This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/authorsrights Author's personal copy Quaternary International 343 (2014) 117e135 Contents lists available at ScienceDirect Quaternary International journal homepage: www.elsevier.com/locate/quaint Cultural change or continuity in the late MSA/Early LSA of southeastern Ethiopia? The site of Goda Buticha, Dire Dawa area David Pleurdeau a, Erella Hovers b, *, Zelalem Assefa c, Asfawossen Asrat d, Osbjorn Pearson e, Jean-Jacques Bahain a, Yin Man Lam f a Département de Préhistoire, UMR 7194, CNRS Muséum national d’Histoire naturelle, France Institute of Archaeology, The Hebrew University, Mt. Scopus, Jerusalem, Israel Department of Anthropology, Archaeobiology Program & Human Origins Program, National Museum of Natural History, USA d Department of Earth Sciences, Addis Ababa University, Ethiopia e Department of Anthropology, University of New Mexico, USA f Department of Anthropology, University of Victoria, Canada b c a r t i c l e i n f o a b s t r a c t Article history: Available online 27 March 2014 Goda Buticha is a newly discovered cave site in southeastern Ethiopia, containing MSA and LSA cultural material, faunal remains, beads, and human skeletal remains. A 2.3 m-deep sedimentary sequence records two occupational phases separated by a sharp chronological hiatus, in the Upper Pleistocene (w 43e31.5 ka cal BP) and in the mid- Holocene (7.8e4.7 ka cal BP). Faunal remains suggest changes in paleoecological conditions that are in agreement with patterns documented in regional speleothembased reconstructions. The lithic assemblage at the base of the sequence is clearly MSA, with Levallois production, unifacial and bifacial points, relatively large debitage and use of local raw materials, associated with a microlithic component. The overlaying LSA assemblage contains diagnostic artifacts (backed microliths and bladelet production), with ubiquitous use of obsidian and MSA elements that appear in the Holocene. In the absence of indications for post-depositional mixture, the apparent cultural continuity of MSA elements from the Upper Pleistocene into the Middle Holocene at Goda Buticha may represent yet another variation of the elusive MSA/LSA transition. Goda Buticha is a key site for reevaluating the dynamics and tempo of this transition in eastern Africa. Ó 2014 Elsevier Ltd and INQUA. All rights reserved. Keywords: MSAeLSA transition Upper Pleistocene Eastern Africa Southeastern Ethiopia Microliths 1. Introduction Intact, well-stratified archaeological deposits that preserve bone and can be dated radiometrically are crucial for our understanding of the Middle Stone Age (MSA) and earliest Later Stone Age (LSA) in East Africa. These periods and geographic region are thought to be highly important for the origin and spread of modern humans as well as the emergence of important cultural innovations (Clark, 1988; Klein, 1995; Lahr and Foley, 1998; Ambrose, 1998a; McBrearty and Brooks, 2000; Walter et al., 2000; Clark et al., 2003; Bruggemann et al., 2004; Forster and Matsumura, 2005; Marean and Assefa, 2005; McDougall et al., 2005; Beyin, 2006, * Corresponding author. E-mail addresses: dpleurd@mnhn.fr (D. Pleurdeau), hovers@mscc.huji.ac.il, ehovers@gmail.com (E. Hovers), zassefa@gmail.com (Z. Assefa), asrata@geol.aau. edu.et (A. Asrat), ompear@unm.edu (O. Pearson), bahain@mnhn.fr (J.-J. Bahain), ymlam@uvic.ca (Y.M. Lam). http://dx.doi.org/10.1016/j.quaint.2014.02.001 1040-6182/Ó 2014 Elsevier Ltd and INQUA. All rights reserved. 2011b; Behar et al., 2008; Klein, 2008; Tishkoff et al., 2009; Rose et al., 2011; Delagnes et al., 2012). Unfortunately, only a handful of stratified sites document the MSA and LSA in East Africa, for example Mumba rockshelter in Tanzania (Mehlman, 1989; DiezMartín et al., 2009; Gliganic et al., 2012; Eren et al., 2013); Enkapune Ya Moto (Ambrose, 1998b) and Lukenya Hill in Kenya (Merrick, 1975; Barut-Kusimba, 2001) and Porc-Epic Cave in Ethiopia. The Mesozoic limestone of the Antalo Formation, which crops out extensively in the southern wall of the Rift Valley around Harar (Ethiopia) (Bosellini et al., 2001), contains numerous karst caves suitable for paleoecological work (Asrat et al., 2007) and for the preservation of bone in archaeological deposits (Assefa, 2006). Previous work on Paleolithic deposits in the region focused mainly on exploration and excavations at the cave site of Porc Epic (Teilhard de Chardin, 1930; Breuil, 1934; Teilhard de Chardin et al., 1940; Breuil et al., 1951; Clark and Williams, 1978; Clark and Williamson, 1984; Brandt, 1986; Pleurdeau, 2003; Assefa, 2006; Assefa et al., 2008), with other sites mentioned briefly. Author's personal copy 118 D. Pleurdeau et al. / Quaternary International 343 (2014) 117e135 A survey conducted by Assefa et al. (2014) in 2007 and 2009 visited previously reported localities as well as new cave sites, documenting the potential for datable speleothems, other paleoenvironmental archives and rock art. Caves were examined for MSA remains that could contribute to the discussion about the dispersal of modern humans out of Africa as well as cultural evolution in Africa. Goda Buticha (Buticha cave, in the local Oromo language), containing thick deposits with both LSA and MSA artifacts, faunal bones and human remains, is a newly discovered site that had not been reported by earlier workers in the region (Fig. 1). 2. The site Buticha cave is located close to the hamlet of Kunama in the administrative district of Serkama village (Dire Dawa district), situated high in a cliff made up of crystalline limestone of the Gebredare Series (Fig. 1). It is a three-chambered karst cave with two mouths opening to roughly NE- and NW-oriented galleries, joined by a third small and dark SE-oriented chamber. The cave is a remnant of a larger system, which has been long destroyed, along NEeSW oriented faults and fractures related to the main Ethiopian Rift. The main gallery opens to NortheNorth West onto the slope. The site contains small rock paintings in dark pigment in the northeastern gallery. In contrast, all the in situ MSA and LSA lithic artifacts, fauna, human remains and ornaments were found in the northwestern gallery (Fig. 2). The sediments containing the Paleolithic material extend laterally into the SE chamber gallery. A large pit close to the cave’s entrance excavated some years ago by a local villager, and then used by porcupines, revealed a sediment column ca. 1.5 m thick, in which fauna and lithic artifacts appeared in situ. Cleaning of the pit revealed a long sequence of intact sediments. The southern wall of the pit was used as a stratigraphic guide for the test excavation conducted at the site. Save for the sediments from the very top 40 cm of the section, which were excavated in square-meter units, excavation was conducted in 50  50 cm quadrants. For the purpose of the test excavation, arbitrary excavation spits of 5e10 cm each were used, with constant evaluation of the lithostratigraphic conditions (changes in color and texture). On the whole an area of 2 m2 was excavated down to bedrock, which was exposed on the western side of the excavation (Figs. 2 and 3) during two field seasons in 2008 and 2011. A 230-cm sedimentary sequence has been exposed, with all archaeological remains >2 cm plotted in a 3D grid system. All the excavated sediment was dry-sieved through a 2 mm mesh. In total a sediment volume of 5 m3 was excavated during these two seasons. 3. Geochronology and sedimentary history Macroscopically, two major sedimentary complexes were distinguished, within which 14 sub-horizontal layers as well as 2 large indurated pockets have been identified (Fig. 3). Fig. 1. General location of Goda Buticha (Dire Dawa area, Ethiopia). Views of the cave from outside and within show its elevation relative to the local topography. Author's personal copy 119 D. Pleurdeau et al. / Quaternary International 343 (2014) 117e135 3.1. Complex I (upper complex) A 50e60 cm thick series of 7 interstratified white, light pink, and black fine-grained silts underlies a 10e30 cm thick layer of very recent organic material (straw, wood) in which many hyena coprolites were found. This series is consistent with anthropogenic ashes from fireplaces. The sediments are rich in long bone shaft fragments. Remains of micromammals were found throughout the ash layers, in some cases calcined and concentrated into dense pockets. A single piece of pottery was found at an elevation of 180e 170 cm above datum. At the base of the Upper Complex, there is a distinct 10e20 cm thick horizon that contains a large number of micromammal remains and some artifacts. Bedrock has been reached within square A1, where the cave wall is inclined toward the central part, some 80 cm above the datum. A limited test pit in square B1 revealed that artifact-bearing sediment continues down to at least 30 cm. Sediments of Complex II contain a rich lithic assemblage (MSA and LSA components), faunal remains, human remains, and several ostrich eggshell beads. Forty-four charcoal samples were collected from in situ contexts, at elevations ranging 230e50 cm above datum. Fourteen samples, all derived from the stratigraphic Lower Complex, were dated using Accelerator Mass Spectrometry (AMS) radiocarbon dating methods. The dates were calibrated using the IntCal 13 timescale (Reimer, 2013) (Table 1, Fig. 3). Table 1 14 C ages on charcoal from Buticha Cave according to elevation and grid. Elevation Grid square A1a/A1b 1 170e160 2 3 160e150 150e140 4 140e130 5 130e120 6 120e110 7 8 110e100 100e90 9 70e60 10 60e50 A1c/A1d B1a/B1b B1c/B1d 4130  30 4676  144 cal BP SacA#27480 (z ¼ 170) 6410  30 7347  71 cal BP SacA#27479 (z ¼ 141) *6200  30 7119  119 cal BP SacA#35205 (z ¼ 132) **6940  40 7797  118 cal BP beta#269596 **6370  50 7300  122 cal BP beta#269595 *6920  50 7791  126 cal BP beta#255123 *6470  40 7375  82 cal BP beta#269594 **5590  50 6380  88 cal BP beta#255124 ***6430  40 7350  74 cal BP beta#269597 ***6440  50 7351  80 cal BP beta#255122 *29,680  230 33,843  405 cal BP beta#269593 **31,590  270 35,497  595 cal BP beta#255121 42,500  1000 46,273  2084 cal BP1 SacA#35207 (z ¼ 69) 38,870  670 42,997  1053 cal BP SacA#35206 (z ¼ 54) Calibrated dates are in boldface. Calibration (IntCal 13 curve) at 1 sigma done with Online OxCal 4.2 (https://c14.arch.ox.ac.uk/login/login.php?Location¼/oxcal/OxCal.html). *Asterisks correspond to the presentation of samples in Fig. 3. 3.2. Complex II (lower complex) Underlying a 10 cm thick brown clay layer (Layer IIa/b) is a 170e 180 cm thick deposit of light brown clay. The increasing presence of limestone clasts (5e10 cm large) from an elevation of w100 cm above the datum led us to separate Complex II into two main layers (IIc and IId). The lowest part of Complex II (between 0 and 50e70 cm above the datum) in square B1 is largely indurated and concreted (IIf). Another large indurated pocket is distinguishable, with a distinctive orange color (IIe). The formation of these concretions is most likely post-depositional. The dates indicate episodes of human occupation in the late Upper Pleistocene from 42,500  1000 14C BP (46,273  2084 cal BP) to 29,680  230 (33,843  405 cal BP) in Layers IIf and IId and during the mid-Holocene from 6920  50 (7791  126 cal BP) to 4130  30 (4676  144 cal BP) in layer IIc and IIb. However, the available radiocarbon dates suggest a major hiatus in the accumulation of the fine-grained sediments of the Lower Complex (II), with the top of Layer IId dated to w29.5e 32 ka 14C BP (33.5e36 ka cal BP) and the lower part of layer IIc dated to w5.5e7 ka 14C BP (6.3e7.8 ka cal BP). This gap corresponds to the sedimentological change between Layer IId and Layer IIc. Author's personal copy 120 D. Pleurdeau et al. / Quaternary International 343 (2014) 117e135 OSL dating analyses are in progress at the Institut de Recherche sur les matériaux, University of Bordeaux, France. The very preliminary results are consistent with these radiocarbon dates and support this general chronostratigraphy (C. Tribolo, personal communication). 4. The lithic artifacts The characteristics of the lithic assemblage provide an essential line of evidence from which to define the cultural stratigraphy at the site and to assess changes in technological strategies. A total of 5463 artifacts were collected in 2008 and 2011 (Table 2). The average density is w1100/m3. However, 87% of the artifacts are concentrated in the Lower Complex, 30e150 cm above datum. In this part of the sequence, the lithic density reaches up to w2000 artifacts/ m3. The artifacts are well preserved. Some 200 large (>20 mm) chert and basalt items show indications of fire damage. In the basal part of the sequence, 15% of the large pieces are burned. Given the nearly identical trends that we observed in the samples from 2008 to 2011, we use the detailed analysis of the 2008 sample to describe the main characteristics of the assemblage and their temporal trends. Table 2 Vertical distribution of lithics in Goda Buticha according to excavation sub-squares. The part of the sequence with the main concentration of lithics is shaded. See text for details. Elevation above datum (cm) Upper Complex I Lower Complex II Layers IIaeIIc Layers IIdeIIf 190e200 180e190 170e180 160e170 150e160 140e150 130e140 120e130 110e120 100e110 90e100 80e90 70e80 60e70 50e60 40e50 30e40 20e30 10e20 0e10 Total Square A1 Square B1 a b c d 4 a 1 10 60 26 7 17 43 106 34 89 98 95 77 103 111 85 18 34 11 5 22 56 85 30 41 55 59 121 120 190 21 22 10 21 28 53 103 38 97 31 27 24 74 30 35 5 14 61 57 23 29 17 9 27 65 159 136 55 55 6 675 790 159 58 961 905 49 104 54 43 25 43 24 19 25 417 Total b 27 27 9 66 62 26 27 52 31 34 40 20 37 40 25 0 21 544 c 0 1 26 6 1 14 106 47 d 3 109 5 37 20 24 9 14 48 55 58 15 53 8 37 31 52 142 42 24 29 11 449 722 24 32 7 81 N 8 165 277 65 28 194 519 522 274 326 402 293 419 429 659 479 201 153 25 27 5463 % 0 3 5 1 1 4 10 10 5 6 7 5 8 8 12 9 4 3 0 0 100 Fig. 2. Plan and longitudinal profile of the SW gallery of Goda Buticha and location of the excavation. The bold line of the grid refers to the stratigraphic section of Fig. 3. Author's personal copy D. Pleurdeau et al. / Quaternary International 343 (2014) 117e135 121 Fig. 3. Synthetic stratigraphic section EeW (transversal) of A/B and 1/2 bands of Goda Buticha. Radiocarbon samples are marked by numbers corresponding to Table 1, where the dates and lab references are given. Throughout the sequence, all stages of lithic production are well represented, with the exception of the initial stage of decortication. Debris (i.e., flakes smaller than 20 mm and all chunks) are abundant and constitute 62% of the lithics (Table 3), indicating that much of the last stages of the knapping process were conducted on site, whereas the presence of various types of core trimming elements suggests on-site core rejuvenation. Author's personal copy 122 D. Pleurdeau et al. / Quaternary International 343 (2014) 117e135 Table 3 Vertical distribution of typo-technological categories. The part of the sequence with the main concentration of lithics is shaded. Elevation below datum (cm) Pebbles, nodules N 190e200 180e200 170e180 160e170 150e160 140e150 130e140 120e130 110e120 100e110 90e100 80e90 70e80 60e70 50e60 40e50 30e40 20e30 10e20 0e10 Total % 1 1 1 1 1 1 1 1 2 7 5 0 Debris Cores Chips <20 mm Chunks Sub-total N N % 6 87 34 15 39 97 141 101 80 145 66 135 98 87 149 20 22 2 6 1330 4 1 6 30 16 6 8 25 23 12 4 19 53 2 1 210 55 66 63 89 45 45 60 59 50 63 57 72 67 72 77 83 55 21 100 62 N Debitage (>20 mm) % 5 1 4 2 1 13 7 6 6 5 4 3 1 4 5 1 5 3 3 3 2 3 1 1 3 2 61* 2 N % 1 3 31 16 2 52 125 79 57 72 75 49 43 36 32 53 4 17 8 100 27 23 30 11 51 44 30 32 41 28 31 21 24 22 20 17 39 57 755 30 Retouched items N 2 10 3 18 7 6 3 16 17 10 9 20 14 11 13 5 4 3 6 7 6 5 7 9 5 9 3 2 2 2 5 14 141 6 In this table only 61 cores are reported. Another 8 cores with retouch were included with tools. In other analyses 69 cores are discussed. Fig. 4. Potential lithic sources in the area of Goda Buticha and Porc Epic. % Total N % 1 11 137 54 18 101 281 260 180 176 270 157 204 152 148 264 24 44 14 6 2502 100 100 100 100 100 100 100 100 100 100 100 100% 100 100 100 100 100 100 100 100 100 Author's personal copy 123 % 0 0 1 0 0 0 0 0 0 0 0 0 0 0 1 5 0 0 7 0 1 19 1 2 13 1 2 N 0 0 1 2 22 3 3 3 1 2 0 0 1 2 0 0 8 0 0 0 2 39 2 2 3 2 1 4 3 9 7 2 3 1 % N 237 3 1 1 12 12 15 12 15 14 15 20 17 32 51 3 6 7 1 20 2 3 5 1 2 1 1 3 1 1 138 1 11 6 12 8 12 3 11 5 21 38 2 7 N 2 0 0 0 1 0 0 0 0 0 0 1 3 20 12 5 61 169 161 110 86 150 85 100 73 65 127 10 24 3 4 1265 4 2 33 5 2 0 5 1 6 5 6 1 28 190e200 180e200 170e180 160e170 150e160 140e150 130e140 120e130 110e120 100e110 90e100 80e90 70e80 60e70 50e60 40e50 30e40 20e30 10e20 0e10 Total 1 % % 756 11 104 39 2 17 71 65 43 59 91 53 68 53 28 35 7 7 3 0 100 76 72 11 17 25 25 24 34 34 34 33 35 19 13 29 16 21 0 30 % N % 100 0 2 2 6 12 4 6 7 9 5 10 10 11 22 19 13 14 50 0 9 1 N % nb 0 0 1 0 0 2 1 2 1 1 0 1 1 0 0 0 0 0 0 17 1 0 0 0 0 0 1 4 2 7 5 4 2 5 3 14 14 8 16 0 17 6 % N N 0 0 15 22 28 60 60 62 61 49 56 54 49 48 44 48 42 55 21 67 51 Obsidian Extrusive Rocks Basalt Limestone Quartz Endogenous Rocks Quartzite Flint Sedimentary Rocks Elevation above datum (cm) Table 4 Distribution of raw materials. The part of the sequence with the main concentration of lithics is shaded. 4.2.2. Débitage The technological characteristics of the debitage are consistent with observations on the cores. Throughout the sequence, lithic reduction is geared towards the production of flakes. In general, débitage items seem to have resulted from various unipolar production methods. The products are often elongated and include laminar and lamellar items. Frequencies of blades and bladelets are low in the basal part of the sequence (13%e30% of the débitage in elevations below 70e80 cm) compared to over 50% towards the top of the Lower Complex (at 90e150 cm above datum) (Table 5; Fig. 6). This increased elongation of artifacts co-occurs with a decrease in % Indeterminate Rocks Other volcanics. 4.2. Lithic technology 4.2.1. Cores The 69 cores recovered from the sequence constitute 3% of all artifacts (7% if chips and debris are excluded). The frequencies of the various core types do not show a directional change across the stratigraphic sequence (Table 5). Thirty-two (46%) of all cores are on obsidian and 31 (45%) are on chert. Levallois flaking is one of the main core reduction strategies observed at Goda Buticha, consistent with our knowledge of MSA technological practices. Levallois cores (n ¼ 34, Fig. 5:1e3), including 9 flakes used as cores due to their suitable geometry (e. g., Delagnes, 1995; Hovers, 2007, Fig. 5:3), occur throughout the whole sequence. In about half of the cases, the Levallois cores bear some residual cortex, even though they tend to be small (average length is 32 mm). Half of the cores show centripetal preparation for preferential flake removals and the remainder were exploited using a unipolar recurrent flaking method. Another quantitatively important group consists of cores with several flaking surfaces (n ¼ 30), which are by definition (Boëda, 1995) non-Levallois. It is difficult to identify the flaking methods involved in their reduction. The majority of cores in this group are very small. They are highly fragmented or heavily exploited (Fig. 5). In a few cases it was possible to identify that a slightly convex flaking surface was used for recurrent removals (i.e., discoidal), which then was turned into a striking platform for exploiting an opposite, convex surface. Single platform (prismatic and pyramidal) cores, accounting for 7.2% of the cores (5/69), are more suggestive of Upper Paleolithic/ LSA technologies. This is corroborated by the presence of crested blades in the assemblage (Fig. 5:6; see below). Some of the prismatic cores were semi-rotated in order to produce blades and bladelets (Fig. 5:4). Pyramidal cores are small (<20 mm) and were used for bladelet production from a flat prepared striking platform. N Total The raw materials used throughout the sequence are dominated by chert (51%), obsidian (30%), basalt (9%), and quartzite (6%). The remainder consists of quartz and limestone (Table 4). Only a small number of artifacts (n ¼ 30) show extensive cortical or natural surfaces. As a result it is impossible to assess whether raw materials derive from secondary (e.g., transported; river bed cobbles) or primary (i.e., in situ geological exposures) sources. Chert and sandstone-quartzite in Buticha Cave are most likely of local origins. The cave is situated within chert-bearing limestone beds, and the sandstone-quartzite is probably derived from the conglomerate layers which cap the karstic limestone in the region (Fig. 4). The Gara Adi silicic volcanic dome, located only about 15e 20 km to the north, is the closest possible primary source for the obsidian. All other known potential sources are more than 100 km away (Negash and Shackley, 2006; Negash et al., 2011; Vogel et al., 2006); more accurate sourcing awaits geochemical analyses. 1 11 137 54 18 101 281 260 180 176 270 157 204 152 148 264 24 44 14 6 2502 4.1. Raw materials 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 D. Pleurdeau et al. / Quaternary International 343 (2014) 117e135 Author's personal copy 124 Table 5 Distribution of cores and blank types larger than 20 mm. The part of the sequence with the main concentration of lithics is shaded. a Total Flakes N 190e200 180e200 170e180 160e170 150e160 140e150 130e140 120e130 110e120 100e110 90e100 80e90 70e80 60e70 50e60 40e50 30e40 20e30 10e20 0e10 Total Blanksa Cores 6 3 1 14 8 6 6 6 4 4 1 4 5 1 69 % N 1 3 23 14 Blades % 0 0 14 15 0 2 9 8 8 7 6 6 7 2 10 8 0 5 0 37 72 42 32 34 52 40 34 35 29 48 4 13 7 100 75 52 70 0 66 47 42 45 40 52 60 60 71 71 77 100 68 70 7 520 55 N Bladelets % 3 2 0 25 2 10 100 18 18 19 28 16 20 18 11 12 0 3 0 16 20 147 16 1 1 2 2 10 27 19 20 14 20 12 6 6 2 N Points % 1 1 0 0 32 5 0 14 25 27 14 33 18 10 18 0 10 8 0 5 10 172 18 14 1 8 38 27 10 28 18 7 10 4 5 N 1 1 3 1 3 2 5 1 2 19 Indeterminates % N 0 0 0 0 0 0 1 1 4 1 3 3 0 10 2 3 0 0 0 2 % 1 4 2 1 2 3 2 3 1 19 N % 0 0 0 0 0 0 1 4 0 2 1 3 5 4 7 0 0 5 0 1 4 44 20 2 56 153 101 71 85 100 67 57 49 41 62 4 19 10 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 2 946 100 Refers here to unretouched artifacts and to the blank types of retouched items. Indeterminate blanks correspond to he majority of unifacially- and bifacially-retouched points, (see text for details). D. Pleurdeau et al. / Quaternary International 343 (2014) 117e135 Elevation above datum (cm) Author's personal copy D. Pleurdeau et al. / Quaternary International 343 (2014) 117e135 Fig. 5. Cores. 1 e Levallois core; 2, 5 e Fragmented and exploited cores (technology unclear); 3, 4 e Single Platform cores. Fig. 6. Elongated pieces, blades and bladelets. 125 Author's personal copy 126 D. Pleurdeau et al. / Quaternary International 343 (2014) 117e135 suggesting that elongated blanks were preferentially selected for secondary modification. The frequency of backed microliths, often used as diagnostic markers of the LSA, increases in the upper levels of the site (Fig. 9) compared to the lower Complex (n ¼ 1 in Layer IId). The fluctuating frequencies of retouched points do not show clear temporal trend. On the whole there are no clear differences in the typotechnological composition throughout the section. MSA elements (Levallois products, large blades, retouched points) cooccur or are interstratified with bladelets and backed microliths (Fig. 9). size. Frequencies of small (<20 mm) bladelets increase along the Lower Complex. This trend is observed for all the raw materials, but most notably in obsidian. The proportion of obsidian bladelets increases to 65% of the unretouched débitage in the 130e140 level. These products are associated with recurrent unipolar flaking, from either a single débitage surface (Levallois sensu lato) or from semirotated cores. Several crested blades/bladelets and overshot flakes and blades (Fig. 6:1, Fig. 7:12) first appear in Layer IIc. The overshot pieces tend to be laterally convex with uni- or bipolar dorsal face scars indicating the removal of series of parallel-sided bladelets, yet the guide-ridges on the dorsal faces and the lateral edges of the blades do not follow a clearly distinct flaking, reflecting lack of standardization of flaking methods. These observations are consistent with the low frequencies of Upper Paleolithic/LSA-type cores, the flaking of which is more standardized (see above). Levallois products (n ¼ 40), recognizable by their shape (often oval), thinness, butt preparation and the organization of dorsal face scars, appear in higher frequencies in the lower part of the sequence (Layers IId and IIe), but are also present in Layers IIaeIIc, which are dated to the mid-Holocene. These elements consist mainly of flakes (n ¼ 28), blades (n ¼ 2) and points (n ¼ 5; Fig. 7). Levallois flakes bear mostly centripetal and perpendicular scar patterns (on 20 flakes and 2 blades), some showing evidence of unipolar (n ¼ 4) and bipolar (n ¼ 4) recurrent flaking methods (Fig. 7:7e12). In addition, éclats débordants and outrepassants (i.e., intentional core maintenance elements) indicate the rejuvenation of Levallois cores by the re-shaping of lateral and distal convexities (Fig. 7:9, 12). 5. Faunal remains Some 340 mammalian faunal remains were collected throughout the sequence at Goda Buticha. Of these, ca. 10% are dental samples, another 10% are long bone shaft fragments, and the rest are unidentifiable bone fragments. Many of the remains are in friable condition, with poor surface preservation, which prevented the identification of anthropogenic or carnivore modifications. Therefore, analysis of the currently available restricted sample focused on taxonomic identifications of sufficiently preserved elements. Future studies of an augmented sample will focus on the quantitative aspects of the faunal assemblage. A diversity of taxa has been identified, including Procaviidae, Lagomorpha, Equus, Phacochoerus, small carnivore, and several types of bovids including bovini, antilopini, Redunca, and Tragelaphus. No domesticated fauna (e.g., sheep and goat) were recovered from any stratigraphic level, including the Upper Complex. The small sample of identifiable fauna from Goda Buticha contributes to our understanding of the local paleoenvironmental record, as it contains some open woodland indicators and suggests ecotone conditions including a source of permanent water that could support edaphic grasslands. The majority of taxa listed above are represented by one or two 4.2.3. Retouched artifacts The retouched pieces (n ¼ 141) constitute w5% or w14% of the assemblage (inclusive and exclusive of debris, respectively). These include 8 retouched cores and 10 micro-flakes. Obsidian was the preferred raw material for retouch (Table 6; Fig. 8). Table 6 Distribution of raw materials by retouched types. Flint Quartzite Basalt Obsidian Other Volcanic rocks Total Side-scraper Notch & denticulate Backed microlith 16 2 4 27 1 1 8 1 1 49 (34.8)a 4 (2.8)a Point End-scraper Burin Truncation Flat retouch Retouched flakes/blades 13 5 1 5 19 8 1 3 2 2 6 15 21 (14.9)a 30 (21.3)a 1 (0.7)a 3 (2.1)a 2 (1.4)a 2 (1.4)a 24 (20.6)a Total N % 38 4 15 83 1 26 3 8 62 1 141 100.0 2 c ¼ 50.8881, p-value ¼ 0.01827. a Absolute frequencies and percents (in parentheses) out of the total number of tools. The frequent formal tool categories are: side-scrapers (n ¼ 50), usually with fine retouch; unifacial (n ¼ 19) and bifacial (n ¼ 11) points, with one or two modified edges; and backed microliths (n ¼ 21), with typically curved backed edges. The rest of the tools are retouched flakes/blades exhibiting continuous retouch of various styles. Some 88% of the tools occur Layers IIc and IId, in elevations of 60e140 cm above datum (Fig. 9). Besides the backed microliths (see also Leplongeon, 2014), which are always made on bladelets, elongated blanks (i.e., blades or bladelets) were used in another 38 cases. These counts do not include the blanks for the retouched points, which in most cases could not be identified due to extensive retouch. These frequencies differ from those observed in the unretouched débitage (Table 5), dental samples, the reduncines being the only exception with a slightly higher representation in the middle-lower sections. This sample is too small to assess the subsistence activities of the occupants of the site. Nearly one hundred fragments of ostrich eggshell (some with a maximum dimension of 4e5 cm) occur throughout the sediment column. Notably, their frequencies decline with depth. Similarly, micro-faunal remains, sometimes heavily burned, appear throughout the sequence, yet their relative frequencies as well as proportion of burned specimens decrease with depth (see above). At 197e194 cm above datum microfauna remains were densely concentrated in a distinctive lens. A detailed analysis of this sample is currently underway and will be published elsewhere. Author's personal copy D. Pleurdeau et al. / Quaternary International 343 (2014) 117e135 127 Fig. 7. Levallois points and flakes. 6. Hominin remains The excavations in 2008 and 2011 at Buticha Cave yielded ten hominin specimens, which derive from the Lower Complex (Fig. 3). One specimen, a fragment of the cranial vault, derives from layer IId and therefore dates to at least 35 ka cal BP. It may be as old as 42 ka, given the date obtained on a charcoal sample from this elevation in an adjacent grid square (see Section 3), whereas the other nine pieces originated from the Holocene deposits. The hominins are represented by five permanent teeth or tooth fragments, two small fragments of cranial vault, a proximal pollical phalanx, two proximal pedal phalanges, and a metacarpal diaphysis (Fig. 10). In the descriptions that follow, we selected a variety of comparative samples. The teeth are compared to two samples from Nubia, the Epipaleolithic (Jebel Sahaba and Wadi Halfa) and agricultural samples published by Calcagno (1989). The Nubian dentitions show a marked reduction across the transition to agriculture. Additional comparisons were made to Jacobson’s (1982) data on Author's personal copy 128 D. Pleurdeau et al. / Quaternary International 343 (2014) 117e135 Fig. 8. Retouched artifacts: 1, 3 e bifacial points; 2 e unifacial point; 4 e retouched crested blade; 5 e end-scraper on a retouched flake; 6 e thinned side-scraper; 7 e denticulate. the teeth of native South Africans (as an example of a generalized, modern sub-Saharan sample) and reports from the literature on dentition of MSA hominins (most of whom had large teeth relative to living Africans). The postcranial remains are compared to data from Pearson (1997), who measured samples of skeletons of recent humans from around the world (including four African populations). Tooth 4 (BC-188): a lower left permanent canine recovered in 2011 from sediments from square B1d at 146 cm above datum (Layer IIc). Crown dimensions: 5.9 mm bucco-lingual (BL) by 6.3 mm mesio-distal (MD). These dimensions are quite small BL and fairly small MD compared to Calcagno’s (1989) Nubians. Tooth 1: a left I1 recovered in 2008 from square A1a from sediments between 130 and 127 cm (Layer IIc) above datum. Crown dimensions: 7.1 mm (BL) by 9.6 mm (MD). These correspond to a fairly wide MD width but unremarkable BL length. Proximal Pollical Fragment: A small, slender proximal left pollical phalanx recovered in 2008 from sediments from square A1a at 127 to 120 cm above datum (Layer IIc). Measurements: maximum length: 26.3 mm; maximum breadth of the proximal end: 10.2 mm. The specimen fits within many modern populations (Pearson, 1997), but indicates a small (female?) individual. Metacarpal Diaphysis (BC-510): Metacarpal diaphysis, probably a left metacarpal IV, excavated in 2011 from square A1b at 127 cm above datum (Layer IIc). The fragment is 37.5 mm long; dimensions at its approximate midshaft are 6.7 mm dorso-volar by 6.1 mm medio-lateral. Tooth 2: a left M3 recovered in 2008 from square B1a 123 cm above datum (Layer IIC). Crown dimensions: 10.3 mm BL by 9.7 mm MD. These dimensions are small compared to all of Calcagno’s (1989) samples. Tooth 5 (BC-541): Fragment of a left upper canine recovered in 2011 from sediments from square A1b at 121 cm above datum (Layer IIc). The crown is too fragmentary to measure its BL or MD dimensions. Cranial Vault Fragment 2 (BC-394): Parietal fragment, recovered in 2011 from sediments from square B1b 119 cm above datum (Layer IIc). Maximum dimensions: 34.5  19.3 mm. The fragment is thin, ranging between 5.4 and 5.1 mm, and bears traces of the sagittal suture. These facts suggest it came from a late adolescent or young adult. Author's personal copy D. Pleurdeau et al. / Quaternary International 343 (2014) 117e135 129 Fig. 9. Diachronic change of retouched pieces. Tooth 3 (BC-UN-25): a heavily worn right P4 recovered in 2011 from sediments from square B1d from 116 cm above datum (Layer IIc). The crown measures 8.1 mm BL by 7.1 mm MD. Relative to Calcagno’s (1989) samples of Nubians, these dimensions are somewhat small MD and quite small BL. Proximal Pedal Phalanx 1 (BC-500): A complete right proximal pedal phalanx, likely of ray III, excavated in 2011 from square A1b 130 cm above datum (Layer IIc). The specimen’s maximum length is 23.7 mm. Proximal Pedal Phalanx 2: A complete left proximal pedal phalanx, likely from ray IV, recovered in 2012 from sediments excavated and sieved in 2011 from square A1d at 120e130 cm above datum (Layer IIc). The maximum length of the phalanx is 19.3 mm. Fig. 10. Hominin remains from Goda Buticha A) Tooth 1 (unnumbered); B) Tooth 2 (unnumbered); C) Tooth 3 (BC-UN-25); D) Tooth 4 (BC-188); E) Tooth 5 (BC-541); F) Cranial vault fragment 1 (unnumbered); G) Cranial vault fragment 2 (BC-394); H) Proximal pollical phalanx 1 (unnumbered); I) Proximal pedal phalanx 1 (BC-500) [proximal pedal phalanx 2 is not shown]; J) Metacarpal shaft (BC-510). All the teeth are scaled to the size of the upper scale bar; the fragments of cranial vault and postcranial bones are scaled to the size of the lower scale bar. Author's personal copy 130 D. Pleurdeau et al. / Quaternary International 343 (2014) 117e135 Cranial Vault Fragment 1: Fragment of cranial vault lacking the internal table of bone, recovered in 2008 from sieved bones and lithics from sediments from square A1a 50 to 40 cm above datum (Layer IId). The maximum dimensions of the specimen are 32.6  23.4 mm, with a maximum preserved thickness of 6.2 mm. With one exception of the BL breadth of tooth 1 (the left I1), the teeth are small in size compared to South Africans from the Dart Collection (Jacobson, 1982) as well as agricultural or Epipaleolithic Nubians (Calcagno, 1989). Teeth from MSA contexts at other sites in Africa tend to be large relative to these comparative samples (Grine et al., 1991, 2000; Rightmire and Deacon, 1991, 2001; Grine and Henshilwood, 2002; White et al., 2003; Verna et al., 2013), although a small number of notably small-toothed individuals are also known (Singer and Wymer, 1982; Bräuer and Mehlman, 1988; McCrossin, 1992; Lam et al., 1996; Royer et al., 2009). The postcranial specimens are average to small in size relative to those of recent Africans. Hominin postcranial fragments from the MSA (for which the most numerous sample comes from South Africa, followed now by Ethiopia (Pearson et al., 2008a, 2008b) display a large amount of variation in size, from medium- to small-sized individuals (e.g., the first metatarsal from Klasies River (Rightmire et al., 2006) and the new pedal phalanges from the Howiesons Poort or late MSA from Diepkloof (Verna et al., 2013)) to large ones (e.g., the second metatarsal from Klasies River main site (Rightmire et al., 2006) and the femur from Blind River (Wang et al., 2008). The medium to small sizes of the postcranial remains from Buticha fit within this broad spectrum of variability. 7. Beads Five ostrich eggshell (OES) beads were found in the upper part of Complex II (Layer IIc) (Fig. 11). Some beads are well-executed and finished, with a perfectly circular perforation and perimeter, when others are made from thicker eggshell, with roughly shaped edges and irregular central perforations. The differences in workmanship can be attributed to the beads being discarded during different stage of manufacture (for steps of OES bead manufacturing, see Orton, 2008; also Kandel and Conard 2005). 8. Discussion 8.1. Goda Buticha in its immediate geographic context Dating the dense sequences in cave sites in eastern Africa has been far from a straightforward matter. Goda Buticha is no exception. Nine stratigraphically consistent radiocarbon dates from Layer IIc, which contains the richest archaeological remains, place this occupation within a short time range between 6940  40 14C BP (7797  118 cal BP) and 5590  50 14C BP (6380  88 cal BP). The archaeological material from Layer IIc has a clear LSA aspect, but includes also some MSA elements. The two younger dates from the top of Layer IId indicate human presence in the cave around 34e35,000 years ago (Table 1). These dates represent a terminal age for the associated MSA cultural finds, while dates from Layer IIf indicate MSA occurrences around 42e46,000 ka cal BP. Since MSA deposits extend down to bedrock (7 cm below datum, i.e., w80 cm below the dated level), earlier dates may still be obtained from the basal section. The taxonomic composition of large mammals at Goda Buticha suggests a relatively closed and wet habitat around the cave in the late Upper Pleistocene. Within the small faunal sample, Redunca is the most common taxon throughout the sequence. Independent of the type of transporting agent (i.e., human or carnivore), this suggests the presence in the cave’s vicinity of a perennial river (as indeed is the case today) or any other permanent source of water that can support moist edaphic grasslands. The presence of Equus (mostly teeth) and Tragelaphus in different levels suggests a nearby presence of an open woodland environment. Together, these lines of evidence point to ecotone conditions that are preferable to certain large mammals, including large bovines. Our understanding of environmental conditions during the occupations of Goda Buticha stands to benefit from speleothem studies in nearby cave sites (Assefa et al., 2014). These emerging paleoclimatic archives are also useful for understanding the complex paleoclimatic record in eastern Africa (e.g., Anyah and Semazzi, 2007; Blome et al., 2012). In the larger Horn of Africa region, the time around w50 ka marks a major wet event that continued until 31 ka (Gasse and Street, 1978; Gasse et al., 1980), followed by generally dry conditions which persisted till the rapid onset of the African Humid Period at about 11.6 ka cal BP (Foerster et al., 2012). Humid conditions persisted during the early Holocene until they gradually changed to dry conditions starting at about 5 ka cal BP. Ongoing investigations on the speleothems from the Mechara area (w170 km SW of Goda Buticha) are consistent with this trend, showing warm and wet conditions in the SE Ethiopian region during the Early to Middle Holocene followed by dry conditions during the Late Holocene (Asrat et al., 2007, 2008; Baker et al., 2010). Wetter conditions inferred from the sedimentological and faunal records of Goda Buticha are generally consistent with such climatic shifts in the region. Goda Buticha was located in an ecological ecotone on the lower flanks of the escarpment, at the relatively low elevation of 1382 m asl. This raises the hypothesis that the exploitation of lowelevation ecotones in the Late MSA was made possible through the inferred climatic amelioration represented in the site’s record, similar to the argument presented by Ambrose (1986) regarding Eburran sites during the Holocene. Within the site itself, the contact between Complex I and Complex II (w180 cm above datum; Fig. 3) represents a sedimentological boundary separating anthropogenic ashes from geogenic silts. At 170e160 cm above datum, a horizon of brown silt contained LSA archaeological material together with MSA components. The main raw material represented is obsidian, which was used to make small tools, including microliths with fine and backed retouch (Leplongeon, 2014). Obsidian cores are correspondingly small. At least some lithic reduction was carried out on site, as evidenced by pieces of core trimming elements (e.g., éclats outrépassants) indicative of core rejuvenation. Notably, in the upper part of Complex II (150e110 above datum) small bladelet cores, made on chert, begin to appear, associated with typical unifacial and bifacial MSA points. The sedimentary fine-grained matrix of Complex II is relatively homogeneous and was deposited through a single depositional regime, yet the radiocarbon dates point to a chronological hiatus between Upper Pleistocene (IId) and mid-Holocene (IIc) deposition (Fig. 3, Table 1). This hiatus, expressed granulometrically by a higher frequency of large clasts in the matrix of Layer IIc, is confirmed by preliminary results of ongoing OSL dating (Tribolo, personal communication). The cultural material in the lower part of Complex II (IId) is dominated by MSA lithic artifacts with few LSA components. At w70e60 cm above datum and down to bedrock, the small component disappears almost entirely, obsidian becomes rare, and the artifacts tend to be large (4e5 cm), clearly Levallois and almost without retouch. Thus the main tendencies observed in the lithic assemblages throughout the Goda Buticha sequence can be summarized as follows: —Artifacts in the sequence of Goda Buticha are generally small, and items longer than 60 mm are absent. The artifacts Author's personal copy D. Pleurdeau et al. / Quaternary International 343 (2014) 117e135 131 Fig. 11. Three beads from the top of the Lower Complex (Layer IIc). The upper and lower rows show the outer and inner surfaces of the beads, respectively. manufactured on basalt and quartzite, which are available locally as large nodules/cobbles, are the largest ones in the assemblage. —The frequency of obsidian generally declines with depth, although some types of MSA points are made on this raw material. —Obsidian cores appear to have been more intensively exploited (average length of ca. 21 mm compared to 40 mm for chert cores). Additionally, obsidian was the preferred raw material for retouch (62% of the retouched artifacts are made on obsidian, vs. only 30% of the total number of artifacts made of this raw material). These two traits are consistent with possible longdistance transport of obsidian as a preferred raw material (see above). Ostrich eggshell (OES) beads from Goda Buticha can now be added to the finds known from only a handful of sites in eastern Africa. Such beads constitute an important element in the context of identifying cultural changes in the Late Pleistocene in the Horn of Africa. At the rock shelter of Enkapune Ya Muto in Kenya OES beads (Ambrose, 1998b: fig. 3) were dated directly or by associated charcoal samples to w40 ka. At Kisese II rockshelter (Tanzania), OES beads are associated with a transitional MSA/LSA industry (Inskeep, 1962) and were directly dated to 31,480 BP (Deacon, 1966). In Mumba Cave (Tanzania) OES beads have been assigned to the LSA with dates of w49 ka (Mehlman, 1989, 1991; Diez-Martín et al., 2009; Gliganic et al., 2012). Direct AMS radiocarbon dating of these beads has yielded multiple and slightly younger dates, between 29 ka and 33 ka (Conard 2004, Weiß, 2000, both cited in Conrad 2005). In Goda Buticha OES beads derive from sediments with a Holocene age, yet they are associated with both LSA artifacts as well as stone tools with MSA affinities. Two alternative hypotheses are considered to explain the change in the relative proportions of MSA and LSA material throughout the Goda Buticha sequence. One explanation implicates natural formation processes and post-depositional mixture of sediments and of MSA and LSA elements, thus creating a ‘phantom’ of a transitional cultural phase. However, no major disturbances were observed macroscopically. We remain careful regarding the occurrence of MSA features in Layer IIc, waiting for formal sedimentological and chronological analyses that will speak to stratigraphic integrity and the potential effects of post-depositional processes (including artifact migration through the sequence). Nevertheless, our macroscopic observations, as well as the radiocarbon chronology, lead us to suspect that such effects, if existent, have been very localized and that the archaeological elements of Layer IIc (as well as the others) derive from an in situ context. An alternative hypothesis is that the sequence of Goda Buticha represents a facet of the material culture variability that is inherent to the late MSA and the LSA in the southeastern Ethiopian region. This hypothesis cannot be evaluated based on the record of a single site. Porc Epic Cave, where a late MSA cultural sequence has been described in detail (Clark and Williamson, 1984; Pleurdeau, 2003, 2004, 2005a, 2005b; Assefa, 2006; Assefa, 2006; Assefa et al., 2008), provides an opportunity to compare the two cultural sequences on a regional scale, as only 30 km separate the two sites. In both sites the same raw materials were used. Obsidian was used more frequently in Goda Buticha (30% vs. 8% in Porc Epic), and basalt is more common in Porc Epic (17% vs. only 9% in Goda Buticha) even though it can be found overlying the sedimentary succession in the region. Such differences suggest different raw material-specific strategies of transport and use within and between localities. In both sites ‘provisioning of place’ (as defined by Kuhn, 1995) seems to have taken place, with tested blocks, nodules or partly decorticated cores imported into the sites, where they were further reduced (as attested by the various types of core trimming elements) to near-exhaustion (average core size in Porc Epic is 40 mm vs. <30 mm in Goda Buticha). Obsidian was preferentially used for the production of bladelets at Porc Epic, while at Buticha it was the preferred raw material for all types of retouched artifacts (Table 6). Technologically, both sites demonstrate the presence of the same four technological schemes in similar proportion. These include Levallois flaking methods (centripetal for preferential flakes and unipolar recurrent), and very low frequencies of semirotated and full blade technologies. Throughout the Porc Epic MSA sequence there are common technological (Levallois and discoid schemes) and typological (e.g., retouched points) features. With the exception of the lower unit at the site, these are associated with elements which are argued to be “typical” LSA characteristics such as backed bladelets, without any significant directional changes throughout the sequence. In the absence of evidence for Author's personal copy 132 D. Pleurdeau et al. / Quaternary International 343 (2014) 117e135 Fig. 12. Location map of MSA and LSA sites in the Horn of Africa mentioned in the text. (image modified from NASA Visible Earth) significant bioturbation or mixing it was argued that MSA and LSA characteristics coexisted in the late MSA assemblages of Porc Epic (Pleurdeau 2005b; but see Leplongeon, 2013, 2014). The relatively small sample size that is currently available for Goda Buticha precludes a robust identification of temporal trends throughout the sequence. Nevertheless, the densest part of the sequence seems to show a similar picture. A change was recorded only at the top of Complex II, dated to the Holocene, where retouched points disappear while microliths and backed bladelets persist. Both Porc Epic and Goda Buticha date to the late Upper Pleistocene. Age estimates for Porc Epic are controversial. They range from MIS 4 (61,202  958, 61,640  1083, and 77,565  1575 a) using obsidian hydration (Michels and Marean, 1984) to radiocarbon determinations >43,200 33,700  00 BP and 35,600  350 BP on opercula from various levels of the site (Assefa et al., 2008). The latter dates are comparable to those from unit IId of Goda Buticha. However, direct gamma-spectrometry dating of the hominin mandible, discovered in 1933, yielded a date of w50 ka (Y. Yokoyama and C. Falgueres, pers. comm.). The local record represented by these two sites may be linked on the basis of its microlithic and laminar products to comparable sites at the regional scale in the Horn of Africa. 8.2. A broader perspective from the Horn of Africa Only a handful of Upper Pleistoceneeearly Holocene sequences in the Horn of Africa provide relevant regional and chronological comparisons to the sequence of Goda Buticha (Fig. 12). Among these occurrences are the open-air sites of Aladi Springs (Clark and Williams, 1978; Gossa et al., 2012), K’one (Kurashina, 1978) and Bulbula in Ziway-Shala area (Ménard et al., submitted) in Ethiopia, as well as MSA and LSA sites from Buri Peninsula in Eritrea (Beyin, 2011a, 2013); the cave of Midhishi 2 (Brandt and Brook, 1984; Gresham, 1984) and the rock shelter of Laas Geel 7, both in Somaliland (Diaz, 2009, and references therein; Gütherz et al., submitted); and the cave site of Mochena Borago (Brandt et al., 2012, and references therein) in Ethiopia. With the exception of Mochena Borago, chronological comparisons of these sequences with Goda Buticha are tentative given that the relevant radiocarbon dates were obtained prior to the 1980s, or due to excavators’ concerns regarding the stratigraphic integrity of the sites. The lithic assemblages of these sites have been assigned to different archaeological taxonomies. K’one is perceived as an MSA assemblage, but the (potentially contemporaneous) sequence of Mochena Borago, dated 54e43 ka, was not assigned to either a MSA, LSA or a transitional entity due to the absence of marked changes throughout the sequence (Brandt et al., 2012). Aladi Springs and Midhishi 2 (stratigraphic unit CSUb) are believed to be transitional MSA/LSA assemblages, whereas the lower complex at Laas Geel 7 is assigned to the Hargeisian (as defined by Clark, 1954; Gütherz et al., submitted). The sites of Lake Besaka which contain the “Ethiopian Blade Tool Tradition” (Brandt, 1982, 1986) post-date 19,000 BP, as radiocarbon measurements on OES (22,080  305, 19,460  205, and 19,280  215) derive from a culturally sterile layer underlying the archaeological material (Brandt, 1986:63). The high degree of variability within these lithic assemblages clearly has contributed to taxonomic confusion or even contradictions. It is important therefore to try and parse the sources of such variability. While a detailed discussion of each assemblage is unwarranted here, some chronological (if the dates are accepted) as well as typo-technological patterns seem to emerge in the context of this broader overview. Author's personal copy D. Pleurdeau et al. / Quaternary International 343 (2014) 117e135 Typo-technologically, all the sites exhibit the use of Levallois flaking methods in Pleistocene horizons, combined with a continuous presence of retouched points, particularly in K’one (Kurashina, 1978) and Midhishi 2 (Brandt and Brook, 1984). Midhishi 2 is exceptional among these sites in that microliths and backed bladelets are not present in the older Pleistocene-age deposits and only occur in CSUb (Brandt and Brook, 1984). Furthermore, Gossa et al. (2012) note a diachronic trend of artifact size diminution in Aladi Springs, similar to Goda Buticha. In one case, at Gobdera, were typical (though not exclusive) MSA elements (centripetal cores and facetted platforms) recognized in an assemblage defined as LSA (Pillipson, 1977). The largest and best described assemblage from K’one (K’one 5), 200 km west from Goda Buticha, provided a single radiocarbon date of 14,670  200 BP, but Kurashina (1978:359) suggests an age between 40 and 30 ka based on stratigraphic extrapolation with studied geological sections. This assemblage is not compared easily with any of the others mentioned above, since it contains only 0.7% retouched tools, the rest being debitage and cores (Kurashina, 1978:323). Kurashina identified the use of Levallois flaking systems, including an important presence of the type I Nubian technology that has recently been promoted as a cultural marker of Late Pleistocene out of Africa dispersals (Beyin, 2013; Van Peer, 1998; Rose et al., 2011; Crassard and Hilbert, 2013). There are no indications for microlithic elements in this assemblage. Further away in the Horn of Africa, the best dated and reported site is the cave of Mochena Borago (Brandt et al., 2012). Several stratigraphic sequences were described and dated independently due to the complex site stratigraphy. These indicate occupations dated to 53,224  2662 cal BP (lower T-group), 45,164  982 cal BP, 43,480  443 cal BP, and 41,159  783 cal BP (upper T-group, Sgroup and R-group, respectively). The overlaying occupational episode dates to the early Holocene (7589  689 cal BP). Thus Mochena Borago, despite its distance from the eastern part of the Horn, depicts the same temporal gap in human occupation. Brandt et al. (2012) note the close typo-technological similarities across the T- and S-group assemblages. Flaking systems consist of various Levallois methods inclusive of the Nubian technology in the upper T-group assemblage. The majority of cores belonged to the group dubbed “SDM” (single-double-multiple platform cores), but their frequencies decline in the S-group assemblage concurrent with the emergence of bipolar technology. The latter is comparable to the broadly contemporaneous assemblage from layer V at Mumba Cave in Tanzania (Diaz-Martín et al., 2009; Gilganic et al., 2012; Eren et al., 2013). Frequencies of backed pieces increase through time since their first appearance in upper T-group assemblages at the expense of retouched points, resulting in an increase through time in richness and evenness of the typological classes of retouched artifacts (including burins, drills/awls, notches, non-standardized unifacial and fewer bifacial points). It appears that part of the variability of the lithic assemblages discussed here derives from situational and specific aspects of the assemblages, related to raw material availability, mobility patterns and site functions. The prolific use of obsidian in K’one, Aladi Springs and Mochena Borago reflects their proximity to the sources of raw material, but assemblage compositions (i.e., frequencies of debitage, cores, CTE and retouched items) differ according to site functions. Whereas K’one may be an obsidian knapping station, Mochena Borago displays the full range of lithic reduction and secondary modification. At Porc Epic, Goda Buticha and Laas Geel, on the other hand, obsidian is one of several raw materials, and is used preferentially for some types of artifacts. Site provisioning strategies also appear to differ according to the distances of a given site from the various raw material sources. Specifically, Porc Epic 133 and Goda Buticha reflect site provisioning and on-site knapping independently of specific raw materials. Other aspects of the lithic assemblages seem to present broad similarities, among which is the co-occurrence of Levallois with semi-rotated and single platform cores early on in MIS 3 assemblages and through MIS 2, as well as the general trend (rather than sudden appearance) of microlithization and the occurrence of backed items in tandem with MSA points. The early dates assigned recently to OES beads from Enkapune Ya Moto and Mumba, which led to the definition of these assemblages as LSA, further confound the issue. The shift from the Middle to the Later Stone Age remains a poorly recognized process both in terms of its temporal extent and of its telltale material culture markers. 9. Concluding remarks In the context of the paleoenvironmental records from the Horn of Africa, the faunal remains from Goda Buticha suggest that the occupations of the site took place during relatively wetter periods. The occupational hiatus between 35 ka and 8e6 ka corresponds to a dry period in this region (Umer et al., 2004). A similar occupational gap seems to have occurred in the majority of sites from this time span, without evidence for human presence between w40 ka and the late/terminal Upper Pleistocene (18,790  340 for CSUb in Midhishi 2, 11,079  160 BP in Aladi Springs), or early Holocene (the upper complex at the rock shelter 7 of Laas Geel (Gütherz et al., submitted) as well as Mochena Borago). The sequence from Goda Buticha presents a rare case in which MSA elements are associated with LSA materials dated to the Holocene. Based on current data, site formation disturbances in Goda Buticha seem to have been localized at most. Therefore, the pattern observed in the site’s sequence needs to be considered when interpreting the shift from the MSA to LSA in the Horn of Africa. The limitations of the late Pleistocene record in the Horn of Africa and dating ambiguities of the available sites must be borne in mind. Having said that, we suggest that in view of the occupational gap that appears to temporally disjoint LSA and MSA assemblages in stratified sites, material culture similarities across this chronological gap may not represent homologies derived from a MIS 3 common ancestry or from different regional ‘anestral’ sites. Instead, the potential effect of historical and demographic processes of movement within, out of, and into Africa should be investigated as possible causes for the general similarities and regional differences during and after “The Dispersal” of modern humans from Africa. Within this framework, Goda Buticha is a key site for reevaluating the dynamics and tempo of the change from MSA to LSA lifeways in this region. Acknowledgments We thank the Ethiopian Authority for Research and Conservation of Cultural Heritages (ARCCH) for permission to survey in the area and to excavate in Goda Buticha. This research was supported by grants to ZA from the National Geographic Society (grants # 8110-06 and 8510-08) and to ZA and DP from the Wenner-Gren Foundation (Grant # ICRG e 102). We would like to acknowledge the efforts of Tilahun G/Selassie, Workalemahu Bekele, Hadis, DeJene Dendana Gulti, Alice, as well as the local people living near the site for their assistance with the excavation. We are also very thankful to Cécile Chapon, Simon Puaud, Marion Hernandez and Chantal Tribolo, for their current support for sedimentological and OSL dating analyses, and to Sébastien Nomade for providing some radiocarbon analysis facilities. We are grateful to the French Center for Ethiopian Studies for providing logistical support and for funding part of the post-excavation analyses. We thank the Tourism Author's personal copy 134 D. Pleurdeau et al. / Quaternary International 343 (2014) 117e135 and Culture Offices of the Eastern Harerghe, Western Harerghe, the Dire Dawa Administration, and Harari National Regional State for fieldwork administrative support. We are grateful to the reviewers of the paper for their very helpful suggestions and to Christian Tryon for discussions and for directing us to some important bibliographic references. References Ambrose, S.H., 1986. Hunteregatherers adaptations in non-marginal environments: an ecological and archaeological assessment of the Dorobo model. Sprache und Geschichte in Afrika. Herausgegeben an den Universitäten Köln and Bayreuth 7 (2), 11e42. Ambrose, S.H., 1998a. Late Pleistocene human population bottlenecks, volcanic winter, and differentiation of modern humans. Journal of Human Evolution 34, 623e651. Ambrose, S.H., 1998b. Chronology of the Later Stone Age and food production in East Africa. Journal of Archaeological Science 25, 377e392. Anyah, R.O., Semazzi, F.H., 2007. Variability of East African rainfall based on multiyear RegCM3 simulations. International Journal of Climatology 27, 357e 371. Asrat, A., Baker, A., Leng, M., Gunn, J., Umer, M., 2008. Environmental monitoring in the Mechara Caves, Southeastern Ethiopia: implications for speleothem paleoclimate studies. International Journal of Speleology 37, 207e220. Asrat, A., Baker, A., Mohammed, M.U., Leng, M.J., Calsteren, P.V., Smith, C., 2007. A high-resolution multi-proxy stalagmite record from Mechara, Southeastern Ethiopia: palaeohydrological implications for speleothem palaeoclimate reconstruction. Journal of Quaternary Science 22, 53e63. Assefa, Z., 2006. Faunal remains from Porc-Epic: paleoecological and zooarchaeological investigations from a Middle Stone Age site in southeastern Ethiopia. Journal of Human Evolution 51, 50e75. Assefa, Z., Pleurdeau, D., Hovers, E., Asrat, A., Pearson, O.M., Duquesnoy, F., Lam, Y.M., Tsion, C., 2014. Survey and explorations of caves in southeastern Ethiopia: MSA and LSA archaeology and rock art. Quaternary International. http://dx.doi.org/10.1016/j.quaint.2013.07.132. Assefa, Z., Yirga, S., Reed, K.E., 2008. The large-mammal fauna from the Kibish Formation. Journal of Human Evolution 55, 501e512. Baker, A., Asrat, A., Fairchild, I.J., Leng, M.J., Thomas, L., Widmann, M., Jex, C.N., Dong, B., Van Calsteren, P., Bryant, C., 2010. Decadal-scale rainfall variability in Ethiopia recorded in an annually laminated, Holocene-age, stalagmite. The Holocene 20, 827e836. Behar, D.M., Villems, R., Soodyall, H., Blue-Smith, J., Pereira, L., Metspalu, E., Scozzari, R., Makkan, H., Tzur, S., Comas, D., 2008. The dawn of human matrilineal diversity. The American Journal of Human Genetics 82, 1130e1140. Beyin, A., 2006. The Bab al Mandab vs the Nile-Levant: an appraisal of the two dispersal routes for early modern humans out of Africa. African Archaeological Review 23, 5e30. Beyin, A., 2011a. Earlyemiddle Holocene human adaptations on the Buri peninsula and Gulf of Zula, coastal lowland of Eritrea. Azania: Archaeological Research in Africa 46 (2), 123e140. Beyin, A., 2011b. Upper Pleistocene human dispersals out of Africa: a review of the current state of the debate. International Journal of Evolutionary Biology. Beyin, A., 2013. A surface Middle Stone Age assemblage from the Red Sea coast of Eritrea: implications for Upper Pleistocene human dispersals out of Africa. Quaternary International 300, 195e212. Blome, M.W., Cohen, A.S., Tryon, C.A., Brooks, A.S., Russell, J., 2012. The environmental context for the origins of modern human diversity: a synthesis of regional variability in African climate 150,000e30,000 years ago. Journal of Human Evolution 62, 563e592. Bosellini, A., Russo, A., Assefa, G., 2001. The Mesozoic succession of Dire Dawa, Harar Province, Ethiopia. Journal of African Earth Sciences 32, 403e417. Brandt, S.A., 1982. A Late Quaternary Cultural and Environmental Sequence from Lake Besaka, Southern Afar, Ethiopia (Ph.D. thesis). University of California, Berkeley. Brandt, S.A., 1986. The Upper Pleistocene and early Holocene prehistory of the Horn of Africa. African Archaeological Review 4, 41e82. Brandt, S.A., Fisher, E.C., Hildebrand, E.A., Vogelsang, R., Ambrose, S.H., Lesur, J.P., Wang, H., 2012. Early MIS 3 occupation of Mochena Borago Rockshelter, Southwest Ethiopian highlands: implications for Late Pleistocene archaeology, paleoenvironments and modern human dispersals. Quaternary International 274, 38e54. Brandt, S.A., Brook, G., 1984. Archaeological and paleoenvironmental research in northern Somalia. Current Anthropology 25 (1), 119e121. Bräuer, G., Mehlman, M.J., 1988. Hominid molars from a Middle Stone Age level at the Mumba rock shelter, Tanzania. American Journal of Physical Anthropology 75, 69e76. Breuil, H., 1934. Peintures rupestres préhistoriques du Harrar (Abyssinie). L’Anthropologie 44, 473e483. Breuil, H., Teilhard de Chardin, P., Wernert, P., 1951. Le paléolithique du Harrar. L’Anthropologie 55, 219e228. Bruggemann, J.H., Buffler, R.T., Guillaume, M.M., Walter, R.C., Von Cosel, R., Ghebretensae, B.N., Berhe, S.M., 2004. Stratigraphy, palaeoenvironments and model for the deposition of the Abdur Reef Limestone: context for an important archaeological site from the last interglacial on the Red Sea coast of Eritrea. Palaeogeography, Palaeoclimatology, Palaeoecology 203, 179e206. Calcagno, J.M., 1989. Mechanisms of Human Dental Reduction: a Case Study from Post-Pleistocene Nubia. Department of Anthropology, Univ of Kansas. Clark, J.D., 1954. The Prehistoric Cultures of the Horn of Africa. An Analysis of the Stone Age Culture and Climate Succession in the Somalilands and Eastern Parts of Abyssinia. Cambridge University Press, Cambridge. Clark, J.D., 1988. The Middle Stone Age of East Africa and the beginnings of regional identity. Journal of World Prehistory 2, 235e305. Clark, J.D., Beyene, Y., WoldeGabriel, G., Hart, W.K., Renne, P.R., Gilbert, H., Defleur, A., Suwa, G., Katoh, S., Ludwig, K.R., 2003. Stratigraphic, chronological and behavioural contexts of Pleistocene Homo sapiens from Middle Awash, Ethiopia. Nature 423, 747e752. Clark, J.D., Williams, M.A.J., 1978. Recent archaeological research in Southeastern Ethiopia. 1974e1975. Annales d’Ethiopie, 19e44. Clark, J.D., Williamson, K.D., 1984. A Middle Stone Age occupation site at Porc Epic Cave, Dire Dawa (east-central Ethiopia). African Archaeological Review 2, 37e71. Conard, N., 2005. An overview of the patterns of behavioural change in Africa and Eurasia during the Middle and Late Pleistocene. In: Backwell, L., d’Errico, F. (Eds.), From Tools to Symbols. Witwatersrand University Press, Johannesburg, pp. 294e332. Crassard, R., Hilbert, Y.H., 2013. A Nubian complex site from Central Arabia: implications for Levallois taxonomy and human dispersals during the Upper Pleistocene. PLoS ONE 8 (7), e69221. Deacon, J., 1966. An annotated list of radiocarbon dates for sub-Saharan Africa. Annals of the Cape Provincial Museum 5, 5e84. Delagnes, A., 1995. Variability Within Uniformity: Three Levels of Variability Within Levallois System. The Definition and Interpretation of Levallois Technology, pp. 201e212. Delagnes, A., Tribolo, C., Bertran, P., Brenet, M., Crassard, R., Jaubert, J., Khalidi, L., Mercier, N., Nomade, S., Peigné, S., 2012. Inland human settlement in southern Arabia 55,000 years ago. New evidence from the Wadi Surdud Middle Paleolithic site complex, western Yemen. Journal of Human Evolution 63, 452e474. Diez-Martín, F., Domínguez-Rodrigo, M., Sánchez, P., Mabulla, A.Z.P., Tarriño, A., Barba, R., Prendergast, M.E., Luque, L., 2009. The Middle to Later Stone Age technological transition in East Africa: new data from Mumba Rockshelter Bed V (Tanzania) and their implications for the origin of modern human behavior. Journal of African Archaeology 7, 147e173. Eren, M.I., Diez-Martin, F., Dominguez-Rodrigo, M., 2013. An empirical test of the relative frequency of bipolar reduction in Beds VI, V, and III at Mumba Rockshelter, Tanzania: implications for the East African Middle to Late Stone Age transition. Journal of Archaeological Science 40, 248e256. Foerster, V., Junginger, A., Langkamp, O., Gebru, T., Asrat, A., Umer, M., Lamb, H., Wennrich, V., Rethemeyer, J., Nowaczyk, N., 2012. Climatic change recorded in the sediments of the Chew Bahir basin, southern Ethiopia, during the last 45,000 years. Quaternary International 274, 25e37. Forster, P., Matsumura, S., 2005. Did early humans go north or south? Science 308, 965e966. Gasse, E., Street, F.A., 1978. Late Quaternary lake-level fluctuations and environments of the northern Rift Valley and Afar region (Ethiopia and Djibouti). Palaeogeography, Palaeoclimatology, Palaeoecology 24, 279e325. Gasse, F., Rognon, P., Street, F.A., 1980. Quaternary history of the Afar and Ethiopian Rift lakes. In: Williams, M.A.J., Faure, H. (Eds.), The Sahara and the Nile, pp. 361e400. Gliganic, L.A., Jacobs, Z., Roberts, R.G., Domínguez-Rodrigo, M., Mabulla, A.Z., 2012. New ages for Middle and Later Stone Age deposits at Mumba rockshelter, Tanzania: optically stimulated luminescence dating of quartz and feldspar grains. Journal of Human Evolution 62, 533e547. Gossa, T., Shale, Y., Negash, A., 2012. A reassessement of the Middle and Later Stone Age lithic assemblage from Aladi Springs, southern Afar Rift, Ethiopia. Azania 47 (2), 210e222. Gresham, T., 1984. An Investigation of an Upper Pleistocene Archaeological Site in Northern Somalia. Unpublished M.A. thesis. The University of Georgia, Athens, GA. Grine, F.E., Henshilwood, C.S., 2002. Additional human remains from Blombos Cave, South Africa:(1999e2000 excavations). Journal of Human Evolution 42, 293e 302. Grine, F.E., Henshilwood, C.S., Sealy, J.C., 2000. Human remains from Blombos Cave, South Africa:(1997e1998 excavations). Journal of Human Evolution 38, 755e 765. Grine, F.E., Klein, R.G., Volman, T.P., 1991. Dating, archaeology and human fossils from the Middle Stone Age levels of Die Kelders, South Africa. Journal of Human Evolution 21, 363e395. Gutherz, X., Diaz, A., Ménard, C., Bon, F., Douze, K., Léa, V., submitted. The Hargeisan revisited: lithic industries from shelter 7 of Laas Geel, Somaliland and the transition between the Middle and Late Stone Age in the Horn of Africa. Quaternary International. Hovers, E., 2007. The many faces of cores-on-flakes: a perspective from the Levantine Mousterian. In: McPherron, S.P. (Ed.), Cores or Tools? Alternative Approaches to Stone Tool Analysis, pp. 42e74. Inskeep, R., 1962. The age of the Kondoa rock paintings in light of recent excavations at Kisese II rock shelter. In: Mortelmans, G., Nenquin, J. (Eds.), Actes du IVe Congres Panafricain de Préhistoire de l’étude du Quaternaire, Sciences Humaines, 40, pp. 249e256. Tervuren. Author's personal copy D. Pleurdeau et al. / Quaternary International 343 (2014) 117e135 Jacobson, A., 1982. The Dentition of the South African Negro. Higginbotham Inc, Alabama. Kandel, A.W., Conard, N.J., 2005. Production sequences of ostrich eggshell beads and settlement dynamics in the Geelbek Dunes of the Western Cape, South Africa. Journal of Archaeological Science 32 (12), 1711e1721. Klein, R.G., 1995. Anatomy, behavior, and modern human origins. Journal of World Prehistory 9, 167e198. Klein, R.G., 2008. Out of Africa and the evolution of human behavior. Evolutionary Anthropology: Issues, News, and Reviews 17, 267e281. Kuhn, S.L., 1995. Mousterian Lithic Technology: an Ecological Perspective. Princeton University Press Princeton. Kurashina, J., 1978. An Examination of Prehistoric Lithic Technology in East-Central Africa, Ph. D. University MIcrofilm International, Ann Arbor. Kusimba, S.B., 2001. The Early Later Stone Age in East Africa: xcavations and lithic assemblages from Lukenya Hill. The African Archaeological Review 18, 77e123. Lahr, M.M., Foley, R.A., 1998. Towards a theory of modern human origins: geography, demography, and diversity in recent human evolution. Yearbook of Physical Anthropology 41, 137e176. Lam, Y.M., Pearson, O.M., Smith, C.M., 1996. Chin morphology and sexual dimorphism in the fossil hominid mandible sample from Klasies River Mouth. American Journal of Physical Anthropology 100, 545e557. Leplongeon, A., 2013. La Transition Middle Stone Age/Later Stone Age en Afrique de l’Est (Éthiopie). Unpublished Ph.D. thesis. Museum national d’histoire naturelle, Paris. Leplongeon, A., 2014. Microliths and the Middle to Later Stone Age transition: new data from Porc-Epic and Goda Buticha Cave Sites, Ethiopia. Quaternary International. http://dx.doi.org/10.1016/j.quaint.2013.12.002. Ménard, C., Bon, F., Dessie, A., Bruxelles, L., Douze, K., Fauvelle-Aymar, F.X., Khaldi, L., Lesur, J. and Mensan, R., submitted. Late Stone Age variability in the Main Ethiopian Rift: new data from the Bulbula river, Ziway-Shala basin. Quaternary International. Marean, C.W., Assefa, Z., 2005. The Middle and Upper Pleistocene African record for the biological and behavioral origins of modern humans. African Archaeology, 93e129. McBrearty, S., Brooks, A.S., 2000. The revolution that wasn’t: a new interpretation of the origin of modern human behavior. Journal of Human Evolution 39, 453e563. McCrossin, M.L., 1992. Human molars from Later Pleistocene deposits of Witkrans Cave, Gaap Escarpment, Kalahari margin. Human Evolution 7, 1e10. McDougall, I., Brown, F.H., Fleagle, J.G., 2005. Stratigraphic placement and age of modern humans from Kibish, Ethiopia. Nature 433, 733e736. Mehlman, M.J., 1989. Later Quaternary Archaeological Sequences in Northern Tanzania. University of Illinois. Mehlman, M.J., 1991. Context for the emergence of modern man in eastern Africa: some new Tanzanian evidence. In: Cultural Beginnings: Approaches to Understanding Early Hominid Lifeways in the African Savanna, pp. 177e196. Merrick, H.V., 1975. Change in Later Pleistocene Lithic Industries in Eastern Africa. University of California, Berkeley. Michels, J.W., Marean, C.A., 1984. A Middle Stone Age occupation site at Porc-Epic Cave, Dire Dawa (east-central Ethiopia), Part II. The African Archaeological Review 2, 37e71. Negash, A., Shackley, M.S., 2006. Geochemical provenance of obsidian artefacts from the MSA site of porc-Epic, Ethiopia. Archaeometry 48, 1e12. Negash, A., Brown, F., Nash, B., 2011. Varieties and sources of artefactural obsidian in the Middle Stone Age of the Middle Awash, Ethiopia. Archaeometry 53 (4), 661e673. Orton, J., 2008. Later Stone Age ostrich eggshell bead manufacture in the Northern Cape, South Africa. Journal of Archaeological Science 35, 1765e1775. Pearson, O.M., 1997. Postcranial Morphology and the Origin of Modern Humans. State University of New York. Pearson, O.M., Fleagle, J.G., Grine, F.E., Royer, D.F., 2008a. Further new hominin fossils from the Kibish Formation, southwestern Ethiopia. Journal of Human Evolution 55, 444e447. Pearson, O.M., Royer, D.F., Grine, F.E., Fleagle, J.G., 2008b. A description of the Omo I postcranial skeleton, including newly discovered fossils. Journal of Human Evolution 55, 421e437. 135 Phillipson, D.W., 1977. The excavation of Gobedra rockshelter. Azania 12, 53e82. Pleurdeau, D., 2003. Le Middle Stone Age de la grotte du Porc-Épic (Dire Dawa, Éthiopie): gestion des matières premières et comportements techniques. L’Anthropologie 107, 15e48. Pleurdeau, D., 2004. Gestion des matières premières et comportements techniques dans le Middle Stone Age africain: les assemblages lithiques de la grotte du Porc-Épic (Dire Dawa, Éthiopie). British Archaeological Reports Limited. Pleurdeau, D., 2005a. Human technical behavior in the African middle stone age: the lithic assemblage of Porc-Epic cave (Dire Dawa, Ethiopia). African Archaeological Review 22, 177e197. Pleurdeau, D., 2005b. The lithic assemblage of the 1975e1976 excavation of the Porc-Epic Cave (Dire-Dawa, Ethiopia). Implications for the East African Middle Stone Age. Journal of African Archaeology 3, 117e126. Reimer, P., 2013. IntCal13 and Marine13 Radiocarbon Age Calibration Curves 0e 50,000 years cal BP. Radiocarbon 55, 1869e1887. Rightmire, G.P., Deacon, H.J., 1991. Comparative studies of Late Pleistocene human remains from Klasies River Mouth, South Africa. Journal of Human Evolution 20, 131e156. Rightmire, G.P., Deacon, H.J., 2001. New human teeth from Middle Stone Age deposits at Klasies River, South Africa. Journal of Human Evolution 41, 535e544. Rightmire, G.P., Deacon, H.J., Schwartz, J.H., Tattersall, I., 2006. Human foot bones from Klasies River main site, South Africa. Journal of Human Evolution 50, 96e103. Rose, J.I., Usik, V.I., Marks, A.E., Hilbert, Y.H., Galletti, C.S., Parton, A., Geiling, J.M., \vCernỳ, V., Morley, M.W., Roberts, R.G., 2011. The nubian complex of Dhofar, Oman: an African middle stone age industry in southern Arabia. PloS One 6, e28239. Royer, D.F., Lockwood, C.A., Scott, J.E., Grine, F.E., 2009. Size variation in early human mandibles and molars from Klasies River, South Africa: comparison with other middle and late Pleistocene assemblages and with modern humans. American Journal of Physical Anthropology 140, 312e323. Singer, R., Wymer, J., 1982. The Middle Stone Age at Klasies River Mouth in South Africa. University of Chicago Press Chicago. Teilhard de Chardin, P., 1930. Le Paléolithique en Somalie fran\ccaise et en Abyssinie. L’Anthropologie 40, 331e334. Teilhard de Chardin, P., Breuil, H., Wernert, P., 1940. Les Industries lithiques de Somalie française. L’Anthropologie 49, 497e522. Tishkoff, S.A., Reed, F.A., Friedlaender, F.R., Ehret, C., Ranciaro, A., Froment, A., Hirbo, J.B., Awomoyi, A.A., Bodo, J.-M., Doumbo, O., Ibrahim, M., Juma, A.T., Kotze, M.J., Lema, G., Moore, J.H., Mortensen, H., Nyambo, T.B., Omar, S.A., Powell, K., Pretorius, G.S., Smith, M.W., Thera, M.A., Wambebe, C., Weber, J.L., Williams, S.M., 2009. The genetic structure and history of Africans and African Americans. Science 324, 1035e1044. Umer, M., Legesse, D., Gasse, F., Bonnefille, R., Lamb, H.F., Leng, M.J., Lamb, A.A., 2004. Late Quaternary climate changes in the Horn of Africa. In: Battarbee, R.W., Gasse, F., Stickley, C.E. (Eds.), Past Climate Variability through Europe and Africa. Developments in Paleoenvironmental Research 6, pp. 159e180. Van Peer, P., 1998. The Nile corridor and the Out-of-Africa model: an examination of the archaeological record. Current Anthropology 39, S115eS140 supplement. Verna, C., Texier, P.-J., Rigaud, J.-P., Poggenpoel, C., Parkington, J., 2013. The Middle Stone Age human remains from Diepkloof Rock Shelter (Western Cape, South Africa). Journal of Archaeological Science 40, 3532e3541. Vogel, N., Nomade, S., Negash, A., Renne, P.R., 2006. Forensic 40Ar/39Ar dating: a provenance study of Middle Stone Age obsidian artifacts from Ethiopia. Journal of Archaeological Science 33 (12), 1749e1765. Walter, R.C., Buffler, R.T., Bruggemann, J.H., Guillaume, M.M., Berhe, S.M., Negassi, B., Libsekal, Y., Cheng, H., Edwards, R.L., Von Cosel, R., 2000. Early human occupation of the Red Sea coast of Eritrea during the last interglacial. Nature 405, 65e69. Wang, Q., Tobias, P.V., Roberts, D.L., Jacobs, Z., 2008. A re-examination of a human femur found at the Blind River Site, East London, South Africa: its age, morphology, and breakage pattern. Anthropological Review 71, 43e61. White, T.D., Asfaw, B., DeGusta, D., Gilbert, H., Richards, G.D., Suwa, G., Howell, F.C., 2003. Pleistocene Homo sapiens from Middle Awash, Ethiopia. Nature 423, 742e747.