Scientists have produced the first global maps of the density and biomass of arbuscular mycorrhizal fungi, revealing an underground system estimated to contain 110 quadrillion kilometers of microscopic fungal threads.
Published at 1:34 p.m. EDT
The largest living network most people will never see is woven through the soil beneath forests, grasslands, farms and backyards around the world.
Scientists have now mapped that hidden fungal infrastructure on a global scale for the first time, estimating where it is most densely concentrated, how much carbon it contains and which landscapes may be placing it at greatest risk.
The international research team focused on arbuscular mycorrhizal fungi, commonly known as AM fungi. These microscopic organisms form partnerships with the roots of approximately 70 percent of plant species, supplying plants with water and nutrients while receiving carbon produced through photosynthesis.
Their threadlike cells, called hyphae, branch through the soil and dramatically expand the area plants can use to find essential resources. When combined across Earth’s topsoil, those fungal threads have an estimated total length of approximately 110 quadrillion kilometers.
That is nearly one billion times the distance between Earth and the sun.
The researchers also estimated that the networks contain approximately 300 megatons of carbon, equivalent to roughly four to six times the combined mass of every living human.
The peer-reviewed study, titled “Global density and biomass of arbuscular mycorrhizal fungal networks,” was published June 11 in the journal Science. It was led by Justin Stewart of the Society for the Protection of Underground Networks, known as SPUN, in collaboration with researchers from institutions including AMOLF in the Netherlands and the University of Sheffield. Science
The study does not mean scientists located a single, uninterrupted organism encircling the planet. The map represents modeled estimates of the combined density and biomass of countless AM fungal networks living in soils across the world.
That distinction matters. “Global fungal network” is a useful description of the planetary infrastructure created by these fungi, but the map is not evidence that every tree and plant on Earth is connected to one continuous underground web.
What scientists have created is something more precise: a kilometer-by-kilometer estimate of where AM fungal infrastructure is concentrated and where it has been diminished.
Building a map of an invisible world
Mapping microscopic fungi presented a challenge unlike charting a forest, river or mountain range.
Fungal hyphae are often thinner than a human hair and can be difficult to distinguish from other microscopic structures in soil. They also grow, retract and reorganize in response to plants, water, temperature and nutrients.
The researchers assembled fungal-density data from more than 16,000 soil cores described in 322 previous studies. Those samples represented environments ranging from tundra and deserts to tropical forests, temperate grasslands and agricultural fields.
The team then developed machine-learning models that combined the soil observations with environmental information to estimate AM fungal network density in regions without direct measurements.
To improve those estimates, scientists at AMOLF used robotic imaging technology to examine more than 300,000 living fungal hyphae grown under laboratory conditions. The imaging allowed the researchers to calibrate the relationship between the visible fungal structures and their underlying biomass.
The completed Mycorrhizal Infrastructure Map provides estimates for individual areas of approximately one square kilometer across terrestrial Earth, excluding ice caps and places where the available information was insufficient for a reliable prediction. EurekAlert!
The models predicted an average fungal hyphal density of approximately 4.4 meters per cubic centimeter of soil in the upper 15 centimeters of the ground. In other words, a very small volume of healthy soil can contain a surprisingly long system of living fungal filaments.
Lead author Justin Stewart said in a publicly released statement that a teaspoon of soil could contain as much as 10 meters, or about 32 feet, of mycorrhizal network. EurekAlert!
What these fungi do for plants
AM fungi are not simply decomposers or mushrooms growing near plant roots.
They enter into an exchange with plants. The plant provides sugars and other carbon compounds created through photosynthesis. The fungus uses its hyphae to collect phosphorus, water and additional nutrients from soil spaces that plant roots cannot easily reach.
Healthy mycorrhizal networks may expand a plant’s effective foraging area by as much as 100 times. They can also provide more than 80 percent of the phosphorus used by some plants, according to information released with the study.
This partnership is ancient. Mycorrhizal fungi are believed to have helped early plants establish themselves on land hundreds of millions of years ago. Plants had access to sunlight and could manufacture carbon-rich compounds, but their primitive roots were inefficient at gathering nutrients. Fungi supplied those nutrients in exchange for energy.
Modern ecosystems still depend on versions of that same biological trade.
The relationship influences plant growth, soil structure, nutrient cycling and resistance to environmental stress. Hyphae bind soil particles and create pathways that can affect the movement of water. When fungal tissue dies or releases compounds into the ground, some of its carbon may remain in the soil.
This is why scientists increasingly describe mycorrhizal fungi as part of Earth’s living infrastructure rather than treating them as a minor component of the underground environment.
A system tied to the carbon cycle
The new map also contributes to the scientific effort to understand how carbon moves between plants, fungi, soil and the atmosphere.
Researchers estimate that AM fungal networks transfer the equivalent of approximately 4 billion metric tons of carbon dioxide into soil each year. That figure is equivalent to about 11 percent of annual carbon dioxide emissions caused by human activity.
It does not mean the fungi permanently cancel 11 percent of emissions.
Carbon delivered underground can remain in fungal biomass, enter other soil organisms, become chemically stabilized or eventually return to the atmosphere through respiration and decomposition. How long the carbon stays in the soil depends on local conditions, including temperature, moisture, soil chemistry, microbial activity and land management.
The study therefore should not be interpreted as evidence that fungal networks offer a simple substitute for reducing fossil fuel emissions.
Their climate importance lies in their role within the larger carbon cycle. Damaging fungal networks can interfere with the movement and retention of carbon in soils, while protecting healthy ecosystems may help preserve natural processes that have regulated the planet for millions of years.
The map gives researchers a new baseline for monitoring those processes.
Grasslands emerge as fungal strongholds
One of the most striking findings involves grasslands.
The researchers estimate that grassland ecosystems contain approximately 40 percent of the world’s AM fungal biomass. Particularly dense networks were predicted in South Sudan’s flooded grasslands, Florida’s Everglades and parts of the Tibetan Plateau. EurekAlert!
Forests often receive the greatest public attention in discussions of conservation and carbon storage. The new findings show that grasslands also contain enormous biological systems beneath the surface, even when their aboveground vegetation appears comparatively simple.
That importance is paired with vulnerability.
Grasslands are among the planet’s least protected ecosystems. They are being converted to agricultural use at a rate that researchers involved in the project say is approximately four times faster than the rate of forest conversion.
When grasslands are plowed, heavily fertilized or converted into intensively managed cropland, their fungal communities can be disrupted. The resulting loss is difficult to see because much of the affected life remains underground.
A field may continue to appear green while its hidden biological infrastructure has become substantially thinner.
Agricultural land showed lower network density
The study predicted that large-scale croplands have approximately half the AM fungal network density found in less disturbed natural ecosystems.
The comparison does not prove that every form of agriculture reduces fungal density by the same amount. Nor does it establish that a single farming practice is responsible for the difference.
Intensive tillage can physically break fungal hyphae. Some fertilizers can reduce a plant’s dependence on fungal partners, causing the plant to direct less carbon toward the relationship. Crop selection, pesticides, soil compaction, irrigation and the absence of living roots between growing seasons may also influence mycorrhizal communities.
However, the authors caution that scientists still lack enough information to determine exactly how individual practices affect fungal function across different soils and climates.
Katie Field, a professor of plant-soil processes at the University of Sheffield and a contributor to the study, said in a university statement that the reduction in cropland density appears alarming. She also stressed that researchers still know relatively little about how different environmental conditions and farming methods influence fungal health, carbon storage and nutrient cycling. The University of Sheffield
That caution strengthens rather than weakens the importance of the map. It shows where researchers should conduct direct measurements and agricultural trials instead of assuming that all fungal networks respond identically.
A different map came first
The new research builds on a related global mapping project published in Nature in July 2025.
That earlier study used nearly 25,000 geolocated soil samples containing more than 2.8 billion fungal DNA sequences from 130 countries. It mapped predicted hotspots of mycorrhizal fungal richness and endemism, meaning areas likely to contain many species or unusually rare communities.
Researchers found that less than 10 percent of predicted mycorrhizal richness hotspots fell within protected areas. The results also showed that aboveground plant diversity does not reliably identify underground fungal diversity. Nature
The two maps measure different things.
The 2025 Nature study asked which places are likely to contain the greatest diversity of mycorrhizal fungi. The 2026 Science study asked how physically dense AM fungal networks are and how much biomass they contain.
An ecosystem can have a great deal of fungal biomass without necessarily possessing the highest species diversity. Conversely, an area can contain rare fungal species without having the greatest total length of hyphae.
Effective conservation planning may need both types of information.
The map is a model, not a direct photograph
Despite its unprecedented scale, the new map contains substantial uncertainty.
Scientists did not extract and measure every fungal thread on Earth. They used soil observations, laboratory imaging, environmental data and machine learning to make predictions for unsampled regions.
Some parts of the world have been studied far more extensively than others. Europe and North America generally have denser scientific sampling coverage than remote tropical forests, politically unstable regions or isolated areas of Central Asia and Africa.
The model can identify where fungal networks are likely to be dense, but local field sampling is still required to confirm conditions on the ground.
The global estimate of 110 quadrillion kilometers should therefore be understood as a scientifically developed approximation, not a literal inventory of every living hypha.
The value of the work comes from creating a testable global baseline. Future soil surveys can improve the model, reveal changes and identify places where its predictions were wrong.
Why the map matters now
Fungi have frequently been omitted from conservation plans designed around visible animals, plants and landscapes.
Protected-area boundaries may preserve forests while overlooking rare fungal communities outside those borders. Climate policies may count tree carbon without measuring the underground organisms that receive and redistribute a substantial portion of it. Agricultural programs may track crop yields without monitoring changes in the living infrastructure supporting soil fertility.
The Mycorrhizal Infrastructure Map gives governments, researchers and land managers a new tool for changing that approach. Its underlying data have been made available for monitoring, restoration planning and further research.
The map could help scientists choose locations for new soil sampling, compare fungal conditions across farming systems and identify landscapes where development could damage exceptionally dense networks.
It also changes the scale at which people can think about fungi.
A mushroom is only the visible reproductive structure of some fungal species. The much larger organism often exists as a branching body beneath the ground. Across the planet, the combined length of those microscopic bodies is almost impossible to comprehend.
Scientists have not discovered the network for the first time. Humans have lived above it throughout our history, and plants have depended on fungal partnerships for hundreds of millions of years.
What has changed is our ability to see it.
Robotic microscopes, international soil databases and machine-learning models have turned an invisible biological system into something that can be measured, mapped and potentially protected.
The result is a new picture of Earth. Beneath the familiar map of cities, rivers, farms and forests lies another geography, made from living threads that move nutrients, support plants and influence the planet’s carbon cycle.
For the first time, that hidden world has a global address.
Reporting and sourcing disclosure
This article is based on the peer-reviewed study “Global density and biomass of arbuscular mycorrhizal fungal networks,” published in Science on June 11, 2026, DOI: 10.1126/science.adu4373. It also draws on the 2025 Nature study “Global hotspots of mycorrhizal fungal richness are poorly protected,” statements released by SPUN, and research summaries from AMOLF and the University of Sheffield.