关键词: 灌丛, 土壤细菌群落, 微生物间相互作用, 功能预测

Abstract: We analyzed the composition, structure, and potential functions of soil bacterial communities in different shrubs in the semiarid region of northwest China, including Cotoneaster acutifolius, Lonicera japonica, and Cornus alba. The bacterial composition was measured with high-throughput sequencing of the 16S rRNA gene. The correlations between soil bacterial communities and soil physicochemical factors and extracellular enzyme activities were analysed using redundancy analysis and the Mantel test. The results showed that: (1) Soil physicochemical factors and enzyme activities varied significantly among the different shrub types (P<0.05). The Cotoneaster acutifolius soil exhibited the highest levels of total carbon, total nitrogen, nitrate, and available phosphorus. (2) Soil bacterial communities differed significantly among the three shrub types (P<0.01). The relative abundances of Actinobacteriota, Methylomirabilota, Proteobacteria, Firmicutes, Myxococcota and Gemmatimonadota differed significantly among the three shrub types (P<0.05). (3) The metabolic functions of bacterial communities predicted by the FAPROTAX database indicated the highest abundance of functional populations in Cotoneaster acutifolius soil. The analysis of co-occurrence network revealed a more complex linkage between bacterial phyla and functional populations in Cotoneaster acutifolius soil. (4) Redundancy analysis and Mantel test indicated that soil catalase activity, alkaline protease activity, and total nitrogen were the main factors influencing bacterial community structure in different shrub soils, playing a crucial role in driving bacterial community composition and structure. In summary, we found that soil fertility and bacterial community function level of Cotoneaster acutifolius were higher than the other vegetation types, suggesting that Cotoneaster acutifolius could be prioritized for vegetation restoration in the study area.

Key words: shrub, soil bacterial community, interaction among microorganisms, function prediction