Stacking the odds: light pollution may shift the balance in an ancient predator–prey arms race

Summary Artificial night lighting threatens to disrupt strongly conserved light‐dependent processes in animals and may have cascading effects on ecosystems as species interactions become altered. Insectivorous bats and their prey have been involved in a nocturnal, co‐evolutionary arms race for millions of years. Lights may interfere with anti‐bat defensive behaviours in moths, and disrupt a complex and globally ubiquitous interaction between bats and insects, ultimately leading to detrimental consequences for ecosystems on a global scale. We combined experimental and mathematical approaches to determine effects of light pollution on a free‐living bat–insect community. We compared prey selection by Cape serotine bats N eoromicia capensis in naturally unlit and artificially lit conditions using a manipulative field experiment, and developed a probabilistic model based on a suite of prey‐selection factors to explain differences in observed diet. Moth consumption by N . capensis was low under unlit conditions (mean percentage volume ± SD: 5·91 ± 6·25%), while moth consumption increased sixfold (mean percentage volume ± SD: 35·42 ± 17·90%) under lit conditions despite a decrease in relative moth abundance. Predictive prey‐selection models that included high‐efficacy estimates for eared‐moth defensive behaviour found most support given diet data for bats in unlit conditions. Conversely, models that estimated eared‐moth defensive behaviour as absent or low found more support given diet data for bats in lit conditions. Our models therefore suggest the increase in moth consumption was a result of light‐induced, decreased eared‐moth defensive behaviour. Policy implications . In the current context of unyielding growth in global light pollution, we predict that specialist moth‐eating bats and eared moths will face ever‐increasing challenges to survival through increased resource competition and predation risk, respectively. Lights should be developed to be less attractive to moths, with the goal of reducing effects on moth behaviour. Unfortunately, market preference for broad‐spectrum lighting and possible effects on other taxa make development of moth‐friendly lighting improbable. Mitigation should therefore focus on the reduction of temporal, spatial and luminance redundancy in outdoor lighting. Restriction of light inside nature reserves and urban greenbelts can help maintain dark refugia for moth‐eating bats and moths, and may become important for their persistence.