Population Dynamics of Methanogenic Archea in Co-Digestion Systems Operating Different Industrial Residues for Biogas Production

: This study aimed to analyze the population dynamics of methanogenic archaea in codigestion systems operated under different concentrations of industrial waste such as ricotta whey and brewery waste sludge in association with bovine manure. It was believed that the association of these residues from the food industry combined with bovine manure can contribute to improve the production of biogas. To identify the archaea, DNA extractions and the sequencing of the 16s rRNA gene were performed from 38 samples of inﬂuents and efﬂuents. The results indicated that Methanosaeta and Methanosarcina were predominant in the co-digestion of ricotta cheese whey and that Methanosaeta , Methanocorpusculum , and Methanobrevibacter prevailed in the co-digestion of residual brewery sludge. The three ricotta cheese whey biodigesters demonstrated efﬁciency in methane production; in contrast, residual sludge of brewery biodigesters only showed efﬁciency in the system operated with 20% co-substrate. of NH 3 -N during the bovine manure biodigestion in a full-scale biodigester. The RCW biodigesters showed close values for the ammoniacal nitrogen concentration of the digested.


Introduction
Anaerobic biodigestion (AB) is an important tool for the treatment of organic waste used to generate byproducts (biogas and biofertilizer) from substrate stabilization, contributing to the production of renewable energy [1][2][3][4]. AB is a metabolic process with multiple syntrophic relationships that occur between microorganisms in the steps of hydrolysis, acidogenesis, acetogenesis, and methanogenesis under specific environmental conditions [5,6].
Methanogenesis is a crucial step to methane (CH 4 ) production [7], and methanogenic archaea are the key microorganisms of this step. These species are divided into three groups, according to the substrate used for their growth. Hydrogenotrophic methanogens use hydrogen (H 2 ) and carbon dioxide (CO 2 ) as substrates; acetoclastic methanogens use acetate structures; and the methylotrophic methanogens use methyl groups [8].
The identification of these microorganisms in samples of biodigesters is possible through metagenomics analysis [9]. As a technique that is independent of cultivation, metagenomics allows the identification of the microbiological diversity from a complex environmental sample [10,11]. This is possible by the investigation of the hypervariable regions (V1 to V9) of the 16s rRNA of prokaryotes.
An increase in the number of studies involving methods for improving the AB was seen in recent years [12], with emphasis on anaerobic co-digestion (AcoD) [13]. AcoD is a

Experimental Design for the Analysis of Archaea Dynamics
The experiment lasted 165 days and was divided into three phases called competent inoculum (phase 1), acclimatization (phase 2), and anaerobic co-digestion (phase 3). The graphic scheme of the sample preparation is presented in Figure S1.
During the competent inoculum phase, seven biodigesters were completely supplied with diluted BM, reaching their full capacity (60 L). Flame tests were carried out daily after the start of biogas production. The collection of biogas for chromatographic analysis was performed after positive flame tests for biogas production. The substrate remained in the biodigester until the concentration of biogas in the system reached at least 60% CH 4 [25].
Subsequently, the acclimatization phase started. In this phase, one biodigester was selected as control and continued the supply exclusively with BM. Three biodigesters started the supply with the mixtures of RCW + BM and other three with mixtures of RSB + BM (Table 1). Biodigesters were identified according to the name and concentration of the co-substrate used in AcoD. The daily supply (2 L/day) was performed according to the values of the mixtures indicated in Table S1 with the supply, the same volume of effluent (2 L) was removed, keeping the system operating at its maximum capacity. The anaerobic co-digestion phase lasted 120 days. Daily supply continued according to the concentrations indicated in Table 1. The control biodigester worked only with BM, so the term co-digestion does not apply to it.

Analytical Techniques
Ammoniacal nitrogen (NH 3 -N) and pH analysis of effluents were performed every 2 weeks. The pH meter Tec-3MP (Tecnal, Piracicaba, Brazil) was used to measure the pH values of effluents, and NH 3 -N was determined according to the standard methods and methodologies [26].
The analysis of the biogas concentration was performed weekly to measure the CH 4 concentrations in biogas samples. These procedures were conducted by the Chromatography Laboratory at Embrapa using the Agilent 7820 A Chromatograph System and the EzChrom Elite interface software [27]. As indicated by Collins, et al., 1997 [27], a split-splinter type 50 [24], with adaptations.

Collection and Conservation of Samples
During the acclimatization phase, one influent sample was collected and after the CH 4 production (15 days), the effluent samples from the BM biodigester were collected. All effluent samples were collected on sampling valves. Before the samples were collected, the sampling valves were open for some minutes to flush out the valves. Thereafter, the effluent samples for microbial analysis were collected in sterile glass bottles and immediately sealed to maintain the anaerobic condition.
During the anaerobic co-digestion phase, influent samples were collected only for the biodigesters that operated in the co-digestion system (RSB20, RSB40, RSB80, RCW20, RCW40, and RCW80). In this phase, the collection of effluent samples occurred weekly until the end of the experiment.
After each collection, the samples were frozen at −80 • C. The samples were submitted to a freeze-drying process, which was conducted by the Chromatography Laboratory using the Liotop model L120 freeze dryer (Liobras, São Carlos, Brazil). After freeze-drying, all samples were stored in Falcon 50 mL conical tubes.

Sample Selection for Molecular Microbiology Analysis
To characterize the acclimatization phase, samples of the influent and effluent from the control biodigester were selected. From the material collected in the sampling valves, sample pooling was made to obtain greater representativeness of the content of the biodigesters. Therefore, each effluent collection had a single pooled sample representative of the biodigester instead of three samples. The preparation of sample pooling of effluent samples occurs from the lyophilized material.
Seventy mg of lyophilized effluent was weighed corresponding to the material collected from the V1 sampling valve and the same volume for the effluent of V2 and V3 valves. Then this material was homogenized and transferred to a microtube, representing a pooled sample with 210 mg of volume. The graphic scheme of the sample preparation is presented as a Figure S2.
For a better representation of this phase, the selection of at least one collection/month for each biodigester was prioritized. The effluent samples corresponding to the collections of days 1, 36, 78, 99, and 120 of the anaerobic co-digestion phase were chosen as representatives for molecular analysis. Sampling pooling of the selected effluent samples was also performed.
The characterization of the anaerobic co-digestion influents was also made through a lyophilized sample pooling. To characterize the RCW influents, were weighed 70 mg of the 20% RCW sample, 70 mg of the 40% RCW sample, and the same volume for the sample with 80% of RCW. Then this material was homogenized and transferred to a microtube, representing a pooled sample with 210 mg volume. The same procedure was carried out for RSB influents.

DNA Extraction
A pre-preparation of samples was made using enzyme lysis buffer (500 mM NaCl; 50 mM Tris-HCl, pH 8; 50 mM EDTA; 4% SDS) in association with mechanical lysis using the "Mini-Bead Beater 16" (BioSpec Products ® , Bartlesville, OK, USA) with zirconia beads (0.1 mm and 0.5 mm) [28,29]. Subsequently, the samples were submitted to the QIAamp DNA stool mini kit protocol (QIAGEN, Heidelberg, Germany), following the manufacturer's recommendations. The determination of concentration and purity of DNA was performed in NanoDrop™ 1000 Spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA).

Metagenomic Analysis-Sequencing and Bioinformatics
The extracted DNA was sent to a service provider company responsible for executing the sequencing and bioinformatics steps. For amplification of the polymorphic region (V4) of the 16S rRNA gene, PCR was conducted in triplicate using the following oligonucleotides: 515 F-'TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGGTGCCAGCMGCCGCGGT AA' and 806 R-'GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGGACTACHVGG GTWTCTAAT' [30,31]. The bioinformatics analysis was performed on the QIIME2 platform, version 2019.7 [32]. The sequences were filtered by quality and grouped into Operational Taxonomic Units (OTUs) using 97% of the identity between them. The sequences were also compared with the Silva 132 database [33] for taxonomic analysis. An OTU table selected by genus is presented as a Table S1.

Statistical Analysis
To investigate an association between the performance of the biodigesters and the composition of the archaeal microbiome, a principal component analysis (PCA) was performed. The PCA tried to identify if the selected physicochemical variables (concentration of CH 4 and NH 3 -N, and pH value) could justify the variation in the occurrence of OTUs in the biodigesters. This analysis was conducted on the statistical software JMP 14.0 (SAS Institute Inc., Cary, NC, USA).

Archaea Acclimatization
To analyze the structure of the archaeal community, the taxonomic classification of sequences was performed to the genus level. In the influent samples, the most abundant genus was Methanobrevibacter (Figure 1), showing a high relative abundance for the three samples. The relative frequency of Methanobrevibacter in the influent samples of BM ( Figure 1A), RCW ( Figure 1C), and RSB ( Figure 1D) were 94%, 95%, and 92%, respectively.
The abundance of Methanobrevibacter in the BM supply samples is in accordance with previous studies that mention it as the most profuse hydrogenotrophic methanogen in the bovine ruminal system [34][35][36]. Therefore, the high frequency of Methanobrevibacter in influent samples of RSB and RCW may be justified due to it being prepared with mixtures containing high amounts of bovine manure.
Senés-Guerrero, et al., 2019 [37], compared the abundance of this genus in rumen and biodigester samples, highlighting the taxon predominance in the rumen sample and its subsequent disappearance in the biodigester. Similarly, after 15 days of HRT from the acclimatization phase, the analysis of the BM effluent ( Figure 1B) had a reduction in the Methanobrevibacter frequency, from 94% in the initial sample to 3% in the digestate.
The reduction of this genus was followed by an increase in Methanosarcina, Methanocorpusculum, and Candidatus Methanoregula abundance. In influent samples, the genera Methanosarcina and Methanocorpusculum showed a relative abundance of 1% and 0.5%, respectively, while the genus Candidatus Methanoregula was not identified. After 15 days of HRT, the relative abundance of the genera Methanosarcina, Methanocorpusculum, and Candidatus Methanoregula increased to 64%, 25%, and 5.5%, respectively.
It was considered that the hydrogenotrophic pathway was predominant during the acclimatization phase. This is because the increase in the abundance of the facultative acetoclastic methanogen, Methanosarcina [38,39], occurred concomitantly with the increase of two exclusively hydrogenotrophic genera, Methanocorpusculum, and Candidatus Methanoregula [40,41]. According to Venkiteshwaran, et al., 2017 [39], the Methanosarcina genus has less affinity for acetate, reinforcing the idea that the hydrogenotrophic pathway was the propellant of methanogens during the acclimatization phase. Furthermore, previous stud- ies suggest that the hydrogenotrophic pathway is the main route for methane production from a biomass with bovine manure [42]. The abundance of Methanobrevibacter in the BM supply samples is in accordance with previous studies that mention it as the most profuse hydrogenotrophic methanogen in the bovine ruminal system [34][35][36]. Therefore, the high frequency of Methanobrevibacter in influent samples of RSB and RCW may be justified due to it being prepared with mixtures containing high amounts of bovine manure.
Senés-Guerrero, et al., 2019 [37], compared the abundance of this genus in rumen and biodigester samples, highlighting the taxon predominance in the rumen sample and its subsequent disappearance in the biodigester. Similarly, after 15 days of HRT from the acclimatization phase, the analysis of the BM effluent ( Figure 1B) had a reduction in the Methanobrevibacter frequency, from 94% in the initial sample to 3% in the digestate.
The reduction of this genus was followed by an increase in Methanosarcina, Methanocorpusculum, and Candidatus Methanoregula abundance. In influent samples, the genera Methanosarcina and Methanocorpusculum showed a relative abundance of 1% and 0.5%, respectively, while the genus Candidatus Methanoregula was not identified. After 15 days of HRT, the relative abundance of the genera Methanosarcina, Methanocorpusculum, and Candidatus Methanoregula increased to 64%, 25%, and 5.5%, respectively.
It was considered that the hydrogenotrophic pathway was predominant during the acclimatization phase. This is because the increase in the abundance of the facultative acetoclastic methanogen, Methanosarcina [38,39], occurred concomitantly with the increase of two exclusively hydrogenotrophic genera, Methanocorpusculum, and Candidatus Methanoregula [40,41]. According to Venkiteshwaran, et al., 2017 [39], the Methanosarcina genus has less affinity for acetate, reinforcing the idea that the hydrogenotrophic pathway was the propellant of methanogens during the acclimatization phase. Furthermore, previous studies suggest that the hydrogenotrophic pathway is the main route for methane production from a biomass with bovine manure [42].

Physicochemical Analysis of Effluents from the Anaerobic Co-Digestion Phase
The methane concentration (% v/v CH4) in biogas samples remained above 54% during the experiment, except for the RSB40 biodigester that declined the methane concentration from the 36th day of anaerobic co-digestion, reaching a concentration of

Physicochemical Analysis of Effluents from the Anaerobic Co-Digestion Phase
The methane concentration (% v/v CH 4 ) in biogas samples remained above 54% during the experiment, except for the RSB40 biodigester that declined the methane concentration from the 36th day of anaerobic co-digestion, reaching a concentration of 40% methane. The RSB80 biodigester declined from the start process of anaerobic co-digestion and stopped biogas production on the 99th day ( Figure 2A).
The pH did not show high variations and remained close to neutrality in all biodigesters, presenting an overall average of 7.1 (±0.24) ( Figure 2B). According to Lemmer et al., 2017 [49], it is possible to achieve the optimal methane yield keeping the pH value in the system between 6.8 and 7.2. From this result, it is possible to infer that the biodigesters showed favorable conditions for the survival of methanogenic archaea [50].
It is possible to observe in Figure 2C a gradual increase in the NH 3   The averages of the ammoniacal nitrogen concentrations were 460 mg/L NH 3 -N (±78) in RCW20, 394 mg/L NH 3 -N (±58) in RCW40, and 315 mg/L NH 3 -N (±85) in RCW80. The results differ from those reported by Comino, et al., 2012 [46], during the co-digestion of cattle slurry and cheese whey, in which values over 800 mg/L of NH 3 -N were found. This may have a relation to the type of cheese whey used and to the addition of limewater to stabilize the RCW pH [24]. The slight increase of NH 3 -N in BM and RCW biodigesters did not cause toxicity to the system since the increase of this compound did not affect methane production. The RSB biodigesters showed a high concentration of ammoniacal nitrogen (>700 mg/L NH 3 -N) from the first digested collection. The average concentration of RSB20 was 949 mg/L NH 3 -N (±116). 40% methane. The RSB80 biodigester declined from the start process of anaerobic codigestion and stopped biogas production on the 99th day (Figure 2A).   The RSB40 and RSB80 biodigesters showed the highest concentrations of ammonia nitrogen, with average values of 1331 mg/L NH 3 -N (±121) and 2952 mg/L NH 3 -N (±501), respectively. It is possible that this considerable increase in the ammoniacal nitrogen concentration is related to the decline in the methane concentration of these biodigesters.

Archaea Identification during Anaerobic Co-Digestion
The improvement in the generation of biogas rich in methane is the central point of AcoD processes, in which an appropriate selection of co-substrate and mixing ratio is necessary [51]. Inappropriate choices can lead to an imbalance in the system, limiting the methane generation. The results showed in Figure 3 reveal the dynamics of methanogenic microbiota in different biodigesters during the anaerobic co-digestion phase. A high abundance of Methanosarcina genus is also observed in the BM samples during the initial anaerobic co-digestion phase with a relative abundance of 76% on the first day ( Figure 3A). However, there is a notable reduction in the frequency of this genus during biodigestion, from 62% on the 36th day to 14% on the 120th day. Simultaneously, there is a considerable increase of Methanosaeta from 6% on the 36th day to 42% on the 99th day, which registered its highest frequency.  Methanosarcina and Methanosaeta are responsible for producing a large part of the CH4 generated during anaerobic digestion [8]. Individuals of the Methanosaeta genus are mandatory acetoclastic methanogens, able to contribute to more than 60% of CH4 generated through the oxidation of acetic acid [52][53][54]. This variation in the prevalence of Methanosarcina is consistent with the results of Song, et al., 2015 [55], observing the codigestion of a pretreated wheat straw with cattle manure, which shows a higher frequency of this group in the initial period of biodigestion and its subsequent decline. The expressive increase in the relative abundance of Methanosaeta may be indicative of its affinity for acetate, suggesting the acetoclastic pathway as the propulsive pathway of BM methanogenesis. However, with the increase of Methanosaeta, there is a slow increase of some hydrogenotrophic genera (Candidatus Methanoregula, Methanocorpusculum, and Methanospirillum). The increase of these genera, concomitant with the increase of Methanosaeta, may be evidence of the coexistence of hydrogenotrophic and acetoclastic Methanosarcina and Methanosaeta are responsible for producing a large part of the CH 4 generated during anaerobic digestion [8]. Individuals of the Methanosaeta genus are mandatory acetoclastic methanogens, able to contribute to more than 60% of CH 4 generated through the oxidation of acetic acid [52][53][54]. This variation in the prevalence of Methanosarcina is consistent with the results of Song, et al., 2015 [55], observing the codigestion of a pretreated wheat straw with cattle manure, which shows a higher frequency of this group in the initial period of biodigestion and its subsequent decline. The expressive increase in the relative abundance of Methanosaeta may be indicative of its affinity for acetate, suggesting the acetoclastic pathway as the propulsive pathway of BM methanogenesis. However, with the increase of Methanosaeta, there is a slow increase of some hydrogenotrophic genera (Candidatus Methanoregula, Methanocorpusculum, and Methanospir-illum). The increase of these genera, concomitant with the increase of Methanosaeta, may be evidence of the coexistence of hydrogenotrophic and acetoclastic pathways [54]. During the analysis of wetland samples, Zhang, et al., 2019 [54], found evidence that these two methanogenic pathways can coexist although one pathway may be more significant for methane production than the other one.
In this case, it was not possible to stipulate what was the exact propulsive pathway for methanogenesis from the 78th day on the BM biodigester, although it is believed that acetogenesis may have prevailed, based on the high abundance of the Methanosaeta genus. It was also noted that the RCW archaeal microbiota was quite similar to that found in the BM biodigester ( Figure 3A-C). The relative abundance of Methanosarcina on the first day of anaerobic co-digestion phase was 66% for RCW20, 66% for RCW40, and 83% for RCW80.
As in the BM biodigester, the RCW20 and RCW40 biodigesters presented an increase in the relative abundance of Methanosaeta along with a slight increase of hydrogenotrophic methanogens during co-digestion.
This coexistence between acetoclastic and hydrogenotrophic methanogens again raises the question about which pathway promotes the methanogenesis in RCW20 ( Figure 3B) and RCW40 ( Figure 3C) co-digestion. Differently, the RCW80 ( Figure 3D) biodigester did not show a considerable increase in hydrogenotrophic methanogens. However, the genus Methanosaeta was dominant during anaerobic co-digestion, reaching 66% relative abundance. Chen, et al., 2015 [56], demonstrated the prevalence of Methanosaeta over Methanosarcina in mesophilic biodigesters supplied with dairy residues under high organic loading rates. Saha, et al., 2019 [57], described similar results during the digestion of sludge obtained in a WWTP and augmented with a mixed waste of fruit, where the prevalence of Methanosaeta over Methanosarcina was observed. However, Chen, et al., 2017 [58], demonstrated that the prevalence of Methanosaeta can occur due to the competitiveness of this genus under high acetate concentrations.
Based on the results, it is assumed that acetoclastic methanogenesis prevailed during the AcoD of ricotta cheese whey due to the abundance of acetoclastic over methanogenic genera. A deep study of the bacterial metagenome that was found in this study may provide us with a more effective answer on this issue. The biodigesters that used RSB as AcoD substrate differed from each other and the BM biodigester. Methanosarcina and Methanocorpusculum were the main representatives of the archaeal community in RSB20 ( Figure 3E), indicating that these two genera led the methane production. The co-occurrence of these two genera suggests that the biodigester RSB20 followed the hydrogenotrophic pathway for methane production.
A noticeable change in the dynamics of archaeal microbiota was observed in the RSB40 biodigester ( Figure 3F). The beginning of RSB40 AcoD is dominated by members of Methanosarcina and Methanocorpusculum genera with 55% and 32% of relative abundance respectively. Subsequently, there was a reduction in the abundance of these two genera and prevalence of Methanobrevibacter genus. Methanobrevibacter was also abundant in the RSB80 biodigester ( Figure 3G). Associated with the increase of this taxon, it is also possible to observe a slight increase of Methanosphaera genus, a hydrogenotrophic methanogen equally abundant in the gastrointestinal tract of ruminants [35,59]. Principal component analysis ( Figure 4) revealed that these two genera, Methanobrevibacter and Methanosphaera, are strongly related to the increase of the NH 3 −N concentration.
Although ammoniacal nitrogen is crucial to microbial growth when in high concentration, it is considered an inhibitor of the system [60]. The NH 3 −N increase is important evidence to justify the discrepant occurrence of Methanobrevibacter in RSB40 and RSB80 samples, followed by a reduction in the relative abundance of the other genera. High concentrations of this compound have been associated with Methanobrevibacter survival in previous studies [61,62].
Nevertheless, our results support those of Nguyen, et al., 2019 [8], who highlighted acetoclastic methanogens, the main responsible for methane production, as vulnerable to NH 3 −N increase. The predominance of hydrogenotrophic methanogen under high NH 3 −N concentration may be considered as a response to the change from acetoclastic methanogenesis to hydrogenotrophic methanogenesis [63]. abundance respectively. Subsequently, there was a reduction in the ab two genera and prevalence of Methanobrevibacter genus. Methanobre abundant in the RSB80 biodigester ( Figure 3G). Associated with the incr it is also possible to observe a slight increase of Methanosphaera genus, a h methanogen equally abundant in the gastrointestinal tract of ruminants component analysis (Figure 4) revealed that these two genera, Metha Methanosphaera, are strongly related to the increase of the NH3−N conce Although ammoniacal nitrogen is crucial to microbial growt concentration, it is considered an inhibitor of the system [60]. The N important evidence to justify the discrepant occurrence of Methanobreviba RSB80 samples, followed by a reduction in the relative abundance of Probably, the interruption in the production of biogas by the RSB80 biodigester before the end of the experiment is due to the increase of NH 3 −N in the system, once high ammoniacal nitrogen values can also prevent the growth of microbiota, affecting the AcoD process [64,65]. The PCA (Figure 4) also revealed that the genus Methanocorpusculum has a connection with the pH value. Zhou, et al., 2016 [66], point Methanocospusculum as the main genus in laboratory-scale mesophilic reactors adjusted to pH 7. However, despite the PCA result, our study does not provide enough evidence to prove a relationship between the occurrence of Methanocospusculum and the pH variation. A heatmap ( Figure 5) was created for a better visualization of the genera that dominated each biodigester. Therefore, it was observed that BM and RCW biodigestion is dominated by the genera Methanosarcina and Methanosaeta. RSB biodigestion, on the other hand, presents the genera Methanosarcina, Methanocorpusculum, and Methanobrevibacter as the propellants of methanogenesis. An OTU table selected by genus is presented in Table S1. Dhungana, et al., 2022 [67], studied the anaerobic co-digestion process, seeking to optimize to increase the recovery of resources from organic waste. The results indicate that the use of an optimized mixture of co-substrates together with the summer starter can remarkably improve energy generation; these results corroborate what we also have in our findings, where the co-substrate, in the case of RCW in the mixture was beneficial in the stability and productivity of the AcoD bioprocess.
between the occurrence of Methanocospusculum and the pH variation. A heatmap ( Figure  5) was created for a better visualization of the genera that dominated each biodigester. Therefore, it was observed that BM and RCW biodigestion is dominated by the genera Methanosarcina and Methanosaeta. RSB biodigestion, on the other hand, presents the genera Methanosarcina, Methanocorpusculum, and Methanobrevibacter as the propellants of methanogenesis. An OTU table selected by genus is presented in Table S1. Dhungana, et al., 2022 [67], studied the anaerobic co-digestion process, seeking to optimize to increase the recovery of resources from organic waste. The results indicate that the use of an optimized mixture of co-substrates together with the summer starter can remarkably improve energy generation; these results corroborate what we also have in our findings, where the co-substrate, in the case of RCW in the mixture was beneficial in the stability and productivity of the AcoD bioprocess.

Conclusions
It was observed that biodigestion of bovine manure may assume the acetoclastic and hydrogenotrophic pathway and sometimes this path can coexist. The three treatments containing RCW showed good results in the production of CH4, proving its efficiency in mixtures containing up to 80% of it. Methanosarcina and Methanosaeta genera are the most abundant in the RCW systems. The AcoD of the RSB is mostly conducted by hydrogenotrophic archaea, especially the Methanocorpusculum and Methanobrevibacter genera, and the facultative Methanosarcina, indicating that the hydrogenotrophic pathway is dominated the CH4 production. The RCW biodigesters have shown better results than the RSB biodigesters. The biodigester that worked with 20% RSB showed a satisfactory result, indicating a possibility in the use of small proportions of this residue in the AcoD process. E-supplementary data of this work can be found in the online version of the paper.

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/su141811536/s1, Figure S1: Experimental design of the project. Biodigesters named according to the co-substrate concentration used; Figure S2: Scheme of sample pooling preparation for DNA extraction. (A) Pooling of lyophilized effluents from 70 mg of the sample collected in each biodigester valve (V1, V2, and V3). (B) Pooling of lyophilized influents from 70 mg of each supply mixture sample (20%, 40%, and 80%). Table S1: OTU table classified by genus level presenting the most abundant genera in influent and effluent samples. The column classified as "Others" contains OTUs that were not possible to be classified by gender or those with small numbers of representatives.