Archaeology

3.1 Excavations

Figure 3.1 – Conjoining remains from Omo I recovered from the 1967 and 2001 field excavations at KHS.

Through archival photos, film footage, field maps from the original expedition and extensive surveys, the team was able to locate the site at which the Omo I and II fossils were recovered in 1967. These included KHS (Kamoya’s Hominid Site) and PHS (Paul’s Hominid Site), as well as two new sites dubbed AHS (Awoke’s Hominid Site) and BNS (Bird’s Nest Site). KHS, AHS and BNS were excavated and surface collected. PHS, from which Omo II derived, yielded some scattered stone tools around the site at lower elevations, but did it not produce any additional fossil remains.

An abundance of fossils in Member I deposits supports the original conclusion that Omo II derived from the upper part of Member I. Surface collection and re-excavation of KHS, the findspot of Omo I, revealed new skeletal elements that complimented the original Omo I finds, including a femoral shaft fragment recovered in 2001 that articulated neatly with the distal end of a femur recovered during the 1967 fieldwork from the 1967 excavations (Figure 3.1). Other newly recovered elements included a large partial innominate bone preserving the lower part of the ilium, the acetabulum and the ischium, a talus and several phalanges. From AHS, the team recovered a nearly complete hominid tibia, lacking the proximal articular surface, and the distal end of a fibula. AHS also yielded extraordinarily dense concentrations of faunal and lithic remains. BNS initially attracted the attention of expedition members because of vertebrate fossils exposed on its surface, but further investigation proved such finds to be sparse. Lithics at BNS seem to reflect the residues of numerous small, possibly independent flintknapping events.

Table 3.1 - Main fossil hominin and archaeological localities in the Kibish Formation, Lower Omo Valley.

3.2 Stratigraphy and Chronology

Early attempts to date the Kibish Formation and associated fossils were unable to do so within any degree of certainty due to the technological limits of radiometric dating in the 1960s. Radiocarbon dating of Nile oyster shells and Th230/U234 of Member I produced dates of greater than 39,900 yr and around 130,000, respectively. The same radiometric techniques yielded similar results for Member III and thus confirmed the antiquity of Member I, but did not come any closer to constraining the age of the Omo Kibish hominins.

Figure 2 - Composite stratigraphy of the Omo Kibish Formation (modified from McDougall, et al. 2005).

In trying to clarify the age of the Kibish fossils, 40Ar/39Ar dates were obtained for Member I, III, and IV of the Kibish Formation. Radiocarbon dates for Member IV confirmed an age of 10,000 years or less for parts of the formation. Pumice crystals from the middle of Member III were dated to 104 kya. Member I was dated to 195 kya just below the level of the PHS site, showing it to be significantly older than the Uranium-Thorium date originally reported in 1969. The lower sections of each member seem to have been laid down rapidly during brief periods of several hundreds or thousands of years, probably as a series of annual flooding events that deposited thin layers of sediment followed by long periods of non-deposition and erosion. Therefore, even though Omo I and Omo II overlie Member I, it is likely that the dates for the fossils are close to those obtained for the pumices.

Dates for the members of the Kibish Formation are remarkably similar to ages of sapropel deposits in the eastern Mediterranean. Sapropels are organic-rich deposits in the Mediterranean that reflect high rates of outflow from the Nile River as a result of intensification of African monsoons and astronomical changes in the orientation of the earth’s axis. Members of the Kibish Formation reflect periods of high outflow of the Omo River, which along with the Nile, begins in the Ethiopian highlands. This suggests that the Kibish Formation members and sapropels are the result of the same climatic cycles.