H₂S Emission and Microbial Community of Chicken Manure and Vegetable Waste in Anaerobic Digestion: A Comparative Study

Abstract

In order to solve the problem of H₂S corrosion in biogas utilization, it is necessary to understand the characteristics and mechanisms of H₂S production in chicken manure anaerobic digestion (CMAD) and vegetable waste anaerobic digestion (VWAD). In this study, lab-scale batch tests of CMAD and VWAD were conducted for 67 days at 35 °C. The results showed that sulfide was found to be the major form of sulfur in CMAD (accounting for 90%) and VWAD (70%). The average concentration of H₂S was 198 ± 79 ppm in CMAD and 738 ± 210 ppm in VWAD. Moreover, 81% of total H₂S was produced at 20 days of methane production in CMAD, but 80% of total H₂S was produced in the first day in VWAD because of the rapid production of biogas and fermentation acidification. The sulfide ion equilibrium model could universally and feasibly predict the H₂S production in CMAD and VWAD. The abundance of Firmicutes, Bacteroidetes, Proteobacteria and Euryarchaeota accounted for about 95% of the total microbes in both CMAD and VWAD; the influence of the fermentation stage on the microbial community was greater than that of the difference between CM and VW; the abundance of SRB was 0.01~0.07%, while that concerning organosulfur compounds fermentation was 22.8~30.5%. This study indicated that the H₂S concentration of CMAD biogas was more than five times that of VWAD because CM is alkalescent but VW is acidic.

1. Introduction

Copious amounts of chicken manure (CM) and vegetable waste (VW) are produced in China, reaching 100–210 million tons and 269 million tons, respectively, in 2015 [1,2]. Improper treatment could result in a foul odor, exposure to toxins and pathogens, etc. Anaerobic digestion (AD) technology has been described as sustainable and is widely applied in the recycling of biogas and fertilizer from CM and VW. The corrosive character of the H₂S in biogas is detrimental with respect to using such biogas in engines or boilers. It could be sensible to also mention its influence on systems which are used for upgrading the biogas to biomethane. Moreover, one of the future ways of using biogas could be its reforming and use for subsequent synthesis. For such novel applications, H₂S is very problematic due to its detrimental influence to catalysts [3,4,5]. These issues could potentially be overcome by the in situ extraction of problematic substances during AD. Therefore, obtaining further knowledge on the exact processes responsible for the production of these odorants is critical.

Numerous studies of the characteristics of odors produced during the AD of municipal sludge were conducted in the past two decades, thereby identifying H₂S as an important contributor of odors and corrosion. Municipal sludge AD typically results in the production of 500–4000 ppm H₂S, a byproduct of the hydrolysis of organic sulfurs and reduction of inorganic sulfates [6,7,8]. Numerous studies showed that there were two biochemical ways to form sulfide (including S²⁻, HS⁻, and H₂S) in anaerobic microbial process: one is the reduction of sulfur to an oxidation state such as sulfate by sulfate reduction bacteria (SRB), which can compete with methanogens for substrates such as H₂ and acetate during AD; the other is the decomposition of organic sulfur such as protein, methionine, and methyl mercaptan by bacteria and archaea [9,10,11,12].

However, studies on the H₂S characteristics of the AD process targeting VW and CM are limited. The differences in metabolism in the anaerobic digestion of CMAD and VWAD, with respect to H₂S generation, are not clearly understood. To answer the issues above, in the present study, the abiotic factors and microbial community were detected to compare the characteristics and metabolism of H₂S emission in CMAD with those in VWAD.

2. Materials and Methods

2.1. Experiment Design and Sample Preparation

The present study fermented two kinds of raw materials for batch processing (35 °C), namely, chicken manure (CM) and vegetable waste (VW). Three replicates (three batch biogas reactors) were designed for each raw material. Figure 1 shows the schematics of the reactor. The metering bottles (Shuniu Glass Co., Ltd., Chengdu, China), biogas producer(Shuniu Glass Co., Ltd., Chengdu, China), and storage(Shuniu Glass Co., Ltd., Chengdu, China) all have an equal working volume of 1 L. The volume of fermentation liquor was 0.8 L, its concentration was 5.9%, and the inoculum to raw material total solids (TS) ratio was 2:1. CM was sourced from a chicken farm in Cixi City, Zhejiang Province, China; VW was sourced from a farmers’ market on a university campus in Beijing, China. The inoculum was naturalized using raw materials (CM and VW) over a period of 40 days. The profiles of the raw material and inoculum are reported in

Table 1. Characteristics of chicken manure (CM), vegetable waste (VW), and inoculum.

Items	CM	VW	Inoculum
TS (%)	27.1 ± 1.7	7.2 ± 0.4	5.2 ± 0.3
VS (%) (dry basis)	72.3 ± 2.5	87.5 ± 1.5	56.5 ± 2.2
Ash (%) (dry basis)	27.7 ± 1.6	12.5 ± 2.2	43.5 ± 2.1
ORP (mV)	−327 ± 19	152 ± 15	−279 ± 12
DO (mg/L)	0.15 ± 0.10	7.37 ± 1	0.16 ± 0.11
pH	7.37 ± 0.16	5.08 ± 0.22	8.5 ± 0.10
Protein (%) (dry basis)	3.9 ± 0.5	11.4 ± 1.3	1.5 ± 0.15
S-sulfide (mg/kg) (dry basis)	2566 ± 354	264 ± 14	8750 ± 279
S-sulfate (mg/kg) (dry basis)	1922 ± 272	1527 ± 220	653 ± 37
S-total (mg/kg) (dry basis)	7509 ± 480	3152 ± 317	9576 ± 643
S-protein (mg/kg) (dry basis)	388	1139	19

. H₂S concentrations and biogas yield were measured each day during peak biogas production and every four days during non-peak production. On days 0, 6, 17, 39, and 67, sludge specimens from the biogas fermenter were even moved into sterilization tubes(Anhui Vincennes Medical Device Co., Ltd., Hefei, China). Samples were stored at −20 °C until extraction of DNA, following which the abiotic factors were determined.

2.2. Determination of Abiotic Factors

The H₂S concentrations in biogas were measured using gas chromatography with a Fission-Product Detector (FPD) (GC-14B, Shimadzu Co., Ltd., Kyoto, Japan); the FPD, inject port, and chromatographic column were all set to 240 °C, 100 °C, and 60 °C, respectively; and the carrier gas was nitrogen. CH₄ and CO₂ were measured with gas chromatography using a Thermal Conductivity Detector (TCD) (GC-14C, Shimadzu Co., Ltd., Kyoto, Japan); inject port, chromatographic column (5 A molecular sieve), and TCD were all adjusted at 60 °C, 60 °C, and 100 °C, respectively; and argon was used as the carrier gas. Sulfate was tested with ion chromatography (type: INTEGRION-HPIC; detector: conductance; separation column: IonPac AS11/HC; guard column: IonPac AG11/HC; leachate: 30 mmol/L KOH; flow velocity: 1 mL/min; current of suppressor: 75 mA; sample size: 25 μL). Using methylene blue spectrophotometry, total sulfides (included S²⁻, HS⁻, and H₂S in fermentation sediment and liquor) and supernatant sulfides were tested. Supernatant sulfide is the sulfide in the upper layer of fermentation broth remaining in centrifugal tube after 5 min. An elemental analyzer was used to test the total sulfur (S-total) by following the manufacturer’s instructions (Flashsmart, Thermo Fisher SCIENTIFIC, Waltham, MA, USA). A Multimeter HACH instrument (HQ40d, HACH)(HACH Water Analysis Instrument (Shanghai) Co., Ltd., Shanghai, China) was used to measure the amount of dissolved oxygen (DO), pH, and oxidoreduction potential (ORP).

2.3. Calculation Formula

2.3.1. H₂S Production Prediction Calculation Formula

The prediction of H₂S production is based on the sulfide equilibrium model (Supplementary File S1), and the key formulas include:

$$C_{H2S}=34S_T/(32\left(1+\frac{K_{S1}}{{10}^{-pH}}+\frac{K_{S1}{K}_{S2}}{{10}^{-2pH}}\right))$$

(1)

In Equation (1) C_H2S represents the H₂S concentration in fermentation liquid (mg/m³), Ks₂ = 1.3 × 10⁻⁷ and Ks₁ = 7.1 × 10⁻¹⁵ are the secondary and primary H₂S ionization constants respectively, and S_T represents dissolved sulfide (mg/m³).

$$C=E{C}_{H2S}$$

(2)

$$P=C{P}_{Biogas}$$

(3)

In Equations (2) and (3), C is the H₂S concentration in biogas (mg/m³), E represents Henry’s constant (0.686 kPa at 35 °C), P (uL/day) represents the daily production of H₂S, and P_Biogas (mL/day) stands for the daily production of biogas.

2.3.2. Formulas for Biogas and H₂S Production Potential

Biogas yield potential was counted based on following formula, which was obtained based on the experience of many batch fermentation studies by others and ourselves [13]:

$$A=(B-C-D)/E$$

(4)

where A: Biogas yield potential of CM/VW; B: biogas cumulant of CM/VW; C: biogas cumulant of inoculum; D: biogas cumulant of samples of sludge removed; E: total solid of CM/VW.

The potentials of H₂S yields of CM/VW are counted based on the following formula:

$$A=({\displaystyle \sum }_{i=0}^n{a}_i\times b_i-{\displaystyle \sum }_{i=0}^n{c}_i\times d_i)/E$$

(5)

where A: H₂S production potential of CM/VW; a_i: biogas production of CM/VW at day i; b_i: H₂S concentration of CM/VW at day i; c_i: inoculum biogas yield at day i; d_i: inoculum H₂S concentration at day i; E: CM/VW total solids. (Note: The samples of the H₂S concentration test were collected from the gas bomb.)

2.4. Characterization of the Microbial Community

Before DNA extraction and high-throughput sequencing, the specimens from reactor replicates were intermingled. Genomic DNA was extracted from samples using cetyltrimethylammonium bromide. The primers (341F (5′-CCTAYGGGRBGCASC AG-3′)/806R (5′-GGACTACNNGGGT ATCTA AT-3′)) were used to amplify the V3–V4 regions of the archaea and bacteria 16S genes [14]. Different samples were identified in labels using barcode sequences. Polymerase chain reaction (PCR) of DNA was performed by sequencing company (Novo gene Biotechnology Co., Ltd. Beijing, China). The PCR products were mixed to the equidensity ratio. The sequencing libraries were generated guided by the manufacturer product description, followeing which the identity codes of each sequence were added. To generate superior-quality data, an Illumina Sequencer called Hiseq PE2500 was utilized, which was subsequently processed using the Qiime2 software program [15]. Cutadapt was used to trim adapter sequences and barcodes [16], which were then truncated at 210 bp and 220 bp, respectively, for reverse and forward reads, merged with 20 bp overlap in DADA2 to form ASV table [17]. At 97 percent similarity to the Naïve Bayes approach, the Greengene database was used for taxonomy classification [18]. For functional predictions, the FAPROTAX database was utilized [19], with vsearch [20] clustering ASVs at 97 percent against the Greengene database used to create taxonomy.

3. Results and Discussion

3.1. Characteristics of CM, VW, and Inoculum

Table 1 shows that sulfide and sulfate were the main forms of sulfur in CM, where S-sulfide accounted for 34.2% of S-total and S-sulfate accounted for 25.6%, while S-protein accounted for 5.2%; CM was weakly alkaline (pH 7.37). Sulfate and protein were the main chemical forms of sulfur in VW, where S-sulfate accounted for 48.5% of S-total and S-protein accounted for 36.1%, while sulfide levels were observed to be very low and were thus ignored due to the extremely high ORP (152 mV) and DO (7.37 mg/L); VW was acidic (pH 5.08). A total of 88.0% of S-total was accounted for by inoculum sulfide due to the lower and middle layers being in a strictly anaerobic state (ORP: −291 mV).

Based on the analysis above, and independent of focusing on CMAD or VWAD, sulfide was the major form of sulfur at the starting time because the inoculum took up two-thirds of the total solid in the fermentation liquor.

3.2. Production Characteristics of Biogas, CH₄, and H₂S

3.2.1. Biogas and CH₄

For CMAD, one biogas production peak appeared from day 12 to day 36 and the CH₄ content of 45% on day 12 increased to 75% on day 32, with an average of 54.9% (Figure 2a–d and

Table 2. Production potential and concentration of biogas and H₂S.

	CM	VW
Biogas production potential (m3/ton (TS))	314 + 92	240 + 78
Average methane content (%)	54.9 + 3.1	49.7 + 2.2
H2S production potential (g/ton (TS))	90 ± 37	288 ± 87
Average H2S concentration (ppm)	198 + 79	738 ± 210

). The biogas production potential of CM was 314 m³/t (TS) (Table 2).

For VWAD, two biogas production peaks appeared (Figure 2a). The first one was observed on day 1, whereas the second peak was observed on day 52 and disappeared on day 60 (Figure 2a,b). Due to acidification (Figure 3a), the content of CH₄ fell to less than 5% on day 1, following which it grew from 45% on day 20 to 70% on day 60 (Figure 2c,d). In addition, the content of H₂ on day 1 was 30%, following which it declined to 0.5% after day 2. Therefore, the biogas’s average CH₄ content was calculated to be 49.7% (Table 2). The biogas production potential of VW was 240 m³/t (TS) (Table 2).

The inoculum generated small amounts of biogas (Figure 2a–d), indicating the incomplete digestion of organic matter prior to measurements.

Based on the analysis above, the fermentation time of CMAD was shorter than that of VWAD, and the biogas production potential and CH₄ content of CM were higher than those of VW.

3.2.2. H₂S Emission

H₂S concentrations for CMAD were between 156 and 492 ppm from day 1 to day 42, and there was a decline in the final concentrations to 72–112 ppm between days 46 and 67 (Figure 2e). The average concentration and production potential of H₂S were calculated as 198 ± 79 ppm and 90 ± 37 g/t (TS), respectively (Table 2). Small amounts of H₂S were produced at day 1, and most H₂S was produced when the CH₄ production peak appeared between day 12 and 36 (Figure 2a,c,e), which means the production of H₂S was synchronous to that of CH₄.

A general declining trend of H₂S concentrations in VWAD was evident, with a particularly pronounced decline from 1960 ppm on day 1 to 280 ppm on day 4, with the final concentrations reaching 132–230 ppm between days 8 and 67 (Figure 2e). The H₂S average concentration and production potential were calculated to be 738 ± 210 ppm and 288 ± 87 g/t (TS), respectively (see Table 2). On day 1, 80% of the H₂S was produced, whereas 20% was produced during days 2 and 67 (Figure 2e). This result was similar to those of previous studies on anaerobic digestion of food waste. For example, H₂S concentrations were measured within full-scale AD at 600–1500 ppm by Kang et al. [21] and Cen et al. [22].

H₂S concentrations of the inoculum control were 112–311 ppm at the peak biogas production time (days 1 to 28). There was a subsequent decline in concentrations to 10 ppm on day 32, following which they stabilized at a light concentration between days 36 and 67.

The results showed that 80% of H₂S was released during the CH₄ yield peak in CMAD, and the production of H₂S was synchronous to that of CH₄. For VWAD, the majority of the production of H₂S occurred during the initial fermentation stage, and the peak production times of H₂S and CH₄ were not synchronous. The average H₂S concentrations in CMAD were substantially lower than those in VWAD, and the H₂S production potential of CM was substantially lower than that of VW.

3.3. The Relationship between H₂S, Sulfide, and pH

For CMAD, the pH increased gradually from 7.3 on day 0 to 8.9 on day 67 (Figure 3a). The supernatant sulfide concentration dropped dramatically from 95 mg/L on day 0 to 23 mg/L on day 6, following which it declined to 69 mg/L on day 17, and then fell to 19 mg/L on day 39, and grew again to 71 mg/L on day 67. The results of present study showed that about 1/25 of the supernatant sulfide in CMAD can be transformed into soluble sulfide.

Based on the pH values and the soluble sulfide concentration mentioned in the last paragraph, and using the sulfide equilibrium equations Equations (1)–(3), H₂S production can be estimated. Figure 3c shows the estimated values of H₂S production. In general, there was consensus between the predicted and measured values. The equilibrium model in particular predicted 88% of H₂S from day 1 to 28; one-variable linear regression analysis showed R² was 0.718 between actual value and predicted value from day 1 to day 68. It is worth noting that the sulfur in H₂S only accounted for 1.8% of the total sulfide, i.e., only a small amount of sulfide in the liquor phase was transferred to biogas in CMAD. Different to our study model, Peu et al.’s [9] study showed that the H₂S emission model in food waste AD was established based on ratio of sulfur and carbon, but this study did not provide an accurate proportion of sulfur’s different chemical forms.

For VWAD, there was a sharp decline in pH from 7.5 on day 0 to 6.3 on day 1, following which there was an increase from 6.7 on day 6 to 8.6 on day 67 (Figure 3a). The supernatant sulfide concentration fell dramatically from 97 mg/L on day 0 to 27 mg/L on day 6, following which it declined and stabilized at 66–72 mg/L between day 17 and 39 and decreased again to 30 mg/L on day 67. According to Table 1 and the TS ratio of VW to inoculum (1:2), the sulfide concentration of VWAD was 329 mg/L; thus, the proportion of the supernatant sulfide in relation to the total sulfide was between 8.2 and 29.4%. This result is similar to Belle et al.’s study [23]: their research found that the concentration of H₂S in the biogas peaked at 3300 ppm on the first day of fermentation, while the pH dropped to 6.8; however, by the sixth day, the concentration of H₂S dropped by 1300 ppm, while the pH increased to 7.5.

Based on the pH values and soluble sulfide concentration from last paragraph, using Equations (1)–(3), the H₂S production can be estimated. Figure 3d shows that the predicted and measured values were similar, with 97% of H₂S on day 1 predicted by sulfide ion equilibrium model; one-variable linear regression analysis showed R² was 0.973 between the actual value and predicted value from day 1 to day 68. This result indicated that the majority of H₂S produced on day 1 was a result of the high supernatant sulfide concentration due to the acidification of the system and the rapid production of biogas. It is worth noting that the sulfur of H₂S accounted for only 3.6% of the total sulfide, i.e., only a small amount of sulfide in the liquor phase was transferred to biogas in CMAD and VWAD.

Based on the analysis above, there were very high concentrations of sulfide in CMAD and VWAD, and only a small amount of sulfide in the liquor phase was transferred to biogas. The characteristics of H₂S production between CMAD and VWAD differed, but the sulfide ion equilibrium mode could accurately predict the H₂S production process in AD with these two materials. The rate of biogas production and pH were the key factors for the release of H₂S. The emissions of H₂S observed on day 1 were not the result of biochemical processes, but rather physical and chemical processes.

3.4. Microbial Community Structure and Function

3.4.1. Microbial Community Structure at Phylum and OTUs Levels

As can be seen from Figure 3a, no matter CMAD or VWAD, the main microbial groups were Firmicutes, Bacteroidetes, Proteobacteria, and Euryarchaeota at the phylum level; these four phyla accounted for about 95% of the total microbes.

For CMAD (see Figure 4a,b), the abundance of Firmicutes, which contained genera (such as Clostridium) which can hydrolyze cellulose and hemicellulose [24,25], accounted for 72% at the initial period, then gradually declined by 47% at 39 d; it reamined stable until 67 d. The abundance of Bacteroidetes, which contains genera (such as Galbibacter) which can ferment glucose and produce H₂/acetate [26], gradually increased from 18% at 0 d to 33% at 39 d; it remained at this level until 67 d. The abundance of Proteobacteria, which contains genera (such as Desulfomicrobium) which can reduce sulfate [27], was 7%, fell to 2% at 17 d, and was between 1.5 and 2.5% at from 39d to 67d. The abundance of Euryarchaeota, which contains genera (such as Methanosarsina, Methanobrevibacter) which can produce methane [28,29], was less than 1% at first 17 days, increased to 12% at 39 d, and then gradually decreased to 2% at 67 d. It is worth noting that the characteristics of the microbial community in the inoculum are similar to that of CM.

For VWAD (see Figure 4a,b), the abundance of Firmicutes was between 60 and 68% at a high abundance level before 17 d, indicating that Firmicutes was relatively adaptable to the acidic environment (pH = 6.5); it dropped to 47% at 39 d and was back to 71% at 67 d. The abundance of Bacteroidetes did not change much between 18 and 25% over the whole fermentation process. The abundance of Euryarchaeota was 6~7% before 17 d and decreased to 3% between 39 and 67 d. The abundance of Euryarchaeota decreased from 2% at the beginning to 0.2% at 17 d, increased to 18% in the early stage of the methane production peak (39 d), and decreased to 1% at the end of the CH₄ production peak (67 d).

3.4.2. Principal Component Analysis (PCA)

As can be seen from Figure 4b, during the intial stages of fermentation, the microbial community structures of A0, B0, and C0 were similar because the raw material and inoculum were mixed at the starting time and the microorganisms were mainly from the inoculum. By day 17, the community structures of A17, B17, and C17 were different: A17 was in the slow production phase of CH₄ (pH: 7.4), B17 was in the acidification phase (pH: 6.5), and C17 was at the peak production period of CH₄ (pH: 7.8). After 39 days of fermentation, the microbial community structures between A39, B39, and C39 were similar because the 39th day is the peak period or close to the peak period of CH₄ production for different fermented raw materials. In the final stage of fermentation, the microbial community structures between A67, B67, and C67 were similar because the fermentation of each raw material was already in the later and organic matter was not available, the microorganisms were lacking a food source. In summary, the influence of the fermentation period on the microbial community was greater than that of the type of raw material.

3.4.3. Microbial Community Function Concerning Sulfide Metabolism

For the reduction in sulfate in CMAD, the abundance of the gene concerning sulfate reduction was relatively stable and low, between 0.01 and 0.07% during the whole process (

Table 3. Abundance (%) of genes with a metabolism function based on FAPROTAX predictions.

Functional Groups	Chicken Manure AD (%)				Vegetable Waste AD (%)				Inoculum (%)
	A0	A17	A399	A67	B0	B17	B399	B67	C0	C17	C39	C67
Methanogenesis	0.21	0.26	6.97	0.20	1.70	0.39	7.79	1.54	0.26	5.34	0.09	0.34
Sulfate respiration	0.03	0.07	0.01	0.04	0.03	0.04	0.01	0.01	0.05	0.04	0.06	0.03
Sulfur respiration	0.48	0.95	0.64	1.12	0.43	2.83	1.28	1.56	0.46	0.51	1.62	0.92
Sulfite respiration	0.02	0.06	0.01	0.03	0.03	0.02	0.01	0.01	0.03	0.04	0.05	0.03
Thiosulfate respiration	0.01	0.02	0.01	0.01	0.02	0.02	0.01	0.00	0.02	0.02	0.01	0.01
Respiration of sulfur compounds	0.52	1.05	0.66	1.18	0.50	2.89	1.30	1.57	0.54	0.57	1.69	0.96
Dark sulfide oxidation	0.00	0.00	0.00	0.06	0.01	0.01	0.01	0.01	0.01	0.00	0.04	0.06
Dark sulfur oxidation	0.00	0.00	0.00	0.06	0.00	0.01	0.01	0.01	0.01	0.00	0.04	0.05
Dark thiosulfate oxidation	0.00	0.00	0.00	0.06	0.00	0.01	0.01	0.01	0.01	0.00	0.05	0.06
Dark oxidation_of_sulfur_compounds	0.00	0.00	0.00	0.06	0.01	0.01	0.01	0.013	0.01	0.00	0.04	0.06
Fermentation	30.05	22.85	6.27	10.28	22.27	22.38	9.45	8.82	27.15	12.53	8.82	8.013

Note: A0, B0, and C0 represent the chicken manure AD, vegetable waste AD, and day 0 samples of inoculum, respectively.

); i.e., sulfate reduction occurred from the start to the end of the experiment. Thus, the biochemical process of the conversion of sulfate in CM to sulfide/H₂S was slow and continuous. For protein decomposition in CMAD, the genes linked to fermentation and playing roles in anaerobic acidification and decomposition of macromolecular organic matter (e.g., amino acids) exhibited high abundances between 22.8 and 30.5% before day 17 (Table 3). They subsequently fell to 6.27–10.28% after day 39. Thus, these genes were identified as important for the transformation of organic sulfur (such as protein) to sulfide between days 0 and 17.

For sulfate reduction in VWAD, the abundance of sulfate reduction genes was relatively stable and low, between 0.01 and 0.05% during the process (Table 3), indicating that sulfate reduction occurred from the start to the end of the experiment. Thus, sulfate in VW (high sulfate) was slowly and continuously converted to sulfide/H₂S, which provides important evidence for the fact that the large production of H₂S on the first day was caused by the sulfide reaching chemical equilibrium and not by sulfate reduction. Similarly, another study showed that the AD of food waste can convert sulfate to sulfide/H₂S in a long-term biochemical process and that large amounts of H₂S were produced on the first day by the sulfide ion equilibrium [13,30]. For protein decomposition in VWAD, the genes linked to fermentation and which play roles in the anaerobic acidification and decomposition of macromolecular organic matter (e.g., amino acids) exhibited a high abundance, between 22.7 and 22.8%, before day 17 (Table 3). Subsequently, they fell to 8.82–9.45% after day 39. Therefore, these genes are important for the transformation of S-protein to S-sulfide between days 0 and 17. This indicates that the large production of H₂S on the first day was not caused by protein decomposition. Another study showed that food waste AD can convert protein to sulfide/H₂S in a long-term biochemical process [13,30].

Generally, independent of focusing on CMAD or VWAD, the abundance of sulfate reduction bacteria (SRB) and microbes concerning organosulfur compounds (such as amino acids and protein) fermentation were stable in the fermentation process: the abundance of SRB was 0.01~0.07%, while that concerning organosulfur compounds fermentation was 22.8~30.5%.

4. Conclusions

In this study, H₂S production characteristics of CMAD and VWAD were observed. The H₂S production potential of CM was substantially lower than that of VW because it was easier to produce a large number of organic acids from VW than from CM. The rapid production of biogas and fermentation acidification were important reasons for the release of H₂S at the start of CMAD. The sulfide equilibrium model in AD was capable of predicting H₂S production in both CMAD and VWAD. Furthermore, the abundance of sulfate reduction bacteria (SRB) and microbes concerning the fermentation of organosulfuric compounds were stable in the fermentation process: the abundance of SRB was 0.01~0.07%, while that concerning the fermentation of organosulfuric compounds was 22.8~30.5%. This study indicated that the H₂S concentration of CMAD biogas was more than five times that of VWAD because CM is alkalescent and VW is acidic; therefore, the potential strategy for controlling H₂S in situ in CMAD is different than in VWAD. In the future, it is important to systematically compare H₂S emission differences between acidic material (such as vegetable waste and food waste) and alkaline material (such as pig manure) in AD.