Reactor operation
The influent and effluent concentrations of ammonia and nitrite as well as their removal rate were shown in Fig. 2. On Phase I, the minimum removal rate of reactor SA was 95%, and became stable at 80d. On Phase II and Phase III, the removal rate of nitrite was kept above 99% without any fluctuation (Fig. 2a). As to reactor SB, 0.02-0.03mg/l ammonia could be detected in the effluent and the minimum removal rate was 89% on Phase I. On Phase II and Phase III, the removal rate of urea was stable. Its ammonia removal rate showed no evident fluctuation when the loading rate increased.
Although different nitrogen sources were used in reactor SA and reactor SB, the general trend were similar. On Phase I, the removal rate of ammonia and nitrite showed fluctuations. On Phase II and Phase III, there was little fluctuation and only nitrite could be detected in the effluent.
Nitrification rate determination
Different nitrogen sources had different effects on the nitrification rate in reactors, which was shown in Fig. 3. On Phase I, two reactors have similar nitrification rate at around 0.19 mg N·g-1 biofilm · d-1. At the end of Phase II, the nitrification rate in reactor SA was 0.4 mg N·g-1 biofilm · d-1, 1.48 times higher than that of at the beginning of Phase II (0.27 mg N·g-1 biofilm · d-1). The nitrification rate in reactor SA was 0.6 mg N·g-1 biofilm · d-1 at the end of Phase III, which was 3.16 times higher than Phase I, reaching 0.6 mg N·g-1 biofilm · d-1. The nitrification rate in reactor SB grew faster than that of in reactor SA. At the end of Phase II, the nitrification rate in reactor SB was 0.62 mg N·g-1 biofilm · d-1,1.72 times higher than that at the beginning of Phase II (0.36 mg N·g-1 biofilm · d-1). At the end of Phase III, it reached 1.29 mg N·g-1 biofilm · d-1, 6.79 times higher than that of on Phase I.
The nitrification rates in reactor SB was higher than that in reactor SA on three Phase during the experiment. At the end of Phase III, the nitrification rates in reactor SB was 2.15 times higher than that of in reactor SA.
The abundance of ammonia-oxidizing microorganisms in reactors
Nitrite and urea had a great impact on the enrichment of ammonia-oxidizing microorganism communities. The abundances of AOA, AOB, comammox and the relative abundance of ammonia-oxidizing microorganisms were shown in Fig. 4. From Phase I to the end of the Phase III, the amoA gene copy number of AOA in reactor SA increased from 4.46×108copies/ (g biofilm) to 5.33×108copies/ (g biofilm) (Fig. 4a); the amoA gene copy number of AOB increased from 1.56×107 copies/ (g biofilm) to 2.86×108 copies/ (g biofilm) (Fig. 4j); and the amoA gene copy number of comammox increased from 5.81×108copies/ (g biofilm) to 9.04×109 copies/ (g biofilm) (Fig. 4g). In reactor SB, the amoA gene copy number of AOA was 5.12×108 copies/ (g biofilm) on Phase I and increased to 9.52×109 copies/ (g biofilm) on Phase III (Fig. 4b); the amoA gene copy number of AOB increased from 1.09×107 copies/ (g biofilm) to 2.54×108 copies/ (g biofilm) (Fig. 4f); the amoA gene copy number of comammox increased from 5.32×108 copies/ (g biofilm) to 5.34×1010 copies/ (g biofilm) (Fig. 4h).
In reactor SA, the initial proportion of comammox to the total amoA was 56%. It accounted for 67% at the 150th day and increased to 92% at the 390th day (Fig. 4b). In reactor SB, the initial proportion of comammox’s amoA to total amoA was 50% and increased to 85% at the 390th day. The abundance of comammox in reactor SB (5.34×1010 copies/ g biofilm) was higher than that in reactor SA (9.04×109 copies/ g biofilm). The proportion of comammox to the total ammonia oxidation microorganisms in reactor SA (92%) was higher than that in reactor SB (85%).
Structural differences in microbial communities
Different nitrogen sources had different effects on the structure of bacterial communities (Fig. 5). On Phase I, the results of high-throughput sequencing in reactor SA and reactor SB were similar. At the beginning of Phase II (the 90th day), the relative abundance of Nitrospirae in reactor SA increased from 0.94% to 5.3% and to 9.70% at the end of Phase II. The relative abundance of Nitrospirae on Phase III (20.60%) was 21.91 times higher than Phase I (0.94%) (Fig. 5a). Reactor SB also showed the same trend. The relative abundance of Nitrospirae increased from 0.89% on Phase I to 17.21% on Phase II. On Phase III, it reached 22.40%, 25.20 times higher than that on Phase I (Fig. 5b).
In addition to Nitrospirae, the relative abundance of Proteobacteria in reactor SA and SB decreased by 6.1% and 10.9%, respectively on Phase III. The relative abundance of Acidobacteria reduced from 29.8% to 8.48% in reactor SA, and 29.21% to 6.13% in reactor SB.
The Pathways of microbial metabolism
Different nitrogen sources would affect the nitrogen metabolism of microorganisms’ genes, and thus their TPM value. After enrichment with nitrite and urea in reactors for 390 days, the TPM of function genes had changed. Based on the annotated metagenomic sequencing by KEGG database, the metabolic pathways of the nitrogen cycle were shown in Fig. 6. Nitrification, nitrogen fixation, denitrification, assimilation nitrate reduction, and dissimilation nitrate reduction pathways could be found in both reactors. However, neither reactor SA nor SB contained hzs that could convert ammonia into hydrozine or hdh that could convert hydrozine into nitrogen, that means the TPM values of genes (nar, nap, nr, nas, nir, nrf, nit-6) related to nitrate reduction and nitrite reduction in reactor SA were higher than those in reactor SB. The TPM values of both amo and hao in reactor SB were higher than those in reactor SA, whereas the nxr in reactor SA was higher than that in reactor SB. The genes with higher TPM in reactors SA and SB showed comparatively high similarity. Due to different nitrogen sources, nxr in reactor SA (TPM = 5099) outnumbered that in reactor SB (TPM = 4071) during the nitrification process, and the amo (TPM = 3915) (Fig. 6), ure (TPM = 3638) and urt (TPM = 4106) (Table S2) in reactor SB had higher TPM values than amo (TPM = 2707), ure (TPM = 3002) and urt (TPM = 3505) in reactor SA.