Quaternary International 300 (2013) 32e47

Contents lists available at SciVerse ScienceDirect

Quaternary International

journal homepage: www.elsevier.com/locate/quaint

Human dispersal across diverse environments of Asia during the Upper Pleistocene

Nicole Boivin a, Dorian Q. Fuller b, Robin Dennell c, Robin Allaby d, Michael D. Petraglia a,* a School of Archaeology, Research Laboratory for Archaeology and the History of Art, University of Oxford, Dyson Perrins Building, South Parks Road, Oxford OX1 2HU, UK b Institute of Archaeology, University College London, 31-34 Gordon Square, London WC1H 0PY, UK c Department of Archaeology, University of Sheffield, Northgate House, West Street, Sheffield S1 4ET, UK d School of Life Sciences, University of Warwick, Wellesbourne, Warwick CV35 9EF, UK article info abstract

Article history: The initial out of Africa dispersal of Homo sapiens, which saw anatomically modern humans reach the Available online 15 January 2013 Levant in Marine Isotope Stage 5, is generally regarded as a ‘failed dispersal’. Fossil, archaeological and genetic findings are seen to converge around a consensus view that a single population of H. sapiens exited Africa sometime around 60 thousand years ago (ka), and rapidly reached Australia by following a coastal dispersal corridor. We challenge the notion that current evidence supports this straightforward model. We argue that the fossil and archaeological records are too incomplete, the coastal route too problematic, and recent genomic evidence too incompatible for researchers not to remain fully open to other hypotheses. We specifically explore the possibility of a sustained exit by anatomically modern humans, drawing in particular upon palaeoenvironmental data across southern Asia to demonstrate its feasibility. Current archaeological, genetic and fossil data are not incompatible with the model presented, and appear to increasingly favour a more complex out of Africa scenario involving multiple exits, varying terrestrial routes, a sub-divided African source population, slower progress to Australia, and a degree of interbreeding with archaic varieties of Homo. Ó 2013 Elsevier Ltd and INQUA. All rights reserved.

1. Introduction at a time of improved climatic conditions in Marine Isotope Stage 5 (MIS 5) (Frumkin et al., 2011). In the absence of early fossil finds Archaeological discoveries in the Levant have long played a further afield and in light of certain primitive morphological fea- critical role in debates about human evolution. Situated at the tures, the Levantine MIS 5 exit has generally been seen as a failed western edge of the Asian landmass, the small region has, since the dispersal (Mellars, 2006a, 2006b; Shea, 2008; Oppenheimer, 2009, late 1920s, yielded key fossil finds of both Homo sapiens and Homo 2012). This notion is offered support by recent molecular analyses neanderthalensis (Shea, 2003, 2008). Interpretation of the Levantine of mitochondrial DNA (mtDNA), which suggest a successful sites has never been straightforward, however, with problems of expansion of H. sapiens out of Africa only at around 60 thousand chronology and a shared Mousterian technology, for example, years ago (ka) (e.g. Macaulay et al., 2005; see below). generating numerous opportunities for controversy. While findings The apparent agreement of archaeological and genetic findings in the region were initially taken to support a gradual transition has led to the emergence of a consensus view regarding the dis- between H. neanderthalensis and H. sapiens, improved chronologies persal of modern humans out of Africa. This view holds that have indicated significant overlap between the two types of human, H. sapiens left Africa in MIS 4 or early MIS 3, and colonized Australia on the order of 25e50,000 years. This recognition has seen the within perhaps as little as several thousand years by rapidly skirt- Levantine fossil remains come to be drawn upon as a major strand ing the coasts of the Indian Ocean. While this may be a reasonable of evidence in support of a replacement model for modern human hypothesis, its uncritical and increasingly widespread multi- origins, with the Levantine H. sapiens fossils now widely inter- disciplinary acceptance suggests the need for more careful preted as representing a first foray of modern humans out of Africa, appraisal. Here we will first examine current fossil, archaeological and genetic findings concerning the dispersal of H. sapiens from Africa, critically appraising these sources of evidence. We then * Corresponding author. consider how this evidence sits in relation to palaeoenvironmental E-mail address: [email protected] (M.D. Petraglia). evidence and a hypothesized dispersal of modern humans across

1040-6182/$ e see front matter Ó 2013 Elsevier Ltd and INQUA. All rights reserved. http://dx.doi.org/10.1016/j.quaint.2013.01.008 N. Boivin et al. / Quaternary International 300 (2013) 32e47 33 southern Asia. As we discuss below, a number of sites in the Arabian Most pressing of all, however, is the problem of identifying Peninsula and South Asia have begun to yield Middle Palaeolithic straightforward signatures of modern human behaviour. In assemblages dating to MIS 5, suggesting that H. sapiens may have accordance with traditional European-derived models, the arrival moved into these geographic zones during climatic ameliorations. of modern humans in a particular region was once seen to be traced The archaeological and palaeoenvironmental evidence is supported relatively easily, through the appearance in the archaeological by new genetic findings that challenge the orthodox view that record arrival of evidence for art, symbolism, and complex Upper H. sapiens replaced archaic species of Homo in Asia. While there Palaeolithic technologies, as seen in Europe around 45e40 ka. seems little need to invoke a multiregional model for the evolution Recent years have, however, seen a thorough deconstruction of of H. sapiens to explain these new genetic findings, more nuanced the notion of a straightforward ‘modern human revolution’. It has models of population structure and a degree of assimilation are been recognised that so-called modern human behaviours, as required to explain a suite of recent autosomal and ancient DNA traced through their apparent archaeological manifestations: 1) do studies (to which we return to below). Overall, the picture of not necessarily arrive as part of a package; 2) do not necessarily modern human origins, which seemed at one point to be crystal- exhibit continuity through time; and, most significantly, 3) are not lising into a fairly comfortable orthodoxy, is now looking decidedly always present in modern human populations. There is nothing in more complex (Dennell and Petraglia, 2012). The time is ripe for terms of material culture attributes that is unique to H. sapiens and consideration of alternatives to a straightforward MIS 4 or 3 exit of, also universal among modern humans in Africa, Eurasia and Aus- and complete replacement by, modern humans, and in particular to tralasia. It is thus the case that with material culture as with fossils, examine the plausibility of earlier exits across southern Asia during absence of evidence need not necessarily indicate evidence of MIS 5. absence. Contrary to the notion of the sudden revolutionary appearance 2. Absence of fossils and artefacts of a new behavioural package with the emergence of modern humans, a major synthesis of behavioural indices for Middle Stone Evidence for the late dispersal of modern humans out of Africa, Age Africa revealed the gradual and non-contiguous emergence of and replacement of local archaic species of Homo, is by no means modern traits (McBrearty and Brooks, 2000). Other parallel syn- straightforward to interpret. Outside the sphere of molecular theses for South Asia (James and Petraglia, 2005) and Sahul genetics, the story rests entirely on evidence from fossils and arte- (Brumm and Moore, 2005; O’Connell and Allen, 2007; Habgood and facts. Each class of evidence has its own problems, but both suffer in Franklin, 2008) have revealed a similar trend. The Australian evi- particular from regional biases in archaeological research. While dence is particularly interesting in that H. sapiens appears to have industrialised regions with long histories of archaeological research arrived without an African ‘package’ of innovations, and to have like Europe and the Levant have seen more comprehensive study, acquired these independently and piecemeal. Examination of the equally if not more critical regions like the Arabian Peninsula, South South-East Asian record has also demonstrated the inadequacy of Asia, East Asia and Southeast Asia have received relatively uneven traditional markers of modern human behaviour, such as refined coverage, and in some cases, minimal investigation. blade and bladelet technology, body ornamentation and mobiliary The earliest well-dated fossils of modern humans outside of and parietal art (Barker et al., 2007; Moore and Brumm, 2007), and Africa, excluding the Levant, are in Laos (c. 46 ka) (Demeter et al., has focused instead on the appearance of, for example, new sub- 2012), Sunda (c. 46 ka) (Barker et al., 2007), Australia (c. 40 ka) sistence strategies and modes of engagement with the landscape (Bowler et al., 2003), Sri Lanka (c. 38 ka) (Deraniyagala, 1992; Perera with the arrival of H. sapiens (Summerhayes et al., 2010). Such et al., 2011) and India (c. 20 ka) (Kennedy, 2000). In Arabia, modern studies have demonstrated the Eurocentric biases of models of human fossils have not been dated to earlier than the Holocene behavioural modernity, which created a set of global expectations (Martin, 2007). Taken at face value, these dates might be inter- from a regional record. In reality, there is no single material trait preted to indicate a dispersal of modern humans from east to west that is inevitable in modern human assemblages. rather than from out of Africa. The paucity of well dated fossils has What this means for Out of Africa models is that cognitive not, however, prevented palaeoanthropologists from drawing modernity may have emerged simultaneously or even before meaning from the observed patterns. The earliest dated fossils from (Kingdon, 1993; Foley and Lahr, 1997) morphological modernity, Sunda and Sahul are frequently taken to suggest not a minimal date and the appearance of traditional indicators of behavioural mod- for the exit of modern humans out of Africa, but an approximate ernity in the archaeological record may mark demographic or social date for Out of Africa. Since all the other dates for intervening changes (Shennan, 2001; Petraglia et al., 2009; Powell et al., 2009), regions in the west are younger, and thus clearly unreliable as rather than cognitive ones. These new ideas about modern human indicators of initial dispersal, the dates from Sunda and Australia, in behaviour suggest the possibility that H. sapiens may have been particular, have come to assume significance in attaching a chro- fully modern back to 200 ka or more, and exited Africa prior to the nology to the narrative of Out of Africa. It needs to be emphasised, appearance of, or without bringing along, more complex tech- however, that the earliest fossils do not necessarily date Asia’s first nologies or archaeologically-attested symbolic practices. We can- modern human colonization, let alone reflect a date for the exit of not therefore rely on these forms of material evidence to trace H. sapiens from Africa, contrary to some influential models that dispersals any more than we can rely on fossils as indicators of the have been proposed. first anatomically modern humans to leave Africa. New under- Artefact assemblages also present their own challenges. standings of behavioural modernity furthermore open up the Archaeological sites are subject to alterations and transformations possibility that the H. sapiens populations that exited Africa and are in depositional contexts, and these play out exceptionally harshly in documented in the Levant in MIS 5 may have been fully cognitively the hyper-arid and tropical environments that today dominate and behaviourally modern (contra Mellars, 2006b). some of the most plausible out of Africa dispersal routes. Even the most well-preserved objects, stone tools, are often difficult to find 3. Coastal route out of Africa? in well stratified, datable contexts e the Arabian Peninsula, for example, has yielded a wealth of Palaeolithic surface finds Another problem with the current consensus view for Out of (Petraglia and Alsharekh, 2003), but until recently, very few Africa is its increasing focus on a coastal route of dispersal. Part of stratified assemblages (Groucutt and Petraglia, 2012). the problem is that two related but separate hypotheses have 34 N. Boivin et al. / Quaternary International 300 (2013) 32e47 frequently been conflated. One argues that, along with a northern subsequently of an intensity sufficient to lead to local over- exit towards Europe and Siberia around 45 ka, H. sapiens also fol- exploitation and to ultimately spur dispersal. Not only is exploita- lowed a more southerly route out of Africa at an earlier date, dis- tion of marine resources sufficiently simple to have been persing through southern Asia (Lahr and Foley, 1994, 1998). The competently undertaken by archaic humans (Cortés-Sánchez et al., other suggests that modern humans ‘coasted’ out of Africa 2011) and even non-human primates (Malaivijitnond et al., 2007), (Stringer, 2000; Bulbeck, 2007), travelling along coastal corridors the probability that humans were at this early stage marine spe- around the Indian Ocean rim to reach Australia by 45 ka. The latter cialists is, based on ethnographic and archaeological parallels, idea has a long history, appearing in various manifestations since extremely low (Bailey, 2009). The many inland Upper Pleistocene Sauer (1962) suggested the potential role of coastlines as corridors sites in Arabia, South Asia, East Asia and Southeast Asia argue much of dispersal for early humans. Subsequent discussions by Kingdon more for a more flexible subsistence approach in which terrestrial (1993), and by Field and colleagues (Field and Lahr, 2005; Field plant and animal resources continued to play a major role in human et al., 2007), explored the possibilities of a southern, coastally- lifestyles. Even early modern human sites that are described as focused dispersal route for modern humans out of Africa, all or ‘coastal’ may furthermore demonstrate a terrestrial focus. Despite part of the way along the Indian Ocean rim. The idea of a coastal being one of the Upper Pleistocene human occupation sites closest route has been embraced by a number of geneticists, who have to the coast, for example, Niah Cave in Borneo possesses a faunal drawn upon it to explain certain patterns in the genetic record of assemblage dominated by terrestrial and freshwater taxa, and lacks human diversity (Macaulay et al., 2005; Thangaraj et al., 2005; evidence for the use of marine resources (Barker et al., 2007). Oppenheimer, 2009). Coastal sites may indeed be lacking from the record due to sub- The gist of the coastal dispersal hypothesis (which should be mergence along shallow continental shelves in many areas, distinguished from the more generic southern dispersal hypoth- although surveys in areas with tectonically raised reef formations esis) is that maritime environments offered particularly favourable and former shorelines have not produced evidence for marine- environments for human occupation, and provided a dispersal oriented archaeological sites either (Korisettar, 2007; Bailey, corridor that enabled rapid movement along the Indian Ocean rim 2009). Even if we grant that most sites are submerged, the pres- of initial colonising populations. This single dispersal event corre- ence of many inland sites stresses the importance of terrestrial sponded with the arid conditions of MIS 4 (Lahr and Foley, 1994; foraging adaptations over the marine specialization that is implicit Stringer, 2000; Forster and Matsumura, 2005; Mellars, 2006a; in coastal dispersal models, particularly those that imply a rapid Bulbeck, 2007; Oppenheimer, 2009), when desert dominated in dispersal where repeated overexploitation of resources led to northern Africa, the Arabian Peninsula, and the northwestern part movements further down the coast. of the Indian subcontinent. Such a model argues that humans were Equally problematic is the idea that coasts were particularly rich able to draw upon the rich resources of coastal environments, and attractive environments for early humans. While certain manifesting a maritime adaptation that led them to rapidly skirt coastal environments are favourable habitats, other coastal areas continents (Oppenheimer, 2009), and to develop the seafaring are effectively ‘deserts’ lacking entirely in critical freshwater technology that eventually enabled colonisation of Sahul (Forster resources. The notion of a ‘coastal oasis’ e the idea that freshwater and Matsumura, 2005). Research on the Red Sea coast of Eritrea springs appeared on emerged continental shelves at times of low and in coastal South Africa has suggested human adaptation to a sea level (Faure et al., 2002 ) e is central, often implicitly, to most coastal marine environment by MIS 5, and indicated to some that models of coastal dispersal, yet lacks empirical support (Bailey, the emergence of modern human behaviour may be linked to a 2009). Even if valid, the coastal oasis model does not postulate more intense and complex exploitation of marine resources the kind of uniformly equable coastline that such a dispersal (Deacon, 1989; Walter et al., 2000; Marean, 2010). Several recent necessitates (Sauer, 1962; Stringer, 2000). Instead, it suggests the genetic studies of mtDNA diversity, meanwhile, have suggested a probability of rather dispersed and non-interconnected wetland single exit of modern humans out of Africa, as well as a rapid dis- environments in places where underground springs exist. There are persal to Australia (Forster, 2004; Forster and Matsumura, 2005; also many coastal regions wholly devoid of shellfish or any obvious Macaulay et al., 2005; Oppenheimer, 2009). These features have marine sources that could be easily gathered or otherwise obtained. been seen as consonant with a coastal route, both because coasts Furthermore, the idea that coasts were, in contrast to inland areas, are increasingly viewed as a kind of ‘highway ’, enabling rapid dis- particularly stable environments is open to question. Westley and persal, but also because desert expansion in many inland regions Dix (2006) summarise evidence for coastal instability during the during the relevant time period would have created barriers to Upper Pleistocene, and emphasise that present day coastlines, and dispersal, forcing populations to move along coasts. our ideas about coasts, are the product of relatively unique envi- Despite the popularity of the coastal dispersal hypothesis, there ronmental circumstances during the exceptionally stable Holocene are numerous reasons why the idea should be approached with a interglacial. greater degree of caution. One is that archaeological evidence In sum, while coastal regions may have been used as dispersal indicative of a maritime adaptation in early humans is extremely routes in prehistory, both for H. sapiens and earlier hominid exits limited. At present, relatively secure evidence for use of marine out of Africa, the notion of a coastal superhighway between Africa resources by early H. sapiens appears in South Africa (Marean, and Australia that enabled unexpectedly fast dispersal is simplistic 2010). Shell beads found in North Africa (d’Errico et al., 2009) and derives little if any support from empirical data. While prob- indicate visitation to the coast and not coastal adaptation neces- ably a key necessary corollary of the currently popular notions of an sarily. The claim for an association between marine resources and MIS 4 exit for Out of Africa, and a rapid dispersal to Australia, it may archaeological sites in Eritrea (Walter et al., 2000) also include lag be argued that the coastal dispersal model has a number of sig- deposits and natural death assemblages (Bruggemann et al., 2004), nificant deficits and unproven assumptions. which do not necessarily demonstrate human subsistence activities (Bailey, 2009). Even if we accept that a maritime adaptation is 4. The genetics of Out of Africa plausible, however, problems remain with the model. Bailey (2009) has noted the implausibility of the idea that the emergence of If fossil and archaeological evidence offer little support for a modern human cognition ushered in a sudden awareness of the rapid coastal dispersal of H. sapiens out of Africa at c. 60 ka, genetic richness of coastal resources, whose consumption was data is equally problematic. While a consensus date of c. 60 ka is N. Boivin et al. / Quaternary International 300 (2013) 32e47 35 often cited for the mtDNA data, recent literature actually offers a Plagnol and Wall, 2006) (but see Li et al., 2008) studies suggest that range of dates for human expansion out of Africa, anywhere populations in Africa were structured, with numerous small and between 85 and 45 ka (Oppenheimer, 2003, 2009, 2012; Macaulay separately evolving populations present in the continent at the et al., 2005; Kivisild et al., 2006; Cabrera et al., 2009; Endicott et al., time of the out of Africa dispersal. Metapopulation models, which 2009; Soares et al., 2009; Rasmussen et al., 2011). Variation is due, argue for the presence of multiple populations, and their absorp- in part, to the values used to calibrate the mtDNA clock (e.g., Scally tion of some admixture before neighbouring extinctions, are more and Durbin, 2012). The problems associated with calibrating the compatible with the wide range of TMRCA estimates for autosomal mtDNA clock have been noted, including in particular the reliance loci than the straightforward single origin model of out of Africa on problematic assumptions about the timing of the divergence (Harding and McVean, 2004). Cranial data offer support for a more between humans and chimpanzees, as well as the constancy of the complex out of Africa scenario involving multiple dispersals, molecular clock (Ho and Larson, 2006; Endicott et al., 2009). Fur- probably by varying routes, at different times and/or from a sub- ther research is required to improve confidence in molecular esti- divided African population (Lahr, 1996; Schillaci, 2008; Gunz et al., mates of human evolutionary timescales. 2009). Another under-appreciated issue is the anomalous nature of the Archaic admixture in the modern human genome in Eurasia is genetic evidence for a rapid spread of modern humans from Africa now clear. While studies of mtDNA (Richards et al., 1996; Krings to Asia. Echoing the fossil date anomaly, the mtDNA branch lengths et al., 1997; Handt et al., 1998; Serre et al., 2004; Green et al., for sampled populations are longest for those which are farthest 2008) and the Y-chromosome (Jobling and Tyler-Smith, 2003) east, in Near Oceania, and shortest in the Asian areas that have not indicated admixture, for example, it was widely recog- would have been encountered first (Merriwether et al., 2005; nized that autosomal DNA might tell a different story. Indeed this Oppenheimer, 2009). The real problem, however, is that the varia- has proven to be the case, with evidence of archaic ancestry sur- tion in branch lengths suggests that a single genotype engaged in facing for the microcephalin gene (Evans et al., 2006), the tau the expansion actually existed for 30 ka, which does not support a (MAPT) locus (Hardy et al., 2005), the b-globin gene (Harding et al., rapid expansion. The anomaly can be explained by what we call an 1997), the immunity genes CD209 and CD209L (Barreiro et al., ‘M buffer’ effect (see Supplementary material A) which implies that 2005), the dystrophin gene (Labuda et al., 2000) and others (Wall the branch ages we observe are considerable underestimates of the et al., 2009). Direct contact between archaic and modern humans time of arrival of the genotype to these areas. Such anomalously is also supported by autosomal data for modern human head lice, long-lived genotypes have been directly observed through ancient Pediculus humanus (Reed et al., 2004). More recently, a draft DNA in species such as the Iberian lynx (Rodríguez et al., 2011). sequence of the Neanderthal genome has been published that Much more problematic, however, is the challenge posed by indicates gene flow from Neanderthals into the ancestors of non- new genomic studies that offer perspectives on evolutionary his- Africans (Green et al., 2010). The study also demonstrated that tory that sometimes contrast markedly with those of mitochondrial Neanderthals shared more genetic variants with present-day and Y-chromosome-based investigations. Recent analysis of the humans in Eurasia than with present-day humans in sub-Saharan frequency of alleles of SNPs (single nucleotide polymorphisms) Africa, suggesting that the gene flow occurred before the diver- from the International HAPMAP Project, for example, suggests that gence of Eurasian groups from one another. The main known region the out of Africa expansion may actually need to be modeled as two in which H. sapiens and H. neanderthalensis co-existed prior to the or more bottlenecks, rather than the single bottleneck suggested by separation of European and Asian lineages of human is the Levant mtDNA studies (Keinan et al., 2007; Wall et al., 2009). The authors (but see Shea, 2008). Given that the overlap between these species of one study (Keinan et al., 2007) suggest that a two bottleneck in the Levant may have lasted only until c. 75 ka, this might be seen model provides the potential best fit to the SNP data, with the first as supporting an exit in MIS 5. However, as our distribution map for at w140e80 ka, potentially reflecting the MIS 5 exit identified in Neanderthals (Fig. 1; Supplementary material B) suggests, a broad the Levant, and the second dating to the Last Glacial Maximum and swath of western and central Eurasia may have been home to potentially associated with demographic upheavals at this time. A Neanderthals from MIS 5 until MIS 3, suggesting a range of possible bottleneck dating to anywhere around 60 ka is not presently sup- regions of co-occupation. Subsequent regionally specific admixture ported by such studies. is suggested by the presence of ancient alleles that are found only or Even more challenging is the presence of autosomal genes predominantly in certain regions e for example, certain European- whose estimated ancestry extends well beyond the conventionally localised alleles for microcephalin in Europe (Evans et al., 2006), accepted date for the speciation of modern humans. While a single and Asia-localised alleles for b-globin dating to >200 ka (Harding origin model predicts that all shared polymorphisms trace their et al., 1997) e and by the presence of basal clades composed origins back to a single deme in Africa, several regions of the entirely of Asian sequences, as for the RRM2P4 region (Cox et al., nuclear genome, both on the X chromosome and on the autosomes, 2008). The Neanderthal findings were soon after accompanied by have deep ancestries that are incompatible with a simple, single a second hominin draft genome, that of the Denisovans of Siberia origin model (Garrigan and Hammer, 2006). These include, for (Krause et al., 2010; Meyer et al., 2012). The finding of gene flow example, the RRM2P4 region of the X chromosome that has an from Denisovans into some modern human populations of Mela- estimated Time of the Most Recent Common Ancestor (TMRCA) of nesia suggest that admixture occurred after the dispersal of 2.33 million years and an apparent East Asian ancestry (Cox et al., H. sapiens into Asia, and probably somewhere in South or Southeast 2008) and the microcephalin D allele, which has a TMRCA of 1.1 Asia (Stewart and Stringer, 2012). million years (Evans et al., 2006). These kinds of findings do not Recent revision of the human genomic mutation rate provides support a straightforward single origin model (see also Harding another line of nuclear DNA evidence that problematizes mtDNA- et al., 1997; Eswaran et al., 2005; Harpending and Eswaran, 2005; based Out of Africa estimates, while at the same time offering a Relethford, 2008; Weaver and Roseman, 2008), and have been chronology that supports evidence for modern human and archaic interpreted as reflecting population structure within Africa and/or admixture. Direct measurements of the nuclear genomic mutation interbreeding with archaic species of Homo outside of Africa. Cer- rate in contemporary humans using next generation sequencing tain mitochondrial (Behar et al., 2008) and autosomal (Harris and technology have indicated a value that is approximately half of that Hey, 1999; Labuda et al., 2000; Harding and McVean, 2004; Satta previously estimated from fossil calibration (Scally and Durbin, and Takahata, 2004; Barreiro et al., 2005; Garrigan et al., 2005; 2012). When applied to the analysis of the time of separation of 36 N. Boivin et al. / Quaternary International 300 (2013) 32e47

Fig. 1. Hominin populations during MIS 5. H. sapiens was probably the main hominin resident in Africa by this time. Neanderthals prior to MIS 5 were distributed in Europe and the Levant, but expanded their range across Asia thereafter (Howell, 1999). H. erectus sensu stricto was probably confined to East Asia by MIS 5 as the Ngandong H. erectus assemblage now appears to have been older (Indriati et al., 2011). Because of the lack of diagnostic skeletal evidence from Arabia and South Asia between 125 and 40 ka, the identity of their inhabitants is unknown, but, we argue, likely to have been H. sapiens. The identification of H. sapiens in Laos helps to support this argument. Note the possible areas of overlap between H. sapiens and the southern limits of Neanderthals and H. erectus which would allow for interbreeding e as indicated by recent studies and possibly by fossil evidence from Zhirendong and other Chinese caves. The distribution of “Denisovans” is unknown as they are indicated only by the ancient DNA of a phalange from Denisova Cave, Siberia, and modern genetic studies of SE Asians and Melanesians. Present indications are that they are probably an East Asian sister population of Neanderthals, but little more is known at present. For Neanderthals, the purple area denotes their likely “core” distribution. The adjoining blue areas show their likely maximal extent, northwards in warmer episodes, and southwards when conditions in the core area deteriorated, or when opportunities arose for expansion. Their easternmost distribution towards NE China and the Pacific coast of NE Asia is shown as uncertain. Filled circles: H. sapiens: 1) Jebel Irhoud, Morocco; 2) Skuhl, Qafzeh, Israel; 3) Herto, Ethiopia; 4) Niah, Borneo; Filled stars: Neanderthals: 5) Tabun, Kebara (Israel); 6) Shanidar (Iraq); 7) Teschik Tasch (Uzbekistan); 8) Denisova, Okladinov (Russian Siberia); Open Stars (other): 9) Hatnora (India) (Homo species indeterminate); 10) Ngandong (H. erectus, Middle Pleistocene). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

African and non-African populations, the revised rate yields an Out anatomically modern humans outside of Africa. Fossil evidence at of Africa chronology in the range of 130e90 ka, suggesting that the Skhul and Qafzeh indicates their presence in the Levant from per- Levantine MIS 5 artefactual evidence represents either permanent haps as early as 130e120ka(Grun et al., 2005; Shea, 2008), i.e., the or repeated temporary occupation by modern humans (Scally and early part of MIS 5 (5e). Durbin, 2012: 748). Younger estimates of Out of Africa on the The implications of the Levantine evidence are hotly debated. basis of mtDNA data may, it is argued, derive from processes of later On the one hand, certain researchers describe the Levantine gene flow and drift/selection and/or complex demographic factors occupation by H. sapiens in MIS 5 as localized and temporary, and including bottlenecks (Scally and Durbin, 2012: 751). argue that these humans were not entirely cognitively modern. Though citing occupation dates that span approximately 40e 5. Evaluating Out of Africa 20,000 years, Mellars (2006b) describes the Levantine occupation as a ‘short-lived event’. The Levantine Mousterian, often charac- The earliest fossils of H. sapiens are identified in East Africa at c. terized by centripetal Levallois reduction methods and a variety of 195e160 ka (White et al., 2003; McDougall et al., 2005), though tool types (e.g. points, scrapers, notches, burins), is viewed as craniometric studies of African forms described as our species show archaic technology, and inadequate to the task of adaptive com- considerable phenotypic diversity (Gunz et al., 2009). Fossil evi- petition with Neanderthals, who are perceived to have replaced dence thus suggests that the emergence of H. sapiens occurred modern humans in this region (Shea, 2003, 2008; Mellars, 2006b). towards the end of MIS 7, just prior to the onset of the colder and Others, however, note the likelihood that Neanderthals and mod- drier conditions of MIS 6. The return to these conditions led to an ern humans overlapped for a significant time period in MIS 5 in the expansion of deserts, likely creating barriers to dispersal out of Levant, and argue for elements of modern human behavior in spite Africa of this new species. If speciation estimates are correct, this of the apparently ‘limited’ technology (Grun et al., 2005). The therefore means that other than a brief window at the end of MIS 7, recent Neanderthal genome findings (Green et al., 2010) may there may have been no real opportunity for modern humans to support the notion of an overlap. Furthermore, the suggestion that leave Africa until MIS 5 (though see below; Dennell, 2009: 469). At use of Middle Palaeolithic technologies reflects a less competitive this time, territorial ranges in Africa appear to have expanded in H. sapiens, which had to await full cognitive modernity and the association with environmental amelioration, and Middle Stone related development of prismatic blades and more advanced pro- Age sites seem to be relatively abundant and well distributed in jectile weaponry to outcompete Neanderthals, would seem to East Africa (Basell, 2008) and the Saharan belt in MIS 5 (Smith et al., reflect a traditional bias in the interpretation of the material record. 2007; Hill, 2009), providing the basis for a dispersal both within The variable appearance of advanced lithic technologies among and out of Africa (Osborne et al., 2008; Drake et al., 2011). As Asian populations of H. sapiens indicates the strong likelihood that indicated, it is indeed in MIS 5 that we find the first evidence for stone tool technologies rather reflect particular adaptive responses N. Boivin et al. / Quaternary International 300 (2013) 32e47 37 to environmental circumstances and demographic parameters. The 2010; Ozkan et al., 2010). So it is not necessarily surprising to find shift to Early Upper Palaeolithic technology in the Levant appears to that as further research is conducted, the picture of Out of Africa have been associated with a reduction in prey size (Shea, 2008), has become more complex. Continued research supports an origin suggesting that either demographic pressure (Stiner et al., 1999, of modern humans in Africa, and both expansion out of Africa and 2000) or environmental change, rather than cognitive develop- replacement of local populations. However, the apparent simplicity ment, drove the appearance of new technologies. A demographic of this model is tempered by plausible evidence for a structured argument has recently been put forward for the appearance of African population, multiple exits, archaic admixture and the Upper Palaeolithic technology and other elements of material cul- existence of an increasing number of late archaic survivals in Asia. ture in western Eurasia, in this case based around the idea that These considerations, together with data suggesting that archaeo- population needs to be of a certain density to retain cultural logical sites in southern Asia correspond with modern humans innovations (Powell et al., 2009). Available evidence suggests that extending back earlier than appreciated, suggest the need to parallels to Upper Palaeolithic technology do not emerge in South explore other models for Out of Africa. Below we present argu- Asia until c. 35 ka, which is much too late to reflect an initial ments for an earlier and more complex dispersal pattern for expansion of modern humans, and here as well, demographic modern humans, drawing in particular on relevant palaeoclimatic, arguments have been forwarded for the appearance of more vegetation, faunal and genetic data and a set of maps that recon- advanced technologies (Petraglia et al., 2009). struct vegetation patterns across Eurasia in MIS 4 and MIS 5 (see The first out of Africa exit by H. sapiens is seen as part of a Figs. 2e5, Table 1, Supplementary material C). limited range expansion of North African human and faunal pop- ulations (Tchernov, 1992; Lahr and Foley, 1994; Rabinovich and Tchernov, 1995; Tchernov, 1996). Mellars (2006b) and other Table 1 Summary of vegetation zones mapped in Figs. 2e5. scholars assume that this is a localized human expansion based on the absence of similar fossils further afield. Yet, as discussed ear- Vegetation zone Resource expectations lier, the fossil record is clearly biased; the absence of modern 1. Mediterranean vegetation. Abundant fruits and nuts and seeds, human fossils in Arabia for MIS 5 is meaningless given that fossils including wild almonds, acorns, Pistacia. of Homo do not appear in the regional record until the Holocene. 2. Mediterranean park-steppe, Many edible grain resources, some nuts Meanwhile, a recent multivariate analysis of cranial morphometric the classic ‘Fertile Crescent’ and fruits as above. Large herbivore data suggests affinity between early modern humans from the vegetation famous for wild herds. Levant and terminal Pleistocene/early Holocene human pop- cereal progenitors such as ulations from Australasia that is potentially reflective of a suc- wheat and barley. 3. The dry steppe of the Middle Some edible seeds, but generally lower cessful MIS 5 exit (Schillaci, 2008). From a palaeoenvironmental East and southern Central Asia. value and higher effort than above, and perspective, it is difficult to identify factors that might have pre- some tubers. Herbivore herds smaller vented dispersals further afield. Dispersal barriers both into and than above. within the Arabian Peninsula are at a minimum during MIS 5, 4. Desert Minimal edible plant and animal resources. No/low human especially in MIS 5e and MIS 5a, when humidity and rainfall were populations expected. higher (Vaks et al., 2007; Fleitmann and Matter, 2009; Preusser, 5. Sub-Desert/Sahel vegetation, Small seeded grasses and forbs occur, 2009) and when terrestrial environments are marked by lakes generally of the Sudano- while seeds and fruit of many trees and and river systems (Parker, 2009; Petit-Maire et al., 2010; Sindhian floristic province shrubs are available. Rosenberg et al., 2011). Genetic (mtDNA) data suggest that the Herds of herbivores occur but these are often of smaller herd size and/or body dispersal from Africa to Arabia of at least one other primate, the size than for zones 2 or 7. Hamadryas baboon, in MIS 5 (Fernandez, 2009), though we would 6. Riverine corridors: marshes and Wild grains of grasses, sedges, and approach this chronology with caution in light of calibration and gallery forests. other aquatics as well as tubers are other issues. Glacial/stadial episodes, like MIS 4 and most of MIS 6, frequent. Small game, birds and fish would have been characterized by reduced monsoon rainfall and are readily available, with some larger game, especially in dry season. greatly expanded deserts (the Sahara, Arabia and Thar deserts), 7. Tropical savannah/woodland- Wild grass grains and other seeds are with the desert barriers much more extensive and uninviting than grass mosaic, including the numerous, including wild progenitors those of the present day. By contrast the interstadials, especially tropical evergreen zones of of most tropical cereals; some tubers the more pronounced upswings of MIS 5e and MIS 5a would have India. are available and numerous edible seasonal fruits. Large herbivore herds. seen reduced deserts, which would furthermore have been sig- 8. Dry tropical woodland, including Numerous fruits, nuts and tubers. nificantly impinged upon by richer vegetation zones. dry deciduous zones of India. Edible seeds of herbaceous plants include wild pulses. Large herbivores 6. Modeling environments and movements are frequent but in small groups. Smaller game present but may be hard to catch. Despite the popularity in recent years of an Out of Africa model 9. Moist tropical woodland and Wild grass grains and other seeds are featuring a single exit, a date of c. 60 ka for a rapid dispersal, and the grassland mosaic, in particular numerous, including seasonal wild rice complete replacement of archaic species, it is clear that this model that of eastern India and Ganges in local wetland; some tubers are is in need of revision given new datasets from genetics, archae- plain in which Dipterocarpaceae available and numerous edible seasonal including Shorea are frequent; fruits. Large herbivore herds. Smaller ology, and palaeoanthropology. This parallels conceptual develop- this zone extends to Southeast game present but may be hard to catch. ments in domestication studies, where genetics and archaeology Asia. initially supported a rapid, localized model for agricultural origins, 10. Moist tropical woodlands, Wild seeds are sparse in time and space; but gradually came to recognise more complex and protracted including both moist deciduous tree-nuts and fruits may be locally and processes involving multiple domestication events for many crops zones like those in India and seasonally abundant but are likely to be true tropical rainforests. unpredictable in time and space. Larger (Brown et al., 2009). Ancient DNA studies, both of domesticated game is rare and smaller game is harder plants and animals, and humans, have also begun to reveal the to catch. Very low human population importance of recent demographic processes, and their ability to densities expected. mask earlier genetic patterns (Haak et al., 2005; Larson et al., 2007, (continued on next page) 38 N. Boivin et al. / Quaternary International 300 (2013) 32e47

Table 1 (continued ) Meher-Homji, 2001). Despite this, many quantitative models of fl Vegetation zone Resource expectations past vegetation essentially atten the earth and focus on temper- 11. Tropical montane vegetation. Some edible fruits, seeds and tubers, ature and rainfall in relation to latitude and longitude (e.g., Adams but generally only localized and and Faure, 1997; Prentice et al., 2000). For such quantitative mod- seasonal. Few large game, and hard els, the diversity of vegetation zones is greatly simplified even for to catch small game. Low human modern baseline maps. We have avoided this here through a less population densities expected. quantitative, if admittedly more subjective approach, drawing 12. Warm temperate hill and Many seasonal edible fruits and nuts, sub-montane vegetation, including acorns. Tubers occur and vegetation maps that take into account regional and topographic including the vegetation of some edible grains and seeds. Large variations evidenced in modern maps. In addition, we have taken the Himalaya foothills, and herbivores are frequent but in small into account modern disjunct distributions of species or sister the evergreen broadleaf forests groups; smaller game present but species, since such disjunction imply past periods of more or less of southern China. may be hard to catch. continuous distributions between two regions, such as between 13. High elevation coniferous Few seasonal nuts and fruits. Minimal forests and desert/the Tibetan seed or tuber resources. Sparse smaller northeast of India and Sri Lanka or the savannas of Africa and India plateau. game. Low human population density (Asouti and Fuller, 2008:72e73). Such disjunctions have long been expected. recognized in plant geography classifications, such as the “Nubo- 14. Mixed evergreen-deciduous Many seasonal edible fruits and nuts, Sindian” and the “Eritreo-Arabian” floristic territories (e.g., Zohary, (mesophytic) forests. including acorns. Some edible seeds, “ ” fl but tubers rarer than in zone 12. Large 1973) or recognition of the Indo-African element in the ora of herbivores are frequent but in small India (e.g., Legris, 1963; Awasthi, 1995). The maps provided here groups; smaller game present but must therefore be regarded as initial hypotheses that require sys- may be hard to catch. tematic testing, through both more palaeoenvironmental proxy 15. Mixed conifer and temperate Seasonal edible fruits and nuts, but records and quantitative modeling. To our knowledge, such deciduous fewer than in zone 14, and with higher toxicity. Edible seeds, seasonal and reconstructions have not been previously attempted for these localized. Large herbivores frequent, periods and geographical scales. as well as hard to catch small game.

6.1. Taking steps eastwards Vegetation reconstructions have been made at the macro-scale, based on examining broad patterns of modern vegetation maps and Entry into the Arabian Peninsula from Africa was possible by relating these to available paleobotanical and palaeoclimatic evi- two main routes (see Fig. 2). The Sahelian vegetation of grasses and dence sources. Synoptic vegetation descriptions, even in the mod- shrubs would have extended far northwards into the Sahara and ern day, are idealized simplifications, although of heuristic value across much of Arabia. Such environments can be expected to have in recognizing that vegetation structure and composition are extended across the regions bordering the Red Sea, from Sudan’s strongly associated with climatic parameters of temperature and Jebel Elba through Egypt’s Jebel Abu Harba to the higher ground in rainfall (Huggett, 1995). Climatic parameters are regionally and the Sinai, providing both a corridor into the Mediterranean envi- locally affected by topography and geographical coordinates, as ronments of the Levant and a potential land route around the Red detailed maps of modern vegetation make clear (e.g., Wang, 1961; Sea (see also Fig. 3 in Petraglia et al., 2010). A Bab el Mandab route at

Fig. 2. Vegetation reconstruction for western Eurasia in MIS 5. See Table 1 for details of vegetation zones. Labelled sites are: (A) Qafzeh; (B) Es-Skhul; (C) Jubbah Palaeolake; (D) Shi’bat Dihya 1 (MIS 3); (E) Jebel Faya; (F) Aybut; (G) 16R Dune; (H) Jwalapuram; (I) Site 50. N. Boivin et al. / Quaternary International 300 (2013) 32e47 39

Fig. 3. Vegetation reconstruction for western Eurasia in MIS 4. See Table 1 for details of vegetation zones. Site labels are as for Fig. 2. the southern end of the Red Sea was possible, with parallel vege- correspond with favourable environmental conditions on the Ara- tation zones existing on either side of this strait. Palaeoshoreline bian Peninsula, such as at the transition between MIS 6 and 5e (e.g., reconstructions over the past 150 ka indicate that there were a Armitage et al., 2011). Once outside Africa, movement beyond the number of periods when short sea crossings (ranging from <5km Levant or southern Arabian Peninsula was likely to have been rel- to <15 km) across the Red Sea were possible, thus suggesting there atively straightforward in MIS 5. The disjunct distribution of is no reason to invoke seaworthy boats or seafaring technology and a number of tree and shrub species including many food species skills (Lambeck et al., 2012). Moreover, potential short sea crossings that humans might have relied upon across Africa, the Yemeni

Fig. 4. Vegetation reconstruction for eastern Eurasia in MIS 5. See Table 1 for details of vegetation zones. Labelled sites (continued from Figs. 2 and 3) are: (G) 16R Dune; (H) Jwalapuram; (I) Site 50; (J) Liujiang; (K) Huanglong; (L) Niah; (M) Huon; (N) Malakunanja; (O) Nauwalabila. 40 N. Boivin et al. / Quaternary International 300 (2013) 32e47

Fig. 5. Vegetation reconstruction for eastern Eurasia in MIS 4. See Table 1 for details of vegetation zones. Site labels are as for Fig. 4. highlands and South Asia suggests the likelihood that their ranges connection with H. sapiens in northeast Africa. Middle Palaeolithic have contracted and that they would have occurred in much more industries situated in northern Arabia, on the shores of the Jubbah frequent patches in MIS 5. This would have included taxa with Palaeolake, date to MIS 5 and MIS 7 (Petraglia et al., 2011, 2012a). edible fruits, such as drumstick trees (Moringa spp.), desert dates The Jubbah Palaeolake occurs in the Nefud Desert (Fig. 2), indicat- (Balanites aegyptiaca), sebesten plums (Cordia spp.), karira (Cap- ing that hominins penetrated this marginal environment in an paris decidua), and tamarind (Tamarindus indica) (see Asouti and ameliorated period, where phytolith studies have demonstrated Fuller, 2008). In addition, a similar range of edible wild millet- the presence of a grassland with some trees. grasses, wild legumes and, on forest margins, yams (especially The existence of two potential routes from Africa to Arabia, and Dioscorea bulbifera) and other tubers (e.g., Arisaema spp.) are shared the presence of several main dispersal corridors through the Ara- across such distributions. Attractive fauna for hunting, such as bian Peninsula, suggest the possibility for multiple exits and more gazelles and ostriches, would also have occurred through these complex dispersal scenarios. Favourable conditions in the Arabian savannah-Sahelian corridors. Thus humans adapted to the northern Peninsula, especially in the north, may have increased the range savannah woodlands and the Sahel in Africa would have been able overlap between Neanderthal and modern human populations, to expand through familiar environments out of Africa and across presenting opportunities for interbreeding (Fig. 1). The potential Arabia to the similar environments that framed a reduced Thar range of Neanderthals in MIS 5 likely extended well into present- Desert. Riverine systems would furthermore have flowed east day Iran, Afghanistan and Pakistan, and possibly south into the across much of Arabia, providing additional dispersal corridors Arabian Peninsula as well (Supplementary material B). (Edgell, 2006; Petraglia, 2007, 2011; Petraglia et al., 2012a). Lakes In contrast, MIS 4 saw widespread expansion of the desert in and rivers in such humid stages as MIS 5 were present in areas now northern Africa and the Arabian Peninsula (Fig. 3). This would have characterized as desert in Arabia (Parker, 2009; Petit-Maire et al., impeded exit out of Africa other than via the Bab el Mandab, where 2010; Rosenberg et al., 2011) and India (Achyuthan et al., 2007). coastal environments of unknown but potentially quite variable Archaeologically, such MIS 5 dispersal routes are not yet clearly quality would have been encountered. Attractive grassland envi- supported by artefactual evidence. Though Middle Palaeolithic sites ronments would have been located considerably inland from the are abundant in Arabia, the majority have been identified from coasts, and probably at high elevations, such as in the Yemeni surface contexts and remain undated (Petraglia and Alsharekh, highlands and the hills of southwest and eastern Oman peninsula. 2003; Groucutt and Petraglia, 2012). New information from Relict populations appear to be present at Jebel Faya and Shi’bat stratified archaeological sites, however, is beginning to emerge. Dihya across MIS 4 and extending into MIS 3, in both the Persian Jebel Faya rockshelter, situated about 55 km from the Persian Gulf, Gulf region and in the Yemeni uplands (Armitage et al., 2011; has been dated to c. 125 ka (MIS 5e) (Armitage et al., 2011). The Delagnes et al., 2012). Pollen data furthermore suggest that in drier recovery of cordiform handaxes, foliates, Levallois and discoidal periods like MIS 4, there was a reduction in the mangrove envi- cores and retouched tools is argued to be reminiscent of East ronments that often support rich coastal resources. Thus, although African industries, implying the early movement of H. sapiens into the lowered sea levels of MIS 4 would have made crossing the Bab Arabia. The most convincing case for a movement from Africa el Mandab (and Arabian/Persian Gulf) easier, the coastal deserts on comes with the recovery of abundant ‘Nubian’ sites situated along either side are likely to have presented few attractions to huntere riverine drainages in the Dhofar region of Oman (Rose et al., 2011). gatherers. This is not to say that exit was not possible out of Africa Here, a Nubian assemblage was dated to 106 ka (MIS 5c), suggesting in MIS 4 (see Field and Lahr, 2005; Field et al., 2007), only that it N. Boivin et al. / Quaternary International 300 (2013) 32e47 41 seems far less likely, and dependent on a number of coastal envi- the date of arrival of H. sapiens to the subcontinent. Archaeological ronmental parameters whose existence is debatable. evidence is more suggestive, particularly in light of recent findings. The recovery of numerous Middle Palaeolithic sites in many basins of 6.2. Passage to India peninsular India, together with stratified occurrences post-dating Late Acheulean assemblages (Misra, 2001), indicates that hominins Subsequent to dispersal out of Africa and across Arabia, palae- were present in the region for a considerable period of time. oenvironmental data suggest a number of potential routes that Archaeological excavations at Jwalapuram, in the Jurreru Valley of H. sapiens may have followed into the Indian subcontinent (Fig. 2). southern India demonstrate that Middle Palaeolithic hominins were These can be simplified into three broad routes, recognizing that presentbyc.78ka(Petraglia et al., 2007, 2012c). Analysis of cores from both hybrid and multiple routes are also possible and indeed likely. the Jwalapuram localities indicates affinities with sub-Saharan Afri- Perhaps the most feasible route is the Sahel (marked as 5 on Fig. 2) can MSA assemblages produced by H. sapiens (Clarkson et al., 2012). vegetation corridor, a route through environments that would have The morphology and technology of the Jwalapuram cores may be featured many seasonal seed foods and recurrent medium-sized distinguished from technologies manufactured by Neanderthals in game like gazelle and ostrich. The many disjunct tree species Southwest Asia and Europe. Though the technology remains poorly between the dry tropical woodlands of India and Africa imply that described, the potential presence of Middle Palaeolithic assemblages these vegetation zones were connected through intervening Arabia dated 130e109 ka in the 16R Dune in the Thar Desert (Achyuthan in wetter periods in the past. These disjuncts include important food et al., 2007; Singhvi et al., 2010) and c. 125e75 ka in the Iranamadu sources that could have contributed to subsistence (e.g., desert Formation of Sri Lanka (Deraniyagala, 1992; Abeyratne, 1996)also dates: Balanites, drumstick trees: Moringa, tamarind: Tamarindus). suggest MIS 5 occupations that may represent H. sapiens, although of The Sahel corridor route may have necessitated a sea crossing at the course an archaic species cannot be ruled out (Petraglia et al., 2010). Strait of Hormuz to present-day southern Iran, although a route While lithic technology does not provide a straightforward or defin- across or around the Persian Gulf is also feasible (Rose, 2010). A itive signature of modern human presence, in this case it does appear second possibility is that there was an expansion of H. sapiens from to necessitate that we consider the possibility that H. sapiens was the Levant, via the Tigris/Euphrates riverine corridor that would present in South Asia by MIS 5a, if not earlier. eventually become a major trade conduit in the region. Such riverine Palaeoenvironmentally, India appears to have been amenable to corridors are always rich in small game, seeds foods and tubers, human habitation in MIS 5 and into MIS 4 (Fig. 2). India had although they might require rather different adaptations than the favourable environments, such as broad areas of savannah (7) and Sahelian corridor. A third, perhaps less likely route is via the Indo- dry tropical woodlands (8), along with moist tropical woodland and Iranian plateau. This would have necessitated following the Medi- grassland mosaic (9) vegetation in the Ganges region. All of these terranean park-steppe (classic Fertile Crescent vegetation) into the environments support a diversity of bovids and cervids, larger Zagros Mountains (see Nasab et al., 2013). This vegetation possibly herbivores (e.g., elephant, rhino), and numerous food plants. They extended across parts of the Iranian plateau or down to coast, then may also be regarded as quite well-buffered, in the sense that cli- onto India. This is perhaps the least likely route for various reasons matic oscillations would have opened woodlands and shifted but involving: 1) topographic complexity; 2) the need to adapt to sig- not removed resources. While climatic fluctuation might have nificantly different environments and novel flora; and 3) the high extirpated local populations across much of more marginal envi- potential that Neanderthals dominated this region (see Fig. 1, ronments like Arabia, populations in India may have been able to Supplementary material B). The range overlap with Neanderthals is persist through such oscillations. The inherent ‘habitability’ of potentially important in light of the evidence for Neanderthal gene India, combined with a possible long-term modern human pres- flow into H. sapiens. Both of the latter two routes, and possibly also ence in the region, may go some way towards explaining the evi- the first, would likely have entailed ecological overlap with Nean- dence for higher long-term population size in India than in derthals. So it is possible that regions east and perhaps even south neighbouring regions (Atkinson et al., 2008 ; Petraglia et al., 2009). of the Levant would have provided opportunities for admixture If there was indeed an entry of H. sapiens into India in MIS 5, it is between modern humans and Neanderthals. possible, indeed probable, that modern humans encountered With a greatly reduced Thar Desert throughout MIS 5, entering archaic species in the region. A cranial fossil find from the Narmada the Indian subcontinent would have been relatively straightfor- River valley supports the presence of an archaic human species ward compared to an entry in MIS 4 (Petraglia et al., 2012b). The (initially identified as Homo erectus or Homo heidelbergensis,now Sahel corridor that originated in North Africa and extended along viewed as an indeterminate species of Homo) in the region some- the southern Arabian Peninsula, subsequently skirting the Iranian time in MIS 6 or 5 (Athreya, 2007; Patnaik et al., 2009). Acheulean and Makran coasts, also likely framed a reduced Thar desert, sites are certainly present in the early to middle stages of the Middle intergrading into the lusher savanna woodlands of the Deccan Pleistocene, indicating the long-term presence in the region of one plateau. This Sahel corridor thus provides a key potential and or more archaic species (Petraglia, 2010). Recent dating of Late previously unrecognized dispersal route between Africa and South Acheulean sites to c. 140e130 ka in the Son valley of northern India Asia. The Thar Desert itself need not necessarily have been a barrier (Haslam et al., 2011) demonstrates that at least one archaic species in MIS 5; palaeoenvironmental evidence from the desert in MIS 5 of Homo continued to inhabit India at the MIS 6/5 transition. The indicates a high percentage of C4 plants, increased summer rains coincidence with the earliest potential dates for Middle Palaeolithic and a strong monsoon (Achyuthan et al., 2007). And, indeed, the sites in the Thar Desert at 130e109 ka is obviously suggestive of an region can be expected to have had numerous rivers, and perhaps early human presence. It is possible that new competition from playa lakes at this time. Based on Early Holocene parallels, we recently arrived anatomically modern humans led or contributed to expect that while many playas were quite saline, a few lakes would archaic extinctions at this time. Admixture with H. sapiens prior to have been freshwater recharged by subsurface drainage to the east the extinction of archaic hominins cannot be ruled out. that was increased under stronger monsoons. Is there fossil or archaeological evidence to support the presence 6.3. Expansion into East and Southeast Asia of modern humans in South Asia in MIS 5? As indicated, fossils are lacking and skeletal remains of modern humans for India and Sri In many ways, the Indian subcontinent demonstrates significant Lanka are relatively young, and almost certainly tell us nothing about ecological similarity in comparison with the environments of 42 N. Boivin et al. / Quaternary International 300 (2013) 32e47

Africa. India is dominated by savannah and Sahel type vegetation, relatively late in East and Southeast Asia. These include Homo as well as vegetation zones whose differences from these are not floresiensis, which survived on Flores until perhaps 17 ka radical, and similar food resources are attested to by disjunct taxa. (Westaway et al., 2009), a possible Upper Pleistocene archaic The region just east of India, however, features significantly dif- population in South China (Curnoe et al., 2012) and an as yet ferent environments that may have posed new challenges to dis- unknown new hominin species from southern Siberia, which persing humans (Fig. 4). The northeast Indian region of Assam and existed as late as 50e30 ka (Krause et al., 2010). adjacent areas are characterized by high hills with dense tropical If expansion east from India was significantly slowed by ecotone forests, and seasonally flooded and swampy river valleys. Tropical change, then the opening up of this landscape in MIS 4 (Fig. 5) may rainforests generally support low human population densities (cf. have had a significant impact on H. sapiens dispersal patterns. Bird Bailey and Headland, 1991) and necessitate specific rainforest et al. (2004) note that in the tropics, glacial periods (and periods of adaptations (e.g., techniques for hunting hard to catch arboreal transition between interglacial and glacial periods) may have been game such as monkey, see Piper et al., 2008; Perera et al., 2011). times of radiation, while interglacial periods were times of con- Swamps may have supported lower populations due to higher solidation. Palaeoenvironmental data provide evidence that a cor- disease loads (cf. Sattenspiel, 2000), and if not true barriers would ridor from east India through to insular Southeast Asia effectively nonetheless have reduced local populations. The impression of a opened up at the beginning of MIS 4, as the moist tropical wood- major barrier between the regions of India and Indochina is rein- lands of western Indochina transitioned into a moist tropical forced by biogeographical and genetic studies suggesting that a woodland and grassland mosaic landscape. Further east, this cor- wide range of animals are differentiated or speciated across this ridor would likely have opened onto a broad tropical savannah/ region. A number of species have their range limits at the Brah- woodland-grass mosaic stretching across much of mainland maputra River, including various species of mongoose (Herpestes Southeast Asia. This has already been referred to as a ‘savannah edwardsii, and the crab-eating mongoose, Herpestes urva), the corridor’ and has been suggested to have constituted a key dis- small-toothed palm civet, the large Indian civet, the hog-badger, persal route into the region for modern humans (O’Connor and the Bengal fox, the Asian elephant, the pygmy hog, the hispid Veth, 2000; O’Connor et al., 2001; Bird et al., 2004, 2005; hare, and various closely related primate species (e.g., the golden O’Connor, 2007; Wurster et al., 2010). The presence of H. sapiens in langur, the hoolock gibbon and the stump-tailed macaque) (Corbet Laos by MIS 3 is secure (Demeter et al., 2012), and although dating and Hill, 1992; Vidya et al., 2005; Veron et al., 2007). The Brah- is a problem, populations may be present in MIS 4. More open maputra also corresponds to four bird sister-species pair bounda- environments may therefore have enabled rapid colonization of the ries (Ripley and Beehler, 1990). The Brahmaputra itself, however, is region down to the end of the Sunda plate. Effective dispersal probably less as a physical barrier to gene flow than a marker beyond this point may have had to await the evolution of an between two significantly distinct ecotones that are the true bar- appropriate maritime adaptation. On the other hand, though entry riers to genetic exchange (Tosi, 2007). Other species differences into Sahul is often taken to have occurred no earlier than 45 ka correlate to a region east of the Brahmaputra, in present-day (O’Connell and Allen, 2004), occupations on the Huon peninsula Myanmar, suggesting that biogeographic barriers in this region and in northern Australia (e.g., Malakunanja, Nauwalabila) may be are multiple. Examples of species separated by a Burmese biogeo- as old as 60e50 ka (Roberts et al., 1994), and in the Philippines as graphic barrier include the mongoose species Herpestes javanicus early as 67 ka (Mijares et al., 2010). and Herpestes auropunctatus, and three bird sister-species pairs Patterns of archaeological, palaeoenvironmental and other data (Ripley and Beehler, 1990). Other examples are the larger mammals for China are complex and are beyond the scope of this article. We that are restricted to Southeast Asia, such as orangutans, gibbons, would only note that the development of the Upper Palaeolithic is sun bears and tapirs (and also the pandas of South China); none of young (35e30 ka) (Qu et al., 2012) and the earliest unambiguous these ever entered South Asia, just as horses, camels and giraffids presence of H. sapiens in China appears to be no older than c. 40 ka never entered Southeast Asia from South Asia. This pattern of based on existing fossil and archaeological evidence (Shang et al., barriers, along with a range of important human genetic (Metspalu 2010; Norton and Jin, 2009), though there are potential hints of an et al., 2004), cultural (e.g., milk drinking, Simoons, 1970), techno- earlier presence of H. sapiens fossils at sites such as Huanglong Cave logical (e.g., the Movius Line, Movius, 1948) and domestic crop/ (Liu et al., 2010a), Liujiang (Shen et al., 2002; Wu, 2004) and Zhir- animal traits (Londo et al., 2006; Fuller, 2007; Larson et al., 2010) endong (Liu et al., 2010b). Much of China is dominated by broadleaf that are or have also been differentiated across the Brahmaputra/ woodlands, rich in tree nuts and some tubers, many of which re- Burma boundary, suggests that it has long constituted one of the sources extend into the Himalayan foothills. While acorns and other most significant terrestrial biogeographic barriers of the Old World. nuts are potentially rich resources, they would have necessitated the Accordingly, while it is not beyond the realm of possibility that development of cultural techniques of processing, which could have East and Southeast Asia will prove to have been colonized prior to contributed to a delay in the colonization of East Asia. Genetically, a MIS 4, it would not be surprising to find that modern human col- degree of Asia-based interbreeding between newly arriving modern onization east of the Brahmaputra/Burma region is delayed, per- humans and archaic species is, as we have seen, suggested by data haps even until vegetation patterns change significantly in MIS 4. from certain genes in the nuclear genome of contemporary humans While H. sapiens obviously displays significant behavioural flexi- (Harding et al., 1997; Cox et al., 2008; Wall et al., 2009). bility and complexity, novel, more densely forested tropical land- scapes of the type encountered in Burma would likely have 7. Discussion and conclusion required significant new behavioural adaptations and toolkits. Factors such as these may have discouraged expansion eastwards The question of whether the earliest H. sapiens fossils found in unless population pressure arose or environments changed. the Levant reflect a failed dispersal of modern humans, or just the Archaeological evidence to date seems to support this hypothesis. first step in a journey that brought them as far as India by the end of Technological differences between India and Indochina/Southeast MIS 5, remains an open one. While mtDNA coalescence ages of 70e Asia are significant throughout the Pleistocene. The colonization of 50 ka are typically cited for the timing of modern human Australia does not appear, on current evidence, to date to earlier movement out of Africa, the evidence in support of such a model, than 60 ka. Perhaps also significant is the fact that an increasing along with such linked elements as the notion of a single exit and number of archaic species of Homo appear to have survived until rapid coastal dispersal, is actually minimal. Patterns in the fossil, N. Boivin et al. / Quaternary International 300 (2013) 32e47 43 archaeological and genetic data do not conclusively support a c. Adams, J.M., Faure, H., 1997. Preliminary vegetation maps of the world since the Last 60 ka date for Out of Africa, and other readings of the evidence, Glacial Maximum: an aid to archaeological understanding. Journal of Archaeological Science 24, 623e647. as well as other models, are possible. Here we have presented a Armitage, S.J., Jasim, S.A., Marks, A.E., Parker, A.G., Usik, V.I., Uerpmann, H.-P., 2011. new analysis of the available evidence, together with a novel The southern route “out of Africa”: evidence for an early expansion of modern e attempt to model Pleistocene environments on the basis of humans into Arabia. Science 331 (6016), 453 456. Asouti, E., Fuller, D., 2008. Trees and Woodlands of South India: Archaeological botanical vegetation reconstruction techniques. These Perspectives. Left Coast Press, Walnut Creek, California. demonstrate that an MIS 5 exit is equally if not more plausible Athreya, S., 2007. Was Homo heidelbergensis in South Asia? A test using the than the MIS 3 or 4 exit that is currently widely advocated for Narmada fossil from central India. In: Petraglia, M., Allchin, B. (Eds.), The Evo- lution and History of Human Populations in South Asia. Springer, Dordrecht, Out of Africa (see also Hetherington and Reid, 2010). In addition, pp. 137e170. our analysis suggests that more complex scenarios for the dispersal Atkinson, Q., Gray, R., Drummond, A., 2008. mtDNA Variation predicts population of modern humans should be favoured. These potentially involved size in humans and reveals a major Southern Asian chapter in human pre- history. Molecular Biology and Evolution 25 (2), 468e474. a structured African population of H. sapiens, multiple exits, mul- Awasthi, A., 1995. Plant Geography and Flora of Rajasthan. Deep and Deep Pub- tiple routes of dispersal, and a degree of interbreeding with archaic lications, New Delhi. species of Homo as anatomically modern humans moved east- Bailey, G., 2009. The Red Sea, coastal landscapes, and hominin dispersals. In: wards. Given the current ambiguity of much of the evidence for Petraglia, M., Rose, J. (Eds.), The Evolution of Human Populations in Arabia. Springer, Dordrecht, pp. 15e37. modern human dispersals out of Africa, it is critical that researchers Bailey, R.C., Headland, T.N., 1991. The tropical rain forest: Is it a productive envi- remain open to alternative readings of the available data. We have ronment for human foragers? Human Ecology 19, 261e285. focused here on the possibility that the modern human exit Barker, G., Barton, H., Bird, M.I., Daly, P., Datan, I., Dykes, A., Farr, L., Gilbertson, D., Harrisson, B., Hunt, C., Higham, T., Kealhofer, L., Krigbaum, J., Lewis, H., recorded by fossil evidence in the Levant in MIS 5 does not repre- McLaren, S., Paz, V., Pike, A., Piper, P., Pyatt, B., Rabett, R., Reynolds, T., Rose, J., sent a failed dispersal, and that in fact our species was not only in Rushworth, G., Stephens, M., Stringer, C., Thompson, J., Turney, C., 2007. The ‘ ’ the Levant but also the Arabian Peninsula during this marine iso- human revolution in lowland tropical Southeast Asia: the antiquity and behavior of anatomically modern humans at Niah Cave (Sarawak, Borneo). tope stage, and spread to India before the Toba eruption at 74 ka Journal of Human Evolution 52, 243e261. (Petraglia et al., 2007). Another valid hypothesis we do not explore Barreiro, L., Patin, E., Neyrolles, O., Cann, H., Gicquel, B., Quintana-Murci, L., 2005. here is that H. sapiens was able to leave Africa in MIS 6 via a The heritage of pathogen pressures and ancient demography in the human innate-immunity CD209/CD209L region. American Journal of Human Genetics grassland corridor (Frumkin et al., 2011; see also Scally and Durbin, 77, 869e886. 2012). Yet another is that our species dispersed out of Africa shortly Basell, L., 2008. Middle Stone Age (MSA) site distributions in eastern Africa and after its first appearance c. 195 ka, in MIS 7 (Dennell and Roebroeks, their relationship to Quaternary environmental change, refugia and the evo- lution of Homo sapiens. Quaternary Science Reviews 27, 2484e2498. 2005: 1102). One other possibility is that there were several, sep- Behar, D.M., Villems, R., Soodyall, H., Blue-Smith, J., Pereira, L., Metspalu, E., arate dispersals of our species out of Africa (Dennell and Petraglia, Scozzari, R., Makkan, H., Tzur, S., Comas, D., Bertranpetit, J., Quintana-Murci, L., 2012). At the same time, we acknowledge that major demographic Tyler-Smith, C., Wells, R.S., Rosset, S., The Genographic Consortium, 2008. The changes occurred in MIS 4 and MIS 3, perhaps explaining the rel- dawn of human matrilineal diversity. American Journal of Human Genetics 82, 1130e1140. atively young mtDNA coalescence age in living populations. The Bird, M.I., Hope, G., Taylor, D., 2004. Populating PEP II: the dispersal of humans and increasing evidence for complexity as well as the clear patterns of agriculture through Austral-Asia and Oceania. Quaternary International 118e e bias for all records, whether archaeological, fossil or genetic, sug- 119, 145 163. Bird, M.I., Taylor, D., Hunt, C., 2005. Palaeoenvironments of insular Southeast Asia gests the need for an open mind to multiple scenarios for Out of during the Last Glacial Period: a savanna corridor in Sundaland? Quaternary Africa, as well as for more rather than less complex models of Science Reviews 24, 2228e2242. H. sapiens dispersal across Eurasia. Bowler, J., Johnston, H., Olley, J., Prescott, J.R., Roberts, R.G., Shawcross, W., Spooner, N., 2003. New ages for human occupation and climatic change at Lake Mungo, Australia. Nature 421, 837e840. Brown, T., Jones, M., Powell, W., Allaby, R., 2009. The complex origins of domes- Acknowledgements ticated crops in the Fertile Crescent. Trends in Ecology and Evolution 24, 103e 109. The field and analytical studies from which this article has Bruggemann, J.H., Buffl er, R.T., Guillaume, M.M.M., Walter, R.C., von Cosel, R., Ghebretensae, B.N., Berhe, S.M., 2004. Stratigraphy, palaeoenvironments and emerged have been funded by the British Academy, the European model for the deposition of the Abdur Reef Limestone: context for an important Research Council (no. 295719), the Leakey Foundation, the Lev- archaeological site from the Last Interglacial on the Red Sea coast of Eritrea. erhulme Trust, the National Geographic Society, the Natural Envi- Palaeogeography, Palaeoclimatology, Palaeoecology 20, 179e206. ronment Research Council, and the Society for South Asian Studies. Brumm, A., Moore, M.W., 2005. Symbolic revolutions and the Australian archaeo- logical record. Cambridge Archaeological Journal 15 (2), 157e175. We thank three anonymous reviewers for helping us to make Bulbeck, D., 2007. Where river meets sea: a parsimonious model for Homo improvements on this article and we are grateful to Michael Has- sapiens colonization of the Indian Ocean rim and Sahul. Current Anthropology e lam for assisting with input of some references. We thank James 48, 315 321. Cabrera, V.M., Abu-Amero, K.K., Larruga, J.M., Gonzalez, A.M., 2009. The Arabian Blinkhorn and Huw Groucutt for inviting us to the Oxford confer- Peninsula: gate for human migrations out of Africa or cul-de-sac? A mito- ence and for their close reading of the text. chondrial DNA phylogeographic perspective. In: Petraglia, M., Rose, J. (Eds.), The Evolution of Human Populations in Arabia. Springer, Dordrecht, pp. 79e87. Clarkson, C., Harris, C., Jones, S., 2012. Continuity and change in the lithic industries Appendix A. Supplementary materials of the Jurreru Valley, India, before and after the Toba eruption. Quaternary International 258, 165e179. Corbet, G.B., Hill, J.E., 1992. The Mammals of the Indomalayan Region: a Systematic Supplementary material associated with this article can be Review. Oxford University Press, Oxford. found in the online version, at http://dx.doi.org/10.1016/j.quaint. Cortés-Sánchez, M., Morales-Muñiz, A., Simón-Vallejo, M.D., Lozano-Francisco, M.C., 2013.01.008. Vera-Peláez, J.L., Finlayson, C., Rodríguez-Vidal, J., Delgado-Huertas, A., Jiménez- Espejo, F.J., Martínez-Ruiz, F., Aranzazu Martínez-Aguirre, M., Pascual- Granged, A.J., Mercè Bergadà-Zapata, M., Gibaja-Bao, J.F., Riquelme-Cantal, J.A., López-Sáez, J.A., Rodrigo-Gámiz, M., Sakai, S., Sugisaki, S., Finlayson, G., Fa, D.A., References Bicho, N.F., 2011. Earliest known use of marine resources by Neanderthals. PLoS One, e24026. Abeyratne, M., 1996. Multidating Studies of Archaeological Sites. Australian National Cox, M., Mendez, F., Karafet, T., Pilkington, M.M., Kingan, S., Destro-Bisol, G., University, Canberra. Strassmann, B., Hammer, M., 2008. Testing for archaic hominin admixture on Achyuthan, H., Quade, J., Roe, L., Placzek, C., 2007. Stable isotopic composition of the X chromosome: model likelihoods for the modern human RRM2P4 region pedogenic carbonates from the eastern margin of the Thar Desert, Rajasthan, from summaries of genealogical topology under the structured coalescent. India. Quaternary International 162e163, 50e60. Genetics 178, 427e437. 44 N. Boivin et al. / Quaternary International 300 (2013) 32e47

Curnoe, D., Ji, X., Herries, A.I.R., Bai, K., Tacon, P.S.C., Zhende, B., Fink, D., Green, R.E., Krause, J., Briggs, A.W., Maricic, T., Stenzel, U., Kircher, M., Patterson, N., Yunsheng, Z., Hellstrom, J., Yun, L., Cassis, G., Bing, S., Wroe, S., Shi, H., Li, H., Zhai, W., Hsi-Yang Fritz, M., Hansen, N., Durand, E., Malaspinas, A.-S., Parr, W.C.H., Shengmin, H., Rogers, N., 2012. Human remains from the Jensen, J., Marques-Bonet, T., Alkan, C., Prufer, K., Meyer, M., Burbano, H., PleistoceneeHolocene transition of Southwest China suggest a complex evo- Good, J., Schultz, R., Aximu-Petri, A., Butthof, A., Hober, B., Hoffner, B., lutionary history for East Asians. PLoS One, e31918. Siegemund, M., Weihmann, A., Nusbaum, C., Lander, E., Russ, C., Novod, N., d’Errico, F., Vanhaeren, M., Barton, N., Bouzouggar, A., Mienis, H., Richter, R., Hublin, J.-J., Affourtit, J., Egholm, M., Verna, C., Rudan, P., Brajkovic, D., Kucan, Z., Gusic, I., McPherron, S.P., Lozouet, P., 2009. Additional evidence on the use of personal Doronichev, V., Golovanova, L., Lalueza-Fox, C., de la Rasilla, M., Fortea, J., ornaments in the Middle Paleolithic of North Africa. Proceedings of the National Rosas, A., Schmitz, R., Johnson, P., Eichler, E., Falush, D., Birney, E., Mulliken, J., Academy of Sciences of the United States of America 106, 6051e16056. Slatkin, M., Nielsen, R., Kelso, J., Lachmann, M., Reich, D., Paabo, S., 2010. A draft Deacon, H.J., 1989. Late Pleistocene palaeoecology and archaeology in the southern sequence of the Neandertal genome. Science 328, 710e722. Cape, South Africa. In: Mellars, P., Stringer, C. (Eds.), The Human Revolution: Groucutt, H.S., Petraglia, M.D., 2012. The prehistory of the Arabian Peninsula: Behavioural and Biological Perspectives on the Origins of Modern Humans. deserts, dispersals, and demography. Evolutionary Anthropology 21, 113e125. Edinburgh University Press, Edinburgh, pp. 547e564. Grun, R., Stringer, C., McDermott, F., Nathan, R., Porat, N., Robertson, S., Taylor, L., Delagnes, A., Tribolo, C., Bertran, P., Brenet, M., Crassard, R., Jaubert, J., Khalidi, L., Mortimer, G., Eggins, S., McCoulloch, M., 2005. U-series and ESR analyses of Mercier, N., Nomade, S., Peigné, S., Sitzia, L., Tournepiche, J.F., Al-Halibi, M., Al- bones and teeth relating to the human burials from Skhul. Journal of Human Mosabi, A., Macchiarelli, R., 2012. Inland human settlement in southern Arabia Evolution 49, 316e334. 55,000 years ago: new evidence from the Wadi Surdud Middle Paleolithic site Gunz, P., Bookstein, F., Mitteroecker, P., Stadlmayr, A., Seidler, H., Weber, G., 2009. complex, western Yemen. Journal of Human Evolution 63, 452e474. Early modern human diversity suggests subdivided population structure and a Demeter, F., Shackleford, L.L., Bacon, A.-M., Duringer, P., Westaway, K., complex out-of-Africa scenario. Proceedings of the National Academy of Sci- Sayavongkhamdy, T., Braga, J., Sichanthongtip, P., Khamdalavong, P., Ponche, J.- ences of the United States of America 106, 6094e6098. L., Wang, H., Lundstrom, C., Patole-Edoumba, E., Karpoff, A.-M., 2012. Ana- Haak, W., Forster, P., Bramanti, B., Matsumura, S., Brandt, G., Tanzer, M., Villems, R., tomically modern human in Southeast Asia (Laos) by 46 ka. Proceedings of the Renfrew, C., Gronenborn, D., Alt, K.W., Burger, J., 2005. Ancient DNA from the National Academy of Sciences of the United States of America 109, 14375e first European farmers in 7500-year-old Neolithic sites. Science 310, 1016e1018. 14380. Habgood, P., Franklin, N., 2008. The revolution that didn’t arrive: a review of Dennell, R.W., 2009. The Palaeolithic Settlement of Asia. Cambridge University Pleistocene Sahul. Journal of Human Evolution 55, 187e222. Press, Cambridge. Handt, O., Meyer, S., von Haeseler, A., 1998. Compilation of human mtDNA control Dennell, R., Petraglia, M., 2012. The dispersal of Homo sapiens across southern Asia: region sequences. Nucleic Acids Research 26, 126e129. how early, how often, how complex? Quaternary Science Reviews 47, 15e22. Harding, R., McVean, G., 2004. A structured ancestral population for the evolution of Dennell, R., Roebroeks, W., 2005. An Asian perspective on early human dispersal modern humans. Current Opinion in Genetics & Development 14, 1e8. from Africa. Nature 438, 1099e1104. Harding, R., Fullerton, S., Griffiths, R., Bond, J., Cox, M., Schneider, J., Moulin, D., Deraniyagala, S.U., 1992. The Prehistory of Sri Lanka. Archaeological Survey Clegg, J., 1997. Archaic African and Asian lineages in the genetic ancestry of Department, Colombo. modern humans. American Journal of Human Genetics 60, 772e789. Drake, N.A., Blench, R.M., Armitage, S.J., Bris- tow, C.S., White, K.H., 2011. Ancient Hardy, J., Pittman, A., Myers, A., Gwinn-Hardy, K., Fung, H., de Silva, R., Hutton, M., watercourses and biogeography of the Sahara explain the peopling of the Duckworth, J., 2005. Evidence suggesting that Homo neanderthalensis con- desert. Proceedings of the National Academy of Sciences of the United States of tributed the H2 MAPT haplotype to Homo sapiens. Biochemical Society Trans- America 108, 458e462. actions 33, 582e585. Edgell, H., 2006. Arabian Deserts: Nature, Origins and Evolution. Springer, Harpending, H.C., Eswaran, V., 2005. Tracing modern human origins. Science 309, Dordrecht. 1995. Endicott, P., Ho, S.Y.W., Metspalu, M., Stringer, C., 2009. Evaluating the mitochon- Harris, E., Hey, J., 1999. X chromosome evidence for ancient human histories. Pro- drial timescale of human evolution. Trends in Ecology and Evolution 24 (9), ceedings of the National Academy of Sciences of the United States of America 515e521. 96, 3320e3324. Eswaran, V., Harpending, H.C., Rogers, A.R., 2005. Genomics refutes an exclusively Haslam, M., Roberts, R.G., Shipton, C., Pal, J.N., Fenwick, J., Ditchfield, P., Boivin, N., African origin of humans. Journal of Human Evolution 49, 1e18. Dubey, A.K., Gupta, M.C., Petraglia, M., 2011. Late Acheulean hominins at the Evans, P., Mekel-Bobrov, N., Vallender, E., Hudson, R., Lahn, B., 2006. Evidence that Marine Isotope Stage 6/5e transition in north-central India. Quaternary the adaptive allele of the brain size gene microcephalin introgressed into Homo Research 75, 670e682. sapiens from an archaic Homo lineage. Proceedings of the National Academy of Hetherington, R., Reid, R.G.B., 2010. The Climate Connection: Climate Change and Sciences of the United States of America 103, 18178e18183. Modern Human Evolution. Cambridge University Press, Cambridge. Faure, H., Walter, R., Grant, D., 2002. The coastal oasis: ice age springs on emerged Hill, C., 2009. Stratigraphy and sedimentology at Bir Sahara, Egypt: environments, continental shelves. Global and Planetary Change 33, 47e56. climate change and the Middle Paleolithic. Catena 78, 250e259. Fernandez, C., 2009. Bayesian coalescent inference from mitochondrial DNA varia- Ho, S.Y.W., Larson, G., 2006. Molecular clocks: when times are a-changin’. Trends in tion of the colonization time of Arabia by the hamadryas Baboon (Papio ham- Genetics 22, 79e83. adryas hamadryas). In: Petraglia, Rose, J. (Eds.), The Evolution of Human Howell, F.C., 1999. Paleo-demes, species clades, and extinctions in the Pleistocene Populations in Arabia. Springer, Dordrecht, pp. 89e100. hominin record. Journal of Archaeological Research 55, 191e243. Field, J.S., Lahr, M.M., 2005. Assessment of the southern dispersal: GIS-based Huggett, R.J., 1995. Geoecology: an Evolutionary Approach. Routledge, London. analyses of potential routes at Oxygen Isotopic Stage 4. Journal of World Pre- Indriati, E., Swisher, C.C., Lepre, C., Quinn, R.L., Suriyanto, R.A., Hascaryo, A.T., history 19, 1e45. Grün, R., Feibel, C.S., Pobiner, B.L., Aubert, M., Lees, W., Antón, S.C., 2011. The age Field, J.S., Petraglia, M., Lahr, M.M., 2007. The southern dispersal hypothesis and the of the 20 meter Solo River Terrace, Java, Indonesia and the survival of Homo South Asian archaeological record: examination of dispersal routes through GIS erectus in Asia. PLoS One, e21562. analysis. Journal of Anthropological Archaeology 26, 88e108. James, H., Petraglia, M., 2005. Modern human origins and the evolution of behavior Fleitmann, D., Matter, A., 2009. The speleothem record of climate variability in in the later Pleistocene record of South Asia. Current Anthropology 46, S3eS27. Southern Arabia. Comptes Rendus Geosciences 341, 633e642. Jobling, M., Tyler-Smith, C., 2003. The human Y chromosome: an evolutionary Foley, R., Lahr, M.M., 1997. Mode 3 technologies and the evolution of modern marker comes of age. Nature Reviews Genetics 4, 598e612. humans. Cambridge Archaeological Journal 7, 3e36. Keinan, A., Mullikin, J., Patterson, N., Reich, D., 2007. Measurement of the human Forster, P., Matsumura, S., 2005. Did early humans go north or south? Science 308, allele frequency spectrum demonstrates greater genetic drift in East Asians 965e966. than in Europeans. Nature Genetics 39, 1251e1255. Forster, P., 2004. Ice Ages and the mitochondrial DNA chronology of human dis- Kennedy, K.A.R., 2000. God-Apes and Fossil Men. University of Michigan Press, Ann persals: a review. Philosophical Transactions of the Royal Society of London Arbor. Series B 359, 255e264. Kingdon, J., 1993. Self-made Man and His Undoing. Simon & Schuster, London. Frumkin, A., Bar-Yosef, O., Schwarcz, H., 2011. Possible paleohydrologic and paleo- Kivisild, T., Shen, P., Wall, D., Do, B., Sung, R., Davis, K., Passarino, G., Underhill, P., climatic effects on hominin migration and occupation of the Levantine Middle Scharfe, C., Torroni, A., Scozzari, R., Modiano, D., Coppa, A., de Knijff, P., Paleolithic. Journal of Human Evolution 60, 437e451. Feldman, M., Cavalli-Sforza, L., Oefner, P., 2006. The role of selection in the Fuller, D., 2007. Non-human genetics, agricultural origins and historical linguistics evolution of human mitochondrial genomes. Genetics 172, 373e387. in South Asia. In: Petraglia, M., Allchin, B. (Eds.), The Evolution and History of Korisettar, R., 2007. Toward developing a basin model for Paleolithic settlement of Human Populations in South Asia. Springer, Dordrecht. the Indian subcontinent: geodynamics, monsoon dynamics, habitat diversity Garrigan, D., Hammer, M., 2006. Reconstructing human origins in the genomic era. and dispersal routes. In: Petraglia, M., Rose, J. (Eds.), The Evolution of Human Nature Reviews Genetics 7, 669e680. Populations in Arabia. Springer, Dordrecht, pp. 69e96. Garrigan, D., Mobasher, Z., Severson, T., Wilder, J., Hammer, M., 2005. Evidence for Krause, J., Fu, Q., Good, J., Viola, B., Shunkov, M., Derevianko, A., Paabo, S., 2010. The archaic Asian ancestry on the human X chromosome. Molecular Biology and complete mitochondrial DNA genome of an unknown hominin from southern Evolution 22, 189e192. Siberia. Nature 464, 894e897. Green, R.E., Malaspinas, A.-S., Krause, J., Briggs, A.W., Johnson, P., Uhler, C., Krings, M., Stone, A., Schmitz, R., Krainitzki, H., Stoneking, M., Paabo, S., 1997. Meyer, M., Good, J., Maricic, T., Stenzel, U., Prufer, K., Siebauer, M., Burbano, H., Neandertal DNA sequences and the origin of modern humans. Cell 90, 19e30. Ronan, M., Rothberg, J., Egholm, M., Rudan, P., Brajkovic, D., Kucan, Z., Gusic, I., Labuda, D., Zietkiewicz, E., Yotova, V., 2000. Archaic lineages in the history of Wikstrom, M., Laakkonen, L., Kelso, J., Slatkin, M., Paabo, S., 2008. A complete modern humans. Genetics 156, 799e808. Neandertal mitochondrial genome sequence determined by high-throughput Lahr, M.M., 1996. The Evolution of Modern Human Diversity. Cambridge University sequencing. Cell 134, 416e426. Press, Cambridge. N. Boivin et al. / Quaternary International 300 (2013) 32e47 45

Lahr, M.M., Foley, R., 1994. Multiple dispersals and modern human origins. Evolu- Movius, H.L., 1948. The Lower Palaeolithic cultures of southern and eastern Asia. tionary Anthropology 3, 48e60. Transactions of the American Philosophical Society 38 (4), 329e420. Lahr, M.M., Foley, R., 1998. Towards a theory of modern human origins: geography, Nasab, H.V., Clark, G.A., Torkamandi, S., 2013. Late Pleistocene dispersal corridors demography, and diversity in recent human evolution. Yearbook of Physical across the Iranian Plateau: a case study from Mirak, a Middle Paleolithic site on Anthropology 41, 137e176. the northern edge of the Iranian Central desert (Dasht-e Kavir). Quaternary Lambeck, K., Purcell, A., Flemming, N.C., Vita-Finzi, C., Alsharekh, A.M., Bailey, G.N., International 300, 267e281. 2012. Sea level and shoreline reconstructions for the Red Sea: isostatic and Norton, C., Jin, J., 2009. The evolution of modern human behavior in east Asia: tectonic considerations and implications for hominin migration out of Africa. current perspectives. Evolutionary Anthropology 18, 247e260. Quaternary Science Reviews 30, 3542e3574. O’Connell, J.F., Allen, J., 2004. Dating the colonization of Sahul (Pleistocene Larson, G., Albarella, U., Dobney, K., Rowley-Conwy, P., Schibler, J., Tresset, A., Australia-New Guinea): a review of recent research. Journal of Archaeological Vigne, J.-D., Edwards, C., Schlumbaum, A., Dinu, A., Balacsescu, A., Dolman, G., Science 31, 835e853. Tagliacozzo, A., Manaseryan, N., Miracle, P., Van Wijngaarden-Bakker, L., O’Connell, J.F., Allen, J., 2007. Pre-LGM Sahul (Pleistocene AustraliaeNew Guinea) Masseti, M., Bradley, D., Cooper, A., 2007. Ancient DNA, pig domestication, and and the archaeology of early modern humans. In: Mellars, P., Boyle, K., Bar- the spread of the Neolithic into Europe. Proceedings of the National Academy of Yosef, O., Stringer, C. (Eds.), Rethinking the Human Revolution. McDonald Sciences of the United States of America 104, 15276e15281. Institute for Archaeological Research, Cambridge, pp. 395e410. Larson, G., Liu, R., Zhao, X., Yuan, J., Fuller, D., Barton, L., Dobney, K., Fan, Q., Gu, Z., O’Connor, S., 2007. New evidence from East Timor contributes to our understanding Liu, X.-H., Luo, Y., Lv, P., Andersson, L., Li, N., 2010. Patterns of East Asian pig of earliest modern human colonisation east of the Sunda Shelf. Antiquity 81, domestication, migration, and turnover revealed by modern and ancient DNA. 523e535. Proceedings of the National Academy of Sciences of the United States of O’Connor, S., Veth, P., 2000. The world’s first mariners: savanna dwellers in an America 107, 7686e7691. island continent. In: O’Connor, S., Veth, P. (Eds.), East of Wallace’s Line: Studies Legris, P., 1963. Le vegetation de L’Inde: Ecologie et Flore. In: Travaux de la Section of Past and Present Maritime Societies in the Indo-pacific Region. A.A. Bal- Scientifique et Technique, vol. 6. Institut Français de Pondicherry (Pondicherry). kerma, Rotterdam, pp. 99e137. Li, J., Absher, D., Tang, H., Southwick, A., Casto, A., Ramachandran, S., Cann, H., O’Connor, S., Aplin, K., Spriggs, M., Veth, P., Ayliffe, L., 2001. From savannah to Barsh, G., Feldman, M., Cavalli-Sforza, L., Myers, R., 2008. Worldwide human rainforest: changing environments and human occupation at Liang Lemdubu, relationships inferred from genome-wide patterns of variation. Science 319, Aru Islands, Maluku (Indonesia). In: Kershaw, P., David, B., Napper, T., Penny, D., 100e104. Brown, J. (Eds.), Bridging Wallace’s Line. Catena Verlag, Reiskirchen, pp. 279e Liu, W., Wu, X., Pei, S., Wu, X., Norton, C., 2010a. Huanglong Cave: a Late Pleistocene 306. human fossil site in Hubei Province, China. Quaternary International 211, 29e41. Oppenheimer, S., 2003. Out of Eden: the Peopling of the World. Constable, London. Liu, W., Jin, C., Zhang, Y., Cai, Y., Xing, S., Wu, J., Cheng, H., Edwards, R.L., Pan, W., Oppenheimer, S., 2009. The great arc of dispersal of modern humans: Africa to Qin, D., An, Z., Trinkhaus, E., Wu, X., 2010b. Human remains from Zhirendong, Australia. Quaternary International 202, 2e13. South China, and modern human emergence in East Asia. Proceedings of the Oppenheimer, S., 2012. A single southern exit of modern humans from Africa: National Academy of Sciences of the United States of America 107, 19201e before or after Toba? Quaternary International 258, 88e99. 19206. Osborne, A., Vance, D., Rohling, E., Barton, N., Rogerson, M., Fello, N., 2008. A humid Londo, J., Chiang, Y.-C., Hung, K.-H., Chiang, T.-Y., Schaal, B., 2006. Phylogeography of corridor across the Sahara for the migration of early modern humans out of Asian wild rice, Oryza rufipogon, reveals multiple independent domestications Africa 120,000 years ago. Proceedings of the National Academy of Sciences of of cultivated rice, Oryza sativa. Proceedings of the National Academy of Sciences the United States of America 105, 16444e16447. of the United States of America 103, 9578e9583. Ozkan, H., Willcox, G., Graner, A., Salamini, F., Kilian, B., 2010. Geographic distribution Macaulay, V., Hill, C., Achilli, A., Rengo, C., Clarke, D., Scozzari, R., Cruciani, F., and domestication of wild emmer wheat (Triticum dicoccoides). Genetic Taha, A., Shaari, N.K., Raja, J.M., Ismail, P., Zainuddin, Z., Goodwin, W., Resources and Crop Evolution. http://dx.doi.org/10.1007/s10722-010-9581-5. Bulbeck, D., Bandelt, H.-J., Oppenheimer, S., Torroni, A., Richards, M., 2005. Parker, A., 2009. Pleistocene Climate Change in Arabia: Developing a Frame- Single, rapid coastal settlement of Asia revealed by analysis of complete work for Hominin Dispersal over the Last 350 ka. In: Petraglia, M., Rose, J. mitochondrial genomes. Science 308, 1034e1036. (Eds.), The Evolution of Human Populations in Arabia. Springer, Dordrecht, Malaivijitnond, S., Lekprayoon, C., Tandavanittj, N., Panha, S., Cheewatham, C., pp. 39e49. Hamada, Y., 2007. Stone-tool usage by Thai long-tailed macaques (Macaca fas- Patnaik, R., Chauhan, P., Rao, M.R., Blackwell, B., Skinner, A., Sahni, A., Chauhan, M.S., cicularis). American Journal of Primatology 69, 227e233. Khan, H., 2009. New geochronological, paleoclimatological, and archaeological Marean, C.W., 2010. When the sea saved humanity. Scientific American 303, 54e61. data from the Narmada Valley hominin locality, central India. Journal of Human Martin, D.L., 2007. Bioarchaeology in the United Arab Emirates. Arabian Archae- Evolution 56, 114e133. ology and Epigraphy 18, 124e131. Perera, N., Kourampas, N., Simpson, I.A., Deraniyagala, S.U., Bulbeck, D., McBrearty, S., Brooks, A., 2000. The revolution that wasn’t: a new interpretation of Kamminga, J., Perera, J., Fuller, D.Q., Szabo, K., Oliviera, N.V., 2011. People of the the origin of modern human behavior. Journal of Human Evolution 39, 453e ancient rainforest: Late Pleistocene foragers at the Batadomba-lena rockshelter, 563. Sri Lanka. Journal of Human Evolution 61, 254e269. McDougall, I., Brown, F., Fleagle, J., 2005. Stratigraphic placement and age of Petit-Maire, N., Carbonel, P., Reyss, J.-L., Salanville, P., Abed, A., Bourrouilh, R., modern humans from Kibish, Ethiopia. Nature 433, 733e736. Fontugne, M., Yasin, S., 2010. A vast Eemian palaeolake in Southern Jordan Meher-Homji, V.M., 2001. Bioclimatology and Plant Geography of India. Scientific (29N). Global and Planetary Change 72, 368e373. Publishers, Jodhpur. Petraglia, M.D., 2007. Mind the gap: factoring the Arabian Peninsula and the Indian Mellars, P., 2006a. Going East: new genetic and archaeological perspectives on the subcontinent into Out of Africa models. In: Mellars, P., Bar-Yosef, O., Boyle, K., modern human colonization of Eurasia. Science 313, 796e800. Stringer, C. (Eds.), Rethinking the Human Revolution. McDonald Institute for Mellars, P., 2006b. Why did modern human populations disperse from Africa ca. Archaeological Research, Cambridge, pp. 383e394. 60,000 years ago? A new model. Proceedings of the National Academy of Sci- Petraglia, M.D., 2010. The Early Paleolithic of the Indian Subcontinent: hominin ences of the United States of America 103, 9381e9386. colonization, dispersals and occupation history. In: Fleagle, J., Shea, J., Grine, F., Merriwether, D.A., Hodgson, J., Friedlaender, F., Allaby, R., Cerchio, S., Koki, G., Baden, A., Leakey, R. (Eds.), Out of Africa I: The First Hominin Colinization of Friedlaender, J., 2005. Ancient mitochondrial M haplogroups identified in the Eurasia. Springer Press, Dordrecht, pp. 165e179. Southwest Pacific. Proceedings of the National Academy of Sciences of the Petraglia, M.D., 2011. Archaeology: Trailblazers across Arabia. Nature 470, 50e51. United States of America 102, 13034e13039. Petraglia, M.D., Alsharekh, A., 2003. The Middle Palaeolithic of Arabia: implica- Metspalu, M., Kivisild, T., Metspalu, E., Parik, J., Hudjashov, G., Kaldma, K., Serk, P., tions for modern human origins, behaviour and dispersals. Antiquity 77, Karmin, M., Behar, D.M., Gilbert, M.T.P., Endicott, P., Mastana, S., Papiha, S.S., 671e684. Skorecki, K., Torroni, A., Villems, R., 2004. Most of the extant mtDNA boundaries Petraglia, M., Korisettar, R., Boivin, N., Clarkson, C., Ditchfield, P., Jones, S., Koshy, J., in South and Southwest Asia were likely shaped during the initial settlement of Lahr, M.M., Oppenheimer, C., Pyle, D., Roberts, R., Schwenninger, J.-L., Arnold, L., Eurasia by anatomically modern humans. BMC Genetics 5, 26. White, K., 2007. Middle Palaeolithic assemblages from the Indian subcontinent Meyer, M., Kircher, M., Gansauge, M.-T., Li, H., Racimo, F., Mallick, S., Schraiber, J.G., before and after the Toba super-eruption. Science 317, 114e116. Jay, F., Prüfer, K., de Filippo, C., Sudmant, P.H., Alkan, C., Fu, Q., Do, R., Petraglia, M., Clarkson, C., Boivin, N., Haslam, M., Korisettar, R., Chaubey, G., Rohland, N., Tandon, A., Siebauer, M., Green, R.E., Bryc, K., Briggs, A.W., Ditchfield, P., Fuller, D., James, H., Jones, S., Kivisild, T., Koshy, J., Lahr, M.M., Stenzel, U., Dabney, J., Shendure, J., Kitzman, J., Hammer, M.F., Shunkov, M.V., Metspalu, M., Roberts, R., Arnold, L., 2009. Population increase and environmental Derevianko, A.P., Patterson, N., Andrés, A.M., Eichler, E.E., Slatkin, M., Reich, D., deterioration correspond with microlithic innovations in South Asia ca. 35,000 Kelso, J., Pääbo, S., 2012. A high-coverage genome sequence from an archaic years ago. Proceedings of the National Academy of Sciences of the United States of Denisovan individual. Science 338, 222e226. America 106, 12261e12266. Mijares, A.S., Détriot, F., Piper, P., Grün, R., Bellwood, P., Aubert, M., Champion, G., Petraglia, M., Haslam, M., Fuller, D., Boivin, N., Clarkson, C., 2010. Out of Africa: new Cuevas, N., De Leon, A., Dizon, E., 2010. New evidence for a 67,000-year-old hypotheses and evidence for the dispersal of Homo sapiens along the Indian human presence at Callao Cave, Luzon, Philippines. Journal of Human Evolution Ocean rim. Annals of Human Biology 37 (3), 288e311. 59, 123e132. Petraglia, M.D., Alsharekh, A.M., Crassard, R., Drake, N., Groucutt, H., Parker, A., Misra, V.N., 2001. Prehistoric human colonization of India. Journal of Biosciences 26 Roberts, R., 2011. Middle Paleolithic occupation on a Marine Isotope Stage 5 (4), 491e531. lakeshore in the Nefud Desert, Saudi Arabia. Quaternary Science Reviews 30, Moore, M.W., Brumm, A., 2007. Stone artifacts and hominins in island Southeast 1555e1559. Asia: new insights from Flores, eastern Indonesia. Journal of Human Evolution Petraglia, M.D., Alsharekh, A., Breeze, P., Clarkson, C., Crassard, R., Drake, N.A., 52, 85e102. Groucutt, H.S., Jennings, R., Parker, A.G., Parton, A., Roberts, R.G., Shipton, C., 46 N. Boivin et al. / Quaternary International 300 (2013) 32e47

Matheson, C., al-Omari, A., Veall, M.-A., 2012a. Hominin dispersal into the Nefud Shang, H., Tong, H., Zhang, S., Chen, F., Trinkhaus, E., 2010. An early modern human Desert and Middle Palaeolithic settlement along the Jubbah Palaeolake, from Tianyuan Cave, Zhoukoudian, China. Proceedings of the National Academy Northern Arabia. PLoS One 7 (11), e49840. of Sciences of the United States of America 104, 6573e6578. Petraglia, M.D., Groucutt, H., Blinkhorn, J., 2012b. Hominin evolutionary history in Shea, J.J., 2003. The Middle Paleolithic of the east Mediterranean Levant. Journal of the Arabian Desert and the Thar Desert. In: Mol, L., Sternberg, T. (Eds.), World Prehistory 17 (4), 313e394. Changing Deserts: Integrating People and Their Environment. The White Horse Shea, J.J., 2008. Transitions or turnovers? Climatically-forced extinctions of Homo Press, Isle of Harris, pp. 61e82. sapiens and Neanderthals in the east Mediterranean Levant. Quaternary Science Petraglia, M.D., Ditchfield, P., Jones, S., Korisettar, R., Pal, J.N., 2012c. The Toba vol- Reviews 27, 2253e2270. canic super-eruption, environmental change, and hominin occupation history Shen, G., Wang, W., Wang, Q., Zhao, J.-X., Collerson, K.D., Zhou, C., Tobias, P., 2002. in India over the last 140,000 years. Quaternary International 258, 119e134. U-series dating of Liujiang hominid site in Guangxi, Southern China. Journal of Piper, P.J., Rabett, R.J., Bin Kurui, E., 2008. Using community, composition and Human Evolution 43, 817e829. structural variation in terminal Pleistocene vertebrate assemblages to identify Shennan, S., 2001. Demography and cultural innovation: a model and its implica- human hunting behaviour at the Niah Caves, Borneo. Bulletin of the Indo- tions for the emergence of modern human culture. Cambridge Archaeological pacific Prehistory Association 28, 88e98. Journal 11 (1), 5e16. Plagnol, V., Wall, J., 2006. Possible ancestral structure in human populations. PLoS Simoons, F., 1970. The traditional limits of milking and milk use in southern Asia. Genetics 2, e105. Anthropos 65, 547e593. Powell, A., Shennan, S., Thomas, M., 2009. Late Pleistocene demography and the Smith, J.R., Hawkins, A.L., Asmerom, Y., Polyak, V., Giegengack, R., 2007. New age appearance of modern human behavior. Science 324, 1298e1301. constraints on the Middle Stone Age occupations of Kharga Oasis, Western Prentice, I.C., Jolly, D., BIOME 6000 participants, 2000. Mid-Holocene and glacial- Desert. Egypt Journal of Human Evolution 52, 690e701. maximum vegetation geography of the northern continents and Africa. Jour- Singhvi, A.K., Williams, M.A.J., Rajaguru, S.N., Misra, V.N., Chawla, S., Stokes, S., nal of Biogeography 27, 507e519. Chauhan, N., Francis, T., Ganjoo, R.K., Humphreys, G.S., 2010. A w200 ka record Preusser, F., 2009. Chronology of the impact of Quaternary climate change on of climatic change and dune activity in the Thar Desert, India. Quaternary continental environments in the Arabian Peninsula. Comptes Rendus Geo- Science Reviews 29, 3095e3105. sciences 341, 621e632. Soares, P., Ermini, L., Thomson, N., Mormina, M., Rito, T., Rohl, A., Salas, A., Qu, T., Bar-Yosef, O., Wang, Y., Wu, X., 2012. The Chinese Upper Paleolithic: geog- Oppenheimer, S., Richards, M., 2009. Correcting for purifying selection: an raphy, chronology, and techno-typology. Journal of Archaeological Research. improved human mitochondrial molecular clock. American Journal of Human http://dx.doi.org/10.1007/s10814-012-9059-4. Genetics 84, 740e759. Rabinovich, R., Tchernov, E., 1995. Chronological, paleoecological and taphonomical Stewart, J.R., Stringer, C.B., 2012. Human evolution Out of Africa: the role of refugia aspects of the Middle Paleolithic site of Qafzeh, Israel. In: Buitenhuis, H., and climate change. Science 6074, 1317e1321. Uerpmann, H. (Eds.), Archaeozoology of the Near East. Backhuys Publishers, Stiner, M., Munro, N., Surovell, T., Tchernov, E., Bar-Yosef, O., 1999. Paleolithic Leiden, pp. 5e44. population growth pulses evidenced by small animal exploitation. Science 283, Rasmussen, M., Guo, X., Wang, Y., Lohmueller, K., Rasmussen, S., Albrechtsen, A., 190e194. Skotte, L., Lindgreen, S., Metspalu, M., Jombart, T., Kivisild, T., Zhai, W., Stiner, M., Munro, N., Surovell, T., 2000. The tortoise and the hare: small-game use, Eriksson, A., Manica, A., Orlando, L., De La Vega, F.M., Tridico, S., Metspalu, E., the broad-spectrum revolution, and Paleolithic demography. Current Anthro- Nielsen, K., Ávila-Arcos, M.C., Moreno-Mayar, J.V., Muller, C., Dortch, J., pology 41, 39e79. Gilbert, M., Lund, O., Wesolowska, A., Karmin, M., Weinert, L., Wang, B., Li, J., Stringer, C., 2000. Coasting out of Africa. Nature 405, 24e27. Tai, S., Xiao, F., Hanihara, T., van Driem, G., Jha, A., Ricaut, F., de Knijff, P., Summerhayes, G., Leavesley, M., Fairbairn, A., Mandui, H., Field, J., Fore, A., Migliano, A., Gallego Romero, I., Kristiansen, K., Lambert, D., Brunak, S., Fullagar, R., 2010. Human adaptation and plant use in Highland New Guinea Forster, P., Brinkmann, B., Nehlich, O., Bunce, M., Richards, M., Gupta, R., 49,000 to 44,000 years ago. Science 330 (6000), 78e81. Bustamante, C., Krogh, A., Foley, R., Lahr, M., Balloux, F., Sicheritz- Pontén, T., Tchernov, E., 1992. The Afro-Arabian component in the Levantine mammalian Villems, R., Nielsen, R., Wang, J., Willerslev, E., 2011. An Aboriginal Australian fauna: a short biogeographical review. Israel Journal of Zoology 38, 155e192. genome reveals separate human dispersals into Asia. Science 334, 94e98. Tchernov, E., 1996. Rodent faunas, chronostratigraphy and paleobiogeography of the Reed, D., Smith, V.S., Hammond, S., Rogers, A.R., Clayton, D., 2004. Genetic analysis southern Levant during the Quaternary. Acta Zoologica Cracoviensis 39, 513e of lice supports direct contact between modern and archaic humans. PLoS 530. Biology 2, e340. Thangaraj, K., Chaubey, G., Kivisild, T., Reddy, A.G., Singh, V.K., Rasalkar, A.A., Relethford, J., 2008. Genetic evidence and the modern human origins debate. Singh, L., 2005. Reconstructing the origin of Andaman Islanders. Science Heredity 100, 555e563. 308, 996. Richards, M., Corte-Real, H., Forster, P., Macaulay, V., Wilkinson-Herbots, H., Tosi, A., 2007. Primate distribution patterns: analyses of fossils, molecules, Demaine, A., Papiha, S.S., Hegdes, R., Bandelt, H.-J., Sykes, B., 1996. Paleolithic behavioral ecology, and paleoenvironment. Journal of Mammalian Evolution and Neolithic lineages in the European mitochondrial gene pool. American 14, 213e216. Journal of Human Genetics 59, 185e203. Vaks, A., Bar-Matthews, M., Ayalon, A., Matthews, A., Halicz, L., Frumkin, A., 2007. Ripley, S.D., Beehler, B.M., 1990. Patterns of speciation in Indian birds. Journal of Desert speleothems reveal climatic window for African exodus of early modern Biogeography 17, 639e648. humans. Geology 35, 831e834. Roberts, R.G., Jones, R., Spooner, N., Head, M.J., Murray, A., Smith, M.A., 1994. The Veron, G., Patou, M., Pothet, G., Simberloff, D., Jennings, A., 2007. Systematic status human colonisation of Australia: optical dates of 53,000 and 60,000 years and biogeography of the Javan and small Indian mongooses (Herpestidae, bracket human arrival at Deaf Adder Gorge, Northern Territory. Quaternary Carnivora). Zoologica Scripta 36, 1e10. Science Reviews 13, 575e583. Vidya, T., Fernando, P., Melnick, D., Sukumar, R., 2005. Population genetic structure Rodríguez, R., Ramírez, O., Valdiosera, C.E., García, N., Alda, F., Madurell- and conservation of Asian elephant (Elephas maximus) across India. Animal Malapeira, J., Marmi, J., Doadrio, I., Willerslev, E., Götherström, A., Arsuaga, J.L., Conservation 8, 377e388. Thomas, M.G., Lalueza-Fox, C., Dalén, L., 2011. 50,000 years of genetic uni- Wall, J., Lohmueller, K., Plagnol, V., 2009. Detecting ancient admixture and esti- formity in the critically endangered Iberian lynx. Molecular Ecology 20, 3785e mating demographic parameters in multiple human populations. Molecular 3795. Biology and Evolution 26, 1823e1827. Rose, J.I., 2010. New light on human prehistory in the Arabo-Persian Gulf Oasis. Walter, R., Buffler, R., Bruggemann, J.H., Guillaume, M., Berhe, S., Negassi, B., Current Anthropology 51, 849e883. Libsekal, Y., Cheng, H., Edwards, R.L., von Cosel, R., Néraudeau, D., Gagnon, M., Rose, J.I., Usik, V.I., Marks, A.E., Hilbert, Y.H., Galletti, C.S., Parton, A., Geiling, J.M., 2000. Early human occupation of the Red Sea coast of Eritrea during the last Cerny, V., Morley, M.W., Roberts, R.G., 2011. The Nubian Complex of Dhofar, Oman: interglacial. Nature 405, 65e69. an African Middle Stone Age Industry in Southern Arabia. PLoS One, e28239. Wang, C.-W., 1961. The Forests of China. Harvard University, Cambridge. Rosenberg, T.M., Preusser, F., Fleitmann, D., Schwalb, A., Penkman, K., Schmid, T.W., Weaver, T., Roseman, C., 2008. Genetic evidence and the modern human origins Al-Shanti, M.A., Kadi, K., Matter, A., 2011. Humid periods in southern Arabia: debate. Evolutionary Anthropology 17, 69e80. windows of opportunity for modern human dispersal. Geology 39, 1115e1118. Westaway, K.E., Morwood, M., Sutikina, T., Moore, M.W., Rokus, A., van den Satta, Y., Takahata, N., 2004. The distribution of the ancestral haplotype in finite Bergh, G.D., Roberts, R.G., Saptomo, E.W., 2009. Homo floresiensis and the late stepping-stone models with population expansion. Molecular Ecology 13, 877e Pleistocene environments of eastern Indonesia: defining the nature of the 886. relationship. Quaternary Science Reviews 28, 2897e2912. Sattenspiel, L., 2000. Tropical environments, human activities, and the transmission Westley, K., Dix, J., 2006. Coastal environments and their role in prehistoric of infectious diseases. Yearbook of Physical Anthropology 43, 3e31. migrations. Journal of Maritime Archaeology 1, 9e28. Sauer, C., 1962. Seashore e primitive home of man? Proceedings of the American White, T., Asfaw, B., DeGusta, D., Gilbert, H., Richards, G., Suwa, G., Howell, F.C., 2003. Philosophical Society 106, 41e47. Pleistocene Homo sapiens from Middle Awash, Ethiopia. Nature 423, 742e747. Scally, A., Durbin, R., 2012. Revising the human mutation rate: implications for Wu, X., 2004. On the origin of modern humans in China. Quaternary International understanding human evolution. Nature Reviews Genetics. http://dx.doi.org/ 117, 131e140. 10.1038/nrg3295. Wurster, C., Bird, M.I., Bull, I., Creed, F., Bryant, C., Dungait, J., Paz, V., 2010. Forest Schillaci, M., 2008. Human cranial diversity and evidence for an ancient lineage of contraction in north equatorial Southeast Asia during the Last Glacial Period. modern humans. Journal of Human Evolution 54, 814e826. Proceedings of the National Academy of Sciences of the United States of Serre, D., Laganey, A., Chech, M., Teschler-Nicola, M., Paunovic, M., Mennecier, P., America 107, 15508e15511. Hofreiter, M., Possnert, G., Paabo, S., 2004. No evidence of Neandertal mtDNA Zohary, M., 1973. Geobotanical Foundations of the Middle East. Swets & Zeitlinger, contribution to early modern humans. PLoS Biology 2, e57. Amsterdam. N. Boivin et al. / Quaternary International 300 (2013) 32e47 47

Further reading Neumann, K., 1991. In search of the green Sahara: palynology and botanical macro- remains. Palaeoecology of Africa 22, 203e212. Prabhu, C., Shankar, R., Anupama, A., Taieb, M., Bonnefille, R., Vidal, L., Prasad, S., Campbell, D., Hammond, H., 1989. Floristic Inventory of Tropical Countries. New 2004. A 200-ka pollen and oxygen-isotopic record from two sediment cores York Botanic Garden, New York. from the eastern Arabian Sea. Palaeogeography, Palaeoclimatology, Palae- Davis, S., Heywood, V., Hamilton, A., 1994. Centres of Plant Diversity: a Guide and oecology 214, 309e321. Strategy for Their Conservation. IUCN, Cambridge. Rattray, J., 1960. The Grass Cover of Africa. FAO, Rome. Dennell, R.W., Martinón-Torres, M., Bermudez de Castro, J.M., 2011. Hominin vari- Roebroeks, W., 2006. The human colonisation of Europe: where are we? Journal of ability, climatic instability and population demography in Middle Pleistocene Quaternary Science 21, 425e435. Europe. Quaternary Science Reviews 30, 1511e1524. Shi, Y., Kong, Z., Wang, S., Tang, L., Wang, F., Yao, T., 1993. Mid-Holocene climates Djamali, M., de Beaulieu, J.-L., Shah-nosseini, M., Andrieu-Ponel, V., Ponel, P., and environments in China. Global and Planetary Change 7, 219e233. Amini, A., Akhani, H., Leroy, S.A.G., Stevens, L., Lahijani, H., Brewer, S., 2008. Stevens, L.R., Djamali, M., Andrieu-Ponel, V., de Beaulieu, J.-L., 2012. Hydroclimatic A late Pleistocene long pollen record from Lake Urmia, NW Iran. Quaternary variations over the last two glacial/interglacial cycles at Lake Urmia, Iran. Research 69, 413e420. Journal of Paleolimnology 47, 645e660. Djamali, M., Baumel, A., Brewer, S., Jacksoon, S.T., Kaderit, J.W., Lopez-Vinyallonga, S., van Campo, E., Duplessy, J., Rossignol-Strick, M., 1982. Climatic conditions deduced Mehregan, I., Shabanian, E., Simakova, A., 2012. Ecological implications of Cous- from a 150-kyr oxygen isotope-pollen record from the Arabian Sea. Nature 296, inia Cass. (Asteraceae) persistence through the last two glacialeinterglacial 56e59. cycles in the continental Middle East for the Irano-Turanian flora. Review of van Zeist, W., Bottema, S., 1991. Late Quaternary Vegetation of the Near East. Palaeobotany and Palynology 172, 10e20. Reichart, Wiesbaden. Hayashi, R., Takahara, H., Hayashida, A., Takemura, K., 2010. Millennial-scale vege- Walter, H., 1971. Ecology of Tropical and Subtropical Vegetations. Oliver and Boyd, tation changes during the last 40,000 yr based on a pollen record from Lake Edinburgh. Biwa, Japan. Quaternary Research 74, 91e99. Wasylikowa, K., 2005. Palaeoecology of Lake Zeribar, Iran, in the Pleniglacial, Hope, G., Kershaw, A.P., van der Kaars, S., Xiangjun, S., Liew, P.-M., Heusser, L., Lateglacial and Holocene, reconstructed from plant macrofossils. The Holocene Takahara, H., McGlone, M.S., Miyoshi, N., Moss, P., 2004. History of vegetation e and habitat change in the Austral-Asian region. Quaternary International 118e 15, 720 735. White, F., 1983. The Vegetation of Africa: a Descriptive Memoir to Accompany the 119, 103e126. UNESCO/AETFAT/UNSO Vegetation Map of Africa. UNESCO, Paris. Li, B., Sun, J., 2004. Vegetation and climate environment during the Late Pleistocene in Whitmore, T., 1984. Tropical Rain Forests of the Far East. Clarendon Press, Oxford. Loess Plateau, China. Geographical Research 23 (5), 641e648 (in Chinese). Yu, G., Chen, X., Ni, J., Cheddadi, R., Guiot, J., Han, H., Harrison, S., Huang, C., Ke, M., Mason, K., 1946. Western Arabia and the Red Sea. Naval Intelligence Division, Kong, Z., Li, S., Li, W., Liew, P.-M., Liu, G., Liu, J., Liu, Q., Liu, K.-B., Prentice, I., London. Qui, W., Ren, G., Song, C., Sugita, S., Sun, X., Tang, L., van Campo, E., Xia, Y., Xu, Q., Migahid, A., 1998. Flora of Saudi Arabia. University Libraries, King Saud University, Yan, S., Yang, X., Zhao, J.-X., Zheng, Z., 2000. Palaeovegetation of China: a pollen Riyadh. data-based synthesis for the mid-Holocene and Last Glacial Maximum. Journal Moore, A., Hillman, G., Legge, A., 2000. Village on the Euphrates: from Foraging to of Biogeography 27, 635e664. Farming at Abu Hureyra. Oxford University Press, New York.