Camera trap data of mammals from Baluran National Park

Dholes (Cuon alpinus) are endangered large carnivores found in scattered populations in Asia. One of the main threats to dholes is the decreasing prey availability throughout their distribution range. In the present study we used camera trap data collected over six years to investigate the temporal activity patterns of dholes and their putative prey species in Baluran National Park in Java, Indonesia. We also explored the overlap in activity between dholes and the park’s other remaining large carnivore the Javan leopard (Panthera pardus melas), as well as humans. Furthermore, we investigated potential differences in activity patterns between dholes in packs and dholes roaming in pairs or alone. We found a high temporal overlap between dholes and their wild ungulate prey species (ranging from D=0.66–0.90), with the lowest overlap observed between dholes and bantengs (Bos javanicus) (D=0.66), and the highest between dholes and muntjacs (Muntiacus muntjak) (D=0.90). A very low overlap was found between dholes and domestic cattle (Bos indicus) (D=0.27) whereas a moderately high overlap was found between dholes and leopards (D=0.70) and dholes and humans (D=0.62). We found a significant difference in activity patterns between dholes in packs and dholes roaming alone or in pairs (D=0.78, p=0.01). Single/pairs of dholes were more active both during the day and at night, whereas packs were predominantly active around sunrise and sunset. The high overlap with humans potentially has a negative effect on dhole activity, particularly for dispersing individuals, and the low overlap with domestic species questions the extent to which dholes are considered to predate on them.

Methods
Camera trapping was conducted from March-December in consecutive years 2015-2020. Camera traps were deployed in a grid system of 1x1 km covering a total area of 130 km² with some grids used in multiple years. The camera traps used were Bestguarder (Bestguarder, China) and Covert (Covert Scouting Cameras, LLC, Russellville, KY). Cameras were installed approx. 30-45 cm from the ground with the camera position to an animal trail at a distance of 2.5 m following. Pictures were sorted and annotated following Sanderson and Harris (2013) and analysed in the software R v. 4.0.3 (R Development Core Team, 2020) using the packages ‘Activity’ v.1.3.3 (Rowcliffe, 2022) and ‘overlap’ v.0.3.4, (Meredith and Ridout, 2021).