Wildlife crossings look like simple bridges and tunnels, but the engineering choices follow animal behaviour closely. A structure placed without regard to that behaviour is usually ignored.
Placement matters more than structure
Animals cross roads where they were already moving. Planners therefore begin with collision records, tracking data and terrain rather than with convenient construction sites.
Movement concentrates along river corridors, ridgelines and the seams between habitat types. A crossing on one of those lines is used heavily; one placed a short distance away may not be found.
This is why retrofitting an existing highway is harder than building a new one. The road alignment is fixed, and the ideal location may sit under a bridge that already exists for other reasons.
Overpasses and underpasses suit different animals
Open-country species that rely on sight lines, including deer, pronghorn and many large grazers, prefer wide vegetated overpasses where they can see across before committing.
Forest species and many carnivores accept underpasses readily, and small mammals, amphibians and reptiles use culverts that would be useless to anything larger.
Dimensions follow from that. An underpass that feels like a tunnel to a bear is a comfortable route for a bobcat, so a single road may need several structure types.
Fencing is what makes a crossing work
The structure is only half the design. Without fencing, animals continue to cross wherever they reach the road, and the crossing sees light use.
Fencing runs for a considerable distance either side and funnels movement toward the opening. Where the fence ends, collisions concentrate, so the ends are tapered or tied into natural barriers.
Escape ramps are added so animals that get onto the carriageway can leave it. Jump-outs are one-directional, allowing exit without providing entry.
How use is measured afterwards
Cameras and track beds record which species use a structure, how often, and whether individuals cross or merely approach and turn back.
Early monitoring often shows hesitation, particularly among older animals accustomed to the road. Use typically rises over several years as younger individuals learn the route.
That lag matters for evaluation. Judging a crossing in its first season understates it, which has led to structures being written off before they were fairly assessed.
Why crossings cost less than they appear to
Crossings are expensive to build, and the comparison that justifies them is not against zero but against the recurring cost of collisions.
Vehicle damage, emergency response, insurance and human injury accumulate on a busy road segment year after year, and large-animal strikes are the most serious of these.
On roads with high collision rates, the avoided costs can approach the construction cost over the life of the structure, which is the argument that persuades transport agencies rather than conservation ones.