R. Potts et al., Paleolandscape variation and Early Pleistocene hominid activities: Members1 and 7, Olorgesailie Formation, Kenya, J HUM EVOL, 37(5), 1999, pp. 747-788
Paleolandscape research tests for variation in the spatial distribution of
hominid artefacts and establishes the association of hominid activities wit
h paleoenvironmental features over distances of 100s to 1000s of meters. Th
is approach requires (1) precise definition of narrow stratigraphic interva
ls based on sedimentary criteria that can be documented over a broad area,
and (2) excavation of these intervals in order to establish taphonomic and
paleoenvironmental contexts. In this report, excavations of three target in
tervals within the early Pleistocene deposits (992 to 780 ka) of the Olorge
sailie basin are described. Assessment of time-averaging and paleolandscape
structure shows that each target interval represents a relatively brief pe
riod (less than or equal to 1000 yrs) and exhibits a unique distribution of
environmental features (e.g., topographic gradients, channels, soil develo
pment). Stone artefacts and fossilized animal bones are distributed nonrand
omly in each interval, and include clusters that were five to 293 times mor
e densely concentrated than the laterally equivalent background scatter. A
paleosol in upper Member 1 preserves a relatively continuous distribution o
f artefacts and fossils, in contrast with the more patchy distribution in t
wo intervals of lower Member 7. We infer that the difference between the tw
o members reflects a real variation in hominid land use-either a response t
o local environmental differences or perhaps a change through time in homin
id interaction with the environment. By expanding the comparative analysis
to diverse basins, it should be possible to test for broader evolutionary c
hange in hominid activities. Examples drawn from East African Pliocene and
early Pleistocene sites suggest that evolutionary change in land use entail
ed (1) wider ranging of hominids and longer distances of stone transport, (
2) expansion of tool-assisted behaviors to a wider diversity of environment
al settings, and (3) more strongly focused placement of particular artefact
forms (e.g., bifaces) in different areas of the landscape in response to s
pecific environmental features, such as lava outcrops, stream channels, and
lake margins.