In the spring of 2016, biologists in Wyoming's Red Desert fitted a GPS collar to a mule deer doe and logged her as Deer 255. That summer she walked 242 miles: north up the sagebrush of the Green River Basin, through working ranches and Bureau of Land Management ground, across highways and under fences, swimming the Snake River, climbing the low north end of the Teton Range, and fording the Fall River in the corner of Yellowstone, to a meadow four miles from Island Park, Idaho. Then her collar failed, and the researchers wondered if it had been a fluke.
It was not. Recaptured in 2018, she repeated the same route, and kept repeating it. By the time she died in the Red Desert in 2024, at ten years and ten months, her collars had logged nearly 3,300 miles across fourteen migrations of the longest documented land migration in the lower 48 states — about 242 miles one way. She never varied the route, only how far up it she went. National park, national forest, private ranch, highway right-of-way: to her it was one route, and her herd is only as safe as the weakest mile of it.
That is the argument running under Ben Goldfarb's Crossings and, before it, James Cheshire and Oliver Uberti's Where the Animals Go. Uberti says it as "to protect animals you must protect where they go." The road-ecology version of that sentence — ours, not a quotation from the book — swaps where for how: to protect wildlife, you must protect how they move.
what a wildlife migration corridor is
A wildlife migration corridor is the specific strip of land a population uses, season after season, to move between the ranges it needs to survive the year. It is not a metaphor and not a general "habitat linkage." For Deer 255 it was a drawable line, narrow in places to a few hundred yards through a mountain pass or a gap between fences, that the herd has learned across generations and cannot easily relearn.
It helps to separate three things people run together. The corridor is the land and water the animals move through. The crossing is the structure, an overpass or underpass or retrofitted culvert, that lets them survive a road cutting that land. Roadkill is what shows up when the route is severed and nothing was built. This series has a post for each: the hub on wildlife corridors as the infrastructure of survival, an explainer on what a wildlife crossing actually is, and the ledger in the carcass is the invoice. This post is about the movement itself, and why the movement is the thing to protect.
three kinds of movement, one requirement
Animals move on three timescales, and each one is a separate way for a fenced-in landscape to fail.
| movement | example | timescale | what breaks it |
|---|---|---|---|
| seasonal migration | Mule deer, elk, and pronghorn between winter valleys and summer high country; caribou; wildebeest | Weeks to months, twice a year, same route | A subdivision on winter range, a fenced highway across the route, an energy field on a stopover |
| daily crossing | A bear between den and berry patch, a bobcat between woodlots, an owl hunting a road verge at night | Hours, every day | A four-lane road with traffic heavy enough that the animal cannot get across alive |
| genetic exchange | A young male panther or mountain lion dispersing to find unrelated mates | A few individuals every few years | A freeway that stops even those rare crossings, so the population breeds with itself |
Seasonal migration is the one people picture: the herd surfing the spring green-up uphill and sliding back down ahead of the snow. Matt Kauffman, who leads the Wyoming Migration Initiative, puts it plainly: deer are migrating for a couple of months, not a couple of days, and they rely on the route itself, not just the two ends. Elk winter range in Colorado's mountain valleys is the same story from the other end of the year; the range is the bottleneck and the route to it is the fuse.
Daily crossing is quieter and kills more. Most of what dies on a road shoulder is not a migrant. It is an animal that lived on both sides of the road and had to get across every night to make a living.
Genetic exchange is the slowest and the one that ends populations without a single carcass. A herd can be fenced into perfectly good habitat and still be dying, because the handful of dispersers who used to carry new genes in cannot arrive. Florida's panthers were down to a few dozen inbred animals before biologists brought in Texas pumas in 1995 to restore the genetic variation isolation had stripped out; the Caloosahatchee crossing picks that story up. The mountain lions penned into California's Santa Monica Mountains by the 101 freeway are the same problem with a different zip code.
Until recently most of these routes had never been drawn. The Global Initiative on Ungulate Migration, hosted by the UN's Convention on Migratory Species, launched in 2021 because most hooved-mammal migrations had never been mapped in enough detail to protect, and it published the first living atlas of those routes in 2024. You cannot protect a line you cannot draw.
the island trap
Here is the failure mode the famous line is aimed at. We protect places we can see: a park with a gate, a preserve with a sign, a scenic valley that photographs well. Boundaries get drawn for politics and views. Animals draw theirs for forage, snow depth, and mates. The two maps overlap less than anyone funding the first one assumes.
Deer 255 summered inside one of the most protected landscapes on Earth and wintered more than 200 miles away on public desert and private ranchland, crossing more jurisdictions along the way than usually plan together for her. Wyoming designated the Sublette corridor vital habitat in 2016, and the collaring itself was USGS, the University of Wyoming, BLM, and WGFD in one room — proof it can be done, not proof it is done. Protect Yellowstone perfectly and you have protected her for about three months a year.
A park without a crossing is an island. An island goes extinct.
| what you fund | what it is | what it protects | what it cannot do alone |
|---|---|---|---|
| park or preserve | A bounded place with a legal designation | The end of a route: summer range, a calving ground, a scenic core | Deliver the animals to it, or let them leave when the season turns |
| corridor land | Easements, fee purchase, fence removal along a mapped route | The route itself: stopovers, pinch points, the miles between ranges | Get an animal across a busy highway alive |
| crossing structure | An overpass, underpass, or culvert retrofit, with fencing that funnels animals to it | Survival at the one point where route meets pavement | Mean anything if the land on either side is developed |
| flyway | A continental aerial route with stopover wetlands | Birds that need to refuel every few hundred miles | Hold together if a single key stopover is drained or paved |
Read the last column top to bottom. Each row is a way to spend real money on wildlife and still lose the population. That is the island trap: funding pieces of a route and calling the route done.
the invisible infrastructure
Goldfarb's contribution in Crossings is to make the barrier visible. By his count there are roughly 64 million kilometers of roads on Earth, enough to circle the planet 1,600 times, with about a million kilometers added every year. Roads are the most extensive thing humans have ever built and the one we look at least. Their effect does not stop at the pavement: noise and light push animals away from a band far wider than the road itself, and a route that is technically open can be functionally closed because nothing is willing to approach it.
The point for anyone holding a conservation budget is that the road network is not a side issue next to habitat loss. It is the mechanism by which habitat that still exists stops working. The engineering answer, and where it does and does not work, is the job of the crossing explainer. This post only needs the diagnosis: the routes were there first, we built through them, and a route that cannot be traveled protects nothing.
why it matters, in order
Genetics. A population that cannot exchange individuals with its neighbors loses variation with every generation, and no amount of protected acreage inside the fence can replace it. Connectivity is what lets a herd or a pride remain itself over a century. This is the intrinsic reason, and it comes first.
Climate shift. As temperatures move, ranges move with them, upslope and poleward. A species that can follow its climate survives; one boxed into a preserve at yesterday's elevation does not. The route is the adaptation mechanism. Fixed boundaries protect the past.
Hunter opportunity. In the American West the migratory herds are the huntable herds, and hunting license dollars and federal excise taxes on gear are much of what pays the state biologists who collar deer like 255 and map where they go. Mule deer across the West are down roughly 40 percent in two decades. A severed migration is a lost season for the people who fund the mapping.
Crash reduction. Federal estimates put large-animal collisions in the United States at one to two million a year. That is the bill humans already pay for severed routes, and the reason some crossings pay back in avoided crashes. It is last on this list on purpose. It is the invoice, not the reason. Price is a bridge to protecting the animal, never a statement of what the animal is worth.
protect the path, not only the postcard
The practical turn is small and it changes what gets funded. Stop asking "what place do we protect" and ask "what route do we hold."
For a land trust or steward, an easement on a mapped pinch point is worth more than the same acreage anywhere else, because it is the mile the whole herd has to pass. For a foundation, the grant that closes a corridor is worth more than the grant that adds a scenic parcel to a park that already has one. For a watershed council or landscape partnership, the route is the reason your jurisdictions have to be in the same room: no single agency owns it, which is exactly why it stays chronically underfunded.
This is where ensurance fits. It treats the route as a present-tense asset with a condition to fund, not a cause to donate to. Two live syndicates split the work the way this post splits it: wildlife-corridor.syndicate coordinates funding for the land, easements and acquisition and fence removal along mapped routes, and wildlife-crossing.syndicate coordinates funding for the structures where those routes meet a road. A certificate funds a named route directly; a coin funds the protocol broadly, with proceeds routed to agents like these. Honest stage: the syndicates are minted, one corridor certificate is live on the hub, and volumes are small. The instruments exist. The routes are waiting on the money.
Deer 255 never asked which jurisdiction she was in. The job is to make sure the next collar tells the same story.
Read next: who pays for a wildlife crossing follows the money that already covers the crash, and where it should go instead. The hub, wildlife corridors: the infrastructure of survival, holds the live instruments. Explore the places already carrying routes on the protocol.
the series
- wildlife corridors: the infrastructure of survival
- what a wildlife crossing actually is
- the carcass is the invoice
- to protect wildlife, protect how they move (this post)
- who pays for a wildlife crossing
- a bridge without the land is still an island