Fossil hominin shoulders support an African ape-like last common ancestor of humans and chimpanzees

Abstract

Reconstructing the behavioral shifts that drove hominin evolution requires knowledge of the timing, magnitude, and direction of anatomical changes over the past ∼6-7 million years. These reconstructions depend on assumptions regarding the morphotype of the Homo-Pan last common ancestor (LCA). However, there is little consensus for the LCA, with proposed models ranging from African ape to orangutan or generalized Miocene ape-like. The ancestral state of the shoulder is of particular interest because it is functionally associated with important behavioral shifts in hominins, such as reduced arboreality, high-speed throwing, and tool use. However, previous morphometric analyses of both living and fossil taxa have yielded contradictory results. Here, we generated a 3D morphospace of ape and human scapular shape to plot evolutionary trajectories, predict ancestral morphologies, and directly test alternative evolutionary hypotheses using the hominin fossil evidence. We show that the most parsimonious model for the evolution of hominin shoulder shape starts with an African ape-like ancestral state. We propose that the shoulder evolved gradually along a single morphocline, achieving modern human-like configuration and function within the genus Homo. These data are consistent with a slow, progressive loss of arboreality and increased tool use throughout human evolution.

Keywords: developmental simulation; geometric morphometrics; phylomorphospace; rotator cuff; scapula.

Fig. 1.

Alternative models of the hominin–panin last common ancestor (LCA). The branching pattern of living apes as inferred from genomic data is agreed upon; however, reconstruction of ancestral nodes differs among researchers. (A) In the African ape (AA) model, hominins derive from a knuckle-walking African ape-like LCA, typically conceived of as chimpanzee-like. (B and C) Ape convergence (AC) models hypothesize that the LCA is either a more generalized great ape or an unknown primitive ancestral Miocene ape. AQ, arboreal quadrupedalism. (D and E) Phylogenetic trees used to model these differences, with evolution from a common morphotype represented by a polytomy. Hominin phylogeny based on ref. .

Fig. 2.

Comparative and evolutionary scapula shape morphospace. (A and B) Representative hominoid scapulae and fossil hominins, shown scaled to identical vertebral border length. (C) Developmental simulations of DIK 1-1 (A. afarensis) using growth vectors from alternate proposed LCA morphotypes. (Top) Simulations. (Bottom) Simulations with DIK 1-1 scapula overlay (transparent blue), scaled to the same size to highlight shape differences. (D) PC1 and -2 morphospace with mean specimen warped along each axis to show associated shape changes. PC1 describes orientation of the spine relative to the blade whereas PC2 describes differences in the borders of the supraspinous fossa. Dashed arrows show the direct evolutionary vector from hypothetical ancestral states (AA, African ape mean; AC, MAA mean) to modern humans. Points, individual specimens; dark ellipses, 90% confidence interval (CI) of the mean; light ellipses, 90% CI of the sample; red points, DIK 1-1 developmental simulations. (E and F) Phylogenetic reconstruction for the AA (tree length = 0.085, P < 0.0001) and AC (tree length = 0.100, P < 0.0001) models illustrates alternative predictions for branching patterns and ancestral states within the morphospace.

Fig. S1.

Extended shape analyses. (A and B) Principal components analysis (PCA) using only the extant species. Phylogenetic reconstructions are shown as branching patterns, with ancestral nodes at branch joints. Dashed lines show the most direct evolutionary path between hypothesized ancestral and descendant states. AA, African ape model; AC, ape convergence model. (C) Canonical variates analysis (CVA) showing the first two axes. Distribution of species and associated shape vectors are similar to those identified in the PCA (Fig. 2) but show greater separation due to the methodology. (D) Phylogenetic reconstructions within the CVA morphospace using a method of squared-change parsimony on the CV1 and -2 axes while using alternative tree hypotheses (Fig. 1 D and E). Results show that a simple morphocline separates Pan and Homo in CV1 for the AA model, and a two-step process in CV1 and -2 for the AC model. (E) Results of the between-group principal components analysis (bgPCA) are largely identical to the PCA and CVA although there is reduced discrimination of African apes, suggesting that Pan and Gorilla may be even more similar when sample size effects are taken into account. (F) The phylogenetic reconstruction within the bgPCA morphospace is also concordant with other ordination results. Color coding is the same as in Fig. 2D. Red dots represent different developmental simulations of DIK 1-1 (A. afarensis).

Fig. S2.

Partial least squares (PLS) results. (A) Distribution of covariation between blade and spine shows that the majority of shape variation is associated with the first axis. (B) Plot of PLS1 axes for blade and spine variation shows a highly significant association of shape variation covariation (R_v = 0.633, P < 0.0001). (C) Associated PLS shape vector direction and magnitudes for positive (+0.10) and negative (−0.15) shown on the mean configuration (red, blade landmarks; blue, spine landmarks).

Fig. S3.

Extended PCA results. (A) Visualization of PC3 shape changes from positive (Top) to negative (Bottom). Gorilla and Pongo differ from all other taxa in having moderately wider scapula beyond those identified in PC1. (B) Three-dimensional scatterplot of PC1, PC2, and PC3. Dashed lines show approximate direct evolutionary vectors connecting hypothesized ancestral and descendant species for the AA and AC models. (C) Two-dimensional scatterplot of PC1 vs. PC3 with examples of Gorilla and Pongo shown. (D) Reconstructed phylogenetic trees from the PC1 and PC3 morphospace from Fig. 2. Note that the PC3 axis is reversed in A and B relative to the trees shown in C and D.

Fig. S4.

PC scores versus time. (A–D) Regressions of PC1 shape scores on both estimated divergence time (ancestral nodes) and geologic age (fossils) illustrate alternative predictions for the timing and strength of selection in the hominin lineage on spine and glenoid orientation.

Fig. S5.

Procrustes distance trees. (A) Tree generated using the unweighted pair group method with arithmetic mean (UPGMA) method. (B) Tree generated using the neighbor joining (NJ) method. In both cases, A. afarensis is associated with Gorilla whereas either H. ergaster or H. neanderthalensis is associated with H. sapiens. Procrustes distances and significance values are found in Table S1.

Fig. 3.

Model of shoulder shape evolution. Scapula morphospace is reconstructed at individual time horizons (t_n) for the phylogeny shown. The ancestral hominoid condition is reconstructed to be similar in shape and configuration to Pongo (t₀). Pongo shares with Lagothrix a penchant for slow, cautious movements through high forest canopy, including frequent bouts of pronograde suspensory locomotion (–61). This similarity suggests that derived “suspensory” postcranial characteristics of Pongo shared with other apes are partially convergent, consistent with evidence from Sivapithecus (16). In this model, hylobatids evolved a more cranially oriented spine and glenoid from this morphotype, which is predicted to fall in the intermediate space (t_1–2). African apes evolved a unique blade shape with cranial spine (t₁), subsequently diversifying into Pan and Gorilla lineages (t₂). Hominins retained the ancestral African ape blade shape, but the angle of the spine relative to the vertebral border gradually shifted (t_3–4), consistent with realignment of the shoulder musculature due to selection associated with more lateralized activities and/or reduced reliance on overhead activities (e.g., climbing).

Fig. S6.

Scapula landmarks. Descriptions and locations of the 3D landmarks used in this study are shown on a representative P. troglodytes scapula (Left) shown in dorsal and medial view.

Fig. S7.

Growth vectors used for developmental simulation. Multivariate regression of group shape on log centroid size for the taxa shown. Individual growth vectors were generated by performing the same multivariate regression analysis using the Procrustes coordinates from the group superimposition but performed on each species alone. The Procrustes coordinates of A. afarensis (DIK 1-1) were simulated at an adult size corresponding to an increase of log centroid size of 0.75, consistent with achieving an adult outcome (dashed ellipses) in each of the taxa used.