Anthropogenic noise and light impact wildlife behavior, distribution, movement, and population fitness and survival. Traffic noise and light can inhibit wildlife use of areas adjacent to roads, impair wildlife perception of traffic risks, and cause a barrier effect to wildlife occurrence and movement well beyond road edges. Planning the location and design of new wildlife crossings and enhancements of existing structures currently does not take into account traffic impacts on wildlife approaching the structures. We developed a statewide model of vehicle light intrusion into areas adjacent to roads to aid locating and designing wildlife crossings. We used ArcGIS to model the “light-scape” at high-resolution around all California state highways. We used 1/10 mile post-miles as the model sources of traffic illumination and modeled propagation across adjacent landscapes to 1 km. We populated the models in the field (I-80) using short-distance transects of light measurements employing a low-light meter. We measured light during a quarter-moon, after civil twilight, at varying known distances from the highway. Individual vehicles generated up to 2.0 lux of light at and near the roadside (~6-16 m from vehicle lanes). We also assumed a 10 lux maximum roadside illuminance, based on studies by the Insurance Institute of Highway Safety. With a 2 lux starting illumination, brightness exceeding a new moon (~0.01 lux) and full moon (0.1) were modeled to occur 163 m and 64 m, respectively from a low traffic-volume highway across a flat landscape. At 10 lux at the highway-edge, brightness exceeding a new moon and full moon were modeled to occur 334 m and 126 m from the highway edge. These distances are well outside the typical approach zones to wildlife crossings, suggesting potential impacts to crossings. We found that the model may not predict illumination well in the presence of dense vegetation, where the model tended to over-estimate illumination. We are currently using vegetation maps to modify the modeled illumination. The model did predict dark places well (topographic dips), highlighting paths to wildlife crossing locations. We also buffered the 354,000 km of public roads in California by 64 m and 163 m, resulting in a potential vehicle-illuminated area of 29,098 (new moon) to 55,489 (full moon) square km, approximately 6.9% to 13% of the state’s area. Because predators are sensitive to ambient light conditions (Barrientos et al., 2023; Sergeyev et al., 2023), we are comparing (and will present) GPS locations of collared mountain lions, bobcat and mule deer to modeled illumination near state highways. Although our focus is on wildlife impacts, this model would be useful for community planning and environmental analysis associated with delivering transportation projects. We are sharing these data through a web-portal supported by a grant from the Wildlife Conservation Network.
Statewide Modeling of Traffic Illumination of Surrounding Habitat
ICOET 2025
Abstract