Colours of the night: spectrum-specific impacts of light pollution on biota

Artificial light at night (ALAN) disrupts ecosystems by altering natural light cycles and affecting species’ physiology and behavior. This meta-analysis examines how different wavelengths and intensities of ALAN impact organisms. Broad-spectrum “cool” light, enriched with blue and ultraviolet radiation, strongly disrupts circadian rhythms, melatonin production, and nocturnal activity. However, against common belief, broad-spectrum "warm" light is almost as impactful as broad-spectrum "cool" light without ultraviolet radiation, even though physiological processes are mostly influenced by short wavelengths. The impact of ALAN is not consistently dose-dependent, as even dim light often causes substantial biological disruptions. Thus, effective mitigation strategies require tailored solutions to specific ecological contexts and should generally avoid nocturnal illumination unless clearly needed, as there is no single “safe dose” and “safe spectrum” of ALAN.

Methods
We conducted literature searches in electronic academic databases: Web of Science and Scopus for peer-reviewed papers published in English until 30 August 2023. We used the following terms TITLE-ABS-KEY ( ( "Artificial light*"  OR  "Light* pollution"  OR  "Light* at night"  OR  "night time light*" )  AND  ( "light intensity"  OR  "spectral composition"  OR  "spectral sensitivity"  OR  "colo*r"  OR  "wavelength"  OR  "LED light"  OR  "HPS light"  OR  "red light"  OR  "blue light" )  AND  ( "species"  OR  "ecosystem*"  OR  "ecology"  OR  "community composition"  OR  "phototaxis"  OR  "flight to light"  OR  "stress"  OR  "abundance"  OR  "behav*"  OR  "reproduction"  OR  "predat*"  OR  "activity"  OR  "physiolog*"  OR  "melatonin"  OR  "development"  OR  "trophic"  OR  "biomass"  OR  "pollination"  OR  "phenolog*"  OR  "photosynt*"  OR  "budburst"  OR  "leaf"  OR  "flowering" ) ). The search returned 1694 studies (after duplicates removal). We also checked the bibliography of eligible articles to extract all relevant papers and therefore included 64 additional studies. The articles were screened for eligibility and we selected studies that met the following inclusion criteria: (1) laboratory and field studies testing biological effects of ALAN on organisms (excluding humans); (2) studies with a control group (dark night conditions) and treatment groups exposed to ALAN with illuminance up to 180 lx; (3) studies with at least 2 replicates per treatment and providing information on light types used and light levels (if necessary, we converted non-photometric quantities lighting units to illuminance in lux, see details below); (4) studies containing data on means, sample size and estimation of variation. If only box plots were presented, we extracted the median and interquartile range. In total, 314 studies (with 4 905 effect sizes) met our inclusion criteria.