An Early Pleistocene hominin mandible from Atapuerca-TD6, Spain

Preservation and Age at Death. The specimen catalogued as ATD6-96 was recovered in a planar section of <1 m², near the area excavated during the 1994–1997 field seasons (1). ATD6-96 is a left half of a gracile mandible of an adult individual (Hominid 7) with the premolars and molars in place (Fig. 1). The specimen is broken at the level of the lateral incisor–canine septum, and the left genial apophysis is preserved. Some post-mortem fractures are observed at the region between the corpus and ramus and near the neck of the condyle, but there is no noticeable distortion. Molar (M)3 is fully erupted and exhibits a minimal wear facet at the mesial marginal ridge. During restoration, the corpus and ramus were separated, and the roots of the M3 were observed directly. These roots are at stage Rc of tooth formation (11). Thus, according to modern human standards of dental development, the age at death of this probably female individual was 17–18 years. However, considering recent evaluations of dental development in H. antecessor (12) an age at death of 15–16 is more convincing.

Description (Figs. 1 and 2). In lateral view, the basal and alveolar margins of ATD6-96 are almost parallel. The height of the corpus decreases slightly from the level of the mental foramina to the level of the lateral prominence (Table 1). As in the specimen ATD6-5 (13) the mandibular body of ATD6-96 is gracile, and exhibits a smooth lateral prominence, that is placed at the level of M2. The swelling of the lateral prominence is greater at the middle area of the corpus, equidistant from the alveolar and basal margins. The posterior part of the lateral prominence is separated from the alveolar margin by a sulcus extramolaris. The sulcus extramolaris is 6.0 mm wide at the level of M3. The torus lateralis superior, torus marginalis, and tuberculum marginalis posterior are barely expressed and only perceived by touch. The tuberculum marginalis anterior is absent. Judging by the preserved portion of the mandibular body, it seems that no mentum osseum was developed in ATD6-96. Four mental foramina are present: The bigger one lies below the level of premolar (P)3 and approximately halfway between the alveolar and basal borders. This foramen opened laterally. The more distal foramen is located at the level of the interalveolar P4/M1 septum, 18.8 mm from the alveolar border, and is directed posteroinferiorly. Two smaller foramina lie below the P3/P4 interalveolar septum, 6.2 and 12.8 mm away from the alveolar border. Despite the small size of the M3 (Table 1) as seen in norma lateralis, the anterior border of the ramus rises at the level of the distal half of M3, and there is no retromolar space. The trigonum postmolare is oblique and measures 7.0 mm. The trigonum postmolare is clearly delimited by the crista pharyngea (or endoalveolaris) and by a less marked crista buccinatorix. The maximum width of the buccinator gutter (the insertion area of the anterior fasciculi of musculus temporalis) is 12.5 mm.

The masseteric fossa is shallow, and the gonion profile is regular, following a smooth curve. The incisura semilunaris is 12.9 mm deep and describes an asymmetric trajectory, the posterior slope being longer and gentler than the anterior slope. The lowest point of the incisura semilunaris is located closer to the coronoid process. The crest defining the mandibular notch meets the condyloid process in the lateral part of the articular surface. The condyloid process is 22.3 mm long and 10.0 mm wide. In posterior view, this process shows a gentle medial lateral convexity. The neck is robust, and the fossa subcondylea is very broad but not especially deep. The posterior border of the ramus is medially flexed because of the action of the sphenoid mandibular ligament.

ATD6-96 lacks an alveolar prominence and a torus mandibularis. As in ATD6-5, the mylohyoid line is defined by a step between the flat and vertical subalveolar plane and a moderately hollowed subalveolar fossa. The mylohyoid line runs from just below the trigonum postmolare forwards and downwards to disappear near the plane of M1. At the M3 level, this line is ≈9.0 mm below the alveolar margin and finishes 15.0 mm below this margin. At the canine level, the planum alveolare is almost vertical. At the level of M3, a smooth step of the basal margin marks the transition from the corpus to the ramus and the limits between the subalveolar fossa and the pterygoid fossa.

The mylohyoid groove is 20.0 mm long. A thin bony sheet bridges ≈4.5 mm of its middle section. The mesial end of this groove lies at the level of the trigonum postmolare, 16.4 mm from the alveolar margin and 10.2 mm from the basal margin. The mylohyoid groove forms an angle of 42° with the alveolar margin. A lingula mandibulae is present.

A photographic restoration of the entire ATD6-96 mandible (Fig. 3) shows an index of the alveolar arcade of ≈100, thus clearly surpassing the highest values obtained in the Middle Pleistocene hominin sample and the Neandertals (14). The pterygoid fossa is shallow. Marked muscular impressions for the attachment of the pterygoid muscle are present at the angulus mandibulae. The medial pterygoid tubercle is well developed.

The posterior teeth are relatively small in comparison to those of Hominid 1 from TD6, the holotype of H. antecessor (15). ATD6-96 shows the P3 > P4 and M1 < M2 size sequences. The crown of P3 is asymmetrical and exhibits a vestige of buccal cingulum. P4 shows a moderately developed talonid. As seen in a computerized tomography scan, both premolars have a single root (Fig. 4), the one of P3 showing a bifurcation (Tomes' root) of the lower half, with buccal and lingual components. The upper half of the root of the third molar is fused and exhibits an expanded single canal (mesotaurodontism).

As additional evidence of cannibalism in the Aurora Stratum (16), ATD6-96 shows seven clear, subparallel cut marks on the external aspect of the coronoid apophysis, probably for detaching the mandible from the skull.

Before comparing the dimensions of the TD6 mandibles with those of other specimens, it is necessary to note that judging by the remarkable size difference between the teeth of ATD6-5 (Hominid 1) and ATD6-96 (Hominid 7), the former was probably a male, whereas the latter mandible belonged to a female. The dimensions of the upper canine of Hominid 1 are at the high end of the range of the genus Homo, which supports the sex status of this individual. On the other hand, the mandible ATD6-5 belonged to an adolescent whose M3 was at the gingival stage of eruption. According to the Rosas and Bermúdez de Castro estimates (13), this mandible could have reached a height at M1 of ≈30 mm, whereas its corpus thickness would have remained similar. In H. sapiens, adult corpus thickness is reached at the time of M2 eruption, which means that in H. antecessor both males and females probably had small mandibles (low and narrow) in comparison with Homo ergaster (East African Early and Middle Pleistocene specimens), Homo mauritanicus, and Homo erectus (Table 2). Especially high are the mandibles from the North African sites of Tighenif and Sidi Abderrahman. Homo heidelbergensis males also have mandibles higher than those of H. antecessor, whereas only the height of the Chinese Middle Pleistocene specimens (mean value) is similar to that of the TD6 mandibles. The width of the Chinese mandibles from Zhoukoudian, Lantian, or Hexian is also comparable with that of the TD6 mandibles, and H. antecessor and Homo pekinensis share a similar robustness index.

H. antecessor exhibits a primitive structural pattern in the mandible that is shared with all African and Asian Homo species (Tables 3 and 4). Some features of the TD6 mandibles, such as the position of the M3 in relation to the ramus, the inclination of the retromolar area, or the gonion profile, are slightly derived from the primitive status, as is also seen in some specimens assigned to H. ergaster, H. mauritanicus, H. pekinensis, and H. heidelbergensis. From the taxonomic and phylogenetic points of view, these traits are only useful in distinguishing H. antecessor from the Java specimens (H. erectus in our narrow vision of this taxon), which maintained a primitive pattern in the mandible. The low angle of inclination of the mylohyoid groove in Sangiran 9 might be an exception. The interpretation of this feature is, however, difficult. In the TD6 mandibles, the angle is small (35° in ATD6-5), thus differentiating these individuals from most hominins except for the Sangiran specimen and the immature individual KNM-WT 15000 (13).

The absence of an alveolar prominence, the almost vertical inclination of the planum alveolare, and the moderately hollowed posterior subalveolar fossa are signs of the gracility of the TD6 mandibles, contrasting with the robustness of the African specimens and some European mandibles, such as Mauer and Arago 13.

The TD6 and Chinese mandibles show some minor differences, which might disappear with more representative samples. The lateral prominence is stronger in the Zhoukoudian specimens, which also exhibit a mandibular angle that is clearly everted and forms a prominent crest, which is especially marked in the male mandible GI (17). In the Lantian mandible, the torus lateralis superior and torus marginalis are more conspicuous, and the alveolar prominence is slightly developed (17). It is also interesting to mention the presence of multiple foramina in the Chinese specimens, as in ATD6-96. Nevertheless, this feature is also present in specimens like Mauer or the Tighenif mandibles (18). As in ATD6-96, the planum alveolare is much less conspicuous in the Zhoukoudian and Lantian mandibles than in the Early and Middle Pleistocene European and African specimens.

The differences in dimensions and robustness between the TD6 mandibles and the East and North African mandibles cast doubt on the African origin of H. antecessor. In contrast, our comparative analysis suggests looking toward the Asian continent. In this respect, it is relevant to mention some data that remained unpublished in 1997, when the new species was named (10), and that are relevant to this discussion. The partial cranium Nanjing I, recovered in 1993–1994 from the Hulu Cave (Tangshan Hill, eastern central China), shows clear modern midfacial traits similar to those observed in the specimen ATD6-69 (19). Wang and Tobias (20) also found similarities between Nanjing I and the Zhoukoudian hominins. Geochronological dates, combined with ecological and paleoclimatic evidence, indicate that the Nanjing skull is ≈600 thousand years old (21). Furthermore, the Locality 1 levels at Zhoukoudian, which yielded most hominin specimens, are now considered at least 800 thousand years old (22). Thus, these Chinese hominins may be contemporaneous with or slightly younger than the TD6 hominins. If the Gran Dolina and Chinese populations are phylogenetically related, they should share a common ancestor that also had a modern midfacial pattern and a gracile mandible. In the cranium, this hypothetical common ancestor would have had a low and flat temporal squama, and an unfused styloid process. These traits would have been retained in the Asian hominins but lost in the TD6 hominins, who exhibit a fused styloid process, a convex temporal squama, and probably a significant increase in cranial capacity (19). The Ceprano calvaria (Italy), which has been tentatively assigned to H. antecessor (23), exhibits a convex temporal squama and a cranial capacity of about 1,057 ml (24). Interestingly, TD6 and Zhoukoudian are the only hominins that have a zygomaxillary tubercle before the Upper Pleistocene (19).

On the other hand, none of the mandibular features considered apomorphic of the European Middle and Early Upper Pleistocene hominin lineage (25) are present in ATD6-96. This conclusion is consistent with observations made in the TD6 cranial and dental remains (19, 26) and suggests that the TD6 hominins do not belong to this lineage. Some H. heidelbergensis specimens, such as Mauer and Arago 13, which exhibit some European apomorphic traits, also retained great robustness, and a conspicuous planum alveolare, which forms a small angle (generally <50°) with the alveolar plane (27). Thus, a probable African origin for the European Middle Pleistocene populations is the most reasonable hypothesis (28).

ATD6-96 was found in association with some lithic objects, whose technical characteristics (Mode 1) are similar to those of the tools found during the 1994–1996 field seasons in the Aurora Stratum (1). In contrast, the Sima de los Huesos and Trinchera Galería Middle Pleistocene sites, also located in Sierra de Atapuerca, have yielded human remains attributed to H. heidelbergensis (29) and associated with Mode 2. Thus, it is probable that the early hominin population of Europe was replaced or genetically absorbed during the Middle Pleistocene by another population, probably coming from Africa, that carried the Mode 2 technology and is well represented by specimens, such as Arago, Atapuerca-Sima de los Huesos, Boxgrove, Mauer, or Petralona (30, 31).

We have made a comparative study of the mandible ATD6-96 recovered in 2003 from the Aurora Stratum, TD6 level, of the Gran Dolina site in Sierra de Atapuerca. The size and shape of this mandible reinforce the taxonomic identity of H. antecessor and are consistent with the hypothesis that this species is closely related to the lineage leading to H. sapiens (10). However, the present evidence does not support the hypothesis of a phylogenetic relationship between the TD6 hominins and the European lineage leading to the Neandertals. On the other hand, observations on TD6 hominins and some Pleistocene Chinese specimens, like Nanjing I, suggest the hypothesis that the Gran Dolina hominins have an Eurasian origin, although the relationships between H. antecessor and other Eurasian populations or species cannot be determined on present evidence. We consider that more information on the Gran Dolina and other contemporaneous hominins will be necessary before revising the phylogenetic position of H. antecessor.

Author contributions: E.C., J.M.B.d.C., J.L.A., E.A., M.B., A.B., I.C., T.C., J.C.D., J.v.d.M., M.M., A.O., A.P.-G., J. Rodríguez, X.P.R., A.R., J. Rosell, R.S., J.V., and J.M.V. designed research, performed research, contributed new reagents/analytic tools, analyzed data, and wrote the paper.

Abbreviations: Mn, molar n;Pn, premolar n.

We thank Giorgio Manzi and Ian Tattersall for helping to improve the quality of the manuscript, Lucía Gomez-Polín for the cleaning and restoration of ATD6-96, and Ana Muela and Susana Sarmiento for photography and image computer processing. The computerized tomography scan images were obtained in the Sant Joan Hospital from Reus (Tarragona, Spain). The Atapuerca research project is supported by Ministerio de Educación y Ciencia Coordinated Projects BOS2003-08938-C03-01, -02, and -03, the Consejería de Cultura y Turismo of the Junta de Castilla y León, and the Fundación Atapuerca.