CENSUSING BOBCATS USING REMOTE CAMERAS

Abstract

We estimated bobcat (Lynx rufus) density for 3 different locations in northern California using active infrared-triggered cameras. Using differences in pelage pattern as well as other physical characteristics, we identified individual bobcats from photographs, and used mark-recapture techniques to estimate population density. Camera density affected the precision of population estimates. The same population was estimated using camera densities of 0.5, 1, 2, 4, 6, and 8 cameras ⋅ km-2. Higher camera densities resulted in more captures and recaptures of bobcats and, conse- quently, in more precise density estimates. Similarly, the number of photo-captures and recaptures increased with increasing study duration. Increasing the area sampled resulted in the capture of more individuals but did not increase the percentage of recaptures. While some locations captured multiple bobcat photographs (e.g., 15 at 1 station), these photos tended to be recaptures of the same individual. There were no more than 2 individuals photo-captured at any 1 camera location. Bobcat density varied among habitat types as predicted. We estimated density as 0.27 bobcats ⋅ km-2 (s = 0.16) overall in an area in the northern Sacramento River Valley and as 0.35 bobcats ⋅ km-2 (s = 0.56) in a steep and rocky canyon within the area. At a 3rd site in the Coast Range, the estimate was 0.39 bobcats ⋅ km-2 (s = 1.44). Bobcats were more diurnal where human activity was less common. In addition, photo-capture was significantly higher along roads and trails without an attractant than it was off-trail with an attractant.