91av

SAVE for the occasional earthquake, the ground beneath our feet seems pretty stable. Eurasia and North America aren’t about to dash off anywhere, and we trust that our favourite shore will always face west when we watch the sunset. Convinced? Your faith may be misplaced.

If some scientists are to be believed, there are times when the spinning planet loses its balance, forcing continents and the mantle beneath to slide round the Earth like thick chunks of crust on a lava lake. As a result, the Arctic could start to swelter under a tropical Sun as Earth’s land masses rearrange themselves. And the needle in your compass-which once pointed so reliably north-could swing round to point west or even south.

Though this strange phenomenon, called “true polar wander”, would take millions of years to play out, that’s a geological eyeblink next to the hundreds of millions of years usually needed for continents to drift a long way. If true polar wander really exists, continents could move up to a metre per year, drastically changing global climate and the evolution of life.

In 1997, a team at Caltech in Pasadena proposed that true polar wander had triggered the great “Cambrian explosion” of animal species about 530 million years ago. If geobiologist Joseph Kirschvink and geologist David Evans are right, Earth’s continents slid across one-quarter of the planet’s face within 15 million years. That mass migration would have rearranged ocean currents, altered sea levels by 100 metres or more, and forced ecosystems to adapt to rapidly changing climate conditions. This notion has caused fierce controversy ever since. “It’s admittedly a nutty idea, but it’s wonderfully elegant,” says Kirschvink. “It explains not just one or two things, but a half-dozen peculiar things about one of the most important events in the history of the biosphere.”

And, within the past year, other teams have published evidence for two pulses of rapid true polar wander during the Cretaceous period about 100 million years ago when the dinosaurs reigned. Those motions were less wild than the huge swing suggested for the Cambrian, but if they happened today they would carry London to the latitude of Casablanca in a few million years-10 times as fast as the continents usually move.

That speed-a 100-metre sprint rather than an evening stroll-makes some researchers deeply sceptical. It just seems so unlike the Earth we know to send its continents flying. “Kirschvink’s hypothesis is based on a brilliant idea, but it is not proven,” says geologist Rob Van der Voo of the University of Michigan. Others go further. “These ideas are pretty bizarre,” says Joseph Meert of the University of Florida in Gainesville. “You have to play pretty fast and loose with the data to get very rapid true polar wander. Everything I’ve seen would argue against it.”

True polar wander is known to happen on Earth at least on a small scale. It’s triggered by a slight imbalance in the Earth’s shape. Basketballs, planets and other things that spin are never perfectly spherical. There’s always some lumpiness, such as ridges on the basketball or continents on the planet. The most stable way to rotate such an object is to place the biggest lumps on its equator, farthest from its axis of rotation. That’s why a short, fat spinning top is harder to knock over than a tall, skinny one.

A non-spherical body can be spun around a different axis, but only briefly. Consider an object elongated along one axis-called a “prolate spheroid”. An American football is a good example. Spinning the football about its long axis leads to a pleasing spiral pass through the air, but it’s unstable. A slight throwing error or a nudge from the wind will send the ball fluttering off-course. “That’s why we pay our quarterbacks so much to make a football spin around its least stable axis of rotation,” says Evans, now at the University of Western Australia.

As it turns out, says Evans, certain alignments of Earth’s continents make the planet a prolate spheroid as well. At various times in geological history, the continents have drifted into a continuous mass called a supercontinent, surrounded by a global ocean. The most recent one was Pangaea, which broke up about 200 million years ago. A previous one, Rodinia, probably existed between about 1 billion and 700 million years ago. Supercontinents form giant caps of land that cover a little less than one hemisphere of Earth. The caps themselves alter the planet’s balance a tad. But what really throws things out of whack, say polar wander advocates, is the hot rock of the underlying mantle as it piles up beneath the crust.

It’s hot beneath the supercontinent because there’s no subduction to cool down that part of the mantle. Ordinarily, the mantle is cooled where the planet’s cold sea floor dives under another plate at troughs called subduction zones. However, a supercontinent prevents these cold slabs from piercing the mantle except around its edges. So the mantle beneath the supercontinent heats up and a plume of warm rock rises from deep within the planet, just as boiling water in a pot sinks around the edges and rises in the middle. As time goes on and the supercontinent remains in place, more and more material accumulates beneath it, unable to escape. The planet becomes unbalanced, like a celestial American football (see Diagram).

Then the fun begins. To stabilise Earth’s rotation, the planet’s outer skin and most of the mantle shear off above the liquid outer core, pushing the excess mass to the equator. If the centre of the supercontinent starts near a pole, it can shift across nearly a quarter of the planet’s surface-90 angular degrees-within 10 to 15 million years.

Eventually, the rising heat beneath the supercontinent pierces the land and splits it open, forming new seas as the fragments drift apart. But for a while afterwards, Evans says, the same plume of hot mantle material keeps churning upwards, affecting the balance of the globe. As a result, sporadic bouts of speedy true polar wander may be followed by what Evans calls an “oscillatory” phase, until the planet’s internal mass gets smoothed out. That’s what Kirschvink and Evans believe happened during the Cambrian. “Geophysical legacies of a supercontinent can persist for hundreds of millions of years after it breaks up,” says Evans.

Advocates of true polar wander base their claims on faint traces of magnetism preserved in rocks. Some rocks contain metallic grains that aligned with the Earth’s magnetic field when the rock was still molten, revealing where in the world it formed. However, there are confounding effects. Weathering, heat or chemical reactions can “overprint” the original signature in a rock with that of a more recent magnetic field. “It’s like leaving your favourite audio tape on the radiator and coming back to find your music destroyed,” Van der Voo says. “Heat can erase the record.” Still, palaeomagnetists have developed exquisitely sensitive ways to sort out these complications and estimate how the locations of Earth’s magnetic poles have changed over time.

In their first paper on the Cambrian explosion (Science, vol 277, p 541), Kirschvink, Evans and Robert Ripperdan of the University of Puerto Rico analysed magnetic data from each of the four major continents of that time. The records from these chunks of land, the precursors of today’s continents, appear to show that the poles moved 90 angular degrees across the Earth’s surface between about 535 million and 520 million years ago-the equivalent of swinging from the North Pole to the equator. Since then, Kirschvink says, further analysis of rocks in what are now Australia and Siberia have strengthened the argument. “It’s bulletproof palaeomagnetic data,” he says.

He is especially excited that the movement coincided with dramatic changes on Earth. For instance, models by physicists Jerry Mitrovica and Jon Mound of the University of Toronto predict that sea levels should rise or fall by up to 200 metres as continents move towards or away from a pole. Records from sedimentary rocks show that the ocean did flood or recede in a few key places at the right time, Kirschvink says.

Then there’s the potential link with the extraordinary flowering of new life during the Cambrian. Nearly all basic designs we see in modern animals arose in the Cambrian, over a period of 20 to 30 million years. Kirschvink and Evans think true polar wander was the trigger. “A continent going from the pole to the equator would have vastly disrupted the oceanic circulation patterns,” Kirschvink says. Imagine shutting off the Gulf Stream, he says. Northern Europe would get cold in a hurry. As land masses moved about during the Cambrian, changes like this would come in quick succession. Climate swings would split apart ecosystems, forcing new adaptations to arise at a rate perhaps 25 to 50 times faster than usual. In addition, sea level changes may have isolated marine and coastal populations, allowing them to diversify.

Reaction to the team’s theory was swift. Geophysicists who create models of the planet’s interior found it reasonable-their simulations of the Earth’s lower mantle suggested that it might flow quickly enough to produce the shearing motions of true polar wander. Palaeomagnetists, however, were less sure. A few, such as Richard Gordon of Rice University in Houston, Texas, and Van der Voo, had published earlier papers on episodes of true polar wander, but these shifts were much slower and occurred during other epochs. They note that the speed of the Cambrian motions depends critically on the dates attached to the magnetic data. “I get nervous about how strong the timing constraints are,” says Gordon. He’s open to the idea but not yet sold on it.

The harshest critics of the hypothesis, Meert and his co-worker Trond Torsvik of the Geological Survey of Norway, are less charitable. Torsvik and Meert dispute how some of the Cambrian continents moved, especially Baltica (now Scandinavia and part of Russia) and Siberia. “Kirschvink’s hypothesis is rigorous: it requires every continent to show 90 degrees of motion in 15 million years,” says Meert. According to Torsvik’s data, he notes, the hypothesis fails on this requirement. “Baltica doesn’t even come close, and Siberia falls quite a bit short. There’s nothing convincing.”

Kirschvink finds it difficult to respond politely. For example, he says that one of the three new locations derived by Torsvik for the pole in Baltica is a magnetic overprint from subsequent heating of the rock.

Last year’s studies have fuelled the debate, pointing to quick bursts of true polar wander during the Cretaceous. First, marine geophysicist William Sager of Texas A&M University and geochronologist Anthony Koppers of the Scripps Institution of Oceanography reported that 84 million years ago, the continents tumbled about 20 degrees within two million years (Science, vol 287, p 455). They based their conclusion on undersea volcanic structures called seamounts. These giant lava piles erupt from the ocean floor and often form islands, preserving Earth’s magnetic field direction as they cool. Water gradually erodes them, but scientists aboard ships can still measure their old magnetisations by towing instruments over them. Sager and Koppers found that two groups of seamounts in the North Pacific appear to be exactly the same age, but their magnetic records point in markedly different directions. True polar wander, they say, is the most logical cause.

“Most of the palaeomagnetic community questions this,” says John Tarduno of the University of Rochester. It is notoriously difficult to get accurate magnetic directions and ages from seamounts. Even Sager isn’t sure whether the observations are rock solid. “It’s data you might not use if you could get at the right formations on land,” he admits.

Tarduno also doubts the second study: a report by palaeomagnetist Michel Prévot of Montpellier University in France and his colleagues (Earth and Planetary Science Letters, vol 179, p 517). Prévot’s team also found signs of a 20-degree tilt, but this one occurred 110 million years ago and over a period of less than 10 million years. However, Tarduno’s analysis of a different set of rocks from the same interval, with Rochester colleague Alexei Smirnov, reveals no true polar wander. Each group is sticking to its guns. “The debate is very vigorous because we are unable to be sure which interpretation is the best,” Prévot sighs.

There’s no sign of the debate over the Cambrian explosion cooling any time soon. As Meert puts it: “The older rocks are, the more likely it is that something has happened to them. And yet for some reason it’s easier for people to accept a bizarre world hypothesis as we go further back in time.”

And let’s not neglect the future. A new supercontinent is emerging, which Evans calls “SuperAsia”. Australia, Arabia and Africa are all converging on Asia. The resulting swath of land might be big enough to put a lid on the mantle and throw another curve into the paths of Earth’s poles. But we have several hundred million years to prepare, so for now you can trust your compasses.