Data from: Acoustic monitoring reveals contrasting responses of two endangered primates to primary forest degradation in the Ecuadorian Chocó

Source
DataCite
Published
2026
Country
Ecuador
License
Creative Commons Attribution 4.0 International
Themes
Threatened Species, Mammals, Forests

Get the data from DataCite

About this dataset

We applied passive acoustic monitoring (PAM) combined with a deep learning-based classification algorithm to detect vocalizations of two endangered primates—the Ecuadorian mantled howler monkey (Alouatta palliata aequatorialis) and the brown-headed spider monkey (Ateles fusciceps fusciceps)—across primary, secondary, and agricultural landscapes in the Canandé and nearby watersheds. We selected 178 plots at the centre of which we deployed an Autonomous Recording Unit (ARU): BAR-LT recorder (Frontier Labs) (175 plots) or a Song Meter Mini 2 Li-ion recorder (Wildlife Acoustics) (23 plots), equipped with a downward-facing omnidirectional microphone mounted approximately 1.70 m above the ground. Recorders captured 2-min files every 15 min throughout the day.Training data included recordings from ARUs and field collections. The vocalizations were then segmented into 3-second clips to train...

How the data was collected

It covers five main stages.First, it imports and preprocesses the data.Second, it provides a descriptive overview of the dataset, including summary statistics on detection counts and site-level occurrences for each species and a map of the study area overlaid with ARU sampling locations.Third, it fits GAMs separately for the Brown-headed Spider Monkey and the Ecuadorian Mantled Howler Monkey. at 500 m and 2000 m buffers; forest cover derived from ESA WorldCover, JRC TMF2, and a combination of both products at 500 m and 2000 m buffers; canopy height from Lang et al.

Browse sustainability datasets on SusanHub