Analysis and valorisation of multi-sources data for fauna road casualties modelisation
Avoiding, reducing, and mitigating for the risk of collisions between transport vehicles and wildlife has become a major issue in a context of increasing artificialization of landscapes, fragmentation of natural environments, and intense development of infrastructure networks and traffic (CGDD/SoeS, 2016) in both industrialized and developing countries (Schmidt & Zumbar, 2008; Jones et al., 2015; Laurance & Balmford 2013). Road traffic killsevery year several hundred thousand, even millions of animals, all species combined, on French roads (Glista et al., 2009; van der Ree et al., 2015; Rytwinski et al., 2016).
The work proposed in this exploratory project is part of this objective. Various initiatives and projects have started in France (Cerema: Bretaud, 2013; Guinard, 2013; UMS PatriNat/MNHN: Billon et al., 2014 and 2015; Berlioz, 2018, both institutions on the COMERCAR exploratory project) and internationally (see the many examples of work in the proceedings of the IENE 2016 and 2018 symposia) to work and analyse, according to a standardised protocol, road mortality data collected along roads. In France, these actions have been deployed, for the most part, in partnership with the DIR network and a few departments and/or regions.

At the same time, a large amount of opportunistic data was collected by the associative networks of local naturalists, implemented, for example, in the Visionature tool (LPO/Faune France). A first analysis of the influence of environmental factors and road and traffic profiles was attempted in 2018 (CEFE/LPO AuRA: Pérard, 2018).
Methodological and technical difficulties remain in all theses works, which are based on radically different principles and methods even though it examines the same problem. This exploratory project has the ambition to unite them by establishing a partnership between the actors of these two types of approach.

The objective of the work proposed in this call for projects is to identify the complementarity of opportunistic road-kill data (association volunteers) and protocol data (DIR/CD patrol boats) in order to understand and predict spatial patterns of wildlife mortality on the road. The specific objective is to statistically exploit these data sets from multiple sources in a study area already identified as a test area. The medium- and longer-term objectives are to better understand which road profiles, vegetation of verges associated with biological traits and environmental factors (specific habitat, adjacent land use types) modulate wildlife mortality in order to determine and propose local solutions for limiting effects of mortality and/or mitigation measures on the major road-kill location and sections. Different types of data with traffic intensity, neighbouring lanscape" at the local, regional and even national levels, and will eventually lead to the establishment of an appropriate strategy for their acquisition for all ILTEs.

Through this project, it is necessary to take into account different vertebrate species and taxa. However, depending on the size of the individuals crushed, the more or less carcass surveyand the ability to identify, the density along the roads (Fahring & Rytwinski, 2009), birds (large species including raptors) and several mammal species (hedgehog, marten and large ungulates) will be prioritised.