Community diversity and microbiota composition
Overall, 7,463,182 raw sequence reads were obtained from 104 water and sediment samples. After quality filtering, 1,450,613 clean reads were obtained and then clustered into 1352 ASVs. To determine whether the sampling depth was sufficient to give an overview of the microbiota, rarefaction curves were generated for the number of ASVs per individual or species (Fig. S1). The alpha diversity indices of microbiota in water and sediment are presented in Fig. S2. Taking into account the richness, Shannon, Simpson, Pielou, Chao, and ACE indices, microbial alpha diversity was higher in Hanjiang River than in the Weihe River and Qinling tributaries. In Weihe River, all the alpha diversity indices in autumn were lower than those in spring.
The microbiota in water samples comprised 14 bacterial phyla and 73 bacterial families. Proteobacteria, Bacteroidetes, Actinobacteria, and Firmicutes were identified as the dominant phyla, accounting for >98% of all sequence reads. In addition, there were 16 dominant families (including Comamonadaceae, Flavobacteriaceae, Cytophagaceae, Moraxellaceae, and Enterobacteriaceae), constituting >84% of all sequence reads (Fig. S4a). The microbiota in sediment samples comprised 14 bacterial phyla and 75 bacterial families. Proteobacteria, Bacteroidetes Firmicutes, Planctomycetes, Actinobacteria, Acidobacteria, and Cyanobacteria were identified as the dominant phyla, amounting to >96% of all sequence reads. In addition, there were 23 dominant families (including Comamonadaceae, Flavobacteriaceae, Xanthomonadaceae, Rhodobacteraceae, Rhodocyclaceae, and Sphingomonadaceae), constituted >81% of all sequence reads. A notable finding was that large numbers of Acinetobacterspp. occurred at the W9, W10, and W11 sites, which mainly fell under agricultural and urbanized land uses in spring (Fig. S4).
The ASVs from the water and sediment samples were clustered into distinct groups between autumn and spring following UPGMA clustering based on Bray-Curtis distances (Fig. S4). The ASVs from water samples of most sites in Weihe River were clustered into one group in autumn. The cluster distances of ASVs in Qinling tributaries and Hanjiang River were close. Similar results for ASV clusters were obtained for sediment samples (Fig. S4b). NMDS ordination results showed significant differences in the compositions of microbial communities between different river areas (water in autumn: ANOSIM R = 0.265,P = 0.001, stress = 0.093; water in spring: ANOSIM R = 0.243,P = 0.001, stress = 0.104; sediment in autumn: ANOSIMR = 0.294, P = 0.001, stress = 0.145; sediment in spring: ANOSIM R = 0.362, P = 0.001, stress = 0.122). The UPGMA tree and NMDS clustering results of microbiota in water samples were consistent with the PCoA results of land-use types in Weihe River, Qinling River, and Hanjiang tributaries (Fig. 3, Fig. 2a).