Forest ecosystem restoration: Initial response of spotted owls to partial harvesting

• We compare habitat selection by spotted owls before vs. after partial harvesting.
• Most owls used stands scheduled for harvest in proportions less than availability.
• Most owls used harvested stands in proportion to availability.
• We model factors that distinguish between used and unused harvested stands.
• Harvest area, nest distance and midstory basal area influenced probability of use.

Conservation planning for spotted owls (Strix occidentalis) hinges upon retaining late-successional and old-growth forests. This strategy is to be supplemented over time by creating structural conditions found in such forests using innovative silviculture in less well-developed forests. Recent research indicates that spotted owls often hunt for prey or may nest in relatively young or mid-seral forest stands that were thinned or partially harvested in previous decades, but little information has been available to evaluate short-term direct responses (<5 year) by spotted owls to such practices. We used selection ratios to compare the frequency of nocturnal use by radio-tagged northern spotted owls (S.o. caurina) and California spotted owls (S.o. occidentalis) ⩽2 years before and ⩽2 years after 150 forest stands were thinned or partially harvested within 1200 m of nest sites of 19 owl home ranges in 5 study areas in western Oregon and northern California. We used logistic regression to investigate habitat and environmental factors that distinguished between 89 stands that were used and 115 stands that were not found used by radio-tagged owls for up to 2 years after treatment via a broad range of partial-harvest or thinning prescriptions within 2400 m of nest sites. Before harvest, radio-tagged owls generally used stands scheduled for harvest treatment in proportions significantly less than availability. After harvesting, selection ratios increased (n = 4), remained the same (n = 4), or decreased (n = 2) among 10 owl pairs for which we acquired sufficient telemetry data both before and after harvesting. Across all owls and all post-harvest conditions, the overall selection ratio increased after harvesting, suggesting that many of the harvests were benign or may have resulted in improved habitat. The probability of use of thinned or partially-harvested stands increased with harvest-unit size, decreased with distance from nest sites, and varied with the intensity of harvest and among forest types as represented by study areas. We found only limited evidence for a positive effect of retained basal area of large trees (⩾66 cm diameter at breast height [dbh]), probably because many treated stands contained no such large trees prior to harvest. We found a quadratic relationship with retained basal area of mid-story conifers (10–65 cm dbh), such that harvested stands that contained 25–35 m2/ha basal area of such mid-story trees were more likely to be used, holding other factors constant at their means. We also found evidence for a positive influence of proximity to riparian zones on probability of use of harvested stands. Although we did not obtain information on prey abundance or foraging efficiency, our study suggests that judicious applications of partial-harvest forestry, primarily commercial thinning, have the potential to improve foraging habitats for spotted owls.