|Abstract or Summary
- Several large wildfires in southwestern Oregon during the summers of 2001 and 2002 provided the opportunity to investigate the impacts of wildfire on northern spotted owls (Strix occidentalis caurina). I used radio-telemetry and demographic surveys to describe demographic performance and habitat selection of spotted owls in the areas burned by the Biscuit, Quartz and Timbered Rock Fires. Demographic surveys were conducted from 2003 – 2006 at the 3 fires. From September, 2004 – August, 2006, 26 spotted owls were monitored with radio-telemetry at the Quartz and Timbered Rock Fires and their surrounding areas.
I investigated differences in occupancy rates between the South Cascades Demography Area and the Timbered Rock Study Area from 1992 – 2006 using occupancy models in program MARK. Occupancy was similar at the Timbered Rock and South Cascades from 1992 – 2002 but occupancy declined rapidly following the Timbered Rock Fire when compared to unburned landscapes at the South Cascades. I also investigated the impacts of fire severity and habitat on occupancy at the Biscuit, Quartz and Timbered Rock Fires. Occupancy at all 3 fires declined from 2003 – 2006. Initial occupancy was positively influenced by the amount of roosting and foraging habitat with low severity burn within the core (β = 0.08, 95% C.I. = -0.02 – 0.17) and negatively influenced by the amount of hard edge within the core (β = -0.33, 95% C.I. = -0.77 – 0.10). Extinction rates increased in a curvilinear manner as the amount of unsuitable habitat within the core increased (β = 2.15, 95% C.I. = 0.25 – 4.05) and as the amount of edge increased (β = 0.20, 95% C.I. = -0.01 – 0.41). Colonization rates were positively influenced by the amount of nesting, roosting and foraging habitat that received a low severity burn within the core (β = 0.08, 95% C.I. = 0.02 – 0.15).
Demographic surveys were used to determine the number of young fledged per pair of spotted owls. I found no significant differences in productivity of spotted owl pairs in burned landscapes at the Biscuit, Quartz and Timbered Rock Fires and unburned landscapes at the South Cascades. Survival was estimated in program MARK using known fates modeling of radio-telemetry data. Annual survival rates of spotted owls that resided within the fire or had recently emigrated out of the fire were lower (0.64, 95% C.I. = 0.37 – 0.84) than owls that resided outside the fire (1.00, 95% C.I. = 1.00 – 1.00).
Annual home ranges of spotted owls in this study were on average 248.46 ha larger than home ranges observed in the same area prior to wildfire (t = -2.85, df = 32, p = 0.01). However, home ranges of spotted owls that resided inside the fire were not significantly different than owls that resided outside the fire (t = 0.72, df = 18, p = 0.48). Differences in home ranges of individual owls were best explained by the amount of hard edge within the 95% fixed kernel home range. Annual home ranges increased as the amount of hard edge within the home range increased (β = 30.71, SE = 2.65, p < 0.01).
Logistic regression was used to assess selection of habitats in relation to early seral forests. Nesting, roosting and foraging habitat with low, moderate or high severity burn was selected by spotted owls in post-fire landscapes. Furthermore, roosting and foraging habitat with a moderate severity burn was also selected. Three habitats were used in a similar manner to early seral forests including; roosting and foraging habitat with low or high severity burn and salvage logged areas. Non-habitat was the only habitat that was commonly avoided. Several abiotic factors were important in determining post-fire habitat selection. Owls selected areas closer to hard edges, perennial streams and lower in elevation than random locations.