Modeling the current distribution of and predicting suitable habitats for threatened species support the species conservation and restoration planning process. Therefore, the purpose of this study was to model the actual distribution and predict environmentally suitable habitats for Carex alatauensis S.R.Zhang 2015, a locally threatened native grass species on the Qinghai–Tibet Plateau. To realize this objective, based on the geographical samples within the natural distribution of C. alatauensis, the dominant climatic factors in its potential distribution range were analyzed using the maximum entropy (MaxEnt) model. The results showed that the average values of the area under the receiver operating characteristic curve (AUC) of the training data were 0.833 ± 0.044, which indicated that the accuracy of the MaxEnt model was pretty high for modeling potential distribution regions of C. alatauensis. The combined results from the Jackknife test and the presented contribution of environmental variables revealed that the annual precipitation, the growth season precipitation, and the precipitation of the driest month were the key climatic factors that restricted the distribution of C. alatauensis on the Qinghai–Tibet Plateau. It is predicted that the potential distribution area of C. alatauensis on the Qinghai–Tibet Plateau is 1.96 × 10⁶ km², and the most suitable area is 3.7 × 10⁵ km², mainly located in the Qilian Mountains, the Himalayas, and the Qingtanggula Mountains.

In order to investigate the characteristics and driving factors of keystone species change in vegetation communities, we conducted a network analysis to analyze the complexity of grassland plant communities and keystone species. We conducted a study with six plots of degraded grassland, intact grassland, and various ages of artificial grassland. We systematically analyzed the characteristics of vegetation community and soil physical–chemical properties under different treatments. Our findings revealed that the 5-year-old artificial grassland exhibited lower biodiversity, as indicated by a Shannon–Weiner index of 1.70 and a Pielou’s evenness index of 0.72. In comparison to the degraded grassland, soil organic carbon and nitrate nitrogen levels in the 5-year-old artificial grassland increased by 22% and 13%, respectively, while soil density decreased by 9%. However, despite the relatively stable plant community in the 21-year-old artificial grassland, it had not fully recovered to the level of natural grassland due to the complexity of the plant community and the shift in keystone species. Therefore, it is crucial to consider the impact of soil bulk density and nutrient levels on the complexity of the plant community when restoring degraded alpine meadows through the establishment of artificial grassland.

This study aims to uncover the structural and functional characteristics of rhizosphere soil bacterial communities of Kengyilia thoroldiana under five types of topographic habitats in the source region of the Yellow River, and to explore the interaction mechanisms between bacterial communities and plant-soil factors, thereby providing microbiological support for the ecological restoration of Kengyilia thoroldiana artificial grasslands in alpine desert grassland. In this study, high-throughput sequencing technology was employed to compare the species composition, diversity, interaction networks, and functional characteristics of rhizosphere bacterial communities of Kengyilia thoroldiana across five topographic habitats in the source region of the Yellow River. Additionally, Mantel tests and redundancy analysis (RDA)) were conducted to explore the key environmental factors driving the structure of bacterial communities. The results showed that habitat differences significantly influenced the community characteristics of Kengyilia thoroldiana and soil physicochemical properties. The plant height, coverage, biomass, and soil carbon, nitrogen, and phosphorus contents were highest in habitats H2 and H5, while they were lowest in habitats H1 and H3. In contrast, soil pH and electrical conductivity exhibited an opposite trend. At the bacterial community level, the number of operational taxonomic units (OTUs) in habitat H5 reached 1917, with α-diversity indices such as Shannon, Ace, and Chao1 being 6.13, 1820.85, and 1844.80, respectively, significantly higher than those in habitat H1. Cluster analysis revealed that habitat H3 formed a distinct group, while the bacterial community structures in the remaining four habitats were similar. Functional prediction indicated that chemoheterotrophy and aerobic chemoheterotrophy were the dominant functions across all habitats, with functional expression values exceeding 9300 in habitats H2, H4, and H5. Redundancy analysis confirmed that soil pH and SOC were the key factors driving the structure of rhizosphere bacterial communities of Kengyilia thoroldiana. In summary, topographic habitats influence the growth of Kengyilia thoroldiana plant communities by shaping soil environmental heterogeneity, thereby regulating the structure and function of rhizosphere bacteria associated with Kengyilia thoroldiana.