The Pleistocene locality of Kanjera, Western Kenya: stratigraphy, chronology and paleoenvironments

Abstract

Kanjera is well known as the source of controversial hominid fossils collected by L. S. B. Leakey in the 1930s. Since 1935, the context of fossils and artifacts from the locality has been in doubt, due to a claim that sediment slumping had commingled materials from stratigraphic units of different ages. A careful re-examination of the geology demonstrates that the Kanjera deposits consist of approximately 37 m of volcaniclastic, fluvial, mudflat and lacustrine sediments that we assign to three major units: the Kanjera Formation, the Apoko Formation, and the Black Cotton Soil. Outcrops cover approximately 2 km2 in two adjacent areas, the Northern and Southern Exposures. Fossils and artifacts are found in primary contexts through much of the stratigraphic column, and extensive trenching failed to reveal any sediment slumping that would have disturbed these contexts. Faulting, rapid lateral facies changes, and an erosional unconformity between the Kanjera and Apoko Formations result in complex geological relationships. Magnetostratigraphic and faunal determinations indicate that the Kanjera Formation is approximately 1·5-0·5 Ma, the Apoko Formation younger than 0·5 Ma, and the Black Cotton Soil latest Pleistocene to Holocene. The hominid sample is derived from the Black Cotton Soil except for Leakey's Hominid 3, which probably was an intrusive burial into Kanjera Formation Bed KN-2. The Theropithecus oswaldi type specimen originated from KN-2a and is dated between 1·1 and at most 1·76 Ma. The Kanjera Formation provides the youngest known records of Metridiochoerus andrewsi and Deinotherium bozasi at about 1·0 Ma.