As a subset of the human–environment system, the tourism ecosystem focuses on the complex dynamics and interactions between tourism activities and the natural environment. Among these, tourism ecological security (TES) is one of the core issues in the study of tourism ecosystems, aiming to balance economic development and ecological environment protection. Currently, the risk early warning of TES has not received widespread attention, and there is an urgent need for a tourism ecological safety risk early warning system to achieve TES monitoring, risk assessment, and decision support. Therefore, this study established a comprehensive TES evaluation system, systematically analyzed the evolution of TES in Sichuan Province from 2010 to 2022, and used the geographical detector to reveal the influencing factors and driving mechanisms of TES. Based on these achievements, an early risk warning system for TES was established based on the Bayesian network model, simulating the response of TES under single-variable and multi-variable scenarios. The research results reveal that TES changes with environmental changes, resource utilization and consumption, and the development of the tourism industry, and there are differences in the driving factors of TES under different conditions. There is a synergistic effect between the influencing factors of TES, and there is a threshold effect in the regulation of tourism ecological safety, revealing the efficiency and limitations of different regulatory strategies. The early risk warning model for TES based on the Bayesian network has high prediction accuracy and can provide effective support for the management and regulatory policies of TES.

Alpine treeline and shrubline ecotones are climatically sensitive transition zones where vegetation shifts strongly influence belowground microbial processes. Soil bacteria and fungi, as core component of the soil microbiome, play vital roles in nutrient cycling and plant–soil interactions within these fragile ecosystems. However, the structure and diversity of soil microbial communities across the treeline–shrubline transition remain poorly understood. Here, we investigated soil bacterial and fungal communities across treeline and shrubline ecotones in two mountain on the eastern Tibetan Plateau. We further examined how soil physicochemical properties shaped microbial community assembly. Our results demonstrated that the community composition of both bacteria and fungi differed significantly between the treeline and shrubline ecotones, while the Shannon index showed no significant variation. At the phylum level, Proteobacteria, Actinobacteriota, and Acidobacteriota dominated bacterial communities, while Ascomycota and Basidiomycota were the predominant fungal phyla. Both the network complexity of soil bacterial and fungal communities changed significantly across ecotones. Specifically, bacterial network complexity increased significantly toward the shrubline, whereas fungal network complexity declined. Bacterial community compositions were co-regulated by both environmental and vegetation factors, while fungal community compositions were only regulated by soil pH. Redundancy analysis revealed that soil organic carbon, pH, and moisture were the primary drivers of bacterial community (38.17%), whereas vegetation cover, soil organic carbon, and moisture explained the largest proportion of fungal community (44.79%). Our findings reveal the distribution patterns and underlying shift mechanisms of microbial communities between the treeline and shrubline ecotone. These insights are crucial for mountain biodiversity conservation and for improving predictions of forest responses to climate change.