Climate Change Reshapes the Habitat Suitability and Niche Stability of Argentina anserina (L.) Rydb. in Three Ethnic Prefectures of Sichuan Province Based on an Ensemble Model

Climate change may substantially alter the habitat suitability and spatial stability of highland edible and medicinal plant resources. Argentina anserina (L.) Rydb. is an important wild resource plant in western China, but its potential distribution and future climate-change response in the three ethnic prefectures of Sichuan Province remain insufficiently understood. In this study, we used an ensemble species distribution modeling framework based on 31 spatially filtered occurrence records and 12 environmental predictors to evaluate the current and future habitat suitability of A. anserina under SSP126, SSP370, and SSP585 scenarios for the 2050s and 2090s. The ensemble model showed acceptable overall discrimination ability and was used to identify broad-scale suitability patterns. Mean diurnal range, precipitation of the driest quarter, isothermality, and topsoil pH were the main environmental predictors, indicating that thermal variability, dry-season moisture availability, and edaphic conditions jointly shaped habitat suitability. Under current conditions, the total suitable habitat area was predicted to be 15.57 × 104 km2, accounting for 57.2% of the study region, with high-suitability habitats mainly concentrated in the northern, northwestern, and central areas. Future projections suggested an overall decline in suitable habitats, especially in high-suitability areas, although the magnitude varied among scenarios and periods. Spatial change and centroid migration analyses indicated that future suitable habitats may undergo contraction and westward or northwestward redistribution rather than uniform expansion. Niche overlap analysis further showed moderate to relatively high overlap between current and future environmental spaces, suggesting partial niche stability with scenario-dependent shifts. Given the limited field-based occurrence records, these findings should be interpreted as broad-scale spatial guidance rather than precise local predictions.