Tropical forest roads drive clearing for kilometres, while health effects go unstudied

How roads affect forests

The review included 65 studies: 19 on forest loss, 47 on ecological effects and six on mitigation. Most came from Malaysia, Indonesia and Thailand.

Nine studies found greater forest-cover loss closer to roads, where road density or length was higher, or after construction or completion; no study found less loss. In New Guinea, loss associated with secondary roads extended as far as 17.7 kilometres. Structural edge effects reached 34 metres at a logging road and 350 metres at a highway, while soil effects remained largely on the road surface.

Reptiles or amphibians led roadkill counts in all seven surveys that recorded every vertebrate group. Deaths were concentrated on short road sections near water. Paved roads reduced crossings by elephants and gibbons and were linked to genetic differentiation in some species. Mammals and a large snake used crossing structures, but these findings came from single spatial comparisons. No study reported a human health outcome; fire, landslides and water quality were not measured.

Why the distance matters

The findings indicate that roads affect tropical forests at two different scales. Access-related clearing can extend for kilometres, while forest-edge changes, wildlife deaths and barriers to movement generally occur over tens to hundreds of metres. The review therefore argues that assessments and mitigation should match the scale of the specific effect, and that measures intended to reduce harm need to be tested before and after they are introduced.

Evidence and gaps

This was a systematic review of English-language journal articles published from 2015 onward, identified through Scopus, PubMed and OpenAlex. Two reviewers independently screened the records and full texts, and the evidence was assessed with established appraisal and certainty frameworks. The review included 65 studies, but the evidence was uneven: forest-loss and ecological findings generally had moderate confidence, while evidence for mitigation rested on single spatial comparisons and was rated very low. Most studies came from Malaysia, Indonesia and Thailand. The review found no human-health studies and no measurements of fire, landslides or water quality, so it cannot establish the effects of roads in those areas.