Quantifying levels of animal activity using camera trap data

Rowcliffe and colleagues extended activity analysis beyond pattern shape to estimate the proportion of time animals spend active, a quantity needed for the Random Encounter Model and for behavioral ecology. By fitting flexible circular models to detection-time distributions and assuming animals are fully active around their peak, they derived an estimate of overall activity level. This activity-level parameter feeds directly into recognition-free density estimators and improves their accuracy. The method is implemented in the activity R package and is widely used to characterize how much of the day a species is on the move. The paper is a key methodological reference linking camera-trap timestamp data to both behavior and density estimation for unmarked species.