Journal of Human Evolution 82 (2015) 34e50

Contents lists available at ScienceDirect

Journal of Human Evolution

journal homepage: www.elsevier.com/locate/jhevol

A geometric morphometric analysis of hominin lower molars:

Evolutionary implications and overview of postcanine dental variation
Aida Go
mez-Robles a, *, Jose

María Bermúdez de Castro b, María Martino

n-Torres b,
Leyre Prado-Simo
n c, Juan Luis Arsuaga d
a
Department of Anthropology, The George Washington University, 800 22nd St. NW, Washington, DC 20052, USA
b
n sobre la Evolucio
Centro Nacional de Investigacio
n Humana, Paseo de la Sierra de Atapuerca S/N, 09002 Burgos, Spain
c
Stomatology Department, Dentistry Faculty, University of Granada, Spain
d
Centro Mixto UCM-ISCIII de Evolucio
n y Comportamiento Humanos, Avd. Monforte de Lemos 5, Pabello
n 14, 28029 Madrid, Spain

a r t i c l e i n f o a b s t r a c t

Article history: Lower molars have been extensively studied in the context of hominin evolution using classic and
Received 3 March 2014 geometric morphometric analyses, 2D and 3D approaches, evaluations of the external (outer enamel
Accepted 17 February 2015 surface) and internal anatomy (dentine, pulp chamber, and radicular canals), and studies of the crown
Available online 1 April 2015
and root variation. In this study, we present a 2D geometric morphometric analysis of the crown anatomy
of lower first, second, and third molars of a broad sample of hominins, including Pliocene and Lower,
Keywords:
Middle, and Upper Pleistocene species coming from Africa, Asia, and Europe. We show that shape
Atapuerca
variability increases from first to second and third molars. While first molars tend to retain a relatively
European Middle Pleistocene
Generalized Procrustes analysis
stable 5-cusped conformation throughout the hominin fossil record, second and third molars show
Dental anthropology marked distal reductions in later Homo species. This trend to distal reduction is similar to that observed
Homo antecesor in previous studies of premolars and upper second and third molars, and points to a correlated reduction
Sima de los Huesos of distal areas across the whole postcanine dentition. Results on lower molar variation, as well as on
Homo heidelbergensis other postcanine teeth, show certain trends in European Pleistocene populations from the Atapuerca
Neanderthals sites. Middle Pleistocene hominins from Sima de los Huesos show Neanderthal affinities and strong
dental reduction, especially in the most distal molars. The degree of dental reduction in this population is
stronger than that observed in classic Neanderthals. Homo antecessor hominins from Gran Dolina-TD6
have primitive lower teeth that contrast with their more derived upper teeth. The evolutionary impli-
cations of these dental affinities are discussed in light of recent paleogenetic studies.
© 2015 Elsevier Ltd. All rights reserved.

Introduction generally accepted utility of dental traits in the classification of

The central role of dental anthropology in the broader frame-
work of paleoanthropology is demonstrated by the profusion of
articles aiming to address taxonomic and phylogenetic questions
via dental morphometry. Dental evolution is subject to several in-
fluences, ranging from developmental constraints related to the
serially homologous nature of the dentition to functional con-
straints related to occlusion (Go
mez-Robles and Polly, 2012).
Nevertheless, a significant phylogenetic signal remains in dental
morphology (Caumul and Polly, 2005), and this is the basis of the
* Corresponding author.

mez-Robles).
E-mail address: aidagomez2@yahoo.es (A. Go
specimens and species, as well as in describing possible evolu-
tionary scenarios (Suwa et al., 1994, 1996; Bailey, 2002a, 2004;
Bailey and Lynch, 2005; Guatelli-Steinberg and Irish, 2005; Kaifu
et al., 2005; Moggi-Cecchi et al., 2006; Martino
n-Torres et al.,
2007a, b; Moggi-Cecchi and Boccone, 2007; Bailey et al., 2009;

mez-Robles et al., 2013). Among studies
Benazzi et al., 2011b; Go
evaluating dental variation, analyses of lower molar morphology
are especially common in the literature, and they range from classic
morphometric analyses (e.g., Wood and Abbott, 1983; Wood et al.,
1983; Bermúdez de Castro and Nicola
s, 1995) to quantitative
studies of form and shape variation (Benazzi et al., 2011a), some-
times based on 3D reconstructions (Skinner et al., 2008). More
recently, some microCT-based studies have focused on the root
morphology of mandibular molars (Kupczik and Hublin, 2010;
Emonet et al., 2012).

http://dx.doi.org/10.1016/j.jhevol.2015.02.013
0047-2484/© 2015 Elsevier Ltd. All rights reserved.

mez-Robles et al. / Journal of Human Evolution 82 (2015) 34e50
A. Go 35

Given the relevance of dental morphology in understanding Nevertheless, a systematic evaluation of hominin lower molars in
hominin evolution, we initiated in 2006 a systematic evaluation of the context of our 2D geometric morphometric series of analyses is
postcanine dental variation in crown morphology (Martino
n-Torres still lacking (Martino
n-Torres et al., 2006; Go
mez-Robles et al.,
et al., 2006). Our evaluations have included a comprehensive rep- 2007, 2008, 2011b). Therefore, the main objective of this study is
resentation of the hominin fossil record, but they have paid special to complete the geometric morphometric analysis of the post-
attention European Pleistocene groups. Our series of papers has canine dentition in order to investigate a more comprehensive
provided evidence that several dental traits typify Middle and scenario of hominin dental evolution during the Plio-Pleistocene.
Upper Pleistocene European groups. These features include a As in previous cases, our analysis is focused on the evaluation of
change towards symmetry and distal reduction in upper premolars the European fossil record and the relationship of European fossils
(Go
mez-Robles et al., 2011b); a very characteristic skewed with modern humans, although Asian and African specimens are
morphology in upper first molars that can be traced back to the late also included to place European variation into a broader context. As
Lower Pleistocene (Go
mez-Robles et al., 2007, 2011a; see also the final article in our series of papers, we discuss the implications
Bailey, 2004); a strong reduction of distal cusps in upper second of dental variation across the whole poscanine dentition, with
and, especially, third molars (Go
mez-Robles et al., 2012); an special emphasis on the evolutionary relationships of the Lower
acquisition of nearly symmetric shape and a strong reduction of the and Middle Pleistocene fossil samples from the Atapuerca sites of
talonid in lower first premolars (Go
mez-Robles et al., 2008); and a Gran Dolina-TD6 and Sima de los Huesos.
general maintenance of an asymmetric conformation in lower
second premolars (Martino
n-Torres et al., 2006). Dental traits Material and methods
related primarily to dental reduction are shared in general terms by
Homo sapiens and Homo neanderthalensis (also referred to as Material
modern humans and Neanderthals throughout the text, respec-
tively), although the specific features of these reductions differ in A sample of 132 lower first molars (M1), 126 lower second
both species. These features are the ones that allow for the differ- molars (M2), and 108 lower third molars (M3) was analyzed
entiation of these two species (and sometimes of their ancestral (Tables 1e3). The study of fossil teeth by means of 2D geometric
populations) in spite of the general reduction that drives dental morphometrics based on image analysis ensured the availability of
evolution in both cases.
Dental morphology shows a wide range of variation within Table 1
List of lower first molars included in this study.a
species, and its analysis across the closely related species that
constitute the hominin clade does not allow for the identification of Species n Specimens
unmistakably delineated species-specific morphologies. However, A. afarensis n¼6 AL128-23, AL145-35, AL266-1, AL288, AL333-
several trends can be identified and linked to particular population 1a, LH2
dynamics either via random factors such as genetic drift (Weaver A. africanus n ¼ 4 MLD2, STS24, STW498, Taung
P. robustus n ¼ 4 SK23, SKW5, SKX4446, TM1517
et al., 2007), or specific direct or indirect selective pressures
P. boisei n ¼ 2 KNM-ER3230, KNM-ER15930
(Grine, 1986; Bermúdez de Castro, 1989; Macho and Moggi-Cecchi, H. habilis s. l. n ¼ 6 KNM-ER1802, KNM-ER5431, OH7, OH13, OH16,
1992; Bermúdez de Castro and Nicola
s, 1995; Bermúdez de Castro STW151b
et al., 2003b; Irish and Guatelli-Steinberg, 2003; Lozano et al., H. ergaster n¼3 KNM-ER820, OH22, KNM-WT15000
2008). Regardless of the origin of morphological diversification, it is Dmanisi n¼2 D211, D2735
H. erectus n ¼ 11 Zhoukoudian: B3.9, 36, 137
evident in most studies of dental morphology that the ranges of Sangiran: S1b, S6, S7-20, S7-42, S7-43, S7-61,
variation of different species overlap, although they can be distin- S7-62, S7-76
guished in their mean values/morphologies. This causes taxonomic H. antecessor n¼3 Atapuerca-TD6: ATD6-5, ATD6-93, ATD6-96
classification based on dental morphology to be a probabilistic task European n ¼ 22 Arago: A13, A40, A89
H. heidelbergensis Atapuerca-SH: AT-2 (Ind 2), AT-101 (Ind 3), AT-
that will give rise to the allocation of specimens to species with
1759 (Ind 6), AT-141 (Ind 10), AT-286 (Ind 11),
certain probability (Bailey et al., 2009). AT-300 (Ind 12), AT-2276 (Ind 14), AT-2779 (Ind
Our previous studies of upper and lower premolars, as well as 16), AT-829 (Ind 18), AT-576 (Ind 19), AT-4318
upper molars, have revealed not only particular morphological pat- (Ind 20), AT-607 (Ind 23), AT-1458 (Ind 24), AT-
terns typical of European groups (Homo antecessor from Gran Dolina- 3934 (Ind 25), AT-561 (Ind 26), AT-792 (Ind 27),
AT-950 (Ind 31)
TD6, European Homo heidelbergensis, and H. neanderthalensis) but Montmaurin
also a markedly derived character state of the Middle Pleistocene Pontnewydd 11
population from Atapuerca-Sima de los Huesos (SH) compared with H. neanderthalensis n ¼ 22 Arcy-sur-Cure (Grotte de l'Hye ne): 3
other European Middle Pleistocene groups and classic Neanderthals Arcy-sur-Cure (Grotte du Renne): 35

mez-Robles et al., 2012; Martino
n-Torres et al., 2012, 2013). The Ehringsdorf
(Go
Hortus: H1262 (Ind II), H988 (Ind V), H IV
derived state of hominins from the SH collection is especially Krapina: B, C, D, E, G, J, 79 (Ind L), 84 (Ind N), 81
apparent in the most proximal premolars and the most distal molars. (Ind P), 80, 105
These morphological peculiarities are especially puzzling in light of Le Moustier 1
the recent DNA analysis of one SH hominin (Meyer et al., 2014), Malarnaud
Petit-Puymoyen (two specimens)
which has revealed a higher genetic mitochondrial affinity with Saint Ce
saire
Denisovans (Reich et al., 2010; Meyer et al., 2012) than with Nean- Fossil H. sapiens n ¼ 10 Qafzeh: 4, H4
derthals. In terms of dental morphology, the scarce morphological Abri Pataud
information on Denisovans shows that they are morphologically Isturitz: 106
Les Rois (three specimens including Les Rois A)
different from both Neanderthals and modern humans, with metric
Saint Germain-La Rivie re: B3, B4, B5
and shape features similar to those observed in Homo habilis and Recent H. sapiens n ¼ 37 Heidenheim (AMNH): 15 individuals
Homo erectus (Reich et al., 2010). La Torrecilla (LPA-UGR): 22 individuals
We have previously evaluated lower second molar morphology a
Ind: individual; AMNH: American Museum of Natural History (New York, USA);
in the context of methodological approaches useful to deal with LPA-UGR: Laboratory of Physical Anthropology, University of Granada (Granada,
evolutionary novelties and losses (Go
mez-Robles et al., 2011c). Spain).
36
mez-Robles et al. / Journal of Human Evolution 82 (2015) 34e50
A. Go

these relatively large samples. The analysis of crown morphology in Table 3

these dental samples was based on occlusal photographs captured List of lower third molars included in this study.a

with a Nikon D80 or with a Nikon D1H digital camera fitted with an Species n Specimens
AF Micro Nikkor 105 mm, f/2.8D lens. The occlusal surface of molars A. afarensis n¼7 AL188-1, AL266-1, AL288, AL333w-57, AL333-60,
was placed as parallel to the lens as possible in order to minimize AL400-1a, LH4
the error due to the orientation of molars in 2D projections A. africanus n ¼ 8 MLD4, MLD18, MLD19, STS52, STW3, STW14,
(Gharaibeh, 2005), which has been calculated to be approximately STW384, STW498
P. robustus n ¼ 3 SK23, SKW5, TM1517
3% of the distance between individual landmarks and the centroid
P. boisei n ¼ 4 KNM-ER729, KNM-ER3230, KNM-ER15930, Peninj:
of the landmark configuration in lower premolars (Go
mez-Robles 1
et al., 2008). This value is similar to the 1e4.5% error rate re- H. habilis s. l. n ¼ 3 OH4, OH13, OH16
ported by Bailey et al. (2004) in measuring cusp base areas in upper H. ergaster n ¼ 2 KNM-BK67, KNM-ER992
Dmanisi n ¼ 1 D211
first molars, with comparable values in intra- and inter-observer
H. erectus n ¼ 5 Zhoukoudian: A2, G1.6, 52, 131
evaluations. Error rates in placing landmarks are lower in molars Sangiran: S1b
than in premolars because their flatter occlusal surfaces make their H. antecessor n ¼ 2 Atapuerca-TD6: ATD6-96, ATD6-113
correct orientation less problematic (Go
mez-Robles, 2010). Taxo- European n ¼ 23 Arago: A13, A89
nomic assignments followed the same guidelines laid out in pre- H. heidelbergensis Atapuerca-SH: AT-1 (Ind 1), AT-811 (Ind 4), AT-13
(Ind 7), AT-4147 (Ind 12), AT-222 (Ind 16), AT-2277
vious articles of this series and are shown in Tables 1e3.
(Ind 18), AT-607 (Ind 23), AT-2385 (Ind 24), AT-
3943 (Ind 25), AT-30 (Ind 26), AT-143, AT-599, AT-
942, AT-1468, AT-1473, AT-1945, AT-2273, AT-
Methods
2777, AT-3182
Montmaurin
Dental morphology was studied by means of a set of landmarks Pontnewydd: 2, 21
located at cusp tips and groove intersections, and a set of equidis- H. neanderthalensis n ¼ 14 Arcy-sur-Cure (Grotte de l'Hye ne): 5
tant sliding semilandmarks located at the molar periphery (Fig. 1). Breuil: 3
Guattari: 2, 3
The employed landmarks were: 1) mesial or anterior fovea, or most
Hortus: H13 (Ind VI), H695 (IndV)
mesial extreme of the central groove; 2) intersection between the Krapina: 106 (Ind E), H, 7 (Ind L), 9
central groove and the mesiobuccal groove; 3) intersection be- Petit-Puymoyen
tween the central groove and the lingual groove; 4) intersection Saint Ce
saire
Shanidar: 2
Vindija: 206
Table 2 Fossil H. sapiens n ¼ 6 Qafzeh: H4
List of lower second molars included in this study.a Abri Pataud
Isturitz: 107
Species n Specimens Les Rois (three specimens)
Recent H. sapiens n ¼ 30 Heidenheim (AMNH): 10 individuals
A. afarensis n¼9 AL128-23, AL145-35, AL188-1, AL207-13, AL266-1,
La Torrecilla (LPA-UGR): 14 individuals
AL288, AL333-1b, AL333W-60, AL400-1a
CENIEH: 6 individuals
A. africanus n ¼ 4 MLD2, STS52, STW14, STW498
a
P. robustus n ¼ 4 SK23, SKW5, SKX4446, TM1517 Ind: individual; AMNH: American Museum of Natural History (New York, USA);
P. boisei n ¼ 3 L427-7, KNM-ER15930, KNM-ER3230 LPA-UGR: Laboratory of Physical Anthropology, University of Granada (Granada,
H. habilis s. l. n ¼ 5 KNM-ER1802, KNM-ER5431, OH7, OH13, OH16 Spain); CENIEH: National Research Centre for Human Evolution (Burgos, Spain).
H. ergaster n ¼ 3 KNM-ER992, OH22, KNM-WT15000
Dmanisi n ¼ 2 D211, D2735
H. erectus n ¼ 11 Zhoukoudian: A2, F1.5, G1.6, K1.96, 43, 44, 45 between the central groove and the distobuccal groove, also cor-
Sangiran: S1b, S5, S7-64, S7-65 responding to the most anterior point of the fifth cusp or hypo-
H. antecessor n¼3 Atapuerca-TD6: ATD6-5, ATD6-96, ATD6-113
conulid; 5) tip of the mesiobuccal cusp or protoconid; 6) tip of the
European n ¼ 24 Arago: A10, A13, A69, A89
H. heidelbergensis Atapuerca-SH: AT-1 (Ind 1), AT-3179 (Ind 2), AT- mesiolingual cusp or metaconid; 7) tip of the distobuccal cusp or
271 (Ind 3), AT-169 (Ind 10), AT-557 (Ind 11), AT- hypoconid; and 8) tip of the distolingual cusp or entoconid.
300 (Ind 12), AT-284 (Ind 14), AT-2193 (Ind 15), AT- The evaluation of first molars included the eight landmarks,
2763 (Ind 16), AT-941 (Ind 18), AT-946 (Ind 20), AT- whereas the evaluation of second molars did not include landmark
607 (Ind 23), AT-2396 (Ind 24), AT-3889 (Ind 25),
4 and the analysis of third molars excluded landmarks 1 and 4 due
AT-1756 (Ind 26), AT-3176 (Ind 27), AT-792 (Ind 28)
Montmaurin to the difficulty of identifying them in the most recent hominin
Pontnewydd: 8, 15 species because of their strongly reduced morphology. The location
H. neanderthalensis n ¼ 18 Arcy-sur-Cure (Grotte de l'Hye ne): 4 of landmarks, especially those located on cusp tips, was assessed by
Arcy-sur-Cure (Grotte du Renne): 5, 21
comparing photographs with the original molars or casts. Heavily
Ehringsdorf
Hortus: H IV, H930 (Ind V) worn molars were not included in our analyses, although a certain
Krapina: 6 (Ind A), C, D, 10 (Ind E), G, J, 1 (Ind L), 86, degree of wear was accepted (up to wear stage 3e4; Molnar, 1971;
107 Supplementary Online Material [SOM] Fig. 1). Apart from these
Le Moustier 1 landmarks, we digitized 39 sliding semilandmarks in counter-
Petit-Puymoyen

saire
clockwise direction. The first semilandmark was located at the
Saint Ce
Fossil H. sapiens n¼8 Qafzeh: H4, 11 point of the mesial margin directly opposite to the mesial fovea, or
Abri Pataud at the point of the mesial margin intersected by the hypothetical
Grimaldi prolongation of the central groove. Landmarks and semilandmarks
Les Rois (three specimens)
were digitized using TpsDig2 (Rohlf, 2005). The 39 semilandmarks
Saint Germain-La Rivie re: 3
Recent H. sapiens n ¼ 32 Heidenheim (AMNH): 13 individuals
were slid in order to minimize the Procrustes distance between
La Torrecilla (LPA-UGR): 19 individuals each individual and the target represented by the consensus
a morphology of the sample (Pe
rez et al., 2006). We used this crite-
Ind: individual; AMNH: American Museum of Natural History (New York, USA);
LPA-UGR: Laboratory of Physical Anthropology, University of Granada (Granada, rion to slide semilandmarks instead of other possible criteria (Gunz
Spain). et al., 2005) because previous studies of hominin lower molars have

mez-Robles et al. / Journal of Human Evolution 82 (2015) 34e50
A. Go
Figure 1. Landmarks and semilandmarks used in the lower first molar analysis.
Landmark 1 has not been included in the lower second molar analysis. Landmarks 1
and 4 have not been included in the lower third molar analysis. B: buccal; L: lingual;
M: mesial; D: distal.
demonstrated a better performance of this approach in a similar
context (Go
mez-Robles et al., 2011c).
After performing a generalized Procrustes analysis (GPA) in or-
der to remove all non-shape variation corresponding to position,
size, and orientation (Rohlf and Slice, 1990), several analyses were
carried out to assess the main morphological trends observed in the
hominin clade. A first exploratory step was carried out using a
relative warps analysis (Bookstein, 1991), which in this case is
equivalent to a principal components analyses (PCA) of Procrustes-
superimposed landmark coordinates. The PCAs were carried out on
complete samples and also on reduced samples including only
Homo species and grouping Dmanisi hominins, Homo ergaster, and
H. erectus under H. erectus s. l. The purpose of repeating analyses on
reduced samples was to ascertain if important morphological fea-
tures distinguishing Homo species might be masked by including
Australopithecus and Paranthropus species. A canonical variates
analysis (CVA), together with a comparison of species mean shapes
based on permutation tests, was employed in order to evaluate the
performance of lower molar morphology in taxonomic classifica-
tion and the general morphological distinctiveness of the three
most recent (and best represented) hominin species:
H. heidelbergensis, H. neanderthalensis, and H. sapiens. The CVAs
used the first ten principal components obtained in a PCA of these
species (based in turn on GPAs including only these three species).
This number of principal components was chosen to ensure that
the number of variables was below the number of individuals per
group in all cases. Percentages of correct classification were cross-
validated by excluding each specimen from the analysis and later
testing the group to which they are assigned. Technical details of
these analyses are explained in previous articles in this series
(Go
mez-Robles et al., 2007, 2008, 2011b, 2012).
The relative contribution of allometric change to shape variation
was evaluated by means of a multivariate regression of shape var-
iables on size. Size was measured as the logarithm of the centroid
size, defined as the squared root of the sum of the squared distances
between a set of landmarks and their centroid or gravity center.
37
Allometric change was evaluated both intraspecifically (static
allometry) in H. heidelbergensis, H. neanderthalensis, and H. sapiens,
and interspecifically (evolutionary allometry). For the latter,
phylogenetic non-independence of data was accounted for by
regressing independent contrasts of shape on independent con-
trasts of size (Klingenberg and Marug
an-Lobo
n, 2013). Independent
contrasts are calculated as weighted differences between pairs of
sister species or nodes, and they were based on two different
phylogenetic trees to explore the effect that the uncertainty in
defining phylogenetic relationships has on these estimates. The
first phylogenetic tree included Paranthropus as a monophyletic
group and used a model of late divergence between Neanderthals
and modern humans, whereas the second phylogenetic tree
considered Paranthropus as paraphyletic and used a model of early
divergence between Neanderthals and modern humans (SOM
Fig. 2). Both trees also differed in the geometry of the relation-
ships between H. ergaster, H. erectus, and the Dmanisi hominins
(which were included as a separate species to increase the number
of data points). Many other modifications of these phylogenetic
trees are possible, but we restricted our analyses to these two
because the aim of this study is not to do an exhaustive comparison
of the effect of phylogenetic uncertainty in the estimation of allo-
metric effects. Analyses were carried out with TpsRelw (Rohlf,
1998) and MorphoJ (Klingenberg, 2011).
Results
General morphological trends
M1 The first principal component (PC) corresponding to the
morphological analysis of lower first molars reflects the reduction
of the fifth cusp, which is accompanied by a reduction of the
mesiodistal diameter giving rise to a more squared morphology, as
well as a change from the typical Dryopithecus pattern (also called
Y5 pattern) to a cruciform groove pattern (Fig. 2: top chart). The
second component shows mainly the distinction between
mesiodistally elongated molars (positive values) versus
buccolingually broad molars (negative values). The main
morphological change across the second PC affects the general
proportion of the molars (rectangular shape versus approximately
squared shape in positive and negative values, respectively), but
this PC does not correspond with strong changes either in cusp
proportions or in the groove pattern.
The most notable observation in the distribution of individuals
in the PC space corresponding to PC1 and PC2 is the presence of two
outliers (two H. sapiens molars with only four cusps). This 4-cusp
conformation is part of the normal variability of H. sapiens, even
though this pattern is a minority in the studied sample. Apart from
this, the PCA graph shows a complete overlapping of different
species, including Australopithecus, Paranthropus, and Homo species
(Fig. 2: top chart).
Lower first molars show a very homogeneous distribution of
variance such that no eigenvalue is clearly predominant versus the
other ones. Actually, PC1 and PC2 account only for 32.78% of total
morphological variance. Therefore, PC3 and PC4 have also been
evaluated, and their combination does reveal certain interspecific
organization (Fig. 2: bottom chart). The most important feature of
the PCA graph showing PC3 against PC4 is the separation of
H. neanderthalensis and H. heidelbergensis from other hominins due
to their exclusive positive values (with only one exception corre-
sponding to the Le Moustier 1 molar) on PC3. Positive scores on PC3
are associated with a general mesiodistally elongated shape, a well-
developed fifth cusp associated with a cruciform groove pattern,
and a mesial displacement of the anterior fovea caused by the
almost universal appearance in Neanderthal and pre-Neanderthal
38
mez-Robles et al. / Journal of Human Evolution 82 (2015) 34e50
A. Go

Figure 2. Principal components analysis (relative warps analysis) of lower first molar morphology in the complete hominin sample. Top: PC1 versus PC2. Bottom: PC3 versus PC4.
Orientation of TPS-grids is the same as in Figure 1. Relevant specimens discussed in the text are indicated.

lower molars of a very well-defined mid-trigonid crest (Bailey,
2002a; Martino
n-Torres, 2006; Bailey et al., 2011; Martino
n-
Torres et al., 2012; Martínez de Pinillos et al., 2014). Austral-
opithecus, Paranthropus, and early Homo molars tend to plot in the
negative halves of the graphs for both PC1 and PC3, without a clear
pattern of distribution along PC2 and PC4 (Fig. 2). Homo erectus also
shows mainly negative values for PC1 and PC3 that are indicative of
their general primitive morphology in lower first molars.
Homo antecessor shows morphologies close to the consensus of the
sample.
These patterns of variation are slightly different from the ones
yielded by PCAs including only species of the genus Homo, a result

mez-Robles et al. / Journal of Human Evolution 82 (2015) 34e50
A. Go 39

that is also driven by the exclusion from this reduced sample of the
H. sapiens outliers (Fig. 3). However, analyses of the complete and
reduced samples are similar in not separating species in the com-
bination of the first and second PCs, but in the combination of PC3
and PC4. When excluding Australopithecus and Paranthropus spe-
cies as well as the H. sapiens outliers, PC1 and PC2 are interchanged
such that the reduction of the hypoconulid is recovered by PC2.
M2 The first PC corresponding to the shape analysis of lower
second molars reveals not only the reduction of the fifth cusp
observed in the most recent hominin species but also a general
reduction of the distal half of the molars associated with the
appearance of an X-pattern (Fig. 4: top chart). On the contrary,
negative scores on PC1 correspond to mesiodistally long molars
with five well-developed cusps and the primitive Dryopithecus

Figure 3. Principal components analysis (relative warps analysis) of lower first molar morphology in the reduced hominin sample (excluding Australopithecus and Paranthropus
species, and H. sapiens outliers). Top: PC1 versus PC2. Bottom: PC3 versus PC4. Orientation of TPS-grids is the same as in Figure 1.
40
mez-Robles et al. / Journal of Human Evolution 82 (2015) 34e50
A. Go

pattern. Negative scores on PC2 correspond to buccolingually
broad molars with peripherally located cusp apices and a
cruciform pattern, whereas positive scores on this PC correspond
to more elongated molars along the mesioedistal axis with
more centrally located cusp apices and a mesially located
anterior fovea.
The first PC shows a gradual separation of earlier molars
belonging to Australopithecus, Paranthropus, and early Homo versus
recent molars belonging to later Homo species (Fig. 4: top chart).
The second PC shows primarily the separation between modern
humans and European groups (mainly SH hominins), whereas
earlier species occupy the whole morphospace along PC2. The three
molars belonging to H. antecessor plot close to one another at the
negative extreme of PC1, thus highlighting the well developed 5-
cusped morphology typical in lower molars of the oldest Euro-
pean populations, similar to those observed in earlier species of the
genus Homo and even in some Australopithecus and Paranthropus
specimens. The three H. ergaster molars have intermediate scores
for PC2, but are distributed along the complete morphospace for
PC1, with the only exception of the region of the plot corresponding

Figure 4. Principal components analyses (relative warps analyses) of lower second molar morphology (top) and lower third molar morphology (bottom) in the complete hominin
samples. Only the first two principal components are represented. Orientation of TPS-grids is the same as in Figure 1. Relevant specimens discussed in the text are indicated.

mez-Robles et al. / Journal of Human Evolution 82 (2015) 34e50
A. Go
to strongly reduced molars. Both lower second molars from the
Dmanisi site plot at the negative half of the graph for PC1, but they
differ along PC2.
For lower second molars, morphological variations recovered by
PC1 and PC2 in the analysis of the complete and reduced samples
are very similar (Fig. 5: top). The distribution of the different spe-
cies in the morphospace is also largely comparable in both analyses,
with European groups and H. sapiens separated along PC2, and
H. antecessor plotting at the negative extreme of PC1.
M3 Similar morphological patterns are observed in the analysis of
lower third molars. The first PC shows the variation from unre-
duced molars with five well-defined cusps to strongly reduced
molars without a hypoconulid (c5) and with a small hypoconid
and entoconid, which are sometimes obscured by the presence of
accessory cusps (Fig. 4: bottom chart). The second PC also reflects
the contrast between approximately squared molars with
peripherally located cusp tips versus mesiodistally elongated
molars with centrally compressed cusp apices. The exclusion in
these molars of the landmark corresponding to the mesial
foveaddue to the difficulty in identify it in strongly reduced
molarsddoes not allow for an evaluation of its position.
However, an a posteriori evaluation demonstrates that many
molars plotting on the positive half of the graph for PC2 show a
mid-trigonid crest (and an associated mesial displacement of the
anterior fossa).
In the PCA graph corresponding to M3 morphology, the most
negative scores correspond to P. boisei molars, which are generally
the most expanded ones (both in size and in cusp number). The
positive half of the graph for PC1, however, includes not only molars
belonging to the most recent species but also some belonging to
Australopithecus afarensis, Australopithecus africanus, H. habilis, H.
erectus, H. ergaster, and H. antecessor (Fig. 4: bottom chart).
Homo ergaster and Dmanisi molars plot close to the center of the
graph, whereas both H. antecessor third molars differ in their
morphology: ATD6-113 has a well-defined 5-cusped conformation
that makes it plot at the upper left quadrant (close to some
H. habilis molars). However, ATD6-96 has a strongly reduced
morphology that makes it plot at the lower right quadrant, in the
area mainly occupied by H. sapiens (although some Neanderthal,
Australopithecus, and H. erectus molars are also present in this area).
Removing Australopithecus and Paranthropus molars has a
noticeable effect in the PCA of the selected M3 sample (Fig. 5:
bottom chart). Although PC1 still recovers distal reduction, the
negative extreme values do not show the broadly expanded talonid
area typical of Paranthropus third molars (especially of P. boisei). The
second PC, however, shows a very similar distinction between
mesiodistally short or elongated molars, which is useful to distin-
guish H. sapiens from H. neanderthalensis and H. heidelbergensis
molars.
Discriminant ability of lower molar shape
The assignment tests based on canonical variates analyses of the
three most recent groups (H. heidelbergensis, H. neanderthalensis,
and H. sapiens) reveal that the lower first molar has the highest
discriminant ability from the three lower molars. Discriminant
ability decreases towards the second and third molars, as intra-
specific ranges of variation increase. The CVA graphs show how the
ranges of distribution of the three species are generally well
defined in lower first molars, whereas they are blurred in second
and third molars (Fig. 6). As the range of variation of the three
species converge towards distal teeth, Neanderthals lose their
morphological identity and their percentage of correct assignment
drops to approximately 60% in second and third molars (Table 4).
41
Percentages of correct assignment are highest in the three cases
for H. sapiens and lowest for H. neanderthalensis, with intermediate
values for H. heidelbergensis. This reflects in part the morphological
distinction of Neanderthal and SH molars due to the very anterior
position of the mesial fovea, as well as the more peripheral position
of cusp apices in H. sapiens (versus the more central position
observed in Neanderthals and, especially, in SH hominins).
Homo neanderthalensis and H. heidelbergensis molars are barely
differentiated in the position of the mesial fovea, but their main
differences are the mesiodistally more elongated shape and the
more central placement of cusp tips observed in H. heidelbergensis.
The observed relatively high percentages of correct classifica-
tion are indicative of significant differences in mean shape between
the three species in the three molar positions. Procrustes distances
between H. heidelbergensis and H. sapiens are greater than the
distances between Neanderthals and H. heidelbergensis and be-
tween Neanderthals and modern humans for the three molars
(Table 5). Neanderthals and modern humans are expected to show
the greatest shape differences because the phylogenetic distance
between both species (the amount of time during which they have
evolved independently) is longer that the phylogenetic distance
between either of them and H. heidelbergensis, regardless of the
phylogenetic position of the latter. Therefore, this result highlights
the morphological uniqueness of the H. heidelbergensis sample,
mostly represented by Sima de los Huesos hominins, which is
especially evident in third molars.
Allometry
Allometric variation shows different patterns when evaluated at
inter- and intraspecific levels (Table 6). Interspecifically, the three
molars show highly significant allometric variation when evaluated
without considering phylogenetic relationships. These analyses
show that allometric variation is stronger in second and third
molars than in first molars. These results are more comparable to

mez-Robles
the ones presented in the other papers in this series (Go
et al., 2007, 2012), and they show that allometric variation is
generally smaller in lower molars than in upper molars. However,
upper and lower first molars show low allometric variation, with
similar values in both cases. Phylogenetically corrected analyses
yield different results when using different phylogenetic trees, this
being the main limitation of this approach in the context of hom-
inin evolution. Lower first molars may or may not show significant
evolutionary allometry depending on the employed phylogenetic
tree, whereas second molars show significant allometry in both
cases and third molars do not show significant allometry in either
case (Table 6). It is important to note, however, that these results
are susceptible to change if a different topology of evolutionary
relationships across species (including different datings for species
and nodes) were proved to be more accurate. It has to be noted as
well that these results are strongly influenced by evolutionary re-
lationships among H. habilis, H. ergaster, Dmanisi hominins, and
H. erectus, which are far from resolved (e.g., Lordkipanidze et al.,
2013).
Other authors have addressed this problem by restricting allo-
metric analyses to closely related species, i.e., Neanderthals and
modern humans (Bailey et al., 2014). Similarly, we also evaluated
intraspecific allometric variation to avoid the potential confound-
ing effect of interspecific differences and phylogenetic uncertainty.
Intraspecific allometric change is significant or marginally signifi-
cant for the three molars in H. heidelbergensis, an effect that is
partially influenced by the inclusion in these samples of larger and
morphologically less reduced molars from Arago, together with
smaller and reduced molars from SH. Similarly, allometric effects
are significant for H. sapiens lower first molars, and they are also
42
mez-Robles et al. / Journal of Human Evolution 82 (2015) 34e50
A. Go

Figure 5. Principal components analyses (relative warps analyses) of lower second molar morphology (top) and lower third molar morphology (bottom) in the reduced samples
(excluding Australopithecus and Paranthropus). Only the first two principal components are represented. Orientation of TPS-grids is the same as in Figure 1.

influenced by the inclusion in this sample of fossil and recent shortening of the mesiodistal diameter caused by the reduction of
H. sapiens, the former being larger and less reduced. The most talonid cusps, especially of the hypoconulid. This reduction is
important morphological change associated with size variation in associated in second and third molars with the frequent observa-
the three molars at both inter- and intraspecific levels is a tion of þ and X groove patterns.

mez-Robles et al. / Journal of Human Evolution 82 (2015) 34e50
A. Go
Figure 6. Canonical variates analysis of lower first, second, and third molars (repre-
sented from top to bottom). TPS-grids represent the deformation from the mean shape
of the sample to the mean shape of H. heidelbergensis (bottom on each plot),
H. neanderthalensis (top on each plot), and H. sapiens (right on each plot). Orientation
43
Discussion
M1
Some particular morphological traits, such as certain additional
cusps, crests, or foveas (Irish, 1998; Bailey, 2002b; Guatelli-
Steinberg and Irish, 2005), and general size (Wood and Abbott,
1983; Suwa et al., 1994; Bermúdez de Castro and Nicol
as, 1995)
can help to distinguish lower first molars of different species.
However, general morphology (measured as general proportions of
molars, number of main cusps, relative area of these cusps, and
organization of main grooves) shows relatively low variation
throughout the hominin fossil record in these molars (Fig. 7).
Several studies trying to evaluate morphological differences in
lower molars of robust and gracile australopiths have observed
certain trends, but not a clear separation of groups (Wood et al.,
1983; Wood and Abbott, 1983; Suwa et al., 1994, 1996; Skinner
et al., 2008). The similarity of lower first molars of Neanderthals
and modern humans has been previously highlighted, with only
some exceptions, the most important being the very frequently
observed mid-trigonid crest in Neanderthals (Bailey, 2002a, b). This
trait is also expressed with a very high frequency in the SH popu-
lation (Martino
n-Torres, 2006; Martino
n-Torres et al., 2007b, 2012).
Quantitative analyses of the cervical and crown outline of Nean-
derthal and modern human first molars have also demonstrated
that lower first molars yield a less clear separation of both groups as
compared with upper first molars (Benazzi et al., 2011a). It has been
suggested, however, that certain crown outline features can be
useful to differentiate Neanderthal and modern human M1s once
size information is regressed out, including a reduction of the
buccodistal outline and a straighter lingual side in modern humans
(Benazzi et al., 2011a).
This relative stability in lower first molar morphology is also
reflected in the homogeneous distribution of variance across
principal components (Go
mez-Robles and Polly, 2012). Actually, no
separation of species or genera is observed in the combination of
the two principal components that account for the highest per-
centage of variance. Different species overlap along PC1 and PC2,
where individual distribution does not correspond to a taxonomic
differentiation. The third principal component, however, does show
interspecific separation. This component mainly reflects the sepa-
ration of H. heidelbergensis and H. neanderthalensis molars due to
the appearance of þ and X groove patterns. Non Y5-patterns appear
to be more common in the SH population (44.4%; Martino
n-Torres
et al., 2012) than in Neanderthals (3% in Bailey, 2002b; and 17.9% in
Martino
n-Torres et al., 2012) or in modern humans (0e20% in Scott
and Turner, 1997; 0e12.5% in Bailey, 2002b; and 22.7e27.6% in
Martino
n-Torres et al., 2012). The second trait that causes a sepa-
ration between Neanderthal and H. heidelbergensis molars and the
other species is the relative displacement of the mesial fovea to-
wards the mesial margin of the molars due to the frequent presence
of the mid-trigonid crest (Bailey, 2002a, b; Martino
n-Torres, 2006;
Martino
n-Torres et al., 2007b, 2012; Bailey et al., 2011; Martínez de
Pinillos et al., 2014). This crest is observed not only in almost all the
Neanderthals and H. heidelbergensis, but also in some molars from
Sangiran (H. erectus), Dmanisi, and Gran Dolina-TD6 (H. antecessor;
Martino
n-Torres et al., 2007a). In some of these cases, however, the
mid-trigonid crest is interrupted by the central groove, the course
of which is attenuated at the point of intersection with the crest.
Arago molars show the traits observed in SH molars to a lesser
of TPS-grids is the same as in Figure 1. Fossil and recent H. sapiens are represented with
different symbols on the plot, but they have been grouped together in canonical var-
iates analyses.
44
mez-Robles et al. / Journal of Human Evolution 82 (2015) 34e50
A. Go

Table 4
Percentage of correct classifications based on canonical variates analyses.

M1 M2 M3

Original Cross-validated Original Cross-validated Original Cross-validated

H. heidelbergensis 95.5% 81.8% 83.3% 75.0% 79.2% 70.8%

H. neanderthalensis 86.4% 77.3% 83.3% 61.1% 71.4% 64.3%
H. sapiens 93.6% 89.4% 95.1% 95.1% 86.1% 77.8%
Total 92.3% 84.6% 89.2% 81.9% 81.1% 73.0%

Table 5 morphological distinction (5-cusped molars in early Homo versus

Mean shape comparisons between H. heidelbergensis, H. neanderthalensis, and 4-cusped molars in later Homo) that may or may not be reflected in
H. sapiens.a
the morphospace depending on the set of landmarks employed
H. heidelbergensis- H. heidelbergensis- H. neanderthalensis- (Go
mez-Robles et al., 2011c). When landmarks demarcating the
H. neanderthalensis H. sapiens H. sapiens fifth cusp are not explicitly included, 5- versus 4-cusped molars
M1 PD ¼ 0.021 PD ¼ 0.029 PD ¼ 0.026 show a gradient across the morphospace, but not a clear disconti-
P <0.001 P <0.001 P <0.001 nuity (Go
mez-Robles et al., 2011c). Following Wood and colleagues
M2 PD ¼ 0.022 PD ¼ 0.038 PD ¼ 0.034
(Wood et al., 1983; Wood and Abbott, 1983), lower molar shape in
P ¼ 0.009 P <0.001 P <0.001
M3 PD ¼ 0.033 PD ¼ 0.047 PD ¼ 0.028 robust taxa is characterized by a combination of a relative reduction
P ¼ 0.004 P <0.001 P ¼ 0.003 of the trigonid area and a relative enlargement of the talonid area.
a
PD: Procrustes distance between the mean shape of listed species. P: p-value
This trend is especially evident when comparing early hominin
based on 10,000 random permutations (testing the null hypothesis of equal mean molars with late hominin molars. Later species show a very marked
shapes in the compared species). reduction of the talonid cusps, which can contribute a significant
lower proportion to the total area of the molar than the trigonid
cusps.
degree, as demonstrated by their marginal location within the None of the Neanderthal second molars in our sample shows a
range of variation of H. heidelbergensis (Fig. 2). 4-cusped configuration, although two molars from SH and one
from Pontnewydd have only four cusps (compared with 72.5% of
M2 H. sapiens). While this may be the result of a sampling bias, it un-
derscores the strong trend to reduction of SH dentitions, which
The results of the classic works by Wood and colleagues reveal seems to be even stronger that that observed in Neanderthals
that shape and size variation of lower second molars corresponding (Bermúdez de Castro and Nicol

mez-Robles et al., 2012).
as, 1995; Go
to Plio-Pleistocene African species show a continuous distribution, Extremely reduced lower second molars with only four cusps and
the extremes of which correspond to P. boisei, and H. habilis and non-Y patterns have also been described in H. erectus specimens
H. ergaster (Wood et al., 1983; Wood and Abbott, 1983). This from Sangiran dated to the Lower Pleistocene or to the early Middle
gradient is also observed in our results, although this variation is Pleistocene (Zanolli, 2013). Three second molars from Arago are
masked due to the inclusion of Asian and European later groups. located far from most SH molars in the PCA plot, and they delim-
Actually, the inclusion of later Homo species reveals a clear itate the area of distribution of H. heidelbergensis in the

Table 6
Inter- and intraspecific allometric analysis.a

Interspecific without phylogeny Independent contrasts (tree 1) Independent contrasts (tree 2) H. heidelbergensis H. neanderthalensis H. sapiens

M1 4.28% 13.26% 18.37% 7.79% 6.54% 7.30%

P <0.001 P ¼ 0.153 P ¼ 0.024 P ¼ 0.076 P ¼ 0.153 P ¼ 0.001
M2 7.08% 21.30% 29.76% 10.63% 6.17% 3.28%
P <0.001 P ¼ 0.001 P ¼ 0.008 P ¼ 0.014 P ¼ 0.390 P ¼ 0.194
M3 9.74% 7.97% 10.31% 9.10% 13.07% 2.79%
P <0.001 P ¼ 0.433 P ¼ 0.267 P ¼ 0.024 P ¼ 0.105 P ¼ 0.455
a
Percentage of shape variance explained by allometric effects. P: p-value based on 10,000 random permutations.

Figure 7. Morphological diversity of lower first molars. H. habilis (A: STW 151), H. heidelbergensis (B: AT-3934), H. neanderthalensis (C: Petit Puymoyen), and H. sapiens (D: Saint
re B3). Orientation of molars is the same as in Figure 1. Scale bars are 5 mm.
Germain-La Rivie

mez-Robles et al. / Journal of Human Evolution 82 (2015) 34e50
A. Go
morphospace (Fig. 4: top chart). As in most mandibular teeth,
H. antecessor has a primitive conformation in lower second molars,
characterized by the large size of the distal half and by the gener-
alized presence of additional cusps (Bermúdez de Castro et al.,
1999b, 2008; Fig. 8).
Both molars from Dmanisi show very different morphologies.
D211 has a general rounded shape and a Y5 pattern. The hypo-
conulid (c5) begins at the central fossa, before the point of inter-
section between the central groove and the lingual groove, a
morphology that is not observed in australopiths. This groove
configuration is also typical of some H. erectus and it is similar to
that observed in some H. habilis, such as OH16. The lower second
molar of D2735, however, has a rectangular and mesiodistally
elongated shape. The evaluation of mesiodistal and buccolingual
diameters (Martino
n-Torres et al., 2008; Macaluso, 2010) reveals
that these different shapes are the result of the shorter mesiodistal
and larger buccolingual diameters of D211. Although not
completely clear due to the presence of additional cusps that
obscure the identification of the main ones, D2735 seems to have a
cruciform groove pattern (Martino
n-Torres et al., 2008). The exis-
tence of secondary cusps (c6 and c7) that distort the location of the
main ones, then altering the groove pattern, is also observed in
Australopithecus, Paranthropus, and H. habilis (Wood and Abbott,
1983; Wood et al., 1983; Tobias, 1991). The derived aspect of
OH22, which plots at the area of the morphospace only occupied by
recent species, is also apparent in Wood et al. (1983). In their
principal components analyses (based either on relative cusp areas
or on the morphology of the groove pattern), OH22 appears as an
outlier and plots far from other African molars, including H. ergaster
KNM-ER992.
M3
Lower third molars show a different pattern of reduction than
that observed in upper third molars (Go
mez-Robles et al., 2012).
Reduction in upper molars is very consistent in H. heidelbergensis,
H. neanderthalensis, and H. sapiens: when only one cusp is reduced
or lost, this is usually the hypocone; if a second cusp is reduced, that
reduction affects the metacone; mesial cusps (protocone and par-
acone) are more stable (Kondo and Yamada, 2003; Takahashi et al.,
2007). Cusp reduction in lower third molars is less consistent and
shows a weaker relationship with size reduction than that observed
in upper third molars (Williams and Corruccini, 2007; Go
mez-
Robles et al., 2012). Actually, H. sapiens lower third molars do not
show significant intraspecific allometric variation, whereas in up-
per third molars both the Neanderthal and modern human lineages
show significant and similar intraspecific allometric trajectories
(Go
mez-Robles et al., 2012).
In spite of this unordered pattern of cusp reduction and the high
intraspecific variation, significant differences in mean shapes of
H. heidelbergensis, H. neanderthalensis, and H. sapiens have been
Figure 8. Morphological diversity of lower second molars. H. antecessor (A: ATD6-113), H. heidelbergensis (B: AT-1), H. neanderthalensis (C: Arcy-sur-Cure: Grotte de l'Hye
recent H. sapiens (D: LT102). Orientation of molars is the same as in Figure 1. Scale bars are 5 mm.
45
identified. However, the taxonomic utility of these differences is
limited. One of the most evident observations related to the lower
third molar analysis is the strong structural reduction of the SH
hominins. The Procrustes distance between the mean shape of
Middle Pleistocene European molars (H. heidelbergensis) and
H. sapiens is considerably higher than the distance observed be-
tween Neanderthals and modern humans. Lower third molars from
Arago plot at a marginal area of the distribution range of
H. heidelbergensis, and they have been described as intermediate in
morphology between Lower Pleistocene European groups and the
SH molars (Bermúdez de Castro et al., 2003a). In spite of the
primitive morphologies observed in the lower dentition of
H. antecessor (Bermúdez de Castro et al., 1999b; Go
mez-Robles
et al., 2008; Go
mez-Robles, 2010), the hemi-mandible ATD6-96
(Carbonell et al., 2005) shows one of the most reduced third molars
observed in the complete sample (including recent H. sapiens). This
molar has a nearly circular shape, and cusps are reduced to small
enamel ridges (Fig. 9B). This molar is located within the range of
distribution of H. sapiens in the PCA plot, close to other earlier
molars that also show a certain degree of reduction (Fig. 4: bottom
chart). The other third molar belonging to H. antecessor, which is
located in the mandibular fragment ATD6-113 (Bermúdez de Castro
et al., 2008), has a much more primitive aspect with five well-
defined cusps arranged in a Y-pattern, in spite of the complex
array of secondary grooves (Fig. 9A). The third molar of the Plio-
Pleistocene mandible D211 is strongly reduced in size, but it still
preserves the five main cusps plus two accessory cusps (c6 and c7;
Margvelashvili, 2006; Martino
n-Torres et al., 2008), thus demon-
strating that size reduction in this case is not associated with
reduction in cusp number. As for australopiths and early Homo, it
has been shown elsewhere that there is considerable overlap in
their ranges of variation and that complex morphologies with
accessory cusps are observed in these groups (Wood et al., 1983).
Serial variation and EvoDevo
First molars are less variable than second and, especially, third
molars. The morphological stability of first molars can be the result
of a canalization process reducing morphological variation or of
stabilizing selection operating in first molars due to their central
developmental and functional role (Go
mez-Robles and Polly, 2012).
The strongest morphological change observed in lower molars is
the reduction of distal cusps, which have been suggested to
constitute a developmental module partially independent of that
formed by mesial cusps of the three molars (Hlusko et al., 2004).
Distal reduction in third molars is first observed in early repre-
sentatives of the genus Homo, although, in these groups, strongly
reduced molars appear only occasionally. The proportion of
reduced versus non-reduced molars seems to be reversed in recent
Homo species, with a majority of individuals losing one or more
cusps, although a certain proportion retain the primitive 5-cusped
ne 4), and
46
mez-Robles et al. / Journal of Human Evolution 82 (2015) 34e50
A. Go
Figure 9. Morphological diversity of lower third molars. H. antecessor (A: ATD6-113; B: ATD6-96), H. heidelbergensis (C: AT-2385), and recent H. sapiens (D: CENIEH 32). Orientation
of molars is the same as in Figure 1. Scale bars are 5 mm.
conformation. This reflects an increasing trend towards reduction
that was initiated in the Lower Pleistocene. This trend is especially
evident in the later developing cusps (distal), thus pointing to a
direct link between developmental dynamics and evolutionary
change (Jernvall, 2000).
Although with species-specific particularities, both Neanderthal
and modern human lineages show a strong trend towards distal
reduction in their premolars and distal molars (Go
mez-Robles and
Polly, 2012). One possible explanation for this is inheritance from
an also distally reduced common ancestor (Go
mez-Robles et al.,
2013). It can also be hypothesized, however, that distal reduction
has evolved in parallel in both lineages. If this were the case, either
shared adaptive pressures or developmental constraints would be
causing the same morphological variation in both lineages. It has
been suggested that restrictions related to space in the jaw and/or
developmental timing can be involved in the structural reduction of
the distal cusps of later developing teeth (premolars and distal
molars; Bermúdez de Castro, 1989; Go
mez-Robles et al., 2012;
Go
mez-Robles and Polly, 2012). Space restrictions could play a
role in reduction in modern humans, but it is unlikely to be oper-
ating in the reduction observed in the Neanderthal lineage. Time
restrictions can be the result of a postdisplacement event causing a
delay in the onset of dental development and giving rise to a
paedomorphic molar that does not completely develop its potential
(Bermúdez de Castro, 1989; Jernvall, 2000). This heterochronic
phenomenon could be related to the appearance of a modern life
history pattern (Bermúdez de Castro, 1989; Go
mez-Robles et al.,
2012), a trait that is thought to have appeared approximately
1 Ma (Bermúdez de Castro et al., 1999a, 2010). If a modern pattern
of dental development already existed in the late Lower Pleistocene
and it is related to dental structural reduction, there would be a
common ancestry to this reduction, even if it has further developed
following slightly different trajectories in the phylogenetic
branches leading to Neanderthals and modern humans. The
morphological trend observed in our series of papers (general distal
reduction in both lineages, but with fine-scale particularities
differentiating both groups) is fully consistent with this hypothesis.
Overview of postcanine dental variation
As this is the last article in our series of papers analysing variation
in postcanine teeth, the final part of this discussion is devoted to
summarizing the most relevant phylogenetic implications of these
morphological trends across the whole postcanine dentition. The
most important interpretations of these morphological patterns in
the context of the interaction between evolution, development, and
function have been discussed elsewhere (Go
mez-Robles and Polly,
2012). Here we focus on the most critical phylogenetic inferences,
some of which have been outlined in the context of ancestral state
reconstruction (Go
mez-Robles et al., 2013), paying special attention
to the European Pleistocene populations coming from Sima de los
Huesos and Gran Dolina-TD6 sites.
Sima de los Huesos fossils Although some teeth from SH have
transitional morphologies between Lower Pleistocene populations
and Neanderthals (i.e., first molars and second premolars), some
dental classes in the SH hominins show more derived morphol-
ogies than in Neanderthals. This is especially evident in distal molar
morphologies, where the proportion of individuals showing
reduction of one or more cusps is not only higher than that found in
Neanderthals but is also higher than the frequencies of reduction
found in many modern human populations (Scott and Turner, 1997;
Bailey, 2002b; Martino
n-Torres, 2006; Martino
n-Torres et al., 2012).
Moreover, Bermúdez de Castro and Nicola
s (1995, 1996) have noted
the extreme reduction in size of the postcanine dentition in the SH
hominins, a feature that is likely related to a reduction in the
morphological complexity of these teeth (Williams and
Corruccini, 2007). Apart from these high frequencies of reduction
of the distal molars, the SH hominins seem to show more derived
morphologies in upper and lower first premolars (Go
mez-Robles
et al., 2008, 2011b). This observation highlights again the greater
variability corresponding to the elements most distant from the
first molars (Go
mez-Robles and Polly, 2012)
The derived morphology of the SH hominins has made some
authors (Endicott et al., 2010; Stringer, 2012) question the
revised chronology of ca. 600 kyr (thousands of years ago)
(Bischoff et al., 2007), and favor a more conservative chronology
of ca. 350 kyr (Bischoff et al., 2003). Recent new data have
yielded a minimum age estimate for the SH hominins of ca. 430
ka (Arnold et al., 2014; Arsuaga et al., 2014). This discrepancy in
SH chronology, however, does not shed light on the strong dental
reduction of the SH hominins in comparison with Neanderthals.
If a progressive and gradual change towards more derived mor-
phologies (from a presumed less reduced ancestral condition) is
expected along the Neanderthal lineage, it is difficult to explain
that the morphology of the SH teeth is even more derived than
that of Neanderthals in aspects related to dental reduction. This
holds even in light of the recently published higher DNA affinity
of the SH hominins with Denisovans than with Neanderthals
(Meyer et al., 2014). First, previous DNA analyses have demon-
strated that affinities based on mitochondrial DNA do not
necessarily match affinities observed through the high-coverage
analysis of the nuclear genome. Mitochondrial DNA analysis first
revealed Denisovans to be the sister group of the
H. neanderthalensis-H. sapiens clade (Reich et al., 2010), whereas
nuclear DNA has shown Denisovans to be the sister group of
Neanderthals but not of modern humans (Meyer et al., 2012).
Second, even if future DNA analyses of the SH hominins
demonstrated that they are indeed closer to Denisovans than to
Neanderthals, this would still imply that the SH hominins have a
closer phylogenetic relationship with Neanderthals than with
modern humans since, as mentioned above, Denisovans and
Neanderthals form a clade to the exclusion of modern humans
(Meyer et al., 2012). In terms of dental morphology, Denisovan
upper third molar size and shape are not only quite different

mez-Robles et al. / Journal of Human Evolution 82 (2015) 34e50
A. Go
from those observed in Neanderthals and modern humans (Reich
et al., 2010) but also very different from those of the Sima de los
Huesos hominins (Personal observation).
The derived condition of the SH hominins can be observed not
only in general conformations but also in discrete dental traits
(Martino
n-Torres et al., 2012). A possible explanation for this is that
phylogenetic signal (Klingenberg and Gidaszewski, 2010) is relaxed
in the evolution of the most distal and mesial elements of the ar-
cade, and/or that dental morphology of the SH hominins shows
extreme morphologies within the probably higher range of varia-
tion of Middle Pleistocene groups, perhaps associated with other
biological particularities of this population. However, it is impor-
tant to note that many other skeletal features in the SH hominins do
not show such derived character states as dental morphology
(Arsuaga et al., 2014). For example, cranial (including temporal) and
facial morphology (Arsuaga et al., 1997; Martínez and Arsuaga et al.,
1997), as well as some postcranial features (Carretero et al., 1997;
Go
mez-Olivencia et al., 2007; Lorenzo, 2007; Go
mez-Olivencia,
2009), show intermediate or primitive conditions with respect to
later Neanderthals. This mosaic pattern of evolution would corre-
spond to a scenario where Neanderthal dental features would have
arisen early on in Neanderthal evolution together with other traits
related to masticatory functions (Arsuaga et al., 2014), giving rise to
a period during which dental morphology changed only slightly or
even faded in its Neanderthal apomorphies. Not surprisingly,
incipient Neanderthal dental features seem to be one of the first
acquisitions of Lower Pleistocene European populations (see
below). The clear derived state of the SH dentitions, as well as the
incipiently derived morphology of Gran Dolina-TD6 hominins to-
wards Neanderthal conformations (Go
mez-Robles et al., 2013),
reveals that dental traits can be leading the Neanderthalization
process in populations that remain undifferentiated (TD6) or not
fully differentiated (SH) in other traits. This model is plausible if
considering the correlation of dental evolution in different dental
classes, which would be evolving as a strongly integrated unit
(Go
mez-Robles and Polly, 2012).
Gran Dolina-TD6 hominins Gran Dolina-TD6 fossils were proposed
to belong to the last common ancestral species of Neanderthals and
modern humans (H. antecessor) according to certain
synapomorphic traits shared by the three species (Bermúdez de
Castro et al., 1997): a high convex anterior border of the temporal
squama, a gracile mandibular corpus without alveolar
prominence at the M1 level, a cranial capacity above 1000 cc, a
modern midfacial topography, and a nearly vertical and anteriorly
positioned incisive canal (Bermúdez de Castro et al., 2004).
Further detailed monographic studies of cranial and postcranial
remains of H. antecessor either supported the position of
H. antecessor as the last common ancestor of H. neanderthalensis
and H. sapiens (Arsuaga et al., 1999; Rosas and Bermúdez de
Castro, 1999) or were unable to rule it out (Carretero et al., 1999;
Lorenzo et al., 1999). Later studies showed similar conclusions
(Carbonell et al., 2005; Bermúdez de Castro et al., 2008; García-
Gonz
alez et al., 2009; Go
mez-Olivencia et al., 2010), although
some questions have been raised about the evolutionary
continuity or discontinuity of Gran Dolina and Sima de los
Huesos populations (Bermúdez de Castro et al., 2003a).
Nevertheless, the study of dental morphology has revealed
the existence of certain similarities between H. antecessor and
H. neanderthalensis (and European H. heidelbergensis), but not
between H. antecessor and H. sapiens (Martino
n-Torres, 2006,
2006; Martino
n-Torres et al., 2007a, 2007b; Go
mez-Robles
et al., 2007, 2011a, b; Go
mez-Robles, 2010). These studies show
a higher resemblance in upper dentition morphology between
H. antecessor and Neanderthals, although the morphology of
lower teeth tends to keep a rather primitive pattern (Bermúdez
47
de Castro et al., 1999b). This can be the result of the stronger
integration and more static pattern of evolution undergone by
the mandibular dentition versus the weaker integration of the
maxillary dentition, which is associated with a less static
evolutionary pattern (Go
mez-Robles and Polly, 2012). In addi-
tion, although it has been claimed that the mandibles from Gran
Dolina-TD6 lack all the apomorphies characteristic of Neander-
thals (Rosas and Bermúdez de Castro, 1999; Carbonell et al.,
2005; Bermúdez de Castro et al., 2008), the ATD6-96 mandible
(the only H. antecessor mandible that preserves the complete
ramus) displays a well-developed medial pterygoid tubercle on
the inner surface of the ramus (Carbonell et al., 2005). This trait
has been proposed to be an apomorphy of the Neanderthal
lineage (Rak et al., 1994; Weaver, 2009), although a recent study
has shown that it can be observed at low frequencies in taxa
other than Neanderthals (Bermúdez de Castro et al., 2015).
Additionally, humeral morphology in H. antecessor has recently
been demonstrated to show a Neanderthal derived morphology
(Bermúdez de Castro et al., 2012).
Part of this apparent disagreement can be explained by the fact
that some characters supporting the existence of a H. antecessor-H.
neanderthalensis-H. sapiens clade are based on highly fragmentary
evidence (e.g., morphology of the temporal squama and cranial
capacity). Moreover, the high cranial capacity of H. antecessor has
been supported for some time by the alleged attribution of the
Ceprano calvaria (with an adult cranial capacity of approximately
1185 cc) to H. antecessor (Manzi et al., 2001). This evidence,
however, was weakened by more recent dating of the Ceprano
fossil to ca. 0.45 Ma (millions of years ago) (Muttoni et al., 2009).
Regarding midfacial topology, classic concerns have focused on the
juvenile status of the ATD6-69 maxilla (e.g., Rightmire, 2001). All
in all, the meaning of facial morphology of H. antecessor remains
controversial. While some studies point to clear similarities with
H. sapiens (Lacruz et al., 2013), others suggest that those similar-
ities are part of a generalized pattern of facial architecture shared
with Middle Pleistocene populations (Vialet, 2005, 2009; Freidline
et al., 2013), or they question the phylogenetic utility of the canine
fossa, claiming that the depressed infraorbital morphology is an
allometric secondary result of small size (Maddux and Franciscus,
2009).
These observations emphasize that, while it is possible to match
individual derived characters from the Gran Dolina-TD6 hominins
with later fossil species, it is not easy to know which of these
characters have the most phylogenetic valence. It is important to
stress that genetic coalescence, population separation, and
phenotypic differentiation of two taxa represent chronologically
distinct events (Hublin, 2009). Different characters may coalesce in
the ancestral species in different ways, giving rise to different
scenarios depending on the studied trait. Some characters (those
with a strong phylogenetic signal) will mirror the topology of the
real phylogeny, whereas others will follow atypical coalescence
patterns. If some traits diversify before separation of the lineages,
the ancestral species will show phenotypic polymorphisms, leading
to the identification of apomorphic traits of one of the lineages in
the incipiently diversifying populations of the ancestral species.
This scenario would be facilitated by geographic fragmentation of
the ancestral species in different (and perhaps isolated) areas
(Dennell et al., 2011).
Focusing on dental morphology, the suite of traits that has
been the subject of detailed analysis in this series of papers, the
phylogenetic position of H. antecessor can be interpreted in
different ways. First, it can be proposed that H. antecessor re-
mains are located at the very beginning of the Neanderthal
lineage. However, this hypothesis is in deep disagreement with
divergence times suggested by the majority of molecular studies
48
mez-Robles et al. / Journal of Human Evolution 82 (2015) 34e50
A. Go
(e.g., Krings et al., 1997, 1999; Ovchinnikov et al., 2000; Ho et al.,
2005; Green et al., 2010). The second possibility is that
H. antecessor is a dead evolutionary lineage (replaced or geneti-
cally absorbed during the Middle Pleistocene by other pop-
ulations, Schwartz and Tattersall, 2005) sharing some phenetic
similarities with Neanderthals. An intermediate explanation
would set H. antecessor as part of the first populations where
incipient Neanderthal characters would be arising, although
reproductive isolation from other penecontemporary pop-
ulations (caused by geographic isolation in Western European
regions) would not be complete (Go
mez-Robles et al., 2013).
Conclusions
The utility of hominin lower molars for taxonomic and phylo-
genetic purposes differs across the three molars. While first molars
show low variability among species but high classification ability,
second and third molars show increased ranges of variation and
decreased ability to correctly classify individuals. Lower first molars
are generally morphologically stable throughout the hominin fossil
record, but lower second and third molars show a clear trend to
distal reduction in later species that parallels distal reductions
observed in upper and lower premolars and upper second and third
molars. Although allometry is involved in the described morpho-
logical changes, allometric variation in lower molars is substantially
weaker than in upper molars. This may be related to the higher
stability of the lower dentition as compared with the upper denti-
tion. Although a parsimonious explanation for the pattern of molar
reduction observed in late Homo species is the inheritance from an
already reduced common ancestor, changes in developmental
timing can also be advocated to support a parallel reduction in
different lineages. Patterns of variation observed in this study, as
well as in previous papers in this series, evince some particularities
of European Pleistocene populations. Within their clear similarities
with Neanderthals, the Sima de los Huesos hominins show a strong
dental reduction, especially in the most distal molars. In fact, the SH
hominins have more reduced second and third molars than classic
Neanderthals. Homo antecessor hominins from TD6 in general have a
primitive lower dentition (with the exception of one extremely
reduced third molar) that contrasts with their more derived,
Neanderthal-like upper teeth. These dental affinities, together with
the information coming from non-dental evidence, support an early
model for the origin of Neanderthals. In order for this model to be
reconciled with the evidence coming from paleogenetic data, it may
be necessary to accept complex evolutionary dynamics and in-
teractions between late Lower and Middle Pleistocene populations.
Acknowledgments
We are grateful to members of the Atapuerca research team and
to researchers in the institutions where we have studied dental
fossils: O. Kullmer, B. Denkel, and F. Schrenk (Senckenberg Insti-
tute); D. Lordkipanidze, A. Vekua, and G. Kiladze (Georgian National
Museum); H. de Lumley, M.A. de Lumley, and A. Vialet (Institut de

ontologie Humaine); P. Mennecier (Muse
Pale
e de l'Homme); I.
Tattersall, G. Sawyer, and G. García (American Museum of Natural
History); L. Jellema and Y. Haile-Selassie (Cleveland Museum of
Natural History); and M. Botella (Laboratory of Physical Anthro-
pology, University of Granada). We are also grateful to the Associate
Editor and the anonymous reviewers for their helpful and
constructive comments. This work was partially supported by
funding by the Spanish Ministry of Science and Innovation (Project
CGL2009-12703-C03-01-02-03). Fieldwork at Atapuerca is sup-
ported by the Consejería de Cultura y Turismo (Junta de Castilla y

n) and the Fundacio
Leo
n Atapuerca.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.jhevol.2015.02.013.
References
Arnold, L.J., Demuro, M., Pare
s, J.M., Arsuaga, J.L., Aranburu, A., Bermúdez de
Castro, J.M., Carbonell, E., 2014. Luminescence dating and palaeomagnetic age
constraint on hominins from Sima de los Huesos, Atapuerca. Spain. J. Hum. Evol.
67, 85e107.
Arsuaga, J.L., Martínez, I., Gracia, A., Lorenzo, C., 1997. The Sima de los Huesos crania
(Sierra de Atapuerca, Spain). A comparative study. J. Hum. Evol. 33, 219e281.
Arsuaga, J.L., Martínez, I., Lorenzo, C., Gracia, A., Mun ~ oz, A., Alonso, O., Gallego, J.,
1999. The human cranial remains from Gran Dolina Lower Pleistocene site
(Sierra de Atapuerca, Spain). J. Hum. Evol. 37, 431e457.
Arsuaga, J.L., Martínez, I., Arnold, L.J., Aranburu, A., Gracia-Te
llez, A., Sharp, W.D.,
Quam, R.M., Falgue res, C., Pantoja-Pe
rez, A., Bischoff, J., Poza-Rey, E., Pare

s, J.M.,
Carretero, J.M., Demuro, M., Lorenzo, C., Sala, N., Martino
n-Torres, M., García, N.,
Alca
zar de Velasco, A., Cuenca-Besco
s, G., Go

mez-Olivencia, A., Moreno, D.,
Pablos, A., Shen, C.-C., Rodríguez, L., Ortega, A.I., García, R., Bonmatí, A., Ber-
múdez de Castro, J.M., Carbonell, E., 2014. Neandertal roots: cranial and chro-
nological evidence from Sima de los Huesos. Science 344, 1358e1363.
Bailey, S.E., 2002a. A closer look at Neanderthal postcanine dental morphology: I.
The mandibular dentition. New Anat. 269, 148e156.
Bailey, S.E., 2002b. Neandertal dental morphology: implications for modern human
origins. Ph.D. Dissertation, Arizona State University.
Bailey, S.E., 2004. A morphometric analysis of maxillary molar crowns of Middle-
Late Pleistocene hominins. J. Hum. Evol. 47, 183e198.
Bailey, S.E., Lynch, J.M., 2005. Diagnostic differences in mandibular P4 shape be-
tween Neandertals and anatomically modern humans. Am. J. Phys. Anthropol.
126, 268e277.
Bailey, S.E., Pilbrow, V.C., Wood, B.A., 2004. Interobserver error involved in inde-
pendent attempts to measure cusp base areas of Pan M1s. J. Anat. 205, 323e331.
Bailey, S.E., Weaver, T.D., Hublin, J.-J., 2009. Who made the Aurignacian and other
early Upper Paleolithic industries? J. Hum. Evol. 57, 11e26.
Bailey, S.E., Skinner, M.M., Hublin, J.-J., 2011. What lies beneath? An evaluation of
lower molar trigonid crest patterns based on both dentine and enamel
expression. Am. J. Phys. Anthropol. 145, 505e518.
Bailey, S.E., Benazzi, S., Hublin, J.-J., 2014. Allometry, merism, and tooth shape of the
upper deciduous M2 and permanent M1. Am. J. Phys. Anthropol. 154, 104e114.
Benazzi, S., Coquerelle, M., Fiorenza, L., Bookstein, F., Katina, S., Kullmer, O., 2011a.
Comparison of dental measurement systems for taxonomic assignment of first
molars. Am. J. Phys. Anthropol. 144, 342e354.
Benazzi, S., Douka, K., Fornai, C., Bauer, C.C., Kullmer, O., Svoboda, J., Pap, I.,
Mallegni, F., Bayle, P., Coquerelle, M., Condemi, S., Ronchitelli, A., Harvati, K.,
Weber, G.W., 2011b. Early dispersal of modern humans in Europe and impli-
cations for Neanderthal behaviour. Nature 479, 525e528.
Bermúdez de Castro, J.M., 1989. Third molar agenesis in human prehistoric pop-
ulations of the. Canary Islands. Am. J. Phys. Anthropol. 79, 207e215.
Bermúdez de Castro, J.M., Nicola
s, M.E., 1995. Posterior dental size reduction in
hominids: the Atapuerca evidence. Am. J. Phys. Anthropol. 96, 335e356.
Bermúdez de Castro, J.M., Nicol
as, M.E., 1996. Changes in the lower premolar-size
sequence during hominid evolution. Phylogenetic implications. Hum. Evol. 11,
205e215.
Bermúdez de Castro, J.M., Arsuaga, J.L., Carbonell, E., Rosas, A., Martinez, I.,
Mosquera, M., 1997. A hominid from the Lower Pleistocene of Atapuerca, Spain:
possible ancestor to Neandertals and modern humans. Science 276, 1392e1395.
Bermúdez de Castro, J.M., Rosas, A., Carbonell, E., Nicola
s, M.E., Rodríguez, J.,
Arsuaga, J.L., 1999a. A modern human pattern of dental development in Lower
Pleistocene hominids from Atapuerca-TD6 (Spain). Proc. Natl. Acad. Sci. 96,
4210e4213.
Bermúdez de Castro, J.M., Rosas, A., Nicola
s, M.E., 1999b. Dental remains from
Atapuerca-TD6 (Gran Dolina site, Burgos, Spain). J. Hum. Evol. 37, 523e566.
Bermúdez de Castro, J.M., Martino
n-Torres, M., Sarmiento, S., Lozano, M., 2003a.
Gran Dolina-TD6 versus Sima de los Huesos dental samples from Atapuerca:
evidence of discontinuity in the European Pleistocene population? J. Archaeol.
Sci. 30, 1421e1428.
Bermúdez de Castro, J.M., Martino
n-Torres, M., Sarmiento, S., Lozano, M.,
Arsuaga, J.L., Carbonell, E., 2003b. Rates of anterior tooth wear in Middle
Pleistocene hominins from Sima de los Huesos (Sierra de Atapuerca, Spain).
Proc. Natl. Acad. Sci. 100, 11992e11996.
Bermúdez de Castro, J.M., Martino
n-Torres, M., Carbonell, E., Sarmiento, S., Rosas, A.,
Van der Made, J., Lozano, M., 2004. The Atapuerca sites and their contribution to
the knowledge of human evolution in Europe. Evol. Anthropol. 13, 25e41.
Bermúdez de Castro, J.M., Pe
rez-Gonza
lez, A., Martino
n-Torres, M., Go
mez-
Robles, A., Rosell, J., Prado, L., Sarmiento, S., Carbonell, E., 2008. A new early
Pleistocene hominin mandible from Atapuerca-TD6, Spain. J. Hum. Evol. 55,
729e735.
Bermúdez de Castro, J.M., Martino
n-Torres, M., Prado, L., Go
mez-Robles, A., Rosell, J.,

pez-Polín, L., Arsuaga, J.L., Carbonell, E., 2010. New immature hominin fossil
Lo
from European Lower Pleistocene shows the earliest evidence of a modern
human dental development pattern. Proc. Natl. Acad. Sci. 107, 11739e11744.

mez-Robles et al. / Journal of Human Evolution 82 (2015) 34e50
A. Go
Bermúdez de Castro, J.M., Carretero, J.M., García-Gonza
lez, R., Rodríguez-García, L.,
Martino
n-Torres, M., Rosell, J., Blasco, R., Martín-France
s, L., Modesto, M.,
Carbonell, E., 2012. Early Pleistocene human humeri from the Gran Dolina-TD6
site (Sierra de Atapuerca, Spain). Am. J. Phys. Anthropol. 147, 604e617.
Bermúdez de Castro, J.M., Quam, R., Martino
n-Torres, M., Martínez, I., Gracia-
Te
llez, A., Arsuaga, J.L., Carbonell, E., 2015. The medial pterygoid tubercle in the
Atapuerca Early and Middle Pleistocene mandibles: Evolutionary implications.
Am. J. Phys. Anthropol. 156, 102e109.
Bischoff, J.L., Shamp, D.D., Aramburu, A., Arsuaga, J.L., Carbonell, E., Bermudez de
Castro, J.M., 2003. The Sima de los Huesos hominids date to beyond U/Th
equilibrium (>350 kyr) and perhaps to 400e500 kyr: new radiometric dates.
J. Archaeol. Sci. 30, 275e280.
Bischoff, J.L., Williams, R.W., Rosenbauer, R.J., Aramburu, A., Arsuaga, J.L., García, N.,
Cuenca-Besco
s, G., 2007. High-resolution U-series dates from the Sima de los
Huesos hominids yields 600 kyrs: implications for the evolution of the early
Neanderthal lineage. J. Archaeol. Sci. 34, 763e770.
Bookstein, F.L., 1991. Morphometric Tools for Landmark Data. Cambridge University
Press, Cambridge.
Carbonell, E., Bermúdez de Castro, J.M., Arsuaga, J.L., Allue, E., Bastir, M., Benito, A.,
Ca
ceres, I., Canals, T., Díez, J.C., van der Made, J., Mosquera, M., Olle
, A., Pe
rez-
Gonza
lez, A., Rodríguez, J., Rodríguez, X.P., Rosas, A., Rosell, J., Sala, R.,
Vallverdú, J., Verge
s, J.M., 2005. An Early Pleistocene hominin mandible from
Atapuerca-TD6. Spain. Proc. Natl. Acad. Sci. 102, 5674e5678.
Carretero, J.M., Arsuaga, J.L., Lorenzo, C., 1997. Clavicles, scapulae and humeri from
the Sima de los Huesos Site (Sierra de Atapuerca, Spain). J. Hum. Evol. 33,
357e408.
Carretero, J.M., Lorenzo, C., Arsuaga, J.L., 1999. Axial and appendicular skeleton of
Homo antecessor. J. Hum. Evol. 37, 459e499.
Caumul, R., Polly, P.D., 2005. Phylogenetic and environmental components of
morphological variation: skull, mandible, and molar shape in marmots (Mar-
mota, Rodentia). Evolution 59, 2460e2472.
Dennell, R.W., Martino
n-Torres, M.a., Bermúdez de Castro, J.M., 2011. Hominin
variability, climatic instability and population demography in Middle Pleisto-
cene Europe. Quatern. Sci. Rev. 30, 1511e1524.
Emonet, E.-G., Tafforeau, P., Chaimanee, Y., Guy, F., de Bonis, L., Koufos, G., Jaeger, J.-
J., 2012. Three-dimensional analysis of mandibular dental root morphology in
hominoids. J. Hum. Evol. 62, 146e154.
Endicott, P., Ho, S.Y., Stringer, C., 2010. Using genetic evidence to evaluate four
palaeoanthropological hypotheses for the timing of Neanderthal and modern
human origins. J. Hum. Evol. 59, 87e95.
Freidline, S.E., Gunz, P., Harvati, K., Hublin, J.-J., 2013. Evaluating developmental
shape changes in Homo antecessor subadult facial morphology. J. Hum. Evol. 65,
404e423.
García-Gonz
alez, R., Carretero, J.M., Rodríguez, L., Go
mez-Olivencia, A., Arsuaga, J.L.,
Bermúdez de Castro, J.M., Carbonell, E., Martínez, I., Lorenzo, C., 2009. Etude
finities and evolutionary history of the early Pleistocene hominids of Java:
analytique d'une clavicule comple te de subadulte d'Homo antecessor (site de
Gran Dolina, Sierra d'Atapuerca, Burgos, Espagne). L'Anthropologie 113.
Gharaibeh, W.S., 2005. Correcting for the effect of orientation in geometric mor-
phometrics studies of side-view images of human heads. In: Slice, D.E. (Ed.),
Modern Morphometrics in Physical Anthropology. Kluwer Academic/Plenum
Publishers, New York, pp. 117e143.
Go
mez-Olivencia, A., 2009. Estudios paleobiolo
gicos sobre la columna vertebral y
la caja tora
cica de los humanos fo
siles del pleistoceno, con especial referencia
a los fo
siles de la Sierra de Atapuerca. Ph.D. Dissertation, Universidad de
Burgos.
Go
mez-Olivencia, A., Carretero, J.M., Arsuaga, J.L., Rodríguez-García, L., García-
Gonza
lez, R., Martínez, I., 2007. Metric and morphological study of the upper
cervical spine from the Sima de los Huesos site (Sierra de Atapuerca, Burgos,
Spain). J. Hum. Evol. 53, 6e25.
Go
mez-Olivencia, A., Carretero, J.M., Lorenzo, C., Arsuaga, J.L., Bermúdez de
Castro, J.M., Carbonell, E., 2010. The costal skeleton of Homo antecessor: pre-
liminary results. J. Hum. Evol. 59, 620e640.
Go
mez-Robles, A., 2010. Ana
lisis de la forma dental en la filogenia humana. Ten-
dencias y modelos evolutivos basados en me
todos de morfometría geome
trica.
Ph.D. Dissertation, Universidad de Granada.
Go
mez-Robles, A., Polly, P.D., 2012. Morphological integration in the hominin
dentition: evolutionary, developmental, and functional factors. Evolution 66,
1024e1043.
Go
mez-Robles, A., Martino
n-Torres, M., Bermúdez de Castro, J.M.,
Margvelashvili, A., Bastir, M., Arsuaga, J.L., Pe
rez-Pe
rez, A., Estebaranz, F.,
Martínez, L.M., 2007. A geometric morphometric analysis of hominin upper first
molar shape. J. Hum. Evol. 53, 272e285.
Go
mez-Robles, A., Martino
n-Torres, M., Bermúdez de Castro, J.M., Prado, L.,
Sarmiento, S., Arsuaga, J.L., 2008. Geometric morphometric analysis of the
crown morphology of the lower first premolar of hominins, with special
attention to Pleistocene Homo. J. Hum. Evol. 55, 627e638.
Go
mez-Robles, A., Bermúdez de Castro, J.M., Martino
n-Torres, M., Prado-Simo
n, L.,
2011a. Crown size and cusp proportions in Homo antecessor upper first molars.
A comment on Quam et al. 2009. J. Anat. 218, 258e262.
Go
mez-Robles, A., Martino
n-Torres, M., Bermúdez de Castro, J.M., Prado-Simo
n, L.,
Arsuaga, J.L., 2011b. A geometric morphometric analysis of hominin upper
premolars. Shape variation and morphological integration. J. Hum. Evol. 61,
688e702.
Go
mez-Robles, A., Olejniczak, A.J., Martino
n-Torres, M., Prado-Simo
n, L., Bermúdez
de Castro, J.M., 2011c. Evolutionary novelties and losses in geometric
49
morphometrics: a practical approach through hominin molar morphology.
Evolution 65, 1772e1790.
Go
mez-Robles, A., Bermúdez de Castro, J.M., Martino
n-Torres, M., Prado-Simo
n, L.,
Arsuaga, J.L., 2012. A geometric morphometric analysis of hominin upper sec-
ond and third molars, with particular emphasis on European Pleistocene pop-
ulations. J. Hum. Evol. 63, 512e526.
Go
mez-Robles, A., Bermúdez de Castro, J.M., Arsuaga, J.-L., Carbonell, E., Polly, P.D.,
2013. No known hominin species matches the expected dental morphology of
the last common ancestor of Neanderthals and modern humans. Proc. Natl.
Acad. Sci. 110, 18196e18201.
Green, R.E., Krause, J., Briggs, A.W., Maricic, T., Stenzel, U., Kircher, M., Patterson, N.,
Li, H., Zhai, W., Fritz, M.H.-Y., Hansen, N.F., Durand, E.Y., Malaspinas, A.-S.,
Jensen, J.D., Marques-Bonet, T., Alkan, C., Prüfer, K., Meyer, M., Burbano, H.A.,
Good, J.M., Schultz, R., Aximu-Petri, A., Butthof, A., Ho € ber, B., Ho€ffner, B.,
Siegemund, M., Weihmann, A., Nusbaum, C., Lander, E.S., Russ, C., Novod, N.,
Affourtit, J., Egholm, M., Verna, C., Rudan, P., Brajkovic, D., Kucan, Z., Gusic, I.,
Doronichev, V.B., Golovanova, L.V., Lalueza-Fox, C., de la Rasilla, M., Fortea, J.,
Rosas, A., Schmitz, R.W., Johnson, P.L.F., Eichler, E.E., Falush, D., Birney, E.,
Mullikin, J.C., Slatkin, M., Nielsen, R., Kelso, J., Lachmann, M., Reich, D., Pa €€
abo, S.,
2010. A draft sequence of the Neandertal genome. Science 328, 710e722.
Grine, F.E., 1986. Dental evidence for dietary differences in Australopithecus and
Paranthropus: a quantitative analysis of permanent molar microwear. J. Hum.
Evol. 15, 783e822.
Guatelli-Steinberg, D., Irish, J.D., 2005. Early hominin variability in first molar dental
trait frequencies. Am. J. Phys. Anthropol. 128, 477e484.
Gunz, P., Mitteroecker, P., Bookstein, F., 2005. Semilandmarks in three dimensions.
In: Slide, D.E. (Ed.), Modern Morphometrics in Physical Anthropology. Springer,
New York, pp. 73e98.
Hlusko, L.J., Maas, M.L., Mahaney, M.C., 2004. Statistical genetics of molar cusp
patterning in pedigreed baboons: implications for primate dental development
and evolution. J. Exp. Zool. 302B, 268e283.
Ho, S.Y.W., Phillips, M.J., Cooper, A., Drummond, A.J., 2005. Time dependency of
molecular rate estimates and systematic overestimation of recent divergence
times. Mol. Biol. Evol. 22, 1561e1568.
Hublin, J.-J., 2009. The origins of Neandertals. Proc. Natl. Acad. Sci. 106,
16022e16027.
Irish, J.D., 1998. Ancestral dental traits in recent sub-Saharan Africans and the ori-
gins of modern humans. J. Hum. Evol. 34, 81e98.
Irish, J.D., Guatelli-Steinberg, D., 2003. Ancient teeth and modern human origins: an
expanded comparison of African Plio-Pleistocene and recent world dental
samples. J. Hum. Evol. 45, 113e144.
Jernvall, J., 2000. Linking development with generation of novelty in mammalian
teeth. Proc. Natl. Acad. Sci. 97, 2641e2645.
Kaifu, Y., Baba, H., Aziz, F., Indriati, E., Schrenk, F., Jacob, T., 2005. Taxonomic af-
dentognathic evidence. Am. J. Phys. Anthropol. 128, 709e726.
Klingenberg, C.P., Gidaszewski, N.A., 2010. Testing and quantifying phylogenetic
signals and homoplasy in morphometric data. Syst. Biol. 59, 245e261.
Klingenberg, C.P., 2011. MorphoJ: an integrated software package for geometric
morphometrics. Mol. Ecol. Resour. 11, 353e357.
Klingenberg, C.P., Maruga
n-Lobo
n, J., 2013. Evolutionary covariation in geometric
morphometric data: analyzing integration, modularity, and allometry in a
phylogenetic context. Syst. Biol. 62, 591e610.
Kondo, S., Yamada, H., 2003. Cusp size variability of the maxillary molariform teeth.
Anthropol. Sci. 111, 255e263.
Krings, M., Stone, A., Schmitz, R.W., Krainitzki, H., Stoneking, M., P€ aa€bo, S., 1997.
Neandertal DNA sequences and the origin of modern humans. Cell 90, 19e30.
Krings, M., Geisert, H., Schmitz, R.W., Krainitzki, H., P€ €bo, S., 1999. DNA sequence of
aa
the mitochondrial hypervariable region II from the Neandertal type specimen.
Proc. Natl. Acad. Sci. 96, 5581e5585.
Kupczik, K., Hublin, J.-J., 2010. Mandibular molar root morphology in Neanderthals
and Late Pleistocene and recent Homo sapiens. J. Hum. Evol. 59, 525e541.
Lacruz, R.S., de Castro, J.M.B., Martino
n-Torres, M., O'Higgins, P., Paine, M.L.,
Carbonell, E., Arsuaga, J.L., Bromage, T.G., 2013. Facial morphogenesis of the
earliest Europeans. PLoS ONE 8, e65199.
Lordkipanidze, D., Ponce de Leo
n, M.S., Margvelashvili, A., Rak, Y., Rightmire, G.P.,
Vekua, A., Zollikofer, C.P.E., 2013. A complete skull from Dmanisi, Georgia, and
the evolutionary biology of early Homo. Science 342, 326e331.
Lorenzo, C., 2007. Evolucio
n de la mano de los homínidos. An
alisis morfolo
gico de
los fo
siles de la Sierra de Atapuerca. Ph.D. Dissertation, Universidad Complu-
tense de Madrid.
Lorenzo, C., Arsuaga, J.L., Carretero, J.M., 1999. Hand and foot remains from the Gran
Dolina Early Pleistocene site (Sierra de Atapuerca, Spain). J. Hum. Evol. 37, 501e522.
Lozano, M., Bermúdez de Castro, J.M., Carbonell, E., Arsuaga, J.L., 2008. Non-
masticatory uses of anterior teeth of Sima de los Huesos individuals (Sierra de
Atapuerca, Spain). J. Hum. Evol. 55, 713e728.
Macaluso, P.J., 2010. Variation in dental remains from Dmanisi, Georgia. Anthropol.
Sci. 118, 31e40.
Macho, G.A., Moggi-Cecchi, J., 1992. Reduction of maxillary molars in Homo sapiens
sapiens: a different perspective. Am. J. Phys. Anthropol. 87, 151e159.
Maddux, S.D., Franciscus, R.G., 2009. Allometric scaling of infraorbital surface
topography in Homo. J. Hum. Evol. 56, 161e174.
Manzi, G., Mallegni, F., Ascenzi, A., 2001. A cranium for the earliest Europeans:
phylogenetic position of the hominid from Ceprano, Italy. Proc. Natl. Acad. Sci.
98, 10011e10016.
50
mez-Robles et al. / Journal of Human Evolution 82 (2015) 34e50
A. Go
Margvelashvili, A., 2006. The morphological description of the dental remains from
the early Paleolithic site of Dmanisi (Georgia). M.Sc. Dissertation, Instituto
Polite
cnico de Tomar-Universidade de Tra
s-os-Montes, Portugal, and Uni-
versitat Rovira i Virgili, Spain.
Martínez de Pinillos, M., Martino
n-Torres, M., Skinner, M.M., Arsuaga, J.L., Gracia-
Te
llez, A., Martínez, I., Martín-France
s, L., Bermúdez de Castro, J.M., 2014.
Trigonid crests expression in Atapuerca-Sima de los Huesos lower molars: in-
ternal and external morphological expression and evolutionary inferences. C. R.
Palevol 13, 205e221.
Martínez, I., Arsuaga, J.L., 1997. The temporal bones from Sima de los Huesos site
(1984e1994). J. Hum. Evol. 33, 283e318.
Martino
n-Torres, M., 2006. Evolucio
n del aparato dental en homínidos: estudio de
los dientes humanos del Pleistoceno de la Sierra de Atapuerca (Burgos). Ph.D.
Dissertation, Universidad de Santiago de Compostela.
Martino
n-Torres, M., Bastir, M., Bermúdez de Castro, J.M., Go
mez, A., Sarmiento, S.,
Muela, A., Arsuaga, J.L., 2006. Hominin lower second premolar morphology:
evolutionary inferences through geometric morphometric analysis. J. Hum.
Evol. 50, 523e533.
Martino
n-Torres, M., Bermúdez de Castro, J.M., Go
mez-Robles, A., Arsuaga, J.L.,
Carbonell, E., Lordkipanidze, D., Manzi, G., Margvelashvili, A., 2007a. Dental
evidence on the hominin dispersals during the Pleistocene. Proc. Natl. Acad. Sci.
104, 13279e13282.
Martino
n-Torres, M., Bermúdez de Castro, J.M., Go
mez-Robles, A., Bastir, M.,
Sarmiento, S., Muela, A., Arsuaga, J.L., 2007b. Gran Dolina-TD6 and Sima de los
Huesos dental samples: preliminary approach to some dental characters of
interest for phylogenetic studies. In: Bailey, S.E., Hublin, J.-J. (Eds.), Dental
Perspectives on Human Evolution. Springer, Berlin, pp. 65e79.
Martino
n-Torres, M., Bermúdez de Castro, J.M., Go
mez-Robles, A.,
Margvelashvili, A., Prado, L., Lordkipanidze, D., Vekua, A., 2008. Dental remains
from Dmanisi (Republic of Georgia): morphological analysis and comparative
study. J. Hum. Evol. 55, 249e273.
Martino
n-Torres, M., Bermúdez de Castro, J.M., Go
mez-Robles, A., Prado-Simo
n, L.,
Arsuaga, J.L., 2012. Morphological description and comparison of the dental
remains from Atapuerca-Sima de los Huesos site (Spain). J. Hum. Evol. 62,
7e58.
Martino
n-Torres, M., Spe va

ckova
, P., Gracia-Te
llez, A., Martínez, I., Bruner, E.,
Arsuaga, J.L., Bermúdez de Castro, J.M., 2013. Morphometric analysis of molars
in a Middle Pleistocene population shows a mosaic of ‘modern’ and Neander-
thal features. J. Anat. 223, 353e363.
Meyer, M., Kircher, M., Gansauge, M.-T., Li, H., Racimo, F., Mallick, S., Schraiber, J.G.,
Jay, F., Prüfer, K., de Filippo, C., Sudmant, P.H., Alkan, C., Fu, Q., Do, R.,
Rohland, N., Tandon, A., Siebauer, M., Green, R.E., Bryc, K., Briggs, A.W.,
Stenzel, U., Dabney, J., Shendure, J., Kitzman, J., Hammer, M.F., Shunkov, M.V.,
Derevianko, A.P., Patterson, N., Andre
s, A.M., Eichler, E.E., Slatkin, M., Reich, D.,
Kelso, J., P€
a€abo, S., 2012. A high-coverage genome sequence from an archaic
Denisovan individual. Science 338, 222e226.
Meyer, M., Fu, Q., Aximu-Petri, A., Glocke, I., Nickel, B., Arsuaga, J.-L., Martinez, I.,
Gracia, A., de Castro, J.M.B., Carbonell, E., P€ €bo, S., 2014. A mitochondrial
aa
genome sequence of a hominin from Sima de los Huesos. Nature 505,
403e406.
Moggi-Cecchi, J., Boccone, S., 2007. Maxillary molars cusp morphology of South
African australopithecines. In: Bailey, S.E., Hublin, J.-J. (Eds.), Dental Perspec-
tives on Human Evolution. Springer-Verlag, Berlin, pp. 53e64.
Moggi-Cecchi, J., Grine, F.E., Tobias, P.V., 2006. Early hominid dental remains from
members 4 and 5 of the Sterkfontein Formation (1966e1996 excavations):
catalogue, individual associations, morphological descriptions and initial
metrical analysis. J. Hum. Evol. 50, 239e328.
Molnar, S., 1971. Human tooth wear, tooth function and cultural variability. Am. J.
Phys. Anthropol. 34, 175e190.
Muttoni, G., Scardia, G., Kent, D.V., Swisher, C.C., Manzi, G., 2009. Pleistocene
magnetochronology of early hominin sites at Ceprano and Fontana Ranuccio,
Italy. Earth Planet. Sci. Lett. 286, 255e268.
Ovchinnikov, I.V., Gotherstrom, A., Romanova, G.P., Kharitonov, V.M., Liden, K.,
Goodwin, W., 2000. Molecular analysis of Neanderthal DNA from the northern
Caucasus. Nature 404, 490e493.

rez, S.I., Bernal, V., Gonza
Pe
lez, P.N., 2006. Differences between sliding semi-
landmark methods in geometric morphometrics, with an application to hu-
man craniofacial and dental variation. J. Anat. 208, 769e784.
Rak, Y., Kimbel, W.H., Hovers, E., 1994. A Neandertal infant from Amud Cave, Israel.
J. Hum. Evol. 26, 313e324.
Reich, D., Green, R.E., Kircher, M., Krause, J., Patterson, N., Durand, E.Y., Viola, B.,
Briggs, A.W., Stenzel, U., Johnson, P.L., Maricic, T., Good, J.M., Marques-Bonet, T.,
Alkan, C., Fu, Q., Mallick, S., Li, H., Meyer, M., Eichler, E.E., Stoneking, M.,
Richards, M., Talamo, S., Shunkov, M.V., Derevianko, A.P., Hublin, J.-J., Kelso, J.,
Slatkin, M., Pa €a
€bo, S., 2010. Genetic history of an archaic hominin group from
Denisova Cave in Siberia. Nature 468, 1053e1060.
Rightmire, G.P., 2001. Patterns of hominid evolution and dispersal in the Middle
Pleistocene. Quatern. Int. 75, 77e84.
Rohlf, F.J., 1998. TpsRelw. TpsSeries. Department of Ecology and Evolution, SUNY,
Stony Brook, New York.
Rohlf, F.J., 2005. TpsDig2. TpsSeries. Department of Ecology and Evolution, SUNY,
Stony Brook, New York.
Rohlf, F.J., Slice, D., 1990. Extensions of the Procrustes method for the optimal su-
perimposition of landmarks. Syst. Zool. 39, 40e59.
Rosas, A., Bermúdez de Castro, J.M., 1999. The ATD6-5 mandibular specimen from
Gran Dolina (Atapuerca, Spain). Morphological study and phylogenetic impli-
cations. J. Hum. Evol. 37, 567e590.
Scott, G.R., Turner, C.G., 1997. The Anthropology of Modern Human Teeth: Dental
Morphology and its Variation in Recent Human Populations. Cambridge Uni-
versity Press, Cambridge.
Schwartz, J.H., Tattersall, I., 2005. The Human Fossil Record, Craniodental
Morphology of Early Hominids (Genera Australopithecus, Paranthropus, Orrorin),
and Overview. Wiley-Liss, New York.
Skinner, M.M., Gunz, P., Wood, B.A., Hublin, J.-J., 2008. Enamel-dentine junction
(EDJ) morphology distinguishes the lower molars of Australopithecus africanus
and Paranthropus robustus. J. Hum. Evol. 55, 979e988.
Stringer, C., 2012. The status of Homo heidelbergensis (Schoetensack 1908). Evol.
Anthropol. 21, 101e107.
Suwa, G., Wood, B.A., White, T.D., 1994. Further analysis of mandibular molar crown
and cusp areas in Pliocene and early Pleistocene hominids. Am. J. Phys.
Anthropol. 93, 407e426.
Suwa, G., White, T.D., Howell, F.C., 1996. Mandibular postcanine dentition from the
Shungura Formation, Ethiopia: crown morphology, taxonomic allocations, and
Plio-Pleistocene hominid evolution. Am. J. Phys. Anthropol. 101, 247e282.
Takahashi, M., Kondo, S., Townsend, G.C., Kanazawa, E., 2007. Variability in cusp size
of human maxillary molars, with particular reference to the hypocone. Arch.
Oral Biol. 52, 1146e1154.
Tobias, P.V., 1991. Olduvai Gorge. The Skulls Endocasts and Teeth of Homo habilis.
Cambridge University Press, Cambridge.
Vialet, A., 2005. La face supe
rieure et moyenne des hominide
s fossiles depuis le
Ple
istocene infe
rieur re
cent. Ph.D. Dissertation, Muse
um National d'Histoire
Naturelle, Paris.
Vialet, A., 2009. L'e
volution de l'homme en Eurasie: nouveaux fossiles, nouveaux
outils d'analyse. L'Anthropologie 113, 245e254.
Weaver, T.D., 2009. The meaning of Neandertal skeletal morphology. Proc. Natl.
Acad. Sci. 106, 16028e16033.
Weaver, T.D., Roseman, C.C., Stringer, C.B., 2007. Were Neandertal and modern
human cranial differences produced by natural selection or genetic drift?
J. Hum. Evol. 53, 135e145.
Williams, B.A., Corruccini, R.S., 2007. The relationship between crown size and
complexity in two collections. Dent. Anthropol. 20, 29e32.
Wood, B.A., Abbott, S.A., 1983. Analysis of the dental morphology of Plio-Pleistocene
hominids I. Mandibular molars: crown area measurements and morphological
traits. J. Anat. 136, 197e219.
Wood, B., Abbott, S.A., Graham, S.H., 1983. Analysis of the dental morphology of
Plio-Pleistocene hominids II. Mandibular molars e study of cusp areas, fissure
pattern and cross sectional shape of the crown. J. Anat. 137, 287e314.
Zanolli, C., 2013. Additional evidence for morpho-dimensional tooth crown varia-
tion in a new Indonesian H. erectus sample from the Sangiran Dome (Central
Java). PLoS ONE 8, e67233.