Research paper
Morphological study of mandibular Canis remains from the El Portalón (Sierra de Atapuerca) Holocene site. Implications for the origin of the domestic dog in prehistory
RAQUEL BLÁZQUEZ-ORTA
Departamento de Geodinámica, Estratigrafía y Paleontología. Grupo Ecosistemas Cuaternarios, Facultad de Ciencias Geológicas, Universidad Complutense de Madrid. Jose Antonio Novais, 12, 28040 Madrid, España; rborta@ucm.es
LAURA RODRÍGUEZ
Área de Antropología. Departamento de Biodiversidad y Gestión Ambiental. Edificio Facultad de Ciencias Biológicas y Ambientales. Universidad de León; León, España; lrodg@unileon.es
Laboratorio de Evolución Humana, Departamento de Historia, Geografía y Comunicación, Universidad de Burgos, Plaza de Misael Bañuelos s/n, 09001 Burgos, España.
MARIÁN GALINDO-PELLICENA
Centro Mixto UCMISCIII de Evolución y Comportamiento Humanos. C/Monforte de Lemos 5, pab 14, 28029 Madrid, España.
Museo Arqueológico Regional. Pz de Las Bernardas s/n, 28801 Alcalá de Henares, Madrid, España; mangeles.galindo@fgua.es
Fundación General de la Universidad de Alcalá de Henares. C/Imagen 3, 28801 Alcalá de Henares, Madrid, España.
NURIA GARCÍA
Departamento de Geodinámica, Estratigrafía y Paleontología. Grupo Ecosistemas Cuaternarios, Facultad de Ciencias Geológicas, Universidad Complutense de Madrid. Jose Antonio Novais, 12, 28040 Madrid, España; nugarcia@ucm.es
Centro Mixto UCMISCIII de Evolución y Comportamiento Humanos. C/Monforte de Lemos 5, pab 14, 28029 Madrid, España.
Corresponding author
EXTENDED ABSTRACT
INTRODUCTION
El Portalón de Cueva Mayor (Sierra de Atapuerca, Burgos, Spain) is one of the most important Holocene archeo-paleontological sites of North Meseta. This site shows evidence of human occupation from the Upper Palaeolithic until the Middle Ages (Carretero et al., 2008). The Bronze Age constitutes a very important legacy in the central region of the Iberian Peninsula. The rich stratigraphic context and the abundant archaeo-palaeontological information of this site (Carretero et al., 2008; Pérez-Romero et al., 2015, 2016) are ideal for studying the different aspects of the domestication process of the wolf (Canis lupus Linnaeus, 1758). Several morphological and molecular analyses demonstrate the wolf as the ancestor of the dog (Germonpré et al. 2009, 2014). Although DNA studies have been carried out to distinguish between dogs and wolves (Lindblad et al., 2005; Druzhkova et al., 2013; Thalmann et al., 2013) it is essential to additionally analyse the differences between dogs and wolves from a morphometric perspective because of the lack of evidence of a specific haplotype in domestic dogs. Here, we study the size and the shape of the mandibles through traditional and geometric morphometric techniques. The traditional method (TM) describes simple changes in size and shape through such measurements as length, width, and height. Geometric morphometry (GM) (2D) allows more complex observation of changes of shape through visual representations of morphological variation (Zelditch et al., 2004).
This work aims to analyse mandibular features of Canis from Middle-Early Bronze Age at the El Portalón site. Also, we compare these remains with current specimens of Iberian wolves (Canis l. signatus), and domesticated dogs (Canis l. familiaris) to study evolutionary changes related to domestication.
EL PORTALÓN SITE
El Portalón (Fig. 1) constitutes one of the entrances to the Cueva Mayor-Cueva del Silo karst system (Carretero et al., 2008; Ortega et al., 2008; Pérez-Romero et al., 2015, 2016). Although several excavations were carried out at this site during the 1970s, the current team defined a stratigraphic sequence with a radiometrically dated chronology of 30 ka – 1000 years BP (Carretero et al., 2008). This sequence is divided into two sedimentary units: a lower unit (Upper Pleistocene) and an upper unit (Holocene). The Holocene contains abundant deposits that record different occupations including during the Mesolithic, Neolithic, Chalcolithic, Bronze Age, Iron Age, Roman and Medieval periods (Carretero et al., 2008; Pérez-Romero et al., 2016). Here, we study mandibles of Canis from Middle Bronze Age (C14 dates from 3330 ± 70 a 3560 ± 50 BP, 1440-2030 calibrated BC and 3490 ± 40 to 3680 ± 40 BP, from 1910 to 2190 cal. BC) and Early Bronze Age (C14 dates from 3630 ± 40 to 3760 ± 40 BP, from 1890 to 2290 cal. BC) (Carretero et al., 2008; Pérez-Romero et al., 2016; Galindo-Pellicena et al., 2017).
MATERIAL AND METHODS
The analysed material (Tab. 1) included 17 mandibles from Iberian wolves, 8 modern dogs, and 3 canid remains (Fig. 2A–C) from the El Portalón site. These remains are practically complete.
Traditional morphometry. We follow the measurements of Von den Driesch (1976). We measured 29 variables on mandibles (Fig. 3, Tab. 2) using a digital calliper to the nearest 0.01 mm. We performed a factorial analysis of principal components through STATISTICA (version 12.5). The M1 is a relevant element to distinguish between the dog and its agriotype, the wolf (Clutton-Brock, 1962; Daza-Perea, 2019). We carried out three student-T analyses to examine if a significant difference exists in the length of M1 between modern wolves and dogs and fossils from the El Portalón site. Also, we performed two bivariate analyses to observe the implication between the M1 size (length and width) and another with the mandible length (Infr–Pc) and M1length.
Geometric morphometry. We took photographs of all mandibles (with the same orientation and placement) in lateral view and placed 32 landmarks in different regions (Fig. 4, Tab. 3) using tpsDig 2 (version 2.30: Rohlf, 2017). Landmarks are anatomically homologous points that allow a better analysis of shape (Zelditch et al., 2004). Bookstein (1991) describes three types of landmarks: Type I (classic craniometric points/discrete juxtapositions of tissue, for example, points of contact between bones), Type II (points of maximal curvature) and Type III (extreme points). In this study, most of the digitized landmarks are of Type I and II due to their more conventional biological meaning. We used MorphoJ (version 1.06d; Klingenberg, 2011) to carry out the analysis. We perform a Procrustes superposition to homogenize the data (Rohlf and Marcus, 1993). This consisted of translating, scaling and rotating landmark configurations to eliminate information unrelated to shape (Zelditch et al. 2004). We thus excluded size, location, and orientation differences, and then transformed the original landmarks in Procrustes space, generating new analogous points. We perform a Principal Components Analysis (PCA) to reduce the dimensionality of the data and to simplify its representation. Also, we used wireframe graphs for a better display of the shape changes. The combination of the PCA and wireframe graphs allows us to analyse the variability of the mandibles.
RESULTS
Traditional morphometry. We contribute the basic statistics of the modern and fossil sample (Tab. 4). The Student-T analyses (Tab. 5) between modern dogs and wolves (Tab. 5A) and the analyses between modern wolves and fossils from the El Portalón site (Tab. 5B) show a significant difference in the length of M1. However, the analyses between modern dogs and fossils from the El Portalón site (Tab. 5C) indicate there is not a significant difference in the length of M1. In Figure 5 we supply two bivariate analyses. In Figure 5A, the modern wolves have the greatest length of M1 than modern and fossils dogs. In Figure 5B, the modern sample have a similar mandible size, but the wolves have the longest M1 lengths, and the dogs from the El Portalón site have shorter lengths of mandible and M1.
For the Factorial Analysis, we obtained two factors (Tab. 6) that represent 86.24% of the total variance. The first factor explains 77.42% of the total variance, while the second factor explains 8.82%. The variables (Tab. 7) in Factor 1 are related to measurements that express the length and height of the mandibles whereas the more relevant variables in Factor 2 are measurements related to molar series (M1–M3) and the carnassial length (L M1, Ltg M1, Ltn M1,) and transverse diameter of the canine (ØTC). We carried out a scatterplot with these factors to study the variability of the specimens (Fig. 6). We interpret the positive values of Factor 1 as longer and larger mandibles. The positive values (modern wolves) of Factor 2 are interpreted as mandibles with larger molar series (by the greatest size of M1). The results obtained suggest that modern specimens are distinguished by a greater mandible size. Also, modern dogs are situated in negative values from Factor 2, with a lesser length of molar series (by the lesser size of M1). On the contrary, modern wolves have the greatest molar series and M1 lengths. We selected mandibles of modern dogs following a size criterion (similar size and shape to wolves). Even so, there is a difference between the M1 size from modern wolves and dogs (Tab. 5, Fig. 5). The remains from El Portalón are distinguished by a smaller mandible size comparable to modern dogs and a smaller M1 length comparable to modern samples.
Geometric morphometry. We analysed the first four Principal Components (PC) (Tab. 8) because they show the maximum shape variation (78.28% of the total variance). PC1 accounted for 44.40% of the total variance and is the one that better analysed the shape variation and discriminates between modern wolves and dogs. The modern sample tended to have more positive scores on this component and fossils from the El Portalón site are located in the negative score space. PC1 (Fig. 7) represent a change since (-) shorter mandibles, with a shorter molar series length (by the shorter size of M1) and a longer ascending ramus size to (+) longer mandibles, with a greater M1 size and molar series and a smaller ascending ramus. The negative zone of PC2 (18.85%) reflects the curvature of the mandible and a lesser robust angular process. The shape differences between the wireframes at the positive extreme of PC3 (8.81%) shows shorter dental cusps (paraconid and protoconid) of M1 and a coronoid apophysis oriented to aboral. The positive zone of PC4 (6.21%) describe mandibles with a shorter ascending ramus height, a greater height of the mandible and shorter M1. Figure 8 shows the PCAs of the four first Principal Components analysed. The results suggest that specimens CMI-A6-48-4, CMI-A8-51.1 and CMI-B6-88-1 are distinguished by a slightly longer mandible shape, with greater height and size of ascending ramus and greater molar series length (by the greatest size of M1). The shape of the specimens CMI-A6-48-4 and CMI-B6-88-1 reflect the curvature of the mandible, a greater height in the mandible (zone of M1), a lesser height of the paraconid and protoconid of M1 and a lesser robust of the angular apophysis. Also, their coronoid apophysis is oriented to aboral. On the contrary, the shape of the specimen CMI-A8-51.1 is distinguished by a straighter mandible, a greater height in the mandible (zone of M1), and a lower height of the paraconid and protoconid of M1. Also, their coronoid apophysis is oriented to the aboral.
DISCUSSION AND CONCLUSSION
Considering wolves and dogs are similar morphologically the differences observed in the traditional and geometric techniques are small. Combining both, we obtain a more complete morphometric analysis and a comparison between techniques and their effectiveness in extracting morphological information in the study of the mandible. The results obtained with traditional analysis (Fig. 6) suggest that modern dogs are distinguished by a greater mandible size like modern wolves. This is expected because we selected mandibles of modern dogs following a size criterion (similar size and shape to wolves). Even so, there is a difference in the molar series between modern wolves and dogs due to a greater M1 length (Tab. 5, Fig. 5). M1 is one of the elements more diagnostics for discriminating between dogs and wolves (Clutton-Brock, 1962; Daza-Perea, 2019). The remains from El Portalón are distinguished by a smaller mandible size comparable to modern dogs and a smaller M1 length comparable to modern samples. On the other hand, GM places the remains of the El Portalón site beside modern wolves in PC1 (Fig. 8). This indicates that fossils and modern wolves have similar morphologies of M1, especially in the protoconid.
From traditional morphometry (factors, t-student and bivariate analysis), we can taxonomically classify the fossils of the El Portalón site as being like dogs. However, the remains also conserved diagnostic wolf characteristics (during the Middle-Early Bronze Age) such as the shape of the mandibles being more similar to wolves with respect to the protoconid of M1 and the mandibular region, for example. The shorter length of mandibles from the modern sample within the El Portalón site coincide with the small size of dogs of this period (De Grossi Mazzorin y Tagliacozzo, 1997, 2000; Sanchis y Sarrión, 2004).
Key words: Neurocranium, splanchnocranium, atlas, axis, sacrum, AnthracotHolocene, Morphology, Geometric Morphometry, Dog, Bronze Age.
How to cite: Blázquez-Orta, R., Rodríguez, L., Galindo-Pellicena, M. & García, N. 2022. Morphological study of mandibular Canis remains from the El Portalón (Sierra de Atapuerca) Holocene site. Implications for the origin of the domestic dog in prehistory. [Estudio morfológico de los restos mandibulares del género Canis en el yacimiento holoceno de El Portalón (Sierra de Atapuerca). Implicaciones en la aparición del perro en la prehistoria]. Spanish Journal of Palaeontology, 37(1), 53-68.
Received 2 December 2021, Accepted 14 March 2022, Published online: 6 May 2022
https://doi.org/10.7203/sjp.24124
(In Spanish only)
