Around the
summer of 2003 in the La Sal Mountains of Utah, mule deer began to turn into
zombies.
Or, at
least they began to act like zombies. They started losing weight, salivated
constantly, and began to walk in listless circles. They grew apathetic and then
stopped running from humans.
At first
only a few sick animals turned up in annual surveys of harvested deer, but the
numbers grew. Testing confirmed the fears of wildlife managers, Chronic Wasting
Disease or CWD. The prion disease produces lesions in the brain that change the
animals' behavior. “We call them zombie deer,” says Martha Garlick, PhD, SD
Mines math professor.
At first
CWD shows no symptoms. It progresses over the course of a few years, but once
contracted it’s always fatal. CWD is highly contagious and it has ravaged deer
and elk populations across the American West.
Understanding
the rate of spread is crucial to stopping any disease. This is where Garlick’s work comes
in, she is teamed up with wildlife biologists, mathematicians, and
statisticians at Utah State University and Colorado State University. The team
is part of a National Science Foundation grant to improve computer models that
can help predict how animal populations move.
“I love math anyway, but, it’s
really cool to actually apply this to something real world. It’s exciting to
predict things about animal movement that will help wildlife managers who care
for these populations.”
Garlick loves to hike and spend
time outdoors but doesn’t spend her days chasing deer and elk in the field,
rather she relies on the rich data sets accumulated from years of GPS collars
that have been fitted to wild animals.
She uses GPS tracking records,
combined with landscape images to build models of how animals travel across
various landscapes. Deer, for example, can move at different rates on rocky
slopes than they can in thick trees, or open grasslands, or mazes of cedar
brush. Researchers use the animal movement data to assign a number to different
landscapes. Animals tend to take the easiest path to resources, like green
grass and water, and they are often confined by barriers like a fence, a
mountain snowline, or a major river. There are many variables to consider when
building the algorithms that can accurately model something so complex. For
example, male deer move in different patterns than females, so they require
some slightly different math.
One of Garlick’s favorite parts of
the job is seeing students get involved in this research. “It’s fun to get
undergraduates excited about this work,” she says. “In math modeling, it’s
sometimes difficult to actually make it fit. You try and keep it simple but
putting in the details so that it mirrors what’s really going on, this is an
exciting challenge.”
The
models Garlick is developing not only have applications in tracking deer and
elk in the mountain west but can also be used to predict the movement of harbor
seals in Alaska, or the spread of the wolf populations introduced into places
like Yellowstone National Park. As the research continues and the models inch
closer to predicting reality, the real-world applications of this work will
continue to grow.