COS 46-2
Reducing wildlife-vehicle collisions by targeting ecological patterns and human behavior

Tuesday, August 11, 2015: 1:50 PM
323, Baltimore Convention Center
Brian A. Crawford, Daniel B. Warnell School of Forestry & Natural Resources, University of Georgia, Athens, GA
John C. Maerz, Warnell School of Forestry and Natural Resources, University of Georgia, Athens, GA
Clinton T. Moore, USGS Patuxent Wildlife Research Center, Athens, GA
Terry M. Norton, Georgia Sea Turtle Center, Jekyll Island, GA

Roads represent a public asset and a pervasive anthropogenic threat to wildlife. Durable management solutions should consider wildlife ecology in conjunction with human behavior in order to mitigate human-wildlife conflicts. Road mortality is well-documented in diamondback terrapins (Malaclemys terrapin) across the majority of their range, including the 8.7-km causeway to Jekyll Island, GA where adult mortality rates are predicted to cause population declines if left unmitigated. From 2009 to 2014, researchers at the University of Georgia and Georgia Sea Turtle Center used intensive road surveying to identify spatial (hot spots) and temporal (hot moments) peaks of terrapin crossing activity and vehicle collisions and exploit these as targets for management. We evaluated impacts of two management strategies on terrapin mortality on the causeway under a modified before-after control-impact design using Bayesian mixed models. The two management strategies were 1) a hybrid roadside barrier designed to prevent terrapins access to the road and 2) flashing warning signage - the first of their kind ever to target turtles - designed to increase driver awareness and reduce the likelihood of vehicles striking an animal on the road.


After placing the barrier at a known hotspot of terrapin crossing in 2011, we estimated a 61% reduction in annual crossing activity compared to no predictable change at two control sites. We installed flashing warning signage at each end of the causeway in 2013 and activated signs for two out of three hours during daily activity peaks around high tide. The probability of mortality given an attempted crossing was significantly lower during periods when signs were flashing (95% Credibility Intervals: 0.29–0.39) relative to periods when signs were present but turned off (0.41–0.54) and prior to installation (0.52–0.59). Our findings support the use of complimentary strategies that target wildlife and drivers at road-impacted sites throughout the terrapin’s range, and our approach serves as a model for investigating patterns of vehicle collisions and designing management based on ecological targets of other species. We will incorporate these predicted effects into a conservation decision-making framework to mitigate road impacts on terrapins while satisfying drivers.