Sabine Pfeffer

Swedish University of Agricultural Sciences, Umeå, Sweden

Researchers, conservationists and stakeholders all over the world are confronted with an important question: What is a reliable, time and cost efficient method to monitor changing wildlife communities? Which census method is most suitable for reliable population estimates? Is there one method that would work well for several species? To investigate these questions, our article published in Remote Sensing in Ecology and Conservation compared dung counts and camera trapping to estimate population densities of moose (Alces alces) and roe deer (Capreolus capreolus) in northern Sweden.

For camera trapping, we tested the random encounter model (REM) which can estimate densities without the need to recognise individual animals. However, the REM requires the estimation of several complex parameters in comparison to the rather traditional and simple dung count method. To make the REM easier and more attractive to be used by citizen scientists, field managers and non-governmental organizations, we tested different simplification options in terms of estimates of detection distance and angle (raw data versus modelled) and of daily movement rate (camera trap-based versus telemetry-based).

To estimate detection distance from the pictures, we marked distances of 5 m, 10 m, and 15 m in front of the camera’s centre with small red ribbons in trees. The example picture of a male roe deer shows these red distance markers and gives an idea of how certain distances and angles can be calculated via trigonometry.

Example picture (Reconyx HC 500) of a male roe deer taken with red distance markers at 5, 10, and 15 m. The white dashed line in the centre field of view was used as reference point for distance calculations, r represents distance between the camera and animal (white dotted line) and α the angle of first detection (white dotted angle).

Often it is difficult to find dung and/or specify the correct species and some of these challenges emerged in our results. We found that dung counts and camera trapping estimated similar densities for moose, which were also comparable to independent estimates from local managers. This follows our expectations since moose dung is large and can be identified and detected with relative ease. In contrast, dung counts appeared to underestimate roe deer densities when compared to camera trapping. Dung of roe deer is relatively small which decreases its detectability. The large overlap in dung morphology among roe deer, red deer and fallow deer may also influence density estimates from dung counts. This highlights the value of camera trapping in multi-species ungulate communities. Estimates of detection distance and angle from modelled versus raw camera data resulted in nearly identical outcomes. However, the telemetry-derived daily movement rate for moose and roe deer resulted in much higher density estimates than the camera trap-derived estimates.

We suggest that the REM with the simplifications that we developed may be a robust complement to dung counts when monitoring ungulate communities, particularly when dung is difficult to find and/or difficult to assign to species. Moreover, we show that the parameters that are needed to run an REM can be estimated from the camera trap images in a relatively simple way. This means that the method holds great potential for citizen science-based programs (e.g., involving hunters) that can track the rapidly changing European wildlife landscape. We suggest to include camera trapping into management programs, where the analysis can be verified via web-based applications.