Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Multiple hominin dispersals into Southwest Asia over the past 400,000 years


An Author Correction to this article was published on 10 January 2022

This article has been updated


Pleistocene hominin dispersals out of, and back into, Africa necessarily involved traversing the diverse and often challenging environments of Southwest Asia1,2,3,4. Archaeological and palaeontological records from the Levantine woodland zone document major biological and cultural shifts, such as alternating occupations by Homo sapiens and Neanderthals. However, Late Quaternary cultural, biological and environmental records from the vast arid zone that constitutes most of Southwest Asia remain scarce, limiting regional-scale insights into changes in hominin demography and behaviour1,2,5. Here we report a series of dated palaeolake sequences, associated with stone tool assemblages and vertebrate fossils, from the Khall Amayshan 4 and Jubbah basins in the Nefud Desert. These findings, including the oldest dated hominin occupations in Arabia, reveal at least five hominin expansions into the Arabian interior, coinciding with brief ‘green’ windows of reduced aridity approximately 400, 300, 200, 130–75 and 55 thousand years ago. Each occupation phase is characterized by a distinct form of material culture, indicating colonization by diverse hominin groups, and a lack of long-term Southwest Asian population continuity. Within a general pattern of African and Eurasian hominin groups being separated by Pleistocene Saharo-Arabian aridity, our findings reveal the tempo and character of climatically modulated windows for dispersal and admixture.


As the only land bridge between Africa and Eurasia, Southwest Asia occupies a unique position for understanding key stages of human evolution and the peopling of the planet. Changing environmental and ecological conditions at the shifting interface between the Saharo-Arabian and Palaearctic biomes strongly influenced patterns of human demography through the isolation, diversification and subsequent mixing of populations4,6,7,8,9,10,11 (Supplementary Information, section 1). A prominent example concerns the geographical context of Neanderthal–sapiens admixture. Although it has been suggested that this occurred in Southwest Asia owing to the ubiquity of Neanderthal ancestry in humans outside Africa6, ‘on-the-ground’ evidence for admixture, or even spatial and temporal contemporaneity with H. sapiens, has remained elusive in the region. One reason for this is the severely fragmented nature of Southwest Asian palaeontological, palaeoenvironmental and archaeological records. This has in turn limited our ability to overcome problematic generalizations regarding the palaeoanthropological record of Southwest Asia and address key questions about the extent to which hominin occupations of the region were continuous, the role of hominin dispersals into and within the region, and how these dispersals and interactions between hominin populations related to changes in biogeography, environment and ecology.

Research in Southwest Asia has traditionally focussed on deeply stratified cave sequences in the Levantine winter-rainfall woodland zone11,12,13,14,15 (Fig. 1, Supplementary Information, section 1). This has led to a detailed record for the woodlands, a southern extension of the Palaearctic biome4,16. However, in the past decade, research in the Arabian Peninsula has begun to document hominin occupations of the arid Saharo-Arabian biome during episodically wetter periods characterized by grasslands, lakes and rivers5,10,17,18,19,20,21,22,23,24,25,26,27,28,29. Emerging patterns of spatially divergent cultural-evolutionary developments in Southwest Asia include a young (less than 200 thousand years ago (ka)) Acheulean in central Arabia24, a technology typically associated with earlier hominins such as Homo erectus. There are also repeated manifestations of distinctive local characteristics, commonly interpreted as autochthonous developments, in ‘refugial’ areas in southern Arabia22,23.

Fig. 1: Palaeoanthropological sites in Southwest Asia dating to between 50 and 500 ka.
figure 1

a, Sites plotted on a map of modern biomes. b, Sites plotted on map of modern precipitation. c, Sites plotted on model of MIS 5e precipitation34, as an illustration of the changes that occurred during humid periods.

Despite these advances, the few reported sites in interior and northern Arabia20,24,25,26,27,28,29 (Supplementary Information, section 1) have small sample sizes of artefacts, and are often raw material procurement and workshop sites, with a very different character to the cave and rockshelter ‘living sites’ that dominate the Levantine woodland record. The absence of permanent fluvial systems and deeply stratified cave sequences in Arabia has hampered the construction of long-timescale archaeological and hydroclimatic sequences. This has limited efforts to recognize important patterns in archaeological and palaeontological records associated with changes in hominin distribution, demography and behaviour.

Here we report multiple palaeolake sedimentary sequences with associated lithic (stone tool) assemblages and fossil fauna in the Nefud Desert of northern Arabia, representing the first detailed long-timescale record of hominin occupations in Arabia (Figs. 13). Khall Amayshan 4 (KAM 4) consists of a series of superimposed lake sequences within a single interdunal basin (Extended Data Figs. 1, 2, Supplementary Information, sections 2, 3). This site, currently unique in Arabia, has preserved a record analogous to the detailed fluvial archives preserved in regions such as northwest Europe. Additionally, we present further evidence for multiple hominin occupations from excavated sites dating to Marine Isotope Stage (MIS) 7 and MIS 5 from the nearby Jubbah palaeolake basin. Together the KAM 4 and Jubbah assemblages show that there were multiple hominin dispersals into Arabia over the last 400,000 years, in association with a unique hydroclimate record.

Fig. 2: The chronology and environmental context of hominin occupations in northern Arabia.
figure 2

a, Al Marrat 327. b, Jebel-Qattar 128. c, Al Wusta20. d, JSM 1 (present study). e, Central (C), Northeast (NE), Northwest (NW), Southwest (SW), Southeast (SE) and South (S) lakes at KAM 4. US, an age for sands underlying the lake (that is, a maximum age for overlying phase of lake formation); IL, in lake (direct date on sediments within lake-related deposits). Black arrows pointing to the left (Southeast and South Lakes) reflect that the luminescence ages provide maximum ages, and the overlying lakes are younger. Filled symbols show quartz ages and open symbols show feldspar ages. f, East Mediterranean sapropel record35, insolation36 (grey), monsoon index37 (black) and oxygen isotope record38 (blue). Southern Arabian humid periods are defined by speleothems in green21. Luminescence ages are presented with 1σ uncertainties and the single U-series age is presented with a 2σ uncertainty.

Fig. 3: Stone tools from KAM 4 and Jebel Umm Sanman 1 (JSM 1).
figure 3

From left to right: assemblage A, KAM 4 (approximately 400 ka), assemblage B, KAM 4 (approximately 300 ka), assemblage C, KAM 4 (approximately 200 ka), JSM 1 (approximately 75 ka), assemblage E, KAM 4 (approximately 55 ka?). Scale bar, 1 cm.

Each KAM 4 palaeolake deposit is stratigraphically similar, being predominantly composed of massive or finely-laminated carbonate rich marls overlying sands (Extended Data Figure 3, Supplementary Information, section 3). The similarity of these marls, each formed by a discrete lake phase, implies that the palaeoenvironment of KAM 4 was broadly similar during successive humid phases. The sediments are comparable to other palaeolake deposits from the western Nefud Desert18,20, but are notable for their stratigraphically distinct and superimposed character, and abundance of associated lithics and fossils. The sediments at the site are fine-grained (sands, silts and marls), reflecting deposition under low-energy or still-water conditions. Larger clasts (gravels) are absent, emphasizing the lack of higher-energy current flow processes feeding the basin during sediment accumulation. Reworking of lithics and fossils from the surrounding landscape into these lake bodies is therefore highly unlikely. Consequently, we argue that the assemblages of lithics and the fossils found in association with these deposits are in situ, which was confirmed by excavations in the case of the Northwest Lake.

The unique KAM 4 record has survived owing to migrating sand dunes that moved in a conveyor-belt-like fashion across the basin, protecting older parts of the sequence from erosion and preventing the mixing of the distinct archaeological and palaeontological assemblages associated with each lake phase. KAM 4 provides the first long-term composite sequence for the later Middle Pleistocene and Late Pleistocene in Arabia, with each phase of hominin occupation associated with a broadly similar environment and lithic raw material availability.

The oldest deposit at KAM 4, the Central Lake, is dated by luminescence to 412 ± 87 ka (Fig. 2, Extended Data Figs. 2, 3, Supplementary Information, section 5). The Central Lake deposits are also heavily iron-stained compared with other deposits at the site, attesting to their greater antiquity within the basin. The Central Lake is stratigraphically overlain by the southernmost edge of the Northeast Lake, which yielded luminescence ages of 337 ± 39 ka and 306 ± 47 ka. Although the age estimates for both the Central and Northeast Lakes have large uncertainties, we emphasize evidence for regional aridity in the millennia either side of both MIS 11 and 9 (Fig. 2), making attribution of Central Lake to MIS 11 and Northeast Lake to MIS 9 parsimonious. The Northwest Lake, which partly overlies the Northeast Lake, is dated by a suite of luminescence estimates on carbonate-rich sands bracketed between two phases of marl deposits to between 192 ± 20 ka and 210 ± 16 ka. A direct U-series age estimate of a bovid fossil from the same layer produced a consistent date of 205 ± 2 ka (2σ) (Supplementary Information, section 6). The Northwest Lake can therefore be correlated with MIS 7, the final humid phase of the Middle Pleistocene. The Southwest Lake has a luminescence estimate of 143 ± 10 ka, and therefore dates either to late MIS 6 or, less probably, to the transition to MIS 5. Sands underlying the Southeast Lake, which overlies the Central and Northeast Lakes, yielded luminescence estimates of 159 ± 11 ka to 149 ± 9 ka, similar to the South Lake at 168 ka ± 12 ka to 142 ± 13 ka. These maximum age estimates derived from underlying sand indicate that the Southeast and South Lakes date to either the final millennia of the Middle Pleistocene or, more probably, to the subsequent Late Pleistocene. As discussed below, we hypothesize on archaeological grounds that the South Lake dates to MIS 5 and the Southeast Lake to early MIS 3.

The Jubbah record, consisting of excavated stratified lithics, enables us to further extend the occupation sequence of the region. Substantially enlarged excavations at Jebel Qattar 1 (JQ 1) increased the sample size of lithics dating to 211 ± 16 ka reported by ref. 28 by 250%. At Jebel Umm Sanman 1 (JSM 1), four new trenches were placed immediately west of earlier test excavations28. The JSM 1 trenches revealed deep (more than 1.5–2.5 m) stratigraphic sequences, comprising a series of silty sands with variable frequencies of local gravel clasts. Luminescence dating indicates that the lower part of the JSM 1 sequence dates to 130 ± 10 ka, whereas the upper part, in which lithics were found, dates to approximately 75 ka (77 ± 7 ka, 72 ± 6.4 ka) (Supplementary Information, section 5).

Each phase of lake formation (apart from the Southwest Lake) at KAM 4 is associated with a distinct lithic assemblage (Fig. 3, Extended Data Figs. 48, Supplementary Information, section 7). Assemblage A (Central Lake, approximately 400 ka) consists of handaxes and associated debitage (Extended Data Fig. 4) and is the oldest dated Acheulean assemblage in Arabia. It shows the production of small and refined handaxes produced by shaping (façonnage) of angular slabs of quartzite. Assemblage B (Northeast Lake, approximately 300 ka) is also characterized by the production of small handaxes (Extended Data Figs. 5, 6). These handaxes are rather homogeneous in their technology and morphology, being small and pointed. Core reduction technology to produce flakes is also present in low frequencies in assemblage B, mostly characterized by preferential Levallois reduction. The subsequent assemblage C (Northwest Lake, approximately 200 ka) shows a Middle Palaeolithic technology. Lithics recovered from the surface and from excavations show a complete absence of handaxe manufacture, and a focus on Levallois technology, often centripetal in character (Extended Data Fig. 7), but somewhat diverse (Supplementary Information, section 7). Assemblage D (Southeast Lake, approximately 125–75 ka) and assemblage E (South Lake, approximately 55 ka) are both of Middle Palaeolithic character—assemblage D has a focus on centripetal Levallois technology and assemblage E has a somewhat diverse technology, but with a strong component of unidirectional-convergent preparation to produce convergent Levallois flakes.

With the enlarged excavations at JQ 1, the assemblage dating to approximately 210 ka has a clearly Middle Palaeolithic character; Levallois flakes are present, and bifacial technology is absent (Supplementary Information, section 8). The JSM 1 assemblage from approximately 75 ka is the largest excavated and dated Palaeolithic assemblage from northern Arabia, and shows a clear focus on centripetal Levallois reduction, with 83% of Levallois flakes having centripetal scar patterns (Extended Data Fig. 9).

This unique record of hydroclimate and associated hominin occupations demonstrates that Acheulean Lower Palaeolithic technology was present during late Middle Pleistocene wet phases, with Levallois technology being present in the final stage of the Acheulean. Assemblages showing similarities to the Acheulo-Yabrudian of the Levantine woodlands have not been identified in Arabia, highlighting distinct trajectories within Southwest Asia. From MIS 7, Arabian Middle Palaeolithic assemblages appear with each phase of increased precipitation, showing varying technological foci in terms of the reduction methods used, from varied Levallois in MIS 7, to centripetal Levallois in MIS 53,20, and unidirectional-convergent in MIS 323,27.

KAM 4 assemblage C has technological characteristics—such as frequent centripetal Levallois flaking—that are more similar to those of the East African Middle Stone Age (MSA) than to the contemporaneous Levantine early Middle Palaeolithic (Supplementary Information, sections 7, 8). Principal components analysis (PCA) of late Middle Pleistocene Levallois flakes shows that KAM 4 assemblage C falls between Omo Kibish-AHS in East Africa30 and Misliya in the Levant31. For the Late Pleistocene, PCA distinguished between H. sapiens-associated assemblages such as Omo Kibish-BNS30 and Al Wusta in Arabia20, and the Neanderthal-associated Levantine assemblages from Kebara15 and Tor Faraj32. KAM 4 assemblage D and JSM 1 orientate to the MIS 5 H. sapiens assemblages, whereas the negative score on the second principal component for KAM 4 assemblage E orientates it to MIS 4–3 Levantine Neanderthal assemblages.

Animal fossils (primarily vertebrates) from KAM 4 allow us to reconstruct the palaeoenvironmental and biogeographical context of hominin occupations. Hippopotamus fossils had previously been reported in Arabia from MIS 5 contexts (fer example, in refs. 20,33). KAM 4 shows that hippopotamuses were also present during MIS 7 and, provisionally, MIS 9 (Supplementary Information, section 10). We also identified hippopotamus in the surface scatter of fossils at the nearby site of Ti’s al Ghadah. The repeated presence of hippopotamuses, which are obligate semi-aquatic mammals that require permanent water bodies several metres deep, provides powerful evidence for the extent of environmental amelioration during repeated ‘green Arabia’ pluvial phases. In addition, the KAM 4 palaeontological assemblages contribute to a growing corpus of evidence indicating that Arabian mammal fauna had stronger affinity with Africa in the Middle and Late Pleistocene than with the Levantine woodland zone4,33. The presence of African bovid taxa such as Syncerus and Hippotragus in northern Arabia indicates the repeated establishment of contiguous regions of grasslands across North Africa and Arabia with abundant freshwater, providing dispersal routes for a variety of species, including hominins. Arabia, however, also features Eurasian and endemic taxa (Supplementary Information 10), indicating that it was a key biogeographical nexus between Africa and the rest of Eurasia that may have also comprised an important interaction zone for hominins.

The northern Arabian late Middle Pleistocene lithic assemblages likewise show greater similarities to African assemblages than to those of Levantine woodland zone sites. The continued production of large handaxes and cleavers in central Arabia at the time the Middle Palaeolithic had appeared in northern Arabia24 indicates high levels of population structure at this time, perhaps to the extent of different hominin species occupying the region. In MIS 5, it seems that much of Northeast Africa and Southwest Asia shared similar material culture, consistent with widespread dispersals of H. sapiens20. Subsequently the cooling and aridification of the last glacial cycle led to the fracturing and decline of populations. Renewed dispersals, perhaps including Neanderthals from the north, occurred during the partial amelioration of early MIS 3 (around 59–50 ka). Comparatively stable environmental and ecological conditions in areas such as the Levantine woodland fostered the development of distinctive localized material culture phases11. By contrast, the record of interior northern Arabia indicates pulses of occupation during episodic phases of increased environmental humidity, seemingly followed by repeated regional depopulation under increasing aridity.

We have identified at least five pulses of human dispersal into northern Arabia, each associated with a phase of decreased aridity. The differences in material culture between these phases—with two phases of Acheulean technology and then three distinct forms of Middle Palaeolithic—suggests that diverse hominin populations, and probably even species, were expanding into the region at different times (we discuss the implications of our findings further in Supplementary Information, section 11). The emerging palaeoanthropological record of Arabia highlights the  dynamism and regional distinctiveness of Middle and Late Pleistocene hominin demography and behaviour in different parts of Southwest Asia. These processes were intimately connected to regional climatic changes. The available record emphasizes pulsed, long-ranging terrestrial dispersals followed by local variation, and finally population contraction. Given the temporal overlap of radically different technologies within Arabia, and the biogeographical evidence for faunal mixture, it is possible that some of the hominin admixture processes identified by genetic analyses occurred in this region. Arabia, and Southwest Asia more generally, is therefore a key region for unravelling not only the increasingly complex history of how our species spread beyond Africa, but more broadly, how our species’ recent success relates to a longer history of hominin dispersals, regional developments and admixture, which occurred in a context of marked environmental oscillation.


Site identification and survey

KAM 4 was initially identified through remote sensing analysis26 (Supplementary Information, section 2). Two main seasons of research were conducted at the site (2014 and 2017) as part of the Palaeodeserts/Green Arabia Project. The site was systematically surveyed with pedestrian transects. Using a total station and Trimble XRS Pro Differential Global Positioning System, the topography of the site was recorded in detail, and all points of interest (stone tools, fossils and sedimentary features) were recorded and entered into a geographic information system. JSM 1 and JQ 1 were first identified in 201128. We carried out renewed excavation of the sites in 2013. With JQ 1, the stratigraphic sequence was already understood, so the aim was simply to increase the sample size of lithics. At JSM 1, the original excavations had rapidly hit bedrock, so renewed excavations were conducted slightly further west in the hope of identifying deeper stratigraphic sequences.

Stratigraphy and sedimentology

Sections for sedimentary analysis and luminescence sampling were excavated for each of the palaeolake phases at KAM 4 (Supplementary Information, section 3). The Northwest Lake was identified as having the best potential to recover buried material, as fossils and lithics appeared to be emerging from sediments, so four trenches (1–4) were dug here. These trenches and the excavations at JSM 1 and JQ 1 (Supplementary Information, section 4) were conducted using single-context excavation methods. All sediments were dry sieved through 5-mm mesh. The focus of this paper is on the archaeological assemblages, and not detailed palaeoenvironmental analysis so our sedimentary description consists of field observations from logging sections. Fossils from KAM 4 have previously been reported33.

Chronometric dating methods

We used luminescence (OSL on quartz and pIRIR on feldspar) methods to date the sedimentary deposits at KAM 4 and at JSM 1 (Supplementary Information, section 5). These measure the time since sediments were last exposed to sunlight. Opaque metal tubes were hammered into cleaned sections, transported to the UK and analysed as described in Supplementary Information, section 5. Environmental dose rates were calculated using location and overburden density (cosmic rays), field gamma spectrometry (gamma), and thick-source beta counting (beta). A bovid tooth (KAM16/85) was recovered from unit 3 of the Northwest Lake at KAM 4. A direct age was obtained using the U-series dating method, which dates the moment uranium is incorporated into the fossil. Powdered samples of both enamel and dentine tissues were drilled from the tooth at Griffith University, and U-series analyses were subsequently carried out at the University of Queensland. While it was initially planned to combine with electron spin resonance analyses, the U-series results obtained showed that the tooth was not suitable for that purpose (Supplementary Information, section 6).

Lithic analysis

Lithics (stone tools) from the excavations at all sites and from the systematic transect survey at KAM 4 were studied using the methodology described previously in refs. 9,20 and in Supplementary Information, sections 7, 8. Our initial focus was on describing the basic typo-technological features of the assemblages. We selected illustrative examples for photography, 3D scanning, and illustration. For the Middle Palaeolithic samples, we carried out a full metric and attribute analysis following the above references and references therein. As well as allowing the description of the assemblages in quantitative terms, we focussed on the characteristics of Levallois flakes from these assemblages as a way to compare assemblages. We did this both in terms of univariate features (such as dorsal scar patterns), and using PCA to compare the morphology of Levallois flakes between the assemblages (Supplementary Information, section 8).

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this paper.

Data and code availability

Data for the PCA analysis are archived at All other relevant data are included in the paper and Supplementary Information.

Code availability

Code for the PCA analysis are archived at

Change history


  1. Dennell, R. The Palaeolithic Settlement of Asia (CUP, 2008).

  2. Dennell, R. From Arabia to the Pacific: How our species colonised Asia (Routledge, 2020).

  3. Groucutt, H. S. et al. Rethinking the dispersal of Homo sapiens out of Africa. Evol. Anthropol. 24, 149–164 (2015).

    Article  Google Scholar 

  4. Stewart, M. et al. Middle and Late Pleistocene mammal fossil records of Arabia and surrounding regions: Implications for biogeography and hominin dispersals. Quatern. Int. 515, 12–29 (2019).

    Article  Google Scholar 

  5. Petraglia, M. D., Groucutt, H. S., Parton, A. & Alsharekh, A. Green Arabia: human prehistory at the cross-roads of continents. Quatern. Int. 382, 1–7 (2015).

    Article  Google Scholar 

  6. Green, R. E. et al. A draft sequence of the Neanderthal genome. Science 328, 710–722 (2010).

    ADS  CAS  Article  Google Scholar 

  7. Drake, N. A., Blench, R. M., Armitage, S. J., Bristow, C. S., White, K. H. Ancient watercourses and biogeography of the Sahara explain the peopling of the desert. Proc. Natl Acad. Sci. USA 108, 458–462 (2011).

    ADS  CAS  Article  Google Scholar 

  8. Blome, M. W., Cohen, A. S., Tryon, C. A., Brooks, A. S. & Russell, J. The environmental context for the origins of modern human diversity: a synthesis of regional variability in African climate 150,000–30,000 years ago. J. Hum. Evol. 62, 563–592 (2012).

    Article  Google Scholar 

  9. Scerri, E. M. L., Drake, N. A., Jennings, R. & Groucutt, H. S. Earliest evidence for the structure of Homo sapiens populations in Africa. Quatern. Sci. Rev. 101, 207–2016 (2014).

    ADS  Article  Google Scholar 

  10. Breeze, P. S. et al. Palaeohydrological corridors for hominin dispersals in the Middle East ~250–70,000 years ago. Quatern. Sci. Rev. 11, 155–185 (2016).

    ADS  Article  Google Scholar 

  11. Malinsky-Buller, A. & Hovers, E. One size does not fit all: group size and the late middle Pleistocene prehistoric archive. J. Hum. Evol. 127, 118–132 (2019).

    Article  Google Scholar 

  12. Shea, J. J. The Middle Paleolithic of the East Mediterranean Levant. J. World Prehist. 17, 313–394 (2003).

    Article  Google Scholar 

  13. Hovers, E. The Lithic Assemblages of Qafzeh Cave (Oxford Univ. Press, 2009).

  14. Hovers, E. & Belfer-Cohen, A. On variability and complexity: lessons from the Levantine Middle Paleolithic record. Curr. Anthropol. 54, S337–S357 (2013).

    Article  Google Scholar 

  15. Meignen, L. & Bar-Yosef, O. Kebara Cave, Mt. Carmel, Israel, Part II: The Middle and Upper Palaeolithic Archaeology (Peabody Museum Press, 2020).

  16. Holt, B. G. et al. An update of Wallace’s zoogeographic regions of the world. Science 339, 74–78 (2012).

    ADS  Article  Google Scholar 

  17. Groucutt, H. S. & Petraglia, M. D. The prehistory of Arabia: deserts, dispersals and demography. Evol. Anthropol. 21, 113–125 (2012).

    Article  Google Scholar 

  18. Rosenberg, T. M. et al. Middle and Late Pleistocene humid periods recorded in palaeolake deposits of the Nafud desert, Saudi Arabia. Quatern. Sci. Rev. 70, 109–123 (2013).

    ADS  Article  Google Scholar 

  19. Parton, A. et al. Middle-Late Quaternary palaeoclimate variability from lacustrine deposits in the Nefud Desert, Northern Arabia. Quatern. Sci. Rev. 202, 78–97 (2018).

    ADS  Article  Google Scholar 

  20. Groucutt, H. S. et al. Homo sapiens in Arabia by 85,000 years ago. Nature Eco. Evo. 2, 800–809 (2018).

    Article  Google Scholar 

  21. Nicholson, S. et al. Pluvial periods in southern Arabia over the last 1.1 million-years. Quatern. Sci. Rev. 229, 106112 (2020).

    Article  Google Scholar 

  22. Armitage, S. J. et al. The southern route “Out of Africa”: evidence for an early expansion of modern humans into Arabia. Science 331, 453–456 (2011).

    ADS  CAS  Article  Google Scholar 

  23. Delagnes, A. et al. Inland human settlement in southern Arabia 55,000 years ago. New evidence from the Wadi Surdud Middle Paleolithic site complex, western Yemen. J. Hum. Evol. 63, 452–474 (2012).

    Article  Google Scholar 

  24. Scerri, E. M. L. et al. The expansion of later Acheulean hominins into the Arabian Peninsula. Sci. Rep. 8, 17165 (2018).

    ADS  Article  Google Scholar 

  25. Scerri, E. M. L. et al. Middle to Late Pleistocene human habitation in the western Nefud Desert. Quatern. Int. 382, 200–214 (2015).

    Article  Google Scholar 

  26. Breeze, P. S. et al. Prehistory and palaeoenvironments of the Western Nefud, Saudi Arabia. Arch. Res. Asia 10, 1–16 (2017).

    Google Scholar 

  27. Jennings, R. P. et al. Human occupation of the northern Arabia interior during early Marine Isotope State 3. J. Quat. Sci. 31, 953–966 (2016).

    Article  Google Scholar 

  28. Petraglia, M. D. et al. Hominin dispersal into the Nefud Desert and Middle Palaeolithic settlement along the Jubbah palaeolake, Northern Arabia. PLoS ONE 7, e49840 (2012).

    ADS  CAS  Article  Google Scholar 

  29. Scerri, E. M. L., et al. The expansion of Acheulean hominins into the Nefud Desert of Arabia. Sci. Rep. 11, 10111 (2021).

    ADS  CAS  Article  Google Scholar 

  30. Shea, J. J. The Middle Stone Age archaeology of the Lower Omo Valley Kibish formation: excavations, lithic assemblages, and inferred patterns of early Homo sapiens behaviour. J. Hum. Evol. 55, 448–485 (2008).

    Article  Google Scholar 

  31. Zaidner, Y. & Weinstein-Evron, M. The emergence of the Levallois technology in the Levant: a view from the early Middle Palaeolithic site of Misliya Cave, Israel. J. Hum. Evol. 144, 102785 (2020).

    Article  Google Scholar 

  32. Henry, D. O. Neanderthals in the Levant. Behavioral Organization and the Beginning of Human Modernity (Continuum, 2003).

  33. Stewart, M. et al. A taxonomic and taphonomic study of Pleistocene fossil deposits from the western Nefud Desert, Saudi Arabia. Quatern. Res. 95, 1–22 (2020).

    ADS  CAS  Article  Google Scholar 

  34. Otto-Bliesner, B. L. et al. Simulating Arctic climate warmth and icefield retreat in the Last Interglaciation. Science 311, 1751–1753 (2006).

    ADS  CAS  Article  Google Scholar 

  35. Zhao, Y. et al. Variations of the Nile suspended discharges during the last 1.75 Myr. Palaeogeog. Palaeoclim. Palaeoecol. 311, 230–241 (2011).

    ADS  Article  Google Scholar 

  36. Berger, A. & Loutre, M. F. Insolation values for the climate of the last 10 million years. Quatern. Sci. Rev. 10, 297–317 (1991).

    ADS  Article  Google Scholar 

  37. Grant, K. M. et al. A 3 million year index for North African humidity/aridity and the implication of potential pan-African humid periods. Quatern. Sci. Rev. 171, 100–118 (2017).

    ADS  Article  Google Scholar 

  38. Lisiecki, L. E. & Raymo, M. E. A Pliocene–Pleistocene stack of 57 globally distributed benthic δ18O records. Paleoceanography 20, 1–17 (2005).

    Google Scholar 

Download references


We thank the Heritage Commission, Ministry of Culture, Saudi Arabia for fieldwork support and permission to conduct this research. The research was funded by the Max Planck Society, the European Research Council (295719 to M.D.P.), the British Academy (H.S.G.), the Leverhulme Trust (ECF-2019-538 to S.B., ECF-2019-538 to P.S.B. and PG-2017-087 to S.B., E.A., S.J.A. and M.D.P.), the Research Council of Norway, through its Centres of Excellence funding scheme, SFF Centre for Early Sapiens Behaviour (SapienCE), project number 262618 (S.J.A.), the Natural Environmental Research Council (NERC) through its London DTP studentship funding (R.C.-W.), The Nature and Science Researchers Supporting Project (NSRSP-2021-5), DSFP, King Saud University, Riyadh, Saudi Arabia (A.M.A.), the Leakey Foundation (M.S.) the Australian Research Council (FT160100450 to J.L. and FT150100215 to M.D.), and the Spanish Ramón y Cajal Fellowship (RYC2018-025221-I to M.D.). We thank L. Clark-Balzan for assistance with the luminescence dating, I. Cartwright for lithic photography and M. O’Reilly for assistance with figures. We thank the museums listed in the Supplementary Information for access to comparative collections.


Open access funding provided by Max Planck Society.

Author information

Authors and Affiliations



Conceptualization: H.S.G., A.M.A. and M.D.P. Data collection: all authors. Luminescence dating: E.A. and S.J.A. U-series dating: G.J.P. and M.D. Palaeoenvironmental analysis: T.S.W., E.A., R.C.-W., P.S.B., S.J.A., N.D., J.L., A.P. and I.C. Lithic analysis: H.S.G., E.M.L.S. and J.B. PCA: W.C.C. Writing: all authors.

Corresponding authors

Correspondence to Huw S. Groucutt or Michael D. Petraglia.

Ethics declarations

Competing interests

The authors declare no competing interests.

Additional information

Peer review information Nature thanks Robin Dennell, Peter deMenocal and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data figures and tables

Extended Data Fig. 1 KAM 4 looking south, team members are visible on the Northwest Lake


Extended Data Fig. 2 Plan of Khall-Amayshan-4, showing multiple phases of lake formation, associated with lithic and fossil assemblages


Extended Data Fig. 3 Stratigraphic logs and chronometric dates for KAM 4, JSM 1 and JQ 1


Extended Data Fig. 4 Assemblage A handaxes from Central Lake (MIS 11), KAM 4.

Scale: 1 cm.

Extended Data Fig. 5 Assemblage B handaxes from Northeast Lake (MIS 9), KAM 4.

Scale: 1 cm.

Extended Data Fig. 6 3D scans of assemblage B (MIS 9) handaxes from KAM 4


Extended Data Fig. 7 Middle Palaeolithic artefacts from KAM 4 and JSM 1.

Top left: KAM 4 assemblage C (MIS 7) Levallois flake, top right: JSM 1 Levallois flake (MIS 5), bottom row: Levallois points from assemblage E (MIS 3?), KAM 4. Scale: 1 cm.

Extended Data Fig. 8 Levallois flakes from assemblage C, KAM 4 (MIS 7).

Scale: 1 cm.

Extended Data Fig. 9 Levallois flakes from JSM 1 (MIS 5, ca. 75 ka).

Scale: 1 cm.

Supplementary information

Supplementary Information

This file contains Supplementary Methods, Discussion, Tables 1–26, Figs. 1–39 and references

Reporting Summary

Peer Review File

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Groucutt, H.S., White, T.S., Scerri, E.M.L. et al. Multiple hominin dispersals into Southwest Asia over the past 400,000 years. Nature 597, 376–380 (2021).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing