China Released Horses Into A Barren Desert With No Grass—Nobody Predicted This Outcome

追加: 2026-06-03

[00:00:03]It's their first taste of freedom.

[00:00:05]Researchers are releasing several Przewalski horses into the wild. It's the only truly wild >> December 1986, 11 horses step off a cargo plane into a desert so barren the local ranchers no longer called it land. They called it the place where nothing grows. No grass, no water, winters that hit 40 below zero. China had just spent millions flying these zoo-born animals from the other side of the world to release them here.

[00:00:33]And every serious ecologist on the planet called it a death sentence. A BBC crew flew in to film the release. They packed up and left after two days convinced they were witnessing a disaster in real time. What those 11 horses did next nobody believed possible. 11 horses against a dead desert because these are not ordinary horses. They are Przewalski's horses, the last truly wild equine species left on the face of the earth. They had been extinct in this region for 17 years.

[00:01:04][music] The Chinese government had just spent millions of dollars flying them in from zoos in East Germany and the United States only to drop them into a desert that devours 3,000 km of new terrain every year. The best ecologists already wrote off the project as a death sentence. These were animals that had spent their entire lives eating hay from buckets and drinking from automatic waters. Now they were being released into a place where winter temperatures plunge to 40 below zero and summer peaks hit 40° C.

[00:01:36]The nearest water source could be 15 km away buried under 10 cm of solid ice. The horses themselves look almost alien.

[00:01:46]Stocky bodies, stiff mohawk-like manes that bristle [music] straight up like something out of a child's drawing.

[00:01:53]Faint zebra stripes running down their legs. They look less like modern animals and more like those cave paintings our ancestors scratched onto stone walls 40,000 years ago. These are Przewalski's horses, >> [music] >> the only wild horse species left on Earth. Experts in desert survival gave them, at best, a 15% chance of surviving the first winter. One ecologist wrote in Nature magazine that the whole project was throwing good money at dead land.

[00:02:20]Local officials in Jimusar County were already drafting public statements to explain the inevitable disaster. And to understand why anyone would attempt something so insane, you have to understand what the Junggar Basin had actually become by 1986. Because this was not a simple conservation story.

[00:02:39]This was about saving an entire region of China from becoming completely uninhabitable.

[00:02:44]>> [music] >> A wasteland that was swallowing China.

[00:02:47]The Junggar Basin was not always a wasteland. For 10,000 years, wild horses [music] had roamed these steppes. The entire ecosystem had evolved to the rhythm of their hooves. And then, in 1969, they vanished. Not gradually, completely. The last confirmed sighting was recorded by a Soviet geological survey team. A few hoof prints in dry mud, a handful of brown hairs caught in desert grass. Then silence. And here is the part nobody saw coming. What followed was an ecological collapse on a scale that left even the most pessimistic scientists stunned. Without horses to pound the surface, the soil hardened into something researchers started calling biological concrete.

[00:03:31]Rain still fell on the region, 10 to 15 cm every year. [music] But instead of soaking into the Earth, the water slid across the impermeable crust [music] and evaporated within hours. Seeds could not break through the surface to germinate. Nitrogen levels plummeted so brutally that by 1974, scientists were comparing soil samples from the basin to lunar dust. The material was still there. The grains of sand and silt still looked the same under a microscope. The basin had simply become biologically dead. And then the desert started to move. Satellite images from 1982 showed the Gobi expanding at 70 km per decade, swallowing grasslands, villages, and entire farmlands. Roads had to be relocated. Rail lines were buried. Beijing, over 1,500 km away, suffered dust storms that lasted for days. Hospitals recorded a 300% increase in respiratory emergencies. People were choking on a desert they could not even see, breathing dirt from a place 1,500 km away. So, China tried everything. The Great Green Wall project, launched in 1978, aimed to plant a forest barrier 4,500 km long along the country's northern border. But, here is the catch. The trees needed more water than the desert could offer. Water tables, stable for centuries, dropped 12 m in some regions.

[00:05:01]>> [music] >> The trees themselves, planted in neat rows like soldiers, began to die from the roots up as they could not reach the ever deeper water. Entire villages had to be relocated when their wells dried up. Engineers [music] tried sand-fixing grids made of straw. They tried planting drought-resistant shrubs. Nothing held the line.

[00:05:21]>> [music] >> By 1985, the situation was a full-blown crisis. Agricultural productivity in Xinjiang province had fallen 34%. The desert was advancing on Beijing faster every year. Computer models projected that, at the current rate, the Chinese capital would be virtually uninhabitable by 2050. And that is the disaster a small group of Chinese ecologists, >> [music] >> led by a researcher named Yong Chen Ming, were staring at when they walked into a government meeting in 1985 and proposed something that sounded like fantasy. Yong stood before officials who had spent two decades watching tree after tree die, watching wells go dry, watching dust devour their capital. And he told them to stop. Stop fighting the desert with bulldozers and tree plantations. Bring back the animal that had kept this place alive since the last ice age. Bring back the wild horse. The room, by all accounts, did not applaud. But the numbers were so bad, the failures so absolute, that within months his plan was approved, which left exactly one question hanging in the air. How could zoo-born animals, hand-fed by humans for three generations, survive in a place that was swallowing cities? The last truly wild horse, the Przewalski's horse, is not just any horse. If you look at its DNA, you will find 66 chromosomes instead of the 64 found in every domestic horse. It is the only equine species on Earth that has never been domesticated, never broken, never tamed. While humans bred Arabians for speed and Percherons for strength, Przewalski's horses [music] stayed exactly the same as when our ancestors painted them on cave walls in the deep past. They are a living window into the ice age. But here is the wild part. In 1985, that window was almost closed.

[00:07:14]Only about 900 of these animals remained on the entire planet, and every single one of them lived in a zoo.

[00:07:20]>> [music] >> The entire species descended from just 12 to 14 individuals captured between 1899 and 1903 by Russian explorers working on commission for European zoo collectors. After World War II, >> [music] >> the global population hit rock bottom.

[00:07:35]31 horses, of which only nine could breed. At one terrifying moment, the species survived in only two places on Earth. The Munich Zoo and the Prague Zoo. Every living Przewalski's horse today carries that genetic bottleneck in its cells. A chromosomal scar from being millimeters from vanishing forever. The mere fact that any were left to send to China in 1986 was in itself a small miracle. China's plan drafted in 1985 and launched in 1986 had three phases.

[00:08:07]First, import horses from Western zoos and acclimatize them at a new breeding center built in Jimsar. Second, move them into huge semi-wild enclosures, essentially teaching zoo animals how to be wild again.

[00:08:20]Third, release them into the Kalameli Nature Reserve for 1,000 km of some of the harshest terrain in Asia. Among the first horses moved to the Jimsar enclosures was a young colt with an immaculate white blaze running down his face. The keepers called him Hong.

[00:08:37]Remember that name. Now is where things get interesting. Hong's argument to the government was not just that these horses would survive. It was that they would actively rebuild the ecosystem.

[00:08:49]"These horses do not just eat grass," he said. "They design entire landscapes.

[00:08:55]And once you understand how the whole story stops sounding like madness. It all comes down to three things.

[00:09:01]>> [music] >> The first is brute force. Each horse weighs between 270 and 360 kilos. When they run, their hooves strike the ground with over 5 and 1/2 kilos of pressure per square centimeter. More force than a jackhammer. In a place like the Junggar Basin where the soil had turned into that biological concrete, that is exactly the force needed to crack the hardened crust and let water back in.

[00:09:25]But here is what machines cannot do.

[00:09:28]The horses are selective. They circle around young, fragile plants. They only pound bare soil and dead vegetation.

[00:09:36]They know the difference. Then there is the way they move. GPS tracking would later reveal that Przewalski's horses cover about 15 km a day in what mathematicians call fractal patterns.

[00:09:47][music] They graze intensely in one spot for hours, then suddenly shift kilometers away, and later return. It looks random. It is not. This pattern creates a disturbance mosaic.

[00:10:00]Short grass here, tall grass there.

[00:10:03]Bare patches in between. Each tiny microhabitat ends up supporting different species of insects, birds, and small mammals. The horses are not just eating.

[00:10:14]They are designing a tapestry. And pay attention to the third, which is the part that sounds most absurd. Turns out that horse dung is not waste. It is a delivery system. Analysis of Przewalski's horse droppings revealed viable seeds from more than 20 different plant species. And the journey through the horse's gut [music] boosts germination rates by up to 300%.

[00:10:36]The acids in their digestive system scarify the tough seed coats, breaking dormancy in seeds that would otherwise have remained locked for years. A single gram of dung contains up to 7.4 viable seedlings. That is double the dispersal rate of cattle and eight times that of sheep. And because these horses cover those 15 km every day, they are seeding vast areas, building genetic corridors between plant populations that had been isolated for decades. Each pile of dung is, in a very real sense, a self-watering pre-fertilized seed packet. That was Young's proposal. That was the theory.

[00:11:14]The problem was that the horses about to be released had spent three to four generations behind bars. They did not know how to find water in a desert. They did not recognize a wolf's howl [music] as danger. They had never experienced a temperature below freezing or above 32°.

[00:11:31]They had lost what researchers called the culture of survival.

[00:11:35]The learned behaviors passed down from mother to foal that no amount of raw genetic programming can replace. So, the question that would decide whether Young was a visionary or a fool was simple.

[00:11:45]Had three generations of zoo life erased the instincts buried in these animals or were they still there >> [music] >> somewhere waiting? Before I go on, before I tell you what happened when these horses actually stepped onto the floor of that desert, do me a small favor. If the story is interesting to you, hit the subscribe button and turn on the bell. Every week we dig into stories like this, the kind of stories that rewrite what we thought we knew about how the natural world really works. And the next part of the story is the part nobody believed possible. What those zoo-born animals did in their first winter. What one colt did on the eighth day of the release. And the number the researchers ended up calculating for what a single one of these horses is worth to the planet.

[00:12:29][music] Stay with me. The payoff is coming. A winter that should have killed them. The first months in Jimsar were a battle that nearly proved the critics right.

[00:12:39]Within weeks botfly infestations hit 100% of the herd and these were not ordinary parasites. The larvae burrowed into the horses' stomachs creating painful lesions that wrecked their digestion just when they needed every calorie to survive. Then came the first winter and the first winter was apocalyptic. Temperatures dropped to 42° below zero. The horses, accustomed to heated stables, huddled against freezing winds that cut across the desert like knives. Their distinctive Mohawk manes froze into ridges of ice so thick that some horses could not lift their heads off the ground. Snow buried what little grass there was under a meter of powder.

[00:13:22]And here is where it gets brutal. The Jimsar staff, Young's team, >> [music] >> watched through binoculars and were forbidden to intervene unless the whole herd risked dying. Seven horses developed severe frostbite. Two mares miscarried in the cold. The international press was already writing the project's obituary. Conservation magazines were preparing postmortems of what they assumed would be a textbook failure. Young himself, as he later admitted, was bracing for the worst. But something else, something nobody had scripted, was starting to happen on the frozen steppe. The foals began to break the ice with their hooves to reach the water beneath, a behavior never observed in captivity in any zoo in the world.

[00:14:06]The mares started to form protective formations around their foals during storms, using their own bodies as windbreaks and rotating positions when one mare got too cold. The horses learned to dig through the snow with a sweeping motion of their forelegs, uncovering grass that no domestic horse would have thought to look for. They started to read the wind, recognizing which hillsides had been swept clear of snow and which held deep drifts. And that is putting it mildly. They were not just surviving. They were remembering.

[00:14:37]Generations of zoo life had not erased what was written in their bones. The instincts were still there, dormant, waiting. The desert was waking them up.

[00:14:47]By spring 1987, the change accelerated beyond anything projected. Grass species that had not been seen in the basin in 17 years began to sprout in the areas where the horses had grazed.

[00:15:00]Compensatory growth, the phenomenon by which grasses regrow faster and thicker after being clipped, increased biomass by 43% in grazed areas compared to ungrazed controls. The horses were not destroying the land.

[00:15:14]They were stimulating it. The soil itself began to change in ways that stunned the scientists monitoring it. In that first year alone, moisture retention jumped 23%.

[00:15:25]Not because rainfall had increased, [music] not because the climate had changed, but because the horses' hooves had hammered about 10,000 tiny craters into every hectare of ground. Each one of those depressions became a miniature reservoir, trapping rainwater that would otherwise have slid across the hard crust and evaporated within hours. The horses were rebuilding the basin's plumbing one hoof print at a time. The second year brought another surprise. Nitrogen levels shot up 37%.

[00:15:55]The horses' dung was not just fertilizing the soil. It was reintroducing the gut bacteria that the soil had been missing for decades. An entire microbial community that had vanished when the wild horses vanished.

[00:16:07]By the third year, botanists were documenting something that should have been impossible.

[00:16:12]22 plant species had returned out of nowhere. Their seeds, apparently, had been lying dormant in the soil for nearly two decades, waiting for the right conditions to germinate. The basin had not been dead.

[00:16:26]It had been sleeping, waiting for the one species that knew how to wake it up.

[00:16:30]But the thing is, surviving one winter is not restoring a population. The horses had to do it again over and over. And the second winter, when it came, would test whether that first miraculous season was the beginning of something real or an unrepeatable fluke. The ecosystem awakens. It was not a fluke. The horses survived the second winter and the third and the fourth. And as soon as the plants returned, everything else followed them. Dung beetles, absent from the basin for 15 years, came back by the millions. They buried the horse dung, further enriching the soil, accelerating the recovery loop. Field surveys began recording 43 insect species that had not been documented in the region in decades. The insects attracted birds.

[00:17:19]Larks, Isabelline wheatears, and sandgrouse moved in to feed. Small mammals followed the birds. Jerboas, pikas, ground squirrels. And then the predators returned.

[00:17:31]Foxes, steppe eagles, and even the occasional lone gray wolf began appearing on camera traps. Each new species added another layer of complexity, another feedback loop driving the recovery. By the late 1990s, the Jim Star herd had grown large enough and the basin had recovered enough for Young's team to make a decision that a decade earlier would have sounded insane. They were going to take down the fences. In August 2001, the project entered its boldest phase.

[00:18:02]>> [music] >> 27 horses were selected for full release into the Kalamely Nature Reserve. No fences, no supplemental feeding, no human intervention of any kind. Just horses in 4,000 km of desert that seen their kind in 32 years. Among the 27 was Hong, the white-blazed colt, >> [music] >> now grown and leading his own harem. The scientists fitted GPS collars, set up camera traps, and held their breath.

[00:18:30]For the first week, the horses stayed within 500 m of the release point as if tethered to invisible chains. They moved in circles. They grazed nervously. They kept looking back toward the enclosures.

[00:18:44]And then came the eighth day. The eighth day.

[00:18:47]>> [music] >> It was just after dawn. The wind, according to field notes, was blowing steady from the northwest. The GPS team at the monitoring station was watching the screen almost by routine when one of the dots on the map started to move.

[00:19:02]Then the dots around it began to move.

[00:19:05]Hong had risen and his harem was lining up behind him. He was not wandering. He was not grazing. He was walking north in a straight line as if he knew exactly where he was going. The dots kept moving, 5 km, 10, 15. The monitoring team started making calls. When Hong stopped, he had led his harem 30 km north to a water source that was on none of the project's maps.

[00:19:32][music] Nobody had told him it was there. No previous generation of his bloodline had walked that path in their entire lives. He simply knew. The team drove out the next morning to confirm what the GPS had already told them. And there it was, a spring-fed water hole half frozen tucked against a low ridge. Hong stood beside it as if he had visited it his whole life. How he knew remains one of the great unsolved mysteries of the operation. Some ancestral memory, some scent on the wind, some pattern in the geology that horses can read [music] and we cannot. Within a month, the horses had carved out territories. Stallions were marking boundaries with dung piles deliberately placed on high ground.

[00:20:18]Natural birth rates hit 87% higher than most zoo populations achieved. And the foals born there in the wild exhibited behaviors that captive-born horses simply do not show.

[00:20:30]They could distinguish 23 different plant species by smell. They reacted to the shadow of a golden eagle by freezing instantly. They grew thicker winter coats than their parents. And the landscape change was already visible from space. Satellite images showed green patches expanding outward from the horses' core territories like slow-motion explosions. The water table, which had been falling 3 cm every year since 1970, stabilized in 2003 and then it started to rise. The horses had reignited a hydrological pump that had been off for a generation. The number that broke the model.

[00:21:09]What was happening in the Junggar Basin shattered everything China thought it knew about fighting deserts.

[00:21:15]For 25 years, the Great Green Wall had been built on a single premise.

[00:21:19]Plant trees to stop sand. By 2010, China had planted 66 billion trees at a cost of 47 billion dollars.

[00:21:28]>> [music] >> And in many regions, the trees had made the underlying problem dramatically worse.

[00:21:33]Sucking water from the soil faster than the land could replenish it. And here is the part that should have changed everything.

[00:21:40]While China's colossal tree planting program was literally drying out the desert, the horses were doing the exact opposite. They were bringing the water back. Their hooves created infiltration points that allowed the scant rainfall to penetrate deep into the soil >> [music] >> rather than evaporating off the crust.

[00:21:57]Their grazing maintained grasslands, which need 90% less water than forests.

[00:22:03]A single horse improved water retention across 50 hectares every year. Then someone ran the numbers and the figures stopped sounding like ecology and started sounding like economics. When ecological economists sat down to calculate how much a single Przewalski's horse was actually worth to the surrounding ecosystem, they came up with a number nobody expected. Soil restoration valued at $12,000 per horse per year. Seed dispersal, 8,000. Water retention, 15,000. Carbon sequestration, 7,000.

[00:22:36]Biodiversity support, 5,000. Total, $47,000 per horse per year. When they calculated the cumulative total, the economists were the next group after the BBC crew and the conservation magazines >> [music] >> to need a second review of their results before they trusted them. Because by 2023, those 11 horses had become 534.

[00:22:59]The herd was effectively managing 26,700 [music] hectares of land, an area larger than many national parks, generating $25 million in ecological benefits every year. And that was a conservative estimate that did not include tourism, scientific research value, or what the basin was now worth in carbon offsets. Among those 534 horses were by then the descendants of Hong, dozens of them, scattered across the reserve in harems of their own. The dead land they had been released into now supported more than 20 plant species that had not grown there in decades. And the desert that had been expanding thousands of square kilometers every year was now retreating, pulling back at a rate of 47 hectares each year in the areas where these horses roamed. The model spread rapidly.

[00:23:49]Mongolia released horses in Hustai National Park in 1992, and by 2023, the population there reached 490 individuals, with the surrounding steppe showing the same restoration patterns as Xinjiang.

[00:24:02]Kazakhstan launched its Golden Steppe Project in 2004, shipping horses from Prague and Berlin to the Kostanay region. Eight years later, there were over 100. Now, nearly a thousand have been released to roam Central Asia. Russia has been quietly studying the Chinese data. Each project kept proving the same principle. You cannot engineer an ecosystem from the top down with human designs and concrete. You have to restore it from the bottom up with evolutionary solutions that have already been tested for thousands of years. China is now expanding the Kalamely Reserve by another 350 km. The official target is to reach 2,500 wild horses by 2040.

[00:24:46]Computer models suggest a population of that size could restore over 100,000 hectares of degraded land, trap 43,000 tons of carbon every year, and support over 300 species of plants and animals that currently exist nowhere else in China. If they pull it off, the Junggar Basin will become the first desert in human history to be reversed not with tree plantations, not with irrigation, not with engineering, but with the hooves of wild horses. The 11 animals that stepped off that cargo plane in 1986 were not just a few endangered animals relocated. They were a biological technology refined by evolution over hundreds of thousands of years, capable of designing and maintaining an entire landscape. We just had to remember how to turn it on. We just had to stop trying to do it ourselves and let the experts, the ones with hooves, take over. Przewalski's horses force us to ask an uncomfortable question. How many of the environmental disasters we are facing right now could be solved by bringing back the original engineers we drove to extinction, rather than throwing trillions of dollars at human-designed solutions that do not [music] work?

[00:25:54]The bison that shaped the American prairie, the wolves that keep Europe's forests in check, the elephants that cultivate Africa's savannas. [music] We have spent the modern era trying to engineer nature. Nature had already designed the solution.

[00:26:10]>> [music] >> We just killed it and then forgot the recipe. Drop your thoughts below in the comments. If you could bring back one extinct ecosystem engineer to any place on Earth to repair the land we have ruined, what species would you choose and why? I read every comment. And if you want more stories about animals that have literally reshaped continents, the subscribe button is your next click.

[00:26:32]>> [music] >> The next story we are working on might be the wildest one yet.