In last week's issue of Science, Paul Silver (of DC's own Department of Terrestrial Magnetism at the Carnegie Institution) and Mark Behn (formerly a post-doc at Carnegie, and now at Woods Hole in Massachusetts) published a paper putting forward an intriguing idea: maybe plate tectonics proceeds in fits and spurts.

Silver and Behn note that most of the world's subduction zones are located in the circum-Pacific belt, and that the Pacific is getting smaller over time. (Subduction destroys oceanic crust, and since the Earth presumably isn't increasing or decreasing in volume, subduction in the Pacific is balanced by seafloor spreading elsewhere, like the Atlantic.)

The Pacific is "predicted" to close in about 350 million years, assuming that the tectonic plates continue to move at about the same rate they're moving now. The death of the Pacific would come as the Americas smash into eastern Asia and Australia, raising up a Himalayan/Appalachian style mountain belt. Silver and Behn posit that this would basically end subduction on planet Earth for a time. This was a startling idea to me at first, but then I thought, "Why not?" Then I thought, "I wish I'd thought of that."

My understanding of mountain belts comes from the Appalachians, which built up in three successive episodes called orogenies. Check out the diagram below (from the excellent textbook Essentials of Geology by Steve Marshak, that I use in my Physical Geology course at NOVA) and follow along so you can see why this new concept startles me a bit (but in a good way):

There used to be a big ocean basin off the "east" coast of North America that closed via subduction over the course of the Paleozoic Era. This extinct ocean goes by the name of Iapetus. This was not a simple event: it was more like a pile-up on the highway than a simple head-on collision. This ancient ocean basin was not just empty ocean. It also included islands and small chunks of continental crust ("microcontinents" like modern-day Madagascar). First a subduction zone developed out there in the ocean, closing a portion of it. This brought a chain of volcanic islands closer & closer to North America. The islands hit North America (around 460 million years ago), in a mountain-building event called the Taconian ("Taconic") Orogeny. Once that had happened, a new subduction zone developed on the ocean side ("outboard") of the islands/accreted terranes. That began to close another part of the Iapetus Ocean. Around 360 million years ago, that episode of subduction ended when a microcontinent (dubbed Avalonia) smacked into North America. This collision caused more mountains to rise: the Acadian Orogeny. Then yet another subduction zone, outboard of the newly accreted Acadian terrane, kept the closure of the Iapetus Ocean going, until finally the continent on the other side of the ocean (Africa) smashed into North America, raising more mountains. This is the Alleghenian Orogeny (sometimes spelled "Alleghany"), which really crumpled up the landscape, starting around 300 million years ago. The moment the Iapetus died was the moment Pangea was born.

I go into all this because the model of plate tectonic convergence the Appalachians display is one that says collisions between plates don't stop the overall convergent forces. As soon as one subduction zone is snuffed out, a new one develops outboard of the continent, where the weaker, denser oceanic crust gets shoved downward.

But does it actually work that way all of the time? Silver and Behn suggest maybe not. Maybe it's an "on-again, off-again" affair. They cite among their evidence an earlier orogeny, the Grenville Orogeny, which sutured together many continents at a much earlier time (about a billion years ago). When that collision had ended, the supercontinent Rodinia was born. Silver and Behn note a lack of volcanic activity around the world for hundreds of millions of years after the Grenville Orogeny (most volcanoes are caused by subduction). Rodinia did eventually break up amid much volcanic activity (including the eruption of the mid-Atlantic's infamous Catoctin Formation), and giving birth to the Iapetus Ocean basin in the process -- but that didn't happen for a long time after Rodinia got assembled. What gives? Does that mean subduction was inactive during that period?

They also offer a modern example: India and the Himalayas. 20 million years ago, India was a microcontinent out in the Indian Ocean, with a pavement of oceanic crust separating it from Eurasia. India moved north, the oceanic crust got subducted, and eventually India plowed into Eurasia, raising the Himalayas. But why hasn't a new subduction zone developed south of India? That would be what would happen if India's orogeny were following the Appalachian example.

Maybe plate tectonics has periods of intense activity (lots of subduction), but then has periods where it's "clogged up," and the movement of the plates slows. Eventually heat builds up in the underlying mantle (the source of plate movement) to the point where the mantle begins to convect more vigorously, and the plates start getting dragged around again. It's kind of a cool notion. I'd be interested to hear what you think about it. Please post any thoughts you have in the comments section below.

The whole idea reminds me of the concept of punctuated equilibrium, a model of biological evolution which bucked the long-standing notion (originated by Darwin himself) that evolution proceeded slowly and methodically over time. Thanks in part to an eye-opening appreciation of the Earth's immense age, the prevailing wisdom was that evolution was gradual, smooth.

Then (in 1972) Niles Eldridge and Steve Gould published a landmark paper that suggested otherwise. Instead of "gradualism," they argued, changes in populations of living organisms may have happened suddenly, experiencing a lot of change in a short period of time. Once equilibrium was achieved, the new status quo was preserved as a non-dynamic scene for a long time. (See image at left, which came from Wikipedia).

They cited the fossil record as their primary evidence: most of the change seen in fossils is a sudden switch of biological "regimes," with new fossils showing up, lasting a while, and then abruptly vanishing. I'm oversimplifying here, but I hope the analogy is clear: if evolution can do it, why not plate tectonics? Is there any reason to think plate tectonic motion couldn't happen in spurts of more activity followed by periods of quiescence? Ponder it...

Reference: Silver, Paul G., and Behn, Mark D., 2008, Intermittent plate tectonics?: Science, v. 319, p. 85-88, doi: 10.1126/science.1148397.

For those without a subscription to Science, you can read the press release about Silver and Behn's work that Carnegie put out by visiting their website.

Labels: appalachians, evolution, iapetus, plate tectonics