kept both ways at once — stopped in a freezer, moving in the wild — until the stopped one rejoined the moving one
The sourdough library on the previous page keeps a thing two ways at once: feeding it, which keeps it alive by replacing it, and freezing its strains, which keeps it exact by stopping it. The two copies sit side by side and never meet. This page is about a thing that was given both treatments — and where, decades later, the two copies did meet.
The stopped copy lives in San Diego. The Frozen Zoo, founded in 1972 by Kurt Benirschke at what is now the San Diego Zoo Wildlife Alliance, banks living cells in liquid nitrogen at about −320 °F. It was begun on a collector's principle: preserve material before you know what it will be for. At its fiftieth anniversary the Alliance put the bank at more than 11,500 individuals from 1,337 species — cells, embryos and gametes from mammals, birds, reptiles, amphibians and fish. Nothing in it is fed. Nothing in it ages, in the ordinary sense. It is a zoo whose animals are all, deliberately, paused.
The moving copy is a species. Black-footed ferrets — North America's only native ferret, a nocturnal hunter of prairie dogs — were thought extinct until a small colony was found in Wyoming in 1981. Eighteen animals were brought into a captive breeding programme; only seven of them left descendants. For the next four decades, every living black-footed ferret — hundreds in captivity, hundreds more reintroduced to the wild — descended from those seven. A living chain, kept moving the ordinary way, birth by birth, and narrowed to a bottleneck seven animals wide.
One ferret stood outside that chain. Willa was a wild female from that last colony. She died in 1988 without offspring, so by the living chain's arithmetic her genes ended with her. But tissue from her had been sent to the Frozen Zoo, where it stayed, stopped, while the seven-founder chain ran on without it. A later genomic study found her genome carried roughly three times more unique variation than the average in the living population — the paused copy had quietly become the wider one.
Two tracks, then one. Step through it. The top line is the living chain; the bottom line is Willa, stopped. Watch how long each runs before they are the same line again.
1981 · the last colony is found in Wyoming
Living: a breeding programme begins. · Willa: alive, wild, unrecorded in any pedigree.
Dates and names below are the public record: U.S. Fish & Wildlife Service announcements and Revive & Restore's project log. The first clone could not breed; the joining happened through her two later copies, not through her.
In June 2024 Antonia — a clone of Willa, born the previous year — gave birth at the Smithsonian's National Zoo & Conservation Biology Institute, after mating with a male called Urchin. Three kits were born; two survived, a female, Sibert, and a male, Red Cloud. They were the first offspring of a cloned endangered animal in a conservation programme, and the first black-footed ferrets descended from eight founders rather than seven.
Read that slowly, because it is the whole finding. Willa did not become a founder when she was collected, or when she was frozen, or even when she was cloned. She became one thirty-six years after her death, on the day her genes produced living young inside the chain she had missed. The freezer had not held a spare copy of the species. It had held a participant — one who arrived at the breeding programme a generation late, by the stopped route, while everyone else arrived on time by the living one.
So the holder's two methods turn out not to be alternatives after all, at least not always. The sourdough library's feeding and freezing never rejoin: the jar is eaten from, the freezer sample is not thawed into lunch. But here the stopped version was kept precisely so that it could rejoin, on a day nobody in 1988 could have scheduled — cloning did not exist as a practical tool when Willa's tissue was banked. The collection was made for reasons its collectors said, in advance, they did not yet understand. The understanding arrived thirty-two years later, wearing Willa's face, in a facility in Colorado; the rejoining took four years more, and a second and third copy, before it produced kits.
What the two tracks hold, between them, is a small correction to the idea of a founder. A founder is usually counted at the start, among the living, in the year the chain begins. Willa shows the count is not closed at the start. Anything kept stopped beside a living chain — faithfully, at −320 °F, for longer than the species' ordinary generation time many times over — is a founder whose founding is still pending. It counts, or it doesn't, on a date in the future that the keeping itself cannot name.