PDF ( .pdf )

PAGE 1

LateNeandertalsinSoutheasternIberia:Simadelas PalomasdelCabezoGordo,Murcia,SpainMichaelJ.Walkera,JosepGibertb,MarianoV.Lo ´ peza,A.VincentLombardic,AlejandroPe ´ rez-Pe ´ rezd,JosenaZapataa, JonOrtegaa,ThomasHighame,AlistairPikef,Jean-LucSchwenningere,Joa ˜ oZilha ˜ of,andErikTrinkausg,1aA ´ readeAntropologíaFísica,DepartamentodeZoologíayAntropologíaFísica,FacultaddeBiología,CampusUniversitariodeEspinardo,Universidad de Murcia,30100Murcia,Spain;bInstitutPaleontolo ´gic‘‘Dr.M.Crusafont,’’Carrerdel’EscolaIndustrial23,08201Sabadell,Barcelona,Spain;c2584Blossom Lane,NewCastle,PA16105;dSeccio ´ndeAntropología,DepartamentodeBiologíaAnimal,FacultaddeBiología,UniversitatdeBarcelona,AvenidaDiagonal 645,08028Barcelona,Spain;eResearchLaboratoryforArchaeologyandtheHistoryofArt,UniversityofOxford,OxfordOX13QY,UnitedKingdom;fDepartmentofArchaeologyandAnthropology,UniversityofBristol,43WoodlandRoad,BristolBS81UU,UnitedKingdom;andgDepartmentof Anthropology,WashingtonUniversity,SaintLouisMO63130 ContributedbyErikTrinkaus,November7,2008(sentforreviewOctober17,2008)MiddlePaleolithicfossilhumanremainsfromtheSimadelasPalomas insoutheasternIberia(datedto < 43,000–40,000calendaryears beforepresent)presentasuiteofderivedNeandertaland/orretained ancestralmorphologicalfeaturesinthemandibularsymphysis,mandibularramus,dentalocclusalmorphology,anddistalhandphalanx. Thesetraitsarecombinedwithvariationinthemandibularcorpus, discretedentalmorphology,toothrootlengths,andanteriordental sizethatindicateafrequencydifferencewithearlierIberianandmore northernEuropeanNeandertals.ThePalomasNeandertalstherefore conrmthelatepresenceofNeandertalsassociatedwiththeIberian persistenceoftheMiddlePaleolithic,butsuggestmicroevolutionary processesand/orpopulationcontactwithcontemporaneousmodern humanstothenorth.dentition mandible MiddlePaleolithic postcraniaIthasbecomeapparentthatthetransitionfromaEurope populatedbyNeandertalstooneoccupiedbyearlymodern humansduringmarine(oxygen)isotopestage(MIS)3consisted ofawestwardspreadofmodernhumans,variablyabsorbingat leastsomelocalNeandertalpopulations(1,2).However,the detailedaspectsofthispopulationprocessremainobscureasa resultofthescarcityofwell-datedlateNeandertalsandearly modernhumans,despitetherecentdirectdatingofseveral specimensfrombothsamples(3–9).Ifwearetounderstandthe paleobiogeographyofthisprocess,andhencethepotential patternsofinteractionsbetweenthesetwomorphologically definedhumangroups,thenitisessentialtodocumentthe biologyofboththeearliestmodernhumansandthelatest Neandertals. IthasbeenrecognizedforsometimethattheMiddlePaleolithic,generallypresumedinEuropetohavebeentheproductof Neandertals,persistedsubstantiallylongerinIberiasouthofthe Pyrenees(southofthe‘‘EbroFrontier’’)thanelsewherein Europe,to 34,000( 34ka)calendaryearsbeforepresent(cal BP)( 30karadiocarbonyearsbeforepresent(14CBP)](10; supportinginformation(SI) Appendix ,Fig.1).Eventhoughthere areIberianNeandertalremainsthathavebeenreferredtothis age(11),theirpurportedlateagehasbeenplacedindoubt(12). Atpresent,themostrecent,securelydated,diagnosticNeandertalfossilfromtheregionistheOliveira1middlemanual phalanxat 43,500calBP( 39ka14CBP)(13),aboutthesame ageasthenorthernSpainElSidro ´nremains(7).Itisinthis contextthatwepresentaseriesofNeandertalremainsfromthe upperlevelsoftheSimadelaPalomasinsoutheasternSpain. ResultsContextofthePalomasHumanRemains.TheSimadelasPalomas (Rock-DoveHole)isakarsticshaft(37°47 59 N,0°53 45 W) inthePermo-Triassicmarbleofahill(CabezoGordo)inTorre Pachecotownship,Murcia,reaching312ma.s.l.,overlookingthe MarMenorcoastallagoonoftheMediterraneanSea( SIAppendix ,Figs.1and2).Themainshaftis18mdeep,opening beneathoverhangingrockat 123ma.s.l.( SIAppendix ,Figs.3 and4).Theshaft’sbrecciatedcontentswerelargelyemptiedby 19thcenturyminers,wholeftasedimentcolumndownoneside andscatteredfossiliferousrubbleonthehillside.Systematic collectionofdisturbedremainsandexcavationoftheupper brecciadepositswerebegunbyM.J.W.andthelateJ.Gibert afterdiscoveryofacrushedfacialskeleton(Palomas1)fromthe uppermostbrecciabyJ.C.Blancoin1991(14,15).Fossilhuman remainshavebeencollectedfromtheminerubble(1992–1999) andexcavatedinsitu(1994–present)( SIAppendix ,Table1). Anumberoftheinsituhumanremainscomefromabove, within,andslightlybelowafusiformlensofdark-graysediment (burnt,accordingtoX-raydiffractionandfluorescenceanalyses).Itattainedamaximumthicknessof 20cmintheangleof anL-shapedexcavationareaintheUpperCutting,coveringa thin,oblongmarbleslab( 40 30 10cm)betweenlevels2k and2m.Thelenspeteredout,bothneartheopenshaftand belowtheentrancewhereitcoveredpartofthefootofa downwardandinwardlysteeplysloping(30°–40°)e ´boulisor screeofmarbleblocks 50kg.Amoreconsistentandwidespreadlowerlayerofdark-graysedimentliesbeneathlevel2l, fadingawaybesidetheopenshaft(Fig.1; SIAppendix ,Figs.5 and6). Onlyhumanbonesandteethfromsedimentspostdatingthe e ´boulis(fromlevelsnodeeperthan2l)aredealtwithhere. Althoughtheselatersedimentscontainfewlargerocksandwere laiddownhorizontally,theyoftenhaveacoarsetextureimplying thattheywerewashedfromthehillsideintothecaveanddown thesteepinteriorscreeslope;theslopehaddevelopedneitheran erodedsurfacenoracalcretecrust,incontrasttothethincalcite depositonthehillsidethateventuallysealedoffthee ´boulis. TheselatersedimentscontainedMiddlePaleolithicartifacts, animalbones(someofwhichareburnt),andpollenindicativeof mildclimaticconditions(14,16,17).Becauseofcarbonate precipitation,thee ´boulisisinpartcementedintoirregular brecciatedconglomeratemasses,inwhichthereareMiddle Paleolithicartifacts,faunalremains,andvariablyarticulatedand crushedhumanremains.ThehumanfossilsincludethePalomas 92and96partialskeletons,thePalomas93to95teeth,andthe remainsofatleast2moreindividuals,includingcraniawith mandibles. Authorcontributions:M.J.W.,J.G.,andE.T.designedresearch;M.J.W.,J.G.,M.V.L.,A.V.L., A.P.-P.,J.Zapata,J.O.,T.H.,A.P.,J.-L.S.,J.Zilha ˜o,andE.T.performedresearch;M.J.W., M.V.L.,J.Zapata,J.O.,T.H.,A.P.,J.-L.S.,andE.T.analyzeddata;andM.J.W.,T.H.,A.P.,J.-L.S., J.Zilha ˜o,andE.T.wrotethepaper. Theauthorsdeclarenoconictofinterest.1Towhomcorrespondenceshouldbeaddressed.E-mail:trinkaus@artsci.wustl.edu. Thisarticlecontainssupportinginformationonlineat www.pnas.org/cgi/content/full/ 0811213106/DCSupplemental . ©2008byTheNationalAcademyofSciencesoftheUSAwww.pnas.org cgi doi 10.1073 pnas.0811213106PNASDecember30,2008vol.105no.5220631–20636 ANTHROPOLOGY

PAGE 2

ChronologicalAgeofthePalomasHumanRemains.Theageofthe humanfossilsofconcernhereisprimary,becausetheissueisthe morphologyofthelatestNeandertalsinEurope.ThePalomas artifactsareallMiddlePaleolithic,madeonretouchedflakesof flint,quartz,rockcrystal,andmarble(14–16,18).Givencurrent reliabledatesforIberianMiddlePaleolithicassemblages(10, 19),thisindicatesanageof 34kacalBP( 30ka14CBP)for thedeposits. CollagenpreservationispoorinthePalomashumanremains, anddirect14Cdatingofthehumanboneshasnotbeenpossible. However,acombinationofdatingtechniquespermitschronologicalcontroloftheUpperCutting(see SIAppendix ,Figs.7and 8,forthestratigraphicpositionsofthesamples). AburntfaunalbonedirectlyadheringtothePalomas59 mandibleinlevel2fprovidedadateof34,450 60014CBP (OxA-10666)[95%CI:40,950–37,622calBPusingtheCariaco0614Ccurve(20)].Aburntlagomorphbonefromthedeeper level2lprovidedastatisticallyidenticalageof35,030 27014C BP(OxA-15423)(95%CI:40,986–38,850calBP).Giventhe naturesofthedatingspecimensandthemoderatelyhighC:N atomicratiosforthesesamples(albeitanexpectedelevation giventheburning),itislikelythattheseagesfallclosertothe upperlimitsofthe95%confidenceintervalsforthesedates( SI Appendix ,SectionIII).Thisinterpretationofthesedatesfrom theuppersedimentfillissupportedbytheassociatedpalynology (17),whichshowsrelativelytemperateconditions,probably duringGIS9andthereforeslightlybeforetheonsetofthe severelycoldHeinrich4oscillation 40kacalBP(21). Toassessfurthertheagesofthesedepositsandthestratigraphicallyoldere ´boulisdeposits,U-series(LA-ICP-MS)dates wereobtainedfrom3bones( SIAppendix ,SectionIV).Adate of43,800 750calBP(APSLP4)isfromafaunalspecimenfrom level2i;itisslightlyearlierthanthelikely14Ccalibratedranges fromlevels2fand2l.Twomucholderandstatisticallysimilar U-seriesdatescomefromthestratigraphicallydeeper,steeply slopinge ´boulis:oneonaPalomas96metacarpalfromlevel2e of54,000 3,850calBP(APSLP1)andtheotheronafaunal bonefromlevel2lof51,000 1,250calBP(APSLP6).These U-seriesestimatespresumethatthesampleswereclosedsystems,whichcannotbeverified.Theyshouldthereforeberegardedonlyascorroboratingthe14Candpaleoclimatological assessmentsoftheageoftheuppersedimentfillandtheearlier ageofthee ´boulislevel,includingsomemixedmaterialinthe burntdark-graysedimenthorizon. Inaddition,asedimentsample(X2509)fromthetopoflevel 2kdirectlyoverlyingthemarbleslabandhencebelowthe dark-graysedimentlevelinthenortheasterncorneroftheUpper Cuttingwasdatedbyusingopticallystimulatedluminescence (OSL)( SIAppendix ,SectionV).Thesampleprovidedanage estimateof54,700 4,700calBPforthesesediments.This determinationincreasesconfidenceintheU-seriesdatesfor stratigraphicallysimilarspecimensandhenceintheageofthe stratigraphicallyyoungerdepositsdatedby14C. Thesedatingassessmentsthereforecombinetoindicateanage fortheuppersedimentfill 40kacalBPbutpossiblyslightly older(i.e., 43kacalBP).ThesubsampleofthePalomashuman remainsfromthisportionofthesedimentincludes63elements, 54or85.7%ofwhichareisolatedteethortoothfragments.They makethesePalomasNeandertalsthemostrecent,andlargest, sampleofsouthernIberianlateNeandertalscurrentlyknown. TheotherNeandertalscloseinagearetheInitialUpper PaleolithiconesfromSpy(9),andprobablySaint-Ce ´saire, Arcy-Renne,andVindijaG1(22–24).ThesePalomasfossilsare alsoapproximatelythesameageastheearliestmodernhumans inEurope(4,25),albeitattheotherendofEurope.Iftheearliest phasesoftheAurignacianwereindeedmadebymodernhumans (26,27),thenthePalomasremainsshouldoverlapintimewith modernhumansascloseasthenorthernPyrenees(10).ThePalomasHumanRemains.ThePalomasfossilhumanremains thereforeconsistof3samples.Therearetheundatedand isolatedremainsdiscoveredintheminers’rubble.Therearethe partialface,partialskeletons,andisolatedremainsfromthe brecciatede ´boulis.Andtherearethe63isolatedremainsfrom theexcavateddepositsatorabovethelevelsdatedto 40–43ka calBP. Thediagnosticremainsfromthefirst2samplescanallbe attributedtoLatePleistoceneNeandertals.Fortheremainsout ofcontext,therelevantaspectsincludesupraorbitaltoruspresence(Palomas11,12,and62),retreatingmandibularsymphyses andlateralcorpusthickness(Palomas6and23),incisorand molarocclusalmorphologyandincisorrootlength(Palomas24 and50),manualmiddlephalanxbreadth(Palomas65),and femoraldiaphysealshape(Palomas52).Forthebrecciated remains,theisolatedteethappearundiagnostic,andPalomas96 andtheotherassociatedskeletonsarestilllargelyinbreccia.Yet, thePalomas1mandiblehasaretreatingsymphysis,aretromolar space,aprominentcoronoidprocess,andanasymmetrical mandibularnotch.ThePalomas92partialskeletonalignswith theNeandertalsindistalhumeral,proximalulnarandfemoral diaphysealmorphology,aswellasinferredbodyproportions. AlthoughmanyoftheisolatedremainsfromPalomasare Fig.1. SchematicdrawingofthecurrentprolesoftheUpperCutting(see SIAppendix ,Figs.5and6).( A )Thee ´boulisabovenorthwesterncornerofthe excavatedcutting.( B )Levels2m-2obrecciacontaininghumanbones.( C )Projectionofthee ´boulisscreeslope,whichislessperceptibleinthisprolethanithad beeninnowremovedsectionsparalleltoitintheforeground.( D )Uppermostlimit(levels 2h-2i)oflensofburntsedimentmainlyinthenortheasternareaof thecutting.( E )Lowerlimitofburntashysediment(levels2m-2o)inthenorthernandeasternareaofthecutting.20632www.pnas.org cgi doi 10.1073 pnas.0811213106 Walker etal.

PAGE 3

undiagnosticastohumangroupinaLatePleistocenecontext, therearesufficientindicatorstoaligntheseremainswiththe Neandertals.Innocasedoanyoftheremainsexhibituniquely derivedcharacteristicsofmodernhumans(cf.,ref.28). GiventheconcernwithlateNeandertalpaleobiology,the considerationsherearelimitedtotheremainsfoundinsituator abovethe40–43kacalBPlevels.Thecomparativesamples consistofMIS5–3westernEurasianNeandertals,MIS5Middle Paleolithicearlymodernhumans(MPMH),circum-MediterraneanEarlyUpperPaleolithicmodernhumans(EUP)( 33ka calBP),andwesternEurasianMiddleUpperPaleolithicmodern humans(MUP)( 33–24kacalBP).TheMandibles.Fourpartialmandibleswerefoundinsituinthe youngerlevels(Palomas49,59,80,and88)(Fig.2).Palomas59 isaleftcorpuslackingthefullsymphysis,andtheothersare variablycompleteimmaturespecimens. ThepreservedboneofPalomas59indicatesthatithada relativelyverticalsymphysisbutnoprominentdevelopmentof eitheratubersymphyseosorlateraltubercles[probablymentum osseumrank3(29)].TheimmaturePalomas49hasasimilaror moreretreatingsymphysealprofile.Thedentalarcadeisonly intactforPalomas49,andaswithotherveryyoungNeandertals (30),itsbi-dc1externalarcadediameter(35.2mm)isbeyond thoseofsimilarlyagedearlymodernhumans(Fig.3),including thatoftheearlierAurignacianLaQuina-Aval4mandible( 30.0 mm).Palomas59hadaretromolarspace,andPalomas80hasa prominentcoronoidprocessandanasymmetricalmandibular notch,butanopenmandibularforamen.Mostofthesefeatures alignthemprincipallywiththeNeandertalsamongMIS5–3 humans(8). Atthesametime,thementalforaminaoftheadultPalomas 59(atP4M1)andtheinfantPalomas49and88(atdm1)are moderatelymesial(8,31),butthejuvenilePalomas80mandible hasanunusuallymesialmentalforamen,becauseitwasdistinctly mesialoftheP3P4break.Thereislittledifferenceinlateral mandibularcorpusheightbetweenNeandertalsandearlymodernhumans(Kruskal–Wallis P 0.251),butthereisasignificant ( P 0.0002)differenceinbreadth( SIAppendix ,Table4).The otherPalomasmandibles(1,6,and23)arewithotherNeandertalsincorpusbreadth(Fig.4).Palomas59,however,isamong theEUPandMUPmodernhumans;itscorpusbreadthis2.21 standarddeviationsfromtheNeandertalmean(withtheKebara 2highoutliertrimmed,1.97standarddeviationswithKebara2 included).TheDentitions.Theabundanceofisolatedteeth,plusthosein Palomas59and80,makeseveralobservationspossible.Allfour ofthemaxillarycentralincisors(I1s:Palomas34,73,79,and90) exhibitmoderatetomarkedlabialconvexity,andthe3lingually preservedoneshavelargemarginalridgesandlingualtubercles (Fig.5).The2maxillarycanines(C1s:Palomas35and74)have verysmalllingualtubercles,andPalomas35haslittleifany shoveling(Fig.5).Ofthe4P4s(Palomas57,59,78,and87),3 eachexhibitatransversecresteventhoughonlythePalomas59 crestispronounced(Fig.6).Allofthemhaveamesially displacedmetaconid,and3haveextralingualcusps.Onlyoneof theP4s,Palomas59,haslingualasymmetry,butitprobably lackedextralingualcusps.Yet,therearenoconsistentassociationsbetweenthesetraitsacrossthe4PalomasP4s.Ofthe5first andsecondlowermolars(M1sandM2s:Palomas29,80,and84, withoneeachfromPalomas59),allexhibitanteriorfoveabut3 lackmidtrigonidcrests(Fig.7).TheoneM2(Palomas36)hasa skewedprofileandcentrallyplacedcusps. MostofthesedentalocclusaltraitsoccurinboththeNeandertalsandotherPleistocene(andrecent)humansamples,and Fig.2. OcclusalviewoftheimmaturePalomas(SP)49mandibularcorpusand lateralviewsofthePalomas59,80and88mandibles.Palomas59isin norma lateralis ,andPalomas80and88areintheplanesoftheirlateralcorpori.Scale incentimeters. Fig.3. Bivariateplotofexternalbi-deciduouscanine(dc1)dentalarcade breadthversusdevelopmentalageforimmatureLatePleistocenehuman mandibles.Symbols:blackdiamond,Palomas49;graycircles,Neandertals; blacksquares,MPMH;blacktriangle,EUPmodernhuman(LaQuina-Aval4); opentriangles,MUPmodernhumans.Agesarebasedondentalcalcication relativetoextanthumans. Fig.4. Bivariateplotofmandibularcorpusbreadthversusheightatthe mentalforamen,forLatePleistocenematuremandibles.SymbolsasinFig.2; numberedsymbolsareforPalomas1,6,23,and59,therst3ofwhichare geologicallyolderthanPalomas59orundated.Theearlymodernhumans withhighcorpusbreadthsareQafzeh9andSkhul4(MPMH),NazletKhater2 (EUP),andCro-Magnon1(MUP);thehighNeandertaloutlierisKebara2.Walker etal. PNASDecember30,2008vol.105no.5220633ANTHROPOLOGY

PAGE 4

notallNeandertalsexhibitallofthem(32).However,theseare alltraitsthatoccurinhighfrequenciesamongtheNeandertals, andNeandertalsinparticularhavehighfrequencies,compared withotherPleistocenesamples,ofthesetraitsoccurringin combination.OfthesePalomasteeth,onlytheI1sandtheM2exhibitthefullsuiteoffeaturesincreasinglyconsideredcharacteristicoftheNeandertals. Rootlengths,especiallyofanteriorteeth,havebeenshownto largelydifferentiateNeandertalsandUpperPaleolithichumans (33);additionaldata( SIAppendix ,Table5)indicateasignificant comparativesampledifferenceinrootlengthsforallbuttheI2. Ofthe13Palomasanteriorteethprovidingrootlengths,4are belowtheNeandertalranges,4are 2fromtheNeandertal means,and77%arebelowtheNeandertalmeans. ThePalomas59M1exhibitssupraradiculartaurodontism,and theM2haspronouncedradicularendotaurodontism,thelatter especiallybeingcharacteristicoftheNeandertals(34).Yet,the Palomas29M2lacksanypulpchamberexpansion. Neandertalsandearlymodernhumanshavesimilarpostcaninedentaldimensions(35),buttheformerhavegreaterI1and I2labiolingualcrowndiameters(Kruskal–Wallis P 10 6for each)butcontrastlessinC1breadths(Kruskal–Wallis P 0.025) ( SIAppendix ,Table6).OneoftheinsituPalomasI1s(Palomas 21)and3ofthe5C1s(Palomas26,54,59)arebelowthe Neandertalrange,andtheremainderofthePalomasanterior mandibularteethareatorbelowtheNeandertalmeans[2-tailed Wilcoxon P 0.047(I1),0.312(I2),0.013(C1)].ThePostcrania.Inthepostcrania,despitemultipleelements( SI Appendix ,Table1),only1diagnosticboneisstratigraphically secureinthemorerecentdeposits,thePalomas28distalhand phalanx.Ithasabroad,roundedapicaltuberositylackingungual spines(Fig.8),anarchaic Homo configuration(36,37).Neandertaldistalphalangealbreadthsaresignificantlyabsolutely broaderthanthoseofalmostallearlymodernhumans(Fig.8; Kruskal–Wallis P 0.00003)( SIAppendix ,Table7),andmost ofthemarebroaderrelativetophalangeallengthdespitethe relativelylongerdistalphalangesoftheNeandertals(38).The distalbreadthofthePalomas28phalanx(9.7mm)fallsabsolutelyandrelativelyamongthosearchaichumans(Fig.8and SI Appendix ,Table7). DiscussionNeandertalAffinities.Theseconsiderationsoftheinsituhuman remainsfromtheupperlevelsoftheSimadelasPalomas confirmthattheyarebestseenaslatesouthwesternEuropean Neandertals.Thereisasuiteoffeatures,includingmandibular symphysealconfiguration,ramalshape,dentalocclusalmorphology,andmanualdistalphalanxshape,thatplacesthemwith archaic Homo andseparatefromearlymodernhumans.Moreover,theI1sandtheP4s,M2,M1sandM2sexhibitapparently derivedNeandertalocclusaltraitsorcombinationsoftraits. Otherretainedplesiomorphousaspectslostamongearlymodern humansorautapomorphoustraitsoftheNeandertals(28)are notpreservedorevidentonlyonundatedorolderPalomas fossils.Theseconsiderationsshouldnonethelessbesufficientto Fig.6. OcclusalviewsofPalomas(SP)mandibularsecondpremolars(P4s). Scaleinmillimeters. Fig.7. OcclusalviewsofPalomas(SP)molars.SP36,maxillaryM2;SP59, mandibularM1andM2;SP80,mandibularM2;SP29,mandibularM2;SP84, mandibularM1.Scaleinmillimeters. Fig.8. Bivariateplotofdistalphalangealbreadthversusarticularlengthfor Palomas28andLatePleistocenesamples.SymbolsasinFig.2.Givenuncertaintiesindigitassignmentforisolatedray2–4distalphalanges,valuesare averagedforthoseindividualpreservingmultipledistalphalangestoprovide anindividualvalue. Fig.5. LingualviewsofPalomas(SP)maxillarycentralincisors(I1s)(SP34,79 and90)andmaxillarycanines(C1s)(SP35and74).Scaleinmillimeters.20634www.pnas.org cgi doi 10.1073 pnas.0811213106 Walker etal.

PAGE 5

confirmthatat 40–43kacalBP,insoutheasternIberia,the MiddlePaleolithicpopulationconsistedofNeandertals.MorphologicalVariability.Atthesametime,itisevidentthatthe latePalomasNeandertalsexhibitacomplexmixofNeandertal andmore‘‘modern’’features.Threeofthemandibleshave mentalforaminathataremoderatelymesialforNeandertals,but oneisunusuallyso.DistinctivedentalfeaturessuchaslargeC1lingualtubercles,M1andM2midtrigonidcrest,P4transverse crest,lingualtubercles,mesialmetaconidandlingualasymmetry,longanteriortoothroots,anteriorcrowndimensions, andlowermolartaurodontismarereducedorvariablypresent inthesample.NoneoftheP4shasall4ofthe‘‘Neandertal’’ configurations. ItispossibletofindindividualNeandertalteethormandibles thatexhibitoneormoreofmostthese‘‘non-Neandertal’’ aspects,andbiologicalvariationacrosstheNeandertals(includingtrendsthroughtimeandclinalvariationinspace)hasbeen noted(7,39–42).However,thelevelofvariationinthese featuresinthePalomaslateMiddlePaleolithicsampleisunusual foragroupofNeandertals.Itispossiblethatthesecontrastswith otherNeandertalsrepresent:( i )lateNeandertalgeneticdriftin thedirectionofmodernhumanmorphologythroughisolationby-distanceinthecul-de-sacofsouthernIberia,( ii )anadaptive shifttolocalenvironmentalconstraintsinsomeofthesefeatures, and/or( iii )theproductofgeneflowfromearlymodernhuman populationstotheimmediatenorth. Thefirstexplanationwouldemphasizeregional(perhaps clinal)variationamongtheNeandertals( cf. ,refs.7and40). SecurelydatedearlierMIS4–3Neandertalsfromsouthofthe Pyrenees[e.g.,Banyoles(43,44),CovaNegra(45),Gegant(46), Valdegoba(47),Zafarraya(48),plustheearlierPalomasremains]exhibittheNeandertalpatternincomparableelements (widelateralmandibularcorpori,relativelyposteriormental foramina,largeanteriorteeth,anteriorfoveaeandmidtrigonid crestsonM1–M3,andasymmetricalP4swithmesialmetaconids andlingualtubercles)withminorvariationinmentalforamen positionandacoupleofthedentaltraits.Thesefossilsthusimply thatthePalomasvariationisnotmerelytheresultoflong-term isolationofIberianNeandertalsthroughtheearlierLatePleistocene.Moreover,suchisolationwouldnotnecessarilyexplain thepresenceofautapomorphousmodernhumantraits(e.g.,a narrowmandibularcorpus,reducedC1lingualtubercles,short anteriortoothroots,orsmallanteriordentalcrowns)ina Neandertalpopulation.Thesecondexplanationwouldimply thatsomeofthesemandibularanddentalfeaturesconferan advantageonthesepopulations,althoughvariationinsomeof thefeaturesmaywellbeselectivelyneutral.ThethirdconsiderationwouldrequirethattheearliestAurignacianofthe northernPyreneesandelsewherebetheproductofmodern humans,toprovideageographicallyproximatesourceforgene flow.Itwouldalsoinvokehumanbiologicalcontactacrossthe ‘‘EbroFrontier’’duringatimewhenthereislittleevidencefor culturaldiffusion(10).Inthisscenario,theculturalcontrasts acrossthe‘‘EbroFrontier’’wouldbeduetobehavioralchoices, possiblyecologicallydriven,ratherthanisolationofthesouthern Iberianpopulations. Conclusion ThehumanremainsfromtheSimadelaPalomasinsoutheastern IberiathereforedocumentthepresenceofNeandertals,relativelylateintheMiddlePaleolithic.Theyhelptosubstantiate thattheMiddlePaleolithicoftheregionwastheproductof Neandertals,eventhoughdiagnostichumanremainsassociated withtheverylatestphasesofthistechnocomplexinEurope remainelusive.AtthesametimethatthePalomashumans exhibitasuiteofderivedNeandertalfeaturesandarchaic Homo configurationslongsincelostamongearlymodernhumans,their morphologicalvariationindicatesthattheydeviatefromthe expectedNeandertalrangesofvariation.Thispatternmaybe resultofgeneticdriftinrelativeisolation,directionalchangeor, perhapsmorelikely,populationcontacttothenorth.ACKNOWLEDGMENTS. WethankthelandownersofCabezoGordoandthe managementoftheCabezoGordoS.A.quarryforallowingtheexcavationsto occur,T.RodríguezandJ.L.PoloforprovidingX-raydiffractionanduorescence analyses,andthemanycolleaguesandvolunteersovertheyearsforenthusiasticallyprovidingassistanceandexpertise.ThisworkwassupportedbytheSpanish government(CGL2005/02410/BTE;BOS/2002/02375;PB/98/0405;PB/92/0971),the Murciangovernment(PSH93/52;05584/ARQ/07;CTC/DGC/SPH/063/2001;CCE/ DGC/IPH/SAR0/1998;CCE/DGC/IPH/SAR/1997;CCE/DGC/IPH/SAR/011/1996;CCE/ DGC/IPH/SAR/1995;CCE/DGC/IPH/SAR/1994;PSH/93/52),themunicipalityofTorre Pacheco,andtheEarthwatchInstitute(1994–2001).Analysisofthehuman remainswassupportedbytheSpanishgovernment(CGL2007–60802/BTE) (A.P.-P.)andWashingtonUniversity(E.T.).1.SmithFH,Jankovic ´I,Karavanic ´I(2005)Theassimilationmodel,modernhumanorigins inEurope,andtheextinctionoftheNeandertals. QuatInt 137:7–19. 2.TrinkausE(2007)EuropeanearlymodernhumansandthefateoftheNeandertals. Proc NatlAcadSciUSA 104:7367–7372. 3.SchmitzRW, etal. (2002)TheNeandertaltypesiterevisited:InterdisciplinaryinvestigationsofskeletalremainsfromtheNeanderValley,Germany. ProcNatlAcadSciUSA 99:13342–13347. 4.TrinkausE, etal. (2003)AnearlymodernhumanfromthePes ¸teracuOase,Romania. ProcNatlAcadSciUSA 100:11231–11236. 5.WildEM, etal. (2005)DirectdatingofEarlyUpperPalaeolithichumanremainsfrom Mladec . Nature 435:332–335. 6.TrinkausE(2005)Earlymodernhumans. AnnuRevAnthropol 34:207–230. 7.RosasA, etal. (2006)PaleobiologyandcomparativemorphologyofalateNeandertal samplefromElSidro ´n,Asturias,Spain. ProcNatlAcadSciUSA 103:19266–19271. 8.SocaruA,Dobos ¸A,TrinkausE(2006)EarlymodernhumansfromthePes ¸teraMuierii, BaiadeFier,Romania. ProcNatlAcadSciUSA 103:17196–17201. 9.SemalP, etal. (2008)NewdataonthelateNeandertals:DirectdatingoftheBelgian Spyfossils. AmJPhysAnthropol ,inpress. 10.Zilha ˜oJ(2006)ChronostratigraphyoftheMiddle-to-UpperPaleolithictransitioninthe Iberianpeninsula. Pyrenae 37:7–84. 11.HublinJJ,BarrosoRuizB,MedinaLaraP,FontugneM,ReyssJL(1995)TheMousterian siteofZafarraya(Andalucia,Spain):DatingandimplicationsonthePaleolithicpeoplingprocessesofwesternEurope. CRAcadSci 321:931–937. 12.MichelV, etal. (2003) ElPleistoceneSuperiordellaCuevadelBoquetedeZafarraya , edBarroso-RuízC(ConsejeríadeCultura,JuntadeAndalucía,Sevilla),pp115–155. 13.TrinkausE,MakiJ,Zilha ˜oJ(2007)MiddlePaleolithichumanremainsfromtheGrutada Oliveira(TorresNovas),Portugal. AmJPhysAnthropol 134:263–273. 14.WalkerMJ, etal. (1998)TwoSESpanishMiddlePalaeolithicsiteswithNeanderthal remains:SimadelaPalomasdelCabezoGordoandCuevaNegradelEstrechodelRío Quípar(MurciaProvince). InternetArchaeol 5.Availableathttp://intarch.ac.uk/ journal/issue5/).AccessedOctober24,2008. 15.WalkerMJ, etal. (1999)ExcavationsatnewsitesofearlymaninMurcia,Simadelas PalomasdelCabezoGordoandCuevaNegradelEstrechodelRíoQuípardela Encarnacio ´n. HumEvol 14:99–123. 16.WalkerMJ(2001) AVeryRemotePeriodIndeed ,edsMillikenS,CookJ(Oxbow, Oxford),pp153–159. 17.Carrio ´nJS, etal. (2003)Glacialrefugiaoftemperate,MediterraneanandIbero-North Africanorainsouth-easternSpain:NewevidencefromcavepollenattwoNeanderthalmansites. GlobEcolBiogeog 12:119–129. 18.WalkerMJ, etal. (2004) SettlementDynamicsintheMiddlePalaeolithicandMiddle StoneAge2 ,edConardNJ(Kern,Tu ¨bingen),pp461–511. 19.Zilha ˜oJ,PettittP(2006)OnthenewdatesforGorham’sCaveandthelatesurvivalof IberianNeanderthals. BeforeFarming 2006(3):95–122. 20.HughenK,SouthonJ,LehmancS,BertrandaC,TurnbullJ(2006)Marine-derived14C calibrationandactivityrecordforthepast50,000yearsupdatedfromtheCariaco Basin. QuatSciRev 25:3216–3227. 21.SepulchreP, etal. (2007)H4abrupteventandlateNeanderthalpresenceinIberia. EarthPlanetSciLett 258:283–292. 22.MercierN, etal. (1991)ThermoluminescencedatingofthelateNeanderthalremains fromSaint-Ce ´saire. Nature 351:737–739. 23.BaileySE,HublinJJ(2006)DentalremainsfromtheGrotteduRenneatArcy-sur-Cure (Yonne). JHumEvol 50:485–508. 24.HighamT,BronkRamseyC,Karavanic ´I,SmithFH,TrinkausE(2006)Reviseddirect radiocarbondatingoftheVindijaG1UpperPaleolithicNeandertals. ProcNatlAcadSci USA 103:553–557. 25.Zilha ˜oJ, etal. (2007) RethinkingtheHumanRevolution ,edsMellarsP,BoyleK, Bar-YosefO,StringerC(McDonaldInstituteofArchaeology,Cambridge,UK),pp 249–262. 26.Henry-GambierD,MaureilleB,WhiteR(2004)Vestigeshumainsdesniveauxde l’AurignacienanciendusitedeBrassempouy(Landes). BullMemSocAnthropolParis 16:49–87.Walker etal. PNASDecember30,2008vol.105no.5220635ANTHROPOLOGY

PAGE 6

27.BaileySE,HublinJJ(2005)WhomadetheEarlyAurignacian?Areconsiderationofthe Brassempouydentalremains. BullMemSocAnthropolParis 17:115–121. 28.TrinkausE(2006)ModernhumanversusNeandertalevolutionarydistinctiveness. Curr Anthropol 47:597–620. 29.DobsonSD,TrinkausE(2002)Cross-sectionalgeometryandmorphologyofthemandibularsymphysisinMiddleandLatePleistocene Homo.JHumEvol 43:67–87. 30.MallegniF,TrinkausE(1997)AreconsiderationoftheArchi1Neandertalmandible. J HumEvol 33:651–668. 31.CoqueugniotH(1999)Lecra ˆned’ Homosapiens enEurasie:Croissanceetvariation depuis100,000ans. BritArchaeolRepS 822:1–197. 32.BaileySE(2006)Beyondshovel-shapedincisors:Neandertaldentalmorphologyina comparativecontext. PeriodBiol 108:253–267. 33.BaileySE(2005) CurrentTrendsinDentalMorphologyResearch ,edZadzinskaE (UniversityofLodzPress,Lodz),pp201–210. 34.KallayJ(1970) Krapina1899 1969 ,edMalezM(JugoslavenskeakademijeznanostiI umjetnosti,Zagreb),pp165–176. 35.TrinkausE(2004)in Miscala ´neaenHomenajeaEmilianoAguirre ,edRubioS(Museo Arqueolo ´gicoRegional,Alcala ´deHenares),pp393–398. 36.TrinkausE(1983) TheShanidarNeandertals (Academic,NewYork). 37.RosasA(1985)FalangeshumanasdelaSimadeslosHuesos(CuevaMayor),Sierrade Atapuerca(Burgos).Estudioanatomicoycomparativo. ActasIVCongEspan ˜ol AntropolBiol 4:557–566. 38.VillemeurI(1994) LaMaindesNe ´andertaliens (CentreNationaldelaRecherche Scientique,Paris). 39.SmithFH(1984) TheOriginsofModernHumans ,edsSmithFH,SpencerF(Liss,New York),pp137–209. 40.TrinkausE(1991)LeshommesfossilesdelaGrottedeShanidar,Irak:E ´ volutionet continuite ´parmileshommesarchaïquestardifsduProche-Orient. L’Anthropol 95:535– 572. 41.TrinkausE,ChurchillSE,RuffCB,VandermeerschB(1999)Longboneshaftrobusticity andbodyproportionsoftheSaint-Ce ´saire1Cha ˆtelperronianNeandertal. JArchaeol Sci 26:753–773. 42.Caramelli etal. (2006)AhighlydivergentmtDNAsequenceinaNeandertalindividual fromItaly. CurrBiol 16:630–632. 43.MarotoJ(ed)(1993)LamandíbuladeBanyolesenelcontextdelsfo `ssilshumansdel Pleistoce `. CtrInvestArqueolGirona 13:1–194. 44.Gru ¨nR, etal. (2006)ESRandU-seriesanalysesofenamelanddentinefragmentsofthe Banyolesmandible. JHumEvol 50:347–358. 45.ArsuagaJL, etal. (2007)NewNeandertalremainsfromCovaNegra(Valencia,Spain). JHumEvol 52:31–58. 46.DauraJ, etal. (2005)ANeandertalmandiblefromtheCovadelGegant(Sitges, Barcelona,Spain). JHumEvol 49:56–70. 47.QuamRM, etal. (2001)HumanremainsfromValdegobaCave(Hue ´rmeces,Burgos, Spain). JHumEvol 41:385–435. 48.Barroso-RuízC,LumleyMAde,CaparrosM,VerduL(2003) ElPleistoceneSuperiordel laCuevadelBoquetedeZafarraya ,edBarroso-RuízC(ConsejeríadeCultura,Juntade Andalucía,Sevilla),pp327–387.20636www.pnas.org cgi doi 10.1073 pnas.0811213106 Walker etal.