The central goals of this project are to digitize all of Gabon's vertebrate collections and to create a software platform that leverages machine-learning algorithms to identify, store, and analyze a large dataset of camera-trap photos. All of these data will be made public on the Gabon Biodiversity portal and through GBIF.org.
Collections and records of Gabonese biodiversity are scattered across the country in several institutions:
- The collection at the National School of Water and Forests (ENEF): 2,000 fish and 200 reptile and mammal specimens
- The Institute for Agronomic and Forest Research (IRAF): 10,000 fish specimens.
- The Tropical Ecology Research Institute (IRET) has the oldest collection of fishes and amphibians.
- The Smithsonian collection in Gamba: 860 fish specimens, 900 herpetofauna, 280 mammals, 90 birds
- The National Parks Agency (ANPN) and international partners (Panthera, University of Stirling) manage an archive of several million photographs (estimated > 1 million occurrence records) of terrestrial mammals and birds from camera trap data.
These various collections are not all digitized and, in the case of camera trap data, are not labeled to the species level. These valuable data are, therefore, inaccessible to the scientific community. The need to work together and pool resources have led the above instituitions to form a partnership to create a national biodiversity database.
Over two years, the project team aims to digitize all vertebrate specimens across various collections (13,500 records), and to fully mobilize existing camera trap data (1 million records) to create a national database and publish it to GBIF. By investing BID funding in training local experts and developing open-source software, this project can have a lasting impact for decades beyond this project.