Assessing bird avoidance of high-contrast lights using a choice test approach: implications for reducing human-induced avian mortality.

Background: Avian collisions with man-made objects and vehicles (e.g., buildings, cars, airplanes, power lines) have increased recently. Lights have been proposed to alert birds and minimize the chances of collisions, but it is challenging to choose lights that are tuned to the avian eye and can also lead to avoidance given the differences between human and avian vision. We propose a choice test to address this problem by first identifying wavelengths of light that would over-stimulate the retina using species-specific perceptual models and by then assessing the avoidance/attraction responses of brown-headed cowbirds to these lights during daytime using a behavioral assay.

Methods: We used perceptual models to estimate wavelength-specific light emitting diode (LED) lights with high chromatic contrast. The behavioral assay consisted of an arena where the bird moved in a single direction and was forced to make a choice (right/left) using a single-choice design (one side with the light on, the other with the light off) under diurnal light conditions.

Results: First, we identified lights with high saliency from the cowbird visual perspective: LED lights with peaks at 380 nm (ultraviolet), 470 nm (blue), 525 nm (green), 630 nm (red), and broad-spectrum (white) LED lights. Second, we found that cowbirds significantly avoided LED lights with peaks at 470 and 630 nm, but did not avoid or prefer LED lights with peaks at 380 and 525 nm or white lights.

Discussion: The two lights avoided had the highest chromatic contrast but relatively lower levels of achromatic contrast. Our approach can optimize limited resources to narrow down wavelengths of light with high visual saliency for a target species leading to avoidance. These lights can be used as candidates for visual deterrents to reduce collisions with man-made objects and vehicles.