Abstract: To clarify the differences in community structure of soil microorganisms in different habitats under the background of severe arsenic contamination, we collected soil samples in four different habitats (bare land, farmland, grassland and forest) near an arsenic-contaminated site in Hezhou, Guangxi. The 16S rRNA Illumina high-throughput sequencing technology and molecular ecological network were used to compare the structure and composition of soil microbial communities, the interactions among microbial species, and the correlations between microorganisms and environmental factors. The results showed that soils of the four habitats were acidic, with the bare land soil being the most acidic. The diversity and richness of microbial communities were lowest in bare land soils and highest in farmland soils. With respect to relative abundance at the bacterial phylum level, the most critical phylum was Actinobacteriota in farmland and Chloroflexi in the other three habitats. The analysis of molecular ecological network structure showed that the network of farmland soil had the lowest number of nodes, connections, and modules, with simpler connections and lower system stability, while forest soil and grassland soil had higher connections and average connectivity, with more complex connections among microbial species and higher system stability. Soil microorganisms in the four habitats mainly showed cooperative relationships. Results of redundancy analysis showed soil pH, available K, total N, and soil organic matter contents significantly influenced soil microbial community structure.

Key words: arsenic, soil microbial community, molecular ecological network, high-throughput sequencing