Origins of the Iberomaurusian in NW Africa: New AMS radiocarbon dating of the Middle and Later Stone Age deposits at Taforalt Cave, Morocco

Journal of Human Evolution 65 (2013) 266e281 Contents lists available at SciVerse ScienceDirect Journal of Human Evolution journal homepage: www.elsevier.com/locate/jhevol Origins of the Iberomaurusian in NW Africa: New AMS radiocarbon dating of the Middle and Later Stone Age deposits at Taforalt Cave, Morocco R.N.E. Barton a, *, A. Bouzouggar b, J.T. Hogue a, S. Lee c, S.N. Collcutt d, P. Ditchfield c a Institute of Archaeology, University of Oxford, 36 Beaumont Street, Oxford OX1 2PG, UK Institut National des Sciences de l’Archéologie et du Patrimoine, Hay Riad, Madinat Al Irfane, Angle rues 5 et 7, Rabat-Instituts, 10 000 Rabat, Morocco Research Laboratory for Archaeology and the History of Art, University of Oxford, Dyson Perrins Building, South Parks Road, Oxford OX1 3QY, UK d Oxford Archaeological Associates Ltd., 1 Divinity Road, Oxford OX4 1LH, UK b c a r t i c l e i n f o a b s t r a c t Article history: Received 5 February 2013 Accepted 1 June 2013 Available online 24 July 2013 Recent genetic studies based on the distribution of mtDNA of haplogroup U6 have led to subtly different theories regarding the arrival of modern human populations in North Africa. One proposes that groups of the proto-U6 lineage spread from the Near East to North Africa around 40e45 ka (thousands of years ago), followed by some degree of regional continuity. Another envisages a westward human migration from the Near East, followed by further demographic expansion at w22 ka centred on the Maghreb and associated with a microlithic bladelet culture known as the Iberomaurusian. In evaluating these theories, we report on the results of new work on the Middle (MSA) and Later Stone (LSA) Age deposits at Taforalt Cave in Morocco. We present 54 AMS radiocarbon dates on bone and charcoals from a sequence of late MSA and LSA occupation levels of the cave. Using Bayesian modelling we show that an MSA nonLevallois flake industry was present until w24.5 ka Cal BP (calibrated years before present), followed by a gap in occupation and the subsequent appearance of an LSA Iberomaurusian industry from at least 21,160 Cal BP. The new dating offers fresh light on theories of continuity versus replacement of populations as presented by the genetic evidence. We examine the implications of these data for interpreting the first appearance of the LSA in the Maghreb and providing comparisons with other dated early blade and bladelet industries in North Africa. Ó 2013 Elsevier Ltd. All rights reserved. Keywords: Maghreb North Africa Microlithic bladelet industry Introduction Recent phylogenetic studies of mtDNA haplogroups M1 and U6 have proposed that modern human populations in North Africa originated from groups that had migrated into this region from Southwest Asia (Maca-Meyer et al., 2003; Olivieri et al., 2006; Gonzalez et al., 2007). However, the nature, timing, and geographical spread of such a back-migration are still a matter of considerable debate (Pennarun et al., 2012). On the one hand, some studies propose an early dispersal of M1 and U6 lineages into North Africa at w40e45 ka (thousands of years ago) (Olivieri et al., 2006), while others suggest multiple events with a major expansion of the U6 lineages in the Maghreb w22 ka (Maca-Meyer et al., 2003; Pereira et al., 2010). Bound up with these models is the proposal that the geographical patterns of the haplogroups can be shown to coincide * Corresponding author. E-mail address: nick.barton@arch.ox.ac.uk (R.N.E. Barton). 0047-2484/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jhevol.2013.06.003 with major technological shifts in the archaeological record. One of these relates to sub-haplogroup U6a1 and its posterior clade U6a1a, with coalescence ages of w22 ka, which may be associated with the appearance of a culture known as the Iberomaurusian (Maca-Meyer et al., 2003). This microlithic bladelet industry is significant because it represents potentially the earliest Later Stone Age technology in the Maghreb (Morocco, Algeria, Tunisia). The genetic studies therefore also highlight the issue of whether the Iberomaurusian was a truly indigenous development to the Maghreb or whether it reflects a general spread of people and traditions from Cyrenaica with older roots in Southwest Asia. While the published genetics research provides useful models for understanding the early peopling of North Africa by modern humans, considerable caution must be exercised in interpreting these data. One issue concerns underlying assumptions regarding the timing of dispersal events that are heavily dependent on the methodology used to estimate molecular divergence values and DNA mutation rates (Endicott et al., 2009; Scally and Durbin, 2012). Indeed there is still a huge disparity between the age of U6 and R.N.E. Barton et al. / Journal of Human Evolution 65 (2013) 266e281 some of its individual clades, such that the age of U6a7 is consistently estimated as being older than that of U6 overall (Olivieri et al., 2006; Pereira et al., 2010). A further challenge is to assess whether any of the proposed demographic models can be demonstrated by testing them against empirical evidence in the archaeological and palaeontological records. For example, did the arrival of modern humans in the Maghreb result in abrupt changes in the archaeological record (replacement model) or were changes brought about independently and within the context of long-term population continuity (Debénath et al., 1986; Linstädter et al., 2012)? Until now, it has been difficult to assess such claims, because of the absence of high precision dating records for key regions such as the Maghreb and the scarcity of well-stratified archaeological sequences with associated human fossil remains. A site of major significance that can help remedy this situation is that of Grotte des Pigeons at Taforalt. The cave is located in the Beni-Snassen Mountains, in northeastern Morocco (Fig. 1), and has been the subject of recent excavations that provide a long and largely unbroken sequence of archaeological deposits from w12 to 110 ka (Bouzouggar et al., 2007; Clark-Balzan et al., 2012), covering the period of the proposed arrival in the region of modern humans. The sediments include both extensive Aterian and Iberomaurusian occupation, with cemetery evidence (Bouzouggar et al., 2006, 2007, 2008; Barton et al., 2007; Taylor et al., 2011; Humphrey et al., 2012). In this paper we report on the upper part of the sequence comprising the terminal Middle (MSA) and Later Stone (LSA) Age deposits of the cave. The 54 AMS (accelerator mass spectrometry) radiocarbon dates provide the first well constrained record for the appearance of the Iberomaurusian in northwest Africa. They also provide a basis for comparing the genetics-derived chronology and enable an independent test of the timing of the transition from the MSA in this region. Among the most intriguing elements in this research are those that concern the nature and origins of the Iberomaurusian. The techno-complex is found very widely distributed across North Africa and is associated with cemeteries containing skeletal remains of robust modern humans attributed to Mechta-Afalou types (Camps, 1974; Lahr, 1996; Irish, 2000; Humphrey and Bocaege, 2008). The Iberomaurusian is particularly well documented in cave, rock-shelter, and open-air sites in the Mediterranean coastal zone of the Maghreb with a distribution that potentially extends 267 into Cyrenaica (McBurney, 1967; Barker et al., 2008) and Egypt (Phillips, 1972). The Iberomaurusian lithic industry is typified by microlithic backed bladelets and, apart from its geographically wide distribution, is significant because it marks a diagnostically clear change from Middle Palaeolithic/Middle Stone Age technologies in the Maghreb (Lubell, 2001; Bouzouggar et al., 2008). Many specialists assign the Iberomaurusian to the Epipalaeolithic (Roche, 1963; Barton et al., 2007; Olszewski et al., 2011) but, despite the extraordinary wealth and density of findspots in the Maghreb and over a century of research, relatively little is known about how or where it originated. Various theories have been proposed for the cultural origins of the Iberomaurusian. The term itself derives from the fusion of two words ‘Ibero’ (meaning Spanish) and ‘Maurusian’ (referring to Mauretania tingitana, the name first given by the Romans to northern Morocco and western Algeria). The definition was introduced by Pallary, who used it to draw attention to similarities between lithic industries in Spain and Morocco that contained “une profusion de très petites lames à dos retouché et à pointe très aiguë” (Pallary, 1909). The implied link with southern Europe was dismissed by later archaeologists who recognised stronger African affinities and adopted alternative names reflecting regional sources such as ‘Oranian’ (Gobert and Vaufrey, 1932) and ‘Mouillian’ (Goetz, 1941) from type locations in Algeria. However, the term Iberomaurusian has always persisted in the literature and, for reasons of taxonomic priority, we shall continue to use it here. Divergent with these views was an idea put forward by McBurney (1967), that the Oranian/Iberomaurusian had arisen out of an ‘Upper Palaeolithic’ industry known as the Dabban, represented at the Cyrenaican site of Haua Fteah and which may be of Near Eastern origin. But a major anomaly in this scheme was that the Iberomaurusian appeared to be earlier in the Maghreb than for the rest of North Africa (McBurney, 1977; Close, 1986). More recently, it has been suggested that the development of the Iberomaurusian was part of a much wider, pan-regional phenomenon resulting in the appearance of backed bladelet technologies across much of North Africa and the Near East around 20e23 ka BP (Close and Wendorf, 1990; Vermeersch, 1992; Godfrey-Smith et al., 2003; Goring-Morris and Belfer-Cohen, 2003). However, this theory neither adequately addressed the possibility of an early Iberomaurusian in the Maghreb nor inherent differences in the tool typologies at Upper Nile Figure 1. Distribution of Iberomaurusian sites. 1. Cap Rhir, 2. El Khenzira, 3. Contrebandiers, 4. El Harhoura II, 5. Dar es-Soltan I, 6. Ghar Cahal, 7. Kehf El Hammar, 8. Hattab II, 9. Ifri El Baroud, 10. Ifri n’Ammar, 11. Kifan Bel Ghomari, 12. Taforalt, 13. La Mouillah, 14. Rachgoun, 15. Columnata, 16. Cap Ténès, 17. Rolland, 18. Rassel, 19. Oued Kerma, 20. El Hamel, 21. ElOnçor, 22. Afalou Bou Rhummel, 23. Tamar Hat, 24. Taza, 25. Ouchtata localities, 26. Horizon Collignon. 268 R.N.E. Barton et al. / Journal of Human Evolution 65 (2013) 266e281 sites such as Wadi Halfa, Gebel Silsila and Deir el Fakhouri, which would make such comparisons less likely (Vermeersch, 1992). A more radical proposal visualised links between the Epigravettian industries of the Italian peninsula and the Iberomaurusian (Camps, 1974; Ferembach, 1985; Debénath, 2003). Although some similarities exist in the typology of the industries, there are many other lines of evidence (palaeontological, genetic, dating) that disfavour an origin of the Iberomaurusian in the Epigravettian of Italy or Sicily (Mannino et al., 2011, 2012). Part of the problem in assessing the Iberomaurusian have been continuing doubts and ambiguities over the correct dating of samples and the integrity of their cultural associations. According to published studies, the earliest conventional radiocarbon ages for the Iberomaurusian come from Grotte des Pigeons (Taforalt), Morocco and Tamar Hat in Algeria. At Taforalt, Roche (1976) recorded two very early ages from ‘terre charbonneuse’ (charcoalrich sediments) of 21,900  400 BP (Gif-2587) and 21,100  400 BP (Gif-2586). But, for reasons that will be discussed below, both of these are now regarded as highly doubtful. Elsewhere in the Maghreb, the oldest radiocarbon date recorded for the Iberomaurusian is (MC-822) 20,600  500 BP from Layer 84/5 at Tamar Hat (Saxon et al., 1974). However, despite the fact that this was one of seven dates ranging from 20,600  500 to 16,100  360 BP, five of which came from stratified contexts of the 1973 excavations, scepticism remains over the use of bulked materials, which may be susceptible to cross-contamination. This is a common problem with other dated sites in the Maghreb whose ages are based on bulked samples and thus subject to mixing of carbonised material of potentially different ages. Outside the Maghreb, the best dating for the oldest Iberomaurusian still comes from the Haua Fteah in Cyrenaica, where layers excavated by McBurney can be shown to be no older than two radiocarbon dates of 16,070  100 BP (GrN-2586) and 18,620  150 BP (GrN-2585) (Close, 1986). Nonetheless the dating was again based on bulked charcoal samples and therefore susceptible to similar doubts over reliability. Allied to the uncertainties with chronology have been questions concerning the stratigraphic relationship between the Iberomaurusian and older industries. In caves in the Maghreb with wellpreserved sedimentary sequences, the Iberomaurusian can often be shown to overlie Aterian deposits. However, the nature of continuity or discontinuity between these two techno-complexes has been a matter of longstanding debate. Early archaeologists such as Antoine (1937) saw no appreciable gap between the Aterian and Iberomaurusian but this was not widely accepted and was gradually replaced by a consensus in favour of a hiatus separating the Aterian from the Iberomaurusian of between five to ten thousand years (Close, 1980, 1988; Debénath et al., 1986). Even so, such interpretations relied on questionable or minimum radiocarbon estimates that seemed to show the Aterian occupying a relatively short chronology from 40 to 20 ka BP (Bordes, 1976e1977; Debénath et al., 1986; Texier et al., 1988; Debénath, 2000). Newer studies based on luminescence, uranium series and AMS dating have now led to a drastic revision of this timescale with much earlier ages for both the oldest and latest Aterian occurrences (Bouzouggar et al., 2007; Roset and Harbi-Riahi, 2007; Richter et al., 2010; Clark-Balzan et al., 2012; Jacobs et al., 2012). For areas outside the Maghreb, a slightly different picture has emerged. In Cyrenaica, the Iberomaurusian can be demonstrated to lie directly above the Dabban ‘Upper Palaeolithic’ industry (McBurney, 1967; Barker et al., 2008), while a similar industry is also documented in northwestern Libya (Garcea, 2004). Thus, two potentially contrasting cultural-historical models have been proposed for geographically adjacent areas: one for the Maghreb exemplified by an Aterian-Iberomaurusian succession, and one outside this region to the east with an ‘Upper Palaeolithic’ industry as a cultural forerunner of the Iberomaurusian. In this paper, we re-examine the evidence for the dating of the sequence in the Maghreb and consider some of its wider implications. The aims of this paper, therefore, are four-fold:  To provide a first high precision Accelerator Mass Spectrometry (AMS) radiocarbon dating record for Taforalt in the Maghreb, spanning the Iberomaurusian (LSA) and the most recent prebladelet technology (MSA).  To identify (dis)continuities in the stratified occupation sequence, using a combination of sedimentological data and depth-age modelling based on Bayesian analyses and a Poisson process deposition model.  To assess relationships of any gaps in settlement and cultural shifts at Taforalt with environmental fluctuations by comparing their timing to existing archaeological and palaeoclimatic records in North Africa and globally.  To use these data to examine more widely the implications for theories of modern human dispersal in North Africa proposed in the genetic evidence. Site setting, stratigraphy and archaeological context Grotte des Pigeons at Taforalt (34 480 3800 N, 2 240 3000 W) is located at 720 m above mean sea level overlooking the Zegzel Valley in the Beni Snassen mountain range (Fig. 2). The bedrock in this area comprises steeply folded Permo-Triassic dolomitic limestones, with the cave itself having formed by rekarstification in a zone of earlier travertines and fluvial conglomerates, constituting a more ancient deep karstic fill. The currently accessible cave, with a large entrance opening to the northeast, has a floor area within the drip line of w400 m2. Today the site lies w40 km from the Mediterranean coast, currently within the ‘thermo-Mediterranean’ biozone (Blondel and Aronson, 1999). Major excavations were undertaken at Taforalt Cave in 1944e 1947, 1950e1955, and 1969e1977 (Roche, 1953, 1963, 1967, 1969, 1976), with further investigations taking place during the 1980s (Raynal, 1980; Courty et al., 1989) and a new phase of excavations that was begun in 2003 (Bouzouggar et al., 2006, 2007, 2008; Barton et al., 2007; Taylor et al., 2011). The new phase of work involved excavating from standing profiles left by Roche and other previous excavators. The aim was to collect fresh dating samples and to investigate the archaeological deposits that spanned a combined depth of over 10 m, and which contained rich Aterian hearth layers overlain by a 4-metre thick sequence of Iberomaurusian deposits. The latter included midden layers and an assemblage of partial skeletons recovered from two burial areas investigated between 1952 and 1955 by Roche (Ferembach et al., 1962; Roche, 1963). Part of the new work involved the investigation of an area contiguous with long sections dug by Roche on the south side of the cave, which we refer to as Sector 8 (Fig. 2). Other new trenches were located on the north side of the cave (Sector 9) and in the cemetery area (Sector 10). The sedimentary sequence for Sector 8 is shown in Fig. 3 and summarised schematically in Fig. 4. In the main part of the cave, Roche (1963, 1976) identified at least 17 sedimentary units (layers I-XVII) containing archaeological finds, which he subdivided into cendreuses (ashy) and argillo-sableuses (clayey sands), with a distinct separation between the upper grey ashy series (levels IeVIIII) and a lower series of reddish brown to yellow clayey sands (IXeXVII). Like Roche, we recognise the same major stratigraphic division between the Grey Series (ashy deposits) and the underlying Yellow Series (clayey sands) but we differ in the finer descriptive detail of the sedimentary succession. The Grey Series comprises an approximately 4 m thickness of dominantly anthropogenic ‘midden’ deposits (ash, R.N.E. Barton et al. / Journal of Human Evolution 65 (2013) 266e281 269 Figure 2. Plan of Taforalt showing earlier Roche excavations and main sectors of the recent field campaigns. charcoal, bone and snail shell debris, burnt limestone, etc). Locally within Sector 8, it is possible to maintain precise and detailed lithostratigraphic control, despite the dominantly lenticular sedimentation mode resulting from constantly shifting human activity in occupation areas across the site. The results of new excavation and sampling are reported in two separate but broadly equivalent sequences, some 3 m laterally apart, and excavated in 2003e05 (units G88-100) and 2009e10 (layers 1e29), the latter located slightly deeper into the cave (Fig. 4). In contrast to the Grey Series, the Yellow Series is finely laminated throughout, indicating emplacement by wash processes, with greater lateral continuity in stratigraphic units. These sediments are always dominated by fine to medium sands, with varying amounts of dolomitic limestone debris and minor peaks of quartzitic ‘grit’. Whilst there is an archaeological presence at most levels, only rarely does the anthropogenic input reach concentrations high enough to be reflected in the actual lithostratigraphy (e.g., the persistent traces of burning in the distinctive horizons of Units Y3 and Y5). Three of the many points of interest in this approximately 2 m thickness of Yellow Series deposits are singled out for comment here. First, there is an increasingly strong quartzitic coarse silt component upwards in Unit Y2, with plastic deformation phenomena present below the sharp upper boundary. Second, the middle portion of Unit Y4 has a more massive structure (lacking clear boundaries and significant erosion planes), with few stones and, again, a peak in quartzitic coarse silt. Third, from the top boundary of Y5 downwards (Y13 being the lowest unit reported here), it can be observed that erosion planes (often irregular) between units and the lamination within units are more strongly characterised than in overlying deposits. The possible implications of these features of the Yellow Series will be discussed below. 270 R.N.E. Barton et al. / Journal of Human Evolution 65 (2013) 266e281 this has been done on strictly defined typological and technological criteria and the phases can be correlated with specific chronostratigraphic units. The bladelet assemblages that make up the Iberomaurusian are markedly different from the flake assemblages that underlie them. These bear no obvious resemblance either in raw material or technology to the Iberomaurusian. A full study of the lithic assemblages will be published elsewhere (Hogue, in prep.) but the preliminary findings are summarised below. Grey Series IB3 Iberomaurusian lithic assemblages (layers 2e29) Figure 3. Main section on the south side of the cave (Sector 8) showing the Grey Series overlying the Yellow Series. Scale is 2 m long. (Photo: Ian Cartwright; copyright: Institute of Archaeology). In his monograph of the site, Roche (1963) described the archaeological sequence for the equivalent of Sector 8. The Iberomaurusian finds were analysed by layer but little attempt was made to explain variability across the sequence that was only alluded to briefly (Roche, 1963). In a subsequent publication, Roche subdivided the Iberomaurusian into two main phases: a phase ancienne and a phase classique (Roche, 1976). This followed the tripartite model developed for the Iberomaurusian by Camps (1974), though the latest phase evoluée of the Iberomaurusian was missing at Taforalt. Roche’s subdivisions of ancienne and classique were based on variation in the gross categories of tools (i.e., relative proportions of end-scrapers, burins, notches, backed bladelets, etc.) (Roche, 1963), with little, if any, attention given to variability within the artefact classes themselves. His scheme was also difficult to replicate because his classification did not take account of the more widely accepted typology devised by Tixier (1963) for the Epipalaeolithic of the Maghreb. In all, Roche described Iberomaurusian artefacts from eight of the grey ashy layers (couches cendreuses) and from nine of the underlying clayey sand layers (couches argilo-sableuses) (Roche, 1976). The ages on charcoals obtained from these stratigraphic units ranged from 10,800  400 BP to 16,420  190 BP, and included one from the cemetery in the grey ashy layers, which yielded an age of 11,920  240 BP. None of these dates overlapped with the two appreciably older ages on burnt sediments from lower down in the clayey sands e also attributed by him to the Iberomaurusian e with ages of 21,100  400 BP and 21,900  400 BP, respectively. Unfortunately there is no record of the artefacts so the association remains a matter of conjecture. Equally, our own studies of the same sequence excavated and drawn by Roche (actually surviving in our Area 3, see Fig. 2) have failed to verify any of his suggested ‘cultural’ boundaries. In addition, neither of the two radiocarbon dates were on charcoals and it is questionable whether determinations obtained on burnt earth would be sufficiently free of extraneous sources of carbon to have provided a reliable age for the deposit. Elsewhere in the same paper, Roche (1976) briefly refers to the direct superposition of the Iberomaurusian over the Aterian and a suggestion that the contact between these two cultural horizons could be dated by a radiocarbon date on Helix snail shells of (Gif-2276) 32,370 þ 1890/-2470 BP. Again, it is impossible to verify the stratigraphic position of any of these samples as no field notes survive. According to the analysis of lithics from the new excavations in our Sector 8, we would now subdivide the Iberomaurusian (sensu lato) into three distinctive phases. In contrast to the earlier work, Only a small degree of variation is present in the lithic assemblages recovered from the sedimentary layers of the Grey Series ashy deposits. Unretouched flakes and bladelets made with a soft hammer, and struck from single platform and opposed platform bladelet cores, are recorded. However, they occur in lower proportions than in the underlying deposits. There is a high frequency of burning within the assemblage. The raw materials consist of small fine-grained siliceous river cobbles known to derive from locations no further than about 25 km from the site. Sources closer to the site may have been exploited but these do not seem to include the streambed immediately below and in front of the cave. The retouched tools (Fig. 5) are dominated by backed bladelet types to the exclusion of almost all other tool classes. These pieces fall most commonly in the categories of curve-backed bladelets (Tixier types 56 through 59) and longer and more elongated pointed straightbacked bladelets (types 45e51). There is also evidence for the use of the microburin technique. End-scrapers, notched and denticulated pieces, and simple retouched flakes and blades are observed in small numbers, whilst burins and most other tool forms are absent from the assemblage. Upper Yellow Series IB2 Iberomaurusian lithic assemblages (Unit Y1) In contrast to the younger phase, bladelet and flake debitage is more common from this unit, which also includes greater numbers of cores. Similar small river cobbles were brought to the cave for knapping. The preferred method was to split the cobbles longitudinally before preparing a unidirectional crest down one edge of the split piece. Removal of blanks proceeded from one end of the core, with knapping sometimes switched to the other end, rather than regularly alternating between the two platforms. Bladelets show low levels of platform abrasion or other forms of preparation. The most noticeable difference from the assemblages from phase 3 is extensive use of the microburin technique. Microburins are very common in these levels (Fig. 5); some of them are quite large, and comprise predominantly distal types with the notch formed on the left lateral margin. Evidence for the use of the microburin technique is also observed in the high proportions of La Mouillah points, a backed bladelet form that retains the microburin facet either distally or proximally. Previous debates have focused on whether this type represents a formal tool in itself or is an intermediate stage in the manufacture of backed bladelets (Neeley and Barton, 1994; Olszewski et al., 2011). With the exception of the La Mouillah points, the range of retouched pieces within the Upper Yellow Series is similar to that found in the overlying Grey Series and there is clear technological continuity across the GreyeYellow Series transition. Lower Yellow Series IB1 Iberomaurusian lithic assemblages (Units Y2eUpper Y4) The assemblages from these layers are also characterised by bladelets in a range of fine-grained siliceous raw materials knapped R.N.E. Barton et al. / Journal of Human Evolution 65 (2013) 266e281 271 Figure 4. Schematic section through the Grey and Yellow Series deposits in Sector 8, illustrating different excavation columns and AMS radiocarbon dates (see Tables 1 and 2). from river cobbles. There is a high degree of similarity with phase 2 in the core-reduction procedures and in the use of soft hammer percussion but a strong divergence can be seen in the method of retouching bladelet tools from those in the phases above. There is no evidence of the microburin technique in the Lower Yellow Series. The distal ends of retouched bladelets are also rarely modified. Instead, the most common tool forms are obtuse-ended backed bladelets (Tixier’s type 67), with semi-abrupt or abrupt backing down one margin, and Ouchtata bladelets (Tixier’s type 70), which carry fine retouch down all or part of one edge (often the proximal portion). Pointed backed forms also occasionally occur (Fig. 5). Lower Yellow Series N-LF non-Levallois flake assemblages (units lower Y4eY11) There is a change in the nature of the lithic assemblage(s) from Lower Y4, which becomes much clearer from below this in Y6. The raw materials in these layers are represented by mainly coarsergrained metamorphic rocks. There are no blades or bladelets and backed tools are all but absent except in Y4. The debitage consists of flakes detached from non-Levallois cores and the method of percussion is dominated by non-organic hammer techniques (hard and soft stone). Amongst the few tools are side scrapers and 272 R.N.E. Barton et al. / Journal of Human Evolution 65 (2013) 266e281 Figure 5. Three phases of the Iberomaurusian. Top row: IB3 Curve-backed and elongated straight-backed bladelets; Middle: IB2 curve-backed points and ‘La Mouillah points’ (right); Bottom: IB1 Backed point (far left), Ouchtata bladelets (second and third from right), and obtuse-ended backed bladelets. Scale in cm. (Drawings by J.T Hogue). thinning flakes that would appear to derive from large bifacial tools (Fig. 6). Methods and results of a high precision radiocarbon chronology AMS radiocarbon samples Samples for AMS radiocarbon dating for this study were taken from Sector 8, the southeastern section of the central part of the cave, where both Iberomaurusian and underlying layers with a nonLevallois industry were present. A total of 52 radiocarbon determinations of cut-marked bones (collagen fraction) and charcoals were therefore available from the same section (Sector 8) of the cave. The majority of samples were individual large charcoals, with recorded spot heights and all can be confidently identified to stratigraphic unit (Table 1). Four were taken from sediment blocks of 10e 15 cm thickness. The entire sequence runs from 0.22 m to 5.77 m below Site Datum. Three radiocarbon determinations were duplicates, measured as quality control at ORAU (Oxford Radiocarbon Accelerator Unit). All samples were pre-treated to remove potential contaminants using standard ORAU protocols for each sample type (Brock et al., 2010). All radiocarbon determinations are cited as ‘BP’, where BP stands for years before 1950 AD; and calibrated radiocarbon or modelled ages on the calendar scale are cited as ‘Cal BP’. Two additional determinations (shown in Table 2 but not in Table 1), from the separate lithostratigraphic section in Sector 9 at Taforalt, represent the earliest occurrences of the Iberomaurusian yet known from this site. These were sampled and prepared in exactly the same manner as for the other determinations. However, they are excluded from the age/depth model because they come from a separate area of the site but are mentioned here to add weight to the overall study and to render the test of the ‘gap hypothesis’ as robust as possible. R.N.E. Barton et al. / Journal of Human Evolution 65 (2013) 266e281 273 Figure 6. Non-Levallois flake industry at Taforalt. Side scrapers (TAF 4056 in quartzite and TAF 6544 in chert), retouched flake (Taf 4163 in quartzite). Scale in cm. (Drawings by M. Grenet). Bayesian analysis Bayesian analysis is used in this study to combine archaeological information (prior probability) with calibrated dates (likelihoods) to improve the precision and accuracy of the chronology (posterior probability). The computer program OxCal v.4.1 (Bronk Ramsey, 2009a) was used for chronology construction. Two priors for modelling the archaeological information, i.e., the deposition sequence, were employed. These were a Poisson process model (P_Sequence; Bronk Ramsey, 2008) and a uniform phase model. The first was used with depth (in metres) information of individual samples and the second was used to model samples (from sediment blocks) without exact spot heights, but with their positions constrained by using the known depth limits of the sediment blocks. Note that, in Fig. 4 ‘apparent’ depth at the section face is shown as the vertical dimension. However, in the Baysesian analysis (cf. Table 1) this has been corrected to ‘true’ depth to reflect the position of samples in a vertical stack, with each unit represented as the typical average value of its observed thickness range. The two different deposition models were cross-linked at the height limits of the sediment blocks. The model averaging approach was used to estimate an appropriate value for k in the P_Sequence model (Bronk Ramsey and Lee, 2013). Duplicate measurements were combined using the R_Combine function in OxCal to produce a weighted average. Formal outlier analysis was also utilised to account for outliers in the radiocarbon scale and the calendar scale. The outlier models are specified to allow the possible shifts in the specified scale to be drawn from a long tail Student t distribution. The outliers can be in the scale of anywhere between 100 and 104 years. These are the models recommended by Bronk Ramsey (2009b) for general purposes when the scale of the possible offsets is unknown. When employed, the overall model is not affected by the odd extreme outlier. Each measurement is assigned a prior probability of 5% of being an outlier. Posterior chronology The posterior chronology for the Iberomaurusian sequence in Sector 8 is shown in Fig. 7 and Table 2. The chronology was modelled using multiple P_Sequence functions due to the presence of different types of sediments and hiatuses in the sequence (Fig. 8). In terms of the outlier analyses applied, no outliers were found in the radiocarbon scale, and only one (OxA-23411) was found to yield 95% probability of being an outlier in the calendar scale. Such findings support the validity of the individual radiocarbon determinations themselves (Staff et al., 2011). The model suggests that the boundary between the Yellow and Grey Series occurred in the range of 15,190e14,830 Cal BP (95.4% probability) (Table 2). In the Yellow Series, the difference in age between samples OxA-22788 (Unit Y1) and OxA-16267 (Unit Y2) was calculated to be between 1245 and 2102 years, and 2677 and 3830 years between samples OxA-16273 (upper Unit Y4) and OxA16271 (lower Unit Y4) (95.4% probability; Figs. 9 and 10). Discussion The AMS record of 54 dates for Taforalt provides the largest coherent set of radiocarbon determinations yet available for this period in the Maghreb and is an important baseline for understanding the development of this LSA technology and its relationship with stratigraphically older industries here and across North Africa. The unmodelled ages indicate a timespan of at least 9000 calendar years for Iberomaurusian occupation, beginning abruptly and with no obvious antecedents at 22,093e21,420 Cal BP (the earliest Sector 9 sample at two s) and ending in this cave (Sector 8) at 12,698e12,548 Cal BP (at two s), though younger ages at other sites indicate a prolonged existence in the region (Bouzouggar et al., 2008; Linstädter et al., 2012). One of the significant implications of the new results is the fresh light they cast on the chronological relationship between the Iberomaurusian and an earlier technology characterised by flakes made using a non-Levallois technique. In the past, the assumption had been that pre-Iberomaurusian industries in this region shared a uniquely Aterian affinity. This idea was directly endorsed by Roche, who believed that there was a “superposition directe entre l’Atérian final .et l’Epipaléolithique très ancien.” at Taforalt (Roche, 1976: 157e158), with a break possibly of relatively short duration in between. The observation was repeated by Debénath et al. (1986), who recognised the same succession in the Témara caves in western Morocco but with a potentially longer hiatus of up to 5000 years. In our excavations at Taforalt, we could find no evidence of such a superimposition and we now suspect that, even if Roche’s observations were consistent, he failed to recognise both the vertical and lateral complexity of the sedimentary sequences. We would suggest that such an interpretation is also no longer tenable for sites in the Témara area (Contrebandiers, El Mnasra, Dar es-Soltane) where redating would imply a much wider gap between the Aterian and the Iberomaurusian (Schwenninger et al., 2010; Jacobs et al., 2012). At 274 R.N.E. Barton et al. / Journal of Human Evolution 65 (2013) 266e281 Table 1 Un-calibrated Radiocarbon determinations from Grotte des Pigeons (Sector 8) at Taforalt, Morocco. Unit Grey series Yellow series Sample OxA-24111 OxA-13479 OxA-23404 OxA-13480 OxA-13516 OxA-24112 OxA-13517 OxA-24113 OxA-23405 OxA-23407 OxA-23406 OxA-23408 OxA-23409 OxA-23410 OxA-13477 OxA-23411 OxA-13478 OxA-22902 OxA-22904 OxA-22787 OxA-22785 OxA-22784 OxA-24109 OxA-22786 OxA-22903 OxA-22905 OxA-22788 OxA-16267 OxA-22907 OxA-22906 OxA-22908 OxA-16268 OxA-13519 OxA-22909 OxA-16272 OxA-16269 OxA-16270 OxA-13518 OxA-16242 OxA-16273 OxA-16271 OxA-16274 OxA-22910 OxA-16275 OxA-24110 OxA-13607 OxA-16243 OxA-16244 OxA-13556 OxA-16276 OxA-16278 OxA-16277 14 C determination 10,680 10,935 10,870 10,950 11,065 11,165 10,990 11,540 11,615 11,465 11,445 11,545 11,890 12,405 12,675 13,060 12,495 12,370 12,490 12,545 12,500 12,660 12,605 12,200 13,045 12,665 12,850 14,005 14,230 14,135 14,110 14,515 13,905 14,140 14,630 15,790 16,285 17,085 16,630 17,515 20,420 20,630 20,030 20,560 20,800 22,200 22,890 25,860 25,760 26,550 29,310 29,160 Uncertainty Av depth (m) Layer Species Ref. 45 40 45 45 45 45 45 50 50 50 55 55 55 55 50 65 50 50 50 55 55 70 55 55 50 50 55 60 55 55 55 60 55 55 60 60 65 65 75 75 90 90 90 90 120 90 120 150 140 140 160 160 0.22 0.36 0.36e0.48 0.36e0.48 0.48 0.53 0.54 0.69e0.61 0.69e0.61 0.94 0.94 1.14 1.56 2.96 3.50 3.51 3.88 3.97e3.88 3.97e3.88 3.97e3.88 3.97e3.88 3.97e3.88 3.98 4.02 4.05 4.08 4.12 4.16 4.17 4.18 4.25 4.26 4.30 4.35 4.36 4.55 4.68 4.72 4.77 4.90 4.96 5.11 5.11e5.16 5.16 5.17 5.24 5.34 5.48 5.61 5.63 5.77 5.77 L2 G88 L3 G89 G89 L4 G90 L6 L6 L8 L8 L11 L15 L25 G97 L28 G99 G99 G99 G99 G99 G99 G100 Y1 Y1 Y1 Y1 Y2 Y2 Y2 Y2 Y2 Y2 Y2 Y2 Y2 Y3 Y3 Y4 Y4 Y4 Y5 Y6 Y6a Y6 Y6 Y6b Y6d Y? Y7 Y12 Y12 Ammotragus Pinus sp. Pinus sp. Pinus sp. Pinus sp. Ammotragus Dicotyledonous Juniperus/Tetraclinus Gazella Juniperus/Tetraclinus Juniperus/Tetraclinus Pinus sp. Pinus sp. Juniperus/Tetraclinus Conifer Juniperus/Tetraclinus Juniperus/Tetraclinus Conifer Conifer Conifer cf. Juniperus cf. Juniperus Bos cf. Cedrus cf. Juniperus cf. Arbutus Conifer Tetraclinus articulata cf. Juniperus Conifer cf. Arbutus Tetraclinus articulata Juniperus/Tetraclinus Conifer Quercus sp. Juniperus sp. Pinus sp. Quercus sp. Dicot unidentified Pinus sp. Pinus sp. Conifer cf. Cedrus Pinus sp. Panthera Taxus sp. Juniperus sp. Pinus sp. Quercus sp. Pinus sp. Cupressus sempervirens Cupressus sempervirens 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 The dating samples are presented by averaged depths from the surface; where ranges shown, this indicates that the samples came from sediment blocks of known thickness. many of these and other sites the Iberomaurusian is separated from earlier (‘Middle Palaeolithic’/MSA) occupation by archaeologically sterile layers (Linstädter et al., 2012). The nature of the first ‘radiocarbon hiatus’ in the cultural sequence at Taforalt is of particular interest because of the implications for the presence/absence of human populations in northwest Africa at that particular time. According to the modelled Sector 8 radiocarbon ages, the gap in occupation between the Iberomaurusian (IB1) and underlying non-Levallois flake (N-LF) industry seems to have been around 3800 years at two s confidence interval. However, this separation may have been of shorter duration because of two slightly older ages for the Iberomaurusian in Sector 9 of Taforalt. These age estimates cannot be incorporated directly into the same age-depth model for the deposits of Sector 8 (although they can be linked stratigraphically within the Yellow Series). Taking the calibrated ages for the oldest Iberomaurusian of 22,093e21,420 Cal BP (2s) and the uppermost levels of the non-Levallois flake industry of 24,769e23,940 Cal BP (2s) considerably reduces the break to potentially no more than 1900 years. These observations on the hiatus are also underpinned by sedimentological data. As referred to above, from lower Y4 downwards, the units that contain the non-Levallois flake industry are characterised by strong bedding features, both in unit boundaries and internal lamination. The immediately overlying layers display no major disconformities and (even though there are small-scale signs of bioturbation) the sediments are more massive in structure and lack distinct boundaries. This part of the sequence covers a thickness range of 10e40 cm and is largely devoid of lithic artefacts. More concentrated signs of occupation with identifiably Iberomaurusian lithics occur above this in upper Y4 that feature fine sands, slightly gritty in places, with smaller stones present in bands. These observations do not suggest major gaps in the depositional sequence, certainly not one corresponding with the ‘radiocarbon hiatus’. 275 R.N.E. Barton et al. / Journal of Human Evolution 65 (2013) 266e281 Table 2 Radiocarbon determinations, their un-modelled calibrated ages, (IntCal09) and the modelled posterior age highest probability density (HPD) ranges at 68.2% and 94.5%. Sample Sector 8 Sector 9 OxA-24111 OxA-13479 OxA-23404 OxA-13480 OxA-13516 OxA-24112 OxA-13517 OxA-24113 OxA-23405 OxA-23407* OxA-23406* OxA-23406/07** OxA-23408 OxA-23409 OxA-23410 OxA-13477 OxA-23411 OxA-13478 OxA-22902 OxA-22904 OxA-22787 OxA-22785* OxA-22784* OxA-22784/85** OxA-24109 YG OxA-22786 OxA-22903 OxA-22905 OxA-22788 OxA-16267 OxA-22907 OxA-22906 OxA-22908 OxA-16268 OxA-13519 OxA-22909 OxA-16272 OxA-16269 OxA-16270 OxA-13518 OxA-16242 OxA-16273 OxA-16271 OxA-16274 OxA-22910 OxA-16275 OxA-24110 OxA-13607 OxA-16243 OxA-16244 OxA-13556 OxA-16276 OxA-16278* OxA-16277* OxA-16277/78** OxA-16260 OxA-16240 14 C date Uncertainty Un-modelled (Cal BP) Modelled (Cal BP) 68.2% 95.4% 68.2% 95.4% Ref 12,662e12,581 12,765e12,665 12,868e12,735 12,893e12,759 13,031e12,897 13,065e12,935 13,076e12,937 13,361e13,281 13,374e13,286 12,709e12,554 12,835e12,636 12,913e12,684 12,944e12,699 13,085e12,841 13,098e12,884 13,100e12,886 13,423e13,161 13,427e13,098 13,425e13,349 13,476e13,380 13,796e13,646 14,540e14,291 14,755e14,596 14,759e14,600 14,866e14,691 14,951e14,771 14,974e14,785 14,986e14,794 13,473e13,299 13,579e13,346 13,856e13,505 14,622e14,166 14,844e14,444 14,849e14,450 14,935e14,571 15,051e14,670 15,055e14,664 15,063e14,669 14,991e14,797 15,096e14,932 15,110e14,940 15,144e14,987 15,204e15,022 15,372e15,050 15,515e15,076 17,126e16,974 17,161e17,020 17,190e17,055 17,479e17,241 17,530e17,242 17,727e17,311 17,962e17,409 18,002e17,625 18,943e18,797 19,544e19,407 19,987e19,542 20,076e19,786 20,763e20,430 24,400e23,963 24,688e24,345 24,764e24,367 24,840e24,479 24,903e24,555 26,814e26,292 28,023e27,561 30,670e30,358 31,016e30,707 31,186e30,996 15,065e14,670 15,171e14,831 15,190e14,830 15,233e14,888 15,342e14,955 15,476e14,995 15,686e15,010 17,204e16,898 17,237e16,950 17,268e16,979 17,532e17,131 17,587e17,158 17,763e17,195 17,985e17,350 18,032e17,459 19,208e18,721 19,801e19,310 20,181e19,445 20,264e19,597 21,160e20,360 24,529e23,859 24,852e24,236 24,890e24,279 24,943e24,378 25,020e24,439 27,374e26,146 28,112e26,962 30,943e29,707 31,120e30,504 31,287e30,880 1 2 3 4 5 6 7 8 9 10 11 10/11 12 13 14 15 16 17 18 19 20 21 22 21/22 23 34,430e33,461 21,560e21,370 21,794e21,480 34,522e33,322 21,746e21,242 22,040e21,431 10,680 10,935 10,870 10,950 11,065 11,165 10,990 11,540 11,615 11,465 11,445 11,456 11,545 11,890 12,405 12,675 13,060 12,495 12,370 12,490 12,545 12,500 12,660 12,562 12,605 45 40 45 45 45 45 45 50 50 50 55 37 55 55 55 50 65 50 50 50 55 55 70 44 55 12,644e12,568 12,879e12,710 12,808e12,646 12,894e12,718 13,085e12,890 13,135e12,962 12,940e12,738 13,432e13,315 13,566e13,362 13,385e13,272 13,378e13,254 13,373e13,271 13,440e13,314 13,830e13,673 14,631e14,164 15,183e14,886 16,135e15,261 14,935e14,255 14,560e14,141 14,924e14,251 15,015e14,568 14,949e14,256 15,175e14,788 15,010e14,606 15,089e14,659 12,698e12,548 12,942e12,649 12,899e12,615 13,051e12,646 13,108e12,747 13,220e12,865 13,070e12,684 13,527e13,258 13,636e13,310 13,445e13,200 13,433e13,173 13,430e13,211 13,562e13,263 13,906e13,492 14,958e14,110 15,447e14,606 16,406e15,190 15,057e14,201 14,910e14,068 15,051e14,196 15,125e14,237 15,066e14,201 15,494e14,524 15,160e14,245 15,204e14,261 12,200 13,045 12,665 12,850 14,005 14,230 14,135 14,110 14,515 13,905 14,140 14,630 15,790 16,285 17,085 16,630 17,515 20,420 20,630 20,030 20,560 20,800 22,200 22,890 25,860 25,760 26,550 29,310 29,160 29,236 18,005 18,185 55 50 50 55 60 55 55 55 60 55 55 60 60 65 65 75 75 90 90 90 90 120 90 120 150 140 140 160 160 114 75 75 14,140e13,957 16,045e15,251 15,180e14,870 15,505e15,095 17,157e16,920 17,487e17,147 17,400e17,009 17,381e16,974 17,861e17,550 17,062e16,863 17,402e17,017 17,939e17,673 19,240e18,795 19,547e19,397 20,386e20,182 19,888e19,568 21,126e20,536 24,512e24,207 24,812e24,459 24,150e23,770 24,744e24,375 24,960e24,556 26,857e26,311 28,017e27,106 30,850e30,483 30,744e30,395 31,208e31,032 34,461e33,660 34,428e33,432 34,445e33,560 21,560e21,343 21,893e21,499 14,482e13,827 16,372e15,183 15,431e14,589 15,873e14,973 17,416e16,819 17,615e16,991 17,523e16,921 17,500e16,901 17,933e17,256 17,179e16,785 17,527e16,925 18,046e17,492 19,305e18,740 19,807e18,965 20,505e19,955 20,094e19,474 21,283e20,465 24,769e23,940 24,970e24,345 24,310e23,546 24,940e24,252 25,121e24,431 27,501e26,173 28,104e26,936 30,994e30,342 30,926e30,290 31,291e30,936 34,567e33,422 34,509e33,297 34,520e33,406 21,804e21,194 22,093e21,420 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 51/52 60 61 Samples marked with an asterisk (*) are duplicates and two asterisks (**) are their combined ages. The sample named “YG” indicates the modelled age for the Yellow/Grey series transition at 4 m. Another critical issue in evaluating the archaeological evidence concerns the attribution of the non-Levallois industry and whether it differs significantly from the Aterian, found in deeper layers of the site. In reality, the differences are reasonably clear-cut and are helped by consistencies in the Aterian typology and technology in various sites throughout the Maghreb. For example, a particular feature of the Aterian is the use of the Levallois technique and associated tools, which usually include a range of scrapers, pedunculate points and bifacial foliates made on diagnostically Levallois products including blades (Bouzouggar and Barton, 2012). In contrast, such items are notably absent in the industry in the lower sequence in Sector 8, which includes occasional flake tools, but none of them are made on obvious Levallois blanks. Unfortunately, as there are no layers with Aterian artefacts in Sector 8, a direct stratigraphic relationship with the non-Levallois industry cannot be demonstrated for this part of the site. However, in Sector 9, accessible sections have revealed a thick sequence of sediments that includes Iberomaurusian layers near the top (with the reported 276 R.N.E. Barton et al. / Journal of Human Evolution 65 (2013) 266e281 Figure 8. Apparent sedimentation rates, Yellow Series in Sector 8. The modelled C14 date ranges (95.4%) are plotted by the two lines against depth in the sequence, so that sedimentation rate approximates to the slope of the lines. Human input tends to increase the slope (sedimentation rate), particularly at the strongly anthropogenic Units Y3, Y5 and Y7. Most instances of apparent low slope are likely to be ‘jumps’ (as indicated in the diagram) caused by demonstrable erosion events. In this respect, the Y8e 11 sequence could be shown as a series of small steps. Only the markedly decreased slope within Unit Y4 appears to represent a truly significant slowing of sedimentation (see main text for discussion). Personal communication). For the moment, Taforalt provides the only securely dated sequence for a clear cultural succession preceding the Iberomaurusian and showing that an Aterian Levalloisdominated technology was probably replaced by one of nonLevallois type. We believe this model can be tested more widely and has major ramifications for interpreting other lithic artefact sequences in the Maghreb. Turning to the Iberomaurusian, although the deposits at Taforalt represent a thick and fairly continuous record of human occupation, there are in fact subtle variations in the cultural sequence. The clearest example is the switch from IB1 with marginally backed (‘Ouchtata’) blades and bladelets to IB2 dominated by microlithic backed bladelets. The actual transition between the two phases is marked by a sharp sedimentary contact between Units Y2/Y1. This is clearly an erosive boundary so it is impossible to know whether Figure 7. Age-depth model for Taforalt cultural sequence. Posterior AMS chronology is modelled with one P Sequence function for samples with spot heights. AMS dates). The layers immediately beneath this are not archaeologically rich but include undiagnostic flakes in raw materials similar to those in the non-Levallois industry of Sector 8. At the base of the Sector 9 excavation trench and separated from the lowest Iberomaurusian by at least 130 cm are a series of charcoal-rich laminated hearth deposits. These contain clear examples of Aterian retouched tools (including a bifacial foliate) and Levallois debitage. An OSL date of 37.57  3.42 BP (Clark-Balzan, Personal communication) has been obtained from the deposits and represents potentially the youngest age so far recorded for the Aterian at this site. No radiocarbon dating has yet been undertaken on the charcoals, but if the age estimate is correct it may give greater credence to an OSL date of 30.9  2.5 ka years for the Aterian at Wadi Noun in southern Morocco (Weisrock et al., 2006; Wengler, 2010), and to ESR dating of around 39  4 ka years on the later Aterian at Mugharet el ‘Aliya in northern Morocco (Wrinn and Rink, 2003), where new radiocarbon dates are also anticipated (Tuross, Figure 9. Modelled chronological break between MSA non-Levallois flake technology and the oldest Iberomaurusian. R.N.E. Barton et al. / Journal of Human Evolution 65 (2013) 266e281 Figure 10. Modelled chronological break between Iberomaurusian IB1 and IB2 industries in Y2 and Y1, respectively. the time gap of 857 years (between 15,686e15,010 BP and 17,204e 16,898 BP at 2s) in the radiocarbon model is more apparent than real. Yet, observations of the sediments towards the top of Unit Y2 reveal some unusual geological phenomena, which may provide clues about the environmental conditions during the later formation of this unit. A raised silt content would suggest that wash input was capturing aeolian dust from the exterior surroundings. It appears too that the signal of regional aridity implied by these processes evidently came to a brusque end with a phase of instability indicated by plastic deformation (wetness and possibly even transient ground freezing) in the uppermost part of this unit. The wider consequences of these observations will be explored further below. Also of relevance is the temporal relationship between the IB2 and IB3 assemblages. Here the techno-typological changes are relatively minor and relate mostly to the highly varied appearance of the microlithic bladelet forms in IB3. These introductions seem to have been made during an uninterrupted period of cave use, and we would therefore hypothesise that such differences in typology may represent a drift in the intensity and nature of activities across the site. One of the key aims of this paper has been to assess whether gaps in settlement and shifts in lithic cultural phases can be correlated with phases of rapid climatic change. Palaeoclimatic records available from Atlantic and the Alboran Sea marine cores (Combourieu Nebout et al., 2002; Sánchez-Goñi et al., 2002; Fletcher and Sánchez-Goñi, 2008) confirm that major climatic shifts occurred in the western Mediterranean during the Late Pleistocene, and these could have had a significant impact on human populations. Of particular interest are so-called Heinrich Events marked by the incursion of cold polar surface waters into the Mediterranean (Moreno et al., 2005). These appear to have been associated with phases of greater aridity in Iberia and northwest Africa as documented in pollen sequences from the Alboran Sea cores (Combourieu Nebout et al., 2002; Sánchez-Goñi, 2006; Jiménez-Espejo et al., 2007; Fletcher and Sánchez-Goñi, 2008) and increased input of windborne dust from the Sahara (Moreno et al., 2002; Moreno, 2012). In contrast, the periods following Heinrich episodes appear to have been relatively warm and humid, as for example the ‘Last Glacial Maximum’ phase after HE2 (Penaud et al., 2010) and the phase of major warming at the beginning of Greenland Interstadial 1e, the latter marked by a rise in sea surface temperatures and the diversion westward of moisture bearing winds bringing higher precipitation to the Maghreb (Moreno et al., 2005; Rodrigo-Gámiz et al., 2011). Climate modelling also suggests that annual rainfall may have fallen below 100 mm per year during 277 certain Heinrich Events (Sepulchre et al., 2007). A further source of information comes from the direct comparison between the radiocarbon dated sequence at Taforalt with the NGRIP ice-core timescale (Fig. 7), which allows broad correlation with DeO (DansgaardeOeschger) stadial-interstadials of the Greenland record (Bond et al., 1993; Dansgaard et al., 1993; Rodrigo-Gámiz et al., 2011). Beginning with the Grey Series, the radiocarbon model confirms that this unit began accumulating close to the start of Greenland Interstadial 1e and continued throughout most of GI1, a phase of relatively humid conditions. However, it is also clear that the Grey ash deposit is heavily anthropogenic (midden) and that the accumulation rate was extremely rapid (at c. 1.7 m/kyr) so that any obvious palaeoenvironmental signal may have been effectively swamped. Only the presence of cedar (Cedrus sp.) charcoals upwards from Unit G89 (Ward, 2007) would imply more ‘montane’ (cooling) conditions, and this could indicate the onset of Younger Dryas or Greenland Stadial 1 at the very top of this sequence (Fig. 7). One other potential sign of variation occurs near the base of the Grey Series where the remains of Barbary ground squirrels (Atlantoxerus getulus) suggest cooler, drier conditions (Bouzouggar et al., 2008). Moving downward (backwards in time) into the Yellow Series, clearer evidence of climatic variation (instability) occurs in the sediments of upper Y2 and here it is interesting to note that the dating model places this period of deposition within the same time span as Heinrich Event 1 (HE1). It also marks the point at which IB1 assemblages are replaced by those of IB2, and may be evidence that the dislocation in the cultural signature was influenced by climatic change. However, it is difficult to ascertain the length of break between the two Iberomaurusian phases because of the erosive unconformity. A more marked example of climate-cultural change occurs earlier in the Yellow Series sequence. Here there is a clear time lag between the earliest Iberomaurusian and the N-LF industry. The stratigraphic expression of this ‘gap’ is not, as in the younger example, an erosion plane. Rather, there is a body of sediment that is generally finer than the norm and which contains significant silt (dust). It would therefore appear that the sedimentation rate dropped drastically (to less than 0.05 m/kyr), probably in a significantly arid and cool period. It is perhaps relevant that the ‘gap’ calculated here includes the time span of HE2, marking the onset of Greenland Stadial 2 (GS2) and following the more variable and probably often more moist environmental conditions normally reconstructed for the preceding GS3. The latter may be reflected in the strong bedding structure present in the underlying Sector 8 Units Y13eY5 at Taforalt. Some indication of cooler climatic conditions is also implied by a rise in cedar (Cedrus sp.) charcoals in Y65, which disappears by the middle of Y4 (Ward, Personal communication). Lastly, what are the implications for the genetic theories of modern human dispersal in North Africa? Two of the most recent scenarios put forward in those studies have suggested that the Iberomaurusian either: 1) originated in populations that gradually spread westward from Cyrenaica, or 2) developed independently somewhere in the Maghreb and was subsequently transmitted eastwards (and possibly west) via an expansion of haplogroups U6a1 and U6a1a lineages. Archaeologically, in order to satisfy the conditions of the first hypothesis, one would expect logically that the oldest expression of this ‘culture’ should occur in the east. Despite new claims for an earliest age at Haua Fteah of w19 ka (Barker et al., 2010), we would suggest caution in the interpretation as this is so far based on dates on two shells from so-called transitional Layers XIVeXV of the old excavation. Equally, suggestion of early dating of sites in Upper Egypt (Close, 2002) are based on old charcoal dates from open sites, whilst the oldest age estimate from western Libya at Ain Shakshuk of 16,750  60 BP (Barich and 278 R.N.E. Barton et al. / Journal of Human Evolution 65 (2013) 266e281 Garcea, 2008) is still substantially younger than the oldest Iberomaurusian recorded in the Maghreb. Secondly, if the progenitor of the Iberomaurusian was an industry with retouched blades and chamfered pieces like the Dabban then it follows that, given the earlier presence of Iberomaurusian in the Maghreb, there should also be signs of the Dabban or similar blade industry in this region. Here again, the criteria are not clearly met as there are no immediately pre-Iberomaurusian blade industries yet known in the Maghreb. So far the most westerly findspot of a blade industry that might fit this description is in the Jebel Gharbi of northwestern Libya, with estimated ages of w30 ka BP (Garcea, 2004). At the Haua Fteah, the earliest Dabban can now be firmly anchored by tephro-chronology to just before w40 ka BP (Lowe et al., 2012). The dating is significant because it brings the Dabban more closely into line with other early blade technologies such as the Ahmarian of the Near East (Belfer-Cohen and Goring-Morris, 2003; Marks, 2003) and at Nazlet Khater in Egypt (Vermeersch, 2010), which are of broadly similar age. However, it is unlikely on technological grounds that there was any demographic connection between these regions (Marks, 1975; Iovita, 2009). It also follows that any interactions between early Upper Palaeolithic blade technologies did not extend very far if at all into the Maghreb. The second scenario, which envisages the earliest occurrence of the Iberomaurusian in the Maghreb, offers a better fit with the archaeological data but raises other important issues. One of these concerns the relationship of the Iberomaurusian with older technologies in this region. As has been shown above, there are no obvious antecedents for the blade and bladelet industries in the Maghreb and it is generally agreed that sterile layers separate the Iberomaurusian from anything earlier, with no suggestion of cultural continuity (Bouzouggar et al., 2008; Linstädter et al., 2012). A question therefore arises whether the replacement of non-Levallois flake assemblages by the Iberomaurusian is equivalent to the Middle Stone Age (MSA) to Late Stone Age (LSA) transition seen in other parts of Africa (Clark, 1959; McBrearty and Brooks, 2000; Mitchell and Barham, 2008). Unfortunately, the only useful comparisons that can be made are with the earliest bladelet technologies in regions that are geographically distant from the Maghreb. For example, in northeast Africa and the Nile Valley, bladelet industries become prevalent after w25 ka (Schild and Wendorf, 2010). But in these regions, the situation is made more complex because the Middle Palaeolithic/MSA includes both Levallois flake assemblages (Nazlet Safaha 2) and blade technologies (Taramsa Hill, Nazlet Khater), which were present from as early as MIS 3 (Van Peer, 2004; Vermeersch, 2010). Nonetheless, it is significant that there are no reported cases of either backed bladelets or ‘ouchtata’ forms in the MSA blade assemblages (Vermeersch, 2010). Further afield, the systematic manufacture of bladelets seems to have begun somewhat earlier. For instance, in East Africa, assemblages with a major bladelet component can be dated to 45 ka BP at Mochena Borago rockshelter (southwest Ethiopian Highlands) (Brand et al., 2012) and to 46 ka BP at Enkapune ya Muto (Kenya) (Ambrose, 1998). In other areas of sub-Saharan Africa at Mumba rockshelter, northern Tanzania, newly reported OSL ages suggest that the production of bladelets became abundant around 49.1  4.3 ka BP (Gliganic et al., 2012), while at Border Cave microliths made on opposed platform cores are common from 44,000 to 42,000 Cal BP (Villa et al., 2012). At each of these sites, there are signs of continuous sedimentation and occupation with underlying layers, which seem to imply gradual cultural transitions. The only area that might offer some parallels with the Maghreb is in the southern margins of southern Africa, where from the end of Marine Isotope Stage 3 (w25 ka) microlithic assemblages made on quartzites become manifest and are succeeded by systematic bladelet production in Robberg-type assemblages from around w19 ka (Klein, 1974; Wadley, 1997; Mitchell, 2002). Thus, over much of Africa standardised microlithic bladelet production (LSA) can generally be seen to supercede technologies characterised by more variable flake and blade manufacture (MSA), although the adoption of changes was by no means synchronous across the whole continent (Mitchell, 2008). On the basis of these considerations, we would suggest that the MSAeLSA template can also be applied to North Africa but that the introduction of bladelet technologies was considerably delayed in the Maghreb and their appearance by w22 ka Cal BP and subsequent spread may partly be explained by the demographic expansion of sub-clades of U6. However, it does leave several questions unanswered: why did these innovations emerge when they did in the Maghreb, and did they arise in response to palaeoclimatic shifts (e.g., Greenland Stadial 2), or were these innovations linked to subsequent demographic rise or the result of the influx of new peoples into this region following the disappearance of the MSA? These and other questions can ultimately only be answered by focusing research on other sites in the Maghreb similar to Taforalt, which have long sequences that cover a comparable time span. Conclusions In this paper we have presented the first high precision record of AMS dates for the Late Pleistocene Maghreb, providing a framework for understanding the development of the Iberomaurusian, the oldest backed bladelet LSA technology in northwest Africa. In examining the dating evidence at Taforalt, several gaps in the sequence were noted, including one of possibly as little as 1900 calendar years separating the first appearance of the Iberomaurusian at 22.0e21.4 ka Cal BP from the underlying nonLevallois flake technology, tentatively attributed to the MSA. A further gap in dating (but this time also coinciding with an erosive unconformity) can be seen between the earliest Iberomaurusian industry with ‘Ouchtata’ retouched blades (IB1) and one above containing microlithic backed bladelets (IB2). The duration of this gap may have been of the order of one to two thousand years and confirms that fully developed microlithic components had emerged in the Iberomaurusian by 15.5e15.0 ka Cal BP. A major accumulation of ashy midden deposits can be identified at Taforalt at 15.2e 14.2 ka Cal BP and use of the cave in the Iberomaurusian was shown to have continued until about 12.6 ka Cal BP (Tables 1 and 2). Assessing the relationship of these gaps to potential environmental shifts has been possible with reference both to cave sedimentological data and palaeoclimatic records for the western Mediterranean. Using broad comparisons with available oceanic and atmospheric records, it has been observed that the disappearance of the latest MSA at around 24 ka Cal BP may have been coincident with pronounced cooling of Atlantic and western Mediterranean waters in Heinrich Event 2 (Penaud et al., 2010), which would have produced a marked increase in continental aridity (Moreno et al., 2005). In contrast, the emergence of the Iberomaurusian seems to have taken place against the relatively warm and moist conditions indicated for the LGM at these latitudes (Penaud et al., 2010). At Taforalt, finer silts were observed in the intervening archaeologically sterile sediments between the proposed MSA and LSA levels that have been interpreted as a signal of increased regional aridity. Higher in the sequence, the erosive hiatus separating the Iberomaurusian phases IB1 and IB2 can be shown to coincide with Heinrich Event 1, although oceanic studies have suggested that this was not necessarily as severely arid as the earlier Heinrich episode (Penaud et al., 2010) and might be correlated with increased marine palaeoproductivity (Rodrigo-Gámiz et al., 2011). The latter could have had major beneficial consequences for populations living on and near the coast. Finally, we do R.N.E. Barton et al. / Journal of Human Evolution 65 (2013) 266e281 not believe it purely fortuitous that the timing of midden accumulation at Taforalt (Grey Series) occurred so close to the beginning of Greenland Interstadial 1e. During this period, the projected northwards migration of the Inter-Tropical Convergence Zone and increased monsoon activity would have brought much wetter conditions to this area (Rodrigo-Gámiz et al., 2011). Such circumstances might explain or may have given rise to some behavioural changes that occurred in the Iberomaurusian in this period (e.g., evidence of increased dietary breadth, greater sedentism, use of the deeper part of the cave site as a cemetery). In conclusion, while present genetic models for the dispersal of humans in North Africa are largely inconclusive, we would suggest that the new dating evidence supports an independent origin of the Iberomaurusian in the Maghreb (certainly by 22 ka Cal BP if not earlier). The latter implies innovation and transmission of new ideas that may have arisen in the Maghreb at a time of environmental instability or that were transmitted via rapid population movements from an area fringing the Maghreb. However, it does not seem likely that the impetus for change came from areas in the south on the Atlantic margin where only younger phases of the Iberomaurusian are so far known (e.g., the Agadir region: Bouzouggar, Personal observation). Parallel developments may have led to the appearance of backed bladelet technologies in Libya and Cyrenaica, but here these seem to have been more deeply rooted in the Dabban. Clearly this picture may change in the light of fresh evidence but our proposal for the MSA-LSA transition in the Maghreb provides a robust model that is capable of extensive testing. Acknowledgements We would like to thank Alison Wilkins and Michael Athanson for their help in producing Figs. 1 and 2, and to Michel Grenet for the artefact drawings in Fig. 6. We are also grateful to Peter Mitchell and Phillip Endicott for their comments on earlier drafts of this paper and to two anonymous reviewers for useful points of clarification and criticism. This project has been funded by the British Academy, Oxford University, Protars P32/09-CNRST (Morocco), NERC (NER/T/S/2002/00700 and NE/E015670/1) and the Leverhulme Trust (F/08 735/F). References Ambrose, S.H., 1998. Chronology of the Later Stone Age and food production in East Africa. J. Archaeol. Sci. 25, 377e392. Antoine, M., 1937. Notes de préhistoire marocaine: XIII e la question atéroibéromaurusienne au Maroc: historique et mise au point, 11è année. Bull. Soc. Préhist. Maroc., 45e58. Barker, G., Basell, L., Brooks, I., Cartwright, C., Cole, F., Davison, J., Farr, L., Grün, R., Hamilton, R., Hunt, C., Inglis, R., Jacobs, Z., Legge, T., Leitch, V., Morales, J., Morley, I., Morley, M., Pawley, S., Pryor, A., Roberts, R., Reynolds, T., el-Rishi, H., Simpson, D., Twati, M., van der Veen, M., 2008. The Cyrenaican Prehistory Project 2008: the second season of investigations of the Haua Fteah cave and its landscape, and further results from the initial 2007 fieldwork. Libyan Stud. 39, 175e223. Barker, G., Antoniadou, A., Armitage, S., Brooks, I., Candy, I., Connell, K., Douka, K., Drake, N., Farr, L., Hill, E., Hunt, C., Inglis, R., Jones, S., Lane, C., Lucarini, G., Meneely, J., Morales, J., Mutri, G., Prendergast, A., Rabett, R., Reade, H., Reynolds, T., Russell, N., Simpson, D., Smith, B., Stimpson, C., Twati, M., White, K., 2010. The Cyrenaican Prehistory Project 2010: the fourth season of investigations of the Haua Fteah cave and its landscape, and further results from the 2007e2009 fieldwork. Libyan Stud. 41, 63e88. Barich, B.E., Garcea, E.A.A., 2008. Ecological patterns in the Upper Pleistocene and Holocene in the Jebel Gharbi, Northern Libya: chronology, climate and human occupation. Afr. Archaeol. Rev. 25, 87e97. Barton, R.N.E., Bouzouggar, A., Bronk-Ramsey, C., Collcutt, S.N., Higham, T.F.G., Humphrey, L.T., Parfitt, S., Rhodes, E.J., Schwenninger, J.L., Stringer, C.B., Turner, E., Ward, S., 2007. Abrupt climatic change and chronology of the Upper Palaeolithic in northern and eastern Morocco. In: Bar-Yosef, O., Mellars, P., Stringer, C., Boyle, K. (Eds.), Rethinking the Human Revolution: New Behavioural and Biological Perspectives on the Origins and Dispersal of Modern 279 Humans. Research Monographs of the Macdonald Institute, Cambridge, pp. 177e186. Belfer-Cohen, A., Goring-Morris, A.N., 2003. Current issues in Levantine Upper Palaeolithic research. In: Goring-Morris, A.N., Belfer-Cohen, A. (Eds.), More than Meets the Eye. Studies on Upper Palaeolithic Diversity in the Near East. Oxbow Books, Oxford, pp. 1e12. Blondel, J., Aronson, J., 1999. Biology and Wildlife of the Mediterranean Region. Oxford University Press, Oxford. Bond, G., Broecker, W.S., Johnsen, S.J., McManus, J., Labeyrie, L., Jouzel, J., Bonani, G., 1993. Correlations between climate records from North Atlantic sediments and Greenland ice. Nature 365, 143e147. Bordes, F., 1976e1977. Moustérien et Atérien. Quaternaria 19, 19e34. Bouzouggar, A., Barton, R.N.E., 2012. The identity and timing of the Aterian in Morocco. In: Hublin, J.-J., McPherron, S. (Eds.), Modern Origins: a North African Perspective. Springer, Dordrecht, pp. 93e105. Bouzouggar, A., Barton, R.N.E., Collcutt, S.N., Parfitt, S., Higham, T.F.G., Rhodes, E., Gale, R., 2006. Le Paléolithique Supérieur au Maroc: apport des sites du NordOuest et de l’Oriental. In: Sanchidrián, J.-L., Márquez, A., Fullola, J.M. (Eds.), La Cuenca Mediterránea durante el Paleolítico Superior (38.000 e 10.000 años). IV Simposio de Prehistoria Cueva de Nerja. Fundación Cueva de Nerja & Málaga UISPP, Málaga, pp. 138e150. Bouzouggar, A., Barton, N., Vanhaeren, M., d’Errico, F., Collcutt, S., Higham, T., Hodge, E., Parfitt, S., Rhodes, E., Schwenninger, J.L., Stringer, C., Turner, E., Ward, S., Moutmir, A., Stambouli, A., 2007. 82,000 year-old shell beads from North Africa and implications for the origins of modern human behavior. Proc. Natl. Acad. Sci. 104, 9964e9969. Bouzouggar, A., Barton, R.N.E., Blockley, S., Bronk-Ramsey, C., Collcutt, S.N., Gale, R., Higham, T.F.G., Humphrey, L.T., Parfitt, S., Turner, E., Ward, S., 2008. Reevaluating the age of the Iberomaurusian in Morocco. Afr. Archaeol. Rev. 25, 3e19. Brand, S.A., Fisher, E.C., Hildebrand, E.A., Vogelsang, R., Ambrose, S.H., Lesur, J., Wang, H., 2012. Early MIS 3 occupation of Mochena Borago rockshelter, Southwest Ethiopian Highlands: Implications for Late Pleistocene archaeology, paleoenvironments and modern human dispersals. Quatern. Int. 274, 38e54. Brock, F., Higham, T., Ditchfield, P., Bronk Ramsey, C., 2010. Current pretreatment methods for AMS radiocarbon dating at the Oxford Radiocarbon Accelerator Unit (ORAU). Radiocarbon 52, 103e112. Bronk Ramsey, C., 2008. Deposition models for chronological records. Quatern. Sci. Rev. 27, 42e60. Bronk Ramsey, C., 2009a. Bayesian analysis of radiocarbon dates. Radiocarbon 51, 337e360. Bronk Ramsey, C., 2009b. Dealing with outliers and offsets in radiocarbon dating. Radiocarbon 51, 1023e1045. Bronk Ramsey, C., 2010. OxCal Program, vol. 4.1.7. Radiocarbon Accelerator Unit, University of Oxford, UK. <http://c14.arch.ox.ac.uk/embed.php?File¼oxcal. html>. Bronk Ramsey, C., Lee, S., 2013. Recent and planned developments of the program OxCal. Radiocarbon. in press. Camps, G., 1974. Les Civilisations Préhistoriques de l’Afrique du Nord et du Sahara. Doin, Paris. Clark, J.D., 1959. The Prehistory of Southern Africa. Pelican, London. Clark-Balzan, L.A., Candy, I., Schwenninger, J.-L., Bouzouggar, A., Blockley, S., Nathan, R., Barton, R.N.E., 2012. Coupled U-series and OSL dating of a Late Pleistocene cave sediment sequence, Morocco, North Africa: Significance for constructing Palaeolithic chronologies. Geochronology 12, 53e64. Close, A.E., 1980. Current research and recent radiocarbon dates from northern Africa. J. Afr. Hist. 21, 145e167. Close, A.E., 1986. The place of the Haua Fteah in the Late Palaeolithic of North Africa. In: Bailey, G.N., Callow, P. (Eds.), Stone Age Prehistory. Studies in Memory of Charles McBurney. Cambridge University Press, Cambridge, pp. 169e180. Close, A.E., 1988. Current research and recent radiocarbon, dates from northern Africa, III. J. Afr. Hist. 29, 145e176. Close, A.E., 2002. Backed bladelets are a foreign country. Archeol. Pap. Am. Anthropol. Assoc. 12, 31e44. Close, A.E., Wendorf, F., 1990. North Africa at 18,000 BP. In: Gamble, C., Soffer, O. (Eds.), The World at 18,000 BP. Unwin Hyman, London, pp. 41e57. Combourieu Nebout, N., Turon, J.-L., Zahn, R., Capotondi, L., Londeix, L., Pahnke, K., 2002. Enhanced aridity and atmospheric high-pressure stability over the western Mediterranean during the North Atlantic cold events of the past 50 k.y. Geology 30, 863e866. Courty, M.-A., Goldberg, P., Macphail, R.I., 1989. Soils and Micromorphology in Archaeology. Cambridge University Press, Cambridge. Dansgaard, W., Johnsen, S.J., Clausen, H.B., Dahl-Jensen, D., Gundestrup, N.S., Hammer, C.U., Hvidberg, C.S., Steffensen, J.P., Sveinbjornsdottir, A.E., Jouzel, J., Bond, G., 1993. Evidence for general instability of past climate from 250-kyr icecore record. Nature 364, 218e220. Debénath, A., 2000. Le peuplement préhistorique du Maroc: données récentes et problèmes. L’Anthropologie 104, 131e145. Debénath, A., 2003. Le Paléolithique supérieur de Maghreb. Praehistoria 3, 259e280. Debénath, A., Raynal, J.-P., Roche, J., Texier, J.-P., Ferembach, D., 1986. Stratigraphie, habitat, typologie at devenir de l’Atérien Marocain: Donneés récentes. L’Anthropologie 90, 233e246. Endicott, P., Ho, S.Y.W., Metspalu, M., Stringer, C., 2009. Evaluating the mitochondrial timescale of human evolution. Trends Ecol. Evol. 24 (9), 515e521. Ferembach, D., 1985. On the origin of the Iberomaurusians. A new hypothesis. J. Hum. Evol. 14, 393e397. 280 R.N.E. Barton et al. / Journal of Human Evolution 65 (2013) 266e281 Ferembach, D., Dastugue, J., Poitrat-Targowla, M.J., 1962. La Nécropole Epipaléolithique de Taforalt (Maroc Oriental): Étude des Squelettes Humains. Edita Casablanca, Rabat. Fletcher, W.J., Sánchez-Goñi, M.F., 2008. Orbital- and sub-orbital-scale climate impacts on vegetation of the western Mediterranean basin over the last 48,000 yr. Quat. Res. 70, 451e464. Garcea, E.A.A., 2004. Crossing deserts and avoiding seas: Aterian North Africane European relations. J. Anthropol. Res. 60, 27e53. Gliganic, L.A., Jacobs, Z., Roberts, R.G., Domínguez-Rodrigo, M., Mabulla, A.Z.P., 2012. New ages for Middle and Later Stone Age deposits at Mumba rockshelter, Tanzania: optically stimulated luminescence dating of quartz and feldspar grains. J. Hum. Evol. 62, 533e547. Gobert, E.G., Vaufrey, R., 1932. Deux gisements extrêmes de l’Ibéromaurusian. L’Anthropologie XVII, 449e490. Godfrey-Smith, D.I., Vaughan, K.B., Gopher, A., Barkai, R., 2003. Direct luminescence chronology of the Epipalaeolithic Kebaran site of Nahal Hadera V, Israel. Geoarchaeology 18, 461e475. Goetz, C., 1941. Note d’Archéologie préhistorique Nord-Africaine sur un foyer oranien de la sablière d’El-Kçar. Bull. Soc. Préhist. Fr 38, 262e265. Gonzalez, A.M., Larruga, J.M., Abu-Amero, K.K., Shi, Y., Pestano, J., Cabrera, V.M., 2007. Mitochondrial lineage M1 traces an early human backflow to Africa. BMC Genomics 8, 223. Goring-Morris, A.N., Belfer-Cohen, A. (Eds.), 2003. More than Meets the Eye: Studies on Upper Palaeolithic Diversity in the Near East. Oxbow Books, Oxford. Humphrey, L.T., Bocaege, E., 2008. Tooth evulsion in the Maghreb: chronological and geographical patterns. Afr. Archaeol. Rev. 25, 109e123. Humphrey, L., Bello, S.M., Turner, E., Bouzouggar, A., Barton, N., 2012. Iberomaurusian funerary behaviour: evidence from Grotte des Pigeons, Taforalt, Morocco. J. Hum. Evol. 62, 261e273. Iovita, R.P., 2009. Re-evaluating connections between the Early Upper Paleolithic of Northeast Africa and the Levant: technological differences between the Dabban and the Emiran. In: Shea, J.J., Lieberman, D. (Eds.), Transitions in Prehistory: Essays in Honor of Ofer Bar-Yosef. Oxbow Books, Oxford, pp. 125e142. Irish, J.D., 2000. The Iberomaurusian enigma: North African progenitor or dead end? J. Hum. Evol. 39, 393e410. Jacobs, Z., Roberts, R.G., Nespoulet, R., El Hajraoui, M.A., Debénath, A., 2012. Singlegrain OSL chronologies for Middle Palaeolithic deposits at El Mnasra and El Harhoura 2, Morocco: Implications for Late Pleistocene human environment interactions along the Atlantic coast of northwest Africa. J. Hum. Evol. 62, 377e394. Jiménez-Espejo, F.J., Martínez-Ruiz, F., Finlayson, C., Paytane, A., Sakamoto, T., Ortega-Huertasa, M., Finlayson, G., Iijimaf, K., Gallego-Torres, D., Fa, D., 2007. Climate forcing and Neanderthal extinction in southern Iberia: insights from a multiproxy marine record. Quatern. Sci. Rev. 26, 836e852. Klein, R.G., 1974. Environment and subsistence of prehistoric man in the southern Cape Province, South Africa. World Archaeol. 5, 249e284. Lahr, M.M., 1996. The Evolution of Modern Human Cranial Diversity: a Study in Cranial Variation. Cambridge University Press, Cambridge. Linstädter, J., Eiwanger, J., Mikdad, A., Weniger, G.-C., 2012. Human occupation of Northwest Africa: a review of Middle Palaeolithic to Epipalaeolithic sites in Morocco. Quatern. Int. 274, 158e174. Lowe, J., Barton, N., Blockley, S., Ramsey, C.B., Cullen, V.L., Davies, W., Gamble, C., Grant, K., Hardiman, M., Housley, R., Lane, C.S., Lee, S., Lewis, M., MacLeod, A., Menzies, M., Müller, W., Pollard, M., Price, C., Roberts, A.P., Rohling, E.J., Satow, C., Smith, V.C., Stringer, C.B., Tomlinson, E.L., White, D., Albert, P., Arienzo, I., Barker, G., Bori c, D., Carandente, A., Civetta, L., Ferrier, C., Guadelli, J., Karkanas, P., Koumouzelis, M., Müller, U.C., Orsi, G., Pross, J., Rosi, M., Shalamanov-Korobar, L., Sirakov, N., Tzedakis, P.C., 2012. Volcanic ash layers illuminate the resilience of Neanderthals and early modern humans to natural hazards. Proc. Natl. Acad. Sci. 109, 13532e13537. Lubell, D., 2001. Late Pleistocene Early Holocene Maghreb, Africa. In: Peregrine, P.N., Ember, M. (Eds.), 2001. Encyclopedia of Prehistory, vol. 1. Kluwer, New York, pp. 129e149. Maca-Meyer, N., González, A.M., Pestano, J., Flores, C., Larruga, J.M., Cabrera, V.M., 2003. Mitochondrial DNA transit between West Asia and North Africa inferred from U6 phylogeography. BMC Genet. 4, 15. Mannino, M.A., Thomas, K.D., Leng, M.J., Di Salvo, R., Richards, M.P., 2011. Stuck to the shore? Investigating prehistoric hunter-gatherer subsistence, mobility and territoriality in a Mediterranean coastal landscape through isotope analyses on marine mollusc shell carbonates and human bone collagen. Quatern. Int. 244, 88e104. Mannino, M.A., Catalano, G., Talamo, S., Mannino, G., Di Salvo, R., Schimmenti, V., Lalueza-Fox, C., Messina, A., Petruso, D., Caramelli, D., Richards, M.P., Sineo, L., 2012. Origin and diet of the Prehistoric hunter-gatherers on the Mediterranean island of Favignana (Egadi Islands, Sicily). Plos One 7, e49802. Marks, A.E., 1975. The current status of Upper Palaeolithic studies from the Maghreb to the northern Levant. In: Wendorf, F., Marks, A.E. (Eds.), Problems in Prehistory: North Africa and the Levant. Southern Methodist University Press, Dallas, pp. 439e458. Marks, A.E., 2003. Reflections on Levantine Upper Palaeolithic studies: past and present. In: Goring-Morris, A.N., Belfer-Cohen, A. (Eds.), More than Meets the Eye. Studies on Upper Palaeolithic Diversity in the Near East. Oxbow Books, Oxford, pp. 249e264. McBrearty, S., Brooks, A.S., 2000. The revolution that wasn’t: a new interpretation of the origin of modern human behavior. J. Hum. Evol. 39, 453e563. McBurney, C.B.M., 1967. The Haua Fteah (Cyrenaica) and the Stone Age of the SouthEast Mediterranean. Cambridge University Press, Cambridge. McBurney, C.B.M., 1977. Archaeology and the Homo sapiens sapiens Problem in Northern Africa. Derde Kroon-Voordracht. Gehouden voor de stichtung Nederlands Museum voor Anthropologie en Praehistorie, Amsterdam. Mitchell, P.J., 2002. The Archaeology of Southern Africa. Cambridge University Press, Cambridge. Mitchell, P.J., 2008. Developing the archaeology of marine isotope stage 3. S. Afr. Archaeol. Soc. Goodwin Ser. 10, 52e65. Mitchell, P.J., Barham, L.S., 2008. The First Africans: African Archaeology from the Earliest Toolmakers to Most Recent Foragers. Cambridge University Press, Cambridge. Moreno, A., 2012. A multiproxy paleoclimate reconstruction over the last 250 kyr from marine sediments: the northwest African margin and the western Mediterranean sea. In: Hublin, J.-J., McPherron, S.P. (Eds.), Modern Origins: a North African Perspective. Vertebrate Paleobiology and Paleoanthropology. Springer, Dordrecht, pp. 3e18. Moreno, A., Cacho, I., Canals, M., Prins, M.A., Sánchez-Goñi, M.F., Grimalt, J.O., Weltje, G.J., 2002. Saharan dust transport and high-latitude glacial climatic variability: the Alboran Sea record. Quatern. Res. 58, 318e328. Moreno, A., Cacho, I., Canals, M., Grimalt, J.O., Sánchez-Goñi, M.F., Shackleton, N., Sierro, F.J., 2005. Links between marine and atmospheric processes oscillating on a millennial timescale. A multi-proxy study of the last 50,000 yr from the Alboran Sea (Western Mediterranean Sea). Quatern. Sci. Rev. 24, 1623e1636. Neeley, M.P., Barton, C.M., 1994. A new approach to interpreting late Pleistocene microlith industries in Southwest Asia. Antiquity 68, 275e288. Olivieri, A., Achilli, A., Pala, M., Battaglia, V., Fornarino, S., Al-Zahery, N., Scozzari, R., Cruciani, F., Behar, D.M., Dugoujon, J.-M., Coudray, C., SantachiaraBenerecetti, A.S., Semino, O., Bandelt, H.-J., Torroni, A., 2006. The mtDNA legacy of the Levantine early Upper Palaeolithic in Africa. Science 314, 1767e1770. Olszewski, D., Schurmans, U., Schmidt, B., 2011. The Epipaleolithic (Iberomaurusian) from Grotte des Contrebandiers, Morocco. Afr. Archaeol. Rev. 28, 97e123. Pallary, P., 1909. Instructions pour la Recherche Préhistorique dans le Nord-Ouest de l’Afrique. In: Mémoires de la Société Historique Algérienne, vol. 3. Jourdan, Algiers. Penaud, A., Eynaud, F., Turon, J.L., Blamart, D., Rossignol, L., Marret, F., LopezMartinez, C., Grimalt, J.O., Malaizé, B., Charlier, K., 2010. Contrasting paleoceanographic conditions off Morocco during Heinrich events (1 and 2) and the Last Glacial Maximum Original. Quatern. Sci. Rev. 29, 1923e1939. Pennarun, E., Kivisild, T., Metspalu, E., Metspalu, M., Reisberg, T., Behar, M.D., Jones, C.S., Villems, R., 2012. Divorcing the Late Upper Palaeolithic demographic histories of mtDNA haplogroups M1 and U6 in Africa. BMC Evol. Biol. 12, 234. Pereira, L., Silva, N.M., Franco-Duarte, R., Fernandes, V., Pereira, J.B., Costa, M.D., Martins, H., Soares, P., Behar, D.M., Richards, M.B., Macaulay, V., 2010. Population expansion in the North African Late Pleistocene signalled by mitochondrial DNA haplogroup U6. BMC Evol. Biol. 10, 390. Phillips, J., 1972. North Africa, the Nile Valley, and the problem of the Late Palaeolithic. Curr. Anthropol. 13, 587e590. Raynal, J.-P., 1980. Taforalt. Mission préhistorique et paléontologique française au Maroc: rapport d’activité pour l’année. Bull. Archéol. Marocaine 12, 69e71. Richter, D., Moser, J., Nami, M., Eiwanger, J., Mikdad, A., 2010. New chronometric data from Ifri n’Ammar (Morocco) and the chronostratigraphy of the Middle Palaeolithic in the Western Maghreb. J. Hum. Evol. 59, 672e679. Roche, J., 1953. Note préliminaire sur les fouilles de la grotte de Taforalt (Maroc Oriental). Hespéris 40, 89e116. Roche, J., 1963. L’Epipaléolithique Marocaine. Fondation Calouste Gulbenkian, Lisbon. Roche, J., 1967. L’Aterian de la grotte de Taforalt (Maroc oriental). Bull. Archeol. Marocaine 7, 11e56. Roche, J., 1969. Les industries paléolithiques de la grotte de Taforalt (Maroc oriental). Quaternaria 11, 89e100. Roche, J., 1976. Cadre chronologique de l’Epipaléolithique marocain. In: Actes du IXè Congrès de l’UISPP: Chronologie et Synchronisme dans la Préhistoire CircumMéditerranéenne, pp. 153e167. Rodrigo-Gámiz, M., Martínez-Ruiz, F., Jiménez-Espejo, F.J., Gallego-Torres, D., NietoMoreno, V., Romero, O., Ariztegui, D., 2011. Impact of climate variability in the western Mediterranean during the last 20,000 years: oceanic and atmospheric responses. Quatern. Sci. Rev. 30, 2018e2034. Roset, J.-P., Harbi-Riahi, M., 2007. El Akarit: Un Site Archéologique du Paléolithique Moyen dans le Sud de la Tunisie. Editions Recherches sur les Civilisations, Paris. Sánchez-Goñi, M.F., 2006. Intéractions végétation-climat au cours des derniers 425 000 ans en Europe occidentale. Le message du pollen des archives marines. Quaternaire 17, 3e25. Sánchez-Goñi, M.F., Cacho, I., Turon, J.-L., Guiot, J., Sierro, F.J., Peypouquet, J.-P., Grimalt, J.O., Shackleton, N.J., 2002. Synchroneity between marine and terrestrial responses to millennial scale climatic variability during the last glacial period in the Mediterranean region. Clim. Dyn. 19, 95e105. Saxon, E.C., Close, A., Cluzel, C., Morse, V., Shackleton, N.J., 1974. Results of recent excavations at Tamar Hat. Libyca 22, 49e91. Scally, A., Durbin, R., 2012. Revising the human mutation rate: implications for understanding human evolution. Nat. Rev. Genet. 13, 745e753. Schild, R., Wendorf, F., 2010. Late Palaeolithic hunter-gatherers in the Nile Valley of Nubia and Upper Egypt. In: Garcea, E.E.A. (Ed.), South-Eastern Mediterranean Peoples between 130,000 and 10,000 Years Ago. Oxbow Books, Oxford, pp. 89e125. Schwenninger, J.-L., Collcutt, S.N., Barton, R.N.E., Bouzouggar, A., El Hajraoui, M.A., Nespoulet, R., Debénath, A., Clark-Balzan, L., 2010. Luminescence chronology for R.N.E. Barton et al. / Journal of Human Evolution 65 (2013) 266e281 Aterian cave sites on the Atlantic coast of Morocco. In: Garcea, E.E.A. (Ed.), South-Eastern Mediterranean Peoples between 130,000 and 10,000 Years Ago. Oxbow Books, Oxford, pp. 18e36. Sepulchre, P., Ramstein, G., Kageyama, M., Vanhaeren, M., Krinner, G., SánchezGoñi, M.-F., D’Errico, F., 2007. H4 abrupt event and late Neanderthal presence in Iberia. Earth Planet. Sci. Lett. 258, 283e292. Staff, R.A., Bronk Ramsey, C., Bryant, C.L., Brock, F., Payne, R.L., Schlolaut, G., Marshall, M.H., Brauer, A., Lamb, H.F., Tarasov, P., Yokoyama, E., Haraguchi, T., Gotanda, K., Yonenobu, H., Nakagawa, T., Suigetsu 2006 Project Members, 2011. New 14C determinations from Lake Suigetsu, Japan: 12,000 to 0 cal BP. Radiocarbon 53, 511e528. Taylor, V.K., Barton, R.N.E., Bell, M., Bouzouggar, A., Collcutt, S., Black, S., Hogue, J.T., 2011. The Epipalaeolithic (Iberomaurusian) at Grotte des Pigeons (Taforalt), Morocco: a preliminary study of the land Mollusca. Quatern. Int. 244, 5e14. Texier, J.-P., Huxtable, J., Rhodes, E., Miallier, D., Ousmoi, M., 1988. Nouvelles données sur la situation chronologique de l’Atérien du Maroc et leurs implications. C.R. Acad. Sci. Paris 307, 827e832. Tixier, J., 1963. Typologie de L’Épipaléolithique du Maghreb. Arts et Métiers Graphiques, Paris. Van Peer, P., 2004. Did Middle Stone Age moderns of sub-Saharan African descent trigger an Upper Palaeolithic revolution in the Lower Nile Valley? Anthropologie 42, 215e226. Vermeersch, P.M., 1992. The Upper and Late Palaeolithic of northern and eastern Africa. In: Klees, F., Kuper, R. (Eds.), New Light on the Northeast 281 African Past: Current Prehistoric Research. Heinrich-Barth-Institut, Köln, pp. 100e153. Vermeersch, P.M., 2010. Middle and Upper Palaeolithic in the Egyptian Nile valley. In: Garcea, E.E.A. (Ed.), South-Eastern Mediterranean Peoples between 130,000 and 10,000 Years Ago. Oxbow Books, Oxford, pp. 66e88. Villa, P., Soriano, S., Tsanova, T., Deganof, I., Higham, T.F.G., d’Errico, F., Backwell, L., Lucejko, J.J., Colombini, M.P., Beaumont, P.B., 2012. Border Cave and the beginning of the Later Stone Age in South Africa. Proc. Natl. Acad. Sci. 109, 13208e13213. Wadley, L., 1997. Rose Cottage Cave: archaeological work 1987 to 1997. S. Afr. J. Archaeol. Sci. 93, 439e444. Ward, S., 2007. Reconstructing Late Pleistocene Vegetation Dynamics from Archaeological Cave Sites in the Western Mediterranean: Links with Climate and Cultural Changes. Ph.D. Dissertation. University of Oxford. Weisrock, A., Wengler, L., Mathieu, J., Ouammou, A., Fontugne, M., Mercier, N., Reyss, J.-L., Valladas, H., Guery, P., 2006. Upper Pleistocene comparative OSL, U/ Th and 14C datings of sedimentary sequences and correlative morphodynamical implications in the South-Western Anti-Atlas (Oued Noun, 29 N, Morocco). Quaternaire 17, 45e59. Wengler, L., 2010. A new relationship between Mousterian and Aterian in North Africa. In: Conard, N.J., Delagnes, A. (Eds.), 2010. Settlement Dynamics of the Middle Palaeolithic and Middle Stone Age, vol. III. Kerns Verlag, Tübingen, pp. 67e80. Wrinn, P.J., Rink, W.J., 2003. ESR dating of tooth enamel from Aterian levels at Mugharet el ‘Aliya (Tangier, Morocco)’. J. Archaeol. Sci. 30, 123e133.