The Lifestyle-Dependent Microbial Interactions Vary Between Upstream and Downstream of the Three Gorges Dam

Abstract

Dams represent the most significant anthropogenic disturbance to global rivers. Previous studies have shown that free-living and particle-attached microbes exhibited differentially in river and reservoir ecosystems. However, little is known about the dam’s effect on their co-occurrence patterns. Here, a random matrix theory (RMT)-based network approach was used to construct microbial ecological networks for free-living and particle-attached communities in the immediate vicinity of the Three Gorges Dam (TGD), based on a high-throughput sequencing of 16S rRNA gene. Microbial distribution pattern showed that differences caused by lifestyle (free-living vs. particle-attached) were greater than those caused by geographic position (upstream vs. downstream of the TGD). Network analysis revealed higher connectivity and a lower number of modules in the overall downstream networks. Furthermore, considering the lifestyle, the network structures and properties for free-living and particle-attached microbes were different between upstream and downstream of the dam. Specifically, free-living communities located upstream of the dam exhibited a more complex co-occurrence pattern than the particle-attached communities, whereas the opposite was true for those located downstream of the dam. This variation indicated a strong impact of the dam on microbial interactions for microbes with similar lifestyle in the vicinity of the dam. We identified 112 persistent operational taxonomic unit (OTU)-level species that stably coexisted regardless of lifestyle and geographic positions. These persistent species occupied 21.33–25.57% of the total nodes in each network, and together with their first neighbors, they contributed more than 50% of the nodes and edges belonging to each network. Furthermore, we found that taxonomic affiliations for central nodes (with high degree) varied in these persistent species sub-networks. Collectively, our findings expand the current understanding of the dam’s effect on species interaction variation patterns for free-living and particle-attached communities in the vicinity of the dam, which are more complex than traditional alpha and beta microbial diversity.