A road cuts a once-continuous forest into two patches. Animals (the drifting spheres) roam their home patch and, at a rate set by dispersal drive, attempt to disperse toward the other side — the natural behaviour that keeps a metapopulation genetically mixed. If a wildlife corridor (the raised green overpass) exists and has spare capacity, they cross through it safely. Otherwise they risk a direct road crossing, where traffic can kill them before they make it across. Every corridor arrival also refreshes that patch's genetic health, which recovers toward full when immigrants keep arriving and decays when a patch sits isolated for too long — the textbook signature of inbreeding depression in a fragmented population. Health throttles the birth rate directly, so an isolated patch quietly shrinks even with no extra deaths.
P(safe crossing) = corridor has capacity ? corridor : road (risk ∝ traffic·dt)
health(t+dt) = health + recoveryRate·arrivals − decayRate·(1−recentArrivals)
births/dt ≈ growthRate · health · pop · (1 − pop/capacity)
- Starting population — combined animals across both patches when the simulation resets.
- Corridor width — 0 removes the corridor entirely (every crossing is a road gamble); higher values raise how many animals can cross safely at once.
- Road traffic — the death probability per second for an animal exposed on the open road; a corridor bypasses this risk completely.
- Dispersal drive — how often animals attempt to leave their home patch at all, i.e. how mobile the species naturally is.
Real-world relevance: this is why conservation agencies build wildlife overpasses and underpasses across highways — a narrow but present corridor restores enough gene flow to prevent the slow genetic and demographic collapse that isolated fragments otherwise suffer.