Ridge suction drives plume-ridge interactions

Abstract

Deep-sourced mantle plumes, if existing, are genetically independent of plate tectonics. When the ascending plumes approach lithospheric plates, interactions between the two occur. Such interactions are most prominent near ocean ridges where the lithosphere is thin and the effect of plumes is best revealed. While ocean ridges are mostly passive features in terms of plate tectonics, they play an active role in the context of plume-ridge interactions. This active role is a ridge suction force that drives asthenospheric mantle flow towards ridges because of material needs to form the ocean crust at ridges and lithospheric mantle in the vicinity of ridges. This ridge suction force increases with increasing plate separation rate because of increased material demand per unit time. As the seismic low-velocity zone atop the asthenosphere has the lowest viscosity that increases rapidly with depth, the ridge-ward asthenospheric flow is largely horizontal beneath the lithosphere. Recognizing that plume materials have two components with easily-melted dikes/veins enriched in volatiles and incompatible elements dispersed in the more refractory and depleted peridotitic matrix, geochemistry of some seafloor volcanics well illustrates that plume-ridge interactions are consequences of ridge-suction-driven flow of plume materials, which melt by decompression because of lithospheric thinning towards ridges. There are excellent examples: (1) The decreasing La/Sm and increasing MgO and CaO/Al2O3 in Easter Seamount lavas from Salas-y-Gomez Islands to the Easter Microplate East rift zone result from progressive decompression melting of ridge-ward flowing plume materials. (2) The similar geochemical observations in lavas along the Foundation hotline towards the Pacific-Antarctic Ridge result from the same process. (3) The increasing ridge suction force with increasing spreading rate explains why the Iceland plume has asymmetric effects on its neighboring ridges: both topographic and geochemical anomalies extend < 400 km along the slower (20 to 13 mm/yr northward) spreading South Kolbeinsey Ridge, but > 1500 km along the faster (20 to 25 mm/yr southward) spreading Reykjanes Ridge. (4) The spreading-rate dependent ridge suction force also explains the first-order differences between the fast-spreading East Pacific Rise (EPR) and the slow-spreading Mid-Atlantic Ridge (MAR). Identified mantle plumes/hotspots are abundant near the MAR (e.g., Iceland, Azores, Ascension, Tristan, Gough, Shona and Bouvet), but rare along the entire EPR (notably, the Easter hotspot at ~ 27°S on the Nazca plate). Such apparent unequal hotspot distribution would allow a prediction of more enriched MORB at the MAR than at the EPR. However, the mean compositions between MAR-MORB and EPR-MORB are the same in terms of incompatible element abundances, and are identical in terms of Sr-Nd-Pb isotopic ratios. This suggests similar extents of mantle plume contributions to EPR and MAR MORB. We consider that the apparent rarity of near-EPR plumes/hotspots results from fast spreading. The fast spreading creates large ridge suction forces that do not allow the development of surface expressions of mantle plumes as such, but draw plume materials to a broad zone of sub-ridge upwelling, giving rise to random distribution of abundant enriched MORB and elevated and smooth axial topography along the EPR (vs. MAR). One of the important implications is that the asthenospheric flow is necessarily decoupled from its overlaying oceanic lithospheric plate. This decoupling increases with increasing spreading rate.