The origin of volcanic activity occurring far from tectonic-plate boundaries has been a subject of contention. The latest geodynamic model offers a fresh take on the matter. See Letter p.386
In this issue, Liu and Stegman1 present a hypothesis for the generation of volcanic centres that might change our view of how plate tectonics influences the distribution of volcanic activity on Earth.
The theory of plate tectonics describes the uppermost of Earth's layers as made up of rigid plates, the relative motions of which are confined to narrow plate boundaries. The boundaries come in three types: divergent, where plates move away from one another and create systems such as mid-ocean ridges; convergent, where one plate slides beneath another, forming subduction zones; and transform margins, where plates slide past one another, as in the San Andreas Fault system.
Plate tectonics successfully explains most of Earth's geological features. For example, volcanism at mid-ocean ridges can be explained by decompression melting associated with passive upwelling of hot (asthenospheric) mantle in response to plate divergence. Volcanism at subduction zones can be described by a combination of two effects: partial melting of the mantle, driven by return flow in the mantle wedge overlying the subduction zone, and melting-point depression, caused by the influx of water released from the descending plate of the subduction zone.
Volcanoes that occur far from plate boundaries — for example, intraplate magmatism — are more difficult to explain with plate tectonics. Some intraplate volcanic systems, such as the Hawaiian volcanic chain in the Pacific plate and the Yellowstone volcanic field in North America, migrate along tracks that seem independent of plate-boundary processes. The effusive but short-lived outpourings of basalts, known as flood basalts, some of which are so large that they cover substantial areas of continents or even entire plates, are also not easily described by the interaction of slowly moving plates.
One popular view is that intraplate magmatism is driven by narrow mantle upwellings (plumes) originating from a hot thermal layer at the core–mantle boundary2. Therefore, the expression of plumes at Earth's surface should be independent of plate motions2. Flood basalts are thought to record the initial impingement of the anomalously hot plume head, whereas the volcanic track, known as the hot-spot track, records the passage of the plate over the plume's tail3. For example, the eruption of the Steens–Columbia River flood basalt about 17 million years ago is thought to represent the initiation of the currently active Yellowstone hot-spot track, and so is conjectured to fit into the plume theory4,5.
However, the eruption area of the Steens–Columbia River flood basalt is oriented north–south, perpendicular to the Yellowstone track. In addition, the geochemistry of the flood basalt differs from that of the Yellowstone volcanics6,7, complicating the plume hypothesis. Alternatively, the Steens–Columbia River flood basalt could be associated with extension of the upper plate behind the Cascades volcanic arc8 (back-arc spreading). But this phenomenon does not seem to explain the sudden appearance of the Steens–Columbia River flood basalt.
In their study, Liu and Stegman1 (page 386) propose that the Steens–Columbia River flood basalt is a natural consequence of slowing convergence between the North American plate and the ancient Farallon plate. This slow-down was presumably associated with the approach of a mid-ocean ridge between the Farallon and Pacific plates 20 million years ago, now manifested as the active Juan de Fuca ridge. The authors performed geodynamic calculations with initial and boundary conditions constrained by observed relative plate motions and plate age. They show that stretching and eventual tearing of the Farallon plate accompanied the slow-down of convergence, resulting in detachment of the Farallon plate.
Liu and Stegman find that the model that best reproduces the presumed current location of the Farallon plate, as constrained from seismic tomography, predicts tearing to have begun about 16 million to 17 million years ago, when the Steens–Columbia River flood basalt initiated. Dynamic pressures generated from this tear resulted in rapid mantle upwelling through this gap in the slab, driving a magmatic flare-up that mimics the structural trend of the Steens–Columbia River flood basalt (Fig. 1).
If Liu and Stegman's model is correct, the implication is that some intraplate magmatism can be explained by the development of gravitational instabilities within subducting slabs. In their model, thermal upwelling is still responsible for flood basalts, but unlike traditional plumes, which derive from the lowermost mantle, an upper-mantle origin is implied. There are, however, some features that remain unresolved. For example, the model does not provide a good explanation for the eastward migration of the Yellowstone hot-spot track, the high ratio of helium-3 to helium-4 in Yellowstone volcanics7 or the presence of a seismic-velocity anomaly extending into the lower mantle beneath Yellowstone9. And it may not explain the isotopic signatures seen in the Steens–Columbia River flood basalt.
If accurate, Liu and Stegman's model should apply to other locations where slab tears have occurred. Such a tear clearly happened in central California about 20 million years ago, because the last remnants of the Farallon plate were captured on the coast of California, but the rest of the Farallon is no longer present beneath the state10. There is evidence of a flare-up in basaltic magmatism east of central California, for example on the border of California and Nevada, during this time. But the magnitude does not seem comparable to that of the Steens–Columbia River flood basalt, suggesting that different boundary conditions might need to be considered in the authors' model.
In any case, Liu and Stegman's study is pertinent because it draws more attention to subducting slabs in generating intraplate magmas. The following examples might be considered. Where a young oceanic plate is subducting, a slab tear, accompanied by large-volume magmatic flare-ups, should develop because young plates are difficult to subduct. This hypothesis may apply to the eastern Pacific. By contrast, when an old oceanic plate is subducting, a long segment of the slab might be expected to stagnate temporarily in the transition zone between the upper and lower mantle11. The juxtaposition of cold slab material against hot mantle at depth would generate small-scale thermal upwellings along the edges of the slab12. These upwellings could generate widespread basaltic magmatism far from the subduction trench, as seen in northeastern China. If the edges of the slab are migrating relative to the upper plate, hot-spot tracks could be generated13.
We note that all of these upwellings are sourced in the upper mantle and likely to be superimposed on magmatism associated with back-arc spreading; thus a complicated pattern of magmatism is expected. Should a subducting slab penetrate deep into the lower mantle, upwellings might be expected even further from plate boundaries.
In conclusion, there is reason to speculate that intraplate magmas might be intimately linked to subducting slabs12,14. In other words, it is conceivable that plate tectonics generates many intraplate magmas. Differences in the magnitude and locations of intraplate magmas may simply be controlled by the scale of subducting slabs. The debate over whether deep-seated thermal plumes exist15 remains unresolved because these narrow upwellings are difficult to detect. An alternative approach is to map out the geometry and length scale of subducting slabs, which may be easier to detect by various geophysical methods. Liu and Stegman's model shows how downwellings, such as subduction, must be considered when understanding the origin of upwellings and their associated magmatic activities.
Liu, L. & Stegman, D. R. Nature 482, 386–389 (2012).
Morgan, W. J. Nature 230, 42–43 (1971).
Richards, M. A., Jones, D. L., Duncan, R. A. & DePaolo, D. J. Science 254, 263–267 (1991).
Smith, R. B. et al. J. Volcanol. Geotherm. Res. 188, 26–56 (2009).
Hooper, P. R., Camp, V. E., Reidel, S. P. & Ross, M. E. Geol. Soc. Am. Spec. Pap. 430, 635–668 (2007).
Leeman, W. P., Schutt, D. L. & Hughes, S. S. J. Volcanol. Geotherm. Res. 188, 57–67 (2009).
Graham, D. W. et al. J. Volcanol. Geotherm. Res. 188, 128–140 (2009).
Carlson, R. W. & Hart, W. K. J. Geophys. Res. 92, 6191–6206 (1987).
Schmandt, B. & Humphreys, E. Earth Planet. Sci. Lett. 297, 435–445 (2010).
Nicholson, C., Sorlien, C. C., Atwater, T. A., Crowell, J. C. & Luyendyk, B. P. Geology 22, 491–495 (1994).
Zhao, D. Phys. Earth Planet. Inter. 146, 3–34 (2004).
Faccenna, C. et al. Earth Planet. Sci. Lett. 299, 54–68 (2010).
James, D. E., Fouch, M. J., Carlson, R. W. & Roth, J. B. Earth Planet. Sci. Lett. 311, 124–135 (2011).
James, D. E., Fouch, M. J., VanDecar, J. C. & van der Lee, S. Geophys. Res. Lett. 28, 2485–2488 (2001).
Anderson, D. L. Geol. Soc. Am. Spec. Pap. 388, 31–54 (2005).
About this article
Magmatic expression of tectonic transition from oceanic subduction to continental collision: Insights from the Middle Triassic rhyolites of the North Qiangtang Block
Gondwana Research (2020)
Tracing the origin of continental HIMU-like intraplate volcanism using magnesium isotope systematics
Geochimica et Cosmochimica Acta (2016)
Origin of hydrous alkali feldspar-silica intergrowth in spherulites from intra-plate A2-type rhyolites at the Jabal Shama, Saudi Arabia
Journal of African Earth Sciences (2016)
Pliocene–Quaternary basalts from the Harrat Tufail, western Saudi Arabia: Recycling of ancient oceanic slabs and generation of alkaline intra-plate magma
Journal of African Earth Sciences (2015)
Late Cenozoic intraplate volcanism in Changbai volcanic field, on the border of China and North Korea: insights into deep subduction of the Pacific slab and intraplate volcanism
Journal of the Geological Society (2015)