Middle Paleolithic human remains from the Gruta Da Oliveira (Torres Novas), Portugal

Abstract

Additional Middle Paleolithic human remains from layers 17, 18, and 22 of the Gruta da Oliveira, Portugal consist of a proximal manual phalanx 2 (Oliveira 5), a partial postcanine tooth (Oliveira 6), a humeral diaphysis (Oliveira 7), a distal mandibular molar (Oliveira 8), and a mandibular premolar (P(3) ) (Oliveira 9). Oliveira 5, 6, and 8 are unremarkable for Late Pleistocene humans. The Oliveira 7 right humerus is moderately robust or the individual had the stocky body proportions of other European (including Iberian) Neandertals. The Oliveira 9 P(3) has a large and symmetrical crown and lacks a distal accessory ridge and accessory lingual cusps, overlapping both Neandertal and recent human ranges of variation. It contrasts with at least recent human P(3) s in having relatively thin enamel. These join the Oliveira 1 to 4 remains in further documenting early MIS 3 Neandertal morphology in western Iberia.

Full text

Middle Paleolithic Human Remains from the Gruta da
Oliveira (Torres Novas), Portugal
John C. Willman,
1
Julia Maki,
1
Priscilla Bayle,
2
Erik Trinkaus,
1
* and Joa˜ o Zilha˜o
3
1
Department of Anthropology, Washington University, Saint Louis, MO 63130
2
UMR 5199 PACEA, Universite
´de Bordeaux, CNRS, MCC, INRAP, Talence F-33405, France
3
Departament de Prehisto
`ria, Historia Antiga i Arqueologia, Universitat de Barcelona/ICREA, Facultat de Geografia
i Histo
`ria, c/ Montalegre 6, Barcelona 08001, Spain
KEY WORDS Neandertal; late Pleistocene; teeth; humerus; phalanx; enamel; Europe
ABSTRACT Additional Middle Paleolithic human
remains from layers 17, 18, and 22 of the Gruta da Oli-
veira, Portugal consist of a proximal manual phalanx 2
(Oliveira 5), a partial postcanine tooth (Oliveira 6), a
humeral diaphysis (Oliveira 7), a distal mandibular molar
(Oliveira 8), and a mandibular premolar (P
3
) (Oliveira 9).
Oliveira 5, 6, and 8 are unremarkable for Late Pleisto-
cene humans. The Oliveira 7 right humerus is moderately
robust or the individual had the stocky body proportions
of other European (including Iberian) Neandertals. The
Oliveira 9 P
3
has a large and symmetrical crown and
lacks a distal accessory ridge and accessory lingual cusps,
overlapping both Neandertal and recent human ranges of
variation. It contrasts with at least recent human P
3
sin
having relatively thin enamel. These join the Oliveira
1 to 4 remains in further documenting early MIS 3
Neandertal morphology in western Iberia. Am J Phys
Anthropol 149:39–51, 2012. V
V
C2012 Wiley Periodicals, Inc.
Until recently. Iberia contributed relatively little to
Neandertal paleontology (Garralda, 2006). However, a
series of discoveries has greatly increased the Iberian
Neandertal presence (Quam et al., 2001; Lorenzo and
Montes, 2001; Arsuaga et al., 2001b, 2007; Barroso-Ruı
´z
et al., 2003; Daura et al., 2005, Rosas et al., 2006; Sar-
rio
´n, 2006; Trinkaus et al., 2007; Walker et al., 2008,
2012), confirmed their late presence south of the Pyre-
nees (Walker et al., 2008; cf., Hublin et al., 1995), docu-
mented aspects of their paleobiology (Lalueza et al.,
1993; Arsuaga et al., 2001a; Rosas et al., 2006; Walker et
al., 2011b,c; Estalrrich et al., 2011), and raised issues
regarding Neandertal variability (Rosas et al., 2006;
Walker et al., 2008, 2011b; Alca
´zar de Velasco et al.,
2011).
In this context, we present additional Middle Paleo-
lithic human remains from the Gruta da Oliveira, Torres
Novas, Estremadura, Portugal. Previously described
human specimens from Oliveira (Trinkaus et al., 2007)
confirmed that these remains derived from Neandertals,
especially Oliveira 1 and 3. The additional remains come
from deposits of the same or similar age and contribute
to both Neandertal paleobiology and variability within
Iberia.
THE GRUTA DA OLIVEIRA
The Gruta da Oliveira (39830@23@N, 8836049@W) is
a collapsed entrance of the multilevel karstic system
associated with the spring of a tributary of the Tagus,
the Almonda River, located 100 km NE of Lisbon. Dis-
covered in 1989, the site contains a Middle Paleolithic
sequence excavated over a thickness of 7 m (Fig. 1).
Extensive, updated discussions of the stratigraphy and
dating are in Angelucci and Zilha
˜o (2009), Zilha
˜o et al.
(2010), and Hoffmann et al. (in press). As of 2011, our
understanding of the site’s formation can be summarized
as follows.
The baseline for the accumulation of the deposits is a
stalagmitic crust for which a terminus post quem of
107 ka is provided by the U-Th age of a candlestick
stalagmite growing on top of it in the Crivo area of this
level of the karstic system. Above, there are 3mof
deposits, the surface of which is exposed in the Cone
Moustierense, the first discovered area of the site, from
the bottom, via speleo-archeological exploration of the
inner karst passages. A Uranium series age of 62.5 ka
was obtained on horse (Equus caballus) tooth enamel
from a sample retrieved therein in 1990, under Early
Uptake assumptions (Zilha
˜o and Mckinney, 1995); the
date is the average of two measurements on the same
sample: 70,250 69,000 and 53,000 15,600/25,300 (1 r
errors). The surface of the Cone Moustierense corre-
sponds to the interface between layers 25 and 26 of the
stratigraphic succession that has been undergoing top–
down archeological excavation since the collapsed en-
trance was opened in 1991. This succession can be subdi-
vided in five main blocks (Fig. 1), from the bottom up:
Grant sponsor: FCT (Fundac¸a
˜o para a Cie
ˆncia e Tecnologia)
PTDCHIS-ARQ0981642008; Grant sponsor: Grant sponsor: STEA
(Sociedade Torrejana de Espeleologia e Arqueologia); Grant sponsor:
IPA (Instituto Portugue
ˆs de Arqueologia); Grant sponsor: RENOVA
(Fa
´brica de Papel do Almonda); Grant sponsor: NSF, the Wenner-
Gren and Leakey Foundations, C.N.R.S.
*Correspondence to: Erik Trinkaus, Department of Anthropology,
Washington University, Saint Louis, MO 63130, USA.
E-mail: trinkaus@artsci.wustl.edu
Received 10 January 2012; accepted 9 April 2012
DOI 10.1002/ajpa.22091
Published online 21 May 2012 in Wiley Online Library
(wileyonlinelibrary.com).
V
V
C2012 WILEY PERIODICALS, INC.
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 149:39–51 (2012)

layers 23–25 are roof collapse and associated sedi-
ment fill, with a low density of artifacts and faunal
remains scattered in homogeneously over a thick-
ness of 1.5 m;
layers 20–22, a 65-cm thick block of sandy, cave
earth sediments, formed during a period of relative
structural stabilization of the roof and walls of the
cave, allowing its use for habitation, as documented
by a large hearth, 1.5 m in diameter, whose con-
tents and associated spatial scatter are layer 21;
layers 15–19, a 1.3-m thick block of deposits formed dur-
ing a period of destabilization that saw multiton bould-
ers collapse over the surfaces of layers 20, 19, and 15; at
this time, human occupation took place in the extant
cave porch (the ‘‘Exterior’’), with run-off and slope proc-
esses leading to syn- or post-depositional, short-distance
displacement of finds towards the excavated area;
layers 9–14 accumulated in the depression formed
inward of the huge boulder fallen over the surface of
layer 15 in grid units N-R/12–15, with human occu-
pation taking place in the ‘‘Side Passage’’ and the
‘‘27-S Chamber’’, as documented by a hearth in grid
units L20–21, layer 14, and the associated dense
concentration of lithics, fauna, charcoal, and burnt
bone;
layers 7–8 formed on top of and outward from the N-
R/12–15 boulder, with a tail extending into the inner
areas, which by layer 9 times had become almost
filled-up; layer 8 corresponds to the latest human
occupation of the site, while layer 7, archeologically
sterile, contains faunal remains probably accumu-
lated by carnivores and is capped by flowstone, in
turn buried under a thick colmatation breccia.
Radiocarbon dates on burnt bone, plus U-series results
on layer 8 bone samples, place layers 7–8 37 ka cal BP,
and layers 9–14 beyond 43.5 ka cal BP. The remains
described here come from deeper in the stratigraphy,
within early marine isotope stage (MIS) 3. The Oliveira
Fig. 1. Gruta da Oliveira. Top: schematic topographic profile of the collapsed entrance and the passages leading to it (left), and
site plan at the elevation of layer 10 with the excavation grid (right); the locations of the relevant speleothem samples and the des-
ignation of the site’s different zones are shown. Bottom: schematic stratigraphic profile with indications of the main stratigraphic
blocks and available dating results.
40 J.C. WILLMAN ET AL.
American Journal of Physical Anthropology

5 phalanx and Oliveira 6 tooth were found at the inter-
face between layers 17 and 18, and the Oliveira 7 hu-
merus was in the upper part of layer 18, 11 cm away
from and 7 cm below the previously described Oliveira 3
humerus from layer 18. Oliveira 8 and 9 were recovered
25 cm from each other in layer 22, the latter 9 cm
above the former [i.e., given the dip of the deposit and
the approximate positioning of Oliveira 8 (see below), at
about the same elevation].
MATERIALS AND METHODS
This article is a presentation of newly excavated Late
Pleistocene human remains, and as such it describes
them in the context of relevant comparative samples.
They are compared to three samples, Late Pleistocene
(MIS 5 to 3) western Eurasian late archaic humans
(Neandertals), southwest Asian early Late Pleistocene
(MIS 5c) middle Paleolithic modern humans (MPMH),
and Eurasian and North African Interpleniglacial (MIS
3) Early and Mid Upper Paleolithic modern humans
(EUP/MUP). The one exception to this is the dental dis-
crete trait frequencies. Most of those data come from
Bailey (2006), and include a few southern African MSA
(MIS 5–4) late archaic humans (which some consider to
be early modern humans) with the MPMH; also, the
upper Paleolithic sample includes both MIS 3 and 2
specimens. As with the southwest Asian MPMH, the
African MSA teeth appear to mostly retain plesiomor-
phic traits and therefore should not bias the compari-
sons. The MIS 2 teeth are notable mostly for reductions
in molar cusp numbers, in the direction of recent
humans. Comparative data for P
3
dental tissue propor-
tions are principally from recent humans, plus data for
two Neandertals.
The linear measurements follow the Martin system
(M-#) (Bra
¨uer, 1988), supplemented as needed. Dental
discrete traits follow the ASUDAS system (Scott and
Turner, 1997) as modified and expanded for later Pleisto-
cene humans by Bailey (2002, 2006). Occlusal wear
scores follow Molnar (1971) and Smith (1984).
The Oliveira 7 distal humeral diaphysis was scanned
on a Siemens Somatom Definition CT Scanner at the
Washington University School of Medicine (kVp: 120;
effective mass: 400 mA 3s; rotation time: 1; pitch: 0.7;
columnation: 16 30.3 mm; slice thickness: 400 lm;
reconstruction increment: 100 lm), producing 1,291 sli-
ces. Geometric cross-sectional parameters (areas and sec-
ond moments of area) were computed from the slice at
35% of length using SLICE/SLCOMM (Nagurka and
Hayes, 1980; Eschman, 1992) after visual reconstruction
of missing medial diaphyseal bone (see below).
The Oliveira 9 P
3
was micro-CT scanned using a
Scanco VivaCT 40 scanner at the Washington University
School of Medicine (70 kVp; 114 lA; 500 projections).
The final volume was reconstructed with an isotropic
voxel size of 21 lm. A buccolingual section was virtually
produced through the P
3
(visually reconstructing the
worn cusp tip) using Amira 5.2.2 software (Visage Imag-
ing) and following the method described by Feeney et al.
(2010) (Fig. 9). Physically and virtually produced sec-
tions closely approximate one another (Olejniczak and
Grine, 2006), and therefore recent human comparative
data were tabulated from both physical (Saunders et al.,
2007; Smith et al., 2008) and virtually generated
(Feeney et al., 2010; Smith et al., 2012) sections. The re-
sultant measurements (Table 6) follow Martin (1985)
and were derived from the section using MPSAK 2.9
software (Dean and Wood, 2003).
Comparisons are made principally using the standar-
dized linear residuals (raw residual/standard deviation
of the pooled residuals) from the reduced major axis
(RMA) regression line through the pooled comparative
sample trimmed of outliers. Significance is based on non-
parametric (Kruskal-Wallis) assessments [computed
using NCSS 7.1 (Hintze, 2010)]. Summary statistics are
nonetheless provided as: mean 6SD (N). Exact chi-
square P-values for the comparative discrete trait distri-
butions were computed with Stat-Xact 4.0 (Mehta and
Patel, 1999). Within sets of comparisons, P-values are
evaluated using a sequentially reductive multiple com-
parison correction, following Rice (1989) and Proschan
and Waclawiw (2000).
Oliveira 5—Proximal manual phalanx
Preservation and identification. The partial proximal
diaphysis and base of a human right second manual
proximal phalanx (Olv P15-478) was discovered during
excavation on July 3, 2008 at the base of layer 17 (Fig.
2). The base is complete with minor surface abrasion to
the dorsal margin of the metacarpal facet but primarily
to the dorsal capsular attachment area. From the intact
subchondral bone margins radially and ulnarly, it does
not appear that more than a couple of tenths of a milli-
meter of the dorsal articular facet were lost. The capsu-
lar insertion area has lost 0.5–1.0 mm of bone. The dor-
sal diaphysis extends distally for 18.9 mm from the mid-
dorsal articular facet, but the palmar extent is only 10.7
mm from the palmar middle of the facet. There are thin
irregular concretions on the metacarpal facet and
between the proximal palmar tubercles for the metacar-
pophalangeal collateral ligaments, but neither of them
obscures morphology nor affects measurements. The
maximum preserved length is 20.9 mm.
Fig. 2. Oliveira 5 right manual proximal phalanx 2 in dorsal
(upper left), palmar (upper right), and proximal (below) views.
Scale bar: 1 cm. [Color figure can be viewed in the online issue,
which is available at wileyonlinelibrary.com.]
41OLIVEIRA MIDDLE PALEOLITHIC HUMANS
American Journal of Physical Anthropology

The bone is identified as a manual proximal phalanx
given its evenly concave articular facet, size and the
larger radioulnar than dorsopalmar dimensions of the
proximal facet and epiphysis. It is from the right second
ray given the modest interosseus tubercle on the right
base (when viewed dorsoproximally) and the prominent
one on the left base. The radial versus ulnar tubercle dif-
ferential is greater than seen on third proximal pha-
langes.
There is no evidence of the proximal epiphyseal fusion
line and, therefore, the bone derives from an individual
minimally in the second half of the second decade post-
natal and probably fully mature (Greulich and Pyle,
1971). There are no pathological lesions.
Morphology. The metacarpal facet is transversely ovoid
with a concavity that is principally radioulnar. The max-
imum depth of the concavity, from the tangent to its ra-
dial and ulnar margins, is 1.4 mm. The ulnar base is
minimally expanded for the capsular attachment and
the associated tendon for the second palmar interosseus
muscle. The radial side has a large and prominent swel-
ling for the insertion of the tendon of the first dorsal
interosseus muscle, evident especially dorsoradially. The
preserved portion of the diaphysis has dorsally thin cort-
ical bone, 1.2 mm thick at the most distally preserved
mid-dorsal margin.
The Oliveira 5 phalangeal base is average to moder-
ately small (Table 1). The one comparison which has the
potential to provide paleobiological information, between
the maximum and articular breadths (Fig. 3), should
reflect the hypertrophy of the insertion for the first dor-
sal interosseus, assuming that articular breadth is an in-
dependent variable related to body size. There is no dif-
ference between the Neandertals and the EUP/MUP
remains in this comparison. The MPMH (all from Qaf-
zeh) have relatively larger maximum breadths, espe-
cially Qafzeh 3. The comparative samples are different
at P50.010, driven by the high values for the MPMH
sample (Neandertal – EUP/MUP Wilcoxon P50.649). It
is unclear why the MPMH sample should be largely sep-
arate from the others, given the generally (but not uni-
versally) reduced hypertrophy of their hands relative to
the Neandertals (Niewoehner, 2001; Maki and Trinkaus,
2011).
In this comparison, Oliveira 5 falls along the higher
margin of the Neandertal—EUP/MUP distribution,
emphasizing its moderately large first dorsal interosseus
tubercle, most similar to the Qafzeh specimens. Its
residual value of 1.23 is above all of the other Late Pleis-
tocene specimens except Qafzeh 3 and 7.
Oliveira 6—Postcanine tooth fragment
Preservation and identification. Oliveira 6 (Olv P16-
812) is a fragment of a heavily worn permanent premo-
lar or molar, discovered on July 3, 2008 at the base of
layer 17 (Fig. 4). The tooth retains 6.5 mm of the side of
the crown and exposed dentine extending up to 4.1 mm
from the external enamel. At its most complete margin,
the crown height is 2.9 mm. There is 10.8 mm of a
crushed partial root encrusted in matrix. There is no oc-
clusal enamel, only the remains of a worn dentine basin
(probably Smith wear Stage 7 and Molnar wear Stage
6).
The retained enamel has no hypoplasias or other path-
ological lesions. There is, however, a rectangular chip of
enamel missing from the middle of the preserved occlu-
sal margin, 1.8 mm wide and extending 0.8 mm from the
occlusal margin (Fig. 4). The margins are rounded, espe-
cially on the right side (with the occlusal surface up) and
the internal margin towards the dentine. The chip was
therefore lost antemortem, with subsequent attrition in
the margins. It does not have the hollowed out and
etched form of a carious lesion. It appears to be an ante-
mortem traumatic chipping of the enamel border that
occurred after it was weakened by wear. Similar enamel
chipping is evident on other Late Pleistocene teeth (e.g.,
Frayer and Russell, 1987; Mallegni, 1995; Hillson, 2006;
Doboset al., 2010). The tooth provides little other pale-
ontological information.
TABLE 1. Comparative metrics of the Oliveira 5 second proximal manual phalanx base (mm)
Maximum height (M-10) Maximum breadth (M-8) Articular height (M-10a) Articular breadth (M-8a)
Oliveira 5 (11.3) 16.7 10.0 12.7
Neandertals 11.9 60.9 (11) 16.1 61.8 (11) 10.2 60.8 (12) 13.2 61.1 (11)
MPMH 12.5 60.8 (4) 17.2 61.4 (4) 10.6 61.0 (4) 12.9 60.9 (4)
EUP/MUP 12.4 60.8 (8) 17.3 61.2 (13) 10.6 61.0 (8) 13.5 60.7 (9)
K-W P-value 0.350 0.299 0.614 0.469
Mean 6SD (N) for the comparative samples. Kruskal–Wallis P-values across the three comparative samples. Estimated value is in
parentheses. MPMH, Middle Paleolithic modern humans; EUP/MUP, Early and Mid Upper Paleolithic modern humans.
Fig. 3. Bivariate plot of second proximal manual phalanx
base maximum versus articular breadths. RMA line through
the trimmed pooled comparative samples (PMB 51.73 3PAB –
6.50; r50.870; N523; comparative sample Kruskal-Wallis P:
0.0097). MPMH, Middle Paleolithic modern humans; EUP/MUP,
Early and Mid Upper Paleolithic modern humans. [Color
figure can be viewed in the online issue, which is available at
wileyonlinelibrary.com.]
42 J.C. WILLMAN ET AL.
American Journal of Physical Anthropology

Oliveira 7—Distal humeral diaphysis
Preservation. The Oliveira 7 humerus (Olv O16-422)
was discovered during excavation on July 12, 2008 in
the upper portion of layer 18. It is a distal diaphysis
from the distal end of the deltoid tuberosity anteriorly to
proximal olecranon fossa (Fig. 5). The diaphysis was
cleaned extensively with a 5.0% and then 7.5% solution
of acetic acid, which removed little of the red-brown car-
bonate encrustation that covers principally the antero-
medial surface; given that the encrustation is thin and
hides little morphology, further efforts to chemically or
mechanically clean the bone were deemed inadvisable.
There are no age indicators preserved on the specimen,
but its overall size and morphology imply full maturity.
There are no pathological lesions. Its maximum pre-
served length is 129 mm.
The humerus has a 1–2 mm wide longitudinal crack
from the proximal end of the lateral pillar posteriorly to
the posterolateral edge towards midshaft; it is 70 mm
long. There is a larger gap along the medial side of the
medial supracondylar area, 3.5 mm wide, which then
rotates to the posteromedial margin as it goes proximally
up to the posterior proximal break. Despite the size of
this medial-posteromedial opening, the CT images con-
firm that there was no expansion of the diaphysis along
that region, only a loss of bone with variable matrix
infilling (Fig. 5).
Proximally the bone has an irregular break trans-
versely across the diaphysis on the anterior three-quar-
ters of the bone. Posteriorly, there is a piece 22 mm
long and 17 mm wide missing, distally to the postero-
medial gap. The anterolateral margin of the bone at the
break has the distal deltoid tuberosity, extending 10
mm from the proximoanterior break. The proximal break
is therefore close to midshaft. Distally, there is the be-
ginning of the midline depression for the olecranon fossa,
but it does not extend to the capsular attachment line.
The medial side is broken off at the proximal end of the
olecranon fossa, but the lateral side extends slightly fur-
ther and retains most of the lateral supracondylar crest.
The edges proximally and distally exhibit the angular
breaks of dry bone, similar to the Oliveira 3 distal hu-
merus and distinct from the rounded margins of the
gnawed Oliveira 4 tibial diaphysis. There is no carnivore
damage to the bone. Linear measurements and mid-
distal cross-sectional geometric parameters are in Table 2.
Overall size. The bone preserves the distal deltoid tu-
berosity and the proximal olecranon fossa, which enable
estimation of its original anatomical lengths; the dis-
tance between them is 105 mm (61–2 mm, depending
on interpretation of the distal deltoid tuberosity margin).
This distance is moderately well correlated with articu-
lar length in a pooled sample of recent humans (N540)
plus six Late Pleistocene humeri (five Neandertals plus
Skhul 4) (ArtLen 51.715 3OlvLen 1128; r
2
50.614).
The resultant estimate for Oliveira 7 (308 613 mm) has
a large SE
est
(4.6%) and 95% CI (282–334 mm). The
same sample predicts a maximum length of 315 613
mm (95% CI; 288–342 mm) for Oliveira 7 (MaxLen 5
1.749 3OlvLen 1132; r
2
50.613). Scaling of the dia-
physeal rigidity (Fig. 6A) suggests that the true value of
its length lies within the upper half of the 95% CI,
unless it was exceptionally robust.
The maximum length mean estimate is above the over-
all mean of a Neandertal sample (Table 3); it is within
the male range (305–337 mm, N56) and well above
that of the females (272–286 mm). It is also below the
means of the MPMH and EUP/MUP samples. Yet, the
95% CI for Oliveira 7 encompasses half of the MPMH
values, 71.4% of the EUP/MUP ones, and all of the
Neandertal values except for the three females.
Fig. 4. Oliveira 6 postcanine partial crown and root in exter-
nal (below) and occlusal-external (above) views. Scale bar: 1 cm.
[Color figure can be viewed in the online issue, which is avail-
able at wileyonlinelibrary.com.]
Fig. 5. Oliveira 7 distal right humeral diaphysis in anterior
(A), posterior (P), medial (M), and lateral (L) views; scale bar: 5
cm. Mid-distal (35%) diaphyseal cross-section (CT slice); scale
bar: 2 cm. [Color figure can be viewed in the online issue, which
is available at wileyonlinelibrary.com.]
43OLIVEIRA MIDDLE PALEOLITHIC HUMANS
American Journal of Physical Anthropology

Morphology.
Surficial morphology. Oliveira 7 presents little surface
morphology of note. The lateral supracondylar crest is
distinct and extends 1.5 mm from the distal break to
the level of the minimum diaphyseal circumference. The
distal deltoid tuberosity is moderately rugose but proj-
ects minimally from the diaphyseal contour. At the prox-
imal break, it is 9.7 mm wide, suggesting an original
maximum breadth of 10–12 mm. The distal end is insuf-
ficiently preserved to assess the pillar thicknesses or
whether it had a septal aperture.
Diaphyseal hypertrophy. The distal minimum circumfer-
ence of 68 mm is modestly above the means of the com-
parative samples (Table 3). A mean robusticity index
using distal circumference and maximum length (21.6) is
within the Neandertal range (18.7 – 23.9) but outside
the ranges of variation of the early modern human sam-
ples (MPMH max: 19.7; EUP/MUP max: 20.6). Using the
minimum 95% CI value (19.9) for Oliveira 7 places it at
the top of the early modern human variation.
The midshaft does not provide a full cross-section, but
the mid-distal (35%) cross-section is largely present (Fig.
TABLE 2. Dimensions of the Oliveira 7 humerus
Maximum length estimate (mean; 95% CI) (M-1) (mm) 315; 95% CI: 288–342
Articular length estimate (mean; 95% CI) (M-2) (mm) 308; 95% CI: 282–334
Distal minimum circumference (M-7) (mm) 68.0
Midshaft maximum diameter (M-5) (mm) (25.0)
Midshaft minimum diameter (M-6) (mm) 19.3
Mid-distal (35%) total area (mm
2
) 348.1
Mid-distal (35%) cortical area (mm
2
) 304.5
Mid-distal (35%) maximum second moment of area (mm
4
) 10,963
Mid-distal (35%) minimum second moment of area (mm
4
) 8,327
Mid-distal (35%) polar moment of area (mm
4
) 19,290
Fig. 6. Metric comparisons of the Oliveira 7 humeral diaphysis to Late Pleistocene right humeri. A: 35% polar moment of area
(J) versus length. B: 35% cortical (CA) versus total subperiosteal (TA) areas. C: 35% maximum (I
max
) versus minimum (I
min
) second
moments of area. D: midshaft maximum versus minimum external diameters. The horizontal line in A provides the 95% CI for
humeral length. MPMH, Middle Paleolithic modern humans; EUP/MUP, Early and Mid Upper Paleolithic modern humans. [Color
figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
44 J.C. WILLMAN ET AL.
American Journal of Physical Anthropology

5). It has been possible to visually reconstruct the miss-
ing portions from the CT slice (slice 935) located at 35%
of the predicted articular length, filling in the posterome-
dial and posterolateral bone following the preserved con-
tours. The endosteal contour was visually separated
from the matrix in-filling.
A comparison of the 35% polar moment of area (as an
overall reflection of diaphyseal rigidity) to humeral
length (Fig. 6A) places the mean value for Oliveira 7
among the more robust of the Neandertal humeri, with a
significant difference across the comparative samples (P
50.007). With a standardized residual of 1.52 for Oli-
veira 7, only Shanidar 4 provides a relatively higher
value (2.07), but other specimens (especially La Chap-
elle-aux-Saints 1 and Kebara 2 at 1.33) are close. Its
mean estimate is well above those of the early modern
humans, although it is close to the [possibly pathologi-
cally elevated (Crevecoeur, 2008)] value (1.04) for Nazlet
Khater 2. The longer length from its 95% CI places it
among the stronger of the early modern humans, but the
shorter length limit would make it exceptionally robust
(standardized residual of 2.74).
These two reflections of humeral diaphyseal robust-
ness can reflect either generalized humeral diaphyseal
hypertrophy, in Oliveira 7 and the Neandertals in gen-
eral, or they may be largely due to the stockier body pro-
portions of the Neandertals relative to early modern
humans (Holliday, 1997). It is not possible to determine
the body proportions of Oliveira 7 (or the other Oliveira
Neandertals), but two more southern Iberian Neander-
tals, Palomas 96 and probably Palomas 92 (Walker et al.,
2011a,b), exhibit the ‘‘arctic’’ body proportions of more
northern European Neandertals. Among recent humans
(Ruff, 2000; Shackelford, 2005), consistent patterns of
robustness are achieved only if humeral diaphyseal
strength is scaled to bone length times body mass. Given
this, the comparisons of diaphyseal circumference or po-
lar moment of area to length alone can indicate strong
arms and/or ‘‘arctic’’ body proportions for Oliveira 7.
A further (partial) reflection of diaphyseal hypertrophy
is cortical area relative to subperiosteal area (Fig. 6B).
There is little difference across Late Pleistocene right
humeri (P50.196). Oliveira 7 falls along the upper mar-
gins of the distribution, close to the highest relative val-
ues for Neandertals and early modern humans; its
standardized residual of 1.31 is close to those of Dolnı
´
Ve
˘stonice 41 (1.31), Krapina 173 (1.24), Nahal ‘En-Gev 1
(1.28) and Regourdou 1 (1.44).
Diaphyseal shape. Diaphyseal shape can be assessed
using the perpendicular second moments of area at the
mid-distal diaphysis and external diameters near mid-
shaft. In the former, there is no difference across the
comparative samples (P50.965), and Oliveira 7 is in
the middle of the distribution (standardized residual:
20.04) (Fig. 6C). At midshaft, the Oliveira 7 minimum
breadth can be measured directly and the maximum di-
ameter can be estimated by extending the posterior bor-
der proximally. The resultant values (Table 2) place Oli-
veira 7 in the overlap of the Neandertal and the EUP/
MUP samples (Fig. 6D). The larger MPMH humeri fall
with the rounder Upper Paleolithic humeri, but the
smaller ones cluster with the Neandertals. As a result,
the three samples are significantly different (P50.016)
and the Neandertal and EUP/MUP ones are especially
contrasting despite some overlap (Wilcoxon P50.003).
Oliveira 7 has a positive standardized residual (0.43),
but one nonetheless exceeded by Mladec
ˇ24, Pavlov 1,
Pr
ˇedmostı
´4, the three smaller Skhul 2, 7 and 15, and
71.4% of the 14 Neandertals.
Oliveira 8—Lower molar
Preservation and identification. The Oliveira 8 tooth
(Olv N16-373) was found during excavation on July 1,
2010 in layer 22. It was encased in extremely hard brec-
cia removed during excavation with a Spit electro-pneu-
matic hammer; it was discovered as the tool flaked off
the brecciated deposit, exposing the sliced specimen. The
portions recovered include a partial crown from the oc-
clusal surface to the cervix and half of a root still
encased in the breccia (Fig. 7). The partial crown fits
cleanly onto the root with minor loss of cervical material.
The root is exposed internally; it was left in the matrix,
since it would be nearly impossible to extract it intact.
Extended search failed to locate the remainder of the
tooth; it probably shattered from the impact.
The tooth is a distal portion of a mandibular molar,
probably from the right side. There is a complete cusp
preserved, between portions of two other cusps. The
three cusps conform to the pattern seen in the distal por-
tions of lower molar crowns, in which the intact one is
the hypoconulid (cusp 5) and the two partially preserved
cusps would be the hypoconid (cusp 3) and the entoconid
(cusp 4). The sulci between the middle cusp and the
other two cusps converge in the buccolingual middle of
the occlusal surface, deviate slightly to the right and
then mesially. The exposed root has a single canal and
converges to a closed apex. It curves apically away from
the break relative to the occlusal plane, or distally for a
lower molar. It also curves slightly to the left when
viewed mesially.
Given the orientation of the root and the positions of
the cusps and sulci between them, the tooth is best con-
sidered as right, making the more intact of the two more
mesial cusps the hypoconid. The distal crown lacks an
interproximal wear facet. The occlusal cusps have only a
slight polishing of the enamel (Smith and Molnar wear
Stage 2), indicating that it is an M
3
or a more mesial
molar prior to the eruption of the next distal molar.
The root apex of the distal root is completely closed
[Moorrees et al. (1963) Ac], which indicates a median
TABLE 3. External osteometric comparisons of the Oliveira 7 humeral diaphysis to distributions of Late Pleistocene samples
Maximum length (mm) Distal minimum circumference (mm) Robusticity index
Oliveira 7 315 (95% CI; 288–342) 68.0 21.6 (95% CI: 19.9–23.6)
Neandertals 306.5 617.7 (13) 64.4 66.6 (11) 21.0 61.8 (9)
MPMH 355.1 625.2 (4) 65.9 63.1 (4) 18.6 61.2 (4)
EUP/MUP 335.3 624.3 (28) 62.9 64.6 (20) 18.8 61.3 (18)
K-W P-value \0.001 0.697 0.010
Given length estimation for Oliveira 7, the 95% CI is provided for both the length and the robusticity index. MPMH, Middle Paleo-
lithic modern humans; EUP/MUP, Early and Mid Upper Paleolithic modern humans.
45OLIVEIRA MIDDLE PALEOLITHIC HUMANS
American Journal of Physical Anthropology

age [9 years if an M
1
,[14 years if an M
2
, and [20
years if an M
3
, based on recent EuroAmerican standards
(Shackelford et al., 2012) despite indications of relatively
rapid M3 formation in Late Pleistocene humans (Tomp-
kins, 1996) [note that the absolute dental developmental
rates of Late Pleistocene humans are unknown (Gua-
telli-Steinberg, 2009) but are probably within the ranges
of variation of recent humans (Shackelford et al., 2012)].
The tooth is therefore a lower right molar of indeter-
minate number. Yet, it derives from the same excavation
layer as Oliveira 9, a lower right P
3
recovered nearby
(see below). It is possible that both teeth derive from the
same mandible, given their stratigraphic contemporane-
ity and their modest degrees of occlusal wear. The Oli-
veira 9 P
3
has more occlusal wear than Oliveira 8, as
well as distinct interproximal facets. It also had achieved
apex fusion stage either A
1/2
or A
c
, indicating an age-at-
death 12–13 years (see below). If the two teeth are
from one arcade, then it is unlikely that Oliveira 8 rep-
resents an M
1
and most likely it is an M
2
or M
3
, given
its minimal occlusal wear and no distal interproximal
facet.
In addition to the excavation damage, the crown lost a
chip of enamel during fossilization from the distolingual
corner, at the end of the sulcus between the entoconid
and the hypoconulid. There is a small hypoplastic pit on
the buccal face of the hypoconid, 4.4 mm from the cervix;
unworn buccal crown height on the hypoconid is 6.2 mm.
Around the distal crown, 2.0 mm from the cervix, there
is a shallow groove with traces of calculus. The groove
appears to be a mild linear dental enamel hypoplasia.
These hypoplasias join the relatively high frequencies of
such defects on the distal molars among the Neandertals
(Ogilvie et al., 1989).
Morphology. Oliveira 8 is unlikely to have had four
cusps, similar to almost all Late Pleistocene M
1
s and
M
2
s, all Neandertal M
3
s and 70% of early modern
human M
3
s (Table 4). If the three distal cusps are cor-
rectly identified, then it lacks a cusp 6 (stage 0); the
common configuration across Late Pleistocene samples.
Note that the only significant difference in these discrete
traits in the Late Pleistocene is in whether M
2
s and M
3
s
have only 4 cusps; if Oliveira 8 is an M
3
, its cusp num-
ber would align it more with the Neandertals.
The buccolingual breadth at the mesial break is 9.3
mm, and the original crown breadth is likely to have
been 10–11 mm. A crown diameter in that range is mod-
est but unexceptional for a Late Pleistocene lower molar
(Trinkaus, 2004).
Oliveira 9—Lower first premolar (P
3
)
Preservation and attrition. A lower first premolar
(P
3
) (Olv N15-383) was discovered on July 7, 2010 in the
hard breccia facies of layer 22 that had to be excavated
with power tools (see above) (Fig. 8). Its two pieces rejoin
without distortion at the cervix. The root apex is absent,
and the tapering suggests that 1–2 mm were lost. There
is also an area of the disto-occlusal quarter of the buccal
crown, 535mm
2
, where the enamel has been
slightly etched. The lingual half of the lingual cusp
(metaconid) is absent; it does not obscure the size of the
tubercle, but it necessitates minor estimation of the
crown buccolingual diameter and contour.
The occlusal surface of the protoconid has been planed
off through normal occlusal wear, resulting in exposure
of a central, 0.5 mm diameter, area of dentine (Smith
wear Stage 2b; Molnar wear Stage 3). There is a flat
Fig. 7. Oliveira 8 right mandibular molar (probably M
2
or M
3
). Left: distal root and crown in mesial view. Upper right: distal
crown in occlusal view. Lower right: the crown in distal view. Scale bars: 1 cm. [Color figure can be viewed in the online issue,
which is available at wileyonlinelibrary.com.]
46 J.C. WILLMAN ET AL.
American Journal of Physical Anthropology

mesial interproximal facet in the middle of the occlusal
half of the mesial crown, 3.2 mm high and 2.4 mm wide.
The middle of it still covers the dentine, to a minimum
enamel thickness of 0.7 mm. The mesial side has there-
fore been planed off substantially. Distally, there is a flat
interproximal facet 2.9 mm high and 4.1 mm wide. How-
ever, it is on the distolingual side of the crown, along the
distal marginal ridge. It does not go onto the mid-distal
extent of the tooth, such that there is no reduction in
the crown mesiodistal diameter from distal interproxi-
mal wear (only from mesial interproximal wear).
Although readily apparent in occlusal view, it did not
significantly change the crown contour of the tooth.
Given the normal mesial but lingually displaced distal
interproximal facets, there was some dental rotation. Ei-
ther the distal P
3
was buccally displaced relative to the
P
4
, or the mesial P
4
was lingually displaced. Given that
the mesial facet is in the normal position for the C
1
, the
latter scenario is probable. Rotated premolars are com-
mon among the Neandertals, being particularly frequent
in the Krapina sample; they are also present in both
Middle and Upper Paleolithic modern humans (Hillson,
2006; Rougier et al., 2006).
The damaged root apex indicates that it should have
been developmentally at least at A
1/2
and probably A
c
.
The former would provide a median age of 13 years,
and the latter would indicate an age [12 years, based
on recent EuroAmericans (Shackelford et al., 2012).
Morphology. Except for damage to the lingual metaco-
nid, the occlusal morphology of Oliveira 9 is intact. The
buccal surface is rounded with a bulge on the cervical
half and largely straight in mesial view on the occlusal
half. The protoconid appears to have had a slightly lon-
ger distal than mesial shoulder. There is a complete, if
concave, transverse crest from the protoconid to the
metaconid (Grade 1), but there is neither a mesial nor a
distal accessory ridge (Grade 0). There are well-formed
mesial and distal marginal ridges, but the mesial one is
traversed next to the metaconid by a distinct but small
mesial lingual groove (Grade 1). The metaconid is cen-
trally located on the lingual side (Grade 0), and there is
little asymmetry of the crown (Grade 0); the lingually
displaced distal interproximal facet has little effect on
the scoring of the crown as essentially symmetrical.
There are no accessory lingual cusps (Grade 1).
These discrete traits of the P
3
, taken individually,
place Oliveira 9 within the ranges of variation of the
Late Pleistocene samples (Table 5). It differs from the
Neandertals principally in its symmetry, the one feature
in which the samples are significantly different,
although there is considerable overlap across the sam-
TABLE 4. Percent present (#/N) for discrete traits of Late Pleistocene lower molars, for which Oliveira 8 provides data
a
Oliveira 8 Neandertals MPMH1MSA Upper Paleol. P-value
b
M
1
4 cusps Absent 1.9% (1/53) 0.0% (0/12) 2.4% (1/42) 0.393
Cusp 6 Absent 30.8% (8/26) 0.0% (0/7) 14.8% (4/27) 0.020
M
2
4 cusps Absent 2.2% (1/45) 10.0% (1/10) 41.7% (10/24) *\0.001
Cusp 6 Absent 48.2% (13.5/28) 0.0% (0/3) 20.0% (4/20) 0.013
M
3
4 cusps Absent 0.0% (0/25) 28.6% (2/7) 30.0% (6/20) *0.001
Cusp 6 Absent 41.7% (5/12) 40.0% (2/5) 44.4% (8/18) 0.107
a
Comparative data from Bailey (2006), with the addition of data from Kostenki, Oase, Palomas, La Quina-Aval, Sunghir, and Val-
degoba. Note that the MPMH1MSA sample includes specimens from Qafzeh, Skhul and southern African MSA sites, and the
Upper Paleolithic sample includes Early, Mid, and Late Upper Paleolithic specimens.
b
Exact Chi-square P-values provided across the comparative samples.
* Significant at P\0.05 after a multiple comparison correction.
Fig. 8. The Oliveira 9 right P
3
in (left to right) mesial, distal, buccal, lingual and occlusal views. Scale bars: 1 cm. Brackets on
the occlusal view indicate the positions of the interproximal facets. [Color figure can be viewed in the online issue, which is avail-
able at wileyonlinelibrary.com.]
47OLIVEIRA MIDDLE PALEOLITHIC HUMANS
American Journal of Physical Anthropology

ples. It should be noted that although Bailey (2006) and
the expanded data set in Table 5 score Neandertal P
3
s
as overwhelmingly ‘‘asymmetrical,’’ Martino
´n-Torres et
al. (2012) considered them to be only slightly asymmetri-
cal. Oliveira 9 also differs from most Neandertals in its
absence of accessory lingual cusps and especially a distal
accessory ridge.
It has been possible to reasonably visually reconstruct
the missing crown portions for a buccolingual slice
through the cusp tips (Fig. 9) and estimate enamel and
dentine areas (Table 6). Limited comparative Late Pleis-
tocene data for P
3
s are available, but more substantial
data exist for samples of recent humans. The Oliveira 9
coronal dentine area (including the pulp space) and
enamel-dentine junction length are similar to the larger
mean values for the recent human samples, but they are
modestly smaller than the mean of two other Neandertal
P
3
s. Yet, its enamel area is matched only by the mean of
the small EuroCanadian teeth, resulting in low average
(AET) and relative (RET) enamel thicknesses, even
below the mean of the two other Neandertal P
3
s. How-
ever, for the samples for which recent human distribu-
tions are available (Feeney et al., 2010; Smith et al.,
2012), the Oliveira 9 values are within their ranges of
variation. It is nonetheless likely that Oliveira 9 follows
the general Neandertal pattern of having low RET, at
least relative to recent humans (cf., Olejniczak et al.,
2008; Bayle et al., 2010; Crevecoeur et al., 2010; Smith
et al., 2012).
The Oliveira 9 root is moderately broad in its cervical
half, and then it tapers modestly to the apical break.
There is a single shallow and smooth longitudinal groove
on both its mesial and distal sides. There is no separa-
tion of the root into two (Grade 0 or 1), the common later
Pleistocene and recent human pattern (Martino
´n-Torres
et al., 2012).
The Oliveira 9 crown buccolingual breadth as pre-
served is 8.8 mm, but the damage to the metaconid
means that it was slightly larger. Rounding out the lin-
gual contour provides a crown breadth of 9.4 mm. This
is associated with a (reduced) mesiodistal length of 7.9
mm (estimated original length: 8.3 mm), cervical
length and breadth of 5.8 and 7.8 mm, respectively, and
a maximum root buccolingual diameter of 7.8 mm. The
attritionally reduced buccal crown height is 8.2 mm; the
buccal root length is currently 13.3 mm, but it was origi-
nally 1–2 mm longer.
The one meaningful comparative external tooth mea-
sure, given wear and available comparative data, is
crown breadth. The comparative samples are signifi-
cantly different from each other (P50.006) (Fig. 10),
and the Oliveira 9 estimate of 9.4 mm places it above
the Neandertal mean (9.1 60.7 mm, N550) and near
the upper limits of the early modern human samples
(MPMH: 8.8 60.5 mm, N58; EUP/MUP: 8.6 60.5
mm, N527). It is a moderately large P
3
for the Late
Pleistocene.
Possible associations
Oliveira 5 to 7 derive from the same stratigraphic
complex as Oliveira 3 (layers 17 and 18). Oliveira 3 and
7 are right humeri, and both preserve the proximal olec-
ranon fossa; they therefore derive from different individ-
uals. Oliveira 5 and 6 could derive separately or together
from either of these individuals. Oliveira 8 and 9, both
right mandibular teeth, come from the same layer and
could well derive from the same hemi-mandible if Oli-
veira 8 is an M
2
or M
3
.
SUMMARY
These fragmentary human remains from layers 17, 18,
and 22 of the Gruta da Oliveira provide further informa-
tion on Neandertal skeletal and dental morphology
during early MIS 3 in Iberia. The previously described
TABLE 5. Lower P
3
occlusal discrete traits for Oliveira 9 and Late Pleistocene comparative samples, counted in terms of presence of
the trait [percent (#/N)]
a
Oliveira 9 Neandertals MPMH1MSA Upper Paleol. P-value
b
Mesial accessory ridge Absent (0) 27.8% (5/18) 0.0% (0/2) 12.5% (1/8) 0.312
Distal accessory ridge Absent (0) 90.9% (20/22) 50.0% (1/2) 100% (9/9) 0.085
Extra lingual cusps None (1) 30.8% (12/39) 16.7% (1/6) 6.7% (1/15) 0.020
Metaconid mesial Central (0) 13.5% (5/37) 50.0% (2/4) 18.8% (3/16) 0.078
Transverse crest Present (1) 91.2% (31/34) 75.0% (3/4) 76.5% (13/17) 0.047
Asymmetry Absent (0) 91.7% (22/24) 75.0% (3/4) 52.9% (9/17) *0.003
Mesial lingual groove Present (1) 62.1% (18/29) 25.0% (1/4) 53.3% (8/15) 0.040
a
Comparative data from Bailey (2006) with the addition of observations from Palomas, Sunghir, and Valdegoba. Trait grade scores
for Oliveira 9 are provided in parentheses. Note that the MPMH1MSA sample includes specimens from Qafzeh, Skhul, and south-
ern African MSA sites, and the Upper Paleolithic sample includes Early, Mid, and Late Upper Paleolithic specimens.
b
Exact chi-square P-values provided across the comparative samples.
* Significant at P\0.05 after a multiple comparison correction.
Fig. 9. 3D virtual reconstruction of the Oliveira 9 P
3
in oc-
clusal view (left) showing the location of the buccolingual slice
(right) used in the analysis of dental tissue proportions. Scale
bar: 5 mm. [Color figure can be viewed in the online issue,
which is available at wileyonlinelibrary.com.]
48 J.C. WILLMAN ET AL.
American Journal of Physical Anthropology

Oliveira 1 to 4 remains were attributed to the Neander-
tals based on the relative breadth of the Oliveira 1 pha-
lanx, the thin pillars of the Oliveira 3 distal humerus,
and the evenly convex contours of the Oliveira 4 tibial
diaphysis near midshaft (Trinkaus et al., 2007). These
additional remains confirm their Neandertal affiliation
through the relatively thin enamel of Oliveira 9
[although early modern human variation in this feature
is poorly known (cf., Bayle et al., 2010)] and the probable
diaphysis to length proportions of Oliveira 7 and implied
body proportions. The other remains and other aspects
of these specimens are morphologically undiagnostic.
ACKNOWLEDGMENTS
The collection of the comparative data has been made
possible by curators too numerous to mention individu-
ally. The CT imaging was done through the Center for
Clinical Imaging Research and the Department of Ortho-
pedic Surgery, Washington University School of Medi-
cine. Authors are grateful to all.
LITERATURE CITED
Alca
´zar de Velasco A, Arsuaga JL, Martı
´nez I, Bonmatı
´A. 2011.
Revisio
´n de la mandı
´bula humana de Ban
˜olas, Gerona,
Espan
˜a. Bol R Soc Esp Hist Nat Sec Geol 105:99–108.
Angelucci D, Zilha
˜o J. 2009. Stratigraphy and formation proc-
esses of the Late Pleistocene deposit at Gruta da Oliveira,
Almonda Karstic System, Torres Novas, Portugal. Geoarch-
aeology 24:277–310.
Arsuaga JL, Martı
´nez I, Lorenzo C, Quam R, Carretero M, Gra-
cia A. 2001a. Las estrı
´as del incisivo de Cova Negra. In: Villa-
verde V, editor. De Neandertales a Croman
˜ones. El Inicio del
Poblamiento Humano en las Tierras Valencianas. Valencia:
Universitat de Vale
`ncia. p327–328.
Arsuaga JL, Martı
´nez I, Villaverde V, Lorenzo C, Quam R, Car-
retero JM, Gracia A. 2001b. Fo
´siles humanos del paı
´s valen-
ciano. In: Villaverde V, editor. De Neandertales a Croma-
n
˜ones. El Inicio del Poblamiento Humano en las Tierras
Valencianas. Valencia: Universitat de Vale
`ncia. p265–322.
Arsuaga JL, Villaverde V, Quam R, Martı
´nez I, Carretero JM,
Lorenzo C, Gracia A. 2007. New Neandertal remains from
Cova Negra (Valencia, Spain). J Hum Evol 52:31–58.
Bailey SE. 2002. Neandertal dental morphology: implications
for modern human origins. Ph.D. Thesis, Arizona State Uni-
versity, Tempe.
Bailey SE. 2006. Beyond shovel-shaped incisors: Neandertal
dental morphology in a comparative context. Period Biol
108:253–267.
Barroso-Ruı
´z C, Lumley MA de, Caparro
´s M, Verdu
´L. 2003.
Los restos humanos Neandertalenses de la Cueva del Boquete
de Zafarraya. In: Barrosos-Ruı
´z C, editor. El Pleistocene Supe-
rior del la Cueva del Boquete de Zafarraya. Sevilla: Conseje-
rı
´a de Cultura. p327–387.
Bayle P, Macchiarelli R, Trinkaus E, Duarte C, Mazurier A, Zil-
ha
˜o J. 2010. Dental maturational sequences and dental tissue
proportions in the early Upper Paleolithic child from Abrigo
do Lagar Velho, Portugal. Proc Natl Acad Sci USA 107:1338–
1342.
Bra
¨uer G. 1988. Osteometrie. In: Knussman R, editor. Anthro-
pologie I. Stuttgart: Fischer Verlag. p160–232.
Fig. 10. Boxplot of lower P
3
crown buccolingual breadths for
Oliveira 9 (Olv9), Neandertals (Nean), Middle Paleolithic mod-
ern humans (MPMH), and Early and Mid Upper Paleolithic
modern humans (EUP/MUP).
TABLE 6. Comparison of dental tissue dimensions and proportions
a
from a reconstructed buccolingual slice (Fig. 9) for Oliveira 9,
compared to Neandertal and recent human P
3
sample mean values
N
Enamel
area (mm
2
)
Coronal dentine
area (mm
2
)
Enamel-dentine junction
length (mm) (EDJ)
Average enamel
thickness (mm) (AET)
Relative enamel
thickness (RET)
Oliveira 9
b
1 (15.27) (35.76) (18.76) (0.81) (13.61)
Neandertals
c
2 17.61 38.87 19.68 0.89 14.38
Indonesians
d
7 18.49 62.51 35.11 65.07 18.62 61.41 0.99 60.09 16.76
Northern Europeans
d
13 18.93 62.34 35.87 62.63 18.63 60.87 1.02 60.13 16.97
Southern Africans
d
9 18.30 63.05 30.93 65.73 17.16 61.53 1.06 60.09 19.18
EuroCanadians
e
45 15.63 25.74 15.05 1.04 22.36
Recent humans
f
17 18.08 33.16 17.85 1.01 17.78
a
The measurements and proportions include: the area of the enamel cap (mm
2
), the area of the coronal dentine (including the pulp
space) (mm
2
), the enamel-dentine junction (EDJ) length (mm), the average enamel thickness (AET) defined as the quotient of
enamel area and EDJ length (mm), and the relative enamel thickness (RET) defined as the AET divided by the square root of the
coronal dentine area (as a surrogate for tooth size) and multiplied by 100. Standard deviations are available only for some of the
measurements in Feeney et al. (2010).
b
A portion of the cuspal enamel and dentine has been reconstructed given occlusal wear (see Figs. 8 and 9). The values are in
parentheses, indicating this estimation.
c
Data from Smith et al. (2012).
d
Data from Feeney et al. (2010). The RET values have been calculated from the mean values for coronal dentine area and AET pro-
vided by Feeney et al. (2010).
e
Data from Saunders et al. (2007). Their male and female mean values have been combined. The AET value is calculated from the
mean values of enamel area and EDJ length. The RET is the mean of the combined male and female samples.
f
Mixed recent and living human samples. Data from Smith et al. (2008).
49OLIVEIRA MIDDLE PALEOLITHIC HUMANS
American Journal of Physical Anthropology

Crevecoeur I. 2008. E
´tude anthropologique du squelette du
Pale
´olithique Supe
´rieur de Nazlet Khater 2 (E
´gypte). Leuven:
Leuven University Press.
Crevecoeur I, Bayle P, Rougier H, Maureille B, Higham T, van
der Plicht J, De Clerck N, Semal P. 2010. The Spy VI child: a
newly discovered Neandertal infant. J Hum Evol 59:641–656.
Daura J, Sanz M, Subira
´ME, Quam R, Fullola JM, Arsuaga
JL. 2005. A Neandertal mandible from the Cova del Gegant
(Sitges, Barcelona, Spain). J Hum Evol 49:56–70.
Dean MC, Wood B. 2003. A digital radiographic atlas of great
apes skull and dentition. In: Bondioli L, Macchiarelli R, edi-
tors. Digital archives of human paleobiology. Milan: ADS Sol-
utions. CD-ROM.
DobosA, Soficaru A, Trinkaus E. 2010. The prehistory and pa-
leontology of the Pestera Muierii, Romania. Etud Rech
Arche
´ol Univ Lie
`ge 124:1–122.
Eschman PN. 1992. SLCOMM Version 1.6. Albuquerque:
Eschman Archeological Services.
Estalrrich A, Rosas A, Garcı
´a-Vargas S, Garcı
´a-Tabernero A,
Santamarı
´a D, de la Rasilla M. 2011. Subvertical grooves on
interproximal wear facets from the El Sidro
´n (Asturias,
Spain) Neandertal dental sample. Am J Phys Anthropol
144:154–161.
Feeney RNM, Zermeno JP, Reid DJ, Nakashima S, Sano H, Bahar
A, Hublin JJ, Smith TM. 2010. Enamel thickness in Asian
human canines and premolars. Anthropol Sci 118:191–198.
Frayer DW, Russell MD. 1987. Artificial grooves on the Krapina
Neanderthal teeth. Am J Phys Anthropol 74:393–405.
Garralda MD. 2006. Los Neandertales en la Penı
´nsula Ibe
´rica.
Munibe 57:289–314.
Greulich WW, Pyle SI. 1971. A radiographic standard of refer-
ence for the growing hand and wrist. Cleveland: Case West-
ern Reserve University Press.
Guatelli-Steinberg D. 2009. Recent studies of dental develop-
ment in Neandertals: implications for Neandertal life histor-
ies. Evol Anthropol 18:9–20.
Hillson SW. 2006. Dental morphology, proportions and attrition.
In: Trinkaus E, Svoboda JA, editors. Early modern human
evolution in central Europe: the people of Dolnı
´Ve
˘stonice and
Pavlov. New York: Oxford University Press. p179–223.
Hintze J. 2010. NCSS 7.1.20. Kaysville: NCSS.
Hoffmann D, Pike AWG, Wainer K, Zilha
˜o J. New U-series
results for the speleogenesis and the Palaeolithic archaeology
of the Almonda karstic system (Torres Novas, Portugal). Qua-
tern Int (in press).
Holliday TW. 1997. Body proportions in Late Pleistocene Europe
and modern human origins. J Hum Evol 32:423–447.
Hublin JJ, Barroso-Ruı
´z C, Medina Lara P, Fontugne M, Reyss
JL. 1995. The Mousterian site of Zafarraya (Andalucia,
Spain): dating and implications on the Palaeolithic peopling
process of Western Europe. C R Acad Sci II 321:931–937.
Lalueza C, Pe
´rez-Pe
´rez A, Chimenos E, Maroto J, Turbo
´nD.
1993. Estudi radiogra
`fic i microsco
`pic de la mandı
´bula de
Banyoles: patologies i estat de conservacio
´. In: Maroto J, edi-
tor. La mandı
´bula de Banyoles en el context dels fossils
humans del Pleistoce
`. Girona: Centre d’Investigacions Arqueo-
lo
´giques. p135–144.
Lorenzo JI, Montes L. 2001. Restes Ne
´andertaliens de la Grotte
de ‘‘Los Moros de Gabasa’’ (Huesca, Espagna). In: Zilha
˜oJ,
Aubry T, Carvalho AF, editors. Les premiers hommes mod-
ernes de la Pe
´ninsule Ibe
´rique. Trabalhos de Arquelogia . Lis-
boa: Instituto Portugue
ˆs de Arqueologia, Vol.17. p77–86.
Maki J, Trinkaus E. 2011. Opponens pollicis mechanical effec-
tiveness in Neandertals and early modern humans. PaleoAn-
thropology 2011:62–71.
Mallegni F. 1995. The teeth and the periodontal apparatus of
the Neandertal mandibles from the Guattari Cave (Monte Cir-
ceo, Lazio, Italy). Z Morphol Anthropol 80:329–351.
Martin LB. 1985. Significance of enamel thickness in hominoid
evolution. Nature 314:260–263.
Martino
´n-Torres M, Bermu
´dez de Castro JM, Go
´mez-Robles A,
Prado-Simo
´m L, Arsuage JL. 2012. Morphological description
and comparison of the dental remains from Atapuerca-Sima
de los Huesos (Spain). J Hum Evol 62:7–58.
Mehta C, Patel N. 1999. StatXact 4.0.1 for Windows. Cambridge
MA: Cytel Software Corp.
Molnar S. 1971. Human tooth wear, tooth function and cultural
variability. Am J Phys Anthropol 34:175–190.
Moorrees CFA, Fanning EA, Hunt EE. 1963. Age variation of
formation stages for ten permanent teeth. J Dent Res
42:1490–1502.
Nagurka ML, Hayes WC. 1980. An interactive graphics package
for calculating cross-sectional properties of complex shapes. J
Biomech 13:59–64.
Niewoehner WA. 2001. Behavioral inferences from the Skhul/
Qafzeh early modern human hand remains. Proc Natl Acad
Sci USA 98:2979–2984.
Ogilvie MD, Curran BK, Trinkaus E. 1989. Incidence and pat-
terning of dental enamel hypoplasia among the Neandertals.
Am J Phys Anthropol 79:25–41.
Olejniczak AJ, Grine FE. 2006. Assessment of the accuracy of
dental enamel thickness measurements using microfocal X-
ray computed tomography. Anat Rec A 288:263–275.
Olejniczak AJ, Smith TM, Feeney RNM, Macchiarelli R, Mazu-
rier A, Bondioli L, Rosas A, Fortea J, de la Rasilla M, Garcia-
Tabernero A, Radovc
ˇic
ˇJ, Skinner MM, Toussaint M, Hublin
JJ. 2008. Dental tissue proportions and enamel thickness in
Neandertal and modern human molars. J Hum Evol 55:12–
23.
Proschan MA, Waclawiw MA. 2000. Practical guidelines for
multiplicity adjustment in clinical trials. Control Clin Trials
21:527–539.
Quam RM, Arsuaga JL, Bermu
´dez de Castro JM, Dı
´ez JC, Lor-
enzo C, Carretero JM, Garcı
´a N, Ortega AI. 2001. Human
remains from Valdegoba Cave (Hue
´rmeces, Burgos, Spain). J
Hum Evol 41:385–435.
Rice WR. 1989. Analyzing tables of statistical tests. Evolution
43:223–225.
Rosas A, Martı
´nez-Maza C, Bastir M, Garcı
´a-Tabernero A,
Lalueza-Fox C, Huguet R, Ortiz JE, Julia
`R, Soler V, de
Torres T, Martı
´nez E, Can
˜averas JC, Sa
´nchez-Moral S,
Cuezva S, Lario J, Santamarı
´a D, de la Rasilla M, Fortea J.
2006. Paleobiology and comparative morphology of a late
Neandertal sample from El Sidro
´n, Asturias, Spain. Proc Natl
Acad Sci USA 103:19266–19271.
Rougier H, Crevecoeur I, Wolpoff MH. 2006. Lower third premo-
lar rotation in the Krapina dental sample. Period Biol
108:269–278.
Ruff CB. 2000. Body size, body shape, and long bone strength in
modern humans. J Hum Evol 38:269–290.
Sarrio
´n I. 2006. Hallazgo de un parietal humano del tra
´nsito
Pleistoceno medio-superior procedente de la Cova del Bolomor,
Tavernes de la Valldigna, Valencia. Arch Prehist Levantina
26:11–23.
Saunders SR, Chan AHW, Kahlon B, Kluge HF, FitzGerald CM.
2007. Sexual dimorphism in the dental tissues in human per-
manent mandibular canines and third premolars. Am J Phys
Anthropol 133:735–740.
Scott GR, Turner CG II. 1997. The anthropology of modern
human teeth. Dental morphology and its variation in recent
human populations. Cambridge, UK: Cambridge University
Press.
Shackelford LL. 2005. Patterns of geographic variation in the
postcranial robusticity of Late Upper Paleolithic humans.
Ph.D. Thesis, Washington University, Saint Louis.
Shackelford LL, Harris AES, Konigsberg LW. 2012. Estimating
the distribution of probable age-at-death from dental remains
of immature human fossils. Am J Phys Anthropol 147:227–
253.
Smith BH. 1984. Patterns of molar wear in hunter-gatherers
and agriculturalists. Am J Phys Anthropol 63:39–56.
Smith TM, Olejniczak AJ, Reh S, Reid DJ, Hublin JJ. 2008.
Enamel thickness trends in the dental arcade of humans and
chimpanzees. Am J Phys Anthropol 136:237–241.
Smith TM, Olejniczak AJ, Zermeno JP, Tafforeau P, Skinner
MM, Hoffmann A, Radovc
ˇic
˙J, Toussaint M, Kruszynski R,
Menter C, Moggi-Cecchi J, Glasmacher UA, Kullmer O,
Schrenk F, Stringer C, Hublin JJ. 2012. Variation in enamel
50 J.C. WILLMAN ET AL.
American Journal of Physical Anthropology

thickness within the genus Homo. J Hum Evol 62:395–
411.
Tompkins RL. 1996. Relative dental development of Upper
Pleistocene hominids compared to human population varia-
tion. Am J Phys Anthropol 99:103–118.
Trinkaus E. 2004. Dental crown dimensions of middle and late
Pleistocene European humans. In: Rubio S, editor. Miscala
´nea
en Homenaje a Emiliano Aguirre III: Paleoantropologı
´a.
Alcala
´de Henares: Museo Arqueolo
´gico Regional. p393–398.
Trinkaus E, Maki J, Zilha
˜o J. 2007. Middle Paleolithic human
remains from the Gruta da Oliveira (Torres Novas), Portugal.
Am J Phys Anthropol 134:263–273.
Walker MJ, Gibert J, Lo
´pez MV, Lombardi AV, Pe
´rez-Pe
´rez A,
Zapata J, Ortega J, Higham T, Pike A, Schwenninger JL, Zil-
ha
˜o J, Trinkaus, E. 2008. Late Neandertals in southeastern
Iberia: Sima de las Palomas del Cabezo Gordo, Murcia, Spain.
Proc Natl Acad Sci USA 105:20631–20636.
Walker MJ, Ortega J, Lo
´pez MV, Parmova
´K, Trinkaus E.
2011a. Neandertal postcranial remains from the Sima de las
Palomas del Cabezo Gordo, Murcia, southeastern Spain. Am J
Phys Anthropol 144:505–515.
Walker MJ, Ortega J, Parmova
´K, Lo
´pez MV, Trinkaus E.
2011b. Morphology, body proportions and postcranial hyper-
trophy of a female Neandertal from the Sima de las Palomas,
southeastern Spain. Proc Natl Acad Sci USA 108:10087–
10091.
Walker MJ, Zapata J, Lombardi AV, Trinkaus E. 2011c. New
evidence of dental pathology in 40,000 year old Neandertals.
J Dent Res 90:428–432.
Walker MJ, Lo
´pez-Martı
´nez MV, Ortega-Rodriga
´n
˜ez J, Haber-
Uriarte M, Lo
´pez-Jime
´nez A, Avile
´s-Ferna
´ndez A, Polo-Cama-
cho JL, Campillo-Boj M, Garcı
´a-Torres J, Carrio
´n-Garcı
´a JS,
San Nicola
´s-del Toro M, Rodrı
´guez-Estrella T. 2012. The exca-
vation of buried articulated Neanderthal skeletons at Sima de
las Palomas (Murcia, SE Spain). Quatern Int 259:7–21.
Zilha
˜o J, Angelucci D, Argant J, Brugal JP, Carrio
´n JS, Car-
valho R, Fuentes N, Nabais M. 2010. Humans and hyenas in
the Middle Paleolithic of Gruta da Oliveira (Almonda karstic
system, Torres Novas, Portugal). In: Baquedano E, Rosell J,
editors. 1
a
Reunio
´n de cientı
´ficos sobre cubiles de hiena (y
otros grandes carnı
´voros) en los yacimientos arqueolo
´gicos de
la Penı
´nsula. Madrid: Museo Arqueolo
´gico Regional. p299–
308.
Zilha
˜o J, Mckinney C. 1995. Uranium-thorium dating of lower
and middle Paleolithic sites in the Almonda karstic system
(Torres Novas, Portugal). In: Actas 3
a
Reunia
˜o do Quaterna
´rio
Ibe
´rico. Coimbra: Universidade de Coimbra. p513–516.
51OLIVEIRA MIDDLE PALEOLITHIC HUMANS
American Journal of Physical Anthropology