Optimal, self-organized, behaviors in the living plasmodial network, Physarum polycephalum

Relying on hydraulics for information transfer, and with no known system for coordinating behaviors between its different parts, the slime mold, Physarum polycephalum, nevertheless builds active networks that solve difficult optimization problems, such as efficiently linking together food sources. How, do Physarum networks achieve this? I will report our efforts to map Physarum’s behaviors, starting with the set of internal flows that an active network can create via actively contracting and expanding its plasmodial tubes. We construct a space of flow modes, ranked by their efficiency in creating long distance fluid flows, and find that real networks robustly select a small number of optimally efficient flow modes, switching between local optima in response to new cues from its environment. Moreover, we show that efficient internal transport self-organizes if the plasmodia simply respond to pressures created within the network, allowing apparently intelligent behaviors to emerge from the physics of internal transport.