Wildlife struggle in an increasingly noisy world.

We live on an ever more-populous planet, pulsating with human-generated noises of every description. The most ubiquitous noisemaking structures we produce are trafficladen roads (Fig. 1), which already criss-cross much of the Earth and are projected to increase in length by some 25 million km by midcentury: enough to encircle the planet more than 600 times (1). For wildlife, the challenges of living in a world increasingly swamped by such infrastructures are only going to worsen. Road ecology as a discipline (1–4) has been galvanized by the realization that we are presently experiencing the most explosive era of infrastructure expansion in human history (1). A growing body of research is revealing the myriad ways that roads can affect wildlife and ecosystems, sometimes opening a Pandora’s box of environmental problems, such as illegal hunting, encroachment, wildfires, and land speculation (5–7). Even where such activities are controlled, roads can still cause marked local changes in the abundance and behavior of wildlife, via edge effects, road kill, and vehicle noise and pollution (2–4). One of the challenges facing road ecologists is to understand the specific causes and mechanisms underlying the highly varied responses of different wildlife species to roads (8). Are certain species shunning roads and their vicinity because of a desire to avoid road clearings or edge effects, or do they fear the sound or sight of fast-moving vehicles? And, are species that persist near roads relatively healthy and unaffected, or do they suffer from stresses provoked by recurring roadrelated disturbances? In PNAS, Ware et al. (9) use a “phantom road” approach to disentangle the effects of road noise from other consequences of road creation. The authors established a ∼0.5-kmlong array of loudspeakers along a ridgeline in a conifer forest-shrubland mosaic in southwestern Idaho, and then played road noises at a volume comparable to a suburban street (increasing ambient noise from an average of 37 dB to a louder 48 dB). Along their phantom road, they captured migrating songbirds in the autumn, using mist-nets, as the birds weremoving southward toward their wintering grounds. Ware et al.’s (9) study elegantly incorporated two types of experimental control. The