Predicting mule deer migration paths without GPS collars

How do researchers understand where big game animals migrate across vast landscapes each spring and fall? That’s the question biologists at the University of Wyoming and the Idaho Department of Fish and Game asked in a study published in the journal Methods in ecology and evolution.

Recent advances in technology have enabled biologists and wildlife managers to track ungulates such as moose and mule deer using GPS collars that reveal the animals’ migration routes. However, collars are expensive and logistically difficult to deploy, making it difficult to make a comprehensive inventory of the corridors required by the herds.

Now a research team has found a promising way to predict where mule deer are likely to migrate without having to collar new animals.

“We were surprised at how well we could predict most deer movements, suggesting that migratory mule deer appear to follow rules that balance the cost of movement with the benefits of forage access, rather than random movement,” says Tristan Nuñez, who led postdoctoral work at the Wyoming Cooperative Fish and Wildlife Research Unit of the US Geological Survey at the University of Wyoming.

Nuñez addressed a question that biologists have long pondered: can we predict migration corridors in areas without animals with GPS collars, using the information they have already learned about the habitats through which tracked animals migrate?

Previous research has relied on GPS data to map migratory corridors, which has proven to be a powerful tool for science-based management and conservation. However, Nuñez and his co-authors hoped they could identify the migration paths based on environmental information or habitat quality alone.

To answer this question, the research team first created models that estimated the movement of a herd of mule deer based on terrain, snowmelt, the intensity of human development and new grass growth. They then compared these predicted corridors to the actual migration routes of 130 mule deer from three GPS-collared herds in Idaho and Wyoming.

The team’s models acted like a wayfinding app on smartphones, identifying the best route for navigating between two points. For mule deer, however, the predicted corridor that best matched actual spring movements was not a straight line or shortest distance between their summer and winter ranges. Instead, the deer generally preferred routes with hilly terrain, shrubby vegetation, and less human development.

“The ability to predict migration routes at an accurate scale will save a tremendous amount of time and money, and ultimately be more useful for wildlife management in Idaho,” said Mark Hurley, co-author of the study from the Idaho Department of Fish and Game.

Traditionally, wildlife managers have relied on GPS data from collared animals to define migratory routes critical to healthy ungulate populations. However, detailed knowledge of seasonal migrations depends on years of time-consuming and expensive data collection. “Even after we’ve collared thousands of animals, we’ve probably only fully described a fraction of the migratory routes across the state,” says Hurley.

In recent years, biologists and wildlife managers have used GPS-collared datasets to map the migrations of more than a hundred herds of ungulates across the western United States.

The work, along with the current study, was made possible by a partnership known as the Corridor Mapping Team. The team was formed in 2018 by the US Geological Survey in response to Secretariat Order 3362 to promote the mapping and conservation of migratory corridors for ungulates.

State wildlife agencies, tribes, and federal researchers use GPS motion data to map migrations and aid in regional big game management and conservation. The predictive migration models are an alternative to costly and labor-intensive animal tracking that benefits the conservation of migrations that have not yet been mapped.

Ungulated animals such as mule deer migrate across the western United States each spring and fall, in line with environmental cues tied to food. But as the human footprint expands in the West, migratory herds increasingly face obstacles such as new subdivisions, energy development, impermeable fences and busy roads on their long journeys.

The new computer algorithms used to predict migration routes will be freely available, allowing wildlife managers in areas where herds have never been collared to better understand where important migrations might occur. Mapped migrations can then show where to make fences more deer-friendly, avoid subdivisions, or build flyovers to ease passage over busy highways and keep large areas open to ungulate migration.

Working with members of the Corridor Mapping Team, Nuñez hopes to extend the work to other big game species, herds and landscapes in the western United States and beyond to better understand how their movement preferences vary. In addition, he says, the models can help shed light on how climate change will affect ungulate migration in the future.