Evolution of the East African Rift System from trap-scale to plate-scale rifting

Many continental rifts are subjected to volcanism in tandem with rifting, which has raised a long-standing debate about whether magmatism is the cause or the consequence of plate fragmentation. To re-evaluate this chicken-and-egg question, we took advantage of five decades of research on the East African Rift System (EARS), the largest active continental rift on Earth, to explore the spatial and temporal relationship between rifting and magmatism. By comparing the co-occurrence of tectonics and volcanism since the Eocene with the present-day seismicity, we delimit the EARS as a ~ 5000 km-wide zone of volcano-tectonics made of four branches affecting not only East Africa but also the Mozambique channel and Madagascar. We then developed a quality filtering procedure of published radiometric ages in order to build two independent, robust, and comprehensive age compilations for magmatism and rifting over this extended EARS. Our thorough quality-checked selection of ages reveals that the EARS presents two distinct regimes of volcanism. Since the Upper Eocene, the rift system was affected by (1) pulses of volcanism in 500–1000 km-wide areas, and (2) a discontinuous but remarkably simultaneous volcanic activity, scattered along the four branches of the EARS since 25–27 Ma. Combining this spatio-temporal evolution of volcanism with a critical review of the timing of rifting, we show that the tectonics of the EARS evolves through time from trap-scale to plate-scale rifting. Until the Middle Miocene, extension structures first developed following flood basalt events and plateau uplifts. Then, volcanism resumed synchronously all over the EARS at ca. 12–12.5 Ma, followed by a general extensional deformation. This evolution, which cannot be explained by the sole action of a plume or of tectonics, is therefore interpreted in an intermediate way in which the EARS results from (1) extensive stresses acting on the African lithosphere in the long-lived context of the Gondwana breakup and (2) an overall complex mantle upwelling dynamics arising from the African Large Low Shear Velocity Province (LLSVP). We propose that extension stresses affecting the African lithosphere also modulate the melting of mantle anomalies and/or the collection of magma through the Pan-African belts. This influence explains the synchronous occurrence of many magmatic and tectonic events in the EARS and at the boundaries of the Nubia and Somali plates. Finally, our results suggest that the source of extension stresses affecting the African plate probably evolved from a dominant far-field origin to prevailing variations of gravitational potential energy (GPE) and a diverging basal shear of the Nubia and Somali lithosphere. This change would stem from an increase of the mantle flux in the Middle Miocene, yielding a change in the EARS’ dynamics from trap-scale to plate scale rifting.