Recovery of Candidatus Brocadia fulgida Persisted After a 3-Month Starvation Period: Nitrogen Removal Performance and Microbial Marker Identification.

The recovery of partial nitritation and anammox (PN/A) process after starvation remains challenging. Substrate limitation disrupts microbial activity and delays the restoration of nitrogen removal performance. Therefore, PN/A nitrogen removal performance was evaluated in a 20-L laboratory-scale continuous stirred-tank reactor (CSTR) containing sponge carriers across four operational phases. The four phases included 3-month starvation, gradual substrate reintroduction under suspended-growth conditions, transitional stabilization with sponge media, and attached-growth recovery under higher nitrogen loading without external acetate supplementation. The laboratory reactor was operated at a nitrogen loading rate of 0.4-0.7 kg N/(m3·d), dissolved oxygen below 0.1 mg/L, and without organic carbon addition. The nitrogen removal efficiency, microbial community composition, and microbial markers of the laboratory-scale reactor were compared with those of two well-operated full-scale systems in Korea and the United States. Total nitrogen removal efficiency exceeded 90% across all three systems when the ammonia-to-nitrite ratio (NH₄+:NO2 -) was maintained between 1:1 and 1:1.5 at low BOD/TN ratio of < 0.05. Next-generation sequencing (Illumina MiSeq, 16S rRNA amplicon profiling at a sequencing depth of 41,120-57,097 reads/sample) was employed to determine the relative abundance of nitrogen-transforming microbial communities, specifically the anammox group, across the three systems. The results showed that the anammox bacteria in all three systems belonged to Candidatus Brocadia. Following gene-specific PCR amplification of the anammox community, denaturing gradient gel electrophoresis (DGGE) identified Candidatus Brocadia fulgida. This species exhibits high environmental adaptability and is commonly detected in full-scale systems operated for > 5 years. Molecular analysis indicated that Candidatus Brocadia fulgida persisted in the system throughout the 3-month starvation period, facilitating the re-establishment of the attached-growth PN/A process. The successful recovery of the laboratory-scale reactor yielded nitrogen removal performance comparable to that of the full-scale PN/A systems, demonstrating that the Brocadia group serves as a keystone taxon during process recovery.