sign in sign up
<<
Home , Burin (lithic flake), Hafting, Middle Paleolithic

Evolutionary Anthropology 24:149–164 (2015)

ARTICLE

Rethinking the Dispersal of Homo sapiens out of

HUW S. GROUCUTT, MICHAEL D. PETRAGLIA, GEOFF BAILEY, ELEANOR M. L. SCERRI, ASH PARTON, LAINE CLARK-BALZAN, RICHARD P. JENNINGS, LAURA LEWIS, JAMES BLINKHORN, NICK A. DRAKE, PAUL S. BREEZE, ROBYN H. INGLIS, MAUD H. DEVES, MATTHEW MEREDITH-WILLIAMS, NICOLE BOIVIN, MARK G. THOMAS, AND AYLWYN SCALLY

Current fossil, genetic, and archeological data indicate that Homo sapiens ori- (Table 1). These models empha- ginated in Africa in the late Middle . By the end of the Late Pleisto- size different hypotheses concerning cene, our species was distributed across every continent except Antarctica, factors such as when dispersals setting the foundations for the subsequent demographic and cultural changes of began, how many occurred and which the . The intervening processes remain intensely debated and a key routes were followed. Recent models theme in hominin evolutionary studies. We review archeological, fossil, environ- have largely fallen into two broad cat- mental, and genetic data to evaluate the current state of knowledge on the dis- egories, emphasizing Marine Isotope persal of Homo sapiens out of Africa. The emerging picture of the dispersal Stage (MIS) 5 (early onset dispersal process suggests dynamic behavioral variability, complex interactions between model) or post-MIS 5 (late dispersal populations, and an intricate genetic and cultural legacy. This evolutionary and model) time frames (Table 1). This, historical complexity challenges simple narratives and suggests that hybrid however, is not a rigid dichotomy. For models and the testing of explicit hypotheses are required to understand the example, models proposing an early expansion of Homo sapiens into Eurasia. onset to dispersal are consistent with subsequent post-MIS 5 dispersals having also played an important role A variety of dispersal models remain unclear, but a variety of in patterns of diversity. (Table 1) address the period between recent studies add important new the widely accepted African origin of data. FOSSIL EVIDENCE Homo sapiens by around 200-150 ka Whereas earlier models focused on and the arrival of our species at the assessing the geographical origins of Hominin fossil remains provided margins of the Old World, including our species based on fossil data, more the initial foundations for the Out Australia, Siberia, and northwest recent approaches seek to combine of Africa model.3 Future fossil dis- , by 50-40 ka.1–4 The evolu- fossil, genetic, archeological, and coveries in Southern Asia have the tionary, demographic, and cultural paleoenvironmental data to illumi- potential to radically transform our processes between these milestones nate the nuances of dispersal into understanding of that dispersal. Early

Huw S. Groucutt, Eleanor M. L. Scerri, Michael D. Petraglia, Ash Parton, Laine Clark-Balzan, Richard P. Jennings, Laura Lewis, Nicole Boivin School of , Research Laboratory for Archaeology and the of Art, University of Oxford, Oxford, OX1 3QY, United Kingdom. Corresponding author: Huw Groucutt. Email: [email protected] Geoff Bailey, Robyn H. Inglis, Maud H. Deve` s, Matthew Meredith-Williams Department of Archaeology, University of York, York, YO1 7EP, United Kingdom James Blinkhorn McDonald Institute for Archaeological Research, University of Cambridge, Downing Street, Cambridge CB2 3DZ Nick A. Drake, Paul S. Breeze Department of Geography, King’s College London, WC2R 2LS, United Kingdom Maud H. Deve` s Laboratoire Tectonique, Institut de Physique du Globe, Paris, 75252, Mark G. Thomas Research Department of Genetics, and Environment, University College London, London, WC1E 6BT, United Kingdom Aylwyn Scally Department of Genetics, University of Cambridge, Cambridge, CB2 3EH, United Kingdom

Key words: ; Homo sapiens; dispersal; demography; genetic ancestry

VC 2015 Wiley Periodicals, Inc. DOI: 10.1002/evan.21455 Published online in Wiley Online Library (wileyonlinelibrary.com). 150 Groucutt et al. ARTICLE

TABLE 1. Summary of Selected Key Models for the Dispersal of Homo sapiens out of Their Place of Origin, Presented in Broadly Chronological Order of Formulationa

Examples of key Model Inferred timing of dispersal Description references Models focusing on Homo sapiens’ origins Asian cradle model Unclear Asia as birthplace of 5 H. sapiens The SW Asian/NE African Unclear SW Asia and NE Africa as 6 cradle model cradle in which H. sapiens evolved and from which it subsequently dispersed Multiregional model Throughout the Pleistocene H. sapiens evolved simultane- 7 ously in several parts of the world, with species integrity maintained by recurrent Recent African Origin (RAO) 100-40 ka H. sapiens evolved in Africa, 3,4 model perhaps in one region such as East Africa, and subsequently dispersed RAO and hybridization model 100-40 ka Accepts RAO, but infers 8–10 greater levels of hybridiza- tion with other hominin species in Eurasia Assimilation model 100-40 ka H. sapiens evolved in Africa, 11,12 but subsequent spread represents gene flow rather than replacement Variants of Late Dispersal Model Multiple dispersals model MIS 4 (70 ka) MIS 4 dispersal by southern 13 then MIS 3 (50 ka) route to Australia, then MIS 3 dispersal of populations by northern route model 50-45 ka Successful out of Africa 14,15 occurred after 50 ka, with derived such as projectiles MIS 4 single coastal dispersal 75-60 ka Structure of mtDNA tree inter- 16 model preted as indicating disper- sal around MIS 4 Single coastal dispersal with 60-50 ka A single dispersal out of Africa 17 geometric technology followed a coastal route, model marked by a trail of geo- metric and symbolic artifacts Variants of Early Onset Dispersal Model Early onset multiple dispersal Beginning in MIS 5 (130-75 ka), Multiple dispersals out of 2,18 model also MIS 3 (55-45 ka) key Africa, associated with climatic “windows of opportunity.” Early dispersals associated with MP technology “Jebel Faya” model 130 ka Dispersal out of Africa across 19 southern Arabia with Levallois, . and bifacial technologies in MIS 5e “Nubian” model by 106 ka Dispersal out of Africa by MIS 20 5c marked by presence of “beaked” (“Nubian”) Leval- lois technology in Arabia aKey references may considerably postdate the initial formulation of particular models. ARTICLE Rethinking the Dispersal of Homo sapiens out of Africa 151

Homo sapiens was morphologically suggest that interbreeding between which is provisionally assigned to variable.4,21 Traits that characterize the species took place elsewhere, that Homo sapiens.28 Several sites in Homo sapiens include: neurocranial small populations of Homo sapiens China are claimed to demonstrate the globularity; a divided supraorbital survived into MIS 4 in the Levant, or presence of Homo sapiens by MIS 5. torus and central and lateral portions; that Homo sapiens reoccupying the However, these sites, which often the face retreated below the forepart Levant in MIS 3 encountered late have poor stratigraphic and chrono- of the brain; a bony chin, even in . logical control, have produced taxo- infants; a gracile tympanic bone; the Vast areas of Asia have yet to pro- nomically ambiguous specimens.25 absence of an iliac pillar; and a short, duce any Pleistocene hominin fos- 3 26,27 thickened superior pubic ramus. Yet sils. Sites further east face dating GENETIC EVIDENCE even at a single site, morphological problems and taxonomic ambiguities variability can be striking. Omo- associated with elements such as The first reconstructions of human Kibish 1, for instance, strongly teeth and foot bones that are not genetic ancestry were based on data expresses the derived features of strongly diagnostic of species.25,28,29 from mitochondrial DNA (mtDNA), Homo sapiens, whereas Omo-Kibish While the relatively understood Y, and a small number 30–32 2, which is believed to be of similar Levantine record may provide a null of nuclear loci. Much of what age, is much more archaic.3 Given hypothesis for demographic change they revealed, such as evidence of a small samples, it is difficult to tell if across a wider area of Southwest recent African origin of Homo sapi- the variation of early Homo sapiens Asia,24 this must be qualified by the ens, remains central to our under- represents intrapopulation variation standing today. However, recent or the existence of several highly developments have changed the structured populations by the later nature of the genetic evidence for Middle Pleistocene. Nevertheless, the the fossil record is most and dramatically fossil record is most parsimoniously parsimoniously inter- increased its scope. New sequencing interpreted as demonstrating the technologies and computational piecemeal development of Homo sapi- preted as demonstrating approaches have enabled large-scale ens in Africa during the later Middle the piecemeal develop- whole-genome analyses of human Pleistocene. populations, while the ability to The earliest known Homo sapiens ment of Homo sapiens in recover ancient DNA sequences from fossils from outside Africa are found Africa during the later fossils has extended our view of in the Levant, one of the few relatively Middle Pleistocene. genetic diversity by tens of millennia intensively studied areas in Asia, at into the past. the sites of Skhul (120-90 ka) and These developments have led to Qafzeh (100-90 ka).3 These fossils important revisions in how we under- display numerous derived traits, with stand the ancestry of and atypical ecological features of the 33 a small number of primitive (archaic) other hominins. For example, it is Levant as a Mediterranean biome in features. Subsequently, Homo sapiens clear that this ancestry involves a a region more widely characterized are present in the Levant from around much greater degree of demographic by the particularities of the much 43 or 42 ka,22 and perhaps at around complexity than could previously be 23 larger Saharo-Arabian biome. 55 ka (75.2–33.6 ka) at Manot . resolved, with evidence of pervasive Dennell has reviewed the fossil In the latter case, however, the age gene flow and admixture between pop- record for Homo sapiens between Ara- ulations.9,10,34,35 The relationship estimates come from a calcitic patina 25 or crust that covers the calvaria; bia and Australia. The oldest South between genetic ancestry and demo- hence, these are minimum dates. The Asian fossils found, from the Sri Lan- graphic history is less straightforward specimen may be considerably older kan cave of Fa Hien, date to 33-30 than it was often assumed to be, and and/or may not reflect dispersal from ka. In Southeast Asia, the oldest fos- requires more sophisticated inferential Africa. Stalagmites from the site dem- sils are from the cave of Tam Pa Ling approaches. Inferences based on the 29 onstrate a hiatus in speleothem for- in Laos and date to  65-45 ka. genealogy of a single genetic locus, mation between late MIS 5 and MIS This age admits the possibility that such as the mtDNA tree, can be prob- 3. fossils have been dis- Homo sapiens either left Africa earlier lematic, particularly for older events. covered in the Levant and elsewhere than the Upper Paleolithic model sug- Such a genealogy represents one ran- in Asia dating to 70-50 ka.24,25 Fossil gests or that dispersal was extremely dom outcome of the genealogical pro- data are consistent with archeological rapid, as hypothesized by coastal dis- cess, the shape of which is only weakly discoveries in suggesting that in the persal models (Table 1). Several new constrained by demography. More- , reliably dated Nean- but preliminary findings suggest that over, while a simple tree is inadequate derthals are present in the Levant Homo sapiens may have arrived in to describe the complexity of human only after MIS 5, when Homo sapiens, Southeast Asia earlier than was previ- ancestral demography, genealogies possibly aside from the ously thought. The site of Callao Cave are always strictly tree-like. cranium, appear to be absent.21 This in the Philippines has produced a The timing of the dispersal of apparently asynchronous timing may hominin metatarsal dating to 67 ka, Homo sapiens out of Africa is a case 152 Groucutt et al. ARTICLE

Box 1. The mtDNA Genealogy and the Chronology of Dispersal out of Africa

Human mtDNA exhibits a geneal- Africa.93 This has led to the argu- been tree-like (Fig. 1A).13 If, ogy in which all haplotypes (unique ment that the MRCA of L3 is an instead, we allow for gene flow DNA sequences) in present-day non- upper boundary on the exit from between the ancestral African and Africans are placed within a clade or Africa, and that the mtDNA geneal- non-African populations after an haplogroup, the most recent com- ogy is incompatible with an earlier earlier initial divergence (Fig, 1B, mon ancestor (MRCA) of which has presence outside Africa for the C), then it is possible for L3 to been dated to 79 to 60 ka.36 This ancestors of present-day humans. have arisen during this period in haplogroup, named L3, encompasses This argument, however, rests one or the other population and several others that are found in mod- largely on the assumption that still be found in both populations ern Africans, predominantly in East human demographic history has today. Indeed, evidence for ongoing

Box Figure 1. Alternative models of the relationship between the mtDNA genealogy and demographic history. Models B and C illustrate the possibility of an early divergence of African and non-African ancestors to, with subsequent gene flow, potentially congruent with fossil and archeological evidence of a dispersal of Homo sapiens out of Africa ca. 100 ka. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

TABLE 2. Demographic parameters and ms commands used for coalescent simulation, Representing exponential growth

starting 15 ka in all three populations from an ancestral Ne of 1,500 to a present-day Ne of 25,000 (corresponding to auto- somal Ne growth from 6,000 to 100,000, assuming equal male and female Ne); African ancestral Ne of 2,500 100ka; non- African Ne of 500 from 55–40 ka in model A and 80–40 ka in models B and C; for model B, migration of 5 individuals per generation out of Africa 60–50 ka; for model C, migration of 1.2 individuals per generation into Africa 100–50 ka. Relative likelihood is the proportion of simulations for which the age of the L3 node was less than or equal to 80 ka, relative to this value for model A. Model Relative likelihood A Late dispersal from Africa 55 ka 1.0 ms 600 20000 -t 1.25 -I 3 200 200 200 -T -eN 0 10 -eG 0 56.27 -eG 0.05 0 -ej 0.1167 3 2 - en 0.1333 2 0.2 -ej 0.1833 2 1 -en 0.3333 1 1 45369 44223 59953 B Early dispersal 120 ka with gene flow from Africa ca 55 ka 0.6 ms 600 20000 -t 1.25 -I 3 200 200 200 -T -eN 0 10 -eG 0 56.27 -eG 0.05 0 -ej 0.1167 3 2 - en 0.1333 2 0.2 -em 0.1667 2 1 100 -em 0.2 2 1 0 -en 0.2667 2 0.32 -en 0.3333 1 1 -ej 0.4 2 1 45539 39872 63447 C Early dispersal 120 ka with subsequent gene flow back into Africa 1.3 ms 600 20000 -t 1.25 -I 3 200 200 200 -T -eN 0 10 -eG 0 56.27 -eG 0.05 0 -ej 0.1167 3 2 - en 0.1333 2 0.2 -em 0.1667 1 2 2 -en 0.2667 2 0.32 -em 0.3333 1 2 0 -en 0.3333 1 1 -ej 0.4 2 1 4984 41383 33507 ARTICLE Rethinking the Dispersal of Homo sapiens out of Africa 153

In each simulation, the coales- cent tree was inspected and the youngest node ancestral to all non- Africans and some Africans was identified as the node equivalent to L3. (Note that there will always be such a node in any genealogy, but under an arbitrary or unstructured demography it will often coincide with the global root.) Box Fig. 2 shows distributions of the age of this node for each model. Relative empirical likelihoods of L3  80 ka are given in Table 2. The models simulated represent Box Figure 2. Distribution of the age of the L3 node in 20,000 simulations for each of plausible demographic the models listed in Table 2. The vertical dashed line at 80 Ka indicates the maximum based on genetic and other evi- estimated age of L3 in the human mtDNA tree. [Color figure can be viewed in the dence for human . They online issue, which is available at wileyonlinelibrary.com.] demonstrate the range of variation in the age of the L3 node that one gene flow has been inferred in To explore these possibilities, we might expect under similar scenar- genome-wide analyses.39,94 simulated examples of each of the ios. However, they are not formally It has been argued that such models in Box Fig. 1 using a coales- fitted to the observed human alternative models can be dis- cent approach and investigated the mtDNA genealogy; indeed, a value counted on the grounds that if they resulting distributions of L3 node of 80 ka or less is unlikely under all were true some non-L3 haplotypes age. Simulated mtDNA genealogies these models (only 9% of simula- would be found in present-day were generated for three popula- tions under Model A). It would be non-Africans.17 However, model B tions representing present-day Afri- possible to infer parameters maxi- involves the fixation of migrant lin- can, European, and Asian samples, mizing this likelihood, but infer- eages from Africa within the ances- with the European and Asian popu- ence based on one node in a single tral non-African population lations diverging at 35 ka, and the genealogy would not be robust. following secondary gene flow. initial African/non-African split Nevertheless, these results show This occurs in mtDNA, but not nec- occurring at 55 ka in model A and that alternative models can be con- essarily at autosomal loci due to 120 ka in models B and C. For structed involving an early disper- their much higher effective popula- model B, we simulated a short sal out of Africa for which a recent age for L3 is not substantially less tion size (Ne), and reflects the fact period of strong migration (5 indi- that mtDNA genealogies are poten- viduals per generation) out of likely than under an equivalent tially more susceptible to such Africa from 60–50 ka, while model model of late dispersal. Without migration and introgression events. C featured an extended period of further evidence of the geographi- Model C involves a recent coales- weaker back-migration (1.2 individ- cal extent and structure of human cence of mtDNA lineages outside uals per generation) into Africa populations during this period, an Africa combined with migration of from 100–50 ka. Full demographic absence of non-L3 haplotypes out- one or more lineages back to parameters and simulation com- side Africa today cannot be Africa. Both of these possibilities mands are listed in Table 2. For regarded as conclusive evidence are made more likely by a low each model, we generated 20,000 against a dispersal of Homo sapiens out of Africa beginning by MIS 5. ancestral non-African Ne. Further, simulations of 600 samples using whole-genome analyses have shown ms95; increasing the number of It is also worth noting that a more that the ancestors of present-day samples had a negligible effect on recent analysis incorporating non-Africans experienced a pro- results since the vast majority of ancient DNA estimated an age of 95-62 ka for L3,43 representing a found reduction in autosomal Ne coalescent events occur recently. to below 3,000 for much of the Conservatively, we took a present- substantial overlap with MIS 5. period between MIS 5 and the day tgen of 30 yr as valid for the Holocene.43,46 whole of human prehistory.96

in point (Box 1). It has been argued mtDNA and Y chromosomal trees ever, as Box 1 shows, this argument, that the chronology and spatial dis- are inconsistent with dispersal any which assumes straightforward cor- tribution of branches of the human earlier than 60-50 ka.17,36,37 How- respondence between genealogical 154 Groucutt et al. ARTICLE trees and demographic history, is rapid population growth following it, cast doubt on the hypothesis of an not valid under plausible alternative which may or may not correspond to a earlier human exodus. This finding models of divergence, with gene flow major migration event such as a popu- also suggest that genetic signals between subpopulations over tens of lation dispersal. This topology also of population divergence before millennia. Support for such models coincides with the Ne minimum of the 60 ka relate to substructure within comes in part from recent evidence non-African bottleneck inferred from Africa. However, there are alterna- that much of the population struc- whole-genome analysis.46 Future anal- tive possibilities: ture in Africa is of surprisingly yses combining widely sampled (1) Since the method used to date ancient provenance.38 Further sup- mtDNA, Y-chromosomal, and genome- introgression assumed only a single port comes from genome-wide infer- wide data should provide a more episode of gene flow, earlier episodes ences of a gradual divergence with powerful means of inferring recent may have been undetected. (2) Nean- ongoing gene flow between African demographic processes. derthals may have ranged farther and non-African ancestors during Returning to earlier events, the north before 70-50 ka, so that Homo MIS 5.39 Indeed, with a revised esti- emerging picture is one in which by sapiens encountered them only some- mate of 0.5 x 109 bp21 y21 for the MIS 5 Homo sapiens existed within time after leaving Africa as part of a nuclear genomic rate, various subpopulations, differing in secondary Homo sapiens migration or whole-genome demographic studies their size and degree of genetic con- Neanderthal expansion southward. favor an older time scale and a more tact, and perhaps distributed over a (3) The vast majority of post-60-ka complex process of dispersal out of wide area. At least one of these human lineages may descend from a Africa.40 Evidence of this rate has included the ancestors of present- second wave of Homo sapiens out of come primarily from sequencing day non-Africans and was character- Africa at that time (model B in Box 1), studies of de novo . Some ized by low Ne. Based solely on in which case earlier episodes of concerns have been raised about present-day genetic data, which has interbreeding may have left a mini- the influence of false negatives on weak geographical resolution when mal genetic legacy. (4) Contact 41 such data. However, not only looking distantly into the past, it is between Homo sapiens and Neander- have more than a dozen such stud- difficult to say how far one or more thals may have occurred earlier but ies now arrived at similarly low val- of these populations might have unproductively, perhaps due to low 42 ues, but independent evidence extended into Asia before 60 ka. A hybrid viability or fertility, or other has also come from comparing major reason for this is the preva- reproductive obstacles.50 ancient and modern human DNA.43 lence of subsequent migration and The lower rate is more consistent gene flow, not only within Eurasia, ARCHEOLOGICAL EVIDENCE with inferences for the timing of but also from Eurasia back into recent events such as the diver- Africa and within Africa.47,48 Such Several dispersal models claim genceofNativeAmericanandEast events weaken the correlation support from patterns in archeologi- Asian populations.39 between present and ancestral haplo- cal and, particularly, lithic data These considerations are not to dis- type distributions and depend on (Figs. 1 and 2; Table 1). In evaluating pute the continuing value of mtDNA ecological and environmental factors the evidence underlying these claims, as a source of information on human that are challenging to model. it must be recognized that multiple evolution, particularly for more recent It may therefore be that some of processes, in this case branching events. It is still more widely sampled these questions can be resolved only (cultural inheritance and spread), than are the autosomes (chromo- by a combination of archeology and blending (cultural diffusion between somes 1–22) and more amenable to ancient DNA sequencing.49 Ancient populations), and convergence (inde- ancient DNA studies. mtDNA also has DNA has already been transformative pendent reinvention), can produce a smaller effective population size (Ne) in revealing interbreeding between similar forms of material culture (around one-quarter of the mean auto- Homo sapiens and other hominins, (equifinality). The first two of these somal value, depending on certain including Neanderthals, , processes can alternatively be demographic factors), meaning that and perhaps other archaic popula- described as and the latter patterns of diversity in mtDNA tions.9,10,49 In particular, one episode as analogy. sequences respond more rapidly to of Homo sapiens-Neanderthal inter- There are many examples of con- demographic changes. Thus, mtDNA breeding has now been dated to 60-50 vergence in . Partic- trees can be informative about more ka, based on the clear signature it left ular care must be taken when likely recent demographic history where in the ancient genomes of MIS 3 indi- drivers of independent reinvention, there are numerous uncoalesced viduals from Siberia.35,43 This has such as the constraints of , branches in the tree, albeit with the implications for the question of exist. In other ways, however, arche- caveats previously mentioned. For when Homo sapiens left Africa. ological data can provide robust sig- example, the “star-like” topology of Forexample,ifitrepresentsthe nals of dispersal. A key example non-African branches of the mtDNA earliest episode of interbreeding, relates to Australia, which, in con- and Y chromosome trees around 50 and if we expect interbreeding to trast to Asia, with its multiple homi- ka44,45 suggests an acutely reduced have begun as soon as humans left nin species, was peopled, as far as is non-African Ne at this time, as well as Africa, this finding would seem to known, only by Homo sapiens. ARTICLE Rethinking the Dispersal of Homo sapiens out of Africa 155

Figure 1. Selected lithics (stone ) from East and for MIS 5 (1–11) and MIS 3 (12–25). 1–4: iconic MIS 5 Middle Paleolithic (Middle ) lithic types and techniques of North Africa; 1–2: tanged or pedunculated points, widely thought to have been hafted tools68; 3, 4: “beaked” (“Nubian”) Levallois cores62; 5–11: other common components of North (5–7, from Bir Tarfawi, Egypt)64 and East (8–11, from BNS, Omo Kibish, )63; African MIS 5 MP assemblages; 5, 8, 9: recurrent centripetal Levallois cores; 6, 10: centripetally prepared Levallois flakes; 7, 11: retouched points. Late MP cores (12, 13) and retouched points (14, 15) and backed microlithic (16) from Mochena Borago, Ethiopia, 50 ka.55 17–25: Early Late Paleolithic (Later Stone Age) lithics from Enkapune Ya Muto, Kenya  50-40 ka; 17: end and side retouched flake; 18–24: backed flakes or ; 25: .103

Available data suggest that Homo The Evidence of MIS 3 Dispersals Most prominently, the appearance of sapiens had reached Australia before new lithic technology and other 50 ka.51 The Australian archeological Various models cite archeological aspects of material culture tradition- record provides a key minimum age data as indicating the dispersal of ally described as “Upper Paleolithic” for dispersal out of Africa. Homo sapiens into Asia 70-40 ka. in the Levant from 47/45 ka has 156 Groucutt et al. ARTICLE

Figure 2. Selected lithics from Southwest and South Asia from MIS 5 (1–12) and MIS 3 (13–27). 1–4: . 1: centripetally pre- pared preferential Levallois core, Jebel Faya, UAE, 125 ka; 2: bifacially flaked , Jebel Faya104; 3,4: beaked (or Nubian) Levallois cores from TH-59, Oman, probably MIS 520; 5–8: Qafzeh Cave, Israel, 100-90 ka; 5: recurrent centripetal Levallois core; 6: centripetally prepared preferential Levallois core; 7: Levallois flake; 8: side retouched Levallois flake.65 9–12: MIS 5 lithics from Jwalapurum 22, India, 75 ka. 9: recurrent centripetal Levallois core; 10: centripetally prepared preferential Levallois core; 11: tanged or pedunculated flake; 12: retouched point66; 13–16, typical artifacts of the Levantine Late Middle Paleolithic, Dederiyeh Cave, Syria, 60 ka. 13,14: unidirec- tional convergent Levallois cores; 15,16: Levallois points with unidirectional convergent preparation.105 17–20, Late Middle Paleolithic lithics from Jwalapurum 3 and 20, India, 55-30 ka, 17: centripetally prepared preferential Levallois core; 18: recurrent centripetal Leval- lois core; 19, 20: Levallois flakes.66 21–27: Early Upper Paleolithic lithics from the Levant,  40 ka. 21: blade core; 22–26: points; 27: blade.15 ARTICLE Rethinking the Dispersal of Homo sapiens out of Africa 157 been seen as evidence of the rearrival Late Paleolithic in South Asia was fac¸onnage reduction methods found of Homo sapiens from Africa (Fig. 2: dominated by blade and microblade at the Arabian site of Jebel Faya and 21–27).15 Key aspects of this include production, with microlithic tech- claims that these are similar to fea- the hypothesis that this dispersal nology becoming widespread from tures of the East and Northeast Afri- reflects the invention of projectile 38/35 ka. The oldest microlithic can Middle Paleolithic (Fig 2: 1–2).19 technology in Africa.15 Such develop- site in Sri Lanka, Fa Hien-lena, dat- An alternative model instead stresses ments perhaps gave Homo sapiens a ing to 38 ka, contains only nongeo- the discovery of “beaked” (or selective advantage over Neanderthal metric forms.59 “Nubian”) Levallois technology, previ- populations in Eurasia. However, the While these findings do not fit eas- ously best known from Northeast extent to which this model can be ily with the notion of microlithic Africa (Fig. 1: 3–4), in Arabia (Fig 2: 3– generalized beyond the Levant is technology as a unique marker of 4).20 The former emphasizes a rather currently unclear. the dispersal out of Africa, Mellars broad combination of features; the lat- Microlithic or geometric technolo- and colleagues emphasize the similar ter highlights one aspect of technology gies were variably present through- range of shapes, such as crescents, that may represent convergent (inde- out the African Middle Paleolithic lunates, and trapezoids, of some sub- pendent) evolution. The notion that (MP). (Here we use “Middle Paleo- sequent South Asian microliths and the “Nubian complex” is a spatially lithic” to include the synonymous the earlier African forms.17 To Mel- and temporally restricted technocom- and “Late Paleo- lars and colleagues, these similarities plex becomes problematic with the lithic” as a way of describing in shape are best explained by discovery of similar technologies from assemblages traditionally designated branching and blending processes of Mauritania to the Thar Desert via as Upper Paleolithic, Later Stone cultural interaction and dispersal. South Africa.67 The distinctive Age, or Microlithic). Mellars and col- Advocates of the notion of dispersal “beaked” or “Nubian” cores were first 17 62 leagues argue for dispersal into from Africa at around 60-50 ka also described by Seligman, who thought southern Asia by a coastal route cite the apparent distribution of that the shape of the median-distal before the origin of the Upper Paleo- “symbolic” artifacts such as beads and ridge was similar to the shape of a lithic in the Levant. They cite the dis- incised ostrich eggshell.17 Cited exam- tortoise beak. We propose use of the tribution of microlithic and ples include Batadomba-lena at 35 ka morphologically descriptive term geometric technologies around the and Jwalapurum 9 at 20-12 ka.58,59 “beaked” instead of “Nubian,” which Indian Ocean rim as providing evi- Mellars and colleagues argue that older implies an automatic association of dence of a single dispersal of Homo examples, as well as early examples of this technology with Nubia; emerg- sapiens out of Africa 55-50 ka (for microlithic technology, have been con- ing evidence may be consistent with example, Fig 1: 17–25). The Howie- cealed by sea-level rise.17 the convergent reinvention of this sons Poort (HP), emphasized by Mel- technology. 17 lars and colleagues, is an early MIS 5 DISPERSALS An alternative lithic Out-of-Africa example of technologies commonly signal may be the spread of East Afri- described as microlithic or geometric Archeological findings in southern can technologies, with a trail of simi- that become temporarily a common Asia have been interpreted as indi- lar assemblages linking East Africa,63 part of assemblages. However, recent cating early dispersals of Homo sapi- Northeast Africa,64 the Levant,65 and studies, emphasize the diversity of ens out of Africa (that is, by MIS 5, as far east as India by late MIS 5 (Fig. HP assemblages, for example in core 130-75 ka) (Table 1). While the 1: 5–11, Fig. 2: 5–12).51,66 The search reduction methods and in features of Middle Paleolithic in Europe is asso- for an archeological “smoking gun” retouched tools.52–54 ciated with Neanderthals, in Africa it for dispersal out of Africa is chal- In East Africa, microliths occur in not only temporally overlaps with lenged by the diversity of lithic tech- low frequencies from around 50 ka most of the period in which Homo nology within Africa before dispersal. and subsequently increase in fre- sapiens was present, but also charac- Scerri and colleagues, for instance, quency.55 For example, a single com- terizes the early expansion of Homo have demonstrated spatially struc- plete crescent was found at Mochena sapiens to the Levant.62–65 Given that tured lithic variability in MIS 5 North Borago (Fig. 1: 16) within a pattern available data indicate that humans Africa,68 correlating with modeled of general technological continuity. were in Southeast Asia and Australia ecozones rather than traditional Earlier claims face taphonomic before the origin of the Late Paleo- “industrial” nomenclatures such as and chronological problems.52 The lithic in Africa and Asia, there is a “Aterian” and “Nubian complex.” This onset of the Late Paleolithic in strong indication that at least an finding is interpreted as indicating East Africa appears to have been early phase of dispersal out of Africa structured (subdivided) populations initially characterized by a reduc- was associated with MP lithic by MIS 5. Such inferences represent tion in the size of lithics and an technology.66,67 Elucidating variabili- an important archeological finding increase in bipolar reduction. In ty within MP technologies is there- that can be factored into models using South Asia, there also appears to be fore of great importance for genetic data. a gradual and complex transition understanding dispersal.68 The South Asian MP shows consid- from the Middle to Late Paleo- One recent model emphasizes the erable technological continuity from lithic.56–61 The early phase of the combination of Levallois, blade, and later MIS 5 through MIS 3 (Fig. 2: 9– 158 Groucutt et al. ARTICLE

12, 17–20),66,67 suggesting that homi- ronmental data for North Africa has ductive environments. Field data, nin population continuity in the shown that MP sites tend to be most however, remain minimal. mosaic environments of South Asia technologically similar to nearby South African sites demonstrate occurred through the Toba super- sites except where they were con- the use of molluscs, fish, and sea eruption of 75 ka. It is possible, how- nected by rivers (see Box 2).68 An mammals exploited from the sea- ever, that there was also an earlier in alternative perspective hypothesizes shore alongside terrestrial foods, but situ Lower to MP transition, that coastal routes were key.17 they lack evidence of offshore activ- although this possibility requires fur- Environmental variation influences ity.83 Most of these findings relate to ther analysis. South Asian MP dispersal not only in terms of its MIS 5/4, but earlier examples may assemblages dating to MIS 5 effects on factors such as net primary be concealed by sea level rise. The beaked (“Nubian”) Levallois technol- productivity, which leads to demo- only other reported example of possi- ogy and other components common graphic fluctuation, but also by open- ble early coastal subsistence outside in the African record,66,67 but lack ing and closing routes, such as South Africa is the MIS 5 Eritrean the kind of technology associated through the generally arid Saharo- site of Abdur,79 where shells from with Neanderthals in at least the Arabian zone.75,76 The history of shellfish apparently collected for Levant (Fig. 2: 13–16) and central research cautions against directly cor- food have been reported alongside Asia.52 Key reviews of Asian paleoan- relating environmental and demo- stone tools and terrestrial mammals thropology include those by Den- graphic processes. For example, in a beach deposit. However, without nell,69 who focused on the pre-MIS 5 paleoclimate data from Lake Malawi further supporting details or evi- period, and Rabett,70 who addresses suggesting that “mega-droughts” dence of other marine resource use, the post-MIS 5 period. occurred in sub-Saharan Africa has Abdur’s place in the record of Pleis- In the case of the Levant, MP tech- been cited as a key mechanism tocene coastal exploitation is 84 nology is found alongside various explaining the apparent dispersal out unclear. There is nothing in these indications of complex behavior, of Africa at around 60 ka.17 However, findings to indicate that marine including deliberate burials, beads, subsequent research has demon- resources were uniquely associated and the use of , features that strated that the MIS 5 mega-drought with anatomically modern humans arguably articulate the record more ended much earlier, by 85 ka.77 or that they supported marine- closely with that of Africa than While numerous archives in the focused paleoeconomies. Other hom- Europe.24,71 Some of the earliest Saharo-Arabian belt attest to dra- inins, and even nonhominin prima- examples of such behavior, which matic increases in humidity during tes, also exploited marine foods on 85 are actually found at the non-African periods such as MIS 5,75,76 early MIS the seashore. site of Skhul (130-100 ka),71 and 3(60-50 ka) also witnessed a signif- Only in Sahul is there clear evidence more robustly at the slightly younger icant humid phase in Arabia.78 This of the conjunction of colonization, site of Qafzeh (100-90 ka), suggest MIS 3 wet period may have provided seafaring, and heavy dependence on that by at least MIS 5 Homo sapiens the context for a renewed phase of marine resources during the Pleisto- was capable of complex, including dispersal out of Africa and/or the cene. Occurring only in the archipel- symbolic, behavior.72,73 Such be- expansion of refugial populations ago environments of Wallacea and the havior was expressed in a variable already within Arabia. The evidence Bismarck islands, this may reflect a manner,perhapsinrelationto of interior humidity and hominin unique combination of circumstances: environmental or demographic fac- occupations some distance from the abundant bamboo driftwood for rafts; 74 tors. The current lack of evidence coast in Arabia in MIS 3 suggests an archipelago environment with of symbolic behaviors in areas such that a coastal route need not have favorable winds and currents where as Arabia may reflect factors such been exclusively followed (Box 2). It land is rarely out of sight; a depauper- as lack of research in a given area also suggests that terrestrial disper- ate island fauna and little available ter- or the fact that most sites discov- sals need not have been limited to restrial food; a rich and varied supply ered consist of raw material pro- MIS 5. of marine resources; and uplifted curement and early stage reduction coastlines preserving occupied during low sea level above the present localities. Coastal Dispersal Models shoreline.86–89 Once humans reached ENVIRONMENTS AND Recent have witnessed inten- landfall in Sahul, they rapidly moved sified debate about the role of coast- inland, leaving little evidence of pro- DISPERSAL ROUTES lines in hominin evolution and longed coastal settlement even on the The final major component of dispersal.17,27,79–83 At one extreme, tectonically uplifted coastlines of debate over the dispersal of Homo proponents of purely coastal routes northern New Guinea. sapiens out of Africa concerns the see those routes as providing a It is possible that analogous condi- routes taken and how they correlate mechanism for fast, directional, pop- tions existed on paleoshorelines else- more widely with ecological condi- ulation expansion along an ecologi- where around the Indian Ocean, but tions. Some have emphasized terres- cally uniform coastal highway.17 that these have been submerged by trial dispersal routes.27 For example, According to this view, MIS 3 coastal sea level rise.17 Testing this hypothe- combining archeological and envi- regions were highly stable and pro- sis requires surveys both underwater ARTICLE Rethinking the Dispersal of Homo sapiens out of Africa 159

Box 2. Paleoclimate, Paleohydrology and the Distribution of Middle Paleolithic Sites

The distribution of MP sites in the modeled 130 ka precipitation data ing has shown that the number of Saharo-Arabian belt provides derived from the down-scaled Com- recorded MPs declines considerably insights into Late Pleistocene land- munity Systems Model (CCSM 3) as distance from major rivers scape use, given that at least the data.100,101 Box Fig. 1 shows that increases, with 74% of MP sites early phase of dispersal out of Africa most MP sites in this region occur within 30 km of large paleodrainage was associated with MP toolkits. As in areas of increased rainfall; how- systems.99 This is ongoing research,

Box Figure 1. The distribution of Middle Paleolithic sites across East Africa, the Saharo-Arabian belt, and India, plotted on a mod- eled precipitation map for the last interglacial (MIS 5) with positions of major paleolakes (dark blue areas) and paleorivers, which form extensive riparian corridors (blue lines). The Neanderthal range line shows the estimated extent of Neanderthal dispersal from the north. The map shows that Middle Paleolithic sites are commonly located in interior regions and that their presence in typically arid areas can be explained by the humid climate conditions of periods such as MIS 5, which activated paleohydrologi- cal networks and potentially transformed major deserts into savannah grasslands and shrublands (green areas) containing numer- ous freshwater lakes and rivers. The paucity of sites in Pakistan and eastern almost certainly reflects research history rather than a real pattern. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

shown in Box Fig. 1, vast areas of ever, it also demonstrates the per- but initial results are congruent the generally arid Saharo-Arabian sistence of an arid desert belt even with hypotheses that fluvial net- belt were potentially transformed during times such as MIS 5e. Paleo- works formed important dispersal into grassland and savanna environ- rivers formed potentially crucial corridors, as Arabian MP sites are ments by increased rainfall during corridors (and filters) through these generally located much closer to several humid periods, including arid environments, as demonstrated major paleorivers than would be those of MIS 5. Box Fig. 1 combines for the Sahara during MIS 5e.68,102 expected if they were randomly paleohydrological data63,97–99 with In Arabia, paleohydrological model- distributed.99 and on land proximal to coast- Late Pleistocene regression high- scapes that have been submerged lines.81,89 A broad delineation of the lights the considerable variability in (Fig. 3). Intensive surveys in some F3 continental shelf around the Indian the width and topography of the areas proximal to narrow coastal Ocean approximating a sea level coastal shelf, and therefore in the shelves, such as the Dhofar coast of position of 2100 m at the maximum area and nature of Pleistocene land- Oman (Fig. 3D), have failed to reveal 160 Groucutt et al. ARTICLE

Figure 3. Map showing topography and bathymetry (SRTM30PLUS). Areas in yellow correspond to currently submerged land that would have been exposed when sea level was 100m lower than present. Colored dots correspond to key archeological sites (, MSA/LSA transitional, and South Asian Late Paleolithic) emphasized by the model of Mellars and coworkers.17 The inserts show that the landscape can be very different from place to place and that there is no “typical” coastal environment. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Late Pleistocene marine-focused great ecological diversity, as for their been combined with terrestrial archeology.20 The presence of Lower marine resources. Others are barren resources and the use of the hinter- Paleolithic and material or inaccessible because of the land through seasonal movements or indicates that this absence is not encroachment of steep mountain in symbiosis with inland commun- simply a question of preservation. ranges, estuarine mudflats, thick for- ities, the more typical pattern in Holo- Parts of the coastal margin of east- est, or desert. cene ethnographic and archeological ern Arabia have experienced tectonic Examples of Holocene Homo sapi- records.81 uplift, including during the Pleisto- ens marine specialist economies are Archeological sites that appear cene,90 but no archeological evidence rare. They typically occur in archipel- close to the coast on large-scale consistent with the expectations of ago environments with highly pro- maps are often much further inland. purely coastal models has been iden- ductive marine resources accessed For instance, the average distance tified on these landforms either. through advanced technologies, from the modern coast of the 31 Conversely, a large area of sub- including seafaring and food storage, Howiesons Poort (HP) sites we ana- merged coastal shelf occurs between often at high latitudes alongside lyzed is 105 km, despite a research East Africa and southwest Arabia, unproductive or inaccessible terres- bias toward coastlines. HP sites are making it a key area for further sur- trial environments (notably in Nor- up to 350 km from the modern coast vey both underwater and on neigh- way, the northwest coast of North (Rose Cottage Cave). Post-HP sites in boring land.89,91 America, and Tierra del Fuego). The South Africa are typically even fur- Coastal regions are extremely vari- possibility that similar economies ther from the coast than are HP able both in space and through time. existed on now-submerged Pleisto- sites. The East African sites that Some regions are attractive as much cene shorelines cannot be ruled out, have been argued to support a because of favorable conditions for but in any case such economies are coastal dispersal 55 ka17 are, on proximal terrestrial resources, likely to have been very patchy in average, 600 km from the coast; including abundant water supplies, their distribution, as in the Holocene. early South Asian Late Paleolithic river estuaries, equable climate con- To the extent that marine resources sites are also typically far inland. In ditions, a long growing season, and were exploited, they are likely to have Sri Lanka, early microlithic sites are ARTICLE Rethinking the Dispersal of Homo sapiens out of Africa 161

TABLE 3. Tabulation of the Relative Strength of Fossil and Archeological Evidence of the Presence of Homo sapiens in Southern Asia and Australia for selected time periodsa

Levant Arabia South Asia Southeast Asia Australia

Date (ka) Fossil Arch. Fossil Arch. Fossil Arch. Fossil Arch. Fossil Arch. 40 111 111 1 111 111 111 111 111 50 1 11 1 11 111 11 1 111 60 11 - 111111 75 111 11 111 -- 100 111 111 11 1 1 -- 140 11 -- a111 5 strong evidence, 11 5 moderate evidence, 15 weak evidence, - 5 relatively good evidence of absence, gray cell 5 uncertain/insufficient information to assign to one of these categories. For details see text.

generally inland and clearly demon- major cultural and demographic wise, similarities between lithic strate terrestrial subsistence, particu- changes around 50 to 30 ka. assemblages, as well as other sorts of larly the of monkeys.59 We suggest that accumulating data archeological data, can be explained Much remains to be learned about increasingly support a hybrid model by different mechanisms. A strong the role of coastal or marine resour- whereby early expansions were archeological argument for dispersal ces and habitats in the Pleistocene. essentially swamped by subsequent would involve the correlated appear- However, evidence of a dramatic ones. In terms of dispersal routes, ance of a package of several ele- to coastal ecologies and we suggest that populations em- ments of material culture. It is clear dispersal along coasts at or before ployed behavioral flexibility and that many key cultural features 50 ka is currently lacking. It is adaptation to use a range of different evolved convergently, among them likely that coastlines provided patches ecologies, including interior savan- Levallois and blade technology, as of favorable habitats, but that these nahs and the coast. Accumulating well as tangling or pedunculation. were discontinuous in space and evidence of early population struc- We interpret available data as indi- time. Rather than limiting dispersal ture and multiple population interac- cating the repeated and independent models to the strictly dichotomous tions indicates that simple models evolution of microlithic technology, “interior” or “coastal,” we suggest for the dispersal process are no lon- but acknowledge that testing this, as that, alongside a range of other Late ger sufficient. A key point is that the well as notions of early dispersals Pleistocene habitats, from the semi- Asian paleoanthropological record with Middle Paleolithic technology, arid to rainforests, the use of coastal remains extremely poor, as illus- requires comparative analyses of ecozones is best seen as part of the trated by the recent discovery of assemblages in Africa and Asia. behavioral flexibility of Homo sapiens. cave paintings dating to at least 40 Quantitative analysis of attributes are ka in Indonesia.92 one method that, critically, can derive Further work is needed to under- technological insights from typologi- DISCUSSION AND stand what constraints genetic and cally indistinct artifacts.54,68 CONCLUSIONS archeological data place on models Different regions have different of ancestral population dispersal strengths and weaknesses in deter- Our review of evidence relating to across the Middle East and South mining the presence of Homo sapi- the timing and routes of dispersal of Asia. We have shown that inferences ens (Table 3). The Levant seems to Homo sapiens out of Africa shows from single-locus genetic data need feature a genuine occupational hia- that fossil, genetic, and archeological to be based on an understanding of tus. The data from Arabia remain data are currently consistent with the relationship between demogra- somewhat ambiguous, with the several different models. There is phy and genealogical stochasticity, strongest archeological evidence con- much more uncertainty in the timing as embodied in coalescent or other sisting of MIS 5 lithics displaying and character of this dispersal than population genetic models. Just as similarities to African and Levantine is generally suggested by those who archeology has largely, but not material associated with Homo sapi- argue that the first successful disper- entirely, transcended the “culture- ens fossils, while post-MIS 5 Arabian sal, occurred 50 (610) ka.1,14,17 history” approach, according to lithics are culturally ambiguous.26 In Uncertainties remain about the ex- which and tool types were South Asia, fossil data are currently tent of cultural and biological conti- simplistically seen as direct proxies absent before 35 ka, while archeo- nuity in Homo sapiens populations for populations, so genetic analyses logical data provide moderate indica- outside Africa from MIS 5 onward, must avoid the “gene-history 5 tions of the presence of Homo as well as about the meaning of the population history” paradigm. Like- sapiens from MIS 5.66,67 In Southeast 162 Groucutt et al. ARTICLE

Asia, the archeological data are the formulation of clear hypotheses 6 Field H. 1932. The cradle of Homo sapiens. ambiguous regarding the presence of that make explicit predictions about J Archaeol 36:426–430. 7 Thorne AG, Wolpoff MH. 2003. The multire- Homo sapiens until 50 ka at sites patterns in different types of data. gional evolution of humans (revised paper). Sci such as Niah Cave.70 A series of ear- Enhancement of the southern Asian Am 13:46–53. lier fossils from across Southeast fossil and archeological records 8 Brauer€ G. 1992. Africa’s place in the evolu- tion of Homo sapiens. In: Brauer€ G, Smith F, Asia provide possible support for the remains critical, alongside the appli- editors. Continuity or replacement? Controver- presence of Homo sapiens back to cation of more ancient DNA sequenc- sies in Homo sapiens evolution. Rotterdam: 100 ka, and more securely to 70/60 ing. The interpretation of such future Balkema. p 83–98. ka.28,29 In Australia, archeological 9 Green RE, Krause J, Briggs AW, et al. 2010. findings will most robustly be A draft sequence of the Neandertal genome. data indicate the presence of Homo achieved within the context of multi- Science 328:710–722. sapiens by 50 ka, with the earliest disciplinary collaborations. 10 Reich D, Patterson N, Kircher M, et al. secure fossil evidence dating to 45 2011. Denisova admixture and the first modern ka. Cumulatively, these data demon- human dispersals into southeast Asia and Oce- ACKNOWLEDGMENTS ania. Am J Hum Genet 89:516–528. strate that Homo sapiens were in 11 Smith F. 1992. The role of continuity in Southwest Asia by 120 ka and We acknowledge funding support modern human origins. In: Brauer€ G, Smith F, Southeast Asia by 50 ka. The arche- from the European Research council editors. Continuity or replacement? Controver- sies in Homo sapiens evolution. Rotterdam: ological and fossil data between (ERC) to M. D. Petraglia (Advanced Balkema. p 145–156. these points can currently be inter- Grant 295719, “PALAEODESERTS: 12 Trinkaus E. 2005. Early modern humans. preted in different ways and, as we Climate Change and Hominin Evolu- Ann Rev Anthropol 34:207–230. have outlined here, genetic data have tion in the Arabian Desert: Life and 13 Lahr MM, Foley RA. 1998. Towards a theory of modern human origins: geography, demog- been subject to questionable inter- at the Cross-roads of the Old raphy and diversity in . pretations that require explicit mod- World”), G. Bailey/G. C. P. King Yearbk Phys Anthropol 41:137–176. eling to be formally tested. (Advanced Grant 269586, “DISPERSE: 14 Klein RG. 2009. The human career: human biological and cultural origins. Chicago: Chi- A number of predictions can be Dynamic Landscapes, Coastal Envi- cago University Press. made from various models and ronments and Human Dispersals”), 15 Shea JJ. 2011. The archaeology of an illu- tested in future research. Regarding and N. Boivin (Starter Grant 206148, sion: the Middle-Upper Paleolithic transition in the question of whether Homo sapi- the Levant. In: Le Tenesor JM, Jagher R, Otte “SEALINKS: Bridging Continents R, editors. The Lower and Middle Paleolithic in ens successfully dispersed into Asia Across the Sea”), all under the “Ideas” the Middle East and neighbouring regions. before 60 ka, several hypotheses specific Programme of the 7th Frame- ERAUL 126. Liege: UniversitedeLi ege. p 169– and expectations can be posed. 182. work Programme (FP7). M. G. 16 Oppenheimer S. 2012. A single southern exit These include the discovery of pre-60 Thomas is supported by a Wellcome of modern humans from Africa: before or after ka Homo sapiens fossils in Asia out- Senior Investigator Award in Medical Toba? Quat Int 258:88–99. side the Levant, as well as the dem- Humanities (Grant number: 100713/Z/ 17 Mellars P, Gori KC, Carr M, et al. 2013. onstration of similarities in material Genetic and archaeological perspectives on the 12/A). E. L. M. Scerri and J. Blinkhorn initial modern human colonization of southern culture reasonably explained by thank the Fondation Fyssen. L. Lewis Asia. Proc Natl Acad Sci USA 110:10699–10704. branching or blending in Africa and acknowledges the support of the Arts 18 Reyes-Centeno H, Ghirotto S, Detroit F, Asia before 60 ka. We expect further et al. 2014. Genomic and cranial phenotype and Humanities Research Council data support multiple modern human disper- analyses of the genomic divergence (UK, #513691), the Wenner-Gren sals from Africa and a southern route into Asia. between African and non-African Proc Natl Acad Sci USA 111:7248–7253. Foundation (#8684), and the Boise ancestors to reveal signals of gene 19 Armitage SJ, Jasim SA, Marks AE, et al. Fund (University of Oxford). P. Breeze flow and population substructure at 2011. The southern route “out of Africa”: evi- is funded by NERC studentship NE/ dence for an early expansion of modern this time and for ancestral demogra- J500306/1. We thank C. Stringer for humans into Arabia. Science 331:453–456. phy inferred from genetic data out- 20 Usik VI, Rose JI, Hilbert YH, et al. 2013. discussion on the morphological defi- side Africa to reflect dispersal into Nubian Complex reduction strategies in Dho- nition of Homo sapiens. far, southern Oman. Quat Int 300:244–266. Asia from MIS 5. 21 Gunz P, Bookstein FL, Mitteroecker P, et al. A difficulty in inferring the routes 2009. diversity suggests and chronology of Homo sapiens’ dis- REFERENCES subdivided population structure and a complex out-of-Africa scenario. Proc Natl Acad Sci USA persal out of Africa is that it requires 1 Willoughby PR. 2007. The evolution of mod- 106:6094–6098. ern humans in Africa; a comprehensive guide. integration of many different sources 22 Douka K. 2013. Exploring “the great wilder- of information, each with its own New York: Altimira. ness of prehistory”: the chronology of the Mid- 2 Petraglia MD, Haslam M, Fuller DQ, et al. ambiguities and assumptions. In dle to the Upper Palaeolithic transition in the 2010. Out of Africa: new hypotheses and evi- Northern Levant. Mitt Gessell Urgeschichte 22: addition, quite distinct processes can dence for the dispersal of Homo sapiens along 11–40. generate very similar patterns of vari- the Indian Ocean rim. Ann Hum Biol 37: 288– 23 Hershkovitz I, Marder O, Ayalon A. 2015. 311. ation in both genetic and archeologi- Levantine cranium from Manot Cave (Israel) 3 Stringer C. 2011. The origin of our species. foreshadows the first European modern cal data (equifinality). If dispersal out London: Penguin. humans. Nature 520:216–219. of Africa occurred in several waves, 4 Stringer C. 2014. Why we aren’t all multire- 24 Shea JJ. 2003. The Middle Paleolithic of the then it was neither exclusively “early” gionalists now. Trends Ecol Evol 29: 248–251. East Mediterranean Levant. J World Prehist 17: nor “late,” but both. Clarification of 5 McCown TD, Keith A. 1939. The Stone Age of 313–394. Mt. Carmel, vol 2. The fossil human remains 25 Dennell R. 2014. Smoke and : the this important issue requires better from the Levalloiso-. Oxford: Clar- fossil record for Homo sapiens between Arabia cross-disciplinary understanding and endon Press. and Australia. In: Dennell R, Porr M, editors. ARTICLE Rethinking the Dispersal of Homo sapiens out of Africa 163

Southern Asia, Australia and the search for 44 Underwhill PA, Kivisild T. 2007. Use of Y 61 Deraniyagala SU. 1992. The prehistory of human origins. Cambridge: Cambridge Univer- chromosome and mitochondrial DNA popula- Sri Lanka: an ecological perspective. Colombo: sity Press. p 33–50. tion structure in tracing human migrations. Department of Archaeological Survey. 26 Groucutt HS, Petraglia MD. 2012. The pre- Ann Rev Genet 41:539–564. 62 Seligman CG. 1921. The older Palaeolithic history of the Arabian Peninsula: deserts, dis- 45 Wei W, Ayub Q, Chen Y, et al. 2013. A cali- age in Egypt. J R Anthropol Inst GB Ireland 51: persals and demography. Evol Anthropol 21: brated human Y-chromosal phylogeny based 115–153. 113–125. on resequencing. Genome Res 23:388–395. 63 Shea JJ. 2008. The Middle Stone Age 27 Boivin N, Fuller DQ, Dennell R, et al. 2013. 46 Li H, Durbin R. 2011. Inference of human archaeology of the Lower Omo Valley Kibish Human dispersal across diverse environments population history from individual whole- formation: excavations, lithic assemblages, and of Asia during the Upper Pleistocene. Quat Int genome sequences. Nature 475:493–496. inferred patterns of early Homo sapiens behav- 300:32–47. 47 Pagani L, Kivisild T, Tarekegn A, et al. 2012. ior. J Hum Evol 55:448–485. 28 Mijares SR, Detroit F, Piper P, et al. 2010. Ethiopian genetic diversity reveals linguistic 64 Wendorf F, Schild R, Close AE, et al. 1993. New evidence for a 67,000--old human stratification and complex influences on the Egypt during the Last Interglacial: the Middle presence at Callao Cave, Luzon, Philippines. Ethiopian gene pool. Am J Hum Genet 91:83– Paleolithic of Bir Tarfawi and Bir Sahara East. J Hum Evol 59:123–132. 96. Plenum: New York. 29 Demeter F, Shackelford L, Westaway K, 48 Gurdasani D, Carstensen T, Tekola-Ayele F, 65 Hovers E. 2009. The lithic assemblages of et al. 2015. Early modern humans and morpho- et al. 2014. The African genome variation pro- Qafzeh Cave. Oxford: Oxford University Press. logical variation in Southeast Asia: fossil evi- ject shapes medical genetics in Africa. Nature 66 Clarkson C, Jones S, Harris C. 2012. Conti- dence from Tam Pa Ling, Laos. PLoS ONE 10: 517:327–332. nuity and change in the lithic industries of the e0121193. 49 Prufer€ K, Racimo F, Patterson N, et al. Jurreru Valley, India, before and after the Toba 30 Cann R, Stoneking M, Wilson A. 1987. Mito- 2014. The complete genome sequence of a eruption. Quat Int 258:165–179. chondrial DNA and human evolution. Nature Neandertal from the Altai Mountains. Nature 67 Blinkhorn J, Achyuthan H, Petraglia MD, 325:31–36. 505:43–49. et al. 2013. Middle Palaeolithic occupation in 31 Hammer MF. 1995. A recent common 50 Currat M, Excoffier L. 2014. Strong repro- the Thar Desert during the Upper Pleistocene: ancestry for human Y chromosome. Nature ductive isolation between humans and Nean- the signature of a modern human exit out of 378:376–378. derthals inferred from observed patterns of Africa? Quat Sci Rev 77:233–238. 32 Mountain JL, Jin AA, Bowcock AM, et al. introgression. Proc Natl Acad Sci USA 108: 68 Scerri EML, Drake NA, Jennings R, et al. 1993. Evolution of modern humans: evidence 15129–15134. 2014. Earliest evidence for the structure of from nuclear DNA polymorphisms. In: Aitken 51 Clarkson C, Smith M, Marwick B, et al. n.d. Homo sapiens populations in Africa. Quat Sci MJ, Stringer CB, Mellars PA, editors. The origin The archaeology, chronology and stratigraphy Rev 101:207–216. of modern humans and the impact of chrono- of Madjedbebe (Malakunanja II): a site in 69 Dennell RW. 2009. The Palaeolithic settle- metric dating. Princeton University Press: northern Australia with early occupation. ment of Asia. Cambridge: Cambridge University Princeton. J Hum Evol. In press. doi:10.1016/ Press. j.jhevol.2015.03.014. 33 Veeramah KR, Hammer MF. 2014. The 70 Rabett RJ. 2012. Human adaptation in the impact of whole-genome sequencing on the 52 Groucutt HS, Scerri EML, Lewis L, et al. Asian Palaeolithic: hominin dispersal and reconstruction of human population history. n.d. assemblages and models for the behaviour during the Late Quaternary. Cam- Nat Rev Genet 15:149–162. dispersal of Homo sapiens out of Africa. Quat bridge: Cambridge University Press. Int. In press. doi:10.1016/j.quaint.2015.01.039 34 Pickrell JK, Reich D. 2014. Toward a new 71 Vanhaeren M, d’Errico F, Stringer C, et al. history and geography of human genes 53 Porraz G, Texier PJ, Archer W, et al. 2013. 2006. Middle Paleolithic shell beads in Israel informed by ancient DNA. Trends Genet 30: Technological successions in the Middle Stone and . Science 312:1785–1788. Age sequence of Diepkloof , West- 377–389. 72 McBrearty S, Brooks AS. 2000. The revolu- ern Cape, South Africa. J Archaeol Sci 40:3376– 35 Seguin-Orlando A, Korneliussen TS, Sikora tion that wasn’t: a new interpretation of the ori- 3400. M. 2014. Genomic structure in Europeans dat- gin of modern human behavior. J Hum Evol ing back to least 36,000 years. Science 346: 54 Lewis L, Perera N, Petraglia M. 2014. First 39:453–563. technological comparison of Southern African 1113–1118. 73 Shea JJ. 2011. Homo sapiens is as Homo Howiesons Poort and South Asian Microlithic 36 Soares P, Alshamali F, Pereira JB, et al. sapiens was. Curr Anthropol 52:1–35. industries: an exploration of inter-regional vari- 2012. The expansion of mtDNA haplogroup L3 ability in microlithic assemblages. Quat Int 74 Powell A, Shennan S, Thomas M. 2009. Late within and out of Africa. Mol Biol Evol 29:915– 350:7–25. Pleistocene demography and the appearance of 927. modern human behavior. Science 324:1298– 55 Brandt SA, Fisher EC, Hildebrand EA, et al. 1301. 37 Pearson OM. 2013. Africa: the cradle of 2012. Early MIS 3 occupation of Mochena Bor- modern people. In: Smith FH, Ahern JCM, edi- ago rockshelter, southwest Ethiopian high- 75 Blome MA, Cohen AS, Tryon CA, et al. tors. The origins of modern humans: biology lands: implications for Late Pleistocene 2012. The environmental context for the origins reconsidered. Hoboken: John Wiley & Sons. p archaeology, paleoenvironments and modern of modern human diversity: a synthesis of 1–43. human dispersals. Quat Int 274:38–54. regional variability in African climate 150,000– 30,000 years ago. J Hum Evol 62:563–592. 38 Campbell MC, Tishkoff S. 2010. The evolu- 56 James HVA, Petraglia M. 2005. Modern tion of human genetic and phenotypic variation human origins and the evolution of behavior in 76 Drake NA, Blench MA, Armitage SJ, et al. in Africa. Curr Biol 20:R166–173. the later Pleistocene record of South Asia. Curr 2011. Ancient watercourses and 39 Schiffels S, Durbin R. 2014. Inferring Anthropol 46(suppl):S4–S27. of the Sahara explain the peopling of the desert. Proc Natl Acad Sci USA 108:458–462. human population size and separation history 57 Petraglia M, Clarkson C, Boivin N, et al. from multiple genome sequences. Nat Genet 2009. Population increase and environmental 77 Lane CS, Chorn BT, Johnson TC. 2013. Ash 46:919–925. deterioration correspond with microlithic inno- from Toba supereruption in Lake Malawi 40 Scally A, Durbin R. 2012. Revising the vations in South Asia ca. 35,000 years ago. shows no volcanic winter in East Africa at 75 human mutation rate: implications for under- Proc Natl Acad Sci USA 106:12261–12266. ka. Proc Natl Acad Sci USA 110: 8025–8029. standing human evolution. Nat Rev Genet 13: 58 Clarkson C, Petraglia MD, Korisettar R, 78 Parton A, White TS, Parker AG, et al. n.d. 745–753. et al. 2009. The oldest and longest enduring Orbital-scale climate variability in Arabia as a 41 Fu Q, Mittnik A, Johnson PLF, et al. 2013. A microlithic sequence in India: 35,000 years of potential motor for human dispersals. Quat Int. revised timescale for human evolution based on modern human occupation at the Jwalapuram In press. doi:10.1016/j.quaint.2015.01.005. ancient mitochondrial genomes. Curr Biol 23: Locality 9 rockshelter. Antiquity 83:326–348. 79 Walter RC, Buffler RT, Bruggemann JH, 553–559. 59 Perera N, Kourampas N, Simpson IA, et al. et al. 2000. Early human occupation of the Red 42 Segurel L, Wyman J, Przeworski M. 2014. 2011. People of the ancient rainforest: Late Sea coast of Eritrea during the Last Intergla- Determinants of mutation rate variables in the Pleistocene foragers at the Batadomba-lena cial. Nature 405:65–9. human germline. Ann Rev Genomics Hum rockshelter, Sri Lanka. J Hum Evol 61: 254– 80 Erlandson JM. 2001. The archaeology of Genet 15:47–70. 269. aquatic : paradigms for a new mil- 43 Fu Q, Moorjani P, Jay F, et al. 2014. 60 Blinkhorn, J. 2014. Late Middle Palaeolithic lennium. J Archaeol Res 9:287–350. Genome sequencing of a 45,000-year-old mod- surface sites occurring on dated sediment for- 81 Bailey GN, Milner NJ. 2002. Coastal hunters ern human from western Siberia. Nature 514: mations in the Thar Desert. Quat Int 350:95– and gatherers and social evolution: marginal or 445–449. 104. central? Before Farming 3-4:1–15. 164 Groucutt et al. ARTICLE

82 Bailey GN, Flemming N. 2008. Archaeology 90 Kusky T, Robinson C, El-Baz F. 2005. Terti- elevation data. EOS, Trans Am Geophys Union of the continental shelf: marine resources, sub- ary-Quaternary faulting and uplift in the north- 89:93–94. merged landscapes and underwater archaeol- ern Oman Hajar mountains. J Geol Soc 162: 99 Breeze P, Drake NA, Groucutt HS, et al. n.d. ogy. Quat Sci Rev 27: 2153–2165. 871–888. Remote sensing and GIS techniques for recon- 83 Marean CW. 2014. The origins and signifi- 91 Bailey GN, Deve`s M, Inglis RH, et al. n.d. structing Arabian palaeohydrology and identify- cance of coastal resource use in Africa and Blue Arabia: Palaeolithic and underwater sur- ing archaeological sites. Quat Int. In press. Western Eurasia. J Hum Evol 77:17–40. vey in SW Saudi Arabia and the role of coasts doi:10.1016/j.quaint.2015.01.022. 84 Bruggemann JH, Buffler RT, Guillaume in Pleistocene dispersal. Quat Int. In press. 100 Hijmans RJ, Cameron SE, Parra JL, et al. MMM, et al. 2004. Stratigraphy, palaeoenviron- doi:10.1016/j.quaint.2015.01.002. 2005. Very high resolution interpolated climate ments and model for the deposition of the 92 Aubert M, Brumm A, Ramli M, et al. 2014. surfaces for global land areas. Int J Climatol Abdur Reef limestone: context for an important Pleistocene cave art from Sulawesi, Indonesia. 25:1965–1978. archaeological site from the last interglacial on Nature 514:223–227. 101 Jennings RP, Singarayer J, Stone E, et al. the Red Sea. Palaeogeogr, Palaeoclimatol, 93 Behar DM, van Oven M, Rosset S, et al. n.d. The greening of Arabia: an ensemble of cli- Palaeoecol 20:179–206. 2012. A “Copernican” reassessment of the mate model simulations infers multiple oppor- 85 Stringer CB, Finlayson JC, Barton RNE, human mitochondrial DNA tree from its root. tunities for human occupation of the Arabian et al. 2008. Neanderthal exploitation of marine Am J Hum Genet 90:675–684. Peninsula during the Late Pleistocene. Quat mammals in Gibraltar. Proc Natl Acad Sci USA 94 Gronau I, Hubisz MJ, Gulko B, et al. 2011. Int. In press. doi:10.1016/j.quaint.2015.01.006. 108:14319–14324. Bayesian inference of ancient human demogra- 102 Drake NA, Blench MA, Armitage SJ, et al. 86 Irwin G. 1992. The Prehistoric exploration phy from individual genome sequences. Nat 2011. Ancient watercourses and biogeography and colonization of the Pacific. Cambridge: Genet 43:1031–1034. of the Sahara explain the peopling of the Cambridge University Press. 95 Hudson RR. 2002. Generating samples desert. Proc Natl Acad Sci USA 108:458–462. 87 O’Connor S, Ono R, Clarkson C. 2011. under a Wright-Fisher neutral model of genetic 103 Ambrose SH. 1998. Chronology of the Pelagic fishing at 42,000 years before the pres- variation. Bioinformatics 18:337–338. Later Stone Age and food production in East ent and the maritime skills of modern humans. 96 Fenner JN. 2005. Cross-cultural estimation Africa. J Archaeol Sci 25:377–392. Science 334:1117–1121. of the human generation interval for use in 104 Bretzke K, Conard NJ, Uerpmann H-P. 88 O’Connell JF, Allen J, Hawkes K. 2010. genetics-based population divergence studies. 2014. Excavations at Jebel Faya: the FAY-NE1 Pleistocene Sahul and the origins of seafaring. Am J Phys Anthropol 128:415–423. shelter sequence. Proc Sem Arabian Stud 44: 69–82. In: Anderson A et al., editors. The global origins 97 Crassard R, Petraglia MD, Drake NA, et al. and development of seafaring. Cambridge: 2013. Middle Palaeolithic and Neolithic occupa- 105 Nishiaki Y, Kanjo Y, Muhesen S, et al. McDonald Institute. p 57–68. tions around Mundafan palaeolake, Saudi Ara- 2012. The temporal variability of Late Levan- tine Mousterian lithic assemblages from Deder- 89 Bailey GN, King GCP, Deve`s M, et al. 2012. bia: implications for climate change and iyeh Cave, Syria. Eurasian Prehist 9:3–27. DISPERSE: dynamic landscapes, coastal envi- human dispersals. PLoS ONE 8:e69665. ronments and human dispersals. Antiquity 86: 98 Lehner B, Verdin K, Jarvis A. 2008. New 334. http://antiquity.ac.uk/projgall/bailey334. global hydrography derived from spaceborne VC 2015 Wiley Periodicals, Inc.