The prehistory of the Arabian peninsula: Deserts, dispersals, and

The Prehistory of the Arabian Peninsula: Deserts,Dispersals, and DemographyHUW S. GROUCUTT AND MICHAEL D. PETRAGLIA

As a geographic connection between Africa and the rest of Eurasia, the Ara-bian Peninsula occupies a central position in elucidating hominin evolution anddispersals. Arabia has been characterized by extreme environmental fluctuationin the Quaternary, with profound evolutionary and demographic consequences.Despite the importance of the region, Arabia remains understudied. Recentyears, however, have seen major developments in environmental studies and ar-cheology, revealing that the region contains important records that should play asignificant role in future paleoanthropological narratives.1–3 The emerging pictureof Arabia suggests that numerous dispersals of hominin populations into theregion occurred. Populations subsequently followed autochthonous trajectories,creating a distinctive regional archeological record. Debates continue on the re-spective roles of regional hominin extinctions and population continuity, with thelatter suggesting adaptation to arid conditions.

There exists a striking imbalancebetween the clear importance of Ara-bia and what is actually knownabout the region. The incorrect per-ception of Arabia as an unchanging

desert has certainly played a role inthe lack of serious research. In somecases, cultural and political issueshave hampered research, as, forexample, in contemporary Yemen.Numerous archeological surveyshave been conducted but, untilrecently, these have been unsyste-matic. Most prehistoric sites consistof lithic scatters from surface con-texts. As a result, knowledge of Ara-bian prehistory often comes fromsites potentially, but not necessarily,representing multiple phases of occu-pation, which lack absolute datesand paleoenvironmental information.Attribution to cultural phases hastypically reflected typological analy-ses of biased collections. The absenceof a pre-Holocene hominin fossil re-cord in Arabia precludes definitiveidentification of the manufacturersof the widespread lithic industries.These problems must be recognizedand used to guide future research.Nevertheless, the Arabian recordconstitutes an important and under-studied dataset. The recent discoveryof stratified archeological sites inSaudi Arabia, Yemen, Oman, and theUnited Arab Emirates demonstrates

the research potential of the region,as do increasingly detailed environ-mental records.Early western explorers of Arabia

recognized stone tools, often in asso-ciation with ancient lake beds.4 Theseearly findings were followed by thediscoveries of archeologists,5 whichsignificantly advanced from the 1970sonward, with large-scale surveys suchas the Comprehensive Survey of theKingdom (1976-1981) in Saudi Ara-bia.6,7 Analogous developments tookplace in southern Arabia.8–11 Over thelast decade, systematic surveys andinterdisciplinary excavations havegreatly increased our understandingof prehistoric Arabia,2,3,12–14 as havedevelopments in paleoenvironmentalstudies.15–17 Research is now extend-ing to the coastal waters around Ara-bia18 and to the use of remote sensingtechniques.19

In the paleoanthropological litera-ture, the Arabian Peninsula oftenserves as a useful blank on the mapin which to draw hypothetical andrather abstract dispersal arrows. Thishas particularly reflected patterns ofglobal genetic variation.20,21 In suchmodels, the specific paleoenviron-mental and biogeographical charac-teristics of Arabia have little role, yetit is precisely such contexts that arecritical to defining patterns of homi-nin dispersal and adaptation. Asgenetic studies of Arabian popula-tions have increased in scale, theyreveal a complex pattern.22–25 Suchstudies show that modern Arabianpopulations are mostly derived fromWestern Asia, reflecting dispersalssince the Last Glacial Maximum(LGM). In some areas, however,there are relatively high levels of‘‘African’’ lineages, which have gener-ally been attributed to historical

ARTICLE

Huw Groucutt is a doctoral candidateat the School of Archaeology, Universityof Oxford. His research focuses on theMiddle Paleolithic of Arabia and LatePleistocene hominin dispersals. He hasrecently conducted excavations in SaudiArabia and is undertaking a comparativeanalysis of lithic technology in Africa andSouthwestern Asia. Email: [email protected]

Michael Petraglia is Professor of HumanEvolution and Prehistory, Senior ResearchFellow, and Co-Director of Centre forAsian Archaeology, Art & Culture, Schoolof Archaeology, University of Oxford. Heis also a Senior Research Fellow, LinacreCollege (Oxford), and a member of theHuman Origins Program, SmithsonianInstitution. Email: [email protected]

Key words: Arabia; paleoenvironments;

Paleolithic; lithic technology; hominins

VVC 2012 Wiley Periodicals, Inc.DOI 10.1002/evan.21308Published online in Wiley Online Library(wileyonlinelibrary.com).

Evolutionary Anthropology 21:113–125 (2012)

processes such as slavery.26,27

Genetic evidence, then, is poorlyplaced to elucidate the position ofArabia in the dispersal of homininpopulations. Likewise, the extinctionof regional populations means theywill not be represented in contempo-rary genetic structure. In this situa-tion, archeology, in the context ofpaleoenvironmental fluctuation,offers a key way to elucidate the dis-persal of hominin populations intoArabia and their subsequent evolu-tionary and cultural trajectories.

GEOGRAPHY AND CLIMATE

The vast size of Arabia, at more

than 3 million km2, suggests that

hominin evolutionary processes have

been complex and regionally vari-

able. The land mass consists of a

range of topographical and environ-

mental settings, including highlands,

lowlands, a coastline some 7,000 km

long, and an interior with numerous

paleorivers and paleolakes (Fig. 1).

Understanding the variable topogra-

phy, ecology, and geography of Ara-

bia is therefore a necessary back-

ground to elucidate changes in homi-

nin demography.28,29

To the north of Arabia there are nofundamental barriers to hominin dis-persal, given favorable paleoenviron-mental conditions. The SinaiPeninsula offered a connection toAfrica.30 On the eastern edge of Ara-bia, the Persian Gulf today is a shal-low sea but, for most of the Pleisto-cene, was a large river valley, leading

some to suggest that it formed an im-

portant population refugia.31 The pos-

sibility of hominins crossing the Bab

al Mandab has been much debated.32

Today the Red Sea here is less than 30

km wide with small islands, meaning

that the maximum single water cross-

ing is less than 18 km. Isotopic studies

suggest that there has not been a land

bridge since the Miocene,33 but

during times of low sea level the gap

would have been significantly

reduced. However, Arabia was con-

cordantly arid during these times.

Consequently, some have stressed the

temporal lag between environmental

amelioration and sea level rise2;

others emphasize the hypothesis of a

significant role for ‘‘coastal oases.’’34

Modern observers are familiar withthe dry side of Arabia, most iconi-cally demonstrated by the Rub’ alKhali, which, at �600,000 km2,forms the world’s largest sandy de-sert. In the east, the Wahiba Sandsarea is of importance, having beenwell studied in paleoenvironmentalterms.35,36 Contemporary Arabia,however, has some areas, such as theAsir-Yemen highlands and Oman’sDhofar, that receive considerablymore precipitation, The landscape ofArabia reveals numerous signs ofmore humid conditions in the past.

Figure 1. Geography of the Arabian Peninsula. The locations of Pleistocene fossil localities and key sites for paleoenvironmental recon-struction are shown; note the southern and coastal concentration of the latter. Black lines depict major paleorivers.28 (Basemap cour-tesy of NASA’s Earth Observatory.) [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

114 Groucutt and Petraglia ARTICLE

Vast paleoriver systems cut acrossthe Peninsula, most starting in thewestern highlands and extendingeast to the Persian Gulf (Fig. 1). Forinstance, the now dry Wadi al Batindrained a large area of northernSaudi Arabia and left a huge alluvialfan around modern Kuwait, some300 m deep and covering 190 by 130km.28 Such paleorivers can clearly beseen in satellite images.37 Paleolakesof various sizes are found across Ara-bia, from the relatively small to thevast, such as the �2,000 km2

Mudawwara paleolake on the borderof Saudi Arabia and Jordan.13,38,39

The environmental evolution ofArabia primarily reflects the varyingcontributions of precipitation fromtwo key weather systems: winterrains from the Mediterranean and,most importantly, the summer rainsof the Indian Ocean monsoon sys-tem. The latitudinal movement of theInter-Tropical Convergence Zone(ITCZ), and with it the monsoonalrains, is the single most importantenvironmental variable in Arabianprehistory. As the ITCZ moved north,Arabia sprang to life as lakes refilledand rivers flowed. Paleoenvironmen-tal scientists have made considerableprogress in elucidating this andrelated processes.15–17,40

Environmental oscillation betweenwet and dry periods can be studiedat two levels, first at a broad ‘‘glacial’’(generally dry) and ‘‘interglacial’’(with wet phases) level and second interms of more short-term change.Figure 2 gives an overview of envi-ronmental change in Arabia andrelated archeological evidence. At abroad level, evidence indicates signif-icant increases in precipitation ineach interglacial period back to MIS11.41 However, most of our knowl-edge covers the last �350 kyr, andprimarily the younger half of this, asshown in Figure 2.15,17 Glacial peri-ods were often characterized by aridconditions. There are, however, com-plications to this dichotomy. In MIS6, for instance, recent evidence dem-onstrates at least short phases ofincreased precipitation,15 which con-trast with traditional views of this asa hyperarid period.42 MIS 5, particu-larly MIS 5e, 5c, and 5a, has beenshown by various studies to have

been particularly wet, with extensivespeleothem, calcrete, and lake forma-tion.13,17,43 A pronounced deteriora-tion occurred with MIS 4; however,few records are known for this pe-riod, so we are dealing primarilywith an absence of evidence (forexample, of speleothem formation).MIS 3 has been labeled the ‘‘debatedpluvial.’’15 Many radiocarbon dateshad suggested a prolonged humidperiod �30-25 ka,44,45 but this is not

reflected in records such as speleo-thems and probably reflects prob-lems with the original radiocarbondates.15,46 A recent study of paleo-lakes demonstrated that some keyexamples previously dated to MIS 3actually date to MIS 5.16 MIS 2,encompassing the LGM, seems tohave been generally extremelyarid.35,47 The most recent majorhumid period dates to �10-6 ka. This‘‘Holocene wet phase’’48–51 is bestknown from southern Arabia, wheremost studies have been conducted, butprecipitation in northern Arabia wasmore than 300% greater than at pres-ent.52 A period of aridification followed,with evidence of abrupt drying.53

Several recent studies have demon-strated, within this broad pattern ofglacial-interglacial oscillation, the ex-istence of short-term wet phases.Presumably, some of the longer wet

periods also included short dryphases. Recent studies in Saudi Ara-bia and the United Arab Emiratesdemonstrate a short-lived wet phaseat �55 ka.39,54 Another short pluvialperiod occurred �14 ka, not long af-ter the LGM.55,56

Much progress has been made inelucidating paleoenvironmentalchange in Arabia, but many ques-tions remain. A key question con-cerns the relative contributions ofthe two weather systems. This hasimplications with regard to the open-ing of dispersal routes. Within inter-glacial periods, what were the dryperiods, such as MIS 5b and 5d, like?And what was the impact of theshort, wet phases in terms of demog-raphy and dispersals? Only recentlyhas the chronology of these short-term fluctuations begun to be under-stood. While records such as speleo-thems provide excellent overallrecords, they seem to not be sensitiveenough to detect short-term pluvialepisodes.

FLORA AND FAUNA

Plants and animals provide infor-mation that is relevant to under-standing hominin occupations. Thecontemporary biome of Arabiareflects its position as an interfacebetween biogeographical zones, par-ticularly the Palearctic and Afrotropi-cal. Changing environmental condi-tions increased the influence of oneor the other of these, with the result-ing mix then filtered through aridphases that gave a distinctly Arabiancharacter to many taxa. One particu-larly important example are thebaboons, Papio hamadryas, whichare found in southwestern Arabia.These are the only wild baboonsfound outside sub-Saharan Africa.The most recent genetic study toexamine the timing of baboon dis-persal into Arabia reported coales-cence ages of �333 ka, 216 ka, and105 ka.57 While such dates comewith rather wide error margins, theysignificantly coincide with intergla-cial periods (Fig. 2). It has beenclaimed that the northerly Arabianpopulations are more geneticallysimilar to African populations,57 per-haps suggesting a Sinai dispersal

The latitudinalmovement of the Inter-Tropical ConvergenceZone (ITCZ), and with itthe monsoonal rains, isthe single mostimportant environmentalvariable in Arabianprehistory. As the ITCZmoved north, Arabiasprang to life as lakesrefilled and riversflowed.

ARTICLE Groucutt and Petraglia 115

route. However, another studyobserved the opposite pattern,58

although issues with sample size are

notable. Through such research, the

wider floral and faunal context of

hominin dispersals into and occupa-

tion of Arabia can be elucidated.Arabia has a rich Miocene fossil

record,59 but few younger fossils areknown. Thomas and colleagues60 dis-covered Pleistocene fossils (n ¼ 139)at three paleolake contexts in the AnNafud desert. They suggested thatthe fauna had an Afro-tropical char-acter. Taxa represented include fish,tortoise, carnivores (Crocuta crocuta,Panthera cf. gombaszoegensis, Vulpescf. vulpes), elephant, horse, hippo(Hexaprotodon sp.), and giant buffalo(Pelorovis cf. oldowayensis). The rep-resented taxa and isotopic valuesfrom teeth indicate a savannah-likeenvironment around the lakes.60 Theoryx present at the An-Nafud fossillocalities has been described as aspecifically arid-adapted form,61 indi-

cating the potential survival of cer-tain mammals in dry periods. Thediscoverers’ suggestion of an EarlyPleistocene date must be taken withcaution, as it was not based on abso-lute dates; moreover, the three local-ities need not be synchronous.McClure62 reported fossils of a vari-ety of species, including gazelle, hip-popotamus, oryx, ostrich, and buf-falo from the area of the Mundafanpaleolake, which was recentlyredated to MIS 5.16 Many sites pre-serve Holocene faunal remains.63

THE LOWER PALEOLITHIC

Arabia has a rich Lower Paleolithicrecord (Figs. 3 and 4).7 SouthwesternSaudi Arabia seems to have a partic-ularly large number of Lower Paleo-lithic sites, although it is unclear towhat extent this reflects survey andpublication bias. In comparison toyounger Paleolithic sites, those of theLower Paleolithic appear to extend

further into the western and north-western fringes of the Rub’ al Khalidesert, perhaps indicating the occur-rence of more extended wet phasesin the earlier Pleistocene. The paleo-environmental conditions of the EarlyPleistocene are unclear, but presum-ably followed a pattern of oscillationsimilar to that seen more recently, butthis was probably less extreme in itsfluctuation. Unfortunately, few excava-tions have taken place, so that thechronology and internal variability ofthe Lower Paleolithic is essentiallyunknown. Localities appear to be cor-related with hilltops, ridges, and terra-ces, and are often located in closeproximity to raw material sources.Several claims have been made for

‘‘Oldowan-like’’, or ‘‘Mode 1’’ assemb-lages in Arabia.64–69 For instance, at201-49 near Shuwayhitiyah in north-ern Arabia, Whalen and colleaguescollected lithics from 16 localities.66

Their collection was dominated byheavy-duty tools (choppers, polyhe-drons, and such), with a fairly smallbifacial component (Fig. 3). Like-wise, claims for ‘‘pre-Acheulean’’ siteshave been made in Yemen. From Al-Guza Cave, Amirkhanov68 reports anOldowan ‘‘pebble industry’’ lacking abifacial component. However, aswith other purportedly Oldowanassemblages, care needs to be takento differentiate artifacts from geo-facts. Similarly, Jagher reports thatintensive efforts to relocate the doz-ens of purportedly Oldowan sitesaround Huqf, Oman,70 did not resultin the discovery of any artifacts.71

Chauhan67 reports the discovery ofOldowan-like lithics on Perim Island,a small island on the Yemeni side ofthe Bab al Mandab. If confirmed, thismay suggest the crossing of the Bab alMandab in the Early Pleistocene.The presence of hominins using

Oldowan-like, or Mode 1, technologyin Arabia is certainly possible, giventhe findings at sites such as Dmanisito the north. However, problems per-vade the Arabian finds. To take theexample of the Shuwayhitiyah local-ities, the assemblages include lithicsthat are clearly recent (probably Hol-ocene); in addition, the collectionstrategy of the discoverers is unclear,and the use of quartz as a rawmaterial may be a significant factor

Figure 2. Arabian environmental, faunal, and archeological records over the last 150 kyr.Thick vertical blue lines represent key humid periods, thin grey lines represent short humidepisodes. Note the correlation of archeological assemblages with periods of increasedrainfall, with the possible exception of Assemblage A at Jebel Faya. Environmental infor-mation and figure structure adapted from Rosenberg and coworkers16; chronology ofbaboon dispersal after Fernandes57; key archeological sites.2,3,13,14,92 [Color figure can beviewed in the online issue, which is available at wileyonlinelibrary.com.]


in determining artifact form. Moregenerally, arguments based solely ontypology carry many inherent limita-tions. Care needs to be taken to dis-tinguish geofacts from artifacts. Also,excavations are required to recoverchronometrically datable assemb-lages that can be placed in a paleoen-vironmental context. If they are takenat face value, it is significant thatmore Oldowan-like sites are known inArabia than in North Africa, perhapsindicating a significant role for Arabiain the early hominin colonization ofEurasia.72 While Whalen and col-leagues believed that the Shuwayhi-tiyah collection is strongly reminis-cent of the Developed Oldowan ofEast Africa,66 comparative technolog-ical analysis is required to elucidate

the relationship of the ArabianLower Paleolithic to that of sur-rounding regions. Also, excavationsneed to be done to clarify the demo-graphic processes of populations inthe Peninsula.

We are on firmer ground with theAcheulean, as sites have been identi-fied across Arabia. Spectacular‘‘Acheulean landscapes’’ are found atDawadmi and Wadi Fatimah, in cen-tral and western Saudi Arabia,respectively. At Dawadmi, various

Acheulean sites were identified in

association with an andesite and rhy-olite dyke.73 Two sites here, 206-76

and 207-68, were subsequently exca-

vated.74,75 The discovers rather spec-ulatively suggested distinct ‘‘func-

tional activity areas,’’ supposedly

demonstrating eight different activ-ities. Six uranium-thorium dates for

calcareous matter adhering to theartifacts produced minimal ages of201-61 ka.75:22 The surface collectionand excavation of these sites pro-duced thousands of lithics, but therehave been no detailed technologicalcomparisons of these lithics toassemblages from surroundingregions. An important factor at theDawadmi sites is the presence ofabundant evidence of early stagereduction. Giant cores demonstrateprocurement and reduction of rawmaterial along the dykes.7,76 The 32Acheulean sites identified near theRed Sea in Wadi Fatimahappear,77,78 based on selective sur-face collections, to have a typologicalstructure similar to those excavatedat Dawadmi and also are locatedclose to sources of raw material. Inboth cases, flakes are the dominanttype, with characteristic bifaces andcleavers making up around 1% of theassemblages (Fig. 3). Uranium-tho-rium dating of calcareous nodules atsite 210-351 produced an age ‘‘in therange of 200,000 years.’’78:78

The Dawadmi and Wadi Fatimahsites share similarities in technologyand landscape position, being closeto raw material sources and associ-ated with riverine systems, whichpresumably functioned as dispersalcorridors.76 The distinctive Acheu-lean bifaces and cleavers at thesesites are generally characterized bydeep flake scars; only rarely are theyhighly symmetrical (Fig. 3). Measure-ments of biface elongation (length towidth ratio) comfortably place Ara-bian assemblages within the Acheu-lean, their mean elongations fallingbetween those of assemblages inAfrica and India.79 Whalen’s excava-tions revealed the presence of buriedLower Paleolithic sites in Arabia.Future excavations at these andother localities are likely to be highlyinformative. At present, the onlyabsolute dates available reflect mini-mum ages.75,78 Comparative designa-tions, typically classifying the Arabianmaterial as ‘‘Middle Acheulean,’’74,75,78

are premature. Detailed statements onthe spatial and temporal relationshipsof the Arabian assemblages to those ofsurrounding regions require consider-ably more research.

Figure 3. Above: view from the top of the jebel at Dawadmi, looking west over Acheu-lean site 206-76 (where vehicles are located). This is the location of one of the first Paleo-lithic site excavations in Arabia. Acheulean sites and artifacts are distributed at the baseof the jebel (over a distance of ca. 10 km) and below the andesite dykes, which servedas raw material sources for stone tool manufacture. Below: selected handaxes fromDawadmi, manufactured from andesite. [Color figure can be viewed in the online issue,which is available at wileyonlinelibrary.com.]


THE MIDDLE PALEOLITHIC

This is the most well-represented

Paleolithic phase in Arabia (Fig. 4),

and its study has seen significant

recent advances. As in the Lower

Paleolithic, sites are often associated

with raw material sources and lacus-

trine and riverine systems, and arelocated at strategic positions in thelandscape. Sites are found acrossArabia except within modern sandy

desert areas. The widespread surfacerecord guided initial conceptions ofMiddle Paleolithic variability6 andcan now be anchored to stratifiedsites excavated in the last five years.Rose and colleagues’ systematic sur-vey in Oman’s Dhofar reveals therich density of Middle Paleolithicsites found when modern surveytechniques are applied.3

Traditionally, the Middle Paleo-lithic of Arabia has been viewed from

a Levantine and European perspec-tive, with sites often described as‘‘Mousterian.’’ This inappropriateframework, combined with the tradi-tional lack of stratified sites, made itdifficult to address spatial and tem-poral variability and hampered thefactoring of the Arabian record intodebates such as that on Out-of-Africadispersals. Nevertheless, importantinformation was gathered at the hun-dreds of surface sites identified.These sites demonstrate the range oftypo-technological variability foundin the Arabian Middle Paleolithicand give some insight into the wayhominins related to the landscape,while stratified sites have the benefitsof a stronger chance of assemblagecontemporaneity and allow bothabsolute dating and paleoenviron-mental contextualization.The Arabian record demonstrates

particular characteristics, includingthe often somewhat irregular, multi-platform character of cores, which isassociated with features such as ageneral paucity of platform prepara-tion. A significant feature of theArabian Middle Paleolithic is thewidespread presence of a bifacialcomponent, which was confirmedby the excavation of Jebel Faya(Fig. 5).2 Rose80 has stressed thesignificance of bifacial technology inOman, suggesting that it impliesLate Pleistocene population connec-tions with sub-Saharan Africa. Sur-veys have shown significant regionalvariation in Arabia. The productionof points, for instance, seems to bevariable, being common in Dhofar,among other sites,3 but seemingly rarein other areas. Zarins and coworkers,73

for example, perhaps reflecting theirLevantine-oriented expectations, foundthe lack of points between Riyadh andJeddah ‘‘disconcerting.’’ Similarly, therepresentation of Levallois technologyappears to be variable in Arabia,although this, in part, reflects evolutionfrom typological to volumetric concep-tions. For instance, as Crassard pointsout, the ‘‘polyhedrons’’ and ‘‘discoids’’illustrated by Whalen and Schatte81,82

are in fact recurrent centripetal Leval-lois cores. Various ‘‘industrial’’ designa-tions have been proposed,6 but theseare problematic. For instance, the pos-sibility that at the single site of Bani

Figure 4. Lower and Middle Paleolithic sites of Arabia. Note that some points representgroups of sites. (DEM data courtesy of the International Centre for Tropical Agriculture37;basemap image courtesy of Nick Drake and Paul Breeze, Kings College, London.) [Colorfigure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]


Khatmah, an assemblage dominatedby tanged points and scrapers from thewestern fringe of the Rub’ al Khalirelates to the otherwise North AfricanAterian needs to be tested by excava-tions and comparative technologicalanalysis.83 A reanalysis of the BaniKhatmah lithics suggests that the Arabian‘‘Aterian’’maydate to theHolocene.84

A report on the Middle Paleolithicsite of Shi’bat Dihya 1 and others inthe Wadi Surdud of western Yemenis to be published shortly.14 A key de-velopment is that this site is nowdated to �55 ka, in contrast to initialestimates of �80-70 ka.12,85 Morethan 5,000 lithics having lengthsgreater than 2-cm, as well as faunalremains, were excavated from a sin-gle thin layer covering 21 m2. Isotopicand paleobotanical data indicate that

the site was occupied during rela-tively arid conditions. The timing ofoccupation correlates with the shortwet phase previously mentioned at�55 ka (Fig. 2), which occurredbetween more arid periods on eitherside. The lithic technology reflects avariety of reduction schemes for theproduction of flakes, blades, andpoints. Levallois technology is pres-ent, but not common. No immediateparallels are found with contempo-rary African or Levantine assemb-lages. Other sites in Wadi Surdud aredated to younger periods of MIS 3.The overall picture is of a distinctiveregional variant, an autochthonousdevelopment from a presentlyunknown ancestral population. Else-where in Yemen, more Levallois-ori-ented point production perhaps hints

at earlier connections with the Levantor may reflect convergent evolution ina similar habitat.11,81,86

At the site of Jebel Faya (FAY-NE1), excavations uncovered strati-fied Paleolithic assemblages.2 Pub-lished information addresses the topthree Paleolithic assemblages (la-beled A to C with increasing depth).Assemblage C (�500 lithics, see Fig.5) is dated by optically stimulated lu-minescence (OSL) to 95 6 13 ka, 1236 10 ka, and 127 6 16 ka, althoughthe authors indicate that the latterdate may be problematic.2 This lastinterglacial assemblage demonstratesa variety of reduction strategies.These include the production of volu-metric blades and Levallois debitage,as well as bifaces, and a variety ofretouched forms. Similar assemb-lages have been discovered nearby insurface contexts,87,88 showing thatthis technological package character-izes the Middle Paleolithic of theArabian side of the eastern PersianGulf, at least during MIS 5. Whilethe closest technological parallels arefound with northeast Africa, it is notimpossible that assemblage C repre-sents an autochthonous develop-ment, or indeed a dispersal from, forinstance, the Indian subcontinent.89

In the overlying assemblages (A andB), bifacial and Levallois reduction isapparently absent except for a fewconvergent flakes in B, which are sim-ilar to Levallois points.2 Neitherbacked nor microlithic technologiesare represented. Assemblage B isundated, but on stratigraphic groundswas deposited between 90 and 40 ka,while A is dated to 40.2 6 4.0 ka and38.6 6 3.1 ka, confirming homininoccupation of eastern Arabia in MIS3. Both assemblages B and A are char-acterized by diverse reduction strat-egies, centered on the production offlakes. In assemblage B, core typesinclude Kombewa and radial, whichare also found in assemblage A, whichis dominated by orthogonal multiplat-form cores. Neither A nor B has a sub-stantial retouched component.A third spatially and temporally

specific Middle Paleolithic industry isfound in Dhofar, Oman. Rose andcolleagues3 identified 110 sites withtechnology very similar to that of theMIS 5 Nubian Complex in northeast-

Figure 5. Selected Arabian Middle Paleolithic artifacts. 1) Levallois core from RASA 2004-149-1, Yemen81; 2) and 3) Levallois cores from Jubbah, Saudi Arabia13; 4) unifaciallyretouched point from Jubbah, Saudi Arabia13; 5) Levallois flake from assemblage C, FAY-NE1, UAE2; 6) Levallois point from Aybyt Thania, Oman3; 7) Levallois point from AybutAuwal, Oman3; 8) bifacial foliate from FAY-NE1, UAE.2


ern Africa. Previously rare examplesof Nubian-like cores had been identi-fied,86 but the new discoveries consti-tute a widespread and seemingly ho-mogenous industry. The sites arefound exclusively on the Nejd Plateauand have not been identified near thecoast. All of the sites take the form ofsurface assemblages, with the excep-tion of Aybut Auwal. Here a smallnumber of artifacts, including a diag-nostic Type 1 Nubian core, werefound in situ, while many otherlithics appear to have eroded fromthe same sediments onto the surface.Two OSL samples from the samestratigraphic unit as the excavatedlithics produced age estimates of106 6 9 ka and 107 6 9 ka, correlat-ing with MIS 5c. Lithics from AybutAuwal and three other sites were ana-lyzed in detail. At all of the sites, Nu-bian Type 1 cores are dominant. Thecore technology of the Dhofar Nubianis remarkably similar to the diagnos-tic cores of the Nubian Complex inNorth Africa. Debitage is dominatedby flakes, although blades are alsocommon. Aside from points, side-scrapers are the most numerous tool.A bifacial component was not identi-fied at any site.In Oman, another industry, which

Rose10 labeled the ‘‘Nejd Leptolithic,’’is widespread. This is characterizedby the production of large bladesfrom unidirectional cores with littleplatform preparation. Blanks wererarely retouched. An interesting fea-ture of this industry is the frequentpresence of lipped platforms, whichare often taken as an indication ofthe use of a soft hammer. Rose andcolleagues3 suggest, on geomorpho-logical grounds, that the Nejd Lepto-lithic is younger than at least someof the MIS 5 Nubian sites. Theseassemblages appear similar to somefrom Oman described by Jagher.71

We might see these large-bladeassemblages as a distinctive regionalpost-MIS 5 development in Oman.The situation in Saudi Arabia is

less clear than that in southern Ara-bia, where most recent research hasfocused. The authors of this paperare conducting research at theJubbah paleolake in northern SaudiArabia,13 an area highlighted by aninitial reconnaissance survey in the

1970s.90 At the site of Jebel Qattar 1,we identified a lithic assemblagecharacterized by centripetal Levalloisand discoidal reduction, in associa-tion with a calcrete dated by OSL to75 6 5.0 ka. As well as notches, den-ticulates, and scrapers, we found aninvasively retouched unifacial point(Fig. 5). Field work in 2011 led to theidentification of several stratifiedMiddle Paleolithic sites around thepaleolake. Our analysis is ongoingand promises to cast light on thecharacter of the Middle Paleolithic ininterior northern Arabia.

As research gathers pace in Arabia,we are now beginning to be able tocorrelate paleoenvironmental and ar-cheological records for the MiddlePaleolithic and move toward compari-sons with surrounding regions. At abroad level, features such as the pres-ence of a bifacial component at manyArabian Middle Paleolithic sites argu-ably orients the Peninsula towardAfrica, where a faconnage componentwas frequent in the Middle Stone Age(MSA).80 In contrast, the LevantineMiddle Paleolithic seems to lack anyevidence of faconnage technology.This regional divergence took placemore than 200 kyr ago. The increasingevidence of variability in the ArabianMiddle Paleolithic suggests thatmultiple dispersals into the areaoccurred, followed by regional autoch-thonous trajectories. Where enoughdata are available, the Arabianassemblages suggest connections toAfrica in MIS 5.2,3 Subsequently, as

seen most clearly at Jebel Faya andWadi Surdud, and perhaps with otherexamples such as the large bladeassemblages of Oman, the outlines ofpost-MIS 5 regional autochthonoustrends are coming into focus. So far,evidence of assemblages similar to theHowiesons Poort and other precociousfacies of the MSA is absent in Arabia,while the predominantly unidirectional-convergent production of Levalloispoints similar to that in the early andlate phases of the Levantine MiddlePaleolithic is known only from Yemen.

THE LATE PALEOLITHIC

Pre-Holocene evolutionary and cul-tural trajectories are poorly under-stood in Arabia. The character of theLate Paleolithic, by which we meanthe period �40-10 ka, which else-where is typically associated withblade-dominated or microlithicassemblages and frequent evidence ofsymbolism and other complex behav-iors, is one of the great mysteries ofArabian archeology. As discussed in arecent review,91 such evidence isephemeral in Arabia. With the usualcaveats in mind, such as a generalreliance on surface sites, a generousinterpretation suggests a possibleLate Paleolithic component at a smallnumber of sites focused in northwest-ern Arabia (Fig. 6). It must bestressed that such attributions areproblematic. The only excavated anddated Late Paleolithic site, Al Hatab,has lithics unlike those of the Epipa-leolithic/Upper Paleolithic elsewhere,and again appears to reflect an au-tochthonous southern Arabian devel-opment (Fig. 7).92

‘‘Upper Paleolithic’’ sites have beendefined by some on primarily typolog-ical grounds, emphasizing the pres-ence of types such as burins andblades64,73; others stress ‘‘interme-diary’’ levels of ‘‘patination, materials,and workmanship.’’ In Yemen, Amir-khanov9 used both of these perspec-tives to define a number of ‘‘UpperPaleolithic’’ sites. These sites needmore study, particularly absolute dat-ing and technological analysis. Super-ficially, many appear actually to becompatible with a Middle Paleolithicdesignation. One interesting assem-blage is the material from Faw Well

As research gatherspace in Arabia, we arenow beginning to beable to correlatepaleoenvironmental andarcheological recordsfor the Middle Paleolithicand move towardcomparisons withsurrounding regions.


in southwestern Saudi Arabia. Thisassemblage, which seems to be unlikeany other in Arabia, but lacks anyabsolute dates, is characterized byblade and microblade production.Edens93 studied part of the assem-blage and suggests possible similar-ities with the Upper PaleolithicAhmarian industry in the Levant.More classically ‘‘Epipaleolithic-like’’sites have a position similar to thoseof the ‘‘Upper Paleolithic.’’ The more

convincing examples are located inthe far north of the Arabian Penin-sula, perhaps indicating brief incur-sions from the Levant.64,94

The excavations at Al-Hatab inOman uncovered a lithic assemblagedating to �13 ka and possiblyextending into the earliest Holocene.That assemblage has typo-technolog-ical similarities to material from thesites of Ras Aın Noor and Dhanaqr,which the discoverers suggest reflects

an autochthonous development oflaminar technologies extending backto perhaps MIS 4.92 The assemblageincludes hard hammer blades, bifa-cial foliates, burins, and endscrapers.The Arabian record seems to lack

specific features relating to lithictechnology and other features of theAfrican Late Stone Age. There arepossible hints at occasional contactswith the laminar industries of theLevantine Upper Paleolithic but, inthe absence of detailed technologicalanalyses, these are broad-scale com-parisons. These features, such as theassemblage from Faw Well, perhapssuggest that dispersals from the Le-vant occurred during the short wetphases of MIS 3. It is possible thatsuch populations survived throughthe LGM in southern Arabia. This isperhaps supported by the distinc-tively Arabian character of assemb-lages at sites like Al-Hatab, indicat-ing regional autochthonous develop-ments.92 It is possible that the sitesin southern Arabia represent a con-tinuation of occupation from MIS 5.Surface sites in the north and westare poorly studied, and indeed mayeven date to the Holocene.

INTO THE HOLOCENE

The early to middle Holocene ofArabia has implications for generalpatterns of occupation and dispersalsin Arabia, as well as adaptation toarid environments. Considerablymore detailed archeological and envi-ronmental records are available forthe Holocene than are for the Pleis-tocene. Many absolute dates areavailable and at least 28 sites pre-serve faunal remains.63 Widespreadrock art provides new insights intohuman behavior, although its studywould be helped by an evolutionfrom the stylistic to more systematicand geographic information system(GIS)-based analyses. In addition,many of the key themes we have dis-cussed in relation to the Paleolithicalso characterize debates such asthat about the origins of the ‘‘Neo-lithic’’ in Arabia.The large number of early to mid-

dle Holocene sites probably demon-strates a substantial demographicincrease, associated in particular with

Figure 6. Late Paleolithic and Early/Middle Holocene sites of Arabia. Note that some pointsrepresent groups of sites. (DEM data courtesy of the International Centre for Tropical Agricul-ture37; basemap image courtesy of Nick Drake and Paul Breeze, Kings College, London.)[Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]


the Holocene wet phase (Fig. 6). Anumber of autochthonous develop-ments can clearly be seen in HoloceneArabia (Fig. 8). These include the de-velopment of new reduction strat-egies such as the ‘‘Wa’sha method’’ inYemen11,95 and the development of

other forms of projectile points in

the Neolithic,96 including ‘‘Fasad

points.’’97 Nonlithic developments

include the domestication of animals

and the development of seafaring.98

Debate continues on whether theHolocene occupation of Arabiareflects population dispersal into thearea,63,97 probably from the Levant,or was a primarily indigenous devel-opment.31 In terms of lithic technol-ogy, for instance, the distinctiveFasad points of southern Arabia havebeen seen as a derivation from Le-vantine early Neolithic technology97

or as an autochthonous developmentfrom Arabian terminal Pleistocenetraditions.92 The discovery of a singleFasad point dating to �13 ka at AlHatab can be taken as supportingthe latter position,92 but it is debata-ble how much a single, unique exam-ple of anything can show. The pres-ence of apparently closely pre-Neo-lithic sites in Yemen is intriguingand again may also support the latter

perspective.99 This may be a ratherabstract dichotomy, as numerous de-mographic scenarios are possible,including both population continuityand population dispersals into thearea. In fact, genetic evidence seemsnow to demonstrate clearly somepopulation continuity from the ter-minal Pleistocene into the Holocene,with additional dispersals from theLevant at �10 ka.22 The Neolithic inArabia has a distinctive characterthat reflects adaptations to the aridenvironment. The use of domesti-cated animals seems to have been ofconsiderably greater importancethan domesticated plants. Sometimelater, perhaps around the third mil-lennium BCE, the domestication ofcamels occurred in Arabia. Throughsuch adaptations to arid environ-ments, people were able to survive inextreme areas and began to tran-scend the ancient correlationbetween demography and environ-mental amelioration.

DISCUSSION AND CONCLUSION

In recent years there has been a pro-nounced acceleration in the pace ofarcheological research and discoveryin Arabia, yet the region remains

remarkably poorly understood. How-ever, newly emerging information cor-relates the human and environmentalstories (Fig. 2). We hypothesize thatmuch of Arabian population history iscyclical; that is, with increased precip-itation, hominin populations repeat-edly expanded into the area. They sub-sequently followed autochthonoustrajectories before becoming extinct,at least until the Late Pleistocene/Hol-ocene, when adaptations to arid con-ditions developed.Given the severity of environmen-

tal fluctuation, which, as Figure 2shows, is increasingly well under-stood, it is reasonable to assume thatthere has not been long-term popula-tion continuity in Arabia at an evolu-tionary scale. This emphasizes theneed to understand the Arabian re-cord in an interregional context. Forthe Lower Paleolithic it is too earlyto make definitive statements in thisregard, but Homo heidelbergensis andHomo erectus probably played a role,as possibly did early Homo. Therehas been little focus on this phase inrecent years, as attention has focusedon modern humans, but there isclearly great potential to elucidatethe Lower Paleolithic in Arabia.The key development in recent

years has been the demonstration,particularly in Dhofar and at JebelFaya,2,3 that Arabia contains African-like Middle Paleolithic assemblagesdating to MIS 5. This strongly sug-gests that dispersals took place atthis time, although fossil evidencewould help to substantiate this hy-pothesis. For example, the archeolog-ical evidence from Dhofar suggestspopulation dispersal from northeastAfrica by MIS 5c,3 a time whenbaboons also seem to have dispersedinto Arabia,57 and when paleoenvir-onmental evidence demonstrates awet phase. This shows how an inter-disciplinary perspective can elucidatethe Paleolithic occupation of Arabia(Fig. 2). Further excavations areneeded to distinguish between differ-ent hypotheses. For instance, doesthe occupation at Jebel Faya repre-sent a continuous occupation of thePersian Gulf area between �125 and40 ka, as the discoverers suggest,2 ordoes it rather indicate repeated dis-persals into the area?

Figure 7. Selected Arabian Late Paleolithic artifacts. 1) backed blade; 2) double platformcore; 3) single platform core, Faw Well, Saudi Arabia93; 4) blade; 5) blade core, Al Hatab,Oman.92


In general, indications of possiblepopulation connections to surround-ing regions remain rather specula-tive, and this is compounded by theabsence of fossil evidence. We sug-gest, however, that the emerging pic-ture suggests a general lack of con-nections between Africa and Arabiaafter MIS 5, the last interglacial.Instead, there are perhaps indica-tions that Arabia sometimes saw

connections to the Levant. The mod-ern Arabian genomic structure isdominated by lineages reflectingpost-LGM population movementsfrom the north.22,23 These dispersingpopulations mixed with existing pop-ulations in a manner that remains tobe firmly understood. The MiddlePaleolithic record is increasinglydemonstrating, probably in commonwith other prehistoric phases, that

after dispersals into the area, envi-ronmental deterioration divided pop-ulations into refugia. These include

the Yemeni highlands and the Per-

sian Gulf. The combination of

regionalization, an arid environmen-

tal setting, and small population

sizes will have all played a significant

role in determining the character of

human adaptations. To take the

lithic evidence, variability will

express demographic, raw material,

functional, and cultural factors. The

balance between these remains to be

understood. Hence, relations to sur-

rounding areas and between regions

within Arabia remain obscure. The

key point is that advances in empiri-

cal data need to be complemented by

theoretical developments such as

those of behavioral ecology, rather

than by a simplistic or ‘‘empiricist’’

perspective that merely compares the

basic morphology of lithic artifacts,

for instance, while ignoring their

context.To conclude, we highlight the sig-

nificance of the Arabian record in

casting light on the development ofadaptations to extreme environ-

ments. The Arabian record is, in fact,

uniquely positioned to elucidate suchdevelopments, as populations dis-

persed into the Arabian Peninsula

from neighboring ‘‘hot spots’’ such as

East Africa and were trapped by thedesiccation of the routes they had

originally followed. Given wide-

spread public and scholarly interest

in and concern about contemporary

climate change, the long-term

records of human occupations and

environmental change in Arabia are

of great importance. With recent

advances, archeologists and our col-

leagues in related disciplines have

begun to outline the chapters of the

Arabian story. Now it is time to fill

in the pages.

ACKNOWLEDGMENTS

We thank HRH Prince Sultan binSalman bin Abdul-Aziz Al Saud,President of the Saudi Commissionfor Tourism and Antiquities (SCTA),and Professor Ali Al-Ghabban, VicePresident of SCTA, for permission to

Figure 8. Selected Arabian Early/Middle Holocene artifacts. 1) Neolithic arrowhead fromJubbah, Saudi Arabia; 2) fluted point from Manayzah, Yemen11; 3) Fasad point from Nadal-Thamam, UAE97; 4) Fasad point from FAY-NE1, UAE97; 5) Wa’shah core from HDOR 538,Yemen95; 6) bifacial foliate from HDOR 538, Yemen11; 7-10) Neolithic points from Khuz-mum, Yemen11; 11) Wa’shah point from HDOR 538, Yemen.95


conduct research in the Kingdom ofSaudi Arabia. We acknowledge Abdul-lah Alsharekh for his unwavering sup-port of our research, as well as RemyCrassard and Adrian Parker for dis-cussions on the Arabian evidence, andJames Blinkhorn for commenting ona draft of this paper. We acknowledgethe financial support of the LeakeyFoundation, the National GeographicSociety, and the Arts and HumanitiesResearch Council (UK, doctoral stu-dentship to H.G.). We appreciate theconstructive comments of John Flea-gle, John Shea, Frank Preusser, andthree anonymous reviewers.

REFERENCES

1 Petraglia MD, Rose JI, editors. 2009. The evo-lution of human populations in Arabia: paleo-environments prehistory and genetics. Nether-lands: Springer.

2 Armitage SJ, Jasim SA, Marks AE, et al. 2011.The southern route ‘‘out of Africa’’: evidence foran early expansion of modern humans into Ara-bia. Science 331:453–456.

3 Rose JI, Usik VI, Marks AE, et al. 2011. TheNubian complex of Dhofar, Oman: an AfricanMiddle Stone Age industry in southern Arabia.PLoS ONE 6:e28239.

4 Philby HSJ. 1933. Rub’ Al Khali: an accountof exploration in the Great South Desert of Ara-bia under the auspices and patronage of hisMajesty ‘Abdul’Aziz ibn Sa’ud, King of theHejaz and Nejd and its dependencies. Geogr J81:1–21.

5 Caton-Thompson G, Gardner EW. 1939. Cli-mate, irrigation, and early man in the Hadhra-maut. Geogr J 93:18–35.

6 Petraglia MD, Alsharekh AM. 2003. The Mid-dle Palaeolithic of Arabia: implications formodern human origins, behaviour and disper-sals. Antiquity 77:671–684.

7 Petraglia MD. 2003. The Lower Paleolithic ofthe Arabian Peninsula: occupations, adaptations,and dispersals. J World Prehist 17:141–179.

8 de Maigret A. 1986. Archaeological activities inthe Yemen Arab Republic, 1986, East West36:376–470.

9 Amirkhanov H. 2006. Stone age of south Ara-bia. Moscow: Nauka.

10 Rose JI. 2006. Among Arabian sands: defin-ing the Paleolithic of southern Arabia. Ph.D.dissertation. Dallas:. Southern Methodist Uni-versity.

11 Crassard R. 2008. La prehistoire du Yemen:diffusions et diversites locales a travers l’etuded’industries lithiques du Hadramawt. BAR,1842. Oxford: Archaeopress.

12 Delagnes A, Macchiarelli R, Jaubert J. 2008.Middle Paleolithic settlement in Arabia: firstevidence from a complex of stratified archaeo-logical sites in western Yemen. Presented at theAnnual Meeting of the Palaeoanthropology,Vancouver.

13 Petraglia MD, Alsharekh AM, Crassard R,et al. 2011. Middle Paleolithic occupation on aMarine Isotope Stage 5 lakeshore in the NefudDesert, Saudi Arabia. Quaternary Sci Rev30:1555–1559.

14 Delagnes A, Brenet M, Crassard R, et al. n.d.The Middle Paleolithic assemblage of Shi-batDihya 1 (Wadi Surdud site complex, Yemen). JHum Evol. In review.

15 Parker AG. 2009. Pleistocene climate changein Arabia: developing a framework for hominindispersal over the last 350 ka. In: Petraglia MD,Rose JI, editors. The evolution of human popula-tions in Arabia: paleoenvironments, prehistoryand genetics. Netherlands: Springer. p 39–49.

16 Rosenberg TM, Preusser F, Fleitmann D,et al. 2011. Humid periods in southern Arabia:windows of opportunity for modern human dis-persal. Geology 39:1115–1118.

17 Fleitmann D, Burns SJ, Pekala M, et al.2011. Holocene and Pleistocene pluvial periodsin Yemen, southern Arabia. Quaternary Sci Rev30:783–787.

18 Bailey GN, Flemming NC, King GCP, et al.2007. Coastlines, submerged landscapes andhuman evolution: the Red Sea Basin and the Far-asan Islands. J Island Coastal Archeol 2:127–160.

19 Kenney D. 2011. The ‘‘works of the old men’’in Arabia: remote sensing in interior Arabia. JArchaeol Sci 38:3185–3203.

20 Quintana-Murci L, Semino O, Bandelt, HJ,et al. 1999. Genetic evidence of an early exit ofHomo sapiens sapiens from Africa through east-ern Africa. Nat Genet 23:437–44.

21 Macaulay V, Hill C, Achilli A, et al. 2005.Single, rapid coastal settlement of Asia revealedby analysis of complete mitochondrialgenomes. Science 308:1034–1036.

22 Cerny V, Mulligan CJ, Fernandes V, et al.2011. Internal diversification of MitochondrialHaplogroup R0a reveals post-Last Glacial Maxi-mum demographic expansions in south Arabia.Mol Biol Evol 28:71–78.

23 Abu-Amero K, Ali H, Gonzalez A, et al.2009. Saudi Arabian Y-chromosome diversityand its relationship with nearby regions. BMCGenet 10:59.

24 Abu-Amero KK, Larruga JM, Cabrera VM,et al. 2008. Mitochondrial DNA structure in theArabian Peninsula. BMC Evol Biol 8:45.

25 Rıdl J, Edens CM, Cerny V. 2009. Mitochon-drial DNA structure of Yemeni population: re-gional differences and the implications for dif-ferent migratory contributions. In: PetragliaMD, Rose JI, editors. The evolution of humanpopulations in Arabia: paleoenvironments, pre-history and genetics. Netherlands: Springer. p69–78.

26 Abu-Amero KK, Gonzalez AM, Larruga JM,et al. 2007. Eurasian and African mitochondrialDNA influences in the Saudi Arabian popula-tion. BMC Evol Biol 7:32.

27 Richards M, Rengo C, Cruciani F, et al.2003. Extensive female-mediated gene flowfrom sub-Saharan Africa into near eastern Arabpopulations. Am J Hum Genet 72:1058–1064.

28 Edgell HS. 2006. Arabian deserts: nature ori-gin and evolution. Netherlands: Springer.

29 Vincent P. 2008. Saudi Arabia: an environ-mental overview. London: Taylor and Francis.

30 Waldmann N, Torfstein A, Stein M. 2010.Northward intrusions of low- and mid-latitudestorms across the Saharo-Arabian belt duringpast interglacials. Geology 38:567–570.

31 Rose JI. 2010. New light on human prehis-tory in the Arabo-Persian Gulf Oasis. CurrAnthropol 51:849–883.

32 Bailey G. 2009. The Red Sea, coastal land-scapes, and hominin dispersals. In: PetragliaMD, Rose JI, editors. The evolution of humanpopulations in Arabia: paleoenvironments, pre-history and genetics. Netherlands: Springer. p15–138.

33 Fernandes CA, Rohling EJ, Siddall M. 2006.Absence of post-Miocene Red Sea land bridges:biogeographic implications. J Biogeogr 33:961–966.

34 Faure H, Walter RC, Grant DR. 2002. Thecoastal oasis: Ice Age springs on emerged con-tinental shelves. Global Planet Change 33:47–56.

35 Preusser F, Radies D, Matter AA. 2002.160,000-year record of dune development andatmospheric circulation in southern Arabia.Science 296:2018–2020.

36 Radies D, Preusser F, Matter A, et al. 2004.Eustatic and climatic controls on the develop-ment of the Wahiba Sand Sea, Sultanate ofOman. Sedimentology 51:1359–1385.

37 Jarvis A, Reuter HI, Nelson A, et al. 2008.Hole-filled seamless SRTM data V4, Interna-tional Centre for Tropical Agriculture (CIAT).Available online at http://srtm.csi.cgiar.org.

38 Petit-Maire N, Carbonel P, Reyss JL, et al.2010. A vast Eemian palaeolake in southern Jor-dan (298N). Global Planet Change 72:368–373.

39 Parton A, Parker AG, Farrant AR, et al.2010. An early MIS3 wet phase at palaeolakeAqabah: preliminary interpretation of themulti-proxy record. Proc Sem Arabian Studies40:267–276.

40 Preusser F. 2009. Chronology of the impactof Quaternary climate change on continentalenvironments in the Arabian Peninsula. C RGeosci 341:621–632.

41 Blechschmidt I, Matter A, Preusser F, et al.2009. Monsoon triggered formation of Quater-nary alluvial megafans in the interior of Oman.Geomorphology 110:128–139.

42 Anton D. 1984 Aspects of geomorphologicalevolution: paleosols and dunes in Saudi Arabia.In: Jado AR, Zotl JG, editors. Quaternary Periodin Saudi Arabia. II. Sedimentological, hydro-geological, hydrochemical, geomorphological,and climatological investigations of westernSaudi Arabia. Vienna: Springer Verlag. p 275–296.

43 Rosenberg TM, Preusser F, Blechschmidt I,et al. 2012. Late Pleistocene palaeolake in theinterior of Oman: a potential key area for thedispersal of anatomically modern humans out-of-Africa? J Quaternary Sci 27:13–16.

44 McClure HA. 1976. Radiocarbon chronologyof late Quaternary lakes in the Arabian Desert.Nature 263:755–756

45 Schulz E, Whitney JW. 1986. Upper Pleisto-cene and Holocene lakes in the An Nafud,Saudi Arabia. Hydrobiologia 3:175–190.

46 Immenhauser A, Dublyansky YV, Verwer K,et al. 2007. Textural, elemental, and isotopiccharacteristics of Pleistocene phreatic cavedeposits (Jabal Madar, Oman). J Sediment Res77:68–88.

47 Clark ID, Fontes JC. 1990. Paleoclimaticreconstruction in northern Oman based on car-bonates from hyperalkaline groundwaters. Qua-ternary Res 33:320–336.

48 Radies D, Hasiotis ST, Preusser F, et al.2005. Paleoclimatic significance of Early Holo-cene faunal assemblages in wet interdunedeposits of the Wahiba Sand Sea, Sultanate ofOman. J Arid Environ 62:109–125.

49 Neff U, Burns SJ, Mangini A, et al. 2001.Strong coherence between solar variability andthe monsoon in Oman between 9 and 6 kyrago. Nature 411:290–293.

50 Fleitmann D, Burns SJ, Mudelsee M, et al.2003. Holocene forcing of the Indian monsoonrecorded in a stalagmite from southern Oman.Science 300:1737–1739.


51 Parker AG, Preston G, Walkington H, et al.2006. Developing a framework of Holocene cli-matic change and landscape archaeology forsoutheastern Arabia. Arab Archaeol Epigr17:125–130.

52 Engel M, Bruckner H, Pint A, et al. n.d. Theearly Holocene humid period in NW Saudi Ara-bia: sediments, microfossils and palaeo-hydro-logical modelling. Quatern Int. In press.

53 Arz HW, Lamy F, Patzold J. 2006. A pro-nounced dry event recorded around 4.2 ka inbrine sediments from the northern Red Sea.Quaternary Res 66:432–441.

54 McLaren SJ, Al-Juaidi F, Bateman MD, et al.2009. First evidence for episodic flooding eventsin the arid interior of central Saudi Arabia overthe last 60 ka. J Quaternary Sci 24:198–207.

55 Goudie AS, Colls A, Stokes S, et al. 2000. LatestPleistocene and Holocene dune construction atthe north-eastern edge of the Rub Al Khali, UnitedArab Emirates. Sedimentology 47:1011–1021.

56 Ivanochko TS, Ganeshram RS, Brummer GJA,et al. 2005. Variations in tropical convection as anamplifier of global climate change at the millen-nial scale. Earth Planet Sci Lett 235:302–314.

57 Fernandes CA. 2009. Bayesian coalescent infer-ence frommitochondrial DNA variation of the col-onization time of Arabia by the hamadryas ba-boon (Papio hamadryas hamadryas). In: PetragliaMD, Rose JI, editors. The evolution of human pop-ulations in Arabia: paleoenvironments, prehistoryand genetics. Netherlands: Springer. p 89–102.

58 Winney BJ, Hammond RL, Macasero W,et al. 2004. Crossing the Red Sea: phylogeogra-phy of the hamadryas baboon, Papio hamadryashamadryas. Mol Ecol 13:2819–2827.

59 Whybrow PJ, Hill A, editors. 1999. Fossil verte-brates of Arabia. NewHaven: Yale University Press.

60 Thomas H, Geraads D, Janjou D, et al. 1998.First Pleistocene faunas from the Arabian Pen-insula: An Nafud Desert, Saudi Arabia. CR AcadSci IIA 326:145–152.

61 O’Regan HJ, Turner A, Bishop LC, et al.2011. Hominins without fellow travellers? Firstappearances and inferred dispersals of Afro-Eurasian large-mammals in the Plio-Pleisto-cene. Quaternary Sci Rev 30:1343–1352.

62 McClure HA. 1984. Late Quaternary paleo-environments of the Rub al Khali. Ph.D. disser-tation. London: University College London.

63 Drechsler P. 2009. The dispersal of the Neo-lithic over the Arabian Peninsula. BAR 1969.Oxford: Archaeopress.

64 Parr PJ, Zarins J, Ibrahim M, et al. 1978. Pre-liminary report on the second phase of the North-ern Province survey 1397/1977. Atlal 2:29–30.

65 Zarins J, Murad AJ, Al-Yish KS. 1981. Thecomprehensive archaeological survey program.The second preliminary report on the south-western province. Atlal 5:9–42.

66 Whalen N, Ali J, Sindhi H, et al. 1986. ALower Pleistocene site near Shuwayhitiyah innorthern Saudi Arabia. Atlal 10:94–101.

67 Chauhan PR. 2009. Early Homo occupationnear the Gate of Tears: examining the paleoanthro-pological records of Djibouti and Yemen. In: HoversE, Braun DR, editors. Interdisciplinary approachesto the Oldowan. Netherlands: Springer. p 49–59.

68 Amirkhanov H. 2008. Cave al-Guza the multi-layer site of Oldowan in South Arabia. Moscow:TAUS.

69 Whalen N, Davis P, Pease D. 1989. EarlyPleistocene migrations into Saudi Arabia. Atlal12:59–75.

70 Whalen N, Zoboroski M, Schubert K. 2002.The Lower Palaeolithic in southwestern Oman.Adumatu 5:27–34.

71 Jagher R. 2009. The Central Oman Paleo-lithic Survey: recent research in southern Ara-bia and reflection on the prehistoric evidence.In: Petraglia MD, Rose JI, editors. The evolu-tion of human populations in Arabia: paleoen-vironments, prehistory and genetics. Nether-lands: Springer. p 139–150.

72 Lahr MM. 2010. Saharan corridors and theirrole in the evolutionary geography of ‘‘Out ofAfrica 1.’’ In: Fleagle JJ, Shea JJ, Grine FE,Baden AL, Leakey RE, editors. Out of Africa 1:the first hominin colonization of Eurasia. NewYork: Springer. p 27–46.

73 Zarins J, Whalen N, Ibrahim M, et al.1980. Comprehensive archaeological surveyprogram: preliminary report on the centraland southwestern provinces survey, 1979 Atlal4:9–36.

74 Whalen N, Sindhi H, Wahida G, et al. 1983.Excavation of Acheulean sites near Saffaqah inal-Dawadami (1402/1982). Atlal 7:9–21.

75 Whalen N, Siraj- Ali J, Davies W. 1984.Excavations of Acheulean sites near Saffaqah,Saudi Arabia. Atlal 8:43–58.

76 Petraglia MD, Drake N, Alsharekh A. 2009.Acheulean landscapes and large cutting toolassemblages in the Arabian Peninsula. In: Pet-raglia MD, Rose JI, editors. The evolution ofhuman populations in Arabia: paleoenviron-ments, prehistory and genetics. Netherlands:Springer. p 103–116.

77 Whalen N, Killick A, James N, et al. 1981.Saudi Arabian archaeological reconnaissance1980: preliminary report on the western prov-ince survey. Atlal 5:43–58.

78 Whalen N, Siraj-Ali J, Sindhi H, et al. 1988.A complex of sites in the Jeddah-Wadi FatimaArea. Atlal 11:77–87.

79 Shipton C, Petraglia MD. 2010 Inter-conti-nental variation in Acheulean bifaces. In: Nor-ton C, Braun R, editors. Asian palaeoanthropol-ogy. London: Springer. p 49–56.

80 Rose JI. 2004. The question of Upper Pleis-tocene connections between East Africa andSouth Arabia. Curr Anthropol 45:551–555.

81 Crassard R. 2009. The Middle Paleolithic ofArabia: the view from the Hadramawt region,Yemen. In: Petraglia MD, Rose JI, editors. Theevolution of human populations in Arabia: pale-oenvironments, prehistory and genetics. Nether-lands: Springer. p 151–168.

82 Whalen N, Schatte K. 1997. Pleistocene sitesin southern Yemen. Arab Archaeol Epigr 8:1–10.

83 McClure HA. 1994. A new Arabian stone toolassemblage and notes on the Aterian industryof North Africa. Arab Archaeol Epigr 5:1–16.

84 Scerri E. n.d. The Aterian in Arabia: a re-analysis of the evidence. Proc Sem ArabianStudies. In press.

85 Macchiarelli R. 2008. From Africa to Asiathrough Arabia: models, predictions and wit-nesses of first phases of human settlement. Pre-sented at the Conference on the First GreatMigrations of Peoples, UNESCO, Paris.

86 Inizan ML, Ortlieb L. 1987. Prehistoire dansla region de Shabwa au Yemen du Sud (R.D.P.Yemen). Paleorient 13:5–22.

87 McBrearty S. 1999. Earliest stone tools fromthe Emirate of Abu Dhabi, United Arab Emi-rates. In: Whybrow PJ, Hill A, editors. Fossilvertebrates of Arabia. New Haven: Yale Univer-sity Press. p 373–388.

88 Rose JI, Petraglia MD. 2009. Tracking theorigin and evolution of human populations inArabia. In: Petraglia MD, Rose JI, editors. Theevolution of human populations in Arabia: pale-oenvironments, prehistory and genetics. Nether-lands: Springer. p 1–12.

89 Petraglia MD. 2011. Trailblazers across Ara-bia. Nature 470:50–51.

90 Garrard A, Harvey CPD, Switsur VR. 1981.Environment and settlement during the UpperPleistocene and Holocene at Jubba in theGreat Nefud, Northern Arabia. Atlal 5:137–148.

91 Maher LA. 2009. The Late Pleistocene ofArabia in relation to the Levant. In: Petraglia MD,Rose JI, editors. The evolution of human popula-tions in Arabia: paleoenvironments, prehistoryand genetics. Netherlands: Springer. p 187–202.

92 Rose JI, Usik VI. 2009. The ‘‘Upper Paleo-lithic’’ of south Arabia. In: Petraglia MD, RoseJI, editors. The evolution of human populationsin Arabia: paleoenvironments, prehistory andgenetics. Netherlands: Springer. p 169–185.

93 Edens CA. 2001. A bladelet industry insouthwestern Saudi Arabia. Arab ArchaeolEpigr 12:137–142

94 Adams RM, Parr PJ, Ibrahim M, et al. 1977.Saudi Arabian archaeological reconnaissance,1976: preliminary report on the first phase ofthe comprehensive archaeological survey pro-gram. Atlal 1:21–40.

95 Crassard R. 2008. The ‘‘Wa’shah method’’:An original laminar debitage from Hadra-mawt, Yemen. Proc Sem Arabian Studies38:3–14.

96 Charpentier V, Inizan ML. 2002. Fluting inthe Old World: the Neolithic projectile points ofArabia. Lithic Tech 27:39–46.

97 Uerpmann HP, Potts DT, Uerpmann M.2009. Holocene (re-)occupation of easternArabia. In: Petraglia MD, Rose JI, editors. Theevolution of human populations in Arabia:paleoenvironments, prehistory and genetics.Netherlands: Springer. p 205–214.

98 Boivin N, Blench R, Fuller DQ. 2009.Archaeological, linguistic and historical sourceson ancient seafaring: a multidisciplinary approachto the study of early maritime contact andexchange in the Arabian Peninsula. In: PetragliaMD, Rose JI, editors. The evolution of human pop-ulations in Arabia: paleoenvironments, prehistoryand genetics. Netherlands: Springer. p 251–278.

99 Fedele FG. 2009. Early Holocene in thehighlands: data on the peopling of the East-ern Yemen plateau, with a note on thePleistocene evidence. In: Petraglia MD, RoseJI, editors. The evolution of human popula-tions in Arabia: paleoenvironments, prehis-tory and genetics. Netherlands: Springer. p215–236.

VVC 2012 Wiley Periodicals, Inc.


Documents

The prehistory of the Arabian peninsula: Deserts, dispersals, and