Artificial light at night is increasingly recognized as an important environmental disturbance that influences the fitness and habitats of numerous species. However, its effects on large vertebrates, and their interactions, remain unclear. A better understanding of these effects is critical as large vertebrates are widely valued for ecological and economic reasons, yet are being exposed to unprecedented light pollution in the wake of land-use change. Here, we combined satellite-derived estimates of light pollution, with GPS-data from cougars (Puma concolor; n = 117, collected 2002 – 2015), mule deer (Odocoileus hemionus; n = 486, collected 1999 – 2018), and locations of cougar-killed deer (n = 1,562 carcasses, collected 2000 – 2018), to assess the effects of light exposure on mammal behavior and predator-prey relationships across wildland-urban gradients in the southwestern United States. We first tested which landscape features, habitat characteristics, traditional human footprint metrics (e.g., road density), and nightlight values were most associated with locations where cougars successfully killed mule deer within each of 18 study areas throughout the region. We then evaluated how nightlight altered the space use and movement of both deer and cougars, while accounting for a suite of environmental and human-related factors using integrated step selection functions.
Results/Conclusions
We found that nightlight had the greatest effect of all covariates on where cougars killed deer at the wildland-urban interface. Our results indicate that mule deer used anthropogenic environments to access forage and were more active at night than their wildland conspecifics. It some cases the anthropogenic landscapes also helped shield deer from predation from cougars. However, as cougar tolerance of nightlight increased relative to wildland individuals, cougars used the relatively darkest areas in the wildland-urban interface to successfully hunt. Both species exhibited functional responses to light pollution at fine scales; individuals with less light exposure increasingly avoided illuminated areas when exposed to greater radiance, whereas deer living in the wildland-urban interface selected for elevated light levels. Our findings illustrate how remotely sensed estimates of a growing global sensory pollutant can be used to better understand the nuanced relationship of a generalist predator and their primary prey. We conclude that integrating estimates of light pollution into ecological studies provides crucial information about how the dynamic human footprint can alter animal behavior and ecosystem function across spatial scales.