Abstract : The East African Rift System (EARS) is the best example worldwide of an active magmatic rift. It extends over more than 3500 km from the Afar province to the Mozambican Gulf, cutting through the Ethiopian and East Africa elevated plateaus that are the topographic expression of one or several Cenozoic mantle plumes beneath this part of Africa. Extensional strain initited along the eastern magmatic branch of the EARS at ca. 30 My in the Afar, and then propagated southwards by linkage of discrete magmatic cells centered on the Kenyan dome, to reach South Kenya 8 My ago. There, the North Tanzanian Divergence (NTD) corresponds to a dramatic change in structural style, associated with the abrupt disappearance of magmatism southwards. The transition zone from the NS axial valley into three diverging rift arms (Eyasi, Natron-Manyara and Pangani), dominated by tilted basement fault blocks, is outlined by a 200 km-long transverse volcanic chain, including major Neogene volcanic edifices such as the Ngorongoro, Mount Meru and Kilimanjaro.
Our work presents new structural and geochemical results on magmatic rocks in the NTD, and its main originality is to integrate complementary approaches involving field investigations, petrological-mineralogical analyses, geochronology, Sr-Nd isotopic and trace elements geochemistry. That allows us to also precise the build-up history and mechanisms of a number of major volcanoes of the NTD (Ngorongoro, Mount Meru, Kilimanjaro) during Plio-Quaternary times. Applying this approach to a wider scale, and compiling published radiometric data with our own K-Ar age determinations, allows us to establish a kinematic rift propagation model to the NTD for the last 8 My. It is proposed that both magmatism and strain shifted eastwards along the transverse volcanic chain, up to the Pangani arm, from 8 My to Present, with the onset of magmatic activity in the whole NTD at 2.5 My. But, the main contribution of our study is devoted to the Kilimanjaro long-lived edifice that forms the prominent structure in the EARS, as a whole. Seventeen new K-Ar ages provide constraints for précising time relationships between its three main magmatic centres, and emphasis has been put on the Kibo central vent that recorded the last building phases in the time range 492-165 ky. In addition, new petrological and geochemical results on Kilimanjaro mafic lavas bring some new insights on the nature of mantle sources that are thought to be heterogeneous lithospheric material with residual amphibole and garnet. Two different metasomatic events, Precambrian and Plio-Quaternary in age, are inferred to have governed the evolution of the mantle sources, with the infiltration of melt from the underlying Neogene plume.
These conclusions, extended at the scale of the NTD for young (<1 My) primitive lavas, highlight differences in the nature of the petrogenetic processes that gave rise to synrift magmatism throughout the NTD. To the East, lavas typically show geochemical compositions derived from the fertilization of the stretched lithosphere by the sub-actual activity of the Kenya plume. On the opposite, lavas from the western part of the NTD present geochemical signatures inherited from a Precambrian metasomatized lithospheric mantle, in agreement with results on mantle xenoliths. The nature, location and timing of lithospheric metasomatism in the magmatic segment of the NTD are likely to be highly controlled by structural inheritance, and more particularly by the presence of the Tanzanian Archean craton and large-scale transverse structures in the surrounding Proterozoic orogenic belts.