Dissolved organic matter (DOM) plays an important role in the evolution of microbial communities. Meanwhile, ecological water delivery is an important feature of Baiyangdian Lake. To explore how the structure of the aerobic denitrification bacteria community responds to DOM during the water delivery period, the DOM components of water were examined and high-throughput sequencing of aerobic denitrification bacteria was performed. The results showed significant differences in DOM concentration in Baiyangdian Lake, with the estuary area exhibiting lower DOM concentrations. The water exhibited strong autogenous source, while DOM in the estuary area had a higher molecular weight and degree of humification. Three protein-like substances (C1, C2, and C4) and one humic-like substance (C3) were identified through PARAFAC. The protein-like substances accounted for the major proportion of DOM, which was consistent with the results of fluorescence regional integration (FRI). The genera of the water body were mainly in the Protebacterice phylum, including Cupriavidus, Aeromonas, Thauera, Shewanella, and Pseudomonas. Meanwhile, Cupriavidus, Thauera, Shewanella, Agrobacterium, and Pseudomonas were the main indicator species, according to random forest (RF) analysis. Through network analysis, 35 key nodes of the network were obtained, belonging to Thauera, Cupriavidus, and Unclassified_bacteria, respectively. Redundancy analysis (RDA) showed that a humic-like substance was the main environmental factor regulating the whole structure of the aerobic denitrification bacterial community, while protein-like substances played important roles in changes to the indicator species and key nodes of the community. Overall, protein-like substances could provide an important reference for selecting carbon sources during the screening of efficient and cold resistance aerobic denitrification bacteria that are adapted to actual water bodies.
Keywords: Baiyangdian Lake; aerobic denitrification; bioinformatics analysis; dissolved organic matter (DOM); network analysis; parallel factor analysis(PARAFAC).