This is the same top-down map every wildlife crossing gets drawn on for planning: two forest patches, a highway between them, and every animal that leaves its patch either detours to the crossing structure or dashes straight across the open road. Which one it does, and whether it survives, follows two real road-ecology relationships.
Structure-use probability:
p_use = clamp01(base + 0.60·(W/60) + 0.28·(V/100) − 0.10·(Q/1800))
W = structure width (m), V = vegetation cover (%)
Q = traffic volume (veh/h), base = 0.17 (bridge) / 0.09 (underpass)
At-grade dash survival (Poisson gap acceptance):
T = road_width / dash_speed ≈ 14 / 3.2 ≈ 4.4 s
λ = Q / 3600 (vehicle arrivals per second)
P(survive) = e^(−λ·T)
- Structure controls — width and vegetation cover raise puse, matching field studies (e.g. Banff National Park) where wider, better-vegetated overpasses see far higher use than narrow culverts; green bridges are used more readily than underpasses at the same width.
- Traffic volume — a higher vehicle arrival rate λ both suppresses structure use slightly (noise/light spillover) and sharply lowers the survival odds of any animal that still attempts an at-grade dash, since P(survive) decays exponentially with λ.
- No structure — every crossing is an at-grade gap-acceptance gamble; watch the roadkill counter climb as traffic volume rises.
Drag the map to pan and scroll (or pinch) to zoom — the underlying counts and probabilities are identical wherever you're looking; this is the same logic conservation planners use to size wildlife crossings on new highway corridors.