Data analysis
To evaluate the differences in land-use type among groups, land-use data were visualized by principal coordinates analysis (PCoA) based on Bray-Curtis distances. Pearson’s correlation coefficients were used to evaluate the impacts of different land-use types on the environmental factors. Both analyses were performed using the “vegan ” package in R software (versions 4.1.2) (R Core Team, 2021).
Microbial alpha diversity was calculated using the “vegan ” package in R software (versions 4.1.2) (R Core Team, 2021). The phylogenetic tree was constructed using the unweighted pair group method with arithmetic mean (UPGMA) based on Bray-Curtis distances with the “ggtree ” package in R software (versions 4.1.2) (R Core Team, 2021) and visualized with iTOL(versions 6) (https://itol.embl.de). Microbial community composition was visualized using the nonmetric multidimensional scaling ordination (NMDS) method based on Bray-Curtis dissimilarities. Analysis of Similarities (ANOSIM) was used to evaluate the degree of separation in microbial communities between groups. Furthermore, a regression analysis was used to determine the Bray-Curtis dissimilarity of microbial communities between sample pairs; subsequently, the environmental factors and land-use types were plotted against the community dissimilarity. The impacts of land-use type and environmental factors on the Bray-Curtis dissimilarity were evaluated using the Mantel’s test and Pearson’s rank correlation.
Microbial functional groups were predicted using the FAPROTAX database (Louca et al., 2016), which is suitable for functional annotation prediction of biogeochemical cycle processes (especially the carbon [C], hydrogen [H2], N, P, sulfur [S] cycles, and other element cycles) in environmental samples (e.g., oceans and lakes). The FAPROTAX database was established based on published and validated literature (Sansupa et al., 2021). The relative abundances of predicted functional groups were subjected to heatmap analysis using the “pheatmap ” package in R software (versions 4.1.2) (R Core Team, 2021). To assess the impacts of environmental factors and land-use types on microbial functional groups, the heatmap and Pearson’s rank correlation were visualized based on Bray-Curtis dissimilarities. Linear discriminant analysis (LDA) effect size (LEFSe) analysis was conducted on the functional groups to identify biomarkers of anthropogenic disturbance.