Scientists have produced the first global map of the fungal networks threaded through Earth's topsoil, and the numbers reframe the climate ledger. The microscopic threads that partner with plant roots extend an estimated 110 quadrillion kilometers — roughly 68 quadrillion miles — beneath our feet, and they move an estimated 4 billion tons of CO2-equivalent into soils each year. That flow is equivalent to roughly 11% of all human-related carbon-dioxide emissions, quietly happening in the top few inches of dirt.
The research, published in Science by an international team coordinated through the Society for the Protection of Underground Networks (SPUN), reframes a question most climate conversations have ignored: what happens to global carbon cycling if the living infrastructure that moves it gets plowed under?
The conventional climate story leaves out the soil
Public conversation about carbon tends to stay above ground. Forests, emissions, oceans, ice. The soil itself usually shows up as a passive substrate — something plants grow in, not something doing work. The new mapping data argues the opposite.
The fungal biomass involved is not trivial. Researchers estimate these networks hold around 300 megatons of carbon — four to six times the combined mass of every human being alive. And the partnership is ancient. About 70% of plant species on Earth depend on arbuscular mycorrhizal fungi to exchange nutrients for sugars, a relationship that predates trees.
The conventional wisdom has been that protecting forests is the highest-leverage move for land-based carbon strategy. The new findings complicate that picture without contradicting it. Forests matter. So does what's underneath grasslands — and grasslands have been treated as agricultural raw material for centuries.
How researchers mapped the unseen
The dataset behind these maps is unusually large for soil science. The team compiled measurements from more than 16,000 soil cores collected worldwide and analyzed over 300,000 living fungal hyphae under laboratory conditions. Machine-learning models extrapolated those measurements to produce global predictions of where networks are dense, where they're sparse, and where they're most at risk.
Dr. Justin Stewart of SPUN put the scale in tactile terms, noting that a single teaspoon of soil can contain up to 10 meters — about 32 feet — of fungal network. The full study in Science builds on a 2025 global analysis of mycorrhizal biodiversity in Nature, which first cataloged the diversity of these underground communities at planetary scale.
The interactive Mycorrhizal Infrastructure Map released alongside the paper lets anyone scroll through the predicted densities. The visual is striking: dense bright bands trace flooded grasslands in South Sudan, the Everglades in Florida, and the Tibetan plateau — places that rarely appear in mainstream climate discussion.
Grasslands are disappearing four times faster than forests
This is where the findings get uncomfortable. Roughly 40% of Earth's arbuscular mycorrhizal infrastructure sits beneath grasslands. Those same grasslands are being converted into cropland at roughly four times the rate that forests are being cleared. They're also among the least protected ecosystems on the planet — 95% of the biodiversity hotspots identified for these fungi lie outside any protected area.
The mechanism of damage is straightforward. Tillage breaks the hyphal threads. Heavy fertilizer use makes the plant-fungus partnership less necessary, so plants stop feeding the fungi. Monoculture cropping narrows the diversity of fungal partners. The cumulative result: croplands carry about 50% lower network densities than comparable wild ecosystems, with reduced capacity to store carbon and cycle nutrients.
That's a measurable subtraction from the climate ledger, and it happens quietly. Nobody sees a hyphal network collapse the way they see a forest burn.
What this means for food systems
The instinct to read this as another indictment of farming misses the point. The food system has to feed roughly 8 billion people. The relevant question is which agricultural practices preserve underground networks and which destroy them — and whether the practices that preserve them can scale.
Some answers already exist. Reduced tillage, cover cropping, diverse rotations, and lower synthetic-fertilizer dependence all correlate with healthier mycorrhizal communities. These practices sit under the umbrella of regenerative agriculture, a label that has been stretched in marketing but points at something real in the soil-science literature. A field that's never plowed and grows multiple species in rotation will host a different fungal community than a continuously tilled corn-soy operation.
The economic friction is also real. Farmers operating on thin margins can't easily absorb a transition period, and current commodity pricing doesn't reward soil carbon. The carbon being moved underground by fungi has no market value to the person whose tractor is doing the moving — or undoing it.
Why fungi keep getting left out
Dr. Toby Kiers, SPUN's executive director, has argued that fungi have been ignored in climate and conservation for too long. The institutional reasons are mundane. Conservation funding flows toward charismatic species and visible habitats. Climate policy is built around measurable above-ground stocks of carbon. Soil microbiology requires expensive sampling and slow analysis. None of this lines up with how money or attention moves.
There's also a knowledge gap that the new mapping starts to close. Dr. Corentin Bisot, a biophysicist at the AMOLF research institute, has pointed to the role of high-resolution imaging, machine learning, and robotics in making the invisible legible. Until you can see something at planetary scale, it's hard to argue for protecting it at planetary scale.
Earlier work in Nature established the diversity baseline. The new Science paper adds the infrastructure layer — how much, where, and how degraded. Together they give policy a target it didn't have before.
What happens next
Including fungi in conservation planning would mean treating soil as habitat rather than substrate. Practically, that could look like extending protected-area designations to grasslands with high mycorrhizal density, adjusting agricultural subsidies to reward soil-network preservation, and adding underground biomass to national carbon accounting. None of those moves are simple. All of them would change which land gets plowed and which doesn't.
For the curiously conscious reader, the takeaway isn't a new product to buy or a new diet to follow. It's a reframe. The carbon math of food includes what happens to the dirt the food grew in, and the dirt is doing more work than most climate models give it credit for. Choices that support diversified, less-tilled growing systems — whether at the policy level, the grocery level, or the local-farm level — feed something that 4 billion tons of annual carbon movement suggests is worth feeding.
The mapping is a beginning, not a conclusion. The researchers have made the data openly available. Whether anyone with land-use authority decides to act on it is the next question.
