Genesis of calcrete and related carbonate rocks in the southern Kenya Rift
Pleistocene to Recent calcretes, which are common in the semi-arid parts of the Kenya Rift Valley, provide clues to interpret regional paleoclimate and paleohydrology. A summary of the calcrete occurrences in or near the southern Kenya Rift Valley is provided to give regional context. At Lake Magadi and Olorgesailie, calcretes mantle the land surface, inhibiting erosion of the Pleistocene lake and lake-marginal sediments. The calcretes represent a period of no net sedimentation and form an important land surface throughout the basin. The several calcretes in the area have not been studied in detail to determine their relationship to other sediments in the basin though attempts have been made at calcrete differentiation based on underlying sediments. Several whole-rock samples were taken around the basin predominantly at the south of Lake Magadi to compare the calcretes with each other and their surrounding sediments. At Lake Magadi an extensive calcrete up to 50 cm thick lies at the interface between the Oloronga Beds (780 to 300 ka) and the overlying High Magadi Beds (~23000 to 9000 y BP). The relationship of this calcrete, called the Oloronga-capping calcrete, with the Green Beds (100 to 23 ka) is not obvious in the field but the calcretes are assumed to be older. Based on its chemistry and field relationships, the commonly pisolitic calcrete was formed within the Oloronga Beds though is likely much younger. Chondrite-normalized REE patterns for the calcrete and underlying sediments are consistent with sediment derivation from the Magadi Trachytes, indicating drainage patterns similar to those at present. Based on calcrete microfabrics and stable isotope data the dominant process of formation was through periodic calcite precipitation at or near the groundwater table under dry conditions following a wetter period. Pisoids formed through cycles of circumgranular cracking around reworked calcrete, and around calcite-rimmed lithic or monocrystalline clasts, induced partly due to pressure of calcite crystallization and root growth. Evidence for biotic activity during formation is scarce; rare fungal hyphae and spores were detected under SEM and root-related microfabrics are rare. The data indicate dominantly groundwater calcrete that was overprinted by and possible overprinted upon pedogenic calcrete. The groundwater table during calcrete formation was higher than at present; however, the lake was constricted and low and the calcrete developed marginal to the lake. Laminar calcrete that developed on local trachytic horsts is difficult to date, but is predominantly abiogenic and may have taken many thousands of years to form. Samples from elsewhere in the basin as well as an in-depth study of the Oloronga Beds would help clarify these possible interpretations. Olorgesailie is a fossil rich area and an important site for the study of human origins. Understanding the paleoenvironmental context of the hominin fossils and archeology is important for evolutionary studies. The Olorgesailie Formation (1.2-0.5 Ma) has yielded several archaeological finds and is composed of lacustrine, lake marginal and fluvial sediments, and paleosols. It is overlain by the post-Olorgesailie Formation sediments, which are the focus of this study. At the time of this study the stratigraphy of the post-Olorgesailie Formation sediments had not been presented. Part of this thesis provides some details for the calcretes within the sequence and presents a model for their formation. Three calcretes in a stratigraphic section were sampled as well as the intervening sediments. Using several types of analytical data, a model is presented to explain their origin. The three calcretes are distinct, predominantly pedogenic, and reflect fluctuating hydrological conditions, as implied by silica cementation in the middle calcrete. The upper capping calcrete may reflect a very shallow water table. The last study site targets calcite and silica-cemented conglomerate along the southeast shoreline of Nasikie Engida, a small lake to the north of Lake Magadi. Initially thought to be a calcrete, the deposit at Nasikie Engida, which overlies the High Magadi Beds, is inferred to be a beachrock, which is unusual since Nasikie Engida is not prone to high-energy conditions implied by the conglomerate. This study was undertaken to determine how the deposit formed. Bulk rock samples were taken parallel to shoreline and analysed using geochemistry and microscopy. The conglomerate is likely an alluvial fan deposited reworked locally along the shoreline. The conglomerate lies west of a col at the lowest point between Nasikie Engida and Lake Magadi, suggesting that connection of the lakes and associated hydrological and hydrochemical changes might have been important in forming the deposit. Much of the conglomerate is composed of authigenic components like silica clasts and laminar carbonate clasts. These laminar carbonate clasts are potentially stromatolitic. The silica clasts, which have an internal spherulitic fabric, and silica cement warrant further study.
calcretes, petrography, diagenesis, Kenya Rift Valley, Magadi, Quaternary
Master of Science (M.Sc.)