This is the same worm-colony population model as the 3D bin view, drawn as a 2D parameter-space heatmap instead of a rendered tank. Every pixel of the top panel is a (temperature, moisture) pair, shaded by the colony's net per-capita growth rate at that pair — so instead of watching worms wander a box at one condition, you see the entire tolerance landscape the colony could be living in, all at once:
dN/dt = r_max · f(T) · f(M) · N · (1 − N/K) − d(T) · N
f(T) = exp( −((T − 25) / 9)² ) optimum ≈ 25 °C
f(M) = exp( −((M − 75) / 20)² ) optimum ≈ 75% moisture
d(T) = 0 for 5 °C ≤ T ≤ 35 °C
d(T) = 0.35·(35−T)/5 below 5 °C or above 35 °C (cold/heat kill)
Green shading marks (T, M) pairs where the colony is net growing; red marks pairs where it is net dying (the cold/heat-kill bands). The white dot is the bin's current condition — drag the sliders and watch it slide across the landscape in real time, with a fading trail of where it has recently been. Because the landscape itself barely moves (it only shifts slightly as the crowding term (1 − N/K) changes with population), a slider drag reads as pure geometry: moving the dot into the red band is the exact same cold/heat kill event as the 3D tank's die-back, just seen as a position in tolerance-space instead of a camera view of the bin. The bottom strip chart scrolls worm population, waste backlog and cumulative castings over the last simulated days, so you can see the same feeding/growth/casting cycle the 3D tank's layer heights show.
- Temperature — the landscape's horizontal axis; growth peaks near 25 °C and the red kill bands appear below 5 °C and above 35 °C.
- Moisture — the vertical axis; growth peaks near 75%, the "wrung-out sponge" optimum.
- Feed rate — doesn't move the dot (it isn't part of the growth landscape) but drives the waste/castings traces on the strip chart below.
- Carrying capacity K — set by bin volume (≈ 2,000 worms); as the population approaches it the whole green region dims slightly, since crowding slows growth even at perfect temperature and moisture.
Real-world relevance: this is the same population model small-scale and commercial vermicomposting operations use to size bins and predict castings yield from feedstock and climate control.