Walk into almost any forest and beneath your feet, threading through the soil, is a network of fungal filaments called hyphae, many of them fused to tree roots in a relationship called mycorrhiza. This part is not in dispute. Somewhere between 80 and 90 percent of temperate and boreal forest tree species form these partnerships[1], and the exchange at their core has been demonstrated repeatedly with radioactive and stable isotope tracers: the fungus delivers soil nitrogen and phosphorus to the tree, and the tree delivers photosynthesized carbon to the fungus[2]. That much is textbook mycology, verified across decades of experiments.

What is far less settled is the more romantic claim that made mycorrhizal networks famous outside plant science departments: that these fungal threads link separate trees into a single, forest-wide communication system — a "wood wide web" — through which mature trees recognize their own offspring, send them extra carbon, and warn neighbours of insect attack.

Where the popular version outran the evidence

In February 2023, a team led by forest ecologist Justine Karst published a systematic review in Nature Ecology & Evolution that traced this narrative back to its sources — and found the foundation considerably shakier than the popular retelling suggested[3]. The team examined three specific claims that recur in books, documentaries and news coverage:

  1. That common mycorrhizal networks (fungal networks linking multiple trees) are widespread in forests.
  2. That resources move through these networks in ways that measurably improve seedling performance.
  3. That mature trees preferentially send carbon and defence signals to their own offspring through the network.

For the first claim, the review found that field evidence was too limited or had too many alternative explanations to support a general statement that such networks are widespread. For the second, existing studies did not support the claim that network-mediated transfer measurably benefits seedlings. For the third — the "mother tree" idea most associated with popular science writing on this topic — the reviewers found no peer-reviewed, published evidence at all[3].

Reviewing how often later papers accurately represented the original field studies they cited, Karst and colleagues found that fewer than half of the statements made about those studies in papers published in 2022 were accurate.[3]

The mechanism behind that drift has a name in the review: positive citation bias. A study with modest, caveated findings gets cited by a slightly more confident summary, which gets cited by a more confident summary still, until the eventual popular account states as settled fact something the original researchers described as tentative. It is a useful case study in how a real, interesting finding — trees and fungi do exchange resources — can accumulate an extra layer of narrative that the underlying data never supported.

What is actually well established

None of this means the underlying biology is unimpressive; if anything, the verified mechanics are more interesting for being precise. Using radioactive phosphorus-33, nitrogen-15 and carbon-14 as tracers, researchers have directly measured the movement of these elements between fungal hyphae and host root tissue, confirming genuine two-way nutrient exchange at the cellular level[2]. Thermodynamic modelling of the transport pathways suggests phosphate moves out of the fungus via proton-coupled transporters rather than passive diffusion — an active, energy-spending process on the fungus's part, not a leak[4]. In beech forests, researchers have traced recently photosynthesized carbon to the exact ectomycorrhizal root tips where fungal-delivered nitrogen was concentrated, showing the exchange is spatially coordinated within an individual root system, not diffuse[5].

At the scale of a single tree and its fungal partners, in other words, mycorrhizal symbiosis is a well-characterized, actively regulated nutrient market. It is the leap from "a tree trades with its fungal partners" to "a forest is a single cooperating superorganism, communicating and providing for its young through that partnership" where the science currently runs out ahead of the evidence.

Where the record is uncertain

Justine Karst, the review's lead author, has been careful to note that the paper does not claim mycorrhizal networks linking multiple trees do not exist — some have been physically mapped in specific forest plots. The finding is narrower: that claims about how widespread these networks are, and what they do for seedlings and "parent" trees specifically, are not yet supported by the volume or quality of evidence that popular accounts imply[3]. Further controlled field research, rather than reinterpretation of older studies, is what the review's authors say is needed to settle the question either way.

For a publication that treats uncertainty as something to explain rather than smooth over, this is a useful example of why that distinction matters. The mycorrhizal fungi beneath a forest floor are doing real, measurable, mutually beneficial work with the trees they are attached to. Whether they are running a forest-wide social network is, for now, still an open question — and the honest answer is more interesting than the settled one would have been.