Trabecular bone anisotropy and orientation in an Early Pleistocene hominin talus from East Turkana, Kenya

Journal of Human Evolution 64 (2013) 667e677

Contents lists available at SciVerse ScienceDirect

Journal of Human Evolution journal homepage: www.elsevier.com/locate/jhevol

Trabecular bone anisotropy and orientation in an Early Pleistocene hominin talus from East Turkana, Kenya Anne Su a, *, Ian J. Wallace b, Masato Nakatsukasa c a

Department of Health Sciences, Cleveland State University, Cleveland, OH 44115-2214, USA Interdepartmental Doctoral Program in Anthropological Sciences, Stony Brook University, Stony Brook, NY 11794, USA c Laboratory of Physical Anthropology, Kyoto University, Kyoto 606-8502, Japan b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 17 September 2012 Accepted 10 March 2013 Available online 16 April 2013

Among the structural properties of trabecular bone, the degree of anisotropy is most often found to separate taxa with different habitual locomotor modes. This study examined the degree of anisotropy, the elongation, and primary orientation of trabecular bone in the KNM-ER 1464 Early Pleistocene hominin talus as compared with extant hominoid taxa. Modern human tali were found to have a pattern of relatively anisotropic and elongated trabeculae on the lateral aspect, which was not found in Pan, Gorilla, Pongo, or KNM-ER 1464. Trabecular anisotropy in the fossil talus most closely resembled that of the African apes except for a region of high anisotropy in the posteromedial talus. The primary orientation of trabeculae in the anteromedial region of KNM-ER 1464 was strikingly different from that of the great apes and very similar to that of modern humans in being directed parallel to the talar neck. These results suggest that, relative to that of modern humans, the anteromedial region of the KNM-ER 1464 talus may have transmitted body weight to the midfoot in a similar manner while the lateral aspect may have been subjected to more variable loading conditions. Ó 2013 Elsevier Ltd. All rights reserved.

Keywords: Cancellous bone Astragalus Gait Ankle Bone functional adaptation

Introduction It is well accepted that Plio-Pleistocene hominins frequently engaged in terrestrial bipedalism, but it remains contentious whether or not their bipedalism was mechanically different from that of modern humans (e.g., Lordkipanidze et al., 2007; Bennett et al., 2009; Haile-Selassie et al., 2012) and/or if they routinely practiced arboreal locomotion (e.g., Stern, 2000; DeSilva, 2009; Green and Alemseged, 2012; Venkataraman et al., 2013). Fossil tali have figured prominently in discussions of early hominin locomotor behavior because the shape and arrangement of the articular surfaces of the talus are thought to provide much information about the structure and function of the entire foot (e.g., Wood, 1974; Latimer et al., 1987; DeSilva, 2009). One aspect of talar morphology that may shed additional light on the locomotor behavior of early hominins is the architecture of the trabecular bone contained within the cortical shell. A recent study by DeSilva and Devlin (2012) concluded that trabecular architecture in the hominoid talus has little value in distinguishing among species. Here, we

* Corresponding author. E-mail addresses: [email protected] (A. Su), [email protected] (I.J. Wallace), [email protected] (M. Nakatsukasa). 0047-2484/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jhevol.2013.03.003

report on data from a large sample of modern human, chimpanzee, gorilla, and orangutan tali, which in contrast does display significant differences among species that may relate to locomotor differences and therefore may be useful for inferring locomotor behavior from fossils. In previous comparative studies of trabecular bone structure among primates, the degree of anisotropy (DA), which describes the extent to which trabeculae are aligned into one or more directions, has stood out as being most able to distinguish among species whose locomotor repertoires involve different habitual joint kinematics. In species with stereotypic locomotor repertoires with joint motion primarily within a particular plane, trabecular bone tends to display greater anisotropy, whereas in species with more diverse locomotor repertoires involving varied joint kinematics, trabecular bone tends to display greater isotropy (Fajardo and Müller, 2001; MacLatchy and Müller, 2002; Ryan and Ketcham, 2002; Maga et al., 2006; Griffin et al., 2010; Saparin et al., 2011; but see Carlson et al., 2008). The plate- or rod-like geometry of trabeculae may also be useful for inferring talar loading and locomotor patterns as plate-shaped trabeculae have been shown to develop primarily in joint regions that sustain high mechanical loads, whereas rod-shaped trabeculae tend to develop in regions that experience lower magnitude loads (Ding et al., 2002). In addition to differences in trabecular anisotropy, the

668

A. Su et al. / Journal of Human Evolution 64 (2013) 667e677

specific primary spatial orientation of talar trabeculae might be expected to differ among species with different habitual joint loading patterns during locomotion (Odgaard et al., 1997; Ryan and Ketcham, 2005; Gosman and Ketcham, 2009). This expectation is based on multiple controlled loading experiments involving animal models that have demonstrated that altering joint loading orientation can cause corresponding adjustments in trabecular orientation (Pontzer et al., 2006; Barak et al., 2011). The capacity of trabecular bone to adjust and realign itself throughout life according to its customary mechanical environment is well documented (Pontzer et al., 2006; Barak et al., 2011). However, it is not the case that trabecular architecture results solely from adaptation to habitual loads. Other factors such as genetics and developmental history can also influence trabecular structure (Judex et al., 2004; Wallace et al., 2012). Thus, trabecular architecture in fossils cannot be assumed a priori to reflect functional loading. Rather, it is necessary to first determine if talar trabecular structural parameters differ among extant hominoids in ways that are consistent with presumed differences in habitual ankle joint loading patterns. Insights gleaned from the results of such comparisons can then be used to interpret talar trabecular architecture in fossils. If early hominins exhibited distinct forms of terrestrial bipedalism and/or engaged in diverse locomotor behaviors including both bipedalism and arboreal activities, then it is reasonable to infer that their talocrural joints would have been subject to different habitual loading patterns than modern humans. This should be detectable in talar trabecular structure given trabecular bone’s responsiveness to mechanical signals. With this in mind, we compared the trabecular architecture of KNM-ER 1464, an Early Pleistocene hominin talus from East Turkana, Kenya, with the talar trabecular architecture of extant hominoids (modern humans, chimpanzees, gorillas, and orangutans) in order to gain insight into the habitual loads that this fossil may have experienced during life. This specimen has exceptional preservation (Fig. 1), and its external morphology displays a unique ‘enigmatic’ mosaic of primitive and derived features (Gebo and Schwartz, 2006), suggesting that this individual may have been adapted for mixed forms of locomotion and/or displayed a unique form of bipedalism. We reasoned that if talar trabecular bone anisotropy and orientation are correlated with habitual patterns of ankle joint loading in extant hominoids, and therefore potentially useful for inferring locomotor behavior from fossils, then the following would be expected:

1. Humans have a greater overall degree of trabecular anisotropy than non-human hominoids and the regional pattern of anisotropy across the talus differs among species. 2. The primary trabecular orientation in the talus differs among species in ways consistent with observed differences in habitual ankle joint postures. Specifically, apes are expected to differ from humans in the anterior regions, based on observations of weight bearing on a highly dorsiflexed ankle during climbing (DeSilva, 2009). Also, apes are expected to differ from humans in the lateral regions, based on the varus angle of the ankle joint during terrestrial locomotion, and greater load transmission from the fibula (Marchi, 2007). If the fossil talus belonged to a hominin with a bipedal gait like that of modern humans that did not frequently engage in arboreal activities, then it should display a similar pattern of anisotropy and primary trabecular orientation. Materials and methods KNM-ER 1464 was discovered in Area 6A of the Koobi Fora Formation, 8 m below the Lower Ileret Tuff, and has been securely dated to 1.59
0.05 Ma (millions of years ago) (McDougall et al., 2012). The specimen is typically assigned to Paranthropus boisei based on its stratigraphic association with craniodental remains characteristic of that species (Wood and Constantino, 2007). Primitive aspects of KNM-ER 1464 include its deeply grooved trochlea and strongly developed fibular facet; derived aspects include its overall large size and low talar angle (Gebo and Schwartz, 2006; DeSilva, 2009) (Fig. 1). The comparative sample used in this study consisted of tali from adult modern humans (Homo sapiens; n ¼ 17), chimpanzees (Pan troglodytes; n ¼ 20), gorillas (Gorilla gorilla; n ¼ 14), and orangutans (Pongo pygmaeus; n ¼ 13) from collections at the Cleveland Museum of Natural History, the American Museum of Natural History, and the Smithsonian National Museum of Natural History. The modern human tali are from twentieth century Americans (HamanneTodd Collection). The non-human tali are primarily from wild-shot specimens. We analyzed female hominoids in order to minimize body mass differences among the species. All specimens lacked signs of any skeletal pathology or traumatic injury to the limbs. Talar trabecular architecture for all specimens, including the fossil, was assessed using high-resolution computed tomography (CT). KNM-ER 1464 was scanned using a XCT Research SA þ scanner

Figure 1. The KNM-ER 1464 talus is well preserved and displays a mixture of primitive features (deeply grooved trochlea, prominent fibular facet) and derived features (large size and low talar angle). (A) Anterior view, (B) Superior view.

A. Su et al. / Journal of Human Evolution 64 (2013) 667e677

(Stratec Medizintechnik, Pforzheim, Germany) at a resolution of 64 mm. In all CT images of the fossil, the trabecular struts appeared well defined and undamaged, and there were very few high radiodensity deposits that could have represented post-mortem mineral inclusions (Fig. 2). Other fossil hominin tali (KNM-ER 813, KNM-ER 1476, and KNM-ER 5428) were also scanned but trabecular bone was not visible. The extant primate tali were scanned using an eXplore Locus SP scanner (General Electric Healthcare, London, ON, Canada) at a resolution of 45 mm. Both the fossil and extant tali were oriented to standardized positions based on the horizontal supratalar plane of the ankle joint (Latimer et al.,1987), such that in sagittal plane view, the base of the neck and the most posterior point of the trochlear surface were on the same horizontal plane, and in the coronal and transverse planes, the superiormost points of the medial and lateral trochlear rims were level with the horizontal plane. The 16-bit three-dimensional images were imported into Amira software (Visage Imaging, San Diego, CA, USA) where each specimen was segmented into nine roughly cubic regions. To define these regions, the trochlear surface was divided into a three by three grid by measuring the maximum linear mediolateral dimension, and the arc between the posteriormost and anteriormost points, and dividing them into thirds. The trabecular bone immediately deep to the trochlear cortex in each region was isolated and saved as a separate volume. A Gaussian filter (s ¼ 1) was applied to reduce noise in the images. Each volume was then imported into Quant3D software (Ryan and Ketcham, 2002) for quantitative analysis of trabecular structure. The volume of interest (VOI) was defined as the largest centered sphere to fit completely within each region of trabecular bone (Fig. 3). Trabecular bone within the VOI was binarized into bone/non-bone using an adaptive, iterative threshold technique (Ryan and Ketcham, 2002). Trabecular structure was quantified using the star volume distribution (SVD) algorithm. The SVD method is based on the measured length of the longest uninterrupted line from a point lying within trabecular bone to a boundary between bone and air, repeated for a series of uniformly distributed orientations and multiple random points (Cruz-Orive et al., 1992; Odgaard et al., 1997). From these data, a fabric tensor is derived that describes how the moment of inertia of bone varies with orientation. From the fabric tensor, three eigenvectors and three corresponding eigenvalues are derived that describe bone distribution. The eigenvectors represent the orientation in three-dimensional space of the primary, secondary, and tertiary material axes. The corresponding eigenvalues (s1, s2, s3) represent the relative magnitudes of each of the three material axes. They are defined such that s1 þ s2 þ s3 ¼ 1 and s1 > s2 > s3. The degree of anisotropy (DA) was calculated as 1  (s3/s1), such that the values are bounded between 0 (perfect isotropy; i.e., a solid sphere of bone) and 1 (perfect anisotropy; i.e., bony struts aligned in a single plane) (Harrigan and Mann, 1984; Doube et al., 2010).

669

Figure 3. The trabecular bone immediately underlying the talar trochlea was segmented into nine spherical volumes of interest.

Although we acknowledge that another conventional way to report DA is (s1/s3), with a lower boundary value of 1 but no upper boundary (Fajardo and Müller, 2001; Ryan and Ketcham, 2002), we chose the convention bounded between 0 and 1 for ease of statistical analysis. Trabecular elongation index (E), calculated as 1  (s2/ s1), distinguishes between rod-shaped and plate-shaped trabeculae by indicating the extent of preferred orientation of trabeculae in the major plane defined by the primary and secondary eigenvectors. If the DA is close to 1, concurrent E values closer to 0 denote more plate-shaped trabecular struts, and E values closer to 1 denote more rod-shaped struts. If the DA is close to 0, values of E are less meaningful. Values of DA close to 0 can describe either a volume with numerous thin trabeculae that are randomly oriented or a volume that is completely filled with bone, both morphologies resulting in a lack of dominant orientations. Although some stereological measurements of trabecular bone are dependent on the spatial resolution and source energy of the CT images and the level of mineralization of bone tissue, eigenvalue ratios are unlikely to be affected by the relatively small difference in scanning resolution for KNM-ER 1464 (64 mm) and the extant primate tali (46 mm) (Müller et al., 1996; Sode et al., 2008). The orientation of the primary eigenvector that describes the primary axis of bone distribution is also unlikely to be affected by the difference in scanning resolution for KNM-ER 1464 and the extant

Figure 2. Representative CT images of the trabecular bone architecture of modern human, KNM-ER 1464, chimpanzee, gorilla and orangutan tali in the (a) coronal plane, (b) transverse plane, and (c) sagittal plane.

670

A. Su et al. / Journal of Human Evolution 64 (2013) 667e677

primate tali. Furthermore, patterns of these measurements across the bone can be discussed with greater confidence, presuming that any systematic imaging error is uniformly distributed throughout each data set. ANOVA was used to compare DA and E first among species (Homo, Pan, Gorilla, Pongo) within each regional subset of the talus, then among regions (A-L, A-C, A-M, C-L, C-C, C-M, P-L, P-C, P-M) within each species. Post hoc pairwise comparisons were conducted with GameseHowell tests. Analyses of DA and E were performed in SPSS 16.0 (Chicago, IL, USA) with significance a < 0.05. Because the dispersion of primary trabecular orientations within each species was more elliptical in shape (Fig. 7), the distribution was not well represented by the Fisher confidence cone, which assumes circular symmetry about the mean vector. Thus eigenvector-based statistics were used to provide a more meaningful description of the species mean and variation in principal orientation (Fisher et al., 1987). In this method, each vector endpoint is thought of as being a mass and the mean principal direction for each species is the axis about which the moment of inertia of the masses would be least (maximum eigenvector). The semiaxes of the 95% confidence ellipse are proportional to the intermediate and minimum eigenvalues. The intraspecific variation in primary orientations is described by a strength parameter z (zero indicates random, isotropic distribution, values >4 indicate strong clustering) and a shape parameter g (0 indicates an elongated girdle-shaped distribution, values of >5 indicate a circular distribution) (Woodcock, 1977). Statistical analyses of orientation data were performed with Palaeomag-Tools v. 4.2 (Lancaster University, UK).

all regions (Table 2, Figs. 4 and 6). Trabecular E was also significantly greater in humans in almost all lateral and anterior regions (except versus the anterocentral region of orangutans) (Table 2, Figs. 5 and 6). The African apes (chimpanzees and gorillas) did not differ from each other in DA or E in any region except P-C. The African apes had significantly greater DA than orangutans in many regions but lower E particularly in the C-M region. Among extant species, humans displayed highly significant differences in DA and E among regions of the talus (Table 3, Figs. 4 and 5). In humans, the Lateral and MLCentral regions were more anisotropic than the Medial region. These regional differences are in contrast to those of chimpanzees, gorillas, and orangutans where few differences in anisotropy were detected among regions (Table 3, Fig. 4). The trabeculae in the lateral regions were significantly more elongated and rod-shaped in modern humans. The Anterior region in chimpanzees and gorillas had significantly lower elongation than the more posterior regions. Orangutans had the fewest differences in DA and E among regions. The magnitude of DA and E in the anterior two-thirds of KNMER 1464 most closely resembled that of the non-human hominoids (Figs. 4 and 5). In the posterocentral and posteromedial regions, however, the fossil displayed relatively high DA, which more closely resembled modern humans (Figs. 4 and 6). Trabecular orientation The primary orientation of trabeculae for each region are plotted on stereoplots in Fig. 7 and summarized in Table 4. Each stereoplot depicts the 95% confidence ellipse and scatter within species in the trend (0e359 with 0 being Anterior) and plunge (0e90 with 0 at points along the perimeter of the circle and 90 directed inferiorly and at the center of the circle) of the primary orientation of trabeculae in each region. The orientation of trabeculae in the modern human and KNMER 1464 tali differed from that of extant non-human hominoids in the anterocentral (A-C) and anteromedial (A-M) regions (see also Fig. 2). In these regions, the trabecular orientation in humans was consistently directed in an anteroinferior line generally parallel with the talar neck. The non-human primates exhibited greater variability (lower z, Table 4) as well as differently directed primary orientations. Chimpanzee and gorilla trabeculae were directed more medially/inferomedially in region A-C and inferoposteriorly in region A-M. Orangutan trabeculae were directed posteriorly in both A-C and A-M. In the posterolateral (P-L) region, the primary orientation in modern humans also differed in forming a tight cluster (g ¼ 6.83, z ¼ 4.79) directed posteroinferiorly toward the posterior calcaneal facet. In contrast, the primary orientation in gorillas, orangutans, and KNM-ER 1464 were directed slightly anteroinferiorly (see also Fig. 2). The transversely directed

Results Trabecular anisotropy and elongation The degree of anisotropy (DA) and the elongation index (E) for each region of the talus are given in Table 1 and shown in Figs. 4 and 5, respectively. Additionally, these data are plotted on shape diagrams (Benn, 1994; Graham and Midgley, 2000) in Fig. 6 for better visualization of the interplay between DA and E in describing trabecular shape. In these shape diagrams, data points toward the top apex indicate more isotropic trabeculae and data points toward the bottom indicate more anisotropic trabeculae. Data points toward the bottom left apex indicate more plate-shaped trabeculae and data points toward the bottom right apex indicate more rodshaped trabeculae. Table 2 presents the results of the ANOVA analysis and post hoc comparisons among species within each region of the talus. Table 3 presents the ANOVA results and post hoc comparisons among regions within each species. The DA of the trabecular bone in the human talus was significantly greater than that of the other extant hominoids in virtually Table 1 Means and standard deviations of DA and E for each region of the talus. Homo (n ¼ 17) DA E A-L A-C A-M C-L C-C C-M P-L P-C P-M

0.88 0.91 0.68 0.87 0.85 0.81 0.80 0.85 0.73

(0.04) (0.04) (0.08) (0.04) (0.08) (0.09) (0.05) (0.09) (0.14)

0.62 0.38 0.50 0.59 0.38 0.35 0.66 0.41 0.27

(0.13) (0.11) (0.11) (0.12) (0.12) (0.13) (0.07) (0.11) (0.09)

Pan (n ¼ 20) DA 0.62 0.68 0.54 0.64 0.57 0.65 0.54 0.57 0.61

(0.10) (0.09) (0.13) (0.11) (0.10) (0.10) (0.15) (0.11) (0.12)

E 0.24 0.27 0.23 0.28 0.38 0.30 0.26 0.39 0.33

(0.13) (0.12) (0.14) (0.14) (0.14) (0.11) (0.09) (0.11) (0.14)

Gorilla (n ¼ 14) DA E 0.69 0.62 0.55 0.65 0.67 0.67 0.58 0.70 0.61

(0.08) (0.10) (0.11) (0.13) (0.11) (0.09) (0.11) (0.08) (0.12)

0.28 0.24 0.23 0.30 0.43 0.38 0.33 0.40 0.32

(0.11) (0.13) (0.10) (0.12) (0.12) (0.10) (0.11) (0.07) (0.12)

Pongo (n ¼ 13) DA E 0.44 0.56 0.54 0.48 0.57 0.49 0.53 0.49 0.37

(0.14) (0.16) (0.14) (0.13) (0.18) (0.12) (0.16) (0.16) (0.11)

0.24 0.29 0.33 0.28 0.29 0.15 0.29 0.26 0.17

(0.13) (0.16) (0.13) (0.15) (0.17) (0.06) (0.12) (0.13) (0.09)

KNM-ER 1464 DA E 0.66 0.68 0.48 0.81 0.61 0.72 0.58 0.87 0.89

0.37 0.31 0.34 0.43 0.26 0.19 0.35 0.11 0.22

A-L ¼ anterolateral, A-C ¼ anterocentral, A-M ¼ anteromedial, C-L ¼ centrolateral, C-C ¼ central, C-M ¼ centromedial, P-L ¼ posterolateral, P-C ¼ posterocentral, P-M ¼ posteromedial.

A. Su et al. / Journal of Human Evolution 64 (2013) 667e677

671

Figure 4. Boxplots of the degree of anisotropy (DA) across nine regions of the talus in each species. The distribution of DA in chimpanzees, gorillas, and orangutans is more homogenous across the anterior regions whereas that of humans displays a pattern of relatively increased DA in the anterolateral and anterocentral regions. KNM-ER 1464 resembles the pattern seen in chimpanzees and gorillas except for uniquely having much greater DA in the posterocentral and posterolateral regions.

trabeculae seen predominantly in chimpanzees seem to extend laterally toward the fibular facet. Discussion In this study, we found that modern human tali had trabecular bone that was overall more anisotropic than that of extant non-

human hominoids and displayed greater interregional variation in the degree of anisotropy. These results are consistent with multiple previous studies of trabeculae in skeletal elements of the foot that found overall greater anisotropy in humans relative to other hominoids (Maga et al., 2006; Griffin et al., 2010; DeSilva and Devlin, 2012). Anisotropically-arranged trabecular bone has been shown to be relatively weaker when loaded in non-primary loading

Figure 5. Boxplots of elongation (E) across nine regions of the talus in each species. Humans differ in pattern from the great apes in having much more elongated trabeculae along the lateral talus than other regions. KNM-ER 1464 also displays this pattern of more elongated trabeculae in the lateral regions.

672

A. Su et al. / Journal of Human Evolution 64 (2013) 667e677

Figure 6. Ternary diagrams of trabecular shape indices for each region of the talus. Within each region, each point plots the degree of anisotropy index (DA) and the elongation index (E) value. Lines were drawn to enclose the distribution of each species for comparative visualization. In these diagrams, data located toward the top apex indicate more isotropic trabeculae and those found near the bottom indicate more anisotropic trabeculae. Data toward the bottom left apex are more plate-shaped trabeculae and those toward the bottom right apex are more rod-shaped trabeculae. The DA of the trabecular bone in the human talus was significantly greater than that of the other extant hominoids in virtually all regions. Trabecular E was also significantly greater in humans in the anteromedial and all lateral regions. A-L ¼ anterolateral, A-C ¼ anterocentral, A-M ¼ anteromedial, C-L ¼ centrolateral, C-C ¼ central, C-M ¼ centromedial, P-L ¼ posterolateral, P-C ¼ posterocentral, P-M ¼ posteromedial.

Table 2 Interspecific ANOVA and p-values for post hoc pairwise comparisons in each region of the talus. F

DA A-L A-C A-M C-L C-C C-M P-L P-C P-M E A-L A-C A-M C-L C-C C-M P-L P-C P-M

p

Homo versus

Pan versus

Gorilla versus

Pan

Gorilla

Pongo

Gorilla

Pongo

Pongo

56.54 37.21 5.67 35.13 20.84 26.31 17.85 31.71 21.48

0.000 0.000 0.002 0.000 0.000 0.000 0.000 0.000 0.000

0.000*** 0.000*** 0.002** 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.042*

0.000*** 0.000*** 0.007** 0.000*** 0.000*** 0.001** 0.000*** 0.000*** 0.077

0.000*** 0.000*** 0.024* 0.000*** 0.000*** 0.000*** 0.000*** 0.000*** 0.000***

0.116 0.399 0.990 0.993 0.063 0.974 0.867 0.001** 1.000

0.002** 0.117 1.000 0.005** 1.000 0.001** 0.996 0.427 0.000***

0.000*** 0.641 0.993 0.008** 0.337 0.001** 0.818 0.003** 0.000**

34.64 3.65 19.15 21.45 2.27 12.03 59.50 6.12 5.50

0.000 0.017 0.000 0.000 0.089 0.000 0.000 0.001 0.002

0.000*** 0.019* 0.000*** 0.000*** 0.999 0.564 0.000*** 0.909 0.475

0.000*** 0.022* 0.000*** 0.000*** 0.714 0.932 0.000*** 0.995 0.606

0.000*** 0.319 0.005** 0.000*** 0.388 0.000*** 0.000*** 0.013* 0.027*

0.773 0.960 0.999 0.994 0.669 0.178 0.227 0.947 0.999

1.000 0.963 0.151 0.999 0.461 0.000*** 0.823 0.039* 0.003**

0.849 0.846 0.098 0.988 0.114 0.000*** 0.852 0.012* 0.008

A-L ¼ anterolateral, A-C ¼ anterocentral, A-M ¼ anteromedial, C-L ¼ centrolateral, C-C ¼ central, C-M ¼ centromedial, P-L ¼ posterolateral, P-C ¼ posterocentral, P-M ¼ posteromedial. ***: p < 0.001, **: p < 0.01, *: p < 0.05.

directions than isotropically-arranged bone (Homminga et al., 2002). The orientation of trabeculae in the tali of modern humans versus apes differed most markedly in the anteromedial region. As qualitatively noted by previous authors, in modern humans the orientation in the anteromedial region is directed toward the talar neck (Takechi et al., 1982; Pal and Routal, 1998; Ebraheim et al., 1999; Athavale et al., 2008), whereas in apes it is directed posteroinferiorly. The orientation in the posterolateral region in humans was angled consistently slightly posteriorly toward the posterior calcaneal facet, while the non-human hominoids did not consistently show this distinctive pattern. Our results differ slightly from those of the recent study by DeSilva and Devlin (2012). DeSilva and Devlin (2012) found that both humans and chimpanzees do not differ in DA across regional quadrants of the talus, except with statistically greater DA in the posterolateral quadrant. We found that indeed chimpanzees show a relatively homogenous distribution of DA across the talus, with only the anterocentral region having statistically greater DA versus other regions. But we found humans to have strong regional differences, with the DA in the anterocentral, anterolateral, and centrolateral regions statistically greater than other regions and the anteromedial region lower than other regions (Table 3, Fig. 4). Methodological differences may explain these disparities. DeSilva and Devlin partitioned their analysis of the talar body into four quadrants while we had a finer partition of nine regions. The coarser four-quadrant analysis may have moderated some details of trabecular architecture. Also, their sample size for each species was

Table 3 Intraspecific ANOVA and p-values for post hoc pairwise comparisons among regions of the talus. DA Homo A-L A-C A-M C-L C-C C-M P-L P-C

A-L A-C A-M C-L C-C C-M P-L P-C Gorilla A-L A-C A-M C-L C-C C-M P-L P-C Pongo A-L A-C A-M C-L C-C C-M P-L P-C

p < 0.001 A-M 0.000*** 0.000***

F ¼ 3.792 A-C 0.673

p < 0.001 A-M 0.445 0.013*

F ¼ 3.413 A-C 0.514

p ¼ 0.001 A-M 0.015* 0.686

F ¼ 2.461 A-C 0.445

p ¼ 0.017 A-M 0.637 1.000

Homo C-L 1.000 0.089 0.000***

C-C 0.945 0.188 0.000*** 0.985

C-M 0.137 0.007** 0.006** 0.198 0.857

P-L 0.001** 0.000*** 0.001** 0.001** 0.500 1.000

P-C 0.975 0.315 0.000*** 0.994 1.000 0.877 0.590

P-M 0.009** 0.001** 0.920 0.012* 0.079 0.597 0.499 0.091

C-L 0.999 0.979 0.196

C-C 0.822 0.029* 0.996 0.470

C-M 0.982 0.997 0.078 1.000 0.208

P-L 0.597 0.037* 1.000 0.314 0.999 0.157

P-C 0.855 0.039* 0.994 0.516 1.000 0.245 0.998

P-M 1.000 0.625 0.660 0.996 0.950 0.959 0.776 0.962

C-L 0.985 0.998 0.401

C-C 0.999 0.958 0.161 1.000

C-M 0.994 0.944 0.101 1.000 1.000

P-L 0.071 0.956 1.000 0.733 0.425 0.331

P-C 1.000 0.351 0.008** 0.945 0.989 0.959 0.040*

P-M 0.531 1.000 0.874 0.994 0.938 0.922 0.993 0.384

C-L 0.997 0.835 0.956

C-C 0.489 1.000 1.000 0.843

C-M 0.982 0.882 0.978 1.000 0.887

P-L 0.811 1.000 1.000 0.990 1.000 0.997

P-C 0.987 0.966 0.997 1.000 0.962 1.000 1.000

P-M 0.905 0.035* 0.056 0.423 0.060 0.234 0.147 0.419

A-L A-C A-M C-L C-C C-M P-L P-C Pan A-L A-C A-M C-L C-C C-M P-L P-C Gorilla A-L A-C A-M C-L C-C C-M P-L P-C Pongo A-L A-C A-M C-L C-C C-M P-L P-C

F ¼ 24.877 A-C 0.000***

p < 0.001 A-M 0.160 0.057

F ¼ 4.126 A-C 0.999

p < 0.001 A-M 1.000 0.993

F ¼ 5.520 A-C 0.997

p < 0.001 A-M 0.905 1.000

F ¼ 2.651 A-C 0.994

p ¼ 0.011 A-M 0.662 0.997

C-L 0.997 0.000*** 0.475

C-C 0.000*** 1.000 0.100 0.001**

C-M 0.000*** 0.998 0.023* 0.000*** 0.997

P-L 0.970 0.000*** 0.001** 0.445 0.000*** 0.000***

P-C 0.001** 0.995 0.310 0.003** 0.998 0.865 0.000***

P-M 0.000*** 0.063 0.000*** 0.000*** 0.080 0.487 0.000*** 0.007**

C-L 0.976 1.000 0.939

C-C 0.060 0.167 0.043* 0.474

C-M 0.813 0.988 0.706 1.000 0.567

P-L 1.000 1.000 0.997 0.998 0.066 0.926

P-C 0.014* 0.042* 0.010* 0.230 1.000 0.259 0.008**

P-M 0.538 0.852 0.435 0.988 0.962 0.999 0.666 0.851

C-L 1.000 0.970 0.734

C-C 0.050 0.015* 0.001** 0.116

C-M 0.355 0.122 0.013* 0.604 0.933

P-L 0.958 0.650 0.231 0.997 0.392 0.961

P-C 0.045* 0.015* 0.000*** 0.125 0.998 0.994 0.482

P-M 0.991 0.808 0.413 1.000 0.347 0.925 1.000 0.439

C-L 0.999 1.000 0.983

C-C 0.993 1.000 0.999 1.000

C-M 0.388 0.166 0.006** 0.205 0.214

P-L 0.978 1.000 0.995 1.000 1.000 0.040*

P-C 1.000 1.000 0.860 1.000 1.000 0.269 0.999

P-M 0.758 0.364 0.027* 0.446 0.417 1.000 0.154 0.593

A. Su et al. / Journal of Human Evolution 64 (2013) 667e677

Pan

E F ¼ 15.353 A-C 0.246

A-L ¼ anterolateral, A-C ¼ anterocentral, A-M ¼ anteromedial, C-L ¼ centrolateral, C-C ¼ central, C-M ¼ centromedial, P-L ¼ posterolateral, P-C ¼ posterocentral, P-M ¼ posteromedial. ***: p < 0.001, **: p < 0.01, *: p < 0.05.

673

674

A. Su et al. / Journal of Human Evolution 64 (2013) 667e677

smaller, which may have also precluded findings of statistical significance. Moreover, in each of the quadrants they included trabecular bone of the inferior aspect of the talus while we focused only on the trabeculae immediately deep to the superior articular surface. The trabeculae of the inferior talus, particularly that of the posterolateral region, may be influenced more by intraarticular forces in the subtalar joint rather than the talocrural joint. These inferior trabeculae may have contributed to their finding of greater DA in the posterolateral region. Although trabecular structure is not modulated solely by mechanical signals, it is reasonable to expect that the inter- and intraspecific patterns in talar trabecular structure detected in the current study were shaped, at least to some extent, by the loads engendered by the ankle joints in these species during locomotion. Indeed, the observed patterns are generally consistent with what is currently known about the functional loads sustained by hominoid ankles. The overall greater anisotropy of trabecular bone in modern human tali accords well with our more stereotypical locomotor repertoire involving ankle joint loading that is generally limited to a single plane. Non-human hominoids, on the other hand, sustain more diverse ankle joint loads as they navigate both arboreal and terrestrial environments (Hunt, 1991; Hunt et al., 2004), which would be expected to result in less oriented trabecular bone (MacLatchy and Müller, 2002; Ryan and Ketcham, 2002; Hébert et al., 2012). In terms of intraspecific regional structural variation,

the pattern of more anisotropic, elongated rod-shaped trabeculae in the lateral aspect of the modern human talus is conceivably related to habitual loads being lower and/or more predictable in this region than on the medial aspect in comparison with the pattern in other extant hominoids. In-vivo internal joint loads at the ankle are difficult to measure (Michelson et al., 2001; Tochigi et al., 2006; Potthast et al., 2008), but clinical support for this non-homogenous load pattern in humans includes the finding that osteochondral lesions occur most often on the medial talus (Raikin et al., 2007). Indeed, mechanical osteopenetration tests have confirmed that trabecular bone of the medial human talus demonstrates greater bone strength than the lateral regions (Hvid et al., 1985). The more homogenous pattern of isotropy found in the nonhuman hominoids is consistent with the relatively more variable nature of their ankle joint loading during locomotion. Indeed, although apes exert higher plantar pressures on the medial side of the foot during climbing (DeSilva, 2009), they exert higher pressures on the lateral side of the foot during quadrupedal locomotion (Vereecke et al., 2003). The primary orientation of trabeculae in the anteromedial region of modern human tali toward the talar neck is likely related to habitual load transfer directed from the medial aspect of the trochlea inferiorly to the talar head and navicular as part of the medial longitudinal arch. In contrast, the posteroinferior orientation of trabeculae in this region among the non-human hominoids may

Figure 7. Distribution and 95% CI of the primary trabecular eigenvector direction in each region of the talus. Each region is represented by the lower hemisphere of a sphere, where each primary orientation eigenvector is depicted as having its origin at the center of the sphere and tip (marker) extending inferiorly to the lower surface of the sphere. In the central-region (C-C), all species have similarly vertically oriented trabeculae. In the anterocentral (A-C) and anteromedial region (A-M), the orientation in humans was consistently directed in an anteroinferior line generally parallel with the talar neck while the non-human primates exhibited greater variability. In the posterolateral region, the primary orientation in modern humans also differed from other species in forming a tight cluster directed posteroinferiorly toward the posterior calcaneal facet. A-L ¼ anterolateral, AC ¼ anterocentral, A-M ¼ anteromedial, C-L ¼ centrolateral, C-C ¼ central, C-M ¼ centromedial, P-L ¼ posterolateral, P-C ¼ posterocentral, P-M ¼ posteromedial.

A. Su et al. / Journal of Human Evolution 64 (2013) 667e677 Table 4 Species means and variation in primary trabecular orientation. n A-L

A-C

A-M

C-L

C-C

C-M

P-L

P-C

P-M

Homo Pan Gorilla Pongo KNM-ER Homo Pan Gorilla Pongo KNM-ER Homo Pan Gorilla Pongo KNM-ER Homo Pan Gorilla Pongo KNM-ER Homo Pan Gorilla Pongo KNM-ER Homo Pan Gorilla Pongo KNM-ER Homo Pan Gorilla Pongo KNM-ER Homo Pan Gorilla Pongo KNM-ER Homo Pan Gorilla Pongo KNM-ER

1464

1464

1464

1464

1464

1464

1464

1464

1464

17 20 14 13 1 17 20 14 13 1 16 20 14 13 1 17 20 14 13 1 17 20 14 13 1 17 20 14 13 1 17 20 14 13 1 17 20 14 13 1 17 20 14 13 1

Mean direction Trend ( ) Plunge ( ) 210.9 214.1 215.0 239.9 200.4 28.8 80.3 81.3 209.4 5.5 350.6 144.5 143.0 187.8 356.9 231.0 248.2 239.6 256.4 249.7 296.8 109.8 133.4 199.3 298.9 30.6 51.4 73.5 144.9 18.1 207.3 260.0 284.4 291.2 38.8 233.0 296.7 276.8 232.1 340.6 35.3 30.7 23.9 6.5 7.2

55.4 47.8 48.8 21.7 59.0 66.9 48.6 52.2 37.4 58.2 46.8 67.4 73.9 33.0 50.2 59.5 55.4 60.9 63.0 70.4 84.8 76.6 85.6 84.1 79.6 61.5 64.1 73.6 71.6 55.6 73.1 30.2 61.8 58.6 68.3 70.3 75.0 78.7 70.1 68.0 74.0 42.9 56.7 68.1 29.8

Data distribution Shape g Strength z 1.86 0.57 0.40 1.04

5.26 4.06 5.44 2.23

0.40 0.48 0.36 0.86

5.88 3.05 3.29 2.63

3.03 0.58 2.68 0.37

3.82 2.05 2.44 2.76

5.07 0.51 1.15 2.74

4.77 4.27 5.11 3.59

0.64 2.69 1.67 0.87

5.83 3.85 5.37 3.41

0.77 0.67 1.25 0.97

4.38 5.33 4.10 2.48

6.83 0.21 0.55 1.95

4.79 3.63 4.60 2.64

0.80 2.07 1.18 1.30

5.85 4.43 5.69 3.02

0.22 1.80 0.65 0.34

2.57 2.69 4.25 2.07

The trend is the orientation within the transverse plane with 0 directed anteriorly and 180 directed posteriorly. The plunge is the angle of incline with 0 lying within the transverse plane and 90 directed inferiorly. The shape (g) of the distribution varies from 0.0 (elongated distribution) to 6.0 or greater (highly circular distribution). The strength (z) of the distribution varies from 0.0 (isotropic distribution) to 6.0 or greater (highly concentrated).

be associated with loads sustained by a highly dorsiflexed ankle during climbing (DeSilva, 2009). In the anterocentral region, the more oblique orientation of trabeculae in the non-human hominoids is consistent with the previously observed greater medial deviation of their talar necks (Day and Wood, 1968). This orientation may thus be associated with an oblique transfer of load from the lateral aspect of the talocrural joint to the talar head during the observed varus ankle posture in apes (Latimer et al., 1987; DeSilva, 2009). In humans, the slightly posterior orientation of trabeculae in the posterolateral region of the talus toward the posterior calcaneal facet is perhaps associated with habitual load transfer directed from the lateral aspect of the trochlea to the posterior calcaneus. This morphology may be linked with the high magnitude load during human heelstrike (Hunt et al., 2001). The non-human hominoids did not consistently show this distinctive pattern, a fact which may be related to the absence of similar habitual rearfoot loading in these species (Schmitt and Larson, 1995).

675

Assuming that the patterns of trabecular structure observed in the extant hominoids examined in the current study were influenced by locomotor behavior, they provide insights into the functional loading history of KNM-ER 1464. The overall degree of trabecular anisotropy and elongation in KNM-ER 1464, particularly in the anterior two-thirds of the talus, most closely resembled that of the non-human hominoids and was different from the high level of anisotropy found in modern humans (Fig. 6). This suggests that this early hominin (and possibly the species to which it belongs; i.e., Paranthropus boisei) used a bipedal gait that was distinct from that of modern humans and/or habitually engaged in more diverse forms of locomotion than modern humans, such as arboreal activities. However, the fossil did display a relatively higher, more modern human-like anisotropy in the posterior region, a pattern that was also detected by DeSilva and Devlin (2012) in their analysis of tali of Australopithecus africanus. Interestingly, the orientation of trabeculae in KNM-ER 1464 is most similar to that of modern humans in the anteromedial region, suggesting that this hominin may have habitually transferred weight in a similar way from the leg, through the talus, to the midfoot. This evidence supports some analyses of external morphology that indicate the possible presence of a medial longitudinal arch in Plio-Pleistocene hominins (Lamy, 1986; Latimer and Lovejoy, 1989; Gebo, 1992). However, the dissimilarity in orientation between KNM-ER 1464 and modern humans in the posterior regions further suggests that if this early hominin was a committed biped, then it likely practiced a form of bipedalism that was mechanically different from that of modern humans (Gebo and Schwartz, 2006). Cotter et al. (2009), in their study of trabecular structure in the T8 vertebral body across hominoids, found a strong negative relationship between DA and volumetric bone volume fraction. Therefore, it is conceivable that the degree of anisotropy in the modern human tali examined here was related less to stereotypy of ankle joint loading and more to the overall lower activity levels of recent humans relative to wild hominoids. The modern humans analyzed (Hamann-Todd Collection) were young adult early twentieth century women who likely worked in service and labor occupations requiring long, repetitive periods of standing or walking (Davis, 1972). Therefore, these humans were by no means physically inactive, but nor were they engaged in patterns of physical activity that closely resemble human foragers. Footwear worn by women in our sample might also have influenced their ankle joint loading and trabecular structure. Versus being barefoot, wearing footwear during walking and running has been shown to result in reduced inversion/eversion and abduction/adduction range of motion (Morio et al., 2009; Wolf et al., 2009) and greater ground reaction forces (De Wit et al., 2000; Divert et al., 2005; Sacco et al., 2010). Therefore, comparison with trabecular bone in an unshod, pre-industrial population may have been more appropriate. However, the enormity of the Hamann-Todd Collection allowed for the selection of a large sample of non-pathological individuals to fit narrow criteria of age and weight. With these results as a baseline, future studies of other populations of modern humans would be useful to assess the effect of locomotor activity level and pattern on anisotropy, especially populations that frequently engage in climbing (Venkataraman et al., 2013). The possibility exists that the pattern of greater anisotropy in the human talus detected in this study was due, at least in part, to genetics or other non-mechanical factors, rather than localized biomechanical adaptation alone. Interspecific comparative studies of the ontogenetic development of trabecular bone morphology have not yet been reported, so it is unknown if DA is innately greater in humans versus other hominoids. Non-mechanical explanations are also reasonable for the primary orientation patterns detected. For example, the anteroinferior trabecular orientation in

676

A. Su et al. / Journal of Human Evolution 64 (2013) 667e677

the anterior region of human tali may simply be a consequence of the progression of ossification from the center of the talar neck posteriorly to the body (Hubbard et al., 1993; Fritsch et al., 1996). Evidence has been unclear whether the gross adult trabecular orientation is established early in development (Skedros et al., 2004, 2007; Cunningham and Black, 2009; Abel and Macho, 2011) or later in association with the increased demands of weight bearing (Tanck et al., 2001; Ryan and Krovitz, 2006). Future research aimed at examining the degree to which talar trabecular bone is influenced by loading versus genetic background might analyze an ontogenetic series of specimens across closely related taxa. If human talar trabecular bone is strongly influenced by locomotion, then changes in talar morphology should take place at one to two years of age at the onset of walking. Finally, it is important to note that although the resolution of the fossil CT data was well within the resolution needed to accurately characterize trabecular morphology (Müller et al., 1996), comparing CT images from different systems and with different resolutions can be problematic, particularly with potential beam hardening effects associated with imaging highly mineralized bone (Tafforeau et al., 2006). Therefore, our characterization of trabecular bone structure in KNM-ER 1464 relative to other hominoids must be considered tentative. Ultimately, the provenance of trabecular bone structure in modern humans as a morphologically primitive, derived, or epigenetic feature will require further analysis of fossil and pre-industrial specimens. Acknowledgments We thank Emma Mbua and Fredrick Kyalo Manthi at the National Museums of Kenya for facilitating the scanning of KNM-ER 1464. For assistance with the data collection of other specimens, we thank Lyman Jellema at the Cleveland Museum of Natural History, Eileen Westwig at the American Museum of Natural History, and Amit Vasanji and Richard Rozic at Image-IQ, Inc. Thanks also to Luci Betti-Nash and Ashley Gosselin-Ildari for their photographic assistance. This study was supported by the National Science Foundation, Wenner Gren Foundation, L.S.B. Leakey Foundation, and Turkana Basin Institute. References Abel, R., Macho, G.A., 2011. Ontogenetic changes in the internal and external morphology of the ilium in modern humans. J. Anat. 218, 324e335. Athavale, S.A., Joshi, S.D., Joshi, S.S., 2008. Internal architecture of the talus. Foot Ankle Int. 29, 82e86. Barak, M.M., Lieberman, D.E., Hublin, J.J., 2011. A Wolff in sheep’s clothing: trabecular bone adaptation in response to changes in joint loading orientation. Bone 49, 1141e1151. Benn, D.I., 1994. Fabric shape and the interpretation of sedimentary fabric data. J. Sediment. Res. A 64, 910e915. Bennett, M.R., Harris, J.W.K., Richmond, B.G., Braun, D.R., Mbua, E., Kiura, P., Olago, D., Kibunjia, M., Omuombo, C., Behrensmeyer, A.K., Huddart, D., Gonzalez, S., 2009. Early hominin foot morphology based on 1.5-million-yearold footprints from Ileret, Kenya. Science 323, 1197e1201. Carlson, K.J., Lublinsky, S., Judex, S., 2008. Do different locomotor modes during growth modulate trabecular architecture in the murine hind limb? Integr. Comp. Biol. 48, 385e393. Cotter, M.M., Simpson, S.W., Latimer, B.M., Hernandez, C.J., 2009. Trabecular microarchitecture of hominoid thoracic vertebrae. Anat. Rec. 292, 1098e1106. Cruz-Orive, L.M., Karlsson, L.M., Larsen, S.E., Wainschtein, F., 1992. Characterizing anisotropy: a new concept. Micron Microsc. Acta 23, 75e76. Cunningham, C.A., Black, S.M., 2009. Development of the fetal ilium e challenging concepts of bipedality. J. Anat. 214, 91e99. Davis, R.H., 1972. Black Americans in Cleveland from George Peake to Carl B. Stokes, 1796e1969. Associated Publishers, Washington. Day, M.H., Wood, B.A., 1968. Functional affinities of the Olduvai Hominid 8 talus. Man 3, 440. De Wit, B., De Clercq, D., Aerts, P., 2000. Biomechanical analysis of the stance phase during barefoot and shod running. J. Biomech. 33, 269e278.

DeSilva, J.M., 2009. Functional morphology of the ankle and the likelihood of climbing in early hominins. Proc. Natl. Acad. Sci. 106, 6567e6572. DeSilva, J.M., Devlin, M.J., 2012. A comparative study of the trabecular bony architecture of the talus in humans, non-human primates, and Australopithecus. J. Hum. Evol. 63, 536e551. Ding, M., Odgaard, A., Linde, F., Hvid, I., 2002. Age-related variations in the microstructure of human tibial cancellous bone. J. Orthop. Res. 20, 615e621. Divert, C., Mornieux, G., Baur, H., Mayer, F., Belli, A., 2005. Mechanical comparison of barefoot and shod running. Int. J. Sports Med. 26, 593e598. Doube, M., K1osowski, M.M., Arganda-Carreras, I., Cordelières, F.P., Dougherty, R.P., Jackson, J.S., Schmid, B., Hutchinson, J.R., Shefelbine, S.J., 2010. BoneJ: free and extensible bone image analysis in ImageJ. Bone 47, 1076e1079. Ebraheim, N.A., Sabry, F.F., Nadim, Y., 1999. Internal architecture of the talus: implication for talar fracture. Foot Ankle Int. 20, 794e796. Fajardo, R., Müller, R., 2001. Three-dimensional analysis of nonhuman primate trabecular architecture using micro-computed tomography. Am. J. Phys. Anthropol. 115, 327e336. Fisher, N.I., Lewis, T., Embleton, B.J.J., 1987. Statistical Analysis of Spherical Data. Cambridge University Press, Cambridge. Fritsch, H., Schmitt, O., Eggers, R., 1996. The ossification centre of the talus. Anat. Anzeiger. 178, 455e459. Gebo, D.L., 1992. Plantigrady and foot adaptation in African apes: implications for hominid origins. Am. J. Phys. Anthropol. 89, 29e58. Gebo, D.L., Schwartz, G.T., 2006. Foot bones from Omo: implications for hominid evolution. Am. J. Phys. Anthropol. 129, 499e511. Gosman, J.H., Ketcham, R.A., 2009. Patterns in ontogeny of human trabecular bone from SunWatch Village in the prehistoric Ohio Valley: general features of microarchitectural change. Am. J. Phys. Anthropol. 138, 318e332. Graham, D.J., Midgley, N.G., 2000. Graphical representation of particle shape using triangular diagrams: an excel spreadsheet method. Earth Surf. Proc. Landf. 25, 1473e1477. Green, D.J., Alemseged, Z., 2012. Australopithecus afarensis scapular ontogeny, function, and the role of climbing in human evolution. Science 338, 514e517. Griffin, N.L., D’Août, K., Ryan, T.M., Richmond, B.G., Ketcham, R.A., Postnov, A., 2010. Comparative forefoot trabecular bone architecture in extant hominids. J. Hum. Evol. 59, 202e213. Haile-Selassie, Y., Saylor, B.Z., Deino, A., Levin, N.E., Alene, M., Latimer, B.M., 2012. A new hominin foot from Ethiopia shows multiple Pliocene bipedal adaptations. Nature 483, 565e569. Harrigan, T.P., Mann, R.W., 1984. Characterization of microstructural anisotropy in orthotropic materials using a 2nd rank tensor. J. Mater. Sci. 19, 761e767. Hébert, D., Lebrun, R., Marivaux, L., 2012. Comparative three-dimensional structure of the trabecular bone in the talus of primates and its relationship to ankle joint loads generated during locomotion. Anat. Rec. 295, 2069e2088. Homminga, J., McCreadie, B., Ciarelli, T., Weinans, H., Goldstein, S., Huiskes, R., 2002. Cancellous bone mechanical properties from normals and patients with hip fractures differ on the structure level, not on the bone hard tissue level. Bone 30, 759e764. Hubbard, A.M., Meyer, J.S., Davidson, R.S., Mahboubi, S., Harty, M.P., 1993. Relationship between the ossification center and cartilaginous anlage in the normal hindfoot in children: study with MR imaging. Am. J. Roentgenol. 161, 849e853. Hunt, A., Smith, R., Torode, M., Keenan, A., 2001. Inter-segment foot motion and ground reaction forces over the stance phase of walking. Clin. Biomech. 16, 592e600. Hunt, K.D., 1991. Positional behavior in the Hominoidea. Int. J. Primatol. 12, 95e118. Hunt, K.D., Russon, A.E., Begun, D.R., 2004. The special demands of great ape locomotion and posture. In: Russon, A.E., Begun, D.R. (Eds.), The Evolution of Thought. Cambridge University Press, Cambridge, pp. 172e189. Hvid, I., Rasmussen, O., Jensen, N., Nielsen, S., 1985. Trabecular bone strength profiles at the ankle joint. Clin. Orthop. Relat. R. 199, 306e312. Judex, S., Garman, R., Squire, M., Donahue, L.R., Rubin, C., 2004. Genetically based influences on the site-specific regulation of trabecular and cortical bone morphology. J. Bone Miner. Res. 19, 600e606. Lamy, P., 1986. The settlement of the longitudinal plantar arch of some African PlioPleistocene hominids: a morphological study. J. Hum. Evol. 15, 31e46. Latimer, B., Ohman, J.C., Lovejoy, C.O., 1987. Talocrural joint in African hominoids: implications for Australopithecus afarensis. Am. J. Phys. Anthropol. 74, 155e175. Latimer, B., Lovejoy, C.O., 1989. The calcaneus of Australopithecus afarensis and its implications for the evolution of bipedality. Am. J. Phys. Anthropol. 78, 369e386. Lordkipanidze, D., Jashashvili, T., Vekua, A., de Leon, M.S.P., Zollikofer, C.P.E., Rightmire, G.P., Pontzer, H., Ferring, R., Oms, O., Tappen, M., Bukhsianidze, M., Agustí, J., Kahlke, R., Kiladze, G., Martinez-Navarro, B., Mouskhelishvili, A., Nioradze, M., Rook, L., 2007. Postcranial evidence from early Homo from Dmanisi, Georgia. Nature 449, 305e310. MacLatchy, L., Müller, R., 2002. A comparison of the femoral head and neck trabecular architecture of Galago and Perodicticus using micro-computed tomography (mCT). J. Hum. Evol. 43, 89e105. Maga, M., Kappelman, J., Ryan, T.M., Ketcham, R.A., 2006. Preliminary observations on the calcaneal trabecular microarchitecture of extant large-bodied hominoids. Am. J. Phys. Anthropol. 129, 410e417. Marchi, D., 2007. Relative strength of the tibia and fibula and locomotor behavior in hominoids. J. Hum. Evol. 53, 647e655. McDougall, I., Brown, F.H., Vasconcelos, P.M., Cohen, B.E., Thiede, D.S., Buchanan, M.J., 2012. New single crystal 40Ar/39Ar ages improve time scale for

A. Su et al. / Journal of Human Evolution 64 (2013) 667e677 deposition of the Omo Group, Omo-Turkana Basin, East Africa. J. Geol. Soc. Lond. 169, 213e226. Michelson, J., Checcone, M., Kuhn, T., Varner, K., 2001. Intra-articular load distribution in the human ankle joint during motion. Foot Ankle Int. 22, 226e233. Morio, C., Lake, M.J., Gueguen, N., Rao, G., Baly, L., 2009. The influence of footwear on foot motion during walking and running. J. Biomech. 42, 2081e2088. Müller, R., Koller, B., Hildebrand, T., Laib, A., Gianolini, S., Ruegsegger, P., 1996. Resolution dependency of microstructural properties of cancellous bone based on three-dimensional m-tomography. Technol. Health Care 4, 113. Odgaard, A., Kabel, J., van Rietbergen, B., Dalstra, M., Huiskes, R., 1997. Fabric and elastic principal directions of cancellous bone are closely related. J. Biomech. 30, 487e495. Pal, G., Routal, R., 1998. Architecture of the cancellous bone of the human talus. Anat. Rec. 252, 185e193. Pontzer, H., Lieberman, D.E., Momin, E., Devlin, M.J., Polk, J.D., Hallgrimsson, B., Cooper, D.M.L., 2006. Trabecular bone in the bird knee responds with high sensitivity to changes in load orientation. J. Exp. Biol. 209, 57e65. Potthast, W., Lersch, C., Segesser, B., Koebke, J., Brüggemann, G.-P., 2008. Intraarticular pressure distribution in the talocrural joint is related to lower leg muscle forces. Clin. Biomech. 23, 632e639. Raikin, S.M., Elias, I., Zoga, A.C., Morrison, W.B., Besser, M.P., Schweitzer, M.E., 2007. Osteochondral lesions of the talus: localization and morphologic data from 424 patients using a novel anatomical grid scheme. Foot Ankle Int. 28, 154e161. Ryan, T.M., Ketcham, R.A., 2002. The three-dimensional structure of trabecular bone in the femoral head of strepsirrhine primates. J. Hum. Evol. 43, 1e26. Ryan, T.M., Ketcham, R.A., 2005. Angular orientation of trabecular bone in the femoral head and its relationship to hip joint loads in leaping primates. J. Morphol. 265, 249e263. Ryan, T.M., Krovitz, G.E., 2006. Trabecular bone ontogeny in the human proximal femur. J. Hum. Evol. 51, 591e602. Sacco, I.C., Akashi, P.M., Hennig, E.M., 2010. A comparison of lower limb EMG and ground reaction forces between barefoot and shod gait in participants with diabetic neuropathic and healthy controls. BMC Musculoskel. Dis. 11, 24. Saparin, P., Scherf, H., Hublin, J.J., Fratzl, P., Weinkamer, R., 2011. Structural adaptation of trabecular bone revealed by position resolved analysis of proximal femora of different primates. Anat. Rec. 294, 55e67. Schmitt, D., Larson, S.G., 1995. Heel contact as a function of substrate type and speed in primates. Am. J. Phys. Anthropol. 96, 39e50.

677

Skedros, J.G., Hunt, K.J., Bloebaum, R.D., 2004. Relationships of loading history and structural and material characteristics of bone: development of the mule deer calcaneus. J. Morphol. 259, 281e307. Skedros, J.G., Sorenson, S.M., Hunt, K.J., Holyoak, J.D., 2007. Ontogenetic structural and material variations in ovine calcanei: a model for interpreting bone adaptation. Anat. Rec. 290, 284e300. Sode, M., Burghardt, A.J., Nissenson, R.A., Majumdar, S., 2008. Resolution dependence of the non-metric trabecular structure indices. Bone 42, 728e736. Stern, J.T., 2000. Climbing to the top: a personal memoir of Australopithecus afarensis. Evol. Anthropol. 9, 113e133. Tafforeau, P., Boistel, R., Boller, E., Bravin, A., Brunet, M., Chaimanee, Y., Cloetens, P., Feist, M., Hoszowska, J., Jaeger, J.-J., Kay, R.F., Lazzari, V., Marivaux, L., Nel, A., Nemoz, C., Thibault, X., Vignaud, P., Zabler, S., 2006. Applications of X-ray synchrotron microtomography for non-destructive 3D studies of paleontological specimens. Appl. Phys. A 83, 195e202. Takechi, H., Ito, S., Takada, T., Nakayama, H., 1982. Trabecular architecture of the ankle joint. Surg. Radiol. Anat. 4, 227e233. Tanck, E., Homminga, J., Van Lenthe, G., Huiskes, R., 2001. Increase in bone volume fraction precedes architectural adaptation in growing bone. Bone 28, 650e654. Tochigi, Y., Rudert, M.J., Saltzman, C.L., Amendola, A., Brown, T.D., 2006. Contribution of articular surface geometry to ankle stabilization. J. Bone Jt. Surg. Am. 88A, 2704e2713. Venkataraman, V.V., Kraft, T.S., Dominy, N.J., 2013. Tree climbing and human evolution. Proc. Natl. Acad. Sci. 110, 1237e1242. Vereecke, E., D’Aout, K., de Clercq, D., van Elsacker, L., Aerts, P., van Elsacker, L., 2003. Dynamic plantar pressure distribution during terrestrial locomotion of bonobos (Pan paniscus). Am. J. Phys. Anthropol. 120, 373e383. Wallace, I.J., Tommasini, S.M., Judex, S., Garland Jr., T., Demes, B., 2012. Genetic variations and physical activity as determinants of limb bone morphology: an experimental approach using a mouse model. Am. J. Phys. Anthropol. 148, 24e35. Wolf, P., List, R., Ukelo, T., Maiwald, C., Stacoff, A., 2009. Day-to-day consistency of lower extremity kinematics during walking and running. J. Appl. Biomech. 25, 369. Wood, B.A., 1974. Homo talus from East Rudolf, Kenya. J. Anat. 117, 203e204. Wood, B., Constantino, P., 2007. Paranthropus boisei: fifty years of evidence and analysis. Yearb. Phys. Anthropol. 50, 106e132. Woodcock, N.H., 1977. Specification of fabric shapes using an eigenvalue method. Geol. Soc. Am. Bull. 88, 1231e1236.