Bacterial and archaeal community structure involved in biofuels production using hydrothermal- and enzymatic-pretreated sugarcane bagasse for an improvement in hydrogen and methane production

Abstract

Sugarcane bagasse (SCB) was used as a lignocellulosic substrate, combining the co-production of H₂ (Stage I) and CH₄ (Stage II) by a dark fermentation process in batch reactors. Hydrothermally- and enzymatic (Aspergillus niger)-pretreated SCB were applied as substrate sources. Two fermentative inocula (In1 and In2) were used in Stage I and a methanogenic inoculum in Stage II (In3), comprising in total three experimental series in relation to Stage I: A (In1), B (In1 plus In2), and C (In2). The final metabolites (solid, liquid, and gaseous fractions) from Stage I were used for CH₄ production (Stage II). The SCB pretreatment employed was favorable for biogas and organic acids production. Higher H₂ and CH₄ yields were obtained in C (4.3 and 6.3 mmol g⁻¹ SCB, respectively). For all conditions, the H₂ production occurred primarily via an acetic acid route. The predominance of cellulolytic enzyme producers (Enterococcus and Clostridium) may have favored the H₂ and subsequent CH₄ production; this last was produced mainly from members of the Methanoregulaceae and Methanosaetaceae families. Furthermore, homoacetogenic bacteria (Acetobacterium, Clostridium, Eubacterium, Holophaga) were also identified in both stages. The synergistic action of these microbial groups promoted the hydrolysis of SCB as well as hydrogen and methane production.

---

1 Introduction

Agricultural biomasses are important platforms for the production of biofuels from renewable raw materials.¹ Hydrogen (H₂) and methane (CH₄) are important energy sources, which can be produced by the anaerobic digestion of different organic wastes.² Given that the dark fermentation for hydrogen production from organic substrates only occurs partially, the combination with methane production can benefit energy recovery.³ In addition, compared to liquid fuels, such as ethanol, hydrogen and methane can be readily separated from the liquid phase, which can reduce the process costs.²

Two-stage processes have previously already been employed for CH₄ production,⁴ for the separation of the hydrolysis/acidogenesis phase from the methanogenesis phase, and for improving each process independently.⁵ Studies into hydrogen and methane production using two-stage fermentation processes have been carried out using a variety of substrates, such as food waste,³ solid household waste,⁴ and sludge from a wastewater treatment plant.⁶

However, few studies have been carried out on the co-production of H₂ and CH₄ from sugarcane bagasse (SCB). Some of the most studied raw materials are sorghum,⁷ corn cob,⁸ olive pulp,⁹ mushroom,^10,11 and potato waste.¹² SCB is a carbohydrate-rich substrate that has been used for hydrogen production.^13,14 The biotechnology processes used in biomass conversion are more ecologically correct and capable of generating energy as well as the thermochemical and electrochemical processes.¹⁵ In this way, the use of SCB for energy generation through biological processes can reduce the environmental impacts caused by the combustion, thus reducing the CO₂ emission indices.

The design and optimization of hydrogen production from hydrothermally (200 °C for 10 min at 16 bar)-pretreated SCB was evaluated using a response surface methodology and anaerobic consortium from an upflow anaerobic sludge blanket reactor (UASB) treating vinasse.¹⁶ The authors observed a maximum hydrogen production of 17.7 mmol L⁻¹ at 60 °C with 3.0 g L⁻¹ of yeast extract, with a predominance of acetic and butyric acids.

Among the microorganisms involved in the H₂ production were Clostridium genus, like Clostridium butyricum, C. acetobutylicum, C. saccharoperbutylacetonicum, C. pasteurianum, and C. cellulolyticum.^17–21 Hydrogen and organic acids production were evaluated with autochthonous (Clostridium and Paenibacillus from SCB) and allochthonous (Clostridium, Bacillus and Enterobacter) bacteria from SCB (pretreated in an autoclave at 121 °C, 1.5 kg cm⁻² for 15 min) in batch reactors.²² The authors observed a maximum production of 23.1 mmoL H₂ L⁻¹ under optimized conditions of 7.0 g SCB per L and pH 7.2 at 37 °C through the acetic acid (1.57 g L⁻¹) pathway. A higher hydrogen production of 50 mmol L⁻¹ was observed when a co-culture of Clostridium thermocellum and Thermoanaerobacterium aotearoense was employed with 40 g L⁻¹ of alkali-pretreated SCB (3% NaOH, 55 °C for 3 h).²³

Meanwhile, due the complexity of SCB components, it is evident that this lignocellulosic biomass is difficult to be fully utilized by only a single bacterial strain. A microbial community, such as anaerobic sludge, is thus usually used as an inoculum source for hydrogen or methane production.^3,4 However, it is still difficult to control and optimize the hydrogen-producing microbial consortium, as the hydrogen production rate and yield are not always high. It is therefore essential to obtain a stable and efficient microbial community that is able to improve the hydrolysis of biomass and hydrogen production.

SCB is composed of hemicellulose, lignin, and cellulose, as well as other minor components. Hemicellulose and lignin fractions protect cellulose fibers from microbial cellulases,²⁴ which are essential in the conversion of complex carbohydrates to fermentable glucose. However, the conversion of lignocellulosic substrates into H₂ requires many steps, including substrate pretreatment, enzymatic hydrolysis, and fermentation.

In a study of hydrogen production from mushroom compost by Clostridium thermocellum, the authors supplemented the culture medium with recombinant β-glucosidases and obtained an increase in hydrogen production of 37%.¹⁰ In contrast, in the present study, commercial enzymes were not used for the hydrolysis of SCB, which makes the biogas production process more attractive due to its reduced costs since the cellulolytic enzymes were obtained from both, the A. niger and the inoculum.

The purpose of the pretreatment of recalcitrant substrates is to change or remove the lignin and/or hemicellulose structure, decrease the crystallinity of cellulose, and increase the substrate surface area in order to benefit the enzymes accessing the structure of cellulose for the better recovery of fermentable sugars.²⁵

According to da Cruz et al.,²⁶ hydrothermal pretreatment eliminates these barriers by solubilizing hemicellulose and breaking the lignin structure. Furthermore, cellulose is also decrystallized, thereby allowing the cellulase enzymes to access those microfibers. This technology for SCB pretreatment is of particular interest because fibers are heated in water, the only solvent used, which eliminates the cost of a catalyst and reduces the waste generation process.²⁷

Pretreatment employing a hydrothermal system consists of keeping SBC in pressurized liquid hot water and promoting rapid depressurization, after achieving a severity parameter, for the purpose of breaking down the lignocellulosic structure to supply the cellulolytic and fermenting microorganism with access to the cellulose structure.²⁸ Therefore, the hydrothermal pretreatment of SCB could solubilize more than 80% of the hemicellulose and more than 40% of the lignin.²⁹ In a study into the hydrothermal pretreatment of SCB (reaction temperature and time between 160 °C and 200 °C and 5–20 min, respectively) for ethanol production, the authors observed in the solid fraction about 5–22% xylan, 22% Klason lignin, and 44–55% glucan.²⁶

Usually, the process of the enzymatic treatment of lignocellulose consists of two steps: primarily, the lignocellulosic biomass is pretreated to destroy its complex structure; and then, the substrate is converted by the enzymes to fermentable sugars.³⁰

Accordingly, this conversion of the lignocellulosic substrate to glucose requires the presence of cellulolytic enzymes. Since fungal pretreatment is a low cost and environmentally friendly pretreatment method, it has attracted increased importance in recent years.³¹ Furthermore, enzymatic pretreatment is conducted under mild conditions and several inhibitory compounds are not generated.³² Therefore, combining hydrothermal pretreatment with enzymatic pretreatment is a very interesting proposition.

The main disadvantages of biological pretreatment are the substantial loss of carbohydrates that may occur, long residence time, necessity for careful control of the growth conditions, the need to avoid contamination, and the space required, which limit its applications. To overcome these limitations, biological treatments are usually used in combination with other treatments.³³ The combination of fungal treatment with liquid hot water was studied to evaluate the hydrolysis of the lignocellulosic biomass Populus tomentosa.³⁴

Among the microorganisms applied in the biological treatment of lignocellulosic substrates, fungi have gained distinction, especially the filamentous fungi that adapt more easily because they are less demanding and do not require large amounts of available water and as their growth is stimulated by the formation of hyphae. Among these, the genus Aspergillus is considered superior to other genera in relation to the production of cellulolytic and hemicellulolytic complexes.³⁵

Although the association of SCB pretreatments has already been studied,^23,36–38 knowledge of the association of the hydrothermal and enzymatic pretreatment of SCB using A. niger as well as the effect of the inoculum source and phylogenetic knowledge of the community are still lacking for H₂ and CH₄ production in a two-stage anaerobic digestion process from lignocellulosic biomass.

Therefore, the purpose of this research was to study the feasibility of the co-production of H₂ (Stage I) and CH₄ (Stage II) from pretreated sugarcane bagasse (SCB) using a two-stage anaerobic digestion process. The metabolites (organic acids and H₂) generated at the end of Stage I served as substrates in Stage II, for CH₄ production. The pretreatment of SCB was conducted in a hydrothermal system combining enzymatic pretreatment with the fungus Aspergillus niger. The resulting solid fraction from both pretreatments was used as a substrate in the H₂/CH₄ assays. Moreover, the microbial community structure in the H₂ and CH₄ production reactors was assessed, using the Illumina MiSeq platform for the bacteria and archaea domains, as well as the relation with the metabolites produced in both operational stages.

2 Material and methods

2.1 Sugarcane bagasse pretreatment

The SCB used in this study was supplied by São Martinho sugar mill (Pradópolis, SP, Brazil). The SCB composition was 31.4% cellulose, 36.6% xylan, 5.0% arabinan, 24.8% lignin, and 1.2% ash.³⁹ The SCB (5 g) was pretreated using a 100 mL capacity hydrothermal system, 20 bar pressure at 200 °C for 10 min,¹⁴ followed by enzymatic pretreatment with the fungus Aspergillus niger. At the end of both pretreatments, the solid fraction was employed as a substrate in the co-production of H₂ and CH₄.

In this study, A. niger strain (ATCC 10577) acquired from the André Tosello Foundation (Campinas, SP, Brazil) was used. The lyophilized strain was reactivated by adding 0.5 mL sterile distilled water to the tube (30 min); then the sample was inoculated in solid medium (20 g L⁻¹ malt extract, 20 g L⁻¹ glucose, 5 g L⁻¹ peptone, 15 g L⁻¹ agar, and pH 7) and incubated at room temperature for approximately 7 days or until total growth on the plate for subsequent application in the SCB pretreatment.

For the enzymatic pretreatment, the hydrothermally pretreated SCB (solid fraction) was used as a carbon source in Mandels and Weber culture media.⁴⁰ For this, 1 mL L⁻¹ of Tween 80 and 5 g of SCB was added into an Erlenmeyer flask. This substrate was autoclaved at 121 °C for 20 min and then 5 mL of sterile culture medium was added to it. Subsequently, a spore suspension with 1 × 10⁸ spores per mL of A. niger was added to the Erlenmeyer flask and incubated at 37 °C with periodic shaking for 7 days. At the end of this period, the SCB pretreated hydrothermally and enzymatically by the fungus was used as a substrate in the H₂ and CH₄ production assays.

2.2 Anaerobic mixed consortia

The first inoculum (In1) was obtained from an anaerobic lagoon of a pulp and paper mill's wastewater treatment plant (WWTP) in Suzano (São Paulo, Brazil). To obtain the cellulolytic and fermentative consortium, 10% (v/v) of the sludge was inoculated in cellulose anaerobe medium⁴¹ containing 2 g L⁻¹ cellulose for 24 h. The second inoculum (In2) was a fermentative and cellulolytic consortia (Clostridium, Bacillus, Bacteroides, and Paenibacillus) obtained in the Biological Processes Laboratory, University of São Paulo (Brazil), using cellulose as the substrate.⁴² This consortia was inoculated in cellulose anaerobe medium⁴¹ containing 2 g L⁻¹ cellulose for 24 h.

The methanogenic inoculum (In3) was a granular sludge obtained from the upflow anaerobic sludge blanket reactor (UASB) used in the treatment of poultry slaughterhouse wastewater operated in mesophilic conditions. To obtain the methanogenic consortium, 10% (v/v) of the sludge was inoculated in cellulose anaerobe medium⁴¹ containing 2 g L⁻¹ cellulose for 72 h. After the incubation period, the inocula were centrifuged at 6000 rpm, 27 °C for 10 min, the supernatant were discarded and the concentrated biomasses were resuspended in cellulose anaerobe medium,⁴¹ which were the inocula of the H₂ and CH₄ production assays.

2.3 Culture media and biogas production assays

Batch reactors (Duran® flasks of 250 mL) were operated in triplicate. The reactors contained 90 mL of cellulose anaerobe medium,⁴¹ 10 mL of inoculum (10% v/v), and 2 g L⁻¹ of pretreated SCB as the substrate. All the assays were performed in mesophilic conditions (37 °C), pH 6.0, and under stirring at 100 rpm. The total fermentation times were 196 h and 464 h for Stages I and II, respectively.

The cellulose anaerobe medium⁴¹ used for the cultivation of the microbial consortia was supplemented with yeast extract (1 g L⁻¹). This medium contained the following compounds: NaHCO₃ (2.1 g L⁻¹); NH₄Cl (0.68 g L⁻¹); KH₂PO₄ (0.18 g L⁻¹); (NH₄)₂SO₄ (0.15 g L⁻¹); MgSO₄·7H₂O (0.12 g L⁻¹); K₂HPO₄ (296 mg L⁻¹); CaCl·2H₂O (61 mg L⁻¹); FeSO₄·7H₂O (21 mg L⁻¹); nitriloacetic acid (15 mg L⁻¹); NaCl (10 mg L⁻¹); MnSO₄·H₂O (5 mg L⁻¹); CoCl₂·6H₂O (1 mg L⁻¹); ZnSO₄·7H₂O (1 mg L⁻¹); CuSO₄·5H₂O (0.1 mg L⁻¹); KAl(SO₄)₂·12H₂O (0.1 mg L⁻¹); H₃BO₃ (0.1 mg L⁻¹); Na₂MoO₄·2H₂O (0.1 mg L⁻¹); and vitamin solution (5 mL L⁻¹). The vitamin solution was composed of: pyridoxine HCl (10 mg L⁻¹); calcium DL-pantothenate (5 mg L⁻¹); lipoic acid (5 mg L⁻¹); nicotinic acid (5 mg L⁻¹); p-amino benzoic acid (5 mg L⁻¹); riboflavin (5 mg L⁻¹); thiamine HCl (5 mg L⁻¹); biotin (2 mg L⁻¹); folic acid (2 mg L⁻¹); and vitamin B12 (0.1 mg L⁻¹).

The solid fraction of pretreated SCB (2 g L⁻¹) was used as the substrate source in all reactors instead of 2 g L⁻¹ cellulose. Three experimental series were performed: A (10% of In1), B (5% of In1 + 5% of In2), and C (10% of In2) (Table 1). The reactor headspace was purged with nitrogen (N₂ 100%) for 10 min in order to provide an anaerobic environment. The final metabolites (solid, liquid, and gaseous fractions) of Stage I (H₂ production) were used for the subsequent CH₄ production stage (Stage II). Therefore, the reactor of Stage I was inoculated with 10% (v/v) of inoculum 3 (In3 – methanogenic) using a syringe so that there was no loss of biogas, to form the reactor from Stage II. The initial pH of the methanogenic stage was adjusted to pH 7 with NaOH (10% w/v). For this, an aliquot of the liquid fraction at the end of Stage I was withdrawn with a syringe and the final pH was measured and then NaOH (10% w/v) was added to the reactors, also with a syringe.

Table 1 Experimental conditions of batch reactors with SCB according to the inoculum source^a

Experimental series	Stage I	Stage II
a In – inoculum; In1 – anaerobic lagoon of a wastewater treatment plant; In2 – fermentative inoculum; In3 granular ludge from UASB reactor used in the treatment of poultry slaughterhouse wastewater.
A	In1	In3
B	In1 + In2	In3
C	In2	In3

---

2.4 Physicochemical and chromatographic analyses

The accumulated hydrogen and methane production were monitored by gas chromatography (GC) 2010 (Shimadzu, Japan) equipped with a thermal conductivity detector (TCD) and a Carboxen 1010 PLOT column (30 m × 0.53 mm) according to Motteran et al.⁴³ H₂ and CH₄ yields were calculated by the amount of these biogas accumulations in the headspace per amount of carbohydrate consumed.

The quantification of volatile organic acids (VOA) was carried out for all reactor conditions at the end of each stage (Stage I and Stage II), by high-performance liquid chromatography (HPLC), equipped with a refraction index detector (RID-10A), UV diode array detector (SPD-M10Avp), LC-10ADvp Pump, CTO-20A oven, SCL 10 Avp controller, and Aminex HPX-87H column, 300 mm × 7.8 mm (BioRad). The mobile phase consisted of H₂SO₄ (0.01 N) at a 0.5 mL min⁻¹ flow rate.⁴⁴ Alcohols were determined by gas chromatography (GC) equipped with a HPINNOWAX column and flame ionization detector using H₂ as the carrier gas with synthetic air and N₂ as the auxiliary gas.⁴⁵

Determination of the total volatile solids (TVS) and pH were conducted following standard methods for the examination of water and wastewater,⁴⁶ whereas carbohydrates determination was carried out indirectly by the phenol–sulfuric acid method using glucose as the standard.⁴⁷

2.5 Kinetic analysis

The experimental data (H₂ and CH₄ values) were adjusted to the average values obtained from triplicates using the Origin® 9.0 software (OriginLab, Northampton, MA). The data of the accumulated H₂ and CH₄ production were adjusted using the Gompertz equation, modified by Zwietering (eqn (1)).⁴⁸where: P = potential H₂ and CH₄ production (mmol L⁻¹), R_m = H₂ and CH₄ production rate (mmol h⁻¹), t = time (hours), e = Euler's number (2.71828), λ = lag phase time, i.e., phase that precedes the onset of H₂ and CH₄ production (hours).

2.6 Microbial community analysis

The analyzes of molecular biology were processed by Illumina MiSeq, and for that, the biomass samples collected from the batch reactors fed with SCB at the end of the operation (Stages I and II) were washed in a phosphate buffer and centrifuged at 6000 rpm for 10 min. The pellets were stored at −20 °C. DNA was extracted according to Griffiths et al.⁴⁹ Polymerase chain reaction (PCR) of the 16S rRNA gene fragments employed the primers 968F and 1401R for the Bacteria domain,⁵⁰ while for the archaea domain, the primers 1100F and 1400R⁵¹ were used.

The phylogenetic diversity of the microbial consortium (A I, C I, A II, and C II) was analyzed by Illumina MiSeq sequencing.⁵² Primers 515F (5′-barcode-GTGCCAGCMGCCGCGG-3′) and 806R (5′-GGACTACHVGGGTWTCTAAT-3′) were used to amplify the V4 regions of the bacteria 16S ribosomal RNA gene using a GeneAmp PCR System (ABI company, USA). Amplicons were purified using calibrated Ampure XP beads. Then, the pooled and purified PCR product was used to prepare the Illumina DNA library. Sequencing was performed at MR DNA (https://www.mrdnalab.com, Shallowater, Texas, USA) on a MiSeq following the manufacturer's instructions. Sequence data were processed using an MR DNA analysis pipeline (MR DNA, Shallowater, Texas, USA).

The sequences were joined, depleted of barcodes, then sequences <150 bp and sequences with ambiguous base calls were removed. Sequences were de-noised, OTUs generated, and chimeras removed. Operational taxonomic units (OTU) were defined by clustering at 97% similarity cutoff. Final OTUs were taxonomically classified using Ribosomal Database Project (RDP-II) Classifier and the National Center for Biotechnology Information (NCBI) (https://rdp.cme.msu.edu; https://www.ncbi.nlm.nih.gov). Finally, the raw reads were deposited into the NCBI as a Sequence Read Archive (SRA) database, under Bioproject number PRJNA387459, BioSample SAMN07152251, SRX2841069 (Experiment), SRR5583156 (Run).

3 Results and discussion

3.1 H₂ and CH₄ production from pretreated SCB

In the degradation of lignocellulosic biomass, H₂ and CH₄ production are influenced by several environmental conditions, mainly temperature and pH.^4,6,53,54 A temperature of 37 °C and initial pH 6.0 were applied for all conditions. Fig. 1 and 2 show the variations of H₂ and CH₄ production from 2 g L⁻¹ pretreated SCB in the two-stage dark fermentation process. The reactors were operated for 660 h and different behaviors with respect to H₂ and CH₄ production were verified for the batch reactors studied.

Fig. 1 Accumulated H₂ production from reactors fed with 2 g L⁻¹ pretreated SCB.

Fig. 2 Accumulated CH₄ production from reactors fed with 2 g L⁻¹ pretreated SCB.

For the conversion of lignocellulosic material (such as SCB) to fuel, the cellulose and hemicellulose must be broken down into their equivalent monomer carbohydrates so that microorganisms can use them in their biological route.⁵⁵ In this study, the H₂ production suggested that the pretreated SCB by hydrothermal plus enzymatic processes facilitated the access and use of sugars by the fermentative bacteria. According to Taherzadeh and Karimi,⁵⁶ the hydrothermal pretreatment of lignocellulosic biomass improves the available and susceptible surface area of cellulose molecules and makes them more accessible to hydrolytic enzymes. In this way, the pretreatment methods applied in this study were effective in breaking and/or disrupting the SBC structures in order to increase the access of cellulolytic enzymes and fermenting-bacteria, thereby aiding the hydrolysis of sugars for acids conversion and H₂ production.

The presence of microorganisms in WWTP inoculum (In1), as well as the presence of the cellulolytic-fermentative consortium (In2), slowly promoted the cellulose degradation from SCB producing H₂ in all stages. This fact was evidenced by H₂ observed in all inoculum conditions A (In1), B (In1 + In2), and C (In2) and CH₄ only in Stage II.

In Stage I, H₂ consumption was observed for conditions B I and C I. The development of homoacetogenic bacteria favored the consumption of H₂ under these conditions since CH₄ production was not observed in this stage. Nine genera with homoacetogenic representatives were identified (Acetobacterium, Clostridium, Eubacterium, Holophaga, Moorella, Ruminococcus, Sporomusa, Thermoanaerobacter, and Treponema). According to Drake et al.,⁵⁷ members from these genera may have both a H₂-producing and consuming behavior. Thus, it is likely that these microbial groups are homoacetogens that play an essential role in acetate formation under the operational conditions employed in this study.

The homoacetogens produce specific enzymes that catalyze the formation of acetyl-CoA, which is then transformed to acetate in catabolism or to cell carbon in anabolism.⁵⁸ In the present study, higher concentrations of acetic acid (Table 6) were observed in conditions with higher H₂ consumption (B I and C I), thus confirming the presence of homoacetogenic metabolism.

In Stage II, H₂ consumption was also verified, however, with the presence of CH₄ for all conditions. In this case, the hydrogenotrophic methanogens were probably the most likely responsible for this decrease. After all, under mesophilic or thermophilic conditions, studies have revealed that homoacetogenesis cannot strongly compete with hydrogenotrophic methanogenesis since the latter produces more energy.^59,60 Even though the occurrence of OTUs related to H₂-producing bacteria, such as Clostridium, Enterobacter and Bacillus in this study, lower H₂ production was observed under A II, B II, and C II, compared to the previous stage (Stage I). It is, thus, likely that H₂ consumption by homoacetogens (Table 6) and/or methanogens (Table 8) could explain this finding.

Table 2 shows the kinetic parameters for H₂ and CH₄, such as potential production (P), maximum production rate (R_m), and λ. The mixture of inoculum 1 and inoculum 2 (condition B) did not favor higher H₂ production compared to the assays using the inocula separately (conditions A and C); probably, due to the lower amount (5%) employed of each inoculum, which may have promoted slower growth of the interest microorganisms.

Table 2 Parameters obtained from the modified Gompertz equation adjustment

Condition	H2				CH4
	P (mmol L^−1)	R m (mmol L^−1 h^−1)	λ (h)	R ^2	P (mmol L^−1)	R m (mmol L^−1 h^−1)	λ (h)	R ^2
a A – inoculum 1, B – inoculum 1 + inoculum 2; C – inoculum 2; I – Stage I and II – Stage II. * Maximum values not adjusted in modified Gompertz equation.
A I	3.54 ± 0.2	0.59 ± 0.6	7.5 ± 3.9	0.95	—	—	—	—
A II	2.0 ± 0.1	0.57 ± 0.2	36.7 ± 3.1	0.89	4.8 ± 0.1	0.04 ± 0.004	698.4 ± 5.7	0.98
B I	2.64 ± 0.04	0.32 ± 0.04	3.1 ± 0.6	0.98	—	—	—	—
B II	0.8 ± 0.05	0.29 ± 0.04	37 ± 3.7	0.87	6.1 ± 0.3	0.03 ± 0.007	634.99 ± 20.2	0.89
C I	5.47 ± 0.1	0.22 ± 0.01	16 ± 0.8	0.99	—	—	—	—
C II	0.52 ± 0.03	0.21 ± 1.8	37 ± 2.4	0.88	7.6 ± 0.2	0.1 ± 0.01	668.6 ± 4.5	0.97

---

In Stage I, 3.5 ± 0.2 mmol L⁻¹, 2.6 ± 0.04 mmol L⁻¹, and 5.5 ± 0.1 mmol L⁻¹ of H₂ were observed for conditions A (In1), B (In1 + In2), and C (In 2), respectively (Table 2). Moreover, in Stage II, the maximum H₂ production was 2.0 ± 0.1 mmol L⁻¹, 0.8 ± 0.05 mmol L⁻¹, and 0.5 ± 0.03 mmol L⁻¹ for the same conditions, respectively. H₂ consumption was observed in all the studied conditions of both stages, except in A I. The presence of homoacetogenic bacteria and/or methanogenic archaea contributed to this reduction on H₂ concentration. The homoacetogenic bacteria are very versatile organisms and are able to convert a variety of substrates to acetic acid as the main end product observed.⁵⁸ Acetic acid was the main end product here with a little production of other metabolites for all conditions studied (Table 4).

Table 3 Parameters of specific biogas production and biogas yields in batch reactors^a

Condition	Parameters
	H2 yield (mmol H2 g^−1 substrate)	CH4 yield (mmol CH4 g^−1 substrate)	Specific production H2 (mmol L^−1 g^−1 TVS)	Specific production CH4 (mmol L^−1 g^−1 TVS)
a A – inoculum 1, B – inoculum 1 + inoculum 2; C – inoculum 2; I – Stage I and II – Stage II.
A I	3.12	—	0.94	—
A II	1.67	3.85	0.70	1.66
B I	2.06	—	0.62	—
B II	0.71	4.75	0.28	2.20
C I	4.26	—	1.81	—
C II	0.51	6.30	0.20	2.94

---

Table 4 Organic acids and alcohols produced under each set of experimental conditions^a

Condition	VOA (mg L^−1)				Alcohol (mg L^−1)
	Acetic acid	Butyric acid	Propionic acid	Acetic acid (%)	Butanol	Ethanol
a VOA – volatile organic acid; SCB – sugarcane bagasse; A – inoculum 1, B – inoculum 1 + inoculum 2; C – inoculum 2 I – Stage I and II – Stage II.
A I	460	50	50	68	24.4	19.3
A II	760	70	100	75	—	—
B I	5880	520	50	84	—	—
B II	4500	620	1260	68	23.8	—
C I	9270	480	700	87	24.1	—
C II	30	10	—	36	—	—

---

It was reported that the H₂ produced from butyrate oxidation reacts rapidly with CO₂ to form acetate by homoacetogenesis.⁶¹ In the present study, the higher concentrations of acetic acid to the detriment of butyric acid in all the studied conditions is consistent with this statement.

No CH₄ was detected in Stage I; however, 4.8 ± 0.1 mmol L⁻¹, 6.1 ± 0.3 mmol L⁻¹, and 7.6 ± 0.2 mmol L⁻¹ CH₄ were observed in Stage II for conditions A, B, and C, respectively. The methane produced, under conditions A and B, originated mainly from hydrogenotrophic methanogenic archaea, due to the decrease in H₂ production and low consumption of acetic acid (Table 4). In this study, representatives of methanogenic archaea were distributed in four orders (Methanobacteriales, Methanomassiliicoccales, Methanomicrobiales, and Methanosarcinales), totaling 82%, 55.8%, 62%, and 64.6% of the sequences for samples from conditions A I, C I, A II, and C II, respectively.

Condition C (In2) provided a higher CH₄ potential production (25.3 ± 0.9 mmol L⁻¹) in Stage II. Probably, In2 (fermentative and cellulolytic consortia) enabled a greater conversion of soluble sugars to H₂ and organic acids, which promoted higher CH₄ production. Thus, in this condition, the CH₄ was produced mainly by methanogenic acetoclastic archaea since the acetic acid produced in Stage I was practically all consumed in CH₄ formation (Table 4).

Values of λ below those observed by Lay⁶² were observed in this study (Table 2). The authors evaluated H₂ production in a batch reactor operated at 37 °C, pH 7, with cellulose as the substrate (12.5 g L⁻¹ and 50 g L⁻¹). These authors used anaerobic sludge as the inoculum and solid residues as the substrate and observed λ values ranging from 3.74 to 4.26 days.

In the present study, λ ranged from 3.1 h in condition B I (In1 + In2; Stage I) to 37 h in B II (In1 + In2; Stage II). The condition that used In2 (fermentative and cellulolytic consortia) as the sole inoculum source was the best strategy for H₂ (Stage I) and CH₄ (Stage II) production, although the rates were lower compared to conditions A (In1 – anaerobic lagoon of WWTP) and B (In1 + In2). Since In2 is less diverse than In1, related to the microbial community, the production and consumption of metabolites were slower, leading to a lower R_m and higher P.

In a study of H₂ production, Soares et al.³⁷ employed 2 g L⁻¹ of hydrothermally pretreated SCB and the same culture medium of this study (cellulose anaerobe). The authors observed a λ of 17.5 h and 4.44 mmol H₂ per L when a thermophilic sludge was employed as the inoculum and the reactors were operated at 55 °C. According to these authors, the higher λ verified for the SCB assays when compared to the glucose assay (9.4 h) was due to the lower availability of sugars in the SCB. In the present study, the association of enzymatic pretreatment to the hydrothermally pretreated SCB may have improved the availability of sugars in the lignocellulosic biomass; thus, promoting lower values of λ in Stage I.

The H₂ production rates (R_m) in this study were similar to those observed by Ren et al.,⁶³ who evaluated the effects of cellulose concentration on H₂ production rates using cow manure as the inoculum. The authors observed production rates of 0.42 mmol L⁻¹ h⁻¹ (5 g L⁻¹ cellulose) and 0.65 mmol L⁻¹ h⁻¹ (10 g L⁻¹ cellulose), while in the present study, the highest values for 2 g L⁻¹ SCB were 0.59 mmol L⁻¹ h⁻¹ (A I), 0.57 mmol L⁻¹ h⁻¹ (A II), and 0.3 mmol L⁻¹ h⁻¹ (B I) (Table 2).

Although Ren et al.⁶³ presented similar R_m values, the substrate conversion values were 0.084 and 0.065 mmol H₂ L⁻¹ h⁻¹ for 5 and 10 g L⁻¹ of cellulose. In this study, these values were much higher, namely 0.29, 0.28, and 0.15 mmol H₂ L⁻¹ h⁻¹ for the reduced applied substrate (2 g L⁻¹). The inocula employed in this study were more efficient in converting a more complex substrate to H₂, and also the combination of both pretreatments (hydrothermal and enzymatic) had an effective effect on biogas production.

Rabelo⁶⁴ evaluated H₂ production using the same inoculum as in the present study (In1) and observed a much lower H₂ production rate (0.05 mmol L⁻¹ h⁻¹) when 2 g L⁻¹ cellulose without enzyme addition was used as the substrate. The enzymatic and hydrothermal pretreatment of SCB in the present study may have contributed to a higher H₂ production rate (0.59 mmol L⁻¹ h⁻¹), compared to that observed by Rabelo.⁶⁴

The maximum H₂ (4.26 mmol g⁻¹ SCB) and CH₄ (6.30 mmol g⁻¹ SCB) yields were obtained for condition C (In 2) in stages I and II, respectively (Table 3). Liu et al.⁴ observed a similar H₂ yield (43 mL g⁻¹ VS added or 5 mmol g⁻¹ VS) in a study of H₂ and CH₄ production from household solid waste (HSW) in a two-stage fermentation process. The HSW consisted mainly of food residue, paper and garden waste. Instead, the CH₄ yield achieved by these authors was slightly lower (500 mL g⁻¹ VS or 4 mmol g⁻¹ VS).

The specific H₂ and CH₄ productions are shown in Table 3. The highest specific H₂ production was observed for condition C I (1.81 mmol L⁻¹ g⁻¹ TVS). A decrease in H₂ production was observed with increasing stages (Stage I to II), with a reduction of 25.7%, 54%, and 88.8% for conditions A, B, and C, respectively. The highest specific CH₄ production (2.94 mmol L⁻¹ g⁻¹ TVS) was also observed for condition C (Stage II) (Table 3).

Rabelo⁶⁴ evaluated H₂ and CH₄ production using the same inoculum as in the present study (In1) and observed a lower specific CH₄ production (0.25 mmol L⁻¹ g⁻¹ TVS) when 2 g L⁻¹ cellulose (without enzyme addition) was used as the substrate. In this study, the addition of another inoculum source, In3 (granular sludge obtained from the UASB reactor used in the treatment of poultry slaughterhouse wastewater), in Stage II may have favored the higher specific CH₄ production. In addition, it is noteworthy that the use of physical (hydrothermal) and biological (enzymatic) pretreatments conjugated in SCB may have increased the availability of sugars in Stage I, which were converted to acids, generating more CH₄ at the end of Stage II, compared to in ref. 65

In this study, the pretreatment with A. niger provided the necessary enzymes for the SCB breakdown, different from that carried out by ref. 66, in which external enzymes were added directly in to the culture medium. Xie et al.⁶⁶ evaluated H₂ and CH₄ production from potatoes by a two-phase anaerobic fermentation. Better performances of 271 mL H₂ per g TVS (22 mmol L⁻¹ g⁻¹ TVS) and 158 mL CH₄ per g TVS (14 mmol L⁻¹ g⁻¹ TVS) were achieved when the residue was pretreated by α-amylase and glucoamylase. Furthermore, the H₂ production process was more interesting due to its reduced costs since the microorganisms that produce these enzymes were acquired from the inocula and the biological pretreatment.

Ratti et al.^67,68 also used commercial enzymes (cellulose, Sigma-Aldrich®) for the hydrolysis of cellulose in H₂ employing batch reactors assays. The authors used as an inoculum source pretreated (acid and heat shock) and leached rumen fluid. The authors observed that the inoculum pretreatment may have inhibited the cellulolytic microorganisms, since H₂ was detected only in the presence of the enzyme. On the other hand, the pretreatment applied in the present study did not inhibit the cellulolytic-fermentative microorganisms, making this process more attractive for the generation of biogas (H₂ and CH₄) in two stages. The biological process in two or more stages for sequential H₂ and CH₄ production has been considered as an alternative to improve the economic viability of waste treatment.⁶⁹

3.2 Production of soluble microbial products

At an initial pH of 6.0, acetic acid was the main metabolite formed in all studied conditions (Table 4), suggesting that fermentation occurred mainly through this route. The organic acids produced (such as acetic, butyric, and propionic) at the end of Stage I were available for use in Stage II, thus favoring the development of acetoclastic methanogens represented by members of the genus Methanosaeta (Methanosarcinales order) (A II: 21.8% and C II: 21.5%, respectively) and, according to Whitman et al.,⁷⁰ archaea belonging to the family Methanosaetace when using only acetate as a catabolic substrate. In this study, the proportion of acetic acid varied from 62% to 81%.

A higher acetic acid concentration at the end of Stage I was observed for conditions B I (5880 mg L⁻¹) and C I (9270 mg L⁻¹). For condition A I (In1), H₂ production was stable and the production of acetic acid was lower (460 mg L⁻¹). In addition, decreases of 8%, 90%, and 68% in the concentration of H₂ (headspace) were observed under conditions A I, B I, and C I, respectively, without CH₄ production. At the end of Stage II, there was a total consumption of the H₂ produced. It is likely that in all cases there was the development of homoacetogenic bacteria. According to Lay (2001), this group may consume 11–43% of the H₂ (in liquid medium) present in the reactors, since they produce acetic acid from CO₂ and H₂.⁶² Acetic acid-producing bacteria, which may also consume the H₂, were identified in this study; for example sequences similar to Acetobacterium, Clostridium, and Holophaga.

The higher concentration of acetic acid availability at the end of Stage I provided in Stage II a potential CH₄ production of 4.8 ± 0.1 mmol L⁻¹, 6.07 ± 0.3 mmol L⁻¹, and 7.6 ± 0.2 mmol L⁻¹ for A II, B II, and C II, respectively. The lowest acetic acid concentration in C II (30 mg L⁻¹) corresponded to a higher CH₄ yield (6.3 mmol g⁻¹ SCB) (Table 3).

The pretreatment (hydrothermal plus enzymatic) applied on SCB was feasible for breaking down the complex molecules of SCB, thereby generating fermentable sugars, which were then converted to organic acids. The acetic acid route for H₂ production was the most evident in the present study. Some other short chain organic acids may have been generated in Stage I, such as lactic acid, although this was not detected in the present study, since its conversion to other acids may have occurred. According to Kapdan et al.,¹⁵ this acid may follow the conversion pathway to produce acetic, butyric, and propionic acids. Thus, the prevalence of these acids as the main end products occurred in this stage (Fig. 3).

Fig. 3 Scheme of soluble metabolites production in Stages I and II for H₂ and CH₄ production from pretreated SCB.

Thus, lactic acid may be related to another H₂ production pathway by microorganisms such as lactic acid bacteria (LAB), since these organisms are commonly present in the autochthonous community of sugarcane.^71,72 Members from LAB were identified in this study. The most representative LAB was the genus Enterococcus (0.5–28.9% of total relative abundance), while in a reduced proportion (<1% of total relative abundance), the genera Granulicatella, Lactobacillus, and Streptococcus were also identified.

According to Moreira et al.,⁷³ the diversity of the produced metabolites (acids) in the sugars fermentation may be due to competition among the strains present in the more complex microbial consortium (condition B). The glucose consumption from the polysaccharides from the SCB breakdown (hydrothermal plus enzymatic pretreatment) and their processing routes in acids and H₂ in Stage I and for CH₄ in Stage II can be verified in Fig. 3.

Rabelo⁶⁴ identified Desulfovibrio sp. (sulfate reducing bactéria-SRB) when using the same inoculum as in the present study (In1) for H₂ production from cellulose. According to these authors, the sulfate present in the culture medium and the acetic acid generated may have favored the growth and maintenance of these groups in the H₂ production reactors.

By using Illumina MiSeq sequencing 1.13% (A I), 0.08% (C I), 1.97% (A II), and 3.89% (C II) of the sequences were related to SRB, totaling 19 genera. Among these, the most representative genera were Desulfitobacterium (0.1–0.95%), Desulfosporosinus (0.03–3.28%), Desulfotomaculum (0.02–0.14%), and Desulfovibrio (0.01–0.42%). These genera were also identified as SBR in other studies.^74–76

In all conditions, butyric and propionic acids were also observed, but in smaller proportions (Table 4). Butanol and/or ethanol were observed at low concentrations compared to acetic acid in conditions A I, B II, and C I (Table 4).

3.3 Illumina sequencing analysis

3.3.1 Illumina result analysis and diversity indices. Using Illumina MiSeq technology, the raw data totaled 1 027 023 sequences from the four samples (Table 5). After trimming, singletons were removed, and the resulting sequences, with an average length of 272 bp, were used to determine a total of 40 and 2128 OTUs (operational taxonomic units) used in the taxonomical classification for archaea and bacteria domains, respectively.

Table 5 Molecular biology analysis, taxonomic results, richness estimation, and diversity index from samples A and C from stages I and II^a

Sequencing result analysis	A I		C I		A II		C II
	Arch	Bact	Arch	Bact	Arch	Bact	Arch	Bact
a bp–bases pair; OTU – operation taxonomic units; A – inoculum 1; B – inoculum 1 + inoculum 2; C – inoculum 2; Arch – archaea domain; Bact – bacteria domain.
Good's estimated coverage (%)	0.919	0.999	1.000	0.999	0.999	0.999	0.999	0.998
Sequence length (bp)	271		271		272		273
Total sequences (trimming data)	111	278 605	52	308 665	4813	235 537	3910	135 336
Total of OTUs (taxonomical classification)	27	1131	16	843	38	1740	36	1537
Richness estimation
Chao1	36.17	1432	25.33	1243	38.5	1863	37	1749
Diversity index
Shannon (H)	2.64	3.84	2.38	2.72	2.19	4.35	2.37	4.53
Simpson (D)	0.88	0.93	0.88	0.87	0.82	0.94	0.87	0.97

---

Sequences were classified in two phyla for the archaea domain (Crenarchaeota and Euryarchaeota) present in all the samples studied. A total of 25 phyla were observed in the bacteria domain, including 22 in A I, 19 in C I, 25 in A II, and 24 in C II. In the two conditions of the H₂ stage process (A I and C I), the most prevalent phylum was Firmicutes (79.47% and 55.65%, respectively). According to Ratti et al.⁷⁹ and Santos et al.,⁸⁰ this phylum contains the main microorganisms that produce H₂. Moreover, the anaerobic members of the phylum Firmicutes are also known to be producers of organic acids with short chains of 3 and 4 carbons, like propionic acid and butyric acid,⁸¹ which justifies the metabolic routes found in this study (Fig. 3 and Table 4).

The Shannon–Weiner (H) and Simpson (D) indices of the microbial consortium were analyzed (Table 5). Comparing the samples from the H₂ stage process (Stage I), the microbial community in A I (3.84) had a higher diversity than C I (0.93). The bacterial community was composed of 209 OTUs with a predominance of the families Enterococcaceae (28.8%) and Clostridiaceae (23.5%), both from phylum Firmicutes. Enterococcaceae was also observed in C I with a similar relative abundance (24.1%), but there was a predominance of Porphyromonadaceae from phylum Bacteroidetes (28.8%).

Bacteria genera belonging to the family Enterococcaceae are Gram-positive and catalase-negative cocci, being also ovoid or spherical in shape, and the cell wall comprises the diamino acid lysine. These organisms are non-endospore-forming and have a low G + C content (33–46%) in the DNA. They are facultative anaerobic and chemo-organotrophic.⁸²

Members of the family Porphyromonadaceae have a variety of cellular morphologies and chemotaxonomic behaviors. Many species of this family are indigenous of the human and animal gastrointestinal tract and oral cavity, nonetheless some species are usually related with a variety of human and animal infections.⁸³

Comparing the samples from CH₄ stage process (Stage II), the microbial community had similar values of diversity indices (Table 5), the bacterial community from A II and C II was composed of 296 and 298 OTUs with a predominance of the family Rikenellaceae (phylum Bacteroidetes) (19.9%) and Thermotogaceae (phylum Thermotogae) (11.9%), respectively.

3.3.2 Bacteria community analysis. In this study, there was a higher relative abundance of six bacterial phylum (Fig. 4 and 5). The most representatives were Bacteroidetes (1.1–40.3%), Chloroflexi (0.1–11.6%), Firmicutes (20.5–79.5%), Proteobacteria (10.8–17.1) Spirochaetes (0.1–9.1%), and Thermotogae (0.1–11.9%). The highest prevalence of the phylum Firmicutes was largely due to the presence of three families, Clostridiaceae (6.3–23.5%), Enterococcaceae (0.5–28.9%), and Ruminococcaceae (3.1–17.1%).

Fig. 4 Interaction of the main bacterial genera involved in the two-stage biofuel production from sugarcane bagasse by dark fermentation.

Fig. 5 Relative abundance of bacterial phylum, family and genus, in reactors from conditions A I (white), C I (gray), A I (cross stripes), and C II (black).

The Firmicutes phylum comprises some microorganisms able to degrade lignocellulosic residues,⁸⁴ and according to Bergquist et al.,⁸⁵ the Firmicutes together with Actinobacteria phyla comprise 80% of the cellulolytic bacteria. The Actinobacteria and Bacteroidetes phyla include microorganisms able to degrade lignocellulosic substrates.⁸⁶ These microbial groups show potential cellulolytic activity, producing enzymes that are capable of degrading or initiating the breakdown of complex carbon compounds, such as cellulose and hemicelluloses (mainly in soils), thus representing a key step in the carbon cycle,^85,87 which justifies their considerable presence in terms of the relative abundance of these phyla in this study.

Streptomyces was the genera belonging to the Actinobacteria phylum with a greater relative abundance (A-II 0.06% and C-II 0.05%) and Bifidobacterium thermophilum (A-I 1.71%; –I 0.02%; A-II 0.03%; C-II 0.02%) as well as representatives of the Micromonosporaceae family (A-II 0.03%). Both of these genera may have played an important role in the ecology of the reactors in the biogas production. Streptomyces genus can produce antibiotics,⁸⁸ restricting the growth of other microorganisms, Micromonospora species can produce several hydrolytic enzymes, thus playing an active role in the degradation of organic matter.⁸⁹

Within the Clostridiaceae family (Firmicutes phylum), 23.1% (A I), 6.3% (C I), 6.1% (A II), and 7.2% (C II) of the reads were related to the Clostridium genus. Many anaerobic bacteria can produce H₂ from carbohydrate-rich organic wastes. Species such as C. buytricum,⁹⁰C. thermolacticum,⁹¹C. pasteurianum,²⁰C. paraputrificum,⁹² and C. bifermentants⁹³ are obligate anaerobes and form spores. Clostrida members produce H₂ during the exponential growth phase.

Rabelo⁶⁴ also observed sequences related to Clostridium (67 clones) when the author used the same inoculum as in this study (In1) for H₂ and CH₄ production using cellulose as a substrate in anaerobic batch reactors. Furthermore, sequences similar to Desulfovibrio (6 clones), Raoultella (2 clones), and Klebsiella (2 clones) were observed to a lesser extent.

The higher diversity observed in A I (H′: 3.84) may be behind the occurrence of H₂-consuming bacteria, which may explain the lower H₂ yield observed when compared with C I (Table 5). The presence of members from the genera with a homoacetogenic behavior was greater in A I (33.72% of the sequences) compared with C I (14.26% of the sequences) (Table 6). However, the higher H₂ production rate in A I (0.59 mmol L⁻¹ h⁻¹) may have provided a stable H₂ level during Stage I, although H₂ consumption by these microorganism may also have occurred. On the other hand, the lower H₂ production rate in C I (0.22 mmol L⁻¹ h⁻¹) promoted a sudden decrease in H₂ production, although this production reached higher values (Fig. 1). This diversity of members with possible homoacetogenic behavior favored the low values of H₂ yield compared to the theoretical one.

Table 6 Relative abundance of homoacetogens genera identified by illumina MiSeq platform from conditions A I, C I, A II, and C II

Genus	A I (%)	C I (%)	A II (%)	C II (%)
Acetobacterium	0.00	0.00	0.01	0.00
Clostridium	23.13	6.26	6.13	7.20
Eubacterium	0.02	0.01	0.08	0.13
Holophaga	0.00	0.00	0.02	0.03
Moorella	0.00	0.00	0.03	0.06
Ruminococcus	10.33	7.85	1.19	0.54
Sporomusa	0.23	0.14	0.06	0.07
Thermoanaerobacter	0.00	0.00	0.00	0.01
Treponema	0.00	0.00	0.23	0.26
Total	33.72	14.26	7.74	8.31

---

The presence of homoacetogens in Stage II was less evident than in Stage I (Table 6), contributing to the hypothesis that the methanogens were favored at this Stage and were the main organisms responsible for the decrease in H₂ production as well as in its consumption. In this study, homoacetogens were probably unable to compete effectively for H₂, as they show quite poor growth kinetics compared to methanogens.^94,95

Even though the bacterial community of condition B was not phylogenetically assessed, it was possible to infer that there was also the presence of these homoacetogens organisms, since the inoculum source consisted of the mixture of In1 (used in condition A I) and In2 (used in condition C I). The remarkable presence of the genus Clostridium, in condition A I may be related mainly to H₂ production and not to homoacetogenesis, since representatives of this genus may have both behaviors. Some genera from the Clostridiacea family may have had a direct influence on the biogas production, either for the direct H₂ production, or for the degradation of sugars, such as cellulose, hexoses, and glucose, generating organic acids (acetic acid and formic acid) as substrates for the methanogenic archaea (Fig. 4).

The genera Brassicibacter (A I 0.06%, C I 0.01%, A II 0.29%, and C II 0.14%) and Anaerosporobacter are capable of producing acetic acid and formic acid, which may have occurred in Stage II, as well as propionic acid, ethanol,⁹⁶ and H₂ by Anaerosporobacter.⁹⁷ The species of these genera are also capable of using fructose, glucose, and sucrose as a carbon source, which may have been generated by the enzymatic hydrolysis in the biological pretreatment and by the enzymatic action of other microorganisms from SCB.⁹⁶

However, although the genus Lutispira found in this study is capable of using a few carbohydrates, it cannot directly ferment arabinose, cellobiose, cellulose, fructose, galactose, glucose, glycerol, lactose, maltose, mannose, ribose, starch, sucrose, xylose, among others, and thus members of this group are restricted to the availability of other energy sources.⁹⁸ Likewise the genus Oxobacter are not able to use sugars, amino acids, organic acids and alcohols as energy sources; however, the fermentation of pyruvate by these microorganisms generates acetic acid and CO₂. When CO₂ is catabolically used, acetic acid and butyric acid are generated as metabolic products.⁹⁹

In the same way, the genus Caloramator has ethanol and acetic acid as end products in the glucose degradation.¹⁰⁰ In the present study, the acetic acid produced will contribute to the methane production in Stage II. The differential of the genera Fonticella and Geosporobacter is that these microorganisms are able to use cellobiose, besides fructose, sucrose, and glucose as electrons donors,¹⁰¹ with formic acid, acetic acid, ethanol, and CO₂ as the main end products from the fermentation of these sugars.¹⁰²

The genus Fervidicella (family Clostidiacea) is capable of fermenting various types of carbohydrates, among them glucose, and produces ethanol, acetic acid, CO₂, and H₂. However, it does not use organic acids as a source of carbon and energy.¹⁰³ In the same way, members of the genus Saccharofermentans are able to use hexoses, polysaccharides, glycoses, and alcohols, generating acetic acid, lactic acid, fumaric acid, CO₂, and H₂. However, in cellulose degradation, the representatives of this microbial group do not generate organic acids, being therefore dependents on the action of the enzymatic treatment of other microorganisms.¹⁰⁴ Probably, this group benefited from the enzymatic treatment of SCB used in this study.

Furthermore, a higher relative abundance of the hydrogenotrophic methanogenic Methanobacterium was observed in this condition, thereby allowing higher H₂ consumption (40.4% – A I and 7.7% – C I). Another genus identified mainly in samples from Stage I was Enterobacter (Firmicutes phylum; Enterobacteriacea family). Members from this genus are facultative anaerobes and the quantity of H₂ produced is equivalent to that from Clostridum members.^90,105 Both aforementioned genera were observed primarily in condition A I (Fig. 5).

A high H₂ yield was observed when Enterobacter sp. was used as the inoculum with 16.5 g L⁻¹ xylose in batch reactors (2.0 ± 0.05 mol H₂ per mol substrate).¹⁰⁶ The higher value obtained by these authors was probably due to the simplicity of the substrate and to the higher concentration used when compared to that used in the present study (2 g L⁻¹ g⁻¹ SCB). Some bacteria from this genus have the ability to degrade cellulosic substrates, since cellulase activity has been observed in some members of the Entrobacteriaceae family.¹⁰⁷

With their higher relative abundance in this study, members from Enterococcus genera (Firmicutes phylum) were observed mainly in the H₂ production stage (A I and C I). In the present study, two species belonging to Enterococcus may be related to SCB degradation and H₂ production in stage I, with E. faecalis (21.2% and 17.9%) and E. villorum (7.6% and 6.1%) identified in samples from A I and C I, respectively.

The potential of Enterococcus gallinarum G1 to degrade cellulose and produce H₂ under mesophilic conditions (37 °C) and pH 6.5 was evaluated.¹⁰⁸ The authors noted that cellulose hydrolysis and H₂ production occurred with 5 h of λ time and H₂ continued to be produced in the following 60 h. The H₂ yield observed was 2.38 mmol H₂ g⁻¹ cellulose and the end liquid products were primarily acetate, propionate, and butyrate. Hu and Zhu¹⁰⁹ and Hu et al.¹¹⁰ also reported that acetic acid and butyric acid were the predominant metabolites.

In the present study, acetate, propionate and butyrate were also the main metabolites generated. Moreover, a higher H₂ yield was observed for both conditions (A I and C I). Although the substrate employed in this study was more complex (SCB), the higher microbial diversity used as the inoculum may have favored the conversion of SCB to easily degradable carbohydrates, thereby promoting the higher H₂ yield (A I – 3.12 mmol H₂ per g SCB added and C I – 4.26 mmol H₂ per g SCB added).

A bacterium able to degrade cellulose was isolated from the gut of Microcerotermes diversus.¹¹¹ According to the authors, three cellulose degrading bacteria isolated belonged to the genera Acinetobacter, Pseudomonas and Staphylococcus. Members from the genus Acinetobacter were also observed in this study, mainly in samples from Stage I, 4.8% and 14.4% for A I and C I, respectively.

Furthermore, some members from this genus have the ability to degrade hemicellulose, an important component of SCB. According to Lo et al.,¹¹² xylanase is the main enzyme to hydrolyze hemicellulose and to originate fermentable sugars (five-carbon) for H₂ production. A cellulolytic bacterium, Acinetobacter junii F6-02, was isolated from soil in Taiwan and this strain was able to produce xylanase and cellulase extracellularly with high efficiency.

Within this genus, two species were identified, Acinetobacter genomo sp. 3 (2.5% – A I and 10.2% – C I) and Acinetobacter soli (1.9% – A I and 4% – C I). The high proportion of this genus on C I may have led to the higher degradation of cellulose and hemicellulose from SCB, which provided a higher H₂ yield in this condition.

Many cellulolytic microorganisms have been described by their capacity to hydrolyze lignocellulosic substrates for energy production. Those cellulolytic organisms include the bacteria Ruminococcus, Bacillus, Streptomyces, and Bacteroides.¹¹³ These genera were also observed in the present study. Among them, Ruminococcus had a higher relative abundance (10.3%, 7.8%, 19.2%, and 5.3%), followed by Bacillus (0.6%, 5%, 0.3%, and 0.3%), Bacterioides (0.1%, 0.1%, 2.5%, and 2.3%), and finally, Streptomyces (0%, 0%, 0.06%, and 0.05%) for samples from A I, C I, A II, and C II, respectively. Consequently, these microbial groups may have used the cellulose from the sugarcane bagasse as an energy and catabolic source.

Another group, probably involved in cellulose degradation, was found at a high proportion in the CH₄ stage. The Geotoga genus (Thermotogales order) was observed mainly in A II (10.1%) and C II (11.1%). Thermotogales are very important due to their ability to metabolize simple and complex carbohydrates, such as sucrose, cellulose, maltose, and xylose.¹¹⁴ The SCB used in this study served as a cellulose source for this bacteria group.

Ruminiclostridium is an anaerobic and cellulolytic bacteria observed mainly in C I (9.1%). This mesophilic bacterium metabolizes cellulose and some hemicellulosic polysaccharides into fermentable sugars.¹¹⁵ The substrate employed in this study (SCB) may have favored the development of members from this genus. More noticeable in the samples from CH₄ production (A II and C II), the genus Rikenella, members of the genus Bacteroides, were found in several different digestive tracts and probably indigenous from the inoculum used in the CH₄ stage.

The characteristics of the culture medium may have favored some microbial groups. In fact, approximately, 0.7% (A I), 28.8% (C I), 1.3% (A II), and 0.6% (C II) of the sequences found were associated with Proteiniphilum genus. Proteiniphilum is an obligately anaerobic Gram-negative bacterium. Members from this genus grow better under mesophilic conditions and the yeast extract (YE) and peptone can be used as energy sources. P. acetatigenes was first observed and isolated from an upflow anaerobic sludge blanket-type anaerobic digester.¹¹⁶ The presence of YE (1 g L⁻¹) in the culture medium, in this study, may have caused the development of this species. Furthermore, higher concentrations of acetic acid (9262.3 mg L⁻¹) and propionic acid (700.1 mg L⁻¹) were observed for the condition with a higher relative abundance of P. acetatigenes (C I).

A total of 250 bacterial species were classified in the biomass of the reactors. In this study, 13 of the most common species (>1% of the classified sequences) accounted for 51.6%, 80.6%, 16.1%, and 17.8% of the classified sequences for reactors from conditions A I, C I, A II, and C II, respectively (Table 7). In the four samples, the difference in the bacterial community composition was evident with a higher abundance of Enterococcus faecalis and Proteiniphilum acetatigenes in the H₂ production stage (A I and C I, respectively) and Geotoga petraea in the CH₄ production stage (A II and C II).

Table 7 Relative abundance of the identified bacterial species from conditions A I, C I, A II, and C II

Species	A I (%)	C I (%)	A II (%)	C II (%)
Acinetobacter genomo sp. 3	2.52	10.16	0.07	0.09
Acinetobacter soli	1.91	4.05	0.04	0.07
Bacillus badius	0.10	4.17	0.03	0.04
Bifidobacterium thermophilum	1.71	0.02	0.03	0.02
Candidatus cloacimonas acidaminovorans	0.01	0.02	0.52	3.39
Clostridium lactatifermentans	1.46	0.07	0.26	0.06
Clostridium metallolevans	2.86	0.03	0.82	0.33
Clostridium termitidis	0.46	9.08	2.42	1.74
Enterobacter hormaechei	10.99	0.13	0.04	0.08
Enterococcus faecalis	21.25	17.93	0.51	0.35
Enterococcus villorum	7.61	6.15	0.23	0.17
Geotoga petraea	0.09	0.10	10.06	11.09
Proteiniphilum acetatigenes	0.67	28.75	1.07	0.38
Others	48.35	19.36	83.89	82.19
Total	100	100	100	100

---

3.3.3 Archaea community analysis. The role of methanogenesis in anaerobic digesters is the terminal step in the transformation of complex organic materials to CH₄ (mineralization). This process consists of a complex interaction of hydrolytic and fermentative organisms, fatty acids-oxidizing-bacteria (often syntrophic), and acetogenic bacteria, prior to the activity of methanogens.¹¹⁷

All known methanogens belong to the Euryarchaeota phylum.¹¹⁸ In this study, the phylum Euryarchaeota represented a large majority of the sequences identified in the biomass under all conditions. There was a predominance of four archaeal families (Fig. 6). The majority of representatives in all samples were Methanobacteriaceae (7.9–55.8%), Methanoregulaceae (22.5–52.9%), and Methanosaetaceae (4.5–25%).

Fig. 6 Relative abundance of archaeal phylum, family and genus in the reactors under conditions A I (white), C I (gray), A II (cross stripes), and C II (black).

In the four samples studied, 11 genera were classified. The biomass from condition A I contained higher populations of Methanobacterium (50.4%) and from C I, A II, and C II higher populations of Methanolinea (34.6%, 47.1% and 35.6%, respectively).

Methanogens belonging to the genus Methanobrevibacter were also observed in this study. Similar to both genera, Methanobacterium sp. and Methanolinea, they can also use the end products of bacterial fermentation to produce CH₄.^119,120 All these genera produce CH₄ from formate and H₂/CO₂, but not from acetate. In this study, it is possible that the H₂ produced in the last stage (A II: 2 ± 0.1 mmol H₂ L⁻¹ and C II: 0.5 ± 0.03 mmol H₂ L⁻¹) served as substrates for the development of those genera in the reactors (Fig. 5).

The genus Methanosaeta are specialized in metabolizing acetate, reflecting a high affinity for the compound.¹²¹ This genus has been observed to dominate the methanogenic community in anaerobic reactors.¹²² In this study, representatives of this genus were uniformly distributed between samples A II (21.8%) and C II (21.5%). This genus was responsible for the conversion of acetic acid to CH₄ from Stage I to II with a reduction of 55.7% and 99.8% for A II and C II, respectively.

A total of nine archaeal species were classified in the biomass of the reactors (Table 8), which accounted for 14.4%, 11.5%, 8.9%, and 17.7% of the classified sequences for the reactors under conditions A I, C I, A II, and C II, respectively. Differences in the composition of the archaeal community in the four samples were evident with the predominance of Methanosphaerula palustris, Methanobacterium subterraneum and Candidatus methanomassiliicoccus intestinalis in the CH₄ production stage (A II and C II).

Table 8 Relative abundance of classified archaeal species under conditions A I, C I, A II, and C II

Species	A I (%)	C I (%)	A II (%)	C II (%)
Candidatus methanomassiliicoccus intestinalis	1.80	3.85	0.35	10.51
Candidatus methanoregula boonei	0.00	0.00	0.17	0.20
Candidatus nitrosocaldus	4.50	1.92	0.50	1.20
Candidatus nitrosocaldus yellowstonii	0.00	1.92	0.02	0.13
Methanobacterium bryantii	1.80	0.00	0.12	0.05
Methanobacterium subterraneum	2.70	1.92	2.80	1.51
Methanosaeta concilii	0.90	0.00	1.06	1.05
Methanosaeta pelagica	0.90	0.00	0.08	0.00
Methanosphaerula palustris	1.80	1.92	3.84	3.09
Others	85.6	88.5	91.0	82.3
Total	100	100	100	100

---

Methanosphaerula palustris is strictly anaerobic, mesophilic, and mildly acidophilic.¹²³Methanobacterium subterraneum is an autotrophic, halotolerant methanogen.¹²⁴ Both species use H₂/CO₂ for CH₄ formation and do not use acetate. The affiliation of Ca. Methanomassiliicoccus intestinalis to a large cluster of sequences retrieved from paddy soils, freshwater, and marine sediments suggest their recent adaptation to gut environments.¹²⁵

The generation of biofuels from renewable resources, mainly lignocellulosic biomass, is possible from the majority of low-priced and abundant non-food materials that can be obtained from plants. Thus, lignocellulosic materials have the potential to provide second-generation biofuels.¹²⁶ The production of H₂, bio-oils, biogas, alcohols, and biodiesel from renewable biomass, such as SCB, have been a major research focus around the world in order to enhance petroleum fuels and reduce environmental pollution.⁵⁵

The application of two-stage anaerobic digestion and hydrothermal and enzymatic pretreatment, is an excellent strategy to improve H₂ and CH₄ production, since the metabolites produced in one stage could subsequently be used in the next operational stages. It was also observed that the microbial community adapted to the conditions imposed, thus favoring each operational stage, optimizing the SCB degradation system in a two-stage process. Integrating H₂ and CH₄ production in the same reactor, in different stages, provides a higher potential to use the SCB to produce bioenergy. The highly diversified microbial consortia used in this study allowed high yields of both H₂ and methane production.

4. Conclusion

The anaerobic conversion of SCB to H₂ and CH₄ is an attractive alternative to renewable energy. This is a rare study that investigated the combination of the hydrothermal and enzymatic pretreatment of SCB through the activity of A. niger, with biogas measurements, by-products analyses, and phylogenetic characterization of the microorganisms involved. The combination of both pretreatments was adequate in the hydrolysis of SCB, since H₂ was detected in all the studied conditions. The mixed culture obtained from wastewater treatment plant sludge (inoculum 1) was a favorable source for H₂ production using SCB, as well as the cellulolytic and fermentative culture (inoculum 2). However, the mixture of inoculum 1 and inoculum 2 did not favor H₂ production. The highest H₂ and CH₄ potential production was obtained when inoculum 2 (fermentative and cellulolytic consortia) was employed in Stage I (C I), obtaining 5.47 mmol L⁻¹ and 7.60 mmol L⁻¹, respectively. The metabolite generated in the first stage fermentation contributed to methane production in the second one. Acetic acid was the major metabolite formed under all the conditions studied, suggesting that fermentation occurred via this route. Microbial consortium able to produce cellulase and xylanase enzymes, such as Enterobacter, Enterococcus, Acinetobacter, Clostridium, and Geotoga, were identified in this study. In the methanogenic stage, three genera of hydrogenotrophic methanogens (Methanolinea, Methanobrevibacter, and Methanobacterium) and one genus of acetoclastic methanogens (Methanosaeta) were identified. Furthermore, homoacetogenic bacterias, such as Acetobacterium, Clostridium, Eubacterium, Holophaga, Moorella, Ruminococcus, Sporomusa, Thermoanaerobacter, and Treponema were also identified. This microbial group played an essential role in Stage I, consuming a large part of the H₂ produced with acetic acid generation. These microorganisms act in synergy for the hydrolysis of SCB and H₂ and CH₄ production. Methanogenesis and homoacetogenesis play decisive roles in the efficiency of H₂ production; thus, further studies are needed to elucidate the interaction of these microbial groups in the degradation of cellulose wastes and in bioenergy production. Therefore, the employment of a consortium containing members from Clostridium, Bacillus, Bacteroides, and Paenibacillus genera (inoculum 2) for hydrogen production and a methanogenic consortium for subsequent methane production represents a promising insight into the anaerobic digestion of hydrothermal- and enzymatic-pretreated SCB, which provides an example of the production of biogas and value-added by-products from lignocellulosic biomass.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This project was funded by Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) process number 2013/20196-7 and 2009/15984-0.

References

1. B. Kamm and M. Kamm, Appl. Microbiol. Biotechnol., 2004, 64, 137–145.
2. Y. Lu, Q. Lai, C. Zhang, H. Zhao, K. Ma, X. Zhao, H. Chen, D. Liu and X. H. Xing, Bioresour. Technol., 2009, 100, 2889–2895.
3. S. K. Han and H. S. Shin, J. Air Waste Manage. Assoc., 2004, 54, 242–249.
4. D. Liu, D. Liu, R. J. Zeng and I. Angelidaki, Water Res., 2006, 40, 2230–2236.
5. V. Blonskaja, A. Menert and R. Vilu, Adv. Environ. Res., 2003, 7, 671–678.
6. C. Ting and D. Lee, Int. J. Hydrogen Energy, 2007, 32, 677–682.
7. G. Antonopoulou, H. N. Gavala, I. V. Skiadas, K. Angelopoulos and G. Lyberatos, Bioresour. Technol., 2008, 99, 110–119.
8. Y. C. Lo, G. D. Saratale, W. M. Chen, M. Der Bai and J. S. Chang, Enzyme Microb. Technol., 2009, 44, 417–425.
9. E. C. Koutrouli, H. Kalfas, H. N. Gavala, I. V. Skiadas, K. Stamatelatou and G. Lyberatos, Bioresour. Technol., 2009, 100, 3718–3723.
10. B.-B. Hu and M.-J. Zhu, Int. J. Hydrogen Energy, 2017, 42, 7866–7874.
11. H.-N. Lin, Y.-T. Wang and M.-J. Zhu, Int. J. Hydrogen Energy, 2017, 42, 26687–26694.
12. H. Zhu, A. Stadnyk, M. Béland and P. Seto, Bioresour. Technol., 2008, 99, 5078–5084.
13. R. P. Ratti, T. P. Delforno, I. K. Sakamoto and M. B. A. Varesche, Int. J. Hydrogen Energy, 2015, 40, 6296–6306.
14. J. K. Braga, L. A. Soares, F. Motteran, I. K. Sakamoto and M. B. A. Varesche, Int. J. Hydrogen Energy, 2016, 41(48), 22812–22823.
15. I. K. Kapdan and F. Kargi, Enzyme Microb. Technol., 2006, 38, 569–582.
16. L. A. Soares, J. K. Braga, F. Motteran, I. K. Sakamoto, E. L. Silva and M. B. A. Varesche, Water Sci. Technol., 2017, 76(1–2), 95–105.
17. H. Yokoi, S. Mori, J. Hirose, S. Hayashi and Y. Takasaki, Biotechnol. Lett., 1998, 20, 895–899.
18. H. L. Chin, Z. S. Chen and C. P. Chou, Biotechnol. Prog., 2003, 19, 383–388.
19. M. Ferchichi, E. Crabbe, G.-H. Gil, W. Hintz and A. Almadidy, J. Biotechnol., 2005, 120, 402–409.
20. G. Liu and J. Shen, J. Biosci. Bioeng., 2004, 98, 251–256.
21. J. K. Braga, A. A. Abreu, F. Motteran, M. A. Pereira, M. M. Alves and M. B. A. Varesche, Waste Biomass Valorization, 2017, 1–11.
22. C. A. B. S. Rabelo, L. A. Soares, I. K. Sakamoto, E. L. Silva and M. B. A. Varesche, J. Environ. Manage., 2018, 223, 952–963.
23. J. Cheng and M. Zhu, Bioresour. Technol., 2013, 144, 623–631.
24. N. Reddy and Y. Yang, Trends Biotechnol., 2005, 23, 22–27.
25. M. Balat, H. Balat and C. Öz, Prog. Energy Combust. Sci., 2008, 34, 551–573.
26. S. H. da Cruz, B. S. Dien, N. N. Nichols, B. C. Saha and M. A. Cotta, J. Ind. Microbiol. Biotechnol., 2012, 39, 439–447.
27. S. Kumar, U. Kothari, L. Kong, Y. Y. Lee and R. B. Gupta, Biomass Bioenergy, 2011, 35, 956–968.
28. S. G. Allen, D. Schulman, J. Lichwa, M. J. Antal, M. Laser and L. R. Lynd, Ind. Eng. Chem. Res., 2001, 40, 2934–2941.
29. Y. Gao, J. Xu, Y. Zhang, Q. Yu, Z. Yuan and Y. Liu, Bioresour. Technol., 2013, 144, 396–400.
30. H. Jørgensen, J. B. Kristensen and C. Felby, Biofuels, Bioprod. Biorefin., 2007, 1, 119–134.
31. B. C. Saha, N. Qureshi, G. J. Kennedy and M. A. Cotta, Int. Biodeterior. Biodegrad., 2016, 109, 29–35.
32. M. J. Taherzadeh and K. Karimi, BioResources, 2007, 2, 707–738.
33. C. N. Hamelinck, G. van Hooijdonk and A. P. Faaij, Biomass Bioenergy, 2005, 28, 384–410.
34. W. Wang, T. Yuan, K. Wang, B. Cui and Y. Dai, Bioresour. Technol., 2012, 107, 282–286.
35. C. L. Aguiar and T. J. B. Menezes, Bol. Cent. Pesqui. Process. Aliment., 2000, 18, 57–70.
36. T. Zhang and M.-J. Zhu, Bioresour. Technol., 2016, 214, 769–777.
37. L. A. Soares, J. K. Braga, F. Motteran, I. K. Sakamoto, P. A. S. Monteiro, P. Seleghim and M. B. A. Varesche, Waste Biomass Valorization, 2018, 1–14.
38. Z. Zhu, M. Zhu and Z. Wu, Bioresour. Technol., 2012, 119, 199–207.
39. F. Ahmad, Methane Production in Response to Sulfuric Acid and Hydrogen Peroxide Assisted Hydrothermal Pretreatment of Sugarcane Bagasse, São Carlos, 2017.
40. M. Mandels and J. Weber, Cellulases and Their Applications, Advances in Chemistry, 1969, ch. 23, vol. 95, pp. 391–414.
41. R. Atlas, Handbook of Microbiological Media, Boca Raton, 4th edn, 2010.
42. C. A. B. S. Rabelo, L. A. Soares and M. B. A. Varesche, Austin Clinical Microbiology, 2017, 2, 1–6.
43. F. Motteran, J. K. Braga, E. L. Silva and M. B. A. Varesche, J. Environ. Sci. Health, Part A: Environ. Sci. Eng., 2016, 51, 1288–1302.
44. E. D. Penteado, C. Z. Lazaro, I. K. Sakamoto and M. Zaiat, Int. J. Hydrogen Energy, 2013, 38, 6137–6145.
45. J. K. Braga and M. B. A. Varesche, Am. J. Anal. Chem., 2014, 05, 8–16.
46. APHA, Standard Methods for the Examination of Water and Wastewater, 20th edn, 2005.
47. M. Dubois, K. A. Gilles, J. K. Hamilton, P. A. Rebers and F. Smith, Anal. Chem., 1956, 28, 350–356.
48. M. Zwietering, I. Jongenburger, F. Rombouts and K. Van’t, Riet, 1990, 56, 1875–1881.
49. R. I. Griffiths, A. S. Whiteley, A. G. Oapos Donnell and M. J. Bailey, Appl. Environ. Microbiol., 2000, 66, 5488–5491.
50. U. Nübel, B. Engelen, A. Felske, J. Snaidr, A. Wieshuber, R. I. Amann, W. Ludwig and H. Backhaus, J. Bacteriol., 1996, 178, 5636–5643.
51. Y. Kudo, FEMS Microbiol. Ecol., 1997, 22, 39–48.
52. B. Si, Z. Liu, Y. Zhang, J. Li, X.-H. Xing, B. Li, N. Duan and H. Lu, Int. J. Hydrogen Energy, 2015, 40, 3191–3200.
53. Y. Y. Li, H. Sasaki, K. Yamashita, K. Seki and I. Kamigochi, Water Sci. Technol., 2002, 45, 143–150.
54. Y. Ueno, H. Fukui and M. Goto, Environ. Sci. Technol., 2007, 41, 1413–1419.
55. R. Chandra, H. Takeuchi and T. Hasegawa, Renewable Sustainable Energy Rev., 2012, 16, 1462–1476.
56. M. J. Taherzadeh and K. Karimi, Int. J. Mol. Sci., 2008, 9, 1621–1651.
57. L. Drake, H. L. Kusel and C. Matthies, in The Prokaryotes, New York, 2006, pp. 354–420.
58. G. Diekert and G. Wohlfarth, Antonie van Leeuwenhoek, 1994, 66, 209–221.
59. R. Conrad and M. Klose, FEMS Microbiol. Ecol., 1999, 30, 147–155.
60. H. H. Fang and X.-S. Jia, Water Res., 1999, 33, 1791–1798.
61. V. Siriwongrungson, R. J. Zeng and I. Angelidaki, Water Res., 2007, 41, 4204–4210.
62. J. J. Lay, Biotechnol. Bioeng., 2001, 74, 280–287.
63. N. Q. Ren, J. F. Xu, L. F. Gao, L. Xin, J. Qiu and D. X. Su, Int. J. Hydrogen Energy, 2010, 35, 2742–2746.
64. C. A. B. S. Rabelo, Avaliação de consórcio microbiano obtido a partir de lodo de estação de tratamento de efluentes de indústria de papel e celulose visando à produção de H₂ a partir de celulose, UNIVERSIDADE DE SÃO PAULO, 2014.
65. C. A. B. S. Rabelo, L. S. Botta and M. B. A. Varesche, Proc. XI DAAL - Lat. Am. Work. Symp. Anaerob. Dig, Havana, Cuba, 2014.
66. B. Xie, J. Cheng, J. Zhou, W. Song and K. Cen, Int. J. Hydrogen Energy, 2008, 33, 5006–5011.
67. R. P. Ratti, L. S. Botta, I. K. Sakamoto and M. B. A. Varesche, Int. J. Hydrogen Energy, 2013, 38, 9707–9717.
68. R. P. Ratti, L. S. Botta, I. K. Sakamoto, E. L. Silva and M. B. A. Varesche, Biotechnol. Lett., 2014, 36, 537–546.
69. M. Kim, C. Liu, J. Noh, Y. Yang, S. Oh, K. Shimizu, D. Lee and Z. Zhang, Int. J. Hydrogen Energy, 2013, 38, 8648–8656.
70. W. B. Whitman, D. R. Boone, Y. Koga and J. Keswani, in Bergey's manual of systematic bacteriology, ed. G. M. G. D. R. Boone and R. W. Castenholz, New York, 2nd edn, 2001, pp. 211–213.
71. Y. Sobrun, A. Bhaw-Luximon, D. Jhurry and D. Puchooa, Adv. Biosci. Biotechnol., 2012, 03, 398–407.
72. F. C. O. Gomes, C. L. C. Silva, C. R. Vianna, I. C. A. Lacerda, B. M. Borelli, Á. C. Nunes, G. R. Franco, M. M. Mourão and C. A. Rosa, Braz. J. Microbiol., 2010, 41, 486–492.
73. F. S. Moreira, R. G. Machado, B. B. Romão, F. R. X. Batista, J. S. Ferreira and V. L. Cardoso, Process Biochem., 2017, 58, 60–68.
74. H. J. Laanbroek, H. J. Geerligs, L. Sijtsma and H. Veldkamp, Appl. Environ. Microbiol., 1984, 47, 329–334.
75. M. Pester, N. Bittner, P. Deevong, M. Wagner and A. Loy, ISME J., 2010, 4, 1591–1602.
76. T. Aüllo, A. Ranchou-Peyruse, B. Ollivier and M. Magot, Front. Microbiol., 2013, 4, 362.
77. F. E. Mosey, Water Sci. Technol., 1983, 15, 209–232.
78. S. R. Harper and F. G. Pohland, in Microbiology and Biochemistry of Strict Anaerobes Involved in Interspecies Hydrogen Transfer, Springer US, Boston, MA, 1990, pp. 387–390.
79. R. P. Ratti, L. S. Botta, I. K. Sakamoto, E. L. Silva and M. B. A. Varesche, Biotechnol. Lett., 2014, 36, 537–546.
80. S. C. Santos, P. R. F. Rosa, I. K. Sakamoto, M. B. Amâncio Varesche and E. L. Silva, Int. J. Hydrogen Energy, 2014, 39, 9599–9610.
81. S. Venkata Mohan, L. Agarwal, G. Mohanakrishna, S. Srikanth, A. Kapley, H. J. Purohit and P. N. Sarma, Int. J. Hydrogen Energy, 2011, 36, 8234–8242.
82. J. M. Whiley and R. A. Hardie, in Bergey's Manual of Systematic Bacteriology The Firmicutes, ed. D. Jones, P. De Vos and G. M. Garrity, Springer US, New York, USA, 2nd edn, 2009, pp. 655–711.
83. M. Sakamoto, in The Prokaryotes, Springer Berlin Heidelberg, Berlin, Heidelberg, 2014, pp. 811–824.
84. C. R. Carere, R. Sparling, N. Cicek and D. B. Levin, Int. J. Mol. Sci., 2008, 9, 1342–1360.
85. P. L. Bergquist, M. D. Gibbs, D. D. Morris, V. S. J. Te’o, D. J. Saul and H. W. Morgan, FEMS Microbiol. Ecol., 1999, 28, 99–110.
86. C. Guevara and M. M. Zambrano, FEMS Microbiol. Lett., 2006, 255, 52–58.
87. A. Pourcher, L. Sutra, I. I. Hébé, G. Moguedet, C. Bollet, P. Simoneau and L. Gardan, FEMS Microbiol. Ecol., 2001, 34, 229–241.
88. J. Bérdy, J. Antibiot., 2005, 58, 1–26.
89. E. A. Barka, P. Vatsa, L. Sanchez, N. Gaveau-Vaillant, C. Jacquard, H.-P. Klenk, C. Clément, Y. Ouhdouch and G. P. van Wezel, Microbiol. Mol. Biol. Rev., 2016, 80, 1–43.
90. H. Yokoi, A. Saitsu, H. Uchida, J. Hirose, S. Hayashi and Y. Takasaki, J. Biosci. Bioeng., 2001, 91, 58–63.
91. C. Collet, Int. J. Hydrogen Energy, 2004, 29, 1479–1485.
92. D. Evvyernie, K. Morimoto, S. Karita, T. Kimura, K. Sakka and K. Ohmiya, J. Biosci. Bioeng., 2001, 91, 339–343.
93. C. C. Wang, C. W. Chang, C. P. Chu, D. J. Lee, B.-V. Chang, C. S. Liao and J. H. Tay, Water Res., 2003, 37, 2789–2793.
94. V. Peters, P. Janssen and R. Conrad, FEMS Microbiol. Ecol., 1998, 26, 317–324.
95. A. Stams, Antonie van Leeuwenhoek, 1994, 66, 271–294.
96. M. X. Fang, W. W. Zhang, Y. Z. Zhang, H. Q. Tan, X. Q. Zhang, M. Wu and X. F. Zhu, Int. J. Syst. Evol. Microbiol., 2012, 62, 3018–3023.
97. H. Jeong, Y. W. Lim, H. Yi, Y. Sekiguchi, Y. Kamagata and J. Chun, Int. J. Syst. Evol. Microbiol., 2007, 57, 1784–1787.
98. H. Shiratori, H. Ohiwa, H. Ikeno, S. Ayame, N. Kataoka, A. Miya, T. Beppu and K. Ueda, Int. J. Syst. Evol. Microbiol., 2008, 58, 964–969.
99. F. A. Rainey, in Bergey's Manual of Systematics of Archaea and Bacteria, John Wiley & Sons, Ltd, Chichester, UK, 2015, pp. 1–4.
100. C. D. Ogg and B. K. C. Patel, Int. J. Syst. Evol. Microbiol., 2009, 59, 95–101.
101. N. Klouche, M.-L. Fardeau, J.-F. Lascourreges, J.-L. Cayol, H. Hacene, P. Thomas and M. Magot, Int. J. Syst. Evol. Microbiol., 2007, 57, 1757–1761.
102. B. Fraj, W. Ben Hania, A. Postec, M. Hamdi, B. Ollivier and M.-L. Fardeau, Int. J. Syst. Evol. Microbiol., 2013, 63, 1947–1950.
103. C. D. Ogg and B. K. C. Patel, Int. J. Syst. Evol. Microbiol., 2010, 60, 1394–1400.
104. S. Chen, L. Niu and Y. Zhang, Int. J. Syst. Evol. Microbiol., 2010, 60, 2735–2738.
105. Y. Nakashimada, M. A. Rachman, T. Kakizono and N. Nishio, Int. J. Hydrogen Energy, 2002, 27, 1399–1405.
106. C. Long, J. Cui, Z. Liu, Y. Liu, M. Long and Z. Hu, Int. J. Hydrogen Energy, 2010, 35, 6657–6664.
107. A. Chatterjee, Y. Cui, Y. Liu, C. K. Dumenyo and A. K. Chatterjee, Appl. Environ. Microbiol., 1995, 61, 1959–1967.
108. A. Wang, L. Gao, N. Ren, J. Xu and C. Liu, Biotechnol. Lett., 2009, 31, 1321–1326.
109. B.-B. Hu and M.-J. Zhu, Microb. Cell Fact., 2017, 16, 77.
110. B.-B. Hu, M.-Y. Li, Y.-T. Wang and M.-J. Zhu, Chem. Eng. J., 2018, 335, 979–987.
111. Z. Pourramezan, G. R. Ghezelbash, B. Romani, S. Ziaei and A. Hedayatkhah, Microbiology, 2012, 81, 736–742.
112. C. M. Lo, Q. Zhang, N. V. Callow and L. K. Ju, Bioresour. Technol., 2010, 101, 717–723.
113. P. Rapp and F. Wagner, Appl. Environ. Microbiol., 1986, 51, 746–752.
114. S. A. Van Ooteghem, S. K. Beer and P. C. Yue, Appl. Biochem. Biotechnol., 2002, 98–100, 177–189.
115. J. Ravachol, R. Borne, I. Meynial-Salles, P. Soucaille, S. Pagès, C. Tardif and H.-P. Fierobe, Biotechnol. Biofuels, 2015, 8, 114.
116. S. Chen and X. Dong, Int. J. Syst. Evol. Microbiol., 2005, 55, 2257–2261.
117. B. Schink, Microbiol. Mol. Biol. Rev., 1997, 61, 262–280.
118. A. Söllinger, C. Schwab, T. Weinmaier, A. Loy, A. T. Tveit, C. Schleper and T. Urich, FEMS Microbiol. Ecol., 2016, 92, fiv149.
119. G. N. Jarvis, C. Strömpl, D. M. Burgess, L. C. Skillman, E. R. Moore and K. N. Joblin, Curr. Microbiol., 2000, 40, 327–332.
120. H. Imachi, S. Sakai, Y. Sekiguchi, S. Hanada, Y. Kamagata, A. Ohashi and H. Harada, Int. J. Syst. Evol. Microbiol., 2008, 58, 294–301.
121. M. S. M. Jetten, A. J. M. Stams and A. J. B. Zehnder, FEMS Microbiol. Lett., 1992, 88, 181–197.
122. L. Raskin, D. Zheng, M. E. Griffin, P. G. Stroot and P. Misra, Antonie van Leeuwenhoek, 1995, 68, 297–308.
123. H. Cadillo-Quiroz, P. Browne, N. Kyrpides, T. Woyke, L. Goodwin, C. Detter, J. B. Yavitt and S. H. Zinder, Genome Announcements, 2015, 3, e01280-1–5.
124. S. Kotelnikova, A. J. L. Macario and K. Pedersen, Int. J. Syst. Bacteriol., 1998, 48, 357–367.
125. G. Borrel, H. M. B. Harris, N. Parisot, N. Gaci, W. Tottey, A. Mihajlovski, J. Deane, S. Gribaldo, O. Bardot, E. Peyretaillade, P. Peyret, P. W. O'Toole and J.-F. Brugere, Genome Announcements, 2013, 1, e00453-13.
126. M. Simpson-Holley, A. Higson and G. Evans, J. Chem. Eng., 2007, 163, 46–59.

---