Scapular breadth does not discriminate suspension from clambering in hominoids: A response to Spear and Williams

Abstract

Spear and Williams (2018) are incorrect in stating that we (Selby & Lovejoy, 2017) argued that "a mediolaterally narrow scapula is unrelated to suspensory locomotion and therefore. . .early crown hominoids could not have been suspensory." We clearly stated that scapular breadth is related to suspension, but is not a primary determinant of suspension. We posited instead that scapular narrowing was a fundamental change in the bauplan of primates that abandoned strict above branch quadrupedality in favor of more diverse arboreal locomotor postures, especially above and below branch palmigrade clambering.And, of course, early crown hominoids could have been suspensory; the point of import is, however, that they were most likely not. As we have argued elsewhere, the morphology of the wrist, elbow, foot, hip, and thigh, as well as the limb proportions of Ardipithecus, all point to a included long bone lengths (approximating Spear and Williams' GMusing femoral head, proximal tibia breadth, anteroposterior and proximodistal humeral head breadths, distal humeral articular surface breadth, lengths of the femur, tibia, humerus, and radius, SWGM) and one that did not (i.e., only joint dimensions of the same bones) to compare to body mass. When each was log transformed for linearity, the joint surface GM better approximated body mass (r 5 0.986 vs. 0.978, both p < 0.05). Not surprisingly, there is a similar pattern for mediolateral breadth and spine angle when either size correction method is used (compare Figure 1a, b).Spear and Williams also point to a significant correlation between size-corrected mediolateral breadth and spine angle in their atelid sample, suggesting that both characters serve as adaptations for suspension. We found a similar correlation between these two variables in both ceboids (r 5 20.844, p < 0.05) and cercopithecoids (r 5 20.652, p < 0.05), but not in hominoids (r 5 0.134, p > 0.05). Such a lack of association between blade shape and spine orientation in hominoids (see also Green, Spiewak, Seitelman, & Gunz, 2016) suggests that mediolateral narrowing and spine angulation evolved independently, as we argued previously.We tested this lack of association further by employing a second, but equally reasonable vicar for scapular breadth: the transect from the scapula's superior angle to its glenoid (Figure 1c, 1d). Note that FIG URE 1 Mediolateral blade breadth and spine angulation in anthropoids. Spine angle (SPINANG) is compared to (a) scapular breadth from spine at vertebral border to glenoid (SPGLN) normalized by a geometric mean (GM) of fore-and hind-limb joint dimensions (see Selby and Lovejoy 2017) and by (b) a GM approximating that used by Spear and Williams (SWGM); (c) superior border (SUPBOR) length normalized by GM, and (d) by vertebral border (VERTBOR). Regression lines and R 2 values are provided in a and b for each superfamily. Scapular breadth does not discriminate hominoids based on either metric or method of normalization 2 | SELBY AND LOVEJOY 198 SELBY AND LOVEJOY normalization of this ...