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 ...
5 mediterranean basins. centre. NORTH-WESTERN ATLANTIC BASINS. 317
5  mediterranean basins. centre.  NORTH-WESTERN ATLANTIC BASINS.  317
Site Map numbering Site Abric de El Pastor 56 Cova del Bolomor 57 Casa Corona 58 El Collado 59 Cova Beneito 60 Coves Santa Maira 61 Cova Foradà 62 Cova Negra 63 Cueva de la Cocina 64 La Cueva Negra del Estrecho Río Quípar 65 El Salt 66 Cova e les Cendres 67 Malladetes 68 Cova Matutano 69 Parpalló 70 Map numbering Sima de las Palomas del Cabezo Gordo 71 Tossal de la Font 72 Tossal de la Roca 73
Site  Map numbering  Site  Abric de El Pastor  56  Cova del Bolomor  57  Casa Corona  58  El Collado  59  Cova Beneito  60...
MEDITERRANEAN BASINS. CENTRE. C.M. Hernández*, B. Galván*, C. Mallol*, J. Machado*, F.J. Molina**, L. Pérez***, J.V. Morales****, A. Sanchís*****, P. Vidal****, A. Rodríguez****** Abric de El Pastor in the Neanderthal Occupation of the Alcoy valleys, Alicante (Spain) Abric de El Pastor is a Middle Palaeolithic site located in the Miocene limestone hills forming Serra Mariola in the mountains of Alicante, at 820 m above sea level. It is located on the right bank of the ravine Barranc del Cinc, on the upper part of the hillside, facing north-east and with a surface area of nearly 60 m2. It is an ancient karst tube formed in a bed of conglomerate limestone, in the Tortonian bioclastic calcirudites. 1. M. Brotons’ excavations The first archaeological excavations in Abric de El Pastor were carried out in 1953 by M. Brotons, a bookseller from Alcoy interested in the local archaeology and who received technical advice from Professor F. Jordá for his work, as reflected in the documentation held in the C. V. M. Municipal Archaeological Museum. This consists of numerous letters dealing with methodology with recommendations about the excavation and the treatment of the archaeological record, also of historiographical interest because of references to the Middle Palaeolithic in Mediterranean Spain; in addition to the excavation logbook, stratigraphic sketch, inventory and drawings of the finds. The stratigraphic notes and some photos have been able to determine that the excavated area was in the central part of the rock-shelter and only affected the upper part of the deposit (about 50cm). Brotons thought he had reached the end of the sequence. Indeed, it was generally believed that the excavation had exhausted the potential * of the deposit, which had yielded rich lithic and faunal assemblages. 2. Current research in Abric de El Pastor The study of the site re-started in 2005 with an examination of the objects found by Brotons and a new series of excavations, initially aimed at establishing their stratigraphic context and the chronocultural sequence, as far as its state of conservation permitted, within a full pluri-disciplinary study of the Neanderthal occupation of the mountains of Alicante (Galván et al., 2007-08, 2008; Morales and Sanchís, 2009; Molina et al., 2010). The lithic assemblage recovered by Brotons is abundant and varied (2,430 pieces), formed by varieties of detritic flint from the surroundings of the site (Molina et al., 2010). Complete chaînes opératoires were identified, above all Levallois, and even several refits were achieved. This attests knapping in situ and the integrity of the lithic assemblage (Galván et al., 2008). The retouched elements form a large proportion of the assemblage (n = 543) with a predominance of side-scrapers and a significant group of pointed objects (n = 64), of which 27 display diagnostic impact fractures (Fig. 1a and b) (Galván et al., 2007-08). The archaeozoological record from the old excavations is formed by a faunal assemblage with an anthropic origin, with a choice of prey consisting mainly of ibex, red deer and horses, as well as the Mediterranean tortoise, which was a common resource (Morales and Sanchís, 2009). U.D.I. de Prehistoria, Arqueología e Historia Antigua. Grupo de Investigación Sociedades Cazadoras Recolectoras Paleolíticas. Universidad de La Laguna ** Dpto. de Prehistoria, Arqueología, Historia Antigua, Filología Griega y F. Latina. Universidad de Alicante. *** Dpto. de Historia e Hª del Arte, IPHES. Universitat Rovira i Virgili **** Dpto de Prehistoria y Arqueología. Universidad de Valencia ***** Museo de Prehistoria de Valencia. S.I.P. ****** Dpto. de Prehistoria, Historia Antigua y Arqueología. GEPEG. Universidad de Barcelona. 319
MEDITERRANEAN BASINS. CENTRE.  C.M. Hern  ndez , B. Galv  n , C. Mallol , J. Machado , F.J. Molina  , L. P  rez   , J.V. M...
320 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. Lithic material: a and b, pointed artefacts with impact fractures from the Brotons collection; c: refit in Mariolatype flint, from SU IVb, demonstrating the almost total reduction of a block.
320  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
MEDITERRANEAN BASINS. CENTRE. 2.1. The stratigraphic sequence The current excavations cover an area of 42 m2 (70% of the total surface area). The sedimentary sequence uncovered to date is 1.5m thick and has been divided into six stratigraphic units, according to macroscopic and micromorphological textural parameters (Fig. 2a and b). The main origin of the sediment is the disgregation of the conglomerate bed with cobble-stones and quartzitic fossil limestone gravel that outcrops in the rock-shelter. The matrix consists of micritic calcite in a good state of conservation, with little evidence of solution. In general there is very little clay, which is only present as fine coverings and filling the porosity. From top to bottom, the sequence is as follows: – SU VI: This has only been identified in a stratigraphic sounding; it is still in the process of being excavated and its thickness is unknown. It is a silty-sandy sediment, with 40% of clasts and gravel, very similar to the following unit. It has yielded a dense accumulation of materials (flint, fauna and charcoal) and several hearths, next to the wall of the rock-shelter (Fig. 2b). – SU V: This has also been identified in the stratigraphic sounding. The silty-sandy sediment, with a proportion of gravel similar to SU VI, is about 25cm thick. It displays evidence of illuviation and repeated reducing-oxidising processes, indicating a regime of greater humidity than in upper layers (Fig. 2b). – SU IV: Identified all over the excavated area, with a mean thickness of about 70cm, it consists of seven sub-units in which discontinuous layers of cobble-stones (IVa, IVc, IVe and IVg) alternate with clayey-sandy silts, sand and fine-grained gravel (IVb, IVd and IVf). Several remains of human occupations, probably of a short duration, have been discriminated in it (Machado et al., 2013) (Fig. 2a and 2b). – SU III: This stratigraphic unit has equally been recognised all over the excavation area, with a thickness varying from 3 to 12cm. It consists of rounded gravel with no matrix, which suggests a period of sub-aerial exposure, in which the rock-shelter was affected by run-off processes. The archaeological remains are very scanty (Fig. 2b). – SU II: This is a relict sediment occurring in the north-west sector of the excavated area, trun- cated by SU I. Its calcitic matrix, cemented in places, contains a large number of solution cavities and its appearance is powdery, due to abundant spots of organic material. The archaeological evidence in this unit is limited to a faunal assemblage associated with a hearth. – SU I: Several episodes of Holocene sedimentation, in which six sub-units (Ia to If) have been discriminated, are all in an erosional contact with the Pleistocene deposit. They fill large hollows (4m in diameter), connected with animalherders’ use of the rock-shelter. They contain Middle Palaeolithic remains in a secondary position, from the erosion of the upper part of the Pleistocene sequence, some of which is conserved as breccia, adhered to the wall up to 0.7m above the modern surface (Fig. 2a). Most of the remains retrieved by Brotons must have come from SU I, whose characteristics coincide with this researcher’s description of his Level A, although he did not realise this was a Holocene layer with Palaeolithic objects in a secondary position. His excavations also partially affected the stratigraphic units II, III and IVa. 2.2. Hearths The excavations at Abric de El Pastor have been able to study one hearth in SU II, two in IVa, three in IVb, one in IVc and five in IVd, as well as further evidence in SU VI. In general, they occupy the centre of the rock-shelter, near the entrance and beneath the line of the overhang. They vary from 0.3 to 1m in diameter. The ash has only been conserved minimally, as shown by micromorphological analysis and FTIR, and therefore these hearths are recognised by their thermal impact on the substrate, whether it is rocky or sandy. They played a structural role in the organisation of the area (Machado et al., 2013) and are characterised by a very rich anthracological record attesting the use of a wide range of plants in the activities connected with the fires (Vidal, per. comm.). 2.3. Human occupations The excavation over the area of SU IV has made it possible to recognise certain aspects of the dynamics of the occupation by the groups who lived in the rock-shelter. SU III and II are archaeologically very poor and affected by Brotons’ excavation and the animal-herders’ activities. Similarly, SU V and VI are only known partially in the sounding. However, the 321
MEDITERRANEAN BASINS. CENTRE.  2.1. The stratigraphic sequence The current excavations cover an area of 42 m2  70  of the ...
322 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 2. a) Stratigraphic sequence SU I to IVd. b) Stratigraphic sequence in the sounding from SU IVd to VI.
322  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
MEDITERRANEAN BASINS. CENTRE. objects found in SU IVa, IVb and IVc are evidence for between four and six brief occupations of low intensity (Machado et al., 2013). The absence of thermoaltered flint in these units susceptible to dating by TL has not allowed a chronological determination, and therefore it can only be said that they are older than 75 ± 10ka, the result of a TL determination for materials in the Brotons collection (N. Mercier, per. comm.). All these occupations are very similar in the use of the space (central hearth or hearths, activities around them and an empty space at the back of the rock-shelter). However, they are different in terms of the importance of the activities carried out. In SU IVa and IVc, animal processing predominated, especially the consumption of Testudo hermanni, whereas knapping tasks are practically unknown. In contrast, these tasks predominate in IVb (Fig. 1c). SU IVd, Josep Fernández Peris*, Virginia Barciela**, Ruth Blasco***, Felipe Cuartero****, Laura Hortelano*****, Pablo Sañudo****** Acknowledgements This communication forms part of the research being carried out in the framework of the project: I+D+I HAR2012-32703, The disappearance of Neanderthal groups in the central region of Mediterranean Iberia. A methodological proposal for an approach to the historical process and the palaeo-environmental background (MINECO-FEDER). We would like to thank Camilo Visedo Moltó Archaeological Museum, Alcoy Town Corporation and the General Directorate for Cultural Heritage of the Government of Valencia for their support of the archaeological research. Bolomor Cave (Tavernes de la Valldigna, Valencia, Spain) Location and geographical context Bolomor cave is located on the southern slope of the Valldigna valley, on the right bank of Bolomor Canyon, near the Valencian town of Tavernes de la Valldigna. Valldigna is a flat-bottomed valley with an east-west orientation. The river Vaca runs through the valley, nourished by the springs of the nearby limestone massifs. The cavity, about 100 m above sea level, is part of the series of karst forms along the northern side of the Mondúver massif, opposite the limestone alignments of the Sierra de Les Agulles and the Sierra de Corbera. The coast lies northeast of the site, almost * currently being excavated, reflects a similar pattern, although with a larger number of hearths. perpendicular with the valley, whose base extends almost at sea level. The cave was formed by the intense karstification and fracturing of the Santonian limestone in a northeast-southwest direction. It is located on a vertical cliff that forms part of the southern flank of the tectonic depression of the Valldigna. It has a surface area of approximately 35 × 17 m (600 m2). Its Universal Transverse Mercator (UTM) coordinates are scale 1:25,000: ETRS89 and area 30N 737919E 4329998N, according to sheet no. 770-4 of the National Geographic Institute. Valldigna is located in one of the wettest areas of Valencia, with a typical Mediterranean climate and an annual rainfall of over 700 mm. The regional to- SIP (Servei d’Investigació Prehistòrica), Museu de Prehistòria, Diputació de València, C/Corona, 36, 46003 Valencia, España. E-mail: josep.fernandez@bolomor.com ** Departament de Prehistòria, Universitat d’Alacant, Carretera de Sant Vicent del Raspeig, S/N. 03690 Sant Vicent del Raspeig, España. E-mail : virginia.barciela@gmail.com *** The Gibraltar Museum, 18-20 Bomb House Lane, PO Box 939, Gibraltar. E-mail: rblascolopez@gmail.com **** Departamento de Prehistoria y Arqueología, Universidad Autónoma de Madrid, Laboratorio de Arqueología Experimental, Campus Cantoblanco, 28049 Madrid, España. E-mail: felipe.cuartero@uam.es ***** Departament de Prehistòria i Arqueologia, Universitat de València. Blasco Ibáñez, 28 46010 Valencia, España. E-mail: lahorte25@hotmail.com ****** Universitat Rovira i Virgili (URV), Campus Catalunya, Avinguda de Catalunya, 35, 43002 Tarragona, España. E-mail: pablo.sanudo@bolomor.com 323
MEDITERRANEAN BASINS. CENTRE.  objects found in SU IVa, IVb and IVc are evidence for between four and six brief occupation...
324 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD pography plays an important role in this record: the mountains advance towards the sea and block the humid winds from the northeast, thus favouring rainfall. The vegetation surrounding Bolomor Canyon has adapted to summer droughts and is evergreen type with small, leathery leaves. No more than 100 years ago, Mondúver was covered by a dense holm oak wood, but today, crop fields and shrubs, including rosemary, thyme, Cistus and Chamaerops, predominate. The holm oaks have disappeared as a result of a long process of degradation. However, Bolomor Canyon is a magnificent example of a vegetation refuge in which flowering ash (Fraxinus ornus), laurus (Viburnum tinus), Mediterranean honeysuckle (Lonicera implexa), sarsaparille (Smilax aspera), Mediterranean buckthorn (Rhamnus alaternus), Cade juniper (Juniperus oxycedrus) and Phoenician juniper (Juniperus phoenicea) grow as representatives of the ancient Mediterranean forest. History of excavations and development of research The importance of the cavity as an archaeological and paleontological site was recognized in the 19th century in the references of Juan Vilanova i Piera in 1868. However, it was geologist Leandro Calvo who provided specific data, including a brief description of the stratigraphy (more details are given in Barciela et al., 2013). The inhabitants of Valldigna at the time already knew about the cavity and for the inhabitants of the village it was where the “Cementerio de los Moros” was located. During the first half of the 20th century, various naturalists and researchers visited the cavity. These included Henri Breuil, who deposited the first materials extracted from the cave at the Institut de Paleontologie Humaine of Paris in 1913. Then, the Comisión del Colegio de Doctores of Madrid explored the cavity in 1923 in search of human remains. However, from 1930, the thick sheet of stalagmites at the base of the cave was mined, affecting a significant part of the archaeological site. This work mainly destroyed the central half of the sedimentary filler, exposing an important sequence over 14-m thick. For this reason, the current excavation is being carried out in the side and central cores, creating an artificial division of the site in the areas of excavation depending on the geographic orientation: the Western, Eastern and Northern sectors. Since the end of the 1980s, the syntheses of the Middle and Lower Palaeolithic in Valencia highlighted the significant research opportunities that this site could offer. But it was not until 1989, that the current phase of systematic research and excavation began, co-directed firstly by Fernández Peris and P. Guillem Calatayud and directed by J. Fernández Peris since 2008 through the Prehistoric Research Service (SIP) of the Regional Council of Valencia and the Ministry of Culture of the Regional Government of Valencia. Stratigraphy, paleoclimatic phases and dating The sedimentary filler in the cave of Bolomor mainly consists of colluvial allochthonous material, which has been deposited through open channels in the walls and ceilings. In addition to these contributions are local gravitational contributions from clastic rock landslides or weathering processes. The sedimentary series that rests on the cretaceous rock starts with thick lithochemical levels in the form of stacked sheets of calcite. In these, layers of pure crystallized carbonate alternate with others that include detrital rock without fauna remains. This deposition covers the whole floor of the cavity from the entrance to the innermost point with castings adapted to an irregular topography and stalagmite formations from clastic rock drops. New materials are stacked over this first layer of filling with a subhorizontal overhang and variable thickness depending on the location (between 4 and 8 m). The karstic deposition of Bolomor cave presents paleoclimatic implications that come from the data based mainly on the sedimentology (Fumanal, 1993, 1995; Guillem, 1995, 1996; Martínez Valle, 1995, 2001; Fernández Peris, 2007). These values translate into a sequence that is summarised from base to ceiling in four paleoclimatic phases (Fig. 1): – Bolomor phase I (levels XVII, XVI and XV): This is a baseline sequence corresponding to a mild climate cycle with a certain level of humidity, at least seasonally, during which materials from external areas accumulated and sediment brecciation is recorded. Level XV offers a gradual separation of the connection. This is the base with the introduction of materials that were weathered in areas at a certain distance from their final point of disposition. The coeval atmosphere seems to have been mild (without the harshness of active physical weathering of the surroundings) and wet, at least seasonally, allowing the dissolution-concretion of carbonates in the environment. Dating of level XVIIa was obtained by amino acid racemisa-
324  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  pogr...
MEDITERRANEAN BASINS. CENTRE. Figure 1. Stratigraphy and dating of Bolomor cave (Tavernes de la Valldigna, Valencia, Spain) tion (AAR) of 525±125 ky1 of the enamel of a tooth belonging to the Equus ferus. The magnetic susceptibility analyses are rejuvenated at the baseline, relegating it to MIS 9 (Ellwood, unpublished). More samples from level XVII are being analysed using AAR at the Biomolecular Stratigraphy Laboratory of Madrid. – Bolomor phase II (levels XIV and XIII): This was a climatic period with interstadial characteristics, mild-warm and seasonally very wet, which caused periodic flooding of the cave. A gradual but very pronounced change in the climatic conditions is marked by the installation of level XIVa and b. The environmental parameters changed significantly and the previous severe manifestations remitted completely, giving way to the action of gentle water streams that introduced certain material and a probable differentiation of soil horizons, given that the vertical migration of carbonates in this unit is evident. It was a mild-warm phase and seasonally very wet. Net contact with the next level (XIII) suggests that the overhead stalac1 Dating performed by G. Belluomini, Dipartimento di Scienzedella Terra from La Sapienza University, Rome. tite covering collapsed. Two dating values were obtained for level XIV using thermoluminescence (TL): 233±35 ky and 225±34 ky2. And at level XIII, the AAR dating values provided a chronology of 228±53 ky3. Therefore, Bolomor Phase II is placed within MIS 7. For level XIV, two thermoluminescence (TL) dating values were obtained: 233±35 ky and 225±34 ky4. In level XIII, the AAR dating values provide a chronology of 228±53 ky5. Therefore, Bolomor phase II is situated within the MIS 7. – Bolomor phase III (levels XII to VII): This was a climate episode with cool and wet os2 3 4 5 Dating performed by W. Stanska-Prószzynska and H. Prószzynska-Bordas, Sedimentology Laboratory at the Facultyof Geography and Regional Sciences, University of Warsaw, Poland. Dating performed byT. Torres, Biomolecular Stratigraphy Laboratory, Madrid. Dating performed by W. Stanska-Prószzynska and H. Prószzynska-Bordas, Sedimentology Laboratory at the Facultyof Geography and Regional Sciences, University of Warsaw, Poland. Dating performed by T. Torres, Biomolecular Stratigraphy Laboratory, Madrid. 325
MEDITERRANEAN BASINS. CENTRE.  Figure 1. Stratigraphy and dating of Bolomor cave  Tavernes de la Valldigna, Valencia, Spai...
326 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD cillations, which gradually evolved towards a harsher and drier situation (level XII). Then, this situation seems to have remitted, with a mild and very wet climate finally setting in (level VIII-VII). In level XII, AAR dating values were obtained that put it in an approximate chronology of 185 ka3. It is therefore placed between the late MIS 7 and MIS 6. – Bolomor phase IV (levels VI to I): This represents the upper section of the sequence with mild and wet oscillations typical of the last interglacial period. It formed a globally mild period, with slightly accentuated cool periods (levels VI to III), which led to the accumulation of small rock fragments as a result of weathering of the cavity’s dome by the action of cryoclastic processes. High humidity caused partial flooding of the cave and brecciation of the sediment. Also, a climate entailing cyclic conditions seems to have been widespread; that is, there were even cooler periods (levels VI, V-IV-III) during which small subangular detrital material accumulated and was then disrupted, giving way to a flow of constant and highly carbonated streams that brecciated the ceiling of levels VI, III and Ic. In general, it was a globally mild period with slightly accentuated deteriorations, during which high and oscillating levels of humidity persisted. This phase is linked to MIS 5e. An absolute dating of 121±8 ka2 was obtained via TL for level II. M. P. Fumanal’s sedimentology studies were continued by Brooks B. Elwoood of Louisiana State University. He obtained a magnetic susceptibility event curve (MSEC) that covered the entire chronostratigraphy. The correlation with the sedimentary cycles was satisfactory and added higher paleoclimatic value to it. Magnetic susceptibility (MS) relates the magnetic properties preserved in the sediments of Bolomor cave with the paleoclimate estimates and correlations with other sites. Using MS and cyclostratigraphy (CS) presents a paleoclimatic structure based on this method (MSEC) through graphic correlation with the marine oxygen isotope record (MIS). In Bolomor cave, this study included a temporary continuous sequence from MIS 9 to MIS 5e. The fauna record The fauna identified at Bolomor cave has resulted in the identification of 20 species of macromammals; the distribution throughout the sequence and relative frequencies indicate the development of slight changes to the environment surrounding the site. The biostratigraphic sequence is mainly characterized by the presence of red deer (Cervus elephus) and wild horse (Equus ferus), and more occasional records at certain times of species such as aurochs (Bos primigenius), fallow deer (Dama sp.) tahr (Hemitragus sp.), megaloceros (Megaloceros giganteus), steppe rhinoceros (Stephanorhinus hemitoechus), wild boar (Sus scrofa), Barbary macaque (Macaca sylvanus), wild ass (Equus hydruntinus), elephant (Palaeoloxodon antiquus), hippopotamus (Hippopotamus amphibius) and beaver (Castor fiber) (Table 1). The presence of carnivores in the cavity is sporadic, both in anatomical representation and frequency in the bone record. Fossil remains of Ursus arctos, Ursus tibetanus, Canis lupus, Panthera leo, Lynx pardina, Vulpes vulpes and Meles meles have been identified (Martínez Valle, 1995, 2001; Sarrión and Fernández Peris, 2006). It is also important to highlight the presence of small animals such as rabbit (Oryctolagus cuniculus), hare (Lepus sp.), birds (Passeriformes, Galliformes, Corvidae, Columbidae, Phasianidae, Anatidae), tortoise (Testudo hermanni) and occasionally fish (Salmonidae), throughout the sequence. The proportion and relative frequencies vary from level to level (e.g., Blasco et al., 2013; Sanchís Serra, 2010). Human consumption of small prey, in the form of cut marks, cremation, deliberate fracturing and human bite marks have been identified throughout the sequence. In some cases, the proportion of these animals exceeds 60% in MNI, as is the case in sub-level XVIIc (Blasco et al., 2013) (Fig. 2). Micro-vertebrates The micro-fauna sampling performed by Guillem (1995, 2001) provided a total of 1,124 individuals from 12 species of micro-mammal (insectivores and rodents) for the western sector of the site: Erinaceus europaeus (common hedgehog), Sorex minutus (Eurasian pygmy shrew), Sorex sp., Neomis sp. (tailed shrew), Crocidura suaveolens (lesser white-toothed shrew), Talpa europaea (mole), Sciurus vulgaris (squirrel), Eliomys quercinus (garden dormouse), Allocricetus bursae (hamster), Arvicola sapidus (water vole), Microtus brecciensis (Cabrera’s vole) and Apodemus sp. (field mouse). None of the species found in Bolomor indicate extremely cold conditions, which coincides with a peri-Mediterranean environment that softens Quaternary climate oscillations. Allocricetus bursaeis the taxon that is most closely connected to dry and cold con-
326  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  cill...
MEDITERRANEAN BASINS. CENTRE. NISP* Ia Ib-c II III Carnivora indet. 2 3 V VI VII VIII XI XII XIII 2 2 5 1 2 Ursus arctos Canis cf. lupus IV 1 Macaca sylvana 1 XV XVIIa XVIIc 1 1 2 2 2 Vulpes vulpes 2 4 2 Panthera leo spelaea 1 3 2 Lynx pardina 1 3 Meles meles 2 Castor fiber P. antiquus 2 4 1 S. hemitoechus 2 1 28 4 1 Equus hydruntinus 3 2 1 H. amphibius 4 2 Sus scrofa 17 1 Cervidae indet. 20 4 5 3 8 1 165 11 41 77 56 5 2 10 8 16 3 46 2 7 1 2 2 2 2 2 6 1 65 2 3 1 Equus ferus 2 115 3 4 2 9 3 1 2 91 271 18 7 55 647 18 Bos primigenius 146 11 12 61 213 16 3 Capridae indet. 36 2 4 5 19 5 4 27 13 51 50 177 132 2 35 2 1 24 22 7 4 12 23 28 6 262 3 325 5 6 Cervus elaphus 17 55 16 Dama sp. 5 4 Megaloceros giganteus 135 182 1156 620 457 1 1 Hemitragus bonali Hemitragus cedrensis 4 1 1 4 121 2 Oryctolagus cuniculus 167 28 5 52 789 297 1 Lepus sp. 1 5 Passeriformes 25 13 5 Galliformes 19 8 8 Phasianidae 24 18 Anatidae 9 10 4 1 Cygnus olor Anas sp. 29 Aythya sp. 34 Corvidae 20 Pyrrhocorax sp. 6 Columba sp. 34 Strigidae 21 16 202 1 Aves indet. 32 3 Testudo hermanni 465 10 17 9 67 18 526 84 Bufo sp. 2 4 4 Pisces 2 22 4 4 1 1 Table 1. Number of specimens identified throughout the stratigraphic sequence of Bolomor cave. *NISP = Number of Identified Specimens. The data on levels Ia, Ib-c, II, III, V, VI, VIII, XIII and XV were extracted from Fernández Peris (2007) and the data on levels IV, XI, XII, XVIIa and XVIIc were taken from Blasco (2011). The data on Lepus sp. were extracted from Sanchís and Fernández Peris (2011). 327
MEDITERRANEAN BASINS. CENTRE.  NISP   Ia  Ib-c  II  III  Carnivora indet.  2  3  V  VI  VII  VIII  XI  XII  XIII  2  2  5 ...
328 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 2. a) Cut marks on a rabbit (Oryctolagus cuniculus) tibia from sub-level XVIIc; b) cut marks on a swan (Cygnus olor) tibia from level XII; c) cut marks on a red deer (Cervus elaphus) femur from sub-level XVIIc; d) cut marks on a rabbit (Oryctolagus cuniculus) jaw bone from sub-level XVIIc; e) cut marks on tortoise (Testudo hermanni) remains from level IV; f) cut marks on a long bone from a large animal from level IV. ditions, along with the Sorex minutus, which had environmental requirements like those of the current Medio-European, along with the Erinaceus europaeus and Talpa europaea that penetrate certain Mediterranean environments inhabited by the Microtus brecciensis. The rest of the species were ubiquitous and related to strict requirements, such as the formation of woods (Sciurus vulgaris) or the presence of waterways (Arvicola sapidus). Micro-mammal remains recovered from Bolomor cave were mainly introduced by small carnivores (Vulpes vulpes) depositing droppings in the cavity and by owls (Strix aluco) regurgitating the parts of the prey that their stomachs could not digest (hair, bones, etc.). Other biotic remains The presence and malacology study of small bivalve seashells contributes to the assessment of possible changes to the coastline and their relationship with climatic oscillations. The origin of these elements could be deposition in the form of pellets by seabirds (cormorants, shearwaters, storm petrels, seagulls, etc.) that feed on fish and shellfish. These small molluscs were documented at levels V, VII, XII, XIII, XIV and XV of the sequence, and their characteristics indicate moments of mild climate and great- er proximity to the sea (Fernández Peris, 2007). This documentation also includes remains of teeth and fish spines (level I, IV and XI). Other inland molluscs (snails), such as the Rumina and Melanopsis, indicate a wet environment in their corresponding levels (I, IV, VIIb and XIIIc). Unfortunately, the sediments of Bolomor cave have been pollen sterile so far (Dupré and Carrión, 2001). However, the presence of fossilised seeds throughout the sedimentary sequence of the cavity has allowed us to carry out paleocarpology analyses. These have documented the presence of two plant species: Celtis australis (European nettle tree) and Prunus spinosa (blackthorn). Both species must have colonised different places. The European nettle tree requires a certain level of humidity in the ground and is usually associated with elm woods, while the blackthorn is common in oak woods. The Celtis australis could be part of the elm wood that would circle the riverside of the Vaca (Bolomor phases II and IV) (Fernández Peris, 2007). Furthermore, the Prunus spinosa would have been associated with an oak wood that could run along the depression of the Valldigna and occupy the shade of the mountains, at least during the cooler periods of the MIS 5e (Bolomor phase IV). The recovery of thermo-modified plant remains in the form of charcoal also allowed study of the plants using anthracology. Remains have
328  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
MEDITERRANEAN BASINS. CENTRE. been recovered from the levels in which anthropic combustion structures have been identified (levels II, IV, XI and XIII). They are currently being analysed. The lithic industry The lithic industry of Bolomor is a lithic flake assemblage or techno-complex pertaining to the Middle Palaeolithic. There is little variety throughout the sequence and it is marked by the occupational characteristics of each level. However, the industries seem to have taken on a certain level of complexity in the most recent periods of the sequence, when occupation also seems have been more intense (Fernández Peris, 2007). In this respect, it is difficult to establish elements of change in the manufacturing of the knapped tools and their relationship with behavioural aspects. The main raw materials used were flint, limestone and quartzite. These came from sea, colluvial and fluvial stones from the immediate surroundings of the site and from more distant areas, such as the Xùquer basin and the Serpis basin (around 15 km from the site) (Fernández Peris, 2007; Fernández Peris et al., 2008). The technological characteristics of the industry allow it to be defined as not Levallois, not faceted and not laminar, consisting of a wide range of highly retouched lithic flake tools. There are no bifaces, cleavers, trihedral or knapped stones in the any of sequences. In general, the lithic record reflects a techno-complex of small flakes (presence of nonlaminar microliths), scrapers, denticulate and varied retouched tools. The pieces show intense reuse in the higher levels (Fernández Peris, 2007; Cuartero, 2008). Macro-tools consisting of a wide range of flake formats made of limestone, without retouching or with simple retouching, are documented in some levels only. This applies to level XII, where the lithic assemblage is almost exclusively large limestone flake with little transformation. Hearths The excavations carried out to date have provided remains of fire use in levels II, IV, XI and XIII (Fernández Peris et al., 2013) (Fig. 3). In the first level, possible ash castings between gaps have been documented; these are the result of hearth cleaning that was done in preparation for new fires. In level IV, the remains of four hearths were found, which left evidence in the form of reddish thermo-modified sediment (rubefaction layer). One of them includes stones thermo-modified at their bases. Level XI falls within MIS 6 and contains six simple hearths without internal structures. As indicated by the analyses and reproductions in progress, the hearths seem to correspond to short combustions (Fernández Peris et al., 2007). Finally, two hearths have been documented in level XIIIc. Both have clear internal structures, which show that the area and the structures were prepared before the fires were lit. Focus 1 was built on base of blocks or dallage de pierre of varying sizes and calcareous origins. These structural blocks appear in direct contact with the layer of carbonaceous sediment and are connected to the thermo-modified sediment. The preparation of focus 2 was different. It is located inside a depression or basin with blocks that run along the edge of the deepest side, with some at the bottom. This basin is 76-cm long by 41-cm wide, with a maximum depth of 5 cm compared to the flat area; it is higher at the southeast end and its depth decreases towards the northwest, where the depression is very shallow. We obtained a dating based on amino-acid racemisation of molluscs collected in the surrounding area of the hearths that gives a chronology of 228±53 ky. This makes it the oldest controlled use of fire in the Iberian Peninsula and in Southern Europe (Fernández Peris et al., 2013). The Palaeoanthropology The human fossils in Bolomor cave published to date comprise six pieces of bone and tooth. Some of the remains were found during the sieving of disturbed sediment by the quarry work carried out in the 1930s, and others have been recovered in the excavation process and, therefore, have a clear stratigraphic location. The human fossil assemblages found comprise a very small and fragmented sample in which dental elements dominate over postcranial elements (Arsuaga et al., 2013): – HCB 01: Fibular shaft measuring 48.7 mm corresponding to an adult. This remain was recovered through systematic excavation of level III (MIS 5e), with an ante quem dating of 121±18 ky. – HCB 02: Lower left molar (M1) corresponding to a child (around 5 years old). It was found in level IV (MIS 5e), with an ante quem dating of 121±18 ky. – HCB 03: Upper right deciduous molar (dm2) corresponding to a 6–9-month-old 329
MEDITERRANEAN BASINS. CENTRE.  been recovered from the levels in which anthropic combustion structures have been identi   ...
330 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 3. Hearths at Bolomor cave: a) hearths on level IV (c. 130 ky); b) hearths on level XI (170 ky); c) hearths on level XIII (250 ky) undergoing excavation; d) fine flake focus I, level XIII; e) focus I section level XIII. baby. It was recovered from the sediment modified by the quarry work. covered from the sediment modified by the quarry work. – HCB 04: Lower left canine (C1) very similar to modern humans. It came from the sediment disturbed in level I sub, wherein intrusive modern day elements have been documented. – HCB 07: Fragment of parietal bone (109 × 116 mm) from the sediment modified by the old quarry work at the site. The gap in the parietal bone could be related to the carbonated sediment from level VI, putting it in the MIS 5e with a dating of around 130 ky. It is currently being studied and the findings published. – HCB 05: Upper left canine (C1) with a very similar morphology to the dental remains from Sima de los Huesos in Atapuerca (Burgos, Spain) and from Krapina (Croatia). This piece was found in the clearance of Ia-Sub XIII in contact with the clearance of XIII. The gap attached to it could be located in level XIII. – HCB 06: Fragment of parietal bone (22 × 18 mm) with a coronal suture. This was re- All of the pieces, except HCB 05, could be assigned to MIS 5; of these, HCB 03, HCB 04 and HCB 06 could be more modern, as they were located in a clearance context with some post-Palaeolithic items. Specimen HCB 05, however, given the sedimentologic characteristics associated with it and where it was found (Clearance XIII), is re-
330  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
MEDITERRANEAN BASINS. CENTRE. lated to levels XII-XV and its location in MIS 7 cannot be dismissed. This remain is metrically different from the Neanderthals and is well related to the canines from Sima de los Huesos, in Atapuerca (Burgos, Spain), and Krapina (Croatia). According to Arsuaga and collaborators (2013), even the less common traces (presence of the entoconulid of the M1 or lack of structural details in C1), are frequently found in European human fossils from the Middle Pleistocene. Therefore, it might be attributable to the Homo heidelbergensis-Homo neanderthalensis lineage. Javier Fernández-López de Pablo* Magdalena Gómez Puche** Marco Aurelio Esquembre Bebia*** Bolomor excavation is part of the archaeological excavations program of Servicio de Investigación Prehistórica (SIP) (Museo de Prehistoria de Valencia) under the authority of Diputación de Valencia (España). Research project is financed by the Ministry of Science and Innovation CGLBOS-2012-34717. Ruth Blasco is a Beatriu de Pinós-A post-doctoral scholarship recipient from Generalita de Catalunya and co-finaced by the European Union through Marie Curie Actiones, FP7. Casa Corona (Villena, Alicante, Spain) 1. Location The site of Casa Corona is located in the town of Villena, close to the administrative border between the provinces of Alicante and Albacete, 80 km from the coast (UTM 680306 4282783, zone 30N, Datum ETRS 89, 502 m.a.s.l). The site lies at the centre of a broad natural corridor known as the Villena-Caudete plain, with an average altitude of 500 m.a.s.l. Quaternary materials, continental aeolian formations, endorheic lagoons and ponds dominate the landscape. The site covers the top of a partially cut Late Pleistocene dune (Fig. 1). The site was discovered in 2006 by the staff of the Archaeological Museum of Villena. In 2008, the company Arpa Patrimonio S.L. excavated the site over three months, in a rescue program resulting from the construction of the Madrid-Alicante AVE high-speed rail network (Fernández-López de Pablo et al., 2013). In 2013, a new excavation phase was started as a part of a multi-year fieldwork program coordinated by the IPHES. In this chapter, we will * ** *** Acknowledgements synthesize the results obtained on the Early and Late Mesolithic occupation phases (Fig. 2) from the 2008 excavations. In addition, we will complement this synthesis with new stratigraphic information from the 2013 excavations. 2. Stratigraphic record and occupation phases The 2013 excavations have characterized the succession of litho-stratigraphic units that form the site’s sedimentary context. From the bottom to top: • SU 303: a 3 m deep stratum of pale-yellow consolidated siliceous sands form the base of the archaeological stratigraphy which lacks archaeological materials. From a geomorphological and sedimentary point of view, it corresponds with a continental dune dated to the Late Upper Pleistocene (FernándezLópez de Pablo et al., i. p.) (Fig. 2 and Table 1). This correlates with the SU 127 from the 2008 Arpa Patrimonio S.L. Excavations (Fernández-López de Pablo et al., 2013). Institut Català de Paleoecologia Humana i Evolució Social (IPHES) jfernandez@iphes.cat Institut Català de Paleoecologia Humana i Evolució Social (IPHES) mgomez.puc@gmail.com Arpa Patrimonio S.L. Avenida Rodalet 23A. San Vicente del Raspeig C.P.03690. (Alicante). arpapatrimonio@gmail.com 331
MEDITERRANEAN BASINS. CENTRE.  lated to levels XII-XV and its location in MIS 7 cannot be dismissed. This remain is metric...
332 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. Aerial photograph of the site after the 2008 excavation. The Pleistocene dune (light brown) is exposed after the excavation of the plough soil. In the south (dark grey) there is a dried inter-dune depression pond. The black square indicates the area where the Early and Late Mesolithic features were located. • SU 301: Dark greyish-brown sandy and semi-compacted stratum with a variable depth ranging from 30-40 cm. It lies in gradual contact over the SU 303. Its upper limit has been post-depositionally disturbed by agricultural uses. It corresponds with an Early to Middle Holocene paleosol formed into a small talweg at the top of the dune. The SU 301 overlies most of the combustion structures and presents archaeological materials of the different ceramic and preceramic occupational phases. This corresponds with the SU 101, 102 and 103 of the Arpa Patrimonio S.L. Excavations. • SU 300: light-brown unconsolidated plough-soil stratum of 20-40 cm depth. It overlies SU 301 and contains archaeological materials from different periods. The Early (Notches and Denticulates Mesolithic) and Late Mesolithic structures and archaeological deposits have been documented at the northern sector of the site, which extends up 565 m2. This sector also records archaeological materials of subsequent occupation phases. That is the case of 6 Chalcolithic and partially conserved pit -structures (one of them is the remaining part of a hut); and a pit that contained an Early Post-Cardial Neolithic ceramic ware. The two oldest occupational phases (Early and Late Mesolithic) have yielded 16 combustion structures and two burial pits. The combustion structures are hearth pits that cut the stratum of Late Upper Pleistocene sands (SU 303). Their average dimensions are 0.7 m in diameter and 0.2 m in depth (Fernández-López de Pablo et al., 2011). The sedimentary fill of the hearth pits is formed by carbonaceous sediment of burnt sands contain-
332  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
MEDITERRANEAN BASINS. CENTRE. ing fire-cracked stones, lithic artefacts, burnt land snails and faunal remains. The radiocarbon dates obtained from two combustion structures indicate different occupational phases (Table 1, Fig.3). The chronology of the combustion structure 4 (SU 142) is the oldest one (Beta-330866, 10312 ± 40 BP). Its calibration ranges overlap the Younger Dryas. Thus, its chrono-cultural attribution should correspond with the Epipaleolithic or Epimagdalenian (in a broad sense). In contrast, the combustion struc- Figure 2. Cumulative plan of the northern sector with the location of the dated features. 333
MEDITERRANEAN BASINS. CENTRE.  ing    re-cracked stones, lithic artefacts, burnt land snails and faunal remains. The radio...
334 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD ture 8 (Beta-323497, 8520±50) is dated during the Early Holocene, in the Boreal chrono-zone and the Early Mesolithic period. New radiocarbon determinations are needed to establish the chronology of the remaining combustion structures. In addition to the combustion structures, the 2008 excavations uncovered two burial pits containing two primary inhumations: the first one, known as burial 1, is the burial of a woman aged 35-40 years old; and the second one, is the primary interment of a child aged 12-18 months (Fernández-López de Pablo et al., 2013). The radiocarbon dates on individual samples of bone collagen gave a very similar chronology; burial 1 (Beta-272856, 7070 ±40) and burial 2 (OxA- Context V-2392-27, 7116 ±32 BP). The calibration ranges of both radiocarbon dates partially overlaps the 8.2 kyr cal BP cold event. Finally, a radiocarbon date on a shell lip fragment of a land snail recovered in the filling sediment of burial pit 1 (UE 128) yielded an Early Holocene age (Beta330865, 8898±40). Such a chronology is placed into the time window defined by the radiocarbon dates of the combustion structures UE 142 and UE166. In turn, that date confirms our previous observations about the fact that the filling sediments of the burial pits incorporated debris of previous Mesolithic occupations. Despite the long occupational sequence, the record of lithic artefacts is quantitatively discrete. 14 C BP age 2 s Cal BP age AMS 10312±40* 12383-11960 Beta-330865 AMS 8898±40* 10190-9892 Quercus sp. Beta-323497 AMS 8520±50 9551-9454 LM Human bone OxA-V-2392-27 AMS 7116 ±32 8007-7886 LM Human bone Beta-272856 AMS 7070 ±40 7972-7800 Affiliation Sample Lab. Ref Method UE 142 EPIMAG S. candidissima Beta-330866 UE 128 EM S. candidissima UE 166 EM Burial 2 Burial 1 Table 1. AMS radiocarbon dates of different archaeological features at Casa Corona. (**) corrected subtraction of 210 years of local limestone effect (Yanes et al., 2013). C dates after the 14 Figure 3. Average probability of calibrated radiocarbon distributions of Casa Corona obtained with CalPal (Weniger and Jöris 2004) and the Intcal13 calibration curve (Reimer et al., 2013). Up: Periodisation model of the Iberian Mediterranean region during the Postglacial (Aura et al., 2011); Bottom: In blue 18O variations from the Greenland GISP2 ice see core (Stuiver, Grootes 2000). In red: Sea Surface Temperature estimation from the Alborán Sea (Cacho et al., 2001).
334  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  ture...
MEDITERRANEAN BASINS. CENTRE. A total number of 3,322 lithic artefacts are associated with both Mesolithic occupation phases, just 29 of them are retouched. The Early Mesolithic phase is characterized by flake debitage and the selection of thick flakes to be transformed in notched and denticulated tools. On the other hand, the Late Mesolithic phase presents 12 geometric microliths: 11 trapezes with step retouch and a “Cocina” type triangle. However, the evidences of blade debitage are meaningless so far (1 core and 76 bladelets). custrine phase 2, which witnessed short-term fluctuations with the alternation of permanent and ephemeral water bodies. On the other hand, 8 pierced Columbella rustica shells with dorsal perforation have been recovered. This species of marine gastropod is common in many ornamental assemblages dated between the Early and Late Mesolithic periods in the Iberian Mediterranean region (Álvarez, 2008). The land snail specimens of the SU 142, dated by radiocarbon during the Younger Dryas chronological span, present a d13C and d18O isotopic composition that indicate dryer and colder palaeoenvironmental conditions than those reported in the specimens recovered from the SU 166 and 128, dated during the Early Holocene (Yanes et al., 2013). On the other hand, the stable isotope analysis of d13C and d18O of Sphincterochila candidissima land snail shells from different combustion structures of Casa Corona and Arenal de la Virgen sites provide additional information on the paleovegetation and paleoathmospheric conditions at the time of the human occupations (Yanes et al., 2013). 3. Paleoecology The Upper Vinalopó valley contains two lake deposits –the Villena lagoon and the Salinas Playa lake– placed at 4 and 15 km away from Casa Corona. Both palaeoenvironmental records provide different chronological resolution about the local vegetation and the responses of hydrological systems responses to Pleistocene-Early Holocene transition climate changes and, particularly, the Bond events. Thus, the study area offers a unique opportunity of correlating the palaeoenvironmental dynamics inferred at the natural deposits with the chronological framework of the human occupations during the Postglacial period (FernándezLópez de Pablo et al., 2011a). The first occupation of Casa Corona was coeval to the Younger Dryas, an episode of climatic deterioration that is well defined in the pollen record of the Villena lagoon. The pollen zone D2 reflects a significant increase on the Artemisia and Ephedra taxa, associated with a general decrease in the tree pollen (Yll et al., 2003). In the Villena lagoon, the beginning of the Holocene corresponds with pollen subzone 3, characterized by the increase of tree pollen, mainly pine, and the abrupt decrease of the Pseudochiazea and Chenopodiaceae taxa, thus, reflecting more temperate environmental conditions. According to the ostracod assemblages of the Salinas playa lake (Roca and Julià 1997), the Early Mesolithic occupations at Casa Corona were coeval to the la- 4. Subsistence The vertebrate zooarchaeological remains are badly preserved at the Casa Corona site due to taphonomic biases. The site’s sedimentary context –dominated by siliceous sands– and the combustion processes, explain the bad preservation of bones, most of them very small (<1cm) diaphyseal fragments. At both Mesolithic phases, only 12 individuals have been identified at taxonomic level (Morales 2008). The presence of three Bos primigenius remains is qualitatively significant. On the other hand, amongst the small prey, we have identified remains of lagomorphs and turtles (Emididae/ Bataguridae). Such a taxonomic spectrum is consistent with the zooarchaeological information from other Mediterranean openair Mesolithic sites located on coastal wetlands such as El Collado (Fernández-López de Pablo et al., i.p.). Unlike vertebrates, land snails are very well represented in the 14 Mesolithic combustion structures at Casa Corona (Fernández-López de Pablo et al., 2011a). Two edible land snail species clearly dominate the land snail assemblages: Sphincterochila candidissima (90,67%, n=2844) and Iberus alonensis (4,23%, n=116). The biometric analyses based on the maximum width of the S. candidissima specimens show that immature individuals were not represented in the malacological assemblages. The spatial association of both species with 335
MEDITERRANEAN BASINS. CENTRE.  A total number of 3,322 lithic artefacts are associated with both Mesolithic occupation pha...
336 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD the combustion structures at Casa Corona and the neighbour site of Arenal de la Virgen uncover a systematic exploitation pattern of this class of invertebrate resources during the Early and the Late Mesolithic periods. The seasonal ethology of both species, which hibernate and aestivate underground or in crevices, suggest that the gathering and consumption took place during the spring and/or fall seasons. The palaeodietary evidence, based on stable isotope d13C and d15N data of bone collagen from the burials, provides complementary information about the Late Mesolithic subsistence patterns (Fernández-López de Pablo et al., 2013). The human remains present d13C (-19.3 and -18.5) and d15N (8.4 and 11.6) values compatible with a diet based on terrestrial resources. There is not a clear isotopic signal suggesting a minimal contribution of marine or fresh water proteins. However, the child individual presented a slightly higher d15N value, due to the breastfeeding effect. 5. Funerary record Two Late Mesolithic burials containing the human remains of two individual inhumations were located northwards of the area delimited by the combustion structures (Fig.4). The distance between both burials is of 7 m, which, in addition to their radiocarbon age overlapping, indicates that they were part of the same cemetery. The identification of the burial pits from the beginning of the excavation process has allowed the documentation of the individual’s position regarding the funerary space as well as their relationship with other Mesolithic occupation remains. Burial 1 is a primary individual interment of an adult woman aged 35-40 years. The skeleton has a poor state of preservation due to the acidity of the soil. However, the epiphyses of the long bones as well as the vertebral bodies are preserved, and were placed in a supine position with the legs flexed and the face towards the left. The skeleton is oriented in a NE-SW direction. The position of this individual in the burialpit, with the upper extremities flexed into the abdominal region and the shoulders hunched, is forced. The thorax is open, indicating that there was enough space around it for this to happen during the stages of decomposition. However, the position of the clavicles indicates a lack of space during decomposition. At the same time, the pelvis is totally closed, and the lower extremities flexed, still leaning against the wall of the grave itself. All these evidences –the persistence of narrow joints, the forced position of the shoulders and the location of the right arm with respect to the ribs– seem to indicate that this individual had been wrapped in a shroud. A pierced shell bead of Columbella rustica was likely deposited as a grave good. The grave pit was filled after the cadaver deposition. The filling sediments contained residues of previous Mesolithic occupations such as burnt land snails, charcoals, lithic debris and small dyaphesal fragments of faunal bones. Burial 2 is a primary individual interment of a child aged 12-18 months. The skeleton also has a poor state of preservation. The individual was placed in a supine position and oriented NW-SE. The skull is raised, with the jaw closely articulated and vertebral arches inside, denoting a position forced by the presence of a perishable object (not preserved) such as a pillow under the skull. Such an arrangement allowed the face of the individual to look towards the left shoulder, which leaned in a higher position than the right one. The orientation of the humerus denotes a forced position of the shoulders. The ribs on the right side belong to a closed thorax. Everything seems to reveal that the dead individual was placed in the pit wrapped. A flint blade, deposited by the side of the right femur, is the only grave good clearly documented. In addition to this item, some burnt land snails and lithic debris were randomly found across the burial pit fill, as well as small fragments of charcoal and burnt bones. 6. Perspectives There are different reasons that make of Casa Corona a key site for progressing in the knowledge of the open-air sites in the Iberian Mediterranean region into its palaeoecological and cultural contexts. First, because the geomorphological context of the site on the top of a continental dune located at the centre of a valley, the alluvial and colluvial sed-
336  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  the ...
MEDITERRANEAN BASINS. CENTRE. Figure 4. Late Mesolithic burials of Casa Corona. imentary processes were not responsible for the site’s formation. Thus, the resulting archaeological record of the human frequentations to this site has generated an extensive palimpsest with different occupation phases. The oldest known occupation is dated during the Younger Dryas, thus, establishing an ante quem chronology for the dune formation. Afterwards, along the Early Holocene, the site witnessed different frequentation episodes that generated an archaeological record composed of different combustion structures. Finally, during the Late Mesolithic, the site witnessed short funerary activity. Considering the site’s long occupational sequence, the identification of chronologically narrower occupation events on the basis of micro-spatial and micro-stratigraphic analyses is our main aim in future works. Also, the study of new lake records should improve the available palaeoenvironmental information and its chronological resolution on a local scale. Both research lines would enhance the interpretation of the impact of the late Pleistocene and Early Holocene environmental changes on human settlement areas. Finally, we would like to stress the relevance of the site for the knowledge of the last Late Mesolithic episodes and the Neolithisation process. In the central Iberian Mediterranean region, the radiocarbon dates of the human burials of Casa Corona (Fig. 4) fill the 7 centuries chronological gap between the last Late Mesolithic contexts and the first Early Neolithic sites. The location of new Late Mesolithic burials and its interdisciplinary palaeoanthropological study are two of the main objectives of the new phase of excavations at the site started in 2013. 337
MEDITERRANEAN BASINS. CENTRE.  Figure 4. Late Mesolithic burials of Casa Corona.  imentary processes were not responsible ...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 338 José Aparicio Pérez * El Collado (Oliva- Valencia) 1. Introduction Around 9,000 BC, a small group of maybe five or six humans with obvious signs of kinship installed their abode a crag or rock shelter near the present day Oliva, in an area known as El Collado, where they remained for about 3,500 years until the middle of the 6th millennium BC. After the excavation work, which began in 1987, we were able to complete the subsequent research process with archaeological, anthropological and radiocarbon analyses. The comprehensive anthropological study was led by Dr. Domingo Campillo Valero, Head of the Palaeopathology and Palaeoanthropology Laboratory at the Barcelona Archaeological Museum, and involved a large team of anthropologists, radiologists, restorers and photographers who studied the purely bone-related aspects such as teeth, illness and palaeopathologies, food science/ eating habits, etc., using conventional and also advanced methods such as isotopic analysis (VV.AA, 2008: 179-344). The archaeological and historical analysis was conducted by the undersigned (VV.AA, 2008: 28-91 and 347-359). The chronology defined by archaeological methods was initially backed up by four radiocarbon datings, as well as others. Two important aspects must be noted. Firstly, exploration of the site was not exhausted and much of what remains is still buried under tons of fertile soil dumped there by the property owner, Mr. Bolinches. Secondly, the site was completely disfigured from its original configuration during the occupation. Weathering processes underway from its abandonment until its discovery certainly altered it, but a much greater impact was undoubtedly caused by subsequent human action including its transforming into cropland with staggered horizontal contoured terraces, and the use of rock from around the site, either loose or dug from limestone outcrops. * On the basis of this assumption and the remains in the vicinity, we believe that this human settlement was established beneath a limestone crag with a shelter-like concavity inside or nearby, with an easterly aspect, i.e. facing the Mediterranean Sea, which during the period was probably slightly higher than its current level, with a marsh, fen or aigua-moll covering what is now the coastal plain. The site is on the slope of a terminal elevation of a low-lying mountain range, a few hundred metres above current sea level and about seven kilometres in a straight line from the coast, a little more than one metre above the start of the marsh. A basal concavity in the same place as the site, inside the probable gap in the cliff or shelter, proof of its existence, facilitated their installation on this steep hillside, which ends at a saddle or collado (hence the name) overshadowed by Montanyeta de Santa Ana, a hill between the site and the marsh. 2. Chronology We shall first clarify the chronology used as the basis for the subsequent considerations. Long before radiocarbon dating became available, we defined the chronology of this settlement on the sole basis of archaeological data, i.e. derived from the lithic material and on the basis of our own structure of the Mesolithic. Later, following anthropological studies, Dr. Campillo Valero became concerned abut this issue and arranged for two radiocarbon analysis to be conducted on human bones from burial site XIII. More recently, following further dental analyses, further datings were suggested to assist confirmation of the previous dates, which we accepted in order to dispel any shadow of doubt, although it had never crossed our minds to question the authenticity and chronology of the human remains (Tab 1 and 2). Arqueólogo. Director de la Sección de Estudios Arqueológicos y Prehistóricos de la Real Academia de Cultura Valenciana. Académico C. de la Real de la Historia. joapa2005@hotmail.com P.O.BOX 2260. En 46080 VALENCIA
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  338  Jos ...
MEDITERRANEAN BASINS. CENTRE. The subsequent analyses were also conducted in the Radiocarbon Dating Laboratory, Faculty of Chemistry of the University of Barcelona, under Dr. Joan S. Mestres, with the following results: Burial XIII (Campillo Individual XIII), from layer 4 of LEVEL II, equivalent to layer 3 of the first excavation, was performed between 7649 and 7570 BC. Burial IV, also from layer 3, was around 8690 BC, and burial VI, from the same layer, was from 8080 BC. Given that the grave was accessed from layer 2 or at least the first layers of Level II –due to the dismantling or disturbance of layer 1 and its mixture with other material during the farming and terracing operations, this layer or level can be dated as no earlier than the middle of the 6th to the start of the 7th or the middle of the 8th ID DATING UNCERTAINTY CALIBRATION 2d cal. A.C BURIAL XIII UBAR 280 7570 160 6804-6066 BURIAL XIII UBAR 281 7649 120 6799-6234 ID DATING UNCERTAINTY CALIBRATION 2d cal. A.C BURIAL IV UBAR 927 8690 100 8188-7551 BURIAL VI UBAR 928 8080 60 7299-6780 Table 1. Datings for burial XIII Table 2. Datings for burials IV and VI Figure 1. Comparative chart of calibrated dates. millennium BC. With the 1000 year period from the middle of the 8th to the middle of the 9th millennium BC designated to layer 3 or Level III, the basal m-r or terra rossa layer can be attributed to the previous period, which ended around 8500 BC. The following graph shows the calibrated datings (Fig. 1) We shall now see if this chronology is feasible on the basis of the lithic material. 3. Lithic material Conscientious washing and sieving allowed us retrieve almost all the archaeological material contained in the soil matrix at the site, consisting of 339
MEDITERRANEAN BASINS. CENTRE.  The subsequent analyses were also conducted in the Radiocarbon Dating Laboratory, Faculty o...
340 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD tools, artefacts and lithic remains, just one bone tool, mastological and mollusc fauna, haematites or ochre and ceramics. The lithic material was essentially flint and stone (see detailed inventory in Table 3). Accordingly, focusing on the flint material, 11,887 items were found (see table), which may seem large but in our opinion is not so, bearing in mind the volume of sieved earth and that 11,204 are debitage flakes and flakelets, leaving 683 tools as such, including 125 cores and 278 blades and bladelets. Tools thus only comprise 4.98% of the flint material, an obviously low figure which indicates the low specific weight of stone tools in the daily activities at this site, with the exception of a certain type. Figure 2. Stratigraphic section This is undoubtedly a microlithic assemblage, despite these being substrate items, sidescrapers, denticulates, retouched flakes, or those of Palaeolithic origin– endscrapers or burins, with the occasional larger format item amongst them. This feature is characteristic of the postulated period. The stone tools and their grouping, together with the soil features, the anthropological remains with their respective radiocarbon datings and the faunal remains allow us to structure the industrial, environmental and cultural sequence of the site in the following manner. Level IV and layers 4 and 5 correspond to the initial settlement which left traces, m-r and red
340  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  tool...
3 1 4 2 3 4 2 ENDSCRAPERS MICRO SCRAPERS BURINS MICRO BURINS STRANGLED BLADES /BLADELETS SIDECRAPERS 23 1 STONE 56 52 70 70 1.271 2 1.195 40 1 8 1 1 1 1 3 5 6 7 C2 19 19 602 1 567 11 1 3 2 1 5 6 5 C3 SEASON I 5 5 18 15 1 2 C4 3 3 10 9 1 C5 42 42 1.816 1.702 50 16 8 4 3 5 4 6 3 15 C1 57 23 34 537 529 2 1 1 2 2 C2 12 12 288 2 274 7 1 1 3 C3 Iª AI-B-T- BII-BIII 4 4 210 205 1 2 1 1 C4 1 1 204 187 6 1 2 2 1 2 1 1 1 IIª N-I Table 3. Overview of retrieved material LEGEND: SUP = Superficial T = Embankment E = Burial B- E= Deep burial N = Level C = Layer B/C = Clay/Pottery V= Various 23 13 1 B/C TOTALS 1 38 646 616 14 1 2 1 2 1 4 5 C1 OCHRE 56 3 1.060 1.018 12 980 697 FLAKES 42 TOTALS 980 576 BLADES/BLADELETS VARIOUS 17 36 1 5 POINT WITH NOTCHES AWLS/DRILLS 2 2 GEOMETRICS 10 1 6 3 1 BACKS & BACKED EDGES 7 1 4 2 6 Sup 4 1 2 52 All Iª. Sup. B12-T CORES FLINT Sup. 6 6 736 3 701 14 1 3 2 4 2 2 4 C1 8 8 382 366 2 1 2 1 1 1 3 5 C2 4 4 328 316 4 2 2 4 C3 19 7 12 421 410 4 2 1 1 1 2 C4 V. V. 48 48 C5 1 1 89 83 3 1 2 EIX SEASON II.. N-II 2 2 565 549 8 1 1 2 4 EXII 2 2 169 167 1 1 EXIII 2 2 217 206 4 1 2 2 2 XIII B-E IIª N-III 224 1 213 2 1 2 1 4 IIª N-III 3 1 2 331 310 11 1 1 3 1 1 1 2 IIª N-IV 392 36 9 347 11.887 24 11.204 278 12 58 30 14 6 7 21 39 39 30 125 TOTALS MEDITERRANEAN BASINS. CENTRE. 341
3  1  4  2  3  4  2  ENDSCRAPERS  MICRO SCRAPERS  BURINS  MICRO BURINS  STRANGLED BLADES  BLADELETS  SIDECRAPERS  23  1  S...
342 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD basal earth, as shown in the stratigraphic section (Fig.2). This period corresponds to Mesolithic I-B (10,000-8,500 BC), chronologically after Mesolithic I-A which in turn followed the Late Magdalenian. One endscraper, 2 burins, 1 microscraper, 3 blades and backed blades, 1 geometric, 1 point with notches, 1 sidescraper, 3 borers/drills and 12 bladelets confirm the above. Level III and Layer 3 correspond chronologically to Dyras III, (8,500 – 7,500 BC), while technological and culturally they corrspond to the Mesolithic II or “Sauveterr” in our structure. This level has yielded 1 endscraper, 6 micro-endscrapers, 2 microburins, 8 burins, 2 backs and backed blades, 3 points with notches, 3 sidescrapers, 2 borers/drills and 20 blades and bladelets. The Mesolithic III-A took place between 7500 and 6500 BC, during the Pre-boreal – with a slight increase in cold spells within a general trend of rising temperatures, lower rainfall and more frequent widespread droughts. During this period, snail collection became widespread in the hinterland areas, with snail and shell middens in coastal areas. This period corresponds to Level II and Layer 2, which yielded 13 endscrapers, 9 microscrapers, 8 burins, 1 microburin, 2 blades/ bladelets with opposite notches, 2 geometric blades, 2 points with notches, 10 sidescrapers, 3 borers/drills, 44 blades and bladelets and over 2,000 flakes and flakelets. The first dated burial (individual IV) in this period took place around 6,740 BC. Level I and Layer 1 correspond to the Boreal period, 6.500/6.000 BC and the start of the Atlantic, from 6000 to 5500 BC in this area. The former was accompanied by a considerable rise in the sea level as temperatures rose and the icecaps melted, which triggered a new marine encroachment, the Versilian in the Mediterranean, accompanied by the submergence of coastal areas and changes in the coastline and the landscape, while local rainfall remained low. The coastal wetlands were reactivated and invaded much of the Gulf of Valencia. Individual 6 was buried around 6,130 BC. Level I and Layer 1 yielded 8 endscrapers, 8 microscrapers, 8 burins, 6 microburins 1 blade/ backed blade, 3 flakes/strangled flakelets, 10 points with notches, 20 sidescrapers, 2 borers/ drills and 70 blades and bladelets. The superfi- cial Atlantic layer yielded 2 endscrapers, 4 microscrapers, 1 burin, 7 microburins, 4 geometric, 1 point with notches, 10 sidescrapers 42 blades and bladelets. Individual XIII lived during the Atlantic. 4. Human remains We have reached the following conclusions from a study by Dr. Campillo Valero (Aparicio, 2008) and his team: 15 individuals were located in the excavated material. It is unlikely that all of them were buried, as we do not know if the zone marked as the perimeter of the site includes the entire original site area or whether part of it was either removed by erosion or the above-mentioned farmwork, and also because ultimately it became impossible to fully excavate the subsoil. However, it is reasonable to assume that not all those who lived at this site were buried here, even if we accept that the earliest date corresponds to the first burial, i.e. from the 2nd century of the 7th millennium BC. Considering that during the 1,300 year period, assuming that the occupation was seasonal and that the group or community was small, at least 200 or 250 people may well have died here in this period, which begs the question what happened to the rest? This issue is impossible to resolve on the basis of the data at hand. Perhaps in the future, if a full excavation of the subsoil can be completed, the permanent or seasonal nature of the site will be ascertained and, if others are found nearby, we might be able to find an answer. From the first layers of level II, either above or within level III, and on level II, the corpses were placed in shallow graves, with dimensions that matched the volume of the remains. In the case of 1, the volume was quite small as it was a bundle of bones; even smaller in the case of X where only a skull and a few bones were buried. In conjunction with the discovery of other remains scattered across the site, this suggests that some of the buried remains were disintegrated when the graves were opened or stumbled on, or the floor was dug up for some purpose. Their shallow nature was presumably due to crushing by the inhabitants and their continuous heavy pressure on the floor. A large rock with a sloping lateral surface was used in the burial of Individual 11, discovered
342  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  basa...
MEDITERRANEAN BASINS. CENTRE. when the pit was opened. The corpse had been placed leaning against this rock and was thus in a near-vertical position. The lack of several bones in the case of skeletons still in anatomical connection is not easy to explanation, with the exception of number XIII, which lay beneath an olive tree which we had to remove to excavate the area beneath. When digging the hole for this tree, the bones of a lower limb were scattered across the terrace and then disappeared as a result of crushing or weathering. Some of the skeletons were deliberately protected by stones on or around them, as in the case of II, III and IV, VI, XIII and large rocks in the case of XIV. Almost all were laid on their left side except for VI and V and possibly III, which were lying on the right. In all cases, the legs were folded, the hands were placed together with the arms placed on the upper or lower chest. Perhaps the most noteworthy aspects of the anthropological analysis are the following: The disinterred individuals or graves revealed: one 9 month old perinatal; 2 sub-adult juveniles aged 15-18; 1 young adult aged 18-20; 6 adults, 3 of them aged 20-25, 1 aged 30-35, and 2 aged 35-40; and 2 mature individuals over 45 years of age. The average age of death is thus 29 years, with 58.33% to 66.66% in the 20 to 40 year range. We may thus conclude that neither infancy nor puberty were the most critical period, in this case situated between 15 and 25 years– the average age amongst the prehistoric populations of the time. The exact causes of death could not be determined, however we did detect few cases of caries and frequent – but resolved– injuries. We were able to determine the sex of 6 males and 3 females, with another 2 doubtful. The average male height was 1636 cm and 1554 cm for women, quite tall in comparison with contemporary and later data, particularly in the case of the females. The anthropological features include a mesocephalic component with brachycephaly, quite rare in Spain and absent in compared series which are later in most cases, with a doubtful evolutionary process. Differences also occur in the skull, which is more robust in this case, accentuating the head. These individuals have low eye sockets and broad faces, suggesting a link to the Central European Mesolithic and subsequent remains from the Valencia region, an aspect which should be consid- ered in studies of a possible indigenous evolution. The mandible is narrower than might be expected, especially considering the large, robust teeth which in the anthropological study seem reminiscent of Neanderthals. This would be consistent with the sub-nasal prominence, interpreted as a persistent archaic feature which underscores the possible Neanderthal lineage and also opens up a suggestive line of research which we wish to begin immediately. The large number of Wormian bones points to probable consanguinity, and endogamy as a useful social aspect. Dental features and other elements found at the sites lead to the conclusion that plants formed a minor part of the diet while there was a high consumption of animal protein, either marine or terrestrial. Analysis of middens Cuerda and Gasull has led to the suggestion that shells were consumed as well. On the other hand, after the initial anthropological studies, Dr. Subirà –an efficient and constant collaborator with a highly skilled team of collaborators– studied the diet on the basis of stable isotopes (Subirà et al., 2003, García et al., 2006). These studies concluded that protein intake of animal origin, both marine and terrestrial, only formed 25% of the diet, with little difference between sexes –both were omnivores– although there were some differences in minor dietary preferences amongst individuals (Chimeno et al., 1992). However, this begs the question that, if animal protein intake was only 25%, there is no evidence of other products which increased it and plant intake was low, what was the food source which completed the rest of their diet? This is a very difficult question which remains in the air. A list of species which formed the terrestrial meat diet has been published elsewhere (VV.AA., 2008). This list reflects the range available in this area according to other studies. The aquatic species certainly included turtles, whose shells were found in abundance on all levels at the site. 5. Conclusion The following historic reconstruction is based on the aforementioned details: Around 9,000 BC, a small group of people, probably five or six, moved from a relatively close settlement to what was probably a shelter or recess at the foot of a limestone cliff on a hillside 343
MEDITERRANEAN BASINS. CENTRE.  when the pit was opened. The corpse had been placed leaning against this rock and was thus ...
344 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD near the modern day town of Oliva. Their aim was to exploit the surrounding territory which included marine resources –the coastline was quite nearby at this time– and land resources as well –in the surrounding plains and mountains there were herds of different quadrupeds, despite the deleterious effects of environmental changes caused by the Interglacial. The sea at the time was a source of fish, Cardium edule and some other species, while the coastal plain was not yet a swamp. Their lithic technology was inherited from the Magdalenian but limited to common types– scrapers, burins, backs and backed blades. The use of bone had been abandoned. These individuals may have arrived from one of the other sites nearby which were abandoned in this period, such as Cova Foradá or El Capurri, both in the Oliva municipality, or from the site in Camp de Sant Antoni, barely 500m away in a straight line. Between 8,500 and 7,500 BC, their stone tools underwent technological changes, gradually incorporating geometric items for hunting or fishing – arrows or spears in the former case, harpoons in the latter. After 7,500 BC, geometric items became widespread and something extraordinary happened: they begin to bury their dead, in the case of these small communities, perhaps amongst individuals who had close blood or kinship ties. The dead had to be protected– they were no longer regarded as inert, worthless refuse but the remains of loved ones. Society became humanised and anthropocentrism began– human beings –men and women– were now at the centre of their world and indeed the world of all, by now a widespread concept and sentiment. This conceptual progression was also linked to human representations in the local Levantine prehistoric art, where it first appeared, coinciding with the burial phenomenon. This small human group did not consume large amounts of seafood or animal products, nor did they need many stone tools for hunting or fishing, reflected in the items accumulated found throughout the occupations of the site. Around 6500 BC, the sea level began to rise slowly and flooded the entire coastal area up to the base of the slope at the foot of the cave, transforming the area into a great marsh or fen that contained abundant lagoon molluscs and turtles. The geometric tools became diversified and new types appeared. In the middle of the 6th millennium, higher –now intense– rainfall probably forced the occupants to abandon the site and search for a new cave or perhaps huts on the plain. We believe they moved to the foot of the slope alongside the marshy plain, as a Neolithic or perhaps Protoneolithic site containing a midden and hand-made pottery has been excavated recently in this area, now in the urban heart of Oliva.
344  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  near...
MEDITERRANEAN BASINS. CENTRE. Elisa María Domènech Faus*; Mª Mercè Bergadà Zapata**; José Antonio Riquelme Cantal***; José Luis Vera-Peláez; Mª Carmen Lozano-Francisco****; Consuelo Roca de Togores Muñoz*****; Rachel Wood****** The Upper Paleolithic From Cova Beneito (Muro, Alicante, Spain) 1. Introduction 2. The Excavation In The Rock Shelter Exterior Cova Beneito opens in a powerful package of gray dolomites around l’Altdels Volcadors (Serra Benicadell). Placed 650 meters above sea level and is oriented west-southwest. It is a large rock shelter, whose overhang has detached and the left wall unbolts in a cavity of 6x8 meters. Archaeological work in the new sector allowed for outlining the boundaries of the occupation area during the Upper Paleolithic, a fact which proves the existence of a settlement of hunters more significant than assumed in previous studies during the late twentieth century. As it progress- From 1980 to 1990, work on the cavity revealed a long sequence with different levels attributed to Mousterian, Middle and initial Upper Paleolithic (Iturbe et al., 1993). In 1993, we proceeded to review the interior north and east profiles in order to complete the sedimentological and palynological studies. The excavation of 1994 clearly displayed the differential contact between the C units, sterile from an archaeological point of view, and the first layers of the Mousterian horizon (Domenech, 2005). In 1999, a new stage began with the excavation of the south profile, in danger of collapse, and near the entrance of the cavity. Following the recent documentation review and materials provided at that time, three Paleolithic phases were differentiated: Gravettian, Solutrean and Solutrean-Gravettian. From that moment on, an open exterior sector is posed under the great shelter protection. In this new step, the main lines of research have focused on the verification of the actual settlement dimensions in order to complete the stratigraphic sequence, paleoenvironment, and paleoeconomics studies, for example, analysis of occupancy levels and obtaining reliable absolute dating. * ** *** **** ***** ****** Figure 1. Stratigraphic section from Cova Beneito Ajuntament de Muro d’Alcoi, Placeta Molina, 4, Muro d’Alcoi-03830 edomenech@vilademuro.net; Museu Palau Comtal, El Pla s/n, Cocentaina-0382 patrimoni@cocentaina.org SERP. Departament de Prehistòria, Història Antiga i Arqueologia. Universitat de Barcelona. C/ Montalegre 6-8, Barcelona- 08001 bergada@ub.edu Universidad de Córdoba,Departamento de Geografia y Ciencia de la Tierra, Plaza del Cardenal Salazar, 3, 14003- CÓRDOBA, jriquelme@uco.es Gaia Museum S.L., Avda. María Victoria Atencia s/n, 2901-MÁLAGA gaiadidactica@gmail.com MARQ Museo Arqueológico de Alicante Pza. Dr. Gómez Ulla s/n Alicante-03013 SPAIN crocat@diputacionalicante.es AResearchSchool of Earth Sciences, Australian National University, Canberra, 0200, Australia; ResearchLaboratory for Archaeology and the History of Art, University of Oxford, Oxford OX1 3QY, United Kingdom rachel.wood@anu.edu.au 345
MEDITERRANEAN BASINS. CENTRE.  Elisa Mar  a Dom  nech Faus   M   Merc   Bergad   Zapata    Jos   Antonio Riquelme Cantal  ...
346 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD es in breadth and depth, large blocks that define the occupation area appear, especially around the Solutrean and Solutrean-Gravettian. In addition, several areas of combustion have been located where flint has been treated, as well as preferred knapping zones, with a presence of large traces of preparation. The obtained sequence begins with a superficial level of medieval and Bronze Age materials. Followed by a minor sterile layer, then level II is developed. This one was tentatively attributed to Solutrean-Gravettian period, according to absolute dating and by analogy with other data obtained in the interior excavation. The level’s bellows are III and IV, which allowed for dating and outlining of some parts, is ascribed to the Solutrean, but was difficult to define given the scarcity of retouched, and especially the absence of, diagnostic spare parts. Upper Paleolithic sequence is ended in the V exterior level with clearly Gravettian industries. 2.1. Stratigraphic sequence, sedimentary evolution and Paleoenvironment In 2009, the sedimentary record studies restarted and focused on the exterior of the cave in the field called “Excavation I”, the north profile, with an output of more than two meters (Fig. 1). The aim of the current research focuses on the analysis of the filling formation, the human occupations interaction with the environment, and finally the postdepositional evolution of the site and the components that comprise it. Thus, it will be possible to frame Cova Beneito data in the Upper Paleolithic evolution in the IberianMediterranean area. The methodology we used involved the stratigraphic-sedimentary field description and application of micromorphology levels V to I; although the data presented here are still quite preliminary (Domenech et al., 2012). Figure 2 a. The stratigraphic sequence of Cova Beneito. Excavation Sector I. Northern Profile. b. Detail of the falling blocks from level III.c. Micromorphological sample from north profile.
346  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  es i...
MEDITERRANEAN BASINS. CENTRE. This record begins with level V, Gravettian of carbonated clayey, silty sand blocks and pebbles. Its formation process responds with a colluvial depositional period characterized by block falls, mainly located at the ceiling level; almost in contact with the overlaying level IV. This brought a change in the general shelter morphology. Subsequently, there was percolation of carbonated water and saturation of sedimentary fill. Below, the level IV is attributed to the Solutrean. It has a silty sand clays matrix, pebbles and blocks with erosional contact respect to the level III; corresponding with a colluvial process less marked than level V. This horizon has obtained an absolute dating of 18,275 ± 175 ky BP setting this stretch of sequence. To relate the paleoenvironment episodes documented with Upper Pleistocene chronostratigraphic scale used globally, dating has been calibrated by CalPal 2007_HULU program (Weninger et al., 2004) with the purpose of placing it between 22,580-21,380 ky calBP. Therefore, it was possible to ascribe it to the GS-2c period, categorized by cold, but not severe, weather conditions. Subsequently, an erosive process occurs. Following this hiatus, begins the sediment reactivation with Solutrean level III. This one has presence of pebble, blocks and silty clays, also caused by a colluvial process and block falls. The conditions of this deposit are more stringent and cold than the others from level IV. This second drop of blocks, produce another important change on the roof shelter morphology. The level II is related to the Solutrean-Gravettian cultural period. Furthermore, it has silty clays with sand, whose depositional conditions are different from those of previous levels. It corresponds to a process of diffuse gullies, with minor competence and continual slow movement of the laminar type. This level is dated in 16,180 ± 140 ky BP that correspond to the calibration range of 19,820 and 18,900 ky calBP, which would correlate with the moist and fresh episodes of GS-2b. Finally, an erosive process that ends with the silting Holocene colluvial fill type, level I, pebbles with silty clays sands with abundant remains of plant origin, in which a tank weathering occurs through today. 2.2. Absolute dating The absolute dates collected for Upper middle Paleolithic phases were obtained by Thermoluminescence, 14C and Accelerator mass spectrometry (AMS) (Table. 1). The first, were achieved in the radiochemical dating laboratory from the Universidad Autónoma de Madrid. Samples for AMS and 14C dating were sent to the laboratory of the University of Uppsala. (Domenech et al., 2012) There is corresponding dating of skeletal remains excavated in the Gravettian levels during the 1999 season. In this sample, no bone was found with marks as they eroded surfaces and concretions. Furthermore, from the 64 selected remains only 4 contain> 0.5% N, and of these, only three can be dated through collagen. To identify bone with sufficient collagen, the nitrogen content was measured by using an Elemental Analyzer adapted to Isotope-ratio mass spectrometry (IRMS), driven in continuous flow mode (Brock et al., 2010.2013). Collagen extraction was attempted only on those containing more than 0.5% nitrogen. This process was followed by the protocol of treatment with an acid wash to demineralize, exogenous removal of carbonates and humid acids, followed by gelatinization, filtration and ultrafiltration (Vivaspin 15,30 KDA molecular weight) (Brock et al., 2010). The lyophilized samples were burned in the EA to allow the measurement of the carbon and nitrogen stable isotope. The radiocarbon years were measured with the Accelerator Mass Spectrometer. The dates have been calibrated with IntCal 13 curve (Reimer et al., 2013) in Oxcal v4.2. (Bronk, 2009). From the data obtained, we must mention two circumstances. The first is related to samples with quality assurance indicators outside the ac- LEVEL II Ua-32243 16180±140 BP (bone) LEVEL III Sample A MAD-3916 TL 18025±1623 BP Sample B MAD-3917 TL18167±1631 BP LEVEL IV Ua-32244 18275±175 BP (charcoal sediment) Table 1. Absolute dates for the exterior excavation. 347
MEDITERRANEAN BASINS. CENTRE.  This record begins with level V, Gravettian of carbonated clayey, silty sand blocks and peb...
348 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Sample Context Yield (mg) Yield (%) %C d13C d15N C:N OxA- Date (BP) Error Calibrated date (cal BP, 95,4% probability) From To BEN-7 Lev. 42 6,59 0.7 42.1 -19.6 6.5 3.2 X- 2366-7 25150 200 29675 28730 BEN-16 Lev. 40 13,5 1,3 44,5 -18,4 5,4 3,5 21842 23730 180 28208 27522 BEN-29 Lev. 32 12,5 1,5 46,2 -18,6 4,5 3,3 21843 23180 160 27719 27183 BEN-8 Lev. 42 5,1 0,5 Table 2. The Gravettian dating results for the levels from 1999 campaign Figure 3. Compared graphics for the radiocarbon dates cepted archaeological collagen-OXA-X 23667 range, which contains 0.7% collagen (only 1% difference acceptable). The second is the high amount of C: N of 3.5 dating ref. OXA-21842 accepted, for instance the index is usually between 2.9 and 3.4. Nonetheless, these samples are only slightly outside the acceptable ranges. Additionally, the consistency of three dates dismisses both limitations (Table 2) (Fig. 2). 2.3. Typology and lithic technology The overall lithic remains in the levels attributed to Upper Paleolithic have posted a total of 7,609; of those 226 have been used as a blank for the retouched pieces, representing 2.9% of the total. For the technological study (Table 3), 4,209 remains were selected. For the typological study, the retouched artifacts have been included in five major groups: scrapers, burins, flat retouched and abrupt retouched tools, microburin and fractured pieces. Also, another various group, composed by simple retouch objects, retouched fractures, denticulate and borers, among others (Table 4). The three groups best represented in the Solutrean and Solutrean-Gravettian, II and III / IV levels are scrapers, abrupt retouched artifacts and various; while in the Gravettian, level V, highlight the abrupt retouched stone tools (Fig. 3). Nonetheless, the two groups reflecting the qualitative differences are flat retouched and fractured pieces. The first, connects to levels II and III / IV; although neither the diagnostic characteristics pieces of these phases appear. While the second links to levels III / IV and V, with similar percentages and parts with resembling morphological type characteristics. The artifacts with abrupt retouch are significant in the Gravettian, among which are the gravettes and bladelet with dejected edge, including double truncation bladelet linked with the presence of microburin. The various group soffers greater variability in the Solutrean levels with significant retouched tools and borers. The archaeological and technological analysis of the three units, in progress, documents the existence of various knapping processes in the production of two types of utensils, blades and flakes.
348  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Samp...
MEDITERRANEAN BASINS. CENTRE. Level II Levels III-IV Level V Scrapers 27,66% (13) 19,32% (23) 10% (6) Burins 8,51% (4) 8,40% (10) 12,66% (7) Flat retouch 4,25% (2) 3,36% (4) - Microburin - 1,68% (2) - Abrupt retouch 14,89% (7) 24,37% (29) 45% (27) Chipped pieces 6,38% (3) 13,44% (16) 13,33% (8) Divers 38,29% (18) 29,41 (35) 20% (12) Table 3. Totals of lithic evidence analyzed Level II Levels III-IV Level V TOTAL SAMPLE Total sample 1403 4629 1577 7609 Technologic analysis 837 1759 620 3216 Retouuuch pieces 47 (3,3%) 119 (2,57%) 60 (3,80%) 226 (2,9%) Table 4. Typological features of Cova Beneito. In blade manufacturing, several methods were used to obtain elongated tools with measurements and diverse morphologies. Among them are identified knapping methods by applying standard methods of volumetric operation; as well as bipolar knapping using the anvil technique, similar to the manufacturing of chipped pieces. These productions were virtually unchanged throughout the sequence, with a peculiar proliferation of bipolar knapping in Solutrean-Gravettian and Gravettian levels. Major changes are observed in flakes manufacturing between different chrono-industrial phases. The Levallois method is clearly defined in Solutrean-Gravettian and Solutrean periods used for making armor with flat retouch. In addition, the presence of exploited cores without prior shaping is observed. The raw material used was flint and limestone. 2.4. Zooarchaeology The material used has provided a total of 8,537 bone fragments, pooled in two levels, Solutrean and Solutrean-Gravettian. Of those, 2,132 (24.97%) have been determined, while the remaining 6,405 (75.03%) residues have not been attributed to any species due to its high degree of fragmentation. Figure 4. Retouched materials. Level V. 349
MEDITERRANEAN BASINS. CENTRE.  Level II  Levels III-IV  Level V  Scrapers  27,66   13   19,32   23   10   6   Burins  8,51...
350 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Solutrean level This level has provided a total of 1,234 bone fragments, only 225 (18.23%) have been identified both anatomically and zoologically, forming the number of identified specimens (NISP). The identified mammal species are: Equus sp. (horse) A carpal (semilunar) belonging to an adult has been identified. Cervus elaphus (deer) Similar to the horse, the deer is also represented by a single bone fragment, in this case a 3rd phalanx. Capra pyrenaica (mountain goat) The mountain goat has provided a total of 8 certain bone fragments (3.56%), representing a minimum of 2 individuals (12.5%). Age cohorts are represented, both juvenile and adult. The anatomical detail is marked by the presence of a greater number of head skeletal portions, followed by the appendicular. fragments (1.94%) representing a minimum of 2 individuals (1.79%) adults. Skeletal fragments best represented are appendicular, followed by the cranial and axial. This is the third most important ungulate species in the site. In the1st phalanx a longitudinal fracture is observed, from which is possible to obtain the bone marrow. Although the identified material has been scarce, some measures have been obtained, particularly from teeth. The measurements acquired are within the variation of this species in the southeast peninsular (Martínez, 1996). Cervus elaphus (deer) This species is represented by a total of 48 specific fragments (2.52%) expressing a minimum of 4 individuals (3.57%).The only age cohort found is the adult. The anatomical analysis has shown the prevalence of appendicular skeletal parts, followed by the cranial and axial. The bone material studied is highly fragmented, responding to butcher process and bone marrow removal (Pérez, 1992). Finally, we note the presence of 1st and 2nd phalanx digested, the first partially and the other entirely, implying the presence of carnivores in the cavity. Capra pyrenaica (mountain goat) Oryctolagus cuniculus (rabbit) The rabbit is the best species represented in the analyzed set a total of 215 fragments determined (95.56%) corresponding to a minimum of 12 subjects (75%).The skeletal portions better represented are the appendicular, followed by the cranial and axial. The bones are fractured and, in most cases, burned so is difficult to propose an intrusion. Fractures on rabbit bones indicate modalities that can only be attributed to human intervention (Pérez, 1992). Solutrean-Gravettian Level The corresponding recovered bone material at this level rises to a total of 7,303 fragments. From this pooled, have been recognized anatomically and zoologically 1,907 (26.11%) of them, that conform the numbers of identified specimens (NISP).The documented mammalian species at this level are as follows: Equus sp. (horse) The horse has provided a total of 37 specific The mountain goat is the best represented ungulate on the site. A total of 78 bone fragments (4.09%) have been recorded; these represent a minimum number of 6 individuals (5.36%). Age cohorts represented are both juvenile and adult, with undistinguishable dominance of one over the other. In this case, the anatomic elements are marked by the presence of a greater number of appendicular portions, continued by cranial (which shows the number of isolated teeth) and axial. Frequently, there is presence of fire traces in the bones. Furthermore, in this case, bitten short bones (patella) and digested (1st phalanx) by carnivores have been found. The identification of these scarce carnivore marks would suggest, possibly a timely existence of these taxa in the cavity. This situation would be directly related to a temporary occupation by man. Sus scrofa (wild boar) Its presence in the bone sample is particularly uncommon. Currently, we only have identified a fragment belonging to the 1st phalanx, possibly next to an adult individual at this age.
350  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Solu...
MEDITERRANEAN BASINS. CENTRE. Oryctolagus cuniculus (rabbit) It is the best represented species in the bone assemblage with a total of 1,740 determined fragments (91.24%) corresponding to a minimum number of 97 individuals (86.61%). The most abundant skeletal portions are the appendicular, after the cranial and axial one. The bones are fractured, and in most cases, burned, so an intrusion is not recorded. In addition to the presence of several cut marked bone, we have found rare evidence indicating how small carnivores or raptors, transported some rabbits into the cavity, documented by the finding of a humerus with small carnivore bites. Nonetheless, there is an overwhelming amount of fragments showing clear human activity. The fractures found on rabbit bones from Cova Beneito, except a few of them as we mentioned above, indicate modalities that can only be attributed to human intervention Lepus granatensis (hare) Like the wild boar, the hare bone sample is especially sparse. Only a fragment of femur belonging to an adult individual has been identified. Pardina Lynx (lynx) Finally, this carnivore is represented by a total of 2 bone fragments (0.11%) belonging to the same adult (0.89%). The determined bone portions belong to appendicular skeleton. The scarcity of bones does not allow for many insights. Nevertheless, the fractured humerus recovered could point to the hunting and consumption by man. It is the only certain carnivore bone sample. The bone sample recovered in Cova Beneito belonging to the Solutrean and Solutrean-Gravettian levels indicate the unclear differences between periods. This statement is based on the presence of different taxa species, as well as the hunting and consumption activity by human groups occupying the cavity in both periods. The continued presence of man in some sites could be a rising point in the same occurrence of carnivores, and therefore a decrease in its role as accumulators of bones. This fact, together with specialization in hunting rabbits, deer and mountain goat, will continue this pattern throughout the Upper Paleolithic period in the peninsular south. The abundance of rabbits, in the levels analyzed, should not overestimate their intake as food, as the hunting of one deer or a mountain goat would be the equivalent biomass to a high number of rabbits. Subsequently, the archaeological record indicates hunting animals with different biomass levels change their treatment in terms of cutting and consumption. For ungulates, anatomical parts would be consumed near the place of capture, while others would berelocated to settlements, because midsize and large prey cannot be totally consumed immediately. On the other hand, rabbit stalking, permits use of a different model for its availability throughout the year as adjusted biomass meat to daily needs of a person. 2.5. Malacological Study Paleontological study was performed on fossils and organic remains of mollusks obtained in the 1999 Upper Paleolithic excavations deposited in the MARQ and 2001-2005 campaigns. In 1999, 16 species of mollusks have been identified, six belong to terrestrial lungfish found in their natural habitat; two fluvial (Melanopsis lorcana and Theodoxus baeticus), eight marinas, six bivalves (Glycymeris insubrica, Pecten maximus, Pecten jacobaeus, Chlamysvaria, Cerastoderma edule and Acanthocardia sp.); a gastropod (Nucella lapillus) and a scaphopoda (Antalisina equicostatum) between the chrono-cultural sections from Solutrean to Gravettian. In all cases, it can be said that the mollusks have been used for a symbolic purpose. In the case of fluvial gastropod Melanopsis lorcana shell, has an anthropogenic hole to wear as a necklace, or other ornament. Similarly, with the shell Insubrica Glycymeris, the umbo is perforated, possibly for use it as a pendant. In the 2001-2005 campaign, 17 species of the terrestrial pulmonary have been identified and as in the previous case were found in their natural context; along with three marine species, one gastropod (Littorina obtusata) and two bivalves (Pecten maximus and Pecten sp.) Knowing the specific features of the cave, out of the reach of the tides during the studied periods, it can be said that all the fauna found there were collected at the sea or beach level by man and transported to the cavern; except Melanopsis lorcana and Theodoxus baeticus which came from a nearby stream, also collected by man. Regarding the paleobiogeography of the species under study, it can be said that all mollusks found currently belong to the Mediterranean, with the exception of Nucella lapillus and Pecten maximus. The first one, during the Gravettian, exclusively Atlan- 351
MEDITERRANEAN BASINS. CENTRE.  Oryctolagus cuniculus  rabbit  It is the best represented species in the bone assemblage wi...
352 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD tic, comes from Portugal. The second one, located in the Solutrean levels, can be found nowadays in the Atlantic and Alboran Sea, but not in Alicante. The presence of cold-water species in these latitudes is consistent with the glacial maximum in the case of Solutrean and Gravettian, and is included in this Würm glacial cold period. This is the logical explanation for the existence of these species in Cova Beneito through these chronocultural segments from the end of Upper Pleistocene. The free swimming (FS) mollusks (Pilgrim and scallops) live on mobile funds, along with the sandy-muddy paleoenvironment (SMU) (gastropod mollusks, bivalves and scaphopods from benthic infralittoral floor). Their presence demonstrates how human population living in Cova Beneito throughout the beginning and middle of the Upper Paleolithic had a common coast line with sand, mud or clay, which indicates mobile backgrounds, with an almost entire absence of rocky coast species. The exception is the Nucella lapillus, Gravettian single copy, attesting the presence of an obstacle, or rocky reef, on the Alicante coast during this chrono-cultural segment. It should be mentioned, regarding the malacofauna found at the site during the Upper Paleolithic, that almost all mollusks found were used for symbolic purposes. Most of them have holes in the umbo or lugs (for the case of shells from bivalves and scallops, respectively) to be used as beads and earrings;gastropods, such as Lorca Melanopsis, Nassarius gibbosulus, Nassarius corniculus and Theodoxus baeticus, with roughly spherical holes near the labrum, resemble a string of beads. Figure 5. Tooth located at Level III In other cases, the gastropods appear polished all around, outer labrum and columella, such as the shell of Nucella lapillus. The presence of Scaphopods is dominated by Antalisinae quicostatum species, drill holes, or anthropogenic punctures, were discovered in specimens. Nevertheless, these tusk shaped shells were used as necklace or jewelries beads, since they show two holes one oral and another aboral. These shells, which were used as symbolic ornaments, can be picked up on the beach, and generally have a bright white color. They are present in all chrono-cultural segments studied in Cova Beneito. Where do the mollusks of these excavations come from? They emanate from the sandy-muddy infralittoral (SMU) paleoenvironment and are free swimming (FS) bivalves. The most logical explanation would be that the shell collectors had gathered on the beach, tide level, with valves already disjointed, but given the fragmentation of the shells of pectinidae, a priori hypothesis cannot dismissed. Regarding Littorina obtusata, it comes from the same tide level, the supralittoral and mesolittoral levels, so their gathering could be straight from the rocks on the Alicante coast during the Solutrean. On the other hand, they could also have been collected directly from the beach, likewise in the tidal zone. 2.6. Study of a tooth In the case of dental research, here is outlined the study of one tooth found in level III, from the Solutrean period on Cova Beneito (Fig. 4).
352  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  tic,...
MEDITERRANEAN BASINS. CENTRE. Radiological techniques, specifically Scanning Electron Microscopy, were used to examine the formation of roots, and the pulp chamber, for information about diet. The tooth sample corresponds to a left upper second molar (27) of the permanent teeth, which belonged to a 12-year-old preadolescent of indeter- J. Emili Aura Tortosa * Coves de Santa Maira (Castell de Castells, Alacant) Les Coves de Santa Maria are located at the suntrap of serra d’Alfaro (1166 m), at the head of the Girona river. The cavity is configured around a breach inclined to the NW opening in the rocky wall of the Eocene reefs sandstones in the external Prebetic at an altitude of 600 m above sea level, 14 m above the current riverbed, and 25 km from the coast. Its denomination includes two sectors: the Boca Oeste (SM-W) and the lower entrance, called Corral del Gordo (CG), as that was its role until recently. 1. Lithoestratigraphy and Chronology The correlation between the lithoestratigraphic sequences of the mentioned sectors is included in Fig.1 The description from the base to the top of the sedimentary sequence is the following: In CG, the known sequence starts with Unit GCII, corresponding to a sandy-silty section with little presence of human activity, in accordance with the micro morphology analysis (Verdasco, 2002). According to the encompassed materials and radiocarbon dating, it was formed between the Last Maximum Glacial and the Younger Dryas, dating between 24 -13 ky cal BP. The upper unit CGI, is placed discordantly in respect to the CGII. It is of Holocene origin and corresponds to a Neolithic “fumier” against the wall; radiocarbon dating places it in the Atlantic Period (6.7 - 6.1 ky calBP). * minate sex. It shows slight wear on the occlusion surface. Through microscopic traces of wear analysis (pits and striations), high density of enamel microstriations with a flat trend have been observed, related to a diet mostly based on very abrasive vegetable consumption without an elaborate processing, such as in populations of similar chronology. Five sedimentary units have been identified at the west entrance. A large share of the upper deposits were formed from fallen materials due to gravity, as water action is weak or absent according to the sedimentological analysis made by Jesús F. Jordá. Unit SM5 was deposited at the end of OIS2, during GI-1 (13.5 – 13 ky calBP), and it is formed by centimetric thick laminations of mixed colors which rest on a speleothem. Unit SM4 is formed by shales with sands, little gravel and small limestone pebbles of less than 3cm. Unit 4 basis could be related to the end of Younger Dryas, while the upper sequence corresponds to the Pre-Boreal, dating between 11-3 - 10.2 ky calBP. Unit SM3 presents a remarkable dip towards the inner part of the cavity. Five sub-units have been identified with a similar grain size and a variable coloring. Pebbles and big limestone blocks appear on the basis, separate from unit 4. Radiocarbon dating places it at the Boreal (10,2 - 9,5 ky calBP). Unit SM2 has an irregular geometry with remains of organic laminations; it shows erosional contact over the lower unit. Unit SM1 presents a chaotic appearance, tabular geometry and it is strongly erosive over the underlying unit. The three upper units present important processes of bioturbation linked to rodent activity and their predators. (jeaura@uv.es). Dept. de Preshistòria i Arqueologia. Universitat de València. 353
MEDITERRANEAN BASINS. CENTRE.  Radiological techniques, speci   cally Scanning Electron Microscopy, were used to examine t...
354 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. Coves de Santa Maira. In the upper part, planimetry of the archaeological site. At the bottom, chronoestratigraphy, lithoestratigraphy and archaeological stratigraphy of the sedimentary sequences of Boca Oeste (SM-W) and Corral del Gordo (CG) sectors.
354  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
MEDITERRANEAN BASINS. CENTRE. 2. Paleoenvironment Anthracological remains studied by Y. Carrión (2005) show that in the cavity´s environment open vegetation existed with a predominance of Pinus nigra/sylvestris during a large part of MIS2. During GI-1, Juniperus sp. reached an important presence, causing an increase of Quercus sp. (deciduous and evergreen) during the Pre-Boreal and highlighted during the Boreal (Fig. 1). Riparian species (cf. Salix– Populus) and thermophile species (cf. Olea europaea) also show an increase from the Holocene. Changes of marine species brought to the site could have a paleogeographic reading. Thus, until the Pre-Boreal, bivalves species correspond to a sandy substrate –mainly Pectinidae sp. and Cardídae sp.– while from Boreal, they are made of rocky substract (Mytilus sp.). On the other hand, icthyofauna shows an increase in species linked to estuaries (sea breams and mullets) during the Boreal. These variations reflect the eustatic changes of the coastal morphology. The identified herbivores are common in other mid-mountain places and in this chronology; however, carnivores allow a certain serial numbering. he skull of de Cuon alpinus, whose direct dating proves its permanence during MIS 2 and apendicular remains of Panthera sp., come from the basal third of CGII. Lynx pardina has been identified in both sectors, and Felis silvestris and Vulpes vulpes are more common as the Holocene advances. 3. Techno-economical features In Fig. 2 a selection of variables is included, attempting to summarize changes of the use of the cavity along the MIS 2 and the Early Holocene. Occupations of the CGII, dated between 24-23kycalBP, contain Solutrean diagnostic elements –heat treatment and bifacially retouched leaf points– and high blade productions. A second occupation episode (CGII 5-2), dated between 17-13 ky calBP, shows a clear increase of micro-blade productions. In both cases, consumption and abandonment phases are registered, contrasting the reduced number of lithic blanks and by-products to pendant ornaments. The bone points collection is small: only double and mono-bevelled based points. The top dating of CGII is linked to the ones of the bottom of the known sequence for SM-W, although both sectors appear quite different. In the lithic productions of SM5 and SM4, there is a higher diversity of raw materials including the use of limestone. An integrated production is now recognized whose sup- Figure 2. Coves de Santa Maira.1: Solutrean broken foliaceus; 2-4: personal ornaments; 5: oxide platelets with repeated incisions; 6: awl; 7-8: personal ornaments; 9: pebble with percussion and polished marks; 10-11: retouched micro-flakes, usedas projectile; 12: human skull fragment with anthropogenic marks; 13: pebble used for ocher enhancement;(n 1 - 4 Solutrean CG-II; n 5-6 Epipalaeolithic, SM-4; n 7-13: Denticulate-notches Mesolithic, SM-3). ports will be used for two large tools groups: scrapers –truncated pieces and microliths. The by-products remains are numerous, being documented throughout the technical process. The identification of domestic industry on bone –needles and awls– contrasts with a limited use of the antler. Another remarkable feature of these occupations is the recorded ocher working: from the collecting of raw material to the processing pebbles (Fig. 2: 5 and 13). From the Boreal, occupations located in SM3 become more episodic with moments of little ‘carnivore´ contribution (Morales, 2013). This change coincides with a lithic industries break, blade and 355
MEDITERRANEAN BASINS. CENTRE.  2. Paleoenvironment Anthracological remains studied by Y. Carri  n  2005  show that in the ...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 356 bladelet productions being drastically reduced for the benefit of flakes, frequently over limestone (Miret, 2007). The input of flat edges from the river and the use of slappers and pebbles suggest new working processes, including ocher processing. Except for some retouched micro-splinters, projectiles have not been identified and bone industry is limited to punches (Aura et al., 2006). From an economic point of view, Capra pirenayca and Cervus elaphus are the basic species –in this order–. In the MIS 2 occupations, few remains of Equus sp. are identified and as the Holocene advances, forest-environment species are incorporated (Rupicapra rupicapra, Capreolus capreolus and Sus scrofa). SM-W has also provided information about the use of vegetables and the transport of marine mollusks and fish to inland sites (Aura et al., 2005). These resources, together with the lagomorphs, have established tendencies and make up the sense of for- José Aparicio Pérez* 4. Human remains In the thick unit SM3 “loose human remains” have been identified with anthropogenic processing and depleting marks of at least three individuals according to the study of Mª P. de Miguel (Fig. 2: 12) (Aura et al., 2009).On the other hand, the only evidences of the Early Holocene identified in CG correspond to a set of human and charcoal remains concentrated in an area of less than 1m2. Its dating places them at the end of the Boreal. This sudden appearance of human remains and burial rites in the Mesolithic start-up constitutes an emergent feature at a regional level, accompanied by direct evidences of relationships with the coast which is also confirmed by the isotope analysis of human remains (Salazar-García et al., 2014). Cova Foradà (Oliva. Valencia) 1. Introduction Cova Foradà is located in the west part of a low hill that, together with others, forms part of the set called Muntanyetes de Oliva, the last foothills of Serra de Mustalla over the coastal floodplain of the Gulf of Valencia, in the town of Oliva, Valencia. The setting where it lies is called Racó de Gisbert. This opened in the limestone rock of the karstic system of Serra de Mustalla, offering two openings: the western one is the entrance, and the eastern one was opened after the beginnings of the Holocene when the vault broke in the deeper part of the cavity. The name refers to this structural particularity, meaning holey cave. At the time of its discovery as an archaeological site in the early 1970s, the cave appeared to be * mulations such as intensification and diversification for the Iberian final Palaeolithic and Mesolithic. a small cavity, rather like a shelter, about 6m deep and 7m wide. The reason was an inner weathering of the limestone rock which produced the subsidence of the back vault that we pointed out, and also the subsidence of the front, or entrance, vault in a length of about 25 m and a width of around 20 m. The back blocks remained unchanged due to handling and extraction difficulties, while the blocks of the western vault had been rolled to the bottom of the ravine in order to take advantage of the rock for lime processing in a furnace, installed for this purpose, which still maintains its infrastructure in situ. The importance of Cova Foradà de Oliva as an archaeological site was proved after the findings during 39 years of research and studies. First of all, a powerful and thick stratigraphic sequence must be pointed out (Fig. 1) which, apart from minor remains on the surface, offers a Arqueólogo Director de la Sección de Arqueologia y Prehistoria de la Real Academia de Cultura Valenciana. Académico C. de la Real de la Historia. P.O. BOX. 2260-46080 VALENCIA joapa2005@hotmail.com
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  356  blad...
MEDITERRANEAN BASINS. CENTRE. Figure 1. Section with the stratigraphy. continuity to life that start in the first levels of the Mesolithic, take roots in the Paleolithic, and whose density and capacity could lead to the extent of the Middle Paleolithic despite the lack of clear dating. Even so, there is a possibility of finding former occupations of the cavity by going deeper, without reaching the bedrock or the sterile riverbed. Between the Mesolithic and the Mousterian we find, without a solution of continuity, the whole sequence known as Upper Paleolithic, Magdalenian, Solutrean and Gravittean-Aurignacian, not having the thickness of Parpalló but with a large quantity of discoveries, taking into account the small surface in which the activity took place. From early Leptonian, Mousterian starts with many lithic and faunal discoveries, just like in former levels, differentiating from Parpalló by the presence of the Middle Paleolithic which is nonexistent in the cavity of Gandía. The difference with Cova Negra de Xàtiva is also significant because of the lack of Upper Paleolithic levels. The singularity of the Middle Paleolithic levels is determined by the anthropological discoveries corresponding, at least for the moment, to three individuals among whom one, the CF10, matches with a Neanderthal specimen with a whole skull, a large part or the complete rib cage, vertebrae and ribs, maybe many of the upper limbs and some yet to be identified from the lower limbs. At this point, we have decided to intensify the investigation of the whole archaeological site, an essential requirement, adding to the anthropological studies already done and those which are being done currently, such as faunal, edaphic, lithic, palynological, environmental, economical, etc., which are more and more necessary in order toreach a holistic vision about the site´s activity. 2. Ecosystem The most remarkable feature of the territory’s ecosystem is the bio-diversity, as it participates, at present, in four environments: the maritime, due to the proximity to the coast, although we can´t forget the fluctuations along the Pleistocene and the Holocene; the mountain, with the whole Serra de Mustalla that we pointed out; the inner aquatic, due to the presence of the river Bullens whose drains run at the foot of the hill where it is located; and the mentioned Serra, located between the municipalities of Oliva and Pego, which also feeds the Las Aguas gap, marsh, or lagoon depending on the time of the year. The ecosystem is clearly reflected in the remains of the consumed fauna which have been recorded in the archaeological activities in the cave which we will mention later. The richness of this ecosystem explains the permanence and continuity of this cavity’s habi- 357
MEDITERRANEAN BASINS. CENTRE.  Figure 1. Section with the stratigraphy.  continuity to life that start in the    rst level...
358 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD tat, frequented during the Middle Ages, the Iberian Age, the Bronze Age, and consistently from the Mesolithic to the Middle Paleolithic, despite not knowing the moment of the first occupation due to the impossibility of reaching the bottom of the cavity. 3. Archaeological Activities The explorations made in 1975 brought together a batch of lithic material found by an amateur archaeology group which delivered it to us for examination. We observed that, for the most part, the material belonged to the Middle Paleolithic or Mousterian. In 1977 we made the first dig campaign, digs which are still being performed. In 2013 we carried out the XXX campaign. However, the first years were discouraging. The first campaign confirmed the existence of Mousterian levels, powerful and rich, both in industry and fauna, but a certain number of signs made us suspect of the contamination on the mentioned levels in the western part –charts E and D, 8 and 9 respectively of those first activities–, confirming the first dating with C14, done at the CSIC, showing the existence of Magdalenian levels. Digging in the range of charts K, J, I, H, G, F, E and D, 7, 8, 9, also in the west part, as well as the removal of levels, very deep into this part when digging charts I11, I12 and I13, fully confirmed this fact. Such an intense removal could be ascribed to treasures hunts in the Middle Ages, an activity that awakened an interest in this whole area and which we thought corresponded to the site´s entire surface. Given these circumstances and as a last option, we considered that if the large existent blocks in the east part, irremovable for their dimensions, corresponded to the wrecked vault, they would seal and correctly preserve the sedimentation and its archaeological content in the deeper part of the cavity. This comes from the Bronze Age as, among the blocks, the existence of burials was detected, confirming the dig and the dating with C14 of the CSIC. The digging of charts A 12-13 and 14; a 12-13, 14, 15 and 16; b 12, 13, 14 and 16 and c 12, 13, 14 and 15 confirmed our hypothesis, and behind a thick humus layer of 50cm, among the blocks and under their layers with the burials´ remains, fruitful archaeological levels arose, both in thickness and in content, exceeding all our hopes. The digging´s continuity as we have pointed out, has allowed us, through the thirty campaigns done since 1988, to configure a powerful and rich stratigraphy (Fig. 1) from the Mesolithic to the Lower Mousterian, established as follows: Mesolithic I with the line coast within the economical limits and the subsistence area –Mesolithic I with the line coast pushed away and outside the limits –Magdalenian with plenty of significant bone industry –Solutrean with Parpallean cutout points –Medium Solutrean – proto-Solutrean –Aurignacian –Gravettian and Mousterian (Fig. 2). Figure 2. Lithic tools from the Mousterian level. Currently, the survey is in full Mousterian, with plenty of material and extraordinary fauna but without signs of the proximity of the cavity´s base, thus the necessity to follow up with a sur-
358  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  tat,...
MEDITERRANEAN BASINS. CENTRE. vey to reach the bottom area, the initial settlement, and in such a way finish the survey, starting later with digging in extension from this area, is necessary. The discovery of human remains in the Mousterian level increased the interest and the richness of the content, reinforcing the protection fence with a cover over the unprotected sedimentation by the vault. 4. Human Remains In the year 2000, during the XX digging campaign, we foundin layer 28 and chart C14/C15, a piece of human jawbone and a piece of skull, both studied by D. Campillo; M.E. Subirà; E. Chimenos; A. Pérez and S. Vila (see Cypsela, nº 14, pages 143-148, Barcelona, 2002), whose conclusions were the following: “The pieces definitely match two individuals, an adult and a child”. The neurocranium fragments are not-excessively swelled, and there is a small fragment corresponding to a very developed torus frontalis due to its morphology and thickness. The preserved fragment of the jawbone, although is damaged, affirming that both the alveolar process and the nostril are quite broad, as well as its vestibule. The teeth are bulky and show in the x-ray the existence of a moderate taurodontism. All the arguments expressed are consistent with a Neanderthal diagnosis, probably a female. In the digging campaign in 2010, it was decided to continue with the digging of charts C-14 and C-15 where in the year 2000 some remains of a jawbone and skull fragments of a Neanderthal were removed. This is why the digging team joined forces with the anthropologists M. Eulàlia Subirà and Jordi Ruiz, both belonging to the Unitatd’Antropologia Biològica of the Universitat Autònoma de Barcelona (UAB), and who in the last years have conducted the study of human remains of the cave in coordination with Gala Gómez Merino and Carlos Lorenzo from the Institut de Paleoecologia Humana I Evolució Social (IPHES), Tarragona. On 9 August, the remains of quite a complete skull of a Neanderthal were discovered. In the following days, the chances of discovery of the upper part of a Neanderthal skeleton that included the skull to the first sacral vertebrae rose. The discovery is important because the skeleton was very complete and the bone joints were in close anatomic connection, that is to say with connections among the bones similar to those while alive, with no displacements at all. In most Neanderthal discoveries in caves, the remains are limited, fragmentary, and scattered, with signs of having been moved and eaten by predators. The different digging campaigns in the cave have proved the presence of hyenas which alternated with the Neanderthals in the use of the cavity. The digging was really slow at every moment because the bones were immersed in calcareous formations. This is why it was decided to extract the whole cemented block, including the immersed remains, for later digging in a laboratory. The fossils’ condition is very delicate. They are quite fragile and are cemented into a block of a very carbonated sediment. This is why, before being dug in the laboratory, they were submitted to a computerized axial tomography (CAT) and other image treatments in order to know the preservation condition of the bone remains, facilitating in this way the block´s digging in the laboratory. To extract and clean the remains from the block, mechanical equipment under binocular loupe was used. The treatments started, and are still continuing, in the Restoration Laboratory of the IPHES in Tarragona headed by Gala Gómez Merino, which has the necessary facilities for dealing with this type of bone material,. Furthermore, corresponding samples for later studies were taken. Once the bones are unlocked, the anthropological study, headed by Dr. M. Subirá of the UAB in coordination with the IPHES members, will take place. Together with the study of human remains, complete paleontological, sedimentological, antracological and palynological studies will also be handled, and the existent dating number will be expended by the C14 or by using other means which the C14 cannot reach. Dr. Eudald Carbonell has promised his complete collaboration from the IPHES and his management. 5. C14 Dating We now offer the dating set obtained throughout these years from the different remains subjected to the corresponding analysis. 359
MEDITERRANEAN BASINS. CENTRE.  vey to reach the bottom area, the initial settlement, and in such a way    nish the survey,...
360 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Cova Forada (Oliva) C-575 9.645 ± 327 12.081-10.000 7.695 ± 327 10.101-8.050 C-277 12.500 ± 800 16.855-12.871 Layer 7. Son. I 10.550 ± 800 14.905-10.921 Cova Forada (Oliva) Cova Forada (Oliva) C-276 ó C-126 11.500 ± 1.000 16.127-10.787 Layer 4. Son. I 9.550 ± 100 14.177-8.837 CSIC-1492 6.196 ± 34 BP Charcoal 5.279-5.046 BC Layer 2, part E o back part. CSIC-1493 5.633 ± 31 BP Charcoal 4.533-4.363 BC Layer 1, superficial Part E o back part. 16.960 ± 100 BP UBAR-935 / CNA 089 Fauna bones 16.960 ± 100 BP 18.133 18.355-18.255 Marjal de Pego (essential on the eco- [14 Samples UBAR] system next to the cavity) From 1.660 ± 50 to 10.120 ± 460 = 13 samples UBAR-45 28.240 = 1 sample Table 1. C14 Dating. 6. General Conclussions As formerly explained, Cova Foradà de Oliva constitutes one of the most complete archaeological sites and therefore, one of the most important to the study of Prehistory on a national level. In our autonomous region, it is possible that the site might be one with the largest stratigraphy comparable to the total of the strata of Cova Negra (Xàtiva) and Cova del Parpalló, (Gandia). It is one of those places which seems to acquire a larger extent after each archaeological exploration campaign –taking place regularly since 1977–. This particular site, so named because of becoming a crossing cavern after its vault´s partial detachment, quintessentially embodies the prototype of a Prehistoric site. Cova Foradà is one of those places which will always provide news. Among its countless virtues, it has a stratigraphic richness that makes the cave exclusive, and this is something just a few sites can maintain. Although early Medium Paleolithic strata have been reached in the last campaigns, the fact that the basal rock has still not been achieved is an opportunity that may bring many surprises. Among the exhumed remains, prolific and abundant lithic tooling points out consistent morphologicality with the different periods represented by the stratigraphy. It also appears that a large quantity of fauna remains: some heavy mammals, but above all little rodents, among which the rabbit stands out due to its abundance. Coprolites, and other vestiges related to the periods of animal occupation combined with human groups, have also appeared. However, this cavity is famous for the discovery of Neanderthal remains, among which the half-fossilized body of an individual is a highlight, and up to now represents one of the most important discoveries added to other remains formerly found in our Peninsula. The vestiges of this specimen appeared in 2010, in a small niche emerging naturally from the east wall at the bottom part of the cave, about 7m deep from the reference point –0–. They were
360  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Cova...
MEDITERRANEAN BASINS. CENTRE. immediately set, the whole skull, part of the face, some vertebrae and part of the rib cage. All these bone vestiges were almost petrified by the action of the carbonate exchanging and the incipient mineralization inherent to such ancient remains. They were part of a concrete block in which mineral and bone parts were a whole. As it was the only way to recover the set without damaging it, the entire dug block had to be extracted by removing part of the cavity´s stone support. In order to be transported and studied in a place with the necessary means for the objective, a box-nest was prepared to make movement and handling easier. Thus, the petrified remains of the Neanderthal body were moved to the IPHES laboratory, in Tarragona, where the restoration team could start a micro-digging, which still continues today, to separate the human parts from the mineralized matrix. Formerly, in the year 2000, a fragment of an upper jawbone had appeared, being attributed to a Neanderthal individual by the anthropologist team headed by Dr. Eulalia Subirà. The different studies done on this bone piece gave very significant data of anthropologic and V. Villaverde*, P. M. Guillem**, R. Martínez-Valle**, A. Eixea* Throughout the years, there have been many studies covering a complete investigation of the archaeological site´s sequences. Samples of the different grounds were taken, the slopes deposition sequences were delimited, C14 dating was done, samples for palynological studies were kept, remains of presumably extinct animals were collected, vestiges of remote human activity were recorded and, in short, connections between all these discoveries were sought. During 2011 and 2012, some ashes and coal from two fireplaces with an estimated dating close to 100,000 years was recovered, and recently in 2013, a limestone piece was identified as a probable bear head with traces of parallel incisions resulting from a human-induced action, this will give rise to future studies as the piece is considered to be related to some ritual or symbolic activity. Cova Negra Location and background archaeological research Cova Negra is located on the left bank of the Albaida River after it passes through the Estret de les Aigües, in the town of Xàtiva (Valencia, Spain). Cova Negra is a clearly visible cavity. Its large dimensions, 18 metres wide and approximately 18 metres high at the entrance, and 25 metres long, provide enough light and offer good conditions for habitability (Fig. 1). * ** pathological nature which provided new proofs about the traditions of these societies. The use of sticks to palliate the pain produced by different processes of gingivitis was recorded and the results were published in a prestigious scientific magazine. J. Vilanova i Piera was the first to identify Cova Negra as an archaeological site. He cited it in his work “Origen, naturaleza y antigüedad del Hombre” in 1872 and included in one plate an illustration of one molar of Equus which was recovered at the site. G. Viñes conducted the first systematic archaeological fieldwork, between 1928 and 1933. Little data was obtained from this works, although a hu- Departament de Prehistòria i Arqueologia. Universitat de Valencia IVACOR. Generalitat Valenciana 361
MEDITERRANEAN BASINS. CENTRE.  immediately set, the whole skull, part of the face, some vertebrae and part of the rib cage...
362 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD dressing the nature of occupation, since from the previous campaigns, the information available only allowed a general interpretation of the recovered lithic assemblages and the faunal remains. As G. Viñes pointed out in one of the few publications concerning his excavations, Cova Negra presents severe disturbances and has been affected by numerous removals. Its location in a crossing point, its systematic occupation, and the loose consolidation of the sedimentary deposit have facilitated all these alterations that seriously affect the interpretation of its stratigraphy and chronology. In 2013, a new stage of archaeological fieldwork was set up, aimed once again at specifying the chronological issues of the deposit and deepening the nature of the lower level occupations, barely excavated in the campaigns from 1980 to 1991. These new works have been favoured by relevant progress made in the Paleolithic period in a regional context, mainly through data sets provided by excavations carried out at Cova del Bolomor (Tavernes de la Valldigna), El Salt (Alcoi) and Abrigo de la Quebrada (Chelva). Furthermore, revision of the chronological sequence proposed in the 1990s has become necessary, particularly in the light of new biostratigraphic data. Figure 1. General view of Cova Negra. man parietal bone was discovered and subsequently studied and assigned to the Neandertal species by M. Fusté. Materials recovered during the course of the excavations were enough to corroborate the relevance of the site and propitiate one preliminary attribution of the lithic assemblages to the Mousterian period. Additionally, the possibility that in the upper part of the deposit could be Capsian materials was suggested. Between 1950 and 1957, the Servicio de Investigación Prehistórica resumed fieldwork campaigns after a first evaluation and revision of Viñes’ results carried out by F. Jordá. In the course of this work, besides F. Jordá, D. Fletcher, E. Pla, V. Pascual and J. Alcacer also participated. Results led to the same chronocultural attribution, although the upper part of the sequence was now ascribed to the Lower Aurignacian. Finally, between 1981 and 1991, excavations were undertaken, again directed by V. Villaverde and aimed at clarifying the stratigraphy and ad- Because these works are currently in progress and the new seriation proposal has not yet been published, in the present text we will refer to the already published results, referring only the general traits that we continue reviewing. Stratigraphic sequence and chronology of the deposit (Fig. 2) The current interpretation of the stratigraphic sequence from Cova Negra comes from the work of M.P. Fumanal. Succinctly, it comprises 15 levels grouped in six sedimentary phases. Hereafter, we will describe their main characteristics: • Cova Negra Phase A. It is the first depositional phase represented by level XV, more than one metre thick. This is an allochthonous level of fluvial origin. This phase corresponds to a floodplain where clay materials were seasonally deposited in the cave by the overflow of the Albaida River. Consequently, the riverbed would be higher above than the present time, which in turn means that processes of terrace formations and down
362  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  dres...
MEDITERRANEAN BASINS. CENTRE. Figure 2. Cova Negra. Stratigraphic profile comparing different sectors excavated during the 1980s. cut of the river would have started at that time. From a paleoclimatic point of view, the absence of cold climate conditions indicators should be mentioned. Phase A is archaeologically sterile due to the flooding episodes that did not allow permanent human habitation of the site. ditions of high moistness, which had already weakened the cave dome. Human occupation happened sporadically and is associated with an increase in the organic components. One TL dating on the sediment corresponding to the base of level XV, gave the result of 117± 17 ky BP. • Cova Negra Phase B. This phase corresponds to levels XIV and XIII and possesses a totally different sedimentary characteristic compared to the previous phase. Along 170 cm, materials are accumulated in a yellow, loose sandy matrix that incorporates a discrete frequency of very altered coarse fraction on the base. This fraction increases towards the top of the deposit as a result of frost weathering processes. In between Phase A and Phase B there was an episode of ceiling collapse, indicating the first cold sign, where large blocks probably fell because of previous dissolution processes occurring in con- • Cova Negra Phase C corresponds to level XII and represents an interruption of the previous conditions. There is an almost total reduction of pebbles and gravels and an absolute dominance of fine-grained fraction, with a clay component that contrasts sharply with the sandy component of the limestone where the cave is situated. These conditions indicate a slightly rigorous climate with seasonal humidity pulses allowing soil formation processes typical of a warm climate. Sediment of Level XII has been dated by TL obtaining a chronological range between 107± 16 ky and 96± 14 ky. 363
MEDITERRANEAN BASINS. CENTRE.  Figure 2. Cova Negra. Stratigraphic pro   le comparing different sectors excavated during t...
364 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD • Cova Negra Phase D. After an abrupt contact, a new accumulation with different traits occurred. It has been interpreted as an erosive phase. This phase includes levels XI to V and its thickness ranges between 90 and 130 cm. The most characteristic feature is the persistent inclusion of angular clasts, fractured under cold environmental conditions, with milder pulses and variable degrees of humidity. A more rigorous climate is observed at levels XI, IX (very pronounced), VII and V. This cold trend decreases at level X (more sharply), VIII and VI (with possible anthropic influences). Therefore, at this phase there is a rhythmic oscillation on climatic conditions, where the effects of deterioration periods are mitigated intermittently, but always inside the dominant cold stage. TL dating obtained on thermoaltered flint from level VIII, gave 255 ± 20 ky and 206 ± 23 ky, also from level V there is a result of 235 ± 21 ky. Additionally, 14C dating from level V provided a radiocarbon date greater than 28.7 ky and less than 34.4 ky. • Cova Negra Phase E. It can be observed clearly in the west sector profile, where it represents a clear interruption compared to the previous phase. It corresponds to level IV that, after an abrupt contact (hiatus or erosive phase), is composed of an alteration level where a paleosol has preserved a clayed horizon, enriched secondarily by carbonates in the form of millimetre-sized nodules of CaCO3. Climatic conditions of this episode within a morphogenetic calm correspond to warm temperatures and seasonal rainfall where weathering erosion processes are replaced by gentle soil formation. This phase is 30 -35 cm thick in this sector. Two TL dates of sediments exist providing ages of 53.8 ± 8 and 50 ± 8 ky BP. • Cova Negra Phase F. This phase culminates the sedimentary and climatic sequence of Cova Negra. It is characterized by rigorous climatic conditions. This phase comprises levels III to I averaging around 60 cm thickness. It corresponds to a significant environmental aridity stage. Mechanical erosion prevails again, but clasts now have sharp edges. The end of this phase coincides with a superficial and disturbed level of variable thickness that includes abundant materials from Middle Paleolithic and some artefacts from Upper Paleolithic. These evidences seem to correspond to final periods of the regional Upper Paleolithic and, at least in the area excavated in the 1950s and the 1980s, they represent low intensity occupations, indicating that human occupation of the cave was interrupted during a relatively long period. Several factors have influenced the chronological interpretation of the sedimentary and climatic data described so far. First, the influence of the research tradition that from the initial work assigned the deposit to Würm chronology based on the study of faunal and lithic assemblages. Also, the identification of advanced traits on some lithic artefacts from the upper level contributed to support this chronological hypothesis. However, the excavation system in the 1950s clustered different levels in the same artificial unit with too much thickness. Finally, the limited knowledge of chronostratigraphic and paleoenvironmental issues in a regional context permitted this chronological proposition to be maintained. Accordingly, during the 1980s and 1990s, Phase A was correlated to the interglacial period Riss/Würm or the beginning of Würm (OIS 5e or 5d), considering that TL dating obtained supported this chronology. Phase B was related with subphases OIS 5 d-b, or stadial Würm I. Following the same logic, Phase C was assigned to OIS 5a, or interstadial Würm I-II. Taking into account its more rigorous climatic conditions, Phase D was associated with OIS 4, or Würm II, dismissing both TL dates on thermoaltered flint and 14C radiocarbon dates from level V. Finally, Phase E, with interstadial characteristics, was assigned to Würm interstadial, while the upper section of the sequence equivalent to Phase F, characterized by aridity and climatic rigour, would suggest a correlation with Heinrich 3 and 4 climatic events, considered a demonstration of the persistence of Middle Paleolithic until OIS 2. Ultimately, most of the archaeological sequence in Cova Negra was placed at the beginning of Upper Pleistocene, while its end would correspond to the Middle Paleolithic period that would have continued until late chronologies contemporaneous with the beginning of the Upper Paleolithic in other regions. This proposal is currently being reviewed in the light of new biostratigraphic indications provided
364  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD      ...
MEDITERRANEAN BASINS. CENTRE. by the revision of the faunal assemblages and the chronological implications derived from the fauna collection study. Furthermore, contributions from relevant sites such as Bolomor, el Salt and Quebrada have improved advances in the knowledge of regional chronological and biostratigraphical context. In 2013, new archaeological fieldwork started again in Cova Negra allowing a reconsideration of the stratigraphic sequence on the site, particularly concerning the postdepositional processes that affected the upper layers, and also providing a more precise sedimentary correlation of the levels at the different sectors of the site. It should be stressed that much of the upper section of the site presents a mixture of different deposits that refers to the Upper Paleolithic and, probably, to a classical Middle Paleolithic, corresponding to OIS 3 and 4. This interpretation is supported not only by lithic materials and some bone industry recovered during cleaning tasks on disturbed levels at the south sector (that can be clearly assigned to Gravettian, Solutrean and Magdalenian periods), but also by archaeological materials retrieved in the 1950 and 1980 campaigns of the same chronology, and some others Middle Paleolithic artefacts such as a Soyons point, which refers to late Middle Paleolithic in south-eastern France. This alteration of the upper levels was already pointed out by Viñes in his work in the 1920s, and, together with the pronounced slope towards the exterior of the cave, explains the presence of recent materials in most of the levels excavated in the 1950s, since this disturbed level, in some parts reaches, more than one metre of thickness. Bearing in mind the available biostratigraphic data and the existing information about lithic industry, a new chronological interpretation for the cave deposit can be proposed. Succinctly, it seems that Phase F in Cova Negra should be subdivided in two sections: the upper part, variable in thickness and completely disturbed, which can be assigned to different periods of the Upper Pleistocene; and some levels, in primary position and separated by an abrupt contact, which could be assigned to the Middle Pleistocene or the beginning of the Upper Pleistocene. The identification of Microtus breccensis has been determinant to support this interpretation. According to that proposal, the other phases of the sequence at Cova Negra should be correlated with Middle Pleistocene, and they beginning should not be situated far from OIS 7. Concerning the chronological evaluation formulated in previous works, a necessary change of perception should be assessed. This will have partial consequences in some other issues, such as the nature of the occupation of the cave and the evaluation of the human remains recovered, as well as a review of the archaeological materials from the campaigns of Viñes and Jordá. The human fossil record The human remains of Cova Negra are some of the richest Middle Paleolithic collections in the Iberian Peninsula and their study has been published in recent works (Arsuaga et al., 2007). Some part of the collection corresponds to disturbed levels and its chronology remains uncertain. However, other remains were recovered during excavations in the 1950s and their position in the sequence can be assigned imprecisely to the upper half of the deposit that corresponds to phases D, E and F. Only one mandibular fragment associated with one molar could be referred to phase C or D. Figure 3. Cova Negra. Right upper central incisor, mandibular fragment and right lower second deciduous molar. 365
MEDITERRANEAN BASINS. CENTRE.  by the revision of the faunal assemblages and the chronological implications derived from t...
366 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD with the superficial disturbed level, it is difficult to establish a chronological attribution. Nevertheless, considering the new chronological proposal, we can recognize this collection as one of the most important evidences of the oldest Neandertal human remains in the Iberian Peninsula, with a similar chronology to the neighbour sequence of the Bolomor site. All the remains that were assigned to a level or layer in the Viñes and Jordá campaigns correlate with OIS 6 and 5. It should not be ruled out that some human fossils associated with the disturbed level could even correspond to OIS 4 or 3. The morphological traits do not result discordant with the proposal that moves the chronology of the collection to the end of Middle Pleistocene. The significant representation of infant individuals should also be stressed, especially considering the lack of these kinds of evidences in other assemblages with similar chronology along the European recent Middle Pleistocene and the beginning of Upper Pleistocene. Characteristics of the lithic industry Figure 4. Cova Negra. A. Quina industry from lower levels; B. Levallois industry from upper levels. The total number of remains recovered until now comes to 26. From the excavations of G. Viñes there are two parietals. One occipital, one incisor, one mandibular fragment associated to one deciduous molar, two femora, one radius, one fibula and two metatarsal bones were recovered at the excavations carried out by F. Jordá. Also, during the excavations undertaken by V. Villaverde, nine cranial fragments, corresponding to right and left parietals, one deciduous molar, one premolar, and one molar were retrieved. Also, we should add the unpublished human remains from the campaign in 2013, one parietal fragment, one cranial fragment, one premolar and one incisor. In total, according to its position in the sequence and its morphological characteristics, a minimum number of seven individuals can be established: two adults, one juvenile and four infants. Taking into account the position of the remains in the sequence and the relation of some of them In the sequence of Cova Negra two different flaking methods have been observed: discoidal debitage and Quina retouch associated with the earliest phases (phases A, C and the lower part of D), and Levallois debitage identified in the latest phases (phases F, E and upper part of D). In any case, in phase B Levallois debitage has also been recorded in a representative proportion (Fig. 4). From an industrial point of view, retouched material is dominated by sidescrapers that reach very high percentages in the whole sequence. Plane and wide blanks predominate in the upper levels, the presence of deviated and transversal sidescrapers should also be highlighted, while thick sidescrapers with Quina and semi-Quina retouch, together with sidescrapers with bifacial retouch, sidescrapers with thinned back and limaces, are present in Quina and discoidal assemblages. Flint is the prevailing raw material, although some fine-grained quartzite and limestone materials with good flaking qualities exist too. In general terms, the low representation of cores is remarkable, as well as lithic debris and scars from retouching. This data, associated with a high proportion of retouched materials, indicates a limited representation of the initial phases of debitage and a large presence of the final
366  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  with...
MEDITERRANEAN BASINS. CENTRE. stages of the chaîne opératoire. These circumstances reinforce the idea of short-term occupations focused on the exploitation of some resources, including previously knapped lithic materials. Faunal assemblages and their paleoenvironmental and chronological implications Keeping in mind the complete sequence, it is possible to assess jointly the faunal remains recovered in the campaigns of the 1950s and 1980s, although the inherent constraints of the 1950 fieldworkshould be taken into account considering both the excavation system and the NW-SE dipping orientation. Cervidae are present along the whole sequence and are represented by three species: red deer (Cervus elaphus), fallow deer (Dama sp.) and roe deer (Capreolus capreolus). Red deer is particularly abundant in the lower levels. It is medium sized and has morphological traits close to the simplicidens type from the beginning of the Upper Pleistocene. Fallow deer also appear along the entire sequence, although with a moderate representation. Its anatomical proportions refer to the inferior limited proportions of Dama dama geiselana from the end of the Middle Pleistocene and beginning of the Upper Pleistocene. The remains of roe deer are less numerous, and are limited to some of the upper levels. Goat is also abundant in the upper levels. Its bones could be assigned to two species: tahr (Hemitragus sp) and ibex (Capra sp). Tahrs from the lower levels of Cova Negra belong to the species Hemitragus bonali, however the specimens of the upper levels (II-III) present minor dimensions and morphological traits characteristic of the species cendrensis, a well-represented species in the southeast of France, from the end of the Middle Pleistocene and the beginning of the Upper Pleistocene. The large amount of tahr in Cova Negra also reinforces its chronological attribution to the final of Middle Pleistocene and beginning of the Upper Pleistocene. Equidae, in the same way as other species mentioned above, are also present in the whole sequence, with values that vary substantially among levels. Horse (Equus caballus) is best documented species, whereas other remains of small sizes classified as Equus indet. might correspond to European ass. Other identified species, although with few remains, are auroch (Bos primigenius), narrow-nose rhinoceros (Stephanorinus hemitoechus), an undetermined rhinoceros, and wild boar (Sus scrofa). Regarding carnivores, up to eight species have been identified in recent campaigns: grey wolf (Canis lupus), dhole (Cuon alpinus), fox (Vulpes vulpes), lynx (Linx pardina), leopard (Panthera pardus), wildcat (Felis silvestris), brown bear (Ursus arctos) and hyena (species undetermined). In addition, the presence of lion (Pantera spelea), macaque (Macacus silvana) and straight-tusked elephant (Palaeoloxodon antiquus) registered in the campaigns of the 1950 should be included. Also, on the lower levels, there are abundant coprolites, probably from hyena. Finally, among the small mammals the presence of abundant remains of lagomorphs, especially European rabbit (Oryctolagus cuniculus), should be included, and a reduced number of hare (Lepus capensis), and beaver (Castor fiber), are documented at levels IX, III and II. Among the groups of birds, the number of species represented is very high, reaching 33 species. The most represented are rock dove (Columba livio/oenas), chough (Pyrrhocorax), and red-legged partridge (Alectoris rufa). Materials from the 1980’s campaigns enable us to specify trends at phases D, E and F in Cova Negra. On the one hand, there is progressive increase of caprines, dominated bytahr, and a gradual decrease of horses and bovinae. Regarding cervidae species, they are relatively abundant in levels IV and V, levels, where the bovinae also gain importance. In relation to the most meaningful species from an environmental point of view, we can only cite the wild boar, present at levels V and IIIb, the roe deer, present al level IIIb, the narrow-nose rhinoceros, documented at level IIa, and the beaver, present at levels IIIa and II. The above mentioned species do not facilitate going deeply into the environmental characteristics associated with these tendencies. Only the large number of horses at level VI can be pointed out to indicate the existence of open landscape, combined with forested areas suitable for deer. In the same way, roe deer could confirm the presence of a certain forested prairie. Levels V and IV, where an increase of deer and a decrease of horses and bovinae is registered, could correspond to moments of forested area expansion. Level IIIb has a low number of deer, while 367
MEDITERRANEAN BASINS. CENTRE.  stages of the cha  ne op  ratoire. These circumstances reinforce the idea of short-term occ...
368 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD tahr remains increase. This last species is associated with cool and moist climatic conditions, something that is consistent with the presence of roe deer in the same level, as well as the presence of brown bear, typical of forested habitats. Meanwhile, at level IIIa, the presence of narrow-nose rhinoceros could be signalling the existence of relatively extent open spaces. Finally, level II presents a high percentage of caprines, where tahr is dominant, but brown bear and beaver are also present. In any case, that the presence of carnivores could be responsible for the accumulations in the cavity and that their actions could distort the faunal spectrum should be taken into consideration. Also, the proximity of the river could explain the presence of gallery forest formations close to the cave, being shelter for particular species. Analyses of microfaunal remains have helped to refine the previous conclusions and above all to specify the chronology of the deposit. On the basis of the taphonomic study over remains recovered, it has been possible to establish that nocturnal birds of prey and carnivores were the main accumulator agents of rodents and insectivores. Specifically, the study of murid molars, which have better supported the postdepositional alterations, has enabled establishment not only of the presence of fox, strix and owl at different levels on the sequence of Cova Negra, but also to associate these accumulations to a lower or higher intensity of human occupation and several factors responsible for postdepositional alterations. In total, microfaunal remains from the 1980 campaigns rose 24,044 units. Distribution by sectors and levels is not equal, since lower levels present larger quantities, while the upper ones, corresponding to the west sector, are the ones with fewer remains. These differences are related to inherent transport factors considering the layers’ arrangement, but they are also associated with the different human occupation intensity, which was already mentioned. The location of Cova Negra, close to the Albaida river, origin of a large number of insects, and its latitude and altitude, provided, during cool oscillations of the Early Würm, different populations of Quiroptera shelter in the cave. A total of 15 different species have been documented in recent excavations deposits. Among the most significant remains there are deciduous teeth, distal epiphysis, not fused to diaphysis, and fetuses that refer to the moment of raising the young by colonies of some species such Rhinolophus ferrumequinum (Greater horseshoe bat), R. euryale (Mediterranean horseshoe bat), R. mehelyi (Mehely’s horseshoe bat), Myotis daubentonii (Daubenton’s bat), M. nattereri (Natterer’s bat), M. myotis (greater mouse-eared bat) and Miniopterus schreibersi (common bent-wing bat). Because of the calm that these species need in their reproductive cycle, their presence in the cave only could be explained if human occupation had very low intensity, which becomes significant data in order to interpret the nature of the occupation by Neandertal populations. On the other hand, the presence of very old adults together with infants seems to indicate the existence of hibernation colonies of particular species, such as Rhinolophus and Plecotus, since both, after the end of the winter period and in the same winter, a certain mortality rate occurred at these ages. All these Quiroptera species as well as rodents, build chronostratigraphic sequences that allow the comparison of Cova Negra with other Middle and Upper Pleistocene archaeological sites in a regional context, and also provide key information to envisage the climatic conditions developed during the formation of levels where they appear. Thus, at levels XIV and XIII (Cova Negra phase B) the presence of species that need moist and cool conditions has been registered, although the identification of Sorex sp., Sorex minutus and Allocritetus bursae might indicate a slight nuance of this cold. Level XII (Cova Negra phase C) presents a typical association of mild and moist environmental conditions. There, the insectivores disappear and the relation between murids and voles is typical of mild environmental conditions. Levels XI to V (Cova Negra phase D) present some associations pointing to an evolution from cold and moist to cold and dry conditions, alternated with some mild oscillations. In that sense, levels XI and IX are characterized by the equilibrium between voles and murids, with the presence again of Sorex minutus, while at level X insectivores disappear, indicating an amelioration of temperatures. Level VIII has a substantial postdepositional alteration caused by fire, whereas levels VII to V have low number of remains. Climatic characteristics of Level IV (Cova Negra phase E) cannot be defined from microfaunal remains either, since they have poor representation. Meanwhile, at levels III to I (Cova Negra phase F), Sorex disappears, Allocritetus bursae persists and there is a proportion between murids (39.5) and voles (26.3), indicating drier climatic conditions. In any case, it should be
368  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  tahr...
MEDITERRANEAN BASINS. CENTRE. noticed that Quiroptera associations reflect mild climatic conditions, typical of the Mediterranean climatic zone at the present time. From a biostratigraphic point of view, the absence of Terricola duodecimcostatus and Microtus arvalis in Cova Negra, species that are documented in the Upper Pleistocene deposits of El Salt and Abrigo de la Quebrada, and the relevant presence of M. Brecciensis, point to a Middle Pleistocene chronology for the Cova Negra deposit. On the other hand, the reduced length of Allocricetus bursae correzensis m1 and the presence of Arvicola aff. sapidus in Cova Negra, as well as its morphological characteristics suggest late phases of the Middle Pleistocene, without ruling out the inclusion of levels II to III at the start of the Upper Pleistocene. Level I, being very thick and disturbed, would incorporate Upper Pleistocene materials, both from the Middle Paleolithic and the Upper Paleolithic. The nature of the occupation of the Cova Negra site The information regarding spatial information is limited to the data provided by the excavations of the 1980s in the west sector and also, some general inferences can be made from the quantification of materials recovered in campaigns in the 1950s. In general terms, occupations in the lower half of the sequence, as well as in the upper section of the sequence, were short-term and spaced between long periods of abandonment. In any case, findings density is particularly low at Cova Negra Phase B (levels XIV and XIII). Data available from the level III west sector refine the occupation system. Within an area around 12m2 different ashen spots, associated with fire-cracked stones and rubified sediments were located. These very simple hearths overlapped, creating a real palimpsest that, in turn, imply repeated occupation patterns in a level without a high density of lithic and faunal remains. In fact, some bones show complex superpositioning taphonomic processes. Specifically, data of carnivore toothmarks and anthropogenic cutmarks, suggesting a rapid succession of episodes of human and carnivore occupation. Short-term occupation, associated with accumulations of hunted animals and consumption activities around hearths, were followed by abandonment episodes and reoccupation of the cave by carnivores, responsible both for the scavenging of the food leftovers discarded by Neanderthals and for the incorporation of new faunal remains consumed, particularly tahr bones (Hemitragus sp.). The construction of simple hearths, a low quantity of animals, especially herbivores, and the partially represented reduction sequence of lithic materials, indicate a model of short and sporadic occupations occupying relatively limited surface areas. In this case, an ellipsoidal area, close to the west wall of the cavity and delimited by large fallen blocks, where organic remains and human occupation evidences are concentrated. This spatial pattern is repeated in the whole cave. Profiles exposed in different archaeological campaigns allow us to notice how accumulations of hearths and organic depositions are concentrated at particular points varying along the sequence. The use of the space was always restricted to relatively limited surface areas. The number of faunal remains recovered at sectors excavated in the 1950’s campaigns corroborates the same pattern. Density of bone findings at level IIIB is 17.5 items per m2, and that of lithic remains is 2.5 per m2.In the 1950’s campaigns, Level III, excavated in a total surface of around 130 m2, yielded a bone density of 1.9 items per m2, and 3.9 items per m2 of lithic remains. Ultimately, the anthropic evidences show spatial variations that also validate a shortterm spatially variable occupation pattern concentrated in the most sheltered area of the cave, the southwest sector. This data, that suggests small groups of hominids, is completely consistent with the existence of carnivore accumulations of large and small herbivores, and the repeated presence of colonies of bats for hibernation and raising their young in the cave. 369
MEDITERRANEAN BASINS. CENTRE.  noticed that Quiroptera associations re   ect mild climatic conditions, typical of the Medi...
370 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Oreto García Puchol*, Joaquim Juan Cabanilles**, Sarah B. McClure***, Josep Lluís Pascual Benito****, Bermat Martí Oliver *****, Manuel Pérez Ripoll******, Joan Bernabeu Aubán*******, Salvador Pardo Gordó********, Lluís Molina Balaguer*********, Yolanda Carrión Marco*********, Agustín Diez Castillo********** The Last Hunter-Gatherers in Cueva de la Cocina (Dos Aguas, Valencia, Spain) 1. Presentation J. Fortea (1973) characterised the “facies Cocina” of the Geometric Epipalaeolithic, attributed to the “Tardenoid” tradition, based on the outstanding archaeological record obtained at the eponymous site of Cueva de la Cocina. This large cave, measuring 20 x 15m, is at 405m a.s.l., in the last foothills of the Sierra del Caballón, between the River Júcar and the coastal plain of Valencia. It is situated in a valley with difficult access (La Canal), at the modern base of Ventana Ravine, which forms part of the drainage of the valley through a vertical exit to the Jalón Ravine. A large lithic assemblage including nearly two thousand geometric armatures, a hundred portable art objects consisting of plaquettes with engravings and remains of paint (geometric linear art) (Fig. 1), and a series of painted motifs on one of the cave walls, interpreted by L. Pericot (1945) as Levantine art and by J. Fortea (1974) as geometric linear art, have captured attention at this key archaeological site for understanding the socio-ecological dynamics of the last hunter-gatherers and first farmers in the western Mediterranean. Research at Cueva de la Cocina goes back to the 1940s, when L. Pericot carried out a series of excavations from 1941 to 1945. The last year’s fieldwork provided the fullest stratigraphic sequence by including Neolithic levels in the upper part of the deposit (Pericot 1945). Cocina’s international Figure 1. Engraved plaquette (geometric linear art) from L. Pericot’s excavations. recognition came when J. Fortea (1973) examined the remains from the 1945 season as part of a wider study of the micro-blade and geometric lithic assemblages of the last hunter-gatherers on the Mediterranean side of the Iberian Peninsula. He interpreted the archaeological sequence at Cocina * Investigadora Programa Ramón y Cajal. Departament de Prehistòria i Arqueologia. Universitat de València. oreto.garcia@uv.es ** Museu de Prehistoria de València –S.I.P.–. joaquim.juan@dival.es *** Department of Anthropology. The Pennsylvania State University. sbm19@psu.edu **** Museu de Prehistoria de València –S.I.P.–. josep.ll.pascual@uv.es ***** Museu de Prehistoria de València –S.I.P.–. bernat.marti@dival.es ****** Departament de Prehistoria i Arqueologia. Universitat de València. manuel.perez@uv.es ******* Departament de Prehistoria i Arqueologia. Universitat de València. juan.bernabeu@uv.es ******** Departament de Prehistoria i Arqueologia. Universitat de València. salvador.pardo@uv.es ********* Departament de Prehistoria i Arqueologia. Universitat de València. lluis.molina@uv.es ********** Departament de Prehistoria i Arqueologia. Universitat de València. yolanda.carrion@uv.es ********** Departament de Prehistoria i Arqueologia. Universitat de València. agustin.diez@uv.es
370  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Oret...
MEDITERRANEAN BASINS. CENTRE. in four phases, two initial Mesolithic phases (A and B, corresponding to Levels I and II in Cocina) and two Neolithic phases (C and D, or Levels III and IV), where the latter were interpreted in terms of the neolithisation of the Mesolithic substrate. At the same time, Fortea began a series of annual excavations that lasted from 1974 to 1981. The results were presented in a brief summary that anticipated their potential interest (Fortea et al., 1981). However, most of the documentation obtained in L. Pericot’s excavations and Fortea’s fieldwork has never been published. 2. New Research Perspectives record. In addition, Cocina has yielded an ensemble of Mesolithic portable art objects that is unique in Iberia, and should be regarded as reflecting the symbolic world of the pre-Neolithic groups. Their presence might evoke the particular significance of the place/cave as a special social space. The confirmation of the age of a burial partially excavated in the 1943 season, at the base of the sequence, would add new elements to this interpretation. The hearths described by the excavators would have been the focus of the domestic and social area, located in the front half of the cave. The Neolithic levels at Cocina reflect a series of poorly-defined occupations yielding some im- A full study of the site within the context of neolithisation in Mediterranean Iberia has been started with the project HAR2012-33111. Several significant aspects of this process converge in Cueva de la Cocina. A recent programme of radiocarbon dates situate the Mesolithic sequence between the mid seventh millennium and first half of the sixth millennium cal BC (Juan Cabanilles and García Puchol 2013; Perrin et al., i.p.). When calibrated, the most recent dates overlap slightly with the oldest dates for the first Neolithic occupations in the central-southern part of Valencia (obtained at such sites as Mas d’Is and En Pardo). The Mesolithic sequence is characterised from bottom to top by a regular and standardised flint blade technology which follows a pattern that is common to the late Mesolithic in the western Mediterranean (Castelnovian, upper Capsian) and aimed to made trapezoidal geometric projectiles using the micro-burin technique (Binder et al., 2012; García Puchol and Juan Cabanilles 2012). In the initial phase (Cocina 1 or Phase A, second half of the seventh millennium cal BC) asymmetrical trapezes with abrupt retouching and concave edges predominate, whereas on the turn of the sixth millennium, Cocina or Muge-type triangles with concave sides increase in numbers (Cocina II or Phase B). In this way, the site was visited repeatedly over nearly a millennium, within a mobile strategy of territorial use, possible connected with coast-interior movements as suggested by the significant number of malacological and ichthyological remains. However, the hunting of ibex, and also red deer and wild boar, was the main form of subsistence procurement. Lithic reduction and food-processing are the main daily activities documented in the material Figure 2. Chrono-cultural sequence at Cueva de la Cocina, after J. Fortea and O. García-Puchol. Ideal stratigraphic section in Sector E1 based on L. Pericot’s excavation logbooks. 371
MEDITERRANEAN BASINS. CENTRE.  in four phases, two initial Mesolithic phases  A and B, corresponding to Levels I and II in...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 372 pressed cardial sherds, characteristic of the earliest Neolithic in the region (second half of the sixth millennium cal BC), overlying the Mesolithic levels. Later visits in the Chalcolithic and Bronze Age have also been documented. J. Fortea proposed an explanation based on the neolithisation of the local population with the introduction of certain Neolithic technological elements, such as the pottery, but hardly any changes in subsistence strategies until the Chalcolithic. A recent reappraisal of the same sector studied previously by Fortea and Pericot (Sector E1 in 1945), focusing on the techno-typological assessment of the lithic assemblage from a diachronic approach (García Puchol 2005) has returned to Pericot’s observations and impressions about stratigraphic differences between the levels with pottery and the previous Mesolithic layers. The hypothesis to be tested in the new re-examination of the data, through a programme of chrono-stratigraphic and bio-archaeological analysis, relates to the interpretation of a break between the Mesolithic Michael Walker*†, Mariano LópezMartínez**†, María Haber-Uriarte***† © All rights reserved. ** *** † † The expansion of the earliest Neolithic towards inland Valencia is marked by finds of a small number of impressed potsherds at sites in the Caroig massif and Sierra del Caballón (including Cueva de la Cocina) and rock paintings classed as Early Schematic Art, reminiscent of decoration on the impressed ware (Martínez Rubio and Martorell Briz 2012). These elements attest the Neolithic advance in a region where a well-structured Mesolithic population existed just a short time before, according to the available radiocarbon dates. The research described here has been carried out in the framework of the project “MESO COCINA: the last hunter-gatherers and the neolithisation paradigm in the western Mediterranean” (HAR201233111) funded by the Ministry of the Economy and Competitiveness of the Government of Spain, and the ANR research program “The last hunter-gatherers of Western Europe” (PI: Pierre Allard). Cueva Negra del Estrecho del Río Quípar (Caravaca de la Cruz, Murcia, Spain) Cueva Negra is 10 km S of Caravaca de la Cruz, lying at 740 m a.s.l. (metres above sea level) and 40 m above the R. Quípar where it flows northwards out of a gorge (“Estrecho”) below the hamlet of La Encarnación (Fig. 1). The large rockshelter contains a noteworthy depth of Pleistocene * and Neolithic or the identification of continuity in the terms proposed by Fortea (Fig. 2). sediments cursorily explored in 1981 (MartínezAndreu et al., 1989). It lies in Upper Miocene (Tortonian) biocalcarenite rock on the right-hand side of the narrow gorge through which the R. Quípar Gorge descends before joining the R. Segura which reaches the Mediterranean Sea 110 km Departamento de Zoología y Antropología Física, Facultad de Biología, Universidad de Murcia, Campus Universitario de Espinardo Edificio 20, 30100 Murcia, España. Correo electrónico: mjwalke@gmail.com Tfnº: 34-620-267104 Calle Pintor Joaquín 10-4º-I, 30009 Murcia, España. Correo electrónico: marianolopez@hotmail.com Tfnº: 34-630-408806 Departamento de Prehistoria, Arqueología, Historia Antigua, Historia Medieval y Ciencias y Técnicas Historiográficas, Facultad de Letras, Universidad de Murcia, Campus Universitario de La Merced, Calle Santo Cristo 1, 30001 Murcia, España. Correo electrónico: mariahaber@pi-ma.es Tfnº: 34-629-756183 Directores de la excavación, MUPANTQUAT, Murcian Association for the Study of Palaeoanthropology and the Quaternary, Asociación Murciana para el Estudio de la Paleoantropología y del Cuaternario, http:www.mupantquat.com (Museo Arqueológico de Murcia, Avenida Alfonso X El Sabio 7, 30008 Murcia, España), toda correspondencia a: Secretario de MUPANTQUAT, M.López Martínez Grupo de Investigación E005-11 de Ciencias Experimentales de la Universidad de Murcia, “Quaternary Palaeoecology, Palaeoanthropology and Technology” (Inv.Resp., Dr.J.S.Carrión García, Departamento de Biología Vegetal, Facultad de Biología, Universidad de Murcia, Campus Universitario de Espinardo Edificio 20, 30100 Murcia, España)
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  372  pres...
MEDITERRANEAN BASINS. CENTRE. E of the site that nevertheless is but 75 km N of the southern Murcian coast. Systematic excavation began in 1990 and 25 field seasons have taken place. For some years neither the chronology nor the complexity of the Pleistocene geology were understood correctly. Inaccuracies and mistaken interpretations in earlier publications were corrected in the 2013 revision (Walker et al., 2013) that supersedes them all and significant aspects of it are summarized here (earlier publications cited here are preceded by indicating they contain some unreliable information, usually a chronological attribution that is too young, sometimes incorrect faunal assignation, occasionally a geological error). The 5 m-deep Pleistocene sedimentary fill (Fig. 2) is assigned by magnetostratigraphy to the Matuyama magnetochron >0.78 Ma (Scott and Gibert, 2009). Optically-stimulated sediment luminescence implies >0.5 Ma and mammalian biochronology indicates >0.7-<1 Ma (Walker et al., 2013): e.g. the extinct Arvicolid rodents Mimomys savini, Microtus (Iberomys/Terricola/Pitymys) huescarensis huescarensis, Pliomys episcopalis, Allophaiomys (Microtus/Euphaiomys) cf. chalinei, Stenocranius (Microtus) gregaloides; the extinct Cervids Megaloceros aff. savini and Dama cf. nestii vallonnetensis; the Rhinocerotid Stephanorhinus cf. etruscus; the Equid Equus altidens, etc. Sediment micromorphology shows the fill represents near-horizontal, gradual, intermittent fluviatile accumulation (Angelucci et al., 2013) with no significant horizontal or vertical discontinuities (pace Jiménez-Arias et al., 2011). Mammals, birds (including waterfowl), reptiles and amphibians corroborate pollen (Carrión et al., 2003) typical of mild (MIS-21), damp, fluvio-lacustrine environments. Anne Eastham identified >60 bird species (Walker et al., 1998) implying nearby biotopes of (1) lakes and rivers with temperate woodland, (2) open mixed woodland, (3) open grassland and Figure 1.Cueva Negra del Estrecho del Río Quípar in its surroundings.. 373
MEDITERRANEAN BASINS. CENTRE.  E of the site that nevertheless is but 75 km N of the southern Murcian coast. Systematic ex...
374 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 2. Cueva Negra del Estrecho del Río Quípar: Plan and sections.. heath, and (4) craggy mountainsides. That suggests the site was frequented owing to its well-favoured position in surroundings with noteworthy biodiversity, though it may have been taken over by birds whenever flooding required mammals to abandon the cave, perhaps seasonally. Cueva Negra “pre-Neanderthal” (Homo cf. heidelbergensis) teeth (Fig. 3) give several measurements (Table 1) outwith modern ranges; e.g. large antero-posterior measurement at the neck, or cervix, between crown and root (cervical vestibulo– or bucco-lingual dimension), typical in Neanderthals and archaic humans and often interpreted as a “buttressing” adaptation to using front teeth as a vise. Extreme tooth wear (attrition) of Cueva Negra front teeth, exposing dentine and the pulp– or root canal (rendering lingual crown height unreliable and uninformative) is also typical of Neanderthals and archaic humans, perhaps caused by using front teeth as a vise; in modern humans tooth wear and exposure of dentine occurs mostly on crowns of back teeth, only rarely on front teeth. One incisor tooth crown has a “shovel” form (a broad vertical scoop) on its internal (lingual) surface; “shovelling” is common on Neanderthal incisors. A canine tooth with occlusal attrition that exposed the pulp canal has a root that is much longer than in modern humans though comparable in length with some Neanderthal canines (e.g. from Grotte d’Hortus). The two premolars lack transverse crests and are thus unlike Neanderthals (though 17% of Sima de los Huesos premolars lacks them also). The most important findings at Cueva Negra concern human activity ca. 0.Ma. Two aspects are of especial interest. First, undoubted evidence of fire has been uncovered (Fig. 4), sealed within a 4.5 m-deep layer of sediment 5 m back from the present entrance, perhaps still further back 0.8 Ma if rock overhanging the entrance has undergone erosion since then. Geoarchaeological investiga-
374  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
MEDITERRANEAN BASINS. CENTRE. swamps, enabling dismemberment and roasting). This does not mean they could reproduce or control fire; there is a dearth of archaeological evidence for hearths or fire-pits before 0.5 Ma. Nevertheless, fire ca. 0.8 Ma supported hominin cognitive versatility, techno-manual dexterity, and palaeoeconomic extractive behaviour in long-vanished Western European palaeoecological and palaeobiogeographical contexts. Cueva Negra exemplifies those aspects; Palaeolithic finds imply resources were exploited as far away as 40 km downstream and 30 km upstream from the site (Zack et al., 2013). That range is unsurprising given that ≥1.3 Ma early humans had begun migrating into Western Europe from northern Africa or western Asia, and therefore they could not have been congenital stick-in-the-muds even though, plausibly, their preferred habitats were localities with abundant biodiversity to hand (cf. ! Walker et al., 2006). Figure 3. Cueva Negra del Estrecho del Río Quípar: Handaxe, human teeth. tions of the sediment suggest combustion (Angelucci et al., 2013) and recent geochemical analysis supports that. Since 2011 excavation has found both thermally-altered, lustreless chert, with potlid fractures and conjoined splintering caused by thermal shock to both nodules and artificiallystruck flakes, and also charred burnt animal bone and white calcined fragments showing conjoined lengthwise long-bone spalling typical of circumferential shrinkage after thermal volatilization of organic components (Walker et al., 2013). Recent taphonomical analysis and electron microscopy of bone fragments attribute discolouration to burning, not to post-depositional mineral staining, and both Fourier Transform infrared spectroscopy and electron spin resonance analysis of chert and bone imply firing temperatures ca. 550ºC (Walker et al., in preparation). A fire-place is not a hearth. The Cueva Negra humans could have brought glowing brands left by a forest fire into the cave to establish and tend a fire where rain or wind would not put it out. They may well have been less afraid of fire outside than other animals they saw fleeing from it (which could have led them to play with fire in order to drive animals towards natural death-traps, such as The excavated Palaeolithic assemblage includes a bifacially-flaked “Acheulian” limestone hand-axe, though it mostly consists of small chert, limestone or quartzite artifacts (<60 mm long), knapped on site, often by bipolar reduction or repetitive centripetal flaking of small discoidal (“Levallois”) cores, and often showing marginal retouch that is mainly steep-angle (>50º) and sometimes abrupt (“Mousteroid”), and very occasionally invasive or semi-invasive low-angle (<30º) (from the hand-axe to very small chert flakes <30 mm long). Serrated, notched or denticulate edges occur, and pieces bearing one or two large notches are common. Some flakes and several flattish or laminar subrectangular fragments were knapped to give steep abrupt (“Mousteroid”) edge-retouch (Fig. 5). Steep retouch on a piece of flattish laminar chert can transform its perpendicular edge to give an acute angle useful for cutting or scraping. It is plausible to see that as being very different indeed from abrupt retouch of “scrapers” in most Mousterian assemblages where steep retouch applied to thin feathered flakes could spare them from accidental breakage by snapping during use or may have been applied to resharpen a cutting tool. Well-formed feathered flakes with striking platforms and bulbs of percussion are fairly uncommon at Cueva Negra, whereas small fragments of laminar chert abound. Many of the small artifacts seem to have much in common with those from the penecontemporaneous Cat- 375
MEDITERRANEAN BASINS. CENTRE.  swamps, enabling dismemberment and roasting . This does not mean they could reproduce or co...
376 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 4. Cueva Negra del Estrecho del Río Quípar: Deep layer with thermally altered remains. Table 1 Cueva Negra fossil human teeth by type, metre-square, layer and spit (right), and measurements taken (below) right mandibular permanent medial incisor B1i(1) left maxillary permanent lateral incisor B2f(1-2) left maxillary permanent canine C3e(2c) right mandibular anterior premolar C2e(3ñ) left mandibular anterior premolar C1a(1-2) anterior permanent tooth root C4g(2c) incisoapical height 23,0 mm 25,2 mm 27,2 mm 21,9 mm 22,9 mm no crown mesiodistal crown dimensión 5,7 mm 7,6 mm 7,0 mm 11,8 mm 6,2 mm no crown buccolingual crown dimensión 7,9 mm 9,6 mm 7,8 mm 8,0 mm 7,6 mm no crown buccal crown height 7,7 mm 9,6 mm 8,2 mm 9,4 mm 10,4 mm no crown lingual crown height 6,1 mm unreliable unreliable unreliable unreliable no crown buccal height of root 15,3 mm 7,9 mm 19,0 mm 14,3 mm 14,6 mm 22,5 mm mesiodistal dimension at neck 5,4 mm 5,8 mm 6,5 mm 4,6 mm 5,5 mm 7,3 mm buccolingual dimension at neck 7,7 mm 7,8 mm 8,0 mm 7,7 mm 7,0 mm 8,1 mm maximal mesiodistal dimension of root 5,0 mm 5,3 mm 5,3 mm 3,7 mm 5,0 mm 5,5 mm maximal buccolingual dimension of root 8,3 mm 8,7 mm 8,0 mm 6,5 mm 6,9 mm 7,6 mm Table 1. Cueva Negra fossil human teeth by type, metre-square, layer and spit (right), and measurements taken (below).
376  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
MEDITERRANEAN BASINS. CENTRE. alan site of Vallparadís (Martínez et al., 2010) and from the Italian site of Isernia La Pineta, rather than with assemblages ≥1 Ma from Atapuerca and Orce. From a descriptive viewpoint of stone-knapping techniques the assemblage may be called “AcheuloLevalloiso-Mousteroid” (Walker et al., 2006; Walker et al., 2013; Zack et al., 2013); this descriptive technomethodological approach can be uncoupled, viewed epistemologically, from prescriptive ontological typologies influenced by relative chronological inferences drawn tautologically from conjectural “culture history” and pseudo-evolutionary conjectures about Palaeolithic technology. Once freed from the dead hand of traditional perspectives, other aspects of the assemblage come to the fore. Several small retouched artifacts seem to fall into overlapping groups, in contrast to some other Spanish Early Pleistocene assemblages that have been called “Oldowan” due to perceived similarity to African ones; the term is inappropriate at Cueva Negra because, unlike typically Oldowan artifacts in Africa, nearly all those excavated at Cueva Negra are <60 mm in size. Steep retouch is seen on many pointed pieces; some are flattish pieces and could be regarded as fine points, “awls”, or “perforators”, whereas others resemble thick “Tayac points” described often in Middle and early Late Pleistocene European assemblages. Pointed artifacts include “becs”, small chunks of chert from each of which there projects incongruously a delicate elongated tiny spur, or “beak” (fancifully bringing to mind a small bird head with its beak). There are also many steeply-keeled fragments; some resemble steep scrapers on short stumpy cores, whereas others, knapped into elongated keeled planoconvex shapes resembling garden slugs (“limaces”) may be called “proto-limaces”. Beaks and slugs could be interpreted as convergent steep scrapers, or where both ends are pointed they could be envisaged as thick double points. However, researchers at 0.7 Ma Isernia La Pineta argue that its beaks and slugs are what were left behind after their reduction by bipolar knapping to remove extremely small flakes used as unretouched tools, backing their argument up with microscopical use-wear analysis and experimental knapping (Crovetto, 1994; Crovetto et al., 1994; Peretto, 1994; Peretto et al., 2004). Flakes produced by bipolar knapping occur at Cueva Negra, though they are yet to be quanti- fied because quantification of bipolar elements depends on whether carinated pieces with notches, spurs (beaks) and planoconvex double-ended slugshaped pieces, were outcomes, first and foremost, of bipolar core-reduction to remove usable flakes, or whether, instead, they were primarily fashioned intentionally for use as implements themselves. The two possibilities need not be mutually exclusive because comparable pieces have been interpreted as implements, sometimes supported by microscopical use-wear analysis; an extensive literature exists with references to “limaces”, “becs” and “microperforators” from Pleistocene and Holocene lithic assemblages Europe, Africa, and North and South America. Most Cueva Negra artifacts are “expedient”, frequently of “informal” shape, implying “opportunistic” or “eclectic” technological behaviour. They bring to mind the different blades of a Swiss knife. It is perhaps unsurprising that retouch is seen as often on stone fragments as on well-made flakes struck from prepared discoidal cores by recurrent repetitive centripetal flaking, given that at 0.8 Ma secant-plane control of knapping was in its infancy worldwide. It should be borne in mind that such cores are known from 1.3 Ma in Africa (de la Torre et al., 2004) where hand-axes have even greater antiquity (1.7 Ma) and that both of these involved bifacial flaking albeit with different formal secant-plane implications (asymmetrical and symmetrical, respectively), though more eclectic informal (“Oldowan”) tool-making continued alongside them. Extraction of regular flakes by recurrent repetitive centripetal flaking of prepared discoidal cores is demanding in both cognitive and technical terms (Coolidge and Wynn, 2005); the putative flakes are, as it were, “hidden” from view (like the yolk inside a hen’s egg, so to speak), and “unimaginable” simply from looking at the external shape of the stone before the reduction sequence begins. Evidence at Cueva Negra of both bifacial handaxe production and recurrent repetitive centripetal flaking of prepared discoidal cores, together with a diverse range of small artifacts, implies manual dexterity, technical aptitude and cognitive versatility. This raises the question of how those who frequented Cueva Negra 0.8 Ma perceived and exploited their surroundings, particularly where they obtained raw materials for stone tools. Two different matters are relevant. First, how far were the different possibilities of different rocks perceived? Secondly, were outcrops available then that nowadays afford suitable stone? 377
MEDITERRANEAN BASINS. CENTRE.  alan site of Vallparad  s  Mart  nez et al., 2010  and from the Italian site of Isernia La ...
378 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD The hand–axe shows 30 fresh bifacial extractions on a flattish limestone cobble with some cortex still present and a similar cobble had 15 unifacial fresh extractions along one side. Both probably were obtained from fluvio-lacustrine gravels, though X-ray diffraction and petrography indicate that their grey-blue micritic limestone (94% calcite; 6% quartz) originated in Lower Jurassic (Lias) rocks (Walker et al., 2006). An unworked cobble from Cueva Negra is a dismicrite containing 10% quartz, radiolarian fragments, and filamentous planctonic fragments, characteristic of Middle Jurassic (Dogger) strata. Lower and Middle Jurassic beds are exposed in mountainsides upstream from the site. Another unworked limestone cobble from the site lacks quartz, being oolitic sparite (oosparite). Two limestone cobbles from a small conglomerate outcrop 0.8 km E of the site also lack quartz, one being composed of cryptocrystalline limestone pellets of organic faecal origin, the other of sparite cement with microscopical fossils. The aforementioned conglomerate outcrop is an Upper Miocene (Tortonian) marine conglomerate, containing complete Ostreid and Pectinid sea-shell fossils, that was an inshore Tethys Sea deposit of cobbles and stones eroded out of the nearby mountainside and later cemented by CaC03 when intense Upper Pliocene and Early Pleistocene neotectonic activity lifted up the mountains and their erstwhile sea-shore to its present height of 750 m a.s.l. (the Tortonian strata dip strongly to the SW and 0.8 km away lie at 730 m a.s.l. below Cueva Negra; the steep dip was not taken into account in some early publications, leading to mistaken interpretations). The cemented cobbles and stones are of limestone, chert and quartzite; several were taken at Cueva Negra and at the outcrop Palaeolithic artifacts have been picked up similar to those of the rock-shelter, including a small prepared discoidal chert core. The chert raw material includes eroded frangible tabular nodules derived from chert blocks or slabs of subparallelepiped shape. They are best described as “fissible” (Stein, 1981: 537) because hammering on them often fails to elicit conchoidal fractures or produce feathered flakes with well-developed convex bulbs of percussion. If hammering does not simply shatter the chert blocks into very small chips and fragments, it may split them open along fissible flat planes defined by inter- nal structure or impurities, and produce flattish sub-rectangular laminar fragments available for modifying as tools. Loose chert cobbles and blocks (some upto 0.3 m across weighing 5 kg) abound <5 km S of Cueva Negra from 770 up to 890 m a.s.l. on the flanks of mountains from whose crags of Jurassic limestone chert was eroded. Massive continental lateral erosion took place in the Upper Pleistocene and initial Pleistocene, leaving high-altitude vestiges of an erstwhile vast gravel spread 100-120 m thick (“raña”; “glacis”) containing conglomerate bands, the base of which lies ca. 45 m above the river today (Walker et al., 2013; Zack et al., 2013). Later on, much of that ancient gravel was displaced both laterally and longitudinally and redeposited, owing to ongoing Early Pleistocene erosion induced by falling base level as uplift continued. The process gave rise, in two or perhaps three depositional cycles, to horizontally-bedded outcrops of gravels and fluvio-lacustrine conglomerate that abound at relative heights of 5-25 m above the river today. These outcrops often include noteworthy fossil vestiges of local lakes. The matter is complicated by unequal uplift of the sides of the active longitudinal shear fault along which the R. Quípar runs (unequal uplift saved the Cueva Negra sediments from riverine erosion), and by unequal spatiotemporal activity of those faults normal to it which probably determined development or drainage of hanging lakes upstream in the upper Quípar valley (where it is called the Rambla de Tarragoya). In short, there is a wealth of possible secondary or even tertiary sources of eroded Jurassic chert, but just which were available to Cueva Negra chert-knappers called out for forensic research. There are also a few small primary chert outcrops. One is of radiolarite ca. 40 km downstream from the rock-shelter where a radiolarite scraper was found in 2013. Another is of lightbrown tabular chert at Río Caramel, a tributary of the R. Guadalentín that joins the R. Segura near Murcia; the upper Quípar is separated by a watershed from the Guadalentín river system. It has not been possible yet to determine whether light-brown chert at Cueva Negra came from the outcrop. At high altitude on the watershed itself there is a very small outcrop of biogenic chert that probably formed in a freshwater Pliocene lake, which has the unusual frondose
378  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  The ...
MEDITERRANEAN BASINS. CENTRE. cactus-like form not unlike that of East African Lake Magadi flint; at least one flake at Cueva Negra may well have come from the outcrop. Primary chert outcrops seem not to have been exploited intensively. Most Cueva Negra chert originated in Jurassic rock strata, albeit obtained from secondary or even tertiary gravel or conglomerate accumulations. Which ones? Petrographically the cherts look much alike. Might help come from chemical finger-printing of chert? Thanks to collaboration with the University of Arizona help came from the laser-ablation inductively-coupled plasma mass-spectrometry of trace elements present in several of the Earth’s crustal rocks (Sc, V, Cr, Co, Zn, Ga, Ge, Rb, Sr, Y, Zr, Nb, Cs, Ba, La, Ce, Pr, Nd, Sm). Chert samples were taken from a number of outcrops for comparison with chert from Cueva Negra (Zack et al., 2013). Multivariate factorial analysis of the data show most chert fragments analyzed from Cueva Negra resemble chert sampled at the Tortonian conglomerate 0.8 km to the E (Fig. 6). A few Cueva Negra fragments more closely resemble samples from fluvio-lacustrine gravel outcrops upstream in the Rambla de Tarragoya as far as the headwaters of the valley (where we even picked up a “proto-limace”) ca. 25 km S of the site. Plausibly, variation in trace-element composition cherts that had formed at separate localities or times in the Jurassic is reflected in compositional differences between outcrops of gravels that received chert nodules eroded from the mountainsides nearest by. Some of those cherts were taken to Cueva Negra. Trace-element characterization indicates how far humans ranged. Research at Cueva Negra throws new light, including fire-light, on the cognitive versatility, manual dexterity and technical aptitude of early humans ca. 0.8 Ma in S.E. Spain. They exploited their surroundings in a competent fashion that implies precise knowledge and accurate awareness of what was available for survival. Research continues both in the field and laboratory at this intriguing late Early Pleistocene site. Acknowledgements: We thank all research collaborators named in the text or references as well as palaeontologists Drs. A. Ruiz Bustos, J. van der Made, X. Murélaga Beirucua and archaeologists Drs. D.A. Roe and I. Martín-Lerma for their kind help. 379
MEDITERRANEAN BASINS. CENTRE.  cactus-like form not unlike that of East African Lake Magadi    int  at least one    ake at...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 380 B. Galván*, C.M. Hernández*, C. Mallol*, J. Machado*, A. Sistiaga*, F.J. Molina**, L. Pérez***, R. Afonso*, M.D. Garralda****, N. Mercier*****, J.V. Morales******, A. Sanchís*******, A. Tarriño********, J.A. Gómez*********, A. Rodríguez**********, I. Abreu*, P. Vidal****** El Salt. The Last Neanderthals Of The Alicante Mountains (Alcoy, Spain) Since 1986 archaeological investigations have been conducted at the site of El Salt (Alcoy, Alicante) attempting to deepen the knowledge of its Paleolithic record from an integrated multidisciplinary perspective. The aim is to understand the MIS 3 Neanderthal population in the central region of the Iberian Mediterranean. The research comprises a wide range of studies, including analysis of site formation processes and the application of archeostratigraphic methods in the dissection of archaeological palimpsests (Machado et al., 2011) with the aim of recognizing and describing settlement patterns, activity areas and the role played by environmental factors. This enclave is located in a mountain area which has yielded several Middle Paleolithic sites in caves, rock shelters and open-air surfaces. All of them are between the Middle Pleistocene (Abric Pastor) and the end of MIS 3. Some of these places have provided a broad and diversified archaeological record, suitable to address one of the most controversial issues in current Paleolithic research: Neanderthal disappearance. Despite many studies that have tried to address this phenomenon in Eurasia, macro-regional approaches have not solved the problem, demonstrating the need to promote regional research. In fact, an increasing number of late Middle Paleolithic deposits are reflecting significant variability. In this sense, current research in the Iberian Peninsula offers a complex scenario for the final stage of * ** *** **** ***** ****** ******** ******** ********* ********* the Middle Paleolithic and the early Upper Paleolithic (Zilhão et al., 2006; Jennings et al., 2011; Baena et al., 2012; Maroto et al., 2012). Research by our team in El Salt has recently provided data in conflict with previous models advocating persistence of the Neanderthal population in the Iberian Peninsula. This leads to questioning the role of Iberia as a refuge area for the last Neanderthals groups (Mallol et al., 2012; Wood et al., 2013; Hernández et al., 2013; Garralda et al., i. p. and Galvan et al., i. p.). This paper summarizes the current state of investigations, with particular attention to the time frame, the sedimentary and paleoenvironment framework, and archaeological indicators, in order to recognize and sequence the context of Neanderthal occupations in the Alicante Mountains during MIS 3. The site of El Salt The site is located at the head of the Serpis River, 680 meters above sea level at the confluence of courses and Barchell Polop, two small tributaries. It is now an outdoor locality of about 300 m2situated at the foot of a 38 m-high limestone wall covered with travertine (Fig. 1b). The wall corresponds to a thrust fault of Paleocene limestone over Oligocene conglomerates that led to the installment of a large travertine U.D.I. de Prehistoria, Arqueología e Historia Antigua. Grupo de Investigación Sociedades Cazadoras Recolectoras Paleolíticas. Universidad de La Laguna Dpto. de Prehistoria, Arqueología, Historia Antigua, Filología Griega y F. Latina. Universidad de Alicante. Dpto. de Historia e Hª del Arte, IPHES. Universitat Rovira i Virgili U.D. de Antropología Física. Facultad de Biología. Universidad Complutense de Madrid UMR 5060 CNRS-Université de Bordeaux Dpto de Prehistoria y Arqueología. Universidad de Valencia Museo de Prehistoria de Valencia. S.I.P. Dpto. de Geografía, Prehistoria y Arqueología. Universidad del País Vasco Agencia Estatal de Meteorología. Valladolid. Dpto. de Prehistoria, Historia Antigua y Arqueología. GEPEG. Universidad de Barcelona.
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  380  B. G...
MEDITERRANEAN BASINS. CENTRE. Figure 1. a) Stratigraphic sequence, b) Overview of the site. formation stretching more than 2.5 km at the foothills of Sierra de Mariola. It is a vast travertine construction linked to the Barchell River. The watercourse was blocked by the limestone, forming a lake whose waters would spill over the tall wall in the form of cascades and waterfalls. This paleolake has been identified by the existence of paleolacustrine travertine and limestone sediments resting on Miocene marls in the area above the site, known as “Cases del Salt”. The space occupied by humans at the bottom of the wall was sheltered by a large travertine overhang, which at times of maximum development covered almost the entire excavated surface. The onset of the projection still remains in place and several associated fallen blockshave been found on the ground about 8 and 10 m away from the wall. The blocks are stratigraphically correlated with the archaeological sequence, indicating that the falls occurred throughout the Neanderthals occupation period. The human occupation surface next to the wall was relatively flat, with water from the travertine system possibly flowing downhill at about 13 m from the wall, forming ramps and stepped platforms down to the river. In sum, El Salt exhibits a strategic position amidst various biotopes (plain, mountain, river valley, springs and a lacustrine-marshy environment). It is immersed in a mountainous area rich with a diversity of resources. General description of the archeostratigraphic sequence and chronological framework: The archeosedimentary deposit of El Salt is 6.3 m thick. It was initially studied by P. Fumanal, who distinguished 13 lithostratigraphic units (Fumanal, 1994) (Fig. 1a). Subsequent work led to subdivision of the sequence into 5 segments, according to their macroscopic textural features, 381
MEDITERRANEAN BASINS. CENTRE.  Figure 1. a  Stratigraphic sequence, b  Overview of the site.  formation stretching more th...
382 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD as well as their archaeological content. The description of each segment is presented below, from base to top: 1. Unit XIII: This is a lithochemical layer of unknown thickness (>0.50 m), it is archaeologically sterile, dated at the top by J. Bischoff (J. Geol. Survey, USA) to 81.5 + / – 2.7ky and 80.1 + / – 4 ky (MIS 5a) using the Th / U method. It represents a subhorizontal platformat the foot of the fault and has been exposed in different parts of the site. The archaeological deposit rests on this platform. To date, the only evidence of human activity associated with this unit consists of a small area of the platform exhibiting thermal alteration (subcircular, 0.60 m diameter) at the top of the unit, in direct contact with overlying SU XII. 2. Units XII to IX: This segment has an average thickness of 1.5 m and is characterized by dark brown, loose, fine sands containing abundant combustion residues. These units yielded an accumulation of large travertine blocks, weighing several tons, along with a large number of smaller sized-blocks, interpreted as the first episode of overhang collapse. The accumulation rests directly on the platform (SU XIII). On the other hand, Units SU XII to IX are characterized by the presence of frequent combustion features, many of them have been identified as structures; simple hearths of various sizes (0.20 to 1 m diameter). They are primarily located near the travertine wall. These features are usually associated with rich archaeological assemblages composed by abundant anthropogenic faunal remains, lithic objects and limestone pebbles with marks of use. This segment of the archeosedimentary deposit represents a dense palimpsest of recurrent human occupations. Preliminary results of our integrated studies point to the existence of activity areas, these correspond to the generic model of “hearth-related assemblages” (Vaquero and Pastó, 2001; Dorta et al., 2010).This part of the sequence has been dated by thermoluminescence (TL) to between 60.7 +/- 8.9 and52.3 +/ky (Galvan et al., i. p.). 3. Units VIII to Lower V: The thickness of this segment ranges between 1.5-2.8 m and is composed of silty and sandy clay. It is characterized by a gradual decrease in archaeological evidence. For instance, only seven combustion features were documented, compared to 54 hearths in SU IX and X. There is also a significant decrease in the lithic and faunal record. Another accumulation of large blocks was identified at the top of SU VI, indicating further roof collapse. This segment has bring in TL dates of 52.3 + / – 4.6 ky to 47.2 + / – 4.4 ky. Lower SU V has been dated by OSL to 45.2 + / – 3.4 ky (Galvan et al., i. p.). 4. Upper Unit V: Only 0.50 m of sediment has been preserved, as this unit was partially truncated by an erosive episode of Holocene age. It comprises generally massive sandy silts, capped by a gravelly deposit (the top 0.20 m).This unit is archaeologically sterile except in the mentioned upper coarse part, where two flint blades were found together with several undiagnosed small lithic objects and a small combustion feature. The sterile segment of the unit has been dated by OSL to 44.7 + / – 3.2 ky (Galvan et al., i. p.). 5. Units IV to I: This is 1.3 m thick segment consisting of an accumulation of pebbles and gravel in a silty clayey matrix. This Holocene deposit is in erosional contact with the one above. It contains reworked materials of different time periods; mainly late Upper Paleolithic, Epipaleolithic / Mesolithic and early Neolithic. Human Occupations Older Than 50 ky (Su XII To Ix) In general, human populations in El Salt seem to respond to a pattern of short duration as manifests different temporal indicators, mainly inferred from the spatial distribution of hearth, units of raw material and lithic refitting. Therefore aspects as the presence of clearly delineated accumulations around combustion structures, with little interaction among them are interpreted as inferred from the relations between their respective materials, the incompleteness of the operational chains, quantitative shortages and unidirectional of refit-
382  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  as w...
MEDITERRANEAN BASINS. CENTRE. travertine wall. Hearths were simple (fire made on the ground), large (1 m diameter) and small (0.60 - 0.20 m). They are often found in clusters. Figure 2. Structure of SU combustion Xa. Plan and profile ting. The evidence supports recycling of lithic or the importance of geogenic against anthropogenic sediment inputs. At microstratigraphic scale, there is a low proportion of anthropogenic elements versus geogenic and biogenic soil material (detrital sand, humified organic matter and animal excrements). The position of the thermally altered objects associated with the combustion structures are basic to establish relationships of synchrony or diachrony among the materials constituting the archaeological palimpsests. Field observations and experimental work have enabled us to determine that the black layers of combustion structures from El Salt were produced by the carbonization of organic matter in the soil on which the fires were made (Mallol et al., 2013). The surfaces of these layers correspond to human occupation surfaces. Consequently, they become fundamental archeostratigraphic landmarks for the study of archaeological site formation processes. From a diachronic perspective, the remains of combustion exhibit spatial recurrence. The hearths were preferentially located near the The combustion structures studied (54) (Mallol et al., 2013) show very good states of preservation (Fig. 2). Some were found intact, including a millimeter-thick layer of calcitic ash with a pseudomorphic cellular structure corresponding to pine wood. . Others are dismantled and display structural features that suggest trampling. According to the latter, the phytolith content and lipid analysis data, the black layers of the studied hearths, which range in thickness from 1 to 3 cm, have a low degree of thermal alteration (<400°). Microscopically, they are mainly composed of soil organic matter (SOM; humified leaves, stems, roots, woody tissue, amorphous organic particles, fungi and spores) in a sandy matrix bioturbated by worms and abundant microfaunal bone and coprolite fragments of different animals, as well asseed coats from the fruits of Celtis sp. The representation of flint microflakes, charcoal and bones of meso and macrofauna in the black layers is low and reflects weak anthropogenic impact. The overlying and underlying sediment has a lighter color due to the scarcity of microscopic charred plant remains. The study of travertine thermal alteration, limestone, flint and fauna reveals that they reached temperatures of 700 °C and 800 °C. Nevertheless the midrange for heat modification is around the 450 °C-500 °C, according to the type of hearths documented in these archaeological units. In this part of the sequence, episodes of abandonment in the overall context of the stratumhas been perceived despite its homogeneous appearance; its recognition during the excavation process is highly difficult. This has been made possible by a multianalytical approach based on the study of the horizontal and vertical distribution of the soil by the use of GIS, the micromorphological analysis of archaeological series associationsand sedimentary lipid markers due recognition, as well as the presence / absence of chemical markers of human presence. Integrated treatment of this information has isolated these episodes of abandonment from the identification of small levels concerning 1 and 3 cm thick throughout the intervention area. 383
MEDITERRANEAN BASINS. CENTRE.  travertine wall. Hearths were simple     re made on the ground , large  1 m diameter  and s...
384 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 3. Lithic industry. A) Middle Paleolithic. B) Upper Paleolithic.
384  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
MEDITERRANEAN BASINS. CENTRE. The remains of 50 ky previous occupations reflect, although not exclusively, diversified and preferential use of local resources. The lithic raw materials, hunting registration and identified fuels, show a minimum holding territory that is confined to Alcoy valleys environment, characterized by its high biodiversity and accessibility to supply distances around 10 km. Flints found in El Salt comes mostly from siliceous formations identified in Mesozoic and Cenozoic levels of the Alicante Prebaetic system, formed in marine platform environment (flinttype Mariola, Upper Campanian- Maastrichtian) or reef / toward reef (flint type Serreta-ilerdean or Beniaia, probably ilerdean). The Serreta type is the most widely used not only in this field, but in all places, from the Middle Paleolithic to the Neolithic (Molina et al., 2011). Its acceptance occurred in the Oligocene detrital deposits, where the blocks freed by erosion and resedimented result easier to collect with the technology of Neanderthal groups. During conducted geoarcheological surveys it has been possible to locate capturing areas of this lithological variety with signs of Middle Paleolithic exploitation in Penella. This site is located on the northern slope of La Serreta and Aigüeta Amarga, both within 3 to 5 km radio from El Salt (Molina et al., 2011). Several tools, retouched or not, were added to the site already configured, while some were collected and reused in the settlement itself. Knapping processes have also been identified, usually ascribed in the Levallois methods and linked to the development of subsistence activities (butcher and vegetable processing, as well as leather and wood working in different states) (Rodríguez et al., 2002) and repair actions of cutting edges (Fig. 3a). A comprehensive set of limestone pebbles with different traces of human manipulation (staking, linear impacts, cuts, wear and fractures) interpreted in the clearest cases as hammers, anvils and retouches increases the evidentiary record concerning to highlighted activities. The faunal assemblage performs similar pattern to other deposits of the Iberian Mediterranean lift from MIS 3, characterized by exploitation centered on goats (Capra pyrenaica), deer (Cervus elaphus) and horses (Equus ferus and Equus hydruntinus). Along with these and in a timely manner, we have identified a use of small prey (lagomorphs, Oryctolagus cuniculus).Nonetheless, most of all Leporidae of El Salt relates to the con- tribution of raptors. Similarly, the presence of bovine (Bos primigenius), turtles (Testudo hermanni) and carnivores is timely. Particularly of the latter a unique carnivore dental remaining indeterminate has been recognized, without any evidence of its action on the bone assemblage. Traces of defleshed bone and intense fracturing reflect faunal exploitation in order to exploit all possible nutrients. The discarded after consumption are distributed throughout the excavation area. There is especially an accumulation near the travertine wall, where the size of the fragment also decreases compared to the outer zone. In the same way, thermal alteration is another constant phenomenon in all archaeological faunal linked to heated waste management and the use of residues as fuel. Our experimental studies helped to raise the possibility that some of the material provided in previous occupations is affected unintentionally by different combustion events made after its deposit. Neanderthal occupations after 50 ky This period is represented by the stratigraphic units VIII to lower V and its main feature is the progressive decrease of human impact, compared to the significant increase of geogenic processes in the formation of archaeological sedimentary deposit (Galvan et al., i.p.). In the stratigraphic unit VIII four combustion structures have been identified and only three in the SU VII, despite widely excavated area (40 m2). The lower incidence of hearths across this deposit stretch entails a coarser distribution of the materials, without recognizing clear accumulation as described for the oldest sedimentary units. The faunal and lithic record, especially the limestone pebbles with traces of human manipulation also show a gradual downsizing of resources (Galvan et al., i.p.). At the roof of SU VI a second episode of falling blocks from the big rock shelter overhang is recognized, which marks a turning point not only in the sediment dynamics, also significantly limited since anthropogenic contribution. From this unit, the only evidence of fire is represented by anthracological remains (less abundant than in the previous segment). The lithic, faunal remains, limestone, and travertine also present thermal alteration marks. This reinforces the idea of a gradual 385
MEDITERRANEAN BASINS. CENTRE.  The remains of 50 ky previous occupations re   ect, although not exclusively, diversi   ed ...
386 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD and steady process of decline in the Neanderthal footprint at the site among 52.3 + / – 4.6 ky BP and 45.2 + / – 3.4 ky BP, coinciding with the end of Heinrich 5event. The upper section of this sequence displays sedimentological traces indicative of an abrupt change. Indeed, the SU V is lithological different from the rest of the deposit. It is configured as a massive sand calcite layer, very fine, fresh and wellsorted to fossilize blocks and mega blocks from the collapse originated during the second episode of roof collapse. This unit must be combined with a high sedimentation rate, given the fresh state of the sand calcite derived from the breakdown of rocky bed and massive structure that presents an assemblage. According to the chronostratigraphic framework provided by TL dates, this process may be hypo- thetically associated with severe aridification described for the central region of the Iberian Mediterranean during the H5 (Sánchez-Goñi and Harrison 2010).A dry or semi-dry climate context is consistent with the interruption of travertine and limestone bedrock. Both would be subject to a recurring solution and precipitation under humidity conditions. Added to this is the identification of authigenic gypsum crystals in SU VI and V as evidence pointing to the mentioned arid or semi-arid conditions. In this framework is ascribed the finding of six dental remains(I1 right, P3 right, P4 right, M1 right, M2 right and third molar remains, also right) located at the base of the SU V lower in a proper Middle Paleolithic archaeological context (Fig. 4). Its morphometric study has attributed a Neanderthal right hemimaxilla to a young individual (Garralda et al., i.p.). The chronology of these fossils falls be- Figure 4. Neanderthals human remains. Plan showing the location of the teeth remains (SU V lower).
386  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  and ...
MEDITERRANEAN BASINS. CENTRE. tween 47.2 + / – 4.4ky (base of SU V lower) and 45.2 + / – 3.4 ky (roof SU V lower). This suggests that they represent some of the last Neanderthals in the region. The Upper Su V: The Disappearance Of Human Occupation Evidence From The Middle Paleolithic This unit has been identified only in the upper zone of the site, along the travertine wall. Its thickness exceeds 0.60 m in some places. These are subhorizontal beds and fine silts and very fine sands with massive structure and diffuse contacts. Specifically some facies of coarse sand and gravel (1-3 mm), with small blocks of limestone (10 cm) scarce and isolated within the thin deposit of this unithave been documented. The entire assemblage goes in the slope direction and out of the wallpackage. Its contacts to wall and roof are fuzzy. The OSL dating of the sediment is 44.7 + / -3.2 ky. This deposit has been excavated in an area of about 30m2, and throughout its development any evidence of human occupation has not been located. At the roof of this unit, in diffuse contact, continues a deposit of about 0.30 m thick and a denser heterometric texture, consisting of silt, coarse sand and gravel. It incorporates lots of fragments and small blocks of limestone and travertine platelets. This deposit thins towards the wall and nozzle in the course of the slope, but its components do not show a preferred orientation. It was excavated in an area of 10 m2 where it has been able to recover a small lithic assemblage composed by two laminar and some technologically undifferentiated flakes, next to small structure of combustion associated to abundant and dispersed anthracological material (about 700 units) under study. These few elements, sparsely diagnostics from the technological and typological point of view, constitute the first evidence of human presence after the discontinuity described above. In neighboring Cova Beneito (Muro de Alcoy, Alicante) Middle Paleolithic sequence also crown a thick sterile deposit, similar to that referred to in the El Salt on which subsequent stratigraphic units assigned to early Upper Paleolithicare developed. This pattern of discontinuity has been described in other Iberian sites (Mallol et al., 2012), suggesting the existence of a depopulation of southern peninsular territory after the Neanderthals’ disappearance and before the first signs of the presence of anatomically modern humans groups. El Salt subsequent to Middle Paleolithic El Salt archeosedimentary sequence ends with a powerful erosive deposit where, 37 lithostratigraphic facies evidencing the succession of different erosion processes have been recognized. The first, very energetic and fast, hit part of the Pleistocene sedimentation. Nonetheless, subsequent events of variable energy, partially removed the gravel deposit itself, as the channel morphology show in the different layers that comprise it. The clasticfraction of this deposit filled the karst system of El Salt with sediment. Its output abroad took place after the reactivation of the water circulation. This led to the opening of a cavity in the upper third of the travertine wall and expulsion from their sediment content. The lithic record provides no doubts about the presence in El Salt of Magdalenian materials, Epipaleolithic / Mesolithic in the sequence (Fig. 3b). Between laminar and microlaminar cores there are diagnostic examples by standardization in the blank obtaining; thus as by the use of soft hammer and pressure flaking. A number of macroutensils are related to macrolithic denticulate facies typical of Mesolithic between the IX and early VIII millennium BP. Bipolar cores and ecaillées from the Gravettian and the Solutrean of Hispanic Mediterranean are also present in this data. The most problematic issue is to attribute specific materials to the early stages of the Upper Paleolithic, especially consideringits current redefinition after the controversy between the Aurignacian and the Gravettian (De la Peña, 2013) and the recent detection of anearly Upper Paleolithic in determinate, well described in Cova Gran (Martinez et al., 2012). The record seems to reveal a ceramic(continuous or discontinuous) occupation that can be bound between the Neolithic IA and the Bell-Beaker horizon transition (5500-2500 BP). The existence of fragments from large containers, especially am- 387
MEDITERRANEAN BASINS. CENTRE.  tween 47.2         4.4ky  base of SU V lower  and 45.2         3.4 ky  roof SU V lower . Th...
388 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD phora-like indicate El Salt could have some habitat stability. The transition of Bell-Beaker pottery is providing the terminus post quem for the deposition of gravels, being newer material enclosed therein. Acknowledgments This article is part of the research undertaken in the project I + D + I HAR2012-32703: The dis- Valentín Villaverde*, Didac Román**, Rafael Martínez-Valle***, E. Badal*, P.M. Guillem***, M. Pérez-Ripoll*, M.M. Bergadà****, C. Real*, M. Borao* Thanks to Camilo Visedo Moltó Archaeological Museum, to the Excmo. Ayuntamiento de Alcoy and the General Direction of Cultural Heritage of the Generalitat Valenciana the support to archaeological research. Cova de les Cendres Cova de les Cendres opens to the sea at the cliffs of Punta de Moraira, 60m above sea level. It is a cavity of a certain extent, consisting of an outdoor space with a high vault presenting lots of cenital detachment blocks, particularly on its south wall, and another indoor space reached through a narrowing between the vault and the sedimentary filling of the surface whose Figure 1. Cova de les Cendres. The cave´s view. * ** *** **** appearance of Neanderthals groups in the central region of the Iberian Mediterranean. A methodological proposal for approaching the historical process and paleoenvironmental framework. (MINECO-FEDER). final part is partially loaded by the sedimentation (Fig. 1). The archaeological excavation performed in this archaeological site have always been carried out in the inner part, reaching a surface of about 600 m2. H. Breuil visited and identified the place as an archaeological site in 1913. However, the first archaological activities were carried out by E. Llobregat in 1974 and 1975 obtaining an important archaeological sequence especially relevant concerning the Neolithic. From the revision of the materials by J. Bernabeu and the identification of a harpoon and several pieces of Paleolithic typology, the possibility of the archaeological site to conserve Pleistocene chronology levels emerged, which is why a survey was practiced in 1981. These activities confirmed, with new materials, the importance of the Magdalenian levels and demonstrated the interest of the archaeological work. After some surveys in later years, the archaeological activities focused on Pleistocene levels have continued from 1995 until now. Archaeological activities of Pleistocene levels started with a stratigraphic survey in squares A-17 Departament de Prehistòria i Arqueologia. Universitat de Valencia. Departament de Prehistòria i Arqueología. Universitat de Valencia. TRACES UMR-5608. UNiversité de Toulouse-Le-Mirail. IVACOR. Generalitat Valenciana SERP. Universitat de Barcelona.
388  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  phor...
MEDITERRANEAN BASINS. CENTRE. and B-17, which has continued up to now (Villaverde et al., 1999)advancing the filling features in the two sectors in which the excavation was further developed (Villaverde et al., 2010).In sector A, in the area in which the excavation of Neolithic levels was made, the extension covers between 10m2 and 12m2, depending on the levels. All the same, the packages corresponding to the Upper Magdalenian have been studied while the archaeological excavation of the Lower and Medium Magdalenian was taking place. In sector B, the archaeological work area has fluctuated between 7m2 and 9m2 and, although part of the Magdalenian levels was conserved at some points, the storage pits and the Holocene erosions contribute to the information being primarily limited to the Solutrean and the Gravettian. Pleistocene stratigraphic sequence Two complete datings have been obtained, one in squarE-13, with a result of 12,740 ± 100 BP, and other in square A-17 with a result of 13,320 ± 170 BP, in this case through traditional C14 and several charcoals, so it is much less reliable. Level X is preserved in a few areas, as has been pointed out, as a consequence of the Holocene erosions, the Neolithic silos, or the erosive process previous to the deposition of level IX. The structure is lamninated and sterile in archaeological terms. The formation seems to be a consequence of flood processes of the cavity, with deposition of fine sediment stored in calm conditions, probably carried by soft streams from the highest areas of the cavity. It is important to point out that the stratigraphy presents a very strong dip line from the entrance to the back, promoting this type of processes in high-humidity stages. The Paleolithic stratigraphic sequence of Cova de Cendres contains a total of 9 levels including, from the roof to the wall, Magdalenian, Solutrean and Gravettian occupations. These packages have a thickness of 3 meters, without reaching the filling basis (Fig. 2). The most recent Pleistocene level is in an erosive contact with the first Neolithic level. Therefore, the stratigraphy presents a sedimentary hiatus from the Epigmagdalenian to the geometric Mesolithic. The Holocene sequence goes from level I to VII with a thickness of 3.5 m and includes a rich sequence of different periods of the Neolithic, the Transition of Bell Beaker Culture and the Bronze Age. Level VIII remains only in some areas, as a great deal of the Neolithic contact with the Pleistocene sequence is preceded by erosive processes. In addition, during the ancient Neolithic, silos were practiced in some areas of the cavity affecting different Magdalenian levels depending on the importance of the previous Holocene erosion. Where itis preserved it has a thickness of 5-7 cm , a red brown color and a highlimestone, cobble, sand, gravels, fraction of little size with the presence of small slabs. Its cultural adhesion is uncertain as the archaeological activity was always done on a very small surface. Level IX has an unequal thickness reaching 3040cmin some areas. The structure is massive and includes a few medium to large-sized blocks. It has erosive contact with regard to the underlying unit that sometimes remained totally dismantled. It encloses materials of the Upper-final Magdalenian. Figure 2. Cova de les Cendres. Right-sagittal stratigraphic profile of square A-17. Levels XI to XVIA. 389
MEDITERRANEAN BASINS. CENTRE.  and B-17, which has continued up to now  Villaverde et al., 1999 advancing the    lling fea...
390 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Level XI is light-brown in color and includes plenty of little and medium-sized limestone cobbles and gravels of bu-angular to angular morpohology, with a variable thickness of between 10cm and 25cm. It coincides with an intense settlement stage and many charcoals and bone remains are included in its matrix. It presents a laminated structure enclosing several consecutive combustion structures in the sector where it was excavated, protected by the peripheral position of several large-sized blocks, certainly favoring the preservation. The package has been dated on different squares. They correspond for sure to this stage the following dating: 13,220 ± 50 BP, 13,120 ± 60 BP, 13,280 ±50 BP and 13,350 ±50 BP, all of them obtained via the AMS system from a single charcoal. The materials correspond to the Upper Magdalenian with many harpoons and other pieces of antler and bone industry. Level XII is in an erosive contact with the former and includes two sub-levels. The higher one –A–, with a thickness of about 10cm, is grayish brown in color made by clayey silty sands enclosing limestone pebbles; the lower one –B–, with a thickness of about 24 cm, it is also in erosive contact with the former. Its structure is laminated with a clayey silty texture, a variable color and a presence of limestone fraction of a varying size on its base. The industry corresponds to the Medium and Lower Magdalenian. Its excavation has only been carried out in a survey of a small area but different dating have been obtained via AMS system confirming this attribution: 13,400 ± 50 BP, 13,690 ± 120 BP, 14,510 ± 50 BP, 14,850 ± 100 BP, 15,630 ± 60 BP and 16,030 ± 60 BP. Level XIII is also in erosive contact with the former; it´s grayish brown color and laminated structure with a variable morphology and thickness in lateral terms, encloses different combustion structures and pulsations of rich coarse limestone, medium and small-sized. Its thickness reaches 3540cm with a marked dip towards the back of the cavity. The dating obtained via AMS are numerous and confirm its attribution to the Evolved Solutrean with the following results: 16,790 ± 60 BP, 17,210 ± 60 BP, 17,230 ± 130 BP, 18,750 ± 130 BP and 18,920 ± 180 BP. Level XIV is blade-structured and in erosive contact with the former, alternating orange and blackish lands, as a consequence of the human settlement processes. The coarse material is moderated, with a small size. The thickness presents important lateral variations, with dislocation phenomena and erosive scars which seem to have a post-depositional origin that is difficult to evaluate, as in some areas it is not preserved. A dating corresponding to the level, as it has been removed directly from the profile, has provided a result of 20,200 ± 80 BP. It is therefore the Middle Solutrean, although its industrial mark is conditioned by the narrowness of the excavated surface and the difficulty of the underlying level splitting. In addition, the identification on the archaeological works is difficult. Level XV presents similar conditions to the former, with a lower thickness and it is also very affected by post-depositional movements. The contact with the XIV is normal but erosive with the XVI. The color is grayish and has a clayey silt sedimentary component. Two datings place it between the 20,800 ± 110 BP and the 21,230 ± 80 BP, indicating a settlement stage of the Lower Solutrean. Level XVI constitutes at the moment the basis of the archaeological sequence. The thickness is high and to descriptive effects it has been divided in three. The sub-level XVI-A reaches 10-15cm of thickness. It is light brown color and with a massive structure; sub-level XVI-B is of laminated structure, with a loamy component and marked black and light beige variations; its contact with the underlying sub-level is in line with a pulsation of a greater proportion of medium-sized limestone and cobbles; and sub-level XVI-C is also laminated structure, with very pronounced color alternations, loamy and the base seems to be in line with a larger presence of coarse fraction and a block of a certain size. It has only been excavated in a small area, so its features should be specified in greater detail in the future. The thickness reaches 20cm. The three sub-levels have provided materials of a clear adhesion to the Gravettian and are well dated, with the following results: 21,880 ± 100 BP, 23,350 ± 100 BP, 23,860 ± 100 BP, 23,920 ± 100 BP, 24,080 ± 150 BP, 24,240 ± 220 BP, 25,600 ± 140 BP and 25,850 ± 260 BP. Features of the industries Lithic industry of the Upper Magdalenian at Cova de les Cendres is characterized by microblade group predominance –62,4%–. Burins –7,9%–, scrapers –7%– and edge-retouched pieces –6,7%– also reach certain importance, but with values far from the micro-blade group. The rest of the typological groups are below 4% and the presence of chipped pieces is hardly mentioned.
390  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Leve...
MEDITERRANEAN BASINS. CENTRE. In the micro-blade group, backsides are thin and direct in almost 90% of the cases, and little blades pointed trough the retouch are the 19.3% of the group and 12% of the total retouched. We could add to these pieces a large deal of little bladelets presenting a natural pointing, with just one retouched edge that does not reach the distal extreme –9.6% of the micro-laminar group and a 6% of the total retouched–. Furthermore, bladelets with reverse retouches reach a high percentage -17.4% of the micro-laminar group and a 1.9% of the total– some of them being pointed. Another fact to be pointed out is the small size of the micro-laminar tooling, a large amount of pieces being classified as very microlithic A, 15% of the 140 retouched little blades –types 84 to 91– are under 10mm, and 65% are not larger than 15mm, showing us the smallest micro-blade tooling of the Upper Paleolithic and the Regional Mesolithic. As for bone industry, its richness must be pointed out, both for the tooling and manufacturing wastes. Among the recovered pieces, the large amount of harpoons is especially interesting (Fig. 3), many of them were recovered in the levels corresponding to this period and some pieces are out of context or integrated into Neolithic levels as a consequence of the silos made during this stage. In total, there are 20 pieces representing one of the most important sets of all of the Iberian-Mediterranean coast. Harpoons offer an important variability in typometrical and morphological terms: the lengths are between 73 mm and 163 mm, 12 pieces are made of horn and 8 are made of bone, the oval section –10 models–, quadrangular –5–, flat –3– and round –1–, and the barbs number is between 2 and 12. On the other hand, the bases do not present variations with respect to the shaft, showing only in some cases simple or double extremity bevel. Only some lateral swelling can be pointed out, previous to the first barb. Finally, except for one case, the barb do not exceed the base width, giving them quite a different appearance from the Cantabrian Magdalenian harpoons (Román and Villaverde, 2012). Most of the harpoons correspond to level XI –11 models–, another 3 correspond to level IX and the rest do not have a defined stratigraphic context. The rest of the bone industry is composed by batons, needles, points of single or double bevel, some points à base raccourcie and many broken point fragments. The ornament is also plentiful, the most part through the drilling of gastropods, mainly Theodoxus fluviatilis, and through bivalve´s shells and atrophied deer´s canine teeth. Figure 3. Cova de les Cendres. Upper Magdalenian bone industry. As the Medium and Upper Magdalenian records in Cova de les Cendres are limited to the survey’s information, the existent one is restricted. The lithic industry is dominated again by the micro-laminar group, but with a greater share of scrapers over burins. In addition, there is an average percentage of substrate pieces. Inside the micro-laminar group, the presence of bladelets with inverse and direct retouches and the existence of truncated pieces of backside stand out, allowing a distinction from the higher levels. The presence of points of single bevel and batons deserve to be highlighted, some of them with curvilinear and angular decoration (Villaverde et al., 2010 and 2012). In the current state of the investigation, only the fact that inside level XII it is possible to establish a micro-laminar stage between the Medium Magdalenian and Upper Solutrean can be pointed out; the distinction of the Badegoulianin Parpalló is significant in regards to the Lower Magdalenian. In any case, it is necessary to wait for the results of the current archaeological activities to precise this part of the sequence in more detail. 391
MEDITERRANEAN BASINS. CENTRE.  In the micro-blade group, backsides are thin and direct in almost 90  of the cases, and lit...
392 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD From this moment and in sector B, the sequence of Cova de les Cendres presents really complex sedimentary conditions determined by the identification of many erosive processes in the form of fairways and buckets, level thickness, and inclination differences as a consequence of the presence of a stalagmite flow projected in the left sagittal part of sector B causing a dipping of the packages opposite to the sector´s general trend, and a high level of bioturbation. All these circumstances make it very difficult to establish, during the excavation process, the stratigraphy detail. This problem affects levels XIII, XIV and XV. These reasons explain why the information related to Solutrean is evaluated on global terms, without specifying the features of the diachrony that the assessment of the obtained information suggests or the assessment of the attribution of industry in levels XIV and XV I as just an attempt partly conditioned by the absolute chronology and some quality pieces. In level XIII, related to Evolved Solutrean, the recovered material exceeds the 5,200 lithic remains, although the number of splinters is very high –64.8%–. While flakes are the main support, the negative dorsal mainly bladed and the balance between full production flakes and the beginning of exploitation and preservation show that the operational chain is intended for the procurement of laminar supports. However, this circumstance is not contrary to the fact that 45% of the flakes have been retouched mainly for the manufacturing of scrapers, burins and pieces with one or both edges retouched. Laminar flakes reach much more measured values, less than blades, but their presence is significant and probably linked to a small raw material, conditioning the length of the obtained supports. The restructuring by retouch is high, especially in the lower half of the package where they involve 39.4% of the supports. A high percentage of the supports correspond to the full production stage –70%–. Strictly blade supports reach a value slightly higher than laminar flakes, especially in the lower half of the package. The consistency of the blade production is significant in quantitative terms: the 56.4% of the pieces are blade and the 87.2% collecting the pieces over laminar flakes and bladelets. Most part of pieces correspond to the full production stage –about the 80%–. As in the former supports, the sizes are small and the selection of elongated supports correspond very precisely to the manufacturing of shouldered points. Micro-laminar supports have very similar quantifications to the blades, and they mainly correspond to the full production stage –70/80%This part of the production is essentially linked to the retouched little blades and shouldered points. The cores are well recorded, with a great presence of the ones made on nodule, dominating the exploited in one or two opposite directions. As for the retouched material, totally296 pieces are collected adding another 73 with retouches of use. The Solutrean set brings 12 laurel leaves points which are present along the whole package, 15 shouldered points, some of them broken, and 7 pointes à face plane. Altogether this represents 11.4% of the retouched material. The Solutrean flat-retouch is limited to 20 pieces, 6.4%. Among the most classic Solutrean pieces, it´s worth noting the presence of a pedunculate point of bifacial retouch, covering just the dorsal face. The parallel of this piece in Parpalló and Ambrosio show that we are not at the end of the evolved Solutrean. Other significant pieces are a willow leaf, a unifacial point similar to the Badegoule type, and an elongated and narrow laurel-leaf point with thin denticulate edges. Shouldered points present an important size variation. They are produced from high-quality laminar supports; the co-existence of a large format, with lengths between 4 and 6 cm and widths between 1 and 1.2cm, and other small format, between 2 and 3cm long and .6 and 0.8cm wide, can be seen. The rest of the industry is characterized by a scraper index –15.5– higher than the burin index –9.1–, a large number of pieces with one or two edges retouched –29.4– as well as chipped pieces –10.1– and a considerable presence of micro-blade tooling –7.4– especially in the higher part of the level. Finally, one of the most interesting novelties of this package is the richness of the bone industry (Fig. 4) and the ornament industry. The number of perforated mollusks comes to 70 pieces, with 14 different recorded species, although the Theodoxus fluviatilis concentrates the highest proportion of the specimens –42.8%–. As for the bone and antler industry, we have 19 pieces with a predominance of double antler or bone points –6 specimens–, a rounded basis point, three of polygonal basis, a flat point, two points a single bevel, two perforated needles, two double thin points, one punch and an unclassifiable fragment.
392  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  From...
MEDITERRANEAN BASINS. CENTRE. The information obtained in Cova de les Cendres suggests the inclusion of the level XIII industry into the Solutreo-Gravettian (Villaverde et al., 2010). With regard to levels XIV and XV, it is possible that the presence of some pointes à face plane could indicate the existence of a Middle or Lower Solutrean stage, although with the information currently available it is not worthwhile to go forward with this subject. The industrial description of the Gravettian in Cova de les Cendres has been made from the information provided by the survey that is currently being extended through the excavation of sector B (Villaverde and Román, 2004). Although it is possible to observe some differences from the sequential point of view in the carving processes, especially between sub-level XVI-A and the two remaining, XVI-B and XVI-C, and more particularly in the importance of the blade carving, the limited nature of the set –2,088pieces– does not give much importance to this face. We will just comment on the higher proportion of flakes in sub-level XVI-A and a larger presence of microblade supports in sub-levels XVI-B and XVI-C. The laminar supports of the three sub-levels present very similar alteration by retouching percentages: 40-45% of the blades and 35-40% of the bladelets. Blades are perfectly positioned between Figure 4. Cova de les Cendres. Bone Industry of the Evolved Solutrean. 20 and 35mm, reaching a maximum of 45mm, while bladelets are between 13 and 32 mm. The retouched material presents a predominance of the micro-blade group –26.5%–, ahead of the backed group –19.6%–, amongst them Gravette and micro-Gravette points can be highlighted. The presence of Cendres-type points is significant, made from a slight modification of blades with a certain format and a natural direction. Pieces with retouched edges and chipped pieces have a good presence, reaching 11.8% of the retouched. It´s also worthwhile highlighting that burins beat the scrapers, the flat ones and those with flint knapping on truncated pieces standing out. As for the bone industry, it´s worth while highlighting its richness, although the typology is focused on double points of round-basis or polygonal points. Among the ornaments we find two species of marine gastropods (Fossarus ambiguus and Littorina obtusata), several Dentalium sp., and two deer´s atrophied canine teeth, one perforated and other with a groove for the suspension. Fauna: economic and paleoenvironmental implications Fauna of Magdalenian levels at Cova de les Cendres (Martínez Valle, 1996) presents a high percentage of rabbit remains. This fact, common in most part of the Upper Paleolithic archaeological sequence and the whole area of the Iberian Mediterranean, is particularly important at the moment of establishing models of economic behavior of the human groups at that time. The evaluation of the recovered materials related to a level XI combustion structure at Cendres is significant in this regard. 95% of the remains correspond to this specimen and the most part of the long bones present cut marks from humans. It´s very possible, as almost every bone has chipping marks -jaws, ribs, vertebrae, scapula, pelvis, all the long bones and even some metatarsus- that these marks are related to meat procurement for its preservation through drying techniques (Villaverde and Madrínez Valle, 1995). Considering the rest of the recorded species, deer is the best represented with a domain of remains belonging to infant, young and sub-adult individuals. The goat is also present, in a much smaller remains number, as well as the horse, and among the carnivores, the lynx, the fox and the wildcat. The enhancement of lynx remains, whose consumption can 393
MEDITERRANEAN BASINS. CENTRE.  The information obtained in Cova de les Cendres suggests the inclusion of the level XIII in...
394 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD be seen from the marks, is a feature that will have continuity along the whole sequence. Finally, the presence of some seal remains must be pointed out. In level XI, Microtus arvalis denotes the development of fresh weather conditions (Guillem, 1996). These same features emerge from the study of the bird fauna, where the record of Pyrrhocorax graculus, considered as a clear sign of fresh weather conditions, is a highlight. During the Solutrean, and compared to the former levels, a decrease in the contribution of bone remains to the archaeological site seems to take place, and an increase of the marks made by carnivores and nocturnal birds is recorded. The Solutrean sample analyzed is composed of 6,953 bone remains where the rabbit prevails just as in the former levels –Figure 4–. The macro-mammals fauna is dominated by the red deer followed by the goat and, in smaller percentages the horse, the auroch, and some carnivores. Red deer hunting is focused on the adult individuals, good size males predominating, and goats and horses, where adults also prevail. The richness of carnivores´ remains is surprising, especially those of the lynx, whose remains present again torn marks and fractures similar to those observed on herbivores bones indicating human consumption. The information provided by the Gravettian levels at Cova de les Cendres is especially relevant because it constitutes one of the contributions to this period at Mediterranean level. First of all, the importance of the hunting and the consumption of small prey in the Gravettian period of the Mediterranean area is determined, particularly the Orictolagus cuniculus (Pérez Ripoll, 2004).Their percentage values are similar to the ones of the rest of the Upper Paleolithic sequence. Secondly, the information of Cova de les Cendres offers a higher diversity of the consumed fauna as the most significant novelty with regard to the known information in other archaeological sites of a similar chronology. Thus, as in other sites a clear predominance of red deer and goat remains can be seen, with total values overtaking 80% of the recognized macro-fauna, in the levels of Cova de les Cendres the predominance of these two species is nuanced, as other species like the horse, the auroch and the wild boar reach a certain presence. This information may indicate both a less detailed model than in later periods and a different occupation rhythm of the cavity, with sojourns of shorter duration. Finally, in view of the modest sample analyzed, it´s difficult to determine the ages of the hunted prey, except for indicating that the hunting of sub-adult individuals in Cendres was occasional, with a higher incidence in young and adult individuals. The Gravettian sequence of Cova de les Cendres has not provided any micro-mammal taxon of very cold weather as the environmental requirements of the analyzed species are included in the bio-climatic thermo-Mediterranean and supraMediterranean levels (Tormo, 2010). The best represented specie in every Gravettian sub-level is Apodemus sylvaticus, followed by Terricola duodecimcostatus and Microtus cabrerae. On the basis of sub-level XVI-C a domain of Apodemus sylvaticusis observed, and together with Eliomys quercinus would indicate the presence of Mediterranean deciduous forests and scrubland areas. The high percentage of microtinos which require on its biotope open spaces plenty of vegetation and a certain degree of edaphic humidity is also important. In Central Gravettian, an increase of humidity from the remarkable presence of microtinosis is seen, above all due to the emergence of Arvicola sapidus. The lack of Crocidura russula also confirms the existence of wet weather conditions. In the higher part of sub-level XVI-A a decrease of humidity seems to be recorded due to the disappearance of Arvicola sapidus and Terricola duodecimcostatus. The increasing of arboreal stratum of the black pine and the undergrowth, to the detriment of open spaces, interpreted in the antracological study, is consistent with the recorded species in this level. Antracological data evaluation In Cova de les Cendres during the Gravettianan absolute predominance of Mediterranean pine remains of cold ecology is recorded, such as the Pinus nigra together with junipers and shrubs like woody fabaceae, labiates, etc. The most thermopile plants like the Aleppo pine –Pinus halepensisor the Quercus, both deciduous and evergreens, have few remains. The current distribution of these pine species and their ecologic features allow an interpretation about the environmental conditions at the end of MIS3. If we consider the altitude and the latitude where the archaeological sites of the
394  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  be s...
MEDITERRANEAN BASINS. CENTRE. eastern coast are located, we can postulate that the Pinus nigra would be on the lower limit of its distribution area, so the average annual temperature could lie between 14ºC and 12ºC at best, and the rainfalls about 400-600 mm in view of the absence of hydrophilic plants. Also, Mediterranean pines are perfectly adapted to the summer drought because they store a large quantity of water in the trunk. During the Solutrean a similar flora to the Gravettian is recorded, but the taxa proportions change. The charcoals show an open landscape ruled by junipers and woody fabaceae, although black pines are still the dominant trees. The pine forest reduction may indicate a decrease in precipitations, opposite to the situation in Cova de Valentín Villaverde* Finally, during the Magdalenian period the vegetation used by human groups is still the one of the black pine forests which during the Late Glacial were recovered at the expense of the junipers and the fabaceae scrubs. It is worth noting the continuous curve of evergreen Quercus, the presence of some deciduous Q. and sporadically, some Aleppo pine (Pinus halepensis). This group of plants could be the prelude of a gradual weather improvement of the Late Glacial (Villaverde et al., 2010). Cova de les Malladetes Location and background Cova de les Malladetes –Malladetes cave– is in the Mondúver massif, in the Barx municipality. This small cave measuring approx. 135 m2 has three NW-facing apertures: the main mouth, another tiny aperture and a small East facing aperture. Its high location, 600 m asl, provides commanding views of the Barx polje and the Valldigna depression. The first archaeological work conducted at the site dates back to the period between 1946 and 1949, when excavations by the Prehistoric Research Service were led by Ll. Pericot and F. Jordá. A substantial area was dug, divided into 19 sectors covering more than two-thirds of the potential space. The base of the infill, 4.90-5.50 m below the surface, was reached in the central sections (D, E and F). Less progress was made in others, only reaching 3.15 m in sectors A and B, 3.40 m in sector I and even less depths in the others. The material retrieved from these digs has only * les Cendres regarding to the continuous curve of evergreen Quercus, so it is difficult to choose one of the two options in view of the high tolerance of this flora that is really well-adapted to the Mediterranean droughts. been studied partially. F. J. Fortea’s Ph.D thesis included an analysis of the packages at the top of the sequence detected in the first test pit excavated in 1946 and sectors F, G and I (Fortea, 1973). This study assessed the Epigravetian and its relationship to the Neolithic. More recently, P. de la Peña (2013) did the same with research focused on the Gravetian. Her study, also part of a Ph.D. thesis, included the 1946 test pit, sectors E, D, F, G, H and I, and the East pit excavated by F.J. Fortea and F. Jordá in 1970 (Fortea and Jordá, 1976). I. Davidson’s 1980 Ph.D. thesis included the study of the fauna from the same test pit (Davidson, 1989). The East test pit, located in the sectors corresponding to sectors L and M, along with another in the western part of the cave, adjacent to sector C, formed the second stage of the Malladetes excavations. In this case, the main results were published in 1976 by the authors of the excavation. These two test pits facilitated the sedimentological definition of the infill, a palynological study and a pal- Departament de Prehistòria i Arqueologia. Universitat de València. 395
MEDITERRANEAN BASINS. CENTRE.  eastern coast are located, we can postulate that the Pinus nigra would be on the lower limi...
396 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD aeoclimatic interpretation of the sequence by M.P. Fumanal (Fumanal, 1988) and M. Dupré (Dupré, 1988). Major stratigraphic and palaeoenvironmental features (1970 dig) (Fig1.) The East test pit shows a stratigraphic evolution that was decapitated at the top when three layers in this sector were excavated in 1949. The detailed description is compulsory reading as this site, scarcely 2 km from Parpalló, has a clear correspondence with the infill of the latter site and permits its sequence to be compared with the other earlier location, with less stratigraphic detail of the infill (Fortea and Jordá, 1976). Level XIV, the base of the infill, is 30-35 cm deep. Little anthropogenic evidence is limited. This level has a massive sedimentary structure, a fine fraction of clayey-silt sand and clear signs of gelifraction. It has a very pale brown colour. The site was formed under intense cold, with an input of autochthonous material under fluctuating humidity. The pollen record shows scarce arboreal vegetation, corroborating this description. Level XIII corresponds to more moderate and moister conditions, with heavy runoff. There is little anthropogenic evidence on this level as well. The vegetation shows the presence of a few thermophilic taxa. Level XII shows a clear decrease in the coarse fraction, in response to runoff erosion facilitated by winter rainfall. The shallow depth might indicate a deceleration of the sedimentary buildup. This level has been dated at 29,690 ± 560 BP. Anthropogenic material remains scarce. There is more arboreal vegetation, primarily pines. Level XI presents abundant evidence of gelifraction. It is approx. 25 cm deep, also with little evidence of human presence. The vegetation is indicative of low humidity. Level X shows a change from the previous pattern. It is also roughly 25 cm deep. The coarse fraction has angled edges which indicate a phase of frequent but not intense cold, under semiarid conditions. It is associated with the start of a more intense human occupation. The lack of arboreal taxa corroborates the harsh climate. Level IX, 20 cm deep, marks a slight impoverishment of the occupation. The coarse fraction still has sharp edges and the sand fraction is larger. It shows signs of intense cold, seasonal rain and little vegetation cover. Level VIII, 15 cm deep, has a large coarse fraction, with strong signs of occupation in the form of ash and burned material. It is similar to level XI, with less intense freezing in winter. The cold conditions are reflected in the vegetation, with a decrease in Oleaceae and the arboreal fraction. Level VII is shallow (approx. 5 cm) with a massive structure and a moderate coarse fraction. This level marks a crisis in the occupation. The formation indicates slightly damp and moderately cold conditions, or an infill of the gaps in the coarse fraction by the fine fraction from the level above. In Level VI, 15 cm deep, the coarse fraction is not particularly large, with a calcium carbonate crust and a rounded appearance. There is a new intensification of its occupation, dated at 21,710 ± 650 BP. This level seems to reflect cold conditions, less intense than the previous level, with gentle runoff. There is a larger proportion of arboreal pollen. Figure 1. Malladetes cave. Stratigraphic profile of eastern test pit, 1970. (Fortea and Jordá 1976). Level V, 15 cm deep, dark grey-brown, has been dated at 20,140 ± 460 BP. The coarse frac-
396  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  aeoc...
MEDITERRANEAN BASINS. CENTRE. tion has a highly evolved aspect with a secondary limestone layer. Its features suggest a more moderate climate with a better rainfall distribution. Level IV corresponds to a uniform thin (5cm) brown layer with many angular platelets, indicative of gelifraction and a harsh, dry climate. It is archaeologically sterile. No pollen from thermophilic plants has been detected. Level III is 25 cm deep. It has a massive structure, a brown colour and a moderate presence of coarse fraction. Moisture conditions seem to have favoured the vegetation cover which slowed runoff. It has been dated at 16,300 ± 1,500 BP. The vegetation reflects these conditions, with a large proportion of arboreal vegetation. A comparison between Level XI, linked by Fortea to the previous levels, and the adjacent sectors must include some of the backed items. As suggested by P. de la Peña, this could mark the start of the Gravettian sequence at the site. The period spanning level X to level VIII contains a rich lithic assemblage with technical and typological features that can be correlated with the Gravettian. Specific features include the presence of backed items with a relatively good representation of Gravettes and microgravettes or bladelets as well as the blade and microblade knapping technique with unipolar and bipolar cores. Splintered items, interpreted in some cases as cores used to remove splinters, are also present. Level II is 20 cm deep. It has a brown colour, asmaller percentage of coarse fraction and no aggregate in the fine fraction. It suggests fresh conditions with little moisture. The vegetation cover is stable with the presence of hygrophylous taxa. Level I, with a partially disturbed upper section, was not studied sedimentologically by M.P. Fumanal. Industrial sequence and cultural features The interpretation of the Malladetes infill is primarily based on the analysis of the stone industry found in test pits excavated in 1970 by Fortea and Jordá. This work and subsequent papers mentioning the site (Aura et al., 2006, Villaverde, 2013) also considered data from the excavations in 1940s. However, the difficulty of correlating some sectors with others and certain inconsistencies in the sequential distribution of materials, especially in the Solutrean sequence and higher levels, makes this correlation purely indicative. The start of the Malladetes sequence, as shown in levels XII-XIV, yielded a poor lithic record in both the 1940’s excavations and the 1970’s test pits, with the exception of a small batch of bone items (Fig. 2) whose typology –three broad-based spear tips made from antlers and two bone awls– brings the Evolved Aurignacian to mind. The dating for level XII is generally consistent with this designation, as is the lack of Gravettian-type items. In this period there was little human presence at the site but no lack of carnivore material. Figure 2. Malladetes cave. Broad based Aurignacian points. 397
MEDITERRANEAN BASINS. CENTRE.  tion has a highly evolved aspect with a secondary limestone layer. Its features suggest a m...
398 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD characteristic narrowing effect found in the early phases of the Mediterranean artistic sequence. Although its correlation with the 1970’s stratigraphy is difficult, it can definitely be attributed to the Gravetian on the basis of the associated material. Secondly, layer 12 of Sector E yielded important material including a human infant occipital, reconstructed from numerous fragments (Fig 4). This human fossil is associated with several charcoal rests and deer antlers, all in a small recess in the wall forming the boundary of the sector. A sample of the charcoal, collected and preserved in association with the remains, was identified as Pinus nigra and dated at 25,210 ± 1120 BP. This is one of the few remains from this chronology located on the Iberian Peninsula, with traits of anatomically modern human populations (Arsuaga et al., 2012). Figure 3. Malladetes cave. Traced uro engraved on Gravettian plaquette (layer 13, sector D). This layer was the source of two major discoveries in the 1940’s excavations. Firstly, the only figurative engraving found at the site (Fig. 3) was found in layer 13 of sector D. One face of the limestone plaquette shows an incomplete bovid, with straight horns and a clearly narrowing mouth, engraved with simple, shallow lines. The rump articulated with the foreleg creates a The period corresponding to levels VI-III is separated from the previous section by a poorly defined level detected in the 1970 excavation, and is difficult to correlate with the 1940’s material. Nevertheless, it shows a series of characteristic features found in Iberian Solutrean facies. The first Solutrean presence at this site, identified by a pointe à face plane, seems to have been during the Lower Solutrean. The dating obtained for this level is not inconsistent, although there is a large degree of indetermination. Bifacial retouch –without the characteristic features of the evolved Solutrean– permits the attribution of this part of the sequence to the Middle Solutrean. Sterile level IV may cor- Figure 4. Malladetes cave. Gravettian infant occipital (layer 12, Sector E).
398  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  char...
MEDITERRANEAN BASINS. CENTRE. respond to the start of the Upper Solutrean at the Parpalló site, while levels III and II show the typical typological evolution in the SolutreanGravettian. Level I is difficult to classify as it contains Solutrean-Gravettian material but, as indicated above, was partially stripped and also contained mixtures of material. C. Olària* A Reference model for the Upper Magdalenian of the Mediterranean flank o the Iberian Peninsula. Cova Matutano (Vilafamés, Castellón, Spain) Summary: The Matutano Cave completes our perspective of the upper-late Magdalenian on the Mediterranean flank of the Iberian Peninsula, evolving in four occupation stages with absolute chronology, habitation structures and a remarkable representation of portable art in all stages. Keywords: Magdalenian Epimagdalenian, portable art. Site and location This site is located within a town, Vilafamés (Castellón), 40º06’50” north and 3°38’ east of the Madrid meridian, on the National Mapping Institute’s “Vilafamés” Map 616” (E/1:50.000). The cave is on the western slope of a mountain, Tossal de la Font, at the end of the Les Altures de les Contestes range composed of Dogger Jurassic limestones which left a karst system with intricate underground networks. Matutano Cave is 351 m asl, roughly 6 km inland from the coast. Interestingly, this cave is at the rear of a house which was once an oil mill entered from the northwest. This construction disturbed the lip and the upper levels of the cave, which was used as a pen. Its 105m² are distributed in a single space (Fig 1). Opposite the site is a flat area, Pla de Vilafamés, an endorheic basin dating from the Pleisto-Holo* Considerations in recent years suggest that the top of the Malladetes sequence may well include not only Holocene material but also Magdalenian or Epimagdalenian levels, as suggested by some of the items from this phase. More detailed analysis is required for further clarification. cene and defined as an authentic polje, which currently holds water like a lagoon. Occupation phases Matutano IV The first occupation of the cave, probably seasonal, from spring to early autumn, was during the initial Upper Magdalenian. Its absolute chronology is (I-11312): 13.960 ± 200 BP– (UGRA-225): 13.370 ± 260 BP. The climate during this period was cold and wet. This phase includes Levels 5 and 6 of Sector 1 and Levels 6 and 7 of Sector 2. The open landscape included herbaceous vegetation alternating with small conifer forests. The identified living structures include small sub-circular hearths, post holes and cists. The most numerous fauna was rabbit, followed by a considerably smaller number of hare and deer. Remains of goat, horse, wildcat and wolf were also found. There was a low presence of hedgehogs. Bird remains are from partridge, hazel grouse, bustard, chough, dove, little owl and blackbird. The malacofauna includes marine taxa– Pecten jacobeus, Crastoderma, Turritella, Cerithium and parts of Cardiidae. No terrestrial taxa were found. The fish fauna remains are probably mullet. The lithic industry in this phase was clearly predominated by burins, along with the presence of large flakes used to obtain endscrapers, a remarkable presence of denticulates and Univeristat Jaume I. Laboratori d’Arqueologia prehistòrica. olaria@uji.es 399
MEDITERRANEAN BASINS. CENTRE.  respond to the start of the Upper Solutrean at the Parpall   site, while levels III and II ...
400 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. Outside view next to oil mill, and fracture lip. Photo Jordi Mestre. sidescrapers and a significant use of simple retouch along with retouch with burin. No abrupt retouch has been found in the analyzed lithic assemblage. The bone industry is characterised by the presence of harpoons and spears on deer antler. The presence of perforated needles has been detected in this phase. Tool decorations are linear and geometric votes. The manufacturing technique on the bone objects was found to be by rotation. The majority of the tools are classified as borers. There is a surprising amount of traces of red ochre on this bone industry. Other stone tools include hammers, retouchers, smoothers and sheets of red sandstone. The only adornments are Pecten jacobeus and Cerastoderma edulee pendants, bored by rotation. There is also a remarkable presence of portable art, superficially engraved or “graffitied” pebbles bearing figurative representations of zoomorphic horse heads and unidentified hindquarters along with linear strokes, all with abundant remains of ochre. Matutano III The second occupation, quite possibly seasonal –spring and autumn–, is from the early-full Upper Magdalenian. Its absolute chronology is between (UGRA-208): 13,220 ± 270 BP and (UGRA201): 12,460 ± 180 BP, a milder, moist climatic interphase. This phase was identified on Levels 4 and 5 of Sector 2 and Levels 5 and 6 of Sector 3 (Fig. 2). In this meadow landscape, there was a significant increase in coniferous forests and a moderate presence of Evergreen oak. The identified occupation structures include hearth floors, holes for posts or props and aligned stones which formed low partitions leading out from the cave wall, used as dumps for ash and other waste. Here as well there was an intense presence of rabbit and a smaller proportion of hare, while the presence of other species is higher. There was an increase in the number of hunted partridges, followed by rock pigeon. Amongst the malacological remains, there was a significant collection of Pecten followed by Glycimeris. Terrestrial taxa were scarce, with the exception of Sector 3 where there was a notable presence of Iberus. Unidentified vertebrae of marine fish fauna have also been found. The lithic industry continued the former traditions, with a predominance of burins. The base material was smaller, with a tendency to use blades. There was a significant presence of simple retouch and burin
400  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
MEDITERRANEAN BASINS. CENTRE. includes the red-legged partridge, rock partridge, bustard, a significant presence of raptors such as the Lammegier vulture, imperial eagle and golden eagle, and also native species such as the rock pigeon and the red billed chough. The mollusc fauna consisted of marine species Pecten and Naticarius, while the only inland species detected on Level 3 was Iberus, while vertebrae of unidentified fish species fauna were found. In the lithic industry, there was a considerable increase of abrupt blades, backed bladelets, abrupt retouches, backed points and undifferentiated abrupts. The bone industry included spears, harpoons, burnishers, perforated items and grooved decorations. The stone industry featured the presence of retouchers in Sector 3, sandstone slabs and few hammerstones. The portable art included figures representing horse and bovid heads and hindquarters, as well as drawn lines, while the adornments were still Glycimeris and Pecten jacobeus pendants with traces of ochre. Matutano I Figure 2. Stratigraphy of section A from Sector 3. retouch. Abrupt retouch began on backed blades. In the bone industry, the use of deer antlers was reduced and replaced by bone, and there was also a decline in harpoons. The stone industry yielded quite similar materials to Phase IV, retouchers, smoothers, hammerstones and sandstone slabs. Zoomorphs, particularly fawns and deer hindquarters, were still etched on pebbles. Ornaments were limited to shells such as Glycimeris with remnants of ochre, perforated by rotation and abrasion. The fourth occupation of this site, considered to be annual, with an intense hunting season between March and October. The absolute dates are (UGRA244): 12,520 ± 350 BP, (UGRA-241): 11,590 ± 150 BP and (UGRA-243): 11,410 ± 610 BP, from the end of the Upper Magdalenian or Epimagdalenian. The climate was warm and moist, and the landscape had a greater tree cover along with rangelands and scrub, alternating with grassland in open areas. This Phase corresponds to the surface level and Level 1 of Sector 1, Levels 1, 2 and 3 of Sector 2, and the surface level and Matutano II This level is the third occupation of the cave, probably seasonal (March to October). This Phase corresponds to an evolved Upper Magdalenian. Its absolute dating is (I-11326):12,390 ± 190 BP. The climate was moderately warm and moist, with a considerable increase in tree cover. Domestic structures correspond to medium sized hearths, in some cases adjacent with floors, as well as post holes for props and cists. The diversity of fauna species was greater, particularly notable in increased deer hunting and a significant presence of hedgehog. The bird fauna Figure 3. Pebble from Level 4, Sector 2, perforated and engraved with a doe or fawn. 401
MEDITERRANEAN BASINS. CENTRE.  includes the red-legged partridge, rock partridge, bustard, a signi   cant presence of rapt...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 402 Levels 1 and 2 of Sector 3. The habitation structures were identified by large hearths and low walls used as partitions and spatial divisions. The fauna remains include a broad species diversity and specialized deer hunting. The bird fauna is predominantly partridge followed by great bustard and pheasant, and just one crow. Marine molluscs were primarily Pecten, Glycimeris and Cerastoderma, while terrestrial molluscs were Iberus Pseudo fachea and remains of Rumina. The fish fauna vertebrae might be from Mugil sp. The lithic industry includes a significant number of abrupts, an abundance of abrupt retouches and few burins, with a wide diversity of typological groups. The bone industry included spears and punches on antlers, although the majority were manufactured on bone, particularly in the case of domestic tools: burnishers and wedges. The stone industry showed an increase in the number of hammerstones, retouchers, smoothers J. Emili Aura Tortosa* Valentín Villaverde Bonilla** The publication of this text coincides with the centenary of the authorization by the la Junta Superior de Excavaciones y Antigüedades (Real Orden de 17-10-1914) to excavate at the Parpalló and Calaveres sites requested by H. Breuil. Unfortunately the global situation prevented him from undertaking the work. A year earlier, Breuil dug a shallow test pit, following the notes drafted by J. Vilanova i Piera, and yielded the first decorated plaquette with engravings. In 1927, Luis Pericot García joined the University of Valencia after a brief period in Paris, coinciding with the establishment of the Servei d’Investigació Prehistòrica of Valencia. He promoted and undertook three seasons of excavations between 1929 and 1931 in conjunction with the service’s first director, I. Ballester Tormo. * Conclusions Cova Matutano is an interesting settlement from the end of the Upper Palaeolithic. It completes our knowledge of this period on the Mediterranean coast of Iberia (Olària, 1999) and the overall development of the Upper Magdalenian at Mediterranean sites such as Tossal de la Roca (Cacho, Jordá et al.,,i.p., 2001) and Cendres (Villaverde, Martínez, et al., i.p., 1999), amongst others. Its interesting range of portable art provides added interest to this chrono-cultural framework from the initial Upper Magdalenian to the Epimagdalenian (Olària, 2008). Cova del Parpalló (Gandía, Valencia) 1. Presentation ** and sandstone slabs. Portable art on pebbles included linear motifs and schematic zoomorphs (Fig. 3), while adornmets were made from Glycimeris and Detalium. The first results published by Pericot confirm the importance of Parpalló, a site that had already received some attention since the last 30 years of the 19th century. In the first stage, a deep sequence of 7-9 metres was organized into two large episodes (Magdalenian and pre-Magdalenian levels), noting some issues to be taken up in subsequent work. This was southern Europe’s first documented a Palaeolithic Art painted and engraved plaquettes and identified a complete evolution of the Solutrean and Magdalenian. This initial information suggested connections with the FrancoCantabrian region, with Levantine Art and with North Africa (Pericot, 1942). Breuil was very interested in having a reference site in the Mediterranean region of the Iberian Peninsula, as reflected in his numerous references to Pericot’s work at Parpalló included in the 1937 reissue of his study of “Les subdivisions du Paléolith- jeaura@uv.es Dept. de Prehistòria i Arqueologia. Universitat de València. valentin.villaverde@uv.es. Dept. de Prehistòria i Arqueologia. Universitat de València.
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  402  Leve...
MEDITERRANEAN BASINS. CENTRE. ique supérieur et leur signification”. Iberia’s role as a bridge between Europe and Africa was already a central topic of discussion in wich any new data becoming argument for the dissemination processes, its routes and its scope. In this context, Parpalló joined the restricted list of sites used in the first half of the 20th century as a basis for organizing the Upper Palaeolithic sequence in Europe. A comprehensive monograph on the work in Parpalló was finally published in 1942, coinciding with the global conflict. The World War once again delayed the impact of the results and gave rise to certain misgivings about the sequence and evolution of Parpalló, which have been clarified by various reviews of the material which began in recent decades and are still underway. on the cave walls itself. Work by J. M. Arias, E. Portell and A. Velasco, supervised by V. Villaverde, began the process of updating the documentary material and reappraising the site’s artistic sequence. We should also mention the related study of part of the ornamental material (B. Soler), the study of human remains (J.L. Arsuaga, I. Martínez) and colouring materials (Roldán et al., 2013). 2. Stratigraphy and chronology Cova del Parpalló –Parpalló Cave– lies on the southern flank of Montdúber, 440 m asl and Most of the authors who have worked with the material found at Parpalló coincide on two outstanding qualities: its ongoing validity due to its ability to open up new perspectives, and the cumulative nature of results which need to be checked with new sequences. The last 30 years of the 20th century marked a turning point in our understanding of this site. Palaeoeconomic studies and the first radiocarbon datings by I. Davidson coincided with a review of the lithic industries at Parpalló and its regional context by J.M ª. Fullola Pericot. These results gave rise to an initial profile of the economic system and the regional settlement in the Mediterranean area (Davidson, 1989), while an outline of regional relations with the Palaeolithic in the Cantabrian area and southern France was based on a sequence ranked into three major technological complexes: Gravettian, Solutrean and Magdalenian (Fullola, 1979). Archaeological seriation, primarily based on typological data, new perspectives in palaeoeconomics, radiocarbon chronology and the regional relationships, steered several revisions of the Gravettian (J. L. Miralles), the final evolution of the Solutrean (V. Villaverde, J. L. Peña and Mª. J. Rodrigo) and the Magdalenian (J. E. Aura). More recently, a revision of the Upper Solutrean by from a technological perspective (Tiffagom, 2006) has begun to extend to the first Magdalenian indications in the Badegoulien facies. Another outstanding contribution has been a technological and functional analysis of Magdalenians endscrapers (Jardon, 2000). Changes in material culture have often been linked to evolutionary processes in graphic expressions on decorated plaquettes and more recently, Figure 1. Cova del Parpalló. Stratigraphy of the Talud sector, reconstructed from graphics by L. Pericot (SIP archive), showing levels, radiocarbon datings and main archaeological divisions 403
MEDITERRANEAN BASINS. CENTRE.  ique sup  rieur et leur signi   cation   . Iberia   s role as a bridge between Europe and A...
404 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD roughly 9 km from the current coastline. The only documented stratigraphy worth a tentative reconstruction is the “Talud”. This sector was excavated during the final season (1931) with the stratigraphic section exposed. In conjunction with the data from the previous digs, this led to two changes to the previously used excavation procedures, as the layers were no longer 25 cm thick as mentioned in the monograph, and their profile was irregular, giving rise to different orientations and slope angles (Fig. 1). Analysis of the photographic series held at the Prehistory Museum of Valencia identified three possible discordant contacts in the Talud stratigraphy, including changes in the dip of the levels, a decrease in the coarser fraction from ceiling to wall, and the detection of cross laminations in the central package. Although this is a limited reconstruction, it is of undeniable interest, as it permits the description of four major episodes that match the main archaeo-stratigraphic divisions (Aura, 1995). The description of these four episodes from wall to ceiling is as follows: I Massive, horizontal disposition with large angular blocks at the base and whitish nodules. Contact with the following section is erosive and corresponds to layers 29 to 27, which encompass material from the Gravettian and the early Solutrean. II Graphic documentation shows that this is the most complex section of the stratigraphy. It is a succession of layers whose lenticular arrangement and orientation are reminiscent of an encrusted basin geometry. The coarse fraction is smaller than the section below and contact with the section above is brusque. It corresponds to layers 26 to 1112, with Solutrean materials. III Fines, lighter in colour and cobbles scattered more or less horizontally. Coincides with layers 11-12 to 6 and includes Badegoulien material. IV Massive sedimentation and scattered cobbles. The top of the section is obscured by what seems to be mudflow. Corresponds to layers 5 to 1, with Magdalenian materials. These layers are presumed to match palimpsests of human occupations, whose rates of accumulation match what is known from other sites. This aspect is also worthy of analysis in the regional settlement context. The density of the anthropogenic material is uneven, with clear differences between the above-mentioned sections. In the Talud sector, a rate of 10 retouched tools per m3 of sediment is only surpassed after the full Solutrean, with more variable ratios during the Upper Solutrean (8 to 88 retouched objects per m3), and up to 260 items in the Badegoulien occupations and 350 retouched tools during the Magdalenian. Recently, it has been able to identify a paleontological context in a gallery wall located archaeological sequence in which quite complete remains of a dhole are associated with several skeletal parts of Capra pyrenaica bearing bite marks and fractures. Unfortunately, this context could not be dated and we do not know its chronological position with respect to the early Gravettian occupations. The series of radiocarbon datings by I. Davidson using collagen samples of bone and antler reappraised the age of the regional solutreanization process and a more southerly dating for a Magdalenian assemblage. New datings for Talud, in this case bone pre-treated with ultrafitration, ages the previous results by 0.5 to 1 ky (Fig. 1). 3. Archaeological sequence and techno-economic transformations The four sedimentary sections and densities are discussed for the Talud section match the technocomplexes described for Parpalló: a Gravetien level at the wall, followed by a thick Solutrean deposit, and a Magdalenian layers in which has identified an initial block related to the Badegoulien. a) The first human occupations correspond to groups of anatomically modern humans who, on the basis of documentation from the nearby Malladetes site (Villaverde et al., 2010), are presumed to have manufactured Gravettian tools. These occupations are below 7.25 m and have a low density of anthropogenic material, which contrasts with the quality of the raw material employed in regular, well-sized blade products to manufacture a number of endscrapers, Gravettes, microgravettes and backed bladelets. The occasional blade produced with a hard hammerstone, with convergent preparation found in the final stages of the Gravettian. The bone industry is limited to awls and double points with a rounded base. The position of the Gravettian and early Solutrean at the Parpalló site, its radiocarbon chronol-
404  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  roug...
MEDITERRANEAN BASINS. CENTRE. ogy and the relationships between the two technocomplexes are open to conjecture, delimiting debate about a critical episode in which new technical traditions were formed. b) Heat treatment of flint, pressure retouch, bifaciality and tanged points are the most characteristic features of the Solutrean (Fig. 2). Its evolution at the Parpalló site and also at Malladetes has been used as a basis to build a chronological model for the Iberian Solutrean. – A Lower Solutrean level was defined between 7.25 and 6.25 m, characterized by flat faced points and the disappearance of tools manufactured with abrupt retouch. The human occupations still have a low density of material and a radiocarbon chronology amongst the oldest Solutrean dates. – The Middle Solutrean is situated between 6.25 and 5.25 m. This period documents heat treatment and flat pressure retouch applied Figure 2. Cova del Parpalló. Solutrean lithic and bone industry. to bifacial lanceolates points; primarily laurelleaf blades with a convex base and asymmetrical items, although flat faced points have not disappeared completely at this stage. Faceted Solutrean and laminar knapping are the basic techniques, with an increase in bone tools, still focused on double points with a rounded base and the first items with a polygonal base, mostly in bone. – The Upper or evolved Solutrean coincides with one of the most complex sections of the Parpalló sequence. Pericot identified an evolutionary process between 5.25 m and 4 m, summarized in the replacement of the barbed and tanged points produced using Solutrean retouch by shouldered points with abrupt retouch, the incorporation of bone industry parallel to the disappearance of stone points. This section has been revised in several papers, all of coincide in the proposal of a sequence detailed in several horizons: 405
MEDITERRANEAN BASINS. CENTRE.  ogy and the relationships between the two technocomplexes are open to conjecture, delimitin...
406 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD • Upper or evolved Solutrean I (5.25 to 4.75 m) in which Solutrean reduction, heat treatment and pressure retouch techniques are joined by a new morphotype: barbed and tanged points; • An evolved Solutrean II or SolutreanGravetian I is described above this general level (4.75 to 4.50 m), in which there is a substantial decline in Solutrean features parallel to an increase in blade production used to shape shouldered points. • An evolved Solutrean III or SolutreanGravetian II (4.50 to 4.25 m), in which the above features are refined. • Finally, a Solutrean-Gravetien III or final level (4.25 to 3.50 m) is added to the evolved Solutrean III, at the expense of the first Magdalenian level described by Pericot. In this section there is a substantial drop in shouldered points and a significant increase in antler simple bevel points. This seriation has sought to describe the transformation of the Solutrean at the Parpalló site, although its impact undoubtedly spread beyond the regional sphere (Aura and Jorda Pardo, 2013; Villaverde et al., 1998). Moreover, the possible incongruence between a long-term Solutrean evolution and the presence of a bone industry –a characteristic Magdalenian feature– can only be resolved by analyzing its incorporation in the context of acculturation-relationship processes between groups which were already fully Magdalenian to the north and still Solutrean further south. When added to this sequence, the first radiocarbon datings lent weight to the perspective that in Parpalló –and also in southern Iberia as a whole–, the Solutrean remained persistent, which directly affects the analytical framework of the early Magdalenian technocomplex. The Solutrean-Magdalenian “transition” (Aura et al., 2012) is another critical point in the Parpalló sequence. As with the early Solutrean, there was a significant drop in the number of sequences that record this period, possibly caused by erosion linked to the Lascaux interstadial. The assessment of the impact of these processes on the archaeostratigraphy and radiocarbon chronology of the earliest Magdalenian is currently under review. Pericot identified the first four stages of the French Magdalenian in Parpalló on the basis of changes in bone points, focusing on the shape of the bases and also decorative motifs (Fig. 3). However, a revision of the lithic and bone industries from the first 3.50 m of the sequence has identified Badegoulien traits in the early Magdalenian lithic industries (part of Pericot’s Magdalenian II and III), confirmed by technical, typological and decorative convergences with the French Badegoulien bone industries. The Parpalló-Talud lithic industries have facilitated the description of two major episodes with different productions and typologies. c) An early Magdalenian on Badegoulien facies has been described above level 3,50 m on Talud layer 11, in fact a “Parpalló type” of Badegoulien, characterized by lithic production aimed at short, broad flakes and blades which were used in turn to produce tools bearing evidence of continuous sharpening. Endscrapers, sidescrapers, notch-denticulates and raclette are the most common tools. Bladelets and microflakes are scarce in Parpalló due to the recovery techniques that were used, however there are several types of cores (on carenated endscrapers, burin-cores and others) which point to smallsized microflake production. The typological and technomorphological evolution of this Badegoulien at the Parpalló site has been divided into two phases: Badegoulien A (Talud layers 11 through 9) and Badegoulien B, (layers 8-6). Bone industry, primarily using deer antler, show a significant increase and a degree of continuity with the end of the Solutrean. The presence of spears with a long, often concave single bevelled edge, double tips and rounded bases, and some flat tips and rods are the most common types. There are two significant incised decorations: broad bevelled herringbone decorations of the Le Placard type, and a second group of motifs that was largely concentrated into the Badegoulien B, consisting of arms, angles and zigs-zag compositions. The datings obtained with prior ultrafiltration applied to bone samples situate start of the Badegoulien phase in Talud at 18.5 ky BP. This result has triggered debate about the above-mentioned Solutrean persistence model, the scope of which extends beyond the site itself. d) The last episode described for the Talud sector corresponds to the Magdalenian blade-
406  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD      ...
MEDITERRANEAN BASINS. CENTRE. Figure 3. Cova del Parpalló. Magdalenian lithic and bone industry. microblade knapping industries. These productions are the basic component of industries which included large percentages of microliths tools and a significant presence of burins. The only dating for Pericot’s Magdalenian III (section 1.70-1.50m) is around 14 ky BP, nearly 2000 years after the current datings for Cendres’ Lower Magdalenien. The bone industry in this block is also manufactured from antler material that was prepared using grooving technique. Points and rods are the predominant groups, with an increased variation of base material (pointed, rounded, single and double bevelled edges) and sections, primarily angular (square-rectangular and triangular) in larger numbers than round sections in some layers. This toolkit, widely distributed in other sites, included barbed points after 14 – 13.5 ky BP. The Parpalló sequence probably continued into historic periods, as material attributed to the Mes- olithic, Neolithic and even Roman times has been found here (Aura, 1995). 4. Human remains In addition to the material in the Vilanova i Piera collection, which is probably from the top levels or the upper galleries and thus has an uncertain chronology, Pericot’s excavations also discovered scattered human remains in the Magdalenian levels consisting of at least 3 individuals and one burial in the central-east sector which corresponds to the deep levels of the Solutrean, although logically, the burial could have been undertaken from the upper levels. It is unclear whether the remains were in a grave or covered with piles of stones and earth. The published description only mentions the discovery of a skull and a humerus fragment, possibly from 407
MEDITERRANEAN BASINS. CENTRE.  Figure 3. Cova del Parpall  . Magdalenian lithic and bone industry.  microblade knapping in...
408 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD the same individual. The identification of stones around the body, extended to slabs and remnants of a home that could coat the skull, are the only known elements on their context. However, the material found between layers 6.50-6.25m included two tibia whose biometric dimensions suggest that they may be from the young female buried in Parpalló. 5. Parietal and portable art: the importance of the sequence The collection of portable art in Parpalló is one of the reference points for European Palaeolithic art. The site has contributed more than 5000 decorated plaquettes which span the entire Palaeolithic sequence (Villaverde, 1994). This is particularly important bearing in mind the scarcity of pre-Magdalenian figurative references elsewhere in Iberia. The collection is enriched with bone items bearing Solutrean and Magdalenian figurative representations, along with the documented rock art on the north wall of the main chamber in the area bounded to the west by the Talud sector. Many of the decorated plaquettes were painted, with and without engraved, which heightened the importance of the collection. Diachronically, the first plaquettes are from the Gravettian. This is a set of items with figurative representations. In some cases their cultural designation is dubious, while the position of two items in the sequence leaves no doubt about their dating, as they show similar conventions to an item with an identical chronology in the neighbouring Cova de Les Malladetes, thus confirming this chronology for the first known figurative portable art in the Spanish Mediterranean region. Analysis of the technical and stylistic components of these items in comparison with those at the start of the Solutrean sequence has found a high degree of congruence throughout this long period. This coherence includes the conventions used to draw limbs and heads, body proportions and the articulation of the different parts of the depicted animals. The legs are drawn in three ways on these levels: arched legs, legs drawn with three strokes and legs drawn with two parallel lines which diverge at the extremities. Within the diversity of the zoomorphs in this phase, the heads show solutions with a greater chrono- logical precision: disproportionately small goat heads with long curved horns, combined with details of the ears; a trilinear convention in the depiction of female deer heads, and a falling termination in equine noses. Finally, a narrowing body in the contact area between the belly and the forelimbs, associated with body predominated by juxtaposed anatomical parts and scant attention to naturalistic articulations, and a larger proportion of straight biangular perspectives in depictions of the extremities, as well as massiveness and a tendency for disproportion are all characteristic traits of this early phase of preMagdalenian art. The techniques include the use of double engraved line and a well documented use of paint, including a variety of colours and combinations with engrave. Red, obtained from iron oxide, and black, both organic and mineral, in this case manganese oxide, have been detected in a recent study of the pigments from these periods (Roldán et al., 2013). The above-mentioned parietal engraved horse can be linked to this phase on the basis of its stylistic features and its position in the sedimentary infill. The association with the barbed sign helps to confirm the “unity” of the early pre-Magdalenian graphic phenomenon in South-western Europe. The differences of these plaquettes from the rest of the Solutrean sequence, allow the possibility of organizing the pre-Magdalenian art, distinguishing between an old and a more recent phase, associated with the most advanced period of the Middle Solutrean and the evolved Solutrean. This proposal is based on the quantitative entity of the 2481 Solutrean platelets found at the Parpalló site, which includes 386 zoomorphs, well distributed between the earlier phase and the period corresponding to the more recent pre-Magdalenian art. The fact that 64 painted zoomorphs (95.5% of zoomorphs in this technique) correspond to pre-Magdalenian art is also indicative of the importance of the art from this period in the Parpalló sequence (Fig. 4). The distinctive traits include the abandonment of the biangular perspective of legs, a greater attention to the articulation of the body parts, and with regard to technique, multiple and repeated strokes. Some of the representations of head and legs not only serve to profile the features of this phase, but also to draw parallels with the scope of
408  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  the ...
MEDITERRANEAN BASINS. CENTRE. Figure 4. Cova del Parpalló. 1: pre-Magdalenian portable art; 2: Magdalenian portable art. SE France: a triangular start to feet, prolonged in a linear stroke, and goat heads that are open at the top, with horn insertions drawn with two strokes in this space. Mention should also be made of the importance of scenes and animations in both pre-Magdalenian phases, including the well-known examples of a fawn drinking from a doe, a doe with two associated fawns and a lynx lunging at the neck of a ibex. The lynx and the two possible dholes found on Middle Solutrean platelets are amongst the few exceptions to an iconography heavily influenced by the importance of deer, ibex, aurochs and horses. Rectangles, cross-hatching and strip lines are signs that became frequent in the recent preMagdalenian. The Magdalenian cycle in Parpalló began with a quite impoverished phase with abundant figures drawn according to somewhat archaic conventions, and soon gave way to a type of art in which the most characteristic feature is its lack of con- nection to prevailing contemporary trends in the Cantabrian and Pyrenean areas. There are few features that permit the figures from these stages to be distinguished from the more recent pre-Magdalenian, although some aspects do show noticeable differences and nuances in the proportions, the articulation of the anatomical parts, a greater attention to internal details and a more detailed execution of the morphology. This change affected the depiction of deer ears, replacing the trilinear convention with V-shaped ears or, in some Upper Magdalenian items, with naturalistic ears, in some cases executed from a single angle perspective. The same details are seen in several figures which either have an absolute profile or a more naturalistic style including details of auroch horns, deer antlers, eyes, mouths and sex, as well as more attention to the correct perspective. There are also elements with a clearly pictographic component, seen most clearly in the depiction of a young animal. The decline of the use of paint in 409
MEDITERRANEAN BASINS. CENTRE.  Figure 4. Cova del Parpall  . 1  pre-Magdalenian portable art  2  Magdalenian portable art....
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 410 the zoomorphs was accompanied by the appearance of composite lines and a rich succession of highly complex variations of signs, some with precise indications of their chronology, such as crosshatching with internal divisions. Wild boar, found on a plaquette showing three individuals, a partridge, a duck, an otter and several canids are particularly interesting species as they all Michael Walker*†, Mariano LópezMartínez**†, María Haber-Uriarte***† © All rights reserved. ** *** † † Recent discoveries of Magdalenian portable art in the Mediterranean area has shown a tendency towards stylization at the end of the regional artistic cycle, which seems to have prevailed over a more proportioned, detailed component in the Upper Magdalenian. Sima de las Palomas del Cabezo Gordo (Torre Pacheco, Murcia, Spain) This vertical cave system was formed by karst solution of a Triassic marble hill overlooking the Mediterranean. In 1991 a spelaeologist descending the 18-m deep entrance shaft (Fig. 1) extracted a fossil (SP1) which, once cemented breccia was removed, comprised Neanderthal maxillae connected to the mandible, with almost all their adult teeth. Subsequent systematic excavation (Walker et al., 2012a) uncovered 3 undisturbed Neanderthal partial skeletons (SP96, SP92, and a child, SP97) with several skeletal parts in anatomical position (including cranio-mandibular articulation, femoro-pelvic articulation, elbow, rib-cage, vertebral column, shoulder girdle, foot bones, etc.), Mousterian Palaeolithic artifacts, and animal bones (some charred), all lying deeply in a cemented rock tumble within the upper part of * appear in the Parpalló collection of Magdalenian art but are generally rare in Palaeolithic art. an 18-m deep wall of brecciated sediments that was left exposed by miners after they took out most of the shaft’s sedimentary fill ca. 1900. SP96 and SP97 have crania and mandibles, unlike SP92. Because SP1 had lain near SP92 it might be SP92’s head. Excavation uncovered SP96 lying with elbows flexed and hands touching the forehead. Computer-assisted tomography revealed hand-bones in breccia adhering to the forehead of the SP97 child (Walker et al., 2012b) which lay underneath SP96, perhaps its parent. The position of the upper extremities implies intentional arrangement before rigor mortis had set in; it is recorded at some other Mousterian sites (Defleur, 1993). Individual articulated skeletons are very important indeed because they enable far better precise and accurate estimates of body size and Departamento de Zoología y Antropología Física, Facultad de Biología, Universidad de Murcia, Campus Universitario de Espinardo Edificio 20, 30100 Murcia, España. Correo electrónico: mjwalke@gmail.com Tfnº: 34-620-267104 Calle Pintor Joaquín 10-4º-I, 30009 Murcia, España. Correo electrónico: marianolopez@hotmail.com Tfnº: 34-630-408806 Departamento de Prehistoria, Arqueología, Historia Antigua, Historia Medieval y Ciencias y Técnicas Historiográficas, Facultad de Letras, Universidad de Murcia, Campus Universitario de La Merced, Calle Santo Cristo 1, 30001 Murcia, España. . Correo electrónico: mariahaber@pi-ma.es Tfnº: 34-629-756183 Directors of the excavation, Murcian Association for the Study of Palaeoanthropology and the Quaternary, MUPANTQUAT web-site http:www.mupantquat.com (Murcia Archaeological Museum, Avenida Alfonso X El Sabio 7, 30008 Murcia, Spain), all correspondence to MUPANTQUAT Secretary M.López Martínez Murcia University Experimental Sciences Research Group E005-11 “Quaternary Palaeoecology, Palaeoanthropology and Technology” (c/o Dr. J. S. Carrión García, Department of Plant Biology, Biology Faculty, Murcia University, Campus Universitario de Espinardo Edificio 20, 30100 Murcia, Spain)
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  410  the ...
MEDITERRANEAN BASINS. CENTRE. Figure 1. Neanderthal skeleton SP 96 (“Paloma”). proportions than estimates derived from pooled statistical analyses of a given bone type (e.g. femur, tibia, or humerus) from an assemblage of bones from different people who may well not be identifiable individually in it. SP96 was a short Neanderthal woman (”Paloma”) <20 years old, with the typically female wide greater sciatic notch of the pelvic basin which underwent post mortem distortion (that “virtual” reconstruction will correct from tomographs). The skeleton is about 85% complete which permits precise and accurate morphological measurements for estimating body proportions (e.g. crural index, etc.) which are undoubtedly Neanderthal and robust (“hyperpolar”). Remarkably, its stature is one of the shortest known for Neanderthal adults (Walker et al., 2011a). The child skeleton SP97 lay underneath “Paloma” (her child?). Also short in stature, SP92 was probably <25 years old (Walker et al., 2011b). Short, too, was the owner of SP77, a femoral head from looser sediment that had banked up against cemented rock tumble and contained scattered Neanderthal remains, including mandibular fragments of a baby, a child, and an adolescent female (Walker et al., 2010a; Walker et al., 2008); this sediment contained lenses with signs of burning. Fragments of 3 more Neanderthal mandibles were found sieving rubble left by miners. Neanderthal finds include many teeth and bone fragments. In all, at least 9 Neanderthals are represented at the site. The excavated Neanderthal remains corre- spond to a time about 50,000 years ago; different scientific methods, Fig.2, give estimates implying >40,000-<60,000 (U-series, TL, 14C: for details, see Walker et al., 2012b). Palaeopalynology indicates cool moist conditions though with persistence of species ill-adapted to resist frost (Carrión et al., 2003), and a time before the cold Heinrich 4 episode may be inferred. Excavated with the SP97 child were the only articulated bones of large animals found at the site so far, viz., two sets of horse ankles (calcaneum, talus, cuboid), one group, cemented by CaCO3 to SP97’s skull, had undergone burning, the other, unburnt, beneath SP97’s trunk, included additionally a distal tibial fragment, and a third horse talus lay underneath SP97. Two leopard metacarpal bones lay near SP97’s skull and 2 leopard hind-paws, with metatarsal and phalangeal bones in anatomical articulation, lay 0.5 m from SP97’s skull in a similar level; no doubt preserved thus by the same process that maintained anatomical connexion in the 3 skeletons. It is unlikely horse ankles and leopard paws were fortuitous accumulations, given absence around SP97 of other body parts of those species. A heavily burnt leopard temporal bone found in mine rubble implies Neanderthal intervention; a large premolar tooth implies presence of cave hyaenas. Flakes and spalls from flint-knapping lay close to SP97. The site has provided some carefully prepared Levallois points on flat triangular flint flakes with finely-retouched 411
MEDITERRANEAN BASINS. CENTRE.  Figure 1. Neanderthal skeleton SP 96     Paloma    .  proportions than estimates derived fr...
412 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 2. La Sima de las Palomas del Cabezo Gordo: Stratigraphical position of articulated Neanderthal skeleton SP 96 (“Paloma”) within cemented rock tumble (broken lines) and situations of dated materials and two other Neanderthal mandibles (see Walker et al., 2012a). margins, surely effective for hunting with thrusting spears. Flint outcrops are unknown on Cabezo Gordo. Some flint quite likely came from a hydrothermal outcrop inspected ca. 25 km to the S though it is hardly the source of the bulk of the Palaeolithic assemblage. Despite lack of clear-cut signs of intentional interment, large stones might have been thrown over SP96, SP92 and SP 97 to deter leopards and hyaenas from disturbing the corpses. This is a prosaic interpretation. of sediment, wind and rain may scatter bones where carnivorous animals and birds have failed (exceptionally, skeletons of creatures trapped in caves or swamps escape from being scattered). At La Sima de las Palomas 3 articulated skeletons from 50,000 years ago lay close together. It raises a conjecture that behavioural or cultural impingement occurred, implicating individuals other than the 3 deceased. Maybe it is an instance of Neanderthals attending to their dead, albeit with a prosaic motive. Rarely are Neanderthal skeletons uncovered in anatomical connection. When animals die their soft parts decompose quickly, aided and abetted by various organisms (carnivorous animals or birds, insects, saprophytic fungi, bacteria, etc.), after which skeletons come apart. Long before they can get buried by natural deposition The 3 Neanderthal skeletons and the rock tumble over them lay on a thin bed of extraordinarily hard conglomerate, though it contained some Palaeolithic artifacts and charred bone fragments. Beneath it coarse sediment has been excavated to a depth of 2 m so far, containing many bones of red deer, horse and other herbivores, many of
412  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
MEDITERRANEAN BASINS. CENTRE. which are charred, as well as rabbits and other small animals, including mandibular fragments of two porcupines (Hystrix brachyura: Rhodes et al., 2013). Tortoise seems to have played a part in the diet (Morales-Pérez and Sanchis-Serra, 2009). Characteristically Mousterian flint artifacts are present. Several smooth rounded cobbles were undoubtedly brought by Neanderthals to the site from stream gravels in the plain below. Andreu Ollé*,**, Palmira Saladié*,**, Josep Vallverdú*,**, Isabel Cáceres**,*, Jan van der Made***, Isabel Expósito*,**, Francesc Burjachs****,*,** Lucía LópezPolín*,**, Carlos Lorenzo**,* Maria Bennàsar*,**, Domingo Carlos Salazar-García*****,******, Carme Olària*******,******** Being larger than some hammer-stones from the site they might have been used to pound or grind minerals (perhaps haematite; the Cabezo Gordo marble contains veins of magnetite and other iron ores) or foodstuff. Vegetable food at La Sima de las Palomas is suggested both by phytoliths discovered in calculus on some Neanderthal teeth (Salazar-García et al., 2013) and two examples of dental caries (Walker et al., 2010b). La Cova de Dalt del Tossal de la Font Geographical location Background research The cave of Tossal de la Font is situated in a complex structural karstic formation located on one slope close to the town of Vilafamés, 25 km north of Castelló de la Plana. Its dimensions reach more than 2 km of interior paths (Castelló, 2003), and it is developed in Jurassic brecciated dolomites. Archaeological fieldwork conducted on the site between 1982 and 1987 uncover an important karstic filling from the Upper Pleistocene (Gusi et al., 1983; 1987), as well as a set of Holocene occupations (Gusi and Aguilella, 1998). The Pleistocene evidences retrieved, whose date was estimated around 90,000 years old, included a full faunal list and a restricted lithic assemblage, besides two human fossils assigned to Neanderthals (Arsuaga and Bermúdez de Castro, 1987; Arsuaga et al., 2001), and a tooth frag- The site of Cova de Dat del Tossal de la Font is on the upper section of this karstic system. Its UTM coordinates are (ETRS89) 30N X=751380, Y=4444419, at 357 metres a.s.l. * Institut Català de Paleoecologia Humana i Evolució Social (IPHES), C/ Marcel·lí Domingo s/n (Edifici W3), Campus Sescelades, 43007 Tarragona-Spain. ** Àrea de Prehistòria, Universitat Rovira i Virgili (URV), Av. Catalunya 35, 43002 Tarragona-Spain. *** Dpto. de Paleobiología, Museo Nacional de Ciencias Naturales, C.S.I.C., José Gutiérrez Abascal 2, 2800****** Madrid-Spain. **** Institució Catalana de Recerca i Estudis Avançats (ICREA). ***** Dept. of Archaeology, University of Cape Town (Cape Town, South Africa). ****** Research Group on Plant Foods in Hominin Dietary Ecology, Dept. of Human Evolution, Max-Plank Institute for Evolutionary Anthropology (Leipzig-Germany). ******* Laboratori d’Arqueologia Prehistòrica, Dept d’Història, Geografia i Art, Universitat Jaume I (UJI). Campus Riu Sec, 12071 Castelló de la Plana-Spain. ******** Servei d’Investigacions Arqueològiques i Prehistòriques. Diputació Provincial de Castelló (SIAP). P.O. 31******. E-12080 Castelló de la Plana-Spain. 413
MEDITERRANEAN BASINS. CENTRE.  which are charred, as well as rabbits and other small animals, including mandibular fragmen...
414 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. Synthetic log of the Pleistocene talus cone deposits from Tossal de la Font. Legend: 1, fenestral porosity; 2, cemented zone; 3, horizontal stratification; 4, cracking with decarbonation and load structures under clasts; 5, fragments of stalagmites; 6, massive to crudely stratified deposit; 7, blockss and gravels; 8, micro laminated stalagmite layer; 9, microlaminated stalagmite empty; 10, secondary unconfromity; 11, listric fissure. ment (Olària et al., 2007). Between 2004 and 2012, in a collaborative research initiative among UJI, SIAP and IPHES, a second period of archaeological fieldwork was carried out on the site. The main aims were: (I) to review the Pleistocene deposit stratigraphy; (II) to date it; and (III) to extend the excavation area, increasing the available data in order to enhance the paleoenvironmental context, the taphonomic processes, and the contextualization of zooarchaeological and technological remains found in the first period of excavations (Saladié et al., 2010). The ultimate depletion of the fertile deposit, the big effort required to reach it, the increasing cementation of the same deposit, and the reiteration of the fossils appearing, led to the end of this second period of fieldwork in 2012 (Gusi et al., 2013). The stratigraphic sequence The fossiliferous deposit that presents a major tilt and limited lateral continuity is situated on the cave entrance, just a few metres from the access, and fills a narrow fringe between the cave wall and a talus cone originated by large endokarstic boulders. The archaeopaleontological units have been established considering the different deposits accumulated through fissures and marked by decarbonation. Two big sets have been identified. In each of them gravitational sedimentation on high slopes has been produced. The entrance deposits have limestone boulders mixed with sandy clay, probably coming from the breccia cemented with large boulders and speleothems that indicate the
414  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
MEDITERRANEAN BASINS. CENTRE. closing of the point of entry. Each unit has this chemical deposit on its top (Fig. 1). relating with defleshing activities and intentional bone breakage have been identified (unit IIa). The most abundant paleontological record is on unit IIa. This unit presents a truncated stalagmitic crust that appears transformed in clasts at unit IIb. These sedimentations and emptying processes characterize the filling of that cave-fissure, where a steep slope and mass wasting movements played a key role in the opening of new points of entry and in the creation of the space to collect breccia deposits with embedded macrovertebrates and lithic remains from the Upper Pleistocene. Carnivore activity, on the contrary, is well represented, and can be traced through different marks such as punctures, pits, scores, crenulated and jagged edges, digested bones. Gnawed marks primarily appear over red deer bones indicating the action of a big predator, such as a hyena, lion, or bear. Some leporid remains also show punctures and scores pointing out chewing activities of carnivores such as lynx or fox (Fig.2, k-n). Although no new human remains have been recovered, the section of the old level E. where they appear in, located in the original publication (Gusi et al., 1983: 17) clearly corresponds with unit IIa. Human remains Pending on new results, U/Th dates are available for two breccia sealed by stalagmitic crusts and situated on top of unit IIa: 61,846 ±585 years BP and 56,014 ±484 years BP. The beginning of the isotopic stage 3 can be quite an accurate chronological assignation for this unit, since both dates are allegedly from the same stalagmitic crust. The archaeological record On the first campaigns, the identification of several rodent species allowed the site to be placed in a chronological framework between the late interglacial Riss-Würm and the beginning of Würm glacial stage (Gusi et al., 1987). At the most recent campaigns, in unit IIa, Iberomys brecciensis has been identified,a characteristic species of the Middle Pleistocene that existed until the Upper Pleistocene. Apart from micromammals, and unit IIa always being the richest and most diverse regarding bone assemblages, the faunal list includes: Lynx spelaea, Felis sylvestris, Crocuta crocuta, Equushydruntinus, Rhinocerotidae, Sus scrofa, Cervus elaphus, Capra/ Hemitragus sp., Bovinae indet., Chelonia and Oryctolagus cunniculus. Cervus elaphus is by far the best represented taxa, and is also the species that shows more variability in the representation of skeletal parts. Distal parts of limbs predominate in faunal assemblages. They are segments that often appear anatomically connected, or located at a few centimetres. Both the rest of the appendicular bones and the axial skeleton are under-represented, and display a more scattered spatial distribution. Anthropogenic activity over faunal remains is very scarce. Some red deer bones with cutmarks, The three human fossils recovered (Fig. 2, o-p) are a distal fragment from a left humerus, a fragment of right coxal and a fragment of tooth. The part that has been preserved in the humerus (CTF1) corresponds to a third distal that has the whole joint and displays several characteristics that makes it alike to the European Middle Pleistocene populations represented at Sima de los Huesos and also to the Neanderthals (Arsuaga et al., 2001). From the coxal fragment (CTF-2) only the upper part of the femoral joint remains and part of the ischial spine. Although their fragmentary conditions do not allow the distinction of taxonomic traits, the presence of a very pronounced supra-acetabular groove and the thickening of the greater sciatic notch edges should be stressed. The tooth fragment (CTF-3) corresponds to a maxilar molar from a child. Lithic industry The available lithic assemblage is very scarce. In total, there are 8 small flint artefacts, 2 from unit Ic and f and 6 from unit IIa (Fig.2, a-j). Half of these tools are simple flakes. Among retouched flakes there are two points, one sidescraper and one denticulate sidescraper. Half of the artefacts have developed patinas, but they surfaces do not display other postdepositional modifications visible (such as erosion or false retouch). One of the flakes shows fine cracks from heat damage. Finally, an single artefact of quartzite, a broken pebble with evident marks of being used as hammerstone has been recovered. Little can be said about the tecnhnotypological characterization of the lithic assemblage, beyond to roughly assigning it to a Mousterian context. Neither is there enough data to infer occupational interpretations based on lithic assemblage. Nevertheless, the collection has a great taphonomic value, since it ap- 415
MEDITERRANEAN BASINS. CENTRE.  closing of the point of entry. Each unit has this chemical deposit on its top  Fig. 1 .  re...
416 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 2. Flint flakes from unit Ic and f (a,b) and from unit IIa (c,e); Retouched flint flakes from unit IIa; denticulate sidescraper (f), sidescraper with marginal retouch (g) and points (i, j); quartzite pebble broken (d), and with percussion marks (e), unit Ic and f; red deer phalanx with a vacuum in its proximal part, unit Ib (k); punctures over coxal bones produced by a small size carnivore (l and m: units Ic and f, and IIc and f, respectively); puncture provoked by a big size carnivore over a vertebra from a medium size animal, unit IIa (n); fragment of human humerus CTF-1 (o); and fragment of human coxal bone CTF-2 (p).
416  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
MEDITERRANEAN BASINS. CENTRE. pears closely related with fossils, and demonstrates the development of human activities in the cave itself or in the immediate surroundings of the cave entrance. Concluding remarks The information gathered led us to interpret that fossils in Tossal de la Font are in secondary position. This location was the result of short distance transportation, since there is low occurrence of modifications that could indicate friction over the substrate such as trampling, rounding and abrasion marks on Carmen Cacho*, Jesús Jordá Pardo** El Tossal de la Roca The archaeological site of Tossal de la Roca is situated in Vall d’Alcalá (Alicante), at 640 m.a.s.l. and 20-25 km away from the present coastline. This west oriented rockshelteris situated in a mountainous area in a foothill on the left bank of the Penegrí creek, one of the small tributaries that configures the short current network of the Serpis River fluvial system. The archaeological sequence, excavated by our team between 1981 and 1999, ranges from the Late Upper Pleistocene (end of OIS 2) to the first third of the Holocene (beginning of OIS 1). It has a discontinuous sequence that spans from Upper Magdalenian to the Mesolithic with trapezes documented in two different areas of the rockshelter. The interior sector (Fig.1) contains several Upper Magdalenian levels (III and II int), and Final Magdalenian (I int), whereas the exterior sector includes diverse occupations from the Notched and Denticulates Mesolithic (II b and II a ext.) and Geometric Mesolithic (I ext.). Radiocarbon dates have provided, the following chronological span for the stratigraphic units, from bottom to top: 17200-16310 cal BP (level III int.), 15550* ** bone surfaces. The fossil association responds to the addition of different events, each of them could include in turn several causes of death and taphonomic processes. In the case of Tossal de la Font, the result is that fossils present a twofold history, one biostratinomic due to exokarstic factors and another fossildiagenetic one caused by endokarstic agents. The first one is characterized mainly by carnivore action and, to a lesser extent by hominid action over some carcasses. The second one can be observed by the presence of dissolution, chemical corrosion produced by the sediment, and also generalized processes of cementation and formation of breccia deposits. 14040 cal BP (level II int.), 13780-13580 cal. BP ( level I int.), 10550-9410 cal. BP ( level IIb ext.), 9510-8640 cal BP ( level IIa ext.), 8560-8230 cal BP ( level I ext.) (Cacho and Jordá 2009: 222-227; Jordá and Cacho, 2008). Palynological and antracological analysis as well as micromammal studies from Tossal de la Roca indicate that level III (int.) developed in temperate and moist climate conditions, conversely, cold and arid conditions are documented for level II (int.); vegetation cover is reduced to conifer at both levels. In the upper part of level II and especially at level I (int.), a significant climate change is reflected, pointing at milder and moister climatic conditions that contribute to an open environment vegetation development and to a large variety of species, where Quercus gain importance at the expenses of conifers; other thermophilic taxa as Juglans, Ulmus, Betula are documented. From level IIb (ext.) there is a great advance of the Mediterranean forest (oak forest) with a remarkable presence of Quercus, which becomes the prevailing species in the subsequent periods (levels IIa and I ext.), together with the increase of other taxa indicative of some humidity. Departamento de Prehistoria. Museo Arqueológico Nacional. Serrano 13. 28001 Madrid. Spain. Departamento de Prehistoria y Arqueología. Universidad Nacional de Educación a Distancia. Senda del Rey 7. 28040 Madrid. Spain. 417
MEDITERRANEAN BASINS. CENTRE.  pears closely related with fossils, and demonstrates the development of human activities in...
418 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD characteristics and quality of raw materials. Both strategies are oriented to bladelets production, lately transformed by pressure debitage on backed bladelets. Retouched artefacts assemblage from these stratigraphic units show a clear predominance of tools made from bladelets, with a large typological variety at level III. Also to be noted is the presence of some scalene triangles. Endscrapers are the second most relevant type of this assemblage, while burins, mainly dihedral, appear in a very low proportion (Cacho and Martos, 2004: 100-101). Figure 1. Stratigraphy of the interior sector. Tossal de la Roca. Rabbit is the most frequent species of Tossal, as it happens in other Mediterranean Magdalenian sites, but considering its low energetic value it should be considered as a complementary diet intake, ibex hunting and, to a lesser extent deer hunting, being the main nutritional food source. From the Geometric Mesolithic, deer consumption increases over ibex, and new species such as chamois or wild boar are incorporated. According to the taphonomic study, ibex and deer hunted during the Magdalenian occupations would have been partly consumed in the hunting place and only some of its portions would be transported to the site. This behaviour will change in the Notched and Denticulates Mesolithic period and even more during the Geometric Mesolithic when the entire prey was carried to the site. Besides this prevailing species, other resources were consumed such as partridges and some fish such as trout and eel during the Magdalenian period, together with the gathering of some fruits and leguminous plants during the Mesolithic occupations (Cacho, 1995: 95-96). Lithic industry is overwhelmingly made of flint from the close surrounding, approximately in a radius of 4 km of distance. Technological study from the Upper Magdalenian levels reveal reduction strategies with predetermined debitage from standardized cores (prismatic), and also a flexible system (amorphous core) that adapts to the After the Late Magdalenian characterized by an increase of endscrapers and a low diversification of tool made from bladelets, there is a hiatus in the Tossal de la Roca sequence that will start again with level IIb in the outside sector, assigned to the Notched and Denticulates Mesolithic. Lithic industry from that stratigraphic unit, dated between 10580 and 9390 cal BP, reflects significant changes in regard to the Late Magdalenian. On the retouched artefact assemblages, bladelets tool decreases at the same time that short endscrapers increase, but the appearance of notches and denticulates in large proportion (20-40%) is the most outstanding change. These artefacts are always very thick and present a typical retouch known as Campiñoide (direct-abrupt/ stepped and bifacial) in most cases. From a technological point of view these lithic assemblages (levels IIb and Ia ext.) configure a flake industry –where in turn flakes are used as core-like flakes–, in addition microblade tools are really scarce limited to little projectile points for hunting. This flake industry with thick notched and denticulates artefacts persist during the Geometric Mesolithic, where trapezes on bladelets are manufactured (Cacho and Jordá, 2009: 230-233). Bone industry is scarce and appears rather fragmented. In the Magdalenian assemblages the presence of assegai should be stressed, some bone points and one needle, but the most relevant artefact is a complete bone harpoon. This has four insinuated teeth though not totally exempt, and a straight base without any kind of retaining mechanism. Similar harpoons to the one of Tossal de la Roca have appeared on other Upper Magdalenian records from the Mediterranean coast of the Iberian Peninsula such as Matutano or Cendres (Cacho and de la Torre, 2005: 259-261; Cacho et al., 2001: 85-87). On Mesolithic levels bone industry is very limited and only consists of some bone points and
418  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  char...
MEDITERRANEAN BASINS. CENTRE. some spatula. Without stratigraphic context, several slabs with zoomorphic representations where cervids and caprines depictions prevailhave been recovered, almost always isolated, although there are some overlapping cases (Fig.2). Schematism is one of the most characteristic traits of this representations, especially on limbs, whereas in the design of head and neck some anatomical details are incorporated (Cacho and Ripoll, 1987: 54). In addition, one pendant on bone spatula with geometric decoration has been recovered (D’Errico and Cacho, 1994). As a novelty, traceology studies have suggested that partridge wings could be used as personal ornaments during the Magdalenian period (Sánchez and Cacho, 2010). Figure 2. Decorated slab from Tossal de la Roca. The most relevant contribution of Tossal de la Roca is the multidisciplinary study of one of the most complete archaeological records from the Upper Pleistocene and the beginning of the Holocene in the Mediterranean coast of the Iberian Peninsula. Research on this site situated in a regional context enables assessment of the environmental transformations which occurred at this transitional period which forced the last hunthergatherer populationsinto a steady adaptation process. This process would have implied major territorial control and deep transformations of food resources procurement strategies, technology, and artistic expression (Cacho and Jordá, 2009: 233234). 419
MEDITERRANEAN BASINS. CENTRE.  some spatula. Without stratigraphic context, several slabs with zoomorphic representations ...
6 southern mediterranean coast, guadalquivir river and betic intramontane basins NORTH-WESTERN ATLANTIC BASINS. 421
6  southern mediterranean coast, guadalquivir river and betic intramontane basins  NORTH-WESTERN ATLANTIC BASINS.  421
Site Map numbering El Aculadero 74 Ardales 75 Bajondillo 76 Cueva Ambrosio 77 Cueva de Nerja 78 Cueva and abrigo del Ángel 79 Cueva del Boquete de Zafarraya 80 Cuenca de Guadix – Baza and Conjunto de Orce 81 Las Grajas de Archidona 82 El Pirulejo 83
Site  Map numbering  El Aculadero  74  Ardales  75  Bajondillo  76  Cueva Ambrosio  77  Cueva de Nerja  78  Cueva and abri...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Manuel Santonja*, Alfredo Pérez-González* El Aculadero (Puerto de Santa María, Cádiz) 1. Initial assessment of El Aculadero and the 1973 – 1980 excavation campaigns The El Aculadero site, located in the northernmost corner of the Bay of Cádiz (southern Spain), was initially discovered and surveyed by Claude Viguier in collaboration with Claude Thibault, under the supervision of François Bordes. The first publications interpreted this discovery as the first clearly European Oldowan site (Bordes and Viguier 1971; Bordes and Thibault 1977), comparable to the known sites in the North Africa Casablanca region. El Acudalero was excavated from 1973 until 1980 under the leadership of Claude Thibault and Ángeles Querol. Their main objectives were to confirm the absence of Acheulian technocomplex in the site and thoroughly study of the stratigraphic sequence. The 127 m2 opened only yielded lithic material in quartzite (91%) and other rock –schists, quartz, flint and limestone–, all found on the same level as the pebble industry. The assemblage consisted of 934 varied knapped pebbles –in some cases probably elementary cores–, 573 retouched flakes, 133 cores, and 1160 flakes and fragments (Querol and Santonja 1983). The tools on flake consisted of pieces with generally non-intense retouch, although some 60 items were defined as sidescrapers or denticulates. There were no Acheulean tools –neither bifaces, nor cleavers, nor trihedral pics–. The most significant exploitation schemes in the cores were discoidal (45 items) and in some cases Levallois (Fig. 2). The presence of tools indicated that this was not exclusively spot of primary raw material captation (Querol and Santonja 1983). However, the lack of debris and the low ratio between products and knapped blank –two flakes per three cores– indicated poor conservation of the initial ensemble, * Figure 1. Plio-Pleistocene stratigraphic sections on the El Aculadero seacliff (Querol and Santonja 1983, cf. fig. 6, p. 23). Continental units IV and V are deposited on marine units bed I, II and III (columns 3, 4 and 5). El Aculadero site (CT) is located at the bottom of the continentals units IV and V. which may have partly disappeared due to erosion processes. However, the deficit of flakes, particularly non-cortical flakes, may be due to their deliberate transfer to more specialized sites (Santonja and Pérez González 2010). Centro Nacional de Investigación sobre la Evolución Humana. (CENIEH); 09002 Burgos. manuel.santonja@cenieh.es 423
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Manuel Santonja , Alfredo P  rez-Gonz  lez...
424 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 2: 1: Levallois flake. 2: Retouched Levallois flake. 3 and 11: Sidescraper on pebble. 4 and 6-9: Sidescrapers. 5: Denticulate. 10 and 13: Levallois cores. 12: pointed pebble-tool.
424  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS In 1983, findings from the study of this industrial assemblage led to the suggestion that the technology employed here was comparable to that observed in the Iberian Middle Pleistocene sites. A new stratigraphic interpretation was particularly focused on the highly erosive contacts between the Pliocene sea levels and the location of the site in the stratigraphic sequence (Fig. 1). These discontinuities suggested that the correct age could be in the second half of the Middle Pleistocene (Querol and Santonja 1983: 244-246). 2-10 µ grain size of the mineral fraction. The Radiochemical Dating Lab at the Autonomous University of Madrid provided the following results (Santonja and Pérez-González 2010): – Sample D: Collected from the yellow sand level on the paleosol developed, in the lowest part of UIV. A red deposit (2.5 YR 4/8) with fine-medium grain sand and silt-clay content (33%). Dated at 110507 + 7481 BP. – Sample C: From the cemented dune above 2. Interpretations of the Lower Palaeolithic in this area since 1983 Studies in recent years have detected Acheulean series including handaxes and cleavers on terraces of the Guadalete River such as Laguna de Medina (Giles et al., 1996). Other terraces of the same river have yielded Acheulean ensembles attributed to the Middle Pleistocene, a comparable situation to Spain’s Central Meseta (Plateau), where Acheulean industries range chronologically from MIS 13 to MIS 6 (Santonja and PérezGonzález 2010a). On the other hand, several assemblages with similar industrial series to El Aculadero have been reported in the vicinity of the Bay of Cadiz and further west. The number of such sites has gradually increased, and in some cases have been proposed ages near the end of the Middle Pleistocene (Giles et al., 1994). At some of these sites, a Mousterianlike development of tools on flake has been found along with discoidal cores and knapped pebbles of the El Aculadero type. 3. El Aculadero dating In the El Aculadero stratigraphic sequence, continental units (UIV and UV) are superimposed on others (UI, UII and UIII) of a marine origin or influence (Fig. 1). The site is located on a quartzite and quartz– gravel level at the base of Unit IV (CT, Fig. 1). This level, along with an underlying sandy level, form a cover-glacis layer which also contains a red soil developed on the sandy layer from before or after the level containing the pebbles and industry (Querol and Santonja 1983). In this edaphic-sedimentary context, four samples were taken for optically stimulated luminescence dating (OSL) –additive dose method–, in the pebble and industry level. This well preserved dune can be clearly identified with Zazo’s Unit 11 (Zazo et al., 1983). A medium grain sand deposit with scarcely clay-silt fraction (<2%), light brown colour (7.5YR 6/4). Dated at 62914 + 5094 BP. – Sample B: Collected in the north of the excavation site, in medium to fine sands with less than 22% silt + clay, brown colour when moist (7.5YR 5/4). Its stratigraphic position is equivalent to the basal part of the top of UV (Section 5, Fig 1). Dated at 40493 + 2434 BP. – Sample A: Recent dune consisting of loose sand, to top of UV (Section 5, Fig 1), light brown colour when moist (7.5 YR 6/4). Fine to medium grain sand and almost no silt + clay fraction (<1%). Dated at 7000 + 466 BP. These four samples permit an objective estimation of the age of the site, situated between samples D and C. The results concur with others conclusions. For Zazo (1989: 119, Fig. 3), the continental materials on the El Puerto de Santa María seacliff are later than the Trafalgar marine episode, dated by Th/U at 100 ky BP. This chronology matches the OSL date of 110.5 + 7.5 ky BP in the UIV sands affected by pedogenesis which led to a well-structured and rubefacted soil during a warm episode, probably corresponding to the Eemian interglacial. If this edaphogenesis did affect the level containing gravels and industry, the age of the site would be no less than 110 ky BP. Otherwise it could be subsequent to the Eemian interglacial. OSL dating of Sample C, collected from the cemented dunes, was 62.9 + 5.0 ky BP, marking then the youngest limit of the El Aculadero industry. 425
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  In 1983,    ndings from the study of this ...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 426 5. Conclusions The OSL dating presented herein defines the age of the El Aculadero site as either in the Eemian interglacial, which ended around 110 ky BP, or in the MIS 4, before c. 63 ky BP. These results rule out not only the earlier datings of the industry, but also the postulated modern ages including the Holocene. On the Atlantic coast of southern Iberia, a group of comparable sites to El Aculadero have been recorded, all characterized by retouched tools on flake Pedro Cantalejo*, Jose Ramos**, GerdChristian Weniger***, Martin Kehl **** Maria del Mar Espejo***** Cueva de Ardales, Province of Malaga The Ardales cave (UTM 337.110/4.082.540) is located in a mountain called Cerro de la Calinora 565 m a.s.l. near the village of Ardales about 50 km north of the Mediterranean coast. It was discovered in 1821 after an earthquake exposed a cave entrance previously sealed by colluvial deposits. From 1852 on, the cave was opened to local tourism without recognizing its prehistoric finds. It was Henri Breuil who recognized the Palaeolithic heritage during a visit in 1918 and who first studied the rock art of Ardales cave (Breuil 1921). The site then lost attention for decades. After a time of scientific standstill, research restarted in 1990 (Ramos et al., 1992). A detailed complete documentation of the artistic inventory was finished after more than 10 years of study in 2005 (Cantalejo et al., 2006). From 20112013, first limited excavations where conducted in the entrance area (Fig. 1) and coring outside the front of the cave took place (Ramos et al., 2014). The site is a multi-branched karstic system that is separated into five areas: Area I (Sala * along with shaped pebble tools. At El Aculadero, the first stages of production are well documented, while at other sites the sub-phases of full production and consumption, represented by significant percentages of retouched tools suggest a more varied range of activities. All of these sites have a common technological identity. This is a local Middle Palaeolithic facies, stretching along a timeline starting in the initial Upper Pleistocene or perhaps earlier. On the basis of known data, it is later than the Acheulean sequences on the Guadalete terraces. del Saco), area II (Sala de las Estrellas), area III (Galería de los Laberintos), area IV (Calvario) and area V (Galerías Altas). This later area was discovered in 1981 by speleologists. It is a separate cave system above the main cave area. Today the Galerías Altas are accessible only by a narrow fissure that can be reached from the Galeria del Arquero by climbing vertically about 18 m high. The natural entrance to this area was probably sealed by a slide in the late Holocene. Coring outside of the original entrance area of the Galerías Altas in 2011 provided evidence for such an entrance. This part of the cave has not yet been analysed systematically. But, burials from Copper Age and Palaeolithic rock art have been recorded from short expeditions into the Galerías Altas. By sealing off the entrance, prehistoric surfaces including dispersed artefacts and burials have been conserved perfectly. The galleries of Ardales cave have a total extension of more than 1,500 m. The cave is accessible today via a system of stairs, which was constructed by its first owner Doña Trinidad Grund in the mid- Ayuntamiento de Ardales-Cueva de Ardales. Avenida de Málaga 1, 29550 Ardales, Málaga, España, pedrocantalejo@gmail. com ** Universidad de Cádiz. Facultad de Filosofía y Letras. Avda. Gómez Ulla s.n. 11003 Cádiz. jose.ramos@uca.es. *** Neanderthal Museum. Talstr. 300, 40822 Mettmann, Germany; weniger@neanderthal.de **** University of Cologne, Institute of Geography. Albertus-Magnus-Platz, 50923 Cologne, Germany; kehlm@uni-koeln.de ***** Cueva de Ardales. 29550 Ardales, Málaga, España. Investigadora, Grupo PAI-HUM-440, mariadeespejo@gmail.com
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  426  5. C...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Figure 1. The entrance cone of Ardales cave. Test excavations are marked by red figures (1-4). The red square marks Zone 2 with a multi-layered stratigraphy dle of the 19th century. The stairs were cut into a steep sediment cone that stretches over 20 m from the opening of the cavity down to the Sala de Estrellas (Fig. 1). The cone is the result of frequent sediment deposition from the slope above the cave entrance. Ardales cave is outstanding in Southern Iberia for its numerous examples of paintings and engravings from the Upper Palaeolithic. Breuil described in his first analysis about 20 animal figures from 10 panels. Until today, 1010 pictorial artefacts have been described from 252 different sites (Cantalejo et al., 2006 and 2014 b). These include 787 signs, 106 unclassified motifs, 98 animal figures, 10 human figures and 9 hand stencils – positives as well as negatives. All kind of surfaces were used for artistic expression: walls, ceilings, grounds, speleothems and blocks. Within the animal representations cervids dominate (64%) followed by equids (26%), others are statistically of minor importance. From the cervides 85% represent females. Two depictions of birds, including a flamingo (Fig. 2), one reptile and a fish are noteworthy. The chronology of the rock art is divided by stylistic analysis into three stages. The initial phase is supposed to represent a Gravettian chronology followed by the middle phase equal to the Solutrean and a final phase attributed to the Magdalenian. Beside the pictorial artefacts an important number of additional finds linked to the human use of the cave have been conserved. These are stone or bone tools placed near the panels with rock art, which might have been used for engravings, paste of red and brown pigments, stone containers used as pigment palettes (Fig. 3) and stone lamps used for artificial lighting (Fig. 4). To date 13 artefacts probably linked to the artificial illumination of the cave were found. Beside mobile stone lamps, immobile lamps made by reshaping of stalagmite bases and very limited micro combustion areas indicated by small heaps of charcoal are documented (Cantalejo et al., 2014 a). A research program for residue analysis by Raman 427
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Figure 1. The entrance cone of Ardales cav...
428 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 3. Profile of Zone 2. The bottom of the sequence immediately above the third stalagmitic crust is dated to 15,945 ± 60 yrs BP. The top of sequence is dated to 3,718 ± 40 yrs BP (all dates uncalibrated). Figure 2. Engraving of a flamingo scanned with a structured light scanner (Breuckmann SmartScan). spectroscopy and radiocarbon dating of charcoal of these items is in progress. Limited test excavations conducted from 2011-2013 at the entrance cone give first results for the absolute chronology of Ardales cave. The entrance cone of the cave (Fig. 1) was originally covered by a stalagmitic crust that was partly destroyed during the construction work of the stairs in the 19th century. In archaeological zone 2 (Fig. 1 and Fig. 5) a multi-layered stratigraphy could be recognized. Below the first and the second stalagmitic crust sediments belonging to the Holocene could be detected together with a small number of ceramics, lithics and fragments from a human cranium. Three radiocarbon dates from charcoal (COL 1640: 3,718 ± 40 yrs BP; COL 1637: 3,621 ± 35 yrs BP; COL 1636: 3,885 ± 36 yrs BP) indicate a chronology for the Copper Age. Below the second crust Pleistocene sediments were found. A radiocarbon date from a charcoal sample immediately above the third crust (COL 1639: 15,945 ± 60 yrs BP) suggests a Solutrean age. Bone fragments and a burin were associated with the sample. Sediments continue below the third crust. Manual drilling was performed using a soil auger, which reached a depth of 35 cm below the crust. The chronology of the sequence is supported by pollen analysis that indicates from the third crust to the top an increase of temperature and humidity (Ruiz Zapata and Gil García 2014). The Holocene occupation of the cave is further supported by radiocarbon dates from test excavations of zone 4. Here charcoal samples from a nearby combustion feature indicate the same chronology as the upper part of zone 2. In zone 3, two charcoal samples gave results of more than 50 kyrs BP. The very small test excavation gave an additional result from a bone fragment of Felis silvestris of Holocene age (COL 2011.1.1: 5,562 ± 48 yrs BP). The contradicting dates need further sampling and expansion of the test excavation. Confirmation of a Middle Palaeolithic chronology would be in agreement with the observation that during the remodelling in front of the cave for the construction of the entrance building lithic artefacts of Middle Palaeolithic technology were discovered. Additionally, 200 m downslope of the actual cave entrance a Middle Palaeolithic surface
428  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Figure 4. Top part of a stalagmite that was struck from the stalagmite, turned around and then used as lamp site La Cucarra has delivered a rich lithic assemblage. The recent test excavations in the entrance cone of Ardales cave give new insight into the human occupation of the site. A Holocene and late Pleistocene occupation could be confirmed. Due to the steep topography of the entrance hall no suitable space for an occupation was available in the mouth of the cave in pre- historic times. Only in front of the cave or inside at the foot of the steep cone plane surfaces were available. A study of lighting conditions by ray tracing gives evidence that the cone area inside the cave was basically without day light in prehistoric times (Hoffmeister et al., i.p.). Occupation and movement in the cave was possible only with the help of artificial light by fire or lamps. The number of finds that document artificial illumination in the cave represent an important sample for further analysis. Holocene occupation was probably linked to burial activities, while Pleistocene occupations were probably related to art activities. A great number of surface finds like lithics and other artefacts as well as human bones and animal bones are still in original position. Some of them are coated by thin stalagmitic layers. The ongoing study of their distribution in relation to the network of paths and items of rock art will give important insight into the human use of the cave. The blend of a rich record of rock art distributed over the whole interior with an excellent surface preservation of many areas of Ardales cave are a most valued resource for the study of past human behaviour. Figure 5. Top part of a stalagmite that was used as container for pigment. 429
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Figure 4. Top part of a stalagmite that wa...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 430 Miguel Cortés Sánchez*, María D. Simón Vallejo** Bajondillo cave (Torremolinos, Malaga) Abstract Bajondillo cave is an archaeological site with a sequence that ranges between c. MIS 7 and V millennium BP, with occupations from the Middle Palaeolithic, Upper Palaeolithic and late prehistoric period. The data obtained from its sedimentary deposits and paleogeographic analysis of the western area of the bay of Malaga are essential for understanding the paleo/climate-environmental evolution of the coast of Malaga during the recent Quaternary period. The archaeological sequences permit the chronoculture of the recent Quaternary period in the south of the Iberian Peninsula to be established. Figure 1. Bajondillo cave: Overview of records from the chronostratigraphical sequence. * ** Prehistory and Archaeology Department. Faculty of Geography and History, University of Seville, c/. Doña María de Padilla s/n. 41004. Seville. mcortes@us.es Archaeological Museum of Frigiliana, c/. Cuesta del Apero, 10. 29788-Frigiliana (Malaga) (Spain). simonmd63@gmail.com
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  430  Migu...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Figure 2. Lithic industries, marine molluscs and barnacles. 431
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Figure 2. Lithic industries, marine mollus...
432 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Keywords: Recent Quaternary period. Chronocultural sequence. Middle Palaeolithic. Upper Palaeolithic. Neolithic Introduction The western sector of the bay of Malaga has a morphosedimentary record of the different periods of the Neogene-Quaternary period that permits the geomorphic sequence of the north eastern foothills of Sierra de Mijas to be established (fan delta; erosion and sedimentary shaping, as cones and alluvial fans; river terraces; travertine construction and coastal deposits or coastal benches), the reference geographic framework for the populations that used this area as an area for gathering resources, mobility, etc. Bajondillo cave is currently located in the area where the best microclimate conditions of the bay of Malaga converge. It has wide and guaranteed availability, from season to season and from year to year, of environmental resources, for example, springs, raw materials for knapping or the development of a rich ecotone (mountain range, proximity to the river mouth and estuary), which explain the use of the area by Neanderthals and modern human beings. From a tectonic point of view, the area can be considered essentially stable since the mid-Middle Pleistocene, so that the impact of the glacio-eustatic dynamic during the recent Quaternary period can be analysed, which designed similar coastlines to the current (Holocene) and others up to 8-10km away, c. 21-18ka (all dates BP). Bajondillo was discovered (1989) in the travertine construction in Torremolinos during construction work that destroyed areas of the upper sedimentary section (Bj/0-Bj/9). The site was excavated to an average of around 13m2, reaching a maximum stratigraphic depth of almost 6m. The western profile was subsequently surveyed in 2000 and 2002, obtaining different chronological and paleo-environmental information (Fig. 1). The sequence possibly started in MIS 7 and shows the typical climate-environmental dynamic of a coastal influence and the recent Quaternary period. The area never experienced the profound effects of other more northern sequences and allowed a shelter to exist where various species of thermophile are quartered. Figure 3. Lithic industries. Middle Palaeolithic (Bj/17 to Bj14). Sequence Bajondillo has 20 archaeological layers whose chronological sequence has been established by 27 dating processes (14C/AMS, TL and U/Th) of 13 levels (Fig. 1). The travertine where the cavity is located is deposited by constant contact with beach sands at a time that has not yet been well defined but probably within an advanced-late phase of the Middle Pleistocene, perhaps MIS 7, the cavity being filled at the base by a lithochemical plug at the bottom of the cavity, sealed with a stalagmite crust (Bj/20) on which a detrital deposit started to be deposited between MIS 6 (Bj/19) and Early Helocene. The earliest archaeological evidence documented is two knapped chopping tools and some lithic tools below Bj/20 and >150ka (Fig. 2.16).
432  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Keyw...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Middle Palaeolithic (Bj/19-Bj/13), c. 15038/28.5ka. From an economic point of view, the Neanderthal populations took advantage of an ecotone which retains an important tree-shrub cover and incorporates a wide variety of vital resources (water, biotic and abiotic). The Mousterian lithic industries developed Levallois production techniques, predominant, but also discoid, among others, with the presence of ramification and miniaturization. Bj/19-Bj/18 (sedimentary episode [ES]/M) deposited under humid and mild conditions (Fig. 2.19). It is noteworthy that, along with large fauna (deer and aurochs), consumption of marine molluscs is also documented (Fig. 2), in a similar age (c.150ka) to the oldest documented in the world (Cortés et al., 2011), in this case by modern humans. Bj/17-Bj/15 (ES/N-O) indicate colder conditions with fluctuations in humidity. The lithic industry at these levels presents a variety corresponding to Typical Mousterian (Fig. 3). Bj/14 (ES/P), in a colder and drier context, Mousterian of Denticulate assemblages tools are deposited. Upper Palaeolithic (Bj/13-Bj/5. Fig. 4). The oldest are detected in the section Bj/13Bj/11(ES/Q-E), generated in a less cold and dry context than in Bj/14. The industry can be attributed to the Aurignacian (fig. 4A). The technological change compared to the Middle Palaeolithic is abrupt, the disappearance of the last techno-cultural examples from the Middle Palaeolithic seems to occur, as in the rest of the sites in the south of the Iberian Peninsula, with the epigonal occupations of the Middle Palaeolithic, as a phenomenon of collapse, rapid and synchronous disappearance with the development of the Mediterranean Aurignacian in c. 30-28/AMS-28.5ka/TL. B/10(US/S), generated under cold and drier conditions than in Bj/11, it contributed Gravettian industries (c. 26-21.5 ky/TL) (Fig. 4B). Bj/9-Bj/6 deposited in a wet and cold context (US/T), followed by a mild and humid one (US/V), and they contributed industries typical of the Solutrean (fig. 4C). The earliest full Iberian Solutrean phase (Bj/9) and c. 21-19/AMS/21.518,1/TL ky with the presence of artefacts with flat retouch. Bj/8-Bj/6 present evolving Solutrean industries (c. <19/AMS/18-16/TL ky) with notched pieces and high percentages of burins, many of them with truncated retouching. Bj/5(US/V), deposited in wet and cold conditions. There are no diagnosis objects to define it but it falls within the advanced-final Late Glacial Maximum, perhaps c. 13-12ka. Bj/4-Bj/3(ES/W) deposited in the Holocene and chronologies corresponding to the Epipaleolithic and Mesolithic. Bj/2-Bj/0(US/X-Z) are attributable to late prehistoric period (Neolithic-Chalcolithic), c. 7.55ka. The Neolithic seems fully formed, with agricultural activities that leave their mark in the form of seeds of cultivated species, sickle elements or leaving their mark in the pollen record. Bajondillo also has functional studies of collections from the Middle Palaeolithic, of phytoliths, raw materials for knapping or a fragmented example of portable art. Systematic archaeological surveying around Bajondillo (20km2) has allowed us to locate outcrops of lithic raw materials similar to some industries recovered at the site. Conclusions The formation of the travertine that gives rise to Bajondillo is earlier than MIS 6 (Bj/19), perhaps the warm MIS 7 period. The coastal strip, whose variable amplitude was dependent on the glacio-eustatic dynamic of the bay of Malaga, is part of the subsistence territories, at least from the end of the Middle Pleistocene. This effect should undoubtedly be correlated with the high suitability of the environment and high availability of mineral and food resources in the geographical area defined by the travertine deposits of Torremolinos, the southern slopes of Sierra de Mijas and the river Guadalhorce; a geographical area that seems to offer sufficient resources for the territory to be integrated and visited, for more than 150 thousand years, in the core of the subsistence and mobility strategies of the hunter-gather communities (neanderthals and modern human beings) and subsequently crop-livestock farmers. From MIS 6 and up until 36/38 ky, levels (Bj/19Bj/14) are deposited with c. Mousterian industries, from which time we find technology systems from the Upper Paleolithic, at least Aurignacian, Gravettian and Solutrean, with some regional peculiarities. Since it was discovered, the site has degraded quickly, which puts the record kept under serious threat. 433
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Middle Palaeolithic  Bj 19-Bj 13 , c. 1503...
434 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 4. Lithic industries. Upper Palaeolithic (Bj711 to Bj/6).
434  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Sergio Ripoll López, Francisco J. Muñoz Ibañez* The Solutrean Station of The Cueva de Ambrosio (Vélez Blanco, Almería, Spain) Abstract: The last investigations made during the excavation of the different levels at this important site of study the Solutrean in the southeast of the Iberian Peninsula, have permitted us to make some precisions on its chronostratigraphic position in the Upper Pleistocene. The calibration of a new radiocarbon date for level IV (Upper Solutrean) and six other new dates (five of them AMS) for level II (Final Upper Solutrean) allow us to place these two cultural phases between the Greenland Stadial GS3 (end of OIS 3) and the end of the Greenland Intersatial GI 2, with a greater occupation of level II after the Heinrich Event H2, agreeing with the intersadial that come just before the last glacial maximum (GS 2). This new data clearly modifies the dates existing before, making the whole record much older. On the other hand, the discovery of decorated panels with engravings and paintings, recovered from Upper Solutrean sediments, allow us to place these representations, with great precision, in the Medium Solutrean (level VI) that must be placed between the GI 5 and the GI 3. The excavation of the area that we call the microstratigraphy, included in level II (Final Upper Solutrean or Evolved Upper Solutrean), has provided 21 double layers of hearts, one of them with a stone structure and thousands of very typical solutrean flint implements that include the characteristic barbed and tanged points, shouldered points and leaf points. The investigations carried out in the Laboratorio de Estudios Paleolíticos with these arrowheads, show that they must be thrown with a bow. Key words: Solutrean, rock shelter, radiocarbon dates, southeastern Iberian Peninsula. 1. Introduction When Breuil was digging in 1911 with Motos on a remote site in the southeast of the Iberian * Figure. 1 Location Map and picture of The Cueva de Ambrosio (Almería, Spain) Peninsula, he could never have imagined that almost 100 years later the levels of occupation that fill this huge cavity would still not be clear. The work carried out by E. Ripoll in the 1960´s could not complete the stratigraphical sequence either, and the long excavation campaigns run by S. Ripoll from 1981 to the present, could not achieve this objective (real campaigns for digging: 1981,1982, 1983, 1986, 1990 1992, 1994 and 2002). Laboratorio de Estudios Paleolíticos. Departamento de Prehistoria y Arqueología. Universidad Nacional de Educación a Distancia. Senda del Rey, 7. 28040 Madrid (Spain) sripoll@geo.uned. es, fjmunoz@geo.uned.es 435
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Sergio Ripoll L  pez, Francisco J. Mu  oz ...
436 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 2. Geo-Archaeology The site is placed in the southeast edge of the external areas of the Betic Mountains (Fig1). The neogen sediments where the cave opens are limestone rocks of the Upper Burdigalian-Lower Langhian, made by biomicritasalgales rich in fossil remains, sometimes loamy, that integrate redeposit rests of the close mentioned materials; they surface in east to west direction belts and they are affected by post-mantle tectonics, although they are not raid by any other mode. The Cueva de Ambrosio is located at the head of the Moral stream, on a peak 1,060 m above sea level. The mentioned limestone rocks are affected by a deep fracturing which conditions their geometry and the processes acting over them. The orientation is S-SW and the dimensions are a maximum of 15m high in the entrance and 17 m to the inside part of the shield´s edge. 3. Chronostratigraphy The stratigraphy sequence of the Pleistocene deposits are composed by two clear lithostratigraphic units (Jordá and Carral, 1988). The lower one is made by thin-sterile materials produced by mud-flow from the inside of the karst that constitutes the beginning of the cavity´s sedimentation without reaching the rock´s substrate. The upper one is made by high energy deposits resulting from gelifraction processes with insertions of deposits mainly from human activity related to the Solutrean occupational levels of the shelter – Ambrosio II: Upper Evolved Solutrean; Ambrosio IV: Upper Solutrean; Ambrosio VI: Medium Solutrean (Fig.2). This unity presents an alternation of high energy sands and conglomerates in the roof, and ends in a cemented fracture which filled the sequence until a subsequent breakdown occurred. In order to place Ambrosio´s sequence in the cronoestratigraphic scale, the radiocarbon dating obtained until this moment have been calibrated to sigma 2 probably 95%– through the calibration curve calPal 2007 Hulu, included in the program calPal (March Version 2007) (Weninger et al., 2007). Also, to place it in the cronoestratigraphic and archaeological context of the Upper Pleistocene in the Southern Peninsula, we have integrated our dates in the set of radiocarbon dating coming from other sites with a similar chronology (between 27000 and 19000 years BP). We have considered 30 valid dating subjected to Figure 2. Statigraphic profile view with the three Solutrean levels during the 1986 campaign. the calibration through the calPal 2007 Hulu curve (Weninger et al., 2007). The taken dates come from the southeastern Mediterranean sites in the Peninsula (Malladetes, Parpalló, Cova Beneito, Ratlla del Bubo, Cendres) (Villaverde et al., 1998, 1999; Villaverde, 2001), southern Mediterranean (Nerja and La Pileta) (Aura et al., 2006; Sanchidrián et al., 2001) and Portuguese Atlantic (Buraca Grande, Vale Boi, Caldeirao, Lagar Velho, Salemas, Lapa da Rainha and Vale Almoinha) (Bicho, 2004; Pettit et al., 2002). The problem arises when comparing the three first dates of the Ambrosio sequence achieved by the usual procedure of the 14C (Ripoll, 1988) with the eight recently obtained, not only the conventional ones (3) but also the AMS ones (5), five of them already published (Ripoll, 2006) and other four unpublished. The first ones are markedly more recent than the second ones, as can be seen in schedule 1.
436  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  2. G...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Level Culture Sample Procedure Cod. 14 C (BP) Data) cal. BP Data (2 s; 95% prob.) Ref. Ambrosio II (SSE) Charcoal Conventional Gif-7276 16500 ± 280 20500 –19140 calBP Ripoll,1988 Ambrosio IV SS Charcoal Conventional Gif-7275 16620 ± 280 20540 –19260 calBP Ripoll,1988 Ambrosio VI SM Charcoal Conventional Gif-7277 16590 ± 1400 23180 –17020 calBP Ripoll, 1988 Ambrosio II Generic (SSE) Charcoal Conventional Gif-9883 19250 ± 70 23570 –22490 calBP Ripoll, 2006 Ambrosio II Layer 1 (SSE) Charcoal AMS GifA-95576? 20150 ± 200 24550 –23550 calBP Ripoll, 2006 Ambrosio II Layer 1 (SSE) Charcoal AMS GifA-95577 19950 ± 210 24320 –23400 calBP Ripoll, 2006 Ambrosio II Layer 2 (SSE) Charcoal AMS GifA-A-II.2 19170 ± 190 23630 –22310 calBP Ripoll, unpublished Ambrosio II Layer 4 (SSE) Charcoal Conventional Gif-A-II.4 19110 ± 90 23450 –22330 calBP Ripoll, unpublished Ambrosio II Layer 6 (SSE) Charcoal AMS GifA-A-II.6 19300 ± 190 23680 –22440 calBP Ripoll unpublished Ambrosio II Layer 9 (SSE) Charcoal AMS GifA-A-II.9 13740 ± 140 ANÓMALA Ripoll unpublished Ambrosio IV SS Charcoal Conventional Gif-9884 21520 ± 120 26270 –25230 calBP Ripoll, 2006 Table 1. Radiocarbon dating. The most ancient date of the three first ones (Gif-7277) presents a clear defect of precision, so it should be excluded in a strict analysis; as for the other two, they are consistent with each other, although they do not match with those obtained recently. When every date has been found in the same laboratory with a difference of almost 20 years, we finally choose the recent ones as they Figure 3. Stratigraphic profile with specific dating of every level. 437
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Level  Culture  Sample  Procedure  Cod.  1...
438 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD come from a clearly defined and registered archaeological context (Fig. 3.). Among the eight dates recently obtained, one of them (GifA-A-II.9) is clearly anomalous inside the sequence, so it will not be taken into account. The other six dates of the level Ambrosio II were obtained in a microstratigraphy constituted by an overlapping of prehistoric bonfires, and among them soft reversals can be seen, probably due to the nature of the dated coals, that, unfortunately, have not been subjected to an anthracology study before being dated. However, they present a good cluster; therefore they can be considered as a whole dating the Ambrosio´s Solutrean Upper Evolved. The date of the level Ambrosio IV offers no problem at all, and allows a precise dating of the Upper Solutrean. Unfortunately, in this new dating-series we do not have any of lower level, Ambrosio IV, that contains materials of the Medium Solutrean. The level Ambrosio IV (Upper Solutrean) is located in the beginnings of the cold stage Greenl and Stadial 3 GS 3) (Björk et al., 1998), in a moment in which the temperature of the Alboran´s Sea fluctuated between 11º C and 14º C (Cacho et al., 1999, 2001), during the first half of the EH 2. At a regional scale, the level Ambrosio IV(Fortea and Jordá, 1976) is placed between the date of the Upper Solutrean of Malladeted and the most ancient of the Medium Solutrean at Vestíbulo de la Cueva de Nerja (NV. 9). The level Ambrosio II (Upper Evolved Solutrean) develops during the second half of the GS3 and the event of Heinrich 2 (H 2), a moment in which the lowest temperatures of the Alboran´s Sea take place during the Final Upper Pleistocene, around 10º C, and ends during the warm inter-stage of the Greenland Interstadial 2 (GI 2) (Björk et al., 1998), with Alboran´s sea temperatures of 12º C (Cacho et al., 1999, 2001). Regionally, this level is placed in a fork defined by the dating of the Solutrean artistic expressions at Cuevas de Nerja and La Pileta (Sanchidrián et al., 2001), and the dating of the Upper Solutrean at Nerja in the southern Mediterranean sea (Aura et al., 2006) and other Portuguese (Bicho, 2004; Pettit et al., 2002) and Valencian sites (Villaverde et al., 1998, 1999; Villaverde, 2001). 4. Mediterranean shouldered points Both Cueva de Ambrosio and Cova Parpalló articulate the industrial sequence of the Extra Cantabrian Solutrean, as they are the only sites with plenty of lithic assemblage and especially Solutrean points (Muñoz, 2000). In the Medium Solutrean (Level IV) the Solutrean tools are marked by the consolidation of the bifacial knapping, prevailing novaculite laurel above the points of flat-side that still maintain an important weight. At the end of this period, a tendency to peduncle in novaculite laurel can be seen; their morphologies adopt a typology of transition between the novaculite laurel subtype “H” fins points (Smith, 1966) and peduncles. Bifacial Solutrean pieces are detected in the archaeological register for the first time. They are drafts of novaculite laurel whose fabrication process would not be completed. In the Upper Solutrean (Level IV), flat-side points are still present, although in an occasional way. The novaculite laurel, which maintain their importance in the whole tooling of plane retouch, get smaller and thinner and sometimes have a rectilinear edge. In the Solutrean arrow points, the idea of peduncles technical is settled, timidly tested at the end of the former stage. Thus, fins points and peduncles become the distinctive tooling of this period, after the firsts drafts made at the end of the Medium Solutrean. The last type of arrow point representing this period is the Mediterranean notch point, with increasing importance on the whole of arrow points. In the Upper Evolved Solutrean (Level II), the plane-invasive retouch loses its preeminence in favour of the abrupt retouch. The novaculite laurel are still present, although their number is significantly reduced. Although the fins points and peduncles are still important, the most remarkable tooling in all this period is the Mediterranean notch point, outnumbering any other arrow point of the Solutrean group. In addition, the tooling on little leaves presents a remarkable advance in regard to the former period. The functional and experimental research done on fins points and peduncles and notch points of abrupt retouch show their use as arrow points, which means that the appearance of the bow and arrow technique could be traced back to the beginning of the Upper Solutrean in the ex-Cantabrian region of the Iberian Peninsula (Muñoz 2000; Muñoz and Márquez 2006; Muñoz et al., 2012), (Fig. 4).
438  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  come...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Figure 4. Upper Solutrean and Upper-Evolved Solutrean Points. A: Fins points and peduncles and fastening proposal in arrow shaft. B: Abrupt retouch of notch points and fastening proposal in arrow shaft. 5. Taphonomy According to traditional interpretations assuming that all the taxa related to lithic industries appearing in an archaeological site are the consequence of human intervention, the marrow accumulation of Ambrosio could be considered an example of an archaeological site for specialized hunting of lagomorphs, as they constitute more than 90% of the recognized individuals. However, the taphonomic research shows that they had a varied contribution made by birds, carnivores and humans. A small number of lagomorphs died of natural causes due to the falling of blocks from the shelter´s roof. Considering all these circumstances and calculating the amount of meat they could bring to the site, Ambrosio was actually not a hunting lodge specialized in rabbit hunting, because other animals like the horse, the goat, or the deer provided more meat. The analysis shows that all the ungulates, as well as foxes, were meat processed by humans. After these were taken apart, carnivores would occupy the shelter scavenging offal of what was left. This shows that human occupation was not continious, so in certain moments the site was abandoned, facilitating carnivores benefiting from the waste they left. Among the ungulates there is a diversified strategy over the goat, the horse and the deer. However, the analysis of seasonality has observed different hunting strategies. A diversified hunting exists of the deer, the horse and the goat in the most benign seasons for the Medium and Upper Solutrean, and other more specialized hunting of the goat in the hardest moments of winter. The duality of these two hunting strategies seems to be conditioned by resource availability. The goat would be the only animal hunted in winter because its adaptability to a variety of means allowed it to live in an immediate environment to Ambrosio during the hardest weather conditions of the year. On the other hand, the deer and the horse went to lower valleys in winter, and during the most benign seasons they climbed to Ambrosio´s environment to take advantage of the mountain pastures. 439
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Figure 4. Upper Solutrean and Upper-Evolve...
440 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 5. View of the set of panel II in which we have identified a total of 26 engraved and painted figures 6. Prehistoric Rock Art The archaeological site gave us a great surprise on 10 September 1992, when we found the first carved figures of Ambrosio cleaning a smooth surface where the reference point 0 is placed. Up to now, we have identified 35 representations engraved and/or painted, mainly zoomorphous including equidae, although there are also a bird, a bovine and some idiomorphs (Fig. 5). However, the remarkable figures are the incredible horse painted in red ochre, left-oriented and of a length of 92cm, and an immensely realistic, engraved and painted human face. Nobody could imagine that in a shelter in the southeast of the Iberian Peninsula parietal representations could exist, and besides that, they were covered with sediments from the different archaeological levels. Rare are the seasons in which we find parietal representations covered by archaeological levels that allow a precise dating, and this is the case of Ambrosio. Although panel I is today outdoors, at the time it was covered by the intact levels placed a few centimeters to the left, which were removed by uncontrolled diggers, as well as the natural breakdown of the trench cuts opened by E. Ripoll in the 1960´s (Fig. 6). All the levels filling the shelter reached a power of 4.97 meters, and would be covering all the embellished surfaces. Panel I-A would be covered by Fini Pleistocene levels on its upper level, as well as by the strata I (sterile), II Upper-Evolved So- Figure 6. All the representations of Ambrosio are covered by archaeological levels. Its position, perfectly established, allows dating the set with great precision in two different cultural horizons. Panel IA is located in the Upper Solutrean, while panel IB and panel II, where the figure subject of this text is placed, match up to Medium Solutrean lutrean with an updated dating that ranges from 19250 and 20150 BP and III (sterile), being made from the occupational ground corresponding to level IV, or Upper Solutrean, with a new dating of 21520 ± 120 BP. On the other side, panels I-B, II and III, which have a lower position respecting the former descriptions, would be covered by level IV, Upper Solutrean and by level V (sterile), and would have been made from the cultural horizon of the Medium Solutrean, which is much more ancient than the human level formerly described and that has a radiocarbonic dating of 23180 calBP. In a region where Paleolithic paintings are really poor, even non-existing, these representations are surprisingly classic. The Cueva de Ambrosio is one of the few places with absolutely dated parietal Paleolithic art in the Iberian Peninsula. It is also located in the Mediterranean area where the placards collection of Parpalló (Valencia) (Pericot 1942; Villaverde 1994) are always mentioned, and we can also find outdoor figures as the place is a shelter, not a deep cave without natural lighting. These parietal figures offer great interest due to the importance, the artistic quality and the location in the southeast of Spain, and are certainly higher than the ones presented in the few sets of Paleolithic Art of the Mediterranean area.
440  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS The discovery of these figures fills the vacuum existing in this area in the geographic dispersion of quaternary-parietal art in the Iberian Peninsula, only represented by hammering equidae of Paleolithic style in Piedras Blancas (Almería) (Martínez 1986/87). In Europe there are only four Paleolithic sites with the characteristics of an absolute dating, as the representations are covered by archaeological levels, and one of these is the cave of La Viña (Asturias), where some naturalistic representations covered by Medium-Evolved-Cantabrian Magdaleniense levels were found (Fortea 1981 and 1990). In the year 2001 the discovery of several parietal representations in the Parpalló (Valencia) was announced (Beltrán 2002). Although nowadays these figures are at a great height, respecting to the current ground, at the time they were covered by the archaeological levels dug by Professor L. Pericot in the 1930´s (Pericot 1942); these are an equidae, a goat and several disjointed lines, as well as a quadruped in red ochre located above the area where Pericot´s Medium and Upper Solutrean were placed. In the cave of El Mirón (Cantabria), some non-figurative representations have been found, covered by the archaeological levels of MediumCantabrian Magdaleniense (García et al., 2012). In the grotto of La Tête du Lion in France (Francia) (Combier 1972 and 1977), pictorial representations were not properly covered by the strata, but the systematic excavation carried out in the paintings base provided the tooling, “pencils” and charcoals used to make them allowed the dating. From this moment on, the figures set found in Ambrosio will be added. The frequent discoveries of Prehistoric art stations outdoors or in archaeological sites, like in the present case, will introduce important changes on the ideas generally assumed referred to the geographical distribution, both for Paleolithic and postPaleolithic art. The imposed schedules by important investigators have provoked these areas to be taken into account as exceptions contradicting objective pragmatisms, easily handling, and that marginally incorporated too many updates which later remained outdated because of the non-resolution of the real problem. Abnormal assumptions were accepted and increased, forcing us to make a revision of these problems admitted up to now as undisputed. These issues must be deeply addressed in scientific meet- ings that shed light on the distribution and dating of Peninsular Paleolithic art. 7. The Solutrean in the south of the Iberian Peninsula The Solutrean in the Iberian Peninsula has a great industrial polymorphism that is reflected in the remarkable differences between the Cantabrian and the ex-Cantabrian areas. These two areas would be limited by an imaginary axis going from Portugal to the southeast of France via Madrid. The main reasons to mark the existing differences are essentially industrial and typological, as in other aspects, like the artistic, they offer many common points. These dissenting points arise to a greater extent at the end of the Solutrean, with the existence or the absence of the plane retouch over notch’s points, as well as the existence or the absence of fins points and peduncles which start to appear in the cultural period of the Upper Solutrean, although at Parpalló they appeared in the Medium Solutrean. Inside the extra-Cantabrian area, the Catalan Solutrean in the archaeological sites of L’Arbreda, ReclauViver, El Cau de les Goges, Davant Pau, etc., offers a certain originality, due to the presence of the plane retouch over notch points as well as other points with an incipient peduncle; together with other features, it could be considered different from the rest of the Spanish Mediterranean, but at the same time it is difficult to “accommodate it into the classic sequence” (Soler, 1986 ). The fauna in almost every archaeological site is mainly dominated by lagomorphs (Oryctolagus cuniculus), related to goats, deer and equidae, as well as other mammals; the proportion of this composition varies in order of the environment of the seasons. . Traditionally, the Solutrean has been considered one of the best known periods of the cultural sequence of the Spanish Mediterranean Paleolithic. However, while it is true that there are lots of archaeological sites with some integrated pieces of the Solutrean group, just a few have an enough representative industrial series to precise the exact moment of the occupation. This is the case of the Cave El Parpalló in Valencia and Cueva de Ambrosio in Almería. Only a few sites have a complete and updated study including all the range of analysis that allow fixing its chronostatigraphic position. 441
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  The discovery of these    gures    lls the...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 442 J. Emili Aura Tortosa*, Jesús F. Jordá Pardo** Cueva de Nerja (Maro, Nerja, Málaga) Introduction Cueva de Nerja is a large cave with thick archaeological deposits in its outer chambers: Vestíbulo Hall, Mina Hall and Torca Hall. It also contains an important ensemble of Palaeolithic and Neolithic art in its inner chambers and varied documentation of funerary practices attributed to the Palaeolithic, Epi-Palaeolithic and Neolithic periods. Located in the extreme south-west of Europe, on the coast of the Alborán Sea, this region has yielded some of the oldest evidence of the use of marine resources by humans. The mutual visibility of southern Europe and North Africa has led to proposals about the existence of relationships between both shores during the Palaeolithic, from simple formal and technical convergences. The data used in this description come from the project directed by Professor F. Jordá Cerdá between 1979 and 1987 about the Palaeolithic and Epi-Palaeolithic in Cueva de Nerja (Aura et al., 2002, 2010). The bibliography lists the most important texts and recent articles. Stratigraphy and chronology The sedimentary sequence in the deposit, obtained by correlating the lithostratigraphic and archaeological sequences in Mina (NM) and Vestíbulo (NV) Halls, consists of twelve stages of sedimentation and erosion corresponding to seven litho-stratigraphic units separated by five stratigraphic discontinuities (Jordá and Aura, 2009) (Fig. 1). These litho-stratigraphic units have been precisely placed on the Upper Pleistocene chronostratigraphic scale by a series of radiocarbon determinations calibrated with the curve CalPal 2007 HULU, using the CalPal software (2013 version). From top to bottom, the chronostratigraphic sequence is as follows (Jordá and Aura, 2009): – Nerja Stage 1 (Unit 1, Levels NV13, NV12 and NV11). Located in NV, this unit overlies a thick speleothem and comprises autochthonous detritic sedimentation which took place between 29,940 and 28,480 cal BP, at the end of the OIS 3a, at a time that coincided with the end of Heinrich event 3 (H3), which would cover the GS 5and GI 4. It includes Gravettian industrial remains. In its base level (NV13, coprolites attributed to Crocuta crocuta spelaea indicate the absence of humans in the cave at the start of the sedimentary record (Arribas et al., 2004). This unit may tentatively be associated with at least a part of the basal levels in NM (NM19, NM18 and NM17). – Nerja Stage 2: A stratigraphic hiatus detect- ed in NV and estimated to have lasted about 3000 years, caused by an erosive process related to GI3 and the start of GS3. – Nerja Stage 3: (Unit 2, Levels NV10, NV9 and NV8). Detected in NV, these are deposits with characteristics of a cold but not extreme climate, dated from 25,6 to 20 cal BP. The sedimentation in the two lower levels (NV10 and NV9) is associated with GS3. An erosive contact has been recognised between NV9 and NV8, and the formation of NV8 also took place in a slightly cold and wet climate, which became drier towards the top of the level. This internal hiatus in Unit 2 may be related to the GI2, while its upper section (NV8) can be linked to the GS2c, in the Last Glacial Maximum. The archaeological materials are Solutrean. – Nerja Stage 4: A stratigraphic hiatus caused by erosive processes together with a possible absence of sedimentation, which led to the non-existence of as much as 4000 years in the record in NV. It corresponds to the cold pulse at the start of the GS2a. * Departament de Prehistòria i Arqueologia, Universitat de València, Avda. Blasco Ibañez 28, E-46010 València, Spain. ** Laboratorio de Estudios Paleolíticos, Dpto. de Prehistoria y Arqueología, Facultad de Geografía e Historia, Universidad Nacional de Educación a Distancia, Paseo Senda del Rey 7, E-28040 Madrid, Spain. (jjorda@geo.uned.es) (jeaura@uv.es)
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  442  J. E...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Figure 1. Cueva de Nerja. Chronostratigraphy, lithostratigraphy and archaeological stratigraphy in the sedimentary sequences in the Vestibule and Mine Halls. 443
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Figure 1. Cueva de Nerja. Chronostratigrap...
444 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD – Nerja Stage 5: (Unit 3, NV7, NV6 and NV5; NM16, NM15 and NM14). Detected in NV, NM and NT, this unit is dated between 14,820 and 13,600 cal BP, coinciding with the GI or Late Glacial period, with mild climate conditions. Level NM15, characterised by a large accumulation of cryoclasts reflects one of the coldest moments in the sequence at Nerja and might be correlated with the GI1b. Nerja Stage 5 finished with the sedimentation in NM14 and NV5, with temperate characteristics, during the GI1a. This unit contains a Mediterranean upper Magdalenian occupation. – Nerja Stage 6: A fluvial erosive phase which removed 600 years from the record, between GI1a and GS1. – Nerja Stage 7 (Unit 4, NV4, NM13 and NM12). It is characterised by a large accumulation of remains of Mytilus edulis forming an anthropic shell midden in NV (Jordá et al., 2011a). This unit developed from 12,950 to 11,360 cal BP, coinciding with the GS1 and the beginning of the OIS1, in the early Holocene. It contains remains of an Epi-Magdalenian occupation. – Nerja Stage 8: An erosive phase responsible for a stratigraphic hiatus that lasted for a little less than 3000 years in the lower Holocene. – Nerja Stage 9: (Unit 5, NV3, NV2 and NV1; NM11 to NM7). The base of this unit has yielded a few remains of the geometric Mesolithic and is located in the time between 8530 and 7920 cal BP, in the climate optimal in the mid-Holocene, at the base of the Atlantic chrono-zone. The middle and top section of the unit corresponds to levels with early and middle Neolithic materials, dated between 7490 and 6820 cal BP, in the middle-late part of the Atlantic chrono-zone. – Nerja Stage 10: An erosive phase detected in NM. – Nerja Stage 11: (Unit 6, NM6 to NM2). Deposits with late Neolithic materials, dated between 6910 and 5050 cal BP, in the last part of the Atlantic chrono-zone and the start of the Sub-Boreal, in the mid-Holocene. Chalcolithic levels are found at the top of this unit, which in Torca Hall is dated to 4820-3600 cal BP (Sub-Boreal). – Nerja Stage 12: (Unit 7) Breccia and speleothems formed in the late Sub-Boreal and early Sub-Atlantic, during a temperate pulse in the upper Holocene. Palaeogeography, palaeotemperature and palaeoenvironment The study of the palaeogeographic evolution of the area around Cueva de Nerja reveals marine transgression in which several episodes can be identified in connection with the human occupations inside the cave (Jordá et al., 2011b). During the GS2b (Between 20 and 19 ky cal BP), in the Last Glacial Maximum, the coastline was at –120m, which left a coastal strip uncovered extending over 4.5 km from the modern shore. After this time, the general warming and ice melt caused sea level to rise. This, in the GS2a, coinciding with the H1 (between 16.5 and 16 ky cal BP), the coastline was at –90 m, and 3.5 km from the modern position, whereas in the GI1 and GS1 (between 14 and 12 ky cal BP) the sea was at –70 m and 3 km away. The rise in level continued until it stabilised at –50m in the Pre-boreal and Boreal (from 11 to 8.5 ky cal BP), with the sea 1km away from the modern shore, and it rose again in the Atlantic and Sub-boreal with the sea level at –20 m and the shore 400m away from its current position. A correlation has been proposed between the high resolution curve of the Sea Surface Temperature (SST) provided by the MD95-2043 core in the Alboran Sea (Cacho et al., 2001) located to the south-east of Nerja with the different stages of sedimentation recorded inside Cueva de Nerja (Jordá et al., 2011b): Nerja 1, with a SST between 10 and 14ºC (last cold episode in the OIS 3a); Nerja 3, with a SST between 12 and 13ºC (GS3, GI2 and GS2c); Nerja 5, with SST between 12 and 14ºC (Late glacial or GI1); Nerja 7, with a SST reaching a minimum of 12ºC (GI1 or Younger Dryas); Nerja 9, with a SST between 18 and 20ºC (Atlantic); Nerja 11, with a SST between 18.5 and 19.5ºC (Atlantic and Sub-Boreal); and Nerja 12, with SST about 18 or 19ºC (Sub-Boreal/Sub-Atlantic). The anthracological study for NV was able to propose the evolution in the palaeovegetation around the cave (Aura et al., 2010). In the late OIS3 (Nerja 1), charcoals of Pinus nigra and P. pinea predominate, together with ligneous Fabaceae and several kinds of shrubs like juniper and rockrose. The area where the two pine species meet
444  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD      ...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS is between the Meso-Mediterranean and SupraMediterranean bio-climatic zones, in areas with a mean annual temperature of between 8 and 15ºC, which reflects the conditions around Nerja in the Gravettian, with a regime of low precipitation: an annual mean of about 400-600mm. In the Last Glacial Maximum (GS3 to GS2), the Solutrean levels (Nerja 3), the same vegetation community existed with a tendency towards a reduction in black pine and an expansion in Fabaceae and other shrubs and low plants, with a similar proportion of stone pine to the previous period. In this way, the coastal strip around Nerja displayed Meso-Mediterranean conditions, although the black pine at higher altitudes indicates a colder vegetation type. In the Late Glacial period or GI1 (Nerja 5), the anthracological studies indicate an improvement in the climate with a decrease in black pine, and the presence of the Aleppo pine and Quercus. It was an open shrub vegetation, dominated by ligneous Fabaceae and with stone pine located probably in the coastal strip. During the Epi-Magdalenian occupation (Nerja 7) in the Younger Dryas and early Holocene, the species diversified, with the appearance of warmth-loving trees such as Olea europaea, mastic, the strawberry tree, box and flax-leaved daphne among the shrubs, while the Fabaceae are still well-represented, black pine continues to decline and a decrease in stone pine is also noted. These typical conditions of the thermo-Mediterranean bio-climatic zone appear to establish themselves on the coast around Nerja from the Epi-Magdalenian onwards. The improved climate of the Late Glacial period continued in the emerged areas during the GS1, while the sea water maintained its stadial conditions, allowing the entry of cold-tolerating Atlantic species. The archaeozoological study also provides palaeoclimate and palaeogeographic data for the late Pleistocene and early Holocene (Aura et al., 2010). Its results indicate that aurochs, horses and chamois were marginal species, while ibex continued to expand throughout the sequence, only surpassed in the number of remains by rabbit, most of which had an anthropic origin. Red deer was relatively important in the Pleniglacial, while wild boar is found from the Late Glacial period onwards. At the base of the sequence, coprolites of Crocuta crocuta spelaea are associated with osseous remains of Equus sp., Bos sp. and red deer (Arribas et al., 2004). Small carnivores (lynx, wild cat, fox and possibly wolf) are not common and appear throughout the sequence. In the base unit was also found a single remain of Testudo hermanni identified in NV, which can be added to the turtles (Emys orbicularis and Mauremys leprosa) identified in NM during the Pleistocene-Holocene transition (Morales and Sanchís, 2009). The earliest evidence of marine mammals is found at the base of Unit 2: two remains of Phoca vitulina. During the Late Glacial period (Nerja 5 and 7) the monk seal has been identified, as well as dolphins (Delphinidae and Delphinus delphi). In NM, the remains of two whale barnacles have been found. Their host was Eubalena australis, and the entry of this whale in the Mediterranean poses palaeo-climate issues (Álvarez et al., 2014). Nerja contains one of the most complete invertebrate, bird and marine fish records in the south of Europe (Aura et al., 2002): thousands of remains of molluscs, fish, echinoderms, crustaceans, cephalopods and sea birds. The data provided by NV allows an analysis of trends over time. This in Nerja Stages 1 and 3, marine molluscs make up less than 20% of the total, and continental gastropods dominate in all the assemblages, coinciding with the practical absence of fish, echinoderms and sea birds. At the top of Nerja 3 (upper Solutrean) this trend is inverted, so that after this time marine molluscs predominate together with an extraordinary number of fish remains, in both Nerja 5 and Nerja 7 units. During the Last Glacial Maximum occupations, several mollusc species indicating cold conditions have been classified (Pecten maximus, Nucella lapillus, Littorina obtusata, etc.). In the Late Glacial levels, Pinguinus impennis has been identified, as well as guillemot and the little auk. These coincide with fish species that now have a northern distribution (Melanogrammus aeglefinus and Pollachius pollachius) (Rodrigo, 1991). Techno-economic transformations Changes in the surroundings of Cueva de Nerja were closely tied to the impact of eustatic oscillations around the continent. Their variations conditioned the size of the emerged land and movements along the main route of communication on the Mediterranean seaboard of Iberia: the coastal corridor. These palaeogeographic changes affected the possible procurement of resources and lithic raw material, as reflected by the composition of the tools and the palaeo-faunal and palaeo-botanical assemblages (Aura et al., 2010). Thus, the variations in the density of remains, in the technology and ty- 445
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  is between the Meso-Mediterranean and Supr...
446 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD pometry of lithic implements, and in the terrestrial and marine origin of the biotic remains attest aspects of the site history, and also of the distribution and economy of the Palaeolithic hunter groups who lived in one of the southernmost parts of Europe. The techno-economic data provided by Cueva de Nerja can be arranged in three main cycles, with some internal details: – The first corresponds to the full Glacial occupations, dated in NV to between 30 and 20 ky cal BP. Their main common trait is a careful blade production (select raw materials and mean length of 50-70mm) obtained from prismatic and flat cores with unipolar reduction series on their widest face. The presence of unweathered cortex indicates the raw material was procured in a primary position, at a distance of over 30 km from the site, and certain diversity has been seen in the flint groups, while jasper and rock crystal reach their highest percentages. After this first characterisation, the descriptors used in Fig. 2 indicate that the density of materials, fauna and tools are able to discriminate two episodes corresponding to Units 1 and 2. • The Gravettian occupations (Unit 1) have yielded the lowest density of lithic objects (150-800 per cubic metre of sediment excavated), basically documenting the phases of use and abandonment (Fig. 2). The lithic projectiles are limited to narrow points, backed in some cases, and a few bone points, found in low densities (3 points/m3). Among the adornments, some gastropod species now have an Atlantic distribution area (cf. Littorina). The taphonomic study of the fauna has identified a carnivore occupation at the base of Unit 1, but the presence of remains with an anthropic origin from NV13 onwards. The most outstanding trait of these assemblages is the mean percentage of lagomorph remains and the high proportion of red deer, which has been linked to the size of the coastal strip. Marine fauna is scarce, limited to a few remains of fish and molluscs, in comparison with the large number of terrestrial gastropods. • The Solutrean occupation (Unit 2) has yield- ed a greater density of lithic remains (9001200 remains/m3) in which all the phases have been documented, from manufacture to use and repair. Blade production is still important, but from NV9 onwards Solutrean reduction methods and heat treatment increase significantly, coinciding with the pres- ence of leaf-shaped points and pieces with flat retouching. In the upper section (NV8 and NV8/s) a decrease in blade production is noted and an increase in backed micro-blade tools, obtained from differentiated productions. The increase in the number of broken and fire-altered projectiles is considerable, made both from lithics (uni and bifacial leafshaped points) and from bone and antler, mainly points with a rounded base (4 points/ m3). The objects of adornment are still made mostly from gastropods. Finally, terrestrial fauna displays an increase in ibex and lagomorphs, and a smaller diversity of species that is partly compensated by the identification of four remains of seals (Phoca vitulina and Monachus monachus). At the top of Unit 2, coinciding with the end of the Last Glacial Maximum, the percentage of marine fauna and pine nuts increases (Badal, 2001). • Although a hiatus is found between both episodes, the techno-morphological and typological traits indicate significant continuity between the Gravettian (NV13-NV11) and the so-called Solutrean A (NV10), and objects similar to those in Vale Comprido have been identified (Fig. 3). – The second of the cycles corresponds to the late Glacial occupations, dated between the end of the H1 event and the Younger Dryas (15-11.5 ky cal BP). These are occupations conserved in the three outer chambers in the cave (Vestíbulo, Mina and Torca), and have yielded a greater density of lithic remains (1200-1800 remains/m3) and diversity of resources exploited (Fig. 2). One common element in these occupations is the manufacture of lithic tools from flint cobbles collected in a secondary position, above all on beaches, according to their shape. Polished cortex is seen on 20% of the knapped lithic materials and many of these cortical fragments are altered thermally. Several thick-grained rocks (limestone, quartzite and metamorphic stone), obtained from cobblestones, were used as well as flint, and macro-lithic tools were made with them. The reduction systems were aimed at obtaining blades and bladelets (20-35 mm mean length) and the different productions can be associated with the flint quality. The best quality raw materials were used for the most regular blade products, using a soft hammer, whereas a hard hammer was mostly used for more robust and irregular blades and bladelets, in
446  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  pome...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Figure 2. Cueva de Nerja. Summary of bio-archaeological and techno-economic data in the Upper Palaeolithic, EpiPalaeolithic and Mesolithic occupations. raw materials of a poorer quality. Most of the microblades are regular, some made from a flake blank, while others, according to their cross-section and profile, could have been extracted from burin-cores and nucleiform endscrapers. A production of short and elongated flakes has also been identified, using cores with a natural back and a uni and bi-polar reduction system. Many of the Magdalenian domestic tools (endscrapers, retouched pieces, notches...) are products shaped from blade cores. The osseous assemblages in these occupations display greater variability and also density (from 6 to 40 pieces/m3, depending on the sector and level). They are dominated by pointed objects, especially double thin and short points, thought to be gorges or straight fish-hooks. The five harpoons and over half the needles come from the Magdalenian levels, while the flat points, the chisels and remaining needles come from the Epi-Magdalenian occupations. Manufacturing waste indicates that 447
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Figure 2. Cueva de Nerja. Summary of bio-a...
448 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 3. Cueva de Nerja. 1-9 Solutrean lithic industry; 10, 13-14, Gravettian lithic industry; 11, pierced and broken Lynx pardina canine tooth; 12: cirripede carina (Pollicipes pollicipes) with suspension notches (1-9: NV10-NV8; 10-12: NV13NV11).
448  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Figure 4. Cueva de Nerja. 1-2, knapped cobbles from the shell-midden; 3-4, Epi-Magdalenian bone industry; 5, cobble stone with geometric engravings; 6-9 upper Magdalenian bone industry (1,2 and 3: NV4; 4: NM13; 5: NM16: 6-9: NV7NV5). Top right: remains of marine fish from level NV4. 449
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Figure 4. Cueva de Nerja. 1-2, knapped cob...
450 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD bone was used to make the thin points and needles, whereas antler was used for the points, most of the harpoons and the chisels (Fig. 4). large number of monk seal remains (Pérez and Raga, 1998) is only second in importance to ibex and is much larger than the number of red deer remains. The changes in the toolkit are accompanied by a clear reorientation towards exploiting the marine environment. In addition, the differences in the marine species that are seen between the chambers also raise questions about the duration and seasonality of the occupations. Contrasting with the diversity in the marine species, the capture of land mammals was still directed towards ibex, with the structural complement of rabbit. – After another hiatus, the third cycle corresponds to the middle Holocene and its main justification is its date within the Atlantic chrono-zone, as it corresponds to two different periods: Late Mesolithic and Neolithic. These are occupations with a low density of lithic remains and an unequal standard of documentation. Within this second cycle, two episodes can be established that unite both sequential and functional elements. The first corresponds to the Mediterranean upper Magdalenian (with harpoons) occupations (15-13.5 ky cal BP: NV7-NV5 and NM16-NM14). A careful blade production, the highest densities of lithic and bone remains in the whole sequence and five harpoon fragments are some of its specific traits. Objects of ornament are sill made from shells and some pierced teeth; above all from Cyclope neritea, C. pellucida and Theodoxus fluviatilis. Marine fauna now becomes more common: molluscs, echinoderms, coastal and estuary fish and marine mammals. These occupations have provided the largest numbers of remains of cetaceans and whale balanoids These indicate a use made of beached animals, whose size would determine whether they were taken to the cave or not: cranial remains and vertebrae in the case of dolphins, portions of skin and meat in the case of the whales. The second episode represents evolution in Magdalenian technical systems and is dated between 13 and 12.5 ky cal BP, at the time of the Younger Dryas. The lithic assemblages reflect an increase in more expedient production and the presence of a significant macro-tool assemblage made from cobbles: hammer-stones, knapped cobbles, grinders and millstones transmit the processing of a wide range of resources and new ways of using the site (Fig. 4). The bone assemblage is restricted to some points, chisels and gorges, while ornments are made from Cyclope neritea, together with Nucella lapillus and Trivia arctica. The marine fauna becomes even more abundant, and one of the oldest shell-middens on the continent accumulated (NV4). Together with the bivalves, the number of fish remains and migratory character of some species marks some differences from the Magdalenian occupations. Among the mammals, the The Mesolithic could not be isolated stratigraphically, as intrusions and Neolithic pits seriously affected it. However, some lithic products (flakes), morphotypes (trapezes, micro-burins and Montbani-type blades) and short-life radiocarbon dates indicate human presence in the cave between 8.5 and 7.9 ky cal BP, although not with the density of the previous periods (500-700 lithic objects/m3). Finally, the Neolithic is dated to between 6.9 and 5 ky cal BP and has yielded a similar density of remains (about 600 remains/m3). Short-life dates indicate a gap of several centuries after the Mesolithic, with the sudden appearance of pottery, domestic species and new lithic reduction systems: pressure knapping associated with the heat treatment of the cores. The osseous industry also exhibits new morphotypes (spoons, tubes, ovicaprine punches) made by fracturing the bones. At the time of this third cycle, Columbella rustica became widely used as an object of adornment on both shores of the Mediterranean. Palaeolithic art and burials Several references exist to Palaeolithic burials in Cueva de Nerja, while in recent Prehistory the cave was repeatedly used as a necropolis. In 1963-64, four “Solutrean” burials were excavated in the Vestíbulo Hall. However, a direct radiocarbon determination for a remain belonging to Individual 1 or A, indicated a Neolithic age. In 1982, an Epi-Palaeolithic burial was excavated in Torca Hall. The individual was identified as a female. In 1984, some phalanges and metatarsals were recovered in NV, for which a tentative age of 18-12 ky BP was proposed, according to which these burials could correspond to Palaeolithic. Apart from these descriptions of articulated remains, “isolated human bones” have been found in Palaeolithic and Epi-Palaeolithic levels, coinciding with a general trend observed in Mediterranean regions after the Late Glacial period. In the deposits in Mina and Vestíbulo Hall, a few portable art objects have been found: plaquettes
450  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  bone...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS and cobblestones with linear incised motifs and a naturalistic representation of a bird, as well as pieces of ochre and tools connected with its processing. This kind of evidence is more common in the Solutrean and Epi-Magdalenian occupations, with which most of the Palaeolithic art inside the cave has been associated. The recent publication by J.L. Sanchidrián of the use of hollows and Pecten sp. valves as possible points of illumination, indicates a close correlation between the occupations in the outer chambers and the use of deep parts of the cave in the full Glacial period. Cecilio Barroso Ruiz *, Anne Marie Moigne **, Miguel Caparrós *****, Vincenzo Celiberti **, José García Solano ***, Guadalupe Monge ****, Antonio Monclova * The Acheulian at Cueva and abrigo del Ángel (Lucena, Córdoba) Introduction Cueva and abrigo del Angel is a Middle to Upper Pleistocene site situated in the south of the Iberian Peninsula, in the municipality of Lucena (Córdoba, Spain). It is located in the Sierra de Aras (also named Sierra de Araceli), at 620 metres a.s.l. Its coordinates are 37º 24’ 22’’ Lat. - 4º 24’ 59’’ Long. The archaeological fieldwork started in 1995. Until 2014, the General Administration of Cultural Heritage from the Junta de Andalucía has approved seven campaigns on this site. Geological setting The lithostratigraphic series of the Sierra de Araceli belong to the External Meridional Subbetic Domain of the Betic Cordilleras, and comprises materials from the Triassic to the latest Quaternary period. Triassic materials are represented by red, green and bluish-purple clays that become masses of gypsum on the top of the sequence. These Triassic materials do not appear at the surface of Araceli massif * The inner chambers are known as the Tourist Galleries and the High Galleries. These two decorated areas are differentiated topographically, but also by the number of motifs and iconographic themes they contain. In the former, most of the simple signs are associated with a few animal figures (horses, ibex and deer). They are pre-Magdalenian in age, possibly beginning in the Gravettian. In the latter galleries, a mono-thematic composition is known as the “Shrine of the Dolphins”. Animals and linear motifs have been described in other parts of the cave too, including a black linear phase, largely of Holocene age. directly, although they probably constitute the base of the stratigraphic series (López Chiclano, 1985). On the contrary, two chronostratigraphic units are well represented both in the Sierra and in the surrounding mountain: a) Mesozoic unit that comprises mainly limestones, dolomites and carbonated marl from the Jurassic-Cretacic period. It is the most relevant relief of the sierra, affected by various faults and diaclases, as the result of which important karstification processes have given rise to the formation of sinkholes and cavities, such as Cueva del Angel. b) Cenozoic unit composed by marl, biocalcarenite and most recent detrital sediments; it has very sharp contact with the former unit. Miocene deposits are composed of more than 100 metres thick of white and green marls and biocalcarenites, whereas the Quaternary period is well represented in vast clay plains resting on the top of the above mentioned materials, corresponding to fluvial, glacis and piedmont deposits. Fundación Instituto de Investigación de Prehistoria y Evolución Humana. Plaza del Coso 21, 14900 Lucena (Córdoba, España) email: barroso.cecilio@gmail.com ** Muséum National d’Histoire Naturelle. Centre Européen de Recherches Préhistoriques de Tautavel. *** Departamento de Prehistoria y Arqueología. Universidad de Granada. **** Departamento Cristalografía, Mineralogía y Q.A. Facultad de Química. Universidad de Sevilla. ***** Département de Préhistorie, Muséum National d’Histoire Naturelle, 75013 París, France. 451
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  and cobblestones with linear incised motif...
452 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Concerning its geomorphological structure, the Mesozoic Unit presents an alternation between soft and hard deposits that give a stepped landform to the northwestern sector of the Sierra de Araceli. The loamy soils are often cultivated and the harder ones support scrubland formations. On the other hand, the Cenozoic unit has a morphology of very eroded domes, also cultivated, that appear as very extensive outcrops (López, 1985). J/K section stratigraphy Transverse stratigraphic profile in J/K zones is the most relevant area discovered so far at the Cueva del Angel. It goes from sector 5 to the middle area of sector 8 (Fig.1) and is 365 cm thick. This profile includes great archaeological and paleontological material that consists mainly of faunal remains, mostly burned, and lithic assemblages, especially worked on flint. Furthermore, secondary calcite precipitations fill any little hollow in the whole profile in a homoge- Figure 1. Stratigraphic profile of J/K zone. neous and accentuated way. These precipitations are typical of the vadose zones of every karst system and come from the alternation of moist and dry periods. In humid periods, water rich in bicarbonate predominates and subsequently is completely filtered by percolation processes along the profile. Whereas in dry periods evaporation prevails and consequently carbonates precipitate again provoking the cementation of the entire stratigraphic profile. Once more these alternations of moist and dry periods conditioned the formation of several generations of speleothem layers that will cover any vacuolar porosity (Monge, 2012). At the J/K profile, twenty stratigraphic levels have been established (Fig.1) based on the observation and description of their particular traits. Levels have been differentiated on the base of the following categories: lithology, colour, texture, consistency, fine-grained structure, coarse fraction proportion, porosity and bioturbation. Also, the presence of anthropic features in each level has been taken into account (Table 1).
452  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Conc...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Porosity Coarse fraction/ finegrained fraction (%) Anthropic features Soft to hard Low 20/80 Scarce Soft to hard Low 1/99 Slightly abundant Stratigraphic levels Depth (cm) Munsell Colour Texture Structure Consistency I 208-250 Pinkish gray (7`5 YR 7/2) Loamy* Blocky II 214-258 Very dark greyish brown (10 YR 3/2) Loamy Granular III 240-300 Dark reddish brown (7`5 YR 4/2; 5 YR 3/2) Loamy Granular Hard High 20/80 Slightly abundant IV 249-307 Pinkish gray (5 YR 6/2) Sandy loam Blocky Loose High 0/100 Scarce V 294-318 Pinkish gray to dark brown (5 YR 6/2; 7`5 YR 4/2) Loamy Platy Soft _ 0/100 Scarce VI 253-355 Dusky red (7`5 YR 3/2; 2`5 YR 3/2) Loamy Granular Soft Low 1/9 Slightly abundant VII 364-381 Red (2`5 YR 5/7) Sandy loam* Granular Soft _ 0/100 Scarce VIII 348-390 Dark brown to dark reddish gray (7`5 YR 3/2; 5 YR 4/2) Sandy loam Granular Soft Low 30/70 Very abundant IX 363-422 Dark brown to dusky red (7`5 YR 3/2; 2´5 YR3/2) Sandy loam Granular Hard Low 50/50 Very abundant X 374-434 Dark reddish brown (7`5 YR 4/2; 5 YR 3/2) Clay loam Granular Soft to hard Low 40/60 Very abundant XI 386-436 Pale red (2`5 YR 6/2) Sandy loam* Granular Hard Low 5/95 Scarce XII 398-438 Gray (2´5 Y 5/0) Sandy loam* Granular Soft _ 1/99 Scarce XIII 400-449 Pinkish gray (5 YR 6/2) Sandy loam* Granular Hard Low 20/80 Scarce XIV 416-443 Pinkish gray (7`5 YR 7/2) Sandy loam Granular Hard _ 20/80 Very abundant XV 405-470 Pale red to pink (2´5 YR 6/2; 2´5 YR 8/4) Sandy loam Granular Hard Low 40/60 Very abundant XVI 423-480 Pale red (2´5 YR 6/2) Loamy Granular Hard _ 30/70 Slightly abundant XVII 436-493 Pinkish gray (7`5 YR 7/2) Loamy* Granular Hard _ 30/70 Slightly abundant XVIII 462-539 Pinkish gray to pale red (5 YR 7/2; 2´5 YR 6/2) Silt loam Blocky Hard _ 20/70 Slightly abundant XIX 494-550 Reddish brown (5 YR 5/4) Silt loam Granular-Blocky Very hard _ 1/9 Slightly abundant XX 546-580 Reddish brown to pink (5 YR 5/4; 5 YR 7/4) Silt loam Granular Plastic _ 20/50 Slightly abundant Table 1. Descriptive characteristics of the stratigraphic levels. Looking at the arrangement of the stratigraphic levels, one of the most noticeable facts is the strong dip direction that levels VII to XV have towards the east (see Fig.1), as well as the small but abundant fractures that dominate in the speleothem system situated in the excavation area of the karst complex and provoke the vertical displacement of some speleothems regarding others. Furthermore, the small cavity (Covacha del Angel) as much as its extension in the sinkhole (Sima del Angel), are part of a fault, in turn integrated in a major fault system on the Sierra de Araceli. Consequently, it seems evident that tectonic processes have occurred once the archaeological record had been deposited. This point has been made quite certain because the original roof and walls from the original cavity are absent, although their existence has been registered in the present floor with frequent speleothem structures. Neither should we exclude a suction effect produced by the sinkhole itself on sediments and gravitational blocks, since it is obvious that both parts of the karst system, the small cave and the sinkhole, were connected in the past. The twenty stratigraphic levels present a subhorizontal arrangement or dipping towards the east (levels VII to XV) with convex morphology that suggest postdepositional collapse phenomena (Monge, 2012). The colouration of all levels varies between different tones of brown, gray, pinkish and reddish, with a loamy to sandy loam texture, according to USDA (Staff, 2010), and a remarkable granular soil structure. Concerning the consistency it is very heterogeneous because it varies according to how each level has been affected by secondary calcite precipitations. Porosity is non-existent or very scarce (if it exists, is always vacuolar), and contacts between levels are neat. The percentage of coarse fraction, as well as the frequency of archaeological material, present major variability. Considering these two variables, three big stratigraphic units have been differentiated (see Fig.1). 453
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Porosity  Coarse fraction     negrained fr...
454 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD The upper unit (I) comprises levels I to VI, and it extends from top (–215 cm) to –350 cm thick. It presents little archaeological material and also scarce and scattered coarse fraction. The prevailing texture is loamy and exhibits different types of structure: blocky, granular and platy. Porosity is quite variable (always of vacuolar type), and brown and reddish are the dominant colour. The middle unit (II) develops from –350 cm to –450, includes levels VII to XV and comprises a large amount of burned bones lithic tools and abundant coarse fraction from carbonated origin. Its colours vary among brown, grey and reddish. Although in a global concept this unit is a breccia deposit, considering the fraction lower than 4 mm, the dominant texture is sandy loam with fine subordinated granulometry; the structures of all levels is granular and its scarce porosity–when present- is of vacuolar type. This unit is the most homogeneous of the three units defined. The lower unit (III) develops from –450 cm to the base of the deposit, includes levels XVI to XX. The prevailing colours vary between pinkish grey and reddish brown. Its texture is sandy loam and much more fine-grained in the bottom part; no porosity has been observed and its structure varies between granular and blocky. This unit presents the lowest proportion both of archaeological material and coarse fraction that again appears disperse. Considering these two descriptive characteristics, which are the ones that reflect a major variability, Unit III is very similar to Unit I. Archaeological site description The site is situated in Mesozoic karstic unit of the Sierra de Araceli, and downcut in the Lías dolomitic limestone formation in the External Meridional Subbetic Domain; its description presents some complexity. The site is located in the southwestern slope of the sierra, at around 100 metres high regarding its base, so it cannot be considered a real mountainous site. The archaeological site presents three differentiated, although interconnected, parts. Firstly, the “platform”, an open-air area that would have been the remnant of a cave that probably collapsed before hominids came to this place. Archaeological excavations were undertaken in this area in 1995. Located a few metres from the platform, there is a small cave of approximately 60 m2 that still has not being excavated, although it was looted in the past. And finally, the sinkhole situated beneath the small cave and the platform. It presents a narrow morphology with vertical 100 metres depth walls. It is filled at the bottom with substantial detrital accumulations that from a 70 m high dejection cone. In 2009, a 30 metre long tunnel was drilled connecting the exterior, almost the base of the sierra, with the inner part of the sinkholes, in order to further study and better understand the nature of the human occupation and also to open the place to tourism in the future. Fieldwork excavations The first archaeological campaign at Cueva del Angel was undertaken in 1995-1996 authorized by the regional government. From these years until 2014, the General Direction of Cultural Heritage of Junta de Andalucia has authorized seven archaeological campaigns on the site: two short and very specific campaigns in 1995-1996; other two short archaeological excavations in 2002-2003; two campaigns in the context of a research project (2005 and 2008), and lately, one new fieldwork excavation in 2013. The archaeological fieldwork has been funded, mainly by the municipality of Lucena, as well as by Consejería de Cultura and Cuevas and Sima del Angel Foundation. Until 2013, the archaeological excavations have been focused at the platform sector. The excavation area has been partitioned into a system of Cartesian coordinates of 1 m2 squares with letter and number axes. The following zones have been excavated: J7, K5, K6, K7, K8, F8, G8, H8, I8 and J8. In 1996 a mining well (L7/L8) from the end of XVIIIth century was discovered. It has 2 m diameter and 5 m depth and the refreshing of its profiles has uncovered a sedimentary sequence along five grid units, all of them situated in the transversal axis of J/K sector. At the present time, excavation has also started in the longitudinal axis in the grid units of the number eight. The platform area has less than 300m², where at least 100 m2 are Pleistocene deposits, the rest correspond to limestone outcrops and especially speleothems (stalagmite, wall speleothem, etc.) which indicates that a closed cavity existed in the past and subsequently –at an undetermined moment– the roof and walls of the cave collapsed leaving the inner part exposed to open air. It has
454  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  The ...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS been hypothesized that on the north wall there could have existed traces of a kind of ceiling configurating some rockshelter-like form where hominids installed a hearth of more than 3 m in diameter. This perfect place for shelter and the rich faunal and botanical environment in the surroundings of the site, favoured the permanence of the Acheulean occupation in this place. Spatial and taphonomical analysis of faunal assemblages, as well as typological and technological studies of lithic industry recovered at the platform area have raised the hypothesis that this space was used for skinning, fracturing, disarticulation, filleting and defleshing animal food transported to the site by hominids, and then cooked on fire and consumed there. Knapping is also documented in this place and is represented by non-modified flakes used as knives. Lithic tools present few variety, sidescrapers predominate (typical for fur working), and notches and denticulates have a very low percentage. Bifacial pieces have pointed shape edges; in some cases this pointed extremity is broken due to the impact of breaking bones. Truncations, burins and endscrapers are anecdotic. It is worth mentioning that all this lithic type distribution is closely related with the specific economic activity developed on the site, and that would explain the absence of numerous typologies typical from the Acheulean toolkit. In 2013, the archaeological excavation in the sinkhole was undertaken by means of 6m² test pit. Previously, a geophysics survey using electronic tomography was done. This survey determined the presence of a 10 m deep deposit with Holocene and Pleistocene origins. In 1 m2 of the test pit more than 800 fragments of human bones have been recovered. All of them belong to Homo sapiens and are associated to Neolithic, Chalcolithic and Bronze Age pottery. It has been proposed as a preliminary hypothesis that these human remains were discarded at the sinkhole. Stratigraphy in the J/K sector The sedimentary deposit is more than 5 m thick; so far 20 stratigraphic units (SU) have been identified without reaching the bedrock at the base. SUII, SUIII, SUXVI and SUXVII are the levels that contain more paleontological and archaeological materials. In the remaining stratigraphic units, although the presence of bones and lithic industry is evident, they are less abundant. No aban- donment episode has been detected at any level, which reinforces the hypothesis of an intense and continuous occupation of the site. Paleontological remains At the present times, an amount of approximately 8000 fossil bone remains compose the paleontological record of the Cueva del Angel. Bone density per m2 is very high; it almost doubles the lithic industry density. The bone assemblage is very rich, being represented at every one of the 17 excavated levels. The most important taxonomic groups found in the faunal spectrum are large mammals; mediumsized and small mammals although appreciable are scarcer. Fossil record has good conservation without evidence of weathering alteration or biological or chemical postdepositional processes. It is characterized by a high degree of fossilization, a recurrent presence of small concretion, big fragmentation and an extended occurrence of bone cremation. The main biases of the collection are due to the previous selection of the species that lives on the biotopes surrounding the site. In the order of Perissodactyla, the horse of Cueva del Angel (Equus ferus) (Fig.2) presents similar characteristics to the pre-Würmien horse of the Iberian Peninsula. Stephanorhinus hemitoechus corresponds to an evolutionary stage of the second half of the Middle Pleistocene. The presence of Dama indicates a record previous to the beginning of the last Glaciation. In that context, Capra only appears on the base of the sequence and could be the oldest species of the Iberian Peninsula. There are also Mediterranean taxa represented: Equus hydruntinus, Cervus elaphus, Bos primigenius and Palaeoloxodon antiquus. One of the particularities of the fossil assemblage is the good representation of Sus scrofa, this species is generally scarce in the Middle Pleistocene records; also, the presence of Bison pricus has been documented in one of its southernmost locations in Europe. Carnivores are represented by 4 taxa: Ursus arctos, Canis lupus, Felis silvestris and Lynx pardinus spelaeus. This Iberian type lynx ought to be used as a transition form in the anagenetic southern lineage. The best represented taxa in the fossil assemblage are Equus, Bos/Bison and Cervus. All of them appear in the whole sequence. Large species present a vari- 455
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  been hypothesized that on the north wall t...
456 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 2. Equid silhouette indicating anatomical parts transported to the site by hominids able proportion always bigger then cervids number, although this last species remains constant. Regarding the mortality profile, adults compose the most abundant group, while juveniles are only a few, except in the case of Stephanorhinus, characterized by major presence of infant individuals. The presence of carnivores is very discrete in the whole collection as well as their evidences on bone remains. Conversely, a large proportion of cutmarks has been documented and intentional breakage patterns that point to an anthropogenic origin for explaining the accumulation of faunal assemblages. The action of carnivores would be limited to the scavenging of bones previously discarded by humans. Skeletal representation of large and medium mammals is mainly composed by long bones, cranial remains (basically mandibles) and, to a lesser extent, hip bones. The axial skeleton is more scarce and mainly represented by ribs. The presence of joints is very punctual, whereas the presence of phalanx is slightly higher. This differential preservation does not respond to the bone nature that can suggest alteration and postdepositional destruction processes, but is the product of anthropic action. Anatomic representation of large and medium-size mammals is very regular along the whole sequence and it is the result of a specific strategy oriented to bone marrow exploitation. The distribution pattern showed by cutmarks, the skeletal representation and the absence of consumption patterns typical from carnivores, indicate a primary access by hominids over the animals’ carcasses of the three prevailing species. Taking into account the large proportion of adult animals, all the evidences point to well developed hunting strategies
456  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS (Gaudzinski and Roebroeks, 2000). Nevertheless, the anatomic representation of Sus, Palaeoloxodon and, probably Stephanorhinus, indicates secondary nutrient sources. Thus, the meat supply comes primarily from cinegetic strategies, with a secondary access over certain small preys. Carcasses are generally dismembered and cut into pieces in the hunting place and then the parts with more nutritional value are transported to the site. Among others, highly splintered long bones and cranial remains are the most represented, also with a regular presence of phalanx indicatives of the bone marrow extraction. For instance, the skeletal representation of cervids is quite complete along the whole sequence. Butchering and fracturing evidences suggest that all the carcasses processing and exploitation, from skinning and evisceration to filleting and periosteum scraping has been documented. The presence of retouchers on different bones of cervid, horse and auroch evidenced the use of lithic industry in situ. Bone marrow exploitation is one of the main activities recurrently documented with the same pattern along the whole sequence. Its use is systematic, as recursive fragmentation method of hemimandible, first phalanxes, hip bones, scapula and even calcaneus demonstrate. Food production strategies follow hardly the same pattern along the whole stratigraphic sequence. Paleoecological context A preliminary paleoclimatic indicator of the environment of the site is provided by the herpetofauna assemblage of the Cueva del Angel, which is characterized by the presence of taxa typical of the Mediterranean domain (Barroso et al., 2011; Barroso et al., 2012). Blanus cinereus is the best squamous represented; ars the genus Chalcides also appears, large size lizards such as Timon lepidus and several small species, such as genus Podarcis and Lacertidae ind., among others. Snakes are represented by colubridae Coronella sp. and Malpolon monspessulanus, and probably cf. Hemorrhois hippocrepis. Chelonia remains correspond to Testudo hermanni, one species that nowadays is only present in the Catalan region, although until the Upper Pleistocene it had a wider geographical distribution in the Iberian Peninsula as their presence in Zafarraya Cave demonstrates (Barroso and Bailon, 2003). Amphibia taxa represented are Bufo bufo, Bufo calamita, Discoglossus (indet.) and Alytidae (indet.). The two first species are currently widely distributed in the Iberian Peninsula where they occupy a great diversity of habitats. The present geographical distribution of the majority of these species has a climatic threshold linked to temperature and summer insolation: average annual temperature greater than 10º C, minimum average temperature of summer months greater than 21º C and average annual insolation of between 2500 and 3000 hours. Among the large mammal remains present at the Cueva del Angel, large herbivores are the most represented taxonomic group, while the presence of carnivores although appreciable is more modest, and the presence of rabbits is very scarce (Barroso et al., 2011; Barroso et al., 2012). The faunal assemblage is dominated by the horse Equus ferus, followed by large bovids B. primigenius/ B. priscus and cervids C. elaphus and D. dama with, although less abundant, a good representation of the suid S. scrofa, the rhinoceros S. hemitoechus, the brown bear U. arctos and the lynx L. pardinus spelaeus. The elephant P. antiquus and the wolf Canis lupus are scarce while the Ibex Capra sp. is practically inexistent. Given the latitude of the site and the average size of the species identified, smaller than the ones of Northern Europe specimen, the collection correlates with those from the end of the Middle Pleistocene. This assemblage in the Cueva del Angel corresponds to an accumulation of anthropic origin during a long period, and is not necessarily representative of a palaeo-biodiversified ambient, although the abundance of large hypsodont herbivores, associated with cervids and boars, reflects a mixed environment of wooded grasslands, probably with a more humid climate than today. Lithic industry Lithic assemblage of the Cueva del Angel recovered along seven archaeological campaigns, sum sup more than 5000 coordinated pieces and with a precise stratigraphic position (Fig. 3). However, aimed to better characterize the industry collection, some pieces come from the early cleaning operations of disturbed sediments which covered the site prior to excavation. Lithic artefacts are abundant in the whole stratigraphy; no level is sterile of that archaeological material, which indicates a continuous process of occupation of the site (Barroso et al., 2012). The lithic assemblage is relatively well preserved, although some flint pieces are desilicified and, around a third part of the artefacts have evidences of fire exposition. 457
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS   Gaudzinski and Roebroeks, 2000 . Neverthe...
458 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 3. Lithic industry: A. Endscraper on flake. B. and C. Flint bifacial sidescraper. D. and E. Double tool: denticulate lateral sidescraper and flint transversal sidescraper. f. Discoidal core with centripetal removals. G. Bone retoucher (pictures by V. Celiberti).
458  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Flint, quartzite and limestone are the three main raw materials employed. Flint is overwhelmingly dominant (more than 97%). It is very abundant in the same region where the Acheulean habitat is located. However, the basic procurement area are the fluvial terraces of the Genil River, 14 km from the site; for that reason, raw material comes on the form of pebbles and only occasionally as tablets and blocks, originally from siliceous stratified levels in limestone. Only 2% of the industry has been knapped in quartzite, whose close procurement areas are the fluvial terraces of the Guadalquivir domain, 40 km north. Limestone knapping is solely sporadic or opportunLevels Tools. retouched I Han axes istic; it represents around 0.47%. Both quartzite and limestone come in the form of pebbles. All the elements of the chaîne opératoire are represented in the lithic assemblage of the Cueva del Angel: entire rocks and pebbles (manuports), as well as percussion tools, knapped pebbles –although extremely rare–, bifacial elements and all the debitage products, cores, flakes, knapping waste products and retouched tools. The typological representation of the lithic industry remains quite steady along the whole stratigraphy (Tables 2 and 3). >2 cm flakes <2 cm flakes Lame/ small lame Cores Debris 64 148 168 10 25 II 30 64 33 4 III 63 118 41 IV 109 293 V 47 VI 61 VII Choppers & chopping tool nº % 269 684 12,28 1 21 153 2,75 9 8 60 299 5,37 169 12 17 142 1 747 13,41 55 20 2 5 22 1 152 2,73 166 87 6 5 233 560 10,05 30 85 31 5 3 36 191 3,43 VIII 26 65 21 2 7 18 139 2,49 IX 67 210 146 15 12 144 594 10,66 X 51 1 130 46 9 3 75 315 5,65 XI 18 1 64 10 2 3 16 115 2,06 XII 27 74 23 3 22 149 2,67 XIII 38 1 98 38 3 5 89 272 4,88 XIV 13 1 59 24 2 3 40 142 2,55 XV 107 1 237 96 8 17 111 577 10,36 XVI 23 31 8 3 2 22 89 1,60 XVII 8 49 6 1 2 8 74 1,33 XVIII 2 2 0,04 IND 45 35 67 12 318 5,71 TOTAL 829 46 2.013 979 100 % 14,9 0,8 36,1 17,6 4 2 Table 2. Lithic tool categories identified 1 1 151 8 93 272 1.336 4 5.572 1,7 4,9 24,0 0,1 100 459
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Flint, quartzite and limestone are the thr...
460 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Artefact types Nº % Endscraper 14 2,0 Burin 13 1,9 Drill 2 0,3 Truncated piece 13 1,9 Clactonian notch 59 8,6 Shoulder retouched 41 6,0 Multiple notch 4 0,6 Bec 16 2,3 Doble bec 1 0,1 Lateral denticulate 15 2,2 Transversal denticulate 7 1,0 Lateral sidescraper 294 42,9 Transverse sidescraper 85 12,4 Double sidescraper 76 11,1 Triple sidescraper 9 1,3 Convergent scraper 26 3,8 Points 3 0,4 Quinson point 4 0,6 Protolimace 2 0,3 Tayac point 2 0,3 686 100 TOTAL Group Nº % Upper Paleolithic type 42 6,1 Notched tools 143 20,8 Sidescrapers group 490 71,4 Points 11 1,6 686 100 Table 3. Groups of retouched tools on flakes. Non-modified flakes are largely the dominant category of the assemblage (53.7% of the total). They are mostly larger than 2 cm (36.1%), while the smallest ones often coming from retouched tool, are less numerous (17.6%). Traceological analyses have not been systematically done yet in the lithic collection of the Cueva del Angel. However some use-wear traces and irregular notches have been observed, which prove their use as knives. Handaxes and thick tools on flakes are also present along the whole sequence. They alternatively appear in small or medium size. More than one third of handaxes are made of flint, some in quartzite and very few of limestone. Handaxe tools were configured mainly on flakes or fractured pebbles, and some still have more or less cortical residue. Almost all of the handaxes show relatively thin pointed extremities, but with a low degree of convergence and oval, lanceolated, subtriangular and cordiform morphologies. The assemblage of large tools also includes a single chopper and a trihedral pick, both in quartzite. Most of the cores were knapped from flint and some from quartzite. These large primary supports would have been reduced on the site by intense debitage and, consequently the identification of original core supports is made difficult by the intensity of the reduction process. Operational schemas were directed towards progressively smaller blanks as volumes were repeatedly reduced using the flake-core technique. Most of the cores present low average dimensions and numerous removal negatives. Out of the analyzed cores (272), almost half were found in a precise stratigraphic position. Their frequency represents 5% of the total lithic material within the stratigraphy. Recurrent unipolar flaking, flaking on alternate surfaces and bifacial discoidal flaking are the main techniques employed at the site. Bifacial discoidal flaking has been most commonly observed for quartzite pieces, while the recurrent unipolar appears mainly in flint pieces. Concerning technological analysis a specific technique aimed to thinning blanks that
460  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Arte...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS may relate to Kostienky type thinning has been well documented. This recurrent unipolar reduction produce very thin flakes from natural or prepared platforms previously obtained from convex or planar extraction surfaces. Recurrent orthogonal reduction where each successive recurrent knapping sequence is followed by a change in striking platform (direction) –provoking exploited surfaces to sometimes become in turn striking platforms– is developed combining the production of centripetal or even partial discoidal flakes. This technique also occasionally produces blades and/or bladelets. A few pyramidal cores are present, whereas Levallois flaking methods are absent on the site. There is an almost total absence of cortical flakes, in contrast a large number of artefacts of the Kombewa type have been recovered, thus one cannot rule out the possibility that part of the raw materials were introduced into the site as large flakes or preconfigured cores, with the initial reduction stage knapped out of the archaeological site. The abundance of éclat débordant and éclat outrepassé illustrates intentional systematization in maintaining convex exploitation surfaces. Final knapping stages have produced small or very small flakes, smaller than 2 cm long. Small retouched tools are abundant along the stratigraphy and have been produced from flakes. Lateral sidescrapers represent the dominant typological group (75%), with lateral single sidescrapers being the most numerous (more than 290 pieces over a total of 294). Transverse side scrapers are also well represented, although double sidescrapers and composite scrapers show a low frequency. Sidescrapers most often present direct retouch, sometimes inverse or bifacial; retouch are thin, semi thick or flat, with 10% of the scrapers shaped by semi-Quina and Quina retouch. Edge morphology was most often convex, and sometimes rectilinear and concave. Notched tools (notches, denticulates and becs) are the second most numerous retouched tools, representing 21% of the total. In this group denticulates, retouched notches and single Clactonian pieces are the most frequent types, whereas single denticulates and becs are less frequent. At least, two convergent-edge denticulates may be assimilated to Tayac points. Combined Upper Paleolithic tool types are less frequent, they represent barely more than 6%. Within this group, endscrapers, burins and truncations are the most numerous. Pointed tools in general are scarce, and include Tayac points, four Quinson points and two proto-limaces. One of the outstanding characteristics of the Cueva del Angel industry is the frequency of flakes and retouched tools with thinned edges. Such thinning is observed on support bases but also on their lateral and distal edges. Another specific technological trait is the removal of flakes from the retouched tool edges, which produces a very characteristic morphology rarely observed in others sites. This could be a distinctive hallmark of the Cueva del Angel industry. Knapping patterns at the Cueva del Angel reflect exhaustive, well standardized and economic use of relatively fine quality materials. Early phases of knapping are not represented in the assemblage since initial shaping was performed outside of the cave, probably on the catchment area where a singular branching operational schema was practiced. This schema was based on repeated application of recurrent unidirectional, often radial, knapping from prepared striking platforms. This economical method sometimes produced cores with a morphology akin to Levallois forms. Raw material procurement is essentially local, which is a typical characteristic of the Acheulean and Mousterian technological behaviour around other sites in Western Europe. The Cueva del Angel lithic assemblage appears to fit well within the regional diversity of well developed final Acheulean industry, and its technological particularities can be interpreted as one more expression of the regional variability already observed at other sites of the Iberian Peninsula,as elsewhere in Western Europe at the end of the Middle Pleistocene. Dating During the first archaeological campaign in the platform of Cueva del Angel, the geologist Joaquín Rodríguez Vidal, carried out a first sample selection aimed at dating. Samples were taken from calcite flowstones to be dated by U/Th. The seven samples taken were processed at the Institut de Paléontologie Humaine in Paris by members of the team of C. Falguères (Botella et al., 2006). Some of the samples were taken to obtain geological ratings (LU9501, LU9502, LU9503, LU9505 and LU9506), and two were aimed to date the Pleistocene deposit (LU9504 and LU9507). At these moments, and after a deep- 461
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  may relate to Kostienky type thinning has ...
462 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD ened knowledge of the site, it has been detected that none of the samples are reliable, because there is no certain attribution to secure location in the topography. Thus all the results have been rejected, including the sample LU9504 that gave a result of 121+11/-10 ky (SU VIII) and has been used as a chronological indicator for the faunal and lithic assemblages of the site. Currently, new dating is being undertaken by Alfredo Pérez and C. Falguères. Conclusions The significant use of fire is one of the most remarkable characteristics of the site. Around 90% of the faunal bone remains are burnt. Mineralogical changes experimented by fossil bones have been recently investigated (Monge et al., 2014). Eighteen samples of fossil bones from J/K sector have been analyzed. In the light of analysis and subsequent studies, the existence of different thermal events of anthropic origin has been proved, not only based on the colour spectrum but also in the increase of the cristallinity of the phosphated phases and the apparition of authigenic rare phosphates as the whitlockite. Besides, diagenetic processes have been observed, including phosphate authigenic. Also, the presence of secondary calcite has been established and manganese oxide precipitations. The analytical evidence indicates that reached temperatures ranged 650ºC to 700ºC, in the large stratigraphic units I and II. So far, no analysis has been done in unit III, thus it is not clear that fire existed or not at that moment. It is clear that meat roasting was one of the most frequent activities, although some smoking strategies aimed at better preservation of the foodcannot be ruled out (Patou-Mathis, 1996). On the other hand, the large degree of fragmentation and the absence both of epiphyses and spongy tissues could indicate the use of bones as fuel (Costamagno et al., 2005; Théry-Parisot et al., 2005). The stratigraphic continuity of the sequence with neither sedimentary hiatus nor human abandonment episodes (Barros et al., 2011), the low carnivore intervention, together with the continuous use of fire and the persistence of butchering patterns, suggest a long, intense and continuous occupation of the site. Cueva del Angel would be a place where most of the daily activities took place and also would act as central point from where all the logistic strategies of raw material procurement would occurred. The faunal spectrum and the skeletal representation indicate the exploitation of a rich and varied ecological niche. Concerning the lithic industry, the tool typology, the technology, etc., fit well within the regional diversity of a well developed final Acheulean industry in the European continent. Knapping patterns at the Cueva del Angel reflect exhaustive, well standardized and economic use of raw materials. It has been observed that raw materials came to the site as preconfigured cores, probably being preliminary flaking done in the procurement area. Ninety-five percent of cores were knapped on flint. Recurrent unipolar flaking, as well as bifacial discoidal flaking are quite significant. Cores were almost depleted, which give rise to small size flakes. After analyzing 272 cores with a reliable stratigraphic context, the presence of Levallois flaking method should be ruled out, as we already mentioned in 2006, when the studied lithic assemblage was composed of 667 pieces (Botella et al., 2006). In the first years of research at Cueva del Angel, lithic assemblage from its upper levels was assigned to a Late Acheulean and to a Mousterian of Acheulean tradition periods. Conversely, in recently published papers (Barroso et al., 2011; Barroso et al., 2012), when the excavation had progressed and sufficient lithic material had been recovered, it was determined that no changes occurred along the whole sequence, the typometry of pieces present the same values, handaxe typology and size alternate indistinctly along the sequence, sidescrapers typology and retouched over them does not change and core typology and exploitation remains the same. In other words, from the base to the last occupation level, the same techno typological pattern continues, and that corresponds exclusively to an Acheulean industry, excluding completely the presence of Mousterian industries. The presence of Acheulean industries in the Iberian Peninsula is basically confined to the fluvial ambience of major rivers. These industries are made with quartzite cobbles (choppers, cleavers, pics, handaxes, etc…) and, technologically are far from the one encountered at Cueva del Angel. Probably, industries from fluvial terraces respond to a different scheme of raw material procurement and transformation developed in the same place, whereas industry from Cueva del Angel responds to a unitary scheme composed by the
462  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  ened...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS habitat and the procurement area. Inhabitants of Cueva del Angel made an extraordinary use of small tools as flakes, also typologically diverse sidescrapers were common. In the future, further dating of levels with lithic industry on fluvial terraces as well as Acheulean levels in caves, will help to ascertain the relationship between both contexts. At the moment, the reference sites that can be compared with Cueva del Angel are Galería in Cecilio Barroso *, Miguel Caparrós **, Deborah Barsky ***, Anne Marie Moigne ****, Antonio Monclova * Regarding the chronology, the rejection of dates used so far prevent the establishment of a precise chronological framework of the Cueva del Angel, but there is no doubt that the analysis currently being done will age considerably the date of 120 ky years available so far. Boquete de Zafarraya cave: A Neanderthal site in southern Iberia Introduction Boquete de Zafarraya cave (hereinafter C.B.Z.) is in the Alcaucín municipality, Málaga Province (southern Spain). It is 1020 m asl, facing S-E at the foot of a cliff, with a steep slope below it. The coordinates are: 36º 56` 58``N – 4º 7`40`` W. This cave in Sierra de Alhama overlooks an extremely irregular landscape of limestone ranges and very narrow valleys boxed in between them. Scarcely a few dozen metres away, however, the landscape opens onto the mountain pass, Boquete de Zafarraya, which at 900 m asl connects the Mediterranean coast with inland Andalucia (Fig. 1). This landscape mosaic is completed by the 22 km² Zafarraya polje, formed from an intra-montane depression of karst and tectonic origin filled with basically Quaternary sediments, powerfully contrasting with its surroundings by its extensive subhorizontal landscape (Barroso, C. et al., 2006). The prehistoric Boquete de Zafarraya cave site was discovered by a member of this team (C.B.R.) in 1979. During the first stage of the archaeological excavations (1981-1983), * Atapuerca (Carbonell et al., 1999), Cova del Bolomor (Fernández Peris, 2007) and Galeria Pesada (Marks et al., 2012). led by C. Barroso and F. Medina Lara, we discovered levels containing Mousterian industry, a rich and varied quaternary wildlife and two quaternary Neanderthal fossils (Zaf. 1) and (Zaf. 2) (Barroso et al., 1983; Barroso et al., 1984.). The second archaeological stage at the site continued from 1990 to 1994. B.Z.C. is the last vestige of a cave which collapsed and whose morphology has been heavily affected by powerful slope erosion. At the mouth, a large arch-shaped porch rises 30 metres, measuring approx. 10 metres at the base. On the north wall there is a narrow 22 m long gallery, barely 2 m. in the widest sections, developed from a fault plane. The cave is in stratified pisolitic and pseudo-oolitic limestone. This gallery is the source of the quaternary deposit. The infill was preserved by the carbonation processes inside which cemented much of it, preventing the intense erosion which took place throughout the area, heavily affected by the steep slope. Neanderthal fossils B.Z.C. is perhaps one of Iberia’s most important sites of Homo neanderthalensis fossils (Bar- Fundación Instituto de Investigación de Prehistoria y Evolución Humana. Plaza del Coso 21, 14900 Lucena (Córdoba, España) email: barroso.cecilio@gmail.com ** Département de Préhistorie, Muséum National d’Histoire Naturelle, 75013 París, France *** Departement d’història i història de l’Art. Universitat Rovira i Virgili. **** Muséum National d’Histoire Naturelle. Centre Européen de Recherches Préhistoriques de Tautavel. 463
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  habitat and the procurement area. Inhabita...
464 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. Neanderthal Capra pirinaica hunting scene in Boquete Zafarraya (author: Antonio Monclova Bohorquez). roso et al., 2006 b). There are at least 9 individuals (2 children, aged 14 months to 12 years and 7 adults, four almost 20 years old and two aged 25 and 30, and one undetermined). There is a notably poor representation of infants and a heavy adult mortality. The most surprising feature, however, is the “significant” presence of Neanderthal bones (16 fossils) in a cave which was visited sporadically by these hominids, who spent very short periods there, computed as a few days each time at the most judging by spatial analysis of the lithic industry and the fragments of animal bones bearing obvious signs of anthropic manipulation. Anatomically, the fossils include: is practically intact, with 13 teeth preserved. This jawbone was studied in depth by Maria Antoinette de Lumley, who established the phylogenetic position of Zafarraya Man in the group of so-called “classic” Neanderthals. Zafarraya retains some features described for the Ante-Neanderthals, particularly amongst those discovered in Banyoles and Montmaurin, although no features permit their connection to modern populations. Mandible Zaf. 4-5-18, discovered in 1990 and 1992, is broken into 3 pieces, with evidence of fire exposure, related to the presence of a Mousterian hearth in the quadrat Q-18. All three fragments are from an adult mandible. The proportions and morphology are comparable to Zaf. 2. Cranial elements: 2 mandibles (Zaf. 2 and Zaf.4-5-18) (Fig. 2) and 5 separate teeth (Zaf. 12 Zaf. 16, Zaf. 20, Zaf. 23 and Zaf. 24). Upper limbs: 1 scapula (Zaf. 6), 1 humerus (Zaf. 22). Mandible Zaf 2, discovered in 1983, was initially attributed to a Homo sapiens neanderthalensis male aged between 25 and 30. The jawbone The specimen discovered in 1990 is the lower half of the glenoid cavity of a right scapula. Its blackish surface evidences exposure to fire and, like other human remains, is associated with the
464  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS hearth in quadriculate Q-18. Comparison with the Krapina scapulae shows it is amongst the variations attributed to males. Humerus Zaf. 22 was discovered in 1992 at the rear of the gallery on a disturbed superficial level, and thus lacks stratigraphic context. Morphological and metric analysis has attributed this fossil to the Neanderthal group, although certain parameters such as maximum length, minimum perimeter and poor robustness are also present in modern humans. Morphologically, the humerus matches the Neanderthals from Cova Tossal de la Font and Grotte d’ Hortus. The proportions indicate that this was a fairly robust and unequivocally female Neanderthal. Lower limbs: 3 femurs (Zaf. 1-15 and 16), 1 tibia (Zaf. 27) 1 coxis (Zaf.17) and 1 foot phalange (Zaf. 3). Femur Zaf. 1 was found during the 1982 digs. Only the upper half of a right femur was preserved, in which part of the upper epiphysis and the upper half of the 225 mm long diaphysis can be appreciated. The femoral head and trochanters were broken. The initial study (see note) attributes this to a Neanderthal adult male aged 30-49 years. It has a relatively strong size ratios for a Neanderthal, measuring 159 cm to 162cm in height. The medial bone fragmentation and the large splinter extracted from the diaphysis are clearly anthropogenic. Metric and morphological analysis concludes that Zaf. 1 is distinct from Neanderthal femurs in the lack of a vertical incurvation. On the other hand, the cortical thickness, antero-posterior flatness of the shaft and the lack of a posterior pilaster are comparable to Neanderthals. The less pronounced muscular relief than other fossils suggests that these were less robust individuals than the classic Neanderthals populations in southwest France and the Middle East. Femurs Zaf. 15 and 26 (Fig.2). Found heavily fragmented and burned in the Q-18 hearth during the 1990-1992 excavations. Zaf. 15 preserves the shaft of a left human femur. Both the anterior and posterior faces have been reconstructed from 11 fragments. Zaf. 26 retains part of the diaphysis of a right femur, reconstructed from 9 fragments. Its robustness is quite different from Zaf.1: Zaf. 15 and 26 have slender shafts and weaker anteroposterior and transversal diameters than Zaf.1. The major difference in strength between the three femurs could reflect Figure 2. Neanderthal fossils in Boquete de Zafarraya cave. From Left to right A: Mandible Zafarraya 2 (photo. Cecilio Barroso Ruíz), B: Burnt mandible 4-5-18, C: Burnt Femur Zafarraya 26, D: Burnt Tibia Zafarraya 27, E: Burnt femur Zafarraya 15 (B, C. D and E: photos. Rafael López Gómez. sexual dimorphism, with FAZ. 1 belonging to a slim male individual while the other two femurs are from two females. Zaf. 27, a right human tibia (Fig. 2), consists of 10 burned fragments found in the Q-18 hearth. The lower half of the shaft and a fragment of the upper half have been preserved. It seems to be from a graciles individual, whose overall proportions suggest an adult female, with the diameter in the midsection of the shaft close to the Ferrassie 2 individuals, attributed to a female, and smaller than those of the Chapelle-aux-Saints tibia, attributed to a male. The phalanx of the right foot (Zaf. 3), located during the 1983 dig, is incomplete, al- 465
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  hearth in quadriculate Q-18. Comparison wi...
466 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD though it may be attributed to a proximal phalanx of the second toe of the right foot. Its morphological and metric characteristics suggest that this may be Neanderthal phalanx. It belongs to an adult, as the overall dimensions of the head are comparable to the homologous Shanidar 8 phalanx, attributed to a female. The coxis (Zaf.17) was located during the 1992 dig in quadrat Q-18, on the edge of the pit that shaped the Mousterian hearth. It is well preserved and permits the examination of the symphysis face, the horizontal and descendant branches of the pubis. The break at the branches is old and covered with concretion. It is attributed to a male individual estimated to have been 20-21 years old. Thorax: 1 rib (Zaf. 19). This human right rib appeared in the 1992 dig in the context of the Q-18 hearth. It is from a right C7, two-thirds of which are well preserved. It is characterized by its thickness, slight helical torsion and faces angled obliquely at the base, which probably gave the bottom of the rib cage an open form, contrary to modern men. It is attributed to an adult Neanderthal male. Mousterian industry A rich and stone tool industry has been collected from all seven distinct Archeostratigraphical Units (AU) attributed to the Mousterian (Barroso et al 2006 c.) The global assemblage comprises 813 items; mostly knapped in flint as well as some quartzose-sandstone. Lithic raw materials were collected from primary and secondary sources not exceeding 12 km away. Outcrops situated some 11 km away provided the finest flint and were preferred to other nearby sources. The assemblage is well preserved overall but some flakes display post-depositional edge damage. Patina attributed to fire exposure is observed on only 17 items. The lithic type distribution per-AU varies significantly according to occupation intensity and the extension of excavated surfaces but no noteworthy technological or typological differences are detected within the sequence. Knapping waste (L> 2 cm) dominates the assemblage and the average flake/core ratio is 23. This core deficiency is to be put into relation with the paucity of small-sized flakes and flakes with residual cortex, all suggesting that much of the knapping and shaping occurred off-site. Many of the 539 knapped items (flakes, blades, points) were obtained by Levallois production methods (194 pcs., Levallois technological index excluding pseudo-Levallois points= 36%). No important variation in the Levallois index is observed within the sequence: the entire assemblage is Levallois dominant. A ratio of 1 Levallois product for every 2 to 3 standard products is constant regardless of the raw material used. Levallois flakes are numerically prevalent (161 pcs.) compared to blades (25 pcs.) and points (8 pcs.). Recurrent centripetal is the most common Levallois method attested from flakes and cores, although lineal, unipolar recurrent and bipolar methods have also been identified. A few symmetrical preferential flakes reflect a final phase of recurrent centripetal Levallois schemes with dorsal surfaces bearing traces of their core’s prepared convex surface. The Levallois flakes often present prepared, facetted platforms (LevFI = 48,6, FI = 16,6). Among the 23 cores unearthed at the site, only 4 are clearly of Levallois conception: 3 centripetal recurrent and 1 unipolar recurrent. A further three cores present convex extraction surfaces analogous to Levallois-type production strategies. Alternative core reduction modes appear nonstandardized, mainly because the matrices are exhausted. These are uni, bi and multi-directional and polyhedral. Most cores are knapped from flint nodules rather than flakes and the assemblage includes only three double ventral flakes and two cores on flakes. Blades were mostly made from the finest flint and their overall index is quite low (BI = 8,1%). Elongated supports were rarely obtained by methods other than Levallois (NLevBI =5,5%, LevBI= 12,8%). Use of a soft percussion instrument is only sporadically recognized. The BDZ industry is further characterized by a high frequency of core-edge flakes (Fr.: éclat débordant) and pseudo-Levallois points, both typically obtained during different phases of Levallois and discoid core surface convexity maintenance.
466  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  thou...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Nearly one quarter of the flakes > 2 cm are retouched into various tool types. While no selectivity of supports is evidenced, the Levallois typological index is to be considered elevated (41,4 %). The retouched toolkit includes: Mousterian points, scrapers (42%), some notched tools (30 %), trimmed and truncated instruments, as well as some Upper Paleolithic types. Quina retouch is rare and there are no handaxes and only one chopper-like tool. Given the techno-typological features outlined above, the stone industries from the Boquete de Zafarraya Mousterian levels are ascribed to a Levallois dominant typical Mousterian. the sedimentary sequence, along with evidence of humans and carnivores, clearly show an alternation in the occupations. The remains of aurochs, deer, wild boar and horse indicate that hominids frequented the Zafarraya polje moorland where these animals were present. Some of the rodents and small birds also came from these moist wooded areas. The faunal remains from the different levels suggest a concomitant presence between carnivores and Neanderthals, with several possible scenarios. There may have been mutual periodic exclusions in which Neanderthals were replaced Large animal vertebrates The faunal remains found at the Zafarraya cave site (Barroso et al., 2006) provide information about various aspects of the Neanderthal lifestyle. First, the list of fauna matches a fresh episode in the middle of the last glaciation. The abundance of Cuon alpinus and Panthera pardus (Fig. 3) is unusual at European sites, permitting its comparison with caves in the Basque Country. The vast majority of the sites where panther (245 remains, 10 individuals) or dhole are well represented have been dated in the Hengelo episode. Ursus arctos (Fig 3), Crocuta crocuta, Felis silvestris, Lynx pardina and Vulpes vulpes form the classic range of Upper Pleistocene carnivores. The herbivores do not point to a specific biochronology except for the apparently robust horse, attributed to Equus c. germanicus which became extinct 30,000 years ago. A substantial majority is mountain fauna, particularly goats (78% of the identified remains) and fallow deer, which are associated with their usual predators: panther, lynx and dhole. The frequent meso– to supra-Mediterranean environment has been described on the basis of palynological analysis as cold and very dry for a herbaceous steppe stratum with a few pines. The sedimentological sequence dated between 49 and 38 ky corresponds to several alternations of more or less moist climates with a homogeneous vegetation throughout the Mousterian occupation of the cave. The microfauna also describes a colder climate than today at 1000 m asl. The cave provided shelter for these large mammals on several occasions. The 46 archaeo-stratigraphic units defined from Figure 3. Carnivore fossils in Boquete de Zafarraya cave. From left to right o A: Panthera Pardus mandible, B: Linx pardina ulna, C: Panthera pardus maxillary, D: Panthera Pardus mandible, Ursus arctos ulna (Photos: Rafael López Gómez). 467
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Nearly one quarter of the    akes   2 cm a...
468 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD by carnivores or vice-versa as the cave occupants. Another scenario is that these two types of predators were not mutually exclusive and may have coexisted, or perhaps the presence of carnivores might even be a direct consequence of the humans. The analysis of different interventions by panthers, lynx and dholes suggests a more intense involvement by the former, which is also consistent with the number of individuals at all levels. The presence of dholes on the upper levels of the infill is more sporadic and their remains show greater pressure by predators. Skeletal representation of goats indicates that the carcasses were brought to the site, although the bias is not uniform in all levels. The abundant remains of this species (2667 items in the stratigraphy) show all stages of impact by carnivores on virtually every part of the skeleton. However, carnivore marks are associated with evidence of anthropological activity. Evidence of carnivore activity is particularly notable on the bones of other carnivores such as dholes, but also on the remains of a hyena and also a bear discovered on the low levels of the infill, on the “levels containing panther”. Deer and wild boar were probably prey for large carnivores. Human activity is visible on many goat items. Like the large carnivores, Neanderthals frequented the cave when hunting goats in this area. The high numerical presence of juveniles indicates the hunting season, which ran from spring to summer. Goat butchering is attested by the numerous cut marks on skulls, phalanges, axial skeletons and long bones. Deliberate fracturing of the jaws and long bones is quite common. Other species such as deer, from a much broader area, were also carried intact to the cave, and similarly have numerous marks and deliberate, systematic breakage. In this case, evidence of human activity predominates over the marks of small carnivores, which are superimposed on the cut marks. The carcasses of large herbivores such as aurochs, horses and wild asses (Equus hydruntinus) are quite incomplete. They may have been dismembered previously, with the most nutritious parts brought to the site. Evidence of carnivore impact on these remains is generally weak. The fossil association on the archaeological levels of Boquete de Zafarraya cave is indicative of frequent occupation by Neanderthals in summer, at the same time as carnivores exploited the cave. The mortality of juvenile deer and carnivores coincides with that of juvenile capridae, clear evidence of intervention by carnivores and humans at the same time. The considerable anthropic presence at the Zafarraya site, evidenced by the stone tools found in every archaeostratigraphic unit, seems to have posed no impediment to panthers frequenting this cave. The abundance of goats and rabbits in the environs must have proved a particular challenge for both groups. Palaeoecology The Zafarraya cave environment was generally inhabited by rock-dwelling species of large mammals directly related to the morphology of this landscape, as well as other species more characteristic of woodlands, areas with abundant water, the polje, sheltered zones and mountain areas with lush vegetation, all propitiated by higher moisture levels than the current local conditions. The best represented carnivore species in Zafarraya cave are dholes and leopards (Fig. 3), both rare at most European Pleistocene sites (all the more so in caves at this altitude). The dhole, typical of the warmest periods of the Eemian interstadial, is robust, closer to the currently species found in semi-desert and mountain landscapes, where it hunts in groups. The leopard occupied several environments including forests and abrupt areas such as Zafarraya where herbivores hunted medium-small sized herbivores. Hyena and fox, more robust than the Mediterranean fox, were scarce while the lynx was slightly more numerous. Bears, lynx and weasel were quite robust, indicating a temperate climate. Of all the carnivore taxa, only brown bear, bobcat and lynx are characteristic of forest habitats, while the rest were more or less ubiquitous (Barroso et al., 2006e; Monclova et al., 2012.). The mountain goat is the most abundant herbivore in Zafarraya Cave, probably a preferential prey for Neanderthals, dholes and leopards. Relatively gracile, this goat is different from the larger preWürm species and the Mediterranean goats associated with Mousterian sites. The continuous presence of deer indicates that it was hunted during summer in the polje woods near the cave. Chamois seek winter refuge in forests when the peaks and pastures are covered with snow, but in Zafarraya they appear in more benign periods, probably because the
468  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  by c...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS moister climate at the time reduced the snow cover in the higher areas. The auroch appears frequently throughout the Pleistocene on the Iberian Peninsula. Its minor presence in Zafarraya could be associated with flooded areas in the cave near the polje. Wild boar are usually quite rare in the Pleistocene, and while scarce in Zafarraya, they show a preference for forested and, more particularly, wet areas. Horses, also rare, are represented by the robust Equus caballus germanicus, known in Western Europe during the first half of the last glaciation. Its relative absence could be related to the site’s elevation. The wild ass, frequent since the last interglacial and typical of the Mediterranean region since the beginning of the Middle Pleistocene, could survive in rugged areas like Zafarraya (Monclova et al., 2012). Given that the few mammal taxa at Zafarraya reduce the representativity of the cenograms, in some of them data from several archaeo-stratigraphic units have been used to clarify the site’s environmental and climatic evolution. On the basis of the preferences shown by different taxa for open/closed, wet/dry environments, these cenograms point to a moist climate which only underwent minor variations throughout the stratigraphic sequence. The lack of a dividing line between the data on the large and small body mass of all Zafarraya mammals indicates that the habitat was closed. When carnivores (average sized species) are removed from the analysis, a dividing line does appear between macro– and micro-mammals, indicative of the presence of open spaces on the edge of the forests in a temperate and permanently moist climate. Evidence of a closed environment not far from the rocky cave landscape, in a climate tending to be moist, matches the polje. The various biotopes around Zafarraya formed a landscape mosaic in which different forest fauna taxa coexisted with other inhabitants of more open and even more humid environments such as streams and wetlands, which only retreated to the polje in the driest periods. Through the successive archaeostratigraphic units in Zafarraya, an abundance of hypsodont animals alternated with braquidonts, an indication of mountain + grassland or forested environments, respectively. The climatogram drafted to compare Zafarraya’s major archaeo-stratigraphic units clusters species by their ecological affinities (giving pri- ority to the moisture gradient). It illustrates the climate trend through the sedimentary infill of the cave as the fauna composition changed in the stratigraphic sequence. Some results are difficult to interpret. While the goat/deer or bison/ deer ratio reflects variations in the intensity of the cold at many Mousterian sites, the altitude of the Zafarraya site hinders this pattern, and thus the continued presence of goats does not necessarily indicate variations in the intensity of the cold but rather oscillations linked to the presence or absence of other species. Similarly, although the presence of horses might indicate alternating colder or drier climates, here they are not only rare but their remains are associated with archaeo-stratigraphic units with highly uneven moisture levels according to other records. In ungulates, neither chamois nor aurochs show variations related to the frequency of the presence of deer in the stratigraphic sequence, while the presence of wild boars, usually associated with increased moisture, does match the increase in deer and chamois, also in a moister period (Monclova et al., 2012). Finally, variations in the proportions of the macromammals through the Zafarraya stratigraphic sequence seem to confirm a moist, temperate climate, cooler than today, with a wetter initial phase when the forests spread, but generally characterized by Mediterranean conditions. The various Neanderthals occupations occurred during the more or less fresh and more or less temperate periods, always at the start of or during the summer, when leopards and dholes arrived. Taphonomy The taphonomic and zooarchaeological analysis of Zafarraya cave gave rise to a proposed model in which the Neanderthal occupants of the cave became specialist hunters of goats, the most abundant species in the record (Barroso and de Lumley, H. 2006). However, the wide diversity of herbivore prey that inhabited the environment and the parallel presence –and continued occupation– of carnivores and Neanderthals prevent this model from fully reflecting the interaction between the two. Their overlap in the cave may have occasionally led to competition for food resources in the environs. The interactions between Neanderthals, carnivores and herbivorous prey has been the subject of a recent study (Caparros et al., 2012). 469
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  moister climate at the time reduced the sn...
470 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD There are three possible scenarios: 1) Neanderthals predominated over carnivores and excluded them from the surrounding ecological system, or vice versa. 2) Once the Neanderthal and carnivore niches overlapped, one excluded the other from the immediate ecological environment. 3) Once the two niches overlapped, neither was able to exclude the other and they coexisted. In order to determine which of the three was most likely, we studied the overlapping Neanderthal/carnivore occupation of the cave from the perspective of their interaction with the herbivore prey items and also their competition for Capra, the most abundant resource. Using the zooarchaeological analysis and the known ecology of the taxa which fed in the vicinity of the cave, a path analysis (Wright, 1968) was applied to the faunal data to define the likely fauna acquisition strategies by Neanderthals and carnivores. Initially, the applicable scenario seems to be a complete overlap of the Neanderthal and carnivore niches, both cohabiting the cave environment. The analysis shows that: 1) Although the continued presence of Capra made it a prey for both groups, Panthera was the main predator, alternating with the Neanderthals in the cave occupation. 2) Rupicapra was overwhelmingly hunted by Panthera and rarely by Neanderthals. 3) Neanderthals were the main accumulators of Cervus, hunted at some distance from the cave and probably scavenged by the carnivores. 4) Large herbivores such as horses and aurochs were hunted by Neanderthals and secondarily by carnivores. In the vicinity of the cave, Panthera was clearly more efficient than the Neanderthals, who were excluded when this carnivore was present (Capra and Rupicapra remain in the record). The large herbivores that roamed the Zafarraya polje probably offered good opportunities for cooperative hunting by Neanderthals (accumulation of Cervus and others). If there was indeed coexistence, as evidenced by scavenging, it must have been during short periods of random or seasonal occupation. Dating Following the discovery of the B.Z.C. site, preliminary studies were presented in several publi- cations, where the age of the Mousterian levels were estimated to be more recent than other Middle Palaeolithic sites in the region. This first estimate was based on a biostratigraphic study which compared rodent populations found in the Zafarraya sediments and those in Cariguela. This latter site was considered contemporary to “Würm II”, permitting the inference that Zafarraya could be the “Würm II-III” or “Würm III” interstadial. In 1989, Barroso et al., forwarded the open hypotheses that the Zafarraya fossils might represent Europe’s last Neanderthal populations. The first series of radiocarbon datings from the Mousterien levels of B.Z.C. were presented in 1995 (Hublin et al., 1995). Two methods were used, C-14 (based on the acid fraction and the collagen) and the U-Th method, using herbivore bone samples. The ages for C-14 (collagen) ranged from 29.8 + 0.6 ky (Gif-9140-II) and higher levels of 31.8 + 0.55 ky (Gif/LSM9140-I) level 8, while the acid fraction, for the same samples gave 23.6 + 1 + 0.1 ky and 22.0 ky. The U-Th ages are stratigraphically consistent, ranging from 25.1 + 1.3 ky on level I-3 to 31.7 + 3 ky on level I-8. Level D, the location of one of the Neanderthal fossils (Zaf. 2), was dated by U-Th at 33.4 + 0.2 ky. These absolute dates showed that between 30 ky and 33 ky BP, groups of Neanderthals survived the pressure of A.M.H. in a sort of “refuge” at the southern extreme of Iberia. The hypothesis of the “last Neanderthals” in Zafarraya thus seemed to be confirmed at this site, which gave rise to the paradigm. A new set of absolute datings began almost a decade later using several methods (Michel et al., 2003, Michel et al., 2006). C-14 datings (AMS) of 20 charcoal samples by ORAU yielded quite mixed results because the collection methodology for the 1996 samples was not excessively rigorous. Samples from stratigraphic sections in grids which were never dug, the source of the problem were: Q-9, P-9, R-9, P-20 and R-1. These samples, with the exception of R-16, yielded aberrant dates. The charcoal samples located and extracted from the stratigraphic sections under controlled conditions during the dig were: P10395, Oxa9001, P10397, P10398, Oxa-9002 and P10400. This methodological control makes these samples highly reliable with respect to their spatial location and stratigraphic context.
470  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Ther...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Eight faunal remains were chosen for C-14 (AMS) dating, although half the samples were rejected due to lack of collagen. The samples at depths from 168 cm to –171 cm yielded three ages ranging from (OxA 8411) 26.36 + 0.44 ky (Z6d) to (OxA 8999) 33.3 + 1.2 ky (Z8os). A sample from –223 cm yielded an incongruous dating, (OxA 9000) 30.65 + 0.65 ky (Z69os). U-Th alpha spectrometry was applied to 25 herbivore teeth. The results were not conclusive, as six ranged chronologically from 29,300 + 1.4-1.3 (Z4cem) to 39,700 + 2.2 (Z126os) at depths between –133 cm and –239 cm. Eight of the samples were between 40.9 + 1.3 ky (Z2os) and 60 + 3.7-3.6 ky (Z205os) at similar depths. (EU) and 24 + 0.4 ky (LU) at –165cm, while the oldest (Z126) 44 + 0.5 ky (EU) and 45 + 0.6 ky was from –178 cm. The weighted mean of the most consistent results yielded by this method, i.e. above 30 ky, is 35 + 0.3 to 39 ky + 0.3 ky. The age of Neanderthal mandible Zaf. 2 could not be defined by high resolution gamma spectrometry direct dating due to its low uranium content. The imprecise conclusion was that it must be at least 42 ky old. The U-Th thermal ionization mass spectrometry (TIMS) was applied to 9 samples, two of which were invalidated. The remaining 7 showed a range between 16.88 + 0.535 ky (Z6d.a) at 171cm and 53,345 + 5,025 (Z696e) at –133 cm. At the same depth, alpha spectrometry dated the –170 cm samples at 31.6 + 203 and for –133cm, at 32.8 + 0.201 –0.2 ky. The obvious contradictions prevent conclusive results from being reached. In 2013, an article was published on the Zafarraya datings based mainly on the calibration of the 2006 AMS dates (Michel et al., 2013, (Table 1). Samples OxA-7117 were 7 metres from Neanderthal jaw Z-2 (found at a depth of –194 cm, and in the same stratigraphic unit D, with a calibrated C-14 age of 38.7 ky to 44,117 ky. In 1995, level D level was dated at 33.4 ± 0.2 ky, which proved fundamental in consolidating the theory of the last Neanderthals in Zafarraya. The ESR (EU, early uptake) and ESR (LU, linear uptake) method was used with 6 horse and goat teeth. The result was (Z62) 24 + 0.3 ky Sample OxA-9002 is from UF35 (-213 –218 cm), the floor of the Neanderthal occupation where several human remains were found: Zaf.4- The final conclusion was that the results showed a chronological range from 30 ky to 45 ky, thus leaving the possible late age of Neanderthals in southern Spain unresolved. SAMPLE LABORATORY Nº DEPTH (cm.) DATA (C-14 ky BP) DATA (C-14 cal. ky BP) Q-14, 3 UE31 P10395 – 182 FAILED – P-11, 70 UG36 D OxA– 7117 – 194 36,9 + 3,0 38,700 – 44,117 Q-14,1 UE34 P10398 – 210 FAILED – R-16, 1 UE? OxA-9002 – 217 34,6 + 0,8 38,763-40,604 Q-17, 202 UG42 E OxA-7120 – 234 30,9 + 1,3 34,038 – 36,839 Q-18, 167 UG42 E OxA-7135 – 235 14,93 + 0,09 18,262 – 18,489 Q-17, 3 UG44 P10400 – 240 FAILED – Table 1. Charcoal samples in precise stratigraphic context, with AMS radiocarbon dating. Both uncalibrated ages (BP) and calibrated ages (cal. BP) are shown, as per Michel, V 2013. 471
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Eight faunal remains were chosen for C-14 ...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 472 5-18, Zaf.15, Zaf.16, Zaf.27, Zaf.17 and Zaf.19. This level has been dated at 38,763 to 40,604 ky. Both this and the previous data seem quite reliable, allowing the Zafarraya Neanderthal fossils to be dated between 38.7 ky and 44 ky. threshold of the method, showed an age greater than 46.7 ± 0.7 ky, i.e., roughly 13,000 years older. The most recent publication on this issue (Wood et al., 2013), present results from Zafarraya. The author applied extremely strict protocols such as ultrafiltration pretreatment (AF) of bone collagen to prevent contamination of the processed samples. Previously, the nitrogen content (% N) of the selected bone sample is measured to determine the collagen content, which is used to decide whether or not to radiocarbon date the sample. Only 6 out of 33 analysed samples contained a % N that would permit collagen to be obtained, and only three of these specimens could be dated, less than 10% of all the selected samples. This methodology was also used at other sites along the Mediterranean coast: Les Cendres (3 samples), Mallaetes (2 samples), El Niño (23 samples), Quebrada (9 samples), Sima de las Palomas (1 sample), El Salt (43 samples), Nerja (25 samples) and Gorham’s Cave (73 samples), all with negative results. Two stages were used in the Zafarraya analysis. In the first stage, bone samples previously dated by C-14 (AMS) at the ORAU (OxA-8024, OxA-8999, OxA-9000, OxA-8411) with results published in Michel et al., 2003 and Michel et al., 2006.) were chosen. Only sample Z8os, which was subject to the AF protocol, yielded a dating of 33.3 + 1.2 ky. The rest had a low % N and were rejected. In the second stage, 29 bone samples from the stratigraphy were analysed. Only two had an acceptable %N, ZAF2 (0.8) and ZAF5 (1.5). and 0.8 Another three samples, ZAF3 (0.7) ZAF7 (0.6) and ZAF8 (0.6), were on the borderline. The AF protocol was applied to these samples, yielding results in ZAF2 (46.3 ± 2.5 ky) and ZAF7 (49.3 ky). The latter, almost on the These newer results clearly question not only the chronological framework on which the evidence of the Middle Palaeolithic in the Spanish Mediterranean was based, but also the very existence of the last Neanderthals in their “refuge” in southern Iberia. This paradigm is thus now under serious challenge, and the existence of these late hominids is difficult to sustain at present. After Zafarraya, Gorham’s Cave was presented as further near-indisputable proof for the paradigm, however in the light of the data provided above, this can no longer be sustained. Other challenges or scenarios must be investigated to explain what seems to have been a “depopulation” of southern Iberia from 40 ky until the arrival of the Gravettian groups. The work by both Michel and Wood underscores the age of the Mousterian levels at Zafarraya, between 38.7 ky and 49.3 ky. THE LOWER AND MIDDLE PLEISTOCENE IN THE GUADIX-BAZA AND ORCE BASIN Introduction and geological context Robert Sala Ramos* The Guadix-Baza region spans a large area in south-eastern Iberia. It was a vast intramontane endorheic basin between the Miocene and the end of the Middle Pleistocene, when regional tectonics led to the formation of the Guadiana Menor River which drained the former lake system. The first * hominins who settled in Western Europe occupied this area at a time when the Guadix sub-basin was crossed by a hydrological system, descending from Sierra Nevada and Sierra de Cazorla and flowing into a lake in the Baza sub-basin in the eastern part of the region. This 3,000 km2 basin lies between the Internal and External Zones of the Betic Range. The Cenozoic infill was initially composed of marine Universitat Rovira i Virgili. Institut Catala de Paleoecologia Humana i Evolucio Social.
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  472  5-18...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Figure 1. Geological map of the Iberian Peninsula and location of the Guadix-Baza basin. (© Oriol Oms). Figure 2 Palaeogeography of the ancient Guadix-Baza river-lake basin at the time of the first human settlement in Orce (© César Viseras, Viseras et al., 2006). sediments, and covered during the upper Miocene by continental deposits. The Orce sector is at the north-eastern end of the basin, with outcrops of Pliocene and Pleistocene alluvial and lacustrine sediments assigned to the Baza formation (Vera et al., 1984, Soria et al., 1987; García 1997, Oms et al., 2009). The area contains many palaeontological and archaeo-palaeontological records spanning the same period and beyond, including remains from the Upper Pleistocene and the Holocene. Three sites in the Orce municipality have an Early Pleistocene record, with evidence of the oldest known human presence in Western Europe. They are providing deep knowledge about primitive human adaptation to the continent and the evolution of these hominins’ behaviour. These three sites, Venta Micena, Barranco León and Fuente Nueva 3, lay on the south-eastern shore of the ancient Baza Lake. The former contains a purely palaeontological record while the other two contain evidence of the earliest known human presence in Western Europe. Most of this chapter focuses on these three sites. However, in order to provide a clearer vision of the prehistoric wealth and importance of the entire region, we have also included four other sites in the Guadix-Baza basin which testify to the diachrony of human evolution in this territory. Figure 3. Reconstructed geomorphological and palaeoecological location of the three main Orce sites (© Oriol Oms). 473
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Figure 1. Geological map of the Iberian Pe...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 474 Bienvenido Martínez Navarro*,**,***, María-Patrocinio Espigares Ortiz****,*****, Sergio RosMontoya****,*****, Paul Palmqvist **** 1. Venta Micena Since the discovery of site of the Venta Micena in 1976, and after nearly four decades of continuous survey and research, the region of Orce (Guadix-Baza Basin, southeastern Spain) has provided one of the most important Early Pleistocene collections of large mammals from the European continent. The Venta Micena stratum is an 80-120 cm thick horizontal layer of micritic limestone whose outcrop can be followed for more than 2.5 Km along the surface (Palmqvist et al., 2008a). Until now, more than 17.000 fossil bones have been unearthed at this site. The chronology of the site has been obtained by combining paleomagnetic data with the biochronological evidences. It is placed within the Allophaiomys ruffoi Zone, located below the Allophaiomys aff. lavocati Zone (ex. Allophaiomys bour- Figure 4. Panoramic of Venta Micena during the field season of 2005. * ** ICREA, Barcelona, Spain Institut Català de Paleoecologia Humana i Evolució Social-IPHES, C/ Marcel.lí Domingo s/n, Edifici W3, Campus Sescelades, 43007 Tarragona, Spain *** Àrea de Prehistoria, Universitat Rovira i Virgili, Avda. Catalunya 35, 43002 Tarragona, Spain **** Departamento de Geología y Ecología, Universidad de Málaga, Campus de Teatinos, 29071 – Málaga, Spain. ***** Museo de Prehistoria y Paleontología, 18858 – Orce (Granada) Spain
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  474  Bien...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS gondiae Zone) and above the Mimomys oswaldoreigi Zone (Oms et al., 2000; Agustí and Madurell, 2003). The large mammal assemblage from Venta Micena is different in composition and younger than the ones from the Italian sites included within the Tasso faunal unit in the Upper Valdarno (Rook and Martínez-Navarro, 2010, and references there in). However, Venta Micena and the Georgian site of Dmanisi, dated around 1.8 Ma (Lordkipanidze et al., 2007), show important similarities between both assemblages, such the presence of the bovid species Soergelia minor (Moyà-Solà, 1987; Vekua, 1995; Buhksianidze, 2005) which is also present at the Greek site of Apollonia-1, named there Soergelia brigittae (Kostopoulos, 1997), and at Italian site of Argentario (Martínez-Navarro et al., 2012). Given that this species is not recorded in the latest Late Villafranchian assemblages, it suggests an earlier age for the Spanish site. Therefore, it is possible to conclude that Venta Micena is situated above the Olduvai subchron, with an age around 1.5±0.1Ma. This chronology is also supported by new data based on ESR methodology (Duval et al., 2011). The study of the bone assemblage from Venta Micena has helped to detect, among the members of the Late Villafranchian fauna, the presence of African origin species, such as the sabre-tooth Megantereon whitei and the mega ungulate Hippopotamus antiquus (Martínez-Navarro, 1991; Alberdi and Ruiz-Bustos, 1985; Martínez-Navarro and Palmqvist, 1995; Palmqvist et al., 2007; MartínezNavarro, 2010). However, it is worth noting that the large mammal assemblages from the Orce sites are basically composed of Holarctic origin species, including equids, rhinos, bovids, cervids, and most of the carnivores, with a faunal list composed by 21 species of large mammals (Table 1). Taphonomic study of Venta Micena shows that most herbivore remains come from carcasses of ungulate prey selectively hunted by hypercarnivores (coursers Homotherium latidens and wild dog Lycaon lycaonoides; ambushers Megantereon whitei and jaguar Panthera cf. gombaszoegensis) and scavenged afterwards by the giant bone-cracker hyena Pachycrocuta brevirostris. Modifications produced by the tooth of this hyena follow specific patterns of bone consumption for Figure 5. Accumulation of bones in the Corte III excavation area of Venta Micena. 475
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  gondiae Zone  and above the Mimomys oswald...
476 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 6. Tooth marks on bones from Venta Micena. 1982 of the skull fragment VM-0, a small and problematic piece that was classified as Homo sp.(Gibert et al.,,1983). However, after a long polemic on the supposed human affinities of this fossil, it was concluded that it belonged to a female of a ruminant without cranial appendages, horn cores or pedicels (Martínez- Navarro, 2002). At the moment, no other clear human remains or other evidence of human activity have been found at the site of Venta Micena (Espigares, 2010). Nevertheless paleoecological study of this assemblage, based on combined taphonomic, biomechanical, ecomorphological and biogeochemical approaches, has provided compelling evidence on the environmental context inhabited by the first human populations that dispersed in Western Europe. VENTA MICENA each anatomical element of the ungulate skeleton. Pits, scores, furrowing and crenulated edges are present in a great number of elements, and the analysis of the fracture edges in long bones are consistent with fractures produced when bones are fresh. These features, together with the conservation pattern of anatomical regions in the bones allow us to confirm that Pachycrocuta was the main accumulator agent of this site. Some taphomonic biases, that support this interpretation, has been detected in the assemblage as: (a) the kleptoparasitism of ungulate carcasses by Pachycrocuta; (b) the transport of ungulate prey as whole carcasses or anatomical parts, depending on the size of the species scavenged; and (c) the selective fracturing of bones as a function of their marrow contents and mineral density. Ecomorphological inferences allowed to classify ungulate species as grazers in open environment (Equus altidens, Hemitragus cf. albus, Bison sp., Praeovibos sp., Hippopotamus antiquus), mixed feeders (Hemibos aff. gracilis, Soergelia minor, Metacervocerus rhenanus and Mammuthus meridionalis) and browsers in forested habitat (Stephanorhinus aff. hundsheimensis and Praemegaceros cf. verticornis). The comparison of the abundance of these ecological categories with those present in modern communities show that the fauna of Venta Micena inhabited a mixed environment with a predominance of open plains in the surroundings of a lake fed by hydrothermal waters, similar to the Rift Valley in East Africa. The presence of several human remains at Venta Micena has been proposed since the discovery in Ursus etruscus Lycaon lycaonoides Canis mosbachensis Vulpes praeglacialis Pachycrocuta brevirostris Megantereon whitei Homotherium latidens Panthera gombaszoegensis Lynx sp. Meles meles. Mammuthus meridionalis Stephanorhinus aff. hundsheimensis Equus altidens Hippopotamus antiquus Bison sp. Hemibos sp. aff. H. gracilis Soergelia minor Praeovibos sp. Hemitragus albus Praemegaceros cf. verticornis Metacervocerus rhenanus Table 1. Large mammals lists from Venta Micena (after Martínez-Navarro, 1991; Martínez-Navarro et al., 2010, 2011, and references therein).
476  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Robert Sala Ramos* Leticia Menéndez Granda*, Sergio Ros Montoya**, Isidro Toro Moyano*** 2. Barranco León Human activity during the Early Pleistocene at Orce has been well described at two sites on the south-eastern edge of the old Baza lake basin. Barranco León and Fuente Nueva 3, first excavated in the 90’s, are two well known sites which have been largely responsible for much of what we know about early human settlement in Western Europe more than one million years ago. Since 2009, the sites are currently under study by a multi-disciplinary, multi-institutional team led by the Catalan Institute of Human Palaeoecology and Social Evolution. The Barranco León site (UTM 548.400 / 4.175.340, 975 m a.s.l.) is located about 3 km from the village of Orce, on the María-Vélez highway. It is located in a north-south gully, at the foot of Sierra Umbria, and ends at the Vélez glen. This area has been well known since the 1980s thanks to the abundance of palaeontological remains (Anadón et al., 1987). The site was discovered in 1983 during a survey in search of micromammals, although it was only in 1992 when A. Arribas did a lithostratigraphic study of the site and discovered the importance of archaeo-palaeontological level BL5 (now level D). This led J. Gibert to apply to the Government of Andalusia for permission to conduct an initial excavation in 1995. Early work Figure 7. Excavation at the Barranco León site (© Jordi Mestre. IPHES). * ** Universitat Rovira i Virgili. Institut Català de Paleoecologia Humana i Evolució Social Departamento de Geología y Ecología, Universidad de Málaga, Campus de Treatinos, 29071 – Málaga, Spain. Muso de Prehistoria y Paleontología, 18858 – Orce (Granada) Spain. *** Museo Arqueológico de Granada. 477
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Robert Sala Ramos  Leticia Men  ndez Grand...
478 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD brought to light several stone tools and remains of horses and hippopotami (Turq et al., 1996). In 1999 and 2000, a new team led by G. Martínez and I. Toro resumed systematic excavation at the site, headed by I. Toro after 2001. The final dig prior to the new research project was in 2005. Since 2010, a new team led by the Institute of Human Palaeoecology and Social Evolution has been responsible for research in the Orce region. A total area of roughly 200m2 has been excavated at Barranco León in the course of all this work. 2.1. Stratigraphy and dating The main archaeological record at the Barranco León site is concentrated in detrital horizons lying on dark clayey and sandy facies which form the basis of the excavation. Above these clay facies there are firstly autochthonous carbonated gravels linked to unchannelled bedload inputs. Resting on these gravels are massive sands with floated cobbles plus fauna and industry. In addition, canaliform fine gravels culminate in material with low-angle cross stratification. The sand facies overlie massive levels with medium and fine gravels decreasing in thickness, fining eastward, suggesting that the direction of the input was oblique (SW-NE) to the current canyon, which is more N-S. The channelled facies gave rise to erosive bases with soft cobbles from the lower units (sand and clay below). These channelled facies derive laterally, eastward to fine sands corresponding to the channel bank containing in situ archaeological material. The current project has highlighted the existence of such fine sediments and the good preservation of the anthropic record therein. The top of this fluvial sequence consists of silty, sandy facies rich in organic matter from the marsh-lake environment. There are several hydroplastic deformations which had a considerable effect on the sand+gravel channel facies. The part of the Barranco León sequence containing the anthropogenic occupation thus shows alternating lake and channel and bank levels, indicating different transgressive and regressive phases in the level of the lake and the human occupation of a channell bank close to that lake. Palaeomagnetism and Electro Spin Resonance (ESR) datings situate Barranco León between 1.2 and 1.4 million years (Oms et al., 2000; Duval et al., 2012), while the mammal biochronology and the more recent dating suggests it is closer to the lower end of this range (Toro et al., 2013). Figure 8. Barranco León stratigraphic sequence (© Oriol Oms).
478  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  brou...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Robert Sala Ramos*, Leticia Menéndez Granda*, Patrocinio Espigares**, Isidro Toro Moyano*** 3. Fuente Nueva 3 The site’s UTM coordinates are 522.490/ 4.174.885, on 1:50,000 Army Sheet 23-28. Discovered in 1991 by Alain Bocquet, there have been several emergency excavations and systematic digs in 1993, 1994 and 1995. Systematic annual excavations took place from 1999 to 2003, as well as in 2005 and 2006. As in Barranco León, the current project began in 2010. In all, roughly 75 m2 have been excavated in two main areas- the majority in the northern part while the area in the south-west quarter is roughly half the size of the other. 3.1. Stratigraphy and dating The stratigraphic sequence containing evidence of anthropogenic activity at this site starts at the bottom with a whitish basal limestone package (level 1) and a nodular and/or brecciated appearance, in irregular contact with the next level. The sequence continues with a level composed of green-grey clay (level 2), then grey-green carbonated clay and white silt (level 3), followed by another level of brecciated limestone (level 4) whose Figure 9. Excavation at Fuente Nueva 3 (© Jordi Mestre. IPHES). * ** irregular surface caused by erosion and brecciation facilitated the formation of pockets which include the clays on level 5. The entire sequence is affected by numerous slikensides which have led to the discovery of fossils in a vertical position. Above and to the west there is a thin layer of dark brown clay (level 6), below a layer of greenish-brown sludge on levels 7-11, which are followed by the limestone ceiling. The aforementioned distinctive levels can be distinguished within the two main upper and lower units of the fossil record. Many of them contain archaeo-palaeontological remains which, for the purpose of clarifying the anthropic activity here, can be assigned to the following individual horizons or levels. In addition to these levels, there are at least two more at a higher level, identified in an extension of the 2013 excavation area. Recent work has increased the number of units containing lithic industry. In addition to Units 1, 2 and 5, where human activity was previously described, remains have been exhumed in technical units 3 and 7, which also contain significant archaeological and palaeontological material. In order to fine-tune the definition of the possible horizons of human activity and/or carnivores, we are currently subdividing Unit 5. This unit –basically greenish loamy sands and shales– can be locally subdivided into 5a, 5b and 5c. At some points, these subdivisions cannot be detected due to the lack of calcrete nodules (one of the indicators), the size of the bones or the shape of the layers, heavily affected by irregularities in the ceiling of Unit 4. The latter factor means that some –or all– of the three subunits can be hard to identify. They are distinguished by the following features: – 5a contains a large amount of palaeontological remains with varying degrees of al- Universitat Rovira i Virgili. Institut Català de Paleoecologia Humana i Evolució Social Departamento de Geología y Ecología, Universiad de Málaga, Campus de Teatinos, 29071 – Málaga, Spain. Museo de Prehistoria y Paleontología, 18858 – Orce (Granada) Spain. *** Museo Arqueológico de Granada. 479
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Robert Sala Ramos , Leticia Men  ndez Gran...
480 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 10. Fuente Nueva 3 stratigraphic sequence. (© Oriol Oms). teration, forming a mass of mainly disarticulated bones (e.g., in the case of the main proboscid tusks). – 5b is a distinctive level due to the large amount of reworked calcrete nodules. – 5c has few nodules, but does contain ‘the’ complete, articulated skeleton of a Mammuthus meridionalis. In the coming years we plan to complete a geological and sedimentological reconstruction to facilitate an integrated 3D view which will be useful not only as a geologic framework for the units containing lithic industry and palaeontological remains, but also to understand the genesis of the site (or at least the agents and natural factors involved). The latest ESR datings from equid teeth are 1.2Ma BP for the fertile levels at this site (Duval et al., 2012).
480  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS María-Patrocinio Espigares Ortiz*,**, Sergio Ros-Montoya*,**, Paul Palmqvist*, Bienvenido Martínez Navarro***,****,***** 4. The faunal assemblage of Barranco León The archaeopaleontological site of Barranco León, dated 1.4 Ma (Toro-Moyano et al., 2013) shows a typical Late Villafranchian assembblage, mostly composed of large mammal species: eight carnivores (one ursid, three canids, two mustelids, one hyaenid, and one felid), one proboscidean, three perisodactyls (two equids and one rhinocerotid) and five artiodactyls (one hippopotamid, two cervids and two bovids) (Table 1). Carnivores are represented only by isolated teeth from adult individuals and are scarcely recorded, whereas herbivores represent the greater part of the bones unearthed. It is especially important the record of a first deciduous molar of early Homo, specimen BL02J54-100, which was found in direct association with lithic artefacts and large mammal bones. It represents the oldest anatomical evidence of human presence in Western Europe. This finding, Figure 11. Antler of Praemegaceros aff. Verticornis (© Sergio Ros). * ** *** **** Figure 12. Third metatarsal of Equus altidens. Figure 13. Occlusal view of a first deciduous molar of Homo sp. recorded at Barranco León (Toro-Moyano et al., 2013). Departamento de Geología y Ecología, Universidad de Málaga, Campus de Teatinos, 29071 – Málaga, Spain. Museo de Prehistoria y Paleontología, 18858 – Orce (Granada) Spain ICREA, Barcelona, Spain Institut Català de Paleoecologia Humana i Evolució Social-IPHES, C/ Marcel.lí Domingo s/n, Edifici W3, Campus Sescelades, 43007 Tarragona, Spain ***** Àrea de Prehistoria, Universitat Rovira i Virgili, Avda. Catalunya 35, 43002 Tarragona, Spain 481
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Mar  a-Patrocinio Espigares Ortiz ,  , Ser...
482 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD combined with the important lithic tool assemblage, confirms that Western Europe was colonized less than 0.5 Ma after the first expansion of Homo out of Africa. From the paleontological point of view this assemblage is very similar to that documented in the nearby localities of Fuente Nueva-3 and Venta Micena. The absence of Soergelia minor, a middle sized, mesodont bovid recorded at Venta Micena, and the presence of the new species of Asian origin Equus sussenbornensis suggests a process of aridification during the late Late Villafranchian, previous to the arrival of the classical Galerian species. It is remarkable the record of megaherbivores, proboscideans are practically absent from this site while hippos are very abundant. This difference with Fuente Nueva-3, where elephants are one of the best represented species, was probably due to the different sedimentary environment in both sites, a palaeochannel for Barranco León and a swampy area for Fuente Nueva-3. In addition to the lithic artefacts and the human tooth, abundant cut-marks and percussion marks, that evidence bone fracturing for accessing marrow contents, have been identified in this site. Cut marks localized on the diaphyses of the major limb bones may be related to defleshing activities. Evidences of disarticulation and evisceration were inferred by cut marks recorded in metapodials, vertebrae and internal surface of ribs. The skeletal elements bearing cut marks are mainly included in the large size category (340-907 kg). Fractured bones show frequently spiral fractures, impact points and flaking produced when they were broken by percussion (Espigares et al., 2008; Espigares, 2010). Evidence of carnivore activity is scarce and consists mainly of scores and pits located in metapodials and carpal/tarsal bones. Comparative taphonomic analyses of this assemblage with the pattern of modification of bones by Pachycrocuta brevirostris developed in Venta Micena, reveal, a similar morphology of the tooth marks than that produced by this hyaenid. Presence of P. brevirostris in Barranco León is supported by the record of abundant coprolites and some isolated teeth, although anthropic activity predominates in Barranco león with a punctual and secondary access of carnivores to these areas. Barranco León D Homo sp. Ursus sp. Canis mosbachensis Lycaon lycaonoides Vulpes cf. praeglacialis Pachycrocuta brevirostris cf. Homotherium sp. Meles meles cf. Pannonictis sp. Mammuthus meridionalis Stephanorhinus hundsheimensis Equus altidens Equus suessenbornensis Hippopotamus antiquus Bison sp. Hemitragus albus Praemegaceros cf. verticornis Metacervocerus rhenanus Table 2. Faunal list (large mammal species) for the archaeological levels of Barranco León (Oms et al., 2000; Martínez-Navarro et al., 2003, 2010; Palmqvist et al., 2005; Toro et al., 2013 and references there in).
482  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  comb...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS María-Patrocinio Espigares Ortiz*,**, Sergio Ros-Montoya*,**, Paul Palmqvist*, Bienvenido Martínez Navarro***,****,***** 5. The faunal assemblage of Fuente Nueva-3 The archaeopaleontological site of Fuente Nueva-3, dated 1.2-1.3 Ma (Martínez-Navarro et al., 1997; Oms et al., 2000; Martínez-Navarro et al., 2010, Duval, 2011), show a typical Late Villafranchian assemblage, similar to that of Venta Micena and is mostly composed of large mammal species: nine carnivores (one ursid, three canids, three mustelids, one hyaenid, and one felid), one proboscidean, three perisodactyls (two equids and one rhinocerotid) and six artiodactyls (one hippopotamid, two cervids and three bovids) (Table 1). It is interesting the absence of Soergelia minor, a middle sized and mesodont bovid that is recorded at Venta Micena, and the presence of new immigrants of Asian origin, especially Equus sussenbornensis (Alberdi, 2010) and Ammotragus europaeus (Moullé et al., 2004). Both species are typical grazers and inhabited open environments, which suggests a process of aridification during the late Late Villafranchian, previous to the arrival of the classical Galerian species. Two archaeopaleontological ensembles have been described in this site: the Upper Level and the Lower Level. The faunal composition of both is the same, the main differences are located in the abundance of lithic artefacts associated with skeletal remains of large mammals (Toro et al., 2011), more abundantly represented in the Lower than in the Upper Level, which suggests a marginal occupation of the latter by the hominins and a difference in the preservation of the anatomical parts. Other intriguing difference is the abundance of Mammuthus meridionalis remains in the Upper Level, including a semi-complete carcase and five tusk, as well as abundant molars and tooth fragments. In addition to the lithic artifacts, there is evidence of anthropic modifications on the bones’ * ** *** **** Figure 14. Semi-complet skeleton of Mammuthus meridionalis. Figure 15. Hemimandible of Stephanorhinus hundsheimensis together with some ribs and fragment bones from large mammals. Departamento de Geología y Ecología, Universidad de Málaga, Campus de Teatinos, 29071 – Málaga, Spain. Museo de Prehistoria y Paleontología, 18858 – Orce (Granada) Spain ICREA, Barcelona, Spain Institut Català de Paleoecologia Humana i Evolució Social-IPHES, C/ Marcel.lí Domingo s/n, Edifici W3, Campus Sescelades, 43007 Tarragona, Spain ***** Àrea de Prehistoria, Universitat Rovira i Virgili, Avda. Catalunya 35, 43002 Tarragona, Spain 483
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Mar  a-Patrocinio Espigares Ortiz ,  , Ser...
484 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD surface, cut and percussion marks were identified in skeletal elements mainly included in the large size category (340-907 kg). Cut marks located on the diaphyses of the major limb bones and on the external surface of the ribs may be related to defleshing activities. There are also cut marks on a pelvis fragment and several metapodials, which were probably produced during the disarticulation of these elements. Finally, there is also a scrape mark which was probably originated during the removal of the periosteum before fracturing the bone. Fracturing patterns in the skeletal elements indicate that fresh bones were broken mainly by percussion. As a result, they show spiral fractures, impact points and flaking (Espigares et al., 2008; Espigares, 2010). Evidence of carnivore activity on the cortical surface is scarce and consists mainly of scores and pits. However, the presence of carnivores in the Upper Level will probably increase in the near future, because more than a 150 coprolites of P. brevirostris have been documented, which show the intense activity of the giant hyenas during the time of deposition of this level, that has provided the oldest evidence of a probable direct competition between Homo and Pachycrocuta, the two major bone modifying and accumulating agents during Early Pleistocene times in Eurasia. The evidence lies in the finding of an incomplete skeleton of M. meridionalis surrounded by 34 coprolites and 17 lithic artifacts. The skewed spatial distribution of these elements, the physical characteristics of the coprolites and the absence of the elephant limbs and cranium suggest that both hominins and hyenas scavenged the carcass of this megaherbivore, following a sequence of consumption in which the hominins arrived first, dismembered and transported the limbs, and probably also the cranium, and later the hyenas consumed the rest of the elephant carcass (Espigares et al., 2013). Fuente Nueva 3 Ursus sp. Canis mosbachensis Lycaon lycaonoides Vulpes cf. praeglacialis Pachycrocuta brevirostris Felidae indet. Lynx sp. Meles meles Pannonictis cf. nestii Mustelidae indet. (small size) Mammuthus meridionalis Stephanorhinus hundsheimensis Equus altidens Equus suessenbornensis Hippopotamus antiquus Bison sp. Ammotragus europaeus Hemitragus albus Praemegaceros cf. verticornis Metacervocerus rhenanus Table 1. Faunal list (large mammal species) for the archaeopaleontological levels of Fuente Nueva-3 (Martínez-Navarro et al., 1997, 2003, 2010; Oms et al., 2000; Palmqvist et al., 2005; Espigares et al., 2013 and references therein).
484  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  surf...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Deborah Barsky *,**, Robert Sala **,*, Isidro Toro ***, Leticia Menéndez Granda*, ** 6. The Lithic industry from Barranco León and Fuente Nueva 3 The Barranco León and Fuente Nueva 3 sites have yielded some of the best preserved and most numerically rich Oldowan stone tool assemblages so far to have been discovered in Eurasia. The industries, excavated in situ from each stratigraphical sequence, are found in clear association with the remains of large and medium-sized herbivores and carnivores (Espigares et al., 2013). The unearthing of this exceptionally rich accumulation (BL=;1 722 pcs.; FN3= 1 264 pcs.) provides the opportunity to analyze one of the oldest known and most comprehensive lithic series in a rigorously controlled archeological, stratigraphical and chronological situation. Ongoing excavations and interdisciplinary studies considerably enlarge potential interpretations of the paleo-environmental context reigning at the time of the appearance of the first inhabitants of Western Europe, some 1,4-1,3 Ma (Toro-Moyano et al., 2010 a, 2010 b). The technological, typological and traceological features of the Orce assemblages are thus founded on a coherent and exceptionally complete data-set which is still increasing in pace with the progression of excavations and research. In future, spatial analysis of the archeological remains will add yet another horizon to what is presently known of the behavior and cognitive levels of the hominins frequenting the Orce paleo-lake basin. The lithic assemblages and their raw materials Only subtle differences are highlighted between the lithic assemblages from BL and FN3 and the two industries are considered to be largely analogous. Both of the toolkits were manufactured from local flint and limestone and they clearly belong to an Mode 1 tradition. Accordingly, they present a low degree of standardization: there are no handaxes, cleavers nor picks and the few existing * Institut Català de Paleoecologia Humana i Evolució Social ** Universitat Rovira i Virgili *** Museo Arqueológico de Granada secondary knapped flakes are attributed to cores rather than to retouched tools (Barsky et al., 2013, 2014). Knapping waste, mainly in flint, is composed of abundant and small-sized flakes. While cores are relatively scarce, all elements of the various reduction and use schemes identified at the sites are represented in the assemblages, proving that these activities were practiced in situ. In addition, a considerable proportion of the larger-sized limestone instruments bear marks of intensive use attributable to percussive or pounding activities. These voluminous tools include non-modified cobbles and blocks marked by traces of percussion and/or breakage impacts. Interestingly, there is a dimensional and functional dichotomy that appears to be rooted in the selective use of the two local raw materials: flint and limestone. These rocks are accessible in the immediate vicinity of the sites. These are the only kinds of rocks identified in the industries to date. Hominin raw material sourcing patterns were, therefore, strictly local. Both bipolar on an anvil and hard hammer methods were used for flake production and sometimes both techniques are recognizable on a single core. The use of bipolar on an anvil knapping is attested by clearly identifiable stigma on some of the cores (opposite impacts/negatives) and some flakes display prognostic traces; such as bulletlike morphologies or opposite impact points on their ventral surfaces. Experimental knapping has confirmed that this method is particularly well adapted to small flake production from tiny, cubeshaped flint matrixes identical to those found near the sites. It should be noted, however, that metersized flint blocks were probably also available from the outcrops and in secondary situations close by. The assemblages (particularly BL) also comprise large, quadrangular, flat limestone slabs displaying percussion stigma on their surfaces. Peripheral breakage with opposite impact points 485
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Deborah Barsky  ,  , Robert Sala   , , Isi...
486 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD are observed on these slabs, buttressing evidence for their use as anvils. Indeed, anvils would have been necessary both for bipolar reduction of stone and to effectively break some of the largersized bones. The industry equally includes the corresponding “anvil fragments” attesting that these activities were carried out on-site. The high item breakage ratio is explained by the force of the blows needed to carry out these activities, and also by the mediocre quality of much of the flint and the limestone. The flint Flint, outcropping from Dogger limestone formations located 5-15 Km to the south, was introduced mostly as flat, slightly rolled nodules originating from secondary deposits in the immediate vicinity of the sites (Turq et al., 1996; Toro Moyano et al., 2010 a; 2010 b). The scarcity of cores relative to flakes as well as a paucity of cortical elements may be related to the morphology of these brute flint nodules (intense breakage, partially covering cortex), but may also indicate that some of the knapping was carried out away from the sites. A few, larger-sized flakes were re-knapped; principally on their ventral surfaces. This could signal an ad hoc behavioral response to competition with other large carnivores present in the landscape and a need to expediently produce small, sharp-edged flakes for rapid and efficient meat acquisition. However, none of the flaked flakes correspond unambiguously with stereotyped retouched tools, such as notches, scrapers or denticulates (Barsky et al., 2010, 2013; Toro-Moyano et al., 2010 a, 2010 b, 2011). These secondary knapped flakes could be linked to bipolar on an anvil production (Zaidner, 2013). Core platforms were not prepared (facetted) and the plane surfaces of removal negatives most often served as platforms; giving way to polyhedral core forms. Knapping episodes were relatively long and the flint cores are generally maximally exploited. The bulk of the flakes correspondingly display unidirectional or orthogonally oriented removal negatives on their dorsal surfaces and smooth striking platforms. Small, cube-shaped cores with opposite removals show crush marks typically produced by this method. A few, somewhat enigmatic products (essentially at FN3) are attributed to pieces esquillées: they likely result from extracting flakes by bipolar method on an Figure 16. Top: small, elongated flint flake (BL 2013.D1.G49 nº 1). Bottom: limestone macro tool (BL 2011 D1.G51.16) with bipolar breakage and flat impact removals on the upper extremity (Photos Jordi Mestre. IPHES). anvil using blows effectuated from a summital crest. The limestone This raw material was collected from local alluvial sources and also from the substructure of the sites themselves. Different kinds of limestone outcrop in the immediate vicinity of the sites and was most probably transported there by water or even locally mined, displaced and collected by hominins: procurement was strictly local in all cases. Most of the archeological limestone cobbles display slightly rolled surfaces and are loosely compacted, indicating short-distance transport. The cobbles, blocks and stones have
486  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  are ...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS bearing on anthropic activities that are somewhat difficult to evaluate, especially since, in some instances (especially at BL), surface alteration impedes adequate readability of the material. In any case, careful study of the limestone (Barsky et al., in progress) reveals that hominins displaced, gathered and used it in different ways that are linked to criteria such as morphology, volume, weight, and petrographic quality. The widely variable nature of the limestone excavated from the archeological contexts has given way to a number of interpretations. At BL, for example, a fossil alluvial bed traversing the site in its southeastern extremity and contemporary to the occupation likely provided the main source of the limestone cobbles. This alluvial level is visible throughout the extension of the Barranco and could also have provided some fragmented flint nodules which erode out of an outcrop upstream. Contrastingly, fine grained siliceous limestone outcrops abundantly around the FN3 site and could have been opportunistically collected- or even locally mined. Limestone cobbles from other (yet to be localized) alluvial sources were also recovered and brought to the site. The presence of mediocre quality and poorly compacted or excessively large limestone items in the sites may therefore be at least partially explained by its introduction by natural forces: alluvial at BL and erosive at FN3. Siliceous limestone was preferentially used for controlled flake extraction and also for the manufacture of some loosely configured macro tools. The latter group includes chopper-like instruments and a few poorly standardized heavy-duty scrapers (as defined by Leakey, 1971). All of these macro instruments bear visible traces of use; mainly accidental removals and crush marks on their crests. At Barranco León, a few of the tools with abundant fracture angles (de la Torre and Mora, 2005) and removal negatives display a polyhedral morphology consistent with a sub-spheroid morpho-type. In any case, the differential quality of the limestone did not always prevent hominins from making some use of even the poorest quality rocks (Fig. 17, bottom right). The assemblages also contain a few considerably well-struck limestone flakes whose presence, alongside some organized core forms (orthogonal, multiplatform), indicates a dual functionality for the limestone: percussion and debitage. Figure 17. Top: small flint flake (FN3 2013.Nivel superior. T92.nº 26) with a cortical striking platform. Bottom: macro limestone tools (left: FN3 2010.R92 nº 98 and right: FN3 2011.P88. nº 28) with lateral unidirectional removals and crush marks on the edges (Photos Jordi Mestre. IPHES). Compared with flint, limestone matrixes often show a lower degree of transformation. The majority of each limestone assemblage is in fact made up of used or slightly modified cobbles and blocks, a large number of which have traces of percussion. The corresponding flakes are bigger than those in flint (3-4 cm). The anthropic nature of the traces on the tools is underlined by the consistency with which they are found to be localized in specific areas on the tools, in accordance with their shape and size. The kinds of traces documented vary considerably in both intensity and morphology. They include: accidental removal negatives, surface scarring, stigmata, irregular retouch, crushing, polish, bipolar breakage impacts, cupula, striations, facetted breakage and fracture angles (Barsky et al., in progress). This trace variability on the limestone macro tools surely indicates that a wider range of activities was being carried out 487
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  bearing on anthropic activities that are s...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 488 at the sites than previously attributed to Mode 1 hominin groups and constitutes a focus of ongoing research efforts. Synthesis The BL and FN3 industries, largely analogous, fall within the variability of the Oldowan or Mode 1 techno-complex sensu lato: local raw material procurement, poorly standardized cores and flakes and very few large or small intentionally shaped items. We underline however that, for each of these defining characteristics the industries show some ‘progressive’ features that could suggest some kind of continuity or evolutionary trend consistent with a move towards more complex technological behavior. This increase in innovative behavior could have been transmitted by progressive social integra- J.M. Jiménez Arenas* 7. Huéscar-1 The Huéscar-1 archaeo-palaeontological site (HU-1, 940 m.a.s.l.) is in the north-eastern sector of the Guadix-Baza intramontane basin, 4 km from Huéscar on the left bank of Barranco de las Cañadas ravine. The site was first excavated under A. Mazo, then by M.T. Alberdi (1986). Surface surveys were conducted by the Sabadell Institute of Palaeontology in the early 80s and again in 2003 by a team led by B. Martínez Navarro. The most productive work in palaeontological terms was the 1986 dig, particularly test pit , while from an archaeological perspective it was the 2003 survey, the only one to detect human presence in the form of stone knapping. Five levels have been defined at HU-1. The three levels containing vertebrates are, from bottom to top, Level 2, composed of superposed sand lenses, which decrease in grain size upwards, yellowish conglomerates, plus carbonates and flint clasts * tion (cultural transmission). We may highlight some of the more significant characteristics underlining the relative sophistication of the hominins present at the paleo lake Baza over a million years ago, such as: a comprehensive knowledge of the potential uses for the raw materials available to them, a clear functional dichotomy established upon this petrographical familiarity applied to optimize its potential, the systematic use of orthogonal knapping strategies for flake production, the capacity to adapt their flake knapping strategies to compensate deficiencies in rock quality, the presence of a few loosely configured macro tools, including heavy-duty scrapers and sub-spheroids, and, finally, a variability of on-site tasks (yet to be defined) is made manifest by the wide range of traces left on their tools. and occasionally Keuper hyacinths of Compostela; Level 3, composed of massive carbonate silts (carbonation increasing upwards) with mottled ochre and millimetric plant residues; and Level 4 composed of yellowish microconglomerates and sands. The published (Alonso Diago et al., 2003) faunal list consists of the following taxa: Pisces: Leuciscus pyrenaicus; Reptilia: Emydidae indet.; Aves: cf. Tachybaptus ruficolis, Anas crecca/A. querquedula, Anas platyrhynchos, Anas clypeata, Anas strepera, Anas sp., Netta rufina, Aythya ferina, Aythya nyroca, Aythya fuligula, Aythya sp., Perdix perdix, Crex cres, Bubo bubo; Soricidae: Soricidae indet.; Rodentia: Eliomys quercinus, Apodemus sp., Paraethomys meini, Castillomys crusafonti ssp., Mimomys savini, Microtus (Pitymys) gregaloides, Microtus (Microtus) brecciensis; Lagomorpha: Oryctolagus sp., Lepus cf. L. granatensis, Leporidae indet.; Carnivora: Canis etruscus, Crocuta crocuta, Vulpes sp., Vulpes prae- Dpto. de Prehistoria y Arqueología, Universidad de Granada, Campus de Cartuja s/n (Granada) 18071
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  488  at t...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS glacialis; Proboscidea: Elephas antiquus, probably the earliest record of that species in Europe; Perissodactyla: Equus altidens, Equus sussenbornensis, Stephanorhinus hundsheimensis; Artiodactyla: Bison cf. schoetensacki, Capra sp., Cervus acoronatus, Dolichodorycerus savini, Sus cf. scrofa, Hippopotamus maior. Figure 18. Knapped flakes from Huéscar-1. Left: flake found in situ on the bed, near Section cut 1. Right: flake found on the surface on the rockfall face. Adapted from Martínez Navarro et al., 2006. There is an extraordinary number of waterfowl remains which, together with the nature of the deposition, suggest that this was a shallow delta lake. The most common taxon amongst the large mammals is the genus Equus. Taphonomic studies have revealed that the accumulated bones, mainly oriented in a N-S direction, show no evidence of alteration by carnivores (neither predators nor scavengers) and few traces of weathering. On the other hand, they do show many marks caused by traction. We can therefore infer that at least part of the large mammal association –perhaps with the exception of Hippopotamus– was transported by flowing water that built up in a shallow delta lake environment. The chronology proposed on the basis of palaeomagnetism (below the Brunhes-Matuyama polarity inversion) and the composition of the faunal assemblage is ~900 ky. The lithic industry consists of three flakes, attributable to the Oldowan or Mode 1, with clear percussion bulbs and no evidence of retouch on their active edges. Two of them “were found on the north side of the stream, about 3 metres from where they had become detached from the rockface, the other in situ, partially adhered to the sediment between sections A and B excavated by Mª. Teresa Alberdi” (Martínez Navarro et al., 2006:55). 489
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  glacialis  Proboscidea  Elephas antiquus, ...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 490 J.M. Jiménez Arenas* 8. Cúllar Baza-1 The Cúllar Baza-1 site (CB-1, 965 m.a.s.l.) is in the northeastern sector of the Guadix-Baza intramontane basin, 2.5 km SE of Cúllar. The three fieldwork seasons (1973, 1975 and 1986), were directed by A. Ruiz Bustos, M. Botella and M.T. Alberdi respectively. The chronology of CU-1 has been defined as ~720 ky on the basis of palaeomagnetism, the biostratigraphic interpretation of the vertebrate association and the stone industry. The stratigraphy was first defined by Ruiz Bustos (1976) and subsequently revised by Alonso Diago et al., (2001) and Torrente Casado (2010). On the basis of the latter author (with additional reference to the denomination in Alonso Figure 19. CB-1 stratigraphy. Left, as proposed by Alonso Diago et al., (2001). Right, as proposed by Torrente Casado (2010). The stratigraphic column is in the centre. Adapted from Torrente Casado (2010). Diago et al., 2001), ten levels can be distinguished in CU-1. The three sub-levels on Level D –the most archaeo-palaeontologically significant– are, from bottom to top, D (C) –black or brown-green * clays and silts, with major plastic tectonic deformations, D’ (C’) –black, massive clayey silt, and D’’ composed of massive, carbonated dark brown sandy silt (D1’). The published faunal list contains the following taxa: Pisces: Leuciscus pyrenaicus; Reptilia: Acanthodactylus cf. erythrurus, Blanus cinereus, Chalcides indet., Chalcides cf. bedriagai, Lacerta sp., Lacerta cf. lepida, Lacerta (Podarcis) indet., Natrix sp., Rhinechis scalaris, Scindidae indet., Testudo sp. B, Timon cf. lepidus; Insectivora: Sorex sp., Neomys sp., Crocidura sp.; Rodentia: Apodemus aff. sylvaticus, Cricetulus (Allocricetus) bursae, Elyomis quercinus, Microtus (Microtus) brecciensis, Arvicola mosbachensis/A. cantiana; Lagomorpha: Lepus cf. granatensis, Oryctolagus cf. cuniculus; Carnivora: Canis etruscus, Crocuta crocuta, Vulpes sp., Vulpes praeglacialis; Proboscidea: Mammuthus trogontherii; Perissodactyla: Equus altidens, Equus sussenbornesis, Stephanorhinus etruscus; Arctiodactyla: Bison cf. schoetensacki, Capra sp., Cervus acoronatus, Dolichodorycerus savini, Sus cf. scrofa. The most common taxon amongst the large mammals is the genus Equus. The site’s sedimentology permits the inference of several major palaeogeographic aspects, with nine sedimentary environments (four suggested by Alonso et al., 2001) which, depending on the depth of the water, are more reductor environments with a greater amount of organic matter (marsh or lake edges) or more open, oxygenised lake environments with a greater input of alluvial matter. The most fertile archaeo-palaeontologicals levels are probably from swamp or lakeside contexts. There is no evidence of taphonomic abrasion, fractures or incisions, or heavy weathering. In contrast, carnivore bite marks, by both hunters and scavengers, are present. The presence of coprolites also indicates in situ activity by hyaenids. Small numbers of articulated elements have been found on all levels. The random Dpto. de Prehistoria y Arqueología, Universidad de Granada, Campus de Cartuja s/n (Granada) 18071
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  490  J.M....
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Figure 20. CU-1 lithic industry. Flakes discovered in 1986 excavation. Adapted from Vega Toscano (1989) and Bonadonna Alberdi (1989). J.M. Jiménez Arenas* 9. Solana del Zamborino site The Solana del Zamborino (SZ) archaeo-palaeontological site, 992 m a.s.l., is in the western sector of the Guadix-Baza intramontane basin, 7 km from Fonelas. This is a terminal level of the Guadix formation. The three fieldwork seasons in 1972, 1973 and 1976 were led by M. Botella. Six levels have been described, three of which contain archaeological and palaeontological remains: from bottom to top, Level A, composed of clays (A1 -grey plastic clay, A21 -green clay), silt (A22 -grey silt) and sand (A23 -very fine sand); Level B, the richest level in fauna and tools, where three types of material have been distinguished, equivalent to lateral changes in facies, very finegrained black clay (B1), sandy loam (B2) and greenish grey clay (B3); and Level C, with two chronological periods (C1 and C2), and within them, lateral changes of facies (C11 -greenish-grey clay with * distribution of the elements suggests that there were no significant alluvial remobilization episodes. The mortality profile of the species shows an L shape, suggesting catastrophic or seasonalattritional mortality on and near the lake banks or marsh areas in a hot, dry climatic context. The meagre lithic industry catalogue consists of a quartzite chopping tool, a metamorphosed dolomite chopper and five manuports in the same rock (1973 excavation), at least two pieces of flint and several worked pebbles (1975 excavation, now misplaced) and nine manuports and two flint flakes (1986 excavation). This assemblage has been assigned to Oldowan or Mode 1 on the basis of both typological (presence of chopper, chopping tool and flakes) and technological criteria (use of hard hammer, lack of retouch). very fine polyhedral structure, C12 -very fine sands, difficult to follow, C21 -prismatic black clay with violet patches, C22 -green clays with a polyhedral structure, C23 -sand and silt, C24- similar clay to C22, with carbonated nodules). The vertebrate faunal association consists of: Reptilia: Lacerta sp., Testudo sp., Anura indet.; Insectivora: Sorex sp., Crocidura sp.; Rodentia: Eliomys quercinus cf. quercinus, Eliomys quercinus cf. lusitanicus, Allocricetus bursae colombierensis, Arvicola sapidus (A. cantiana, según A. Ruiz Bustos), Microtus brecciensis, Apodemus cf. flavicollis; Lagomorpha: Oryctolagus cf. cuniculus, Lepus sp.; Carnivora: Canis cf. lupus, Panthera (Leo) spelaea, Lynx cf. pardina, Felis sylvestris; Proboscidea: Mammuthus trogontherii (present only in level A), Elephas antiquus (levels B and C); Perissodactyla: Equus caballus torralbae (present in levels A, B and C), Dpto. de Prehistoria y Arqueología, Universidad de Granada, Campus de Cartuja s/n (Granada) 18071 491
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Figure 20. CU-1 lithic industry. Flakes di...
492 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 21. SZ site stratigraphy. Adapted from Casas et al., (1976). Stephanorhinus hemitoechus; Arctiodactyla: Cervus elaphus, Capreolus capreolus, Dama sp., Bos (Bos) primigenius, Bos (Bison) priscus, Hippopotamus sp., Sus scrofa; Primates: Macaca sp. The archaeo-palaeontologicals levels have been linked to lake and marsh environments in which the different lithologies correspond to seasonal variations in the water level under a warm, moist climate regime. The predominant taxa are Equus and Bos. The lithic industry consists of more than 1,500 items. Amongst the raw materials, there is an abundance of quartzite and quartz tools but less flint items. Most of the items are retouched, some with soft hammerstones. The most abundant types are scrapers, notches and denticulates, and also the three characteristic Acheulean or Mode 2 formats, i.e., handaxes, cleavers and trihedrals, as well as cores and choppers. All of this led Botella et al., (1976) to frame it in the Mediterranean Late Acheulean. However, Level A has a major feature: a lack of denticulates, a drastic reduction in sidescrapers and a lack of Acheulean types. In conjunction with the replacement of E. antiquus by M. trogontherii, this suggests a hiatus between Level A and the others with a palaeontological and archaeological record. One outstanding aspect of SZ is the large amount of charcoal and the presence of a structure inter- Figure 22. Handaxes found at Solana del Zamborino. Above: chordate handaxe found on the surface. Below: handaxe from Level B. Spots indicate where it was retouched with a soft hammer. Adapted from Botella et al., (1976) and Jimenez-Arenas et al., (2011). preted as a hearth. The circular distribution of the cobbles and the heat disturbance to their inner faces are the main arguments for this interpretation. The chronological interpretation of this site has been reignited in recent years since the proposal of SZ as the source of one of Europe’s oldest evidence of handaxes (720 ky, Scott and Gibert 2009). However, this dating has been seriously questioned since several errors have been detected in the interpretation of the archaeological assemblage (generically framed in an early Middle Palaeolithic context) and significant discrepancies between the faunal list and the original list referred to by the authors of the aforementioned article, with taxa omitted or altered in an attempt to make the faunal assemblage consistent with ages deduced from magnetostratigraphy (Jiménez-Arenas et al., 2011). From an archaeological –both typological and technological– and palaeontological perspective, the Solana del Zamborino site should therefore be attributed to a well advanced point in the Middle Pleistocene (400-300ka).
492  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Juan Manuel Jiménez Arenas*, Ignacio Martín-Lago*,** 10. Cueva Horá (Darro, Granada) Horá Cave is 2.5 km from Darro, in the final eastern foothills of Sierra Harana, 1,217 m a.s.l. Discovered by H. Obermaier in 1916, several excavations have been undertaken since the middle of the 20th century under J. Spahni (1956), M. Pellicer (1957) and M. Botella (1977-1985). The cave is in a limestone block separated from the main range by a major N-NE/SSE fault. Although this site is called a cave, in fact it is morphologically a shelter with two entrances, one facing SE and the other NW. On the steeply sloping floor there are many limestone blocks which have fallen from the roof and the retreating entrance canopy. Excavations headed by M. Botella at the entrance documented at least 61 levels (I-LXI) in more than 14 m of sedimentary infill, all containing traces of human presence according to the authors. Only the sequence containing Levels I to XIX, the first three metres of infill, have been published to date. These levels were grouped into six set of strata which showed that the corresponding palaeoclimate occurred between two warm, humid periods. The first three strata were deposited in a cold and somewhat wet context, while the fourth (levels VIII-IX) coincided with the warmest and wettest period. The fifth set (levels X-XII) was the coldest and driest stage, while the sixth (levels XIIIXVII) was similar to the wettest climatic phase. The cave infill has been inserted chronologically in the Würm. However, the stratigraphic sequence poses several problems for its chronostratigraphic analysis. These issues, as well as criticism of the designation of micromammal species, has had played a decisive role in the attribution of this cave to a specific period in the Upper Pleistocene. * ** The faunal list consists of the following taxa: Rodentia: Eliomys quercinus, Allocricetus bursae, Microtus arvalis agrestis, Microtus cf. dentatus, Microtus (Pitymys) duodecimcostatus, Arvicola sp., Pliomys lenki, Pliomys episcopalis, Apodemus sylvaticus; Carnivora: Lynx pardina, Canis lupus; Perissodactyla: Equus caballus cf. germanicus, Equus hydruntinus, Stephanorhinus hemitoechus; Artiodactyla: Cervus elaphus, Bos o Bison sp., Capra pyrenaica. The principal taxon, Equus, clearly predominates over the others. The lithic industry has not been fully analysed to date, as only partial studies of some of the archaeological levels have been conducted. Overall, flint is the most commonly used raw material in this industry. The upper levels have been classified techno-typologically as a typical Figure 23. Horá Cave stratigraphy (cut 7). Adapted from Botella and Martínez (1979). Dpto. de Prehistoria y Arqueología, Universidad de Granada, Campus de Cartuja s/n (Granada) 18701. Delegación Territorial de Educación, Cultura y Deporte en Granada, Paseo de la Bomba, 11 (Granada) 18071 493
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Juan Manuel Jim  nez Arenas , Ignacio Mart...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 494 Mousterian with a weak Levallois knapping and non-Levallois facies, rich in sidescrapers (Levels I-XIX, XXII, XXV and XXVI), while the lower levels (XLVIII, IL and L) were previously associated with Southern Late Acheulean (Botella et al., 1984). However, analysis of Levels LV and LVI, the oldest studied to date, has led us to the conclusion that Horá Cave could be associated with non-Levallois Mousterian facies predominated by sidescrapers and denticulates, with few handaxes, an indication of its conservative nature in comparison with previous traditions. This industry is comparable to those found at other sites in southern Iberia such as Levels D and E in Carigüela Cave, and Levels I-XIII in Angel Cave. Figure 24. Lithic industry from the base level of Horá Cave. Adapted from Botella and Martínez (1979). Isidro Toro Moyano*, Juan Manuel Jiménez Arenas** The Las Grajas de Archidona cave (Malaga, Spain) 1. Introduction Situated in the south of the Iberian Peninsula right in the geographical centre of Andalusiain the north of the province of Malaga, the cave lies less than one kilometre to the west of the town of Archidona. An extremely powerful regional landscape feature, this is a huge rock shelter (Fig. 1) located at an altitude of 775 metres above sea level. The mouth of the shelter is 66 metres wide, 40 metres deep and 30 high from canopy to floor, which is uneven, quite steeply sloped and covered with large clasts that have broken away from the canopy. In the central section, which is higher, there is a small cave, and the wall of the shelter features numerous hollows and protuberances. The shelter was used in the more recent past as a fold for livestock. Between 1972 and 1976 it was the scene of a series of systematic archaeological digs * ** Figure 1. The Las Grajas de Archidona cave. carried out under the direction of Professor Benito del Rey of the Prehistory Department of Salamanca University (Benito del Rey, 1976). Museo Arqueológico de Granada. Carrera del Darro 41-43. 18010 Granada. isidro.toro@juntadeandalucia.es Departamento de Prehistoria y Arqueología Universidad de Granada. Campus Cartuja. 18003 Granada. jumajia@ugr.es
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  494  Mous...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS The shelter was found to contain diagrammatic cave drawings in and on some of the hollows and protuberances and a necropolis consisting of artificial caves dug out around the threshold of its mouth. In 2009 the General Directorate of Cultural Heritage of the Regional Government of Andalusia helped to start a programme to study the materials recovered during the aforementioned archaeological digs and to publish the findings. This year has seen a continuation of the programme with the general research project entitled “Las Grajas Archidona. Human Settlements, Cultures, Faunas, Floras, Climates and Landscapes during the Middle and Late Pleistocene Central Andalusia”. 2. The Cave. Geological Context.The Archaeological Site The Archidona Mountains are located in the central-western part of the Béticas Ranges, within the area known as the Mid-Subbética. These mountains form part of the series of early to middle Lias limestones with a dolomite base upon which successive layers of white bioclastic and oolitic limestone have been laid down to reach a thickness of approximately 200 metres. The Lias limestones display the affect of speleogenesis due to chemical dissolution or atmospheric phenomena that have resulted in the existence of numerous rock shelters, caves, chasms and other types of cavity and of a karst landscape that is partially plugged by detritic and calcitic deposits. The entire system, shelter, cave, hollows and protuberances forman immense archaeological site that has revealed an extensive record encompassing both ancient and recent prehistory. The archaeological digs were carried out over sixteen square metres in the eastern area of the shelter next to its wall. This area revealed a stratigraphic column composed of nine levels which, although all of them contained archaeo-paleontological remains, level 6 stands out for its greater wealth of finds (Fig. 2). Figure 2. Cross-section (from Sevilla, 1988) showing the stratigraphy. Order Carnivora Order Perissodactyla Order Artiodactyla Order Lagomorpha Order Insectivora Order Rodentia Order Chiroptera Order Chelonia Order Anura 3. The Fossil Record 1. The paleontological record Although it has not been studied in its entirety, partial details have been published showing the presence of five macromammaland many micromammal species, especially Chiroptera, with the remains of Cheloniiand several species of herpetofauna also being found (Tab 1). Order Squamata Species Carnivora indet. Equus caballus Bos sp. Capra sp. Cervus sp. Oryctolagus sp. Erinaceus sp. Talpa sp. Sorex sp. Crocidura cf. russula Allocricetus bursae Microtus brecciensis Apodemus sylvaticus Eliomys querquinus Arvicola cf. sapidus Rhinolophus ferrumequinun Rhinolophus euryale Myotis myotis Myotis emarginatus Myotis cf. bechsteini Papistrellus papistrellus Papistrellus savii Papistrellus kuhlii Plecotus austriacus Eptesicus serotinus Miniopterus schreibersi Emys orbicularis Testudo aff. hermanni Discoglossus pictus Bufo cf. bufo Bufo calamita Rana ridibunda Lacerta cf. lepida Lacertidae indet. cf. Blanus cinereus Coronella sp. Cf. Malpolon monspessulanus Cluber/Elaphe Vipera sp. Table 1. List of fauna found in the Las Grajas cave (from Espigares and Ros, 2011). From the taphonomic and archaeo-zoological point of view (Espigares and Ros, 2011), a presence 495
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  The shelter was found to contain diagramma...
496 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD of mostly medium– to small-sized species can be observed. The analyses of the cortical surface have clearly shown the presenceof cut marks related with the processing of the carcasses of the animals for obtaining energy resources. The skeletal elements were found to have been systematically fractured, above all the long bones of the extremities, for the purpose of extracting the bone marrow. Furthermore, the presence of a significant number of elements showing evidence of charring has been documented. There is a paleoenvironmental and paleoclimaticinterpretation (Sevilla, 1988, Sevilla et al., 2012), that indicates a “dry Iberian” or southern climate, a climate that was slightly cooler and wetter than it is today with less noticeable differences between winter and summer. The chronology of the record (more specifically that of level 6) is established as being halfway through the Middle Pleistocene during a cold period, with this observation being based on the analysis of the morphology and the size of a series of rodent fossils found at the site. We believe that this cold period could correspond to isotopic stage 8 (between 303,000 and 245,000 years ago). 2. The Lithic Industry The raw materials used to support this industry are of local and semi-local origin, the majority of which consists of flint of variable quality, limestone and quartzite. Both the flint and limestone are mainly used for reduction, with the quartzite being associated with percussion. Outcrops of flint and limestone are found in the Jurassic formations of the Archidona Mountains, whereas the nearest quartzite is found at a distance of approximately 6 kilometres from the shelter. This is a uniform site from the technical point of view. All the elements of the different phases of the shaping process (hammer stones, cores, flakes and other reduction debris) are present to a greater or lesser extent. This uniformity extends to the effects of the “chaine operatoire” (operational sequence), with all the phases thereof being present. The “chaines operatoires” are intense and, in many cases, result in the total reduction of the core. Reduction is predominantly carried out using a hard hammerstone, although the use of soft hammerstones is also detected for the manufacture of tools and retouching. The composition of the site is dominated by the flakes and debris left behind by the reduction and retouching processes. The condition of the material found is good and features fresh sharp edges, with cases of double patination due to exposure to the air or rolling being practically non-existent. Large pebbles, nodules and tablets are all used as supports. The cores show signs of recurrent and orthogonal unipolar reduction, with final phases of discoidal reduction (unifacial or bifacial) and some pyramidal cores with centripetal projections. With respect to the flakes, several categories may be observed: Cortical flakes (Type I), namely flakes whose whole upper face is completely covered with the cortex of the stone (primary flakes), extremely cortical flakes (Type II), partially cortical flakes (Type III) and non cortical flakes (Type IV). Tools make up approximately 10% of the finds. These are, fundamentally, flat scrapers, with smaller quantities of denticulate tools and notches; some scratchers and burins, and a few reverse knives. Another of the objects found at this siteis an excellent hand axe manufactured in good quality flint, the cortex of which has been retained on its base and lower face, and which is finished in an extremely careful mannerusing a soft hammerstone around its entire edge (Benito del Rey, 1982). Likewise, there are a number of elements that would appear to have been manufactured using pressure reduction with prior heat treatment. The retouch is semi-abrupt or abrupt, and there are also Quina and semi-Quina type retouches, with there being fewer examples of the flat type. With respect to direction, direct retouch dominates. 4. Conclusion In conclusion, we feel we can safely say that what we have here is an industry based on flakes of a Mousterian appearance typical of the archaic Middle Palaeolithic. These lithic finds dating from halfway through the Middle Pleistocene are flake-based industries, with a low Levallois index, an abundance of scrapers, the presence of denticulate tools and the absence or presence of hand axes, albeit always in very low amounts, and are well documented in the Iberian Peninsula. Las Grajas de Archidona presents one of the oldest examples of the human occupation of caves in the south of the Iberian Peninsula, that is, more than 300,000 years ago. Its fossil record, which would cover the whole of the Middle and Late Pleistocene and Holocene periods, provides information of exceptional value for understanding the behaviour of the hunter-gatherer societies that populated Andalusia through out this long period of time, and the way they relate with and use the resources around them.
496  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  of m...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Miguel Cortés Sánchez*, María Dolores Simón Vallejo**, Francisco José Jiménez Espejo***, José Antonio Riquelme Cantal**** El Pirulejo. A Late Glacial campsite in the Subbetic Sierra (Cordoba, Spain) Summary. The richness and diversity of the record at the El Pirulejo site makes it one of the most important Late Glacial archaeological sequences in southern Iberia. An interdisciplinary project has contextualized Solutrean and -more particularly- Magdalenian levels on the basis of a technological analysis of the material culture and a study of symbolic expressions as well as economic and environmental aspects. Keywords. Late Glacial. Solutrean. Magdalenian. El Pirulejo (Priego, Córdoba) was discovered in 1988 and excavated by M.D. Asquerino in an 8m2 area (fig. 1). After his death, his research team summarized the main discoveries (Asquerino 1991, Cortés et al., 2008.). Palaeoenvironmental data El Pirulejo is in the Subbética Range of Cordoba, near the Salado River, a tributary of the Guadalquivir, on an ecotone where several biotopes come into contact (700-1200m asl). The abundant biotic and abiotic resources probably played a role in its choice as an anthropogenic site during the Late Glacial. The site is located in a cave at the foot of a travertine cascade which was filled with sediment during the Late Glacial-Early Holocene. The pollen record depicts a landscape generally dominated by oaks and a remarkable presence of thermophilic taxa, some still persistent (myrtle, wild olive or mastic) and others relict (chestnut and walnut). Herbaceous steppe formations and heathland appeared at the end of the early Bölling/early Dryas, in a relatively dry and somewhat cold climate. Montane pine forests formed at high altitudes on the surrounding mountains (Pandera and Magina) with Lusita* nean oak woods in shady areas and moist valleys. During the Bølling-Allerød Late Glacial Interestadial, a significant rise in temperatures and humidity encouraged the growth of riparian forests (alder, birch, willows and elms) and Quercus sp. (Holm and Lusitanean oak), especially at the end of the period, accompanied by today’s xerothermophyllous maquis. In the early Holocene – around 10ka calBP– there was a significant increase in the cover of Holm and Lusitanean oak woods. Pollen data define a refuge for mesophilic species (chestnut and walnut) during the Late Glacial and a similar type of vegetation -thermophilic faciation of Betic and basophilous Holm oak woods- maintained down to the present day without major changes. Chronocultural sequence Indeterminate Upper Palaeolithic, (P/6). Possibly a Solutrean level given its position between level P/5 and the underlying travertine rock, attributed to c. 19 ky BP, although it has scarcely been explored. Evolved Solutrean s.l., (P/5). Scarcely explored, although it is known to contain highly indicative elements from this period (foliaceous lithic retouched points). Mediterranean Middle Magdalenian (P/4). Lithic industries defined by high percentages of microflake tools with little diversification and a predominance of inverse retouch, a lack of geometrics, a clear predominance of burins over endscrapers and other groups. The available dating (P/4D, ca. 14.4ka BP) and technical features match the MMM/MSM-A attributes defined by several authors (see Cortés et al., 1998 and references therein) for the Magdalenian sequence Departamento de Prehistoria y Arqueología. Facultad de Geografía e Historia, Universidad de Sevilla, c/. María de Padilla s/n. 41004. Sevilla. mcortes@us.es ** Museo Arqueológico de Frigiliana. c/. Cuesta del Apero, 10. 29788-Frigiliana (Málaga). simonmd63@gmail.com *** Departament of Biogeosciences. Japan Agency for Marine-Earth Science and Technology (JAMSTEC).Natsushimacho, 2-15, 2370061 – Yokosuka (Japan). fjjspejo@jamstec.go.jp **** Departamento de Geografía y Ciencias del Territorio. Universidad de Córdoba. Plaza Cardenal Salazar, s/nº. 14071-Córdoba. jriquelme@uco.es 497
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Miguel Cort  s S  nchez , Mar  a Dolores S...
498 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. El Pirulejo. Stratigraphy. in Mediterranean Iberia. The artefacts on organic matter include a semi-cylindrical rod decorated with incised motifs, similar to items found in more northern areas (fig. 2). To date, level P/4 is the only one with a 14C/AMS dating: 14,250 + 90 BP (Poz21164, charcoal) or 17,458 ± 257 cal BP. Mediterranean Upper Magdalenian (P/3). Scalene bladelets and a geometric shape (isosceles triangle), a novelty which has also been found in a much smaller assemblage than the one available for P/4. The number of bladelets with inverse retouch is much smaller. Mediterranean Final Upper Magdalenian (P/2). Sharp decline in archaeological material and disturbance to the site due to post-palaeolithic structures. Bronze Age. Funerary structures. Contemporary-Modern Era. Heterogeneous deposits and materials. There is also an assemblage of highly symbolic objects, ochre, personal ornaments and Palaeolithic mobile art. The latter includes a remarkable series on hard animal matter (P/5-P/3) including an outstand- ing semi-cylindrical rod (P/4) with an incised motif and clear parallels with Parpalló cave and the LowerMiddle Magdalenian in the Cantabria-Pyrenees area. The other motifs are simpler- individual oblique lines or a cross designed using deep incisions. The portable art from El Pirulejo includes 20 platelets with ochre, 24 engravings (P/5-P/3), 15 of them in P/4. The iconography includes zoomorphs of a goat (1) deer (2) and indeterminate graphics (7). There are few signs (triangles, cross, zigzag and spindle). The material used as ornaments includes 2 fluvial and 20 marine molluscs which, given the location of the site (> 80 km from the coast), points to a mobile/exchange network amongst groups in southern Iberia during the Late Glacial. Economic aspects The analysed faunal assemblage is predominated by rabbit, followed by ungulates and a few carnivores (table 1).
498  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS Figure 2. Worked lithic industry (1-6), platelet with portable art depicting a goat from P/4 (7), decorated semi-cylindrical rod fragment from P/4 (8). The taphonomic data (alteration by heat, fracturing or marks) point to a basically anthropogenic source of the rabbit remains. Death patterns show that with only one exception, all the deer were adults. However, all age groups of ibex have been found, with a predominance of juvenilesadults, along with infant and juvenile wild boars. No seasonal patterns have been detected in their capture. Many ungulate fragments bear signs of intense thermal alteration and defleshing. Carnivore species, mainly recognizable from jaw and teeth remains, are scarce throughout the sequence. The lynx and bobcat remains are broken and burned, indicating that they were hunted and consumed. The inorganic resources show that El Pirulejo was a strategic residential campsite in the Sub-Betic area. Some of these resources were secondary acquisitions in the local area, including cobbles and platelets washed downstream on the Salado or Ophites River 500m away, for use as hammerstones, anvils, grinders and portable art, while flint had to be obtained from further away. This strategy is evidenced by the near-absence of patina on the material, as the reduction and pre-shaping of the cores was done elsewhere. Other types of rock (galene or rock crystal) and seashells (from > 80km away) also support the idea of socio-economic networks and mobility circuits between the coast and the inland amongst the hunter-gatherers of Andalusia during the Late Glacial-Early Holocene. Palaeoanthropology In addition to anthropological remains from the Bronze Age, there is also some material from level P/4 (teeth and skulls). Mitochondrial genetic analysis of two identified individuals has found similar P/5 Taxon Cervus elaphus (deer) Capra pirenaica (ibex) Rupicapra rupicapra (chamois) Sus scrofa (wild boar) Oryctolagus cuniculus (rabbit) Lepus granatensis (hare) Mustela nivalis (wessel) Lynx pardina (lynx) Felis silvestris (bobcat) Vulpes vulpes (fox) Carnívora sp. NRD 3.11 5.78 0.44 90.67 - P/4 MNI 8.00 12.00 4.00 76.00 - NRD 0.89 2.80 0.05 0.11 96.07 0.01 0.03 0.01 0.03 P/3 MNI 0.94 2.13 0.24 0.47 96.26 0.24 0.24 0.24 0.24 Table 1. Percentage of mammals (%) in each of the defined archaeological levels. NRD 0.58 2.10 0.21 96.99 0.04 0.24 024 0.02 P/2 MNI 1.06 2.85 1.06 93.59 0.36 0.36 0.36 0.36 NRD 2.91 1.43 1.49 93.79 0.19 0.19 - MNI 4.95 2.97 2.97 86.14 1.98 0.99 - 499
SOUTHERN MEDITERRANEAN COAST, GUADALQUIVIR RIVER AND BETIC INTRAMONTANE BASINS  Figure 2. Worked lithic industry  1-6 , pl...
500 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD results to other Palaeolithic individuals. Sample 2PI coincides with the consensus sequence, the most common in European populations (20%), while 1PI shows a combination of mutations 16182C-16183C16189C, only present in one individual from Macedonia. These are the only data currently available for Late Glacial human populations in Andalusia. Synthesis El Pirulejo is an exceptional site that contains relevant information about the end of the Palaeolithic in southern Iberia. It is also the oldest Magdalenian record available for this region, and is helping to complete knowledge about the context prior to the widespread use of harpoons. It is also the first site to permit a detailed characterization of the occupation of the interior of this region during the Late Glacial, as to date, the best known sites have been in coastal areas with a heavy economic dependence on marine ecotones (e.g. the Bays of Málaga and Algeciras). However, El Pirulejo is just one example of a settlement process that must have been much more intense judging by evidence from other less well known sites such as Mármoles, Murciélagos, Nacimiento and El Duende, and parietal art sites in the Malaga hinterland (Ardales and Pileta). All this information shows that Late Glacial hunter-gatherer populations in southern Iberia were not restricted to coastal areas, evidenced by famous sites such as Nerja, Hoyo de la Mina and Gorham. El Pirulejo and other sites in the Guadalquivir River basin show that southern Iberia’s main river system played a key role in these populations’ mobility and subsistence in the Betic hinterland. Indeed, there has been a remarkable increase in knowledge about the settlement of inland Iberia during the Late Glacial since research projects began to focus on sites from this era. El Pirulejo definitely shows that Magdalenian technocultural innovations were fully assimilated in inland Andalusia as well, following a similar pattern to other Iberian sites. El Pirulejo is therefore probably an example of the territorial articulation in the Late Glacial linked to the Guadalquivir River system and the exploitation of resources in the Betic area.
500  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  resu...
7 strait of gibraltar NORTH-WESTERN ATLANTIC BASINS. 501
7  strait of gibraltar  NORTH-WESTERN ATLANTIC BASINS.  501
Site Map numbering Abrigo de Benzú 84 Giraltar: Gorham and Vanguard cave 85
Site  Map numbering  Abrigo de Benz    84  Giraltar  Gorham and Vanguard cave  85
STRAIT OF GIBRALTAR José Ramos*, Darío Bernal*, Eduardo Vijande*, Juan Jesús Cantillo*, Antonio Barrena*, Salvador DomínguezBella*, Joaquín Rodríguez Vidal**, Simón Chamorro***, Juan José Durán****, Manuel Abad**; David Calado*****, Blanca Ruiz Zapata******, María José Gil García******, Ignacio Clemente*******, Paloma Uzquiano********, Mila Soriguer*, Antonio Monclova*, Jesús Toledo*, Sergio Almisas* The Benzu rockshelter (Ceuta). Stratigraphic sequence and record of Hunter Gatherer societies of marine resources with Mode 3 technology in North Africa Location, stratigraphy, chronologies The Benzu rockshelter (Ramos et al., 2008; Ramos et al., eds., 2013) is located in the NorthAfrican area of the Strait of Gibraltar, Ceuta. It is situated 230 m from the current coastline, 63 m.s.n.m., close to Algarrobo stream and Ballenera Bay. It is located in Dolomitic marble Alpujarrides of the Triassic Age, with an abrupt typography and almost vertical walls. It has lost a part of its superior cover due to collapsing. Its dimensions are about 15.52 x 6.2 m, with a visor cover. In the southwest end there is a little cave with Neolithic settlements. The archeological deposit has got a surface of about 61.1m2 with a power higher than 5.50 m of cemented carbonates sediment, with calcareous Geological divide and interspersed levels of sediment of calcites. Laterally this thickness reduces up to 1m. Ten strata have been identified, and seven of them have human occupation evidences of the Middle and Upper Pleistocene (Fig. 1). During the cold stages of the Quaternary, sea-level fell more than 120m, generating a huge * platform that is immersed today. The immediate territory of the rockshelter offered the possibility to access large resources: marine, hunting, plants, water sources, lithic. Several studies about micro morphology and bioerosive processes in the environment of the Benzu rockshelter show that its erosive process would be previous to human occupation, before e.i.9. The investigation has been directed by J.Ramos and D.Bernal (University of Cádiz), and researchers belonging to different institutions have participated, also Co-operation agreements between University of Cádiz and the City of Ceuta have been developed. The studies have been done with the authorization of the Ministerio de Cultura. The results of dates have been: level 7 (Th/U, IGM: ±70 ky), level 5 (OSL, Shfd 020136: 168 ± 11 ky), level 3b (Th/U, IGM: 173 ± 10 ky), and level 2 (OSL, Shfd 020135: 254 ±17 ky). In this way, the sediment and archeological sequence is previous to 70Ka and the register of the first human occupation of the rockshelter is before 250 ky. Due to the peculiar characteristics of the site, a specific technology has been used; this has Universidad de Cádiz, jose.ramos@uca.es, dario.bernal@uca.es, eduardo.vijande@uca.es, jesus.cantillo@uca.es, antonio.barrena@uca.es, salvador.dominguez@uca.es, mila.soriguer@uca.es, anmonc@terra.es, jesustoledo86@gmail.com, seralmcru@alum.us.es ** Universidad de Huelva, jrvidal@uhu.es y Universidad de Atacama, manuel.abad@uda.cl *** Instituto de Estudios Ceutíes. Ceuta, schamorrom@wanadoo.es **** Instituto Geológico y Minero de España. Madrid, jj.duran@igme.es ***** Instituto Português do Património Arquitectónico. Faro. Portugal, dcalado@ippar.pt ****** Universidad de Alcalá de Henares, blanca.ruiz@uah.es, mjose.gil@uah.es ******* Institució Milá i Fontanals, CSIC. Barcelona, ignacio@bicat.csic.es ******** Universidad Nacional de Educación a Distancia (UNED), p_uzquiano@hotmail.com 503
STRAIT OF GIBRALTAR  Jos   Ramos , Dar  o Bernal , Eduardo Vijande , Juan Jes  s Cantillo , Antonio Barrena , Salvador Dom...
504 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. The Benzu rockshelter. Stratigraphic profile permitted the removal of blocks that have been finished to dig in laboratory with a micro spatial control of the products (Domínguez-Bella et al., 2012). Pollen, anthracology, terrestrial and marine fauna Pollen analysis has shown that the territorial vegetation was mainly constituted by Cedrus and, in a minor extent, by Pinus. Quercus-p, Olea, Ceratonia and waterside elements, such as Alnus, Salix and Ulmus, have been developed. The shrubby herbaceous has mainly been constituted by steppe elements (Artemisia, Asteraceae and Chenopodiaceae). The shrub layer, with Ericaceae and Juniperus, did not have an important role. This composition shows Mediterranean characters conditions, primarily dry, with pathways of water which are more or less permanent and ponds which favor the development of waterside and aquatic taxa. Throughout the sequence, oscillations and changes in tendency to the decrease of the humidity rate occurred. A cyclicity happened with the installation of a forest including warm and Mediterranean elements together with a varied shrubby herbaceous vegetation and a high representation of waterside taxa and aquatic elements. The Anthracological data have documented plant taxa: Erica sp., in strata 4, and Fabaceae in strata 2. These two taxa have good flammable properties and could be used as fuels. The terrestrial fauna consists of 3,362 bony fragments of medium-sized mammals and splinters, as well as pieces of humerus diaphysis of medium-sized ungulates. They were deposited as a consequence of human activity. Several areas of activity and possible consumption places in 4, 5 and 6 strata have been documented. There are lots of bony fragments presenting burned and deliberate fractures. Bovine ungulates and other mediumsized herbivorous predominate. The marine fauna is documented with 144 fragments, showing an exploitation of coastal resources, underlining the presence of malacofauna –mainly Patélidos molluscs– in all the stratigraphic sequence. There is a clear predominance of the Gastropod Class faced with the Bivalvia, noticing the group of the non-spiraled gastropods and in particular the Patellidae family, followed by Siphonariidae, being Siphonaria pectinata the greatest exponent. In this way, there are specimens of Patella sp., among others. Referring to bivalves, its representativeness is attested to by the presence of some remains of the species Tapes decussatus in level 6, and others belonging to Glycimeridae family. Remains of ictiofauna vertebrae in level 5a are registered –possibly from the Sparidae family.
504  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
STRAIT OF GIBRALTAR Lythic technology From strata 1 to 7, 36,092 samples have been analised. The raw materials basically come from the immediate environment of the site, underlining the compact sandstones –61.71%–, in the face of the red radiolarites rocks –36.37%–. There are other minority lithologies that would set a certain mobility of the human groups that frequented Benzú. There is a predominance of rests of carve –35,322 samples– in the face of refinished products –763–. 523BN1G-Cores, 11,648BP-Flakes and 23,151 ORT-Others rests of carve have been documented. There is a remarked presence of levallois technique and Centripetal-Multipolar Core. Among flakes, the internal ones are the most representative, as well as those from the levallois technique. Tipometric Analysis show that flakes –95.76%– dominate over blades –4.24%–, with a significant presence of small-sized types. Among the retouched products-BN2G, scrapers dominate over notches, denticulates and points –Fig.2–. Five Operational Technique Indirect scheme have been documented: 1-Longitudinal, 2-Unipolar, 3-Centripetal, 4-Bipolar, 5-Multipolar, that show the technique process of the lithic production. The functional study has documented traces of use in almost the 20% of the lithic rests analyzed. The wood working is documented in every archaeological level with more frequency than the activities for meat production and/or leather. However, in level 3 butchering is the most representative activity, and it is the only level where scraping of a hard material from animal origin is documented. There is a significant presence of thermal damage. The use of shrink fitting in an instrument is obvious in level 4. Conclusions The Benzu rockshelter was a place where Hunter gatherer societies of marine resources usually went, doing activities in a seasonal residence with production and consumption processes. We discuss the register and documentation of leveraging practices of marine resources –fish and mollusks–. Species next to the coast have been collected, being an important resource and one of the most ancient evidences of fishing and shell fishing practices by prehistoric society. Figure 2. The Benzu rockshelter. Carved lythic products. a-b: Centripetal-multipolar Core; c: BP– Levallois Flake; d: BN2G.Scraper; e-f: BN2G-Points. (a-b: Strata 1, c-d-e-f: Strata 4 a). The lithic assemblage is clearly mode 3, therefore Middle Paleolithic. The use of fire related to processes of prophylaxis and cleaning has been proved. In conclusion, there are no significant technical differences within the sequence. The technical systems of production and working show different ways of life based on hunting, collecting and exploitation of marine resources. 505
STRAIT OF GIBRALTAR  Lythic technology From strata 1 to 7, 36,092 samples have been analised. The raw materials basically ...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 506 Clive Finlayson*, Ruth Blasco*, Joaquín Rodríguez-Vidal**, Francisco Giles Pacheco***, Geraldine Finlayson*, José María Gutierrez****, Richard Jennings*****, Darren A. Fa*, Jordi Rosell******,*******, José S. Carrión********, Antonio Sánchez Marco*********, Stewart Finlayson*, Marco A. Bernal***** Gibraltar excavations with particular reference to Gorham’s and Vanguard Caves Gibraltar (36°07’13”N 5°20’31”W) is located at the southern end of the Iberian Peninsula, at the eastern end of the Bay of Gibraltar. It is a small peninsula being 5.2 km in length, 1.6 km in maximum natural width and about 6 km2 in total land area. This peninsula forms part of the northern shore of the Strait of Gibraltar, linking the Mediterranean Sea and the Atlantic Ocean (Fig. 1). Currently, the Rock of Gibraltar includes 213 catalogued cavities, at least 26 catalogued as containing archaeological deposits. Among these, Gorham’s Cave is perhaps the most referenced in the research and general literature; however, there are other significant Pleistocene archaeological sites, as Vanguard Cave, Devil’s Tower Rock Shelter, Forbes’ Quarry, Ibex Cave and Beefsteak Cave, among others. Interest in the geology, pre-history and natural history of Gibraltar during the 19th and early 20th centuries Early developments Gibraltar, being a military fortress, acted as a magnet which concentrated individuals who would otherwise not have come to the area. The knowledge accumulated and disseminated by these officers was crucial in highlighting the uniqueness of Gibraltar and the surrounding areas of Spain. Among them, Lieutenant Colonel Willoughby Verner was an intrepid explorer with an insatiable passion for collecting and classifying birds and birds’ eggsVerner was also interested in prehistory. In 1911 he had heard of a cave with paintings in the Ronda area in southern Spain and was responsible for making known the exist- The history of cave research in Gibraltar goes back to the 18th Century.The Reverend John White, brother of the famous Gilbert White of Selborne, who was chaplain at Gibraltar during the 1770s, collected many zoological specimens and kept detailed records, corresponding regularly with his brother and other famous zoologists of the day, in particular Thomas Pennant and Daines Barrington. White wrote a Fauna Calpensis, the first detailed zoological account of Gibraltar, which was sadly never published, with the manuscript now lost (Mullens, 1913). * ** Great interest and excitement about the geology and prehistory of Gibraltar was generated during the 19th Century following the discovery of rich deposits of bone breccia, as well as bones and human artifacts in caves in the limestone of the peninsula. The material recovered was considered to be of such great importance that it attracted the attention of famous names of the day, for example Sir Hugh Falconer and George Busk. As early as 1846 James Smith, who was an officer stationed in the Garrison of Gibraltar and who had become an active member of the Gibraltar Scientific Society, published a paper “On the Geology of Gibraltar” in the Quarterly Journal of the Geological Society of London (Smith, 1846). The Gibraltar Museum, 18-20 Bomb House Lane, PO Box 939, Gibraltar. Depto. Geodinámica y Paleontología, Facultad de Ciencias Experimentales, Campus del Carmen, Universidad de Huelva, 21071 Huelva, Spain *** Gibraltar Caves Project, 18-20 Bomb House Lane, P. O. Box 939, Gibraltar **** Museo Histórico Municipal de Villamartín. Avda. de la Feria s/n, 11650 Villamartín, Cádiz, Spain ***** Research Laboratory for Archaeology, University of Oxford, New Barnett House, 28 Little Clarendon Street, Oxford OX1 2HU, UK ****** Àrea de Prehistòria, Universitat Rovira i Virgili (URV), Avinguda de Catalunya 35, 43002 Tarragona, Spain. ******* IPHES, Institut Català de Paleoecologia Humana i Evolució Social, C/ Marcel·lí Domingo s/n (Edifici W3), Campus Sescelades, 43007 Tarragona, Spain ******** Department of Plant Biology, University of Murcia, Campus de Espinardo, 30100 Murcia, Spain ********* Area of Neogene and Quaternary Faunas, Institut Català de Paleontologia, Campus de la UAB, Cerdanyola del Vallès, Barcelona, Spain
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  506  Cliv...
STRAIT OF GIBRALTAR ence of La Cueva de la Pileta and its Palaeolithic Cave Art. The eminent prehistorians Professor H. Obermaier and L’Abbé Henri Breuil learnt of the existence of the cave and visited it with Verner in 1912. The relationship between Verner and Breuil developed from this contact and is another example of the fortuitous way in which discoveries were often made. Breuil, who was Professor at the Institut de Paléontologie Humaine de Paris, visited Gibraltar in 1914 at Verner’s instigation and, while walking along the north-eastern side of the Rock, commented to Verner that the brecciated talus he had observed should prove fruitful in investigating the existence of prehistoric Man at Gibraltar (Verner, 1919). Breuil returned in 1917 and examined the brecciated talus. At the time he was in the war service of the Naval Bureau and the French Embassy at Madrid and was employed on several occasions as courier between Madrid and Gibraltar (Breuil, 1922). He found animal bones and Mousterian implements but was prevented from exploring further by a military policeman. He returned yet again in 1919, on this occasion with a Governor’s permit to excavate. He found conclusive evidence of use of the site by “Palaeolithic Man” (Breuil, 1922). At Breuil’s instigation, Miss Dorothy Garrod conducted detailed excavations of the site between November 1925 and January 1927 –her results included the discovery of fragments of the skull of a Neanderthal child. Gibraltar’s second Neanderthal had been found, a few hundred metres from the 1848 find in Forbes’ Quarry. The discovery is indirectly attributable to Verner– without his initial discovery of La Pileta and contact with Breuil, the latter might never have visited Gibraltar. Perhaps the individual most responsible for bringing Gibraltar’s caves and deposits to the forefront of scientific research was Captain James Brome, Governor of the Military Prison on Windmill Hill, Gibraltar, between April 1863 and December 1868. His investigations were so detailed and thorough that it prompted scientists such as Falconer and Busk to visit Gibraltar and examine the Rock’s rich deposits. When Brome arrived in Gibraltar, the scientific community had begun to recognize the importance of Gibraltar’s palaeontological deposits, especially the bone breccias. He took up the appointment of Governor of the Military Prison on Windmill Hill, Gibraltar. Windmill Hill is an ancient wave-cut platform at the southern end of the Gibraltar peninsula and it is here that a system of fissure caves (known as Figure 1. Location of Gibraltar at the southern Iberian Peninsula (top) and the present-day Rock of Gibraltar showing the location of Gorham’s (A) and Vanguard Cave (B) on its east face: (A1, A2) Stratigraphic sequence of the outer area of Gorham’s Cave; (B1, B2) new excavations at the Upper part of the stratigraphic sequence of Vanguard Cave. the Genista complex) is to be found. The largest and most important of the system is Genista I which was discovered by Brome. He used convict labour to excavate this deep fissure, which yielded large quantities of bone some of which are thought to be the oldest so far found in Gibraltar. Brome was a thorough researcher and gained the respect of scientists of the day with whom he corresponded and to whom he sent most of what he collected in Gibraltar. The bulk of Busk’s paper (Busk, 1868) on the Gibraltar bone finds is a verbatim account of Brome’s discoveries. The Neanderthal finds The year 1848 saw many momentous events in European politics, but it was also an archaeological 507
STRAIT OF GIBRALTAR  ence of La Cueva de la Pileta and its Palaeolithic Cave Art. The eminent prehistorians Professor H. O...
508 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD watershed. Another momentous event, although unrecognised at the time, was the recognition and curation of a fossilised human cranium found during work at Forbes’ Quarry, Gibraltar (Busk, 1865; Broca, 1869; Sollas, 1907). Although, technically speaking, the child’s skull from Engis in Belgium was the first known discovery of a Neanderthal fossil, some 18 years earlier, its features were less obviously distinct from those of a modern human, and it was over a hundred years before its importance was recognised. In the case of the Forbes’ Quarry discovery, the unusual morphology of the face and vault alone could have been enough to alert an educated observer to its possible significance, but instead fate decreed that today we discuss “Neanderthal Man” (Homo neanderthalensis) rather than “Calpican Man” (“Homo calpicus”) (King, 1864; Keith, 1911). So it was that on the 3rd of March, 1848, a Captain Edmund Flint, secretary of the Gibraltar Scientific Society (at this time renamed the Gibraltar Museum Society) presented a human skull to this body of essentially military officers. The minutes of the meeting simply read: “Presented a human skull from Forbes’ Quarry, North Front, by the Secretary...”. Flint had been in charge of the society’s museum since the 5th June, 1844, and his efforts were recognized in the minutes of 3rd October, 1849. Nobody took much notice of the skull which was promptly put away in the Society’s museum. The skull was in fact that of a Neanderthal but this was not realized until eight years later when another was found in the Neander Valley in Germany. Brome sent the skull to England with his extensive material from the Genista Caves. This material was being examined by Falconer and Busk. In 1864, Busk visited Gibraltar and went to Forbes’ Quarry with Lieutenant Alexander Brown. There they found the matrix in which the skull had been embedded. It was then that Busk pronounced the skull: “to be of a human being of the lowest known organization somewhat analogous to the Neanderthal” (Busk, 1868), a view supported by Falconer: “This human skull yielded by the Rock, appears to us to point to a still higher antiquity of man than even those found in the valley of the Vezere in the south of France. In fact, it is the most remarkable and perfect example of the kind now extant” (Murchison, 1868). The skull was exhibited at the meeting of the British Association in Bath in 1859. Busk subsequently presented the skull to the Royal College of Surgeons in 1868 – it caused a sensation and became known as the Gibraltar Skull. The skull was examined by Professor Sollas and Dr. Sera of Naples and Professor Keith of the Hunterian Museum and it was stated to be of a woman who may have lived 200,000 years ago (Duckworth, 1911). Recently, as part of his research on George Busk (Gardiner, 1999), Professor Brian Gardiner located a review paper by Cook (1997), which referred to two neglected publications of Busk from 1864. These provide further information on the Forbes’ Quarry discovery, and show that Busk was remarkably prescient in identifying some key morphological features of the fossil – in fact he was the first to note the midfacial projection and inflated cheekbones which are now considered one of the most distinctive of Neanderthal characters. As these sources were apparently unknown to Sir Arthur Keith when he described what was known of the early history of the specimen (Keith, 1911), we quote some of the pertinent material from these papers. On the 16th July 1864, Busk wrote a short communication in The Reader (Busk, 1864) entitled “Pithecoid Priscan Man from Gibraltar”. Near the end he stated of the Forbes’ Quarry cranium: “Its discovery also adds immensely to the scientific value of the Neanderthal specimen, if only as showing that the latter does not represent, as many have hitherto supposed, a mere individual peculiarity, but that it may have been characteristic of a race extending from the Rhine to the Pillars of Hercules: for, whatever may have been the case on the banks of the Dussel, even Professor Mayer [a contemporary sceptic regarding the Neanderthal find] will hardly suppose that a rickety Cossack engaged in the campaign of 1814 had crept into a sealed fissure in the Rock of Gibraltar.” In the Bath Chronicle Busk (1864) described the fossil in more detail, making morphological comparisons with “Negro”, Australian and Tasmanian crania. He stated: “The cranium in question, we understand, was originally deposited in a museum of natural curiosities, which at one time existed at Gibraltar, but which it is to be much regretted has of late years been allowed to fall into a state of confusion and neglect […] Its extraordinary peculiarities fortunately struck the notice of Dr Hodgkin in a visit paid by that ethnologist to Gibraltar in the course of last year, in company with Sir Moses Montefiore, and it was at his instance that Captain Browne [Brome?], with his eminent zeal in the cause of science, was induced to procure its being forwarded to us for examination and description. […] it was dug up in the course of some excavations being made in what is
508  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  wate...
STRAIT OF GIBRALTAR termed “Forbes Barrier”, which is situated near the entrance into the fortress from the neutral ground or mainland. […] the Gibraltar skull exhibits not only several of the striking peculiarities of the neanderthal [sic] calvarium but also many others, which from the imperfect condition of that famous specimen, are altogether wanting in it […] In general outline the Gibraltar cranium viewed in profile, bears a strong resemblance to that from the Neanderthal, except that the sapraorbital [sic] projection is not quite so great. The forehead is equally receding, and the great depression in the hinder part of the cranium is equally remarkable in both […] One consequence of the great breadth and convexity of the nasal process of the maxillary bone, combined with the increased width of the nasal opening, is, as it were, to throw forward the entire nasal framework, whilst at the same time the canine fossa […] is entirely filled up, the central portion of the bone rising in a uniform curve on either side, so that the central part of the countenance projects in a very remarkable manner ”. It was the virtual absence of information on the skull and the circumstances of its discovery that led to Dr. W.L.H. Duckworth’s (Cambridge University) visits between 1910 and 1912. His stated objective was: “to learn from personal observation and inquiry , so much as might be possible about the circumstances of the discovery of the now classical ‘Gibraltar Skull’ “ (Duckworth, 1911). Duckworth found very little – the site had been extensively quarried and the cave’s depth reduced (Duckworth, 1911). To make matters worse, a rock fall during Duckworth’s visit sealed off the cave completely. The skull is today in the Natural History Museum in London, transferred from the Royal College of Surgeons. A cast is exhibited in the Gibraltar Museum. Forbes’ Quarry is the subject of a research and conservation project by the Gibraltar Museum. As we have already discussed, a second Neanderthal skull was found much later, in 1926. Dorothy Garrod found the fragmented skull of a child in Devil’s Tower Rock Shelter and took it back to England along with all the material collected from this Mousterian Rock Shelter. This skull is also in the Natural History Museum in London. Today, Forbes’ Quarry is nearly stripped of Pleistocene sediments, but there are lingering pockets of a cemented, shelly sand which, to judge from the remaining matrix on the fossil, may relate to the provenance of the cranium. However, it will only be by direct age estimates using tech- niques such as Electron Spin Resonance on tooth enamel, or Gamma Ray dating on the whole cranium, or OSL on the matrix, that we will eventually determine whether the Forbes’ Quarry Neanderthal dates from the earlier or later part of the Late Pleistocene. The neighbouring site of Devil’s Tower produced the partial skull of a Neanderthal child in 1926 (Garrod et al., 1928), and has greater potential for further excavation and discoveries. It preserves much more Pleistocene sediment than Forbes’ Quarry, and it is possible to relate that sediment to the previous excavations. Neanderthal archaeological sites There are several other sites in Gibraltar which preserve evidence of Neanderthal occupation. One, Ibex Cave, lies high up on the eastern face of the Rock, while four others lie to the southeast, close to the sea near “Governor’s Beach”. The present beach mainly consists of fine limestone blast debris from military tunnelling operations, but there are also cemented remnants of more ancient beaches which presumably accumulated during Oxygen Isotope Stage (OIS) 5. The caves are named (from the south) Bennett’s, Gorham’s, Vanguard and Boat Hoist. Three of these caves (Ibex, Gorham’s and Vanguard) have been excavated since 1994 as part of the Gibraltar Caves Project and the project PalaeoMed. Gorham’s Cave Gorham’s Cave was discovered in 1907 by Captain A. Gorham of the 2nd Battalion Royal Munster Fusiliers, who opened up a fissure at the back of the cavity which bears his name. Subsequently, for convenience, both the cavern and the system of fissures came to be known as Gorham’s Cave. The cave appears to have been forgotten after 1907, although it may have been visited sporadically by military speleologists. However, on 16 March 1945, Lieutenant George Baker Alexander, R.E., a graduate geologist from Cambridge University, arrived in Gibraltar and conducted a thorough geological survey of Gibraltar at this time, concluding with the production of a new geological map of the region (Rose and Rosenbaum, 1990). Alexander became the first person to excavate Gorham’s Cave, along with his companion, Lt. Monke. Both set out to excavate the upper layer of the site. Alexander’s work, how- 509
STRAIT OF GIBRALTAR  termed    Forbes Barrier   , which is situated near the entrance into the fortress from the neutral g...
510 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Waechter’s excavations represented the first large-scale excavations in Gorham’s Cave and established that it contained a record spanning perhaps 100 kya of Middle Palaeolithic, Upper Palaeolithic and Holocene occupation (Waechter, 1951, 1964). Waechter reported the presence of ancient hearths at various levels in the cave, and of faunal material throughout the sequence, dominated by the remains of ibex, rabbit and many species of bird. Unfortunately, many aspects of Waechter’s excavations were never properly recorded or published, and much of the material he recovered has since disappeared. On the other hand, Waechter’s stratigraphic sequence of layers running approximately horizontally east-west must have been simplified considerably compared with the complex reality which has since been observed. Figure 2. Stratigraphic profile of Gorham’s Cave: Schematic profile of the outer sector (middle area of the cave) modified from Collcut (2013) in Barton et al., (2013) (left) and stratigraphic profile of the inner sector (right). ever, was not viewed well by the Gibraltar Museum Committee of the day. At about the same time (spring of 1948), the then governor, Sir Kenneth Anderson, presumably on the advice of Padre Brown, stopped further digging and wrote to the British Museum asking them to continue any further explorations. However, the British Museum had no staff available and the governor’s letter was forwarded to Prof. Dorothy Garrod at Cambridge University, who had excavated Devil’s Tower Rock Shelter in 1927–28. She was unable to undertake the work, and asked Dr John D’Arcy Waechter, fellow of the British Institute of Archaeology, Ankara, to fit the work in with his own programme in Turkey. The second phase of systematic excavations was carried out by a joint team from the Natural History Museum, London, led by Dr Christopher Stringer, and the British Museum, London, and by Ms Jill Cook, who visited Gibraltar in 1989. After preliminary excavations, the work developed as the ‘Gibraltar Caves Project’, jointly directed by the Gibraltar Museum and the Natural History Museum, London. Work until 1997 focused on the outer part of the cave, which had previously been excavated by Waechter (1951, 1964). Since 1997, the project direction has expanded to include the Museo de El Puerto Santa María and the University of Huelva. It is at this stage that the excavations in the inner part of Gorham’s Cave commenced; their first results were published by Finlayson et al., (2006). The excavations in the outer area have been recently described by Barton et al., (2013), shedding light on the sedimentary formation of the cave with a stratigraphic sequence of more than 16 m in thickness (Fig. 2). This sequence is composed mainly of earthy materials covering a cemented beach-rock deposit which presumably accumulated during OIS 5. The nature and sedimentary structures of the sediments filling the cave show a massive aeolian accumulation related to transgressive coastal dunes that migrated during OIS 3 highstand substages and/or cold, arid periods (Jiménez-Espejo et al., 2013). The stratigraphic series include dark-brown organic-rich silty clay, grey sand and irregularly bedded yellowish-brown sand, brown-black organic-rich clay with whitish gritty phosphatic lenses and interbedded, massive, homogeneous, coarse brown sand (Collcut, 2013).
510  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Waec...
STRAIT OF GIBRALTAR Radiocarbon dates of between ca. 29 and 51 kyr BP were obtained for UBSm.7 and BeSm.1; nevertheless, the dates from the underlying LBSmff.1–5 (ca. 42 and 56 kyr BP) seem to suggest that most charcoal fragments could have been derived from lower down the sequence (Higham et al., 2013). The single-grain (SG) optically stimulated luminescence (OSL) chronology and the Bayesian age model yielded an age of MIS 5 near the base of the stratigraphy (119,300±14,800 kyr for CSm; Rhodes, 2013a). The excavations in the inner area exposed an area of ≈29 m2 of bedrock and a stratigraphic sequence formed by four archaeological levels (IV–I from bottom to top; Fig. 2). The stratigraphic composition is different from that of the other sectors, displaying local rock falls, aeolian dust and mainly karstic clay (Finlayson et al., 2006). The sedimentary deposit is thinner (<2 m) than that of the outer area due to the higher position of the cave substrate. Levels I and II correspond to Phoenician and Neolithic horizons, respectively. Level III (mean depth of ≈60 cm) is subdivided into a basal Figure 3. Mousterian tools from Level IV of Gorham’s Cave. Solutrean (IIIb) and an upper Magdalenian (IIIa) horizon. A distinctive feature of the middle part of this level is the high proportion of fallen fragments of angular limestone and speleothem. Levels III and IV are clearly differentiated by their textural composition, since level III consists of sandy sediment with dark brown clay in a sandy matrix, while level IV is characterised by a beige-coloured pure clay horizon (Finlayson et al., 2006). Regarding lithic assemblages, the outer stratigraphic sequence is consistent with the Middle Palaeolithic techno-complexes in its middle and lower part. The knapping technique mainly follows discoid reduction sequences, although a significant increase in laminar flakes coming from bipolar Levallois cores is observed at SSLm.5–6. The last moments of the Middle Palaeolithic seem to be represented by a Levallois point from UBSm.4, as CHm.5 is the first attributed to the Upper Palaeolithic (Barton and Jennings, 2013). In the inner area, level IV corresponds to a Mousterian technocomplex (Giles Pacheco et al., 2012; Shipton et al., 2013; Fig. 3). All the lithics from this level – in flint, sandstone, limestone and others – are made from autochthonous raw materials from the fossil beach deposits near the caves and from the levels of flint immersed in the Jurassic units of the Rock. The characteristics of the assemblage indicate discoidal and Levallois reduction methods. Some cores show unipolar orthogonal and opposite bipolar reductions. The tools from Level IV show a predominance of sides-crapers and denticulates. Notches and abrupt retouches are also represented. The metrics of the flakes seem to be conditioned by the size of the pebbles, especially in the case of flint, since the nodules in the beach breccias are small. In contrast, the technology from the overlying Level III is characteristic of the Upper Palaeolithic, with diagnostic pieces attributable to the Solutrean and Magdalenian (Giles Pacheco et al., 2012). Palaeobotanical (charcoal and pollen) samples from Gorham’s Cave have revealed a diverse Mediterranean landscape during the Middle and Upper Palaeolitihc, covering the stratigraphic sequence of the cave (Carrión et al., 2008). Inferred vegetation types include oak, pine, juniper and mixed woodlands and savannahs, grasslands with heaths, heliophytic matorrals, phreatophytic formations (such as wetlands and riverine forests) as well as a thermomediterranean coastal scrub. The macro-mammals do not show marked fluctuations 511
STRAIT OF GIBRALTAR  Radiocarbon dates of between ca. 29 and 51 kyr BP were obtained for UBSm.7 and BeSm.1  nevertheless, ...
512 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD through time, as they appear taxonomically constant through the stratigraphy with a predominance of two ungulate species – Cervus elaphus and Capra ibex (Currant et al., 2013a). Only the presence of grey seal (Halichoerus grypus) can be interpreted as punctual evidence of a cold phase in CHm (D unit in Waechter, 1951, 1964). At least 33 taxa of amphibians and reptiles have been recovered, including 24 in the inner area, including newts, toads, frogs, tortoises, turtles, lacertid and scincid lizards, geckos and several snakes (Blain et al., 2013). In the outer area, the largest assemblage comes from LBSmcf.11 and involves 21 species. The most frequent specimen is the western spadefoot toad (Pelobates cultripes; Gleed-Owen and Price, 2013a). Regarding small mammals, five species show predominance along the sequence – Oryctolagus cuniculus, Apodemus sylvaticus, Eliomys quercinus, Microtus brecciencis and Terricola (Microtus) duodecimcostatus (López-García et al., 2011; Price et al., 2013). The inner chamber also has an important representation of Myotis myotis (López-García et al., 2011). The outer area has yielded a significant large assemblage of bird species, with at least 90 species (seabirds, ducks, birds of prey, partridges, waders, pigeons, swifts, crows and small passerines), which were registered by Cooper (2013a). Put together with recent finds in the inner chambers (Sánchez-Márco, in prep.) the total Pleistocene avifaunal list for Gorham’s Cave is currently at 142, the highest recorded in any Pal- aeolithic archaeological site. The fossil birds from Gorham’s and Vanguard Caves have been used in the quantification of the habitats outside the cave (Finlayson, 2006) and data from birds, amphibians, reptiles, micro-mammals and intertidal molluscs have been used in climate reconstruction at scales down to seasonal (Finlayson, 2006; Ferguson et al., 2011; López-García et al., 2011; Blain et al., 2013) Finally, Finlayson et al., (2012) showed an association involving the direct intervention of Neanderthals on the wing bones of raptors and corvids, which was interpreted as evidence of extraction of large flight feathers (Fig. 4). Vanguard Cave Vanguard Cave, located on the southeast face of the Rock of Gibraltar, is one of four caves which make up the Gorham’s Cave complex. Vanguard Cave shows a stratigraphic sequence which is less complex than that of Gorham’s Cave (Fig. 1). It contains 17 m of deposits, mainly composed of massive, coarse-to-medium sands intermixed with tabular-to-lenticular units of silts and silty sands (Macphail and Golberg, 2000). Most of the Vanguard sediments are calcareous, with little diagenesis. In the upper area of the cave, the sands are interdigitated with black humic clays, showing evidence of phosphatisation. However, the sediment deposits at Vanguard are generally less phosphatic and organic, and exhibit fewer diagenetic Figure 4. Examples of cut-marks on corvid wing bones from the Middle Palaeolithic levels of Gorham’s Cave: (A) Proximal diaphysis of Pyrrhocorax pyrrhocorax humerus (GOR’96 NO. 299); (B) proximal diaphysis of Pyrrhocorax graculus ulna (GOR’00/B5/NIV/57). Images taken from Finlayson et al., (2012).
512  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  thro...
STRAIT OF GIBRALTAR Figure 5. Section of Vanguard Cave showing excavation areas along the dune deposit. Graph taken from Stringer et al., (2008). changes compared to Gorham’s Cave (Macphail et al., 2013). Five main excavation areas (A–E) were established at different heights of the dune, with A being the highest area and E the lowest one (Fig. 5; see Stringer et al., 2008 and Macphail et al., 2013 for more details). Sectors A and B are described as Upper area, C and D as Middle area and E as Lower area. From these areas, sediment samples were collected for optically stimulated luminescence (OSL). Quartz OSL data, including singlegrain (SG) measurements, indicate that much of Vanguard Cave was filled around the time of OIS 5 (Rhodes, 2013b). This age estimation is older than the OSL chronology reported by Pettit and Bailey (2000), but for Rhodes (2013b), this is primarily due to a difference in dose rate estimation. The OSL date from the uppermost part of the sequence yielded an age of 75 kyr (when the cave was practically silted). This age was obtained from breccia fixed in the wall of the cave; however, erosion phenomena could have altered some superficial sand layers, generating new earlier deposits. Macroscopic analyses of charcoals indicate very little evidence of taxonomical change, with an arboreal landscape apparently dominated by warmclimate vegetation. Pistacea sp. and Olea sp. are registered as thermophilous indicators along the sequence, since both species are located within thermo-Mediterranean bioclimates (<600 m.a.s.l.; Ward et al., 2013). The examination of the reptile and amphibian assemblages carried out by GleedOwen and Price (2013b) yielded a minimum of 17 species in the middle excavation area. A core of four species, including western spadefoot toad (Pelobates cultripes), stripeless tree frog (Hyla meridionalis), Moorish gecko (Tarentola mauritanica) and worm lizard (Blanus cinereus), was detected 513
STRAIT OF GIBRALTAR  Figure 5. Section of Vanguard Cave showing excavation areas along the dune deposit. Graph taken from ...
514 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD at Vanguard Cave. Of these, B. cinereus is obligate thermophile, T. mauritanica does not tolerate extreme cool environments and H. meridionalis is restricted to meso-Mediterranean biomes with an overall dry climate. On this basis, the herpetofaunal remains suggest a remarkable environmental stability, with thermophilous species that are currently restricted to southern Europe (GleedOwen and Price, 2013b). A total of 73 bird species have currently been identified from Vanguard Cave (Cooper, 2013b; Sánchez-Marco in prep.) and are the subject of ongoing ecological and biogeographical analysis. Macro-mammals show little variation, indicating environmental stability during the deposition period. The Middle area is characterised by the presence of ibex (Capra ibex), red deer (Cervus elaphus), wild boar (Sus scrofa) and bear (Ursus arctos), as well as evidence of marine mammals (seals and dolphins). Almost 50% of bones show human-induced damage (e.g. cut-marks, percussion marks) affecting ibex, red deer, wild boar and seal, and only 3% bear carnivore tooth-marks (Currant et al., 2013b). In addition to the terrestrial fauna, two monk seal fossils (Monachus monachus) show human alterations on a proximal phalanx and a scapula (Stringer et al., 2008). In association to terrestrial and marine mammals, molluscan shells such as Mytilus galloprovincialis, Callista chione, Acanthocardia tuberculata, Patella vulgate, P. caerulea and a few barnacles (Balanus sp.) are widely documented. The study of the lithic artefacts reported by Barton (2013) comes from the Upper and Middle areas of the cave and shows clear assignation to the Middle Palaeolithic techno-complex. Lithic industry suggests little variation through the sequence, with little change in the dominance of quartzite over finer-grained cherts. This reflects the more common availability of this material in comparison to other raw materials. Limestone from the cave bedrock was often used to make artefacts, representing a significant expedient behaviour. For Barton (2013), the low diversity of raw material and the limited range of tools in the assemblages are concordant with a succession of short-term human occupations at the cave. Currently, new fieldwork is being carried out at the Upper part of the sequence. Our aim is to determine the complete stratigraphic sequence of the cave, developing a new programme of dates, and to deepen our understanding of Vanguard Cave during the Neanderthal occupation phases (see Fig. 1-B1, B2 for the new excavation).
514  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  at V...
8 central plateau CENTRAL PLATEAU 515
8  central plateau  CENTRAL PLATEAU  515
516 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Site Map numbering Ambrona and Torralba 86 Cuesta de la Bajada 87 Jarama VI 88 Conjunto de Atapuerca 89 Manzanares and Jarama 90 La Peña de Estebanvela 91 Pinedo 92 Pinilla 93 San Quirce 94 Cueva de Maltravieso, Cueva de Santa Ana, Cueva de el Conejar, Vendimia and El Millar 95 Siega Verde 96 Valdegoba 97
516  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Site...
CENTRAL PLATEAU Manuel Santonja*, Alfredo PérezGonzález*, Joaquín Panera**, Susana Rubio-Jara**, Carmen Sesé***, Enrique Soto***, Laura Sánchez-Romero*, **** Ambrona and Torralba archaeological and palaeontological sites, Soria Province 1. Discovery and early research 1.2. Research resumed by F.C. Howell (1960-1963) The Torralba and Ambrona sites (Santonja et al., in: Santonja and Pérez-González, 2005 (eds.): 18-39) are 150 kilometres north east of Madrid on the watershed between the Ebro, Duero and Tagus Rivers, in the southern part of Soria Province. This is a strategic transit zone between the highlands of the Iberian Meseta (plateau) and the Jalón River Valley. An oblique perspective generated by a digital terrain model shows the two sites between reliefs drained by tributaries of the Atlantic slope Duero and Tajo Rivers. To the east, the clear outline of the Jalón River canyon, which flows into the Ebro and eventually into the Mediterranean (Fig. 1). In Spain after the Civil War, in 1936, Lower Palaeolithic research was abandoned almost entirely. Only occasional visits to international congresses awakened memories of Torralba and Ambrona and spurred individual initiatives such as the palynological studies by J. Menéndez Amor and F. Florschütz in 1959 and 1963, which focused on the Middle Pleistocene chronology of the sites (Santonja and Vega, 2002). 1.1. First work by Marquis of Cerralbo (1909-1916) The discovery of both sites began with the detection of large elephant bones in 1888 at the Torralba railway station. Between 1909 and 1913, Enrique de Aguilera y Gamboa (1845-1922), the 17th Marquis of Cerralbo, excavated over 2000 m2 in Torralba. From 1914 to 1916, he continued his work at the Loma de los Huesos in Ambrona site, 2.5 km to the north. His results had a great impact at the time and drew visits by leading researchers. Cerralbo’s conclusions, paradoxically from a creationist’s ideological perspective (Santonja and Vega, 2002), conjugated the key aspects around which the site was later interpreted –organised hunting of elephant herds by a group of human settlers on the banks of a lake– and provided a glimpse of the potential importance of these sites for the study of human behaviour (Isaac, 1977: 3-4). * ** *** **** Contacts between Luis Pericot and Clark Howell at the Panafrican Prehistory Congresses led to the recommencement of the research work, nearly 50 years later of the Cerralbo’s work. Howell proposed systematic full-cover, multidisciplinary excavations for Torralba and Ambrona. By the end of the decade, research into the African Pleistocene had implanted this model, but in 1960 it was a novelty in Europe. When this work began, the Torralba site was estimated to cover approximately 3800 m² (Howell et al., 1962), of which over 2000 m2 were to be preserved, of which 1026 m2 dug between 1961 and 1963 (González Echegaray and Freeman, 1998). Loma de los Huesos in Ambrona was estimated to cover roughly 6000 m2, of which 1243 m2 were excavated in 28 weeks during the 1962 and 1963 digs (Howell, 1965). Both sites were studied by the same team –K.W. Butzer, E. Aguirre, P. Biberson and L.G. Freeman– who used a similar methodology and reached common conclusions. Butzer’s geological survey attributed an identical age and formation processes to both sites and defined a morphosedimentary unity, the “Torralba Formation”, which integrated the strati- Centro Nacional de Investigación sobre la Evolución Humana. (CENIEH); 09002 Burgos manuel.santonja@cenieh.es IDEA (Instituto de Evolución en África), Museo de los Orígenes, Plaza de San Andrés 2, 28005 Madrid. Departamento de Paleobiología. Museo Nacional de Ciencias Naturales. Madrid. Escuela Interuniversitaria de Posgrado en Evolución Humana, Universidad de Burgos. C/ Juan de Austria 1, 09001 Burgos. 517
CENTRAL PLATEAU  Manuel Santonja , Alfredo P  rezGonz  lez , Joaqu  n Panera  , Susana Rubio-Jara  , Carmen Ses     , Enri...
518 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. Geographic location of the Torralba and Ambrona sites in the south of Soria province, 150 km NE of Madrid (Spain). Both sites are in the valley of La Mentirosa arroyo Rivulet, also known as Mansegal, a tributary of the Jalón River which flows into the Ebro. The Bordecorex stream flows into the Duero River, while south of Olmedillas and Ventosa del Ducado, the river network flows into the Henares River (Tagus basin). graphic sequences of the two sites, described at the time as twin (Butzer, 1965). The published studies of the lithic tools, all preliminary, interpreted the industry at either sites as early (Freeman, 1975) or middle (Biberson, 1964) Acheulean. Biberson and Aguirre also noted the existence of worked bone, an issue debated subsequently and pending a systematic review (Domínguez-Rodrigo, in Santonja and PérezGonzález (eds.), 2005: 282-287). This first research stage led to an interpretation of the sites which in some general aspects matched Cerralbo’s imaginative foresight. The Mansegal or La Mentirosa stream valley connected the highlands plateau of the North with the Ebro Basin, and was probably a corridor frequented by herbivores during their seasonal migration (Butzer, 1971). The presence of these herds led groups of hominins –in a display of premonitory behaviour– to burn the veg- etation in order to drive them into swampy zones where weeds and mud hampered the animals’ movement. In these conditions would have been easy to kill them, dismembered in nearby spaces and prepared for consumption (Howell, 1966). 1.3. The Howell-Freeman period (1980-1983) The complete aperture of East Africa to Pleistocene research led Howell to interrupt his Spanish work in 1963. The large interdisciplinary teams which started to work at African sites proved decisive and triggered profound changes to the methods used in Palaeolithic archaeology from the 1970’s onwards. In this context, the interpretations of Torralba and Ambrona were reviewed by Binford, who found no arguments in support of organized hunting, expressed doubts about the presence of fireplaces and questioned whether the areas containing bones had
518  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
CENTRAL PLATEAU remained in a primary position. Even in the absence of different data from those published by Howell and his team, Binford suggested that natural agents and trampling by elephants had caused intense modifications (Binford, 1987). The debate was in full swing when a new phase of research at Ambrona began in 1980. With the addition of 207 m2 dug by E. Aguirre in 1973, by the start of the 1980 season, almost 1450 m2 had been excavated at Ambrona. During this stage, under the joint leadership of Howell and Freeman and the management of M. Almagro –responsible for channelling relations with the Spanish government and facilitating funding for the project under the Spain-US Cultural Cooperation Programme–, an additional 1267 m2 were dug in 203 days. By 1983, 2717 m2 of the estimated 6000 m2 of the Ambrona site had been excavated. In addition, work continued on the hillside opposite Loma de los Huesos (Camp North). Here, although published references are quite vague (Howell and Freeman, 1982), we know that a little over 200 m2 were dug, 55m2 in 1963 and 162 m2 in 1981 and 1983, according to unpublished documents held at the Numantine Museum in Soria. The Ambrona assemblage remained attributed to Butzer’s “Torralba Formation”, with new aspects in the stratigraphic interpretation (Howell et al., 1995), primarily the differentiation of two members, the “Lower complex” and the “Upper complex”. The former included the characteristic concentrations of megafauna –particularly elephant– and Acheulean industry. An intermediate occupation was defined in the central part of the site, with fauna and sporadic industry considered close or equivalent to Camp North, where deer, aurochs and elephant were recorded along with some Acheulean handaxes. At the “Upper Complex”, on the levels of alluvial and colluvial origin, a more frequent and more evolved lithic industry than the Lower complex was detected, also identified as Acheulean. Scarcely any elephant remains were detected, replaced in importance by Equus (Howell et al., 1995). At the end of this stage, the older interpretations of Ambrona’s Lower complex were accepted with certain nuances. The faunal remains were basically regarded as residue from deliberate hunting activity and the processing and consumption of the food. The hominins had also shifted substantial portions of this prey to their base camps, whose location was imagined –in the absence of evidence– to be on high ground overlooking the area, “overlooking the valley” (Freeman, 1994). 2. Current state of research at Torralba and Ambrona 2.1. Digs from 1990 to 2000 and subsequent work The controversy over the nature of these sites continued into the late 1980’s, and the published information was still insufficient to be able to test the hypotheses proposed in the light of the previous excavations. Substantial unknowns still remained about the general sedimentary processes and also the microstratigraphy and the spatial distribution of the remains. It was known for certain that there were still large unexcavated areas in Ambrona and probably in Torralba as well. Consequently, in 1990 –following an excavation permit granted to E. Carbonell in 1988 which was not continued and there are no known results– another project headed by M. Santonja and A. Pérez-González was begun using geoarchaeological, taphonomic, and techno-economic methods. In summary, its aims were to understand the morphodynamic and sedimentary processes in order to contextualize the megafauna assemblages and interpret the human activity in this area (Santonja, 1989). An initial stage between 1990 and 1991 ascertained the basic features of the local geomorphological evolution and situated the sites in a geological framework at the local and regional scale (Pérez-González et al., 1991). Annual digs were conducted at Ambrona from 1993 to 2000, focused on the “Lower complex”, with 688 m2 excavated. In Torralba, work was hampered by the large volume of waste material from previous excavations dumped on the site itself, and thus only limited test pits were dug (Santonja et al., in Santonja and Pérez-González (eds.), 2005: 104-123). In 2001, a vertical electrical sondage was lowered into a sinkhole in Jurassic dolomites 200 m NE of Ambrona, where Pleistocene fauna was recorded. In 2001 and 2002, samples were removed for magnetostratigraphy as well as dating by luminescence and aminoacid racemisation (Parés et al., in Santonja and Pérez-González (eds.), 2005: 190-199; Falguères et al., 2006). Finally, in 2013 the National Human Evolution Research Centre (CENIEH) began fresh digs focused initially on the middle stratigraphic member –partly equivalent to the Upper Complex of Howell (PérezGonzález et al., in: Santonja and Pérez-González, 2005 (eds.): 176-199)–, in Ambrona and Torralba. 519
CENTRAL PLATEAU  remained in a primary position. Even in the absence of different data from those published by Howell and ...
520 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 2. A: Stratigraphy of Ambrona, lower and middle members, in the central sector of Loma de los Huesos. B: Relative geomorphological position of the Torralba and Ambrona archaeological sites. 2.2. Results from Ambrona and Torralba sites between 1990 and 2000 2.2.1. Geomorphology and chronology of the sites Ambrona is at the bottom of the karst valley or polje between Torralba, Ambrona and Conquezuela, while Torralba is in the valley of the Mansegal, built up from the bottom of the polje and set on the + 35 m terrace of the current valley (Pérez- González et al., in Santonja and Pérez-González (eds.), 2005: 176-199). A cut and fill phase of the Mansegal stream, and another of the +35 m terrace development and subsequent incision on this terrace separate the two sites (Fig. 2b), categorically invalidating hypotheses which merged their stratigraphies in a single unified sequence (Butzer, 1965). The chronological distance between Ambrona and Torralba has also been checked by numerical
520  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
CENTRAL PLATEAU dating. The middle stratigraphic unit of Ambrona has been dated at c. 350 ky by ESR/U series (Falguères et al., 2006), suggesting an age of around 400 ky for the lower unit of this site. In the case of Torralba, OSL datings currently in press (N. Mercier) indicate around 200 ky, similar to the 220 ky and 240 for U series (Howell et al., 1995), obtained on a +20-25 m terrace (comparable to Torralba) in the nearby Alto Henares area. This reinforces the geomorphological interpretation which refutes the contemporaneity of the sites and instead suggests a sequence of occupations in the area. The Torralba site at the bottom of the Mansegal River valley is close to MIS 7 (243-192 ky), while the lower and middle stratigraphic units of Ambrona correspond to MIS 11 (424-375 ky) and MIS 10 (374-338 ky) or 9 (337-301 ky) respectively, in positions related to the small ponds and the drainage network which developed at the bottom of the Conquezuela polje. 2.2.2. Ambrona stratigraphy The stratigraphic units defined in Ambrona correspond to fluvial and fluvio-lacustrine environments (Pérez-González et al., in: Santonja and Pérez-González, 2005 (eds.): 176-199), each one with a different preservation potential for remains. A systematic, integrated stratigraphic interpretation has been proposed for the site, which we have called “Ambrona formation”, consists of the lower, middle and upper members. Excavation campaigns from 1993 to 2000 were carried out in the lower member, that includes the following levels (Fig. 2a): – A fluvial (AS1) and another fluvio-lacustrine level (AS4) which contains the highest density of lithic and faunal remains, albeit in a derivative position, partly brought from outside the preserved site. – Level AS3, composed of mud built up at the bottom of a shallow pond with occasionally input via small channels. This level has the best conservation conditions for industry and fauna remains, mainly found in a primary position. – Other minor levels in the lower stratigraphic unit are of a fluvial nature (AS1/2 and AS2). At the top of this unit is the fluvio-lacustrine level AS5. The presence of archaeopalaeontological remains is sporadic in all these cases. All these levels were mapped, defining the vertical and horizontal relationships of the stratigraphy across more than 400 linear metres. While AS1, AS5 and AS6 are spread across almost the entire site, other levels cover smaller areas, resulting in different stratigraphic sequences in each zone. Thus, in the Central and Western sectors there are several areas where AS1/2, AS2 and AS3 are absent and the AS4 level lies directly in contact with AS1. The middle member of the Ambrona formation (AS6) includes fluvio-lacustrine deposits in the Central and Western Sectors and fluvial deposits in the Eastern Sector. The latter are rich in lithic industry and they also contain fauna. In some cases they occur in overbank facies, accumulated in low energy conditions which enabled the remains be found in an almost primary position. The Ambrona formation is completed by the upper member (AS7), composed of channelled and edaphized facies. It is archeologically and palaeontologically sterile. 2.2.3. Ambrona palaeontology The results from the 1990-2000 stage almost coincide with those obtained previously, with some further details, particularly about birds and small vertebrates. The mammal association identified between 1993 and 2000 (Sesé and Soto, in: Santonja and Pérez-González 2005 (eds.): 258281) is the following: Crocidura sp.; Microtus (Iberomys) brecciensis; Arvicola aff. sapidus; Apodemus aff. sylvaticus; Oryctolagus sp.; Canis lupus cf. mosbachensis; Panthera (Leo) cf. fossilis; Palaeoloxodon antiquus; Stephanorhinus hemitoechus; Equus caballus torralbae; Cervus elaphus; Dama cf. dama; Capreolus sp. and Bos primigenius. Other taxa identified previously in the lower Ambrona member must be added to this list (Howell et al., 1995): Vulpes sp., Crocuta crocuta aff. praespelaea and Megaloceros aff. savini. In the middle member, almost the only species recorded to date from the 1993-2000 excavations is Equus caballus and very occasional remains of Palaeoloxodon antiquus. The presence of avian fauna is recorded (Sánchez Marco, in: Santonja and Pérez-González 2005 (eds.): 248-257): Anser anser; Tadorna ferruginea; Tadorna sp.; Anas acuta; Anas strepera; Anas sp.; Mergus merganser; Anseriformes indet.; Fulica cf. atra; Otis tarda and Vanellus vanellus; A herpetological sample (Martínez-Solano and Sanchiz, 521
CENTRAL PLATEAU  dating. The middle stratigraphic unit of Ambrona has been dated at c. 350 ky by ESR U series  Falgu  res ...
522 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD in: Santonja and Pérez-González 2005 (eds.) : 232239) includes: Bufo bufo; Bufo calamita; Discoglossus cf. jeanneae; Hyla arborea; Pelobates cultripes; Pelodytes punctatus; Rana perezi; Rinechis scalaris; Natrix sp.; Lacertidae and Colubridae indet. There was also some ichthyologic remains (Perea and Doadrio, in: Santonja and Pérez-González 2005 (eds.): 240-247) of Chondrostoma arcasii. The macromammal series is considered characteristic of the advanced but not final Middle Pleistocene, with more modern elements than the peninsular faunas from early Middle Pleistocene sites such as Cúllar de Baza I (Granada) or Buenavista, Campo de Tiro and Polígono Industrial (Toledo), characterized by Mammuthus trogontherii. The micromammals show a similar chronology. The morphology and size of Microtus (I.) brecciensis teeth imply a previous age to those found in late Middle Pleistocene populations of the same species. The presence of a relatively large form of the Arvicola aff. sapidus species indicates a more modern phase than the fauna from Cúllar de Baza I, with Arvicola mosbachensis. The evolutionary stage of the Ambrona microfauna corresponds to the third ensemble of Middle Pleistocene associations defined by Sesé and Sevilla (1996), which include fauna such as that found in Áridos. 2.2.4. Palaeoenvironments In addition to the study of the fauna and nanofauna –ostracods–, palynology and biomineralizations –phytoliths– have provided information about the environmental conditions when the deposits of Ambrona formation were accumulated (Baltanás et al., 2005; Ruiz Zapata et al., 2005; Pinilla et al., 2005, in Santonja and Pérez-González (eds.), 2005: 200-231). All the conclusions indicate the existence of climatic constants which were comparable to the current conditions, albeit with certain nuances. The macromammals from the lower member of Ambrona indicate a good representation of forest environments and open lands, with meadows and areas with abundant water in relatively warm and moist climatic conditions. Birds confirm the proximity of wetlands, flooded zones and shallow water bodies. Some species like the common goose and lapwing denote flat or gently undulating grassland. Taxa such as swamphen, coot and northern pintail require thick patches of vegetation around water bodies. The swamphen, a sedentary animal, is incompatible with very low temperatures. The herpetofauna corroborates these interpretations and indicates that the conditions were similar to today, with a more or less contrasted seasonality, less dry summers and less wet springs and winters, with slightly higher winter temperatures. Locally, the presence of ostracods as Leucocythere cf. mirabilis in several levels indicates a lake system in oligotrophic conditions. The taphocenosis found in the ostracods is similar to current conditions in shallow ponds and lakes in southern Europe. Sometimes, Heterocypris salina became predominant at Level AS6 (middle member), suggesting a drier and colder period, a trend also suggested by the dominion of Equus caballus in replacement of Palaeoloxodon antiquus. The silicophytoliths in the lower member suggest a temperate climate. The diatoms are often epiphytes, indicating a frequency of aquatic plants. Biominerals are less abundant at the top of the lower member due to the changed environmental conditions. Silicophytoliths are more abundant, but with many spicules and diatoms reduced to the Amphora genus, reflecting quite stressful environmental conditions for microorganisms. The reduction in the number and variety of biominerals is greater in the middle member: virtually all are silicophytoliths and almost all C3 grasses. Hantzschia amphioxys, a species that can survive in a wide range of saline environments, predominates amongst the diatoms. There are almost no biominerals at the top of AS6, suggesting that the lake may have dried up. Pollen analysis describes vegetation mainly consisting of pine (Pinus), juniper/sabine (Juniperus) and grasses (Poaceas). The riparian taxa such as alder (Alnus), willow (Salix), elm (Ulmus) and characteristic swamp species (Cyperaceas, Ranunculaceas, Typha...) were present. Temperate trees such as deciduous oak (Quercus), birch (Betula), chestnut (Castanea), hazel (Corylus) and walnut (Juglans) were also detected. These results corroborate the predominance of a milder climate than today’s conditions at the time of the lower member accumulation. 2.2.5. Human presence The technical characteristics of the lithic industry and the sources of the raw material have fostered debate about the mode and intensity of the site’s use. With regard to bone industry, recent studies (Domínguez-Rodrigo, in: Santonja and Pérez-González, 2005 (eds.): 282-287) do not support the hypothesis suggested primarily by
522  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  in  ...
CENTRAL PLATEAU Aguirre and Biberson about an intense transformation of elephant bones by shaping and retouch. The possibility that the fragments corresponding to the tip of infantile elephant tusks found at both sites deriving from any kind of manufacturing has also been rejected (Villa and d’Errico, in: Santonja and Pérez-González, 2005 (eds.): 288-305), arguing that they broke off naturally in the course of the elephants’ lives. To understand the significance of the Ambrona lithic industry, we must take into account its taphonomic history (Santonja et al., in: Santonja and Pérez-González, 2005 (eds.): 306-333). AS1 and AS4, the richest levels in the lower member, are fluvial deposits with a degree of energy. The industry they EXCAVATED AREAS IN THE LOWER STRATIGRAPHIC MEMBER OF AMBRONA IN 19932000 Level AS1: 535m2. Only 35m2 complete. Only the surface of the level in the rest contain is not in a primary position, having been dragged from its original positions in the immediate vicinity of the site. This material and the fauna found on the same levels was carried and classified by size by the watercourse. The technological imbalance in the series from AS1 and AS4 are not due to any palaeoeconomic or functional factors but rather to the natural process which formed record. On AS3, however, the industry is mainly in a primary position, albeit with some items deposited along the small streams leading to the pond. It is very low-density and essentially contains final elements of chaînes opératoires with little or no shaping, such as non-retouched flakes selected by size and form, and also bifacial macrotools brought from outside and left on the site. LITHIC INDUSTRY 235 items, including 9 handaxes and 5 tools. Density in 35 m2: 5 items/m2, 1 handaxe/5 m2 and 1.25 tools/m2 (= 5 items/4 m2) Level AS3: 250 m2 72 items, including 2 handaxes and 17 tools Density in 250 m2: 1 item /3.5 m2; 1 handaxe/ 125 m2 and 1 tool/15 m2 Level AS4: 379 m2 339 items, including 1 handaxe and 56 tools. Density in 379 m2: nearly 1 item/m2;1 handaxe in 379 m2 and 1 tool/7 m2 Table 1. Excavated areas and lithic industry in the Lower Stratigraphic Member of Ambrona. The industry at AS1 and AS3 can definitely be ascribed to the Acheulean technocomplex, as with the other levels in the lower stratigraphic member, given that in AS4, hydraulic factor is responsible for the deficit in medium and large format of lithic items, which explains the absence of bifacial tools. The industry in AS6, previously defined as a more advanced Acheulean type than the lower levels (Howell et al., 1995), is characterized by the lack of true handaxes and cleavers, the development of retouched tools on flake and the presence of Levallois debitage (Fig. 3). This level corresponds to the early Middle Palaeolithic (Santonja and Pérez-González, 2006). 2.2.6. Ambrona palaeoeconomy The study of the sedimentation processes has enabled us to establish significant differences in relation to the meaning of the presence of remains in each stratigraphic context. A unified interpretation of the site, accepted until 1993, now seems inappropriate. The process by which the sequence was built up is a millenarian time period, but each level also comprises a major diachrony. Consequently, interpretations contextualised in short time intervals can only be applied to specific stratigraphic and spatial units. Moreover, they are obviously only meaningful for each case in point, and may differ even in areas that are part of the same level. The low lithic density suggests that human activity did not reach great intensity in the lower stratigraphic member. Although the small amount of Acheulean evidence in the surrounding area supports this interpretation (Rodríguez de Tembleque, in: Santonja and Pérez-González, 2005 (eds.): 334- 523
CENTRAL PLATEAU  Aguirre and Biberson about an intense transformation of elephant bones by shaping and retouch. The possib...
524 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 3. Lithic industry of the middle member of Ambrona site. Various sidescrapers: convergent pedunculate (1), angled-convergent (10), alternate angled-convergent (17), alternate with partial low retouch (6), doubles (8 and 12), concave (3), straight opposite a cortical back (11), sub-transversal straight (5), straights with invasive retouch (7 and 14 on débordant flake), straight with stepped retouch and Kombewa removal on ventral face (15) and sub-transversal convex (16). Denticulates (2 and 9). Levallois flakes (4, 12 –sidescraper– and 13). Flint, except for 8 (lidite) and 16 (quartzite).
524  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
CENTRAL PLATEAU 351). Under these circumstances it must be stressed that the Ambrona area was a recurrent point of attraction for several millennia, given that lithic industry has been recorded on all levels in the lower member. The general pattern for the procurement of raw material coincides in all of them, and it must also be noted that both the flint and the quartzite used here was brought from elsewhere, in the case of the flint, from sources up to several dozen kilometres away (Freeman, 1991; Parcerisas, 2006). The relationships between fauna –particularly elephants– and humans has not been established in all cases. The low incidence of freshly broken bones and cut marks indicates that human groups did not play a major role in the accumulation of these fauna remains. Current studies of the behaviour patterns of herds of elephants and other herbivores eloquently define the environments where remains of these animals build up in Africa: around ponds and springs, and during prolonged droughts (Haynes, 1991). A natural scenario such as the one indicated in the previous paragraph is what we propose for Ambrona (Villa et al., in: Santonja and Pérez-González, 2005 (eds.): 352-381). The concentration alpha at Level AS3, mainly consisting of a well-circumscribed adult elephant carcass, has been studied in depth (Fig. 4). This case provides an example of an individual which died from natural causes and was buried in mud, without evidence of human intervention. The presence of lithic industry and the few recorded cut marks indicates that the Palaeolithic groups only acted in the site in a marginal way on the fauna. 2.2.7. Torralba Work at Torralba between 1990 and 2000 has been much more limited than in Ambrona (Santonja et al., in: Santonja and Pérez-González, 2005 (eds.): 104-123). Nevertheless, stratigraphic checks, numerical dating and reviews of the industry confirm that this is an Acheulean site situated chronologically toward MIS 7 (vid. § 2.2.1). Figure 4. Remains of adult male Palaeoloxodon antiquus scattered across less than 60 m2 on Level AS3, Central Sector. Excavation campaign of 1995. Bones not detected in the excavated assemblage marked in black (upper left). 525
CENTRAL PLATEAU  351 . Under these circumstances it must be stressed that the Ambrona area was a recurrent point of attrac...
526 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 5. A: Torralba site. Zone partly dug by the Marquis of Cerralbo (1909-1913) and areas excavated by F. C. Howell (1961-1963). Equidistance of level curves: 25 cm. B: Torralba: composite stratigraphic profile (adapted from Butzer 1965).
526  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
CENTRAL PLATEAU The oldest formations on which the Torralba Pleistocene deposits lie are, as in Ambrona, red clays and gypsum from the Keuper and TriassicJurassic carbonate deposits. Despite the large area of the site, the only detailed stratigraphy is the N-S section in the western sector of the site (Fig. 5a), published by K.W. Butzer (1965). This is a composite profile (Fig. 5b) which starts with up to 3-4 m deep red colluvial facies lying on the Keuper. These deposits disappear to the north and the sequence continues with gray sand facies, interspersed with angular and sub-angular gravel with a carbonate composition, sized 1-3 cm along the major axis, reaching a maximum depth of 1 m in the northern half of profile. At the top there is a fairly continuous unit of grey-green marl, somewhat more sandy at the base, with a maximum depth of roughly 2 m. Above these facies, of a shallow lacustrine nature, there is a red alluvial-colluvial deposit, between 0 and 1.5 m deep. In this sector, the Pleistocene and Keuper levels are affected by reverse faults with movements of more than one metre. The above-mentioned river sand and gravel facies contain the main concentrations of fauna and industry found during the excavations by Cerralbo and Howell. The evidence of fluvial rolling found on the fauna and industry is typical of this type of context, and implies movement or disturbance of some intensity (Sánchez-Cervera et al., e. i. p.). The former interpretation of some of the Torralba stratigraphic units (Freeman and Butzer, 1966; Freeman, 1994) as occupation sites thus seems irrelevant, as they did not take into account the fact that these deposits were built up in fluvial contexts with enough energy to move the material. The same is true of the findings of the taphonomic studies of the Torralba fauna (Díez et al., 1985), which assumed the unitary nature of the assemblages on each level, which is are inconsistent with the secondary position (non-autochthonous) of the remains. The composition of the Torralba macrofauna closely resembles Ambrona, although there is a clear imbalance in the frequency of certain representative taxa such as Equus and Elephas. A parallel of any greater scope would be premature at this stage, given that only a few taxa –including Equus (Prat, 1977), poorly represented in the lower member of Ambrona– have been studied in depth, and the microfauna and small vertebrates –except for birds (Tadorna ferruginea; Mergus serrator; Anatidae indet. and Porphyrio porphyrio)– are still largely unknown also at Torralba. In Torralba, the lithic industry has a density of less than 1 item per m3 throughout the levels excavated by Cerralbo and Howell. As in Ambrona, such low frequencies suggest low-intensity human presence and interventions. The raw material collection patterns are also similar to those observed in Ambrona. The presence of cores and flakes shows that in Torralba, quartzite and flint blocks brought from elsewhere were exploited. But the lack of cortical flakes suggests also that the material brought to the site may have been previously scabbled. Similarly, the lack of cores used to produce the flake supports for handaxes and cleavers indicates that this toolkit was already configured when it was brought to the site (Sánchez-Cervera et al., i.p.). The handaxes and cleavers on flake set lets us include Torralba in the Acheulean technocomplex. However, progressive technological items have also been observed, such as handaxes and cleavers with retouch (bifaces support of tools), making further work necessary to check for technological traditions which may be ascribed to the Middle Palaeolithic, as in the case of Ambrona. In the general context of the European Acheulean, it is important to stress the presence of true flint hachereaux, since there is a clear tendency to link these items closely to the availability of quartzite. In the light of the numerical chronologies now available, Torralba is one of the most recent Acheulean sites in southern Europe. 527
CENTRAL PLATEAU  The oldest formations on which the Torralba Pleistocene deposits lie are, as in Ambrona, red clays and gy...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 528 Manuel Santonja*, Alfredo Pérez-González* Cuesta de la Bajada (Teruel): an early Middle Palaeolithic site 1. Discovery and excavation Cuesta de la Bajada (CB hereafter), situated on the left bank of the Alfambra River upstream from Teruel, was the first open air Middle Pleistocene site to be excavated in the Aragón Region. The identification of fauna and stone tools by Etienne and Nicole Moissenet prompted an initial survey and assessment in October 1990. Between 1991 and 1994, three brief excavations in a 30 m2 area focused on the most accessible zone, now called the Western Sector (SO), which yielded basic information about the site and its environs (Santonja and Pérez-González 2001). Subsequently, a thick accumulation of sterile sediment in the East Sector (SE) was removed, leaving a new surface of 350 m2 ready for excavation, of which 92 m2 were dug in successive campaigns between 1999 and 2011(Santonja et al., e.i.p.). The working methodology, which has remained constant since 1990, has adopted an essentially geoarchaeological approach with the participation of a large multidisciplinary team. 2. Geology and chronology The Alfambra River flows through Cuesta de la Bajada, overlap in the lower section of the site terrace, which show syn-sedimentary thickenings of over 60 m caused by subsidence of the rocky substratum of Pliocene carbonates and gypsums in this valley sector. This terrace is in a central location on the Alfambra 10 fill-strath stepped terraces, with relative heights given in relation to the current river level at +2-3 m (T10, present-day floodplain), +6 m (T9), +13-15 m (T8), +20-25 m (T7), +30-35 m (T6), +40-45 m (T5), +50-53 m (T4, Cuesta de la Bajada site terrace), +65-70 m (T3), +90-95 m (T2), and +103-104 m (T1). At CB there have been frequent lateral changes to the sedimentary facies, a quite common phenomenon in fluvial environments. In CB-SO, we * identified four successive floors of poorly stratified gravels containing large concentrations of fauna and lithic industry, all in non autochthonous positions (Santonja et al., 2000). In CB-SE, which is located in an upper stratigraphic position, we recorded a sequence of fine sediments built up in a pond environment. In this case, the archaeo-palaeontological materials are basically in a primary position. These deposits are in a small depression caused by deformation, located between a cyclical sequence of gravel bars and floodplain silts. At the bottom is a massive level with massive gravel (G) that lacks any archaeological record, successively followed by levels CB3, CB2 and CB1 with 1.5 m thickness (Fig. 1). The composition of the base level CB3 is sandy with high percentages of clays. Level CB2 has a greater presence of gravels (1-3 cm) and granules (2-4 mm) at the base, along with finer sands toward the top. The more recent Level CB1 also consists of a fining-upward sequence with a massive internal structure, albeit with a much finer texture, with granules and pebbles at the base and clays towards the top. Units CB3, CB2 and CB1 are capped with a 1 m-thick series of floodplain facies, denoted level P. All these levels correspond to low-energy sedimentary environments. A multidisciplinary study about the chronology of archaeological level CB3 was conducted using single-grain OSL and ESR dating in sedimentary quartz at the CENIEH lab (Burgos), multigrain aliquot OSL (Institut de Recherche sur les Archéomatériaux, Bordeaux), ESR/U series on Equus teeth (IPH, Paris), and also Amino Acid Racemization (AAR) analysis on Equus teeth (Biomolecular Stratigraphy Laboratory ETSIM, Madrid). According to these datings, the age of the Cuesta de la Bajada site ranges between 250 ky and 450 ky. Considering other dates obtained in nearby river formations, the most likely age of this site is MIS 8 or prior (Santonja et al., submitted). Centro Nacional de Investigación sobre la Evolución Humana. (CENIEH); 09002 Burgos manuel.santonja@cenieh.es
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  528  Manu...
CENTRAL PLATEAU 3. Palaeontology and environment The faunal assemblage recorded to date this site is as follows (Blain, in preparation); Sesé and Soto, in Santonja et al., 2000): – Herpetofauna: Alytes obstetricans; Pelodytes punctatus; Bufo bufo; Bufo calamita; Hyla arborea; Pelophylax perezi; Anura indet.; Lacertidae indet.; Coronella cf. girondica; Vipera sp.; Ophidia indet. – Micromammals: Lagomorpha: Oryctolagus cf. cuniculus; Soricomorpha: Crocidura cf. russula; Rodentia: Eliomys quercinus, Apodemus cf. sylvaticus, Cricetulus (Allocricetus) bursae, Microtus brecciensis, Microtus cf. duodecimcostatus and Arvicola aff. sapidus. – Macromammals: Carnivora: Canis lupus; Proboscidea: Elephas (Palaeoloxodon) antiquus; Perissodactyla: Stephanorhinus cf. hemitoechus and Equus chosaricus; Artiodactyla: Cervus elaphus, Bos primigenius, Rupicapra rupicapra and Capra sp. Almost all these species – rodents, elephant, rhinos, bovids, canids and caprids – are represented by isolated remains. Cervus is most frequent, but Equus is by far the most abundant genus. From a palaeoecological perspective, there are different palaeoenvironments. Among themselves: Arvicola aff. sapidus inhabited the proximity of watercourses with riparian vegetation; Microtus inhabited moist and wet soils. All the species indicate developed vegetation, and most are typical of open landscapes, although Eliomys quercinus and Apodemus cf. sylvaticus are not abundant, suggest the existence of wooded areas nearby. 4. Lithic industry and human activity The raw materials used at Cuesta de la Bajada was flint in some cases, more commonly silicified limestone, present in the Alfambra River’s alluvium, and also quartz and quartzite, which must have been brought from the Guadalaviar River basin, 2 km away. Local ordinary limestone was also used in an almost natural state, in some cases as anvils or hammers. The lithic industry in the two sectors of CB was identical, both technologically and typologically. Figure 1. Stratigraphic section of the Alfambra River terrace T4 and details of the sequence G, CB3, CB2, CB1 and P at the Cuesta de la Bajada site. Key: (1) Mud. (2) Clay and silt. (3) Sand and granules. (4) Gravel. (5) Oxidation level. (6) Carbonate concretions. (7) Colluvial and soil. (8) Cross-stratification. (9) Stratification contact. (10) Covered. In SO, the worked material, dragged along alluvial environments, evidences a degree of size-based selection, and some sections of the operational chains are missing or infra represented. By contrast, in SE, the knapping process is fully represented: hammerstones, raw material which was only tested, cores, all phases of flake production, debris and retouched tools. The cores were managed relatively frequently by means of orderly removal systems –polyhedral, Quina, discoidal and sometimes, Levallois–. In some cases, however, there is evidence of random and low intensity exploitation, adapted to the size and shape of the blank. 529
CENTRAL PLATEAU  3. Palaeontology and environment The faunal assemblage recorded to date this site is as follows  Blain, i...
530 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD None of the assemblages contained any macrotools (handaxes, cleavers, large retouched flakes or trihedral pics). The retouched toolkit matches Mousterian patterns, with a large percentage of sidescrapers and denticulates (Fig. 2). In the light of the dated chronologies, the CB industry should be framed in the European context of the early Middle Palaeolithic. Human activity took place in the general framework of the Alfambra floodplain, either on one of the small channels that ran through it or in relation to the ponds in the shallow depressions caused by surface deformation due to subsidence in the area. In SE, green fractures and numerous cut marks were detected, particularly on long bones of horses and in some cases deers, indicative of systematic human intervention on these animals. Acknowledgements To all our colleagues who participated in the research and excavation of Cuesta de la Bajada: Carmen Sesé and Enrique Soto –National Museum of Natural Sciences–, H. Blain –IPHES, Catalonian Institute of Human Palaeoecology and Social Evolution– (Palaeontology); Manuel Domínguez Rodrigo, José Yravedra and Rebeca Barba Egido –Prehistory Dep., Complutense University, Madrid– (Taphonomy); Blanca Ruiz Zapata –University of Alcalá de Henares– (Palynology); Ch. Falguères –Institute of Human Palaeontology, Paris–, Lee Arnold and Mathieu Duval –CENIEH–, T. de Torres and José E. Ortiz (Biomolecular Stratigraphy Laboratory, School of Mines and Energy, Complutense University, Madrid), N. Mercier – Centre de Recherche en Physique Appliquée à l’ Archéologie, Bordeaux University– (Geochronology); Giacomo Gillani, Eduardo Méndez, Joaquín Panera, Susana Rubio, Juan R. de Tembleque and Borja Sánchez-Cervera (Excavation and Archaeology); Raquel Rojas Mendoza (Lithic drawings and Conservation treatments). Figure 2. Cuesta de la Bajada lithic industry. Sidescrapers (1-10); endscrapers (11-13); flakes from tool retouching (14-15); Levallois flakes (16-17); Levallois cores (18-20).
530  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  None...
CENTRAL PLATEAU Jesús F. Jordá Pardo*, Marta Navazo Ruiz**, J. Carlos Diez FernándezLomana** Jarama VI (Valdesotos, Guadalajara, Castilla – La Mancha) Introduction Work in the Upper Jarama Valley (Valdesotos, Guadalajara, Castilla-La Mancha) began in 1983, when a team led by Dr. Francisco Jordá Cerda discovered the Jarama I and Jarama II sites. In 1985 and 1988, systematic excavations were conducted at Jarama II as part of the Prehistoric Research in the Upper Jarama Valley (Valdesotos, Guadalajara) project, accompanied by archaeological surveys in the Jarama canyon in 1988 which located a new archaeological cave site, Jarama VI, excavated between 1989 and 1993 under the same project. Jarama II yielded lithic and bone industries attributed to the Magdalenian along with a sperm whale ivory statuette depicting a glutton and also several Chalcolithic burials (Adán et al., 1995). In 1992, an emergency dig at Jarama I permitted the recovery of a lithic assemblage attributable to the Magdalenian (Adán et al., 1995). Stratigraphy and chronology Jarama VI, 822 m asl, is a Rock Shelter 23 m above the left bank of the Jarama River, located amongst the Cretaceous karstified dolomites on the S boundary of the Spanish Central Range. It faces N and is partially infilled by sediment. Its lithostratigraphic sequence consists of several units resting on the bedrock (Adán et al., 1995, Jordá Pardo 2007) (Fig. 1). The lowest unit (JVI.3), a cryoclastic deposit containing lithic and bone remains, is well preserved throughout its area and sealed by an intermediate unit, JVI.2, which has an erosive contact and is structured into three sub-units consisting of sands and luttites of obviously fluvial origin. The lower sub-unit (JVI.2.3) is sterile, alternating between sands and luttites deposited by floods. The intermediate silty sub-unit (JVI.2.2) is a floodplain facies containing archaeo* ** logical remains, scattered locally around a small hearth. The top sub-unit of this unit (JVI.2.1) consists of clayey sands with interspersed clasts and abundant archaeological material. The upper unit (JVI.1), separated from JVI.2 by a diastem, consists of gelifracts and reddish sands and silts due to sheet-flood sedimentation, with quartzite, slate and quartz pebbles of anthropogenic origin at the base. Less than 4 m2 of this unit is preserved. It contains abundant archaeological material. The sequence culminates with breccia and speleothems (JVI.K). In 1992, three conventional 14C dates situated the lower units in a range between 41 and 30 ky cal BP (Beta-56639 and Beta-56638), with the upper unit between 29.6 and 26.8 ky cal BP (Jordá Pardo 2001, 2007), although there are doubts about the representativity of the latter (Beta-56640) (Fig. 1). In 2008, AMS 14C dates with ultrafiltration was performed in collaboration with the Oxford Radiocarbon Accelerator Unit on samples of bone fragments bearing anthropogenic marks. Only 3 of the 30 analyzed bones could be dated, two of which (OxAX-2290-56 and OXA-21714) revealed infinite ages beyond 50.2 for unit JVI.1 and 47 ky BP for unit JVI.2, while the third date (OxA-X-2310-22) for subunit JVI .2.2 was 49.4 ± 3.7 ky BP (Wood et al., 2012). At the end of 2010, a fieldwork was conducted at Jarama VI along with researchers from the Neanderthal Museum and University of Cologne in order to collect several samples, some of them for luminescence dating (IRSL). The results are consistent with the ORAU datings, situating the occupations at beyond 50 ky BP (Kehl et al., 2013). Lithic industry The raw material employed here was primarily quartz and quartzite, followed by flint and rock Laboratorio de Estudios Paleolíticos, Departamento de Prehistoria y Arqueología, Universidad Nacional de Educación a Distancia. Ciudad Universitaria. Paseo Senda del Rey 7, E-28040 Madrid, Spain; jjorda@geo.uned.es Área de Prehistoria, Departamento de Ciencias Históricas y Geografía, Universidad de Burgos. Plaza Misael Bañuelos s/n, E-09001 Burgos, Spain; mnavazo@ubu.es; clomana@ubu.es 531
CENTRAL PLATEAU  Jes  s F. Jord   Pardo , Marta Navazo Ruiz  , J. Carlos Diez Fern  ndezLomana    Jarama VI  Valdesotos, G...
532 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Figure 1. Jarama VI. Plan of the rock shelter, showing the excavated area, longitudinal and transversal stratigraphic sections and dating table (illustrations and production, J.F. Jordá). crystal. The former were brought to the cave as cobbles, while the flint arrived already shaped in the form of flakes and retouched flakes. Quartz, quartzite, sandstone, shale and limestone pebbles were also found in the cave but not considered to be part of the technological assemblage. Differences in the operational chains were detected, depending on the type of raw material. The quartz and quartzite cores were in the early stages of reduction, the vast majority on pebble. Only three flint items were found in the whole sequence, all exhausted cores on flake. Discoidal reduction is the most common system. The aim of the reduction was to produce small and micro-sized material (flakes and retouched items). Butts and dorsal faces have cortex, indicating that they were in the early stages of reduction. The flint flakes and retouched flakes have the same size but lack cortex. The best represented types are denticulates and notches in JVI.1 and JVI.3, and sidescrapers in JVI.2. The material includes allochthonous flint sidescrapers whose edges were sharpened at the campsite, while the local materials were used to manufacture denticulates and notches. Other documented types include endscrapers, becs, burins and points (Fig. 2). The technological characteristics of Jarama VI suggest expeditious knapping of local materials, indicating these groups’ adaptability to local resources. On the other hand, flint was an allochthonous material and was therefore part of a more complex mobility strategy. This material was fully exploited, and the items were already knapped when they arrived at the site. We interpret the presence of pebbles as a
532  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Figu...
CENTRAL PLATEAU Fauna Unit JVI.2 unit has provided a near-complete adult human first metatarsal bone (Fig. 2). Although the species is difficult to diagnose, its context, morphology and dimensions permit its attribution to Homo neanderthalensis. The surface shows clear signs of action by a carnivore (depressions and grooves) (Lorenzo et al., 2012). Other more modern human remains (Homo sapiens) have been identified on the surface of the site. Skeletal remains from microvertebrates (Pliomys cf. lenki, Microtus arvalis, Microtus agrestis, Apodemus sylvaticus and Apodemus flavicollis) are abundant in all units at Jarama VI. Beaver, rabbit, several species of birds and some amphibians are also present. None of the above taxa show evidence of hunting or human consumption. Herbivores predominate over carnivores in all units, although the minimum number of individuals is small in all cases. Deer, goats, large bovids and horses, generally adults, are predominant. Analysis of the bone surfaces has found indications of rapid burial with minimal pre– or post-sedimentary alterations. We have found little evidence of exposure to the air, but somewhat more to roots and trampling. Several zones contain bones with signs of abrasion, dissolution and crusting. Moisture is evidenced by the presence of manganese on bones as well as cracks caused by environmental changes. Figure 2. Jarama VI : Human metatarsal (photos and montage, C. Lorenzo) and stone tools from units JVI.1 (a, flake; b and c, retouched flakes), JVI.2 (a, point, b and c convergent sidescrapers) and JVI.3 (3a, c, e and f, flakes, b point, c sidescraper) (illustrations B. Márquez, flake montage, M. Navazo and A. Benito). reserve of raw material, which in strategic terms suggests that there was prior planning for the requirements of these groups, a strategy that was repeated in the different occupation phases of this shelter. Long bones, the main category, bear signs of fresh breakage, epiphysis removal and abundant percussion and cut marks. Incisions suggest that defleshing was the main activity. Very few bones were burned. Several were used as retouchers. Remains bearing carnivore bite marks (probably from wolves and foxes) occur in small numbers. Overlapping marks show that these animals always had access to the prey after the hominids as marginal scavengers, without prolonged habitation in the cave. They left a coprolite and half a dozen digested bones. Conclusion The three archaeosedimentary units at Jarama VI seem to reflect short but repeated occupations by mobile groups of Neanderthals who used the cave in activities associated with the consumption of medium-sized herbivores obtained in the local environs in a temperate climate during OIS 3. 533
CENTRAL PLATEAU  Fauna Unit JVI.2 unit has provided a near-complete adult human    rst metatarsal bone  Fig. 2 . Although ...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD 534 Carbonell, E.a,b,c; Huguet, R.b,a,c,*; Cáceres, I.a,b; Lorenzo, C.a,b,d; Mosquera, M.a,b; Ollé, A.b,a; Rodríguez, X.P.a,b; Saladié, P.b,a ; Vergès, J.M.b,a; García-Medrano, P.b; Rosell, J.a,b; Vallverdú, J.b,a,c; Carretero, J.M.e,d; Navazo, M.f,g; Ortega, A.I. g,h; Martinón-Torres, M.g; Morales, J.I.b,a; Allué; E.b,a; Aramburu, A.i; Canals, A.a,b,n, Carrancho,A.f; Castilla, M.e; Expósito, I.b,a; Fontanals, M.b,a; Francés, M.e; Galindo-Pellicena, M.d,j; García-Antón, D.a,b; García, N.d,j; Gracia, A.d,k; García, R.e; GómezMerino, G.b,a; Iriarte, E.e; LomberaHermida, A.b,a; López-Polín, L.b,a; Lozano, M.b,a; Made van der, J.l; Martínez, I.d,k ; Mateos, A.g; PérezRomero, A.e; Poza, E. d,j; Quam, R.m,d; Rodriguez-Hidalgo, A.b,a,n; Rodríguez, J. g Rodríguez, L.e; Santos, E.e,d; Terradillos, M.k; Bermúdez de Castro, J.M.g; Arsuaga, J.L.d,j Sierra de Atapuerca archaeological sites Introduction Sierra de Atapuerca (Burgos, Spain) is a midaltitude karst range characterised by the subterranean morphology concentrated on its southwestern flank (San Vincente Hill, 1085 m asl). This multilevel karst system, an inactive legacy of old base levels formed during the Plio-Pleistocene, is linked to palaeo-upwelling of the Pico River. It consists of three main levels of sub-horizontal a b ducts which are interconnected by spaces and sinkholes, now hanging +90 m, +70 m and +60 m above the present bed of the Arlanzón River (Ortega et al., 2013, 2014). Only 4.7 km of the accessible ducts in this system are known at present. Around 50 completely infilled cavities have been identified (Ortega, 2009), some of which became exposed when a cutting for a mine railway line between Monterrubio de la Demand and Àrea de Prehistòria, Universitat Rovira i Virgili (URV), Avinguda de Catalunya 35, 43002 Tarragona, Spain. Institut Català de Paleoecologia Humana i Evolució Social (IPHES), C/ Marcel.lí Domingo s/n e Campus Sescelades URV (Edifici W3), 43007 Tarragona, Spain. c. Unidad asociada al CSIC. Departamento de Paleobiología, Museo Nacional de Ciencias Naturales. Calle José Gutierrez Abascal, 2. 28006 Madrid, Spain. d Centro Mixto UCM-ISCIII de Investigación sobre Evolución y Comportamiento Humanos, c/Monforte de Lemos, 5, 28029 Madrid, Spain. e Laboratorio de Evolución Humana (LEH), Dpto. de Ciencias Históricas y Geografía, Universidad de Burgos, Edificio I+D+i, Plaza Misael Bañuelos s/n, 09001 Burgos, Spain. f Área de Prehistoria. Dpto. de Ciencias Históricas y Geografía, Universidad de Burgos, Edificio I+D+i, Plaza Misael Bañuelos s/n, 09001 Burgos, Spain. g Centro Nacional de Investigación sobre la Evolución Humana (CENIEH), Paseo Sierra de Atapuerca, 09002 Burgos, Spain h Grupo Espeleológico Edelweiss, Excma, Diputación Provincial de Burgos, Paseo del Espolón s/n, 09071 Burgos, Spain i Departamento de Mineralogía y Petrología, Facultad de Ciencia y Tecnología, Universidad del País Vasco/EHU, c/ Sarriena, s/n, 48940 Leioa, Spain j Departamento de Paleontología, Universidad Complutense de Madrid, Avenida Complutense s/n, 28040 Madrid, Spain. k Área de Paleontología, Departamento de Geología, Universidad de Alcalá de Henares, 28871 Alcalá de Henares, Spain. l Museo Nacional de Ciencias Naturales (MNCN). Calle José Gutierrez Abascal, 2. 28006 Madrid, Spain. m. Department of Anthropology, Binghamton University (SUNY), Binghamton, NY 13902-6000, USA. n. Equipo Primeros Pobladores de Extremadura, Casa de la Cultura Rodríguez Moñino. Avda. Cervantes s/n, 10003 Cáceres, Spain * Corresponding autor: Institut Català de Paleoecologia Humana i Evolució Social (IPHES), C/ Marcel.lí Domingo s/n e Campus Sescelades URV (Edifici W3), 43007 Tarragona, Spain. E-mail address: rhuguet@iphes.cat
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  534  Carb...
CENTRAL PLATEAU Figure 1. Location of Sierra de Atapuerca sites. Karst map based on original topography by Edelweiss Speleological Group, adapted from Ortega (2009). Green: top level of karst, purple: middle level, pink: bottom level. Villafría was dug in the late 19th century (Ortega et al., 2012). Other cavities remained hidden, although in the course of hundreds of thousands of years, they have been visited by successive settlers in the Sierra de Atapuerca area. Apart from the sites within the karst system, open air campsites with evidence of activity by human groups have been recorded on the slopes and moors around this low mountain range. Sierra de Atapuerca and its occupations are one of Europe’s most important sources of ancient human fossils. They were declared a World Heritage Site by UNESCO in 2000. In this chapter, we will review the research that has been underway at the Sierra de Atapuerca sites for more than thirty years. The excavations and subsequent analysis of several sites, both caves and open air campsites, have found evidence of occupations by hominins groups in different periods, from 1.3 million BP to less than 3,000 years ago. Digs at Sierra de Atapuerca have focused on four different sectors: Trinchera del Ferrocarril, Cueva Mayor, Cueva del Mirador and the open air karst zone. Listed in chronological order of human occupation, the cave sites are Sima del Elefante, Gran Dolina, Galería, Sima de los Huesos, Portalón and Mirador. The outdoor sites include Hotel California, Hundidero, Fuente Mudarra and Valle de las Orquídeas (Fig. 1). 535
CENTRAL PLATEAU  Figure 1. Location of Sierra de Atapuerca sites. Karst map based on original topography by Edelweiss Spel...
536 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Trinchera del Ferrocarril As its name suggests, the Trinchera del Ferrocarril (Railway Trench) is an artificial trench dug during the construction of a mine railway. In plan, the trench is a 500 metre long arc running N-S through the southern part of the Sierra. The primarily limestone walls of the cutting, no more than 20 metres high, contain sectioned cavities which had been filled with sediment of different origin. Three of these cavities –from pit to determine the characteristics of the infill, its archaeological potential and a rough chronology for the sedimentary deposits. Systematic excavation of the site began in 1996, and has continued uninterrupted down to the present day across a 32 m2 excavation area. The stratigraphic succession at Sima del Elefante is 15 m wide, with a 25 metre thick and with high degree of heterogeneity due to lateral and vertical lithological changes. The sedimentary deposit is divided into 21 units, grouped in turn into three sedimentary phases. Phase I is the lowest in the sequence, from TE7 to TE14. Phase II contains units TE15 to TE19, inclusive. Finally, the most recent Phase III comprises Units TE20 and TE21 (Rosas et al., 2001, 2006) (Fig. 2). Palaeomagnetic analysis has detected polarity changes at the basis of unit TE17. Sediments below this unit from TE16 to TE7 have reversed polarity and have been assigned to the Matuyama subchron (> 780 ky) (Parés et al., 2006). This is consistent with the results from the analysis of the U/Th uranium series of a stalagmite sample in the TE16-TE17 contact area, which shows a chronology of more than 400,000 years. Analysis of cosmogenic cores shows that the age of sublevel TE9c in Sima del Elefante is 1.22 ± 0.16 Myr (Carbonell et al., 2008). These dates are consistent with biochronological data (Rofes and Cuenca-Bescós, 2006; Cuenca-Bescós and García 2007; García et al., 2008). Figure 2. Stratigraphic section of Sima del Elefante. Asterisk marks position of Matuyama-Brunhes inversion. Height in metres from Trinchera del Ferrocarril floor. Synthetic stratigraphy shows location of U-Th and cosmogenic nuclide datings (Rosas et al., 2006; Carbonell et al., 2008). south to north Sima del Elefante, Gran Dolina and Galería– have been defined as archaeological sites. Sima del Elefante The Sima del Elefante site is the southernmost cave in the Trinchera del Ferrocarril. The first archaeo-palaeontological work here was in 1986 under Prof. Emiliano Aguirre, consisting of a test On the basis of biochronological material, more recent units containing archaeo-palaeontological records of the site (TE18-TE19) have been attributed to the second half of the Middle Pleistocene, around 250-350 ky (OIS 9-8) (Rosas et al., 2006, Lopez-García et al., 2011). However, uranium series (U/Th) analysis of a stalagmitic crust from the roof of level TE18 has yielded two datings, 307 ± 19 ky and 255 ± 12 ky (Bischoff pers. comm.). These results suggest that level TE18 was formed during OIS 9 and 7, and that the chronology of TE19 is more recent than 255,000 ky. Lower levels have yielded a rich faunal association including small animals such as birds, lagomorphs and beavers, as well as medium and large sized animals (Sánchez Marco, 2004; CuencaBescós and García, 2007; García et al., 2008; Van der Made, 2013) (Tab.1). The climatic and environmental reconstruction on the basis of faunal
536  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Trin...
CENTRAL PLATEAU analysis indicates that the landscape around Sima del Elefante through the lower sequence (Lower Pleistocene) included open habitats dominated by moist, wooded areas, large areas with permanent water (Rosas et al., 2006, Blain et al., 2010 ;). In the upper units, equids remains are predominant, although remains of other herbivores and carnivores have also been found (Rosas et al., 2001; Van der Made, et al., 2003, 2013. Cuenca-García and Bescós, 2007) (Tab.1). For these units, the suggested landscape is a moist forest with open spaces and possibly drier and colder conditions than the Lower Pleistocene units (Rosas et al., 2006; López-García et al., 2011). Evidence of human activity has been documented in the Middle Pleistocene units and also in the oldest units of Sima del Elefante. By 2013, 127 lithic artefacts had been located (Fig. 3). To date, 86 stone artefacts have been recovered from the Lower Pleistocene Phase I units (TE7-TE14). The main raw material is chert (72.1%) of both Cretaceous and Neogene origin. Three quartz objects and some artefacts in Cretaceous limestone have also been found. All the raw materials could have been found within a 2 km radius of the site. The most represented categories are related to knapped products (flakes and flake fragments). However, there is a significant percentatge (34.9%) indeterminable objects. due to the poor preservation of the Neogene chert. Only four cores (one of them a core fragment) have been found amongst the Lower Pleistocene assemblage. These cores bear evidence of short knapping sequences, based on longitudinal removals. The knapped products have different morphologies but are generally small, averaging 32 x 30 x 9 mm for complete flakes. We have also found a few knapping products with centripetal removals. Retouched tools have only been found in units TE13 (n=3) and TE14 (n=1). These retouched flakes are slightly larger than the average non-retouched items, and have been classified as sidercrapers (n=2) and notches (n=2). In 2013, a chert flake was unearthed in unit TE8, the oldest evidence of human activity found to date in Atapuerca. In the Lower Pleistocene units, no pebble tools, handaxes or cleavers have been found. This assemblage has been assigned to Mode 1 (Carbonell et al., 2008; Ollé et al., 2013). Some of the faunal remains (0.6%) from these lower units (TE7-TE14) bear signs of anthropogenic modifi- Figure 3. Archaeo-palaentological material from Sima del Elefante. 1: Cleaver-like tool (Unit TE18, Sandstone). 2: Sandstone handaxe (Unit TEsup), 3: Quartzite point (TEsup), 4: Retouched flake (sidescraper) of Neogene chert (TE19), 5: Retouched flake (sidescraper) of Cretaceous chert (TE13), 6: Cretaceous chert core (TE9c), 7: Cretaceous chert flake (TE9c), 8: Cretaceous chert flake (knapping debris) (TE9c), 9: Homo sp. mandible (TE9c), 10: Macromammal bone with evidence of fresh fracture (arrows) (TE9c), 11: Left, cut marked bovid vertebra (TE9c); right, electron microscope detail of cut marked bovid mandible (TE9c)(Photos: A. Ollé/J.Mestre/R.Huguet/IPHES). cations (cut marks and breakage). Most of the remains with cut marks are from ungulates, specifically deer and bison. We have also identified human activity on small animals such as birds, rabbits and turtles (Blasco et al., 2011, Huguet, 2013). These marks are found in all anatomical areas, from the appendicular skeleton through the axial skeleton to the skull. Some ungulate long bones were fractured by hominins in order to access nutritional resources inside the bone. The distribu- 537
CENTRAL PLATEAU  analysis indicates that the landscape around Sima del Elefante through the lower sequence  Lower Pleistoc...
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD x x x x ssp. x x x ssp. 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 x x x x x x x x x x x x x x x x cf. x x sp. sp. x cf. 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 x x x x cf. x cf. x x x sp. x x cf. x x x x cf. x x x x x cf. x x x x x x cf. cf. sp. cf. x cf. x x x cf. x x x x x x x sp. x x x x x x x sp. cf. x x x x x x x x x x x x x x x x x x x x x x x x sp. x x sp. x x sp. x x x x x x x x x x x x x x sp. ssp. x x x x x x x TE13 x x TE12 x x x x x cf. TE14 TD6.3 TD6.2 TD6.1 TD7 TD8 SH TD10.3 TD10.2 TD10.1 GIIa GIIb sp. sp. sp. x x TD5 TD34-5? TD3TD4 x x x x x x x x x x x x x x x x TE7 x x TE8 x x x x x TE9 x x x x TE10 x x TE11 x x GIII Equus cf. hydruntinus Capreolus priscus Stephanorinus cf. hemioechus Bison sp. (small) Cervus elaphus priscus Equus ferus Dama dama clactoniana Hemitragus bonali Megaloceros solilhacus sspp. Bison schoetensacki Praeovibos cf. priscus Mammuthus sp. Cervidae indet. Cervus elpahus cf. acoronatus Eucladoceros giulii Hippopotamus Stephanorhinus etruscus Sus scrofa Bison cf. voigtstedtensis Macaca Equus altidens Dama vallonnetensis Mustelidae indet. Martes martes Mestela putorius Muestela nivalis Canidae indet. Cuon alpinus europaes Canis lupus Felis sylvestris Panthera sp. Panthera leo Vulpes vulpes Ursus sp. Hyaena sp. Ursus deningeri Lynx pardinus spelaeus Meles meles Homoterium sp. Canis mosbachensis Vulpes praeglacialis Crocuta crocuta Panthera gombaszoegensis Mustela palerminea cf. Baranogale antiqua Lynx cf. issiodorensis Lynx sp. Pannonictis cf. nestii Canis sp.(arnensis/mosbachensis) Vulpes cf.V. alopecoides Ursus dolinensis TE19 538 Table 1. Stratigraphic distribution of carnivores, ungulates, sub-ungulates and primates at Atapuerca Pleistocene sites (from Rodríguez et al., 2011; van der Made, 2013).
PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  x  x  x x...
CENTRAL PLATEAU tion and location of the cut marks and fractures indicate that human groups had different activities in the butchering chain. The identification of these activities shows that hominins had primary access to some of the animals that they consumed. The bird, rabbit and carnivore remains found in anatomic connection in the lower units of Sima del Elefante indicate an excellent state of conservation of the fossils found in this cave. However, macromammal remains are scarce and fragmentary. If hominins had processed their prey inside the cave, we would expect to find a large number of anatomical elements of their prey, but the anatomical representation of animals indicates otherwise. The faunal remains thus suggest that most of the anatomical assemblage is the result of low intensity occupations, possibly located near the cave entrance (Huguet et al., 2007). The units in Phase I of the sedimentary infill generally have similar dynamics to the anthropic record recovered to date. However, special mention must be made of level TE9c (1.22 ± 0.16 Ma). At this level, in addition to indirect evidence of human presence, three hominin fossils were found: a mandible, a phalanx and a humerus fragment. These remains were provisionally attributed to H. antecessor (Carbonell et al., 2008, Bermúdez de Castro et al., 2010a), however after a comparative morphological analysis of the mandible, Bermúdez de Castro et al., (2011) concluded that its attribution to any known taxon is unclear, and thus suggested that it should be referred to as Homo sp. Along with these human remains, faunal remains with signs of anthropic intervention and 33 stone objects were found. Units TE18-TE19 at this site (Middle Pleistocene) have yielded 41 artefacts, the majority from unit TE19 (n=36). Only one stone tool has been found on TE18, along with four objects whose exact origin is unknown, as they were taken from the stratigraphic section in this part of the sequence. Middle Pleistocene material includes a considerable use of sandstone (39.3%) followed by quartzite (34.1%) and Neogene chert (24.4%). Knapped products, retouched flakes and hammerstones predominate in this lithic assemblage, along with four cores. The predominant knapping strategies are unifacial and unidirectional, but there is also evidence of centripetal knapping in some products. To date no evidence has been found on these levels of the use of knapping methods involving a predetermined morphology of the products, such as Levallois or discoidal, for example. Most of the configured tools were found in unit TE19. The largest proportion of morphotypes are sidescrapers and denticulates. There is also a large unifacial pebble tool from TE18, with a similar morphology to a cleaver. A sandstone handaxe was found on TE Sup, as well as a point and a cleaver, both knapped with quartzite. The lithic industry found on the Middle Pleistocene levels has been tentatively ascribed to Mode 2 (Acheulean). The faunal remains found in these upper units have not been analyzed in depth due to the poor state of the material, although a preliminary list of fauna has been drafted (Table 1). However, Rosas et al., (2004) have presented several working hypotheses concerning the origin of the fossil assemblage in unit TE19, including the possibility that this unit acted as a trap for animals attracted by water or fresh grass. The presence of the taxon Ursidae might be related to the use of the cave for hibernation. Gran Dolina The Gran Dolina site (TD) is an 18 metre thick cave infill. Its stratigraphic succession was initially divided into 11 units, TD1 to TD11 from base to top (Gil and Hoyos, 1987; Parés and Pérez-González, 1999. Pérez-González et al., 2001), which were later revised slightly (Rodríguez et al., 2011) (Fig. 4). The first systematic archaeological excavations were carried out between 1981 and 1989 in a 30 m2 area on level TD10 level. Between 1990 and 1991, work focused on TD3-4, the earliest levels with evidence of human activity. A 9 m2 test pit initiated in 1993 confirmed the presence of palaeontological material at every level of Gran Dolina, except for endo-karstic infills on TD1-2. The results of this test pit led to the start of the horizontal excavation of Gran Dolina from 1996 onwards, which covered an area of more than 95 m2 (Fig. 5). In 2001, excavation began on a series of overhangs in the western part of this site due to the threat of their collapse. This work recovered material from levels TD4 to TD10. At present, level TD10.3 (approx. 90 m2) is being excavated horizontally along with the overhang of TD3-4 (approx. 8 m2). 539
CENTRAL PLATEAU  tion and location of the cut marks and fractures indicate that human groups had different activities in t...
540 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD The Lower Pleistocene record Figure 4. a. Stratigraphic section of Gran Dolina. Asterisk marks position of Matuyama-Brunhes boundary. Height in metres from Trinchera del Ferrocarril floor; b. Synthetic stratigraphy shows location of TL, IRSL and ESR/UTh datings, from Falguères et al., (1999); Berger et al., (2008) and Falguères et al., (2013). Legend: (1) Mesozoic limestone from Gran Dolina roof; (2) speleothem (3) mudstone, clayey silt/terra rossa; (4) bat guano; (5) laminated silty clay; (6) calcilutites and calcarenites; (7) gravel and cobbles and clast flow (8) position of fallen cobbles; (9) principal stratigraphic discontinuity; (10) secondary discordance and silt-sand-clay infill; (11) Matuyama-Brunhes boundary; (12) disappearance of Mimomys savini and first appearance of Iberomys brecciensis; (13) Position of Aurora stratum; c, Palaeolatitude of virtual geomagnetic pole of Gran Dolina stratigraphic section. Each point is a mean Fisher direction of individual samples. The Matuyama-Brunhes boundary is in stratigraphic unit TD7 (Parés and Pérez-González, 1999). Figure modified from Ollé et al., (2013). Over 1,300 faunal remains of herbivores and carnivores (Table 1) have been recovered from unit TD3-4. Amongst the latter, the most frequent is the bear species Ursus dolinensis (García and Arsuaga, 2001). These animals used the cave regularly for hibernation, as evidenced by numerous remains and claw marks on the walls. However, this pit was a natural trap for ungulates which died when they fell in. Predator activity around these herbivore carcasses was uncommon. Some of the faunal remains show cut marks, and anthropogenic fractures suggest that human groups entered the cave to exploit the fallen ungulates and thus had primary access. Carnivore tooth marks have also been documented. Remains of felines (Panthera gombaszoegensis) and small dogs (Canis sp.) suggests that these predators may have been responsible for the tooth marks (Rosell, 1998; Huguet et al., 2013). A small collection of tools, primarily quartzite, has been recovered from this sedimentary deposit. The objects show simple working sequences, essentially reduced to obtain flakes from unipolar strategies and rough configuration of cobble choppers (Carbonell et al., 2001, Rodríguez, 2004). Unit TD5 shows different types of operation. The remains documented from this unit have different origins. On the one hand, carnivores were quite active. One of the most important taxa are hyaenids, which occupied the unit as a den, with documented remains of their prey and also some coprolites. In unit TD5, some of the remains also arrived by gravitational processes (Huguet, 2007; Saladié, 2009). In subunit TD6.3, remains of hyena (Crocuta crocuta) and their prey bearing numerous tooth marks and modifications during digestion permit the inference that these animals used the site as a den (Fernández-Díaz, 2013). In both units (TD5 and TD6.3) occupation by hyaenids alternated with ursids, which used the cave to hibernate. Hominins also occupied the cave, alternating with both carnivores to a lesser extent than at the next level up (Saladié, 2009, Fernández-Díaz, 2013). Subunit TD6.2 is the result of an anthropogenic assemblage where a large collection of archaeo-palaeontological remains was found, including more than 180 hominin remains. Thermoluminescence and simulated infrared luminescence dating for this assemblage is 960 ± 120
540  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  The ...
CENTRAL PLATEAU Figure 5. Horizontal excavation of level TD10, 2007 dig (Photos: J. Mestre/IPHES). ky (Berger et al., 2008). Palaeoenvironmental and palaeoclimatic studies indicate that overall, unit TD6 corresponded to a period with an interglacial climate, holartic vegetation and abundant resources (García Antón, 1998; CuencaBescós et al., 1999. Burjachs, 2002; Rodríguez et al., 2011). The human remains found in unit TD6 at Gran Dolina were assigned to a newly described species, Homo antecessor, which was proposed as the last common ancestor of modern humans and Neanderthals (Bermúdez de Castro et al., 1997). During the last decade, the TD6 human hypodigm has increased, permitting advances in the taxonomic and phylogenetic characterization of this species. H. antecessor has a number of features suggesting its “modernity”, such as a cranial capacity of more than 1,000 cc, a modern tooth growth pattern (Bermúdez de Castro et al., 2010b) and a modern face in both its aspect and its growth forms (Bermúdez de Castro et al., 1997, Lacroix et al., 2013.). The TD6 hominins also shared some features of the postcranial skel- eton with Homo sapiens (Carretero et al., 1999). Several features which were hitherto considered typical and unique to the Neanderthal lineage have been identified in the TD6 human remains, including humerus (Bermudez de Castro et al., 2012) and, more particularly, teething (MartinónTorres et al., 2006, 2007; Gomez-Robles et al., 2007, Martinón-Torres et al., 2007). Recent studies suggest that H. antecessor may be a European lineage of Asian origin, close to the divergence point between H. sapiens and H. neanderthalensis (Martinón-Torres et al., 2007, 2011, , Bermúdez de Castro and Martinón-Torres 2013). Zooarchaeological analyses have detected frequent cut marks and anthropogenic bone breakage on the remains found at this level (Fig. 6), indicating that the hominins who occupied TD6.2 actively accessed the prey that they brought to Gran Dolina (Saladié et al., 2011). Remains of H. antecessor are amongst their prey. This is the oldest case of anthropogenic cannibalism known to date (Fernández-Jalvo et al., 1996), which current stratigraphic evidence suggests took place over a 541
CENTRAL PLATEAU  Figure 5. Horizontal excavation of level TD10, 2007 dig  Photos  J. Mestre IPHES .  ky  Berger et al., 20...
542 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD long time sequence (Carbonell et al., 2010). Anthropogenic modifications to H. antecessor and deer on level TD6.2 suggest that the butchering process was the same for both taxa, and that the remains were also discarded on the floor of the habitat in the same way. In this context, the consumption of infants and immature individuals was common. This age profile is similar to the one associated with episodes of intergroup aggression in chimpanzees. These parallels permit cannibalism in TD6 to be linked to low-risk attacks on individuals, possibly in order to defend and expand the resource provisioning territory against other neighbouring groups (Saladié et al., 2012). The H. antecessor remains also show certain peculiarities, such as the lack of activity by carnivores and better preserved axial and brittle bones than the other animals in the assemblage. This feature seems related to the different types of occupation that took place during the formation of TD6.2 and the episodes of cannibalism. This investigation suggests that the Lower Pleistocene hominins were at the top of the food chain and were able to control the competition that might arise from other groups of congeners or other predators in the same ecological niche (Saladié et al., 2014). In this regard, Huguet et al., (2013) propose that the groups of hominins which inhabited Sierra de Atapuerca during the Lower Pleistocene had a high degree of control over their territory and its resources. Evidence of the stone tools in this assemblage is much more representative numerically than the above-mentioned units (Fig. 7) (Carbonell et al., 1999, Rodríguez, 2004, Ollé et al., 2013.). For the first time, there is a full range of suitable rocks for working in the Atapuerca area. Chert -both Neogene and Cretaceous- is the predominant material, followed by quartzite, quartzarenite, sandstone, quartz and limestone. A degree of planning can be observed in the way these resources were managed, with all stages of the lithic production chains present (hammerstones, cores, flakes, retouched flakes and knapping debris). The reduction strategies are varied (longitudinal unipolar, centripetal and occasionally bipolar on anvil), and all seem to be aimed at the systematic production of small and medium format artefacts. This is the first point in the Atapuerca sequence where retouch is used systematically to make tools in the form of denticulates, notches and, to a much lesser extent, sidescrapers. To date, the only evidence of large-format tools is a single sandstone chopper. Level TD7 marks a major shift in the dynamics of the cave. This level consists of limestone gravel, giving way laterally to silt. The action of water thus seems to mark the overall dynamics of the cave in this period. This level has only yielded remains of Stephanorhinus etruscus and Praeovibos in anatomical connection, suggesting that the natural trap in the roof of the cave was reactivated (Rosell and Blasco, 2009). The only lithic item recovered is a small quartz flake. Figure 6. a) Homo antecessor jaw with percussion stigma. b) H. antecessor proximal phalanx with cut marks. c) H. antecessor rib processed during corpse defleshing. d) Striations on H. antecessor tibia fragment produced while Achilles heel was being cut. e) Equine phalanx with cut marks produced during skinning. f) Deer radius with removals during fracturing to access bone marrow (Photos: P. Saladié/ IPHES). The Middle Pleistocene record In the Gran Dolina sedimentary succession, the Matuyama-Brunhes palaeomagnetic boundary is located at the top of unit TD7 (Parés and PérezGonzález, 1995, 1999). The Middle Pleistocene archaeo-palanteological fossil record in Gran Dolina therefore consists of units TD8, TD8-9, TD9 and TD10. Each of these units has specific features which show that the cave was sometimes used by
542  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  long...
CENTRAL PLATEAU carnivores as a den and in others, as a base for Preneandertal occupations. The palaeomagnetic data combined with ESR and uranium series situate unit TD8 at the beginning of the Middle Pleistocene, circa 600 ky BP (Falguères et al., 1999, 2013; Parés and PérezGonzález, 1999). This unit contains a large, diverse range of ungulates and carnivores, predominantly fallow deer (Dama vallonnetensis) and occasional remains of carnivores (Table 1). The sample of fallow deer remains is characterized by the presence of appendicular and cranial items, with abundant carnivore tooth marks. They bear no signs of anthropogenic action and there is no lithic industry. According to Blasco et al., (2011) these features suggest that the primary agent involved in this accumulation of ungulates were hyaenids. However, several factors do not fully correlate with some of the features traditionally used to define these carnivores’ dens. In TD8 there are no immature carnivores, marks related to the final stages of carnivore consumption (e.g., intensive bone chewing, diaphyseal cylinders or hollowing), there is a low proportion of coprolites, an absence of an attritional mortality profile, and also many whole bones. According to Blasco et al., (2011), this variation in the composition of the assemblage from what might be expected in a den is because the assemblage in TD8 might not be the exclusive result of the cave’s use as a hyaenid den, but rather the product of the combination of several types of occupations, in the course of which it was occasionally accessed by other carnivores as well. Units TD8 and TD8/9 are currently a hiatus in the presence of anthropic activity in Gran Dolina. This is the only section with no evidence of material culture, as four lithic industry items have been found in the unit immediately above (TD9) (Ollé et al., 2013), making this the first unit in Gran Dolina with evidence of human activity in the Middle Pleistocene, dated by TL at 480 ± 130ka. TD10 has the largest accumulation of archaeological remains in the entire Atapuerca complex. This unit is divided into 4 lithostratigraphic sub-units, identified from base to top as TD10.4 to TD10.1. The top two units (TD10.1 and TD10.2) are now fully excavated. Both sub-units have yielded large concentrations of archaeo-palaeontological material, with approximately 120,000 faunal remains and 35,000 lithic Figure 7. Lithic industry from TD4 (a-c) and TD6.2 (d-l). a) quartzite unipolar core; b) quartzite flake; c) quartzite chopper; d, e, f) quartzite flakes; g, h, i) retouched flakes (denticulates) in Cretaceous chert; j) quartzite retouched flake -refit of 2 items-; k) sandstone chopper; l) large Neogene chert core, with refit of 2 flakes (white flecks) (Photos A. Ollé/IPHES). items recovered to date. This density of material shows an intense occupation of Sierra de Atapuerca by Preneandertal groups. The high level of activity by these hominins is confirmed by the different assemblages generated by occupations with a similar chronology at the nearby Galería and Sima de los Huesos sites. Geochronological tests of TD10 to date include a TL dating of 430 ± 59 ky for the top of subunit TD10.3, and 543
CENTRAL PLATEAU  carnivores as a den and in others, as a base for Preneandertal occupations. The palaeomagnetic data combi...
544 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD a series of ESR/UTh datings including two for TD10.2 (418 ± 63 and 337 ± 51 ky), one for the base of TD10.1 (379 ± 57ka) and an average of 337 ± 29 ky for its top. However, a slightly discordant average TL date (244 ± 26 ky) has also been obtained for the lower part of unit TD10.2. The archaeo-stratigraphic sequence ends with an archaeologically sterile unit (TD11) dated between 240 ± 44 ky and 55 ± 14 ky (Falguères et al., 1999. 2013, Berger et al., 2008). The technology of subunits TD10.2 and TD10.1 is characterized by diverse, standardized operating sequences and tool configuration (Fig. 8). Centripetal flake removal methods are predominant, along with some hierarchical cores and a somewhat predetermined size and shape in the products. The presence and degree of configuration of large standardized tools (handaxes and cleavers) is less than the assemblage documented at the Galería site, and instead there is a higher incidence of tools on small flakes which are moreover morphologically diverse and standardized. The rich archaeological level documented at TD10.1 could therefore reflect a local evolution from Mode 2 (Acheulean) to Mode 3 (Mousterian) in Sierra de Atapuerca. Finally, the top section of TD10.1 clearly shows a gradual decrease in the use of Gran Dolina. Technologically, it seems to follow the transitional trends identified in the rest of TD10.1. In fauna, TD10.2 is a clear case of specialized hunting focused on the exploitation of bison (Bison sp.), as approximately 95% of the NISP (Number of Identified Specimens) and MNE (Minimum Number of Elements) correspond to these animals. In contrast, in sub-unit TD10.1 there is a broad spectrum of prey (Blasco et al., 2013, Rodríguez-Hidalgo, in progress), with a predominance of ungulates such as deer and horses (Table 1) and also other animals such as small prey and a few carnivores (Blasco et al., 2010, 2013). The information gathered from the analysis of the sub-units in TD10 excavated in their entirety to date suggests a wide variety of preneandertals’ subsistence strategies at the end of the Acheulean. Galería Figure 8. Lithic industry from TD10. a) centripetal chert core (TD10.2); b) unifacial centripetal core in quartzite (TD10.1); c) large bifacial tool in sandstone (TD10-2); d) small quartzite handaxe (TD10-1); e) quartzite double sidescraper (TD10-1); f) quartzite sidescraper (TD10-1); g) Cretaceous chert sidescraper (TD10-1); h) sandstone denticulate (TD10-1); i) Cretaceous chert denticulate; j) quartzite point (TD10-1) (Photos: A.Ollé/IPHES). The Galería site is 50 m south of Gran Dolina. It is divided into three sections: the central section (TG) linked to the north with a hall known as Covacha de los Zarpazos (TZ) and to the south with a vertical conduct known as Trinchera Norte (TN). The first archaeological works at the Galería site began in 1976. Systematic, excavations were developed from 1982 to 1995. From 2002 until 2010 the works were focused on TZ. Since then, the excavations have focused on sector TG-TN. Nowadays, the excavation area affects to more than 40m2. The sedimentary infill at Galería
544  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  a se...
CENTRAL PLATEAU consists of five lithostratigraphic units, identified from bottom to top as GI to GV (Pérez-González et al., 1995, 2001). Archaeologically sterile Unit GI is the oldest, and consists of facies from the interior. Units GII and GIII are rich in lithic and faunal remains. Units GIV and GV ultimately clogged this cave. While the latter were initially sterile, in the latest work phase GIV has yielded over 100 items including stone tools and faunal remains (Fig. 9). The Matuyama-Brunhes transition has been detected in GI, at the base of Galería (PérezGonzález et al., 1999). TL, IRSL and ESR datings for units GII to GIV (Berger et al., 2008, Falguères et al., 2013) suggest that it was formed between 500 and 250 ky. The speleothem that seals unit GIV in the central sector of TG has been dated at 118 +71/-49 ky and 200 ky by means of uranium-series and ESR respectively (Grün and Aguirre, 1987, Falguères et al., 2013). Galería has provided important evidence of human occupation, with a rich Mode 2 or Acheulean lithic assemblage associated with abundant faunal remains (Ollé et al., 2005, 2013, Cáceres et al., 2010, Cuenca-Bescós et al., 2010, Rodríguez et al., 2011, García-Medrano et al., 2013). Two human fossils, a mandible and a skull fragments have been attributed to H. heidelbergensis (Bermúdez de Castro and Rosas, 1992; Arsuaga et al., 1999a). (Fig. 10). The lithic items in Galería were produced from 7 types of raw material, all found within 2 to 5 km of the site (García-Antón et al., 2002. Figure 9. Galería stratigraphic sequence, showing location of luminescence, ESR and Useries samples. Legend: 1) Upper Cretaceous limestones and dolomites (Galería cave wall); 2) Speleothems; 3) Limestone blocks and cobbles; 4) Main stratigraphic unconformities 5) Lateral facies variations, from clay loam to gravels (left side of figure) and from gravels to breccia (right); 6) Cut and fill; 7) Limit of GII Unit layers; 8) Bat guano level; 9) Luminescence samples (Berger et al., 2008); 10) ESR samples (Falguères et al., 2013); 11) U/Th samples (Bischoff, published in Falguères et al., 2013); 12) U/ Th and ESR samples (Grün and Aguirre, 1987); 13) Matuyama-Bhrunes reversal (Pérez-González et al., 1999); 14) Soil; 15) Allostratigraphic levels; 16) Archaeopalaeontological levels. 545
CENTRAL PLATEAU  consists of    ve lithostratigraphic units, identi   ed from bottom to top as GI to GV  P  rez-Gonz  lez ...
546 PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT: THE CURRENT ARCHAEOLOGICAL RECORD Ollé et al., 2005, 2013, García-Medrano, 2011, García-Medrano et al., 2013, 2014; Terradillos, 2010; Terradillos and Rodríguez, 2012). The best represented exploitation methods are multipolar centripetal and unipolar longitudinal, reflected by products and cores. Also, other strategies such as multipolar orthogonal have been documented. Knapped products are the most common structural catergory. Large cutting tools are well represented, although the small and medium formats are predominant (scrapers, denticulate and tips) (Carbonell et al., 2001). In the earliest levels (GIIa), the large tools are mainly made on quartzite cobbles, while from GIIb onwards, the large tools are made on Neogene chert and sandstone flakes (García-Medrano, 2011, GarcíaMedrano, et al., 2014). The main uses of the tools were for butchering, although work on hides and, to a lesser extent, plant material has also been identified (Márquez et al., 2001. Ollé, 2003). The majority of the faunal remains in Galería are from herbivores, with a major presence of deer, horses and, in smaller numbers, bovines and rhinoceros. Carnivore remains are scarce, (Rodríguez et al., 2011) (Table 1). Galería also has a good representation of micromammals (Cuenca-Bescós et al., 2010) and birds (Sánchez Marco, 1999). Figure 10. Above: General view of renewed Galería excavation (J.Mestre/IPHES). Centre (left to right: quartzite handaxe from TG07; quartzite cleaver from TN2B; quartzite sidescraper from GSU11 (Photo: P. García-Medrano/A. Ollé/IPHES). Below (left to right): Long bone fragment of medium-sized animal from GSU10 with cut marks interrupted by carnivore tooth marks; deer sacrum with carnivore tooth marks from TN6 (Photos: I. Cáceres/IPHES). García-Antón and Mosquera, 2007). Neogene chert is most abundant, followed by quartzite and sandstone. Other material such as Cretaceous chert, limestone, quartz and schist, have a minority presence. The operative chains are highly fragmented and the knapping sequences are mainly allochthonous. The knapping inside the cave was aimed to solve specific requirements, and was highly expeditious. Most of the artefacts were produced outside the cave (Mosquera, 1995, Carbonell et al., 2001, Ollé, 2003; The most part of herbivores anatomical representation are axial and cranial elements, with few remains from the appendicular skeleton. These skeletons are from 219 individuals of all ages, with a slightly greater abundance of immature than adult and senile individuals. These remains show abundant evidence of carnivore intervention (tooth marks) and less human intervention (cut marks and fractures). Nevertheless, the faunal association at Galería does not match the expected pattern of an assemblage originated by hominins or carnivores. The taphonomic features suggest that Galería did not meet the environmental conditions appropriated to the establishment of human occupations, nor was it used as a carnivore den (Cáceres et al., 2010). The origin of this accumulation was the vertical conduct in TN, which acted as a natural trap for herbivores. Fallen animals attracted the attention of carnivores and hominins alike to exploit these meat resources (Díez et al., 1999. Huguet et al., 2001. Cáceres,
546  PLEISTOCENE AND HOLOCENE HUNTER-GATHERERS IN IBERIA AND THE GIBRALTAR STRAIT  THE CURRENT ARCHAEOLOGICAL RECORD  Oll ...