Victoria Cabrera Site: A Middle Stone Age site at Olduvai Gorge, Tanzania

Abstract

Olduvai Gorge (Tanzania) is a key site for the study and comprehension of human evolution in East Africa. However, the origin of Homo sapiens and the Middle Stone Age have been poorly understood in the Gorge thus far. In this study, we present the dating, taphonomic, technological and typological analyses of the lithic industry and faunal remains excavated at the Victoria Cabrera Site (VCS) during the 2017 fieldwork season. The stratigraphic sequence of the site contains several levels of fluvial origin, some of them with lithic and faunal remains. Most remains (lithics and bones) are affected by rounding processes. Infrared stimulated luminescence (IRSL) dating has provided ages between 90 and 70 ka BP for these layers that can be related to the Marine Isotope Stage 5a. Lithics are scarce and the dominant raw materials are Naibor Soit quarzite and basalt, followed by phonolite, chert and quartz as the most common raw materials. The lithic technology is based on the production of flakes obtained from prepared cores, with the discoid method and, to a lesser extent the Levallois method being the most frequently used. The retouched blanks are described as “substratum” or “domestic” tools (sidescrapers, notches or denticulates). There is just one heavy duty piece and points are absent. The faunal assemblage includes Alcelaphinae, Antilopinae and Equidae. The major parts of the unidentified faunal remains are size 3 ungulates, followed by carcasses of size 2 and size 4 species. We cannot relate lithic and bones because no cut marks or percussion marks have been identified and carnivore action is scarce. In sum, VCS represents the first accurately dated Middle Stone Age site, with lithic and faunal remains in Olduvai Gorge.

Introduction

Olduvai Gorge has been known on merit as the Cradle of Humankind since 1959, when Mary and Louis Leakey discovered the Zinjanthropus (Paranthropus) skull in the FLK site. Since that time, several research teams and projects have devoted considerable effort to the paleoanthropological, cultural and behavioural knowledge of the first figures of genus Homo: the Oldowan and the Acheulean. However, little attention has been paid to upper beds in Olduvai Gorge, as well as to the origin of our species and its material culture.

Olduvai Gorge is located in the South-eastern Serengeti (northern Tanzania), in the great East African Rift (Fig. 1). The erosion of the Olduvai River, cutting through a plain amidst a line of volcanoes in the South—e.g. Sadiman, Lemagrut, Olmoti or Ngorongoro—and lower-height metamorphic reliefs in the North—like Olongoidjo or Watumi—generated the gorge (Uribelarrea, 2014). The structural landscape remained stable until the Ngorongoro Collapse 2.2 Ma, whilst the plain was affected by external processes (weathering, and fluvial and aeolian sedimentation). Following the basin sedimentation, the Olduvai River eroded the plain generating the gorge we know today, split into the Main and Side Gorges. Throughout that process, an intense tectonic activity as well as further basin sedimentation took place, generally by volcanic processes during the Middle and Late Pleistocene. Thanks to their low-energy sedimentary conditions, the availability of water sources and the primary sources of lithic raw material, the spot was the perfect place for the occupation of the territory by hominin groups. Such an assertion is contrasted with the more than 120 sites and the remains of four hominin species discovered (Paranthopus boisei, Homo habilis, Homo ergaster/erectus and Homo sapiens). Nowadays, the gorge geology can be split into seven beds (Hay, 1976).

Bed I (2.1Ma) in Olduvai Gorge has been well-known since the 1930s. Bed I deposits contain an important amount of sites bearing hominin fossil remains and archaeological assemblages (including fauna and lithic remains) that seem to belong to the first representatives of our genus (Homo) as well as to other contemporary hominin species (Paranthropus boisei) (Leakey, 1971; Uribelarrea, 2014; Domínguez-Rodrigo et al., 2007, 2010; Barboni et al., 2010). Oldowan and Acheulean industries have been discovered in Bed II in sites (e.g. BK or SHK) in association with remains of Homo erectus (Ashley et al., 2014; Domínguez-Rodrigo et al., 2013; Uribelarrea et al., 2014). Beds III and IV have provided more than 40 Acheulean sites (Leakey and Roe, 1994). After that, the Masek Bed was deposited while the great faults affect the gorge move (Leakey and Roe, 1994; Hay, 1976). This Bed represents the sedimentary clogging of the basin. At this juncture, around 0.6 and 0.4 Ma, the first fault movement happened, together with the movement of the five faults affecting the gorge. The result is the creation of the Olbalbal depression—in the east of the gorge—and the creation of the Olduvai River itself, which begins in Lake Ndutu and flows into the newly created depression. This marks the start of Olduvai Gorge due to the exposition of the above-mentioned deposits on account of the erosive process (Hay, 1976; Uribelarrea, 2014). The levels of Ndutu and Naisiusiu Beds will be deposited above them.

The Middle Stone Age (MSA) has been defined by the knapping of prepared cores, especially the Levallois-type and, above all, by retouched pieces, particularly projectile points associated to Homo sapiens (Clark, 1988; McBrearty and Books, 2000; Basell, 2013). Stratigraphically, the MSA in Olduvai falls within the Ndutu Bed, which consists principally of tuffaceous, aeolian, fluvial and colluvial deposits, which are deposited in the final stage of the gorge formation and, which cover the eroded areas by the Olduvai River itself (Hay, 1976). Research on the human occupation of this Bed has been discontinuous and biased in the three most relevant lines of research: chronology, human remains, and archaeology.

One unsatisfactorily resolved issue for Bed Ndutu is its chronology. The initial approach was carried by Hay, who considered that Upper Ndutu might be in the range of 60–32 ka BP, and Lower Ndutu in the range of 60–400 ka BP (Hay, 1976: 152), although he used indirect methods to this end. Manega (1993) is without a doubt, the researcher who has devoted the utmost efforts to discover the chronology of the Ndutu Bed. According to his research, on the basis of the Single Cristal Laser Fusion (SCLF) technique, Upper Ndutu might be located in a range between 210 ± 20 ka BP and 450 ± 40 ka BP and using amino acids around 260-500 ka BP. Manega dates the start of Bed Naisiusiu in 42 ± 10 ka BP by using SCLF. The latter runs counter to Skinner's dating for the same Bed, which was dated in 62 ± 5 ka BP (Skinner et al., 2003)—and it is therefore, far older than Manega's proposal—by using ESR (electron spin resonance).

Several Homo sapiens remains have been found in the Ndutu Bed. The first one is OH-11, a palate and maxillary arch of Homo sp found on the surface in western DK and assigned to the Ndutu Bed (Hay, 1976, pp. 159, Leakey, 1971, pp. 230), which presents a sturdy morphology for the modern standards (Rightmire, 1980, pp. 227). The second one is OH-1968, a skull fragment partially in situ in Upper Ndutu levels (Von Zieten, 2009), but whose origin and study are uncertain. The last one is OH-83, a partial calvaria partially recovered in stratigraphy in the Upper Ndutu levels, close to the PLK site (Reiner et al., 2017) and identified as Homo sapiens. There is no direct dating for any of them, therefore they have been placed in the chronological range proposed by Hay for Ndutu (Hay, 1976, pp. 152) with the uncertainties that, as we have seen, it presents.

The archaeological study of the Ndutu Bed has also been intermittent over the last decades, with only three previous interventions. The first paper on MSA was carried out by Mary Leakey and colleagues, and in it, two sites with MSA industry are presented (Leakey et al., 1972). Both sites were not located on the paper, tough follow-up work placed them in the localities 4b and 26 (Hay, 1976, pp. 159) without further details of planimetry or stratigraphy, one of them belonging to Upper Ndutu and the other to Lower Ndutu (Hay, 1976, pp. 28). The lithics of both sites were presented uniformly, and in there, the presence of prepared cores (Levallois and discoid) and a few retouched tools was outlined. The most used raw material was basalt, and there was a small quantity of phonolite and quartz. The assemblages, although with due precaution, were linked to the Early MSA of Lake Eyasi (Leakey et al., 1972).

The second project related to the MSA was directed by A. Mabulla, co-author in this study, from the University of Dar es Salaam in collaboration with the University of Florida between 1989 and 1990 (Mabulla, 1990). Several localities shedding light on lithic and fauna find-spots were explored—none of them in stratigraphy except for locality 26 (Ndutu type site according to Hay, 1976).

The third research project was led by M. Eren during the course of the 2013 field season and was the most intense study conducted so far. They carried out a systematic survey between the second fault and the Olbalbal depression, besides other selective surveys in the junction of both gorges, resulting in the discovery of 72 find-spots which provided 521 MSA flaked pieces, as well as some faunal remains (Eren et al., 2014).

Even though the two first projects had already made it clear that the Ndutu Bed reported having occupations belonging to MSA, certainly after the work of M. Eren's team, the extraordinary potential which Olduvai Gorge has for the first modern humans in northern Tanzania became clear.

In this paper, we present the first results of Victoria Cabrera Site (VCS), which is the first site in stratigraphy with radiometric dating and several stratigraphic levels with lithic industry and faunal remains found in Olduvai Gorge.