Almost four years ago, when Aaron Mrotek became manager of The Nature Conservancy Patagonia-Sonoita Creek Preserve, he was startled by the amount of roadkill he kept seeing along State Highway 82.
His first thought was “somebody ought to do something about all this killing,” but some months later, as the pace of deaths continued along the highway, a flash hit him: “That someone is me.”
What is needed to cut vehicle-animal collisions, Mrotek was clear, is data about these roadkills, lots of it. Perfect for citizen science, he thought. Local wildlife biologist Eric Herman, was excited about the idea. To provide a protocol for data gathering, Herman drew on his extensive professional work doing wildlife surveys and organizing data.
His protocol first specifies the “path of interest” as a 22-mile portion of Highway 82, from Sonoita’s crossroads to the Nogales International Airport. The protocol goes on to detail the process of conducting surveys, equipment needed, safety measures and data to be collected. For systematically archiving and geolocating the photos and data that road surveys would generate, Mrotek recruited Borderland Restoration Network (BRN) computer whiz Randy Moore to develop an app for the surveyors to install on their phones.
These ideas evolved into a two-year research program in what’s been called “ road ecology.” Launched in June 2025, the data gathering phase was preceded by recruiting and training a number of volunteer “Roadkill Rangers.” In teams of two, they go out on the highway, survey the whole path of interest, stopping as safely as possible at every death location to record info about it.
Once at a death site, the surveyor photographs the carcass, sends a shot to the survey app, then answers the app’s questions about the creature’s species (if detectable) and estimated stage of life – whether mature or not. From the phone’s geolocation info, the app notes time and date and then pinpoints the coordinates of each stop. This info will allow mapping in the future.
Once the photo and data are entered, a member of the team removes the carcass from the roadway, ideally to about 50 feet from the centerline, but topography and brush sometimes prevent that. A bright orange spot of paint is a sprayed on the body, just to let future surveyors know that particular body had already become data.
The Arizona Department of Game and Fish Department has given the program a permit to move dead animals off the roadway. Eric Herman will use the roadkill data to send the agency an annual report, under the terms of the permit.
At the end of a trip, the gathered information is sent to the cloud as ready for entry on a growing Excel spreadsheet. Net result: the Roadkill Ranger survey process transforms those bodies of flesh into bodies of data.
What has the first year of roadkill surveying produced? In an August 2026 online presentation for volunteers and staff, Mrotek opened with a summary and a map of the findings from a major earlier study by Paul Beier of Northern Arizona University, an environmental scientist who has studied wildlife corridors and habitat connectivity. Beier’s research sought alignments of animals movements between the Santa Rita Mountains and ranges to the southeast, routes which cross 82 in the project area. Beier identified three major areas of possible wildlife crossing, and Mrotek’s preliminary analysis showed fairly decent alignment with two of Beier’s “clusters.”
Mrotek delved into many subtleties of the data being gathered, a major one being time of year, with three times the roadkill incidents being reported in the summer as in the winter. Some parts of the road are safer than others, Mrotek explained, such as long straight stretches with good vision and vegetation management. Some species of roadkill might be carried off by predators. And vultures relate to roadkill in several ways: one was seen carrying a snake from centerline to small trees near Patagonia High School; another was hit and killed while feasting on a roadkill creature, and still others will sit on the road eating but fly away safely if a vehicle comes by.
With all this variability in conditions and incidents, Mrotek reminded the group that the program records roadkill incidents but is not a census. Mrotek’s suggestion is that the purpose of the field work is for providing data about where crossings actually occur, not for estimating the numbers of various species. And it doesn’t count the lucky ones who made it across.
In the project’s database, the species officially listed are explained as “simplified” species, because there may be several species of skunk, for example, and the species ID process would have big challenges. Whitetail vs. mule deer is a familiar example of easy-to-identify subspecies, but in entering roadkill data, both are simply accounted as deer.
The data of the first year identify 22 species, plus 18 unidentified carcasses. The five most prevalent were skunk (41), rabbit (31), deer (22), snake (19) and coatimundi (18).
For the first year, there were no bears or pronghorns, nor were there any of three regionally known species of big cats: mountain lions, ocelots and jaguars.
Smaller creatures that might get hit or squashed on 82 include rodents like mice and packrats, amphibians like toads and frogs, arachnids like tarantulas, and reptiles such as snakes and lizards. All species seen are important indicators of biodiversity, but except for the amphibians, the tabulated local reports don’t include these smaller characters.
The ongoing second year of the project will finish in June of 2027. A footnote is that because of the ongoing pavement project for the “south reach” of the survey area, Mrotek recently stopped the roadklll data gathering in that half until resurfacing concludes in mid-November.
After the survey period finishes, all reported death sites will be mapped and the whole survey length analyzed for density of roadkill, clusters of data that are hotspots for designing future wildlife crossings of the highway. The first two years of photos and data will be woven together into a larger, more nuanced picture of the area’s wildlife distributions, habitat, water sources, and migration patterns.
The final quilt of understandings will include proposals for regional wildlife care and habitat management. They will be shared with national experts in wildlife habitat, corridors and the design of safer highway crossings. For building safer wildlife crossings, what comes next in a years-long process would be funding, stakeholder inputs, detailed design, permitting and construction.
Once a wildlife crossing is put in use, there would be continued monitoring of the “users,” and significant reductions in roadkill are expected to result.
Mrotek knows he’s tasking the Rangers with difficult work, and frequently praises their wilingness to do a sometimes messy, even gory, project. “Wouldn’t have happened without them,” he said.
What’s the experience like for the Roadkill Rangers, the ones who find and tally the deaths which result from all that rubber rolling down the road? Since there is a lot of guts on the pavement, the right adjective for experiencing these deaths, whether dried or fresh, big or small, is “visceral.”
In a conversation with a local death doula, Holly Stefanski, she pointed at how the roadkill effort can also be understood as part of one’s own spiritual path. When a human dies slowly, Stefanski’s work as a death doula is to hang out with them through that process, supporting for days, weeks or even months with her comforting presence and bits of assistance.
She spoke of the timeless point when the spark stops. “Later, you come back to regular consciousness with an ever-deepening gratitude for the miracle of life.” Next day, in an email continuing the thought, she wrote “Every life fragile, fleeting and yet so very crucial to the whole. My hope is that when we encounter the death of any creature, we hold it in a space that reminds us that ALL of us matter because in the end that is precisely what we have been and forever only be.”
Author’s note: This writer’s involvement in this story began by volunteering, rather than being on assignment for the PRT. Along with my partner Carolyn Shafer, we joined the program and got trained as Rangers. In the last year, we’ve done a dozen surveys and added in a pair of extra “incidental” reports – for a deer and a javelina.
