Propagation of a melting anomaly along the ultraslow Southwest Indian Ridge between 46°E and 52°20'E: interaction with the Crozet hot-spot?

Regional axial depths, mantle Bouguer anomaly values, geochemical proxies for the extent of partial melting and tomographic models along the Southwest Indian Ridge (SWIR) all concur in indicating the presence of thicker crust and hotter mantle between the Indomed and Gallieni transform faults (TFs; 46 degrees E and 52 degrees 20'E) relative to the neighbouring ridge sections. Accreted seafloor between these TFs over the past similar to 10 Myr is also locally much shallower (> 1000 m) and corresponds to thicker crust (> 1.7 km) than previously accreted seafloor along the same ridge region. Two large outward facing topographic gradients mark the outer edges of two anomalously shallow off-axis domains on the African and Antarctic plates. Their vertical relief (> 2000 m locally) and their geometry, parallel to the present-day axis along a > 210-km-long ridge section, suggest an extremely sudden and large event dated between similar to 8 (magnetic anomaly C4n) and similar to 11 Ma (magnetic anomaly C5n). Asymmetric spreading and small ridge jumps occur at the onset of the formation of the anomalously shallow off-axis domains, leading to a re-organization of the ridge segmentation. We interpret these anomalously shallow off-axis domains as the relicts of a volcanic plateau due to a sudden increase of the magma supply. This event of enhanced magmatism started in the central part of the ridge section and then propagated along axis to the east and probably also to the west. However, it did not cross the Gallieni and Indomed TFs suggesting that large offsets can curtail or even block along-axis melt flow. We propose that this melting anomaly may be ascribed to a regionally higher mantle temperature provided by mantle outpouring from the Crozet hotspot towards the SWIR.