We demonstrate the power of using passive acoustic monitoring (PAM) & machine learning (ML) to survey species, using the example of ruffed lemurs in southeastern Madagascar!



Ruffed lemurs are keystone species in Malagasy rainforests, as they fill vital pollinator and seed disperser roles. However, they are Critically Endangered and limited to fragments of the once-continuous eastern Malagasy rainforest corridor.



We developed an open-source (!) training dataset & CNN model for detecting lemur calls. Then, we compared our PAM/ML pipeline for detecting lemur calls to in-person observations/manual analysis, and found that passive acoustics & machine learning was:

➡ cheaper

➡ more temporally scalable

➡ less labor-intensive

➡ less time-intensive



We also show the first quantitative, published evidence of nocturnal vocal activity for ruffed lemurs! We saw a particular peak in vocal activity (nocturnal & diurnal) during the short mating period.

With incredible collaborators @emmanueldufourq, Lorène Jeantet, Andrea Baden & the incredible folks at Centre ValBio.

Full paper (open-access) here: 

An integrated passive acoustic monitoring and deep learning pipeline for black‐and‐white ruffed lemurs (Varecia variegata) in Ranomafana National Park, Madagascar

American Journal of Primatology is a broad-scope primates journal publishing studies from all areas of the field, from behavioral ecology to molecular genetics.

Wiley Online Library