AI-generated photorealistic illustration · Created by Sellect Daily (AI); not an event photo.
Dear anyone who has looked at a wildlife bridge and wondered whether the animals got the memo: there is some lovely evidence that they did.
Did the animals actually cross?
In an Arizona US Highway 93 case study, cameras recorded 15,134 crossings by nine species over four years. That is measured use of crossing features, not a prediction from a bridge rendering.
The hopeful part has a monitoring record
The U.S. Fish and Wildlife Service’s published case study reports that desert bighorn sheep made up 94 percent of the animals documented at the structures. It also reports sheep-vehicle collisions falling 97 percent from the preconstruction figure. These are results from that particular monitored project, described in a case study cited as 2018—not a nationwide promise or a new 2026 count.
There is something cheering about an improvement that comes with a record of what happened afterward. The story doesn’t have to end at the ribbon-cutting. In this case, the useful follow-up was animals continuing across the highway corridor to reach resources, with fewer sheep-vehicle collisions reported.
A hopeful design becomes a stronger story when someone checks what uses it.
Read the happy number without making it bigger
A crossing count is an activity count. Imagine one sheep going over a bridge, then coming back later: that would be two crossing events involving one animal. This is an invented counting example, not a reconstruction of the Arizona observations. It shows why 15,134 crossings should not be rewritten as 15,134 different animals saved.
In our fictional tabletop scene, Tori moves one sheep token across a model bridge. Mush makes one tally. Tori brings the same token back; Mush adds another mark but leaves the animal-token count alone. One sheep, two trips. The little commuter has been busy.
The collision percentage answers a different question: how did the recorded collision figure change relative to the project’s earlier figure? It doesn’t count individual animals using the bridge, and it doesn’t guarantee the same reduction somewhere else. Keeping those measures separate lets the encouraging result stay encouraging without turning it into a miracle.
For this case study, keep three labels attached: Arizona US93, the reported monitoring period, and the measure being counted. Crossing events and collision reduction are different results.
Mush keeps the two tallies apart. The sheep doesn’t need a second identity just because it came home.
An early deer sighting is good news, too
A separate Fish and Wildlife Service story dated March 15, 2024 describes Idaho’s Cervidae Peak overpass. It reports mule deer on trail cameras only weeks after completion. The story also describes an expected collision reduction, which is a forecast in that report, not a measured outcome we can quietly promote to a result.
That’s a useful comparison when reading a new crossing announcement. A camera sighting can establish that an animal used a structure. A longer monitoring record can answer further questions about use and collisions. Neither needs to pretend to be the other to be worth sharing.
The Idaho report explains another helpful detail: fences can keep animals away from traffic while also obstructing migration. A crossing provides a route through that barrier. It also warns that some animals may still cross the road, so an overpass is no reason for drivers to stop being careful.
Keep one small question for the next update
Our earlier Los Angeles crossing story looks at what a habitat connection is designed to do. These Arizona and Idaho reports let us ask the follow-up: what was actually observed after construction? They are different projects, with different evidence—not results for Los Angeles.
When the next wildlife-crossing photograph appears, look for the place, the observation period and the unit being reported. A deer on camera is a welcome beginning. A careful follow-up can tell an even better story. Mush is ready with the tally sheet; Tori has promised not to count the same commute as a new sheep.
Sources: US93 completed-project case study, cited as 2018, and Cervidae Peak report, March 15, 2024. Checked November 3, 2026. Historical reported outcomes remain dated; no claim of fresh fieldwork or a current nationwide total.