Double-edged sword effects of dissimilatory nitrate reduction to ammonium (DNRA) bacteria on anammox bacteria performance in an MBR reactor

摘要

Dissimilatory nitrate reduction to ammonium (DNRA) bacteria imposing double-edged sword effects on anammox bacteria were investigated in an anammox-membrane bioreactor (MBR) experiencing an induced crashrecovery event. During the experiment, the anammox-MBR was loaded with NII4+-N:NO2---N ratios (RatioNII(4)(+)-N:NO2---N) of 1.20-1.60. Initially, the anammox-MBR removed over 95 of 100 mg/L NH4+-N and 132 mg/L NO2---N (RatioNH4+-N: NO2---N-N = 0.76, the well accepted stoichiometric RatioNH(4)(+)-N: NO2-#-N for anammox) in the influent (Stage 0). Then, we induced a system crash-recovery event via nitrite shock loadings to better understand responses from different guilds of bacteria in anammox-MBR, loaded with 1.60 RatioNII(4)(+)-N:NO2---N with 100 mg/L NO2--N in the influent (Stage 1). Interestingly, the nitrogen removal by anammox bacteria was maintained for about 20 days before starting to decrease significantly. In Stage 2, we further increased influent nitrite concentration to 120 mg/L (1.33 RatioNII(4)(+)-N: NO2- -N) to simulate a high nitrite toxicity scenario for a short period of time. As expected, nitrogen removal efficiency dropped to only 16.8. After the induced system crash, anammox-MBR performance recovered steadily to 93.2 nitrogen removal with a 1.25 RatioNH4+-N:NO2--N and a low nitrite influent concentration of 80 mg/L NO2--N. Metagenomics analysis revealed that a probable causality of the decreasing nitrogen removal efficiency in Stage 1 was the overgrowth of DNRA-capable bacteria. The results showed that the members within the Ignavibacteriales order (21.7) out competed anammox bacteria (17.0) in the anammox-MBR with elevated nitrite concentrations in the effluent. High NO2--N loading (120 mg N/L) further caused the predominant Candidatus Kuenenia spp. were replaced by Candidatus Brocadia spp. Therefore, it was evident that DNRA bacteria posed negative effects on anammox with 1.60 RatioNH(4)(+)-N: NO2--N. Also, when 120 mg/L NO2--N fed to anammox-MBR (RatioNH(4)(+)-N: NO2--N = 1.33), canonical denitrification became the primary nitrogen sink with both DNRA and anammox activities decreased. They probably fed on lysed microbial cells of anammox and DNRA. In Stage 3, a low RatioNH(4)(+)-N: NO2--N (1.25) with 80 mg/L NO2--N was used to rescue the system, which effectively promoted DNRA-capable bacteria growth. Although anammox bacteria's abundance was only 7.7 during this stage, they could be responsible for about 90 of the total nitrogen removal during this stage.