Migratory bats respond to artificial green light with positive phototaxis

Abstract

Artificial light at night is spreading worldwide at unprecedented rates, exposing strictly nocturnal animals such as bats to a novel anthropogenic stressor. Previous studies about the effect of artificial light on bats focused almost exclusively on non-migratory species, yet migratory animals such as birds are known to be largely affected by light pollution. Thus, we conducted a field experiment to evaluate if bat migration is affected by artificial light at night. In late summer, we presented artificial green light of 520 nm wavelength to bats that were migrating south along the shoreline of the Baltic Sea. Using a light on-off treatment, we observed that the activity of Pipistrellus nathusii and P. pygmaeus, the two most abundant migratory species at our site, increased by more than 50% in the light-on compared to the light-off treatment. We observed an increased number of feeding buzzes during the light-on compared to the light-off treatment for P. nathusii. However, feeding activity was low in general and did not increase disproportionately during the light-on treatment in relation to the overall echolocation call activity of bats. Further, P. nathusii were attracted towards the green light at a distance of about 23 m, which is way beyond the echolocation detection range for insects of Nathusius' bats. We therefore infer that migratory bats were not attracted to artificial green light because of high insect densities, but instead by positive phototaxis. We conclude that artificial light at night may potentially impact bat migration in a yet unrecognized way.

Fig 1. Difference in the cumulative number of echolocation calls recorded per night during the dark and light periods.

Differences for P. nathusii (Pn), P. pygmaeus (Pp) and Nyctaloids (Nyct) at the landside pole (A), the central pole with the light on-off treatment (B) and the seaside pole (C). Data are depicted as boxplots with 25 and 75 percentiles as the border of the boxes and whiskers encompassing 5 and 95 percentiles.

Fig 2

Difference in the cumulative number of feeding buzzes (FB) recorded per night (A) and difference in the relative foraging activity per night (B) at the central pole for P. nathusii. Data are depicted as boxplots with 25 and 75 percentiles as the border of the boxes and whiskers encompassing 5 and 95 percentiles.