A long reach. Nevada's Basin and Range terrain, here in a topographic rendition, may be the result of long-ago events off the West Coast.

Dave Stegman/UCSD/SIO and Mike Sandiford/The University of Melbourne

A New Engine for Plate Tectonics

Beneath our feet float enormous slabs of rock. They crash into each other to form mountain ranges, dig deep trenches in the ocean, and tear huge landmasses apart. Yet scientists still aren't clear why these "tectonic plates" move the way they do. A new study may have an answer. A group of geodynamicists drawing on both present-day plate motions and model simulations argues that tectonic plates move fast or slow depending on how they sink into deep-sea trenches, dragging the rest of the plate along in the process.

"It's a nice, really simple concept," says geophysicist Donald Forsyth of Brown University, and it may even be right.

There has been no shortage of theories to explain the pace of tectonic plate motions. Some geophysicists have suggested that Earth's churning mantle drags overlying plates along, like scum on the surface of simmering water. The faster the churning, the faster the plate motion. Others have argued that because older parts of plates have cooled and thus are denser, the speed at which they can sink into the mantle and drag the rest of the plate along with them determines how fast the plate moves.

Geodynamicist Wouter Schellart of Monash University in Melbourne, Australia, and colleagues tested these ideas and found them wanting. Instead, they found that a plate's speed depends on the width of its leading edge that is sinking into a deep-sea trench. Wide-edged plates slip quickly into the mantle at the trench, whereas narrow ones tend to be slow. The group observed the same relation in models simulating plates sinking into the mantle.

In the model simulations, the process works like this: Ocean tectonic plates can slide at an angle into Earth's mantle like a tilted conveyor belt, but they are also sinking vertically. The mantle resists the slab's vertical sinking the way water resists a paddle being drawn through it, strongly across its width with resistance-reducing eddies at its edges. So if a sinking slab is wide, the edge effect is relatively minor, and the slab will slide into the mantle, pulling the plate along with it. A narrow slab, on the other hand, will encounter less resistance and tend to sink rather than pull its plate down the conveyor.

Schellart and his colleagues, who report their findings in the 16 July issue of Science, see changing plate width at work in North America's past. Fifty million years ago, the great Farallon ocean plate, which extended from Alaska to South America, was fast sinking beneath the edge of western North America, sliding hard against the base of the continent, and thereby scrunching up the southern Rocky Mountains. But 30 million years ago, the Farallon plate split in two as its narrower midsection disappeared beneath North America. Being narrower, the remaining plates slowed and no longer pushed up the Rockies, the group infers. And as in the group's models, the mostly sinking slabs began dragging the deep-sea trench to the west. That shift, the researchers say, would have stretched the continent instead of squeezing it, creating Nevada's hummocky Basin and Range terrain.

Plate width as the critical factor in plate motion "is a pretty good concept," Forsyth says. The claim that the changing width of a sinking plate reshaped the western United States is also "plausible, [but] the history is very complicated. It will get people thinking." Seismologist Gene Humphreys of the University of Oregon, Eugene, agrees about the history part. Seismic imaging beneath the western United States shows a mess of slab fragments that don't yet tell much of a story about what happened 30 million years ago. The geodynamicists have their work cut out for them.