Using traffic data to estimate wildlife populations

Wildlife populations are threatened worldwide by, among others, habitat fragmentation and hunting pressure. An important impediment for the large scale, national and regional, management of the populations is the difficulty to quantify population dynamics. The purpose of this study is to present a tool for such estimations which is based on available data in several countries; traffic load and traffic accidents with wildlife. An econometric model is developed, which accounts for landscape characteristics. It is applied to wild boar in Sweden, for which data on traffic load and accidents for different counties and years are available. Landscape characteristics are introduced with direct or indirect effects on population growth. The indirect landscape model gives the best statistical performance, and the results show relatively small differences in calculated intrinsic growth rate among counties but considerable differences in predicted population developments.