Co-occurrence Network Reveals the Higher Fragmentation of the Bacterial Community in Kaidu River Than Its Tributaries in Northwestern China

Abstract

Rivers and their tributaries sculpt the earth's surface, and play an important role in substance circulation and energy flow. Bacteria are involved in most biogeochemical processes in the fluvial ecosystem; however, their pattern distribution in a river and its tributaries has not yet been investigated in detail. In the present study, high-throughput sequencing was employed to examine bacterial communities and their co-occurrence networks between Kaidu River and its nine tributaries in northwestern China. The results obtained demonstrated that both bacterial communities shared a similar dominant sub-community, mainly consisting of Actinobacteria, Bacteroidetes, and Proteobacteria, with Limnohabitans and Variovorax as the dominant genera. In spite of these commonalities, bacterial community structures still significantly differed between these two habitats, which may be related to the distance-related dispersal limitation. Their co-occurrence networks were generally both positively structured. The structural analysis showed that OTUs from the same phyla were more likely to co-occur. Although the keystone genera were taxonomically different between Kaidu River and its tributaries, they both shared common trophic properties in exploiting niches under oligotrophic conditions. We noted that their relative abundances were less than 1%, indicating the over-proportional roles of rare genera in the bacterial community. In addition, the inferred networks showed less nodes and edges, but higher modularity in Kaidu River than its tributaries, suggesting the higher fragmentation of the bacterial community in the mainstream.

Keywords: Kaidu River; bacterial community; co-occurrence network; tributaries.

Fig. 1

Sampling sites in Kaidu River and its tributaries

Fig. 2

Major physicochemical properties in Kaidu River and its tributaries

Fig. 3

Comparisons of bacterial α-diversity (a) and β-diversity (b) between Kaidu River and its tributaries.

Fig. 4

Relative abundance of OTUs at phylum and genus levels in Kaidu River and its tributaries. Only the top seven phyla are shown at each sampling site, and the rest are defined as ‘Others’. At the genus level, only the top six dominants are exhibited. Unknown taxa with significant proportions are grouped as ‘Unclassified’, and the rest are defined as ‘Others’.

Fig. 5

Co-occurrence networks of bacterial communities in Kaidu River. The top 5 modules are presented by different colors. The size of each node corresponds to their relative abundance. Black edges are representative of a positive relationship and red edges of a negative relationship.