falling on bones causing multiple randomly dispersed scratch marks, usually transverse to the length of the bone, as the bones are rubbed against the stones. These marks are also described as trampling marks

by animals (Andrews and Cook 1985), including humans (Domínguez-Rodrigo et al. 2009) due to bones pressed into the rocky substrate. In archaeological contexts, trampling marks strongly mimic cut marks made by stone tools, and distinction between them is especially relevant. Criteria used to distinguish cuts by stone tools from scratches due to trampling are based on orientation and location of striations on anatomical areas (near articular ends, muscle

insertions or tendon attachments), which are congruent with butchering purposes (disarticulation, deeshing,

skinning off or cleaning the bone from fat).

Abrasion (by water or wind) may also produce scratches, although striations are microscopic in size (Fernández--

Jalvo and Andrews 2003). Rounding affects broken edges as well as anatomical protuberances.

Root-marks (linear marks on bones) are the result of roots of vascular plants in symbiosis with fungi or bacteria, and marks are commonly divided into branches and show signs of chemical corrosion at their interior. There are no fossils affected by root-marks at Azokh 1.

Soil corrosion on bone surfaces indicates the side of the bone that has been in direct contact with an acidic ground under constant humidity. When lifted from the sediment during excavation, fossils were marked with a permanent marker placed on the side that had been in contact with the soil.