In the context of global climate change, clarifying the effects of climatic factors on the potential distribution of forage grass species is critical for grassland development and ecological management. In this study, Phleum pratense was selected as the focal species, and an ensemble species distribution modeling framework was developed using the Biomod2 platform. After excluding highly correlated environmental variables, 17 ecologically meaningful predictors were retained to project the potential habitat suitability patterns of P. pratense and their future dynamics. The results indicate that the ensemble model achieved high predictive accuracy. Annual mean UV-B radiation, mean temperature of the driest quarter, and precipitation of the coldest quarter were identified as the primary climatic factors shaping the potential distribution of P. pratense. Under current climatic conditions, highly suitable habitats occupy a relatively limited and fragmented area, whereas low- and moderately suitable habitats predominate. Under future climate scenarios, increasing emission levels are associated with an overall contraction of moderately and highly suitable areas, accompanied by a continuous expansion of unsuitable habitats. In addition, the centroid of highly suitable habitats is projected to shift northwestward. These findings suggest that climate change may substantially alter the potential suitable range of P. pratense by modifying key climatic gradients and stress-period conditions. Overall, this study provides a scientific basis for the utilization of P. pratense germplasm resources and the sustainable development of plateau grassland systems.

Araucaria angustifolia (Bertol.) Kuntze, a critically endangered tree species, plays an irreplaceable ecological role in its native habitats. Under global climate change, identifying the drivers governing its geographic distribution and assessing climate-related threats can provide scientific guidance for the long-term conservation and habitat restoration of this species. In this study, a total of 287 valid occurrence records from 27 countries were compiled. Combined with 14 screened environmental variables, an optimized Maximum Entropy (MaxEnt) model was used to predict the potential suitable habitats of A. angustifolia under historical climate conditions (1970–2000), as well as under low-emission (SSP126) and high-emission (SSP585) scenarios for the future periods of 2050, 2070, and 2090. Under historical climatic conditions, the average training AUC value from 10 replicate model runs was 0.979, indicating excellent and reliable model performance. Globally, the species has 1.91 × 10⁶ km² of moderately suitable habitat and 0.95 × 10⁶ km² of highly suitable habitat, with a total suitable habitat area of 2.86 × 10⁶ km², accounting for only 1.92% of the global terrestrial area. Mean annual temperature (bio1), mean temperature of the coldest quarter (bio11), and annual temperature range (bio7) are the dominant environmental variables shaping the distribution of A. angustifolia, followed by annual precipitation (bio12). Under future climate scenarios, the overall suitable habitats of A. angustifolia exhibit a slight contracting trend, whereas their spatial distribution patterns remain relatively stable.