Wild horses and aurochs-proxies are turned loose on the reclaimed steppe

Begin at the pen gate, at first light, where the grass smells of frost and dung and something older than either. They opened it without ceremony. A Mandate warden named Saltanat Orozova lifted the latch, stepped back, and for a long minute nothing happened at all. Then a lead mare — a stocky, dun-colored thing bred back toward the wild horses that grazed this ground before anyone plowed it — walked out onto the retired steppe as if she had done it every day of her life. The rest of the herd followed, slowly, then all at once.

Eighty-four animals. Forty horses, forty-four of the heavy, back-bred cattle the restoration people call proxies for the aurochs that vanished from this basin centuries ago. They now hold, on paper, some sixty thousand hectares of ground that grew wheat within living memory, and then grew nothing much for two decades after the retreat pulled the farms back toward the water.

"Seeding gave us a carpet," Orozova told me, watching the herd string out along a shallow draw. "It did not give us a mosaic. Grass with no mouths on it becomes one thick thing, and one thick thing burns. We need them to open it — to graze it patchy, to churn it, to carry seed in their gut and their hooves. We are not planting a steppe. We are trying to hire one to plant itself."

The five-year number

Every animal wears a collar. The data runs to a monitoring station outside Turgai and, from there, into the Gaia Ledger, where it becomes a line in a biomass account that basins are ranked by. The Mandate has set a public target: a self-sustaining herd, breeding without supplemental feed or veterinary rescue, five winters from release. Below that number, the release is a subsidy. At or above it, the ground is said to hold on its own.

My colleague Henrik Vantaa would tell you the collars are the point. That a returned steppe is exactly and only what the Ledger can measure of it, and that sentiment is what happens when the numbers get too thin to defend. We have argued this before, and we will again. I will only say that I have stood in the draw where the mare led her herd, and the number for what I smelled there has not been invented yet. That does not mean the Ledger is wrong. It means it isn't finished.

Dr. Emin Karabekov, the population ecologist who designed the founding herd, is careful about the target. "Eighty-four is a genetic knife's edge," he said. "We built in redundancy — the studbook goes back four generations, we have avoided the obvious bottlenecks — but a bad winter, a disease we did not model, a single dry summer that thins the grass before the foals are weaned, and the curve bends down instead of up. We do not get to release them twice and pretend it is the same experiment."

The towns that stayed

The steppe is not empty, whatever the maps suggest. Turgai stayed. So did three smaller settlements built during the retreat, when families pulled upslope and inland and were promised the new ground would be theirs to work. Some of that ground is now inside the release zone.

I met a herder named Bekzat Amanov at the fence line. He was not admiring the mare. He was counting. "Sixty thousand hectares," he said. "Ask them how many were grazed by our animals last spring. Ask them who decided a wild horse needs it more than my sheep do." His grandfather, he said, was moved here on a Mandate promise. Now a different arm of the same Mandate has drawn a line through that promise and called it wilderness.

And who remembers? That's the question I keep in my jar with the soil. The Ledger remembers carbon and biomass. Amanov remembers a grandfather and a promise. Both entries are real, and the release honored only one of them.

Still, the herd moved out onto the ground as if it belonged there, because in the long arithmetic of this basin it once did. Orozova watched until the last collar cleared the rise, then closed the gate she no longer needed and pulled off her gloves.

"Now we find out," she said, "whether they can do the part we can't."