Trophic networks improve the performance of microbial anodes treating wastewater

Abstract

Microbial anodes represent a distinct ecological niche that is characterized mainly by the terminal electron acceptor, i.e., the anode potential, and the substrate, i.e., the electron source. Here, we determine the performance and the biofilm community of anode microbiomes while using substrates of increasing complexity (organic acids or organic acids and sugar or real domestic wastewater) to mimic different, practically relevant, trophic levels. α-Diversity values increased with substrate complexity. In addition, the higher abundance value of Deltaproteobacteria in the biofilms corresponds to higher reactor performance (i.e., COD removal, current density, and Coulombic efficiency). In reactors exploiting real wastewater, the diversity of the planktonic microorganisms was only little affected. Microbiome network analysis revealed two important clusters for reactor performance as well as performance-independent pathogen-containing clusters. Interestingly, Geobacter was not found to be integrated in the network underlining its outstanding individual ecological role in line with its importance for the efficiency of the electron harvest for all reactors. The microbiome analysis of different trophic levels and their temporal development from initial colonization to stable treatment demonstrate important principles for the implementation of microbial anodes for wastewater treatment.

Keywords: Applied microbiology; Environmental microbiology.

Fig. 1

Course of current density (j) and characteristic process parameters (Coulombic efficiency (CE, yellow) and COD removal (ΔCOD, turquoise)) in real wastewater reactors Real_WW 1–5 (a) and defined wastewater reactors TCA_WW 1–3 and Ferm_WW 1–3 (b). The current density (black line) was continuously monitored for all reactors over the course of the experiment, while the other parameters (colored bars) were determined per batch (detailed values in Supplementary Table 1). Note the different scaling of axis for a and b to allow the visualization of differences between reactors. The Real_WW 1–5 reactors were run with real domestic wastewater as the only source of carbon and microorganisms. The TCA_WW 1–3 and Ferm_WW 1–3 reactors were inoculated with 5% real domestic WW in the first batch and run all batches with 0.6 g L⁻¹ COD equivalents of propionate, butyrate and acetate (TCA_WW 1–3) or sucrose, and propionate, butyrate, and acetate (Ferm_WW 1–3). After each batch, the reactor liquid was completely replenished by fresh real domestic WW, respectively, defined as wastewater

Fig. 2

Bacterial community composition in all samples based on nonmetric multidimensional scaling. The color code of the symbols represents the different reactors running with real (Real_WW) or defined wastewater (TCA_WW and Ferm_WW) as well as the sample origin being the fresh domestic wastewater, biofilm, or planktonic phase. A more detailed assignment of the individual reactors can be found in Supplementary Fig. 1

Fig. 3

Bacterial community composition. a Relative abundance of Deltaproteobacteria in the anode biofilms in the Real_WW reactors 1–5 over time. The biofilms formed at the electrode during batch I were not sufficiently dense for sampling and therefore had to be excluded. The relative abundance refers to all sequences in each sample but only genera of Deltaproteobacteria with a contribution of more than 1% are displayed. b Community composition of the anode biofilms after batch VI, including all genera with an abundance of at least 1%. The defined wastewater reactors TCA_WW 1–3 and Ferm_WW 1–3 were inoculated with 5% domestic wastewater for batch I. In contrast, Real_WW reactors 1–5 received fresh real WW in each batch as the only source of carbon and repeated inoculum

Fig. 4

α-Diversity indices in all samples based on the diversity measures Observed genera, Chao1, and Shannon. The color code of the symbols represents the different reactors running with real (Real_WW) or defined wastewater (TCA_WW and Ferm_WW). The shape of the symbols is referring to the sample origin, the fresh domestic wastewater (square), biofilm (circle), or planktonic phase (triangle). The calculation of p values was performed with the R packages vegan, iNext, and RDPutils

Fig. 5

Microbiome interaction network. Interestingly, as a result of this analysis, Geobacter is not included; thus, it is not connected to the other genera. This supports the hypothesis that Geobacter occupies a specific ecological niche within the anode biofilm independent from general COD degradation and other microorganisms. The biggest cluster (I) has the highest importance for the functional performance of the reactors followed by cluster V, while the members of cluster II show negative correlations to reactor performance parameters and are probably responsible for alternative COD degradation pathways being independent of anode respiration