Universal dose–response sensitivity of ecological processes to low levels of artificial light at night

Artificial light at night (ALAN) is an emerging driver of biodiversity change, yet quantitative evidence on how biological responses scale with light intensity remains limited. We compiled and modelled 15 datasets covering a wide range of taxa and processes, including physiological, behavioural, and ecosystem-level responses. Each dataset was fitted with a logarithmic dose–response model, enabling comparison across studies and identification of critical thresholds. Across all datasets, biological responses increased steeply at very low illuminance levels, with the strongest changes occurring in the sub-lux range (<1 lx). This pattern was consistent across organismal groups (fish, birds, insects, plants) and process types, and no clustering by taxon or function was detected. The convergence of response shapes suggests a potentially universal sensitivity to dim artificial light, likely reflecting fundamental properties of photoreceptor physiology. The steep responses at low light intensities contrast with many other environmental stressors, where sigmoidal or threshold dynamics are typical. These findings highlight the need to consider even dim artificial illumination as ecologically significant. Our models further provide quantitative thresholds that can be integrated into ecological risk assessment frameworks. Sub-lux levels emerge as critical exposure ranges, with implications for conservation management and the development of regulatory guidelines. Our study demonstrates that dose–response modelling of biological light sensitivity can uncover generalisable patterns across taxa and processes. Recognising the ecological importance of low-level illumination, e.g. due to skyglow, is essential to inform evidence-based mitigation strategies and to safeguard biodiversity in a world of rapidly expanding night-time lighting.