Limitations to recording larger mammalian predators in savannah using camera traps and spoor

Traditionally, spoor (tracks, pug marks) have been used as a cost effective tool to assess the presence and in some cases the individual identity of larger mammals. Automated camera traps are now increasingly utilised to monitor wildlife, primarily as the cost has greatly declined and statistical approaches to data analysis have improved. While camera traps have become ubiquitous, we have little understanding of their effectiveness when compared to traditional approaches using spoor in the field. Here, we 1) test the success of camera traps in recording a range of carnivore species against spoor in realistic field settings (dirt roads in a South African wilderness reserve); 2) ask if simple measures of spoor size taken by amateur volunteers are likely to allow individual leopards to be tracked in the field and 3) for a trained tracker, ask if this approach may allow individual leopards to be followed with confidence in savannah habitat. We found that camera traps under‐recorded mammalian top and meso‐carnivores when compared with spoor in the field, with camera traps more likely to underrecord the presence of smaller carnivores (civet 64%; genet 46%, Meller's mongoose 45%) than larger (jackal sp. 30%, brown hyena 22%), while leopard was more likely to be recorded by camera trap (all recorded by camera trap only). We found that amateur trackers could be beneficial in regards to recording leopard presence; however the large variance in measurements of spoor taken by volunteers suggests that this approach is unlikely to allow the collection of further information about individual leopards. Nevertheless, the use of simple spoor measurements in the field by a trained field researcher increases their ability to reliably follow a leopard trail in difficult terrain. This allows researchers to glean further data on leopard behaviour and habitat use without the need for complex spoor analysis.