New protocols for visual and acoustic monitoring of animal biodiversity

Sabrina Krief (MNHN Professor) and Hugo Magaldi (MNHN/OFVi Research Engineer), assisted by Raymond Katumba and Harold Rugonge, managers of the Sebitoli Chimpanzee Project (SCP) and 3 SCP field assistants, Daniela Birungi, Robert Asiimwe and Robert Nyakahuma, took part in this mission.

Objectives

1. Field testing of the algorithm developed in the lab for sorting animal taxa detected by CT (portability, speed, simplicity) 
2. Setting up a protocol for monitoring animal biodiversity using camera traps (CT) 
3. Setting up a protocol for monitoring animal biodiversity using passive audio recorders (PAR)
4. Implementation of a field protocol for individual identification by CT for 2 taxa of ‘endangered’ species (chimpanzees, elephants) 
5. Recruitment of new agents to monitor biodiversity using remote sensing (started at the beginning of December 2024 with definition of profiles, publication and interviews). 
6. Technical training for members of the Ugandan team in the new protocols. 

Fiel protocols

At the end of this mission, three protocols were active.

1. A first protocol dedicated to the visual and acoustic monitoring of general animal biodiversity, with 20 locations spread over a regular grid, each equipped with a CT and a PAR.
2. A second protocol for the visual identification of chimpanzees and elephants, with 7 CTs at variable locations.
3. A third protocol for comparing different acoustic landscapes (forest, swamp, edge of forest), with 3 PARs recording continuously.

Recruitment  

To monitor these protocols, the project relies on the Ugandan team previously in charge of the TCs (3 field assistants), and therefore on the SCP's existing local expertise in monitoring animal biodiversity through TCs. The team was strengthened by the recruitment of 2 additional assistants, Oliver Kemigisa and Nathan Alinaitwe, whose 2-week trial period was successfully completed. The integration of the mission into the SCP's activities was made possible thanks to managers Raymond Katumba and Harold Rugonge and their full support for the team. 

A data manager, Innocent Kasekendi, in charge of supervising the field team in liaison with Hugo Magaldi and the research team at the Muséum National d'Histoire Naturelle, was recruited after a two-week trial period. The first candidate considered had more of a laboratory research profile than a field technical profile, so Innocent's integration into the team is the culmination of a new recruitment process begun during the mission and completed last week.

Following the installation of the TCs and RAPs, two weeks of data collection and analysis tests in real-life situations enabled the new team to get up to speed.

Testing the algorithm 

The inter-species classification algorithm developed at the Muséum has been deployed on the station's laptop. It runs on a CPU and without an internet connection via an intuitive user interface, and enables an initial rapid analysis of the video clips collected by the data manager. The Ugandan team has also been trained in the use of a proprietary labelling tool, to simplify the task of annotating the clips.

Perspectives 

The challenge for the coming months is for the field team to maintain and adapt these protocols on site, while analysing the data collected and using it at the MNHN to develop new artificial intelligence tools for acoustic monitoring and visual identification of individuals.