The 366 daily episodes in 2014 were chronological snapshots of earth history, beginning with the Precambrian in January and on to the Cenozoic in December. You can find them all in the index in the right sidebar. In 2015, the daily episodes for each month were assembled into monthly packages, and a few new episodes were posted. Now, the blog/podcast is on a weekly schedule with diverse topics, and the Facebook Page showcases photos on Mineral Monday and Fossil Friday. Thanks for your interest!

Monday, December 4, 2017

Episode 377: The Tepuis

Today’s topic for Episode 377, the Tepuis of South America, was suggested long ago by a listener.

Photo of Mt. Roraima by Jeff Johnson, used under  Creative Commons Attribution-Share Alike 3.0 Unported license.

The Tepuis are huge, high-standing plateaus isolated from their surroundings by near-vertical cliffs. The name means "house of the gods" in the language of the Pemon, the indigenous people who live in the region of northeastern South America where the tepuis are found. They’re especially numerous around the common borders of Venezuela, Brazil, and Guyana, and they include Mount Roraima, the setting for Arthur Conan Doyle’s 1912 novel “The Lost World”—a nearly inaccessible, remote, high, jungle-covered terrain. Doyle imagined the isolated preservation of dinosaurs and other extinct critters in his novel.

You commonly see the tepuis called the oldest plateaus on Earth, with suggestions that they are two billion years old. This is absolutely untrue: the rocks are indeed ancient, but the plateaus themselves as landforms are vastly younger, and there aren’t any dinosaurs – sorry. We’ll talk about both geological aspects of this unique ecosystem, the rocks and the landforms they make.

The area is part of the Guyana Shield, one of the ancient cores of the South American continent called cratons, from a Greek word for strength. Cratons make up the hearts of all the continents. In North America, multiple pieces of somewhat different age underpin most of Canada, with the Superior Craton extending into Minnesota, Wisconsin, and Upper Michigan. South America is made up of two large ancient cratons, the larger one in central and coastal Brazil, and the other, the Guyana Shield, in Venezuela and the Guianas and adjacent parts of northern Brazil.

The rocks that form the near-vertical escarpments of the tepuis were laid down as diverse sandy sediments probably about 1900 to 1500 million years ago, early to middle Proterozoic time. Some of the rocks that make up the Guyana shield are even older, back into Archaean time, more than two and a half billion years old, but they generally underlie the tepuis rather than form them. All these ages are similar to the cratons of the other continents.

Ancient Precambrian rocks have usually undergone multiple episodes of tectonic activity, burial, and heating, so they are mostly metamorphic rocks, which means changed form from their original sedimentary nature. The high cliffs that form the walls of the tepuis are mostly quartzite, not much different from the sand they originally were, and they are among the youngest of the Precambrian rocks in the Guyana Shield. They’re still relatively flat lying, not highly contorted like many ancient rocks, and because they are resistant, that helps them stand high and uneroded. But not completely unerodable. These quartzites even hold caves whose origin is not well understood.

That brings us to the formation of the plateaus themselves. That probably happened really very recently geologically speaking, as the result of erosion. The 6,000 feet of erosion that formed the Grand Canyon happened in just the past 5 or 10 million years or so, just yesterday, geologically, even though the rocks there are hundreds of millions to more than a billion years old. The cliffs of the tepuis are high, but not as high as the Grand Canyon is deep. Angel Falls, the highest in the world, drops off Auyan Tepui almost a thousand meters, 3200 feet. The entire tepui is somewhat higher, about 1,600 meters, almost at one mile or 5,200 feet. So the erosion that carved these cliffs, spectacular though they are, could be quite recent, like the Grand Canyon.

There is some evidence that the plateaus might be older than that, but still much, much younger than the rocks that form them. I’ve seen some suggestions that the tepuis formed as erosional plateaus as long ago as 70 million years. It’s almost impossible for me to believe that, because in what has been an area of abundant rainfall for a long time, I’d expect most topographic features to have been eroded away – or if not away, at least into minimal remnants of their original geography. Most, but not all, of the modern day surficial features we see on earth are not much more than 20 or 30 million years old, and most surficial expressions are much younger, shaped by glaciers, wind, and rivers.

In the Tepuis, there almost certainly was a pile of sedimentary rocks much younger than and on top of the rocks presently exposed, and those later rocks must have been eroded off. That probably began to happen when the South Atlantic ocean started to open during the Mesozoic, 170 or so million years ago, when the margins of the oceanic rift became relatively high-standing and subject to more active erosion. It would not have been until perhaps the last 10 million years or so that the tepuis began to attain their present shapes, with their thousand-meter cliffs. The present-day surface of the tepuis is probably flat because the flat-lying rocks of the ancient Roraima group are so much more resistant than whatever was above them. But, because whatever was eroded away is gone now, we don’t know directly what kind of rocks they were nor how thick they were. It’s not completely impossible for the present-day surface of the tepuis to be a really ancient surface, many tens of millions of years old. But I think that’s unlikely. Even resistant, flat-lying rocks are subject to erosion, and they’re being eroded actively today – the huge waterfalls are evidence of that. Nonetheless, the actual age of the tepuis’ surfaces is still debated.

Another line of evidence for the age of the tepuis comes not from geology, but from evolutionary biology. Patricia Salerno, a biologist at the University of Texas, studied mutation rates in DNA of diverse species of tree frogs to get an idea of how long ago those species had a common ancestor. She and her colleagues, writing in the International Journal of Organic Evolution in 2012, came up with 5.3 million years ago for the common ancestor of tree frog species on four different tepuis. That number jibes well with my arm-waving geologist’s guess that 5 or 6 million years of erosion might produce the landforms we see today. Does that mean the tepuis only formed that recently, isolating the various populations of frogs so that they evolved into the species we see today? Well, maybe, but not necessarily. It’s actually possible, and observations support this, that the modern species – or their ancestor – climbed those cliffs to the lush ecosystems at the tops of the tepuis, which might therefore have existed for millions of years before the frogs inhabited them. So we’re going to leave the exact time of origin of the tepui plateaus hanging – but my own feeling, as a geologist who hasn’t done any research there himself, is that as landforms, they’re likely at the very most 30 or 40 million years old, and perhaps even significantly younger. The rocks that make them up are close to two billion years old – but that does not make the tepuis themselves anything like that old. Read articles that call them two-billion-year-old landforms with skepticism.

Thanks to my friends, geologists Colleen Elliott and Petr Yakovlev for helping me clarify some of my thoughts on this. They are certainly not responsible for any mistakes I’ve made.

—Richard I. Gibson

Link to a comprehensive study of the Roraima Supergroup rocks (PDF):


  1. There is the question of what uplifted the Roraima sandstones ... it's not known, to my knowledge. Could be either pre-Pangea/pre-Gondwana collisions, or more recent collisions involving the poorly understood Caribbean plate. Our own work suggests an age of several tens of millions of years.

    More recent theories, Richard, suggest that the Grand Canyon might be 70 million years old, not 6 or 7; it almost surely predates the uplift of the Colorado Plateau through which it slices.

    Remember that erosion is often FASTER in drier areas, because there is so little vegetation to soak up moisture or bind the soil.

    Remember also that the porous sandstone of the Roraima Formation would allow a lot of water to infiltrate rather than developing surface runoff in streams. Groundwater can dissolve rock, of course – but the silica in sandstones dissolves very slowly. Estimates of erosion rates (actually, rates of retreat of tepui edges) per million years aren't that high.

    Cheers, Tom Givnish (