EARTH SCIENCE Plate motion and mantle plumes A model based on geophysical data from the Indian Ocean suggests that a mantle-plume head may once have coupled the motions of the African and Indian tectonic plates, and determined their respective speeds. S A . R. DIETMAR MÜLLER N ews reports persistently remind us of the effects that plate tectonics has on our daily lives. Earthquakes and tsunamis are consequences of the relentless grinding of tectonic plates past each other. In Earth’s deep mantle, rocks flow like warm toffee, and hot, buoyant material can rise up to the surface, causing volcanic eruptions and ash clouds that can bring air traffic to a halt. Our knowledge of what ultimately drives the plates is still incomplete, however, not least when it comes to understanding the possible influence of mantle plumes. On page 47 of this issue, Cande and Stegman 1 offer a fresh perspective on the possible coupling between plate motions and the forces generated by these plumes. Mantle plumes are hot, cylindrical upwell- ings capped by an enlarged, mushroom-like head. When a rising plume head arrives at Earth’s surface, volcanoes spew out scorching volcanic ashes, gases and voluminous lava. These are quite rare geological events. But when they happen they can cause the break-up of supercontinents, as well as mass extinctions and major ecological changes on land and in the oceans 2 . On the bright side, they are also known to act as a catalyst for emplacing diamonds in the shallow crust 3 . Cande and Stegman 1 extend the catalogue of phenomena attributed to mantle plumes by suggesting that they are able to impart a sub- stantial force on plates — a plume-push force — that can do more than just break up conti- nents. Their hypothesis is that if a rising plume head impinges on the base of a tectonic plate long after supercontinent break-up and disper- sal, its pushing force may result in a substantial transient acceleration or deceleration of plates. Whether a plate adopts the speed of a tortoise or a hare after being hit by a plume depends on how the plume-push force balances out with other forces acting on the plate. Cande and Stegman study the effect of the Réunion plume, whose mushroom head is thought to have reached the surface around 67 million years (Myr) ago, emplacing the huge volumes of volcanic rocks in India known as the Deccan traps 4 . Using a careful analysis of marine geophysical data from the Indian Ocean, they note several phenomena associ- ated with this event: the Indian plate sped up between 68 and 66 Myr ago and subsequently sustained absolute plate-motion speeds of 10–12 centimetres per year for about 15 Myr, with its peak speed reaching about 18 cm yr −1 . Between 52 and 45 Myr ago, India’s motion slowed dramatically to less than 4 cm yr −1 . In addition, Africa’s motion slowed down while India’s sped up, only to return to its previous path and speed after India had decelerated. Figure 1 shows the state of play at 65 and 40 Myr ago, with respective fast and slower motion of the Indian plate. India’s superfast motion has been puzzling. One model ascribed it to loss of India’s conti- nental roots owing to mantle-plume activity when the supercontinent Gondwana originally broke apart 5 . However, such an erosion and melting away of India’s roots, thought to have occurred around 130 Myr ago, cannot be held responsible for the specific acceleration start- ing 67 Myr ago. Moreover, some geodynamic models suggest that the maximum velocity that tectonic plates typically reach is 8 cm yr −1 . The reason for this maximum speed is that plates are largely driven by the pull of slabs subduct- ing in the mantle, especially upper-mantle slabs to which plates are attached 6,7 . The characteristic sinking velocity for upper-mantle slabs is about 7 cm yr −1 , with a maximum of 8 cm yr −1 for the oldest, thickest, least buoyant slabs 6 . This work has been confirmed by sophisticated numerical models for present- day plate motions 8 , which show a near-perfect agreement between observed plate motions and those predicted if plates are driven only by upper-mantle slabs. 10 cm yr –1 0 40 80 120 160 200 240 280 Age of sea floor (Myr) 65 Myr ago 40 Myr ago Figure 1 | The Indian and African plates 65 and 40 million years ago. The black arrows represent absolute plate motions, with the fast motion of India 65 Myr ago, and the slow motion of Africa, corresponding to a time shortly after the Réunion plume head is thought to have arrived at Earth’s surface; the plume head is hypothesized 1 to have affected the plates’ speeds. By 40 Myr ago, the influence of the plume had waned: the Indian plate had slowed down considerably and the African plate had resumed its former direction and speed. Rotated present-day topography is shown on the continents, and sea-floor age is depicted in the ocean basins. The estimated extent of now-destroyed continental margins is represented by the grey areas; mid-ocean ridges and transform faults are shown as black lines; subduction zones are shown as a red line, with teeth on the overriding plate. (Plate reconstructions generated using the GPlates software, data and rotations 11 .) 40 | NATURE | VOL 475 | 7 JULY 2011 NEWS & VIEWS © 2011 Macmillan Publishers Limited. All rights reserved