Abstract
The purpose of the study was to link microbial community composition and chemical properties of various biomass and their resulting digestate residues for their potential use in biogas production or soil enrichment. The order of biogas production, graded from high to low was as follows: corn silage, grass silage, fruit waste, and dairy sewage sludge. Different bacterial families were predominant in different biomass. Corn silage deteriorated as a result of long-term air exposition and may serve as an efficient feedstock substrate for anaerobic digestion. A positive role in plant biocontrol microorganisms found in grass straw residues, and reasonable biogas yield obtained from this substrate suggests the use of grass straw for biogas production and its residues to enrich the soil. Due to potential threat of introducing pathogens into the soil within fruit waste or dairy sewage sludge, or soil acidification by fruit waste repeated use in field application, this biomass should be sanitized prior to soil application. Simultaneously, low biogas yields from fruit waste and dairy sewage sludge substrates make it necessary to transform them in anaerobic digestion with more energetic co-substrates. Tested residues may deliver a robust and wide range of methanogens as inoculum for further anaerobic digestion process.
Download PDF
Full Article
Evaluation of Microbial Community Composition of Dairy Sewage Sludge, Corn Silage, Grass Straw, and Fruit Waste Biomass for Potential Use in Biogas Production or Soil Enrichment
Karolina Oszust * and Magdalena Frąc
The purpose of the study was to link microbial community composition and chemical properties of various biomass and their resulting digestate residues for their potential use in biogas production or soil enrichment. The order of biogas production, graded from high to low was as follows: corn silage, grass silage, fruit waste, and dairy sewage sludge. Different bacterial families were predominant in different biomass. Corn silage deteriorated as a result of long-term air exposition and may serve as an efficient feedstock substrate for anaerobic digestion. A positive role in plant biocontrol microorganisms found in grass straw residues, and reasonable biogas yield obtained from this substrate suggests the use of grass straw for biogas production and its residues to enrich the soil. Due to potential threat of introducing pathogens into the soil within fruit waste or dairy sewage sludge, or soil acidification by fruit waste repeated use in field application, this biomass should be sanitized prior to soil application. Simultaneously, low biogas yields from fruit waste and dairy sewage sludge substrates make it necessary to transform them in anaerobic digestion with more energetic co-substrates. Tested residues may deliver a robust and wide range of methanogens as inoculum for further anaerobic digestion process.
Keywords: 5764
Contact information: Institute of Agrophysics, Polish Academy of Sciences, Doświadczalna 4, 20-290 Lublin, Poland; *Corresponding author: k.oszust@ipan.lublin.pl
INTRODUCTION
In recent years, the use of biomass for various management, such as biogas production or soil enrichment, has received increasing attention and is regarded as a simple commodity that has a value. Since biomass processing is microbially mediated and microorganisms quickly react to the changes occurring in environments, microbial community composition can provide evidence of how biodegradation is proceeding.
Accordingly, as methane is produced by the activity of several microorganisms, gaining knowledge about the ecophysiology of the microbes enhances the understanding of their particular roles (Goswami et al. 2016). Methanogenesis is the key step for methane production (Schink et al. 2017), but the microbiologically controlled hydrolysis of complex macromolecules can be regarded as the rate-limiting step of plant biomass bioconversion (Veeken and Hamelers 1999). These aspects drive the need for in-depth knowledge of the microbial community composition of feedstock substrates. It is also necessary to analyse and control the quality of the physicochemical properties and the microbiological composition of these substrates (Insam et al. 2015). At present, the application of high-throughput sequencing technologies (NGS) (e.g., Illumina sequencing platforms) to 16S rDNA gene amplicon sequencing provides the high resolution required for studying the composition of bacterial and archaeal communities of feedstock substrates (Lim et al. 2017), anaerobic digesters (Vanwonterghem et al. 2014; Bozan et al. 2017), anaerobic digestion residues (Oszust et al. 2015), and soil after residue application as a source of exogenous organic matter (Su et al. 2017).
Among the various types of feedstock biomass available for the production of biogas, corn silage is currently the most favoured, particularly in Central and Western Europe (Oleszek et al. 2016). However, a sufficient yield of corn, as well as cellulose-based substrates obtained from energy plants biomass, e.g., miscanthus, reed canary grass, or Virginia mallow, may be obtained only from fertile soils, which should rather be used for food production, not for energy crops (Giovannetti and Ticci 2016). Moreover, pretreated cellulosic biomass realize high yields that are crucial to commercial success in biological conversion (Wyman et al. 2005). Although many pretreatment methods are known, e.g., ammonia explosion, aqueous ammonia recycle, controlled pH, dilute acid, flow through, lime, or biological approaches (Wyman et al. 2005; Maroušek and Kwan 2013), additional steps, especially enzymes hydrolytic application, may increase the total costs of biogas production (Zheng et al. 2009).
As an alternative for corn or energy plants, organic waste-based type fermentation substrates were previously studied (Oszust et al. 2017). The cited authors indicated that lower biogas yields are obtained from waste, compared with corn silage. However, substrates mixed for co-fermentation produce satisfactory results (Gómez et al. 2016; Böjti et al. 2017). In some countries (Germany, Italy) it is permitted to add energy-rich substrates (up to 20%) to make reactors more efficient (Insam et al. 2015).
In choosing substrates for biogas production, the quantity of these substrates available in the local area matters (Montusiewicz and Lebiocka 2011). The constitution of these substrates should also be taken into consideration, especially the biodegradable compound content and microbiological composition. The methane yields from cow manure, chicken manure, and a straw mixture ratio have been reported (Li et al. 2017). Furthermore, food expellers and sludge can be used for biogas production by anaerobic fermentation (Oszust et al. 2017).
Fruit industry wastes, dairy sewage sludge, and grass, especially those acquired from fallow lands, are readily available and suitable for biodegradation, material recovery, and energy production. Using such substrates can reduce the cost of the process, because there are no transportation expenses. What is more, such substrates are permitted by legislation (Frąc and Ziemiński 2012). Waste from fruit processing residues consists mainly of woody stalks and leaves. These are produced in large quantities and constitute a source of nutrients (cell wall polysaccharides, such as pectin, cellulose, and hemicellulose) (Bouallagui et al. 2005). Similarly, sewage sludge, a byproduct of the dairy industry, is a valuable substrate for methane production because it is rich in fat and protein (Frąc et al. 2014).
The agrochemical value of the residues biomass from a biofermenter was evaluated previously (Kolář et al. 2010). As a fertilizer in general terms it was reported to have higher ion exchange and buffering capacity than the material before anaerobic fermentation (Kolář et al.2008). Advantages of applying biochar from the fermentation residue in crop production were explained previously by Maroušek (2014). Thus, biogas residues may be converted to a variety of value-added byproducts that can be applied to the soil (Chanakya et al. 1999). Thus, the utilization of biogas residue as exogenous organic matter for field applications is a form of soil conditioning to enhance crop yield. The microbial biomass and metabolic activities of soil are comprehensively stimulated after the application of digestate residues; this phenomenon is attributed to the supplementation of the levels of carbon and nutrients in the soil (Frąc 2012). Therefore, the soil microbial response after residue application is currently being evaluated (Van Nguyen et al. 2017). Biogas residues are also valuable products because they contain microbial components, which provide the key factor to ensure the success of anaerobic digestion. To the best of our knowledge, the methanogen composition of residues has not been analysed thoroughly to date. The most recent study of Zhao et al. (2017) highlights the role of microbes in choosing acclimated anaerobic sludge (biogas residues) as microbial and nutritional regulators to improve the biomethanation of fruit wastes. The hypothesis of our work was that microbial community composition of dairy sewage sludge, corn silage, grass straw, fruit waste biomass differs among each other. Moreover, their resulting digestate residues are varied, and elucidating microbial community composition may suggest the most relevant ways to manage or utilize these types of biomass.
The purpose of the study was to link microbial community composition and chemical properties of dairy sewage sludge, corn silage, grass straw, fruit waste biomass, and their resulting digestate residues for their potential use in biogas production or soil enrichment. Therefore, we 1) evaluated the physicochemical properties of feedstock substrates and digestate residues, and 2) determined their core microbial community composition of biogas substrates and digestate residues. It is worth mentioning that the metagenomics of the microbial community in fruit waste and dairy sewage sludge has not been reported before. This research intends to elucidate the role of the biomass microbial community for biogas yield effectiveness, and determine the role of methanogens, which occur in digestate residues, following biogas production. Thus, the biogas yield kinetic study of substrates was linked to the physiochemical parameters and microbial community composition. What is more, both opportunities and threats related to incorporating biomass into the soil are highlighted.
EXPERIMENTAL
Biomass Characteristic
Eight different organic biomasses were evaluated: fruit waste (FW), dairy sewage sludge (DSS), grass straw (GS), corn silage (CS), fruit waste digestate residues (FWR), dairy sewage sludge digestate residues (DSSR), grass straw digestate residues (GSR), and corn silage digestate residues (CSR).
FW – waste from soft fruit processing residues consists mainly of spoiled raspberries, strawberries, and currants peel expeller and pulp flakes after squeezing the juice, as well as woody stalks and leaves;
DSS – waste taken from the dairy company, where it was formed in a mechanical-biological treatment plant as excess sludge after purification of wastewaters from technological lines;
GS – an air-dried mix of the most common grass species found in Poland, namely: Phragmites australis, Poa pratensis, Festuca arundinacea, Festuca rubra, Elymus (Agropyron) repens, Dactylis glomerata, Arrhenatherum elatius, Lolium perrene, Calamagrostis epigejos;
CS – prepared from corn forage ensiled in silos and covered by foil, exposed to air prior to anaerobic digestion
FWR, DSSR, GSR, CSR were generated in anaerobic digestion of FW, DSS, GS, CS, respectively as described below. Each of the biomass samples was transported in a portable refrigerator into the laboratory and afterwards frozen immediately for further analyses.
Anaerobic Digestion
The details of the anaerobic digestion experiment are shown in Table 1. Based on the preliminary results, the initial load for the fermenter was selected. The weighed portion of the 2 to 4 mm feedstock substrate along with the inoculum was placed in a sealed fermentation vessel with a working 500 mL volume. The fermenters (three replications for each substrate) were placed in a 37 °C water bath. The biogas was transferred to a cylindrical, calibrated gas manifold filled with acidified water. The accumulated gas displaces water from the collector to an overflow tank. The level of gas in the collector was recorded every 24 h. An analysis of biogas composition was performed periodically. Fermentation was carried out until there were no significant increases in biogas volume. For this study an anaerobic sediment (obtained from and standardized by the laboratory of Biogaz Zeneris Sp. z o. o., Poznań, Poland) that was starved and nonadapted was used as the inoculum. The term “nonadapted” indicates that the sediment was not, prior to the experiment, supplied with the substrate for which biogas potential was investigated. Biogas composition was determined by (1) biogas analyser GA 2000 Plus (Geotech, Rzeszów, Poland), (2) GFM 410 (GAS DATA Ltd., Coventry, Great Britain); analyses included: CH4, CO2, O2, NH3, H2S and H2), and (3) VARIAN MicroGC – 4900 gas chromatograph (Palo Alto, USA); analyses included: CH4, CO2, O2, N2, H2S, and H2).
Table 1. Anaerobic Digestion Parameters
Physicochemical properties of biomass
Dry matter (d.m.) was evaluated using the gravimetric method, organic dry matter (o.d.m.) was also evaluated using the gravimetric method, pH was measured using the electrometric method, and ammonium nitrogen (AN) was determined with Spectroquant cuvette tests (MERCK). Total nitrogen (TN) was determined using the Kjeldahl method (% d.m.), whereas total carbon (TC) was determined using solid sample mineralization in a furnace and the detection of combustion products in the detector of the central unit of the TOC apparatus, according to (Szarlip et al. 2014). Chemical oxygen demand (COD) was evaluated using the dichromate method; volatile fatty acid (VFA) content was determined using the Spectroquant tube test assay in supernatant after centrifugation (MERCK). The total ion content (Cr, Ni, Cu, Zn, Cd, Pb, Mg, K, Ca, Hg, and P) of the substrates was evaluated using the inductively coupled plasma with mass spectrometry (ICP-MS) method after microwave digestion (Gałązka and Gembal 2015).
Metagenomic Analysis by NGS
DNA isolation
The DNA from different types of biomass was extracted using a FAST DNA Spin Kit for Feces (MPBiomedicals, Santa Ana, CA, USA) according to the protocol, as described previously by Gryta et al. (2017). The amount of DNA was determined by a NanoDrop® 2000 spectrophotometer (Thermo Scientific™, Waltham, MA, USA).
16S rDNA gene amplification and amplicon sequencing
The MiSeq 2000 platform (Illumina Inc., San Diego, CA, USA) was applied to sequence the DNA of microorganisms. Polymerase chain reaction (PCR) was performed with NebNext High-Fidelity 2x PCR Master Mix (New England BioLabs, Ipswich, MA, USA) according to the manufacturer’s protocol and the following primers: 515F (5’-AAT GAT ACG GCG ACC ACC GAG ATC TAC ACT ATG GTA ATT GTG TGC CAG CMG CCG CGG TAA-3’) and 806R (5’-CAA GCA GAA GAC GGC ATA CGA GAT XXX XXX XXX XXX AGT CAG TCA GCC GGA CTA CHV GGG TWTCTA AT-3’) for the v4 region of 16S rDNA (Caporaso et al. 2012). The conditions for the 16S rDNA genes were as follows: 94 °C for 3 min; 35 cycles of 94 °C for 45 s, 50 °C for 60 s, and 72 °C for 90 s; and a final extension at 72 °C for 10 min. The libraries were indexed in TrueSeq (Illumina, San Diego, CA, USA). Sequencing was performed in PE reads 2 x 250 bp with an Illumina v2 MiSeq reagent kit.
Calculations
Processing of the sequencing data
A bioinformatics assay was carried out based on the reference sequence database Greengenes_13_05 (DeSantis et al. 2006) and was performed using an algorithm from Qiime software (Caporaso et al.2010). The analysis included the following steps: demultiplexing of samples and adaptor cutting, quality analysis, taxonomic composition, and diversity analysis. Sequences which were over 97% identical were grouped into one operational taxonomic unit (OTU) using a distance-based OTU program. The application of MiSeq Reporter v2.3 allowed for classifications at a species level. The taxonomy database for the metagenomics workflow was the Illumina version of the Greengenes database (DeSantis et al. 2006).
Statistical analysis
All statistical analyses were performed on operational taxonomic units (OTUs) data > 1% of occurrence at least in one sample. The dendrogram was based on scaled axis bond distances (Ward’s method, within Euclidean distance), with marked boundaries of Sneath’s criteria. Cluster analyses were performed using Statistica software (version 10.0). Dissimilarity displays were made using Similarity Percentages (SIMPER) software. Principal component analysis (PCA) was performed using the PAST 3.14 program (Hammer et al. 2008). Thus, multivariate analysis with ordination was applied to generate a biplot. The Krona visualization tool allowed for the exploration of relative abundances, and confidences within the hierarchies of metagenomics classifications were used (Ondov et al. 2011). Similarity percentages – family contributions were calculated using the PRIMER 7 program (Clarke and Gorley 2006).
RESULTS
Physicochemical Properties of Biomass
Daily gas production increased gradually as the microbial community adapted to the reactor environment. The highest cumulative biogas and methane yield was noted from anaerobic digestion of silages. In CS it was 533 dm3/kg d.m. and 356 dm3/kg d.m., accordingly. However the highest methane content in biogas was achieved after DSS digestion (75.7%). The highest daily biogas production as depicted by constant biogas yield was obtained on the fourteenth day of the anaerobic digestion process (Fig. 1).
Fig. 1. Efficiency of anaerobic digestion. Explanations: please see Table 1. Vertical bars indicate standard deviations, n=3
Table 2. Physicochemical Properties of Biomass
The following chemical properties of the feedstock substrates (fruit waste (FW), dairy sewage sludge (DSS), grass silage (GS), and corn silage (CS)) were evaluated: dry matter (d.m.), organic dry matter, ash, pH, ammonium nitrogen, total nitrogen, total organic carbon, chemical oxygen demand, volatile fatty acids, the content of elements, and biogas composition (Tables 2 and 3).
Table 3. Biogas Yield and Composition
Dry matter content ranged from 8.1% in FW to 60.6% in GS. Organic dry matter content in DSS and GS was relatively constant and reached 86% d.m. In CS and FW, organic dry matter was found to be from 94% to 96% d.m. Ash content in the tested samples was rather low (from 3% to 13%). The highest amount of total nitrogen was measured in DSS (7.9% d.m.), and the lowest amount of total nitrogen was measured in CS and GS (1.39% d.m. and 1.93% d.m., respectively). Substrates showed different C/N ratios. This ratio reached 21.8, 5.1, 16.3, and 30.1 in FW, DSS, GS, and CS, respectively. The VFA after fermentation ranged from 259 mg CH3COOH dm-3 in FW, 285 mg in CS, 384 mg in GS, and 398 mg in DSS. Ammonium nitrogen after fermentation was lower in FW and CS (1181 mg NH4+ dm-3 and 1239 mg NH4+ dm-3) than in DSS and GS (1561 mg NH4+ dm-3 and 1534 mg NH4+ dm-3, respectively). The amounts of such heavy metals as Cr, Ni, Cu, Cd, Pb, and Hg ranged from less than 1 ppm to 23 ppm. Slightly higher amounts of Zn were noted, they ranged from 25 ppm, 44 ppm, 63 ppm, and 89 ppm in CS, FW, GS, and DSS, respectively. Phosphorus content was rather constant in the tested samples. Substrates differed significantly as far as K and Ca content were concerned. Ca amounts of 4137 ppm, 14804 ppm, 1909 ppm, and 29272 ppm were noted in FW, GS, CS, and DSS, respectively. As far as residues ions content was noted to be higher than in substrates. Macroelements like K and P contents increased in residues, however N decreased. For example from 7.9% d.m. in DSS and 2.36% d.m. in FW to 0.02% d.m. in corresponding residues, with disparate d.m. content for the tested biomass.
Biomass Metagenomic Analysis by NGS
The obtained sequences, among feedstocks and resulting digestate residues were classified into 11 phyla (Euryarchaeota, Actinobacteria, Bacteroidetes, Chloroflexi, Cyanobacteria, Fibrobacteres, Firmicutes, Planctomycetes, Proteobacteria, Spirochaetes, and Synergistetes), one unidentified phylum, 16 classes, within 26 orders and 48 families (with 13 separate but not identified) across the entire data set. Although the 16S rRNA sequencing currently represents the most important study target in bacterial ecology, this is biased by the presence of variable copy numbers in bacterial genomes and sequence variation within closely related taxa or within a genome (Větrovský and Baldrian 2013).
The results of cluster analysis shown in Fig. 2 indicated evident differences among the samples when the family level of taxonomic classifications was taken into consideration. A dendrogram with clustering samples taken at the family level according to stringent Sneath’s criteria revealed residues that form groups, which indicates that their microbial communities compositions were similar. Each sample of the substrate represented a separate group when strict criteria are applied. The FW and DSS met the 66% similarity criterion, and they represented one group that indicates that microbial composition might be similar.
Fig. 2. Dendrogram with clustering samples on family level according to the stringent Sneath`s criterion (33%) and less restrictive criterion (66%). Explanations: “f” family non-identified, classification according to order, fruit waste (FW), dairy sewage sludge (DSS), grass straw (GS), corn silage (CS), digestate fruit waste residues (FWR), dairy sewage sludge residues (DSSR), grass straw residues (GSR), and corn silage residues (CSR)
The average dissimilarity displayed for each pairwise combination (Simper approach) of feedstock and residues groups as presented in Table 4 and the indices show how the sample communities differ and what is the particular contribution to this dissimilarity for each particular family. Among residues, the average microbial dissimilarity was low, indicating relatively high resemblance between communities, that ranged between 7.79% dissimilarity for CSR and GSR to 20.09% for DSSR and GSR; whereas in substrates the range went from 85.56% for DSS and FW to 98.14% for CS and GS. When comparing CS and GS, the largest influence on the disparity between those two samples was Acetobacteraceae (30.45%), GS and DSS – Bacillaceae (24.50%), FW and CS – Acetobacteraceae (32.50%), DSS and CS – Acetobacteraceae (32.36%), FW and GS – Streptophyta (27.73%), DSS and FW – Streptophyta (29.26%), DSSR and GSR – Marinilabiaceae (15.30%), CSR and DSSR – Marinilabiaceae (15.06%), FWR and DSSR – Tissierellaceae (12.48%), CSR and FWR – Marinilabiaceae (15.12%), GSR and FWR – Marinilabiaceae (18.30%), CSR and GSR – Propionibacteriaceae (14.38%). This supports the findings of cluster analysis (Fig. 2) with respect to the residues grouped together.
Table 4. Dissimilarity Displayed for Each Pair-Wise Combination of Feedstock and Residues Groups
Table 4 cont. Dissimilarity Displayed for Each Pair-Wise Combination of Feedstock and Residues Groups
Fig. 3. The environment-vector view of the microbial composition biplot on family level (a). Explanations: please see Table 4
Fig. 3 cont. The environment-vector view of the microbial composition narrowed down biplot on family level (b). Explanations: please see Table 4.
Even closely similar microbial composition (dissimilarity <20%, as measured by the percentage of its abundance contribution) may be considered. Nevertheless, this approach provides an insight into the differences in pairwise combinations of samples. The general overview on sample grouping with respect to microbial relative abundance and its taxonomic classification are shown in the biplot exploratory graph (Fig. 3) of principal component analysis (PCA). The corresponding results to cluster analyses and the Simper approach are to be found in PCA. PC1 and PC2 explained 31.66% and 36.84% of the variance. Biplot vectors clearly indicate that families mainly occurred in tested substrates and residues. CS was mostly inhabited by Acetobacteraceae (63.9% of the individuals among the whole community) and Lactobacillaceae (31.8% among all of the revealed families in total); GS by Bacillaceae (49.2%) and Planococcaceae (45.2%), DSS by Xanthomonadaceae (23.7%), Enterococcaceae (16.1%), Nocardioidaceae (9.5%), unidentified family of the Rhizobiales order (5.5%) and others from Proteobacteria (<4.5%) (Caulobacteraceae, Methylocystaceae, Phyllobacteriaceae, Rhodobacteraceae) and Firmicutes (<3.2%) (Lactobacillaceae, Streptococcaceae); FW was primarily inhabited by an unidentified family from the Streptophyta order (51.3%) and an unidentified family from Rickettsiales (12.4%). Residues are differentiated as far as microbial composition is concerned compared to the substrates. The dissimilarity resulted from a higher content of Clostridiales, especially Clostridiaceae (22.5% – 23.9%) than in the substrates (<1.5%). After anaerobic digestion in biogas residues Euryarcheota (Methanosaetaceae primarily from 12.5% to 14.7% of all the families) and Bacteroidetes (Propionibacteriaceae, Marinilabiaceae, Porphyromonadaceae, and two unidentified families) were noted, whereas in substrates, their amount was rather low <0.2%.
Figure 4 (a–d) particularly represents the archaeal community composition of biogas residues (CSR, GSR, DSSR, FWR, respectively). Archaea communities comprised 21% of the whole microbial community tested using the 16S rDNA marker in CSR and GSR, whereas DSSR and FWR were 20% and 22%, respectively. In CSR only Euryarcheota (100%) occurred. This is the methanogens group, and 91% of it consisted of Methanobacteriales (Methanosaetaceae 60%, Methanosarcinaceae 14%, Methanospirillaceae 11%, Methanobacteriaceae 5%) and 9% belonged to an unidentified order. The GSR community composition was found to be similar to CSR (1% dissparity) (Table 4). In FWR, Methanobacteriales (93% of Methanobacteria) consisted of the following families: Methanosaetaceae (73%), Methanosarcinaceae (17%), Methanobacteriaceae (6%), Methanospirillaceae (3%), Methanocorpusculaceae (1%). Among Methanobacteriales of DSSR (90% Methanobacteria) as much as 77% was Methanosaetaceae, 15% was Methanosarcinaceae, 4% Methanospirillaceae, 2% Methanospirillaceae and 2% Methanocorpusculaceae.
DISCUSSION
The production of methane via the anaerobic digestion of organic substrates within the inoculum including methanogenic Archaea is accomplished by the intricate relationship of microbial communities of feedstock substrates and inoculum components. According to Perrotta et al. (2017), inoculum-specific outcomes in the experiment, where different inocula and the same substrate are subjected to anaerobic digestion, suggests the influence of such factors as species-species interaction. We assumed that apparently the relatively analogical situation occurs if an attempt includes the use of the same inoculum for diverse substrates. Thus, occurring interaction depends on both chemical properties of substrates and their inherent microbial community composition. Consequently, biogas yield strictly depends on its chemical composition and on the susceptibility of its organic compounds to decomposition under anaerobic conditions (Vanwonterghem et al. 2014; De Vrieze and Verstraete 2016; Fitamo et al. 2017a).
Fig. 4. Methanogen’s community composition of biogas residues (a) dairy sewage sludge (DSSR), (b) fruit waste (FWR)