What decides whether the molds gather into a few big clusters or stay in a scattered web? Within about 50 ticks the even scatter breaks up into bright yellow-green clusters where molds have piled up, joined by fainter streams.
Try this Pick the Thick Highways preset (900 molds, deposit 60, diffusion 60, evaporation 4) and press Reset, since population only changes at setup. The molds fuse into a few blobs about 15 patches across. Fine Web (150 molds, evaporation 20) leaves only faint, scattered threads.
About this model Where it ends up
A port of NetLogo's Slime, a model of slime mold aggregation. Cells that each drop a chemical and steer toward more of it gather into clusters with no leader, much as cellular slime molds such as Dictyostelium gather by following cAMP.
Each tick, in random order, every mold compares the chemical on the patches 1 step ahead, 45° left and 45° right, turns 45° toward the richer side if either beats straight ahead, wiggles up to 39° either way, steps forward 1 and adds deposit-amount to its patch. Then each patch passes diffusion-rate percent of its chemical evenly to its 8 neighbors, and every patch loses evaporation-rate percent. Only population needs a Reset; the other sliders are read every tick. Clicking a patch adds 60 chemical to it.
NetLogo's version has sliders for sniff-threshold, sniff-angle and wiggle-angle and fixed chemical rates; here the angles are fixed, molds always follow the gradient, and the chemical rates are sliders. Nor is this Jones's (2010) Physarum model, whose agents sense several cells ahead and build networks. These sense one patch ahead and form clusters.
