91����

Photosynthesis is the oldest carbon-capture technology on Earth. For eons, plants have pulled carbon dioxide from the atmosphere and locked carbon — the building block of life on our planet — into their bodies and roots. 

In young forests, the widespread consensus is that this process rapidly pulls, or sequesters, carbon from the atmosphere. As forests mature, more trees start to die, releasing the carbon they captured in their wake. Carbon sequestration, the thinking goes, slowly stalls and old forests eventually release roughly as much carbon into the air as they capture.

Thirty years of measurements taken by 91���� scientists at a remote 550-acre forest challenge this idea.

At Howland Research Forest, located about 30 miles north of Orono, Maine, in the towns of Edinburg and Howland, 98-foot towers rise above the spruce and hemlock canopy. They are topped by instruments that measure carbon dioxide flux — the exchange of the gas between the forest canopy and the atmosphere. The measurements are so precise that they can detect the breath of a technician working nearby.

What they have recorded is a carbon record of exceptional length drawn from a mature, undisturbed forest. The data generated here is reshaping how the world understands forests and their influence on climate. The towers on Howland have been collecting data since 1996, making them among the longest-running records of their kind in the United States, second only to Harvard Forest. 

These findings are used by scientists, educators and land managers worldwide, informing forest management, timber production, carbon budgets, conservation and policy. But the future of this research is uncertain. For decades, the Howland towers were supported through the federally funded AmeriFlux network. As research priorities shifted, that support was interrupted, putting the long-running record at risk. 

A recent $175,000 private gift to the 91���� Foundation — equal to the site’s annual operating cost— has temporarily filled that gap, keeping the research running through next year. Without it, the towers would have gone offline this August, bringing 30 years of continuous monitoring to a standstill. 

While the gift provides a one-year lifeline, it does not solve the underlying challenge. Sustaining this irreplaceable observatory and training ground for 91���� students aspiring to become foresters, conservationists and researchers requires .

A living benchmark

Mature, unmanaged forests are exceptionally rare in New England. Centuries of timber harvest have reset the ecological clock across nearly every landscape. 

At Howland, the forest canopy is dominated by trees between 100 and 200 years old. Some are even older. They had already stood for centuries when Henry David Thoreau passed through central Maine on his way to Mount Katahdin in the 1840s.

A yellow birch documented in the forest’s permanent research plots dates to the mid-1600s — at least 367 years, the oldest of its species on record in Maine. Cedar and hemlock that sprouted in the 1700s stand alongside fallen logs in every stage of decay. Their slow decomposition is a critical part of the carbon cycle that researchers here are still working to fully understand.

Coarse woody debris — the fallen logs and standing dead trees so characteristic of old forests — are largely absent from managed landscapes. At Howland, it is everywhere. 

The forest is surrounded by active timber operations, which makes it something else: a control site. Without Howland as a baseline for what an unmanaged forest looks like, the comparisons that inform forest management across the Northeast would be far less meaningful.

The Northeast Wilderness Trust recognized that value in 2007, when it purchased the 550-acre site and permanently protected it as forever-wild. The protection came at a critical moment: the previous owner had considered resuming active management, which would have disrupted decades of research.

“We permanently protected Howland because ancient forests are rare and have so much to teach us,” said Shelby Perry, the Northeast Wilderness Trust’s wildlands ecology director. “Howland shows us the unique value of wild places amid managed landscapes. What researchers learn here informs forest policy at a global scale. Keeping this land wild and keeping the research going are inseparable.” 

For these findings, time is priceless

When a carbon flux tower was installed at Howland in 1996, it became a founding site of the AmeriFlux network, which has spread to more than 500 monitoring stations across the Americas. 

“We measure temperature, wind speed and direction, and the air’s carbon dioxide, water and methane concentration every tenth of a  second. That’s over 315 million data series a year,” says Roel Ruzol, 91���� research associate and Howland Forest’s site manager. 

Ruzol keeps the towers running, coordinates access for researchers and works with Fraver to eliminate what he calls “noise,” or anomalies in the data created by rain, snow, power fluctuations or a bird perching near the equipment. They then convert these measurements into precise half-hour flux rates, that is, how much carbon the forest absorbs and releases. The findings are online for anyone to access.

The Howland record is now 30 years old. The forest has stored, on average, nearly 3.5 tons of carbon dioxide per acre per year. That was not unexpected, but the long-term trend was surprising, even to the researchers who built the flux network. The rate of carbon uptake is increasing over time. This old forest has not plateaued; it’s accelerating sequestration.

“If you considered just the first 10 years, you would see a slight decrease and might conclude the forest was declining as a carbon sink,” said Shawn Fraver, associate professor of forest ecology at the 91����, who has conducted research at Howland since 2015. “However, when viewed over 30 years, we actually see an increase. That long-term view completely changes the story.”

The trend held through climate extremes. The 30-year record spans the warmest, wettest and driest years in the past 125 for Maine. Scientists do not yet know why, but the findings, which were in the Journal of Geophysical Research: Biogeosciences, provide the basis for experimental research that can provide actionable insights. 

Another study co-led by Fraver, whose research program is partially supported by the Maine Agricultural and Forest Experiment Station and the National Science Foundation, aims to identify the mechanisms by which forests switch from methane sources to sinks and back. Previous research showed that forest soils emit methane in wet conditions. Waterlogged soils favor the microbes that generate it. But those studies largely focused on landscapes that are known methane sources, ignoring sites like Howland that, depending on soil moisture, alternate between being a source and a sink. 

The documented source-sink switch at Howland has led to a National Science Foundation-funded project, now entering its final year, which attracted researchers from institutions across the country — San Diego State University, North Carolina State University, Arizona State University, and the Woodwell Climate Research Center in Massachusetts, as well as the 91����. The findings are expected to meaningfully revise how methane is accounted for in forest carbon budgets.

The data from Howland have been downloaded more than 16,000 times since being made publicly available in 2007. Researchers around the world have used Howland and the data gathered there to build climate models, calibrate satellite data, advise policymakers and teach the next generation of forest scientists.

A living lab for forest management

Howland’s scientific value extends beyond the conservation land itself. Three flux towers now operate across the area: two on Northeast Wilderness Trust land and a third on property managed by American Forest Management (AFM), a forestry consulting company that has been managing harvest activity nearby in recent years. “Not every forest company would be willing to cooperate with us on this. It’s a huge benefit to us as researchers,” Fraver said. 

The AFM tower, established in 2012, has been tracking carbon dioxide fluxes continuously through a period of active shelterwood harvest. With the Howland towers providing an unmanaged comparison site, researchers have a rare before-and-after opportunity that almost never presents itself in landscape-scale ecology.

A newly funded $243,000 study from the will extract maximum value from that natural experiment. Combining forest inventories, LiDAR-derived biomass estimates, ground and tower-level flux measurements and carbon modeling will give managers the most comprehensive picture ever assembled of how partial harvests affect the forest carbon budget.

“American Forest Management has always believed that good forestry and good science go hand-in-hand. Hosting the tower on our client’s land is one of the most tangible ways we can contribute to that.” said Jeremy Miller, Region Technical Manager at AFM and alum of 91����’s School of Forest Resources. “The NSRC study will give us an idea of how partial harvesting affects carbon dynamics in the short and long term.”

The results will give forest managers who want to incorporate carbon objectives into their planning, alongside timber production, a quantitative foundation to build on.

A proving ground for people and technology

Howland also serves as a key research site for 91���� graduate students, along with undergraduate field technicians who spend summers gaining hands-on experience with advanced instrumentation and long-term ecological research methods. 

Chris Hettwer, who earned his master’s degree from the School of Forest Resources in 2025, said the opportunity to work at Howland helped draw him to Fraver’s lab. 

“Working at Howland was an instrumental experience that fundamentally shaped me as a scientist,” says Hettwer, who is now pursuing a Ph.D. at the University of Chicago. “In addition to learning the technical field skills for researching ecosystem processes, I gained a deeper understanding of how scientific research is conducted, communicated, and built cumulatively over time. Much of what I do now is grounded in what I learned at Howland.”

A network of approximately 150 permanent research plots distributed around the flux towers provides a training ground unlike any classroom. 

So does a 7.4-acre research plot, established by NASA in 1989 to ground-truth satellite instruments. This project led the site to be photographed from space more than anywhere else on the planet at the time. Every tree above 10 centimeters in diameter was mapped and measured. In 2015 and 2025, Fraver and his research team repeated the inventory and took core samples from 10% of the trees to track growth trends. 

That plot revealed the 367-year-old yellow birch. It also told the story of a forest shaped by centuries of disturbance: spruce budworm outbreaks, wind storms, selective harvests in the 1800s and the slow return of a mature forest carpeted with deadwood and moss. 

Howland is where emerging scientists and technology learn to read the forest.

One more year

Maintaining Howland’s flux towers, sensors and core infrastructure costs approximately $175,000 per year. That annual cost was historically covered through the federally-funded AmeriFlux network, but shifting research priorities have left the site without consistent support.

“Without this gift, the AmeriFlux tower would have shut down this August,” Fraver said. “We have funds to continue another year. But our latest research, the student experience, the experimental work on managed versus unmanaged forests — all of it depends on the towers being operational. If the tower goes offline, we lose the context this monitoring data provides, and everything that has been built on it.”

The recent private gift ensures operations through next year, but it does not establish a long-term solution. Keeping one of the longest carbon dioxide flux records in the Americas running, and preserving decades of continuous data, requires sustained annual investment.

“Science is a continual and ever-building process. The continuity of monitoring at Howland is a big part of what makes this data so powerful,” Hettwer said.  

The science at Howland is answering questions that matter in Maine and globally — how much carbon do mature forests sequester, how harvest practices affect forest-atmosphere interactions, and whether forest soils remove methane. These are not strictly academic questions. The findings help forest managers and policymakers make data-driven decisions in a changing world.

“The questions we are asking now about carbon, methane and how managed and unmanaged forests compare will offer critical insights for forest management in the coming decades,” Fraver said. “This forest is capable of providing those answers. It just needs the towers to keep running.”

Those interested in supporting monitoring at the Howland Research Forest may donate or contact Elizabeth Erickson,  senior director of philanthropy at the 91���� Foundation, at elizabeth.erickson@maine.edu or 207.581.1145. 

Contact: Erin Miller, erin.miller@maine.edu