Pulsed artificial light at night alters moth flight behaviour

Abstract

Vehicle headlights create pulsed artificial light at night (pALAN) that is unpredictable, intense and extends into previously dark areas. Nocturnal insects often have remarkable low-light vision, but their slow pupillary light responses may leave them vulnerable to pALAN, which has important ecological consequences. To test this, we exposed nocturnal moths-important pollinators and prey-to four pALAN treatments. These comprised 'cool' and 'warm' lights, either emitted from phosphor-coated light-emitting diodes (LEDs) or RGB (red-green-blue) LEDs, matched in colour (CCT) and intensity to human vision. We assessed the initial behavioural response, likely crucial to the survival of an organism, of 428 wild-caught moths comprising 64 species. We found that exposure to a cool phosphor-coated LED light pulse increased instances of erratic flight and flight-to-light that are likely detrimental as they increase the risks of impact with a vehicle, predation or excess energy expenditure. Our findings suggest that pALAN can cause a wide range of behavioural responses in nocturnal moths, but that the most harmful effects could be minimized by reversing the current shift towards high CCT (cool) phosphor-coated LED car headlights. Lower CCT or RGB alternatives are likely to provide benefits for road safety while reducing ecological harm.

Keywords: artificial light at night; flight behaviour; headlights; moths; pulsed light.

Figure 1.

Spectral sensitivity of (a) cones in humans [33,34], with long- (lw), middle- (mw) and short-wavelength-sensitive cones (sw), referred to based on the peak sensitivity, presented; and (b) photoreceptors of the nocturnal elephant hawkmoth (Deilephila elpenor) [35,36], with ultraviolet (UV), blue (B) and green photoreceptor (G) sensitivities shown. Additionally, green photoreceptor sensitivity typically used for achromatic contrast calculations is depicted by the ‘G alone’ line; and (c) irradiance plots of experimental treatments. Graph (c) was rescaled to max = 1.

Figure 2.

(a) Frequency of initial behaviours under all pALAN treatments and control, displayed in a mosaic plot, where the width of the column indicates the number of repeats. (b) Latency (log-transformed) of moths’ responses under the different treatments. The lines in the box plots represent the medians; the box represents the interquartile range (IQR), the upper (Q3) and lower (Q1) quartiles; the whiskers show the range of the data (minimum and maximum value excluding outliers, respectively, with dots showing outliers). The letters above the plots indicate the following: A indicates the spectrum differences between phosphor-coated LEDs and RGB LEDs, with p = 0.003; B indicates the statistical comparison (Bonferroni post-hoc test) between cool and warm phosphor-coated LEDs, with p = 0.001, and C indicates the statistical comparison between cool and warm RGB LEDs, with p = 0.392.