Archaeology of the Pleistocene-Holocene transition in Portugal : Synthesis and prospects

Contents lists available at ScienceDirect Quaternary International journal homepage: www.elsevier.com/locate/quaint Archaeology of the Pleistocene-Holocene transition in Portugal: Synthesis and prospects Cristina Gameiro a,∗ , Thierry Aubry a,b , Francisco Almeida c , Luca Dimuccio a,d , Sónia Gabriel a,k,l , Rita Gaspar e,f , Sérgio Gomes g , Ramón Fábregas Valcarce f , Sofia Figueiredo m , Carmen Manzano h , João Marreiros i,j , Cláudia Oliveira e,k , André Tomás Santos a,b , Maria João Silva 1 , João Pedro Tereso a,e,k , Pedro Xavier m a UNIARQ - Centre for Archaeology, School of Arts and Humanities, University of Lisbon, Alameda da Universidade - Campo Grande, 1600-214, Lisboa, Portugal b Fundação Côa Parque, Portugal c Taungurung Clans Aboriginal Corporation, Australia d Centre of Studies in Geography and Spatial Planning (CEGOT), FLUC, Department of Geography and Tourism, Portugal e MHNC-UP, Natural History and Science Museum of the University of Porto, Portugal f GEPN-AAT, Grupo de Estudios para a Prehistoria do Noroeste Ibérico, Arqueoloxia, Antigüidade e Territorio, University of Santiago de Compostela, Spain g CEAACP, University of Coimbra, Portugal h Arqueologia & Património, Portugal i TraCEr, Laboratory for Traceology and Controlled Experiments, MONREPOS. Archaeological Research Centre and Museum for Human Behavioural Evolution. RGZM, Germany j ICArEHB - Interdisciplinary Center for Archaeology and Evolution of Human Behaviour, University of Algarve, Portugal k InBIO, Research Network in Biodiversity and Evolutionary Biology (Associated Laboratory), CIBIO - Research Center in Biodiversity and Genetic Resources/University of Porto, Portugal l LARC, Direção Geral do Património Cultural, Portugal m Lab2PT, Laboratório de Paisagens, Património e Território, Minho University, Portugal ARTICLE INFO Keywords: Tardiglacial Chronostratigraphy Palaeoenvironment Lithic technology Art ABSTRACT The Tardiglacial of Portugal has been associated with the Magdalenian culture and lithic industries characterized by tool miniaturization, a diversity of microlith types, and the absence of a intentional blade production. The technological characterization, the chronology and the phasing of the Portuguese Magdalenian have been defined based on data recovered from open-air sites of the Estremadura region (Central Portugal). This paper presents an overview of the research undertaken over the last twenty-five years, including results from research and preventive archaeology fieldwork outside this region, namely in the Côa, Sabor and Vouga Valleys (northern Portugal), as well as in the Guadiana Valley and Algarve regions (southern Portugal). Our chronological boundaries are the Greenland Stadial 2-1b and the 8.2 ka event, from Early Magdalenian to Early Mesolithic. Regarding vegetation, deciduous Quercus underwent expansion during the warm phases of the Tardiglacial and retracted during cold ones, when pines increased. After the Solutrean, the faunal assemblages show a decrease in the variability of the represented species and an increase in fish, birds, small mammals and rabbits (Oryctolagus cuniculus). Concerning the cultural sequence, the Middle Magdalenian remains uncharacterised. After the Upper Magdalenian, and thenceforward, the use of local raw materials and of cores-on-flakes (burin or carinated endscraper type) for bladelet production gradually increased. In terms of lithic armatures typology, a four-stage sequence can be discerned: 1) Upper Magdalenian with axial points rather than backed bladelets, quite common in previous phases; 2) Final Magdalenian with an increase in the diversity of armature types; 3) Azilian with geometric microliths, curved backed points (Azilian points) and Malaurie points, and 4) Early Mesolithic without retouched bladelet tools or at best a persistence of Azilian armature types. There were some changes in the Palaeolithic rock art of the Douro basin between phase 3 (Final Magdalenian) https://doi.org/10.1016/j.quaint.2020.03.018 Received 18 November 2019; Received in revised form 20 February 2020; Accepted 12 March 2020 ∗ Corresponding author. E-mail addresses: [email protected], [email protected] (C. Gameiro). 1 Independent researcher Quaternary International xxx (xxxx) xxx–xxx 1040-6182/ © 2020 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/). Please cite this article as: Cristina Gameiro, et al., Quaternary International, https://doi.org/10.1016/j.quaint.2020.03.018 and phase 4 (Late Azilian): figurative animal representations give place to animal depictions characterized by their geometrical bodies, often filled-in, and red deer becomes the best-represented animal. 1. Introduction The origins of Palaeolithic research in Portugal are closely connected to the French school, which accounts for the adoption of the terminology and the phasing of the ‘classic’ Périgord sequence (Zilhão, 1997a; Gameiro, 2012). Heleno (1944) searched for the “European origins of the Portuguese people” and the presence in Portugal of French researchers (Breuil, Zbyszewski and Roche), and their collaboration in a number of archaeological projects, reinforced this connection. According to Zilhão (1997a,b), the techno-typological variability and the dates obtained for the Portuguese Estremadura define Early (19,500–16,000 cal BP), Upper (between 14,500 and 13,000 cal BP) and Final Magdalenian phases (13,000–11,500 cal BP), whose characterization, however, does not follow the ‘classic’ Magdalenian subdivisions of Southwestern France (Breuil, 1913; Sonneville-Bordes, 1960). In Portugal, the absence (or scarcity) of remains asigned to the Middle Magdalenian is arguably due to a sedimentary hiatus (Zilhão, 1997a). The early phase is divided into two diachronic technological facies (Cabeço de Porto Marinho and Cerrado Novo), while the final phase includes the Rossio do Cabo facies, characterized by a production scheme of small twisted bladelets with marginal retouch, and the Carneira facies, defined by Malaurie points, Azilian points and trapezoid geometrics. According to Zilhão (1997a,b) proposal, the Final Magdalenian would end during the Pre-Boreal (now Greenlandian stage). During the Early Mesolithic, at Areeiro III site, the lithic industry is characterised by carinated “endscraper” cores for the production of Dufour and Areeiro bladelets. Despite the existence of curved backed points, an Azilian index fossil, the absence of bone industry, as well as the technological continuity between the Franco-Cantabrian Magdalenian and the Azilian, made the differentiation of an Azilian period impossible (Zilhão, 1997a, p. 45). Bicho (1994) initially recognized two technological facies only, Rio Maior, which spans the entire Magdalenian, and Carinated, which starts around 12,500 cal BP and would still be present in Pre-Boreal Areeiro III, which he classifies as Epipaleolithic, thus stressing the technological continuity with the previous period. This author has since abandoned this facies concept, and used Early, Late and Final Magdalenian instead, emphasizing the absence of data supporting a Middle Magdalenian (Bicho, 1997). Lately, Bicho and Haws (2012), argued that Portuguese Magdalenian is a homogeneous cultural reality, with no ruptures in terms of technology, subsistence or settlement and spanning the 20,000–10,000 cal BP interval. Based only on the Cabeço de Porto Marinho sequence, Marks and Mishoe (1997) established a distinction between Early Magdalenian and Late Magdalenian, the latter phase being differentiated by an increase in the typological variability of microliths. The scarcity of tools on blades, when compared with the rest of Southwestern Europe, is used as an argument to individualize this group since this technical choice is unrelated to the availability of flint, which is abundant in the vicinity of the site (Marks, 2000). In recent years, the study of the final Upper Palaeolithic blank production methods in the lower Côa Valley and Portuguese Estremadura (Gameiro, 2012) has confirmed the phasing proposed by Zilhão (1997a), particularly the existence of a production of small bladelets of the Areeiro type using cores-on-flake (carinated “endscrapers” and “burins”) during the ca. 13,500–12,000 cal BP interval, and the existence of the Rossio do Cabo facies. Lately, the use of the term Azilian has been proposed for the Côa Valley industries (Aubry et al., 2017), formerly called “Carneira facies” Final Magdalenian (Zilhão, 1997a,b). Araújo (2016b), in her thesis on the Early Mesolithic of Portugal, established a chrono-cultural framework for this cultural phase, confirming the rarity of retouched bladelet tools in the open-air sites dated to the Early Holocene, with the exception of the lithic assemblage of Areeiro III, associated with radiocarbon dates of ca. 9500-9000 cal BP and showing a strong technical and typological convergence with the Rossio do Cabo facies. Nevertheless, the available data do not allow a clear definition of the characteristics of the transitional lithic industries. 2. Materials and methods This paper presents a synthesis of paleoenvironmental data and summarizes some aspects of material culture (lithic industries and art) dating from between 20,000 and 8200 cal BP, i.e. between GS-2.1b and the 8.2 ka event (Rasmussen et al., 2014). Nevertheless, the rare Early and Middle Magdalenian, as well as the Early Mesolithic sites, are only mentioned when deemed useful to understanding the long-term cultural process; therefore, focus will be centered on the period between 14,000 and 11,500 cal BP. A significant number of Tardiglacial sites have been identified in recent years, not only as a result of ongoing research projects in Estremadura, the Côa Valley and the Algarve regions, but mostly in the scope of preventive archaeology works: in the Sabor Valley (north of the Côa Valley), in the Vouga Valley (halfway between Estremadura and the Côa Valley) and in the Guadiana Valley (to the South) (Fig. 1). However, existing data continue to be partial and resulting from single occupation open-air sites; presently, a complete stratigraphic sequence covering the Pleistocene-Holocene transition (14,000 to 10,500 cal BP) is yet to be identified. The available Paleobotany data provide a broad perspective of the Portuguese territory during the period in question but not enough detail is yet available to establish the correlation with climate change events documented in the Greenland ice-core records (Rasmussen et al., 2014). An overview on the faunal assemblages known to date will be presented as well. Nevertheless, the degree of preservation of bone material varies according to region; apart from Fariseu (Côa Valley) no other sites from the Iberian massif have preserved faunal remains. Only in the Estremadura (Ocidental Meso-Cenozoic basin) and Algarve (Meridional Meso-Cenozoic basin) is faunal preservation documented. The study of lithic industries follows the chaîne opératoire approach and is presented according to the six main geographic units referred above (represented in Fig. 1), individualized according not only to their history of research, but also in terms of their distinct lithological environments. The map in Fig. 1 shows virtually all the known Late Pleistocene and Early Holocene sites. Sites featuring sequences with both Late Pleistocene and Early Holocene occupations, as well as those where chronology within this time span is yet to be defined, were included in the Pleistocene-Holocene transition category. 3. Paleoenvironmental data 3.1. Paleoclimate During the Late Pleistocene, human societies were affected by deglaciation, sea-level rise and climatic instability in the North Atlantic, Europe and Greenland at a global or planetary scale (Naughton et al., 2016). Understanding human cultural change and establishing a correlation between the observed technical and cultural adaptations and paleoclimatic history has been a major challenge for prehistorians. During the last decades, considerable progress has been made in the C. Gameiro, et al. Quaternary International xxx (xxxx) xxx–xxx 2 (caption on next page) C. Gameiro, et al. Quaternary International xxx (xxxx) xxx–xxx 3 study of climate changes, which were more frequent than previously thought (Rasmussen et al., 2014). For the period of concern here, a multi-proxy study comparing terrestrial and marine data from Northwest Iberia improved the characterization of its different climatic phases (Naughton et al., 2016): Oldest Dryas, which was a complex event with three phases, (a) extremely cold/relatively wet, (b) cool/ dry, and (c) warmer/increased moisture; Bølling-Allerød interstadial (increase in temperature and precipitation); Younger Dryas (cold and dry; winter precipitation in northern and central Western Iberia); Holocene interglacial. A recent study by S. Davis (2019) on the variation in the size of rabbit bones - both modern (from Europe) and archaeological (from Portugal) - indicates that during the last glaciation the present-day Portuguese territory was some 7 or 8° colder than nowadays. 3.2. Chronostratigraphy Estremadura remains the best-known area when it comes to correlating sedimentary sequences and palaeoenvironmental events (Angelucci, 2002; Cunha et al., 2006; Aubry et al., 2011, 2017). Levels corresponding to the Early, Upper and Final Magdalenian have been dated by radiocarbon and thermoluminescence in different loci of the archaeological site of Cabeço de Porto Marinho (Fig. 1-n°29), within a sequence of colluviated aeolian sands (Zilhão, 1997a). The other openair sites are single-phase occupations with no preservation of spatial structure (Zilhão, 1997a; Angelucci, 2002; Almeida, 2008; Aubry et al., 2011, 2015). Cave occupations are scarcer than in Upper Palaeolithic chronologies: the cave sites known or claimed to have been used during the Magdalenian are Gruta do Caldeirão (Fig. 1-n°26), Buraca Grande (Fig. 1-n°28), Lapa do Picareiro (Fig. 1-n°23), Lapa do Suão (Fig. 1n°15) and Lapa dos Coelhos (Fig. 1-n°22), to which a couple of rockshelters, e.g., Abrigo 1 de Vale dos Covões (Fig. 1-n°29), and Abrigo das Bocas (Fig. 1-n°17), can be added (Bicho et al., 2003; Almeida et al., 2004; Aubry et al., 2011, 2017; Gameiro et al., 2017). One of the following situations can be observed in these sequences: thin sedimentary deposits accumulated during the Bølling-Allerød stabilization phase, Fig. 1. - Map of Late Pleistocene and Early Holocene archaeological sites in Portugal, according to data obtained from the Endovélico National Database, Cultural Heritage, Ministry of Culture. Late Pleistocene sites: 1-Vale Santo 4; 2-Praia da Galé; 3-Vale Boi; 4-Monte do Januário; 5-Cruz da Pedra; 6Monte da Ribeira; 7 -Chancudo 3; 9-Cabeço dos Cinco Pinheiros; 10-Baião; 11-Vale da Mata; 12-Cerrado Novo; 13-Pinheirocas 2; 14-Rossio do Cabo; 15-Lapa do Suão; 16-Pinhal da Carneira; 18-Vascas; 19-Cabeço de Porto Marinho; 20-Cadoiças; 21-Bairrada; 22-Lapa dos Coelhos (Almonda); 25-Olival Fechado; 26-Gruta do Caldeirão; 29Abrigo 1 de Vale de Covões; 33-Ínsula II; 34-Cardina I/II; 35-Fonte do Olmo; 36-Devesa 3; 39-Foz do Medal; 40-Telheiro da Barreira; 45-Pedra do Patacho; PleistoceneHolocene transition sites: 8-Malhada do Mercador 2; 17-Abrigo Grande das Bocas; 23-Lapa do Picareiro; 28-Buraca Grande; 30-Rôdo; 31-Bispeira 8; 32-Vau; 37Quinta da Barca Sul; 38-Fariseu; 46-Monte Roncanito 21; 58-Volta do Cocão; Early Holocene sites: 24-Costa do Pereiro; 27-Cruz da Areia; 41-Armação Nova/Rocha das Gaivotas; 42Barranco das Quebradas 1; 43-Castelejo; 44-Palheirões do Alegra; 47-Barca do Xerez de Baixo; 48-S. Julião; 49-Pinhal da Fonte; 50-Cabeço do Curral Velho; 51-Ponta da Vigia; 52-Toledo; 53-Vale Frade; 54-Areeiro III; 55-Fonte Pinheiro; 56-Casal do Papagaio; 57-Prazo; 59-Magoito; 60-Cova da Baleia; 61Vale Sá. Fig. 2. Stratigraphic layout of the archaeological site of Fariseu (Côa Valley) (Aubry, 2009). C. Gameiro, et al. Quaternary International xxx (xxxx) xxx–xxx 4 culturally attributed to the Upper/Final Magdalenian (Lapa dos Coelhos, Abrigo 1 de Vale dos Covões and Lapa do Picareiro) or palimpsest levels containing materials of different chronologies (such as Buraca Grande and Lapa do Suão). At Gruta do Caldeirão (Zilhão, 1997a), a sedimentary gap was identified, postdating the Solutrean and corresponding to the period between 21,000 and 17,500 cal BP characterized by a humid cool climate phase, in correspondence with the GS-2.1b/GS-2.1a of Rasmussen et al. (2014). This general pattern might explain the rarity of Early Magdalenian contexts, the redeposition of Solutrean materials in alluvial plain contexts, documented at Olival da Carneira, Arneiro and Passal (Zilhão, 1997a). Later, a sedimentary hiatus corresponding to the period between 16,000 and 14,500 cal BP might account for the difficulty in isolating contexts ascribable to the Middle Magdalenian. From the Allerød onwards, the number of known sites increases significantly. A new cold episode, probably GS1 (Rasmussen et al., 2014), likely led to an increase in erosive processes between 13,000 and 11,500 cal BP, which may account for the erosional scattering of such Final Magdalenian open-air contexts as Bairrada. In the Côa Valley, the Fariseu site (Fig. 1-n°38), located at the interface between alluvial plain and slope, yielded a paleoenvironmental and archaeostratigraphic (Fig. 2) record covering the time span between 18,500 and 11,500 cal BP (Mercier et al., 2006; Aubry, 2009; Aubry et al., 2010, 2015). The micromorphological study (Sellami, 2009) and the radiocarbon, TL and OSL datings obtained at this site (Mercier et al., 2006; Aubry, 2009) support the idea of a correspondence between mobiliary and rock art representations, as well as correlation between the alternating alluvial and slope sedimentation phases and the climate oscillations of the Tardiglacial identified in deep-sea records (Aubry et al., 2010). Cryoclasts present in layers 7/8 and layer 4 of the sequence may be related to Heinrich 1 (ca. 16,000–14,000 cal BP) and Heinrich 0 (12,000–11,500 cal BP) events, while the alluvial deposits from layers 5 and 6 would correspond to the climate improvement of the Bølling-Allerød interstadial. The TL and OSL dates of Fariseu layer 4 (Table 1), obtained on fragments of burnt quartzite and sediment, are consistent with their stratigraphic position and in agreement with the 14 C dates on bone samples (Aubry, 2009). The formation of this unit by slope processes, resulting from cryoclastic degradation of metasediments, mostly accumulated at the base of the slope, at the edge of the alluvial plain, is identical in nature to layer 3 of Quinta da Barca Sul (Fig. 1-no. 37) (Aubry et al., 2010); the respective archaeological contents are similar as well (Aubry, 2009; Gameiro, 2009, 2012). 3.3. Vegetation dynamics in the Pleistocene-Holocene transition Pollen sequences covering the Pleistocene-Holocene transition were recovered in 13 sites (Supplementary material -Table 1). Eight sites are located in the Serra da Estrela mountain range; the mainland areas of southern Portugal and the central coast were not sampled. Still, and overall, sequences are distributed throughout the study area and sampling sites are diverse, allowing the detection of a variety of ecological trends. Data from Dryas I was retrieved only in Lagoa de Marinho, in the NW, where low values of arboreal pollen (deciduous Quercus and Pinus sylvestris) and high values of shrubby and herbaceous taxa such as Erica, Poaceae and Artemisia suggest an open mountain vegetation (Ramil Rego et al., 1998). There is no data regarding lower elevation areas. The beginning of the Bølling-Dryas II-Allerød is usually characterized by an increase in arboreal vegetation, namely deciduous Quercus in the NW and Pinus (mostly P. sylvestris but also some P. pinaster) in the sandy soils of the southwest coast (Queiroz, 1999). The Serra da Estrela sequences indicate an increasing presence of trees (Betula and Pinus) and shrubs near the sampling site (van der Knaap and van Leeuwen, 1997). The expansion of Quercus in the NW seems to have been slightly interrupted during Dryas II, and P. sylvestris expanded, decreasing later on, in the Dryas III, during which there is an expansion of grasslands due to climate forcing (Ramil Rego et al., 1998). At Lagoa do Golfo and Lagoa de Santo André, in the SW, P. sylvestris dominated throughout this period. During the Allerød, there is a marked expansion of woodlands, from north to south; deciduous Quercus dominate these woodlands, except in the SW coast, where pines continue to dominate and expand, with an understorey of Ericaceae (Santos and Sánchez-Goñi, 2003). But even here there is an expansion of deciduous Quercus and several hygrophilous and mesophilous taxa such as Salix, Alnus and Corylus (Queiroz, 1999). In mountain areas of central Portugal, such as Serra da Estrela, there is a decline in Poaceae and, overall, pioneer taxa (van der Knaap and van Leeuwen, 1997). We would stress that in the southernmost sequence, CM5 Beliche, thermomediterranean taxa such as Olea, Pistacia and Phillyrea remain absent (Fletcher et al., 2007). The expansion of woodlands was abruptly interrupted during Dryas III. The declining temperatures and dryness of this period led to major responses in the vegetation. The decline of deciduous oak woodlands is visible in all pollen sequences, leading to an increase in xerophytic taxa, mostly herbaceous plants. Some sites present specific trends, due to their location. At Lagoa de Marinho, in the NW, even P. sylvestris values decrease significantly (Ramil Rego et al., 1998), but in the highmountain sequence of Serra da Estrela there is an increase in Pinus and Artemisia (van der Knaap and van Leeuwen, 1997). The Chã das Lameiras pedoanthracological data demonstrate that, along with the increasing presence of xerophytic herbaceous communities documented by palynology, there is an increasing presence of shrubby Fabaceae (López-Sáez et al., 2017). In the south-western and southern coasts, the abrupt decline in Quercus is accompanied by an increase of Pinus. At Cabeço do Porto Marinho, in central Portugal, charcoal analyses Table 1 TL/OSL dates for Portuguese Magdalenian and Pleistocene-Holocene transition sites (after Aubry, 2009; Bicho and Haws, 2012; Gaspar et al., 2016b). Site Layer Lab ref. Method Material Age (ka) Quinta da Barca Sul 3 (UA8) GifTLQBS4 TL Burnt quartzite 11.9 ± 1.1 Quinta da Barca Sul 3 (UA7) GifTLQBS3 TL Burnt quartzite 11.6 ± 1.2 Quinta da Barca Sul 3 (UA5) GifTLQBS2 TL Burnt quartzite 12.7 ± 1.0 Fariseu 4b Cobble 17 TL Burnt quartzite 11 ± 1.1 Fariseu 4c Cobble 2 TL Burnt quartzite 10.8 ± 1.7 Fariseu 4e Cobble 8 TL Burnt quartzite 11.8 ± 0.9 Fariseu 6 Sedc6 OSL Sediments 15.2 ± 1.6 Fariseu 7 Cobble 24 TL Burnt quartzite 13.7 ± 1.0 Fariseu 7 Cobble 9 TL Burnt quartzite 14.3 ± 1.1 Fariseu 8 Sedc8 OSL Sediments 18.4 ± 1.6 Carneira II 60–70 TL Flint 10.5 ± 0.7 CPM 13 TL Flint 14.1 ± 1.1 Lagoa do Bordoal LEPT2 OSL Quartz 14.8 ± 2.9 Foz do Medal 1034 X6566 OSL Sediments 12.35 ± 0.9 C. Gameiro, et al. Quaternary International xxx (xxxx) xxx–xxx 5 Table 2 Radiocarbon dates for Portuguese Magdalenian, Pleistocene-Holocene transition sites and Early Holocene (after Zilhão, 1997a; Monteiro-Rodrigues and Angelucci, 2004; Aubry et al., 2008a; Valente and Carvalho, 2009; Bicho and Haws, 2012; Pereiro, 2013; Araújo et al., 2014; Sousa and Gonçalves, 2015; Araújo, 2016b; Carvalho et al., 2016; Benedetti et al., 2019 and unpublished data from Rôdo and Vau provided by Carmen Manzano and Sérgio Gomes). Because of the reservoir effect, dates on shell were not considered. CPM = Cabeço de Porto Marinho; VCA1 = Abrigo 1 de Vale de Covões; BXB= Barca do Xerez de Baixo * 2σ calibration according to OxCal v4.3.2 Bronk Ramsey (2017); r-5; IntCal 13 Atmospheric curve (Reimer et al., 2013). Site Layer Lab Ref. Sample Date BP Date cal BP (2σ) * CPM-I Lower SMU-2015 Charcoal 16,340 ± 420 20,770–18,795 Caldeirão Fa ICEN-69 Bone 15,170 ± 740 20,355–16,620 CPM-IIISW Lower WK-3126 Charcoal 16,180 ± 290 20,232–18,868 CPM-I Lower ICEN-542 Charcoal - hearth 15,820 ± 400 20,150–18,309 Caldeirão Eb ICEN-70 Bone 14,450 ± 890 19,966–15,243 CPM-IIIS Lower SMU-2668 Charcoal 14,050 ± 850 19,385–14,666 CPM-II Middle SMU-2476 Charcoal 15,410 ± 195 19,108–18,223 CPM-VI Lower SMU-2634 Charcoal 15,420 ± 180 19,070–18,275 CPM-IIIS Lower WK-3127 Charcoal 15,040 ± 210 18,747–1,7829 Olival Fechado 2 Beta-179279 Charcoal - Pinus sylvestris 13,460 ± 80 16,488–15,942 Buraca Grande 9 OxA-5522 Baguette demi-ronde 13,050 ± 100 15,957–15,287 Rôdo 6/NA1 17C/0829 Leguminosae 13,050 ± 40 15,832–15,383 Lapa do Picareiro G OxA-5527 Charcoal - Pinus 12,320 ± 90 14,824–14,011 VCA-1 5 Beta-201013 Bone 12,340 ± 50 14,686–14,102 Bispeira 8 5 17C/0838 Leguminosae 12,010 ± 60 14,686–14,102 CPM-I Upper ICEN-687 Charcoal 12,220 ± 110 14,675–13,779 Lapa dos Coelhos 4 GrN-18377 Charcoal - Pinus sylvestris 12,240 ± 60 14,465–13,955 Lapa do Picareiro F Wk-6677 Charcoal 12,210 ± 60 14,336–13,860 VCA-1 8b Gif-12080 Charcoal 12,220 ± 24 14,223–14,008 VCA-1 8b UA33482 Charcoal 12,050 ± 70 14,088–13,749 CPM-IIIS Middle ICEN-689 Charcoal - hearth 11,810 ± 110 13,945–13,427 Lapa do Picareiro F Wk-4219 Charcoal 11,780 ± 90 13,772–13,442 Lapa do Picareiro E-Middle Wk-5431 Charcoal 11,700 ± 120 13,764–13,296 Galeria da Cisterna 3 OxA-11129 Phalanx - Capra sp. 11,755 ± 80 13,720–13,445 CPM-I Upper SMU-2011 Charcoal 11,680 ± 60 13,709–13,380 Bispeira 8 4 17C/0837 Leguminosae 11,680 ± 60 13,709–13,380 Lapa do Picareiro E-Lower Wk-4218 Charcoal 11,550 ± 120 13,605–13,119 Lapa dos Coelhos 3 GrN-18376 Bone - Cervus elaphus 11,660 ± 60 13,597–13,340 Buraca Grande GC2 Gif-96307 Bone 11,390 ± 110 13,446–13,066 Caldeirão Eb ICEN-72 Bone 10,700 ± 380 13,316–11,393 CPMIIIT Upper ICEN-690 Charcoal - hearth 10,940 ± 210 13,270–12,428 CPMIII Upper ICEN-545 Charcoal 11,160 ± 130 13,261–12,748 CPMII Upper SMU-2637 Charcoal 11,110 ± 130 13,198–12,724 Pinhal da Carneira 4 SMU-2635 Charcoal - hearth 10,880 ± 90 12,995–12,670 Lapa do Suão 7 Gx-27590 Charcoal 10,900 ± 70 12,961–12,692 Galeria da Cisterna 3 GrA −9722 Bone 10,820 ± 60 12,810–12,651 VCA1 5 UA33479 Bone 10,540 ± 95 12,695–12,140 Fariseu 4 Beta −213130 Bone 10,510 ± 40 12,601–12,244 Bocas 1 0+ ICEN-900 Bone - Bos primigenius 9880 ± 220 12,150–10,671 CPM VI Middle SMU-2636 Charcoal 10,160 ± 80 12,100–11,404 Bocas 1 Fundo ICEN-901 Bone 10,110 ± 90 12,047–11,324 VCA1 8b UA24593 Bone 10,010 ± 90 11,940–11,240 Magoito 1B C ICEN-82 Charcoal 9910 ± 100 11,767–11,171 Fariseu 4 Ua-32645 Bone 9830 ± 130 11,759–10,781 VCA1 4 Beta-201014 Charcoal 10,020 ± 40 11,718–11,316 VCA1 7 UA33480 Bone 9725 ± 75 11,264–10,786 Casal do Papagaio Base ICEN-369 Bone 9710 ± 70 11,247–10,787 Casal do Papagaio Middle ICEN-372 Charcoal 9650 ± 90 11,220–10,739 Vau 5/NA2 17C/0813 Leguminosae 9680 ± 40 11,210–10,805 Magoito 1B ICEN-577 Charcoal 9490 ± 60 11,083–10,580 Prazo C5A GrA-15861 Charred Pinecone fr. 9410 ± 70 11,068–10,425 Prazo C5–C6/1 Ua-20495 Charcoal Fraxinous sp. 9525 ± 70 11,126–10,595 VCA1 8b UA33481 Bone 9315 ± 90 10,724–10,254 Rôdo 6/NA1 17C/0825 Leguminosae 9360 ± 40 10,696–10,443 Cardina 4.2 Beta-460529 Bone 9220 ± 30 10,400–10,240 Cardina 4.2 Beta-460528 Bone 9160 ± 30 10,500–10,260 Rôdo 6/NA1 17C/0824 Leguminosae 9050 ± 40 10,254–10,176 Fariseu 4 GX-32147-MAS Chamois Teeth 8930 ± 80 10,235–9765 Areeiro III Structure 1 ICEN-546 Charcoal 8570 ± 130 10,125–9275 Areeiro III Main level ICEN-547 Charcoal 8860 ± 80 10,199–9679 Ponta da Vigia Hearth 2 Sac-1747 Charcoal - Pinus pinaster 8850 ± 90 10,198–9629 Areeiro III Structure 2 ICEN-494 Charcoal 8850 ± 50 10,168–9710 Ponta da Vigia Hearth 2 ICEN-51 Charcoal 8730 ± 110 10,153-9535 Ponta da Vigia Hearth 3 Sac-1741 Charcoal - Pinus pinaster 8670 ± 80 9913–9501 BXB Structure E OxA-13266 Charcoal - Q. coccifera 8729 ± 36 9887–9556 Cruz da Areia Structure 3 Beta 343555 Charcoal – Pinus pinaster 8720 ± 40 9820–9551 Areeiro III Main level ICEN-688 Charcoal 8380 ± 90 9533–9137 BXB Combustion area Beta-120607 Charcoal 8640 ± 50 9732–9527 BXB Structure A OxA-13406 Charcoal - Q. coccifera 8150 ± 40 9252–9006 BXB Structure A OxA-13265 Charcoal - E. arborea 8248 ± 35 9401–9091 (continued on next page) C. Gameiro, et al. Quaternary International xxx (xxxx) xxx–xxx 6 indicate the presence of pines and a shrubby vegetation, including some indicators of thermicity, such as Olea europaea and Arbutus unedo (Zilhão et al., 1995). Some of these taxa have also been identified at nearby Lapa do Picareiro (Bicho et al., 2003). The onset of the Holocene is marked by new and abrupt changes in vegetation. From north to south there is a great expansion of Quercus. However, as opposed to the Late Glacial climatic warmings, evergreen Quercus soon gains some relevance and in the southernmost sequence of CM5 Beliche it is even more abundant than deciduous oaks (Fletcher et al., 2007). Overall, south-western sequences still document high values of pines but there is a replacement of P. sylvestris by P. pinaster (Queiroz, 1999) although the former continues to exist in the area during the Mid-Holocene (Soares and Silva, 2018; Monteiro, 2018). In central and northern Portugal, deciduous Quercus thrive. In the NW they became largely dominant as the Holocene deciduous oak forests were rapidly established (Ramil Rego et al., 1998). The sequences of Serra da Estrela suggest a rapid expansion of oaks at low altitudes and an expansion towards higher altitudes, taking several centuries to reach the sampling sites (van der Knaap and van Leeuwen, 1997). Here, the expansion of heathlands was slow while at Chã das Lameiras there is an expansion of Fabaceae and a decrease in P. sylvestris (López-Sáez et al., 2017). Even in northeast Portugal, nowadays with marked Mediterranean climate, archaeopalynological data from Prazo dating from the mid-10th millennium cal BP onwards suggest a dominance of deciduous Quercus, with high values of Olea and Pistacia, among other thermophilous shrubby taxa. Charcoal data attested the local presence of both deciduous and evergreen Quercus, as well as P. pinea/pinaster, A. unedo, among others (Monteiro-Rodrigues et al., 2006). At Foz do Medal, located slightly to the north, charcoal from secondary deposits dating from the mid-9th millennium cal BP document the presence of evergreen Quercus, P. pinaster and Fraxinus, among others (unpublished data). All this data not only suggest a rapid expansion of Holocene forests but also that they differed in composition, depending on regional conditions and previous ecological history. 3.4. Fauna The animal remains referred to herein were recovered from levels corresponding to the Magdalenian, Azilian and Early Mesolithic (ca. 20,000–8200 cal BP) at fourteen sites distributed in four mainland areas of Portugal: Fariseu (Côa Valley); Vale dos Covões, Gruta do Caldeirão, Lapa do Picareiro, Lapa dos Coelhos, Lapa do Suão, Abrigo das Bocas, Costa do Pereiro, Toledo, Vale Frade, Pinhal da Fonte (Estremadura region); Barca do Xerez de Baixo (Guadiana Valley); Vale Boi and Barranco das Quebradas (Algarve). A common trait of the faunal assemblages is the prevalence of mammal remains. These represent between 90 and 100% of the animal remains found at most sites - excepting the Early Mesolithic sites of Toledo (close to 86%), Vale de Frade (less than 77%) and Barranco da Quebradas where the assemblages are exclusively shellfish (Table 3). The ungulates found at Portuguese archaeological sites dated to the last 30,000 years include red deer, roe deer, wild boar, aurochs, equids, ibex and chamois. The latter two species seem more important during the Solutrean and are related to rockier landscapes, higher altitudes and colder environments (Bicho and Haws, 2012). Deer (Cervus elaphus) and wild boar (Sus scrofa) are abundant at Lapa do Picareiro, Vale Boi and Toledo (Supplementary material – Table 2). Aurochs is the most represented species at Barca do Xerez de Baixo, an Early Mesolithic butchering site. Red deer, horse and rabbit were also found at this inland site (Valente, 2013; Araújo, 2016a, 2016b). As shown in Table 3, Magdalenian faunal assemblages document the consumption fish, birds, rabbits and molluscs (Chauviére, 2002; Davis, 2002: 42; Callapez, 2003; Bicho et al., 2003; Davis et al., 2007; Gabriel and Béarez, 2009: 332; Gabriel, 2011; Roselló and Morales, 2010; Moreno-Garcia, 2011; Regala, 2011; Bicho and Haws, 2012; Manne et al., 2012; Valente, 2013; Araújo, 2016a, 2016b; Carvalho et al., 2016; Gameiro et al., 2017). Most of these were also recorded in Solutrean levels at Caldeirão and Lapa dos Coelhos, as well as in Early Mesolithic sites of Portuguese Estremadura (Dupont and Araújo, 2010; Gabriel, 2011; Araújo et al., 2014; Carvalho et al., 2016). Mammal remains with butchery cut marks and other impact traces likely suggest that bone marrow extraction may have occurred, and bones may have been used to produce tools (Gabriel and Béarez, 2009). Rabbit (Oryctolagus cuniculus) is the most abundant taxon at Fariseu, Caldeirão, Lapa do Picareiro, Lapa dos Coelhos and Lapa do Suão (Supplementary material - Table 2). It is always problematic to establish whether rabbits were brought by humans or result from natural accumulation by predators (Davis, 2019; Hockett, 1999; Hockett and Bicho, 2000; Hockett and Haws, 2002; LLoveras et al., 2011). Nevertheless, taphonomic studies have demonstrated that, at least since the Solutrean, rabbits were hunted and part of human diet (LLoveras et al., 2011). Fish taxa document the exploitation of the freshwater environments close to Fariseu, Lapa dos Coelhos and Lapa do Picareiro (Bicho et al., 2003; Gabriel and Béarez, 2009; Roselló and Morales, 2010). Fishes included in the marine division, found both in marine-estuarine milieus, are present at Gruta do Caldeirão, Toledo, Vale de Frade and Vale Boi (Table 3) (Zilhão, 1997b; Gabriel, 2011). In the Caldeirão cave, the four teeth of Sparus aurata identified should correspond to adornments (cf. Zilhão, 1997a: 154) probably exchanged or brought from the littoral. The presence of clupeids at Lapa do Picareiro (Bicho et al., 2003) cannot be used to suggest incursions to the coast for fishing, as some of the species in the Clupeidae family include anadromous fish, such as the allis shad and twait shad, which migrate into fresh water to spawn. Besides shellfish that could be procured for food in marine-estuarine environments (Vale Boi, Gruta do Caldeirão, Vale de Frade, Toledo, Pinhal da Fonte and Barranco das Quebradas) (Table 3/Supplementary material - Table 2) (Callapez, 2003; Dean et al., 2011; Manne et al., Table 2 (continued) Site Layer Lab Ref. Sample Date BP Date cal BP (2σ) * BXB Structure A OxA-13264 Charcoal - E. arborea 8250 ± 37 9402–9091 Vale Sá Structure Gif-10348 Charcoal 8500 ± 110 9765–9143 Costa do Pereiro 2 WK-35997 Bone - Cervus elaphus 8680 ± 80 9693–9548 Costa do Pereiro 1b WK-30215 Bone – Capra pyrenaica 8564 ± 37 9560–9482 Cova da Baleia Structure CBL-N7-5-L2, Beta-295903 Charcoal 8460 ± 50 9540–9333 Prazo Hearth C6 GrA-15986 Charcoal Quercus sp. 8370 ± 70 9526–9142 Prazo Hearth C5 GrN-26402 Charcoal 8380 ± 60 9525–9258 Prazo Hearth C5 CSIC-1621 Charcoal 8397 ± 38 9496–9305 Cova da Baleia Structure V10-52-L2 Beta-295905 Charcoal 8320 ± 40 9464–9141 Cova da Baleia Structure CBL-R10-24-L3 Beta-295904 Charcoal 8250 ± 50 9409–9033 Cova da Baleia Structure CBL-L7-2-L3 Beta-294170 Charcoal 8230 ± 50 9400–9029 São Julião A Q1,B1 ICEN-179 Charcoal 8120 ± 100 9401–8663 C. Gameiro, et al. Quaternary International xxx (xxxx) xxx–xxx 7 Table 3 Animal remains found in levels corresponding to the Magdalenian and Early Mesolithic (20,000–8200 cal BP). For sake of comparison we use the number of remains identified per taxa (NISP) found in each site/layer. "P" indicates that a given taxa is present, but no inventory is published in detail. Shellfish are also listed in terms of presence/absence, as the existing data is mostly presented counting the minimum number of individuals found (MNI) or weight. (*) Indicates the presence of fluvial organisms. Site label: VB= Vale Boi; CLD= Caldeirão; S7-9 and S4-6 = Lapa do Suão (Layers 7–9; 4–6); PIC-F/G and PIC-E/L = Lapa do Picareiro (Layers F/G; E/L; E/U); LC-4 and LC-3 = Lapa dos Coelhos (Layer 4 and layer 3); VCA1-3/4 = Vale dos Covões 1 (Layers 3 and 4); BCS= Bocas); FAR-4 = Fariseu (Layer 4); CP= Costa do Pereiro; TOL = Toledo; VFR= Vale de Frade; PF= Pinhal da Fonte; BXB= Barca do Xerez de Baixo; BQ= Barranco das Quebradas 1 and 3 (Layer 3 and 5/10). Data collected from: Chauviére (2002); Bicho et al. (2003); Callapez (2003); Davis (2002): 42; Davis et al. (2007); Gabriel and Béarez (2009): 332; Roselló and Morales (2010); Zilhão et al., 2010: 7; Gabriel (2011); Moreno-Garcia (2011); Regala (2011); Bicho and Haws (2012); Manne et al. (2012); Valente (2008), 2010 & 2013; Carvalho et al. (2016); Araújo (2016a,b); Gameiro (2017). Data for VCA 1 is still unpublished (S. Gabriel unpublished data). Taxonomic list for terrestrial taxa adopts Bencatel et al. (2019) taxonomic reference list. Early/Midlle Magdalenian Upper Magdalenian Final Magdalenian Azilian Early Mesolithic VB CLD S-7/9 PIC-F/G PIC-E/L LC-4 PIC-E/U S-4/6 LC-3 VCA1-3/4 BCS FAR-4 CP TOL VFR PF BXB BQ1-3 BQ3-5/10 MAMMALS Insectivora (Eilipotyphla) - Insectivores _ _ _ _ _ _ _ _ _ _ _ _ _ 25 1 _ _ _ _ Carnivora - Carnivores 9 66 18 _ _ _ _ 1 1 _ _ _ _ 12 5 _ 23 _ _ Artiodactyla - Even-toed ungulates 135 86 43 276 28 9 34 8 8 P P 14 40 190 7 _ 470 _ _ Perssodactyla - Odd-toed ungulates 51 6 1 _ _ _ _ _ _ _ P 1 9 17 _ 10 _ _ Rodentia - Rodents _ 1 _ _ _ _ _ _ _ _ _ 3 _ _ _ _ _ _ Lagomorpha 163 5265 12191 9638 594 132 717 377 218 P P 79 21 760 110 _ 205 _ _ Medium/large mammals indet. _ _ _ _ _ _ _ _ _ _ _ 638 _ 668 72 _ _ _ _ Small mammals indet. _ _ _ _ _ _ _ _ _ _ _ 182 _ _ _ _ _ _ _ Micromammals indet. _ P _ _ _ _ _ _ _ _ _ 6 _ _ _ _ _ _ _ Total Mammals 358 5424 12253 9914 622 141 751 386 227 _ _ 922 62 1664 212 _ 708 _ _ % Total NISP 98,62 99,27 100,00 100,00 100,00 90,38 100,00 99,74 98,70 _ _ 98,40 100,00 86,04 76,81 _ 99,58 _ _ BIRDS _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Birds _ 28 P P P _ P P _ _ _ 1 _ 67 5 _ _ _ _ Total Birds _ 28 _ _ _ _ _ _ _ 1 67 5 _ _ _ _ % Total NISP _ 0,51 _ _ _ _ _ _ _ 0,11 3,46 1,81 _ _ _ _ REPTILES _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Reptiles _ _ _ _ _ _ _ _ _ _ _ _ _ 58 _ _ _ _ _ Total Reptiles 28 _ _ _ _ _ _ _ 1 67 5 _ _ _ _ % Total NISP 0,51 _ _ _ _ _ _ _ 0,11 3,46 1,81 _ _ _ _ AMPHIBIANS _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Amphibians _ _ _ _ _ _ _ _ _ _ _ _ _ 2 22 _ _ _ _ Total Amphibians _ _ _ _ _ _ _ _ _ _ _ _ 2 22 _ _ _ _ % Total NISP _ _ _ _ _ _ _ _ _ _ _ _ 0,10 7,97 _ _ _ _ FISH _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Freshwater/Anadromous/Catadromous _ _ _ P P 15 P _ P _ _ 13 _ _ _ _ _ _ _ Marine-Estuaine _ 4 _ _ _ _ _ P _ _ _ 143 37 _ _ _ _ Total Fish _ _ _ _ _ 15 _ _ _ _ _ 13 143 37 _ _ _ _ % Total NISP _ _ _ _ _ 9,62 _ _ _ _ _ 1,39 7,39 13,41 _ _ _ _ INVERTEBRATES _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Terrestrial _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 88 _ Freshwhater _ 1 _ _ _ _ _ _ _ _ _ 1 _ _ _ _ _ 1 _ Marine-Estuaine 5 7 _ _ _ _ _ _ _ _ _ _ _ _ P P 3 8975 16982 Shellfish P P P P P P P P* P _ _ _ _ _ _ _ _ (continued on next page) C. Gameiro, et al. Quaternary International xxx (xxxx) xxx–xxx 8 2012; Valente, 2008, 2010; 2013; Araújo, 2016a); molluscs were also used for shell beads, as the presence perforated Littorina obtusata indicates (Callapez, 2003; Chauviere, 2002; d’ Errico and Vanhaeren, 2002; Regala, 2011). 4. Lithic industries The widespread of technological approaches as well as the identification of sites outside the classic Estremadura region, during the last twenty-five years, is reason enough to address the issue focusing on six main geographic units (shown in Fig. 1). Apart from Estremadura and Algarve, situated in areas of limestone substrate, the other regions are situated in the Iberian massif, where only quartz and siliceous metamorphic rocks are available. In Table 4 a summary of the main characterizing traits of the lithic production is displayed. 4.1. Estremadura Estremadura is the best-known region, because its limestone substrate, and hence the existence of caves, shelters and local availability of flint motivated its survey since the 19th century. However, the sedimentary gaps mentioned above do not allow neither a detailed characterization of the Solutrean-to-Early Magdalenian transition nor the definition of a Middle Magdalenian phase. These hiatuses seem to correlate with climate changes (Zilhão, 1997a, 1997b; Aubry et al., 2008a, 2011; Gameiro, 2012) and do not imply the changes in population dynamics and demographic density that some have argued for (Bicho and Haws, 2012). In this region, flint is always the most used raw material, due to the existence of sources located at < 10 km from the sites. Other locally available rocks, such as quartzite, which is often used for expedient flake debitage, were also used; its percentage can reach up to 20% (Lapa dos Coelhos layer 3). Good quality quartz was used for bladelet debitage in Lapa dos Coelhos layer 4, where it represents up to 36% of the assemblage (Gameiro, 2012; Gameiro et al., 2013). Only four sites contain occupations that have been ascribed to the Early Magdalenian: CPM, Caldeirão, Cerrado Novo and Vascas, and only the first two have radiometric dates (Table 2). Bladelet tools are not very diversified; pointed bladelets are absent and marginal retouch is used to regularize the edges of blanks with a rectangular tendency (> 40%) (Fig. 3). Cores-on-flake are used, in proportions that remain under 45%, to produce bladelet blanks Flake and bladelet-oriented debitage account for 80% and 20%, respectively, of the lithic tool-kits. Inverse retouch is common on retouched flakes and scrapers (Fig. 4). The Middle Magdalenian has not been characterized yet, since there are only two sites with remains ascribable to this phase: Buraca Grande, with a directly dated semi-circular baguette (Aubry and Moura, 1993) (Table 2), that is hard to associate with other elements from its level of provenience; and, based on its dating (Table 2), the open-air site of Olival Fechado, were a production of backed points or microgravettes coexists with marginally backed bladelets obtained from cores-on-flakes (burin-like) (Silva, 2003). The Upper Magdalenian is documented at eight sites: CPM (loci I and IIIs), Lapa dos Coelhos layer 4, Galeria da Cisterna, Gruta do Caldeirão, Picareiro, Vale da Mata, Vascas and Lapa do Suão. In Lapa dos Coelhos layer 4, a high percentage of microgravettes was identified; they coexist with rectangular backed bladelets (Fig. 5.1). The blanks were obtained through the exploration of cores-on-flakes (burin-like) (Fig. 5.5). There is an independent laminar production and a production of small flakes through bipolar debitage on anvil during the final reduction stages (Gameiro, 2012; Gameiro et al., 2013). The Rossio do Cabo Final Magdalenian facies, identified at three sites (Pinhal da Carneira, Rossio do Cabo and Lapa dos Coelhos), can be distinguished from the previous phase by the absence of autonomous laminar production and the increased use of cores-on-flake for bladelet debitage (Fig. 6). At Lapa dos Coelhos, from the Upper Magdalenian Table 3 (continued) Early/Midlle Magdalenian Upper Magdalenian Final Magdalenian Azilian Early Mesolithic VB CLD S-7/9 PIC-F/G PIC-E/L LC-4 PIC-E/U S-4/6 LC-3 VCA1-3/4 BCS FAR-4 CP TOL VFR PF BXB BQ1-3 BQ3-5/10 Total Invertebrates 5 8 _ _ _ _ _ _ _ _ _ 1 _ _ _ _ 3 9064 16982 % Total NISP 1,38 0,15 _ _ _ _ _ _ _ _ _ 0,11 _ _ _ _ 0,42 100,00 99,47 TOTAL NISP 363 5464 12253 9914 622 156 751 387 230 _ _ 937 62 1934 276 _ 711 9064 17072 C. Gameiro, et al. Quaternary International xxx (xxxx) xxx–xxx 9 yielded a large lithic assemblage; no faunal remains were preserved and study of these sites is still ongoing (Gameiro et al., 2018). At Rôdo, a colluvial deposit (SU006) accumulated on top of the river terrace contained a lithic assemblage and some anthropogenic stone structures that may correspond to combustion features. The top of the sequence (SU003) was disturbed by post-depositional phenomena (ravine channels, recent reforestation and road construction). Excluding some blade fragments and microlithic types of probable Neolithic age, the lithic industry shows uniform techno-typological characteristics. Debitage is oriented towards the production of flakes and bladelets. Cores-on-flake (Fig. 10: 6 and 7) were exploited along their edges (mostly burin but also endscraper types) to produce bladelets; bipolar debitage on anvil is also frequent (Fig. 10: 3,4 and 5). Retouched tools are rare but flint unguiform endscrapers, marginally backed bladelets (Fig. 10: 2) and curved backed bladelets (Azilian points: Fig. 10:1) were identified. Quartz, quartzite and other coarse-grained cobbles were used for expedient flake debitage or transformed into denticulates, notches or scrapers and macro-tools. Regarding the chrono-cultural affiliation of these evidences, we would point out that the stratigraphic resolution of this open-air site doesn't support the construction of a solid sequence correlating material culture and radiocarbon dates (Table 2). Furthermore, we would also stress that: i) radiocarbon data suggest an occupation of the site that might have begun during the Middle Magdalenian and lasted until the Early Mesolithic; ii) the lithic assemblage does show Final Magdalenian techno-typological characteristics, which also support a possible Azilian occupation of the site. At Vau archaeological site two clearly differentiated chronological horizons were identified: Holocene layers (SU001; 002 and 003) containing a few typically Upper Palaeolithic artefacts, as a result of postdepositional disturbance (ravine channels); and Pleistocene layers (SU005 and 013), which contain lithic artefacts only. A few marginally retouched bladelets, microliths and one Teyjat or Ahrensburg point (Fig. 11) were recovered atop the Gravettian occupation. Prismatic and “flake-edge cores” were used in bladelet debitage schemes and bipolar knapping on anvil is frequent. Reffiting indicates unipolar or centripetal expedient debitage in flake reduction sequences. This lithic assemblage suggests a short Tardiglacial occupation, which can be ascribed to Final Magdalenian despite the existence of a Mesolithic date (Table 2), which is not consistent with the lithic industry. Moreover, the top of the stratigraphic sequence shows evidence of erosion and was disturbed by recent farming-related activities, which hinder a better definition of the site's early Holocene occupation sequence. Four negative structures were identified at Bispeira 8; some may have been used as fireplaces. Most of the artefact assemblage was recovered in SU002, which corresponds to the sedimentary unit covering all the negative structures and is a palimpsest containing artefacts of different ages. Structures 1 and 2 provided Late Paleolithic radiocarbon dates but reduced technological information: two quartzite flake cores in SU004, and one quartzite flake core, one quartz core, two quartz flakes and one quartzite flake in SU005. Structures 3 (SU020/027) and 4 (SU021/024/025) yielded Neolithic radiocarbon dates, one ceramic fragment and 14 lithic artefacts including one flint blade fragment, five flint bladelets, two quartz flakes and another flake on a coarse-grained rock. At all three sites, most of the stone tools were produced using local raw materials such as quartz and other coarse-grained rocks gathered from the nearby alluvial gravels. By comparison with Vau's Gravettian layer, which features a high percentage of flint and silcrete, the more recent occupations of Vau, Rôdo and Bispeira 8 bear witness to a slight decrease in flint and silcrete use between the Early Upper Palaeolithic and the Tardiglacial, similar to the pattern previously identified at Cardina (Côa Valley) (Aubry et al., 2016). The percentage of retouched tools, including index fossils such as armatures, is very low, which makes chrono-cultural attribution difficult. Based on lithic technology, the Rôdo human occupation can be ascribed to Final Magdalenian and Azilian; nevertheless, undergoing work on spatial analysis will test the existence of a Late Middle Magdalenian or Upper Magdalenian occupation as suggested by the 13,050 ± 40 BP (15,832–15,383 cal BP) 14 C dating. The Vau and Bispeira 8 sites seem to have been occupied during Final Magdalenian and Neolithic. 4.5. Guadiana Valley Between 1998 and 2002, Pleistocene occupations were identified in the scope of the mitigation works required by the construction of a dam in the Guadiana River basin (Southern Portugal) (Almeida, 2013). Prior to this work, clusters of macrolithic material from this southeast quadrant of the country, mostly made on quartzite, labelled Languedocian by H. Breuil were all that could be considered of Pleistocene age (Breuil, 1917). Since then, the term Languedocian has been used as a short-hand designation devoid of chronological significance (Raposo and Silva, 1984) applied to an expedient technological scheme, dependent on the locally available raw material and probably motivated by functional imperatives (Almeida et al., 2013; Araújo and Almeida, 2013). In this river basin, dry spells alternate with torrential flooding, which accounts for the erosion of low-elevation sedimentary formations and for the preservation of sites only on high alluvial or colluvial platforms. The intense agricultural exploitation of some platforms with a potential for sedimentary preservation originated the formation of palimpsests that are difficult to disentangle. At the same time, the absence of macro-organic remains has hampered radiometric dating and made it impossible to obtain data on diets and paleoenvironments. The study of lithic materials and their comparison with the cultural sequence identified in Estremadura (Zilhão, 1997a) allowed the identification of at least two Magdalenian sites: Monte da Ribeira 9 (Mourão) and Malhada do Mercador I-SW (Mourão) (Gameiro, 2012; Almeida et al., 2013). The human occupations of Chancudos 3 and Monte Roncanito 21 may date to the end of the Tardiglacial or the beginning of the Holocene, but the rough appearance of the lithic industry does not allow certainty. The absence of flint in this region underpins the dominance of quartz and quartzite, which account for more than 90% Fig. 11. Vau – possible Teyjat ou Ahrensburg point (?) on regional Bajocian flint. Photo by C. Manzano. C. Gameiro, et al. Quaternary International xxx (xxxx) xxx–xxx 16 of the raw materials, resulting in uncharacteristic industries and hindering inter-regional comparison. The existence of quartz veins in the vicinity of Monte da Ribeira explains the predominance of this raw material: 73% of the total. Flint only reaches 1.57%, but lydite, a local fine-grained rock, is more abundant, at 4.44%. Laminar production is absent and most blanks are flakes produced expediently, with no investment in the conformation of mostly unipolar volumes. Most tools are notches, denticulates and retouched flakes. The endscraper group is not very significant (14%). Bladelet tools account for 19% of the artefacts and are mostly composed of fragments of flint backed bladelets; bladelet production is rather careful. It was possible to identify two exploitation modalities: convergent exploitation originating pyramidal cores and parallel exploitation originating cores with rectangular debitage surfaces. Although some cores provide evidence for bladelet production on quartz, no retouched bladelets on this raw material were identified. The existence of microgravettes (Fig. 12:25) and backed bladelets of rectangular tendency was considered an archaism supporting the attribution of this site to an Early or Upper Magdalenian phase (Gameiro, 2012; Gameiro and Almeida, in press). The site of Malhada do Mercador I-SW, is interpreted as a single occupation site, despite the identification of different SU's. The stratigraphic differences result from recent farming. Quartz and quartzite materials account for 90% of the total, distributed equally among these raw materials, at 45% and 46% respectively. Around 33% of the tool-kit is composed of common tools (notches, denticulates and scrapers) with endscrapers reaching ca. 18% of the tools. Quartz and quartzite flakes are the most abundant type of blanks and are obtained by expedient, unipolar methods, taking advantage of the morphology of the volumes: flattish cobbles exploited from the thicker side, producing short flakes always partially cortical; the more globular cobbles are exploited from their natural dihedrons, resulting in larger flakes. The bladelet group accounts for 20% of the total; these blanks are made on flint, quartz and other indeterminate but fine-grained raw materials. Cores-on-flake are rare and the bipolar debitage on anvil, especially on quartz, is frequent. There is no deliberate blade production. The occurrence of trapezoids, Areeiro bladelets, marginally backed bladelets and a Malaurie point resulted in an ascription to the Azilian (Gameiro and Almeida, in press). The sites of Chancudos 3 and Monte Roncanito 21 share some typological characteristics with sites ascribed to the Magdalenian (the presence of endscrapers, for example). At Monte do Roncanito 21, quartzite debitage is closer to the debitage schemes identified at Barca do Xerez, a site located upstream, on a platform overlooking the Guadiana River, and ascribed to the Early Mesolithic (with radiocarbon dates in the 9700-9200 cal BP interval; Araújo, 2016b). The Early Mesolithic site of Barca do Xerez is interpreted as an animal processing site and knapping activities are simple, expedient and clearly adapted to the morphology of local quartzite cobbles in order to produce flakes (Araújo and Almeida, 2013; Araújo, 2016b). The absence of absolute datings does not allow for solid conclusions. However, two hypotheses can be put forward: we are dealing with contemporaneous contexts and the observed differences have a Fig. 12. Monte da Ribeira – Upper Magdalenian? 1 - truncated backed point, 2–24 – backed bladelets fragments and complete exemplars; 25 – microgravette; 26 – trapeze; 27 - pièce esquillée; 28 and 29 – denticulated tools. Flint and other indeterminate local raw materials. Drawings by K. Monigal C. Gameiro, et al. Quaternary International xxx (xxxx) xxx–xxx 17 functional explanation or these differences may have chronological significance and Monte Roncanito 21 may document a phase predating the Early Mesolithic, probably already Holocene, but with a persistence of Palaeolithic technological elements (Almeida et al., 2013). 4.6. Algarve So far, the available data for the Pleistocene-Holocene transition period in Southern Portugal are limited and mostly concern the southwestern region of the Costa Vicentina (Fig. 1). As a result of various research projects carried out since the 1990s, Tardiglacial occupations have been recorded at the rock-shelter of Vale Boi and the open-air sites of Praia da Galé, Lagoa do Bordoal, Ponta Garcia and Vale Santo 4, (Mendonça, 2009; Bicho et al., 2009, 2010). The open-air sites of Barranco das Quebradas, Rocha das Gaivotas and Castelejo are of Holocene age (Bicho, 2003; Stiner et al., 2003; Carvalho and Valente, 2005; Valente and Carvalho, 2009; Valente, 2008, 2010). Except for Vale Boi, where the Magdalenian occupation overlies a Solutrean level, and Lagoa do Bordoal, with an OSL dating of 14,800 years ago, the remaining sites are unstratified (eroded, reworked or surface scatters; Mendonça, 2009) and the lithic assemblages recovered are small and typo-technologically uncharacteristic. Therefore, Tardiglacial age is possible but remains unconfirmed. Thus, a preliminary summary of the human occupation in the region between 14,000 and 8000 cal BP is feasible (Dean et al., 2011) but must bear this caveat in mind. The trend towards microlitization starts with the Solutrean. The Tardiglacial industries are characterised by the exclusive use of local raw materials (quartz, graywacke, quartzite and flint) and are aimed at the expedient production of flakes and chips, the latter produced by bipolar debitage on anvil. Elongated blanks are rare and except for denticulated and notched bladelets there are no armatures (Mendonça, 2009). Two phases were identified in this region: Early/Middle Magdalenian and Final Magdalenian. A functional classification of the sites was proposed: Vale Boi would correspond to a base camp, Ponta Garcia and Vale Santo 4, where only surficial materials were identified, would correspond to procurement and knapping sites, and Lagoa do Bordoal and Praia da Galé would be temporary camps for the seasonal exploitation of hunting and aquatic resources (Bicho et al., 2010; Mendonça, 2009). Barranco das Quebradas is located on a small coastal ravine perpendicular to a seashore characterized by limestone outcrops; five archaeological loci were discovered here. Locus 1 and Locus 3 belong to the earliest Holocene. The small, 40 to 120 cm-thick shell-midden deposits feature a limited number of uncharacteristic lithics; the mollusc taxa represented are Monodonta lineata, Patella, Mytillus, Pollicipes and Thais haemastoma species (Bicho et al., 2009, Stiner et al., 2003; Valente, 2008, 2010). The Rocha das Gaivotas site is a shell-midden context featuring several occupations dating from the Early Mesolithic to the Early Neolithic. Interpreted as a raw material procurement site, located about 100 m from a flint source, Early Mesolithic layer 3 yielded two combustion structures and a very small number of artefacts. The predominant mollusc taxa are the same as at Barranco das Quebradas (Carvalho and Valente, 2005; Valente, 2010). Castelejo, located between Barranco das Quebradas and Rocha das Gaivotas, has a human occupation dating to ca. 10,500 cal BP and is characterized by a very small number of lithics; besides molluscs, the faunal remains include fish and rabbit remains (Soares and Silva, 2004; Valente and Carvalho, 2009). 5. Art Based on the five most represented themes in the Palaeolithic rock art of the Douro basin and the territory to the south of the Tagus River, a study applying multiple correspondence analysis and ascending hierarchical classification revealed the existence of at least four classes of horse and aurochs, as well as three classes of ibex, red deer stag and Fig. 13. - Fariseu rock 4 (drawing by Fernando Barbosa). Panel with phase 2 and 3 figures. Note the phase 2 horse (in red), reapropriated and partially superposed by a phase 3 male aurochs (in orange). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.) C. Gameiro, et al. Quaternary International xxx (xxxx) xxx–xxx 18 red deer doe depictions (Santos, 2017:136–164). The study of the parietal stratigraphy of rock art panels with more than one class and the sedimentary context of several sites in the Côa Valley allowed us to infer that these classes correspond to at least four phases of artistic production (Santos, 2017:129–255). In the scope of this article, we shall focus on phases 3 and 4. Class 3 horses and aurochsen can be ascribed to Phase 3, as well as part of the ibexes, red deer stags and does of Class 2. These animal depictions feature naturalistic forms and well-proportioned bodies; the heads frequently show anatomical details (e.g., the mouth, ears, outlined nostrils, almond-shaped eyes). All four legs are usually represented, often in uni-angular profile. Hooves are also commonly represented, as are the chromatic variations of the coat and integument on heads and limbs. Incision is the most common technique, in its simple and repeated forms, although pecking and abrasion have also been identified (Santos, 2017:138–139, 144, 150, 155, 159, 163). Phase 3 figures are usually found in panels where motifs from Phase 2 also occur. Nevertheless, it is sometimes difficult to accurately distinguish the figures of each phase, especially when dealing with red deer stags and does or ibexes. Actually, class 2 of these themes (red deer stags/does or ibexes) may belong to the same phases as the aurochsen (phase2) and horses (phase 3) of classes 2 and 3. The cultural attribution of Phase 3 was based on stylistic comparison (Fig. 13). The most evident parallels are found in the FrancoCantabrian region, in contexts ascribed to the Middle and Final Magdalenian (Santos, 2017: 233–245), generally corresponding to figures catalogued as “style IV” by Leroi-Gourhan (Leroi-Gourhan, 1995 [1965], 283–288). These representations are also very similar to an important group of figures, found in some portable rock art objects from the Magdalenian layer of Foz do Medal site (eg. Figueiredo et al., 2015a,b; Figueiredo et al., 2016), located in the Sabor Valley, approximately 27 km northeast of the Côa Valley. A terminus post quem is provided by the dating of Phase 2 to a period between the Upper Solutrean and the Magdalenian, based on style and geoarchaeological context (Santos, 2017:178–180). Due to the low number of motifs from Phases 2 and 3, if considered separately, as well as the high frequency of rocks in which these phases occur together, their study was performed jointly (Santos, 2017:381–389). In the Côa Valley, the thematic distribution of this group is as follows: horse – 24.9%; aurochs – 19.8%; red deer stag – 14.5%; ibex – 13.6%; humans – 5.1%; red deer doe - 4%; chamois - 1%; feline – 0.6%. There are also isolated representations of fish, birds and perhaps bison. Indeterminate and incomplete figures reach 15.1%. If one adds to these figures the pre-existing depictions found in the same panels (Santos, 2017, vol. II, table 6.14), aurochs is the most represented motif (23.2%), followed by horse (22.3%), ibex (17.9%), red deer stag (11.4%), red deer doe (3.6%), humans (2.7%) and chamois (0.2%). The indeterminate and incomplete figures reach 16.1%. In relation to the preceding phases, the non-figurative repertoire increases and graphic units are more diversified (Santos, 2017, vol II, 485). At Foz do Medal, the thematic distribution is somewhat different, mostly corresponding to caprine representations, followed by horses and aurochsen, red deer being very rare; humans are represented by a single figure (Figueiredo et al., 2015a:1576). Differences in relation to the Côa Valley parietal art may be due to several factors: only 92 figurative motifs occur in layer 1055 of Medal (as opposed to 470 in phases 2 and 3 of Côa), in which two different graphic facies may well be subsumed (Figueiredo et al., 2015a:1577). The figures ascribed to Phase 3 denote important relations between Northern Portugal and the Cantabrian region. But, above all, they indicate a rupture of contacts with the South. In fact, representations broadly attributed to the Magdalenian in Southern Portugal (such as Fratel and Gardete in the northern bank of the Tagus River (Gomes, 2010:476), Porto Portel in the Guadiana basin (Baptista and Santos, 2013:220–226) and possibly some of the Escoural cave representations (e.g. Gomes, 2002) feature similar characteristics to depictions from Southern Iberia, where the sequence of Parpalló is an important reference (Villaverde Bonilla, 1994; Bicho et al., 2007). Phase 4 animal depictions of the Douro basin are characterized by their geometrical bodies, often filled-in. The heads are represented without any internal detail. The four legs of the animals are usually represented in oblique or straight biangular perspective. Incision is the most common technique, both in the figures’ outlines and in the fillingin of the bodies. Nevertheless, pecking, scraping and red painting also occur (Aubry et al., 2017:407–409). This type of rock art is well dated due to an important collection of portable art composed of 89 objects, Fig. 14. Canada do Inferno rock 41 (drawing by Fernando Barbosa). Motifs 1 to 8 probably date from phase 3, while motifs 9 to 12 probably date from phase 4. Besides the differences in style, phase 4 motifs are situated on a marginal area of the panel, as compared to the other motifs, which confirms that phase 4 motifs were executed after the older motifs. C. Gameiro, et al. Quaternary International xxx (xxxx) xxx–xxx 19 Fig. 15. Vale de José Esteves rock 13 (detail of a drawing by Fernando Barbosa). Note the prominence of the cervids, a characteristic of phase 4, to which this panel should be ascribed. C. Gameiro, et al. Quaternary International xxx (xxxx) xxx–xxx 20 featuring the kind of representations we have just described. It was unearthed in Fariseu layer 4 and dated by radiocarbon, TL and OSL to between 12,000 and 10,000 years ago. In cultural terms, it is attributed to the Late Azilian (Santos et al., 2018). Phase 4, which is very well documented in the Spanish northern sub-meseta (Bueno Ramirez et al., 2007), is also found in other sites of Northern Portugal, such as Pedra de Asma 7 in the Sabor Valley (Figueiredo, 2014:20–22). Given the morphological similarities between the incised figures of Côa Valley's Phase 4 and other pecked motifs from the north of Portugal, traditionally ascribed to the Holocene period, we consider that these animals all belong to the same period (Santos et al., 2018:54-55). Among them would be Cabeço do Aguilhão (Figueiredo et al., 2015b), Passadeiro (Sanches and Teixeira, 2014) and Parada (Teixeira, 2016). Some of these sites also feature deep incisions, locally called unhadas do diabo (lit.: the devil's scratches), which, according to some authors, can also be dated to this phase (Sanches and Teixeira, 2014). As opposed to what happened during Phase 3, southern Portugal features a series of Phase 4 representations very similar to those we have been discussing. This is the case of some pecked figures of the Tagus Valley, namely those belonging to the old sub-phase of the naturalistic period, as defined by Varela Gomes (Gomes, 2007: 87, 91–102), as well as some incised figures of the Guadiana Valley, namely at Moinhola (Baptista and Santos, 2013:147–149) and Molino Manzánez (Collado Giraldo, 2006:224–225; Santos et al., 2018: 57). The "striated" figures of Escoural cave should also be attributed to this phase (Collado Giraldo, 2006:379; Santos et al., 2018:60). The most represented animals during Phase 4 are red deer and ibex, in the Tagus Valley (Gomes, 2007:113) as much as in the Côa Valley (Santos et al., 2018:55–56). The latter is particularly relevant, because comparison with previous phases is possible. While the horse/aurochs pair, followed by ibex, dominates in the Magdalenian, the Azilian of Phase 4 is largely dominated by red deer (Fig. 15), followed by ibex. Style change follows the same pattern. While Phase 3 has parallels in the Franco-Cantabrian zone, the Azilian graphic conventions are widespread across all of Southwestern Europe, from Portugal to Italy and from southern Spain to northern France (Santos et al., 2018:56–64). This graphical tradition seems to develop in Western Europe between 14,000 and 10,500 cal BP (Fig. 14) (Santos et al., 2018:68). Phase 4 may also have included schematic, not filled-in shapes (Santos et al., 2018:66). They have been identified, for example, in the Murat shelter of Southwest France (e.g. Lorblanchet, 1985, 1989, 1996), Peña de Estebanvella in the Spanish northern meseta (e.g. García-Díez, 2013) and eventually in the Côa Valley (Santos et al., 2018:66). Regarding the Medal collection, a recently initiated project intends to study the representations of layer 1055. Until the integral study of the collection is carried out, the characteristics of the most naturalistic representations against the more schematic ones, their respective numerical value and whether they correspond or not to the same moment are not clear, opening the road to multiple hypothesis, among them, the possibility that the transition between the most naturalistic figures and the more schematic ones could have occurred during the interstadial period. 6. Discussion The existence of stratigraphic discontinuities between 20,000 and 18,000 cal BP and between 14,000 and 12,500 cal BP (Zilhão, 1997a) makes it more difficult to understand the diachronic variability of the lithic tool-kits and the paleoenvironment of the human groups who occupied the present-day Portuguese territory during the Tardiglacial. Although the number of known sites has increased significantly over the last 25 years, sites dated to the Early Magdalenian are still poorly known, the Franco-Cantabrian Middle Magdalenian (Langlais, 2007) remains uncharacterised and most sites are dated to the timespan of the Bølling/Allerød and Dryas III (Gameiro, 2012; Gameiro et al., 2013). The pollen sequences of continental Portugal (Supplementary material - Table 1) document vegetation responses to the major climatic fluctuations of the Late Glacial and Early Holocene. Besides the expansion and retraction of woodlands, changes in their composition are also indicated by the available data, which are rather unevenly distributed across the territory. In most of the sequences, deciduous Quercus expands during the warm phases of the Tardiglacial and declines in the cold periods, during which pines were widespread. However, pine is the main taxa in Southwestern Portugal throughout the whole Pleistocene-Holocene transition, despite the variations in their overall percentage in the sequence and in the proportion of Pinus sylvestris and Pinus pinaster, the latter expanding during the warmer phases. The expansion of thermophillous taxa such as Olea europaea, Pistacia and Arbutus unedo during the warmer phases also shows considerable variation, depending on site location. Their significant and quick expansion during the Early Holocene is only justified by their survival in refugia throughout the territory during the colder phases, as documented by pollen and charcoal data. The number of sites where a faunal record was preserved is very small (Supplementary material - Table 2), which hinders any interpretations of changes in subsistence strategies and the possible relationship with paleoenvironmental evolution. However, the study of faunal assemblages from cave sequences such as Caldeirão (Davis, 2002; LLoveras et al., 2011), Lapa do Suão (Haws and Valente, 2006), Lapa do Picareiro (Bicho and Haws, 2012) and Lapa dos Coelhos (Gameiro et al., 2017) allowed us to observe, after the Solutrean, a decrease in the variability of the represented species and an increase of fish, as well as birds, and small mammals, particularly rabbits (Oryctolagus cuniculus), reaching percentages above 90% in the Final Magdalenian levels of Lapa dos Coelhos for example (Gameiro et al., 2017). This has already been pointed out (Davis, 2002, 2019; Hockett and Haws, 2002; Stiner et al., 2003; Zilhão, 1997a): « the emphasis on rabbits may have been the result of changes on their ecology as well as of the invention of new hunting technologies (such as snares or nets) or of an increase in human population numbers» (LLoveras et al., 2011:2448). Deer (Cervus elaphus) and wild boar (Sus scrofa) are also present during the Magdalenian and continue to represent a significant proportion of the hunted game in interior areas until the Early Mesolithic (Araújo et al., 2014; Araújo, 2016b; Carvalho et al., 2016), when marine and estuarine species occur but on coastal areas only (Dupont and Araújo, 2010; Araújo, 2016a). It has been proposed that, during the Magdalenian, settlement and subsistence strategies would be closer to the forager type, as opposed to the collector-oriented strategies of the Solutrean (Zilhão, 1997a). These changes in diet and in subsistence strategies, as well as the appearance of new technologies, should had have an ideological impact, of which the thematic changes in art between the Magdalenian and the Azilian is the more evident feature. In fact, red deer not only becomes the most depicted species, but for the first time a single species dominates the graphic corpus by far (aurochsen, ibexes and horses are represented in similar numbers during the Gravettian and the first half of the Solutrean, the same happening with horses and aurochsen during the phases dated to between the Upper Solutrean and the final Magdalenian). Inter-regional comparisons (Table 4) are only available for the Final Magdalenian. Based on the archaeological record from different regions, recent studies (Gameiro, 2012) have argued that regional variability reflects the adaptation of local groups to different raw material environments, but it has been demonstrated that some technological options were, nonetheless, maintained during the Tardiglacial for extensive periods of time. Moreover, variability in the typology of microliths might relate to the last phase of transformation (i.e. retouch) rather than to blank production strategies (Gameiro, 2012; Gameiro et al., 2013). The use of cores-on-flakes (burin or carinated endscraper type) increased from 20,000 to 11,000 cal BP. The use of this debitage strategy enables an efficient production of longer, narrower but thicker C. Gameiro, et al. Quaternary International xxx (xxxx) xxx–xxx 21 bladelets, naturally pointed and with a straight profile. This increase was accompanied by a typological diversification of armatures. The use of cores-on-flakes has not been identified in Fariseu layer 4 (Côa Valley) and in the sites of the Guadiana Basin (Malhada do Mercador I and Monte da Ribeira) it does not reach the proportions recorded in the Estremadura sites. In the Sabor, Côa and Vouga Valleys and in the Guadiana Basin, the predominant use of quartz to make small microlithic blanks (bladelets and chips) explains the use of bipolar debitage on anvil, which is effective in these cases (Fig. 10: 4). The use of hyaline quartz crystals for bladelet production (Fig. 8: 5 and Fig. 9: 1–3) was identified in the Sabor (Gaspar et al., 2016a), Côa and Vouga Valleys (Gameiro, 2009, 2012; Gameiro et al., 2018); this strategy was also identified in these regions during the various phases of the Gravettian (Aubry, 2009; Klaric, 2009; Gameiro et al., 2018). In the Guadiana Basin, lydite and jaspers, sometimes gathered in the form of small slabs, were also similarly exploited, from a natural edge or dihedral (Gameiro, 2012; Almeida et al., 2013). Probably, this use of natural edges or dihedrals, requiring no more than the preparation of the striking platform, attests to the same technical concept: minimum preparation of the volumes for maximum profitability in the production of small, easily transportable blanks. With regards to differences in technical use, there appears to be an increase in the use of soft-stone hammers (Gameiro, 2017) and pressure retouch (Gameiro, 2012) from the Final Magdalenian onwards. 7. Conclusions Concerning the transition between the end of the Pleistocene and the beginning of the Holocene, opinions have been divided between researchers who observe a technological rupture at an earlier moment (ca. 11,500 cal BP) (Zilhão, 1997a) and others who see continuity until 10,000 cal BP (Bicho and Haws, 2012). The use of the territory and subsistence options seem to reflect more variation, probably correlated with environmental changes. The expansion of forests and the increase of flooded areas, originated by sea rise, favoured costal circulation, as can be inferred from Early Holocene site location and subsistence choices (Araújo, 2016a, 2016b). Taphonomic issues must be taken into account as well, such as the differential preservation of bone tools, which are fundamental in the phasing of the Franco-Cantabrian area, but also the role of rates of sediment accumulation and of climate oscillations between the cold phases of GS-1 (Rasmussen et al., 2014). The existence of sedimentary hiatus, the difficulties in discerning the contemporaneity of human occupations due to a radiocarbon plateau (Naudinot et al., 2019), the absence of radiocarbon dates for some of the sites, the existence of assemblages featuring uncharacteristic industries and, above all, the non-existence of a complete stratigraphic sequence covering the Pleistocene-Holocene transition (14,000 to 10,500 cal BP) keep us from attaining a more accurate understanding of this transition period. Nevertheless, in terms of armature typology, a four-stage sequence can be discerned: 1) Upper Magdalenian with axial points rather than backed bladelets, quite common in previous phases; 2) Final Magdalenian with an increase in the diversity of armature types; 3) Azilian with geometric microliths, curved backed points (Azilian points) and Malaurie points, and 4) Early Mesolithic without retouched bladelet tools or, at best, a persistence of Azilian armature types. For the time being, there does not seem to be enough resolution to confirme a Late Azilian phase (with curved backed points) followed by a Laborian phase (with Malaurie points), as previously suggested (Aubry et al., 2017). It has been said that after 14, 500 cal BP lithic industries show a tendency towards regionalization and follow different evolution lines, as compared to the rest of SW Europe (Zilhão, 1997a,b). Nevertheless, the effects of GS-1 climatic change might have restored extra-regional contacts allowing the import of typologies from other regions (Aubry et al., 2017). Blank production strategies remain similar between Final Magdalenian and Azilian but armature typology displays types suggestive of contacts and closer relationships with a broader region of Southwestern Europe: curved backed points (Azilian points) and backed and truncated points (Malaurie points) are present. Another (rare) example would be the Ahrensburg point (Fig. 11) recovered at the Vau site. On the other hand, the Early Holocene period seems to be characterized by a trend towards regionalism: simple expeditive debitage using local raw materials. The expansion of contacts during GS-1 (Rasmussen et al., 2014). is also evident in art. In fact, as referred above, a striking morphological contrast between the northern Iberian animal depictions and their souhern counterparts can be observed during the Magdalenian. In Côa and Sabor, as in the Franco-Cantabrian zone, the naturalistic shapes attached to Leroi-Gourhan's style IV are common, while in the southern sites of Portugal, as in the remaining meridional sites of the Iberian Peninsula, no such figures are known to exist. During GS-1 however, the same animal figures with geometric bodies, generally filled-in, are found not only in the northern and southern sites of Iberia but also in other regions of Southwestern Europe (Santos et al., 2018). This overview of the Late Pleistocene-Holocene transition in the Western façade of Iberia aims to summarize existent data but also to demonstrate that, despite the significant enhancement of research during the last twenty-five years, we still don't have enough data to accurately characterise the beginning of the Holocene. Acknowledgments We wish to thank Mathieu Langlais, Didac Roman and Celia Fat Cheung for the invitation to participate in this volume. We thank J.P. Ruas for the photos and A. Lucena for English text translation and revision. A special thanks to Simon Davis, Cidália Duarte, João Zilhão and to the anonymous reviewers whose comments helped improve and clarify this manuscript. Financial support for this research has been provided by FCT - Fundação para a Ciência e a Tecnologia in the framework of projects: Upper Paleolithic and Preventive Archaeology in Portugal: challenges and opportunities (PTDC/HAR-ARQ/30779/2017) and Neanderthal to Anatomically Modern Human transition in the Côa Valley: Environments, Symbolism and Social networks (PTDC/EPH-ARQ/ 0326/2014). Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.quaint.2020.03.018. References Almeida, F., 2008. Big Puzzles, Short Stories: advantages of refitting for micro-scale spatial analysis of lithic scatters from Gravettian occupations in Portuguese Estremadura. In: Aubry, T., Almeida, F., Araújo, A.C., Tiffagom, M. (Eds.), Espace et temps: Quelles diachronies, quelles synchronies, quelles échelles? Procedings of the UISPP meeting (4-9 Setembro, Lisboa, 2006). BAR International Series 1831, Oxford, pp. 69–79. Almeida, F., 2013. Testemunhos do Paleolítico no Regolfo do Alqueva. Resultados do Bloco 1 do Plano de Minimização de Impactes. EDIA/DRCALEN [s.l.]. Almeida, F., Angelucci, D., Gameiro, C., Correia, J., Pereira, T., 2004. Novos dados para o Paleolítico Superior final da Estremadura Portuguesa: resultados preliminares dos trabalhos arqueológicos de 1997-2003 na Lapa dos Coelhos (Casais Martanes, Torres Novas). Promontoria, Faro. 2 (2), 157–192. Almeida, F., Gameiro, C., Pereira, T., Quelhas, A., 2013. O paleolítico superior. In: Almeida, F. (Ed.), Testemunhos do Paleolítico no Regolfo do Alqueva. Resultados do Bloco 1 do Plano de Minimização de Impactes. EDIA/DRCALEN, pp. 93–245 [s.l.]. Andrade, M., Matias, H., Garcia, M., 2013. Under Pressure. Remontagens de pedra lascada no contexto de intervenções de emergência: o caso dos sítios do Tardiglaciar da área de Leiria. Arqueol. Hist. 62–63, 39–50. Angelucci, D., 2002. The geoarcheological context. In: In: Zilhão, J., Trinkaus, E. (Eds.), Portrait of the Artist as a child. The Gravettian Human skeleton from the Abrigo do Lagar Velho and its Archaeological Context, Trabalhos de Arqueologia, vol. 22. Instituto Português de Arqueologia, Lisboa, pp. 58–91. Araújo, A.C., 2003. O Mesolítico inicial da Estremadura. In: Gonçalves, V. (Ed.), Muita gente, poucas antas? IPA, Lisboa, pp. 101–114. C. Gameiro, et al. Quaternary International xxx (xxxx) xxx–xxx 22 Araújo, A.C., 2016a. The significance of marine resources during the early mesolithic in Portugal. In: Dupond, C., Marchand, G. (Eds.), – Archéologie des chasseurs-cuillers maritimes. De la function des habitats à l’organisation de l’espace litoral. Paris, Societé préhistorique française. SPF, Paris, pp. 127–143. Araújo, A.C., 2016b. Une histoire des premières communautés mésolithiques au Portugal. BAR. Araújo, A.C., Almeida, F., 2013. Barca do Xerez de Baixo: um testemunho invulgar das últimas comunidades de caçadores-recolectores do Alentejo interior. Edia, Beja. Araújo, A.C., Moreno-García, M., Gabriel, S., 2014. Para além de Toledo. Outros dados, novas revisões e algumas reflexões sobre o Mesolítico antigo do litoral da Estremadura. In: Revista Portuguesa de Arqueologia. vol. 17. DGPC, Lisboa, pp. 5–34. Aubry, T., 2001. L'occupation de la basse vallée du Côa pendant de Paléolithique Supérieur. In: Zilhão, J., Aubry, T., Carvalho, A.F. (Eds.), Les premiers hommes modernes de la Péninsule Ibérique. Actes du colloque de la commission VIII de l'UISPP. Vila Nova de Foz Côa, 22-24 de Outubro de 1998. IPA (Trabalhos de Arqueologia 17), Lisboa, pp. 253–280. Aubry, T., 2009. 200 séculos da história do Vale do Côa: incursões na vida quotidiana do caçadores-artistas do Paleolítico, vol. 52 IGESPAR (Trabalhos de Arqueologia, Lisboa. Aubry, T., Almeida, M., Dimuccio, L., Gameiro, C., Neves, M., Klaric, L., 2008a. Caracterisation et discontinuites des registes pedosedimentaires de l'Occident peninsulaire entre 30 000 et 10 000 BP: implications sur l’interpretation archéologique. In: Aubry, T., Almeida, F., Araújo, A.C., Tiffagom, M. (Eds.), Espace et temps: Quelles diachronies, quelles synchronies, quelles échelles? Procedings of the UISPP meeting (4-9 Setembro, Lisboa, 2006). BAR International Series 1831, Oxford, pp. 9–21. Aubry, T., Barbosa, F., Gameiro, C., Luís, L., Matias, H., Santos, A.T., Silvestre, M., 2015. – De regresso à Cardina 13 anos depois. Resultados preliminares dos trabalhos arqueológicos de 2014 no Vale do Côa, vol. 18. Revista Portuguesa de Arqueologia, pp. 5–26. Aubry, T., Carvalho, A., Zilhão, J., 1997. Arqueologia. In: Zilhão, J. (Ed.), Arte Rupestre e Pré-História do Vale do Côa: Trabalhos de 1995-1996. Ministério da Cultura, Lisboa, pp. 74–209. Aubry, T., Dimuccio, L., Almeida, M., Neves, M., Angelucci, D., Cunha, L., 2011. Palaeoenvironmental forcing during the Middle-Upper Palaeolithic transition in central-western Portugal. Quat. Res. 75, 66–79. https://doi.org/10.1016/j.yqres. 2010.11.002. Aubry, T., Dimuccio, L., Mèrce Bergadà, M., Sampaio, J., Sellami, F., 2010. Palaeolithic engravings and sedimentar environments in the Côa river Valley (Portugal): implications for the detection, interpretation and dating of open-air rock art. J. Archaeol. Sci. 37, 3306–3319. Aubry, T., Gameiro, C., Santos, A.T., Luís, L., 2017. Existe Azilense em Portugal? Novos dados sobre o Tardiglaciar no Vale do Côa. In: Arnaud, J.M., Martins, A. (Eds.), Actas do II Congresso de Arqueologia da Associação dos Arqueólogos Portugueses. AAP, Lisboa, pp. 403–418. Aubry, T., Gameiro, C., Mangado Llach, J., Luís, L., Matias, H., Pereiro, T., 2016. Upper Palaeolithic lithic raw material sourcing in Central and Northern Portugal as an aid to reconstructing hunter-gatherer societies. J. Lithic Stud. 3 (2). https://doi.org/10. 2218/jls.v3i2.1436. Aubry, T., Luís, L., Mangado, J., Matias, H., 2012. We will be known by the tracks we leave behind: exotic lithic raw materials, mobility and social networking among the Côa Valley foragers (Portugal). J. Anthropol. Archaeol. 31 (4), 528–550. https://doi. org/10.1016/j.jaa.2012.05.003. Aubry, T., Moura, H., 1993. Nouvelles données sur les occupations paleolithiques de la region de Redinha (Serra de Sicó, Portugal). In: 3° Reunião do Quaternário Ibérico. Coimbra. GTEQ/AEQUA, pp. 439–449. Aubry, T., Neves, M., Almeida, M., Mangado Llach, J., 2008b. Modalidades de aprovisionamento em matérias-primas líticas no Baixo Mondego durante o Holocénico: dados novos e revisão de séries arqueológicas. In: In: Hernández, M., Soler, J., e López, J. (Eds.), Actas do IV° Congreso del Neolítico Peninsular MARQ, Alicante, pp. 258–265. Baptista, A.M., Santos, A.T., 2013. A arte rupestre do Guadiana português na área de influência do Alqueva, vol. 1 EDIA & DRCALEN (Memórias d'Odiana, nova série, Beja. Benedetti, M., Haws, J., Bicho, N., Friedl, L., Ellwood, B., 2019. Late Pleistocene Site Formation and Paleoclimate at Lapa Do Picareiro. Portugal, pp. 1–29 Geoarchaeology. Bencatel, J., Sabino-Marques, H., Álvares, F., Moura, A.E., Barbosa, A.M., 2019. Atlas de Mamíferos de Portugal. Évora. Bicho, N., 1994. The end of the palaeolithic and the mesolithic in Portugal. Curr. Anthropol. 35 (5), 664–674. Bicho, N., 1997. Spatial, technological, and economic organization after the last glacial maximum in Portuguese prehistory. In: Fullola, J., Soler, N. (Eds.), El món mediterrani desprès del pleniglacial (18 000-12 000). Museu d'Arqueologia de Catalunya, Girona, pp. 213–223. Bicho, N., 2000. Technological Change in the Final Upper Palaeolithic of Rio Maior, vol. 8 ARKEOS, Tomar. Bicho, N., 2003. A importância dos recursos aquáticos na economia dos caçadores do Paleolítico e Epipaleolítico do Algarve. Xelb 4, 11–26. Bicho, N., Carvalho, A.F., Gonzalez Sainz, C., Sanchidrián, J.L., Villaverde Bonilla, V., Straus, L.G., 2007. The upper Paleolithic rock art of Iberia. J. Archaeol. Method Theor 14 (1), 81–151. Bicho, N., Cascalheira, J., Cortés, M., Gibaja, J., Évora, M., Manne, T., Marreiros, J., Mendonça, C., Pereira, T., Regala, F., 2009. A ocupação humana durante o Plistocénico Final no Algarve Ocidental. In: Boski, T., Moura, D., Gomes, A. (Eds.), VII reunião do Quaternário Ibérico Livro de Resumos. Universidade do Algarve, Faro, pp. 171–174. Bicho, N., Haws, J., 2012. The magdalenian in central and southern Portugal: human ecology at the end of the Pleistocene. Quat. Int. 272–273, 6–16. Bicho, N., Haws, J., Almeida, F., 2011. Hunter–gatherer adaptations and the Younger Dryas in central and southern Portugal. Quat. Int. 242 (2), 336–347. Bicho, N., Haws, J., Hockett, B., Markova, A., Belcher, W., 2003. Paleoecologia e ocupação humana da Lapa do Picareiro: resultados preliminares. Rev. Portuguesa Arqueol. 6 (2), 49–81. Bicho, N., Manne, T., Cascalheira, J., Mendonça, C., Évora, M., Gibaja, J., Pereira, T., 2010. O Paleolitico superior do sudoeste da Península Ibérica: o caso do Algarve. In: Mangado, X. (Ed.), El paleolítico superior peninsular. Novedades del Siglo XXI. SERP, Barcelona, pp. 219–238. Breuil, H., 1913. Les Subdivisions du paléolithique supérieur et leur signification. Congrès international d'anthropologie et d'archéologie préhistoriques - compte-rendu de la XIVème session, vol. 1. Albert Kündig, Genève, pp. 165–238 1912, tome. Breuil, H., 1917. Glanes Paléolithiques anciennes dans le Bassin du Guadiana. L'Anthropologie. Paris. 2 (1–2), 1–19. Bronk Ramsey, C., 2017. Methods for Summarizing Radiocarbon Datasets. Radiocarbon 59 (2), 1809–1833. https://doi.org/10.1017/RDC.2017.108. Published online by Cambridge University Press: 20 November 2017. Bueno Ramirez, P., Balbín Behrmann, R., Alcolea González, J., 2007. Style V dans le bassin du Douro. Tradition et changement dans les graphies des chasseurs du Paléolithique Supérieur européen. L’ Anthropologie 111, 549–589. Callapez, P., 2003. Moluscos marinhos e fluviais do Paleolítico superior da Gruta do Caldeirão (Tomar, Portugal): evidências de ordem sistemática, paleobiológica e paleobiogeográfica. Rev. Portuguesa Arqueol. 6 (1), 5–15. Carvalho, A.F., Valente, M.J., 2005. Novos contextos conquíferos pré-históricos na Costa Vicentina. In: Actas do 2.° Encontro de Arqueologia do Algarve (Xelb, 5). Câmara Municipal de Silves, Silves, pp. 9–26. Carvalho, A.F., Valente, M.J., Marreiros, J., 2016. In: Mosaic Adaptations of Early Holocene Hunter-Gatherers in Central Portugal: Lithic and Faunal Evidence from the Open-Air Site of Costa Do Pereiro (Torres Novas), vol. 63. pp. 169–187 Quartär. Chauviere, F.-X., 2002. Industries et parures sur matières dures animales du Paléolithique supérieur de la grotte de Caldeirão (Tomar, Portugal). Rev. Portuguesa Arqueol. 5 (1), 5–28. Collado Giraldo, H., 2006. Arte rupestre en la cuenca del Guadiana: el conjunto de grabados del Molino Manzánez (Alconchel—Cheles), EDIA (Memórias d'Odiana—Estudos arqueológicos do Alqueva, vol. 4 Beja. Cunha, L., Almeida, M., Neves, M., Dimuccio, L., Aubry, T., 2006. Contributo da sequência cultural pleistocénico-holocénica para a compreensão da génese e evolução do canhão fluviocársico do Vale das Buracas, vol. 3. Publicações da Associação Portuguesa de Geomorfólogos, Coimbra, pp. 69–75. Davis, S.J.M., 2002. The Mammals and Birds from the Gruta Do Caldeirão, Portugal, vol. 5. Revista Portuguesa de Arqueologia, pp. 29–98. Davis, S.J.M., 2019. Rabbits and Bergmann's rule: how cold was Portugal during the last glaciation? Biol. J. Linn. Soc. XX, 1–24 2019. Davis, S.J.M., Robert, I., Zilhão, J., 2007. Caldeirao cave (Central Portugal) - whose home? Hyaena, man, bearded vulture. Cour. Forschungsinst. Senckenberg 259, 213–226. Dean, R., Valente, M., Carvalho, A., 2011. The mesolithic/neolithic transition on the Costa Vicentina, Portugal. Quat. Int. 264. https://doi.org/10.1016/j.quaint.2011.10. 024. d'Errico, F., Vanhaeren, M., 2002. Criteria for identifying red deer (Cervus elaphus) age and sex from their canines. Application to the study of upper palaeolithic and mesolithic ornaments. J. Archaeol. Sci. 29 (2), 211–232. Dupont, C., Araújo, A.C., 2010. Ressuscitar o concheiro mesolítico de Toledo (Lourinhã, Portugal): uma lixeira repleta de invertebrados marinhos. ISSN 1134–6787, 53–62 Férvedes (Vilalba, Lugo), 6. Figueiredo, S., 2014. O concelho de Mogadouro e a revelação de 20 000 anos de arte rupestre. In: I Encontro de Arqueologia de Mogadouro. Mogadouro, Abril de 2013. Actas. Município de Mogadouro, Mogadouro, pp. 9–31. Figueiredo, S., Nobre, L., Xavier, P., Gaspar, R., Carrondo, J., 2016. First approach to the chronological sequence of the engraved plaques of the Foz do Medal alluvial terrace in Trás-os-Montes, Portugal, ARPI. In: Arqueología y Prehistoria del Interior peninsular, 4 (extra. Homenaje a Rodrigo de Balbín Behrmann). Universidad de Alcalá de Henares, Madrid, pp. 64–77. Figueiredo, S., Xavier, P., Nobre, L., Giraldo, Collado, 2015a. Placas móveis com grafismos rupestres paleolíticos do Terraço do Medal (Nordeste, Portugal): uma primeira análise a técnicas e estilos. In: In: Collado Giraldo, H., Garcia Arranz, J.J. (Eds.), Symbols in the Landscape: Rock Art and its Context. Proceedings of the XIX International Rock Art Conference IFRAO 2015 (Cáceres, Spain, 31 August — 4 September 2015) 37. Instituto Terra e Memória [Arkeos, Tomar, pp. 573–1588. Figueiredo, S., Xavier, P., Silva, A., Neves, D., Dominguez Garcia, I., 2015b. The Holocene transition and post-palaeolithic rock art from the Sabor Valley. In: Medina-Alcaide, M.-Á., Romero Alonso, A.J., Ruiz-Marquez, R.M., Sanchidrián Torti, J.L. (Eds.), Sobre rocas y huesos: las sociedades pré-históricas y sus manifestaciones plásticas. Universidad de Córdoba y Fundación Cueva de Nerja, Córdoba, pp. 192–203. Fletcher, W.J., Boski, T., Moura, D., 2007. Palynological evidence for environmental and climatic change in the lower Guadiana valley, Portugal, during the last 13 000 years. Holocene 17, 481–494. Gabriel, S., 2011. A exploração dos recursos ictícos. In: Araujo, A.C. (Ed.), O concheiro de Toledo no contexto do Mesolítico Inicial do litoral da Estremadura Lisboa: Ministério da Cultura. Instituto de Gestão do Património Arquitectónico e Arqueológico, pp. 127–144. Gabriel, S., Béarez, P., 2009. Caçadores-pescadores do Vale do Côa: os restos de fauna do sítio do Fariseu. In: Aubry, T. (Ed.), - 200 séculos da História do Vale do Côa: incursões na vida quotidiana dos caçadores-artistas do Paleolítico. IGESPAR (Trabalhos de Arqueologia 52), Lisboa, pp. 331–339. C. Gameiro, et al. Quaternary International xxx (xxxx) xxx–xxx 23 Gameiro, C., 2009. Utensílios e suportes microlíticos do Magdalenense final no Vale do Côa: o exemplo da U.E. 4 do Fariseu. In: Aubry, T. (Ed.), - 200 séculos da História do Vale do Côa: incursões na vida quotidiana dos caçadores-artistas do Paleolítico. IGESPAR (Trabalhos de Arqueologia 52), Lisboa, pp. 256–268. Gameiro, C., 2012. La variabilité régionale des industries lithiques de la fin du Paléolithique Supérieur au Portugal. [PhD Thesis. Université de Paris IPanthéonSorbonne. Gameiro, C., 2017. A utilização do percutor em pedra branda na debitagem lamelar durante o Magdalenense na Estremadura portuguesa/Soft hammerstone percussion use in bladelet debitage during the Magdalenian at Portuguese Estremadura. J. Lithic Stud. 4 (3). https://doi.org/10.2218/jls.v4i2.2532. Gameiro, C., 2018. A tecnologia lítica do fim do Tardiglaciar no Centro de Portugal: o exemplo do Abrigo 1 de Vale de Covões (Soure), vol. 2. Ophiussa, Lisboa, pp. 5–24. Gameiro, C., Aubry, T., Almeida, F., 2013. A variabilidade regional das indústrias líticas do final do Paleolítico Superior em Portugal. In: Arqueologia em Portugal – 150 anos actas do Congresso da Associação dos Arqueólogos Portugueses, Lisboa. Associação dos Arqueólogos Portugueses, Lisboa, pp. 277–287. Gameiro, C. & Almeida, F., (in press). Existiu um povoamento tardiglaciar na Bacia do Guadiana? Revisão dos dados. Actas da Mesa redonda: Transição PleistocénicoHolocénico, 13 de outubro de 2018, Museu Arqueológico do Carmo, Lisboa. Gameiro, C., Davis, S., Almeida, F., 2017. A sequência estratigráfica da lapa dos coelhos: funcionalidade e subsistência ao longo do pleistocénico superior no sopé na serra de aire (Portugal). In: Arnaud, J.M., Martins, A. (Eds.), Arqueologia em Portugal 2017Estado da questão. Associação dos Arqueólogos Portugueses, Lisboa, pp. 357–374. Gameiro, C., Gomes, S., Manzano, C., Costa, B., Ameijenda, A., Oliveira, L., MonteiroRodrigues, S., Gomes, A., Oliveira, C., Tereso, J., Matias, H., Aubry, T., 2018. The Pleistocene-Holocene Transition: New Data from the Sites of Rôdo, Vau and Bispeira 8 (Vouga Valley, Portugal). Poster 8th Annual Meeting of the European Society for the study of Human Evolution. https://docplayer.net/87563946-European-societyfor-the-study-of-human-evolution.html p.94. García-Díez, M., 2013. La expresión gráfica de La Peña de Estebanvela (Segovia) en el contexto de los últimos grupos cazadores-recolectores europeos. In: In Cacho Quesada, C. (Coord. de), Ocupaciones magdalenienses en el interior de la Península Ibérica. La Peña de Estebanvela (Ayllón, Segovia). Junta de Castilla y León/ CSIC, pp. 471–514. Gaspar, R., Ferreira, J., Carrondo, J., Silva, M.J., 2016a. The use of quartz during the upper paleolithic and early mesolithic in Sabor Valley (NW Iberia): the Foz do medal case. Quat. Int. 424, 98–112. Gaspar, R., Ferreira, J., Carrondo, J., Silva, M.J., García-Vadillo, F.J., 2016b. Open-air gravettian lithic assemblages from northeastern Portugal: the Foz do medal site. Quaternary International 406. Elsevier & INQUA, Amsterdam, pp. 44–64. Gomes, M.V., 2002. Arte rupestre em Portugal — perspectiva sobre o último século. Arqueologia 2000. Balanço de um século de investigação arqueológica em Portugal, vol. 54. AAP (Arqueologia & História, Lisboa, pp. 139–194. Gomes, M.V., 2007. Os períodos iniciais da arte do Vale do Tejo (Paleolítico e Epipaleolítico), Cuadernos de Arte Rupestre, vol. 4. Gobierno de la Región de Murcia, Murcia, pp. 81–116. Gomes, M.V., 2010. Arte rupestre do Vale do Tejo. Um ciclo artístico-cultural pré e protohistórico. Faculdade de Ciências Sociais e Humanas, Lisboa (Tese de Doutoramento, policopiada). Haws, J., Valente, M.J., 2006. Animal carcass utilization during the late upper paleolithic occupation of lapa do Suão, Portugal. In: Haws, J., Brugal, J.-P., Hockett, B. (Eds.), Paleolithic Zooarchaeology in Practice. BAR international Series, pp. 29–37. Heleno, M., 1944. O problema Capsense: contribuição portuguesa para a sua revisão. Comunicação ao Instituto de Arqueologia na sessão de Abril de 1944, Lisboa. Hockett, B., 1999. Taphonomy of a Carnivore-Accumulated Rabbit Bone Assemblage from Picareiro Cave, Central Portugal. Journal of Iberian Archaeology 1, 251–256. Hockett, B., Bicho, N., 2000. The rabbits of Picareiro cave: small mammal hunting during the late upper Palaeolithic in the Portuguese Estremadura. J. Archaeol. Sci. 27, 715–723. Hockett, B., Haws, J., 2002. Taphonomic and methodological perspectives of leporid hunting during the Upper Palaeolithic of the Western Mediterranean basin. J. Archaeol. Method Theor 9, 269–302. Klaric, L., 2009. Les systèmes de production de supports d’armatures et leur place dans la gestion de ressources lithiques : une voie privilégiée pour la compréhension des sociétés gravettiennes de la Vallée du Côa. In: Aubry, T. (Ed.), - 200 séculos da História do Vale do Côa: incursões na vida quotidiana dos caçadores-artistas do Paleolítico. IGESPAR (Trabalhos de Arqueologia 52), Lisboa, pp. 247–256. Klaric, L., Guillermin, P., Aubry, T., 2010. Des armatures variées et des modes de productions variables. Gall. Prehist. 51, 113–154. Langlais, M., 2007. Dynamiques culturelles des sociétés magdaléniennes dans leurs cadres environnementaux. Enquête sur 7 000 ans d’évolution de leurs industries lithiques entre Rhône et Èbre. [PhD thesis : Université de Toulouse II/Universitat de Barcelona. Leroi-Gourhan, A., 1995. In: Préhistoire de l'art Occidental. Nouvelle édition revue et augmentée, vol. 1 Citadelles & Mazenod (L'Art et les Grands Civilisations, Paris 1965. LLoveras, L., Moreno-Garcia, M., Nadal, J., Zilhão, J., 2011. Who brought in the rabbits? Taphonomical analysis of mousterian and solutrean leporid accumulations from Gruta do Caldeirão (Tomar, Portugal). J. Archaeol. Sci. 38, 2434–2449. Lorblanchet, M., 1985. Premiers résultats de nouvelles recherches à l'abri Murat, Préhistoire Quercinoise, 2, Cabrerets, Association des "Amis du Musée A. Lemozi. pp. 58–94. Lorblanchet, M., 1989. De l'art naturaliste des chasseurs de rennes à l'art géométrique du Mésolithique dans le Sud de la France. In: Colóquio Internacional de Arte Préhistórica — Nos 25 anos da Gruta do Escoural. Câmara Municipal de Montemor-oNovo, Montemor-o-Novo, pp. 95–124 (Almansor, 7). Lorblanchet, M., 1996. Du Magdalénien à l'Azilien en Quercy. In: La vie préhistorique Dijon, pp. 282–285 Éditions Faton. López-Sáez, J., Figueiral, I., Cruz, D., 2017. Palaeoenvironmental and vegetation dynamics in Serra da Nave (Alto Paiva, Beira Alta, Portugal) during the Late Pleistocene and the Holocene. Estudos Pré-Históricos, 17. In: Proceedings Mesa-Redonda A Préhistória e a Proto-história no Centro de Portugal: avaliação e perspectivas de futuro (Mangualde, 26 - 27 November 2011), pp. 11–23. Manne, T., Cascalheira, J., Évora, M., Marreiros, J., Bicho, N., 2012. Intensive subsistence practices at Vale Boi, an upper paleolithic site in southwestern Portugal. Quat. Int. 264, 83–99. Marks, A., 2000. Upper Paleolithic occupation of Portugal: residents Vs. Visitors. In: Rodrigo Balbín et al.li, Coord - Paleolítico da Península Ibérica. Vol. II - Actas do 3° Congresso de Arqueologia Peninsular (UTAD/Vila Real, Setembro de 1999). ADECAP, Porto, pp. 343–349. Marks, A., Mishoe, M., 1997. The magdalenian of Portuguese Estremadura. In: Fullola, J., Soler, N. (Eds.), El món mediterrani desprès del pleniglacial (18 000-12 000). Museu d'Arqueologia de Catalunya, Girona, pp. 225–232. Mendonça, C., 2009. A Tecnologia Lítica No Tardiglaciar Do Algarve. [Master thesis. Universidade do Algarve]. Mercier, N., Valladas, H., Aubry, T., Zilhão, J., Jorons, J.-L., Reyss, J.-L., Sellami, F., 2006. Fariseu: first confirmed open-air Palaeolithic parietal art site in the Côa Valley (Portugal). Antiquity 80. Monteiro, P., 2018. Economia de recoleção da madeira para combustível dos últimos caçadores-recolectores de Muge: estudo antracológico dos concheiros mesolíticos do Cabeço da Amoreira e Cabeço da Arruda (Santarém, Portugal). PhD thesis. Faculdade de Ciências Humanas e Sociais, Universidade do Algarve. Monteiro-Rodrigues, S., Angelucci, D.E., 2004. New data on the stratigraphy and chronology of the prehistoric site of Prazo (Freixo de Numão). Revista Portuguesa de Arqueologia, vol. 7. Instituto Português de Arqueologia, Lisboa, pp. 39–60 1. Monteiro-Rodrigues, S., Figueiral, I., López-Sáez, J.A., 2006. Indicadores paleoambientais e estratégias de subsistência no sítio pré-histórico do Prazo (Freixo de Numão – Vila Nova de Foz Côa – norte de Portugal), III Congresso de Arqueologia de Trás-osMontes. Alto Douro e Beira Interior – Debates no Vale do Côa. ACDR de Freixo de Numão, Vila Nova de Foz Coa, pp. 96–119. Moreno Garcia, M., 2011. Exploração dos recursos faunísticos de origem terrestre. In: Araujo, A.C. (Ed.), O concheiro de Toledo no contexto do Mesolítico Inicial do litoral da Estremadura Lisboa: Ministério da Cultura. Instituto de Gestão do Património Arquitectónico e Arqueológico, pp. 99–124. Naudinot, N., Fagnart, J.-P., Langlais, M., Mevel, L., Valentin, B., 2019. In: Les dernières sociétés du Tardiglaciaire et des tout débuts de l'Holocène en France. Bilan d’une trentaine d’années de recherche, vol. 59 Gallia Préhistoire. http://journals. openedition.org/galliap/1394Gallia. Naughton, F., Sanchez Goñi, M.F., Rodrigues, T., Salgueiro, E., Costas, S., Desprat, S., Duprat, J., Michel, E., Rossignol, L., Zaragosi, S., Abrantes, F., 2016. Climate variability across the last deglaciation in NW Iberia and its margin. Quat. Int. 414, 9–22. https://doi.org/10.1016/j.quaint.2015.08.07. Pereira, D.I., Azevedo, T.M., 1995. Evolução quaternária do Graben da Vilariça (Trás-os Montes, NE Portugal). 20. Caderno do Laboratorio Xeolóxico de Laxe, pp. 123–137. Pereira, T., Andrade, C., Costa, M., Farias, A., Mirão, J., Carvalho, A.F., 2016. Lithic economy and territory of Epipaleolithic hunter–gatherers in the Middle Tagus: the case of Pena d'Água (Portugal). Quat. Int. 412, 135–144. https://doi.org/10.1016/j. quaint.2015.08.081. Part A. Pereira, T., Carvalho, A., 2015. Abrupt technological change at the 8.2 ky calBP climatic event in Central Portugal. The Epipaleolithic occupation of Pena d’Água Rock-shelter. Comptes Rendus Palevol 14, 423–435. Pereiro, T., 2013. O sítio Mesolítico Antigo da Cruz da Areia: uma abordagem (Geo) Arqueológica. Master Thesis. Faculdade de Ciências da Universidade de Lisboa. Queiroz, P., 1999. Ecologia Histórica da Paisagem do Noroeste Alentejano. PhD thesis in Biology. Faculdade de Ciências, Universidade de Lisboa. Ramil Rego, P., Muñoz-Sobrino, C., Rodríguez-Guitián, M., Gómez-Orellana, L., 1998. Differences in the vegetation of the north Iberian Peninsula during the last 16,000 years. Plant Ecol. 138, 41–62. Raposo, L., Silva, A.C., 1984. O Languedocense: ensaio de caracterização morfotécnica e tipológica. O Arqueólogo Português. Série. 4 (2), 87–166. Rasmussen, S., Bigler, M., Blockley, S., Blunier, T., Buchardt, S., Clausen, H., Cvijanovic, I., Dahl-Jensen, D., Johnsen, S., Fisher, H., Gkinis, V., Guillevic, M., Hoek, W., Lowe, J., Pedro, J., Poop, T., Seierstad, I., Steffensen, J., Svensson, A., Vallelonga, P., Vinther, B., Walker, J., Wheatley, J., Winstrup, M., 2014. A stratigraphic framework for abrupt climatic changes during the Last Glacial period based on three synchronized Greenland ice-core records: refining and extending the INTIMATE event stratigraphy. Quat. Sci. Rev. 106, 14–28. Regala, F., 2011. Os adornos do paleolítico superior de Vale Boi (Vila do Bispo - Algarve). Master thesis in Archaeology In: Faculdade de Ciências Humanas e Sociais, Universidade do Algarve. Reimer, P., Bard, E., Bayliss, A., Beck, J., Blackwell, P., Ramsey, C., Van der Plicht, J., et al., 2013. IntCal13 and Marine13 Radiocarbon Age Calibration Curves 0–50,000 Years cal BP. Radiocarbon 55 (4), 1869–1887. https://doi.org/10.2458/azu_js_rc.55. 16947. Roselló, E., Morales, A., 2010. Lapa dos Coelhos: informe sobre los restos de peces. In: Mata, Almonte, Coord, E. (Eds.), – Cuaternario y Arqueologia: homenaje a Francisco Giles Pacheco, Cádiz, Asociación Profesional del Patrimonio Histórico-Arqueológico de Cádiz (ASPHA). /Servicio de Publicaciones de la Diputación Provincial de Cádiz, pp. 159–167. Sanches, M. de J., Teixeira, J.C., 2014. O abrigo do Passadeiro, Palaçoulo (Miranda do Douro). Um caso de estudo de gravuras rupestres dos inícios do Holocénico no Nordeste de Portugal. 35. Porto, Portvgalia, pp. 61–75 FLUP. Santos, A.T., 2017. A arte paleolítica ao ar livre da bacia do Douro à margem direita do Tejo: C. Gameiro, et al. Quaternary International xxx (xxxx) xxx–xxx 24 uma visão de conjunto, PhD thesis in Archaeology. Faculdade de Letras, Universidade do Porto. Santos, A.T., Barbosa, A.F., Aubry, T., García Díez, M., Sampaio, J.D., 2018. O final do ciclo gráfico paleolítico do Vale do Côa: a arte móvel do Fariseu (Muxagata, Vila Nova de Foz Côa). Portvgalia 39, 3–92. Santos, L., Goñi, M., 2003. Lateglacial and Holocene environmental changes in Portuguese coastal lagoons 3: vegetation history of the Santo Andre coastal area. Holocene 13 (3), 459–464. Sellami, F., 2009. Les données de la séquence stratigraphique du site de Fariseu: processus de déposition et d’érosion des dépôts en limite de la plaine alluviale de la Vallée du Côa. In: In: Aubry, T. (Ed.), - 200 séculos da História do Vale do Côa: incursões na vida quotidiana dos caçadores-artistas do Paleolítico, vol. 52. IGESPAR (Trabalhos de Arqueologia, Lisboa, pp. 103–108. Silva, C.M., 2003. O Olival Fechado no contexto peninsular. Master Thesis in Archaeology. Faculdade de Letras, Universidade de Lisboa. Silva, A.F., Rebelo, J.A., Ribeiro, M.L., 1989. Notícia explicativa da Folha 11C e Torre de Moncorvo. Serviços Geológicos de Portugal, Lisboa. Soares, J., Silva, C., 2004. Alterações ambientais e povoamento na transição MesolíticoNeolítico na Costa Sudoeste. In: Tavares, A., Tavares, M., Cardoso, J. (Eds.), Actas do Congresso Evolução Geohistórica do Litoral Português e Fenómenos Correlativos: Geologia, História, Arqueologia e Climatologia. Universidade Aberta, Lisboa, pp. 397–423. Soares, J., Silva, C., 2018. Living in the southwest Portuguese coast during the late mesolithic: the case study of Vale marim I. J. Archaeol. Sci.: Rep. 18, 1011–1025. Sonneville-Bordes, D., 1960. Le Paléolithique supérieur en Périgord. Delmas, Bordeaux. Sousa, A.C., Gonçalves, V., 2015. Fire walk with me. O sítio de Cova da Baleia e as primeiras arquitecturas domésticas de terra no Centro e Sul de Portugal. In: Gonçalves, V., Diniz, M., Sousa, A.C. (Eds.), 5.° Congresso do Neolítico Peninsular. Actas. UNIARQ, Lisboa. Stiner, M., Bicho, N., Lindly, J., Ferring, C.R., 2003. Mesolithic to Neolithic transitions: new results from shell-middens in the western Algarve, Portugal. Antiquity 77 (295), 75–86. Teixeira, J.C., 2016. O abrigo de Parada, um sítio de arte rupestre do Vale do Sabor (Alfândega da Fé, Bragança, Trás-os-Montes). In: In: Sanchez, M. de J., Cruz, D. J. da (Eds.), Atas da II a Mesa Redonda "Artes Rupestres da Pré-história à Proto-história". Estudo, conservação e Musealização de maciços rochosos e monumentos funerários (Porto, Faculdade de Letras, 10, 11 e 12 de Novembro de 2011), vol. 18. CEPBA [Estudos Pré-históricos, Viseu, pp. 41–70. Valladas, H., Mercier, N., Froget, L., Jorons, J.-L., Reyss, J.-L., Aubry, T., 2001. TL dating of upper palaeolithic sites in the coa valley (Portugal). Quat. Sci. Rev. 20, 939–943. Valente, M.J., 2008. As últimas sociedades de caçadores-recolectores no Centro e Sul de Portugal (10.000 - 6.000 anos BP): aproveitamento dos recursos animais. unpublished Phd thesis. Faro: Universidade do Algarve. Valente, M.J., 2010. O Barranco das Quebradas (Vila do Bispo) no contexto dos concheiros mesolíticos no Sudoeste português. 10. XELB, Silves, pp. 15–38. Valente, M.J., 2013. Arqueozoologia da Barca do Xerez de Baixo. In: Araújo, A.C., Almeida, F. (Eds.), Barca do Xerez de Baixo: um testemunho invulgar das últimas comunidades de caçadores-recolectores do Alentejo interior. Edia, Beja. Valente, M.J., Carvalho, A.F., 2009. Recent developments in Early Holocene huntergatherer subsistence and settlement: a view from south-western Iberia. In: In: McCartan, S., Schulting, R., Warren, G., Woodman, P. (Eds.), Mesolithic Horizons: Papers Presented at the Seventh International Conference on the Mesolithic in Europe, Belfast 2005, vol. 2. Oxbow Books, Oxford, pp. 312–317. Villaverde Bonilla, V., 1994. Arte paleolítico de la Cova del Parpalló. Estudio de la colección de plaquetas y cantos con grabados y pinturas. Servei d'Investigació Prehistòrica, Valencia. van der Knaap, W., van Leeuwen, J., 1997. Late Glacial and early Holocene vegetation succession, altitudinal vegetation zonation, and climatic change in the Serra da Estrela, Portugal. Rev. Palaeobot . Palynol. 97, 239–285. Zilhão, J., 1997a. O Paleolítico Superior da Estremadura Portuguesa. Colibri, Lisboa. Zilhão, J., 1997b. The Paleolithic settlement of Portuguese Estremadura after the last glacial maximum. In: Fullola, J., Soler, N. (Eds.), El món mediterrani desprès del pleniglacial (18 000-12 000). Museu d’Arqueologia de Catalunya, Girona, pp. 233–242. Zilhão, J., Marks, A., Ferring, C., Bicho, N., Figueiral, I., 1995. The Upper Paleolithic of the Rio Maior basin (Portugal). Preliminary results of a 1987-1993 PortugueseAmerican research project. Trab. Antropol. Etnol. 35 (4), 69–82. C. Gameiro, et al. Quaternary International xxx (xxxx) xxx–xxx 25