The American Goshawk (Astur atricapillus) is a forest-dwelling raptor on the Fremont-Winema National Forest in south-central Oregon, where pre-project nest surveys are conducted prior to management. Recent budget constraints and reductions in U.S. Forest Service seasonal personnel have increased pressure on wildlife monitoring programs. This thesis presents an approach for improving survey efficiency by incorporating technology. A distribution model was developed using predictor variables derived from publicly available remote sensing data of important nest site characteristics for goshawks. The model was refined iteratively across two field seasons (2024–2025) using confirmed and historic nest locations and over 900 hours of broadcast acoustic survey effort. Vegetation type was the largest contributor to model predictions (43.1%), while distance to streams showed the highest permutation importance (49.7%), indicating proximity to riparian features as the strongest independent predictor of nest occurrence. Eleven of twelve detected nests occurred in areas with high model-predicted suitability scores, and detection success was substantially higher when survey effort was concentrated in high-suitability areas. Autonomous recording units (Wildlife Acoustics Song Meter Minis) were deployed across project areas during the 2023 and 2024 field seasons and analyzed using detection software with trained observer validation. ARUs with positive goshawk detections were significantly closer to active nest sites than those without detections. Results support deploying ARUs in high-quality habitat early in the breeding season to guide subsequent survey effort, and in low-suitability areas as a lower-cost alternative to in-person surveys. Together, these results demonstrate that distribution modeling and autonomous recording technology can increase the efficiency of goshawk monitoring programs, enabling wildlife managers to maintain survey quality while reallocating limited resources.
