An oasis in the desert: The potential of water sources as camera trap sites in arid environments for surveying a carnivore guild

Abstract

Discussions regarding the importance of accounting for detection probability have long been present in ecological literature. Various studies have demonstrated the influence of survey design on detection probabilities, and whilst the placement of camera traps along roads is a commonly used survey design, it has shown to be biased towards certain species. In arid environments, water sources have the potential to be efficient sites for camera trap placement. We compared the influence of a water source camera trap survey design on the detection probabilities of a guild of seven carnivore species, in comparison detection probabilities from camera traps along roads, on arid, commercial farmland in southern Namibia. Results showed detection probabilities for all species to be higher at water, with the water source design producing shorter latencies of detections and higher naive occupancy estimates for most species. However, for species with unique markings, the 2 water source design produced lower proportions of images suitable for individual identification. As detection probabilities of all species were influenced in a positive manner, we suggest placing camera traps at water sources in arid environments to be an effective survey design. However, for surveys requiring individual identification, placing camera traps along roads may be more suitable. Highlights Comparison of two camera trap survey designs for detecting a guild of seven carnivores in Namibia. Camera traps placed at permanent water sources compared to traps placed along roads. Detection probabilities higher for all species, latency until detection first shorter and naive occupancy estimates higher for most species at water sources. Proportions of images suitable for individual identification higher from road camera traps. Water source camera trap survey design effective for multiple carnivore species in arid environments, unless individual recognition is required, for which road survey design is more suitable.

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

“…Technology and analysis techniques for camera-trap data [4–6] in biodiversity monitoring [7, 8], regional occupancy and species richness surveys [9, 10], and single species and community ecology studies [11–14] have evolved rapidly. Because many threatened species have extremely low detection probabilities [3, 15, 16], it is common practice to increase sample sizes and ensure statistically robust results by placing cameras at species-preferred sites [16–19] or extending survey duration [20, 21]. As current conservation efforts shift away from focusing on large, charismatic species to include smaller species that have not traditionally been the primary focus of research [1, 14], the problem of limited sample size is further compounded by naturally lower detection probabilities of small bodied species [22], and the influence of interference competition from larger predators [23].…”

Activity patterns and resource partitioning: seven species at watering sites in the Altun Mountains, China

“…Desert animals actively use watering sites which are also high risk areas for predation (Valeix et al, 2009). Camera trapping at these sites is a practical approach for investigations into the animals of arid regions (Edwards et al, 2016). Moreover, the use of camera trapping data of animal behavior and activity could be a feasible method to study community structures and interspecific competition, and to evaluate overlap or partition of niches.…”

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

“…Despite the relatively low number of leopard events, model-averaged detection probability was estimated at 0.24 (± SE 0.07), which is above the recommended threshold of 0.15 required to provide robust occupancy estimates (Mackenzie et al, 2002). Such a relatively high detection probability is likely to result from placing camera traps at artificial water sources which have been shown to be effective in surveying a variety of carnivore species in an arid environment (Edwards, Gange & Wiesel, 2016). It is recommended that camera traps are placed in random and passive sites for occupancy surveys (Rovero et al, 2013).…”

“…Surveys placing camera traps at water sources have the potential to record a wide variety of species, particularly in arid environments, where surface water is scarce (Edwards, Gange & Wiesel, 2016). They therefore represent particularly promising datasets for which to investigate the potential of utilising non-target data, and given the diversity of species visiting water sources, a chance to run occupancy models on species very different from those the camera traps were originally deployed for.…”