AI anticipates spring bird arrival as Arctic climate shifts

[Cara J]

As temperatures rise and the Arctic climate becomes more unpredictable, birds could be following the shifting spring and showing up to nest off schedule. Researchers recently developed machine learning approaches to analyze sound recordings to estimate when songbirds will arrive for the breeding season. They hope the technology can also be harnessed to assess climate change’s effects on other wildlife in other places. The techniques did a good job of replicating the work of humans on the ground, said Ruth Oliver, first author on the study published in Science Advances. “Deploying them in the future would be important for understanding how climate change might be impacting the timing of bird arrival on breeding grounds,” she said. The timing has implications for the birds’ reproductive success and population, she said. This new approach “could help us tease out” which environmental conditions most affect changes in their timing. A graduate student at Columbia University, Oliver and her colleagues aimed to investigate how increasing variability in the onset of the Arctic spring related to when migratory songbirds flew in to reproduce. They worked with 1,200 hours of vocalizations by many songbird species, including Gambel’s white-crowned sparrows (Zonotrichia leucophrys) and Lapland longspurs (Calcarius lapponicus), [...]

View the full article