Necks arched backward, John O’Keefe and Greta VanScoy eye the upper branches of a black gum tree in the middle of Massachusetts. All around them, acorns pinball to the forest floor as leaves rustle in the mid-September wind. This susurration invites the kind of abstract thought that so often accompanies walks in the woods. But the minds of O’Keefe and VanScoy remain grounded in the concrete—in numbers—as they look skyward at the canopy.

“It’s over 50,” O’Keefe observes.

“It is over 50,” VanScoy says. “I don’t know if it’s 70.”

“I was thinking 65.”

“Sixty-five is perfect.”

Sixty-five percent, that is—their estimate of the leaf coloration in this tree, a shock of red-orange in an area that, on just the second day of fall, is still mostly green.

VanScoy jots down the number on a sheet attached to an orange clipboard. Next to it, as she will for the dozens of other trees on this grid, she notes their observation for the tree’s leaf fall, too—they settle on 7 percent—before she presses on with O’Keefe to survey a red oak, a dominant species in this forest, and more maples, birches, gums and ashes.

For more than three decades, O’Keefe has studied the same trees along this two-and-a-half-mile loop. His study, the “Phenology of Woody Species at Harvard Forest,” has tracked leaf budburst, development, coloration and drop since the spring of 1990, relying on nothing but his own eyeballing to collect data every week during the spring and fall.

“There’s virtually no measurement,” he says. “It’s all guess.”

John O’Keefe and Greta VanScoy
John O’Keefe and Greta VanScoy assess a black gum tree in mid-September. Benjamin Cassidy

In an era of drone surveys and time-lapse camera readouts of forest growth, the shoe-leather study has endured as a rare and vital long-term look at how seasonal biological phenomena change over time. The project has helped other scientists recognize that the interdisciplinary branch of science known as phenology—or the study of recurring natural events—can illuminate the effects of climate change, inspiring a network of researchers across the country.

“The kind of things that John did caught the attention of people like me and others in the United States,” says Mark D. Schwartz, a founder of the USA National Phenology Network who’s aimed to emulate O’Keefe’s research in the Midwest.

The phenoclimatologist at the University of Wisconsin-Milwaukee acknowledges that studies by foot can’t cover as much ground, or as many species, as satellites or other forms of remote sensing. And O’Keefe says that some scientists still might dismiss his subjective study as “natural history.” But others appreciate that he’s kept the most important variables—the trees and the data collector—almost entirely static over his many years of record-keeping.

“John is an incredibly well-calibrated instrument,” says Andrew Richardson, an ecologist at Northern Arizona University who worked with O’Keefe while at Harvard University. “It’s really his skill and diligence that make the record what it is.”

But now VanScoy will do that calibrating. Earlier this year, O’Keefe handed over data collection duties to the meticulous researcher and educator at the Harvard Forest. VanScoy understands she’s replacing someone who essentially became a control for a natural experiment. During the spring and fall, she’s welcomed O’Keefe’s eye as she’s made observations, trying to see the forest as he does.

“He’s the bar,” VanScoy said early on during the phenology study walk that September afternoon.

O’Keefe, for his part, offers her reassurance about how, in time, one’s observations become more consistent, and the lessons of the “ground truth”—the physical measurements that often complement aerial photo data—begin to reveal themselves.

“I have a PhD in forest ecology, but I’ve learned so much more just by making the close observations of individual plants and walking the same trail for years,” he says. “That’s where you really learn.”

Their route begins next to a pasture in the Harvard Forest, a research haven established in 1907 by Harvard University. Just over an hour’s drive west from Cambridge in the sylvan town of Petersham, the 4,000-acre property hosts as many as 100 research projects at a time, with scientists examining everything from carbon exchange to biodiversity to ants as they draw from Wi-Fi in the woods and other infrastructural advantages. The forest is one of the most studied in the world.

Harvard Forest
A view from the Harvard Forest’s walk-up research tower shows the oak canopy in peak fall color. David R. Foster

In 1988, the same year the Harvard Forest became a Long-Term Ecological Research site, O’Keefe arrived in Petersham fresh out of a forest ecology PhD program at the University of Massachusetts Amherst. In grad school, he helped an adviser set up a phenological study of deciduous trees, and he felt compelled to develop his own to help him “get outside.” Choosing a site he could visit regularly, he started monitoring 33 species of trees and shrubs at the Harvard Forest in the spring of 1990, taking extended lunch hours (and then some) to complete the work.

Early on, newspaper reporters reached out to see when O’Keefe thought the leaves would “peak” in color during New England’s famous fall foliage season. But few asked about why that date changed every year.

“What was not really paid attention to, or of interest [to them], was the whole effect of climate change on it,” O’Keefe recalls as he and VanScoy trudge among a string of maples and birches.

That changed in the late 1990s, as media and academic interest picked up. A study in 1997 by University of Montana researchers, for instance, used his visual observations to help model the length of growing seasons in the continental U.S. amid climate variability.

After the turn of the century, as others downloaded and cited his work, O’Keefe was a co-author of studies on the presence of anthocyanins, or red pigments, in Harvard Forest trees; on the tendency of understory species to develop leaves earlier to access sunlight; and on the rise of the forest’s net carbon uptake amid warming.

O’Keefe and other researchers at Harvard also focused on the intersection of climate change and foliage forecasting. In 2014, they used his observations of eight anthocyanin-producing species to complete what one scientist at Appalachian State University deemed “the first scientific analysis of the impacts of weather on the timing and duration of fall color, at least for trees that turn red in the fall.” The study, published in PLOS One, found that the amount of autumn color—calculated by the length of time and number of leaves with red, not their brilliance—would increase for most species through 2099 under conditions projected by the Intergovernmental Panel on Climate Change. Still, the researchers identified significant variance among the species: Black cherry, red maple and white ash tended to experience peak coloration earlier, while oaks and sugar maples blazed later.

Richardson was a co-author of that study and a frequent collaborator with O’Keefe, albeit while pioneering a more modern method of phenology research. In 2008, the principal investigator of the PhenoCam network set up a digital time-lapse camera on an environmental measurement station tower in Harvard Forest, making it one of the first sites to host technology that now captures plant life cycles at more than 700 locations around the world. Each camera sends a live picture every 30 minutes from sunrise to sunset to a server at Northern Arizona University, where researchers then update greenness data they derive from the images and post the photos online. In essence, it allows researchers to monitor color change in autumn and budburst in the spring from afar.

“Phenology is kind of this Victorian-era science. It’s really focused on the natural history of organisms, and you can imagine people like Henry David Thoreau going out and looking at the leaves in their woods and making copious phenological notes,” Richardson says, “and I think with PhenoCam we kind of bring that Victorian science into the 21st century.”

Which isn’t to say he’s discounting O’Keefe’s on-the-ground work whatsoever. “When you look at the start- and end-of-season dates that we estimate from the PhenoCam images, and compare that to John’s dates,” Richardson says, “the correlation between the two is incredibly strong.”

Harvard Pond
Trees blush near the Harvard Pond in autumn. David R. Foster

On the trail, O’Keefe and VanScoy are casually fastidious. One rule, for example, is that branches without buds aren’t counted when the researchers estimate the proportion of leaf fall; only those with buds factor into their judgments, the percentages they deliberate over tree after tree after tree.

But they don’t speak in purely scientific terms. Along the way, they revel in the abundance of Concord grapes, the trickle of streams, the aroma of slimy cap mushrooms and the crunch of a good acorn season.

“The data is super interesting,” VanScoy says, “but probably one of the most interesting things about this is everything you write in the margin.”

She’s curious to see how non-native pests and pathogens will change the forest in the years to come. Hemlock woolly adelgids and emerald ash borers have already altered its landscape, she notes as they survey the final trees around the Harvard Forest’s Fisher Museum. “It’s going to look different,” VanScoy says. “I don’t think it’ll be that long before it looks different.”

O’Keefe, meanwhile, long ago hypothesized that he’d see signs of an earlier spring and later fall in the Harvard Forest over time, as has been discovered elsewhere on a warming planet. But while fall has shifted later in Petersham, spring hasn’t budged much. “To me, that’s actually more interesting and informative, because you have to ask the next question, which is, ‘Why?’ And you have to dig a little deeper.”

Or walk, as it were, a little longer. It’s just someone else’s turn now.

Get the latest Science stories in your inbox.