WILD EARTH WATCHUnderstanding Nature Through Evidence
Home / Threats / Roads & Infrastructure
Threat Profile

Roads & Infrastructure

Roads fragment habitat and kill wildlife directly — an estimated one million vertebrates a day in the U.S. alone, from a road network covering 60 million kilometers worldwide.

Published May 2026 Last reviewed July 2026 Evidence level Strong Reading time 6 min

Overview

Roads, pipelines, transmission corridors and other linear infrastructure fragment habitat, act as barriers to animal movement, and kill wildlife directly through collision. Roads alone are estimated to cover about 60 million kilometers globally and affect roughly 20% of total U.S. land area through their direct and edge effects.

Established fact

An estimated one million vertebrate animals are killed by vehicles every day in the United States. A global roadkill dataset spanning 1971–2024 across 54 countries recorded mammals as the largest share of documented roadkill (61%), followed by amphibians (21%), reptiles (10%) and birds (8%).

Source: Scientific Data (Nature), Global Roadkill Data, 2024

Documented Impacts

Direct mortalityRoadkill is the single greatest source of mortality for 28% of studied mammal species — in some cases exceeding all other causes of death combined.
Barrier effect and fragmentationInfrastructure blocks dispersal and recolonization, isolating subpopulations from the wider metapopulation — a slower-acting but often more consequential impact than direct mortality.
Secondary pollutionRoads contribute light and noise pollution, road-salt and vehicle-derived water contamination, and tire-derived microplastics and chemicals into adjacent habitat.
Established fact

Global estimates suggest hundreds of trillions of insects, including billions of pollinators in North America alone, are killed on roads annually — a scale of mortality that is easy to overlook because individual insect deaths aren't tracked the way vertebrate roadkill is.

Source: bibliometric synthesis, Urban Ecosystems, 2025
Editorial analysis

Mitigation exists and works: well-designed wildlife crossings (culverts, overpasses) can reduce roadkill by 80–95% where properly sited. The gap isn't a lack of solutions — it's that crossings are frequently built without rigorous, species-specific evaluation of whether a given design actually works for the local wildlife, which means some infrastructure marketed as a conservation measure may not deliver much benefit in practice.

Uncertainty & Evidence Gaps

Vertebrate roadkill counts are reasonably well-documented in regions with active monitoring programs, though global insect mortality figures are far rougher extrapolations. The effectiveness of specific wildlife-crossing designs varies by species and site, and is under-studied relative to how widely such crossings are now being built.