Mountain lion habitat use relative to human activities in and around Redwood National and State Parks of northwest California

Substantial increases in mountain lion-human conflicts in the western United States have created a need for quantitative assessment of lion (Puma concolor) resource selection at multiple scales. Using a fixed kernel estimator, multi-year seasonal home ranges were calculated for six adult male and three adult female radio-collared lions monitored from September 1998 to June 2001 in and adjacent to Redwood National and State Parks (RNSP). Mean multi-year wet season home ranges were not significantly different in size from dry season (P = 0.87). At the patch level, based on stepwise logistic regression models developed for nine individuals, lions were located significantly closer to streams (five of nine lions), low use roads (seven of nine) and medium use roads (five of nine), and farther from high use roads (four of nine) than random. All lions (nine of nine) were located significantly farther from active timber harvests than random. Three of five lions were located significantly farther from low use roads during the dry season, or period of greatest human activity in RNSP and adjacent private timberlands, than they were during the wet season, or period of least human activity. At the landscape level, the percentage of mature forest was significantly less (X2= 3.18, P = 0.07) in 50% fixed kernel home ranges than in 95%. Lion 50% fixed kernel home ranges were more fragmented due to significantly (X 2 = 6.68, P = 0.04) greater forest — non-forest edge density and a greater number of forest habitat patches per unit area, compared to 95% fixed kernel home ranges. Humans killed three of six collared male lions during the 34 months of the study. Two male lions, whose home ranges were adjacent to the deceased lions', significantly shifted their home range 18 and 21 km south within two months following these mortality events into the recently vacated territories. In both cases, such that the shift was away from RNSP lands and the home ranges occupied one year post mortality were almost entirely on adjacent private timberlands. The habitat in home ranges occupied by the two transitioning lions following mortalities of their neighbors was characterized by greater edge density of forest and non-forest habitat types. In order to minimize lion-human conflicts in RNSP, visitor attractions should not be constructed in areas dominated by young forest with a great amount of non-forest edge that provide optimal conditions for lion prey. Low and medium use roads, if used by visitors on foot, and riparian areas should receive signage indicating the increased likelihood of encountering a lion.