The earliest cut marks of Europe: a discussion on hominin subsistence patterns in the Orce sites (Baza basin, SE Spain)

Ancient evidence of human presence in Europe is recorded in several Early Pleistocene archaeopalaeontological sites from Spain, France and Italy. This is the case of Barranco León (BL) and Fuente Nueva-3 (FN-3), two localities placed near the town of Orce (depression of Baza and Guadix, SE Spain) and dated to ~1.4 Ma. At these sites, huge assemblages of Oldowan tools and evidence of defleshing, butchering and marrow processing of large mammal bones have been recovered together with a deciduous tooth of Homo sp. in the case of level BL-D. In this study, we: (i) describe in detail the anthropic marks found in the bone assemblages from these sites; (ii) analyse patterns of defleshment, butchery and marrow processing, based on the modifications identified in the cortical surface of the fossils; and (iii) discuss on the subsistence strategies of the first hominins that inhabited the European subcontinent during Early Pleistocene times.


Supplementary Discussion S1
Although the diet of early Homo probably included a broad spectrum of food resources, "meat made us humans" is a recurrent topic in any debate on the evolution of the human genus (Stanford, 1999;Bunn, 2007;Hardy 2010;Bunn et al., 2017;Hardy et al. 2017;Prado-Nóvoa et al., 2017). This debate is largely based on the evolutionary trend documented in the human fossil record to increasing encephalization, which runs in parallel to a trend to decreasing the size of the post-canine teeth (Jiménez-Arenas et al., 2012. The latter trend reflects the progressive adaptation of hominins to a higher-quality, easier-to-digest diet, which resulted in a shortening of the gastrointestinal tract. This forced the members of Homo to a greater consumption of meat compared with the australopithecines (Aiello and Wheeler, 1995;Bunn, 2001;Dunsworth and Walker, 2002).
Humans differ from other anthropoids in being the most genuine omnivorous species of the order Primates. As such, we show a number of anatomical and physiological adaptations to a more carnivorous diet than that of the great apes, our closest living relatives, and these adaptations probably arose in Africa at the origin of the genus Homo ~2.6 Myrs ago. For example, the gastrointestinal tract of H. sapiens is 40% shorter than the one expected for a primate of our size (Chivers and Hladik, 1980;Henneberg et al., 1998). As a result, the ratio of intestinal length to body length of humans (5:1) is lower than the one of baboons (8:1), not to speak on those of truly herbivorous mammals like the horse (12:1) and cattle (20:1), but is included within the range of values depicted by the extant carnivores (6-4:1). Similarly, our ratio of gastrointestinal surface area to body surface area (0.8:1) is close to the value of carnivores (0.6:1) and lower than those of baboon (1.1:1), horse (2.2:1) and cattle  (Chivers and Hladik, 1980;Milton, 2003). In fact, the small intestine and colon represent in humans 67% and 17% of the total volume of the gastrointestinal tract, respectively. In contrast, this proportion is reversed in the great apes, ranging between 14-28% and 52-54%, respectively (Mann, 2000). The consequence of having a shorter colon than those of other primates is that humans have a shorter food-passage time, which decreases the absorption of fibrerich vegetal foods (Milton and Demment, 1988). Moreover, the enterocytes of the human digestive tract absorb with preference the iron linked to the haemoglobin and the porphyrin compounds, while the assimilation of ionic iron from plants is reduced by 50-70% due to the presence of phytates and phenolic compounds (Cook, 1990;Pereira and Vicente, 2013). For this reason, a vegan diet does not satisfy the human daily demands of iron (1.5 mg per day) and must be supplemented.
Dietary quality relates inversely in primates with body mass and with metabolic rate per unit mass, which allows the great apes to subsist with a diet in which vegetal matter represents between 87% and 99% (Milton, 2003). However, the human diet is more digestible and more energy-rich (in kJ per day and kg of body mass) than the one expected from our metabolic rate, which is the typical for a primate of our body mass (Leonard and Robertson, 1996;Henneberg et al., 1998). This relates to the high maintenance costs of the nervous tissue, which sums up to 20-25% of the basal metabolic rate in humans compared with 8-10% in chimpanzees (Leonard and Robertson, 1994).
According to the "expensive-tissue hypothesis" (Aiello and Wheeler, 1995), brain expansion competed with other tissues that have high maintenance costs, such as the heart, kidneys and splanchnic organs (liver and gastrointestinal tract). In this way, the trend to increasing encephalization in the genus Homo required to allocate to the nervous system a greater fraction of the energy budget invested in body maintenance. This required to cut the metabolic expenditure of other tissues with high maintenance costs. The size of the human heart could not be reduced, because a biped stance requires more heart pressure than a quadrupedal one to supply the brain with blood. Similarly, any reduction in either kidney size or energetic expenditure would have reduced the maximum concentration of urine excreted. A more dilute urine would in turn have been problematic for early Homo in the open habitats, where drinking opportunities are scarce and thermoregulatory requirements place considerable demands on the water budget (Wheeler, 1991). In the case of the liver, the size of this organ is constrained by the particular energy requirements of the brain, which uses glucose exclusively as its fuel. For these reasons, given that the gut is the only metabolically expensive organ that could be reduced in size, the expansion of the human brain forced a shortening of the gastrointestinal tract compared to the dimensions expected for the digestive system in a great ape of our size, and this caused a shift to a more carnivorous diet compared to the australopithecines (Aiello and Wheeler, 1995). Moreover, a number of taeniid tapeworms that in their adult stage are characteristic parasites in carnivorous mammals use Homo sapiens as a definitive host. These parasites belong to the Taenia solium and T. saginata-T. asiatica subclades, and probably represent the result of two independent host shifts to hominins in sub-Saharan Africa. This suggests that an omnivorous diet, dependent on scavenging bovid prey taken by large felids and hyaenids, provided the ecological context for the evolution of Taenia specialized in early Homo as a definitive host (Hoberg et al., 2001). relates to the classical debate on Homo as a "hunter" or as a "scavenger" (for review and references, see Blumenshine, 1995;Blumenshine et al., 2007;Domínguez-Rodrigo et al., 2007Espigares et al., 2013;Martínez-Navarro et al., 2014;Rodríguez-Gómez et al., 2016;Martínez-Navarro, 2018).

Meat consumption by early
Since the early eighties, two competing hypotheses on carcass acquisition by hominins have been subject to debate. A number of authors (e.g., Binford, 1981Binford, , 1985Blumenschine, 1986Blumenschine, , 1987Blumenschine, , 1991Blumenschine, , 1995Blumenschine et al., 1994;Capaldo, 1997;Selvaggio, 1998;Arribas and Palmqvist, 1999) considered that early Homo with Oldowan tools had secondary access to ungulate meat and marrow through passive scavenging of partially defleshed carcasses abandoned by their primary predators. In contrast, other researchers (e.g., Bunn and Ezzo, 1993;Domínguez-Rodrigo, 1999;Domínguez-Rodrigo and Barba, 2006;Domínguez-Rodrigo et al., 2007;Bunn and Pickering, 2010) have suggested that the hominins had primary access to fully fleshed carcasses. The reasoning of the latter is that although the extant large felids do not consume bone marrow contents, which are then available for scavengers like the hyenas, they exploit intensively the body of their prey and consequently, the flesh in the abandoned carcasses uses to be very scarce. In their view, this contradicts the interpretation of the finding of abundant cut marks related to defleshing activities in the Early Pleistocene sites as an evidence of secondary access by the hominins to these carcasses. Instead, the cut marks would reflect a primary access to the ungulate carcasses by the hominins, with carcass acquisition resulting from hunting more likely than from power, confrontational scavenging or active searching (Domínguez-Rodrigo et al., 2007. However, it is worth noting that this interpretive context is based on experimental studies performed on modern leopards and lions, which do not apply to the Early Pleistocene hominin populations with Oldowan tools that lived in Europe or in eastern and southern Africa. In the Early Pleistocene, the predator guild was dominated in the Old World by two species of sabre-tooth cats, Megantereon whitei and Homotherium latidens (Felidae, Machairodontinae), as documented in the Orce sites (Martínez-Navarro and Palmqvist, 1995Palmqvist, , 1996Arribas and Palmqvist, 1999;Palmqvist et al., 2007Palmqvist et al., , 2011. The craniodental and postcranial anatomy of these felids, which lack of modern analogues, suggests that they were able to hunt larger ungulate prey relative to their body size compared with the living felines. For this reason, the sabre-toothed cats probably exerted a higher predation pressure on the juveniles of megafauna and exploited their prey to a lesser extent than the extant pantherine cats. This would have resulted in greater amounts of flesh abandoned in the carcasses Palmqvist, 1995, 1996;Arribas and Palmqvist, 1999;Palmqvist et al., 2003Palmqvist et al., , 2007Palmqvist et al., , 2008aPalmqvist et al., , 2008bPalmqvist et al., , 2011Ripple and Van Valkenburgh, 2010;Van Valkenburgh et al., 2016;Martín-Serra et al., 2017). Such resources would in turn be available for passive scavengers such as the giant, short-faced hyena Pachycrocuta brevirostris, which behaved at Orce as a strict scavenger (Palmqvist and Arribas, 2001;Palmqvist et al., 2011), and even for the hominins, as documented in FN-3 (Espigares et al., 2013).  Palmqvist, 1995, 1996;Arribas and Palmqvist, 1999;Palmqvist et al., 2007). This African predator had relatively enlarged sabres and showed an extreme reduction of the premolar teeth, as well as some anatomical differences in the skull related to the relative efficiency of the biting muscles at the level of the lower carnassial. Such craniodental features suggest that M. whitei was able to kill efficiently large ungulate prey relative to its own size. Given its dentition, extremely specialized for meat slicing, this predator only consumed soft tissues. As a result, it left variable amounts of flesh on the carcasses of the ungulates hunted and all within bone nutrients intact, which could be subsequently scavenged by the shortfaced hyenas and the hominins. Moreover, the persistence of sabre-tooths in Europe helps to explain the success of the Oldowan tools in this continent, where the Oldowan/Acheulean transition took place one million years later than in Africa (Palmqvist et al., 2005). The reason is that the sharp flakes characteristic of Oldowan assemblages were fully appropriate for scavenging the carcasses partially defleshed by the sabre-tooths and the cores would have been useful for breaking the bones for accessing their marrow content (Plummer, 2004). In fact, the marked seasonality that characterized temperate Europe for most of the Pleistocene, with cooler and drier conditions than those of subtropical Africa, made the availability of scavengeable ungulate carcasses a key resource for the hominins and may have assisted the initial dispersal of Homo from Africa (Turner and Antón, 1998;Arribas and Palmqvist, 1999;).
In contrast, the new pantherine cats that arrived in Europe during the mid-Pleistocene revolution exploited their kills more fully than the sabre-tooths, which resulted in the loss of a regular source of scavengeable carcasses for the bone-cracking hyenas and the hominins. Under these new ecological circumstances, the trophic niche that they exploited vanished. This forced the hominins toward behavioural and technological improvement, which resulted in the development of the Acheulean tools of Homo heidelbergensis (Arribas and Palmqvist, 1999). The destiny of the giant hyena P.
brevirostris, constrained by its enormous size and highly specialized anatomy, was extinction (Palmqvist et al., 2007(Palmqvist et al., , 2011. In the case of the hominins that inhabited the depression of Baza and Guadix, although the scavenging of ungulate carcasses partially defleshed and abandoned by sabre-tooth cats would have enhanced their survival during Early Pleistocene times Palmqvist, 1995, 1996;Arribas and Palmqvist, 1999;Martínez-Navarro, 2004Palmqvist et al., 2005Palmqvist et al., , 2007Palmqvist et al., , 2011Espigares et al., 2013;Rodríguez-Gómez et al., 2016, 2017, this does not preclude other procurement strategies (e.g., opportunistic hunting of small-sized mammals, kleptoparasitism, or even scavenging of carcasses of very large animals that died from other causes). The latter interpretation is supported by evidence of competition between Homo sp. and P.
brevirostris for the exploitation of the carcass of an old individual of Mammuthus meridionalis in the Upper Level of FN-3. The skeleton of this elephant is dismembered and surrounded by flint flakes and coprolites, which suggests a sequential pattern of carcass consumption by hominins and hyenas (Espigares et al., 2013).
In addition, the use of a mathematical model that allows evaluating the sustainability of the community of secondary consumers, based on the biomass of primary consumers potentially available, provides relevant data on this hominin population (Rodríguez-Gómez et al., 2016). Three scenarios were modelled for estimating the population density of Homo sp. in the depression of Baza and Guadix: "hunter", "scavenger" and "mixed behaviour". The results obtained suggested that a scavenging behaviour would have been the optimal scenario for this population, as it would allow to hold 0.10-0.14 individuals per km 2 during a year, values that are close to the densities of modern hunter-gatherers (range: 0.03-0.24 individuals per km 2 ). In As noted in the main text, the faunal assemblages of BL and FN-3 record no evidence of consumption of other animal resources (e.g., rodents, rabbits, birds or herpetofauna) except the two cut-marked plastron fragments from BL-D. In spite of this, the practice of a strategy based on small game is documented in many Middle and Late Pleistocene sites, although it is less frequent during the Early Pleistocene (Speth and Tchernov, 2002;Steadman et al., 2002;Blasco, 2008;Braun et al., 2010;Huguet et al., 2013;Rufá et al., 2014). If we focus on the turtles, evidence of consumption of these animals is also frequently preserved in a number of Middle and Late Pleistocene sites from Israel, Italy, Spain, or South Africa (Blasco, 2008;Blasco et al., 2013), and has been described in modern hunter-gatherer populations (Klemens and Thorbjarnason, 1995). However, apart from the two plastron fragments from Orce, there is only evidence of the exploitation of these resources during Early Pleistocene times in East Turkana (Braun et al., 2010;Archer et al., 2014) and Sima del Elefante (Blasco et al., 2011). Similarly, consumption of fish and crabs has been documented in the early Middle Pleistocene site of Gesher Benot Ya´aqov, dated at ~750 ka (Alperson-Afil et al., 2009).
Additional information on the paleodiet of ancient human populations can be retrieved from microwear analysis. In this sense, the microwear texture complexity of the hominin teeth recovered in Olduvai suggests a broader range of foods available for Homo ergaster than for Australophitecus afarensis, A. africanus, H. habilis or Paranthropus boisei (Ungar et al., 2012). The teeth of the Dmanisi hominins, which are associated with stone artefacts of Oldowan tradition that are similar to those of early Homo in Africa, preserve a microwear pattern that falls within the range of variation for African H. ergaster. This suggests a similarity in their foraging behaviour (Pontzer et al., 2011). Given that the record of Homo sp. at Orce is limited to the deciduous molar tooth from level BL-D, no microwear analysis has been performed in the sites studied here. Although the taxonomic determination of this tooth at the species level is not possible (Toro-Moyano et al., 2013), an early Homo seems an appropriate candidate to inhabit the depression of Baza and Guadix in the Early Pleistocene, given the chronological proximity with the Dmanisi hominins and the presence of Oldowan lithic tools. In this sense, and carefully, the inferences derived from microwear studies performed on the Dmanisi teeth could be also extrapolated to the Orce hominins.