If you have ever knelt in a Douglas-fir forest in the Pacific Northwest and peeled back the duff to expose the roots beneath, you have probably seen them: fine white threads, thinner than a human hair, fanning out through the dark soil like a nervous system. For most of the twentieth century, those threads were treated as plumbing, a passive network for moving water and minerals. We now know they are something stranger: a living, continent-spanning marketplace where plants trade sugar for nutrients, and where the chemistry of one tree can ripple through dozens of others within hours.
In June 2026, a team led by Stewart and colleagues published the first quantitative global map of this hidden infrastructure, drawing on more than 16,000 soil cores and 322 field studies to estimate the total reach and carbon cost of mycorrhizal fungi worldwide [1]. Their headline number is difficult to grasp without a sense of scale: roughly 110 quadrillion kilometres of fungal hyphae thread through the world's soils. That is enough filament, laid end to end, to span the distance from the Earth to the Sun about three-quarters of a billion times. The same synthesis estimates that these fungi hold around 300 megatons of carbon at any given moment, a living stock roughly four to six times the mass of every human on Earth.
What Mycorrhizal Fungi Actually Do
The partnership at the heart of this story is older than roots themselves. Roughly 80% of vascular plant species swap sugars produced by photosynthesis for nutrients and water collected by fungal hyphae, a relationship that likely allowed plants to colonise land more than 400 million years ago. The fungi extend the effective reach of a root system by orders of magnitude. In the top 15 centimetres of soil, hyphae run for roughly 4.4 metres per square centimetre on average, and a single teaspoon of grassland soil can hold up to 10 metres of arbuscular mycorrhizal network.
What gets exchanged is surprisingly concrete. The plant side hands over glucose, and sometimes fatty acids and amino acids, on a regular basis. The fungal side delivers nitrogen, phosphorus, potassium, and water, plus trace minerals like zinc and copper. In pot experiments, plants connected to richer fungal networks grow larger on the same amount of sunlight, and they recover faster from drought. The fungi are not just couriers; they are also chemists, secreting enzymes that liberate nutrients from organic matter that plant roots cannot crack open on their own.
There are several distinct guilds of mycorrhizal fungi, and they matter in different ways. Arbuscular mycorrhizal fungi, the oldest group, partner with most crop plants and tropical trees. Ectomycorrhizal fungi dominate the boreal and temperate forests of the Northern Hemisphere, including pines, oaks, birches, beeches, and Douglas-firs. Ericoid and orchid mycorrhizas are more specialised, but together these guilds cover nearly every terrestrial biome on Earth.
The First Global Atlas of Forest Symbioses
To understand how the Stewart 2026 numbers translate to real landscapes, it helps to pair them with the 2019 map from Steidinger and colleagues, which used more than 1.1 million forest plots from over 70 countries to chart where the dominant mycorrhizal types sit on the globe [2]. The pattern is striking. Ectomycorrhizal forests form a band across the boreal zone and stretch down through the mountain ranges of North America, Europe, and East Asia. Arbuscular mycorrhizal forests dominate the tropics, the temperate broadleaf zones of the Southern Hemisphere, and most of the world's savannas.
This matters for climate science. Forests dominated by ectomycorrhizal fungi tend to store more carbon in the soil than arbuscular-dominated forests, partly because the fungal tissues themselves are richer in compounds that resist breakdown, and partly because the bacterial communities that decompose organic matter behave differently in their presence. Knowing which fungal guild owns which patch of forest lets modellers predict how carbon stocks might shift as the climate warms and tree ranges migrate poleward.
The "Wood Wide Web" and Its Critics
The phrase "Wood Wide Web" entered scientific writing in 1998, and over the following two decades it became a popular shorthand for a much more specific set of claims: that trees subsidise their neighbours through shared fungi, that dying trees send resources to seedlings, and that warning signals about insect attack can race through mycelial threads to plants that have not yet been touched. The original tracer studies, beginning in the late 1990s, used carbon isotopes labelled with carbon-13 or carbon-14 to follow sugars from one plant into another.
By 2023, a formal reanalysis of the underlying experiments had catalogued a striking range of results. Some studies found substantial transfer of labelled carbon from a donor plant into a receiver; in others, the figure was indistinguishable from zero. The reanalysis concluded that the variation was real, often driven by species, season, and soil phosphorus levels, but that the popular image of forests as cooperative chat rooms overstated what the evidence could support. Common mycelial networks exist. The fungi do carry signalling molecules. Whether those signals amount to intentional communication between trees remains, for now, an open question with mixed empirical support.
What Remains Genuinely Uncertain
Three questions sit open in the literature. First, how much of the carbon that fungi receive from one plant actually reaches a second plant, as opposed to being respired or locked into fungal tissue? Field measurements vary by an order of magnitude depending on species, season, and soil chemistry. Second, are the warning signals that have been observed in potted seedlings robust in mature forests, where roots intertwine with dozens of fungal species at once? Third, how will these networks respond to warming, nitrogen deposition, and the loss of old-growth structure? Early evidence suggests the more specialised partnerships, especially ectomycorrhizal ones, are also the most vulnerable.
What I find genuinely impressive, sitting with this as someone who has spent a career digging in forest soils, is how much of the story was invisible to us a generation ago. The Stewart 2026 map is a milestone, but it is a snapshot, not a portrait. The next decade of work will tell us whether this vast, quiet infrastructure is as resilient as it looks, or whether the conversations we thought we heard were mostly our own projections into the dark.