How many mantle plumes in Africa? The geochemical point of view

The association of anomalous topographic swells and widespread Cenozoic volcanism within the African plate (Hoggar, Tibesti, Darfur, Ethiopian highlands, Kenyan dome) may reflect either the involvement of one, or several, deep mantle plumes, or, alternatively, be attributed to tectonic processes involving only the lithosphere or the shallow asthenosphere. We present here new helium isotopic measurements that, added to existing data, allow us to restrict the spatial extent of a high-3He component (up to 20 Ra) to the Ethiopia–Afar volcanic province, for places where large volumes of Oligocene pre-rift flood basalts and ignimbrites erupted within a short (1–2 Ma) time interval. All other investigated African volcanic provinces display MORB-type, and/or continental lithosphere-like, 3He / 4He signatures (7 ± 2 Ra) often modified by a contribution of crustal He. The distribution of He isotopic signatures in Africa, together with other isotopic (Sr–Nd–Pb) tracers measured in Miocene to Plio–Quaternary alkaline lavas in East-Africa, is fully consistent with the occurrence of two types of mantle plumes: (i) a large, deep-sited mantle plume characterized by a high-3He signature, possibly originating from the core–mantle boundary according to seismic mantle tomography, which triggered the flood basalt eruptions 30 Ma ago and which subsequently interacted with shallower mantle sources to produce the syn-rift volcanism of the Ethiopia–Afar province; and (ii) a second-order type of shallow mantle upwelling, presumably originating from depths shallower than 400 km as suggested by seismic wave imaging, distinct from the main Afar plume and disseminated within the African plate under the uplifted and rifted swells. The above conclusions do not support the view of a unique large mantle plume feeding all Cenozoic African volcanic provinces. The fact that high-3He signals are associated with the largest lava volumes erupted in Africa since the beginning of the Cenozoic argue against models advocating a shallow origin for high 3He / 4He signatures. Instead, they confirm that such signatures characterize hot material coming from the deep mantle.