Los Villares locality (Ruidera, Castilla-La Mancha, Spain): a new Middle Pleistocene fossil assemblage from the Southern Iberian Plateau with possible evidence of human activity

7 Cuaternario y Geomorfología ISSN: 0214-1744 ISSNe: 2695-8589 www.rediris.es/CuaternarioyGeomorfologia/ C y G Derechos de reproducción bajo licencia Creative Commons 3.0. Se permite su inclusión en repositorios sin ánimo de lucro. https://doi .org/10.17735/cyg.v36i1-2.90422 Los Villares locality (Ruidera, Castilla-La Mancha, Spain): a new Middle Pleistocene fossil assemblage from the Southern Iberian Plateau with possible evidence of human activity Yacimiento de los Villares (Ruidera, Castilla-La Mancha, España): un nuevo conjunto fósil del Pleistoceno Medio en la Submeseta Sur Ibérica con posible evidencia de actividad antrópica García-Martínez, D.(1), (2), (3), (4)*; Duval, M.(2), (5); Zhao, J.-X.(6); Feng, Y.(6); Wood, R.(7), Huguet, R.(8), (9), (10), Cifuentes-Alcobendas, G. (11), (12); Palancar, C. A.(4); Moya-Maleno, P.R. (3), (13) (1) Physical Anthropology Unit, Department of Biodiversity, Ecology, and Evolution, Faculty of Biological Sciences, Complutense University of Madrid, Madrid, Spain. Email: [email protected] (2) Centro Nacional de Investigación sobre la Evolución Humana (CENIEH), Paseo Sierra de Atapuerca, 3, 09002 Burgos, Spain. (3) Centro de Estudios del Campo de Montiel (CECM), Plaza Mayor s/n, 13328 Almedina, Castilla-La Mancha, Spain. (4) Paleoanthropology Group, Museo Nacional de Ciencias Naturales (CSIC), José Gutiérrez Abascal 2, 28006 Madrid, Spain. (5) Australian Research Centre for Human Evolution (ARCHE), Griffith University, Nathan, QLD 4111, Australia. (6) Radiogenic Isotope Facility, School of Earth and Environmental Sciences, The University of Queensland, Brisbane, QLD 4072, Australia. (7) Research School of Earth Sciences, Australian National University, Building 142 Mills Road, Acton, ACT 2601, Australia. (8) IPHES-CERCA, Insitut Català de Paleocologia Humana i Evolució Social, Zona Educacional, 4, Campus Secelades URV (Edifici W3) E3, 43700 Tarragona, Spain. (9) Departament d’Historia i Historia de l’Art. Universitat Rovira i Virgili (URV). Avinguda de Catalunya, 35, 43002, Tarragona, Spain (10) Unit associated to CSIC. Departamento de Paleobiología, Museo Nacional de Ciencias Naturales, c/ José Gutiérrez Abascal 2, 28006 Madrid, Spain. (11) IDEA (Institute of Evolution in Africa), University of Alcalá de Henares, Covarrubias 36, 28010, Madrid, Spain. (12) Area of Prehistory, Department of History and Philosophy, University of Alcalá de Henares, C/ Colegios 2, 28801 Alcalá de Henares, Madrid, Spain (13) Área de Prehistoria, Facultad de Geografía e Historia, Complutense University of Madrid, c/Profesor Aranguren S/N, 28040 Madrid, Spain. * Corresponding author. 8 Cuaternario y Geomorfología (2022), 36 (1-2), 7-35 Summary We present the discovery of a Middle Pleistocene fossil assemblage at Los Villares locality (Ruidera, Ciudad Real, Castilla-La Mancha), which has possible evidence of associated human activity. The potential of the site has been evaluated through multidisciplinary research including taxonomy, anatomy, deep learning, and direct dating of fossil remains. A surface study carried out in 2017, over a very limited area (2 m2) on the slope of one of the Ruidera lakes led to the discovery of more than 50 fossil specimens, including cranial (mainly teeth) and postcranial remains. This rich assemblage is dominated by the remains of Caprinae, although the presence of some small or medium carnivore remains also stands out. The identification of a cut mark, tested with Convolutional Neural Networks, suggests the presence of human activity within the bone assemblage. Several fossils were directly dated using a multi-technique approach involving radiocarbon, U-Th, and ESR methods. The results constrain the fossil assemblage to between 300 ka and 400 ka, positioning Los Villares as one of the first Middle Pleistocene localities identified in the Upper Guadiana basin, on the Southern edge of the Southern Iberian Plateau. These promising initial results show the great potential of the site to contribute to filling a gap of knowledge in the Pleistocene archaeo-paleontological record of the Iberian Peninsula. Nevertheless, we also acknowledge the need for systematic excavations in the future, not only to obtain a better idea of the lateral and stratigraphic extension of the fossil assemblage and its complete taxonomic composition, but also to confirm the human presence at the site. Key words: paleontology; Southern Iberian Plateau; Middle Pleistocene; direct dating; taphonomy; cut mark. Resumen Presentamos aquí el descubrimiento de un conjunto fósil del Pleistoceno Medio encontrado en Los Villares (Ruidera, Ciudad Real, Castilla-La Mancha), con posible evidencia asociada de actividad humana. El potencial del yacimiento ha sido evaluado a través de una investigación multidisciplinaria que incluye anatomía, Deep learning y datación directa de restos fósiles. Un estudio de los materiales superficiales, donados recientemente y procedentes de un área muy limitada (2 m2) en la ladera de una de las lagunas de Ruidera, condujo al descubrimiento de más de 50 especímenes fósiles, incluidos restos craneales (principalmente dientes) y postcraneales. Este rico conjunto está dominado por restos de Caprinae, aunque también destaca la presencia de algún resto de pequeño o mediano carnívoro. La identificación de una marca de corte, testada con técnicas estadísticas utilizando Redes Neuronales (Convolutional Neural Networks), en un resto pone de manifiesto la presencia de actividad antrópica en el conjunto. Además, varios fósiles fueron datados directamente por medio de un enfoque de múltiples técnicas que involucran métodos de Radiocarbono, U-Th y ESR, proporcionando un rango cronológico de entre 300 ka y 400 ka para los fósiles. Este es, que sepamos, uno de los primeros yacimientos fechados del Pleistoceno medio en la Meseta del Sur de la Península Ibérica, especialmente en la cuenca alta del Río Guadiana. Estos prometedores resultados iniciales demuestran el gran potencial de la localidad de Los Villares para contribuir a llenar un vacío de conocimiento en el registro arqueológico-paleontológico del Pleistoceno Medio de la Península Ibérica. No obstante, reconocemos la necesidad de realizar excavaciones sistemáticas en el futuro para tener no solo una mejor idea de las extensiones laterales y estratigráficas, así como de la composición taxonómica del conjunto fósil, sino también para confirmar la presencia humana en el sitio. Palabras clave: paleontología; Submeseta Sur Ibérica; Pleistoceno medio; datación directa; tafonomía; marca de corte. 9 Cuaternario y Geomorfología (2022), 36 (1-2), 7-35 1. Introduction The Iberian Peninsula is characterized by an abundant and diverse Pleistocene archeo-paleontological record, with fossil assemblages mostly found in karstic (caves, chambers, or shelters), fluvial (river channels or floodplains), and marshy or lacustrine environments (Arribas Herrera & Jordá Pardo, 1999; Jordá Pardo, 2008a). While these localities cover the entire Peninsula, they are mostly concentrated in several geological units (Jordá Pardo, 2008b) in both mountainous areas such as the Cantabrian Mountains and the Pyrenees, the Central System, the Iberian, and Coastal-Catalan Ranges and the Baetic System, and the major river basins of the Tagus, Guadiana (including the Campo de Calatrava sub-basin) and Guadix-Baza. However, whilst Middle Pleistocene fossil deposits are frequent in these mountainous regions, only the Guadix-Baza basin contains numerous large mammal assemblages, and sites in the two basins (the Tagus and Guadiana Basins) in the “Southern Iberian Plateau” are much less proliferous, leaving a substantial gap in the record. For example, in the Cantabrian Mountains and Pyrenees, El Sidrón (Rosas et al., 2012), El Castillo (Valdés, 1984) and Lezetxiki (Baldeón, 1993) have yielded significant Pleistocene assemblages of faunal and human remains, as have, Pinilla del Valle (Baquedano et al., 2012) and Jarama VI (Jordá Pardo, 2007) in the Central System and the Atapuerca karst complex (Arsuaga et al., 1999; Bermúdez de Castro et al., 1997), Torralba and Ambrona sites (Falguères et al., 2006), Cueva de los Casares (Barandiarán & Altuna, 1973) and Cova de Bolomor (Fernández Peris et al., 2008) in the Iberian and Coastal-Catalan Ranges. Similarly, among the three basins mentioned by Jordá Pardo (2008b?), the Guadix-Baza is represented by a substantial number of fossil sites including Fonelas P-1 (Arribas Herrera et al., 2001), Solana del Zamborino (Álvarez-Posada et al., 2017), and the Orce archaeo-paleontological complex (Arribas Herrera & Palmqvist, 1998; Titton et al., 2020; Toro-Moyano et al., 2013). In contrast, the Tagus and Guadiana basins in the “Southern Iberian Plateau” are much less proliferous. The Tagus Basin is represented by fossil sites such as Áridos (Blain et al., 2014; Santonja et al., 2001), TAFESA (Sesé, 2010), or Perales del Río (Sesé & Soto, 2002) and Jordá Pardo (2008a) lists only Pliocene fossil sites such as Las Higueruelas in the Upper Guadiana basin. This reflects a real bias in the archeo-paleontological record of the Southern Iberian Plateau, and more specifically in the Upper Guadiana Basin (García-Martínez, 2019). In this context, the discovery of a new Pleistocene fossil assemblage at Los Villares, i.e. within the Upper Guadiana basin in the Southern Plateau, is of particular interest. Here we report the initial results from a multidisciplinary investigation including taxonomy, taphonomy, and geochronology designed to evaluate the potential of Los Villares. 1.1. Context 1.1.1. Overview of the evidence of the human presence in the Iberian Peninsula during the Middle Pleistocene While the Early Pleistocene archaeological and fossil record of the Iberian Peninsula is quite rich and diverse, it shows a significant hiatus from the end of the Early Pleistocene to around 500 ka, well into the Middle Pleistocene, where there is little evidence to support the presence of hominin species. Around 500-400 ka, a human species very different to the earlier Homo antecessor, both in terms of body size and shape (Arsuaga et al., 1999, 2015; Carretero et al., 2004) and associated typo-technological assemblages (Ollé et al., 2013), has been well documented in Atapuerca (Sima de Los Huesos and Gran Dolina TD10) and Gruta de Aroeira (Daura et al., 2017) at the end of the Tagus Basin (Portugal). This species, now considered as “pre-Neanderthal” (Arsuaga et al., 2014), produced a Mode 2 technology associated with systematic and directional carcass processing, includ- 10 Cuaternario y Geomorfología (2022), 36 (1-2), 7-35 ing hunting events like in Gran Dolina TD10 (García-Medrano et al., 2015; Rodriguez-Hidalgo et al., 2017). The rest of the Middle and Late Pleistocene is well documented by the presence of Neanderthals in the Iberian Peninsula, with Mode 3 technology and numerous fossil sites all around the Iberian Peninsula (Baquedano et al., 2012; Fernández Peris et al., 2008; Rodriguez-Perez et al., 2017; Rosas et al., 2006; Rosas et al., 2017; Walker et al., 2011). However, whilst the Iberian Peninsula is characterized by an overall dense and diverse Early-to-Late Pleistocene archaeological and fossil record, the Southern Plateau, and the Guadiana Basin, in particular, is clearly contrasting with the big picture, with only few Middle Pleistocene archeo-paleontological localities identified so far. 1.1.2. The Plio-Pleistocene fossil record from the Southern Plateau and the Guadiana Basin The Southern Plateau (Meseta sur) of the Iberian Peninsula is an extended plain of moderFigure 1: Geographical setting of the Tagus (green) and Guadiana (brown) Basins. The main fossil sites identified in the Upper Guadiana basin (yellow circles) are highlighted. 1: Bonete y Piedrabuena (Pliocene); 2: Las Higueruelas (Pliocene); 3: Valverde de Calatrava I (Lower to Middle Pleistocene) y II (Pliocene); 4: Cueva de los Toriles (Middle Pleistocene); 5: Los Villares (Middle Pleistocene); 6: El Provencio (Middle Pleistocene); 7: Fuensanta del Júcar (Lower to Middle Pleistocene). Modified from https://commons.wikimedia.org/, CC BY-SA 3.0. Figura 1: Marco geográfico de las cuencas del Tajo (verde) y Guadiana (marrón). Se destacan los principales yacimientos fósiles identificados en la cuenca del Alto Guadiana (círculos amarillos). 1: Bonete y Piedrabuena (Plioceno); 2: Las Higueruelas (Plioceno); 3: Valverde de Calatrava I (Pleistoceno Inferior a Medio) y II (Plioceno); 4: Cueva de los Toriles (Pleistoceno Inferior a Medio); 5: Los Villares (Pleistoceno Medio); 6: El Provencio (Pleistoceno Medio); 7: Fuensanta del Júcar (Pleistoceno Inferior a Medio). Modificado de https://commons.wikimedia.org/, CC BY-SA 3.0. 11 Cuaternario y Geomorfología (2022), 36 (1-2), 7-35 ate altitude (mostly around 600-700 m.a.s.l.) compared with its Northern part, which is dissected by the present-day fluvial networks of the Tagus and Upper Guadiana rivers (Fig. 1). It is bordered by mountain ranges such as the Central System (to the North), the Iberian Chain (to the Northeast), and the Baetic System, and Sierra Morena (to the South) (Casas-Sainz & de Vicente, 2009). The geography of this area is characterized by large plain terrains along the Tagus and Guadiana basins, except for the Toledo Mountains area, which roughly separates the two basins. From a geological point of view, the easternmost part of the Southern Plateau is mainly composed of Neogene and Quaternary deposits, except for some islets within the Campo de Montiel area which are formed by large limestone plaques from the Mesozoic era and settled on much folded Paleozoic substrates giving rise to more rugged landscapes (Casas-Sainz & de Vicente, 2009; de la Cruz, 2013; Montero González, 2003; Jiménez Ramírez & Chaparro Sabina, 1989; Ramírez et al., 1982). In the westernmost part of the Southern Plateau, Precambrian-Paleozoic substrates are most common (Casas-Sainz & de Vicente, 2009). Given these geological and geographical features, Plio-Pleistocene fossil assemblages identified in the region are found mostly associated with either the fluvial deposits of the Tagus and Upper Guadiana basins or the karstic sedimentary infilling systems in the easternmost part (García-Martínez, 2019). Focusing on the Guadiana Basin, a few and relatively unexplored Pliocene localities have been described in the literature, such as Las Higueruelas (Badiola et al., 2007), Bonete y Piedrabuena (Torres & Mazo, 1991) and Valverde de Calatrava II (Alberdi et al., 1984), both in the Campo de Calatrava area (Fig. 1). Their Pliocene chronology has been mostly inferred from biochronology. Lower-to-Middle Pleistocene localities have been also reported, such as Valverde de Calatrava I, Fuensanta del Júcar and El Provencio (Domínguez-Solera et al., 2020; Mazo et al., 1990) (Fig. 1). The latter two were found in fluvial deposits associated with the Júcar and Záncara Rivers, in terraces positioned +60 m and +15-16 m above the current river channel, respectively, and belonged to the Guadiana fluvial network during Pleistocene times (Mazo et al., 1990; Santisteban & Schulte, 2007). Based on the identification of the large mammal taxa Mammuthus meridionalis or Hippopotamus amphibius major at Valverde de Calatrava I and Fuensanta del Júcar, age estimates of around 1.3-0.8 Ma (Aguirre, 1989; Mazo, 1999) and 1 Ma (Mazo et al., 1990) were inferred respectively. The archeo-paleontological site of El Provencio (Cuenca) provided fossil remains initially attributed to Mammuthus meridionalis, suggesting a similar Lower Pleistocene chronology to the Fuensanta del Júcar assemblage (Mazo et al., 1990). However, a Middle Pleistocene age is more likely as a recent study suggested that these remains could be instead attributed to Mammuthus trogontherii, a species covering a younger time range, and one ESR date obtained from the lowermost and sterile stratigraphic unit provides a burial age of around 800 ka (Domínguez-Solera et al., 2020). Finally, García-Martínez et al. (2020) and Megía García et al. (2020) recently reported the discovery of a fossil assemblage and associated Middle Paleolithic lithic tools at the cave site of Cueva de Los Toriles (Carrizosa, Castilla-La Mancha) (Fig. 1), with a potential Middle Pleistocene chronology (unpublished data). 1.2. Los Villares locality In April 2009, P.R.M.M. was informed of the discovery of bone fragments on one of the slopes of the Los Villares house state, located on the “La Colgada” lake, on the right bank of the Parque Natural de las Lagunas de Ruidera (Ciudad Real, Castilla-La Mancha) (Fig. 2). It was not until a decade later that a multidisciplinary team was able to gather and provide the first formal evaluation of this new site, which is presented here for the first time to the scientific community. The Ruidera lakes are a series of 14 lakes dammed with tuff barriers along ~15 km of a 12 Cuaternario y Geomorfología (2022), 36 (1-2), 7-35 Figure 2: Geographical setting of Los Villares locality (Ruidera, Ciudad Real, Castilla-La Mancha). Legend: a) Spain, b) Castilla-La Mancha province, c) Campo de Montiel area. Built with QGIS 3.12.3 -GNU Free Licenseand free wms data from Instituto Geográfico Nacional (IGN), Spain). Figura 2: Marco geográfico de la localidad de Los Villares (Ruidera, Ciudad Real, Castilla-La Mancha). Leyenda: a) España, b) Provincia de Castilla-La Mancha, c) Zona de Campo de Montiel. Construido con QGIS 3.12.3 -GNU Free Licensey datos wms gratuitos del Instituto Geográfico Nacional (IGN), España). 13 Cuaternario y Geomorfología (2022), 36 (1-2), 7-35 headwater of the Guadiana River, that incise the Mesozoic carbonates forming the high plateau of Campo de Montiel (Ordóñez et al., 2005, González Martín et al. (1987) in the southern part of the Southern Plateau. This exceptional karstic landcape, is similar to the Plitvice lakes in Croatia, mentioned in Roman times (Álvarez et al., 2007; González Martín, 2007; González Martín et al., 2004; Planchuelo, 1952), and has important natural resources such as fish farms, game animals and perpetual pastures (Moya-Maleno, 2011). The current morphology of the lakes dates from the Holocene, around 6,000 years ago (González Martín et al., 1987; Montero González, 2003). However, geomorphological studies suggest the constant transformation of the area over time, both by natural episodes, such as the earthquake of the 16th century, and human action (Hijano, 2013; López Sanz, 1993; Martín, 2000), leading to modification and, sometimes, the disappearance of some lakes (Álvarez et al., 2007; Marín Magaz, 2007). The paleontological site of Los Villares is located on the eastern slope above the La Colgada lake at 808 meters above sea level (m.a.s.l.), i.e. about 21 meters above the current lake level (787 m.a.s.l.) and about 80 meters from the shore on a 45º slope. During the construction of a local road, more than 50 excellently-preserved faunal fossils were identified and surface-collected in 2017 by locals from an area of about 2 m2 (Fig. 3). 2. Material and methods 2.1. Fossil assemblage Anatomical and taxonomic identifications of all fossil remains recovered from Los Villares (N=57) were based on the comparative material collections from the Museo Nacional de Ciencias Naturales (MNCN-CSIC). Of the 57 elements, 53 were clean and individualized (Figs. 4, 5), while three were still embedded in breccia (Fig. 6). The last specimen was only studied through pictures since it was recovered in 2009 and access to the original fossil could not be obtained (Fig. 7). 2.2. Preliminary taphonomic study First, we analyzed the breakage pattern of the bones using the Villa and Mahieu model (Villa & Mahieu, 1991). In addition, the surface of the bones was inspected with a binocular magnifier to identify and describe potential marks resulting from human activity. Once identified, we carried out a deep learning approach to confirm their human origin. To accurately identify the type of taphonomical alteration (i.e., cut mark, tooth mark, trampling mark, etc.) present in the specimen, an approach combining deep learning and computer vision was employed. Deep learning and Figure 3: Fossiliferous sediments exposed after a road construction at the house state of Los Villares (Ruidera, Ciudad Real), in 2009. Figura 3: Sedimentos fosilíferos expuestos tras la construcción de una carretera en la urbanización Los Villares (Ruidera, Ciudad Real), en 2009. 14 Cuaternario y Geomorfología (2022), 36 (1-2), 7-35 Figure 4: Examples of dental fossil remains recovered from Los Villares locality (see descriptions in Table 5). Figura 4: Ejemplos de restos fósiles dentales recuperados en la localidad de Los Villares (ver descripción en la Tabla 5). Figure 5: Examples of skeletal elements recovered from Los Villares locality (see descriptions in Table 5). Figura 5: Ejemplos de elementos óseos recuperados en la localidad de Los Villares (ver descripción en Tabla 5). 15 Cuaternario y Geomorfología (2022), 36 (1-2), 7-35 Computer Vision algorithms based on Convolutional Neural Networks (CNN) have recently been successfully used to classify different types of bone surface modifications (e.g., (Byeon et al., 2019; Cifuentes-Alcobendas & Domínguez-Rodrigo, 2019; Domínguez-Rodrigo et al., 2020). These algorithms have been able to outperform human analysts in accuracy rates while providing more replicable results and more objectivity than previous methods (Domínguez-Rodrigo et al., 2020). Here, we used transfer learning to guarantee optimal feature extraction of the models with relatively small sample sizes. The models used were both with sequential (VGG16 and VGG19) and parallel (ResNet50, DenseNet201, and InceptionV3) architectures, and their accuracy rates ranged from 89% to 92% in our preliminary tests. The reader may refer to Domínguez-Rodrigo et al. (2020) (ResNet50, VGG16, VGG19, and InceptionV3 models) and Abellán et al. (2021) (DenseNet201) for more information on Artificial Intelligence model building and tuning. Additionally, we selected a Rectified Linear Unit (ReLU) as the activation function and a Stochastic Gradient Descent (SGD) as an optimizer, since these two have become the baseline of CNNs for their efficiency and good overall performance (Goodfellow et al., 2016; Ketkar, 2017). Figure 6: Examples of brecciated fossil specimens (red arrow) recovered from Los Villares locality (see descriptions in Table 5). Figura 6: Ejemplos de especímenes fósiles brechificados (flecha roja) recuperados de la localidad de Los Villares (ver descripciones en la Tabla 5). Figure 7: Felinae mandible recovered from Los Villares locality. Only pictures could be studied from this specimen since we did not have access to the original specimen (see descriptions in Table 5). Figura 7: Mandíbula de Felinae recuperada de la localidad de Los Villares. Solo se pudieron estudiar fotografías de este espécimen ya que no tuvimos acceso al espécimen original (ver descripciones en la Tabla 5). 22 Cuaternario y Geomorfología (2022), 36 (1-2), 7-35 Table 5: Basic description of the fossil remains collected from Los Villares locality. Key: N/A = not applicable; Indet. = Indeterminate; * = possible attribution, to be confirmed in the future. Tabla 5: Descripción básica de los restos fósiles recolectados en la localidad de Los Villares. Clave: N/A = no aplicable; indet. = Indeterminado; * = posible atribución, por confirmar en el futuro. Anatomy Part Laterality Taxa Development RV001 Teeth Partial Indet. Caprinae Adult RV003 Teeth Proximal Left Caprinae Adult RV004 Teeth Partial Indet. Caprinae Adult RV032 Teeth Partial Indet. Caprinae Adult RV044 Teeth Distal Left Caprinae Adult RV046 Teeth Distal Left Caprinae Adult RV047 Teeth Complete Left Caprinae Adult RV048 Teeth Complete Indet. Caprinae Adult RV049 Teeth Proximal Indet. Caprinae Adult RV050 Teeth Proximal Left Caprinae Adult RV051 Teeth Complete Indet. Caprinae Adult RV007 Calcaneus Complete Indet. Caprinae Adult RV022 Calcaneus Body N/A Caprinae Immature RV029 Calcaneus Distal Left Caprinae Adult RV017 Scapula Diaphysis Indet. Caprinae Adult RV011 Phalanx Shaft Indet. Caprinae Adult RV012 Phalanx Proximal Right Caprinae Adult RV005 Femur Proximal Right Caprinae Adult RV006 Femur Proximal Left Caprinae Immature RV002 Humerus Distal Left Caprinae Adult RV031 Humerus Distal Left Caprinae Immature RV038 Humerus Distal Right Caprinae Adult RV009 Metacarpal Indet N/A Caprinae Adult RV055 Phalanx Indet. Left Caprinae Adult RV010 Tibia Proximal Left Caprinae Adult RV021 Tibia Indet. N/A Caprinae Adult RV025 Tibia Distal Left Caprinae Adult RV030 Tibia Diaphysis Indet. Caprinae Immature RV053 Tibia Distal Left Caprinae Adult RV014 Ulna Distal Left Caprinae Immature RV018 Ulna Proximal Left Caprinae Immature RV020 Ulna Partial Indet. Caprinae Adult RV013 Vertebra Indet. Indet. Caprinae Immature RV041 Metacarpal Indet. Indet. Caprinae Indet. RV019 Scapula Indet. Indet. Caprinae* Adult RV008 Phalanx Indet. Indet. Caprinae* Adult RV037 Femur Distal Left Caprinae* Indet. RV027 Calcaneus Distal Left Caprinae* Adult 23 Cuaternario y Geomorfología (2022), 36 (1-2), 7-35 minimum number of 5 Caprinae individuals. Additionally, 1 Felinae, 1 Cervidae, and 1 Lagomorpha were also assessed, totaling a minimum number of 8 individuals and 4 taxa in the assemblage. 3.2. Preliminary taphonomic study So far, the fossil record recovered at the Los Villares site is scarce. However, we were still able to analyze the fracture pattern in 10 long bones found at the site. The analysis of the fracture outlines shows that most of the bones have irregular edges. The relation between the shaft length and preserved length shows that no bones preserve more than half of their length. Our study indicates that the long bones were fractured when they were already dry, and overall show no evidence of human activity. The only exception is the remain RV-017, which does show potential evidence of human-made alterations (i.e., cut mark; Fig. 10). This possibility was statistically evaluated through an artificial intelligence (AI) approach. Results show that all the AI models consistently agree in classifying this bone surface modification (BSM) as a cut mark (Table 6). Moreover, four out of the five models RV056 Teeth Indet. Indet. Carnivore* Adult RV043 Tibia Indet. Indet. Cervidae Adult RV054 Mandible Proximal Right Felinae Adult RV016 Rib Indet. Indet. Indet. Indet. RV024 Coxal bone Distal Right Indet. Indet. RV042 Scapula Partial Indet. Indet. Indet. RV033 Indet. Proximal Right Indet. Indet. RV034 Indet. Partial Indet. Indet. Indet. RV035 Indet. Partial Indet. Indet. Indet. RV036 Indet. Partial Indet. Indet. Indet. RV039 Indet. Partial Indet. Indet. Indet. RV040 Indet. Partial Indet. Indet. Indet. RV028 Tibia Partial Indet. Indet. Indet. RV023 Vertebra Proximal Indet. Indet. Indet. RV026 Vertebra Complete Left Indet. Indet. RV045 Femur Hemimandible Indet. Indet. Adult RV052 Femur Complete Indet. Lagomorpha Adult RV015 Long bone Partial Indet. Large Carnivore Indet. Indet. RV057 Metacarpal Various Indet. Small Carnivore Indet. Adult Figure 10: Detailed picture of the potential humanmade cut mark found on fossil RV-017. Figura 10: Imagen detallada de la posible marca de corte hecha por humanos encontrada en el fósil RV-017. 24 Cuaternario y Geomorfología (2022), 36 (1-2), 7-35 achieve confidence rates of 99% or higher, strengthening the reliability of this identification. Also, both parallel and sequential model architectures achieve similar results, showing that they are equally suitable for these classification problems and thus, both should be considered when conducting AI analyses. 3.3. Geochemical characterization of the breccia Grain size analysis shows that the sediment is silty sand, almost equally dominated by the sandy (44.1 %) and silty (41.5%) fractions, while clay represents about 14.4 % of the total (Table 7). Quantitative XRF analysis shows that it is largely dominated by Ca (45.4%) and Si (11.01 %), most likely resulting from carbonates and silicates, respectively (Table 8). The non-negligible amount of Al (4.0%) may illustrate the presence of clays, in agreement with grain size analysis, and the presence of feldspars. 3.4. Dating results 3.4.1. Radiocarbon dating Although theoretically bone with 0.2 %N should contain the ≥ 1 % collagen required for radiocarbon dating, (Brock et al., 2012) found bone containing ≥ 0.7 % N had a 73 % chance of containing sufficient collagen when extracted and cleaned with an ultrafiltration protocol. Unfortunately, samples from Los Villares contained between 0.02-0.07 %N and were not datable with radiocarbon. 3.4.2. LA U-series analyses High-resolution LA U-series analyses were performed to (i) evaluate the spatial distribution of U-series data across dental tissues of samples #546, #547, and #547, and (ii) evaluate the suitability of sample #546 for combined U-series/ESR dating. Table 6: Classification values for the bone surface modifications (BSM) image analyzed. The table shows the confidence of each model architecture when classifying the BSM into the different three classes. The categories definition can be found in Domínguez-Rodrigo et al. 2020. Tabla 6: Valores de clasificación para la imagen de modificaciones de la superficie ósea (BSM) analizada. La tabla muestra la confianza de cada modelo de arquitectura al clasificar el BSM en las tres clases diferentes. La definición de las categorías se encuentra en Domínguez-Rodrigo et al. 2020. Tooth mark Cut mark Trampling mark DenseNet201 21.79% 77.20% 0.99% VGG16 0% 100% 0% VGG19 0.04% 99.95% 0.01% InceptionV3 0.20% 99.78% 3.21·10-5% ResNet50 0.73% 99.25% 1.79·10-5% Table 7: XRF results from the brecciated sediment showing the relative proportions of various major elements in their oxidized state. LOI = loss of ignition. Tabla 7: Resultados XRF del sedimento brechificado que muestran las proporciones relativas de varios elementos principales en su estado oxidado. LOI = loss of ignition. Amount (%) SiO211.05 Al2O34.04 Fe2O31.78 MnO 0.027 MgO 0.39 CaO 45.39 Na2O0.03 K2O0.58 TiO20.21 P2O5 1.25 SO30.02 LOI 35.24 25 Cuaternario y Geomorfología (2022), 36 (1-2), 7-35 Sample #547 shows apparent ages that are systematically >100 ka across dentine and enamel: the first enamel domain (EN1-EN2) analyzed, and the dentine return overall consistent ages of ~127 ka and ~129 ka, respectively, while the second enamel domain yields younger estimates of ~106 ka. The asymmetric shape of the age profile suggests that there is a more recent uranium uptake overprint from one side of the tooth. In comparison, sample #548 shows much younger age estimates, although overall consistent across dentine (~51 ka) and enamel (~54 ka) (Table 3). Finally, LA analysis of sample #546 returns somewhat consistent U-series data within enamel and dentine domains. No recent overprint is observed within the tooth. The uranium concentration varies between 0.2 and 0.4 ppm in enamel, and between 26 and 30 ppm in dentine. Similarly, the apparent U-series ages vary within a relatively narrow range (dentine) or are all 1-σ consistent (enamel). Pseudo bulk mean values of ~119 ka and ~171 ka are obtained for the enamel and dentine, respectively. In contrast, the cement returns more scattered U-series data, with apparent U-series age estimates that are close or even beyond secular equilibrium, indicating that the tissue is experiencing uranium leaching. While the LA raster CE2 provides an apparent age result of 319.8 ± 38.9 ka (Table 3) that could be interpreted as a minimum age constraint for the fossil, the occurrence of uranium leaching in this tissue (LA raster CE1) suggests that the reliability of this result should be considered with caution. All apparent U-series age estimates should be regarded as minimum age constraints (provided there is no evidence for uranium leaching), since uranium uptake may sometimes be significantly delayed after the death of the organism. Consequently, a minimum age of ~130 ka, ~50 ka, and ~170 ka may be proposed for samples #547, #548, and #546, respectively. Additionally, these results indicate that the age of the fossil specimens lies beyond the Radiocarbon time range. Finally, U-series data collected for sample #546 show that (i) the enamel tissue has very low uranium concentrations (<0.5 ppm), and (ii) there is no evidence of uranium leaching in dentine and enamel tissues (i.e., finite age can be calculated), unlike in cement. These two observations indicate that the tooth is overall suitable for ESR dating: sample preparation for ESR dating was therefore focused on an area of the tooth without cement. 3.4.3. ESR dose evaluation Measurement precision achieved is overall excellent (0.8%), resulting in a DE repeatability of < 2 % (Table 3). DE value obtained over the full irradiation dose range is 303 ± 13 Grays (Gy) (Fig. 9). Goodness-of-fit is excellent (adjusted r2 > 0.99). The corresponding Dmax/ DE ratio value falls within the recommended range by Duval and Grün (2016) for a DE value of this magnitude (5 < Dmax/DE < 10). Additional dose-response curve (DRC) fitting was performed using data weighting by 1/s2, yielding virtually the same DE value (Table 3). This illustrates the negligible impact of data weighting on the fitting outcome. Table 8: Granulometric results obtained from the brecciated sediment using the Robinson pipette method. Tabla 8: Resultados granulométricos obtenidos del sedimento brechificado utilizando el método de la pipeta de Robinson. Size class Fraction (µm) retained (% weight) Sand 2000-500 30.3 500-212 6.6 212-53 7.2 Total sand 44.1 Silt 53-20 13.0 20-2 28.5 Total silt 41.5 Clay <2 14.4 26 Cuaternario y Geomorfología (2022), 36 (1-2), 7-35 3.4.4. Combined U-series and ESR age calculations Solution U-series analyses of the bulk powdered dental tissues return apparent age estimates of ~74 and ~138 ka (Table 4). As minimum age constraints for the fossil, they are not incompatible with the LA results, which are somewhat older for both tissues. The differences observed between solution and LA analyses most likely result from the spatial heterogeneity of U-series data within the tooth, which may sometimes be significant since the samples were collected from different parts of the tooth. Combined U-series/ ESR age calculations were performed using solution U-series analytical results since the powdered enamel sample is the same that was independently analyzed by ESR, while the dentine sample was directly attached to this enamel sample. Age calculation based on the US model (Grün et al., 1988) yields a combined US-ESR estimate of 346 +23 -21 ka for tooth #546 (Table 4). Modeled p-values indicate a recent uptake in the enamel (>0), and a somewhat early uptake in the dentine instead (p< 0). Dose rate evaluation indicates that the beta component from the dentine and the gamma component represents about 30% of the total value. The cosmic dose rate is also not negligible, representing about 20 % of the total value. Three main sources of uncertainty may be identified. First, the absence of in situ dosimetry precludes a proper evaluation of the true gamma dose rate. Given the non-negligible weight of this component in the dose rate evaluation, we anticipate that any heterogeneity from the surrounding sedimentary environment, may potentially significantly impact the calculated age result. However, this is presently impossible to quantify. In contrast, the influence of both long-term depth and long-term water content can be roughly assessed. Age simulations using depth values of 0.5, 2 and 5 m return US-ESR estimates of ~338 ka, ~358 ka, and ~386 ka, respectively, i.e., -2%, +3%, and +12 % compared to the value initially calculated. In all cases, the calculated values remain 1-σ consistent with the age of 346 +23 -21 ka. Variation of the long-term water content by ± 5% and ± 10% compared with the initial value, return USESR estimates that vary by ± 3-4% and ± 7-8%, respectively. Again, all the calculated ages remain 1-σ consistent with the age initially calculated. In summary, even when using somewhat extreme long-term depth and water content values, the US-ESR age remains between 300 and 400 ka. This chronology would position the fossil assemblage between MIS 9 and 11 (Lisiecki & Raymo, 2005). 3.5. Significance of Los Villares fossil assemblage in the Southern Plateau and the Spanish archeo-paleontological record To the best of our knowledge, this is the first directly dated fossil assemblage of the Middle Pleistocene age reported in the Guadiana Basin (Southern Plateau), filling a gap in the Pleistocene fossil record of the Iberian Peninsula (Fig. 11). In the Northern Plateau, the Atapuerca Complex (Burgos) includes Galería TG8 level dating >350 ka (Demuro et al., 2014; Pérez-González et al., 2001), Sima de los Huesos dated to around 450 ka (Arsuaga et al., 2014) and the Gran Dolina TD10 level with an age of between 320-370 ka (Falguères et al., 1999; Moreno et al., 2012). In the surroundings of the Iberian chain, there are Middle Pleistocene fossil sites linked to river terraces such as Torralba and Ambrona complex (Soria, Castilla y León) and Cuesta de la Bajada (Teruel), at ages of 314-366 ka for AS6 from the Lower Complex of Ambrona (AS1 to AS6; Falguères et al., 2006)), and 243-337 ka (for Cuesta de la Bajada; Santonja et al., 2014). In the upper Tagus Basin, Áridos 1 and 2 have yielded microand large mammals as well as Acheulean lithic artifacts (Santonja et al., 2001 and references therein). They belong to the stratigraphic unit Arganda I, which has been chronologically constrained to between 295 and 365 ka based on ESR and Amino Acid Racemization methods (Panera et al., 2011). 27 Cuaternario y Geomorfología (2022), 36 (1-2), 7-35 Downstream in the Tagus Basin, the Gruta da Aroeira (Portugal) contains Mousterian levels that include a partial human skull and have been dated to 390-436 ka (Daura et al., 2017). In the Coastal-Catalan Ranges, levels XVII-XV from Cova del Bolomor (Fernández Peris et al., 2008), have been dated around 347-242 ka. Finally, in the South of the Iberian Peninsula, the Acheulean assemblage and associated fauna excavated from the site of Solana del Zamborino (Guadix-Baza basin) have been recently chronologically constrained to between around 300-480 ka using magnetostratigraphy (Álvarez-Posada et al., 2017). Moreover, from an archeological perspective, the identification of a cut mark on one of the fossils suggests a human presence in the area around 300-400 ka. The chronology of the Los Villares site falls within a period of special interest, in which several Spanish localities have provided very distinct lithic assemblages attributed to either Acheulean or Middle Paleolithic technocomplexes. For some authors, this suggests that different human groups might have been coexisting in the Iberian Peninsula during the Middle Pleistocene (e.g., Méndez-Quintas et al., 2018; Santonja et al., 2014). For example, Cuesta de la Bajada (Teruel) site has yielded a Middle Paleolithic lithic industry and associated fauna possibly correlated to the MIS 9, while the Ambrona archaeo-paleontological site, located on the NE edge of the Central System has produced an Acheulean assemblage of somewhat close chronology (e.g., Duval, 2018; Falguères et al., 2006). Among the archeological sites in the region of Los Villares, the Upper Guadiana Basin, it is worth highlighting El Sotillo (Arroyo & de la Torre, 2013; Ciudad Serrano et al., 1983), Albalá (Arroyo & de la Torre, 2013; Santonja et al., 1977), El Martinete (Querol & Santonja, 1983), Porzuna (Vallespí et al., 1979) and Dehesa and Molino del Emperador (López-Recio et al., 2001; Morín de Pablos et al., 2007). While a large number of Acheulean stone tools have been found in all (e.g., > 1000 bifaces in Porzuna), Molino del Emperador is the only site with a supposedly clear chronostratigraphic context: it is located on the terFigure 11: Main Middle Pleistocene localities in the Iberian Peninsula with similar ages to Los Villares locality. Figura 11: Principales localidades del Pleistoceno Medio en la Península Ibérica con edades similares a la localidad de Los Villares. 28 Cuaternario y Geomorfología (2022), 36 (1-2), 7-35 race +13/16m of the Guadiana River, which has been dated by OSL to around 150 ka (López-Recio et al., 2004). However, this age should be treated with caution, given some methodological issues recently reported by Garcia-Vadillo et al. (2021). The chronologies of all other sites are based on the typological assignment of the lithic assemblages and are therefore more difficult to securely place in time (Fernández et al., 2005). 4. Conclusions The results from this multidisciplinary study show the Los Villares fossil site is likely to be a significant archeo-palaeontological archive in the Southern Iberian Plateau. Not only does it have a high density of well-preserved skeletal and dental remains, but it also has a crucial chronology and may contain evidence of human activity. As such, it will fill a gap of knowledge in an area characterized by a very limited Pleistocene fossil and archeological record compared to other Iberian regions (García-Martínez, 2019). Our multi-technique dating approach provides the first direct age constraints for the archeo-paleontological assemblage, positioning Villares around MIS 9-11 i.e., within a time range of great interest for the understanding of human migrations and occupations in the Iberian Peninsula during the Middle Pleistocene. However, we would also like to clearly acknowledge the existing limitations of the present work to avoid any overinterpretation of these promising results. For example, the ESR chronology is based on a very limited number of samples analyzed, and there is inherent uncertainty associated with the gamma dose rate reconstruction and the evaluation of long-term depth and water content. Nevertheless, considering most sources of uncertainty, our simulations suggest that the age of the fossil specimen remains systematically between 300 and 400 ka. In addition, the presence of a cut mark suggests possible human activity at the site, although other hypotheses such as carnivore origin cannot be ruled out until new additionnal evidence of human presence can be provided. Based on these first results, there is now a crucial need to collect further data about the spatial and stratigraphic extension of the fossil assemblage through a proper systematic excavation campaign. In particular, future investigations will focus on the detailed description of the stratigraphic and geological context. This will enable us to determine whether the fossil assemblage represents a single depositional event. Acknowledgments We thank José Ramos Sanz, president of Los Villares house state (Ruidera, Ciudad Real), and Antonio Ruiz Reinosa, guard of the house state, for their kindness and collaboration regarding the preservation of the archaeo-paleontological heritage in Los Villares. We also thank Dr. Carlos Arroyo and Dr. Javier Campos Fernández de Sevilla for the news about the findings. We are grateful to Salvador Jiménez and Dr. Santiago Domínguez for their initial input on the paleontology, archeology, and geology of the area. María Jesús Alonso Escarza, Leticia Miguens Rodríguez and Javier Iglesias Cibanal (CENIEH), and Faye Liu (UQ) provided invaluable support for the ESR and U-series dating analyses. Aspects of the dating analyses have been funded by the Spanish Ramón y Cajal Fellowship RYC2018-025221-I granted to MD. U-series dating analyses were carried out within the framework of the existing Brisbane Geochronology Alliance between Griffith University, the University of Queensland, and the Queensland University of Technology. The “Juan de la Cierva Formación” program (FJCI-2017-32157), from the Spanish Ministry of Science and Innovation, funds DGM. Grain size and XRF analyses were both performed at CENIEH by Leticia Miguens Rodríguez and Javier Iglesias Cibanal (Technical report I-2020-009-GE), and by Ana Alvaro Gallo (Technical report AM I-2020016), respectively. 29 Cuaternario y Geomorfología (2022), 36 (1-2), 7-35 References Abellán, N., Jiménez-García, B., Aznarte, J., Baquedano, E., & Domínguez-Rodrigo, M. (2021). Deep learning classification of tooth scores made by different carnivores: achieving high accuracy when comparing African carnivore taxa and testing the hominin shift in the balance of power. Archaeological and Anthropological Sciences, 13(2), 1-14. https://doi. org/10.1007/s12520-021-01273-9 Aguirre, E. (1989). Vertebrados del Pleistoceno continental. Mapa Del Cuaternario de España, Escala 1: 1.000 000, 47-69. Alberdi, M. T., Jiménez, E., Mazo, A. V., Morales, J., Sesé, C., & Soria, D. (1984). Paleontología y biostratigrafía de los yacimientos villafranquienses de Las Higueruelas y Valverde de Calatrava II (Campo de Calatrava, Ciudad Real). Actas I Reunión de Estudios Regionales de Castilla-La Mancha, 3, 255-277. Álvarez, M., Cirujano, S., Montero González, E., Roja, C., Rodrigo, M., Piña, E., Rodríguez, J. C., Soriano, Ó., Aboal, M., & Marín, J. P. (2007). Ecología acuática y sociedad de las lagunas de Ruidera. CSIC, Madrid. Álvarez-Posada, C., Parés, J. M., Sala, R., Viseras, C., & Pla-Pueyo, S. (2017). New magnetostratigraphic evidence for the age of Acheulean tools at the archaeo-paleontological site “Solana del Zamborino” (Guadix-Baza Basin, S Spain). Scientific Reports, 7(1), 1-9. https:// doi.org/10.1038/s41598-017-14024-5 Arribas Herrera, A., & Palmqvist, P. (1998). Taphonomy and paleoecology of an assemblage of large mammals: hyaenid activity in the lower Pleistocene site at Venta Micena (Orce, Guadix-Baza Basin, Granada, Spain). Geobios, 31, 3-47. https://doi.org/10.1016/ S0016-6995(98)80056-9 Arribas Herrera, A., Riquelme, J. A., Palmqvist, P., Garrido, G., Hernández, R., Laplana, C., Soria, J. M., Viseras, C., Durán, J. J., & Gumiel, P. (2001). Un nuevo yacimiento de grandes mamíferos villafranquienses en la Cuenca de Guadix-Baza (Granada): Fonelas P-1, primer registro de una fauna próxima al límite PlioPleistoceno en la Península Ibérica. Boletín Geológico y Minero, 112(4), 3-34. Arribas Herrera, A., & Jordá Pardo, J. F. (1999). Los mamíferos del Cuaternario kárstico de Guadalajara (Castilla-La Mancha, España). In la huella del Pasado. Fósiles de Castilla-La Mancha (pp. 327-353). Array. Arroyo, A., & de la Torre, I. (2013). Acheulean large flake technology in Campo de Calatrava (Ciudad Real, Spain). Archaeology, Ethnology, and Anthropology of Eurasia, 41(4), 2-10. https:// doi.org/10.1016/j.aeae.2014.07.002 Arsuaga, J. L., Carretero, J.-M., Lorenzo, C., Gómez-Olivencia, A., Pablos, A., Rodríguez, L., García-González, R., Bonmatí, A., Quam, R. M., Pantoja-Pérez, A., Martínez, I., Aranburu, A., Gracia-Téllez, A., Poza-Rey, E., Sala, N., García, N., Alcázar de Velasco, A., CuencaBescós, G., Bermúdez de Castro, J. M., & Carbonell, E. (2015). Postcranial morphology of the middle Pleistocene humans from Sima de los Huesos, Spain. Proceedings of the National Academy of Sciences, 112, 11524-11529. http://www.pnas.org/content/ early/2015/08/27/1514828112.abstract. https://doi.org/10.1073/pnas.1514828112 Arsuaga, J. L., Lorenzo, C., Carretero, J. M., Gracia, A., Martínez, I., García, N., Bermudez de Castro, J. M., & Carbonell, E. (1999). A complete human pelvis from the middle Pleistocene of Spain. Nature, 399(6733), 255-258. https:// doi.org/10.1038/20430 Arsuaga, J. L., Martínez, I., Arnold, L. J., Aranburu, A., Gracia-Téllez, A., Sharp, W. D., Quam, R. M., Falguères, C., Pantoja-Pérez, A., Bischoff, J., Poza-Rey, E., Parés, J. M., Carretero, J. M., Demuro, M., Lorenzo, C., Sala, N., MartinónTorres, M., García, N., Alcázar de Velasco, A., … Carbonell, E. (2014). Neandertal roots: Cranial and chronological evidence from Sima de los Huesos. Science, 344(6190), 1358-1363. https://doi.org/10.1126/science.1253958 Badiola, E. R., Mazo, A. V., & Ruiz, P. R. (2007). El yacimiento de Las Higueruelas, Alcolea de Calatrava (Ciudad Real): procesos diagenéticos y volcanismo asociado. Estudios Geológicos, 63(2), 67-86. https://doi.org/10.3989/ egeol.07632194 Baldeón, A. (1993). El yacimiento de Lezetxiki (Gipuzkoa, País Vasco): los niveles musterienses. Munibe, 54, 3-97. Baquedano, E., Márquez, B., Pérez-González, A., Mosquera Martínez, M., Huguet Pamiès, R., Espinosa, J. A., Sanchez Romero, L., Panera, J., & Arsuaga, J. L. (2012). Neandertales en el Valle del Lozoya: los yacimientos Paleolíticos del Calvero de la Higuera (Pinilla del Valle, Madrid). Mainake, 33, 83-100. Barandiarán, I., & Altuna, J. (1973). La Cueva de Los Casares (Riba de Saelices, Guadalajara). Excavaciones Arqueológicas en España. (Vol. 76). Ministerio de Educación y Ciencia. 30 Cuaternario y Geomorfología (2022), 36 (1-2), 7-35 Bermúdez de Castro, J. M., Arsuaga, J. L., Carbonell, E., Rosas, A., Martínez, I., & Mosquera, M. (1997). A Hominid from the Lower Pleistocene of Atapuerca, Spain, Possible Ancestor to Neandertals and Modern Humans. Science, 276, 1392-1395. https://doi.org/10.1126/ science.276.5317.1392 Bermúdez de Castro, J. M., Martinón-Torres, M., Gomez-Robles, A., Prado-Simon, L., MartinFrances, L., Lapresa, M., Olejniczak, A., Carbonell, E., Gracia, A., Olejniczak, A., PradoSimon, L., Gomez-Robles, A., Lapresa, M., Carbonell, E., Arsuaga, J. L., & Bermudez de Castro, J. M. (2011). Early Pleistocene human mandible from Sima del Elefante (TE) cave site in Sierra de Atapuerca (Spain): a paleopathological study. J Hum Evol, 61(1), 12-25. https:// doi.org/10.1016/j.jhevol.2011.01.004 Bermúdez de Castro, J. M., Martinón-Torres, M., Martín-Francés, L., Modesto-Mata, M., Martínez-de-Pinillos, M., García, C., & Carbonell, E. (2017). Homo antecessor: The state of the art eighteen years later. Quaternary International, 433, 22-31. https://doi.org/10.1016/j. quaint.2015.03.049 Blain, H.-A., Santonja, M., Pérez-González, A., Panera, J., & Rubio-Jara, S. (2014). Climate and environments during Marine Isotope Stage 11 in the central Iberian Peninsula: the herpetofaunal assemblage from the Acheulean site of Áridos-1, Madrid. Quaternary Science Reviews, 94, 7-21. https://doi.org/10.1016/j. quascirev.2014.04.009 Brock, F., Wood, R., Higham, T. F. G., Ditchfield, P., Bayliss, A., & Ramsey, C. B. (2012). Reliability of nitrogen content (% N) and carbon: nitrogen atomic ratios (C: N) as indicators of collagen preservation suitable for radiocarbon dating. Radiocarbon, 54(3-4), 879-886. https:// doi.org/10.1017/S0033822200047524 Byeon, W., Domínguez-Rodrigo, M., Arampatzis, G., Baquedano, E., Yravedra, J., Maté-González, M. A., & Koumoutsakos, P. (2019). Automated identification and deep classification of cut marks on bones and its paleoanthropological implications. Journal of Computational Science, 32, 36-43. https://doi.org/10.1016/j. jocs.2019.02.005 Carretero, J. M., Arsuaga, J.-L., Martinez, I., Quam, R. M., Lorenzo, C., Gracia, A., & Ortega, A. I. (2004). Los humanos de la Sima de los Huesos (Sierra de Atapuerca) y la evolucion del cuerpo en el género Homo. In E. Baquedano (Ed.), Homenaje a Emiliano Aguirre (Vol. 4, pp. 120-136). Museo Arqueologico Regional. Casas-Sainz, A. M., & de Vicente, G. (2009). On the tectonic origin of Iberian topography. tectonophysics, 474(1-2), 214-235. https://doi. org/10.1016/j.tecto.2009.01.030 Chetlur, S., Woolley, C., Vandermersch, P., Cohen, J., Tran, J., Catanzaro, B., & Shelhamer, E. (2014). cudnn: Efficient primitives for deep learning. ArXiv Preprint ArXiv:1410.0759. Cifuentes-Alcobendas, G., & Domínguez-Rodrigo, M. (2019). Deep learning and taphonomy: high accuracy in the classification of cut marks made on fleshed and defleshed bones using convolutional neural networks. Scientific Reports, 9(1), 1-12. https://doi.org/10.1038/ s41598-019-55439-6 Ciudad Serrano, A., García Serrano, R., Caballero Klink, A., & Francia Villajos, A. (1983). Materiales paleolíticos de “El Sotillo”. Museo de Ciudad Real, Estudios y Monografías, 8. Clark, T. R., Roff, G., Zhao, J., Feng, Y., Done, T. J., & Pandolfi, J. M. (2014). Testing the precision and accuracy of the U-Th chronometer for dating coral mortality events in the last 100 years. Quaternary Geochronology, 23, 35-45. https://doi.org/10.1016/j. quageo.2014.05.002 Daura, J., Sanz, M., Arsuaga, J. L., Hoffmann, D. L., Quam, R. M., Ortega, M. C., Santos, E., Gómez, S., Rubio, A., & Villaescusa, L. (2017). New Middle Pleistocene hominin cranium from Gruta da Aroeira (Portugal). Proceedings of the National Academy of Sciences, 114(13), 3397-3402. https://doi.org/10.1073/ pnas.1619040114 de la Cruz, M. A. S. (2013). La delimitación del Campo de Montiel: principales enfoques y problemáticas. Revista de Estudios Del Campo de Montiel, 3, 51-84. de Lombera-Hermida, A., Bargalló, A., TerradillosBernal, M., Huguet, R., Vallverdú, J., GarcíaAntón, M.-D., Mosquera, M., Ollé, A., Sala, R., & Carbonell, E. (2015). The lithic industry of Sima del Elefante (Atapuerca, Burgos, Spain) in the context of Early and Middle Pleistocene technology in Europe. Journal of Human Evolution, 82, 95-106. https://doi.org/10.1016/j. jhevol.2015.03.002 Demuro, M., Arnold, L. J., Parés, J. M., PérezGonzález, A., Ortega, A. I., Arsuaga, J. L., Bermúdez de Castro, J. M., & Carbonell, E. (2014). New luminescence ages for the Galería Complex archaeological site: resolving chronological uncertainties on the acheulean record of the Sierra de Atapuerca, Northern 31 Cuaternario y Geomorfología (2022), 36 (1-2), 7-35 Spain. PLoS One, 9(10), e110169. https://doi. org/10.1371/journal.pone.0110169 Domínguez-Rodrigo, M., Cifuentes-Alcobendas, G., Jiménez-García, B., Abellán, N., PizarroMonzo, M., Organista, E., & Baquedano, E. (2020). Artificial intelligence provides greater accuracy in the classification of modern and ancient bone surface modifications. Scientific Reports, 10(1), 1-11. https://doi.org/10.1038/ s41598-020-75994-7 Domínguez-Solera, S. D., Moreno, D., & PérezGarrido, C. (2020). A new complete sequence from Lower to Middle Paleolithic: El Provencio complex (Cuenca, Spain). Quaternary International, 566, 39-56. https://doi.org/10.1016/j. quaint.2020.04.053 Duval, M. (2015). Electron Spin Resonance Dating of fossil tooth enamel. In W. J. Rink & J. W. Thompson (Eds.), Encyclopedia of Scientific Dating Methods (pp. 239246). Springer Netherlands. https://doi. org/10.1007/978-94-007-6304-3_71 Duval, M. (2018). Sobre el potencial de la resonancia paramagnética electrónica como herramienta geocronológica en contextos geoarqueológicos: un resumen de 30 años de investigación en la Península Ibérica. Boletín Geológico y Minero, 129(1-2), 35-57. https:// doi.org/10.21701/bolgeomin.129.1.002 Duval, M., & Grün, R. (2016). Are published ESR dose assessments on fossil tooth enamel reliable? Quaternary Geochronology, 31, 19-27. https://doi.org/10.1016/j. quageo.2015.09.007 Duval, M., Grün, R., Parés, J. M., Martín-Francés, L., Campaña, I., Rosell, J., Shao, Q., Arsuaga, J. L., Carbonell, E., & Bermudez de Castro, J. M. (2018). The first direct ESR dating of a hominin tooth from Atapuerca Gran Dolina TD-6 (Spain) supports the antiquity of Homo antecessor. Quaternary Geochronology, 47, 120-137. https://doi.org/10.1016/j. quageo.2018.05.001 Falguères, C., Bahain, J.-J., Pérez-González, A., Mercier, N., Santonja, M., & Dolo, J.-M. (2006). The Lower Acheulian site of Ambrona, Soria (Spain): ages derived from a combined ESR/U-series model. Journal of Archaeological Science, 33(2), 149-157. https://doi. org/10.1016/j.jas.2005.07.006 Falguères, C., Bahain, J.-J., Yokoyama, Y., Arsuaga, J. L., Bermúdez de Castro, J. M., Carbonell, E., Bischoff, J. L., & Dolo, J.-M. (1999). Earliest humans in Europe: the age of TD6 gran Dolina, Atapuerca, Spain. Journal of Human Evolution, 37(3-4), 343-352. https://doi.org/10.1006/ jhev.1999.0326 Fernández Peris, J., Barciela, V., Blasco, R., Cuartero Monteagudo, F., & Sañudo Die, P. (2008). El Paleolítico Medio en el territorio valenciano y la variabilidad tecno-económica de la Cova del Bolomor. Treballs d’arqueologia, 14, 141-169. Fernández, V., Arteaga Cardineau, C., Baena Preysler, J., Escalante García, S., González Martín, J. A., López Recio, M., Marín Magaz, J. C., & Morín de Pablos, J. (2005). III. El Pleistoceno y las industrias paleolíticas de la cuenca alta y media del río Guadiana. In Los primeros pobladores de Castilla-La Mancha (pp. 142-190). Ferràndez-Cañadell, C., Ribot, F., & Gibert, L. (2014). New fossil teeth of Theropithecus oswaldi (Cercopithecoidea) from the Early Pleistocene at Cueva Victoria (SE Spain). Journal of Human Evolution, 74, 55-66. https:// doi.org/10.1016/j.jhevol.2014.02.020 García-Martínez, D. (2019). ¿Dónde están los fósiles manchegos? El sesgo en el registro paleontológico del Pleistoceno de La Mancha. Revista de Estudios Del Campo de Montiel, Extra 3, 17-46. https://doi.org/10.30823/ recm.02019116 García-Martínez, D., Valenciano, A., Suárez-Bilbao, A., Palancar, C. A., Megía García, I., Moreno, D., Campaña, I., & Moya-Maleno, P. R. (2020). New remains of a primitive badger from Cueva de los Toriles (Carrizosa, CastillaLa Mancha, Iberian Peninsula) suggest a new quaternary locality in the southern Iberian plateau. Journal of Iberian Geology. https:// doi.org/10.1007/s41513-020-00127-y García-Medrano, P., Ollé, A., Mosquera, M., Cáceres, I., & Carbonell, E. (2015). The nature of technological changes: The Middle Pleistocene stone tool assemblages from Galería and Gran Dolina-subunit TD10. 1 (Atapuerca, Spain). Quaternary International, 368, 92-111. https://doi.org/10.1016/j. quaint.2015.03.006 García-Vadillo F.J., Duval M., Canals-Salomó A., Rodríguez-Álvarez X.-P., Garcia-Garriga J., Carbonell-Roura E. (2021). Contexto crono-estratigráfico y cultural del conjunto lítico de Base Menacho (cuenca del río Guadiana, Badajoz, España): primeros resultados. Cuaternario y Geomorfología 35 (3-4), pp. 147-173. https:// doi.org/10.17735/cyg.v35i3-4.89904 Gibert, J., Gibert, L., Ribot, F., Ferràndez-Canadell, C., Sánchez, F., Iglesias, A., & Walker, M. J. (2008). CV-0, an early Pleistocene human pha-