To Track a Forest’s Recovery, Artificial Intelligence Just Listens
By “listening” to forest soundscapes and identifying animal species, AI can be used to evaluate biodiversity efforts, according to a new study.
A squirrel monkey is seen in the Ecuadorian Amazon rainforest in 2009.
Photographer: Alejandra ParraThe sounds of a forest — from the guttural croaking of frogs to the chirpy trills of birdsong — can be used to track biodiversity recovery, according to a study published today in the journal Nature Communications. Paired with artificial intelligence, soundscapes can provide a low-cost way to measure the health of an ecosystem.
Researchers looked at reforestation projects in a part of Ecuador where 90% of the tropical forest had been lost to logging. “We could only describe that the trees [were] regrowing,” says Jörg Müller, a professor of animal ecology at the University of Würzburg in Germany and lead author of the study. “But what is happening to the animals?”