That is the question asked by a recemt paper in Folia Primatologica. I don’t have access but Discovery News has the story. Apparently, the researchers analyzed the capitates of a number of hominoids and hominins. According to Discovery News:
The scientists observed that the Australopithecus anamensis wrist bones exhibited pressure loads associated with modern arboreal animals. The analyzed Australopithecus afarensis bones conversely showed stress loads comparable to those of more terrestrial species, including modern humans.
The researchers concluded that the important shift in early hominid lifestyle happened around the time when A. afarensis first emerged.
It’s likely that Australopithecus anamensis walked on the ground at times too, but Macho [the lead author - afarensis] points out that “form follows function.”
For those who have access the paper can be found here and this is the abstract:
This pilot study explored whether the redirection of stress through trabeculae within morphologically constrained capitates provides information about habitual/positional behaviours unavailable from the study of external morphology alone. To assess this possibility, an experimental finite element approach was taken, whereby no attempt was made to reconstruct the actual magnitudes and loading conditions experienced by the capitates in vivo. Rather, this work addressed fundamental biological questions relating to bone plasticity, i.e. internal versus external bone morphology. The capitates of 7 species with different and – in the case of fossils – inferred locomotor behaviours were selected. Virtual models of capitates were created, scaled to the same size and subjected to the same theoretical load. In the first set of analyses, models were assigned the material properties of bone throughout, whereas in the second set, models were assigned 11 different material properties representing the trabecular architecture derived from high-resolution CT. Species with arboreal behaviours consistently redirected loads towards the ulnar aspect of the capitate when trabeculae were introduced, while terrestrial species, and the bipedal Homo, redirected stress towards the radial side. From these preliminary analyses, it is tentatively concluded that Australopithecus anamensis habitually engaged in arboreal behaviours, whereas Australopithecus afarensis did not.