Comparing the effectiveness of camera trapping, driven transects and ad hoc records for surveying nocturnal mammals against a known species assemblage

Management and conservation rely on surveys to determine which species are present but some species, such as nocturnal mammals, can be very difficult to detect. Camera trapping, driven transects, and ad hoc records are well-established noc- turnal survey methods, but their detection effectiveness has not been tested against known species presence. We tested the effectiveness of these methods with a known assemblage of nocturnal African mammals. We compared camera traps placed at targeted and random locations, driven transects, and informal ad hoc records against one another and the 20 predominately nocturnal medium- and large-sized mammals known to be on site. Species-specific detection patterns, detection curves for each method, and detection effort required per species, were compared to determine the effectiveness of each method. Each of the methods detected at least 60% of known nocturnal mammal species over a 20-month period: driven transects and ad hoc records were the most successful (16/20 80%), followed by targeted camera traps (15/20 75%) and random camera traps (12/20 60%). Jaccard’s Coefficients of Community Similarity peaked at 0.78, showing notable differences in the communities sampled with different methods. There was considerable variation in effort required to detect each species for each method, and the number of species detected with specific levels of effort was also highly variable between methods. We suggest that mixing methods, using different camera placement including use of random locations, and incorporating cost-neutral ad hoc records into formal survey efforts can all greatly increase the overall effectiveness and efficiency of species detection