Relationship between wildfire, salvage logging, and occupancy of nesting territories by northern spotted owls.

摘要

The northern spotted owl (Strix occidentalis caurina) is one of the most intensively studied raptors in the world; however, little is known about the impacts of wildfire on the subspecies and how they use recently burned areas. Three large-scale wildfires in southwest Oregon provided an opportunity to investigate the short-term impacts of wildfire and salvage logging on site occupancy of spotted owls. We used Program MARK to develop single-species, multiple-season models of site occupancy using data collected during demographic surveys of spotted owl territories. In our first analysis, we compared occupancy dynamics of spotted owl nesting territories before (1992-2002) and after the Timbered Rock burn (2003-2006) to a reference area in the south Cascade Mountains that was not affected recently by wildfire. We found that the South Cascades had greater colonization probabilities than Timbered Rock before and after wildfire ( beta =1.31, 95% CI=0.60-2.03), and colonization probabilities declined over time at both areas ( beta =-0.06, 95% CI=-0.12 to 0.00). Extinction probabilities were greater at South Cascades than at Timbered Rock prior to the burn ( beta =0.69, 95% CI=0.23-2.62); however, Timbered Rock had greater extinction probabilities following wildfire ( beta =1.46, 95% CI=0.29-2.62). The Timbered Rock and South Cascades study areas had similar patterns in site occupancy prior to the Timbered Rock burn (1992-2001). Furthermore, Timbered Rock had a 64% reduction in site occupancy following wildfire (2003-2006) in contrast to a 25% reduction in site occupancy at South Cascades during the same time period. This suggested that the combined effects of habitat disturbances due to wildfire and subsequent salvage logging on private lands negatively affected site occupancy by spotted owls. In our second analysis, we investigated the relationship between wildfire, salvage logging, and occupancy of spotted owl territories at the Biscuit, Quartz, and Timbered Rock burns from 2003 to 2006. Extinction probabilities increased as the combined area of early seral forests, high severity burn, and salvage logging increased within the core nesting areas ( beta =1.88, 95% CI=0.10-3.66). We were unable to identify any relationships between initial occupancy or colonization probabilities and the habitat covariates that we considered in our analysis where the beta coefficient did not overlap zero. We concluded that site occupancy of spotted owl nesting territories declined in the short-term following wildfire, and habitat modification and loss due to past timber harvest, high severity fire, and salvage logging jointly contributed to declines in site occupancy.