The influence of camera trap flash type on the behavioural reactions and trapping rates of red deer and roe deer

Maik Henrich, Sandra Niederlechner, Max Kröschel, Stefanie Thoma, Carsten F. Dormann, Florian Hartig & Marco Heurich

Abstract

Camera traps have become an important tool in wildlife monitoring. However, an issue in interpreting their data in statistical analyses of population densities, demography or behaviour is that the probability of detecting the target animals and their behaviours may vary depending on environmental and methodological factors. A specific problem is the type of flash used in the camera trap, as animals may react differently to different flash types and change their avoidance or habituation behaviour accordingly over time. Here, we provide the first systematic test of the impact of flash type in studies of red deer (Cervus elaphus) and roe deer (Capreolus capreolus), based on an analysis of behavioural responses to white, standard infrared and black flashes during 900 camera trap deployments in the Bavarian Forest National Park and the Northern Black Forest. The results revealed that both deer species were more likely to react to standard infrared than to black flash, but trigger delays prevented comparisons to white flash. Red deer reacted more frequently to camera traps than did roe deer, and responses were more common in the Northern Black Forest than in the Bavarian Forest National Park, where hunting is severely restricted. Contrary to our expectations, camera trapping rates of free‐ranging deer did not significantly decline over time for any flash type or species. Despite the lack of evidence for avoidance behaviour, we recommend the use of black flash for behavioural studies of deer populations to minimize the risk of introducing a source of disturbance whereas infrared and white flash are equally suitable for determinations of demographic parameters.

What the paper shows and why it matters

(AI-generated)

Camera traps are considered non-invasive, yet their night-time flash may disturb animals. After reports that white flash disturbed animals, it has increasingly been replaced by infrared (IR) light, in the belief that these longer wavelengths disturb less, although many nocturnal mammals can see IR light. Reactions also matter for the data: animals that avoid a camera, or get used to it, change how often they are detected. That would violate the assumption of capture–recapture methods and of the random encounter model (an estimator for unmarked populations) that cameras do not affect animal movement. How far flash type changes behaviour and trapping rates is unclear and probably species-specific; the study addresses this for red deer and roe deer.

The study finds that the flash type of a camera trap changes what deer do at the moment of detection, but not how often they are detected over time. It compared white, standard IR and black flash (IR of longer wavelength) in about 900 deployments in two German forest areas (Bavarian Forest National Park, Northern Black Forest). Red deer and roe deer looked at the camera and fled more often at standard IR flash than at black flash, whose reactions resembled those in daytime photos without flash; the study reads this as indicating that deer can perceive standard IR light, or at least the red glow of its LEDs, and suggests that they do not see black flash. White flash could not be compared because of trigger delays. Trapping rates of free-ranging deer did not decline for any flash type or species, so flash-specific avoidance was not detected.

**Figure 2.** Proportions of events during which animals looked at the camera trap and exhibited flight responses. From Henrich et al. (2020), *Remote Sensing in Ecology and Conservation*, licensed under CC BY-NC-ND 4.0. Click the image to enlarge.

The study concludes that all three flash types can be used for studies of population parameters, with the choice following range and photo quality, whereas black flash suits behavioural studies because it disturbs least. It suggests that a brief, harmless disturbance may have no lasting effect, and that individual red deer rarely return to the same site, which makes avoidance unlikely but also makes changes in visitation hard to detect. The behavioural effect of white flash was not measured, so its suitability rests on trapping rates alone.