Roads are more than infrastructure. They have far-reaching impacts on the environment, reshape landscapes, and kill millions of animals every year.
With more than 21 million kilometres worldwide — roughly one million in France alone — roads cover 2% of the national territory, and the network keeps growing. Their effects on wildlife extend well beyond their physical footprint: habitat destruction, landscape fragmentation, traffic noise, light pollution, heavy metals, and contaminated runoff reaching waterways...
Ecologist Richard Forman has described road networks as the sleeping giants of environmental science.
Forman & Alexander (1998) · Spellerberg (1998) · Meijer et al. (2018)European roads are estimated to kill around 29 million birds and 192 million mammals every year, though these figures do not cover all species. Estimates exist for other parts of the world (Brazil, the United States, Canada), but none has yet been produced at the scale of France.
These global figures conceal widely varying realities: some species lose less than 1% of their population on roads, others more than 10%. Mortality is often selective. For many species, collisions disproportionately affect young individuals or those dispersing, which can disrupt the demographic structure of a population even when the total number of casualties appears low.
Grilo et al. (2020) (Europe) · González-Suárez et al. (2018) (Brazil) · Loss et al. (2014) (United States) · Bishop and Brogan (2013) (Canada) · Moore et al. (2023)Collisions are not random. They cluster at spatial hotspots — areas of wildlife presence and movement corridors — and follow temporal peaks: spring amphibian migrations, autumn deer rutting season, the nocturnal or crepuscular activity of most mammals...
What makes prediction difficult is that spatial and temporal factors do not act in isolation; they interact with the biological traits of each species: population density, movement speed, ranging behaviour, reactions to vehicles... It is the combination of these three dimensions — space, time, and biology — that determines collision risk, and this interdependence explains why findings from one study do not always replicate in other contexts.
Pagany (2020) · Grilo et al. (2020) · Gunson et al. (2011)The most effective solution remains the combination of wildlife crossing structures (underpasses, amphibian tunnels, green bridges, etc.) with guiding fences that direct animals towards them. A recent meta-analysis shows mortality reductions of up to 83%, provided these structures are placed in the right locations. Speed reductions in sensitive areas have also produced conclusive results in several studies.
Beyond these direct interventions, findings are more disappointing: most chemical repellents, reflectors, and warning signs have failed to demonstrate reproducible effectiveness. New approaches are under investigation, such as acoustic deterrents, enriching underpasses to encourage amphibians to use them, and detection systems for large fauna on roads.
Rytwinski et al. (2016) · Denneboom et al. (2021) · Glista et al. (2009) · Testud et al. (2020) · Druta and Alden. (2020)