The nature and chronostratigraphy of Quaternary pyroclastic accumulations from lake Barombi Mbo (West-Cameroon)

The K/Ar dating of a limburgitic flow indicates that the Barombi Mbo maar was active a million years ago. The morphology of the maar and the unusual high K 2 O/Na 2 O ratio of the limburgite suggest either that this age is too old or that this flow is not the last activity of the maar. Twenty three meters of sediments were cored from the bottom (core BM6) of the 110-m-deep lake which currently fills the maar. They are composed of clayey and sideritic laminae with interstratified ash layers in the lower half of the core. These ashes, composed of slightly vesiculated sideromelane with olivine and labradorite phenocrysts, have been dated between 21,000 and 11,000 yr. B.P. They show no alteration and correspond to aerial fallout of alkaline (sodic) basaltic affinity, which is common in the area, but are distinct from the mafic magmas which generated the maar. The origin of the tilting and uplifting of a 1.5-m sedimentary column in the lower part of the core remains controversial. Holocene-age pyroclastites are rare and consist of a mixture of granulitic xenocrysts and typically transitional tachylite similar to volcanics from the northeast (Manengouba). Magnetite microspherules occurring in Holocene layers have no volcanic equivalent. They could be related to biogenic or pedogenetic processes induced by more humid climatic conditions.