Global map estimates mycorrhizal hyphae in soil

Beneath most land plants run the hyphae of arbuscular mycorrhizal fungi. Until now, however, there was no global, comparable estimate of their density. An international team gathered results from more than sixteen thousand soil cores, used over three hundred thousand laboratory measurements of hyphae, and combined the data with information on climate, soil and land use. On that basis it built a model predicting the distribution of the networks in the topsoil layer of the world's soils.

The estimated total length is about 110 quadrillion kilometres of hyphae, and the associated biomass contains roughly 300 megatonnes of carbon. The authors also model a flow of around four billion tonnes of carbon dioxide equivalent per year into the soil by way of these fungi. The figures are enormous, but they do not describe one continuous "underground internet". They are the sum of countless local networks of many species and individuals.

The model indicates that about forty per cent of the total length falls in grassland ecosystems. For arable land the predicted density was roughly half that in comparable uncultivated habitats. There may be several reasons: ploughing tears hyphae apart, fertilising changes the profitability of the symbiosis, and simplified crop rotation limits the continuity of hosts. The result does not, however, say that every agricultural practice always reduces mycorrhiza by exactly half.

The greatest caution concerns converting hyphae into climate. Carbon passed by a plant to a fungus does not automatically remain in the soil for centuries. Some returns to the atmosphere through respiration and decomposition, and the residence time depends on minerals, soil aggregates, moisture and the organisms that feed on hyphae. An estimate of flow is not the same as durable removal of emissions. The map also has weaker data coverage in some regions, which increases local uncertainty.

The value of the work lies in establishing a reference point. It is now possible to ask where loss of habitat or agricultural intensification damages particularly dense networks, and where practices that limit ploughing and maintain living roots help to rebuild them. Further field measurements will improve the model. For a reader, the most important thing is to distinguish a striking length from ecological significance: hyphae are infrastructure for exchanging nutrients, but their role in storing carbon requires a balance of inputs, losses and time.

The map can also show where samples are most lacking. A new core from a region of high uncertainty is often more valuable than another measurement from a well-described station. The authors provide a framework for updating, so the value of the estimate should grow through the transparent addition of data, rather than by treating the first version as a definitive atlas of the underground.

Evidence status — 29 August 2026

The Science article was published on 11 June 2026. Its model combines 322 studies, more than 16,000 soil cores across nine biomes and automated measurements of over 300,000 hyphae. The result — about 1.10 × 10¹⁷ kilometres of living arbuscular-mycorrhizal hyphae in topsoil and roughly 300 ± 60 megatonnes of biomass — is a global estimate, not a direct measurement of one connected organism. The authors explicitly identify poorly sampled ecosystems. The map matters as a testable baseline for carbon-cycle work and monitoring priorities, not as an exact hyphal counter for any chosen site.