Effect of sludge age on methanogenic and glycogen accumulating organisms in an aerobic granular sludge process fed with methanol and acetate

Abstract

The influence of sludge age on granular sludge formation and microbial population dynamics in a methanol- and acetate-fed aerobic granular sludge system operated at 35°C was investigated. During anaerobic feeding of the reactor, methanol was initially converted to methane by methylotrophic methanogens. These methanogens were able to withstand the relatively long aeration periods. Lowering the anaerobic solid retention time (SRT) from 17 to 8 days enabled selective removal of the methanogens and prevented unwanted methane formation. In absence of methanogens, methanol was converted aerobically, while granule formation remained stable. At high SRT values (51 days), γ-Proteobacteria were responsible for acetate removal through anaerobic uptake and subsequent aerobic growth on storage polymers formed [so called metabolism of glycogen-accumulating organisms (GAO)]. When lowering the SRT (24 days), Defluviicoccus-related organisms (cluster II) belonging to the α-Proteobacteria outcompeted acetate consuming γ-Proteobacteria at 35°C. DNA from the Defluviicoccus-related organisms in cluster II was not extracted by the standard DNA extraction method but with liquid nitrogen, which showed to be more effective. Remarkably, the two GAO types of organisms grew separately in two clearly different types of granules. This work further highlights the potential of aerobic granular sludge systems to effectively influence the microbial communities through sludge age control in order to optimize the wastewater treatment processes.

Figure 1

(A) Evolution in time of methane concentration measured just after the anaerobic feeding period when the aeration started (filled circle), average anaerobic solid retention time (days) (open diamond) and biomass concentration (g VSS l⁻¹) (filled triangle) in an aerobic granular sludge reactor fed with acetate and methanol at 35°C. (B) Nitrogen species in the effluent: ammonium in the effluent (open diamonds), nitrate in the effluent (open triangles), nitrogen removal efficiency (closed squares). Arrow indicates when the solid retention time of 24 days was introduced. Lines are shown to indicate trends.

Figure 2

(A) black granules at the start of the experiment, (B) black and white granules after 120 days of operation at reduced SRT, and (C) white granules after 250 days with a solid retention time of 24 days fed with acetate (85% COD) and methanol (15% COD). Average granule size distribution for A, B and C were 1.7, 0.9 and 0.6 mm respectively.

Figure 3

Methane measured in the headspace during anaerobic batch tests with methanol on selected white and black granules taken at day 140.

Figure 4

DGGE on bacterial 16s rDNA gene isolated from black and white granules (day 120) from an aerobic granular sludge reactor extracted with the protocol described in the standard extraction kit from (Mobio, USA) and the adjusted liquid nitrogen (N₂) extraction method (as described in Experimental procedures).

Figure 5

The phylogenetic tree was calculated using maximum-likelihood algorithm implemented in ARB (RAxML module). Full sequences from the SSU115-NR99 database were used for calculation together with an SSUref: bacteria filter (resulting in 1125 base pairs). DGGE bands and clones with a variable length (311–962 base pairs) were added later using parsimony algorithm. Bootstrap (250rxn) was performed but not shown. Sequences shown were deposited into GenBank under Accession No. KP064478–KP064500.

Figure 6

Microscopic examination of crushed white granules showing the tetrad-forming Defluviicoccus-related organisms in phase contrast.

Figure 7

FISH performed on a black granule with (A1) phase-contrast, (A2) GAOmix targeting γ-Proteobacteria (fluos), (A3) EUBmix targeting all bacteria (Cy5); white granule (B1) phase contrast, (B2) DF 2 targeting α-Proteobacteria (Cy3), (B3) EUBmix targeting all bacteria (Cy5).