Monitoring Mountain Lions in the Desert Southwest: Spatial Density Estimation and Results of a Novel Hair Sampling Technique

Large carnivore monitoring is a difficult endeavor, and mountain lions (Puma concolor), with their nondescript pelage, secretive nature, and vast home ranges, are no exception. We conducted a 2-part study on a 528 km² study area in southwestern New Mexico to (1) evaluate a new technique for noninvasive collection of genetic samples and (2) obtain a density estimate for the population under study. We modified Belisle foothold traps to collect hair from free-ranging mountain lions and obtained 7 hair samples in 1618 trap nights. All samples yielded enough DNA for species-level identification and 6 of the 7 (85.7%) provided individual genotypes. We monitored the hair traps with remote cameras, which indicate 13 occasions in which mountain lions passed within 2 m of a hair trap without triggering it. From these records, we have developed strategies to reduce misses and increase hair sample collection rates. We additionally combined photos from hair trap sets with photos from 25 independent cameras in a 100 km² subsection of the study area for use in a spatial mark-resight model augmented with GPS data from 3 collared individuals. Our resulting density estimate was 0.67 (95% BCI: 0.31–1.17) mountain lions/100 km², which we believe to be negatively biased due to high heterogeneity of detection probability among marked individuals. The study area hosts long-term mountain lion monitoring, so we recommend implementing the spatial models across historical and future data sets to evaluate population trends; continued use of the hair traps will build a data bank of genotypes and promote their use in other studies