The daily destruction of nature’s carbon stores is happening right before our eyes, as forests are ravaged by catastrophic wildfires and vast tracts of wildlands are cleared for agriculture. But even greater stores of carbon lie hidden beneath our feet, and they too are under threat.

The world’s soils are a gigantic carbon sink that has so far played a vital, outsized role in mitigating humanity’s excessive carbon emissions. But climate change, industrialized agriculture and other human activities threaten to degrade global soil carbon storage — maybe dangerously so.

Preserving the ecosystem services of this subterranean environment is crucial to meeting global net zero commitments.

The global soil carbon sink: An uncertain future

Earth’s top 2 meters (6 feet) of soil contain an astonishing 2.5 trillion metric tons of carbon — more than is held in living vegetation and the atmosphere combined. But this is not a static store: Soils are constantly gaining and losing carbon as plants, fungi, bacteria and animals act upon them. An estimated 60 billion metric tons of carbon flow in and out of the soil every year, more than three times the scale of human emissions.

“That flux in and out of the soils can be the biggest determinant of atmospheric CO₂ levels,” says Jonathan Sanderman, a senior scientist at the Woodwell Climate Research Center in the U.S. But as the climate warms, the balance of carbon flowing in and out of soil is being disrupted, with potentially disastrous consequences.

Estimates suggest that global soils could lose 50 billion metric tons of carbon by 2050 — roughly 15% of projected human carbon emissions over the same period. If soils cease being a carbon sink and become a net carbon source, that could trigger positive feedback loops that accelerate climate change, just as we race to decarbonize our societies.

Concerningly, the potential for major soil carbon losses has largely not been factored into climate policy negotiations.

This calls into question whether emissions cuts under the Paris climate agreement will be sufficient to achieve international targets (such as limiting warming to less than 2° Celsius above pre-industrial levels) if soils become a major additional carbon source in a hotter world.

Mixed findings: Assessing soil carbon sink health

For decades, scientists have been regularly monitoring the amount of carbon uptake by ecosystems on land, known as the “land sink,” using a combination of direct measurements and computer models to assess how carbon moves through different parts of the Earth system. But these studies offer a mixed picture of land sink health.

Two recent studies reported alarming drops in the land sink’s uptake of carbon, while other analyses have found relatively constant rates of sequestration. Unfortunately, no study has yet been able to dig into the status of the global soil carbon sink directly.

However, recent research led by Yinon Bar-On, an assistant professor at the Weizmann Institute of Science in Israel, found that most carbon taken up by the land sink is being stored in non-living material. This important analysis showed that while the lank sink sequestered approximately 35 billion metric tons of carbon between 1992 and 2019, only 1 billion metric tons was stored in living vegetation. “There’s a mismatch between what we can observe from the fluxes between land and the atmosphere, and what we can observe that is building up in vegetation,” explains Bar-On.

The missing component, the study concluded, is non-living organic matter. This includes soil, leaf litter, deadwood and lake sediment, as well as human-made carbon stores, such as harvested wood products and landfills. Many of these carbon stores are not included in models of the global carbon cycle.

Frustratingly, researchers don’t fully understand the mechanisms that influence the soil carbon cycle. Without that understanding, we can’t hope to develop models that represent these processes accurately. And what we don’t know could bring unpleasant future surprises.

The ebb and flow of soil carbon

At a very basic level, we do understand that carbon is added to the soil by decomposing organic matter. Plants take up CO₂ to fuel photosynthesis, and some of that carbon is added to the soil when dead leaves, wood and roots decay. Fungi also play an important role, storing carbon in their intricate networks of branching filaments known as hyphae.

One important group known as mycorrhizal fungi form beneficial partnerships with plants, providing water and nutrients in exchange for carbon. “Symbiotic fungi can contribute to soil carbon directly through their own biomass in their underground networks … but probably their biggest role is how they interact with plants,” explains Heidi-Jayne Hawkins, a researcher at the University of Cape Town in South Africa.

Hawkins and colleagues estimate that symbiotic fungi store more than 13 billion metric tons of carbon each year — equivalent to about a third of humanity’s annual CO₂ emissions from fossil fuels.

But as all this carbon is stored, large amounts of carbon are also being lost from the soil due to the activity of microbes, which use it to fuel their growth and metabolism, releasing CO₂ in the process.

Critically, no one yet knows how much a warming world will impact carbon-storing fungi and carbon-releasing microbes.

Monitoring climate change’s complex effects on soil carbon

Laboratory experiments have shown that warmer temperatures increase activity of the enzymes that microbes use to break down soil carbon. Based on this simple relationship between temperature and microbial activity, scientists expect that in a warmer world, soil microbes will release more carbon into the atmosphere.

On the other hand, rising atmospheric CO₂ levels stimulate plant photosynthesis, meaning that plants sequester more carbon — an effect known as CO₂ fertilization. Thanks to this warming-induced higher photosynthesis rate, the land carbon sink has actually grown since the start of the industrial revolution, mopping up more and more of humanity’s emissions.

These opposing effects of elevated temperature and CO₂ fertilization on soil carbon have been factored into climate models, such as those used by the U.N.’s Intergovernmental Panel on Climate Change (IPCC). But there’s a catch: These models can’t currently capture the true complexity of the soil carbon sink, in part because scientists don’t fully understand the mechanisms that influence soil carbon gains and losses.

We know, for example, that not all soil carbon is consumed by microbes, but it’s less clear what protects long-term carbon stores, and this is an active area of research. “Over the last two decades, there’s been a big evolution in our thinking about how carbon is stabilized in the soil,” explains Sanderman.

It was once thought that, over time, soil carbon formed complex compounds that are inaccessible to microbes. But more recent research suggests it’s mainly physical barriers, such as binding to clay particles, that protects soil carbon from hungry carbon-expelling microbes.

Scientists are now working to “translate some of this new process-based knowledge into models that could give us a better idea of whether [future] soils are going to be a net [carbon] sink or source,” says Sanderman.

Questions about long-term warming responses

Another major gap in understanding is how soil ecosystems will change over the long term. Microbe species living in soils (of which there may be millions) differ in how efficiently they use carbon for growth, and this determines what proportion of the carbon they consume is released into the atmosphere.

Field experiments have found that warming can change the composition of soil microbial communities, potentially altering the balance of soil carbon gains and losses in unpredictable ways. If microbial communities shift toward more heat-tolerant species that use carbon more efficiently, for example, then soil carbon losses might stabilize over time.

Researchers need long-term field experiments to find out. But getting funding for multidecade artificial-warming experiments is challenging, so there have been relatively few such studies to date.

One exception is a soil-warming experiment at the Harvard Forest in central Massachusetts, U.S., that has been run since 1991 by Jerry Melillo from the Marine Biological Laboratory in Massachusetts, and colleagues. After 26 years, forest plots warmed by 5°C (9°F) had lost 17% of the carbon stored in the top 60 centimeters (24 inches) of soil. But that carbon wasn’t lost at a steady rate. Instead, researchers identified a cyclical pattern of carbon release associated with major reorganizations of the soil microbial community.

This underscores the complexity of soil microbe communities and the importance of long-term soil carbon studies.

“It’s really important to maintain the existing experiments and keep them running as long as possible, because they’re only going to generate more and more useful information over time,” says Andrew Nottingham, a soil scientist at the University of Leeds in the U.K.

Deep soils: An alien environment for soil microbiologists

Most soil carbon research to date has focused on the top meter of soil, so we know very little about what happens deeper. This matters because deep soils have unique characteristics that might respond differently to warming.

The chemical composition of deep soils is distinct from surface soils; they’re also often low in oxygen, richer in clay and have higher moisture content.

“It’s almost like an alien environment for microbiologists in terms of understanding the carbon dynamics,” says Nottingham. Deep soils also tend to hold much older carbon stores, which, if lost, may be more difficult to replenish.

To improve understanding of deep soil carbon dynamics, the International Soil Carbon Network has established a coalition of soil ecologists to conduct long-term studies.

Tropical soils: More sensitive to warming than once thought

There are also geographical gaps in our soil understanding. The huge carbon stores in the Arctic permafrost have been extensively studied, as have temperate soils in Europe, North America and Asia. But tropical soils have largely been overlooked.

This is partly for logistical reasons — a lack of infrastructure and funding in the Global South has made field experiments more challenging there. There’s also been a long-standing scientific misconception that assumed tropical soils would be less affected by climate change.

“There’s this old paradigm that all tropical soils are geologically old and strongly weathered, and therefore infertile and low in organic matter,” Nottingham explains. “What’s often overlooked is the fact that tropical soils can be extremely diverse, and even the weathered soils can hold a lot of carbon.”

Researchers now know that about a third of all soil carbon resides in the tropics. Tropical forests exchange more CO₂ with the atmosphere than any other land-based ecosystem, and as much as 50% of that carbon comes from soil. So, even a fractional increase in tropical soil carbon emissions could have a big impact on the global climate, requiring countries to up their net-zero commitments.

In 2016, Nottingham established the first long-term tropical soil-warming experiment, in a lowland forest in Panama. The SWELTR experiment uses underground cables to warm the top 1.2 m (4 ft) of soil by 4°C (7°F), mimicking end-of-the-century warming predicted for the tropics.

After two years, soil microbes in warmed plots showed increased enzyme activity, and soil CO₂ emissions increased by 55%. This contrasts with the results of a similar study on temperate forest soils in California, which reported just a 35% increase in soil CO₂ emissions.

Based on lab experiments and thermodynamic theory, scientists had expected warming would cause greater increases in soil carbon emissions in cooler, low-latitude soils. The SWELTR results challenged that long-held assumption. “We’re finding out that tropical soils have a much more important role in Earth’s carbon cycle and are much more sensitive to warming than we previously recognized,” Nottingham explains.

This preliminary finding could pose a problem for policymakers: Global soil carbon models used for climate projections have up to now largely been based on data from temperate soils. The SWELTR results suggest we may be dramatically underestimating the scale of short-term soil carbon losses due to climate change, though more study is needed.

Land management: An opportunity to boost soil carbon stocks

Knowing the mechanisms of tropical soil carbon loss is important, Nottingham says, not just to help predict climate change impacts, but to “improve understanding of how to better manage soils in the tropics for carbon sequestration.” Improving farming and forestry practices could offer an opportunity to boost soil carbon, not only in the tropics but also in temperate regions.

Globally, land conversion for agriculture has released 110 billion metric tons of carbon from the top layer of soil over the last 12,000 years, while existing land management practices have had a 10 times greater impact on soil carbon than climate change so far.

“There’s a whole suite of practices that land managers can implement that could start rebuilding some of that lost carbon,” says Sanderman, such as minimizing soil disturbance, planting perennial crops, or using cover crops to keep the soil surface covered with vegetation year-round. With the right management practices, degraded soils could sequester as much as a billion metric tons of additional carbon each year.

Climate-smart farming methods can store about 0.3 metric tons of carbon per hectare per year on average, but some farms achieve much better results than others. “We don’t really know why some soils seem to gain a lot more carbon than others,” says Sanderman. Working out how mineralogy, soil structure and land management practices interact to store carbon in the soil could help us tailor farming methods to maximize soil carbon sequestration.

Scientists agree that restoring damaged soils and protecting remaining wild ecosystems from degradation will be key to ensuring that soils continue being an ally in the fight against climate change. Those efforts could prevent society’s decarbonization progress from being overshadowed by catastrophic soil carbon losses.

Banner image: Climate change threatens to disrupt the soil carbon cycle, potentially releasing billions of tons of carbon into the atmosphere and overshadowing international decarbonization efforts. Image by USDA-NRCS/Catherine Ulitsky via Flickr (CC BY 2.0).

Citations:

Jackson, R. B., Lajtha, K., Crow, S. E., Hugelius, G., Kramer, M. G., & Piñeiro, G. (2017). The ecology of soil carbon: Pools, vulnerabilities, and biotic and abiotic controls. Annual Review of Ecology, Evolution, and Systematics, 48(1), 419-445. doi:10.1146/annurev-ecolsys-112414-054234

Crowther, T. W., Todd-Brown, K. E., Rowe, C. W., Wieder, W. R., Carey, J. C., Machmuller, M. B., … Bradford, M. A. (2016) Quantifying global soil carbon losses in response to warming. Nature, 540, 104–108. doi:10.1038/nature20150

Hawkins, H., Cargill, R. I., Van Nuland, M. E., Hagen, S. C., Field, K. J., Sheldrake, M., … Kiers, E. T. (2023). Mycorrhizal mycelium as a global carbon pool. Current Biology, 33(11), R560-R573. doi:10.1016/j.cub.2023.02.027

Ciais, P., Ke, P., Sitch, S., Chevallier, F., & Liu, Z. (2025). Low latency carbon budget analysis reveals a large decline of the land carbon sink in 2023 and 2024. doi:10.5194/egusphere-egu25-51

Friedlingstein, P., O’sullivan, M., Jones, M. W., Andrew, R. M., Hauck, J., Landschützer, P., … & Zeng, J. (2024). Global Carbon Budget 2024. Earth System Science Data Discussions, 2024, 1-133. doi:10.5194/essd-17-965-2025

Pan, Y., Birdsey, R. A., Phillips, O. L., Houghton, R. A., Fang, J., Kauppi, P. E., … Murdiyarso, D. (2024). The enduring world forest carbon sink. Nature, 631(8021), 563-569. doi:10.1038/s41586-024-07602-x

Bar-On, Y. M., Li, X., O’Sullivan, M., Wigneron, J., Sitch, S., Ciais, P., … Fischer, W. W. (2025). Recent gains in global terrestrial carbon stocks are mostly stored in nonliving pools. Science, 387(6740), 1291-1295. doi:10.1126/science.adk1637

Melillo, J. M., Frey, S. D., DeAngelis, K. M., Werner, W. J., Bernard, M. J., Bowles, F. P., … Grandy, A. S. (2017). Long-term pattern and magnitude of soil carbon feedback to the climate system in a warming world. Science, 358(6359), 101-105. doi:10.1126/science.aan2874

Chambers, J. Q., Tribuzy, E. S., Toledo, L. C., Crispim, B. F., Higuchi, N., Santos, J. D., … Trumbore, S. E. (2004). Respiration from a tropical forest ecosystem: Partitioning of sources and low carbon use efficiency. Ecological Applications, 14(sp4), 72-88. doi:10.1890/01-6012

Nottingham, A. T., Meir, P., Velasquez, E., & Turner, B. L. (2020). Soil carbon loss by experimental warming in a tropical forest. Nature, 584(7820), 234-237. doi:10.1038/s41586-020-2566-4

Hicks Pries, C. E., Castanha, C., Porras, R. C., & Torn, M. S. (2017). The whole-soil carbon flux in response to warming. Science, 355(6332), 1420-1423. doi:10.1126/science.aal1319

Sanderman, J., Hengl, T., & Fiske, G. J. (2017). Soil carbon debt of 12,000 years of human land use. Proceedings of the National Academy of Sciences, 114(36), 9575-9580. doi:10.1073/pnas.1706103114

Beillouin, D., Corbeels, M., Demenois, J., Berre, D., Boyer, A., Fallot, A., … Cardinael, R. (2023). A global meta-analysis of soil organic carbon in the Anthropocene. Nature Communications, 14(1). doi:10.1038/s41467-023-39338-z

Paustian, K., Larson, E., Kent, J., Marx, E., & Swan, A. (2019). Soil C sequestration as a biological negative emission strategy. Frontiers in Climate, 1. doi:10.3389/fclim.2019.00008

Tropical forest roots show strain as changes aboveground filter below

FEEDBACK: Use this form to send a message to the author of this post. If you want to post a public comment, you can do that at the bottom of the page.