A mycorrhizal network under a Kansas prairie and a mycorrhizal network under a Finnish spruce stand are not running the same trade.
Both are fungi living with plant roots. Both move mineral nutrients and water into plants and take plant carbon in return. But they come from fungal lineages separated by hundreds of millions of years, they enter the root in different ways, they fail for different reasons, and in the first global map of where they are richest, one of them sits inside protected areas at roughly a third the rate of the other.
This post is for land stewards, agency staff, researchers, and landscape partnerships writing plans that say "soil biology" or "fungal network." Underground, that phrase resolves into at least three different jobs. Fund the wrong one and the money does competent work on the wrong trade.
what a mycorrhizal network actually is
A mycorrhizal network is the living connection between plant roots and mycorrhizal fungi: hyphae — threads much finer than a root hair — that extend a plant's reach through soil and exchange nutrients and water for the plant's carbon. Where those hyphae colonize more than one plant, they link them. The word is Greek for fungus-root.
This is physical infrastructure, and it has now been measured. Stewart et al., in Science (June 2026), published the first global map of arbuscular mycorrhizal hyphal density: about 1.10×10¹⁷ kilometres of living hyphae in the top 15 centimetres of soil — roughly 110 quadrillion kilometres — weighing an estimated 300 ± 60 million tonnes, four to six times the mass of every living human. In ordinary soil that works out to metres of hyphae in a single cubic centimetre.
"Mycorrhizal" is a job description, not a taxonomic group. And a third major type of fungus under the same ground is not mycorrhizal at all.
three underground trades, not one
| type | who it partners with | where it dominates | what conversion does |
|---|---|---|---|
| arbuscular mycorrhizal (AM) | living roots of about 70% of land plant species, including most food crops (SPUN, A Hidden Infrastructure). Only ~350 described species in Glomeromycota; the partnership is roughly 450 million years old | grasslands, rangeland, cropland, drylands, most tropical forest. Grasslands alone hold about 40% of global AM hyphal biomass | Tillage severs hyphae. Heavy phosphorus fertiliser makes the trade unnecessary and plants stop paying for it. Fungicides hit the fungus directly. Stewart's maps predict cropland at about half the hyphal density of wild systems — a modelled association, not a verdict on any one field |
| ectomycorrhizal (EcM) | living roots of perhaps 2% of plant species — but those species are the dominant trees: pines, spruces, oaks, beeches, birches, eucalypts, dipterocarps. Roughly 60% of forest tree stems globally (Steidinger et al., Nature 2019) | temperate and boreal forest, dipterocarp forest in Southeast Asia, scattered tropical stands | Clear-cutting removes the hosts, and old-growth networks built over centuries do not reassemble on a rotation. Nitrogen deposition shifts communities toward pollution-tolerant generalists. Warming moves tree ranges faster than the fungi follow |
| saprotrophic fungi (not mycorrhizal) | dead matter, not living roots — wood, litter, dead roots | every ecosystem with dead organic material; richest where large deadwood stays on the ground | Removing deadwood and litter removes the substrate. White rot fungi are the main organisms that fully break down lignin; lose the deadwood and that lane thins |
The table is not the kingdom. Ericoid mycorrhizal fungi run their own trade in acidic, peaty ground — the reason blueberries, cranberries, and heath can live where almost nothing else will. And in 2025, a study in Communications Biology reported that dark septate endophytes, a melanized group long filed as "present, function unclear," formed connections between plants and moved water between them — functions previously credited to mycorrhizal fungi alone. One study in one system. The honest read is that the underground is probably larger than the mycorrhizal map, not that the map is wrong.
the ones people get wrong
AM fungi are not small EcM fungi. AM hyphae penetrate root cells and build arbuscules, branched exchange structures inside the cell wall. EcM fungi sheathe the root tip and grow between cells, never inside them. Different architecture, different chemistry, different failure mode.
A stocked stand is not a rich network. A forest can be cruised, mapped, and certified and still be running a thin, generalist EcM community after decades of nitrogen load. Timber inventory measures the host, not the partner.
A mushroom is not the network. Fruiting bodies are the reproductive surface of a fungus whose body is the mycelium in the soil. Most AM fungi never produce anything you can see standing up, which is part of why they went unmapped so long.
the protection gap is not one gap
Van Nuland et al., in Nature (July 2025), mapped AM and EcM richness and endemism at 1 km resolution using 2.8 billion DNA sequences from about 25,000 soil samples across 130 countries. Only about 9.5% of mycorrhizal richness hotspots fall inside protected areas. Plants and animals are roughly three times better covered.
Split by type, the gap stops being one number:
| group | share of richness hotspots inside protected areas |
|---|---|
| ectomycorrhizal (EcM) | ~13.9% |
| arbuscular mycorrhizal (AM) | ~5.1% |
| all mycorrhizal hotspots | ~9.5% |
That split is not mysterious. Protected areas were largely drawn around forest, mountain, and remote country — the geography that happens to hold EcM. AM's strongholds are grassland, rangeland, and farmed land, some of the most converted and least protected ground on Earth. Those are also the places holding most of the hyphal biomass Stewart's team measured.
Two different arguments follow from that, in two different places. Grassland is the biome carrying roughly 40% of AM hyphae with almost none of the protection — see what grasslands actually are. Forest is where the EcM trade lives and where inventory already exists for the trees but not for the partner — see what forests actually are.
Two honest caveats. A 1 km pixel is about 100 hectares: these maps are a regional screen, not a parcel survey. And the taxonomy underneath is thin — as of March 2025 the IUCN Red List had assessed roughly 1,300 fungal species, of which 411 are at risk. We are naming the trade before we have named most of the traders.
name the type before you write the plan
The type tells you what has to stay living. The programs that already work on that cover live in other guides — do not treat this table as a second grasslands or forests playbook.
| if the living network is | what has to stay living | next read |
|---|---|---|
| AM under grassland and rangeland | continuous living roots — the cover that is the grassland | what grasslands actually are |
| AM under cropland | living roots and less disturbance between cash crops | you don't have a soil problem |
| EcM under forest | host trees standing long enough for the network to persist | what forests actually are |
| saprotrophs in deadwood and litter | deadwood and litter left where they fall | the saprotroph row in the types table above |
A bird-cover easement and a fibre plan can both be good work and still say nothing about the trade underneath. Naming the type is what turns "protect soil biology" into a condition someone can keep.
frequently asked questions
what is a mycorrhizal network?
A mycorrhizal network is the living connection between plant roots and mycorrhizal fungi. Fungal threads called hyphae extend through soil, deliver mineral nutrients and water to the plant, and take carbon in return. Where the same fungal network colonizes several plants, those plants are physically connected underground.
what is the difference between arbuscular and ectomycorrhizal fungi?
Arbuscular mycorrhizal (AM) fungi grow inside root cells, partner with about 70% of land plant species, and dominate grasslands, croplands, and tropical soils. Ectomycorrhizal (EcM) fungi wrap the root tip from outside, partner with a small share of plant species that happens to include most dominant temperate and boreal trees, and dominate forest soils. AM is one ancient lineage; EcM arose independently many times across the fungal tree.
do all underground fungi do the same job?
No. Mycorrhizal fungi trade with living roots. Saprotrophic fungi decompose dead material and are the only organisms that fully break down lignin. Endophytes live inside plant tissue and buffer stress. Losing one group is not compensated by the others — they are different functions, not spare capacity.
what you can actually fund
The maps are the science layer. The Society for the Protection of Underground Networks (SPUN) is the nonprofit behind both global datasets cited here — it samples soil, sequences fungal DNA, and publishes where the networks are rich and how badly covered they are. It tells capital where. It is not the payor, and a map is not protection.
Protection is a condition on a named place, held over time. Name the type, then read the cover that hosts it: what grasslands actually are, what forests actually are, and the pillar. Explore if you want a place. Who pays is a later post.
The network is not a receipt, a credit, or a bag of inoculum — that argument has its own post at a carbon receipt is not a living fungal network. The fungi trade whether or not anyone maps them or buys anything. Naming the type is how the money finds the right underground.