Travels of America

Half a billion years ago, a large chunk of North America went missing. That chunk has now turned up in the Andes of Argentina.

By Tim Appenzeller|Sunday, September 01, 1996
For Bill Thomas, the hills and hollows of the up-country South are home ground. I grew up speaking genuine Appalachian, says the tall, quiet geologist from the University of Kentucky. For nearly 30 years Thomas has been tramping around the Appalachians, trying to understand why they appear to stop in Alabama--only to resume hundreds of miles to the west, in Arkansas, as the Ouachita Mountains. That gap makes room for the broad Mississippi Valley and ultimately for the Gulf of Mexico, and Thomas has been asking how it formed. He never expected the answer would take him 4,000 miles south to the arid foothills of the Andes.

Then again, Ricardo Astini never thought he’d have to leave those foothills, called the Precordillera, for the thickets and piney woods of northern Alabama. Astini, a geologist at the University of Córdoba in Argentina, speaks genuine Spanish, a fast and animated version. He has spent a decade trying to understand how that tract of limestone hills--so weirdly distinct from its surroundings in fossils and rock types--ended up snuggled against the Andes in western Argentina. But by last year Astini was picking over rocks in Alabama, and Thomas was planning a field trip to South America.

What brought these two unlikely collaborators together wasn’t an academic exchange program or yearnings for a drastic change of scene; it was the realization that they were working on the same problem. More than half a billion years ago, Thomas and Astini have now shown, a block of crust 500 miles square broke away from North America, drifted across an ocean, and welded itself to South America. The result was a gap in the North American coast that eventually became the Gulf of Mexico, and a large tract of foreign crust in South America that became the Precordillera.

The result, also, is support for a new map of the world of 500 million years ago. Geologists have tended to picture the continents as tango dancers, sometimes glued together, sometimes stepping apart, but always paired off with the same partner--North America with Africa, for example. But just about the time Thomas and Astini were tracing the history of the Precordillera and the Gulf of Mexico, another pair of geologists were conceiving a new view of ancient geography. In it, continents that would later be strangers faced each other across oceans that no longer exist. And in it, a land swap between the east side of North America and the west side of South America--once barely conceivable--was no more than routine.

Thomas, to start, was interested in only one sliver of that earlier world: the ancient East Coast. The southern Appalachians form a series of gentle curves, each one set a little farther to the west than its neighbor to the north. Those offsets, Thomas realized in the late 1970s, are a blurry image of sharp zigzags in the old edge of the continent--of the jagged rift along which Laurentia, as ancient North America is called, tore away from some other continent. Later an ocean called the Iapetus rolled through the gap; and later still, other blocks of crust slammed into Laurentia, shoving the ancient coast inland and raising the modern Appalachians.

All along the eastern flank of the mountains, from Virginia to Alabama, Thomas could see the stratigraphic signature of the rift that became the Iapetus. There were conglomerate layers that had formed when sand and gravel washed into the developing rift; there were limestones-- consisting of fossilized sea creatures--that had been deposited after the sand and gravel, once the rift had widened into ocean. From the age of the limestones Thomas could tell that the process was already well under way by 540 million years ago. At every zigzag in the ancient coastline, the story was the same--except at the biggest one.

That one is at the southern tip of the Appalachians in Alabama, where a fault called the Alabama-Oklahoma transform joins them with the Ouachitas, 360 miles northwest near the Arkansas-Oklahoma line. Rocks retrieved from boreholes along the fault, mostly by oil prospectors, made it clear to Thomas that the Ouachitas are a continuation of the Appalachians, and that before they were mountains they too lay on the Iapetus coast. But there was one problem: there was no evidence that a rift had formed in the Ouachitas at the same time--no 540-million-year-old limestones, for instance.

Apparently, Thomas realized, the great continental tear that created the edge of Laurentia didn’t zigzag west through Oklahoma at first; it continued straight south from Alabama. Only later did it jump inland along the Alabama-Oklahoma transform, which may have been a line of weakness formed during the original continental breakup. Thomas thinks that jump occurred around 545 million years ago. That’s the age other geologists have assigned to extensive layers of basalt and other volcanic rocks--now mostly buried under younger sediments--at the western end of the Alabama- Oklahoma fault. And that’s when, according to Thomas, the land in Oklahoma started rifting, allowing magma to well up through the fractured crust.

The land to the east of this Ouachita rift was not a continent; that continent, North America’s eastern partner, had already departed. What it had left behind was a 500-mile-square chunk of crust, stretching from southern Alabama west to Oklahoma and south beyond Houston, and bounded on the south and east by the Iapetus. As the Ouachita rift widened, this square of land broke away and began sliding east along its northern boundary, the Alabama-Oklahoma transform. Limestones buried in two ancient valleys just north of the fault, Thomas found, record how the ocean flooded into the valleys as the crustal block slid out of the way. By 515 million years ago, it had sailed clear of North America and was drifting free in the Iapetus.

The departed block left a gap that Thomas calls the Ouachita embayment--and that the rest of us know as the Gulf of Mexico. Although other landmasses later filled the gap--in the collisions that raised the Ouachitas--later still the added crust broke away along the original faults, creating the Gulf. Explaining the origin of that absence of land was Thomas’s only goal at first. This block was something I just needed to dispense with, he recalls. I was asked several times, ‘Where did it go?’ My attitude, somewhat cynically, was ‘I don’t care.’

But now we think we know where it went.

Finding a bit of one continent in another isn’t so startling these days. Since the late 1960s, geologists have known that Earth’s surface is in constant motion, as new ocean floor congeals from magma at volcanic ridges, then trundles away from them at an inch or two a year. As it moves, the ocean floor can sweep along continents, as well as smaller landmasses and islands. The smaller bits eventually get plastered onto the leading edges of continents, giving them a fringe of exotic terranes. California, for instance, is nothing but. Finding yet another terrane along the Andes would cause no particular stir.

But finding one that started off as a piece of North America is another matter. To have said five years ago that the Precordillera was a part of North America was something most people just weren’t willing to contemplate, says Eldridge Moores of the University of California at Davis. Few were willing to consider a map on which North and South America were close enough half a billion years ago to trade a piece of crust. Few people thought seriously at all about geography that ancient.

Until about five years ago, geologists rarely pushed their maps of Earth beyond Pangaea, the supercontinent that broke up some 200 million years ago. Because the oceans that opened then are the oceans of today, it’s easy to trace how the continents were arranged in Pangaea, just by working backward from their tracks in the ocean floor. Those tracks say that the East Coast of North America once nuzzled up against North Africa, while South America’s eastern shoulder fitted into the hollow of West Africa. Before Pangaea, though, we don’t have any ocean floor to guide us, says Ian Dalziel (pronounced dee-ell) of the University of Texas at Austin. So Pangaea is the oldest paleogeography that any of us will totally agree with.

When geologists have tried to look back before Pangaea, they have tended to re-create it again and again--to put eastern North America somewhere opposite North Africa, separated by an ocean that closed when Pangaea formed. In that picture, the Atlantic is only the latest in a series of oceans that have come and gone between North America and Africa. Dalziel calls this scheme yo-yo tectonics.

And a few years ago he and Moores put the yo-yo on the shelf. It started with Moores, who studies the Great Basin, the Sierra Nevada, and other parts of the American West. Like every other geologist who works in the region, he had seen signs that some matching landmass farther west had rifted away 650 or 700 million years ago. You always worried, ‘Where was that piece?’ says Moores. People were always trying to find pieces in the Northern Hemisphere that might fit. None of them quite did.

Then in 1989, Moores ventured into the Southern Hemisphere--to Antarctica, on a field trip organized by Dalziel. I knew almost nothing about Antarctica at the time, he says. But there was a lot of talk on the ship about Antarctica, and they gave out these National Geographic maps. I like looking at maps, and I spent a lot of time looking at this one.

Illuminating as that map and the field trip itself were, though, there was no flash of insight on the ice sheet. That came several months later, back home in the Davis library. There Moores happened on a paper by two Canadian geologists proposing that 700 million years ago, Canada and Australia were joined together. The Canadians had stopped their analysis at the U.S. border, but Moores didn’t. At the same time Canada was supposedly linked to Australia, he knew, Australia was linked to Antarctica. And Canada, as always, was closely tied to the United States. As Moores recalls it: I thought, ‘Aha.’

Moores guessed he had found North America’s lost western partner. The part the Canadians cared about had once been connected to Australia, all right, but Moores’s own professional territory had been connected to East Antarctica. Some quick research in the library convinced him that rocks reminiscent of a 1.8-billion-year-old belt in the American Southwest do indeed peep through the ice in the Transantarctic Mountains, the original edge of Antarctica. Moores called his new hypothesis sweat (Southwest U.S.-East Antarctica), and one of the first people he tried it out on was Dalziel.

Eldridge made this leap of faith and put pen to paper and faxed me this map and asked me, ‘Is this crazy?’ recalls Dalziel. Dalziel thought not; he had been musing along the same lines. And now he started thinking about North America’s long-lost eastern partner--the landmass that had rifted away 540 million years ago to create Bill Thomas’s zigzag coast.

Geologists had generally assumed that partner was Africa, which later slammed back into North America to raise the Appalachians--the yo-yo model. But in Peru, of all places, Dalziel had found rocks that contradicted that assumption. They seemed to have been deformed in a mountain-building episode more than a billion years ago--at the same time as rocks in Labrador. If Labrador and Peru were thousands of miles apart at the time, as the standard scenario required, that would have been a remarkable coincidence. But it made perfect sense, Dalziel realized, if the northeastern corner of North America had once nestled into the sharp bend in Peru’s coast--if South America rather than Africa had been North America’s eastern partner before Pangaea. Using software that allowed him to play continental matchmaker on a computer, Dalziel compared the outlines of the potential partners and also checked the relict magnetism in their rocks, which is a rough indicator of their latitude at the time the rocks formed. I asked myself, ‘Was it paleomagnetically reasonable for Laurentia to be down next to South America?’ Yes, indeed, it was.

Thanks to Moores and Dalziel, then, there was a whole new map of Earth before Pangaea. Until 750 million years ago, in this view, North America was near the South Pole, wedged between Antarctica and Australia on one side and South America on the other, in a supercontinent called Rodinia. After that, according to Dalziel’s simulations, North America went through Houdini-like contortions to escape from its partners, culminating in a 250-million-year-long end run up the west side of South America. Clearing the northern end of that continent, it faced off for the first time against North Africa, which was to become its neighbor in Pangaea.

The end run worked on a computer screen, but was there evidence for it in the real world? Dalziel proposed that geologists look for some. He said they might find North America’s calling cards--pieces of crust it had deposited on other continents during its wanderings. As it happened, Argentine scientists had already found one years before.

Even now the precordillera is a realm apart. For nearly 500 miles, north to south, it rises from the vineyards of western Argentina, in pale cliffs of limestone and shale, to peaks that pierce 14,000 feet. Here and there the crumpled terrain is cut by a braided river carrying snowmelt from the Andes, which nourishes a fringe of grass and willows. Acacia bushes and cactus claim the rest of the level ground, however, and there’s little enough of that: the Precordillera is a desert tipped on its side. Most of it is a geologist’s dream, with exposed layer cakes of rock that rise a thousand feet and more and reveal hundreds of millions of years of history.

More than half a billion years ago, according to those rocks, the land here was not a fractured, buckled desert. Long before the geologic violence that uplifted the Andes, the Precordillera lay flat, forming shallows and tidal flats in a warm inland sea. And the fossils that now pepper its limestone cliffs--crustacean-like trilobites and the winglike shells of brachiopods--indicate that sea was nowhere near South America. In such features as the shape of a trilobite’s head or the flare of a brachiopod’s shell, the fossils differ subtly but unmistakably from typical South American ones. As early as 1965, according to geologist Victor Ramos of the University of Buenos Aires, Argentine paleontologists were saying, Jesus Christ, those trilobites are North American.

The fossil experts had little idea how the intruders might have gotten there. But Ramos came up with one. In 1981, at a symposium in the United States, he heard speakers explain how to recognize exotic terranes from the traces of seafloor along their boundaries. I opened my eyes, because the kind of evidence they were talking about was the kind we had, Ramos recalls. The thing that had amazed me was the pillow lavas on both sides of the Precordillera. Pillow lavas form at seafloor volcanoes when magma is rapidly quenched in cold water. Their presence on both sides of the Precordillera suggested it had once been surrounded by ocean. To Ramos, it had to be a terrane.

The North American fossils indicated where that terrane had come from. So did its thick limestone layers, which resembled ones in the northern Appalachians--or so Ramos thought. In 1984 he proposed that the Precordillera might have broken free of North America’s East Coast half a billion years ago and collided with South America. Later it became landlocked when another chunk of crust swept in from an unknown source to the west.

At a time when Moores and Dalziel had not yet redrawn the pre- Pangaea map, Ramos’s scenario was mind-boggling: it required the Precordillera to have sidestepped thousands of miles across the globe. In Argentina, Ramos’s proposal caused a big debate, recalls Ricardo Astini. Lots of geologists were mad at him because it was a revolutionary idea. In North America the response was mainly silence.

But Astini and his paleontologist colleagues at the University of Córdoba, Luis Benedetto and Emilio Vaccari, were actually working in the Precordillera. They had seen its strangeness for themselves and understood that it seemed to call for strange explanations. In the mid-1980s, they set out to test Ramos’s idea. They soon found rocks and fossils that traced the details of the Precordillera’s surprising voyage.

The fossils showed the voyage was swift. Until the early Ordovician Period, 480 million years ago, they are quintessentially Laurentian. Then their aspect starts to change. The trilobites and brachiopods that filled the shallow seas of the Precordillera for the next 15 million years or so are as distinctive, to paleontologists’ eyes, as the plants and animals of certain islands are today. The Precordillera was apparently a kind of Madagascar--a large island in a now-vanished ocean (albeit an island that was mostly water-covered). By 465 million years ago, though, the island had docked in South America: the fossils from then on look identical to those found on the continent.

From the rocks, Astini and his colleagues extracted a blow-by- blow account of the collision. On the east side of the Precordillera, its leading edge dipped under the South American coast, forming a submarine trough that collected thick beds of sediment that can still be seen today, now crumpled and uplifted. On the west side, earthquakes produced by the collision shook loose bus-size blocks of limestone, sending them hurtling into the sea; today those blocks can still be seen, trapped in other sedimentary rocks. In the central Precordillera, gaps in the rock sequence show how the collision thrust the limestone well above sea level, where it could be worn down by weather. Finally, not long after the Precordillera snuggled up against South America, gravel deposits sifted down onto its eastern edge from mountain glaciers on the continent; South America was in the grip of an ice age 440 million years ago. North America was still tropical. But the Precordillera’s career as a tropical reef was over.

Working in its parched landscape, Astini and his colleagues could piece together its whole journey--except for its starting point. Ramos had suggested the northern Appalachians, but the rocks and fossils there didn’t seem a perfect match. Astini’s group started looking at rocks from elsewhere in the Appalachians. Then in 1992 they got a lucky break. The North-South connection was made for them by Christopher Schmidt, a geologist from Western Michigan University.

Schmidt came to Córdoba to work on another topic, Astini recalls, and after a while we met each other. I talked to him about our project and the probable foreign origin of the Precordillera.

Bill Thomas takes up the thread: Chris and I had been working together, and he had become pretty familiar with my story about the origin of the Gulf of Mexico. Ricardo was describing the Precordillera to Chris, and Chris said, ‘Hey, I know where it came from.’

Chris encouraged me to be in contact with Thomas, adds Astini. I read his 1991 paper on a terrane that went out of Laurentia that has no name. It should be somewhere in the world, I thought.

So Astini began comparing strata from the southern Appalachians with those of the Precordillera; he measured the size of the missing piece of North America against the Precordillera; he considered the mirror-image histories of departure and arrival. He knew right away he was seeing the beginning of a beautiful intercontinental friendship.

When you look at some of the sections in Alabama where the southern Appalachians outcrop, or you look at samples from boreholes, they really match one-to-one what we have in the northern Precordillera, Astini says. It’s incredible--the colors, the thicknesses of the rock all match. Where you have green shales, we have green shales. Where you have red shales, we have red shales. It’s really incredible to travel so far and have the same strata. It becomes credible, though, if you accept that the Appalachians and the Precordillera were once connected.

Last fall that idea received about as much ratification as a scientific hypothesis can hope for. Dalziel and three other geologists organized a meeting to discuss how the Precordillera might fit into the geography of the ancient world. Researchers came from all over the world to the town of San Juan, on the eastern edge of the Precordillera. They took field trips out into the limestone mountains to see the fossils and strata for themselves; they listened to Ramos, Astini, Thomas, and their colleagues present their cases. And in the end they reached a unanimous verdict: the Precordillera was the long-lost piece of the Gulf Coast. As one geologist, George Viele of the University of Missouri, put it: If this isn’t Bill Thomas land, we have a lost continent floating out there like the Flying Dutchman.

The pre-Pangaean Earth, meanwhile, was beginning to look like Moores-Dalziel world. It was hard to see how North America’s southeast coast could have handed land off to South America’s west coast if the two hadn’t passed by each other. And it was hard to see how North America could have made an end run around South America without having previously been part of Rodinia, the southern supercontinent, in the way Moores and Dalziel had envisioned. Just how close the two continents passed after separating is in dispute; Dalziel is convinced they actually collided again, exchanging the Precordillera in the process. But most researchers at the conference were convinced by Astini’s fossil evidence that the Precordillera had crossed the Iapetus on its own, as an island. From the time it took for the fossil changeover, they even estimated how wide the ocean must have been: 1,200 to 1,800 miles.

The enthusiasm for North-South land swaps ran so high in San Juan that the Precordillera came to seem but the clearest example of the process. Ramos now thinks the crust straight west of the Precordillera, in the high Andes, was also once a southern extension of North America, which followed 100 million years behind and locked the Precordillera into South America. Some terranes may even have traveled the other way. A tract of alien rock trapped in the Piedmont of North Carolina, for instance, looks suspiciously South American to some researchers, as does the Oaxaca region of southern Mexico. It sounds as though we had to have some traffic police, Moores joked at the meeting. We had all these pieces that would have collided if they hadn’t kept to the right.

What I think things may have looked like, he adds, is what you see in the western Pacific today. And what you see are chunks of continents, and volcanic island arcs, and you see arcs colliding and continents running into arcs and this incredible complexity that’s changing very fast. Fast-changing complexity is not what the world of half a billion years ago used to look like; it used to be a blank. But that was before the new era of North-South cooperation.
Comment on this article

Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Collapse bottom bar

Log in to your account

Email address:
Remember me
Forgot your password?
No problem. Click here to have it emailed to you.

Not registered yet?

Register now for FREE. It takes only a few seconds to complete. Register now »