The Wood Wide Web: The Network That Feeds the Forest

Start with what is unarguable, because it is remarkable enough on its own. Nearly all plants, roughly 90 percent of all land plant species, live in intimate partnership with fungi at their roots, a relationship called mycorrhiza, meaning simply 'fungus-root.' It is a genuine mutual exchange: the fungus threads its fine filaments, called hyphae, far out into the soil, hoovering up phosphorus, nitrogen, and water that the plant's own roots cannot reach, and hands them to the plant; in return, the plant pays the fungus in sugar, feeding it something like 4 to 20 percent of the carbon it captures from sunlight. This is not a fringe arrangement. It is how most of the plant kingdom actually eats, and it is ancient: fossils from the Rhynie chert in Scotland show plants already partnered with these fungi around 410 million years ago, among the very first life to colonize dry land. There are two main styles. In most plants the fungus grows right inside the root cells; in the trees of temperate and northern forests, oaks, pines, birches, firs, the fungus instead wraps each root tip in a dense sheath and weaves between the cells in a structure called the Hartig net. Either way, the root and the fungus become, functionally, one organ.

The leap from 'plants have fungal partners' to 'plants are networked to each other' is small and also well established. Because a single fungus can partner with several plants at once, its threads can physically link the root systems of two or more separate plants into a shared, continuous web of fungal tissue. That literal, physical connection is what the phrase 'wood wide web' actually, defensibly points to, and it has been directly observed. In one striking piece of work, researchers used DNA fingerprinting to map an entire real network in a Douglas fir forest and found it was 'scale-free,' the same mathematical shape as the internet or a social network: a few big, old, heavily-connected 'hub' trees, and many young trees hanging off them by just a link or two. So the wiring is real, and it is not even randomly arranged. Every honest part of the popular story starts here, on firm ground. The trouble begins only when we ask what, exactly, flows along those wires, and why.

The whole modern story traces to a single elegant experiment. In 1997, the Canadian forest ecologist Suzanne Simard and her colleagues published a paper in Nature with the plain title 'Net Transfer of Carbon Between Ectomycorrhizal Tree Species in the Field.' Crucially, this was done in a real forest, not a laboratory pot. Simard's team fed paper birch and Douglas fir seedlings two different traceable isotopes of carbon, then measured where the carbon ended up. It moved: carbon flowed through the shared fungal network from one tree species to the other, and did so in both directions. This was a genuine, peer-reviewed, field result, and it is the bedrock the entire wood-wide-web concept rests on. Two points of precision matter enormously, though, and they are usually lost in the retelling. First, the net flow ran from the sun-lit birch to the shaded fir, which is exactly what you would expect from simple physics: carbon, like water, tends to move from where there is more of it to where there is less, down a 'source to sink' gradient. It was not evidence of a tree 'choosing' to help another. Second, the now-famous phrase 'wood wide web' was not Simard's. It was coined by an editor at Nature, writing up her result with a nod to the World Wide Web. The metaphor, in other words, arrived attached to the science almost from the very first day, and the two have been tangled together ever since.

In the decades since, real laboratory and field work has genuinely expanded what we know the network can do, and this middle layer of findings is solid even where its forest-scale importance is still argued over. Fungal networks carry a real nitrogen-transfer pathway between plants. When bean plants are attacked by aphids, unattacked neighbors linked to them by fungal threads have been shown to switch on their own chemical defenses in advance, apparently cued by a signal traveling through the shared network, and a similar defense relay has been demonstrated between tomato plants. Seedlings that tap into an established network tend to survive and grow better than those left to start alone. And richer fungal diversity underground measurably boosts the productivity, stability, and even the plant biodiversity of the community above. There is no serious doubt that mycorrhizal fungi are keystone players in how ecosystems work. What remains genuinely open is a narrower and more stubborn question: how much the direct tree-to-tree transfer of resources through these networks actually matters, out in a real forest, at a scale that changes which trees live and die.

Before turning to the contested claims, it is worth pausing on just how physically vast a single fungus can be, because it makes the idea of an underground network viscerally real. In the Malheur National Forest in Oregon lives a single fungal individual, a honey fungus called Armillaria ostoyae, whose body spreads through the soil across an estimated 9.6 square kilometers, roughly 2,400 acres. By area, it is one of the largest living organisms on Earth, all of it a single genetic individual, mostly invisible, threaded through the forest floor and occasionally sending up clusters of mushrooms. That particular fungus happens to be a tree-killing parasite rather than a helpful partner, so it is not itself a mutualistic network. But it dramatizes the key physical fact underneath this whole subject: a single fungal body really can reach across a forest. The extensive, soil-spanning growth that makes a giant Armillaria possible is the same kind of reach that lets mutualist fungi wire many separate trees together. The stage on which the wood wide web plays out is genuinely enormous.

The network's reach shapes real forest chemistry, too, in ways that matter beyond the trees themselves. Forests dominated by the sheath-forming (ectomycorrhizal) fungi store markedly more carbon in their soils than other forests, because those fungi compete with decomposers for nitrogen and slow the rotting of dead matter, an effect with real weight for climate carbon models. And at the far edge of the network's biology sits one genuinely strange fact: around 400 plant species have given up photosynthesis entirely. Ghostly white plants like the aptly named Indian pipe, and many orchids, have no green in them at all; they survive by plugging into a fungal network and siphoning off carbon that the fungus obtained from nearby green trees. They are, in effect, plants that have learned to hack the network and live as freeloaders on it. This peculiar lifestyle has evolved independently more than 40 separate times, which tells you that exploiting the shared network for a free meal is not a freak accident but a recurring opportunity that evolution keeps rediscovering. The network is real, it is large, and things really do move through it. All of which makes it doubly important to be careful about the beautiful claims it cannot yet bear.

Here is the heart of the matter, and the reason this article exists. The most beloved parts of the wood-wide-web story, the parts that made it a phenomenon, are precisely the parts the evidence does not currently support. The centerpiece is the 'mother tree': the claim that the largest, oldest trees act as network hubs and preferentially funnel carbon and defense signals to their own genetic offspring, recognizing kin and nurturing it. In 2023, three scientists, Justine Karst, Melanie Jones, and Jason Hoeksema, published a careful review in Nature Ecology and Evolution examining what the published evidence actually shows, and their findings were sobering. They reported that the specific claim, that mature trees preferentially send resources to their own offspring through these networks, has no peer-reviewed published evidence supporting it. Not weak evidence: none, at the level the popular story asserts. They also found that the physical existence of these networks in real forests is far less established than the confident retellings imply, only around five studies have ever actually mapped a mycorrhizal network in the field, across just two forest types and two of the world's roughly 73,300 tree species. And reviewing more than 1,500 papers, they found that the rate at which unsupported, cooperative-sounding claims were being cited approvingly had roughly doubled over 25 years, a snowballing of optimism outrunning the data. Notably, the authors pointed out that some of those overstated citations came from their own earlier work, an honest act of scientific self-correction, not a hit job.

The deflation goes further, and it is worth being specific rather than vague. Where carbon transfer between trees has actually been measured, the amounts are usually tiny, on the order of a fraction of a percent up to a few percent of what the giving tree makes, quite possibly too little to matter to the receiving tree's survival at all. Worse for the tidy story, several experiments claiming network transfer have not fully ruled out that the resources moved by another route entirely, directly through touching roots, or by simple diffusion through the soil, rather than through the fungal threads specifically. And the picture of a warm cooperative commons is contradicted by the network's own darker traffic: fungi appear to run something more like a market, steering scarce nutrients toward whichever tree pays them the most carbon, while trees can cut off underperforming fungal partners, and the very same threads that carry food can also carry poisons between plants or let a parasitic fungus prey on a weak seedling. None of this erases the network. It reframes it. The forest floor is not a nursery run on kindness; it is a churning, competitive, deeply interdependent marketplace, and, tellingly, this caution is not new or purely external. Simard herself co-authored a paper back in 2006 with a wry French title translating to 'Dangerous Liaisons,' flagging that a purely cooperative reading was too simple. The field's own skepticism is older than the backlash.

So the refusals need to be stated plainly, precisely because the temptation to believe otherwise is so strong. No, trees do not 'talk' to each other in any intentional sense; the transfers and signals that have been measured are driven by concentration gradients and ordinary chemistry, by physics and metabolism, not by will, choice, or awareness. No, a 'mother tree' does not consciously decide to feed her children; that specific claim, as we saw, has no supporting evidence at the level it is usually told. And no, the forest is not a giant brain or a conscious superorganism 'thinking' through its fungal wiring; that comparison, which appears throughout popular books on the subject, is an evocative metaphor, offered as a metaphor by its more careful authors, and it must not be mistaken for a demonstrated mechanism. The word 'web,' the word 'communication,' even the phrase 'mother tree,' are all figures of speech laid over a set of chemical and physical processes, and they quietly smuggle in intention that the biology does not contain. Refusing that smuggled intention is not cynicism. It is the only way to see the real thing clearly. Because the real thing needs no embellishment at all. Under the floor of nearly every forest on Earth runs an immense, ancient, living fungal network, hundreds of millions of years old, physically linking most of the plants on the planet to the fungi that feed them and, often, to one another. Resources and even warning signals really do move across it. That we are still arguing about exactly how much, and to whose benefit, is not a disappointment. It is what a genuine, unfinished scientific frontier looks like, and it is far more interesting than the fairy tale.