Cranium Mandible

Cranium Mandible



8 mm Teeth

'Standardized Minimum Animal Units'

Figure 7.6 Relative frequencies of the main skeletal elements of horses and bovids in different stratigraphie levels of Combe Grenal, calculated as in Fig. 7.5. Bovids include remains of both Bos and Bison; owing to the scarcity of remains of bovids throughout the sequence, all of the remains of these taxa from the Wurm II levels (layers 11-35) have been combined. From Chase 1986a: Fig. 8, Tables A48-51.

of horse remains in these levels shows some similarities to that of reindeer in the same levels, there are three important differences: first, the ratio of head and tooth remains to those of postcranial bones is much higher for the horse than for reindeer; second, horse remains show an unusually high frequency of pelvic bones, substantially higher than those for any other species in any other level of the site; and third, there is a curious over-representation of bones from front limbs (especially in levels 25-23) over those of rear limbs. The significance of the latter features remain enigmatic, although it may be recal led that a similar over-representation of pelvic remains was also documented in many of the assemblages of red deer bones in the Würm I levels of the site. Whatever the significance of this feature in economic or car-case-processing terms (cf. Chase 1986a: 54) there is a marked contrast between relative frequencies of pelvic remains of horses and reindeer recorded in these particular levels.

The second distinctive pattern of horse remains is seen in the later part of the Würm II levels (layers 19-11) corresponding with the final Quina-Mousterian levels and the overlying levels of Denticulate Mousterian.

The total assemblage of horse bones from these levels is small (including only 19 post-cranial remains, together with 528 teeth) but nevertheless shows a number of idiosyncratic features suggesting a major change in carcase utilization patterns (Fig. 7.6). In the first place, these levels show an even higher proportion of head and tooth remains than those documented in the immediately underlying levels - with remains of heads and jaws now massively dominating the faunal assemblage. Second, the postcranial bones consist overwhelmingly of a single skeletal element -i.e. ulnae from the lower front limbs. With the exception of isolated specimens of radii, car-pals, metapodials and phalanges, all other parts of the skeleton are entirely lacking. Even allowing for the exceptionally small sample size, this suggests a curiously different pattern of skeletal representation from that documented in any of the underlying levels of the Combe Grenal sequence.


Finally, Chase presents some interesting if limited data on the representation of large bovid remains - i.e incorporating the remains of both aurochs (Bos primigenius) and steppe bison (Bison prisons) which are effectively impossible to separate for the majority of bones. The remains of these species are too sparse in most parts of the sequence to allow a detailed, layer by layer analysis, and to obtain a sample of adequate size Chase had to combine material from all parts of the Würm II levels into a single composite sample. Because of this and the impossibility of differentiating between the remains of Bos and Bison, any detailed analysis of these remains would, as Chase points out, be rather premature. Nevertheless, the general pattern which emerges from the combined Würm II sample appears to show several features which are in many ways similar to those recorded for red deer in the Würm I levels and to some extent those of horse in levels 25-20 (Fig. 7.6). In general, remains from the lower limbs (particularly the radiocubitus, ulna and tibia) dominate those of the upper limbs (femur and humerus) and there is also an over-representation of remains from the head and teeth. These patterns again emphasize the more marginal parts of the carcase in preference to the main meat-bearing bones -especially in view of the total absence of scapulae and virtual absence of ribs and vertebrae, which in the case of these very large animals are extremely large and robust bones, unlikely to be easily destroyed or overlooked in excavations. Perhaps the most curious feature of the bovid remains is the virtual absence of extreme distal portions of limbs, represented by the metapodials and phalanges - again, large and easily identified bones which are unlikely to be destroyed or missed in excavations. This contrasts with the patterns documented for almost all other species in the site and must surely indicate either that the feet of these large bovids were rarely brought into the Combe Grenal site or that they were systematically removed or discarded elsewhere.

Skeletal patterns at other sites

One of the central questions in this context is how far the particular patterns of bone element frequencies documented by Chase at Combe Grenal can be regarded as typical of Middle Palaeolithic faunal assemblages from southwestern France and how far they may be in some way specific to this particular location. Detailed information from other sites in the region is still limited. The best data come from the studies by Guy Laquay (1981) on several of the faunal assemblages recovered from Bordes' excavations in the Würm I levels at Pech de l'Aze sites II and IV (Bordes 1972, 1975a) and from the studies of Stephane Madelaine (1990) on the assemblages excavated by Denis Peyrony from the sites of Le Moustier, La Ferrassie, Les Mer-veilles and Gare de Couze and now stored in the National Museum of Prehistory at Les

• Combe Grenal Q Other sites

Figure 7.7 Relative proportions of postcranial remains to teeth recorded for different species in Mousterian faunal assemblages from southwestern France. The data show clearly that remains of horses and bovids are represented in general by much lower overall frequencies of postcranial remains to teeth than those of both red deer and (especially) reindeer. Infilled circles indicate data from Combe Grenal. From Chase 1986a; Laquay 1981; Madelaine 1990 and other sources.

• Combe Grenal Q Other sites

Figure 7.7 Relative proportions of postcranial remains to teeth recorded for different species in Mousterian faunal assemblages from southwestern France. The data show clearly that remains of horses and bovids are represented in general by much lower overall frequencies of postcranial remains to teeth than those of both red deer and (especially) reindeer. Infilled circles indicate data from Combe Grenal. From Chase 1986a; Laquay 1981; Madelaine 1990 and other sources.

Eyzies. Further analyses have been provided by Binford (1988) for the material from the major occupation horizon (layer VIII) at the Grotte Vaufrey and by Delpech (in press) for the material from levels 2-5 at Le Regourdou. In common with Combe Grenal all these sites are either cave or rock-shelter locations situated within or closely adjacent to major river valleys (Fig. 8.1).

Few of these sites have provided sufficiently large samples of faunal remains to permit the kind of detailed analyses provided by Chase for the Combe Grenal assemblages and for the majority of assemblages we are reduced to some rather gross comparisons based mainly on overall frequencies of cranial (mainly teeth) as opposed to postcranial parts of the skeleton (Fig. 7.7). For the assemblages recovered from earlier, pre-war excavations (Le Moustier, La Ferras-sie, Les Merveilles and Gare de Couze) it should also be kept in mind that the precise standards of faunal recovery employed during the excavations are unknown - although from the composition of the surviving collections there is no reason to think that the material was collected in a significantly more selective or biased way than in more recent excavations. For the two cave sites (Le Regourdou and Grotte Vaufrey) there is the added complication that a part of the faunal material almost certainly derives from intermittent use of the caves as carnivore dens, rather than from residues of human occupations (Binford 1988; Delpech, in press). Even so, there are some interesting and significant patterns which emerge from these studies, which provide some useful comparisons with the better documented material from Combe Grenal (Fig. 7.7).

1. In at least five of the assemblages documented here remains of reindeer are represented by relatively high ratios of postcranial bones to teeth, strongly reminiscent of those documented throughout most of the occupation levels at Combe Grenal, especially from the main Quina-Mousterian levels in layers 25-20 (cf. Figs 7.5, 7.7). This can be seen, for example, in the assemblages from the Quina or Ferrassie-Mousterian levels at Le Regourdou (layer 2), Les Merveilles (lower layer) and Pech de TAze IV (layer II), as well as in the MTA levels at both Gare de Couze and Les Merveilles (upper layer). The consistency of this pattern, now documented in so many sites suggests strongly that reindeer carcases in general were treated differently from those of most other species, and were more often introduced into the sites as relatively complete carcases rather than as more selective parts. Some of the possible implications of this are discussed in the following section.

2. A second, sharply contrasting pattern can be seen in remains of the much larger species of large bovids (either Bos or Bison) and horse. Detailed analyses of these taxa are frequently hampered by small sample sizes but for almost all the faunal assemblages discussed here, remains of these species are consistently represented by much lower ratios of postcranial bones to teeth than for reindeer (Fig. 7.7) - for example in the assemblages from the Typical and Charen-tian Mousterian levels at both Pech de 1'Aze II (layers 4B and 4C2) and Pech de l'Aze IV (layer 12), and in the MTA levels at Gare de Couze and Le Moustier (layers G and H). The clear implication is that these large species were usually introduced into the sites with a much heavier emphasis on heads and jaws, rather than on the postcranial carcase - once again recalling patterns recorded for the same species at Combe Grenal. That this pattern cannot be dismissed simply as an artefact of selective preservation (i.e. as indicating better survival prospects for teeth than bones in occupation deposits) is clearly demonstrated by the fact that remains of reindeer recovered from precisely the same occupation levels (at Gare de Couze, Les Merveilles and Pech de l'Aze IV) show effectively the reverse of this pattern. The only exception is the small assemblage of horse remains recovered from the Rissian level at Grotte Vaufrey, which shows a predominance of postcranial bones and an exceptionally low incidence of head and teeth remains. As Binford (1988) has pointed out, however, this could conceivably be either an artefact of small sample sizes or, more likely, an indication that the horse carcases were introduced into the site mainly by carnivores, rather than by humans.

3. Remains of red deer show a pattern which appears to be generally closer to that of the larger herbivore species (i.e. bovids and horse) than to reindeer. In five of the documented assemblages (layers J3a and X at Pech de 1'Aze IV, layer 4B at Pech de 1'Aze II, layer VIII at Grotte Vaufrey and layer J at Le Moustier) there is a much stronger emphasis on heads and teeth than on the postcranial skeleton (Fig. 7.7) - again recalling Chase's results for the same species in the majority of the occupation levels at Combe Grenal (Fig. 7.5). One notable exception is the assemblages from levels 2-5 at Le Regourdou, where remains of red deer are represented overwhelmingly by various postcranial bones and with only sparse representation of heads and teeth (Delpech, in press). This raises the possibility of a significantly different pattern of utilization of red deer carcases at this site which could possibly be related to the peculiar structural or even ritualistic features reported by Bonifay (1964) from these levels. However, Delpech warns that many of the red deer remains at Le Regourdou could just as easily result from carnivore activity in the cave, rather than from the relatively ephemeral episodes of human occupation. To regard the data from Le Regourdou as an exception to the normal pattern of red deer representation in southwestern French sites would almost certainly be premature from the evidence to hand.

4. Remains of species other than those discussed above are too poorly represented in the majority of southwestern French sites to allow any reliable assessment of patterns of bone element frequencies. Laquay (1981), however, has presented some interesting if limited data on remains of roe deer recovered from level 52 at Combe Grenal and layer J4 at Pech de FAze IV. According to Laquay each assemblage is characterized by a ratio of postcranial to cranial remains which is much higher than those recorded for either red deer or horse in the same levels, and which is comparable with the patterns of reindeer remains in the various sites discussed earlier. Similar patterns have been documented, on rather small samples, for the remains of wild boar from two sites - Le Regourdou (layer 2) and Pech de 1'Aze IV (layer J3a) (Laquay 1981). It may be significant that both of these are relatively small species, with body weights not too different from those of reindeer (Van den Brink 1967). While it would be premature to draw too many conclusions from such limited samples, it may well be that the carcases of these smaller animals were exploited and processed by Middle Palaeolithic groups in a broadly similar way to those of reindeer.

Was this article helpful?

0 0

Post a comment