Motorways induce various impacts on neighbouring bird populations. Mortality of birds by traffic is one of the most important impacts. To measure mortality and to evaluate factors which influence it, it is necessary to take into account biases affecting carcass surveys. Study area: motorways A10 Virsac - St-Jean-d'Y, A837, A89 Périgueux, A64 (eastern Toulouse) (total length: 261 km). Carcass surveys (capture-mark-recapture) were conducted by car and short sections were sampled by foot. Analyses showed that taxinomic orders and carcass weight, age of carcass and season (i.e. scavengers activity especially high on spring) influence carcass surveys. Barn Owl and some Turdidae species were the most killed by traffic. After having measured density of Passeriformes living close to the motorway sections, we analysed the following biological factors (GLM): density, diet, wing shape, learning ability, migratory/sedentarity, fear Flying Initial Distance (FID). Results: FID was the only factor, with density, that influenced bird mortality. We also studied (GLM) following environmental factors: motorway transversal profile, traffic intensity, land cover crossed by the motorways sections and verges vegetation layers. Results: Strigiformes were more killed on motorways sections with combined verges vegetation layers, they were less killed on sunken profiles such as Passeriformes which were more killed on verges with trees. Efficient mitigation measures that should reduce mortality for both Strigiformes and Passeriformes taxons are thus sunken profiles for at least earth acoustic berms and cutting or avoiding planting trees on hotpoints of bird casualties.
