ROBERT SALA RAMOS (EDITOR) EUDALD CARBONELL JOSÉ MARÍA BERMÚDEZ DE CASTRO JUAN LUIS ARSUAGA (COORDINATORS) PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD BURGOS, 2014
ROBERT SALA RAMOS  EDITOR   EUDALD CARBONELL JOS   MAR  A BERM  DEZ DE CASTRO JUAN LUIS ARSUAGA  COORDINATORS   PLEISTOCEN...
Assistant Editors: María Gema Chacón Navarro Marcos Terradillos Bernal Amèlia Bargalló Ferrerons Cristina Vega Maeso Any reproduction, distribution, public communication or transformation of this book can only be done with the permission of its authors, with the exceptions permitted by law. Please contact CEDRO (Centro Español de Derechos Reprográficos, www.cedro.org) if you need to photocopy or scan any part of this book. Published thanks to Junta de Castilla y León through Fundación Siglo para las Artes y el Turismo de Castilla y León. Photos from the covert: 1. Handaxe from Galería, 2. Skull 5 from Sima de los Huesos, 3. and 4. Detail of Gran Dolina TD10-1, 5. Jaw of ursus dolinensis from Gran Dolina TD5, 6. Cutmarks from Gran Dolina, 7. Point from Sima del Elefante, upper levels, 8. Aerial view from Trinchera del Ferrocarril, and 9. Laboratory of microfauna, Arlanzón river (photos: IPHES) © Texts: The authors and the Fundación Atapuerca © Images: The creators and the Fundación Atapuerca © Current Edition: The Fundación Atapuerca and the Universidad de Burgos Publisher: UNIVERSIDAD DE BURGOS SERVICIO DE PUBLICACIONES E IMAGEN INSTITUCIONAL Edificio de Administración y Servicios C/ Don Juan de Austria, nº 1 09001 BURGOS – SPAIN FUNDACIÓN ATAPUERCA Carretera de Logroño nº 44 09198 Ibeas de Juarros (Burgos). ISBN: 978-84-92681-87-7 (Printed Edition) 978-84-92681-88-4 (e-book) Legal Deposit: BU-206. – 2014 Photocomposition: Rico Adrados, S.L. (Burgos) Print: Rico Adrados, S.L. (Burgos)
Assistant Editors  Mar  a Gema Chac  n Navarro Marcos Terradillos Bernal Am  lia Bargall   Ferrerons Cristina Vega Maeso  ...
5 CONTENTS Contents PREFACE ............................................................................................................................................ 9 MAJOR PHYSIOGRAPHIC, GEOGRAPHIC AND ECOLOGICAL REGIONS 15 18 Monforte de Lemos ensemble ........................................................................................................ 26 Valdavara ...................................................................................................................................... 31 Cantabrian mountains and coastline ........................................................................................... 35 Cueva de Aitzbitarte III and IV ..................................................................................................... 37 Antoliñako Koba ........................................................................................................................... 41 Axlor ............................................................................................................................................ 45 Los Azules .................................................................................................................................... 49 Cabo Busto ................................................................................................................................... 52 El Castillo ..................................................................................................................................... 55 Cueva de la Güelga ....................................................................................................................... 60 Cueva de Las Caldas ..................................................................................................................... 64 Cueva Morín ................................................................................................................................. 72 Ekain ............................................................................................................................................. 78 Cueva del Esquilleu ...................................................................................................................... 82 El Pendo ....................................................................................................................................... 87 Cueva de la Riera .......................................................................................................................... 92 La Viña ......................................................................................................................................... 95 Labeko Koba ................................................................................................................................. 99 Lezetxiki ....................................................................................................................................... 105 Cueva del Mirón ........................................................................................................................... 110 Santimamiñe ................................................................................................................................. 114 El Sidrón ....................................................................................................................................... 122 Sopeña .......................................................................................................................................... 3. 13 Cova Eirós .................................................................................................................................... 2. North-western Atlantic basins ...................................................................................................... As Gándaras de Budiño ................................................................................................................. 1. 129 Ebro Valley, Pyrenees and Pre-Pyrenees ....................................................................................... 133 Abauntz ........................................................................................................................................ 135
5  CONTENTS  Contents  PREFACE ..............................................................................................
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 6 Arrillor .......................................................................................................................................... 141 Atxoste ......................................................................................................................................... 148 Cova del Parco .............................................................................................................................. 152 Roca dels Bous .............................................................................................................................. 159 Cova Gran de Santa Linya ............................................................................................................ 162 Cova de l´Estret de Tragó .............................................................................................................. 167 Fuente del Trucho ......................................................................................................................... 171 Fuentes de San Cristóbal ............................................................................................................... 179 Gabasa .......................................................................................................................................... 181 Kanpanoste Goikoa ....................................................................................................................... 188 Forcas ............................................................................................................................................ 192 Nerets, Cova de les Llenes ............................................................................................................. 196 Martinarri ..................................................................................................................................... 200 Mendandia .................................................................................................................................... 208 Mediterranean basins. North of the Ebro River. ........................................................................... 219 Abric Romaní ............................................................................................................................... 4. 204 Montsant valley ensemble ............................................................................................................. 221 La Cansaladeta .............................................................................................................................. 236 Cinglera del Capelló ..................................................................................................................... 238 Reclau Viver ensemble .................................................................................................................. 246 St. Julià de Ramis Pleistocene ensemble ........................................................................................ 256 Middle Pleistocene ensemble Montgrí, La Selva and Puig d’en Roca: Puig d´en Roca, Cau del Duc de Torroella de Montgrí, La Selva, Cau del Duc d´Ullà, Can Garriga, Pedra Dreta, Can Rubau and La Jueria ..................................................................................................................... 260 Cova de l’Arbreda ......................................................................................................................... 266 Cova del Gegant ........................................................................................................................... 276 Cova del Rinoceront ..................................................................................................................... 281 Els Vinyets .................................................................................................................................... 284 Barranc de la Boella ....................................................................................................................... 287 Molí del Salt ................................................................................................................................. 295 Cova de les Teixoneres, cova del Toll ............................................................................................. 302 Vallparadís .................................................................................................................................... 5. 232 La Cativera ................................................................................................................................... 308 Mediterranean basins. Centre ....................................................................................................... 317 Abric de El Pastor ......................................................................................................................... 319 Cova del Bolomor ......................................................................................................................... 323 Casa Corona ................................................................................................................................. 331 El Collado ..................................................................................................................................... 338 Cova Beneito ................................................................................................................................ 345 Coves Santa Maira ........................................................................................................................ 353 Cova Foradà .................................................................................................................................. 356
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  6  Arrill...
CONTENTS Cova Negra ................................................................................................................................... Cueva de la Cocina ....................................................................................................................... 370 La Cueva Negra del Estrecho del Río Quípar ............................................................................... 372 El Salt ........................................................................................................................................... 380 Cova de les Cendres ...................................................................................................................... 388 Cova de les Malladetes .................................................................................................................. 395 Cova Matutano ............................................................................................................................. 399 Cova del Parpalló .......................................................................................................................... 402 La Sima de las Palomas del Cabezo Gordo .................................................................................... 410 Tossal de la Font: La Cova de Dalt, La Cova de Baix ..................................................................... 413 Tossal de la Roca ........................................................................................................................... 417 Southern Mediterranean coast, Guadalquivir River and Betic intramontane basins .................. 421 El Aculadero ................................................................................................................................. 423 Ardales .......................................................................................................................................... 426 Bajondillo ...................................................................................................................................... 430 Cueva de Ambrosio ....................................................................................................................... 435 Cueva de Nerja ............................................................................................................................. 442 Cueva and abrigo del Ángel .......................................................................................................... 451 Cueva del Boquete de Zafarraya ................................................................................................... 6. 361 463 Guadix-Baza basin and Orce ensemble: Orce: Venta Micena, Barrranco Leon, Fuente Nueva 3, Huescar 1, Cullar-Baza-1, Solana de Zamborino, Cueva Hora ....................................................... Las Grajas de Archidona ............................................................................................................... 497 Strait of Gibraltar ......................................................................................................................... 501 Abrigo de Benzú ........................................................................................................................... 503 Gibraltar: Gorham and Vanguard caves ......................................................................................... 8. 494 El Pirulejo ...................................................................................................................................... 7. 474 506 Central plateau ............................................................................................................................. 515 Ambrona and Torralba .................................................................................................................. 517 Cuesta de la Bajada ....................................................................................................................... 528 Jarama VI ...................................................................................................................................... 531 Atapuerca ensemble: Gran Dolina, Galería, Sima del elefante, Sima de los Huesos, Portalón and Cueva de El Mirador ..................................................................................................................... 534 Manzanares and Jarama: San Isidro, Las Delicias, Orcasitas, Transfesa, Prepesa, Áridos, Valdocarros, El Cañaveral, Los Ahijones and Los Berrocales .............................................................................. 561 La Peña de Estebanvela ................................................................................................................. 568 Pinedo ........................................................................................................................................... 574 Pinilla: Camino Cave, Navalmaíllo Rockshelter, Buena Pinta Cave, Ocelado Rockshelter, Descubierta Cave ................................................................................................................................ 577 San Quirce .................................................................................................................................... 584 Cueva de Maltravieso, Cueva de Santa Ana, Cueva de El Conejar, Vendimia and El Millar ........... 587 7
CONTENTS  Cova Negra ........................................................................................................
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 8 Siega Verde ................................................................................................................................... 601 Valdegoba ..................................................................................................................................... 608 Palaeolithic art in the north of Spain ............................................................................................ 611 Introduction .................................................................................................................................. 613 Altamira ........................................................................................................................................ 615 Cueva de Altxerri .......................................................................................................................... 628 Cueva de Ekain ............................................................................................................................. 630 Santimamiñe ................................................................................................................................. 633 Cueva de Covalanas ...................................................................................................................... 634 La Garma ...................................................................................................................................... 636 El Pendo ........................................................................................................................................ 644 El Castillo ..................................................................................................................................... 647 Las Chimeneas .............................................................................................................................. 651 La Pasiega ..................................................................................................................................... 653 Las Monedas ................................................................................................................................. 656 Hornos de la Peña ......................................................................................................................... 658 Chufín .......................................................................................................................................... 659 El Pindal ....................................................................................................................................... 661 Llonín ........................................................................................................................................... 663 La Covaciella ................................................................................................................................ 666 Tito Bustillo .................................................................................................................................. 667 La Lluera ...................................................................................................................................... 672 Cueva de la Peña de Candamo ...................................................................................................... 674 10. Post-Pleistocene art from the Iberian Levant ................................................................................ 679 Roca dels Moros de El Cogull, Abrigo de Perellada IV, Abrigo de Cabra Feixeta, Abrigos d’Ermites en la Serra de la Pietat, Conjunto d’Ermites, Cova Centelles, Cova del Puntal, Cingle de la Mola Remigia, Cova Remigia, Racó Gasparo, Racó Molero, Abrigos de la Joquera, La Saltadora, Abrigo d’en Melia, Abrigo del Cingle del Barran de l’Espigolar, Barranco de la Valltorta, La Sarga, Abrigo del Lucio, Barranco Moreno, Cueva de la Araña, El Abrigo de la Pareja, Cañaica del Calar, Fuente del Sabuco, La Risca, El Milano, Cantos de la Visera en Monte Arabí, Cueva de los Grajos, Estrechos de Santoge, Lavaderos de Tello, Peñón de la Tabla del Pochico, Prado del Azogue, Cueva de la Vieja, Minateda, Abrigo de los toros de las Bojadillas, Solana de las Covachas, Abrigo de los Oculados, Cueva del Tío Modesto, Peña del Escrito, Selva Pascuala, Abrigo de Selva Pascuala, Marmalo, Los Trepadores, Tia Mona, Tio Garroso, Los Chaparros, Cocinilla del Obispo, Doña Clotilde, Prado del Navazo, Toros de la Losilla, Cañada de Marco, Val del Cahrco del Agua Amarga, Barranco Hondo, Roca dels Moros de Calapatá, Calapatá, Chimiachas, Piezarrodilla, Cerrao I and I I, Cabras blancas, Prado de las Olivanas, Muriecho, Fariceu ......................................................................... 681 9.
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  8  Siega ...
preface
preface
PREFACE Our group has decided to update the information available about Spanish palaeohistory for the UISPP congress. We thought that this international congress would be an ideal occasion to collate and publish two volumes on our remote history. The first volume looks at the first palaeoccupations in the prehistoric record of the Iberian Peninsula. As you know, prehistory research is making spectacular progress thanks to fieldwork, lab work and also teamwork undertaken by archaeologists, biologists, geologists and botanists. New theoretical proposals by experts in life, earth and social sciences, as well as modern methodologies and applied technologies are all helping to shape new insights into our history, and thus build on our previous knowledge and interpretations. Yet none of this can be done without hard data. Despite this great panoply of disciplines, it would be impossible to provide a consistent explanation of the eco-social process of human evolution without archaeological excavations. In this volume, we wish to highlight above all the empirical data that allow we archaeologists and prehistorians to advance “a posteriori” interpretations of the structures and systems of past life. This realisation of the need to update the information widely available about fieldwork is what has led us to edit this volume about prehistory on the Iberian Peninsula, particularly in Spain. We want to present historians –and also history buffs and all other human specimens with a thirst for knowledge– with the facts that allow us to concoct the outlines needed to apply reverse engineering to reconstruct the palaeoecological evolution of our genus in a particular territorial context. We want to present information about the evolution of the Homo genus and its environment, which means that both the natural environment and the historic environment are the focus of this work. An analysis of the current state of fieldwork and initiatives at prehistoric sites makes an indis- pensable contribution to a coherent understanding of the historic sequence of a territory. It also allows us to bring this type of scientific inquiry up to date, and places us in a better position to understand the nature of the species through the records of the past. In recent times, the massive volume of results studied and published in specialist journals has made it necessary to draft sequential summaries in which all the data from all the research is accessible in a single monographic volume. This facilitates the necessary socialization of the work of specialists in different periods of history. Direct access to the current state of prehistoric archaeology –either in print or digital format– improves our ability to plan and discuss the past, the present and to possess a grounding to build the future. The cascade of empirical data contained in the collective effort of this first volume is, in our opinion, of incalculable value: it is the result of the work of the majority of the Spanish teams who are working in the field and have helped to make the knowledge of this country’s prehistory one of the most thorough and up to date in the world. Bringing information about prehistoric archaeological work up to date in an orderly, concise manner in order to avoid an intractable volume has led us to the concept and structure of this book. We have arranged the sites on the basis of a regional grid, which has allowed us to organize and synthesize our discourse, and also avoid having to repeat maps and other graphic material which would make the text less palatable. This is a reference work in which the reader can find essential information about the archaeological sites which the editors believe are the most significant for our prehistory. We apologize if this updated summary does not include every site currently in the process of investigation and publication. Our intention has been to produce a significant, 11
PREFACE  Our group has decided to update the information available about Spanish palaeohistory for the UISPP congress. We ...
12 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD up to date synthesis –necessarily abbreviated– of the current state of prehistory on the Iberian Peninsula. We do not know if this has been achieved, but whatever the case, we still think the effort was necessary and our commitment has been to coordinate this work. The articles on the sites, all of them synthetic and systematized, open a window onto the empirical reality of our archaeological knowledge and hence the true state of our knowledge, avoiding interpretations on the basis of data and knowledge that has not been published and well checked. This volume is thus an effort that we wish to share with all those who are interested in the study and the analysis of the past. We also hope it will serve us to update our own work and make it more visible in other areas outside the strict realm of prehistory. The transformation of the social sciences into the sciences of socialized knowledge is an evolutionary perspective of the way we understand the world. Intervening in and influencing the state of knowledge about the history of populations is a strategic decision that has shaped the powerfully empirical nature of this proposal. When another synthesis is written, the information will probably have changed, but the underlying essence –our interest in sharing what we know with our species– will remain. This has been our intention. The reader has the last word. EUDALD CARBONELL Vice-president of Fundación Atapuerca Director of Institut Català de Paleoecologia Humana i Evolució Social
12  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  up to...
1 north-western atlantic basins NORTH-WESTERN ATLANTIC BASINS. 13
1  north-western atlantic basins  NORTH-WESTERN ATLANTIC BASINS.  13
Site Map numbering As Gándaras de Budiño 1 Cova Eirós 2 Monforte de Lemos ensemble 3 Valdara 4
Site  Map numbering  As G  ndaras de Budi  o  1  Cova Eir  s  2  Monforte de Lemos ensemble  3  Valdara  4
NORTH-WESTERN ATLANTIC BASINS. Arturo de Lombera-Hermida *,**,***, Xosé Pedro Rodríguez *,** The Paleolithic site of As Gándaras de Budiño (O Porriño, Pontevedra) The Paleolithic site of As Gándaras de Budiño is located in O Porriño industrial park (Pontevedra) (UTMX: 531.407; UTMY: 4.661.631). It is situated in the valley of the river Louro, which runs through the southern sector of the Galician Meridiana Depression, delimited on each side by the granite heights of Faro de Budiño to the east and Serra do Galiñeiro to the west. Discovered in 1961 by Henri Nonn, in 1963 excavation and survey work began under the guidance of Dr. Emiliano Aguirre (1964). At the same time as the excavation work was being carried out, a geological and stratigraphic study of the site and of several coastal and continental deposits was conducted in order to correlate the geomorphology of Budiño with its regional context (Butzer, 1967). The archaeological excavations proved the existence of large accumulations of material, identifying workshops areas and the presence of hearths. According to D. Echaide, the lithic assemblages would correspond to the Acheulean. Due to the methodological approach used in the excavations, As Gándaras de Budiño becomes the emblematic site of Paleolithic research in Galicia. However, the association of the Acheulean lithic assemblages with the surprising carbon dating obtained from two carbon samples, within the Upper Pleistocene and a priori confirmed through stratigraphic correlation of the levels of As Gándaras with the deposits from the southeast coast, would make it one of the most controversial sites (Aguirre and Butzer, 1967). Given this problem, in 1979 Julio M. Vidal Encinas resumed excavation between 1980 and 1982, opening several trenches near the area ex* cavated by Emiliano Aguirre and recovering many materials associated with fluvial and colluvia formations, similar to those identified previously (Vidal, 1982). Finally, in 1991 rescue archaeological excavations were carried out under the guidance of Dolores Cerqueiro Landín. Given the problematic and controversial chronological interpretation of this site, many review papers have dealt with the stratigraphy and geomorphological context of As Gándaras de Budiño (Ramil et al., 1993; Gracia et al., 2004), and of its lithic assemblages (de Lombera et al., 2011; Méndez, 2007; 2008). Synthesis papers based on the stratigraphic descriptions permit the different levels to be correlated and the sequence of the depositional processes observed at the As Gándaras de Budiño site to be reconstructed (Ramil et al., 1993). Initially, the Louro basin was filled by various levels of arkoses and clays during the Upper Miocene. Then, the basin was captured by the dynamic of the Miño in the Pleistocene, forming the fluvial level sequence (T+67m; T+56m; T+33m; T+24m) and small lateral channels associated with the Louro valley, level T+24m being the level related to the site (fluvial levels). Following an episode of microfracturing and dismantling of the fluvial deposits and clay baselevels, the lower colluvium deposited, formed by two units (levels 6a and 6b). The lithic industries classified as Acheulean are located in this deposit. Following a second episode of erosion, dismantling and dispersion of the lower colluvium, sedimentation of the upper colluvium occurs, now over the underlying colluvium, over the source fluvial levels, over tertiary clay mate- IPHES, Institut Català de Paleoecologia Humana i Evolució Social, C/ Marcel.lí Domingo s/n. Campus Sescelades, (Edifici W3), Tarragona 43007 ** Area de Prehistoria, Universitat Rovira i Virgili (URV), Av. Catalunya 35, Tarragona 43002 *** Grupo de Estudos para a Prehistoria do Noroeste (GEPN). Dpto. de Historia I, Universidade de Santiago de Compostela. Praza da Universidade 1, Santiago de Compostela 15782 15
NORTH-WESTERN ATLANTIC BASINS.  Arturo de Lombera-Hermida  ,  ,   , Xos   Pedro Rodr  guez  ,    The Paleolithic site of A...
16 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. Stratigraphy of the As Gándaras de Budiño site and possible stratigraphic correlations between the different levels. rials. This contains some industries as a result of disassembly of the lower levels (vg. Level 2 of Locus I). Finally different paedogenesis phenomena occur on the upper colluvium levels (Fig. 1). The presence of lithic assemblages is constant throughout the sequence. The uniformity of the technology of the assemblages recovered from the different levels and trenches was identified from the very start. Adhering to the morphogenetic processes of the site, the origin of the lithic materials must be related to Acheulean settlements only, which are identified in the fluvial sequences associated to T+24m. However, the materials associated to the upper colluvium showed certain volumetric and technological differences (lower incidence of Large Cutting Tools (LCT), smaller format) that, added to its stratigraphic position, led some authors to consider them as evidences of later settlements, initially linked to the Mousterian (vg. Level 2 of Locus I) (Vidal, 1982). Taphonomic reviews of colluvium levels show the existence of volumetric selection and mechanical alteration from rolling of the artefacts in the assemblages. At the same time, these assemblages show clear convergence at a technological level with those identified at the fluvial levels (de Lombera et al., 2011; Méndez, 2008). For this reason, these small variations between the different sites must be understood as a consequence of the material resedimentation processes. The lithic industry is defined by usage of local and nearby resources, the proportion of quartzite and quartz varying between the different sites (63%26% and 73.93%-35.5%, respectively). Rock crystal appears to a lesser extent (3.2%-0.7%), related to quartz formations. Quartz is an abundant resource in the Louro basin, the source of good quality quartzite, predominant in the assemblages, is located on the terraces of the river Miño, 3 km to the south. The assemblages are characterised by the dominance of knapping products, generally small and medium sized (30-60 mm), particularly the quartz items. Regarding the knapping strategies, the more expeditious types that performance longitudinal and bifacial reduction series (especially in quartz) predominate but with a significant presence of centripetal and discoidal cores and products, whereas Levallois products are barely represented (around 2%). The opportunistic resource of bipolar knapping on quartz is also documented, witnessed through various knapping products, as well as large blocks interpreted as anvils. The presence of handaxes and cleavers (Fig. 2), although consistent, varies between the different sites. The majority of the handaxes are made on flake with amygdaloidal morphologies, evenly edges and silhouettes trimming, mainly using a hammerstone. There is a smaller representation of cleavers and picks. Regarding light duty tools, denticulate, scrapers and notched tools stand out.
16  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figur...
NORTH-WESTERN ATLANTIC BASINS. Figure 2. Handaxes and cleavers recovered from the excavations of Emiliano Aguirre (Fotograph: Museo Municipal de Vigo. Quiñones de León). These assemblages show a clear difference as to the management of raw materials based on their knapping quality, quartzite being used mainly for making LCT, medium and large implements, and in more complex knapping methods. Quartz, however, becomes more important for small blanks and tools. Although for the quartz tools all the phases of the chaîne opératorie seem to be represented, this is fragmented for the quartzite tools, basically represented by flakes, cores in an advanced stage of reduction and LCT. Conclusions The technical characteristics of the site of As Gándaras de Budiño place it within the peninsular Acheulean assemblages from the second half of the Middle Pleistocene. In this respect, the absolute dating obtained show the different morphogenetic processes of the Louro basin but not the age of the industries. Its location in the Meridiana Depression gives it a strategic settlement within the mobility of the Atlantic Façade, that also explain the large number of findings and their wide distribution in the area. 17
NORTH-WESTERN ATLANTIC BASINS.  Figure 2. Handaxes and cleavers recovered from the excavations of Emiliano Aguirre  Fotogr...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 18 Arturo de Lombera-Hermida *, **, ***, Xosé Pedro Rodríguez *, **, Ramón Fábregas-Valcarce*** Cova Eirós archaeo-palaeontological site, Triacastela, Lugo 1. Cova Eirós site. Location and background The few limestone formations in north-western Spain and the primary focus of research on river terraces and rock shelters have resulted in a small number of documented cave sites in this area, amongst which Cova Eirós (Eirós Cave) ranks amongst the most important in palaeontological and archaeological terms. The Cova Eirós site is located in a village, Cancelo, in the Triacastela municipality (Lugo) (UTM X: 646.855; UTM Y: 4.736.428). The cave entrance is on the NNW slope of Monte Penedo in Serra do Ouribio, 780 metres asl and 25 metres above a stream, Arroyo de Bezcos. The Cova Eirós karst system is part of the Cándana limestone Series, formed during the Lower Cambrian. The cave is 104 metres long, with a mouth that currently measures 2 metres high by 3.5 metres wide. The entrance narrows after the first 7 metres into a 15 metre long neck, followed by the cave’s largest space, the “Main –or Mammoth– Hall”, 15 metres long, 6 metres wide at the most and up to about 5 metres high. The cave then continues inwards in a NNW direction in a series of three overlapping levels of galleries (Grandal, 1993) which are almost entirely clogged by clayey sediment interspersed with stalagmitic crusts or floors. From the outset Cova Eirós became renowned as an outstanding palaeontological site due to the presence of bear remains. In the late 1980s, UDC conducted several digs in the middle and end sections of the cave (1988, 1989 and 1991), which recovered approx. 4,000 bones from at least 43 bears (Grandal, 1993). Based on these findings, Cova Eirós became one of the most important sites on the Iberian Peninsula for Ursus spelaeus remains. * AMS 14C analysis dated a bear bone at 24,090 ± 440 BP (Grandal and Vidal, 1997), consistent with the dating of a stalagmitic crust below the fertile level in the final gallery of the cave, (25,233 ± 5,027 BP) (U series) (Grandal, 1993). The most recent datings on bear bones extend the time range to a period between 24,000 and 31,000 BP (Pérez et al., 2011). In addition, two datings of the stalagmite crusts in the central corridor and the final gallery have yielded 117,252 ± 75,438 BP and 97,051 ± 15,426 BP respectively, although caution is advised due to their wide standard deviation (Grandal and Vidal, 1997). In 1993, the first archaeological dig in Cova Eirós was a 1 m2 test pit at the cave entrance, in which five archaeological levels were identified and 550 lithic tools were recovered, all initially attributed to the Middle and Upper Palaeolithic. Following these reports, in 2008 a new phase in the exploration of the Cova Eirós occupations was begun (Fábregas et al., 2009). Initially, two test pits were dug at the mouth: Pit A at the entrance (4 m2), which included the 1993 pit in one of its survey squares, and a second pit on the outer embankment (Cata B). The quantity and quality of the recovered material and the stratigraphic potential of the sediment led to the extension of the excavation area to nearly half the inhabitable space of the entrance sector (21 m2) (Fig. 1). 2. The Pleistocene sequence in Cova Eirós The 16 metre long entrance sector decreases in height to 0.8 m at the neck that leads to the interior part of the complex. A steep embankment coincides with the vertex of the mouth. The in- IPHES, Institut Català de Paleoecologia Humana i Evolució Social, C/ Marcel.lí Domingo s/n. Campus Sescelades, (Edifici W3), Tarragona 43007 ** Area de Prehistoria, Universitat Rovira i Virgili (URV), Av. Catalunya 35, Tarragona 43002 *** Grupo de Estudos para a Prehistoria do Noroeste (GEPN). Dpto. de Historia I, Universidade de Santiago de Compostela. Praza da Universidade 1, Santiago de Compostela 15782
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  18  Artur...
NORTH-WESTERN ATLANTIC BASINS. Figure 1. Location of the Cova Eirós site. Plan of the cave and the current archaeological dig with a topographic reconstruction of the entrance sector (2009). 1) Entrance sector 2) Great Hall 3) Final Gallery, Palaeontological interventions. habitable space is thus reduced to this vertical between the ledge and the interior neck. On the west wall, however, eroded remains of a cave wall mark the former entrance area. Geophysical surveys using Georadar in this sector have detected a 3.4 m deep stratigraphic infill. Excavation has only reached the –120/140 cm level below the present floor, and occupations prior to the currently identified may thus be expected to appear in the future. Six archaeological levels have been identified in the stratigraphic sequence, with two types of sedimentary dynamics. The lower levels are associated with in situ sedimentation, basically linked to gravitational input (Levels 4, 3 and 2), while the upper levels are from wind-borne matter (Levels 1A and 1B). Levels B and C levels are linked to the dynamics underway outside the cave (the slope) (Rodríguez et al., 2011) (Fig. 2). The stratigraphic sequence described for the site is the following: –Surface Level. Topsoil. Very loose and heavily bioturbated organic soil. Limestone clasts produced by rockfalls from the ceiling. First indication of disturbed archaeological material. This level also contains a perimeter stone paving which delimited two medieval silos. Potential: 10-15 cm. –Level B: Whitish-yellow silty matrix, well compacted, containing angular limestone gravel measuring 5-13 mm along the major axis. Abrupt, irregular upper limit and gradual, diffuse lower limit. A 30 mm deep layer of very dark sediment has been identified at the bottom of this level. A considerable amount of archaeological material has been retrieved from Archaeological Level B. The upper limit of this layer, defined by limestone blocks measuring 40-140 mm on the longest axis, has a predominantly south slope. It might be related to one of the most recent processes in the reduction of the cave cornice. AMS radiocarbon dating of the archaeological level is 12,060 ± 50 BP (Beta – 308859) (Fábregas et al., 2012). –Level C: A heavily compacted clay matrix with gravel and clasts. A steep south-facing slope follows the topography of the embankment, and intersecting levels 1, 2, 3 and 4. This is interpreted as a disturbance of the levels inside the cave. –Level 1: The upper part consists of slightly compacted fine yellow sand with a subhorizontal 19
NORTH-WESTERN ATLANTIC BASINS.  Figure 1. Location of the Cova Eir  s site. Plan of the cave and the current archaeologica...
20 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 2. Stratigraphic profile West of the dig in Cova Eirós laminar stratification (1A). The top is archaeologically sterile. At the bottom, the sand becomes orange, more compacted and a parallel horizontal stratification (1B). The top and bottom limits of the level are net. Depth: 10 to 23 cm. –Level 2: Orange clay-loam, more compact than the previous level. Two distinct facies: at the top (level 2a), purer sands, without concretions, while the lower limit is marked by the presence of small limestone plates. At the base (sub-level 2b) several small subangular limestone clasts (3-5 cm), heavily carbonated, with a crust in the southern sector. Net ondulating base level. Archaeological level. 34-45 cm deep. C14 AMS radiocarbon dating: 31,690 ± 240 BP (Beta – 254280) –Level 3: Brown clay with small-sized limestone gravel. More compact and homogeneous than the overlying level. Contact with level 2 is net and undulating, indicating the existence of a gap between the two levels. Archaeological level. 20-35 cm deep. OSL dating: 84,807 ± 4919 BP (MAD-5612BIN). –Level 4: Composed of clays and fine yellow sands, heavily carbonated with sub-rounded limestone and schist gravels. Minor lateral variations. Archaeological level: currently under excavation. Unknown depth. 3. Middle Palaeolithic occupations Two levels attributed to the Middle Palaeolithic, the most substantial records in the study of this period in Galicia, have been identified in the lower part of the sequence. Level 4 Level 4 has a high density of archaeological material, currently the richest of the sequence. The excavation and analysis of the data is still underway, however preliminary results point to certain aspects. As in most of the identified occupations in Cova Eirós, the lithic assemblage is dominated by quartz (90.1%), with quartzite playing a secondary role (9.7%). Knapping products are over-represented, with few cores and retouched tools on flake (sidescrapers, denticulates, etc.) which, in conjunction with the scarcity of corticality in the products, indicates that this assemblage represents the final stages of reduction. From the technological characteristics of the product, we can deduce that the predominant knapping methods are longitudinal and orthogonal strategies, although discoidal and Levallois quartz products have been identified amongst the higher quality items, particularly in the finegrained quartzite. The technological characteris-
20  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figur...
NORTH-WESTERN ATLANTIC BASINS. tics of the lithic assemblage on this level suggest that it is Mode 3. The faunal assemblage includes remains of bears, deer and horses. Several show evidence of human intervention (fracturing and cut marks) and thermal alteration. Several bone fragments have been attributed to Coelodonta sp. The most significant discovery on this level was 5 metres from the current entrance: a combustion structure with a small (40 x 25 cm) oval focal area (Fábregas et al., 2012). A large concentration of tools and skeletal remains, many of them with evidence of heat impact, was identified lying around this hearth. taxonomic definition. Some of the bones show clear cut marks and fractures, pointing to the anthropic factor as the main agent in butcheringrelated activities, skinning and access to bone marrow. Taxonomically, Cervus elaphus has the highest representation, followed by Rupicapra rupicapra and Capreolus capreolus. Carnivores include numerous Ursus spelaeus, with many neonatal tooth remains. The presence of Canis lupus and Vulpes vulpes is also documented, indicating the alternating use of the cave by hominids and carnivores. 4. Upper Palaeolithic sequence Level 3 Level 2 The initial excavation of this level was heavily affected by the presence of medieval silos, restricting its area to 1.5 m2. The size of the current work area permitted excavation in the entire zone to commence, allowing us to gather more information about the Neanderthal occupations in Cova Eirós. The level 2 lithic assemblage contains a high percentage of small-sized quartz industry (93.3%). Although the type of raw material and the large number of knapping fragments hinder an accurate techno-typological definition, certain changes in the supply strategies –with the appearance of small amounts of rock crystal and flint– and the presence of blades in rock crystal ascribe this assemblage to be attributed to the initial Upper Palaeolithic. C14 dating places these occupations in the Aurignacian. The detected lithic assemblage consists of 702 items. The predominant raw material is quartz (90%), followed by fine grain quartzite (8%). The identified reduction methods are Levallois, discoidal and orthogonal, with some evidence of Kombewa (Fábregas et al., 2009; Rodríguez et al., 2011). As in the previous level, the operational chains are fragmented, with a clear predominance of knapped products and few cores. The presence of retouched items –sidescrapers, denticulates and becs– is greater here (9.9% ). This aspect is more obvious in the fine-grained quartzite, an allochthonous lithic resource, represented by Levallois flakes and points, as well as sidescrapers bearing evidence of treatment as curated tools (Fig. 3). The characteristics of the lithic assemblage on Level 3 suggest its attribution to Mode 3, with a differential management of lithic resources found in north-western Iberia in the Middle Palaeolithic, defined particularly by the relationship of Levallois techniques to fine-grained quartzite and the use of more expedient methods with quartz, although quartz Levallois products have also been found. Functional analysis of this assemblage has identified hunting-related activities (broken spear tips), butchering and hide treatment at the site (Lazuén et al., 2011). The faunal assemblage is characterized by a high percentage of fragmentation, hindering its The faunal record includes Cervus elaphus and Capreolus capreolus and a remarkable representation of carnivores, with several remains of Canis lupus and Panthera pardus, as well as both bear species which hibernated in the cave (Ursus arctos and U. spelaeus). The faunal remains show a high level of fragmentation and some have obvious cut marks and fresh fracture. The low density of lithic tools, the identification of fragmented operational chains and the higher incidence of carnivore taxa point to occupations with little impact, perhaps related to short, repeated stays, alternated with the cave’s use as a den by carnivores and ursids. Level 1 The material is concentrated at the base of this stratum, where 729 archaeological items have been recovered (Fábregas et al., 2010; Rodríguez et al., 2011). The lithic assemblage is predominantly quartz (86.1%), while flint and rock crystal have a greater representation (4.8 % and 6.1 %, respectively), parallel to an increased blade and bladelet component. Two types of chaines opéra- 21
NORTH-WESTERN ATLANTIC BASINS.  tics of the lithic assemblage on this level suggest that it is Mode 3. The faunal assembla...
22 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 3. Levallois points (a) and flakes (b) in quartzite from the Middle Palaeolithic levels (Level 3 and 4). toires have been detected at this level, depending on the type of raw material and the purpose of the reduction. In the local quartz and quartzites, more expeditious strategies (longitudinal or centripetal) were used to obtain flakes, while in the case of rock crystal and flint, production focused on blade flakes, with microblade cores and bladelets in this material as well as a few backed items (Fig. 4). While the quartz prisms are of local origin, the flint items, only found in elements from the end of the knapping sequences, are from further afar, possibly linked to the above-mentioned outcrops 12-14 km to the NE. The technical nature of the assemblage and the datings currently available ascribe this level to the end of the Gravettian. spect to the previous levels, particularly ursids, a trend documented at other sites in Cantabria. One outstanding discovery on this level is a pendant made from a canine tooth of a small carnivore (Fábregas et al., 2010). The faunal assemblage consists of 135 items. Rupicapra rupicapra and Cervus elaphus still predominate amongst the herbivore species, but there is a significant decrease in the carnivores with re- Archaeological level B is linked to the final retreat of the cave ledge, defined by a line of limestone slabs and blocks, some over 30 cm in diameter. The palaeosol indicates an embankment The top of level 1 (1A) is a succession of small layers of fine and coarse sands of wind-borne origin. This level is sterile, indicating that the cave was probably not occupied by either humans or animals during the harshest periods of the Last Glacial Maximum, contemporary to the formation of nearby glaciers (O Queixadoiro). Level B
22  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figur...
NORTH-WESTERN ATLANTIC BASINS. dynamic on a steep slope towards the archaeological south. Moving inwards, this level merges with the current organic floor, and thus the preserved area is quite small (barely 3 m2). Radiocarbon dating places this event right at the end of the Magdalenian (Fábregas et al., 2010). The lithic assemblage shows a clear specialization, with a high percentage (40.6%) of rock crystal, second only to quartz. There are two different chaines opératoires in the reduction strategies: the production of flake items in quartz and quartzite with longitudinal, orthogonal and, to a lesser extent, bipolar techniques; and on the other hand, specialized production centred on the exploitation of rock crystal prisms for bladelets and backed elements. Given that all parts of the sequence are represented here (cores, volumetric adjustment elements, retouched products, etc.), it follows that these items were produced in situ. Faunal remains –including a deer antler– are quite scarce, possibly due to the more organic component of the sediment. Level C Level C is in the outer part of the embankment, with a steep south slope (lines 20-21). It contains a high density of lithic and bone material. These items show a mixture of features, evidence of the disturbance of this assemblage (Fábregas et al., 2009). Taphonomicaly, the bones are in different stages of fossilization and wear. Within the lithic assemblage, Levallois flakes in quartz and quartzite akin to those recovered from levels 4 and 3 have been identified, along with rock crystal and flint blades and bladelets and flakes of the type found on levels 1 and 2. This level is therefore considered to be the result of the dismantling and disturbance of the levels inside the cave at the top of the embankment due to erosion during the retreat of the cave ledge. 5. Recent Prehistory and Early Medieval occupations While most of the identified occupations at the site have been ascribed to various periods in the Middle and Upper Palaeolithic, there is also evidence of this cave’s use in subsequent periods. In recent prehistory, it was used as a burial site, judging by the Bell-beaker ceramic material recovered at the entrance (Fábregas et al., 2012), and the human remains found inside the cave (Corridor), dated at the height of the Bronze Age (Ua-38121, 3151 ±31 BP). These funerary activities can be contextualised in the late Bell-beaker horizons and the middle of the second millennium BC. Finally, the surface level contained several storage structures (silos, UA1 and UA2), a hearth (UA06) and a perimetral pavement related to agro-pastoral activities in Cova Eirós from the late 10th to the 15th century AD (Teira et al., 2012). Figure 4. Core and bladelets in rock crystal, and flint bladelet from Cova Eirós, levels 1 and B. 23
NORTH-WESTERN ATLANTIC BASINS.  dynamic on a steep slope towards the archaeological south. Moving inwards, this level merg...
24 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 6. Mobile and rock art The recurrent presence of Upper Palaeolithic communities in eastern Galicia, the discovery of mobile art and, most importantly, documentation of rock art in northern Portugal indicated that rock art could probably be found in Galicia as well. In 2011, several paintings and engravings were discovered in Cova Eirós, the first Palaeolithic rock art detected in north-western Iberia (de Lombera and Fábregas, 2013). Previously, the only evidence of Palaeolithic art in this region was the Férvedes II stone pendant in Xermade (Lugo) and Dentalium shells found on Lower Magdalenian levels at Valdavara 1 (Becerreá, Lugo). Mobile art Several items of mobile art have also been found at the Cova Eirós site (Fábregas et al., 2010; 2012). A small pendant in a perforated canine, probably from a fox (Vulpes vulpes), was found on the Gravettian level (Level 1). Several bone industry remains were also found, including a double pointed speartip from the interfaces of an Upper Palaeolithic level (1C). Both sides are decorated with a zigzag pattern composed of several parallel, discontinuous lines, whose closest equivalent would be Magdalenian items from the Altamira and El Pendo sites (Fábregas et al., 2012). Rock art There are several parietal art forms inside Cova Eirós from the Upper Palaeolithic. The ongoing characterization, classification and dating work of these motifs will permit a precise definition of the chrono-cultural coordinates of these Palaeolithic expressions. Eleven decorated panels have been identified to date. One of the features of the series –also a constraint for their study– is the poor state of the images due to heavy washing of the paintings and also the large amount of recent graffiti which has affected many of the motifs. The largest concentration is in the Main or Mammoth Hall, the widest space of the cave, 15 metres long and up to 5 m. high. Many painted and engraved motifs are on the walls at low or mid-height. Although figures have been identified in all sectors of this hall, the majority are on the west wall (Panels I to VI), looking inwards on the right. This sector contains the most complex themes and panels, with associations of black paintings and engravings, especially on Panel III. On the opposite wall is Panel VIII, with highly complex concepts, distribution and number of engravings, alternating signs with geometric and zoomorphic subjects. Graphic work has also been located in different recesses in the cave and the south-east gallery, an indication of the wide dispersal of these manifestations. Half of the motifs detected to date are painted elements, followed by engravings (46.4 %) and the possible use of reliefs or natural highlighting of the rock (3.6 %). However, engraved figures are most probably under-represented in the counts, as many lines have been classified together. It is also important to note that many black lines or dots may be the result of repeated visits to the interior zones of the cave during recent prehistory, early medieval and contemporary periods. The motifs are generally small and heavily influenced by the spaces and surfaces available in the cave. Quantitatively, the main themes are painted dots and lines; thin engraved lines, both individual and in sets; zoomorphs, both painted and engraved (bovids and possibly deer, equids and carnivores), many of them incomplete (partial representations of cervical-dorsal lines, hindquarters, etc.) followed by the representation of signs. The predominant painting technique is black. FT– Raman spectrometry analysis has identified the use of charcoal as a pigment. Engravings include thin, shallow lines, in some cases striated and others associated in a dense, variegated manner. Finally, the possible use of natural enhancements of the rock surface with morphologies reminiscent of animal silhouettes, in which specific dots or lines serve to highlight certain anatomical parts of the animal, have also been documented. There is a great technique and thematic homogeneity in this art. According to technical and stylistic studies currently underway, certain techno-morphological characters permit a working hypothesis for the chrono-cultural context of this art. The presence of numerous thin linear, composite and striated engravings superimposed on other motifs, the depiction of zoomorphs with elongated bodies and members, simplified representations of limbs, bodies with filled innards (e.g. bovid Panel I, Fig. 5) and the small-sized figures all seem to suggest the end of the Magdalenian/ transition to the Mesolithic. However, the infor-
24  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  6. Mo...
NORTH-WESTERN ATLANTIC BASINS. mation now available prevents us from ruling out the possibility of older motifs. 7. Conclusions The Cova Eirós site contains the most complete stratigraphic sequence now available for the study of the Middle and Upper Palaeolithic in Galicia. It allows a direct comparison of evolving technologies, subsistence strategies, adaptation to and exploitation of the territory between the Neanderthals and the H. sapiens of north-western Iberia. Cova Eirós is a reference point in tracing the history of the settlement of Galicia’s eastern ranges (Serras Orientais) and its relationship to the open air settlements and rock-shelters elsewhere in north-western Iberia. The archaeological record and the cave art at this site point to a similarity and even convergence with historical processes identified along the Cantabrian Coast. Acknowledgements The archaeological work underwayat Cova Eirós is part of the research project entitled, “Poblamiento durante el Pleistoceno medio/Holoceno en las comarcas orientales de Galicia” (HUM200763662, HAR/2010-21786 financed by the Ministry of Economy and Competition. Figure 5: Photograph filtered with D-Stretch and tracing of bovid and cervid zoomorphs on Panel I. 25
NORTH-WESTERN ATLANTIC BASINS.  mation now available prevents us from ruling out the possibility of older motifs. 7. Concl...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 26 Xosé Pedro Rodríguez*,** , Arturo de Lombera-Hermida *,**,***, Ramón Fábregas-Valcarce*** Paleolithic occupations in the Monforte de Lemos Basin (Lugo, Galicia) Introduction The information available about the Paleolithic in the north west of the Iberian Peninsula, particularly inland Galicia, is scarce and corresponds to accidental and sporadic finds. The only finding recorded for the Monforte de Lemos Basin (province of Lugo) was a handaxe in Vilaescura (Sober) in the mid twentieth century. Systematic research of the area started in 2006 as a consequence of accidental finding by an amateur but the findings from recent years have demonstrated the existence of a significant Paleolithic settlement in this area. The Depresión de Monforte, irrigated by the river Cabe, is a tertiary basin surrounded by higher Paleozoic and Hercynian areas, which reach 600m high to the west (Chantada area) and 1,600m to the east (Serra do Courel). The average height of the Cabe valley is 290 metres above sea level. The origin of the Monforte basin, with a surface area of 175 Km2, is tectonic, following the Hercynican fault lines that exist in a WNW-ESE direction. Following a neotectonic episode and subsequent fluvial rearrangement, the Pleistocene sediment linked to the paleo-channels and alluvial fans covered the banks with tertiary silts and clays of a lacustrine environment. These Quaternary deposits, arranged in a sequence of flat surfaces, are identified as river terraces, glacis and pediments (Ameijenda 2011). Arqueological surveys Systematic surveys, carried out between 2006 and 2010, were conditioned by the dense vegetation, which restricted the extension of the explored area. However, more than eighty open air artefact scatters were discovered, which correspond to * Lower, Middle and Upper Paleolithic. These locations have different entities, from sites with dozens of artefacts per square metre to others with just one isolated finding (Fig. 1). In addition to the surveys, test pits were dug in places whose concentration of artefacts could suggest the presence of archaeological sites in a stratigraphic context (As Lamas and Valverde in the Monforte Basin, and Pedras in the granite plateau of O Saviñao) (Fábregas et al., 2009; 2010). Findings were also discovered in stratigraphic context in the sites of O Regueiral and Áspera (Rodríguez et al., 2008). The geo-archaeological work has allowed us to reconstruct a relative chronological framework in accordance with the characteristics of the Quaternary surfaces and the technological interpretations of the lithic assemblages (de Lombera et al., 2011). The majority of the sites located could be assigned to Mode 2 or Acheulean lithic industries. Based on the topography and morphometrics of the Quaternary deposits (considering river terraces, glacis and pediments), 7 levels of erosion were identified on the sides of the basin (Fig. 1) (Ameijenda, 2011). The majority of the archaeological findings are located in intermediary levels (N4 and N5), whereas their presence in the other levels is scarcer, except in N1. The oldest evidence, in accordance with the morphotechnic characteristics and the sedimentary contexts, are located in Chao de Fabeiro (Erosion level 7), on the north bank of the Basin (Fig. 1). At this site, 26 knapped tools were recovered, mainly in quartzite. The configuration strategies are focused in the production of pebble tools. The presence of handaxes and chopping tools and the orthogonal, longitudinal and, to a lesser extent, centripetal reduction sequences suggest their ascription to Mode 2. The majority of Mode 2 sites are located in the intermediary surfaces (erosion levels N5-N4). In N5, the most significant scatters are those of Chao Vilar IPHES, Institut Català de Paleoecologia Humana i Evolució Social, C/ Marcel.lí Domingo s/n. Campus Sescelades, (Edifici W3), Tarragona 43007 ** Area de Prehistoria, Universitat Rovira i Virgili (URV), Av. Catalunya 35, Tarragona 43002 *** Grupo de Estudos para a Prehistoria do Noroeste (GEPN). Dpto. de Historia I, Universidade de Santiago de Compostela. Praza da Universidade 1, Santiago de Compostela 15782
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  26  Xos  ...
NORTH-WESTERN ATLANTIC BASINS. Figure 1: Location of the sites at the Monforte Basin. The squares indicate the Mode 2 findings; the circles, Mode 3 findings and the triangles, Mode 4 findings. Black indicates sites with a higher density of artefacts. The dots indicate isolated findings. The erosion levels are also specified (N1 to N7). The numbers correspond to significant sites: 1, Chao Fabeiro; 2, Chao Vilar; 3, As Lamas; 4, O Reguerial; 5, Valverde; 6, Costa Grande; 7, San Mamede. 27
NORTH-WESTERN ATLANTIC BASINS.  Figure 1  Location of the sites at the Monforte Basin. The squares indicate the Mode 2    ...
28 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD (1-2 and 3), with 62 items (Fábregas et al., 2009). As Lamas (UTM: 621.332, 4.711.619) is located in erosion level N4, where five archaeological points were discovered, which provided 241 pieces (Fig. 2.4-7). Two test pits were dug here in 2009: the test pit II revealed two archaeological levels, related to the colluviums that covered the fluvial sediments, dating by OSL in 39866 ± 3554 BP (Level 3) and 38947 ± 3150 BP (Level 2) (Fábregas et al., 2010).The first level (N-II) was related to a Middle Palaeolithic occupation and the second (N-III) provided artefacts with technical characteristics and alterations (oxides), which were very similar to the pieces recovered on the surface. Its presence in a colluvium from the Upper Pleistocene should be considered the result of an episode of erosion that destroyed old sediments during the Heinrich event 4. In these sites, the configuration strategies focus on handaxes, cleavers and trihedral picks. The majority of the handaxes are knapped on pebble and generally show a high amount of cortex on the surface. On the flake tools, the retouches create continuous and denticulate dihedral edges; sidescrapers (17.9%) and denticulate tools (13.2%) dominating over notch tools, becs and endscrapers. The most common reduction sequences are unidirectional unifacial / bifacial and centripetal, followed by the orthogonal method. The presence of the discoid method is minimum (5% of cores) and the Levallois method has only been identified in one core in Chao Vilar-II. Given the raw material used (quartzite pebbles) and its availability, the cores are knapped using natural platforms and their reduction sequences are short (60-70% of cores abandoned in initial or intermediary stages of reduction). Flakes usually show a high amount of cortex on the dorsal surface and faceted striking platforms are rare (3%). In the Mode 2 sites, the raw material used is quartzite, followed by quartz. The presence of fluvial “neocortex” in almost all of the artefacts suggests exploitation of the secondary deposits of Quaternary age along the basin. Given that the artefact scatterings are directly on these surfaces, access to raw material was immediate and quick. The Mode 3 sites are mainly located in erosion levels 1 and 2, particularly O Regueiral, As Gandariñas, Susao, San Mamede, Gullade and Level II of Test pit 2 of As Lamas (Fábregas et al., 2007, 2009, 2010). In these site Large Cutting Tools (LCT) are almost non-existent, whereas the small lithic flake tools take on more importance. The reduction strategies are dominated by the centripetal and discoid methods. The only archaeological records in a stratigraphic context were identified in O Regueiral and As Lamas Test pit II). In the first, located in an alluvial fan, 32 artefacts were recovered (Fig. 2.8, 2.10, 2.12), some of them in a colluvium dated by the OSL in 69446 ± 5472 BP. In the site at As Lamas (Test pit 2), Level 2 is linked to the upper colluvium but the rolling and size of the lithic tools suggest a primary context of disposition (Fábregas et al., 2010). In the Mode 3 sites, centripetal and discoid cores dominate (Fig. 2.9-10), while the orthogonal or longitudinal methods are less common. Some cores and tools in fine-grained quartzite could be related to the Levallois method, as they show surface hierarchisation in the reduction. However, use of the Levallois method in Monforte is quite limited, whereas discoid products are more common, particularly those related to the final stages of small quartz core reduction (Fig. 2.9). These lithic assemblages are dominated by flakes that often show centripetal disposition of the negative scars on the dorsal surface and the presence of dihedral and faceted striking platforms (Fig. 2.11-12). The configuration sequences focus on flake tools, denticulate tools (45.5%) dominating over side scrapers (27.2%) and other tools like end scrapers and points. Some large tools, like handaxes, could be present but there are very few (Fig. 2.8). In Mode 3 sites, the use of fine-grained quartzite increases in response to the most demanding reduction methods (Levallois and discoid) and the increased standardisation of flake tools. In the sites at Valverde, Costa Grande-III, Áspera and Pedrouzos de Mourelos, Mode 4 lithic assemblages were discovered. Unlike the Upper and Middle Palaeolithic sites, these are located in high places and not on the Quaternary surfaces. Their lithic industry stands out for the presence of blade technology and a wide range of raw materials, particularly rock crystal and flint. The most significant site is Valverde, located on a hillside 350 metres above sea level (UTM: 624.434, 4.713.497). The artefacts were made of quartz, fine-grained quartzite, rock crystal and flint (Rodríguez et al., 2008). Their techno-typological characteristics (cores and laminar and micro-laminar products, and leaf-shaped points), allow it to be ascribed to the Solutrean. Along with these chaînes opératoires from the Upper Palaeolithic, knapped in good quality raw materials (rock crystal, fine-grain quartzite and flint), bipolar knapping and the discoid method focus on the production of quartzite and quartz implements with good cutting edges (de Lombera et al., 2012).
28  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD   1-2 ...
NORTH-WESTERN ATLANTIC BASINS. Figure 2: Lithic industry of the Monforte Basin (Mode 2 and Mode 3). 1, Quartzite handaxe (Chao Fabeiro, Mode 2); 2, Quartzite chopper Chao Fabeiro, Mode 2); 3, Quartzite handaxe (Chao Vilar, Mode 2); 4 and 5, quartzite side scrapers (As Lamas, Modo 2); 6, quartzite notch (As Lamas, Mode 2); 7, Quartzite handaxe (As Lamas, Mode 2); 8, Quartzite handaxe (O Regueiral, Mode 3); 9, quartzite discoidal core (Gullade III); 10, quartzite discoidal core (O Regueiral, Mode 3); 11 and 12, quartzite flakes (Gándara Chá and O Regueiral, Mode 3); 13 and 14, retouched quartzite flakes (Gandariñas, Mode 3). 29
NORTH-WESTERN ATLANTIC BASINS.  Figure 2  Lithic industry of the Monforte Basin  Mode 2 and Mode 3 . 1, Quartzite handaxe ...
30 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 3: Lithic industry of the Monforte Basin (Mode 4). Valverde site (1-10): 1 and 2, leaf-shaped projectile fragments in flint (1) and quartzite (2); 3, micro-laminar core in rock crystal; 4 and 5, backed items in rock crystal; 6, 7 and 9: laminar and micro-laminar fragments in rock crystal (6) and quartzite (7 and 9); 8, bipolar core in quartz; 9, flint flake; 11, quartzite blade core, Áspera site.
30  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figur...
NORTH-WESTERN ATLANTIC BASINS. Conclusions The Monforte de Lemos Basin plays a strategic role due to its geographical location, as it is located in the natural route connecting inland Galicia to the western Meseta (de Lombera et al., 2011). It is also framed by the two main fluvial systems of the NW of the Iberian Peninsula (the Miño and Sil) that constrain structural mobility through this territory. Secondly, due to its low altitude, the average temperatures of the valley of the Sil and the Monforte Basin during the glacial period were warmer than in the surrounding areas and comparable to those recorded on the coast. Therefore, the Monforte Basin could serve as refuge area during glacial periods, whereas the mountain and inland re- Susana Alonso Fernández *, Manuel Vaquero *,**, Alicia Ameijenda Iglesias *** gion, with drier and colder conditions, would be covered by icecaps and steppes. The convergence of these geological, geographical and topographical characteristics could explain the high concentration of human settlements identified during the Middle and Upper Pleistocene. Acknowledgements The archaeological excavations in Monforte de Lemos are part of the research project “Poblamiento durante el Pleistoceno medio/Holoceno en las comarcas orientales de Galicia” (HUM2007-63662, HAR/2010-21786 of the Ministry for the Economy and Competition). La Cova and Valdavara 3 (Becerreá, Lugo) The archaeological activity carried out since 2007 in the municipality of Becerreá (Lugo) has produced an extensive archaeological record corresponding to different chronocultural periods, including times almost unknown before in Galicia, such as the late Upper Pleistocene and the Middle Magdalenian. The work was carried out in two sites: La Cova de Valdavara, in which a well-preserved sequence was found, which includes levels from the Late Prehistory and late Upper Paleolithic, and Valdavara 3, another cave deposit with remains of the late Upper Pleistocene (Vaquero et al., 2011). La Cova de Valdavara and Valdavara 3 are situated in the karst located on the right bank of the river Naron (known in this area as Cruzul), one of the tributaries of the left bank of the river Navia, at 120m and 220m, respectively, above the current channel of the river. They are part of a cave system located in limestone formations called the Calizas de Vegadeo, from the Lower-Middle Cambrian, outcropping along a wide stretch plotted NNWSSE between the Palaeozoic formations of the Dominio del Manto de Mondoñedo, where sandy and slate rocks predominate (Vera, 2004). Some of these caves, like Cueva de Furco or Cova da Venta (Fernández Rodríguez, 1993; Grandal, 1991), were already known for having provided some isolated archaeological or paleontological remains. La Cova de Valdavara was discovered in the 1960s by a group of amateurs from Becerreá. The archaeological excavations were carried out between 2007 and 2013, working in three sites with different sequences and deposits: Valdavara 1, which is the original cavity found in the 1960s, Valdavara 1-2, which corresponds to the outer slope and Valdavara 2, another small cavity located barely 6 m below Valdavara 1. In Valdavara 1 a sequence of almost one–and-ahalf metres deep has been documented, consisting of two main stratigraphic units (Fig. 1). The upper unit corresponds to the Late Prehistory (Vaquero et * Institut Català de Paleoecologia Humana i Evolucio Social (IPHES), Escorxador s/n, 43003 Tarragona ** Area de Prehistoria, Universitat Rovira i Virgili, Avinguda Catalunya 35, 43002 Tarragona *** Universidade de Santiago de Compostela (USC) 31
NORTH-WESTERN ATLANTIC BASINS.  Conclusions The Monforte de Lemos Basin plays a strategic role due to its geographical loc...
32 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. Stratigraphic sequence of Valdavara 1. al., 2009). The study of the flint and quartz lithic industry, the domestic fauna and pottery recovered in this upper unit demonstrate varied use of the cavity during the Late Neolithic-Chalcolithic. The recovery of decorative seashells (Dentalium) infers mobility across the territory of these populations and possible contact with coastal population. Similarly, the recovery of small human remains, such as phalanges and teeth of different individuals, suggest repeated use of the cavity for burial purposes, probably as a primary burial site (Vaquero et al., 2009). The variety of micro-mammals recovered is consistent with other Holocene-Chalcolithic associations from the north of the Iberian Peninsula and suggests a more Mediterranean climate (Blain et al., 2009). The lower unit corresponds to the late Upper Paleolithic, when it seems that the cavity was repeatedly occupied between 16,800 and 18,700 cal. years BP (Table 1) (Vaquero et al., 2009). The evidence documented, above all in level 4, is coherent with the characteristics of the Lower and Middle Magdalenian of the Cantabrian coast: laminar flake knapping, clear dominance of burins among the retouched objects, portable art, decorative bone industry and decorative objects: including an assemblage of seashells and a pendant made from an atrophic red deer canine tooth (Fig. 2). Particularly significant is the recovery of a deciduous tooth in this level, which is the oldest human remain found in Galicia to date. A significant cultural change seems to occur at the base of this packet, characterised by the appearance of quartzite lithic objects. The wide variety of species of micro-mammals recovered has provided a large amount of environmental and paleo-environmental data for the period and suggests cooler conditions than in the upper unit (Lopez-García et al., 2011). At the site of Valdavara 1-2, under a top level that was significantly altered by post depositional processes and which contained very heterogeneous materials, a more homogeneous unit appeared (level C) with lithic artefacts in flint and quartz, indicating some technical strategies aimed at obtaining flakes. Dating of the associated remains of fauna indicates an Early Holocene chronology for this level (Fábregas et al., 2010). This dating, along with the characteristics of the lithic assemblage, is coherent with attributing this level to the Macrolithic Mesolithic, which is documented for the first time in Galicia. Below this unit, we have identified another level (level D), characterised by the predominance of quartzite flakes that, if their chronology were confirmed, would correspond to the oldest occupation documented in Valdavara cave. In Valdavara 2 a level was located at the top of the sequence containing the remains of at least three infant individuals, devoid of any associated material culture objects, but whose dating puts them in the chronological context of the Middle Bronze Age (Vaquero et al., 2008, Vaquero et al., 2009). Another stratigraphic unit appeared below this level with a fossil record characterised by the presence of a large number of carnivore remains but with no evidence whatsoever of human presence.
32  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figur...
NORTH-WESTERN ATLANTIC BASINS. Loc. Level Material Lab.ref. Years 14C Years cal. BP Years cal. BC Val 1 2 Human bone Beta-235727 4410 ± 40 5160-4840 3210-2890 Val 1 4 Bone Beta-235728 13,770 ± 70 17,080-16,880 15,130-14,930 Val 1 4 Bone Beta-235726 14,630 ± 70 17,890-17,730 15,940-15,780 Val 1 6 Bone Beta-257849 15,120 ± 70 18,700-17,820 16,750-15,870 Val 1-2 C Bone Beta-257850 8920 ± 50 10270-9830 8320-7880 Val 1-2 C Bone Beta-259199 8890 ± 60 10250-9770 8300-7820 Val 2 3 Human bone Beta-235729 3270 ± 40 3600-3400 1650-1450 Val 2 3 Human bone Beta-235730 3250 ± 40 3600-3360 1650-1410 Table 1. Radiocarbon dating of Valdavara cave. Calibration at 2s (p= 95%) has been performed using the cal Pal-2007-Hulu curve (Weninger and Jöris, 2004). Between 2009 and 2011 an rescue excavation was carried out in a cavity exposed due to the consequences of the blasting of a quarry 100 metres from la Cova de Valdavara, which was named Valdavara 3. The large variety of fauna recovered from this site (lion, leopard, hyena, rhinoceros, bison, bear, fox, wolf, boar, roe deer, chamois, fallow deer), the characteristics of the micro-fauna and the first radiometric data suggest a chronology of early Upper Pleistocene, between 100,000 and 120,000 years. Although it is mainly a palaeontological assemblage, the discovery in situ of lithic industry stratigraphically associated to the fauna confirms human presence in the site. Although the chronology of the deposit is yet to be verified, this data places Valdavara 3 as an important reference regarding the oldest settlement in Galicia. Figure 2. Personal decorative objects found in the Magdalenian levels of Valdavara 1. 33
NORTH-WESTERN ATLANTIC BASINS.  Loc.  Level  Material  Lab.ref.  Years 14C  Years cal. BP  Years cal. BC  Val 1  2  Human ...
2 cantabrian mountains and coastline NORTH-WESTERN ATLANTIC BASINS. 35
2  cantabrian mountains and coastline  NORTH-WESTERN ATLANTIC BASINS.  35
Site Map numbering Site Map numbering Cueva de Aitzbitarte III and IV 5 Cueva del Mirón 21 Antoliñako Koba 6 Santimamiñe 22 Axlor 7 El Sidrón 23 Los Azules 8 Sopeña 24 Cabo Busto 9 El Castillo 10 Cueva de la Güelga 11 Cueva de Las Caldas 12 Cueva Morín 13 Ekain 14 Cueva del Esquilleu 15 El Pendo 16 Cueva de la Riera 17 La Viña 18 Labeko Koba 19 Lezetxiki 20
Site  Map numbering  Site  Map numbering  Cueva de Aitzbitarte III and IV  5  Cueva del Mir  n  21  Antoli  ako Koba  6  S...
CANTABRIAN MOUNTAINS AND COASTLINE. Jesús Altuna*, Koro Mariezkurrena*, Joseba Ríos** La Cueva de Aitzbitarte III and IV (Basque country) In Aitzbitarte Hill (Rentería, Basque Country) there is a set of caves of which the two largest, III and IV, have been excavated in recent decades. The latter was worked on in the 1960’s by J. M. Barandiaran (1961). E. Harlé (1908) found reindeer remains there for the first time in the Iberian Peninsula. Recently, number III has been worked on by J. Altuna (Altuna et al., 2011). Both caves are located 7 km from the current coast in a straight line. Cave III develops to the southwest. It has a large vestibule which leads to a big hall that is 60m long and averages 15m wide. Recent excavations have been conducted at the entrance of this hall and in its deep zone. Here we report the last campaigns carried out at the cave entrance. Currently, the excavations carried out in the profound area are under study. Stratigraphy (Fig. 1) showed a Middle Paleolithic level (VI), a mixed Middle Paleolithic and Aurignacian level (Vb inferior), an Evolved Aurignacian (Vb central), three Gravettian levels (upper Vb, Va, IV), and two other levels (III and II) with Solutrean chronology but linked to the Gravettian technocomplex. Level I shows an admixed composition of reworked materials. The level Vb central has five 14C AMS dates made in Uppsala. Four of them have provided dates to 33,605 ± 1165 to 31,000 ± 835 and an other to 28,010 ± 600 BP. The Va level has five dates the oldest one being 31,210 ± 860 and four more of 28,950 ± 655 to 26,350 ± 475. In addition the IV has six dates, five from 28,320 ± 605 to 24,240 ± 365 and an other to 22,420 ± 290. The III has seven dates, grouped from 22,580 ± 295 to 18,400 ± 215. Finally, level II also has one which dates to 19,765 ± 220 (Ua-37959). The sedimentological study has shown that the sequence was formed under climate conditions characterized by low temperatures and humidity. Level VI is the warmest assembly. The level V, especially Vb, is gradually wetter and colder. Level IV is also damp and * ** slightly warmer, while in level III the harshest conditions have been recorded, ameliorating gradually towards the end of the sequence. The pollen study indicates a predominance of open landscapes with sparse tree representation (less than 6%).Level VI shows more arboreal presence (5.6%) with Pinus, Juniperus and Betula. The Vb central level is very cold (tree cover 1.6%) with Pinus and Juniperus. The upper Vb shows weak tree recovery (2.5%) with the same species. The Va also indicates a weak recovery in Betula. The IV displays an overall increase in Juniperus (3%).Due to the lack of spore pollen conservation, information was not obtained on levels III and II. A total of 16,556 Micromammal remains were recovered using a 1 mm mesh sieve. The greatest part of them come from levels Vb to III. Represented species are similar in all levels, Microtus gr. Agrestis-arvalis being the most abundant, followed by Microtus oeconomus. Pitymys pyrenaicus is the most represented species of the genus Pitymys, and there is also evidence of Pliomys lenki relict species. This species representation indicates a climate characterized by cold and moist conditions, and an open landscape with scarce wood patches. This information, obtained by sedimentological, polinical and micromammal analyses, is also confirmed by macrommal representation, with a good representation of open landscape animals and the presence of species usually linked with colder climates, such as Rangifer tarandus or Alopex lagopus. Bird fossil assemblages also correspond with these environmental conditions identifying the species Lagopus mutus, which lives today in the upper edge of coniferous forests and rocky high mountain areas in the Alps and the Pyrenees. The lithic and bone industry reveals transformations between the Evolved Aurignacian (Vb central) and the beginning of the Gravettian (Vb superior Va Centro de Custodia e Investigación de los materiales Arqueológicos de Gipuzkoa. 20015 San Sebastián (GOAZ). Altuna@arkaios.com Centro Nacional de Investigación sobre la Evolución Humana. 09002 Burgos. joseba.rios@cenieh.es 37
CANTABRIAN MOUNTAINS AND COASTLINE.  Jes  s Altuna , Koro Mariezkurrena , Joseba R  os    La Cueva de Aitzbitarte III and ...
38 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. Stratigraphic units from Aitzbitarte cave III and IV) and provides data characterizing human occupations around 20,000 BP (levels III and II), which have a difficult cultural attribution in the Cantabrian regional framework. Level VI was excavated in 1 m2, and the scarce material recovered can only be attributed to Middle Paleolithic technocomplexes, without more precision. The upper part of this level is mixed with the lower part of Vb (Vb inferior), so the industrial complex shows an anomalous nature mixing elements from Middle Paleolithic and Evolved Aurignacian. Level Vb central displays intense bladelet production from cores, flake edges and Vachons type cores, and also unipolar production of wide and flattened blades. The types of flint used were obtained in nearby outcrops, either the Urgonian, located in Aitzbitarte hill’s limestone, or a Flysch variety from Gaitxurizketa (<10 km N).The use of flint from outcrops located between 30 and 100 km (Flysch flint from Barrika- NW- or Bidache- N) and flint coming from more distant outcrops in Chalosse (130 km, N), Salies de Bearn (100 km, N), Urbasa (90 km, S) or Treviño (125 km SW) has also been documented. The most represented formal tools are Aurignacian blades and the typical Aurignacian scrapers. Burins exhibit great variations of functions and morphologies, the most characteristic being the burin-cores (Vachons, Dihedral and Multiple) and burins on truncation. Substratum tools and splintered pieces are also very abundant, while the bladelet tools are scarce, the Dufour type retouched bladelets are the most represented type. There is also plenty of barely configured macrolithic industry. Bone industry is scarce and non-diagnostic, composed mainly by lissoirs and retouchers. Also found in this level was one engraved schist slab with nonfigurative depictions that connects with the first Cantabrian graphic traditions (Fig. 3 and 3b). The assemblage can be characterized as palimpsest of occupations with intense and varied activities, for instance the processing of fauna. The industrial characteristics along with the dates obtained on the level, allow a cultural affiliation to the Evolved Aurignacian. The upper section (Vb superior) shows a change from level Vb central. The manufacturing of bladelet is made from unipolar prismatic cores; also laminar knapping is more regular, producing narrow blades. The retouched assemblage shows a decrease of typical Aurignacian tools and an increase of burin types, with some examples of Noailles burins (Fig. 2). From the bone industry the awls and the bone points stand out. At the level Va the production of unipolar blades and bladelets from prismatic cores, pyramidal cores and burin-cores is remarkable. The most characteristic retouched tools are the burins, dihedral and on truncation, with some typical Noailles, and the splintered pieces. The bone technology shows a limited presence of pointed artifacts, possibly bone points, alongside other “domestic” tools.
38  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figur...
CANTABRIAN MOUNTAINS AND COASTLINE. Figure 2a. Platelet engraved from Aurignacian level Figure 2b. Platelet engraved. Drawing.. No remarkable difference was noted in the knapping strategies between levels IV and Va. The presence of diverse tool types is notable, with some Noailles burins and splintered pieces, followed by scrapers, truncated blades and backed tools. The only evidence of bone industry is one bone point fragment. The assemblage composition also seems to be related to domestic activities. The upper part of level Vb and levels Va and IV, are techno-typological and chronologically correlated with an early Gravettian phase. The decrease in the use of local flint (Urgonian or Gaitxurizketa) is significant in comparison with Evolved Aurignacian, as is the increase of flint coming from more distant sources, such as Chalosse or Bidache. Level III shows a great variability in production systems identifying production of narrow and wide bladelets, both from cores or flake, and production of flakes and the importation of already produced blades. The retouched assemblage is composed of burins of various types, including Noailles and Busqué, backed bladelets, scrapers and two geometrics. The bone tool assemblage is the richest and most varied of the entire sequence, with two exceptional needle fragments and two bone points with flattened circular sections. In addition there is one pendant made of canine fox and one fragment of bone bead. This level corresponds to multifunctional occupations, including domestic activities and the preparation and restoration of hunting tools. Level II is characterized by the production of different sized bladelets, especially from pyramidal cores and probably from Busqué type burin-cores; while flake and blade production seems to have tak- en place outside the cave. Regarding the retouched tools, most represented types are burins, specially Noailles and Busqué types. On the contrary, scrapers are scarce and there are not clear foliate types. Pieces with abrupt retouch, especially truncated and backed tools, are relatively abundant, and maybe some of them could be interpreted as fragments of shouldered points. The bone industry shows a pos- Figure 3. Noailles burin from Gravettian levels 39
CANTABRIAN MOUNTAINS AND COASTLINE.  Figure 2a. Platelet engraved from Aurignacian level  Figure 2b. Platelet engraved. Dr...
40 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD sible bone point fragment and bird bone tube with transverse linear engravings. Levels III and II chronologically coincide with the beginning of regional Solutrean: nevertheless the industrial assemblage shows distinct characteristics closer to the Gravettian technocomplex. For that reason, it seems an original industry variety linked with other assemblages as Amalda (V) or Ermittia (V). Meanwhile, level I appears extremely altered offering unreliable data. Level III (recent Gravettian) has provided four human teeth belonging to three children under 13 years of age and one adult. It is, so far, the oldest known remains of our species in the Basque Country. The identified evidence is: one D2 corresponding to newborn between 12 - 21 months, one M1 belonging to a 5.5 - 6 year old child, one M2 from another child aged between 11- 13 years and one P4 from an adult. The macromammals assemblage is composed of 30,261 remains, generally very fragmented, of which 2,149 have been taxonomically identified. The large number of undetermined specimens responds to the 2 mm water sieving and to the exhaustive recovery of fragments during the excavation. The macromammals are the subsistence basis for the human groups that occupied the cave. Most represented species are the bovine (among them Bos primigenius and Bison priscus), Cervus elaphus and Rupicapra rupicapra. The presence of other ungulates such as Rangifer tarandus, Capreolus capreolus and Capra pyrenaica, is scarce, some of these species being absent in certain levels. In the Vb central level (Evolved Aurignacian) Cervus elaphus dominates, followed by bovine and Rupicapra rupicapra. The Ursus spelaeus also has a large representation. The upper level Vb (Early Gravettian) shows a similar faunal spectrum, leading subsequently to bovine increase during Early Gravettian (Va and IV). Finally on levels III and II bovine presence decreases, and Rupicapra rupicapra increases. Reindeer is represented by a few remains in levels Va, IV, III and II. There are numerous anthropogenic traces on the bones (fractures, multiple incisions, traces of fire in changing degrees). There are a few traces of Carnivore bites, probably on abandoned human faunal remains. An Ursus spelaeus ulna with cut-marks recovered in Vb central level is also remarkable. On the other hand, level IV is notorious for its large hearths and for the large amount of chopped and burned small bone fragments. The elevated proportion of spongy bone fragments with high fat content, indicates that they probably served as fuel during the harshest periods with low tree cover. The Bird assemblage is small, with 137 determinable remains, mostly concentrated in levels III (82) and IV (32). Identified remains correspond basically to Lagopus mutus, Perdix perdix and Pyrrhocorax. While these findings are compatible with their introduction in the cave by humans, at the moment no anthropogenic modifications have been observed in the bone fragments. Also, the appearance of rock species, especially at level III, indicates that the cavity was not permanently occupied by humans at that moment. Fossil fish assemblage is also small, with 156 remains. The most abundant genus is Salmon, especially trout. The presence of Anguilla anguilla, Trisopteros minutus (III), Blenius sp. (Va and IV), Scomber sp. (Va) and Platichys flesus (Vb) is significant. This latter species occupied the seawater areas of river mouths and could be caught in the Oiartzun River. The mollusks assemblage is composed by shells of marine origin, typical of the Bay of Biscay, such as: Patella vulgate (l. I, II and III), Littorina littorea (l. II, III and IV), L. obtusata, L. fabalis (both in l. III), Stramonita haemastomsa, Mytilus sp., Pecten maximus and Antalis sp. (four of them found in l. II). Three shells have been transformed into hanging ornaments. In the Aitzbitarte III exterior sequence an interesting archeoestratigraphic sequence has been documented, environmental evolution and transformations in human behavior and culture, including the site function, subsistence strategies, and tool procurement, have been observed. At this time the research focused in the interior part of the cave is very advanced, specifically in the rich Gravettian levels with Noailles burins, where subsistence strategies were based almost exclusively on bovine hunting (over 90% of the remains). The level IV, of la Cueva de Aitzbitarte located above the III, has a poorly defined Aurignacian level, one Solutrean, two Magdalenian and another Azilian level. Also in the deepest part of the cave paleolithic paintings have recently been found (Garate et al., 2013). In comparison with cave III, the absence of Gravettian occupations in cave IV is noteworthy, and likewise the late Paleolithic occupations represented in this cave (Magdalenian and Azilian) are absent from cave III.Antoliña Cave (Antoliñako koba in Basque) (DATUM ETRS 89 x: 528.685.46 Y: 4.801.186,83 Z: 285) is placed within a basin downstream of Oka or Gernika River, in area called Urdaibai (Bizkaia). It is close to the Santimamiñe classic archaeological site (Fig. 1: 1). Its mean height above sea level is 285 m; strategically orientated in the southwest direction, allowing visual control in most part of the Urdaibai basin and access to the coastal valleys of the rivers Lea and Artibai.
40  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  sible...
CANTABRIAN MOUNTAINS AND COASTLINE. Mikel Aguirre* Antoliñako Koba (Gautegiz-Arteaga, Bizkaia) Antoliña cave (Antoliñako Koba in Basque) (DATUM ETRS 89 x: 528.685.46 Y: 4.801.186,83 Z: 285) is located in a basin downstream of the Oka or Gernika River, in an area called Urdaibai (Bizkaia). It is close to the Santimamiñe classic archaeological site (Fig. 1:1). Its average height above sea level is 285 m; it is strategically orientated toward the southwest, allowing visual control of most parts of the Urdaibai basin and access to the coastal valleys of the Lea and Artibai rivers. The archaeological excavation was carried out between 1995 and 2008. It has allowed the identification of a major stratigraphic sequence of Aurignacian, Gravettian, Upper Solutrean, Lower advanced Magdalenian occupations, evidence of Upper Magdalenian and Azilian (Aguirre, 1996, 2001; Aguirre et al., 2001). The archaeological record and paleo environmentare the most complete and richest in the Urdaibai between c.35, 000 BP and 14,500 BP, with some well-defined stratigraphic hiatuses in sequence with essentially anthropogenic accumulation dynamics. Are they demographic contractions in the eastern Bay of Biscay? Is this a change of strategy in territorial exploitation? There have been some contributions to the origin of the siliceous raw materials from Antoliña (Tarriño et al., 1998; Tarriño, 2006: 136-139) and the paleoenvironment from microvertebrates sequence (Zubeldia et al., 2006). Currently, an interdisciplinary archaeological study is underway. The configuration of the caves’ sedimentary fill has significant lateral changes in their character and vertical development. In the west room (Fig. 1: 2), stratigraphy has little impact due to the proximity to the rock base; we call this Upper Sedimentary Set (USS) and Middle Sedimentary Set (MSS). Meanwhile, the Lower Sedimentary Set (LSS) was of considerable depth in the northern part of the gallery and lobby with archaeological levels absent. Here we will expose, from wall to roof, different documented archaeological horizons (Fig. 2). * Horizon 1 – The earliest evidence of Antoliña occupation is located in a confined level between stalagmite filtering (Fig 2) from LSS, Sbl-P (brown sandy silt with phosphates) where one marmot was identified among the poor faunal evidence, along with one scraper and a few other lithic elements. There is currently no dating for these strata, but there is some geochemical analysis (Yusta et al., 2005). These levels belong to the lower half of this sedimentary base assembly. Observation has been done of forming speleothems, blocks and limit clasts, as well as different sandy layers, sometimes brecciated brown silty clay, and packages of very compact orange clays mixed with phosphates. All of this it is archaeologically sterile except for the mentioned stratum. Horizon 2 – The top half of LSS is a thick phosphate with various clay packages and speleothems. It contains the archaeological first level of a certain entity, composed of brown sand with abundant phosphate lumps (Sb-P). This horizon displays Aurignacian industry, the most significant being the simple retouching (in denticulate and scrapers) with a few Dufour flakes. Horizon 3 – It is covered by orange sands and phosphates (So-P) with similar archaeological lithic material and scarce fauna; poorly preserved at these levels, dating is pending. Its roof differentiates from upper zone in color, texture and small bioturbation. Horizon 4 – Above, it is defined by one reddish brown compact sand package (Sbk), with infrequent clasts or corroded blocks and countless lateral variations (Table 1). We have data near the roof, 30,640 ± 240 BP (Beta-251304) and one on charcoal 29,990 ± 230 BP (GrA-23,898) in the lower level section. This horizon contains hue evolved Aurignacian industries, with significant percentages of substrate groups, rare Dufour bladelet and ordinary bone industry. This stratum, containing compact brown silt (L) and sand with blocks (Lbk lower / Sbk), forms the basis of the MSS. UNED, CA Bergara, San Martin Agirre Plaza 4, 20570 Bergara (Gipuzkoa) maguirre@bergara.uned.es 41
CANTABRIAN MOUNTAINS AND COASTLINE.  Mikel Aguirre   Antoli  ako Koba  Gautegiz-Arteaga, Bizkaia   Antoli  a cave  Antoli ...
42 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. 1. Location of Antoliña. 2. Map of the excavation. 3. Antoliña Gravettian engraving on a pebble.4-7. Retouched falt Solutrean artifact from Antoliña. 7. Indeterminate ungulate rib with a distal tip end.
42  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figur...
Cultural att. Aziliense Lower Magd. Upp. Solu. Upp. Solu. Gravettian Gravettian Evol. Aurig. Ref. Level Lanc/ Upper Lgc Lower Lgc Lmb Lmc Lab/Sab Upper Lmbk/Smbk Lower Lmbk/Smk Material Bone Bone C14AMS C14AMS C14AMS Bone Bone C14AMS Bone C14AMS Bone Bone C14AMS Bone Bone Bone Charcoal Bone C14AMS Bone C14AMS Bone C14 Bone Bone C14AMS C14AMS Bone C14 C14AMS Bone C14 Bone Methhod C14AMS C14 C14AMS C14AMS C14AMS Beta-251304 GrA-23898 Beta-230279 GrN-23786 Beta-251300 Beta-251299 Beta-230282 Beta– 251303 Beta– 215542 Beta-233766 GrN-23785 Beta-230284 Beta-251301 GrN-23784 GrN-23783 Beta-230280 Beta-230281 Beta-215543 Beta-215544 Ref. lab. Table 1. Dating of Antoliñako koba and its calibration (2007 Calpal HULU, Weninger et al., 2010). Middle Upper Sedimentary ensemble 30640 ± 240 BP 29990 ± 230 BP 27520 ± 190 BP 27390 ± 320 BP 27100 ± 190 BP 26720 ± 180 BP 26710 ± 180 BP 26140 ± 150 BP 26080 ± 200 BP 22640 ± 120 BP 19280 ± 120 BP 19020 ± 120 BP 17340 ± 100 BP 14680 ± 100 BP 14680 ± 80 BP 14630 ± 70 BP 14580 ± 70 BP 10800 ± 40 BP 10220 ± 40 BP BP Date 34823 ± 355 34253 ± 196 32109 ± 222 32064 ± 277 31822 ± 154 31486 ± 273 31469 ± 284 31046 ± 343 31011 ± 356 27358 ± 388 23044 ± 283 22879 ± 303 20801 ± 301 17998 ± 378 18003 ± 374 17954 ± 379 17829 ± 314 12777 ± 63 11932 ± 128 CalBP 34467 – 35178 34057 – 34449 31887 – 32331 31786 – 32341 31667 – 31976 31212 – 31759 31185 – 31753 30702 – 31389 30655 – 31367 26969 – 27746 22760 – 23327 22575 – 23182 20500 – 21102 17619 – 18376 17628 – 18377 17574 – 18333 17515 – 18143 12713 – 12840 11804 – 12060 68% range calBP CANTABRIAN MOUNTAINS AND COASTLINE. 43
Cultural att.  Aziliense  Lower Magd.  Upp. Solu. Upp. Solu. Gravettian  Gravettian  Evol. Aurig.  Ref. Level  Lanc  Upper...
44 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 2. Test of provisional stratigraphy chrono-climatic correlation in Antoliña, such as radiocarbon and curves of temperature variation and partial processing of column sampling microvertebrates (Zubeldia et al., 2006). Horizon 5 – After a relatively short sedimentary hiatus (with no interfaces defined in the west room, but clearer in the north gallery) it develops a package with high archaeological material density at the base, compact yellowish brown silt and sand with abundant blocks (Lbk upper / Sybk) attached to Gravettian with Noailles bur- ins (Aguirre, 2013). In the bone industry there are outstanding oval section assegais, with morphologies similar to the Isturitz. As for its implications, one small hammer-abraseur with deer morphology is remarkable (Fig. 1: 3) (Aguirre and Gonzalez Sainz, 2011). We have six dates found with Accelerator mass spectrometry (AMS) and 14C,
44  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figur...
CANTABRIAN MOUNTAINS AND COASTLINE. as one conventional, staggered consistently between c. 25,800 and 27,700 BP (c. 31,000-32,300 calBP). The archaeological material, Gravettian type with Noailles, becomes scarcer in the upper sections, with yellow silt and sand with blocks (Lyb / Syb). Toward the roof net erosional contact occurs. It has been dated to 22,640 ± 120 BP (Beta233766). Horizon 6 – The USS begins over the Tardiglaciar erosive difference. It has significant Upper Solutrean occupations: the lower one of brown silt with clasts (Lbc) and the upper with brown silt blocks (Lbb). Both abundant with flat retouched pieces (Fig. 1 4 7), as well as a predominance of concave bases (about fifty) and uncommon little lateral notches. Noailles burins are present. There are three dates: 19,280 ± 120 BP (GrN-23,785) and 19,020 ± 120 BP (Beta-230284) for Bsc; and 17,340 ± 100 BP (Beta-251301) for the upper Lbb. Horizon 7 – After another hiatus, with erosion signs on the Lbb roof, it follows grayish silt with clasts (Lgc), already in surface in one lobby section and the west room. They are greatly affected by various conditions (cattle, illegal) reaching the Solutrean levels in most of the excavated González-Urquijo, Jesús*; Ibáñez, Juan José**; Lazuén, Talía***, Mozota, Millán** Horizon 8 – Unable to establish sedimentary differentiation with upper section Lgc, (Lgc upper), in some exceptional locations it is possible to find greater density of small clasts and mollusks between the archaeological material. It is dated (Beta-215543) to 10,800 ± 40 BP. In marginal areas of the northwest room and small gallery zones of black (negro: n) silty-clay and silt with clasts (Lcnc), overlapping Lgc, a speleothems roof is conserved, sealed and dated in 10,220 ± 40 BP (Beta215544), data that is very close to the previous, but not coinciding. During screening of scrambled sediment,several fragments of Magdalenian harpoons, hardly attributable to such late dates, were recovered. Logically, this fact shows occasional visits during the Magdalenian to the central section of Lgc. The composition of archaeological material of upper Lgc and Lcnc is similar (comprised of backed elements and analogous variety of mollusks) assigning it to Azilian. Axlor Research History The site of Axlor (Dima, Bizcaia) is located on the northern Atlantic coast of the Iberian Peninsula. The cave mouth is located at about 320 m above sea level on an interior valley near one of the crossing points between the Atlantic and Mediterranean basins. The site was discovered by archae* area. The bone industry evidence comprises sub-triangular and square sections and spatulate points. There are four dating between 14,510 and 14,780 BP (17,515 and 18,377 calBP), two of them completed with AMS and the others with conventional methods. The lower section of this horizon corresponds to Lower Magdalenian (Lgc lower). ologist J.M. Barandiarán in 1932 and was excavated by him between 1967 and 1974. Barandiarán described nine stratigraphic layers and defined levels III to VIII as fertile, all of them with Mousterian lithic industries (Barandiarán, 1980). The faunal remains of big mammals were studied by J. Altuna (1989). The lithic and bone industries were analyzed by A. Baldeón in his doctoral thesis Instituto de Prehistoria (IIIPC) / Departamento de Ciencias Históricas, Universidad de Cantabria, Avda de los Castros, s/n, 39005 Santander, gonzalje@unican.es ** Institució Milá i Fontanals, CSIC, C/Egipciaques, 15 08001 Barcelona, ibanezjj@imf.csic.es and millanm@imf.csic.es *** PACEA (CNRS-Université de Bordeaux), Allée Geoffroy de Saint Hilaire 33615 PESSAC t.lazuen@pacea.u-bordeaux1.fr 45
CANTABRIAN MOUNTAINS AND COASTLINE.  as one conventional, staggered consistently between c. 25,800 and 27,700 BP  c. 31,00...
46 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD (1985) and extensively published later (Baldeón, 1999). A. Baldeón classified all levels as Charentian Mousterian –with some variations for the two lower levels– having abundant sidescrapers. Stratigraphic Sequence In 2000, the excavation of the Axlor site was restarted by a team led by J. González-Urquijo and J.J. Ibáñez. The new excavation took place in contemporary levels to those of the sequence excavated by Barandiarán. Roughly, levels B to N (Fig. 1) correspond to levels III to VIII of the excavation of the 1960’s and 1970´s (González-Urquijo et al., 2005, 2006). This sequence which chronologically corresponds to the final stages of the Middle Paleolithic. Level D is dated to 42,010 + 1,280 BP (Beta-144262) and the lower levels of sequence (M and N) date to 47,500 BP or earlier. The sequence, recognized until then, was completed, at the base, with a mass of yellow clay: level IX. The new excavations have revealed a sedimentary deposit stratigraphically located below the original sequence, with two sublevels containing archaeological material deposited around OIS 4-5. A deep drilling, down to -6.80 marks, revealed a huge sterile filling. Also, the remains of a layer of the early Upper Paleolithic were located in the upper section of the sequence (level A) corresponding to Figure 1. Axlor Stratigraphy. the base of level II of J.M. Barandiarán, which had been considered sterile at the time. At the lower levels (Axlor R), the lithic industry is scarce (n = 414), and is knapped in flint (46.5%), quartz (27%), silicified mudstone (14%), limestone (8%), quartzite, limonite, and sandstone. In the immediate surroundings of the site, the chances to find useful rocks for knapping are limited to silicified mudstone, limestone, limonite, and sandstone. These materials are relatively easy to locate in both primary and secondary position in the nearby scree and waterways, less than 1 km (0.62 miles) from the cave. The best quality silicified mudstone is found in the formations of the Supraurgonian black Flysch a few kilometers north of Axlor. However, blocks usable for knapping can be found in Albian age outcrops, a few hundred meters from the site. Quartz can be discovered on the edges of Biscay synclinorium about 5-10 km north from the site. We do not know the origin of the quartzite found in very low amounts in fluvial deposits near Axlor. The flint basically comes from the coastal Flysch, located 35-40 km from the site. The bulk of the set is composed of débitage residues (65%) and flakes (28%), whereas cores and retouched tools barely reach 1% and 5.8%, respectively. It should be noted, however, that out of the 116 flakes in the collection 48 (41%) correspond to rejuvenation flakes. The most represented utensils are the sidescrapers which are of different
46  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD   1985...
CANTABRIAN MOUNTAINS AND COASTLINE. types: lateral, transversal, and double, four of them featuring Quina or Quina-like retouch. A different management of flint over other raw materials can be observed in level R. Flint artifacts are smaller and much more often retouched. An impor- Figure 2. Quina sidescrapers. tant part of them are in fact rejuvenation flake from Quina sidescrapers, which sometimes have been, in turn, retouched and used (Lazuén and González Urquijo, i. p. a). The functional study of a sample of lithic tools of this layer suggests the development of 47
CANTABRIAN MOUNTAINS AND COASTLINE.  types  lateral, transversal, and double, four of them featuring Quina or Quina-like r...
48 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD a variety of tasks, including butchery, as well as work on non-woody plants, dry hide and wood (Lazuén and González Urquijo, in press b). At the bottom of the modern sequence (levels M and N), the raw materials used for the lithic production are similar to those of the lower layers: flint (45%), quartz (30%), silicified mudstone (20%), quartzite, and others. Flint comes from outcrops to the north, in the coastal Flysch in the form of flakes and tools of medium or large size (> 5 cm) or small Levallois-type cores. These cores produced flakes of up to 1-1.5 cm as can be observed in the negatives of the latest extractions. A significant portion of the tools at the site are points of Mousterian type, to be used as projectiles. The silicified mudstone is exploited often with Levallois techniques for the production of larger flakes, which are often retouched as sidescrapers. Production schemes followed with quartz are less defined. The faunal remains of level N are composed of deer 74% (Altuna, 1989), with 20% of animals from rocky goat, chamois areas and almost no presence of large bovid and horse remains. An important behavioral feature recognized in Axlor is the intensive use of bone hammers, mainly diaphysis fragments. Nearly a thousand of them have been recovered in the campaigns conducted to date and about 500 have been technically and functionally analyzed (Mozota, 2012). Out of these, 73 are from level N and 92 from level M. These tools are obtained from faunal remains, without strict selection criteria, and without an intentional manufacture, unlike what happens in upper layers. They are mainly used for various lithic retouching techniques. At these levels, especially in level N, the presence of abundant and well-preserved combustion structures is to be noted. The latest part of the sequence corresponding to the final Middle Paleolithic (levels D, C and B) has a quite different tool management strategy. Most of the material is flint (over 60%) followed way behind by silicified mudstone and quartz (between 10% and 15% each material). At these layers the flint comes from three different sources, the coastal Flysch, the Urbasa mountain range, and the Treviño outcrops (González Urquijo et al., 2005). There are hardly any cores or evidence of block knapping. Large flint artifacts are imported to the site. These are obtained through a Quina-type production of thick implements displaying a dorsal surface. These flakes are retouched as Quina sidescrapers at the site, where they are heavily used and reshaped to reach very small sizes at the time they were abandoned (Fig. 2). Some of the rejuvenation flakes, specifically the largest, appear to be the result of an intentional, or at least a preferential, selection to produce a new generation of tools. As has been demonstrated through the analysis of a large sample of retouched artifacts from level D (n = 917, Frías, 2013), the most abundant type–about 75%– are the sidescrapers, which are very often double or multiple. The final retouched tools are very small in size. This is the result of the intensive exploitation and the use of rejuvenation which account for about 30% of the retouched implements. Unlike what was initially observed with a small sample (n = 50, Rivers, 2007), the average size of the final sidescraper reached only 2.3 x 2.2 x 0.8 mm and many of them are less than 2 cm in any their two dimensions. The bone hammers are very abundant at these levels. According to the analyzed sample (over 200, Mozota, 2013), these are obtained after a careful selection of formats, or perhaps an intentional production of fragments during big mammals food processing. Its use is more specialized (with variations in the types of tools and predominant use of retouching tools in “Quina” tasks). In level B, these are especially massive, being adapted to the extraction of the bigger rejuvenation flakes. The faunal remains of levels D, C and B are far more diversified, showing a progressive increase in the presence of horse and bovids to the detriment of deer. In the most recent layer of the Middle Paleolithic –level B–, deer, goat, large bovids, and horse are distributed in almost balanced shares in the spectrum of ungulate species (Altuna, 1989; Castaños, 2005). Balance Most of the occupations at the Axlor site occurred during the Middle Paleolithic and the most intense took place along OIS 3.The behavior that is best reflected in these occupations during OIS 3 is the variability in the technical organization of the Neanderthal societies of the period. This variability can be observed in the range of hunted animals, the formation and use of the bone industry, the management of lithic raw materials –with highly variable percentages for the different types of rocks and flint deposits in different regions–, and in the repertoire of knapping techniques or deductible territorial mobility patterns.
48  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  a var...
CANTABRIAN MOUNTAINS AND COASTLINE. Manuel R. González Morales* Los Azules cave (Cangas de Onís, Asturias) Location and description: Los Azules cave is located on the southern slope of Mount Llueves, today a dense forest, in Contranquil, Cangas de Onís, roughly at 30 m. above the current course of the Sella River. Its name comes from the farm property on the riverbank. It consists of a complex of small cavities on the same face of Cretaceous Albian-Lower Cenomanian limestone outcrops. The cave that has been exca- Figure 1. Azilian harpoons from Los Azules cave: 1-2: Level 5; 3-4: Level 3g; 5-9: Level 3f. (According to Fernández-Tresguerres and Junceda 1994) * vated have two mouths –originally defined as Los Azules cave I (west) and II (east)– separated by a rocky pillar, which open to a single inner space. When discovered it was almost completely filled with sediment, with a thick top layer dumped by landslides from the slope, as was the case with the other adjacent cave mouths of the complex. Discovery and excavation: The archaeological site was discovered accidentally in 1971 by Alberto Blanco Castaño and Francisco de la Roz Soto, who found an Azilian harpoon and some quartzite and flint flakes. Formal excavation began in 1973 under Juan Fernández-Tresguerres after finding that the site had been plundered by clandestines. In the first two campaigns became clear that the site contained a wide range of Azilian occupations and extraordinary materials, including several characteristic harpoons (Fig. 1) and rich lithic industry. At the end of the 1974 field season, the discovery of bones from a human foot on the edge of the excavated area led to a larger-scale excavation in 1975, which exhumed a burial (Fig. 2) in the Azilian level, with exceptionally well preserved remains from different skeletal parts. Subsequent excavations until the mid-90s extended the excavated area to the inner area of Los Azules cave II, and test pits were dug on the platform outside, in front of each mouth and on the bench between them. Instituto Internacional de Investigaciones Prehistoricas de Cantabria, 39005 Santander, Spain 49
CANTABRIAN MOUNTAINS AND COASTLINE.  Manuel R. Gonz  lez Morales   Los Azules cave  Cangas de On  s, Asturias   Location a...
50 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 2. Plan of Los Azules cave burial (As per Fernández-Tresguerres 1976, sketch by Lorenzo Arias Páramo). Stratigraphy: The infill of the two cave mouths shows a relatively turbulent sedimentary history, with successive phases of water erosion and infills of clay from landslides, interstratified with clast episodes facilitated by exposure to external agents and the poor quality of the rock. The cultural sequence described by the excavators, with the exception of the aforementioned disturbances, is as follows: Level I: Sterile. Level 7 (layers a-c): Magdalenian. Level 2: Late Azilian. Level 8: Magdalenian. Level 3 (layers a-h): Late Azilian. Level 9: Magdalenian. Level 4: Sterile. Level 10: Magdalenian. Level 5 (layers a-b): Early Azilian. Level 11: Magdalenian. Level 6: Late Magdalenian. Level 12: Magdalenian. The cave was almost completely filled up at the time of its discovery, and the surface layer formed a mound at the entrance. This is most probably due to a clay flow slopedown slope that ultimately fossilized the cave. The upper layers of level 3 (3a-d) had a limited distribution at the entrance and outside the cave, while the lower layers (especially 3d2 and 3e) extended throughout the cave and contained abundant ash, organic matter and evidence of intense human activity. From layer 3f onwards, occupations were found towards the back of the cave, also with large amounts of ashes. Level 4, a period when the cave is not occupied, consists of yellowish clay with abundant angular limestone clasts across the entire area of the cave. Level 5, like the lower beds of Level 3, appear at the rear of the cave, filling a large erosive depression that affected Upper Magdalenian levels 6-9. These levels experienced both the erosive effect of water flowing deep into the cave and also the intense use of the space by Azilians, who dug ditches, pits and various other structures. The deepest levels (10-12) have been tentatively assigned to the Magdalenian, without further details available at present. Various structures have been found here as well, including layers of cobbles, enclosure walls, caches of ochre and others.
50  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figur...
CANTABRIAN MOUNTAINS AND COASTLINE. Lithic and bone industries: According to Fernandez-Tresguerres, the raw material management pattern was characterized by an increased selection of materials located near the site throughout the Azilian period. Less common in the Early Azilian, this trend became more marked during the ‘Azilianization’ of the lithic industry. The classic Azilian progressed from careful selection to the exploitation of areas closer to the site. The Magdalenian levels showed an almost constant use of higher quality flint, a tendency that persisted into the Early Azilian. There was a massive accumulation of knapping debris in the classic Azilian, indicative of constant lithic manufacture during the occupation, including a significant presence of ultra-local quartzite collected from the banks of the Güeña and Sella Rivers, although it had limited use in the manufacture of tools, primarily limited to substrate types (denticulates, notches and retouched flakes), with a smaller number of endscrapers made from the same material, and almost no bladelets or points. Flint, on the other hand, was widely used to manufacture endscrapers, bladelets and points, with a predominance of mediocre or poor quality radiolarite. Higher quality flint is present, but in much lower proportions than in previous periods. Fernández-Tresguerres states that the retouched tools in the Early Azilian at Los Azules cave show an industry that is well defined by “small elongated double backed points, in some cases tending to be straight and sometimes showing flat retouch on the distal dorsal face, reminiscent of the shape of late Sauveterre points”. Their appearance was preceded by very short, thick points. “The rest of the industry is less distinctive from the Upper/Late Magdalenian, where small, somewhat rounded endscrapers and abundant backed bladelets are observed, often linear and in some cases double backed. Burins become less frequent. Denticulates and notches are as common as in any of the levels of this site”. What is most noteworthy about Los Azules, however, is the bone industry, particularly the 105 Azilian harpoons. The sequence in this cave reflects the evolution of harpoon types from the oldest classic Azilian items (level 3 h), usually with more barbs than their more modern counterparts, in which the perforation tends to be closer to the centre of the base, later shifting towards the centre of the shaft in the central layers of level 3 (g, f and e), then returning to the centre of the base. The four harpoons from Level 5 (early Azilian) are worthy of special mention. Two are complete, one of them has a round perforation at the base – lacking in the other one–, both of them infrequent features in harpoons from this period. However, the outstanding aspect is the decoration of one of the complete harpoons –an extraordinary piece with seven barbs– and also one of the fragments, an exceptional feature for this period, specially considering the particularities of the decoration. Both items share the same decoration, based on oblique lines with short strokes connected to them, a pattern also found in decorative pendants that are chronologically midway between the late Magdalenian and the Azilian. In the case of the complete item, the initial decoration was covered by another motif based on strips of two parallel lines with the space between them filled by short oblique etched lines, a motif which also extends to the barbs. This second decorative phase almost exactly matches the decoration on a similar harpoon fragment found in La Lluera cave, more than 60 km away. In addition to the harpoons, excavations of the upper levels of the classic Azilian unearthed an sagaie decorated on most of its central flank with series of short oblique lines, and a spatula made from a heavily polished deer metapodial –retaining part of the articulation– with a blunt point with one side covered with finely etched lines of dots. Azilian burial: During the controlled excavation of a burial in 1975, it was found to contain not only human remains but also a remarkable range of grave goods. The body had been laid on its back in a shallow depression along the west wall of the cave, dug into layers 3b-d. The base seemed to have been dusted with ochre, with several pebbles delimiting the right side. A large limestone slab was placed on the knees of the deceased, and the whole grave had been covered with a pile of stones and earth. This adult male, more than 40 years old and 1.75 m tall, had possibly suffered bone disease from an early age, which seriously hindered his ability to walk. According to Fernandez-Tresguerres, he could made a minor contribution to the group’s economy, yet he reached a relatively advanced age –evidence of a strong sense of group solidarity– and was given a unique burial, perhaps due to a specific 51
CANTABRIAN MOUNTAINS AND COASTLINE.  Lithic and bone industries  According to Fernandez-Tresguerres, the raw material mana...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 52 role that he may have played in this social group. Several items recognizable as grave goods were placed around or over the corpse: stone tools, two harpoons, lithic raw material –cores and hammerstones– and deer antler fragments. Two piles of large, carefully stacked Modiolus sp. shells and a badger skull were placed beside the left leg. An even more striking complement was a series of pebbles painted with black points, several of them delimiting the head of the grave. This complex burial, an exceptional discovery for the Azilian in southwestern Europe, is one of the few elements that provide a glimpse of the spiritual and social world of the last Palaeolithic hunter-gatherers. Conclusion Los Azules is undoubtedly the most important Azilian site on the Iberian Peninsula and a key loca- José Adolfo Rodríguez Asensio* Cabo Busto: A Middle Pleistocene Site Cabo Busto is located on the edge cliff of the Western Asturian wave-cut platform and was excavated by J.A. Rodríguez Asensio between 1993 and 1997. Two levels of human occupation have been identified in the stratigraphic sequence, level II and level V. Both belong to the Acheulean period but correspond with two different chronological spans, separated by several thousand years. The Cantabrian wave-cut platform (locally named Rasa) is a flat coastal 5 km wide platform extending from the mountain foot hills of the ancient coastline to the actual sea cliff that is situated in this area, 60 metres above sea level. This platform descends from Cape Peñas, where it is * tion for the definition of the Late Magdalenian/Azilian sequence in southwestern Europe. Still pending the detailed publication of its stone and bone material and information about its sedimentology, fauna and environment, it is nevertheless clear that the series of harpoons –more than the sum of all those found at the rest of the Azilian sites in Iberia– can play a major role in determining the evolution –and extinction– of this particular technology, and act as an extraordinary basis for comparisons. The possible excavation of the Magdalenian levels or the extension of work on the Azilian layers at some point in the future would undoubtedly provide a unique source of information for research into this late Pleistocene period. The richness of materials at the site and the diversity of structures linked to the use and habitation of the cave will make a valuable contribution to our understanding of the lifestyles of the hunter-gatherers at the end of the Palaeolithic in the transit to the Mesolithic. 100 metres high, into the coast of Burela, the result of the geo-morphological process of tilting. Altogether, this sector provides one of the most relevant habitability areas at the moment of the first human population arrival to the North of the Iberian Peninsula. Several rivers in the region flow into this platform. More precisely, along their fluvial terraces, different settlements from this period have been documented. Numerous characteristics provide this area with excellent habitability conditions: it is an area where movement is easy and it offers a very rich hunting area with plenty of water. Some of the most important and well-known Lower Paleolithic sites such Área de Prehistoria. Dpto. de Historia. Universidad de Oviedo. Campus de El Milán. 33011 Oviedo. adolfo@uniovi.es
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  52  role ...
CANTABRIAN MOUNTAINS AND COASTLINE. chopping tools and a minimal presence of small flakes and cores. They are primitive type materials made with elemental technology and poorly defined shape. Based on the chronological interpretations of other similar fluvial deposits in the area dated around interglacial MIS 11, and the recent findings of lithic assemblages, a minimum relative chronology can be established for level V of Cabo Busto, corresponding with MIS 13-11, between 500 and 300 ky, during the Acheulean period. Figure 1. Stratigraphy of Cabo Busto excavation showing the two archaeological levels, II and V. as Bañugues, Tenrero, and Louselas are located at this coastal platform. The Esvariveror Canero river, running along the lower part of the rasa before flowing into the Cantabrian Sea, is responsible for the transportation of the geological deposits where the oldest human occupation remains have been found. Thus, it seems quite likely that this archaeological finding would not have been in situ, but that they have been swept along the floodplains. The earliest level, or Busto V, has been ascribed to the Middle Pleistocene (Fig. 1). It is the oldest archaeological level documented in Asturias, and represents the first human occupation in fluvial terraces of the Cantabrian region. In the first archaeological fieldwork, the technotypological sequence could not have been defined due to the scarce number of lithic items. However, in a recent rescue excavation carried out in 2013 due to refurbishment in a nearby pond, a large collection of lithic materials were retrieved and the old archaeological stratigraphy was also identified. The lithic assemblage, still under study, is composed primarily of massive pieces of end-products, crude bifacial tools, large retouched flakes without a clear typology, Once these deposits settled down, ponds emerged in the landscape and, precisely at the edge of the water reservoir, remains of human occupation have been found. This evidence corresponds to level II in the archaeological sequence of Cabo Busto and techno-morphologically have been dated in the Upper Acheulean, that in Asturias region has a relative chronological span corresponding to the Riss-Würm interglacial (Fig. 1). Level II offers different lithic assemblages made of another type of quartzite, the Ordovician quartzite, the most common variety together with the fine-grained sandstone that was used for lithic reduction in the Acheulean, both in this site and the whole settlements in Asturias. This type of raw material produces crude and primitive artefacts, in contrast to fine-grained quartzite, it is extremely tough and very difficult when knapping, but that produces very accurate tools. Typologically and technologically, the lithic collection is more elaborated and has a better shaped finish. Among the tools we can identify: a broad set of medium size subtriangular or amygdaloidal shape bifaces, also a broad and rich collection of cleavers knapped from large flakes that predetermined their morphology, very few trihedral picks, and a significant set of small size flake tools, like notched and denticulate flakes, a wide number of scrapers, Mousterian points, and, even if they are scarce, some tools obtained using llevallois techniques, like piercing tools, knives and retouched artefacts. The few core are just pebbles, although some prepared core also appear. The paucity of any kind of pebble tools proves that it was employed in a specific chaîne opératoire aimed at producing bifacial artefacts and primary flake-blank. The small size of the cores and the low frequency of preparation techniques reinforce that fact. 53
CANTABRIAN MOUNTAINS AND COASTLINE.  chopping tools and a minimal presence of small    akes and cores. They are primitive ...
54 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD ping intentionally. Closely related with the above mentioned uniformity comes a low functional specialization. Most of the lithic tools are polyvalent and might have been used in multiple, various, and sometimes opposed functions. These lithic assemblages can be better described as polyfunctional artifacts, although some groups such as notches and denticulates could point out certain degrees of specialization. Figure 2. Chaîne opératoire of Cabo Busto site. The sourcing of raw materials would be available at the same platform where outcrops of quartzite and sandstone can be found. Also, although exceptionally, the nearby beaches can provide large pebbles and blocks for producing thick tools such as bifaces and cleavers. The chaîne opératoire of Cabo Busto is simple, highly uniform, and with a low functional specialization. Its simplicity can be observed in the technical processes of production since the shape of pebbles determines the type and morphology of the tools. Concerning the high uniformity, the low technical and morphological variability indicates low diversity. Types are repeated and the unique variations are the result of the raw material characteristics more than the product of knap- Therefore, the chaîne opératoire of Cabo Busto site is characterized by the immediacy of the strategies in the knapping process and the use of tools, and by the localism of raw material procurement and distribution strategies. Immediacy and localism can be observed in the sourcing, knapping and use of lithic artefacts. The catchment area of cobbles and pebbles can be found at the same site. Once the large blanks are transported to the campsite, flake debitage and finishing tools are used at the campsite as well. Finally, tool use would have also occurred at the campsite and we think that most of the tools were produced to obtain new tools made not of stone but of perishable materials such as wood. In short, at the Cabo Busto platform, an archaeological sequence with two different prehistoric levels has been identified: Busto II and Busto V. The first one corresponds to the Upper Acheulean and presents similar characteristics to other known sites at the Asturian wave-cut such as Bañuges, Tenrero, and Louselas. The second level represents, so far, the only case of Middle Pleistocene human occupation at the Northern Iberian Peninsula. Also, the different raw materials employed allows discernment of both levels as unambiguously different episodes, being fine-grained quartzite used in the early occupation and Ordovician quartzite used by the Acheulean population (Rodríguez Asensio, 2001).
54  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  ping ...
CANTABRIAN MOUNTAINS AND COASTLINE. F. Bernaldo de Quiros * A. Neira Campos * J.M. Maillo Fernández ** El Castillo cave 1. Introduction El Castillo cave is in the Puente Viesgo municipality (Cantabria), on the hill of the same name near the La Pasiega, Las Monedas and Las Chimeneas sites. This series of decorated sites makes Castillo Hill the largest known set of Palaeolithic art. The cave was discovered by H. Alcalde del Rio in 1903. Archaeological sediment covered the entire current entrance area, forcing early explorers to enter on their hands and knees (Fig 1). A H. Alcalde del Rio’s first excavations unearthed the Magdalenian levels. A subsequently visit by Prince Albert I of Monaco led the recently established Institut de Paleontologie Humaine to commission H. Breuil and H. Obermaier, along with H. Alcalde del Rio, to conduct new excavations from 1910 until 1914. During this time, they discovered a long stratigraphy ranging from the Lower Palaeolithic to the Metal Ages, one of the longest known sequences in Europe. This succession of levels was largely responsible for H. Breuil’s first outline of his subsequent subdivisions of the Upper Palaeolithic. 2. Exploration of “El Castillo cave” El Castillo cave is one of the longest known archaeological sequences, with a complete succession of Palaeolithic occupations which have been subdivided into 25 units. These include Lower and Middle Palaeolithic archaeological units (Mousterian Unit 20 to Unit 26) and all the complexes of the Upper Palaeolithic: one * ** Figure 1. Plan of “El Castillo” cave. Transitional Aurignacian (Unit 18), Early Aurignacian (Unit 16), two Gravettian (Unit 14 and 12), one Solutrean Unit 10), one Cantabrian Lower Magdalenian (Unit 8), one Upper Magdalenian (Unit 6) and one Azilian (Unit 4). The total depth of the sequence was calculated by H. Obermaier to be 18 to 20 m (Fig. 2). The stratigraphy included “sterile” archaeologically interlayers which isolated the series of occupations from each other. Obviously one of the problems for the initial study of the site was the system used in 1910/14. Contemporary documents show that H. Obermaier collected the material in geological strata which contained the Área de Prehistoria, Universidad de León, Campus de Vegazana, León. 26071 fberg@unileon.es Dpto. Prehistoria y Arqueología, UNED. C/ Paseo Senda del rey, 7. 28040 Madrid. 43°17’32”N, 3°57’53”W Zona 30 X 421.678 Y 4.793.734 55
CANTABRIAN MOUNTAINS AND COASTLINE.  F. Bernaldo de Quiros   A. Neira Campos   J.M. Maillo Fern  ndez     El Castillo cave...
56 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Unit 16 Figure 2. Stratigraphy of “El Castillo” cave. same type of sediment and also archaeological vestiges corresponding to a particular cultural unit (Obermaier, 1914/1925). As a result, some layers such as the Azilien and Solutrean pose no difficulties for analysis, while others evidenced a high density of occupations, which in a modern excavation would represent a process of gradual industrial transformation within certain cultural units. This is the case of Upper Magdalenian Unit 6. When H. Obermaier sketched out the stratigraphy, he clearly showed two levels of “hearths”, as confirmed by Oxford AMS dating (see below). 3. New excavations After analysing the documents and materials from the old excavation (Cabrera 1984), in 1980 we began to clear the debris and vegetation from the cave entrance. Since then, we have continued to study the stratigraphy and the site with an interdisciplinary team led by V. Cabrera and F. Bernaldo de Quirós, since 2004 by the latter. After uncovering what was left of the site, we found that the layers with the best potential for complete analysis corresponded to the sequence ranging from the first Upper Palaeolithic occupations, Obermaier’s Units 16 and 18, to the base. This 5 m high series spreads out from the cave, and contains a high density of materials. Moreover, it is isolated by 40 cm of silt from the Middle Palaeolithic occupations. The material culture in Unit 16 is small in number but in qualitative terms can be included without difficulty in the Archaic Aurignacian technocomplex. The lithological set is predominated by fine-grained quartzite, followed at a great distance by flint, coarse-grained quartzite or quartz. There are two sets of lithic technology systems: laminar and flakes. Laminar schemes are divided into prismatic operative schemes, and carinated scrapers and carinated burins (Cabrera Valdés et al., 2002). As in other Archaic Aurignacian complexes in the region, there was a continuity in the exploitation of prismatic cores, which began with the production of blades and ended with bladelets. Both items are rectilinear with little curvature, and in the latter case, lack torsion. This feature is the same in the bladelets produced by the other above-mentioned schemes. In the case of the most common scheme, carinated scrapers, the extraction surface is broad and hence the supports are broad, straight and lack torsion. Flakes are scarce but characterized by a discoidal type of production. Typologically, Dufour bladelets (standard, with quite minor or denticulate retouch) account for 26% of the total, followed by substrate items (14%). The scrapers are Aurignacian, and one busqué type of burin stands out. Unit 18 Stratigraphically, Unit 18 was subdivided into 18a, 18b, 18c. It lies between two sterile (17 and 19) units which are the result of two collapses of the cave cornice. Unit 19 seals Unit 20 (Middle Palaeolithic). It consists of a large fan of big blocks which form an external buttress on top of which are loam-sand clays –yellowish brown with horizontal furrows, in some cases due to runoff– marked by discontinuous layers of gravel and sand in the same clay matrix. Above this sediment are levels 18b and 18c of variable depth, depending on the zone in the cave. Level 18c appears primarily in the longitudinal section. It consists of black sub-horizontal materials with very thin layers (<1 cm.) of charcoal but no evidence of thermal alteration or reddening. This has led us to suggest that it was formed by material cleared from hearths in other parts of the cave (Cabrera and Bernaldo de Quirós, 1984, Bernaldo de Quirós et al., 2008,
56  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Unit ...
CANTABRIAN MOUNTAINS AND COASTLINE. 2010). This interpretation is also supported by its marginal location in the cave and the high proportion of lithic microdebris in comparison with the items discovered. Level 18b consists of a dense concentration of bones along with lithic industry, predominantly limestone, as well as quartzite and sandstone hammerstones and a smaller number of quartzite and flint flakes. The fauna largely consists of cranial elements, jawbones and remains of axial skeletons, which leads us to interpret this as a primary processing site for animal remains, with large, easy replaceable tools. Limestone is ideal raw material for this purpose. Both levels have a brown clay matrix, characterized by a greater abundance of organic matter and less detrital calcareous elements than Unit 19. They contain medium-sized angular limestone blocks, scattered or in irregular groups, resulting from para-sedimentary rockfalls. The sterile level 18a, possibly indicative of the point prior to the rockfall, sits above these levels. The lithic industry found in these levels is consistent with the material collected by H. Obermaier, although the difference in the occupied areas must not be overlooked. In both sub-levels, the retouching material consists of endscrapers, both simple and carinated, and few burins (primarily dihedral). Material found in the substrate such as sidescrapers and denticulates is also important. Aurignacian blades on laminar flakes (common in the Cantabria region) are also present. Raw materials include different varieties of quartzite and flint. Quantitatively, however, the predominant material is black Jurassic limestone, exogenous to the cave, found in large boulders in the surrounding valleys. This material has a high degree of alteration, particularly in excavated sector 18c. In 18b, it is almost all in the form of debris, cores and flakes amongst faunal remains. However, some of the least altered sectors of level 18c have yielded a carinated scraper and a dihedral burin, permitting the supposition that part of the abundant material made unrecognizable by subsequent alteration may have been characteristic tools. Technologically, lithic production is predominated by discoidal operational schemes, with two well defined methods: unifacial and bifacial. Débitage began with the extraction of cortical flakes in two directions: chordal and centripetal. It is curious to note that the negatives of the thinner cores not as sharp as the thicker ones because secant exploitation is not possible in the former, and are thus produced in a sub-parallel direction to the crest separating either side of the core. This makes the final shape of the core quite similar morphologically to recurrent centripetal Levallois cores. The technique used throughout the débitage sequence is direct percussion with a hard hammerstone. A more discrete laminar operational scheme has been identified in the case of bladelets and flakelets from burin-like or unipolar pseudo-prismatic fine-grained quartzite cores. Blades were produced from limestone using prismatic and carinated burin schemes, and to a lesser extent from carinated scrapers. The technique was direct percussion with hard and also soft hammerstones. The bone industry is scarce but significant. On level 18c we found two distal fragments of staghorn speartips, a bone fragment fishhook resembling those found on the Aurignacian levels, and an awl on a horn flake. In addition, an antler handle was discovered on 18b (Tejero et al., 2005; Tejero and Bernaldo de Quirós, 2008) (Fig. 3). The discovery of this bone industry enables these levels to be linked to H. Obermaier’s digs, in particular his Aurignacian D (V. Cabrera’s Unit 18) where the set of ten spears and several bones with marks was found (Cabrera Valdés, 1984). Level 18c also yielded evidence of symbolism with a distal fragment of a chisel bearing a series of short, rectilinear incisions on the left edge of the upper face, oriented transversely to the longitudinal axis of the item (Cabrera et al., 2001). Figure 3. Handle from level 18b 57
CANTABRIAN MOUNTAINS AND COASTLINE.  2010 . This interpretation is also supported by its marginal location in the cave and...
58 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD We also found a mesial fragment of an ungulate metapod bearing a series of incisions on the upper face: three deep marks with an irregular contour, two of them parallel and perpendicular to the longitudinal axis of the item, while the third is in a divergent, oblique direction. More interesting is a flat bone fragment with lines painted on its upper side which form a figurative representation, interpreted as an animal head, facing the right side of the preserved fragment. SEM composition analysis detected the presence of natural graphite. Level 18b includes several items, most notably a proximal fragment of a hyoid bone, possibly from Cervus elaphus, with lines scratched and painted in black on its upper face (Cabrera et al., 2001; Tejero et al., 2005; Tejero et al., 2008). The theme has been interpreted as an animal foreleg. Analysis of the pigments in the painted lines indicates the presence of manganese, suggesting that it was drawn with the same manganese “pencil” used for the incision which left the marks found inside the groove. Interestingly, this is not a unique case. The use of instruments to scratch and draw at the same time has been detected in several figures at the cave in Chauvet. The same level has yielded a triangular sandstone flake with four lines etched on the flattest surface of one of the faces, while the dorsal face has a natural concavity. The incisions have a U-section and seem to have been made with the thick edge of a stone tool. The two levels of this Unit have been attributed to a “transitional Aurignacian”, an industrial complex which for us is the oldest phase of the Upper Palaeolithic, comparable to others such as Châtelperronian, Jermanovician, Bohunician, Neronian, etc., regardless of the human species which may have produced them. Many Mousterian elements are still present here, but new traits appear, possibly indicating the presence of groups of modern humans and the ensuing crisis in the Middle Palaeolithic. These complexes may therefore be considered as “transition industries” which show different local solutions to an early presence of new populations. sionally characterized as Mousterian. One characteristic of its industry on almost every level is lithic production using discoidal schemes in bifacial and unifacial modes. The Levallois methods are present on some levels such as 20e, focused on the production of laminar flakelets or bladelets (Sánchez Fernández and Bernaldo de Quirós, 2008). There is also a small output of blades from Levallois cores and also from unipolar cores. Some of these cores bear evidence of alternating bifacial retouch. The lithic industry on level 20e is characterized typologically by an average index of sidescrapers, few denticulates and little Quina retouch. It can be classified as standard Mousterian. Our new excavation work has also found similar cleavers on flake to those unearthed in the old digs. These items led F. Bordes to describe the Mousterian levels of this and other Cantabrian assemblages as vasconian, although the entity of the facies was refuted for the Spanish sites. Unit 21 Although the study of this Unit is still in a preliminary state, it has yielded one item which can be interpreted from a symbolic perspective: a 5.7 cm long quartzite pebble decorated with a line of four pitted points on its outer face and a fifth one above them in the centre of the line (Cabrera et al., 2004, Bernaldo de Quirós, 2006) (Fig. 4). The nature of the pittings precludes any functional interpretation. Unit 20 Unit 20, still under analysis, has been divided into 20 a/b, 20c, 20d and 20e. It can be provi- Figure 4. Decorated pebble from Level 21
58  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  We al...
CANTABRIAN MOUNTAINS AND COASTLINE. 4. Chronology One of the fundamental missions in our review of the cave was to define a time frame for each of the cultural assemblages in the stratigraphy. We now have datings for almost all the units in the cave, mainly from bones sampled from the same cut during the 2002 clearing process, or the material at the American Museum of Natural History, where several boxes of sediment samples (including archaeological material) were archived in the 1920’s. Four us, this collection is a time capsule, since the attributions of the archaeological units were defined by Obermaier himself. In other cases, especially in Units 18 and 20, samples were taken during our excavations. For the Upper Magdalenian, present in Unit 6, datings by I. Barandarian for art objects are 10,310 ± 120 BP for the upper sub-level and 12,390 ± 220 BP for the lower sub-level (Barandiaran, 1988). However, a rhinoceros remain from this Unit has been dated at 31,800 ± 600, showing that fossil collection is older than it might seem (Bernaldo de Quirós et al., 2006). For Unit 8, attributed to the Cantabrian Lower Magdalenian with its characteristic scapula etched with deer heads and strong connections to rock art, there are two datings, both unpublished. One is a bone fragment dated at 15,540 ± 70 BP (Beta 242618) found in the stratigraphy during the 2002 clearing process, while the other is from a selected scapula fragment held at the IPH, which yielded 15,160 ± 70 BP (Beta 242620). Both dates are close to those obtained at other sites from scapulae etched with deer heads. To learn more about the Unit 10 dates, we sent a sample from the American Museum of Natural History collection to BETA Analytic, which returned a date of 19,260 ± 90 (Beta 242619). The same process was repeated with the Gravettian units. The dates for Unit 12 ranged from 24,070 BP to 25,920 BP and for Unit 14, 29,600 BP and 29,740 BP (Bernaldo de Quirós et al., 2012). Due to complexity of Units 18 and 20 and the implications for their organization, a series of extremely important datings for the reinterpretation of the Middle-Upper Palaeolithic Transition were done. Most were in AMS C14 and ESR at three laboratories: Tucson, Oxford and Gif-sur-Yvette (Cabrera Valdés and Bischoff 1989, Cabrera Valdés et al., 1996). The specimens were taken from different zones of the site and different excavation seasons, all independent from each other. Material from the National Archaeological Museum in Madrid and the AMNH was also dated. Despite the relatively large number of samples, the diversity of laboratories and the methods used, all results are quite consistent. We currently have over 20 datings between 40,000 and 45,000 BP for Unit 18. We also have C14 datings for Unit 20, most ESR. All are staggered between 41,000 and 49,000 BP (Liberda et al., 2010). For Unit 22 there is one ESR dating at 59,100 BP, and for Unit 23, a stagmitic crust which seals the base sections (Units 24, 25, and 26), there are two: 89,000 and 92,200 BP (Rink et al., 1995, 1997). 5. Resource management The management of the resources used by the populations who occupied El Castillo cave is another of the interests that have driven our work. The possibilities afforded by a site of this nature cannot be ignored. One of the research lines at the site has been the changes in resource and land use. We first did an analysis of the fauna seasonality, since the information about the season when the animals were hunted down and their age is taken from the growth marks on their teeth. For the El Castillo site, we selected 159 deer teeth (the most numerous species) from levels 18b, 18c (Transitional Aurignacian), 20a, 20b and 20c (Mousterian with cleavers). The most relevant results are that during the Mousterian, animals were captured from late autumn until spring (Pike Tay et al., 1999). During the Transitional Aurignacian, they were hunted from winter and throughout spring. All ages are present in both cases, with more young adults, suggesting similar fauna resource management solutions in both cases, centred on the winter, which is consistent with an aggregation model in which the individuals in the group shared their resources at this time of year. Several studies of the fauna are currently under way to check these results and expand our knowl- 59
CANTABRIAN MOUNTAINS AND COASTLINE.  4. Chronology One of the fundamental missions in our review of the cave was to de   n...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 60 edge about the lifestyles of the human groups at this exciting time. 6. Conclusions The El Castillo cave site is one of the most important records of the Middle and Upper Palaeolithic on the Iberian Peninsula and indeed in Europe. The presence of a stratigraphy represent- Mario Menéndez*, Gerd-Christian Weniger **-***, David Álvarez-Alonso1, María de Andrés-Herrero ***,Eduardo García *, Jesús F. Jordá *, Martin Kehl ****, Julio Rojo *, José M.Quesada *, Isabell Schmidh ** La Cueva de la Güelga. Cangas de Onís. Asturias Introduction La Cueva de la Güelga, whose name in the local language refers to wet and shady sites, opens to the heart of a limestone mountain valley, forming a cul-de-sac. A stream flows from the current cave aperture and has configurated a karst system with corresponding terrace drain caverns that were successively occupied during the Middle and Upper Paleolithic. This group of rock shelters and caves has been divided into different sectors for investigation, which has developed from 1989 to the present. A-B and C areas are located on the lower terrace, occupied during the Magdalenian and Solutrean. At the top is D sector, with occupations attributed to Châtelperronian, Aurignacian and Mousterian. This valley, closed in itself, has provided numerous lithic remains in surfaces, mostly attributable to Mode 3, surely exponents of intense and prolonged occupations. It is located 200 m above sea level, and along with Buxu and Azules caves, is a core site in the middle reaches * ing all stages of human presence dating back more than 300,000 years enables a wide range of working hypotheses to be tested, both historically and in other disciplines (palaeontology, climatology, etc.). Our studies have focused on the transition period from the Middle to the Upper Palaeolithic. They have contributed several aspects that challenge –and indeed will continue to challenge– current views. This confirms the importance of the site and the opportunities it presents. of the Sella River, territorially linked with other coast sites, 15 km away, around the Ribadesella Bay (Menéndez, 2003). Areas A, B and C (Upper Paleolithic): Located around the current cave entrance, they show remains of an intense Solutrean occupation swept by the river into the karst. The only evidence from the upper Solutrean, industry also present in neighboring Buxu Cave, are gap vestiges attached to the wall of the shelter and in situ layer (Area C), with notch points and concave bases. Also, A and C were excavated and assigned to Cantabrian Lower Magdalenian or Magdalenian III occupation, from the so-called Juyofacies. The lithic and especially the bone industry, display the existence of a group of hunters specialized in deer (55%), chamois (24%) and goats (20%), probably during the middle months of the year (spring / summer), which left at layer 3 an excellent collection of art mobilier. The hyoid hanging of deer must be highlighted, decorated assegai Departamento de Prehistoria y Arqueología, Universidad Nacional de Educación a Distancia. Ciudad Universitaria, Paseo Senda del Rey 7, E-28040 Madrid, Spain; mmenendez@geo.uned.es; dalvarez@gijon.uned.es; egarciasmail@gmail.com; jjorda@geo.uned.es; juliorojo@juliorojo.jazztel.es; jmquesada@geo.uned.es; ** Neanderthal Museum. Talstr. 300, 40822 Mettmann, Germany; schmidt@neanderthal.de; weniger@neanderthal.de *** University of Cologne, Institute of Prehistoric Archaeology. Albertus-Magnus-Platz, 50923 Cologne, Germany mdeandres@neanderthal.de, **** University of Cologne, Institute of Geography. Albertus-Magnus-Platz, 50923 Cologne, Germany; kehlm@uni-koeln.de
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  60  edge ...
CANTABRIAN MOUNTAINS AND COASTLINE. and bones emphasizing an adult deer tibia fragment, three heads of the same animal synchronously recorded, all done with fresh bone, but with very different styles and conventions. This occupation of the late Magdalenian is well dated, around to 14 key BP (Fig.1). results (Jordá et al., 2013; Menéndez et al., 2014). We will summarize the current geoarchaeological results and hypotheses in future work. The geoarcheological D sector sequence from La Güelga comprises a series of levels generated by both anthropogenic and natural processes. These natural processes detected by the sedimentological analysis, highlight the gravitational collapse of large blocks, gelifraction and diffuse gullies of very low energy (Jordá Pardo et al., 2013). The micromorphologic analysis of D interior area identifies features that indicate the nature in situ of both the Mousterian (L9) and Aurignacian (L5-L6) levels, whereas in the Châtelperronian levels (L1 and L2) the traits indicated were emplaced by processes of creep after a roof block fall and aren’t significantly compacted by trampling. Chronological invertion seems to confirm this hypothesis. D Interior: The sequence excavated so far consists of nine archaeological layers deposited in slope (Fig. 2), into the cave, under a strong surface layer (S1 and S2) (Quesada and Menéndez, 2009; Jordá et al., 2013). Figure 1. Engraved Magdalenian tibia. Area D (Transition MP / UP): In the middle terrace, around11 m above the current stream bed, appears a cave entrance excavated since 2000, which was filled in by debris from a collapsed cornice that fell and formed a slope. Overall, nine archaeological layers were excavated inside the shelter, which we have called D interior. The result was a Châtelperronian – Aurignacian – Mousterian sequence, separated by periods of collapse and abandonment of the cavern (Quesada and Menéndez 2009). As stratigraphic variations in certain areas were observed and the Aurignacian interlayer was the utmost interest to the transition paradigm MP / UP, in 2005 a new excavation zone was determined to be opened on the outdoor terrace under the large blocks of the old collapsed shelter. This area, which has been called D exterior, provided intense Mousterian occupation. In 2012, in collaboration with the Neanderthal Museum in Mettmann and the University of Cologne (Germany), led by G-Ch.Weniger, micromorphological analysis of D zone was done on interior and exterior levels, to contrast them with the sedimentological Châtelperronian (L1 and L2): Layers 1 and 2 form a sedimentological unit in slope into the cave interior. It was only useful for excavation 3.7 m2. A flint laminar industry was found, having noted the presence of two Châtelperron points, and another assemblage of quartzite flakes, such as scrapers and denticulates. The presence of lithic manufactured the absence of bone artifacts and 14C studies (Table 1) encouraged us to define this set as Châtelperronian, considering the possible underlying Aurignacian as an interstrafication. Recent dating of the lower level (L5) and sedimentological and microstratigraphic analysis carried out by the University of Cologne does not ensure that this level is in situ. Under level 2 a fringe of stone blocks detached from the shelter and a layer of clay and silt appear from the outside. Levels 3 and 4 are almost sterile. Aurignacian (L5 and L6): Under a line of stone blocks (L5) appears a clay layer (L6); shown in situ by the sedimentological and microstratigraphic analyses. This unit has provided a few anthropic remains, although very typical. The lithic assemblage, mostly laminar, is made on flint and quartzite. There are nosed scrapers, one Aurignacian blade and retouched flakes. Regarding bone industry, several flattened oval section awls were found, a moothed mesial fragment of assegai and one deer phalanx whistle. The chronology (pend- 61
CANTABRIAN MOUNTAINS AND COASTLINE.  and bones emphasizing an adult deer tibia fragment, three heads of the same animal sy...
62 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 2. Overall stratigraphy of la Cueva de la Güelga. ing new dates) places it into 38ky calBP. Despite the reduced sample, the homogeneity and conventional characters of the assemblage, the absence of contradictory elements, preliminary dating and stratigraphic position indicate an undoubted Aurignacian presence. Below this layer, a long period of cave abandonment (L7 and L8) is documented. Mousterian: It is represented in D interior by level 9, showing intense human presence evident in combustion remains, wealth lithic industry (Mode 3), with Levallois pieces and animals bones with fleshing traces; as well as the possibility of setting spatial occupation patterns. This inside occupation matches on open air the terrace level with the 4B layer from D exterior. Both have provided a typically Mousterian lithic accumulation, consisting of local quartzite flakes retouched, denticulate and scrapers, as well as Levallois points. All phases of the operational chain are present, mostly discoid and also Levallois. Scarce flints remain, Piloña type show relationships with other sites, such as Sidrón cave in the same river basin. Premolar (15) and several human dental fragments, with Neanderthal morphology, were found. The14C dating with pretreatment by ultra filtration (OxA) places this occupation in the period 55/44 ky in OIS 3c, between H6 and H4 events (Menéndez et al., 2009; Quesada and Menéndez, 2009; Jordá et al., 2013). The fauna recovered, around 70,000 remains, show deer (66%) and chamois (31%) predominance, along with uncommon species and diverse ecosystems, such as mammoth (Mammuthus primigenius), panther (P. pardus), megaloceros, rhino, wolf, boar, etc. suggesting a recurrent and prolonged use of the site by the Neanderthal populations of the River Sella basin. Conclusions 1. The assignment Châtelperronian sediments (L1 and L2) are displaced. Their sedimento-
62  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figur...
CANTABRIAN MOUNTAINS AND COASTLINE. Zone Level Culture Material Procedure Code BP Date Deviation Indoor D 2 ¿Châtelperronian? Bone with marks AMS + untrafiltration COL2014 37429 Indoor D 2 ¿Châtelperronian? Bone with marks AMS + untrafiltration OxA-27958 Indoor D 5 Aurignacian Bone with marks AMS + untrafiltration Indoor D 9 Mousterian Bone with marks Indoor D 9 Mousterian Outdoor D 4b Outdoor D 95% probability calibrate Date CalPal 2007 Hulu INTCAL 13 302 42780 –41460 calBP 42320 –41400 calBP 40300 1200 45910 –42070 calBP 45890 –42090 calBP Beta377233 33610 220 41730 –35570 calBP 38720 –37200 calBP AMS + untrafiltration OxA-19244 43700 800 49020 –44540 calBP 48740 –45300 calBP Bone with marks AMS + untrafiltration OxA-19245 44300 1200 50660 –44380 calBP Out range calibration Mousterian Bone with marks AMS + untrafiltration OxA-20122 47400 2700 Out range calibration Out range calibration 4b Mousterian Bone with marks AMS + untrafiltration OxA-20123 >43200 Outdoor D 4b Mousterian Bone with marks AMS + untrafiltration OxA-20124 48500 3500 Out range calibration Out range calibration Outdoor D 4b Mousterian Bone with marks AMS + untrafiltration OxA-20125 >43600 Table 1. Datations of Cueva de la Güelga calBP, before an intense Mousterian occupation (45/48 ky calBP). logical quality is not enough to defend such a significant hypothesis as interlayer Aurignacian. Future work should pursue an explanation for their stratigraphic position and timing (42/45 ky calBP). 3. There is a long period of abandonment between Aurignacian and Mousterian occupations (7/10 ky). 2. There is an Aurignacian presence, with little information, but with a timeline around 38 ky 4. The lower Magdalenian occupation provided an excellent collection of portable art. 63
CANTABRIAN MOUNTAINS AND COASTLINE.  Zone  Level  Culture  Material  Procedure  Code  BP Date  Deviation  Indoor D  2    C...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 64 Mª Soledad Corchón Rodríguez* La Cueva de las Caldas (Priorio, Northern Spain) 1. La Cueva de las Caldas in the context of the Nalón valley The Oviedo basin, where La Cueva de las Caldas is located, forms part of the central-northern part of the Cantabrian region of Asturias and the western end of the central depression of Asturias. It is a region of Mesozoic and tertiary materials, extending towards the coast, characterized by a landscape of deep karst modelling. This territory forms the known limit of Cantabrian Upper Paleolithic deposits, as Paleozoic rocks spread westwards to the western Asturias-Leon area and no outdoor settlements are known. In terms of geomorphology, the surroundings of Las Caldas correspond to a depression whose evolution is linked to the encasement of the river Nalón. The middle section of this valley, the network of tributaries and streams that flow into the Nalón, have formed a landscape of small sheltered valleys, running laterally to the main valley, in one of which the cave is located, on the right bank of the river. It consists of a karstic complex developed in mountain limestone (Visean-Namurian, Lower Carboniferous), composed of two connected cavities –Caldas I and Caldas II–, structured in a complex network of galleries and channels over 1km long. The main entrance (Caldas I) faces SWW, and its geographic coordinates are 5º 54´ 723´´ W, 43º 20´ 123´´ N, 160 metres above sea level. Towards the exit of Las Caldas valley, 800 m from the site, there are medicinal mineral hot springs that give the cave its name; not far from it, between 5 and 20 km away, another eighteen springs with therapeutic properties are known. At the end of the Pleistocene these circumstances, in addition to the variety of ecosystems created by the significant contrasts in altitude, generated a shelter environment with a wide range of ecosystems in the surroundings of Las Caldas: high mountains (Sierra del Aramo, 8 km ways, with peaks of 1,700 m.), low-lying hills (Peña Avis: 410 m) and sheltered valleys. This data helps to explain the large concentration of Palaeo* lithic sites in the middle Nalón valley: more than 20 caves and rock shelters occupied between the Aurignacian and Azilian, 13 of them with parietal art . In the main cavity (Caldas I) the cave preserves one of the most significant stratigraphic records of the Solutrean and Magdalenianin south east Europe. The site has been open excavated to 24m2, representing all of the occupied topographic units: Sala I, Sala II, Pasillo I and Corte exterior. The stratigraphic sequence covers a time range of c.10,000, from the start of the regional Solutrean (24185 ± 370 calBP, Middle Solutrean) to the Late Magdalenian (14936 ± 342 calBP). The thickest Solutrean stratigraphic units correspond to Sala I, with 17 levels that cover the entire sequence of the Cantabrian Solutrean (level 3 to base 19). Above this series are remains of eroded Magdalenian levels, one of them Late Magdalenian (level 2A). Sala II, however, only preserves one Late Solutrean level (level XIV) resting on the limestone floor, as the oldest levels were evacuated towards sala II when the hypogeum river. A thick filler of 16 Magdalenian levels was deposited on this level, with all of the stadials represented: upper, early and recent middle, upper and late Magdalenian. This deposit covers from the Lascaux interstadial to the Alleröd, and to date the levels there are 26 dating results14C (Table 1). 2. The Solutrean stratigraphy of Caldas I and the outdoor engravings The Solutrean record of Las Caldas is a reference in the Cantabrian region but its conditions of preservation are not the same in the 4 areas excavated. The outer hall, altered by the collapse of the overhang that covered the entrance, only preserves one Upper Solutrean level (level III); the same applies to Sala II, with just one level (level XIV). In the Pasillo and Sala I, although the levels do not always coincide, the record is very extensive: levels 3-18 and 3 – base 19, respectively. Chair in Prehistory. University of Salamanca, Department of Prehistory, c/Cervantes s/n, 37002 Salamanca, Spain. scorchon@usal.es
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  64  M   S...
CANTABRIAN MOUNTAINS AND COASTLINE. Lab. Ref.. C BP 14 CalBP_CalPal 2007-HULU (68% range calBP) Level / Sector Classification Ua-15318 20250 ± 235 (AMS) 24185 ± 370 (23814 – 24555) 15 (Sala I) Middle Solutreen Ly-2428 19510 ± 330 23340 ± 468 (22872 – 23808) 16 (Topera) Middle Solutreen Ly-2426 19480 ± 260 23296 ± 413 (22882 – 23709) 12b (Pasillo) Middle Solutreen Ly-2425 19030 ± 320 22857 ± 404 (22452 – 23261) 12t (Pasillo) Middle Solutreen Ly-2429 19000 ± 280 22843 ± 379 (22464 – 23222) 18 (Topera) Middle Solutreen Ly-2424 19390 ± 260 23199 ± 398 (22801 – 23597) 9 (Pasillo) Upper Solutreen Ly-2423 18310 ± 260 21960 ± 388 (21571 – 22348) 7 (Pasillo) Upper Solutreen Ua-15316 18305 ± 295 (AMS) 21949 ± 412 (21537 – 22361) 11 (Sala I) Upper Solutreen Ua-15315 17945 ± 370 (AMS) 21541 ± 603 (20938 – 22144) 9 (Sala I) Upper Solutreen Ua-4302 17380 ± 215 (AMS) 20837 ± 358 (20478 – 21195) XIVc (Sala II) End of the Solutreen Ly-2422 17050 ± 290 20405 ± 495 (19910 – 20900) 4 (Pasillo) End of the Solutreen Ly-2421 18250 ± 300 21904 ± 424 (21480 – 22328) 3 (Pasillo) End of the Solutreen Ua-4301 15165 ± 160 (AMS) 18324 ± 273 (18051 – 18597) XIII (Sala II) Lower Magdalenian Ua-4300 14835 ± 130 (AMS) 18156 ± 282 (17874 – 18438) XII inf (Sala II) Lower Magdalenian Ua-2735 14495 ± 140 (AMS) 17635 ± 282 (17353 – 17917) XII (Sala II) Lower Magdalenian Ua-2734 13755 ± 120 (AMS) 16881 ± 230 (16651 – 17111) XI (Sala II) Lower Magdalenian Ua-10188 13370 ± 110 (AMS) 16297 ± 436 (15860 – 16733) IX (Sala II) Middle Magdalenian Ua-10189 13640 ± 150 (AMS) 16604 ± 393 (16211 – 16997) VIII (Sala II) Middle Magdalenian Ly-2936 13310 ± 200 16220 ± 475 (15745 – 16695) VIII (Sala II) Middle Magdalenian Ly-3318 12869 ± 160 15571 ± 512 (15059 – 16083) VII (Sala II) Middle Magdalenian Ua-10190 13650 ± 140 (AMS) 16641 ± 363 (16277 – 17004) VIc (Sala II) Middle Magdalenian Ly-2427 13400 ± 150 16314 ± 454 (15860 – 16768) IV/III (Sala II) Middle Magdalenian Ua-10191 13185 ± 155 (AMS) 16114 ± 441 (15672 – 16555) IIIb-IIIc (Sala II) Middle Mag / Upper Mag Ua-10192 12960 ± 190 (AMS) 15775 ± 529 (15245 – 16304) II (Sala II) Upper Magdalenian Ua-10193 12595 ± 125 (AMSA) 14936 ± 342 (14594 – 15278) I (Sala II) Upper Magdalenian Ua-10194 12590 ± 120 (AMS) 14931 ± 337 (14593 – 15268) -II (Sala II) Upper Magdalenian Table 1. Cueva de Las Caldas radiocarbon dates (Weninger, B., Jöris O., Danzeglocke, U. 2007: Calpal – Cologne University Radiocarbon Calibration Package) 65
CANTABRIAN MOUNTAINS AND COASTLINE.  Lab. Ref..  C BP  14  CalBP_CalPal 2007-HULU  68  range calBP   Level   Sector  Class...
66 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Another relevant aspect is the existence of linear and ideomorph engravings on the right wall, next to the entrance and lit by the daylight. The Solutrean engravings are deep linear grooves in a regular series. Their stratigraphic dating in the Middle Solutrean is of additional interest, as they are split by a fracture line and there are large blocks from the collapse of the right wall, which occurred during the deposit of level 15, according to the excavation carried out in Pasillo I. In the middle of the Pasillo area, the Solutrean engravings are replaced by Magdalenian engravings; the most recent, fusiform, overlapping others of fine multiple lines with claviformideomorphs and a female styling. A large quartzite plaquette engraved with a similar fusiform, from the base of the Middle Magdalenian, is a solid chronological reference for the former, overlapping the rest (Corchón et al., 2009a). initial Cantabrian Solutrean (middle of the European sequence) is dated at 24185 ± 370 calBP (AMS, level 15, camber I), 23340 ± 468, 23296 ± 413 and 22857 ± 404 calBP (ordinary C14: levels 16, base 12 and ceiling 12, corridor). 2.1. Characterisation of the Solutrean levels At the beginning of the Upper Solutrean the previous very wet conditions continued (Sala I, levels 12-11); but as it progressed the climate turned very cold and wet (GS-2c), the levels near the entrance present cryoturbation (levels 9-8), and cold steppe fauna appears throughout the section (levels 9-4: mammoth and reindeer). It is dated at 23199 ± 398 and 21904 ± 424 calBP (levels 9 and 7). The Late Solutrean returns dates of 20837 ± 358 calBP for Sala II (AMS, n. XIVc), a similar date to those obtained for the Pasillo (levels 4-3, more altered due to their proximity to the entrance). During this disposition, the climate is extremely wet and cold with flooding processes and partial erosion of the deposits. The excavations reveal that, based on current data, they are the oldest base levels on the Cantabrian coast. They were deposited under very wet and cold conditions in the Late Glacial Maximum (LGM: levels base 19, 19, 18). As the Middle Solutrean progressed, the environment changed to mild and very wet (GI 2), with intermittent episodes of flooding in which the cave was abandoned. These are the conditions recorded for levels 17 to 13 (Sala I) and 17 to 11 (pasillo I). This The fauna in the Solutrean levels is very abundant (84,465 remains), 14,579 of which can be identified; according to the authors of the study (Altuna and Mariezkurrena), it is the richest record of Solutrean remain in the Iberian Peninsula. There are a large number of anthropic alterations (fracture marks related to obtaining bone marrow and stripping the flesh or carving) in all levels, demonstrating that the vast majority of the ungulates were contributed by humans. Figure 1. Karstic complex (Caldas I and II). Stratigraphy of Sala II.
66  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Anoth...
CANTABRIAN MOUNTAINS AND COASTLINE. The dominant species among the macro-mammals in the 17 Solutrean levels is deer (> 65%). This is followed by horse, which at the start of the Middle Solutrean exceeds 20% (level base 19), although its presence is minimum (3-4.6%) in the Upper and late Solutrean (levels 8 and 4-5). Then the mountain goat exceeds and equals the horse in some levels in the centre of the sequence; and the chamois, absent in the Middle Solutrean, increases its presence towards the end (levels 4-5: 18.9%). There are some roe remains in the Upper Solutrean. Regarding bovine species, they are represented throughout the sequence but the low number and fragmented remains do not allow us to determine whether they are aurochs or bison. Reindeer are found at the beginning of the Upper Solutrean (levels 12, 11, 10), and in a level from the middle (level 15); but do not exceed 0,2% of the remains identified. This scarcity is coherent with the general data for the Cantabrian region, where its rarity is perceived in the sites located to the west (Asturias), while the frequency is higher in the territory more to the east (Guipúzcoa). Las Caldas confirms this data, given the richness of its fauna material and the fact that the majority of the sequence corresponds to a particularly cold period (UMG). Due to its relative rarity, 20 remains of mammoth tooth are interesting, which coincide with the most coldest phase of the Upper and Late Solutrean (levels 9, 8, 7 and 3); and fragments of the tusk of this species were used as the raw material for making a rod (level Middle Solutrean18), and portable art (engraved tusk, levels base 10-11; two pendants, levels 9 and 8, Upper Solutrean). The first tooth remains were identified by E. Aguirre and F. Poplin, and later by J. Altuna. In addition to the ungulates already mentioned, the record is completed by remains of carnivores in the Middle and Upper Solutrean (cave bear, fox, badger, marten, leopard and lynx). Even scarcer are marmot, rabbit (a distal end of humorous from this animal shows incisions from removing the flesh) and hare. 2.2. Solutrean level raw materials, technology and industries The subsistence strategies of the Solutrean and Magdalenian groups in Las Caldas are similar in terms of extensive logistics mobility to procure exotic raw materials and exchange cultural items. The Solutrean and Magdalenian portable art used unusual local minerals, such as amber and lignite, to make necklace beads. The study of the lithic raw materials allows the traditional routes travelled and the extensive territo- Figure 2. Upper Solutrean. Layout of the fauna and laurelleaf blades of various sizes (Sala I). ries visited by the Palaeolithic groups in the Nalón in order to gather high quality siliceous rocks to be identified. A. Tarriño has identified fifteen local siliceous rocks in Las Caldas, including flint, lacustrine flint (Cenozoic), Jurassic flint (Mesozoic), carboniferous flint from mountain limestone (Paleozoic), Barrios quartzite and Paleozoic rock crystal. All of these are found in the vicinity of the cave, between 2 and 15km away. In addition to these, the Solutrean and Magdalenian levels show high quality allochthonous flint that constitute real lithological markers, demonstrating that these materials were transported from distant source areas (Corchón, Martínez and Tarriño, 2009). These are Flysch flint, brought to the cave from Vizcaya (Barrica) and the French Pyrenees (Bidache), 150 and 300 km away; Urbasa flint (Navarre), from 380km away and Treviño flint (Álava), 350km 67
CANTABRIAN MOUNTAINS AND COASTLINE.  The dominant species among the macro-mammals in the 17 Solutrean levels is deer    65...
68 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD from the cave. Finally, Chaloss flint (SW France) has been identified, the nearest source area of which is 550km away, representing the most distant Palaeolithic siliceous rock known to date. Later, new allochthonous and local varieties have been identified, one of them very suitable for laminar knapping: Piloña flint transported to Las Caldas from the eastern basin of Oviedo, 40-50 km away (Tarriño et al., 2013). The knapping technology and techniques applied to the Solutrean leaf-shaped blades are known thanks to the discovery of numerous tools in various stages of reduction, which have allowed two operational sequences to be reconstructed, one for knapping laurel-leaf blades and the other for notched points (Corchón et al., 2013a). In the Middle Solutrean (levels base 19 –13, Sala I), from the oldest levels the thin bifacial laurel-leaf blades coexist, often knapped using pressure flaking and frequently in high quality foreign flint (Treviño, Urbasa, Flysch), with thick bifacial leaf-shaped blades with a rounded base and irregular percussion flaking, in many cases discarded in the preliminary knapping phases. A unique case is the use of tools that are barely operative due to their fragility: three leaf-shaped blades and a racloir delicately knapped in rock crystal (level13 and 14), as well as a bifacial laurel-leaf blade in flint associated to a necklace consisting of 7 deer canines at the base of the sequence (level base 19). However, leaf-shaped blades are only abundant in two levels (14 and 18: 18%), being scarce in the rest (3-5.5%). Other characteristic types are points a face plane (unifacial flat retouch points), which are often pieces in progress, and leaf-shaped racloirs for which foreign flint was also reserved. There is an abundance of retouched blades and scrapers (3%-5%), some on Solutrean blades, always prevalent over few and mediocre burins. To summarise, even though the record is smaller than in other levels (440 tools), due to the interruptions and washing caused by flooding, in the Middle Solutrean domination of the technique is noticeable, the laminar trend of tools and the widespread use of Solutrean pressure retouch, which characterises the Upper Solutrean levels, are not fully developed. This is reflected in the many shapeless cores knapped in local quartzite and flint and the high percentages of substrate also show lower technical expertise (>25% recess, denticulate and chipped tools and scrapers) and the diverse tools (40% flake and other atypical retouched tools), in all levels. In the Upper Solutrean (level12-7, chamberI) the record is very rich: 13,125 lithic materials recovered, 1,049 of which are tools (8%) and 96 cores (0.73%). The structure of the common tools does not show any major changes compared to the previous stadial and there are even more scrapers than burins; in all levels, the percentages are similar. Some thick leafshaped blades in quartzite, discarded in the knapping process have been reused as burins. Substrate still has a significant weight (14.6 – 19%), as do retouched blades (4.62 – 9.09%). The high rate of laminar flake is the most characteristic feature, consistent with the abundance of leaf-shaped blades knapped from large flakes and laminar supports. The laurel-leaf, willowleaf, notched points and other leaf shapes in progress range between 43% and 38%, revealing the boom in flat retouch techniques across the section. Laurelleaf blades, unifacial and bifacial, have very contrasting dimensions, with varied techniques for attaching handles-convex, straight, concave, asymmetrical base, revealing a high level of specialisation and their adaptation to different types of prey. As regards the late Solutrean, even though there is a large sample (5,036 remains), it is considered less representative due to the erosion and flooding that occurred during and after the levels were deposited. The tools (608) reach a high percentage (12.07%), but this increase is due to the high rate of substrate. The Scrapers Index/ Burins Index ratio is inverted in favour of the latter; thick scrapers on flake appear and retouched blades are abundant, particularly towards the end (level 3). The number of leaf-shaped blades decreases to 6% (levels 3-6), in accordance with the process of gradually abandoning knapping of leafshaped blades in the late Solutrean in contrast with the Upper Solutrean (39% and 43%, levels 10 and 8, respectively). To sum up, the late Solutrean (levels XIV and 6-3) shows worsening of the lithic types, with few laurel-leaf blades, flakes with fine side retouches and combined tools (scraper-truncated blade, burin-truncated blade). A few racloirs, perforators, denticulate tools and some atypical raclettes complete the tools. Bone types, which are scarce and uncharacteristic of the Middle Solutrean (assegais, rods and engraved smoothers) and of the Late Solutrean (some assegais, awls), demonstrate the creativity in the Upper Solutrean. The assegais show different formats for attaching the handle –bevelled, bi-point, with central flattening–, accompanied by rods and needles. With these, two unique pieces in the Cantabrian Solutrean have also been recovered: a large spear thrower from antler (183.3 x 17.23 x 8.78 mm), unique (level 11c), with a distal hook made by cutting and polishing and with the shaft not entirely polished and with traces of scraping and grooving and a notched point
68  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  from ...
CANTABRIAN MOUNTAINS AND COASTLINE. made of bone from the tibia of a deer, preserved complete (level 12 ceiling: Corchón et al., 2013b). There are also many pendants, made from various materials: ivory slabs (Upper Solutrean), necklace beads in bone, antler, amber, ivory and lignite (Middle and Upper Solutrean), with perforated bones and teeth and incisors in all levels. 3. The Magdalenian sequence in Las Caldas 3.1. The Lower Magdalenian The period between 18324 ± 273 and 16881 ± 230 calBP is occupied by the Lower Magdalenian, an extremely complex episode in the Cantabrian. In La Cueva de las Caldas the sedimentology study shows that it corresponds to the transition between the Lascaux (level XIII) to very cold and wet conditions (levels lower XII, XII, XI), with erosive and solifluction processes and in one case, with the sliding of the sediment section (level lower XII) (Fig. 1). According to Altuna and Mariezkurrena, the ungulate fauna is dominated by Cervus elaphus (>40%) in all levels, reaching 61 and 80% (lower XII and XIII). This is followed closely by Equus ferus, with a large number of remains in all levels. If we assess the meat supplied by the horse, its importance grows in the Lower Magdalenian, reaching 70% in some levels (level XII, XI). However, goat, roe and chamois have a low presence, as do bovine (aurochs-bison). These only have a significant presence in levels XIII and lower XII (4.3 and 4.6% of remains). Obviously these values increase if we look at the meat supplied, as it is a large animal that supplies an abundance of meat. As data of particular interest, level lower XII contained a Coelodonta antiquitatis remain. The Rangifer tarandusis present in levels XII and lower XII (2 and 1 remains, respectively, and one antler from level XII), in accordance with the aforementioned harshness of the environment. Roe appears in both levels (later cf.). Among the nonungulates, wolf, fox and marmot are just as rare. The four levels show industries with similar characteristics (facies of the west of Asturias), with more laminar tools and better fracturing than those in the centre of the region (eastern Cantabria-Asturias: facies Juyo). Of particular interest are the selection of high quality allochthonous flint for laminar knapping, the large number of burins on blades, retouched blades, back edge bladelets and scalene triangles, which exceed 40% of the total in some levels. The bone industry shows various types of assegai –single-bevelled and grooved–, rods and needles and many bone objects show line pairs, series of parallel lines, indents and other engraved symbols. The figurative subject is limited to an antler engraved with deer and a horse with a synthetic design, in level XII (Corchón, 1994). 3.2. The Middle Magdalenian Levels IX-IV of Sala II were deposited over a thick deposit of sterile clay from flooding (>40cm, level X), which separates the section from the neighbouring Lower Magdalenian (Fig.1). The sedimentology study (Hoyos, 1995) and the material culture show that they correspond to two successive stadia. The early Middle Magdalenian (level IX-IV) is a very cold and wet episode (GS2) with a thick deposit (>70 cm) that offers representations of cold steppe fauna engraved on slabs: Rangifer tarandus, Mammutus primigenius and Coelodonta antiquitatis. Accordingly, the fauna consumed includes reindeer remains at the base (necklace of eleven serrated incisors, level IX) and in the ceiling (level VI) of the section. In the latter level, particularly cold, boar and reindeer appear together, reproducing the reindeer-roe association observed in the Lower Magdalenian. In this respect, Altuna says (Corchón, in process) that the simultaneous presence of conflicting communities of ungulates, which do not currently exist in any biotype nor among the fauna in the sites in open areas of central and western Europe during the Würm glacial stage, is not uncommon in the Cantabrian region. Its complex terrain creates a labyrinth of valleys with sunny and shady spots in the same valley, where cold steppe areas can coexist with deciduous woods. The palaeoeconomy is based on the exploitation of deer (61%), horses (19%) and capridae (Capra and Rupicapra, 15%), with some bovine and carnivores. Visits to the coast provide marine mammals to the archaeological record (Physeter macrocephalus, Globicefala melas, Halichoerus gryphus), the majority teeth perforated to be used as pendants and a wide collection of molluscs (Pecten maximus, Nucella lapillus, Littorina obtusata, Mytilus, Teredonavalis). One data of interest refers to the association of the sperm whale with Coronula diadema (level VIII), an ectoparasite of large cetacean that proves that the meat from these animals was transported to and consumed in the cave. In addition, the mollusc Teredo navalis proves that wood that had been submerged in the sea was transported to Las Caldas (Corchón et al., 2008). The advanced Middle Magdalenian (level V-IV) was deposited under harsh and very wet 69
CANTABRIAN MOUNTAINS AND COASTLINE.  made of bone from the tibia of a deer, preserved complete  level 12 ceiling  Corch  n...
70 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD knapped on blades. Retouch and Aurignacian blades account for 10% and Dufour and backed blades represent 16%. The proportion of laminar flake among the knapped remains is low, which could be explained by the use of the blades and laminar flake for knapping tools. The advanced Middle Magdalenian, equally representative (>24,000 lithic remains), provide high levels of laminar flake, supports are second and blades/bladelets reach 40%. Similarly, 49% of cores are laminar (prismatic, pyramidal). Among the tools, burins (8%) barely exceed scrapers, some laminar, with many retouched blades (8%) and back edge bladelets with some triangles. The bone industries and mobiliary art are particularly typical in the early Middle Magdalenian (Fig.4), offering many Pyrenean type items: relief, sculptures, trimmed contours, rims, spear throwers, pointed horses teethand engravings. With them appear protoharpoons, forked assegais, denticulate points, semi-cylindrical rods, spatulas and perforated canes, extensively engraved. The engraved lithic plaque, very numerous (c. 100 to 200 per level), reproduce, in addition to the typical animals horse, deer, goat, bison, aurochs, reindeer, mammoth, rhinoceros), symbolic topics: acephalous, isolated hands or legs and beast-like anthropomorphs, carrying a pack on his back or doing an activity (sitting, squatting or with extended arms). Figure 3. Early Middle Magdalenian: engraved sperm whale’s tooth, engraved seal and dolphin teeth, sea molluscs, remains of Coronula Diadema. conditions, during the transition to the Late Glacial Maximum (GI Ie), with erosion and solifluction displacement processes. The fauna records the disappearance of steppe species and a notable reduction in large ungulates: Equusferus (1.5-2.3%) and Bos/Bison (<4%). Specialisation in deer (55%), goat (34%) and chamois (8-9.5%) is evident. The early Middle Magdalenian ceiling is dated at 16641 ± 363 calBP (n.VIc) and its base (level IX) at 16297 ± 436 calBP, the latter probably rejuvenated by the intense humidity and carbonation of the level. The advanced Middle Magdalenian (level IV: 16314 ± 454 calBP) is also coherent with the dating of the next level, transitional to the Upper Magdalenian (level IIIb-c: 16114 ± 441 cal BP). The early Middle Magdalenianlithic industry with more than 50,000 remains shows a high proportion of burins (20%), particularly dihedral burins (14%), doubling the number of scrapers, some The advanced Middle Magdalenian in terms of symbolic expression translated the palaeocological changes described. The Pyrenean volumetric models and techniques (relief, sculptures, rims, trimmed contours, etc.), the steppe fauna and the symbolic theme disappear or are very rare. In bone objects, forked assegais and protoharpoons become rarer but semicylindrical rods with linear decorations increase, as do needles, with a notable presence and variety (3%). The portable art shows a wide variety of complex symbols (arches, spindles, inlaid angles, eyes, rhombus, etc.), but animal symbols are limited: goat, horses, deer and salmonidae, with a synthetic design and lightly outlined limbs, appendages and manes (Fig.4). 3.3. Upper and Late Magdalenian The Upper Magdalenian has been dated at 14936 ± 342 calBP in Sala II. Like the rest of the site in the Nalón valley, they are laminar industries that have become significantly smaller. They are characterised by the abundance of burins, the continuity of the bone point types (cylindrical and oval assegais, bevelled, round and forked based, semi-cylindrical rods with
70  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  knapp...
CANTABRIAN MOUNTAINS AND COASTLINE. Figure 4. Early Middle Magdalenian portable art (1-7: plaquettes engraved with anthropomorphs, bone with engravings of horse legs and spear thrower with relief bison hoof) and advanced Middle Magdalenian (8: horse pelvis with engraved horses). tuberculated decoration), with one-sided harpoons. In the most recent level, late Magdalenian (level 2A, Sala I), a harpoon with a centre perforation characteristic of the end of the sequence in the Cantabrian was collected (Corchón, 2007). In the portable art, zoomorphology symbols become rarer, and in general the highly elaborate decorations, decreasing engraved slabs. At the same time, linear decorations, usually regular series of short incisions engraved on the shaft of assegais, rods and harpoons, which combine the functional quality with simple decorative graphic schemas. Conclusions The data provided by research at La Cueva de las Caldas cave shows that the middle Nalón valley could have acted as a shelter during the UMG and Late Glacial Maximum. Las Caldas shows that, since the beginning of the regional Solutrean, regional organisation has arisen in the subsistence strategies. This is demonstrated by the mobility of the social groups that occupied the valley in order to collect resources and raw materials, which were sometimes transported to the cave over significant distances. In the early Middle Magdalenian this mobility implied the existence of long-distance cultural contacts, exchanges and the distribution of objects, the portable of which puts their origin in the SW French Pyrenees. The rapid fall in the levels of occupation and cultural evidence in the Nalón valley as the Upper Magdalenian advanced is a little-known phenomenon. The break in these flows of cultural diffusion over time with the onset of milder climates are perhaps related to the glacier discharges and mass landslides documented in the High Nalón valley (Jiménez, 1997), which could have affected the middle section of the Nalón valley in warm and wet climate cycles, given that this drained the whole Macizo Central of Picos de Europa and its peak hosted significant Würm glaciers (Jiménez, 1996). 71
CANTABRIAN MOUNTAINS AND COASTLINE.  Figure 4. Early Middle Magdalenian portable art  1-7  plaquettes engraved with anthro...
72 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Maíllo-Fernández, José Manuel*; Arteaga, Carlos**; Iriarte-Chiapusso, María-José***; Fernández, Antonio****;, Wood, R.*****, Bernaldo de Quirós, Federico******. Cueva Morín (Villanueva de Villaescusa, Cantabria) Introduction Cueva Morín is one of the key sites to understand the evolution of the Palaeolithic in Cantabrian Spain, thanks to its long stratigraphy. It contains many of the Middle and Upper Palaeolithic techno-complexes, and has provided a series of unusual finds in the region, in connection with the ways of life and funerary realm of its Palaeolithic inhabitants. The cave is located in Villanueva de Villaescusa (Cantabria) and is also known as Mazo Moril and Cueva del Rey. It formed in a small hill of Urgonian (Aptian) limestone, in the Solia drainage basin, at 57m above sea level and 22m above the Obregón rivulet, 6 km from the modern coastline in the Bay of Santander. The cave entrance faces north-east, and leads into a short cave (GonzálezEchegaray and Freeman, 1971). Cueva Morín is not the only cave in the hill, which in fact contains a complex series of passages, above all another cave on a lower level, called Cueva del Oso, where surveying found some lithic artefacts on the surface (Serna et al., 2001). History of research on the deposit The cave was made known to the scientific community by H. Obermaier and P. Wernert in 1910. After several visits to the cave, two years after the discovery, in 1912, J. Carballo and P. Sierra carried out a small pit which remained unrecorded until O. Cendrero later published some of the materials the two researchers had found (Cendrero, 1915). From 1917 to 1919, J. Carballo continued working in the cave and undertook what can be considered the first serious and systematic excava* ** *** **** ***** ****** tion of the deposit. At this time, the Upper Palaeolithic and two Middle Palaeolithic levels were excavated (Carballo, 1923). In turn, in 1918 after he had finished his fieldwork, Carballo invited the Count of Vega del Sella to excavate the site. These new excavations, which lasted two years, were soon published by the Count (Vega del Sella, 1921), and he also informed about the Middle and Upper Palaeolithic occupations. The site was abandoned until the mid-1960s, when from 1966 to 1969, a Spanish-American team led by the Professors J. González Echegaray and L.G. Freeman carried out further excavations. These were some of the first “modern” excavations in Palaeolithic archaeology at that time, introducing new excavation techniques and with the participation of an inter-disciplinary team to assess the totality of data obtained from the deposit (González-Echegaray and Freeman, 1971, 1973). This research, as well as the application of new excavation methods, also contributed a complete and revised sequence of the different occupations at the site, which included for the first time, a clear and well-defined Chatelperronian level. The stratigraphic sequence comprising the deposit after this fieldwork consisted of 22 levels, attributed to the following periods: Azilian (Level 1), Magdalenian (Level 2), upper Solutrean (Level 3), Gravettian (Levels 4 and 5b), evolved Aurignacian (Levels 5a), early Aurignacian (Levels 6 and 7), archaic Aurignacian (Levels 8 and 9), Chatelperronian (Level 10), archaeologically sterile (Levels 18 to 21), and Mousterian (Level 22). In addition to this significant cultural sequence, González Echegaray and Freeman’s excavations uncovered evidence of two dwelling structures belonging respectively to the Mousterian and Au- Dpto. Prehistoria y Arqueología, UNED. C/ Paseo Senda del rey, 7. 28040 Madrid. jlmaillo@geo.uned.es Dpto. de Geografía y ordenación del territorio, Área de Geografía Física. Universidad Autónoma de Madrid. IKERBASQUE /Área de Prehistoria, Universidad del País Vasco. Departamento de Geografía, UNED. Research School for Earth Sciences, Australian National University, Canberra, Australia Área de Prehistoria, Universidad de León.
72  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Ma  l...
CANTABRIAN MOUNTAINS AND COASTLINE. rignacian, and exhumed three pseudomorphs (one of them complete) and the badly altered grave of a fourth (Freeman, 1971a, 1973; Freeman and González-Echegaray, 1973, González-Echegaray and Freeman, 1978). Thirty-five years later, it was necessary to update information about Cueva Morín, in relation with new methods and data introduced during that time, from both methodological and epistemological approaches to the science. For this reason, one of the authors (J. M. Maillo-Fernández), together with J. González Echegaray as co-director, carried out a research project which included a small archaeological study of a section of the deposit in 2005, in order to determine its sedimento- logical and geomorphologic characteristics, as well as obtain environmental (palynology) and chronocultural information about the deposit. The small size of the section left by the old excavations and their delicate state of conservation were deciding factors when selecting the area to excavate, in order to preserve as much of this important section as possible. Stratigraphy, geomorphology and chronology The most complete stratigraphic sequence in the deposit (22 levels although not all of them are archaeological) was attained by the 1966-69 excavation (Table 1). Level Composition Thickness (cm) Cultural attribution 1 Sandy-silty, 7,5 YR 3/2 2-20 Azilian 2 Silty-sand with gravel, 10 YR 2/2 5-10 Magdalenian 3 Silt, 7.5 YR 3/2 2-8 Upper Solutrean 4 Sandy silt, 5 YR 3/2 5-20 Gravettian 5 Sandy silt with gravel, 5 YR 2/1 15-30 Gravettian and final Aurignacian 6 Silty, 5 YR 3/3.5 20-30 Early Aurignacian 7 Sandy silt with clay lenses, 5 YR 2/1 10-18 Early Aurignacian 8 Silty-sand, 5 YR 3/3 10-20 Proto-Aurignacian 9 Silty-sand with fine gravel, 10 YR 3.5/3 5-8 Proto-Aurignacian 10 Clay, 10 YR 3/3 2-5 Chatelperronian 11 Sandy-silty, 2.5 YR 2/0 8-18 Mousterian 12 Silt with gravel, 7.5 YR 3/2 12-20 Mousterian 13 Clayey-silt, 7.5 YR 2/0 5-8 Mousterian 14 Silty-clay, 10YR 3.5/3 7-10 Mousterian 15 Clayey-silt, 10 YR 3.5/3 15-20 Mousterian 16 Silty-sand, 10 YR 5/6 12-15 Mousterian 17 Sandy-silt with gravel, 10 YR 4.5/5 12-22 Mousterian 18 Sandy-silt, 7.5 YR 4/4 45-70 Sterile 19 Sandy-silty, 7.5 YR 4/4 12-22 Sterile 20 n/a 2-5 Sterile 21 n/a 7-10 Sterile 22 n/a 2-5 Indeterminate Palaeolithic Table 1. Archaeological levels in Cueva Morín 73
CANTABRIAN MOUNTAINS AND COASTLINE.  rignacian, and exhumed three pseudomorphs  one of them complete  and the badly altere...
74 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. Plan of the cave. The sedimentological study carried out by H. Laville and M. Hoyos (Laville and Hoyos, 1994) revealed some issues in the chrono-climatic sequence and the composition of some levels, in comparison with earlier work (Butzer, 1981). Later, J. Sanguino and collaborators, indirectly by using the other researchers’ work, even cast doubts on the nature of some of the archaeological levels (Sanguino and Montes, 2005). Therefore, an understanding of the formation of the sedimentary deposit in the cave was one of the main objectives of the fieldwork carried out in Cueva Morín in 2005. The sedimentological study Figure 2. Stratigraphy of Cueva Morín. performed with samples collected then has characterised sediments of two different kinds in the main section of the deposit (IB-IXB in the classic nomenclature) (Fig. 1). On the one hand, those near the cave entrance of external origin and with the classic formation process of fill in rock-shelters and caves. On the other, sediments related with water flowing from inside the cave to outside (Fig. 2). This multi-episodic flow may have alternated with the occupation in the cave, as no erosive scars are seen between the sediments of the two parts. As Count of Vega del Sella was able to observe in his excavation, it formed a meander whose channel-lag deposits may be situated around square IX, where the cave turns and the material is larger grained. The outer side of the meander, where the finer sediment is deposited, can be traced in the section of square J (Fig. 1). This current may have been active at different times in the sedimentary history of the cave, as the end of the cave is blocked by calcite and sub-angular and sub-rounded cobbles, as if it was an underground “point bar”. It was abandoned, allowing later occupations, as suggested by the finds made in the central part of the cave in previous archaeological studies and as is seen in the whole main stratigraphic section. Chronology is the Achilles heel of the deposit. Several attempts have been made to date its human occupations with very different results and leaving part of the stratigraphy unsampled (Table 2). The first attempt was during González Echegaray and Freeman’s excavations (Stuckenrath, 1978). The results were partly contradictory with the stratigraphy and generally rejected by the scientific community. However, attention should be paid the determinations from Level 8, proto-Aurignacian, as they bear no relation to the stratigraphic section or the excavation in that area, where the industries have been
74  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figur...
CANTABRIAN MOUNTAINS AND COASTLINE. Level Techno-complex Material Context Lab. Ref. Date (BP) Calibration range (cal BP, 95.4% probability range) from Reference to 5s Gravettian Charcoal Excavation SI 953 20120 ± 340 BP 25194 23452 Stuckenrath, 1978 7 Early Aurignacian Charcoal Excavation SI 954 31500 ± 880 BP 37985 33975 Stuckenrath, 1978 7 Early Aurignacian Charcoal Excavation SI 955 28680 ± 840 BP 34381 31171 Stuckenrath, 1978 7 Early Aurignacian Charcoal SI 955a 27260 ± 1500 BP 35220 28610 Stuckenrath, 1978 8 Proto-Aurignacian Charcoal Morín III SI 956 27710 ± 1300 BP 34925 29420 Stuckenrath, 1978 8 Proto-Aurignacian Charcoal Section (ABA) GifA96263 36590 ± 770 BP 42432 39734 MaílloFernández et al., 2001 8 Proto-Aurignacian Charcoal Section OxA19084 40060 ± 350 BP 44399 43052 Maroto et al., 2012 8 Proto-Aurignacian Charcoal Upper part of Morín I SI 952 27630 ± 540 BP 33022 30820 Stuckenrath, 1978 8 Proto-Aurignacian Charcoal Idem, soluble part in NaOH SI 952a 27360 ± 740 BP 33445 30214 Stuckenrath, 1978 9 Proto-Aurignacian Charcoal Section (ABA) GrA33891 33.430 +250 -230 BP 38470 36879 Maroto et al., 2012 10 Chatelperronian Charcoal GrA33823 29.380 +260 -240 BP 34033 32984 Maroto et al., 2012 10 Chatelperronian Charcoal Section SI951 27800 ± 560 BP 33235 30935 Stuckenrath, 1978 10 Chatelperronian Charcoal SI951a >30340 BP 11 Mosterian Charcoal Section (ABA) OxA19083* 41800 ± 450 BP 46022 44406 Maroto et al., 2012 11 Mosterian Charcoal OxA19459* 43600± 600 BP 48343 45648 Maroto et al., 2012 11 Mosterian Charcoal Section (ABA) GifA96264 42000 ± 730 BP 46922 44120 MaílloFernández et al., 2001 Idem, soluble part in NaOH Section (Acid only) Idem, soluble part in NaOH Scetion (ABOx-SC) Stuckenrath, 1978 N/A Table 2. Radiocarbon dates from Cueva Morín. Stuckenrath’s determinations (1978) were given in years BC supposing a half life of 5370 years. They have been recalculated following Stuiver and Polach (1977) using a half life of 5568 years and placing it on a time scale before 1950. These recalculated dates have been used in the calibration. The dates with an asterisk were obtained from the same sample. The determinations have been calibrated against IntCal13 (Reimer et al., 2013) in OxCal v.4.2. (Ramsey, 2009). 75
CANTABRIAN MOUNTAINS AND COASTLINE.  Level  Techno-complex  Material Context  Lab. Ref.  Date  BP   Calibration range  cal...
76 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD defined. These dates correspond to the area of the “burials” which are associated with this Level 8. They are all in a span of time between 27 and 26 ky BP. In our research project, we carried out a series of determinations with material taken from the section, and these have provided dates for the Mousterian of 40 ky BP and for Level 8 of 36.5 ky BP (MailloFernández et al., 2001). Recently, Maroto and collaborators have re-analysed samples we have provided by ultrafiltration, also obtaining disconcerting results, in which the date for Level 10 (Chatelperronian) is very similar to the one obtained by Stuckenrath (Maroto et al., 2012 and Table 2). Unfortunately the fragment was very poorly conserved and only the carbon prepared before the dating process could be analysed. It therefore only gives a minimum age for this level. Once again, technical improvement induced us to carry out a new attempt with material from the 1966-69 collection, with the ultrafiltration method. The low collagen content in the samples meant that results could not be obtained. Cultural sequence Cueva Morín is an important site for Upper and Middle Palaeolithic studies in the region owing to the long cultural sequence conserved in its deposit: 7 Mousterian levels, 1 Chatelperronian, 2 proto-Aurignacian, 2 early Aurignacian, 1 evolved Aurignacian, 2 Gravettian, 1 Solutrean, 1 Magdalenian, and 1 Azilian. The Azilian level (Level 1) has not been dated, although grosso modo this techno-complex can be situated between 11,500 and 9,500 BP (Tresguerres, 2004). The layer is located near the cave entrance, and contains a dense number of objects including micro-blade tools, mostly consisting of retouched bladelets, micro-gravettes and numerous endscrapers (González Echegaray, 1971a). Level 2 corresponds to the upper Magdalenian and is characterised by a large number of burins, truncated pieces, and carinated and nucleiform endscrapers, together with a large proportion of microlithic tools (backed bladelets, denticulates, Dufour, geometric microliths, etc.). The osseous assemblage is abundant, with fragments of sagaie points with a circular cross-section, some of them with central groove, decorated flattened rods, and pendants made from bone and red deer canine teeth (González Echegaray, 1971a). The upper Solutrean (Level 3) is a thin layer where the lithic assemblage is not particularly rich. However, shouldered points, and to a lesser extent willow and laurel leaf points are relatively abundant. Therefore, this level can be clearly attributed to the upper Solutrean in Cantabrian Spain (González Echegaray, 1971a). The Gravettian is represented by two layers (Levels 4 and upper 5), with a large number of backed pieces (Gravette points, micro-gravettes), truncated pieces, and some Noailles burins in the most recent level. No recent studies have examined the technotypology in these two levels in depth, nor is any date available to situate them in the Cantabrian Gravettian, as the one obtained by Stuckenrath (1978) is clearly anomalous. Recently, the Gravettian in Cueva Morín has been identified as belonging to a late stage in which Noailles burins are scarce and backed pieces more abundant (Peña, 2011). Similarly, the evolved Aurignacian at Cueva Morín (lower Level 5) is in need of a techno-typological reappraisal. The evolved Aurignacian is a polymorphic techno-complex in Cantabrian Spain, which hinders its characterisation (Cabrera et al., 2004). Thus, in Cueva Morín, the lithic assemblage differs from the early Aurignacian in the smaller number of carinated endscrapers and a greater abundance of thick-nosed endscrapers. However, the osseous assemblage is abundant and meaningful in this level, with spindle and flat-shaped points together with the survival of split-based points. Levels 6 and 7 in the deposit correspond to the early Aurignacian. The industry is characteristic of this techno-complex, with abundance of carinated pieces, two types of blade debitage, one for blades from prismatic cores with unipolar exploitation and another for bladelets made from carinated cores (Arrizabalaga, 1995; Cabrera et al., 2004). However, the bone industry is poor and not diagnostic. The proto-Aurignacian (Levels 8 and 9) have been reappraised recently from the technological and typological viewpoints (Arrizabalaga, 1995; MaílloFernández, 2003). It is a markedly micro-blade industry, with abundant Dufour bladelets. The blade were extracted from prismatic cores with unipolar reduction, in a continuum between blades and bladelets. The relative importance of substrate tools (sidescrapers, denticulates and notches) should be stressed, as well as flake reduction schemes (Maíllo-Fernández, 2012). Level 9, because of techno-typological and sedimentological issues, may have suffered taphonomic alterations, affecting its industrial integrity.
76  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  de   ...
CANTABRIAN MOUNTAINS AND COASTLINE. There is no doubt that the most important level, the one that has attracted most attention to Cueva Morín, is Level 10, corresponding to the Chatelperronian. When it was identified, it was used to address the problem of the Middle-Upper Palaeolithic transition, at that time associated with the “Aurignacian-Mousterian” (González Echegaray, 1969, 1971a). The lithic assemblage has been revised on several occasions since then (Arrizabalaga, 1995, Maíllo-Fernández, 2003). However, the level and its integrity have been discussed by several scholars. K. Butzer, who carried out the first sedimentologival study of the cave, concluded that this level contained parts colluviated by later cryoturbation (Butzer, 1981: 146), probably based on the incoherence of the dates (Stuckenrath, 1978). In the review carried out by H. Laville and M. Hoyos, it was thought that it had been eroded towards the interior and scalloped towards the exterior by cryoturbation processes (Laville and Hoyos, 1994: 204). The total integrity of the level has recently been questioned following a re-interpretation of the data given by Laville and Hoyos, who thought the Cueva Morín, like nearly all the caves in Cantabrian Spain, was not apt for determining a regional palaeo-climate sequence. In this new interpretation, the scalloped part was taken as an exclusive indicator of solifluction and, taken together with the thinness of the level (2 – 5cm), it was concluded that Level 10 “is an erosive contact, altered by water currents in Levels 11 and 9” (Muñoz and Montes, 2003: 206). We carried out the first examination of the section since 1969. This allowed us to verify that Level 10 is visible in the outer part of the stratigraphic section, exactly as identified by Butzer, Laville and Hoyos, and as represented in the relevant monographs (González Echegaray, 1971b; Butzer, 1981: 143). In addition, the sedimentological curve of the Level 10 reveals that it was deposited in a low energy process and not by solifluction. The scalloping is due to a load deformation processes, which also caused scrolling, which is only found at the top of the level, although certain higher energy water action, which produced a large scroll, cannot be ruled out. The Chatelperronian level has also been questioned from the point of view of the lithic assemblage; its existence has even been denied, arguing that it is a mixture “in equal parts” of Aurignacian and Mousterian material (Muñoz and Montes, 2003: 206). We have already reasoned at length against this hypothesis (Maíllo-Fernández, 2007a, 2008), but we might repeat that it is very hard for us to un- derstand how the mixture of two lithic assemblages that belong to two well-defined techno-complexes (Aurignacian and Mousterian) can result in a third, completely different one (Chatelperronian). The cultural sequence at Cueva Morín ends with a series of Mousterian levels. The most recent, Levels 11 and 12, classified as denticulate Mousterian, is characterised by discoid and partly Levallois production, with a small but significant micro-blade production (Maíllo-Fernández, 2001, 2007b). The lower section (Levels 13-17), where the predominance of cleavers is the most significant trait (Mousterian variety traditionally known as Vasconian), was attributed to the Mousterian of Acheulian tradition in the first studies (Freeman, 1971b). Dwelling structures and burials The 1966-69 excavation at Cueva Morín did not only reveal a chrono-stratigraphic sequence of vital importance in Cantabrian Spain, but also two dwelling structures and a series of burials. The dwellings appeared in the central part of the cave, but belonged to different cultural periods. The oldest corresponded to Level 17 (Mousterian) and had been cut through by Vega del Sella’s excavations. It is 6.6m2 in size, and consists of a line of stones forming a curved area with sediment inside it clearly different from the rest of the level. No remains of hearths were detected inside it (Freeman, 1973). In Level 8, proto-Aurignacian, another area was interpreted as a dwelling structure. Partially destroyed by the excavations in the early twentieth century, which do not allow its width to be determined, it was nearly three metres long and rectangular. There was no evidence that the structure had been covered, at least with posts. The sediment had been dug out to a maximum depth of 27cm, and against the innermost wall there were remains of a hearth, possibly in a pit, and on the opposite side a step 125cm long and 50cm wide, interpreted as a bench (Freeman, 1971a). Associated with the Aurignacian, at the back of the cave, after an area interpreted by the excavators as a wall with wooden posts, were found a series of mounds that held the burials of four individuals. The most peculiar thing about these burials is that they did not contain the skeletal remains of the bodies. Instead, in the decomposition processes, they had 77
CANTABRIAN MOUNTAINS AND COASTLINE.  There is no doubt that the most important level, the one that has attracted most atte...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 78 turned into adipose and later been covered by sediment, so that the remains were turned into a positive mould that was difficult to interpret. The most complete and most recent is Morín I, found in a grave 210 x 52 cm in size. Although it is not easy to interpret, it seems that this individual was lying on its left side with flexed arms and slightly flexed legs. The excavators interpreted that its head had been amputated and deposited in the lower part of the grave. In the mount covering the body, the remains of two small hearths were found (Freeman and González Echegaray, 1973). The remains called Morín II were limited to a dark, greasy and plastic substance, similar to that in the other graves, but it could not be interpreted Jesús Altuna* Koro Mariezkurrena* further. Beneath Morín I and partly destroyed by it, Morín III only consisted of two legs. Finally, Morín IV was limited to the partial remains of the grave and the mound covering it, as it had been destroyed by the other burials. Acknowledgements This study is dedicated to the memory of Professor Joaquín González Echegaray and Professor Leslie G. Freeman, who were truly responsible for the scientific understanding of Cueva Morín. Ekain cave (Deba, Basque Country) Ekain cave (Deba, Basque Country), a cave famous for its rock art ensemble in its interior (Altuna, 1996), possesses an archaeological deposit at its entrance. This has been excavated in two stages; between 1969 and 1975, first directed by J. M. de Barandiarán and later by J. Altuna (Altuna and Merino, 1984) and then between 2009 and 2011, directed by J. Altuna (Altuna, 2009). The cave is located at the confluence of two valleys, each with a small stream, in an area with biotopes of steep crags. The two streams together flow into the River Urola one and a half kilometres downstream, at a point 8 kilometres from the modern coastline. In this point, the landscape has changed to gentle hills. The deposit in the entrance of Ekain, 5m thick, consists of 12 levels (Fig. 1). The lowest Levels XII and XI are totally barren, both archaeologically and palaeontologically. Level X is very rich in remains of Ursus spelaeus. Human influence is minimal and only a few signs of Chatelperronian remains have been identified. * Figure 1. Stratigraphy of the archaeological deposit at Ekain cave. Centro de Conservación e Investigación de los materiales Arqueológicos de Gipuzkoa (GOAZ). Paseo de Zarategi 84-88. 20015. San Sebastián. altuna@arkaios.com
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  78  turne...
CANTABRIAN MOUNTAINS AND COASTLINE. Level IX also contains quite a large number of Ursus spelaeus remains, but only a little evidence of the Gravettian. Its base was dated by radiocarbon to over 30,600 BP (I-11056). These two levels contain very little knapping waste, which indicates that the artefacts found in them had been made elsewhere. The cave was therefore visited sporadically by people who did not usually live there. The site of Irikaitz, located 2km from Ekain, with Gravettian remains in its upper part, might be connected with this. Level VIII was probably deposited in the Würm III/IV interstadial, as it no longer contains Ursus spelaeus remains. The lithic assemblage is scanty and not diagnostic. There are a few faunal remains, above all of Rupicapra rupicapra. The presence of Sus scrofa and Capreolus capreolus indicates climate amelioration in this level, which has been dated to 20,900 ± 450 BP (I-13005). Level VII (Cantabrian lower Magdalenian) is the first level displaying intense human occupation. The sedimentological study shows that the base of the level was deposited in conditions of moderate humidity and relatively high temperatures, while these changed to cold and humid at the top of the layer. Five radiocarbon determinations have been obtained, dating to between 16,500 and 15,400 BP (all by Isotopes, Westwood, New Jersey). The pollen analysis shows, at the base of the level, a proportion of 12% Pinus, 3% Corylus and a smaller presence of Alnus, Betula and Quercus t. pedunculata, as well as an abundance of Ericaceae and filicales triletes. This appears to correspond to the late Lascaux interstadial. The conditions worsen and the number of taxa decreases, but the climate improves again at the end of the deposition of the level, which suggests the Bölling. The most common small mammal in the coldest phases is Microtus oeconomus, and Arviciola terrestris in the less cold phases. Microtus gr. agrestis-arvalis and Talpa europea also occur in significant numbers. The lithic assemblage displays specialisation with a large number of microlithic backed bladelets used to make hunting implements. This suggests a seasonal occupation of the cave. However these occupations were prolonged as indicated by the fact that knapping was performed in the cave. Knapping waste is abundant. Equally, the readaptations of decorticating flakes and core flakes, as well as some burins and their spalls, confirm this. The osseous assemblage clearly exhibits the particular characteristics of the Cantabrian lower Magdalenian. The most characteristic tools because of their chrono-stratigraphic position are assegai points with a square cross-section and split base, and a bi-pointed object. They belong to a single type, pointed artefacts, which seems to suggest a particular activity (hunting). The presence of antler waste rods means that certain artefacts were manufactured in situ. The ungulate remains clearly support the seasonality indicated by the industry. The base of animal subsistence was red deer. The age analysis of the fawns and other young ungulates demonstrates that they were hunted in the first month of life (June) and others also in a mild season of the year. None were hunted in winter. This raises the matter of the base site from which Ekain was used as a hunting post. Two hours away on foot is the site of Urtiaga, occupied throughout the year, with Erralla an hour and a half away (Altuna, Baldeón and Mariezkurrena, 1985). Both sites possess thick lower Magdalenian levels. The analysis of the skeletal parts in the deposit shows that the animals were brought to the site whole, which is logical bearing in mind that many of the prey were fawns killed in their first month of life. The osseous industry in this Level VII at Ekain displays similarities with Level F in Urtiaga. Level VI (upper-final Magdalenian) formed in its lower part in very cold and less humid conditions than the previous level. It includes the most evidence of frost-shattering in the whole deposit. The upper part of level formed in a cold and dry climate. The pollen studies reflect a decline in arboreal and filical species in the lower part, in which deciduous trees are practically non-existent. Carduaceae acquire their highest proportions, which Ericaceae reach a minimum level in the severest conditions (low humidity and extreme cold) in the sequence at Ekain. The level is dated by radiocarbon to 12,050 ± 190 BP (I-9240). However, the upper part of the level indicates a slight improvement, probably corresponding to the Alleröd. The significant decrease in Ericaceae, together with the lesser tree cover, produced a decline in the Cervus elaphus population and consequently these animals were replaced as prey by Capra pyrenaica. The faunal analysis suggests that the site was still occupied seasonally. The skeletal profiles show that 79
CANTABRIAN MOUNTAINS AND COASTLINE.  Level IX also contains quite a large number of Ursus spelaeus remains, but only a lit...
80 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD the prey was brought to the cave whole less often than in Level VII. This is understandable because the preferred prey was adult ibex and the hunting sites would have been of more difficult access. The animals were butchered where they were hunted and selected parts were brought back to the cave. The faunal assemblage also includes evidence of Rangifer tarandus and Lepus timidus, as well as salmon vertebrae. It is interesting to note that the rock art in the cave includes a painting of a salmon. plays a deep longitudinal incision next to the start of the barbs on one side and profound incisions in the barbs. Another exhibits fine incisions on one of its sides, forming a decorative V-shaped motif. On the opposite side, seven short and shallow transversal incisions are similar to the so-called “hunting-marks”. This specimen is similar to a harpoon found in Level D at the site of Urtiaga. The lithic assemblage further supports this seasonality. The most abundant implements are still microliths, with few medium-size tools such as burins, scrapers or denticulates. The percentage of burins increases slightly, which may be connected with the greater importance of the osseous industry in this level. The small backed tools display a change within this level, as in its lower part (Level VIb) there are no backed points, whereas the upper part (Level VIa) does contain some. Two of the assegai points are decorated with schematic depictions of an ibex viewed from the front, like some figures in the rock art ensemble. The same level yielded an engraved sandstone plaquette representing three animals: a male ibex, a stag and a horse (Fig. 2). The ibex is the most outstanding figure as its head is depicted in great detail. Its horn displays the two typical curves of the Pyrenean species, on which the growth rings are indicated by a series of transversal lines. The red deer, drawn with finer lines, displays antlers in which the two base points, the middle points in each antler and the wider crown are indicated. The third figure, a horse, is less conspicuous and perfect than the other two. The osseous assemblage also exhibits two stages, as there are no harpoons in Level VIb, while there are in Level VIa. The harpoon assemblage in VIa is homogeneous. Their barbs are on one side, and they are somewhat flattened. The barbs are large and separated, with few of them in relation to the shaft and with deep incisions. One of them has a flat shaft, with a single row of barbs. It dis- Level V reflects more humid and less cold conditions than Level VI, although the reduction in arboreal cover and filicales persists. The lithic industry is poorer, but continues to include backed points and bladelets, scrapers, burins and denticulates. The assemblage from this level is not very characteristic but appears to represent the end of the upper Palaeolithic in the cave. This level does not contain a single remain of marine molluscs. Figure 2. Plaquette from Level VI at Ekain cave.
80  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  the p...
CANTABRIAN MOUNTAINS AND COASTLINE. Levels IV and III, dated in the Azilian, clearly show the climate amelioration in the Pre-Boreal, with very humid conditions and clear rise in the temperatures. Pine, which had been the main tree species, was slowly replaced by deciduous taxa such as Corylus and Alnus, with a presence of Juglans, Tilia, Quercus, Betula, Fagus, Cupresaceae and a large percentage of Ericaceae, which reaches its maximum values. Microtus oeconomus disappears definitively from among the small mammals, while the bat Myotis myotis appears. Red deer was the preferred prey over ibex. The representation of Capreolus capreolus increases and Sus scrofa appears in the Ekain sequence together with Meles meles. Shellfish were gathered, as shown by the large number of remains of Patella and Osilinus lineatus. These resources were easily obtainable at this time, as the coastline approached the cave. Figure 3. Contour découpé of a bird. The lithic assemblage continues to indicate specialisation in hunting. The osseous industry includes a base of a typical Azilian flat harpoon with eye-shaped perforation. (Altuna 2009, 2010 and 2011) in reference to two aspects of the deposit. Level II formed in very dry and warm conditions. It has been dated to 9540 ± 210 BP (I11666). The lithic industry is comparable with Level III but with a more advanced Mesolithic component, including some triangles and the appearance of the micro-burin technique. Shellfish gathering reached its greatest importance in the whole sequence at Ekain. The species represented are Osilinus lineatus, Patella vulgata, P. depressa, P. aspera and Mytilus edulis. Remains of ungulates are scanty, with evidence of Cervus elaphus, Capra pyrenaica and Sus scrofa. The presence of 11 remains of Bos taurus is striking, and these were undoubtedly introduced afterwards. A potsherd was also found in this level, and must have equally been a later intrusion. Together with the remains of Bos taurus, this indicates that a new excavation needed to be carried out in an intact area, in order to clarify these intrusions. Level I is the modern surface layer. Modern Excavations New excavations were carried out from 2009 to 2011 and these provided some significant data First, affecting Levels II, III and IV, a series of remains of human infants and fragments of hand-made pottery were found. One of the sherds displayed button-shaped decorations and another had finger-nail impressions. One fragment of an infant’s mandible, found in Level II, refitted with another piece from Level III. This mandible has been dated to 4960 ± 60 BP (Ua36854). The pottery and Bos taurus remains found in Level II in the old excavations must be ascribed to this assemblage. When the base of Level VI was excavated, a contour découpé of a bird, made from a bovine rib, was found (Fig. 3) Altuna and Mariezkurrena, 2013). This displays detailed internal shaping and has been dated to 13,862 ± 129 BP (Ua39108). Both the date and the characteristics of the find correspond to the middle Magdalenian. This occupation in this phase, which must have been very short-lived in Ekain cave, was not detected by the old excavation. Level VI begins immediately beneath it. This representation is unique as, to date, no figures of birds have been found on this kind of object. Additionally, it was made from a bovine rib, rather than from a hyoid bone, which is also unusual in contours découpés. 81
CANTABRIAN MOUNTAINS AND COASTLINE.  Levels IV and III, dated in the Azilian, clearly show the climate amelioration in the...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 82 Javier Baena Preysler * Elena Carrion Santafé ** Cueva de El Esquilleu: a new point of reference for the Cantabrian Mousterian As a result of collaboration between the Autonomous University of Madrid and the Consejería de Cultura of the government of Cantabria, in the context of the project HUM2004-04679/HIST “CONTEXTO CRONOLÓGICO Y CULTURAL DEL FINAL DEL PALEOLÍTICO MEDIO EN EL NORTE PENINSULAR”, a series of field seasons were carried out in Cueva de El Esquilleu in Castro-Cillorigo (Cantabria) from 1997 to 2006. The state of conservation of the sequence in Cueva de El Esquilleu, and the method used (giving priority to determining the sequence and sampling), has enabled the collection of a large amount of data from the faunal, pollen, anthracological, sedimentological, stratigraphic and, naturally, archaeological records. The research has also Figure 1. Cueva de El Esquilleu (Cantabria, Spain) and its surroundings. * ** included a taphonomic study, technical and functional analysis, the finds of human remains, the micro-spatial reconstruction by phytolith analysis, and the study of processes of procurement of lithic resources and the functional relationship of the site with its environment. In short, this new sequence is one of the most complete Middle Palaeolithic sites in Iberia. El Esquilleu is a unique deposit. Although it is not at a great altitude (280m above sea level, and 68m about the River Deva), the montane environment predominates. It is a large rock-shelter, in a mostly limestone area (Valdeteja formation). It is a rugged location, different from most of the Cantabrian valleys that were occupied, and similar to only a small number of Mousterian sites in northern Spain, such as Axlor (Ríos Garaizar, 2012). The geomorphologic setting is a small basin between mountains, bounded to the south by the steep slope of the mountain valleys, which were glaciated at their heads, and very difficult connections with surrounding valleys (Fig. 1). The general nature of the Cantabrian relief, with its tendency towards dividing up the territory, is more pronounced in the Deva basin, with the connection to the coastal strip through a narrow gorge. The climate is also slightly milder than in other parts because of its particular location. The length of the occupation (more than 50,000 years) and the stratigraphic sequence are unique in Cantabrian Spain. The excavation, formally a sounding, was carried out from 1997 to 2006, with two test excavations perpendicular to each other, one of them transversal to the rock wall. The large surface area excavated was 14m2. In the sounding transversal to the cave wall, a depth of 4.2m was reached, and 41 stratigraphic levels were defined, 34 of them archaeologically fertile and in succession (Fig. 2). Over 100,000 lithic remains were retrieved, and about 25% of these have been described and studied from the technical viewpoint (as well as the initial typo- Dpto. Prehistoria y Arqueología, Universidad Autónoma de Madrid, Campus Cantoblanco, 28049 (Madrid, Spain) Subdirección General de Museos Estatales, Ministerio de Cultura, 28071 (Madrid, Spain)
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  82  Javie...
CANTABRIAN MOUNTAINS AND COASTLINE. IV 19,300±100 BP III Bone OxA-19968 19,310±80 BP III B Bone OxA-19246 20,810±110 BP 44,100±1300 BP Charcoal AA37882 36,500±830 BP Charcoal Beta149320 39,000±300 BP OxAX-2297-31 OxA-20320 49,400±1300 BP Charcoal OxA-20318 53,400±1300 BP Charcoal OxA-19993 >54,000 BP Charcoal OxA-20319 >58,500 BP XVIII Charcoal OxA-19993 49,700±1600 BP XIX-1 OxA-19967 OxA-19966 XVII Bone Bone XVII III OxA-19965 XVII 12,050±130 BP Bone 40,110±500.420 BP 43,700±1400 BP XVII-2 Charcoal AA-29664 GrA-33816 XVII-1 Charcoal Bone Charcoal XIII III AA37883 XIF 3640±90 BP Charcoal VI-2 GrA-33829 GrA-35065 VI-1 Charcoal Charcoal VI III GrA-35064 VIF The chronology at El Esquilleu covers a time from the early OIS 3, including OIS 3c and OIS 3b, the H5, H4 and H3 cold events in the middle part of OIS 3a, and probably part of OIS 2. Its levels cover a full span of time (>53 ky BP to <20 ky BP) (Jordá Pardo et al., 2009; Baena et al., 2012; Maroto et al., 2012; Yravedra and Uzquiano, 2013). The available radiocarbon dates situate the sequence quite well between about 60 and 20 ky BP. From a conventional perspective, the El Esquilleu sequence extends beyond the natural climate limit associated with the end of the Middle Palaeolithic and enters in Würm III (H4 in terminology of Heinrich, 1988). The whole sequence is Mousterian, with a long series of determinations: Charcoal V logical assignation and the general identification of their attribution). GrA-35064 IV Figure 2. Stratigraphical sequence in Cueva de El Esquilleu. Eastern profile Charcoal 22,840±280-250 BP 23,560±120 BP Charcoal OxA-19085 39,280±340 BP XIX-2 Charcoal OxA-19086 >54,600 BP XIX-3 Charcoal 39,600±400 BP XIX-4 Charcoal XXI-I Charcoal OxAV-2284-29 OxAV-2284-30 OxA-20321 XXId Burnt clay Mad3299 XXIb Burnt clay Mad3300 (TL) 51,034±5114 BP (TL) 53,491±5114 BP 30,250±500-430 BP 34,380±670 BP 52,600±1200 BP 39,650±450 BP >59,600 BP Table 1. El Esquilleu dates. The sequence is special not only because of the late age of its upper levels but also because of its early chronological start (53 ky BP for Level XXIb) in a total sequence of 41 levels. No other Middle Palaeolithic site is known in Cantabria or in the rest of Iberia, with such a continual and repeated occupation of the same place. Expressed in terms of the number of generations, if the occupation had been continual (and it was undoubtedly seasonal and possibly intermittent), these would amount to 1,600. The dates obtained for El Esquilleu are reasonably coherent, bearing in mind the variation in types of samples, methods and laboratories. Some of the 83
CANTABRIAN MOUNTAINS AND COASTLINE.  IV  19,300  100 BP  III  Bone  OxA-19968  19,310  80 BP  III B  Bone  OxA-19246  20,8...
84 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD differences can be explained as laboratory errors. Thus, for example, the different dates for Level XVII (53,400 ± 1300 BP, Oxa-20318, charcoal; and >58,500 BP, OxA-20319, charcoal) and for Level XIF (36,500 ± 830 BP, AA 37882, AMS charcoal; and 34,380 ± 670 BP, AA 37883, AMS) were obtained from a single sample that was divided up. Other cases can be regarded as inconsistent, such as (1) the three dates obtained for Level VI with results between 44 and 40 ky BP that do not match the chronological sequence of the occupation; and (2), those for Level XIX, for which the new determinations (30 ky BP) contradict the first ones obtained. However, the date for Level III (12,050 ± 130 BP, AA29664, AMS bone) is only atypical. It fits in the archaeological sequence and the samples have passed the stress tests carried out in specific studies (Jordá Pardo et al., 2009; Maroto et al., 2011). Stratigraphically, it may be supposed that the sequence entered markedly colder conditions after Level VI, which may correspond to H3c (c. 30 ky BP). The new dates obtained for Level III (19,300 ± 100 and 19,310 ± 80 14C ky BP and 20,810 ± 110 14C BP), performed within a monographic project involving a critical reappraisal of many of the dates obtained in Iberia, support its validity. However, other clearly aberrant results, such as one recently obtained for Level III, 3640 ± 90 BP, may be explained by the intrusion of Holocene charcoal (Maroto et al., 2011) or the presence of a very atypical Upper Paleolithic industries. In this way, radiometric coherence is observed in the dates around 20 ky BP for the end of the sequence. They were obtained in different laboratories, using different pre-treatment methods (including ultrafiltration) and display C13 values that confirm the quality of the sample (Maroto et al., 2011). Although many of the dates thought to be recent have been put back in time, in general terms El Esquilleu is still within an increasingly select group. However, some unknown issues cannot be ruled out, such as a possible contamination of the samples owing to post-depositional processes or percolation problems. T aphonomical studies (Yravedra and González Castanedo, 2013) have suggested that the dates may have been obtained from bones which had not been handled by humans, and which are therefore not archaeological material. Methodologically (i.e., De la Rasilla and Santamaría, 2013), the limitations of the radiometric method itself have been pointed out, because of calibration deficiencies (Jöris et al., 2011), and the divergent results caused by the specific treatment and analytical protocols (Bird et al., 2010). Technologically (Vaquero, 2013), other studies based on the nature of the lithic assemblage (whose expedient and indeterminate character has been highlighted) suggest that Level III could be a facies with an atemporal cultural attribution. In any case, a relativist position of the dates can be used, where, independently of their exact result, their relative position in the sequence can be considered (Carrión et al., 2013). In addition, the internal analysis of the industries and their relationship with the environment in chaîne opératoire terms provides interesting information about changes in the relationship between the group and their surroundings (Carrión et al., 2008). El Esquilleu is also special because its sequence starts at an early date. There are few sites dated before 40-45 ky BP in the Cantabrian Mousterian. The classic chronologies of Castillo Level 22, Pendo XVII and Lezetxiki V have been seen to be disputable and older dates are very rare, with only El Sidrón (c. 50 ky BP; De la Rasilla et al., 2013) and in a nearby region, Cueva Corazón (96.95 ky BP, by TL, Díez et al., 2008). Most of the occupations that are known are concentrated in a time near the late Mousterian: Mirón, Covalejos, Arrillor, Sopeña, Morín, Amalda and Axlor (Hoyos et al., 1999; Sanquino and Montes, 2005; Straus and González Morales, 2001; Maroto et al., 2011, etc.). An attempt has been made to explain this circumstance by the limitation of the radiocarbon method itself (Santamaría and de La Rasilla, 2013), as the risk of rejuvenating the result increases exponentially with an increase in the age of the sample. The litho-stratigraphic study divided the sequence into four sections according to their composition and the agents involved in their formation, by integrating taphonomic and geo-archaeological criteria, X-Ray diffraction, environmental scanning electron microscopy, and thermoluminescence (Jordá, 2008). From bottom to top, these four sections are ESQ-D (Levels XXXI to XLI), ESQ-C (Levels XII to XXX), ESQ-B (Levels I to XI) and ESQ-A (covering breccia and speleothems; the cave, which was partially filled, was initially sealed by this large formation; Jordá Pardo et al., 2009). The human occupation is located in the central Units B and C. Unit C coincided with sedimentation in the rock-shelter by diffuse run-off; the upper central Unit B consists of clasts and frost-shattered rocks
84  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  diffe...
CANTABRIAN MOUNTAINS AND COASTLINE. with a contribution of diffuse run-off and heavier flow at the top. The characteristics of the matrix in the upper section, according to some authors, would have favoured the vertical movement of contaminating material (Santamaría and de La Rasilla, 2013). The micro-morphological analysis (Mallol et al., 2010) confirms the good state of conservation in Unit B, although it is affected in its internal structure by cryoturbation, and moderate diagenetic action in Unit C (with levels revealing intense anthropic action). In this way, in neither of these sections have any general processes been detected that might have altered the deposit in any considerable way. Below these, possibly 9m of barren sedimentation is characterised by a massive clayey matrix with highly altered limestone clasts, formed in endokarst conditions (Mallol et al., 2010). Phytolith analysis (Cabanes et al., 2010) provided surprising results. The sample from the upper litho-stratigraphic section (Unit B) yielded the worst sample, but the moderate diagenesis in Unit C, and the upper levels in Unit B, favoured the conservation of a larger number of remains, which had initially been interpreted as an accumulation of ash (Jordá Pardo et al., 2008). Only the detailed micro-stratigraphic study revealed a succession of hearths, with thick accumulations of ash, burnt bones and artefacts. In this way, it has been suggested that beds of grass were related to a large central area of hearth, with a possible selection of grass species, depending on their properties. The charcoal record is irregular. Acceptable data were obtained from Level XI to Level XVII, although samples were taken as far as Level XXII. Thus, species have been identified for a span of time dated grosso modo between 53 and 36 ky (Uzquiano et al., 2012). The study confirms the general presence of pine in the sequence, although between Levels XIV and XI, the percentage of pine decreases in favour of a greater variety of species, including Sorbus aria, steppe-type scrub, Juniperus, Betula and a large range of shrubs (possibly chosen to light the fires). This decline in pine may be explained by environmental change, which meant that fuel had to be gathered over a larger area and shrubs used in larger quantities (Uzquiano et al., 2012). The data do not seem to indicate that firewood ever became scarce. Pollen data (Ruíz Zapata and Gil García, 2005) confirm the predominance of pine throughout the sequence, and this occasionally alternates with river- bank species in optimal climate conditions, accompanied by herbaceous plants. A brief reconstruction indicates that between the old Levels XXX and XX (>50 ky BP) a possibly colder phase was represented by an open vegetation with Asteracae, Poaceae and Chenopodiaceae, and a significant presence of pine as well as some Betula and Cupresaceae. However, in the phase between Levels XXX and XIV, the hearths noticeably affected pollen conservation and therefore the record consists mainly of pine, birch, Cupresaceae and Asteracae. These levels are dated grosso modo before 39 ky BP. Between 39 and 34.5 ky BP, the palynological study reflects an optimum in humidity conditions, as pine is accompanied by a larger range of mesophile plants, especially Betula. This circumstance, which is found at other Cantabrian sites dated in a similar time, concords with the anthracological data obtained at El Esquilleu (Uzquiano et al., 2010). The archaeozoological record at El Esquilleu agrees with what might be expected in the area (Yravedra, 2006; Uzquiano et al., 2012). Ibex and chamois (Rupicapra rupicapra) are the most common species, while bovids (aurochs) occasionally appear in Levels XIII and XIF, and cervids between Levels V and XIV. The presence of these animals probably denotes connections with areas on the coastal strip. In addition, in the levels in which the seasonality of the hunting could be reconstructed (basically from Level XIV to Level III, owing to lesser fragmentation of the bones), this took place mostly in the summer, with prolongations in late spring or early autumn. Only in Level XI were caprids hunted also in winter (Uzquiano et al., 2012; Yravedra, 2006), which may be interpreted as a change in the seasonality pattern and a more residential use of the cave. In general terms, therefore, it may be affirmed that the environmental data from Level XI to Level VI indicate that the site was more closely integrated in the environment, the occupation was more stable seasonally and there was a larger range of prey, occasionally including species not found in the immediate surroundings, and a wider and more varied plant catchment area. The pollen data indicates greater humidity and a certain expansion of trees in addition to pine. The provenance of raw materials also supports a changing strategy in the use of the environment in the central levels, in contrast with the lower occupations in Levels XIX to XVI. The raw materials were transported over distances of up to 30km, with connections to the 85
CANTABRIAN MOUNTAINS AND COASTLINE.  with a contribution of diffuse run-off and heavier    ow at the top. The characterist...
86 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD coast, in lithologies involving fragmented chaînes opératoires, suggesting great mobility. From Level XIV onwards, the selection of raw material increased (quartzite cobbles from secondary aggregate deposits; Manzano et al., 2005) and the cave acted as a central place. El Esquilleu and other sites like El Habario and El Arteu (Carrión, 2002) were then operating in a functionally combined way. However, from Level VIII, the procurement used new deposits and lithologies in the Deva valley to the south of site, and to the south-east and tributaries, and the site increasingly made use of local resources towards the end of the sequence. very few anthropic marks (Yravedra and González Castanedo, 2014). The importance of carnivore action in the formation of the deposit has equally been seen at another important site in northern Spain (Level VII in Cueva de Amalda), and is defined both by distinctive marks and the differential presence of skulls and distal limb bones. This would directly explain the presence of many of the caprids in these levels (Rupicapra rupicapra and Capra pyrenaica in the case of El Esquilleu), in comparison with larger animals (such as equids and red deer) whose origin is thought to be necessarily anthropic (as in any case they would be smaller carnivores, like hyenas or foxes). In the upper levels, a detailed t phonomical study indicates an increase in carnivore activity from Level V onwards, especially in Levels IV and III, to such an extent that they may have become the main accumulating agent. The bone material in Level III displays The detailed study of the material found in the hearths, which mostly appear in the litho-stratigraphic Unit C in El Esquilleu, is also significant. Hearths have been identified in Levels XXI, XXIII, XXV, XXVIII and XXIX. Charcoal from the hearth in Level XXI was dated by TL to 53,491 ± 5114 BP. This is the oldest date in the series, corresponding to the earliest occupation. The number of burnt bones and the degree of fragmentation gradually increases from Level XIV to the base of the sequence. The degree of combustion of the bones is high and quite unmistakeable evidence of intentional breakage is occasionally seen (Level XXI), indicating the bones were used as fuel in the hearths (Yravedra and Uzquiano, 2013; Uzquiano et al., 2010). Although the pollen data reflects conditions of open vegetation with a predominance of grasses, the anthracological study showed that there would not have been an extreme scarcity of plant matter or of species for lighting the fires (for which Ericaceae and Fabaceae were used at El Esquilleu). Nor is the active selection of the more favourable bones for fuel seen at the site. This suggests that it was a hygienic habit, in which organic waste was thrown on to the hearths (Yravedra and Uzquiano, 2013). Figure 3. Lithic artefacts from El Esquilleu (I). Levels III to VI, (II) Levels VII to IX, (III) Levels XI to XV, and (IV) lower levels. (Drawn by E. Carrión) Practically all the levels have yielded remains of microfauna, except the three lowest levels (XVIII to XXX) and Levels XVI, XXI and XXIII, affected by diagenesis and solution processes. Level VIII has provided the largest assemblage (Sesé in Baena et al., 2005; Sesé in Uzquiano et al., 2012). From the environmental viewpoint, none of the taxa in the sequence indicate extremely cold conditions. Between Levels XXVI and XIV, only indeterminate rodents are found (arvicolids). However, from Level XI onwards, a larger range of species (Eliomys quercinus, Pliomyslenki, Microtus cf. lusitanicus, Microtus arvalisagrestis, etc.) reflects the more temperate conditions between 34.3 and 36.5 ky BP, as detected in pollen
86  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  coast...
CANTABRIAN MOUNTAINS AND COASTLINE. sequences at other sites in Cantabrian Spain, like Covalejos and Labeko Koba. Finally the study of the archaeological material has provided an important sequential collection of materials, enabling a unique techno-economic study which surely corresponds to adaptive processes within the environment around the site. In this sense, a surprisingly close correlation is found between the dominant technical sequences, the investment in energy in the changing models of raw material pro- Ramón Montes Barquín* The stratigraphic sequence of the El Pendo cave (Escobedo de Camargo, Cantabria, Spain) 1. Introduction The El Pendo cave was discovered for science in 1878 by Marcelino Sanz de Sautuola (1880). Since then it has been the object of numerous archaeological actions, such as those performed by Juan Vilanova y Piera (at the end of the 19th century), J. Carballo and B. Larín (during the first decades of the 20th century) and Martínez SantaOlalla (the excavations of 1953-1957), among many others. Three monographs mark the research into the cave up to the present day: that published by Carballo and Larín, in 1933, that of J. González Echegaray (1980) regarding the excavations carried out by Martínez Santaolalla, and that published by Montes and Sanguino (2001) about the actions performed between 1994 and 2000, during which the cave paintings were discovered. To these can be added numerous works on partial aspects and references in joint studies. The international importance acquired by El Pendo is due, without doubt, to the International Field Archaeology Courses designed and delivered as a consequence of the excavations of 1953-1957 under the direction of the Spanish archaeologist Julio Martínez Santa-Olalla, courses that led to European and North American archaeologists developing an * curement, and the climate fluctuations in the environment, especially in the last parts of the sequence (Baena et al., 2012). In addition, the artefacts display great techno-typological development (Fig. 3), especially in the levels with a predominance of the Quina (Levels XX and XV-XIf), Levallois (Levels IXI and X-VIII) and discoid (Levels XXX-XXI and VI-III) techniques. Level XVII displays a clearly different nature with certain predominance of blades. In all cases, however, the dominant schemes are found together with secondary ones. intense research activity. Leading figures from the world of prehistoric archaeology of the time, such as Cheynier and the husband and wife team of Leroi and Gourhan, personally directed the team on the excavation of the stratigraphic deposit located inside the cave, in which up to 18 different strata were documented and –in theory– covered the period from the onset of the Middle Palaeolithic to the Bronze Age. Unfortunately, these studies were not published at the time, and it was Joaquín González Echegaray who, in 1980, published a report on the excavations and a scientific interpretation of the site based on different studies, the geology section of which was written by K.W. Butzer (1980). The interpretation, theoretically generous and at the same time very possibly frugal as far as practicality is concerned, put forward by K.W. Butzer (1980; 1981) regarding its stratigraphic column transformed El Pendo into a site of iconic proportions (on a par with the sequences of El Castillo and Cueva Morín) when it comes to establishing the general sequence of the Cantabrian (and Iberian Peninsular) Palaeolithic. However, in recent years Butzer’s idyllic interpretation has been seriously questioned and doubt cast on the value of this sequence (Montes and Sanguino, dirs. 2001; Montes et al., 2005). Moreover, these latest studies have redefined the series, which would, Unidad Técnica del Itinerario Cultural del Consejo de Europa Caminos de Arte Rupestre Prehistórico. RCDR C/San Martín del Pino, 16 – 3 bajo. 39011 Peñacastillo-Santander. rmontes@prehistour.eu. 87
CANTABRIAN MOUNTAINS AND COASTLINE.  sequences at other sites in Cantabrian Spain, like Covalejos and Labeko Koba. Finally...
88 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD in reality, consist of a total of 33 levels, with the base layer being subjected to a whole host of dating methods that not only show the existence of remains of up to 84 ky BP, but also the presence of severe chronocultural inversions that represent a clearly anomalous post-depositional accumulation when it comes to establishing a sequence of some interest (a fragment of spear –or rod– that is clearly post-Palaeolithic in appearance was found at the base of the sequence…). 2. From Classical Stratigraphic Benchmark To Inconsistent Erudite Reference. The mouth of the El Pendo cave is situated in the village of Escobedo, which in turn forms part of the municipal district of Camargo (Cantabria, northern Spain). It lies within the karst massif of Alto del Peñajorao, one of the region’s many minor coastal ranges. The landscape in the area is relatively gentle and rolling, typical, in other words, of the coastal fringe of northern Spain, being composed, as it is, of Mesozoic materials from the Lower Cretaceous and forming part of the so-called Urgonain Complex. The limestones that interest us here have been attributed to the Aptian Age. The karstification of the limestone has given rise to numerous single and compound sinkholes. As it happens, the El Pendo cave is situated in the north face of one of the many compound sinkholes to have formed in this system. A precise topographic survey (carried out by Luque in 2001) places the floor of the cave’s entrance at a height of 90 metres above sea level, while the height of the surface area of the place where the archaeological digs carried out between 1953 and 1957, and the more recent excavations of 1994 to 1997 were started, is 71 metres above sea level (representing a 19-metre fall over the barely 35 linear metres between the cave entrance and the dig site). The “classical” sequence (namely the 18 levels identified by Butzer) is based on a series of cuts made in the left-hand side of the enormous rubble cone that starts a few metres in front of the fence erected at the back of the cavity’s vestibule, extends over 80 metres beyond the aforementioned barrier (Fig. 1) and ends just before reaching the area in which the red paintings were discovered during the 1997 dig. González Echegaray (1980) and his collaborators interpret the sequence in accordance with the following series of human occupations: Bronze Age (level 0), Azilian (level I), the end of the Magdalenian (level II), “late Aurignacian” (levels III and IV), Gravettian (V and Va), “evolved Aurignacian” (VI), “Aurignacian I” (VII), “lower Perigordian (VIII), “archaic Aurignacian” (VIIIa and VIIIb), “denticulate Mousterian” (VIIId), Mousterian (IX and X), “denticulate Mousterian” (XI and XII), “typical Mousterian” (XIII and XIV), Mousterian (XV), “denticulate Mousterian” (XVI), “Unidentified and yet to be identified industry (XVII and XVIII)”. Figure 1. Location and footprint Regarding the characteristics and interpretation of this deposit over time, we can refer to Montes et al (2005). The most recent interpretation of this sedimentary deposit highlights the existence of a broad range of post-depositional processes, which are visible and may be documented in the cuts made during the excavations carried out, and concludes that, whatever the case may be, the area in which digs were undertaken in the 20th century and where the “classical sequence” was documented is nothing more
88  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  in re...
CANTABRIAN MOUNTAINS AND COASTLINE. Figure 2. Cross-section of the cave than a monumental rubble cone brought in from an original accumulation area most likely situated in the Pleistocene vestibule of the grotto (Fig. 2). In other words, what we have here is a derived site with an infinite number of geo-archaeological problems that render it invalid as a reference for the establishment of a sequence of any chronostratigraphic value. action of the stream that evacuates the compound sinkhole, of which the cave system is a natural drain (currently, the grotto’s active watercourse is documented as being at around 20 metres below the mouth used for accessing the site). In fact, the current floor of the sinkhole is above many of the levels that form the base section of the site. Few doubts currently remain regarding the fact that the general sedimentological dynamic of the site has, to a great extent, been conditioned by the The calcareous elements that make up the deposit would appear to have come from rock falls caused by thermoclastic processes, but these would not have 89
CANTABRIAN MOUNTAINS AND COASTLINE.  Figure 2. Cross-section of the cave  than a monumental rubble cone brought in from an...
90 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD occurred on the vertical plane of the sequence. The only blocks that could have fallen on this would have been those loosened from the roof of the cave by processes of chemical solution rather than cracking and fracturing. The mechanical detachment point would have to be brought back to the area close to the mouth of the cave. Finally, the alteration of the detrital elements would have been caused not only by post-sedimentary processes, but also by the length of time these elements were exposed to the weather and the climatic conditions that prevailed while they were being transported. Three principal agents were involved in the formation of the sedimentary deposits of El Pendo cave: the slope of the accumulation cone that protrudes from the rock shelter at the mouth of the cave, the stream that drains the compound sinkhole and the topography of the cone. These three elements have acted together, with one or other of them being more influential at different times, to create the deposit we know today (Fig 2). a) The slope. Starting in the vestibule of the cave is a large debris cone made up of big blocks that stretches into the cave’s interior to form an accumulation cone. The distance from the point at which the slope begins in the vestibule of the cave to the area containing the main sequence column of the site is 34.6 metres, with the gradient being 44 degrees. Given the steepness of the slope, the action of gravitational flows is extremely feasible. b) The stream that drains the sinkhole of El Pendo cave acts as a very effective transportation agent, capable of dislodging the detritic materials contained within it. Inside the cave it currently flows beneath the archaeo-sedimentary deposits, but its height at the point at which it disappears is a few metres above the levels of the shaft, and this was shown by the topographical survey. As was seen during the heavy rainstorm of 1983, under certain circumstances the stream is not capable of evacuating all the water that accumulates in the small basin of the sinkhole. This results in it overflowing and using the upper level of the cave, where the geoarchaeological deposit is situated, to drain off the water. The sedimentary record provides irrefutable evidence of other times when the water carried by the stream penetrated the cave. c) Finally, the topography of the cone has determined the routes taken by the flows and the areas in which the materials have accumu- lated. The rocky escarpment of the entrance has undergone considerable chemical and mechanical alteration processes. The fallen materials have formed a rubble cone. These fallen blocks have changed the directions in which the waters flow, thereby causing lateral changes in the storm water runoff, which might explain the deposition of some materials in one area of the cave and not in others. These three agents serve to develop the interpretation of the depositional sequence of the El Pendo cave and the sedimentary factors that have resulted in the stratigraphy that we can see today. The study of the materials recovered during the digs carried out between 1994 and 1997 on the levels of the base of the sequence (in “the shaft”), especially of those obtained from levels 25 and 26 (which provided a significant number of finds), also provided data that showed them to be clearly inconsistent and, therefore, an unreliable chronocultural record. Apart from a few anecdotal, and at the same time tremendously symptomatic questions, such as the discovery of a fragment of spear point, or rod, in level 32, or of lithic finds that have been dubiously ascribed to the Mousterian technocomplex in many of the sequence base levels (theoretically dated as pertaining to the Middle Palaeolithic), the internal study of the lithic series, based on the operational sequence analysis methodology, supports the existence of notable anomalies. The palynological and paleontological data also revealed countless anomalies, in fact far too many to mention in detail here (Montes, Sanguino, dirs., 2001). In addition to geological, paleoecological and archaeological studies, the most recent excavations included the use of absolute dating methods such as ESR, Thermoluminescence and Uranium/Thorium series, although Carbon 14 was not used, given the purported chronology of the lower sequence being worked on. The programme of absolute dating methods was, without doubt, essential for establishing the hypothesis that what we have here is a monumental archaeological misunderstanding. Its results clearly show the inconsistency of the El Pendo sequence that the geological data provided in such an overwhelming manner (Fig. 3). 3. The El Pendo site. A current assessment Unfortunately, the sedimentary deposits of the El Pendo cave brought to light by the excavations of 1953-57 are the result of a major post-depositional that includes three elements: the slope of the rubble cone in question, the stream that
90  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  occur...
CANTABRIAN MOUNTAINS AND COASTLINE. drains the sinkhole –at the bottom of which lies the cave– and the ever-changing topography of the rubble cone in which the different cuts have been made. The entire sedimentary dynamic that has given rise to the depositional sequence can be explained by the action of these three agents. In the geo-archaeological studies undertaken following the theory put forward by K.W. Butzer (Hoyos and Laville, 1982; Montes and Sanguino, 2001; Montes et al., 2005), a great deal of evidence has been compiled and documented regarding a source of energy powerful enough to carry and transport huge quantities of materials and inflict a great deal of erosion while doing so. Episodes involving the total or partial flooding of the grotto have also been documented. The analysis of the 33 levels which, in reality, go to make up the series, and the contacts between them, provides interesting information about the nature and origin of the sequence; this in one way can be interpreted as a geo-archaeological stratigraph that is valid as a benchmark series. Moreover, the different absolute dating methods used have confirmed the existence of significant anomalies as regards the existence of an authentic chronocultural sequence. The study of the archaeological and paleontological materials obtained from the base of the sequence have proved the limited internal cultural coherence of many of the levels. The dynamic interpretation of the stratigraphic levels and the sedimentology shows that it is the slope movement phenomena that are mainly responsible for the formation of the deposit. The description of the processes, level by level, causes a loss of perspective, but if we examine the sequence as a whole, the El Pendo deposit is –in general terms– a macro sequence with positive grain selection. Successive infill processes have made the slope less steep, thereby reducing the potential energy, which explains the positive trend of the grain selection. The sedimentary differences are the result of the varying forces of the gravitational processes. Likewise, most of the contacts between the levels are erosive, which implies, on the one hand, an absence of the sequence’s time record and, on the other, the incorporation of material, whether archaeological or not, from the underlying part of the level above. Several examples of these erosions occur. In short, the stratigraphy of the El Pendo cave cannot be used as a reliable source of knowledge for the Upper Pleistocene in the area excavated in the 20th century. The mixtures of archaeological material from different levels means this site cannot be considered as a benchmark in the paleoclimatic and Figure 3. Stratigraphic cut and absolute dating methods chronocultural sequence of the Cantabrian Region despite the weight of its Historiography. Although different degrees of contamination exist in the stratigraphy, and not all the levels must be treated in the same way, the extremely large body of research to have used data from the El Pendo levels, in any section of its sequence, must be severely questioned, mainly because no guarantee exists of the synchronism of the elements contained by each level. Despite everything, it has also been possible to document the presence of some areas in the exterior shelter (not inside the cave) that would appear not to have suffered major alterations and offer hope for future investigations (Montes and Sanguino, dir., 2001) which, under all circumstances, should be carried out away from the area we know as “the classic El Pendo site”, namely, in the area that is still covered by the debris left by the collapse of the overhang of the shelter. 91
CANTABRIAN MOUNTAINS AND COASTLINE.  drains the sinkhole    at the bottom of which lies the cave    and the ever-changing ...
92 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Lawrence Guy Straus*,** , Geoffrey A. Clark*** La Riera Cave (Posada de Llanes, Asturias) The 1976-1979 excavation of La Riera Cave (Fig. 1), coming a decade after that of Cueva Morín in Cantabria by J. González and L.G. Freeman and done in association with M.R.González Morales, represented the continuation of a international, collaborative effort to modernize Paleolithic research in Cantabrian Spain. These Hispano-American, projects, like those that followed in El Juyo (by Freeman and González Echegaray) and Mirón (by Straus and González Morales) (both in Cantabria), were designed and conducted withexplicitly interdisciplinary,p aleoanthropological,problem-oriented foci that guided excavation methodologies, analyses and interpretations, fundamentally diverging from the solely culture-historical raison d’être of many traditional excavations in Spain and elsewhere. Figure 1. La Riera in its surroundings. * ** *** The complete, monographic publication of the La Riera research (Straus and Clark 1986) has made this one of the most widely cited, extensively debated and frequently restudied excavations in the long history of Upper Paleolithic and Mesolithic research in Cantabrian Spain. A surprisingly small cave, La Riera has nonetheless played an unusually significant role in the development of scientific (and anthropological) archeology in the Iberian Peninsula, and the profession ultimately owes the importance of this site to its discovery, initial excavation and monographic publication to the Conde de la Vega del Sella, in whose footsteps the authors respectfully followed 60 years later. La Riera Cave is located on the narrow coastal strip of eastern Asturias at 43°25’31” N x 4°52’ W x 30 m above sea level. The low cave mouth faces west from the south slope of the low La Llera “rasa” ridge, which runs parallel to the shore west-east between the valley of the Bedón Riverand the town of Llanes. Between this ridge and the steep Sierra de Cuera range (maximum elevation: 1315 m at only 7 km from the shore), there is a depression that is drained via the Calabres stream that runs underground through a karstic system immediately adjacent to and slightly below La Riera. The Calabres resurges at the Niembro inlet 1.5 km north of La Riera and the open ocean coastline is at 1.75 km, while it would have been less than 10 km from the site during the Last Glacial Maximum. Other cave sites in this ridge include Cueto de la Mina (only 50 m from La Riera), Balmori, Tres Calabres, Bricia,etc., giving the area around the town of Posada de Llanes one of the densest concentrations of Paleolithic and Mesolithic sites (including minor cave art loci) in all of Iberia. Department of Anthropology, University of New Mexico, Albuquerque, NM 87131 USA Instituto Internacional de Investigaciones Prehistoricas de Cantabria, 39005 Santander, Spain School of Human Evolution and Social Change, Arizona State University, Tempe,AZ 85287 USA
92  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Lawre...
CANTABRIAN MOUNTAINS AND COASTLINE. The mouth of La Riera was completely filled with archeological deposits capped by a rare intact Asturian shell midden (“conchero”) at the time of its discovery by Vega del Sella probably during his excavation of Cueto de la Mina in 1914-15. The Conde excavated the outer part of La Riera in 1917-18 and published a monograph on it and nearby Balmori in 1930. He uncoverd a sequence of Solutrean, Magdalenian, Azilian and Asturian layers. His Magdalenian clearly included both Lower and Upper phase components, indicated respectively by the presence of nuecleiform endscrapers and quadrangular cross-section geometrically engraved antler points on the one hand and antler harpoons on the other. The Magdalenian horizon also included “archaic-looking” macrolithic artifacts some of which the Conde (improbably) attribued to an Acheulean deposit that had washed into the cave from the slope above the cave at this time –despite that fact that he cited the presence of other similar artifacts among the unquestioned Magdalenian assemblage. La Riera was a key site for the Conde’s definition of a new Mesolithic “culture”, the Asturian– also characterized by the presence of “crude” macrolithic implements (i.e., cobble picks), sometimes misinterpreted by later scholars as also being of Acheulean age (see discussion in Clark 1976, 1983). The talus slope in front of La Riera was tested and a concreted conchero remnant sampled by Clark in 1969 as part of his dissertation research on the Asturian (Clark 1976, 1983). This research yielded two radiocarbon dates that were among the first ever run for this techno-complex, clearly showing it to be postPleistocene and pre-Neolithic in age. The 1970s excavation aimed to gather and analyze data to reconstruct the environments of the late Last Glacial and early Postglacial and to use artifactual and faunal evidence to elicit information about and to suggest explanations for variations in hunter-gatherer adaptations as their uses of the cave changed through time against the backdrop of changing conditions. Anthropological hypothesis testing was a keystone of the research and the agenda was explicitly processual, both directors having been students of L.G.Freeman at the University of Chicago and directly and indirectly influenced by the thinking of L.R.Binford and the “New Archeology” of the late 1960s-1970s. In this context, much reliance was placed on radiocarbon (as opposed to Bordesian cumulative percentage graphs of retouched stone tool frequencies) to date levels, while diagnostic artifacts such as Solutrean points, Magdalenian and Azilian harpoon types, certain Magdalenian sagaie forms, and Asturian picks were acknowledged generally to be temporally diagnostic. Consequently La Riera was the first Upper Paleolithic/Mesolithic site in Cantabria (and indeed in Iberia) for which a major investment was made to procure large numbers of 14C determinations–albeit with inevitable contradictions and inconsistencies, due to the number of different labs that did the dating, the use of both bone and charcoal, and the inability at the time to remove as many contaminants as can be done now, and the possibility of mixing by cryoturbation and prehistoric human activities such as hearth pit digging. Other dates on shells (corrected for marine reservoir effect) were later published by A.Craighead (1999), generally confirming the ages of the oldest Solurean, the Upper Magdalenian, the Azilian and the Asturian conchero. The 7-10 m² excavation (of a remnant of intact deposits left by the Conde in the interior of the cave) was done by fine “dissection” of units thought to approximate more or less horizontal “living floors”. This resulted in the definition of 36 levels and lenses in contrast to the Conde’s 4 horizons, and in fact most of the Asturian (save small remnants) had earlier been removed, so that the new excavation mainly sampled the Solutrean, Magdalenian and (more locally) Azilian units, for a stratigraphic thickness of about 1.8 m (plus a 60 cm-deep sondage dug into the basal clay deposit (Levels 1-3). The resulting culturestratigraphic sequence included a pre-Solutrean (“Aurignacian” or, more likely, undiagnostic “Gravettian”) component (Levels 1-3, probably > 20 uncal. BP), several Solutrean layers with shouldered, concave base laurel and willow leaf points (Levels 4-17, 20-17 uncal. kya), Initial and Lower Magdalenian (Levels 18-19, 16.5-15.2 uncal. BP), Upper Magdalenian (Levels 20-24, 13-11.5 uncal. kya), Azilian (Levels 26-27, 11.5-10.5 uncal. kya), and Asturian (9.0-6.5 uncal. BP). Finds were piece-plotted in 3D and all sediments were screened in water through fine mesh (Fig. 2). The following analy- 93
CANTABRIAN MOUNTAINS AND COASTLINE.  The mouth of La Riera was completely filled with archeological deposits capped by a r...
94 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD ses are published in the monograph: sedimentology (H. Laville), speleothems (R. Harmon), palynology (Arl. Leroi-Gourhan), macrobotanicals (K. Cushman), features and lithic technology (Straus and Clark), lithic raw materials (J. Ordaz, L. Suáraz and R. Esbert with Straus and Clark), osseous industry (M.González Morales), mammalian faunas (J. Altuna), fish (M. Menéndez de la Hoz with Strans and Clark), marine mollusks (J. Ortea), oxygen isotopes (M. Deith and N.J. Shackleton), human remains (M. D. Garralda). There are several background/objectives, synthesis, statistical, interpretive/conclusion chapters by Straus and Clark. Some of the main conclusions of the research included: There is no basis for subdividing the Solutrean into artifact-based phases and the late Solutrean intergraded with the early Magdalenian through a process of “desolultreanization”, thus questioning the “reality” of these concepts as separate “cultures”. The role of the site changed notably through time (e.g, from transitory, specialized camp for hunting ibexon the nearby cliffs of the Sierra de Cuera, to major multi-functional residential hub with diversified technologies and features associated with many hunting and gathering activities, plus others such as parietal and portable art creation, to a dump for shells and other bulk garbage); The process of subsistence intensification through both situational specialization and overall diversification, earlier thought to have begun in the Magdalenian, started here with the Solutrean and included the large-scale collection of shellfish, some fishing, and the increasingly intensive hunting of ibex and red deer, including the (ultimately Figure 2. Stratigraphy of La Riera Cave, L.G. Strauss and G.A. Clark, (Eds.), Anthropological Research Papers, 36, Tempe, Arizona, 1986; G.A. Clark.
94  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  ses a...
CANTABRIAN MOUNTAINS AND COASTLINE. counter-productive) taking of ever larger numbers of young individuals, as well as the addition of woodland-adapted species (boar, roe deer) as Late Glacial and early Post-glacialconditions permitted. There was no clear relationship between some of the changes in site use and major climatic shifts, but demographic pressure was seen as a key motor in driving subsistence change and presumably many of the main technological developments especially in the area of weap- David Santamaría*, Elsa Duarte*, María González-Pumariega**, Lucía Martínez*, Paloma Suárez*, Javier Fernández de la Vega*, Gabriel Santos***, Tom Higham****, Rachel Wood*****, Marco de la Rasilla* It is a large rock shelter with a surface area of approximately 225 m2 and ~30 m long (Fig. 1), which preserves a wide stratigraphic and cultural sequence dated in the Middle and Upper Paleolithic (Fig. 2) and many parietal engravings that are partially covered by the stratigraphy (Fortea, 1994). ** *** **** ***** In short, despite undoubted shortcomings, La Riera helped significantly to “change the nature of the conversation” about the meaning of inter-assemblage and even inter-period variability in Spain for a generation of researchers. La Viña rock shelter (Asturias, Spain) The site is located in La Manzaneda, 9 km south of Oviedo, in the middle basin of theNalónriver. Facing S-SE, the rock shelter opens up inaVisean-Namurianlimestone measuring ~200m long and ~30m high and around 500m from the right bank of the river Nalón. The UTM30 ETRS89 coordinates of the site are X = 270725.79Y = 4799477.68 Z = 292 metres above sea level. * onry (e.g., Solutrean points, antler sagaies with backed bladelet inserts, antler harpoons, and the presumed invention of traps, nets and maybe weirs, as well as of new, more efficient hunting strategies and tactics). The shelter was discovered in 1978 by A. J. Gavelas (1981) and a little later the Prehistory Department of the University of Oviedo surveyed the rock shelter, confirming its archaeological and artistic interest. The excavations, directed by J. Fortea and integrated in the Proyecto de Investigación Nalón medio (Middle Nalón Research Project), started in 1980 and lasted through to 1996 (Fortea,1981, 1990, 1992, 1995, 1999 and 2001). The cultural episodes, the physical and biological environments, the rock engravings, the portable art and the mineral pigments are currently being studied. The archaeological interventions focused on two sectors of the shelter, called central and western, coinciding with the areas where the parietal Área de Prehistoria. Departamento de Historia. Facultad de Filosofía y Letras. Universidad de Oviedo, c/ Teniente Alfonso Martínez s/n, 33011, Oviedo (España) santamariadavid@uniovi.es duarteelsa@uniovi.es lucia_satis@hotmail.com psuarez.ferruelo@ gmail.com mrasilla@uniovi.es Consejería de Educación, Cultura y Deporte del Principado de Asturias, Apartado de correos nº 29, 33590 Ribadedeva. maria. glez-pumariegasolis@asturias.org Departamento de Ingeniería Cartográfica y del Terreno, Facultad de Ciencias, Universidad de Salamanca, Pza. de la Merced s/n 37008 Salamanca (España) gsd@usal.es Research Laboratory for Archaeology and the History of Art, University of Oxford, South Parks Road, Oxford OX1 3QY, United Kingdom thomas.higham@rlaha.ox.ac.uk Research School of Earth Sciences, Australian National University, 1 Mills Road, Canberra 0200, Australia rachel.wood@anu.edu.au 95
CANTABRIAN MOUNTAINS AND COASTLINE.  counter-productive  taking of ever larger numbers of young individuals, as well as th...
96 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. Floor plan and stratigraphic section of the western sector engravings are concentrated, and on two cores next to the wall of the shelter, which preserve remains of settlements posterior to level IV, which forms the current floor. Central Sector. Located between lines10 and 14, its surface area is approximately 30 m2. The whole area was excavated up to the topof level V and to the bedrock atline14. It presents a wide stratigraphic and cultural sequence. From top to bottom: Strata I to III. Identified in Core 4, next to the north wall of the shelter (squareH-15).Attributed to the Tardiglacialand the Holocene (Fortea, 1990). Stratum IV. Attributed to the Middle Magdalenian. Dated at 13,300 ± 150 (Ly-3317) and 13,360 ± 190 BP (Ly-3316) (González-Morales et al., 1989; Fortea, 1990; Duarte, 2010). Contact with the underlying unit is erosional unconformity. It is very rich in lithic and bone industry and portable art.
96  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figur...
CANTABRIAN MOUNTAINS AND COASTLINE. Stratum V. Sub-divided into several levels in the area not covered by the overhang. Classified as Upper Solutrean with notched and concave base projectile points (Fortea, 1990). Stratum VI. Sub-divided into three levels (VIa, VI band VIc). Level VI awas assigned to aphase previous to the UpperSolutrean –Middle Solutrean– and levels VI band Vic to Gravettian, the latter, VIc, with Noailles burins (Fortea, 1992). Stratum VII. Collapse of the overhang; only outside the shelter. Sterile from an archaeological point of view. Stratum VIII. Attributed to the Aurignacian (Fortea, 1995). Stratum IX. The techno-typological analysis of the lithic industry puts it in the oldest Aurignacian (Suárez, 2013). Stratum X. Classified as undetermined initial Upper Paleolithic (Suárez, 2013). Few lithic remains. Stratum XI. Limestone bedrock. Archaeologically sterile. Western sector. Located in lines 23 and 27, next to the western wall of the shelter. Its surface area is approximately ~17 m2. It was open excavated up to the topof level V and to the bedrock at squares F-25 to F-27 and G-25 to G-27. It also presents a wide stratigraphic and cultural sequence partially dated by radiocarbon, conventional and AMS (Santamaría, 2012; Wood et al., 2014, Tab. 1). From top to bottom: Stratum I. Identified in Core 1. From the Holocene, it presents an industry with a low diagnostic typology. Stratum II. Also identified in Core 1, without cultural assignment (Fortea, 1990). Stratum III. Excavated in lines I and J and in Core 1. Provisionally classified as Upper Magdalenian (Fortea, 1990). Figure 2. Selection of the archaeological materials. 1-5. Magdalenian, 6-8. Solutrean, 9-12. Gravettian, 13-14. Aurignacian, 15-16. Mousterian. Drawings: 1-9, 14 and 16 E. Duarte, 10-12 L. Martínez, 13 and 15 D. Santamaría. tected a few Noailles burins, so this level best fits in to an advanced Gravettian phase. Stratum VIII. Gravettian with Noailles burins, microgravette pointsand pedunculated points similar to the shape of the Font-Robert point (Fortea,1992). Stratum IV. Corresponds to stratum IV of the central sector. Attributed to the Middle Magdalenian (Fortea,1990, Duarte, 2010). Stratum IX.Gravettian with Noailles burins, microgravette points, many burins and a knapped calamite fossil deliberately modified (Fortea, 1992; Martínez and Rasilla, 2013). Dated at 24,680 ± 130 BP (OxA-21688). Stratum V. Similar to V in the central sector. Attributed to the Upper Solutrean (Fortea,1990). Stratum X. Gravettian with Noailles burins (Fortea, 1992; Martínez and Rasilla 2013). Stratum VI. Middle Solutrean with points á face plane (unifacial flat retouch) and laurel-leaf points (Fortea,1990, Fernández de la Vega and Rasilla, 2012). Stratum XI. Aurignacian with keeled and nosedendscrapers, busqué burins and Dufour subtype Roc-de-Combebladelets. This level belongs to the late Aurignacian (Martínez, 2010) and has been dated between 27,900 ± 280 (OxA-X-2290-19) and 30,600 ± 370 BP (OxA-21687). Stratum VII. End of the Gravettian with Gravette points, microgravettes points and backed bladelets (Fortea, 1992). Current research has de- 97
CANTABRIAN MOUNTAINS AND COASTLINE.  Stratum V. Sub-divided into several levels in the area not covered by the overhang. C...
98 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Stratum XII. Aurignacian with keeled and nosedendscrapers, busqué burins, Aurignacian blades and many Dufour subtype Roc-de-Combebladelets. Ascribed to the late Aurignacian (Santamaría, 2012). Two antler points of flat section were recovered from F-26. Dated at 31,500 ± 400 (OxA-21689) and 31,600 ± 400 BP (OxA-21678). Stratum XIII.This level overlaps with level XIII basalin squaresG-25 to G-27 and XIII infin line F. Assigned to early Aurignacian (Santamaría, 2012), with many keeled endscrapers, some Aurignacian blades and Dufour subtype Dufour bladelets. A split base point ofelliptic section was recovered in square F-27. The middle and top section of this level have been dated between 30,650 ± 360 (OxA21845) and 31,860 ± 680 BP (GifA-95463). The interfaces XIII-XIII basa land XIII-XIII infhave been dated at 35.800 ± 1000 (GifA-95550) and 36.500 ± 750 BP (Ly-6390), respectively. Stratum XIII inf. First Aurignacian occupation of the site. This unit was partially deposited over strata XIII basal, XIV*, XIV and IA, reaching the bed rockin some areas of the sector. The contact between this unit and the underlying units (IA, XIV, XIV* and XIII basal) is erosional unconformity. Classified as polymorphic Proto-Aurignacian with microlaminarprismatic cores, keeledendscrapers, and Dufour subtype Dufour bladelets (Santamaría, 2012). Only the contact with the overlaying strata XIII (Ly-6390, vid supra) has been dated. Stratum XIIIbasal. Last Mousterian level of the site. Only preserved in line G and sub-squares 3, 6 and 9 of F-25 and 1 of F-26. Towards the south the Nº Inv Lab Ref. Level Pret. BP Date cal BP Date From To VI-1 Ly-3317 IV C 13300±150 15986±223 16453 15515 VI-2 Ly-3316 IV C 13360±190 16075±285 16664 15469 VI-45 OxA-21688 IX UF 24680±130 28716±152 29016 28403 VI-35 OxAX-2290-19 XI UF 27900±280 31800±376 32586 31190 VI-36 OxA-21686 XI UF 20820±130 - - VI-39 OxA-21687 XI UF 30600±370 34581±342 35285 33922 VI-83 OxA-19195 XI ABA 30130±170 34180±176 34541 33848 VI-72 OxA-21678 XII UF 31600±400 35521±415 36309 34734 VI-73 OxA-21689 XII UF 31500±400 35431±409 36231 34671 VI-4 Ly15/OxA-4092 XIII C 19930±220 - - VI-5 GifA-95463 XIII C 31860±680 35976±828 37846 34545 VI-67 OxA-21705 XIII UF 31160±230 35073±248 35582 34615 VI-68 OxA-21845 XIII UF 30650±360 34621±339 35320 33967 VI-6 Ly-6390 XIII-XIII low C 36500±750 41043±676 42341 39676 VI-7 GifA-95550 XIII-XIII bas C 35800±1000 40412±961 42230 38560 VI-8 GifA-99230 XIII bas C 48100±1600 50650±1820 54424 47305 VI-9 GifA-99231 XIII bas C 37700±590 - - - VI-10 GifA-95537 XIII bas C >39000 - - - VI-85 OxA-19144 XIII bas ABOx >59300 - - - VI-85 OxA-19196 XIII bas ABA >62000 - - - VI-11 GifA-95551 XIV* C >39000 - - - - - Table 1. La Viña dating. In bold, the anomalous dating values. Pre-treatment (Pret). UF: Ultrafiltration. ABA: acid/base/ acid. ABOx: acid/base/oxidation. C: Conventional. The Columns age cal BP, From and Until they are calibrated with the OxCal programme based on the 2013 IntCal curve (Bronk Ramsey, 2009, Reimer et al., 2013).
98  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Strat...
CANTABRIAN MOUNTAINS AND COASTLINE. level forms a wedge changing to XIIIinf. The contact between these units is erosional unconformity. It is Mousterian, rich in lithic industry, with many different types of sidescrapers, some Mousterian and Tayac points, with quite a few denticulate tools, a good Chatelperron point and four bifaces with a transverse edge (Santamaría, 2012). Dated at >62,000 BP (OxA-19196). Stratum XIV*. This unit is preserved in lines25 and 26 of the western sector. It does not overlap with XIV, but both outline a pseudo-horizontal linein section–with similar top and bottom heights. This unit is partially deposited on IB, IA and XIV in erosional unconformity. Mousterian with more sidescrapers than denticulate tools and some Mousterian points (Santamaría, 2012). Dated at >39,000 BP (GifA-95551). Stratum XIV. This unit only appears in lines 26 and 27. Mousterian with more denticulate tools than sidescrapers (Santamaría, 2012). Stratum XV/IA. Present in lines 25 and 27 of the western sector. Mousterian with more denticulate tools than sidescrapers (Santamaría, 2012). Stratum IB. Preserved in sub-squares 7, 8 and 9 of G-25 and 1-2 of G-26.Mousterianwith few lithic tools and a similar proportion of sidescrapers and denticulate tools (Santamaría, 2012). Stratum RA. Altered bedrock. Archaeologically sterile. Alvaro Arrizabalaga * María-José Iriarte-Chiapuso * * The parietal engravings are spread across five areas along the large rock wall, all of them exposed to direct sunlight (González-Pumariega,2013). Several engraving assemblages are currently unearthed c which were documented as the excavations advanced. These have been arranged into two graphic horizons that are successive over time (Fortea, 1994): the first and oldest, associated to the Aurignacian, consists of deeply engraved lines, arranged rhythmically into various groups and the second, Gravettian-Solutrean, consists of figurative (mainly deer) and non-figurative art. Labeko Koba (Arrasate, Gipuzkoa) 1. Introduction The archaeological site of Labeko Koba is located on the southwestern edge of Gipuzkoa, along the upper course of the river Deba and in a highly anthropic environment (almost in the town of Arrasate). Like the whole of Gipuzkoa, it is located in a geographic area that is a mixture * ** The erosional processes identified in the lower section of the western sector, from IB to XIII, have aided vertical and lateral displacement of archaeological materials between the Mousterian and Aurignacian levels. These contaminations have played a significant role in the techno-type configuration of the series studied, generating a local (i.e. culturally linked to the local Mousterian) and gradualtransition from Middle to Upper Palaeolithic, recognisable from at least IA but whose origin is strictly taphonomic or post-depositional. These contaminations are very evident in the Mousterian levels and less obvious or more elusive in the Aurignacian levels, clearer in XIII inf than in XIII, and undetectable in XII (Rasilla and Santamaría, 2011-12, Santamaría, 2012). of Cantabrian, Pyrenean and Aquitaine environments. It was excavated between September 1987 and December 1988, under exceptional circumstances, which could be referred to as for salvage purposes rather than for emergency purposes. The work method was adapted to a certain extent to ensure complete removal of the site before the cavity was destroyed by the Arrasate ring road but Universidad del País Vasco (UPV-EHU). Tomás y Valiente s/n. 01006 Vitoria. alvaro.arrizabalaga@ehu.es Universidad del País Vasco IKERBASQUE (UPV-EHU). Tomás y Valiente s/n. 01006 Vitoria. 99
CANTABRIAN MOUNTAINS AND COASTLINE.  level forms a wedge changing to XIIIinf. The contact between these units is erosional...
100 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD included the guarantees typical of this type of deposit. All of the sediment recovered while the excavation was sieved with water, allowing systematic recovery of almost one hundred percent of the assemblages at the site. In fact, the absolute dating values, the fauna spectrum identified (Fig. 1), the paleoenvironmental analyses and the lithic and bone techno-complexes allowed, in the case of Labeko Koba, the characteristics of the human settlements of the site to be defined quite well.In this respect, it could be said that Labeko Koba is one of the main sites corresponding to the initial Upper Palaeolithic excavated in the Cantabrian environment in recent decades. Sometime after the excavation, advances in the monograph (Arrizabalaga, 1989a, 1989b, 1991, 1992, 1993), a monograph with various analytical studies (Arrizabalaga and Altuna –eds.–, 2000) and a brief summary of its content (Arrizabalaga et al., 2000, 2002, 2003) were published. 2. Circumstances and dating of the deposit The post-depositional circumstances have significantly altered the archaeological remains. The majority of the surface of the site is not protected by the original overhang, so the remains deposited there have been heavily washed by the rain and exposed to the elements. This degradation particularly affects the bone (both fauna and industry) and pollen remains and significantly alters the lithic record. In fact, this circumstance imposes higher restrictions on reading and interpreting the site, much higher than those imposed by the fact that it is an excavation that has been carried out without interruption over a long period. To highlight a particularly symptomatic detail, the main stratigraphic differences of the site occur laterally (depending on whether they are located under the overhang or outside it) compared to the vertical ones, which are more attenuated (especially in the Figure 1. Sample of the fauna recovered. Bone industry. Levels VII-IV (Proto aurignacian and Lower Aurignacian).
100  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  incl...
CANTABRIAN MOUNTAINS AND COASTLINE. area outside the protection of a small rock shelter). Radiocarbon dating of the site levels has been difficult due to the taphonomic characteristics of the deposit, whose sediment has been heavily washed by rain. The dates initially available were much more recent than those estimated for the early Upper Palaeolithic of the region: Level lower IX (Châtelperronian) 34,215 + 1265 BP (Ua 3324); Level VII (Proto-Aurignacian) 31,455 + 915 BP (Ua 3321); Level V (Early Aurignacian) 30,615 + 820 BP (Ua 3322). An article was published recently (Wood et al., 2014) that presents nineteen new dates for the Labeko Koba sequence, which are much more consistent with one another and with the new regional framework for the early Upper Paleolithic after pre-treating the radiocarbon samples (in this case, ultrafiltration). In accordance with this, the dates that have not been calibrated that establish the Châtelperronianinthe site would be 38,100 ± 900 BP (OxA-22562) and 37,400 ± 800 BP (OxA22560); the Proto-Aurignacianwould cover 35,250 ± 650 BP (OxA-21793) to 36,850 ± 800 BP (OxA21766) and the three Early Aurignacian levels, from 35,100 ± 600 BP (OxA-21778) for level VI, from 34,750 ± 750 BP (OxA-21767) to 34,650 ± 600 BP (OxA-21779) for level V, and from 33,600 ± 500 BP (OxA-21768) to 33,550 ± 550 BP (OxA-21780) for level IV. Calibration, as demonstrated in the aforementioned article, shows additional aging of these results by between four and six millennia. Therefore, the probability curve of the oldest dating values clearly exceeds 44,000 cal BP and the most recent dating values, at least 37,000 cal BP. These dates are significantly more consistent with the new regional framework established for SW Europe (France, Italy, Germany). We must add that almost all of the new dating values have been made on bone retouch in order to minimise the risk of dating bone remains brought in by predators or scavengers that therefore do not have any direct relationship with human activity. 3. The stratigraphic sequence of Labeko Koba and its archaeological refit We have evidence of initial visits to the cave by species that would alternate with one another throughout the occupation sequence (various carnivores and humans). Before the entrance that we excavated started to settle, at the bottom there was a pit cave that accessed the lower red karstic of Labeko Koba. In the top section of this pit cave and at the base of the debris cone that originated in it, in 1973 and 1987 three batches of material were recovered (the majority archaeozoological), named Sima (pit cave) (1973), Derrubio Superior (Upper Debris) (1987) and Derrubio Inferior (Lower Debris) (1987). The presence of a fragment of a Châtelperron points in the Sima materials, a section of a burin (in Derrubio Superior) that can be traced back to a burin in sub-level lower IX and the fauna associations in both batches indicate a precise chronological identity between these materials and the assemblage in level IX. It is therefore probable that these materials with no stratigraphy correspond to initial visits to Labeko Kobafrom hyena, cave bear and humans, using the pit cave as a den in the case of the carnivores. Human presence, which is very occasional, is more difficult to explain but could be interpreted in relation to the procurement of biotic resources (meat, skin, antlers, bones, etc.) present in the pit cave through the activity of carnivores, the leading players in the taphocoenosis at the base of the deposit. The presence of many hyaena and bear remains in the pit cave can be explained by the difficulties of climbing back up the slope of the narrow pit cave of Labeko Koba, after throwing animals remains into it in order to eat them. Thus, some of these carnivores would die inside the lower cave. The prey includes horse and deer, species that would be relegated to more secondary positions in the levels in which humans were the principal authors of the fauna contributions. The site sediment originated when the pit cave that fed the debris cone of the lower system became obstructed with silt. We assume that for some time the pit cave was still occasionally activated as a sink hole due to the similarity of the Palaeolithic materials of sub-level upper IX and the Derrubio Superior. However, there is not a single Dufour bladelet among the material (very overabundant in level VII), which leads us to deduce that the pit cave was definitively closed during this period, or in level VIII (sterile from an archaeological point of view). In addition, the sinkhole could have been completely obstructed before the deposition of level IX, later occasionally activating. However, what does seem to be true is that it only drags sedimentary materials included in sublevels lower and upper IX and perhaps VIII too. The early clogging phases of Labeko Koba, which represent almost half of the total archaeological thickness of the site, make up level IX (in 101
CANTABRIAN MOUNTAINS AND COASTLINE.  area outside the protection of a small rock shelter . Radiocarbon dating of the site ...
102 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD some cases it reaches almost two metres thick). This level does not seem to have an anthropic origin and the majority of the material recovered from it consists of fauna remains, accumulated from hyena contributions (that have left part of their bone remains and bite marks in a large part of the series) and other carnivores (Arrizabalaga et al., 2010; Villaluenga et al., 2012). As happened with the archaeological material from the lower gallery network, human presence seems to be limited to occasional visits to procure materials or meat. During these visits, they left some evidence of industry, of opportunistic character (barely knapped and the remains are crude supports or tools probably used as cutting instruments). The arrangement of archaeological remains on the inside and some of the smaller sedimentary differences suggested, from its excavation, differentiating two sections or sub-levels in the core of level IX, separated by a horizon of fragments of a stalagmite slab. Sub-level lower IX has been attributed to a hunting base from the Châtelperronian, mainly due to the significance of three Châtelperronian points in a limited batch of lithic industry (Ríos-Garaizar 2008; Ríos-Garaizar et al., 2012). There is also a very deteriorated fragment of assegai. It is noteworthy that the scarce lithic industry from this sub-level shows typical Upper Palaeolithic characteristics, such as the high number of blades in the assemblage. Similarly the flint sources will be the same as those detected throughout the levels: Sierra de Urbasa and the Treviño syncline (to the south) and the coastal Flysch (to the north) make up the outcrops detected. These circumstances will be common to the entire series of Labeko Koba. The environmental data on sub-level lower IX comes from several pollen and sediment samples and the ecology of the animal species taken to the site by carnivores. A certain convergence is observed in these studies, which indicate dating in the Würm-Les Cottés interstadial for this wet and relatively mild stage with the start of the stratigraphic sequence. The sedimentology study detects a high level of humidity, at the same time as an environment in which it is cold but not very cold. The pollen analyses, carried out on samples from outside the central column, showed the appearance of mesothermophile taxon (such as Castanea). The macro-mammals include a dominant presence (65%) of deer and a lower frequency of hyena than in the upper debris, in a spectrum that indicated a more moderate climate than the upper part of the level. Apart from a few reindeer remains, there are no indicators of a cold climate. As occurs in the rest of the stratigraphic sequence, the micro-mammals return a mild reading for the level and the avifauna does not provide any significant information in this respect. The upper part of level IX occupies the majority of the unit and has almost no industry remains (just five flint flakes and another five possible hammer stones on Irish elk antler). From a cultural point of view, it is difficult to label this sub-level, which is sandwiched between the Châtelperronian (lower IX) and the Proto-Aurignacian (VII), separating them more clearly than the irregular level VIII. However, its environmental characteristics allow us to certify that we have entered the Pleniglacial– Würm III. The sedimentology, palynology and archaeozoology studies indicate intense cold. The sedimentology study also detects a humid environment, with little energy, which decreases at the top of the level. The steppe vegetation and the recrudescence of the climate are noticeable in the pollen record with the domination of Poaceae throughout the level and by the constant presence of Ephedra. Among the ungulates deer dominates and mammoth and woolly rhinoceros emerge strongly in the stratified sequence. Level VIII, which is located in a dispersed fashion in different gentils throughout the cave, is characterised precisely for being archaeologically sterile. Level VII frequently rests directly on the top of level IX, without continuity. Due to this discontinuity, it has not been identified in the samples of the columns and neither do we have any paleoenvironmental information about it. Level VII indicates a clear inflection compared to the underlying level. For the first time in the Labeko Koba sequence, the human being takes on a key role, rivalling the carnivores (mainly bears in this level), to occupy the small space available. Although we cannot identify its rhythm from the record we have, it seems to have occurred in an alternation between human occupations with certain stability over time and the use of the cave by bears as a den. In fact, the lithic industry seems to indicate that this level is the only level in Labeko Koba that presents a fairly complete and closed assemblage, which includes all of the segments of the lithic operating processes. In particular, given the composition of the lithic industry (Fig. 2), this level clearly adheres to the Proto-Aurignacian, more specifically from the variety of the abundant retouch and semi-abrupt bladelets (often Dufour type). In general, the flint sources
102  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  some...
CANTABRIAN MOUNTAINS AND COASTLINE. Figure 2. Sample of the lithic industry. are the same (as in the rest of the sequence) but an advance in provisioning from the outcrops to the south is noticeable in the levels of stable occupation, compared to a higher proportion of northern Flysch flint, common in more sporadic presences. However, the proximity of the outcrops and the Figure 3. Stone with engraved lines. Level VII. awareness of them do not seem to indicate full access, as the use of the raw material throughout the sequence is such that it allows a serious shortage of lithic resources to be detected. Various remains have been recovered from level VII (a stone with engraved lines (Fig. 3), a small ball of amber and various decorative pieces on bone) which indicate that, at least from the ProtoAurignacian, these groups had a symbolic universe similar to that observed throughout the Upper Palaeolithic. As a result of the irregular presence of the remains in the Cantabrian Upper Palaeolithic, we might be encountering the first evidence of this kind described for the northern Iberian Peninsula. From an environmental point of view, the reading of level VII is not unambiguous in the light of the different analyses. The sedimentology study records significant intensification of cold. In the pollen record, the base of the level shows that the cold observed in sub-level upper IX remains stable or increases, at the same time as the level of humidity falls. However, the top section of level VII shows a significant improvement, presenting a milder and wetter landscape, from which steppe taxon disappear. Finally, the archaeozoology study detects a significant change in the proportion of ungulates, probably related to the generalisation of 103
CANTABRIAN MOUNTAINS AND COASTLINE.  Figure 2. Sample of the lithic industry.  are the same  as in the rest of the sequenc...
104 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD the human contribution. In terms of the environment, although different sections have not been distinguished, dominance of woolly rhinoceros remains have been observed at the bottom of the level, in the same way as fawn and boar are mainly located in the upper area, which could help confirm the hypothesis presented by the palynology study. In Labeko Koba, levels VII and V are separated by level VI, which is very poor; culturally it seems to lean more towards level V than level VII due to its industrial composition. Although the lithic industry is not significant, in this level there is a bone object which has been identified as a split-based assegai, an item that delimits the occupation as an initial presence of Early Aurignacian people. A large number of stones and small blocks characterise level VI; they are often cemented by a reactivation of the cavity that dissolved and precipitated some of the carbonates contained in the stones. It is suggestive to see the result of the last phase of significant collapse of the overhang of the cave in this brecciated mass. Sedimentology and palynology studies certify that humidity remained and increased although lower uniformity is recorded in the characterisation of the dominant temperature during the deposit of this unit. While the sedimentology study shows remission of cold in level VI, the pollen record seems to prove a worsening of the climate, with a fully stadial landscape. The archaeozoology study does not provide much information on the environment, but it does show regression in carnivore contributions (which are now small knapped objects) across the level in correlation with the higher importance of human hunting. In this respect, a large increase in the presence of bison (followed by deer and horse) sets the tone of the rest of the sequence, which stands out for several recurring characteristics. These include specialised hunting of bison, which is demonstrated by removing the carnivore remains from the assemblage, the almost non-existence of rock-dwelling animals among the species hunted or the identification of an area that was probably the preferred hunting ground on the plain of the river Aramaio, at the other end of which the Lezetxiki site is located. The chronology of level V is also accurate, given the existence of an almost complete split-based assegai with a flattened section that could correspond to remains of this bone type. The lithic techno-complexes point in the same direction, as they present modal and group levels within the parameters observed for the Early Aurignacian in other levels in the Cantabrian (Morín cave or Polvorín cave, among others). There is still a certain amount of paleoenvironmental information available for level V, even though pollen data comes from isolated samples. The sedimentology study records again a situation of intense cold, as do the aforementioned pollen samples, which is also relevant to the decrease in the ambient humidity. The use of bones as fuel is identified in both this record and in the upper level (level IV), after being carefully fragmented (Yravedra et al., 2005). It could be concluded that, in an environment with little forest cover, this byproduct would be used as fuel, after recovering the marrow that was easier to make use of. The mammoth appears again in the archaeozoological record, and this also demonstrates the climate was worsening. Regarding the economic variables, this level allows us to observe even greater specialisation in hunting bison, followed by horses, deer and other ungulates. Level IV of Labeko Koba is more difficult to date. We have just one absolute dating. Our main argument for ascribing level IV is its close similarity to level V, in every way (environment, sediment, industry, activity). If we look at the lithic record, we can see that the majority of the tests used in the structural dynamics of the lithic industry tend to associate both levels. The lithic structure that returns the best cultural diagnosis (the modal) shows a significant similarity between the two models, even though the semi-abrupt retouch, significantly represented in level V, almost disappears from level IV. It is also possible to reference this Early Aurignacian level with others (units of Gatzarria or Morín cave). The bone industry does not provide definitive data in terms of cultural chronology, we consider level IV a more advanced phase of the Early Aurignacian. At this point, we must stress that we are not referring to an Evolved Aurignacian or évolué, in its French nomenclature, which has a series of characteristics that clearly distinguish it from the Early Aurignacian. Proximity to the model of level V seems to dominate in the assemblage of level IV. For level IV the sedimentology study detects a progressive remission of cold. We do not have any other significant data on the climate at the time of the deposition of the level, or after, as the pollen record is highly disturbed, and the ungulate remains maintain similar conditions to those in level V, in which bison dominate, followed by horse and deer. The last occupations of the site are detected in level III, with the cave almost clogged up. There are no dating values and almost no materials that
104  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  the ...
CANTABRIAN MOUNTAINS AND COASTLINE. could be dated in this level. Furthermore, the industries located in it are particularly poor. For these reasons the only object that can be used to obtain a possible dating for the level comes from the continuity between levels III and IV, both in terms of stratigraphy and other circumstances (the fauna recovered is very similar, very little coherent lithic industry and the sedimentology Alvaro Arrizabalaga * Lezetxiki (Arrasate, Gipuzkoa) 1. History of research Although the Cave of Lezetxiki had been known since 1927, its excavation only began in 1956 after J.M. de Barandiarán’s return from exile. Between 1956 and 1968, alone or in the company of several collaborators (Fernández Medrano, Boucher and Altuna, among others), Barandiarán excavated a surface area of over 100 square metres, to a depth of over nine metres in some places. The sequence included Middle and Upper Palaeolithic levels, and several fossil human remains, not precisely located, owing to the poor conservation of the sequence and the difficulties in interpreting the deposit. The excavation reports were published annually, and specialised studies of the anthropological remains (Basabe), fauna (Altuna and Chaline) and the sedimentological sequence (Kornprobst and Rat) were made known. In order to clarify certain aspects not solved by these publications, since 1996 a new team has been excavating a small sector of the deposit, under the supervision of A. Arrizabalaga and M.J. Iriarte-Chiapusso. This is in an area of about 25 square metres, next to Barandiarán’s southern section, as well as a four square-metre trench on one side, called Lezetxiki II. 2. Chrono-stratigraphic sequence Although the modern fieldwork has maintained the layout of the grid of the original excavations, * study indicates that tempering that started in level IV continues). Judging by this and considering that a sedimentological hiatus does not seem to occur between the start of the stratification and the clogging of the cavity, we could conclude that in terms of chronology, level III is not far from the Upper Aurignacian to which we ascribe levels V and IV. the levels have been numbered differently (using letters, rather than Roman numerals) to avoid confusions. From the bottom to top, the levels dated before the Eemian are Barandiarán’s Units VII and VIII, and Levels M, N, O, P and R in Arrizabalaga’s sequence. Above these, Barandiarán’s Mousterian levels are Units IIIb, IVa, IVc, Va and VI, which correspond to the new units F, G, I and J (Fig. 1) and L, respectively. Finally, the Aurignacian (IIIa or E), Solutrean (D) and Magdalenian (Ia or B) levels complete the sequence. The intricate topography of the cave and the enormous surface excavated explain the discrepancies between the two series, nearly all of which are concentrated in the basal part of the deposit, where Barandiarán only excavated a small side area, whereas the modern excavations have succeeded in discovering a new and older sequence. For the same reasons, the deposit at Lezetxiki includes intensely leached outer sediments, wellprotected areas inside the cave and intermediate transition or rock-shelter areas. The lateral changes in the conditions of the sedimentary fill and the state of conservation of the archaeological materials are consequently very large. This has equally caused great difficulties for the precise dating of the levels (Falguères et al., 2005/2006), despite the numerous attempts with different methods (such as radiocarbon, U/Th, ESR, AAR). According to geochronological and biostratigraphic crite- Universidad del País Vasco (UPV-EHU). Tomás y Valiente s/n. 01006 Vitoria. alvaro.arrizabalaga@ehu.es 105
CANTABRIAN MOUNTAINS AND COASTLINE.  could be dated in this level. Furthermore, the industries located in it are particula...
106 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD working strategy meant the tunnel was excavated from wall to wall. Figure 1: Mandible fragment of a Barbary macaque, from Level J in Lezetxiki II. Reference: Castaños et al., 2011. ria explained in several publications, the base of Lezetxiki II has been attributed to Isotope Stage 6, while Level R in Lezetxiki, the deepest level currently being excavated probably corresponds to Isotope Stage 7. 3. Interpretation of the deposit and significant aspects of the new fieldwork The most recent fieldwork has obtained interesting information enabling a reappraisal of the deposit at Lezetxiki. Some additional research is required, and it is, therefore, likely that in the coming years a series of results of studies and dates will achieve an even better understanding of the site. Consequently, many of the observations made below should be considered provisional and pending verification in an exhaustive report. 3.1. Stratigraphic conflicts This refers to monitoring the stratigraphic circumstances in the excavation opened on the southern side of Barandiarán’s excavation. A surface area of about 20 m2 has been opened (continuous variations in the surface area are caused by changes with depth), in the bands 18, 20, 22 and 24, by projecting the grid system of the original excavation. A square in what would be Band E has been started, which will leave a North-South section, not obtained in the classic excavation as the Our experience has shown that the stratigraphy of the deposit is quite clear and continuous. It is clear in the sense that the interfaces between the levels, which are nearly always thick or very thick, are quite visible and can be identified during the fieldwork. Alternating levels with and without artefacts have been documented, and these helped Barandiarán to follow uniform criteria for the assignation of levels in most of the excavation. It is continuous from the sedimentary point of view, as the sedimentary characterisation is very homogeneous (thick layers of very compact clay, occasionally with calcareous lumps of different sizes) between the new Levels A and L (as far as Barandiarán’s VI, inclusive); and no erosional contacts are seen, breaking this uniformity, except perhaps between Barandiarán’s Levels IIIa and IIIb. Phenomena of stratigraphic alteration (bio– and cryoturbation) are only seen in a few places. The most common of these alterations is mechanical deformation caused by the plasticity of the clay when saturated in water. There has been speculation about the dip of the original levels at Lezetxiki. Our observations have confirmed that they dip significantly towards the south on the north-south axis, resulting in a difference in depth of 3m over the 15m of this axis in Barandiarán’s excavation. However, in our northsouth section in Band E, this dip has disappeared, and the levels are nearly horizontal. It seems that this phenomenon is connected with the opening in the tunnel along a breach on its eastern side in Band 18 (the site was accessed along this breach during the excavation) on the very limit of the area currently being excavated. This breach may have caused mechanical tension on the clayey levels in Lezetxiki, causing the fall of sediment and archaeological materials towards the ravine on the east of the cave. Beyond this point, the lower parts of the tunnel would not have been affected so much by this tension and maintained an approximately horizontal position. Unfortunately, this hypothesis is difficult to verify as the sedimentary record in the affected parts of the tunnel are missing. However, the presence of particular archaeological objects in different levels in the deposit, such as objects incorrectly attributed to the osseous assemblage, and in reality the result of a post-depositional alteration known as charriage à sec, means that it should be taken seriously.
106  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  work...
CANTABRIAN MOUNTAINS AND COASTLINE. In contrast, precise information is available about the dip of the levels in the east-west axis. In the area we have been able to study, the dip has been seen to be extremely complicated, and it is impossible to describe a simple pattern for it. Some levels are sub-horizontal, whereas other dip slightly, or not so slightly, towards the east or the west. In contrast with what used to be thought, the latter dip in the most frequent. Undoubtedly, the effect of the rock-shelter that originated the rock wall makes the sediments accumulate in larger amounts against it, and then slope off towards the wall in the west, on an incline. It is very likely that this effect occurred in other parts of the tunnel (in the bands with even numbers) where the disappearance of the overhang led to the characteristic sedimentation of a rock-shelter and not of a cave. Similarly, although Lezetxiki is obviously a karst cave, its post-depositional development has been very similar to an open-air site as most of the sediment is located outside the rock overhanging the tunnel or the rock wall. For these reasons, three separate records should be considered when interpreting the fill and the evolution of the deposit: as a cave, as a rock-shelter and as an open-air site. These circumstances have caused us to correct our previous assessment of the industrial attributions of the different levels at Lezetxiki. Apart from the central bands in the site, which were excavated first of all and in which the absence of archaeological materials in the surface layer presumably caused a shift in the numbering of the levels in comparison with the bands to the north and south, we have opted to fully respect the attributions of the archaeological materials to the levels established by the excavator during the fieldwork. Precisely because of the lack of fixed parameters in the dip of the levels, a drawing of the stratigraphic assignation of the materials may have led to an error in the definition of the industrial assemblages. Some remarks should also be made about industrial characterisation in the sequence at Lezetxiki, particularly in the transition sub-levels between the Middle and Upper Palaeolithic. Without wishing to prejudge the chronology that will finally be assigned to sub-levels IVc, IVa and IIIa, the presence of a heterogeneous lithic assemblage, with components of ambivalent chaînes opératoires has always struck the attention. This duality in the appearance of the industry has variously been attributed to the archaism of EUP assemblages in northern Spain, cultural and technological mixture of models, the result of severe post-depositional alteration and careless excavation techniques, etc. Our observations cast doubts on the latter suggestion, as we have verified the main anomalous situations detected by Barandiarán’s excavation and assessed the possibility of post-depositional disturbance. Thus, as occurred in squares distant from Level IVc in Barandiarán’s excavations, where blade industries were found together with Levallois points, we have located very similar tools to those that suggested the putative stratigraphic mixing. Equally, when the large number of raw materials other than flint had been stressed, we have been able to increase the proportion of these even more. This is in stratigraphic series often over a metre thick, and where it is not easy to find a simple explanation. In short, in our opinion, industrial characterisation at Lezetxiki requires cultural and technological explanations and the suspicion of mixing between levels over a metre in depth away from each other, separated by industrially barren layers, should not be raised over and over again (depending on the results when they are dated). In this respect, Levels III and IV at Lezetxiki require a new approach to their study, as clichés about the “typical” characterisation of the Mousterian and Aurignacian contribute very little at this and other sites. 3.2. Geochronology The dates currently available for Lezetxiki are, for several reasons, not determinant although they mark some interesting trends in that they suggest the levels are older than was initially thought (Falguères et al., 2006). One of the main objectives of the modern research at Lezetxiki is to provide the sequence with a geochronological framework allowing the levels to be interpreted in a wider regional context. During the recent fieldwork, nine charcoal and bones samples have been taken for radiocarbon dating. The four samples analysed (conventional and mass accelerator C14) have given unsatisfactory results, as the dates are much younger than would correspond to their archaeological context, and therefore they are thought to be aberrant. The paper cited above gives the principal data currently available. However, there are also some new results, involving the use of the racemisation technique and based on bio-stratigraphic criteria, described at length in the latest publications. 107
CANTABRIAN MOUNTAINS AND COASTLINE.  In contrast, precise information is available about the dip of the levels in the east...
108 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD As described on other occasions and explained below, in the absence of sedimentary or archaeological materials to provide a context for the famous Lezetxiki humerus (Fig. 2), the true context of the human fossils found in the old excavation will hopefully be located in Lezetxiki II. As has been pointed out, there is a certain lack of definition in the position of the three fossil human remains found in the classic excavations at Lezetxiki. Whereas the Neanderthal teeth found between Levels III and IV can easily be positioned with their coordinates among all the points in the archaeological levels, and additionally offer no difficulties in their phyletic attribution, this is not possible in the case of the humerus. We know exactly where it was found, but there still are problems in three different lines of interpretation: 1. The phyletic attribution of the humerus is confusing, as in the scientific literature it has been compared both with Neanderthal remains and pre-Neanderthal human types, such as Homo heidelbergensis (Fig. 2). 2. The humerus was found in a volume of sediment with no lithic or osseous remains. Once that sediment had been excavated, it became impossible to date any elements in that context or obtain the palaeo-environmental information. In addition, as the bone is highly mineralised, it cannot be dated directly, and this would in any case be very debatable from the methodological point of view. 3. Since 1998, a small cave we have called Lezetxiki II has been excavated and this is heading towards Cueva de Leibar. This small cave was explored during the original excavations at Lezetxiki, precisely when the human humerus was found in Cueva de Leibar, but because the position of the sieve had made access more difficult, it was not investigated further. We have always suspected that this cave might represent the actual context of the Lezetxiki humerus. As a consequence of all this, we have slowly removed the pile of sediment sieved by the previous excavation to locate the original floor of the cave and open trench, as in some places the sediment had filled the cave to the roof. Each year we have advanced between half a metre and one metre. The 2012 fieldwork completed the excavation of Squares J15, K15, L15 and M15, in a trench one metre wide, four metres long and 320cm deep. In this, we have identified a sequence consisting of: Chalcolithic, barren, Early Upper Palaeolithic, layer of calcite, barren, possible Mousterian, and levels attributed to the MIS5 and MIS6. In 2004, a detailed survey of this cave fixed the stratigraphic relationship between Lezetxiki II and the central passage in Lezetxiki and Cueva de Leibar. As well as a small passage, now inaccessible, directly connecting the central passage in Lezetxiki with Lezetxiki II, it seems clear that there is a direct connection between Lezetxiki II and Leibar. If we continue the trench another metre (Square I15) towards the west and descend 50cm from the current floor level, we should reach the roof of Leibar Cave, in the approximate area in which the Lezetxiki humerus was found. The connection between the two caves, therefore, seems to be granted, and their stratigraphic link is very likely. It remains to be seen whether, in these circumstances, more fossil human remains have been conserved in Lezetxiki II. In any case, the new stratigraphy in Lezetxiki II, parallel to the oldest occupations in the main cave, is of the greatest interest in providing a context and geochronological background for those occupations. 3.3. Neanderthal symbolic behaviour Figure 2. Human humerus from Level VIII in Cueva de Lezetxiki (J.M. Barandiarán’s excavation). Even when it was not known whether opening a small excavation area would obtain any significant archaeological materials to modify the over-
108  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  As d...
CANTABRIAN MOUNTAINS AND COASTLINE. all understanding of Lezetxiki, some impressive malacological remains have been recovered (Fig. 3) (Arrizabalaga et al., 2011), providing an assessment of the symbolic behaviour of the last Neanderthals who occupied the cave. These are two shell fragments at the base of Level III and a further two in Sub-Level IVc. The former (one possibly from a warm-climate marine mollusc and the other a freshwater bivalve) were found precisely in the same context as the two teeth attributed to Neanderthals in Bands 16 and 18 in the excavation. Although it is very difficult to determine whether or not the shells were modified by humans, it is clear that the shells of inedible molluscs were intentionally selected and taken to the site. Both remains are polished and one of them, which is brightly coloured, was obviously selected for strictly aesthetic reasons. The case of the two marine shells found in Level IVc is even more striking because, first, there is no doubt that Neanderthals brought them to the site (the teeth were found nearly a metre above the relative level of this find) and second, it is easier to determine that they were modified by humans, possibly complementing natural processes. These are two fragments of snail shells, probably marine species. One of them conserves the central column and a section of the outer shell wall, so that it could easily be used as a pendant by threading a cord through the gap. The other appears to be a fragment of the helical column of an even larger shell, through which a cord could be threaded, so it could also have been suspended as a pendant. Although the cord could easily have come loose through the inverse path of an open spiral, this may have been solved by blocking it with a large quartz grain (which does not come from the site) in the canal. In both cases, simple microscopic observation can identify striations over the natural polishing and other abrasions compatible with the use of these shells as pendants, following the procedure described above. The presence of these malacological remains in levels where it can be understood that they were brought by Neanderthals suggests that it is necessary to reflect on the symbolic behaviour of this species. Although Neanderthals have traditionally been attributed extremely simple behavioural patterns, including gathering fossils and minerals that “struck their attention” and taking them to their dwellings, in recent years more sophisticated situations have been identified. It is currently believed that the last Neanderthals, at least occasionally, displayed behaviour and attitudes towards ornamental and symbolic elements similar to those recognised in modern humans in the Early Upper Palaeolithic. The debate appears to be focused on why the Neanderthals displayed this behaviour, either because it was apprehended and copied from the first modern humans in Europe, or because it represents cultural evolution intrinsic to the phyletic development of the Homo genus. Figure 3. Fragment of a red Spondylus shell, from Level III at Lezetxiki. Reference: Arrizabalaga et al., 2011. 109
CANTABRIAN MOUNTAINS AND COASTLINE.  all understanding of Lezetxiki, some impressive malacological remains have been recov...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 110 Manuel R. González Morales*, Lawrence Guy Straus*,** Cueva del Mirón (Ramales de la Victoria, Cantabria) Cueva del Mirón (Fig. 1) is a large cave situated in the second foothill chain of the Cantabrian Cordillera in eastern Cantabria, very near the border with Vizcaya. Located at 43 ° 14’ 42” N and 3 ° 27’ 9 ” W and 260 m above sea level on the west-facing cliff of Monte Pando, about 150 m above the valley floor of the Asón River at its confluence of the Calera and Gándara, the 16 m wide x 20 m-high cave mouth dominates the broad intermontane Valle de Ruesga. It is near the crossroads of major north-south and east-west avenues of communication respectively connecting the Cantabrian coast with the Castilian meseta via the Asón and 920 m Los Tornos Pass and the Basque Country with the central coastal zone of Cantabria (now partly occupied by the Bay of Santander) via the Carranza and Ruesga valleys and 674 m Alisas Pass. El Mirón is surrounded by summits of ≥1000 m in elevation, yet is only 20 km from the present shore at the mouth of the Asón (about 25 km from the pleniglacial shore). Probably more or less continuously utilized by humans for residence and (since Neolithic Figure 1. El Mirón. Cave´s view (photo: Alejandro García Moreno). * ** Instituto Internacional de Investigaciones Prehistoricas de Cantabria, 39005 Santander, España. Department of Anthropology, University of New Mexico, Albuquerque, NM 87131 USA.
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  110  Manu...
CANTABRIAN MOUNTAINS AND COASTLINE. times) for sheltering livestock, Cueva del Mirón was identified as an archeological site by H. Alcalde del Río and L. Sierra at the same time (September 1903) that they discovered Covalanas (directly above Cueva del Mirón) and La Haza (below and ca. 300 m north of it) –the second and third Paleolithic cave art sites to be recognized in Spain. Written off by archeologists as being disturbed by modern human activity, the only known (but unpublished) testing of this site –a trench dug across the middle of the inner cave gallery– was done by workers on the orders of civil engineer and amateur archeologist A. García Lorenzo during his construction of the road up the mountainside to Covalanas in the 1950s. Visited in October 1973 by LGS, the cave’s potential as a major site left an indelible impression. The authors directed large-scale excavations in Cueva del Mirón between 19962013 (Straus and González Morales 2012). Cueva del Mirón (Fig. 2) consists of a large, dry, sunlit vestibule measuring 30 m deep x 7-11 m wide x 12-13 m high; a dark, 7-8 m wide inner gallery that is accessible for about 80 m (after which it is filled to the ceiling with alluvial deposits and travertine); and a narrow (3-4 m-wide) connecting passage that is 20 m long and contains a ramp of colluvial-alluvial sediments-an erosional face of the inner cave’s alluvial infilling. The cave has obviously undergone several cycles of deposition and erosion by running water studied by the late W.R. Farrand (2012). Indeed the inner cave trench (cleaned and deepened in 1996) revealed a sequence of Medieval, Bronze Age and Lower Magdalenian visits/occupations whose sediments fill an ancient channel that cut through the alluvium, remnant terraces of which survive along the edges of the inner cave. Above the top of the ramp in the connecting gallery breccia under a travertine remnant adhering to the cave wall yielded flakes and faunal remains dated to the Azilian period and the sedimentary infilling of a niche in the cave wall above there present erosional surface about two thirds of the way down the ramp produced artifacts and bones dating to the Initial Magdalenian. The vestibule (where geophysical prospection has shown there to be some 9 m of sedimentary deposits down to bedrock) was probably emptied of deposits by running water and progressively refilled –mainly with sediments washed down from the in- ner cave alluvium, but also with others washed and blown in from the exterior via the vast cave mouth, together with material deposited by humans and animals over thousands of years. Excavations were conducted in three areas of the vestibule: the Outer Vestibule (9.5 m²), the Vestibule Rear (maximally 17 m², depending on the level), and the Mid-Vestible Connecting Trench (maximally 7.5 m²). Included within the Vestibule Rear area is a large looters’ pit from which some 25 m³ of mixed sediments were removed and dry-screened. This pit had reached the base of the culturally and organically rich Magdalenian sequence, stopping at the top of the Solutrean and thus allowing excavation of a 2-1 m sondage through Solutrean and Gravettian-age levels down to a horizon at the foot of the buried ramp deposit that dates to 41 uncal. kya –Final Middle Paleolithic. Sampling of the Solutrean levels was later expanded to a total of 4 m² after excavation of overlying Magdalenian layers in two more contiguous squares in Vestibule Rear. Also in the Vestibule Rear, excavation in a minimally connected area of 4 m² between the cave wall and a large engraved block yielded a secondary human burial of Lower Magdalenian age sandwiched between layers of the same period. The site has been dated by 84 AMS and conventional radiocarbon dates (Straus and González Morales 2003, 2007, 2010; Straus et al., n.d.a) ranging from AD 1400 to 41,000 uncal. BP. The culture stratigraphic sequence revealed in the Outer Vestibule includes Lower Magdalenian (16.6-15.0 uncal. kya), possible Middle Magdalenian (ca. 15-14.6 uncal. kya), Upper Magdalenian (13-12 uncal. kya), Azilian (ca. 12-11 uncal. kya), extremely poor Mesolithic with hiati (9.5-8.4 uncal. kya), Neolithic (5.7-4.7 uncal. kya), Chalcolithic (4.1-3.8 uncal. kya) and early Bronze Age (3.7 uncal. kya) levels. The Mid-Vestible Trench cuts through possible Solutrean or Initial Magdalenian (17.4 uncal. kya), Lower Magdalenian (15.9 uncal. kya), Upper Magdalenian (12.5 uncal. kya), Azilian (11.6-10.3 uncal. kya), Neolithic (5.8-4.9 uncal. kya) and possible disturbed Chalcolithic and/or Bronze Age deposits. The Vestibule Rear includes traces of Terminal Mousterian (41.3 uncal. kya), Gravettian (27.6 uncal. kya), Solutrean (19.2-18.4 uncal. kya), Initial Magdalenian (17.6-17.0 uncal. kya), Lower Magdale- 111
CANTABRIAN MOUNTAINS AND COASTLINE.  times  for sheltering livestock, Cueva del Mir  n was identified as an archeological ...
112 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 2. Stratigraphic profile of Cueva del Mirón nian (ca. 17-14.9 uncal. kya), possible Middle Magdalenian (ca. 14.1-13.4 uncal. kya), possible Upper Magdalenian (12.5-12 uncal. kya), possible poor Azilian and Mesolithic. It is likely that sloping Neolithic, Chalcolithic and Bronze Age deposits had originally existed in the Vestibule Rear, but had been removed by shepherds to level the area for use as a corral (still in existence in 1996). The fumiers that characterize much of the Holocene sequence have yielded an importance series of paleomagnetic results (Carrancho et al., 2013).The complete sequence of nearly 40,000 years has provided detailed micromammalian evidence of Late Pleistocene and Holocene fluctuations in climate and vegetation (Cuenca-Bescós et al., 2008, 2009), as well as a less continuous palynological record (still under study by M.J. Iriarte). While the Mousterian and Gravettian (González Morales and Straus 2013a) levels only suggest minor human visits at least to the Vestibule Rear area, the Solutrean levels seem to indicate repeated, archeologically but somewhat more visible visits, probably by parties that moved up into the mountains in summer from base camps in the coastal zone and that were heavily involved in ibex and red deer hunting, as suggested by the relatively large numbers (and relative frequencies) of foliate and shoulded points made on diverse kinds of lithics, accompanied rather enigmatically by large numbers of perforated shell, tooth, bone and stone beads (Straus and Gonzalez Morales 2009; Straus et al., 2011a, 2013). Cueva del Mirón has one of the most important sequence of Intial Magdalenian levels in the Cantabrian region, characterized by the eventual disappearance of Solutrean points and the prsence of large, thick, round-section antler points and both large flakes and “archaic” tool types on local non-flint raw materials and bladelets (unretouched, retouched and backed) on excellent quality non-local flints, but without the diagnostic tools of the somewhat older
112  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
CANTABRIAN MOUNTAINS AND COASTLINE. French Badegoulian. It is with this period that the cave began to witness massive, long-term, multi-purpose human occupations, as attested by the masses of faunal remains, lithic debris and tools, hearths, fire-cracked rocks and dark brown-black, charcoal-rich sediments (Straus et al., n.d.b). A broken slate pendant with the engraving of a horse head comes from this period (González Morales and Straus 2013b). Such intensive occupations continued throughout the classic Lower Magdalenian, with thick palimpsests of living floors very similar to those of El Juyo, Altamira, Santimamiñe and El Castillo on or at the edge of the coastal zone. Remains of red deer, ibex (under study by J-M. Geiling) and salmon are extraordinarily abundant, as are stone and osseous artifacts, including large numbers of bladelets, nucleiform scrapers (on bladelet cores) (Straus et al., 2008), sagaies of many types including iconic quadrangularsection ones with geometric or “tectiform” engraved decorations, eyed needles, and an antler spearthrower remarkably similar in form and dimensions to ones from sites in SW France (González Morales and Straus 2005, 2009). There are also “macrolithic” tools made on local nudstone, quartzite and limestone, but the flints (used especially in bladelet manufacture) mostly come from Upper Cretaceous outcrops in the coastal zone of western Vizcaya and eastern Cantabria (at distances of at least 40-50 km from the site) (Rissetto 2009). The Lower Magdalenian levels contain many, often repeatedly re-used hearths filled and surrounded by fire-cracked rocks (some previously used as anvils) and masses of heavily fragmented long bones, suggestive of processing for grease rendering via stone boiling (Nakazawa et al., 2009). Notably these levels have yielded engraved fragments and one whole red deer stag scapula with the engraved and striated image of a hind head and the outline of a bovine head. The style, composition, technique and hind subject are virtually identical to engraved scapulae from several Lower Magdalenian sites in central Cantabria and eastern Asturias (notably Altamira, El Castillo and El Juyo), as well as closely resembling images on the walls of the former two and other caves, thereby defining a regional cultural marker (González Morales and Straus 2009). The DNA studies of salmon and red deer from this and the other Paleolithic horizons of Cueva del Mirón show that this region served as a refugium for these species (as it did for humans) during the Last Glacial Maximum and hence was a source area for the repopulation of more northerly regions of Europe beginning during Oldest Dryas –the time of the Initial and Lower Magdalenian (Consuegra et al., 2002; Meiri et al., 2013). It is to the Lower Magdalenian that belongs the secondary human burial and the fall and at least intial engraving of the 2x1x1 m limestone block behind which the mandible and some 100 other bones of the human were deposited after they had naturally lost their flesh and then been stained with red ochre (Straus et al., 2011b). The sediments used to cover the bones were also stained with red ochre and glitter with hematite cystals (R. Seva, personal communication). The engraved block adjacent to the bones is also stained red and the walls of the vestibule rear are covered with masses of engravings, including the image of a horse, all possibly of Lower Magdalenian based in part on their height above the ground surface at the time (García Díez et al., 2012) It is known that the block had fallen atop a layer (110) dated to 16 uncal. kya (a few centuries before the burial was done behind it) and was engraved on its sheered off flat surface during subsequent years before being covered over by dated Middle and Upper Magdalenian, Azilian and Mesolithic-age levels. The ungulate faunal assemblages of the Middle and Upper Magdalenian and Azilian levels are dominated by red deer and ibex, all hunted during the warm season, suggesting a mobility pattern that included winter residential bases in the coastal zone (Marín 2009, 2010). Unlike the nearby cave sites of El Valle and El Horno, located on the valley floor, the Upper Magdalenian and Azilian occupations of El Mirón seem to have been of low intensity and frequency, with relatively few artifacts (though these do include an antler harpoon and an ochre-stained pebble respectively) (Gonzáez Morales and Straus 2012). The cave was only fleetingly visited during Mesolithic times when settlement was concentrated at shell middens around the newly formed Asón estuary. Quite abruptly, locally well-made, undecorated ceramics (studied by C.Vega [2012]), domesticated sheep/goat, cattle and pig (studied by J.Altuna and K.Mariezkurrena [2012]) and 113
CANTABRIAN MOUNTAINS AND COASTLINE.  French Badegoulian. It is with this period that the cave began to witness massive, lo...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 114 wheat (studied by L. Peña-Chocarro [2012]) appeared in El Mirón as it was reoccupied intensively ca. 4650 cal. BC, making it one of the earliest known Neolithic sites in northern Atlantic Spain. Unlike most cave sites dating to the Chalcolithic and Bronze Age in Cantabrian Spain which are funerary loci, Cueva del Mirón was a major residential place for both humans Juan Carlos López Quintana* Amagoia Guenaga Lizasu* Cueva de Santimamiñe (Kortezubi, Bizcaia). Stratigraphy and human occupations 1. Cueva de Santimamiñe (Kortezubi, Biscay). Geography and first excavations in its archaeological deposit Cueva de Santimamiñe is in the Oka River basin, a coastal valley on the eastern side of the northern Spanish coast, in the Province of Biscay. Its location, on the southern slopes of Mt. Ereñozar (446.5m), is in a strategic position over the valley. The cave entrance, at an altitude of 137m, faces south/southeast, and leads into a large entrance chamber lit by daylight. The entrance and the chamber contain a stratigraphic deposit with an environmental sequence ranging from late isotope stage MIS 3 to about the middle of MIS 1. Santimamiñe also contains a Palaeolithic art ensemble, discovered in 1916 and later studied by H. Breuil (1917), who found further engravings and defined the main graphic units in the decorated chamber. The archaeological deposit has been excavated in two stages: first, from 1918 to 1926, by T. Aranzadi, J.M. de Barandiarán and E. Eguren (Barandiarán, 1976: 11-344); and second, from 1960 to 1962, by J.M. de Barandiarán (Barandiarán, 1976: 345-419). The excavation in the cave entrance reached a depth of over 8m, with a stratigraphic sequence covering the period from the Aurignacian to the Roman age and later (Barandiarán, 1976: 421-475). * and livestock (increasingly cattle), with numerous large pits (some with fire-cracked rocks), masses of ash, ceramics, a pair of arrowheads (in the Chalcolithic) and a copper pin, plus possible slag (in the Bronze Age). There are hints of later visits to the site and indeed it was inhabited by people as recently as the early post-Civil War period. 2. The 2004-2006 study of the archaeological deposit: stratigraphic sequence. After 42 years without studying the deposit at Santimamiñe, in 2004 a new series of archaeological research began (2004-2014), with the aim of reexamining and updating the stratigraphic record at this Basque site. The research was framed within a full study and development programme initiated by the Culture Department in the Foral Deputation of Biscay. This programme includes the re-exploration, mapping and environmental monitoring of the cave, as well as a reappraisal of the archaeological deposit and the Palaeolithic art ensemble. The archaeological work has involved 25 months of fieldwork since 2004, supervised by J.C. López Quintana and A. Guenega Lizasu. The first three years (2004-2006) were spent revising the stratigraphy of the deposit, and a first monographpresentation of the site has already been published (López Quintana and Guenaga, 2011); from 2007 to 2014, the human occupations in the Holocene and Late Pleistocene have been the subject of a larger-scale excavation. This reappraisal of the stratigraphy at Santimamiñe has studied a sequence 6m thick, which has been analysed and interpreted in accordance with the principles of Analytical Stratigraphy Asociación de Arqueología AGIRI / Círculo de Estratigrafía Analítica. Apartado de Correos nº 208, 48300 Gernika-Lumo (Bizkaia). E-mail: arkeoagiri@hotmail.com
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  114  whea...
CANTABRIAN MOUNTAINS AND COASTLINE. Figure 1. Frontal stratigraphic section of the deposit in Santimamiñe and an analytical matrix of the sequence. (Laplace, 1971; Sáenz de Buruaga, 1996). This sequence consists of 27 stratigraphic units, grouped into four series according to geoclimatic and paleoethnological criteria (López Quintana and Guenaga, 2011). These four stratigraphic series will be described from the base to the top (Fig. 1). – Lower palaeontological series (MIS 3/MIS 2) The lowest part of the stratigraphic series at Santimamiñe is 1.05m thick and consists of four stratigraphic units: Lsm-Sa, Arb-o, Arp and Arg-o. The sediment is mainly fine material (clays and silts), in proportions always reaching above 95%, attesting decantation processes in a wet environment (Areso and Uriz, 2011). However, the Arg-o unit has yielded some indicators of a cold environment, suggesting open vegetation with the presence of reindeer. The absence of evidence of human occupation is one of the traits of this series, which contains two levels with palaeontological remains (Arb-o and Arg-o). The faunal repertoire 115
CANTABRIAN MOUNTAINS AND COASTLINE.  Figure 1. Frontal stratigraphic section of the deposit in Santimami  e and an analyti...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 116 of the former is greatly altered while the accumulation in the second evidences the site was used as a shelter by carnivores. Radiocarbon determinations1 (AMS) situate the series between the late isotope stage MIS 3 (Arb-o Unit: 26,890 ± 180 BP, Beta-259132) and the early isotope stage MIS 2 (Arg-o Unit: 20,530 ± 110 BP, Beta-240906). – Flooding Complex (MIS 2) The Flooding Complex is a layer of sediment 2.3 – 2.6m thick, formed by four stratigraphic units: Avp-Sj, Lrg, Arp-Sa and Lsr-Ap. They are all predominated by the fine fractions (silts and clays) and coarse components are absent. The complex is barren from archaeological and palaeontological points of view, and it formed during a mild and wet oscillation which can be dated in the GI-2 interstadial in the NGRIP climate sequence (Rasmussen et al., 2008), between ca. 20,000 and 18,800 BP. – Cryoclastic Middle-Upper series (MIS 2) The Middle-Upper series, 1.8m thick, consists of 12 units with certain stratigraphic complexity: 8 levels with archaeological content, 1 barren level, 2 layers of calcite (T4 and T5) and 1 erosion phase (V-Almp). The sedimentary trait defining this series is the evidence of cryoclastic processes, mainly in Units Csn-Camr, Slnc and Arcp. These three levels have provided indicators of a cold climate of differing intensity, which have been associated with the cold phases GS-2, GI-1d and GS-1 in the NGRIP sequence (Rasmussen et al., 2008). This series includes the Balm and Almp levels, which almost certainly correspond to the NGRIP phase GI-1e. In M. Hoyos’s regional late glacial climate sequence, this would be equivalent to his Cantabrian Phase VI (Hoyos, 1995). – Holocene Upper series (MIS 1) The first stages of the Holocene are represented by the erosional hiatus V-Arcp and slight evidence of human activity (hearth facies H-Sln) included in the calcite layers T2 and T3 dated to 7580 ± 50 BP (Beta-240899). Above this, the stratigraphic unit Slm-Lsm is 0.5m thick, with a preponderance of the fine fraction. The lower part of this group (Slm level) contains evidence of a Neolithic occupation, between ca. 5500 and 5000 BP. After this time, human presence at Santimamiñe becomes increasingly occasional, although the cave was used for burials in the Chalcolithic-Bronze Age (Lsm level). 3. The sequence of occupations at Cueva de Santimamiñe: landscape and environment exploitation strategies from the lower Magdalenian to the Chalcolithic-Bronze Age. The stratigraphic sequence at Santimamiñe hosts seven phases of human occupation, above the flooding complex and corresponding to the climate events GS-2, GI-1, GS-1 and MIS 1. 3.1. Red deer hunters in the late lower Magdalenian The first human occupation at Santimamiñe (Csn-Camr Unit) took place during a cold phase, assignable to the late GS-2 stadial in the NGRIP sequence (Rasmussen et al., 2008) or the early Cantabrian V (Hoyos, 1995). From its base to the top, intensification in human activity in the cave can be seen, parallel to the deterioration in environmental conditions. This took place during a short period of time, judging by the two available C14-AMS determinations: 14,670 ± 80 BP (Beta240904) and 14,650 ± 80 BP (Beta-240905) This series contains the first human occupations in the cave, coinciding with the three phases of climate deterioration. The first occupation at Santimamiñe (Csn-Camr) took place in the late lower Magdalenian, in around 14,700 BP, according to data from the new excavation. Therefore the reappraisal of the stratigraphy at Santimamiñe has not found evidence of Early Upper Palaeolithic occupations (López Quintana and Guenaga, 2011: 56). The Csn-Camr archaeozoological record is polarised towards the hunting of red deer, which make up 91.4% of the total ungulates captured (Castaños and Castaños, 2011). The percentages of chamois and ibex are surprisingly low (4.6% and 3.2%, respectively) in this landscape of crags and steep hillsides. The scanty evidence of fishing or the gathering of molluscs as food supports this idea of an extraordinarily specialised subsistence strategy. All the C14 dates in the text are given in uncalibrated years BP (conventional C14 age). The lithic assemblage displays the highest percentage of retouched tools in the sequence, and equally the lowest proportion of debitage prod- 1
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  116  of t...
CANTABRIAN MOUNTAINS AND COASTLINE. ucts, which means that lithic reduction tasks are scarcely represented at the site. The lithic raw materials include 23.2% of exotic flint in the total (Tarriño, 2011) with sources located between 70 and 180km away (Urbasa, Treviño, Tercis and Chalosse). This is indicative of the great territorial mobility of lower Magdalenian red deer hunters. The taphonomic study of the large mammals proposes an occupation specialised in processing the game, essentially red deer, which were taken to the cave whole to be skinned, butchered, defleshed and finally broken up to obtain the bone marrow (San Pedro and Cáceres, 2011). Antler and boneworking were other important activities at the site, which has yielded a large collection of bi-pointed assegai points with triangular-trapezoidal crosssections. 3.2. Evidence of human occupation during the formation of Almp stratigraphic unit? After the time of the lower Magdalenian red deer hunters, a layer of calcite (Balm) was dismantled, possibly by erosion related to an increase in humidity. The back of the entrance chamber again suffered ponding and displays no sign of human presence, until the humidity decreased relatively (in the middle and upper part of Almp) allowing occasional human occupation events. This level reflects cool wet conditions, probably in the interstadial phase GI-1e of the NGRIP sequence or, on a regional scale, in Hoyos’s Cantabrian VI phase (ca. 13,300-12,700 BP). In this section of the sequence, some stratigraphic alterations have been detected, as material was disturbed by the overlying Slnc level. This means that the archaeological record obtained should be assessed with caution (López Quintana and Guenaga, 2011: 33-36). If the faunal assemblage from Almp is examined as a whole, a significant change is seen from the underlying Csn-Camr level. The Almp unit has provided a diversified assemblage, in which ibex (Capra pyrenaica) is more abundant than red deer, in proportions of 50.3% and 40.4%. The results of the ichthyological study are in agreement with this new strategy of diversification in the consumption of resources. Almp reveals an increase in fishing (mainly of salmonids), as the total of the ichthyological series in the sequence goes from 2.37% in the Csn-Camr Unit to 28.4% in Almp (Roselló and Morales, 2011). 3.3. The Upper-Final Magdalenian: diversification and full use of natural resources The following episode in the stratigraphic sequence at Santimamiñe is the most intense human occupation at the site, coinciding with the coldest and driest phase in the deposit (Slnc). Climatically, it can be included in the GI-1 interstadial of the NGRIP sequence, possibly in the GI-1d cold oscillation (Rasmussen et al., 2008). In Hoyos’s late glacial sequence (Hoyos, 1995), it clearly corresponds to the Cantabrian VII, dated between ca. 12,700 and 11,700 BP. The pollen record reveals the worst conditions in the stratigraphy, with 2% tree pollen (pine, birch and juniper), while Compositae dominate in the herbaceous-shrub layer, together with the appearance of Artemisia (Iriarte, 2011). A C14-AMS determination is available for the lower part of Slnc: 12,790 ± 70 BP (Beta240902). Another C14-AMS determination obtained in the underlying Almp level, (12,250 ± 70 BP: Beta-240903), and regarded as intrusive, may have come from the disturbance of the upper part of Slnc (H1-Slnc), and closely matches the coldest and driest phase of the Cantabrian VII stage. The archaeozoological assemblage in Slnc is diversified in the hunting of ungulates. Red deer (49.1%) is more common than ibex (30.7%), while some rarer species in earlier periods, such as roe deer and large bovids, become more common. Fish (almost exclusively salmonids) are indicative of the model of a diversified resource use, as this level contains the largest number of remains, with 58.58% of the whole ichthyological assemblage at the site (Roselló and Morales, 2011). As regards food processing, a sandstone slab may have been used as a refractory surface in the context of a hearth (Delgado-Raack, 2011). The most common plant species used as firewood is juniper, as well as a large number of indeterminable conifers (Euba, 2011). The Slnc archaeological record reflects a subsistence model aimed towards a full use of natural resources. This diversification in subsistence strategies is associated with a relative reduction in the size of the territory being exploited and an extension in the cycles of cave occupation. Within the River Oka basin, the different ecosystems were used more exhaustively, from the valley bottom and steep crags, to the rivers and estuary, and even the shore, although less intensively because of its distance from the site. In accordance with human occupations of greater intensity and in longer cy- 117
CANTABRIAN MOUNTAINS AND COASTLINE.  ucts, which means that lithic reduction tasks are scarcely represented at the site. T...
118 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 2. Selection of backed points and bladelets from the Slnc level (Upper-Final Magdalenian)
118  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
CANTABRIAN MOUNTAINS AND COASTLINE. cles during this phase, the reappraisal of the Palaeolithic rock art by C. González Sainz and R. Ruiz Idarraga is of interest. In their study, the art at Santimamiñe is interpreted as a synchronic ensemble which was probably produced during the time of the occupations in the Slnc level. They reject the theory that the ensemble is an accumulation of figures produced between the middle Magdalenian and the Azilian (González Sainz and Ruiz Idarraga, 2010: 150-151). are prismatic, which were used to obtain bladelet blanks. The bone assemblage consists of harpoons (with a magnificent example of a Cantabrian harpoon with a single row of barbs and a pierced base) (Fig. 3), as well as assegai points, spatulas, rods, awls, fine points or stilettos, pins and needles (González Sainz, 2011). The Slnc level has also yielded the largest lithic assemblage in the deposit. Indeed, 62.88% of the total lithic assemblage recovered during the 20042006 fieldwork comes from this level, where it reflects a specialisation in backed bladelets (Fig. 2), which make up 38.36% of the retouched elements in the level (López Quintana et al., 2011). In coherence with this, the most common cores The late glacial sequence at Santimamiñe concludes with the Arcp stratigraphic unit, assignable to the final part of the NGRIP GS-1 stadial (Rasmussen et al., 2008) or Cantabrian IX phase in the regional sequence (Hoyos, 1995). Two C14-AMS determinations are available for the bottom and top of the level 10,060 ± 60 BP (Beta-240901) and 10,100 ± 60 BP (Beta-240900), respectively. The pollen re- 3.4. The Azilian level: the last Ice Age hunter-gatherers Figure 3. Selection of harpoons from Santimamiñe: 1924 (3 and 6), 1961 (5), 2005 (2 and 4), 2007 (7) and 2008 (1) field seasons 119
CANTABRIAN MOUNTAINS AND COASTLINE.  cles during this phase, the reappraisal of the Palaeolithic rock art by C. Gonz  lez ...
120 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD cord shows an increase in tree cover, with hazel and oak appearing in the sequence for the first time (Iriarte, 2011). The climate has been defined as cold and wet, with the conditions improving in its most recent phase (Murelaga et al., 2011), perhaps announcing the onset of isotope stage MIS 1. one of the main changes in the new climate conditions, the proliferation of open-air sites, whose clearest example is the site of Pareko Landa, but also at Goienzabal 1, Sollube Txikerra 1, Katillotxu, Garbola and Landabaso (López Quintana and Guenaga, 2009: 110 and 117). The archaeological data suggest a more occasional occupation that in the previous levels (López Quintana and Guenaga, 2011: 439). Red deer is once more the most hunted ungulate (63.7%), while the hunting of ibex (21.9%), roe deer and large bovids (both 6.1%) decreases. The fishing of salmonids also declines significantly, with a change from the 58.58% of the total ichthyological assemblage in the underlying Slnc level, to 7.99% in this unit. Marine molluscs are not found in this level either; the number of sea urchin remains increases but they do not represent an important food resource since the 437 remains belong to a minimum number of only two individuals (Gutiérrez Zugasti, 2011). The stratigraphic reappraisal from 2004 to 2006 found evidence of the use of malacological resources in the estuary, with a predominance of taxa gathered in rocky and muddy zones (Ostrea edulis, 37%, and Scrobicularia plana, 32%) (Gutiérrez Zugasti, 2011). The diversity of usable resources in the estuary increased significantly in the Holocene. The industry in the Arcp level has been defined as a “critical transition episode” within the Santimamiñe deposit (López Quintana et al., 2011: 103). The components of the osseous assemblage disappear. In turn, the lithic assemblage is diverse, characterised by a significant number of denticulates and a decline in the backed bladelets. The percentage of exotic flint is significantly smaller than in the lower Magdalenian (Csn-Camr Unit), as it decreases from 23.2% to 7.8%. The Urbasa (70km away) and Chalosse (180km in a straight line) varieties are found but, in comparison with the lower Magdalenian, the Treviño and Tercis types disappear (Tarriño, 2011). 3.5. The early Holocene: the use of estuary resources The area excavated in 2004-2006 yielded some Mesolithic remains (malacology, fauna and lithics), in the H-Sln Unit and dated to 7580 ± 50 BP (Beta-240899). In contrast, the old excavations documented a thick shell-midden layer (nearly 1m thick), whose lower section (Level IV) without pottery, was attributed to the Mesolithic and yielded a rich collection of archaeological material (Barandiarán, 1976: 429-431). With the Holocene (MIS 1), a changed is perceived at Santimamiñe in the way the cave space was used, with the human habitat now mostly in the entrance and outer area, which was excavated in last century’s fieldwork (Barandiarán, 1976). This is in accordance with 3.6. The middle Holocene: Neolithic farming groups. The climatic conditions for the Neolithic farmers were the wettest in the sequence at Santimamiñe. The formation of the Slm level has been dated by C14-AMS to 5450 ± 50 BP (Beta-240898) and 5010 ± 40 BP (Beta-240897), in the late Neolithic. The pollen study shows the significant spread of tree cover (53% in the middle part of Slm), with a predominance of birch (37%) accompanied by hazel and oak (Iriarte, 2011). The archaeozoological series is quite poor, although domestic species (cattle, sheep/goats and swine) predominating over wild ungulates, mainly red deer, although ibex, chamois and roe deer have also been recorded (Castaños and Castaños, 2011). It is noteworthy that, like the lithic assemblage and the pottery, two-thirds of the ungulate assemblage is found in the upper Lsm level, which indicates the occupation in the entrance chamber in the early Neolithic was very tenuous. At the same time, this level attests the most intense exploitation of marine molluscs (Gutiérrez Zugasti, 2011). Gathering was focused on muddy zones in the estuary, and rocky areas to a lesser extent. The peppery furrow shell Scrobicularia plana (86% MNI) is the most common species, followed by the grooved carpet shell Ruditapes decussatus (5.5%) and oysters (3%). In contrast, evidence of fishing declines drastically (the level provides 2.66% of the ichthyological assemblage at the site), with some changes in taxonomical representation, particularly a decrease in salmonids (Roselló and Morales, 2011). Evidence of charred wood increases remarkably, with a clear predominance of oak, followed by Rosaceae (Prunus sp.) (Euba, 2011). In the lower Slm level, the two main species (oak and
120  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  cord...
CANTABRIAN MOUNTAINS AND COASTLINE. hazel) are accompanied by testimonial evidence of strawberry tree, birch, heath, ash, Pomoideae, willow/poplar, buckthorn and pine, as well as an exceptional fragment of wild olive tree (Olea europaea). The archaeological content of the Slm level is limited, but coherent with a Neolithic assemblage. The best represented lithic typological groups are the denticulates and endscrapers, and in second place, backed bladelets (López Quintana et al., 2011). However, qualitatively, double-bevelled retouched segments, including a triangular bi-truncated piece in the style of a Sonchamp-point, are of particular interest. Some 95.9% of the flint is the local Cretaceous Flysch variety, available within 20km of Santimamiñe. Exotic flint is limited to 4.1%, and represented by solely the Urbasa variety, as those from further afield (Chalosse and Tercis) disappear from the record. This suggests a gradual reduction in the areas covered by the first farming groups. timamiñe. In the neighbouring site of Kobaederra, barley and emmer wheat have been identified in the Neolithic levels, with a date of 5375 ± 90 BP (AA-29110) obtained for a barley grain from Level IV (Zapata, 2005: 557). 3.7. Funerary use in the Chalcolithic-Bronze Age After the Neolithic, human occupation of Santimamiñe becomes increasingly occasional, although in the middle and upper part of the Lsm Unit, dispersed human remains indicate a possibly very sporadic funerary use of the cave (Fig. 4). The anthropological study of the human remains (Herrasti and Etxeberria, 2011), which also included finds made by the old excavations, determined an MNI of six individuals. One of these, discovered in the 2004 field season, has been dated to 3710 ± 40 BP (Chalcolithic-Bronze Age) while its genetic study showed that it was a bearer of the T2b mitochondrial line (Cardoso et al., 2011). One novel point is the identification of six shell fragments with signs of being used as tools (four of Ostrea edulis, and one each of Mytilus galloprovincialis, Ruditapes decussatus and Patella sp.), the first objects of this kind to be published in Cantabrian Spain (Gutiérrez Zugasti et al., 2011). In general, they were used in transversal scraping actions, in some cases interspersed with longitudinal cutting tasks. Except for the Ruditapes decussatus fragment, possibly used to process plant matter, the other shells were used to work a soft or medium-hard animal substance. In short, in the middle Holocene, in mild and wet climatic conditions, the cave of Santimamiñe was used, perhaps occasionally and repeatedly, by farming communities who additionally exploit the environment by hunting ungulates and gathering molluscs in the estuary. For comparison, 1.5km from Santimamiñe, in its Levels III and IV, dated between ca. 5800 and 5600 BP, the cave of Kobaederra displays a predominance of domestic fauna, which makes up 70.2% of the archaeozoological assemblage (Altuna and Mariezkurrena, 2009: 762). To date, no cultivated cereals have been found in the stratigraphic reappraisal at San- Figure 4. Copper chisel with an antler handle, found in Level II (Chalcolithic-Bronze Age) in 1924. 121
CANTABRIAN MOUNTAINS AND COASTLINE.  hazel  are accompanied by testimonial evidence of strawberry tree, birch, heath, ash,...
122 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Marco de la Rasilla*, Antonio Rosas**, Juan Carlos Cañaveras***, Carles Lalueza-Fox****, David Santamaría*, Sergio SánchezMoral*****, Almudena Estalrrich**, Antonio García Tabernero**, Pablo G. Silva******, Enrique Martínez*, Gabriel Santos******, Lucía Martínez*, Elsa Duarte*, Rosa Huguet*******, Markus Bastir**, Javier Fernández de la Vega*, Paloma Suárez*, Ana Belén Díez*, Beatriz Fernández Cascón**, Soledad Cuezva*****, Ángel Fernández Cortés*****, Elena García Antón*****, Concepción Muñoz***, Javier Lario********, Pedro Carrasco******,Pedro Huerta******, Puy Ayarza******, Fernando Álvarez Lobato******, Loreto Rodríguez******, Inmaculada Picón******, Begoña Fernández******, M. Standing******, Carmen Sesé*********, Trinidad de Torres**********, José Eugenio Ortiz**********, Helène Valladas***********, Norbert Mercier***********, Nadine Tisnèrat-Laborde***********, Rainer Grün************, Stephen Eggins************, Thomas Higham*************, Rachel E. Wood************, Ramón Julià**************, Vicente Soler***************, Ernestina Badal****************, Antonio Tarriño*****************, Domingo Carlos Salazar******************, Jesús Alonso*******************.# El Sidrón (Piloña, Asturias) In memory of Javier Fortea Pérez (1946-2009), Manuel Hoyos Gómez (1944-1999) and Olvido Otero González (1908-1938) The well-known eponym El Sidrón has a very special history (Fig. 1). It started with the development of a karstic system between two types of rock (sandstone and Neogene conglomerates) as a result of the flow of a small stream. It continued with the use of the cave as a refuge and a hiding place during the Spanish Civil War and the aftermath and with the presence of some endemic species of bats and cave insects. It ended up as the * Universidad de Oviedo. mrasilla@uniovi.es ** Grupo de Paleoantropología. Dpto. de Paleobiología. MNCN. CSIC. Madrid. arosas@mncn.csic.es *** Universidad de Alicante. jc.canaveras@ua.es **** Inst.de Biología Evolutiva. CSIC/UPF. Barcelona. carles.lalueza@upf.edu ***** Dpto. de Geología. MNCN. CSIC. Madrid. ssmilk@mncn.csic.es ****** Universidad de Salamanca. pgsilva@usal.es ******* Universidad Rovira i Virgili. Tarragona. rhuguet@iphes.cat ******** UNED. Madrid. javier.lario@ccia.uned.es ********* Dpto. de Paleobiología. MNCN. CSIC. Madrid. c.sese@mncn.csic.es ********** LEB. ETSIM. Universidad Politécnica de Madrid. trinidad.torres@upm.es *********** LSCE/IPSL. CEA-CNRS-UVSQ. Francia. helene.valladas@lsce.ipsl.fr ************ Australian National University. Canberra. Rainer.Grun@anu.edu.au ************* ORAU. University of Oxford. UK. thomas.higham@rlaha.ox.ac.uk ************** ICTJA-CSIC. Barcelona. rjulia@ictja.csic.es *************** IPNA-CSIC. Tenerife.vsoler@ipna.csic.es **************** Universidad de Valencia.ernestina.badal@uv.es ***************** Universidad del País Vasco. antonio.tarrinno@gmail.com ****************** Max Planck Institute. EVA. Leipzig. domingo_carlos@eva.mpg.de ******************* FASE. Asturias. jesusalonso002@gmail.com # Javier Fortea Pérez fue investigador principal del proyecto hasta el año 2009.
122  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Marc...
CANTABRIAN MOUNTAINS AND COASTLINE. Figure 1. Top: Development of the karstic system and location of the entrance and the Osario Gallery. Bottom: Osario Gallery floorplan and excavated sectors (2013). container of non-figurative scarce and enigmatic rock art (Pinto, 1975; Rasilla et al., 2011: 189-191) and above all, a significant number of Homo neanderthalensis fossil remains associated to the Mousterian lithic industry. Its incorporation into archaeological and paleoanthropological research was inevitably linked to the devastating war episode referred to above, because, for obvious reasons, the discovery of two jaws in 1994 led to legal proceedings, a police report and administrative proceedings that, after it was proved that they belonged to the Neanderthal species (Prieto et al., 1998, 2001; Rosas and Aguirre, 1999), ended in 1999 with the design and immediate implementation of a research project. The project posed a key question at the beginning: whether the material that appeared in the Osario gallery (Fig. 1) actually came from that place and whether the existing record had been removed. Both cases were proved to be true during the early stages of archaeological excavation. The project also tackled new questions: how did the remains get to this gallery? What is their chronology? What are the physical and paleobiological characteristics of the human fossils? What is the relationship between the fossils and artefacts associated with them? (Fortea, 2003, 2007b, 2007c, Rasilla et al., 2011a, 2011b; Rasilla et al., 2013). Obviously, at the outset it was difficult to predict that the archaeological and anthropological record obtained would greatly exceed expectations, as this is an exceptional site due to the amount and quality of the remains and the results, inferences and interpretations obtained and obtainable. From 2000 to the present, several parts of the karst system were excavated as the aforemen- 123
CANTABRIAN MOUNTAINS AND COASTLINE.  Figure 1. Top  Development of the karstic system and location of the entrance and the...
124 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD tioned questions were answered and according to the requirements that the study itself imposed. The intensity of activity in the Osario Gallery has been constant, GPR techniques even being used from the beginning to verify the burial theory and act accordingly, and it was confirmed that the materials were in secondary position, concentrated almost entirely in the area within sectors 2 and 3 from strip 10 to the north, with some remains in sector 4 (Fig. 1 and 2). Therefore, there was no funeral activity in this gallery. After answering the question of how the record got there, the questions pointed in a different direction, i.e., outside the system. Its character provided clear indications that the original place of deposit was not a settlement in use. Therefore, the area where the human fossils, lithic industry and few examples of associated fauna were deposited would need to be searched in addition to the archaeological site. To answer this question, two complementary activities were carried out. Firstly, geophysical analysis, gravimetric analysis and mechanical boreholes and archaelogical test pits were used to try to find out about the subsoil in the vertical exterior of Osario Gallery and identify shelters, galleries and channels in the karst system directly related to it and currently covered up (Fig. 2). Secondly, various boreholes were made in two shelters in the system, which had a high probability of containing the site: in La Cabañina and La Tumba. In addition, data was collected on the archaeological map made in the council and to survey the surrounding area. The results are different. In relation to Osario Gallery and its vertical exterior, a hypothesis on the filler model has been established by implementing geological and geophysical data from both sites (Cañaveras et al., 2011; Silva et al., 2011), while archaeological boreholes in the shelters have not offered anything related to Mousterian / Neanderthal, although La Cabañina has delivered an interesting but modest collection of lithic material, bones, pottery and fauna –wild and domesticated according to the stage– from the late Upper Paleolithic / Mesolithic, Chalcolithic / Bronze, Iron, Roman and Medieval that support the presence of these groups and, in some cases, the use of caves for their settlement (Rasilla et al., 2011c). Similarly, the study of the lithic industry that was being carried out in El Sidrón from 2005 and a little later in La Viña rock shelter (Santamaría 2006, 2012) revealed an interesting fact related to the raw material that, in parallel, also required prospecting of this abiotic element. There is a type of flint in the area, and even in theconglomerates inside the cave, of which much of the lithic industry of El Sidrón is made but which also appears in small amounts in the Middle and Upper Paleolithic levels of La Viña. This had an immediate corollary, as it was necessary to check whether this flint could be found in other sites and Paleolithic levels in Asturias and Cantabria and at the same time, develop a specific line of research regarding siliceous raw materials (Fortea et al., 2010; Santamaría et al., 2010, 2011). The information collected has allowed this raw material to be named “Sílex de Piloña”, and from now on, studies of prehistoric lithic industry must take into account this type of siliceous rock, as its proven travelling quality makes it a region-wide lithologic trace. Although the Cantabrian region joined these studies late, we are gradually discovering the siliceous raw materials present in Asturias and that could have been used by our ancestors (Duarte et al., e.p.; Tarriño et al., 2013). Inferences from different disciplines and analysis techniques 1. The arrival of the archaeological record (Sánchez-Moral et al., 2007; Silva et al., 2011; Cañaveras et al., 2011; Santamaría et al., 2010, 2011, 2012; Rasilla et al., 2011a; Santos et al., 2012a and b). After the primary position had been discarded and the Osario Gallery deposit had been buried, it was necessary to explain how the accumulated material had got here, as it invariably came from an area outside this one. Initially, the contributions must have come from the southern area, following the direction of the stratification and of the water that, coming from the runoff and the drips, happens in winter and spring inside the gallery. However, the data questioned this hypothesis and transferred the vehicle to channels located in the ceiling. This was because the vast majority of the material is concentrated between frames E-H/10 – E-H/4 and there are a few remains, with a slight increase in the number of fauna, which is always scarce, in squares D-E-F/20-23 (Fig. 1). Even if we project the archaeological evidence onto the floor
124  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  tion...
CANTABRIAN MOUNTAINS AND COASTLINE. Figure 2.Top: Interpretive geological cut of Osario Gallery. The geological features and the most significant geomorphological levels are shown in relation to the geophysical anomalies detected, the mechanical boreholes and the profiles of Osario Gallery (GO) and Main Gallery of El Sidrón cave (P). Bottom: Dissolution mesomorphology in the floor of Osario Gallery. A: Details of how the partitions and shovel shapes control the texture and geometry of the deposits (area F-G / 9). B: Detail of the partitions exposed in area E-F/8-9. C: View from the north of the centre of the gallery (September 2010) and diagram of the stratigraphic series. 125
CANTABRIAN MOUNTAINS AND COASTLINE.  Figure 2.Top  Interpretive geological cut of Osario Gallery. The geological features ...
126 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD plan, we see a cone shape in both cases, which supports an entrance from the outside. This phenomenon can be explained because outside, on one of the edges of a karst polje, there was a rock shelter where the archaeological record was deposited and a streamsinks a few metres down, as currently occur in La Cabañina rock shelter. For some reason, the system got blocked, coinciding with a storm or a stage of significant rainfall, the water rising up to the level of the rock shelter. When it became unblocked at a later stage, the deposit suddenly and very quickly entered the cave via channels, getting trapped in the Osario Gallery (Fig. 2). In this final position there was a post-deposition process focused mainly on the eastern wall favouring the flow of the aforementioned water and in a sinkhole that affected area E/9. Then the rock shelter broke apart and covered the quaternary sediment but its position has been located using geophysical analysis, gravimetric analysis and the corresponding topography. We are trying to access this place in order to check if there are any remains and, if there are, what the deposition process was like, as this would re-open the hypothesis that it had been a burial site. Furthermore, we can probably deduce, based on the position of the record inside the gallery, the arrangement of the elements (fossils, tools and fauna) in the original exterior rock shelter. The majority of the material was located in a specific area that entered via a channel that took it to sectors 2 and 3 of the gallery and a little of the material –particularly the fauna– was in a different area, but nearby, and entered via another channel that took it to sector 4 of the gallery (Fig. 1). 2. Chronology of the fossils and lithic tools (Torres et al., 2010, 2011; Wood et al., 2013). Both the taxa and the techno-complex to which they belong make it necessary to use different absolute dating systems, as at best, since ~55,000 BP,14C cannot be used and because depending on the result, we could be in the centre of the current debate on the Middle Paleolithic/ Upper Paleolithic – Homo neanderthalensis/Homo sapiensor the debate on the persistence of Neanderthal groups in the Iberian Peninsula up to chronologies around 25-24,000 BP. It is important to point out that, firstly, for various reason, not all of the procedures chosen have returned satisfactory results (e.g. Uranium/Tho- rium) and secondly, problems have been brought to light that could arise according to the methods used by the laboratories and also, problems related to subsequent archaeological interpretation. To minimise this impact, several remains were dated using various procedures (AARD, ESR and 14 CAMS), the dates of the first two coinciding quite well and those of the third being very different and out of range (Geochron Laboratories). As the dates could fall within the C14 range and in order to clear up the doubt generated, samples were sent to another laboratory (Beta Analytics). The results (between ~35,000 and ~41,000 BP) placed the El Sidrón record within the aforementioned debate. However, there was news of the dates obtained from a sample sent in 1998 to the Gif-sur-Yvette laboratory (48,500±2600 and 49,200±2500 BP) that significantly changed the vision and interpretation of the El Sidrón record. At the same time, samples were taken to date the sediment using OSL, into which the archaeological and anthropological materials fit quite well, to the ceiling and wall, and they were correlated with those obtained using other procedures. Finally, for the project led by Oxford University various unique European sites, including El Sidrón, were dated and the date of 48,400±3200 BP was returned. As maybe easily inferred, there is a disagreement between the dating of Geochron/Beta and Gif/Oxford and the main reason for this is the pretreatment used to eliminate any contamination. Geochron/Beta used the conventional system whereas Gif/Oxford used more sophisticated protocols, ninhydrin and ultrafiltration, respectively. For this reason, it is more sensible to take on the older dating values; the average value of these is 48,800±1600, and are generally more in agreement with the AAR, ESR and OSL values. 3. The nature of the anthropological collection and its paleogenomics (Rosas and Aguirre, 1999; Lalueza, 2011; Lalueza et al., 2005, 2011, 2012a and b; Rosas et al., 2006, 2007, 2011a, 2011c, 2012, 2013; Bastir et al., 2010; Dean et al., 2013; Hardy et al., 2012; Engelken et al., 2014, Castellano et al., 2014). What is perhaps more worthy of noteis that the existence of thirteen individuals and the practice of anthropophagy has been proved (Fig. 3). This has been possible because various specific molars
126  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  plan...
CANTABRIAN MOUNTAINS AND COASTLINE. and other specific bone parts have appeared and also due to the reliable presence of cut marks and deliberate fracturing of various human bones. We have other examples from the discovery of a large number of Neanderthal individuals in the same site but in this case, it has been possible to study the mitochondrial DNA of twelve of them, which provides a new view of some of their behaviour, such as the movement of females and the higher stability of males (patrilocality) in Nean- derthal groups. Various lines of research are open that will explore the interpretation of this “family” on a scale unseen before now, helped by the good general state of preservation of the fossils. Seven adults (3 , 3 , 1 ?), three teenagers (2 , 1 ¿ ?), two juniors (1 ?, 1¿ ?), one child (1 ?) and a total of ~2100 items give an illustrative source of information about laterality (righthandedness); the use of the mouth as a third hand by all adults and teenagers (grooves in the front Figure 3.Top left: Side view of a jaw and maxilla in anatomical connection. Top right: Cut marks on a humorous. Bottom: Photo and sketch of one of the reconstructed assemblages with a total of twenty-one pieces. 127
CANTABRIAN MOUNTAINS AND COASTLINE.  and other speci   c bone parts have appeared and also due to the reliable presence of...
128 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD teeth); episodes of physiological stress –malnutrition– (at least one hypoplasia coinciding with weaning and, in some cases, two or more); a mandibular abscess on adult 2 that must have caused chewing problems and pain; calculus present in all individuals with traces on a specimen of bitumen and consumption of medicinal plants; estimated height (between 164-171cm and between 153161 cm: an average of 164cm); a neurocranial morphology tending towards brachycephaly; a slight anatomic variation depending on the geographic area, those from the south tending more towards a wider face and less prognathous than those from the north; and their skeletal characteristics correspond to those known as classic Neanderthal. In addition, one adult female was red-haired and fair-skinned; one individual was blood type 0 (variant 001); Neanderthals had language, although at the moment we do not know how complex it was; one specimen could detect bitter taste but needed to eat a large quantity of the product to notice it; and the mitochondrial lineage of twelve individuals could be established. Thus, four adults (1 , 3 ), two teenagers and one junior belong to lineage A; one adult female belongs to lineage B and one adult ( ), one teenager, one junior and one child belong to lineage C. What is significant is that each adult female has a different lineage (A, B, C) and that all adult males belong to the same lineage (A). Therefore, at some time, a culinary practice was carried out on a group that was related to one another to some extent, which gives a very accurate snapshot of not only a biological fact but also a cultural one, their remains being left in the aforementioned rock shelter in a way that they were not affected by carnivores or rodents and could smoothly fossilise until they were discovered. 4. The qualities of the material culture and of the biotic and abiotic resources (Fortea et al., 2003; 2007a, 2007b, 2007c, 2009, 2010; Rasilla et al., 2011a, 2011b; Sesé, 2011; Sanchiz and Mar- tín, 2011; Santamaría 2012; Santamaría et al., 2010, 2011; Duarte et al., e. p.; Tarriño et al., 2013). Associated to the human fossils, around ~400 lithic artefacts with unmistakable Mousterian type-technology conditions and a few fauna remains have been found. The type of raw material –Piloña flint– used in general stands out; the refitting of various lithic pieces that currently represents 20%, but this will increase with new yearly incorporations (Fig. 3); the arrangement of the archaeological record and the scarceness and properties of the macro-mammal fauna (deer, large bovid, chamois, horse, wolf and bear) . The presence of Cretaceous flint in primary and secondary position in the Neogene conglomerates is of significant interest because the Neanderthals used this raw material, and some pieces of quartzite, to make their tools in direct relation to the priority activity deduced from the data exposed: tools for processing their conspecifics. This is based on the fact that this material entered Osario Gallery at the same time and the fossils and the refits also show that cores were collected from the immediate surroundings, they were knapped, different elements were used (flakes, denticulate tools, etc.) and they were left in the same place. A key food element, such as fauna, in principle does not have any relation to the activity considered because it only has marks from carnivores and, therefore, in this case it was not directly related to humans. It reached this deposit through the activity of carnivores and when the aforementioned event occurred, allof the material that was in this part of the site entered at the same time, getting trapped in the natural traps of Osario Gallery. Data suggests that it is a unique cultural fact and occurred quickly over time, focusing on a prominent task; therefore, it is vital to discover –and we are working on this– other Mousterian sites near El Sidrón to document this other part of human activity that is not reflected in the present one.
128  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  teet...
CANTABRIAN MOUNTAINS AND COASTLINE. Ana Cristina Pinto Llona* 25.000 years of Palaeolithic occupation at Sopeña (Asturias, Spain). Introduction Sopeña Sopeña was discovered in 2001. It has provided an ancient and intact sequence of layers that reveals an intense Middle and early Upper Palaeolithic human occupation. Due to the wealth of findings and the sophistication of recording techniques, the excavation progress is relatively slow. Much of what we know so far about Sopeña comes from the analyses of archaeological materials and features uncovered in a small 2x1 test excavation. These advances have been published in different works (Pinto Llona et al., 2005, 2006, 2009, 2012). Here I will summarize published aspects of Sopeña, and the work that I see as most relevant that has been carried out thus far and is currently in. Sopeña is located in the Onís county of northern Spain, bordering the Picos de Europa National Park, at 450 m. above sea level. It opens to the southwest (Fig. 1) and affords an unobstructed view of the valley of the river Güeña. It appears as a limestone shelter of modest dimensions. The entrance is protected by collapsed limestone blocks, covered by thick successive flowstone layers. Observations in situ and current studies (Ground Penetrating Radar) suggest that the observable floor, and the exposed stratigraphy, are the top part of much deeper sedimentary infillings, and thus that the cave is much larger than it appears to be now. All the levels are very archaeologically fertile, and lay almost horizontal. Adjacent to the flowstones, several limestone blocks of about 2 metres high, close the rockshelter to the northwest. On the upper surface of these there are deep linear engravings of the type that is sometimes assigned to the Aurignacian (Fortea Pérez, 2000-2001). On the side of one of these blocks, there is a pecked ithyphallic anthropomorph that is thought to be Neolithic in age. Looking at the plan map of Sopeña, its sediments clearly split in two different units. An imaginary line in the middle of the site, from east to west, vertically cuts the sediment like a knife would a pudding. The stratified archaeological levels on which we are working are south of this line; they completely disappear north of it. In the northern part there is a massive flow of yellowish and unstratified sterile sediments. We are currently digging a small test excavation in this area to better understand the sedimentary history of the site. It remains to be seen if this is a natural or a prehistoric man-made feature. The test excavation Figure. 1. Plan map of Sopeña. The shaded area indicates the test excavation. * In 2001 a 2 x 1 m test excavation was carried out (Fig. 1). The first archaeological layer (Level 0) appeared under more than one metre of cat- c/o J. Villarías, ILLA, CCHS-CSIC, c/ Albasanz 26-28, 28037 Madrid acpintoll@gmail.com 129
CANTABRIAN MOUNTAINS AND COASTLINE.  Ana Cristina Pinto Llona   25.000 years of Palaeolithic occupation at Sope  a  Asturi...
130 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD tle dung. Rockshelters in the area were used, and are still used, as cattle barns. Regularly, the dung is raked and carried to the nearby pastures as fertilizer. This practice might have scrapped away some upper deposits, although the total absence of anachronistic remains does not support this idea. The test excavation was dug under the main fault in the limestone of the rockshelter. This excavation reached 3 metres below Datum, and 16 archaeological levels were exposed. Each exhibits well-defined differences in colour and texture, and in the type, size and density of other materials in it (Fig 2). We stopped the test excavation given the friability of the sediments, and also because of the presence of collapsed blocks in the lower levels that could not be removed without damaging the exposed sequence. Pale levels alternate with darker, some rich in charcoal; the sedimentation is almost horizontal, and the stratigraphic and archaeological sequence doubtlessly continues to greater depths. In view of this neat colour alternation and horizontality, and also of the micro-sedimentary analysis, I do not believe that there are inverted sequences in Sopeña. The test has been dug by the east wall, which dips further to the east as we go deeper; on the upper levels it is not possible to Figure 2. Stratigraphic profile of the west wall of the test excavation. LEVEL LITHIC % Lithic Animal remains % Animal remains TOTAL % Total 0 95 43,38 124 56,62 219 0,53 I 417 26,77 1141 73,23 1558 3,78 II 400 24,43 1237 75,57 1637 3,97 III 1153 20,62 4439 79,38 5592 13,56 IV 41 11,71 309 88,29 350 0,85 V 104 15,27 577 84,73 681 1,65 VI 115 13,79 719 86,21 834 2,02 VII 436 16,50 2206 83,50 2642 6,41 VIII 202 4,57 4220 95,43 4422 10,72 IX 662 7,91 7711 92,09 8373 20,30 X 276 7,78 3271 92,22 3547 8,60 XI 1232 48,52 1307 51,48 2539 6,16 XII 187 18,17 842 81,83 1029 2,49 XIII 130 16,09 678 83,91 808 1,96 XIV 90 7,95 1042 92,05 1132 2,74 XV 319 5,42 5567 94,58 5886 14,27 Table 1. Number of finds per level in the test excavation.
130  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  tle ...
CANTABRIAN MOUNTAINS AND COASTLINE. stand up, and the sediment adjacent to the wall is loose enough. Therefore, dates on materials from this area could offer aberrant results. Although we have tried to select samples from the area furthest away from the wall within test excavation, there is little doubt that dates obtained from materials recovered during the excavations now in course will give a fuller picture. The great abundance of finds compensates for the small size of the site. In the 2 x 1 x 2 metres of the test pit more than 40,000 remains were retrieved. This fact suggests a frequent or even intense use of the site during the Mousterian, the early Upper Palaeolithic and the Gravettian. Table 1 shows the number of finds per level in the text excavation; bone fragments are more frequent in all the levels. We did not identify any “Transitional” type tool in this sondage although they might appear elsewhere in the site. The transition Mousterian-EUP at Sopeña seems abrupt, as some dates, and also the neat qualitative difference in the lithics which are present suggest. The lithic assemblage consists chiefly of debitage by-products and few formal artefacts, especially in the Upper Paleolithic levels. Regarding faunal remains, most bone fragments recovered are not taxonomically identifiable. Although not frequent, cut-marks on bone have been noted, and also carnivore action. We have not identified any human remains, which contrasts with the nearby site of El Level Sidrón (Rosas et al., 2012) where the same time frame has provided many cannibalized Neanderthal remains. Stratigraphy, archaeological attribution and absolute dates Levels I to XI can be assigned to the Upper Palaeolithic: levels I to VII to the Gravettian and levels VIII to XI to the early Upper Paleolithic. Levels XII to XV are Mousterian. Although few in number, we have several absolute dates, listed below, that suggest a relatively rapid sedimentary accumulation for the Gravettian sequence, and a slower one for the Middle Palaeolithic (Pinto Llona et al., 2012). One AMS 14C date from level XII points to this as the most recent Mousterian of the northern Iberian Peninsula (Maroto et al., 2012), and this is followed almost immediately by the early Upper Palaeolithic. The dates obtained for level XII could suggest a Mousterian occupation of some 10.000 years. We consider these chronologies as a guide while more absolute dates are obtained and the current excavation progresses towards the deeper older levels. The excavation method The excavation of the sondage was conducted recording the depth of each level with a home-made Lithic culture Method Dating Lab. Ref. Date BP ky CalPal online CalBC III Gravetian C14 AMS Beta-198144 21.020 ± 100 25.168 ± 377 X EUP C14 AMS Beta-198145 23.550 ± 180 28.496 ± 393 XI EUP C14 AMS Beta-171157 32.870 ± 530 37.359 ± 857 XI EUP C14 AMS GrA-39.760 34.470 + 650 – 450 39.726 ± 891 XI EUP ESR-LU Williams 2005SP02 40.300 ± 4.800 XII Mousterian C14 AMS GrA-35.500 35.500 + 650 – 460 40.336 ± 975 XII Mousterian C AMS Beta-198146 38.630 ± 800 43.052 ± 741 XII Mousterian ESR-LU Williams 2005SP03 49.300 ± 5.300 XIII Mousterian ESR-LU Williams 2005SP05 57.100 ± 12.500 XIV Mousterian ESR-LU Williams 2005SP08 50.400 ± 8.700 XV Mousterian ESR-LU Williams 2005SP10 57.200 ± 12.300 14 Table 2. Absolute dates of Sopeña. Level XI is the earliest Upper Paleolithic level. Level XII is Mousterian. 131
CANTABRIAN MOUNTAINS AND COASTLINE.  stand up, and the sediment adjacent to the wall is loose enough. Therefore, dates on ...
132 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD water level. As a rule, no vertical positions were taken of individual finds. However in the larger surface currently in excavation we have always used state-ofthe-art technologies. We dig by discrete stratigraphic units within each level, using a 50 cm. grid. Using a Total Station we map in three dimensions every individual find. The position of the Total Station is rechecked and corrected several times a day, so the average error is of about 2 mm. This machine is connected to hand-held computers, equipped with GIS software (ArcViewtm, ArcGistm) in the cave, and also connected to a barcode scanner that transmits the correct number of the find to the database. All the archaeological finds bigger than 2 cm have been 3D mapped using this method. It allows the vertical and horizontal mapping of finds in real time, as they are dug. All the sediments are carried to the field laboratory, floated, sieved and sorted. Despite its accuracy and due to the abundance of finds, this method is slow and the excavation progresses relatively slowly. After 10 years of work we have barely started to excavate Gravettian Level IV, in a 6 x 2 metres trench. Concluding remarks There can be no doubt of Sopeña being a site to be taken into account in the debate about the Neanderthal to Cro-Magnon transition in north- em Iberia. Whether with transitional industries or without them, continuous stratigraphic sequences in situ that document this time period are scarce in the north of the Iberian Peninsula. Aside from the excavation and scientific works, I have made a great effort to highlight the value of the site in the perception of the locals, doing everything possible to protect it, and to change the idea that Sopeña is only valuable as an occasional cattle –or, much worse, goat– barn. Local awareness is necessary, a consciousness of the true value of our Heritage and of the effort that we archaeologists carry out to bring it to the light. The returns will always be local, and social. Acknowledgements Our thanks to the Consejería de Cultura del Principado de Asturias, Concejo de Onís, National Geographic CRE, Wenner-Gren Foundation, Institute of Human Origins (ASU), Wings World Quest Foundation, Programa Ramón y Cajal del Ministerio de Ciencia e Innovación, to all the scientific collaborators of the Sopeña Project and to all excavation participants Also to Dr. Carbonell for inviting me to participate in this volume, and to the Editors for their work in coordinating it.
132  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  wate...
3 ebro valley, pyrenees and pre-pyrenees NORTH-WESTERN ATLANTIC BASINS. 133
3  ebro valley, pyrenees and pre-pyrenees  NORTH-WESTERN ATLANTIC BASINS.  133
Site Map numbering Abauntz 25 Arrillor 26 Atxoste 27 Cova del Parco 28 Roca dels Bous 29 Cova Gran de Santa Linya 30 Cova de l’Estret de Tragó 31 Fuente del Trucho 32 Fuentes de San Cristóbal 33 Gabasa 34 Kanpanoste Goikoa 35 Forcas 36 Nerets and Cova de les Llenes 37 Martinarri 38 Mendandia 39 Montsant valley ensemble 40
Site  Map numbering  Abauntz  25  Arrillor  26  Atxoste  27  Cova del Parco  28  Roca dels Bous  29  Cova Gran de Santa Li...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Pilar Utrilla *, Carlos Mazo *, Rafael Domingo ** The Abauntz cave (Arraitz, Navarre, Spain) Introduction Close to the village of Arraitz, the cave lies at around 30 metres above the left-hand bank of the Zaldazaín River. The site has long been known about; it was mentioned by J.M. de Barandiarán as early as 1956. Hydrographically speaking, the cave is located in the Ebro Valley, but the proximity of the Cantabrian Region and the prevailing climatic conditions act as a link between the two watersheds. The cave is strategically positioned between the flatlands to the south and the more rugged and wooded country to the north, lying, as it does, at 650 metres above sea level, close to the Velate Pass, which enables an easy crossing of the Pyrenees. It is a highland area made up of rounded rolling hills. Heavy rainfall (approximately 1,800 millimetres per year) is distributed across all four seasons. The site has been excavated over two phases: between 1976 and 1979 under the direction of Pilar Utrilla (Utrilla, 1982), and again in 1988, 1991 and 1993 to 1996, when Carlos Mazo joined the team to co-direct the dig. Since then the team has focused on laboratory studies, some of which have been of great interest and made a considerable impact in the media. The site’s ten archaeological levels document 50,000 years of recurring visits paid by peoples with different cultural traditions and needs: the cave served as a hunting lodge during the Solutrean, as a permanent settlement during the Mousterian, Middle Magdalenian and Neolithic, was sporadically settled towards the end of the Magdalenian and during the Azilian before being used as a burial ground in the Chalcolithic and as a hideout from the invasions following the collapse of the Roman Empire, thereby providing us with one of the most complete and complex stratigraphic sequences found anywhere in the Ebro Valley (Fig. 1). 1. The Acheulean Tradition in the Mousterian. Bears and Cleavers Signs of the oldest human presence correspond to level h, which was found in 1994 inside the cave, two metres below the preceding archaeological stratum (Solutrean); with around 8 m2 of it be- ing excavated. It contains no structures or hearths, although scorched bones were found. The thickness of the level is approximately 50 centimetres (Mazo et al., 2013). A number of bear teeth were dated by amino acid racemisation (T. Torres and E. Ortiz) at 47±7 ky, which would coincide with the date of AMS (GrA-16.960) >45000: both dates correspond to those of the industry, in which the relative abundance of cleavers suggests an MTA-type facies. Some 2,000 recognisable and extremely wellpreserved animal remains were found. Of these, 81.4% are of Ursus spelaeus, without signs of anthropic activity, which appear to be an accumulation linked with hibernation and death by natural causes. Of the rest, 10% correspond to carnivores (Panthera pardus, Canis lupus, Vulpes vulpes, hyenas) and only 7% to ungulates. Human consumption is documented (cut marks on the bones of Cervus), as is that of carnivores (gnawed bones of Rupicapra), which follows a pattern similar to that detected in the Moros de Gabasa Cave. Far fewer remains have been found of bovids, reindeer, ibex horses or rhinoceros. Almost 90% of the faunal remains are due to the cave being naturally used as a shelter by bears and other carnivores. Of the 42 lithic remains the most abundant are 11 cleavers (26% of the total) (Fig. 2), which were found alongside 2 hand axes, 2 racloirs, 2 scrapers, 3 truncations and 4 retouched flakes. The cleavers are manufactured using limestone, basalt and other volcanic rocks, whereas flint was the raw material of choice for hand axes and other pieces, just as it was at Najerilla (Utrilla and Mazo, 1996d). This brings the Abauntz cave into line with other sites found in the hinterland of the Bay of Biscay coast line (Castillo, Morín, Pendo, Olha, and so forth) where cleavers were manufactured using identical technology (Mazo et al., 2013), which entitles us to revive the term of Vasconiense (Basque) proposed by Bordes and which has today been reinstated by several authors. 2. Passing Solutreans and Gravettian Echoes Some of the materials found in the disturbed level (Gravettian and Vachons projectile points) enable * Área de Prehistoria. Universidad de Zaragoza. C/ Pedro Cerbuna, 12. 50009 Zaragoza ** Área de Prehistoria. Universidad de Zaragoza. Pza. Constitución, s/n. 22001 Huesca 135
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Pilar Utrilla  , Carlos Mazo  , Rafael Domingo     The Abauntz cave  Arraitz, Nava...
136 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. The stratigraphic sequence documented in the Abauntz cave indicating the archaeological levels, most significant materials, dates and cultural allocation. one to suggest the possible presence of a Gravettian level in the cave that has since disappeared due to karst activity (Fig. 3.1). With respect to level f, altered following the sedimentation thereof, it still contains residual elements that would appear to date back to the Upper Solutrean, though the settlement of the site is by no means intense or especially fertile. The highlights of the limited number of finds are a thick flat retouched perforator and some beautiful notched projectile points, most of which are broken, which leads one to believe that the site was used as a temporary hunting camp where broken weapons were replaced and repaired (Fig. 3). This flat retouch is found alongside two projectile points featuring abruptly retouched notches that look like they could have come from the Cave of Salpetriere, thereby affirming the “hinge-like” nature of the site that links it with both Bay of BiscayAtlantic and Mediterranean influences. This is the westernmost find as regards this morphotype, which is known in the foothills of the central Pyrenees (Chaves, Trucho), where the projectile points are “more like those found in the Cave of Salpetriere” than those of the coastal regions (such as those with a curved peduncle found at Ambrosio or Parpalló). This lends added further weight to the idea of a communication route with the north of the Pyrenees via the central passes (such as that of Cerdaña) as well as via the more logical coastal route (Bocaccio and Utrilla 2013, Domingo et al., 2012). This technical duality found at Abauntz would reflect contacts between the peoples of the Ebro Valley with a tradition rooted in the Cave of Salpetriere and those moving along and settling the Cantabrian (Bay of Biscay) corridor running between Aquitania and what we now know as Asturias, in northern Spain. 3. Trans-Pyrenean Hunters of the Middle Magdalenian Level e, the best known and preserved of the stratigraph, represents the only well-documented human presence during the Middle Magdalenianin the Ebro Valley, although it is clearly linked with the Cantabrian-Aquitanian world. It is dated (bone, AMS) at 13500±160 (OxA-5983; 16413±423 cal
136  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES BP). Identified during the first campaigns (Utrilla, 1982), Mazo used it to support his theory of the functionality of the Magdalenian lithic tools, which threw up a number of surprises. The burins had never been used for notching: on the truncated types the active part was the selfsame truncation, which was used for scraping skins; the blunt end of the burin was used to make the tool easier to grasp. Dihedral burins and perforators found in the same area had perforated skins; on the double burin-scraper implements the active part was the scraper blade, with the burin only being used for grasping purposes (Mazo, 1989). Microspatial studies, palynology and functional analysis have made it possible to propose a reconstruction of the activities performed in the two main chambers of the cave (Utrilla and Mazo, 1992; Utrilla et al., 2003). Significant concentrations were documented: dihedral burins and perforators in the entrance area in front of the hearth; scrapers and truncated burins in the right-hand side of Room 1, around a hearth; spears and other bone projectile points in the intermediate passageway and in Room 2. On the other hand, within the northernmost angle of the cave hardly any lithic materials were found, but there was abundant pollen from plants suitable for the preparation of rest areas: bracken and rushes. Regarding the activities performed, we would suggest that the flint was reduced near the entrance, that is, in the best-lighted area; further in were the spaces used for working the skins (perforation and scraping), and right at the back the rest area. The interior room could have been a rest area where the weapons were stacked against the wal lor another work area used for the processing of skins, which could be sheared and/or smoked there. Several small-sized postholes were found between the passageway and Room 2, which leads one to imagine a structure for hanging skins consisting of a framework of posts that would have separated one area from the other. Altuna and Mariezkurrena (1996) documented the only remains of Saiga tatarica found on the Iberian Peninsula, thereby confirming the relations between Abauntz and areas of the northern Pyrenees: 6 bones, barely 1% of one level and dominated by Rupicapra and, to a lesser extent, Cervus and Capra. The authors state that Saiga tatarica reached its maximum extension in Western Europe during the Magdalenian (especially the Middle Magdalenian, when they were to be found in places as close to Abauntz as Isturitz and Dufaure). The remains (5 phalanges and 1 central tarsal) lead one to believe that the animal was not hunted close to the site, but that these bones were attached to a skin brought in by people who came from Aquitania. Figure 2. Cleavers found in level h of the Abauntz cave (Navarre, Spain) (1 to 4) compared with other of the Najerilla River site (La Rioja, Spain) (5 to 8). Figure 3. Vachons projectile point (No. 1) and Solutrean materials from level f 137
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  BP . Identi   ed during the    rst campaigns  Utrilla, 1982 , Mazo used it to supp...
138 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD The large number of bone industry remains could be seen as confirmation that Abauntz would appear to have been dependent upon the large settlement of Isturitz, although it also presents similarities with communities of western Cantabria. The complex decorative motifs that embellish the stag horn rods of Abauntz are almost identical to those found at sites such as Caldas and La Viña in Asturias, Isturitz in Pyrénées-Atlantiques or La Madeleine in the Dordogne, which went to make up a Magdalenian koine along the entire coastline of the Bay of Biscay and in the southeast of France. Abauntz is located at a key strategic point; it is close (and with easy access) to large settlements such as Isturitz. We are able to point out three other decorative motifs found at different sites: a series of triangles framed within parallel lines (chevrons emboités), reversed parentheses and bison heads in profile. (Utrilla and Mazo, 1996 a and c; Utrilla et al., 2013; Duarte et al., 2012). The level also provided two bones engraved with marks in multiples of seven, which evoke calendar counts referring to lunar phases that are relatively simple to follow with a minimum observation of the night sky (Utrilla, 2004; Utrilla and Martínez-Bea, 2008, Mazo et al., 2008): one hyoid horse bone, perforated for suspension, and a mammoth ivory pendant. The hyoid has two series of 13 and 14 deep incised in its sides. The ivory pendant bears five series of marks with the sequence 10-14-14-14-14. This type of pendant has also been found at sites such as La Güelga and Tito Bustillo (2 at each, also in multiples of seven) or at the faraway cave of La Marche (in the French Department of Vienne), where at least three perforated hyoids bearing marks of this type have been found. 4. The Late Magdalenian of Level 2r Dated using C14 at 11760±90 (OxA-5116; 13643±151 cal BP), this is one of the less productive levels with barely a few dozen lithic remains being found. The most important discoveries regarding this period of settlement are three pieces of portable art, namely, three stone blocks engraved with diverse figures. After an intense period of study (Utrilla and Mazo 1996b and c; Utrilla et al., 2004 and 20072008), in 2009 an article that was to have great media impact was published in the magazine JHE claiming that one of them contained Western Europe’s oldest map (Utrilla et al., 2009). Included on one of the faces of block 1 is what we interpret as being the oldest representation of a map showing the immediate surroundings of a prehistoric settlement. A number of engraved lines show what would appear to be the landscape as seen from the cave: a mountain, streams and ravines running down from the hilly area to the plane… and some of the Figure 4. Horse engraved on block 3. animals that inhabit these places: extremely schematic ibex (bodies in profile, head front-on indicated using two superimposed “V” shapes) places around the mountain, two young bovids on what would appear to be the plane… Then there are marks that have been interpreted as paths and fords on what would appear to be a “guide” to the immediate area for the group of humans living there or, given that the block was left abandoned there, for other hunting parties. On the same face a large stag is accompanied by a series of spiral markings which could represent the noises or odours of its bellow. In contrast to the other two blocks, a natural hollow in block 2 could well have been used as a fat or bone marrow lamp. Its most eye-catching feature is the representation of a horse accompanied by several goats that repeat the schematic model found on block 1. Other marks similar to those interpreted as being symbols of the landscape can also be observed: a possible watercourse, a path… Block 3 bears the most carefully executed representation, artistically speaking. Subtriangular in shape, it shows the engraving, with a profusion of anatomical details, of a protome of a horse, the animal that most characterises both static and portable Magdalenian art, especially in the area of the western Pyrenees (Fig. 4). Due to the frequency of its representation, is could be seen as being the “clan symbol” of a human community with strong links to Aquitanian settlements such as Isturitz, where L. Mons (1996) identifies 180
138  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  The ...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES horse or figures, or Duruthy, where the symbolic value of the horse is evident (Cleyet-Merle 1996). 5. The Last Hunters: The Azilian of Level d This level only appears in Room 1. Its formation is linked with currents of water that washed materials from the Magdalenian levels down into it. Snail shells abound in a level that is archaeologically poor, in which around 70 retouched pieces and cores were found. Small reverse blades account for almost 41% of the finds, and unguiform scrapers and Azilian projectile points are also documented for a level dated using conventional C14 at 9530±300 (Ly-1964; 10858±405 calBP). This brief human occupation of the site forms part of a complex chronocultural panorama that has yet to be fully defined (Soto et al., i.p.). 6. The Neolithic of Level b4 Level b4 bears witness to a notable period of human occupation with abundant structures: a number of holes and large hearths. A sterile layer of angular stones (b3) sets this level apart from the one below it. It produced blades marked by use and featuring abrupt retouches, smooth, crudely-made pottery alongside finely-crafted vessels that are black in colour with the surface having been worked with a spatula and burnished, a small polished stone axe head and two abruptly retouched and double-bevelled geometric microliths. Five holes were identified in the central area of Room 1. One could have been used for disposing of the ashes from the adjacent hearth and another contained diverse objects and materials: a pebble-based hammer-pestle-smoothing tool, a small polished hand-axe (less than 3 centimetres in length), a retouched blade and fragments of smooth pottery that once formed part of at least two vessels, one with a rough and the other with a spatulated finish. The level is carbon dated using the conventional C14 technique at 5390±120 (I-11309; 6158±129 cal BP), thereby placing it the Late Neolithic. The area around the cave could well have been a place of temporary residence for groups of Neolithic peoples possibly linked more with animal husbandry than with crop cultivation. 7. A Large Collective Tomb: Chalcolithic Burials. It is impossible to ascertain the exact number of burials carried out at the site due to the fact that many of the bodies were later moved and intentionally burned; the result is a thick layer of scorched bones covering a good part of the surface area. The total would exceed one hundred. These remains have been studied by J. I. Lorenzo, D. Turbón and D. Campillo, and it was E. Fernández (2005) who carried out ge- netic analyses and detected the presence of MiddleEastern and even African lineages. It has been possible to identify four types of burial. The oldest are the shaft tombs, individual or double, which usually include bone spatulas being placed close to the femur of the deceased. One tomb, named “Alberto”, contained a mature individual, curled up and accompanied by two bone spatulas, a flat retouched projectile point and two stone barrel beads. This type of shaft is more frequently found in the passageway and in Room 2. A human bone taken from one of the shafts was dated at 4370±70 (CSIC 785; 5012±124 cal BP). The burned remains occupied a uniform surface area of around 16 m2 in Room 1. It would appear that incinerating the bodies did not form part of the burial ceremony, but that it was carried out some time later, perhaps for reasons of hygiene. Around this large area of scorched bones shafts were found that look like they had been sunk to avoid it. The archaeological material found in these included leaf-shaped projectile points, many with heat cracks, and a number of necklace beads. The level was dated at 4240±140 BP (Ly-1963; 4798±205 cal BP), almost contemporary with the shaft tomb burials. A third type of tomb consists of a stone cist structure composed of blocks sunk vertically in the ground and covered with a large slab (of non-local sandstone). It contained the remains of two adult males, two young women and two children, all with their legs folded. Remains that might have been related with the aforementioned individuals were found in nearby graves: two adult males, a woman of over 30, an infant, a child and a young woman. They were dated at 4025±35 BP. The last type of burial is that of the deceased being deposited without any visible structure. These are the most recent interments and must have been distributed throughout the entire cave. Near the entrance of the cave the recently deceased were laid on top of the burned human remains. The materials found with them differ insofar as they include peduncle and finned projectile points alongside their leaf-shaped equivalents. An outstanding feature is that of the presence of pendants made from wild boar tusks together with diverse adornments, as well as the almost complete absence of ornamental objects in Room 2, which contrasts sharply with the profusion found in the area around the entrance. Those in the entrance were dated at 3975±35 BP and those of the second room at 3900±35 BP. (Fig. 5) 8. Romans Sheltered in the cave during the Fall of the Empire The last human presence recorded in Abauntz cave occurred during the period of instability and social violence towards the end of Roman rule when uncon- 139
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  horse or    gures, or Duruthy, where the symbolic value of the horse is evident  C...
140 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 5. Types of Chalcolithic Burial. trolled armed groups (bagaudae) or bands who were struggling for power sowed terror among the population, either for political or purely subsistence purposes. It was during these hard times that a series of holes were dug (a hoe was found on the site dating back to this activity) into which objects of value were put, including silver rings and more than 300 coins that enable these events to be dated at between the reigns of Constantine and Arcadian: the minimum post quem date is 408, the year in which the most recent coin found was minted. A hiding place used during the barbarian invasions is another viable interpretation. 9. Conclusion The surroundings of the Abauntz cave, which is geographically located in the Ebro Basin, clearly link it with the settlements of the coastal areas of the Bay of Biscay-Aquitaine. It was used, with different functions and intensity, by numerous human groups for a
140  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES period of fifty thousand years. The first of these were Neanderthals, Acheulean Tradition Mousterians, whose occupations alternated with those of bears and other carnivores that sheltered and hibernated in the cave. The cave was sporadically used by groups of hunters during the Upper Palaeolithic: perhaps during the Gravettian and most certainly, although without dates, during the Upper Solutrean, by a human group that combined Atlantic and Mediterranean influences. The Middle Magdalenian period of Abauntz, up until now the only contemporaneous level in the Ebro Valley, is the most intense and significant occupation. It has provided microspatial information (activity and rest areas), functional information (the “anomalous” use of truncated burins and scraper-burins used for scraping) and data about regional contacts via the decorative motifs of its bone industry that strongly bind Abauntz not only with settlements such as the super-site of Isturitz and others of the Aquitanian area, but also with sites such as La Viña and Caldas in Asturias. At the end of the Magdalenian there was a sporadic yet extremely interesting occupation that left us with three magnificent blocks engraved with portable art: a map of the surrounds (block 1), a lamp decorated with similar subjects (block 2) and an isolated representation of a horse that could be the symbol of the Isturitz Clan (block 3). The last hunters to visit the cave were people of the Azilian tradition: the low intensity of this particular human occupation is matched by the poor sedimentological condition of the level. The first non-hunter gatherers present were Neolithic people who possibly used the cave for purposes of animal husbandry. Andoni Sáenz de Buruaga * Arrillor cave (Araba, Basque Country): Climatic and industrial evolution during the Upper Pleistocene 1. Geographical settings Arrillor cave is situated in the southern slope of the Basque Mountains –topographical ridge that * During the Chalcolithic the cave was used as a burial ground for several centuries and features three main interment methods: single shaft, cist or simply depositing the deceased on the surface, with at least one hundred people being laid to rest there. Some bodies were later intentionally burned, probably for hygienic and not ritual purposes. The last specific use to which the cave was put dates back to the beginning of the 5th century AD, when some individuals hid objects of certain value there (silver rings and several hundred coins) against a backdrop of great insecurity and social violence, when the first groups of barbarians entered the Iberian Peninsula. From 1976 to the present day our team has been working on the Abauntz cave, both in the field and in the laboratory. As we have seen, this is one of the most notable prehistoric sites, due to its stratigraphic strength and importance, in southwest Europe. A monograph regarding the first archaeological campaigns (Utrilla, 1982) and dozens of references made in specific or general articles bear witness to the importance of the discoveries made and the conclusions reached. A summary of all of these will appear shortly in Quaternary International as part of the minutes of the Conference held in Bilbao at the end of 2013. All this research has been made possible thanks to the subsidies received from the Regional Government of Navarre and to the help received from successive research project grants awarded by the Spanish Ministry for Economic Affairs and Competitiveness (MINECO), one of which is still in force, namely, HAR 2011-27197: “Broadening New Horizons, Rethinking Ancient sites in the Ebro Valley”. The signatories form part of the Consolidated Research Group “The First Settlers of the Ebro Valley”, H07 of the Regional Government of Aragon. divides the Cantabrian drainage (to the north) and the Mediterranean drainage (to the south)– in the middle of the Gorbea Massif, north of the Araba province. Administratively, Arrillor cave is located Círculo de Estratigrafía Analítica. Departamento de Geografía, Prehistoria y Arqueología. Facultad de Letras. Universidad del País Vasco (UPV-EHU); (andoni.buruaga@ehu.es) 141
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  period of    fty thousand years. The    rst of these were Neanderthals, Acheulean ...
142 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD framed in a broader research programme aimed at discovering Paleolithic sites in the territory of Araba. Since then, seven archaeological campaigns, focused on the cave entrance, have been undertaken between 1989 –1994, and in 1997. The excavated area was divided using a 22 m2 grid, and in some parts reached 5 metres deep. On that substantial archaeological deposit, a significant number of human occupations, particularly Mousterian evidence, occurred along different climatic episodes of the first half of the last Würm glaciation. Also, some data from Magdalenian occupation were even registered in one level (Sáenz de Buruaga, A., 1989, 1990, 1991, 1992, 1993, 1994, 1997). In 1997 regular excavation was terminated. Figure 1. Geographic settings of Arrillor. in the municipality of Murua, in the Zigoitia valley (Fig. 1). Its geographic coordinates are X=521.057, Y=4.761.540, and it is located at 710 metres a.s.l. The entrance cave has an eastern/north-eastern orientation and lies in Urgonian limestone formation. The cave is 150 metres long, and its axis is east-west oriented. The site is situated at the southern end of a rocky spur, on the confluence of two ravines, Asunkorta (from the west), and Errkaseku (from the east); both flow together into the Zallas River, 25 metres in a straight line from Arrillor cave and 12 metres down the slope. This orographic context provides a strategic location because of the hydrographic resources available and its privileged position to monitor game movements. 2. Research on Arrillor cave The first archaeological research was conducted in 1959 by J.M. Barandiarán and D. Fernández Medrano, who excavated a test trench, 3.5 m2 and 1 metre deep, at the entrance of the cave (Barandiarán and Fernández Medrano 1959). The excavation provided a small collection of lithic artefacts and faunal remains of prehistoric appearance, but results at that time did not provide any precise chrono-cultural diagnosis. Thirty years later, in 1989, A. Sáenz de Buruaga set up a systematic study project on Arrillor cave, 3. Sedimentary strata dynamics Sedimentary stratigraphy deposited at Arrillor cave correlates with isotopic stages 3 to 1, and represents diverse events between the Würm II and the Würm IV, according to the chronostratigraphic alpine denomination The sedimentology study is in progress, but the paleoclimatic correlations and the subsequent interpretation established during the excavation is still valid (Hoyos et al., 1999). Moreover, the identification and definition of the levels at the archaeological sequence followed the theoretical basis of “Analytical Stratigraphy” (Laplace, 1971; Sáenz de Buruaga, 1996; Sáenz de Buruaga et al., 1998). The stratigraphic sequence of Arrillor has a depth of ca. 5 metres. It can be divided basically into three sedimentary complexes: one in the bottom part, with cryoclastic and alluvial deposits; one in the middle, of fluvial origin; and one in the upper part, also of cryoclastic formation, with severe postdepositional alterations in the middleupper section (Fig. 2 and Tab. 1). • The lower sedimentary complex encompasses a series of levels with cryoclastic origin (Cag, Blm, Car, Clmg, Cglm and Clm), coexisting with other levels of fluvial origin (Lgj, Ln and Lj). This sedimentary complex is about 50 to 115 cm thick. The presence of this type of deposits with well-differentiated sedimentary components, and its stratigraphic position in the general sequence, suggest its formation was probably during a more advanced cold temperate phase of Würm IIa, in the climatic transition from Würm IIa to Würm IIa/b.
142  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  fram...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES ers of this material, sometimes thicker and sometimes thinner, compose the deposit. At the base, silts replace sands. On this arranged succession of yellowish sandy layers (designated Sa), greyish plastic clays layers (designated Agp), and other reddish sandy compacted layers (designated Srk), 10 sedimentary units (SU) have been identified. Though, in general, Sa layers and Agp layers are archaeologically sterile, Srk layers have yielded diverse archaeological evidences: basically, faunal remains and lithic artefacts that might be related with Mousterian techno-complexes. The most remarkable assemblages have been recovered at the Amk structural assembly (composed by layers Amk, H-Amk, and I-Sa 8), in the lower half of this fluvial formation, and in its upper third at level Smk-l (Fig. 3). Figure . 2. Sedimentary deposits in the anterior frontal sector, of the central area of the excavation. In these cryoclastic levels, several Mousterian lithic artefacts associated with some faunal remains were found. • The middle sedimentary complex is clearly of fluvial origin. It is 245 to 305 cm thick. In general, this part formed as a result of cyclical succession of mechanical weathering, transportation and erosion, and analogous processes of deposition generated under temperate climatic conditions which, in the end, can be correlated with an interstadial phase (probably Würm IIa/b). Sand is the prevailing sedimentary component. Lay- Table. 1. Analytical Matrix of the stratigraphic sequence at Arrillor cave. 143
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  ers of this material, sometimes thicker and sometimes thinner, compose the deposit...
144 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD The Würm IIb strata initially maintains a fine brown sand composition with an increasing presence of gelifracts: first as cobbles (Smc), and later on as boulders (Smb), corresponding here with a particularly cold phase of the stadial. Afterwards, a change in the matrix is registered: sands are replaced by brown silts in a new stratigraphic episode (Lmc). While Smc and Smb have yielded a discrete number of faunal and lithic Mousterian evidences, Lmc provided a more substantial archaeological assemblage of Mousterian affiliation. This lower phase ended with stratigraphic unit (Lam), somewhat representative from a cultural point of view. Furthermore, a new sedimentary hiatus (VLam), originated by a solifluxion flowstone occurred in a Holocene episode, removed, on the one hand, most of the Würm III and IV deposits from the central cave area, and on the other hand, piled up a new yellow silty-clay deposit (Ala). Only in a small lateral part, the stratigraphic evidences of advanced stages of Würm were preserved: first, in a level of cobbles and boulders (Labc), which are archaeologically sterile, and that can be related with middle and late phases of Würm III; and, later, in a yellow silt deposit (La), formed in the Würm IV, that contains lithic evidence of the Magdalenian period (Fig. 4). 4. Human occupations at the cave Figure. 3. Stratigraphic posterior frontal profile at Arrillor: upper and middle deposits on the central area of the excavation. • The upper sedimentary complex is, in general, of cryoclastic origin. It is 65 and 95 cm thick. After a net erosive contact on the base (V-Sa 1), an infill piles up. Two phases must be differentiated at this deposit: a) one in the lower part, which should be related to the stadial Würm IIb; and b) another one in the upper part, formed in the Würm III and IV. On top, a post Pleistocene episode is superimposed. Various archaeological references have been documented in the total 22 stratigraphic features differentiated on the sedimentary deposit of Arrillor cave. Before proceeding further, it is necessary to briefly address two issues about these particular archaeological levels: 1) with the exception of one level considered Magdalenian, the rest refer to different parts of the Mousterian period; and 2) in most of the archaeologically fertile layers, the amount of artifacts is scant and with very low diagnostic value, insufficient to provide a precise morpho-technical characterization: only layers Amk, Smk-l and Lmc, from bottom to top, gather the most significant and solid evidence of human occupation. In that sense, besides lithic artifacts, a very interesting collection of faunal remains has been properly identified (Castaños Ugarte, 2005). This data enhances the information regarding the environmental context of each deposit.
144  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  The ...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES precisely, in this particular context the human molar (AR-1740) of a young hominid aged 9-13 years old was found (Bermúdez de Castro and Sáenz de Buruaga, 1999) (Fig. 5). Also, from this stratigraphic formation, two charcoal samples were AMS dated (45700 ±1200 B.P. (OxA-6084) and 45400 ±1800 B.P. (OxA-6251)) setting it chronologically in the middle interstadial Würm IIa/b. Concerning the faunal remains, there is a predominance of ungulates: red deer and bison being the best represented species; on the contrary, goat and equus remains are scarce, and roe deer and rhinoceros are very rare. Figure 4. Upper deposits in a central area of the excavation. On the bottom, after a net erosive contact, the upper section of the yellow sands belonging to the middle sedimentary complex. • The Smk-l deposit, situated in the upper third of the middle fluvial complex, offers an interesting batch of lithic tools of Mousterian tradition (sidescrapers, point, etc.), and illustrative of the levallois debitage method, knapped almost exclusively on lidite. It could be asserted that the collection is specialized in the exploitation of this raw material (Fig. 6). Its particular techno-typological composition refers to Mousterian industries which are fairly specialized at sidescrapers and points, on large-sized and flat blanks, and using the levallois method. • The structural ensemble Amk, placed on the bottom third of the middle fluvial deposit, includes the following stratigraphical units, from bottom to top: Amk, HAmk, and I-Sa 8. In this group, a significant and varied set of lithic artefacts knapped on flint and quartzite was recovered. Among others, typical Mousterian artifacts such as sidescrapers, points, denticulates, and other pieces that are more morpho-technologically evolved such as scrapers, truncations and burins. These typological trends allow characterization as a polymorphic Mousterian, enriched with leptolithic performs of short-size pieces. The upper part of this sedimentary complex presents a horizon of hearths, of very dark colour, accumulated over the whole area (H-Amk). This horizon contains a large amount of charcoal and bones. More Figure 5. Molar of a young Neanderthal retrieved at the stratigraphic deposit H-Amk and dated around 45000 years B.P. (Picture: Archaeological Museum of Araba Bibat; D.F.A.). 145
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  precisely, in this particular context the human molar  AR-1740  of a young hominid...
146 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 6. Types in lidite from level Smk-l (Picture: Archaeological Museum of Araba Bibat; D.F.A.). The faunal spectrum still shows the predominance of ungulates, although together with bison and red deer, goat becomes more important; Equus on the contrary, remains very scarce. One AMS radiocarbon date of a bone gave 43100±1700 B.P. (OxA-6250), setting the last human occupation preserved in the middle fluvial complex, at a later time in the interstadial Würm IIa/b. • Level Lmc is situated in the bottom half of the upper cryoclastic complex. From a typological industry point of view, this level includes a more substantial lithic collection, which is well differentiated typologically from the previous repertory of the underlying levels Smb and Smc, correlated with a new cold climatic stage on Würm IIb, and associated with Mousterian industrial complex of denticulates, carenage blanks, and preferably flaking of black flint which is locally available and mediocre in quality. Thus, in level Lmc there are also Mousterian industrial complex artefacts (sidescrapers, points, denticulates), together with some evolved morpho-technical types (endscrapers, burins), of acceptable quality of flint, and also of other local raw materials (quartzite, quartz). This industrial complex can be defined as Mousterian of sidescrapers and short-sized flat blanks. The faunal assemblage is still characterized by ungulates, clearly prevailing red deer, followed by the chamois and bison, and in lesser extent rhinoceros, roe deer and equus. The AMS radiocarbon date on a bone from this level provided 37100±1000 B.P. (OxA-6106), which situates the occupation on a later phase of Würm IIb, immediately previous to the moister and temperate episode of Würm II/III.
146  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES 5. Concluding remarks Systematic excavation carried out for 7 campaigns, between 1989 and 1997, on this cave, situated in the northern area of the Araba province, have provided a rich sedimentary deposit that includes significant lithic and faunal assemblage from Upper Pleistocene, as well as a Neanderthal tooth. This deposit that is five metres thick, goes from an early phase of Würm II (or Würm IIa) to the Late Glacial of Würm IV. The most significant sedimentological and archaeological deposits concern the Mousterian period ñparticularly stratigraphic levels Amk, Smk-l and Lmc –, although less representative, isolated Magdalenian evidences were documented. From bottom to top, three sedimentary complexes have been defined: Figure 7. Mousterian hearth form level Smk-l, on the upper section of the middle sedimentary complex of Arrillor. a) The lower complex, of cryoclastic and fluvial origin, includes some levels with Mousterian industries. Regarding climate conditions, this complex is related with the cold stage of the Würm IIa. b) The middle complex: with a fluvial genesis, that preserved a considerable number of Mousterian levels and horizons. This complex developed within the moister and temperate phase of the interstadial Würm II a/b. Some sections of this complex have been dated between 46000 and 43000 B.P. (Fig. 7). c) The upper complex: of cryoclastic formation, contains the last Mousterian industries, and preserves a small number of Late Magdalenian evidence. The sedimentary sequence seems to have been formed from different deposits associated to cold stages of Würm II (or Würm IIb), of Pleniglacial (or Würm III), and the Late Glacial (or Würm IV). One radiocarbon date from the middle-bottom deposits situates it in the late phase of Würm IIb around 37000 B.P. The Arrillor Cave deposit addresses some key questions regarding the evolution process and the climatic and environmental transition from the Early Würm to the Late Würm in the inner territories of the Basque Country. Particularly worth mentioning is the preserved part around ca. 55000 and 35000 B.P. that offers quite remarkable sedimentological evidence. This meaningful sequence should be contrasted with other Mousterian sites nearby, such as Axlor (Bizkaia), and Lezetxiki (Gipuzkoa), all of them in the same environmental unit of the Basque Mountains, forming a triangular arrangement where sites are separated by hardly 20 linear km (Sáenz de Buruaga, 2000: 62). 147
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  5. Concluding remarks Systematic excavation carried out for 7 campaigns, between 1...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 148 Alfonso Alday* The site at Atxoste (Vírgala, Álava) Atxoste is a rock shelter located by the river Berrón, south facing in the middle of an extremely varied landscape with valley and mountain resources. It gave shelter to late Upper Magdalenian communities and although its roof collapsed at the beginning of the Holocene, the conditions of the site resulted in its occupation throughout the Mesolithic (all of its phases are represented), the Neolithic, the deposition of burials being its final use in the Metal Ages. The stratigraphic sequence is more than six metres thick and it is subdivided into several sections, where only the base layer –clay– and the surface –vegetable matter– are not of archaeological interest. We distinguish two major sedimentary units, the first from the late Pleistocene of rapid formation, and the second from the Holocene of slower constitution. The internal coherence of the sediment, the absolute dating and its cultural context guarantee the viability of the deposit that is, for the chronocultural environment in which it is encompassed, the most significant site in the Ebro basin, at least. Excavation of the shelter was carried out, without interruption, between 1995 and 2006 under the direction of A. Alday. Stratigraphic sequence Level VIII sedimentary base of humid, plastic and compact clays from Maestu diapir. Its initial light brown colour turns orangey. It has no archaeological interest. Level VII: 80 cm thick, with a dark loam clay and organic material matrix, which incorporates large blocks encrusted at different times, compressing the soil. Cold and humidity caused waterlogging and frost weathering which fractured the blocks. Thus the level will evolve internally, with different colourations, textures and higher fracturing. Culturally, it corresponds to the late * Upper Magdalenian, its lithic retouching is characterised by an extensive representation of back edged arrowheads and blades and scrapers. It has no noteworthy bone industry and the fauna is very fragmented. Level VI: dense, one–and-a-half metre thick unit whose formation was affected by the collapse of the canopy and subsequent breaking of the blocks, causing various dispositional situations. Its considerable thickness can be separated internally: from its base to –320, characterised by the significance of major fracturing and the compactness of the soil; from –320 to –270, where the soil becomes lighter and takes on a looser texture; from –270 to the top, with more compact, gritty and organic sediments. This separation is coupled with the individualisation of three cultural units: the lower from the laminar Mesolithic, where back edge tools are the most abundant retouched objects; the intermediary Sauveterrian style, where the evolution of back edge arrowheads and the introduction of geometric microliths and splintered pieces is noticeable; the top, which offers a significant industrial change as its industry rests on lithic flake and denticulate tools made from irregular materials (racloirs, perforators or other prototypes); tools that define the Mesolithic of notched and denticulate tools. Level V with a thickness of 15-20 cm, with a loam structure, brown-red colour and constant presence of white and black (carbon) specks. There is a significant proportion of land molluscs and numerous hearths. Culturally, it adheres to the Mesolithic of notched and denticulate tools, offering a morphotechnological evolution compared to the previous episode. Level IV: its thickness ranges from 20 to 25 cm, with clarified soil, a noticeable fall in Helix cepaea nemoralis and elements with higher fractioning. They are more compact and humid sediments. Culturally, its retouched industry adheres to geometric Mesolithic, with more trapezoids than triangles, a presence of mircoburils, back edge blades Área de Prehistoria Universidad del País Vasco (EHU/UPV). Tomás y Valiente s/n 01006 Vitoria-Gasteiz. a.alday@ehu.es
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  148  Alfo...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES and a good selection of substrate elements (with a certain inheritance from lithic flake tools). Level III: of continuous formation but where several units can be distinguished. The lower (IIIb2) of 20 cm of fine black loam, where helix colonies are represented that give a gritty feel to the soil; the intermediary (IIIb1), of 20 cm, with a horizontal disposition, with somewhat lighter loam soil and less granularity; the top, which is 20 cm thick, with finer soil and dark grey loam. This level stands out for being unctuous and greasy. Its base responds to the characteristics of geometric Mesolithic, now with a higher presence of triangles compared to trapezoids. The other two sections correspond to early Neolithic; segments Figure 1. Stratigraphy of the site at Atxoste in double bevel among the retouched lithic industry being very significant; Boquique style patterns, among others, on pottery; the discovery of a large windmill for grinding vegetable matter; the profile of sickles and the identification of wild and domestic fauna. Level II: of varying thickness (between 15 and 35 cm), it presents a mixture of soils and archaeological products from the burial phase (level I) and from the last period it was occupied as shelter (level II). The soil is loose and loamy, brown in colour and with lithic elements from the Chalcolithic. Level I: it conserves part of the organic matter from the surface level, in dry and loose brown soil, without a clear separation with respect to the 149
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  and a good selection of substrate elements  with a certain inheritance from lithic...
150 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD next horizon. Of the burial character ascribed to the Metal Ages, where a couple of bodies were arranged in foetal position, some in partial anatomical connection and the rest disorderly. Surface level: organic matter with a grainy texture, dry and dusty in place, with an irregular presence of blocks. It has no archaeological interest (Fig. 1). Level Code GrA-9786 3360±40 I GrA-9787 3470±40 GrA-24684 3680±50 GrA-24683 4980±50 GrA-6846 4730±50 GrN-22739 4560±110 GrA-9789 6220±50 GrA-13414 6050±40 GrA-13415 6940±40 GrA-13468 7140±50 GrA-13418 7340±50 GrA-14419 6970±40 GrA-13447 7810±40 GrA-13448 8030±50 GrA-15700 8510±80 GrA-15699 8760±50 GrA-15858 9550±60 GrA-35142 9510±50 GrA-22865 11720±70 GrA-22866 11760±70 GrA-23107 11690±80 GrA-22900 11800±60 GrA-13473 8840±50 GrN-26663 9650±150 GrA-35141 9450±50 GrN-26664 9510±150 GrN-26665 9820±150 GrN-26666 11910±170 GrN-26667 11960±180 F2 GrA-19554 12070±60 G GrA-19502 12200±90 H GrA-19870 11730±80 H2 GrA-19503 12540±80 II III IIIb1 IIIb2 IV V VI VIb1 VII D E E2 F Table 1. Carbon dating 14 BP Date Cultural assessment Atxoste is an archaeological deposit that contains information almost without interruption over 10000 years of prehistory; it has a rich inventory of material, flint, pottery and bone, and abundant recordings of fauna and carbon (Fig 2). In the Late Glacial Maximum, Palaeolithic populations overflow their traditional environments of shelter to occupy more interior areas. In the Iberian Peninsula the phenomenon results in new settlements in, for example, the Ebro basin: in this context Atxoste reflects the settlement of populations in the upper reaches. The choice of the place and repeat visits demonstrate the interest of the communities in exploiting an environment where valley and mountain ecosystems coexist. This results in the wide spectrum of fauna recovered during excavation: stag, deer, horse, boar, goat, chamois, wolf, plus smaller fauna and, anecdotally, turtle. However, the territory does not have flint, the material used almost exclusively for making stone tools. It is collected in the outcrops of Urbasa (30km north east), Treviño and Loza (both at 30km south west), from the Cantabrian Flysch (at least 100 km) and in Neolithic times, evaporite was collected from the Ebro (around 100 km south east). The percentages vary from one variety and another according to the characteristics of each episode, reflecting, in any case, the will to exploit the region further. The cultural stages represented refer us to various stages of the late hunter-gatherer. The Traceology define hunting practices, butchering, working hides, wood, bone… Over time, the presence of groups settling in the site stabilised; they extended their activities and in the late stage, they built a cabin next to the wall of the shelter as another sign of their geographic settling. This model lives on in the early stages of the Neolithic, with certain innovations: renewal of lithic tools, the introduction of pottery, crop and livestock domestication (according to the identification of direct taxa or from indirect evidence associated to these practices).
150  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  next...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Figure 2. Archaeological materials at the site in Atxoste 151
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Figure 2. Archaeological materials at the site in Atxoste  151
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 152 Javier Mangado*, José Miguel Tejero*, Josep Maria Fullola*, Maria Àngels Petit*, Marta Sánchez *. Cova del Parco (Alòs de Balaguer, La Noguera, Lleida). The Magdalenian sequence Cova del Parco is located in the pre-Pyrenees at Lleida, in the village of Alòs de Balaguer, in the Noguera region (coord. UTM 31 T – X: 329322; Y: 4642202). The archaeological site is placed 420 m above sea level and 120m above the Segre River, with a north-south orientation. A single 10.5 long by 4.5m wide gallery of triangular shape at the entrance forms the cavity, laterally communicates with a large shelter of 5.5 m by 30 m, is enclosed by a masonry wall of historic period. The discovery of the archaeological site dates back to the middle 1970´s when the first excavations were carried out by professor Joan Maluquer de Motes in 1974, 1981 and 1984, which included almost the complete digging of the upper stratigraphic sequence containing ceramic levels. Ten years after the discovery, in 1984, a 3 m2 trench allowed Dr. Maluquer de Motes to establish a stratigraphic sequence in six strata, finding in the deepest one a set of lithic materials of the Final Upper Palaeolithic (Maluquer de Motes, 1983-1984, 1985; Fullola et al., 1988). Dr. Fullola restarted excavation activities in 1987, starting excavation campaigns which are still annually conducted and headed by several investigators from the Seminari d’Estudis I Recerques Prehistòriques at the University of Barcelona. Figure 1. Location and archaeological site plan. * SERP (Seminari d’Estudis i Recerques Prehistòriques) de la Universidad de Barcelona, área de Prehistoria, Departamento de Prehistoria, Historia Antigua y Arqueología, Facultad de Geografía e Historia, Universidad de Barcelona; calle Montalegre, 6, E-08001, Barcelona.
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  152  Javi...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES respond to rock falls –levels XV and XIV– being sterile from an archaeological point of view. From level XIII to VI low intensity runoff processes, with a high human-induced activity, are produced. Climatically, the levels between XI and VII present a wet environment, except from level IX where a colder pulse is detected. From a palynological point of view, the presence of oak and fern spores in level XII demonstrate warmer and wetter conditions, as in levels VIII and VII where the presence of hazel and willow are recorded for the first time and the presence of oak remains. However, levels X and IX are cold and dry, highlighting the presence of conifers: Pinus and Juniperus, always in low percentages (Fullola et al., 1997). Figure 2. West sequence stratigraphy of the Dr. Maluquertrench analysed by Bergadà (M.M. Bergadà, 1998). 1. Sedimentary Sequence and Palaeoenvironmental Evolution The detailed analysis of the sedimentary sequence and the establishment of the palaeoenvironmental evolution from the “West stratigraphic sequence” led by Dr. Maluquer trench, were possible thanks to the studies of the PhD of M.M. Bergadà in which a sedimentary sequence in fifteen levels was established, one of the first scientific contributions from Cova del Parco to the Upper Palaeolithic in Catalonia (Bergadà, 1998:46-51, Bergadà et al., 1999). In the analysis of the sedimentary sequence two processes were observed: sedimentary as –runoffs and rock falls – and post-depositional –humidity and biological activity-. The lower levels cor- Sedimentation in levels VI to I presents high intensity runoffs and rock falls due to the break up processes of the wall and roof cavity. Also, from level VI to IV sedimentary crust formations can be observed pointing at arid conditions and colder pulses than the former ones. In levels III and II, especially in the first, a gravel deposit, limestone blocks, and conglomerates in cracked states were observed, related to periglacial conditions, coming up from the cavity´s wall and roof breaking up. In level II an increase in humidity was recorded. In level II contact, large limestone blocks and fallen conglomerates can be seen as a consequence of climatic process. Finally, sedimentation of notable strength in level I is due to different intensity runoff processes in the water flow. In the higher part of level I, clastic evidences are located: small rock falls and wall break up. The environmental conditions would be semi-arid with humidity pulses and cold temperatures. In short, locates in levels XI and VII is a wet and warm environment with a cold pulse. From level VI, a semi-arid ambience is detected –storm rainfall regime– with cold pulses. Later, in level III, cold conditions still rule, however in level II humidity increases. In level I, an increase in storm rainfall eroding the slopes is produced –high sedimentation rates– in semi-arid conditions, with humidity and cold episodes (Bergadà, 1998: 79-80). 2. Cova del Parco archaeological Sequence Archaeological materials at Cova del Parco, found during the excavations of Dr. Maluquer de Motes and housed in the Montsec Museum at Ar- 153
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  respond to rock falls    levels XV and XIV    being sterile from an archaeological...
154 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD was the first time that the presence of a “classic” Epipaleolithic sequence in western Catalonia defined by Fortea was recorded with some microbladelets levels to which geometrical levels with triangles and segments were overlapped, together with many microburins. The excavation allowed the establishment of its corresponding cultural and chronological sequence. Under a very ancient Epipaleolithic geometrical horizon –level Ia2, into depths of about –175/– 200 cm– dated from charcoals coming up from two combustion structures –henceforth EC, Estructuras de Combustion in Spanish–: EC11 and EC12 in 10930±100BP (GifA 95562) and 10770±110BP (GifA95563) respectively, and with an industry in which micro-bladelets and geometric elements of Sauveterrian type are documented –triangles and segments– (Fullola et al., 1998), a micro-laminar Epipaleolithic moment had appeared –levels Ib and Ic, into depths of about –200/–230 cm.–; this was dated with charcoal from the EC15a in 11430±60BP (OxA 8656) (Fullola et al., 2004). Figure 3. Some of the Magdalenian hearths at Cova del Parco. tesa de Segre, show the different Neolithic settlements from Cardial to Recent Neolithic, as well as the III millennium Bell Beaker Culture and the Early Bronze Age (Petit, 1996). The excavations carried out by our team from 1987 have allowed the recording of a cultural sequence presenting three stages. Firstly, in the Neolithic levels remain, almost inexistent, the basal part of a storage structure – silo– was recorded, dated with charcoal remains in 6120±90BP (GrN-20058); the ceramic content placed the abandonment in the Epicardial Neolithic, a moment in which it was reused as a landfill with plenty of manure and ashes (Petit, 1996). Secondly, the excavation and the register of Epipaleolithic levels in the archaeological site were developed –from 1993 to 1999-, the presence of which had formerly gone unnoticed. This Finally, under this microlaminar Epipaleolithic and after an abandonment period, we document a very precise stratigraphic moment –a depth between 230/-240 cm– dated in 12605±60 BP (OxA 10796), corresponding to the last Final Upper Magdalenian settlement –level II–; it was separate from the rest of the Magdalenian settlements sequence due to a huge rock fall –about –240/–260 cm.–. After this moment, the Magdalenian settlements sequence, still being excavated today, was developed. 3. Radiocarbon dating of the Magdalenian levels at Cova del Parco The sedimentary sequence analyzed by M. M. Bergadà (1998) was dated from charcoal samples removed from the same section she described. During these years, new dating carried out from recovered charcoals during the excavation process of different EC have allowed us to establish the chronological sequence that we present, hereunder all radiocarbon dates are uncalibrated. Thus, we have distinguished a Final Upper Magdalenian, dated in 12460±60BP (OxA10797) (z-269 cm) and 12560±130BP (OxA10835) (z-271 cm) (the date 13175±60BP corresponding to OxA10798 from charcoal recovered in the inner part of the EC19 (z-273 cm), is not considered valid) from an Upper Magdalenian, placing its beginning in
154  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  was ...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES depths between –280/–285 cm. based on the elongated scalene triangles documentation with different dating: 12995±50BP (OxA13597) (z-285 cm), 13025±50BP (OxA13596) (z-280 cm) and 13095±55BP (OxA17730) (z-293,5 cm). The progressive disappearance of these lithic elements, as well as, the appearance of different technological changes, for example, in the laminar support modules: typological –new spear point types– and functional –in the settlement dynamic, as well as the last radiometric dating obtained: 13255±50BP (OxA29336) (z-322 cm) and 13475±50BP (z-318 cm) establish the hypothesis of detecting the Middle Magdalenian settlements. 4. Cova del Parco Magdalenian settlements main features We are now presenting a synoptic summary about what archaeological excavations at Cova del Parco represent in the Magdalenian. This production will be incomplete as the field work still continues. 4.1. Intrasite dynamic We want to point out that the excavation over a 40m2 extension of the Magdalenian settlements at Cova del Parco is not casual. One of our main priorities in the moment of dealing with the Magdalenian study in this archaeological site has been to develop it from an socio-economic perspective and human behaviors. This is why we carried out an excavation in extension, which allowed us to observe the spatial relations between the multiple traces –artifacts and ecofacts– and the evidenced structures. The combustion structures – EC– deserve a special mention from us, as we consider that the majority of the productive, social, and cultural activities of those communities were carried out around them, this being why they have received special attention throughout these years. However, we must not obviate that our work faces an important problem: the trench-survey carried out in 1984 by Dr. Maluquer de Motes longitudinally divided the archaeological site by its central part, affecting the whole archaeological sequence. This is why we cannot establish with absolute certainty the stratigraphic continuity among the activities developed in each area in which the site has been divided, namely: on the right side, the space properly defined as a cave, and on the left side, the outer field considered a shelter. Furthermore, the trench suffered the irremediable loss of the contextual information attached to the archaeological materials recovered during their excavation, finally forming only a material collection. The Magdalenian settlements at Cova del Parco are characterized by great complexity, highlighted by different elements. Firstly, by the number and diversity of the discovered and excavated structures; we also have the ECs, flat and not-delimited or delimited by a stone crown, or a pave in a bucket, simple or double; and we have the “Knapping Remains Deposits” “KRD”, or “DRT“ in Spanish (Depósito de Restos de Talla)– defined as carving remains accumulation, in a very small surface coming out from a concrete technical process (Mangado et al., 2009, 2010). Secondly, the complexity has been proved from several activities recorded thanks to the typological and functional lithic tool diversity (Calvo, 2004) and over hard animal materials (Tejero, 2005). Both the production and repairing of lithic tools (Langlais, 2004, 2010), osseous tools (Tejero and Fullola, 2008), and the leatherwork in different stages of the operational chain (Calvo, 2004) attest to this. Regarding the Final Upper Magdalenian, the spatial distribution studies of the traces, as well as the analysis of the combustion structures main features –typology, micro-stratigraphy and content– show us an important and multifunctional settlement of the cave´s central area where, together with the hearths´ culinary functionality –attached to many faunal burned remains– other activities arise, mainly of the working type: lithic works/flintknapping, bone, leather... In this way, those areas closer to the walls show their functional marginality and are mainly reserved as buildup of waste areas, as faunal remains of little or void nutritional value and lithic traces which were rejected for manufacturing are recovered in them (Mangado et al., 2006-2007). The outer area, or shelter area, is also characterized for this marginal behavior in which working activities were hardly developed, such as the possible smoking of skin or food (Bergadà, 1998: 77-79). This behavior, so different in the use of spaces regarding the activities recorded in them, will be modified as we break into the Upper Magdalenian. Thus, the EC attached to this moment increase in presence and reuse in the outer area. The shelter´s 155
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  depths between    280    285 cm. based on the elongated scalene triangles document...
156 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD EC shows, at the same time, a higher typological and functional variability. In this way, together with the activities of purely signaling and lighting, lithic carving work activities attached to those structures emerge, which have allowed even technological refitting among different lithic elements. The other structures showing the settlement complexity carried out by the Magdalenian hunter-gatherers are the already mentioned –KRD– which allow us to rebuild the mobility intrasite (Mangado et al., 2006-2007). 4.2. Resource Management Throughout the Magdalenian at Cova del Parco The studies carried out throughout these years have allowed not only a certain intrasite perception towards the Magdalenian occupants working and life space system at Cova del Parco, but also a certain perception towards the outer area extrasite, beyond the archaeological site, to know the territory´s management and its resources by these communities. Accordingly, we have also observed some differences between the Final Upper Magdalenian and the Upper Magdalenian. 4.2.1. Abiotic Resources Along the analyzed period, we have documented the presence of concrete flint types being the object of detailed petro-archaeological features which progressively reveal the way a decrease in the flint types used is variably produced as these further varieties are not represented during the Final Upper Magdalenian, the result of which we may consider a certain regionalization process of the exploitation of the resource. In this way, for the most ancient stages of the studied sequence, so far –Upper Magdalenian-, we have proved little presence of materials coming from long and/or very long distances, exceeding widely the regional state displacement; these materials were introduced in the archaeological site in both engraved bladed supports and configured cores. This discovery, which forms part of one of our ongoing PhD – MS-, and which will shortly be launched, brings to light a behavior of siliceous materials supply over the long-distance axis enclosing both the Pyrenees slopes and some neighboring territories, not strictly Pyrenean, a circumstance showing us a wide mobility of these Upper Magdalenian groups. This mobility will gradually be reduced along the Final Upper Magdalenian, during which the recorded materials are of a regional support, adopting a local resource regime during the last hunter-gatherers settlements from very Early Holocene. The C.O.L for the tool manufacture attached to these siliceous resources, also reveals along the studied sequence a progressive adaptive behavior both to the metric features and to the raw materials knapping used in tool manufacturing. Thus, during the Upper Magdalenian the bladed module, presenting both blades and bladelets produced in the archaeological site from the core reduction sequence of high-quality raw materials, mainly pyramidal and prismatic morphologies and at the same time part of the core maintenance elements –flakes and cortical flakes– for the diverse domestic tool making –endscrapers, side-scrappers, burins, becs, etc– is recovered. Progressively, we observe minor module exploitation of raw materials and therefore a larger number of bladelets rather than blades is produced on the site, the former seeming to be produced as supports or even as finished tools –some endscrapers and burins highlight this– while at the same time, the configuration and maintenance activities of the bladelets´ cores are simplified since the volume exploitation of smaller size raw materials and from poorly modified cores are usually invoked (Langlais, 2010). Consequently, the tool technology is affected by this circumstance and the local flint of a minor knapping quality is progressively used for manufacturing different lithic tool types (Mangado, 2005). Lithic tools of Magdalenian levels at Cova del Parco are widely dominated by the projectile elements throughout the exhumed sequence. Backed bladelets and backed points predominance reveal a wide typological variety in which the presence of truncated backed bladelets highlights, an element we used as a cultural marker to point out the transition between Final and Upper Magdalenian stage. The hunting set is supported by domestic tools showing wide diversity of working activities developed at the archaeological site, both referring to scraping, hammering, and leather work (Calvo, 2004) and to osseous tool production and maintenance (Tejero and Fullola, 2008) 4.2.2. Biotic resources The industry in osseous raw materials at Cova del Parco includes two large subsections. The first one is referred to as domestic and hunting equipment made of bones and deer antlers. Secondly, we have a set of objects of personal ornament,
156  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  EC s...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Figure 4. Upper Magdalenian settlements Lithic and bone industry at Cova del Parco (graphics R. Álvarez). 1,10,16: endscrapers. 2: bec. 3: backed bladelet. 4-9: scalene triangles. 11-13: burins. 14-15: bladelets cores. 17: retouched blade fragment. 18: Dentalium sp. 19: Nasarius sp. 20: Homalopoma sanguineum. 21: Theodoxus fluviatilis. 22: pendant of deer´s atrophied canine tooth. 23: needle. 24: spear point distal fragment. 25: spear point (pointe de sagaie) 157
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Figure 4. Upper Magdalenian settlements Lithic and bone industry at Cova del Parco...
158 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD which, with the exception of two pendants made of atrophied deer canine tooth, were made from different mollusks species shells. Both entities of the Cova del Parco archaeological material register form an important set from the numeric point of view and above all from a qualitative perspective, as they include not only objects but also other “technical” elements –wastes, pieces being processed, blanks–. These last ones are essential to rebuilding the technical operative sequence of the exploitation of organic material of animal origin. This feature constitutes exactly the largest osseous industry (bone and antler) contribution in Cova del Parco to the knowledge of the site´s Magdalenian settlements and, by extension, of the Iberian Magdalenian. Although the osseous industry development is fairly recently related to the lithic technology, its huge capacity has been widely demonstrated for the better knowledge of the paleoethnographic aspects of the hunter-gatherers in the Upper Palaeolithic (Averbouh, 2000). The analysis´ results show that the operative sequence of the bone and antler exploitation is driven to the production of the rod or baguette type blanks through the double longitudinal grooved procedure (Tejero 2005, Tejero and Fullola 2006, 2008, Tejero et al., 2010). In the bone´s case, the blocks to be exploited are plausibly selected among the bone remains removed from the food chain without observing specific fracture patterns of technical nature. The fauna recovered in the Magdalenian levels at Cova del Parco are mainly goat´s remains (Nadal, 1998). The exploited antlers, always from deer (Cervus elaphus), probably come up from shed antler collections if we abide by their modules of thick cortical tissue and the lack of deer presence among the fauna hunted at Parco. The bone and antler blanks are transformed by an overall scraping, respectively by needles and spear points (projectile elements). This correspondence between raw material and type of object, not limited to the Magdalenian, is related to the structural properties of every material, making them more efficient in transformation tasks –bone– or as projectile elements –antler– (Christensen and Tejero i.p). With reference to the personal ornament objects, Cova del Parco has provided a set of more than one hundred pieces, with an important presence of marine gastropods (Homalopoma sanguineum and Cyclopeneritea) and fluvial gastropods (Theodoxus fluviatilis) (Tejero 2005, Estrada 2009, Estrada et al., 2010). The shells of the different mentioned taxa –some of them keep ochre remains– were perforated probably through an indirect percussion. The use of this technique to perforate the shell of very small and relatively thin species requires very precise control of the process. This fact shows a high degree of technical expertise in the Magdalenian inhabitants of Cova del Parco. At the same time, the selection of a few varieties –difficult to work with– among the wide range of mollusks available to the Magdalenian populations at Cova del Parco shows an election guided by cultural criteria and not by a technical availability and/or efficiency. However, most parts of faunal elements recorded during Magdalenian at Cova del Parco corresponds to hunted and consumed faunal remains. Its conservation status is usually fragile, as it appears really fragmented. Despite that, cut marks identification has been possible in many occasions, showing an intensive prey exploitation of mainly Capra pyrenaica. 5. Conclusions The excavations during more than a decade of the Magdalenian levels at Cova del Parco by the SERP of the UB team has been the key for the understanding of the Final Upper Pleistocene settlements in the northeast of the Iberian Peninsula. Both for the large sedimentological and paleoenvironmental sequence and for the quantity and quality of the cultural traces and the exhumed human structures, this archaeological site is an essential reference point for the global study of the Magdalenian settlements in both the Pyrenees slopes. 6. Acknowledgments In recent years, research work has been carried out as part of programme SGR2014-108 of the Generalitat de Cataluña and programme HAR2011-26193 of the MINECO.
158  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  whic...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Rafael Mora *,**, Jorge MartínezMoreno*, Xavier Roda Gilabert *,***, Ignacio de la Torre****, Alfonso Benito-Calvo*****, Miquel Roy *,******, Sofía Samper*, Susana Vega *, Jezabel Pizarro*, Javier Plasencia* The Mousterian site of Roca dels Bous (Lleida, Pre-pyrenees) In a 1973 review, Emili Sunyer mentions an important Mousterian sequence in the first slopes of Pyrenees of Lleida. This short letter which brought us to Roca dels Bous (Mora, 1988), constituted the origin of our investigation of the human settlement located in the Eastern Prepyrennes (Fig. 1A). During these years, the field work carried out at Roca dels Bous, Tragó, Cova Gran de Santa Linya, and recently at Abric Pizarro, show this area to be key to analyse the human presence in the Upper Pleistocene and the Holocene in the Iberian Northeast. Deconstructing Palimpsests Roca dels Bous (X = 321.266, Y = 4.638.067, UTM H31 N ETRS89) is located at Cingle de la Cascalda, an Eocene limestone and Oligocene conglomerates cliff more than 40m high on the right edge of the Segre river, 275m A.S.L. (Fig. 1B). This slope´s deposit is 20m thick and has a fluvial terrace that is minted on its basis over the substrate (Fig. 1C). The excavation is focused on the upper platform (Fig. 1C), where a first level –R3– arose, dated by 14C AMS in 38.8±1.2 ky BP (AA 6481). The excavated sequence reaches 1.5m deep. In a sedimentary level, it is a sequence of little consolidated breccia of sand and shale, plenty of angular medium-small size debris, and large autochthonous blocks coming from the weather erosion of the shelter’s limestone. Up to now, 100 m2 of levels N10 and N12 have been excavated. Other levels have been * Figure 1. A) Roca dels Bous, Cova Gran de Santa Linya and Cova del’Estret de Tragó topographic location; silexand metamorphic rocks´ deposits topographic location at Noguera Prepyrenees. B) Roca dels Bous. Centre d’Estudis del Patrimoni Arqueològic de la Prehistòria (CEPAP). Universitat Autònoma de Barcelona. 08193 Bellaterra, Spain. cepap@uab.cat ** ICREA– Academia Program *** Becario Programa FPI–MINECO **** Institute of Archaeology-University College London 31-34 Gordon Square, WC1H 0PY London, United Kingdom. i.torre@ucl. ac.uk ***** Centro Nacional de Investigación sobre la Evolución Humana (CENIEH). Paseo Sierra de Atapuerca s/n. 09002 Burgos. alfonso. benito@cenieh.es ****** FI DGR– Generalitat de Catalunya Fellowship 159
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Rafael Mora  ,  , Jorge Mart  nezMoreno , Xavier Roda Gilabert  ,   , Ignacio de l...
160 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. C) Archaeological site profile. detected in several surveys, N14 and S9, which could be extended by similar surfaces. The complex site formation processes of the deposit make it difficult to follow these archaeological units’ dispersion. Two alluvial fans are located in the deposit´s ends, one on the E side which articulates the sediments´ income with a slope of 10º-15º to the southwest; the other, on the W side, is sub-horizontal –5º– sloping to the southeast. These riverbeds form a depression covering an important part of the excavated area. These carbonated surfaces homogenize the sediments´ colouring. To surmount the lack of visibility, we assume that lithic and bone remains are sedimentary particles. Their systematic coordination defines accumulations with horizontal and vertical dispersion separated by sterile, delineating surfaces with inclinations and depressions derived from the sedimentary accretion of the lateral cones (Fig. 2A). The large excavated combustion structures show the regular use of fire and confirm the archaeological geometry of these units. Similarly, they allow detection overlapping which involves the settlements´ sequence. These fusion/fission phenomena indicate that the sedimentary rhythms are not homogeneous (Fig. 2A). The archaeological levels are added to short-term events which were staggered during a term scale that is difficult to evaluate (Mora et al., 2008; Martínez-Moreno et al., 2010). Artefacts and Behaviours Contextualizing these processes is essential to analyze the variability observed in the Mousterian artefacts. Determining the origin of raw materials is a priority, and silex and quartzite outcrops feeding Roca dels Bous have been identified. They basically manage metamorphic rocks which abound in the fluvial deposits in this area. Silex rocks are not a local resource, although they appear in two regional outcrops: (1) the Garumnian formation which extends by Montclús and Tragómountains, (2) Serra Llarga (Oligocene) (Fig. 1A) (Roy et al., 2013). There is no lack of rocks in this area to produce artefacts, so the changes in the composition of raw material and how this affects lithic assemblages describes the techno-cognitive and techno-economic environment of these Neanderthal groups. This conception can be evaluated in levels N10 and N12, resulting in remarks which affect the debate about Mousterian variability causes (Mora et al., 2008). In N12, excavated along 105m² and 20cm thickness, 22 hearths and the accumulation of 90 kg of rocks shaping a set of more than 23.500 artefacts are identified, of which all the segments related to the knapping process are present. The metamorphic rocks constitute 80% of the assemblage (Fig. 1C). However, retouched and small expediently knapped flint fragments are selected. These behaviors suggest the transport of finished pieces and small blanks from which little supports are obtained from 15-20 km. Retouched quartzite pieces are large with denticulate edges while the flint ones are small and instruments shaped with continuous fronts (Fig. 2B). N10 suggests remarkable differences. This level follows along 95 m2 with 10cm thickness, where 20 hearths were excavated. A radical decrease in artefacts can be seen –about 2100 pieces– which represent the transport of 11 kg of raw material. Sixty-six percent of the instruments are flint manufactured, although from the weight, the distribution between metamorphic rocks and flint is well-balanced suggesting that the flint artefacts are small, as seen in N12. There are more flint cores than quartzite cores, although they likewise point out a managing from the expedient methods as well as centripetal recurring methods which conform volumes under 5 cm (Fig. 2B). The retouched are preferably shaped of flint –80%–, the denticulate being more frequent than the continuous-edge pieces, as well as the quartzite supports (Martínez-Moreno et al., 2010; Mora et al., 2008). The retouched pieces are usually fragmented and some remounts suggest that they are repaired
160  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Figure 2. A) N10´s excavated hearths distribution. Vertical projection, E-W –up– and N-S –right-, in which the hearths´ overlapping and fusion/fission phenomena inside the level can be discerned. In the N-S projection, the sterile between N10 and N12 can be seen. B) Cores trends on silex and metamorphic rocks extensively exploited until configuring small volumes (up), retouched tools (down). 161
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Figure 2. A  N10  s excavated hearths distribution. Vertical projection, E-W    up...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 162 instruments (de la Torre et al., 2012). Many retouched pieces suggest their recycling. The double patinas identified in some pieces could correspond to artefacts recovered on the site or in the surrounding landscape which are reactivated to obtain new supports (Mora et al., 2008). This intense management does not obey the lack of this material in the area, allowing us to approximate these groups techno-cognitive environment. Likewise, especially N10 suggests short-term activities; the archaeological site served as a stop in the movements between residential displacements. If so, Roca dels Bous represents a web of Neanderthal settlements inside this regional environment in the Prepyrenees of Lleida and Huesca (Mora et al., 2008). Future prospects These arguments, discussed in other contributions (Casanova et al., 2009; Martínez-Moreno et al., 2010; de la Torre et al., 2013), point out that these techno-typological tendencies do not re- Rafael Mora Torcal *,**, Alfonso BenitoCalvo***, Jorge Martínez-Moreno *, Ignacio de la Torre****, Susana Vega Bolivar *, Miquel Roy *,*****, Xavier Roda Gilabert*,******, Sofia Samper Carro * These behaviours related to flint management must be attached to a fragmented chaineo peratoire along a wide techno-temporal scale, converting Roca dels Bous into a privileged place in the movement of Neanderthal groups (Mora et al., 2008; de la Torre et al., 2013). The settlement´s strategic position allows an effective control of the seasonal animal movements, especially equids (horse and wild ass) and deer, between the Ebro Depression and the Pyrenees. This short-term settlement– pattern should be attached to annual cycle short periods in which the environment offered opportunities to obtain prey and transport them to the archaeological site. The inferences from Roca dels Bous, and in general the pre-Pyrenees settlements at Noguera, suggest that this area will have a prominent role in the investigation of the Middle Palaeolithic in the Northeast of the Iberian Peninsula. A key sequence in the Western Mediterranean Prehistory: Cova Gran de Santa Linya (Pre-Pyrenees in Lleida) This large rock shelter was discovered in 2002 during a survey program coordinated by the Centre d’Estudis del Patrimoni Arqueològic de la Prehistòria (CEPAP). This settlement contains a large chrono-cultural sequence covering Middle Paleolithic, Early Upper Paleolithic, Magdalenian, Neolithic and Chalcolithic. The use of this site by hunter-gatherers and farmer-shepherds turned the * spond to techno-economic factors such as the lack of raw materials in the environment. N12 notes the option of using local rocks. place into a key location for analyzing human settlement of the Pyrenees during Prehistory. Geographical situation Cova Gran (X=318541, Y=4643877, UTM H31N ETRS89) is located in the eastern PrePyrenees, in Lleida (see Fig.1A in Roca dels Bous Centre d’Estudis del Patrimoni Arqueològic de la Prehistòria (CEPAP). Universitat Autònoma de Barcelona. 08193 Bellaterra, Spain. cepap@uab.cat ** Programa ICREA- Academia. *** Centro Nacional de Investigación sobre la Evolución Humana (CENIEH). Paseo Sierra de Atapuerca s/n. 09002 Burgos, Spain. alfonso.benito@cenieh.es **** Institute of Archaeology-University College London 31-34 Gordon Square, WC1H 0PY London, United Kingdom. i.torre@ucl. ac.uk ***** Becario Programa FPI – MINECO ****** Becario Programa FI DGR- Generalitat de Catalunya
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  162  inst...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES paper), 385m A.S.L., and in a lateral valley of the Noguera-Pallaresa river where the ravine of Sant Miquel digs into the limestone of the Upper Cretaceous creating a shelter of over 2000 m2 (Fig. 1A). The chrono-stratigraphic sequence of the sedimentary and cultural processes of the deposit is made up from the sectors Ramp, Transition and Platform which are correlated from 40 14C AMS and T1 dates (Fig. 1B). We do not reject the appearance of new chrono-cultural segments. Cova Gran is relevant for contextualizing last 50.000 years of the human presence in the Western Mediterranean (Mora et al., 2011). The longitudinal profile of the shelter permits appreciation of the deposit shaped from two large juxtaposed platforms (Fig. 1B). The sequence of the first one –point west and external to the rockshelter– is defined in the Ramp sector. Transition and Platform sectors record the sedimentary development under the visor. The Ramp Sector: Archaeo-Stratigraphy of the Outer Platform The Ramp sector –R– is a 200 m2 platform with a 20º east gradient. The dug area extends 120 m² and some levels go on outwards. In this area of 2.5 m thickness, the sedimentary units S1 and 497 are differentiated, originated by different climatic processes (Fig. 1C). The basal unit S1 is a set of 2m made by medium and coarse gravels, sand-clayey matrix and limy-angular debris of a gravitational origin which indicate cold conditions. Unit 497 –0,5m– is composed of granular sediments affected by water flow indicating relatively milder environmental conditions. Relevant sin-post depositional processes are not detected (Benito Calvo et al., 2011). indicates several implications. In these stratigraphycally overlayed levels, recovered lithic assemblages correspond to different cultural traditions, with changes affecting raw material, knapping systems, blanks and the retouched tools (Fig. 2A). This technological change redirects to the debate about the Middle/Upper Paleolithic transition, and the possible extinction of H. neanderthalensis in parallel to the dispersion of H. sapiens in Western Europe. S1B summarizes the general trends of the Mousterian in Cova Gran (Martínez-Moreno et al., 2011; Mora et al., 2012). Cores are reduced to obtain centripetal flakes similar to those observed in Tragó and Roca dels Bous (de la Torre et al., 2013). The flint coming from the Garumnian formations adjacent to the site is the most used raw material in all levels. Similarly, metamorphic rocks are transported configuring large pseudoLevallois blanks elaborated outside the site –Fig. 2B-. In other words, whole materials are obtained in the immediate area (see Fig. 1A article about Roca dels Bous). Regarding the retouched tools, denticulates are more common than pieces with continuous edges (Mora et al., 2012). 497D shows important differences. Metamorphic rocks disappear and although the Garumnian flint is the main resource, the contribution of configured supports from Serra Llarga –20 km far away– increases. The knapping system is intended to obtain blade-elongated supports with a low morpho-technical standardization degree. This includes end-scrapers and burins on blade, backed and points on bladelet, artefacts unknown in the In sector R, 8 archaeological levels are stratified punctuated for being sterile. Unit S1 contains levels S1E, S1D, S1C, S1B1 andS1B which correspond to Mousterian, and 497D assigned to an undetermined Early Upper Paleolithic. Sedimentary unit 497, levels 497C and 497A, are attached to other stages of Early Upper Paleolithic cycle. All levels are rich in lithics, bones and hearths. Marine ornaments, especially Nassarius incrassatus (Martínez-Moreno et al., 2011) have been recovered in 497D, 497C and 497A. These artifacts are considered markers of the irruption of H. sapiens in Western Europe. The techno-typological change detected between S1B and 497D –on the roof of the unit S1– Figure 1. A) Cova Gran de Santa Linya. 163
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  paper , 385m A.S.L., and in a lateral valley of the Noguera-Pallaresa river where ...
164 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 2. B) Archaeo-stratigraphic sequence: A) shelter´s floor where the dug sectors are located –Platform, Transition and Ramp– B) Longitudinal transection of the deposit. C) Sectors R, T and P chrono-stratigraphy (see Mora et al., 2011)
164  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Mousterian. Although, an important component of scrapers, denticulates and notches on flake persist –Fig. 2B–. These features do not match with the trends described in the first techno-complexes of the Upper Paleolithic in Western Europe (Martínez-Moreno et al., 2010). Some anomalies are appreciated in the dating of this sector, which can be related to the protocols used in C14 laboratories. However, the dates provided for 497D from charcoal coming from a hearth, place this level in the interval 40-38.5kacalBP. The archaeo-stratigraphic resolution and the chronometrical intervals of these levels contribute to the debate about of Middle/Upper Paleolithic transition (Martínez-Moreno et al., 2010). Archaeo-stratigraphy of the inner platform: Transition and Platform sectors The central platform is a surface extended under the visor of the shelter of about 2000 m2. This delineation restricts the most ancient archaeological levels to sector R –Fig. 1B–. This sequence is known by the Transition sector (T) –a survey of 2x2 m–, and the Platform sector (P), which embraces a dug area of 32 m2. Two units are identified in both sectors: unit N corresponds to the Holocene, unit P is attached to the final of MIS 2 (–Fig. 1B–.) The most ancient human presence is detected during the Last Maximum Glacial (LMG) in the sedimentary unit P of the Platform sector, made by very angular debris and falls from large blocks with limited fine sediments –Fig.1B–. Levels 4P, 5P, 6P and 8P take place in this sequence, with 1.7 m thick, dated between 20.4-18 ky calBP –stage Gs2b–. Points and backed bladelets, burin and endscrapers on blade are associated with massive antler projectiles, needles and perforated gastropods. These artefacts can refer to the Early Magdalenian. Sector T is a survey of 3.5 m depth in which several levels with different technical features from sector P take place –Fig.1B–. A dating on the survey´s basis (17-16.8 ky calBP) allow sector T to be attributed to Middle Magdalenian and possibly Upper Magdalenian levels (Mora et al., 2012). The Holocene sequence N erodes levels P in sectors T and P, creating complex geometries over which farmer-shepherds communities settled from the Early Neolithic, just as it is identified in the Platform sector –Fig. 1B–. The most intense settlement moment occurs during the Late Neolithic –5500-5100 calBP, recording 30 domestic structures –hearths, post-holes and pits– (Mora et al., 2012). Above these settlements, this area is used for stabling sheep-goat during the Late Neolithic (5000-4600 calBP), Calcolithic and Late Bronze Age (3950-3000 calBP). These accumulations, each 0.5 m thick, include several stages of intentional burning in order to condition the place for future visits. These appreciations open new perspectives to analyze the origins of pastoralism in northeast Iberia (Polo et al., 2014). 165
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Mousterian. Although, an important component of scrapers, denticulates and notches...
166 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 2. A) N-S and E-W projections of 497D –triangles– and S1B –points– separated by sterile. B) Lithic artifacts of S1B Middle Paleolithic –on top– and 497D initial Upper Paleolithic –below–. Future prospects Cova Gran de Santa Linya articulates several important research questions in the current scientific debate: the disappearance of the Neanderthals and the appearance of modern humans, hunter-gatherer adaptations during the LMG and the emergence of the first farmers. These issues are essential to analyze human presence in the southern Pyrenees, an area in which important progress is taking place. We think that this is a privileged place to analyze the course of different human groups which occupied this shelter for 50.000 years.
166  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Joel Casanova i Martí *, Rafael Mora Torcal *,**, Xavier Roda Gilabert *,***, Jorge Martínez-Moreno *, Miquel Roy *,****, Susana Vega * The Middle Paleolithic sequence of Cova de l’Estret de Tragó (Lleida, Pre-Pyrenees) Cova de l’Estret de Tragó was discovered during surveys in the Noguera Ribagorçana river, conducted in 1990 by the IEI-Diputació de Lleida (see Fig. 1A in the article Roca dels Bous). This rockshelterhas a surface of 14 x 10 m and is part of karst limestone Fm. Bona (X = 301856 Y = 4644190 UTMH31N ETRS89) to 390 m (Fig. 1A). This position in the first Prepyrenees of Lleida, plays a key association between the Ebro Basin and the Pyrenees (Martínez-Moreno et al., 2004; Casanova et al., 2009). Geographical Context and Chronostratigraphic Eight archaeological units from Middle Paleolithic are identified in Tragó. They are interbedded with sterile sediments allowing individualized analysis. The archeostratigraphy of the deposit is Figure 1. A) Cova del Estret de Tragó currently flooded by the Santa Anna marsh. B) Sequences of the inner rock shelter (78-23 survey) and the excavated area on the deposit platform. The Upper Unit (UU), Middle Unit (MU) and Lower Unit (LU) are positioned. C) Thermoluminescence series sequence. * ** *** **** Centre d’Estudis del Patrimoni Arqueològic de la Prehistoria. Facultat de Lletres. Universitat Autònoma de Barcelona. 08193 Bellaterra. cepap@uab.cat Programa ICREA-Academia Becario programa FPI-MINEC Becario Programa FI-DGR. Generaltitat de Catalunya 167
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Joel Casanova i Mart    , Rafael Mora Torcal  ,  , Xavier Roda Gilabert  ,   , Jor...
168 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD established from the sequences obtained in the survey on the square78-23 and the excavated area (25 m2) and it can be defined in three units (Martínez-Moreno et al., 2004; Casanova et al., 2009) (Fig. 1B): • Lower unit (LU): high energy of fluvial deposit from Ribagorçana Noguerariver; containing medium and large size cobbles within a carbonated sandy matrix. This terrace constitutes the base of the site. • Middle unit (MU): low to medium energy environment composed of clayey sand and silt resulting from the alternation of flooding, causing vertical migration of carbonates that precipitate on archaeological material. At this level of 1.5 m thickness, archaeological units UA3, UA1 and UA2 are excavated. • Upper unit (UU): residual breccia fixed to the shelter wall dismantled in the rest of the deposit. In this survey, 1 m2 and 1.10 m deep, S4, S5, S6 and S7 levels follow. It has not been possible to correlate the archaeological levels of the Middle unit with the Upper unit (Fig. 1C). Eleven Thermoluminescence (TL) dates are available, generating chronometric ranges that frame the occupation of the settlement. The S5, S6 and S7 from the Upper unit levels are assigned to MIS 3 (between 43 ± 4.6 and 52.1 ± 6.7 ky). The date of UA1 (41.7 ± 4.5 ky) suggests this Middle unit level may be related to the Upper unit. The four dates obtained from UA2 are staggered in the range 75.3 ± 7.8 –110 ± 12 ky, indicating this level of 60 cm thickness is formed on the MIS 5. The UA3 has two dates and we accept the corresponding to MIS 5e stage (126 ± 15 ky). This series makes Tragó a key site to analyze the Upper Pleistocene Mousterian settlement in the northeast of the Iberian Peninsula (Casanova et al., 2009). Archaeological record Levels excavated can be considered low resolution palimpsests with thousands of lithic and bones imbricated without apparent order (Table 1). These accumulations would be the result of repeated visits to the shelter at different time periods, interrupted by abandonment phases. There are not identified hearths, but regular fire use in the entire sequence can be recognized by the presence of abundant charcoal, and burned bone and artefacts. Bone carbonation and difficult anatomic and/ or specific bone identification are not able to calculate indices of their presence. The species identified are Vulpes vulpes; Sus scropha, Cervus elaphus, Capreolus capreolus, Bos sp., Capra pyrenaica, Equus caballus and Equus cf. hydruntinus. This eurythermal association describes an environment that integrates meadows with wooded areas insertedin alow/medium mountainous but sharp landscape. Abundant helical fractures on diaphysis denote intense marrow recovery (MartínezMoreno et al., 2004). The lithic assemblage, composed by over 20,000 artifacts, is essential to characterize the Upper Unit Middle Unit S4 S5 S6 S7 UA1 UA2 UA3 Surface (m2) 0,5 1 1 1 23 35 13 Thickness (cm) 20 30 15 15 10 60 15 Hammers 0 4 1 1 3 75 38 Cores 1 10 14 11 15 423 229 Flakes 33 315 184 152 496 6564 2303 Flakes frag. 70 406 502 328 777 7895 2275 Chunks 15 76 185 88 73 947 452 Retouched 16 163 140 41 143 791 274 Table 1. The excavated surface and the average thickness of the archaeological levels from the Upper and Middle Unit with the number of artifacts recovered.
168  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  esta...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Mousterian in this area. Metamorphic and siliceous rocks proliferate in the environment (see Fig. 1A in the article Roca dels Bous). Flint, present in the Montes de Tragó and Serra Larga (10 km), refers to local sourcing. At diachronic levels, the metamorphic rocks are the most important in lower units UA3 and UA2. This trend is reversed in UA1 and S4, S5, S6, S7 where the flint is the majority (Casanova et al., 2009; De la Torre et al., 2013). Figure. 2 A) Macrotools from UA3 related to percussion activities. B) Variability of organized knapping systems showing the configuration of small volumes. C) Expedient cores. 169
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Mousterian in this area. Metamorphic and siliceous rocks proliferate in the enviro...
170 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD All categories related to knapping, including microdebitage, are represented. The debitage is structured from organized expedient technical systems (Fig. 2). Organized systems allow management from schemes referred to as Levallois and discoid methods. The expedient methods are applied to volumes obtaining few blanks. These strategies, present in each sequence, denote the application of complementary technical behaviors depending on specific needs (Casanova et al., 2014). The cores usually measure less than five cm and intensive management of consumption it is observed (Fig. 2).This behavior cannot be explained by the lack of raw materials in the environment, we consider it a technical choice focused on elaborating small artifacts (Casanova et al., 2009).The most common blanks are flakes, points and blades are rare. Blanks were obtained with hard hammers, generating accidents like broken pieces, Siret burins and double bulbs. This continuity in the knapping methods derived several reflections. In the Middle Unit levels the expedient methods are mainly against structured methods, a pattern that persisted in the Upper Unit. This notion of technological stasis denotes a cognitive arrangement in the transmission of technical knowledge that could imply that in this area a stable cultural tradition was developed (Casanova et al., 2009). In UA3 25 cobbles of metamorphic rocks and granite were knapped by façonnage for shaping macrotools artefacts (Fig. 2).In these pieces, modifications over the edges can be seen, relating to percussion activities (Casanova et al., 2014). In the sequence, the retouched percentage is low (Table 1), selecting the flint to retouch pieces; although in UA3 and UA2 metamorphic rocks are more abundant. The most common blanks are short flakes retouched on a single edge (lateral or transverse); while double retouch edges are scarce. Denticulate and notched pieces with simple or abrupt retouch are more numerous than sidescrapers with continuous retouch. Preponderance of denticulate pieces is constant along the sequence. Although pseudo-Levallois retouched el- ements are identified, most of them are made on regular flakes or fragments. Tragó in the Middle Paleolithic Context of Northeast Iberia The repeated use of this area during the Upper Pleistocene between MIS 5e, MIS 5 and MIS 3 should link with the control available from the settlement on the strait of Noguera-Ribagorçana river and floodplains currently flooded by the Santa Ana reservoir. This strategic point would not be unnoticed by the Neanderthal population. The rock shelter centralizes prey acquisition and their passing through the corridor allowing ambush in the wooded areas around the river. Similarly, displacement to the rocky outcrops adjacent to the settlement can be identified. Even though the radiometric record is inaccurate, it cannot confirm their occupation in the MIS 4. This gap could be related to climatic crisis causing the abandonment of this environment. The cyclic occupation/abandonment of the area as a result of environmental factors should be retained as a possibility. The technical continuity from the combination of technical methods, expeditious and organized, present throughout the entire sequence is relevant. We stress the importance of technological stasis notion identified in other sites of the Pyrenean foothills. Likewise, we warn that these technical options articulate a cultural tradition extended into the Upper Pleistocene northeastern Iberia. A number of attributes of this entity are the panoply of knapping methods, orientation to obtain small blank, and the denticulate preferred configuration (Casanova et al., 2009; de la Torre et al., 2013). Under this perspective, Cova Estret of Tragó is a relevant settlement to investigate the Neanderthal lifestyle in the Iberian Peninsula. Acknowledgments We dedicate this article to Joel Casanova i Martí. These lines are a demonstration of our respect, affection, and admiration.
170  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  All ...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Pilar Utrilla Miranda*, Vicente Baldellou Martinez**, Manuel Bea Martinez*, Lourdes Montes Ramirez*, Ramón Viñas Vallverdú*** The cave of Fuente del Trucho (Asque – Colungo, Huesca) 1. Location The cave of Fuente del Trucho is located in the Arpan ravine on the left side of the river Vero next to the spring that gives it its name. The Paleolithic art of the mentioned cave and the Levantine and schematic art of the Arpán shelter, 870 m long –an exceptional case in the Iberian Peninsula– are located in the same ravine. The river Vero runs among canyons throughout 10 km, in this course there are only two natural entries, crossing the vertical walls, to access to the riverbed: Villacantal Bridge and Tozal de Mallata. There are Prehistoric paintings in both of them. A shallow cave, with a large 22 m wide entrance facing the southeast, gives access to a wide room, 24m deep, divided into two dissymmetrical lobes. The smallest one, at the left, has a spherical dome, blackened by organic matter and haze, and an oval window allowing sunlight to enter with zenithal lighting. The floor falls into an oblique calcite flow, several deeply drawn engravings made on it receive direct sunlight at dawn. The second lobe, in semi-gloom, presents its walls and roof covered with paintings, most of them red. The floor today presents naked rock in a great part of the room, with the exception of the right side of the cave, under the tri-lobed signs, where a messy deposit presents lithic materials which correspond to the Upper Paleolithic and the Mousterian. Towards the exterior, moving out from the painted area, the levels present a larger sequence, although in this case they seem to only be attached to the Mousterian. 2. Investigation History The first explorers of Sierra de Guara ravines, particularly Pierre Minvielle, illustrated the archaeological deposit contained in the cave. However, the cave paintings were discovered in 1978 by a team of the Museum of Huesca and the University of Zaragoza headed by Vicente Baldellou. He entrusted the ex* ** *** Universidad de Zaragoza utrilla@unizar.es Museo de Huesca IPHES cavation of the Mousterian levels of the outside area to Anna Mir, who worked in five campaigns from 1979 –Mir, 1987– and the areas corresponding to the Upper Paleolithic were entrusted to Pilar Utrilla, who in 1980 did the first survey of the bottom of the external engravings disrupted by the presence of a very hard crust. In 2005 she restarted the excavation, in cooperation with Lourdes Montes as codirector, proceeding into the right inner area of the cave, at the bottom of the tri-lobed signs (Montes et al., 2006; Utrilla et al., 2010). A new campaign is expected to be started in the summer of 2014. V. Baldellou, director of the cave art surveys in the river Vero for many years, coordinated from the begining the different procedures at Fuente del Trucho, taking care of the study of the parietal exhibitions, first in cooperation with A. Beltrán, in the progress of the Altamira Symposium (Beltrán and Baldellou, 1981; Beltrán, 1993) and some years later with R. Viñas, elaborating direct calques between 1989 and 1991. In the year 2000, S. Ripoll and F.J. Muñoz made photo-documentation of the roof (Ripoll et al., 2001), and since the year 2011, V. Baldellou, M. Bea and P. Utrilla have carried out digital treatment and calques assembling within the project HAR 2011-27197 “Reconsidering ancient archaeological sites. Expanding into new horizons in the Middle Valley Prehistory of river Ebro”. For this purpose, all the old photos which systematically covered every cave´s wall have been digitalized, treated with the D Stretch application for Image J and Photoshop CS5 and collated with the original calques of Baldellou and Viñas. In the Gravettian convention celebrated in Altamira, the two main sets of the cave were published: the roof –panel XV– and the frieze –panel VI and VII– (Utrilla et al., 2013). In this convention –session A11a- the dating results made by U/Th about the cleanest crusts covering some figures –hands, points, horse and tri-lobed– were presented (Hoffman et al., 171
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Pilar Utrilla Miranda , Vicente Baldellou Martinez  , Manuel Bea Martinez , Lourde...
172 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1: Roof floor –left, indicating the painted panels- and ground floor –indicating archaeological pits and the engravings´ location–. Topography J. Angás – Scanner 3D. Patrimony and Industry i.p.), while the Candamo convention focused on the relationship between the archaeological site and the paintings ( Utrilla et al., 2014). The final study is pending on the projected haze cleaning of the dust mucking up the wall now, a splendid white support which would highlight the hundreds of paintings, most of them red, which cover the surface of the roof and the walls for decorative purposes. There are two welldefined areas in Fuente del Trucho: the paintings in a semi-gloom at the large right cave (Fig. 1.1); and the external engravings placed on the slanting floor of the small left lobe (Fig. 1.2). 3. The paintings The provisional paintings stocklist (Ripoll et al., 2001) records 22 panels with a hundred figures. Series of points, tri-lobed signs, hands and horses are the four main painted items, to which a small goat, a deer, an undefined animal –a bear, a bovid, or a horse- and several signs are added. The points series are represented in 8 panels. They appear in horizontal lines shaping a 6 m long frieze in panel VI (Fig. 2), or forming part of more complex motifs represented in both of the walls –panel VII– and the roof –panel XV– (Fig. 3). Panel XI contains red couplet lines, also a feature of ancient Gravettian panels, usually attached to hands and fingering (González Sáinz, 2003). Regarding to the complex series of interpretations on the roof, in 1993 Beltrán proposed that they probably represent the sky dome, constellations of stars maintaining rhythms. In another vein, in 2005 Utrilla suggested that the routes of the Pyrenean ports dividing Gargas and Fuente del Trucho could have surprised the travelers, maybe losing their way or making detours, which could be represented by the complexity of some motifs. In regards to the lineal series of the vertical frieze, in bands of 4 and 5 lines, the representation of the same motif in the Levantine art at the closer cave of Arpán draws attention. Their Paleolithic parallels are found in many caves of the Cantabrian Coast, the
172  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Figure 2. Frieze´s calque –panels VI and VII–. Below, detail of the listing horses, the deer, hands and points where the dating samples by U/Th were taken. most part of them in an ancient chronology (Candamo, Llonín, Pindal, Mazaculos II, La Meaza, Chufin, Porquerizo, Castillo, la Pasiega, La Garma, Castillo or Cullalvera), the French Pyrenees (Niaux, Marsoulas, Trois Frères, Bedeilhac) or the South of the Peninsula, in this case attached to horses (Las Palomas, Atlanterra, El Moro). The date U/Th obtained for the crust covering the series of points in panel VII, more than 31,000 years –the most ancient of the painted set– places this motif in the Aurignacian-Gravettian transition (Hoffman et al., in XVII UISPP World Congress). The hands: about forty negative hands, which could reach one hundred once the walls are cleaned, are recognized (Fig. 3). At least 13 are left hands and 6 are right hands. In some case, the forearm also appears to be painted, and in two more cases the fingers rise so short and separate that they seem to be a bear claw –panel VII (Fig. 3.2). Children´s hands are frequent. Highlighted by its size is a baby´s hand at the bottom of the cave, as is also found in Gargas (Sahly, 1975). The calcite crusts overlapping several hands can be seen (Fig. 3.4) allowing the dating of six cases by U/Th, establishing the most ancient date higher than 27,500 years and being placed in the same dating state as all the Gravettian hands (Hoffman et al., i.p.). Its location presents a concentration in two areas of the vertical wall: the bottom of the cavity, with 18 samples grouped in 3 panels –I, II and III– and the right wall with 16 samples in another eight panels –VIII, IX, XIII, XIV, XVIII, XIX, XX and XXI–. However, the most interesting core is placed in the centre of the roof –panel XV–. There three black hands from children with incomplete fingers appear together with 2 red hands which could be related to the complex series of red points presenting radial motifs. Seen as a whole, the hands at Fuente del Trucho present two peculiar features which make them different from the known hands in the Cantabrian Coast and that, on the other hand, approximate them to the French representations of the Pyrenees north side: 1) there are painted hands in black and 2) a great deal are incomplete, lacking the third finger. In fact, among 173
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Figure 2. Frieze  s calque    panels VI and VII   . Below, detail of the listing h...
174 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 3. Calque of the points and hands series placed in the centre of the roof –panel XV- together with several hand photos. Notice the incomplete fingers in hands 1, 3 and 5, the bear claw appearance of hand number two and the chalky crusts over hand number four where the samples for U/Th were taken. Photos treated with D-Strech.
174  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES the hundreds of hands represented in the Cantabrian Coast caves –mostly at Castillo (56), the Garma (32) and Fuente del Salín (14), and in minor cases at Altamira, Tito Bustillo and Cudón– just two are painted in black; one, negative, from Gran Techo of Altamira and other, positive from Fuente del Salín and only the hand at Cudón presents incomplete fingers. Instead, in the south of the Peninsula we find two black hands with incomplete fingers at Ardales (Málaga) and with an alteration of the little finger, in red, at Maltravieso (Cáceres). However, on the French side black hands are predominant, and at the same time the majority of the incomplete hands are located in the Gargas and Tibiran caves in the same vertical as Fuente del Trucho, on the other side of the Pyrennes. This fact makes us suggest as a hypothesis that real finger loss, especially the longest middle finger, could be due to freezing when crossing the Pyrenees, with a higher incidence occuring in the weaker people, the children (Utrilla, 2005). Valuing the classic interpretation of Leroi Gourhan as a hunting code with bent fingers is not possible at Fuente del Trucho, as most of the incomplete fingers appear on children´s hands. There are some doubts about the superposition between black hands and red points. In Ripoll´s opinion, both red hands and the black ones in panel XV are infra-placed to the red points series (Ripoll et al., 2001). However, Beltrán and Baldellou observed in 1981 that “a black hand is over the red points”. It is not easy to solve this issue as in the left part of the little finger some red points seem to be hidden under the black hands’ halo, while in the index finger the red points appear over a colorless black (Fig. 3.5). At the moment, the U-Th dating (Hoffman et al., i.p.) would lean toward a greater point antiquity, although, as it is an ante quem dating, both issues could be contemporary. The Gravettian crust dating recovering the hands would come after the dates, about 26,000, which have been given the cave Cosquer or the 26,860 ± 460 BP of a bone in Gragas, as the ones calibrated cal BP give dating around the 32,000, which are the ones compared to the U/Th at Fuente del Trucho. A similar dating to ours is given by the charcoal found at the foot of the panel at Fuente del Salín (calibrated about the 27, 000) (García Diez and Garrido, 2012). Regarding the hands at Castillo where the same technology has been used by the same team, the most ancient dates reach the 37,630 in the sample 0-82 (Pike et al., 2012). The horses: seven clear samples and two doubtful ones all painted in red are recognized in the in- ner room. Three of them –two in panel VI and one in panel VIII– are placed in the same frieze, filling an intermediate space between two bands with hand figures, and a fourth sample, in panel IV, is placed in the left of the opposite edge of the frieze in the roof. They seem to be attached to a linear series of points –panel VI– or to digitations –panel VIII– presenting on the two horses´ heads a high compositional similitude both with stiff horsehairs and listed lines on the neck (Fig. 2). This presence of this detail could date them in the Evolved Solutrean, as could be indicated in the manes on listed necks of an engraved horse on stone in the Petite Grotte de Bize which Sacchi –1986– attached to the Upper Solutrean but according to Djindjian –2013– it comes from an ancient excavation – Genson– with mixed materials, so it could belong both to Solutrean and Gravettian. In panel VIII, associated to 5 digitalization, a fifth horse head with the long nape and nostril of a horse appears, from which a crust shaped on its back has been dated showing a possible performance before the 29,000, that is to say, in Gravettian chronology. On the other hand, the total similitude of this figure with the one represented in sector C2 of the Pasiega is surprising: the same long and fallen nape, a forward curved bow in the horsehair, and a double line on the back (Utrilla et al., 2012, fig. 7). This sample is attached to two series of point curves and to a triangular sign (González Sáinz and Balbín, 2002). In panel XII, on the roof, two opposite horses appear, one of them is a complete figure, wounded by a lance or a dart. The lack of details in the inner part, the pronounced cervical-dorsal curve and the legs in open parallel lines shaped like brackets, could classify it into the Middle Solutrean, according to the sequence of Parpalló (Villaverde, 1994). The other sample, only represented by a small elongated head and a long curved neck would fit better with the Gravettian (Utrilla et al., 2012). Deer, goat and bear: In panel VII a colorless “headless horse” was published (Ripoll et al., 2001) but the digital treatment of the figure throughout the DStrecth application shows a small elongated head, a recognizable deer horn and a short tail (Fig. 2). This fact excludes its cataloging as a horse in spite of presenting an identical morphology to the horse in panel VI, with massive hindquarters and legs in open parallel lines (Utrilla et al., 2012). In panel XXI a little goat with an upright open nose and small parallel horns appears, framed by fis- 175
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  the hundreds of hands represented in the Cantabrian Coast caves    mostly at Casti...
176 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD sures and attached to hands and three or four trilobed signs (Utrilla et al., 2012, fig. 9). This little goat is similar to a zoo morph at Nerja (Sanchidrián 1994, fig. 66) for which a pre-Solutrean chronology, according to the lighting traces, is proposed given the dating of 24,130±140 BP (28,532-27.832 calBP) and 20,980±100 BP (25,600-25.060 cal BP) (Medina et al., 2010). This would match with the TL and U/Th dates proposed for other similar goats in panel IV/6 of the Garma about the 26.000 (González Sáinz, 2003); or the female deer with trilinear head at Antoliña in a stone of a Gravettian level dated in en 27,390±320 (31.942-30.840 calBP) (Aguirre, 2007). There is another little goat coming from the Solutrean-Gravettian II of Parpalló –number 18,100, fig. 156-, although it corresponds to the TNT type –triple naturalist trace–, more frequent in the initial Solutrean (Villaverde, 1994). Finally, in the very centre of the line of points in panel VI, a large head is represented which could belong to both a bear and a bovine or a horse (Fig. 2). The tri-lobed signs appear in a highlighted position in two panels: with the shape of a pointed tri-lobed in the frieze of panel VI, attached to the listed-horsehair horse (Fig. 2); and in panel XXI with the shape of three or four signs of semi-circular forehead apparently attached to the little goat, two hands and a point series (Utrilla et al., 2012). The pointed tri-lobed presents several well-visible crusts on its line, over which four datings, the most ancient close to 26,000, have been obtained by the Pike and Hoffman team, confirming again the proposed Gravettian chronology. Regarding its reading, the existence of an elongated sign crossing one of the lobes led Beltrán to identify it in 1993 as a vulvar sign, although it could also correspond to the horse´s belly. The vulvar shapes of Castillo –the scutellum– would be distantly similar motifs, or Micolón –with an inner trace in both cases–, or the triangular signs of ancient sanctuaries, as Pasiega, La Lluera II, Lloseta, Chufin o Maltravieso. However, Eric Robert has not documented on his Franco-Cantabrian Corpus of signs –2006– any sample of the same type as the ones at Trucho –personal communication-. In the opinion of Casado –1979– triangular shapes are more common in the central area of the Peninsula, with the most important core at Ojo Guareña and Maltravieso, in this case related, as in Trucho, to hands and digitations. A last fact to report is the topographic distribution with a main and well-visible position in the panels of Fuente del Trucho, which separates again the prevailing trend of the full Cantabrian signs –quadrilateral and oval– performed “in diverticulum, lateral camarines to the communication routes or in the main composition edges, frequently noting a quest for hiding which contrasts with the wanted visibility for the animal figures” (González Sáinz, 2005). 4. The engravings panel In a central position and in a preferential location, a large bear figure appears in a ball like hibernating position, shaped by a technique combining excision for the body and incision for the head. Also documented is an engraved head of a second bear and a claw and a foot of the same animal, both from an excision technique (Utrilla et al., 2012, fig. 2, number 1 to 4). On the right of the bear, three herbivorous heads in the same vertical can be distinguished. From top to bottom, a horse and a deer looking left and a second horse with a rectangular nose looking right, maybe a feline according to Beltran´s reading –1993– , are identified. It is also difficult to identify the species of the deer: the starting of the webbed horns forward and next to the forehead excludes a deer on the interest of both a reindeer (Beltrán, 1993; Ripoll et al., 2001), or a Megaceros, an animal fitting into the ancient chronology and of which engraved samples in the Meseta (Siega Verde, Cueva del Reno) have been documented (Alcolea and Balbín, 2003). The typical hump we saw in the Cougnac samples is not discerned, although the very small head fits with the Megaceros. The set is clearly atypical, both for the performance technique –the excision on the bear and the claw– and for the subject matter of the animals represented –there are neither female deer of tri-lineal heads nor headless bison nor wild bull– although the bear finds parallels on the engraved claws in Niaux or in external sanctuaries such as Venta La Perra (Arias et al., 1998). 5. The archaeological deposit The archaeological material was obtained from two different places during the excavation carried out by Utrilla and Montes in the 2005 campaign: 1) at the bottom of the tri-lobed signs of panel XXI an altered level emerged, indicating, through the classic “fossil director” of the lithic industry and some absolute dating, the existence of real people from the Initial Upper Paleolithic and the Solutrean (Fig.4). Other-
176  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  sure...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES wise, with some doubts, it might come from the Magdalenian with typological echoes of the former period (4 multiple borers, two of them en étoile) or Final-Upper (little circular endscrapers, others unguiform and two small points with a marked central peduncle, as the Teyjat type). On the other hand, a thermoluminiscent date of 13.244±945 obtained from an endscraper crackled by fire would suggest the presence of hearths in the Middle Magdalenian, but there are no significant bone materials from this period. The bone industry has only provided one sub-circular section of a spear fragment and a longitudinal section shaft fragment with a rounded and polished edge similar to a sample found in the Gravettian at Reclau Viver. 2) at the bear´s bottom, in the deep trace engravings area, outside the cave. Two areas are distinguished there: one modified by the Late Middle Ages structures in shape of circular buckets and hearths paved with stone; and another adjacent one hardly explored which, in principle, could be intact, something we expect to confirm in the 2014 campaign. The first one provided some wheels and glazed pottery and a date of the hearth´s coal of about 1,235±35 BP (GrA-29918), which calibrated to 68% supposing a 776 ± 63 a.C. This fact brings the settlement to the beginnings of the Muslim presence in the upper Vero, and may be related to the moment in which Abderramán I started a well-documented punishment campaign in this area –781–, after Charlemagne’s expedition in the year 778 to Zaragoza. The second area provided altogether six backed bladelets, one of them a hump-backed piece. Recently we have dated a single bone in 31,880±220 BP (Beta 365760) (Cal BC 34,560) whose chronology is Auragnician; in any case, we cannot discard it belonging to an underneath Mousterian level that could reflect a rejuvenation produced by the roots. Together with them, there are 6 endscrapers, 2 borers and 4 scrapers. Our objective for the 2014 campaign is to check if the area is really intact, what comes up from the bone and how they fit with the next sample of Anna Mir excavation which provided two fertile levels with Mousterian materials produced. The typological stock list of the pieces attributable to the Upper Paleolithic from the messy level of the inner survey provides 14 flat, 8 carenated and 4 nosed endscrapers. These last ones lead us to suggest, together with the existence of 15 retouched blades, some of them strangled, the presence of Aurignacian people in the archaeological site. As Gravettian elements –or Magdalenian–, 21 points and little backside leaves among Gravette, micro-Gravette, Vachons points and other short peduncles, more like the Teyjat type than the Font Robert type, could be included. Among the 12 backed bladelets, some present an oblique truncated not getting scalene. There are also 4 écaillées shapes which are present in Gravettian contexts. There are 7 dihedral burins and 8 truncated, one of them multiple, as the burins in Noailles, although they are really large. Four flat invasive retouched pieces and 4 notched points of the Mediterranean type should be attached to the Solutrean, one of them identical in its typometry to the ones found in the close cave of Chaves dated in 19,700 BP, a moment fitting suitably with some horse styles. A more detailed study of the lithic materials can be seen in Montes et al., 2006; Montes and Utrilla, 2008; Utrilla et al., 2010; and Domingo et al., 2012. 6. Chronology In short, the parietal art of Fuente del Trucho is framed within the ancient moments of Paleolithic art, at least within the Gravettian, as the U/Th dating of the crusts overlaying the points series, hands, tri-lobed and some of the horses –with elongated and fallen nose– shows. The stylistic sequence of the first Llonín horizon remains like this (Fortea et al., 2004) and of the archaic panels final part of the Garma (González Sáinz, 2003). The horse with the elongated neck and small head in panel XII would be also archaic, according to the examples of Parpalló and the small goat with the upright and open nose –similar to the zoomorphals “dated” at Antoliña, Nerja, Parpalló or la Garma–. An AMS dating of 20,800±100BP (25,41424.640 calBP), from a bone removed from aremanied level at the small goat´s bottom, could place this occupation in Late Gravettian or Early Solutrean. However, the lithic typology of some fossils heading the deposit also attests to a later Solutrean presence at the cave, highlighted by flat-retouched pieces or cutting points which would indicate a Middle and Upper Solutrean. Due to stylistic criteria fixed at Parpalló (Villaverde, 1994), the Middle period would correspond to some horses, like 177
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  wise, with some doubts, it might come from the Magdalenian with typological echoes...
178 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 4. Lithic industry of the 2005 excavation. Numbers 1 to 17: materials found in the messy level of the inner sequence: 1 to 3 nosed endscrapers; numbers 4 to 9 backsides and peduncle points; 10 to 12 possible Solutrean pieces; 13 lissoir on bone; 14 to 17 Magdalenian typology pieces; numbers 18 to 23: little points and backsides from the excavation at the bottom of the engravings. the one with legs in parallel open lines in panel XII, the one in panel IV or its twin, the deer in panel VII, and maybe the Upper Solutrean could be applied to the listed horses of the frieze, with well-marked details on the horsehair, although Bize´smoble parallel results now insecure (Djindjian, 2013). It has not been possible to determine if the horse in panel VI is overlaid on the dated tri-lobed, whose chronology was better indicated. In conclusion, once the presence of the Gravettian and the Aurignacian at Fuente del Trucho is confirmed, we would stand before an archaeological site in an intermediate location between the forceful seat of Seriñá at Girona (La Arbreda, ReclauViver) and the Cantabrian Coast sites. Fuente del Trucho would as a result be a main point in the east-west mobility of the South Pyrenees.
178  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Jordi Rosell Ardèvol*,** Antoni Canals Salomó*,** Las Fuentes de San Cristóbal (Veracruz, Huesca) Las Fuentes de San Cristóbal was located on the left margin of the Isabéna river passing through a narrow gorge of the same name formed in the sandstones of the Areny Formation, south of the town of Veracruz (Huesca). Its coordinates were 42 ° 19’36 .6 “N 0 ° 34’13 .2” E. The site was located at the base of the old A1605 road between the towns of Serraduy and Beranuy, 820 m above sea level and just 20 m above the present river bed. It was precisely the route of this road which cut the shelter section, leaving only a witness to the north of 2 m wide and 5 high (known as Profile 3 or P3); thus, a hall of 15-20 m2 surface by 2 m high in the south direction (P1 and P2). Fills on both sides had similar characteristics: a low fluvial section and an upper dominated by the contributions of slopes. Part of P1-P2 sediments were used during the construction of the road, so the archaeological work developed between 1998 and 2002 was limited to excavating the witness P3 and the entire surface of the lower levels of P1 – P2. Following this work, the site was destroyed by the redevelopment of the existing road. The stratigraphic section log of the site was described from the Profile 3. Later work consisted of making connections between the two sectors of the settlement. Thus, in P1-P2 5 archaeological levels were discovered, named A-G from top to bottom, and ten different levels in P3 were called consecutively M-V (Fig. 1). Subsequent correlation work revealed correspondence between the two basal levels P1-P2 (F-G) and level V from P3. From an archaeological point of view, the higher levels of P3 included within the slope sediments, were characterized by slightly denser accumulations of artifacts (stone tools, bone fragments, and charcoal). The arrangement of these scored elements seeming paleosurfaces. They could have originally been spread over the surface of the shel* ** ter. Basal levels placed on the fluvial sediments were thicker and characterized by high density anthropogenic objects and some hearths. The raw material used to produce most of the lithic industry was flint (73.5%), distantly followed by micritic limestone (11.3%). The other materials (porphyry, quartzite, sandstone, and quartz) were used in a testimonial way (Menéndez, 2009). The 2,199 elements studied show a representation of all categories of the operative chain, suggesting these knapping activities were performed on the site. Flakes are the most exemplified products being 95% overall. Cores (0.8%) are usually in very advanced stages of exhaustion showing discoid reduction strategies and to lesser extents the Levallois. Retouched elements are rare and account for only 2.2% of the recovered lithic elements. Denticulate and side scrapers are the most common elements, along with some isolated notches. Regarding the faunal remains, the degree of fragmentation is very high, which prevents a greater degree of taxonomic identification. Ungulate remains corresponding to deer (Cervus elaphus) and horses (Equus ferus) were recognized. Evidence of rhinoceros (Rhinocerotidae indeterminate) has also been recognized. The relationship between humans and faunal remains is determined by the presence of some cut marks on bones. The appendicular elements clearly dominate different levels, with occasional dental fragments representing cranial skeletons. We have not identified remains from the axial skeleton (vertebrae and ribs). Neither was the action of carnivores recognized in the cavity. All these elements relate the presence of ungulate traces in the cavity with hunting activities of human groups, and also explain the differential carriage in favor of the limbs of prey and the consequential abandonment of other post-cranial axial parts in the catching field. The identified taxa Àrea de Prehistòria, Universitat Rovira i Virgili (URV), Avinguda de Catalunya 35, 43002 Tarragona, España. IPHES, Institut Català de Palaeoecologia Humana i Evolució Social, C/ Marcel·lí Domingo s/n (Edifici W3), 43007 Tarragona, España. 179
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Jordi Rosell Ard  vol ,   Antoni Canals Salom   ,    Las Fuentes de San Crist  bal...
180 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. The geographical localization of Las Fuentes de San Cristóbal in the Pyrenees context. Overview of the site and stratigraphy of P3. suggest hunter preferences to open environment resources, such as the plains near the bottom of the valley. Even though the steep dominates the landscape, nothing indicates its use by these human groups. Several series made with Accelerator Mass Spectrometry (AMS) and 14 C datings were performed using coal from different levels (Table 1). All of them gave consistent results placing the field in a framework that covers the Middle / Up- FSC-Level Location Method Data Lab. Ref. Bibliography Ref. E P1-P2 C AMS 38.650 600 OxA-19145 Maroto et al., 2012 F P1-P2 C14 AMS 39.290 +490 -410 GrA-33817 Maroto et al., 2012 37.330 +490 -410 GrA-33904 Maroto et al., 2012 36.200 ± 350 OxA-19933 Maroto et al., 2012 38.550 ± 450 OxA-19934 Maroto et al., 2012 14 P1-P2 G P1-P2 C AMS 14 P1-P2 M P3 C AMS 20.220 ± 380 OxA-8591 Rosell et al., 1998 O P3 C14 AMS 27.200 ± 1.000 OxA-8589 Rosell et al., 1998 P P3 C14 AMS 36.000 ± 1.900 OxA-8591 Rosell et al., 1998 14 Table 1. Datings obtained from different levels of Las Fuentes de San Cristóbal.
180  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES per Paleolithic transition. Likewise the lower levels, F-G in P1-P2 and the correlation of V in P3 are placed near 39 ky. The P level from P3, where a Mousterian point was recovered, represents the last assemblage of the site associated with Middle Paleolithic technological complex (Fig. 2). Unfortunately, the O and M levels, dated to 27 and 20 ky, respectively, did not provide any diagnostic information allowing us to assign them to a specific chrono-cultural period. Although the attachment to the Middle Paleolithic level P was initially taken with caution (Maroto et al., 2005; Vaquero et al., 2006), this level demonstrates the existence of human communities in inland areas of the Pre-Pyrenees with some temporal continuity between 40 and 35 ky. The discontinuity appears to have occured after the P level, as the data gap spanning a time frame of about 9,000 years displays. This lack of population could be related to the disappearance of the Middle Paleolithic human populations in the region and a late re-occupation of the territory by groups of the early Upper Paleolithic. Figure 2. Point recovered in the P level dated to 36 ky. Lourdes Montes*, Pilar Utrilla* The cave of Los Moros 1 at Gabasa (Huesca) 1. Presentation and history In the 1980s, a joint project was undertaken between the University of Zaragoza and the Archaeological Museum of Huesca. Between 1981 and 1983 Pilar Utrilla and Vicente Baldellou excavated the well-known cave of Moro de Olvena with Neolithic and Bronze Age levels and in 1984 their joint interventions in two unique sites got underway: the Mousterian cave of los Moros 1 at Gabasa and the cave of Chaves, which in ad* dition to Upper Palaeolithic levels also contained an extraordinary Neolithic deposit, which is now totally razed. The site at Gabasa, five cavities with Prehistoric and Medieval remains, was located in 1982 by M. Badía and visited by Olvena’s team in the same year. In 1983, permission was requested from the Ministry of Culture to excavate all of the cavities but the excavation could not be undertaken until the summer of 1984: Utrilla directed the team that worked on the Mousterian cave and Baldellou directed the excavation of the other Área de Prehistoria. Universidad de Zaragoza. lmontes@unizar.es and utrilla@unizar.es. 181
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  per Paleolithic transition. Likewise the lower levels, F-G in P1-P2 and the correl...
182 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD cavities, whose deposits showed significant alterations. From 1985, the campaigns, 7 in total over 10 years, focused exclusively on the Mousterian site and in 1986, L. Montes replaced Baldellou in the joint direction until the last campaign carried out in 1994. Although it was fenced off, the cave has undergone continuous clandestine interventions, which have led to the almost complete disappearance of the control “balk” respected by excavations for further research. The cave is located on the pre-Pyrenean foothills of Huesca, at the foot of a limestone cliff, which is located to the north of the small town of Gabasa, currently the municipality of Peralta de Calasanz. The cliff is an Eocene limestone syncline that dominates one of the headwaters of the Sosa, a tributary of the left bank of the Cinca and on which the karst of the network of caves has developed, which is currently inactive. The Mousterian cave or Gabasa-1 consists of two small chambers: when it was discovered the outer chamber contained very little sedimentary deposit while the inside, a chasm partially filled with debris, housed the site (Fig. 1 and 2). closed on the outside by a dry stone wall. The upper layers of soil have been repeatedly removed to be used as a fertiliser for the crop terraces that extend along the foot of the cave; in the outer chamber, which has the best living conditions, the deposit has been removed down to a thick stalagmite crust more than a metre thick. In the inner chamber, with a depth of almost three metres, the site is conserved although the fill presented a disturbed surface, particularly in the central area.In its vertical development, this chamber is a bell-type cavity, whose walls maintain carbonate stone edges of varying diameter at different heights, the footprint of water erosion. For this reason, in the site we are presenting, the excavated area greatly exceeds the contour of the walls drawn at the height “0” in the reference plan. The place had been used until recently as a cattle fold and its mouth, directed eastwards, partially L. Montes’s PhD thesis (1988) summarised the techno-typological study of the lithic remains, ex- Over the years, there have been various publications about partial aspects of this site, but a monographic report has not yet been released: The first presentation on the site in Bolskan extracted the first excavation campaigns (Utrilla and Montes 1986). Figure 1. Location of Gabasa1cave at the foot of the cliff that dominates the town
182  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  cavi...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES tending the stratigraphic context of the site to the 1987 campaign. In the same year, the first attempt was made to reconstruct the chronostratigraphy of the cavity (Azanza et al., 1988) and an initial approach to the microfauna (Gil and Lanchares 1987), which was subsequently corrected in unpublished documents (Guillén 1994; Cuenca 2002). The first international presentation on the site was given at a conference on Neanderthals in 1986 in Lieja (Utrilla and Montes 1989). With work more advanced, a new sediment and paleoclimate study was published (Hoyos et al., 1992) whose data was used in a proposal on the end of the Middle Palaeolithic in the Ebrovalley (Utrilla and Montes 1993). A little later, M.ª Fernanda Blasco published her thesis, a key study on the fauna of Gabasa, and an interesting summary of its taxonomic and taphonomic analyses (Blasco 1995 and 1997). Between the two came the presentation at the meeting in Capellades in 1995 on the remains of level g, which analysed in-depth the differences between human prey and the contributions of other carnivores (Blasco et al., 1996); the accumulation of remains of hyena during the last excavation campaigns led to a specific study on this animal (Blasco and Montes 1997). In the meeting in Foz-Coa, the human remains and AMS dating were presented for the first time, with new reflections on the Transition (Lorenzo and Montes 2001 and Montes et al., 2001) The study of pollen preserved in hyena coprolites allowed percolation problems to be solved in the paleoclimate (González-Sampériz et al., 2003 and 2005). M.D. Garralda (2005) included the Neanderthal remains in this cavity in her review of the Iberian Peninsula for the tribute to J. Altuna; it had been previously included in M. Haber’s (2003) PhD thesis on Neanderthals in the Iberian Peninsula. A review of the Ebro Boundary was presented in Lieja, 2001, at the14thUISPP Congress (Utrilla et al., 2004) and a revised and updated version was included in the volume that commemorated the centenary of the excavations of Monte Castillo (Utrilla et al., 2006). In 2006, the tribute to V. Cabrera was published; this presented a review of the Middle Palaeolithic in the Ebro basin, which included a summary of the data on Gabasa (Montes et al., 2006). A detailed technology study of the levels of Gabasa based on a wide collection of the lithic remains from all of the campaigns was presented (Santamaría et al., 2008),taking advantage of the meeting on technology variability organised by the UAB. A review of the hyena remains accompanied by a new dating of level h based on racemisation in dentin was presented at the meeting on the dens of hyena and other carnivores (Utrilla et al., 2010). Other authors have used and published some specific remains from the cave in other studies: Hernández Carrasquilla (2001) published a new species of vulture identified among the bird remains and R. García González (2011) focuses in Gabasa in two separate studies on filiation of the Capra pyrenaica. 2. Stratigraphy The archaeological levels of Gabasa, all of them of Mousterian chronology, were named during the excavation using lowercase letters, from the most recent level a, to the deepest, level h (Fig.2). The latter was presented as an open framework consisting of boulders and stone fragments, in which the fine fraction only remains in situ at the top, and it is accumulated loose through percolation in some points at the bottom, supported by the sub-horizontal walls of the bell-shaped chamber when closed. The colour and texture of the rest of the layers were very similar in appearance and the only significant difference was the size of the clasts. The following proposal summarises the stratigraphic data based on a sedimentology study (Hoyos et al., 1992), a basic count of the lithic remains (Montes 1988 and Santamaría et al., 2008) and of the remains of fauna (Blasco, 1995), and the dates available (Montes et al., 2001; Utrilla et al., 2010). The sedimentology study allows the fill of the cave to be associated to the limestone lithology in which it opens, formed by sub-aerial and underground contributions, except for the fauna remains and the materials contributed by human groups. Its detritus and loose nature favoured occasional percolation of pollen and some fine fraction. The stratigraphic units were identified using Roman numerals and the archaeological levels with lowercase letters. – Unit I (level h). Up to almost 2 metres thick, it is characterised by the abundance of thick elements and within these, by the dominance of great boulders over minor blocks, whereas 183
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  tending the stratigraphic context of the site to the 1987 campaign. In the same ye...
184 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 2. Stratigraphy and site at Gabasa the fine fraction only appears in the top area. It seems to have formed under cold and dry conditions in which freezing, significant in duration but not intensity, was responsible for the gravitational contributionand boulders and stone fragments detached from the walls. The presence of fine fraction in the top is interpreted as a change towards wetter conditions. Mixed in with this rocky framework, in level h, 300 lithic and 1390 fauna remains were recovered. The bones were apported both by humans and carnivores. Dating based on aminoacid racemisation in dentine of a total of 15 horse’s teeth returned an average age of 140000±43000 (LEB 8538-8558). The measurement, as it is inaccurate, puts this section from the deposit in phase OIS 6, taking it out of the dates managed up until then for the entire site. – Unit II (levels g + f). Arranged without sedimentary discontinuity with the former, the sedimentology grouping of levels gand f, an average 50 cm thick, shows a matrix of brown clays with gravel and sand that contain small stone fragments. Level g was distinguished from level f during the excavation by the dis- appearance from the matrix of the small calcareous concretions and scattered charcoal that appeared in it. Regarding the archaeological remains, there is a significant difference between the two layers: while more than 1200 lithic remains correspond to level g, level f yielded just 550, with an excavated area of one quarter smaller. Something similar happens with the fauna; 4194 remains in f compared to 8741 remains in level g. In the latter level, the presence of a hyena den is noteworthy, partly accounting for the large number of animal remains. For level g an AMS date was obtained that only specified an age earlier than 50700 BP (OxA-5675). Its immediate superposition over Unit I, without erosion traces or stratigraphic discontinuity, suggests an old chronology for this series of levels, also in the Middle Pleistocene. – Unit III (level e). This level was formed in the outer chamber, from which, in a clear process of solifluction, it entered the inner cavity, resulting in partial erosion of level f, over which it settled. The clay matrix with gravel and calcareous sand is very similar to the previous
184  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES one but this includes large stone fragments and boulders. Its structure is responsible for the varying thickness: from 70 cm in the areas connecting the two chambers, to an average of 30 cm on the two sides of the expansion fan. The lithic remains add up to almost one thousand, whereas fauna remains amount to 4795. Although an AMS dating specifies only dates prior to 51900 BP (OxA-5674), a conventional C14 dating would put it in the late OIS-3, despite its wide deviation: 46500 +4400/-2200 BP (GrN-12809). – Unit IV (levels d+c). This sedimentology unit is characterised, like the previous two, by its clay matrix, which in this case includes an abundance of stone fragments, platelets and boulders. The difference between the two archaeological levels was established by the higher compactness of level d and its lower clast content compared to the overlaying level c. The lithic remains recovered reached 531 in level d whereas in level c 253 remains were counted; besides, other 215 were attributed to a+c, in some areas where the lack of level b prevented these two layers from being distinguished (vid. infra). The same happens with the fauna remains, which amount to 1862 in level d and 2133 in a+c. In this unit, the upper section, level c, has two AMS dates: >47800 BP (OxA-5673) and >46900 BP (CAMS10290/Beta-68391). Perhaps, the carbon dating obtained for level a+c could also be related to this layer c, given the similarity of the result: >45900 (OxA-5672). – Unit V (level b).The only archaeological level identified was sterile: a carbonate crust, with little development and discontinuous, whose absence creates the aforementioned problem to distinguish between c and a. – Unit VI (level a). Another clay matrix with numerous stone fragments, platelets and boulders, whose only difference with level c, according to the sedimentology (the stones edges were sharpener), we were unable to identify during excavation. As we said, in some areas which did not have level b, the material from a + care considered as a whole: 478 elements from the total of a + c can be singled out for definitely belonging to this level a; as aforementioned, 215 elements had to be attributed imprecisely to a + c. In the fauna, the total of a + c amounted to 2133 elements. Regarding its chronology, in addition to the generic AMS dating prior to 39900 BP (OxA-5671), we must remember the one obtained from another charcoal from a + c, which we have indicated in level c. 3. Paleoclimate characteristation of the record from Gabasa The attempts at paleoclimate reconstruction of the Gabasa deposit were corrected in the aforementioned subsequent studies, qualifying the interpretation of the site. After the initial merely climatic estimate, made using provisional data on the fauna, pollen and sedimentology in the early campaigns (Azanza et al., 1988), a revision of these was undertaken, proposing two options based on the sedimentology analysis (Hoyos et al., 1992),which determined cold conditions for the bordering levels, a and h, whereas the centre of the deposit was considered warm in general, with some cooler oscillations in levels g-f and particularly in e, and with varying humidity: the first option put the development of the sequence in the Würm II (based on the dating of level e, the only one available at the time),whereas the second placed the central unit (levels c-g) in the interstadial Würm II-III, with two bordering levels, level h at the base, placed in Würm II, and level a,at ceiling level, as the start of Würm III. The subsequent study of the fauna carried out by F. Blasco (1995) backed the majority of these observations, particularly in relation to the dryness/ humidity of the conditions that affected the levels of the deposit (except a, dry according to the fauna assemblage). The testimonial presence of Cuon and Leopard confirmed the warm nature of the central unit of the sequence, but being aware of the low determining value of the macrofaunain identifying the oscillations in climate that could have arisen during the period in question. A brief microfauna analysis of a random sample carried out by P. Guillem ratified the determination of the macrofauna and the sedimentology as regards the warm/dryness of the deposit, coinciding with the fauna as regards level a, also dry (Blasco 1995: 60). A little later, the series of AMS dates obtained by R. Hedges at the Oxford laboratory and another by Beta Analytic, although inaccurate (between >39.9 ky BP for level a and >50.7 ky BP for level g) steered us towards the first paleoclimate proposal mentioned 185
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  one but this includes large stone fragments and boulders. Its structure is respons...
186 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD earlier, which put the entire sequence at least in the Würm II (Montes et al., 2001). However, new results from a pollen study of hyena coprolites in this site (González-Sampériz et al., 2003 and 2005), qualified this proposal, keeping the entire Gabasa deposit in Würm II with permanent conditionsof dryness and cold or cool temperatures, except at the top (level a). The proportion of Quercus t. ilexcoccifera in level a, dated at an inaccurate > 39900 BP, would suggest milder and warmer conditions, however, that could be related to the start of the interstadial Würm II-III, as documented in the apparently contemporary levels of Beneyto X, Carigüela V or in the base of Arbreda. The top of the Gabasa sequence could therefore be included in the globally warm conditions of OIS 3, distinguishing it from the rest of the site. The warm nature that the palynology study would assign to this level is not consistent with the cold nature assigned by the sedimentology based on the proliferation of frost weathering inside stone fragments (including frost weathering of platelets). As the study showed (González-Sampériz et al., 2005: 593), this difference could be resolved by considering the presence of platelets as an occasional occurrence and not as something from the general environment. We currently live in a climate considered warm but on some winter days with heavy or prolonged frost, frost weathering of platelets can be generated in the cave at Gabasa. At the start of some excavation campaigns we noticed these types of platelets on the surface of the cave and the consequent scar on the roof of the cave, which proves that it is of recent origin. This could have been normal in the area, the Ebro valley, where marked continentality is one of the most significant characteristics of the climate. Later, the earliest settlement, level h was dated by T. Torres and J.E. Ortiz at 140±43 ky (LEB 85388558) by aminoacid racemisation in the dentine of horse’s teeth (Utrilla et al., 2010). This dating, except for the problem posed by its very wide range, a priori placed the old thick deposit of level h in the early Middle Pleistocene. The new date offered was a surprise and led us to evaluate a more extensive chronology, with longer hiatuses in the filling process: clear discontinuity between units II (g+f) and III (e); interruptions in the sediment in the core of IV (d+c); viscous and mass transport to the inner chamber of levels III (e) and VI (a)… This meant considering a chrono-stratigraphy review of the deposit and suggested that the lower part (lithological levels I and II, or archaeological level sh and g+f) corresponded to an older chronology, from the Middle Pleistocene. The discontinuity that marked the base of level III (e)would represent the start of a new sedimentary phase, whose chronological distance from the previous one we cannot identify using the vague “prior to” dates that we have for the site. The upper units (III to VIII), could be immediately posterior but also more recent, if we would accept the conventional date of level e. 4. The lithic and bone remains and the interpretation of the site In Gabasa 6 human bones attributed to the Neanderthal have been recovered (Utrilla 2000; Lorenzo and Montes 2001; Garralda 2005): a first upper premolar (P3) from level f attributed to a young person; a lower right molar (M2) also from level f, perhaps from a female adult and another lower right molar (M1) from the remained sediments, this one with a toothpick groove and probably from a male adult. Corresponding to level e are a right clavicle, without its apophysis and a first toe phalanx from a left foot and finally, the remained sediments also included a first metatarsal from a adult right foot. The magnificent collection of fauna studied by Fernanda Blasco (1995 and 1997) presents more than 23,000 remains, over half of which can be determined. From these, 23 species of mammal have been distinguished: 9 ungulates (Equus caballus, Equus hydruntinus, Bos sp., Rhinocerotidae, Cervus elaphus, Capra pyrenaica, Rupicapra rupicapra, Capreolus capreolus, Sus scropha); 10 carnivores (Ursus spelaeus, Crocuta spelaea, Panthera spelaea, Panthera leo, Panthera pardus, Canis lupus, Cuon alpinus, Vulpes vulpes, Lynx spelaea, Felis sylvestris), 2 mustelids (Meles meles, Mustela putorius) and 2 lagomorphs (Oryctolagus cunniculus, Lepus sp.) In addition, 20 species of bird and a very significant collection of microfauna were identified. Almost half of the remains allow the species to be identified, which suggests extraordinary conservation of the collection and a low bone processing index. M.F. Blasco’s PhD thesis, expressed very cautiously due to the chronological estimates available at the time of the study, suggests the possible presence of older types of horse and hyena among the remains: possible dating of the lower levels (f, g, h) in the Middle Pleistocene would fit into these approaches. The taphonomic study of these fauna remains, their distribution over the site and the relationship with the human artefacts suggest alternating use of the cave between humans and carnivores. Therefore,
186  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  earl...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Figure 3. Gabasa: scrapersand Mousterian point (no. 10) from levels a+c (1-4), d (5-8) and e (9-12). Figure 4. Gabasa: scraper and limace (no. 4) from levels f (1-4), g (5-8) and h (9-11). the Neanderthals, who mainly hunted horses and red deers, must have occupied the cave in the summer, whereas the carnivores, whose prey marks are present on ibex’s bones, always adults, would occupy the cave at other times. We know that the foals and fawns were hunted by humans from the clear marks (slits, grooves…) left by the flint instruments on the bones, while processing the prey. However, it is not currently possible to determine the pattern (in the same year or in different years) this alternate occupation followed (Blasco 1995; Blasco et al., 1996). lithic flakes, in addition to a few but very good points and slightly exploited cores (Fig. 3 and 4). In the two more recent levels (a and c),the technology and typology change (an increase in certain backed knives), which is reflected in an artificial increase in the group of tools from the Upper Paleolithic and which could fit into the type B Mousterian Acheulean tradition (Montes 1988; Utrilla and Montes 1993). Regarding the instruments recovered, the lithic remains currently analysed typologically relate these assemblages to the facies «typical, rich in scrapers» of the Mousterian: they are basically assemblages with many scrapersand untransformed The lithic assemblage of Gabasa-1 show a high level of technological homogeneity throughout the sequence (Santamaría et al., 2008). The main operating sequence in all levels is discoid. However, two secondary operating sequences have been identified; Levallois and Quina, whose reconstruction and individualisation require more detailed study. Raw 187
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Figure 3. Gabasa  scrapersand Mousterian point  no. 10  from levels a c  1-4 , d  ...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 188 materials would have been supplied from the surrounding area of the site; the selected nodules were tested in situ and then brought into the site. Production of supports was mainly done using simple albeit very economical and fast (discoid) knapping methods and techniques. The supports obtained (cortical products, flakes “à dos débordant”, ordinary flakes and centripetal ones) were mainly transformed into scrapers and, to a lesser extent, denticulated and notched tools, and used, along with unretouched flakes, to process prey (fawn and foals). The low incidence of retouched flakes in the assemblage suggests that some of the tools were brought into the site. The typological differences of level a+c are related to changes in the transformation of supports into tools, characterised by a significant decrease in the group of scrapers and different treatment of the flakes “à dos débordant”, which are not so much transformed into scrapers but are used in a raw state. To summarise, it is a magnificent site that, considering the type of fauna found (ibex, horse and deer dominating over other herbivores and a variety of carnivores, including cave hyena) and the lithic industry recovered, is interpreted as a hunting camp specialising Alfonso Alday* Kanpanoste and Kanpanoste Goikoa sites (Vírgala, Álava, Basque Country) Two shelters are presented together due to their geographic proximity –roughly 200m apart– and their shared stratigraphic sequences resulting from the same activities by Meso-Neolithic communities, parallel to the neighbouring Atxoste site. The shelters are in the southern foothills of Azáceta Pass. Kanpanoste faces west and Kanpanoste Goikoa west-northwest. The first, next to Berrón River, retains a visor measuring almost 12 m long and 2 m wide. The second, more difficult to reach, provides better shelter with a 13m long x 3 m wide roof, although originally it was larger. As usual at sites from this period and zone, their features include a good strategic location, * in fawn and foals, which contradicts the cliché of peremptory and indiscriminate hunting with a lack of specialisation of the Neanderthals. Subsequent processing of the prey (cutting, deboning and preparation of the skins) would be done using the unvaried but abundant lithic material recovered, which would explain the extraction of flakes from cores in situ and the minimum transformation of these supports into specific tools, points and scrapers, which would be brought along already finished. Some big tools knapped on pebbles (hammers, choppers and chopping-tools) could have been used in butchery tasks. Humans would alternate occupation of the cave with other predators, from season to season, particularly the cave hyena and wolf, which would use it as a den and whose prey would have been mainly ibex. Seasonal occupation of the site would explain the minimum preparation of the site (there are barely any stable hearths, although there is a lot of charcoal in the sediment; the failure to remove the angular clasts, some of which were very big and would have made the settlement very uncomfortable) and the conservation of the fauna, truly extraordinary, given that a minimum amount of the prey hunted was consumed in the cave. which facilitated surveillance and access to the open spaces of the Arraia Valley and the mountain landscape in the immediate hinterland. Kanpanoste: Archaeological work by A. Saenz de Buruaga in 1990 distinguished three sedimentary units in a 1 m deep sequence, one of which contains two distinct sections (Fig. 1). Lanh Level: 45cm deep, dipping south-north, consisting of a fine, compact silt-clay matrix, blackish in colour, with fewer larger fraction ele- Área de Prehistoria Universidad del País Vasco (EHU/UPV). Tomás y Valiente s/n 01006 Vitoria-Gasteiz a.alday@ehu.es
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  188  mate...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Figure 1. Top: Stratigraphic section and ecological features of Kanpanoste. Bottom: Stratigraphic section of Kanpanoste Goika. ments and a major component of Helix nemoralis. It is divided into two units for sedimentological and archaeological reasons. The upper unit (Lanhs) includes blocks, while the lower level (Lanhi) does not. The lithic industry in both units is predominated by campiñoide-style notches and denticulates on flakes, followed by denticulates, sidescrapers denticulates, endscrapers and awls. The lack of microliths on the lower horizon and their presence on the level indicate a differential nuance in a changing industry. This unit also contains an interesting collection of perforated nasaridas and an atrophied deer canine which were probably part of a personal adornment. The inventory is completed with a set of stone macrotools for use as hammers/retouchers, scraping or processing plants. The scarce fauna includes wild boar, the predominant species, followed by roe deer red deer and chamois. Culturally, the material can be identified with Mesolithic notches and denticulates. Clag Level: The defining feature of this 18 cm deep level is the considerable volume of edged 189
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Figure 1. Top  Stratigraphic section and ecological features of Kanpanoste. Bottom...
190 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD clasts, associated with a greyish, ash-like silt-clay fraction with a loose structure. It is in erosive contact with the other two strata at both the wall and the ceiling. It is poor in archaeological material, with visits during the Chalcolithic, Neolithic and the late Mesolithic identified from a few denticulates, two geometrics, a speartip with flat retouch and a few flat ceramic fragments. Clam Level: With an average depth of 20 cm, this level contains many angular clasts in a fine brown silt-clay matrix, loose but rough to touch, with a major input of present-day organic matter. This level lack archaeological interest. Level Lanhi Lanhs Code BP Date GrN-22441 8200±70 GrN-22442 7920±100 GrN-22440 7620±70 Table 1. Radiocarbon references of Kanpanoste. Three other analyses have yielded rejuvenated dates due to sample mineralization Kanpanoste Goikoa Archaeological work by A. Alday in 1992 and 1993 detected four sedimentary units, some of which contain several cultural episodes in a one meter thick sequence. Level III-bottom: This level rests on the rocky base of the shelter, limestone with weathering processes which enriched the section with small-sized gravel. Average depth of 20cm, lying on a southnorth dip, and also more gently east-west. Fine, silty, compact matrix with no large components except for the base. Dark brown colour. The retouched lithic industry is not large but significant. Half are notches and campiñoide-style denticulates on flake, accompanied by endscrapers and sidescrapers in the same style. Red deer, auroch and wild boar clearly predominated the identified game. Culturally ascribed to Mesolithic notches and denticulates. Level III– top: A more gradual process than the previous level, with a 20cm potential, a lighter brown soil with a homogeneous silty matrix and an increased proportion of angular clasts, detached from the wall and ceiling of the shelter. Colonies of gastropods, mainly Helix cepaea, are mixed with the sediment, and similarly, numerous charcoal remains resulting of different fires. In fact, a hearth consisting of two rings of limestone blocks organises most of the space. The sedimentation is interrupted by one large and several small blocks which became detached from the roof. Although the notches and campiñoide-style denticulates are the most numerous typological group in the retouched lithic industry, by this stage these items had lost much of their role, replaced by geometric reinforcement (triangles and trapezoids, accompanied by microburins) and backs. Flakes with minimal retouch also made a major contribution. The same fauna spectrum remains, with a slight increase in deer and chamois. This level is culturally ascribed to the geometric Mesolithic. Level II: A steep dip makes its potential (20cm average) vary from one point to another, although it is homogeneous, with very silty soils in a loose, ash coloured structure, with not an excessive amount of angular clasts and limestone slabs. Colonies of terrestrial gastropods were isolated in some sectors, and there is evidence of lit fires in the associations of charcoal and slabs. The worked stone industry has a greater influence than flakes, with a tendency to use simple flakes with marginal retouch and backs as the predominant tools, although geometric items are also present (13% of triangles, trapezoids and double bevel segments, accompanied by microburins) as are endscrapers (10%). Small pottery shards, a few pebbles and slabs, and a grinder for plant processing complete the record. Wild animals combine with ovicaprids, cattle and domestic pigs. Although the homogeneous sediment hinders an internal breakdown of this level, a detailed vertical analysis of the materials has revealed three cultural horizons: geometric Mesolithic at the base (with an industry that has evolved from the top of Level III); Early Neolithic in the middle (with some ceramics, double bevelled segments, a sickle for cutting vegetable and mill); and initial Chalcolithic at the top (indicated by small tips with flat retouch). Level I: The current floor level, 20 cm deep, has a wedge layout, with a dry, dusty silty matrix in the first few centimetres, light brown tending to yellow, and a quite uneven component of blocks and clasts in its 20 cm depth. There are few lithic items (10 retouched tools), ceramics (19 frag-
190  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  clas...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Figure 2. Selection of prehistoric materials of Kanpanoste (de A. Cava). 191
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Figure 2. Selection of prehistoric materials of Kanpanoste  de A. Cava .  191
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 192 ments, one with Boquique decoration), or faunal remains (10 fragments). Level Code BP Date III-lower* GrN-20215 GrN-20455 7620±80 7860±330 III-upper* GrN-20214 GrN-20289 6360±70 6550±260 II* GrN-20213 GrN-20267 3430±60 4350±60 II (domestic cereal) GrA-9790 4550±40 II (domestic ovi-caprid) GrN-202738 4190±100 Table 2. Radiocarbon references for the Kanpanoste Goika site * The laboratory divided the sample into two (carbonaceous and collagen). Due to the small amount of collagen available and hence the breadth of the standard deviation, the dating for the upper Level III should be accepted with some reservations, given the regional context. Carlos Mazo, Pilar Utrilla* The discrete amount of the material inventories for Kanpanoste Goikoa and Kanpanoste (Fig. 2) would rank these sites in a modest position if they were analysed individually. However, their strength lies in their complementarity. They are part of a system of territorial occupation which also involved the Atxoste, Mendandia and Fuente Hoz shelters, as well as several open air sites in Entzia, Urbasa and a few valleys in today’s Alava province. In fact, the Kanpanoste assemblage was used as a basis to describe the Mesolithic notch and denticulate techno-typological unit. The roots of this territory date back to 8600 BP, and remained in place for a millennium. The presence of seashells from both the Cantabrian and the Mediterranean and the circulation of siliceous materials suggest a thoughtful articulation of the space. The features of the lithic industry, flake blanks tending to be carenated and in pieces, denticulate fronts often resulting from recycled material, loss of microlithics, apparent roughness in items which actually follow preset patterns (in which there is a notable lack of hunting equipment) are shared by many sites in the Ebro River basin and the Mediterranean fringe, which shaped a break in the evolution from the early Holocene microflake assemblages to those which characterize the geometric Mesolithic. Forcas I and Forcas II sites The Sierra de Castillo de Laguarres and the Sierra de Torón form a pre-Pyrenean structural unit split by the river Ésera where it passes through the Alto Aragon village of Graus. At this point rise Peña del Morrón (599 m) and Peña de las Forcas (635 m), on either side of the river and up to 300 m above the current riverbed. The latter, located on the left bank of the Ésera and of its tributary the Isábena, represents the most western point of the Sierra del Castillo de Laguarres. The place, which stands on the Aqui* Cultural material Área de Prehistoria. Universidad de Zaragoza. tanian Miocene conglomerates, records the presence of two prehistoric sites, Forcas I and Forcas II, whose stratigraphic deposits show prolonged occupation, almost continuous, from the Lower Magdalenian up to the Late Neolithic, with two subsequent occupations in the Chalcolithic and early Roman Empire (Fig. 1). In both cases, occupation took place under the protection of narrow, very shallow rock-shelters, generated at the base of the conglomerate by the action of the river. The coordinates of Forcas I are X: 280.125;
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  192  ment...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Y: 4.673.241 and those of Forcas II are X: 280.242; Y: 4.673.709, and the height is 471 m and 470 m respectively. It cannot be said that they are sites with a clear archaeological sensitivity (indeed Forcas II is oriented to the north and almost on the shore of a river that flooded it at times, as shown by the levels of sand and silt in the deposit). However, the geography and topography give the site an advantageous position because the Ésera narrows there and the place becomes a checkpoint and obligatory pass between the mountain and the Barasona valley, now covered by a reservoir; because the location provides access to various biotope resources and because the Isábena acts as a transversal pass that connects the Ésera valley with Noguera Ribagorzana to the east. Forcas I was discovered by Jean Vaquer in 1990 and excavated by Carlos Mazo and Pilar Utrilla between 1990 and 1992. It was seriously affected by aggregate mining, which restricted its extension (unknown) to just a narrow 23 linear metre strip of stratigraphic deposit that stuck to the wall of the conglomerate, which at this points runs North to South. In this deposit, 14 stratigraphic units have been recorded (from 4 to 17) with 9 fertile archaeological levels and more than 8000 lithic remains have been recovered. The assemblage represented by the bone industry is very inconspicuous and there are also very few identifiable fauna remains. Occupation starts at level 15, corresponding to a Lower Magdalenian of the classic Cantabrian type, with a date of 14440 ± 70 BP, typologically well authenticated by the presence nucleiform endscrapers and rabots, and which would fit in well with the Lower Magdalenian of the area, particularly with the neighbouring Alonsé cave, with which it shares not only similar radiometric dates but also a similar technology in laminated cores (débitage sur tranche transversale à encoche), which are also present in contemporary levels on the other side of the Pyrenees. Moving up the sequence, levels 14 and 13d are classified as Upper Magdalenian while levels 13a and 11 correspond to the Late Magdalenian. The first two share the aforementioned laminar core technology, while the second two offer clear typological similarity as regards group indices and even primary types, with a variety of scrapers, the presence of burins in a similar proportion and retouched laminar flake. From level 10a, the surface conserved decreases significantly and consequently also the number of elements recovered. Despite the chronological difference between levels 10 and 9, the typological similarity between their industries is almost exact. Figure 1. Stratigraphy of the two rock-shelters at Forcas. In both levels, the lithic industry, although scarce, is based on thumbnail scrapers, consistent with a generic Epipaleolithic or Azilian. Occupation of the site ends at level 7, which was confined to a very small space and offered very few retouched components. Its industrial characterisation is not categorical. Its stratigraphic position and chronology (9360 ± 140 BP) correspond to a microlaminar Epipaleolithic, and the laminar and microlaminar components (which account for more than one third) and a good representation of laminar cores would certainly support this but if we look at the retouched pieces, only two microgravettes would only fit in with it and a 193
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Y  4.673.241 and those of Forcas II are X  280.242  Y  4.673.709, and the height i...
194 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD scalene triangle, which would go well in Sauveterrian industries of the Preboreal. Forcas II is 560 m north of Forcas I, at the confluence of the river Isábena with the Ésera. It was discovered in 1991 and excavated in two phases during 1991-1992 and 1996-1997. It is also a long shelter, which offers more protection than Forcas I, but it also has a very small overhang and faces north. This orientation and its immediate proximity to the river (which flooded it several times, as shown by the flood silt levels) make the site seem like a place that is not suitable for continued occupation, but it is suitable for a temporary settlement, in the summer, as a hunting ground, which was the proposed interpretation. The site had also undergone some destruction and clearing processes and three very different areas are distinguishable: the western area of the long shelter records a stratigraphic sequence consisting of 8 levels, 6 of which are archaeologically fertile. The base unit is I, consisting of fine sand from the river. Between levels Ia and Ic, sterile, is the first occupation of the site, level Ib, which corresponds to a macrolithic Mesolithic with notch and denticulate tools (8650 ± 70 BP) with a poor industry, as is common in old macrolithic contexts (prior to 8500 BP) like those observed in nearby sites, such as Legunova and Peña 14. An industry of crude instruments made with local raw materials such as quartzite and limestone. Levels II (7240 ± 40 and 7150 ± 40 BP) and IV (7000 ± 40 BP) correspond to a geometric Mesolithic, with rough retouched instruments and microburins and they are separated by a sterile level of sand from the river (level III). In level II, always with balanced modules, although there are some long pieces, asymmetric trapezoids and concave sided trapezoids dominate, along with scalene triangles and scalene triangles with a small concave side. In level IV, the symmetric and asymmetric trapezoids appear equally, followed by trapezoids with a small retouched base. Scalene triangles dominate among the triangles. The trapezoids include micro-trapezoids and the triangles include obtuse and both types include some with inverse retouch (Fig. 2). A few segments also appear, either as curved back edge blades or as rough retouch segments. In general, level IV differs from level II in the range of sizes, types and positions of the retouches, perhaps the result of a possible ultra-Pyrenean influence. This level also includes a plaquette engraved with precise geometric designs. In the peninsula Mesolithic similar laminates are found in Cocina II, lo- cated in the same chronological and stratigraphic time as the one in Forcas, that is, at the end of the geometric Mesolithic, at a time immediately prior to the cardium, and in a level, layer 6, in unbroken contact with level 5; now with cardium pottery. However, they differ in the support of the plaquette; that of the Alto Aragon site is smaller, flatter and tabular and the engraving is very shallow compared to the deep lines of the platelets at Cocina. However, the nearest plaquette with the most similar decoration are in southern Italy, on the plaquette of Grotta delle Veneri. In Forcas II, the chronology of the geometric Mesolithic is later, unlike Bajo Aragon and Alto Ebro, and it is not so deeply rooted. There are only 250 years between its appearance in level II and the first pottery in level V, which appeared very early; furthermore, there is no break in the stratigraphy, and, therefore, levels IV and V are in direct contact, closely linked and with no transition. Levels V (6940 ± 90) and VI (6900 ± 45 and 6740 ± 40) represent the moment of old Neolithic transition. As in the previous phase, in level V triangles dominate over trapezoids and, in addition, double bevel retouch appears, which would be exclusive in level VI. Both contain the oldest pottery in the Ebro valley, cardium or impressed ware (undoubtedly exchanged or borrowed from Neolithic groups, perhaps from SE France through the valleys of Tet and Aude and the valley of SegreCinca), but all within a hunter-gather economy. After a moment of abandonment, which corresponds to sterile level 7, consisting of silt, level VIII, the most recent area of this rock-shelter, is clearly Neolithic with drills, sickle elements with a cereal patina and domestic fauna. In the pottery productions, the decoration is restricted to straight cords. In the central area, the most affected by the clearance, a wall canvas consisting of 5 layers of irregular ashlar, some fragments of terra sigillata and a thin walled glass were recorded. The TSH materials offer a dating corresponding to I-III century and the structure might have been used to control the transport routes during the early Roman Empire, and more specifically the river crossing, which has historically been crossed at this point, where there is an old bridge today. Finally, in the eastern area the evidence of occupation starts in level 7 and continues through 6 and 5 and the three levels could belong to the same cultural horizon, considering the high level of coincidence in the technology and forms of the pottery remains, particularly in levels 6 and 5. At the base of level 7, there
194  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  scal...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Figure 2. Evolution of the geometrics in the Mesolithic-Neolithic transition of Forcas II. are human remains, some of which are among the ashes of a circular structure and inside a crack closed with stones, next to a wall, and which are related to a single burial time around 4430 ± 40 BP. These remains were accompanied by pottery fragments, including some examples with elongated and parallel mammae-like lumps (“Veraza” style), similar to others that appeared in the Early Bronze age levels at the nearby cave of Moro de Olvena. The limit between level 7 and the neighbouring level 6 is dated at 3920 ± 30. There are few decorations and the Beaker decorations are included in the Pyrenean group, contemporary in a large part of its development with Ciempozuelos or Salamó (phase II of Aragon), although they could also be placed in phase III, late Beaker, which would coincide with the Pyrenean barbelé type, contemporary to the Tarragon group of Arbolí, as is the case with the cave at Moro de Olvena. 195
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Figure 2. Evolution of the geometrics in the Mesolithic-Neolithic transition of Fo...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 196 Jordi Rosell *,**, Xose Pedro Rodríguez *,** Ruth Blasco***, Edgard Camarós *,**, Maite Arilla *,**, Andrea Picin *,**,****, Eneko Iriarte ***** Nerets and La Cova de les Llenes archaeological sites (Pallars Jussà, Lleida) Nerets and Cova de les Llenes archaeological sites are located in the Pre-Pyrenean area of the Pallars Jussà (Lleida). The current configuration of this area rises from the Alpine Orogeny which formed an important fold of cretaceous materials resulting in two anticlines, one in the south represented by Serra del Montsec (1700 m above sea level) and another in the north, with the Serres de Sant Gervàs and Boumort (2200 m above sea level).From the geomorphologic point of view, the syncline forms a sedimentary basin known as Conca de Tremp (Conca de Baix and Conca de Dalt), where the tertiary landings show several Mesozoic outcrops. The whole area is crossed by an important hydrographic network from north to south whose two main rivers are the Noguera Pallaresa and its affluent the Flamisell. Both rivers rise from glacier valleys of the Axial Pyrenees. The Nerets archaeological site is located in the central part of this basin, specifically in the Mesozoic sandstones of the Arén Formation located to the east of the town, Talarn. This site was discovered by chance, in 1989, by a local amateur thanks to the discovery of surface archaeological material (Rosell and Rodríguez, 1991; Rodríguez and Rosell, 1993). Nerets hill has a height of 625 meters above sea level, with good visibility over the river Noguera Pallaresa. In 1995 an archaeological intervention was carried out, directed by Jordi Rosell. The aim was to conduct a systematic recognition and survey of areas with greater concentrations of lithic materials on the surface. An excavation took place in the lower part of the hill, affecting 6 m2 initially, later expanding to 16 m2. During this intervention, some lithic industry was recovered but no faunal remains were located. The stratigraphic sequence described during the excavation of 1995 was formed by a pack of conglomerates of large heterometrical pebbles * ** *** **** ***** and very rounded polygenic with a matrix of slimes and fine sands from the base to the roof (Fig. 1). Most of the lithic industry was found in this level. A higher metric pack of fine to medium-sized sands with clays was also identified. This layer also provided lithic industry. At some points, this sequence was covered by a carbonate crust dating to U/Th en >75 ky. The current ground was placed on a higher level. The total described sequence was 60cm high. The 1009 lithic objects recovered at Nerets come from the non-systematic survey of 1989 and mainly from the systematic survey and the excavation of 1995 (Rodriguez, 2004) (Table 1). All the material displays great consistency. In this assemblage, sequences of systematic production of flakes and of tool configuration have been identified. The shortage of debitage remains could be due to the fact that most part of the recovered material was recovered on the surface. Among the raw materials, a clear predominance of quartzite is identified –almost 80%– (Tab. 1). Also noted is the use of hornfels, sandstone, and quartz. The rest of the raw materials do not even reach 1%. The presence of some knapping flint objects could be the result of a different dynamic from the rest, as the features of these objects are quite different from the other items. The most used raw materials appear in the present riverbed of the river Noguera Pallaresa or in some of its ancient terraces which are very close to the archaeological site. A differential management of the raw material has been observed: the quartzite is used in both processes of production and instruments configuration; on the other hand, the hornfels is mainly used for shaping large-size instruments of pebble (Fig. 1.1). This rock is used in exploitation processes as it does not offer skills as good as the quartzite. Àrea de Prehistòria, Universitat Rovira i Virgili (URV), Avinguda de Catalunya 35, 43002 Tarragona, España. IPHES, Institut Català de Palaeoecologia Humana i Evolució Social, C/ Marcel·lí Domingo s/n (Edifici W3), 43007 Tarragona, España. The Gibraltar Museum, 18–20 Bomb House Lane, Gibraltar. Neanderthal Museum, Talstrasse 300, 40822 Mettmann, Alemania. Departamento de Ciencias Históricas y Geografía, Universidad de Burgos, Villadiego, s/n, 09001 Burgos, España.
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  196  Jord...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Bn BN1G BN1GC Quartzite 52 Hornfels 17 6,5% 41 BN1GE 5,2% 53 6,7% BP BN2G BN1G Ind. BN2GC 3 0,4% 424 53,3% 67 FRAGS INDET TOTAL BN2GE 8,4% 5 0,6% 146 18,4% 4 0,5% 795 78,79% 18,7% 14 15,4% 3 3,3% 2 2,2% 24 26,4% 5 5,5% 0 0% 22 24,2% 4 4,4% 91 9,02% Sandstone 14 28,6% 9 18,4% 1 2% 1 2% 9 18,4% 2 4,1% 0 0% 12 24,5% 1 2% 49 4,86% Quartz 6 23,1% 0 0% 2 7,7% 0 0% 8 30,8% 0 0% 0 0% 9 34,6% 1 3,8% 26 2,58% Schist 0 0% 0 0% 0 0% 0 0% 3 42,9% 0 0% 0 0% 4 57,1% 0 0% 7 0,69% Limestone 1 20% 1 20% 0 0% 0 0% 2 40% 0 0% 0 0% 0 0% 1 20% 5 0,50% Slate 1 12,5% 0 0% 0 0% 0 0% 4 50% 0 0% 0 0% 3 37,5% 0 0% 8 0,79% Porphyry 2 100% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0% 0 0% 2 0,20% Flint 0 0% 0 0% 1 9,1% 0 0% 4 36,4% 4 36,4% 0 0% 2 18,2% 0 0% 11 1,09% 2 13,3% 1 6,7% 0 0% 3 0 0% 4 26,7% 2 13,3% 15 1,49% Indet 3 Total 96 20% 9,5% 67 6,6% 61 6,0% 6 0,6% 481 20% 47,7% 78 0% 0 7,7% 5 0,5% 202 20,0% 13 1,3% 1009 Table 1. Raw materials and structural categories of objects in Nerets archaeological site (Rodríguez, 2004). Nb=Natural bases (pebbles, cobbles or blocks selected in order to flake them or use them as hammers); NB1G= Negative Bases of first Generation; NB1gc= Negative Bases of first Generation of Configuration (tools on pebble); NB1GE=Negative Bases of first Generation of Exploitation (cores on pebble); PB= Positive Bases (flakes); NB2G= Negative Bases of second Generation; NB2GC= Negative Bases of second Generation of Configuration (retouched flakes); NB2GE= Negative Bases of second Generation of Exploitation (cores on flakes); Frags= Fragments; Indet.= Not determinable objects. The most common exploitation strategy to produce flakes consisted of a bifacial-centripetal knapping prioritizing one of the sides (flaked side) over the other (the one of preparation), with the aim of pre-setting the final morphology of the products (Levallois method) (Fig. 1.3, 1.5). Bifacial-centripetal cores were also recovered without predetermination. The resulting flakes of these centripetal strategies have also been located in the archaeological record (Fig. 1.4, 1.6). Operational Themes, exploiting transverse planes of thick pebbles by using extractions based on horizontal planes, are also implemented. To carry out this kind of process, quartzite is used mainly when the method chosen requires a predetermination of the final product. The objects produced are medium-format flakes with dihedralshaped or trihedral-shaped edges. In the most representative exploitation strategies, all the Operational Units involved in the production processes are present. The existence of cortical objects and flakes to prepare the exploitation of cores show this. On the other hand, cores which were left in different stages of exploitation, from the initial stages to almost exhausted cores, have been recovered (Fig. 1.5). All of these features suggest that the exploitation was developed in the archaeological site. smaller pieces). In the pebble tools, dihedral edges in the distal-transversal area are usually configured, with an upright and/or convex delineation (choppers and chopping tools) (Fig. 1.2). The configuration of lateral-transversal dihedrals and trihedrals are also important (pick) (Fig. 1.1). Transversal dihedrals are the largest part among the retouched flakes. In these kinds of instruments, the second aim is the configuration of lateral dihedrals followed by the configuration of trihedrals and denticulate. From the typological point of view, the side scrapers are predominant among the retouched flakes, with 13 elements (Fig. 1.7), followed by denticulates (n=11) and isolated notches (n=5). There are five featureless abrupts: two denticulates and three continuous. Three end scrapers, two truncates and one burin have also been found. It does not seem to be a selection of a specific type of flake to be configured. In fact, both cortical nonfaceted products (of quartzite and hornfels) and bi-faceted or multi-faceted platforms (of quartzite) with an established morphology are configured. It seems to be a type of selection related to the size of the retouched blanks: the average sizes of retouched flakes are 24 mm longer than the non-retouched ones. The configured items (pebbles and flakes) are 18.3% of Nerets lithic industry (except for the If we add the effectives of instruments of pebble and of flake, we observe that in 68% of the items 197
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Bn  BN1G BN1GC  Quartzite 52 Hornfels  17  6,5  41  BN1GE  5,2  53 6,7   BP  BN2G ...
198 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. Nerets stratigraphic and lithics (Rodriguez, 2004) (1-9) and Cova de les Llenes. (10): 1, Quartzite handaxe (on flake):2, retouched flake with a dihedral transversal edge (cleaver); 3, Quartzite flake (Levallois); 4, Bifacial Core of quartzite (Levallois); 5, Flake of quartzite (Levallois); 6, centripetal-bifacial core of quartzite in the final exploitation stage (Levallois); 7, unifacial-angular of hornfels with distal trihedral (pick); 8, Unifacial pebble of quartzite (chopper); 9, Quartzite flake with a side scraper retouch (on the left side) and a notch (on the right side); 10, Quartzite handaxe.
198  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES dihedral edges of upright or convex delineation have been configured. Adding concave dihedrals (notches), the percentage of dihedral edges reaches 72%. The configuration of transversal dihedrals must be noted. Distal trihedrals have been configured in almost a fifth of all the items. Denticulate edges are not very large (they are present in 6.9% of the items). There are four objects classified as hand axes (3.4% of the items) (Fig. 1.8), and eight items (6.9%) fit into the cleaver morphological type (Fig. 1.9). The absence of faunal remains makes an interpretation of the archaeological site´s role impossible. Nevertheless, with the available lithic record we can suggest that Nerets, due to its easily accessible location on a hill controlling a narrow passage of the river Noguera Pallaresa, was a place usually visited by human groups to develop processes of production and configuration of items potentially usable for processing tasks of faunal resources. Unfortunately, there are hardly criteria, except from the morphotechnical, which assign this site to a determined chronology. According to these criteria, Nerets could be placed in an advanced stage of the Middle Pleistocene, with technology including pebble tools- with some operational standards characteristics of Mode 2 – and complex production strategies (Levallois method) (Rodríguez, 2004). Accordingly, we could place Nerets in a phase of transition from Mode 2 to Mode 3. Recently, similar deposits have been discovered in other basin locations. The industry recovered in these places presents important similarities with the industry discovered at Nerets, which shows an important flow of human groups in this area during the end of the Middle Pleistocene, which may be related to warm periods. Unfortunately, these kinds of deposits have not recovered any information about the climatic or ecological context in which human occupation was developed. This has made prospections move towards the anticlines, looking for caves that have allowed the preservation of other kind of registers beyond the lithic. But the erosion, caused by both the influence of glaciarism at the end of Pleistocene and a very active river network, has caused the previous sedimentary deposits to be conserved only in some caves with very particular features or in those placed in high points far from the rivers. One of these cavities is La Cova de les Llenes (Conca de Dalt). This cave is a 250 m long karstic tube. The current entrance is placed on a cliff of 180 m over the river Flamisell and through the Congost d’Erinyà. The height is about 750 m above sea level. The cave was archaeologically studied for the first time at the end of the 1940´s by Professor Juan Maluquer de Motes (1951), who conducted just one excavation campaign at the entrance of the cave looking for Neolithic materials. The only existing description of the stratigraphic sequence of the cave, which discusses a basal layer made of Pleistocene materials where remains of cave bears appear, is the fruit of this campaign. This description caused the cave to be visited by the current investigation team and, after carrying out a preliminary archaeological action in February 2013, resulted in the first excavation campaign in August 2013. This first campaign was focused on the entrance of the cave where the sample and the stratigraphy of Professor Maluquer de Motes were recovered, and an excavation of 25m2 was initiated. The stratigraphy shows a higher strata of anthropic origin –levels 1 to 6- with a high content of ash and charcoal related to industrial or pre- industrial cremation activities of an unspecified date. Below, level 7 is formed by yellowish-brown shales presenting a mixture of Pleistocene materials and ceramic elements from the Neolithic and the Bronze Age. Level 8 is the first intact figuring; it is formed by green to brownish shales with some blocks of sandstones from the breakage of walls and the roof. Many faunal remains have been recovered on level 8 as well as lithic items corresponding to the beginning of the Middle Paleolithic, with very similar features to the items from Nerets. The preservation of level 8 is related to the development of some stalagmite crusts which blocked the cave entrance. These crusts, today in dating process, were probably formed before the maximum development of the glaciers at the Pyrenees during the MIS 3. Their presence preserved the sediments until the beginning of the Holocene when they were dismantled due to the regression of the cave´s mouth, allowing the entrance of human groups during the Neolithic. Currently, remains of these crusts in the entrance walls can be seen, but at the end of the cavity they remain intact covering all the Pleistocene deposits. The lithic industry in level 8 is mainly made of quartzite and other metamorphic rocks from the Flamisell river and from the Paleogene conglomerate formations of the area (Fig. 1. 10). The flakes are the predominant elements and, together with the few recovered cores, show both Levallois and discoidal reduction sequences. The faunal spectrum is mainly composed of bear remains (Ursus spelaeus), hyenas (Crocuta sp.), tares (Hemitragus sp.), horses (Equus ferus) and 199
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  dihedral edges of upright or convex delineation have been con   gured. Adding conc...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 200 deers (Cervus elaphus). The relationship between the faunal remains and lithic industry is determined by the presence of some cutmarks as well as anthropogenic fractures. However, the most important activities developed in this cave seem to be related to carnivores. Pending geochronologic and paleo-ecological information, the faunal composition and the lithic industry seem to be consistent with the end of the Middle Pleistocene of the Iberian Peninsula. In conclusion, Nerets and La Cova de les Llenes are part of a regional project with the aim of studying Neanderthal groups from the Late Middle Pleistocene in an area placed at the gateway to the Pyrenees. The study of these kinds of archaeological sites is important to understand the capacities of these human communities and their development in marginal areas far from places with great Alfonso Alday* Acknowledgements This research is financed by the projects CGL2012-38434-C03-03, CGL2012-38358, CGLBOS-2012-34717 and HAR2010-18952-C02-01 of the Ministry of Science and Innovation of the Spanish Government. Ruth Blasco is a post-doctoral fellow of the Beatriu de Pinós-A program of the Generalitat de Catalunya, co-financed with the Marie Curie Actions, EU-FP7. Edgard Camarós is a pre-doctoral fellow FI in the Generalitat de Catalunya, co-financed with funds of the European Social Fund. We would like to thank Jordi Fábregas for the help provided with his comments and all the members of the “Tritons” team for the logistic support during the fieldwork. Martinarri rock shelter (Obécuri, Treviño) Martinarri is a south-facing rock shelter with a roof that rises almost 3m above the current floor, and a 15m long and 3-4 m deep area, followed by a large terrace where prehistoric archaeological work has detected prehistoric activity. It lies in a large basin, now covered by dense forest on sandy, hilly terrain where sandstone outcrops containing shelters are the exception. This suggests that the prehistoric communities which settled here had detailed knowledge and control of the zone, as they chose the shelter with the best conditions in the district. The relatively monotonous catchment area consists of gentle hills and gorges, 80% lying between 700 and 900 m asl, which does not match the usual landscape patterns at Meso-neolithic sites, a possible reason for the smaller potential of the stratigraphic and cultural sequence. The flint supply points were probably the chert outcrops in Loza and Treviño (20 km away), Urbasa (30 km) and the coastal Flysch (100 km). * genetic flows of the same period, which the Mediterranean corridor might be. The site is still under excavation, with four annual digs since 2008 directed by A. Alday in a small area which has nevertheless yielded a high density of archaeological material and defined the complete stratigraphic sequence. Stratigraphic sequence and archaeological content Five sedimentary layers have been identified in the shelter, supplemented by several more on the adjacent platform. Each unit is practically horizontal, and basically composed of the substrate sands. Postdeposition phenomena (erosion, gullies, landslides) have not affected the strata, laid down in continuity without sterile units. Several negative structures (mainly post holes) enter each layer (Fig. 1). Surface: A thin layer of humus and loose sands with seemingly recent carbonaceous stains and Área de Prehistoria Universidad del País Vasco (EHU/UPV). Tomás y Valiente s/n 01006 Vitoria-Gasteiz. a.alday@ehu.es.
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  200  deer...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Figure 1. Martinarri stratigraphic sections (costra: crust). a herbaceous layer covering the outer quadrats. Few archaeological materials, including faunal remains, knapping debris, the odd core, a various blade flakes, two endscrapers, retouched blades and an abrupt retouch. Level 100: 15cm deep, with a compacted sandy texture and weathered sandy sections broken off from the shelter. Fifteen small diameter circular/oval holes have been interpreted as the product of inserted stakes, possibly related to the prehistoric level where ceramic material was found on the adjacent terrace. One thousand lithic items have been recovered from this level, the majority knapping debris and blade flakes, along with two dozen backs, a dozen scrapers, notches and denticulates and several micro-triangles and a few segmentiforms. There are over 1000 faunal fragments. Culturally this layer is attributed to the Mesolithic microindustry of Sauveterrian inspiration. Level 101: Up to 23cm deep, with light brown soil enriched with material of increasing size with depth. Evidence of lit fires which reached high temperatures. Abundant prehistoric material: over 5,000 lithic items including 300 blade flakes and over 200 retouched items (half of them blades and backed points, four dozen endscrapers, two retouched blades and a series of abrupt 201
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Figure 1. Martinarri stratigraphic sections  costra  crust .  a herbaceous layer c...
202 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD retouches, burins, notches and denticulates). The assemblage also includes the presence of micro triangles and segments. There are at least 3,000 small faunal fragments. The level shows cultural similarities to the higher level, and is ascribed to the Mesolithic microindustry of Sauveterrian inspiration (Fig. 2). Level 102: 20cm deep, characterized by compacted sand and an increasing larger fraction (many clasts and a few blocks). Its excavation revealed two large sandstone slabs which were part of a hitherto undefined anthropic structure. This is E.U 4, with an oval area of dark soil associated with blocks, interpreted as a hearth. The level has an extraordinary archaeological record consisting of over 6,000 lithic items, the majority knapping debris, many cores and approx. 200 blades without retouch. Amongst the tools (n = 250), the majority are backs (approx. 200), followed by scrapers (n = 60), retouched blades, drills, burins, endscrapers, notches and abrupt retouches. Once again microlithic triangles were collected. The presence of sandstone slabs and cobbles (usually in horizontal association) is also significant. One atrophic deer canine used as a pendant figures amongst more than 4,000 faunal items. This level is ascribed culturally to the Mesolithic microindustry of Sauveterrian inspiration, with typometric and formal variations from the other assemblages. Level 103: This 15 cm deep layer is filled with dry, sandy sediment with no organic elements. Its grainy texture is either compact (a hard to dig breccia) or darker and lenticular, rich in archaeological material. The presence of blocks is irregular. The remarkable material record includes 3,300 lithic items including almost 100 blade flakes, 35 backed flakes, 12 endscrapers, and fewer abrupt retouches, sidescrapers and denticulates. 1,300 faunal fragments were recovered. This level is ascribed culturally to the Upper-late Magdalenian (Fig. 2). A test pit on the open air terrace revealed an even deeper stratigraphy exceeding one metre, which can be subdivided into six units. The most notable feature here is the ceramic material including decorated fragments in Unit B, possibly from the end of Metal Age. We have associated this episode with the above-mentioned posts on level 101. The lower levels maintain the features outlined for the interior of the shelter, with minor nuances. Level Code BP Date 101 Beta – 314962 340±30 102 GrA-46014 8455±45 103 GrA-45940 11890±50 Table 1. Radiochronological references for the Martinarri site. The dating for level 101, based on a charcoal specimen collected in one of the holes, does not match the cultural material recovered, but rather the action of possibly contemporary shepherds. Cultural overview Although fieldwork in Martinarri is still in progress, the material recovered to date shows the quality of the site due to both the richness of its contents and the chronological-cultural periods involved. Martinarri has confirmed an integrated exploitation of today’s Alava province at the end of the Pleistocene, which signalled the definitive ‘colonization’ of inland zones. Not far from this site are other roughly contemporary shelters: Atxoste to the north, and Montico de Charratu and Peña del Castillo to the west. Portugain, Kukuma, Socuevas, Berniollo and possibly Bardallo as well (the latter two open air sites, are a little further away in the same district). One of the features of this series of campsites is the lack of a bone industry, along with backed flakes and endscrapers as the lynchpins of the lithic component, leaving burins to one side, and the substrate –Portugain, linked to the exploitation of the Urbasa siliceous outcrop, diverges from this pattern. The industrial content is compatible with intense hunting. In fact, it is highly likely that the vast Alava plain– whose boundary walls contain other points in this chronology, and the valleys that flank it provided shelter for a wide range of large mammals with a biomass apt for consumption. Moreover, the diversity of occupied spaces and the exploitation of a variety of flint indicate detailed knowledge of this territory
202  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  reto...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Figure 2. Selection of prehistoric materials from Martinarri. 203
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Figure 2. Selection of prehistoric materials from Martinarri.  203
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 204 and interest in its comprehensive exploitation. The settlement dynamics coincide with a pattern found in the neighbouring Aragon, part of a late glacial process common to other parts of Europe, characterized by settlements in new territories from a base in traditional refuge zones. The upper levels reveal a techno-industrial line which has only been identified relatively recently in this area: microblade assemblages with a Sauveterre inspiration (Fig. 2). The appearance of micro triangles and subsequently segments, along with changes in the styles and dimensions of backed tools, justify this identification. By this stage, it was no longer unusual to find double backed tools, worked bases, items which include Alfonso Alday* fine apical retouches opposite the backs or even arched designs. A tendency towards microliths, more so amongst tips than blades, is another striking evolution between the Magdalenian level and those higher up. Martinarri is not an isolated case, since similar changes have been described at the Atxoste and Socuevas sites, which coexisted with other assemblages such as Mendandia and Las Orcillas with industries which seem to follow similar trends. Parco Cave, at the other extreme of the Ebro River basin, a long series of north Pyrenean sites and Cantabrian assemblages such as Ekain must all be taken into account in assessments and interpretations of this collection from Martinarri. Mendandia (Sáseta, Treviño) The roof of this east-facing rock shelter covers roughly 52 m2. Alongside there is a 385 m2 platform on a steep slope overlooking the Ayuda River 40/50m below, at a distance of roughly 100m. Its strategic position provides commanding views along the river gorge and also immediate access to the midaltitude pastures, with a range of local landscapes from valleys, gorges, plateaus and grasslands to forests and abundant wildlife resources. Level V: Surveyed in a 70cm cut, which only showed prehistoric interest at the top. The malleable clayey soil has an orange colour, gradually lightening. Small coarse fraction. The record includes 920 fauna fragments and 196 lithic items, including 6 retouched endscrapers, 1 awl, 2 denticulates, 1 abrupt retouch and 1 sidescraper. One perforated Nassa reticulata was also found. This level is ascribed culturally to the laminar Mesolithic technology. Excavated between 1992-1995 and in 1997 by A. Alday in a 13 m2 area, the site yielded a vast range of material which was classified into five sedimentological divisions and six industrial sections. Level IV: More than 40cm deep in some areas, with a slight dip from west to east. Dark brown with blackish tones, compact silty structure with little large fraction, a wet and greasy aspect, small isolated puddles and breccia. Colonies of helix and frequent presence of charcoal and fire are noteworthy aspects. 47,579 bone fragments were inventoried. The lithic industry includes 11,284 items, with 94 cores and 354 retouched objects (35 endscrapers, 58 awls, 9 burins, 8 abrupt flakes, 2 composite items, 3 backed bladelets, 23 short notches, 139 short denticulates, 4 notches on blades, 1 denticulate on irregular flake, 4 serrated items, 6 abrupt re- Stratigraphic sequence This is a continuous sequence –with no erosive or infertile phases– of eminently human origin and gradual changes in the texture, tone and composition of the sediment. It spans several Mesolithic and Neolithic periods between 8500 and 6400 BP (Fig. 1 and 2). * Área de Prehistoria Universidad del País Vasco (EHU/UPV). Tomás y Valiente s/n 01006 Vitoria-Gasteiz. a.alday@ehu.es.
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  204  and ...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Figure 1. Mendandia stratigraphic section. touches, 4 geometric items, 1 microburin and 57 miscellaneous items). The bone record contains few items, which includes two nasarids. This assemblage is culturally ascribed to the Mesolithic notch and denticulate industry. Level III: A continuous 25cm deep horizon containing 2 cultural entities: dry, silty structure, fine grained, greyish colour. Colonies of terrestrial molluscs are common, and there is clear evidence of fire. 15,562 bone fragments were found in the lower part of this level (III-inf). The lithic industry consists of 3,869 items, with 50 cores and 237 retouched objects (20 endscrapers, 15 awls, 3 abrupt flakes, 11 backed bladelets, 11 short notches, 56 short denticulates, 9 notches on flake, 3 denticulates on flake, 2 abrupt retouches, 33 geometrics, 20 microburins and 54 miscellaneous items). Adornments include atrophied deer canines, Nasa, Natica catena and Cypraea. Culturally ascribed to the geometric Mesolithic. The upper part of level III (III-sup) contained 12,518 bone fragments. The lithic industry included 1,282 items with 106 retouched objects (13 end- scrapers, 6 awls, 1 abrupt on flake, 22 backed bladelets, 4 short notches, 11 short denticulates, 2 notches on flake, 1 denticulate on blade flake, 2 serrated edges with abrupt retouch, 4 abrupt retouches on blade, 18 geometrics, 6 microburins and 16 miscellaneous items). 343 ceramic fragments were counted, with lines and incisions used in decoration. Culturally ascribed to the Early Neolithic. Level II: A 20 cm deep homogeneous level with a brown colour, silty and plastic structure, some clasts, and 4,766 bone fragments. The lithic industry includes 953 items, including 75 retouched objects (5 endscrapers, 3 awls, 10 backed bladelets, 1 notch on flake 1 denticulate on flake, 3 notches on blade, 1 denticulate on blade, 1 serrated edge, 3 abrupt retouches, 21 geometrics, 6 microburins and 19 sidescrapers). There were 794 ceramic sherds, several of them decorated with impressions below the lip, finger-drawn lines and ungulations. This level is culturally ascribed to the Early Neolithic. Level I is the current floor, from 10 to 20cm deep. Brown-grey matrix, malleable, silty composition, initially dry and dusty, then compacted, 205
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Figure 1. Mendandia stratigraphic section.  touches, 4 geometric items, 1 microbur...
206 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD moist and more granulous when excavated. Abundant small clasts, few cobbles and 1,044 bone fragments. The lithic industry consists of 182 items with only 8 retouched objects (two endscrapers, one awl, one denticulate on flake, 2 denticulate on blade, 1 abrupt flake and 1 segment). 33 undecorated ceramic shards were found. This level is culturally ascribed to the Early Neolithic. Level Code BP Date I GrN-22740 6440±40 II GrN-22741 6540±70 GrN-22742 7180±45 GrN-19658 7210±80 Ua-34366 7265±60 GrN-22743 7620±50 GrN-22745 7780±40 GrN-22744 7810±50 GrA-6874 8500±60 III-upper III-lower IV V Table 1. Radiochronological references for the Mendandia site. Lithic and ceramic industry The density of the lithic industry and its coordination with the stratigraphic sequence have played an important role in resolving the evolution of the population at this site. The typological groups have a relatively balanced composition at each level: only level IV shows a rupture between denticulates and scrapers, and between these two elements, miscellaneous and scrapers, and the other categories (Fig. 2). The upper horizons (I to III-inf), characterized by the geometric basis of their industries, show a sequential evolution: double bevelled segments characterize the three most recent episodes –Neolithic–, while triangles and abrupt trapezoids represent the oldest –Mesolithic– period. It is interesting to note the presence of occlusal forms, which individualize the geometrism of the upper Ebro River basin in comparison to other areas of the same basin and the Cantabrian coast, as an example of a personal stylistic development. Level IV has yielded an extraordinary amount of lithic industry, with a proliferation of notches, denticulates and awls on flake and fragments, flake retouch used in the manufacturing process. Although this industry seems to be quite rough, in fact it was well thought out and organized to shape tools with predetermined faces, suitable for woodworking, as proven in traceological analyses. Its techno-typological and chronological concordance with other Iberian sites lends considerable content to the Mesolithic notch and denticulate industry, for which the Mendandia phase is major point of reference. The small amount of material on level V consists of endscrapers and backed blades, indicating its affiliation to the laminar Mesolithic industry. The inhabitants of Mendandia collected flint from outcrops in Loza and Treviño, some 15 km away, Urbasa (35 km away), Flysch (80 km) and the Ebro Evaporite (100 km). Most of the ceramic material is from indeterminate parts of recipients (92.5%) including edges (n = 83), a few handles and lids. Originally they were simple forms such as bols, a few in a closed “S” shape. The decoration shows evidence of technical evolution: the motifs are on edges or lips, with incision used in the earliest periods and imprints in the most recent. The three C14 dates for the upper level III show the surprising antiquity of this record. Perhaps for this reason the assemblage is difficult to identify stylistically, while the ceramic material from higher levels concords well with early Neolithic imprinted ceramics (Table 1). Lifestyles Hunting was one of the most privileged activities in Mendandia. The strategic location of the shelter facilitated the groups’ capture of a wide variety of species, 90% of which were roe deer, deer and aurochs, followed by wild boar, goat, horse and chamois, and an anecdotal presence of fox, wolf, marten, badger, bobcat, rabbit and hare. The estimated age of the prey shows that hunting was mainly practiced in late spring and early summer. The proportion of each anatomical part, supplemented with anthracological data, suggests that certain items were smoked for their (presumed) transfer to another campsite. A genetic study has suggested that some of the old Neolithic bovids may have been domesticated. Altuna and Mariezkurrena also suggest that the age and sex spectrum of the faunal assemblage in this
206  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  mois...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Figure 2. Selection of prehistoric materials in Mendandia. 207
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Figure 2. Selection of prehistoric materials in Mendandia.  207
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 208 herd may evidence quasi-domestication. Palynological data contained the excavation report show that the Neolithic landscape was compatible with agriculture, which would concur with the identification of flint blades used to cut grain. Pilar García-Argüelles* Jordi Nadal* Josep Mª Fullola* The Montsant Valley (Localities of Valle del Montsant Priorat, Tarragona), A key core region for Prehistory in the NE of Iberian Peninsula In recent years, research work has been carried out as part of programme SGR2014-108 of the Generalitat de Cataluña and programme HAR2011-26193 of the MINECO 1. Presentation The middle course valley of the Montsant River, an affluent of the Siurana, the final tributary of the Ebro before its connection to the sea, contains a concentration of prehistoric sites, specifically from the Late Upper Palaeolithic and Epipalaeolithic, that makes it a benchmark area in the Iberian Peninsula when researching these two phases of prehistory. This area is located in the district of Priorato, in the province of Tarragona (Catalonia, Spain), in the NE of the Iberian Peninsula (Fig. 1). The presence of material evidence of the prehistoric population has been known since the 1930s, when Salvador Vilaseca prospected the area and discovered different many flint artefacts from the El Filador and Els Colls shelters. His work in this area, practically all of which was in the municipality of Margalef de Montsant, continued through to the end of the 1960s, with special emphasis on the aforementioned shelter, El Filador, where he carried out different excavation campaigns until 1968. His numerous publications (Vilaseca 1936; 1949; 1953; 1960; 1968; 1973) demonstrated the vital importance of the sector studied for Epipal* Traceology has revealed a wide range of activities performed in this shelter, from butchering, leather and woodwork –linked to the numerous fires–, and flint knapping to the use of materials for colouring. aeolithic times. The work by Javier Fortea in the early 1970s, re-examining the materials from El Filador for his PhD, elevated the site to a preeminent position within the cultural and material evolution of the Epipalaeolithic in the Iberian Peninsula (Fortea, 1973). Starting in 1978, the University of Barcelona began an excavation programme in the middle course of the Montsant. The programme began with the re-excavation of El Filador (1979-1997), and the systematic prospecting in the area soon produced results. Excavation was done at three other sites, the Els Colls shelter (1982-1991), the Boix cave (19831984) and the L’Hort de la Boquera shelter (since 1998); other surface settlements were located, such as L’Hort d’en Marquet or El Planot (the latter with Mousterian materials, located in the highest terrace of the river), and other settlements already mentioned by Vilaseca, and with very positive potential for the future, were surveyed, such as the cova de la Jaia or Tormos d’en Celoni, among many others. In addition to this, in 1981, the only known example of cave art with an engraved Palaeolithic figure in the NE of the peninsula was discovered; this was a figure of a deer, found in the inner galleries of the cova de la Taverna (Fullola and Viñas, 1985; Fig 4.2), and while initially this appeared out of place, it is much more coherent in light of the Late Palaeolithic chronological context of the sites in the Montsant Valley that we present in this paper. In regard to a different matter, the construction of a dam required an urgent excavation in Adscripción de los tres autores: SERP (Seminari d’Estudis i Recerques Prehistòriques) de la Universidad de Barcelona, área de Prehistoria, Departamento de Prehistoria, Historia Antigua y Arqueología, Facultad de Geografía e Historia, Universidad de Barcelona; calle Montalegre, 6, E-08001, Barcelona. Correspondent author: garciaarguelles@ub.edu
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  208  herd...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Figure 1. Location of the Montsant Valley, in the NE of the Iberian Peninsula. the L’Auferí shelter, which was carried out by archaeologists from our group in the early 1990s. All of these excavations clearly demonstrated the richness of the area; in consequence, since 1978, the University of Barcelona, first with research programmes directed by one of us (JMF) and later, since 1986, through the SERP (Seminari d’Estudis i Recerques Prehistòriques), and with different co-directors of the different excavations, with special dedication by the other two authors of this paper (PG-A and JN), has been working continuously in the middle sector of the Montsant, in the municipality of Margalef de Montsant. 2. Geoarchaeological information The Montsant Valley gets its name from the Montsant Range, located to the SW of the Central Catalan Depression. Geographically, the Montsant Range is oriented NE to SW, with a total length of 19 km, and approximately 30 km from the Mediterranean coast. Geologically, Oligocene conglomerates predominate, reaching thicknesses of up to 300 m, with a sandy matrix and calcareous cement. Alternating with the conglomerates are layers of red clays, gypsum, and flint nodules, which erode much more easily and create rounded shapes, which served as shelters under which the Prehistoric peoples lived. In the area between Margalef de Montsant and Bisbal de Falset, the river loses a significant part of its erosion capacity, which has protected the different archaeological sites in the area. Dr. Bergadà (Bergadà, 1998) has proposed an evolutionary and chronological reconstruction of the sedimentary sequence of the middle course of the Montsant River. Prior to 10.950BP Phase 1: alluvial deposit with significant intensity, with several m of gravel and sand, located 209
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Figure 1. Location of the Montsant Valley, in the NE of the Iberian Peninsula.  th...
210 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD mainly on the concave bank of the river. Probably occurred during the Upper Pleistocene, which would correspond to terrace level T2. Level IV of L’Hort de la Boquera and level IX of the Els Colls shelter. Phase 2: period of lower alluvial intensity, with sand deposits. Level VII of the Els Colls shelter and level II of L’Hort de la Boquera. Phase 3: time of flooding with local additions (falling blocks), level IV and V of the Els Colls shelter. From 10,950– 10,050 BP Phase 4: flat flood area (bioturbated sandy silts) with local additions, falling blocks from the ledges of the Els Colls shelter (levels IV, III and II) and from L’Hort de la Boquera (level II), with local runoff contributions. 3a. El Filador The El Filador site is located in the municipality of Margalef de Montsant, to the north of the Priorato district, in the province of Tarragona (E(X):311907.8m-N(Y)4572589.3m UTM 31N/ ETR S89). It is a large shelter approximately 100 m long, although the archaeological site is located in the central part. It is located 15 m above the current level of the Montsant River, on the left bank, approximately 340 masl, just opposite the town of Margalef de Montsant. The intermittent news of El Filador given by Salvador Vilaseca were based on prospecting studies prior to the Civil War and in short campaigns during the 1950s and 1960s (Vilaseca 1936; 1949; 1953; 1968; 1973). Since 1979, the University of Barcelona has been in charge of the excavations, (Fullola and García-Argüelles, 1980), directed through the SERP from 1986 to 1997. The environmental conditions were wet and cold, based on the data from the Els Colls shelter, while L’Hort de la Boquera had less moist conditions. Runoff processes began to operate, level III and IV of L’Hort de la Boquera, and the slope deposits began to form in a semi-arid environment. Our stratigraphic survey (García-Argüelles et al., 2005; Fig 2) shows 11 distinct levels, characterised as follows: From 10,050– 9,000 BP Level 2: 25 cm thick, and also very localized in the northwest zone, where it covers 11 m2. On the whole, ashen grey, with characteristics of intense combustion, and consequently, significant alteration of its composition. This corresponds to the first archaeological level. Phase 5: the river began to carve its path (T1b) with energetic sedimentation of gravel and sand; level XII of the El Filador shelter. Phase 6: lower intensity, levels XI, X, and VIIIIX of the El Filador shelter. Phase 7: flat flood area (bioturbated sandy silt) of this terrace level. This would correspond to the rest of the sequence studied in the El Filador shelter. The runoff processes triggered the creation of alluvial cones. After 9,000BP Phase 8: the river carves its path and forms terrace T1a. 3. Archaeological sites In this section, we will summarise the principal sites in which we have excavated in the area of the middle course of the Montsant River since 1978 and present the most significant findings. Level 1: Thickness of approximately 15 cm; consisting of disturbed earth, with the incorporation of modern materials. In fact, traces of the level remained in the northwest zone only. Level 3: Very thin, with a thickness of approximately 15 cm, and located in the NW sector of the shelter, covering just 8 m2. Level T: This torrential addition overlaid level 4 and covered the entire surface area of the site. This addition came from a lateral flow. Its thickness decreases from one metre, in the SE sector, until it disappears on the opposite side. It also included materials belonging to level 4 in its matrix of pebbles and gravel. Level 4: This level already appeared throughout the entire extension of the site, 20 m2, and in some sectors presented interspersing of torrential additions, mainly in the SE sector. Thickness approximately 25 cm. Level 5-6: This is one of the double levels based on the sedimentological study, but from the archaeological point of view, it was impossible to differen-
210  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  main...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES tiate it. Thickness between 10 and 20 cm; covered 8 m2 and was also concentrated in the NW sector. Level 7: This is one of the thickest levels, 41 cm, and 27 m2 excavated. In the SE zone, this level was connected directly with level 4, and since they had the same composition, it was very difficult to distinguish between them. Level 8-9: Had the same characteristics as level 5-6 and a thickness of 76 cm, although just 28 cm correspond to the archaeological level; the rest was made up of flood sand. Levels 10 and 11, sterile from the archaeological point of view, and that reached the underlying terrace of the Montsant River, T1. Figure 2: Stratigraphic diagram of the El Filador shelter, and cross-sections at metres 4 and 10 (from García-Argüelles et al., 2005: 69). 211
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  tiate it. Thickness between 10 and 20 cm  covered 8 m2 and was also concentrated i...
212 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Culturally, El Filador presents a continuous sequence from microlaminar phases, which today we tend to assimilate into a transition with the end of the Upper Palaeolithic, until the moment of Notches and Denticulates, with some ceramic remains, at the end of the occupation. The lower levels, 8-9, are the ones that appear to correspond to the microlaminar phases, with radiocarbon dating ranging from 11,000 to 10,880 BP uncalibrated; this indicates a range between the twelfth and thirteenth millennium calBC. Backed blades (43%) and backed points (14.4%) predominate. These are followed by burins (16.7%), denticulates (5.2%), and burins (2%). There are also two large non-delineated structures of soil rubificated by fire, as well as different anthropically-contributed stone blocks and two knapping zones (Fig 3.2). Levels 7, 4 and 3 form the largest part of the geometric Epipalaeolithic package. Abrupt retouching predominates, accounting for more than 80%, with few scrapers and even fewer burins. Micro-burins, a clear indicator of the geometric production, account for between 30 and 45% of the retouched elements. Segments of circles and triangles are the geometric forms present; the former predominates over the latter in the initial moments, with this proportion re- versing in level 3 (fig 3.3). There are no trapezoids. In level 7, four sandstone polishers intended for producing arrow shafts were found. There are various pebbles with traces of red paint; in one case, in level 4, up to six red lines were clearly visible (fig 4.3); in another, from level 7, a red stripe covered the entire perimeter; and the other examples were entirely covered in ochre, as if they had been submerged in it. An anvil stone for knapping made up of two stones appeared in level 7, in relation to two combustion structures. Other structures on level 4, in addition to ashes and rubifacted earth at the base, presented a fill which included hundreds of shells from Cepaea nemoralis. On this same level, three slate plates were found, intentionally cut in a double bevel; two of these included a palimpsest of finely engraved lines, with traces of ochre in the grooves; the most plausible interpretation is that they were used as supports for cutting soft materials such as skins, which had been prepared with ochre. Level 4 also produced a bone punch, with an oval cross-section, made of a bovine metatarsal, an exceptional case of conservation in soil that has very negatively affected materials of animal origin. The datings obtained for this set of levels are centred on the 10th millennium BP uncalibrated. Figure 3. 1 –Materials from the Late Upper Palaeolithic of L’Hort de la Boquera, 2– Microlaminar materials from the Epipalaeolithic, levels 8-9 of El Filador, 3– Materials from the geometric Epipalaeolithic, levels 4-7 of El Filador.
212  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Cult...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES It is also important to highlight the presence of two conical shaped basins filled with flint, twelve fragments of ceramic that is difficult to attribute, small fragments of bone, and charcoal. These basins were obviously excavated during a more recent time. Two carbon datings place this in the second half of the ninth millennium BP uncalibrated. From the El Filador sequence, we can infer a series of general reflections, which are described below. Figure 4: 1– Tracking of naturalist figures engraved in a calcareous block, from the Late Upper Palaeolithic from L’Hort de la Boquera, 2– Tracing of the Palaeolithic engraving of a deer in the Cova de la Taverna, 3– Painted pebble from level 4, geometric Epipalaeolithic, from El Filador (58 x41 x 18 mm) Levels 5-6 are interspersed between level 7 and level 4 in one part of the site, in an area of just 10 m2, in the NW sector of the shelter. Their origins indicate slow, periodic flooding, based on finer granulometry than in levels 7, 4, and 3. The occupations were, therefore much more sporadic and localized in one part of the site. Culturally, however, the representation of geometric elements continues to be significant (12.5% of segments and 7.5% of triangles), along with 42.5% micro-burins; backed bladelets and backed point accounted for 15%, and 12.5% endscrapers; burins are still virtually non-existent. Dated at 9,988±97 BP uncalibrated. Level 2 of the El Filador shelter covered only 11 m2 and was located in the NW section of the site. Unlike the rest of the sedimentary package, it was grey, as a result of intense combustion. Industrially, the high percentage of denticulates, especially notches and épines (36%) and side scrapers (13.3%), along with an increase in burins (11%) and the spectacular drop in abrupt retouching (barely 3.9%), means that this level 2 of El Filador enters into the Notches and Denticulates phase recently defined in the Ebro Valley (Alday 2006). The end of the Upper Palaeolithic culture chronologically “merged” with the microlaminar complex and makes it difficult to distinguish the technological identity of some levels or sites through the record. Some settlements are attributed to the Late Upper Palaeolithic or the microlaminar Epipalaeolithic, depending on the subjective criteria of the researchers, due to the lack of definition of the industry recovered and the vagueness of its chronology. In theory, this does not affect the records of El Filador itself, but does affect other nearby sites, such as Els Colls, which we will see later in this paper, with chronologies that immediately precede or coincide with it. Emphasising that it is difficult to distinguish the oldest microlaminar sites from those of the Late Upper Palaeolithic, the microlaminar technical complex probably developed fully (with possible earlier roots) in the 12th millennium BP and would reach its floruit in the 11th millennium BP (depending, we repeat, on the number of absolute datings) to languish throughout the 10th millennium BP, a phase in which there are few datings. The transition from the microlaminar to geometric complex should not be understood as a break, because the appearance of geometric elements does not mean the disappearance of the microlaminar component. On the contrary, we see that the backed elements continue to be very important (24% in level 7, and 24.2% in level 4). And in fact, according to the dating table, the geometric complex began at the same time as the maximum expansion of the microlaminar complex, in the 11th millennium BP. This could lead us to think that the technological innovation represented by the appearance of Epipalaeolithic geometrism would be used initially on certain occasions, and in certain circumstances, or to carry out specific functions, while for others, the preference would be to continue with the backed elements of the microlaminar complex. Finally, the presence of the geometric component was ultimately super- 213
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  It is also important to highlight the presence of two conical shaped basins    lle...
214 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD imposed on the microlaminar deposits with no geometric elements, at the time of the maximum expansion of the Filador facies – which technologically were Sauveterroid – in the 10th millennium BP (García-Argüelles and Nadal, 1998; GarcíaArgüelles et al., 2013). This, in turn, is presented in some sites, with a much lower frequency, until the 9th millennium BP. We need to contradict the assertion by other authors regarding the high percentage of denticulates in all levels of El Filador. This is true only in level 2, of Notches and Denticulates / Neolithic, and we would like to reassert the data: on level 8-9 we have 10 denticulate elements; on level 7, 16; on level 5-6, 1; on level 4, 9 and on level 3, 4 (5.8%). There is not a high percentage of denticulate tools, so El Filador cannot be used to talk about a new facies of Notches and Denticulates, located between the microlaminar and geometric, which other researchers have located in the Ebro river basin and the Central Plateau, at sites such as Forcas (Utrilla and Mazo, 1997), Mendandia, Atxose, Peña 14, or El Ángel (Alday, 2002). El Filador does not have the latest facies of the geometric Epipalaeolithic of the Fortea classification, which is characterised by the presence of trapezoids and the disappearance of the micro-burin technique. In our area, we have la Balma of la Margineda (level 4) and those of the Ebro Valley (Pontet, Botiquería, Forcas) According to the datings of Balma Margineda, its chronology would be the 9th millennium, coinciding with the most modern, and recent datings of the Filador facies – Sauveterroid. In addition to the lack of data for trapezoidal facies in Catalonia, the presence of a series of technical complexes that had not been attributed to the typology established by Fortea has been consolidated, and have been cited as sites with “atypical” industry. With an increasingly expanding record, this group has consolidated itself as the cultural model for the 9th millennium BP, which could explain the lack of sites with trapezoids. The chronologies of these sites extend to the 8th millennium BP. In any case, for the time being, the evolution detected in the Ebro Valley does not appear to be reproduced: Microlaminar Epipalaeolithic, Mesolithic with Notches and Denticulates, and Mesolithic geometric with trapezoids (Utrilla, 2002). The Catalonian record for the 8th millennium BP is still too poor. At El Filador, we are missing the last two facies which could clarify the transition to a Neolithic without datings and with the small quantity of ceramic of level 2. The southern zone of Catalonia still has a gap, as noted by Martí and Juan-Cabanilles (1997:237), despite the fact that there have already been significant new developments in the Pyrenees areas (Petit et al., 1996; Pallarès, Bordas and Mora 1997). The technological and economic model that we observed in El Filador was successful in the area for just over two millennia. After that, there is a chronological and cultural hiatus until the arrival of the Neolithic. Other sites in the area such as El Auferí (Adserias et al., 1996) or L’Hort de la Boquera corroborate this coherent development and an occupation of the territory that leads to a rational exploitation of the biotic and abiotic resources in one of the few areas in which these spatial distribution studies have been able to be carried out in a geographically limited area. 3b. L’Hort de la Boquera This is a small shelter, no more than 9 m long and 4 m deep, that has lost part of its ledge and that conserves an excavatable surface area of no more than 20 m2, half of which is below the conserved section of the ledge. It is oriented S-SE, is located on the right bank of the Montsant River, approximately 25 m above the current level, and approximately 400 masl (E(X):312108.8mN(Y)4573254.3m UTM 31N/ETR S89). Its stratigraphy is arranged on 4 levels, level I, which contains 2 sublevels (Ia and Ib) was formed by the processes of streams and falling blocks from the ledge. This level is archaeologically sterile. Level II is made up of a sandy silt matrix. It has a thickness of 47 cm, and is the only archaeological level at the site in which different habitation moments have been detected, as we will discuss later in this paper. The stratigraphy of the site is completed with level III, which is made up of fine sand that contains some flint remains, but that do not indicate stable human occupation, and level IV, made up mainly of pebbles and gravel, and which rests on terrace T2 of the Montsant Rivers, 24 m above the current level (Fullola, 1978; Bergadà, 1993:157165; García-Argüelles et al., i.p.). In regard to the human activity of level II, so far, more than 30,000 lithic elements have been recovered, mainly made of flint, but there is also a small number of elements made of slate and limestone.
214  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  impo...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES 2.15% of the elements have been retouched and endscrapers and backed elements predominate. Denticulates and truncated tools were found in smaller proportions. Burins play a smaller role, on fracture, on retouch, and dihedrals (Fig 3.1). There are also round slate plates with bevelled edges and markings that are currently being studied. In this sense, there is a large limestone block on which a figure of a bird, possibly a crane, has been engraved, surrounded by other clearer figures that can be interpreted as anthropomorphic (García-Argüelles et al., i.p.; Fig 4.1). It appears that the presence of portable figurative art begins to be constant in the sites of the Late Upper Palaeolithic in southern Catalonia, as in the case of Molí del Salt and Sant Gregori (Falset). Paleo-environmentally, the paleobotanical data is limited to the data provided by the anthracological studies conducted by Dr. Allué. There is a significant presence of Pinus sylvestris type and some elements of Juniperus sp.. In terms of fauna, despite the deficient conservation of bone, they cannot be said to be scarce. Only the presence of Capra pyrenaica and Oryctolagus cuniculus can be cited. The group is completed by different Cepaea nemoralis, which are abundant in El Filador. The gathering of terrestrial molluscs in the Catalonian area is very frequent in the Epipaleolithic sites, but not for Palaeolithic occupations. In this sense, L’Hort de la Boquera is one of the first indicators of the expansion of the dietary spectrum in the northeast of the Iberian Peninsula during the Pleistocene-Holocene transition. The datings of the phases of occupations range from 12,250±60 BP to 11,850±45 BP and 11,775±45 BP, which places it, along with the cultural characteristics that we just mentioned, in a phase that corresponds to the Late Upper Palaeolithic. (Mangado et al., 2010; Fullola et al., 2012). 3c. Els Colls This is a shelter which is more than 50 m long and between 2 and 8 m high, and up to 6 m deep in some sectors. It is oriented towards the SW and is located on the right bank of the Montsant River, 20.7 m above the current level, and approximately 400 masl (E(X):312108.8m-N(Y)4573254.3m UTM 31N/ETR S89). Sedimentalogically, 8 levels have been identified (Bergadà, 1993:186-191); level I was formed recently. Level II is the most archaeologically rich, with more than 9,000 lithic elements. This occupation is settled on fallen blocks of conglomerate (level III). During the excavation, we observed that the blocks of conglomerate were directly above level IV, whose lithic material presented, in many cases, fracturing of the pieces (Bergadà, 1998). This level IV was excavated in a smaller sector of the shelter, and therefore offered fewer lithic elements (1,500). Below level V, to the river terrace, the levels are sterile. Abrupt elements predominate in level II (backed bladelets and backed points). In terms of simple retouched tools, side scrapers are most prevalent, and there are also burins, which in many cases are double burins, and on truncated elements. Fauna is not very abundant and appears to be highly fragmented; most elements belong to large mammals and are splinters whose size indicates intense activity for the exploitation of hunting products (deer, wild goat, etc.). In many cases, the elements present evidence of combustion processes, connected with the appearance of a home on level II. There are traces of pollen, mainly from Pinus and Quercus t.ilex. Three absolute datings have been obtained, two by C14: 10,950±120 BP and 10,050±85 BP, and a third by thermoluminescence 13,000±1,000 BP. (Fullola et al., 1993). In the underlying level IV, we documented a predominance of simple retouching, endscrapers, and denticulates, followed by abrupt retouched tools such as backed bladelets and backed points. The study of the spatial distribution of the archaeological material has determined the existence of an important area of knapping and a combustion structure. There are no traces of fauna. The archaeological levels of the Els Colls shelter present a series of industrial characteristics that are chronologically and culturally homogeneous. The large dimensions of different backed elements could lead us to think of early phases of the Upper Palaeolithic, but the rest of the sites in the valley, the aforementioned datings, and other technological details clearly place the site in the Late Upper Palaeolithic (Rodríguez Baylach, unpublished). 3d. La Cova del Boix The cave is an opening in a cliff on the left bank of the Montsant River. It has a single chamber approximately 25 m wide by 15 m deep, open in the NE direction (E(X):313992.0m-N(Y)4573860.0 UTM 31/ETR S89) (García-Argüelles and Fullola, 2002). It was excavated in 1982/83. The stratigraphy is inverted due to the disturbances caused by the use of 215
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  2.15  of the elements have been retouched and endscrapers and backed elements pred...
216 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD No radiocarbon dating is available for this site, but its industrial characteristics place it closer to the cases mentioned before of Els Colls and L’Hort de la Boquera, which involves a chronologicalcultural attribution to the Late Upper Palaeolithic with the characteristic of a higher proportion of endscrapers with respect to burins (GarcíaArgüelles and Fullola, 2002). V corresponds to the terrace on which the site is settled, which consists mostly of river pebbles. More than 20,000 lithic elements were collected, almost all flint, of which 3.4% are retouched, with endscrapers predominating, followed by elements with abrupt retouching and finally, burins. Traces of fauna are very scarce and very poorly conserved; it can only be said that Capra pyrenaica, along with Cervus elaphus and Oryctolagus cuniculus predominate. There is one reliable radiocarbon dating, which also corresponds to the base of the sequence, level V, which is sterile. The date is 12,317±114 BP, and for the reasons mentioned regarding the location of the sample, it was not taken into consideration by the excavation directors (M.Adserias personal communication). Nevertheless, in our opinion, it gives a post quem date for the archaeological package of level III and is similar to the ones obtained at other sites in the area, with significant technotypological similarities, such as L’Hort de la Boquera and Els Colls. (Adserias et al., 1996). 3e. L’Abric de l’Auferí 3f. Other sites in the middle valley of the Montsant River This is the only site that was excavated as an emergency excavation due to the impact of the construction of the Margalef reservoir. It was excavated in the early 1990s under the direction of the members of our group Maria Adserias and Raül Bartolí. Once again, this is a shelter that opens in the southern direction, on the left bank of the River, approximately 22 m above the current level (E(X):315173.8m-N(Y)4574478.8m UTM 31N/ ETR S89) (Adserias and Bartrolí, 2007). Having presented the principal excavated sites, it is important to remember that, as mentioned in the introduction, that the research by S. Vilaseca and our own excavations have identified many other sites, whether caves, shelters, or outside, in the middle valley of the Montsant River. This makes this one of the areas with the highest concentration of archaeological stations in the northeast of the Iberian Peninsula. the space by carboneros (charcoal makers) who used the cave as a shelter. However, the recovered materials indicated that prehistoric occupation had existed. Just over one thousand lithic elements were recovered, of which 41 are retouched pieces; backed elements clearly predominate, with some endscrapers, which outnumber the burins. There are some traces of fauna, but it is difficult to distinguish between the old elements and the modern additions. Remains from wild animals such as Cervus elaphus or Capra pyrenaica which can be attributed to prehistoric additions have been recovered. In this excavation, two sectors were opened. In one, abundant ceramics were found in the first level, which was one of the few cases of Montsant sites with occupation with Neolithic chronology. The most interesting sector is the one described in sector II, which contains 5 levels. Level I corresponds to the surface level of sandy matrix, with a large quantity of gravel from 1 to 3 cm, and a significant organic component; it incorporates archaeological materials from the lower levels as a result of agricultural disturbances. Level II is located in the parts protected by the shelter ledge. Its matrix is made up of the degraded materials of the wall of the shelter and incorporates archaeological material from the previous level in the contact zone. Level III, divided into two sublevels, presents, in its lower section, the true archaeological level, with many lithic elements but with no ceramic. Level IV, with a silty matrix, only presented a few lithic elements incorporated from the upper level; and level L’ Hort d’en Marquet, located in 1980, is an area with an abundance of flint on the surface and which presents, in one of the areas in which lithic material has been found, stratigraphy that is currently very degraded due to human activities during historic times. In a second level of this sequence, with a thickness of approximately 1015 centimetres, a concentration of archaeological material was detected. This consists of retouched pieces that included scrapers, burins, and knapped dorsal elements. Without absolute dating, and with the data from the other excavated sites (Els Colls and L’Hort de la Boquera), it is very feasible that the site corresponds to the Late Upper Palaeolithic. (Fullola and García-Argüelles, 1980) . El Planot would be another one of these sites. This is a surface site, the only one located on the upper terrace of the river, more than 30 m above the current level, on the borders of the municipalities of Margalef de Montsant and La Bisbal de Falset.
216  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  No r...
EBRO VALLEY, PYRENEES AND PRE-PYRENEES Approximately 50 pieces were recovered, with side scrapers and denticulate tools predominating. For this reason, along with its location on the upper terrace, the site is supposed to correspond to a Mousterian chronology (Fullola and García-Argüelles, 1982-83). 4. Conclusion After describing some of the most important sites, we feel that the archaeological importance of the Montsant Valley has been clearly explained, especially in regard to the final phases of the hunter-gatherer communities of the Late Upper Palaeolithic and the Epipalaeolithic. In this sense, beyond this chronological margin, for now, we have just one site that could be attributed to the Middle Palaeolithic (El Planot) and two occupations that could be attributed to the Neolithic (on the upper levels of l’Auferí and El Filador). It is quite likely that the geographic and agricultural attractiveness of the area for hunter-gatherers was lost, at least in regard to agricultural potential. The majority of the Montsant sites are very homogeneous in regard to a series of characteristics, which include chronology, when reliable datings are available. Most of them are occupations in shelters located on the second terrace level of the river, at heights above the current level that range from 20 to 25 metres. They are located on both sides of the Montsant, but they all coincide at a position of convergence between the river and the end of cliffs that face it, and that mark the start of the ranges between which it runs, the Montsant Range on the left, and the Llena Range on the right. The most recurrent animal species is the Capra pyrenaica. The sites normally have a single phase, with a single archaeological level (Boix, L’Auferí, L’Hort de la Boquera), or several archaeological levels that correspond to the same chronologicalcultural phases (Els Colls). The predominant lithic elements are backed elements or endscrapers, which are always more abundant than burins. Its datings range between the middle of the 8th and middle of the 12th millennia BP, which would make them coincide with the final phases of the Magdelenian. Despite this, though, they do present some specific aspects, such as the aforementioned scarcity of burins and the predominance of endscrapers, the lack of bone industry (although this may be due to postdepositional biases) or in some cases, the presence of figurative portable art on lithic supports (L’Hort de la Boquera), which starts to become generalised in the Tarragona sites of this time (Molí del Salt, perhaps Sant Gregori), and that brings them closer to synchronic sites in the neighbouring zone of Castellón. Based on all of this, we think that it could be attributed to the Late Upper Palaeolithic, coinciding with what we know as the Late Upper Magdalenian, with some regionally-specific characteristics. For now, according to the data on the lithic industry, fauna, and locations of the sites, we suppose that these sites from the Late Upper Palaeolithic in the Montsant Valley could present a similar pattern of occupation: sites close to the river, which would allow the gathering of flint, which was available in abundance in the river, and which were possibly exploited for food resources. They would also facilitate access from the cliffs to the higher areas of the surrounding mountain ranges for specialised hunting of mountain goats, or that took advantage of the intersection between cliffs and the river to wait for this prey when it descended from the mountains to be captured. Finally, and although the site currently represents an exception to the model explained before, we have the El Filador shelter, which, thanks to its stratigraphic sequence and its complete dating series, as well as the reinterpretation of some of its levels, continues to serve as a reference for the systematization of the different facies of the Epipalaeolithic in the northeast area of the Iberian Peninsula, with the presence of microlaminar, geometric, notches and denticulates moments (García-Argüelles et al., 2005; García-Argüelles et al., 2013). 217
EBRO VALLEY, PYRENEES AND PRE-PYRENEES  Approximately 50 pieces were recovered, with side scrapers and denticulate tools p...
4 mediterranean basins. north of the ebro river NORTH-WESTERN ATLANTIC BASINS. 219
4  mediterranean basins. north of the ebro river  NORTH-WESTERN ATLANTIC BASINS.  219
Site Map numbering Site Abric Romaní 41 La Cansaladeta 42 La Cativera 43 Cinglera del Capelló 44 Reclau Viver ensemble 45 St. Julià de Ramis Pleistocene ensemble La Selva and Puig d´en Roca ensemble 46 Montgrí Middle Pleistocene ensembles 47 Cova de l’Arbreda 48 Cova del Gegant 49 Cova del Rinoceront 50 Els Vinyets 51 Barranc de la Boella 52 Molí del Salt 53 Map numbering Cova de les Teixoneres and Cova del Toll 54 Vallparadís 55
Site  Map numbering  Site  Abric Roman    41  La Cansaladeta  42  La Cativera  43  Cinglera del Capell    44  Reclau Viver...
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER. Josep Vallverdú*,**, Ethel Allué*,**, Amelia Bargalló*,**, Isabel Cáceres*,**, Gerard Campeny*,**, María Gema Chacón*,**,***, Maria Joanna Gabucio*,**, Bruno Gómez*,**, Juan Manuel López-García****, Mónica Fernández*,**, Juan Marín*,**, Francesca Romagnoli*,**,*****, Palmira Saladié*,**, Alex Solé*,**, Manuel Vaquero*,**, Eudald Carbonell*,**,******. Abric Romaní (Capellades, Anoia) 1. Location and research history The l’Abric Romaní site is a large rock shelter located on the north side of the travertine cliff known as Cinglera del Capelló (Capellades, Barcelona). La Cinglera is a 60m thick escarpment made from lacustrine-spring travertine mesa formed by a multilayer groundwater springs of the Capellades region. Capellades village lies on this travertine mesa at 300-320m above sea level. Cinglera del Capelló was thoroughly explored by Amador Romaní at the beginning of the twentieth century. A. Romaní was a businessman in the paper industry of Capellades and a naturalist associated to the Institució Catalana d’Història Natural, in turn associated with the Institut d’Estudis Catalans. The Abric Romaní, also known locally as Balma del fossar vell before it was discovered in 1909, is one of the first sites in the Iberian Peninsula in which Mousterian lithic industries were identified. The first excavation campaigns at Abric Romaní were sponsored by the Institut d’Estudis Catalans, first directed by Father Nobert Font i Sagué and then by mining engineer Lluís Marià Vidal. One hundred years after it was discovered, three long periods of research activity and archaeological excavations can be identified. The first runs until the 1930s and its results are related to the argument over the age and the presence of Neanderthals in Catalonia. This phase ends with the * publication of the Abric Romaní findings in the Història de Catalunya d’Antoni Rovira i Virgili, based on the Atlas de Prehistòria made by Amador Romaní, with a depth of 10m (Fig. 1). The second period was led by Dr Eduardo Ripoll and the 5th INQUA Congress held in Spain. The work of Dr Ripoll was based on reviewing the stratigraphy and extending the studies carried out by Amador Romaní. There were also specific studies dedicated to the Upper Paleolithic lithic industries, found in the layer two of the nomenclature of Amador Romaní, by Dr Georges Laplace. Professors Henry de Lumley and Eduardo Ripoll published the stratigraphy review and Mousterian lithic industry in various papers in the first half of the 1960s. The last period of excavation work and research started in 1983 and is still going on. The early years of this project were led by a work group associated to the Universitat Autònoma de Barcelona and the Centre de Recerques Paleoecosocials (CRPES) under the direction of Dr Eudald Carbonell, Artur Cebrià and Dr Rafael Mora. The research programme of this last excavation phase adheres to an approach based on the excavation of large surface which are seemingly very well preserved. The dating of the travertine based on the Uranium series and the pollen preserved in the calcareous sediments have returned very important results on the paleoecology of the Abric Romaní and the region of Capellades. The Abric Romaní scree desposits are an singular archive in continental settings and IPHES, Institut Català de Paleoecologia Humana i Evolució Social, C. Marcel.lí Domingo s/n, Campus Sescelades URV (Edifici W3), 43007 Tarragona, Spain ** Àrea de Prehistòria, Universitat Rovira i Virgili (URV), Avinguda de Catalunya 35, 43002 Tarragona, Spain *** UMR7194 – Département de Préhistoire, Muséum national d’Histoire naturelle, 1, rue René Panhard, 75013 Paris, France **** Sezione di Scienze Preistoriche e Antropologiche, Dipartamento di Studi Umanistici, Università degli Studi di Ferrara, Ferrara, Italy ***** Cattedra di Preistoria, Dipartimento di Storia, Geografia, Archeologia, Arte e Spettacolo, Università degli Studi di Firenze, Via S. Egidio 21, 50122 Firenze, Italy ****** Visiting professor, Institute of Vertebrate Paleontology and Paleoanthropology of Beijing (IVPP), PR China 221
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER.  Josep Vallverd   ,  , Ethel Allu   ,  , Amelia Bargall   ,  , Isabel C  ce...
222 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. Location of the Abric Romaní in the NE of the Iberian Peninsula and sketch of the section of the site with the work carried out by Amador Romaní during the first quarter of the nineteenth century. The stratigraphic sketch of the Abric Romaní was published in the Història de Catalunya in 1928, in which the execution of shaft 1 is noteworthy. complements the global paleoenvironmental archives of the Upper Pleistocene. The excavation and research team from the Universitat Rovira i Virgili in Tarragona, under the direction of Dr Eudald Carbonell, started to work in Abric Romaní during the large surface excavation work in level H in 1989. This team is currently continuing with the work within the Institut Català de Paleoecologia Humana i Evolució Social (IPHES). The interest of the research and development of these excavations is based on the spatial documentation on the structures and the archaeological materials. The large surface excavation work has occupied 9 archaeological levels (form H to P) and three monographs dedicated to levels H, I and J have been published. From archaeological level K, the large surface excavation work is complete and it is not affected by the pits of previous phases of archaeological work. The thickness of the large surface excavation sediment is around 10m deep. In addition, Amador Romaní’s shaft 1 was excavated to an additional 6m deep and its base was dated at 70,000 BP. A mechanical borehole from level P has documented another 30 m of deposits, which is located around 9 m below the last travertine of the rock-shelter. Therefore, the talus scree deposits accumulated at the foot Abric Romaní escarpment is at least 40 m thick. This thick places the base of the talus scree at the foot of the escarpment at 260 m above sea level. This elevation of 260 m is very close to the elevation of the +20-25 m terrace encaised in the valley of the river Anoia.
222  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER. 2. Stratigraphy and chronology The Sedimentary Geology of the talus scree deposits at Abric Romaní escarpment has an evident zonation caused by the dripline of the carbonate curtains in the roof of the archaeological site. In the region of Capellades the carbonate curtains are called Capelló. These carbonate curtains are formed by plants being encrusted by calcite and develop along the lacustrine travertine escarpment concurrently with the alveolar weathering, or tafoni, of the cliff wall. In some archaeological levels, such as levels J and M, the distance between the back wall and the dripline of the rock-shelter can be more than 12 m. Therefore, during the fastest growth periods of the carbonate curtain in the Abric Romaní, there is a wide surface habitable at the foot of the wall protected from the weather by the shelter roof. The stratigraphy of the Abric Romaní has repeatedly been described during various research excavation projects. Amador Romaní distinguished sand stratum and stalagmite stratum. The work of Ripoll and de Lumley recognises three types of sediments–rocks: gravel and sand deposits; reddish calcareous deposits and calcareous deposits (travertine). In the stratigraphic descriptions from the start of the third phase of research, the attention given to the different variants of travertine lithofacies is noteworthy. All of these descriptions are based on observing outcrops of the upper part of the stratigraphic sequence of the rock-shelter and they coincide in describing the three main lithofacies described by Dr Ripoll and Dr de Lumley. The works of Amador Romaní’s in the shaft 1 confirmed the volumetric significance of the travertine and calcareous sediments in the stratigraphic succession of the Abric Romaní rock-shelter. The travertine was widely dated and published in Nature by J.L. Bischoff. This work details the high chronological resolution of the stratigraphic succesion and the potential of the site to illustrate the latest archaeological assemblages of the material culture of the Neanderthals in the northeast of the Iberian Peninsula. The high temporal resolution of the sediment of the Abric Romaní was characterised by its high sedimentation rate, estimated at 0.6m per 1000 years. The dates have been established using the uranium series and situate the deposits of the Abric Romaní in marine isotope stage 3 (40 – 60 ky BP). Later, the same geochronologist published the calendar chronology of the travertine containing the Upper Paleolithic industries in the layer 2, or level A of the current archaeological stratigraphy, and calibrated the radiocarbon chronology of the carbon in this layer (level A). The results confirmed a consistent and reliable group of calendar dates of high temporal resolution for the travertine at the top of the shelter sediment sequence and also confirmed the old age of the Upper Paleolithic at Abric Romaní (42.6 ky BP) . Sampling of shaft 1 also offered the opportunity to research the pollen biostratigraphy of the Abric Romaní. The pollen study was carried out by Dr F. Burjachs and Dr R. Julià from the Institut Jaume Almera in Barcelona. The samples were taken from shaft 1 and different profiles available in the old archaeological pits from the top portion of the stratigraphy of the Abric. The pollen record was divided into 5 bio-zones which register abrupt climate changes. Bio-zones 1, 2 and 3 show variations in the arboreal / non-arboreal content in a chronostratigraphy interval similar to the variation in the isotopic content of the oxygen identified in the ice core samples from Greenland. The upper bio-zones of the sediment sequence are more difficult to correlate. However, the pollen curve for the Abric Romaní shows a singular relation between age models of paleoclimatic change recorded in prehistoric caves and rock shelters and global stratotypes of environmental change based on in marine and glaciological settings. From archaeological level K, large surface excavation in the Abric Romaní started documenting a large area without pits from the old research projects. Two stratigraphic profiles were preserved in the centre of the shelter. However, as the depth of the large surface excavation increased, it was clear that the outcrop of the coveta Nord section of the site was the best stratigraphic pannel to explain the Sedimentary Geology of the rock-shelter (Fig. 2). A few years later, one of the profiles reserved from the shelter wall was excavated while the volume of the other was reduced in order not to affect the spatial documentation of the archaeological levels excavated in large surface. 2.1. The coveta Nord section The roof of the rock-shelter presents the carbonate curtain welded with the deposits of the stratigraphy of the rock-shelter after the level J. This welding determines the increase in the frequency of the endokarstic sedimentary process which forms bio-chemical sediments in the strati- 223
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER.  2. Stratigraphy and chronology The Sedimentary Geology of the talus scree ...
224 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 2. Sketch of the coveta Nord section of Abric Romaní and its synthetic lithostratigraphic logs Legend: a, conglomerates and tuffaceous sands; b, bryophyte bio-constructions; c, crystalline angular gravel; d, cementation and algal laminated bio-constructions . e, calcareous sands and siliciclastic red muds; f, recrystallizated stalagmitic massif; g, boundary surfaces ; f, sequence boundaries; Comments on the column: I, archaeological levels; II, Uranium series dates; III, sequence numbering; IV, Bond cycle boundaries with the D-O Greenland event numbers and the Heinrich events of the North Atlantic with the GISP2 temporal scale ; V, pollen zones. graphic record of the upper part of the Abric Romaní succesion. The rest of the sedimentary succesion could be described by the predominance of the clastic processes alternated with the biochemical processes, particularly near the wall of the shelter. In the dripline, the stratigraphy shows the importance of fragmentation of the curtain of the rock-shelter. These clastic processes were highlighted in the first stratigraphy studies of the last research period, as large blocks in the shaft 1 and used as a criterion to separate stratigraphic ensembles. The growth or shrinkage of the carbonate curtain of Abric influences the location of the dripline. The temporal drifting of the dripline results in clear zoning of the deposits accumulated at the foot of the travertine escarpment: an interior zone, between the back-wall and the dripline of the rock-shelter and an exterior zone, from the dripline to the slope of the talus scree. The coveta Nord section fit as best stratigraphic outcrop for understanding the talus scree deposits (talus d’eboulis) in a rock shelter setting. Many of the facies described as sediments formed by precipitation (travertine) are quickly buried and recrystallization is relatively limited and allows uranium series dating. The coveta Nord section has allowed the stratigraphic description work to be reviewed to set out 5 allostratigraphical units or sequences. These sequences are based on hierarchisation of the discontinuities or boundary surfaces of the sedimentary bodies. The even numbered sequences contain deposits of dominant clastic facies and indicate shrinkage of the carbonate curtain of the shelter curtain. The odd numbered sequences contain dominant calcium carbonate precipitation and represent the Capelló accretion. It is noteworthy that the collapse of the carbonated curtain is time transgressive: it occurs regularly over all sequences.
224  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER. The upper chronostratigraphic horizon of the sequence is the most recent travertine dated in the Abric Romani averaged at 40 ky BP. This discontinuity has a clear lithological change between the most recent travertine and red calcareous and siliciclastic sediments in the coveta Nord section. These deposits have been dated using OSL at 29.5 ± 1.9 ky BP. The chronostratigraphic horizon of the bottom of sequence unit I, which separates sequences I and II, has dates measured at 44.9 ± 2.5 and 44.6 ± 1.5 ky BP by Uranium series. These dates between the last travertines averaged in 40 ky and 44.9 above level E are close to the Greenland interstadial 12 time span, also known as Hengelo interstadial in the pollen biostratigraphy of Grand Pile. Greenland ice cores. Up until now we have been able to group sediment sequences and chronological intervals for stadials and interstadials 9 to 14 and we are now, in the archaeological levels P and Q, very close to reaching Greenland interstadial-stadial 16-17. The environmental change age model provided by the Abric Romaní sequences is independent and the dates measured are calendar dates (U-series dates). These environmental change dates, established by a group of dates and stratum delimited by discontinuities, show better correlation with the cronology of ice-core GISP2 age model than the GRIP age model. The chronostratigraphic horizon separating sequence II from III are the travertine in which archaeological level Jb lies and the average date measured is 50.0 ± 1.6 ky BP. The bottom boundary of the sequence III has a calendar date of 52.5 ± 1 and their chronostratigraphic boundaries, of level K, can be correlated with Greenland interstadial 14, also known as Glinde interstadial. Dating of the base of sequence 4 has not yet been determined but it contains a date group prior to Greenland interstadial 14. Paleoenvironmental research of the Abric Romaní sequence has provided a lot of valuable data for reconstructing the paleoecology of the human occupations. The paleoenvironment reconstruction data comes from several proxies that include paleobotany (palynology and anthracology) and micro-vertebrates (micro-mammals, amphibians and reptiles) that complement and allow habitat and climate aspects to be understood in order to define the landscape in which Neanderthals lived. The sediment sequences of Abric Romaní show the climatic control in the rhythmicity of the carbonate curtain growth and shrinkage. Uranium series dating of the Abric Romaní sequences indicate the formation of calcium carbonate precipitation deposits and fragmentation during the abrupt environmental oscillations, with a magnitude of the time scale of the Bond cycles or long periods of cooling. These long periods of cooling end with a maximum period of cooling correlated with the Henrich events in the North Atlantic. Therefore, the red calcareous and siliciclastic sediments at the top of sequence II, which contains level E, shows the aeolian deposits during the chronology of Henrich 5. At the top of sequence I, although formed by red calcareous and siliciclastic sediments too, from the top with an OSL dating of around Henrich 3, the date determined in the most recent travertine of Abric Romaní, dated at 40 ky, is significant as this date is very close to the chronology of the Heinrich 4 event in the North Atlantic (39 ky BP). To sum up, the chronostratigraphy and the sequences of Abric Romaní cover the oscillations described in the Oxygen isotope stratigraphy in the 3. Paleoenvironment The palynological sequence throughout the phases shows a dominance of Pinus that characterises tree formations during the whole period (Fig. 3). At the base of the sequence, between 70 and 67 ky BP, the data reflects a warm climate phase with a dominance of tree pollen, including Quercus evergreens and Olea / Phillyrea. Between 66 and 59 ky BP gramineae dominate, reflecting a colder phase. Between 57 and 50 ky BP and Pinus and gramineae (Artemisia and Poaceae) dominate with a presence of meso and thermophilic taxa at times, indicating climate oscillations during this period. Between 50 and 47 ky BP the data reflects the dominance of Asteraceae, Poaceae and Artemisia, suggesting steppe vegetation and cold conditions. Finally, around 46 ky a warm climate trend is identified with an increase in Quercus and Olea-Phillyrea (Fig. 3). Anthracological data for levels D to O shows a dominant taxon in the anthrocological set, which is Pinus sylvestris and represents more than 90% of identifiable material. The dominance of this species is conditioned by the selection of this taxon for use as fuel. Level O presents other taxa in addition to Pinus such as as Prunus and Juniperus and in level D, mesophile species in charcoal fragments are identified such 225
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER.  The upper chronostratigraphic horizon of the sequence is the most recent t...
226 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD as deciduous Acer, Quercus sp and other unidentifiable angiosperms, which may reflect a change due to more favourable climate conditions that are consistent with the palynological data. Scots pine forests are the dominant tree formations throughout the sequence, which are characterised for being forest with many clearings without much taxonomic diversity. Microvertebrates at Abric Romani come from the study of levels D, E, J, N and O. The species identified are insectivores: Russula Crocidura, Sorex gr. coronatus and Talpa europaea-araneus, Chiroptera: Miniopterus schreibersii, Pipistrellus pipistrellus, Nyctalus lasiopterus; and rodents: Microtus arvalis, M. agrestis, Iberomys cabrerae, Terricola duodecimcostatus, T. cf. pyrenaicus, Arvicola sapidus, Apodemus sylvaticus and Eliomys quercinus. Similarly, water environments are identified, such as Arvicola sapidus. The herpetofauna study represents amphibians and reptiles providing 3 anuran: Bufo bufo, Bufo calamites, Rana temporaria; and 3 reptiles: Lacertidae indet., Anguis fragilis and Vi- pera aspis (Fig. 3). This data reflect domination of open forest species and species that require humidity and colder climate conditions than the current ones. The paleoenvironmental data on Abric Romani reflects a mosaic landscape: forests; grasslands; riparian forests, which are larger or smaller according to the warmer or cooler climate phases. In general, the sequence data reflects colder temperatures than current ones, which is proven by the presence of certain micro-vertebrates and the distribution of Pinus sylvestris. 4. The lithic industry The general characteristic of all archaeological levels of the Abric Romani is the fragmentation of the operative sequences, although refitting studies shows that full or almost full knapping sequences have been found in some levels. The main objective of the lithic sequences is to obtain as many knapped products as possible and these are there- Figure 3. The biostratigraphy record of the Abric Romaní and the habitats identified by the disciplines dedicated to paleoenvironmental and paleoecological reconstruction
226  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  as d...
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER. fore the predominant category in all archaeological levels of the site. However, cores and retouched objects are scarce in the record, not reaching significant percentages in most of the archaeological levels except level O, where we documented 170 cores. In terms of raw materials, the most used is chert, in different varieties (Fig. 4, D), followed by quartz and limestone, and there are a few in quartzite, porphyry, granite and schist. Lithic raw apcaptation is local and semi-local in a perimeter around the shelter ranging from a few hundred metres to 20 km. Changes in the strategy for raw materials provision within the same environment are observed throughout the sequence. The clearest example is chert; this raw material is dominant throughout the sequence but the percentage changes from one archaeological level and another. In the case of level I, this raw material has very low percentages (level I = 50%) compared to the rest of the sequence. Secondly, there are levels that have significantly higher percentages (F, G, L, O), reaching values of up to 90%, and even monopolising all knapped objects, such as level H. This would reflect the ability of human groups to select the most appropriate strategy from the possible alternatives for supply raw material at any given time. Morphotechnical analysis of the cores and knapped products has allowed the different types of operative sequences performed in the Middle Palaeolithic levels of the site to be reconstructed: tested cores left without late transformation, hierarchical centripetal cores, Levallois method (essentially recurrent centripetal), discoid method, polyhedral morphologies and fragments or flakes with some isolated retouch without any predetermined organisation or schema. The distribution of the different strategies is not homogeneous on all levels. Thus, in the upper levels (level E) there is a tendency for hierarchical strategies (Fig. 4, C). However, in the intermediary levels (levels I, J, K, L, M) there is a preference for non-hierarchical models. At lower levels O and P the existence of hierarchical strategies is identified again, with a high number of cores and knapped products that show the use of the Levallois method. In terms of the production sequences, there is an almost exclusive operating standard. Denticulate morphologies dominate in all levels, especially in the lower levels (90%) (Fig. 4, A). For this reason, after the first studies of the lithic industry, based on the liste-types, the Abric Romani sequence was ascribed to “Mousterian of denticulates”. Large surface excavation has allowed studies on the spatial distribution of lithic assemblages to be carried out (Fig. 4, E). These studies have permitted reconstruction of the spatial distribution of the lithic technical activities and in particular, of movement and transport of raw materials between the lithic activity areas identified within the site. The spatial dimension of the operative sequences shows an anthropic differentiation in different lithic activity areas of the living floors, particularly in the levels made of large lithic assemblages quantity, with spaces that act as convergence points for the different operating processes. By analysing the spatial distribution of the lithic remains in a large part of the sequence of Abric Romaní we have been able to document a very different use of the space, indicating highly complex organisation and activity area contemporaneity in the Neanderthal living floors. Secondly, these studies have also provided diachronic and synchronic data, several episodes of independent occupation being documented for the same level, which we have identified based on the recycling of artefacts. A functional analysis of the use-wear lithic assemblage using a scanning electron microscope has been carried out on a sample of flakes and retouched objects from all archaeological levels in the sequence. The results show that the pieces in which traces of use-wear have been identified were used mainly for transforming animal biomass in butchery activities and in some isolated cases, for technical activities related to transforming vegetable remains (mainly wood). 5. The fauna The skeletal remains of fauna are abundant throughout the sequence of Abric Romani. A total of 13 different taxa have been identified, although deer (Cervus elaphus) and horses (Equus ferus) are the most common animals in all levels of the stratigraphic sequence. Aurochs (Bos primigenius) and chamois (Rupicapra pyrenaica) are also present, the former in the lower part of the sequence and the latter at higher levels. The presence of specimens of rhinoceros (Stephanorhinus hemitoechus) has also been documented at various levels, although through few remains. At level E a femur of an unidentified proboscidea was recovered. Despite the high prominence of her- 227
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER.  fore the predominant category in all archaeological levels of the site. Ho...
228 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 4. Lithic industry at Abric Romaní. A. Chert retouched objects (denticulate tools); B. Pseudo-Levallois lithic flake in limestone; C. Bifacial centripetal cores (discoid and Levallois) in chert; D. Flake and laminar flake in chert; E. Lithic reassemblages in chert, quartz and limestone.
228  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER. bivores in all levels, some carnivore remains have also been recovered (Table 1). These are most abundant in the upper part of the sequence (up to level E) where a cave environment allows the presence of these animals to be associated with the establishment of occasional dens. Apart from the natural intrusion of carnivores in levels B and O, remains of Lynx sp. and Felis silvestris have been documented, respectively, with cut marks and which are the result of the contribution and use by Neanderthals. In terms of taphonomy, the most abundant modifications in the bone assemblage are those related to anthropic use: cut marks and fractured bones. Cremation has also been found very often but it could be related to specific deliberate activities. However, it can be said that cremation was possibly carried out using different processes ranging from preparing food to clearing the floor of the living floor. Carnivore modifications are almost absent from the faunal assemblage. In addition to the diversity, it seems clear that deer and horses were an important part of the diet of the Neanderthals during the settlements at Abric Romaní. The assemblage stands out for the presence of high survival items: craniums, jaws, diaphysis fragments of long leg bones. The first stud- A B Ursus sp. D X E The faunal record for the Abric Romaní shows that Neanderthals were active hunters and that they used complex strategies to catch and transport the carcasses of large ungulates. The low presence of carnivores and their modifications in the assemblage suggests that these animals were not much competition for the Neanderthals. These human groups also developed an intense intake of the entire carcass and combined with the use of fire, they left few items that could be a target in order to be ravaged by carnivores. F G H I J K L X Canis lupus C ies established that this representation was due to the different ways prey was transported. According to this model, deer was transported whole and the axial skeleton of larger animals was left where it was found. Current research indicates that the selective transportation process carried out by Neanderthals was highly complex and variable in terms of anatomy and did not simply involve abandoning the axial skeleton of large animals. Thus, for example, the scarce rhinoceros remains or the single proboscidea remain in level E, seems to be the result of this decision. The low presence does not mean anything more than selective transport of the skeleton of these animals, undoubtedly related with the profitability of the contribution to the camp of one party or another. M N X O P X X X X Panthera leo spelaea X Panthera pardus Lynx sp. X X Felis silvestris X X Crocuta crocuta X X X X X Proboscidea indet. X X X X Stephanorhinus hemitoechus X X Equus ferus X X X X X X X X X X X X X X X X Cervus elaphus X X X X X X X X X X X X X X X X X X X X X X X X X Bos primigenius Rupicapra pyrenaica X X X X X X Table 1. Main fauna taxa in the archaeological levels of Abric Romaní X X X 229
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER.  bivores in all levels, some carnivore remains have also been recovered  Ta...
230 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD new interpretations, providing information on the ways of life and social organisation of the domestic space of the Neanderthals. The methodology for studying these remains is based on analysing the morphology and stratigraphic and spatial position. The preservation of wood in the form of imprints, charred wood and charcoal allows us to interpret, for the first time, how firewood was gathered, the formats collected and how it was used. The fuel used for household maintenance consisted mainly of Scots pine and different sizes and shapes were used. According to data provided by the characterization of the firewood in the level M, the sample from which is the largest and best studied so far, there are two fuel collection methods. Collection would be based on dead wood, therefore, that which occurs physiologically (twigs) would be collected directly, immediately and systematically with little investment and effort while larger firewood that occurs more traumatically (branches and logs) require a greater investment of effort to find and transport, which is offset by the higher performance of these items. Different patterns of fuel use have also been observed, according to the spatial distribution pattern: especially the distance between the wall of the shelter and the human occupation/s density. Figure 5. A wood imprint located in level N. The imprint is clearly defined aswood and seems to have been modified, as it does not have any type of branch. 6. Wood imprints The wood record at Abric Romaní represents a particularly unique case, as the context is a prehistoric habitat. Wood remains have been identified in all of the levels excavated in the large surface excavation (from H to P), its use for various activities being documented; as fuel, the wood imprints are documented as charred firewood on the combustion strucutres or as accumulated reserves; for structural or architectural applications (Fig. 5); for wooden objects; and for tools. The fossils are preserved under very special conditions related to the drip of water from the shelter roof. Both charred wood and imprints formed/buried very quickly by calcium carbonate precipitation. The study of this exceptional record and the possibility of associating it with the other elements of the living floor a various large surface archaeological levels have great potential to generate 7. Combustion structures The record on fire use at the Abric Romaní is further evidence of the exceptional state of preservation of the archaeological assemblages and, particularly, the possibility to study archaeological structures in the record on ancient human groups like the Neanderthals. The combustion structures have been repeatedly documented in the archaeological levels excavated and their spatial documentation in a large surface excavation allows uncommon problems in the Archaeology of the Neanderthals. Use of fire is an activity that is repeated in the levels excavated in the shelter and their spatial record can be allow a comparaison analogue with the use of fire in others prehistoric and current hunter-gatherer (contemporaneous) localities. Studying the combustion structures of the Abric Romaní indicated the existence of a number of intra-occupational episodes in every archaeological level. The spatial distribution of the combustion structures could indicate the juxtaposition of the different activity areas around the fire use (heart related activity areas). The juxtaposi-
230  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  new ...
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER. tion of distinct activity areas is a guide for estimating the site structure in the archaeological record (types of settlement, number of occupants, etc.). The spatial documentation combustion strucutres in the Abric Romaní record supports the argument that the Neanderthals used fire for different purposes. The inventory of the combustion structure documented at Abric Romaní exceeds 200 and these include a range of construction techniques. The most common combustion structures (>80%) are flat and without stones. There are also flat combustion structures with stones; within concavities with carved tails; in small pits with burned stones and sediments; in re-excavated holes. In our opinion, many of these combustion structures could be considered special fire use activity areas. The spatial distribution of most of the flat combustion structures can be described by their distance from the shelter wall according to the each archaeological level. In level I we documented combustion strucutres with accumulations of faunal and lithic remains (hearth related assemblages) without any relationship between them. However, there are 8 combustion structures connected by a few refitted lithic and faunal remains which describe a circle. The combustion structure assemblage in level J shows two modal distances from the shelter wall. One places the use of fire within 0 to 3m of the wall and the other places the combustion strucutres between 6 to 9 metres. Level N combustion strucutres has been arranged based on the principle that certain group of these combustion structures, which are one metre apart, are the inner zone of the site structure during the prehistoric settlement in the rock-shelter. This inner zone does not have any lithic or faunal remains and the combustion strucutres are less than 2 metres from the shelter wall. The other combustion structures in level N are distributed into two arches to 5 and 9 metres respectively from the inner zone of the settlement. Level O contains a large number of combustion structures. There are combustion structures in concavities with burned rocks and very few fauna remains; and combustion structures with lithic industry between 6 and 12 metres from the shelter wall. Large combustion structures are superimposed close to the shelter wall and have elementary hearths 1 metre apart and contain fauna remains and knapped lithic artefacts. The study of micro-artefacts, microstructures and other residue from the molecular scale is relatively limited and is one of the next challenges of the Abric Romani research team. Another challenge for archaeological researchers is to characterise the combustion strucutres that have no lithic artefacts or fauna remains. Among the micro-remains, incorporated during the fire use determined in the Abric Romani combustion structure record, there are coprolites, calcium oxalate, fibre remains, pigments, etc. The thermal modification of the combustion structure sediments, based on the modification of calcite by heat and the combustion residues rich in charcoal, indicates fire use in which low temperatures are reached. Charcoal-rich sediments contain little ash. The low temperatures estimated are based on observing the size of elementary hearths too, observed in the stratigraphic sections of the combustion structures. These are charcoal lenses 20cm in diameter with a matching rubification band of the same size. This size, measured in meridional stratigraphic profiles of these elementary hearths, indicates use of fire in which fuel is limited. There are also combustion structures with large carbonaceous lenses and a rubification band of more than 40 cm but these are less common in the Abric Romani combustion structure record. 231
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER.  tion of distinct activity areas is a guide for estimating the site structu...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 232 Josep Maria Vergès*,**, Andreu Ollé*,** La Cansaladeta Introduction The site of La Cansaladeta is situated in the Roixeles canyon, a natural pass carved in the PreCoastal Catalan mountain range by the Francolí River, which connects the Tarragona coastal plain with the interior depression of Conca de Barberà. The archaeological deposit is located in the upper part of a fluvial terrace, approximately 45-50 metres above the river, at the foot of a small, almost dismantled rock shelter formed in the Middle Triassic limestones (Muschelkalk deposits) at 260 metres a.s.l. Members of the Department of Prehistory of the URV discovered the site in 1998. In 1999 a first excavation was carried out, aimed at assessing the site’s potential. Excellent results brought its integration in the research project “Paleoenvironmental evolution and prehistoric settlement in the Francolí and Gaià Rivers” in 2002 (Angelucci et al., 2003, 2004). Since then archaeological fieldwork is conducted annually. Stratigraphy The sedimentary deposit is 16 metres thick. It is composed of a significant succession of alluvial deposits, covered by hillside deposits product of mass wasting dynamics under the influence of gravity forces. The alluvial sequence, noticeable in the fluvial terrace at 45-50 m, starts with different sized blocks and gravels transported by the river. After that, from bottom to top, there is an alternation of alluvial gravels and sands, alternating with occasional depositions of limestone blocks coming from rocky walls. The last phases of alluvial accumulation consist of sandy layers assorted with substantial blocks. A rubified paleosol is present at the end of the sedimentary cycle opening a steady * ** phase. In the upper part, the alluvial deposit is cut by erosive processes occurring after the river bed was filled and downcut by the river. Subsequent to that, hillside deposits were piled, forming a limestone breccia in silty loamy matrix. The archaeological material appears on the top of the alluvial accumulation, close to the limit with the foothill deposits (Fig.1). The most ancient evidences of human activity have been found at level M, where lithic assemblages were recovered in a sandy layer superimposed on a limestone breccia. However, the first traces of importance regarding human occupation have been documented at level L and K. During this period, the relative height of the site in relation to the bed river protected it from the main stream. The developed paleosol indicates Mediterranean environmental conditions. Also, at that time, the rock shelter had still preserved its brow, since that level contains cryoclastic breccia materials originally coming from that part. These elements point to colder and wetter climate conditions, prior to the formation of the paleosol. Afterwards, the erosive action of the river affected the rock shelter stratigraphy. Archaeological levels J and I belong to this phase. Lowenergy alluvial deposits from the river flooding, and cultural layers result of human occupation during periods with no depositional processes from the watercourse, compose the sedimentary sequence. Later, when the river started to downcut the underlying bedrock, its influence on the rock shelter sedimentation sequence gradually disappeared. After a transition phase, corresponding to level E, the predominance of colluvial hillslope materials and rocks from the wall is documented. This is observed in levels D, C, B IPHES, Institut Català de Paleoecologia Humana i Evolució Social, C/ Marcel·lí Domingo s/n (Edifici W3), Campus Sescelades, 43007 Tarragona, Spain Àrea de Prehistòria, Universitat Rovira i Virgili, Fac. de Lletres, Av. Catalunya 35, 43002 Tarragona, Spain
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  232  Jose...
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER. Figure 1. A. Location of La Cansaladeta site (on the centre), on the top of the fluvial terrace ±45-50 m. of the Francolí River; B and C. Overview of the excavated areas. D. Lithic assemblages in situ, level J (hornfels pebble and flint flakes); E. Altered bone, level K; F. Stratigraphic sequence (modified from Angelucci et al., 2004). 233
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER.  Figure 1. A. Location of La Cansaladeta site  on the centre , on the top o...
234 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD and A, which compose a thick edaphological sequence, the result of a large and steady geomorphological period, probably developed in Mediterranean climate conditions. The rest of the raw material comes from pebbles and cobbles of fluvial origin. Its morphology, in cases such as the quartz, the quartzite and the lidite, is derived from the original geological formation, the Buntsandstein conglomerates. Chronology All the raw materials, except the granite, have been used as blanks for knapping stone objects. Granite, hornfels, quartzite and limestone have been used to make hammers or anvils. The presence of all sized cores, flakes, other debitage waste-products, and some refitting, indicate that most of the chaîne opératoire took place at the campsite. One of the first chronological proxies is the fluvial terrace situated on the top of the sequence that has been ascribed as Final Lower /Early Middle Pleistocene. Another key element providing a relative chronology is the presence in level L of an Hystrix refosa tooth, a kind of porcupine that became extinct around 500,000 years in the Iberian Peninsula (G. Cuenca, pers. com.). On top of that, two numerical dates are available: one ESR/ US date on a rhinoceros tooth fragment from level J, that yielded a minimum age of 340,000 ± 17,000/16,000 years BP (C. Falguères, pers. com.); and another date by thermoluminiscence method on burnt flint form on level D, 372,000 ± 34,000 BP. Accordingly, occupations at La Cansaladeta site can be situated in a chronological span from 300,000 to 500,000 years BP (Angelucci et al., 2004; Ollé et al., 2008). Archaeological record Almost all the archaeological materials of La Cansaladeta site correspond to lithic assemblages (7575 items, 94.8% of the total), whereas faunal remains represent the remaining 5.2%. Lithic and faunal assemblages have been retrieved at 8 of the 10 archaeological levels, while the other two levels (B and A) only had lithic material. Lithic assemblages Raw materials employed for knapping activities can be found both in secondary position on the alluvial deposits of Francolí river at the bottom of the site where they probably were procured, and also in primary position within a minimum catchment area with a radius of 10 km. Flint was largely the most common rock type used in around 82% of the artefacts, followed by hornfels (7.5%), quartzite (3.3%), and other rock types such as limestone, agate, granite, and lidite, with percentages lower than 1%. Flint, from Eocene deposits, appears as nodules of irregular morphology, slightly rounded by fluvial erosion, with a heterogeneous structure and abundant cracking. The lithotechnique assemblage presents a large homogeneity between levels (Fig.2). Unipolar core reduction strategy, together with orthogonal and opposed bipolar core flaking strategies, has been documented. Also centripetal core reduction has been detected, which is poorly standardised. Flint characteristics caused different knapping accidents that, in the end, have influence on the exploitation process. Likewise, the small size of quartz, quartzite, and lidite restricted the length of the lithic reduction sequence, and favoured knapping on an anvil. On the contrary, the large size and homogeneity furnish hornfels pebbles as the chosen raw material to shape large tools or flaked big blanks. Flakes are the most abundant debitage products. Due to the above-mentioned flint characteristics, many of them have irregular morphology and present a lot of knapping accidents. Direct percussion with a hard hammer contributes to increase the knapping accidents percentage, and makes bulbs and marked scars on the core surfaces that restrict the reduction dynamics. Because of that, most of the debitage products are shaping out or preparation flakes with cortical backs. Along the whole sedimentary succession, retouched pieces have a very low frequency, and the morphology variability is scarce. Denticulate objects dominate (notched, denticulate points, épines and denticulate scrapers). Technical procedures observed at La Cansaladeta lithic assemblages make it difficult to ascribe it to a particular technological complex. Nevertheless, the presence of large tools characteristics of Mode 2 (such as cleavers, and picks) is significant, although they are sporadic and limited to the lower levels. On the other hand, reduction techniques aimed at getting debitage products with a prede-
234  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  and ...
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER. Figure 2. 1. Cores (a, from level K, hornfels; b, c, e, from level J, flint; d, from level D, flint); 2. Flakes (a,b,d, from level D; c, e, f, from level J; g, from level K; all made of flint except g, made of quartzite); 3. Retouched flint flakes (a,b,c,d, from level D; e,g,h,I, from level K; f, from level J; j, from level L); 4. Large tools on hornfels, from level J. 235
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER.  Figure 2. 1. Cores  a, from level K, hornfels  b, c, e, from level J,    i...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 236 termined morphology (such as Levallois) are not documented, and retouched flakes do not show any standardized or regular configuration. Thus, the available data, broadly interprets the lithic assemblage of La Cansaladeta as Acheulean, a collection where large sized tools are clearly in the minority. Faunal remains Animal remains were recovered from eight levels, being levels J, K and L, those that gather the vast majority of the faunal record. The osteological collection is characterized by two main trends: the small size of the fossils, and its intense postdepositional alteration. Obviously, the assemblage represents only a small percentage of the whole animals set that might have been deposited during the occupations. This fact limits the information that archeofaunal remains can provide. Nonetheless,different taxa have been Fontanals, Marta **´ * Vergès, Josep Maria *´ ** Morales, Juan Ignacio* Stratigraphy and chronology of occupations The stratigraphic succession is approximately 2 m deep, divided into 8 archaeological levels, identified from base to top as A, B, Bb, C1, C2, C3, C3b and C4, the 7 geoarchaeological units distinguished on the basis of sedimentary and pedo- ** Fire is responsible for the most common alterations in La Cansaladeta faunal assemblages (specially al levels D and L). With some exceptions, intensity of burning damage has been rather medium and did not reach maximum degrees of white calcined bones. This heat alteration has also been identified in lithic artefacts (especially at levels C and D, and to lesser extents, at levels E, J, and K), as rubified areas, fine fire cracks and thermal debris. Although no charcoal fragments or hearths were documented at the site, the amount of burned items, and their iteration along the stratigraphic sequence, suggests that they were the result of human activity rather than natural fire effects.. La Cativera (El Catllar, Tarragona). A Pleistocene-Holocene interface site in southern Catalonia The Cativera archaeological site (El Catllar, Tarragona) (Fig 1), is in a small open shelter on the left bank of the Gaià River, roughly 70 m above sea level. The shelter, at the base of a Miocene calcarenite wall, is 23m wide, with a maximum height of 3 m and a current depth of at least 3 m. * identified: Leporidae remains at levels I, J, K, and L; Cervidae bones at levels K, and L; one Equidae, one Rhinocerotidae and one fish remains at level J; as well as the above-mentioned porcupine at level L. Some of these bones present butchering cut marks and intentional breakages that have an anthropic origin. logical criteria (Fig. 1). The top section, including levels A, B and Bb, consists of a calcareous breccia with a silty-loam matrix built up from fragments of the ceiling and fine sediment on the slope. The sedimentation of the middle and base part, corresponding to the rest of the archaeological levels, is originated in the cyclical alluvial processes associated with the activity of the Gaià River (Angelucci, 2003, 2005). Charcoal sample datings are consistent with the stratigraphic sequence, and situate the site’s chronocultural sequence between the final stages of the Pleistocene and the start of the Holocene IPHES, Institut Català de Paleoecologia Humana i Evolució Social, C/ Marcel·lí Domingo s/n (Edifici W3), Campus Sescelades, 43007 Tarragona, Spain Àrea de Prehistòria, Universitat Rovira i Virgili, Fac. de Lletres, Av. Catalunya 35, 43002 Tarragona, Spain
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  236  term...
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER. Figure 1. Frontal view of the Cativera deposit Level Lab. Ref. Material C14 BP Data A MAD-4645BIN Ceramic 4645± 316 A AA-23367 Charcoal 7979 ± 60 B AA-23368 Charcoal 8860 ± 95 Bb Beta-281623 Charcoal 8230± 40 C1 AA-23369 Charcoal 10370 ± 100 C2 AA-23370 Charcoal 10660 ±120 C3 AA-23371 Charcoal 11230 ± 100 C3b AA-23372 Charcoal 11135 ± 80 Table 1. La Cativera site datings Archaeological material The archaeological items found on all levels consist of lithic industry, iron oxides –not documented on level A–, marine and terrestrial malacofauna, charcoal, fauna and mobile art on level C4. However, there are significant typological differences between the material on level A and the rest of the archaeological levels. The flint and limestone lithic material is by far the most abundant on the levels, although the latter material is in a much smaller proportion. Flint is only involved in the tool production and configuration sequences, while limestone is functionalized directly, without modification. Despite this homogeneity in the use of raw materials through the levels, Level A is predominated typologically by notches and denticulates, and thus referred to as configured items with simple reduction sequences intended for flakes (Fontanals et al., 2009), while in the rest of the sequence, the exploitation systems are predominantly aimed at producing blade-like products, the majority of which were endscrapers and blade/backed blades (Morales et al., 2012, 2013). This profile seems to indicate that in all the occupations, the raw materials were supplied from the terraces of the Gaià River, quite close to the settlement. This source was abundant but generally did not provide good quality material. The average length of the items was never more than 15 cm, and less than 10 cm in the case of good knapping material (Fontanals 2001). Abundant remains of marine mollusc fauna were found at the site, no doubt influenced by its proximity to the coast. To date, 18 different species have been documented. This variation is particularly broad on levels B and Bb, where specimens were clearly used for at least three differentiated purposes: consumption (Mytilus galloprovincialis and Patella caerulea), ornaments (Dentalium vulgare and Cyclope sp.) and as a container for ochre (Glycimeris insubrica and violascens). This differentiated use of marine mollusc fauna is less obvious on level A, where the documented species, judging by their features, seem to have been brought to the site for consumption, as indicated by the presence of Patella caerulea and Cerasthoderma glaucum, the latter quite abundant, unlike level B, although other secondary uses must not be excluded, as in the case of Pecten jacobeus and Insubrica Glycimeris, despite the absence in this case of traces of ochre inside the valves found in assemblage B. Items from several species of terrestrial mollusc fauna have also been documented. On level C3, 21 Cepaea nemoralis shells suggest the use of this species as a food source. 237
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER.  Figure 1. Frontal view of the Cativera deposit  Level  Lab. Ref.  Material...
238 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD The faunal record is quite scarce on all levels and the recovered remains are poorly preserved. This is probably due to the high acidity of the sediment in the rock-shelter and postdepositional taphonomic disturbances to the archaeological remains, mainly caused by roots (Allué et al., 2000). Nevertheless, various skeletal parts of Leporiade and Cervidae taxa were identified on levels A, B and Bb. Despite the above-mentioned bias in the record due to poor preservation, the number of Leporidae items is clearly much higher than others, permitting the assumption of its priority for consumption. Remains of burned wood fuel have been recovered throughout the sequence, although the sample is only representative on the levels excavated horizontally. Several species have been identified, many of them common to levels A, B and Bb, such as Pinus sp, Pinus alepensis, Juniperus and Quercus ilex/coccifera. Their differing degrees of presence on each level and the presence/absence of other species, such as the presence of Acer sp on level A alone and the abundance of conifers and Juniperus sp on levels B, Bb and C1, are proof of climate variations between the different occupation periods. Different evidence of the use and processing of iron oxide has been documented in these series, with the exception of level A. Fragments of this mineral have been found, mostly burned or impregnated in the surface of different species of marine mollusc fauna, a considerable Manuel Vaquero *,**, E. Susana Alonso Fernández **, Ethel Allué **, James L. Bischoff***, Francesc Burjachs**,****, Josep Vallverdú** Conclusions The chrono-cultural sequence in the Cativera rock shelter has provided reference data for the study of the cultural processes between the Late Glacial and the early Holocene in northeastern Iberia. This period has been subject to reinterpretation since the emergence of new archaeological records which, in conjunction with the revision of existing assemblages and datings, permits the construction of a scenario in which the apparent persistence of micro-blade technocomplexes until the start of the Holocene overlaps with the few sauveterroid records, coinciding with the Pleistocene-Holocene transition and an increasingly numerous and better defined sets of notches and denticulates, evidence of a clear temporal and cultural break. The background to this cultural and chronological discontinuity can only be interpreted with new archaeological data and their social, economic and demographic interpretation. Cinglera del Capelló sites (Capellades, Barcelona) Cinglera del Capelló is a 1.5 km long travertine scarp on the right bank of the Anoia River, Capellades, Anoia District, Barcelona Province. At this point, the Anoia River cuts through Cat* ** *** **** number of endscrapers and some natural limestone blocks. It also appears in patches in the sediment. The use of this mineral has also been detected on level C4, where remains of staining material concentrated in the centre of a limestone pebble has been identified, with a set of straight and curved red lines connected physically. alonia’s pre-coastal mountain range in a gorge (Capellades Narrows), a natural link between the inland districts of the Ebro Depression and the Catalonian Pre-Coastal Depression. There are Àrea de Prehistòria, Universitat Rovira i Virgili, Avinguda Catalunya 35, 43002 Tarragona Institut Català de Paleoecologia Humana i Evolució Social (IPHES), Escorxador s/n, 43003 Tarragona US Geological Survey, ms/470, 345 Middlefield Rd, Menlo Park CA 94025 Institució Catalana de Recerca i Estudis Avançats (ICREA), Barcelona
238  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  The ...
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER. several rock shelters in the travertine wall, many of which contain evidence of prehistoric occupation. Most of these sites were first discovered and excavated by Amador Romaní in the early 20th century (Bartrolí et al., 1995). Many only contain material from late prehistory, but in some cases, documented sequences contain Palaeolithic and Mesolithic occupation levels. The latter are the focus of the Abric Romaní-Cingles del Capelló research project, begun in 1983 and still in progress. The most important of these sites is undoubtedly Abric Romaní, the subject of another article in this volume and therefore not discussed here. This paper presents three other shelters in the Cinglera area which have been excavated in recent years: Balma dels Pinyons, Balma de la Costa de Can Manel and Abric Agut (Fig. 1). The sequences documented at these sites mainly cover the second half of the Upper Pleistocene and the start of the Holocene. They provide information about the occupation of Cinglera during the Upper Palaeolithic and Mesolithic, thus complementing the Abric Romaní stratigraphy and permitting the reconstruction of the sequence of environmental and cultural changes for the period from 100 to 7 ky BP, i.e., from the bottom of Romaní to the top of Agut (Bartrolí et al., 2008, Vaquero et al., 2013.). 1. Balma del Pinyons This site was discovered by Amador Romaní. In 1905 he carried out the first excavation, only at the top of the sequence, and identified an archaeological level with several stone items. Between 2000 and 2001, a 3 x 2 m test pit (Vaquero, 2004) documented a sequence at least 6 m deep, dated by several methods (U/Th, luminescence, C14 AMS) at between ca. 52 ky BP and the start of the Holocene (Table 1-3). The stratigraphic sequence includes sedimentary deposits which are not linked to their formation by precipitation of water-borne calcium carbonate. These levels indicate quite different conditions from the predominant climate in the sequences found in the majority of the Cinglera sites. The following units have been distinguished, listed in an upward sequence (Fig. 2): – Unit 7 (2 m). Succession of layers of travertine with a dome morphology, interstratified with sands and gravels. Figure 1. Location of Cinglera Capelló sites (A) and overview of Cinglera (B). 1. Abric Romaní, 2. Balma dels Pinyons, 3. Abric Agut, 4.Balma de la Costa de Can Manel. – Unit 6 (0.5 m). Red silty sands. Contains archaeological level C, dated at 43,185 to 41,745 cal BP. – Unit 5 (0.8 m). Poorly stratified travertine, gravel and sand. 239
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER.  several rock shelters in the travertine wall, many of which contain eviden...
240 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Site Lab. Cod. Stratigraphic level U ppm Agut 00-43 4.4 0.84 2.90 15 7731 ± 370 Agut 00-51 4.4 1.00 2.27 17 9376 ± 453 Agut 00-50 4.6 0.98 2.42 16 9875 ± 390 Agut 00-113 4.6 0.90 2.59 30 10863 ± 326 Agut 00-263 4.6 1.2 2.48 9 10905 ± 375 Agut 01-135 4.8 1.41 2.68 3 14274 ± 200 11064 ± 1480 Agut 00-63 4.8 1.93 1.76 11 13633 ± 527 12672 ± 1200 Can Manel U/238U 234 230 Th/232Th BP date years 01-136 1 1.3 2.85 40 01-134 1 0.6 2.71 3 11900 ± 1600 Can Manel 00-262 2 1.1 2.62 4 13600 ± 700 Can Manel 07-20 5 0.55 2.44 17 33742 ± 260 Can Manel 07-8 6 0.7 2.37 46 35954 ± 239 Can Manel 07-7 Base 0.6 2.33 7.6 39388 ± 264 Pinyons 00-259 1 0.9 2.37 228 8800 ± 600 Pinyons 00-258 1 0.6 2.87 4 9000 ± 2000 Pinyons 00-257 5 1.0 2.15 143 32800 ± 1000 Pinyons 02-15 7 1.7 2.29 96 48100 ± 1300 Pinyons 02-16 7 0.8 2.26 35 9995 ± 650 12700 ± 160 Can Manel BP date years corrected 52700 ± 1900 Table 1. Uranium series dating of travertine samples from Agut, Can Manel and Pinyons. – Unit 4 (0.8 m). Red silty sands, carbonate sands, gravels and blocks. – Unit 3 (1.2 m). Poorly stratified red silty sands, carbonate sands and weathered gravels. Includes archaeological level B. – Unit 2 (0.4 m). Red silty sands and fallen blocks. Archaeological level A is at the top. Unit 1 (1.7 m). Laminated and stratified travertine. Figure 2. Lithostratigraphic columns in Pinyons, Can Manel and Agut, showing stratigraphic units described in the text and the archaeological levels. a. Travertine, conglomerates and carbonate sands; b. Bryophyte bioconstruction; c. Travertine and angular gravel; d. Algal laminated travertine; e. Red sands and silts; f. Stalagmitic dome; g. Boundary surfaces; h. Unit limits. In this sequence, three horizons with evidence of human occupation have been identified, although their archaeological content is quite poor in general and composed exclusively of lithic artifacts and charcoal. The bone record has not been preserved. Level A corresponds to the description by Amador Romaní at the top of the site, immediately below the early Holocene travertine. The excavated area was almost completely plundered by clandestine diggers, and only a few isolated items were recovered. The second archaeological level (level B) was identified at the top of unit 3. It yielded a small assemblage of lithic remains, with evidence of an in situ knapping sequence, although some of the
240  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Site...
MEDITERRANEAN BASINS. NORTH OF THE EBRO RIVER. material seemed to be in a secondary position. The lack of diagnostic elements precludes an accurate chronocultural estimate beyond its attribution to the Upper Palaeolithic on the basis of a few blade elements and the radiometric data. There is somewhat more abundant material on level C, whose technological and dating features suggest its attribution to the Middle Palaeolithic. In any case, the available data indicate quite sporadic and ephemeral human occupations throughout the sequence. 2. Balma de la Costa de Can Manel Unlike most Cinglera sites, this shelter was not documented by Amador Romaní. It was discovered as a consequence of the new Capellades Archaeological Park. Excavations between 2003 and 2006 consisted an 18 m2 survey which yielded a stratigraphic sequence at least 9 m deep. As in the case of Pinyons, this survey did not reach the base of the site. U/Th and 14C AMS dating indicate that this sequence lies between ca. 12 and ca. 39 ky BP (Table 1 and 3). The following stratigraphic units were differentiated (Fig. 2) : – Unit 7. Poorly stratified red silty sands ly- ing directly above a dome-shaped travertine layer. Includes archaeological level F. – Unit 6 (1.5 m). Carbonated micro-stratified sands, travertine with fallen blocks and red silty sands. Includes archaeological level E, with two AMS 14C dates, between 29,805 and 31,485 cal BP. – Unit 5 (0.5-1 m). Stratified carbonate sands and gravels with blocks and travertine blocks. – Unit 4 (0.4 m). Stratified carbonate sands and gravels with red silty sand. Includes archaeological level D, dated at 18,645 to 18,050 cal BP. – Unit 3 (2.5 m). Poorly stratified gravels, car- bonate sands