Chronological reassessment of the Middle to Upper Paleolithic transition and Early Upper Paleolithic cultures in Cantabrian Spain

RESEARCH ARTICLE Ch onological eassessmen o he Middle o Uppe Paleoli hic ansi ion and Ea ly Uppe Paleoli hic cul u es in Can ab ian Spain Ana B. Ma ı ´n-A oyo 1,2 *, Joseba Rios-Ga aiza 3 , Law ence G. S aus 4 , Jenni e R. Jones 1 , Ma co de la Rasilla 5 , Manuel R. Gonza ´lez Mo ales 1 , Michael Richa ds 6 , Jesu ´s Al una 7 , Ko o Ma iezku ena 7 , Da id Ocio 8 1Ins i u o In e nacional de In es igaciones P ehis o ´ icas de Can ab ia, (Uni e sidad de Can ab ia, San ande , Gobie no de Can ab ia), San ande , Spain, 2Le e hulme Cen e o E olu iona y S udies, Depa men o A chaeology, Uni e si y o Camb idge, Camb idge, Uni ed Kingdom, 3Cen o Nacional de In es igacio ´n de la E olucio ´n Humana, Bu gos, Spain, 4Depa men o An h opology, MSC01 1040, Uni e si y o New Mexico, Albuque que, New Mexico, Uni ed S a es o Ame ica, 5Depa amen o de His o ia, Uni e sidad de O iedo, O iedo, Spain, 6Depa men o A chaeology, Simon F ase Uni e si y, Bu naby, B i ish Columbia, Canada, 7Cen o de Conse acio ´n e In es igacio ´n de los Ma e iales A queolo ´gicos y Paleon olo ´gicos de Gipuzkoa, Donos ia/San Sebas ia ´n, Spain, 8Depa men o Ci il and En i onmen al Enginee ing, Impe ial College London, London, Uni ed Kingdom *[email p o ec ed] Abs ac Me hodological ad ances in da ing he Middle o Uppe Paleoli hic ansi ion p o ide a be e unde s anding o he eplacemen o local Neande hal popula ions by Ana omically Mode n Humans. Today we know ha his eplacemen was no a single, pan-Eu opean e en , bu a he i ook place a di e en imes in di e en egions. Thus, local condi ions could ha e played a ole. Ibe ia ep esen s a signi ican mac o- egion o s udy his p ocess. No he n A lan ic Spain con ains e idence o bo h Mous e ian and Ea ly Uppe Paleoli hic occupa- ions, al hough mos o hem a e no p ope ly da ed, hus hinde ing he chances o an adequa e in e p e a ion. He e we p esen 46 new adioca bon da es conduc ed using ul a- il a ion p e- ea men me hod o an h opogenically manipula ed bones om 13 si es in he Can ab ian egion con aining Mous e ian, Au ignacian and G a e ian le els, o which 30 a e conside ed ele an . These da es, alongside p e iously epo ed ones, we e in eg a ed in o a Bayesian age model o econs uc an absolu e imescale o he ansi ional pe iod. Acco ding o i , he Mous e ian disappea ed in he egion by 47.9–45.1ka cal BP, while he Cha ˆ elpe onian las ed be ween 42.6k and 41.5ka cal BP. The Mous e ian and Cha ˆ elpe o- nian did no o e lap, indica ing ha he la e migh be ei he in usi e o an o shoo o he Mous e ian. The new ch onology also sugges s ha he Au ignacian appea s be ween 43.3–40.5ka cal BP o e lapping wi h he Cha ˆ elpe onian, and ended a ound 34.6–33.1ka cal BP, a e he G a e ian had al eady been es ablished in he egion. This e idence indi- ca es ha Neande hals and AMH co-exis ed <1,000 yea s, wi h he ca ea ha no diagnos- ic human emains ha e been ound wi h he la es Mous e ian, Cha ˆ elpe onian o ea lies Au ignacian in Can ab ian Spain. PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0194708 Ap il 18, 2018 1 / 20 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 OPEN ACCESS Ci a ion: Ma ı ´n-A oyo AB, Rios-Ga aiza J, S aus LG, Jones JR, de la Rasilla M, Gonza ´lez Mo ales MR, e al. (2018) Ch onological eassessmen o he Middle o Uppe Paleoli hic ansi ion and Ea ly Uppe Paleoli hic cul u es in Can ab ian Spain. PLoS ONE 13(4): e0194708. h ps://doi.o g/ 10.1371/jou nal.pone.0194708 Edi o : John P. Ha , New Yo k S a e Museum, UNITED STATES Recei ed: Augus 11, 2017 Accep ed: Ma ch 8, 2018 Published: Ap il 18, 2018 Copy igh : ©2018 Ma ı ´n-A oyo e al. This is an open access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal au ho and sou ce a e c edi ed. Da a A ailabili y S a emen : All ele an da a a e wi hin he pape and i s Suppo ing In o ma ion iles. Funding: This esea ch has been suppo ed by he Eu opean Commission h ough FP7-PEOPLE- 2012-CIG (Re : 322112), he Spanish Minis y o Economy and Compe i i eness (HAR2012-33956 and RYC-2011-00695), Can ab ia Campus In e na ional and Uni e si y o Can ab ia o ABMA. The Ins i u o In e nacional de In es igaciones In oduc ion The Middle o Uppe Paleoli hic ansi ion is gene ally associa ed wi h he eplacemen o local Neande hal popula ions by Ana omically Mode n Humans (AMH) in wes e n Eu asia [1]. One o he mos deba ed issues is he p ecise ch onology o he eplacemen o he popula- ion and i s po en ial ela ionship wi h majo biological, en i onmen al and cul u al e en s. Due o ecen e isions o a chaeological, ch onological and geog aphical da a, he spa io- em- po al delimi a ion o he Middle o Uppe Paleoli hic ansi ion in cen al and wes e n Eu ope has imp o ed [1–5]. Me hodological de elopmen s in adioca bon da ing ha emo e young ca bon con amina ion om old o ganic samples [6–7] ha e been c ucial o es ablishing accu- a e ch onologies, especially a he adioca bon limi a ound ~50,000 yea s ago, when he Mid- dle o Uppe Paleoli hic ansi ion ook place. Thus, he e-da ing o se e al key si es in Eu ope has p oduced olde ages han p e iously hough [1]. The iming o Neande hal disappea ance in no hwes e n Ibe ia (~48-46ka uncal BP) and I aly (~44-42ka cal BP) and he age o so- called “ ansi ional indus ies” (Cha ˆ elpe onian and Uluzzian, bo h 44 o 40ka cal BP) ha e been ecen ly delimi ed [1,3]. While he a ibu ion o Neande hals o he Cha ˆ elpe onian in no he n Ibe ia and sou hwes e n F ance is seldom con es ed (bu see [8]), he ques ion o he make s o he Uluzzian is s ill deba ed. Human deciduous ee h ound in Sou he n I aly (G o a del Ca allo) sugges ha AMH we e he make s o he Uluzzian [9], whe eas o he esea che s con inue o a gue o Neande hal au ho ship [10]. Those “T ansi ional echno- complexes” (especially he Cha ˆ elpe onian), ha e been in e p e ed as an e idence o a sho pe iod o coexis ence be ween local and immig an human popula ions [1,11], wi h he ea ly phases o he Au ignacian complex (P o o-Au ignacian and ea ly Au ignacian) being a ib- u ed o he i s mode n humans in Eu ope, as ea ly as 42 ka cal BP, du ing GI10. In Ibe ia, he end o he Mous e ian has been shown o ha e been ea lie in he No h han once hough , while he Ea ly Uppe Paleoli hic in he Sou h (G a e ian) is olde han was p e iously belie ed [12]. Thus, he hypo hesis o a e y ea ly appea ance o Au ignacian in no he n Ibe- ia is no longe suppo ed [2,4]. He e, we ocus on he Can ab ian egion in no he n Ibe ia, whe e he a chaeology, s a ig- aphy and ch onology o se e al key si es—El Cas illo, Labeko Koba, La Viña, Esquilleu and Mo ı ´n—ha e been ecen ly e iewed [2,5,13], p o iding an ini ial amewo k o he Middle o Uppe Paleoli hic ansi ion. Howe e , u he esea ch a mo e ep esen a i e si es in his egion wi h Mous e ian, Cha ˆ elpe onian, Au ignacian and G a e ian ma e ials, ha ei he lacked adioca bon da es o had da es ob ained in he la e 20 h cen u y be o e he de elop- men o igo ous p e- ea men me hods, is needed o secu ely es ablish he ch onology by including mo e a chaeological si es. As pa o a wide p ojec o econs uc he clima e and local and egional en i onmen al condi ions ha la e Neande hals and ea ly AMH aced in he egion du ing Ma ine Iso opic S age 3 (MIS3), using s able iso opic analyses (δ 13 C, δ 15 N, δ 18 O and δ 34 S), i s we pe o med a comple e e iew o he ch onological da a om 13 egional si es ha included 28 Middle and Ea ly Uppe Paleoli hic a chaeological le els. Ou o 51 samples ha we e p ocessed o adioca bon da ing, 5 ailed and he emaining 46 ecei ed an AMS measu emen using new ul a il a ion me hod ha is mo e igo uous in emo ing possible bone collagen con amina ion, om con empo aneous si es which ei he lacked da es o had p e iously p oduced p oblema ic da es. The new 14 C adioca bon da es ob ained in his p ojec , in combina ion wi h o he a ailable ul a il e ed da es om he egion, ha e allowed he de elopmen o an independen , adio- me ic ch onology o p e iously unda ed si es in he Can ab ian egion, which had un il now been a ibu ed only h ough s a ig aphic posi ion and/o ma e ial cul u e cha ac e is ics o he Mous e ian, Cha ˆ elpe onian, Au ignacian o G a e ian echno-complexes. Re-da ing o Ch onological eassessmen o he Middle o Uppe Paleoli hic ansi ion in Can ab ian Spain PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0194708 Ap il 18, 2018 2 / 20 P ehis o ´ icas de Can ab ia is sponso ed by he Uni e si y o Can ab ia, he Go e nmen o Can ab ia and Banco San ande . This pape was mainly w i en du ing a esea ch s ay by one o us (ABMA) a he Le e hulme Cen e o E olu iona y S udies, Depa men o A chaeology, Uni e si y o Camb idge unded by Spanish Minis y o Educa ion, Cul u e and Spo s (Jose de Cas illejo P og am, Re no. CAS15_00054). The unde s had no ole in s udy design, da a collec ion and analysis, decision o publish, o p epa a ion o he manusc ip . Compe ing in e es s: The au ho s ha e decla ed ha no compe ing in e es s exis . sequences ha had p e iously been da ed using con en ional o ea ly AMS 14 C me hods was unde aken, p oducing signi ican changes in p e ious in e p e a ions o ele an si es, some- hing ha has also been e i ied by e iewing he li hic echnocomplexes. The esul s p o ide g ea e empo al p ecision o he Middle o Uppe Paleoli hic ansi ion in he Can ab ian Region by building a Bayesian model using samples ea ed wi h a obus p e- ea men me hodology. The model includes new Mous e ian modelled da es wi hin he ange o he calib a ion cu e and de ines he bounda ies o he s a and end o each o he Ea ly Uppe Paleoli hic echno-complexes. This esea ch p esen s new possibili ies o add essing he iming o he c i ical p ocesses o cul u al change, local ex inc ions and popula ion eplacemen ha happened du ing he Mid- dle-Uppe Paleoli hic ansi ion in no he n Spain and mo e gene ally in Eu ope. The da ing esul s epo ed he e will also se e o discuss he s a ig aphic in eg i y and cul u al a ibu- ion o he si es analysed. Ma e ial Thi een si es we e da ed as pa o his p ojec , including a o al o h ee Middle Paleoli hic and wen y- i e ea ly Uppe Paleoli hic a chaeological le els. These si es a e loca ed in he mode n-day egions o As u ias in he wes , Can ab ia in he cen e and he Basque Coun y in he eas o he Can ab ian Region (Fig 1). These a chaeological le els we e selec ed because o hei a ibu ion o he Mous e ian (n = 3), Cha ˆ elpe onian (n = 1), Au ignacian (n = 11) and G a e ian (N = 13) echno-complexes, p e iously de e mined based on hei s a ig aphic posi ion, ma e ial cul u e, a ailable da ing o a combina ion o hese ac o s. The Mous e ian le els da ed we e Amalda VII [14], Axlo IV [15] and Llonı ´n VI [16–20], while he only Cha ˆ el- pe onian le el was Ekain Xa [21]. The Au ignacian le els we e Ai zbi a e (Ca e III-en ance a ea) Vb cen e [22], Ekain IXb, Cob an e V and VI [23], El Ruso Ca e I IVb [24], El O e o IV, V and VI [25], Mo in 7c [26], Co alejos B (2) and C (3) [27]. Finally, he G a e ian le els Fig 1. Loca ion o si es in he Can ab ian egion, no he n Spain. 1: La Viña, 2: La Gu¨elga, 3: Llonı ´n, 4: Esquilleu, 5: El Cas illo, 6: Co alejos, 7: Ruso I, 8: Mo ı ´n, 9: Cob an e, 10: El O e o, 11: Mi o ´n, 12: El Cuco, 13: A illo , 14: Axlo , 15: Bolinkoba, 16: Labeko Koba, 17: Ekain, 18: Amalda, 19: Ai zbi a e Ca e III. h ps://doi.o g/10.1371/jou nal.pone.0194708.g001 Ch onological eassessmen o he Middle o Uppe Paleoli hic ansi ion in Can ab ian Spain PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0194708 Ap il 18, 2018 3 / 20 we e Ai zbi a e (Ca e III-en ance a ea) IV, Va and Vb uppe , Amalda V and VI, Bolinkoba VI/F [28–29], El Cuco III and Vb [30–31], La Viña VII, VIII, IX and X [32–36] and Llonı ´n V. O he egional si es wi h con empo aneous cul u al a ibu ions, bu unclea s a ig aphy, insu icien a chaeological in o ma ion, o hose ha a e cu en ly s ill unde s udy we e no included he e. We also ook in o conside a ion he adioca bon da es ecen ly published o he egional si es o La Viña, Esquilleu, La Gu¨elga, El Cas illo, Mo ı ´n, El Mi o ´n, A illo and Labeko Koba [2–5,37–38] ha ha e been achie ed using he same ul a il a ion me hod as his s udy, making hem compa able and app op ia e o include in he Bayesian models. We ha e ini ially accep ed he cul u e-s a ig aphic designa ions o indi idual le els p oposed by he exca a o s/analys s o he a ious si es. This b ings wi h i he consequence o inco po a - ing di e ing classi ica o y c i e ia (including di e ing app oaches o “lumping” e sus “spli - ing” in such designa ions) among he many di e en esea che s who s udied he a ious a i ac assemblages, some o which con ain mo e de ini i e empo ally/cul u ally bounded and hus diagnos ic a i ac s han o he s. Howe e , he Bayesian app oach applied he e has he capaci y o challenge hese p io a ibu ions, p omp ing a ein e p e a ion o he li hic assem- blages and he s a ig aphy. Whe e misma ches we e iden i ied, si es we e e isi ed, and i nec- essa y, a ibu ions changed acco ding o ou e iews. In case o doub , da es we e disca ded o a oid any bias in he conclusions. 51 samples we e p ocessed o adioca bon da ing on bones wi h e idence o human manipula ion such as an h opogenic ac u es and/o cu ma ks. Sample in o ma ion o each specimen, including si e, a chaeological le el, con ex , sample numbe , animal species and ana omical elemen is included in S1 Table. The loca ion o he a chaeological ma e ial in egional museums and ins i u ions is indica ed in he desc ip ion o each si e in S1 File. This esea ch implies a ema kable inc ease in he numbe o eliable adiome ic da es in he egion and pe iod o s udy, which cu en ly comp ises 61 de e mina ions [4,5,12–13] [S3 Table]. The selec ion o he auna samples in museum collec ions was based on he ollowing c i e ia: 1) s a ig aphic posi ion o he emains a he si e; 2) selec ion o bones p e e ably di ec ly labelled o assu e hei p o enance wi hin he si e; 3) selec ion o animal bones axo- nomically and ana omically iden i iable and 4) selec ion o ana omical elemen s wi h clea an h opogenic ma ks. Me hods Radioca bon da ing P io o adioca bon da ing, a subsample o all specimens was analysed o ob ain δ 13 C and δ 15 N alues, which also es ablished he s a e o collagen p ese a ion. O he bones p e iously analysed o s able iso opic analyses, specimens yielding >1% collagen we e selec ed o da ing o maximise he likelihood o achie ing a success ul adioca bon da e. All samples me he quali y assu ance c i e ia as shown in S1 Table. Collagen ex ac ion and analysis o adioca - bon da ing we e unde aken a he Ox o d Radioca bon Accele a o Uni (ORAU). Collagen was ob ained ollowing he me hod de ailed by B ock e al [39], which in ol es he demine ali- sa ion o he mine al componen (and any exogenous ca bona es) o d illed bone powe using 0.5M HCL a 5˚C o e nigh , be o e emo al o o ganics and humic acid using 0.1M NaOH solu ion o 30 minu es a oom empe a e (RT), and a inal wash in 0.5M HCL o 1 hou a RT. The collagen was gela inised in a 0.001M HCL solu ion o 20 hou s a 70˚C. EZEE™bio- logical il e s (45–90 μm) we e used o emo e smalle soluble componen s, be o e ul a il a- ion using cleaned 30 kDa MWCO Vi aspin™15 ul a il e s. Combus ion o he collagen using an elemen al analyse (ANCA-GSL), linked o an iso ope a io-mass spec ome e (Se con 20– 20) p oduced Ca bon and Ni ogen s able iso ope da a, and samples we e da ed using an Ch onological eassessmen o he Middle o Uppe Paleoli hic ansi ion in Can ab ian Spain PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0194708 Ap il 18, 2018 4 / 20 Accele a o Mass Spec ome e ollowing con e sion o excess CO 2 in o g aphi e using an i on ca alys [40]. Bayesian modelling Indi idual Bayesian age models could be buil o La Viña (S1 Fig), Co alejos (S2 Fig), El Cuco (S3 Fig), Ekain (S4 Fig) Amalda si es (S5 Fig) and Ai zibi a e (Ca e III) (S6 Fig) using OxCal4.2 so wa e [41] and he INTCAL13 calib a ion cu e [42]. No ch onological models we e buil o he o he si es, because o he limi ed numbe o da es. The Bayesian model enables o modi y he calib a ed P obabili y Dis ibu ion Func ion (PDF) o indi idual da es based on he exis ing ela i e s a ig aphic and o he ela i e age in o ma ion. Bo h new ul a- il a ion da es and p e ious ones ecen ly published o he pe iod o s udy in no he n Spain we e conside ed o comple eness (see S2 Table). The p esence o p oblema ic de e mina ions ha do no ag ee wi h he p io amewo k was conside ed by adop ing a - ype ou lie model wi h an ini ial 5% p obabili y o each de e mina ion o be an ou lie . Likewise, a s- ype ou lie model was used o es he cohe ence o wo adioca bon da es ob ained om he same bone emain [43]. A esolu ion o 20 yea s was assumed, being a easonable balance be ween equi ed accu acy and compu a ional cos s, and a sensi i i y es on he ou lie model and p io bounda y es was conduc ed o ensu e he obus ness o he esul s. A discussion o each a chaeological si e and cul u al a ibu ion acco ding o he da es a e discussed indi idually in S1 File. CQL codes a e also epo ed he e (S1 and S2 Codes). An O de unc ion in he OxCal was used o calcula e he p obabili y ha one PDF p eda ed ano he , p o iding in o ma ion o assess synch onici y and empo al o e lap o indi idual a chaeological le els and cul u al phases in each o he six indi idual si es modelled. In all cases, con e gence was g ea e han 95% and he model ag eemen index close o 100% excep in he case o La Viña, whe e i is s ill clea ly abo e he 60% alidi y h eshold. Once models o indi idual si es we e conduc ed, modelled da es om each indus y (pos- e io PDFs) we e g ouped wi hin a single cul u al phase (inse ed as p io s) wi hou assuming any o de . This was achie ed by placing hem wi hin a single Phase wi h s a and end bound- a ies. Indi idually calib a ed da es coming om si es we e no indi idual si e model could be conduc ed (such as Cob an e Le el VI, Llonı ´n V, Bolinkoba VI/F and Mo in 4), as well as o he exis ing eliable modelled da es in he egion ( om Labeko Koba [4], El Cas illo [5] and La Viña [13]), we e included. These new uno de ed Bayesian age models enabled de e mining a egional PDF o he empo al bounda ies be ween Mous e ian, Cha ˆ elpe onian, Au igna- cian and G a e ian, as well as an easie compa ison be ween si es. The ad an age o his analy- sis in compa ison wi h indi idual models is ha i enables u he cons aining he unce ain y o he bounda ies by assuming ha he onse and disappea ance o cul u es we e egional p o- cesses a he han localised phenomena. This is easonable gi en he known le el o egional in e connec ion du ing La e Pleis ocene in no he n Ibe ia. Likewise, i educes he po en ial bias a one pa icula si e, as pooling oge he se e al si es in one egion se es o inc ease he eliabili y o he esul s, which can be in e p e ed egionally wi h mo e con idence (see S11 Figu e in [3] and see Figu e 8 in [5]). The di e ence be ween he PDFs o he s a and end bounda ies was also calcula ed o es ima e he likely du a ion o he phase. The esul s we e compa ed wi h he G eenland Oxygen Iso opic eco d (NGRIP) [44,45] as a global clima ic eco d o de ine he di e en cul u al pe iods in he Can ab ian egion. All Bayesian models we e un 4–5 imes and esul s compa ed o check consis ency. They disclosed accep able le els o ep oducibili y when compa ed, al hough key bounda y pa ame- e s we e usually wi hin 50–100 yea s o one ano he wi h epe i ion o he model. This is he usual accu acy expec ed when using his app oach, and consequen ly, all da es epo ed he e Ch onological eassessmen o he Middle o Uppe Paleoli hic ansi ion in Can ab ian Spain PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0194708 Ap il 18, 2018 5 / 20 ha e been ounded o he nea es cen u y. Finally, he younges and oldes da es o each cul- u al pe iod we e emo ed and Bayesian models we e e un o es he dependence o he con- clusions on hem. Resul s and discussion Ou o he 51 da es p ocessed o adioca bon da ing, 46 p oduced success ul AMS measu e- men s as hey had enough collagen o p o ide iable esul s. Fou samples ailed due o low collagen yields and one due o a high C:N alue. Collagen was gene ally well p ese ed and only one sample om El Cuco yielded less han 1% collagen, bu i had a C:N a io o 3.39 was wi hin he accep able ange o in i o collagen [46]. O he 46 da es, 3 da es we e beyond he adioca bon limi and 13 showed misma ches wi h hei de ined cul u al a ibu ion p o iding a ch onology younge han ini ially a ibu ed ei he by cul u al a e ac s ound wi hin ha le el o p e ious da ing (S1 Table). Consequen ly, in his s udy, a subs an ial da ase o 30 alid new da es was ob ained and only wi h hose da es Bayesian models we e gene a ed. The 13 da es ha show disc epancies be ween he cul u al a ibu ions de e mined by he new da es ob ained and hose based on adi ional analyses o he a chaeological ma e ials co - espond o 7 o he 13 s udied si es (11 o he 28 le els). This migh be caused by one o se e al easons such as exca a ion me hods (e.g., ailu e o see o iden i y s a ig aphic di e ences, a ibu ions based on e y small, no e y diagnos ic a i ac collec ions om limi ed exca a- ions o es pi s) o cu a o ial issues wi h he aunal ma e ial, as se e al o he egional muse- ums ha e been mo ed epea edly o di e en loca ions o e se e al decades. Fo each si e, possible easons o hose inconsis encies a e discussed in S1 File. Indi idual Bayesian models we e c ea ed o si es whe e samples belonged o mo e han one laye , such as La Viña (S1 Fig), Co alejos (S2 Fig), El Cuco (S3 Fig), Amalda (S5 Fig) and Ai zbi a e III (S6 Fig). Fo Ekain wo da es o Le el IXb we e also modelled (S4 Fig). CQL codes a e p esen ed in S1 and S2 Codes. This allowed, conside ing he s a ig aphy o he si es as p io in o ma ion o he da es, leading o mo e obus esul s. The obus ness o he esul s has also been e i ied by means o a sensi i i y es (see S4 Table), which esul ed in ma ginal changes o he ele an bounda ies be ween cul u es. In se e al a chaeological le els, ypological ch onologies ha e been challenged by he new da ing e idence. Thus, in As u ias, he si e o Llonı ´n Le el V, a ibu ed o he G a e ian [16– 20], yielded wo e y di e en esul s: one cohe en wi h an Ea ly G a e ian a ibu ion (~28k uncal BP) and ano he da e o 20k uncal BP, which could co espond o an ea ly Solu ean o a e y la e phase o he G a e ian, younge han Mo ı ´n Le el 4 [47] and simila o he AMS da es o Le el III om Ai zbi a e III [48]. These esul s sugges ei he a much longe o ma- ion o Le el V (mo e han 8,000 yea s o only 9 cm o s a ig aphy), al hough we canno dis- miss he possibili y o some dis u bance o admix u e wi h o e lying Solu ean Le el IV which emains unda ed. In he Au onomous Communi y o Can ab ia, he new ch onology o El O e o, which was p e iously unda ed, indica es a Magdalenian ch onology (~15–10 ka uncal BP, and he ea e ) ins ead o an Au ignacian one as adi ionally p oposed [25]. Despi e he p esence o appa en ly cha ac e is ic a i ac s [25,49–50], his esul u ges cau ion in he da ing o le els solely using a i ac s ha a e no empo ally diagnos ic. In Le el V o Cob an e, iden i- ied as Au ignacian, despi e i s lack o cha ac e is ic ma e ial cul u e [51] he ob ained da es co espond o he Solu ean ins ead (~18k uncal BP). The appa en absence o cha ac e is ic Au ignacian li hics, and he p esence o a leas one piece wi h pa ial in asi e (Solu ean) e ouch [51], suppo his a ibu ion. Cob an e Le el VI was in e p e ed as A chaic Au igna- cian because o he p esence o cha ac e is ic li hic ools, such as Du ou bladele s, la ge “Au i- gnacian” blades and ca ina ed end-sc ape s, which appea ed alongside Mous e ian-looking Ch onological eassessmen o he Middle o Uppe Paleoli hic ansi ion in Can ab ian Spain PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0194708 Ap il 18, 2018 6 / 20 ools [51]. New wo da es show an incohe ence be ween hem: one wi h a da e beyond he adioca bon limi and ano he wi h a con empo aneous egional Ea ly Au ignacian da e. Cu a ional o s a ig aphical issues could be he eason o his incohe ence as explained in S1 File. In El Cuco, he Le els III and Vb p e iously de ined [30] bo h as G a e ian p o ided Au ignacian (~35k uncal BP) and Mous e ian (~49k uncal BP) da es, espec i ely. The lowe le els o his sequence (VII-XIII) ha e been ecen ly eassessed h ough new 14C AMS da ing on shells and a echnological li hic s udy. These new analyses con i m a Mous e ian a ibu ion o le els p e iously de ined as Au ignacian and an Au ignacian ch onology o le els de ined as G a e ian [31]. In El Ruso, Le el IVb o iginally de ined as E ol ed Au ignacian [52], p o- ided a G a e ian da e (~28k uncal BP). Al hough he small li hic assemblage is no e y diagnos ic, he p esence o a la -nosed end-sc ape and a double ca ina ed end-sc ape was men ioned. In Bizkaia, he G a e ian Le el VI o Bolinkoba p o ided a con empo aneous da e o his egional pe iod, while a second one is much younge (~10k uncal BP), a ibu able o he egional Azilian, sugges ing signi ican mixing o cu a ional p oblem as no diagnos ic Azilian li hic ma e ial has been eco e ed o iden i ied in Le el VI assemblage [53]. In Amalda (Gipuz- koa), h ee da es p o ided younge da es han expec ed [54]. One sample om Mous e ian Le el VII p o ided a da e wi hin a G a e ian ime ange (~28k uncal BP), possibly e lec ing a limi ed admix u e be ween Le els VI and VII [55]. Ano he sample om he G a e ian Le el VI ga e a medie al da e. I s loca ion close o he ca e en ance, whe e he uppe le els we e exposed, poin s owa ds he possibili y ha ma e ials om his o ic imes we e inse ed in o Las Glacial sedimen s. A u he sample om he G a e ian Le el V p o ided a Magdalenian age (~14k uncal BP), maybe de i ing om unknown p oblems du ing exca a ion o cu a ion. In Ai zbi a e ca e III Le el Vb (cen e ) yielded a single da e ha is olde han he no ma i e a ibu ion (likely Ea ly Au ignacian a he han E ol ed Au ignacian) [56], e idencing a pos- sible p esence o an Ea ly Au ignacian a he base o he le el. In Ekain (Gipuzkoa), Le el Xa, unda ed, was in e p e ed as a Cha ˆ elpe onian hun ing camp, wi h a small li hic collec ion, including a single ypical Cha ˆ elpe onian poin , and h ee o he backed blades and bladele s [57]. The new da e o Le el Xa is 34 ka uncal BP, which is a oo young o he egional Cha ˆ- elpe onian, da ed a 37-38k uncal BP in nea by Labeko Koba [4]. The loca ion o he bone sampled a he si e and he signi ican p esence o ca e bea s do no ule ou he possibili y o admix u e caused ei he by hose ca ni o es o by o he pos -deposi ional p ocesses no iden i- ied du ing he exca a ion o Ekain. These disco dances a e pa icula ly ele an o si es p o usely used in he pas o in es i- ga e he cul u al adap a ions du ing he Can ab ian Uppe Paleoli hic and e lec he complexi- ies o a ibu ing a chaeological le els based solely on li hic ypology o echnology. I also e eals he p oblems caused by he a ibu ions based on single da es wi hou aking in o accoun he cha ac e is ics o he a chaeological assemblage. The egional ch onological amewo k o each cul u al pe iod as ob ained om he Bayes- ian age model is p esen ed in Figs 2–4. No egional model o he Cha ˆ elpe onian is included as only ul a il e ed da es a e a ailable o Le el IX lowe o Labeko Koba and his si e has been modelled elsewhe e (see Figu e S20 in [3]). Rega ding he ch onology o he La e Mous e ian in he Can ab ian egion, he da es om Axlo go beyond he adioca bon limi and samples om Llonı ´n Le el VI ailed. Howe e , wo new da es, bo h om Le el VII in Amalda, add impo an new e idence o he exis ing eco d. The upda ed Bayesian model o he Uppe o Middle Paleoli hic ansi ion shows ha he end bounda y PDF o he Mous e ian in he egion is 47.9–45.1ka cal BP (all p obabili y anges a e exp essed a 95,4% he ea e ), con i m- ing ha i occu ed ea lie in he A lan ic zone han in he no heas e n Ibe ia and in wes e n and no he n Eu ope [3]. The appa en end o he Mous e ian came immedia ely be o e and Ch onological eassessmen o he Middle o Uppe Paleoli hic ansi ion in Can ab ian Spain PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0194708 Ap il 18, 2018 7 / 20 du ing GI12 (Fig 2). No Cha ˆ elpe onian da es we e ob ained in his p ojec , despi e Le el Xa o Ekain ha ing o iginally been de ined as Cha ˆ elpe onian based on li hic ypology. Howe e , he da e ob ained om Ekain was oo young o he egional ange o he Cha ˆ elpe onian, which is ound in he Can ab ian Region in Mo ı ´n and Labeko Koba, bu is only eliably da ed in he la e a 42.6–41.4ka cal BP (see SI pp.47 in [3]). The ea lie disappea ance o hose Mous e ian g oups in he Can ab ian egion (no he n A lan ic Ibe ia), while hey s ill su i ed in Ca alonia (no h-eas e n Medi e anean Ibe ia) o a ew mo e millennia un il jus be o e 42 cal BP [2,58], as da ed in Romani and L‘A b eda, migh ha e been caused by as ye unknown ac o s and equi es u he esea ch. The age model da a e eals ha he e was an in e al o 2,700 o 5,800k (95.4% p obabili y), wi h a median o 4,400 yea s be ween he end o he Mous e ian and he s a o he Cha ˆ elpe - onian in he Can ab ian Region (Table 1). The Cha ˆ elpe onian appea s o s a se e al millennia a e he Mous e ian comple ely disappea ed in he egion, while in F ance his “ an- si ional” echnocomplex appea s be o e he end o he Mous e ian [2]. I Neande hals we e esponsible o bo h Mous e ian and Cha ˆ elpe onian echnocomplexes in he Can ab ian Fig 2. Radioca bon da es o he uppe mos da es o he Mous e ian assemblages om he Can ab ian egion (As u ias, Can ab ia and he Basque Coun y) calib a ed agains In Cal13 [42] in OxCal .4.2 [41] assuming each sample has a 5% p io p obabili y o being and ou lie wi hin he gene al - ype ou lie model [43]. All da es ha e been ob ained using he ul a il a ion p o ocol [40]. Re e ences a e gi en in S1 Table. A model = 99.8. deno es p e iously unmodelled da es in oduced as R-da es while he o he s a e modelled da es a indi idual si es in oduced as P io PDFs. h ps://doi.o g/10.1371/jou nal.pone.0194708.g002 Ch onological eassessmen o he Middle o Uppe Paleoli hic ansi ion in Can ab ian Spain PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0194708 Ap il 18, 2018 8 / 20 Fig 3. Radioca bon da es o he Au ignacian assemblages om he Can ab ian egion (As u ias, Can ab ia and he Basque Coun y) calib a ed agains In Cal13 [42] in OxCal .4.2 [41] assuming each sample has a 5% p io p obabili y o being and ou lie wi hin he gene al - ype ou lie model [43]. All da es ha e been ob ained using he ul a il a ion p o ocol [40]. Re e ences a e gi en in S1 Table. A model = 84.2. deno es p e iously unmodelled da es in oduced as R-da es while he o he s a e modelled da es a indi idual si es in oduced as P io PDFs. h ps://doi.o g/10.1371/jou nal.pone.0194708.g003 Ch onological eassessmen o he Middle o Uppe Paleoli hic ansi ion in Can ab ian Spain PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0194708 Ap il 18, 2018 9 / 20 S6 Fig. Radioca bon da es om Ai zbi a e III modelled in OxCal4.2 [41,43] agains INTCAL13 [42]. (TIF) S1 Code. CQL indi idual codes. (DOCX) S2 Code. CQL egional codes. (DOCX) Acknowledgmen s This esea ch has been suppo ed by he Eu opean Commission h ough FP7-PEOPLE-2012- CIG (Re : 322112), he Spanish Minis y o Economy and Compe i i eness (HAR2012-33956 and RYC-2011-00695), Can ab ia Campus In e na ional and Uni e si y o Can ab ia o ABMA. The Ins i u o In e nacional de In es igaciones P ehis o ´ icas de Can ab ia is sponso ed by he Uni e si y o Can ab ia, he Go e nmen o Can ab ia and Banco San ande . The au ho s wan o acknowledge he Museo de P ehis o ia y A queologı ´a de Can ab ia, Museo A queolo ´gico de Bizkaia (specially o D. Ga a e) and he Cen o de Pa imonio Cul u al del Depa amen o de Cul u a del Gobie no Vasco o he acili ies p o ided du ing he sampling and s udying ma e ial. They wa mly app ecia e he signi ican help o a chaeologis P. Rasines, R. Mon es and E. Muñoz du ing sample selec ion. This pape was mainly w i en du ing a esea ch s ay by one o us (ABMA) a he Le e hulme Cen e o E olu iona y S udies, Depa - men o A chaeology, Uni e si y o Camb idge unded by Spanish Minis y o Educa ion, Cul u e and Spo s (Jose de Cas illejo P og am, Re no. CAS15_00054). ABMA and DO a e g a e ul o P o . Robe Foley and P o . Ma a Mi azo ´n-Lah o hei ins i u ional suppo . Au ho Con ibu ions Concep ualiza ion: Ana B. Ma ı ´n-A oyo. Da a cu a ion: Ana B. Ma ı ´n-A oyo. Fo mal analysis: Ana B. Ma ı ´n-A oyo, Joseba Rios-Ga aiza , Da id Ocio. Funding acquisi ion: Ana B. Ma ı ´n-A oyo. In es iga ion: Ana B. Ma ı ´n-A oyo, Joseba Rios-Ga aiza , Jenni e R. Jones. Me hodology: Ana B. Ma ı ´n-A oyo. P ojec adminis a ion: Ana B. Ma ı ´n-A oyo, Jenni e R. Jones. Resou ces: Ana B. Ma ı ´n-A oyo, Ma co de la Rasilla, Manuel R. Gonza ´lez Mo ales, Michael Richa ds, Jesu ´s Al una, Ko o Ma iezku ena. So wa e: Ana B. Ma ı ´n-A oyo, Da id Ocio. Supe ision: Ana B. Ma ı ´n-A oyo. Valida ion: Ana B. Ma ı ´n-A oyo, Law ence G. S aus, Da id Ocio. Visualiza ion: Ana B. Ma ı ´n-A oyo, Joseba Rios-Ga aiza , Da id Ocio. W i ing – o iginal d a : Ana B. Ma ı ´n-A oyo, Joseba Rios-Ga aiza , Law ence G. S aus, Jenni e R. Jones, Da id Ocio. Ch onological eassessmen o he Middle o Uppe Paleoli hic ansi ion in Can ab ian Spain PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0194708 Ap il 18, 2018 16 / 20 W i ing – e iew & edi ing: Ana B. Ma ı ´n-A oyo, Joseba Rios-Ga aiza , Law ence G. S aus, Jenni e R. Jones, Ma co de la Rasilla, Manuel R. Gonza ´lez Mo ales, Michael Richa ds, Jesu ´s Al una, Ko o Ma iezku ena, Da id Ocio. Re e ences 1. Hublin J. The mode n human coloniza ion o wes e n Eu asia: when and whe e? Qua Sci Re . 2015; 118: 194–210. h ps://doi.o g/10.1016/j.quasci e .2014.08.011 2. Ma o o J, Vaque o M, A izabalaga A, Baena J, Baquedano E, Jo da ´J, e al. Cu en issues in la e Mid- dle Palaeoli hic ch onology: New assessmen s om No he n Ibe ia. Qua In e na . 2012; 247:15–25. h ps://doi.o g/10.1016/j.quain .2011.07.007 3. Higham T, Douka K, Wood R, Ramsey CB, B ock F, Basell L, e al. The iming and spa io empo al pa - e ning o Neande hal disappea ance. Na u e 2014; 512: 306–309. h ps://doi.o g/10.1038/ na u e13621 PMID: 25143113 4. Wood RE, A izabalaga A, Camps M, Fallon S, I ia e-Chiapusso MJ, Jones R, e al. The ch onology o he ea lies Uppe Palaeoli hic in no he n Ibe ia: New insigh s om l’A b eda, Labeko Koba and La Viña. J Hum E ol. 2014; 69: 91–109. h ps://doi.o g/10.1016/j.jhe ol.2013.12.017 PMID: 24636733 5. Wood R, de Qui o ´s FB, Maı´llo-Fe na ´ndez JM, Teje o JM, Nei a A, Higham T. El Cas illo (Can ab ia, no he n Ibe ia) and he T ansi ional Au ignacian: Using adioca bon da ing o assess si e aphonomy. Qua In . Fo hcoming; h ps://doi.o g/10.1016/j.quain .2016.03.005 6. Higham T, Jacobi R, B onk Ramsey C. AMS adioca bon da ing o ancien bone using ul a il a ion. Radioca bon 2006; 48: 179–195. h ps://doi.o g/10.1017/S0033822200066388 7. Higham T. Eu opean Middle and Uppe Palaeoli hic adioca bon da es a e o en olde han hey look: p oblems wi h p e ious da es and some emedies. An iqui y 2011; 85 (327): 235–249. h ps://doi.o g/ 10.1017/S0003598X00067570 8. Ba -Yose O, Bo des JG. Who we e he make s o he Cha ˆ elpe onian cul u e? J Hum E ol 2010; 59, 586–593. h p://dx.doi.o g/10.1016/j.jhe ol.2010.06.009 PMID: 20692683 9. Benazzi S, Douka K, Fo nai C, Baue CC, Kullme O, S oboda J, e al. Ea ly dispe sal o mode n humans in Eu ope and implica ions o Neande hal beha iou . Na u e 2011; 479, 525–528. h ps://doi. o g/10.1038/na u e10617 PMID: 22048311 10. Zilhão J, Banks W, d’E ico F, Gioia P. Analysis o Si e Fo ma ion and Assemblage In eg i y Does No Suppo A ibu ion o he Uluzzian o Mode n Humans a G o a del Ca allo. Plos ONE 2015; 10(7): 131–181. h ps://doi.o g/10.1371/jou nal.pone.0131181 PMID: 26154139 11. Ruebens K. Regional beha iou among la e Neande hal g oups in Wes e n Eu ope: a compa a i e assessmen o la e Middle Palaeoli hic bi acial ool a iabili y. J Hum E ol. 2013; 65: 341–362. h ps:// doi.o g/10.1016/j.jhe ol.2013.06.009 PMID: 23928352 12. Wood RE, Ba oso-Ruiz C, Capa os M, Jo da ´JF, Gal a ´n San os B, Higham TF. Radioca bon da ing cas s doub on he la e ch onology o he Middle o Uppe Palaeoli hic ansi ion in sou he n Ibe ia. P oc Na l Acad Sci U S A 2013; 110: 2781–2786. h ps://doi.o g/10.1073/pnas.1207656110 PMID: 23382220 13. Wood R. The con ibu ion o new adioca bon da ing p e- ea men echniques o unde s anding he Middle o Uppe Palaeoli hic ansi ion in Ibe ia. Uni e si y o Ox o d. PhD Thesis. 2011. h ps://o a.ox. ac.uk/objec s/uuid:075d79c6-edb4-4 19-9e34-50a63e7b7 e0 14. Al una J, Baldeo ´n A, Ma iezku ena K. edi o s. La Cue a de Amalda (Zes oa, Paı ´s Vasco). Ocupa- ciones Paleolı´ icas y Pos paleolı´ icas. Donos ia-San Sebas ia ´n: Sociedad de Es udios Vascos; 1990. 15. Gonza ´lez-U quijo JE, Iba ´ñez-Es e ´ ez JJ. Ab igo de Axlo (Dima). A keoikuska: In es igacio ´n a queo- lo ´gica; 2001. pp.90–93. 16. Fo ea J, de la Rasilla M, Rod ı ´guez V. La cue a de Llonı ´n (Llonı ´n, Peñamelle a Al a). Campañas de 1991 a 1994. Exca aciones A queolo ´gicas en As u ias 1987–1990, O iedo: Se icio de Publicaciones del P incipado de As u ias; 1992; 2: pp. 9–18. 17. Fo ea J, de la Rasilla M, Rod ı ´guez V. La cue a de Llonı ´n (Llonı ´n, Peñamelle a Al a). Campañas de 1991 a 1994. Exca aciones A queolo ´gicas en As u ias 1991–1994; 1995; 3: pp. 33–44. 18. Fo ea J, de la Rasilla M, Rod ı ´guez V. La cue a de Llonı ´n (Llonı ´n, Peñamelle a Al a). Campañas de 1995 a 1998. Exca aciones A queolo ´gicas en As u ias 1995–1998; 1999; 4: pp. 59–68. 19. Fo ea J, de la Rasilla M, Rod ı ´guez V. L’a pa ie ´ al e la se ´quence a che ´ologique pale ´oli hique de la g o e de Llonı ´n (Peñamelle a Al a, As u ies, Espagne). P e ´his oi e, A e Socie ´ e ´s. BSPA. 2004; LIX: 7–29. Ch onological eassessmen o he Middle o Uppe Paleoli hic ansi ion in Can ab ian Spain PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0194708 Ap il 18, 2018 17 / 20 20. Fo ea J, de la Rasilla M, Rod ı ´guez V. La cue a de Llonı ´n (Llonı ´n, Peñamelle a Al a). Campañas de 1999 a 2002. Exca aciones A queolo ´gicas en As u ias 1999–2002; 2007; 5: pp.77–86. 21. Al una J, Me ino J. El Yacimien o P ehis o ´ ico de la Cue a de Ekain. San Sebas ia ´n: Sociedad de Es udios. Vascos; 1984. 22. Al una J, Ma iezku ena K, Rios-Ga aiza J. Ocupaciones humanas en la cue a de Ai zbi a e III (Ren- e ia, Paı ´s Vasco) sec o En ada: 33.000–18.000 BP. Vi o ia-Gas eiz: Eusko Jau la i za en A gi alpen Ze bi zu Nagusia; 2011. 23. Rasines del Rı ´o P. Geog a ı ´a, es a ig a ı ´a y c onologı ´a de la cue a de Cob an e. Sau uola: Re is a del Ins i u o de P ehis o ia y A queologı ´a. 2009; 15: 43–47. 24. Muñoz Fe na ´ndez E, Se na Gancedo A. Los ni eles solu enses de la Cue a del Ruso I (Igollo de Cama go, Can ab ia). Sau uola: Re is a del Ins i u o de P ehis o ia y A queologı ´a. 1999; 6: 161–176. 25. Echega ay JG. Cue a del O e o. Exca aciones A queolo ´gicas en España, 53. Mad id: Minis e io de Educacio ´n Nacional Di eccio ´n Gene al de Bellas A es, Se icio Nacional de Exca aciones A queolo ´gi- cas; 1966. 26. Gonzalez Echega ay J. F eeman LG. Cue a Mo ı ´n: Exca aciones 1966–1968. San ande : Pa ona o de Cue as P ehis o ´ icas; 1971. 27. Sanguino Gonza ´lez J, Mon es Ba quı ´n R. Nue os da os pa a el conocimien o del Paleolı ´ ico Medio en el cen o de la Regio ´n Can a ´b ica: La Cue a de Co alejos. San ande : Museo de Al ami a Monog a ı ´as. 2005; 20: 10–38. 28. Ba andia a ´n JM. Bolinkoba y o os yacimien os paleolı ´ icos en la sie a de Ambo o (Vizcaya). Cuade - nos de his o ia p imi i a 5, 2. Mad id: Semina io de His o ia P imi i a del Homb e; 1950. 29. I ia e-Chiapusso MJ, A izabalaga A. El yacimien o a queolo ´gico de Bolinkoba (Abadiño, Bizkaia). C o ´nica de las in es igaciones en la ca idad. Secuencia es a ig a ´ ica y c onologı ´a nume ´ ica, In: I ia e-Chiapusso MJ, A izabalaga A, edi o s. Bolinkoba (Abadiño) y su Yacimien o A queolo ´gico: A queologı ´a de La A queologı ´a Pa a La Pues a En Valo de Su Depo ´si o, a La Luz de Las Exca a- ciones An iguas Y Recien es. Kobie Se ie BAI 6. Dipu acio ´n Fo al de Bizkaia, Bilbao; 2015: pp. 5–88. 30. Muñoz Fe na ´ndez E, Mon es Ba quı ´n R. edi o s. In e enciones a queolo ´gicas en Cas o U diales, omo III. A queologı ´a y a e upes e paleolı ´ ico en las ca idades de El Cuco o Sobe a y La Las illa. Cas o U diales: Excmo. Ayun amien o de Cas o U diales. Concejalı ´a de Medio Ambien e y Pa imonio A queolo ´gico; 2007. 31. Gu ie ez-Zugas i I, Rı ´os-Ga aiza J, Ma ı ´n-A oyo AB, Rasines P, Ma o o J, Jones J, Bailey G, Rich- a ds M. Fo hcoming. A ch ono-cul u al eassessmen o he le els VI-XIV om El Cuco ock-shel e : a new sequence o he La e Middle Paleoli hic in he Can ab ian Region (no he n Ibe ia). Qua In h ps://doi.o g/10.1016/j.quain .2017.06.059 32. Fo ea J. Ab igo de La Viña. In o me de las campañas 1980–1986. Exca aciones A queolo ´gicas en As u ias (1983–1986). 1990; 1: 55–68. 33. Fo ea J. Ab igo de La Viña. In o me de las campañas 1987–1990. Exca aciones A queolo ´gicas en As u ias (1987–1990). 1992; 2: 19–28. 34. Fo ea J. Ab igo de La Viña. In o me y p ime a alo acio ´n de las campañas 1991 a 1994. Exca aciones A queolo ´gicas en As u ias (1991–1994). 1995; 3: 19–32. 35. Fo ea J. Le Pale ´oli hique supe ´ ieu en Espagne, Galice e As u ies, 1991–1995. En M. O e (di .), Le pale ´oli hique supe ´ ieu Eu ope ´en: bilan quinquennal 1991–1996. ERAUL 76. Liège: Uni e si e ´de Lie ´ge; 1996. pp. 329–339. 36. Fo ea J. 1999. Ab igo de La Viña. In o me y p ime a alo acio ´n de las campañas de 1995 a 1998. Exca aciones A queolo ´gicas en As u ias (1995–1998), 4: 31–41. 37. B ad mo ¨lle M. The G a e ian occupa ion o Le el 4 Cue a Mo ı ´n and i s egional con ex . Munibe. 2015; 66: 23–52. 38. S aus LG, Gonza ´lez Mo ales MR. El Mi o ´n Ca e (Ramales, Can ab ia, Spain) da e lis V: Middle Paleo- li hic and Lowe Magdalenian. Radioca bon. 2016; 58(4): 943–945. h ps://doi.o g/10.1017/RDC.2016. 84 39. B ock F, Higham T, Ramsey CB. P e-sc eening echniques o iden i ica ion o samples sui able o adioca bon da ing o poo ly p ese ed bones. J A chaeol Sci. 2010; 37 (4): 855–865. h ps://doi.o g/ 10.1016/j.jas.2009.11.015 40. B onk Ramsey C, Higham T, Bowles A, Hedges R. Imp o emen s o he p e- ea men o bone a Ox o d. Radioca bon. 2004; 46 (1): 155–163. h ps://doi.o g/10.1017/S0033822200039473 41. B onk Ramsey C. Bayesian analysis o adioca bon da es. Radioca bon. 2009; 51(1): 337–360. h ps:// doi.o g/10.1017/S0033822200033865 Ch onological eassessmen o he Middle o Uppe Paleoli hic ansi ion in Can ab ian Spain PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0194708 Ap il 18, 2018 18 / 20 42. Reime PJ, Ba d E, Bayliss A, Beck JW, Blackwell PG, B onk Ramsey C. e al. In Cal13 and Ma ine13 adioca bon age calib a ion cu es 0–50,000 yea s cal BP. Radioca bon. 2013; 55 (4): 1869–1887. h ps://doi.o g/10.2458/azu_js_ c.55.16947 43. B onk Ramsey CB. Dealing wi h ou lie s and o se s in adioca bon da ing. Radioca bon. 2009; 51 (3): 1023–1045. h ps://doi.o g/10.1017/S0033822200034093 44. Ande sen KK, S ensson A, Johnsen SJ, Rasmussen SO, Bigle M, Ro ¨ hlisbe ge R, e al. The G een- land ice co e ch onology 2005, 15–42 ka. Pa 1: cons uc ing he ime scale. Qua Sci Re . 2006; 25: 3246–3257. h ps://doi.o g/10.1016/j.quasci e .2006.08.002 45. S ensson A, Ande sen KK, Bigle M, Clausen HB, Dahl-Jensen D, Da ies SM, e al. The G eenland ice co e ch onology 2005, 15–42 ka. Pa 2: compa ison o o he eco ds. Qua Sci Re . 2006; 25: 3258– 3267. h ps://doi.o g/10.1016/j.quasci e .2006.08.003 46. DeNi o MJ. Pos mo em p ese a ion and al e a ion o in i o bone collagen iso ope a ios in ela ion o palaeodie a y econs uc ion. Na u e. 1985; 317: 806–809. 47. B ad mo ¨lle M. Ho ¨hlenlage des G a e ien—Mus e jungpala ¨o¬li hische Ho ¨hlennu zung am Beispiel des G a e ien No dspaniens. Hambu g: Ve lag D . Ko ač; 2014. 48. Al una J, Ma iezku ena K, de la Peña P, Rios-Ga aiza J. Los ni eles g a e ienses de la cue a de Ai z- bi a e III (Gipuzkoa). Indus ias y aunas asociadas. In: de las He as C., Lashe as J.A., A izabalaga A ´., de la Rasilla M. edi o s. Pensando El G a e iense: Nue os Da os Pa a La Regio ´n Can a ´b ica En Su Con ex o Peninsula Y Pi enaico. Monog a ı ´as Del Museo Nacional Y Cen o de In es igacio ´n de Al a- mi a, 23. Mad id: Minis e io de Educacio ´n, Cul u a; 2013. pp. 184–204. 49. Rios-Ga aiza J, de la Peña P, Maillo-Fe na ´ndez JM, 2013. El inal del Au iñaciense y el comienzo del G a e iense en la egio ´n can a ´b ica: una isio ´n ecno- ipolo ´gica. In: de las He as C, Lashe as JA, A i- zabalaga A ´, de la Rasilla M. edi o s. Pensando El G a e iense: Nue os Da os Pa a La Regio ´n Can a ´b- ica En Su Con ex o Peninsula Y Pi enaico. Monog a ı ´as Del Museo Nacional Y Cen o de In es igacio ´n de Al ami a, N.o 23. Mad id; Minis e io de Educacio ´n, Cul u a, 2013; pp. 369–382. 50. Be naldo de Qui o ´s F. Los inicios del Paleolı ´ ico Supe io Can a ´b ico. San ande : Monog a ı ´as del Museo y Cen o de es udios de Al ami a 8; 1982. 51. Muñoz Fe na ´ndez E, San ama ı ´a San ama ı ´a S. Ana ´lisis de la indus ia lı ´ ica de la cue a de Cob an e. Sau uola 2009; 15: 145–189. 52. Muñoz Fe na ´ndez E. Exca aciones a queolo ´gicas en la Cue a del Ruso I. A ance p elimina . A que- nas. 1991: 61–157. 53. Ba andia a ´n Maes u I. Paleomesolı ´ ico del Pi ineo Occiden al. Bases pa a la sis ema izacio ´n ipolo ´gica del ins umen al o ´seo paleolı ´ ico. Monog a ı ´as A queolo ´gicas. Za agoza: Uni e sidad de Za agoza; 1967. 54. Al una J. Si uacio ´n y desc ipcio ´n de la cue a de Amalda. His o ia de las exca aciones. Desc ipcio ´n del elleno. Es uc u as en el yacimien o. Da aciones de adioca bono. O os yacimien os del alle. In: Al una J, Baldeo ´n A, Ma iezku ena K. edi o s. La Cue a de Amalda (Zes oa, Paı ´s Vasco). Ocupa- ciones Paleolı ´ icas Y Pos paleolı ´ icas. Sociedad de Es udios Vascos, Donos ia-San Sebas ia ´n, 1990. pp. 9–31. 55. Rios-Ga aiza J. O ganizacio ´n econo ´mica de las sociedades Neande ales: el caso del ni el VII de Amalda (Zes oa, Gipuzkoa). Zephy us. 2010; LXV, 15–37. 56. Rios-Ga aiza J, de la Peña P, San Eme e io A. Es udio de las indus ias lı ´ icas y o ´seas de la cue a de Ai zbi a e III (Zona de la en ada). In: Al una J., Ma iezku ena K., Rios-Ga aiza J. edi o s. Ocupa- ciones Humanas En La Cue a de Ai zbi a e III (Ren e ia, Paı ´s Vasco) Sec o En ada: 33.000–18.000 BP. Vi o ia-Gas eiz: Eusko Jau la i za en A gi alpen Ze bi zu Nagusia, 2011. pp. 81–351. 57. Rios-Ga aiza J, A izabalaga A ´, Villaluenga A. Hal es de chasse du Cha ˆ elpe onien de la Pe ´ninsule Ibe ´ ique. Labeko Koba e Ekain (Pays Basque Pe ´ninsulai e). L’An h opologie 2012; 116: 532–549. h ps://doi.o g/10.1016/j.an h o.2012.10.001 58. Camps M, Higham T. Ch onology o he Middle o Uppe Palaeoli hic ansi ion a Ab ic Romanı ´, Ca alunya. J Hum E ol. 2012; 62 (1): 89–103. h ps://doi.o g/10.1016/j.jhe ol.2011.10.010 PMID: 22137586 59. Zilhão J. Ch onos a ig aphy o he Middle- o-Uppe Paleoli hic T ansi ion in he Ibe ian Peninsula. Py - enae 2006; 37(1): 7–84. 60. Lowe J, Ba on N, Blockley S, B onk Ramsey C, Cullen V, Da ies W, e al. Volcanic ash laye s illumi- na e he esilience o Neande hals and ea ly mode n humans o na u al haza ds. P oc Na l Acad Sci U S A 2012; 109 (34): 13532–13537. h ps://doi.o g/10.1073/pnas.1204579109 PMID: 22826222 61. Fedele FG, Giaccio B, Hajdas I. Timescales and cul u al p ocess a 40,000 BP in he ligh o he Campa- nian Ignimb i e e up ion, Wes e n Eu asia. J Hum E ol 2008; 55: 834–857. h ps://doi.o g/10.1016/j. jhe ol.2008.08.012 PMID: 18922561 Ch onological eassessmen o he Middle o Uppe Paleoli hic ansi ion in Can ab ian Spain PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0194708 Ap il 18, 2018 19 / 20 62. Golo ano a LV, Bo iso ich Do oniche V, Elansia Clegho n N, Koulko a M, Sapelko T, Shackley MS. Signi icance o ecological ac o s in he Middle o Uppe Paleoli hic T ansi ion. Cu en An h opolology 2010; 51: 655–691. h ps://doi.o g/10.1086/656185 63. Maı ´llo-Fe na ´ndez JM, Be naldo de Qui o ´s F. A chaic Au ignacian in El Cas illo ca e (Spain): echnol- ogy and ypology composi ion. An h opologie. 2010; 114 (1): 1–25. h ps://doi.o g/10.1016/j.an h o. 2010.01.001 64. Y a ed a J. New Con ibu ions on Subsis ence P ac ices du ing he Middle-Uppe Paleoli hic in No h- e n Spain. In: Cla k J, Spe h J. (Eds). Zooa chaeology and Mode n Human O igins. Ve eb a e Paleobi- ology and Paleoan h opology book se ies. Sp inge ; 2013. pp 77–95. 65. an Klinken GJ. Bone Collagen Quali y Indica o s o Palaeodie a y and Radioca bon Measu emen s. J A chaeol Sci 1999; 26(6), 687–695. h ps://doi.o g/10.1006/jasc.1998.0385 Ch onological eassessmen o he Middle o Uppe Paleoli hic ansi ion in Can ab ian Spain PLOS ONE | h ps://doi.o g/10.1371/jou nal.pone.0194708 Ap il 18, 2018 20 / 20