Abstract: To study the soil fungal community compositions and functional groups under different management practices and slope positions, soil samples derived from the topsoil (0-20 cm) were collected from the upper, middle, and lower slope positions of conventional and organic tea gardens. High-throughput sequencing, functional predictions and the molecular ecological network analysis were performed to investigate the community structures, functional groups and ecological network of soil fungi, respectively. The two-way ANOVA including management practices and slope positions showed that the two-way interaction had no effect on the soil fungal community α-diversity. Regardless of the management practices and the slope positions, the predominant fungal phyla were Ascomycota, Basidiomycota and Mortierellomycota. Management practices did not change the compositions of dominant fungal species, but affected their relative abundance. Principal coordinate analysis (PCoA) shows that the soil fungal community structures differed significantly among different management practices. The community structures of soil fungi at different slope positions were noticeably different under the conventional planting pattern (P<0.05), but were similar under the organic planting pattern (P>0.05). The linear discriminant analysis effect size (LEfSe) analysis shows that 37 biomarkers were very sensitive to the changes in the management, with different management practices enriching for different fungal populations. The numbers of degree, clustering, edges, degree centrality and closeness centrality of the fungal interaction network under the organic planting pattern were all higher than those under the conventional planting pattern, indicating that the fungal networks were more complex in the organic tea garden. Saprotrophic fungi were the dominant fungal group across all tea gardens (66.67%~70.18%). The effects of the management practices on soil fungal functions were obvious. Compared to those under the conventional planting pattern, the average abundance of endophyte-litter saprotroph-soil saprotroph, wood saprotroph and animal pathogen-endophyte-plant pathogen-undefined saprotroph significantly increased, but the average abundance of undefined saprotroph, plant pathogen and animal pathogen-plant pathogen-undefined saprotroph remarkably decreased. The spearman correlation analysis and the redundancy analysis (RDA) show that soil total phosphorus, available phosphorus, total potassium, organic matter, cation exchange capacity and pH were the main factors affecting the abundance and diversity of the soil fungal community. Organic planting drastically changes the structure and the compositions of the soil fungal community, enhances fungal network complexity and stability, and thus is beneficial to maintain the sustainable ecosystem in tea garden soil.

Key words: management practices, slope position, fungal community structure, collinear network analysis, FUNGuild functional prediction