Increases in night light did not decrease aerial insectivore counts relative to other birds on North American survey routes in two sensor eras, 1992-2025

Artificial light at night can decrease flying insects, so it can decrease the food of birds that catch insects in the air. We tested this prediction with North American Breeding Bird Survey counts and a harmonized DMSP/VIIRS night-light series. We used four independent samples: the DMSP era (light 1992-94 to 2011-13, counts 1993-97 to 2010-14) and the VIIRS era (light 2014-16 to 2022-24, counts 2015-18 to 2022-25), each east and west of 100 W (827-1,560 routes per sample). A route-level long-difference model related the change in light to the change in counts of aerial insectivores (swallows, swifts, nightjars, tyrant flycatchers) minus non-insectivores, with BCR fixed effects and observer and survey-date controls. The contrast per doubling of (1 + DN) was +5.4% (95% interval -4.9 to +16.8), +0.3% (-9.7 to +11.3), +2.8% (-2.7 to +8.5) and -4.9% (-10.7 to +1.3). The signs did not agree and no test passed Holm correction. In the eastern samples, the intervals exclude relative decreases larger than about 5%. As a secondary result, other insectivores (mostly forest songbirds) decreased relative to non-insectivores where DMSP-era light increased in the east (-11.6%, -16.2 to -6.6; permutation p = 0.001). This result persisted in five checks but did not repeat in the VIIRS-era east (+0.4%, -2.6 to +3.5). In annual route-year panels, future light predicted counts as well as past light. Light change is a marker of development. Thus, these data cannot separate light from habitat loss, traffic noise or lower detection of songbirds near roads. The analysis does not measure deaths during migration at lit buildings. Code, derived route tables, source hashes and figures are included.

artificial light at night · light pollution · Breeding Bird Survey · aerial insectivores · DMSP · VIIRS · long difference · birds · insectivores · observation bias · null result · North America