Introduction

The process of human dispersal from Africa started before 2 Ma (Coppens 2017; Han et al. 2017a; Malassé Dambricourt et al. 2016). The oldest evidence of the Eurasian peopling seems to suggest a scenario in which the Asian continent was occupied before Europe. The oldest human evidence comes from China and dates from more than 2 Ma (Boëda and Hou 2011; Han et al. 2017b). Around 1.9–1.6 Ma, human presence is attested to in most of Southern Asia (Dennell 2008a, b; Zhu et al. 2008), in Southeast Asia (Falguères et al. 2016; Swisher et al. 1994), in Northern China (Dennell 2015), and in Georgia (Baena et al. 2010; Gabunia 2000; Garcia et al. 2010; Mgeladze et al. 2010; Vekua 2002).

In Southern Europe, several sites display the arrival of Homo before and around 1 Ma (Álvarez et al. 2015; Arzarello et al. 2007, 2015; Barsky et al. 2010; Bermúdez de Castro et al. 2010; Carbonell et al. 2008; de Lombera-Hermida et al. 2016; De Lumley et al. 1988; Despriée et al. 2006, 2010; Huguet et al. 2017; Lopez-Garcìa et al. 2015; Moyano et al. 2011; Muttoni et al. 2011; Parés et al. 2006), but very few human remains have been discovered and actually attributed to Homo sp. (Bermúdez de Castro et al. 2010; Toro-Moyano et al. 2013).

Among these European archeological sites, Pirro Nord 13 (PN13) represents evidence of the oldest human presence in Europe (Lopez-Garcìa et al. 2015). The site, located inside a karst fissure, holds faunal remains in association with lithic artifacts. The nature and characteristics of the fissure filling allow for the statement that the faunal and archeological remains are in a secondary position and that the sediment, within the bones and lithics, has been transported inside the fissure, mainly by water and gravitational processes (Giusti and Arzarello 2016; Karampatsou 2017). Despite site formation processes, the homogeneity of the deposits and the contemporaneity of the artifacts and faunal remains have been clarified by several works (Bagnus 2011; Giusti and Arzarello 2016; Karampatsou 2017), but the mutual relations between the faunal remains and human activity have never been proved before. The aim of this work is to define the relationship between the faunal remains and the lithic industry of Pirro Nord 13 through zooarchaeological study of the faunal remains and use-wear analysis of the lithic industry.

Historical and background of research

The fossiliferous area of Pirro Nord (also known as Cava Pirro or Cava Dell’Erba) is located at the northwestern margin of the Gargano promontory, close to the village of Apricena (FG, Apulia, Italy; 41°4800700 N, 15°2300500 E) (Fig. 1). The paleontological findings come from several karst fissures situated at the top of a Mesozoic limestone formation that has been the object of exploitation by numerous limestone quarries (Fig. 1). During the Pleistocene, these fissures were part of a very complex interconnected karst system (Pavia et al. 2012). Each fissure containing Villafranchian paleontological remains has been named “P” or “PN,” followed by a progressive number. Paleontological studies have been conducted there since the 1970s (Freudenthal 1971), and systematic field investigations have been carried out by several research teams (De Giuli et al. 1986). In the PN13 fissure, some lithic industry was found, together with vertebrate fossils related to the Pirro Nord Faunal Unit (Arzarello et al. 2009, 2012, 2015; Giusti and Arzarello 2016).

Fig. 1
figure 1

Site location: left—position of Pirro Nord fossiliferous area; right—position of the fissure of PN13 inside the quarry (aerial view)

Full size image

Geomorphological and stratigraphic settings

The deposit that preserves the paleontological and archeological remains of PN13 fills a vertical karst fissure that was exposed by the mining activities of the limestone quarry. The fracture opens at the stratigraphic boundary between the Mesozoic limestone and the Pleistocene calcarenite formation, in the wilderness zone of a more complex karst system located in the Dell’Erba quarrying complex (Pavia et al. 2012). The deposit is more than 4 m thick and is characterized by a sandy loam matrix with clasts and blocks of local lithology, showing a chaotic arrangement and distribution. On lithological and textural bases, four sedimentary units (SUs) have been distinguished (Fig. 2). Unit A consists of a thin sandy-clay layer that covers the underlying units almost for their entire lateral extension. Its formation was due to vertical percolation of water that led to the deposition of secondary clay at the top of the sequence; the same process filled the empty interstices between the large blocks that characterize the underlying SUs, determining the deposition of clay and fine sand layers. Units B to D are composed of sandy loam sediment among limestone and calcarenite cobbles and boulders, which show poor sorting and a varying degree of sizes and roundness. Units C and D show an abrupt increase in the number and dimension of clasts and large blocks and a significant increase in the number of fossils and lithic artifacts as well (Giusti and Arzarello 2016). The formation of the deposit was due to a first collapse of the top of the fissure, which resulted in communication of the karst system with the ground surface, which determined the fall of the large blocks of calcarenite and limestone. An applied set of spatial analyses (Giusti and Arzarello 2016), along with recent geochemical analyses and geoarchaeological revision of the site (Karampatsou 2017), suggested that the formation of SUs B, C, and D was the result of a gravitational selection of the chaotic materials carried into the fissure during one or more subsequent events of mud flow or earth flow, triggered by humid and arid weathering.

Fig. 2
figure 2

PN13 stratigraphy. a PN13. Stratigraphic layout across squares 1 and 6 (*): labels indicate the stratigraphic units (SUs), and lines mark their bases. b Details of SU B: chaotic structure of many sub-rounded, less than decimetric clasts in sandy muddy matrix. c Details of SU C: chaotic structure of varying sizes of decimetric clasts and boulders in brown-green muddy sand matrix—pin indicates a Mn-rich lens. d Details of SU D: same as c but with larger boulders and common sandy lenses—see the pin—sometimes structured in thin layers with alterations of sand and mud

Full size image

Considering the stratigraphy and the taphonomical processes, models that describe the formation of the site have been proposed by Bagnus (2011), Berruti (2017), and Giusti and Arzarello (2016). These models assume that animals died closed to the karst sinkhole, and the action of heavy rains transported the faunal remains and lithic artifacts into the fixture, in which they were rapidly buried with mass-wasting processes, such as mud flow or earth flow, also carrying rocks. No post-depositional reworking or mixture of the sedimentary deposit has been observed.

Considering this geomorphological and stratigraphical setting and the re-deposition processes that affected the deposit, the faunal remains and the lithic artifacts found in the PN13 fissure have been studied as a unique assemblage (Arzarello et al. 2009, 2012, 2015; Bagnus 2011; Berruti 2017; Giusti and Arzarello 2016; Lopez-Garcìa et al. 2015).

Faunal remains and biochronology

The PN13 large mammal assemblage is typical of the Pirro Nord Faunal Unit (FU) (De Giuli et al. 1986; Gliozzi et al. 1997; Masini and Sala 2007, 2011). This FU is the last of the Late Villafranchian, and it denotes a first faunal renewal related to the beginning of arid climatic conditions that prevailed during the Galerian. It registers the first occurrence (FAD) of Bison (Eobison) degiulii, Canis (Xenocyon) lycaonoides, Canis mosbachensis, and Equus altidens (Gliozzi et al. 1997; Masini and Sala 2007; Petronio and Marcolini 2013).

The Pirro Nord local fauna is characterized by numerous taxa of reptiles, amphibians, birds, and mammals (Alberdi and Palombo 2013; Angelone 2013; Bedetti 2003; Bedetti and Pavia 2013; Blain et al. 2016; Delfino and Atzori 2013; Delfino and Bailon 2000; Marcolini et al. 2013; Petrucci et al. 2013; Rook and Sardella 2013; Salari et al. 2013). Among large mammals, PN13 is characterized by an assemblage dominated by Pseudodama, cf. P. farnetensis, with, among ungulates, Bison (Eobison) degiulii, Equus altidens, a large size cervid (Praemegaceros?), and Stephanorhinus, cf. S. hunsheimensis. At this site is also present Mammuthus meridionalis and, among carnivores, Canis mosbachensis, Canis, cf. C. (Xenocyon) lycaonoides, Vulpes, cf. V. alopecoides, Pachycrocuta brevirostris, Ursus etruscus, Homotherium sp., Lynx issiodorensis, Felis, cf. F. lunensis, Meles sp., and Martes sp. (Colopi 2016). Among small mammals, the Pirro Nord local fauna is related to the Early Biharian mammal age and the “Mimomys pusillus-Mimomys savini zone” (Gliozzi et al. 1997; Lopez-Garcìa et al. 2015; Masini and Sala 2007). According to the other Pirro Nord localities, the only arvicolid species present at PN13 is Allophaiomys ruffoi. Although no direct dates are available yet, this species has been related to a chronological range between 1.3 and 1.6 Ma, allowing for the hypothesis that Pirro Nord is older than other Western European sites where an ancient human presence has been attested (Lopez-Garcìa et al. 2015).

Lithic assemblage

From 2006 to 2015, 340 lithic artifacts were found in the PN13 fissure related to anthropic activities (231 flakes, 37 cores, and 72 debris) (Table 1). The raw material was collected no more than 7 km from PN13, in secondary positions (riverbeds or slope deposits). The exploited flint pebbles and cobbles have different morphologies and sizes, and they come from the Gargano Cretaceous succession (Arzarello et al. 2009, 2015, 2016; Giusti and Arzarello 2016). In the PN13 lithic assemblage, two main reduction sequences were adopted: an “opportunistic” débitage based on the exploitation of multiple striking platforms (max 5) to produce flakes with different morphologies but always with at least one cutting edge; and a “centripetal” débitage to produce flakes with convergent cutting edges. The “opportunistic” reduction sequence has been adopted for the reduction of largest flint cobbles. This method is an adaptation of the raw material shape to obtain the greatest quantity of blanks with a sharp cutting edge and with the least number of gestures (Arzarello et al. 2015). The “centripetal” reduction sequence could be compared more simply to some S.S.D.A. cores described by Forestier (Forestier 1993)—a “centripetal” debitage for, in most cases, the production of flakes with convergent cutting edges (Arzarello et al. 2015). There is not a clear relationship between the type of flint and the method of débitage adopted for the reduction of the cores, but a relationship between the raw material morphology/size and the knapping method has been underlined (Arzarello et al. 2009, 2015; Giusti and Arzarello 2016). The technique used is always direct percussion by a hard hammer. The retouched instruments are very rare, and only four flakes, issued by an opportunistic method, are characterized by an intentional modification of the cutting edges. All of them were broken, and two show a bending fracture (Arzarello et al. 2009, 2015; Giusti and Arzarello 2016). The macroscopic analysis attests that a good state of preservation characterizes the whole series of artifacts. At this level of analysis, the artifacts have sharp edges and do not show macroscopic evidence of transportation: 35% carry traces of superficial Fe-Mn oxide patinas, as observed on more than 70% of the faunal remains. The only post-depositional alteration identified consists of fractures that affect 20% of the lithic material (Arzarello et al. 2015; Giusti and Arzarello 2016).

Table 1 Composition of the lithic assemblage of Pirro Nord
Full size table

Materials and methods

Zooarchaeological study

The material is stored and available for the study at the Dipartimento di Studi Umanistici, Sezione di Scienze preistoriche ed antropologiche of the University of Ferrara under the permission of the Soprintendenza Archeologia, Belle Arti e Paesaggio per le province di Barletta-Andria-Trani e Foggia. Previous works have shown that the PN13 faunal remains are characterized by frequent post-depositional alterations, such as cracks, hydric abrasion (rounding/polishing), carnivores’ modifications, ferromanganese oxide deposition, and chemical alterations (Arzarello et al. 2012, 2015, 2016; Bagnus 2011; Giusti and Arzarello 2016).

The zooarchaeological study was conducted on 1285 remains from the 2012–2015 excavation campaigns; the taxonomic attribution of this sample refers to Colopi (Colopi 2016), except for the remains belonging to the Testudines order, which are under study.

The remains were grouped into five size categories according to the system of classification developed by Bunn (1986) for the Olduvai faunal assemblages: very large size (> 1000 kg), large size (300–1000 kg), medium size (100–300 kg), small size (20–100 kg), and very small size (< 20 kg) (Bunn et al. 1980; Bunn 1986). When possible, the age of death of the animals (age AT death) was defined. This analysis was based on the stages of eruption of the teeth, the dental wear (Ervynck 1997; Grant 1982; Stiner 1998) and the epiphyseal fusion stages (Brown and Chapman 1991; Eisenmann 1979). The following indices were calculated: NISP (number of identified specimens), MNE (minimum number of skeletal elements), MNI (minimum number of individuals), and %MAU (minimum number of animal units) (Binford 1981; Lyman 1994).

The study of the sample was conducted to detect all taphonomic modifications, paying special attention to hominin and carnivore modifications (cut marks, bone breakage, and tooth marks); these elements were analyzed according to the works of different scholars (Andrews 1995; Behrensmeyer 1978; Cáceres 2002; Fernández-jalvo and Andrews 2003; Fernández-Jalvo and Andrews 2016; López-González et al. 2006; Marín-Arroyo et al. 2014). Butchery traces, with striations of variable length and width, were analyzed using the criteria described by Binford (1981), Shipman and Rose (1983), Nilssen (2000), Domínguez-Rodrigo et al. (2009), Domínguez-Rodrigo et al. (2017), and Soulier and Costamagno (2017).

A bone breakage study was conducted using the method developed by Villa and Mahieu (1991), which considers the type of fracture, the angle, and the edge texture. To discern the breakage traces linked to anthropic activity, the presence of percussion notches and impact flakes, in accordance with the works of Capaldo and Blumenschine (1994) and Pickering and Egeland (2006), was considered. Carnivore tooth marks (pits and scores) were analyzed in view of the type of the tooth marks, the location of these marks on the bones, the type of the bone tissue affected (shaft or end), and the dimensions (length and breadth) of each tooth mark (Delaney-Rivera et al. 2009; Domínguez-Rodrigo and Barba 2006; Domínguez-Rodrigo and Piqueras 2003; Pobiner 2015; Pobiner et al. 2008; Saladié et al. 2014). The zooarchaeological study was conducted with the aid of two different stereomicroscopes: Optika (SZM-2 EWF × 10/× 20, magnification × 0.7–× 4.5) for all of the material and a Hirox digital microscope (K-H 8700) for the study of butchery traces with striations of variable length and width.

Use-wear analysis

This study began with the preliminary evaluation of the entire artifact sample found in the fissure of PN13 during the excavation campaigns from 2006 to 2015 (340 artifacts). Two criteria were applied to select artifacts for the use-wear analysis: presence of at least one useful edge (selected according to the criteria developed by Terradillos-Bernal and Rodríguez-Álvarez (2017); and surface preservation (absence of marked PDSMs (post-depositional surface modifications) (Levi Sala 1986)). The PN13 evaluated sample is finally composed of 63 flint flakes (15 centripetal and 48 opportunistic flakes). For the analysis, each selected artifact was gently washed with warm water and soap, and then it was washed for 3 min in a mixture of demineralized water (75%) and alcohol (25%) in an ultrasonic tank and was open air dried. The use-wear analysis of the PN13 assemblage was conducted with an integrated approach that uses a low power approach (Odell and Cowan 1986) in combination with a high power approach (Keeley 1980). Several works (Berruti and Daffara 2014; Beyries 1987; Burbidge et al. 2014; Cruz and Berruti 2015; Hardy 2004; Lemorini et al. 2014a, b; Moss 1983; Van Gijn 2014; Visentin et al. 2015; Wilkins et al. 2015) have shown, in fact, that the use of both methodologies integrated is more effective and productive.

The analysis of the macro-traces, or the low power approach, provides information about the potential activities carried out (e.g., cutting, scraping, piercing, etc.) and general interpretation of the hardness of the worked materials (Lemorini et al. 2006; Lemorini et al. 2014a; Odell 1981; Semenov 1964; Tringham et al. 1974). The analysis of the micro-traces, or the high power approach, was conducted to provide a more detailed understanding of the activities carried out with the lithic artifacts and to support the diagnosis of the processed materials (Keeley 1980; Lemorini et al. 2006; Lemorini et al. 2014a; Rots 2010; Van Gijn 2014; Ziggioti 2005). The analysis of the lithic artifacts was conducted using three different types of microscope: a Seben Incognita III stereoscopic microscope with magnification from × 20 to × 80 (for the first scanning of the collection), a Leica EZ4 HD stereoscopic microscope with magnification from × 8 to × 40 (for the low power approach analysis) and an Optika B 600 Met metallographic microscope supplied with 5 PLAN IOS MET objectives (× 5–10–20–50–100) (for the high power approach analysis). Adobe Photoshop CS6 Portable (© Adobe) software was used for the treatment of the images since it allows a single image to be built up from several photos taken at different depths of field. To better identify the traces of use present on the artifacts, a reference collection with flakes made of the same raw material used in Pirro Nord was produced (it was also used for other papers (Buccheri et al. 2016)). Several specific activities were then conducted on different materials (skinning, filleting, woodworking, etc.) with the experimental lithic tools to link the use-wear features to tool motions and to the processed materials. In particular, 10 tools were used to carry out butchering activities on wild boar and rabbit carcasses. During the experimentation, un-retouched flakes obtained using the S.S.D.A. method and the hard percussion technique were used. For each of them, the time of use, the direction of the gesture, and the material worked were recorded. After the experiments, each artifact was gently washed with warm water and soap, and then it was washed for 3 min in a mixture of demineralized water (75%) and alcohol (25%) in an ultrasonic tank and was open air dried. Finally, each object was studied using the same method used for the archeological collection.

Results

Zooarchaeology results

Of the 1285 fossil remains present in the sample, 482 were identified anatomically and taxonomically, likely due to the condition of the bones and, in particular, the high fragmentation rate. The large mammal assemblage is coherent with the associations found in other Pirro Nord localities, and it is not contaminated by faunal remains coming from Quaternary karstic fissures, while in the small mammal assemblage, three remains related to the Terre Rosse endemic assemblages (Messinian-Zanclean) have been detected (Colopi 2016; Lopez-Garcìa et al. 2015).

During the study, the remains were divided into five size categories. The very large size regroups adult Mammuthus meridionalis and Stephanorhinus, cf. S. hundsheimensis. The large size contains adult Equus altidens, Bison (Eobison) degiulii, Ursus etruscus, Homotherium sp., young Stephanorhinus, cf. S. hundsheimensis, and Homotheirum sp. In the medium size group were placed young Equus altidens, Ursus etruscus, and Praemegaceros and adults of Pachycrocuta brevirostris and Pseudodama, cf. P. fartenensis. The small size group includes young of Pachycrocuta brevirostris, Canis mosbachensis, Hystrix refossa, Pseudodama, cf. P. farnetensis, and Lynx issiodirensis, and adult of Canis cf., C. (Xenocyon) lycaonoides. Finally, the very small size group contains young Canis, cf. C. (Xenocyon) lycaonoides, Meles sp., Martes sp., Vulpes, cf. V. alopecoides, Felis, cf. F. lunensis, and Testudines. Despite taxonomic variability, the MNIs (Table 2) shows the predominance of fallow deer with 12 individuals, while the other genera present in the sample do not exceed an NMI of 7.

Table 2 NISP, MNE, and MNI of taxa identified at PN13 according to age categories
Full size table

Related to division for age groups, considering the entire sample (Table 2), the finds attributable to adult specimens are the most represented, with 29 individuals, while the juveniles are 15, the infantile are 9, and the senile are 10. In terms of skeletal representation, appendicular and cranial elements are the most common, while axial parts are relatively rare (Table 3).

Table 3 NISP (MNE) according to the weight classes
Full size table

The PN13 assemblage is composed of different sizes of ungulates, mainly deer, equines, and other sizes of very large animals, such as rhinoceroses and proboscideans. Anatomically, all of the segments were conserved in this set, and the appendicular skeleton is best represented, followed by cranial and axial segments. Deer (Pseudodama farnetensis and Praemegaceros) had more complete skeletal representation. Isolated teeth are best represented, followed by unidentifiable long bones, metapods, vertebrae, phalanges, ribs, tarsals, tibiae, mandibles, skulls, femurs, humeri, radii, scapulae, antlers, pelvises, ulnae, carpals, patellae, sterna, and horns. The majority of MNE are represented by large animals, followed by medium-sized, small-sized, very small-sized, and finally very large-sized animals. According to %MAU (Table 4), a biased representation was observed in all of the taxa in the assemblage. The best-represented elements of larger animals are those belonging to the medial and proximal appendicular segments, followed by the axial and cranial, with relatively high proportions. The well-represented elements in the medium-sized category are those inherent to the medial appendicular skeleton, followed by axial and cranial ones. The small-sized animals are greatly represented by elements appertaining to the axial segment, followed by the cranial and appendicular ones, with small proportions of representation.

Table 4 MNE, MAU, and %MAU for the three main weight classes
Full size table

The PN13 faunal remains assemblage is affected by several taphonomic modifications: manganese oxide patina (97.12%), rounding (57.27%), polishing (33.23%), iron oxides (51.83%), and root-etching marks (5.83%) (Fig. 3).

Fig. 3
figure 3

Taphonomic alterations observed on the analyzed fossils of Pirro Nord 13

Full size image

Despite the post-depositional modifications, evidence of human activities (cut marks and breakage patterns) and that of carnivores (tooth marks) were identified in this assemblage, although hominin faunal exploitation evidence is less abundant than carnivore activity. During the study, 14 cut-marked bones were identified (1.08%) (Table 5). This analysis was based on the diagnostic criteria of Domínguez-Rodrigo (Domínguez-Rodrigo et al. 2009; Domínguez-Rodrigo et al. 2012; Domínguez-Rodrigo and Yravedra 2009; Egeland 2012; Merritt 2012), which were used to precisely characterize these cut marks and to distinguish them from those produced by taphonomic processes (such as trampling and weathering): location, orientation, trajectory of the groove, and the presence or absence of micro-striations inside the groove, the shape of the groove (narrow V or wide V), and the presence of flaking on the shoulder. All of them are slicing marks, with lengths varying between 21.42 and 1.36 mm. The majority are obliquely oriented in relation to the major axis of the bone and only in few cases are transverse. A cluster of obliquely oriented cut marks were found on proximal metaphysis of a left metatarsal, previously recognized as unidentified (Colopi 2016) and now attributed to cf. Equus (12D17B4S66). This bone is not well preserved, and it lacks any secure morphological features that allow for species attribution. Nevertheless, part of the small facet for the mesentocuneiform (Hussain 1975) is present, and on the proximal diaphysis portion, a typical Perissodactyla transverse incision for tendon attachment is visible (Fig. 4). On two shaft metatarsal bones of Pseudodama, cf. P. farnetensis, on one Praemegaceros diaphysis shaft metatarsal and on one lumbar vertebra of Pseudodama were found cut marks obliquely oriented (Fig. 5). Three other cut marks are situated on three different rib diaphyses attesting to transverse and oblique orientation (Fig. 6). Finally, four obliquely oriented incisions are identified on four long bone shafts.

Table 5 Bone remains cut marks, morphology, typology and activity by taxon and weight. MCM = number of cut marks, Def = defleshing, Skin = skinning, Diaph. = diaphysis, Epiph. = epiphysis, Metaph. = metaphysis, Prox. = proximal, Ant = anterior, Lat = lateral, A = adult, JA = young adult, L = large size, M = medium size, S = small size, Indet = indeterminate
Full size table
Fig. 4
figure 4

Cut marks: PN13, PN13-D17B4S66; 1 and 2, several oblique incisions, grouped on a left metatarsal of Equus; 3, carnivore tooth mark (score)

Full size image
Fig. 5
figure 5

Cut marks. a PN13-D22B2S69, linear incision, oblique on the lateral dorsal part of the body of a lumbar vertebra of Pseudodama, cf. P. farnetensis. b PN13-D22B1 (1-1), linear, oblique and deep incision on the diaphysis of a metatarsal of Pseudodama

Full size image
Fig. 6
figure 6

Cut marks. a PN13-RIM 1-2, two incisions, parallel, oblique scutted by the edge of the fracture on the diaphysis of a long bone of a medium-sized animal. b PN13-D19B5S73, two incisions, parallel, obliques on limb mid-shaft of a medium-sized animal. c PN13-D19D1S05, cut mark, oblique linear on the lateral part metatarsal of Praemegaceros

Full size image

During the analysis of the sample evidence, intentional bone breakage documented on eight remains was found (Fig. 7): two fragments of limb bone shafts with percussion notches (of two medium-sized animals); four impact flakes, two of them referable to a medium-sized animal (metatarsal bone and humeri of deer); and another one referable to large-sized animal.

Fig. 7
figure 7

Bone breakage evidence. a PN13-D24C2-34, impact flake (large-sized animal). b PN13-D24C2-16, impact flake (medium-sized animal). c PN13-D21B2S59 limb mid-shaft showing percussion notches (medium-sized animal). d PN13-D22B2S70, proximal epiphysis+diaphysis of deer metatarsal bone demonstrating flake parasitism by percussion. e PN13-D18B3S93, limb mid-shaft large-sized animal with percussion notches

Full size image

On the faunal remains, several carnivore tooth marks (pits and scores) were observed (5.56% of the whole sample) (Table 6 and Figs. 8 and 9). The remains referable at the appendicular skeleton (stylopodial and zeugopodial) are those that appear more affected by the carnivore tooth marks. The score dimensions show a range of length of 7.61 mm and a breadth of 1.1 mm, while those of the pits are 3.7 mm for the length and 3.31 mm for the breadth.

Table 6 Carnivore bone damage recognized in Pirro Nord 13 assemblage
Full size table
Fig. 8
figure 8

Carnivore modifications. a PN13-D19C4S116, score on long bone metaphysis (large-sized animal). b PN13-RIM 1-2-2015, Pit on bone fragment (small-sized animal). c PN13-C11B2S02, pit on the dorsal part of a vertebra

Full size image
Fig. 9
figure 9

Carnivore modifications. a PN13-D20C4S125. b PN13-C10B5S82, cervical vertebrates (medium-sized animal) showing consumption traces. c PN13-D22B1S84, digested fragment of metapodial diaphysis (medium-sized animal). d PN13-D22C6S184, hyena coprolith. e PN13-D23C5S119, femur mid-shaft (medium-sized animal) presenting pitting

Full size image

There is also a cut-marked bone (equid metatarsal) that preserves cut marks associated with one score mark made by a carnivore tooth (Fig. 4).

The results obtained analyzing the epiphyses/diaphysis ratio and the percentage of deterioration show a highly significant negative correlation, indicating intense modification caused by carnivores at PN13, although in the case of medium-sized remains, this percentage is low (Table 6).

Use-wear analysis results

The microscopic analysis of the surfaces of the lithic finds, conducted with the aid of a stereomicroscope, showed that 296 of the 340 lithic artifacts studied present post-depositional alterations (PdAs) or PDSMs (Levi Sala 1986): 259 edges crumbling, 208 Fe-Mn oxide patinas, 190 rounding of edges and ridges, 277 polished surfaces, and 58 white patinas. The different PDSMs registered on the lithic artifacts have different intensities, combinations and distributions (Fig. 10).

Fig. 10
figure 10

PDSM alteration on a PN 14 Beta 2007. a On this surface, it is possible to see the presence of white patina (W), polished patina (P), rounding of the edge (R), and Fe-Mn patina (F). b Post-depositional polish on the edge of the artifact

Full size image

Therefore, only 63 artifacts have both criteria (useful edge and surface preservation) needed to be suitable for use-wear analysis.

Among the 63 artifacts selected for the use-wear analysis, only five showed use-wear traces (Fig. 11). They are medium- and small-sized flakes, measuring between 20 and 43 mm in length, 30 and 15 mm in width, and 6–15 mm thick. Four of them show only one used edge, while the other has two active edges. In total, five different used flakes have been found with six functional edges. Some of the functional edges identified present edge removals not due to intentional retouching but rather micro-flaking produced during use (Odell 1981; Semenov 1964). The main difference between the traces of use and post-depositional edge removals is the combination of the trace attributes; the contact with the worked material produces specific combinations of attributes, which rarely are replicated by post-depositional agents (Lemorini et al. 2014a). In any case, considering the antiquity of the site and the formation process of this archeological deposit, only the edges that presented edge removals in association with polishes (Keeley 1980) were considered used. The use wear traces identified on the edges of the artifacts are sufficiently developed to determine the kinematics of the tool use (motion) and the type of the worked material with a satisfactory degree of reliability. Small, diagonally oriented edge removals (attributable at cutting motions) were recognized on the four functional edges belonging to the flakes with only one functional edge, and all of them also presented polishes linked to butchering activities. The polishes identified affect the edge surfaces gradually, and their extension decreases from the outside to the inner part; furthermore, the width of these polishes on the faces of the edges is limited. The limited polish extension on the surfaces usually indicates that the position of the tool relative to the worked material was characterized by a high angle of work. The longitudinal motion and a high angle of work are typical features of cutting actions, which are also typical of butchering activities. Part of the recorded polishes on the edges appears like a rough band associated with, in some cases, diagonal smears that could be the result of contact with soft animal tissue. Furthermore, on the same edge surfaces, there are areas showing small, smooth, and flat spots of polish that could be linked to contact with bone (Fig. 11). On the last two functional edges identified were recognized sub-perpendicular and diagonally oriented edge removals and rough bands of polishes with, in some cases, sub-perpendicular smears, likely linked to mixed actions on soft animal tissue (Fig. 11). Similar characteristics are found on the margins of the instruments used during the experimentation to perform butchering actions on rabbit and wild boar carcasses (Fig. 12), as well as the data presented in numerous scientific works dealing with the same type of traces (e.g., Lemorini et al. 2006; Lemorini et al. 2016; Iwase 2016; Pawlik 2013; Solodenko et al. 2015; Lemorini et al. 2014b).

Fig. 11
figure 11

Use wear traces on PN13 artifacts: 13ϒ2007 (scale bar 1 cm): (a) micro-use wear interpreted as the result of a mixed motion (× 30); (b, c) details of the edge area (× 100, × 200) with a band of rough polish (contact with fleshy tissues). 19ε2007 (scale bar 1 cm); (a) micro-use wear interpreted as the result of a mixed motion (× 30); (b) details of the edge area (× 100) with a band of rough polish (contact with fleshy tissues?). 140B4C72011 (scale bar 1 cm); (a) micro-use wear interpreted as the result of a longitudinal motion (30x); (b, c) details of the edge area (× 100, × 200) with rough polish (contact with fleshy tissues). S107C6D212012 (scale bar 1 cm); (a) details of the area in the square (× 100) with rough polish (contact with fleshy tissues); (b) details of the area in the square (× 200), with rough polish (contact with fleshy tissues). S11B3A2008 (scale bar 2 cm); (a) micro-use wear interpreted as the result of a mixed motion of butchering (× 50); (b; c) details of the area in the square (× 100, × 200) with small, smooth, and flat spots of polish (contact with bone)

Full size image
Fig. 12
figure 12

Use wear traces on experimental artifacts: Sper. 8 (scale bar 1 cm): (a) this use-wear is the result of transversal motion during a butchering activity on a wild boar carcass (× 30); (b, c) details of the edge area (× 100, × 200) with small, smooth and flat spots of polish due to contact with bone. Similar to the traces found on the artifact S11B3A2008. Sper.5 (scale bar 1 cm): (a) this use wear is the result of longitudinal motion during a butchering activity on a rabbit carcass (× 30); (b, c) details of the edge area with a band of rough polish (× 100, × 200) by contact with fleshy tissues. Similar to the traces found on the archeological artifacts 13ϒ2007, 19ε2007, 140B4C2011, and S107C6D212012

Full size image

Discussion

The available data about the first European peopling are not very numerous, but every little piece of information is fundamental to understanding human behavior. As already mentioned, Europe was peopled for the first time before 1 Ma. Some of the oldest sites can give us data only about technical behavior since the bones are not preserved (Peretto et al. 1998a, b; Despriée et al. 2006), but some others, including Pirro Nord 13, allow us to also have information about hominid-carnivore interactions. Because PN13 is not a primary deposition site, the archeological and paleontological assemblage can reflect different origins (Colopi 2016; Giusti and Arzarello 2016; Lopez-Garcìa et al. 2015). The presence of several types of post-depositional alterations on the bones and lithic surfaces confirms that the remains were transported from their original place into the fissure (Arzarello et al. 2012, 2015; Bagnus 2011; Berruti 2017; Giusti and Arzarello 2016). Therefore, in the fissure were collected faunal remains and lithics in relation to each other and findings that have only a chronological relation to each other. All of the materials contained in the PN13 fissure, lithic artifacts and fauna remains, are related to the same time span that can be placed at the end of the Villafranchian (Colopi 2016; Pavia et al. 2012; Bagnus 2011; Berruti 2017; Giusti & Arzarello 2016). Indeed, the paleontological and geoarchaeological studies confirmed that this period was the only time when the material in the fissure accumulated (Berruti 2017; Colopi 2016; Giusti and Arzarello 2016; Karampatsou 2017; Lopez-Garcìa et al. 2015). The relation between fauna and lithics is demonstrated only for a reduced part of the assemblage (Bagnus 2011; Giusti and Arzarello 2016). For these reasons, the analyses conducted on the PN13 remains are mainly qualitative, while the quantitative aspects of these studies have relative meaning. Regarding the use-wear analysis of the lithic assemblage, the only activity that it was possible to identify on the PN13 lithic artifacts is the exploitation of animal resources to obtain soft animal tissues, e.g., muscle, viscera, brains, and marrow. Indeed, the presence of two different types of polishes on the same edge, one linked to contact with bone and the other linked to contact with fleshy tissues, can be related to meat extraction from the animals’ carcasses. The presence of this type of use-wear traces on lithics tools referable to ancient sites is not new; indeed, use-wear traces referable to butchering activities have been found on artifacts from many of the oldest African sites: Kanjera South (± 2.0 Ma, Kenya (Lemorini et al. 2014a), Aïn Hanech (± 1.78 Ma, Algeria, (Sahnouni et al. 2013), and Koobi Fora at ± 1.5 Ma (Keeley and Toth 1981). The presence of use-wear traces linked to butchering activities on lithic tools has also been attested in many ancient European and Asian sites: Monte Poggiolo (Longo 1994; Peretto et al. 1998a, b), level TD6 of Gran Dolina (Carbonell et al. 1999) and Xiaochangliang (Shen and Chen 2000).

Despite the post-depositional modifications and numerous carnivore tooth marks, evidence of human activities (cut marks and breakage patterns) has been recognized in the considered faunal assemblage. All of the cut marks identified are slicing marks discovered on medium and large-sized animals’ bones, mainly equids and cervids. According to the criteria developed by Binford (1981), Nilssen (2000), and Soulier and Costamagno (2017), all of the identified cut marks can be linked to the realization of two different slaughtering activities: defleshing and skinning. The presence of traces linked to intentional bone breakage (such as percussion notches and impact flakes) might testify to the practice of marrow extraction. Interaction between carnivores and hominins at PN13 is demonstrated by a cut-marked bone (equid metatarsal) that preserves cut marks associated with one score mark made by a carnivore (Egeland 2007). Even if different methods to identify the access type (i.e., primary or secondary access by hominids) have been applied (Blasco and Rosell 2009; Blumenschine and Selvaggio 1998; Capaldo 1995; Domínguez-Rodrigo and Barba 2006; Egeland et al. 2004, 2014; Selvaggio 1998) in the considered faunal assemblage, overlapping of cut marks and teeth marks, as well as evidence of anthropic evisceration on the faunal remains, is not present. However, the combined evidence of butchering activities on metapodial parts, axial skeletons, the absence of cut marks on stylopodial and zeugopodial limb mid-shafts (Domínguez-Rodrigo and Barba 2006; Domı́nguez-Rodrigo 1997; Egeland 2012; Galán and Domínguez-Rodrigo 2013; Merritt 2012; Nilssen 2000; Soulier and Costamagno 2017; Val et al. 2017) and the presence of impact flakes and percussion notches on limb mid-shafts (Blumenschine and Selvaggio 1998; Capaldo and Blumenschine 1994; Domı́nguez-Rodrigo 1997; Gaudzinski 1993, 2004; Pickering and Egeland 2006; Yeshurun et al. 2011), seem to attest to secondary access to the carcasses by the hominids of PN13 (Gaudzinski 1993, 2004; Yeshurun et al. 2011). Nevertheless, we do not completely exclude the primary access because of the presence of cut marks mid-shaft on unidentifiable long bones.

Conclusions

The scarcity of faunal remains with anthropic and carnivore marks and the nature of the PN13 fissure do not permit determining with confidence the type of access to the carcasses, although it has been established that hominins competed with carnivores for animal resources. Given the small number of lithic tools with use-wear traces and bone with cut marks, for this assemblage, it was not possible to correlate the results.

Not forgetting the available data and remembering that the evidence is not numerous in any of the oldest European sites, it seems that human behavior was not the same for all sites. Around 1 Ma, human groups preferred primary, secondary and mixed access to animal carcasses. Abundant cut marks have been found on the large mammal bones of Fuente Nueva-3 and Barranco León, and opportunistic scavenger behaviour has been underscored for the hominin populations (Rodríguez-Gómez et al. 2016). In Bois-de-Riquet, anthropic marks have been identified for two long bone shafts, and probable secondary access to the carcass has been supposed (Bourguignon et al. 2016). In Level TE9c of Sima del Elefante, cut marks and breakages have been found on Equidae, Bovidae, and Cervidae specimens; even in this case, skinning, defleshing, and anthropic breakage are the recognized activities. Unlike what is observed for the other sites, in Sima del Elefante, the hominins had primary access to the prey (Huguet et al. 2017). The Vallonnet cave is another example of primary access to herbivores of middle size, but in the site, secondary access is also documented for larger animals and, in particular, for Bison schoetensacki (Echassoux 2009).

In conclusion, due to the nature of the accumulation, we cannot assume that all specimens represent a single event; moreover, the post-depositional alteration made it difficult to obtain exhaustive data. Of the approximately 1200 fossils analyzed, a small number (not much more than 1%) present cut marks, most likely made during skinning, meat removal and breakage of the bones. The use-wear analysis, conducted on 300 lithic specimens, allowed for identifying only 5 artifacts that preserved traces (important data for a such old site) and were used during removal of soft tissues. These findings indicate a similar trend to the archaeozoological study and generally with the actual knowledge about this topic, i.e., that early members of the genus Homo were meat consumers. However, in the case of Pirro Nord, whether they had primary (hunters) or secondary (scavengers) access to carcasses cannot be determined.

In Europe, around 1 Ma, we have a homogeneous substratum for what concerns lithic production, but we have some peculiarities, for example, the shaped elements, the retouched blanks and the utilization of centripetal debitage methods. It is very difficult to define a detailed picture when we face numerically reduced evidence, but at the current state of knowledge, we can certainly underscore how human behavior during the first European peopling was strongly influenced and conditioned by the type and availability of resources.