PALEOCHEMISTRY OF PLIO-PLEISTOCENE LAKE TURKANA, KENYA

The paleochemistry of Plio-Pleistocene Lake Turkana can be estimated by using the chemistry of lakes from the Eastern Rift of Africa as an analogue. Most modern East African lakes occupy closed basins; their chemistries follow an evaporation trend defined by the precipitation of certain mineral phases with increasing alkalinity. Estimates of paleoalkalinity can be used to closely estimate the chemical the chemical composition of ancient lakes. Three methods are used to estimate paleoalkalinity. Diatoms, molluscs, and fish have certain metabolic requirements that are dependent on pH, alkalinity, or calcium levels; thus fauna and flora can be used as paleoalkalinity indicators. Exchangeable cations on clay minerals can also be used because the relative concentrations of sodium and calcium in lake waters are related to alkalinity. Absence or presence of certain minerals also can serve as a paleoalkalinity indicator. Although the latter two techniques give estimates of paleoalkalinity that are averaged over several hundred or thousand years, their estimates agree with the instantaneous estimates based on biologic considerations. This study shows that the earliest lake phase was very fresh and continued until the end of the Kubi Algi Formation. The Lower Member of the Koobi Fora Formation is shown to have been a fresh- to brackish-water lake. From the beginning of Upper Member time (about 1.8 m.y. ago) to the present, the lake occupying the Turkana Depression has varied from a brackish lake that overflowed to a closed basin lake that fell below overflow level and whose alkalinity rose to about 200 meq/l. © 1979.