Vestiges of the Sedimentary Archive of Late Paleolithic Inhumations from San Teodoro Cave: Insights into ST3 Burial and Site Stratigraphy

Abstract

Studies of prehistoric burials are fundamental for understanding cultural human evolution. Those found in the San Teodoro cave (northeastern Sicily) are significant for the discovery at the turn of the 1930s and 1940s of at least four individuals (ST1–ST4). About 15–16 kyr ago, the bodies of ST1–ST4 were intentionally buried, apparently in a manner original to the context of prehistoric burials, namely by covering them with a continuous layer of red ochre found to connect the graves. Since the earliest excavations, plagued by clandestine digging, there is no material memory of the stratigraphic transition from the burial layer to the subsequent anthropogenic deposit through the red ochre, and nothing certain is known about the orientation of ST3, the presence of grave goods and the ochre cover related to this burial. Moreover, there is no exhaustive knowledge of how much is actually left of the anthropogenic layers described in the old literature. Based on field observations and 3D reconstruction of ST3’s skull position and deposits at the San Teodoro site, we provide insights into anthropological issues, such as the rediscovery of the red ochre vestiges that reasonably covered the ST3 burial, and the burial context of this individual, and shed light on what actually remains of the stratigraphic units described in the 1940s.

1. Introduction

Late Upper Paleolithic burials are indeed rare in Europe [1]. In several cases, burials were accompanied by rituals that were reserved for adults, children and infants regardless of their gender. Burial rituals included the use of materials such as red pigments (ochre) and grave goods generally consisting of ornaments and objects of various nature, such as mollusk shells, animal bones, bovid and cervid antlers, ivory pendants and necklaces of perforated teeth [1,2,3,4,5,6,7,8,9]. Out of the 20 burials reported for the European area [1], 11 occur in the Italian territory, 15 considering the Gravettian time [10,11]. Among the European Late Upper Paleolithic burials, two of the oldest reports pertain to the Addaura cave complex [12,13] and the San Teodoro cave [14,15], respectively, along the northwestern and northeastern coast of Sicily. These sites, examples of late Epigravettian culture dating back to about 15–16 kyr ago [12,16,17,18,19], represent important Late Paleolithic sites in the Mediterranean area. San Teodoro is of particular importance for the discovery of seven human individuals, ST1–ST7 [14,15,20,21,22]. In addition, people that frequented the site of San Teodoro possibly were among the earliest humans who arrived in Sicily, although there are some faint hints, e.g., some finds of bifacial tools and other ancient samples mainly from the Agrigento and Palermo provinces, that populations arrived on the island much before the Late Epigravettian time, presumably during the Lower-Middle Paleolithic [23].

Prehistoric burial rituals are essential to the understanding of human cultural evolution, and wherever possible, any traces of burial practices deserve to be preserved and stratigraphically well circumstantiated, above all in critical and endangered contexts. This is the case of the San Teodoro cave (Figure 1), a site of particular relevance to archaeo-anthropological studies of the Late Epigravettians and their burial practices [24]. In fact, at least four individuals (ST1–ST4) have been found to be the subject of single burials, assumed to have occurred at the same time [14,15] and apparently all linked to an unusual use of red ochre as a mortuary practice. Indeed, the whole burial setting at San Teodoro seems to show unusual characters in terms of orientation of the bodies within the cave, the presumed massive abundance and wide extension of the pigment spread, and the supply and mode of use in funerary ritual of red ochre [14,15,25]. This kind of pigment is known to have been used by several human species (Homo erectus, H. neanderthalensis, H. sapiens), from the Paleolithic to the present day, in several settings, not necessarily related to burial rituals. The significance of its usage during prehistory in anthropological contexts—symbolic, ritualistic, artistic or functional—is still shrouded in mystery and subject to different interpretations and contexts [26,27,28,29,30,31,32,33,34,35,36,37,38].

Unfortunately, during the anthropological excavations conducted at the turn of the 1930s and 1940s at San Teodoro cave, several acts of vandalism by clandestine diggers made it difficult to interpret the anthropological contexts accurately, and, to date, partly due to the excavations conducted subsequently (Figure 1E), the stratigraphic context of the archeo-anthropological sequence appears severely impoverished because of the absence of some of the anthropic levels, some of which remain in strongly cemented facies greatly reduced in extent [17]. In addition, the complete stratigraphic setting showing the transition between the inhumation layer and the later anthropic one, marked by an interposed red ochre layer, has not been observable or not exhaustively interpreted since the excavations in the 1940s. These gaps have given rise to doubts about the use of the red ochre and actual relation of this pigment with any specific burial [2,39], and in some cases [8] the relationship between red ochre and burials is not even mentioned among the most emblematic cases of prehistoric funerary rites.

Here we present a new view on the rediscovery of the transition between inhumation and butchery-cooking deposits, complete with the ochre traces as the last vestige of the red pigment layer used for burial ritual at the San Teodoro site. We will discuss the relations of this pigment with the burial of individual ST3 and, on the basis of old literature, show a 3D reconstruction of the position of ST3’s skull and other related bones, allowing interpretation of the mode of burial and orientation of this individual. In addition, thanks to the combination of old literature and new observations, we will provide insights into the relations between anthropological events and stratigraphic units of the site.

Figure 1. (A) General location of the study site (black asterisk), the San Teodoro cave in NE Sicily. (B) Location of the study site (white asterisk) in relation to the stretch of coastline facing. (C) The Jurassic limestones of Pizzo Castellaro where the cave (white triangle) opens. (D) Entrance of the cave. (E) Plan of the cave showing the location of the anthropological excavations (1942 in light blue; prior to 1942 in yellow) and the paleontological excavations [40,41,42]. Positions of excavations are extrapolated from Garilli et al. [25], modified. Planimetry is from the survey by Iemmolo et al. [43].

Figure 1. (A) General location of the study site (black asterisk), the San Teodoro cave in NE Sicily. (B) Location of the study site (white asterisk) in relation to the stretch of coastline facing. (C) The Jurassic limestones of Pizzo Castellaro where the cave (white triangle) opens. (D) Entrance of the cave. (E) Plan of the cave showing the location of the anthropological excavations (1942 in light blue; prior to 1942 in yellow) and the paleontological excavations [40,41,42]. Positions of excavations are extrapolated from Garilli et al. [25], modified. Planimetry is from the survey by Iemmolo et al. [43].

2. The San Teodoro Site: Background and Critical Issues

2.1. Geology

The San Teodoro cave (Figure 1), identified as 5000SI-ME at the Catasto Nazionale delle Grotte d’Italia (National Registry of Italian Caves) [44], is located in northeastern Sicily, very close to the village of Acquedolci (Messina), about 1.2 km from the coastline, at an elevation of 125 m a.s.l. (referring to the entrance), with a total height difference of 17 m [44]. At the time of the Late Epigravettian attendance, in the early last deglaciation, low stand was about −90 m [45], with the nearest shore about six kilometers distant from the site. The cave consists of a tunnel-like chamber of about 1200 m² that narrows slightly inward. At the entrance, the cave is about 5 m high and 15 m wide; inside, it shows a maximum height of 20 m.

The cave opens in the Jurassic limestone of the Longi-Taormina geological Unit, which is part of the Kabilo-Calabride geo-structural Unit in the Nebrodi Mountains, an area affected by a complex tectonic setting [46].

2.2. Site Stratigraphy, Human Remains and Attendance

The first archeo-anthropological excavation at San Teodoro cave dates back to 1937, when Mr. Bonafede, commissioned by the Institute of Geology of the University of Palermo, extracted the first of the seven human individuals (ST1–ST7) discovered at the turn of the 1930s and 1940s [14,15,20,21]. All the inhumated bodies belonged to the Late Epigravettian culture [47]. While ST1–ST5 are reported in the stratigraphic and planimetric surveys provided by Maviglia [14] and Graziosi [15] (Figure 2 and Figure 3), nothing is known about the location of ST6 [20] and ST7 [21], discovered by Maviglia in 1940 [22]. To date, the largest source of information on the stratigraphy and anthropology (with special regards to the burial practices) of the Late Paleolithic deposits of San Teodoro is provided by the works of Maviglia [14] and Graziosi [15]. According to their indications (Figure 2), the ST1–ST4 bodies were buried in the uppermost part (layer E) of the oldest deposit in the cave, those containing, among other vertebrates, bone remains of elephant (in the lower part of layer F) and hyena, both of which were supposedly extinct from Sicily before the first arrival of Homo sapiens on this island. No industry was found in burial layer E, but only very localized fireplace remains, probably predating burials. This means that at an early phase of occupation of the site for human attendance was almost exclusively related to inhumations, and only sporadic frequentations characterized the period before inhumations [15]. The younger part of the cave deposit (anthropogenic layers D–A of Graziosi [15], see Figure 2) possibly recorded a change in the use of the site by the Epigravettians, showing intense human frequentation that resulted in the production of large quantities of cooking and butchering remains, especially with regard to layers D–C [15,17]. Graziosi described layer D as a reddish deposit rich in charcoal, vertebrate bone remains and lithic industries, and distinguished the upper layer C as a large “unico” (single) fireplace with very abundant faunal remains and industries, indicating the period of major human attendance of the site. All the anthropogenic layers (Figure 2), especially layers C and D, tapered toward the southeast to the contact with the large boulder that occupies most of the Bonfiglio 1998 excavation (trench α) [15]. Layers B and A showed a progressive decrease of faunal remains, lithic industries and charcoal [15]. According to Garilli et al. [17] and Forgia et al. [19], except for a patch of heavily cemented deposit exposed on the east wall of the cave, a few meters from the entrance, to date only layer D would remain of the anthropogenic deposit described by Maviglia [14] and Graziosi [15].

An isolated, incomplete skull was found in 1942 together with fragmentary humeri and a few fragmentary vertebrae (individual ST5) in layer B, thus suggesting a younger age than ST1–ST4 based on their respective stratigraphic positions (Figure 2).

In the area affected by the burials, the transition between layers F–E and layers D–A was marked by a bed of red ochre, with an average thickness of 5 cm, which was first observed in the 1938 excavation. To date, since a very small portion rediscovered and analyzed by Garilli et al. [25] was collected to trace its origin, no trace of this ochre has been preserved in situ. The very scarce sample of ochre studied by Garilli et al. [25] was collected in its entirety from a critical position, being almost entirely outcropping on the present floor level, on the edge of the north wall of Bonfiglio’s excavation performed in 1989 (trench α), presumably from the base of layer D [25]. A bacteriological origin has been traced from this small amount of red pigment, providing the earliest documentation of collection of the raw material from deposits formed by iron-oxidizing bacteria, most likely from some springs in the vicinity of the site [25]. A similar way of procuring is known elsewhere only for the Neolithic North American site of Babine Lake in British Columbia [48].

The DNA-based [16,49] and morphological [20,21,22,50,51] analyses available in the literature for the Epigravettians discovered at San Teodoro indicate the prevalence of females (ST1, ST2, ST4, ST5 and ST6) over males (ST3 and ST7). There is, however, a considerable margin of uncertainty for ST2 [18], ST4 [2,15] and especially for ST1 [15,52], who, on anatomical grounds, have been attributed to the male sex [15]. Morphometric analysis conducted by D’Amore et al. [53] on the craniofacial remains of individuals ST1–ST3, ST5 and ST7 suggested affinities with coeval populations from central and southern peninsular Italy. A detailed morphometric dataset performed on all ST1–ST5 bone remains was reported by Graziosi [15], who estimated a height of just over 1.60 m for ST1 and ST4. Dental analysis of ST1, ST2, ST6 and ST7 suggested a frequent non-alimentary use of dentition [22]. An age of 25–35 [15,22] and 40–50 [15] years was assumed for ST1 and ST2, respectively. An advanced age was assigned to ST5 [15]; a full adult age to ST3 and ST7 [22].

All the San Teodoro individuals are housed in Italian institutions: ST1 and ST2 at the Museo G.G. Gemmellaro, Palermo; ST3, ST5 and ST6 at the Museo e Istituto Fiorentino di Preistoria, Florence; ST7 at the Museo Civico di Storia Naturale, Milan.

2.3. Inhumations

Maviglia [14] and Graziosi [15] described for San Teodoro an atypical use of ochre in the mortuary practice compared with other known Paleolithic to Neolithic sites, where the red pigment was found positioned under the body or smeared directly on the body or some of parts of the deceased, on grave goods or simply placed nearby [3,7,27,30,39,50,54,55,56,57,58]. In fact, Maviglia [14] reported an extensive and continuous layer, averaging 5 cm thick, of soil “heavily impregnated with red ochre” that extended 4–5 m and covered the topsoil of the burials of individuals ST1–ST4, with the ST4 grave partially covered. Graziosi [15] reported an extension of about 8 m over the same individuals (Figure 2 and Figure 3). The red layer, which occurred about 20 cm over most of the bodies, or a little more [14], would have been spread as a funerary ritual at least over the ST1–ST4 graves, and only subsequently would have stained some of the skeletons [15]; ST1 had traces of ochre on the right side of the skull; the skull of ST3, found by accident by Maviglia in 1940 apparently due to a small landslide caused by a dripping of the cave vault, was heavily smeared with red ochre. Only a few other bones of this individual were collected [14,15]: the epistrophe, the atlas, some other vertebrae, a scapula, the head of the left humerus and a clavicle. In addition, some phalanges of the left hand were found adhered to the chin, as if the hand had been placed over the mouth (suggesting primary inhumation); other phalanges of the right hand were lower down [15] (p. 103 and pl. 4, Figure 2). In fact, Maviglia’s illustration of the in situ ST3 skull shows two phalanges resting on the lower jaw (Figure 4A). Remains of ST2 and ST4, both looted by clandestine diggers, did not show any trace of the red pigment. Of ST2, covered by a thinner blanket of soil according to Maviglia [14], at a greater depth according to Graziosi [15] (Figure 2), only the skull was recovered. ST4 was largely incomplete of the upper part, including the skull. At the time of discovery, ST1 and ST2 were lying parallel to the cave axis (NW-SE), a short distance from each other. For these individuals, Graziosi [15] reported a north–south orientation, but a northwestern–southeastern orientation seems to be more correct (Figure 3): ST1 lying on his left side, head facing southeast; ST2 supine, head facing northwest. The body of ST4 was interred perpendicular to the cave axis, midway between ST1 and ST2, feet facing northeast, presumably with left hand bones under the gluteal region [15] (see Figure 2 and Figure 3). The orientation of ST3 is not at all clear, nor is the location of the related burial in relation to excavations conducted at the end of the last century and more recently in the early 2000s. Even the ST3 burial was disturbed by clandestine activity [14,15]. Maviglia himself [14] (p. 103), who first discovered this individual, stated that the orientation of ST3 inhumation was not clear. Graziosi [15] dubitatively hypothesized an east–west orientation (northeast–southwest according to the correct orientation of the cave axis). Mussi [2] reported the ST3 orientation as unknown; coherently, Alciati et al. [59] reported that the “original position cannot be determined in any way”. According to the drawings of Maviglia [14] (pl. 2) and Graziosi [15] (Figure 2), ST3 was found a little more than a meter from ST2 and about a meter from the large boulder of the trench α (Figure 3 and Figure 5).

Twelve perforated red deer canines were associated with the ST1 grave goods. Probable grave goods of ST4 consisted of fragments of animal bones, the end of a red deer antler branch, found near the left femur at hand level, and small pebbles [15]. It is possible that the hyena skull found 50 cm from the remains of ST4 may have enriched the grave goods associated with this individual [15].

2.4. Human Stratigraphy

The stratigraphy of the site, above all in relation to the paleoanthropological context of the inhumations, is rather entangled and problematic, partly due to the disturbances perpetrated by clandestine diggers. The site stratigraphy appears patchy in places and in any case not always sufficiently comprehensive, at least by the modern standards required of archaeological excavations today. We have, however, to consider (or reconsider) the pioneering works of Maviglia and Graziosi [14,15,60]. As for the inhumations, these authors indicated that, because the ochre carpet was continuous and extended over ST1–ST4, bodies of these individuals would have been interred at the same time. Due to the absence of lithic industries and food remains in the inhumation layer E, the same authors argued that deposition of the overlying layer D (rich in industries and vertebrate faunal remains) could not have been completed at the time of inhumation or, in other words, that, much likely, only a minor part of this layer had been already deposited. They provided explanations to this hypothesis, based on the fact that if layer D had already been laid down in its full thickness at the time of burial, some remains of this level would have been found inside the graves, and the carpet of ochre on top of the ST1–ST4 burials could not have been laid. Much more recently, this hypothesis has been accepted [18,49], at least in part, although without new insights or explanations that consider the anthropogenic origin as the main driver of the processes that led to the formation of the D–A layers. In the last 15 years, radiometric analyses have been performed on ST1 [12,16], ST2 [49], ST4 and ST5 [18], on charcoals [19] from anthropogenic deposits overlaying inhumation layer E and on a faunal remain (aurochs tooth) from the basal part of layer D [17] from the same collecting area indicated for the charcoals investigated by Forgia et al. [19]. Figure 6 shows a synoptic view of these radiometric analyses.

The age reported for ST5, being comparable to the age of most of the inhumations in layer E, is not consistent with the stratigraphic position, since this individual was collected from layer B. It is possible that ST5, whose remains consist of the skull and some other bones, underwent a secondary burial. Recovery and redeposition of human bones, particularly the skull, is indeed a known practice in prehistory [61,62,63]. Graziosi [15] (p. 139), commenting on the stratigraphic context of ST5 and the scarcity of its bone remains, considered this hypothesis as plausible. However, it cannot be ruled out that ST5 had a primary burial at a time so close to the other earliest burials in layer E that the two inhumation events appear temporally overlapping due to the inadequate accuracy of radiocarbon dating.

The average date cal BP of 13,462 kyr attributed to ST2 [49] is not coherent with the ages of the other dated individuals ST1 and ST4 who were inhumated in the same layer E, nor with Graziosi’s observations that inhumations occurred at the time of the early formation of level D [15]. The radiometric analysis of Garilli et al. [17], in fact, reported an average date cal BP of 14,966 kyr (2σ) and 15,012 kyr (1σ) from layer D for an aurochs tooth, which matches with the age of ST1 and ST4. This age is consistent with the Graziosi’s observation [15] that the inhumations are likely to be chronologically related to the earliest formation of layer D.

3. Materials and Methods

3.1. The Investigated Section

During one of the fieldworks conducted in 2021 in the San Teodoro cave, a stretch of the northwestern section of the trench α performed by Prof. Bonfiglio’s team in 1998, corresponding to E9–G9 excavation squares, was resumed. This section, up to approximately 1.50 m in height and 2.40 m in width, lies at the boundary between the excavation α and the southeastern part (squares H22 and G22) of the excavation ST21 (Figure 5 and Figure 7), begun by a team of researchers from the University of Palermo in 2021. The section is distant a few decimeters from the large boulder (b1 in Figure 2, Figure 3 and Figure 5) in the trench α, the same boulder as drawn by Maviglia [14] (pl. 2) and Graziosi [15] (Figure 1). This location is of particular significance, as it is on the border of the excavations conducted by Maviglia in 1938–1940 and by Graziosi in 1942 (Figure 3). In detail, the part of the section furthest from the cave wall, lying a little more than a meter from the large boulder, is close to the location where the skull of ST3 was found (Figure 3). On the northwestern side, the section is bordered by a drywall raised on the boundary of the Maviglia excavation probably to protect the section (Figure 7A,D).

We considered two reference points for a correct positioning of ST3’s skull based on figure 2 in Graziosi [15]: a distance from the prominence of the northeastern cave wall as appeared at the time of Graziosi’s field working and a comparable distance from the large boulder in the trench α (Figure 3). A further indication could be provided by the distance along the vertical of the section between the vestiges of the red ochre layer and the end of the red deer antler.

3.2. The 3D Reconstructions

With the aim of understanding the orientation of individual ST3 at the time of burial, we tried to provide a 3D reconstruction of the position of parts of the skeleton most indicative of its location. As a starting point, we reconstructed the ST3 skull position based on the photo that Maviglia [14] (pl. 4, Figure 2) took at the time of ST3’s discovery in 1940 and the description he provided (p. 103). The photo (Figure 4A) shows part of the upper and lower jaws, respectively, with seven and four teeth; two phalanges placed on the lower jaw and the mental foramen (Figure 4A). Among the upper teeth visible in the photo, we identified the central incisors. The main aim of our 3D reconstruction (Figure 8 and Figure 9 and Supplementary Model S1) was first to establish the bone orientation as in Maviglia’s photo (Figure 4B), with particular regard to the upper and lower jaws, then to provide indication about orientation of ST3’s skull and body at the burial time. To understand the position of ST3’s body, we have made a more complete reconstruction based also on Maviglia’s description of what was found along with the bones of the head. Among these bones, which include elements of arms, shoulders and trunk, those of the hands certainly play a prominent role, since Maviglia [14] (p. 103) hypothesized that the left hand was placed over the mouth. We also gave attention to Graziosi’s images [15] (pl. 3, figures a and 13) of the ST3 skull as restored in Florence a few years after its sampling. These images do not show the incisors that can, instead, be seen in Maviglia’s photo. It is plausible that these teeth were lost during the different transports undergone by ST3 or during sampling. For the orientation of the lower jaw, we also relied on the identification of the mental foramen (Figure 4A).

The textures used in the reconstruction for Figure 8 and Figure 10 were modified from the files “Tiled PBR material—Dirt Rock 001” by XY01 [64] and “Red_Mud_Stones” by Rob Tuytel [65], “limestone5-bl” available at Free PBR Materials [66], and “skeleton-character-psx” by Vinrax [67]. The skeleton model was modified from “Animated Skeleton Casts Spell with Hands” by LasquetiSpice [68]. Version 3.5.1 of the free and open source BLENDER software was utilized for the bone and skeleton reconstruction. The model in Figure 11 was made using the software SketchUp Make v.13.0.4124.

In order to shed light on some uncertainties related to the stratigraphic context and anthropological issues, such as the location of ST3 and the relationships between ST3 and the recently discovered ochre, we also attempted a reconstruction based on combining the observations of early archaeologists working at the San Teodoro site with recent excavations and fieldwork. In particular, we used the section illustrated and discussed by Graziosi [15] (pl. 3, Figure 1), the planimetric reconstruction accompanied by solid landmarks, and the main features of the revised section exposed on the northwestern side of the trench α [41].

Using the same method as described above, further reconstructions (Figure 10 and Figure 11 and Supplementary Model S2) were made in order to locate the burial area in a more general topographic context within the San Teodoro cave and to attempt correlation of layers preserved at the investigated section with those described by Graziosi [15]. For two of these reconstructions (Figure 10A,B), we considered the current maximum amount of sunlight penetrating the cave, taking into account the topography of the cave entrance, of the Pizzo Castellaro relief, where the cave opens, and the direction of the cave axis. This lighting condition occurs during the summer period, with the sun at 15° on the horizon and an azimuth angle of 315°. As a topographic base for reconstruction, as in Figure 10, we used the survey of the Speleo Club Ibleo—Ragusa (Supplementary Figure S1) performed by Iemmolo et al. [43]. For the solar illumination of the study site, we used SunEarthTools.com [69].

The supplemental interactive models of the 3D reconstructions can be performed in one of the 3D viewers available online, e.g., ONLINE 3D VIEWER [70].

4. Results

The investigation of the resumed section at the boundary of excavation ST21 and trench α provided the following results. From bottom to top, the section (Figure 7A) consists of a 1.30–1.50 m-thick basal deposit of fine yellowish to grayish sediment with sporadic hyena coprolites, sterile as far as Paleolithic lithic industries are concerned. A horizontal, millimeter-thick, brownish-orange layer crosses the basal deposit in the investigated section (Figure 7B,D). This deposit, the same that Graziosi [15] described as layer E, changes rather abruptly to an about 20 cm-thick, brownish-reddish layer with charcoal fragments, lithic industries (quartzites and flints) and food remains. This layer is comparable with layer D described by Graziosi [15]. At the studied section, layer D is overlaid by a gray fine sediments bearing abundant remains of animal bones, presumably representing layer C of Graziosi [15] (see below). This layer appears partly removed by the ST21 excavation. In the southwestern part of the section, bordering the area that was excavated by Maviglia, the transition between layers E and D is interspersed with a lens of red ochre about 15 cm wide and about 3 cm thick just below the base of the upper layer (Figure 7B–D).

The 3D reconstructions we performed based on an original image by Maviglia [14] showed clearly the position of ST3’s skull and the phalanges of his left hand (Figure 8 and Figure 9). These reconstructions also allowed us to formulate hypotheses about the position of ST3’s hands and arms, which may be essential for understanding the burial position of this individual. The position of the skull was found to be evidently tilted toward the right side, while the mandible appeared slightly rotated toward the same side, as if the head was only slightly rotated and likely slightly tilted at the time of burial, and thereafter the skullcap and upper jaw had rested on the right side.

The stratigraphic reconstruction, as in Figure 11 and Supplemental Model S2, made it possible to correlate the layers described by Graziosi [15] (Figure 1), which no longer exists in the area of the 1940 excavation, with those cropping out at the investigated section. This correlation points to the preservation, at least partial, of the anthropogenic layer C and clearly shows that the layer of red ochre found at the boundary of between excavation squares G21 and G9 is compatible and correlatable with that which covered individual ST3.

In a much broader context of the site, reconstructions, as in Figure 10, showed all the main topographic features of the cave, particularly the considerable rise in slope, starting about ten meters from the entrance, the location of anthropological excavations within the cave, and the maximum area of direct solar illumination. This area, which we assume to be not too dissimilar to the maximum sunlight area dating to 15–16 kyr ago, includes that part of the cave affected by the ST1–ST4 burials as well as the place where the few bones of individual ST5 were found.

5. Discussion

5.1. ST3 Inhumation: Orientation and Ritual

Archaeological sites may be altered by cultural and natural processes [8]. As a consequence, intentional human burials can be remarkably affected by disturbances that make it difficult to understand funerary rituals and stratigraphic details. Observation and conservation of undisturbed stratigraphic records can therefore be of great interest in understanding funerary contexts. For the San Teodoro site, one of the most important in the Mediterranean for the discovery of at least four Late Paleolithic burials, most of the information on burial practices goes back to the 1940s articles by Maviglia [14] and Graziosi [15], who first described and interpreted the anthropological significance of this Sicilian site. However, several decades after the discovery of human remains at San Teodoro, some authors looked doubtfully at the burial practice at this site because the original documentation appeared sketchy, partly due to several disturbances perpetrated by clandestine diggers that negatively affected the archaeological record. The rediscovery of red ochre found in situ again sheds light on the stratigraphic context described by Maviglia and Graziosi and provides good evidence that the patch of ochre found near the edge of Maviglia’s excavation was actually part of the carpet that covered individual ST3. The red ochre was, in fact, observed at the transition between a yellowish fine sediment, sterile of lithic industry, namely the layer where ST1–ST4 were found, and the reddish-brownish layer rich in quartzite, flint tools and charcoals, which attests to the early assiduous attendance of the site by Epigravettian people. The relationship between the ochre rediscovered at the investigated section and the ST3 burial is well supported by two pieces of evidence. Following our reconstructions and observations by Maviglia [14] and Graziosi [15], it can be argued that the location of the red ochre is a very short distance from the northeastern limit of Maviglia’s excavation and about 40–50 cm from what is reasonably assumed to be the position of the cranial find of ST3 (Figure 3). Another important hint is the finding of a red deer antler tine, found at about 15 cm below the red ochre within the inhumation level (Figure 7A,C). We interpret this finding as part of the ST3 individual’s grave goods. Significantly, a red deer antler tine was also found as part of the grave goods related to ST4 during the Maviglia’s excavation. The stratigraphic position of this finding is also significant. The red deer antler was close to ST3, in a position compatible with the grave bottom, as Maviglia [14] (p. 102) indicated a shallow burial, about 20 cm deep, much less deep than that of ST1, for which he reported a grave cavity about 50 cm high. The fact that this deer remain was not in the immediate vicinity of ST3, at an estimated distance of about 40 cm from his skull (Figure 7 and Figure 11), does not invalidate our hypothesis. Complete cervid antlers or tines of their antlers are commonly found in prehistoric burials [3,39,57,58,71,72]. These objects were usually laid near or on top of the deceased, but also, though rarely, in their immediate vicinity. At the Saint-Germain-la-Rivière burial, along with other vertebrate remains used as grave goods, a split reindeer antler was found about 60–70 cm from the buried human individual [3].

The hand phalanx found by Forgia et al. [19] near our placement of ST3 may indeed belong to this individual. Recovery of genetic material and DNA analysis from this remain could confirm this reasonable hypothesis.

As pointed out by Maviglia [14] and Graziosi [15], the Epigravettians who buried at least four of their companions at San Teodoro celebrated a rather unusual funeral rite in terms of the way and abundance that ochre was used. In almost all known prehistoric intentional burials, the use of ochre seems to have involved spreading of this red pigment on the grave floor, around or on the bodies (sometimes referred to as “stained burial”), or the deposition of ochre-smeared objects or small portions of ochre also in the pit fill [3,7,8,10,11,36,39,55,56,60,73,74,75]. At the study site, a layer of ochre of the considerable average thickness of 5 cm would have been spread only on the ground covering the ST1–ST4 burials [14,15]. Stratimetric data seem to support this mode of spreading the red pigment. In fact, the ochre found in the investigated section, in the immediate vicinity of the position reconstructed for ST3, a few centimeters above one of the probable grave goods of this individual, seals the inhumation layer E and is immediately followed by the oldest of the anthropogenic layers (layer D). We cannot, however, confirm the abundance/extension of the ochre or even the contemporaneity of the four certainly intentional burials (ST1–ST4) as suggested by Graziosi [15]. On the other hand, it cannot be ruled out that relatively little time passed between burials—at any rate a time short enough not to erase the memory of the burial area—and only later, possibly immediately after the last burial or anyway during the earliest more assiduous frequentations of the site, the ochre layer was spread over at least ST1–ST4, maybe to emotively connect the burials and/or to mark the sepulchral area. In addition, given also the copious availability of the pigment at the disposal of the Epigravettians from San Teodoro [25], it is possible that the red pigment may have been spread in several episodes during a relatively short period. This could have had the purpose of restoring the ochre carpet as a carrier of symbolic meanings.

Spreading of ochre on the top ground covering the burial is a very unusual ritual in the prehistoric record. Indeed, in Paleolithic burial rituals the red ochre was usually sprinkled on or under the deceased, e.g., [1,4,5,73], or within the burial, even in copious amounts together with grave goods as in the Late Paleolithic of European Russia [55]. To our knowledge, only one example of a clearly similar ritual use of red ochre is available in the literature. Among the several Mesolithic burials at Vedbæk, in Denmark, the red pigment was found around the skeletons and on the pit bottom of the GØ: AE double burial. However, at the GØ: AE burial, the pigment was also used as a “flat compact ochre cake” “at the top level, immediately below the marking stone”, “well above the two skeletons half a meter further down” [53]. This modality is comparable to that observed at the San Teodoro site, except for the amount of ochre used, which is much more abundant at this site according to Maviglia [14] and Graziosi [15]. Peterson et al. [53] reported as a further comparable case that of burial 57 in the Meso-Neolithic Zvejnieki ground in northern Latvia [30]. However, it seems that burial 57, although unusual, is not entirely comparable [30] (p. 120, Figure 7). The reason for this mode of using ochre remains unknown. It can be debated whether the use of ochre in burial rituals may be related to symbolic or functional purposes or whether both aspects are involved. However, this debate still remains in the realm of theoretical speculation [26,27,37]. Regarding the symbolic use of ochre in prehistoric burials, one hypothesis—considered hazardous by Marshack [27]—invokes the choice of the color red as a reminder of blood and thus of life [26,30,35,54,76]. In this case, the use of ochre is explained as an attempt to prolong the life of the deceased [72]. Functional aspects of ochre use may be related to its antibacterial, antioxidant and insect repellent properties [33,35,77] or its ability to retard decomposition [78] and neutralize smells [79]. The hypothesis that spreading ochre on the top ground of the burial—as in the case of the study site—would have protected bodies from carrion eaters, such as wild boars and bears, was discarded on experimental grounds [80]. There is, however, a reasonable hypothesis that could be advanced for the case of the study site. In the Finnish Neolithic of Kukkarkoski Comb Ware Burial Ground, where new burials were connected to old ones also practicing the overlapping of graves, some intensive ochre deposits possibly were used to indicate the location of the deceased [32]. It is plausible that ochre was used at San Teodoro with similar purpose, more likely to connect burials. This would explain why the ochre layer was spread continuously over ST1–ST4, even where no burials were present.

The 3D reconstruction of the position of the ST3’s skull based on the Maviglia’s image [15] (pl. 4, Figure 2 and Figure 3) allowed a series of interesting considerations on the orientation of this male individual at the time of his inhumation. The position/orientation of a skull can actually help in understanding the original position at the time of burial and indicate movement depending on decomposition, the nature of the covering sediment and the shape of the grave [81]. When examining the skull, however, it must be considered that the mandible is a labile part, whose movements should be regarded as separated from those of the cranium [81]. In general, after decomposition of the body, the covering sediment represents a valuable support for the skeleton, especially if it is loose and characterized by fine-grained texture, as at the study site [81]. Most likely, therefore, the sediment used to cover ST3’s body supported the skull, which at the time of discovery reasonably reflected the original position of the burial, or allowed only minimal displacements. This is also supported by other bone remains recovered by Maviglia [14] near the skull, including bones in direct connection, such as the clavicle, scapula and the head of the left humerus, and other bones in succession to these, such as the epistrophe, atlas and occipital bone. This anatomical arrangement appears as an orderly arrangement of bones, like the one we have tried to reconstruct in the 3D model in Figure 6, and suggests that no significant taphonomic process disturbed the inhumation of ST3.

The fact that ST3’s skull looks toward the tip of the feet indicates a NE–SW orientation of the body, with the feet facing the southwest wall of the cave, thus parallel to ST4 but with the opposite direction. The slight inclination of the skull to the right should probably be attributed partly to the shape of the pit, which was probably narrow, and partly to the decomposition of tissues that have reduced the skull’s support supplied by the sediment. The presence of two phalanges of the left hand, laid between the upper and the lower jaws, and of some phalanges of the right hand slightly below, drives the position of arms and elbows (Figure 8 and Figure 9 and Supplementary Model S1). These parts of the body should have been superimposed on the ribcage, a position that would have allowed the hands to remain on or near the mouth. In our reconstruction we assumed that ST3 was buried supine, but it is equally reasonable that his body was slightly on the right side given the inclination of the skull. Similar hand positions were found in burial H-29 of the Pre-Pottery Neolithic from Atlit-Yam, Israel [82].

At San Teodoro cave, the area where the inhumations took place and where much of the Epigravettian activity went on might not have been chosen at random. The simulation of solar illumination in Figure 10 shows that ST1–ST4 individuals were buried at a very few meters from the entrance, quite close to the northeastern wall, in the only area of the cave today subject to direct sunlight, an area considerably less extensive than the entire cave surface. For the San Teodoro site, in the absence of evidence to the contrary, it is possible to hypothesize a relationship between the location of burials and the particular sunny lighting conditions of the chosen area. Assuming that at the time of the burials the entrance threshold to the cave was higher [40]—a usual morphology for a cave entrance not intensely altered by anthropogenic activities or erosive processes (see Supplementary Figure S1)—the sunlight beam seen from the interior by the Epigravettians was probably meant to be more striking because of the greater contrast created between the light beam and the dark interior.

5.2. Stratigraphy

The resuming of the section at the boundary between the ST21 and Bonfiglio excavations provided new insights into the current stratigraphy of the anthropogenic deposits in San Teodoro cave. Comparison of our results with past reconstructions allowed us to outline the evolution of major anthropological and depositional events (Figure 12). The section showed the stratigraphic succession as reported by Maviglia and Graziosi at the transition between the inhumation layer and the later anthropogenic layers through the red ochre cover used for the funerary ritual. Our field observations and 3D reconstruction (Figure 7 and Figure 11 and Supplementary Model S2) indicate that, in addition to layer D, at least part of layer C, which was thought to have been lost in the area historically and recently concerned with anthropological excavations, is still preserved.

The radiometric ages of about 16–16.5 kyr BP of Forgia et al. [19], derived from charcoal assumed to form a lens within layer E, may indeed provide a chronological indication of human attendance prior to the inhumation event(s). However, the processes that led to the deposition of the dated charcoal in the northeastern sector of the section studied here, could have a different explanation than that indicated by Forgia and co-authors [19]. Our investigation suggests that the charcoal remains do not form a lens within the inhumation layer, as in the case described by Graziosi [15] for a fireplace intruded in layer E, in the vicinity of ST1’s inhumation (Figure 2 and Figure 3). Rather, as shown in Figure 7B,C, charcoals are within layer D, the base of which can be dated (Figure 6) to about 15 kyr ago [17], in a stratigraphic context not subject to disturbance. No lentiform geometry is in fact observable in the section, no charcoal intrusion into layer E, nor sediment rubification [83,84] under the presumed lens, but just scatters of charcoals above the top of layer E. Charcoal, indeed, is a material easily subject to reworking, especially in anthropogenic contexts such as caves inhabited in prehistoric times. As a consequence, interpretation of radiometric data of cave coals must be supported by microstratigraphic and geodynamic analyses and careful screening of the material subjected to dating [85,86,87]. In our opinion, the dated charcoals probably were redeposited in layer D after reworking from one of the fireplaces described as intruded in layer E [15]. The redeposition presumably occurred due to anthropogenic activities, such as repeated ground restoration resulting in the redistribution of older elements. This is supported by the co-occurrence of lithics, lacking in the fireplace, attesting the earliest attendances [15] (p. 127), and by stratimetric evidence such as the absence in the investigated section of the typical intrusion structures (Figure 7). In addition, we did not observe any evidence of the disturbance reported [19] for the southwestern part of the section. On the contrary, there are several clues supporting an undisturbed environment: (1) the continuity and horizontality (Figure 7B,D) of the thin orange layer, just below the transition zone from the red ochre to anthropogenic layers D in the most southwestern part of the section; (2) the congruence with the succession of layers described by Graziosi [15]; (3) the continuity of the boundary between layer E and layer D, visible along the whole section as well in the other walls of the trench α; and (4) the stratigraphic and spatial position of the ochre and the red deer antler tine, both of which are very likely correlated with the funerary rite of ST3.

In the upper part of the study section, the change in color from reddish brown to light gray marks the transition of layer D to an overlying different layer. The overlying layer is characterized by the occurrence of very abundant bone remains of animals used for food consumption and a much more conspicuous content of lithics in comparison with layer D. These features point to the occurrence of widespread and numerous hearths over the top of layer D [88], possibly as a result of more continuous frequentation of the site by the Epigravettian people. This suggests attributing the gray deposit to layer C of Graziosi [15], who describes it as having formed during the period of maximum occupation of the site also due to the increasing content of lithics and vertebrate animal bones.

6. Conclusions

Some 80 years after the discovery of intentionally inhumated humans during the Late Paleolithic of San Teodoro cave, we provided new insights into the stratigraphy of this site and the burial features concerning the inhumation of the individual known as ST3.

Field observations, combined with 3D reconstructions and data from the pioneering literature, proved to be a valuable tool to detail the location of ST3, as well as its burial context, and to further investigate the stratigraphy of cave deposits related to Epigravettian attendance.

It is very likely that ST3’s body was interred supine or slightly on his right side, with a northeast–southwest direction, feet facing the southwestern wall of the cave, hands and elbows on the ribcage, perhaps because of the narrow shape of the pit. He probably was accompanied by grave goods and the top of his grave cover sprinkled with red ochre.

The section investigated at the boundary between excavations ST21 and Bonfiglio 1998 preserves in situ the vestiges of the complete transition from inhumation layer E to anthropogenic layer D through the red ochre layer, consistent with Maviglia’s and Graziosi’s observations dating to the late 1930s and early 1940s. Our 3D reconstruction of the stratigraphy in the area of the anthropological excavations points to the preservation, probably partial, of the anthropogenic layer C. This layer can also be distinguished in empirical observations of the studied section by its gray color (instead of the reddish color of layer D) and the huge amount of vertebrate and lithic remains. The ochre patch outcropping in the studied section is what remains of the ritual cover of the ST3 grave and testifies to the originality of the ritual use of red ochre in the funerary context of San Teodoro compared to other known prehistoric burials. Possibly, the ochre used at San Teodoro had the function of connecting at least the four inhumations ST1–ST4.

Review of the studied section revealed that the depositional and taphonomic processes that led to the formation of the Paleolithic deposits were governed by activities related to human frequentation of the site.

Located at the edge of the Bonfiglio and Maviglia excavations, the portion of the investigated section with red ochre is at risk of collapse. It constitutes the trace of an important record of the funerary rite of individuals ST1–ST4, to date the only in situ sign of prehistoric inhumations in Sicily. Therefore, we believe that considerable efforts should be made to preserve this remarkable heritage for future generations.