Femoral curvature in Neanderthals and modern humans: A 3D geometric morphometric analysis more

Journal of Human Evolution 60 (2011) 540e548 Contents lists available at ScienceDirect Journal of Human Evolution journal homepage: www.elsevier.com/locate/jhevol Femoral curvature in Neanderthals and modern humans: A 3D geometric morphometric analysis Isabelle De Groote a, b a b University College London, Department of Anthropology, Gower Street, London WC1E 6BT, United Kingdom The Natural History Museum, Palaeontology Department, London, SW7 5BD, United Kingdom a r t i c l e i n f o Article history: Received 24 March 2010 Accepted 23 September 2010 Keywords: Neanderthals Homo sapiens Postcrania Curvature Geometric morphometrics a b s t r a c t Since their discovery, Neanderthals have been described as having a marked degree of anteroposterior curvature of the femoral shaft. Although initially believed to be pathological, subsequent discoveries of Neanderthal remains lead femoral curvature to be considered as a derived Neanderthal feature. A recent study on Neanderthals and middle and early Upper Palaeolithic modern humans found no differences in femoral curvature, but did not consider size-corrected curvature. Therefore, the objectives of this study were to use 3D morphometric landmark and semi-landmark analysis to quantify relative femoral curvature in Neanderthals, Upper Palaeolithic and recent modern humans, and to compare adult bone curvature as part of the overall femoral morphology among these populations. Comparisons among populations were made using geometric morphometrics (3D landmarks) and standard multivariate methods. Comparative material involved all available complete femora from Neanderthal and Upper Palaeolithic modern human, archaeological (Mesolithic, Neolithic, Medieval) and recent human populations representing a wide geographical and lifestyle range. There are significant differences in the anatomy of the femur between Neanderthals and modern humans. Neanderthals have more curved femora than modern humans. Early modern humans are most similar to recent modern humans in their anatomy. Femoral curvature is a good indicator of activity level and habitual loading of the lower limb, indicating higher activity levels in Neanderthals than modern humans. These differences contradict robusticity studies and the archaeological record, and would suggest that femoral morphology, and curvature in particular, in Neanderthals may not be explained by adult behavior alone and could be the result of genetic drift, natural selection or differences in behavior during ontogeny. Ó 2010 Elsevier Ltd. All rights reserved. Introduction When the Feldhofer Neanderthal remains were discovered in the 19th century, researchers noted a marked degree of anterior curvature of the femoral shaft and ascribed it to pathology (Klaatsch, 1901; Boule, 1908; Trinkaus and Shipman, 1993). With the subsequent discoveries of other Neanderthal remains, femoral curvature was considered to be a derived feature of Neanderthals (Klaatsch, 1901; Trinkaus and Shipman, 1993; Churchill, 1998; Golovanova et al., 1999; Czarnetzki, 2000; Weaver, 2003; Yamanaka et al., 2005). Some scholars suggested that Neanderthal curvature is the result of rickets (Ivanhoe, 1970; Ivanhoe and Trinkaus, 1983; Czarnetzki, 2000) or osteomalacia (Czarnetzki, 2000). Because there is no widespread evidence of Neanderthals eating fish (with the exception of shell fish consumption at Gibraltar) (Hockett and Haws, 2005) and because they lived in the northern regions of E-mail address: i.degroote@nhm.ac.uk. 0047-2484/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.jhevol.2010.09.009 Europe, Ivanhoe suggests Neanderthals experienced an insufficient amount of vitamin D in their diet and as a consequence of rickets show skeletal deformities, such as abnormal long bone curvature (Ivanhoe, 1970; Ivanhoe and Trinkaus, 1983; Czarnetzki, 2000). However, the curvature observed in Neanderthals is an accentuation of normal anteroposterior curvature of the diaphysis (Steinbock, 1976) and never assumes the irregular mediolateral curvature associated with rickets (Ivanhoe and Trinkaus, 1983). Neither does rickets explain the observed variation in anterior curvature between modern human populations. Relatively little work has been done to quantify diaphyseal curvature in Neanderthals, but the long claimed distinction in degree of femoral curvature in Neanderthals was challenged by Shackelford and Trinkaus (2002). Shackelford and Trinkaus (2002) suggested that Neanderthals were indistinguishable from Middle Palaeolithic and middle and early Upper Palaeolithic early modern humans in their degree of absolute anterior curvature. Additionally, the Neanderthals and the Middle Palaeolithic individuals from Qafzeh and Skhul were found to exhibit a more distal apex of I. De Groote / Journal of Human Evolution 60 (2011) 540e548 541 curvature (point of maximum curvature) compared with more recent populations (Shackelford and Trinkaus, 2002). They suggested that this could be correlated with measures of bone hypertrophy or an overall decrease in lower-limb robusticity during the Middle to Upper Palaeolithic. The five regional groups (Euroamericans, African-Americans, North and South American Amerindians, Inuits and Japanese) were divided into urban vs. non-urban and were significantly different in femoral curvature. Shackelford and Trinkaus (2002) suggested that the overall decrease in femoral curvature in modern humans was due to a decrease in long-distance mobility. Research from forensic anthropology also suggests that significant differences exist in femoral curvature among modern human populations (Stewart, 1962; Walensky, 1962, 1965; Gilbert, 1975, 1976; Trudell, 1999). Initial studies demonstrated the diagnostic value of femoral curvature in distinguishing among Native American, African-American and Caucasoid American populations (Stewart, 1962; Walensky, 1962, 1965; Gilbert, 1975, 1976; Trudell, 1999). When the research was expanded by increasing the number of populations, no relationship was found among femoral curvature, habitual behavioral patterns and latitudinal position of those populations (Stewart, 1962; Walensky, 1962, 1965; Gilbert, 1975, 1976; Trudell, 1999). The more detailed characterization of curvature with 3D morphometrics has been shown to refine the differences among modern human groups further (De Groote, 2008). When De Groote (2008) compared modern human femora from a geographical and behaviorally diverse human sample, she hypothesized that femoral curvature is related to habitual activity patterns. The highest levels of curvature for the femur were found in populations with the highest terrestrial activity levels. Femoral curvature follows different trends from robusticity, suggesting that it is not necessarily a response to the same loading regime (Ruff et al., 1993, 1994; Trinkaus et al., 1994, 1999; Pearson, 2000; Ruff and Trinkaus, 2000; Shackelford and Trinkaus, 2002; Stock, 2002, 2006; Stock and Pfeiffer, 2004; Carlson et al., 2007; Shackelford, 2007). For the femur, which is loaded proximodistally, curvature facilitates muscle packing (Lanyon, 1980), and curvature may be a compromise between bone strength and predictability of bending strains and material failure (Lanyon, 1980, 1987; Bertram and Biewener, 1988; De Groote, 2008). Because femoral curvature is unrelated to climate (latitude in this analysis), it was suggested that curvature may ultimately be a better indicator of activity levels than crosssectional measures of long bone robusticity. This study tests the hypothesis proposed by Shackelford and Trinkaus (2002) that there is no difference between Neanderthals and modern humans using a new 3D geometric method to quantify anterior femoral curvature (De Groote et al., 2010). There are three main external influences that need to be considered when interpreting the functional meaning of curvature for early modern humans and Neanderthals, which is known to show a wide range of intraspecific variation in modern humans. The first is the effect of body size on curvature. Most mammals show positive allometry with curvature of the femur (Swartz, 1990). Ruff et al. (1997) proposed that Neanderthals are on average 30% larger than recent humans and that early modern humans are about 10% larger than recent modern humans (Ruff et al., 1997). If curvature is related to body mass, it is predicted that Neanderthals will have higher degrees of curvature than both early and recent modern humans. Within modern humans, however, De Groote (2008) found no correlation between body mass and curvature, nor was this noted by Shackelford and Trinkaus (2002) and thus this is not explored further in this study. The second influence that needs to be investigated is the effect of habitual behavior on curvature. Femoral curvature has been suggested to negatively affect the strength of the bone (Bertram and Biewener, 1988), but also to allow for greater muscle packing and placing the muscle vector more parallel to the diaphyseal axis (Lanyon and Bourn, 1979). This is supported by the concavity of the radius and tibia of many mammals with respect to the flexor musculature, allowing for greater volume (Lanyon et al., 1979; Lanyon, 1980). Increased femoral musculature in humans due to increased activity may therefore increase anteroposterior shaft curvature. Modern humans and Neanderthals most likely did not differ in their subsistence strategies and were probably both hunting and scavenging (Lieberman, 1989; Bar-Yosef, 2004; Pearson et al., 2006). Although there may have been differences in hunting practices (Speth and Tchernov, 1998; Marean and Assefa, 1999, 2005), their resource acquisition and overall workload involved high activity levels, and this is apparent in the similarities in their post-crania, such as similar levels of robusticity (Lieberman, 1989; Trinkaus et al., 1989; Hublin et al., 2006). If curvature is a response to activity levels in human populations, it is predicted that Neanderthals, having high activity levels, will display similar levels of degree of curvature to early modern humans and other hunteregatherers. Within modern humans, individuals and populations with lower activity levels exhibited lower degrees of curvature (De Groote, 2008). Thirdly, it is necessary to consider the effect of climate on curvature. Many of the distinctive Neanderthal postcranial features are considered the consequence of a hyperpolar body form (Hublin, 1989; Ruff, 1991; Weaver, 2003; Weaver and Steudel-Numbers, 2005). Neanderthals, being reported as “hyper-polar” (Weaver, 2003), would be predicted to have a higher degree of curvature, than any modern human population. In Neanderthals and modern humans alike, it is expected that if there were a strong effect of climate on curvature, that this would have been established in the population genetically rather than only through individual ontogeny as climatic adaptations in humans are known to become genetically established adaptations over time (Mayr, 1956). The lack of correlation between climate and increased femoral curvature in modern humans, however, suggests that anterior femoral curvature is not a response to cold climate (De Groote, 2008), and can therefore be omitted as a possible explanation for anterior femoral curvature in Neanderthals and early modern humans. If Neanderthals are distinct in their anterior femoral curvature from early modern humans, and early modern humans resemble recent modern humans more than they do Neanderthals (Trinkaus and Shipman, 1993; Churchill, 1998; Golovanova et al., 1999; Shackelford and Trinkaus, 2002; Weaver, 2003; Yamanaka et al., 2005) this would reject the hypothesis by Shackelford and Trinkaus (2002) that Neanderthals and middle and early Upper Palaeolithic modern humans had similar activity levels. It would indicate that Neanderthals were more active than early and recent modern humans or that curvature was a genetically determined character, which is relevant to the current debate surrounding the phylogenetic relationship between Neanderthals and modern humans. Materials and methods The skeletal sample used in this study was divided into three groups: recent modern humans, early modern humans and Neanderthals. The recent modern human sample comes from a geographically and behaviorally diverse collection of recent modern humans (Figure 1). This sample consists of 421 individuals of both historic and pre-historic populations, and was used to capture the range of recent modern human variation as accurately 542 I. De Groote / Journal of Human Evolution 60 (2011) 540e548 Figure 1. Map with recent human sample of 102 adult males, 89 adult females and 230 unsexed adults from behaviorally and geographically diverse populations (sizes: n ¼ 2 to n ¼ 25) from Mesolithice19th century. as possible (Table 1). The sample included males, females and unsexed individuals, and ranged from young (epiphyseal sutures of the femur closed but visible) to old adults. The better preserved of the left or right femur was selected. The Neanderthal sample consisted of all available complete femora. There were six specimens included in the sample. The early modern human sample consisted of ten specimens (Table 2) from the early and middle Upper Palaeolithic. Although this sample is partly different from that used by Shackelford and Trinkaus (2002) due to the nature of the data collected, they cover broadly the same time-period. Both left and right femora were included or casts where the original was unavailable or too fragile. The curvature of the femoral diaphysis was quantified using geometric morphometrics. The relevant analytical approaches Table 1 List of modern humans in the sample. Population African American Alaskan Aleut Alaskan Native Andaman Arizona Australian Bantu Belgian Medieval Belgian Mesolithic Belgian Neolithic British Neolithic Chinese Colorado Native Czech Medieval Danish Medieval Danish Neolithic Egyptian English Medieval English Urban N 14 13 13 14 18 11 1 24 1 22 2 9 2 25 15 19 5 16 21 Population French Medieval English Urban French Medieval French Neolithic Greenland Inuit Kazach Medieval Khoikoi Lapland Natufian New Mexico Ohio Peru Pygmy Russian Eskimo Russian Mesolithic Siberia South Dakota Tasmanian Tierra del Fuego N 13 21 13 19 14 6 10 17 2 9 13 11 4 14 11 16 15 2 2 have been developed by a number of authors and are summarized in O’Higgins (2000) and Gunz et al. (2005), and recently the advantages over linear measurements for the quantification of femoral curvature have been demonstrated (De Groote et al., 2010). Semi-landmarks make it possible to include point and outline Table 2 List of Neanderthals and early modern humans. Neanderthal Original Europe Spy 2 righta La Ferrassie 2 leftb La Chapelle aux Saints rightb Early modern human Original Europe Chancelade rightf Combe Capelle rightf Western Asia Sungir rightg Dolni Vestonice 13 righth Dolni Vestonice 14 righth Dolni Vestonice 16 lefth Levant Ohalo II rightc Qafzeh 9 rightc Cast Europe St. Germain righti Western Asia Kostienki 14 rightj a b c d e f g h i j Cast Europe Le Moustier leftd La Ferrassie 1 left b La Ferrassie 2 right b Neanderthal right e Royal Belgian Intitute for Natural Sciences, Brussels. Musee de l’Homme, Paris. Tel Aviv University. Museum für Vor- und Frühgeschichte in Berlin. Rheinisches Museum in Bonn. Musee du Périgueux. Laboratory for reconstruction, Moscow.  Dolní Vestonice. Musée National du Prehistoire. Kunstcamera St Petersburg. I. De Groote / Journal of Human Evolution 60 (2011) 540e548 Table 3 Definitions of univariate measurements. 543 Maximum length measured along the biomechanical axis.(biomech axis: where the most superior point of the head of the femur and the most lateral point of the greater trochanter describe a 90 angle, the perpendicular line down from the most superior point of the head to the most inferior point on the medial condyle). Neck-shaft angle (Martin N 29) Also collo-diaphyseal angle. Martin N 29. The angle described by the shaft-axis (going through the middle of the shaft) and the neck-axis (going through the middle of the neck) The angle of femoral torsion is the angle made by the axis of the femoral neck with the tangent of the posterior surface of the Torsion (Martin N 28) femoral condyles. Midshaft ratio AP diameter 50%/ML diameter 50%  100 where the midshaft anteroposterior diameter (Martin N 10) and the mediolateral diameter (Martin N 8) is taken at the 50% level. The 0% shaft level is defined as the most inferior edge of the medial condyle; the 100% is the most superior point of the head of the femur. Subpilastric ratio AP diameter 25%/ML diameter 25%  100 where the subpilastric anteroposterior diameter (Weaver, 2003) and the mediolateral diameter (Weaver, 2003) is taken at the 25% level. The 0% shaft level is defined as the most inferior edge of the medial condyle; the 100% is the most superior point of the head of the femur. Epiphyseal breadth ratio Maximum condylar width/length  100 where Maximum condylar width (Martin N 21) is the distance between the point where the medial epicondyle projects maximally (local maximum of a curved surface) and the point where the lateral epicondyle projects maximally (local maximum of a curved surface) Neck length ratio Neck length/length  100 where length of the head-neck axis (Martin N 14) is the length of the axis from the most medial point of the head to the middle of the intertrochanteric line. Midshaft robusticity index AP diameter 50% þ ML diameter 50%/length  100 where the midshaft anteroposterior diameter (Martin N 10) and the mediolateral diameter (Martin N 8) is taken at the 50% level. The 0% shaft level is defined as the most inferior edge of the medial condyle; the 100% is the most superior point of the head of the femur. Head robusticity index SI head diameter þ AP head diameter/length  100 where Head diameter (Martin N 18) is the maximum superoinferior diameter of the femoral head on the edge of the articular surface and the anteroposterior head diameter (Martin N 19) is the maximum anteroposterior diameter of the femoral head on the edge of the articular surface. Femur length (Martin N 1) information in a single analysis and to consider the curves separately or as part of the whole bone morphology. Landmarks and semi-landmarks were collected using a Microscribe 3DX digitizer (Immersion Corporation), a laptop computer, Microsoft Excel and Microscribe Utility Software v.4.0 (MUS v. 4.0). The semi-landmarks were reduced and equidistantly spaced along the diaphysis in MathematicaÒ to create a dataset with a homologous configuration of landmarks and semi-landmarks for each individual (see details in De Groote et al., 2010). After this registration procedure, the configurations were exported into Morphologika 2 (O’Higgins and Jones, 1998) for further analysis. The 3D configurations were analyzed using general procrustes analysis (GPA; also referred to as GLS: generalised least squares) and principal component analysis (Bookstein, 1991; Adams et al., 2004). The principal component scores for each individual for each principal component were used as data in univariate and multivariate analyses and combined with other variables (Bookstein, 1991; Adams et al., 2004; Gunz et al., 2005; Slice, 2005). Some linear measurements were calculated from the landmark configuration and some of those measurements were standardized by size by the calculation of ratios or indices (Table 3) multiplied by 100 to facilitate comparisons. Using indices eliminates the effect of scale on the measurement, although allometric effects are not estimated due to the lack of allometry within modern humans. Also, in the principal component analysis there was no correlation between the principal components and centroid size. ANOVA was used to determine the effect of group membership (Neanderthal, early modern human or recent modern human) influencing curvature. Post-hoc tests were performed to identify differences between the samples. Both a Hochberg’s GT2 (for very different sample sizes, Field, 2000) and a GameseHowell procedure (for small and uneven sample sizes where homogeneity of variance is not assumed for all samples, Field, 2000) were used in SPSS v.15. Only when the result was significant for both post-hoc procedures, was it considered significant. Canonical linear discriminant functions were calculated using the principal component scores and univariate measurements for Neanderthals, early modern humans and recent modern humans. The recent modern human dataset was reduced by using the mean for each population. This resulted in 36 population means used in the discriminant function analysis. Only the two most important principal components were considered for inclusion (Dytham, 1999; Weaver, 2002). Results The changes for anterior femoral curvature along each principal component are visualized using Morphologika 2. The figures presented correspond to the most extreme positive and negative individuals on the scale for each PC. The curves are made up of eight proportional semi-landmarks combined with two landmarks at the end of the curves. Viewing angles were chosen to illustrate similarities and differences most clearly. The curves are shown in lateral view unless otherwise stated. The first three principal components explain 63.7%, 9.62%, and 7.30% of the variance, respectively, (total 80.06%). Subsequent PCs explain minimal amounts of the variation and are not considered further. The first principal component (acurvePC1) reflects variation in degree of anterior curvature or subtense (Figure 2a). The second principal component (acurvePC2) is related to the position of the apex of curvature (Figure 2b). The third principal component (acurvePC3) is the shape of the shaft in anterior view (Figure 2c). Curvature The groups are significantly different for curvature (acurveAllPC1 df ¼ 2; F ¼ 17.933; P 0.001). Neanderthals have the highest degree of anterior curvature, followed by early modern humans. Recent modern humans are the straightest (Figure 3). Statistically, Neanderthals are different from both early and recent modern humans although there is significant overlap in their ranges. Apex of curvature The groups are significantly different for the position of the apex of curvature on the anterior surface (acurveAllPC2 d.f. ¼ 2; F ¼ 5.130; P 0.01). Neanderthals have the lowest apex of curvature and are significantly different from early and recent modern humans (Figure 4). There is no significant difference between the groups for sinusoidal shape of the anterior surface of the femoral shaft (acurvePC3). 544 I. De Groote / Journal of Human Evolution 60 (2011) 540e548 Univariate measurements The groups are significantly different for all univariate measurements (Table 4). The highest F-scores are for head-robusticity (d.f. ¼ 2; F ¼ 16.429; P 0.01), neck-length (d.f. ¼ 2; F ¼ 18.568; P 0.01), neck-shaft angle (d.f. ¼ 2; F ¼ 8.621; P 0.01) and robusticity index (d.f. ¼ 2; F ¼ 8.824; P 0.01). Post-hoc tests indicate that Neanderthals have relatively the largest femoral head, longest neck and largest distal epiphyses (d.f. ¼ 2; F ¼ 8.155; P 0.01) compared with early and recent modern humans, although their midshaft robusticity and neck-shaft angle is comparable with that of early modern humans. Early modern human femora are longer and have lower torsion angles (d.f. ¼ 2; F ¼ 4.420; P 0.015), are more robust, and have higher midshaft (d.f. ¼ 2; F ¼ 3.186; P 0.05) and subpilastric ratios (d.f. ¼ 2; F ¼ 4.399; P 0.02) than do recent modern humans. Early modern human femora have a high midshaft shape ratio, which probably reflects the strong expression of the linea aspera. Neanderthal femora have an almost round shaft at the midshaft level and lack a clear linea aspera. Discriminant function analysis A DFA with cross-validation using the PCs for anterior curvature and apex of curvature and univariate measurements used in the analyses above was used to separate Neanderthals, early and recent modern humans. Function 1 separates best between Neanderthals and modern humans in general, whereas function 2 separates early modern humans from recent modern humans (Figure 5). The variables in Table 5 appear in the order of their discriminating power. Function 1 explains 86.1% of the variance and reflects (ordered according to decreasing correlation between the variable and the function) relative size of the femoral head, degree of anterior curvature, high neck-length ratio, width of the distal Figure 2. Morphological trends for the anterior curve of the femur for Neanderthals, early and recent modern humans. (a) Principal Component 1: lateral view. Negative values are less curved, positive values are more curved. (b) Principal Component 2: lateral view. Individuals with negative values have a more proximal apex of curvature, whereas those with positive values have a more distal apex of curvature. (c) Principal Component 3: anterior view. Negative values are the straightest, whereas positive values indicate a mediolaterally sinusoidal shape. Positive and negative visualisations correspond to the most extreme positive and negative scores for each PC. Figure 3. The anterior curve of the femur for Neanderthals, early and recent modern humans. (Horizontal line ¼ mean, Box ¼ 2 S.D., whiskers: range). The higher values for Neanderthals indicate that they are more curved than the modern humans. I. De Groote / Journal of Human Evolution 60 (2011) 540e548 545 Figure 4. The anterior apex of curvature of the femur for Neanderthals, early and recent modern humans. (Horizontal line ¼ mean, Box ¼ 2 S.D., whiskers: range). The higher value for Neanderthals indicates a lower apex of curvature. epiphyses, low neck-shaft angle and a high subtrochanteric shape ratio. Function 2 explains the remaining 13.9% of variance and reflects a low midshaft and subpilastric shaft shape, low robusticity and a high apex of curvature (Table 5). In the original classification results, 96% was correctly classified. Of the fossils, only Kostienki 14 was classified as a recent modern human whereas the two individuals from the British Neolithic were Table 4 ANOVA results and descriptives for Neanderthals, early and recent modern humans and the univariate measurements of the femur. F Femur length, mm Neck-shaft angle,  Torsion angle,  Midshaft shape ratio Subpilastric shape ratio Epiphyseal Br ratio 8.621 <0.001a Sig. Neanderthal Early Homo sapiens Recent Homo sapiens Neanderthal Early Homo sapiens Recent Homo sapiens Neanderthal Early Homo sapiens Recent Homo sapiens Neanderthal Early Homo sapiens Recent Homo sapiens Neanderthal Early Homo sapiens Recent Homo sapiens Neanderthal Early Homo sapiens Recent Homo sapiens Neanderthal Early Homo sapiens Recent Homo sapiens Neanderthal Early Homo sapiens Recent Homo sapiens Neanderthal Early Homo sapiens Recent Homo sapiens N 8 13 428 8 13 428 8 13 428 8 13 428 8 13 428 8 13 428 8 13 428 8 13 428 8 13 428 Mean 430.25 456.14 426.52 118.68 124.27 127.41 10.43 11.17 16.73 103.02 128.38 114.16 87.63 102.06 88.08 18.87 17.12 17.11 15.85 13.98 13.87 13.66 13.44 12.41 22.35 18.72 18.54 S.D. 32.06 34.17 34.18 5.21 7.63 5.71 14.87 9.02 6.91 14.49 20.95 19.11 9.83 18.80 15.73 1.39 1.18 1.33 2.62 1.09 1.07 1.01 0.93 1.15 1.00 1.37 1.65 classified as early modern human. The DFA with cross-validation was able to correctly classify Neanderthals and recent modern humans relatively successfully with 83.6% (five out of six Neanderthals) and 97.2% (35 out of 36 modern human populations) classified correctly. Early modern humans were classified correctly in 60% of cases (six out of ten), with three considered recent modern human and one considered Neanderthal. Overall, for the three groups together, this gives 88.5% of correct cross-validated classification. 4.420 0.013 3.186 0.042 4.399 0.013 8.155 <0.001a Neck length ratio 18.568 <0.001a Robusticity index 8.824 <0.001a Head robusticity 16.429 <0.001a a Significant at a ¼ 0.05. Figure 5. Canonical Linear Discriminant Function 1 and 2 for Neanderthals, early and recent modern humans. 546 Table 5 Discriminant function coefficients. Function 1 Head robusticity Anterior curvature Neck length ratio Epiphyseal breadth ratio Neck shaft angle Subtrochanteric shape ratio Subpilastric shape ratio Midshaft shape ratio Apex of curvature Robusticity index a I. De Groote / Journal of Human Evolution 60 (2011) 540e548 Function 2 0.355 À0.161 0.084 0.250 À0.001 À0.129 À0.480a À0.441a 0.419a À0.377 0.450a 0.416a 0.348a 0.274a À0.259a 0.157a 0.016 À0.050 0.153 0.208 Largest absolute correlation variable and discriminant function. Discussion The goal of this research was to investigate the differences and similarities in anterior femoral curvature between Neanderthals and modern humans, and to consider these differences in the light of the rest of their femoral morphology. The study found that Neanderthals are distinct from both early and recent modern humans in that they exhibit a higher degree of anterior femoral curvature. De Groote (2008) suggests that in recent modern humans, femoral curvature is a plastic feature that responds to loading of the femur during activity. This supports the hypothesis by Shackelford and Trinkaus (2002) that populations with high activity levels have a high degree of femoral curvature. De Groote (2008) also found that there is a relationship between curvature and robusticity in recent modern humans and concluded that anteroposterior widening of the shaft, a low apex of curvature and high degree of curvature are the result of repetitive loading on the lower limb during subsistence strategy-related terrestrial mobility (Ruff et al., 1994, Ruff, 1987; Larsen et al., 1995; Holt, 2003; Stock and Pfeiffer, 2004), and this hypothesis is supported by the strength circularity indices at the femoral midshaft and their strong correspondence with terrestrial mobility (Stock, 2006). Because of the correlation between subsistence-related activity and curvature in recent modern humans, the prediction was that Neanderthals, being hunteregatherers, would have a high degree of femoral curvature. Moreover, their curvature should be comparable with that of early modern humans because the two groups had broadly similar lifestyles (Trinkaus et al., 1989). Early modern humans and Neanderthals most likely did not differ in their subsistence strategies and were both hunting and scavenging (Lieberman, 1989; Sorensen and Leonard, 2001; BarYosef, 2004; Pearson et al., 2006). Faunal assemblages from occupation and butchery sites show that both groups had early access to animals, and cut-mark patterns indicate a primary reliance on hunting rather than scavenging (Speth and Tchernov, 1998). Trinkaus and Zimmerman (1982) and Klein (2003) have argued that Middle Stone Age people were less adept hunters because they only hunted a few of the available species and that Neanderthals show a high incidence of skeletal trauma because of the risk involved in close range hunting (Trinkaus and Zimmerman, 1982; Berger and Trinkaus, 1995; Klein, 2003). Recent investigations of faunal assemblages have shown that some Neanderthal sites may be dominated by a single prey species, but this is also documented among some modern hunteregatherer societies (Marean and Assefa, 1999, 2005). The reliance on meat for Neanderthals and early modern humans living in temperate and cold regions, such as Europe and Western Asia, was important for survival. Neanderthals and early modern humans living in Europe during Glacial times must have relied on frequent meat acquisition for their diet, as it is likely that plant foods would have been unavailable for consumption during parts of the year. This is confirmed in stable-isotope analyses from sites such as Vindija Cave, Croatia; Scladina, Spy and Engis in Belgium, and Marillac and Saint-Césaire in France (Fizet et al., 1995; Richards et al., 2000, 2001; Bocherens et al., 2001, 2005; Drucker and Bocherens, 2004). Marean and Assefa (1999) argue that the Middle Palaeolithic Neanderthals may not have been less adept hunters than their modern human contemporaries. Instead, they might have been less adept at using and processing carcasses in order to render higher caloric yields, such as fat rendering and storage, which put them at a subtle disadvantage in comparison with modern humans. These disadvantages are not only the lower caloric intake per prey animal, but also the increased personal risk because of more frequent hunting (Marean and Assefa, 1999). This low return on time expended may have resulted in moderately higher activity and mobility levels in Neanderthals compared with early anatomically modern humans. Similarities in lifestyle and subsistence pattern between Neanderthals and the earliest modern humans is also apparent in the archaeological record, where similar species of large animals are found in both Neanderthal and early modern human deposits. Neanderthals were effective hunters (Speth and Tchernov, 1998) and some consider them a top predator in the environment in which they lived (Bocherens et al., 2005). They also hunted in a given region for a longer period of time than modern humans, who were more seasonally mobile (Lieberman, 1989). Although there is some variation overall, Neanderthals and early modern humans were likely very similar in terms of mobility, resource acquisition and overall workload, and this is apparent in their postcranial anatomy (Lieberman, 1989; Hublin et al., 2009). When corrected for size and body proportions, Neanderthals have lower limb bones that were similar in cross-sectional strength to those of modern humans (Trinkaus et al., 1989; Trinkaus and Ruff, 1999a,b; Hublin et al., 2006). This is also reflected in the results on robusticity presented here, which showed no significant differences between robusticity levels of the shaft between Neanderthals and early modern humans. In degree of femoral curvature, however, Neanderthals show a significantly higher degree of curvature and a lower apex of curvature compared with both early and recent modern humans as well as other differences highlighted in the discriminant analysis. The Neanderthal femur is characterized by a range of features that put them apart from modern humans. They have a relatively large femoral head reflecting their high body mass. The relatively longer femoral neck would have given them a more advantaged abductor position (Wolpoff, 1978). Their wide distal articulation reflects the large knees of the Neanderthals, which have been described in detail by Trinkaus and Rhoads (1999) and hypothesized to offer knee extensor mechanical advantage. The lower neck shaft angle of the Neanderthals compared with both early and recent modern humans was also the subject of previous studies and has been found to be the result of elevated mechanical stress levels during ontogeny (Trinkaus, 1993; Anderson and Trinkaus, 1998; Weaver, 2003). These differences would either suggest that despite the archaeological record and the robusticity results (Trinkaus et al., 1989) indicating otherwise, that Neanderthals had higher activity levels or that their overall femoral morphology is the result of other factors such as natural selection in response to the physical environment of this population or genetic drift. Conclusion Neanderthals have femora with a higher degree of anterior femoral curvature than do early modern humans and recent I. De Groote / Journal of Human Evolution 60 (2011) 540e548 547 modern humans. They also have the most distal apex of curvature. They have wide distal epiphyses, large femoral heads, low neckshaft angles (only compared with early modern humans) and are the most robust (significantly different from recent modern humans only). Discriminant function classification successfully distinguished Neanderthals from the recent modern human group and the fossil modern humans. Neanderthals and early modern humans had broadly similar hunteregatherer lifestyles, and their postcranial skeleton was likely subject to the same stresses as modern humans. Neanderthals show a high degree of femoral curvature, as well as other characteristics considered to facilitate activity, reflecting their active lifestyles. Early modern humans display a high degree of femoral curvature but, contrary to Neanderthals, one that is well within the range of variation of modern humans. Although there may have been some variation in the specific subsistence-related activities they performed, there is a widely held view that Neanderthals and early modern humans had similar lifestyles and activity levels. Therefore, the higher degree of femoral curvature in Neanderthals may not only be explained by behavior alone. From a taxonomic and phylogenetic perspective, Neanderthals are distinct in their expression of curvature compared with modern humans, but it remains to be investigated whether the low degree of curvature is a derived recent human trait, or whether a marked degree of curvature is an autapomorphy of Neanderthals. It has also been suggested that certain differences in morphology between Neanderthals and recent modern humans are the result of behavior during ontogeny (Trinkaus, 1993). In order to investigate this further it is necessary to expand this study to an ontogenetic sample. References Adams, D.C., Rohlf, F.J., Slice, D.E., 2004. Geometric morphometrics: ten years of progress following the ‘revolution’. Ital. J. Zool. 71, 5e16. Anderson, J.Y., Trinkaus, E., 1998. Patterns of sexual, bilateral and interpopulational variation in human femoral neck-shaft angles. J. Anat. 192, 279e285. Bar-Yosef, O., 2004. Eat what is there: hunting and gathering in the world of Neanderthals and their neighbours. Int. J. Osteoarchaeol. 14 (3e4), 333e342. Berger, L.R., Trinkaus, E., 1995. Patterns of trauma among the Neandertals. J. Archaeol. Sci. 22, 841e852. Bertram, J.E., Biewener, A.A., 1988. Bone curvature: sacrificing strength for load predictability? J. Theor. Biol. 131 (1), 75e92. Bocherens, H., Billiou, D., Mariotti, A., Toussaint, M., Patou-Mathis, M., Bonjean, D., Otte, M., 2001. New isotopic evidence for dietary habits of Neanderthals from Belgium. J. Hum. Evol. 40 (6), 497e505. Bocherens, H., Drucker, D.G., Billiou, D., Patou-Mathis, M., Vandermeersch, B., 2005. Isotopic evidence for diet and subsistence pattern of the Saint-Cesaire I Neanderthal: review and use of a multi-source mixing model. J. Hum. Evol. 49 (1), 71e87. Bookstein, F., 1991. Morphometric Tools for Landmark Data. Cambridge University Press, Cambridge. Boule, M., 1908. L’homme fossile de la Chapelle-aux-Saints (Corrèze). Anthropologie 19, 519e525. Carlson, K.J., Grine, F.E., Pearson, O.M., 2007. Robusticity and sexual dimorphism in the postcranium of modern hunter-gatherers from Australia. Am. J. Phys. Anthropol. 134 (1), 9e23. Churchill, S.E., 1998. Cold adaptation, heterochrony and Neandertals. Evol. Anthropol. 7 (2), 46e60. Czarnetzki, A., 2000. The significance of pathological changes for judging the morphology of classical Neanderthals. In: Orschiedt, J., Weniger, G.-C. (Eds.), Neanderthals and Modern Humans e Discussing the Transition: Central and Eastern Europe from 50.000e30.000 B.P. Neanderthal Museum, Mettmann, pp. 296e302. De Groote, I., 2008. A Comprehensive Analysis of Long Bone Curvature in Neanderthals and Modern Humans Using 3D Morphometrics. Ph.D. dissertation, University College, London. De Groote, I., Lockwood, C.A., Aiello, L.C., 2010. A new method for measuring long bone curvature using 3D landmarks and semilandmarks. Am. J. Phys. Anthropol. 141 (4), 658e664. Drucker, D.G., Bocherens, H., 2004. Carbon and nitrogen stable isotopes as tracers of change in diet breadth during Middle and Upper Palaeolithic in Europe. Int. J. Osteoarchaeol. 14 (3e4), 162e177. Dytham, C., 1999. Choosing and Using Statistics: a Biologist’s Guide. Blackwell Science Ltd, Malden. Field, A., 2000. Discovering Statistics Using SPSS. Sage Publications Ltd, London. Fizet, M., Mariotti, A., Bocherens, H., Lange-Badre, B., Vandermeersch, B., Borel, J.P., Bellon, G., 1995. Effect of diet, physiology and climate on carbon and nitrogen stable isotopes of collagen in a late pleistocene anthropic palaeoecosystem: Marillac, Charente, France. J. Archaeol. Sci. 22 (1), 67e79. Gilbert, B.M., 1975. Anterior femoral curvature. Am. J. Phys. Anthropol. 42 (2), 303. Gilbert, B.M., 1976. Anterior femoral curvature e probable basis and utility as a criterion of racial assessment. Am. J. Phys. Anthropol. 45 (3), 601e604. Golovanova, L.V., Hoffecker, J.F., Kharitonov, V.M., Romanova, G.P., 1999. Mezmaiskaya cave: a Neanderthal occupation in the Northern Caucasus. Curr. Anthropol. 40 (1), 77e86. Gunz, P., Mitteroecker, P., Bookstein, F., 2005. Semi-landmarks in three dimensions. In: Slice, D.E. (Ed.), Modern Morphometrics in Physical Anthropology. Kluwer Academic, New York, pp. 73e98. Holt, B.M., 2003. Mobility in Upper Paleolithic and Mesolithic Europe: evidence from the lower limb. Am. J. Phys. Anthrop. 122 (3), 200e215. Hockett, B., Haws, J.A., 2005. Nutritional ecology and the human demography of Neanderthal extinction. Quatern. Int. 137 (1), 21e34. Hublin, J.J., 1989. Climatic changes, paleogeography, and the evolution of the neanderthals. In: Akazawa, T., Aoki, K., Bar-Yosef, O. (Eds.), Neandertals and Modern Humans in Western Asia. Plenum Press, New York, pp. 295e310. Hublin, J.-J., Harvati, K., Harrison, T., Pearson, O., Cordero, R., Busby, A., 2006. How different were Neanderthals’ habitual activities? A comparative analysis with diverse groups of recent humans. In: Delson, E., MacPhee, R.D.E. (Eds.), Neanderthals Revisited: New Approaches and Perspectives. Springer, Netherlands, pp. 135e156. Hublin, J.-J., Richards, M.P., Adler, D.S., Bar-Oz, G., 2009. Seasonal patterns of prey acquisition and inter-group competition during the middle and upper Palaeolithic of the Southern Caucasus. In: Delson, E., MacPhee, R.D.E. (Eds.), The Evolution of Hominin Diets. Springer, Netherlands, pp. 127e140. Ivanhoe, F., 1970. Was Virchow right about Neanderthal? Nature 227 (5258), 577e579. Ivanhoe, F., Trinkaus, E., 1983. On cranial deformation in Shanidar 1 and 5. Curr. Anthropol. 24 (1), 127e128. Klaatsch, H., 1901. Das Gliedmaßenskelett des Neanderthalmenschen. Anat. Anz. 19, 121e154. Klein, R.G., 2003. Whither the Neanderthals. Science 299, 1525e1527. Lanyon, L.E., Bourn, S., 1979. The influence of mechanical function on the development of the tibia. J. Bone Joint Surg. 61A (2), 263e273. Lanyon, L.E., Magee, P.T., Baggott, D.G., 1979. The relationship of functional stress and strain to the processes of bone remodelling. An experimental study on the sheep radius. J. Biomech. 12 (8), 593e600. Lanyon, L.E., 1980. The influence of function on the development of bone curvature e an experimental study on the rat tibia. J. Zool. 192 (DEC), 457e466. Lanyon, L.E., 1987. Functional strain in bone tissue as an objective, and controlling stimulus for adaptive bone remodelling. J. Biomech. 20 (11e12), 1083e1093. Larsen, C.S., 1995. Biological changes in human populations with agriculture. Ann. Rev. Anthrop. 24, 185e213. Lieberman, D.E., 1989. Neanderthal and Early modern human mobility patterns: comparing archaeological and anatomical evidence. In: Akazawa, T., Aoki, K., Bar-Yosef, O. (Eds.), Neanderthals and Modern Humans in Western Asia. Plenum, New York, pp. 263e276. Marean, C.A., Assefa, Z., 1999. Zooarcheological evidence for the faunal exploitation behavior of Neandertals and early modern humans. Evol. Anthropol. 8 (1), 22e37. Marean, C.A., Assefa, Z., 2005. The middle and upper Pleistocene African record for the biological and behavioral origins of modern humans. In: Marean, C.A., Assefa, Z. (Eds.), African Archaeology. Blackwell Publishing, Oxford, pp. 333e371. Mayr, E., 1956. Geographical character gradients and climatic adaptation. Evolution 10 (1), 105e108. O’Higgins, P., 2000. The study of morphological variation in the hominid fossil record: biology, landmarks and geometry. J. Anat. 197, 103e120. O’Higgins, P., Jones, N., 1998. Facial growth in Cercocebus torquatus: an application of three dimensional geometric morphometric techniques to the study of morphological variation. J. Anat. 193, 251e272. Pearson, O.M., 2000. Activity, climate and postcranial robusticity: implications for modern human origins and scenarios of adaptive change. Curr. Anthropol. 41 (4), 569e607. Pearson, O.M., Cordero, R., Busby, A., 2006. How different were Neanderthals’ habitual activities? A comparative analysis with diverse groups of recent humans. In: Harvati, K., Harrison, T. (Eds.), Neanderthals Revisited: New Approaches and Perspectives, pp. 135e156. Richards, M.P., Pettitt, P.B., Trinkaus, E., Smith, F.H., Pauvonic, M., Karavanic, I., 2000. Neanderthal diet at Vindija and Neanderthal predation: the evidence from stable isotopes. Proc. Natl. Acad. Sci. 97 (13), 7663e7666. Richards, M.P., Pettitt, P.B., Stiner, M.C., Trinkaus, E., 2001. Stable isotope evidence for increasing dietary breadth in she European mid-Upper Paleolithic. Proc. Natl. Acad. Sci. 98 (11), 6528e6532. Ruff, C.B., 1987. Sexual dimorphism in human lower limb bone structure: relationship to subsistence strategy and sexual division of labor. J. Hum. Evol. 16, 391e416. Ruff, C.B., 1991. Climate and body shape in hominid evolution. J. Hum. Evol. 21 (2), 81e105. Ruff, C.B., Trinkaus, E., Walker, A., Larsen, C.S., 1993. Postcranial Robusticity in Homo I: temporal trends and mechanical interpretation. Am. J. Phys. Anthropol. 91, 21e53. 548 I. De Groote / Journal of Human Evolution 60 (2011) 540e548 Trinkaus, E., Rhoads, M.L., 1999. Neandertal knees: power lifters in the Pleistocene? J. Hum. Evol. 37 (6), 833e859. Trinkaus, E., Ruff, C.B., 1999a. Diaphyseal cross-sectional geometry of near eastern middle Palaeolithic humans: the tibia. J. Archaeol. Sci. 26 (10), 1289e1300. Trinkaus, E., Ruff, C.B., 1999b. Diaphyseal cross-sectional geometry of near eastern middle Paleolithic humans: the femur. J. Archaeol. Sci. 26 (4), 409e424. Trinkaus, E., Shipman, P., 1993. The Neanderthals: Changing the Image of Mankind. Knopf, New York. Trinkaus, E., Zimmerman, M.R., 1982. Trauma among the Shanidar Neanderthals. Am. J. Phys. Anthropol. 57 (1), 61e76. Trinkaus, E., Ruff, C., Churchill, S.E., 1989. Upper limb versus lower limb loading patterns among near eastern middle Paleolithic hominids. In: Akazawa, T., Aoki, K., Bar-Yosef, O. (Eds.), Neanderthals and Modern Humans in Western Asia. Plenum, New York, pp. 391e404. Trinkaus, E., Churchill, S.E., Ruff, C.B.,1994. Postcranial robusticity in Homo II: humeral bilateral asymmetry and bone plasticity. Am. J. Phys. Anthropol. 93, 1e34. Trinkaus, E., Churchill, S.E., Ruff, C.B., 1999. Long bone shaft robusticity and body proportions of the Saint-Cesaire 1 Chatelperronian Neanderthal. J. Archaeol. Sci. 26, 753e773. Trudell, M.B., 1999. Anterior femoral curvature revisited: race assessment from the femur. J. Foren. Sci. 44 (4), 700e707. Walensky, N.A., 1962. A study of femoral bowing in man. Anat. Rec. 142 (2), 289. Walensky, N.A., 1965. A study of anterior femoral curvature in man. Anat. Rec. 151 (4), 559e570. Weaver, T., Steudel-Numbers, K.L., 2005. Does climate or mobility explain the differences in body proportions between Neanderthals and their upper Paleolithic successors? Evol. Anthropol. 14 (6), 218e223. Weaver, T.D., 2002. A Multi-causal Functional Analysis of Hominid Hip-Morphology. Stanford University. Weaver, T.D., 2003. The shape of the Neandertal femur is primarily the consequence of a hyperpolar body form. Proc. Natl. Acad. Sci. U S A 100 (12), 6926e6929. Wolpoff, M.H., 1978. Some implications of relative biomechanical neck length in hominid femora. Am. J. Phys. Anthropol. 48 (2), 143e147. Yamanaka, A., Gunji, H., Ishida, H., 2005. Curvature, length, and cross-sectional geometry of the femur and humerus in anthropoid primates. Am. J. Phys. Anthropol. 127 (1), 46e57. Ruff, C.B., Walker, A., Trinkaus, E., 1994. Postcranial robusticity in Homo III: ontogeny. Am. J. Phys. Anthropol. 93, 35e54. Ruff, C.B., Trinkaus, E., Holliday, T.W., 1997. Body mass and encephalization in Pleistocene Homo. Nature 387 (6629), 173e176. Ruff, C.B., Trinkaus, E., 2000. Lifeway changes as shown by postcranial skeletal robustness. Am. J. Phys. Anthropol. S30, 266. Shackelford, L.L., Trinkaus, E., 2002. Late Pleistocene human femoral diaphyseal curvature. Am. J. Phys. Anthropol. 118, 359e370. Shackelford, L.L., 2007. Regional variation in the postcranial robusticity of late Upper Paleolithic humans. Am. J. Phys. Anthropol. 133 (1), 655e668. Slice, D.E. (Ed.), 2005. Modern Morphometrics. Modern Morphometrics in Physical Anthropology. Kluwer Academic Publishers, New York. Sorensen, M.V., Leonard, W.R., 2001. Neandertal energetics and foraging efficiency. J. Hum. Evol. 40 (6), 483e495. Speth, J.D., Tchernov, E., 1998. The role of hunting and scavenging in Neanderthal procurement strategies: new evidence from Kebara Cave (Israel). In: Akazawa, T., Aoki, K., Bar-Yosef, O. (Eds.), Neanderthals and Modern Humans in Western Asia. Plenum Press, New York, pp. 223e239. Steinbock, R.T., 1976. Paleopathological Diagnosis and Interpretation: Bone Diseases in Ancient Human Populations. Charles C. Thomas Publishers Ltd. Stewart, T.D., 1962. Anterior femoral curvature e its utility for race identification. Hum. Biol. 34 (1), 49e62. Stock, J.T., 2002. Climatic and Behavioral Influences on Postcranial Robusticity among Holocene Foragers. Ph.D dissertation, University of Toronto. Stock, J.T., Pfeiffer, S.K., 2004. Long bone robusticity and subsistence behaviour among later Stone Age foragers of the forest and fynbos biomes of South Africa. J. Archaeol. Sci. 31 (7), 999e1013. Stock, J.T., 2006. Hunteregatherer postcranial robusticity relative to patterns of mobility, climatic adaptation, and selection for tissue economy. Am. J. Phys. Anthropol. 131 (2), 194e204. Swartz, S.M., 1990. Curvature of the forelimb bones of anthropoid primates e overall allometric patterns and specializations in suspensory species. Am. J. Phys. Anthropol. 83 (4), 477e498. Trinkaus, E., 1993. Femoral neck-shaft angles of the qafzeh-skhul early modern humans, and activity levels among immature near-eastern middle Paleolithic hominids. J. Hum. Evol. 25 (5), 393e416.