If you read about Homo habilis you will find out that they retained
arboreal features.
https://www.ncbi.nlm.nih.gov/pubmed/18711733
QUOTE:
Abstract
The Homo habilis OH 62 partial skeleton has played an important,
although controversial role in interpretations of early Homo locomotor
behavior. Past interpretive problems stemmed from uncertain bone length
estimates and comparisons using external bone breadth proportions, which
do not clearly distinguish between modern humans and apes. Here, true
cross-sectional bone strength measurements of the OH 62 femur and
humerus are compared with those of modern humans and chimpanzees, as
well as two early H. erectus specimens-KNM-WT 15000 and KNM-ER 1808. The
comparative sections include two locations in the femur and two in the
humerus in order to encompass the range of possible section positions in
the OH 62 specimens. For each combination of section locations, femoral
to humeral strength proportions of OH 62 fall below the 95% confidence
interval of modern humans, and for most comparisons, within the 95%
confidence interval of chimpanzees. In contrast, the two H. erectus
specimens both fall within or even above the modern human distributions.
This indicates that load distribution between the limbs, and by
implication, locomotor behavior, was significantly different in H.
habilis from that of H. erectus and modern humans. When considered with
other postcranial evidence, the most likely interpretation is that H.
habilis, although bipedal when terrestrial, still engaged in frequent
arboreal behavior, while H. erectus was a completely committed
terrestrial biped. This adds to the evidence that H. habilis (sensu
stricto) and H. erectus represent ecologically distinct, parallel
lineages during the early Pleistocene.
END QUOTE:
Ron Okimoto