Abstract
Biodegradation of digestate liquid fraction was performed in the activated sludge system with acetic acid, flume water, and molasses as external carbon sources. High-throughput sequencing was used to gain in-depth insight into the activated sludge microbial community. The type and amount of carbon source in influent (COD/TN ratio) significantly influenced microbial community structure, especially at the genus level, and thus the biodegradation performance of digestate liquid fraction. The highest total nitrogen and chemical oxygen demand removal efficiencies averaging 85.3% and 88.3%, respectively, were achieved in series with acetic acid and flume water and COD/TN ratio of 10.7 and 11.2, respectively. The microbial diversity in these series averaged at 3.08 and 3.65. The dominant bacteria at the phylum level in series with acetic acid were Proteobacteria and Bacteroidota, and at the genus level Azospira, while in series with flume water they were Bacteroidota and Firmicutes, and Macellibacteroides, respectively.
Download PDF
Full Article
Insights into the Microbial Community Structure in the Biodegradation Process of High-Strength Ammonia Digestate Liquid Fraction in Conventional Activated Sludge System
Aleksandra Chuda,a,* Anna Otlewska,b and Krzysztof Ziemiński a
Biodegradation of digestate liquid fraction was performed in the activated sludge system with acetic acid, flume water, and molasses as external carbon sources. High-throughput sequencing was used to gain in-depth insight into the activated sludge microbial community. The type and amount of carbon source in influent (COD/TN ratio) significantly influenced microbial community structure, especially at the genus level, and thus the biodegradation performance of digestate liquid fraction. The highest total nitrogen and chemical oxygen demand removal efficiencies averaging 85.3% and 88.3%, respectively, were achieved in series with acetic acid and flume water and COD/TN ratio of 10.7 and 11.2, respectively. The microbial diversity in these series averaged at 3.08 and 3.65. The dominant bacteria at the phylum level in series with acetic acid were Proteobacteria and Bacteroidota, and at the genus level Azospira, while in series with flume water they were Bacteroidota and Firmicutes, and Macellibacteroides, respectively.
DOI: 10.15376/biores.18.2.3540-3559
Keywords: Digestate biodegradation; Biological nitrogen removal; External carbon source; Activated sludge; Microbial diversity
Contact information: a: Department of Environmental Biotechnology, Lodz University of Technology, Wolczanska 171/173 Street, 90–530 Lodz, Poland; b: Institute of Fermentation Technology and Microbiology, Lodz University of Technology, Wolczanska 171/173 Street, 90–530 Lodz, Poland;
*Corresponding author: olachuda50@gmail.com
GRAPHICAL ABSTRACT
INTRODUCTION
There is a growing concern about environmental pollution with digestate, which is produced in significant amounts during anaerobic digestion processes (Monfet et al. 2018; Świątczak et al. 2019). A 1 MW biogas plant generates about 10,000 to 30,000 m3 of digestate per year, thus around 180 million tonnes of digestate are produced annually in Europe (Gienau et al. 2018; European Commission 2019). The currently used method of digestate utilisation, which is application as fertilizer, carries the risk of soil acidification, eutrophication of surface waters, and emission of greenhouse gases. In order to avoid exceeding the fertilization limits of nitrogen and phosphorus, digestate can be used as fertilizer only in the growing season. Due to such restrictions, biogas plants must have adequate facilities with a storage capacity of up to 6 months. This results in high costs compared with digestate fertilizing value, in addition to complicated logistic and transport requirements. Therefore, attempts are still being made to introduce alternative digestate processing technologies (Świątczak et al. 2019; Duan et al. 2020; Nyang’au et al. 2022). The first step of digestate treatment is usually its mechanical separation into a liquid and solid fraction (Chuda and Ziemiński 2021a). The liquid fraction, containing several thousand mg/L of ammonia nitrogen, can be treated by biological methods, such as advanced nitrogen removal processes (nitritation-denitritation, nitritation-anammox), that have been recently reported (Monfet et al. 2018; Świątczak et al. 2019). However, because the conventional activated sludge process with autotrophic nitrification and heterotrophic denitrification is still the most common biological method used in on-site wastewater treatment plants, many agro-food factories are considering the possibility of its application for the treatment of the digestate liquid fraction (Yan et al. 2018; Yang et al. 2020).
Activated sludge technology is based on a complex microbial community, mainly bacteria, including those responsible for the removal of nitrogen and phosphorus compounds, as well as microorganisms devoid of direct bioremediation properties but cooperating with other community members (Cinà et al. 2019; Ríos-Castro et al. 2022; Wang et al. 2022). Therefore, a comprehensive understanding of microbial community structure, especially within specific microbial subpopulations, such as nitrifiers, denitrifiers, and their response to influent composition, is necessary for the efficient operation of wastewater treatment plants (Gao et al. 2016; Xu et al. 2018). Many techniques have been developed to study and characterize microbial communities over the past decade. Frequently used methods included those based on cultivation, traditional molecular biotechnology techniques, as well as more advanced high-throughput sequencing techniques, which provide high accuracy in the determination of complex microbial communities and their metabolic pathways (Zhang et al. 2018; Yang et al. 2020). However, despite such work, the relationships between microbial diversity and complex nutrients removal processes in activated sludge systems have not yet been fully explored (Cinà et al. 2019). Most studies of microbial community structure focus on activated sludge from municipal wastewater treatment plants, or plants in which a mixture of industrial and domestic wastewater is treated (Gao et al. 2016; Zhang et al. 2018). Meanwhile, knowledge of the microbial community of activated sludge from industrial wastewater treatment plants remains limited (Yang et al. 2020).
The microbial community is influenced by temperature, pH, oxygen concentration, organic and nitrogen loading rate, as well as the composition and COD/TN ratio of influent. If COD/TN ratio is excessively high, heterotrophic organisms use more available ammonia nitrogen compared to nitrifying organisms, resulting in decreased nitrification efficiency (Gao et al. 2016; Xu et al. 2018, 2019). In contrast, for wastewater with a low COD/TN ratio, such as digestate liquid fraction, the denitrification bacteria, which are mostly heterotrophic, cannot grow, resulting in incomplete denitrification. In such a situation, it is necessary to add an external carbon source to the influent. The most frequently used carbon sources are ethanol, methanol, and acetic acid. This is related to the low molecular weight, which makes them easily broken down by microorganisms and used by denitrifying bacteria. The drawback of these carbon sources is a relatively high price, which significantly increases the operating costs of treatment plants (Xue et al. 2018; Yan et al. 2018; Lan et al. 2022; Liu et al. 2022). Therefore, the application of food industry effluents and by-products as alternative carbon sources has recently attracted a lot of attention. The use of these cheap and organic substance-rich products could not only improve the denitrification process, but it also could contribute to their better management, and thus to the creation of a circular economy plant (Rodríguez et al. 2007; Xue et al. 2018; Yan et al. 2018).
The research attempted to determine the impact of the type and amount of external carbon source in the influent on the structure and diversity of the activated sludge microbial community, and thus to fully understand the complex process of biological nitrogen removal from the digestate liquid fraction. The study compared the composition of microorganisms consortium and the efficiency of digestate liquid fraction biodegradation obtained after the use of flume water and molasses, i.e., alternative carbon sources from the sugar factory with those achieved after adding acetic acid, a conventional carbon source, which was a reference point. No studies were found that focused on changes in the microbial community structure in combination with the efficiency of biological treatment of the digestate liquid fraction with the addition of external carbon sources from the sugar industry. These results could contribute to a better understanding of the biodegradation process of high-strength ammonia nitrogen wastewater in a conventional activated sludge system.
EXPERIMENTAL
Substrates Characteristics
Digestate liquid fraction
The study used the liquid fraction obtained after mechanical separation of digestate in the decanter centrifuge type UCD 305-00-32 (GEA, Warsaw, Poland) and stored at 4 °C. The digestate was collected from a biogas plant processing sugar beet pulp located in the Südzucker sugar factory. The chemical composition of digestate and its liquid fraction are presented in Table 1.
Table 1. Characteristics of Substrates
External carbon sources
In the study, acetic acid with a purity of 80% was used. Acetic acid is a conventional carbon source, often added to intensify the denitrification process (Xue et al. 2018). Flume water and molasses supplied by the sugar factory were also investigated. Flume water circulates in a closed circuit and is used for hydraulic transport and washing of sugar beet. It is rich in organic matter, which consists mainly of sugars resulting from losses at various points of the manufacturing process (Macarie and Mer 2006). Molasses is a dense, dark brown syrup, the most valuable by-product from the sugar industry with a sugar content of around 50% (Silva et al. 2009). The chemical characteristics of the listed carbon sources are summarised in Table 1.
Activated sludge
The activated sludge was taken from an industrial wastewater treatment plant (WWTP) located on the premises of the sugar factory. The wastewater treatment plant operates on the basis of the conventional activated sludge method in a system with preliminary denitrification.
Experimental Set-Up
The biodegradation tests of the digestate liquid fraction were conducted in cooperation with the biogas plant and wastewater treatment plant located in the Südzucker sugar factory. Three laboratory systems were used in the study, containing a separate denitrification chamber (DC), nitrification chamber (NC), and secondary settling tank. The internal recirculation rate between the chambers corresponded to 500% of the inflow. The rate of sludge recirculation from the secondary settling tank to the denitrification chamber was 120%. The concentration of mixed liquor volatile suspended solids (MLVSS) in the activated sludge amounted to 3.2 ± 0.7 g/L, and the age of the sludge averaged 30 days. Influents, consisting of the liquid fraction of digestate diluted with treated wastewater and an external carbon source (ext.), were fed into the denitrification chambers (Fig. 1). Acetic acid was an external carbon source in the first stage of the study involving DAA1 and DAA2 series, flume water in the second stage in DFW1 and DFW2 series, and molasses in the third stage in DMS1 and DMS2 series. The series, which lasted 30 days, differed in the CODext/NO3–N and COD/TN ratios, and thus in the amount of carbon source in the influent. The composition of the influents is presented in Table 2. During the biodegradation process, the total nitrogen loading rate was constant and amounted to 19.40 ± 1.57 mg N/g MLVSS d, while nitrogen load added with the liquid fraction of digestate accounted for 74.86 ± 6.38% of this value. The COD loading rate depended on the type of carbon source and CODext/NO3–N ratio and ranged from 172.99 ± 22.82 (DAA1, DFW1, DMS1 series) to 218.89 ± 29.79 mg O2/g MLVSS d (DAA2, DFW2 and DMS2 series). The method of determining CODext/NO3–N ratio in the influent, and thus COD/TN ratio, as well as the dimensions and operating parameters of the reactors were presented in detail in the authors’ previous article (Chuda and Ziemiński 2021b).
Fig. 1. Diagram with the flow of substrates during biological treatment of the digestate liquid fraction in the activated sludge system
Table 2. Composition of Influents
Analytical Methods
The methodology for analysing concentrations of total nitrogen (TN), ammonia nitrogen (NH4–N), nitrate nitrogen (NO3–N), nitrite nitrogen (NO2–N), chemical oxygen demand (COD), biochemical oxygen demand (BOD5), mixed liquor volatile suspended solids (MLVSS), as well as pH and oxygen content measurement methods were described in a previous article (Chuda and Ziemiński 2021b). The concentrations of volatile fatty acids (VFAs) and sugars were determined using the Agilent 7890A gas chromatograph (Agilent Technologies, Santa Clara, CA, USA) and UV-spectrophotometer (Thermo Scientific Multiskan Go, Thermo Fisher Scientific, Munich, Germany), respectively, following the procedures described by Cieciura-Włoch et al. (2021).
Microbial Community Analysis
DNA extraction
Total genomic DNA was extracted from sludge samples (250 g wet weight) using DNeasy PowerSoil Kit (Qiagen, Germantown, MD, USA) according to the manufacturer’s protocol with some modifications. The Qubit 2.0 fluorometer (Invitrogen/Life Technologies, Carlsbad, CA, USA) was employed for DNA quantity and quality analysis.
Polymerase chain reaction (PCR) amplification and high-throughput sequencing
The amplification of the V3-V4 region of the 16S rDNA was performed with primers 341F and 785R specific for bacteria (Klindworth et al. 2012). The PCR reactions were performed in triplicate with Q5 Hot Start High Fidelity 2x Master Mix (New England BioLabs Inc., Ipswich, MA, USA) under the following conditions: initial denaturation at 95 °C for 2 min, denaturation at 95 °C for 30 s (25 cycles), annealing at 55 °C for 30 s, extension at 72 °C for 1 min, with a final extension at 72 °C for 5 min. The PCR products were purified by AMPure XP magnetic beads (Beckman Coulter Life Sciences Inc., Indianapolis, IN, USA). The amplicon libraries were indexed using Nextera Index Kit (ThermoFisher Scientific, MA, Waltham, USA) following the manufacturer’s protocol.
High-throughput sequencing data analysis
High-throughput sequencing was performed by Genomed S.A. (Warsaw, Poland) on the MiSeq System employing MiSeq Reporter (MSR) v 2.6 software (Illumina, Inc., San Diego, CA, USA) in paired-end technology (PE, 2 × 250 bp). The QIIME 2 bioinformatic platform (https://qiime2.org/) with the SILVA v 138 database was used for the readings classification (Quast et al. 2013). The analysis contained: adapter sequences removal (cutadapt program), quality evaluation of the readings and low-quality sequences removal (cutadapt program), connecting paired sequences (SeqPrep algorithm), clustering based on the selected database of reference sequences (uclust algorithm), and sequence chimeras removal (usearch algorithm) (Edgar 2010; Martin 2011). Operational Taxonomic Units (OTUs) were assigned to taxa at an identity threshold of 97%. The diversity and evenness of the microbial community were analysed with alpha diversity estimators including Shannon (H), Simpson’s (D), and inverse Simpson’s (invD) indices determined using the Mothur program v. 1.30.1 (https://www.mothur.org; Schloss et al. 2009). The similarity of samples was presented on Venn diagrams (Venny 2.1, BioInfoGP Service, Centro Nacional de Biotecnologia, Madrid, Spain), based on specific OTUs for each activated sludge sample on genera level. Canonical correspondence analysis (CCA) was employed to determine the influence of operational parameters on the microbial communities. The CCA was conducted using XLSTAT v. 2022.1.1 (Lumivero, Denver, CO, USA) according to the manufacturer’s protocol based on datasets for OTU profiles and environmental variables.
Statistical Analysis
To determine the differences between concentrations of NO2–N, NO3–N, NH4–N, TN, COD, removal efficiencies of NH4–N, TN, COD, percentage of bacteria, and biodiversity indices for all types of carbon sources and COD/TN ratios, an analysis of variance (ANOVA) test followed by Tukey post hoc was performed. Samples for analysis were taken in triplicate. In all tests, differences were significant only when p < 0.05. All statistical analyses were performed using Statistica 12 program (StatSoft, Krakow, Poland).
RESULTS AND DISCUSSION
Diversity of Microbial Community of Activated Sludge
Following the analysis of the results presented in Table 3, it was found that type and concentration of the external carbon source in the influent affected the microbial community composition and diversity. On average, 108,653 ± 17,409 16S rDNA bacterial sequences were determined in the activated sludge collected from both chambers in series with the lower COD/TN ratios (DAA1, DFW1, DMS1), while it was 102,753 ± 13,180 in series with the higher COD/TN ratios (DAA2, DFW2, DMS2). Taking into account the carbon source type, the mean number of 16S rDNA sequences in the activated sludge from DAA series, in which acetic acid was used, amounted to 104,506 ± 6,273. In series with the alternative carbon sources, flume water (DFW series) and molasses (DMS series), it was 108,754 ± 26,699 and 103,822 ± 7,645, respectively. The OTUs number in the activated sludge from DAA, DFW, and DMS series averaged at 1683 ± 20, 2036 ± 86, and 1503 ± 118, respectively. The Shannon (H), Simpson’s (D), and inverse Simpson’s (invD) indices were in the range of 2.71 to 3.73, 0.07 to 0.17, and 6.02 to 14.94, respectively (Table 3).
Table 3. Alpha Diversity Index of the Microbial Community of the Activated Sludge Collected from DC and NC Chambers in All Series Differing in the Type of Carbon Source and COD/TN Ratio
A high H index, a high invD index, and a low D index indicate greater microbial diversity (Cinà et al. 2019; Yang et al. 2019). In the current study, the highest biodiversity (H index 3.64 to 3.73, invD index 10.63 to 14.94, D index 0.07 to 0.09) was determined in the activated sludge from both chambers in DFW series. This showed that the usage of flume water rich in a variety of organic compounds enabled the growth of more bacterial genera without causing any of them to dominate. It was noticed that in DFW, as well as DAA series, biodiversity was slightly higher with the lower COD/TN ratio. Chutivisut et al. (2018) and Xu et al. (2019) also observed that lower C/NO3– and COD/TN ratios determined a greater bacteria diversity in the population, which meant that various bacteria had a relatively large share among the dominating ones, while the higher ratios allowed the growth of only certain bacterial groups, but in higher proportions. In the described study, the lowest diversity expressed as Shannon (H) index (2.71 to 3.00) was found in DMS1 series in which the carbon source was molasses, and the COD/TN ratio was 9.5. Similar values of the H index, amounting to 2.39 and 3.29, were noted by Xu et al. (2019) during the treatment of synthetic wastewater with glucose and the COD/TN ratio of 10 and 7, respectively (Table 4).
Table 4. Structure and Diversity of Microbial Community in Various Treatment Systems
In the current study, the CCA was performed to determine the potential influence of the operational parameters (COD/TN ratio, CODext/NO3–N ratio, and external carbon source type) on the microbial communities of the activated sludges (Fig. 2). The length of arrows in the ordination plot indicates the strength of the environmental variables associated with the microbial community structure. The filled circles represent the different bacterial communities of activated sludge in both NC and DC chambers. The structure of the microbial community of the activated sludge from both chambers in DMS2 series was strongly related to the type of external carbon source, while in DFW2 series it was related to the COD/TN ratio. The weaker impact of operating parameters on the microbial community was observed in both chambers in DAA series.
Fig. 2. Canonical correspondence analysis (CCA) of high-throughput sequencing data and operational parameters in the bioreactors
Taxonomic Classification of the Microbial Communities
The composition of microbial communities of the activated sludges is shown in Fig. 3. Bacteria dominated all consortia (90.87 ± 3.23%), with remaining taxa belonging to Archaea. At the phylum level, 26 to 35 different bacterial phyla were identified in the activated sludges, whereas only four phyla of Archaea (Nanoarchaeota, Euryarchaeota, Halobacterota, and Iainarchaeota) were detected with a relative abundance being less than 0.05%.
Analysing the results in Fig. 3, it was noticed that Proteobacteria, Bacteroidota, and Firmicutes were the most abundant bacteria at the phylum level in the activated sludges. These bacteria are described in the literature as widely present in conventional wastewater treatment plants and denitrification environments, playing a significant role in removing organic components and nutrients (Gao et al. 2016; Xue et al. 2018; Xu et al. 2019; Wen et al. 2020; Yang et al. 2020; Ríos-Castro et al. 2022).
Fig. 3. Heatmap of bacterial phyla in the microbial community of the activated sludge collected in each series from the DC and NC chambers (given in terms of the number of OTUs)
The exact proportion of Proteobacteria, Bacteroidota, and Firmicutes in the microbial community depended on the type of carbon source and the COD/TN ratio in the influents. Proteobacteria was the most dominant phylum in the activated sludge in both DAA series in which acetic acid was the carbon source, followed by Bacteroidota and Firmicutes. The mean proportion of Proteobacteria, Bacteroidota, and Firmicutes amounted to 46.65 ± 2.34, 16.14 ± 1.76, and 4.48 ± 0.51% in the denitrification chambers, and 52.41 ± 2.07, 10.15 ± 1.37, and 1.25 ± 0.48% in the nitrification chambers, respectively. These results were similar to the amounts of these bacteria in the activated sludge during the treatment of domestic and industrial wastewater conducted by Xu et al. (2018) in the WWTP and the activated sludge collected by Gao et al. (2016) from full-scale municipal WWTP treated domestic wastewater (Table 4).
In DFW and DMS series with flume water and molasses as the carbon source, respectively, the same bacterial phyla were dominant, but their amounts were distributed differently than in series with the addition of acetic acid. The proportion of Proteobacteria, Bacteroidota, and Firmicutes in DFW series averaged at 19.68 ± 2.15, 44.20 ± 2.90, and 29.02 ± 0.09% in the denitrification chambers, and 20.99 ± 0.06, 38.18 ± 3.09, and 33.80 ± 2.35% in the nitrification chambers, respectively. The COD/TN ratio did not have a significant (p > 0.05) effect on the populations of mentioned bacteria in DAA and DFW series. It had a more noticeable effect on the microbial community composition in DMS series with molasses. In DMS2 series, in which the COD/TN ratio was 11.4, the proportion of Proteobacteria, Bacteroidota, and Firmicutes in the activated sludge averaged at 14.63 ± 1.02, 29.35 ± 1.12, and 28.05 ± 1.43% in the denitrification chamber, and 11.03 ± 0.98, 33.45 ± 1.54, and 30.47 ± 1.43% in the nitrification chamber, respectively. In DMS1 series, in which the COD/TN ratio was lower and amounted to 9.5, the proportion of Proteobacteria was 13.54 ± 1.05% in the denitrification chamber and 17.06 ± 1.12% in the nitrification chamber, Bacteroidota 35.26 ± 1.49 and 42.68 ± 1.82%, Firmicutes 48.60 ± 1.74 and 36.80 ± 1.56%, respectively. Xu et al. (2019) conducting the treatment process of synthetic wastewater with glucose defined the proportion of Proteobacteria in the activated sludge at the level of 68.38%, and Bacteroidetes at 14.19% with the COD/TN ratio similar to DMS1 series and equal to 10 (Table 4). Analysing the obtained results, it was noticed that Proteobacteria in DFW and DMS series occurred in lower amounts than in DAA series, whereas percentages of Firmicutes and Bacteroidota were significantly (p < 0.05) higher. This was probably because of the ability of Firmicutes and Bacteroidota to break down complex polysaccharides in hydrolytic and fermentative processes. By contrast, Proteobacteria comprises a broad variety of aerobic, facultative, and anaerobic bacteria, which can utilise various organic substrates to produce volatile fatty acids (Gao et al. 2016; Xue et al. 2018; Xu et al. 2019; Baskaran et al. 2020).
Predominant Bacterial Genera in Biological Treatment of the Digestate Liquid Fraction
The relative abundance of the microbial communities of the activated sludges at the genus level under different external carbon sources and the COD/TN ratios is presented in Fig. 4. In total, 522 genera of bacteria were detected. The article focuses only on the dominant genera and those significant from the point of view of the discussed process.
Analysing the Venn diagram (Fig. 4A through C) showing the similarities of the composition of the activated sludge from the NC and DC chambers at different COD/TN ratios, it was noted that the number of shared OTUs at the genus level in DFW series was the highest at 228 (Fig. 4B). The number of shared genera in DAA (Fig. 4A) and DMS (Fig. 4C) series was lower and amounted to 175 and 108, respectively. The shared OTUs indicated that some bacteria could exist in the activated sludge regardless of the COD/TN ratio and the chamber type. A possible explanation for this observation was the high rate of internal recirculation of oxygenated treated wastewater from the nitrification chamber into the denitrification one. This allowed some aerobic bacteria to survive in the anoxic unit. The unique OTUs specific to the defined environmental conditions were also present in the activated sludges. Their number in DAA series (NC and DC) was 1 to 9, in DFW series 5 to 13, while in DMS series it was in the range of 4 to 15.