Decisions on camera trap data analysis for REM density estimations

[Yorgos Iliopoulos]

Hello 

We have finished field work and successfully took data from 47 cameras for REM use to estimate red deer and wild boar densities. 

1) When calibrating the cameras in the field I used the calibration stick at measured distances from each camera where detection was tested to be closely to 100% and there was a good view of the animals with no major obstacles,  thus defining the detection zone where cameras are working well and then doing the calibration at this area. 

However on some occasions during the operational phase deer pass behind this "line" and the "pre-defined" detection zone  (which was marked in the field). Should I include those events during annotation, but without track and digitize the corresponded animal paths, so the encounter is registered,  but not using those passes for speed and other parameter estimation? or should omit those detections completely?. The density of red deer is low so I think it is crucial to include as many detections as possible to the analysis.

2) The cameras were set in a period of 40 days as the terrain was very difficult and only one person was involved. The first camera was set at 15/1/2026 and the last was collected at 22/6/2026. 

Should I use data from only the "common" operational period or from the whole period where each of those cameras was operational ? Average duration of camera operation was 114 days (sd 24.58). 

3) Is that period of time an issue for estimating the deer density as it could be considered very open or is it acceptable?

Again, as the deer density is low it is good to include as much as possible red deer events in the analysis but of course without compromising accuracy of the estimation too much

Thank you!