Anthropological Review • Vol. 81(3), 225–251 (2018)

Anthropological Review Available online at: https://content.sciendo.com/anre Journal homepage: www.ptantropologiczne.pl

The affinities of antecessor – a review of craniofacial features and their taxonomic validity Francesc Ribot Trafí1, Mario García Bartual2, Qian Wang3

1Museo Municipal de Prehistoria y Paleontología Dr. Gibert, Granada, 2Museo Paleontológico de Elche and Fundación Cidaris, Alicante, Spain 3Department of Biomedical Sciences, Texas A&M University College of Dentistry, Dallas, U.S.A.

Abstract: The phylogenetic affinities of , a hominin dating from the early Middle of , are still unclear. In this study we conducted a comprehensive review of the TD6 hypodigm within the context of the historical development of paleoanthropological issues concerning this . H. antecessor, based on all available craniofacial features to date, displays a midfacial morphology very similar to specimens attributed to Classic , suggesting that H. antecessor is the geographical European variant of Classic H. erectus.

Key words: Atapuerca, Early Homo, Homo erectus, Homo antecessor, midfacial morphology, Middle Pleistocene

Introduction accepted, as some authors have doubts about its validity as a unique taxon (eg, In 1997, Bermúdez de Castro and col- Delson 1997; Wang 1998; Wang and leagues defined a new species, Homo an- Tobias 2000a; Rightmire 2001, 2007; tecessor (Bermúdez de Castro et al. 1997), Stringer 2002; Etler 2004; Rabadà 2005; based on craniofacial and dental remains Ribot et al. 2006, 2007, 2017, in press of ATD6-69 specimen, an immature in- 2018; Cartmill and Smith 2009; Ribot dividual found at the TD6 level of Gran and García Bartual 2016). Dolina (Atapuerca, Burgos, Spain) and One of the main problems for the dated between 0.78 Ma (Carbonell et al. acceptance of this new species 1995) and 0.9 Ma (Parés et al. 2013).The is its diagnosis, which was done on the facial morphology features of ATD6-69 basis of a subadult individual. Rightmire were the hallmarks that allowed Bermú- (1998a) noted that the attribution of a dez de Castro and colleagues to define new species, based solely on the facial this new species, and also place it as a morphology of an immature specimen is common ancestor of Neandertals and problematic, and also questioned wheth- H. sapiens (see “Differential diagnosis of er the presence of a true canine fossa in Homo antecessor”). However, the phyloge- a young individual —one of the diagnos- netic status of H. antecessor is not widely tic characteristics for this taxon— is an

Original Research Article Received January 14, 2018 Revised June 21, 2018 Accepted June 25, 2018 DOI: 10.2478/anre-2018-0020 © 2018 Polish Anthropological Society 226 Francesc Ribot Trafí, et al. appropriate morphological for the characteristics, in the study of Carbonell diagnosis of a new species. In fact, the et al. (2005) they compare 14 mandibular zygomaxillary fragment from Atapuerca, characters from the ATD6-96 ATD6-58 (attributed to an adult individ- with those of different Homo taxa (Table ual), shows a reduced canine fossa, which 4 in Carbonell et al. 2005). The interpre- differs from to the ATD6-69 fragment at- tation of the table shows that most of tributed to an immature individual (Ber- the features present in ATD6-96 are also múdez de Castro et al. 1997; Arsuaga et found in Classic H. erectus (see Taxonomic al. 1999) (Fig. 1). notes). In fact, only one trait can be con- In 2003, a mandible of an adult H. an- sidered as derived, which is the hollowed tecessor was recovered (ATD6-96) (Car- posterior subalveolar fossa. This trait is bonell et al. 2005). From the study of this moderately hollowed in ATD6-96 and in , the authors concluded that: “[…] some H. heidelbergensis, shallow in almost none of the mandibular characteristics consid- all Classic H. erectus individuals, deep in ered apomorphic in the European Middle Pleis- Classic H. erectus from Zhoukoudian, and tocene and Upper Primitive Homo lineage are subtle in H. neanderthalensis and some H. present in ATD6-96. This evidence reinforces heidelbergensis. However, the distinction the taxonomic identity of H. antecessor and drawn by Carbonell et al. (2005) between is consistent with the hypothesis of a close rela- a shallow and moderately hollowed subal- tionship between this species and Homo sapi- veolar fossa, is very subjective and ens” (Carbonell et al. 2005. p. 5674). assigned to either of the two categories While it is not clear if H. antecessor rep- are virtually identical. Moreover, in a re- resents a new taxon based on diagnostic cent review of H. antecessor, Bermúdez de

Fig. 1. Comparison between ATD6-69 —juvenile— (right) and ATD6-58 —adult— (left) Footnote: It can be observed how the canine fossa in ATD6-58 is shallower than that found in ATD6-69. Picture of ATD6-69 courtesy of Roberto Sáez; picture of ARD6-58 taken from Schwartz and Tattersall (2002). Homo antecessor - craniofacial features and their taxonomic validity 227

Castro et al. (2017) did not present any The second one states that H. antecessor derived feature in the H. antecessor mandi- has an Asian origin, in which this tax- ble, with comparison to modern on would be the descendant of H. erectus (see Table 2 in Bermúdez de Castro et al. (Carbonell et al. 2005; Martinón-Torres 2017). et al. 2007), based on the strong similar- It has also been argued that the al- ities of H. antecessor and the Asian Classic leged ancestor-descendant relationship H. erectus in their and midfa- between H. antecessor and H. sapiens, and cial morphology, and the shared presence the recognition of H. antecessor as a new of a zygomaxillary tubercle (Carbonell et taxon, is reinforced by the remodeling al. 2005). Also, these hominins present pattern of the facial bone of ATD6-69, similarities in some dental features, such which shows a strong similarity with that as the “morphological robusticity” of the in H. sapiens (Lacruz et al. 2013). In both anterior dentition (e.g., triangular shovel species, there is a retraction at the sub- shape), and in the posterior dentition, by nasal region (orthognathic face) associat- the simplification of occlusal morphol- ed with the presence of resorptive areas ogy, high frequencies of non-Y groove around the nasomaxillary clivus (Lacruz patterns, the presence of the mid-tri- et al. 2013, 2015). This pattern contrasts gonid crest, the anterior fovea and the with that of KNM-WT 15000 —an imma- transverse crest in the (Mar- ture individual of West Turkana Earliest tinón-Torres et al. 2007). H. erectus (or H. ergaster) (see Taxonomic notes), dated to approximately 1.7 Ma—, Differential diagnosis for Homo which shows the primitive facial bone antecessor remodeling pattern seen in australopiths and early Homo, with presence of areas of The erection of the new species, H. an- deposition at the subnasal region (prog- tecessor, and its relation to H. sapiens was nathic face) (Bromage 1989; Lacruz et al. based on a series of characters on ATD6- 2013, 2015). Thus, Lacruz and colleagues 69 (Bermúdez de Castro et al. 1997), (2013) conclude that the H. sapiens-like which were viewed as derived traits with retraction of the midface in ATD6-69 is respect to primitive Homo (H. rudolfensis, a key aspect of H. antecessor in its recog- H. habilis, Earliest H. erectus [or H. ergas- nition as a new taxon. However, the va- ter] and Classic H. erectus) and to the later lidity of this characteristic to establish appearing European Homo (H. heidelber- taxonomic relations is problematic con- gensis and H. neanderthalensis). Special em- sidering the findings of previous studies phasis was put on the ATD6-69 midfacial (Kurihara et al. 1980; McCollum 2008). morphology and its related nasomaxillary Regarding the origin of H. antecessor, functional complex (Bermúdez de Castro the scientists studying the Atapuerca ma- et al. 1997, 2017; Arsuaga et al. 1999), terials have proposed two different op- including: the coronal orientation of the tions: the first one states that this taxon infraorbital surface with inferoposterior had an African origin, in which H. anteces- slope of its plane; presence of canine fos- sor and Classic H. erectus share a common sa; maxillary inflection oriented horizon- ancestor with the Earliest H. erectus (or H. tally and a projection of the nasal bones; ergaster) (Bermúdez de Castro et al. 1997, arched zygomaticoalveolar crest with a 2017; Arsuaga et al. 1999; Rosas 2000). superiorly positioned zygomatic root plus 228 Francesc Ribot Trafí, et al. the presence of an incisura malaris; spinal midfacial and cranial morphological com- and lateral nasal crests nearly fusing to bination found in the ATD6 hominins is form a lower nasal edge; presence of a zy- considered a modern human synapomor- gomaxillary tubercle; modern facial bone phy that first appeared in the fossil record remodeling pattern; and convex superi- with the appearance of H. antecessor (Ta- or border of the temporal squama. This ble 1, Fig. 2).

Table 1. Craniofacial and mandibular supposedly derived characteristics that are part of the definition of Homo antecessor* and which are discussed in the present study (see also Fig. 1) Diagnostic features of ‘H. antecessor’ discussed in this study 1 Coronal orientation of the infraorbital surface with inferoposterior slope of this plane 2 Presence of canine fossa 3 Maxillary inflection horizontally plus projection of the nasal bones 4 Arched zygomaticoalveolar crest with incisura malaris 5 Spinal and lateral nasal crests nearly fusing to form a lower nasal edge 6 Presence of zygomaxillary tubercle 7 Modern facial bone remodeling pattern 8 Convex superior border of the temporal squama

*Bermúdez de Castro et al. 1997; Arsuaga et al. 1999; Carbonell et al. 2005; Bermúdez de Castro et al. 2017.

Fig. 2. ATD6-69 facial characteristics considered exclusively derived with respect to Homo sapiens, and used to define the species Homo antecessor (Bermúdez de Castro et al. 1997, 2017; Arsuaga et al. 1999) Footnote: a) coronal orientation of the infraorbital surface with presence of a true canine fossa; b) horizon- tally arched zygomaticoalveolar crest, with a high zygomatic root and zygomaxillary tubercle; c) maxillary inflection determined by the infraorbital plate and the lateral nasal wall (area marked in green lines); d) spinal and lateral nasal crests that are very close together and almost fuse forming the lower nasal edge; and modern facial bone remodelling pattern: (+) depository fields; (–) resorptive fields. Modified from Schwartz and Tattersall (2002) and Lacruz et al. (2013). Homo antecessor - craniofacial features and their taxonomic validity 229

Material and methods to this replica during several months, and was allowed to study and compare it with Material the pictures and descriptions of ATD6-69 made by the Atapuerca team. We have Materials for this study include casts, im- also used photos and the original descrip- ages, and description of early Homo and tions of the material. penecontemporary Asian Classic H. erec- tus which present midfacial skeleton re- Taxonomic notes mains, partially or complete. In particular we studies the following hominin casts: In this study, the midfacial morpholo- KNM-ER 1470; KNM-ER 1813; OH 24; gy of H. antecessor is compared with that OH 13; KNM-ER 3883; KNM-ER 3733; of the older Homo species (early Homo KNM-WT 15000; Gongwangling; San- and Earliest H. erectus [or H. ergaster]), giran 17; Zhoukoudian (reconstruction and with the penecontemporary species of Weidenreich —1937a— and Sawyer of Classic H. erectus. The taxa used and and Tattersall —1995 [and Tattersall and their representative specimens are as Sawyer, 1996]—). Also, a sample of 35 follows. extant human skulls from the Laboratory of Paleopathology and 1. Early Homo. We considered as early of the National Museum of Archeology Homo those specimens dated between of Catalonia in Barcelona. Regarding the ~2.4 and 1.4 Ma (Table 2). Although study of H. antecessor remains we used a the classification of early Homo is con- high-quality replica of ATD6-69 (not the troversial, in this work we have opted standard cast that is hosted in many lab- for a more classic classification and, ac- oratories), which was brought by the re- cording to Antón et al. (2014) we distin- searchers from the Atapuerca group for guish between the 1470 group and the a special exhibition at the Miquel Crusa- 1813 group (also called H. rudolfensis and font Institute of in Sabadell H. habilis, respectively, by several authors (Barcelona). One of us (F.R.T.) had access —e.g. Wood 1992; Kimbel et al. 1997;

Table 2. Fossil specimens classified here as Early Homo Specimen Age (Ma) Author A.L. 666-1 2.33 Kimbel et al. 1997 1470 group KNM-ER 1470 2.03-2.09 Joordens et al. 2013 KNM-ER 62000 1.95-1.98 Joordens et al. 2013 1813 group OH 24 1.88 Hay 1976 KNM-ER 1813 ~1.86 Feibel et al. 2009 OH 62 1.75-1.85 Johanson et al. 1987 OH 65 ~1.78 Blumenschine et al. 2003 OH 13 1.60 Hay 1976 KNM-ER 42703 1.44 Spoor et al. 2007 230 Francesc Ribot Trafí, et al.

Strait et al. 1997; Prat 2004; Kimbel 2009; others include it among the australopiths: Schrenk et al. 2015; Simpson 2015). Falk (1986) notes that the cortical sulcal In 1470 group, we include the spec- pattern of ER 1805 has affinities with imens KNM-ER 1470 and KNM-ER ; Tobias (1980) allocated 62000. Leakey et al. (2012) include the specimen to boisei on the KNM-ER 1470 and KNM-ER 62000 in basis of morphology; likewise, Prat the same group because of its morpho- (2002, 2004), based on a cladistic analy- logical similarity. sis with 122 cranial characteristics, con- In 1813 group, we include the speci- cludes that ER 1805 should be included in mens OH 24, OH 62, OH 65, KNM-ER the genus Paranthropus. Finally, Grossman 1813 and KNM-ER 42703. (2009) points out that ER 1805 presents The classification of the OH 65 max- neither the typical morphology of Homo illa is controversial: Blumenschine et al. nor that of Paranthropus, and concludes (2003) and Clarke (2012), they group it that the suite of traits presented in KNM- with KNM-ER 1470, and Clarke (2012) ER 1805 are certainly unique. Due to separates KNM-ER 1470 and OH 65 (H. these different interpretations about this habilis) from KNM-ER 1813 and OH 24 fossil, we prefer to exclude KNM-ER 1805 (Australopithecus cf. africanus). However, from this study. the inclusion of KNM-ER 1470 and OH 65 to the same group has been ques- 2. Earliest Homo erectus (or Homo er- tioned: Spoor et al. (2007) point out that gaster). We consider as Earliest H. erectus the morphology of the KNM-ER (or H. ergaster) those hominins that are in 42703 is very close to that of the large a temporal range of approximately 1.8 to specimens of H. habilis (OH 65 and KNM- 1.3 Ma (Table 3) —classified as Earliest ER 1805) and different from that of H. erectus by Antón 2003. Of the spec- KNM-ER 1470. Also, Leakey et al. (2012) imens studied by us, Antón includes in and Antón (2012) separates OH 65 from the group Earliest H. erectus the African KNM-ER 1470 and KNM-ER 62000, be- specimens KNM-ER 3733 and KNM-WT cause OH 65 is much more prognathic 15000, the remains of Dmanisi and San- subnasally and has more posteriorly posi- giran 4 (we added the Asian fossils Gong- tioned zygomatic roots, the nasoalveolar wangling and Sangiran 17, classified by clivus is arched at the alveolar margin, Antón as Early H. erectus for presenting and its canine alveoli are not part of the more recent chronologies than the Ear- anterior tooth row; the project liest H. erectus, however, the last datings beyond the bicanine line and the palate place the remains of Gongwangling at is parabolic (like KNM-ER 1813 and dif- 1.63 Ma —Zhu et al. 2015— and Sangiran ferent from the ‘U-shaped’ in KNM-ER 17 between 1.30 and 1.25 Ma —Larick et 62000). Therefore, we include OH 65 in al. 2001, 2004; Antón et al. 2007). Thus, 1813 group. Antón joins in the same group the speci- Another specimen whose classification mens from East Africa, those of Dmanisi is difficult is KNM-ER 1805. Many authors and Sangiran 4. Likewise, due to the sim- classify it as Early Homo (ex, Wood 1991, ilar morphology, Kaifu et al. (2010) and 1992; Kimbel et al. 1984, 1997, 2004; Lordkipanidze et al. (2013), attributed Wood and Richmond 2000, Williams fossils from East Africa, the Dmanisi and et al. 2012; Schrenk et al. 2015); however, Sangiran sample to the same group. Homo antecessor - craniofacial features and their taxonomic validity 231

Table 3. Fossil specimens classified here as Earliest Homo erectus (or ) Specimen Age (Ma) Author Africa KNM-ER 3733 1.78 Wood 1991 KNM-ER 3883 1.50-1.65 Wood 1991 KNM-WT 15000 1.50 Walker and Leakey 1993 Middle East (Dmanisi) D2282 ~1.77 Gabunia et al. 2000 Rightmire et al. 2006 D2700 ~1.77 Gabunia et al. 2000 Rightmire et al. 2006 D3444 ~1.77 Gabunia et al. 2000 Rightmire et al. 2006 D4500 ~1.77 Gabunia et al. 2000 Rightmire et al. 2006 Asia Gongwangling 1.63 Zhu et al. 2015 Sangiran 4 >1.5 Swisher et al. 1994 Antón 2003 Antón and Swisher 2004 Larick et al. 2001, 2004 Antón et al. 2007 Sangiran 17 1.25-1.30 Larick et al. 2001, 2004 Antón et al. 2007

The difficulty of classifying taxonom- gaster (Gabunia et al. 2000); H. georgicus ically these hominins is clearly observed (Gabunia et al. 2002); H. erectus georgi- in the disparity of interpretations be- cus (Rightmire et al. 2006); and H. erec- tween different authors. Thus, Wang and tus ergaster georgicus (Lordkipanidze et al. Tobias (2000a) used the term H. ergaster 2013). We have opted for a denomination for hominins with Facial Pattern I, and that combines the one used by Antón Late H. erectus (Classic H. erectus in this (2003) and the one used by Williams et paper) for hominins with Facial Pattern al (2012), and we classify these remains II, making a clear differentiation of the as Earliest H. erectus (or H. ergaster). facial models between both groups. On This group presents a facial mor- the other hand, Williams et al. (2012) phology that distinguishes it from both call the remains of East Africa as H. erec- early Homo and the Classic H. erectus and tus (or H. ergaster). Recently, Tattersall H. sapiens. Such morphology is character- (2015) demonstrates the need to sepa- ized by: a) absence of canine fossa and rate H. ergaster from H. erectus and points presence of sulcus maxillaris; b) zygoma- out the great diversity of classifications ticoalveolar crest oblique or straight; c) for Dmanisi remains: H. erectus (Gabunia absence of incisura malaris; d) nasal bones and Vekua 1995; Henke 1995; Bräuer and little projected; e) weak maxillary inflec- Schultz 1996; Vekua et al. 2002); H. er- tion; f) forward zygomatic root with re- 232 Francesc Ribot Trafí, et al. spect to the lateral nasal margin; g) a low in H. sapiens it is highly flexed. The H. sa- origin of the zygomatic; and h) rectilinear piens flexed face is determined by the ori- superior border of the temporal squama. entation of the infraorbital coronal bone plate, with the lower surface that slopes 3. Classic Homo erectus. In this study, inferiorly and slightly posteriorly (canine Classic H. erectus included specimens fossa), and the more sagittal direction of dated between ~1 Ma and ~0.6 Ma the lateral nasal wall. Along the joint of (hominins classified as Middle H. erec- these two maxillary surfaces a flexion zone tus by Antón, 2003 [Zhoukoudian and is formed. Likewise, the zygomaticoalve- Nanjing], we add the Yuxian skulls). In olar crest is generally curved and often this group, we also included the African presents a malar notch or incisura malaris specimens KNM-OL 45500, OH 12 and (Rak 1983, 1986; Arsuaga et al. 1999). BOU-VP-2/66, although they are not part For this study, we compare the char- of this study, since KNM-OL 45500 is a acteristics described as derived in H. an- frontal and a temporal fragment, OH 12 tecessor, which are six midfacial features, is very fragmented and BOU-VP-2/66 is a plus the facial bone remodeling pattern calvarium (Table 4). —related to midfacial flexion—, and the form of the superior border of the tempo- Anatomical features of the midface ral squama. Moreover, we compare them with those of the most ancestral and The term midfacial is applied to the por- penecontemporaneous hominins (early tion of the anterior skull that is visible in Homo, Earliest H. erectus [or H. ergaster] norma frontalis and that is vertically delim- and Classic H. erectus). ited by the distance ranging from the alve- Coronal orientation of the infraorbital sur- olar plane of the maxilla to the nasion, and face. transversally from zygion to zygion (Pope Maxillary inflection. This inflection is 1991). The topography of this zone is vari- determined by the eversion of the nasal able in the different hominin groups, but lateral wall and the slope downwards and

Table 4. Fossil specimens classified here as Classic Homo erectus Specimen Age (Ma) Author Africa BOU-VP-2/66 ~1 Asfaw et al. 2002 KNM-OL 45500 0.9-0.97 Potts et al. 2004 OH 12 0.78 Tamrat et al. 1995 Antón 2004 Asia Yunxian 0.8-~0.94 Yan 1993 Feng 2008 Feng et al. 2011 De Lumley and Li 2008 Zhoukoudian (lower strata of Locality 1) ~0.8 Shen and Jin 1991 Shen 2001 Shen et al. 2009 Tangshan (Nanjing) ~0.62 Zhao et al. 2001 Homo antecessor - craniofacial features and their taxonomic validity 233 slightly backward of the infraorbital plate and straight-oblique zygomaticoalveolar (Bermúdez de Castro et al. 1997; Arsuaga crest. Likewise, the same morphology et al. 1999). Arsuaga et al. (1999) also of AT-404 is observed in the skull of the add to this morphology the projection of probable H. heidelbergensis of Steinheim the nasal bones, which advances the up- and in some modern specimens (Fig. 3). per border of the nasal rim. Zygomaticoalveolar crest. This crest Canine fossa. The canine fossa has been forms the lower border of the zygomatic defined as an infraorbital depression process of the maxilla and arises from the that encompasses most of the zygomatic lateral wall of the maxilla. The zygomati- process of the maxilla (Mellinger 1940). coalveolar crest can be best seen in frontal Some authors add to this definition that view, and can be determined by the line this infraorbital depression produces a from alveolar plane to the zygomaxillare horizontal incurvation as as an in- (Pope 1991; Koesbardiati 2000). In this curvation of the zygomaticoalveolar crest study we apply the categories defined by (Arsuaga et al. 1999; Maddux and Fran- Etler (1994): oblique (straight and diag- ciscus 2009). This description coincides onally oriented), arched, and horizontal. with what Maureille and Houët (1997) In frontal view, the zygomaticoalveolar call infraorbital depression. Howev- crest is subject to the development of the er, Arsuaga et al. (1996, 1999) suggest incisura malaris and structures related to the inverse by describing the maxillary the malar tubercle (Rak 1983). The incisu- fragment AT-404 from Sima de los Hue- ra malaris may be present or not (Fig. 4). sos (Atapuerca) (Figure 4 in Arsuaga Lower nasal margin. In the formation of et al. 1996, and Figure 7 in Arsuaga et the lower nasal margin three crests are al. 1999), as a possible Homo heidelber- involved: 1) spinal —originates from the gensis, which presents a canine fossa ; a spinal crest can

Fig. 3. Examples of canine fossa with straight-oblique zygomatoalveolar crest Footnote: A) ; B) Probable , Steinheim (flipped image); C) Probable H. heidelbergensis, AT-404, Atapuerca (obtained from Arsuaga et al. 1999); D) Modern human (obtained from Stan et al. 2013). 234 Francesc Ribot Trafí, et al.

Fig. 4. Modern human faces with curved zygomaticoalveolar crest Footnote: A) Absence of incisura malaris. B) With incisura malaris. Collection of the Laboratory of Paleopa- thology and Paleoanthropology, National Museum of Archeology of Catalonia in Barcelona. Photos made by FRT with the authorization of Dr. Domènec Campillo. be located posteriorly, along the poste- the anterior origin of the masseter mus- rior pole of the alveolar process, or an- cle (Rightmire 1998b). It must be distin- teriorly, along the nasoalveolar clivus; 2) guished from the malar tuberosity, which lateral —originates in the lateral margin is a relatively prominent elevation on of the piriform aperture; and 3) turbinal the malar surface parallel to the inferior —originates near the anterior end of the border of the bone, and situated between inferior turbinate (Gower 1923). This the orbital and free margins of the malar margin can be formed by the presence of (Koesbardiati 2000). a single crest or by the combination of Facial bone remodeling pattern. Growth re- two or three, resulting in up to 6 catego- modeling is an integral component of the ries (Gower 1923; Lahr 1994; Franciscus craniofacial growth process (McCollum 2003) (absence of lower nasal margin is 2008). Growth remodeling refers to var- considered as the seventh category). ious combinations of active bone deposi- The above mentioned anterior nasal tion and resorption on the inner and outer spine is the anterior extension of the inci- surfaces of a single bony lamina (McCol- sive crest that forms a spinal process that lum 2008). The presence of areas of bone projects anterior to the alveolar process. resorption in the maxilla and the nasoal- Hominoids and many fossil hominins veolar clivus during ontogeny is associ- do not present this anterior extension, ated with the orthognathic face, whereas therefore, they do not have a true ante- the presence of areas of bone deposition rior nasal spine, although they usually in the maxilla and the nasoalveolar clivus present a small protuberance or tuber in contributes to the formation of the prog- the nasoalveolar clivus that indicates the nathic face (Enlow and Wang 1965; Enlow most anterior union of the nasal septum. 1966; Bromage 1989; McCollum 2008). Zygomaxillary tubercle. This tubercle Superior border of the temporal squama. occupies the lower part of the anterior The superior border of the temporal face of the zygomatic bone, which marks squama may present two morphologies: Homo antecessor - craniofacial features and their taxonomic validity 235 flat or rectilinear, which is associated straight and diagonally oriented zygoma- with a low cranial capacity, and convex ticoalveolar crest; very little nasal lateral or arched, which is associated with in- eversion (little sagittal orientation of the creased cranial capacity. lateral nasal wall); orthognathic subnasal plane; infraorbital foramen close to the Results orbital margin; very broad, flattened na- soalveolar clivus, wich curves backward Early Homo around the P3 roots (well defined P3 jug- um), the canine alveoli are fully part of In early Homo the facial morphology ex- the anterior row; very broad and deep hibits two possible morphotypes, the one palate; relatively narrow upper face; and present in 1470 group, and the one pres- forward extension of the maxillary sinus ent in the 1813 group, on the other hand. to the coronal plane of the inferior nasal However, prior to these two groups, margin. In KNM-ER 1470 the inferior there is the Hadar maxilla A.L. 666-1, nasal margin is formed by the lateral ex- dated ~2.4 Ma and classified as Homo aff. tension of the spinal crest (nasal margin H. habilis (Kimbel et al. 1997). The con- configuration 2 —Franciscus 2003). figuration of this maxilla could indicate Finally, in KNM-ER 1470, the superior the morphological pattern present in the border of the temporal squama is recti- first Homo. Kimbel et al. (1997) describe linear. this fossil as having a nasoalveolar clivus which is not very prognathic; presence 1813 group of corrugations in the alveoli of the cen- tral incisors; slight eversion of the lateral This group includes KNM-ER 1813, nasal margins; lateral nasal crest that ex- KNM-ER 42703, OH 24, OH 65 and OH tends and curves medially on the clivus; 62 and presents a different facial model spinal tubercle (nasal spine not observ- from the previous one: the infraorbital able in lateral view); spinal crest marking plane is oriented coronally and is pos- the lower nasal margin; and superficial teroinferiorly inclined, and presents a depression that forms a rudimentary ca- more extensive lateral maxillary surface nine fossa in the form of a “tear-drop”. (not compressed) than in the previous group, giving rise to an extensive and 1470 group shallow canine fossa (Rak 1983; Kim- bel et al. 1997; Schwartz and Tattersall The first morphology, characterized by 2003); the zygomatic root is positioned KNM-ER 1470 and KNM-ER 62000 dis- posteriorly to the coronal plane (on P4/ plays a coronal, flat and anteroinferiorly M1 in all specimens of this group — oriented infraorbital plane, with a small Leakey et al. 2012), with a higher origin; distance between the anterior maxillary the zygomaticoalveolar crest is arched surface and the root of the zygomatic pro- (KNM-ER 1813, KNM-ER 42703, OH cess (compressed coronal plane), without 24) or straight and diagonally oriented canine fossa and with zygomaticoalveolar (OH 62, but with a very sallow canine crest with low and forward origin (on P3/ fossa —Schwartz and Tattersall, 2003); P4 in KNM-ER 62000 and on P4 in KNM- also, it does not exhibit an incisura ma- ER 1470 —Leakey et al. 2012), and with laris nor a zygomaxillary tubercle (Pope 236 Francesc Ribot Trafí, et al.

1991); the forward extension of the Earliest Homo erectus maxillary sinus to the coronal plane (or Homo ergaster) of the inferior nasal margin is not evi- dent in the 1813 group; but there is In the Earliest H. erectus (or H. ergaster), the presence of maxillary inflection and two midfacial morphologies are found lateral nasal eversion (marked in KNM- related to life stage (Wang and Tobias ER 1813 and somewhat weaker in OH 2000a). The morphology found in imma- 24); the subnasal region is narrow and ture individuals (KNM-WT 15000 and prognathic; the nasal inferior margin is D2700) is characterized by: the orien- weakly developed: in KNM-ER 1813 and tation of the infraorbital region is posi- OH 24 is formed by the lateral expan- tioned in the coronal plane, without be- sion of a small blunt spinal crest that ing anteroinferiorly or posteroinferiorly does not come together with the later- inclined (uniplanar); everted nasal bones al crest (category 2 of Franciscus 2003), and sagittal oriented maxillae lateral to while the lateral crest descends by the the pyriform aperture; and weak maxil- naseoalveolar clivus, whereas in OH 62 lary inflection. The zygomatic process of the spinal crest extends laterally until it the maxilla is flat, and there is a presence merges with the turbinal crest, both sep- of a sulcus maxillaris (partial obliteration arated from the lateral crest (category 3 of the canine fossa, which becomes a sul- of Franciscus 2003). cus that descends from the infraorbital In most early Homo, there is only a foramen). The zygomaticoalveolar crest small midline tubercle on the inferior is arched in D2700 and oblique in KNM- nasal margin instead of an anterior nasal WT 15000, without an incisura malaris and spine. The exceptions are OH 24 (Right- with the absence of a maxillary tubercle. mire 1993; Franciscus and Trinkaus The lower nasal margin in KNM-WT 1988; McCollum et al. 1993) and OH 15000 (Rightmire 1998b) is formed by 62 (Johanson et al. 1987), both exhibit a the spinal crest, which is directed lateral- well-defined anterior nasal spine. KNM- ly towards the lateral crest, but does not ER 1813 presents a small elevation on merge, since the spinal crest is only a few each side of the midline suggesting that millimeters posterior to the lateral nasal in vivo there was an anterior nasal spine margin (category 3 of Franciscus 2003), (Schwartz and Tattersall, 2003). The na- while in D2700 there is no spinal crest. sal spine in OH 24 and OH 62 is not ob- And finally, the presence of nasal spine, servable in lateral aspect. although it is not visible in norma later- In OH 13 and SK 27 the facial bone alis. growth remodeling pattern shows strong The facial morphology found in adult bone deposition on the anterior face, specimens is characterized by marked similar to that found in the australopiths, maxillary flexion (the maxillae lateral and possibly related to nasoalveolar prog- to the pyriform aperture face forward), nathism (Bromage 1989). of the following are also present: a sul- The morphology of the superior bor- cus maxillaris is evident; the zygomati- der of the temporal squama is variable: coalveolar crest is arched or straight hor- rectilinear in OH 24 or curved (convex) izontally; nasal bones project anteriorly in KNM-ER 1813. (nasal prominence); a spinal crest, to a greater or lesser degree of development, Homo antecessor - craniofacial features and their taxonomic validity 237 form the lower nasal margin (except in incisura malaris; there is a significant pos- Sangiran 17). KNM-ER 3733 and Sangi- terior shift in the origin of the zygomatic ran 17 exhibit a zygomaxillary tubercle relative to the lateral nasal margin; pres- (Pope 1991; Leakey and Walker 1985; ence of zygomaxillary tubercle; and evert- Rightmire 1998b; Schwartz and Tattersall ed nasal bones (projected nose). 2003). The existence of a prominent na- In the skulls of Yuxian the midfacial sal spine observable in lateral view is vis- morphology is described by having a cor- ible in Sangiran 4 (Sawyer and Tattersall onally oriented infraorbital plate, a high 1995; Tattersall and Sawyer 1996; Fran- origin of the zygomatic root, the presence ciscus and Trinkaus 1988; Rightmire, of a canine fossa; and a straight zygomati- 1998b; McCollum 2000; Schwartz and coalveolar crest and with a marked incisu- Tattersall 2003) and in Gongwangling ra malaris (Li and Etler 1992). Regarding (Ju-kang 1966; Rightmire 1998b); while the midface, in specimens from Zhouk- the origin of the zygomatic in all Earliest oudian, Wang and Tobias (2000a. p.25) H. erectus (or H. ergaster) is just lateral to stated: “Medially, the portion of the maxilla the lateral margin of the nasal aperture. lateral to the pyriform aperture faces lateral- Incisive corrugations are also observed ly, while the infraorbital part faces anteriorly; in D2282 (Rightmire et al. 2006), D4500 thus they jointly constitute a concave flexion”. (Rightmire et al. 2017) and Gongwan- Effectively, in the new reconstruction gling (Rightmire 1998b). of Classic H. erectus from Zhoukoudian The facial bone growth remodeling (Sawyer and Tattersall 1995; Tattersall pattern has only been studied in the ad- and Sawyer 1996; Wang 1998), the inci- olescent individual KNM-WT 15000, sura malaris is less restricted that in Weid- which shows a primitive pattern, i.e. a enreich’s (1937a) reconstruction, result- flat clivus without corrugations or re- ing in a wider zygomaticoalveolar crest. sorption around the anterior tooth roots This expansion produces a much more (Lacruz et al. 2013, 2015). pronounced infraorbital hollow. Unfor- In all specimens, the superior border tunately, it is impossible to ascertain of the temporal squama is rectilinear, ex- whether this feature is a true canine fossa cept in D3444. or a sulcus maxillaris (seen in the Weiden- reich reconstruction), because the maxil- Classic Homo erectus la used in this reconstruction is eroded. Other maxillae from ZKD lower are The midfacial morphology in this group, also fragmentary, which makes the actual especially in the Chinese Classic H. erectus morphology of the infraorbital region of (Yuxian [Li and Etler 1992], Zhoukoud- the maxilla difficult to observe (Tattersall ian [Wang and Tobias 2000a; Etler 2004], personal communication, 2016). Accord- and Tangshan —Nanjing [Wang and To- ing to Wang (1998) and Wang and Sun bias 2000a,b, 2001; Liu et al. 2005]) is (2000), the Weidenreich reconstruction characterized by: infraorbital plate ori- of the female Classic H. erectus facial skel- ented coronally and posteroinferiorly in- eton involves fragments of cranial skel- clined; the sulcus maxillaris has become a etons from different individuals (both true canine fossa; the maxillary flexion is male and female), a female cranial cap fairly pronounced; the zygomaticoalveo- (skull XI or LII) and a male maxilla (No lar crest is very arched, with a developed V, mentioned as No II); so, the zygomatic 238 Francesc Ribot Trafí, et al. bone had to be turned to some degree to 1813 group— Earliest H. erectus [or H. er- make contact with the maxillary and the gaster] and Classic H. erectus), because it frontal bones. The zygoma faces laterally presents a flexed midface (i.e., coronal to meet the frontal and maxillary bones, orientation of the infraorbital surface; leading to lower and larger orbit. Nev- projecting nose; maxillary inflection — ertheless, as seen in the reconstructed with lateral wall of the nasal opening male skull (Sawyer and Tattersall 1995; everted with respect to the adjacent infra- Tattersall and Sawyer 1996), the general orbital surface—; canine fossa; and hori- topography of the midfacial skeleton as zontally arched zygomaticoalveolar crest reconstructed by Weidenreich (1937a) with high zygomatic root) (Bermúdez de for the Zhoukoudian population is still Castro et al. 1997, 2017; Arsuaga et al. useful. In the Tangshan face, the anterior 1999; Rosas 2000). and posterior surfaces of the zygomatic In extant African hominoids (Goril- process of the maxilla meet each other la and Pan), East African australopiths, in a blunt edge. The edge extends supe- 1813 group and in extant H. sapiens, the riorly and laterally, and then turns infe- infraorbital region is oriented coronally riorly, forming an incisura malaris (Liu et and with a posteroinferior slope, which al. 2005). Tangshan displays, in the lower extends towards the zygomaticoalveo- area of the zygomatic root, a very well-de- lar crest; above the infraorbital foramen veloped canine fossa associated with a and immediately below the lower orbital marked midfacial flexion (Wang and To- border is the transverse buttress, which bias 2000a,b; 2001). forms a ridge that extends medially from Yunxian 2 and Zhoukoudian Local- the body of the zygomatic bone towards ity 1 Classic Homo erectus (crania III, V, the upper part of the nasal aperture. This XI, and XII) displays a temporal squama transverse buttress is most pronounced with a convex superior border (Right- in Gorilla, and gradually decreases to be- mire 1998a; Etler 2004; Wu and Poiri- come rudimentary in H. sapiens. Thus, the er 1995). Conversely, the examination transverse buttress along with the nasal of the Tangshan shows that aperture (more or less pronounced — the temporal squama is relatively high very elongated in Gorilla and modest in and curved (convex) (Liu et al. 2005). H. sapiens) and the posteroinferior slope of the infraorbital region form the ca- Discussion nine fossa. In the species in which the infraorbital plane slope is anteriorly, the The knowledge of patterns of facial mor- transverse buttress is obliterated (i.e., A. phology and topography are important africanus, Paranthropus, and 1470 group) for understanding the variability and evo- and the canine fossa has been modified lutionary relationships of early hominins (maxillary furrow in A. africanus, and (Tobias 1967, 1991; Rak 1983, 1986; maxillary fossula in ) Pope 1991; Rightmire 1998b, 1993; or disappeared (Paranthropus boisei and Wang and Tobias 2000a). In the differen- 1470 group). This could be due to the tial diagnosis of H. antecessor, some previ- anteroinferior compression of the coro- ous scientists have noted that this taxon nal plane caused by the strong anterior is derived from the most primitive forms expansion of the maxillary sinus. In this of Homo (Early Homo —1470 group and respect, however, Arsuaga et al. (1999) Homo antecessor - craniofacial features and their taxonomic validity 239 concluded that the presence of canine group there is no incisura malaris, while in fossa and flexed face are derived charac- Earliest H. erectus [or H. ergaster] this trait ters in H. sapiens that are produced for the is variable, and in Classic H. erectus and H. first time in H. antecessor, and they argued, sapiens it is present). On the other hand, citing Kimbel et al. (1997), that in all species with a high origin of the masse- early Homo and Earliest H. erectus (or H. ter, and a coronally oriented infraorbital ergaster) (e.g., KNM-ER 3733 and KNM- plane, the zygomaticoalveolar crest is WT 15000) a compression of the coronal straight and obliquely oriented (australo- plane is observed resulting in the typical piths and paranthropiths) (Freidline et smooth face in these hominins. But, the al. 2013). real interpretation done by Kimbel et al. The modern human-like facial fea- (1997) in this respect is different, these tures of H. antecessor are ancestral reten- authors point out that in 1470 group tions, relative to later humans, and can there is a strong anterior expansion of the be considered to be part of a generalized maxillary sinus that causes anteroposte- pattern of facial architecture (Wang and rior compression of the coronal plane Tobias 2000a; Freidline et al. 2013; Ribot and, therefore, obliteration of the canine et al. 2017). This conclusion fully coin- fossa, while the anterior surface of the zy- cides with what Rak (1986) calls a gen- gomatic process of the maxilla is practi- eralized face, present in many cally in the same coronal plane as the na- including modern humans and fossil spe- sal opening, resulting in the smooth face; cies of the genus Homo, which is charac- while in 1813 group, the maxillary sinus terized by: an infraobital surface oriented reaches no further anteriorly than the in- in the coronal plane and sloping pos- cisive fossa, and in lateral view, the zygo- terorinferiorly, and at the junction with matic process of the maxilla and the nasal the nasal aperture this inclination forms opening are separated by a considerable the canine fossa; the zygomaticoalveolar topographical interval, and posteriorly crest is curved and joins the body of the positioned zygomatic process roots, thus maxilla at approximately the midpoint of forming the canine fossa, and concluding the dental arch or posterior to it; and the that the latter is the generalized mor- nasoalveolar clivus is convex in cross-sec- phology in hominins (in this description tion. KNM-ER 3733 is also included). The face present in H. sapiens is in many On the other hand, the zygomati- aspects plesiomorphic (in the most basic coalveolar crest morphology is also com- sense), and these characteristics are even plex. Thus, the general curvature of the present very weakly in the East African zygomaticoalveolar crest and the pres- australopiths and early Homo. When each ence or absence of the incisura malaris are characteristic of the middle facial third is correlated characteristics, and the expres- taken independently, or if all of them are sion of these characteristics depend on taken together, the pattern in A. afarensis the height of the masseter origin (Freid- is a prototypical morphology that could line et al. 2013). Species with a low origin evolve by the relatively simple modifica- of the masseter and a coronal orientation tions into the more derived face of H. sa- of the infraorbital plane have an arched piens. The complete transformation of the zygomaticoalveolar crest, and may or plesiomorphic face to the current one is a may not have an incisura malaris (in 1813 gradual process that culminates with the 240 Francesc Ribot Trafí, et al. formation of a more flexed and orthog- In fact, the description of the midfa- nathic face. Thus, the anterior part of this cial morphology of Yunxian Classic H. midfacial region is greatly affected by the erectus (Li and Etler 1992) is to the same bite and occlusion forces of the anterior as the description of ATD6-69 (Bermú- dentition and premolars, while the zygo- dez de Castro et al. 1997): “[…] the face matic and the zygomaticoalveolar crest is flattened and orthognathic with moderate al- are affected by molar loads (Wang et al. veolar prognathism; the maxilla has a distinct 2010; Rodrigues Freire et al. 2014). The canine fossa; the lateral part of the maxilla is strongest stress of these forces (extensive oriented coronally and highly angled to the zy- stress and compressive stress) occurs in gomatic; there is a high origin of the zygomatic the canine fossa and in the frontal pro- root; a horizontal zygomaxillary border and cess of the maxilla, two of the structures a pronounced malar incisures […]” (Li and most related to facial flexion. Postnatal Etler 1992, p. 404). The same morphol- ontogeny could be influenced by epigen- ogy is also found in the Tangshan face etic factors, such as biomechanics, which (Wang and Tobias 2000a,b, 2001; Liu et can elucidate functional differences be- al. 2005). Also, in table 5 it can be ob- tween different facial forms. In this sense, served that the results for the midfacial it should be noted that the orthognathic morphology in H. erectus sensu lato are face is better adapted to absorb the mas- very variable, but if we divide this group ticatory or paramasticatory stress than in Earliest H. erectus (or H. ergaster) and the prognathic face (Wang et al. 2010). Classic H. erectus, it is appreciated that Compared with East African australo- the morphology of the first group is more piths, the face of the 1813 group is some- primitive, while that of the second group how more flexed, with a more everted is more derived, which justifies the divi- wall adjacent to the lateral nasal margin, sion of H. erectus sensu lato in these two and to an infraorbital region more sloped groups. posteroinferiorly, which results in a ca- Therefore, the presence of a flexed nine fossa less restricted (more expand- midface is a plesiomorphic characteristic ed) and less deep. This trend of midface for all the extant African hominoids (Go- flexion increase continues in Earliest H. rilla and Pan), the East African australo- erectus (or H. ergaster) (where the canine piths and early Homo (except for 1470 fossa has been partially obliterated and group) and in H. sapiens. So we agree with has become the sulcus maxillaris), and cul- Wang and Tobias (2000a) who stated minates in Classic H. erectus, where the that the flexed face is plesiomorphic and, combination of a greater posteroinferior therefore does not constitute an effective slope of the infraorbital coronal plane, an taxonomic basis for the proposed new arched zygomaticoalveolar crest with inci- species, H. antecessor. sura malaris and a greater eversion of the Regarding the nasal entrance, ATD6- wall adjacent to the lateral nasal margin 69 presents the category 3 of Gower are linked to the presence of a “true” ca- (Franciscus 2003), the model that oc- nine fossa (Fig. 5). curs primarily in modern humans (the This midfacial morphology present 35.93% in a series of 109 modern skulls in Classic H. erectus is similar to the one [Franciscus 2003]). However, in modern present in H. sapiens, and it is identical to humans all models of combination of that present in H. antecessor (Table 5). crests for forming the lower nasal margin Homo antecessor - craniofacial features and their taxonomic validity 241 - Marked Present Present Homo antecessor Arched/horizontal 3 Curved Coronal with po with Coronal steroinferior slope Present - Homo erectus Marked Present Present Classic Arched/horizontal – Curved steroinferior slope Coronal with po with Coronal Present - a - Homo erec - Homo erga b c Homo antecessor Earliest tus (or ster ) Variable 2 and 3 Rectilinear Coronal with po with Coronal steroinferior slope Variable Variable Sulcus maxillaris Weak/variable - Homo erectus sensu Homo erectus lato Variable 2 and 3/variable Variable Coronal with po with Coronal steroinferior slope Variable Variable Variable Variable - 1813 group Arched/horizontal 2 and 3 Rectilinear Coronal with po with Coronal steroinferior slope Absent Absent Present Weak Examined hominins 1470 group Oblique 2 Rectilinear Coronal with ante - roinferior slope Absent Absent Absent Very weak Very Craniofacial feature Zygomaticoalveolar Zygomaticoalveolar crest Lowernasal margin configuration* Superior border of the temporal squ - ama Orientation of the of Orientation infraorbital surface Incisura malaris Zygomaxillarytu - bercle Canine fossa Maxillary inflection Present in KNM-ER 3733 and Sangiran 17. Present Marked in Gongwangling and Sangiran 17. Marked Present in KNM-ER 3733 and Sangiran 17. Present Table 5. Craniofacial features of examined hominins (see discussion in text) compared with hominins (see discussion in text) 5. Craniofacial features of examined Table a b c *The categories of inferior nasal margin used here are those described in Gower (1923) modified by Franciscus (2003). *The categories of inferior nasal margin used here are those described in Gower (1923) modified by Franciscus 242 Francesc Ribot Trafí, et al.

Fig. 5. Morphology of the midface in hominins compared with ATD6-69 Footnote: In all the images, the red line marks the inclination of the infraorbital plane. A) Early Homo, 1470 group. Upper row (front view), left: KNM-ER 1470, right: ATD6-69. Observe in ER 1470 the absence of a canine fossa, the practical absence of maxillary flexion, and a rectilinear/oblique zy- gomaticoalveolar crest. Bottom row (side view), left: KNM-ER 1470, right: ATD6-69; the infraorbital plane in ER 1470 is oriented anteroinferiorly (uniplanar), which indicates a compressed coronal plane, with low and forward origin of the zygomaticoalveolar crest. B) Early Homo, 1813 group. Upper row (front view), left: KNM-ER 1813, right: ATD6-69. Observe in ER 1813 the presence of extended and shallow canine fos- sa, the low maxillary flexion, and an arched zygomaticoalveolar crest, without incisura malaris. Bottom row (side view), left: KNM-ER 1813 (flip image), right: ATD6-69; the infraorbital plane in ER 1813 is oriented posteroinferiorly, with a slightly forward origin of the zygomaticoalveolar crest. C) Earliest Homo erectus (or Homo ergaster). Top row (front view), left: KNM-WT 15000, right: ATD6-69. Observe in WT 15000 the absence of canine fossa and the presence of sulcus maxillaris, and a rectilinear/oblique zygomaticoalveolar crest. Bottom row (side view), left: KNM-WT 15000, right: ATD6-69; the infraorbital plane in WT 15000 is oriented anteroinferiorly (uniplanar), which indicates a compressed coronal plane, with a slightly backward origin of the zygomaticoalveolar crest. D) Classic Homo erectus. Upper row (front view), left: Tangshan skull (Nanjing), right: ATD6-69. Note in Tangshan the presence of a marked canine fossa, the marked maxillary flexion, and an arched zygomaticoalveolar crest with incisura malaris. Bottom row (side view), left: recon- struction of the Zhoukoudian skull (Sawyer and Tattersall, 1995; Tattersall and Sawyer, 1996), right: ATD6- 69; the infraorbital plane in Zhoukoudian is oriented posteroinferiorly, with a very backward origin of the zigomatic alveolar crest. As can be seen, the morphology of Classic H. erectus is identical to that of ATD6-69. Pictures of ATD6-69 courtesy of Roberto Sáez. Homo antecessor - craniofacial features and their taxonomic validity 243 are present. Furthermore, category 3 is pattern of a modern type (Lacruz et al. ancestral for Homo as it is present in OH 2013; Bermúdez de Castro et al. 2017). 62, KNM ER 3733 and KNM-WT 15000 The midfacial bone remodeling pattern (Franciscus 2003). observed in ATD6-69 is present in H. sa- On the other hand, some researchers piens. In both ATD6-69 and H. sapiens it have shown that Earliest H. erectus (or H. is associated with an orthognathic face, ergaster) nasal morphology is practical- a fully flexed maxillary inflection (which ly identical to that of modern humans anteriorly projects the nasal bones), a (Franciscus and Trinkaus 1988), based coronally oriented infraorbital region on the following characters: a) presence and canine fossa, and a curved zygoma- of anterior nasal spine prominence in ticoalveolar crest with an incisura malaris some specimens (Sangiran 4 [Franciscus (Lacruz et al. 2013). and Trinkaus 1988; Rightmire 1998b; The orthognathic face is associated McCollum 2000; Schwartz and Tatter- with areas of bone resorption during on- sall 2003]; Gongwangling cranium [Woo togeny (Enlow and Wang 1965; Enlow 1966; Rightmire 1998b]); b) smooth na- 1966). In ATD6-69, these resorptive ar- sal cavity entrance with horizontal nasal eas are located in the zones of the naso- sill; and c) posterior spine. Also in Sangi- maxillary clivus, anterolateral maxilla, ran 4 (the only specimen in which it can and in the canine fossa (Fig. 2). Deposi- be observed) a vomeral insertion does tional areas in ATD6-69 are close to the indeed occur above the nasal sill. canine prominence, in the lateral walls of Another important feature of the mid- the nasal aperture and in the anterior part facial morphology of H. antecessor is the of zygomatic (Lacruz et al. 2013) (Fig. 2). presence of a zygomaxillary tubercle. The ATD6-69 and H. sapiens facial bone Arsuaga et al. (1999), Carbonell et al. remodeling pattern is very different from (2005) and Bermúdez de Castro et al. that seen in the australopiths (LH 2, LH (2017) noted that this is a derived fea- 21, AL 333-105, Sts 2, Stw 59, Taung, Sts ture, and that it is observed for the first 24, Sts 57, MLD 2 and Sts 52), early Homo time in the remains of Gran Dolina and (OH 13, SK 27) and Earliest H. erectus (or Zhoukoudian. However, the presence of H. ergaster) (KNM-WT 15000). In all of a zygomaxillary tubercle was noticed for the latter specimens there is a large depo- the first time in KNM-ER 3733 —Pope sition area of the maxilla and the naso- 1991; Leakey and Walker 1985— (dated alveolar clivus, contributing the develop- ~1.8 Ma [Wood, 1991]) and Sangiran 17 ment of the prognathic profile (Bromage —Pope 1991; Rightmire 1998b— (dated 1989; Lacruz et al. 2015). However, the 1.25-1.30 Ma [Larick et al. 2001, 2004; facial bone remodeling pattern in A. sedi- Antón et al. 2007]). Thus, the tubercle ba (MH1), which present moderate prog- precedes the appearance of ATD6 and nathism, shows that the pattern present Zhoukoudian hominins. Therefore, the in this specimen is identical to that in H. presence of zygomaxillary tubercle in H. sapiens, presenting vertical bands of bone antecessor is also a plesiomorphic, not a resorption in the lower face (Lacruz et al. derived trait. 2015). According to the authors, these The last feature of the middle facial vertical bands of resorption are reminis- third described as derived in ATD6-69 cent with the alternating stripes of high is the presence of a bone remodeling and low tension observed in the roots of 244 Francesc Ribot Trafí, et al. the anterior teeth during biting, at least Therefore, the presence of a very sim- as simulated in a human cranial mod- ilar facial bone remodeling pattern in el subjected to finite elements analysis Paranthropus, A. sediba, ATD6-69 and H. (Wang et al. 2010; Lacruz et al. 2015). sapiens could indicate the emergence of Moreover, variability of this feature this model in different moments in hu- should be taken into account. In the man evolution. study by Kurihara et al. (1980), a sam- Regarding the morphology of the tem- ple of 27 modern human skulls aged be- poral bone, According to Arsuaga et al. tween 2 and 14 old was considered, (1999), the ancestral condition would be 20 (74%) show a predominantly resorp- a low temporal squama with a straight tive anterior lower face; in the remaining superior border. According to these au- 7 specimens (26%), most of the anteri- thors, this morphology is found in H. ha- or lower face was found to be deposito- bilis, Earliest H. erectus (or H. ergaster) and ry. Also, in a sample of 33 children and Classic H. erectus; while in H. antecesor, subadult chimpanzees, McCollum (2008) the presence of a temporal squama with found a complete depository pattern in arched superior border represents the de- the anterior lower face in only 6 speci- rived condition, which is also observed mens, while in 21 specimens the pattern in the hominins of the European and Af- was partially resorptive and fully resorp- rican Middle Pleistocene, in those of the tive in four. The same author, in a sample Asian Middle Pleistocene, in Neandertals, of 22 modern human skulls, shows that and in modern humans. In this regard, in 55% “they were found to display large ar- Carbonell et al. (2005) add that H. ante- eas of surface deposition along the nasoalveo- cessor and Classic H. erectus must share a lar clivus and anterior maxilla” (McCollum hypothetical common ancestor (possibly 2008). These studies indicate that there Earliest H. erectus [or H. ergaster]) that is a close relationship between an orthog- retained the primitive condition of a low nathic face and areas of bone resorption and flat temporal squama, and note that during ontogeny, located in the zones of these features are retained in Classic H. the nasomaxillary clivus, maxillary an- erectus. However, the skull of Classic H. terolateral, and in the canine fossa. And, erectus from Bouri (~1 Ma), already pres- conversely, the bone deposition located ents a high and arched temporal squama in these same facial regions is not always (Asfaw et al. 2002, supplementary infor- present. Thus, there is variability in the mation). Likewise, many Classic H. erec- models of resorption and deposition in tus, as Yunxian 2, Zhoukoudian Locality 1 the nasomaxillary clivus, anterolateral (crania III, V, XI, and XII), and Tangshan maxilla and canine fossa, which can be (Wu and Poirier 1995; Rightmire 1998a; present in both orthognathic faces as in Liu et al. 2005; Etler 2004), also present prognathic faces. the same morphology. Thus, a temporal Although it has not been possible to squama with a convex superior border is establish an ontogenetic series for Classic already found in Classic H. erectus from H. erectus, based on final midfacial adult Africa and China. form, it is reasonable to argue that their Therefore, regarding the midfacial ontogenetic process of the midface must and cranial morphology, ATD6 hominins be similar to that found in some modern and Classic H. erectus share the follow- humans (Wang 1998). ing characters: presence of canine fossa, Homo antecessor - craniofacial features and their taxonomic validity 245 arched zygomaticoalveolar crest with in- erectus in the upper lower and middle cisura malaris and a high zygomatic root, Pleistocene of Europe (Wang and Tobias maxillary inflection, and nasal morpholo- 2000a; Rabadà 2005; Ribot and García gy practically identical to that of modern Bartual 2016; Ribot et al. in press 2018). humans; as well as a convex superior bor- Moreover, the low and flexed face with der of the temporal squama. Thus, the fa- slight to absent prognathism in modern cial topography of ATD6-69, considered humans could be more readily developed to be derived (Bermúdez de Castro et al. from a low and flexed face such as that 1997; 2017; Arsuaga et al. 1999) is ac- seen in Classic H. erectus. tually a plesiomorphic trait, because it is already present in the penecontemporary Acknowledgements Classic H. erectus. Moreover, in a study by Freidline et al. (2013) on the evaluation We would like to thank Dr. Vance of the changes in the shape of the devel- Hutchinson, Dr. Esteban Sarmiento, opment of facial morphology in H. anteces- Dr. Bill Jungers, Dr. Carles Ferrández, sor, these authors conclude that most of Dr. Lluís Gibert, Dr. Jordi Galbany, and the facial characteristics similar to those Prof. Xinzhi Wu for their reviews of this of current humans, present in ATD6-69, manuscript and for their useful sugges- are intercorrelated, and they can be con- tions. We thank Dr. Ian Tattersall for sidered ancestral retentions in relation providing information about the maxil- to later humans. They also suggest that lary of H. erectus from Zhoukoudian, Dr. these features were present in varying de- Scott Maddux for his information about grees in more plesiomorphic forms, such the variability of the zygomaticoalve- as Classic H. erectus and H. antecessor, and olar crest, and Dr. Jeffrey Schwartz for have even been retained in recent mod- providing information on the maxillary ern humans. of KNM-ER 1813. We thank Dr. Alejan- dro Pérez-Pérez for allowing us to study Conclusions the casts of hominins of the Depart- ment of Evolutionary Biology, Ecology We conclude that none of the midfacial and Environmental Sciences (Section characters that have been employed to of Zoology and Biological Anthropolo- define H. antecessor are unique to the Gran gy), at the University of Barcelona, and Dolina hominins. Any one of them has Dr. Domènec Campillo for allowing us been found in Classic H. erectus. There- to study the casts of hominins and col- fore, it becomes untenable to separate lections of extant human skulls of the both groups in two different taxa, H. Laboratory of Paleopathology and Pa- antecessor and Classic H. erectus, and the leoanthropology (National Museum of Atapuerca hominins of ATD6 should be Archaeology of Catalunya, Barcelona). all classified as Classic H. erectus (Wang Images of ATD6-69 from figures 1 and 1998; Wang and Tobias 2000a; Etler 5 are courtesy of Roberto Sáez. Images 2004; Rabadà 2005; Ribot et al. 2006, from Zhoukoudian H. erectus are courte- 2007, in press 2018; Ribot and García sy of Dr. Ian Tattersall and Gary Sawyer. Bartual 2016). We also thank the two anonymous re- Therefore, ATD6 hominins would viewers for their comments and sugges- demonstrate the presence of Classic H. tions. 246 Francesc Ribot Trafí, et al.

1996. Evolución humana en Europa: regis- Authors’ contributions tro y debates. Revista Española de Paleon- tología Nº Extraordinario:269-77. FRT was the principal investigator, desi- Arsuaga JL, Martínez I, Lorenzo C, Gracia A, gned and performed the study, conceived Muñoz A, Alonso O, Gallego J. 1999. The and designed the manuscript; MGB ar- human cranial remains from Gran Dolina ranged and analyzed the data, wrote part Lower Pleistocene site (Sierra de Atapuer- of the manuscript; QW wrote part of the ca, Spain). J Hum Evol 37:431-57. manuscript, and was a proofreader. All Asfaw B, Gilbert WH, Beyene Y, Hart WK, authors read and approved the final ma- Rennek PR, WoldeGabriel G, Vrba ES, nuscript. White TD. 2002. Remains of Homo erectus from Bouri, Middle Awash, Ethiopia. Na- Conflict of interest ture 416:317–20. Bailey SE. 2002. A closer look at The authors declare that there is no con- postcanine dental morphology: The man- dibular dentition. Anat Rec 269:148–56. flict of interest. Bermúdez de Castro JM, Arsuaga JL, Car- bonell E, Rosas A, Martínez I, Mosquera M. Corresponding author 1997. A hominid from the Lower Pleisto- cene of Atapuerca, Spain: Possible ancestor Francesc Ribot Trafí, Museo Municipal to neandertals and modern humans. Sci- de Prehistoria y Paleontología Dr. Gibert, ence 276:1392–5. Calle Las Tiendas, sn, 18858 Orce, Gra- Bermúdez de Castro JM, Martinón-Torres nada, Spain M, Martín-Francés L, Modesto-Mata M, E-mail address: [email protected] Martínez-de-Pinillos M, García C, Car- bonell E. 2017. Homo antecessor: The state References of the art eighteen years later. Quaternary International 433:22–31. Antón S. 2003. Natural history of Homo erec- Blumenschine RJ, Peters CR, Masao FT, tus. Am J Phys Anthropol 46:126–70. Clarke RJ, Deino AL, Hay RL, Swisher Antón SC. 2012. Early Homo: Who, When, CC, Stanistreet IG, Ashley GM, McHenry and Where. Current Anthropology 53 Sup- pl 6:S278-98. LJ, Sikes NE, Van Der Merwe NJ, Tactikos Antón C, Potts R, Aiello LC. 2014. Evolution JC, Cushing AE, Deocampo DM, Njau JK, of early Homo: An integrated biological per- Ebert JI. 2003. Late Homo and spective. Science 345:1236828. Hominid Land Use from Western Olduvai Antón SC, Spoor F, Fellmann CD, Swisher CC Gorge, Tanzania. Science 299:1217–21. III. 2007. Defining Homo erectus: Size con- Braüer G, Schultz M. 1996. The morpholog- sidered. In: Henke W, Tattersall I, editors. ical affinities of the Plio-Pleistocene man- Handbook of Paleoanthropology. 1st ed. dible from Dmanisi, Georgia. J Hum Evol Berlin, Heidelberg, New York: Springer‐ 30:445–481. Verlag. Bromage TG. 1989. Ontogeny of the early Antón SC, Swisher CC 3rd. 2004. Early dis- hominid face. J Hum Evol 18:751-73.. persal of Homo from Africa. Ann Rev An- Carbonell E, Bermúdez de Castro JM, Arsuaga thropol 33:271–96. JL, Allue E, Bastir M, Benito A, Cáceres I, Antón SC. 2004. The face of Olduvai Hominid Canals T, Díez JC, Van der Made J, Mos- 12. J Hum Evol 46:337–47. quera M, Ollé A, Pérez-González A, Ro- Arsuaga JL, Martínez I, Carretero JM, Gracia dríguez J, Rodríguez XP, Rosas A, Rosell A, Lorenzo C, Bermúdez de Castro JM. J, Sala R, Vallverdú J, Vergés JM. 2005. An Early Pleistocene hominin mandible from Homo antecessor - craniofacial features and their taxonomic validity 247

Atapuerca-TD6, Spain. Proc Natl Acad Sci in Yunxian County in Hubei. Jianghan Ar- USA 102:5674–8. chaeology 120:57–64. Carbonell E, Bermúdez de Castro JM, Arsua- Franciscus RG, Trinkaus E. 1988. Nasal mor- ga JL, Díez JC, Rosas A, Cuenca-Bescós G, phology and the emergence of Homo erectus. Sala R, Mosquera M, Rodríguez XP. 1995. Am J Phys Anthropol 75:517–27. Lower Pleistocene Hominids and Arti- Franciscus RG. 2003. Internal nasal floor con- facts from Atapuerca-TD6 (Spain). Science figuration in Homo with special reference to 269:826–30. the evolution of Neandertal facial form. J Cartmill M, Smith FH. 2009. The human lin- Hum Evol 44:701–29. eage. Hoboken, New Jersey: Wiley-Black- Freidline SE, Gunz P, Harvati K, Hublin JJ. well. 2013. Evaluating developmental shape Clarke RJ. 2012. A maxilla and changes in Homo antecessor subadult facial other newly-discovered hominid fossils morphology. J Hum Evol 65:404–23. from , Tanzania. J Hum Evol Gabunia L, de Lumley M-A, Vekua A, Lordki- 63:418–28. panidze D, de Lumley H. 2002. Découverte De Lumley H, Li TY, directors. 2008. Le site d’un nouvel hominidé à Dmanissi (Trans- de l’homme de Yunxian: Quyuanhekou, caucasie, Géorgie). C R Paleovol 1:243–53. Quingqu, Yunxian, province du Hubei. Par- Gabunia L, Vekua A. 1995. A Plio-Pleistocene is: CNRS Editions. hominid from Dmanisi, East Georgia, Cau- Delson E. 1997. One skull does not a specie casus. 373:509–12. make. Nature 389:445–6. Gabunia L, Vekua A, Lordkipanidze D, Swish- Enlow DH, Wang S. 1965. Growth and re- er CC 3rd, Ferring R, Justus A, Nioradze modelling of the human maxilla. Am J Or- M, Tvalchrelidze M, Antón SC, Bosins- thodontics 51:446–64. ki G, Jöris O, Lumley MA, Majsuradze G, Enlow DH. 1966. A morphogenetic analy- Mouskhelishvili A. 2000. Earliest Pleisto- sis of facial growth. Am J Orthodontics cene hominid cranial remains from Dman- 52:283–99. isi, Republic of Georgia: , geolog- Etler DA. 1994. The Chinese : ical setting, and age. Science 288:1019–25. New Finds, New Interpretations. PhD the- Gower CG. 1923. A contribution to the mor- sis. Berkeley: University of California. phology of the apertura piriformis. Am J Phys Etler DA. 2004. Homo erectus in East Asia: hu- Anthropol 6:27–36. man ancestor or evolutionary dead-end? Hay RL. 1976. The setting of the gorge and Athena Review 4:37–49. history of the geological investigations. In: Falk D. 1986. Evolution of Cranial Blood Hay RL, editor. The Geology of the Olduvai Drainage in Hominids - Enlarged Occipital Gorge. Berkeley: University of California Marginal Sinuses and Emissary Foramina. Press. Am J Phys Anthropol 70:311–324. Henke W. 1995. Qualitative and quantitative Feibel CS, Lepre CJ, Quinn RL. 2009. Stratig- analysis of the Dmanisi mandible. In: Rad- raphy, correlation, and age estimates for lanski RJ, Renz H, editors. Proceedings of fossils from Area 123, Koobi Fora. J Hum the 10th international symposium on den- Evol 57:112–22. tal morphology. Berlin: Christine und Mi- Feng X. 2008. Stratégie de débitage et mode chael Brünae GbA. de façonnage des industries du Paléo- Johanson DC, Masao FT, Eck GG, White TD, lithique inférieur en Chine et en Europe Walter RC, Kimbel WH, Asfaw B, Manega entre 1 Ma et 400 000 ans Ressemblances P, Ndessokia P, Suwa G. 1987. New par- et différences de la culture de l’homme de tial skeleton of Homo habilis from Olduvai Yunxian et Acheuléen européen. L’Anthro- Gorge, Tanzania. Nature 327:205–9. pologie 112:423–47. Joordens J, Dupont-Nivet G, Feibel C, Spoor Feng XB, Lu CQ, Wang H. 2011. New ad- F, Sier MJ, van der Lubbe JH, Nielsen TK, vance in the study of the Homo erectus site Knul MV, Davies GR, Vonhof HB. 2013. 248 Francesc Ribot Trafí, et al.

Improved age control on early Homo fos- Java, Indonesia. Proc Natl Acad Sci USA sils from the upper Burgi Member at Koobi 989:4866–71. Fora, Kenya. J Hum Evol 65:731–45. Leakey MG, Spoor F, Dean MC, Feibel GS, Kaifu Y, Indriati E, Aziz F, Kurniawan I, Baba Antón SC, Kiarie C, Leakey LN. 2012. New H. 2010. Cranial Morphology and Variation fossils from Koobi Fora in northern Kenya of the Earliest Indonesian Hominids. In: confirm taxonomic diversity in early Homo. Norton CJ, Braun DR; editors. Asian Pa- Nature 488:201–4. leoanthropology: From Africa to China and Leakey REF, Walker AC. 1985. Further homi- Beyond. Springer. nids from the Plio-Pleistocene of Koobi-Fo- Kimbel WH, Rak Y, Johanson DC. 2004. The ra, Kenya. Am J Phys Anthropol 67:135–63. skull of Australopithecus afarensis. Oxford Li T, Etler DA. 1992. New middle Pleistocene University Press. hominid crania from Yunxian in China. Na- Kimbel WH. 2009. The origin of Homo. In: ture 357: 404–7. Grine FE, Fleagle JG, Leakey RE, editors. Liu W, Zhang Y, Wu X. 2005. Middle Pleis- The first humans. Origin and early evolu- tocene human cranium from Tangshan tion of the genus Homo. Springer. (Nanjing), Southeast China: A new re- Kimbel WH, Johanson DC, Rak Y. 1997. Sys- construction and comparisons with Homo tematic Assessment of a Maxilla of Homo erectus from Eurasia and Africa. Am J Phys From Hadar, Ethiopia. Am J Phys An- Anthropol 127:253–62. thropol 103:235–62. Lordkipanidze D, Ponce de León MS, Margve- Kimbel WH, White TD, Johanson DC. 1984. lashvili A, Rak Y, Rightmire GP, Vekua A, Cranial Morphology of Australopithecus af- Zollikofer CPE. 2013. A Complete Skull arensis: A Comparative Study Based on a from Dmanisi, Georgia, and the Evolu- Composite Reconstruction of the Adult tionary Biology of Early Homo. Science Skull. Am J Phys Anthropol 64:337–388. 342:326–31. Koesbardiati T. 2000. On the relevance of the Maddux SD, Franciscus RG. 2009. Allometric regional features of the face in East Asia. scaling of infraorbital surface topography PhD Thesis. Hamburg: Universität Ham- in Homo. J Hum Evol 56:161–74. burg. Martinón-Torres M, Bermúdez de Castro JM, Kurihara S, Enlow DH, Rangel RD. 1980. Re- Gómez-Robles A, Arsuaga JL, Carbonell E, modelling reversals in anterior parts of the Lordkipanidze D, Manzi G, Margvelashvili human mandible and maxilla. Angle Orth- A. 2007. Dental evidence on the hominin od 50:98–106. dispersals during the Pleistocene. Proc Lacruz RS, Bermúdez de Castro JM, Mar- Natl Acad Sci USA 104:13279–82. tinón-Torres M, O’Higgins P, Paine ML, Maureille B, Houeët F. 1997. Nouvelles don- Carbonell E, Arsuaga JL, Bromage TG. nées sur charactèristiques derivées du ma- 2013. Facial morphogenesis of the earliest sif facial superieur des Neandertaliens. An- Europeans. PLoS One 8: e65199. thropol Prehistorie 108 :89–98. Lacruz RS, Bromage TG, O’Higgins P, Toro-Ib- McCollum MA, Grine FE, Ward SC, Kimbel acache V, Warshaw J, Berger LR. 2015. Dis- WH. 1993. Subnasal morphological varia- tinct growth of the nasomaxillary complex tion in extant hominoids and fossil homi- in Au. sediba. Sci Rep 5:15175. nids. J Hum Evol 24:87–111. Lahr MM. 1994. The Multiregional Model of McCollum MA. 2000. Subnasal morpholog- Modern Human Origins: A Reassessment ical variation in fossil hominids: a reas- of its Morphological Basis. J Hum Evol sessment based on new observations and 26:23–56. recent developmental findings. Am J Phys Larick R, Ciochon RL, Zaim Y, Sudijono, Su- Anthropol 112:275–83. minto, Rizal Y, Aziz F, Reagan M, Heizler McCollum MA. 2008. Nasomaxillary remodel- M. 2001. Early Pleistocene 40Ar/39Ar ages ling and facial form in robust Australopithe- for Bapang Formation hominins, Central cus: a reassessment. J Hum Evol 54:2–14. Homo antecessor - craniofacial features and their taxonomic validity 249

Mellinger W. 1940. The canine fossa. Arch of the International Union for Prehistoric Otolaryngol 31:930-7. and Protohistoric Sciences (UISPP). S03- Parés JM, Arnold L, Duval M, Demuro M, 29, Lisbon, Portugal, 4-9 September 2006. Pérez-González A, Bermúdez de Castro JM, Ribot Trafí F, García Bartual M. 2016. Evo- Carbonell E, Arsuaga JL. 2013. Reassessing lución del esqueleto mediofacial en Homo. the age of Atapuerca-TD6 (Spain): new pa- Implicaciones para un modelo de evolu- leomagnetic results. Journal of Archaeolog- ción. In: Ribot F, coordinador. Homenaje al ical Science 40:4586–95. Dr. José Gibert Clols. Una vida dedicada a Pope GG. 1991. Evolution of the zygomatico- la ciencia y al conocimiento de los primer- maxillary region in the genus Homo and its os europeos. Granada, Spain: Publicaciones relevance to the origin of modern humans. Diputación de Granada. J Hum Evol 21:189–213. Rightmire GP, Lordkipanidze D, Vekua A. Potts R, Beherensmeyer AK, Deino A, Ditch- 2006. Anatomical descriptions, compara- field P, Clark J. 2004. Small Mid-Pleisto- tive studies and evolutionary significance cene Hominin Associated with East African of the hominin skulls from Dmanisi, Re- . Science 305:75–78. public of Georgia. J Hum Evol 50:115–41. Prat S. 2002. Anatomical study of the skull Rightmire GP, Ponce de León MS, Lordki- of the Kenyan specimen KNM-ER 1805: a panidze D, Margvelashvili A, Zollikofer CP. re-evaluation o its taxonomic allocation? 2017. Skull 5 from Dmanisi: Descriptive C. R. Paleoevol 1:27–33. anatomy, comparative studies, and evolu- Prat S. 2004. Les premiers représentants du tionary significance. J Hum Evol 104:50– genre Homo, en quête d’une identité. Ap- 79. ports de l’étude morphologique et de l’anal- Rightmire GP. 1993. Variation among early yse cladistique. Bulletins et Mémoires de la Homo crania from Olduvai Gorge and the Société d’Anthropologie de Paris 16:17–35. Koobi Fora region. Am J Phys Anthropol Rabadà i Vives D. Los homínidos de Atapuer- 90:1–33. ca: entre la sinonimia y la paradoja (Sierra Rightmire GP. 1998a. in de Atapuerca, Burgos, España). Batalleria the Middle Pleistocene: The role of Homo 12:143–8. heidelbergensis. Evolutionary Anthropology Rak Y. 1983. The Face. New 6:218–27. York: Academic Press. Rightmire GP. 1998b. Evidence from facial Rak Y. 1986. The Neanderthal: A New Look at morphology for similarity of Asian and Af- an Old Face. J Hum Evol 15:151–64. rican representatives of Homo erectus. Am J Ribot F, García Bartual M, Wang Q. 2018. A Phys Anthropol 106:61–85. comparative study of the craniofacial fea- Rightmire GP. 2001. Patterns of hominid evo- tures defining ‘Homo antecessor’. Ahead Acta lution and dispersal in the Middle Pleisto- Anthropologica Sinica (in press). cene. Quaternary International 75:77–84. Ribot F, Gibert J, García Bartual M. 2007. Rightmire GP. 2007. Later Middle Pleisto- La evolución facial de los homínidos: un cene Homo. In: Henke W, Tattersall I, edi- modelo de evolución darvinista. In: Braga tors. Handbook of Paleoanthropology. 1st JC, Checa A, Company M, editors. XXIII ed. Berlin, Heidelberg, New York: Spring- Jornadas de la Sociedad Española de Pale- er-Verlag. ontología. Libro de resúmenes. Instituto Rodrigues Freire A, Bevilacqua Prado F, Geológico y Minero de España y Univer- Cláudia Rossi A, Yoshito Noritomi P, Hait- sidad de Granada, 2007, 195. [Abstract]. er Neto F, Ferreira Caria PH. 2014. Biome- (Caravaca de la Cruz, 3-6 de Octubre de chanics of the human canine pillar based 2007). on its geometry using finite element analy- Ribot F, Gibert J, Sánchez F, Iglesias A, Gibert sis. Int J Morphol 32:214–20. L. 2006. Homo antecessor: un nouveau tax- Rosas A. 2000. Human evolution in the last on? XV UISPP Congress: World Congress million years — The Atapuerca evidence. 250 Francesc Ribot Trafí, et al.

Acta Anthropol Sin 19 (Suppl.):8–17. Evol 32: 17–82. Sawyer GJ, Tattersall I. 1995. A new recon- Stringer C. 2002. Modern human origins: struction of the skull of Homo erectus from progress and prospects. Philos Trans R Soc Zhoukoudian, China. In: Gibert J, Sánchez Lond B Biol Sci 357:563–79. F, Gibert L, Ribot F, editors. The Hominids Swisher CC 3rd, Curtis GH, Jacob T, Getty and their environment during the lower AG, Suprijo A, Widiasmoro. 1994. Age of and middle Pleistocene of Eurasia. Pro- the earliest known hominids in Java, Indo- ceedings of the International Conference nesia. Science 263:1118–21. of Human Palaeontology (Orce). Granada: Tamrat E, Tohveny N, Taieb M, Opdyke N. Museo de Prehistoria y Paleontología J. 1995. Revised magnetostratigraphy of Gibert y Ayuntamiento de Orce. Plio-Pleistocene sedimentary sequence of Schrenk F, Kullmer O, Bromage T. 2015. The the Olduvai Formation (Tanzania). Palae- Earliest Putative Homo Fossils. In: Henke ogeogr Palaeoclimatol Palaeoecol 114:273– W, Tattersall I, editors. Handbook of Paleo- 83. anthropology. 2nd ed. Berlin, Heidelberg, Tattersall I. 2015. Homo ergaster and Its Con- New York: Springer-Verlag. temporaries. In: Henke W, Tattersall I, edi- Schwartz J, Tattersall I, editors. 2002. The Hu- tors. Handbook of Paleoanthropology. 2nd man Fossil Record, Volume 1: Terminology ed. Berlin, Heidelberg, New York: Spring- and Craniodental Morphology of Genus er‐Verlag. Homo (Europe). New York: Wiley-Liss. Tattersall I, Sawyer GJ. 1996. The skull of Schwartz JH, Tattersall I. 2003. The human “Sinanthropus” from Zhoukoudian, China: a fossil record. Volume 2. Craniodental mor- new reconstruction. J Hum Evol 31:311–4. phology of genus Homo (Africa and Asia). Tobias PV. 1967. The Cranium and Maxillary New York: Wiley-Liss. Dentition of Australopithecus (Zinjanthro- Shen G, Gao X, Gao B, Granger DE. 2009. pusy) boisei En: Olduvai Gorge: Volume 2. Age of Zhoukoudian Homo erectus deter- Cambridge: Cambridge University Press. mined with 26Al/10Be burial dating. Nature Tobias PV. 1980. The natural history of the 458:198-200. helicoidal occlusal plane and its evolu- Shen G, Jin L. 1991. Restudy of the upper age tion in early Homo. Am J Phys Anthropol. limit of Beijing Man cave. Acta Anthrop Si- 53:173–187. nica 10:273–277. Tobias PV. 1991. The Skulls, Endocasts and Shen G. 2001. Current status and prospects Teeth of Horno habilis In: Olduvai Gorge: for dating fossil hominid sites in China. In: Volume 4. Cambridge: Cambridge Univer- Tobias PV, Rath MA, Moggi-Cecchi J, Doyle sity Press. GA, editors. Humanity from African nais- Vekua A, Lordkipanidze D, Rightmire GP, sance to coming millennia. Florence, Italy: Agustı´ J, Ferring R, Majsuradze G, Firenze Univ. Press. Mouskhelishvili A, Noradze M, Ponce de Simpson SW. 2015. Early Pleistocene Homo. León M, Tappen M, Tvalchrelidze M, Zol- In: Muehlenbein MP, editor. Basics in Hu- likofer C. 2002. A new skull of early Homo man Evolution. Elsevier. from Dmanisi, Georgia. Science 297:85–9. Spoor F, Leakey MG, Gathogo PN, Brown FH, Walker AC, Leakey MD. 1993. The Narioko- Antón SC, McDougall I, Kiarie C, Manthi tome H. erectus skeleton. Cambridge: Har- FK, Leakey LN. 2007. Implications of new vard University Press. early Homo fossils from Ileret, east of Lake Wang Q, Sun L. 2000. Eightieth of Pe- Turkana, Kenya. Nature 448:688–91. king Man: Current status of Stan MC, Bordei P, Dina C, Iliescu DM. 2013. and the Zhoukoudian site. Anthropol Rev Characteristics of some anatomical land- 63:19–30. marks on the anterior face of the maxilla. Wang Q, Tobias PV. 2000a. Review of the phy- ARS Medica Tomitana 2:79–83. logenetic position of Chinese Homo erectus Strait DS, Grine FE, Moniz MA. 1997. J Hum in light of midfacial morphology. Acta An- Homo antecessor - craniofacial features and their taxonomic validity 251

thropol Sin 19 (Suppl.):23–33. Woo J-K. 1966. The hominid skull of Lantian, Wang Q, Tobias PV. 2000b. Recent advances Shensi. Vertebrata PalAsiatica 10:1–16. in Chinese palaeo-anthropology. South Af- Wood B. 1991. Koobi Fora Research Project. rican Journal of Science 96:463–6. Vol. 4: Hominid cranial remains. Oxford: Wang Q, Tobias PV. 2001. An old species and a Clarendon Press. new frontier: Some thoughts on the taxon- Wood BA. 1992. Origin and evolution of the omy of Homo erectus. Przeglad Antropolog- genus Homo. Nature 355:783–90. iczny - Anthropological Review, 64:9–20. Wood B, Richmond BG. 2000. Human evo- Wang Q, Wright BW, Smith A, Chalk J, Byron lution: taxonomy and paleobiology. J Anat GD. 2010. Mechanical impact of 196:19–60. loading on the midfacial skeleton Wu X, Poirier FE. 1995. Human Evolution in and its relevance to human evolution. The China: A Metric Description of the Fossils Anatomical Record 293:607–17. and a Review of the Sites. London: Oxford Wang Q. 1998. Comparison of Homo heidel- University Press. bergensis and Chinese Homo erectus. Ph.D. Yan GL. 1993. A preliminary study on mag- Dissertation. Institute of Vertebrate Pale- netic stratigraphy of the geological section ontology and Paleoanthropology, Chinese with the fossil bed of Yunxian Homo of Hu- Academy of Sciences, Beijing. bei. Earth Science-Journal of China Univer- Weidenreich F. 1937a. Reconstruction of the sity of Geosciences 18:221–6. entire skull of an adult female individual of Zhao J, Hu K, Collerson KD, Xu H. 2001. Sinanthropus pekinensis. Nature 140:1010–1. Thermal ionization mass spectrometry Weidenreich F. 1937b. The dentition of Sinan- U-series dating of a hominid site near Nan- thropus pekinensis. Palaeont Sin New Ser jing, China. Geology 29:27–30. D1:1–180. Zhu ZY, Dennell R, Huang WW, Wu Y, Rao Weidenreich F. 1943. The skull of Sinanthropus ZG, Qiu SF, Xie JB, Liu W, Fu SQ, Han JW, pekinensis: A comparative study on a prim- Zhou HY, Ou Yang TP, Li HM. 2015. New itive hominid skull. Palaeont Sin New Ser dating of the Homo erectus cranium from D10:1–484. Lantian (Gongwangling), China. J Hum Williams FL, Schroeder L, Ackermann RR. Evol 78:144–57. 2012. The mid-face of lower Pleistocene hominins and its bearing on the attribution of SK 847 and StW 53. HOMO - Journal of Comparative Human Biology 63:245–57.