Presets
Playback
Apparatus
Model checks
The science
What is modelled
A grazing rabbit divides its time between eating and watching. More watching means the fox is seen from farther away but less food. The grass is richer farther from the warren, which has been grazed bare, but a rabbit far from it needs longer to reach safety. An attack is a race: the rabbit reacts after a short delay and sprints home while the fox runs at it.
Equations
- Mean sighting distance
sdet = s1 (1 + k v), sighting distance exponential. - Time to safety
tb = τ + d / vr; the rabbit escapes if the fox is seen beyondsmin = max(vf τ, vf tb − d), soP = exp(−smin / sdet). - Feeding rate
I = (1 − v)(1 − e−d/L). - Bout fitness
F = I T exp(−λ (1 − P) T)with T = 4 h.
Checks run in the page
- A time-stepped chase Monte Carlo reproduces the escape probability.
- A Poisson attack Monte Carlo reproduces the survival of the bout.
- A golden-section search confirms the grid optimum.
- The analytic marginal benefit of distance matches a finite difference.
Honest simplifications
The chase is one-dimensional and in a straight line, speeds are constant with no acceleration or dodging, the fox never gives up, detection is a single exponential distribution and the fitness is a simple product of food and survival. It shows the shape of the trade-off, not measured rabbit behaviour.
What to try
- Press Fast fox and see the best distance move toward the burrow.
- Press Greedy and far away and read the escape probability.
- Press Many foxes and watch the best vigilance rise.