Understanding acoustic indices as multi-taxa biodiversity and habitat quality indicators

• We devise a conceptual framework aiding the interpretation of acoustic indices. • Indirect and direct drivers alter bio-, geo-, and anthrophony and thus soundscapes. • Evaluation of framework using recordings from a land-use gradient in Madagascar. • Acoustic indices act as multi-taxa biodiversity and habitat quality indicators. • Richness of vocalizing and non-vocalizing taxa are both linked to acoustic indices. Acoustic monitoring captures the sound of nature, revealing biodiversity patterns and underlying drivers. Acoustic indices are increasingly used to analyse obtained recordings, but studies show inconsistent performance of acoustic indices as biodiversity indicators. Here, we develop a conceptual framework on what acoustic indices reflect, including sound sources as well as their direct and indirect drivers. We evaluate our framework using recordings from six land–use types and old–growth forest sites in north-eastern Madagascar. We found the highest correlation (adj. R 2 > 0.51) of Acoustic Diversity Index and Acoustic Entropy with endemic multi-taxa species richness, canopy closure, and vegetation density. The results suggest that including multiple taxa strengthens the association between the indices and species richness, potentially accounting for unobserved vocalising taxa, and that acoustic indices indicate habitat quality, reflecting underlying indirect and direct drivers affecting biodiversity. Yet, some acoustic indices are limited by saturation and their sensitivity is too low to detect small changes in biodiversity (e.g., loss of few species).