On the location of hotspots in the framework of mantle convection

Putative mechanisms that have been proposed to explain intraplate “hotspot” volcanism extensively depart from the early plume theory, and many do not involve deep mantle flow. Here, we look for a relationship between hotspot volcanism and mantle flow using flow models excited by density anomalies inferred from seismic tomography. We show that previously identified major hotspots are preferentially located, to a high degree of statistical significance, above regions of positive divergence of horizontal shear tractions beneath the lithosphere. This observation renders it difficult to discard some contribution of mantle flow as a control on hotspot volcanism and instead suggests that mantle plumes are drawn toward, and conveyed by, mantle upwellings (either active or passive), which are revealed by the positive stress divergence. This allows us to exclude a variety of external or shallow mechanisms for the major hotspots. Because we also find that many secondary hotspots do fall at random locations with respect to mantle flow, we emphasize that alternative processes are also required to trigger the less productive volcanism.