**Background:** Biological soil crusts (BSCs) are critical components of dryland ecosystems, covering approximately half of arid and semiarid drylands. They provide ecosystem services including soil stabilization, nutrient cycling, and water regulation. Bacteria play essential roles in BSC formation and function, but how bacterial communities change with BSC development in alpine sandy regions remains poorly understood. The Gonghe Basin on the Qinghai-Tibet Plateau is a region highly sensitive to climate change, where BSCs are widely distributed in Caragana microphylla shrub communities.
**Methods:** Sampling was conducted in August 2021 in the Gonghe Basin sandy land (100°25′ E, 36°24′ N; elevation >2,900 m; annual mean temperature 2.4°C; annual precipitation 246.3 mm). Five developmental stages of BSCs were collected: bare sand, microbial crusts, algae crusts, lichen crusts, and moss crusts. Each stage had three replicate plots (≥20 m apart), with five subsamples per plot composited. Soil physicochemical properties were measured including SWC, pH, TC, TOC, TN, NO₃⁻, NH₄⁺, TP, and soil texture. Bacterial 16S rRNA gene amplicon sequencing (515F/806R primers) was performed on the Illumina HiSeq platform. OTUs were clustered at 97% similarity. Alpha diversity (observed species, Shannon-Wiener, Chao1) and beta diversity (Bray-Curtis dissimilarity, PCoA, PerMANOVA) were analyzed. Mantel tests assessed correlations between bacterial communities and environmental variables.
**Key Results:** A total of 13,590 bacterial OTUs were detected across all stages. Bacterial diversity significantly increased with BSC development: observed species rose from 3,200 (bare sand) to 4,507 (moss crusts), representing a 40.8% increase. Shannon-Wiener and Chao1 indexes showed the same pattern, peaking in moss crusts. Six phyla dominated (>77% relative abundance): Proteobacteria (24.6%±7.0%), Actinobacteria (15.1%±4.6%), Bacteroidetes (15.2%±3.0%), Acidobacteria (10.6%±6.0%), Cyanobacteria (6.5%±6.2%), and Firmicutes (5.3%±2.3%). Acidobacteria (6.3–20.8%) and Bacteroidetes (11.8–18.6%) were notably more abundant than in other desert regions. With BSC development, copiotrophic bacteria (Actinobacteria, Acidobacteria, Bacteroidetes, Verrucomicrobia, Planctomycetes, Gemmatimonadetes) significantly increased, while oligotrophic bacteria (Proteobacteria, Firmicutes) significantly decreased. Proteobacteria declined from 32.1% (bare sand) to 15.0% (moss crusts); Firmicutes declined from 8.0% to 3.2%. Cyanobacteria peaked in algae crusts (13.3%±2.0%) and was significantly higher than in other stages (p<0.05). At the genus level, Microcoleus (cyanobacteria) was most abundant in algae crusts. Soil properties changed markedly: SWC increased from 2.04% (bare sand) to 13.65% (moss crusts); TOC, TC, TN, and TP increased 4.61, 4.76, 2.65, and 0.39 times respectively from bare sand to moss crusts. Sand content decreased from 95.36% to 75.33%, while silt and clay increased. PCoA showed clear separation of bacterial communities across stages (PerMANOVA: R²=0.422, p=0.001). Mantel tests revealed SWC (R=0.683, p=0.002), TC (R=0.598, p=0.001), TP (R=0.591, p=0.001), clay content (R=0.424, p=0.003), pH (R=0.389, p=0.006), and TN (R=0.388, p=0.001) were significantly correlated with bacterial community composition. SWC exerted the strongest effect.
**Clinical Implications:** While not a clinical study, this research has implications for ecological restoration and desertification control in alpine sandy regions. The findings demonstrate that BSC development stage can serve as an indicator of ecosystem health in alpine sandy lands. The shift from oligotrophic to copiotrophic bacterial communities with BSC development suggests that ecological functions transition from soil stabilization (early stages) to nutrient cycling and material circulation (later stages). Understanding these microbial dynamics is critical for predicting alpine desert ecosystem responses under climate change scenarios and for formulating sustainable land management policies in dryland ecosystems.