Positional behaviour of chimpanzees living in the savannah-mosaic environment of Issa Valley, Tanzania: Insights to the origins of human bipedalism

Habitual bipedal walking is unique to humans amongst primates and its associated morphological features are used to define the human clade (hominins) from other apes over the last 7 million years (Ma). Yet, the evolutionary origins of our bipedal gait remain unknown. The palaeontological record supports an adaptive hominin radiation from a closed (e.g., tropical forest) to a more open and seasonal heterogeneous environment (e.g., savannah-mosaic) as central to the emergence and evolution of terrestrial bipedalism in the human lineage [1]. However, morphological features that are advantageous for arboreal locomotion are present in the forelimbs of many hominins (early and late), generating long-standing debate around the functional significance of these features, and the link between bipedalism and terrestriality. In the absence of direct fossil evidence, extant chimpanzees that live across a habitat gradient provide ideal models to test the “savannah-landscape effect” on ape locomotor behaviour and substrate use [2-3]. Chimpanzee locomotor studies to date, however, have focused only on forest-dwelling communities [4-5], limiting our knowledge of the full range of chimpanzee locomotor behaviour and its application for modelling hominin evolution. Here, we characterize for the first time the positional behaviour and substrate-use of chimpanzees living in an open, dry habitat. Chimpanzees (P. t. schweinfurthii) of the Issa Valley, western Tanzania, live in a savannah-mosaic habitat dominated by open miombo woodland with strips of closed evergreen (riparian) forest: a mosaic that resembles the reconstructed palaeoenvironments of Pliocene hominins. To investigate the influence of an open habitat on positional behaviour and terrestriality we, 1) quantified the frequency of terrestrial and arboreal positional behaviours between the forest and woodland at Issa, and 2) compared our findings to data published on forest-dwelling chimpanzees. Specifically, we tested the hypothe