High resolution reconstructions of the Southwest Indian Ridge, 52 Ma to present: implications for the breakup and absolute motion of the Africa plate

SUMMARY We reconstruct the post-52 Ma seafloor spreading history of the Southwest Indian Ridge at 44 distinct times from inversions of ≈20 000 magnetic reversal, fracture zone and transform fault crossings, spanning major regional tectonic events such as the Arabia–Eurasia continental collision, the Arabia Peninsula’s detachment from Africa, the arrival of the Afar mantle plume below eastern Africa and the initiation of rifting in eastern Africa. Best-fitting and noise-reduced rotation sequences for the Nubia–Antarctic, Lwandle–Antarctic and Somalia–Antarctic Plate pairs indicate that spreading rates everywhere along the ridge declined gradually by ≈50 per cent from ≈31 to 19–18 Ma. A concurrent similar-magnitude slowdown in the component of the Africa Plate’s absolute motion parallel to Southwest Indian Ridge spreading suggests that both were caused by a 31–18 Ma change in the forces that drove and resisted Africa’s absolute motion. Possible causes for this change include the effects of the Afar mantle plume on eastern Africa or the Arabia Peninsula’s detachment from the Somalia Plate, which culminated at 20–18 Ma with the onset of seafloor spreading in the Gulf of Aden. At earlier times, an apparently robust but previously unknown ≈6-Myr-long period of rapid kinematic change occurred from 43 to 37 Ma, consisting of a ≈50 per cent spreading rate slowdown from 43 to 40 Ma followed by a full spreading rate recovery and 30–40° clockwise rotation of the plate slip direction from 40 to 37 Ma. Although these kinematic changes coincided with a reconfiguration of the palaeoridge geometry, their underlying cause is unknown. Southwest Indian Ridge abyssal hill azimuths are consistent with the slip directions estimated with our newly derived Somalia–Antarctic and Lwandle–Antarctic angular velocities, adding confidence in their reliability. Lwandle–Antarctica Plate motion has closely tracked Somalia–Antarctic Plate motion since 50 Ma, consistent with slow-to-no motion between the Lwandle and Somalia plates for much of that time. In contrast, Nubia–Somalia rotations estimated from our new Southwest Indian Ridge rotations indicate that 189 ± 34 km of WNW–ESE divergence between Nubia and Somalia has occurred in northern Africa since 40 Ma, including 70–80 km of WNW–ESE divergence since 17–16 Ma, slow to no motion from 26 to 17 Ma, and 109 ± 38 km of WNW–ESE divergence from 40 to ≈26 Ma absent any deformation within eastern Antarctica before 26 Ma.