Early Neanderthal mandibular remains from Baume Moula-Guercy (Soyons, Ardèche)

Abstract

We provide an ontogenetically-based comparative description of mandibular remains from Last Interglacial deposits (MIS 5e) at Baume Moula-Guercy and examine their affinities to European and Middle Eastern Middle-to-Late Pleistocene (≈MIS 14—MIS 1) Homo. Description of the M-G2-419 right partial mandibular corpus with M_1-3 (15–16.0 years ±0.5 years) and mandibular fragments M-F4-77 and M-S-TNN1 is with reference to original fossils, casts, CT scans, literature descriptions, and virtual reconstructions. Our comparative sample is ontogenetically based and divided into a Preneanderthal—Neanderthal group and a Homo sapiens group. These groups are subdivided into (1) Preneanderthals (≈MIS 14-9), Early Neanderthals (MIS 7-5e), and Late Neanderthals (MIS 5d-3), and (2) Middle (MIS 5) and Upper (MIS 3-Pre-MIS 1) Paleolithic and recent H. sapiens. Standard techniques were employed for developmental age and sex determinations and measurements. The M-G2-419 mandible possesses corpus features that link it most closely with the Sima de los Huesos Preneanderthal and Early Neanderthal groups. These include mental foramen position, number, and height on the corpus, anterior marginal tubercle position, and mylohyoid line orientation. Metrically, the M-G2-419 mandibular corpus is small relative to adults in all groups, but the thickness/height relationship is like the adult condition. The thickness of the corpus is more like Neanderthal children than adolescents. Molar crown features suggest affinities with the Preneanderthal—Neanderthal group. The Moula-Guercy mandibles possess a combination of Neanderthal-associated features that provides insights into MIS 7-5e paleodeme variation and the timing of appearance of MIS 5d-3 Neanderthal facial features.

1 INTRODUCTION

Understanding the origins of the Neanderthal lineage is complicated by the high degree of morphological variation among individuals from Middle Pleistocene localities in Europe and the paucity of the record (Abbazzi et al., 2000; Arsuaga et al., 2014; Daura et al., 2017; de Lumley, 1973; de Lumley, 2015; Hublin, 1982; Mounier, 2011; Mounier et al., 2009; Quam et al., 2023; Roksandic et al., 2011, 2018; Rosas, 1998; Rosas et al., 2019; Rosas & Bermúdez de Castro, 1998; Stefan & Trinkaus, 1998; Verna et al., 2020; Vialet et al., 2018; Williams, 2013; Williams & Cofran, 2016). Mandibular remains from Mauer (Mounier, 2011; Mounier et al., 2009; Schoetensack, 1908), Arago (Condemi, 2001; de Lumley, 1973, 2015; Rosas, 2001; Rosas & Bermúdez de Castro, 1998, 1999) and the Atapuerca localities (Sima de los Huesos, Gran Dolina: Bermúdez de Castro et al., 2017; Quam et al., 2023; Rosas, 1995, 2001; Rosas et al., 2019) have played a significant role in this discussion. While the morphology of these early mandibles presents problems in clarifying the origins of the Neanderthal lineage (Quam et al., 2023; Rosas, 2001), the mandibular sample from the Sima de los Huesos (SH, MIS 12—MIS 11) possesses a set of features suggested to be characteristic of Neanderthals (Rosas, 1995, 2001; Rosas et al., 2019; Rosas & Bermúdez de Castro, 1998). However, the exact time of appearance of other potentially Neanderthal-associated mandibular characters, and the time at which higher frequencies are reached for individual traits considered unique to later Neanderthals, is unclear. Mandibular remains from Eemian (Last Interglacial, MIS 5e) deposits at Baume Moula-Guercy provide new evidence to address these questions.

The Serre (hill) de Guercy is a Kimmeridgian stage (Late—Upper Jurassic) clayey limestone eminence rising ≈250 m above the western bank of the Rhône River in the commune of Soyons, Ardèche. The eastern side of the Serre de Guercy contains numerous karstic cavities (Baume Moula-Guercy, Baume Néron, Grotte des Enfants, and Trou du Renard), some of which have yielded Neanderthal skeletal and cultural remains (Defleur et al., 1992; Defleur et al., 1994; Lepic & Lubac, 1872; Veyrier et al., 1951; Veyrier & Combier, 1951). Systematic excavations at the Baume Moula-Guercy revealed 20 stratigraphic levels (Payen et al., 1990; Saos et al., 2014). A diverse fauna and flora were recovered from Layers XV-XII in association with a Ferrassie Mousterian (Middle Paleolithic) lithic assemblage (Defleur, 1995, 2015; Defleur et al., 1998; Valensi et al., 2012). Excavation of Layer XV revealed the remains of at least six Neanderthal individuals (Guercy 1–6: Richards et al., 2021). These remains were directly associated with the fauna and possess cuts and percussion marks consistent with the perimortem processing of bone for its nutritional content (Defleur, 1995; Defleur et al., 1999; Defleur et al., 2020; Defleur & Desclaux, 2019; Defleur, Dutour, Valladas, Combier, & Vandermeersch, 1993; Defleur, Dutour, Valladas, & Vandermeersch, 1993).

Analyses of the Guercy 1 partial cranium (Richards et al., 2021), a subadult occipital (Richards et al., 2022a), the Guercy 1 endocranium (Richards et al., 2022b), and the Guercy 3 subadult maxilla (Richards et al., 2023) show that these individuals are morphologically consistent with other MIS 7—5e Early Neanderthals and more like Preneanderthals than Late Neanderthals in most features.

An isolated partial right mandibular corpus with M_1-3 (M-G2-419) and two isolated corpus fragments (M-F4-77, M-S-TNN1) were also recovered. The M-G2-419 corpus has tentatively been assigned to Guercy 1, an adolescent female, while the other fragments are tentatively assigned to Guercy 4 (M-F4-77, M-S-TNN1), an older adult male (Richards et al., 2021).

An Eemian age determination for the hominin bearing Layer XV at Baume Moula-Guercy is based on extensive assessment of the stratigraphic sequence and the contained faunal, floral, and cultural components (Crégut-Bonnoure et al., 2010; Defleur et al., 2001; Defleur et al., 2020; Manzano, 2015; Payen et al., 1990; Saos et al., 2014). Further support for an MIS 5e association derives from dates calculated from hominin and faunal teeth following laser ablation U-series and CSUS/ESR dating, respectively (Willmes et al., 2016). Absolute dating of higher layers provides additional support. Samples from Layer IV produced radiocarbon ages of >50 ka, while CSUS/ESR dating of faunal teeth produced an age range of 52 ± 2 to 69 ± 3 ka. These dates associate Layer IV with the top of MIS 4—base of MIS 3 (Willmes et al., 2016). Samples of tephra from Layer VI, subjected to thermoluminescence dating, produced an age of 72 ± 12 ka (=top of MIS 4: Sanzelle et al., 2000). Additional discussions of the stratigraphy, fauna, flora, dating, and cultural component from Baume Moula-Guercy are available in Defleur et al. (2020), Grün and Stringer (2023), Richards et al. (2021), and Williams and Cofran (2016).

Comparative descriptions are influenced by sample composition and how the sample is partitioned into groups. The late Early and Middle Pleistocene European and Middle Eastern hominin record is relatively well populated until the base of MIS 9, mainly due to the large quantity of remains from Atapuerca (Sima de los Huesos, Arsuaga et al., 2014; Gran Dolina, Bermúdez de Castro et al., 2017). In contrast, hominin remains from MIS 8 are extremely rare (Dean et al., 1998). The site of Payre in Southeastern France provides the main evidence (isolated teeth, mandibular and parietal fragments; =base MIS 8—top MIS 7: Condemi et al., 2010; Moncel & Condemi, 2007; Moncel et al., 2008; Valladas et al., 2008; Verna et al., 2020). From MIS 7—MIS 5e, the hominin record is populated but skeletal remains are rare, fragmented, widely distributed, and dominated by those from the Krapina paleodeme (sensu Bräuer et al., 2020; Dean et al., 1998; Howell, 1952, 1957, 1960, 1996, 1999; Richards et al., 2021). During MIS 5d—MIS 5b, dating uncertainties are widespread at hominin sites in Europe (cf. La Ferrassie, MIS 5d-MIS 3; Turq et al., 2008: Molare, MIS 5d-MIS 5a; Mallegni & Ronchitelli, 1989; Tuniz et al., 2012; Spagnolo et al., 2020: Regourdou, MIS 5; Turq et al., 2008) and the Middle East (Tabun, MIS 7-5d; Verna et al., 2010: Skhul, MIS 5e; Grün et al., 2005: Qafzeh, MIS 5d-MIS 5a; McDermott et al., 1993). Further, there are few securely dated European MIS 4 hominin fossils (Keeling et al., 2023), leaving only MIS 3 with a well-populated record. Given these data, Neanderthals chronostratigraphically and biostratigraphically associated with MIS 7—MIS 5e form a temporal isolate within the Neanderthal lineage. This discontinuity in the Pleistocene hominin record has also been recognized by Verna et al. (2020). Temporal gaps, like those preceding and postdating the Early Neanderthal record, can impede our ability to track morphological continuity. However, it is becoming clear that Neanderthals from this time range possess a “total morphological pattern” (LeGros Clark, 1964) that separates them from both earlier European and Middle Eastern Homo (≈MIS 14—MIS 9) and later members of the Neanderthal lineage (MIS 5d—MIS 3: Arsuaga et al., 1997; Couture & Hublin, 2005; Dean et al., 1998; Elyaqtine, 1995, 1996; Guipert et al., 2011; Rougier, 2003; Richards et al., 2021; Richards et al., 2022a; Richards et al., 2022b; Richards et al., 2023; Vandermeersch, 1978; Verna et al., 2020). While numerous identifiers have been applied to designate the coherence of this group and its separation from other Neanderthals (cf. Condemi, 1998; Dean et al., 1998; Elyaqtine, 1995, 1996; Hublin, 1998; Martínez & Arsuaga, 1997; Rougier, 2003), we follow Verna et al. (2010, 2020) and Richards et al. (2021), Richards et al. (2022a), Richards et al. (2022b), and Richards et al. (2023) in referring to these individuals as “Early Neanderthals.” Because analysis of nuclear DNA extracted from the SH remains by Meyer et al. (2016) shows them to be either very early Neanderthals or closely related to Neanderthals, and because they possess some morphological characters suggested to be derived in Neanderthals, we categorize them as “Preneanderthals” (MIS 12—MIS 11). Non-SH European Homo mandibles associated with MIS 14—MIS 10 are grouped separately and termed Middle Pleistocene Homo, as their relationship to the SH Preneanderthals is unclear (Arsuaga et al., 2014; Daura et al., 2017; Quam et al., 2023; Rosas et al., 2019; Rosas & Bermúdez de Castro, 1998; Vallois, 1956; Verna et al., 2020). More recent Neanderthal remains associated with MIS 5d—MIS 3 are categorized as “Late Neanderthals.”

The Moula-Guercy partial mandible M-G2-419 is from an adolescent, while the M-F4-77 and M-S-TNN1 mandibular fragments represent one or more adults. Appropriate comparisons, therefore, are to an ontogenetically based comparative sample.

Early Neanderthal mandibular remains preserving the lateral corpus are few, are mostly incomplete, and sample a broad geographic region (Verna et al., 2020). The adult sample comprises only seven individuals (Ehringsdorf F, Krapina 57–59, La Chaise Bourgeois-Delaunay 1, Payre 15, Tabun II: Condemi, 2001; Piveteau, 1970; Smith, 1976; Verna et al., 2020; Vlček, 1993; Williams, 2006). The subadult sample is limited to nine individuals. These subadults sample a broad developmental range (≈4.0–17.5 years) but have a similar geographic distribution to the adult sample (Ehringsdorf G, Krapina 52–56, La Chaise Abri Suard S13, La Naulette 1, Montgaudier: Duport, 1971; Duport & Vanderneersch, 1976; Leguebe & Toussaint, 1988; Mann & Vandermeersch, 1997; Smith, 1976; Tillier & Genet-Varcin, 1980; Vlček, 1993; Williams, 2006). When this early sample is considered by marine isotope stage, portions of the lateral corpus from MIS 7 comprise two adults (Ehringsdorf F, Vlček, 1993; Payre 15, Verna et al., 2020) and one subadult (Ehringsdorf G, Vlček, 1993). In MIS 6, preserved lateral mandibular corpuses are even more rare, comprising only one adult (La Chaise Bourgeois-Delaunay 1: Condemi, 2001; Rosas & Bermúdez de Castro, 1998) and one subadult (La Chaise Abri Suard S13: Tillier & Genet-Varcin, 1980). Considering lateral mandibular corpus remains potentially associated with MIS 5e, three of the eight mandibles (La Chaise Bourgeois-Delaunay 1, La Naulette 1, Montgaudier) have dating uncertainties. The La Chaise Bourgeois-Delaunay 1 mandible, often considered to be associated with MIS 5e, is likely associated with MIS 6 (Blackwell et al., 1983). The age of the Montgaudier mandible is unclear, but it may be of Riss III age (circa MIS 6: Mann & Vandermeersch, 1997; Marquet, 1987). Patou-Mathis (2000) associated the La Naulette 1 mandible with MIS 5e. Whereas the stratigraphic sequence indicates a potential early age for this specimen, attempts to construct a chronological framework for this locality have met with little success (Toussaint et al., 2011; Toussaint & Pirson, 2006). Given these data, lateral mandibular corpus remains associated with MIS 5e derive solely from the Moula-Guercy and Krapina paleodemes.

The SH Preneanderthal mandibular sample comprises 13 adults and five adolescents (Rosas, 1995, 2001; Rosas et al., 2019). While some SH mandibles are fragmentary, newly excavated fragments have been fitted, resulting in more complete mandibles (Quam et al., 2023). Unfortunately, the subadult SH Preneanderthal sample is limited to adolescents, between 12.5–17.0 years of age, and represents morphometric variation within a single paleodeme. When the Late Neanderthal sample is considered, there are 26 adult and 21 subadult mandibular specimens. However, 15 of the subadults range in age from 2.0 to 10.5 years of age, leaving only Malarnaud 1 (Arsuaga et al., 1989; Condemi et al., 2013; Leroi-Gourhan, 1958; Minugh-Purvis, 1988; Rosas, 2001; Vallois, 1958), Cova del Gegant 1 (Condemi, 2001; Daura et al., 2005; Quam et al., 2015), and Petit-Puymoyen 1 (Minugh-Purvis, 1988) in a developmental age range comparable to that of the Moula-Guercy M-G2-419 adolescent.

The Moula-Guercy mandibular sample, while fragmentary, provides the first securely dated evidence for mandibular evolution in European MIS 5e Early Neanderthals other than that from Krapina. Further, the partial mandibular corpus provides new data on adolescent female Neanderthals. Here we provide a comparative morphological and metric description of the Moula-Guercy M-G2-419, M-F4-77, and M-S-TNN1 mandibular and dental remains. These data are assessed in relation to our ontogenetic sample of Homo mandibular remains from European and Middle Eastern localities in the ≈MIS 14—MIS 1 and recent time ranges. The description is supplemented by a virtual reconstruction of M-G2-419.

2 MATERIALS AND METHODS

2.1 Materials

Moula-Guercy mandibular remains include: (1) a 15–16.0 ± 0.5-year-old right mandibular corpus with erupted RM_1-2 and erupting RM₃ (M-G2-419: Richards et al., 2021), (2) a fragment of an adult right inferior mandibular corpus (M-F4-77), and (3) a fragment of right mandibular corpus with a R_C (M-S-TNN1: Figures 1, 2). For descriptive and comparative purposes, the original fossils were evaluated with reference to the published literature, photographs, CT images, and casts of mandibular remains from European and Middle Eastern Middle-to-Late Pleistocene (MIS 14—MIS ≈ 1) Homo (N = 115; 65 adults, 50 subadults; Supplementary Data, Tables S1 and S2, respectively). Our recent human mandibular sample (N = 60) has a developmental age range of 10 years to adulthood (subadults, 10.1–18.0 years, N = 30; adults, >18.0 years, N = 30). The recent human sample is housed at the Center for Dental History and Craniofacial Study, A.A. Dugoni School of Dentistry, University of the Pacific.

Our Middle-to-Late Pleistocene comparative sample comprises a Preneanderthal—Neanderthal group and a H. sapiens group. These groups were subdivided into time successive subgroups based on marine oxygen isotope stages (MIS). The subgroups include: (1) Middle Pleistocene Homo (MIS 14—MIS 11, N = 6), (2) SH Preneanderthals (MIS 12—MIS 11, N = 18), (3) Early Neanderthals (MIS 8/7—MIS 5e, N = 17), (4) Early/Late Neanderthals (MIS 5d?, N = 1), (5) Late Neanderthals (MIS 5d—MIS 3, N = 45), (6) Middle Paleolithic H. sapiens (MIS 5, N = 5); and (7) Upper Paleolithic H. sapiens (MIS 3—MIS ≈ 1, N = 23).

Our dental sample comprises 542 teeth (M₁ = 217, M₂ = 187, M₃ = 138) and represents a maximum of 237 individuals. The dental sample was divided into groups based on associated MIS stages, as above. The number of teeth in each group is as follows: (1) Middle Pleistocene Homo (MIS 14—MIS 11, N = 23), (2) SH Preneanderthals (MIS 12—MIS 11, N = 18), (3) Early Neanderthals (MIS 8/7—MIS 5e, N = 112); Early/Late Neanderthals (MIS 5d?, N = 6); (4) Late Neanderthals (MIS 5d—MIS 3, N = 156), (5) Middle Paleolithic H. sapiens (MIS 5, N = 45); and (6) Upper Paleolithic H. sapiens (MIS 3—MIS ≈ 1, N = 182).

2.2 Methods

Standard mandibular dimensions were measured following Martin (1928) and Bräuer (1988), or as referenced in the text and tables. Except for the fossil subadults, summary data for the mandibular corpus height, thickness, and index of robusticity (thickness/height × 100) are provided for different tooth positions and the mental foramen region, while the individual values are provided in the Supplementary Data, Excel spreadsheet. We report metric data for the fossil subadults in the text because the record is fragmentary and sample sizes for each age stage are small. Compiled metric values for both the Pleistocene and recent samples are listed by subgroup, MIS association, and individual in the Supplemental Data, Excel spreadsheet.

Summary morphological data related to the mental foramen and anterior marginal tubercle are provided in text tables. Raw data for the latter two features and the mylohyoid line position and orientation are available in the Supplemental Data, Excel spreadsheet. Frequency and position data for mental foramina and anterior marginal tubercles were collected and reported for each mandibular side observed, with the result that some individuals are represented more than once. The number of mental foramina ranged from one to four foramina per side. The position of the main mental foramen and anterior marginal tubercle were recorded relative to the dentition by drawing a straight line through the center of the foramen or greatest elevation of the tubercle and noting the line's position relative to the deciduous or adult dentition. The configuration of the mylohyoid line follows criteria provided by Rosas (2001). Morphological observations for both the Pleistocene and recent samples are listed by subgroup, MIS association, and individual in the Supplemental Data, Excel spreadsheet.

Summary morphological descriptions of teeth retained in the M-G2-419 and M-S-TNN1 mandibular remains are available in Hlusko et al. (2013). Here we expand the morphological and metric observations of the M-G2-419 teeth, while adding observations on the configuration of the pulp cavities of the M₁-M₃. Further morphological studies of the Moula-Guercy dentition require μXCT scans which are not currently available. Morphological observations follow criteria detailed in Patte (1959), Genet-Varcin (1974), Wolpoff (1964, 1979); Bermúdez de Castro, 1988; Bermúdez de Castro et al., 1999; Bermúdez de Castro et al., 2019; Bailey, 2002a; Bailey, 2002b; Martinón-Torres et al. (2012), Garralda et al. (2019), and Martín-Francés et al. (2022). Crown features were scored with reference to the Arizona State University Dental Anthropology System (ASUDAS: Scott & Irish, 2017). Note that in some cases features named and defined in the ASUDAS have been modified but retain the same name (cf. Martinón-Torres et al., 2012). Caution should, therefore, be exercised in the compilation of frequency data by referencing the methods sections to confirm authors' definitions of features. Molar occlusal wear was graded following Molnar's (1971) wear stages.

Measurements of the dentition follow Martin (1928). Dimensions for the comparative sample were compiled from literature sources as listed in the Supplemental Data, Excel spreadsheet (Fossil Dentition and Foss Dent Stats). We compiled values for the mesiodistal length and buccolingual breadth and calculated the crown index. Teeth were compiled by position and assigned to groups based on associated MIS stage, as above. Metric data for the comparative sample are listed in the Supplemental Data by individual and include the MIS stage, group assignment, and references for site dating and metric values.

To assess and delineate the degree of pulp chamber enlargement in the lower molars (taurodontism), we employed criteria established by Keene (1966) for his “Taurodont Index” (TI). The slightly modified version of this index developed by Shifman and Chanannel (1978), as discussed in Constant and Grine (2001), was also used in our assessment.

2.3 Developmental age and sex determinations

We employed isosurface reconstructions and volume renderings of the teeth to determine a developmental age for M-G2-419 and our recent human sample. We only assessed the amount of crown and root calcification in these determinations; the eruption pattern was not considered. Individual tooth scores were obtained by comparison to the dental aging sequence developed by Schour and Massler (1941). A mean age was assessed from the total number of teeth available. We also examined the degree of RM₃ crown and root development relative to the dental aging system of Anderson et al. (1976). The latter method is based on the tooth mineralization standards of Moorrees et al. (1963). Use of this method allowed direct comparisons to the similarly aged Le Moustier 1 Late Neanderthal (Condemi, 2005; Schwartz & Tattersall, 2005; Thompson & Bilsborough, 2005; Thompson & Nelson, 2005) and use of the data compiled for the Moula-Guercy dental sample by Hlusko et al. (2013).

The assessed developmental age for the M-G2-419 partial mandible is 15–16.0 ± 0.5 years (Richards et al., 2021). The R_C embedded in the M-S-TNN1 mandibular fragment has complete root formation and a slight degree of occlusal wear on the canine tip, suggesting a young adult age. The likely antimere of this tooth is a L_C (M-G4-144) that shows no occlusal wear. We suggested that the teeth and mandibular fragment belongs to Guercy 4 (Richards et al., 2021).

Given that a different developmental rate may have been present in Neanderthals relative to recent H. sapiens (Dean et al., 1986, 2001; Granat & Heim, 2003; Macchiarelli et al., 2006; Mahoney et al., 2021; Martín-González et al., 2012; Ramirez Rozzi & Bermudez de Castro, 2004; Rosas et al., 2017; Smith et al., 2007, 2010, 2014), we used the method of Granat and Heim (2003) to calculate a “suggested age” in Neanderthal years. The details of this age assessment can be found in Richards et al. (2021).

Noting the cautions expressed in Genoves (1954, 1969), Weiss (1972), Smith (1980), and Trinkaus (2016), we made an estimation of the sex of the Moula-Guercy M-G2-419 mandible. As discussed in Richards et al. (2021), we employed standard criteria for sex determinations. These criteria include muscle attachment morphology and the overall size, proportions, and thickness of the mandible. The M-G2-419 mandible most likely derives from a late-stage adolescent female. Further, the M-G2-419 mandible was assigned to Guercy 1 during an assessment of the minimum number of individuals (Richards et al., 2021). This likely association with Guercy 1, a late-stage adolescent female partial cranium, strengthens the sex determination assessed from the mandible.

2.4 Virtual reconstruction

A General Electric (GE) BrightSpeed scanner, at the Hôpital Privé Pasteur, Radiology Unit, Guilherand Granges, France, was used to acquire CT scans of the craniodental remains. Scans were acquired in helical mode and reconstructed with a GE “detail” convolution kernel. The slice and interslice thicknesses are 0.625 and 0.3 mm, respectively. Pixel spacing is 0.4883 × 0.4883 mm². For illustrative purposes, the right hemi-mandible was mirror-imaged, and the two halves fit to a virtual reconstruction of the Le Moustier 1 mandible. The pulp chambers of the M₁-M₃ were hand segmented and reconstructed as isosurfaces in Amira 3D.

3 RESULTS

3.1 Immature right mandibular corpus M-G2-419

3.1.1 Comparative morphological description

The M-G2-419 mandibular fragment comprises a nearly complete right mandibular corpus, with RM_1-3 (Figure 1a–d). The mandible has been assigned a developmental age of 15–16.0 ± 0.5 years. This specimen is associated with the Guercy 1 late-stage adolescent female partial cranium. The mandible was fractured such that bone mesial to the socket for the R_C is not preserved. Only the inferior and distolingual portions of the R_C socket and a small extension of the labial cortical plate are preserved. The distal portion of the RP₃ socket is preserved while the RP₄ socket is complete (b-l = 8.6; m-d = 6.1 mm). The mandibular ramus and a portion of the gonial region are missing. A vertical fracture, ≈10 mm posterior to the mandibular foramen, runs superiorly from the inferior border of the corpus. Posterior to this fracture, the ramus and gonial region are missing (Figure 1c). This vertical fracture is joined by a horizontal fracture in the buccal cortical plate that extends posteriorly from the alveolar margin. This fracturing is associated with the loss of the lingual cortex starting from the mesial edge of the RM₃ crypt. From this point, a fracture positioned ≈2 mm inferior to the developing RM₃ roots extends through the base of the mandibular foramen to terminate at the vertical fracture of the ramus (Figure 1c). These fractures resulted in the loss of the superior ramus and exposure of the distal aspect of the developing RM₃ crown and root. Two additional fractures are not associated with bone loss. The first fracture starts at the distobuccal corner of the RM₂ and extends inferiorly around the corpus to terminate near the distolingual RM₃ root. The second fracture in the lingual cortex starts near the anteroinferior extent of the corpus and extends parallel to the mylohyoid line where it terminates near the distal RM₁ root (Figure 1c).

FIGURE 1

Inferior (a), superior (b), medial (c), and lateral (d) views of the M-G2-419 adolescent Early Neanderthal mandible.

The buccal surface of the corpus is well preserved (Figure 1d). Posteriorly, the root of the ascending ramus originates in line with the middle of the RM₃. The position of the anterior ramal root relative to the erupting RM₃ is like the similarly-aged Le Moustier 1 (Figures 1b,d and 2). However, in Le Moustier 1, the root of the ramus is substantially more robust and positioned more laterally. In M-G2-419, the anterior extent of the root of the ramus is nearly coincident with the alveolar ridge. This configuration is not related to the young age of the M-G2-419 mandible, as both Roc de Marsal 1 (Madre-Dupouy, 1992; Tillier, 1983) and Devil's Tower 1 (Tillier, 1982, 1988) have the ramal root placed further laterally, relative to the buccal alveolar ridge. The space between the distal RM₂ and the anterior ramal root is occupied in M-G2-419 and Le Moustier 1 by the developing RM₃; a retromolar gap is not present.

FIGURE 2

Isosurface reconstructions of the M-G2-419 mandible, with the original fossil (gold) and mirror-image (blue) positioned relative to Le Moustier 1 (transparent isosurface). These two individuals have similar developmental ages but differ in their assessed sex. In “a,” note that the ramal root (1), mesial RM₁ (2), and inferior border of the corpus (3) are aligned. Alternatively, note in “b,” the difference in curvature (1) and extent (2) of the anterior mandible between M-G2-419 and Le Moustier 1. This difference suggests a flatter anterior lower face in the M-G2-419 individual. Such a facial configuration is consistent with the adult Early Neanderthals Saccopastore 1 and 2 (Sergi, 1944, 1962) and the juvenile Early Neanderthal maxilla (M-I4-55) from Moula-Guercy (Richards et al., 2023).

Posteroinferiorly, the buccal corpus is flat and lightly marked by a subtriangular masseteric fossa (Figure 1d). The anterior extent of the fossa is positioned at the level of the distal RM₃ crown. In the subadult Early Neanderthal Krapina 53 and the Late Neanderthals Roc de Marsal 1, Devil's Tower 1, Teshik Tash 1, and Le Moustier 1, this region of the corpus is more inflated (superoinferiorly convex) relative to M-G2-419, but the anterior extent of the masseteric fossa is similarly excavated. Alternatively, the lateral corpus of Scladina 4A-1 is flat (Otte et al., 1993; Toussaint et al., 1998; Verna & Toussaint, 2014), as in M-G2-419. Rosas (2001) reports that all SH Preneanderthals except AT-607 (≈14–16 years) have a deeply excavated masseteric fossa. In contrast, all Early and Late Neanderthals examined by Rosas (2001) have either flat (88.5%, 23/26) or shallow (11.5%, 3/26) fossae, respectively.

The alveolar ridge parallels the basal border in M-G2-419, with the basal border being flat from the anterior marginal tubercle to the preserved posterior corpus (=pregonial region). The Krapina Early Neanderthal mandibles are similar to M-G2-419 in having parallel borders (cf. Krapina 53, 55, 57). Although not considered characteristic of Early Neanderthals, a pregonial notch is present in the MIS 6 associated La Chaise Abri Suard S13 (Tillier & Genet-Varcin, 1980). A pregonial notch is also found in the Middle Pleistocene Homo Arago II and Mauer mandibles (Rosas & Bermúdez de Castro, 1999). In the Late Neanderthals, Devil's Tower 1 (Tillier, 1982, 1988) and Roc de Marsal 1 (Tillier, 1984), the inferior border is concave in the region inferior to the distal M₁, forming a pregonial notch. This concavity is absent, and the basal border is flat, in the late-stage adolescent Le Moustier 1, as in M-G2-419. A pregonial notch is present in the adult Regourdou 1, but it is absent in Amud 1 and La Quina H5.

Anterior to the masseteric fossa, the corpus of M-G2-419 has a very slightly developed lateral prominence inferior to the RM₂-RM₃ (Figure 1d). The slightly younger Krapina 53 and adult Krapina 57 Early Neanderthals have somewhat more developed lateral prominences. The Late Neanderthals Teshik Tash 1 and Le Moustier 1 lack significant swelling at the lateral prominence, but both are more developed than M-G2-419. Superior to the lateral prominence, a slightly excavated extramolar sulcus can be discerned. The M-G2-419 extramolar sulcus is less prominent than that observed in the much younger Devil's Tower 1 and substantially less prominent than the sulcus expressed in Le Moustier 1.

Anterior to the lateral prominence, but posterior to the mental foramen, the lateral corpus in M-G2-419 is flat (Figure 1a,b,d). The oblique line arcs anteroinferiorly across this flattened corpus, extending from the anterior root of the ramus to the posterosuperior border of the superior marginal torus. At the level of the mental foramen, the oblique line takes a superiorly directed course, extending to ≈5 mm from the alveolar ridge, whereafter it terminates at the anterior extent of the preserved corpus inferior to the RP₃. The region superior to the oblique line and inferior to the alveolar ridge forms a shallow depression. This depression is superoinferiorly constrained and anteroposteriorly shorter in younger Neanderthals (cf. Devil's Tower 1). With growth, the sulcus enlarges, both anteroposteriorly and superoinferiorly, but is similarly limited to the region distal to the dm₁ or P₃. Further, a lack of inflation of the lateral prominence in M-G2-419 results in a shallow sulcus, while the greater development of the lateral prominence in Le Moustier 1 is associated with an increased depth of the depression. In adult Neanderthals, this depression is taller and deeper, increasing the separation of the oblique line and alveolar ridge (cf. Krapina 58, Regourdou 1, Vindija 206, 231).

The superior marginal torus forms a distinct swelling around the mental foramen (Figure 1d). Similar but less extensive bone deposition around the mental foramen also occurs in Late Neanderthal subadults (cf. Roc de Marsal 1, Devil's Tower 1, Le Moustier 1). The superior marginal torus in M-G2-419 is limited superiorly by the oblique line and associated sulcus. Two distinct but low and rounded ridges extend anteriorly from the superior marginal torus and mental foramen. These ridges diverge as they extend anteriorly from the foramen (cf. Vindija 209, 231; Regourdou 1). The region between these two ridges is slightly depressed, extends from the mental foramen to the anterior extent of the preserved corpus, and likely contained branches of the mental nerve and associated blood vessels.

The single mental foramen in M-G2-419 has a superoinferiorly compressed oval shape (6.3 mm, a-p; 2.7 mm, s-i) and an anteroinferiorly inclined orientation (Figure 1d). A predominantly anteroinferior exit of the neurovascular bundle is indicated by a smooth anterior but sharp posterior border. Rosas (1995) notes that the SH Preneanderthals have a small groove at the posterior border of the mental foramen, indicating that a portion of the neurovascular bundle exited posteriorly. The Late Neanderthal subadults Roc de Marsal 1 and Devil's Tower 1 do not have depressions that would have indicated the course of the mental nerve and vascular bundle. A broad and anteroinferiorly-directed groove extends from the mental foramen in Regourdou 1, as in M-G2-419. However, Vindija 206 and Amud 1 have a posteriorly grooved foramen, like the SH Preneanderthals. Vindija 231 has small grooves extending anteriorly and posteriorly from the foramen. However, there is an accessory foramen situated posterior to the main foramen with a broad groove in the posterior foraminal wall. These features suggest that the main neurovascular bundle would have exited posteriorly, like that in Vindija 206. (Figure 3)

FIGURE 3

The M-S-TNN1 inferior mandibular corpus fragment is shown in an inferior and superior view in (a). The M-F4-77 mandibular fragment and R_C are shown in (b).

In M-G2-419, two small foramina (<1 mm) are located near the mental foramen. Both are ≈4.5 mm from the main foramen, with one anterosuperior and the other posteroinferior to it. We do not consider these microforamina to comprise multiple mental foramina. In the SH Preneanderthals, multiple mental foramina are only present in the AT-2 adolescent; they are not present in the adult sample (Figure 4). However, multiple mental foramina are present in the adult earlier Middle Pleistocene specimens (cf. Mauer, Schoetensack, 1908: Montmaurin—La Niche, Vallois, 1956; Billy & Vallois, 1977; Rosas, 2001; Rosas & Bermúdez de Castro, 1999; Condemi, 2001; Mounier et al., 2009; Vialet et al., 2018: Supplementary Data, Excel spreadsheet). In Early Neanderthals, multiple mental foramina occur at a frequency of 37.5% (3/8) in subadults and 25% (2/8) in adults, with each mandibular side tabulated separately (Table 1). In this group, Montgaudier is the only specimen having more than two foramina, and its geological context is unclear (Mann & Vandermeersch, 1997). In Late Neanderthal subadults, from one to four mental foramina are present, with the frequency of two or more foramina being 61.2% (11/18). In adult Late Neanderthals, 59.0% (13/22) of the sample has two or more foramina, with only Regourdou 1 having three (Table 1, Figure 4). Stefan and Trinkaus (1998) describe La Quina H9 as having four foramina. However, we agree with their assessment that only two are mental foramina; the other two would constitute accessory foramina.

FIGURE 4

Bar graph showing the percentage occurrence of the different numbers of mental foramina in subadult and adult SH Preneanderthals, Early Neanderthals, Late Neanderthals, Middle Paleolithic Homo sapiens, Upper Paleolithic H. sapiens, and recent H. sapiens. These data represent the number of foramina present on each preserved side of the mandible, not the number present per individual.

TABLE 1. Number and position of mental foramina and position of the anterior marginal tubercle in European and Middle Eastern Middle-to-Late Paleolithic (≈MIS 14—MIS 2 to ≈1) Homo and recent humans^a.

Group		Number of mental foramina^b				Principal position of mental foramina^c											Anterior marginal tubercle position^d
		I	II	III	IV	dm1	dm1-dm2	dm2	P3	P3-P4	P4	P4-M1	<M1^e	M1	>M1	M1-M2	P3	P3-P4	P4	P4-M1	M1
M-G2-419^f		✓										✓							✓
Subadult SH Preneanderthals	Total %	3 75.0	1 25.0									3 100							1 25.0	3 75.0
Adult SH Preneanderthals	Total %	15 100										4 25.0		9 56.3		3 18.8		1 4.8	2 9.5	3 14.3	15 71.4
Subadult Early Neanderthals	Total %	5 62.5	2 25.0		1 12.5	1 12.5	2 25.0				1 12.5	1 12.5		3 37.5			1 25.0		1 25.0		2 50.0
Adult Early Neanderthals	Total %	6 75.0	2 25.0									1 16.7		4 66.7	1 16.7					1 20.0	4^g 80.0
Subadult Late Neanderthals	Total %	7 38.9	7 38.9	3 16.7	1 5.6	7 35.0	5 25.0	3 15.0			1 5.0	2 10.0 Tuniz		2 10.0			3 33.3	2 22.2	3 33.3		1 11.1
Adult Late Neanderthals	Total %	9 41.0	11 50.0	2 9.0							5 16.0	5 16.0	1 3.2	15 48.4	3 9.7	2 6.5		2 15.4	2 15.4	3 23.1	6 46.2
Subadult Neanderthals (Early/Late)	Total %	12 46.2	9 34.6	3 11.5	2 7.7	8 28.6	7 25.0	3 10.7			2 7.1	3 10.7		5 17.9			3 25.0	2 16.7	4 33.3		3 25.0
Adult Neanderthals (Early/Late)	Total %	15 50.0	13 43.3	2 6.7							5 13.5	6 16.2	1 2.7	19 51.4	4 10.8	2 5.4	1 5.6	2 11.1	2 11.1	4 22.2	9 50.0
Subadult Middle Paleolithic Homo sapiens	Total %	2 100					1 50.0				1 50.0						1 (?) 100
Adult Middle Paleolithic Homo sapiens	Total %
Subadult Upper Paleolithic Homo sapiens	Total %	5 100				3 60.0	2 40.0
Adult Upper Paleolithic Homo sapiens	Total %	7 100							1 14.3		6 85.7
Subadult Homo sapiens (Middle/Upper Paleolithic)	Total %	7 100				3 42.9	3 42.9				1 14.3
Subadult recent Homo sapiens^h	L R Total %	26 28 54 90.0	4 2 6 10.0						2 2 4 7.0	24 21 45 75.0	2 6 8 13.0	2 1 3 5.0
Adult recent Homo sapiens^h	L R Totals% L R Totals%	29 30 59 98.0	1 1 2.0						2 1 3 2.0 1 1 2.0	22 18 40 23.0 15 8 23 38.0	40 46 86 49.0 10 19 29 48.0	21 21 42 24.0 4 3 7 12.0		2 1 3 2.0

• Abbreviations: L, left side; R, right side; SH, Sima de los Huesos.
• ^a Sample composition, observations by individual, and sources of the observations can be found in the Supplemental Data, Excel spreadsheet. Percentage occurrence values have been rounded, resulting in some instances where the total percentage occurrence is slightly greater or less than 100%.
• ^b Roman numerals are used for the number of foramina present to avoid confusion with the number of individuals with single (I), double (II), triple (III), or quadruple (IV) foramina. Totals in categories I-IV represent the number of preserved mandibular sides with that number of foramina, not the number of individuals. Right/left asymmetry in foraminal number occurs frequently.
• ^c Although some authors record occurrences of mental foramina inferior to a deciduous tooth as being below the adult equivalent (e.g., dm₁ = P₃, dm₂ = P₄), we distinguish between deciduous or adult dentition wherever possible. References to the original literature can be found in the Supplementary Data, Excel spreadsheet.
• ^d Anterior marginal tubercle data derive mainly from Rosas (1995, 2001). We have supplemented these data with observations from other literature sources. Reference to the original literature can be found in the Supplementary Data, Excel spreadsheet. Because there are so few subadults with anterior marginal tubercles, we opted to list occurrences of tubercles inferior to a deciduous tooth as being below the adult equivalent (e.g., dm₁ = P₃, dm₂ = P₄). The anterior marginal tubercle in Qafzeh 11 is described by Tillier (1984) as only a slight elevation of the mandibular corpus below the P₃. Whether this feature constitutes a true anterior marginal tubercle is questionable, given the complete absence of these tubercles in other Middle-to-Late Paleolithic H. sapiens.
• ^e Some foramina are offset slightly from M₁. To record such occurrences, we used a < or >sign to denote that the foramina are slightly mesial (<) or slightly distal (>) to the center of M₁.
• ^f In the M-G2-419 mandible, there are two small foramina near the mental foramen, but we do not consider these to be accessory mental foramina. Although these foramina clearly transmitted branches of the inferior alveolar nerve, they are not directly associated with the mental foramen.
• ^g The anterior marginal tubercle in Payre 15 is described as below the R_C—RP₃ interseptal region (Verna et al., 2020). However, we concur with Quam et al. (2023) that the tubercle is positioned below the RM₁. The position of the tubercle can be easily discerned in the figures accompanying the Payre 15 description (Verna et al., 2020).
• ^h Observations for the recent samples comprise (1) data that derive from Tebo and Telford (1950) (N = 87 mandibles, with observations on the right and left sides) and (2) data collected by the authors for the current study. Raw data for our recent subadult and adult samples are provided in the Supplemental Data, Excel spreadsheet.

The M-G2-419 mental foramen is centered inferior to the P₄-M₁ (Table 1 and Figure 1; Supplemental Data, Excel spreadsheet). A similar positioning is found in all SH Preneanderthal adolescents. In adult SH Preneanderthals, the foramen is most often positioned inferior to the M₁ (56.3%, 9/16). The remainder has the foramen positioned inferior to the P₄-M₁ (25.0%, 4/16) or M₁-M₂ (18.8%, 3/16) (Table 1; Supplemental Data, Excel spreadsheet). Recently, Quam et al. (2023) observed that the mental foramen is positioned inferior to the M₁ in 75% of adult SH Preneanderthals, while the remainder are inferior to the P₄-M₁. However, their tabulation does not distinguish those cases with the foramen positioned inferior to the M₁-M₂ interdental septum, as described by Rosas (2001) and reported here.

In the Early and Late Neanderthal samples, some of the foramina are inferior to the deciduous dentition. Although this is indicated in Table 1, we consider those occurrences inferior to the deciduous molars to be inferior to the premolars here in the text. In subadult Early Neanderthals, the foramen is most often positioned inferior to either the M₁ (37.5%, 3/8) or P₃-P₄ (25.0%, 2/8). Foramina positioned inferior to the P₃, P₄, or P₄-M₁ are also present at a frequency of 12.5% (1/8) each (Table 1). In adult Early Neanderthals, the mental foramen is most often positioned inferior to the M₁ (66.7%, 4/6). In only two cases from this sample does the foramen have a different position (P₄-M₁, 16.7% and distal M₁, 16.7%: Table 1). In Late Neanderthal subadults, the mental foramen is most often found inferior to the P₃ (35%, 7/20) or P₃-P₄ (25%, 5/20). However, mental foramina are also positioned inferior to the P₄ (20%, 4/20), P₄-M₁ (2/20, 10%), and M₁ (2/20, 10%). While SH Preneanderthals aged 12.5–17.0 years do not have the foramen positioned inferior to the M₁, this positioning does occur in the ≈9- to 10-year-old Krapina 53 Early Neanderthal. Alternatively, mental foramina positioned inferior to the M₁ have not been observed until ≈14–15.0 years of age in Late Neanderthals (cf. Malarnaud 1, Cova del Gegant 1). In adult Late Neanderthals, mental foramina most often occur inferior to the center of the M₁ (48.4%, 15/31). However, some are positioned slightly off the center of the M₁, either mesially (3.2%, 1/31) or distally (9.7%, 3/31). A small percentage of foramina are positioned more distally between the M₁-M₂ (6.5%, 2/31: Table 1). Deviations from this pattern are seen, with foramina positioned equally inferior to either the P₄ or P₄-M₁ (16%, 5/31).

In the combined subadult Middle-to-Upper Paleolithic H. sapiens sample, the mental foramen is most often positioned inferior to either the P₃ (42.9%, 3/7) or the P₃-P₄ (42.9%, 3/7) interdental septum. Only Qafzeh 11 has its foramen positioned inferior to the P₄ (Tillier, 1984, 1999). The sample of Middle-to-Upper Paleolithic adults includes only individuals from the Upper Paleolithic. In this sample, the foramen is positioned inferior to the P₄ in every specimen except Pavlov 3, in which the foramen is inferior to the P₃ (Table 1). In recent human adolescents, the mental foramen is most often positioned inferior to the P₃-P₄ (75%, 45/60). Foramina inferior to the P₃ (7.0%, 4/60), P₄ (13.0%, 8/60), or P₄-M₁ (5.0%, 3/60) were also observed. In recent adults, the most usual position of the mental foramen is inferior to the P₄ (48%, 29/60), although foramina inferior to the P₃-P₄ (38%, 23/60) are also observed in high frequency. A small percentage of individuals have the foramina positioned inferior to either the P₃ (2.0%, 1/60) or P₄-M₁ interdental septum (12.0%, 7/60). Comparing our recent human adult data to that of Tebo and Telford (1950) shows similar values, with two exceptions (Table 1). Our sample has a significantly lower percentage of individuals with foramina inferior to the P₃-P₄, but a significantly higher percentage of cases inferior to P₄-M₁.

Directly inferior to the mental foramen and centered on the RP₄, the M-G2-419 mandible has a pronounced and elongated anterior marginal tubercle (cf. submental tubercle of Gorjanović-Kramberger, 1906). The tubercle is 13 mm long (a-p) and demarcated superiorly by a short and shallow intertoral sulcus. The anterior marginal tubercle in M-G2-419 reaches its maximum height inferior to the RP₃-RP₄ interdental septum. Tubercle formation in the Krapina 54 Early Neanderthal subadult is slight. Alternatively, in Krapina 53, the anterior marginal tubercle is developed similarly to M-G2-419, even though its assessed developmental age is younger (≈10–13 years: Gorjanović-Kramberger, 1906; Smith, 1976).

Anterior marginal tubercles have been recorded in all adult European Middle Pleistocene Homo specimens sampled, except for Mala Balanica BH 1, in which the region is damaged (Roksandic et al., 2011, 2018). Of these cases, only Arago II has the tubercle located inferior to the M₁ (Rosas, 2001); the others are positioned equally inferior to the P₄ and P₄-M₁ (Rosas, 2001; Rosas & Bermúdez de Castro, 1998). In SH Preneanderthals, some of the smaller specimens do not possess an anterior marginal tubercle (Rosas, 1995, 2001). The M-G2-419 mandible has a more pronounced tubercle relative to the similarly aged and similarly gracile AT-172 individual. In SH Preneanderthal adolescents possessing an anterior marginal tubercle, it occurs most frequently inferior to the P₄-M₁ (75%, 3/4), with a single case inferior to the P₄ (25%, 1/4: Table 1 and Figure 5). In the adult SH Preneanderthals, 71.4% (15/21) have the tubercle inferior to the M₁, with P₄-M₁ being the next most frequent position (14.3%, 3/21). A small percentage of individuals in this group have the anterior marginal tubercle located inferior to either the P₃-P₄ interdental septum (4.8%, 1/21) or the P₄ (9.5%, 2/21).

FIGURE 5

Bar graph showing the percent occurrence of an anterior marginal tubercle inferior to different tooth positions in subadult and adult SH Preneanderthals, Early Neanderthals, and Late Neanderthals. These data represent the position of the anterior marginal tubercle on each preserved side of the mandible.

In Early Neanderthal subadults, the sample size is small, but the most frequent positioning of the tubercle is inferior to the M₁ (50%, 2/4). Tubercle positioning inferior to the P₃ (25%, 1/4) and P₄ (25%, 1/4) have also been recorded. Although positioning of the anterior marginal tubercle inferior to the M₁ is most frequent in the Early Neanderthal subadults, both specimens with this tubercle position are from Krapina. Tubercle positioning in adult Early Neanderthals shows little variation, with 80.0% (4/5) of the tubercles positioned inferior to the M₁. Only the Krapina 57 adult Early Neanderthal has a more anteriorly positioned anterior marginal tubercle (P₄-M₁). In Late Neanderthal subadults, we recorded the tubercle's position relative to either the deciduous or adult dentition (Supplemental Data, Excel spreadsheet). Here we collapse these categories, counting deciduous molars as the corresponding premolar, and find an anterior marginal tubercle is most often positioned inferior to either the P₃ (33.3%, 3/9) or the P₄ (33.3%, 3/9). Tubercles inferior to the P₃-P₄ interdental septum (22.2%) are the next most frequent. Only a single occurrence of a tubercle inferior to the M₁, accounting for 11.1% of the sample, was observed in the ≈15.0-year-old Cova del Gegant adolescent (Daura et al., 2005). In adult Late Neanderthals, the anterior marginal tubercle, is mainly positioned inferior to the center of the M₁ (46.2%, 6/13) or P₄-M₁ (23.1%, 3/13). Other positions of the tubercle in this group have equal frequencies and are equally divided between the P₃-P₄ and P₄ (15.4%, 2/13). In the Late Neanderthal children Archi 1 (Arnaud, 2015; Ascenzi & Segre, 1971), Devil's Tower 1 (Minugh-Purvis, 1988), and Le Fate 1(+2) (Giacobini et al., 1984; Giacobini & de Lumley, 1983), the tubercle is specifically noted as well developed. None of the Early Neanderthal subadults have been described similarly.

In Middle Paleolithic H. sapiens subadults, Qafzeh 11 has been described as possessing a weakly developed anterior marginal tubercle beneath the P₃ (Tillier, 1984), while Qafzeh 10 lacks tubercle development.

The width of the inferior border of the corpus in M-G2-419 is ≈11.5 mm (includes the anterior marginal tubercle) but thins to ≈4.5 mm at the fractured posterior extent of the corpus. The thickness just posterior to the anterior marginal tubercle is 10.0 mm. The inferior border is narrow and flat laterally, but it angles superomedially along its medial edge to transition into the weakly developed inferior mandibular torus. In the similarly aged Le Moustier 1 Late Neanderthal, the inferior border is wider at the tubercle (13.5 mm), but the border thins posteriorly, as in M-G2-419 and other Neanderthals. Although showing some similarities to M-G2-419, the inferior border in Le Moustier 1 is broader and flatter anteriorly, making it more like adult Late Neanderthals than like M-G2-419. The Moula-Guercy mandible is markedly more gracile, and differences in the construction of the inferior corpus between these individuals likely reflect, in part, the difference in sex assessed for these individuals. A gradual decrease in basal corpus breadth that results in a very thin gonial region is considered a derived character of European Neanderthals (Piveteau, 1957; Rosas, 1992, 2001). Thinning of the gonial region cannot be confirmed in M-G2-419 due to the incomplete inferior border.

The lingual surface of the M-G2-419 mandible is well preserved. The mylohyoid line is faintly marked. This line takes a steep posterosuperior course from the base of the corpus and inferior to the mesial edge of the RP₄ to near the anterior third of the RM₃ crypt. Smith (1976) observed that the anterior mylohyoid line in Krapina Early Neanderthals generally extends from the inferior M₁, although some extend from inferior to the P₄. The terminal position of the mylohyoid line in M-G2-419 is obscured because it intersects a portion of the alveolar margin resorbed in preparation for RM₃ eruption. The mylohyoid line is also interrupted by a narrow groove that parallels the alveolar margin and runs inferior to the mesial RM₃. Following this groove, the mylohyoid line runs a more horizontal course ≈6–7 mm inferior to the intact alveolar ridge (i.e., the non-resorbed region at M₂). The groove in the mylohyoid line likely contained the lingual nerve, although this would constitute a slightly anterior position of the nerve relative to that in recent humans (Sittitavornwong et al., 2017).

The posterior configuration of the mylohyoid line in SH Preneanderthals differs from M-G2-419 and Neanderthals, in general. In the SH Preneanderthal group, the posterior end of the mylohyoid line is inferiorly placed relative to the alveolar ridge (6.1–15.5 mm, AT-505 and 605, respectively; Rosas, 2001), and the terminal portion arcs superiorly to reach the alveolar margin. In Neanderthals, in general, Rosas (2001) notes that the angle and course of the mylohyoid line is most often diagonal, as in SH Preneanderthals. However, the mylohyoid line in some Neanderthals can be seen to take a steeper and straighter course prior to its termination near the alveolar margin. Based on data in Rosas (2001), the mylohyoid line in 87.5% (7/8) of Early Neanderthals (La Chaise BD 1, Ehringsdorf F, Krapina 55, 57–59, Tabun I) is intermediate in position relative to SH Preneanderthals and Late Neanderthals. This morphology is also consistent with that in the Early Neanderthal Payre 15 (Verna et al., 2020). La Naulette 1 is the only Early Neanderthal with a mylohyoid line that is described as “high” and, therefore, like the Late Neanderthals. However, note that 38.9% (7/18: Rosas, 2001) of Late Neanderthals have either a low or medium height level at the M₃. Further, in the subadult Late Neanderthals Roc de Marsal 1 and Devil's Tower 1, the former has a mylohyoid line that is steeply inclined and approximates the alveolar margin, while the latter has an initially steeply inclined mylohyoid line that flattens and terminates ≈8.0 mm from the alveolar margin.

The anterior extent of the mylohyoid line in M-G2-419 flattens such that it runs parallel to the basal border prior to fading into the lingual corpus inferior to the mesial RP₄. An inferior mandibular torus is present but poorly developed. A shallow sulcus separates the inferior border of the corpus from the inferior mandibular torus. This sulcus defines a weakly expressed V-shape between the inferior border and the inferior mandibular torus. This configuration has been noted as a distinctive feature of Neanderthals, in general (Smith, 1976; Suzuki, 1970). In M-G2-419, the region just anterior to this depressed region constitutes the posterior-most extent of the digastric fossa. The inferior mandibular torus is slightly developed anteriorly, and it fades into the corpus inferior to the RM₂-RM₃ interdental region. Further, the anterior origin of the mylohyoid line is positioned ≈2 mm superior to the inferior mandibular torus in this anterior region. In the SH Preneanderthals, the mylohyoid line does not contact the inferior torus and the torus is weakly expressed in the adolescent female AT-172; however, in the adult AT-250 the mylohyoid line connects to the superior portion of the inferior mandibular torus (Rosas, 1995). In the Krapina Early Neanderthals (cf. subadults 53, 54; adults 58–59), the mylohyoid line coincides with the inferior mandibular torus (Smith, 1976). In the subadult Late Neanderthal Roc de Marsal 1, the mylohyoid line intersects the inferior mandibular torus, while the line and torus are separate in Devil's Tower 1 and Le Moustier 1, as in M-G2-419. In adult Late Neanderthals, the usual condition is for the anterior mylohyoid line to intersect the inferior mandibular torus.

In M-G2-419, the inferior mandibular torus and mylohyoid line define the inferior and superior borders of the submandibular fossa, respectively. The inferior mandibular torus also defines the inferior border of the submandibular fossa in the Krapina sample (Gorjanović-Kramberger, 1906). The submandibular fossa in M-G2-419 is a short and shallow, but well-defined, concavity positioned inferior to the RM_1-3 and the degree to which the mylohyoid line overhangs the submandibular fossa is minimal. The Early Neanderthals Krapina 52, 53, 55, and 57 all have a strong medial projection of the corpus at the mylohyoid line, which overhangs the submandibular fossa. Smith (1976) notes that it might be best to refer to these mylohyoid lines in the Krapina sample as ridges, due to their developed nature. Alternatively, Krapina 56 and 58 have less projection of the medial corpus at the mylohyoid line. In the Payre 15 Early Neanderthal (Verna et al., 2020), the mylohyoid line region does not project, and the medial corpus is flatter than M-G2-419. In the subadult Late Neanderthals Roc de Marsal 1, Devil's Tower 1 and Le Moustier 1, the mylohyoid line region strongly overhangs the submandibular fossa. In Late Neanderthal adults, the medial corpus varies from flat with virtually no depth to the submandibular fossa to a strong overhang (cf. Regourdou 1 vs. La Quina H5).

In M-G2-419, the sublingual fossa, like the submandibular fossa, is apparent but weakly expressed. The sublingual fossa (14.0 × 5.0 mm) lies inferior to the P₃-M₁ and is delimited inferiorly by the mylohyoid line.

At the posterior extent of the medial corpus, the pterygoid fossa extends as a shallow concavity from the base of the RM₃ crypt to the anterior edge of the medial pterygoid insertion. A very shallow mylohyoid groove courses inferiorly from the fractured edge near the mandibular foramen. The anterior-most medial pterygoid muscle insertion site is smooth and unmarked by ridges for attachment of medial pterygoid muscle septa, as detailed in Richards et al. (2003): cf. Le Fate 1(+2), Giacobini and de Lumley (1983). In the SH Preneanderthals, Rosas (2001) observed that the pterygoid fossa is shallow (90.9%, 10/11), with only the adult AT-792 noted as having a deep fossa. In Early Neanderthals, he only described Ehringsdorf F (33.3%, 1/3) as having a shallow pterygoid fossa. In the Late Neanderthals, only La Quina H9 and Malarnaud 1 are noted as having a shallow pterygoid fossa (11.8%, 2/17: Rosas, 2001).

3.1.2 Comparative metric analysis

Comparative metrics can be found in Tables 2 and 3, and the Supplemental Data, Excel Spreadsheet. Here we report on our analysis of metric dimensions related to the mental foramen and mandibular corpus at M₂. Refer to the Supplemental Data for a metric analysis of measurements taken at P₄-M₁ and M₁. Metrics for subadult fossils are presented in the text and Supplemental Data but not in the tables.

TABLE 2. Sample size (N), range, mean ($$$ \overline{x} $$$), and standard deviation (sd) for mandibular dimensions (in mm) in adult European and Middle Eastern Middle-to-Late Paleolithic (≈MIS 14—MIS 2/≈1) Homo and subadult and adult humans^a.

		P4-M1 height	P4-M1 thickness	Robusticity index at P4-M1	M1 height	M1 thickness	Robusticity index at M1	M1-M2 height	M1-M2 thickness	Robusticity index at M1-M2	M2 height M69 (2)	M2 thickness M*69b	Robusticity index at M2	M2-M3 height	M2-M3 thickness	Robusticity index at M2-M3
M-G2-419		25.9	12.9	49.8	23.4	13.5	57.7	23.0	13.3	57.8	23.8	14.1	59.2	24.6	15.1	61.4
SH Preneanderthals	N Range $$$ \overline{x} $$$ sd				7 26.7–37.1 31.0 3.53	7 14.4–17.1 15.9 1.03	7 45.0–56.1 51.6 3.90				12 27.2–37.0 31.3 3.28	13 15.6–20.8 17.4 1.33	12 46.8–61.4 56.1 4.48
Early Neanderthals	N Range $$$ \overline{x} $$$ sd				4 31.4–37.9 34.5 3.35	3 16.0–16.7 16.3 0.35	3 43.0–50.5 47.9 4.27	3 30.2–35.6 32.3 4.11	3 14.4–16.9 16.0 1.33	3 44.2–47.7 50.0 4.65	7 27.0–38.5 31.8 4.22	7 14.5–18.0 15.8 1.27	7 44.8–59.3 50.1 5.08	4 28.4–34.0 30.3 2.23	4 14.0–16.5 15.5 1.08	4 41.2–55.9 43.4 15.06
Late Neanderthals	N Range $$$ \overline{x} $$$ sd	4 26.4–31.5 28.1 2.37	3 12.7–15.6 14.1 1.45	3 48.1–58.6 52.4 5.52	11 25.4–35.5 32.1 2.92	12 13.3–16.8 15.1 1.39	11 38.0–55.9 47.4 6.14	7 26.1–32.7 28.8 2.78	7 13.6–17.9 15.2 1.38	6 47.2–59.8 54.3 4.53	12 28.8–35.0 32.9 1.81	14 12.7–19.0 16.3 13.96	12 43.3–57.6 49.3 5.17	9 26.0–33.0 30.1 2.71	8 14.0–19.0 15.8 1.41	8 48.0–60.0 53.3 4.95
Early/Late Neanderthals	N Range $$$ \overline{x} $$$ sd				16^b 24.6–37.9 32.2 3.63	16 12.0–17.2 15.1 1.55	15 38.0–55.9 47.6 5.45	10 26.1–35.6 30.3 3.63	10 13.6–17.9 15.4 1.29	9 44.2–59.8 51.7 5.38	20 27.0–38.5 32.4 3.07	22 12.7–19.0 16.2 1.70	20 43.3–59.3 50.2 5.47	14 26.0–34.0 29.9 2.67	13 14.0–19.0 15.7 1.22	13 41.2–60.0 53.1 5.35
Middle Paleolithic H.s.	N Range $$$ \overline{x} $$$ sd										2 29.0–34.5	2 13.0–17.0	2 38.0–58.6	1 34.5	1 13.0	1 37.7
Upper Paleolithic H.s.	N Range $$$ \overline{x} $$$ sd							7 23.4–33.5 29.4 3.96	7 11.4–16.6 13.4 2.01	7 34.0–70.9 46.9 12.40	3 26.3–28.5 27.6 1.17	3 13.3–17.2 15.1 1.97	3 50.6–61.2 54.6 5.78
Middle-to-Upper Paleolithic H.s.	N Range $$$ \overline{x} $$$ sd							7 23.4–33.5 29.4 3.96	7 11.4–16.6 13.4 2.01	7 34.0–70.9 46.9 12.40	5 26.3–34.5 29.3 3.09	5 13.0–17.2 15.1 1.98	5 38.0–61.2 52.0 11.69
Recent Homo sapiens subadults (10–18 years)^c	N Range $$$ \overline{x} $$$ sd	30 17.1–29.0 24.0 2.74	30 9.0–16.1 12.7 1.68	30 38.5–68.1 53.5 7.99	30 16.0–27.1 22.7 2.65	30 9.7–16.7 14.1 1.66	30 49.6–91.9 62.8 10.28	30 16.7–26.3 22.6 2.65	30 12.4–19.2 15.2 1.89	30 52.8–92.3 68.0 9.76	30 16.5–25.5 21.9 2.51	30 11.6–19.2 16.1 1.66	30 59.8–93.6 74.1 9.31	22 17.4–27.9 24.0 2.60	22 11.1–19.4 15.7 1.69	22 51.3–95.1 65.8 8.85
Recent Homo sapiens adults^c	N Range $$$ \overline{x} $$$ sd	30 20.7–35.0 29.5 3.97	30 11.6–16.8 14.0 1.52	30 36.7–71.1 48.1 6.91	30 20.1–31.7 28.1 2.76	30 13.1–18.1 15.3 1.30	30 42.5–68.6 55.0 6.01	30 20.3–32.8 27.3 2.94	30 13.8–20.0 16.1 1.55	30 45.3–72.4 59.5 6.80	30 20.2–30.7 25.6 2.89	30 15.1–20.7 17.9 1.76	30 50.0–82.9 70.6 8.11	30 19.6–31.7 26.0 2.83	30 14.4–20.6 18.2 1.74	30 54.9–80.2 70.2 5.64

• Abbreviations: $$$ \overline{x} $$$, sample mean; N, sample size; sd, standard deviation; SH, Sima de los Huesos.
• ^a Sample composition, metric values by specimen, and references to literature-derived data are listed in the Supplemental Data, Excel spreadsheet.
• ^b Sample totals for the combined sample include the Early/Late Neanderthal Tabun CI, which is not included in either the Early or the Late Neanderthal tabulations.
• ^c All recent H. sapiens data were collected by the authors for the current study. Metric values and observations are listed in the Supplemental Data, Excel spreadsheet. Refer to the Methods section for information on how developmental ages were determined.

TABLE 3. Sample size (N), range, mean ($$$ \overline{x} $$$), and standard deviation (sd) for mandibular dimensions (in mm) related to the mental foramen for adult European and Middle Eastern Middle-to-Late Paleolithic (≈MIS 14—MIS 2/≈1) hominins and subadult and adult humans^a.

		Height at mental foramen M69 (1)^b	Thickness at mental foramen M69 (3)	Robusticity index at mental foramen M166	Mental foramen-basal margin height^b	Mental foramen-alveolar margin height^b
M-G2-419		25.9	12.9	49.8	10.3	12.4
SH Preneanderthals	N	13	13	13	5	5
	Range	28.8–38.0	14.8–21.8	45.0–57.7	10.9–16.3	16.9–22.5
	$$$ \overline{x} $$$	32.3	16.8	52.0	12.7	18.6
	sd	3.77	1.77	4.28	2.12	2.50
Early Neanderthals	N	6	6	6	1	1
	Range	29.2–42.5	14.7–18.8	38.6–54.8	12.8	14.3
	$$$ \overline{x} $$$	34.1	16.3	48.2
	sd	4.71	1.44	5.74
Late Neanderthals	N	19	19	19	1	5
	Range	26.4–38.0	12.3–16.2	38.4–58.1	17.9	14.5–18.5
	$$$ \overline{x} $$$	32.7	15.4	47.3		16.5
	sd	3.52	1.47	4.87		1.71
Early/LateNeanderthals	N	26^c	26	26
	Range	26.4–42.5	12.3–18.8	38.4–64.7
	$$$ \overline{x} $$$	32.7	15.7	48.4
	sd	4.10	1.45	6.06
Middle Paleolithic Homo sapiens	N	2	2	2
	Range	35.0–36.0	13.2–16.6	36.7–47.4
	$$$ \overline{x} $$$
	sd
Upper Paleolithic Homo sapiens	N	9	9	8
	Range	25.2–36.5	10.0–13.5	34.7–50.0
	$$$ \overline{x} $$$	30.0	12.0	41.5
	sd	3.74	0.98	5.54
Middle-to-Upper Paleolithic Homo sapiens	N	11	11	10
	Range	25.2–36.5	10.0–16.6	34.7–50.0
	$$$ \overline{x} $$$	31.0	12.6	41.6
	sd	4.02	1.64	5.50
Recent Homo sapiens subadults (10.1–18.0 years)^d	N	30	30	30	30	30
	Range	15.9–30.8	9.4–14.4	36.1–63.2	9.0–14.4	6.0–16.1
	$$$ \overline{x} $$$	25.0	11.7	47.1	11.9	11.2
	sd	2.77	1.44	6.58	1.44	2.14
Recent Homo sapiens adults^d	N	30	30	30	30	30
	Range	20.1–34.5	10.4–16.2	31.1–56.3	8.0–17.0	9.8–19.7
	$$$ \overline{x} $$$	29.7	13.3	45.0	13.6	14.0
	sd	3.46	1.61	5.88	2.09	2.60

• Abbreviations: $$$ \overline{x} $$$, sample mean; N, sample size; sd, standard deviation; SH, Sima de los Huesos.
• ^a Sample composition, metric values by specimen, and references to literature-derived data can be found in the Supplemental Data, Excel spreadsheet.
• ^b In many cases, the published values for the height from the alveolar ridge to the mental foramen and the mental foramen to the basal border are confusing, because their combined length is greater than the stated height from the alveolar ridge to the basal border.
• ^c Sample totals for the combined sample include the Early/Late Neanderthal Tabun CI, which is not included in either the Early or the Late Neanderthal tabulations.
• ^d All recent H. sapiens data were collected by the authors for the current study. Metric values and observations are listed in the Supplemental Data, Excel spreadsheet. For the heights from the mental foramen to the basal border and the alveolar margin, we measured the distance from the bottom or top of the mental foramen, respectively. The combined distance can be subtracted from the height at the mental foramen to obtain the height of the foramen. Refer to the Methods section for information on how developmental ages were determined.

In M-G2-419, the height, thickness, and index of robusticity at the mental foramen are 25.9 mm, 12.9 mm, and 49.8, respectively (Tables 2 and 3). Mandibular height at the mental foramen shows little difference between the adult SH Preneanderthals and Late Neanderthals. The Early Neanderthals have only a slightly greater mean height relative to either SH Preneanderthals or Late Neanderthals, but this might be due to a smaller sample size. Corpus thickness at the mental foramen decreases slightly through the adult SH Preneanderthal, Early Neanderthal, and Late Neanderthal sequence. These changes are reflected in a slight, sequential decrease in the robusticity index at the mental foramen (Figure 6). The M-G2-419 specimen falls below the range of variation for height at the mental foramen relative to the SH Preneanderthal adolescents and adults and the Early and Late Neanderthal adults. Alternatively, while falling below the range of variation in SH Preneanderthal and Early Neanderthal adults, the M-G2-419 corpus thickness at this position falls at the bottom of the adult Late Neanderthal range of variation. Considering subadult Early Neanderthals aged 9–17 years, the M-G2-419 height at the mental foramen is most like the 14- to 16-year-old Krapina 54. The height is also nearly identical to the mean value for this group (N = 5, mean = 25.8 mm, SD = 2.14). Alternatively, the thickness at the mental foramen for M-G2-419 falls below the range of variation for this age group, being like the much younger La Chaise Abri Suard S13 (4–5.0 years). Because the M-G2-419 mandible is so thin at the mental foramen, the robusticity index (49.8) falls below the range of variation for subadult Early Neanderthals (Table 3, Figure 6). The robusticity index for these subadults decreases from a range of 61.0–69.5 in ages younger than 11.0 years to 53.4–56.3 for ages from 12 to 17.5 years. Comparing M-G2-419 to the Late Neanderthal subadults reveals a more complex pattern. Values for the height at the mental foramen reach a value comparable to M-G2-419 by 7–8.0 years of age (cf. Combe Grenal 1, 26.9 mm). The mean height for the 7- to 17.0-year-olds (N = 8, mean = 24.5, SD = 2.67, range 21–29.0 mm) does not differ significantly from that of the Early Neanderthal subadults, although the age range is slightly younger in the late group. Considering the thickness at the mental foramen, the Late Neanderthal subadults range from 11.5–13.8 mm until ≈9.0-years of age, whereafter the thickness increases to a range of 15–19.7 mm. The M-G2-419 thickness at the mental foramen is most like values recorded for Le Fate 1(+2) and Scladina 4A-1, with developmental ages of 8–10.0 years. The robusticity index for the Late Neanderthal subadults decreases with age, as in the Early Neanderthals, but the decrease is not as consistent, and it does not decrease to the same extent (Figure 6). Part of this difference may be related to the differing age profiles of the Early Neanderthal and Late Neanderthal samples. In Le Moustier 1, which has a similar developmental age as M-G2-419, the corpus at the mental foramen is much taller and wider and the robusticity index is high for its assessed age.

FIGURE 6

Bivariate plot of the robusticity index at the mental foramen by developmental age for SH Preneanderthals, Early Neanderthals, Late Neanderthals, Middle Paleolithic Homo sapiens, Upper Paleolithic H. sapiens, and recent H. sapiens.

FIGURE 7

Bivariate plot of the robusticity index at the mandibular first adult molar by developmental age for SH Preneanderthals, Early Neanderthals, Late Neanderthals, Middle Paleolithic Homo sapiens, Upper Paleolithic H. sapiens, and recent H. sapiens.

In M-G2-419, the index of the vertical height of the mental foramen from the inferior corpus is 39.8 (height from basal margin to foramen/total height of corpus at the mental foramen × 100). Published data by specimen is minimal for the corpus height inferior to the foramen (Supplemental Data, Excel spreadsheet). The adult Ehringsdorf F Early Neanderthal is the only specimen with an index value (=42.0) and it is similar to M-G2-419. Recently, Quam et al. (2023) provided summary data (means, ranges) for this index in SH Preneanderthals and other Middle-to-Late Pleistocene Homo. Assessment of their data shows the foramen in SH Preneanderthals to be placed slightly lower on the corpus relative to their combined Early—Late Neanderthal sample. However, no significant differences were observed by these authors between any of the groups studied. The index value for the M-G2-419 mandible falls near the middle of the range for adult SH Preneanderthals (range = 35.3–45.7) and in the lower third of their combined adult Early—Late Neanderthal sample (range = 35.8–49.5).

Note, however, that the Quam et al. (2023) study only included mid-to-late-stage adolescents and adults. Considering only adolescent SH Preneanderthals, the height from the basal border to the mental foramen ranges from 9.6 to 11.9 mm, while the range for the distance from the alveolar ridge to the foramen is 15–17.7 mm. These heights in the M-G2-419 mandible are 10.3 and 12.4 mm, respectively (Table 3). So, while the basal height to the foramen is similar to the SH Preneanderthals, the height from the foramen to the alveolar margin is below the range of variation. The only Early Neanderthal that has both values available is Ehringsdorf F (12.8 mm basal border to foramen, 14.3 mm foramen to alveolus) and, although it is an adult, these two dimensions show a similar relationship as in M-G2-419. In our recent H. sapiens sample, the mental foramen is positioned inferior to the center of the corpus in ≈50% of cases. The height of the mental foramen on the corpus shows a minor correlation with age increase.

Mean values for the height and thickness of the corpus at M₂ are quite similar in the adult SH Preneanderthals, Early Neanderthals, and Late Neanderthals, but the mean index of robusticity is higher in SH Preneanderthals due to a slightly thicker, slightly shorter corpus. The M-G2-419 corpus height at M₂ is much shorter than in adult SH Preneanderthals, Early Neanderthals, or Late Neanderthals. The thickness of the corpus at M₂ in M-G2-419 falls just below the range of variation for SH Preneanderthals and Early Neanderthals, although it is closest to the Early Neanderthals. The thickness value is easily encompassed within the range of variation for adult Late Neanderthals. The robusticity index at the M₂ shows the same pattern as at the M₁, wherein the SH Preneanderthals have a higher value, while the Early Neanderthals and Late Neanderthals have similar mean values (Figure 8). The M-G2-419 index of robusticity (59.2) is, again, more like the adult SH Preneanderthals (Table 2, Figure 8).

FIGURE 8

Bivariate plot of the robusticity index at the mandibular second adult molar by developmental age for SH Preneanderthals, Early Neanderthals, Late Neanderthals, and recent Homo sapiens.

Corpus height, thickness, and robusticity index at M₂ in SH Preneanderthal adolescents is only available for AT-23, and these values follow the trends in the adult data (Supplemental Data, Excel spreadsheet). Considering these measurements for the two Early Neanderthal subadults in the sample, the height at M₂ in M-G2-419 is nearly the same as that for Ehringsdorf G (24.0 mm) and La Naulette 1 (23.0 mm). However, the corpus thickness at M₂ in M-G2-419 is much narrower than in either of these Early Neanderthals (range 16–17.0 mm: Supplemental Data, Excel spreadsheet). Both individuals have high indices of robusticity at M₂ relative to the M-G2-419 value and, in this limited sample, the index value does not decrease with an increase in individual age. Examination of corpus height at M₂ for the Late Neanderthal subadults shows that the value for M-G2-419 is near that for Malarnaud 1 (22.0 mm). The thickness of the corpus at M₂ in Late Neanderthal subadults has a range of 14–15.2 mm, with the youngest individual, Zaskalnaya VI-72 (=14.0 years; Kolossov et al., 1975) having a value like M-G2-419. In this group, only Malarnaud 1 and Zaskalnaya VI-72 have index values (68.2 and 70.0, respectively), and both are substantially higher than the M-G2-419 value, although they have relatively similar developmental ages (Figure 8). In sum, the robusticity index in M-G2-419 is lower than in any Early or Late Neanderthal subadult, but it is within the range of SH Preneanderthal adolescents.

Sample sizes for adult Middle—Upper Paleolithic H. sapiens measurements at the mental foramen, M₁, and M₂ are small (Tables 2 and 3, Figures 6-8). The M-G2-419 height at the mental foramen falls at the lowest end of the adult range of variation in this combined sample. Thickness values at the mental foramen show M-G2-419 to fall near the mean. The index of robusticity at the mental foramen for adults in this group is much lower than that for the SH Preneanderthal, Early Neanderthal, and Late Neanderthal sequence (range = 34.7–50.0). The M-G2-419 mandible falls just inside the upper end of this range of index values. Height values at the M₂ in the combined Middle—Upper Paleolithic adults show that M-G2-419 differs substantially from this group (N = 5, range 26.3–34.5 mm: Tables 2 and 3). Alternatively, both the corpus thickness (N = 5, range 13.0–17.2 mm, Table 2) and the index of robusticity at M₂ (N = 5, range 38.0–61.2) for this combined sample comfortably contain the M-G2-419 values. The height values at the mental foramen for subadults in this combined Middle—Upper Paleolithic sample are similar. Values for Qafzeh 4 and 10 bracket the M-G2-419 value, while that for La Quina-Aval 4 is slightly shorter. However, the Qafzeh individuals are younger (6.5–7.5 and 12.0, respectively) while La Quina-Aval 4 is aged at ≈5.5–6.5 years (Verna et al., 2012). Thickness values at the mental foramen and those for the robusticity index in Qafzeh 4 and La Quina-Aval 4 are the most like M-G2-419. Data for height (N = 10, mean = 22.6 mm, SD = 3.132) and thickness (N = 10, mean = 14.5 mm, SD = 1.543) of the corpus at M₁ for subadults in this sample are like those in M-G2-419. However, the index of robusticity for this combined group (N = 10, mean = 65.5, SD = 11.544) is higher than that for M-G2-419. In general, the index of robusticity at M₁ decreases in this group with developmental age, except for Les Rois A (70.6: Minugh-Purvis, 1988). Predmostí 25 (Matiegka, 1934) and Sungir 3 (Minugh-Purvis, 1988) have values for the robusticity index at M₁ that are more like that of M-G2-419.

The M-G2-419 values for corpus height, thickness, and the robusticity index at the mental foramen are like the mean values for adult recent H. sapiens. Mean values for corpus height and thickness at P₄-M₁ and M₁ in recent adults are only slightly greater than those in M-G2-419, but the indices of robusticity are slightly lower. Comparing the mean height, thickness, and index of robusticity values at the M₂ between adult recent H. sapiens and M-G2-419 shows the thickness to be increasing relative to the height. This results in a significant increase in the robusticity index in the recent sample. The earlier H. sapiens sample has mean index values at M₁ and M₂ that are more similar, unlike the situation in recent H. sapiens. This means that the Middle—Upper Paleolithic mandibles have a sequence of index values that are more like those in Neanderthals. In subadult recent H. sapiens, the mean values for the height, thickness, and robusticity index at the mental foramen and M₁ show the same pattern as in the adult sample.

3.1.3 Morphological assessment of the M-G2-419 M₁-M₃

The M-G2-419 M₁-M₂ crowns have a rounded square shape, while the M₃ has a nearly round shape. Crown shape is driven by the M₁-M₂ having generally short mesiodistal lengths relative to their buccolingual breadths (Table 4, Figure 9). In contrast, Martinón-Torres et al. (2012) observed that the ovate-rounded rectangular shape of the SH Preneanderthals is due to their being mesiodistally elongated. The M-G2-419 M₁ shows a slight degree of occlusal wear consistent with Molnar's (1971) Grade 2 wear stage. The M₂ shows very slight rounding of the cusp tips and ridges that equate with Grade 1 wear. The M₃ is unworn and unerupted.

TABLE 4. Mesiodistal and buccolingual crown dimensions (in mm) and crown index for mandibular molars (M₁–M₃) for Middle Pleistocene—Upper Paleolithic hominins from Europe and the Middle East^a.

		M1	M1	M1	M2	M2	M2	M3	M3	M3
		Mesiodistal length	Buccolingual breadth	Crown index	Mesiodistal length	Buccolingual breadth	Crown index	Mesiodistal length	Buccolingual breadth	Crown index
Group		M81	M81 (1)	M81 (1)/M81*100	M81	M81 (1)	M81 (1)/M81*100	M81	M81 (1)	M81 (1)/M81*100
M-G2-419		10.45	(10.3)	98.60	11.1	(11.6)	104.5	10.5	(11.3)	107.60
M-D1-230		10.60	(11.6)	109.40
M-L4-TNN5		10.90	(11.9)	109.20
Middle Pleistocene Homo^b	N	10	10	10	6	6	6	7	7	7
	Range	11.0–13.8	10.5–13.1	84.00–108.26	11.5–14.6	10.9–13.9	88.0–95.20	10.5–13.2	9.7–13.70	82.31–113.22
	$$$ \overline{x} $$$	12.10	11.73	97.08	12.5	11.62	92.62	12.30	11.29	91.91
	sd	0.75	1.02	7.93	1.10	1.20	2.76	0.92	1.35	10.33
SH Preneanderthal	N	8	7	7	4	4	4	6	6	6
	Range	11.0–12.0	10.0–11.4	91.00–100.94	10.7–11.4	9.9–10.8	90.35–94.74	10.5–12.7	9.2–11.3	85.98–91.74
	$$$ \overline{x} $$$	11.33	10.81	96.14	11.08	10.25	92.55	11.13	9.88	88.76
	sd	0.55	0.43	3.12	0.38	0.40	1.93	0.79	0.77	2.20
Early Neanderthal	N	44	43	43	35	34	34	33	31	31
	Range	10.5–13.6	9.0–14.2	82.00–109.4	9.3–14.0	9.8–14.2	75.97–130.11	8.3–13.9	8.0–13.1	80.58–128.64
	$$$ \overline{x} $$$	11.86	11.35	95.55	11.88	11.75	99.95	11.19	11.00	98.79
	sd	0.88	1.10	6.46	1.16	1.02	13.86	1.13	1.27	13.66
Early/Late Neanderthal	N	2	2	2	2	2	2	2	2	2
	Range	10.0-10.4	10.2–10.5	98.08–105.00	11.2–11.5	10.6–10.8	93.91–94.64	10.3–10.9	9.8–10.3	89.91–100.00
Late Neanderthal	N	59	60	59	55	56	55	38	38	38
	Range	8.0–12.5	7.4–11.9	82.22–135.00	10.4–13.7	9.2–12.6	83.30–109.60	9.7–13.0	8.0–13.0	81.50–113.04
	$$$ \overline{x} $$$	11.15	10.67	96.12	11.60	10.93	94.51	11.67	11.14	96.38
	sd	1.06	0.90	8.86	0.73	0.79	6.05	0.73	1.06	8.43
Middle Paleolithic	N	22	21	21	13	13	13	10	10	10
	Range	10.5–13.2	10.5–12.8	86.15–104.07	10.2–12.6	10.0–12.4	88.50–107.48	10.3–12.9	9.9–11.9	84.50–98.06
	$$$ \overline{x} $$$	11.91	11.58	96.89	11.28	11.18	99.27	11.72	10.80	92.29
	sd	0.77	0.64	5.35	0.78	0.74	5.56	0.96	0.73	4.25
Upper Paleolithic	N	70	71	70	67	69	66	36	40	35
	Range	9.8–13.0	9.3–12.0	83.33–117.00	7.4–13.3	8.7–12.4	81.20–126.32	8.9–14.1	9.1–12.9	73.81–110.00
	$$$ \overline{x} $$$	11.52	11.10	96.19	10.90	10.83	100.05	11.34	10.78	95.55
	sd	0.82	0.61	7.24	1.35	0.93	8.98	1.24	0.90	7.52

• Abbreviations: $$$ \overline{x} $$$, sample mean; M₁, permanent lower first molar; M₂, permanent lower second molar; M₃, permanent lower third molar; N = sample size; sd, standard deviation.
• ^a Metric values by individual, MIS stage, and references to literature sources are provided in the Supplemental Data, Excel spreadsheet. We provide the raw observations (Fossil Dentition Tab) and the totals and summary statistics by group (Fossil Dent Stats Tab). Note that some degree of caution should be exercised when using these data. This is because there are some significant differences between observers for some values. To avoid such problems in the mandibular assessments, we compiled data by investigator and then made determinations as to the likely correct values. In compiling the tooth data, we only recorded a single observer's values. In a couple of cases, we opted to change an observer's dimensions when it was clear the value was unusual and other investigators had values that were both consistent across a range of observers and consistent with the range of values for other individuals in a particular sample.
• ^b Refer to the Methods section for the range of MIS stages employed to define individual groups.

FIGURE 9

Bivariate plots of the mesiodistal length and buccolingual breadth of the (a) M₁, (b) M₂, and (c) M₃ for Middle Pleistocene, SH Preneanderthals, Early Neanderthals, Early/Late Neanderthals, Late Neanderthals, Middle Paleolithic Homo sapiens, and Upper Paleolithic Homo sapiens.

The M-G2-419 M₁ has is a deep anterior fovea (Grade 4) due, in part, to a well-developed mesial marginal ridge. Deep anterior foveae are found in 85% of SH Preneanderthals and 81.3% of Neanderthals (Martinón-Torres et al., 2007). Anterior foveae on the M₁ are found in 38.9% of Paleolithic H. sapiens (Martinón-Torres et al., 2012). The M-G2-419 M₁ anterior fovea is bordered distally by a well-developed mid-trigonid crest (Grade 2, Bailey, 2002a; Bailey, 2002b). In the SH Preneanderthals, the mid-trigonid crest is present in 95% of cases and is coupled with a ≥Grade 2 anterior fovea in 85% of cases (Martinón-Torres et al., 2012). In Neanderthals, 53.3% of cases possess a continuous mid-trigonid crest, with 46.4% possessing both a continuous crest and deep anterior fovea. Martinón-Torres et al. (2012) observed that this trait combination is nearly absent in Paleolithic and modern humans.

In addition to the M-G2-419 M₁, there are two isolated M₁'s from Moula-Guercy (M-D1-230 and M-L4-TNN5). All three of these molars are briefly described in Hlusko et al. (2013). These authors described these isolated molars as having deep anterior foveae and Grade 1 mid-trigonid crests. We observed that the isolated teeth have deep (Grade 3) anterior foveae. However, in M-D1-230, the mid- trigonid crest comprises only a low ball-like extension from the protoconid; the central groove separates this enamel extension from the metaconid. In M-L4-TNN5, there is only an extremely attenuated crest-like extension of the protoconid protruding into the central groove. So, unlike the situation in M-G2-419, both isolated M₁'s should be considered to have discontinuous mid-trigonid crests, at best.

A deflecting wrinkle (Scott & Irish, 2017) is not present on the M-G2-419 M₁. Bailey (2002a) found that this feature is generally absent in Neanderthals. A distal trigonid crest (Scott & Irish, 2017) is not present. Continuous distal trigonid crests of Grade 2 join the metaconid and protoconid in 30% of the M₁'s from SH Preneanderthals (Martinón-Torres et al., 2007), while in Neanderthals, they occur at low frequencies (3.6% Bailey, 2002a). The hypoconulid is moderately developed (Grade 3, Scott & Irish, 2017) in M-G2-419. Martinón-Torres et al. (2007) found large to very large hypoconulids to be present in all Middle Pleistocene—Upper Paleolithic individuals studied (Grades 4 and 5) but observed that development of this cusp is highly variable. The M-G2-419 M₁ lacks a C6, but this cusp does occur on the isolated M-D1-230 M₁. Martinón-Torres et al. (2012) recorded a complete lack of a C6 in SH Preneanderthals, while Bailey (2002a) found them to be present in Neanderthals (27.8%) and anatomically modern human (44.4%). On the M-G2-419 M₁, a weakly developed C7 cusp (Grade 1) is interposed between the metaconid and entoconid. A small C7 is also present on the M-D1-230 isolated M₁. A C7 is absent in 50% of the SH Preneanderthals, while equal percentages (15%) of Grades 1–3 C7's are present (Martinón-Torres et al., 2012). Bailey (2002a) observed that C7's are present in 18.8% of Neanderthals and 3.3–61.1% of anatomically modern to recent humans.

Hlusko et al. (2013) described the M-G2-419 M₁ as having an X-5 cusp pattern. However, the cusp pattern should be Y-5 due to the defining connection between the metaconid and entoconid (Scott & Irish, 2017). The full range of cusp patterns (Y, X, +) was observed by Martinón-Torres et al. (2012) in the SH Preneanderthals, while Neanderthals and H. sapiens tended to have a greater percentage of the Y-5 cusp pattern on the M₁. In comparison, the M-D1-230 isolated RM₁ was suggested by Hlusko et al. (2013) to have a Y-5 cusp pattern; however, given that the metaconid and hypoconid are in contact, this should be an X-5 pattern. The isolated M-L4-TNN5 LM₁ also has an X-5 cusp pattern (Hlusko et al., 2013).

The M-G2-419 M₂ has a deep anterior fovea, but it is scored as a Grade 3 due to a less well-developed mesial marginal ridge. The anterior fovea is positioned mesial to a continuous mid-trigonid crest (Grade 2). The combination of a ≥Grade 2 anterior fovea and continuous mid-trigonid crest is found in 66.6% of Neanderthals and 71.4% of SH Preneanderthals (Martinón-Torres et al., 2012). The combination of these two features is less frequent in Middle Pleistocene individuals (33.3%; Martinón-Torres et al., 2012) and H. sapiens. A distal trigonid crest is not present on the M-G2-419 M₂. Distal trigonid crests are mostly absent on the M₂'s of the SH Preneanderthals (Martinón-Torres et al., 2012). In Neanderthals, Bailey (2002a) observed that these crests are not well developed (Grades 1–2) and occur at low frequencies (8%). However, note that Martinón-Torres et al. (2012) scored the distal trigonid crests in Neanderthals as Grades 2 and 3, highlighting the difficulty in quantifying accessory trigonid crest development.

The M-G2-419 M₂ was described as having 4 cusps arranged in a cruciform (+) pattern (Hlusko et al., 2013). However, a hypoconulid is present, and we consider the tooth to have a cruciform (+) 5 cusp pattern. Bailey (2002a) and Martinón-Torres et al. (2012) found that Neanderthals have M₂ cusps arranged in a Y pattern in 78.8% and 70.8% of cases, respectively. In the other groups examined, + patterns were more common than X patterns, and Neanderthals were the only group to have a high incidence of the Y cusp pattern. Although the M-G2-419 M₂'s hypoconulid is present, it is not well developed (Grade 3). Martinón-Torres et al. (2012) scored the hypoconulid as vestigial or absent (Grades 0–1) in 33.4% of SH Preneanderthals, 11.1% of Middle Pleistocene hominins (Grade 1), and 14.2% of Neanderthals (Grade 0). Alternatively, Bailey (2002a) suggests that 100% of Neanderthal M₂'s examined possessed five cusps. An absent or reduced hypoconulid contrasts with it being expressed as a moderately large cusp of Grades ≥3 in Paleolithic H. sapiens (28%) and recent H. sapiens (15.3%; Martinón-Torres et al., 2012). The M-G2-419 M₂ lacks a C6 cusp but a C7 cusp (Grade 1) is present.

A deep anterior fovea is present on the M-G2-419 M₃ (Grades 3–4). A high frequency of Middle Pleistocene individuals (80%), SH Preneanderthals (95.2%), and Neanderthals (89%) express deep anterior foveae on the M₃ (Martinón-Torres et al., 2012). However, these authors also found that this fovea occurs in 53% of Paleolithic H. sapiens and 44.3% of recent humans. A mesial marginal ridge is not present (Grade 0) on the M₃ of M-G2-419. A mid-trigonid crest is present in M-G2-419, but it is not continuous, as it is interrupted by a short and very thin extension of the central groove. A continuous crest was observed to be present at high frequencies in SH Preneanderthals (72.7%) and Neanderthals (70.6%) but absent in the fossil and recent H. sapiens studied by Martinón-Torres et al. (2012). The co-occurrence of an anterior fovea and continuous mid-trigonid crest was observed at high frequencies on the M₃'s of SH Preneanderthals (57.1%) and Neanderthals (64.7%: Martinón-Torres et al., 2012). Alternatively, these features occur together in only one-fifth (1/5) of their Middle Pleistocene sample, and they are absent in their fossil and recent H. sapiens samples. A distal trigonid crest and C6 cusp are absent in the M-G2-419 M₃. In 35% of the Middle Pleistocene, SH Preneanderthal, and Paleolithic H. sapiens sampled, a C6 of grade 2 or 3 was present (Martinón-Torres et al., 2012). These authors found a C6 to be expressed in 50% of Neanderthal M₃'s. A C7 is present on the M-G2-419 M₃, and it is slightly more developed (Grade 2) than on the M₁ and M₂. Martinón-Torres et al. (2012) found distal trigonid crests to occur in 69.5% of the SH Preneanderthals (Grades 1-2) and to be relatively frequent in the Middle Pleistocene group. A distal trigonid crest was observed at low frequencies by these authors in Neanderthal M₃'s, consistent with its absence in M-G2-419. A C7 is variably expressed in SH Preneanderthals, Neanderthals, and Paleolithic and recent H. sapiens (Martinón-Torres et al., 2012). In M-G2-419, a post-metaconulid (Scott & Irish, 2017) appears to be present between the distal metaconid and C7 cusp. The hypoconulid in M-G2-419 is similarly developed as those on M₁ and M₂. In the SH Preneanderthals, the hypoconulid is reduced, with the cusp being absent or only slightly apparent in 34.8% of the group (Grades 0–1: Martinón-Torres et al., 2012). In Neanderthals, the latter authors found that the hypoconulid on the M₃ was absent in two cases (12.5%). In their sample, the hypoconulid was also reduced or absent in 38.9% and 52.4% of Paleolithic H. sapiens and recent H. sapiens M₃'s, respectively.

Hlusko et al. (2013) suggest that the M-G2-419 M₃ has a Y-5 cusp pattern. However, because the protoconid and entoconid are in contact, the cusp pattern should be an X-5 (Scott & Irish, 2017). In the SH Preneanderthals, Martinón-Torres et al. (2012) found that the Y cusp pattern was not well represented and that X and + patterns predominated. In Neanderthals, a Y pattern occurs in 60% of the cases examined by these authors. In their Middle Pleistocene and H. sapiens (fossil and recent) groups, Martinón-Torres et al. (2012) found that the M₃ cusp pattern was predominantly a + or an X pattern, respectively.

Pulp chamber expansion in the M-G2-419 M₁ (TI = 22.3%) is consistent with the cynodont category of Keene (1966), while the M₂ (TI = 26.0%) falls into the hypotaurodont category (Figure 10). In the Shifman and Chanannel (1978) scheme, both the M₁ and M₂ would be classified as hypotaurodont. The degree of taurodontism could not be assessed in the M₃ due to incomplete root development, but the size of the pulp chamber superior to the bifurcation of the roots suggests it would also be hypotaurodont. The M-G2-419 M₁ and M₂ have pulp chambers like that reconstructed for the Krapina 57 M₂ by Kupczik and Hublin (2010). Further, the M-G2-419 M₁ and M₂ roots are not pyramidal or conical, as observed by Kallay (1963) and Kupczik and Hublin (2010) in a large portion of the Krapina specimens. The M-G2-419 M₁-M₂ have two mesiodistally compressed roots that are separate along their entire length.

FIGURE 10

Isosurface reconstructions of the pulp cavity (red) for the M₁, M₂, and M₃ projected onto a transparent volume rendering of the M-G2-419 mandibular corpus.

3.1.4 Metric assessment of the M-G2-419 M₁-M₃

The M-G2-419 M₁ has a mesiodistal length (10.45 mm) that lies at the bottom of the Early Neanderthal range (Table 4; Figure 9). Although the Moula-Guercy M₁ is mesiodistally short, it is only slightly shorter than some other Early Neanderthals (cf. Chateauneuf 2, Krapina 102, La Chaise BD 1). The M-G2-419 mesiodistal length is most like that in Late Neanderthals. Reference to the M-D1-230 and M-I4-TNN5 M₁'s from Moula-Guercy shows them to be slightly longer mesiodistally than M-G2-419 and more like the mean condition for SH Preneanderthals, Early Neanderthals, and Late Neanderthals (Figure 9). The buccolingual breadth of the M-G2-419 M₁ (10.3 mm) is narrow and falls near the low end of the Early Neanderthal range. As a group, the Early Neanderthals tend to have slightly narrower M₁'s relative to their length than either SH Preneanderthals or Late Neanderthals. The M-D1-230 and M-I4-TNN5 M₁'s are buccolingually wider than M-G2-419 and are wide relative to their mesiodistal length when compared to other Early Neanderthals.

The mean mesiodistal length of the M₂ is stable in the groups observed, with only the Middle Pleistocene and Upper Paleolithic samples having slightly higher or lower mean lengths, respectively (Table 4, Figure 9). The M₂ in M-G2-419 has a mesiodistal length (11.1 mm) that lies near the mean length for all groups examined, excepting those from the Middle Pleistocene group. The buccolingual width of the M-G2-419 M₂ (11.6 mm) is like that observed in Middle Pleistocene Homo and Early Neanderthals but slightly narrower than in SH Preneanderthal, Late Neanderthal, and Middle-Upper Paleolithic H. sapiens. Data presented in Figure 9 shows that the M-G2-419 M₂ plots within a subgroup of Early Neanderthals with relatively short mesiodistal lengths in relation to their more typical buccolingual breadths. This subgroup of individuals derives from Krapina, La Chaise Bourgeois-Delaunay, and Tabun, and they tend to lie above the main trajectory of mesiodistal-buccolingual size increase.

The mean values for the mesiodistal length of the M₃ are similar for most of the groups studied. The mesiodistal length of the M-G2-419 M₃ (10.5 mm) lies at the bottom of the Middle Pleistocene, SH Preneanderthal, and Middle Paleolithic samples. Alternatively, the Early Neanderthal, Late Neanderthal, and Upper Paleolithic samples have ranges that incorporate shorter mesiodistal lengths of the M₃ (Table 4). These short mesiodistal lengths tend, however, to be unusual for all the groups studied. Treating these individuals as outliers shows the mesiodistal length of the M₃ in M-G2-419 to plot in the lower third of the SH Preneanderthal, Early Neanderthal, and Late Neanderthal samples. The buccolingual breadth of the M-G2-419 M₃ (11.3 mm) is like that of the Middle Pleistocene, Early Neanderthal, and Late Neanderthal groups. This breadth differs from that of the SH Preneanderthal and Upper Paleolithic groups, as they tend to be slightly narrower (Table 4). When both dimensions are considered, the M-G2-419 M₃ is seen to have a relatively short mesiodistal length paired with a relatively wide buccolingual breadth. This relationship places the M-G2-419 M₃ along the upper edge of the general trend in the data for all groups. This positioning is like that observed in the plot of the M₂, where M-G2-419 plots with a group of Early Neanderthals at the upper edge of the general trend in the data. In this case, however, there are several Late Neanderthals that plot within this cluster, unlike the situation seen for the M₂'s.

Comparing crown indices for the observed groups shows the mean index and range values for the M₁ to be stable, except for some slightly larger values in Late Neanderthals and Upper Paleolithic humans (Table 4, Figure 9). Alternatively, when the M₂ is considered, the mean value for the Middle Pleistocene and SH Preneanderthal groups is lower than that for the Early Neanderthals and Middle and Upper Paleolithic H. sapiens. This indicates that the earliest groups have long crown lengths relative to their breadths. This changes in Early Neanderthal and Middle Paleolithic H. sapiens groups, as crown length to breadth dimensions becomes more equal. Index values for the Late Neanderthals show them to have slightly increased crown lengths relative to their breadths but not to the extent seen in the two earliest groups. The M-G2-419 M₁-M₃ and the isolated M₁'s from Moula-Guercy have high index values that are consistent with values seen in the Early Neanderthals and Middle Paleolithic H. sapiens.

3.1.5 Developmental age and sex

In M-G2-419, the premolar sockets are fully formed, and the RM_1-2 are fully erupted and have completed root formation. The RM₃ is mainly contained within its crypt. However, the distal crown protrudes above the alveolar ridge, suggesting that the crown was erupting through the gingiva. The M₃ roots are developed such that there is ≈1.0 mm of root formation inferior to the bifurcation. Based on dental calcification timings for recent humans (Schour & Massler, 1941), a developmental age of 15–16.0 ± 0.5 years is indicated. Thompson and Nelson (2005) assessed an age of 15.5 ± 1.25 years for the Le Moustier 1 adolescent. This age is based on lower third molars with ≈25% of root growth completed (root length < crown height). This degree of root development equates to ≈14–14.5 years of age in the Moorrees et al. (1963) system. Because the M-G2-419 RM₃ has completed a greater amount of root development, using the same criteria results in an age estimate closer to 16.0 years.

Given the age assigned to the M-G2-419 individual, the masseter, medial pterygoid, temporalis, mylohyoid, and platysma muscle attachments are only weakly demarcated. The lingual and buccal cortices are thin. In general, the corpus is very gracile. When compared to the similarly aged Le Moustier 1 male individual, the M-G2-419 mandible is substantially smaller and more lightly built. This mandible is likely associated with the Guercy 1 cranial vault, which we have suggested belongs to an adolescent female (Richards et al., 2021).

3.2 Isolated mandibular fragment M-F4-77

The M-F4-77 mandibular fragment is the right inferior border of an adult mandibular corpus (Figure 3a). The maximum length is 46.6 mm (a-p). The lingual and buccal cortices are 2.5 and 3.0 mm thick, respectively. The posterior edge of the digastric fossa is not preserved. The digastric fossa of La Quina H5 (Henri-Martin, 1926; Verna, 2006), Regourdou 1 (Piveteau, 1963), and Amud 1 (Condemi, 2001; Rosas, 2001; Rosas & Bermúdez de Castro, 1999) extends posteriorly to approximately the premolar-molar junction, while that in Vindija 206 and 231 only extends to below the P₄ region. Given this relationship and the thickness of the fragment, it can confidently be placed along the inferior border below the molar region. A small portion of the inferior mandibular torus is apparent on the inferolateral surface, as are multiple fine attachment ridges associated with the platysma muscle. The basal corpus is buccolingually wide anteriorly (7.8 mm) but thins gradually to 4.8 mm as the gonial region is approximated. An inferior border that regularly decreases in width and becomes very thin in the gonial region is considered a derived character of European Neanderthals (Piveteau, 1957; Rosas, 1992, 2001). However, as in M-G2-419, the presence of gonial thinning cannot be confirmed in M-F4-77. This mandibular fragment has tentatively been assigned to Guercy 4, a robust adult male (Richards et al., 2021).

3.3 Isolated mandibular fragment M-S-TNN1

The M-S-TNN1 mandibular fragment comprises a small portion of the right mandibular corpus with a fully erupted adult R_C (Figure 3b). The M-G4-144 L_C is likely the antimere of this canine based on crown and root morphology and linear enamel defects (Hlusko et al., 2013). The basal canine alveolus (≈50% preserved), the mesial wall of the RP₃ socket, and a small portion of the lingual cortical plate are preserved on the M-S-TNN1 mandibular fragment. The gentle inferomedial slope of the lingual cortex is like that in La Quina H5 (Rosas, 2001; Verna, 2006; Wolpoff et al., 1981), Zafarraya 1 (Condemi et al., 2013), and Kebara 2 (Rosas, 2001). The slope of the lingual plate differs from both the stronger inferomedial slope in Regourdou 1 and the more vertical slope of the lingual plate in Amud 1 (Condemi, 2001; Rosas, 2001; Rosas & Bermúdez de Castro, 1999). It also differs from the more vertical orientation of the lingual plate found in most recent humans. This mandibular fragment and tooth have been tentatively assigned to Guercy 4. Both the M-S-TNN1 R_C and M-G4-144 L_C have fully formed roots. The R_C has very slight apical wear, while the L_C is unworn (Hlusko et al., 2013).

4 DISCUSSION

4.1 Introduction

Mandibular morphology has played a substantial, but likely outsized, role in discussions of the number of species present in Europe during the Middle Pleistocene, connections between paleodemes comprising the Neanderthal lineage, and the timing of anatomically modern human incursions into Western Europe (Abbazzi et al., 2000; Arsuaga et al., 1997, 2014; Bermúdez de Castro et al., 2017; Cattani et al., 1991; de Lumley, 1973, 2015; Keeling et al., 2023; Lebel et al., 2001; Lebel & Trinkaus, 2002; Mounier, 2011; Mounier et al., 2009; Quam et al., 2023; Roksandic et al., 2011, 2018; Rosas, 1995, 2001; Rosas et al., 2019; Rosas & Bermúdez de Castro, 1998, 1999; Skinner et al., 2016; Stefan & Trinkaus, 1998; Trinkaus, 2007; Vallois, 1956; Verna et al., 2020; Vialet et al., 2018; Wolpoff & Frayer, 2005). Discussion hinges both on the presence or absence of individual features and on unique combinations of these features, which have been suggested to be derived or to occur at high frequencies in Neanderthals. In relation to the M-G2-419 mandibular corpus, these features include the retromolar space, the number of mental foramina, the position of the mental foramen or foramina on the corpus relative to both the overlying dentition and corpus height, the position and development of the anterior marginal tubercle, the position of the lateral prominence, the configuration of the masseteric fossa, the inclination of the mylohyoid line and its vertical position inferior to the alveolar ridge at the M_2-3, submandibular fossa depth, and the relief of the pterygoid fossa (Hublin, 1998; Quam et al., 2023; Rak, 1998; Rosas, 1995, 1998, 2001; Rosas et al., 1991; Rosas & Bermúdez de Castro, 1998, 1999; Stringer et al., 1984; Trinkaus, 1988).

Here, we have undertaken a metric and nonmetric assessment of these features in the Moula-Guercy mandibular remains to determine their affinities to subdivisions within European and Middle Eastern Middle Pleistocene—Upper Paleolithic Homo. Because the comparative description is ontogenetically based, it allows us to go beyond static comparisons to similarly-aged individuals and examine these features throughout the growth cycle.

4.2 Morphological overview

The alveolar ridge and the inferior border of the mandibular corpus in M-G2-419 are parallel, a condition observed consistently in Early—Late Pleistocene mandibles (Vialet et al., 2018). The M-G2-419 masseteric fossa is lightly marked and consistent with the condition in Early Neanderthals and Late Neanderthals. This contrasts with the suggested primitive condition of a deep and well-marked fossa as seen in the SH Preneanderthal paleodeme (Rosas, 2001). A retromolar space (gap) is not present and the extramolar sulcus is poorly developed in M-G2-419. This may be due to the adolescent age of the Moula-Guercy specimen. However, the degree of extramolar sulcus development in M-G2-419 is like that found in Early Neanderthal and Late Neanderthal children, not adolescents. The M-G2-419 lateral prominence is poorly developed and positioned as in the developmentally younger Krapina 53. The Late Neanderthal subadults Teshik Tash 1 and Le Moustier 1 lack significant swelling of the lateral prominence, but in both it is more developed than in M-G2-419.

Quam et al. (2023) observed that the presence of a single mental foramen appears to be primitive for Homo. They suggest that the development of multiple mental foramina, while present in the Middle Pleistocene Homo sample, occurs at a higher frequency in Neanderthals and may represent a derived condition.

The M-G2-419 mandibular corpus has a single ovate mental foramen. In the SH Preneanderthals, the earliest generally accepted members of the Neanderthal clade, a single mental foramen is ubiquitous, as only the adolescent AT-2 has multiple foramina. The Early Neanderthals differ from the SH Preneanderthals by having a higher, but still low, frequency of multiple mental foramina. Two foramina occur in slightly more than a third of subadults, with only Montgaudier 1 having more than two foramina. In adults, the frequency of multiple foramina drops to 25% (2/8). While the occurrence of multiple foramina increases slightly in Early Neanderthals, relative to SH Preneanderthals, the SH sample lacks individuals younger than 12.5 years. Given that foramina appear to coalesce with age, this may account for this difference. In Late Neanderthal subadults, the frequency of multiple mental foramina essentially doubles, with nearly a quarter of them having more than two foramina. Adult Late Neanderthals have a near equal frequency of multiple foramina as observed in the subadult sample, but only Regourdou 1 has more than two foramina.

In Late Neanderthals, the degree of decrease in the frequency of multiple mental foramina as individual age increases is less than that seen in Early Neanderthals. On this basis, the Late Neanderthals clearly differentiate themselves from the SH Preneanderthal and Early Neanderthal groups, inclusive of M-G2-419. The Late Neanderthals show evidence of a developmental shift that results in the formation of multiple foramina in early age stages and an increase in the number of foramina. Multiple foramina are produced when the point at which the mental nerve exits the foramen moves distally along the nerve trunk to expose the multiple terminal fibers. Foramina appear to coalesce in later age stages, such that adults possess a similar frequency of multiple mental foramina as subadults, but the number of foramina per individual is reduced. This developmental shift is a distinguishing characteristic of MIS 5d—MIS 3 Late Neanderthals.

Mental foramen position relative to the dentition varies greatly within the Neanderthal lineage, but differences in frequency of expression are apparent between chronological subgroups. However, this feature appears to be impacted by mandibular size (Rosas, 1995) and dental arcade length (Robinson & Williams, 2010; Williams & Krovitz, 2004).

The mental foramen of M-G2-419 is inferior to the P₄-M₁, making it align most naturally with the SH Preneanderthals. In the Middle Pleistocene Homo sample, only Arago II has the mental foramen positioned inferior to the M₁; the remaining five mandibles have the foramen more anteriorly positioned. All the SH Preneanderthal adolescents have the foramen positioned inferior to the P₄-M₁, as in M-G2-419. Most of the adult SH Neanderthals have the mental foramen positioned inferior to P₄-M₁ or M₁, with a few having a more posterior placement. Given these data, it appears that a slight developmental change has occurred, relative to the Middle Pleistocene Homo sample, where in adult SH Preneanderthals, the inferior alveolar neurovascular bundle now exits the corpus slightly more posteriorly relative to the dentition. In adult and subadult Early Neanderthals, the mental foramen is most often positioned inferior to the M₁. Only Krapina 59 has the foramen positioned posterior to the center of M₁ (Wolpoff, 1980); however, given the large size of this mandible (Smith, 1976), and the Krapina dentition in general (Smith, 1978), the foramen's posterior placement may be due solely to its large size (Robinson & Williams, 2010; Rosas, 1995). In Early Neanderthal subadults, positioning inferior to the P₄-M₁ or more anterior teeth also occurs. Although sample sizes are small, it is of interest that the Krapina subadults have more posteriorly positioned foramina relative to the French Early Neanderthal subadults. Developmental age does not appear to account for this difference. In adult Late Neanderthals, the mental foramen is most often positioned along a transect from the mesial M₁—M₁-M₂ interdental septum. However, approximately one-third of the Late Neanderthal adults have more anteriorly placed foramina (P₄, P₄-M₁). Among the Late Neanderthal subadults, only two have the mental foramen positioned inferior to the M₁, with the more usual location being inferior to the premolars, especially the first premolar. This distribution of foramen positions is partially driven by the developmental age profile of Late Neanderthals, as it contains more young individuals.

In sum, the SH Preneanderthal, Early Neanderthal, and Late Neanderthal adult samples show that the mental foramen is more often positioned slightly posteriorly, relative to the Middle Pleistocene Homo sample. Comparing the SH Preneanderthals to the Early Neanderthals is complicated by the lack of young individuals in the SH sample and the apparent coalescence of foramina with age. However, foramen position in adult Early Neanderthals is like that in the SH Preneanderthals. Alternatively, while the Late Neanderthals share the presence of posteriorly positioned foramina with these two groups, they possess a higher frequency of more anteriorly positioned foramina. These positioning data are consistent with the observed higher frequency of multiple mental foramina throughout the developmental sequence in Late Neanderthals. Condemi (2001) suggested that the high degree of variation in foramen placement in the closely related members of the Neanderthal lineage makes it difficult to assign any major taxonomic significance to this feature. However, Robinson and Williams (2010) concluded that mental foramen position may have taxonomic value in hominoids, particularly at the subspecific level. Given differences observed between the SH Preneanderthals—Early Neanderthals, relative to Late Neanderthals, we suggest that further work may shed more light on subspecific or paleodeme variation in these groups.

Although the significance is unclear, the SH Preneanderthals all appear to have an inferior alveolar neurovascular bundle that exits the mental foramen posteriorly (Rosas, 2001). In M-G2-419, an anterior exit of at least a portion of the neurovascular bundle is clearly marked. The direction of the nerve's exit is variable in other Early Neanderthals and Late Neanderthals, with both conditions represented. Given recent work on the inferior alveolar nerve canal (Yoakum & Terhune, 2023), further work needs to be done to properly assess foramen morphology relative to nerve positioning.

The anterior marginal tubercle of M-G2-419 is positioned inferior to P₄, which is more anterior than in most adult SH Preneanderthals, Early Neanderthals, and Late Neanderthals. It is also more anterior than in most subadult SH Preneanderthals, all of which are adolescents, but more consistent with the positions commonly seen in subadult Neanderthals, many of which are younger. Although M-G2-419 is lightly built, its tubercle is more developed than in the adult SH female AT-172 but is comparable to those in the similarly-aged Krapina 53 and 54. Mandibles assigned to the European Middle Pleistocene Homo sample all possess anterior marginal tubercles, which are usually positioned inferior to the P₄ or P₄-M₁, with only one inferior to the M₁ (Roksandic et al., 2011, 2018; Rosas, 2001; Rosas & Bermúdez de Castro, 1998). Some SH Preneanderthal adolescents do not possess an anterior marginal tubercle, but when present it occurs most frequently inferior to the P₄-M₁. In adults, the tubercle is more posteriorly positioned, with nearly three-quarters of them being inferior to the M₁ (Rosas, 2001). In the SH Preneanderthals, Rosas (2001) has shown that the anteroposterior positioning of the tubercle is impacted by size increase and is correlated with the position of the mental foramen. The position of the anterior marginal tubercle is, then, consistent with the slightly more posterior location of the mental foramen in adults from SH. In Early Neanderthal subadults and adults, the anterior marginal tubercle is most often positioned inferior to the M₁, with two subadults having the tubercle inferior to the premolars. In this case, Early Neanderthals are like SH Preneanderthals in having the tubercle inferior to the M₁. Although the most frequent position of the anterior marginal tubercle is inferior to the M₁ in Early Neanderthal subadults (2/4), both specimens with this position are from Krapina, making these data consistent with those for mental foramen position. Tubercle position in the adult Late Neanderthals is most often inferior to the M₁ (46%). However, the remaining 54% of individuals have the anterior marginal tubercle positioned inferior to a range of more anterior teeth, with P₃-P₄ and P₄ comprising most of the alternate positions. The anterior marginal tubercle in Late Neanderthal subadults is not usually positioned inferior to the M₁, as this position is only known for the poorly developed tubercle of Cova del Gegant 1 (Condemi, 2001). In Late Neanderthals, the more usual position of the subadult anterior marginal tubercle is inferior to the premolars.

Given current data, positioning of the anterior marginal tubercle appears to have shifted posteriorly to a slight degree in the SH Preneanderthals; alternate positions are few. This condition appears to be retained in Early Neanderthals. While the positioning of an anterior marginal tubercle inferior to the M₁ is the most frequent in Late Neanderthals, they show higher percentages of individuals with tubercles in more anterior positions. This observation is consistent with changes in the lower face of Late Neanderthals relative to the SH Preneanderthal and Early Neanderthal conditions (Arsuaga et al., 1997, 2014; Bastir et al., 2007; Bermúdez de Castro et al., 2017; Franciscus, 1995, 2003; Freidline et al., 2013; Heim, 1976, 1978; Holton & Franciscus, 1995; Ponce de León & Zollikofer, 2001; Rak, 1986; Richards et al., 2023; Schwartz et al., 2008; Smith & Paquette, 1989; Smith, 1983; Tattersall, 2006; Trinkaus, 1984; Trinkaus, 1987; Zollikofer et al., 2008). However, the lack of any infants or children in the SH Preneanderthal sample and the small number of adolescents in the Late Neanderthal sample make it difficult to compare tubercle ontogeny across these groups and clarify the significance of a posterior shift in tubercle position. Of further interest is that, while some SH Preneanderthals do not possess an anterior marginal tubercle (Rosas, 2001), comparable data are unavailable for the Early and Late Neanderthals.

There is evidence of a repositioning of the origin of the mylohyoid muscle in Neanderthals. The course of the mylohyoid line in most SH Preneanderthals is said to be intermediate relative to the pattern in Middle Pleistocene Homo and Late Neanderthals (Rosas, 2001). Based on data in Rosas (2001), the mylohyoid line in the majority of Early Neanderthals is intermediate between SH Preneanderthals and Late Neanderthals. The M-G2-419 mylohyoid line would also be categorized as intermediate. In Late Neanderthal adults, the characteristic pattern has the mylohyoid line running at a steeper angle across the corpus and to have its posterior end positioned at or near the alveolar ridge at M₃ (Rosas, 2001). However, in this group, nearly 40% of the individuals have mylohyoid lines that have been categorized as either low or medium in height at the M₃ (Rosas, 2001). Variation in subadult Late Neanderthals also reflects this differential positioning of the mylohyoid muscle. The position of the mylohyoid line reflects variation in several soft tissue structures, which are impacted by overall facial size and orientation. Given available evidence, there is variation in this feature that requires further study to discern its evolutionary significance.

The inferior mandibular torus in M-G2-419 is weakly expressed and separated from the lateral portion of the inferior corpus by a slight sulcus, as in Krapina specimens and Neanderthals in general (Smith, 1976; Suzuki, 1970). Further, in M-G2-419 the mylohyoid line terminates superior to the torus, disappearing into the corpus inferior to the P₄. The mylohyoid line and inferior mandibular torus are separated in SH Preneanderthals, except for a single individual. In both subadult and adult Krapina mandibles, the mylohyoid line transitions into the inferior marginal torus anteriorly. Although this transition usually occurs inferior to the M₁, it also occurs inferior to the P₄ (Smith, 1976), as in M-G2-419. The subadult Late Neanderthals Devil's Tower 1 and Le Moustier 1 have the torus and mylohyoid line separated anteriorly, as in M-G2-419. Alternatively, Suzuki (1970) observed that the mylohyoid line transitions into the inferior mandibular torus anteriorly in Late Neanderthals. New investigations are needed to clarify whether the mylohyoid line descends during growth to the level of the inferior mandibular torus.

Deeply excavated submandibular fossae are thought to be a derived character in Neanderthals, in general (Rosas, 2001). In M-G2-419, the sublingual and submandibular fossae are shallow, and the mylohyoid line minimally overhangs the submandibular fossa. Krapina subadults generally have a large offset between the region superior to the mylohyoid line and the region inferior to the line. However, both Krapina 56 and 58 have less strong offsets (Smith, 1976). A stronger offset between these two regions might be associated with male individuals. However, there is almost no offset in the robust Payre 15 (Verna et al., 2020) and Regourdou 1 (Piveteau, 1963; Trinkaus, 2016) mandibles, both of which are probably males. Further, while a significant degree of sexual dimorphism is present in the Krapina sample, Smith (1976) did not find lingual corpus morphology to be useful in delineating an individual's sex.

The M-G2-419 mandible has a shallow and smooth region anterior to the medial pterygoid attachment site. This lack of topographic detail is consistent with the overall poorly marked nature of the internal corpus and the positioning of the medial pterygoid insertion. Rosas (2001) suggests that an excavated pterygoid fossa may be a derived feature in European Neanderthals. However, the topography of the pterygoid fossa in M-G2-419 is more like that described for the SH Preneanderthals (Rosas, 2001) than like that described for either the Early Neanderthals or Late Neanderthals. The developmentally younger Late Neanderthal subadults Roc de Marsal 1 and Devil's Tower 1 both show a greater extent and depth of the pterygoid fossa than M-G2-419. In contrast, the pterygoid fossa in the adult Early Neanderthal Ehringsdorf F and Late Neanderthals Malarnaud 1 (subadult) and La Quina H9 (adult; Rosas, 2001) are shallow and not well-marked by anatomical detail. There may be a relationship between an anteroinferiorly deep pterygoid fossa and a pregonial notch, which are both present in Roc de Marsal 1, Devil's Tower 1, and Regourdou 1, but absent in M-G2-419.

4.3 Mandibular metric overview

Insights into the configuration of the lower face can be derived from measures of the mandibular corpus at various locations along the tooth row. However, many assessments of mandibular size are limited to only one or two points along the corpus. Here we provide an overview of dimensions taken at the mental foramen and at M₂, while our assessments at the P₄-M₁ and M₁ are provided in the Supplemental Data. The corpus is shorter and narrower at the mental foramen in M-G2-419 than in SH Preneanderthal adults and adolescents, Early Neanderthal adults, and most adult Late Neanderthals. However, the index of robusticity at the mental foramen in M-G2-419 is like that for adults in these three groups. The SH Preneanderthal adolescents and Early Neanderthal subadults have a higher index value relative to M-G2-419. When compared to Early Neanderthal subadults, the M-G2-419 height at the mental foramen is like Krapina subadults, but the corpus is thinner at this position. This relationship can also be seen relative to the Late Neanderthal subadults, where M-G2-419 corresponds more closely to 7- to 8.0-year-old children.

Compared to adult values, the index of robusticity at the mental foramen is consistently high in subadult Early and Late Neanderthals. Both groups have higher index values than M-G2-419. When this index is plotted relative to developmental age, M-G2-419 is positioned slightly below the SH Preneanderthal and Early Neanderthal adolescents (Figure 6). However, there is only a single Late Neanderthal in the age range of M-G2-419 (=Malarnaud 1). Given the overall similarity of the Early Neanderthal and Late Neanderthal adult means and ranges of variation, it is likely that further data from Late Neanderthal adolescents would show them to have indices like those in the SH Preneanderthal and Early Neanderthal samples. However, given current samples, the Late Neanderthal subadults show generally higher index values than the Early Neanderthal subadults, including during adolescence. Sample size for the early group may be driving this difference. Alternatively, it may indicate a different growth pattern, or patterns, in the Late Neanderthals, as suggested by changes in the number and position of the mental foramina and position of the anterior marginal tubercle.

The positioning of the mental foramen on the lower half of the mandibular corpus has been argued to be a derived condition within the genus Homo (Daura et al., 2005; Quam et al., 2023). Quam et al. (2023) found no significant differences between the SH Preneanderthals or their combined Early—Late Neanderthal sample for the height of the mental foramen on the corpus. All foramina were positioned on the lower portion of the corpus. The height of the mental foramen on the corpus in M-G2-419 falls in the middle of the range of variation for the Neanderthal sample reported by these authors.

When corpus height at M₂ is considered, SH Preneanderthal, Early Neanderthal, and Late Neanderthal adults are similar, and the value for M-G2-419 falls below the range of variation for these groups. As for the thickness at the M₂, M-G2-419 falls within the lower end of the Late Neanderthal range. The index of robusticity at the M₂ in M-G2-419 is like that expressed in adult SH Preneanderthals, but near the upper-end of the ranges for both Early Neanderthals and Late Neanderthals. When the subadult record is considered, there is only a single SH Preneanderthal adolescent with an index value, and it is within the adult range. Compared to the two Early Neanderthal subadults with these data, both of which are adolescents, the height at the M₂ is comparable, but the corpus thickness is much narrower in M-G2-419, resulting in a lower index of robusticity. When M-G2-419 is compared to the two available Late Neanderthal subadults with similar age assessments, its height is greater while its width is similar, again resulting in a lower index value.

When M-G2-419 is considered relative to the combined sample of adult Middle—Upper Paleolithic H. sapiens, its height at the mental foramen falls at the lower end of the range of variation, while its thickness is near the mean, leading to an index of robusticity that falls at the top of the range for this group. The early modern human adult sample has generally lower index values than the SH Preneanderthal—Late Neanderthal samples. Corpus height at M₂ in M-G2-419 is substantially shorter than in early modern humans, while the thickness is within the range, once again resulting in an index value at the top of the early modern human range. Considering the small sample of subadult Middle—Upper Paleolithic H. sapiens, the height, thickness, and index of robusticity at the mental foramen in M-G2-419 are like La Quina-Aval 4 (Verna et al., 2012). However, the latter individual is significantly younger than the Moula-Guercy adolescent. When the height and thickness at M₁ are considered, the subadult Middle—Upper Paleolithic sample contains values like those for M-G2-419. However, only some of the ≈10.0- to 12.0-year-olds in this group, represented by Predmostí 25 (Matiegka, 1934), Kostenki 3, and Sungir 3 (Minugh-Purvis, 1988), have similar robusticity indices.

When M-G2-419 is compared to the recent adult H. sapiens sample, the height, thickness, and index of robusticity at the mental foramen, P₄-M₁, and M₁ are similar. Adult values for the index of robusticity are lower than the subadult values, as found in all the groups compared. However, in recent H. sapiens, both adults and subadults have high values for the index of robusticity at the M₂, compared to M-G2-419 and the rest of the comparative sample.

4.4 M-G2-419 M₁-M₃ morphological and metric overview

Cusp patterns on the M-G2-419 lower molars vary (M₁ = Y-5, M₂ = + − 5, M₃ = X-5) but are consistent with those found in all the groups studied. However, our assessment of these patterns differs from that in Hlusko et al. (2013). The most significant difference is that they suggested that the M-G2-419 M₂ lacks a hypoconulid and is, therefore, four cusped. We have identified a reduced hypoconulid on the M₂, making it 5-cusped. This observation is consistent with Bailey's (2002a) finding that 100% of the Neanderthal M₂'s she examined had five cusps. Further, Martinón-Torres et al. (2012) only found the hypoconulid to be absent (Grade 0) in two of the Neanderthal teeth in their sample. Bailey (2002a) also observed that 4-cusped M₂'s are a low frequency trait associated with Upper Paleolithic and recent H. sapiens.

The M-G2-419 M₁-M₂ both have strong mesial marginal crests, while the M₃ lacks this crest. Bailey (2002a) notes that it is not uncommon for the mesial marginal crest to be absent in Neanderthals. All three molars have deep anterior foveae. In the case of the M₁-M₂, these foveae are bounded by continuous mid-trigonid crests. The M₃ differs, as the mid-trigonid crest is very slightly divided by the central groove. The presence of deep anterior foveae and mid-trigonid crests is consistently found at high frequencies on the M₁-M₂ of European Middle Pleistocene Homo, SH Preneanderthals, Early Neanderthals, and Late Neanderthals, although frequencies vary by tooth (Bailey, 2002a; Martinón-Torres et al., 2012). A similar situation has been observed for the M₃, although they are less frequent in the Middle Pleistocene Homo group. This trait combination occurs at low frequencies in fossil and recent H. sapiens.

Because the M-G2-419 M₁ is mesiodistally short and buccolingually narrow, it lies at the lowest end of the range of size variation for Early Neanderthals but near the middle of the range for Late Neanderthals (Figure 9). The two isolated M₁'s (M-D1-230, M-I4-TNN5) from Moula-Guercy also have relatively short mesiodistal lengths but have wider buccolingual breadths. For teeth with similar mesiodistal lengths, they are unusually wide.

The mesiodistal length of the M-G2-419 M₂ is near the mean for SH Preneanderthals, Early Neanderthals, and Late Neanderthals. Alternatively, when the buccolingual breadth is considered, the M-G2-419 M₂ is seen to lie along the outer extent of the SH Preneanderthal and Late Neanderthal samples. When these dimensions are plotted (Figure 9), a subset of Early Neanderthals with wide buccolingual breadths relative to their mesiodistal lengths occupy the space above the SH Preneanderthal and Late Neanderthal clusters. The Moula-Guercy M₂ plots slightly below this group.

When the M-G2-419 M₃ is considered, a similar situation occurs. The M₃ is again mesiodistally short but relatively wide buccolingually. This proportion again results in the M-G2-419 M₃ lying at the base of a cluster of specimens with high breadth values, both relatively and absolutely, in the Early Neanderthal range (Figure 9). However, unlike the situation with the M₂, this group also contains some Late Neanderthals.

Based on M₁ crown index values for M-G2-419, M-D1-230, and M-I4-TNN5, the teeth are generally wide relative to their lengths, resulting in their rounded square shape. Alternatively, the Middle Pleistocene, SH Preneanderthal, and, to a lesser extent, Late Neanderthal M₁'s are narrower relative to their lengths. This results in the SH Preneanderthals having an ovate—rounded rectangular shape (Martinón-Torres et al., 2012). The Late Neanderthals would have crown shapes intermediate to these two extremes. Further, while the Early Neanderthals have M₁ crown indexes that follow the general trend for all the groups, this is not the case for the M₂ and M₃. Index values for the Early Neanderthal M₂'s show that a subgroup of individuals deviates from the general trend of size increase by being mesiodistally short and slightly wide buccolingually. This is also true for the M₃'s but to a lesser extent.

The M-G2-419 M₁-M₂ have a moderate degree of pulp chamber expansion, and it occurs superior to the root bifurcations (supraradicular of Kallay, 1963: Figure 10). The pulp chambers are classified as cynodont (M₁) and hypotaurodont (M₂) in the classification scheme of Keene (1966), while both are hypotaurodont in the scheme of Shifman and Chanannel (1978). The M-G2-419 M₁—M₂ have two mesiodistally compressed roots that are separate along their length. “Prismatic” (radicular) taurodontism is not present. This observation is consistent with that for the SH Preneanderthals, as >50% of the sample is hypotaurodont, with only a single case of hypertaurodontism observed (Martinón-Torres et al., 2012). Within Neanderthals, taurodontism is variably present and variably pronounced (Patte, 1959; Trinkaus, 1984). It is principally the Krapina Early Neanderthals that have a high degree of pronounced taurodontism (Kallay, 1963).

In summary, the Moula-Guercy mandibular remains provide new data regarding MIS 5e Neanderthals and Early Neanderthals generally. The M-G2-419 mandible is morphologically and metrically most closely linked with Preneanderthals from the Sima de los Huesos and members of the Early Neanderthal group. Whereas M-G2-419 and other Early Neanderthals are like SH Preneanderthals in the number and position of mental foramina and position of the anterior marginal tubercle, the Late Neanderthals differentiate themselves from these groups. Among subadult and adult Late Neanderthals, a greater number of mental foramina are observed, as are more individuals with more anterior placements of the mental foramen and anterior marginal tubercle. This indicates that the developmental shift impacting this facial region, first seen in SH Preneanderthals, continues into Early Neanderthals. In the Late Neanderthals, another developmental shift impacts this facial region. The presence of a second shift in development is consistent with the differentially developed face of the Late Neanderthals, relative to the earlier groups, as earlier suggested by Sergi (1962).

In keeping with its subadult age, corpus dimensions of M-G2-419 are small relative to adults, but its index of robusticity is more like adults than subadults. A similar situation occurs in adolescent SH Preneanderthals. In M-G2-419, this is mainly due to a thin corpus that is more like those of Neanderthal children than those of adolescents. The crowns of the M-G2-419 M₁-M₃ have a feature combination that is most frequently found in SH Preneanderthals, Early Neanderthals, and Late Neanderthals, but which is generally either at low frequencies or absent in Paleolithic and recent H. sapiens.

While the morphology of the M-G2-419 mandible and dentition provides insights into the evolution of the Neanderthal face, interpretations are complicated by several factors. The most significant of these factors is that individuals are concentrated in different age ranges in the SH Preneanderthal, Early Neanderthal, and Late Neanderthal groups. Tillier (2011) discussed the impact of this issue on discerning ontogenetic changes in Neanderthals. Difficulties associated with the available samples are further accentuated by the fact that 100% of the variation in our Preneanderthal group and 50% in our Early Neanderthal group represents that within single paleodemes. Late Neanderthals are more variable, likely due to the representation of multiple paleodemes in the record. Whereas further material will solve some of these issues, a reanalysis of mandibular morphology, an expansion of metric data collected on existing specimens, and new and innovative methodologies (Arnaud et al., 2015) would enhance the comparability between existing samples.

AUTHOR CONTRIBUTIONS

Gary D. Richards: Conceptualization; investigation; funding acquisition; writing – original draft; methodology; validation; visualization; writing – review and editing. Rebecca S. Jabbour: Conceptualization; investigation; writing – review and editing; visualization; validation; methodology. Gaspard Guipert: Conceptualization; investigation; writing – review and editing; visualization; validation; methodology. Alban Defleur: Conceptualization; investigation; methodology; validation; visualization; writing – review and editing; project administration.

ACKNOWLEDGMENTS

Fanny Derym, Musée Archéologique de Soyons, provided access to collections under her care. The authors wish to thank Dr. Habib Al Dahdah and his team at the Hôpital Privé Pasteur Radiology Unit, Guilherand Granges and the staff at EQUIPEX NanoID, CEREGE, OSU Pytheas, UMR 7330 CNRS-AMU for providing access to and support with CT scanning. The authors are very grateful for access to CT data that have been made available for download via NESPOS (Stiftung Neanderthal Museum, Germany) and the Croatian National History Museum. Requests for access to CT-derived models of the Moula-Guercy mandibular sample should be forwarded to the Moula-Guercy site director, Dr. Alban Defleur.

FUNDING INFORMATION

This work was funded by, Arthur A. Dugoni, School of Dentistry, University of the Pacific, San Francisco, CA; Grant number: 03-Activity-059.

CONFLICT OF INTEREST STATEMENT

The authors declare no potential conflicts of interest.

Open Research

DATA AVAILABILITY STATEMENT

Data collected from the Moula-Guercy mandible and virtual reconstruction, from our modern human skeletal series, and from the literature are provided within the main text and tables and Supplemental Data. The acknowledgments section contains information on access to CT data employed in this work.