Recent technological and statistical advances have improved our ability to monitor leopards using camera traps. As a result, these cameras are now widely used to determine leopard densities, and to infer population health. However, to determine densities accurately the camera traps need to be deployed in relatively intensive, high-density grids. This makes the method best suited to surveys of small areas.
Through a large-scale camera trapping exercise in the Kruger National Park, South Africa, we are exploring potential indices that could provide a more accurate indication of population health than density alone. These indices may also be possible to obtain from less dense camera trap arrays, making camera trapping more feasible across larger areas.
Over the last year and a half, as part of Lucy Smyth’s PhD, we deployed 10 camera trap grids in and around Kruger National Park. This involved 930 individual camera trap setups, which produced about 500,000 images. Despite genetic and anecdotal evidence of leopard poaching, this leopard populations has not been studied for the last 10 years. Our data therefore provide important baseline information for the area.
We are comparing the leopard densities from Kruger National Park to other potential indices of population status that are more feasible to obtain from camera traps in more expansive areas, and will also help us to develop a more holistic picture of leopard population health. These include population structure, phenotypic resemblance, and timing of activity, which are useful for the following reasons:
Basing policy and management decisions on population density alone means that action is often only taken when leopard populations show declines (by which point the situation is usually already quite dire). So we are looking to see whether changes in population structure may be able to provide an early warning system that could indicate likely population trends before numbers drop appreciably.
Given that leopard rosette and whisker spot patterns are more similar in more closely related individuals, we are investigating whether rosette patterns seen in camera-trap photos could provide an indication of the genetic diversity present within populations.
Finally, a growing bank of evidence is showing that increased levels of anthropogenic disturbance are affecting the proportion of diurnal versus nocturnal activity in many mammal species. We are exploring this in leopards, to see whether daily activity patterns may be indicative of the intensity of disturbance to which a population is exposed.
Leopards are in a lot more trouble than we think.
Leopard populations are in free-fall across much of their range, yet the species is often assumed to warrant little conservation concern. In Kruger National Park, population densities have shown a worrying decline since they were last assessed 10 years ago. Pretoriuskop, which was previously estimated to have the highest density in Kruger, now has one of the lowest leopard densities in the central and southern parts of the park.
Lucy Smyth (iCWild, Panthera), PhD candidate and project contact
Prof Justin O’Riain (iCWild), Supervisor
Dr Guy Balme (Panthera), Supervisor
Dr Neil Midlane (Singita), Supervisor
Ross Pitman (Panthera), Data science
Joleen Broadfield (Panthera), Data science
Collaborators and funders
Click here to view iCWild's full output list, and use the filter tool to view outputs related to Leopard Camera Trapping.