Abstract 198 – Animal learning contributes to both problems and solutions for bear–train collisions

Colleen Cassady St. Clair

Transportation infrastructure frequently causes an ecological trap for bears that are attracted
to roads and railways for foraging and travel opportunities, but these attractants increase the
risk of mortality from collisions. This situation occurred for a threatened population of grizzly
bears (Ursus arctos) in Banff National Park, Canada, where train strikes became a leading cause
of mortality. Several students, collaborators, and I explored this problem by studying rail-
associated food attractants, habitat use of GPS-collared bears, and patterns of past mortality.
Bears appeared to be attracted to grain spilled from rail cars, enhanced growth of adjacent
vegetation and train-killed ungulates with rail use that increased in spring and autumn, and in
areas where trains slowed, topography was rugged, and human density was low. Across
multiple species, mortality was best predicted by higher train speeds, with contributions from
track curvature and proximity to water bodies. Rapid learning by bears may have increased
collision vulnerability via reductions in lethal bear management, changes in ungulate
distribution and abundance, and changes in human activity, but that same learning capacity
might prevent train strikes in future via simple warning devices, such as the one we invented,
that signal approaching trains. Emerging technology associated with electric vehicles might also
contribute to better safety by amplifying and directing engine noise as an audible warning to
reduce strike vulnerability for bears and other wildlife.