Can LiDAR predict acoustic detection distance in heterogeneous forest environments?

The estimation of the distance at which an animal can be detected is important information in wildlife monitoring. To account for imperfect detection of individuals, frameworks like distance sampling or spatial capture-recapture rely on a detection function of the distance. In acoustic monitoring, the detection distance is influenced by sound attenuation between the source and the receiver. This attenuation results from travel of sound waves in the air and the obstacles in their path. In forest environments, vegetation significantly contributes to sound attenuation, and its effect varies with vegetation structure. LiDAR data provide fine-scale information on both horizontal and vertical structure of vegetation. These data are often available for large areas through public online geographic information libraries and can be processed in numerous ways. In this study, we have conducted broadcasting tests along linear transects to model the attenuation due to vegetation as a function of frequency and distance, weighted by LiDAR-derived data. The objective of this work is to adjust an attenuation coefficient to the LiDAR-weighted distance in order to predict detection distance in an array of recorders.