Abstract
Narrow-range endemic plant species are highly sensitive to environmental variability due to their restricted distributions and narrow ecological niches, yet quantitative assessments of such species in Central Asian mountain ecosystem remain limited. This study applied an ensemble species distribution modeling (SDM) approach to assess the ecological constraints and conservation efforts of Lepidium olgae, a strict endemic species of the Nuratau Mountains in Uzbekistan. Species occurrence records from field surveys and herbarium data were integrated with remotely sensed climatic, vegetation, topographic, soil, and atmospheric variables. Parsimonious models (Generalized Linear Model (GLM), Maximum Entropy (MaxEnt), Multiple Adaptive Regression Splines (MARS), Surface Range Envelope (SRE)) were implemented in BIOMOD2 4.3.4, and ensemble predictions were used to reduce algorithmic uncertainty and identify core habitat patterns. Results showed that wet-season precipitation was the dominant driver of species distribution, followed by vegetation productivity (NDVI) and thermal stability, indicating a strong dependence on moisture availability and stable microhabitats. Ensemble projections revealed a highly fragmented potential distribution, with suitable habitats covering only 8% of the reserve area, closely matching the observed distribution of 6.5%. This strong spatial overlap confirms a narrowly constrained realized ecological niche. These findings highlight the critical role of microhabitat stability for the persistence of Lepidium olgae and provide a spatially explicit basis for prioritizing in situ conservation and guiding model informed translocation efforts.