PALEVO 1631-0683 Elsevier S1631-0683(13)00171-1 10.1016/j.crpv.2013.10.008 Research article Human palaeontology and prehistory (Palaeoanthropology) Human palaeontology and prehistory (Palaeonthropology) Trigonid crests expression in Atapuerca-Sima de los Huesos lower molars: Internal and external morphological expression and evolutionary inferences Expression des crêtes du trigonide de molaires inférieures à Atapuerca-Sima de los Huesos : expression morphologique interne et externe et inférences évolutionnistes Martínez de Pinillos Marina marinampg@hotmail.com marina.martinezdepinillos@cenieh.es a Martinón-Torres María a Skinner Matthew M. b c Arsuaga Juan Luis d Gracia-Téllez Ana d e Martínez Ignacio d e Martín-Francés Laura a Bermúdez de Castro José María a Hominid Dental Research Group, National Research Center on Human Evolution (CENIEH), Paseo Sierra de Atapuerca s/n, 09002 Burgos, Spain Hominid Dental Research Group, National Research Center on Human Evolution (CENIEH), Paseo Sierra de Atapuerca s/n Burgos 09002 Spain Department of Anthropology, University College London, 14, Taviton Street, London, WC1H 0BW, United Kingdom Department of Anthropology, University College London 14, Taviton Street London WC1H 0BW United Kingdom Department of Human Evolution, Max Planck Institute for Evolutionary Anthropology, Deutscher Platz 6, 04103, Leipzig, Germany Department of Human Evolution, Max Planck Institute for Evolutionary Anthropology Deutscher Platz 6 Leipzig 04103 Germany Centro Mixto UCM-ISCIII de Evolución y Comportamiento Humanos, Avd. Monforte de Lemos 5, Pabellón 14, 28029 Madrid, Spain Centro Mixto UCM-ISCIII de Evolución y Comportamiento Humanos Avd. Monforte de Lemos 5, Pabellón 14 Madrid 28029 Spain Área de Paleontología, Departamento de Geología, Universidad de Alcalá de Henares, 28871 Alcalá de Henares, Spain Área de Paleontología, Departamento de Geología, Universidad de Alcalá de Henares Alcalá de Henares 28871 Spain 13 3 S1631-0683(14)X0003-5 205 221 © 2013 Académie des sciences. Published by Elsevier B.V. All rights reserved. 2013 Académie des sciences Full (PDF)

Trigonid crest patterning in lower molars is distinctive among Late Pleistocene hominins such as Homo neanderthalensis, fossil Homo sapiens and modern humans. In this paper, we present an examination of trigonid crest patterning in the Middle Pleistocene permanent lower molar sample (n = 62) of Homo heidelbergensis from Sima de los Huesos (SH). Crest expression was assessed from 3D models of the enamel and the dentine surfaces that were produced using micro-computed tomography (microCT). The aims of our analysis are to: 1) characterize the pattern of trigonid crest expression at the outer enamel and enamel-dentine junction surfaces (OES and EDJ) of the SH sample, 2) evaluate the concordance of expression between both surfaces, and 3) place trigonid crest variation in the SH sample into a phylogenetic context. Our results reveal a greater variability in the expression of trigonid crests at the EDJ (14 types) compared to the OES (4 types). Despite this variability, in almost all cases the expression of a continuous mid-trigonid or distal crest at the OES corresponds with the expression of a continuous mesial/mid-trigonid or distal trigonid crest, respectively, at the EDJ. Thus, it is possible to predict the type of trigonid crest pattern that would be at the OES in the case of partially worn teeth. Our study points to increased variability in trigonid crest expression in M3s compared to M1s and M2s. Moreover, our analysis reveals that the SH sample matches broadly the trigonid crest patterns displayed by H. neanderthalensis and differs from those exhibited by H. sapiens, particularly in the almost constant expression of a continuous middle trigonid crest at the EDJ. However, SH hominins also exhibit patterns that have not been reported in H. neanderthalensis and H. sapiens samples. Other aspects of the variability of the trigonid crest expression at the dentine are presented and discussed.

La configuration de la crête du trigonide des molaires inférieures est caractéristique chez les homininés du Pléistocène supérieur, Homo neanderthalensis et Homo sapiens fossile, et aussi chez les humains actuels. Dans cet article, nous présentons l’examen de la configuration de la crête du trigonide d’un échantillon de molaires inférieures permanentes (n = 62) attribuées à Homo heidelbergensis, du site Pléistocène moyen de Sima de los Huesos (SH), en Espagne. L’expression de la crête est établie en modélisant les surfaces de l’émail et de la dentine à l’aide de rendus virtuels 3D basés sur un registre microtomographique (microCT). Le but de notre analyse est : 1) de caractériser le degré d’expression et la configuration de la crête du trigonide à la surface externe de l’émail (OES) et au niveau de la jonction émail–dentine (EDJ) de l’échantillon SH, 2) d’évaluer la concordance d’expression entre les deux surfaces et 3) de replacer le degré de variation de la crête du trigonide observé au sein de l’échantillon SH dans un contexte phylogénétique. En comparaison de celle observée sur l’OES (quatre types), nos résultats révèlent une grande variabilité dans l’expression de la crête du trigonide au niveau de l’EDJ (14 types). Malgré cette variabilité, dans presque tous les cas, l’expression d’une crête du trigonide intermédiaire ou distale continue sur l’OES correspond respectivement à l’expression d’une crête du trigonide mésiale/intermédiaire ou distale continue au niveau de l’EDJ. Ainsi, à partir de l’analyse de la morphologie interne, il est possible de prévoir le type de configuration de la crête du trigonide ayant existé sur l’OES dans le cas de dents partiellement usées. Notre étude souligne une variabilité croissante dans l’expression de ce trait sur les M3s par rapport aux M1s et M2s. En outre, notre étude révèle que le degré de variation des configurations de la crête du trigonide dans l’échantillon de SH s’accorde avec celui observé chez H. neanderthalensis, mais qu’il diffère de celui de H. sapiens, en particulier dans l’expression presque constante d’une crête du trigonide intermédiaire continue au niveau de l’EDJ. Cependant, les homininés de SH révèlent aussi des configurations qui n’ont pas été observées chez H. neanderthalensis et H. sapiens. D’autres aspects de la variabilité dans l’expression de la crête du trigonide au niveau de la dentine sont présentés et discutés.

 Permanent lower molars, Homo heidelbergensis, Sima de los Huesos, Homo neanderthalensis, Homo sapiens, Microtomography, Trigonid crest patterns Molaires inférieures permanentes, Homo heidelbergensis, Sima de los Huesos, Homo neenderthalensis, Homo sapiens, Microtomographie, Configuration des crêtes du trigonide presented Handled by Yves Coppens Introduction

Variation in dental form, as well as the frequency and degree of expression of many dental traits are highly heritable, making teeth more useful than other skeletal elements to assess phylogenetic relationships among fossil hominins and modern humans (Hrdlička, 1923, Kaifu et al., 2005, Martinón-Torres et al., 2007 and Martinón-Torres et al., 2012). The expression of morphological traits has been traditionally recorded in the outer enamel surface (OES), but other researchers, particularly during the last decade, have attempted to characterize the expression of these features at the enamel-dentine junction (EDJ) (Bailey et al., 2011, Macchiarelli et al., 2006, Skinner et al., 2008a and Skinner et al., 2008b). It is assumed that the morphology of the EDJ is largely responsible of the external morphology of a tooth (e.g., Guy et al., 2003, Nager, 1960 and Schwartz et al., 1998). According to Korenhof (Korenhof, 1982 and Scott and Turner, 1997) the EDJ morphology is more evolutionarily conservative than the OES morphology because “the enamel-dentine partition is much more a genetic blueprint of the occlusal anatomy of the teeth” (page 350 from Korenhof, 1982). However, the precise level of concordance between both surfaces is still under study (e.g., Bailey et al., 2011, Skinner et al., 2010, Skinner et al., 2009 and Skinner et al., 2008b).

In the past, in order to analyze the EDJ surface it was necessary to employ destructive techniques to remove the enamel cap, or that the teeth were broken or incomplete (e.g., Corruccini, 1998, Korenhof, 1982, Nager, 1960 and Suzuki and Sakai, 1973). More recently, the increasing availability of high-resolution micro-computed tomography (microCT) allows the virtual separation of the different tissues that compose a tooth, a process that is also called segmentation. From this segmentation, it is possible to obtain accurate three-dimensional (3D) reconstructions of detailed morphological features on both the OES and EDJ in a non-destructive manner. In this sense, dental microCT studies are expanding traditional analyses of discrete morphological traits to include their initial development on the inner enamel epithelium prior to enamel deposition; a methodology which has been shown to greatly improve our understanding of their ontogeny and variability within and among species (e.g., Ortiz et al., 2012, Skinner et al., 2008a and Skinner et al., 2008b).

The expression of trigonid crests on the enamel surface of human molars has revealed certain patterns of variation that seem to be taxonomically and phylogenetically informative (e.g., Bailey, 2002a, Irish, 1998, Scott and Turner, 1997, Turner et al., 1991 and Zubov, 1992a). From an evolutionary point of view, the primitive mammalian cusp pattern in molars was a triangle (Vandebroek, 1967 and Zubov, 1992b) (see Fig. 1.3 and Fig. 1.25 from Hillson, 2005). In humans, as in most primates, the mesial or anterior part of the lower molars is called the trigonid (trigon in upper molars), and the distal or posterior part of the lower molars is the talonid (talon in upper molars) (White and Folkens, 2005). Trigonid crest refers to the expression of a crest that connect the first (protoconid) with the second (metaconid) main cusps in lower molars (Scott and Turner, 1997 and Turner et al., 1991) and it can arise from any lobe segment or from the marginal ridge complex as well (Bailey et al., 2011; see also below).

Following the pioneer work of Korenhof (Korenhof, 1982 and Scott and Turner, 1997), trigonid crests have recently been studied at the EDJ by means of microCT (Bailey et al., 2011 and Skinner et al., 2008b). These studies have confirmed that the distinctive trigonid crest expression at the OES in Neanderthals is matched at the EDJ, both in terms of frequency and morphology (Bailey et al., 2011). In particular, the expression of conspicuous and continuous middle trigonid crests in lower molars has been interpreted as typical of Homo neanderthalensis (Bailey, 2002b and Martinón-Torres, 2006). We now know that this species is not only characterized by the high frequency of expression of middle trigonid crests, but also that the structures underlining this crest in the dentine may be different to the ones that contribute to the middle trigonid crest expression in other hominin species (Bailey et al., 2011). However, evolutionary origin and timing of this distinctive Neanderthal pattern remain unclear, as the EDJ expression of the trait has not been examined in geochronologically older hominin species.

In this study, we aim to investigate the expression of trigonid crests at EDJ of the Atapuerca-Sima de los Huesos lower molar sample by means of microCT. Sima de los Huesos site is a small cavity of approximately 8 m2 × 4 m2 that belongs to the Cueva Mayor-Cueva del Silo karst system (Atapuerca, Spain). This site has provided the largest Middle Pleistocene Homo fossil record coming from the same place and, to date, the human fossils recovered sum up more than 6500 remains, about the 80% of the worldwide human fossil record for the Middle Pleistocene (Bermúdez de Castro et al., 2004a and Bermúdez de Castro et al., 2004b). This extraordinary accumulation gives us the opportunity to study intrapopulation variability in a fossil population (Arsuaga et al., 1997, Arsuaga et al., 1991, Arsuaga et al., 1993 and Martinón-Torres et al., 2012). Sima de los Huesos hominins have been assigned to Homo heidelbergensis, a species that has been interpreted as ancestral to H. neanderthalensis, although the exact relationship between both taxa is still matter of debate (Arsuaga et al., 1997, Hublin, 2009, Martinón-Torres et al., 2012 and Mounier et al., 2009). Our objectives are to:

characterize the pattern of trigonid crest expression at the OES and EDJ of the SH sample;

evaluate the concordance of expression between both surfaces;

place trigonid crest variation in the SH sample into a phylogenetic context and in particular in relation to H. neanderthalensis.

Materials The Sima de los Huesos dental sample

Although the SH dental sample consists of 213 permanent molars, of which 123 are lower molars, we have only included in this study those that were isolated or included in a mandibular fragment small enough to fit in the microCT scanner. In addition, and following Bailey et al. (2011), we excluded those with wear degree higher than category 5 (Molnar, 1971).

 Comparative sample

We compare the SH molars to a sample of H. neanderthalensis specimens and to a sample of early and contemporary H. sapiens, since the latter is considered the sister lineage of H. neanderthalensis (Table 1). The Neanderthal and H. sapiens samples were taken from NESPOS© and ESRF© data bases, from El Mirador Cave – an archaeological site located at the southern side of the Sierra de Atapuerca with human remains from Calcolitic and Bronze Ages (Cáceres et al., 2007) – and from CENIEH's dental collection, composed of clinically extracted teeth from patients of known age and sex, representing a modern Spanish population.

 Methods

Each molar of the SH sample and the contemporary H. sapiens sample, as well as those from El Mirador, was scanned with a Scanco Medical AG Micro-Computed Tomography 80, housed at the CENIEH. The Scanco Medical Micro-CT80 system is characterized by a maximum scan size diameter of 75.8 mm and a maximum scan length of 140 mm, with image matrix from 512 × 512 to 4096 × 4096 pixels. Scans of the H. heidelbergensis teeth from Sima de los Huesos used a peak energy of 70 kV and intensity of scanning of 114 μA. The sample time of one projection is 800 000 μs, and the microCT takes nearly 500 parallel projections for each 180°. The resultant slice thickness ranged from 18 to 36 micrometers (μm). Segmentation of the microCT volume was done semi-automatically with manual corrections using AMIRA 5.3.3 (Visage Imaging, Inc.) and no filters were applied. For the rest of the comparative dental sample, we have obtained the microCT scans from the NESPOS© and ESRF© databases.

 Terminological considerations and scoring procedures

Dental literature has been employing different terminologies to refer to the same anatomical parts (e.g., Bailey et al., 2011, Carlsen, 1987, Korenhof, 1982, Martinón-Torres et al., 2012, Scott and Turner, 1997, Turner et al., 1991 and Wu and Turner, 1993). In order to avoid confusion and to allow the reproducibility of our method among researchers, we specify below the terms and the terminology sources we have used in this manuscript (Fig. 1). These terms and terminology are based on morphology of the OES. We consider a cusp an occlusal projection from the crown (White and Folkens, 2005) that is typically made up of three different segments (also called accessory lobes by Scott and Turner, 1997): mesial, middle or essential, and distal. The middle or essential segment is the lobe bearing the cusp (Carlsen, 1987 and Scott and Turner, 1997), and the free apex on top of the essential segment will be called cusp tip (note that Carlsen, 1987, called it essential cusp, but Scott and Turner, 1997, simply call it cusp).

Arising from these concepts about the basic components of a molar, we can now discuss the different terminologies published so far to refer to the trigonid crests. In 1982, Korenhof analyzed deciduous molars and defined a mesial (sic) trigonid crest – although Wu and Turner (1993) and Bailey et al. (2011) assumed that Korenhof also called it middle trigonid crest – and a distal trigonid crest. These crests were joining the mesial and the distal parts, respectively, of the two main mesial cusps of a lower molar. Wu and Turner (1993) defined a middle and a distal trigonid crest, but they did not find individuals expressing both crests at the same time. According to them, when a complete crest connects the middle portions of the mesial cusps and it lies mesialward, but not on the marginal border, it should be called a middle trigonid crest; otherwise it would be a distal trigonid crest. Scott and Turner (1997) only described a distal trigonid crest, defining it as a crest or ridge that courses buccolingually along the distal aspect of the primitive trigonid, now represented by the protoconid and metaconid.

Finally, in recent studies, Skinner et al. (2008b) and Bailey et al. (2011), scored crests at the EDJ based on their origin and the relative position of one to another, defining a mesial, a middle and a distal trigonid crest. According to them, a mesial trigonid crest could occur between the mesial marginal ridge and the middle trigonid crest or even between cusp tips as far as its position is substantially mesialward within the tooth contour [although according to Carlsen's terminology (Carlsen, 1987) that would be a middle-middle crest]. Although we acknowledge the atypical mesial position of some middle trigonid crests (Fig. 9b from Bailey et al., 2011), in order to standardize terms to discuss anatomy in such a detail, we have preferred to subscribe to the notion the middle or essential segment is the one bearing the cusp tip (Carlsen, 1987 and Scott and Turner, 1997). Within this framework, we propose that the three types of crests should be scored independently of the presence/absence of the other, in a similar way as they are scored at the OES (e.g., Guatelli-Steinberg and Irish, 2005 and Martinón-Torres et al., 2012).

Methodologically, to score the trigonid crests we draw a straight line from the protoconid to the metaconid tip. Any trigonid crest lying distal to this line is classified as a distal trigonid crest (DTC), and over or mesial to this line as a middle (MdTC) or mesial trigonid crest (MeTC) (Fig. 2). If one of the ends of the crest that runs along the mesial aspect of the protoconid and metaconid goes to the mesial marginal ridge we define it as a MeTC. If not, we define it as a MdTC [what Korenhof (1982) calls mesial trigonid crest, in our study is considered as a middle trigonid crest]. We are aware that this method may imply some problems in the case of teeth with more atypical shapes and where the cusp tips are displaced. However, it relies in the anatomical components of a tooth and thus, we believe it is more consistent in the identification of the components. The recognition of the crests depending on the co-expression of other crests and the relative position of one respect to another (Bailey et al., 2011) would be indeed more problematic in the case of displaced cusps or atypical shapes. Based on observations of the study sample it is not possible at the OES to reliably differentiate a mesial (i.e., one in which a part of the crest meets the marginal ridge) from a middle trigonid crest. Therefore, at the OES we only score a middle trigonid crest and a distal trigonid crest, following traditional scoring systems (Korenhof, 1982 and Scott and Turner, 1997). Furthermore, in those cases where an accessory cusp 7 (or metaconulid) is present and spatially belongs to the trigonid, we consider it as part of the distal segment of the metaconid. Summarizing, and taking into account all the anatomical considerations above, at the OES we recognize four basic types of trigonid crest expression (Fig. 3):

type A: a continuous MdTC and an absent or discontinuous DTC;

type B: continuous DTC and absent or discontinuous MdTC;

type C: continuous MdTC and DTC;

type D: absent or discontinuous MdTC and DTC.

Although both antimeres were analyzed, we employed the unilateral count method (Turner, 1987), and in case of asymmetry we have chosen the tooth with the highest degree of expression for the trait because we consider that it has more strongly expressed the genetic signal. Since the analyzed dental features are minimally affected by sexual dimorphism (Turner et al., 1991) both sexes have been pooled together. We report the inventory of the trigonid crest patterns of variation found at the EDJ in the SH molar sample, as well as the frequencies and percentages of the trigonid crests pattern at both the OES and EDJ per molar class. To test for significant correspondence of the TC patterns between the OES and the EDJ among SH, H. neanderthalensis and H. sapiens, a non-parametric Chi-square (PAST, Hammer et al., 2001) test is used. For the statistical analysis of the OES, the data were dichotomized into “presence” and “absence” (we considered that the trait is “present” when the crest is continuous, otherwise we consider that the trait is absent). For the statistical comparison of the EDJ, we compare the frequencies of “continuous”, “discontinuous”, and “absence”. Results were considered statistically significant with a P-value < 0.05.

 Results Trigonid crest types at the OES

Table 2 presents the frequency of expression of a continuous MdTC and DTC at the OES. We observe that the totality of the SH and Neanderthal M1 and M2 samples present a continuous MdTC. For the M3s, this type of crest is present in the 83.3% of the SH sample and the 90% of the H. neanderthalensis. Concerning to H. sapiens, almost half of the sample analyzed presents this expression of a MdTC, with the highest value for M1s (50%) and M3s (50%), followed by M2s (25%). Regarding a continuous DTC at the OES, we observe that for the SH sample, these are only present in the M3s. Meanwhile, the Neanderthal sample is the only one that exhibits continuous DTC in all molar types, with an increase from M1 to M3. In the sample of H. sapiens, there are no examples with continuous DTC.

The results of analyses of trait frequencies at the OES between species are provided in Table 3. If we dichotomize the expression into present (continuous crest) and absent (discontinuous or absent crest) we observe that SH and Neanderthals possess significantly higher frequencies of MdTC presence than H. sapiens for the M1, M2 and total molar sample. For the M3s, the frequency of a MdTC in H. sapiens is lower but they are not significant. Regarding the frequency of DTC presence, once again SH and Neanderthals present higher frequencies for the M3 and for the whole molar sample. In summary, H. sapiens shows significant differences (P < 0.05) when compared with SH and Neanderthal samples; the former possessing a lower frequency of trigonid crest presence than the two latter.

 Trigonid crest types at the EDJ

The variety and combinations of trigonid crests on the dentine surface of the SH sample is much greater than that expressed at the enamel surface. Table 4 presents comprehensive typology and description of all types, which are also illustrated in a schematic form (Fig. 4) and on actual SH specimens (Fig. 5). Table 5 shows that the expression of the continuous MeTC usually increases from the M1 to the M3. H. sapiens are the only group with a discontinuous MeTC. For the SH sample we did not record any M1 with a continuous MeTC. In general terms, the continuous MdTC decreases its expression from the M1s to the M3s. Overall, M1s from the SH sample present this expression of a MdTC. As it was the case for the MeTC, only H. sapiens exhibits a discontinuous MdTC. In relation to the continuous DTC on the dentine, we can observe that this expression is only present in SH and Neanderthals. The H. sapiens sample only presents a discontinuous DTC and only in M1s and M2s. In particular, all the M2s from SH and 7 of 8 Neanderthal M2s present a discontinuous DTC.

Table 6 displays the results of analyses of trait expression (continuous/discontinuous/absent) at the EDJ among species. Concerning the MeTC, it is relevant to note that there are no significant differences among the three groups whereas for the MdTC and the DTC, H. sapiens exhibits significant differences with the other two. SH sample possesses significantly higher frequencies of MdTC expression than H. sapiens for the M1, M2 and total molar sample. Regarding the DTC, SH shows higher frequencies of expression than H. sapiens for the M2, M3 and the total molar sample. Thus, the frequencies of MdTC for the M3s and DTC for M1s in H. sapiens are lower, but not significant. Meanwhile, Neanderthals shows significantly higher frequencies of MdTC and DTC expression than H. sapiens but only for the total molar sample. For the M1s, M2s and M3s the frequencies in H. sapiens once again are lower, but not significant.

 Variability of the trigonid crest types at the EDJ

The four basic types of trigonid expression at the OES contrast with a wider spectrum of crests types and combinations at the EDJ. As seen in Table 7, type A at the OES is associated with higher variability at the EDJ (10 types), followed by type C (6 types at the EDJ), type D (4 types) and type B (1 type). However, and despite this variability, all teeth generally present a strong correlation between the enamel and dentine surface in the sense that a continuous MdTC at the OES corresponds with a continuous MeTC or MdTC at the EDJ, and vice versa. Similarly, a continuous DTC at the OES tends to correspond with a continuous DTC at the EDJ.

Type A at the OES (continuous MdTC with absent or discontinuous DTC) can be represented at the dentine surface by a continuous MeTC or MdTC in combination with a discontinuous DTC, except in Qafzeh 10 (left M2) where a MeTC or MdTC at the EDJ is absent (type 4). Type B (continuous DTC with absent or discontinuous MdTC) at the OES corresponds with an absent or discontinuous MeTC or MdTC and a continuous DTC in all cases. Type C at the OES (continuous MdTC and DTC) corresponds to the co-expression of two continuous crests also at the EDJ, except in AT-100 (left M3) where the MeTC or MdTC at the EDJ is absent -type 3-, AT-811 (right M3) where the DTC at the EDJ is discontinuous -type 14- and in Regourdou 1, Krapina D4 and Krapina D106 that show at the EDJ a type 2 (MeTC continuous with an absent or discontinuous DTC) crest. Type D (absent or discontinuous MdTC and DTC) is also related on the dentine with an absent or discontinuous MeTC or MdTC and a discontinuous DTC in all cases except EQ-H5 (right M1) where the MdTC is continuous at the EDJ -type 11- and MIR4 P22 205 (left M1) where there is a continuous MeTC or type 2 at the EDJ.

 Frequencies of expression for the types at the OES and EDJ by tooth class

We have detailed the frequency of the different types of trigonid crest expression on both enamel and dentine surfaces for the M1 (Table 8), M2 (Table 9) and M3 (Table 10) samples using the individual count method. In the M1s, the highest frequency of expression at the OES for the SH and Neanderthals samples corresponds to type A, with a 100% for the former and all but one specimen for the latter. In contrast, in H. sapiens type A is expressed by half of the sample whereas the other half expresses type D. Interestingly, type D is absent in SH and H. neanderthalensis. At the EDJ, and taking into account that this surface shows more variability than the enamel, we can observe that the most representative trigonid crest types for the SH, Neanderthal and H. sapiens samples are type 8, type 10 and type 4, respectively.

Concerning the M2s (Table 9) and at OES, we can observe the same pattern for the SH and Neanderthal samples, with type A being displayed by the totality of the sample in the former and all but one in the latter. However, for H. sapiens the maximum value at the OES is represented by type D (which in turn is absent in SH and Neanderthal M2s) and the lowest by type A (which in turn was the most frequent type in SH and H. neanderthalensis M2s). At the EDJ, the highest frequencies are for types 11 and 13 for the SH M2s; types 7, 10 and 14 for Neanderthals; and type 4 for H. sapiens, which is present by a little more than half of the sample.

With regard to the M3 samples, there is a greater variability between the trigonid crest types on enamel and dentine than in M1s and M2s (Table 10). For the SH sample, the four types are represented, with type C being the most frequent and type D the least. Neanderthals exhibit the same pattern as the SH sample with the exception that they do not display type B. In H. sapiens, half of the sample presents type A and the other half displays type D. At the EDJ, the highest frequency for the SH M3s corresponds to types 5 and 6 in Neanderthals corresponds to type 2, and in H. sapiens to type 4.

 Discussion Variability and frequency of expression at the EDJ

Despite a greater variability in the expression of trigonid crests at the EDJ (14 types) compared to the OES (4 types), in almost all cases there is concordance between the presence of a continuous mid-trigonid and/or distal trigonid crest at the OES and the presence of a continuous mesial/mid-trigonid and/or distal trigonid crest at the EDJ. As an example, and with only a few exceptions presented below, a type A at the OES (continuous MdTC) always corresponds at the EDJ with types where the MeTC or the MdTC is continuous (types 1, 2, 7, 8, 10, 11, 12, 13, 14). Similarly, a type C at the OES (continuous DTC) always corresponds at the EDJ with types where the DTC – regardless the origin – is continuous. This is particularly useful in the case of worn teeth, because it allows estimation of the type of trigonid crest pattern that would have been present at the unworn OES.

Regarding the distal trigonid crest at the EDJ, Wu and Turner (1993) did not find individuals expressing both a MdTC and DTC, and Bailey et al. (2011) did not observe molars with a distal-distal configuration (“true” distal trigonid crest sensu Bailey et al., 2011). Conversely, our study identified several cases of co-expression of MdTC and DTC, and we have also found cases of “true” distal trigonid crest in SH sample; although they are not very common (2 molars or 4.7%) and only M3s display it (Fig. 6). Concerning a crest that connects middle and distal segments, Bailey et al. (2011) found only a small number of cases (and only in Homo sapiens) whereas in our analysis we have found five M3s from SH sample (11.6%) and four molars from Krapina – one M1, one M2 and two M3s – from H. neanderthalensis (11.8%). However, there are no examples in our H. sapiens sample that present this type of trigonid crest pattern. These discrepancies may be related to the limitations of trying to characterize the variability of a species with small sample sizes and possibly to some differences in the scoring system employed in Bailey et al. (2011) and our study (see Terminological considerations and Scoring procedures section).

Concerning the origin of the crests that conform the middle trigonid crest at the EDJ, we can observe that from the SH molars analyzed (n = 43) there are 31 specimens with MdTC, 20 of which (64.5%) present a middle-middle crest (see Fig. 7) (i.e., a “true” middle trigonid crest sensu Bailey et al., 2011) whereas the remaining 11 (35.5%) present a MdTC of mesial-middle origins. Moreover, in our study we have found that the majority of the molars with a MdTC show a crest that remains high between the cusps (grade 3 from Bailey et al., 2011), a type that Bailey et al., 2011 found in their Neanderthal sample, but not in their modern human sample. In accordance with Bailey et al. (2011), this type of crest was also recorded in our H. neanderthalensis sample (67.6%) and it was absent in H. sapiens. However, grade 3 (sensu Bailey et al., 2011) was also present in 77.4% of the SH group.

In addition, among all the molar classes, the M3s are those with more variability and complexity. Their rather atypical morphologies (i.e., rotation of the cusps within the tooth contour or reduced main cusps in combination with accessory cusps) (see Martinón-Torres et al., 2012), prevented in some cases an easy identification of the morphological features. As an example, in AT-100 the marginal ridge is atrophied and ends at the mesial part of the molar, making it difficult to distinguish whether there is a continuous MeTC or not (Fig. 8). Indeed, most of the cases showing a “mismatch” between surfaces corresponded to atypical teeth. Despite of this inconvenience, there is a general concordance in the expression of continuous crests at the OES and the EDJ, with only some exceptions in which what seems to be a continuous MdTC on the enamel surface actually is a discontinuous MeTC at the EDJ (two for SH, three for Neanderthals and one for H. sapiens sample) (Fig. 9); and with three H. sapiens molars in which what seems to be a discontinuous MdTC on the enamel surface actually is a continuous MdTC at the EDJ (Fig. 10). As stated above, it is uncommon and when it occurs it is usually in atypical teeth and often complicated with the expression of a C7.

 Taxonomical implications

In our study, we have followed some traditional criteria from the ASUDAS (Turner et al., 1991 and Wu and Turner, 1993) to score the trigonid crest at the OES (for more details see Terminological Considerations above), but we have also recorded the full variability expressed at the EDJ to have a more precise knowledge morphological variation in the SH sample. According to Bailey et al. (2002a,b) and Martinón-Torres et al. (2012) the expression at the OES of a continuous MdTC is a typical Neanderthal feature, but this feature is also present in the Middle Pleistocene hominins fossils from Europe like Sima de los Huesos and Arago—but not in Mauer despite being the holotype of the species Homo heidelbergensis. This trait can be also present in earlier species like H. erectus, H. georgicus and H. antecessor but their expression is less pronounced than in Neanderthals and the frequencies tend to be lower (e.g., Martinón-Torres et al., 2007, Martinón-Torres et al., 2008, Martinón-Torres et al., 2012 and Zanolli, 2013). The pattern of expression of this feature at the OES coincides with previous studies about the OES of SH dentitions that have shown that these fossils present the classic Neanderthal combination of traits (Bermúdez de Castro, 1986, Bermúdez de Castro et al., 1999, Martinón-Torres et al., 2007 and Martinón-Torres et al., 2012). Now, using microCT and virtual segmentations, we can ratify those resemblances at the EDJ. Our study reveals that at the EDJ there is a greater number of SH molars with a middle or mesial trigonid crest than without it, that the middle crest is more frequent than the mesial one and that there is a high proportion of lower molars with a continuous mesial (MeTC) or middle trigonid crest (MdTC) and a discontinuous distal trigonid crest (DTC), which means a clear predominance of type A at the OES. This type is also frequent in our Neanderthal sample and very rare in our H. sapiens sample, ratifying the results of Bailey et al. (2011). In addition to the differences in the frequency of expression, there are six trigonid crest types that are expressed in the SH and H. neanderthalensis that are absent in the H. sapiens group (i.e., types 3, 5, 6, 9, 13 and 14). If analysis of larger and geographically diverse modern human samples confirms this difference, it may be of particular utility in the taxonomic assignment of isolated teeth.

Although a direct comparison between our study and Bailey et al. (2011) analysis is not possible because of the different scoring systems employed, it was possible to make some general assessments between both (Table 11). Thus, if we focus on the percentages that show a continuous MdTC - grade 2 and 3 of Bailey et al. (2011) with a discontinuous DTC, we can assert that the almost constant expression of this pattern in our SH and Neanderthals molars in contrast with the low percentages in our H. sapiens sample, is in concordance with the results presented by Bailey et al. (2011) (Fig. 11). Indeed, when a continuous MdTC is present in other group rather than SH or H. neanderthalensis, the crests are low and dip further at the sagittal sulcus (this study; Bailey et al., 2011 and Zanolli and Mazurier, 2013). As previously stated by Bailey et al., 2011, the structures underlining the MdTC in the dentine tend to be different to the ones that contribute to the middle trigonid crest in other hominin species (Bailey et al., 2011). As an example, the expression of a continuous MdTC at the EDJ in other groups such as Pan, A. africanus, Homo sapiens (Bailey et al., 2011) or the North African Middle Pleistocene hominins from Tighenif (Zanolli and Mazurier, 2013) is usually associated to a grade 2. In addition, the expression of a continuous MdTC in H. sapiens is never associated with a grade 3 (this study; Bailey et al., 2011).

We have also identified a wide spectrum of trigonid crest types at the EDJ. Importantly, there are three trigonid crests types (types 3, 9 and 13) that are only present in the SH sample and have not been found in the Neanderthal or the H. sapiens group. Future studies in larger samples are necessary to know whether these SH exclusive types are present in other hominin groups or if they are particular to the SH group. If they are present in earlier hominins it would mean SH retains some primitive features that are lost in classic Late Pleistocene Neanderthals, as it happens with other cranial and postcranial elements (Arsuaga et al., 1997). On the contrary, if it is confirmed that those types are not present in any other hominin group, they should be considered as derived (and maybe autopomorphic) features for the SH hominin group, and they would ratify the highly derived state of this population with regard to H. neanderthalensis (Gómez-Robles et al., 2007, Gómez-Robles et al., 2008, Gómez-Robles and Polly, 2012 and Martinón-Torres et al., 2012).

 Conclusions

In agreement with Bailey et al. (2011) and because the EDJ exhibits a greater number of trigonid crest types than the enamel, we developed a comprehensive scoring system to characterize trigonid crest patterning in the Sima de los Huesos lower molar sample. Our analysis reveals higher variability of trigonid crest expression at the EDJ compared to the OES; however, our results confirm the general concordance between the two surfaces with regard to the expression of continuous mesial/middle and distal trigonid crests (Bailey et al., 2011 and Skinner et al., 2008b). Thus, the manifestation of the trigonid crest patterns at the enamel is mostly determined by the development of the EDJ. This fact is particularly relevant as it allows prediction of the type of trigonid crest pattern at the OES in the case of excessively worn teeth. This also maximizes the number of teeth that can be analyzed, enlarging the sample size particularly in the case of the scarce fossil human remains.

In concordance with previous studies based on the OES morphology, our study confirms that SH displays trigonid crest patterns that have been classically considered as typical of H. neanderthalensis and significantly different from H. sapiens. However, it has also shown some trigonid crest types that were present neither in the H. neanderthalensis sample nor in the H. sapiens sample. Analysis of other Pleistocene hominins will improve the taxonomic and phylogenetic information about these traits, as well as, to make possible evolutionary inferences about the expression of these traits in relation with other ancestral hominins.

 Acknowledgments

We thank all the members of the Atapuerca research team, in particular those who excavate the Sima de los Huesos site, for their laborious and dedicated work. We also acknowledge several people for providing access to the studied material and for their helpful assistance: R. Macchiarelli, P. Bayle and P. Semal from NESPOS society; E. Carbonell and M. Lozano (Institut Català de Paleoecologia Humana i Evolució Social, Tarragona, Spain). We are also grateful to G.R. Scott for valuable revisions and instructive comments that have definitely helped to improve the manuscript. We also recognize the inestimable help of P. Fernandez and E. Lacasa from the Restoration and Conservation Laboratory of the CENIEH, and M.C. Ortega from Centro de Evolución y Comportamiento Humanos (ISCIII, Madrid) for the restoration and conservation of the SH collection. We are also grateful to M. Modesto from the CENIEH for his suggestions in the statistical analysis. The mCT scanner of the SH, Mirador and CENIEH dental collection were performed at CENIEH with the collaboration of CENIEH staff. We want to recognize the inestimable help of two anonymous reviewers and the editor for their constructive comments. Any error is solely the responsibility of the authors.

This research was supported with funding from the Dirección General de Investigación of the Spanish Ministerio de Educación y Ciencia (Project N_ CGL2009-12703-C03-01); Secretaría de Estado, Investigación, Desarrollo e Innovación of the Spanish Ministerio de Economía y Competitividad (Project N_ CGL2012-38434-C03-02); Consejería de Educación de la Junta de Castilla y León (Project CEN074A12-1) and the Leakey Foundation. Fieldwork at Atapuerca is supported by the Consejería de Cultura y Turismo of the Junta de Castilla y León and the Fundación Atapuerca. We are also grateful to the altruistic help of Fundación Caja Rural de Burgos in the investigation issues. M.M-P. has the benefit of a predoctoral grant of the Fundación Atapuerca.

Grant sponsorship: Spanish Ministry of Ciencia e Innovación, Ministry of Culture of the Junta de Castilla y León (Spain), Fundación Atapuerca, Leakey Foundation.

 Arsuaga J.L. Carretero J.M. Lorenzo C. Gracia A. Martínez I. Bermúdez de Castro J.M. Carbonell E. Size variation in Middle Pleistocene humans Science 277 1997 1086–1088 Arsuaga J.L. Carretero J.M. Martínez I. Gracia A. Cranial remains and long bones from Atapuerca/Ibeas (Spain) J. Hum. Evol. 20 1991 191–230 Arsuaga J.L. Martínez I. Gracia A. Carretero J.M. Carbonell E. Three new human skulls from the Sima de los Huesos Middle Pleistocene site in Sierra de Atapuerca, Spain Nature 362 1993 534–537 Bailey S.E. Neandertal Dental Morphology: Implications for Modern Human Origins 2002 Arizona State University Arizona 238 p Bailey S.E. A closer look at Neanderthal postcanine dental morphology: I. The mandibular dentition New Anat. 269 2002 148–156 Bailey S.E. Skinner M.M. Hublin J.J. What lies beneath? An evaluation of lower molar trigonid crest patterns based on both dentine and enamel expression Am. J. Phys. Anthropol. 145 2011 505–518 Bermúdez de Castro J.M. Dental remains from Atapuerca (Spain) I. Metrics J. Hum. Evol. 15 1986 265–287 Bermúdez de Castro J.M. Martinón-Torres M. Carbonell E. Sarmiento S. Rosas A. Van der Made J. Lozano M. The Atapuerca sites and their contribution to the knowledge of human evolution in Europe Evol. Anthropol. Issues News Rev. 13 2004 25–41 Bermúdez de Castro J.M. Martinón-Torres M. Lozano M. Sarmiento S. Muela A. Paleodemography of the Atapuerca-Sima de los Huesos hominin sample: a revision and new approaches to the paleodemography of the European Middle Pleistocene population J. Anthropol. Res. 60 2004 5–26 Bermúdez de Castro J.M. Rosas A. Carbonell E. Nicolás M.E. Rodríguez J. Arsuaga J.L. A modern human pattern of dental development in Lower Pleistocene hominids from Atapuerca-TD6 (Spain) Proc. Natl. Acad. Sci. U S A 96 1999 4210–4213 Cáceres I. Lozano M. Saladié P. Evidence for Bronze Age cannibalism in El Mirador Cave (Sierra de Atapuerca, Burgos, Spain) Am. J. Phys. Anthropol. 133 2007 899–917 Carlsen O. Dental Morphology 1987 Munksgaard Copenhagen, Denmark 181 Corruccini R.S. The Dentino-Enamel Junction in Primate Mandibular Molars. Human Dental Development, Morphology, and Pathology: A Tribute to Albert A. Dahlberg 1998 University of Oregon Anthropological Papers Portland 1–16 Gómez-Robles A. Martinón-Torres M. Bermúdez de Castro J.M. Margvelashvili A. Bastir M. Arsuaga J.L. Pérez-Pérez A. Estebaranz F. Martínez L.M. A geometric morphometric analysis of hominin upper first molar shape J. Hum. Evol. 53 2007 272–285 Gómez-Robles A. Martinón-Torres M. Bermúdez de Castro J.M. Prado L. Sarmiento S. Arsuaga J.L. Geometric morphometric analysis of the crown morphology of the lower first premolar of hominins, with special attention to Pleistocene Homo J. Hum. Evol. 55 2008 627–638 Gómez-Robles A. Polly P.D. Morphological integration in the hominin dentition: evolutionary, developmental, and functional factors Evolution 66 2012 1024–1043 Guatelli-Steinberg D. Irish J.D. Brief communication: early hominin variability in first molar dental trait frequencies Am. J. Phys. Anthropol. 128 2005 477–484 Guy F. Brunet M. Schmittbuhl M. Viriot L. New approaches in hominoid taxonomy: morphometrics Am. J. Phys. Anthropol. 121 2003 198–218 Hammer Ø. Harper D.A.T. Ryan P.D. Past: paleontological statistics software package for education and data analysis Palaeontol. Electro. 4 1 2001 9 Hillson S. Teeth Second revised ed. 2005 Cambridge University Press Cambridge 388 p Hrdlička A. Dimensions of the first and second lower molars with their bearing on the Piltdown jaw and on man's phylogeny Am. J. Phys. Anthropol. 6 1923 195–216 Hublin J.J. The origin of Neandertals Proc. Natl. Acad. Sci. U S A 106 2009 16022–16027 Irish J.D. Diachronic and synchronic dental trait affinities of Late and Post-Pleistocene peoples from North Africa Homo 49 1998 138–155 Kaifu Y. Baba H. Aziz F. Indriati E. Schrenk F. Jacob T. Taxonomic affinities and evolutionary history of the Early Pleistocene hominids of Java: dentognathic evidence Am. J. Phys. Anthropol. 128 2005 709–726 Korenhof C.A.W. Evolutionary Trends of the Inner Enamel Anatomy of Deciduous Molars from Sangiran (Java, Indonesia). Teeth: Form, Function and Evolution 1982 Columbia University Press New York 350–365 Macchiarelli R. Bondioli L. Debénath A. Mazurier A. Tournepiche J.F. Birch W. Dean M.C. How Neanderthal molar teeth grew? Nature 444 2006 748–751 Martinón-Torres M. Evolución del aparato dental en homínidos: estudio de los dientes humanos del Pleistoceno de la Sierra de Atapuerca (Burgos) (Ph.D. Dissertation) 2006 Santiago de Compostela University Spain 626 p Martinón-Torres M. Bermúdez de Castro J.M. Gómez-Robles A. Arsuaga J.L. Carbonell E. Lordkipanidze D. Manzi G. Margvelashvili A. Dental evidence on the hominin dispersals during the Pleistocene Proc. Natl. Acad. Sci. U S A 104 2007 13279–13282 Martinón-Torres M. Bermúdez de Castro J.M. Gómez-Robles A. Margvelashvili A. Prado L. Lordkipanidze D. Vekua A. Dental remains from Dmanisi (Republic of Georgia): morphological analysis and comparative study J. Hum. Evol. 55 2008 249–273 Martinón-Torres M. Bermúdez de Castro J.M. Gómez-Robles A. Prado-Simón L. Arsuaga J.L. Morphological description and comparison of the dental remains from Atapuerca-Sima de los Huesos site (Spain) J. Hum. Evol. 62 2012 7–58 Molnar S. Human tooth wear, tooth function and cultural variability Am. J. Phys. Anthropol. 34 1971 175–190 Mounier A. Marchal F. Condemi S. Is Homo heidelbergensis a distinct species? New insight on the Mauer mandible J. Hum. Evol. 56 2009 219–246 Nager G. Comparison between the spatial relations of the dentin crown relief and the enamel relief of dental crowns Acta Anat. 42 1960 226 Ortiz A. Skinner M.M. Bailey S.E. Hublin J.J. Carabelli's trait revisited: an examination of mesiolingual features at the enamel-dentine junction and enamel surface of Pan and Homo sapiens upper molars J. Hum. Evol. 63 2012 586–596 Schwartz G.T. Thackeray J.F. Reid C. van Reenan J.F. Enamel thickness and the topography of the enamel–dentine junction in South African Plio-Pleistocene hominids with special reference to the Carabelli trait J. Hum. Evol. 35 1998 523–542 Scott G.R. Turner C.G. II. The Anthropology of Modern Human Teeth: Dental Morphology and its Variation in Recent Human Populations 1997 Cambridge University Press Cambridge 382 p Skinner M.M. Evans A. Smith T. Jernvall J. Tafforeau P. Kupczik K. Olejniczak A.J. Rosas A. Radovčić J. Thackeray J.F. Toussaint M. Hublin J.J. Brief communication: contributions of enamel-dentine junction shape and enamel deposition to primate molar crown complexity Am. J. Phys. Anthropol. 142 2010 157–163 Skinner M.M. Gunz P. Wood B.A. Boesch C. Hublin J.J. Discrimination of extant Pan species and subspecies using the enamel-dentine junction morphology of lower molars Am. J. Phys. Anthropol. 140 2009 234–243 Skinner M.M. Gunz P. Wood B.A. Hublin J.J. Enamel-dentine junction (EDJ) morphology distinguishes the lower molars of Australopithecus africanus and Paranthropus robustus J. Hum. Evol. 55 2008 979–988 Skinner M.M. Wood B.A. Boesch C. Olejniczak A.J. Rosas A. Smith T.M. Hublin J.J. Dental trait expression at the enamel-dentine junction of lower molars in extant and fossil hominoids J. Hum. Evol. 54 2008 173–186 Suzuki M. Sakai T. Occlusal surface pattern of the lower molars and second deciduous molar among the living polynesians Am. J. Phys. Anthropol. 39 1973 305–316 Turner C.G. Late Pleistocene and Holocene population history of East Asia based on dental variation Am. J. Phys. Anthropol. 73 1987 305–321 Turner C.G. Nichol C.R. Scott G.R. Scoring procedures for key morphological traits of the permanent dentition: the Arizona State University Dental Anthropology System Kelley M. Larsen C. Advances in Dental Anthropology 1991 Wiley Liss New York 13–31 Vandebroek G. Origin of the cusps and crests of the tribosphenic molar J. Dent. Res. 46 1967 796–804 White T.D. Folkens P.A. The Human Bone Manual 2005 Academic Press Burlington, USA 464 p Wu L. Turner C.G. Brief communication: variation in the frequency and form of the lower permanent molar middle trigonid crest Am. J. Phys. Anthropol. 91 1993 245–248 Zanolli C. Additional evidence for morpho-dimensional tooth crown variation in a new Indonesian H. erectus sample from the Sangiran Dome (Central Java) Plos One 8 2013 e67233 Zanolli C. Mazurier A. Endostructural characterization of the H. heidelbergensis dental remains from the early Middle Pleistocene site of Tighenif, Algeria C. R. Palevol. 12 2013 293–304 Zubov A. The epicristid or middle trigonid crest defined Dent. Anthropol. Newsl. 6 1992 9–10 Zubov A. Some dental traits in different evolutionary lines leading to modern man Dent. Anthropol. Newsl. 6 1992 4–8

Scheme of some of the molar crown structures used in this study. MS: mesial segment; ES: essential segment; DS: distal segment; CT: cusp tip.

Schéma de quelques structures de couronne de molaires utilisées dans cette étude. MS : segment mésial ; ES : segment essentiel ; DS : segment distal ; CT : pointe de la cuspide.

(Color online.) A. Occlusal enamel surface of AT-2271 (M3) from Sima de los Huesos (SH) illustrating a middle trigonid crest (MdTC) and a distal trigonid crest (DTC). B. Occlusal dentine surface from AT-2271 illustrating a corresponding middle trigonid crest (MdTC) and a distal trigonid crest (DTC). C. Dentine from AT-143 (M3) of the SH site where we can see a mesial trigonid crest (MeTC). The green line from protoconid tip to metaconid tip defines the position and classification of the crests.

(Couleur en ligne.) A. Surface occlusale d’émail d’AT-2271 (M3) de SH illustrant une crête du trigonide intermédiaire (MdTC) et une crête du trigonide distale (DTC). B. Surface occlusale de dentine d’AT-143 (M3) du site SH où l’on observe une crête du trigonide médiale (MeTC). C. La ligne verte depuis la pointe du protoconide jusqu’à la pointe du métaconide définit la position et la classification des crêtes.

(Color online.) Top: scheme of the four basic types of trigonid crests on enamel surface identified in our study. Scheme represents the occlusal surface at the outer enamel surface (OES) including the tips of the cusps (represented by the open circles) of the protoconid and metaconid (mmr: mesial marginal ridge). Explanation in the text. Bottom: examples of different types of trigonid crests on the enamel surface: type A (AT-946); type B (AT-2385); type C (AT-2271); type D (AT-943).

(Couleur en ligne.) En haut : schéma des quatre types de base des crêtes du trigonide sur la surface de l’émail identifié dans notre étude. Le schéma représente la surface occlusale au niveau de l’OES incluant les pointes des cuspides (représentées par les cercles ouverts) des protoconide et métaconide (mmr : arête marginale mésiale). Explication dans le texte. En bas : exemples de différents types de crêtes du trigonide sur la surface de l’émail : type A (AT-946) ; type B (AT-2835) ; type C (AT-2271) ; type D (AT-943).

Based on Korenhof, 1982.

(Color online.) Scheme of all types of trigonid crests on dentine surface identified in our study from the Sima de los Huesos (SH) molar sample. Explanation in Table 4. View is towards the occlusal surface of the enamel-dentine junction (EDJ) including the tips of the dentine horn (represented by circles) of the protoconid and metaconid and the mesial marginal ridge (represented by the continuous black line). Dashed lines mean that the crest is absent or discontinuous.

(Couleur en ligne.) Schéma de tous les types de crêtes du trigonide sur la surface de dentine identifiées dans notre étude sur l’échantillon de molaire SH. Explication dans le Tableau 4. La vue est vers la surface occlusale de l’EDJ, incluant les pointes de la corne de dentine (représentées par des cercles) des protoconide et métaconide et de l’arête marginale mésiale (représentée par une ligne continue noire). Les lignes tiretées signifient que la crête est absente ou discontinue.

(Color online.) Examples of different types of trigonid crests on the dentine: type 1 (AT-169); type 2 (AT-1761); type 3 (AT-1945); type 4 (AT-1473); type 5 (AT-2271); type 6 (AT-942); type 7 (AT-141); type 8 (AT-829); type 9 (AT-2438); type 10 (AT-3934); type 11 (AT-2276); type 12 (AT-2270); type 13 (AT-557); type 14 (AT-811). Right molar images (type 2, 5, 6, 7, 9 and 11) have been mirrored to facilitate the comparisons.

(Couleur en ligne.) Exemples de différents types de crêtes du trigonide sur la dentine : type 1 (AT-169) ; type 2 (AT-1761) ; type 3 (AT-1945) ; type 4 (AT-1943) ; type 5 (AT-2271) ; type 6 (AT-942) ; type 7 (AT-141) ; type 8 (AT-829) ; type 9 (AT-2438) ; type 10 (AT-3934) ; type 11 (AT-2276) ; type 12 (AT-2270) ; type 13 (AT-557) ; type 14 (AT-811). Les images de molaires droites (des types 2, 5, 6, 7, 9 et 11) ont été réfléchies par un miroir pour faciliter la comparaison.

(Color online.) Illustration of a “true” distal trigonid crest at enamel-dentine junction (EDJ) and outer enamel surface (OES) in an M3 molar specimen (AT-2385). Virtual reconstruction of dentine and enamel computer models on micro-computed tomography.

(Couleur en ligne.) Illustration d’une « vraie » crête du trigonide distale au niveau de l’EDJ et de l’OES dans un spécimen de molaire M3 (AT-2385). Reconstitution virtuelle de modèles de dentine et d’émail informatisés sur un registre microtomographique.

(Color online.) Illustration of a “true” middle trigonid crest at enamel-dentine junction (EDJ) and outer enamel surface (OES) in an M2 molar specimen (AT-946). Virtual reconstruction of dentine and enamel computer models based on micro-computed tomography.

(Couleur en ligne.) Illustration d’une « vraie » crête du trigonide intermédiaire au niveau de l’EDJ et de l’OES d’un spécimen de molaire M2 (AT-946). Reconstitution virtuelle de modèles de dentine et d’émail informatisés sur un registre microtomographique.

(Color online.) Enamel and dentine surface of a M3 (AT-100) with an atrophied mesial marginal ridge (MMR) (see text for specific mention). Virtual reconstruction of dentine and enamel computer models on micro-computed tomography.

(Couleur en ligne.) Surface de dentine et d’émail d’une M3 (AT-100) avec une arête marginale mésiale (MMR) atrophiée (voir texte pour mention spécifique). Reconstitution virtuelle de modèles de dentine et d’émail informatisés sur un registre microtomographique.

(Color online.) Enamel and dentine surface of a M3 (AT-811) with lack of correspondence (see text for specific mention). Virtual reconstruction of dentine and enamel computer models on micro-computed tomography.

(Couleur en ligne.) Surface de l’émail et de la dentine d’une M3 (AT-811) avec absence de correspondance (voir le texte pour plus de détails). Reconstruction virtuelle de modèles informatiques pour l’émail et la dentine à partir de données microtomographiques.

(Color online.) Enamel and dentine surface of a M1 (EQ-H5) with lack of correspondence (see text for specific mention). Virtual reconstruction of dentine and enamel computer models on micro-computed tomography.

(Couleur en ligne.) Surface de l’émail et de la dentine d’une M1 (EQ-H5) avec absence de correspondance (voir le texte pour plus de détails). Reconstruction virtuelle de modèles informatiques pour l’émail et la dentine à partir de données microtomographiques.

Illustrations of differences among specimens for grades of presence of middle trigonid crest (MdTC) at the outer enamel and enamel-dentine junction surfaces.

Illustration des différences entre les spécimens selon les degrés de présence des crêtes du trigonide intermédiaires (MdTC) au niveau de l’OES et de la EDJ.

Study sample of lower molars.

Liste des molaires inférieures étudiées dans ce travail.

Species Origin N total M1 (n) M2 (n) M3 (n) H. heidelbergensis Sima de los Huesos a 62 22 18 22 H. neanderthalensis Engis b 1 1 – – Gibraltar b 3 2 1 – Ehringsdorf c 1 1 – – Abri Bourgeois-Delaunay c 1 1 – – Regourdou c 6 2 2 2 Abri Suard c 4 3 – 1 Krapina c 20 7 6 7 Hunas c 1 – – 1 Roc de Marsal c 2 2 – – H. sapiens Equus Cave b 2 2 – – Qafzeh b 8 4 4 – Lagar Velho c 1 1 – – El Mirador a 9 3 3 3 CENIEH a 12 4 7 1

CENIEH micro-computed tomography data base.

ESRF© data base.

NESPOS© data base.

Percentages and frequencies of the continuous middle (MdTC) and distal trigonid crest (DTC) at the OES (only one antimere included). We consider the trait is “present” when the crest is continuous. Otherwise we consider that the trait is “absent”.

Pourcentages et fréquences des crêtes du trigonide intermédiaires (MdTC) et distales (DTC) au niveau de l’OES (un seul antimère inclus). Le trait est considéré comme « présent » lorsque la crête est continue. Sinon, le caractère est considéré comme « absent ».

Total molar sample M1 M2 M3 Middle trigonid crest at OES  SH 93.0% (40/43) 100.0% (14/14) 100.0% (11/11) 83.3% (15/18)  H. neanderthalensis 97.0% (33/34) 100.0% (16/16) 100.0% (8/8) 90.0% (9/10)  H. sapiens 39.2% (11/28) 50.0% (6/12) 25.0% (3/12) 50.0% (2/4) Distal trigonid crest at OES  SH 25.5% (11/43) 0.0% 0.0% 61.1% (11/18)  H. neanderthalensis 20.5% (7/34) 6.2% (1/16) 12.5% (1/8) 50.0% (5/10)  H. sapiens 0.0% 0.0% 0.0% 0.0%

Chi-square test among SH, Homo neanderthalensis (NEA) and Homo sapiens (SAP) for each tooth position in order to analyze the differences in the trait frequencies (presence/absence) of MdTC and DTC at the OES.

Test du Chi2 parmi les spécimens de SH, Homo neanderthalensis (NEA) et Homo sapiens (SAP) pour chaque type de dent, afin d’analyser les différences entre les fréquences de trait (présence/absence) des MdTC et DTC au niveau de l’OES.

OES (presence/absence) MdTC DTC NEA SAP NEA SAP SH  M1 – 0.00 a 0.34 –  M2 – 0.00 a 0.22 –  M3 0.62 0.15 0.56 0.02 a  Total 0.42 0.00 a 0.60 0.00 a NEA  M1 0.00 a 0.37  M2 0.00 a 0.20  M3 0.09 0.07  Total 0.00 a 0.01 a

Data correspond to P-values (P < 0.05).

Typology and description of trigonid crest forms identified at the dentine surface of the SH sample.

Typologie et description des formes de crêtes du trigonide identifiées à la surface de la dentine de l’échantillon SH.

Type 1 Continuous MdTC with an absent or discontinuous DTC Type 2 Continuous MeTC with an absent or discontinuous DTC Type 3 Continuous DTC when at least one of the origins is the distal segment, with an absent or discontinuous MeTC or MdTC Type 4 Absent or discontinuous DTC when at least one of the origins is the distal segment, with an absent or discontinuous MeTC or MdTC. The number of distal ridges can be variable Type 5 Continuous MdTC and a continuous DTC arising from the middle segments Type 6 Continuous MdTC and a continuous DTC when at least one of the origins is the distal segment Type 7 Continuous MdTC and a discontinuous DTC when at least one of the origins is the distal segment Type 8 Continuous MdTC and a discontinuous DTC arising from the distal segments Type 9 Continuous MeTC and continuous DTC when at least one of the origins is the distal segment Type 10 Continuous MdTC, a discontinuous DTC when at least one of the origins is the distal segment, and a pronounced distal ridge on one side Type 11 Continuous MeTC or MdTC, a discontinuous DTC when at least one of the origins is the distal segment, and a discontinuous DTC originated from the distal segments Type 12 Continuous MdTC, a discontinuous DTC arising from the middle segments and discontinuous DTC arising from the distal segments Type 13 Continuous MeTC, a discontinuous DTC arising from the middle segments and discontinuous DTC arising from the distal segments Type 14 Continuous MeTC, a discontinuous DTC when at least one of the origins is the distal segment, and a pronounced distal ridge on one side

Frequencies and percentages of the continuous mesial (MeTC), middle (MdTC) and distal trigonid crest (DTC) at the EDJ (only one antimere included). We consider the trait is “present” when the crest is continuous. Otherwise we consider that the trait is “absent”.

Pourcentages et fréquences des crêtes du trigonide mésiales (MeTC), intermédiaires (MdTC) et distales (DTC) au niveau de la jonction émail–dentine (EDJ) (un seul antimère inclus).

Total molar sample M1 M2 M3 Mesial trigonid crest at EDJ  SH 20.9% (9/43) 0.0% 27.2% (3/11) 33.3% (6/18)  H. neanderthalensis 20.5% (7/34) 12.5% (2/16) 25.0% (2/8) 30.0% (3/10)  H. sapiens 7.14% (2/28) 8.3% (1/12) 0.0% 25.0% (1/4) Middle trigonid crest at EDJ  SH 72.0% (31/43) 100.0% (14/14) 72.7% (8/11) 50.0% (9/18)  H. neanderthalensis 76.4% (26/34) 87.5% (14/16) 75.0% (6/8) 60.0% (6/10)  H. sapiens 39.2% (11/28) 58.3% (7/12) 25.0% (3/12) 25.0% (1/4) Distal trigonid crest at EDJ  SH 23.2% (10/43) 0.0% 0.0% 55.5% (10/18)  H. neanderthalensis 14.7% (5/34) 6.2% (1/16) 12.5% (1/8) 30.0% (3/10)  H. sapiens 0.0% 0.0% 0.0% 0.0%

Chi-square test among SH, Homo neanderthalensis (NEA) and Homo sapiens (SAP) for each tooth position in order to analyze the differences in the trait presence (continuous/discontinuous/absence) of MeTC, MdTC and DTC at the EDJ.

Test du Chi2 parmi les spécimens de SH, Homo neanderthalensis (NEA) et Homo sapiens (SAP) pour chaque type de dent afin d’analyser les différences dans la présence de trait (continu/discontinu/absence) des MeTC, MdTC et DTC au niveau de l’EDJ.

EDJ (continuous/discontinuous/absence) MeTC MdTC DTC NEA SAP NEA SAP NEA SAP SH  M1 0.17 0.27 0.17 0.00 a 0.38 0.09  M2 0.91 0.10 0.91 0.03 a 0.22 0.01 a  M3 0.85 0.74 0.61 0.08 0.07 0.00 a  Total 0.97 0.14 0.66 0.00 a 0.08 0.00 a NEA  M1 0.72 0.07 0.49  M2 0.14 0.05 0.06  M3 0.85 0.19 0.12  Total 0.19 0.00 a 0.02 a

Data correspond to P-values (P < 0.05).

Variability of the TC types found at the EDJ and their correspondence to the OES from all the studied groups.

Variabilité des types TC trouvés à la EDJ et leur correspondance au niveau de l’OES de tous les groupes étudiés.

Enamel (OES) Dentine (EDJ) Type A 1 - 2 - 4 - 7 - 8 - 10 - 11 - 12 - 13 - 14 Type B 3 Type C 2 - 3 - 5 - 6 - 9 - 14 Type D 2 - 4 - 7 - 11

Frequency of expression of trigonid crests and their percentages on both enamel and dentine for the M1 permanent molars analyzed of Homo heidelbergensis, Homo neanderthalensis and Homo sapiens.

Fréquence d’expression des crêtes du trigonide et pourcentages associés sur l’émail et la dentine pour les molaires permanentes M1 analysées issues de Homo heidelbergensis, Homo neanderthalensis et Homo sapiens.

SH Homo neanderthalensis Homo sapiens OES types n/% EDJ types n/% OES types n/% EDJ types n/% OES types n/% EDJ types n/% A 14 (100%) 1 1 (7.14%) A 15 (93.75%) 1 0 A 6 (50%) 1 1 (8.33%) B 0 2 0 B 0 2 0 B 0 2 1 (8.33%) C 0 3 0 C 1 (6.25%) 3 0 C 0 3 0 D 0 4 0 D 0 4 0 D 6 (50%) 4 4 (33.33%) Total 14 5 0 Total 16 5 0 Total 12 5 0 6 0 6 1 (6.25%) 6 0 7 2 (14.28%) 7 1 (6.25%) 7 1 (8.33%) 8 4 (28.57%) 8 1 (6.25%) 8 2 (16.66%) 9 0 9 0 9 0 10 3 (21.42%) 10 5 (31.25%) 10 1 (8.33%) 11 3 (21.42%) 11 3 (18.75%) 11 1 (8.33%) 12 1 (7.14%) 12 3 (18.75%) 12 1 (8.33%) 13 0 13 0 13 0 14 0 14 2 (12.50%) 14 0 Total 14 Total 16 Total 12

Frequency of expression of trigonid crests and their percentages on both enamel and dentine for the M2 permanent molars analyzed of Homo heidelbergensis, Homo neanderthalensis and Homo sapiens.

Fréquence d’expression et pourcentages associés des crêtes du trigonide présentes sur l’émail et la dentine pour les molaires permanentes M2 analysées issues de H. heidelbergensis, H. neanderthalensis et H. sapiens.

SH Homo neanderthalensis Homo sapiens OES types n/% EDJ types n/% OES types n/% EDJ types n/% OES types n/% EDJ types n/% A 11 (100%) 1 2 (18.18%) A 7 (87.5%) 1 0 A 3 (25%) 1 0 B 0 2 0 B 0 2 0 B 0 2 0 C 0 3 0 C 1 (12.5%) 3 0 C 0 3 0 D 0 4 0 D 0 4 0 D 9 (75%) 4 9 (75%) Total 11 5 0 Total 8 5 0 Total 12 5 0 6 0 6 1 (12.50%) 6 0 7 0 7 2 (25%) 7 0 8 0 8 1 (12.50%) 8 2 (16.66%) 9 0 9 0 9 0 10 1 (9.09%) 10 2 (25%) 10 1 (8.33%) 11 3 (27.27%) 11 0 11 0 12 2 (18.18%) 12 0 12 0 13 3 (27.27%) 13 0 13 0 14 0 14 2 (25%) 14 0 Total 11 Total 8 Total 12

Frequency of expression of trigonid crests and their percentages on both enamel and dentine for the M3 permanent molars analyzed of Homo heidelbergensis, Homo neanderthalensis and Homo sapiens.

Fréquence d’expression et pourcentages associés des crêtes du trigonide présentes sur l’émail et la dentine pour les molaires permanentes M3 analysées issues de Homo heidelbergensis, Homo neanderthalensis et Homo sapiens.

SH Homo neanderthalensis Homo sapiens OES types n/% EDJ types n/% OES types n/% EDJ types n/% OES types n/% EDJ types n/% A 6 (33.33%) 1 0 A 4 (40%) 1 0 A 2 (50%) 1 1 (25%) B 2 (11.11%) 2 1 (5.55%) B 0 2 3 (30%) B 0 2 1 (25%) C 9 (50%) 3 2 (11.11%) C 5 (50%) 3 0 C 0 3 0 D 1 (5.56%) 4 1 (5.55%) D 1 (10%) 4 1 (10%) D 2 (50%) 4 2 (50%) Total 18 5 3 (16.66%) Total 10 5 1 (10%) Total 4 5 0 6 3 (16.66%) 6 2 (20%) 6 0 7 0 7 0 7 0 8 1 (5.55%) 8 1 (10%) 8 0 9 2 (11.11%) 9 0 9 0 10 1 (5.55%) 10 1 (10%) 10 0 11 1 (5.55%) 11 1 (10%) 11 0 12 0 12 0 12 0 13 2 (11.11%) 13 0 13 0 14 1 (5.55%) 14 0 14 0 Total 18 Total 10 Total 4

Comparison between the data of our samples and those from Bailey et al. (2011) regarding continuous or discontinuous MdTC.

Comparaison entre les données issues de nos échantillons et celles présentées dans Bailey et al. (2011) en ce qui concerne les MdTC continues ou discontinues.

This study Bailey's data SH sample H. neanderthalensis H. sapiens H. neanderthalensis H. sapiens n % n % n % n % n % Enamel (OES)  MdTC continuous (grade 2-3, Bailey et al., 2011) 40 93.02 33 97.05 11 39.28 61 100 8 11.76  MdTC discontinuous (grade 1, Bailey et al., 2011) 3 6.97 1 2.94 3 10.71 – – 13 19.11 Dentine (EDJ)  MdTC continuous (grade 2-3, Bailey et al., 2011) 40 93.02 33 97.05 13 46.42 72 98.63 24 35.29  MdTC discontinuous (grade 1, Bailey et al., 2011) 3 6.97 1 2.94 6 21.42 1 1.36 8 11.76