How Do Trees Communicate? Roots, Fungi and Airborne Chemistry

The best-established channel is volatile organic compounds, gases released from damaged leaves that drift to neighbours. The first suggestion came in 1983 from work on willows and poplars, where trees near damaged ones increased their production of defensive compounds, and it was treated with scepticism for years before being confirmed repeatedly. The mechanism is now well documented in sagebrush, lima beans, maize, tomatoes and several tree species: a plant under attack emits a characteristic mixture, and plants downwind that detect it prime their own defences, producing bitter tannins, toxins or thicker cell walls faster when attacked themselves. A second and more surprising use is recruitment, since some of the emitted compounds attract the predators and parasitic wasps that eat the herbivore doing the damage, which amounts to calling for help. Whether this counts as communication or as eavesdropping on an unavoidable emission is a real argument, and for a plant signalling to its own distant branches, which also happens, the question does not arise.

Underground, almost all trees live in partnership with fungi. The fine threads of mycorrhizal fungi wrap or penetrate the root tips, vastly extending the surface available for absorbing water and minerals, and take sugars from the tree in exchange, a partnership perhaps 400 million years old and found in the great majority of land plants. Because a single fungal individual can connect the roots of many trees, the forest floor contains a physical network linking separate plants, which Suzanne Simard demonstrated carbon moving along in a 1997 study using labelled isotopes between birch and fir. The findings that followed are genuinely striking:

• •Carbon moves between trees along fungal threads, and the direction can reverse with the season, flowing from birch to shaded fir in summer and back in spring
• •Nitrogen, phosphorus and water also move, and seedlings connected to a network of established trees survive better than isolated ones
• •Defence signals travel underground as well as through the air, with attacked plants triggering responses in connected neighbours
• •The network is not evenly connected: large old trees link to many others, which is where the term mother tree came from
• •Some plants exploit the system entirely, and the ghost orchid and Indian pipe have no chlorophyll at all, living wholly on carbon taken from fungi connected to photosynthesising trees

The network became famous under the name wood wide web, and a body of popular writing has described forests as cooperative communities in which old trees deliberately nurse their young and share resources with the sick. A set of critical reviews since 2023, led by Justine Karst, Melanie Jones and Jason Hoeksema, has gone back through the underlying studies and found the support thinner than the popular account implies. Their conclusions are worth stating plainly: common networks are real and widespread, but their structure in natural forests has been mapped in very few places; transfer of carbon between mature trees has been shown, yet the quantities are usually small relative to what a tree makes for itself; the claim that seedlings preferentially benefit is inconsistent across studies; and the idea that trees deliberately direct resources to kin is not established, since the fungus has its own interests and may be the party doing the allocating. Simard disputes parts of the critique. The honest position is that the plumbing exists and is important, and that the intentions attributed to it are not evidence-based.

The habit of describing the network as an internet obscures that the fungus is an organism with its own strategy rather than a cable laid for the trees' convenience. It can be mutualistic, parasitic or somewhere between depending on conditions, it can take more sugar than the nutrients it returns are worth, and where carbon moves from one tree to another it may be passing through fungal tissue for the fungus's own reasons rather than being delivered. Several researchers argue that a fungus connected to many trees benefits from keeping its whole portfolio alive, which would produce apparent generosity without any tree intending anything. This is a less charming story and a more tractable one, because it makes predictions that can be tested, and it does not require plants to have goals.

The practical stake is in forestry. If seedlings establish better when connected to existing fungal networks, then clear-felling, which destroys the network along with the trees, is more damaging than replanting figures suggest, and retaining some mature trees at harvest is worth its cost. If the network moves water during drought, the resilience of a mixed old forest differs from that of a plantation of one age and one species. And the defensive signalling has agricultural applications already being trialled, since a crop that can be primed by an airborne compound to prepare its defences in advance needs less pesticide. None of these depend on trees being altruistic. They depend on the forest being a connected system rather than a collection of separate plants in the same field, which is the part of the claim that the evidence supports firmly.

Signals pass between trees in two established ways: airborne compounds released by damaged leaves, which prime the defences of neighbours and attract predators of the attacker, and mycorrhizal fungi whose threads physically connect the roots of many trees and along which carbon, nutrients, water and defence signals demonstrably travel. The popular picture of old trees deliberately nursing their young has run ahead of the studies, which critical reviews since 2023 have shown to be fewer and less consistent than claimed, and the fungus has interests of its own.