The Institute of Forest Genetics in Rhinelander, Wisconsin, was already studying tree breeding and the effects of ionizing radiation on trees when the Atomic Energy Commission came calling. The commission wanted to predict, as precisely as possible, what a nuclear disaster would do to American forest ecosystems. Similar experiments were underway at the Brookhaven National Laboratory on Long Island, where a pine-oak forest was being irradiated, and in the tropical rainforest of Puerto Rico. Wisconsin's northern hardwood forests represented a third distinct ecosystem type. The Enterprise Radiation Forest project officially started in 1968, when researchers leased 1,440 acres from Oneida County for $8,476 -- a bargain price for twenty years of access to an entire forest. Infrastructure went up: a perimeter fence, a small control building in the southeast corner, and a wooden stand for the radiation source. That source, recharged at Oak Ridge National Laboratory in Tennessee, contained 10,000 curies of caesium-137, which decays with a half-life of 33 years and emits penetrating gamma radiation. When not in use, it sat encased in a lead pig.

Irradiation began on May 3, 1972. For twenty hours each day, the caesium source bathed the surrounding forest in gamma rays. The remaining four hours were reserved for researchers to enter the area and gather data. Thomas Rudolph, the project's director, spent much of his time not in the forest but in meeting rooms, reassuring nervous locals and skeptical members of Congress that the experiment was safe. He had reason to be concerned about public perception: a radioactive source blasting a public forest in rural Wisconsin was not an easy sell, even during the height of Cold War anxiety. The experiment was designed to run for twenty years, long enough to observe multi-generational effects on the forest ecosystem. Instead, federal budget cuts ended the irradiation on October 16, 1972, after just one growing season. Researchers continued gathering data through June 1974, when the project was officially terminated. The site was dismantled, the source removed, and the land returned to public use.

The findings were both predictable and surprising. Radiation exposure in the Enterprise forest fell between the levels measured at Brookhaven's pine-oak forest on Long Island and the Puerto Rican rainforest, a gradient the researchers attributed to the varying complexity of the three ecosystems. Tree trunks proved to be significant radiation shields; species on the far side of large trunks received measurably less exposure. Damage varied dramatically by species and distance from the source. The biggest surprise came in the zone closest to the caesium-137 source. Researchers had predicted total devastation -- a dead zone. Instead, several species of lichen and some vascular plants survived. Lichen communities turned out to be roughly 250 times more resistant to radiation damage than coniferous forests. In the broader forest, researchers catalogued 193 species of vascular plants across 52 families, tagged and observed small mammals for behavioral changes, and measured radiosensitivity by interphase chromosome volume. The data contributed to the emerging science of radioecology, helping calibrate predictions for what nuclear fallout would mean for temperate forests.

The radiation study produced a remarkably thorough census of the forest's inhabitants, a snapshot of north woods biodiversity in the early 1970s. Johanna Clausen recorded 193 species of vascular plants around the primary study site alone, with another 111 species elsewhere in the fenced area. The canopy was dominated by quaking aspen, paper birch, red maple, and sugar maple, while the forest floor hosted wild sarsaparilla, large-leaved aster, sedge, wild lily-of-the-valley, and bracken fern. Thirty species of mammals were identified between 1968 and 1972, from short-tailed shrews and masked shrews to coyotes, red foxes, porcupines, and white-tailed deer. Amphibians included American toads, spring peepers, green frogs, and wood frogs alongside red-backed salamanders. Reptiles ranged from red-bellied snakes and garter snakes to western painted turtles. Frederic Erbisch counted 30 species of lichen. The forest, roughly half wetland and half second-growth deciduous woodland, with trees averaging just 24.4 years old in 1970, was young and thriving -- before, during, and after the radiation.

No long-term effects of the radiation experiment on the forest ecosystem were ever identified. The trees that survived the growing season of 1972 kept growing. The lichens that shrugged off 250 times the dose that would kill a conifer continued to colonize bark and rock. The deer mice and chipmunks tagged by researchers lived out their generations without observable legacy effects. Today the Enterprise Radiation Forest is simply the Enterprise block of the Oneida County Forest, open to the public, unremarkable to anyone who does not know its history. The control building is gone. The fence is gone. The wooden stand that once held 10,000 curies of caesium-137 is gone. What remains is a forest that absorbed a Cold War experiment and carried on, its resilience the most important finding of a study that was cut short before it could fully measure what it set out to learn. The lichens, it turns out, were tougher than the federal budget.