White-Tailed Deer Response to Vehicle Approach: Evidence of Unclear and Present Danger

DeVault TL (2014) White-Tailed Deer Response to Vehicle Approach: Evidence of Unclear and Present Danger. PLoS ONE 9(10): e109988. doi:10.1371/journal.pone.0109988 White-Tailed Deer Response to Vehicle Approach: Evidence of Unclear and Present Danger Bradley F. Blackwell 0 Thomas W. Seamans 0 Travis L. DeVault 0 Elissa Z. Cameron, University of Tasmania, Australia 0 United States Department of Agriculture, Animal and Plant Health Inspection Service, Wildlife Services, National Wildlife Research Center, Ohio Field Station , Sandusky, Ohio , United States of America The fundamental causes of animal-vehicle collisions are unclear, particularly at the level of animal detection of approaching vehicles and decision-making. Deer-vehicle collisions (DVCs) are especially costly in terms of animal mortality, property damage, and safety. Over one year, we exposed free-ranging white-tailed deer (Odocoileus virginianus) to vehicle approach under low ambient light conditions, from varying start distances, and vehicle speeds from 20 km/h to approximately 90 km/ h. We modeled flight response by deer to an approaching vehicle and tested four hypotheses: 1) flight-initiation distance (FID) would correlate positively with start distance (indicating a spatial margin of safety); 2) deer would react to vehicle speed using a temporal margin of safety; 3) individuals reacting at greater FIDs would be more likely to cross the path of the vehicle; and 4) crossings would correlate positively with start distance, approach speed, and distance to concealing/refuge cover. We examined deer responses by quantiles. Median FID was 40% of start distance, irrespective of start distance or approach speed. Converting FID to time-to-collision (TTC), median TTC was 4.6 s, but uncorrelated with start distance or approach speed. The likelihood of deer crossing in front of the vehicle was not associated with greater FIDs or other explanatory variables. Because deer flight response to vehicle approach was highly variable, DVCs should be more likely with increasing vehicle speeds because of lower TTCs for a given distance. For road sections characterized by frequent DVCs, we recommend estimating TTC relative to vehicle speed and candidate line-of-sight distances adjusted downward by (1-P), where P represents our findings for the proportion of start distance by which .75% of deer had initiated flight. Where road design or conservation goals limit effectiveness of line-of-sight maintenance, we suggest incorporation of roadway obstacles that force drivers to slow vehicles, in addition to posting advisory speed limits. - Data Availability: The authors confirm that all data underlying the findings are fully available without restriction. The USDA/APHIS National Wildlife Research Center maintains the policy and practice of formally archiving research-related materials and data at the close of a study specifically for anyone who might wish to revisit aspects of the study or requests from government agencies, non-government agencies, and the public. As a US Federal Agency, the NWRC has the responsibility to make available these materials upon request. Requests for data can be made to Diana L. Dwyer, Supervisory Technical Information Specialist/Unit Leader: . The website for the NWRC library is: http://www.aphis.usda.gov/wps/portal/banner/help?1dmy&urile = wcm%3apath%3a%2Faphis_ content_library%2Fsa_our_focus%2Fsa_wildlife_damage%2Fsa_programs%2Fsa_nwrc%2Fsa_information_services%2Fct_library. Funding: The research was supported by the US Department of Agriculture, Animal and Plant Health Inspection Service, Wildlife Services, National Wildlife Research Center, and the US Federal Aviation Administration (FAA) under agreement DTFACT-04-X-90003. Opinions and conclusions expressed in this manuscript are not necessarily those held by the FAA. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. Animal-vehicle collisions pose not only mortality to the animals involved, but in some cases substantial safety concerns to people and costs associated with collisions. Deer (Cervidae), given their size, general abundance, and association with humans, represent a particularly critical safety problem relative to collisions with vehicles. As an indirect example, based on its claims data State Farm Insurance Company estimates that 1.23 million deer-vehicle collisions (DVCs) occurred in the USA from July 1, 2011 through June 30, 2012, a 7.7% increase over the previous year and representing approximately $4 billion in damages annually (i.e., $3,305 U.S. per incident; https:// www.statefarm.com/about-us/newsroom/2012/10/23/deer-vehicleconfrontations. Accessed 2014 Sep 19). Recent studies have elucidated habitat and landscape-level attributes that might increase the likelihood of DVCs and other animal-vehicle collisions (e.g., obstructive cover proximate to roads [1], [2], [3]; landscape diversity near roads, including lower densities of people [4], [5], [6]). Further, in areas of relatively high human density, but where human presence is essentially non-threatening, wildlife generally show shorter flight-initiation distances (FIDs) [7]. In the context of DVCs, shorter FIDs relative to vehicle approach equate to less time for the deer and driver to react. However, even with a wealth of information on near-road habitat factors that contribute to DVCs, effective reduction of these incidents is multifaceted and at present there is no economically and logistically feasible solution to the problem over large geographic scales. Contributing to the difficulties in managing DVCs and other animal-vehicle collisions is a general lack of understanding about the fundamental causes of such collisions, particularly at the level of animal detection of approaching vehicles and subsequent decision-making [8]. Behavioral response of ungulates to roadbased vehicles can vary relative to age, sex, antler presence/size [9], [10]; traffic volume/speed [10], [11], [12]; vehicle type and noise [13]; habitat features [3], [5], [9], [11], [12], see also [14]; and prior experience (hunting pressure) [15], [16]. We suggest, however, that commonalities in species-specific behavioral adaptations for predator detection and avoidance can provide insight into aspects of DVCs, sensu [17]. For example, white-tailed deer (Odocoileus virginianus) in dense vegetation might flee even when a predator is at considerable distance, likely because of loss of visual contact [18]. In contrast, alert and flight-initiation distances for this species are generally positively correlated with distance to an observed, potential threat, such as the point in which an observer begins an approach [18]. Thus, maintenance of visual contact with a potential threat is key to the white-tailed deers ability to make flight decisions based on r (...truncated)