Metagenomics was used to analyze the phage community structure in the water and sediment of Huangshui River, Weihe River and Luohe River, and the diversity, host and functional potential of the three tributaries of the Yellow River were compared. The results showed that 99.77% of the unknown vOTUs (viral Operational Taxonomic Units) identified from the metagenome could not be clustered with the NCBI Refseq database, indicating a high degree of species novelty in the phage community. A total of 158 genera belonging to 15 families were detected in three tributaries. The endemic phage genera were the most in Huangshui River and Pbi1virus, Badnavirus, Ea214virus and Gammabaculovirus were detected only in this tributary. The phage community structure of the three tributaries was significantly different and the spatial heterogeneity was greater in the sediment samples. Diversified auxiliary metabolic genes (AMGs) indicated that phages might be involved in the cycling of nitrogen, carbon and phosphorus in Huangshui River, Weihe River and Luohe River, but there was no difference in the distribution of AMGs in the three tributaries. The host range of phage community spanned 18 prokaryotic phyla, mainly including Proteobacteria, Actinobacteriota and Bacteroidota. About 1.8% of the phages might infect hosts across phyla. Phages infecting Planctomycetota were enriched in water and sediments of Luohe River, while Chloroflexota-infected phages were relatively abundant in the sediments of Weihe River. There was a significant correlation between the bacterial and phage community and the consistency was higher in water samples.

The authors probed Illumina high-throughput sequencing in Hanjiang River to investigate the community structure, distribution and assembly mechanisms of bacterioplankton integrally and rare taxa (RT), abundant taxa (AT) respectively. Hanjiang River bacteria spanned across 22 phyla, 55 classes, 93 orders, 132 families and 164 genera dominated by Proteobacteria, Actinobacteria and Bacteroidetes. 4035 ASVS (80.2%) were defined as RT. Significant seasonal differences were found of the bacteria community by alpha diversity and this pattern were intensified by RT. Non-metric multidimensional scaling (NMDS) showed insignificant differences of the seasonal and spatial distribution of bacteria community in Hanjiang River. Mantel tests considered dissolved oxygen (DO) and ammonia nitrogen as primary influencing factors of autumn microbiome community. Variance partitioning analysis (VPA) indicated more stress from environmental factors to the bacteria community compared with spatial factors. The holistic bacteria community was dominated by deterministic process, while both RT and AT were mainly affected by stochastic process. Overall, RT was mainly influenced by environmental selection as opposed to AT which were more inclined to stochastic distribution.