Ecosystem-scale methane fluxes in terrestrial uplands diverge from soil-sink expectations

Methane is a potent greenhouse gas, trapping heat in the atmosphere far more effectively than carbon dioxide over a century. For decades, scientists have tracked where this gas comes from and where it disappears. We know that wetlands and human activities like farming release vast amounts of it. To balance the global ledger, researchers have long relied on a specific, quiet assumption about dry land: that the soils of forests, grasslands, and shrublands act as a consistent vacuum, sucking methane out of the air and destroying it through natural bacterial processes. This "soil sink" has been treated as a fixed, reliable fact in climate models, estimated to remove a massive amount of methane every year. If this assumption is wrong, however, our entire understanding of how the planet regulates its temperature could be off balance.

A new study challenges this long-held view by looking at the land not through small, isolated samples, but through the eyes of the entire ecosystem. Researchers led by Sparkle Malone at Yale University, working with data from forty-six monitoring towers across North America, measured the actual exchange of methane between the ground and the sky over several years. Instead of finding the steady, silent cleanup crew they expected, they discovered that most of these upland landscapes are actually leaking methane, albeit slowly. The study suggests that the traditional view of dry land as a net sink is a mistake born from looking at the wrong scale, and that the Earth's natural ability to absorb methane may be far weaker than previously thought.

For years, the belief that dry soils consume methane was built on measurements taken with small chambers placed directly on the ground. These devices cover a tiny patch of dirt, sealing it off to measure how much gas the soil bacteria eat. This method has been the standard for decades, providing a clear, consistent picture of soil biology. But a chamber only sees what happens right at its feet. It cannot see what is happening in the tree roots, the dead wood, or the tiny, hidden pockets of water within the soil that might be producing gas. It also misses the movement of air through the canopy of a forest or the subtle shifts in moisture that happen across a whole field. The researchers suspected that by focusing only on the soil surface, scientists were missing the bigger picture of how these ecosystems actually behave.

To get a truer view, the team turned to the National Ecological Observatory Network, a system of tall towers scattered across the United States. These towers are equipped with sensors that measure air quality at different heights, from the forest floor up into the tree canopy. By comparing the concentration of methane at the top of a tower with the concentration lower down, the researchers could calculate the total flow of gas moving through the entire ecosystem, including the trees, the soil, and the air in between. They analyzed data from 2021 to 2024, covering forty-six sites that represented a wide variety of environments, from tropical forests in Puerto Rico to shrublands in Alaska.

The results were striking. When the researchers looked at the annual balance of methane for these sites, eighty-three percent of them were not acting as sinks. Instead, they were net sources, releasing more methane into the atmosphere than they absorbed. The amount released was small, ranging from a tiny fraction of a gram to just over one gram of carbon per square meter each year, but the direction of the flow was the opposite of what the models predicted. Only seventeen percent of the sites functioned as the expected sinks, and even those were much weaker than the global estimates suggested. The strongest sink they found was in a tropical forest in Puerto Rico, but even there, the amount of methane removed was less than half of what the global budget assumes for soil alone.

The study argues that the discrepancy arises because the soil chamber and the ecosystem tower are measuring two different things. The chamber measures the soil's potential to eat methane, while the tower measures the ecosystem's actual behavior. In the real world, factors that chambers miss—such as plants transporting gas from deep underground, tiny wet spots in the soil where bacteria produce gas, or the movement of gas through tree trunks—can turn a soil that is theoretically a sink into an ecosystem that is a source. The researchers found that even in dry, upland areas, these hidden processes are enough to tip the balance.

This finding does not mean that soil bacteria have stopped eating methane. The study confirms that the soil sink exists, but it shows that it is not the whole story. The ecosystem as a whole is doing something different. When the researchers tried to scale up their findings to the entire globe, the results suggested that upland terrestrial ecosystems might be a net source of methane to the atmosphere, rather than a sink. Depending on how the data was calculated, the global land area could be releasing anywhere from twenty-two to seventy-five teragrams of methane per year. This is a massive shift from the current global budget, which assumes these lands remove between eleven and forty-nine teragrams.

The implications for climate science are significant. If the land is not absorbing as much methane as we thought, then the planet's natural buffer against rising greenhouse gas levels is smaller than we believed. This could mean that methane stays in the atmosphere longer, warming the planet faster than current models predict. The study suggests that the way we build climate models needs to change. Instead of assuming that dry land is a fixed sink, we need to account for the complex, hidden pathways that allow methane to escape.

The researchers are careful to note that their work is based on a specific set of sites in North America, and that more data from around the world is needed to confirm if this pattern holds everywhere. They also point out that the instruments used to measure these tiny flows are incredibly sensitive, and that the signal is often near the limit of what can be detected. However, even when they tested their results against different assumptions and potential errors, the conclusion remained the same: the dominant signal across the network was one of emission, not absorption.

This study serves as a reminder that nature is often more complex than our simplest measurements suggest. By looking at the whole forest rather than just a patch of dirt, scientists have uncovered a hidden layer of methane exchange that challenges our understanding of the global climate. It suggests that the Earth's ability to clean up its own greenhouse gases may be more fragile than we hoped, and that the true balance of the atmosphere depends on processes we are only just beginning to see. The next step for science is to build models that can capture these full ecosystem dynamics, ensuring that our predictions for the future are based on the reality of the whole system, not just a piece of it.