The Wood Wide Web: Forest Communication and the Research Behind It

The Specific Question: What Did Simard Actually Find, and What Does the Critique Say?

In 1997, Suzanne Simard and colleagues published a study in Nature (vol. 388, pp. 579-582) that used carbon isotope tracing to measure carbon movement between Douglas fir and paper birch connected by a shared ectomycorrhizal network in a British Columbia forest. The experimental design applied carbon-13 to birch and carbon-14 to Douglas fir in replicated plots, then measured isotope ratios in the adjacent trees after 20 days. The results showed net carbon transfer from birch to Douglas fir when Douglas fir was shaded, and net transfer in the opposite direction under different light conditions. Bidirectional transfer through the fungal network was detectable and statistically significant.

This is the core finding that earned the research its reputation. It demonstrated that two different tree species, connected through a shared fungal network, could exchange carbon at measurable rates. The ecological implication, that forests are physiologically integrated systems rather than collections of competing individuals, was a genuine shift from the dominant competitive framing of forest ecology at the time.

What followed in popular science was a separate matter. The phrase "wood wide web" entered mainstream discourse around 2016, amplified by a TED talk, a documentary, and Simard's 2021 book "Finding the Mother Tree." The popular framing added claims that large hub trees ("mother trees") deliberately allocate carbon to their seedling offspring, that trees can recognise kin, and that forest networks operate as cooperative intelligence. These claims extend significantly beyond what the 1997 paper and subsequent studies established.

The Karst critique does not invalidate the Simard 1997 finding. Bidirectional isotope transfer through shared mycorrhizal networks has been replicated. What the critique establishes is that the popular framing substantially overstates what the science supports. The distinction matters for this pillar: the mycorrhizal fungi pillar makes its case on mechanism, not on sentimentality, and the mechanism is strong enough without the additions.

The Mechanism: Sink-Source Dynamics, Not Communication

Ectomycorrhizal fungi (ECM) form associations with most forest trees, including Douglas fir, beech, oak, pine, and birch, wrapping root tips in a dense hyphal sheath rather than penetrating root cells (the arbuscular mode). ECM hyphae extend outward from root tips into the surrounding soil matrix, forming dense networks that can span tens of metres and connect multiple trees of the same or different species. The critical point is that the fungal network connects to more than one host simultaneously. A single ECM fungal individual (genet) can colonise root tips of multiple adjacent trees, creating a shared vascular pathway between those trees at the hyphal level.

Carbon movement through this shared network is governed by source-sink dynamics. Carbon in the form of sugars flows from areas of high photosynthetic production (source) to areas of high metabolic demand (sink). A shaded seedling has high demand and low production: it is a strong sink. A canopy-dominant tree in full sun has high production and lower marginal demand: it is a source. When both are connected through a shared ECM genet, carbon follows the concentration gradient from source to sink. No decision is made by any tree. The fungus moves carbon where the chemistry dictates.

The fungal perspective adds a further complication to the altruism narrative. ECM fungi are not neutral conduits. They are organisms with their own metabolic requirements, and they retain a fraction of the carbon they move. There is active research on how much of the inter-tree transfer documented in isotope studies is actually fungal appropriation rather than tree-to-tree transfer. The fungal network is not a pipeline: it is a living system that extracts a toll from every transaction it facilitates.

The distinction between arbuscular mycorrhizal fungi (AMF) and ectomycorrhizal fungi matters here. AMF associate with approximately 80 percent of vascular plant species, including most agricultural crops. ECM associate primarily with forest trees. The wood wide web research, including Simard's work, is almost entirely about ECM forest networks. The findings do not directly translate to agricultural AMF systems. This is covered in depth in the arbuscular vs. ectomycorrhizal page.

The Numbers: Transfer Rates, Network Density, and What the Evidence Actually Shows

The Simard 1997 study reported net carbon transfer representing 3 to 6 percent of daily photosynthate in Douglas fir seedlings growing under shade conditions. This is not a trivial amount for a shaded seedling, but it represents a marginal contribution to the carbon budget of a sun-exposed canopy tree. The ecological significance of inter-tree carbon transfer is contested precisely because the fractions involved are small relative to each tree's own photosynthetic production.

The practical numbers on ECM network density are less contested. In mature temperate forests, the extraradical mycelium of ECM fungi can reach 200 to 400 metres of hyphae per gram of soil in the top 10 centimetres, significantly denser than AMF networks in agricultural soils. A single ECM genet can span hundreds of square metres and connect dozens of trees. The sheer physical density of the network is well-established. What is disputed is the functional significance of the connections in terms of resource allocation and ecological outcome.

The scale of ECM networks is also why agroforestry systems with established tree rows carry high fungal network density and why newly planted trees in bare soil take years to build comparable infrastructure. Cover crops in cover crop management systems can accelerate AMF network establishment in arable contexts, but ECM networks in tree-based systems operate on decade timescales that cannot be short-circuited.

The Practitioner View: What Forest Network Research Means for Managed Land

For a practitioner managing agroforestry, woodlot, or reforestation, the practically useful findings from ECM network research are narrower than the popular framing suggests but still significant. Three findings translate reliably into management decisions.

First: seedling establishment near established trees is consistently better than in open ground, and ECM network access is a plausible partial explanation. When replanting forest after harvest, maintaining nurse trees that can share ECM networks with new seedlings reduces establishment failure rates. The mechanism may be network carbon transfer, or it may be other facilitative effects from the established tree, including shade, litter chemistry, and soil moisture modification. The practical recommendation (retain nurse trees when replanting) is robust even if the mechanism attribution is contested.

Second: ECM network diversity correlates with forest resilience. Forests with multiple ECM fungal taxa associated with their trees show more stable responses to drought, pest, and pathogen stress. The mechanism is redundancy: different ECM taxa perform well under different conditions, and a diverse network maintains function across environmental variation. This is directly analogous to the AMF diversity argument in agricultural contexts. Monoculture fungal networks, like monoculture crops, are brittle under stress.

Third: the network architecture of ECM forests supports the agroforestry productivity case even without the communication narrative. Mature tree systems in alley cropping designs sit at the interface of ECM and AMF systems. Tree rows contribute ECM network density; crop rows operate on AMF networks. Where tree roots and crop roots overlap in the alley zone, the fungal community is richer and more diverse than in either monoculture. This is a functional productivity benefit, measured in nutrient uptake efficiency and drought resilience, that does not require any claim about inter-tree communication. The syntropic agriculture page covers the full tree-crop network integration argument.

The management implication for restoration practitioners is this: protect established ECM networks during harvest and replanting. The network infrastructure in the soil is often more valuable than the above-ground biomass being managed, and it is far harder to restore. Selective harvest that retains nurse trees preserves decades of network investment. Post-harvest biochar application from residue pyrolysis can partially compensate for disrupted ECM habitat by providing pore structure that newly colonising fungi can exploit, but it is not a substitute for preserving the live network: biochar accelerates recolonisation from surviving fragments but cannot replace the decades of community diversity that selective harvest would have preserved. Clear-cutting resets the network to zero and adds ECM reestablishment time to the reforestation timeline, which is already measured in years.

Where It Fits: The Wood Wide Web Debate as a Lesson in Evidence Quality

The wood wide web story illustrates a pattern that appears across regenerative agriculture claims: a genuine and important scientific finding gets amplified into a narrative that overreaches the evidence, which then provokes a correction that is interpreted by critics as invalidating the original finding. Neither the amplification nor the overcorrection is accurate.

The genuine finding: shared mycorrhizal networks connect trees, allow measurable nutrient transfer, and the density and diversity of those networks correlates with ecosystem resilience. This is real, reproducible, and important. The overreach: trees communicate, cooperate, and make decisions through these networks in ways that parallel human social behaviour. This is not supported by the evidence and creates unrealistic expectations for what ECM network management can deliver. The correction (Karst 2023) identifies the overreach precisely. It does not invalidate the genuine finding.

For practitioners working with the root exudate and microbiome page, this distinction matters. Root exudate chemistry does influence which fungal partners colonise a given root zone, and different plant species do release different signal compounds. The question of whether those signals constitute communication in any meaningful biological sense is separate from the management question of whether exudate diversity correlates with fungal network richness. It does, and that is the lever to manage.

The practical lesson from 25 years of wood wide web research is that forest ecosystem function depends on below-ground network infrastructure at least as much as on above-ground canopy structure. Management decisions that damage or destroy the fungal network, including compaction by heavy machinery, bare-soil periods, and fungicide application, carry long-term costs that are not captured in immediate yield measurements. Compost applications in the first year after any soil-disturbance event deliver AMF and ECM spores from the compost's plant-residue community, accelerating fungal network recolonisation from a broader community pool than would arrive naturally from adjacent undisturbed zones alone. The cost accounting for below-ground network damage belongs in every forestry and agroforestry management model, regardless of what one thinks about tree sentience.