Dataset and code: Auger et al. 2026. Chimpanzees and elephants are reliable flagship-umbrella species for eco-labels. Biotropica.

This depository contains the dataset and R scipt that was used in the following article: Auger, C., Montfort, F., Bellón, B., Tibesigwa, J. J., Rugonge, H., Katumba, R., … Krief, S. (2026). Chimpanzees and Elephants Can Be Used as Reliable Flagship-Umbrella Species for Eco-Labels. Biotropica, 58(1), e70143. doi.org   This dataset contains number of independent detection events for observed taxa, as well as anthropic pressure measurements (intensity of illegal activities targeting fauna, intensity of illegal activities targeting flora, distance from the forest edge), and environmental variable (Normalized Difference Vegetation Index, Topographic Wetness Index , elevation), at 109 camera-trap sampling sites in the Sebitoli area of Kibale National Park (KNP), Uganda. KNP (0°13ʹ–0°41ʹN; 30°19ʹ–30°32ʹE) protects 795 km² of mid-altitude moist tropical forest (1,100–1,590 m a.s.l.) characterized by bimodal rainfall. The Sebitoli sector (~25 km²), located in the northern part of KNP, consists mainly of degraded and regenerating forest with only 14% old-growth stands. The Sebitoli Chimpanzee Project (SCP) has been conducting long-term biodiversity monitoring in the area since 2008. For this dataset, 14–19 camera traps were deployed annually, between January 2020 and December 2023, yielding 109 sampling sites, each sampled for at least 15 days. Species identifications were performed to the finest possible taxonomic level, and IUCN Red List status was assigned at the species level. Following standard protocols, independent detection events was defined as consecutive detections of the same taxon separated by ≥30 minutes. We summed independent detection events for each taxa at each CT sampling site. Here we used the Sentinel-2 time series, obtained from Google Earth Engine (earthengine.google.com), to calculate the median Normalized Difference Vegetation Index (NDVI) for the four-year study period. Elevation data was obtain from the Shuttle Radar Topography Mission (STRM) and the Topographic Wetness Index calculated from the SRTM elevation data using the r.topidx tool in GRASS GIS Distance from the forest edge was calculated at each camera trap sampling site as the Euclidean distance to the KNP’s canopy cover limits delineated from satellite imagery. Illegal activities data were collected and geolocated by a SCP team during patrols conducted six days a week, eight hours per day, over the four-year study period. Illegal activities recorded have been grouped into two categories: activities targeting wildlife (wire snares), and activities targeting flora (tree cutting, firewood harvesting, wild Piper guineense harvesting, charcoal burning and tree debarking). To spatially model illegal activities intensity, we used the QGIS Heatmap (Kernel Density Estimation) tool with a 100-meter kernel diameter to generate raster layers for each category, where higher density values corresponded to a greater concentration of illegal activity points.   Code was designed from Niku et al. (2019) using the gllvm package (Niku et al., 2019, 2021; van der Veen et al., 2021, 2023) in R software (v. 4.3.1). Correlation between explanatory variables was tested using Pearson’s correlation test. Correlation matrix, statistically validating pairwise co-occurrence between species, were plotted for significant correlation (p-value<0.05).