Electrical recordings from oyster mushroom mycelium colonizing a wood-shavings substrate reveal that spiking activity is neither uniform nor random. Using an eight-channel star-shaped electrode array, researchers in 2026 found that voltage spikes cluster into bursts whose frequency, amplitude, and duration differ sharply by direction — some sectors of the colony firing ten times more frequently than others. Propagation delays between channels ranged from seconds to hours, and burst origins were reproducibly concentrated in specific angular regions.
The finding reframes mycelial networks as spatially extended excitable media rather than passive nutrient conduits. The directional heterogeneity is the key detail. A homogeneous network would produce statistically similar activity at all recording points; what appears instead is a landscape of preferred ignition sites and biased propagation corridors. The colony does not simply conduct signals — it routes them, using morphological asymmetry and substrate contact geometry to shape where electrical events begin and how far they travel.
Note (2026-11-11): checked against the paper (Adamatzky, arXiv 2601.08099; one author, three recording sessions of up to five days each). The spike-rate difference (more than an order of magnitude between directions), the clustered bursts and the delays from seconds to hours are as in the paper. Three statements are the essay's own. (1) The essay says the colony routes signals 'using morphological asymmetry and substrate contact geometry'; the paper says the anisotropy 'likely reflects' hyphal density, nutrient availability, transport pathways or metabolic demand, and it does not mention substrate contact. (2) The essay says burst origins were reproducibly concentrated in specific angular regions; the paper also reports that the identity of the most active directions can vary between recording sessions, which it reads as dynamic reconfiguration of electrically active regions rather than fixed patterns. (3) The essay says the network physically remodels itself over days and weeks as a fact; the paper did not measure growth or structure, and names growth dynamics only as one possible mediator of the slow propagation. The paper describes the slow spread as unlike fast excitable-wave transmission, and says its findings support the excitable-medium interpretation.
This has consequences well beyond mycology. Excitable media — cardiac tissue, slime molds, chemical reaction-diffusion systems — share the property that local threshold crossings propagate through the medium as waves. But the mycelial case adds a dimension most laboratory excitable media lack: architectural growth. The network physically remodels itself over days and weeks, meaning the routing topology is not fixed but co-evolves with the signals flowing through it. Structure shapes signaling, and signaling shapes structure.
The implication extends to any system where information flows through a substrate that the information itself is building. Infrastructure is never neutral. The geometry of a network does not merely permit communication — it edits it, preferentially amplifying certain messages while attenuating others based on nothing more than the accident of where growth began.