Redundancy in Anaerobic Digestion Microbiomes during Disturbances by the Antibiotic Monensin

Abstract

The antibiotic monensin is fed to dairy cows to increase milk production efficiency. A fraction of this monensin is excreted into the cow manure. Previous studies have found that cow manure containing monensin can negatively impact the performance of anaerobic digesters, especially upon first introduction. Few studies have examined whether the anaerobic digester microbiome can adapt to monensin during the operating time. Here, we conducted a long-term time series study of four lab-scale anaerobic digesters fed with cow manure. We examined changes in both the microbiome composition and function of the anaerobic digesters when subjected to the dairy antibiotic monensin. In our digesters, monensin was not rapidly degraded under anaerobic conditions. The two anaerobic digesters that were subjected to manure from monensin feed-dosed cows exhibited relatively small changes in microbiome composition and function due to relatively low monensin concentrations. At higher concentrations of monensin, which we dosed directly to control manure (from dairy cows without monensin), we observed major changes in the microbiome composition and function of two anaerobic digesters. A rapid introduction of monensin to one of these anaerobic digesters led to the impairment of methane production. Conversely, more gradual additions of the same concentrations of monensin to the other anaerobic digester led to the adaptation of the anaerobic digester microbiomes to the relatively high monensin concentrations. A member of the candidate OP11 (Microgenomates) phylum arose in this anaerobic digester and appeared to be redundant with certain Bacteroidetes phylum members, which previously were dominating.IMPORTANCE Monensin is a common antibiotic given to dairy cows in the United States and is partly excreted with dairy manure. An improved understanding of how monensin affects the anaerobic digester microbiome composition and function is important to prevent process failure for farm-based anaerobic digesters. This time series study demonstrates how anaerobic digester microbiomes are inert to low monensin concentrations and can adapt to relatively high monensin concentrations by redundancy in an already existing population. Therefore, our work provides further insight into the importance of microbiome redundancy in maintaining the stability of anaerobic digesters.

Keywords: Bacteroidales; OP11; anaerobic digestion; antibiotic; dairy manure; disturbance; hindgut; microbiome; monensin; redundancy.

FIG 1

β-Diversity of all anaerobic digester microbiome samples from day 175 to the end of the operating period. (A) Principal-coordinate analysis (PCoA) plot based on weighted UniFrac distance. (B) Distance-based redundancy analysis (db-RDA) plot showing the four measured parameters that best explained the variation observed in the PCoA plot. Methane, methane yield; alkalinity, bicarbonate alkalinity concentration; ammonia, total ammonium concentration; monensin, monensin concentration in the substrate. In both the PCoA and db-RDA plots, the points are colored on a gradient scale representing the duration of the operating period in days.

FIG 2

Performance parameters for the four anaerobic digesters from day 175 to the end of the operating period for specific methane yields (A), tVFA concentrations (B), total ammonium concentrations (C), and bicarbonate alkalinity concentrations (D). Points are colored on a color scale corresponding to the concentration of monensin dosed to the anaerobic digester at that time point. The different periods are marked with gray and white shading from left to right during the operating period: the first gray shading identifies P1; the first white shading identifies P2; the second gray bar identifies P3, and the second white area identifies P4 (see Table S4 in the supplemental material). The fast anaerobic digester was stopped at the end of P3 due to a loss of stability.

FIG 3

β-Diversity of anaerobic digester microbiome samples during different periods. (A) Boxplot of weighted UniFrac distances within microbiomes from a single anaerobic digester throughout a similar operating period. Control anaerobic digester boxplot corresponds to days 203 to 306, when the slow anaerobic digester was fed only control cow manure without monensin addition; the low A anaerobic digester boxplot corresponds to days 203 to 383; the low B anaerobic digester boxplot corresponds to days 203 to 355; the fast anaerobic digester boxplot only includes samples during which the anaerobic digester was subjected to high concentrations of monensin (i.e., days 265 to 306); and the slow anaerobic digester boxplot corresponds to days 306 to 383, when the anaerobic digester was subjected to high concentrations of monensin. (B) Similarity of samples postdisturbance compared to three samples prior to disturbance (prior to P2 on days 175, 189, and 201) based on the weighted UniFrac distance metric. Similarity was calculated as one minus the average weighted UniFrac distance between the sample day and the three sample days prior to disturbance (prior to P2). Colors represent the monensin concentration gradient: dark red, high monensin concentration (1 to 5 mg · liter⁻¹); pink, low monensin concentration (<1 mg · liter⁻¹); white, no monensin in substrate. Periods (P2 to P4) are represented by gray and white shading.

FIG 4

Heatmaps representing relative abundance of major OTUs in anaerobic digester samples during the entire sampling period for the low A anaerobic digester (A), low B anaerobic digester (B), fast anaerobic digester (C), and slow anaerobic digester (D). OTUs represented here reached at least 5% relative abundance in any one anaerobic digester sample. Lowest-level taxonomic data, as well as the OTU ID number, are provided. The sidebar color scale represents the monensin dosing rate to the anaerobic digester. For the fast digester (panel C), the black line in the red bar indicates when the monensin dose was directly increased from 1 to 5 mg · liter⁻¹. For the slow digester (panel D), the black lines in the red bar indicate when the monensin dose was increased (from 1 to 2 mg · liter⁻¹, from 2 to 3 mg · liter⁻¹, from 3 to 4 mg · liter⁻¹, and from 4 to 5 mg · liter⁻¹).

FIG 5

β-Diversity of the fast and slow anaerobic digester microbiome samples during high monensin periods (≥1 mg · liter⁻¹; day 265 to 307 for fast anaerobic digester, and day 307 to 383 for slow anaerobic digester). (A) Principal-coordinate analysis (PCoA) based on weighted UniFrac distance. Time points are colored in a gradient scale corresponding to monensin concentration in the substrate, as indicated by the color sidebar. (B) Taxon biplot showing the most abundant OTUs in these samples (OTUs that reached an average relative abundance of ≥1% across the time period shown). OTUs are shown nearest to the samples that they are most abundant in. Sample time points are not shown as they are displayed in the left-hand PCoA. Point size represents average relative abundance of that OTU across all samples (minimum average relative abundance is 1.0%, maximum average relative abundance is 12.0%). Points are colored by phylum-level taxonomy. Three OTUs discussed in the text are highlighted. OTU 820298, OP11 OTU; OTU 265425, Prevotella OTU; OTU 837605, Bacteroidales OTU.