Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

A comparative analysis of biogas production from tomato bio-waste in mesophilic batch and continuous anaerobic digestion systems

  • Árpád Szilágyi,

    Roles Conceptualization, Investigation, Writing – original draft

    Affiliation Department of Biotechnology, University of Szeged, Szeged, Hungary

  • Attila Bodor,

    Roles Methodology, Visualization

    Affiliations Department of Biotechnology, University of Szeged, Szeged, Hungary, Institute of Environmental Sciences, University of Szeged, Szeged, Hungary, Institute of Biophysics, Biological Research Centre, Szeged, Hungary

  • Norbert Tolvai,

    Roles Investigation

    Affiliation Department of Biotechnology, University of Szeged, Szeged, Hungary

  • Kornél L. Kovács,

    Roles Formal analysis, Writing – review & editing

    Affiliations Department of Biotechnology, University of Szeged, Szeged, Hungary, Institute of Biophysics, Biological Research Centre, Szeged, Hungary, Department of Oral Biology and Experimental Dental Research, University of Szeged, Szeged, Hungary

  • László Bodai,

    Roles Investigation

    Affiliation Department of Biochemistry and Molecular Biology, University of Szeged, Szeged, Hungary

  • Roland Wirth,

    Roles Data curation, Software

    Affiliation Department of Biotechnology, University of Szeged, Szeged, Hungary

  • Zoltán Bagi,

    Roles Validation

    Affiliation Department of Biotechnology, University of Szeged, Szeged, Hungary

  • Ágnes Szepesi,

    Roles Investigation

    Affiliation Department of Plant Biology, University of Szeged, Szeged, Hungary

  • Viktória Markó,

    Roles Methodology

    Affiliation Department of Biotechnology, University of Szeged, Szeged, Hungary

  • Balázs Kakuk,

    Roles Data curation, Formal analysis

    Affiliation Department of Biotechnology, University of Szeged, Szeged, Hungary

  • Naila Bounedjoum,

    Roles Writing – review & editing

    Affiliation Department of Biotechnology, University of Szeged, Szeged, Hungary

  • Gábor Rákhely

    Roles Conceptualization, Data curation, Supervision, Writing – review & editing

    rakhely.gabor@brc.hu

    Affiliations Department of Biotechnology, University of Szeged, Szeged, Hungary, Institute of Environmental Sciences, University of Szeged, Szeged, Hungary, Institute of Biophysics, Biological Research Centre, Szeged, Hungary

A comparative analysis of biogas production from tomato bio-waste in mesophilic batch and continuous anaerobic digestion systems

  • Árpád Szilágyi, 
  • Attila Bodor, 
  • Norbert Tolvai, 
  • Kornél L. Kovács, 
  • László Bodai, 
  • Roland Wirth, 
  • Zoltán Bagi, 
  • Ágnes Szepesi, 
  • Viktória Markó, 
  • Balázs Kakuk
PLOS
x

Abstract

Annually, agricultural activity produces an enormous amount of plant biomass by-product. Many studies have reported the biomethane potential of agro-industrial wastes, but only a few studies have investigated applying the substrates in both batch and continuous mode. Tomato is one of the most popular vegetables globally; its processing releases a substantial amount of by-product, such as stems and leaves. This study examined the BMP of tomato plant (Solanum lycopersicum Mill. L. cv. Alfred) waste. A comparative test revealed that the BMPs of corn stover, tomato waste,and their combination were approximately the same, around 280 mL methane/g Volatile Solid. In contrast, the relative biogas production decreased in the presence of tomato waste in a continuous mesophilic anaerobic digestion system; the daily biogas productions were 860 ± 80, 290 ± 50, and 570 ± 70 mL biogas/gVolatile Solid/day in the case of corn stover, tomato waste, and their mixture, respectively. The methane content of biogas was around 46–48%. The fermentation parameters of the continuous AD experiments were optimal in all cases; thus, TW might have an inhibitory effect on the microbial community. Tomato plant materials contain e.g. flavonoids, glycoalkaloids (such as tomatine and tomatidine), etc. known as antimicrobial and antifungal agents. The negative effect of tomatine on the biogas yield was confirmed in batch fermentation experiments. Metagenomic analysis revealed that the tomato plant waste caused significant rearrangements in the microbial communities in the continuously operated reactors. The results demonstrated that tomato waste could be a good mono-substrate in batch fermentations or a co-substrate with corn stover in a proper ratio in continuous anaerobic fermentations for biogas production. These results also point to the importance of running long-term continuous fermentations to test the suitability of a novel biomass substrate for industrial biogas production.

Introduction

There is a growing interest in renewable energy sources due to the depletion of fossil fuels and their negative effects on the environment. One of the most common non-conventional energy carriers is biogas produced in anaerobic digestion (AD) processes. A wide variety of biomasses, such as maize or grass silage, crop plant and agricultural by-product, wastewater sludge, food processing by-products, domestic organic waste, and manure, have been used as substrates in biogas plants [13]. The stability and productivity of the biogas plants could be affected by diverse substrates. Moreover, the microbial community composition fluctuates and varies with the substrates [1].

AD has an increasing role in the agricultural sector for the energetic utilization of organic wastes [4]. Approximately 5 billion hectares are used for agricultural production worldwide [5], generating large volumes of plant biomass by-products, such as orange, onion, and potato peels, green plant residues, and tomato pomace [6].

Many studies have reported that various kinds of agricultural by-product, e.g., cabbage, capsicum, pepper, cucumber, eggplant, and tomato residue, are suitable for biogas production [4, 7, 8].

A lignocellulosic biomass’ energy potential depends on its composition, such as the lignin, hemicellulose, and cellulose ratio, and the microbial community formed or designed to convert these fibers to energy carrier. However, various plant species might produce antimicrobial agents that negatively affect the microbial activities during e.g. AD [911]. Moreover, the methane yield can be influenced by parameters, such as temperature, pH, volatile fatty acids (VFAs), ammonia, macro- and micronutrients, and potentially toxic compounds. The toxic or inhibitory compounds could modify microorganisms’ activity [2, 12].

Therefore, it will be useful to investigate the optimal mix of substrates and the conditions of fermentation. While the biomethane potential (BMP) of many wastes in batch fermentation experiments have been tested, there are just a few studies comparing the batch fermentation to continuous fermentation [4, 7]. Batch fermentations are suitable in estimating a substrate’s methane yield and biodegradability, but they do not provide information about the long-term effect of the substrate on the fermentation.

Several standard methods, such as DIN 38414 TL8, ASTM D 5210, ISO 11734, ISO 14853, ISO 15985, and VDI 4630, are used to evaluate BMP [13]. These standards prescribe the conditions of fermentation, including the amount of inoculum, substrates, and medium, the definition of the necessary controls, blanks, and replicates, and the specification of the experimental setup [13]. In continuous AD system, the daily feeding of reactors, the removal of fermentation liquid, the harvest of biogas, and the monitoring of the fermentation parameters are carried out periodically over the long term. In this experimental setup, the durable effect of the applied substrate could be examined [14].

Also, the rapid expansion of “omics” approaches has enabled a better understanding of the AD process. Analyzing the structure, composition, and activity of the involved microbial communities and combining the metagenomics and culture-dependent methods are necessary to gain a deeper insight into the fermentation processes, identify the key factors involved in the optimization of the AD process parameters and enhancement of the biogas yield from a more diversified group of plant biomass waste substrates [15, 16].

Tomato (Solanum lycopersicum) is one of the most widely cultivated vegetable crops globally [7]. During its traditional or greenhouse cultivation, harvesting, and industrial processing, most of a tomato plant remains unused, producing a large quantity of waste consisting of peels, seeds, stems, and leaves that reached 473,989 tons in 2017 [17]. Greenhouse cultivation alone produces 15 tons/ha/year of tomato plant waste [18]. Tomato stems and leaves contain various bioactive substances, such as phenols, flavonoids, and glycoalkaloids, such as tomatine and tomatidine, which have possible antimicrobial, antiviral, and antifungal effects [9]. Traditionally, these by-products are returned to the global cycles of goods by composting or burning [19, 20] or discarded to landfill.

Managing these by-products has become a global environmental and economic issue. However, tomato waste is also one of the most underutilized sources of renewable energy [21]. Nowadays, most studies focus on utilizing tomato crops, seeds, and peels for biogas production or carotenoid extraction. Many studies investigated the utilization of various kinds of tomato waste, such as those from groceries, processing plants, tomato sauce, and tomato puree [2226]. There are only few reports on the bioenergetic utilization of stems and leaves from tomato wastes [7, 2629]. A previous study evaluated the BMP potential of two different tomato plant residue, mostly stems and leaves; however, the stems and leaves’ effects on continuous fermentation were not investigated [8].

This study compares the BMPs of tomato stems and leaves–potentially useful substrates–in batch and continuous AD. We applied tomato waste as mono- and co-substrate in both types of fermentation and followed the process parameters with standard analytical techniques. The metagenomic approach was used to monitor the variations in the microbial communities caused by the substrates. This study is the first attempt to carry out a complex investigation of tomato waste (stems and leaves), as a mono- or co-substrate, in batch and continuous AD systems.

Materials and methods

Substrate specification

Tomato (S. lycopersicum Mill. L. cv. Alfred) plant waste (TW) was obtained from the Department of Plant Biology at University of Szeged. Dry corn stover (CS), provided by the Faculty of Agriculture, University of Szeged, was used as the positive control in both batch and continuous AD experiments.

Both, tomato stems and leaves and corn stover came from field experiments and were recovered after the harvest. The air-dried plant materials were milled and sieved with an electric grinder (Retsch SM 100, Haan, Germany). The maximum particle size was 2 mm. In the co-fermentation experiments, CS and TW were combined at a 7:3 ratio (CoS) based on volatile solid and C/N measurements.

Total solid, volatile solid, and C/N ratio measurements

To determine the dry matter content (total solid, TS), the plant materials were kept at 105°C until their weight became constant. The volatile solid (VS) content was determined by placing the dried residues in an incinerator at 550°C for 2 hours. An Elementar Vario MAX CN (Elementar Group, Hanau, Germany) analyzer was used to determine the C/N ratio of the substrates. The temperature of the combustion and post combustion tube was set to 900°C, the temperature of the reduction tube was set to 830°C. After the samples were burnt in the combustion tube the water vapour was separated by a specific adsorption column containing Sicapent (Merck KGaA, Darmstadt, Germany). The components were detected with a thermal conductivity detector. Helium (5.0) was the carrier and flushing gas (Messer Group, Bad Soden, Germany).

Fiber analysis

FIWE 3 Fiber Analyzer (VELP Scientifica, Usmate, Italy) was used to determine the fiber composition of the applied substrates [30].

Inoculum

The AD inoculum was collected freshly from a biogas plant (Zöldforrás Ltd., Szeged, Hungary), which operated with a mixture of pig slurry and maize silage under mesophilic conditions. Prior to utilization, the sludge was filtered through a 2 mm filter to remove particles larger than 2 mm and undigested plant materials.

Batch fermentation

The biomethane potential (BMP) tests of the CS (positive control), TW, and CoS were performed in batch fermentation in 125 mL Wheaton glass serum bottles (Merck KGaA, Darmstadt, Germany). Based on the VS contents, the substrate to the inoculum ratio was set to 2:1 (VDI I.) and 1:1 (VDI II.). Blank samples containing only water and inoculum were prepared to measure the background methane formation. The bottles were sealed with rubber septa and the airspaces were flushed with nitrogen gas (Purity: 5.0, Messer group, Bad Soden, Germany) for 10 minutes. These experiments were performed in triplicates following the VDI 4630 standard. The fermentations were carried out under mesophilic condition (37.0°C ± 0.5°C) in the final volume of 40 mL. The reactor’s initial pH was 7.5 ± 0.15. Methane content in the headspace of the bottles was measured daily for 30 days. 100μl samples were taken with a Gastight Hamilton syringe and analyzed by gas chromatography (GC) (Section Gas chromatography analysis). Each sample was stirred manually before gas analysis. All fermentations were performed in triplicates.

Continuous fermentation

Both tomato waste as a mono-substrate (TW) and co-substrate (CoS) were also tested in continuous fermentation. The fermentations were carried out in 5 L, continuously stirred tank reactors (CSTRs) [31]. The reactors were filled with the inoculum and then operated under a mesophilic (37°C) condition until the residual biogas production dropped to zero. We subsequently started to feed the reactors daily (OLR: 1 g VS/L; particle size: <2mm). The volume and methane content of the biogas were measured every day; the various fermentation parameters, such as pH, electroconductivity, Volatile Organic Acid/Total Inorganic Carbon (VOA/TIC), ammonium-ion cc., were recorded weekly. The volume of the produced biogas was determined by Brooks gas flow meters (5860S/BA1KA0BA0BA1B1, Brooks Instrument, Hatfield, USA) directly connected with the reactors. The methane content of the generated biogas was measured (Section Gas chromatography analysis).

Gas chromatography analysis

The CH4 content was measured daily by an Agilent 6890N gas chromatograph (Agilent Technologies, Santa Clara, United States). The GC was equipped with a HP Molesive 5 Å column (length 30 m, I.D. 0.53 mega bore, film 25 μm) and a thermal conductivity detector. The flow-rate of Argon 5.0 (Linde Group Hungary, Budapest, Hungary) carrier gas was set to 16.8 mL/min. Split injection mode was applied at 0.2:1 and 150°C. During batch fermentation, the headspaces of the bottles were flushed with nitrogen gas (Messer Group, Bad Soden, Germany) for 10 minutes after each sampling and measurement in the case of batch fermentation. In continuous AD systems, gas samples were taken from the reactor’s headspace in every day into a 15 mL gas tight serum bottle and the methane measurements were carried out as we described above. The methane calibration was performed with certified gas mixture ((5% CO2; 5% CH4; 5% H2; 85% N2) Linde plc., Dublin, Ireland)). For the sampling and injection, Hamilton sample lock syringe was used (Merck KGaA, Darmstadt, Germany).

Fermentation parameters

The pH of the samples was measured with a Radelkis OP-211/2 digital pH meter (RADELKIS Kft, Budapest, Hungary). The volatile organic acid and total acid capacity (VOA/TIC) of the samples were determined using a Pronova FOS/TAC 2000 instrument.

High throughput DNA sequencing

A 150 mg sample from each reactor was collected for total community gDNA purification at the start of fermentation (inoculum) and on the 40th day to determine the digester’s microbial compositions. The samples were stored at −20°C for later use. The extractions were performed with a Quick-DNA Fecal/Soil Microbe Kit (Zymo Research Corporation, Irvine, USA). The DNA was first quantified with a Qubit 4.0 fluorimeter (Invitrogen, Waltham, USA), and then, its integrity was examined on 1% agarose gel.

The 16S Metagenomic Sequencing Libraries were prepared according to the Illumina’s protocol. The V3-V4 region of 16S rDNA was amplified by PCR (Illumina, San Diego, USA), with the Illumina_16S_341F and Illumina_16S_805R primer pair [32]. A 25-μL PCR reaction mixture contained 12.5 ng of genomic DNA, 2x KAPA HiFi HotStart Ready Mix, and 0.2 μM of each primer. The PCR’s parameters include initialization at 95°C for 3 min, and 25 cycles of denaturation, annealing, and extension at 95°C, 65°C and 72°C (each for 30 sec), and final elongation at 72°C for 5 minutes. The PCR products were purified, analyzed, indexed, then the libraries were validated and sequenced (Illumina MiSeq platform, MiSeq® Reagent Kit v3 (600 cycles), by Seqomics Ltd., Mórahalom, Hungary). The sequencing data is available on the NCBI Sequence Read Archive (Submission number: SAMN13231577).

Bioinformatics methods for metagenomic analysis

The trimming and quality filtering of the sequencing reads was carried out using the DADA2’s [33] filterAndTrim function (parameters: truncLen = c(240,220), maxN = 0, maxEE = c(2,2), truncQ = 2, rm.phix = T, ctrimLeft = c(50, 55)) in the R environment (R version 3.5.3, Great Truth) [34]. The trimmed and quality-filtered reads were uploaded to the One Codex webserver and classified by the Targeted Loci Database [35]. The phylum and genus-level results were downloaded in CSV format and imported into the R package. The Principal Component Analysis was carried out with FactoMineR [36] and visualized with factoextra [37]. The phyloseq package was used for richness estimation and visualization. The statistical tests were carried out with the DESeq2 (test = "Wald", fitType = "parametric") [38, 39]. The differences were considered significant by adjusting a p-value threshold to 0.01. The results were visualized with ggpubr [40].

Tomatine and tomatidine inhibition test in batch mode

The effects of tomatine and tomatidine were tested in batch experiments using α-cellulose as a substrate. The fermentations were conducted according to the VDI 4630 standards (Sections Inoculum, Batch fermentation and Continuous fermentation; https://www.vdi.de/en/home). 0.1, 1, 10, 100 μg tomatine or tomatidine were dissolved in DMSO and put into the reactors. Positive and solvent controls were used as well. Positive control contained inoculum, water and α-cellulose; the solvent control contained inoculum, water, α-cellulose and DMSO without tomatine or tomatidine. The methane production was measured daily for 16 days, till the methane formation dropped to zero.

Phytotoxicity test

Lettuce is a good model plant for monitoring phytotoxic compounds [41]. The germination/elongation tests of lettuce seeds (Lactuca sativa L.) were performed to estimate the fermentation effluents’ phytotoxicity. Germination inhibition and root elongation inhibition were conducted on a thin layer of filter paper placed on a Petri dish (90 *× 20 mm). Each plate was supplemented with 1, 3, or 5 v/v% fermentation sludge solution in distilled water; an identical volume of distilled water acted as a control. 30 seeds were placed on each plate. The Petri dishes were incubated at 25°C with lids in the dark. The appearance of radicles indicated germination. Measurements were performed after 0, 1, 2 and 3 days of sludge treatments. The number of germinated seeds and the length of the radicles were determined by ImageJ software [42].

Statistical analysis

The data were analyzed by GraphPad Prism 8 Software, and Dunnett’s multiple comparison test was applied in every case.

Results and discussion

Batch fermentation of corn stover, tomato waste, and co-substrate

The BMP of tomato stem and leave residues were determined in batch fermentation experiments. The tomato wastes were used as mono-substrate (TW) and co-substrate (CoS) with CS. Based on CS and TW’s VS-, carbon- and nitrogen contents, CS and TW were mixed into CoS at a 0.7:0.3 ratio. Before fermentation, the VS, TS, carbon content, nitrogen content, C/N ratio, and fiber composition of the substrates was determined (Tables 1 and 2).

The TS contents of TW, CS, and CoS are very similar, while the VS content of TW is 8% lower than that of CS; however, the CS has a C/N ratio at least two times higher than those of TWs (Table 1). The optimal C/N ratio for an AD process ranges from 20 to 30 [43, 44]. The C/N ratios of CS, TW and CoS fall into the optimal C/N range. The main fiber components of the CS, TW, and CoS, i.e., solubles, hemicellulose, cellulose, and lignin, were also determined (Table 2); CS had a higher content of hemicellulose, cellulose, and lignin than TW.

On the other hand, TW’s soluble content was approximately two times higher than that of CS. The lignin content is important to a substrate’s digestibility. A cross-linked, heterogeneous biopolymer, lignin is hard to degrade; it prevents microorganisms and their enzymes from accessing the cellulose and hemicellulose [45]. Therefore, TW’s lower lignin and higher soluble content make it more accessible to microbial utilization.

The cumulative methane yields from CS, TW, and CoS (both VDI I. and VDI II. systems) were statistically indistinguishable, they varied between 245±18 and 289±15 mL/gVS (Fig 1). Tomato crop residue (stems, leaves, vines and residual fruit) was previously tested in batch experiments by Li et al. where the biogas yield was 214.8 mL/gVS with 57.9% methane content (124.4 mL methane/gVS [46], which was substantially lower than our yields.

thumbnail
Fig 1. The cumulative biogas yield of the batch experiments.

The substrates used were corn stover (CS), tomato waste (TW), and co-substrate (CoS, a mixture of CS and TW) with simple (VDI I.) and double (VDI II.) substrate loads. There are no significant differences between the cumulative biogas yield of substrates according to the Dunnett’s multiple comparison test (n = 3, p ≤ 0.05).

https://doi.org/10.1371/journal.pone.0248654.g001

The experiments suggest that both TW and CoS are promising substrates for AD.

We also noted that TW’s methane production rate was higher than CS in the first 5 days, which might be attributed to TW’s higher soluble content and lower lignin concentration (S1 Fig).

Anaerobic digestions of TW, CS, and CoS in a continuous system

TW’s persistent effect on fermentation was tested in continuous AD experiments using mono- (TW) and co-substrate (CoS). In these experiments, CS was also used as a positive control. First, the CSTR were fed with CS, TW, and CoS for three weeks to adapt their microbiota to the substrates. After the adaptation phase, the biogas production and methane contents were measured daily for 56 days.

The daily biogas yields of the continuous AD were 860 ± 80, 290 ± 50, and 570 ± 70mL/(gVS * d) with CS (CSTRCS), TW (CSTRTW), and CoS (CSTRCoS), respectively (Fig 2); the methane concentrations were comparable at 46.01 ± 3.56 (CSTRCS), 46.91 ± 3.77 (CSTRTW), and 48.10 ± 3.96% (CSTRCoS). The volume of the daily produced biogas drastically dropped in the presence of TW, by 66.28%, in the case of TW/CS and by 33.72% in the case of CoS/CS. Throughout the CSTR experiments, the fermentation parameters, i.e., pH, VOA/TIC, conductivity and NH4+ concentrations, remained within the optimal ranges. The pH varied between 7.20 and 7.92. The VOA/TIC value was between 0.1–0.3, and the NH4+ concentrations were less than 1.2 g/L.

thumbnail
Fig 2. The daily biogas production of continuous experiments.

The substrates used were corn stover (CS), tomato waste (TW), and co-substrate (CoS, a mixture of CS and TW). According to the statistical analysis, there was a significant difference between the CS and TW, as well as CS and CoS. n = 56, p ≤ 0.0001 (Dunnett’s multiple comparison test).

https://doi.org/10.1371/journal.pone.0248654.g002

The CSTR results diverge from our batch fermentation experiments and a previous study on BMP [8]. Since both approaches had the other fermentation parameters in the optimal range, their difference might be due to the presence of TW component that reduced biogas production in the CSTR system. A similar phenomenon was also observed by Li and colleagues in solid state, hemi-solid state and liquid AD systems using tomato residues (stalks, leaves and other wastes) [47]. According to their conclusions, the higher loading of tomato residues might cause stress leading to lower methane production. However, TW, including stems and leaves, contains antimicrobial compounds, such as flavonoids, phenols, vitamins, tomatine, and tomatidine [911]. The continuous input of these substances might influence the composition of the AD microbiota. Therefore, we carried out a 16S metagenomic analysis of CSTRs (Section Microbial composition in the continuous fermenters) and tested the effect of tomatine and tomatidine in batch fermentation experiments (Section Tomatine and tomatidine toxicity).

Microbial composition in the continuous fermenters

Bacterial community.

The microbial communities in continuous fermentation were analyzed on day 0 (inoculum) and 40 of the experiment by 16S amplicon metagenomic analyses. The statistics of the crude and processed sequencing reads are shown in Table 3.

The Shannon and Simpson indices were calculated to characterize and compare the microbial diversities (alpha diversity) in the various fermentations (S2 Fig). CSTRTW’s Shannon and Simpson indices were much lower than those of CSTRCS and CSTRCoS, suggesting that the microbial communities in CSTRCS or CSTRCoS fermentations exhibited substantially higher diversities than CSTRTW. Also, CSTRCoS’ diversity indices were more elevated than CSTRTW’s, indicating an increased in diversity when CSTRs were fed with both substrates. In addition, these results are consistent with the finding that TW contains antimicrobial compounds [9, 11].

To further explore TW’s effect on the microbial compositions, the bacterial and archaeal communities were analyzed with Principal Component Analysis (PCA) (Fig 3). The two axes explained 60% of the total variance; the first axis explained 38.2%, and the second axis 21.8%. The CSTRs fed with TW, CS, and CoS formed diverse and discrete clusters. The parallel experiments were grouped together. As expected, the CoS samples were closer to the control (CS) than to the TW samples. The PCA results confirmed the influence of TW on the microbial community composition in AD.

thumbnail
Fig 3. The Principal Component Analysis of CSTRs fed with CS (corn stover), TW (tomato waste), or CoS (co-substrate) at the genus level.

The bigger, colored circles showed the average of technical parallels (1,2,3).

https://doi.org/10.1371/journal.pone.0248654.g003

First, the abundances of the bacterial and archaeal domains present in the various fermentations were compared to each other on the 40th day. The microbial community of the inoculum was also investigated and presented as a reference. The bacterial domain predominated in all three continuous digestions, reaching 94.54% in abundance in CSTRTW, 94.97% in CSTRCS, and 95.72% in CSTRCoS. The archaeal domain followed with 5.46% in CSTRTW, 5.03% in CSTRCS, and 4.28% in CSTRCoS.

The most abundant bacterial phyla were Firmicutes, Bacteroidetes, Proteobacteria, Synergistetes, Spirochaetes, Actinobacteria, Chloroflexi, and Fibrobacteres; the read count of each phylum was higher than 0.7%. These phyla are found to be abundant in AD systems or biogas plants [48]. Firmicutes and Bacteroidetes were the most consistently dominant phyla in all samples. However, the relative read count of each sample was distinct. The Firmicutes accounted for 52.7%, 61.3%, and 48.3% of total reads in the CSTRCS, CSTRTW, and CSTRCoS fermentations, respectively. The second most abundant bacterial phylum, the Bacteroidetes showed 26.7% (CSTRCS), 20.0% (CSTRTW) and 34.3% (CSTRCoS) appearance in the reactors.

The relative read counts of Synergistetes showed an increasing trend when fermenters were fed with TW or CoS; the abundances of this phylum increased by 5.14 and 1.5 times in CSTRTW (8.9%) and CSTRCoS (2.6%), relative to that of CSTRCS (1.7%), respectively. Since several members of Synergistetes are known to ferment polypeptides and organic acids into acetate, hydrogen and carbon-dioxide, as well as to form syntrophic metabolism with hydrogenotrophic methanogens, an increase in their abundance in CSTRTW was presumably due to the higher nitrogen content of TW (Table 1) [49, 50].

TW had the strongest effect on the microbial community composition in certain genera, such as Clostridium (of phylum Firmicutes), Anaerocella (Bacteroidetes), Acetomicrobium (Synergistetes), Herbinix (Firmicutes), and Gracilibacter (Firmicutes). The relative read counts of these genera were 1.88 (Clostridium), 2.80 (Anaerocella), 14.00 (Acetomicrobium), 6.00 (Herbinix), and 3.86 (Gracilibacter) times higher in the CSTRTW than in the CSTRCS fermentations (Fig 4A). The members of the Clostridium genus participate in the steps of hydrolysis, acidogenesis, and acetogenesis in AD [51, 52]. The members of the Anaerocella genus prefer carbohydrates, have a role in protein degradation and produce hydrogen and fatty acids [52]. The bacteria of the Acetomicrobium genus were formerly reported to ferment amino acids and saccharides to produce acetic acid and hydrogen [53]. Very little is known about the Herbinix and Gracilibacter bacteria; they were found to utilize saccharide monomers and produce VFAs and ethanol [16, 54].

thumbnail
Fig 4.

The relative read counts (% of classified and selected reads) of the bacterial (A) and archaeal domain (B) at the genus level. The top 11 phyla and genera belonging to the bacteria domain are presented. Bacterial genera with a relative read-counts lower than 3.0% were classified as others. All classified genera of the archaea domain are represented.

https://doi.org/10.1371/journal.pone.0248654.g004

Moreover, TW negatively affected the abundance of Sedimentibacter (Firmicutes), Bacteroides (Bacteroidetes), Treponema (Spirochaetes), and caused complete disappearance of Ruminoclostridium (Firmicutes) genus in the CSTRTW fermentations (Fig 4A). Less pronounced, but similar tendencies could be observed in the composition of the microbial community in the CoS fermentations. The bacteria of the Bacteroides genus were previously described as plant cell wall degraders [55]. The bacteria of the genus Ruminoclostridium has a well-known role in cellulose degradation, and the Treponema species were found to degrade plant polysaccharides [56, 57]. The genus Sedimentibacter was also observed to be a core genus in household biogas digesters [58]. Taken together, the disappearance of cellulose- and hemicellulose-degrading bacteria from the CSTRTW might have led to a decrease in biogas production.

Archaeal community.

Only two archaeal phyla, Euryarchaeota and Thaumarchaeota, were detected. Euryarchaeota was the predominant phylum in all reactors. In contrast to the TW fermentations containing 0.7% of Thaumarchaeota, the occurrence of this phylum in CS and CoS fermentations was 3.2% and 4.2%, respectively. The Thaumarchaeota, a prevalent group in Archaea, is present in many ecosystems, including soil, biogas plants, and marine and fresh water [59, 60] and play important roles in both nitrogen- and carbon-cycles [61].

The distribution of archaeal sequences at the genus level shows that Methanothrix, Methanosarcina, and Methanoculleus were the predominant archaeal genera in both CSTRCS and CSTRCoS, Methanosarcina and Methanoculleus were the most abundant taxa in the CSTRTW fermentation, and Methanothrix was barely detectable (Fig 4B). At the same time, the relative read counts of Methanosarcina and Methanoculleus increased by 71.0% and 10.7%, respectively, as compared to that of CSTRCS (Fig 4B). Methanothrix species are acetoclastic methanogens that utilize acetic acid to produce methane [62]; therefore, their disappearance is surprising and may be due to their inability to utilize hydrogen and carbon-dioxide as substrates for methanogenesis [63]. The genus Methanosarcina is able to utilize acetate, CO2, and H2 for methane production [48]. The genus Methanoculleus is hydrogenotrophic methanogen utilizing CO2 and H2 for methane production [6466]. According to a previous study [67], methanogens undertake the hydrogenotrophic pathway under stressful conditions. Moreover, the high soluble organic content of the applied substrate could promote the hydrogen-utilizing methanogenes in the reactors [67]. From these data, it might be concluded, that—in TW fermentations—the biodiversity and the number of the dominant genera in the archaeal domain has decreased.

Tomatine and tomatidine toxicity

TW has been found to contain bioactive compounds such as glycoalkaloids, i.e., tomatine and tomatidine, and phenols. Tomatine and tomatidine have been reported to have antimicrobial and antifungal effects under various conditions [911]; therefore, the effect of tomatine and tomatidine on batch fermentation was tested. Tomatidine, had no effect on the fermentation (The methane yields at 0.0000, 0.0025, 0.0250, 0.2500, 2.5000 μg/mL tomatidine concentrations were 523.7±2.8, 522.9±8.5, 524.0±6.2, 525.2±6.5, 523.7±7.0. In contrast, tomatine could inhibit the methane formation at 0.0250, 0.2500, 2.5000 μg/mL concentrations (Fig 5).

thumbnail
Fig 5. The effect of tomatine concentration on methane production in a batch experiment on day 16.

The methane production of the reactors containg 2.5000, 0.2500, 0.0250, or 0.0025 μg/mL tomatine are shown to have significant differences, according to Dunnett’s multiple comparison test (n = 3, p ≤ 0.05).

https://doi.org/10.1371/journal.pone.0248654.g005

Tomatine’s negative effect was relatively stronger at the beginning of the fermentation; then, the effect gradually decreased until the end (S3 Fig). The decrease of the negative effect could be due to the microbial adaptation to tomatine (the time is likely too short) or tomatine’s decomposition during fermentation. Further analysis is necessary to address these possibilities.

Phytotoxicity test

The evaluation of sewage sludge’s toxicity is important for assessing the applicability of the biogas plants’ effluents for agricultural purposes. Nowadays, phytotoxicity tests have attracted more attention and gained wider acceptance [41]. Seed germination and radicle length monitoring are the most common phytotoxicity tests. Hence, we determined germination indexes (GIs) of Lactuca sativa L. seeds and measured the radicle lengths (RLs) to assess the phytotoxicity of the various effluents from CS, TW, and CoS fermentations. Samples were taken from the CW-AD fermenters at the starting point, the midpoint, and the endpoint of the continuous fermentation experiments. GIs and RLs of lettuce were measured on the first, second, and third day. The effluent of CS, TW, and CoS fermentations, at 5%v/v effluent concentration, had no effect on the germination of lettuce seeds on the third day. On the other hand, the effluents applied at 5v/v% had minor negative effects on RL, while 3 v/v% effluents slightly increased the RL in all samples. Therefore, at 3 v/v%, none of the CW-AD effluents were phytotoxic on lettuce, and these effluents could be used in agriculture as a fertilizer.

Conclusions

Tomato waste (TW), i.e., tomato stems and leaves, seems to be a promising substrate for biogas production since it has relatively low lignin and hemicellulose content and a lower C/N ratio than CS. Batch fermentation experiments confirmed that TW could be a promising substrate for biogas production both as mono-substrate or co-substrate with CS. However, in continuous AD systems, the involvement of TW decreased the biogas production. The fermentation parameters of the continuous AD experiments, such as pH, volatile fatty acid content, and NH4+ concentration, were optimal in all cases; thus, TW might have an inhibitory effect on the microbial community. The negative effect of tomatine on the biogas yield was confirmed in batch fermentation experiments, therefore TW should be applied as co-substrate in real applications.

The metagenomic analyses of the microbial composition in the fermentations revealed significant rearrangements, and reduced biodiversity in the microbial community in TW’s presence.

Agricultural usage of the fermentation effluent requires preliminary tests, including phytotoxicity assays. It has been demonstrated that the effluent of TW-containing fermentations had no toxic effect on L. sativa var. capitata germination.

The results demonstrated that tomato waste could be a good mono-substrate in batch fermentations or a co-substrate with corn stover in proper ratio in continuous anaerobic fermentations for biogas production. However, further optimization, including pretreatments or the application of thermophilic conditions, might further improve the process.

These results also point to the importance of running long-term continuous fermentations to test the suitability of a novel biomass substrate for industrial biogas production.

Supporting information

S1 Fig.

A, Kinetics of the produced methane from corn stover (CS) and tomato waste (TW) in batch experiment. B, Cumulative methane production of CS and TW. There are not significant differences between the used substrates according to the Dunnett’s multiple comparison test (n = 3, p≤0.05).

https://doi.org/10.1371/journal.pone.0248654.s001

(TIF)

S2 Fig. α-diversity analysis for the fed-batch reactors fed with Corn Stover (CS), Tomato Waste (TW) and Co-Substrate (CoS).

3–3 technical parallel/reactor.

https://doi.org/10.1371/journal.pone.0248654.s002

(TIF)

S3 Fig. Effect of tomatine on the daily cumulative biomethane production.

The tested concentrations were the following: 2.5000, 0.2500, 0.0250 and 0.0025 μg/mL.

https://doi.org/10.1371/journal.pone.0248654.s003

(TIF)

Acknowledgments

The authors would like to say thank you for the technical support to Sarolta Papp.

References

  1. 1. Buettner C, von Bergen M, Jehmlich N, Noll M. Pseudomonas spp. are key players in agricultural biogas substrate degradation. Sci Rep. 2019;9: 12871. pmid:31492882
  2. 2. Anna Schnürer; Asa Jarvis. Microbiological handbook for biogas plants. 2009.
  3. 3. Deublein D, Steinhauser A. Substrate. In: Deublein D, Steinhauser A, editors. Biogas from waste and renewable resources. Weinheim, Germany: Wiley-VCH Verlag GmbH & Co. KGaA; 2010. pp. 55–84. https://doi.org/10.1002/9783527632794.ch7
  4. 4. Achinas S, Euverink GJW. Theoretical analysis of biogas potential prediction from agricultural waste. Resour Technol. 2016;2: 143–147.
  5. 5. Food and Agriculture Organization of the United Nations (FAO). FAOSTAT: Agricultural land–Area by country, 2018. http://www.fao.org/faostat/en/?fbclid=IwAR0DEhWqvAHl32VfErXOdT203cbM4wdc8Nfw5xCeq6Qt0zZA32Cp0RY-GNE#data/RL/visualize. accessed September 2019.
  6. 6. Schieber A, Stintzing FC, Carle R. By-products of plant food processing as a source of functional compounds—Recent developments. Trends Food Sci Technol. 2001;12: 401–413.
  7. 7. Li Y, Xu F, Li Y, Lu J, Li S, Shah A, et al. Reactor performance and energy analysis of solid state anaerobic co-digestion of dairy manure with corn stover and tomato residues. Waste Manag. 2018;73: 130–139. pmid:29223482
  8. 8. Martínez M, Driller L, Chamy R, Schuch W, Nogueira M, Fraser P. Biogas potential of residues generated by the tomato processing industry under different substrate and inoculum conditions. Acta Hortic. 2017; 151–158.
  9. 9. Silva-Beltrán NP, Ruiz-Cruz S, Cira-Chávez LA, Estrada-Alvarado MI, Ornelas-Paz JDJ, López-Mata MA, et al. Total phenolic, flavonoid, tomatine, and tomatidine contents and antioxidant and antimicrobial activities of extracts of tomato plant. Int J Anal Chem. 2015;2015. pmid:26609308
  10. 10. Mitchell G, Gattuso M, Grondin G, Marsault É, Bouarab K, Malouin F. Tomatidine inhibits replication of Staphylococcus aureus small-colony variants in cystic fibrosis airway epithelial cells. Antimicrob Agents Chemother. 2011;55: 1937–1945. pmid:21357296
  11. 11. Bednarek P, Osbourn A. Plant-microbe interactions: Chemical diversity in plant defense. Science. 2009. pp. 746–748. pmid:19423814
  12. 12. Teodorita Al Seadi. Biogas Handbook. Teodorita Al Seadi, editor. University of Southern Denmark Esbjerg; 2008.
  13. 13. Holliger C, Alves M, Andrade D, Angelidaki I, Astals S, Baier U, et al. Towards a standardization of biomethane potential tests. Water Sci Technol. 2016;74: 2515–2522. pmid:27973356
  14. 14. Kakuk B, Kovács KL, Szuhaj M, Rákhely G, Bagi Z. Adaptation of continuous biogas reactors operating under wet fermentation conditions to dry conditions with corn stover as substrate. Anaerobe. 2017;46: 78–85. pmid:28576713
  15. 15. Kampmann K, Ratering S, Kramer I, Schmidt M, Zerr W, Schnell S. Unexpected stability of Bacteroidetes and Firmicutes communities in laboratory biogas reactors fed with different defined substrates. Appl Environ Microbiol. 2012;78: 2106–19. pmid:22247168
  16. 16. Rettenmaier R, Duerr C, Neuhaus K, Liebl W, Zverlov V V. Comparison of sampling techniques and different media for the enrichment and isolation of cellulolytic organisms from biogas fermenters. Syst Appl Microbiol. 2019;42: 481–487. pmid:31153679
  17. 17. Torres-León C, Ramírez-Guzman N, Londoño-Hernandez L, Martinez-Medina GA, Díaz-Herrera R, Navarro-Macias V, et al. Food waste and byproducts: an opportunity to minimize malnutrition and hunger in developing countries. Front Sustain Food Syst. 2018;2: 52.
  18. 18. Pane C, Celano G, Piccolo A, Villecco D, Spaccini R, Palese AM, et al. Effects of on-farm composted tomato residues on soil biological activity and yields in a tomato cropping system. Chem Biol Technol Agric. 2015;2: 1–13.
  19. 19. Kulcu R. Compost of greenhouse tomato plant residues, wheat straw, separated dairy manure, and the effect of free air space process. Pol. J. Environ. Stud. 2014;23: 1341–1346.
  20. 20. Himmel ME, Ding S-Y, Johnson DK, Adney WS, Nimlos MR, Brady JW, et al. Biomass recalcitrance: engineering plants and enzymes for biofuels production. Science. 2007;315: 804–7. pmid:17289988
  21. 21. Kassab Z, Kassem I, Hannache H, Bouhfid R, Qaiss AEK, El Achaby M. Tomato plant residue as new renewable source for cellulose production: extraction of cellulose nanocrystals with different surface functionalities. Cellulose. 2020;27: 4287–4303.
  22. 22. Saghouri M, Mansoori Y, Rohani A, Khodaparast MHH, Sheikhdavoodi MJ. Modelling and evaluation of anaerobic digestion process of tomato processing wastes for biogas generation. J Mater Cycles Waste Manag. 2018;20: 561–567.
  23. 23. Nugroho A, Sumantri I. Biogas production of tomato sauce wastewater by batch anaerobic digestion. AIP Conference Proceedings. American Institute of Physics Inc.; 2020. p. 030010. https://doi.org/10.1063/1.5140902
  24. 24. Budiyono, Manthia F, Amalin N, Hawali Abdul Matin H, Sumardiono S. Production of biogas from organic fruit waste in anaerobic digester using ruminant as the noculum. MATEC Web of Conferences. 2018. https://doi.org/10.1051/matecconf/201815603053
  25. 25. Ulusoy Y, Ulukardeler AH, Ünal H, Aliba K. Analysis of biogas production in Turkey utilising three different materials and two scenarios. African J Agric Res. 2009. Available: http://www.academicjournals.org/AJAR
  26. 26. Deressa L, Libsu S, Chavan RB, Manaye D, Dabassa A. Production of biogas from fruit and vegetable wastes mixed with different wastes. Environ Ecol Res. 2015;3: 65–71.
  27. 27. Luengo E, Álvarez I, Raso J. Improving carotenoid extraction from tomato waste by pulsed electric fields. Front Nutr. 2014;1: 12. pmid:25988115
  28. 28. Atem AD, Indiveri ME, Llamas S. Biomass storage for further energy use through biogas production. Int J Hydrogen Energy. 2010;35: 6048–6051.
  29. 29. Saev M, Koumanova B, Simeonov Iv. Anaerobic co-digestion of wasted tomatoes and cattle dung for biogas. J Univ Chem Technol Metall. 2009;44: 55–60.
  30. 30. Goering HK, Van Soest PJ. Forage fiber analyses (apparatus, reagents, procedures, and some applications). USDA Agr Handb. 1970
  31. 31. Kovács KL, Ács N, Kovács E, Wirth R, Rákhely G, Strang O, et al. Improvement of biogas production by bioaugmentation. Biomed Res Int. 2013;2013. pmid:23484123
  32. 32. Zheng W, Tsompana M, Ruscitto A, Sharma A, Genco R, Sun Y, et al. An accurate and efficient experimental approach for characterization of the complex oral microbiota. Microbiome. 2015;3: 48. pmid:26437933
  33. 33. Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJA, Holmes SP. DADA2: High-resolution sample inference from Illumina amplicon data. Nat Methods. 2016;13: 581–583. pmid:27214047
  34. 34. R Core Team. R: A Language and environment for statisticalcomputing. Vienna, Austria; 2019.
  35. 35. Minot SS, Krumm N, Greenfield NB. One Codex: A sensitive and accurate data platform for genomic microbial Identification. bioRxiv. 2015; 027607.
  36. 36. Lê S, Josse J, Husson F. FactoMineR: An R package for multivariate analysis. J Stat Softw. 2008;25: 1–18.
  37. 37. Kassambara A, Mundt F, Kassambara, A.; Mundt F. Factoextra: extract and visualize the results of multivariate data analyses. URL http//www.sthda.com/english/rpkgs/factoextra BugReports. 2017.
  38. 38. McMurdie PJ, Holmes S. phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One. 2013;8: e61217. pmid:23630581
  39. 39. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15. pmid:25516281
  40. 40. Kassambara A. ggpubr: “ggplot2” Based Publication Ready Plots. R package version 0.2. https://CRAN.R-project.org/package=ggpubr. https://CRANR-project.org/package=ggpubr. 2018. R package version 0.1.8
  41. 41. Mañas P, De las Heras J. Phytotoxicity test applied to sewage sludge using Lactuca sativa L. and Lepidium sativum L. seeds. Int J Environ Sci Technol. 2018;15: 273–280.
  42. 42. Schneider CA, Rasband WS, Eliceiri KW. NIH Image to ImageJ: 25 years of image analysis. Nature Methods. 2012. pp. 671–675. pmid:22930834
  43. 43. Hills DJ. Effects of carbon: nitrogen ratio on anaerobic digestion of dairy manure. Agric Wastes. 1979;1: 267–278.
  44. 44. Wang X, Lu X, Li F, Yang G. Effects of temperature and Carbon-Nitrogen (C/N) ratio on the performance of anaerobic co-digestion of dairy manure, chicken manure and rice straw: Focusing on ammonia inhibition. PLoS One. 2014;9. pmid:24817003
  45. 45. Raud M, Tutt M, Olt J, Kikas T. Effect of lignin content of lignocellulosic material on hydrolysis efficiency. Agron Res. 2015; 13(2): 405–412.
  46. 46. Li P, Li W, Sun M, Xu X, Zhang B, Sun Y. Evaluation of biochemical methane potential and kinetics on the anaerobic digestion of vegetable crop residues. Energies. 2018;12: 26.
  47. 47. Li Y, Lu J, Xu F, Li Y, Li D, Wang G, et al. Reactor performance and economic evaluation of anaerobic co-digestion of dairy manure with corn stover and tomato residues under liquid, hemi-solid, and solid state conditions. Bioresour Technol. 2018;270: 103–112. pmid:30212770
  48. 48. Detman A, Mielecki D, Pleśniak Ł, Bucha M, Janiga M, Matyasik I, et al. Methane-yielding microbial communities processing lactate-rich substrates: A piece of the anaerobic digestion puzzle. Biotechnol Biofuels. 2018;11. pmid:29387172
  49. 49. Zamanzadeh M, Hagen LH, Svensson K, Linjordet R, Horn SJ. Biogas production from food waste via co-digestion and digestion- effects on performance and microbial ecology. Sci Rep. 2017;7: 17664. pmid:29247239
  50. 50. Westerholm M, Schnürer A. Microbial responses to different operating practices for biogas production systems. Anaerobic Digestion. IntechOpen; 2019. https://doi.org/10.5772/intechopen.82815
  51. 51. Tapadia-Maheshwari S, Pore S, Engineer A, Shetty D, Dagar SS, Dhakephalkar PK. Illustration of the microbial community selected by optimized process and nutritional parameters resulting in enhanced biomethanation of rice straw without thermo-chemical pretreatment. Bioresour Technol. 2019;289. pmid:30290321
  52. 52. Abe K, Ueki A, Ohtaki Y, Kaku N, Watanabe K, Ueki K. Anaerocella delicata gen. nov., sp. nov., a strictly anaerobic bacterium in the phylum Bacteroidetes isolated from a methanogenic reactor of cattle farms. J Gen Appl Microbiol. 2013;58: 405–412. pmid:23337575
  53. 53. Cook LE, Gang SS, Ihlan A, Maune M, Tanner RS, McInerney MJ, et al. Genome sequence of Acetomicrobium hydrogeniformans OS1. Genome Announc. 2018;6. pmid:29954899
  54. 54. Kim W, Hwang K, Shin SG, Lee S, Hwang S. Effect of high temperature on bacterial community dynamics in anaerobic acidogenesis using mesophilic sludge inoculum. Bioresour Technol. 2010;101. pmid:19362824
  55. 55. Weiß S, Tauber M, Somitsch W, Meincke R, Müller H, Berg G, et al. Enhancement of biogas production by addition of hemicellulolytic bacteria immobilised on activated zeolite. Water Res. 2010;44: 1970–1980. pmid:20022352
  56. 56. Gharechahi J, Zahiri HS, Noghabi KA, Salekdeh GH. In-depth diversity analysis of the bacterial community resident in the camel rumen. Syst Appl Microbiol. 2015;38: 67–76. pmid:25467553
  57. 57. Koeck DE, Wibberg D, Maus I, Winkler A, Albersmeier A, Zverlov V V., et al. Complete genome sequence of the cellulolytic thermophile Ruminoclostridium cellulosi wild-type strain DG5 isolated from a thermophilic biogas plant. J Biotechnol. 2014;188: 136–137. pmid:25173616
  58. 58. Rui J, Li J, Zhang S, Yan X, Wang Y, Li X. The core populations and co-occurrence patterns of prokaryotic communities in household biogas digesters. Biotechnol Biofuels. 2015;8: 158. pmid:26413157
  59. 59. Costa OYA, Souto BM, Tupinambá DD, Bergmann JC, Kyaw CM, Kruger RH, et al. Microbial diversity in sugarcane ethanol production in a Brazilian distillery using a culture-independent method. J Ind Microbiol Biotechnol. 2015;42: 73–84. pmid:25404204
  60. 60. Brochier-Armanet C, Gribaldo S, Forterre P. Spotlight on the Thaumarchaeota. ISME Journal. 2012. pp. 227–230. pmid:22071344
  61. 61. Pester M, Schleper C, Wagner M. The Thaumarchaeota: An emerging view of their phylogeny and ecophysiology. Curr Opin Microbiol. 2011;14: 300–306. pmid:21546306
  62. 62. Liu C, Sun D, Zhao Z, Dang Y, Holmes DE. Methanothrix enhances biogas upgrading in microbial electrolysis cell via direct electron transfer. Bioresour Technol. 2019;291: 121877. pmid:31376672
  63. 63. Patel GB, Sprott GD. Methanosaeta concilii gen. nov., sp. nov. (’Methanothrix concilii’) and Methanosaeta thermoacetophila nom. rev., comb. nov. Int J Syst Bacteriol. 1990;40: 79–82.
  64. 64. Cheng L, Qiu TL, Li X, Wang WD, Deng Y, Yin XB, et al. Isolation and characterization of Methanoculleus receptaculi sp. nov. from Shengli oil field, China. FEMS Microbiol Lett. 2008;285: 65–71. pmid:18557787
  65. 65. Wirth R, Kovács E, Maráti G, Bagi Z, Rákhely G, Kovács KL. Characterization of a biogas-producing microbial community by short-read next generation DNA sequencing. Biotechnol Biofuels. 2012;5. pmid:22316115
  66. 66. Bagi Z, Ács N, Böjti T, Kakuk B, Rákhely G, Strang O, et al. Biomethane: The energy storage, platform chemical and greenhouse gas mitigation target. Anaerobe. 2017;46: 13–22. pmid:28341558
  67. 67. Liu Y, Wachemo AC, Yuan HR, Li XJ. Anaerobic digestion performance and microbial community structure of corn stover in three-stage continuously stirred tank reactors. Bioresour Technol. 2019. pmid:31100566