Making the most of by‐catch data: Assessing the feasibility of utilising non‐target camera trap data for occupancy modelling of a large felid

Abstract

Camera traps are an increasingly used tool in ecology, having the ability to capture large numbers of photographic records in short survey periods. For many surveys the number of non-target records outweighs those of focal species, making them a potentially rich and often under-utilised data source. Occupancy analysis of non-target data represents a potential way to optimise survey output, whilst increasing 'return on investment'. This study assessed the feasibility of using non-target data from a Hartmann's mountain zebra Equus zebra hartmannae survey in Gondwana Canyon Park, southern Namibia, for occupancy analysis on leopard Panthera pardus. Using a survey design with 15 camera traps at water sources, 26 leopard events were detected over 72 days. Model fit was adequate and produced a modelaveraged occupancy of 0.64 (S.E. 0.36), and a detection probability of 0.24 (S.E. 0.07).Whilst there was a lack of precision in the final occupancy estimate, the study provided valuable pilot data for future surveys. The results highlight the ability of camera traps to obtain information-rich datasets, which, when properly archived, can be used for providing information on a number of ecological topics, ranging far beyond that which the traps were originally deployed for.

Community-Based Monitoring of Jaguar (Panthera onca) in the Chinantla Region, Mexico

“…The length of our survey (65 days) was appropriate to ensure that occupancy (occupied or not occupied) would not change throughout the survey period (Edwards et al., 2018; Mackenzie et al., 2018; Wang et al., 2019). We found the percentage of forest cover, the survey effort, the capture rate of prey, and NDVI were the variables that best explained the jaguar habitat use.…”

Maximising camera trap data: Using attractants to improve detection of elusive species in multi-species surveys

“…Even though reduced activity is immediate in some species, we also demonstrate that this effect lasts only 1–2 days before levelling off. Given these results, accurate reporting of camera trap protocols is particularly important in the era of big data, where camera trap data from a variety of locations are being used to ask questions about species that were not originally the targets of the survey [66]. While the use of attractant can increase our understanding of species that have not traditionally been the primary focus of research, practitioners should consider that some species may have lower detection probabilities or reduced activity when attractants are deployed before making inferences about their ecological and conservation requirements.…”

Designing an occupancy framework to monitor an endemic rainforest duck: Methodological and modelling considerations

“…First, we estimated the optimal number of replicate surveys ( K ) needed and the number of survey sites ( s ) to achieve an occupancy estimate with a precision of 0.05. This was done following the framework provided by Mackenzie and Royle (2005) and also successfully applied by Edwards et al (2018). We input our ψ and p estimates into Table 1 of Mackenzie and Royle (2005) to estimate K and used equation 3 to estimate s which is: where, s = number of sites …”

“…With this backdrop, we explored the use of occupancy models to assess and monitor the species in the lowland remnant rainforests of Dehing Patkai Wildlife Sanctuary in Upper Assam, India. Occupancy models provide rigorous framework and hence have been widely used across species and habitats (Edwards et al 2018). It was fitting to use an occupancy framework for our study as it requires lower number of survey sites than abundance surveys and hence are less expensive and labour intensive (Mackenzie et al 2006; Edwards et al 2018).…”