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

RESEARCH ARTICLE Chronological reassessment of the Middle to Upper Paleolithic transition and Early Upper Paleolithic cultures in Cantabrian Spain Ana B. Marı ´n-Arroyo 1,2 *, Joseba Rios-Garaizar 3 , Lawrence G. Straus 4 , Jennifer R. Jones 1 , Marco de la Rasilla 5 , Manuel R. Gonza ´lez Morales 1 , Michael Richards 6 , Jesu ´s Altuna 7 , Koro Mariezkurrena 7 , David Ocio 8 1Instituto Internacional de Investigaciones Prehisto ´ricas de Cantabria, (Universidad de Cantabria, Santander, Gobierno de Cantabria), Santander, Spain, 2Leverhulme Centre for Evolutionary Studies, Department of Archaeology, University of Cambridge, Cambridge, United Kingdom, 3Centro Nacional de Investigacio ´n de la Evolucio ´n Humana, Burgos, Spain, 4Department of Anthropology, MSC01 1040, University of New Mexico, Albuquerque, New Mexico, United States of America, 5Departamento de Historia, Universidad de Oviedo, Oviedo, Spain, 6Department of Archaeology, Simon Fraser University, Burnaby, British Columbia, Canada, 7Centro de Conservacio ´n e Investigacio ´n de los Materiales Arqueolo ´gicos y Paleontolo ´gicos de Gipuzkoa, Donostia/San Sebastia ´n, Spain, 8Department of Civil and Environmental Engineering, Imperial College London, London, United Kingdom *[email protected] Abstract Methodological advances in dating the Middle to Upper Paleolithic transition provide a better understanding of the replacement of local Neanderthal populations by Anatomically Modern Humans. Today we know that this replacement was not a single, pan-European event, but rather it took place at different times in different regions. Thus, local conditions could have played a role. Iberia represents a significant macro-region to study this process. Northern Atlantic Spain contains evidence of both Mousterian and Early Upper Paleolithic occupations, although most of them are not properly dated, thus hindering the chances of an adequate interpretation. Here we present 46 new radiocarbon dates conducted using ultrafiltration pre-treatment method of anthropogenically manipulated bones from 13 sites in the Cantabrian region containing Mousterian, Aurignacian and Gravettian levels, of which 30 are considered relevant. These dates, alongside previously reported ones, were integrated into a Bayesian age model to reconstruct an absolute timescale for the transitional period. According to it, the Mousterian disappeared in the region by 47.9–45.1ka cal BP, while the Cha ˆtelperronian lasted between 42.6k and 41.5ka cal BP. The Mousterian and Cha ˆtelperronian did not overlap, indicating that the latter might be either intrusive or an offshoot of the Mousterian. The new chronology also suggests that the Aurignacian appears between 43.3–40.5ka cal BP overlapping with the Cha ˆtelperronian, and ended around 34.6–33.1ka cal BP, after the Gravettian had already been established in the region. This evidence indicates that Neanderthals and AMH co-existed <1,000 years, with the caveat that no diagnostic human remains have been found with the latest Mousterian, Cha ˆtelperronian or earliest Aurignacian in Cantabrian Spain. PLOS ONE | https://doi.org/10.1371/journal.pone.0194708 April 18, 2018 1 / 20 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 OPEN ACCESS Citation: Marı ´n-Arroyo AB, Rios-Garaizar J, Straus LG, Jones JR, de la Rasilla M, Gonza ´lez Morales MR, et al. (2018) Chronological reassessment of the Middle to Upper Paleolithic transition and Early Upper Paleolithic cultures in Cantabrian Spain. PLoS ONE 13(4): e0194708. https://doi.org/ 10.1371/journal.pone.0194708 Editor: John P. Hart, New York State Museum, UNITED STATES Received: August 11, 2017 Accepted: March 8, 2018 Published: April 18, 2018 Copyright: ©2018 Marı ´n-Arroyo et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: All relevant data are within the paper and its Supporting Information files. Funding: This research has been supported by the European Commission through FP7-PEOPLE2012-CIG (Ref: 322112), the Spanish Ministry of Economy and Competitiveness (HAR2012-33956 and RYC-2011-00695), Cantabria Campus International and University of Cantabria to ABMA. The Instituto Internacional de Investigaciones Introduction The Middle to Upper Paleolithic transition is generally associated with the replacement of local Neanderthal populations by Anatomically Modern Humans (AMH) in western Eurasia [1]. One of the most debated issues is the precise chronology of the replacement of the population and its potential relationship with major biological, environmental and cultural events. Due to recent revisions of archaeological, chronological and geographical data, the spatio-temporal delimitation of the Middle to Upper Paleolithic transition in central and western Europe has improved [1–5]. Methodological developments in radiocarbon dating that remove young carbon contamination from old organic samples [6–7] have been crucial for establishing accurate chronologies, especially at the radiocarbon limit around ~50,000 years ago, when the Middle to Upper Paleolithic transition took place. Thus, the re-dating of several key sites in Europe has produced older ages than previously thought [1]. The timing of Neanderthal disappearance in northwestern Iberia (~48-46ka uncal BP) and Italy (~44-42ka cal BP) and the age of socalled “transitional industries” (Cha ˆtelperronian and Uluzzian, both 44 to 40ka cal BP) have been recently delimited [1,3]. While the attribution to Neanderthals of the Cha ˆtelperronian in northern Iberia and southwestern France is seldom contested (but see [8]), the question of the makers of the Uluzzian is still debated. Human deciduous teeth found in Southern Italy (Grotta del Cavallo) suggest that AMH were the makers of the Uluzzian [9], whereas other researchers continue to argue for Neanderthal authorship [10]. Those “Transitional technocomplexes” (especially the Cha ˆtelperronian), have been interpreted as an evidence of a short period of coexistence between local and immigrant human populations [1,11], with the early phases of the Aurignacian complex (Proto-Aurignacian and early Aurignacian) being attributed to the first modern humans in Europe, as early as 42 ka cal BP, during GI10. In Iberia, the end of the Mousterian has been shown to have been earlier in the North than once thought, while the Early Upper Paleolithic in the South (Gravettian) is older than was previously believed [12]. Thus, the hypothesis of a very early appearance of Aurignacian in northern Iberia is no longer supported [2,4]. Here, we focus on the Cantabrian region in northern Iberia, where the archaeology, stratigraphy and chronology of several key sites—El Castillo, Labeko Koba, La Viña, Esquilleu and Morı ´n—have been recently reviewed [2,5,13], providing an initial framework for the Middle to Upper Paleolithic transition. However, further research at more representative sites in this region with Mousterian, Cha ˆtelperronian, Aurignacian and Gravettian materials, that either lacked radiocarbon dates or had dates obtained in the late 20th century before the development of rigorous pre-treatment methods, is needed to securely establish the chronology by including more archaeological sites. As part of a wider project to reconstruct the climate and local and regional environmental conditions that late Neanderthals and early AMH faced in the region during Marine Isotopic Stage 3 (MIS3), using stable isotopic analyses (δ 13 C, δ 15 N, δ 18 O and δ 34 S), first we performed a complete review of the chronological data from 13 regional sites that included 28 Middle and Early Upper Paleolithic archaeological levels. Out of 51 samples that were processed for radiocarbon dating, 5 failed and the remaining 46 received an AMS measurement using new ultrafiltration method that is more rigoruous in removing possible bone collagen contamination, from contemporaneous sites which either lacked dates or had previously produced problematic dates. The new 14 C radiocarbon dates obtained in this project, in combination with other available ultrafiltered dates from the region, have allowed the development of an independent, radiometric chronology of previously undated sites in the Cantabrian region, which had until now been attributed only through stratigraphic position and/or material culture characteristics to the Mousterian, Cha ˆtelperronian, Aurignacian or Gravettian techno-complexes. Re-dating of Chronological reassessment of the Middle to Upper Paleolithic transition in Cantabrian Spain PLOS ONE | https://doi.org/10.1371/journal.pone.0194708 April 18, 2018 2 / 20 Prehisto ´ricas de Cantabria is sponsored by the University of Cantabria, the Government of Cantabria and Banco Santander. This paper was mainly written during a research stay by one of us (ABMA) at the Leverhulme Centre for Evolutionary Studies, Department of Archaeology, University of Cambridge funded by Spanish Ministry of Education, Culture and Sports (Jose de Castillejo Program, Ref no. CAS15_00054). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interests: The authors have declared that no competing interests exist. sequences that had previously been dated using conventional or early AMS 14 C methods was undertaken, producing significant changes in previous interpretations of relevant sites, something that has also been verified by reviewing the lithic technocomplexes. The results provide greater temporal precision for the Middle to Upper Paleolithic transition in the Cantabrian Region by building a Bayesian model using samples treated with a robust pre-treatment methodology. The model includes new Mousterian modelled dates within the range of the calibration curve and defines the boundaries for the start and end of each of the Early Upper Paleolithic techno-complexes. This research presents new possibilities for addressing the timing of the critical processes of cultural change, local extinctions and population replacement that happened during the Middle-Upper Paleolithic transition in northern Spain and more generally in Europe. The dating results reported here will also serve to discuss the stratigraphic integrity and cultural attribution of the sites analysed. Material Thirteen sites were dated as part of this project, including a total of three Middle Paleolithic and twenty-five early Upper Paleolithic archaeological levels. These sites are located in the modern-day regions of Asturias in the west, Cantabria in the center and the Basque Country in the east of the Cantabrian Region (Fig 1). These archaeological levels were selected because of their attribution to the Mousterian (n = 3), Cha ˆtelperronian (n = 1), Aurignacian (n = 11) and Gravettian (N = 13) techno-complexes, previously determined based on their stratigraphic position, material culture, available dating or a combination of these factors. The Mousterian levels dated were Amalda VII [14], Axlor IV [15] and Llonı ´n VI [16–20], while the only Cha ˆtelperronian level was Ekain Xa [21]. The Aurignacian levels were Aitzbitarte (Cave III-entrance area) Vb center [22], Ekain IXb, Cobrante V and VI [23], El Ruso Cave I IVb [24], El Otero IV, V and VI [25], Morin 7c [26], Covalejos B (2) and C (3) [27]. Finally, the Gravettian levels Fig 1. Location of sites in the Cantabrian region, northern Spain. 1: La Viña, 2: La Gu¨elga, 3: Llonı ´n, 4: Esquilleu, 5: El Castillo, 6: Covalejos, 7: Ruso I, 8: Morı ´n, 9: Cobrante, 10: El Otero, 11: Miro ´n, 12: El Cuco, 13: Arrillor, 14: Axlor, 15: Bolinkoba, 16: Labeko Koba, 17: Ekain, 18: Amalda, 19: Aitzbitarte Cave III. https://doi.org/10.1371/journal.pone.0194708.g001 Chronological reassessment of the Middle to Upper Paleolithic transition in Cantabrian Spain PLOS ONE | https://doi.org/10.1371/journal.pone.0194708 April 18, 2018 3 / 20 were Aitzbitarte (Cave III-entrance area) IV, Va and Vb upper, 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. Other regional sites with contemporaneous cultural attributions, but unclear stratigraphy, insufficient archaeological information, or those that are currently still under study were not included here. We also took into consideration the radiocarbon dates recently published for the regional sites of La Viña, Esquilleu, La Gu¨elga, El Castillo, Morı ´n, El Miro ´n, Arrillor and Labeko Koba [2–5,37–38] that have been achieved using the same ultrafiltration method as this study, making them comparable and appropriate to include in the Bayesian models. We have initially accepted the culture-stratigraphic designations for individual levels proposed by the excavators/analysts of the various sites. This brings with it the consequence of incorporating differing classificatory criteria (including differing approaches to “lumping” versus “splitting” in such designations) among the many different researchers who studied the various artifact assemblages, some of which contain more definitive temporally/culturally bounded and thus diagnostic artifacts than others. However, the Bayesian approach applied here has the capacity to challenge these prior attributions, prompting a reinterpretation of the lithic assemblages and the stratigraphy. Where mismatches were identified, sites were revisited, and if necessary, attributions changed according to our reviews. In case of doubt, dates were discarded to avoid any bias in the conclusions. 51 samples were processed for radiocarbon dating on bones with evidence of human manipulation such as anthropogenic fractures and/or cut marks. Sample information for each specimen, including site, archaeological level, context, sample number, animal species and anatomical element is included in S1 Table. The location of the archaeological material in regional museums and institutions is indicated in the description of each site in S1 File. This research implies a remarkable increase in the number of reliable radiometric dates in the region and period of study, which currently comprises 61 determinations [4,5,12–13] [S3 Table]. The selection of the fauna samples in museum collections was based on the following criteria: 1) stratigraphic position of the remains at the site; 2) selection of bones preferably directly labelled to assure their provenance within the site; 3) selection of animal bones taxonomically and anatomically identifiable and 4) selection of anatomical elements with clear anthropogenic marks. Methods Radiocarbon dating Prior to radiocarbon dating, a subsample of all specimens was analysed to obtain δ 13 C and δ 15 N values, which also established the state of collagen preservation. Of the bones previously analysed for stable isotopic analyses, specimens yielding >1% collagen were selected for dating to maximise the likelihood of achieving a successful radiocarbon date. All samples met the quality assurance criteria as shown in S1 Table. Collagen extraction and analysis for radiocarbon dating were undertaken at the Oxford Radiocarbon Accelerator Unit (ORAU). Collagen was obtained following the method detailed by Brock et al [39], which involves the demineralisation of the mineral component (and any exogenous carbonates) of drilled bone power using 0.5M HCL at 5˚C overnight, before removal of organics and humic acid using 0.1M NaOH solution for 30 minutes at room temperate (RT), and a final wash in 0.5M HCL for 1 hour at RT. The collagen was gelatinised in a 0.001M HCL solution for 20 hours at 70˚C. EZEE™biological filters (45–90 μm) were used to remove smaller soluble components, before ultrafiltration using cleaned 30 kDa MWCO Vivaspin™15 ultrafilters. Combustion of the collagen using an elemental analyser (ANCA-GSL), linked to an isotope ratio-mass spectrometer (Sercon 20– 20) produced Carbon and Nitrogen stable isotope data, and samples were dated using an Chronological reassessment of the Middle to Upper Paleolithic transition in Cantabrian Spain PLOS ONE | https://doi.org/10.1371/journal.pone.0194708 April 18, 2018 4 / 20 Accelerator Mass Spectrometer following conversion of excess CO 2 into graphite using an iron catalyst [40]. Bayesian modelling Individual Bayesian age models could be built for La Viña (S1 Fig), Covalejos (S2 Fig), El Cuco (S3 Fig), Ekain (S4 Fig) Amalda sites (S5 Fig) and Aitzibitarte (Cave III) (S6 Fig) using OxCal4.2 software [41] and the INTCAL13 calibration curve [42]. No chronological models were built for the other sites, because of the limited number of dates. The Bayesian model enables to modify the calibrated Probability Distribution Function (PDF) of individual dates based on the existing relative stratigraphic and other relative age information. Both new ultrafiltration dates and previous ones recently published for the period of study in northern Spain were considered for completeness (see S2 Table). The presence of problematic determinations that do not agree with the prior framework was considered by adopting a t-type outlier model with an initial 5% probability for each determination to be an outlier. Likewise, a s-type outlier model was used to test the coherence of two radiocarbon dates obtained from the same bone remain [43]. A resolution of 20 years was assumed, being a reasonable balance between required accuracy and computational costs, and a sensitivity test on the outlier model and prior boundary test was conducted to ensure the robustness of the results. A discussion of each archaeological site and cultural attribution according to the dates are discussed individually in S1 File. CQL codes are also reported there (S1 and S2 Codes). An Order function in the OxCal was used to calculate the probability that one PDF predated another, providing information to assess synchronicity and temporal overlap of individual archaeological levels and cultural phases in each of the six individual sites modelled. In all cases, convergence was greater than 95% and the model agreement index close to 100% except in the case of La Viña, where it is still clearly above the 60% validity threshold. Once models for individual sites were conducted, modelled dates from each industry (posterior PDFs) were grouped within a single cultural phase (inserted as priors) without assuming any order. This was achieved by placing them within a single Phase with start and end boundaries. Individually calibrated dates coming from sites were no individual site model could be conducted (such as Cobrante Level VI, Llonı ´n V, Bolinkoba VI/F and Morin 4), as well as other existing reliable modelled dates in the region (from Labeko Koba [4], El Castillo [5] and La Viña [13]), were included. These new unordered Bayesian age models enabled determining a regional PDF of the temporal boundaries between Mousterian, Cha ˆtelperronian, Aurignacian and Gravettian, as well as an easier comparison between sites. The advantage of this analysis in comparison with individual models is that it enables further constraining the uncertainty of the boundaries by assuming that the onset and disappearance of cultures were regional processes rather than localised phenomena. This is reasonable given the known level of regional interconnection during Late Pleistocene in northern Iberia. Likewise, it reduces the potential bias at one particular site, as pooling together several sites in one region serves to increase the reliability of the results, which can be interpreted regionally with more confidence (see S11 Figure in [3] and see Figure 8 in [5]). The difference between the PDFs of the start and end boundaries was also calculated to estimate the likely duration of the phase. The results were compared with the Greenland Oxygen Isotopic record (NGRIP) [44,45] as a global climatic record to define the different cultural periods in the Cantabrian region. All Bayesian models were run 4–5 times and results compared to check consistency. They disclosed acceptable levels of reproducibility when compared, although key boundary parameters were usually within 50–100 years of one another with repetition of the model. This is the usual accuracy expected when using this approach, and consequently, all dates reported here Chronological reassessment of the Middle to Upper Paleolithic transition in Cantabrian Spain PLOS ONE | https://doi.org/10.1371/journal.pone.0194708 April 18, 2018 5 / 20 have been rounded to the nearest century. Finally, the youngest and oldest dates for each cultural period were removed and Bayesian models were rerun to test the dependence of the conclusions on them. Results and discussion Out of the 51 dates processed for radiocarbon dating, 46 produced successful AMS measurements as they had enough collagen to provide viable results. Four samples failed due to low collagen yields and one due to a high C:N value. Collagen was generally well preserved and only one sample from El Cuco yielded less than 1% collagen, but it had a C:N ratio of 3.39 was within the acceptable range for in vivo collagen [46]. Of the 46 dates, 3 dates were beyond the radiocarbon limit and 13 showed mismatches with their defined cultural attribution providing a chronology younger than initially attributed either by cultural artefacts found within that level or previous dating (S1 Table). Consequently, in this study, a substantial dataset of 30 valid new dates was obtained and only with those dates Bayesian models were generated. The 13 dates that show discrepancies between the cultural attributions determined by the new dates obtained and those based on traditional analyses of the archaeological materials correspond to 7 of the 13 studied sites (11 of the 28 levels). This might be caused by one or several reasons such as excavation methods (e.g., failure to see or identify stratigraphic differences, attributions based on very small, not very diagnostic artifact collections from limited excavations or test pits) or curatorial issues with the faunal material, as several of the regional museums have been moved repeatedly to different locations over several decades. For each site, possible reasons for those inconsistencies are discussed in S1 File. Individual Bayesian models were created for sites where samples belonged to more than one layer, such as La Viña (S1 Fig), Covalejos (S2 Fig), El Cuco (S3 Fig), Amalda (S5 Fig) and Aitzbitarte III (S6 Fig). For Ekain two dates for Level IXb were also modelled (S4 Fig). CQL codes are presented in S1 and S2 Codes. This allowed, considering the stratigraphy of the sites as prior information for the dates, leading to more robust results. The robustness of the results has also been verified by means of a sensitivity test (see S4 Table), which resulted in marginal changes to the relevant boundaries between cultures. In several archaeological levels, typological chronologies have been challenged by the new dating evidence. Thus, in Asturias, the site of Llonı ´n Level V, attributed to the Gravettian [16– 20], yielded two very different results: one coherent with an Early Gravettian attribution (~28k uncal BP) and another date of 20k uncal BP, which could correspond to an early Solutrean or a very late phase of the Gravettian, younger than Morı ´n Level 4 [47] and similar to the AMS dates of Level III from Aitzbitarte III [48]. These results suggest either a much longer formation for Level V (more than 8,000 years for only 9 cm of stratigraphy), although we cannot dismiss the possibility of some disturbance or admixture with overlying Solutrean Level IV which remains undated. In the Autonomous Community of Cantabria, the new chronology of El Otero, which was previously undated, indicates a Magdalenian chronology (~15–10 ka uncal BP, and hereafter) instead of an Aurignacian one as traditionally proposed [25]. Despite the presence of apparently characteristic artifacts [25,49–50], this result urges caution in the dating of levels solely using artifacts that are not temporally diagnostic. In Level V of Cobrante, identified as Aurignacian, despite its lack of characteristic material culture [51] the obtained dates correspond to the Solutrean instead (~18k uncal BP). The apparent absence of characteristic Aurignacian lithics, and the presence of at least one piece with partial invasive (Solutrean) retouch [51], support this attribution. Cobrante Level VI was interpreted as Archaic Aurignacian because of the presence of characteristic lithic tools, such as Dufour bladelets, large “Aurignacian” blades and carinated end-scrapers, which appeared alongside Mousterian-looking Chronological reassessment of the Middle to Upper Paleolithic transition in Cantabrian Spain PLOS ONE | https://doi.org/10.1371/journal.pone.0194708 April 18, 2018 6 / 20 tools [51]. New two dates show an incoherence between them: one with a date beyond the radiocarbon limit and another with a contemporaneous regional Early Aurignacian date. Curational or stratigraphical issues could be the reason for this incoherence as explained in S1 File. In El Cuco, the Levels III and Vb previously defined [30] both as Gravettian provided Aurignacian (~35k uncal BP) and Mousterian (~49k uncal BP) dates, respectively. The lower levels of this sequence (VII-XIII) have been recently reassessed through new 14C AMS dating on shells and a technological lithic study. These new analyses confirm a Mousterian attribution to levels previously defined as Aurignacian and an Aurignacian chronology to levels defined as Gravettian [31]. In El Ruso, Level IVb originally defined as Evolved Aurignacian [52], provided a Gravettian date (~28k uncal BP). Although the small lithic assemblage is not very diagnostic, the presence of a flat-nosed end-scraper and a double carinated end-scraper was mentioned. In Bizkaia, the Gravettian Level VI of Bolinkoba provided a contemporaneous date for this regional period, while a second one is much younger (~10k uncal BP), attributable to the regional Azilian, suggesting significant mixing or curational problem as no diagnostic Azilian lithic material has been recovered or identified in Level VI assemblage [53]. In Amalda (Gipuzkoa), three dates provided younger dates than expected [54]. One sample from Mousterian Level VII provided a date within a Gravettian time range (~28k uncal BP), possibly reflecting a limited admixture between Levels VI and VII [55]. Another sample from the Gravettian Level VI gave a medieval date. Its location close to the cave entrance, where the upper levels were exposed, points towards the possibility that materials from historic times were inserted into Last Glacial sediments. A further sample from the Gravettian Level V provided a Magdalenian age (~14k uncal BP), maybe deriving from unknown problems during excavation or curation. In Aitzbitarte cave III Level Vb (center) yielded a single date that is older than the normative attribution (likely Early Aurignacian rather than Evolved Aurignacian) [56], evidencing a possible presence of an Early Aurignacian at the base of the level. In Ekain (Gipuzkoa), Level Xa, undated, was interpreted as a Cha ˆtelperronian hunting camp, with a small lithic collection, including a single typical Cha ˆtelperronian point, and three other backed blades and bladelets [57]. The new date for Level Xa is 34 ka uncal BP, which is far too young for the regional Cha ˆtelperronian, dated at 37-38k uncal BP in nearby Labeko Koba [4]. The location of the bone sampled at the site and the significant presence of cave bears do not rule out the possibility of admixture caused either by those carnivores or by other post-depositional processes not identified during the excavation of Ekain. These discordances are particularly relevant for sites profusely used in the past to investigate the cultural adaptations during the Cantabrian Upper Paleolithic and reflect the complexities of attributing archaeological levels based solely on lithic typology or technology. It also reveals the problems caused by the attributions based on single dates without taking into account the characteristics of the archaeological assemblage. The regional chronological framework for each cultural period as obtained from the Bayesian age model is presented in Figs 2–4. No regional model for the Cha ˆtelperronian is included as only ultrafiltered dates are available for Level IX lower of Labeko Koba and this site has been modelled elsewhere (see Figure S20 in [3]). Regarding the chronology of the Late Mousterian in the Cantabrian region, the dates from Axlor go beyond the radiocarbon limit and samples from Llonı ´n Level VI failed. However, two new dates, both from Level VII in Amalda, add important new evidence to the existing record. The updated Bayesian model for the Upper to Middle Paleolithic transition shows that the end boundary PDF for the Mousterian in the region is 47.9–45.1ka cal BP (all probability ranges are expressed at 95,4% hereafter), confirming that it occurred earlier in the Atlantic zone than in the northeastern Iberia and in western and northern Europe [3]. The apparent end of the Mousterian came immediately before and Chronological reassessment of the Middle to Upper Paleolithic transition in Cantabrian Spain PLOS ONE | https://doi.org/10.1371/journal.pone.0194708 April 18, 2018 7 / 20 during GI12 (Fig 2). No Cha ˆtelperronian dates were obtained in this project, despite Level Xa of Ekain having originally been defined as Cha ˆtelperronian based on lithic typology. However, the date obtained from Ekain was too young for the regional range of the Cha ˆtelperronian, which is found in the Cantabrian Region in Morı ´n and Labeko Koba, but is only reliably dated in the latter at 42.6–41.4ka cal BP (see SI pp.47 in [3]). The earlier disappearance of those Mousterian groups in the Cantabrian region (northern Atlantic Iberia), while they still survived in Catalonia (north-eastern Mediterranean Iberia) for a few more millennia until just before 42 cal BP [2,58], as dated in Romani and L‘Arbreda, might have been caused by as yet unknown factors and requires further research. The age model data reveals that there was an interval of 2,700 to 5,800k (95.4% probability), with a median of 4,400 years between the end of the Mousterian and the start of the Cha ˆtelperronian in the Cantabrian Region (Table 1). The Cha ˆtelperronian appears to start several millennia after the Mousterian completely disappeared in the region, while in France this “transitional” technocomplex appears before the end of the Mousterian [2]. If Neanderthals were responsible for both Mousterian and Cha ˆtelperronian technocomplexes in the Cantabrian Fig 2. Radiocarbon dates for the uppermost dates of the Mousterian assemblages from the Cantabrian region (Asturias, Cantabria and the Basque Country) calibrated against IntCal13 [42] in OxCal v.4.2 [41] assuming each sample has a 5% prior probability of being and outlier within the general t-type outlier model [43]. All dates have been obtained using the ultrafiltration protocol [40]. References are given in S1 Table. A model = 99.8. denotes previously unmodelled dates introduced as R-dates while the others are modelled dates at individual sites introduced as Prior PDFs. https://doi.org/10.1371/journal.pone.0194708.g002 Chronological reassessment of the Middle to Upper Paleolithic transition in Cantabrian Spain PLOS ONE | https://doi.org/10.1371/journal.pone.0194708 April 18, 2018 8 / 20 Fig 3. Radiocarbon dates for the Aurignacian assemblages from the Cantabrian region (Asturias, Cantabria and the Basque Country) calibrated against IntCal13 [42] in OxCal v.4.2 [41] assuming each sample has a 5% prior probability of being and outlier within the general t-type outlier model [43]. All dates have been obtained using the ultrafiltration protocol [40]. References are given in S1 Table. A model = 84.2. denotes previously unmodelled dates introduced as R-dates while the others are modelled dates at individual sites introduced as Prior PDFs. https://doi.org/10.1371/journal.pone.0194708.g003 Chronological reassessment of the Middle to Upper Paleolithic transition in Cantabrian Spain PLOS ONE | https://doi.org/10.1371/journal.pone.0194708 April 18, 2018 9 / 20 S6 Fig. Radiocarbon dates from Aitzbitarte III modelled in OxCal4.2 [41,43] against INTCAL13 [42]. (TIF) S1 Code. CQL individual codes. (DOCX) S2 Code. CQL regional codes. (DOCX) Acknowledgments This research has been supported by the European Commission through FP7-PEOPLE-2012CIG (Ref: 322112), the Spanish Ministry of Economy and Competitiveness (HAR2012-33956 and RYC-2011-00695), Cantabria Campus International and University of Cantabria to ABMA. The Instituto Internacional de Investigaciones Prehisto ´ricas de Cantabria is sponsored by the University of Cantabria, the Government of Cantabria and Banco Santander. The authors want to acknowledge the Museo de Prehistoria y Arqueologı ´a de Cantabria, Museo Arqueolo ´gico de Bizkaia (specially to D. Garate) and the Centro de Patrimonio Cultural del Departamento de Cultura del Gobierno Vasco for the facilities provided during the sampling and studying material. They warmly appreciate the significant help of archaeologist P. Rasines, R. Montes and E. Muñoz during sample selection. This paper was mainly written during a research stay by one of us (ABMA) at the Leverhulme Centre for Evolutionary Studies, Department of Archaeology, University of Cambridge funded by Spanish Ministry of Education, Culture and Sports (Jose de Castillejo Program, Ref no. CAS15_00054). ABMA and DO are grateful to Prof. Robert Foley and Prof. Marta Mirazo ´n-Lahr for their institutional support. Author Contributions Conceptualization: Ana B. Marı ´n-Arroyo. Data curation: Ana B. Marı ´n-Arroyo. Formal analysis: Ana B. Marı ´n-Arroyo, Joseba Rios-Garaizar, David Ocio. Funding acquisition: Ana B. Marı ´n-Arroyo. Investigation: Ana B. Marı ´n-Arroyo, Joseba Rios-Garaizar, Jennifer R. Jones. Methodology: Ana B. Marı ´n-Arroyo. Project administration: Ana B. Marı ´n-Arroyo, Jennifer R. Jones. Resources: Ana B. Marı ´n-Arroyo, Marco de la Rasilla, Manuel R. Gonza ´lez Morales, Michael Richards, Jesu ´s Altuna, Koro Mariezkurrena. Software: Ana B. Marı ´n-Arroyo, David Ocio. Supervision: Ana B. Marı ´n-Arroyo. Validation: Ana B. Marı ´n-Arroyo, Lawrence G. Straus, David Ocio. Visualization: Ana B. Marı ´n-Arroyo, Joseba Rios-Garaizar, David Ocio. Writing – original draft: Ana B. Marı ´n-Arroyo, Joseba Rios-Garaizar, Lawrence G. Straus, Jennifer R. Jones, David Ocio. Chronological reassessment of the Middle to Upper Paleolithic transition in Cantabrian Spain PLOS ONE | https://doi.org/10.1371/journal.pone.0194708 April 18, 2018 16 / 20 Writing – review & editing: Ana B. Marı ´n-Arroyo, Joseba Rios-Garaizar, Lawrence G. Straus, Jennifer R. Jones, Marco de la Rasilla, Manuel R. Gonza ´lez Morales, Michael Richards, Jesu ´s Altuna, Koro Mariezkurrena, David Ocio. References 1. Hublin J. The modern human colonization of western Eurasia: when and where? Quat Sci Rev. 2015; 118: 194–210. https://doi.org/10.1016/j.quascirev.2014.08.011 2. Maroto J, Vaquero M, Arrizabalaga A, Baena J, Baquedano E, Jorda ´J, et al. Current issues in late Middle Palaeolithic chronology: New assessments from Northern Iberia. Quat Internat. 2012; 247:15–25. https://doi.org/10.1016/j.quaint.2011.07.007 3. Higham T, Douka K, Wood R, Ramsey CB, Brock F, Basell L, et al. The timing and spatiotemporal patterning of Neanderthal disappearance. Nature 2014; 512: 306–309. https://doi.org/10.1038/ nature13621 PMID: 25143113 4. Wood RE, Arrizabalaga A, Camps M, Fallon S, Iriarte-Chiapusso MJ, Jones R, et al. The chronology of the earliest Upper Palaeolithic in northern Iberia: New insights from l’Arbreda, Labeko Koba and La Viña. J Hum Evol. 2014; 69: 91–109. https://doi.org/10.1016/j.jhevol.2013.12.017 PMID: 24636733 5. Wood R, de Quiro ´s FB, Maı´llo-Ferna ´ndez JM, Tejero JM, Neira A, Higham T. El Castillo (Cantabria, northern Iberia) and the Transitional Aurignacian: Using radiocarbon dating to assess site taphonomy. Quat Int. Forthcoming; https://doi.org/10.1016/j.quaint.2016.03.005 6. Higham T, Jacobi R, Bronk Ramsey C. AMS radiocarbon dating of ancient bone using ultrafiltration. Radiocarbon 2006; 48: 179–195. https://doi.org/10.1017/S0033822200066388 7. Higham T. European Middle and Upper Palaeolithic radiocarbon dates are often older than they look: problems with previous dates and some remedies. Antiquity 2011; 85 (327): 235–249. https://doi.org/ 10.1017/S0003598X00067570 8. Bar-Yosef O, Bordes JG. Who were the makers of the Cha ˆtelperronian culture? J Hum Evol 2010; 59, 586–593. http://dx.doi.org/10.1016/j.jhevol.2010.06.009 PMID: 20692683 9. Benazzi S, Douka K, Fornai C, Bauer CC, Kullmer O, Svoboda J, et al. Early dispersal of modern humans in Europe and implications for Neanderthal behaviour. Nature 2011; 479, 525–528. https://doi. org/10.1038/nature10617 PMID: 22048311 10. Zilhão J, Banks W, d’Errico F, Gioia P. Analysis of Site Formation and Assemblage Integrity Does Not Support Attribution of the Uluzzian to Modern Humans at Grotta del Cavallo. Plos ONE 2015; 10(7): 131–181. https://doi.org/10.1371/journal.pone.0131181 PMID: 26154139 11. Ruebens K. Regional behaviour among late Neanderthal groups in Western Europe: a comparative assessment of late Middle Palaeolithic bifacial tool variability. J Hum Evol. 2013; 65: 341–362. https:// doi.org/10.1016/j.jhevol.2013.06.009 PMID: 23928352 12. Wood RE, Barroso-Ruiz C, Caparros M, Jorda ´JF, Galva ´n Santos B, Higham TF. Radiocarbon dating casts doubt on the late chronology of the Middle to Upper Palaeolithic transition in southern Iberia. Proc Natl Acad Sci U S A 2013; 110: 2781–2786. https://doi.org/10.1073/pnas.1207656110 PMID: 23382220 13. Wood R. The contribution of new radiocarbon dating pre-treatment techniques to understanding the Middle to Upper Palaeolithic transition in Iberia. University of Oxford. PhD Thesis. 2011. https://ora.ox. ac.uk/objects/uuid:075d79c6-edb4-4f19-9e34-50a63e7b7fe0 14. Altuna J, Baldeo ´n A, Mariezkurrena K. editors. La Cueva de Amalda (Zestoa, Paı ´s Vasco). Ocupaciones Paleolı´ticas y Postpaleolı´ticas. Donostia-San Sebastia ´n: Sociedad de Estudios Vascos; 1990. 15. Gonza ´lez-Urquijo JE, Iba ´ñez-Este ´vez JJ. Abrigo de Axlor (Dima). Arkeoikuska: Investigacio ´n arqueolo ´gica; 2001. pp.90–93. 16. Fortea J, de la Rasilla M, Rodrı ´guez V. La cueva de Llonı ´n (Llonı ´n, Peñamellera Alta). Campañas de 1991 a 1994. Excavaciones Arqueolo ´gicas en Asturias 1987–1990, Oviedo: Servicio de Publicaciones del Principado de Asturias; 1992; 2: pp. 9–18. 17. Fortea J, de la Rasilla M, Rodrı ´guez V. La cueva de Llonı ´n (Llonı ´n, Peñamellera Alta). Campañas de 1991 a 1994. Excavaciones Arqueolo ´gicas en Asturias 1991–1994; 1995; 3: pp. 33–44. 18. Fortea J, de la Rasilla M, Rodrı ´guez V. La cueva de Llonı ´n (Llonı ´n, Peñamellera Alta). Campañas de 1995 a 1998. Excavaciones Arqueolo ´gicas en Asturias 1995–1998; 1999; 4: pp. 59–68. 19. Fortea J, de la Rasilla M, Rodrı ´guez V. L’art parie ´tal et la se ´quence arche ´ologique pale ´olithique de la grotte de Llonı ´n (Peñamellera Alta, Asturies, Espagne). Pre ´histoire, Art et Socie ´te ´s. BSPA. 2004; LIX: 7–29. Chronological reassessment of the Middle to Upper Paleolithic transition in Cantabrian Spain PLOS ONE | https://doi.org/10.1371/journal.pone.0194708 April 18, 2018 17 / 20 20. Fortea J, de la Rasilla M, Rodrı ´guez V. La cueva de Llonı ´n (Llonı ´n, Peñamellera Alta). Campañas de 1999 a 2002. Excavaciones Arqueolo ´gicas en Asturias 1999–2002; 2007; 5: pp.77–86. 21. Altuna J, Merino J. El Yacimiento Prehisto ´rico de la Cueva de Ekain. San Sebastia ´n: Sociedad de Estudios. Vascos; 1984. 22. Altuna J, Mariezkurrena K, Rios-Garaizar J. Ocupaciones humanas en la cueva de Aitzbitarte III (Renteria, Paı ´s Vasco) sector Entrada: 33.000–18.000 BP. Vitoria-Gasteiz: Eusko Jaurlaritzaren Argitalpen Zerbitzu Nagusia; 2011. 23. Rasines del Rı ´o P. Geografı ´a, estratigrafı ´a y cronologı ´a de la cueva de Cobrante. Sautuola: Revista del Instituto de Prehistoria y Arqueologı ´a. 2009; 15: 43–47. 24. Muñoz Ferna ´ndez E, Serna Gancedo A. Los niveles solutrenses de la Cueva del Ruso I (Igollo de Camargo, Cantabria). Sautuola: Revista del Instituto de Prehistoria y Arqueologı ´a. 1999; 6: 161–176. 25. Echegaray JG. Cueva del Otero. Excavaciones Arqueolo ´gicas en España, 53. Madrid: Ministerio de Educacio ´n Nacional Direccio ´n General de Bellas Artes, Servicio Nacional de Excavaciones Arqueolo ´gicas; 1966. 26. Gonzalez Echegaray J. Freeman LG. Cueva Morı ´n: Excavaciones 1966–1968. Santander: Patronato de Cuevas Prehisto ´ricas; 1971. 27. Sanguino Gonza ´lez J, Montes Barquı ´n R. Nuevos datos para el conocimiento del Paleolı ´tico Medio en el centro de la Regio ´n Canta ´brica: La Cueva de Covalejos. Santander: Museo de Altamira Monografı ´as. 2005; 20: 10–38. 28. Barandiara ´n JM. Bolinkoba y otros yacimientos paleolı ´ticos en la sierra de Amboto (Vizcaya). Cuadernos de historia primitiva 5, 2. Madrid: Seminario de Historia Primitiva del Hombre; 1950. 29. Iriarte-Chiapusso MJ, Arrizabalaga A. El yacimiento arqueolo ´gico de Bolinkoba (Abadiño, Bizkaia). Cro ´nica de las investigaciones en la cavidad. Secuencia estratigra ´fica y cronologı ´a nume ´rica, In: Iriarte-Chiapusso MJ, Arrizabalaga A, editors. Bolinkoba (Abadiño) y su Yacimiento Arqueolo ´gico: Arqueologı ´a de La Arqueologı ´a Para La Puesta En Valor de Su Depo ´sito, a La Luz de Las Excavaciones Antiguas Y Recientes. Kobie Serie BAI 6. Diputacio ´n Foral de Bizkaia, Bilbao; 2015: pp. 5–88. 30. Muñoz Ferna ´ndez E, Montes Barquı ´n R. editors. Intervenciones arqueolo ´gicas en Castro Urdiales, tomo III. Arqueologı ´a y arte rupestre paleolı ´tico en las cavidades de El Cuco o Sobera y La Lastrilla. Castro Urdiales: Excmo. Ayuntamiento de Castro Urdiales. Concejalı ´a de Medio Ambiente y Patrimonio Arqueolo ´gico; 2007. 31. Gutierrez-Zugasti I, Rı ´os-Garaizar J, Marı ´n-Arroyo AB, Rasines P, Maroto J, Jones J, Bailey G, Richards M. Forthcoming. A chrono-cultural reassessment of the levels VI-XIV from El Cuco rock-shelter: a new sequence for the Late Middle Paleolithic in the Cantabrian Region (northern Iberia). Quat Int https://doi.org/10.1016/j.quaint.2017.06.059 32. Fortea J. Abrigo de La Viña. Informe de las campañas 1980–1986. Excavaciones Arqueolo ´gicas en Asturias (1983–1986). 1990; 1: 55–68. 33. Fortea J. Abrigo de La Viña. Informe de las campañas 1987–1990. Excavaciones Arqueolo ´gicas en Asturias (1987–1990). 1992; 2: 19–28. 34. Fortea J. Abrigo de La Viña. Informe y primera valoracio ´n de las campañas 1991 a 1994. Excavaciones Arqueolo ´gicas en Asturias (1991–1994). 1995; 3: 19–32. 35. Fortea J. Le Pale ´olithique supe ´rieur en Espagne, Galice et Asturies, 1991–1995. En M. Otte (dir.), Le pale ´olithique supe ´rieur Europe ´en: bilan quinquennal 1991–1996. ERAUL 76. Liège: Universite ´de Lie ´ge; 1996. pp. 329–339. 36. Fortea J. 1999. Abrigo de La Viña. Informe y primera valoracio ´n de las campañas de 1995 a 1998. Excavaciones Arqueolo ´gicas en Asturias (1995–1998), 4: 31–41. 37. Bradtmo ¨ller M. The Gravettian occupation of Level 4 Cueva Morı ´n and its regional context. Munibe. 2015; 66: 23–52. 38. Straus LG, Gonza ´lez Morales MR. El Miro ´n Cave (Ramales, Cantabria, Spain) date list V: Middle Paleolithic and Lower Magdalenian. Radiocarbon. 2016; 58(4): 943–945. https://doi.org/10.1017/RDC.2016. 84 39. Brock F, Higham T, Ramsey CB. Pre-screening techniques for identification of samples suitable for radiocarbon dating of poorly preserved bones. J Archaeol Sci. 2010; 37 (4): 855–865. https://doi.org/ 10.1016/j.jas.2009.11.015 40. Bronk Ramsey C, Higham T, Bowles A, Hedges R. Improvements to the pre-treatment of bone at Oxford. Radiocarbon. 2004; 46 (1): 155–163. https://doi.org/10.1017/S0033822200039473 41. Bronk Ramsey C. Bayesian analysis of radiocarbon dates. Radiocarbon. 2009; 51(1): 337–360. https:// doi.org/10.1017/S0033822200033865 Chronological reassessment of the Middle to Upper Paleolithic transition in Cantabrian Spain PLOS ONE | https://doi.org/10.1371/journal.pone.0194708 April 18, 2018 18 / 20 42. Reimer PJ, Bard E, Bayliss A, Beck JW, Blackwell PG, Bronk Ramsey C. et al. IntCal13 and Marine13 radiocarbon age calibration curves 0–50,000 years cal BP. Radiocarbon. 2013; 55 (4): 1869–1887. https://doi.org/10.2458/azu_js_rc.55.16947 43. Bronk Ramsey CB. Dealing with outliers and offsets in radiocarbon dating. Radiocarbon. 2009; 51 (3): 1023–1045. https://doi.org/10.1017/S0033822200034093 44. Andersen KK, Svensson A, Johnsen SJ, Rasmussen SO, Bigler M, Ro ¨thlisberger R, et al. The Greenland ice core chronology 2005, 15–42 ka. Part 1: constructing the time scale. Quat Sci Rev. 2006; 25: 3246–3257. https://doi.org/10.1016/j.quascirev.2006.08.002 45. Svensson A, Andersen KK, Bigler M, Clausen HB, Dahl-Jensen D, Davies SM, et al. The Greenland ice core chronology 2005, 15–42 ka. Part 2: comparison to other records. Quat Sci Rev. 2006; 25: 3258– 3267. https://doi.org/10.1016/j.quascirev.2006.08.003 46. DeNiro MJ. Postmortem preservation and alteration of in vivo bone collagen isotope ratios in relation to palaeodietary reconstruction. Nature. 1985; 317: 806–809. 47. Bradtmo ¨ller M. Ho ¨hlenlager des Gravettien—Muster jungpala ¨o¬lithischer Ho ¨hlennutzung am Beispiel des Gravettien Nordspaniens. Hamburg: Verlag Dr. Kovač; 2014. 48. Altuna J, Mariezkurrena K, de la Peña P, Rios-Garaizar J. Los niveles gravetienses de la cueva de Aitzbitarte III (Gipuzkoa). Industrias y faunas asociadas. In: de las Heras C., Lasheras J.A., Arrizabalaga A ´., de la Rasilla M. editors. Pensando El Gravetiense: Nuevos Datos Para La Regio ´n Canta ´brica En Su Contexto Peninsular Y Pirenaico. Monografı ´as Del Museo Nacional Y Centro de Investigacio ´n de Altamira, 23. Madrid: Ministerio de Educacio ´n, Cultura; 2013. pp. 184–204. 49. Rios-Garaizar J, de la Peña P, Maillo-Ferna ´ndez JM, 2013. El final del Auriñaciense y el comienzo del Gravetiense en la regio ´n canta ´brica: una visio ´n tecno-tipolo ´gica. In: de las Heras C, Lasheras JA, Arrizabalaga A ´, de la Rasilla M. editors. Pensando El Gravetiense: Nuevos Datos Para La Regio ´n Canta ´brica En Su Contexto Peninsular Y Pirenaico. Monografı ´as Del Museo Nacional Y Centro de Investigacio ´n de Altamira, N.o 23. Madrid; Ministerio de Educacio ´n, Cultura, 2013; pp. 369–382. 50. Bernaldo de Quiro ´s F. Los inicios del Paleolı ´tico Superior Canta ´brico. Santander: Monografı ´as del Museo y Centro de estudios de Altamira 8; 1982. 51. Muñoz Ferna ´ndez E, Santamarı ´a Santamarı ´a S. Ana ´lisis de la industria lı ´tica de la cueva de Cobrante. Sautuola 2009; 15: 145–189. 52. Muñoz Ferna ´ndez E. Excavaciones arqueolo ´gicas en la Cueva del Ruso I. Avance preliminar. Arquenas. 1991: 61–157. 53. Barandiara ´n Maestu I. Paleomesolı ´tico del Pirineo Occidental. Bases para la sistematizacio ´n tipolo ´gica del instrumental o ´seo paleolı ´tico. Monografı ´as Arqueolo ´gicas. Zaragoza: Universidad de Zaragoza; 1967. 54. Altuna J. Situacio ´n y descripcio ´n de la cueva de Amalda. Historia de las excavaciones. Descripcio ´n del relleno. Estructuras en el yacimiento. Dataciones de radiocarbono. Otros yacimientos del valle. In: Altuna J, Baldeo ´n A, Mariezkurrena K. editors. La Cueva de Amalda (Zestoa, Paı ´s Vasco). Ocupaciones Paleolı ´ticas Y Postpaleolı ´ticas. Sociedad de Estudios Vascos, Donostia-San Sebastia ´n, 1990. pp. 9–31. 55. Rios-Garaizar J. Organizacio ´n econo ´mica de las sociedades Neandertales: el caso del nivel VII de Amalda (Zestoa, Gipuzkoa). Zephyrus. 2010; LXV, 15–37. 56. Rios-Garaizar J, de la Peña P, San Emeterio A. Estudio de las industrias lı ´ticas y o ´seas de la cueva de Aitzbitarte III (Zona de la entrada). In: Altuna J., Mariezkurrena K., Rios-Garaizar J. editors. Ocupaciones Humanas En La Cueva de Aitzbitarte III (Renteria, Paı ´s Vasco) Sector Entrada: 33.000–18.000 BP. Vitoria-Gasteiz: Eusko Jaurlaritzaren Argitalpen Zerbitzu Nagusia, 2011. pp. 81–351. 57. Rios-Garaizar J, Arrizabalaga A ´, Villaluenga A. Haltes de chasse du Cha ˆtelperronien de la Pe ´ninsule Ibe ´rique. Labeko Koba et Ekain (Pays Basque Pe ´ninsulaire). L’Anthropologie 2012; 116: 532–549. https://doi.org/10.1016/j.anthro.2012.10.001 58. Camps M, Higham T. Chronology of the Middle to Upper Palaeolithic transition at Abric Romanı ´, Catalunya. J Hum Evol. 2012; 62 (1): 89–103. https://doi.org/10.1016/j.jhevol.2011.10.010 PMID: 22137586 59. Zilhão J. Chronostratigraphy of the Middle-to-Upper Paleolithic Transition in the Iberian Peninsula. Pyrenae 2006; 37(1): 7–84. 60. Lowe J, Barton N, Blockley S, Bronk Ramsey C, Cullen V, Davies W, et al. Volcanic ash layers illuminate the resilience of Neanderthals and early modern humans to natural hazards. Proc Natl Acad Sci U S A 2012; 109 (34): 13532–13537. https://doi.org/10.1073/pnas.1204579109 PMID: 22826222 61. Fedele FG, Giaccio B, Hajdas I. Timescales and cultural process at 40,000 BP in the light of the Campanian Ignimbrite eruption, Western Eurasia. J Hum Evol 2008; 55: 834–857. https://doi.org/10.1016/j. jhevol.2008.08.012 PMID: 18922561 Chronological reassessment of the Middle to Upper Paleolithic transition in Cantabrian Spain PLOS ONE | https://doi.org/10.1371/journal.pone.0194708 April 18, 2018 19 / 20 62. Golovanova LV, Borisovich Doronichev V, Elansia Cleghorn N, Koulkova M, Sapelko T, Shackley MS. Significance of ecological factors in the Middle to Upper Paleolithic Transition. Current Anthropolology 2010; 51: 655–691. https://doi.org/10.1086/656185 63. Maı ´llo-Ferna ´ndez JM, Bernaldo de Quiro ´s F. Archaic Aurignacian in El Castillo cave (Spain): technology and typology composition. Anthropologie. 2010; 114 (1): 1–25. https://doi.org/10.1016/j.anthro. 2010.01.001 64. Yravedra J. New Contributions on Subsistence Practices during the Middle-Upper Paleolithic in Northern Spain. In: Clark J, Speth J. (Eds). Zooarchaeology and Modern Human Origins. Vertebrate Paleobiology and Paleoanthropology book series. Springer; 2013. pp 77–95. 65. van Klinken GJ. Bone Collagen Quality Indicators for Palaeodietary and Radiocarbon Measurements. J Archaeol Sci 1999; 26(6), 687–695. https://doi.org/10.1006/jasc.1998.0385 Chronological reassessment of the Middle to Upper Paleolithic transition in Cantabrian Spain PLOS ONE | https://doi.org/10.1371/journal.pone.0194708 April 18, 2018 20 / 20