Dakatcha Woodland Biodiversity Assessment

This dataset contains camera trap observations collected during a long-term mammal survey in Dakatcha Woodland, Kilifi County, Kenya, between April 2019 and July 2022. The survey was designed to establish a baseline inventory of medium- and large-bodied terrestrial mammals within and around the A Rocha Dakatcha Nature Reserve and to support conservation planning for this globally important biodiversity hotspot. The survey comprised 122 camera trap deployments using Reconyx HyperFire infrared cameras, representing 6,779 camera-trap days of sampling effort. Cameras were deployed for approximately two months per location before being moved to new sites, providing broad spatial coverage across the study area. Cameras were positioned along wildlife trails or other locations showing evidence of mammal activity and were programmed to capture three photographs per trigger. Images were managed and classified using Wildlife Insights, which was used throughout the project for deployment management, image annotation, species identification, and quality control. The complete project, including all classified images and metadata, is publicly available at app.wildlifeinsights.org. Images were manually classified and identified to the lowest taxonomic level possible. When species-level identification was not possible, records were assigned to the highest confidently identified taxonomic rank. The published dataset includes wild mammals, as well as records of humans and domestic animals detected during the survey. Independent detection events were defined using a one-hour interval between consecutive records of the same taxon at the same camera. The dataset is published as a Camtrap Data Package following the open camera trap data standard, facilitating interoperability and reuse through GBIF and other biodiversity data infrastructures. It provides occurrence records together with deployment metadata, camera locations, sampling effort, timestamps, and associated media references where available. The scripts used to convert the Wildlife Insights export into the published Camtrap Data Package, together with all analyses and figures presented in the accompanying report, are openly available at github.com, providing a fully reproducible workflow from image classifications to the published dataset and report. The survey documented 28 wild mammal species, including globally threatened species such as the Endangered Golden-rumped Sengi (Rhynchocyon chrysopygus) and the Vulnerable Leopard (Panthera pardus). Beyond species occurrences, the dataset supports studies of species distributions, occupancy, relative abundance, activity patterns, community composition, and long-term monitoring of mammal populations in East African coastal forests. A detailed analysis of the survey methods, species accounts, and conservation implications is presented in the accompanying report, Mammal diversity survey in Dakatcha Woodland, Kenya: Results from a four-year camera trap survey (2019–2022).Data have been standardized to Darwin Core using the camtrapdp R package and only include observations (and associated media) of animals. Excluded are records that document blank or unclassified media vehicles and observations of humans.