???ViewItemPage???

???ViewItemOverview_lblAbstractSection???

Continental rifting is an important component of the Wilson cycle, and a process-level description requires integration of constraints from seismic tomography, seismic anisotropy, and non-isostatic topography in addition to geological observations. We discuss the evolution of the East African Rift (EAR) and the European Cenozoic Rift System (ECRIS) and define two end members. 1) a-type rifts, such as the EAR and ECRIS, form parallel to mantle flow above elongated, asthenospheric anomalies of low-seismic velocity, “fingers” (LVF) and stay at embryonic stage with slow extension mostly driven by gravitational potential energy. 2) b-type rifts, such as the Menderes and Corinth Rifts, form perpendicular to mantle flow and lead to oceanisation; the Gulf of Aden, Red Sea and Baikal are intermediate. We then propose a new model for the evolution of the short-lived ECRIS (∼44–33 Ma) in the magma-poor period of transition between the Pyrenean orogeny and Mediterranean back-arc extension. The propagation of a LVF toward the north emanating from the Canaries hotspot, all the way to the Massif Central and the upper and lower Rhine region formed the rift on top of a positive anomaly of non-isostatic topography. Fast slab retreat from the end of the Eocene modified asthenospheric mantle flow, initiating the recent Mediterranean subduction regime with back-arc basin opening and dispersal of the asthenospheric anomaly. From ∼8 Ma, slab retreat successively ceased in the central and western Mediterranean, giving way to compression and resumption of volcanism, possibly related to the reestablishment of a mantle LVF. We conclude speculating on the respective roles of a-type and b-type rifts for plate tectonics more general, including for the Mesozoic fragmentation of Pangea.