Search for a command to run...
Forecasting species’ responses to long-term trends of ecosystem degradation is paramount in shifting disturbance regimes, especially for forest-associated species undergoing steep, range-wide population declines. Large-scale passive acoustic monitoring (PAM) networks provide an unprecedented opportunity to predict how such species may respond to environmental and management pressures through time and over broad spatial scales. Using detections from an extensive PAM network spanning over 4,000 autonomous recording units, we predict the distribution and drivers of habitat use for the varied thrush (Ixoreus naevius), a widespread but declining temperate forest-interior songbird species. We employ boosted regression trees to create fine-scale, range-wide distribution maps of the varied thrush under current and future climate and habitat conditions, providing a strong test case for evaluating how acoustic data, particularly from bycatch recordings, can forecast shifts in habitat suitability. Despite incidental data collection, we detected 5.6 million apparent instances of the varied thrush and manually validated presence at 2,595 automated recording units, with our final model achieving high predictive performance with low deviance (AUC = 0.92). Climatic water deficit emerged as the strongest predictor of occurrence, underscoring the species’ vulnerability to warming and drying conditions. Projected distributions of varied thrush under future high- and intermediate-warming climatic scenarios (simulating RCP8.5 and RCP4.5) revealed broad-scale contractions of suitable habitat by the year 2100, with 70% and 63% of current habitat decreasing in suitability, respectively. Low elevation, vertical forest structure, and dense canopy cover were also strong predictors of occurrence, though to a lesser degree. These results highlight the power of passive acoustic monitoring to guide regional adaptation and forest management frameworks for maximizing biodiversity inference amid rapid ecosystem change.