Highways cut habitat into pieces and kill animals that try to cross. Wildlife crossing structures address both problems, and their design follows from how different species behave at a barrier.
The road is a genetic barrier as well as a hazard
Collisions are the visible cost, but the larger ecological effect is separation. Populations on either side stop mixing, and small isolated groups lose genetic diversity over generations.
Seasonal movement is also disrupted. Many western species migrate between summer range at elevation and winter range in valleys, and a highway sits directly across that path.
A crossing structure is therefore built to restore movement, and reduced collisions are a benefit that follows from it.
Overpasses and underpasses suit different animals
Large ungulates such as elk and deer generally prefer open sightlines and resist entering confined spaces. Wide vegetated overpasses suit them because the animal can see the far side before committing.
Predators and smaller mammals often accept culverts and underpasses, and some actively prefer cover. Structure choice follows the target species rather than engineering convenience.
Amphibians and reptiles need small dedicated passages with damp conditions, since a dry concrete box is an impassable desert at their scale.
Fencing is what makes a crossing work
An animal will not find a structure by chance across miles of open shoulder. Continuous fencing runs along the highway to guide movement toward the crossing point.
Where the fence ends, animals simply walk around it, which concentrates collisions at that spot. Fence length and end treatment determine most of the project's effectiveness.
Escape ramps are added so an animal that gets inside the fenced corridor can leave rather than running the road.
Animals need time to adopt a structure
Use is low in the first seasons after construction. Animals are wary of new features and of the disturbance left by building work.
Adoption grows as individuals learn the route and, for herd species, as that knowledge spreads through the group. Monitoring cameras are used to track the change.
Judging a crossing by its first year understates it, which is why evaluations run for several years before conclusions are drawn.
Where they get built
Placement is driven by collision records and by movement data from collared animals, since both point to where the road intersects an established route.
Cost is high enough that structures are usually built during a scheduled highway reconstruction rather than as standalone projects. The timing of road work often decides the timing of the crossing.
That dependence explains why crossings appear in clusters along recently rebuilt corridors rather than evenly across a state.