Discussion

Presence of Humans in Azokh 1 Cave

Evidence for the presence of humans is recorded in all levels of the excavation in Azokh 1. Fragmentary human fossils have been found of H. heidelbergensis from Unit V and H. neanderthalensis from Unit II (King et al. 2016), and lithic implements made by humans have also been recovered from all levels of this site (Asryan et al. 2016) together with stone tool induced damage on some of the fossil bone.

Carcasses of animals were dismembered and butchered (cut, sawn and scraped). Once animals were free of meat and skin, bones were broken to extract the marrow. Signs of this human induced breakage are cut marks, impact and percussion marks, peeling, conchoidal scar and adhered bone flakes, and these affect 125 fossils (6.7% of total NR in Azokh 1). The few complete fossils found at the site are skeletal elements with low marrow content, and these were left unbroken by humans. The complete sequence of butchering has been observed on cave bears at Unit II. Higher abundances of cut marks have been distinguished on limbs and axial skeleton. Butchering has also been observed on medium sized animals, but small sized animals have less stone tool induced damage. Burnt bones may be assumed to be the result of human action in Azokh 1, because the excavation area from where the burnt fossils have been recovered is far from the cave entrance and they would be unlikely to have been burnt by natural fires. Unit 1 yielded 72 burnt bones, with 10 burnt in Unit II and five in Unit Vm, making 4.6% of the total NR for the site as a whole.

Chewing by humans was found on a single rib fragment from Unit I (Fernández-Jalvo and Andrews 2011). The ends of the rib were bent during human chewing by pushing up or down on the ends of the bone with the hands and holding the ends between the teeth. This type of damage was named as fraying by Pobiner et al. (2007) and experimentally reproduced in humans (Saladié 2009; Fernández-Jalvo and Andrews 2011; Saladié et al. 2013) and chimpanzees (Pickering and Wallis 1997; Plummer and Stanford 2000).

Carnivore Damage

Carnivore tooth marks have been identified on 120 fossils from Azokh 1 (6.4%), but only 30 of them have tooth marks on their broken edges, which suggests that carnivore action was unimportant in producing the breakage at the site. Pinto and Andrews (2004) and Pinto et al. (2005) have done an extensive study of various sites in the Iberian Peninsula with Ursus spelaeus and Ursus arctos as part of the faunas. These authors investigated sites that yielded only cave bears (Troskaeta, Tito Bustillo, Eirós, named monospecific) and compared them with sites where other carnivores were found together with bears (cave bears at Arrikrutz and brown bears in a modern natural trap, Sima de los Osos from Somiedo). Rabal-Garcés et al. (2011) applied the same methodology to the site Coro-Tracito (Huesca, Spain) that is also monospecific.

Two of the monospecific sites, Tito Bustillo and Cova Eiros, have been distinguished as denning areas for female cave bears with young (Pinto et al. 2005), and the bones from these two caves have few small chewing marks, but a range of sizes to over 7 mm (Fig. 10.12). A similar pattern is seen for Coro Tracito (Rabal-Garcés et al. 2011), but Troskaeta has a more uniform distribution, although still with many marks greater than 7 mm. Comparison of these sites with Azokh 1 (Fig. 10.12) shows a lower intensity of tooth marks and more limited range of sizes in the Azokh 1 fossil assemblages. Unit I has not yielded any fossil carnivores (except for reworked bear fossils from Unit II brought into Unit I by modern burrowers), but it has provided the highest abundance of chewing marks which are all smaller than 4 mm. This is similar to the fox-ravaged assemblage from Neuadd (Wales) described by Andrews and Armour-Chelu (1998). Unit I has been compared with Atapuerca TD6, where a small canid of similar size to foxes was identified by Díez et al. (1999), although the species may be different. It is likely that the type of carnivore responsible for the chewed bones in Unit I, at that late stage of the cave infilling, was either dogs of the people that inhabited the cave, or wild jackals or foxes which still live in the area today.

With regard to lower units of Azokh 1 (II to Vm), the range of sizes of puncture marks and grooves is diverse (Fig. 10.8), with minimum values (smaller than 4 mm) on compact bone and larger than 7 mm on cancellous tissues. This may be due to either other carnivores involved in the site (felids, canids, or even mustelids, all of which are recorded in the fossil fauna) or to the presence of different sizes of cave bears: adult males are much bigger than females, and juveniles. Azokh 1 puncture marks have a lower abundance and smaller sizes than those recorded at monospecific U. spelaeus sites (Pinto and Andrews 2002; Pinto et al. 2005; Rabal-Garcés et al. 2011). The minimum dimension of puncture marks on diaphyses, ‘category a’

(Fig. 10.8, ‘pc’), have mean values ranging from 2.6 to 5.5 mm and maximum values between 4.0 and 8.8 mm. These punctures on diaphyses or compact bone are larger than those produced by any extant carnivore including hyenas (mean 1.5–2.24 and max 2.1), lions (mean 1.1–2.2 and max 2.3) or panthers. On the contrary, pit breadth on epiphyses of the Azokh 1 fossil assemblage (pac ‘punctures on articular or on cancellous tissues’) and score/groove breadth (gc or gac, minor axis measured in Azokh 1) provide similar or even smaller values than modern lions or hyenas. The maximum size of punctures recorded from Azokh 1 fossils (minor axis/breadth) is 8.8 mm for puncture marks on diaphyses (pc) and 8.7 mm for those on cancellous bone

(pac). A major axis/length of 17 mm was recorded for one tooth print (Fig. 10.7a). These dimensions are too large for lions, for which the maximum records of punctures are 6.3 mm on epiphyses – minor axis – and 8.16 mm on metaphyses – major axis – (Delany-Rivera et al. 2009).

At Azokh 1, the low number of fossils with carnivore chewing marks, the low proportions bone splinters, and low breakage associated with tooth marks all reject the involvement of carnivore bone breakers such as hyenas or wolves at any level of Azokh 1. Lions are not bone crushers. They leave relatively low numbers of tooth marks on bone, produce few bone splinters, and in particular they leave almost no marks on the limbs of the carcasses of their prey (Dominguez-Rodrigo 1999). Cave bears are also not bone crushers. Bone accumulations, documented by Pinto and Andrews (2004) and Pinto et al. (2005), that are comprised solely of cave bears showed percentage completeness ranging from 42 to 84% (Pinto and Andrews 2004). Furthermore, Haynes (1983) observed that bears could occasionally use their cheek teeth and leave characteristic scratches on the shaft resembling those made by rodents: short and parallel, shallow etched straight score lines

(Haynes 1983, p. 169). Some grooves observed on some fossils from Azokh 1, too far from the edge to be rodent made tooth marks (Fig. 10.7), may fit with this description.

Some differences in the tooth mark sizes can be observed in units from Azokh 1 (Fig. 10.12). Unit Vu in particular is the only unit at Azokh 1 that has a high proportion (21.4%) of tooth marks greater than 7.1 mm, and the distribution of tooth mark sizes is similar to those of Troskaeta and Coro Tracito (Fig. 10.12), which are monospecific sites of cave bears. Unit III has similarities to Arrikrutz (although tooth marks larger than 7.1 mm have not been observed). Arrikrutz is a site where there were mixtures of different sized carnivores chewing the bones, including bears (Pinto and Andrews 2004; Pinto et al. 2005). Units II and Vm (the former with 4.9% of tooth marks greater than 7.1 mm, have distributions of tooth mark sizes similar to that of Unit I with the largest sized tooth marks present, and like Unit I, it is likely that the fossils from Units II and Vm had been chewed by a small carnivores like a fox or jackal and possibly by cave bears as well.

This brings to a controversial subject with regard to the diet of U. spelaeus. Physiological studies based on skull, mandible and tooth morphology have inferred a largely herbivorous diet for this cave bear (Kurtén 1976; Mazza et al. 1995; Mattson 1998; Grandal d’Anglade and LópezGonzález 2005). Figueirido et al. (2009), however, showed indications of omnivorous diet based on morphometric analyses of the skull and dentition of U. spelaeus. Several

studies based on isotopic signals (e.g., Bocherens et al. 1994; Fernández 1998; Vila Taboada et al. 1999, 2001; Fernández et al. 2001) provided strong evidence that Ursus spelaeus was highly herbivorous. Brown bears, on the other hand, have isotopic signals of pure carnivory in spite of their observed omnivorous diet (Bocherens et al. 1997, 2006). In an analysis of cave bear from a cave in Romania, Richards et al. (2008) found that the cave bear teeth of Pestera cu Oase had higher nitrogen isotope values than seen in herbivores, and they could therefore be considered omnivorous. Dental microwear has also provided evidence of an omnivorous diet before dormancy (Pinto et al. 2005; Peigné et al. 2009). Finally there is the evidence discussed here of cave bear sites which have bones preserved with carnivore tooth marks larger than those of any of the usual makers of tooth marks, such as hyenas and wolves, and which lack evidence of other carnivores being present.

The low rates of chewing marks and near absence of bone splinters excludes hyenas or any other bone crusher carnivore being active at Azokh 1. The large size of tooth marks on bone diaphyses also excludes small carnivores and larger species such as lions, which even though they are not bone breakers, produce smaller sized tooth marks on the bones of their prey. These results from Azokh 1, as well as from other sites that only yielded cave bears, leaves little margin for doubt that in some cases cave bears eat meat and chew bones. Environment impoverishment, extreme climate conditions or just population variability might affect the extent to which cave bears have this behavior in the different sites where U. spelaeus is present. Finally, even strict herbivores, such as deer, reindeer, cows or camels, may also chew bones (Sutcliffe 1973, 1977; Brothwell 1976; Johnson 1985), and some populations of deer and fallow deer may do so intensively (Cáceres et al. 2011). This behavior in ungulates stems from nutritional deficiencies in the environment (Grasman and Hellgren 1993). In the case of U. spelaeus, even if they are more herbivorous in some areas (e.g., Bocherens et al. 1994; Fernández 1998; Vila Taboada et al. 1999, 2001; Fernández et al. 2001) than others (Pinto and Andrews 2004; Pinto et al. 2005; Rabal-Garcés et al. 2011; Richards et al. 2008), they are less restricted by their dental morphology than deer to a herbivorous diet. Less work has been done on the Middle Pleistocene cave bear (U. deningeri), but we have shown that it too probably chewed bones and may have been an habitual scavenger (Andrews and Fernández-Jalvo 1997).

Some fossils from Units II, III and Vu are extremely rounded (Fig. 10.9). Fossil size, shape (Fig. 10.4), orientation, skeletal elements (Voorhies groups, Table 10.4) and bone density (Table 10.3) all exclude transport of bones into