Dental remains from Atapuerca

Jose´ Maria´ Bermu´dez de Castro, Antonio Rosas & Marı´a Elena Nicola´s Museo Nacional de Ciencias Naturales, C.S.I.C., Jose´ Gutie´rrez Abascal 2, 28006 Madrid, Spain Received 20 December 1997 Revision received 16 March 1999 and accepted 25 April 1999 Keywords: Gran Dolina, Atapuerca, Lower Pleistocene, teeth Dental remains from Atapuerca-TD6 (Gran Dolina site, Burgos, Spain) A descriptive and comparative study of the human dental remains recovered from level six (Aurora stratum) of the Gran Dolina (TD) site, Sierra de Atapuerca, northern Spain, is presented The TD6 dental sample consists of two deciduous and 28 permanent teeth, belonging to a minimum of six individuals Paleomagnetic analyses of the TD sequence suggest an age for the Aurora stratum of greater than 780 ka The associated fossil macro- and micrommals are consistent with a date for the Atapuerca-TD6 hominids at the end of the Early Pleistocene As a whole, the size of the TD6 teeth is large and comparable to that of the African Lower Pleistocene hominids Most morphological dental traits are plesiomorphic for the genus Homo, whereas the shape analysis has revealed some size-related trends that differ from those seen in early Homo Thus, the relative size increase of the maxillary and mandibular anterior teeth and buccolingual diameter of the first molars, the reduction of the third molars and lower canines, and the P3 >P4 size sequence are all characteristic of the population represented by the TD6 hominids The morphological evidence suggests that the TD6 hominids are closer to African Lower and early Middle Pleistocene hominids than they are to European Middle Pleistocene hominids However, some dental traits of the TD6 hominids, such as the expansion of the anterior teeth, the P3 >P4 size sequence, and the morphology of the lower canine also suggest an evolutionary continuity between the population represented by these hominids and the European Middle Pleistocene population Furthermore, dental evidence indicates that the TD6 hominids are clearly distinct from Asian H erectus Dental evidence also suggests that the TD6 hominids could represent an evolutionary connection between H ergaster and H heidelbergensis  1999 Academic Press Journal of Human Evolution (1999) 37, 523–566 Article No jhev.1999.0323 Available online at http://www.idealibrary.com on Introduction The Sierra de Atapuerca is situated about 010 E and 4220 N It is near the Arlanzo´n river and 14 km east from the city of Burgos, northern Spain A disused railway traverses into the southwestern slope of the Sierra de Atapuerca, where some fissure-fill Pleistocene deposits of three main karstic cavities are exposed in a cutting (Carbonell et al., 1998) One of these deposits, Gran Dolina (TD), has outcrops up to 18 m deep The TD site fills the cavity catalogued as BU-IVA-16 (Martin et al., 1981) A total of 11 0047–2484/99/090523+44$30.00/0 lithological levels, numbered 1–11 from bottom to top, can be distinguished in TD (see Pare´s & Pe´rez-Gonza´lez, 1995) Excavation of a m3 (2·52·5) planar section was begun in 1993 In July of 1994, excavation reached level (TD6), and one of the TD6 strata, the so-called Aurora stratum, yielded a rich faunal and lithic assemblage (Carbonell et al., 1995) Included in this assemblage were more than 30 human fossil remains The top of the Aurora stratum is about m below the Matuyama/Brunhes boundary (Pare´s & Pe´rez-Gonza´lez, 1995) Excavation of the Aurora stratum (6 m2)  1999 Academic Press   ´    524 Table ET AL List of the mandibular, maxillary and dental remains from the TD6 Aurora stratum Inventory number ATD6-5 ATD6-13 ATD6-1 ATD6-2 ATD6-3 ATD6-4 ATD6-6 ATD6-7 ATD6-8 ATD6-9 ATD6-10 ATD6-11 ATD6-12 ATD6-14 ATD6-69 ATD6-48 ATD6-52 ATD6-312 Specimen Hominid Right mandible fragment (M1–M3)  Left maxillary fragment (C–P3)  Left lower C  Left I2  Right P3  Right P4  Fragment of crown of right lower C  Right P3  Right P  Left P4  Right M1  Left M  Right M2  Left maxillary fragment (dc–dm1) Maxilla and left zygomatic bone (left P3, M1, and unerupted M2, M3 and right I2–M1) Left I2 Left I1 Left I2 germ finished in 1996 After sorting the fossils recovered from this stratum, a total of 85 human remains were identified The TD6 human sample includes postcranial (Carretero et al., 1999; Lorenzo et al., 1999), facial and neurocranial (Arsuaga et al., 1999), mandibular (Rosas & Bermu´dez de Castro, 1999), and dental specimens Because of a unique combination of cranial, dental, and mandibular traits in the current TD6 hypodigm we have named a new Homo species, H antecessor, to include these human fossil remains (Bermu´dez de Castro et al., 1997) In this paper it was suggested that the new species could represent the last common ancestor of the Neandertals and modern humans That is, H antecessor would have an Afro-European distribution, and would be the origin of two evolutionary lineages, one of them exclusively European and formed by two chronospecies, H heidelbergensis and H neanderthalensis, whereas the other led to H sapiens, probably in the African continent Previous information concerning dental H1 H2 H3 H4 H5 H6 remains from TD6 were published in Carbonell et al (1995) and Bermu´dez de Castro et al (1997) The aim of this paper is to present a detailed morphological and metrical comparative analysis of these dental remains A discussion of the phylogenetic relationships of the TD6 hominids with regard to the information offered by the dental remains is also presented TD6 dental remains The dental sample from the Aurora stratum from TD6 belongs to a minimum of six individuals (Table 1) The Hominid (H1) is represented by a fragment of the right side of a mandible (Rosas & Bermu´dez de Castro, 1999) with M1, M2, and unerupted M3, a small and poorly preserved maxillary fragment with the left C and P3, as well as eleven isolated teeth Hominid (H2) is represented by part of a left infant maxilla, retaining dc–dm1 The radiographic image of this specimen shows the unerupted crowns of permanent teeth A well preserved subadult maxilla, with the left frontal      process and the left zygomatic bone, forms Hominid (H3) This specimen also preserves in situ the left P3, M1, and unerupted M2 and M3, as well as the right I2, C, P3, P4, and M1 Two moderately worn left I1 and I2, which not fit to each other, represent the evidence of two additional adult dentitions in the TD6 human sample (H4 and H5) Finally, the germ of a left I2 is evidence of another infant individual (H6) Details of the pattern of dental development and an assessment of the physiological age of the subadult hominids from TD6 will form the subject of a future paper Methods We provide in this paper descriptions of the morphology of the TD6 teeth, indicating the main morphological traits of each specimen In order to facilitate the understanding of these descriptions, photographs of the teeth are also included The mesiodistal (MD) and buccolingual (BL) dimensions of the Atapuerca-TD6 teeth were recorded to the nearest 0·1 mm following the technique proposed by Flechier (1975) A caliper with wide, flat and pointed tips, which allowed insertion between teeth still in situ, was used The computed crown area (CCA=MDBL) was calculated from these tooth dimensions When the root (anterior teeth), or at least one element of the root complex (postcanine teeth), was complete, the maximum root length (RL) was also recorded, following the technique described in Bermu´dez de Castro (1988) The surface area of the crowns of the TD6 premolars and molars, as well as the areas of the major cusps of these teeth, were measured from occlusal photographs Each specimen was oriented so that the occlusal plane was perpendicular to the optical axis of a camera fitted up with a Tamron 90 mm F 2·5 lens and a Tamron extension tube 18F In order to get a maximum depth of field, an aperture of f/32 was set The 525 magnification ratio was adjusted to 1:1, and a scale was included in each photograph and placed parallel to the occlusal plane The prints were made at 5 magnification, which proved to be a suitable magnification for using a planimeter (Ushikata X-Plan 360d) The boundaries of the crown, and that of each individual cusp (following the criteria of Wood et al., 1983) were marked with ink on the photographs Following Wood & Abbott (1983), the original mesial and/or distal borders of the teeth with interproximal wear were estimated by reference to the overall crown shape and the buccolingual extent of the wear facets Before obtaining any measurement, the exact enlargement of each print was checked using the planimeter Tests showed intraobserver measurement error to be less than 1% The overall crown size (measured crown area: MCA) was taken to be the sum of the areas of the four or five major cusps) For comparative purposes we have used the MD and BL dimensions of the Homo specimens and samples listed in Table Furthermore, references to these metric data in other hominids included in the genus Australopithecus and Paranthropus were obtained from Johanson & White (1979), Johanson et al (1982), White et al (1994), and Blumenberg & Lloyd (1983) We have also used crown and cusp area data published by Wood et al (1983), Wood & Uytterschaut (1987), Wood & Engelman (1988), and Bromage et al (1995) Apart from the new species represented by the Gran Dolina hominids, we recognize the following Homo species: H rudolfensis, H habilis (s.s), H ergaster, H erectus (s.s) (sensu Wood, 1992), H heidelbergensis (the European Middle Pleistocene hominids), H neanderthalensis, and H sapiens We refer to the African Middle Pleistocene hominids, such as those of Tighenif, Rabat, Sidi Abderrahaman, Thomas I, and Olduvai (OH 22), simply as AFMP, thus avoiding for the moment the commitment of formally 526 Table   ´    ET AL List of the Homo specimens and samples whose mesiodistal and buccolingual tooth measurements are used for comparison with the Atapuerca TD6 dental sample OH 6, 7, 13, 16, 24 and 39 OH 22 UR 501 KNM-WT 15000 SK 27 KNM-ER 820, 992 KNM-ER 803, 808, 1590, 1805, 1813 KNM-ER 3733 KNM-BK 67 and BK 8518 Dmanisi Sangiran S7 Sangiran and B Sangiran 17 Zhokoudian Tighenif 1, and Rabat Casablanca Thomas I Mauer Atapuerca-SH Arago 2, 13, 21, 26, 36, 59 and 60 Petralona Montamaurin (isolated teeth) Montmaurin (mandible) La Chaise (Suard Cave) Lazaret Neandertals from Western and Central Europe and Near East assigning these hominids to a Homo species The state of some of the dental features analysed in this study was noted for each one of these species/groups Furthermore, the state of the dental features was also noted for A afarensis This species was thus used as a reference group, or outgroup in cladistic terms In order to characterize the shape of each specimen and the samples, we calculated the following indices: the ratio of computed crown areas of the anterior and posterior premolars [(CCA) P3/P4 and (CCA) P3/P4], the first and second molars [(CCA) M1/M2 and (CCA) M1/M2], and the lateral incisors and first molar [BL I2/(CCA) M1 and BL I2/(CCA) M1] Tobias (1991) Rightmire (1980) Bromage et al (1995) Brown & Walker (1993) Wolpoff (1971) Leakey & Wood (1973) Leakey et al (1978) Leakey & Walker (1985) Wood & Van Noten (1986) Gabunia & Vekua (1995) Grine & Franzen (1994) Wolpoff (1971) Jacob (1973) Wolpoff (1971) Authors Thoma & Vallois (1977) Howell (1960) Sausse (1975) Howell (1960) Authors Authors Authors, in a cast Billy (1982) Authors Genet Varcin (1975, 1976) Authors This sample includes the specimens listed in Table of Bermu´dez de Castro (1986), as well as the specimens of Tabun (Lumley, 1973) and Shanidar (Trinkaus, 1983) The first molar was used as the reference because this tooth is the most stable of the molar dental field (Dahlberg, 1945) The Crown Shape Index (CI=BL/MD100) was also calculated for the lower and upper teeth We have compared the tooth shape of H1 from TD6 with the following specimens and samples: the Atapuerca-SH (H heidelbergensis), Neandertal (H neanderthalensis), Zhoukoudian (H erectus), and Gran Canaria (H sapiens) samples (which are represented by their corresponding mean values for each variable), as well as the specimens OH (H habilis), KNM-ER 992 and KNM-WT 15000 (H ergaster), Mauer (H heidelbergensis), Tighenif 3, and      Dmanisi In order to analyse the relative proportions of the dimensions of the TD6 teeth in comparison with these specimens and samples these we have processed the data as follows: First, we obtain the value of the expression: TD6 Vi /1/2(TD6 Vi +SiVi) (1) where Vi (i=1 n) is the value of each of the n variables analysed in TD6 and in the Si (i=1 n) specimens or samples In order to include the greatest number of specimens in this analysis, we have used the MD and BL dimensions of the mandibular I2, C, P3, P4, M1, and M2 for the comparison The expression [Equation (1)] represents a size ratio between TD6 and the comparative sample If the value of the variable in TD6 is identical to the value of the variable in one Si, then the expression [Equation (1)] is equal to 1; if the value of the variable in TD6 is greater than the value of the variable in one Si, then the expression [Equation (1)] is >1, and vice versa Then, let us calculate the average of the twelve values obtained by Equation (1) for the comparisons between TD6 and each one of the ten specimens (six) and samples (four): The expression [Equation (2)] is called the ‘‘average dental ratio’’ (ADR) Thus we have obtained ten ADR values If the expression [Equation (2)] is >1, it means that the values of the dimensions of the TD6 are, on average, greater than those of the particular specimen or sample and vice versa This may be illustrated by means of bivariate scattergrams: the ADR obtained for all comparisons (e.g in the X-axis) vs the value of the expression [Equation (1)] (Y-axis) for each variable 527 Obviously, the expressions [Equations (1) and (2)] take the value when Si =TD6; that is to say, when we compare TD6 with itself In this case, the point representing TD6 will be on the line X=Y Regarding the comparison between TD6 and the other specimens and samples, the greater the difference between the expressions [Equations (1) and (2)] for a certain variable Vi, the greater the distance of the specimen or sample Si to the line X=Y for this variable In the bivariate scattergram obtained for any of the Vi variables, when a specimen or sample is located on the upper part defined by the X=Y line it implies that this specimen, or sample, shows a proportionality lesser value than TD6 for the variable at issue and, vice versa, when a specimen, or sample, is placed on the lower part as defined by the X=Y line, it implies that this specimen or sample shows a proportionally higher value than TD6 for the variable considered The variance of the distribution of all TD6 Vi((TD6 Vi +Si Vi)1/2) values represents the so-called WF distance (Ö) between TD6 and one of the specimens or samples (e.g SA) In order to make a global assessment of the tooth shape using the information available from most specimens and samples, and to assess phenetic affinities among all these specimens and samples, we have obtained the shape distance, first defined in Bermu´dez de Castro (1993), as follows: where i=1 n, and SA Vi and BB Vi are the values of the variable Vi in samples, or specimens, SA and SB respectively, and where Ö=0 expresses the identity in shape 528   ´    ET AL      529 Figure Buccal aspect of the deciduous canine and first molar of the maxillary fragment ATD6-14, belonging to hominid from TD6 Scale bar 1·5 cm between SA and SB From the dissimilarity matrix formed using this expression, cluster analyses (UPGMA method) were carried out These analyses were performed using the NTSYS program (Rohlf, 1992) In order to test the goodness of fit of the cluster analyses to the data, the MXC OBLIQUAMP program (Rohlf, 1992) was used to compare the cophenetic matrix and the original matrix that was clustered Furthermore, in order to assess phenetic similarities between the TD6 hominids and other fossils, a principal component analysis (PCA) on 45 specimens (including 25 modern human individuals) was also made using 12 dental variables: the MD and BL of I2–M2 (raw data) The Stratgraphics statistical package (1986) was employed to make this analysis Descriptions Upper deciduous canine The condition of the dc of ATD6-14 is excellent (Figures and 2) Wear has blunted the pointed apex of the crown, but no dentine is exposed Wear also affects the lingual face, flattening the mesial marginal ridge and the central ridge A small mesial interproximal wear facet is present The occlusal outline is a MD-elongated asymmetrical ovoid Near the mesial and distal corners, the occlusal edges exhibit a deep notch, which delimits the main cusp from well-differentiated mesial and distal cusplets or styles In labial view the crown shows a pentagonal shape The apical half of the labial face is strongly convex MD This surface also shows a marked lingual Figure Occlusal view of: (a) ATD6-8, right P4; (b) C and P3 of the maxillary fragment ATD6-13; (c) ATD6-9, left P4; (d) ATD6-7, right P3; (e) ATD6-11, left M1; (f) ATD6-10, right M1; (g) ATD6-12, right M2; (h) ATD6-3, right P3; (i) ATD6-4, right P4 All these teeth belong to Hominid from TD6 (j) dc and dm1 of the maxillary fragment ATD6-14 (Hominid 2) Approximate enlargement 2.85 530   ´    inclination, and is bounded by mesial and distal depressed areas and grooves, which emphasize the independence of the mesial and distal styles The central area of the labial face is crossed by three short vertical grooves, whereas the cervical third of this face exhibits a strong basal prominence Viewed lingually, the styles are bounded by short mesial and distal grooves The broad mesial and distal marginal ridges extend from the apex to the strong basal swelling, but no lingual tubercle is present The lingual face also shows a broad central ridge, which is delimited by mesial and distal longitudinal grooves This face is not shovel-shaped Upper deciduous first molar The crown of the dm1 of ATD6-14 is wellpreserved (Figures and 2) Occlusal wear has smoothed the topography of the lingual slopes of the paracone and metacone, as well as the buccal slope of the protocone There is no evidence of enamel crenulations or secondary fissures There are point dentine exposures on the protocone and the mesiostyle The occlusal outline is a rounded rectangle with a pronounced MB extension, due to the presence of a strong and bulbous molar tubercle The protocone is the largest cusp, followed by the paracone and metacone The hypocone is absent The mesial marginal ridge is strongly developed, and forms a mesiostyle at the buccal half of the occlusal face The mesiostyle is separated from the paracone by a long and broad fovea anterior The crista obliqua is well developed, and not interrupted by the median longitudinal groove The fovea posterior is located at the intersection of the distal part of the median longitudinal groove and a long BL groove, which represents the boundary between the protocone and metacone and a well developed distal marginal ridge The median longitudinal groove also extends on to the distal face In buccal view, a short ET AL groove, which extends below the occlusal rim, delimits the distal slope of the paracone and the mesiostyle The lingual face exhibits some longitudinal shallow grooves and a strong basal prominence In distal view, the molar tubercle sticks out as marked basal prominence Upper lateral incisor The crown of the right I2 of ATD6-69 is in good condition (Figure 3) Incisal wear has produced a slightly convexity of the incisal edge, and exposed a very narrow dentine strip, which is about 3·0 mm long The labial face shows a gentle convexity from the cervical line to the incisal rim In lateral view, the labial face is inclined lingually In lingual view, three clear longitudinal grooves are observed The shallow distal groove represents the limit between the conspicuous distal marginal ridge and a well developed central–distal ridge The distal marginal ridge begins at the occlusal–distal corner and expands cervically to form the lingual basal prominence A central and deeper groove delimits the central–distal ridge and a less prominent central–mesial ridge This groove ends at the basal prominence and forms an incipient lingual fovea The two central ridges end about 2·4 from the incisal edge Finally, a mesial groove delimits the central–mesial ridge and the well developed mesial marginal ridge A notch breaks the continuity between the mesial marginal ridge and the basal prominence The area between the central ridges and the incisal edge is concave, and this tooth exhibits a characteristic shovel shape Some gentle cervical-occlusally directed enamel folds complicate the topography of the labial surface Upper canines The left C of H1 and H3 are in good condition (Figures 1, and 4) The H3 C did not reach the occlusal plane, and the distance from the labial alveolar border to      531 Figure Occlusal aspect of the lateral incisor, canine, third and fourth premolars and first molar of the maxilla ATD6-69, belonging (Hominid 3) from TD6 Scale bar 19 mm the tip of the crown measures 8·5 mm Occlusal wear has rounded the crown of the C of the H1 and exposed a small diamondshaped dentine patch about 2·4 mm long The occlusal outline is an ovoid Viewed labially, the crown is asymmetrical in both the angle of the slope and the length of the mesial and distal occlusal edges, the distal being longer and more inclined than the mesial The upper half of the labial face shows several grooves and gentle enamel folds, which run occlusal–cervically and give a ridged aspect to this face The C of the H1 exhibits a mild hypoplastic line, which is located 4·4 from the buccal cervical line The lingual face is marked by the mesial and distal marginal ridges, the mesial one being more strongly developed Both marginal ridges flow into the swollen lingual basal prominence A central longitudinal groove delimits the broad and shallow mesialcentral and distal-central ridges, which barely jut out from the flattened lingual surface The root of the H1 C is MD compressed, and exhibits broad mesial and distal longitudinal grooves These grooves differentiate the labial and lingual components of the root, the labial being dominant Upper third premolars The left and right P3s of H1 and H3 are in good condition (Figures 1, and 4) The apices of the buccal and lingual roots of the H1 left P3 are missing The buccal root of this specimen was accidentally broken, but it has been restored The H3 left P3 shows slight occlusal wear on the summit and distal edge of the protocone, as well as distinct wear facets on the summit, mesial and distal edges, and lingual slope of the paracone In the right P3 of this individual, occlusal wear has created facets only on the summits of the paracone and protocone, as well as on the mesial edge of the paracone Furthermore, the left P3 exhibits an incipient interproximal wear facet on the mesial and distal aspects of the 532   ´    ET AL Figure (a) Mesial view of the right M2 ATD6-12; (b) buccal aspect of C and P3 of the maxillary fragment ATD6-13; (c) distal view of the right P3 ATD6-7; (d) mesial view of the left P4 ATD6-9; (e) mesial view of the right P4 ATD6-8; (f) distal view of the left M1 ATD6-11; (g) distal view of the right M1 ATD6-10 Scale bar cm ... these dental remains A discussion of the phylogenetic relationships of the TD6 hominids with regard to the information offered by the dental remains is also presented TD6 dental remains The dental. .. to H sapiens, probably in the African continent Previous information concerning dental H1 H2 H3 H4 H5 H6 remains from TD6 were published in Carbonell et al (1995) and Bermu´dez de Castro et al...  ´    524 Table ET AL List of the mandibular, maxillary and dental remains from the TD6 Aurora stratum Inventory number ATD6-5 ATD6-13 ATD6-1 ATD6-2 ATD6-3 ATD6-4