Factors Influencing the Location of Deer-Vehicle Accidents in Iowa

作者： Michael W. Hubbard , Brent J. Danielson , Richard A. Schmitz

关键词: Logistic regression 、 Demography 、 Poison control 、 Traffic accident 、 Geography 、 Deer–vehicle collisions 、 Accident prevention 、 Injury control 、 Stepwise regression 、 Annual variation 、 Ecology (disciplines) 、 Ecology, Evolution, Behavior and Systematics 、 Nature and Landscape Conservation

摘要: Half a million deer-vehicle accidents (DVAs) occur annually on North American roads, yet the possible factors associated with these are not well understood. We examined influence of landuse patterns and highway characteristics Iowa DVAs from 1990-97. Over 25% occurred at 3.4% all mileposts in Iowa. Annually, greatest numbers May November. 2.59-km2 areas around set randomly selected milepost locations (n = 1,284) to identify that may have influenced DVAs. Stepwise logistic regression produced 6-variable model ($chi ^{2}{}_{6}$ 331.543, P 0.001) included 4 landscape variables, number bridges, lanes traffic. A factor classification tree constructed variables revealed 97% ≥4 bridges were "high" DVA sites (≥14 DVAs). The correctly classified 63.3% 245 validation data set. Our results suggest mitigation should be focused high bridges.