Tectonic-plate flexure may explain newly found volcanoes

The reduction of pressure in the mantle is a central reason behind Earth’s volcanism. Mantle rock that rises adiabatically to a low enough pressure melts without any added heat source. The ascent may be actively driven by convection, as in mantle plumes, where hot, buoyant jets from the deep interior are thought to provide a rich source of material to fuel the volcanism. Or it may be passive, as at mid-ocean ridges, where magma forms from the upwelling of rock into the gaps left by the spreading of tectonic plates. Deglaciation can also trigger volcanic pulses, as it did in Iceland at the end of the last ice age when the elastic plate rebounded from the melting of the ice cap. Meteorite impacts that excavate large craters at Earth’s surface are yet another cause of pressure release.

A group of researchers from Japan and the US has now discovered a class of volcanoes that apparently formed from pressure release of a different kind—one based on the flexure produced in one lithospheric plate as it is forced beneath another. Naoto Hirano of the Tokyo Institute of Technology and colleagues found clues to the volcanoes a few years ago during deep-sea exploration off the coast of Japan. They analyzed what turned out to be young basalt—erupted lava—hundreds of kilometers from where the Pacific plate dips below Japan.

“Complete serendipity” is how coworker and geochemist Stephanie Ingle of the University of Hawaii describes their find. The researchers were investigating sediment compositions, not volcanism. And each underwater volcano, or seamount, produced less than 1 cubic kilometer of lava, a millionth of the volume that has erupted from the Hawaiian islands. It was only by tracking similarities in sonar data during their recent survey ¹ that Hirano and his crew could locate the small seamounts and their lava beds. During their expedition, the researchers used a manned submersible—the Shinkai (Japanese for “deep sea”)—to sample basalt from the sea floor.

What makes the discovery off Japan intriguing is that volcanism there is unexpected. The seamounts are young, some less than a million years old, and erupt through a cold plate dating from 135 million years ago. Old plate is thick—estimated to reach nearly 100 km in this region of the Pacific Ocean, far from the tectonic boundaries where volcanism typically occurs. High concentrations of a few trace elements in the basalt signaled that only small amounts of melting had occurred at depths greater than 100 km. Rare-gas compositions confirmed that the melting originated in the upper mantle, or asthenosphere.

Current models of Earth hold that minerals found in the asthenosphere at depths around 150 km below the sea floor should remain solid. But the presence of even small concentrations of volatile materials such as carbon dioxide and water lowers the melting temperature of mantle rock. Melting at those depths would then supply the additional magma needed to percolate to the surface through so thick a plate. The morphology of the basalt dredged up by Hirano and company—in particular, the embedded pores and vesicles—bore out the presence of those volatiles. As the melt rose through the lithosphere, dissolved gases in the magma changed phase to form bubbles—much like carbon dioxide bubbles in a newly opened bottle of beer. Further decompression of the gaseous bubbles provided explosive force to the eruption.

To explain the presence of the volcanoes, arranged on the sea floor in rows parallel to the trench off Japan, Hirano and his colleagues propose that a gentle buckling of the lithospheric plate (east of the outer rise) accompanies the severe bend formed as the Pacific plate slides under Japan, as pictured on page 21. The plate’s slight upward bulge then depressurizes the mantle underneath to create magma, which squeezes through cracks and stress fractures at the bottom of the plate and drains upward.

Most geophysicists imagine that the melt percolates along microscopic grain boundaries, at least initially. Although buoyant, the melt still has to overcome the friction generated as it flows around rocky grains. The eventual formation of a macroscopic crack appears essential to prevent the melt from losing heat and freezing on its way to the surface. The youngest volcanoes appear farthest from the trench, where stress conditions make the plate ripe for fracture. Older ones appear closer to the trench.

The Pacific plate cools nonuniformly as it ages, so it’s no surprise that it might contract and crack in places. Indeed, the floor of the Pacific Ocean is littered with thousands of volcanic seamounts, rifts, and faults, in various orientations and alignments. ² And other volcanoes, such as ones in the islands of western Samoa, are thought to form from a flexed plate, though in a different tectonic context, and display chemical signatures similar to those found in lavas sampled off the Japanese coast. Some theorists argue that lithospheric flexing may even explain a periodicity in the spacing of volcanic islands.

Models support the idea that the lower part of the plate stretches, but Brown University’s Donald Forsyth questions whether the stresses created several hundred kilometers from the subduction zone are indeed large enough to fracture the rock. Why an eruption should occur on a plate whose upper regions are compressed is also puzzling. Magma can, however, force its way to the surface by forming and expanding its own crack, especially if the magma is volatile-rich, as it appears to be here. Still, Forsyth and other researchers agree that the newly found seamounts, small as they are, offer a compelling example of volcanism in part of Earth that is not one of the usual volcanic suspects—a mid-ocean ridge, an upwelling plume, or an island arc.

Together, however, the mechanisms of depressurization and the flux of volatiles may not be enough to explain the eruptions. Based on the geochemistry of the lavas, Hirano argues that the seamounts tap a region of the mantle that contains preexisting small but widely distributed pockets of partial melt. Thermodynamically, that’s a reasonable viewpoint, at least in regions where conditions are hot enough for liquid and solid to coexist, says Caltech’s David Stevenson. And it may prompt researchers to revisit their interpretation of the seismic low-velocity zone, commonly observed below ocean basins, where signals travel significantly more slowly than they do in surrounding regions of the mantle.

Last year, researchers explained the low-velocity zone as possibly due to the rheology of olivine—the upper mantle’s most prevalent mineral. ³ But the presence of small amounts of molten material could also contribute to slow seismic velocities. Minerals in the asthenosphere are unlikely to be perfectly mixed, and the region may well experience thermal gradients and melting anomalies. The implications for volcanism are fascinating: A through-going crack or leaky fault would support volcanism just about anywhere on Earth. Judging from the Japanese seamounts, though, the eruptions might be meager.