A century ago, roughly one in four hardwood trees in the eastern United States was an American chestnut. The species was widespread, practical, and highly regarded for its wood and its annual nut crop. In about forty years, a fungus introduced accidentally on imported seedlings killed an estimated three to four billion of these trees, making it one of the most destructive environmental events in American history. For the past century, the species has survived mainly through small sprouts from living root systems that cannot grow large enough to reproduce. A major genetic rescue effort, which has included a significant scientific setback, is now attempting to bring the tree back.
Before the twentieth century, American chestnuts made up between a quarter and half of the hardwood canopy across large areas of the Appalachian Mountains and eastern forests, from Maine to Georgia and west to the Ohio Valley. The tree was valued for practical reasons. Its wood was straight-grained, lightweight, and notably resistant to rot, making it a preferred material for fences, furniture, railroad ties, and cabins. The bark supported a regional tanning industry, and the nuts, produced in large quantities each fall, fed wildlife and free-range livestock. In many rural Appalachian communities, families collected and sold the nuts as a reliable seasonal source of income.
The loss of the chestnut affected more than the timber and nut trade. Botanists describe the tree as a keystone species, meaning its presence disproportionately shaped the health and structure of the ecosystem around it. Dozens of insect species depended on chestnut leaves and had no alternative host once the trees disappeared. Bears, deer, wild turkeys, and squirrels relied on the annual nut crop before winter. The chestnut was a more dependable food source than the oak trees that later filled much of the ecological gap.
The disaster began quietly in the late nineteenth century. Imported Asian chestnut seedlings, brought to the United States for ornamental purposes, carried a fungus later named Cryphonectria parasitica. That fungus had evolved alongside Asian chestnuts and caused them little harm. The American chestnut, which had never encountered the pathogen, had no resistance at all. The fungus was first identified in 1904 by Hermann Merkel, a forester at the New York Zoological Park, now the Bronx Zoo, where he observed chestnut trees dying from an unfamiliar bark disease.
The fungus kills tissue in the bark and cambium layer, cutting off the flow of water and nutrients above the point of infection. It produces oxalic acid, a compound that kills surrounding plant tissue and allows the fungus to spread. The American chestnut had no biological defense against this chemical attack, leaving even healthy young trees with little chance of survival once infected.
The disease spread faster than foresters expected. From the first documented case in the Bronx in 1904, the fungus moved outward through spores carried by wind, birds, and insects, expanding its range by tens of miles per year in some directions. By the 1920s, it had destroyed chestnut populations across most of the Northeast. By the 1940s, it had swept through the entire natural range, from New England to the southern Appalachians. An estimated three to four billion mature trees died during that period.
People who witnessed the collapse described entire mountainsides of chestnut trees dying within a few seasons. Rural communities that depended on the annual nut harvest watched that source of income disappear within a generation. Newspaper accounts from Appalachian communities in the 1920s and 1930s described hillsides turning a ghostly gray-brown as dead chestnuts stood bare among healthy trees of other species. Efforts to slow the spread through quarantine zones and the removal of infected trees proved ineffective against a pathogen whose spores could travel long distances on the wind.
One biological feature made recovery possible. The fungus usually kills only the above-ground portion of an infected tree, leaving the root system largely intact. Those surviving root systems have spent the past century sending up new sprouts each year in repeated attempts to regrow. Because the fungus is still present in eastern forests, the sprouts typically become infected and die before they can produce their own seeds. Foresters sometimes refer to this as a living-dead phenomenon: a species that technically continues to exist through root systems across millions of acres, producing sprout after sprout for a hundred years, but functionally unable to complete its reproductive cycle.
Because the roots survived, restoration researchers had something to work with. A walker through eastern Appalachian forests will likely pass dozens of these surviving sprout clusters without recognizing them. They appear as low, bush-like growth, rarely more than a few feet tall before blight infection kills them back. They can be mistaken for a different, unremarkable kind of understory plant unless one knows the distinctive toothed leaf shape and unique bark texture. The transformation of a once-dominant forest giant into anonymous knee-high sprouts scattered through the woods has been one of the more poignant aspects of this story.
The first serious, sustained effort to breed resistance into those surviving roots began under the American Chestnut Foundation, established in 1983 to coordinate a long-term restoration strategy. The foundation’s core approach has been a technique called backcross breeding. It involves crossing American chestnuts with Chinese chestnut trees, which are naturally resistant to blight, then crossing the resulting hybrids back with native American chestnuts over multiple generations. The goal is to retain the genes responsible for the Chinese chestnut’s natural blight resistance while eliminating as much of the rest of the Chinese chestnut’s genetic material as possible. The aim is to produce a tree that is mostly American chestnut, about 15 parts out of 16 by the time the breeding program reaches its later generations, while carrying enough inherited resistance to survive blight infection in the wild.
This approach requires unusual institutional patience. Each generation of backcrossing takes years to reach testing age, and each generation must be deliberately exposed to the fungus and assessed for resistance before breeders know which trees to advance to the next round. Decades of this work have produced trees with improved blight tolerance, although reaching resistance levels comparable to the fully resistant Chinese parent species has proven more difficult than early genetic theory suggested.
The scale of the effort is considerable. The American Chestnut Foundation coordinates work across dozens of state chapters, university research orchards, and thousands of individual volunteer growers across the tree’s historical range. Each volunteer maintains a small collection of test trees, recording flowering dates, growth rates, and results from fungus exposure, and exchanging seed and pollen shipments to support genetic diversity in the broader breeding program. This distributed, volunteer-supported infrastructure, sustained for more than four decades, represents a rare example of multi-generational citizen science organized around a single conservation goal.
A parallel scientific approach developed at the State University of New York College of Environmental Science and Forestry, under researchers William Powell and Charles Maynard, working within the American Chestnut Research and Restoration Project. Rather than breeding resistance through generations of hybridization, this team pursued direct genetic engineering. They introduced a single gene from wheat, not from another chestnut species, that produces an enzyme called oxalate oxidase, abbreviated as OxO. This enzyme neutralizes oxalic acid, the same toxic compound the blight fungus relies on to kill surrounding tree tissue. Instead of killing the fungus or preventing infection, this method disarms the fungus’s primary chemical weapon, allowing the infected tree to tolerate and contain the infection rather than die from it. Wheat and several other plants naturally tolerate this compound using similar enzymes.
The resulting genetically modified strain, developed under the working name Darling, showed promising blight tolerance in field trials that began around 2017. The tolerance appeared theoretically stronger than what decades of traditional backcross breeding had achieved. The single-gene approach also had a practical advantage in that it involved introducing one specific, well-understood gene rather than attempting to isolate and recombine a set of resistance-related genes distributed across the genome of a different species. The resulting tree retains almost all of its original American chestnut genetic identity, rather than the approximate 15/16 proportion that traditional breeding can achieve.
By 2020, the genetically modified strain known as Darling 58 appeared close to a significant milestone. Researchers formally petitioned the U.S. Department of Agriculture for deregulation, which would allow the tree to be planted and distributed without the restrictions usually applied to genetically modified organisms. Approval would have made Darling 58 the first genetically modified forest tree approved for release into wild U.S. ecosystems. The petition began a lengthy multi-agency regulatory review, since a genetically modified forest tree intended for release into wild forests, rather than confined agricultural use, represented uncharted territory for federal regulators. The USDA, Environmental Protection Agency, and Food and Drug Administration each examined different aspects of potential environmental impact and safety. Field trials continued throughout the review, with researchers tracking blight tolerance, growth rates, and overall tree health across multiple test sites.
In December 2023, that momentum encountered a serious problem. The American Chestnut Foundation, which had been a close partner with SUNY ESF and provided significant support for the Darling 58 project, announced that it was withdrawing its support entirely. The reason was a classification error in the laboratory. Researchers at SUNY ESF discovered that the trees planted, tested, and reported for years under the name Darling 58 were actually a different genetic strain, now referred to as Darling 54. This strain produces significantly lower levels of the protective OxO enzyme and therefore has weaker blight resistance than the true Darling 58 was believed to possess. Years of published field data submitted to the public and as part of an active federal regulatory petition had described the performance of the wrong tree.
The fallout within the restoration community was significant. Media coverage at the time described the resulting disagreement between researchers defending the underlying science and skeptics questioning whether the entire genetic engineering approach could be trusted, in harsh terms, some describing it as a civil war within the broader chestnut restoration movement. The foundation’s public statements cited the misclassification error and what it described as disappointing performance results from the strain actually growing in the ground.
The mechanisms by which an error this large went undetected for years illustrate how difficult it is to conduct this kind of long-term forestry research with precision. Tracking thousands of individual trees across multiple field sites, each grown from carefully labeled but visually similar seedlings over long periods, creates many opportunities for such record-keeping errors, especially in a research field with no established long-term protocol for genetically modified forest trees to follow.
SUNY ESF acknowledged the error directly while continuing its core research program, now operating under the corrected name Darling 54. Andrew Newhouse, director of the American chestnut project at SUNY ESF, was notably direct in public comments, describing the correction as fixing a labeling error and attaching the correct name to the correct tree, not as a reason to abandon years of accumulated genetic engineering work. According to the program’s 2024 annual progress report, researchers continued to describe Darling trees as showing blight resistance better than any other American chestnut they had studied, even accounting for the identity correction. The team reported hand-pollinating more than 10,000 flowers across over 100 genetically diverse mother trees that year, producing more than 11,000 seeds for continued breeding and selection work.
The regulatory petitions themselves remain formally pending before federal agencies. Researchers have acknowledged that the timing of this kind of unprecedented review, the first genetically modified forest tree proposed for release into the wild, is genuinely uncertain, and that regulators have not yet identified any safety concerns that would justify rejecting the petition outright.
It is worth noting that the American Chestnut Foundation’s broader restoration strategy was never built entirely around a single genetically modified strain. The foundation has continued to describe its overall approach as a deliberately diverse portfolio, maintaining its traditional backcross breeding program independently of the Darling controversy, and continuing to work with a wide network of university partners, state chapters, and individual volunteer growers across the tree’s historical range. This structural decision, distributing the risks of the restoration effort across multiple independent programs rather than betting the entire project on a single genetic solution, is what allowed the broader movement to absorb the Darling 58 setback without collapsing. Traditional backcross trees, chinkapin hybrids, and the corrected Darling 54 genetic engineering program are all still developing simultaneously today. Each represents a different bet on the best way to eventually return a functionally reproducing, blight-tolerant American chestnut population to the eastern forest canopy.
Unresolved concerns remain. No genetically modified forest tree has ever been approved for unrestricted use in wild U.S. ecosystems. The regulatory petitions for the Darling program remain pending without guarantees of approval or a clear timeline. The 2023 misclassification raised legitimate, ongoing questions about data integrity and oversight that extend beyond a correctable clerical error. Some Indigenous communities and independent researchers have separately raised concerns about the environmental risks and cultural implications of deliberately releasing a genetically modified organism into wild forest ecosystems, concerns that remain under active discussion. Even the most established backcross breeding approach has not yet produced trees with resistance reliably comparable to the fully resistant Chinese chestnut parent species used to introduce that resistance. Every current restoration path, whether genetically modified or traditionally hybridized, remains a genuinely promising but not yet fully proven solution to the underlying problem.
The honest state of this story is serious, ongoing work rather than a final victory. What is notable is that bringing a keystone species back from the edge of extinction is not requiring a single triumph but rather four decades of patient, sometimes unglamorous, deeply collaborative work among traditional plant breeders, molecular geneticists, federal regulators, and volunteer growers. The Darling 58 error could easily have become a quiet reason to abandon genetic engineering as a restoration strategy entirely. Instead, the researchers involved chose the harder and more responsible path: correct the public records, acknowledge what happened, and continue the underlying work under its accurate name. That kind of public correction is rare and valuable, and it is exactly what conservation efforts of this scale need.
Species rarely get a second chance at their own history. The fact that the American chestnut is in a position to have that chance at all, root systems stubbornly growing for a hundred years waiting for a version of itself able to survive, is itself a small marvel of persistence that no amount of scientific ambition alone could have created. The effort has required many different kinds of people working in parallel over four decades: molecular biologists sequencing a single gene from wheat, volunteer growers in dozens of state chapters hand-pollinating flowers on trees in their backyards, federal regulators reviewing an unprecedented category without an established guide, and a foundation willing to publicly withdraw support from a partner project rather than quietly continue funding work it no longer fully trusted. No single part was sufficient on its own. It took all of them working imperfectly, sometimes at cross-purposes, to keep a century-old recovery effort alive enough to reach where it is today.
A century ago, the American chestnut was one of the most abundant and useful trees in the eastern United States, feeding wildlife and rural economies alike, before a fungus that entered the country accidentally reduced billions of mature trees to root systems and struggling sprouts in about forty years. For a hundred years since, those root systems have kept trying, sprouting again and again, never managing to grow tall enough to reproduce. Today, an ambitious and sometimes imperfect combination of traditional breeding and genetic engineering is working patiently in parallel across several different paths to give those persistent roots a version of themselves able to survive long enough to finish the job. One of those paths stumbled publicly in 2023, and the researchers involved chose to correct the record rather than hide it and continue working anyway.
When you hear that America’s legendary tree has finally returned, the real story is not a simple triumph. It is a century of unglamorous persistence underground, met finally by sustained, humble human effort to give it the one thing it has been quietly waiting for all that time: a fair chance to grow up. No one involved in this effort is promising a full forest of mature American chestnuts next year or even in the next decade. What they are promising, and what four decades of continuous, sometimes embarrassing, unfinished work suggests, is that there are people still showing up year after year, trying on behalf of a tree that has never stopped trying itself.
