In the east of Brazil, mysterious cones of earth rise from the dry, hard-baked soil. Each of these mounds is about 30 feet wide at its base, and stands six to 13 feet tall. From the ground, with about 60 feet of overgrown land separating each mound from its neighbors, it’s hard to tell how many there are. But their true extent becomes dramatically clear from space.
Using satellite images, Roy Funch from the State University of Feira de Santana has estimated that there are about 200 million of these mounds. They’re arrayed in an uncannily regular honeycomb-like pattern. Together, they cover an area roughly the size of Great Britain or Oregon, and they occupy as much space as the Great Pyramid of Giza 4,000 times over. And this colossal feat of engineering is, according to Funch, the work of the tiniest of engineers—a species of termite called Syntermes dirus, whose workers are barely half an inch long.
Termites are well known for creating elaborate nests, with vast networks of underground tunnels. Many species create skyscraping chimneys atop these lairs to ventilate the underground chambers, and in some African species, these mounds can tower 30 feet high. But the Brazilian mounds are neither chimneys nor nests. They’re just amorphous lumps of soil, with no internal structures. Nothing lives inside them. Instead, “they’re just slag piles,” says Funch.
The local termites create extensive subterranean labyrinths of tunnels, and they eject whatever soil they don’t need on the surface. The process is slow and gradual, but the termites have been at it for millennia. They’re more like a geological force than an organism. Just as rivers or glaciers sculpt the landscape around them, so too have these tiny insects sculpted some 90,000 square miles of Brazil into a junkyard that’s visible from space.
These mounds are known as murundus, and rather confusingly, they’re different from other mounds called campos de murundus. The latter are found in the wet, savanna-like areas of southern and central Brazil instead of the dry northeast, and it’s not clear whether they’re built by termites at all. Campos de murundus have also been well studied for decades; by contrast, the northeastern murundus, though well known by locals, have been largely neglected by scientists.
Funch first saw them three decades ago, when he arrived in Brazil as a Peace Corps volunteer. After an unsatisfying stint doing administrative work in the nation’s capital, he found a more compelling life in Lençóis, a former diamond-mining town in the east. Surrounded by forests, waterfalls, and caves, Funch became a tour guide, national-park director, and environmentalist. And wherever he went, he kept noticing the striking mounds. He wrote about them in local popular-science magazines, but never managed to stoke much interest in them.
Since then, farmers have exposed more of the mounds in an attempt to create more grazing land by clearing local plants. Their efforts were mostly futile: The local soil is too poor for crops or livestock. But they did allow Funch and other researchers to start cataloging the murundus using satellite imagery. “It used to be all green and brown, but around eight years ago, Google Earth sharpened their images, and I could see the mounds that I had known from the ground,” he says.
The honeycomb distribution of the murundus is just one example of mysterious repeating patterns in nature. The famous fairy circles of Namibia and Australia—discs of bare red soil that pockmark miles of low grassland—are another. For years, scientists have argued about the cause of these circles. Some say they’re the work of termites. Others think they’re caused by the grasses themselves, battling for water and nutrients. Still others have argued that it’s a bit of both.
The murundus, Funch says, are likely to be much simpler in origin. “There’s no doubt,” he says, that they’re termite-made. “I’ve seen termites building the mounds with my own eyes.”
By examining murundus that had been sliced open by road-construction teams and probing them using optic fibers, Funch and his colleague Stephen Martin realized that each one begins when a termite colony builds a vertical tube rising straight up from its underground nest. The tube isn’t a chimney—it’s mostly closed at the top except for small side-holes around the rim. As the termite workers excavate their nest, they chuck soil out of these holes, eventually creating a cone. “There’s no engineering involved,” says Funch. “They’re just throwing the stuff out.” The cones aren’t even used as portals to the surface world: When termites forage at night on the forest floor, they emerge through temporary tubes between the murundus that they then seal during the day.
At first, Funch and his team thought each mound was the work of a separate termite colony—but that’s not the case. By collecting soldiers from beneath different mounds and seeing whether they would fight one another, they showed that a single colony can occupy the space beneath many cones. They need that space to survive. These termites live in an extremely dry ecosystem, so the leaves they forage upon drop rarely and erratically. To find enough fallen foliage, they need to expand their networks of tunnels to cover large tracts of forest. And to dig these tunnels efficiently, they need to make several waste mounds.
Funch’s colleague Paul Hanson analyzed sand grains from the center of 11 murundus using a technique that can measure when the grains were last exposed to sunlight. Using this method, he could work out when the first grains were buried, and thus when construction on each mound began. The youngest was 690 years old. The oldest was 3,820. That’s comparable to the oldest known termite mounds elsewhere in the world, and having studied only 11 murundus, Funch thinks far older ones likely exist.
Termites don’t live for 3,820 years, so do murundus lie unused for most of their life span after their creators die? Or, perhaps, do many generations of termites exploit the same mounds?
“Beats the shit out of me, man,” says Funch. “We have no idea.”
We want to hear what you think about this article. Submit a letter to the editor or write to firstname.lastname@example.org.