Risky business or simple solution – Relative abundance indices from camera-trapping

The wild life of tick-borne pathogens

“…Camera traps are a reliable method for measuring the relative abundance of species provided that sampling is rigorously standardized 167 and that differences in sampling effort and detection distance are accounted for 54 , as was the case in our study. Yet, our survey was imperfect in detecting carnivores.…”

“…Avoiding or pruning of vegetation in the FOV may reduce the difference in detectability between sites to some degree, but not entirely. Differences in detection distance in different habitats can be especially problematic when capture rates from camera traps are used to study habitat selection or other variables which are linked to the habitat 167 . Estimating EDD per habitat type and correcting capture rates accordingly can reduce bias in habitat selection studies that use camera trap data.…”

“…This problem can be overcome by averaging measurements over multiple camera locations. The assumption of random movement of animals compared to the camera position is essential for estimating relative abundance 167 as well as for estimating density using the random encounter model (REM) 170,184 . Studies using these methods rely on camera placement in random positions, which ensures random movement of animals relative to the camera position.…”

“…For animals that cannot be distinguished individually, camera traps are often used to derive relative abundance estimates, reasoning that photographic capture rates -the number of visits recorded per unit time -of a species will be proportional to its abundance 166 . However, the use of capture rates as relative abundance indices has received criticism, as capture rates are influenced by other factors besides animal abundance, which can lead to important biases (e.g., Sollmann et al 167 ). Some of these factors are related to non-random movement of the focal species, such as the preferred use of trails.…”

“…However, most factors that bias capture rates are related to the PIR sensor of the camera trap: camera type, camera placement, animal size, vegetation density, temperature and relative humidity all influence the effective range of the sensor 54,168 . Some of these factors can be dealt with using adequate sampling design, for example, using the same camera model and thus the exact same PIR sensors 164 , the same sampling period, similar climate, and random sampling points 167,168 . However, when the goal is to compare different species, or the same species in different habitat types, biases may remain a problem.…”