Camera-shy wildlife and capture bias: behavioral response to TrailMaster 1550 camera traps

Background/Question/Methods Remotely triggered cameras (camera traps) have long been used by wildlife biologists to gain insight into wildlife ecology. From establishing species presence/absence to studying feeding ecology, camera traps are a powerful field-biology tool. However, there has been much conjecture, and little scientific research, addressing possible detection biases caused by wildlife response to the camera units themselves. Traditional camera-trapping methodologies employ 35mm flash cameras; these cameras, when activated, emit noise and produce visible light that may alter wildlife behavior. It is currently unknown the extent to which wildlife are deterred or avoid capture by these systems as a result. By monitoring a ubiquitous camera trap, the TrailMaster 1550, with a silent and minimally invasive quasi-video technology (Reconyx camera-trap system) this project provides insight into potentially data-biasing behavioral responses of wildlife. Reconyx cameras were placed on game trails at three sites in the foothills of the Santa Cruz Mountain range in California for two phases of data collection. In the first phase, the Reconyx system photographed animals for two months and recorded baseline visitation frequencies and behaviors. In the second phase, TrailMaster units were set up, within the field of view of the Reconyx cameras, and left for the subsequent two months. Reconyx images were analyzed for differences in behavior and visitation frequencies between phases I and II.

Results/Conclusions The data show that, for some species, there is a large site-dependent difference in visitation rates between the two phases of this study. For example, mule deer (Odocoileus hemionus) image captures were 3.5 times greater in the absence of TrailMaster cameras when compared to the period when the TrailMaster units were present. In addition, there is a measurable species-dependent increase in vigilance behavior that is associated with the TrailMaster system. Broadly, this study has found that there are identifiable differences in behavior correlated with the presence of the TrailMaster camera systems. These data provide a starting point for re-evaluation of the voluminous historical data collected using TrailMaster and similar camera-trapping systems. In addition there are a variety of situations where TrailMaster systems are optimal or cannot be replaced; in these instances the results of this experiment can help inform future study design and best practices.