Mark Harmon crouches low next to log number 219: a moss-covered western hemlock tree trunk, five meters long, lying dead on the ground in the lush green woods. It’s marked by a thin aluminum tag. The forest ecologist leans in close, his unruly white beard nearly brushing against the decomposing cylinder. Dark, flaky patches on the dull, reddish-brown wood closer to the ground show where fungi have infiltrated the cellulose within. Farther down the trunk, multicolored fungal conks protrude like hard shelves barely big enough for a mouse. A shiny black beetle scurries along the ground, then out of sight under the log. Harmon presses gently on 219 with three fingertips. It’s so spongy that he is reluctant to roll back a chunk of it to reveal what lies underneath. “Oh, I don’t want to destroy it,” he says slowly. “It’s all falling apart.”
Harmon, a longtime faculty member at Oregon State University, has been watching number 219, and more than 500 other logs nearby, decay for 40 years. He has trekked to this site in the H. J. Andrews Experimental Forest, a watershed nestled in Oregon’s western Cascade Mountains, at least 100 times. He drives more than two hours on paved and gravel roads from his home in Corvallis, Ore., then hikes in half a mile through the undergrowth, carrying tape measures, scales, saws and a computer to chronicle the relentless changes. His goal: establish an exhaustive baseline dataset that any scientist could use to test hypotheses about tree decomposition or to compare patterns of decomposition in the Pacific Northwest with those in other regions.
Decomposition can explain how and how fast carbon, captured by plants during photosynthesis, returns to the atmosphere. That process, which plays out at dizzying scales of both space and time, influences the long-term productivity and biodiversity of a forest. Harmon’s findings could influence when, or even whether, forest planners decide to remove dead logs to improve the health of the woods. Decay shapes how wildfire spreads through a timberland, too. Snags (dead but standing trunks) and downed trees also provide habitat for animals.
On supporting science journalism
If you’re enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
Before Harmon and his colleagues launched this log-decomposition experiment, scientists studying the impact of dead wood on the environment primarily looked only at what had already rotted, without understanding the variety of long-term factors that affected the decay. But by the early 1980s Harmon and other researchers realized patterns of decomposition emerged only from detailed tracking of actual logs sustained over decades, like snapshots stitched together into a multidimensional movie. Even after 40 years, Harmon says, ecologists are unearthing new questions: How does temperature affect the activity of decomposers such as brown rot fungi on various wood species? How do changing ecosystems promote or hinder interactions among invertebrates, microbes and wood? At what rate is carbon released from downed wood? This last one is of particular importance because it affects nutrient cycling through soils and roots, as well as climate change.
Mark Harmon gently pulls up a section of a tree carcass to reveal how deeply it has decomposed. For four decades he has been gathering detailed measurements of more than 500 logs at six forest sites, looking at how bugs, fungi and microbes affect decomposition rates, tissue density and sugar concentration. Wood breakdown affects a forest’s resiliency, wildfire spread and carbon dioxide emissions. Harmon’s successors are to collect data for another 160 years.
Harmon is leading the way to answers, but he may never know what they are. He designed the grand project to run for at least 200 years—well beyond his lifespan and those of his immediate successors. Ecologist Jennifer Powers of the University of Minnesota says that Harmon “really thought about long-term processes that shape forests in setting up a study he knew he would never see the end of.”
Most people regard dead trees as a nuisance, a wasted resource or something to trip over. Harmon sees revelation. When he was 21, during a run in the hilly forests of central Massachusetts, he encountered a green log that seemed to glow against the dark wooded backdrop. He had a vision that he would one day run a research effort on log decay. Granted, he wasn’t entirely clearheaded at the time. “It was helped by some substances,” he admits. “But I can still see that log.” For his first major research project, Harmon compared decomposition rates of 10 species of trees killed by fires in the Smoky Mountains. Conifer species, he found, decayed more slowly than deciduous trees, and Quercus prinus, the chestnut oak, decayed the fastest, losing 11 percent of its wood density every year.
In 1979 Jerry Franklin, at the time a forest ecologist at Oregon, visited the Smokies where Harmon, then a graduate student at the University of Tennessee, happened to be his tour guide. Forestry school had taught Franklin that a dead tree was valuable only if it were loaded onto a truck and sent to a mill. “But I came to realize that this wood I had been taught was a waste, a fire hazard and an impediment to travel had a lot of value,” Franklin says. It was still a part of the forest, still boosting biodiversity by providing habitat and returning carbon and nitrogen to the soil, he explains.
Franklin recognized a likeminded soul in Harmon. In 1980 Harmon moved to Oregon to start his doctorate, and a few years later Franklin recruited him to run an experiment simply called the Log Decomposition Project. Harmon described it as a “‘They must be crazy, but maybe not’ kind of project.” He launched the endeavor with specific but far-reaching scientific questions. He wanted to know how widely the decay rate of hardwood differed among tree species, how colonies of microbial decomposers affected that rate, and whether bugs and other invertebrates sped up the process by bringing in the microbes.
The idea of cutting down a bunch of healthy trees in one location and hauling them elsewhere was a hard sell. During one dinner at Oregon State, Harmon listened to the dean of the School of Forestry call the emerging experiment the “most stupid f–ing thing I’ve ever heard of in my life.” People regularly reminded him that he would die before it was finished or that “only an idiot” would wait 200 years for results. Over time, however, funding continued to come through.
An aluminum tag identifies a log number 219, a western hemlock. Researchers have discovered that logs of different species might take anywhere from three to 750 years to fully decompose.
In September 1985, under Harmon’s direction, crews used chainsaws, loaders, grappling hooks, cables and shovels to cut down and drag nearly 100 trees from an area kilometers across to rot at a clearing in the Andrews Experimental Forest. Each fresh log was 5.5 meters long and 45 to 65 centimeters in diameter. Since then, Harmon, nicknamed Dr. Death by the National Science Foundation, has come to this place, known as site 3, again and again to photograph the logs’ physical appearance and to catalog the succession of bugs and other invertebrates that munch on and live within the rotting wood. He uses electronic calipers and scales attached to a laptop to measure lengths, widths, weights and tissue densities, and he carries digital instruments to record the temperature, humidity and air pressure of the forest. He’s also carried thousands of “cookies”—disks several centimeters thick cut from a log’s end—back to the Andrews laboratory to tease out concentrations of sugars and track changes in carbon and nitrogen content. Every sample has been barcoded and stored. After the first collection, considered time zero, Harmon’s team filled an entire walk-in cooler with more than 1,000 cross-sectional samples. His wife, Janice Harmon, a plant ecologist, scanned the barcodes of more than 35,000 entries over time. Plastic bags filled with rotten wood samples—some powdery, some chunky—still inhabit their garage in Corvallis, waiting to be documented.
As the crew set up site 3 in 1985, it also dragged numerous other trees, similarly cut, to five additional sites scattered throughout this forest. One worker navigating a loader at the site, Harmon says, unleashed a string of expletives describing exactly what he thought about the project. In total, the crew placed 530 logs representing four species that dominate this ecosystem: Douglas firs, western hemlocks, western red cedars and Pacific silver firs. Since the beginning of the second Reagan administration, Harmon has been leading scientists, volunteers and students to these hidden glades to measure in close detail the decay of the woody carcasses.
Although weathering, solar radiation and bugs all contribute to breakdown, wood-decomposing fungi do most of the damage. These nearly invisible microorganisms often hitchhike on invertebrates’ backs to get inside the rotting wood. To catalog these populations, Harmon and his colleagues delicately scrape them off the inner walls of log cavities and into sample bags, along with whatever other tiny creatures are hanging around in there.
Today the forest is co-managed by the U.S. Forest Service, Oregon State University and the Willamette National Forest and hosts a variety of silviculture studies. It has become a long-term lab where scientists investigate the effects of disturbances such as floods and fires on a forest ecosystem. “Because we have this [70-year] baseline,” says Mark Schulze, an assistant professor at Oregon State and the forest’s director, “we can really understand these processes.” The experimental forest is the perfect place for Harmon’s work because decay plays out over scales that researchers don’t usually measure. The ongoing project is still revealing new mysteries and has created a small but energetic subfield. Scientists are now measuring tree decay in dozens of similar undertakings on six continents. Researchers in China, Germany, and other countries are probing how the climate, environment and decomposer populations in different regions interact to shape decay. They’re looking at how decay rates vary by species and location, which can shape policies around forest management and habitat protection. They’re feeding data to climate scientists, who can more precisely model the rates at which different kinds of forests may hold or release carbon.
Under some dead trees, the action of decomposers might create new soil, but under others it may leave crumbled clay or sand.
Harmon is widely regarded as the de facto pioneer of the field, having published dozens of relevant papers that have garnered thousands of citations. Hans Cornelissen, a systems ecologist at Vrije Universiteit Amsterdam who in 2012 launched Loglife, a log-decomposition project in the Netherlands that mimics the Oregon study on a smaller scale, calls Harmon the “founding father” of modern wood-decomposition science. Harmon is exacting, even obsessive, about tree decay. His focus is so deeply hardwired that he can’t ignore it. “It’s quite aggravating, actually,” he says as we move away from log 219. He seems resigned, almost exasperated. No matter where he goes or what he’s doing, even on vacation, he says with a sigh, “I’m always seeing dead trees.”
It’s quickly getting warmer on this May morning as we make our way through the ongoing experiment. Other tagged specimens in the distance look like random waves frozen on a green pond. The woods are quiet except for the occasional knock of a downy woodpecker or the distant whistle of a varied thrush. Harmon, now 72 and technically retired, easily straddles thigh-high berms and bobs under fallen conifers. He removes his hard hat and wipes his brow as we come up on a fallen western red cedar, not part of the experiment. In the study’s first couple of decades, Harmon says, the researchers found that the outer sapwood of a western red cedar decayed faster than any part of any other tree they examined. The interior heartwood, however, is the most decay-resistant, which is why it’s often used to build decks and raised-bed gardens for homes. Two extremes in one species. As a result, cedars tend to stand intact until their roots give out, and they crash down all at once—unlike, say, Douglas firs, which tend to splinter in big chunks, leaving standing snags. Trees decay differently when they’ve fallen and are within easier reach of decomposing microbes than when they remain standing.
People reminded Harmon that he would die before the research was finished or that “only an idiot” would wait 200 years for results.
Harmon puts his hard hat back on over his thinning tangle of dark hair. I’m wearing a hard hat, too, because big trees drop big limbs, although I question whether this plastic shell could protect me against a falling widow-maker. A few meters away Harmon points out log number 218, a Douglas fir. Whereas the crumbling hemlock heartwood of log 219 seemed about to implode, this prone Douglas fir was firm enough for us to stand on. The advantage of the Andrews log-decomposition study, Harmon says, is that he and his team know exactly when decay started—not the case for trees downed naturally—which helps them and other scientists more clearly understand the timeline and drivers of decay within and among different species. “We knew that was our opportunity,” he says. “Those were our initial [experimental] conditions.”
Over the past 40 years the mounting measurements have yielded unexpected insights. Deadwood might remain on a forest floor or stand upright as a snag for anywhere from three to 750 years. In a 2020 analysis, Harmon and his colleagues estimated that decay rates can vary by a whopping 244-fold across species and climates. Heliocarpus appendiculatus, a tropical tree better known as a jonote, loses nearly 98 percent of its mass a year, whereas Eucalyptus camaldulensis, the river red gum tree, endemic to Australia, loses only about 0.4 percent a year. Rates can vary within species, too. “You could have parts of trees that could last less than a decade and others up to 1,000 years,” Harmon says.
Another surprise is how drastically deadwood can alter the forest floor. Fallen trees don’t simply rot. Harmon rolls sturdy log 218 away from us to reveal a patch of mineral soil the color of the darkest chocolate. It’s made up mostly of crumbled clay, rocks and sand, as opposed to organic soil, which contains decaying organic matter such as that from trees and leaves. Fungal tendrils twist through the dark brown mat.
Wildfire in August 2023 almost ruined the 40-year decomposition experiment, consuming three of the six log sites. Bright-orange fire moss has quickly colonized some of the burned landscape.
“This forest floor has kind of melted away,” Harmon says. Organic soil digested by fungi or nematodes or bacteria under the log hadn’t been replenished. Yet leaves and branches falling on the log had accumulated and decayed over decades, producing a fertile organic soil on top of the log, where moss and other plants were now growing. “The log has basically elevated the forest floor 50 centimeters off the ground,” Harmon says. Fallen trees shift the chemistry of the soil below and above and, with that, the population of microbes in the environment.
Harmon’s group found that the soil changes the tree, too, as ants and other insects ferry dirt and microbes into the decaying log. Whether a dead tree touches mineral soil, stands as a snag or remains suspended over the forest floor after falling against a living tree can dramatically influence the concentration of carbon it stores. And the mix of the many decay factors influences the likelihood that new trees will take hold in that ground, reshaping the habitat of a forest, which in turn affects the overall health of the region.
A forest dominated by slow-rotting species can hold enormous stores of carbon for decades or centuries, whereas quickly decaying species can release lots of carbon into the air. Extrapolated to a global level, sequestration and emission can significantly affect amounts of atmospheric carbon dioxide and therefore influence climate change.
Knowing these rates is particularly important to climate change modelers, says Jonathan Schilling, whose lab at the University of Minnesota focuses on decomposition and fungi. He has run wood-decomposition experiments in Alaska and New Zealand, among other places. “We’ve got logs rotting all over the place,” he says. In 2024 he and his colleagues compared the decomposition preferences of white rot—fungi that break apart the tough lignin in trees and thus release the carbon dioxide—with those of brown rot, which head for the cellulose, leaving the lignin behind.
“There’s a lot more carbon left behind in the soil for the brown rot mechanism,” Schilling says. That matters because white rot fungi, which prefer warmer forests, are encroaching on northern regions because of changing temperatures and rainfall. The result? More carbon dioxide gets pumped into the air. “There’s a lot of carbon at stake,” Schilling says, “and enough uncertainty that we need to know how that process works.”
The Andrews experiment has inspired many others around the world. In the 2000s Powers launched the first tropical decomposition study, which involved 14 countries, with Harmon’s work as a model. In 2012 Cornelissen and his crew in the Netherlands arranged logs of 25 species in two “tree cemeteries” for his Loglife experiment. Cornelissen has also collaborated on decomposition projects in Romania, Germany and China. In 2024 he worked with Amy Zanne, an ecologist at the Cary Institute of Ecosystem Studies in Millbrook, N.Y., on a review of wood-decomposition studies that explains varying decomposition patterns around the world. Zanne sees a hidden wonderland in decomposition, populated by overlooked, disregarded players that nonetheless have critical roles in an evolving ecosystem. “I love thinking about the underdogs, the underseen things, and how hidden things make the world go round,” she says.
Harmon almost lost the entire Andrews project on August 5, 2023, when lightning struck a tree on Lookout Ridge. Fire spread quickly, and within a few weeks it had incinerated 70 percent of the watershed, nearly 10,500 hectares. It burned through three of the six log-decomposition sites, stripping living trees of leaves and incinerating much of the deadwood, which was a blow. After Harmon and I hike back to the gravel road near site 3, we drive farther up the ridge, get out and walk through the ghostly remains of site 6.
Nearly two years on, this site retains a faint, mephitic whiff of smoke and char. The fire felled giant firs and sculpted cedars, and the burned boles still stand in scorched, abstract shapes. Remaining branches, leafless and thin, glow silver against the dark snags in the sharp afternoon sun. They’ll fall eventually. Harmon squats and cradles a singed aluminum tag, barely readable, identifying a round, blackened wood skeleton as a western hemlock. There are no scurrying beetles in sight.
The 2023 Lookout Fire left hundreds of snags—dead trunks that remain standing. Dead Douglas firs tend to splinter and fall in pieces; cedars tend to stand until their roots give out and then crash down in an instant. Ongoing study of scorched logs should uncover how fire ash and charcoal alter decomposing wood.
Fire changes the game, Harmon says. It can be tragic. Yet it is also an opportunity to see a forest in a new light. “It’s changed, but, you know, it’s going to come back. It’s going to be another manifestation of the same thing,” Harmon says. Unlike in site 3, the organic soil here has all but vanished under ash and charcoal. It’s unclear which of the four species in the log-decomposition project may proliferate most in regions devastated by fire. And even though fire kills trees, it doesn’t remove them. Harmon points to a snag, maybe 30 feet tall, with tiny mushrooms protruding from cracks. He notes a little patch of uncovered mineral soil where seedlings have emerged. New trees will grow with the legacy of snags and downed trees around them, and the new forest may be even more structurally interesting.
The future, of course, is uncertain. The log-decomposition project is one of 27 in the Long-Term Ecological Research network, a collection of large-scale experiments funded by the National Science Foundation probing everything from how expanding cities affect tree-growth rates to how disturbances such as extreme wind, fire and flood shake up an ecosystem. Funding has always been an issue, and Harmon, Franklin, and others worry that recent widespread cuts to federal grants may reach the Andrews experiment. Its timescale might save it. The project requires little maintenance, and the logs will rot whether anyone is watching them or not. For now, someone is.
Harmon retired in 2016 but can’t stay away; he is still churning through enormous datasets to publish papers. He has turned the reins over to two younger researchers at Oregon State: Georgia Seyfried, a soil scientist who studies biogeochemical processes, and Jacob Bukoski, an ecologist who focuses on carbon cycling and climate change mitigation. “I think there’s a real opportunity here,” says Bukoski, who looks forward to working fire into the increasingly complex, emerging view of decomposition.
On our way back to the forest headquarters, Harmon and I cross a narrow stream flanked with hemlocks and pull over. We walk about 45 meters into the forest—over dead logs, under dead logs—and arrive at one of Harmon’s sacred spots, a grove of giant Douglas firs. Their trunks stretch at least two meters across. Their lowest branches are higher than most of the other surrounding treetops. These behemoths are older than the Mona Lisa. We stand silently in the shadow of the living giants for a few moments. With their thick, deep furrows and invisible crowns, they seem invincible and infinite. But that’s an illusion, Harmon says. This scene is a snapshot. After the trees fall—unless they burn—they’ll probably remain intact on the forest floor for another few hundred years, housing bugs, remodeling the forest and eventually sinking softly into the contours of the new woodland, in the shadow of new giants.
Scientists used to assume that decomposition was instantaneous, Harmon says—that when a tree dies, it essentially disappears. “But that’s not true anywhere on Earth, and it’s never been true,” he says. A dead tree is “just a transition to something else.”
