Skip to main content
NCBI home page
Journal of Anatomy logoLink to Journal of Anatomy
. 2012 Sep 5;221(5):394–405. doi: 10.1111/j.1469-7580.2012.01560.x

3D geometric morphometric analysis of the proximal epiphysis of the hominoid humerus

Julia Arias-Martorell 1, Josep Maria Potau 2, Gaëlle Bello-Hellegouarch 1, Juan Francisco Pastor 3, Alejandro Pérez-Pérez 1
PMCID: PMC3482347  PMID: 22946496

Abstract

In this study we perform a three-dimensional geometric morphometric (3D GM) analysis of the proximal epiphysis of the humerus in extant great apes, including humans, in order to accurately describe the functional anatomical differences between these taxa. In addition, a fossil hominin specimen of Australopithecus afarensis was included in a multivariate GM analysis in order to test the potential of this methodological approach for making locomotor inferences from fossil remains. The results obtained show significant differences in proximal humeral morphology among the taxa studied, which had thus far largely remained unnoticed. Based on morphofunctional considerations, these anatomical differences can be correlated to differences in the locomotor repertoires of the taxa, thus confirming that the proximal humerus is suitable for constructing paleobiological inferences about locomotion. Modern humans display markedly divergent features, which set them apart from both the extant great apes and the fossil hominin A. afarensis. The morphology of the proximal epiphysis of the humerus of the latter more closely resembles that of the orangutans, thus suggesting that despite hindlimb adaptations to bipedalism, the forelimb of this taxon was still functionally involved in arboreal behaviors, such as climbing or suspension.

Keywords: 3D geometric morphometrics, hominoidea, humerus, proximal epiphysis

Introduction

The glenohumeral joint and orthogrady in hominoids

The glenohumeral joint is a synovial spherical joint in which the glenoid cavity of the scapula articulates with the humeral head. Possible motions of the glenohumeral joint include virtually all combinations of movements: flexion–extension; abduction–adduction; and axial rotation (Larson, 1988). The principal structure of this joint is the articular surface of the humeral head, which is rounded and inflated, and consists of one-third of a sphere. It articulates directly with its reciprocal part in the scapula, the glenoid cavity, which is actually flattened and far more reduced than the humeral head, these differences in size and shape being the cause of the wide motion range of the joint. The proximal epiphysis of the humerus also contains the insertions of the four rotator cuff muscles that originate on the scapular fossae, and have their insertions on the greater and lesser tubercles of the humerus. The greater tubercle holds the insertions of the supraspinatus, infraspinatus and teres minor, whereas the lesser tubercle holds the insertion of the subscapularis (Aiello & Dean, ).

The most significant evolutionary changes undergone by humeral morphology in hominoid primates are related to the modifications experienced by the shoulder girdle while adapting to orthograde positional behaviors. All extant hominoids (hylobatids, pongids and hominids) display an orthograde body plan characterized by substantial modifications of the axial skeleton (vertebrae, ribs and girdles) as well as of the appendicular skeleton. These derived features, shared by all extant hominoids, include a transversely broad and dorsoventrally shallow rib cage, contrasting with the transversely narrow and dorsoventrally deep rib cages of the pronograde monkeys (Aiello & Dean, ; Larson, 1988; Ward, 1978b). Pronograde posture is the ancestral condition shared by primarily quadrupedal monkeys (Ward, 1978b, 1979; Ward et al. 1993), including the first hominoids – with the possible exception of Morotopithecus (MacLatchy et al. 2009a). Stem hominoids displayed a pronograde body plan suitable for generalized, above-branch quadrupedalism, with orthograde anatomical features having apparently evolved in a mosaic fashion in later hominoids (Ward, 1979; Almécija et al. ). The orthograde rearrangement of the axial skeleton provided the anatomical basis for the migration of the scapula onto the back of the rib cage (Aiello & Dean, ), resulting in a dorsally positioned scapula with a glenoid cavity facing ventro-laterally rather than ventrally (Inman et al. 2009; Aiello & Dean, ; Larson, 1988). Striking modifications have occurred in the Hominoidea scapular shape and anatomical disposition, which have enabled substantial functional modifications of the rotator cuff muscles (Inman et al. 2009; Ashton & Oxnard, 1990; Oxnard, 1996, 2000, 1967; Oxnard & Neely, 1968; Roberts, 2000; Larson & Stern, 2007a, 2007b; Crompton et al. 2001; Larson, 1947, 1945, 1988, 1989; Aiello & Dean, ; Potau et al. 2007). The proximal morphology of the humerus, as part of the glenohumeral joint, has been accordingly modified, although this anatomical region has received far less attention than the scapula in the literature.

The functional morphology of the hominoid proximal humerus

Amongst the morphological features of the proximal humerus in hominoids, humeral torsion has received the most attention in studies of functional morphology (Larson, 1947, 1995, 2007a; Cowgill, 1976; Rhodes, 2011). Traditionally, torsion is considered as an accommodation to the repositioning of the scapula onto the dorsal aspect of the transversely wide rib cage of the hominoids (Larson, 1995). When the forelimb is used in the parasagittal plane, a dorsally positioned scapula and a laterally facing glenoid cavity demand a certain amount of humeral torsion while maintaining the orientation of the elbow joint. Anatomically, torsion is the result of the migration of the greater and lesser tubercles of the humerus, not the result of proximal shaft torsion. Thus, torsion is mainly a by-product of an expansion of the articular surface of the humeral head (Fleagle & Simons, 1945; Rose, 1974; Aiello & Dean, ; Larson, 1988). This produces the displacement of the bicipital groove inwards (Aiello & Dean, ) and the relative constriction of the lesser tubercle in great apes, in comparison with the primitive pattern that lacks humeral torsion (Inman et al. 2009; Aiello & Dean, ), whereas there seems to be negligible variation in the size of the greater tubercle among hominoids (Inman et al. 2009; Fleagle & Simons, 1945; Rose, 1974; Larson, 1988). However, differences in humeral torsion between hominoids suggest that other factors participate in the degree of torsion of each species (Larson, 1947). Humeral torsion among hominoids seems to have been independently acquired in order to meet their locomotor needs (Larson, 1947; Rose, 1974). In quadrupedal knuckle-walking behaviors, humeral torsion is necessary to maintain fully functional arms in the parasagittal plane, as an accommodation to the quadrupedal posture in animals with a dorsally placed scapula (Larson, 1947, 1988). Gibbons present the least degree of torsion among hominoids, and this condition reflects a compromise between the changes caused by the repositioning of the scapula and the position of the elbow of the abducted arm during brachiation (Larson, 1947). In humans, torsion seems to be related to manipulative skills, although differences in humeral torsion exist between modern western and non-western populations, the latter showing lower degrees of humeral torsion and, therefore, a higher degree of humeral retroversion (Larson, 2007b). There is some confusion concerning the use of the terms humeral torsion and humeral retroversion in the literature (Larson, 2007a; Rhodes, 2011). In the primitive condition of the humerus, the proximal epiphysis faces the posterior aspect of the shaft, a condition that is present in non-hominoid primates (Larson, 1995, 2007a). Humeral torsion is a derived feature measured by the degree to which the humeral shaft rotates in a medial direction (Evans & Krahl, 1999; Krahl & Evans, 2011; Krahl, 2009; Larson, 1995). In contrast, humeral retroversion consists of a modified condition of the already twisted humerus in which it is reoriented posteriorly, causing the ‘reduction’ of torsion. Thus, a reduction of humeral retroversion would in fact increase humeral torsion and vice versa (Larson, 2007a,b; Rhodes, 2011).

The proximal epiphysis of the humerus is further characterized by the presence of an extensive articular surface, which enhances the movement in the articulation. The most important feature of this complex is the degree to which the tubercles are rotated to allow for additional articular surface in the transverse plane (Corruccini & Ciochon, 1986; Fleagle & Simons, 1945; Rose, 1974; Aiello & Dean, ; Larson, 1988). An extensive, inflated articular surface is present in the humeral head of hominoids, especially in Hylobates (Jungers, 1944), enabling hominoids to achieve a substantial degree of abduction–adduction and axial rotation even when the joint is in a fully flexed position (Rose, 1974).

Mechanical models of the upper limb have been proposed in many studies (Tuttle & Basmajian, 1978a,b; Larson & Stern, 2007a, 2007b; Schmitt, 1993). According to Larson & Stern (2007a), in arm-raising arboreal primates the supraspinatus is the primary muscle responsible for preventing the proximal displacement of the humeral head caused by the deltoid. However, in motions combining abduction and lateral rotation, the infraspinatus acts as primary synergist of the deltoid. The ability of infraspinatus to contribute to abduction seems to be related to its proximo-laterally oriented insertion in the greater tubercle in hominoids, whereas more terrestrial primates that do not use arm-raising in their primary form of locomotion, such as the cercopithecoids, seem to have more laterally facing insertions for the infraspinatus (Larson, 1989). Larson (2007b) also reported that shape differences in the subscapularis facet reflect a functional differentiation of this muscle affecting the degree of glenohumeral mobility: long and narrow facets show a greater functional differentiation than rounder facets. This is related to the manner in which tendinous fibers of the subscapularis insert on the lesser tubercle: those on the most proximal part of the tubercle seem to be involved in abduction and medial rotation, whereas those on the distal parts are involved in adduction and medial rotation. In this respect, hominoids display a greater degree of functional differentiation, showing the longest and narrowest facets, whilst cercopithecoids seem to have a less specialized subscapularis, as they are reported to show round insertion facets (Larson, 1989).

Geometric morphometrics (GM) and joint morphology

Three-dimensional (3D) numerical studies of proximal humeral morphology have been attempted. However, they have mainly been conducted in the prosthetic design field (Boileau & Walch, 1992; Tanaka, 1984; Aroonjarattham et al. 2009), and none have applied 3D GM techniques, as is the case with other postcranial elements, such as the articular region of the proximal femur (Harmon, 1982, 2007; Holliday et al. 2008), the proximal metatarsal shape (Proctor et al. 2008; Proctor, ), the shape of the proximal tibia (Turley et al. 1985), elbow morphology (Drapeau, 2010), and studies of joint reciprocity (Harcourt-Smith et al. 1998). Here, we present a 3D GM analysis of the proximal epiphysis of the humerus, using newly-defined landmark points, in order to evaluate and quantify differences in the shape of the humeral head between various extant great ape genera, and to try to relate these differences to their respective locomotor repertoires. So far, the proximal epiphysis of the humerus has not been considered an area with the potential for discriminating locomotor adaptations among humans and their hominin ancestors. However, the available literature clearly indicates that locomotor factors determine to some extent the morphology of this anatomical region (Corruccini & Ciochon, 1986; Fleagle & Simons, 1945; Jungers, 1944; Rose, 1974; Larson, 1988, 1989). Thus, the aim of this study was to determine whether humans (Homo sapiens), chimpanzees (Pan troglodytes), gorillas (Gorilla gorilla) and orangutans (Pongo pygmaeus) can be adequately discriminated from each other based on their proximal humeral morphology, so far poorly characterized from a functional perspective but here evaluated by means of a 3D GM approach. Finally, the relevance of this approach as a proxy of locomotor behavior characterization in fossil hominini was tested through an analysis of the proximal humeral head shape of Australopithecus afarensis.

Materials and methods

The Hominoidea sample included in this study consisted of 34 proximal epiphyses of the humerus: ten H. sapiens (five males and five females); nine P. troglodytes (five males and four females); nine G. gorilla (five males and four females); and six P. pygmaeus (one male and five females; Fig. 1; Table 1). The H. sapiens sample, of known age and sex, was obtained from the Unit of Human Anatomy and Embryology of the University of Barcelona (UB, Spain). The non-human samples came from the Museum of Natural Sciences of Barcelona (Barcelona, Spain) and the Anatomical Museum of the University of Valladolid (UVA, Spain). Specimens from both museums were bred in captivity at various zoological parks in Spain and died of natural causes. Captive primates usually exhibit locomotor repertoire frequencies that differ from those in the wild, and thus the anatomical interpretations derived from this sample might require confirmation with wild-bred specimens. Only clean, well-preserved, left humeri of adult individuals were selected, based on museum records when available, or determined by the full epiphyseal fusion of the humerus. High-quality molds were made with a silicon-based impression material (Speedex Putty polysiloxane condensation-type silicon elastomer by Coltène/Whaledent), which included the bicipital groove and the greater and lesser tubercles on the proximal epiphysis of the humerus. High-quality, positive epoxy casts (Ferropur, Feroca Composites) were obtained from the negative molds. The cast (Bone Clones Hominid Series, ref. SC-036D) of the partial left humerus of specimen A.L.-288-1 of A. afarensis (3.2 ma, Hadar, Ethiopia; Johansson et al. 2010) was included in the analysis.

Fig. 1.

Fig. 1

Open in a new tab

Proximal epiphyses of the humeri of the analyzed taxa (from left to right): Gorilla gorilla, Pan troglodytes, Homo sapiens, Pongo pygmaeus and the cast of Australopithecus afarensis, A.L. 288-1r.

Table 1.

Hominoid species analyzed, sample sizes and locomotor pattern classification

Species Male Female N Locomotion
Gorilla gorilla 5 4 9 Terrestrial knuckle-walking
Pan troglodytes 5 4 9 Terrestrial knuckle-walking /arboreal arm swinging
Pongo pygmaeus 1 5 6 Arboreal quadrumanous/arm swinging
Homo sapiens 5 5 10 Bipedalism
Australopithecus afarensis 0 1 1 Bipedalism vs. knuckle-walkingvs. climbing/suspension
Total 16 19 35
Open in a new tab

All the casts were scanned using a laser Next Engine 2000 surface scanner (NextEngine, 401 Wilshire Blvd, Santa Monica, CA 90401, USA) at a scanning resolution of 40 K (2×) points inch−2, with 0.005-inch point separation (ScanStudio HD software, NextEngine). The scanned surfaces were edited and processed with the ScanStudio HD CORE software package (NextEngine, 2006). The resulting scans were saved in Ply file format, which consists of triangular mesh data with additional information about color and texture. This file format was suitable for Landmark Editor (Wiley, ), the software for 3D landmark data acquisition that provided the landmark coordinates required by MorphoJ (Klingenberg, 2007) for 3D GM analyses.

GM: landmark selection and setting

The landmark-based GM procedure is an effective method for capturing information about the shape of an organism and for statistically testing shape differences between groups. It is also a powerful and intuitive visualization method as data are recorded to capture the geometry of the structure under study (Rohlf & Marcus, 1999; Zelditch et al. 2004). Landmarks are multidimensional (2D or 3D) coordinates of morphological markers fitted by superimposition methods, which match homologous landmarks as closely as possible by minimizing the Procrustes summed squared distances between corresponding landmarks, so shape changes can be described as the residuals of the superimposition explained as transformation vectors (Rohlf & Marcus, 1999).

In this study, a total of 26 newly defined, single-point type III (Bookstein et al. 1997; O'Higgins, 1986) landmarks were selected (Fig. 2; Table 2). Although landmark homology cannot be absolutely demonstrated, there is geometric equivalence between them across analyzed specimens (Bookstein et al. 1997; O'Higgins, 1986; Harmon, 1982; Holliday et al. 2008). Six landmarks were defined in the articular surface of the humeral head: L1, the closest point to the bicipital groove on the articular perimeter; L2, the most proximal point of the articular perimeter, close to the supraspinatus insertion; L3, the most distal point of the articular perimeter; L4, the middle point between L2 and L3; L5, the most ventral point of the articular perimeter defined by the intercept of the perpendicular line to L2–L3 through L4; and L6, the most dorsal point of the articular perimeter defined by the intercept of the perpendicular line to L2–L3 through L4. Five landmark points were selected on each insertion area (proximal, distal, medial, lateral and central) of the rotator cuff muscles –subscapularis L7–L11, supraspinatus L12–L16, infraspinatus L17–L21 and teres minor L22–L26 – considering that the supraspinatus muscle insertion is the most proximal one on the greater tubercle.

Fig. 2.

Fig. 2

Open in a new tab

Location of the defined landmark points of the proximal epiphysis of the humerus in (A) superior view, (B) full articular surface view, (C) anterior view and (D) posterior view.

Table 2.

Landmark numeration, categorization (after Bookstein et al. 1997; O'Higgins, 1986) and description

Landmark Type Description
L1 III Closest point to the bicipital groove on the articular perimeter
L2 III Most proximal point of the articular perimeter, close to the supraspinatus insertion
L3 III Most distal point of the articular perimeter
L4 III Middle point between L2 and L3
L5 III Most ventral point of the articular perimeter defined by the intercept of the perpendicular line to L2–L3 through L4
L6 III Most dorsal point of the articular perimeter defined by the intercept of the perpendicular line to L2–L3 through L4
L7 III Most distal point of the subscapularis insertion
L8 III Most proximal point of the subscapularis insertion
L9 III Midpoint between L7 and L8
L10 III Farthest point to humeral head defined by the intercept with the subscapularis insertion perimeter of a perpendicular line to L7–L8 through point L9
L11 III Closest point to humeral head defined by the intercept with the subscapularis insertion perimeter of a perpendicular line to L7–L8 through point L9
L12 III Most anterior point of the supraspinatus insertion
L13 III Most posterior point of the supraspinatus insertion
L14 III Midpoint between L12 and L13
L15 III Farthest point to humeral head defined by the intercept with the supraspinatus insertion perimeter of a perpendicular line to L12–L13 through point L14
L16 III Closest point to humeral head defined by the intercept with the supraspinatus insertion perimeter of a perpendicular line to L12–L13 through point L14
L17 III Most distal point of the infraspinatus insertion
L18 III Most proximal point of the infraspinatus insertion
L19 III Midpoint between L17 and L18
L20 III Farthest point to humeral head defined by the intercept with the infraspinatus insertion perimeter of a perpendicular line to L17–L18 through point L19
L21 III Closest point to humeral head defined by the intercept with the infraspinatus insertion perimeter of a perpendicular line to L17–L18 through point L19
L22 III Most distal point of the teres minor insertion
L23 III Most proximal point of the teres minor insertion
L24 III Midpoint between L22 and L23
L25 III Farthest point to humeral head defined by the intercept with the teres minor insertion perimeter of a perpendicular line to L22–L23 through point L24
L26 III Closest point to humeral head defined by the intercept with the teres minor insertion perimeter of a perpendicular line to L22–L23 through point L24
Open in a new tab

Data analysis

The landmark coordinates of the extant hominoid specimens were analyzed with MorphoJ 3D GM, and then the A. afarensis specimen coordinate information was added to the extant primate dataset for comparison. In both cases, the coordinate raw data subsets extracted from Landmark Editor software were transferred to the MorphoJ statistical package (Klingenberg, 2007) and analyzed. The first step in the morphometric analysis (Kendall, 1982; Bookstein, 1985; Rohlf & Marcus, 1999; Marcus et al. 2000; Dryden & Mardia, 2008; O'Higgins, 1986; Zelditch et al. 2004) consisted of carrying out a General Procrustes Analysis (GPA) that rotated, scaled and superimposed all forms (registration), explaining the differences between them in terms of landmark displacement vectors relative to this registration (Rohlf, 2001; O'Higgins, 1986; Zelditch et al. 2004). The resulting GPA displacement vectors were ordinated by Principal Components Analysis (PCA), reducing the complex multidimensional data into a few eigenvectors that were linear combinations of the landmark displacements, which explained shape variation among samples (O'Higgins, 1986; Zelditch et al. 2004; Klingenberg, 2007).

A Multivariate Regression (MR) analysis of shape on centroid size (CS) was performed to test the amount of shape variance that could be explained by size differences, with Principal Components (PC), indicative of shape, being the dependent variable and size (CS or log CS) the independent one. The null hypothesis predicted that shape was independent of size; MR determines the percentage of variance explained by size and tests its significance (O'Higgins, 1986; Zelditch et al. 2004; Klingenberg, 2007). MorphoJ allows MRs with a permutation test of 1000 randomizations to be performed, and gives the option of pooling the regression within given subgroups as an external variable. A pooled within-groups regression is the ideal way to perform this test when dealing with groups where there must be a correction for size effect among them. We chose to conduct the test with the within-species pooled variable option including four groups: G. gorilla, P. troglodytes, P. pygmaeus and H. sapiens. To explore differences between groups, a Canonical Variate Analysis (CVA) was conducted. Even though PCA and CVA are similar methods to ordinate and explore patterns within the whole sample, the PCA computes variables suitable for examining variation among individuals, whereas CVA creates variables to assess the relative positions of groups in a sample (Zelditch et al. 2004). In our study, group sample sizes were smaller than the number of landmarks per individual. In these cases, CVA tends to overestimate the differences between groups (Klingenberg, 2007). Therefore, we performed a Discriminant Analysis (DA), a statistical method implemented to assure the reliability of group differences found in CVA. The implementation in MorphoJ used Fisher's classification rule, and the reliability of the discrimination between groups was tested with a leave-one-out, jack-knife cross-validation method. Parametric T-square tests, also with permutation, were performed to compare group means within the DA in MorphoJ (Klingenberg, 2007).

Finally, the morphological affinities of the humeral proximal epiphysis of A. afarensis A.L. 288-1r fossil specimen with those of the extant hominoids were analyzed for a preliminary exploration of shape with a PCA.

Results

Analysis of the extant hominoid subset

The PCA for the extant hominoid subset yielded 33 eigenvectors, of which the first three components explained 66.4% of total variance (48.2% PC1, 11.3% PC2 and 6.9% PC3). The remaining eigenvectors were poorly representative (< 6% each) of total variance. The plot of PC1 vs. PC2 (59.5%) showed clear differences between groups, with non-overlapping 95% equiprobable ellipses of group means for most comparisons (Fig. 3). Homo sapiens, with negative values for PC1, clearly differed from G. gorilla and P. troglodytes, which overlapped for PC1 but showed significant differences for PC2. Pongo pygmaeus showed an intermediate position in the graph, around the 0 value for both components, overlapping with neither the H. sapiens nor the Pan/Gorilla clusters. PC3 (Fig. 3) did not discriminate between groups, although P. pygmaeus showed slightly smaller values than the other groups.

Fig. 3.

Fig. 3

Open in a new tab

Scatter plots of the first three PCs – PC1 vs. PC2 (upper plot) and PC1 vs. PC3 (lower plot) – derived from the PCA within MorphoJ. The ellipses include 95% confidence intervals of the means (group centroids) of the extant hominoid samples analyzed: Gorilla gorilla (red rhomboids); Homo sapiens (green dots); Pan troglodytes (dark blue stars); and Pongo pygmaeus (pale blue triangles). The PCA was separately computed including Australopithecus afarensis (asterisk) as a group in order to explore its morphological affinities with the other hominoid groups, clearly clustering within P. pygmaeus dispersion range. The inclusion of A. afarensis in the PC analysis caused almost no displacement of the position of the specimens in the other groups, so only one PCA figure is shown for simplicity.

Multivariate Regression tests (with permutations) between PC and the CS showed that PC1 was not correlated with CS (P = 0.2023), and only 3.05% of total shape variation was attributable to size for the first PC. PC2 was not correlated with CS either (P = 0.1900), although 13.1% of the shape variation was predicted by size for this component, as was also the case for PC3 (P = 0.0685), with 9.9% of shape variation correlated to size differences. Thus, no significant effect of size upon shape was observed.

The CVA provided three eigenvectors explaining 100% of total variance (77.1% CV1, 14.2% CV2 and 8.2% CV3). Permutation tests run over Mahalanobis and Procrustes distance matrices showed significant differences between all group pairs, with P < 0.0015 in all tests (Table 3). Leave-one-out cross-validations showed that the post-hoc probabilities of correct classification did not decrease in the Homo–Pan comparison (100%), and showed a maximum reduction (−27.3%) in the Gorilla–Pan one (Table 4). The plot of CV1 vs. CV2 (Fig. 4) showed a clear differentiation between most of the groups considered. Homo fell on the positive side of the CV1 axis, apart from all the other groups and showing the lowest dispersion (95% ellipse of the sample mean) of all groups. Again, Pongo showed intermediate values between Pan, Gorilla and Homo, although it showed CV1 values closer to those of Pan and Gorilla, groups that could be discriminated for CV2. The first axis (CV1) clearly separated Homo from the rest of the Hominoidea, as well as to a certain extent Pongo from Pan and Gorilla, but showed no differentiation between the two knuckle-walkers. CV2, instead, clearly set Pan apart, on the positive side of the axis, from Gorilla, on the negative one, leaving Pongo and Homo undifferentiated. CV3 showed a clear differentiation of Pongo from the other taxa (Fig. 4). Finally, the DA with cross-validation test, conducted to check the consistency of the differences between pairs of groups, showed significant differences after randomizations for both the Mahalanobis and Procrustes distances for all comparisons except between Gorilla and Pan, for which significant differences were found for the Procrustes distance (P < 0.01) but not for the T-square test (P = 0.092).

Table 3.

Significance values (P) of the Mahalanobis and Procrustes distances between groups using 1000 permutations

Mahalanobis Procrustes
Species Gorilla Homo Pongo Gorilla Homo Pongo
Homo < 0.0001 – – < 0.0001 – –
Pongo 0.0002 < 0.0001 – 0.0004 0.0001 –
Pan < 0.0001 < 0.0001 0.0001 0.0013 < 0.0001 0.0008
Open in a new tab

Table 4.

Percentages of post-hoc correct classifications from the discriminant functions and after leave-one-out cross-validation. (1) From the discriminant CVA; (2) after cross-validation; (3) decrease in correct classification % due to reduced sample sizes

(1) (2) (3)
Gorilla gorilla–Homo sapiens 100 94.8 −5.2
Gorilla gorilla–Pongo pygmaeus 93.4 73.4 −19.6
Gorilla gorilla–Pan troglodytes 94.5 66.7 −27.3
Homo sapiens–Pongo pygmaeus 87.5 68.8 −18.3
Homo sapiens–Pan troglodytes 100 100 0
Pongo pygmaeus–Pan troglodytes 93.4 86.7 −6.3
Open in a new tab

Fig. 4.

Fig. 4

Open in a new tab

Scatter plots showing the dispersion of analyzed specimens and groups (95% confidence interval ellipses of group mean) for CV1 vs. CV2 (upper plot) and CV1 vs. CV3 (lower plot). Note that Homo sapiens (green dots) is the most homogeneous group, whereas the other hominoids show greater dispersions. Pongo pygmaeus (pale blue triangles) shows the highest dispersion. CV1 differentiates Homo and Pongo from the other Hominoidea, CV2 distinguishes the two knuckle-walkers, and CV3 discriminates Pongo from both the knuckle-walkers and humans.

Shape differences between extant hominoids

Shape differences can be described from the CVA by plotting shape transformations between pairs of taxa (Fig. 5). Shape changes along CV1 between Pongo and the knuckle-walkers (Fig. 5a) showed a greater inwards rotation of the articular head of the humerus of Pongo compared with the knuckle-walkers (Fig. 5a-1). The relative positions of the muscle insertion areas on the greater tubercle also differed in the two groups, with Pongo having an aligned disposition of the three attachments, and the distance between the humeral head and the insertions was more similar, due to their progressive proximal–distal separation with respect to the humeral head (Fig. 5a-3). In contrast, knuckle-walkers did not display this alignment of the insertion sites, with the supraspinatus and teres minor insertions being closer to the humeral head, while the infraspinatus insertion was clearly separated from it (Fig. 5a-3). Pongo pygmaeus also showed greater proximity between the subscapularis and supraspinatus insertion areas, resulting in a narrowing of the bicipital groove (Fig. 5a-2). Shape changes along CV1 between Homo and Pongo (Fig. 5b) showed the accentuation in humans of the same inwards rotation of the humeral head affecting the orangutan (Fig. 5b-1). This provides the human articular surface with its characteristic oval shape, by increasing the proximo-distal diameter, whilst that of the orangutan preserves a greater dorso-ventral axis (Fig. 5b-1). Regarding the tubercles, Homo displayed even more aligned insertion areas of the great tubercle compared with Pongo, and also a greater separation between the humeral head and the supraspinatus and teres minor insertions (Fig. 5b-3). Homo showed a markedly higher proximity between the insertion areas of the subscapularis and supraspinatus, thus displaying a further narrowing of the bicipital groove (Fig. 5b-2).

Fig. 5.

Fig. 5

Open in a new tab

Shape changes between group centroids derived from the CVA. The superimposition views (right side) corresponding to the orientations (1, 2 and 3) of the humeri (left side): (1) superior; (2) bicipital groove and (3) major tubercle with humeral head profile. Comparisons shown correspond to: (A) CV1 shape changes between Pongo (red continuous outline) and knuckle-walkers together (blue dotted outline); (B) CV1 shape changes between Pongo (red continuous outline) and Homo (blue dotted outline); (C) CV2 shape changes between Gorilla (red continuous outline) and Pan (blue dotted outline); (D) CV3 shape changes between Pongo (red continuous outline) and all other hominoids together (blue dotted outline). The orientations of the humerus shown have been chosen to optimally highlight the differences between the taxa.

The distribution of groups along CV1 (Fig. 4) seemed to vary depending on the degree of inwards rotation of the humeral head, with knuckle-walkers (blue dotted outline in Fig. 5a-1) showing a less rotated surface than Pongo (red continuous outline in Fig. 5a-1), and Homo (blue outline in Fig. 5b-1) having a more rotated one than Pongo (red continuous outline in Fig. 5b-1). Similarly, a progressive enlargement of the greater tubercle can be observed (Fig. 5a-1, b-1), caused by a greater alignment of the insertion areas in Pongo and, especially, in Homo. This alignment seems to be related to a progressive separation between the humeral head and the insertion areas of supraspinatus and teres minor in both species (Fig. 5a-3, b-3), while the subscapularis insertion maintains its position with respect to the humeral head in the knuckle-walkers and shows a slight displacement towards the great tubercle in Pongo and Homo (Fig. 5a-2, b-2). Together, the rotation of the humeral head and the displacement of the subscapularis have resulted in a narrowing of the bicipital groove, mainly in humans. Shape changes along CV2 (Fig. 4) showed a clear differentiation between Pan and Gorilla, with Pan (blue dotted outline in Fig. 5c-1) showing an inwards rotated humeral head compared with the Gorilla (red outline in Fig. 5c-1), despite the infraspinatus insertion area being similarly positioned in both (Fig. 5c-3). In addition, compared with the Gorilla (Fig. 5c-3), Pan showed a greater separation between the humeral head and the insertion area of the supraspinatus, a greater proximity between the humeral head and the teres minor insertion, and a somewhat larger size of the subscapularis insertion area, despite the position of the subscapularis insertion and the width of the bicipital groove being similar in both (Fig. 5c-2). Finally, sample distribution along CV3 (Fig. 4) explained the larger and more inflated humeral articular head of Pongo (red continuous outline in Fig. 5d) compared with the other hominoid species considered (Fig. 5d-2, d-3).

Comparison of A. afarensis with the extant hominoids

Bone Clones osteological replicas are considered artistic renderings for instructive purposes, not reliable as scientifically accurate reproductions. Thus, before using the A.L. 288-1r cast for comparative purposes, 2D measurements of the proximal humerus epiphysis were obtained from the cast (averaging five repeated measures) and compared with those provided by Johansson et al. (2010) for the original fossil. Deviations of measurements between the original and the cast were computed for two linear measurements and one index: (1) antero-posterior diameter of the head (APD: original 26.8 mm, cast 26.02 mm; −2.91%); (2) medio-lateral diameter of the head (MLD: original 31.1 mm, cast 29.58 mm; −4.89%); and (3) head shape index (MLD/APD: original 1.1604, cast 1.1369; −2.03%). The humeral head MLD/APD index showed a deviation of about 2%, the cast being shorter in the medio-lateral dimension. This deviation, though not negligible, can be considered minimal for the purpose of this preliminary comparison of the shape of the humerus and is unlikely to greatly affect the overall interpretation. Further analyses with more abundant and reliable fossil casts will be required, but the inclusion of this fossil specimen clearly shows the potential of this approach for fossil analyses.

The PCA performed after the GPA including the A. afarensis specimen yielded 34 PCs; the first component (PC1) explained 47.4% of total variance (somewhat below the 48.2% obtained if A. afarensis was not included), the second (PC2) 11.4% (11.3% if not included), and the third (PC3) 6.9% (identical to that not including it). As in the previous analysis, the remaining PCs explained less than 6% each. The GM of the cast of A. afarensis, compared with the extant analyzed hominoids, clearly fell within the 95% equiprobable ellipse of the Pongo sample (Fig. 3), away from both Homo and the knuckle-walkers, with intermediate values for PC1, which was the most discriminant component.

Discussion

The greater degree of inwards rotation of the humeral head of Homo and Pongo corresponds to a major dorsal extension of the articular surface that could be related to functional demands of external rotation of the glenohumeral joint in both taxa. In Pongo pygmaeus, external rotation is especially important during the swing phase of vertical climbing (Larson & Stern, 2007a), while in H. sapiens the external rotation retards contact between the greater tubercle and the acromion during elevation of the arm in the scapular plane (Inman et al. 2009; Basmajian & De Luca, 1963). Moreover, the inwards rotation also results in an increase in the proximo-distal diameter of the humeral head of H. sapiens as well as in an increase in its distal region. This could be related to the neutral position of the humerus when the arm is lowered, as the glenoid cavity of the scapula articulates mainly with the distal region of the humeral head when the arm is in a resting position (Kapandji, 1988). Pongo pygmaeus seems to be less specialized in this respect, but arm-raising involves the contact of the glenoid cavity with the central and lateral surfaces of the humeral head (Kapandji, 1988), so the large and inflated articular surface of Pongo pygmaeus might be a derived anatomical characteristic for enhancing arboreal capabilities, increasing the mobility of the glenohumeral joint (Rose, 1974; Aiello & Dean, ). The knuckle-walkers display a less marked inwards rotation of the humeral head, with a ventral expansion that allows greater amplitude of internal rotation motion of the glenohumeral joint, which is a very important feature in knuckle-walking behavior (Larson & Stern, 2007b). However, P. troglodytes shows greater inwards rotation compared with G. gorilla, which could be related to a higher degree of arboreal locomotion in the former (Larson & Stern, 2007b).

Regarding the tubercles, in P. troglodytes and G. gorilla, the placement of the infraspinatus insertions were comparable, thus suggesting a similar function of this muscle in both species, acting as a stabilizer of the glenohumeral joint and preventing dorsal displacement of the humerus during the stance and support phases of knuckle-walking, at least in P. troglodytes, as no convincing electromyography studies have been conducted in G. gorilla (Tuttle & Basmajian, 1978a,b; Larson & Stern, 2007a, 2007b). Both knuckle-walkers are also characterized by the position of the supraspinatus insertion near the humeral head, when compared with Pongo pygmaeus and H. sapiens. This can be related to the similar function of the supraspinatus as a stabilizer of the glenohumeral joint, together with the infraspinatus, during the stance phase of knuckle-walking (Larson & Stern, 2007a, 2007b). This role as a stabilizer of the glenohumeral joint played by the supraspinatus in P. troglodytes and G. gorilla also results in an increase in the mass of this muscle as seen in the greater size of the supraspinous fossa with respect to the infraspinous fossa, as observed by Schultz (1989) and Roberts (2000).

Compared with knuckle-walkers, Pongo pygmaeus showed a supraspinatus insertion that was more separated from the humeral head. This could be related to a more important function as an elevator of the upper limb of this muscle in Pongo pygmaeus, in accordance with its more arboreal locomotion. The less important role played by the supraspinatus as a stabilizer of the glenohumeral joint in Pongo pygmaeus, compared with the knuckle-walkers, results in a lower relative mass of this muscle, as evidenced by the relatively lower values of the size of the supraspinous fossa reported by Roberts (2000). In this respect, it is important to note that the more arboreal P. troglodytes presents a supraspinatus insertion that is more separated from the humeral head than the more terrestrial G. gorilla, in addition to a lesser relative mass of this muscle (Potau et al. 2007) and a relatively smaller supraspinous fossa (Roberts, 2000). The teres minor recruitment for humeral adduction in arm swinging (Tuttle & Basmajian, 1978a,b; Larson & Stern, 2007a) might be the cause of the unexpected shape and placement of this muscle insertion in Pongo pygmaeus. It is relatively larger and farther placed, relative to the humeral head, than in knuckle-walkers. The enhancement of this muscle might be related to the more arboreal behavior of Pongo pygmaeus, as hoisting and arm swinging are basic mechanisms for a below branch locomotion (Tuttle & Basmajian, 1978a,b; Larson & Stern, 2007a). The infraspinatus insertion displays a similar position with respect to the humeral head both in Pongo pygmaeus and in the knuckle-walkers. Nonetheless, Larson & Stern (2007a) reported the infraspinatus to be an essential muscle for resisting transarticular tensile stresses during pendant suspension and arm swinging. Pongo pygmaeus shared with G. gorilla and P. troglodytes the location and morphology of the subscapularis insertion, a muscle that is extremely important in the stabilization of the glenohumeral joint during the support phase of climbing (Larson & Stern, 2007a), as well as during quadrupedal and knuckle-walking stance phases (Tuttle & Basmajian, 2011).

Homo sapiens showed a morphological pattern of the insertions of the greater tubercle that was radically different from those of the knuckle-walkers and, to a lesser extent, Pongo pygmaeus. As in Pongo pygmaeus, the separation between the supraspinatus insertion area and the humeral head could be related to the relatively smaller size of the supraspinous fossa in H. sapiens (Roberts, 2000; Potau et al. 1969, 1969). This morphology enhances the important function of the supraspinatus muscle as an elevator of the arm (Inman et al. 2009; Basmajian & De Luca, 1963). The increased distance between the teres minor insertion area and the humeral head in humans might be related to the important role of this muscle as external rotator of the glenohumeral joint during elevation of the arm in the scapular plane (Inman et al. 2009; Basmajian & De Luca, 1963).

Although 3D GM comparisons allowed for functional interpretations of morphological variations in humeral head shape and muscle insertion positions, metric comparisons of muscle insertion positions and scapular sizes and shapes are required to test the accuracy of the functional significance presented here.

Evolutionary perspectives

The reported differences in the location of the muscle insertions in the proximal epiphysis of the humerus between the various hominoid species studied, including humans, along with their respective distinctive morphology of the humeral head, provide the required framework for enlightening locomotor inferences in the fossil record. The attempted analysis of the proximal humeral epiphysis of A.L. 288-1 provides useful preliminary information on the locomotor behavior of the extinct species A. afarensis. Its clearly divergent morphology from that of modern humans and its allocation within the Pongo pygmaeus confidence interval in the PCA analysis together provide significant morphometric evidence suggesting that A. afarensis might have practiced arboreal suspensory behaviors, as suggested by various authors (Vrba, 1979; Susman et al. 2003, 1934; Crompton et al. 2007), although contested by others (Latimer et al. 1989; Latimer & Lovejoy, 1987; Latimer, 1986).

These alternative hypotheses have distinct evolutionary implications. The resemblances shown here of the morphological patterns of the proximal epiphysis of the humerus between Pongo pygmaeus and A. afarensis favor the view that bipedal hominins evolved from a generalized arboreal ancestor (McHenry, 2009b; Thorpe et al. 1999; Crompton et al. 2007, 1987; Lovejoy et al. 2009a,b), instead of from a knuckle-walking ancestor, as argued by researchers who support the homology (synapomorphy) of knuckle-walking adaptations in chimpanzees and gorillas (Begun, 2009; Richmond & Strait, 2010; Corruccini & McHenry, 1999; Richmond et al. 2007; Williams, 2010). The greater similarities between Pongo and Gorilla than between Pongo and Pan shown here, suggest that the derived condition of Pan might have appeared independently from that of Gorilla, which might support the contention that knuckle-walking could have evolved twice (Larson, 1995; Dainton & Macho, 2008; Kivell & Schmitt, 1984). However, further and more detailed analyses of fossil apes and hominins (including Ardipithecus) would certainly help to shed new light on this debate. Our results also indicate that modern humans display an extremely derived condition of the glenohumeral joint, which would have adapted to manipulative behaviors requiring greater speed and precision. African apes, on the contrary, display entirely different derived conditions as a result of adaptations to a highly specialized suite of locomotor activities requiring enhanced stability at the glenohumeral joint during knuckle-walking in a joint complex primarily adapted to arboreal behaviors (Larson & Stern, 2007b).

Concluding remarks

The 3D GM study of the proximal epiphysis of the humerus has revealed morphological differences between the extant hominoids studied. These differences reflect adaptations of the shoulder girdle to the locomotor behaviors displayed by each taxon. The comparative morpho-functional analysis of the human proximal epiphysis of the humerus showed that humans display highly derived features, which separate them from all other hominoids, mainly related to the increase in speed and precision required for manipulative behaviors. Orangutans showed a distinct functional anatomy of the proximal epiphysis of the humerus, indicative of locomotor adaptations to arboreal behaviors, whereas gorillas and chimpanzees displayed independently derived morphologies, probably due to their mainly terrestrial locomotor behaviors. However, Pan showed a morphology reflecting both terrestrial and arboreal conditions, although more adapted to arboreality than Gorilla.

Acknowledgments

We thank Eulàlia Garcia (Museum of Natural Sciences of Barcelona) for providing access to specimens under her care, as well as Beatriz Pinilla for her advice on methodology. The authors would also like to thank the two anonymous reviewers for their useful comments, which helped improve the earlier versions of this manuscript. This research was funded by the Spanish Ministry of Education and Science (CGL2010-15340 and CGL2011-22999 to APP), the Generalitat de Catalunya (DURSI 2009SGR-00884 to APP), and the Predoctoral Fellowship Grant Program of the U. of Barcelona (APIF-UB2009/10 to JAM).

References

  1. Aiello L, Dean C. An Introduction to Evolutionary Anatomy. London: Academic Press; 1990. [Google Scholar]
  2. Almécija S, Alba DM, Moyà-Solà S. Pierolapithecus and the functional morphology of Miocene ape hand phalanges: paleobiological and evolutionary implications. J Hum Evol. 2009;57:284–297. doi: 10.1016/j.jhevol.2009.02.008. [DOI] [PubMed] [Google Scholar]
  3. Aroonjarattham P, Jiamwatthanachai P, Mahaisavariya B, et al. Three-dimensional morphometric study of the Thai proximal humerus: cadaveric study. J Med Assoc Thai. 2009;92:1191–1197. [PubMed] [Google Scholar]
  4. Ashton EH, Oxnard CE. The musculature of the primate shoulder. Trans Zool Soc Lon. 1963;29:553–650. [Google Scholar]
  5. Basmajian JV, De Luca CJ. Muscles Alive: Their Functions Revealed by Electromyography. Baltimore: Williams & Wilkins; 1985. [Google Scholar]
  6. Begun DR. Miocene fossil hominids and the Chimp-Human clade. Science. 1992;257:1929–1933. doi: 10.1126/science.1411507. [DOI] [PubMed] [Google Scholar]
  7. Boileau P, Walch G. Three-dimensional geometry of the proximal humerus. J Bone Joint Surg Br. 1997;79:857–865. doi: 10.1302/0301-620x.79b5.7579. [DOI] [PubMed] [Google Scholar]
  8. Bookstein FL. Size and shape spaces for landmark data in two dimensions. Stat Sci. 1986;1:181–222. [Google Scholar]
  9. Bookstein FL, Schäfer K, Prossinger H, et al. Comparing frontal cranial profiles in archaic and modern Homo by morphometric analysis. Anat Rec. 1999;257:217–224. doi: 10.1002/(SICI)1097-0185(19991215)257:6<217::AID-AR7>3.0.CO;2-W. [DOI] [PubMed] [Google Scholar]
  10. Corruccini RS, Ciochon RL. Morphometric affinities of the human shoulder. Am J Phys Anthropol. 1976;45:19–38. doi: 10.1002/ajpa.1330450104. [DOI] [PubMed] [Google Scholar]
  11. Corruccini RS, McHenry H. Knuckle-walking hominid ancestors. J Hum Evol. 2001;40:507–511. doi: 10.1006/jhev.2001.0473. [DOI] [PubMed] [Google Scholar]
  12. Cowgill LW. Humeral torsion revisited: a functional and ontogenetic model for populational variation. Am J Phys Anthropol. 2007;134:472–480. doi: 10.1002/ajpa.20689. [DOI] [PubMed] [Google Scholar]
  13. Crompton RH, Lieberman SS, Oxnard CE. Morphometrics and niche metrics in prosimian locomotion: an approach to measuring locomotion, habitat, and diet. Am J Phys Anthropol. 1987;73:149–177. doi: 10.1002/ajpa.1330730203. [DOI] [PubMed] [Google Scholar]
  14. Crompton RH, Vereecke EE, Thorpe SKS. Locomotion and posture from the common hominoid ancestor to fully modern hominins, with special reference to the last common panin/hominin ancestor. J Anat. 2008;212:501–543. doi: 10.1111/j.1469-7580.2008.00870.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Crompton RW, Sellers WI, Thorpe SKS. Arboreality, terrestriality and bipedalism. Phil Trans Royal Soc B. 2010;365:3301–3314. doi: 10.1098/rstb.2010.0035. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Dainton M, Macho GA. Did knuckle-walking evolved twice? J Hum Evol. 1999;36:171–194. doi: 10.1006/jhev.1998.0265. [DOI] [PubMed] [Google Scholar]
  17. Drapeau MSM. Articular morphology of the proximal ulna in extant and fossil hominoids and hominins. J Hum Evol. 2008;55:86–102. doi: 10.1016/j.jhevol.2008.01.005. [DOI] [PubMed] [Google Scholar]
  18. Dryden I, Mardia KV. Statistical Shape Analysis. New York: John Wiley; 1998. [Google Scholar]
  19. Evans GF, Krahl VE. The torsion of humerus: a phylogenetic survey from fish to man. Am J Anat. 1945;76:303–337. [Google Scholar]
  20. Fleagle JG, Simons EL. The humerus of Aegyptopithecus zeuxis: a primitive anthropoid. Am J Phys Anthropol. 1982;59:175–193. doi: 10.1002/ajpa.1330590207. [DOI] [PubMed] [Google Scholar]
  21. Harcourt-Smith WE, Tallman M, Frost S, et al. Analysis of selected hominoid joint surfaces using laser scanning and geometric morphometrics: a preliminary report. In: Sargis EJ, Dagosto M, editors. Mammalian Evolutionary Morphology: A Tribute to Frederik S. Szalay. Berlin: Springer; 2008. pp. 373–383. [Google Scholar]
  22. Harmon EH. The shape of the hominoid proximal femur: a geometric morphometric analysis. J Anat. 2007;210:170–185. doi: 10.1111/j.1469-7580.2006.00688.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Harmon EH. The shape of the early hominin proximal femur. Am J Phys Anthropol. 2009;139:154–171. doi: 10.1002/ajpa.20966. [DOI] [PubMed] [Google Scholar]
  24. Holliday TW, Hutchinson VT, Morrow MMB, et al. Geometric morphometric analyses of hominid proximal femora: taxonomic and phylogenetic considerations. HOMO. 2010;61:3–15. doi: 10.1016/j.jchb.2010.01.001. [DOI] [PubMed] [Google Scholar]
  25. Inman VT, Saunders JB, Abbot LC. Observations on the function of the shoulder joint. J Bone Joint Surg Br. 1944;26:1–30. [Google Scholar]
  26. Johansson DC, Lovejoy CO, Kimbel WH, et al. Morphology of the Pliocene partial hominid skeleton (A.L. 288-1) from the Hadar Formation, Ethiopia. Am J Phys Anthropol. 1982;57:403–451. [Google Scholar]
  27. Jungers WL. Relative joint size and hominoid locomotor adaptations with implications for the evolution of hominid bipedalism. J Hum Evol. 1988;17:247–265. [Google Scholar]
  28. Kapandji IA. The Physiology of the Joints, Volume 1: Upper Limb. Churchill Livingstone: Edinburgh; 2007. [Google Scholar]
  29. Kendall DG. Shape manifolds, procrustean metrics, and complex projective spaces. Bull London Math Soc. 1984;16:81–121. [Google Scholar]
  30. Kivell TL, Schmitt D. Independent evolution of knuckle-walking in African apes shows that humans did not evolve from a knuckle-walking ancestor. PNAS. 2009;106:14 241–14 246. doi: 10.1073/pnas.0901280106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Klingenberg CP. MorphoJ: an integrated software package for geometric morphometrics. Mol Ecol Resour. 2011;11:353–357. doi: 10.1111/j.1755-0998.2010.02924.x. [DOI] [PubMed] [Google Scholar]
  32. Krahl VE. The torsion of humerus: its localization, cause and duration in man. Am J Anat. 1947;80:275–319. doi: 10.1002/aja.1000800302. [DOI] [PubMed] [Google Scholar]
  33. Krahl VE, Evans GF. Humeral torsion in man. Am J Phys Anthropol. 1945;3:229–253. [Google Scholar]
  34. Larson SG. Subscapularis function in Gibbons and Chimpanzees: implications for interpretation of humeral head torsion in hominoids. Am J Phys Anthropol. 1988;76:449–462. [Google Scholar]
  35. Larson SG. Role of supraspinatus in the quadrupedal locomotion of Vervets (Cercopithecus aethiops): implications for interpretation of humeral morphology. Am J Phys Anthropol. 1989;79:369–377. doi: 10.1002/ajpa.1330790313. [DOI] [PubMed] [Google Scholar]
  36. Larson SG. Functional morphology of the shoulder in primates. In: Gebo DL, editor. Postcranial Adaptation in Nonhuman Primates. DeKalb, Illinois: Northern Illinois University Press; 1993. pp. 45–69. [Google Scholar]
  37. Larson SG. New characters for the functional interpretation of primate scapulae and proximal humeri. Am J Phys Anthropol. 1995;98:13–35. doi: 10.1002/ajpa.1330980103. [DOI] [PubMed] [Google Scholar]
  38. Larson SG. Estimating humeral torsion on incomplete fossil anthropoid humeri. J Hum Evol. 1996;31:239–257. [Google Scholar]
  39. Larson SG. Notes and comments on the definition of humeral torsion: a comment on Rhodes (2006) Am J Phys Anthropol. 2007a;133:819–821. doi: 10.1002/ajpa.20505. [DOI] [PubMed] [Google Scholar]
  40. Larson SG. Evolutionary transformation of the hominin shoulder. Evol Anthropol. 2007b;16:172–187. [Google Scholar]
  41. Larson SG, Stern JT. EMG of Scapulohumeral muscles in the Chimpanzee during reaching and “arboreal” locomotion. Am J Anat. 1986;176:171–190. doi: 10.1002/aja.1001760207. [DOI] [PubMed] [Google Scholar]
  42. Larson SG, Stern JT. EMG of Chimpanzee shoulder muscles during knuckle-walking: problems of terrestrial locomotion in a suspensory adapted primate. J Zool. 1987;212:629–655. [Google Scholar]
  43. Latimer BM. Locomotor adapatations in Australopithecus afarensis: the issue of arboreality. In: Coppens Y, Senut B, editors. Origine(s) de la Bipédie Chez les Hominidés. Paris: CRNS; 1991. pp. 234–244. [Google Scholar]
  44. Latimer BM, Lovejoy CO. The calcaneus of Australopithecus afarensis and implications for the evolution of bipedality. Am J Phys Anthropol. 1989;78:369–386. doi: 10.1002/ajpa.1330780306. [DOI] [PubMed] [Google Scholar]
  45. Latimer BM, Ohman JC, Lovejoy CO. Talocrural joint in African hominoids: implications for Australopithecus afarensis. Am J Phys Anthropol. 1987;74:155–175. doi: 10.1002/ajpa.1330740204. [DOI] [PubMed] [Google Scholar]
  46. Lovejoy CO, Simpson SW, White TD, et al. Careful climbing in the Miocene: the forelimbs of Ardipithecus ramidus and humans are primitive. Science. 2009a;326:70e1–70e8. [PubMed] [Google Scholar]
  47. Lovejoy CO, Suwa G, Simpson SW, et al. The Great Divides: Ardipithecus ramidus reveals the postcrania of our last common ancestor with African Apes. Science. 2009b;326:101–106. [PubMed] [Google Scholar]
  48. MacLatchy LM, Gebo D, Kityo R, et al. Postcranial functional morphology of Morotopithecus bishopi, with implications for the evolution of modern ape locomotion. J Hum Evol. 2000;39:159–183. doi: 10.1006/jhev.2000.0407. [DOI] [PubMed] [Google Scholar]
  49. Marcus L, Corti M, Loy A, et al., editors. Advances in Morphometrics. New York: Plenum Press; 1996. [Google Scholar]
  50. McHenry H. The first bipeds: a comparison of the A. afarensis and A. africanus postcranium and implications for the evolution of bipedalism. J Hum Evol. 1986;15:177–191. [Google Scholar]
  51. O'Higgins P. The study of morphological variation in the hominid fossil record: biology, landmarks and geometry. J Anat. 2000;197:103–120. doi: 10.1046/j.1469-7580.2000.19710103.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Oxnard CE. The functional morphology of the primate shoulder as revealed by comparative anatomical, osteometric and discriminant function techniques. Am J Phys Anthropol. 1967;26:219–240. [Google Scholar]
  53. Oxnard CE. The architecture of the shoulder in some mammals. J Morphol. 1968;126:249–290. doi: 10.1002/jmor.1051260302. [DOI] [PubMed] [Google Scholar]
  54. Oxnard CE. Evolution of the human shoulder: some possible pathways. Am J Phys Anthropol. 1969;30:319–331. doi: 10.1002/ajpa.1330300302. [DOI] [PubMed] [Google Scholar]
  55. Oxnard CE, Neely PM. The descriptive use of neighborhood limited classification in functional morphology: an analysis of the shoulder in primates. J Morphol. 1969;129:127–148. doi: 10.1002/jmor.1051290202. [DOI] [PubMed] [Google Scholar]
  56. Potau JM, Bardina X, Ciurana N. Subacromial space in African Great Apes and subacromial impingement syndrome in humans. Int J Primatol. 2007;28:865–880. [Google Scholar]
  57. Potau JM, Bardina X, Ciurana N, et al. Quantitative analysis of the deltoid and rotator cuff muscles in humans and great apes. Int J Primatol. 2009;30:697–708. [Google Scholar]
  58. Potau JM, Artells R, Bello G, et al. Expression of myosin heavy chain isoforms in the supraspinatus muscle of different primate species: implications for the study of the adaptation of primate shoulder muscles to different locomotor modes. Int J Primatol. 2011;32:931–944. [Google Scholar]
  59. Proctor DJ. University of Iowa; 2010. Three-dimensional morphometrics of the proximal metatarsal articular surfaces of Gorilla, Pan, Hylobates, and shod and unshod humans. Unpublished PhD Dissertation. [Google Scholar]
  60. Proctor DJ, Broadfield D, Proctor K. Quantitative three-dimensional shape analysis of the proximal hallucial metatarsal articular surface in Homo Pan Gorilla and Hylobates. Am J Phys Anthropol. 2008;135:216–224. doi: 10.1002/ajpa.20723. [DOI] [PubMed] [Google Scholar]
  61. Rhodes JA. Humeral torsion and retroversion in the literature: a reply to Larson. Am J Phys Anthropol. 2007;133:820–821. [Google Scholar]
  62. Richmond BG, Strait DS. Evidence that humans evolved from a knuckle-walking ancestor. Nature. 2000;404:382–385. doi: 10.1038/35006045. [DOI] [PubMed] [Google Scholar]
  63. Richmond BG, Begun DR, Strait DS. Origin of human bipedalism: the knuckle-walking hypothesis revisited. Yrbk Phys Anthropol. 2001;116:70–105. doi: 10.1002/ajpa.10019.abs. [DOI] [PubMed] [Google Scholar]
  64. Roberts D. Structure and function of the primate scapula. In: Jenkins F, editor. Primate Locomotion. New York: Academic Press; 1974. pp. 171–200. [Google Scholar]
  65. Rohlf FJ. Shape statistics: Procrustes superimposition and tangent spaces. J Classif. 1999;16:197–223. [Google Scholar]
  66. Rohlf FJ, Marcus L. A revolution in morphometrics. Trends Ecol Evol. 1993;8:129–132. doi: 10.1016/0169-5347(93)90024-J. [DOI] [PubMed] [Google Scholar]
  67. Rose MD. New postcranial specimens of Catarrhines from Middle Miocene Chinji Formation, Pakistan: descriptions and a discussion of proximal humeral functional morphology in anthropoids. J Hum Evol. 1989;18:131–162. [Google Scholar]
  68. Schmitt D. Substrate size and primate forelimb mechanics: implications for understanding the evolution of primate locomotion. Int J Primatol. 2003;24:1023–1036. [Google Scholar]
  69. Schultz AH. Some distinguishing characters of the mountain gorilla. J Mammal. 1934;15:51–61. [Google Scholar]
  70. Susman RL, Stern JT, Jungers WL. Arboreality and bipedality in the Hadar hominids. Folia Primatol. 1984;43:113–156. doi: 10.1159/000156176. [DOI] [PubMed] [Google Scholar]
  71. Susman RL, Stern JT, Jungers WL. Locomotor adaptations in the Hadar hominids. In: Delson E, editor. Ancestors: The Hard Evidence. New York: Alan R. Liss; 1985. pp. 184–192. [Google Scholar]
  72. Tanaka H. Numerical analysis of the proximal humeral outlines: bilateral shape differences. Am J Hum Biol. 1999;11:343–357. doi: 10.1002/(SICI)1520-6300(1999)11:3<343::AID-AJHB6>3.0.CO;2-H. [DOI] [PubMed] [Google Scholar]
  73. Thorpe SKS, Holder RL, Crompton RH. Origin of human bipedalism as an adaptation for locomotion on flexible branches. Science. 2007;316:1328–1331. doi: 10.1126/science.1140799. [DOI] [PubMed] [Google Scholar]
  74. Turley K, Guthrie EH, Frost SR. Geometric morphometric analysis of tibial shape and presentation among Catarrhine taxa. Anat Rec. 2011;294:217–230. doi: 10.1002/ar.21307. [DOI] [PubMed] [Google Scholar]
  75. Tuttle RH, Basmajian JV. Electromyography of Pongid shoulder muscles II. Deltoid, rhomboid and rotator cuff. Am J Phys Anthropol. 1978a;49:47–56. doi: 10.1002/ajpa.1330490109. [DOI] [PubMed] [Google Scholar]
  76. Tuttle RH, Basmajian JV. Electromyography of Pongid shoulder muscles III. Quadrupedal positional behavior. Am J Phys Anthropol. 1978b;49:57–70. doi: 10.1002/ajpa.1330490110. [DOI] [PubMed] [Google Scholar]
  77. Vrba ES. A new study of the scapula of Australopithecus from Sterkfontein. Am J Phys Anthropol. 1979;51:117–130. [Google Scholar]
  78. Ward CV. Torso morphology and locomotion in Proconsul nyanzae. Am J Phys Anthropol. 1993;92:291–328. doi: 10.1002/ajpa.1330920306. [DOI] [PubMed] [Google Scholar]
  79. Ward C. Postcranial and locomotor adaptations of hominoids. In: Henke W, Tattersall I, editors. Handbook of Paleoanthropology Part 2. Berlin: Springer; 2007. pp. 1011–1030. [Google Scholar]
  80. Ward CV, Walker A, Teaford MF, et al. Partial skeleton of Proconsul nyanzae from Mfangano Island, Kenya. Am J Phys Anthropol. 1993;90:77–111. doi: 10.1002/ajpa.1330900106. [DOI] [PubMed] [Google Scholar]
  81. Wiley DF. Landmark Editor 3.0. IDAV, University of California, Davis, ( http://graphics.idav.ucdavis.edu/research/ EvoMorph)
  82. Williams SA. Morphological integration and the evolution of knuckle-walking. J Hum Evol. 2006;58:432–440. doi: 10.1016/j.jhevol.2010.03.005. [DOI] [PubMed] [Google Scholar]
  83. Zelditch ML, Swiderski DL, Sheets HD, et al. Geometric Morphometrics for Biologists: A Primer. New York: Academic Press; 2004. [Google Scholar]

Articles from Journal of Anatomy are provided here courtesy of Anatomical Society of Great Britain and Ireland

RESOURCES