Mycelium Networks — Nature's Underground Internet

What is mycelium?

The vegetative body of a fungus — the part doing the real work of nutrient acquisition — is the mycelium: a branching, anastomosing mass of filaments called hyphae. Individual hyphae are typically 1–10 micrometres in diameter, yet they can extend centimetres per day, collectively covering vast distances underground. The mushrooms we see above ground are merely temporary fruiting bodies; the permanent, metabolically active organism is the mycelial network hidden beneath the surface. A critical structural feature is anastomosis — the fusion of adjacent hyphae into closed loops — which lets a network reroute flow around damaged sections and stay connected even when parts are disrupted, far more robust than any simple branching tree. The largest known example is a single Armillaria ostoyae (honey fungus) colony in Oregon's Malheur National Forest, covering approximately 9.6 square kilometres and estimated to be around 2,400 years old.

The Wood Wide Web

The Wood Wide Web — coined by science writer Merlin Sheldrake and popularised by Suzanne Simard's research — refers to underground mycorrhizal networks that link the roots of multiple trees through shared fungal intermediaries, forming what ecologists call a common mycorrhizal network (CMN). In 1997, Simard and colleagues published results in Nature showing that Douglas fir and paper birch trees exchanged carbon bidirectionally through their common network — shaded seedlings of each species received net carbon from the other, with direction depending on which tree had surplus photosynthate. Network analysis of CMNs reveals a scale-free topology: a small number of highly connected "mother trees" anchor connectivity, and removing them disproportionately fragments the network. Roughly 90% of terrestrial plant species form mycorrhizal associations, and the total length of mycorrhizal hyphae in the top 10 cm of forest soil can exceed 100 km per square metre.

Nutrient transport: mass flow and diffusion

Mycelium moves nutrients through two complementary mechanisms: cytoplasmic streaming (bulk flow driven by turgor pressure gradients, observed at 5–10 µm/s) and Fickian diffusion for small molecules. Flow through hyphal tubes follows the Hagen-Poiseuille equation, the same law governing laminar viscous flow through cylindrical pipes:

Q = (π · r⁴ · ΔP) / (8 · η · L)
Q = flow rate, r = hypha radius, ΔP = pressure difference
η = cytoplasm viscosity (~1–3 mPa·s), L = tube length

The critical detail is the r⁴ dependence: doubling a hypha's radius increases its flow capacity by a factor of 16 — which is why fungi invest in wide "trunk" hyphae for long-distance transport and use fine exploratory hyphae only at the foraging frontier.

Physarum polycephalum: the optimal network builder

Physarum polycephalum is a slime mould — not technically a fungus, but a member of Mycetozoa — that offers the clearest window into how biological networks solve optimisation problems. In the famous Tokyo rail network experiment (Toshiyuki Nakagaki, 2000, Nature), researchers placed oat flakes on a map at the positions of Tokyo's main cities and inoculated the map with Physarum at the Tokyo position. Over 26 hours, the slime mould extended, explored, and retracted, leaving a network of reinforced veins that closely resembled the actual Tokyo metropolitan rail network in both topology and efficiency. The mechanism is a simple local feedback rule — each vein carrying more flux widens, while low-flux veins shrink — combined with Kirchhoff's current law at each node, producing networks that are simultaneously low-cost, short-path, and fault-tolerant.

Electrical signalling in fungi

Experiments by Andrew Adamatzky and colleagues at the University of the West of England, published in 2022, recorded clear electrical potential oscillations in live mycelium of multiple species, sharing characteristics with neuronal action potentials: spike amplitude of approximately 50 mV, spike frequency up to approximately 1 Hz, propagation along individual hyphae, and stimulus-response coupling — adding nutrients or damage near one end triggered spike propagation toward the stimulus. Unlike neurons, hyphae have no axon, no myelin sheath and no synaptic cleft; the signals more closely resemble plant action potentials, as seen in Mimosa pudica or the Venus flytrap. Whether these spikes are genuine information-carrying communication or electrochemical noise from ion transport remains scientifically debated.

Mycelium as a computing substrate

By placing electrodes at multiple points in a growing mycelial mat and applying stimuli, researchers have recorded outputs consistent with AND gates (output only when both inputs are active) and OR gates (output when either input is active), with the logic implemented by the network's own dynamics rather than designed circuitry. The branching patterns of mycelium are governed by activator-inhibitor dynamics — the same class of equations Alan Turing proposed in 1952 for biological pattern formation — mathematically identical to equations used in some neuromorphic computing architectures and spiking neural network models. Practical constraints remain substantial: biological processes run millions of times slower than silicon and the substrate cannot be conventionally reprogrammed.

Sustainable materials and network science lessons

Mycelium composites, made by inoculating agricultural waste like hemp hurds or sawdust with fungal strains, offer thermal insulation around R-3.5 per inch, compressive strength up to 200 kPa, natural fire resistance, and full home-compostability — commercialised by companies like Ecovative Design and MOGU. Mycoremediation exploits fungal enzymes (laccases, peroxidases) that break down lignin and structurally similar pollutants such as petroleum hydrocarbons and synthetic dyes, while fungal cell walls biosorb heavy metals. At the network level, anastomosis gives mycelium high edge connectivity (the same principle behind mesh networking), while the positive feedback loop of flow widening tubes implements a biological form of reinforcement learning with no central controller — network scientists measure node importance with betweenness centrality, and in forest CMNs, hub "mother trees" have disproportionately high centrality, so protecting them may matter more for network resilience than protecting an equal number of random trees.

Frequently asked questions

Try it live

Everything above runs in your browser — open Mycelium Growth and watch nutrient-driven network expansion and anastomosis formation in real time. Nothing is installed, nothing is uploaded.