Throughout most of their range in the Sierra Nevada range in California, populations of the California spotted owl (Strix occidentalis) are declining. Previous studies have associated high (greater than 70%) levels of canopy cover with territory occupancy, owl survival, and higher reproduction. High cover forests, however, are also associated with stand densities and fuel loads that are prone to high-severity fires and mortality from drought and insects. We analyzed the forest structures associated with owl occupancy across 316 territories found in four large (cumulatively 340,000 ha) study areas. Three of these areas were the locations of long-term demographic studies that provided at least 15 years of data on occupancy, reproduction, and survival. The fourth was the entire Tahoe National Forest where we had data on the locations of known occupancy. We used airborne LiDAR to measure canopy cover, distribution of overstory tree heights by strata, and opening sizes and frequencies to analyze vertical and horizontal heterogeneity. We analyzed changes in forest structure continuously with distance from nest sites as well as three set areas: nests (4 ha), protective activity centers (PACS) (121 ha), and territories (314 ha).
Multivariate modeling of habitat variables found territory occupancy was positively associated with the density of tall trees, especially those >48m and to a lesser extent trees 32-48m. The large canopies of these tall trees resulted in high canopy cover (typically >70%). However, owls avoided locations with high canopy cover in lower strata trees (2-16 m) for nests and PACs. Tall tree density gradually decreased out to about 500m (78.5 ha area) from the nest location, when structure became increasingly indistinguishable from the surrounding landscape. Large (>1000m2) gaps were avoided near nests, but were found in the surrounding PACs and territories. With increasing distance from the nesting areas, there was increasing structural heterogeneity among PACs and territories. Multivariate modeling did not find any structural variables or measures of heterogeneity associated with owl reproduction. In our analysis owls appear to select nesting areas for a threshold density of nearby tall trees, but once that threshold is met, differences in forest structure do not affect reproductive success. Our results suggest that managers should work to protect and develop islands of tall trees, but can reduce understory cover and ladder fuels at distances greater than 500m from nests to improve forest resilience.