Abstract
Climate change has profoundly impacted global weather patterns, intensifying extreme events and shifting seasonal trends. Species distribution models (SDMs) are crucial for assessing ecological effects, particularly in forecasting climate-induced shifts. This study examined the distribution of Diploknema butyracea by refining and spatially thinning occurrence data to a 1 km resolution, reducing an initial 413 records to 80 unique localities. Nineteen bioclimatic variables from the CHELSA dataset were analyzed, retaining only those with variance inflation factor (VIF) values below 10. Climatic conditions from past, present, and future scenarios were incorporated to assess habitat suitability over time. To ensure model reliability, block partitioning cross-validation was applied, identifying the LQ model (Rm = 1.5) as the most effective. Key determinants of habitat suitability included bio08 (mean temperature of the wettest quarter), bio06 (minimum temperature of the coldest month), bio07 (temperature annual range), bio12 (annual precipitation), and bio18 (precipitation of the warmest quarter). During the Last Glacial Maximum (~20,000 years ago), suitable habitats were largely confined to the southern lowlands of Nepal. By the mid-Holocene, rising temperatures and increased precipitation enabled expansion into mid-hill regions. The present distribution shows high suitability across mid-hills and lowlands, while future projections (SSP3-7.0) indicate declining suitability in the lowland Terai due to increasing temperatures and precipitation, with a slight increase at higher elevations. These findings underscore the importance of climatic stability in shaping the distribution of D. butyracea and offer valuable insights for conservation planning and forest restoration in the face of climate change.