Forest restoration focused on enhancing resilience is necessary to prevent widespread, fire-driven conversion of many forests to shrubland. However, restoration planning is challenged by limited resources to monitor and understand biodiversity responses to changing disturbances and management intervention. We combined bioregional-scale passive acoustic monitoring, a machine-learning tool, BirdNET, and management-relevant remotely-sensed habitat data to (1) map occurrence of ten avian indicator species across 25,000 km2 of the Sierra Nevada, California, USA, (2) examine the impact of fire history on patterns of occurrence, and (3) identify hotspots of overall richness and richness of fire-averse and fire-affiliated species. We demonstrate that species occurrence can be mapped in the context of policy-ready habitat information that is congruent with the information used by managers in restoration planning. The nuanced responses to fire history we found among indicator species suggest that necessary forest heterogeneity could be enhanced by the restoration of pre-colonial disturbance regimes. Monitoring complementary suites of indicator species with ever-improving bioacoustics technology, and relating their occurrence to relevant habitat metrics, has the potential to transform restoration planning by providing managers with holistic, high-resolution, whole-ecosystem information, thus facilitating adaptive management in an era of rapid environmental change. This dataset includes: RawHabitatData_Rasters.zip - Rasters of habitat data used in analysis PredictedOccupancy_Shapefiles.zip - Shapefiles of species occurrence predictions (model output) Brunketal2023_PolicyReadyInfoCode.R - Code to reproduce analyses in the manuscript ModelingData.zip - Data to reproduce occurrence models PredictionData.zip - Data to predict at interpolated locations