Abstract
According to the IUCN Red List, the endangered eastern North American tree species, butternut (Juglans cinerea L.), has been negatively impacted by an introduced pathogen and declining habitat, hindering conservation efforts. We developed an ensemble model to project spatiotemporal dynamics in suitable habitats for butternut across its native eastern United States (US) range under contrasting emissions scenarios. Our model integrated six algorithms weighted by predictive performance on test data. Predictor variables emphasized temperature, precipitation, topography, and humidity variables influential for butternut based on ecological literature. Across four future periods through the end of the 21st century, the ensemble model projected habitat gains, losses, or stability depending on geographic location and climate model, highlighting variability and uncertainty. While some ensemble projections showed poleward and elevational range shifts per previous climate impact studies, incorporating dispersal limitations in future models could avoid under- or over-estimating shifts. Precipitation seasonality and slope aspect emerged as highly influential variables for projected distribution. Slope gradients may provide local climate refugia amid regional drying. Overall, these results demonstrate complex species-climate interactions across spatiotemporal scales, underscoring the need for adaptive and location-specific conservation strategies attuned to ecological intricacies to ensure biodiversity amid the uncertainties of global change.