Out of the way of the slow-moving pronghorn, we made camp at the edge of a small hill and sat on our sleeping pads to watch them. When dusk fell, much to our surprise, they retreated back down the trails they had made, splashing through Wagon Creek and cantering along the bare sage slopes below. They were returning to the safety of the valley floor for the night. Like road crews clearing a highway, the pronghorn punched trails through snowdrifts all day and then returned at sundown to the lower elevations.

Hall Sawyer, a biologist for the consulting group Western Ecosystems Technology Inc., has been following the Teton pronghorn and the mule deer that share their winter range for over a decade. He's made a startling discovery. "In the last few years, we're recognizing the internal anatomy of migration routes," he says. Migration corridors are less like a conveyor belt where animals get on at one end and off at the other than they are like a chain of hotels linked by strips of highway. The migrating mule deer that Sawyer studied spent 95 percent of their time at the "stopover points" (the hotels) where they rested and foraged, and just 5 percent of it zipping through the "movement corridors" (the highways).

The maps he's generated resemble a rosary: strings of beads with spaces in between, the bulbous stopover locations linked by narrow movement corridors. They offer conservationists a new lens for looking at ungulate migration, important for figuring out how to protect it. Sawyer says conservationists could work on protecting ungulate stopovers the way they've protected refuges for migrating waterfowl through the Midwest.

"The best-case scenario would be no development in migration corridors," Sawyer says. Today, however, no migration corridor in the U.S. is protected from end to end, and development is already slated for some lands where ungulates migrate. Distinguishing stopover points from movement corridors could help managers minimize migration hindrances when they decide where to locate pipelines or wind turbines. "There will be some development. Given that choice, it's better to put it in the movement corridors than the stopover points with the qualification that it doesn't impede migration," Sawyer says. "We need to maintain connectivity."

His finding matched what Jake and I were seeing: The pronghorn spent only a little time navigating the trickiest parts of the corridor, such as these snow-choked slopes. Whenever possible, they camped out in the drier lowlands along the Green River.

The following afternoon, Jake and I worked our way another couple of miles north through deep snow to the pass. We settled down on a dry patch of sagebrush to wait for the pronghorn. Jake napped, but I buzzed with excitement. For the first time during my four seasons following the migration, I was ahead of the animals as they neared the highest point in the corridor, the gateway to the summer range on the far side. A mile to the south of our sunny resting place, 80 pronghorn dozed on a rise. I watched them through binoculars, wondering whether they would try to push higher this afternoon or retreat to the Green River, just as they'd done each of the past several nights.

Finally, around 4 p.m., as if a drill sergeant had shouted a command, the pronghorn jumped to their feet and began testing the snow at the edge of the sagebrush. There is nothing casual about migration. The animals were organized, methodical and efficient. They moved in silence, but with perfect coordination. A doe kicked a trail across a drift to reach a long strip of dry land that paralleled the edge of the forest, leading toward us. The rest of the herd lined up to follow her.

Within moments, the first pronghorn were hurrying by just 30 yards from where we crouched in the shadow of a ponderosa. The mass of tan-and-white bodies whispered between the tree trunks. Their hooves crunched through the snow, and they panted open-mouthed. The leaders broke trail up a drift through an open gate and into the beetle-killed pines at the crest of the divide. The rest of the herd streamed behind them, more than 80, perhaps as many as 150, half of the Teton population.

For half an hour we sat frozen in place, watching the quiet flow of animals as they scurried along the edge of the forest. After the last white butt disappeared into the trees, Jake and I looked at each other in amazement.

"I can't believe we just saw that," I whispered.

"It's like the Serengeti, seeing so many animals move past at once," Jake replied.

Nothing remained of that surge of life except for the narrow paths where pronghorn hooves had punctured the snow, like rows of stitches linking the desert in the south to the summer range in Teton Park.

Emilene Ostlind was a High Country News editorial fellow in the winter of 2011 and now works as a freelance journalist in Lander, Wyo. This story was supported by contributions to the High Country News Enterprise Journalism Fund.