The microbiome of two strategies for ammonia removal with the sequencing batch moving bed biofilm reactor treating cheese production wastewater

Abstract

Cheese production facilities must abide by sewage discharge bylaws that prevent overloading municipal water resource recovery facilities, eutrophication, and toxicity to aquatic life. Compact treatment systems can permit on-site treatment of cheese production wastewater; however, competition between heterotrophs and nitrifiers impedes the implementation of the sequencing batch moving bed biofilm reactor (SB-MBBR) for nitrification from high-carbon wastewaters. This study demonstrates that a single SB-MBBR is not feasible for nitrification when operated with anerobic and aerobic cycling for carbon and phosphorous removal from cheese production wastewater, as nitrification does not occur in a single reactor. Thus, two reactors in series are recommended to achieve nitrification from cheese production wastewater in SB-MBBRs. These findings can be applied to pilot and full-scale SB-MBBR operations. By demonstrating the potential to implement partial nitrification in the SB-MBBR system, this study presents the possibility of implementing partial nitrification in the SB-MBBR, resulting in the potential for more sustainable treatment of nitrogen from cheese production wastewater.

Keywords: AOB and NOB competition; industrial wastewater; microbial community; nitrogen removal; partial nitrification.

FIG 1

Strategy 1—extended aerobic operation. First operational strategy for achieving TAN oxidation with the SB-MBBR system. Carriers taken from the AN/AE SB-MBBR were used to inoculate the NIT_ExtendedAE reactor, which simulates an extended aerobic stage. The AN/AE SB-MBBR continued operation following the inoculation and was also operated for the second strategy (Fig. 2).

FIG 2

Strategy 2—two SB-MMBRs in series. Second operational strategy for achieving TAN oxidation with the SB-MBBR system. The two SB-MBBRs are operated in series, each with an 8-hour hydraulic retention time (HRT) and a total HRT of 16 hours.

FIG 3

Nitrogen concentrations in the NIT_ExtendedAE (A) and the NIT_Series reactor (B). Average and 95% confidence interval of TAN, NO₃⁻, and NO₂⁻. Data presented for the NIT_ExtendedAE reactor were collected over the 810 hours required to achieve TAN oxidation, which was representative of the extended aeration stage (A). Data presented from the NIT_Series were collected from a cycle after steady-state conditions were achieved. Some error bars are small for some data points and thus cannot be seen.

FIG 4

Average and 95% confidence interval (CI) of the cell coverage and the live fraction in the biofilm of each reactor. The cell coverage and live fraction of cells were calculated from the viable and non-viable cells imaged through CLSM. The error bars represent the CIs between different CLSM images for each reactor.

FIG 5

The microbiota in each reactor is distinct. (A) Principal coordinate analysis of reactor microbiotas using the unweighted Unifrac. Permutational multivariate analysis of variance revealed that reactor identity accounted for 78% of the total variation (P < 0.0001). (B) Chao1 and Shannon (C) indices for each reactor. Differences between groups were tested using a Mann-Whitney U test; *P < 0.05. (D) Relative abundances of each reactor carrier at the order level. Orders not present at >3% in at least four samples were merged into “Other.”

FIG 6

Total relative abundances of ANAMMOX-, AOB-, and NOB-associated genera.

FIG 7

Differentially abundant ANAMMOX-, AOB-, and NOB-associated taxa. OTUs identified as differentially abundant between the reactors with taxonomies associated with ANAMMOX, AOB, and NOB genera are highlighted. Where required, a pseudo-count was added to allow for Log10 plotting. OTUs are separated by their function, genus, species, and OTU ID. False discovery rate-corrected P values are displayed based on DESeq2 analysis. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Only significant comparisons are highlighted.