The Pan African Programme (PanAf) was established to study the emergence and change of population diversity in chimpanzees from an evolutionary, ecological and anthropocene perspective. Research is not limited to a single level of biological organisation, but includes studying genetic, microbial, physiological, demographic, resource use, habitat interaction, behavioral and cultural variation among populations. With a very large collaborator network temporary research sites were established at more than 40 locations throughout the chimpanze range. Using a uniform protocol the largest existing data and sample set on chimpanzee population diversity was collected, by using a wide range of non-invasive techniques, including camera trapping, the collection of organic sample material, or transect and plot sampling.
Thanks to this unprecedented community effort a wide range of questions can be addressed with the PanAf, for example ‘what are the conditions under which chimpanzee populations have diversified’, ‘what is the genetic population history of chimpanzees’, ‘which role did Pleistocene refugia play for population diversification’ or ‘are there specific hotspots of and conditions for populations to diversify’?
Contact: Mimi Arandjelovic and Hjalmar Kühl
The IUCN SSC A.P.E.S. database is a central repository for field survey information on great apes that is supported by a very large network of conservation and research institutions and individuals. It is widely used to assess abundance and temporal trends of ape populations, to identify key drivers of density distribution and change, to inform conservation bodies, including IUCN, CITES or UNEP on the state of apes and to conduct a large variety of scientific research projects. Data compiled on this platform are available on request and are handled according to the data access and release policy.
Research projects include the assessment of future scenarios of great ape status under climate and land-use change, the modeling of chimpanzee distribution over the last 120,000 years, or the identification of key factors that enable and limit the persistence of great apes in the Anthropocene.
Contact: Tene Sop
Chimp&See is a citizen science platform for the classification of video clips from camera traps installed in African ape habitat. Chimp&See has a user community of more than 14,000 citizen scientists, who have conducted already more than 300,000 classifications of video clips. Citizen Scientists engage in species classification, individual identification of chimpanzees or detection of health issues. Research has shown that citizen scientist provide high quality data to be used by scientists for further analyses.
Watch videos, identify species and mark the behavior of chimpanzees and other animals to join research on great apes, sympatric species and their environment.
Contact: Mimi Arandjelovic
Zamba Cloud is an open access platform for the AI-based classification of videos from camera traps and was developed in collaboration with DrivenData Inc, USA. At present classification algorithms distinguish 23 classes of African tropical forest species. Zamba Cloud is also very useful for separating videos with no animal observations from those with observations. Zamba Cloud is currently extended to allow retraining for any species and this module along with other features, such as depth estimation will be available in the second half of 2021. Register, upload your videos and see which videos contain observations or which species your cameras have recorded.
Contact: Mimi Arandjelovic and Hjalmar Kühl
Camera trapping has become a widely applied approach for surveying wildlife around the globe. At present, however, it is challenging to conduct large camera trap surveys for the estimation of density and abundance of multiple species. Although, the automated classification of species from camera trap footage has made much progress in the last years, there is no standard approach yet to integrate output from classification systems with inferential methods for estimation of density and abundance. More specifically, false classifications of species identity and species occurrence introduces errors in estimated density and abundance, if not accounted for properly. This project therefore focuses on the interface between species classification output and inference of occupancy, density and abundance by developing approches for semi-automated visual species monitoring.
Contact: Hjalmar Kühl