Enhanced Methane Production in the Anaerobic Digestion of Swine Manure: Effects of Substrate-to-Inoculum Ratio and Magnetite-Mediated Direct Interspecies Electron Transfer

Abstract

Improving the anaerobic digestion (AD) of swine manure is crucial for sustainable waste-to-energy systems, given its high organic load and process instability risks. This study examined the combined effects of substrate-to-inoculum ratio (SIR, 0.1–3.2) and magnetite-mediated direct interspecies electron transfer on biogas production, effluent quality, and microbial community dynamics. The highest methane yield (262 ± 10 mL CH₄/g COD) was obtained at SIR 0.1, while efficiency declined at higher SIRs due to acid and ammonia accumulation. Magnetite supplementation significantly improved methane yield (up to a 54.1% increase at SIR 0.2) and reduced the lag phase, particularly under moderate SIRs. Effluent characterization revealed that low SIRs induced elevated soluble COD (SCOD) levels, attributed to microbial autolysis and extracellular polymeric substance release. Furthermore, magnetite addition mitigated SCOD accumulation and shifted molecular weight distributions toward higher fractions (>15 kDa), indicating enhanced microbial activity and structural polymer formation. Microbial analysis revealed that magnetite-enriched Syntrophobacterium and Methanothrix promoted syntrophic cooperation and acetoclastic methanogenesis. Diversity indices and PCoA further showed that both SIR and magnetite significantly shaped microbial structure and function. Overall, an optimal SIR range of 0.2–0.4 under magnetite addition provided a balanced strategy for enhancing methane recovery, effluent quality, and microbial stability in swine manure AD.

1. Introduction

Swine manure is a major type of organic waste generated in large volumes by the livestock industry and is characterized by high concentrations of both organic matter and nitrogen [1]. If not adequately treated prior to discharge, it can result in a range of environmental issues, including odor emissions, water pollution, and greenhouse gas release. In response, substantial efforts have been directed toward developing technologies that ensure stable treatment and resource recovery of swine manure [2,3,4]. Among these, anaerobic digestion (AD) has gained significant attention as a biogas production method that converts organic compounds into methane and carbon dioxide via four main stages, hydrolysis, acidogenesis, acetogenesis, and methanogenesis, thereby serving both environmental mitigation and renewable energy recovery while contributing to the realization of carbon neutrality [5]. However, the efficiency and stability of AD are influenced by various factors, including the physicochemical characteristics of the input substrate, composition and activity of the microbial community, and operational parameters [6,7].

Swine manure contains slowly degradable organics, including fibrous materials such as cellulose and hemicellulose, which likely delay the initial hydrolysis step and reduce overall biodegradability. Although its lignin content is typically lower than that of ruminant manure, the solid fraction can continue to pose hydrolysis challenges. These physicochemical characteristics underscore the necessity of tailored operational strategies to enhance the AD performance when swine manure is used as a feedstock. Owing to its combination of low biodegradability and high nitrogen content, precise adjustment of operational conditions is essential to ensure process stability. Among these conditions, the substrate-to-inoculum ratio (SIR) is recognized as a critical parameter influencing AD performance [8,9]. SIR represents the balance between the organic load of the substrate and the available microbial biomass, directly influencing biodegradation efficiency, organic acid accumulation, ammonia concentration, and methane yield [10,11]. While low SIRs promote sufficient microbial activity and rapid degradation, excessively low values can potentially limit substrate availability, resulting in reduced methane productivity. Conversely, high SIRs can result in the accumulation of organic acids and ammonia, resulting in acidification and microbial inhibition [10,12]. Therefore, optimizing SIR is particularly crucial for swine manure; however, studies specifically investigating optimal SIR conditions for this substrate remain limited.

To address limitations that cannot be fully resolved solely by SIR adjustment, advanced strategies have been explored to further enhance AD performance. Recently, direct interspecies electron transfer (DIET) has gained traction as a promising approach to enhance AD performance [13,14]. Conventional electron transfer relies on indirect mechanisms involving intermediates, such as hydrogen (H₂) and formate, which are prone to energy losses, delayed reactions, and pH fluctuations [15]. In contrast, DIET enables direct electron exchange between microbial species through conductive materials, thereby accelerating electron flow, enhancing reaction kinetics, and improving overall process stability.

Various conductive materials, such as granular activated carbon (GAC) [16], biochar [17], and carbon nanotube (CNT) [18], have been investigated to facilitate DIET, among which magnetite (Fe₃O₄) has emerged as a particularly promising candidate owing to its high electrical conductivity and biocompatibility [19]. In addition to facilitating DIET between methanogens and iron-reducing bacteria, magnetite indirectly enhances enzymatic activity during hydrolysis and acidogenesis by stimulating microbial growth. It also promotes the degradation of recalcitrant organics and fosters active microbial metabolism. These multifunctional roles make magnetite an effective additive for improving AD performance [20,21].

However, despite its potential benefits, the effectiveness of magnetite in enhancing AD is highly dependent on substrate characteristics and operational parameters, particularly the SIR. Variations in SIR alter the balance between substrate availability and microbial biomass, which consequently influences electron transfer pathways, microbial community structure, and methane productivity. Despite increasing interest, integrated studies examining this interaction remain limited. For example, González-Fernández and García-Encina et al. (2009) evaluated anaerobic digestion of swine slurry under different SIRs and found that an SIR of 1.0 achieved the highest biogas recovery with a stable VFA profile [22]. Córdova et al. (2022) reported that magnetite addition to swine manure digestion enhanced methane content to 60% and achieved a maximum methane yield of 3.82 × 10⁻² Nm³ CH₄/kg VS compared with the control [23]. However, integrated investigations evaluating the combined effects of these two factors remain scarce. Therefore, a systematic evaluation of the interactive effects between SIR and magnetite addition is essential to optimize DIET-driven performance improvements. A combined strategy integrating SIR optimization with DIET-enhancing additives, such as magnetite, could offer a synergistic solution for improving the AD of swine manure, particularly under high organic loading conditions, by enhancing methane yield and maintaining process stability.

Based on this context, this study investigated the effects of varying SIR on biogas production and organic matter degradation during the AD of swine manure. Additionally, this study assessed the feasibility of magnetite supplementation to facilitate DIET and compared its impact under different SIR conditions in terms of process performance and stability. To support this analysis, the molecular weight (MW) distribution of dissolved organic carbon (DOC) and microbial community profiling were conducted to elucidate degradation patterns and microbial interactions. This study provides a comprehensive understanding of the interactive effects of SIR and magnetite on the biological response and microbial ecology in AD systems, offering valuable insights for optimizing and stabilizing swine manure-based AD processes.

2. Materials and Methods

2.1. Preparation of Substrate, Inoculum, and Magnetite

Table 1. Characteristics of the substrate and inoculum used in this study.

	Substrate	Inoculum
TCOD (g/L)	33.5 ± 0.2	68.7 ± 7.1
SCOD (g/L)	16.9 ± 0.4	2.0 ± 0.1
TS (g/L)	21.7 ± 0.3	72.8 ± 0.2
VS (g/L)	13.7 ± 0.2	61.4 ± 0.1
Alkalinity (g CaCO3/L)	23.6 ± 0.5	7.4 ± 0.1
pH	8.0 ± 0.2	7.5 ± 0.1

summarizes the characteristics of the substrate and inoculum used in the experiments.

The substrate was swine manure sampled from the storage tank of a livestock manure treatment facility located in Cheongju-si, Republic of Korea. Its measured properties were as follows: total chemical oxygen demand (TCOD) = 33.5 ± 0.2 g/L, soluble chemical oxygen demand (SCOD) = 16.9 ± 0.4 g/L, total solids (TS) = 21.7 ± 0.3 g/L, volatile solids (VS) = 13.7 ± 0.2 g/L, alkalinity = 23.6 ± 0.5 g CaCO₃/L, pH = 8.0 ± 0.2. The inoculum was obtained from an anaerobic digester operated at a brewery wastewater treatment plant in the same city. Its characteristics were as follows: TCOD = 68.7 ± 7.1 g/L, SCOD = 2.0 ± 0.1 g/L, TS = 72.8 ± 0.2 g/L, vs. = 61.4 ± 0.1 g/L, alkalinity = 7.4 ± 0.1 g CaCO₃/L, pH = 7.5 ± 0.1. Prior to the experiments, the substrate was stored at 4 °C, and the inoculum was kept at room temperature (25 °C) in sealed containers to maintain anaerobic conditions. Both were thoroughly mixed before use in the experiments. The magnetite used in the experiment exhibited a purity of 99% and was purchased from Junsei Chemicals Co., Ltd. (Tokyo, Japan). The particle size of the magnetite powder was within the range of 50–100 nm, enabling effective interaction with microbial cells and dispersion within the anaerobic medium. Prior to use, the magnetite was stored in an airtight container at room temperature (25 °C) in a dry, dark environment to prevent oxidation and moisture absorption.

2.2. Biochemical Methane Potential Test

A batch-type BMP test was conducted to determine the optimal SIR and evaluate the effect of magnetite addition on AD performance. Serum bottles with a total volume of 250 mL were employed, and the working volume was adjusted to 165 mL. The substrate concentration was fixed at 13 g VS/L, corresponding to 0.65 g VS per reactor. To achieve target SIRs of 0.1, 0.2, 0.4, 0.8, 1.6, and 3.2 (g VS substrate/g VS inoculum) [11,12], inoculum sludge was added at 6.50, 3.25, 1.63, 0.81, 0.41, and 0.20 g VS, respectively. For each SIR condition, an additional experimental group comprising 20 mM Fe was prepared to assess the effect of DIET stimulation [20,24]. After adding the substrate, inoculum, and magnetite (if applicable), tap water was added to each bottle to reach the final working volume of 165 mL. The experimental conditions for SIRs, with and without magnetite addition, are summarized in

Table 2. Experimental conditions for BMP test at different SIRs with and without magnetite addition.

	SIRs (g VS/gVS)	Substrate (g VS)	Inoculum (g VS)	Magnetite (mM Fe)	Working Volume (mL)
Without magnetite addition	0.1	0.65	6.50	20	165
	0.2	0.65	3.25	20	165
	0.4	0.65	1.63	20	165
	0.8	0.65	0.81	20	165
	1.6	0.65	0.41	20	165
	3.2	0.65	0.20	20	165
With magnetite addition	0.1	0.65	6.50	20	165
	0.2	0.65	3.25	20	165
	0.4	0.65	1.63	20	165
	0.8	0.65	0.81	20	165
	1.6	0.65	0.41	20	165
	3.2	0.65	0.20	20	165

.

The initial pH of each reactor was adjusted to 7.5 utilizing either 3 N KOH (85%, Duksan Pure Chemicals Co., Ltd., Seoul, Republic of Korea) or 3 N HCl (35%, Duksan Pure Chemicals Co., Ltd., Seoul, Republic of Korea). Anaerobic conditions were established by purging the headspace with high-purity nitrogen gas (99.999%) for at least 2 min, followed by sealing the bottles with butyl rubber stoppers and aluminum caps. The reactors were incubated for 42 days at 38 ± 2 °C within a shaking incubator (LSI-1005R, Daihan Labtech, Seoul, Republic of Korea), which provided continuous horizontal agitation at 150 rpm to ensure adequate mixing and prevent biomass settling. All experiments were conducted in duplicate, and the average values were used for subsequent data analysis.

2.3. Analytical Methods

The following parameters were monitored during the experimental period: chemical oxygen demand (COD), biogas volume and methane composition, total residual organic acid, and MW distribution. TCOD and SCOD were measured via the closed reflux colorimetric method as described in the Standard Methods [25]. Daily biogas production was measured using 50 and 100 mL glass syringes (Labscitech, Corona, CA, USA). The methane content in the biogas was analyzed at each sampling point employing a gas chromatograph (GC, SRI 310, SRI Instruments, Torrance, CA, USA) equipped with a thermal conductivity detector (TCD). The GC was fitted with a HayeSep T column (3 ft × 1/8″), and high-purity nitrogen gas (99.9999%) was used as the carrier gas.

Total residual organic acids were analyzed employing high-performance liquid chromatography (HPLC; LC-2040C MT Plus, Shimadzu, Kyoto, Japan) equipped with an Aminex HPX-87H column (300 mm × 7.8 mm; Bio-Rad, Hercules, CA, USA). The mobile phase was 0.008 N sulfuric acid (98%, Sigma-Aldrich, St. Louis, MO, USA), and the operating conditions included a UV detection wavelength of 210 nm and a flow rate of 0.6 mL/min. The MW distribution of DOC was analyzed under a sample concentration of 20 mg/L DOC employing high-performance size-exclusion chromatography (HPLC-SEC; LC-20A series, Shimadzu, Kyoto, Japan) equipped with a UV detector.

2.4. Microbial Community Analysis

To analyze the microbial community structure, 30 mL of digested slurry was collected from each BMP reactor at the end of the incubation period. The samples were centrifuged at 3000 rpm for 10 min, and 2 mL of the pellet was retrieved for DNA extraction. DNA was initially extracted utilizing the FastDNA™ Spin Kit for Soil (MP Biomedicals, Santa Ana, CA, USA) according to the manufacturer’s protocol (Macrogen, Seoul, Republic of Korea), followed by further purification using the UltraClean^® Microbial DNA Isolation Kit (Mo Bio Laboratories, Carlsbad, CA, USA). PCR amplification was performed utilizing the GS-FLX Titanium emPCR Kit (454 Life Sciences, Branford, CT, USA), and 20 ng of extracted DNA was used in a 50 μL PCR reaction. The bacterial 16S rRNA gene was amplified employing the universal primer set 27F (5′-GAGTTTGATCMTGGCTCAG-3′) and 518R (5′-WTTACCGCGGCTGCTGG-3′). PCR reactions were performed using the FastStart High Fidelity PCR System (Roche, Basel, Switzerland) under the following thermal conditions: initial denaturation at 94 °C for 3 min, followed by 35 cycles of denaturation at 94 °C for 15 s, annealing at 55 °C for 45 s, and extension at 72 °C for 1 min, with a final extension step at 72 °C for 8 min. Amplified PCR products were purified using AMPure^® XP beads (Beckman Coulter, Brea, CA, USA). Sequencing was conducted employing the FLX Titanium platform based on 454 pyrosequencing technology, in accordance with the manufacturer’s protocol. Sequencing data were analyzed using the MOTHUR software 1.36.1 package to identify operational taxonomic units (OTUs), assign taxonomic classification, and perform community-level comparisons. To ensure data quality, sequences containing one or more ambiguous bases were removed. Only reads exhibiting a length of at least 300 nucleotides were retained for downstream analysis.

2.5. Statistical Analysis

Cumulative methane production, methane production rate, methane yield, and lag phase were calculated using the modified Gompertz equation Equation (1) implemented in the SigmaPlot 10.0 (Systat Software Inc., San Jose, CA, USA):

$$M\left(t\right)=R_0\times exp\left\{-exp\left[{\displaystyle \frac{R_0\times e}{M_0}}\times (\lambda -t)+1\right]\right\}$$

(1)

where M(t) represents the cumulative methane production at cultivation time (mL), M₀ denotes the methane production potential (mL), R₀ denotes the methane production rate (mL/d), λ = Lag phase (d), and e is 2.71828.

To evaluate microbial community dynamics between different SIRs in samples with and without magnetite, 16S rRNA gene sequencing data were analyzed using alpha diversity indices (Shannon, Simpson, Chao1) and beta diversity using principle coordinate analysis (PCoA) based on Bray–Curtis dissimilarity metrics at the genus level, using Python Jupyter Notebook 7.2.2.

3. Results and Discussion

3.1. Effects of Substrate-to-Inoculum Ratio and Magnetite Addition on Methane Production

The SIR is known as a critical operational parameter in AD, as it determines the balance between substrate availability and microbial activity. In this study, the influence of varying SIRs (ranging from 0.1 to 3.2) on methane yield and organic matter degradation was investigated. The results are summarized in Figure 1 and

Table 3. Summary of methane yield, methane production rate, and lag phase under varying SIRs in anaerobic digestion of swine manure, with and without magnetite addition.

	SIRs	CH4 Yield (mL/g COD)	CH4 Production Rate (mL/day)	Lag Phase (Day)
Without magnetite addition	0.1	262 ± 10	77 ± 5	-
	0.2	110 ± 7	94 ± 4	-
	0.4	77 ± 6	49 ± 1	-
	0.8	90 ± 7	40 ± 1	3.1 ± 0.3
	1.6	70 ± 10	43 ± 1	11.6 ± 0.1
	3.2	45 ± 9	35 ± 1	17.1 ± 0.2
With magnetite addition	0.1	267 ± 16	106 ± 4	-
	0.2	170 ± 14	75 ± 3	-
	0.4	89 ± 9	43 ± 2	-
	0.8	102 ± 12	45 ± 1	2.5 ± 0.2
	1.6	77 ± 9	26 ± 1	6.2 ± 0.3
	3.2	39 ± 6	17 ± 1	11.5 ± 0.3

.

In the absence of magnetite, the methane yield and production rate declined as the SIR increased. At SIR 0.1, the methane yield reached 262 ± 10 mL/g COD, with a production rate of 77 ± 5 mL/day, whereas at SIR 3.2, these values decreased to 45 ± 9 mL/g COD and 35 ± 1 mL/day, respectively. Additionally, the lag phase extended significantly at a high SIR, reaching 17.1 ± 0.2 days at SIR 3.2. This inhibition is attributed to the accumulation of organic acids and pH drop under excess substrate conditions. In contrast, the addition of magnetite generally enhanced methane production, particularly under low SIR conditions. For instance, at an SIR of 0.1, the methane production rate increased by approximately 38% (106 ± 4 mL/day) compared with that of the control sample (77 ± 5 mL/day), in conjunction with a modest improvement in methane yield to 267 ± 16 mL/g COD. At an SIR of 0.2, the methane yield increased by approximately 54.5% (170 ± 14 mL/g COD vs. 110 ± 7 mL/g COD). Additionally, the lag phase at an SIR of 0.8 was reduced from 3.1 ± 0.3 to 2.5 ± 0.2 days, indicating that magnetite facilitated the earlier onset of methanogenesis.

The observed improvements can be attributed to the conductive properties of magnetite, which facilitate DIET among syntrophic microbial communities. This effect was supported by microbial community analysis (Section 3.3), which showed increased relative abundance of Syntrophobacterium and Methanothrix under magnetite-amended conditions. These taxa are known to participate in syntrophic acetate oxidation and acetoclastic methanogenesis, suggesting that magnetite enhanced DIET and strengthened microbial cooperation, ultimately improving biogas production.

Unlike conventional interspecies electron transfer mediated by intermediates, such as hydrogen or formate, DIET enables more efficient electron flow, resulting in reduced energy losses and enhanced process stability [26]. However, as the SIR increased beyond 1.6, the positive effects of magnetite diminished. At an SIR of 3.2, no significant differences were observed in the methane yield or production rate between the magnetite-supplemented and control groups. This indicates that, under substrate-overloaded conditions, the accumulation of organic acids and ammonia likely suppressed microbial activity to the extent that DIET was no longer effectively promoted by magnetite [27].

In addition to methane production, organic matter removal efficiency was identified as a key performance indicator. TCOD and SCOD were analyzed under each SIR condition, both with and without magnetite addition (Figure 2).

TCOD removal efficiency was highest at low SIRs (40.5 ± 1.3% and 35.2 ± 1.5% at SIRs of 0.1 and 0.2, respectively) but gradually declined to 22.7 ± 0.5% at an SIR of 3.2 (Figure 2a). In contrast, SCOD concentrations exhibited an inverse trend, with values of 17.6 ± 0.7 and 11.7 ± 0.6 g/L at SIRs of 0.1 and 0.2, respectively, decreasing to 3.4 ± 0.2 g/L at SIR 3.2 (Figure 2b). These findings indicate that, at low SIRs, the presence of excessive microbial biomass potentially induces the release of microbial cell-associated compounds, such as extracellular polymeric substances (EPS) and soluble metabolites, thereby resulting in higher SCOD levels. Therefore, effluent SCOD represents not only residual undegraded substrate but also microbial-derived organics, highlighting the complexity of interpreting SCOD as a sole indicator of biodegradation efficiency.

Similar patterns were observed in the magnetite-supplemented reactors. Notably, SCOD concentrations at SIRs of 0.4 and 0.8 were reduced by 42.2 and 31.4%, respectively, compared with that of the control samples. In contrast, the TCOD removal efficiency was marginally lower in the presence of magnetite, indicating that, as microbial activity was enhanced, the rate of substrate degradation was likely reduced marginally. These results demonstrate a trade-off between methane productivity and effluent quality depending on the SIR. Low SIR conditions can enhance methane production; however, they can compromise effluent quality owing to the release of soluble microbial byproducts. In contrast, high SIR conditions can improve effluent clarity but significantly reduce methane yield. Considering both methane recovery and effluent quality characteristics, such as the SCOD and MW distribution, SIR 0.2–0.4 was identified as the most balanced operational window.

3.2. Effects of SIR and Magnetite Addition on Organic Acid Accumulation and MW Distribution

To better understand the previously observed inverse relationship, where increasing SIR was associated with decreased microbial diversity and methane productivity, the organic acid composition and MW distribution were analyzed to gain deeper insight into the underlying metabolic responses under varying SIR and magnetite conditions. Organic acids, produced during AD, serve as essential intermediate metabolites that represent the degree of substrate degradation and metabolic efficiency of the methanogenesis stage. As such, organic acid profiles are widely used as key indicators for diagnosing process stability. In this study, the metabolic characteristics and equilibrium states of biological conversion under varying SIR conditions were quantitatively assessed by analyzing total residual organic acid concentrations and compositions.

As presented in Figure 3, the total organic acid concentrations were highest at SIRs of 0.1 and 0.2, measuring 2399 ± 128 and 1832 ± 61 mg COD/L, respectively. Organic acid levels gradually declined as SIR increased, reaching a minimum of 550 ± 53 mg COD/L at SIR 0.8. Notably, all values remained below 2500 mg COD/L, which is well under the typical inhibition threshold for AD systems (approximately 4000–6000 mg COD/L), confirming the absence of process inhibition induced by organic acids [28].

Acetic acid and propionic acid were identified as the dominant organic acids, with their proportions varying by SIR. At SIRs of 0.1 and 0.2, acetic acid accounted for nearly 100% of total organic acids, indicating that acidogenesis was active, whereas methanogenesis was less efficient, resulting in acetic acid accumulation. In contrast, the proportion of propionic acid increased at SIRs ≥ 0.4; at SIR 3.2, it became the predominant acid. This shift is attributed to the thermodynamically unfavorable nature of propionate oxidation to acetate, particularly under high substrate concentrations [29]. The addition of magnetite significantly influenced organic acid concentration and composition. In all SIR conditions, total organic acids were lower in magnetite-amended groups, with the differences more pronounced at higher SIRs. At SIRs of 1.6 and 3.2, total organic acid concentrations in the control group were 1114 ± 79 and 794 ± 132 mg COD/L, while magnetite-amended reactors exhibited 618 ± 99 and 548 ± 92 mg COD/L, reductions of 44.6 and 30.9%, respectively. Additionally, the magnetite groups exhibited a higher proportion of acetic acid and reduced propionic acid at SIRs ≥ 0.4, suggesting enhanced syntrophic acetate oxidation via DIET [30].

To further explain the apparent contradictory observation of high organic acid concentrations at low SIRs despite elevated methane production, the MW distribution of DOC in the effluent was analyzed, as shown in Figure 4.

At SIR 0.1, the average MW was highest at 1.326 ± 0.21 kDa, gradually decreasing to 0.424 ± 0.11, 0.394 ± 0.01, and 0.362 ± 0.02 kDa at SIRs 0.2, 0.4, and 0.8, respectively. However, at SIRs 1.6 and 3.2, the average MW increased again to 0.651 ± 0.13 and 0.564 ± 0.12 kDa, respectively. These findings indicate that, under low SIR conditions, microbial autolysis or stress-induced responses likely promote the release of high-MW substances, including EPS, which are typically associated with cell decay or environmental stress. In contrast, at high SIRs, the incomplete degradation of excess substrate likely resulted in the accumulation of residual high-MW organic matter in the effluent.

Furthermore, magnetite addition exerted a notable impact on the MW distribution of the effluent. Within the 0.1–0.8 SIR range, the magnetite-amended groups consistently exhibited higher average MWs than the corresponding control reactors, i.e., 1.605 ± 0.16, 1.190 ± 0.04, and 1.004 ± 0.21 kDa at SIRs of 0.1, 0.2, and 0.8, representing increases of 21.1, 180.4, and 177.8%, respectively. This trend suggests that DIET-enhanced microbial activity likely stimulated the production of EPS or induced microbial autolysis [31]. Notably, high-MW fractions (>15 kDa) accounted for 5.9, 4.6, and 3.0% of the total soluble organics at SIRs 0.1, 0.2, and 0.8, respectively, indicating a decreasing contribution of large biopolymers as the SIR increased. Although typical EPS molecules exceed 100 kDa and are not fully represented within the analytical range, the presence and relative enrichment of >15 kDa fractions support the hypothesis of EPS generation in magnetite-enhanced systems [32]. However, at SIRs ≥ 1.6, the differences in MW between magnetite and control groups were minimal, likely owing to reduced microbial activity under substrate-overloaded conditions, which limited the DIET-promoting effects of magnetite.

3.3. Microbial Community Dynamics

3.3.1. Alpha and Beta Diversity and Community-Level Shifts

To assess the ecological characteristics and structural responses of microbial communities during AD, alpha diversity indices, including Shannon, Simpson, and Chao1, were analyzed in conjunction with principal coordinate analysis (PCoA) for beta diversity (

Table 4. Alpha diversity indices (Shannon, Simpson, Chao1) coordinates of bacterial and archaeal communities under different SIRs in samples with and without magnetite.

		Bacteria			Archaea
	SIRs	Shannon	Simpson	Chao1	Shannon	Simpson	Chao1
Without magnetite addition	0.1	2.41	0.85	87	0.97	0.55	48
	0.2	2.29	0.81	88	0.93	0.54	52
	0.4	1.95	0.70	87	0.96	0.52	54
	0.8	1.80	0.67	88	1.08	0.55	47
	1.6	1.40	0.58	71	1.24	0.59	55
	3.2	1.36	0.56	64	1.22	0.58	47
With magnetite addition	0.1	2.37	0.84	85	0.91	0.54	46
	0.2	2.19	0.78	84	0.93	0.54	44
	0.4	1.72	0.65	74	0.98	0.53	51
	0.8	1.71	0.65	65	1.09	0.55	46
	1.6	1.40	0.58	52	1.20	0.58	45
	3.2	1.45	0.58	65	1.36	0.65	55

and Figure 5).

In bacterial communities without magnetite addition, both Shannon and Simpson indices decreased as the SIR increased. Specifically, the highest values were observed at SIR 0.1 (2.41 and 0.85, respectively), and the lowest values were recorded at SIR 3.2 (1.36 and 0.56). This indicates that low substrate loading favored balanced and diverse bacterial populations, whereas high loading resulted in the dominance of specific taxa, thereby reducing overall diversity and evenness. A similar trend was observed under magnetite-supplemented conditions, where Shannon and Simpson indices decreased from 2.37 and 0.84 at SIR 0.1 to 1.45 and 0.58 at SIR 3.2, respectively. In contrast, archaeal communities exhibited the opposite trend. In the absence of magnetite, the Shannon and Simpson indices increased from 0.97 and 0.55 at SIR 0.1 to 1.22 and 0.58 at SIR 3.2, respectively. A comparable pattern was observed in the magnetite-added group, wherein the indices increased from 0.91 and 0.54 (SIR 0.1) to 1.36 and 0.65 (SIR 3.2). These results indicate that, while bacterial communities were more balanced under low SIR conditions and became less diverse as substrate loading increased, archaeal communities achieved greater diversity and evenness at higher SIRs. This is likely attributed to a broader spectrum of methanogenic archaea being activated or maintained under elevated organic loading. The Chao1 index, representing species richness, further supported these patterns. In bacterial communities without magnetite, richness remained relatively high (87–88) under low SIR conditions (0.1–0.8) but declined to 64 at SIR 3.2. Similarly, richness in the magnetite-amended group decreased from 85 to 65 across the same SIR range. In contrast, archaeal richness remained relatively stable across all SIR conditions (44–55), irrespective of magnetite presence, indicating that archaeal species richness was less sensitive to substrate loading. These diversity patterns were consistent with community-level structural changes identified by PCoA (Figure 5).

In the bacterial domain (Figure 5a), distinct divergence along the PC1 axis was observed at low SIR levels (0.1–0.4), particularly between magnetite-amended and control groups. This separation indicates that magnetite influenced bacterial consortia by promoting DIET-mediated interactions and triggering metabolic shifts, such as EPS production. As SIR increased (≥0.4), bacterial communities progressively shifted toward the negative PC1 region, indicating altered structures under substrate-rich conditions. Notably, at high SIR levels (≥1.6), clustering patterns between the magnetite and control groups overlapped substantially, indicating reduced microbial responsiveness to magnetite owing to substrate overloading. In contrast, archaeal communities (Figure 5b) exhibited more conservative shifts across SIR levels, with minimal separation between treatments, reflecting the resilience of methanogenic archaea to environmental variation and their stable contribution to biogas production. Based on these observations, the distinct structural and functional responses of bacterial and archaeal communities to substrate loading and magnetite application were evident. To better understand how specific microbial taxa contributed to these patterns, the relative abundance of dominant genera was subsequently analyzed.

3.3.2. Comparative Analysis of Dominant Microbial Taxa Under Different SIR and Magnetite Conditions

Bacterial community analysis revealed that Clostridium (21.6–63.3%), Levilinea (2.1–16.3%), and Turicibacter (4.8–18.9%) were the dominant genera across all SIR conditions (Figure 6a).

Clostridium and Turicibacter are characterized as fast-growing acidogenic bacteria with high substrate turnover, known for rapid hydrolysis of organic matter and production of various organic acids. Particularly, Turicibacter, a Firmicutes genus specialized in carbohydrate metabolism, exhibits enhanced activity in the presence of simple sugars and organic acids and is known for its adaptability to intestinal and anaerobic environments. In this study, the relative abundances of Clostridium and Turicibacter increased as SIR increased, reaching 63.3 and 17.5%, respectively, at SIR 3.2, indicating their competitive advantage under substrate-rich conditions. Similar trends were observed in the magnetite-amended group, where Clostridium abundance increased from 23.6–43.2% at low SIR (0.1–0.2) to 61.5–62.0% at high SIR (1.6–3.2).

In contrast to these fast-growing fermenters, Levilinea and Bellilinea, known for slower growth and lower substrate turnover, were more prevalent under low substrate conditions. At SIR 0.1 and 0.2, Levilinea accounted for 16.3 and 9.5%, and Bellilinea accounted for 3.4 and 5.9%, respectively. However, their abundances declined sharply with increasing SIR, dropping to 2.1 and 1.5% at SIR 3.2, respectively. This inverse trend relative to Clostridium and Turicibacter indicates that Levilinea and Bellilinea play a greater role in organic acid generation and substrate degradation under substrate-limited conditions and are competitively disadvantaged under substrate-rich environments. This represents a shift in microbial ecological strategy in response to increasing substrate availability, favoring fast-growing fermenters over slow-growing degraders.

Syntrophobacterium, a representative syntrophic bacterium known for converting propionate to acetate and interacting with hydrogenotrophic methanogens, exhibited higher abundance under low SIR conditions (4.9% at SIR 0.1, 2.8% at SIR 0.2). Its abundance declined significantly at high SIR (0.4–0.6%), which corresponds with propionate accumulation and methanogenesis inhibition. Notably, in magnetite-amended reactors, Syntrophobacterium abundance increased substantially compared with the control samples (without magnetite), with a 29.3% higher relative abundance observed at SIR 0.2. These findings indicate that magnetite facilitated DIET and strengthened syntrophic cooperation with methanogens, thereby contributing to the enhanced methane yield. Consistently, at SIR 0.2, the methane yield in the magnetite-amended group increased by 54.5% compared with the control samples. Terrisporobacter, which degrades peptides and amino acids to produce propionate, exhibited increasing dominance under high substrate conditions. In this study, its relative abundance reached 6.4 and 6.8% at SIR 1.6 and 3.2, respectively, implying active protein degradation and propionate accumulation, possibly contributing to methanogenesis inhibition. This trend was consistent with the higher proportion of propionate in total organic acids (70.8% and 100%) and the lower methane yields (70 ± 10 mL CH₄/g COD, 45 ± 9 mL CH₄/g COD) observed at SIR 1.6 and 3.2, respectively.

Alongside these bacterial shifts, archaeal community composition also varied in response to SIR and magnetite amendment (Figure 6b). Methanothrix (47.9–61.0%) and Methanobacterium (25.2–46.4%) were the dominant archaeal genera. Methanothrix, an acetoclastic methanogen that converts acetate to methane, exhibited the highest abundances at SIR 0.4 and 0.8 (61.0 and 59.1%, respectively) in the control samples and similarly high levels in magnetite-amended reactors (59.6 and 60.9%, respectively). This suggests that acetate utilization via the acetoclastic pathway was active under moderate substrate conditions regardless of magnetite presence.

However, at high SIR (1.6 and 3.2), the relative abundance of Methanothrix diverged between groups. In non-amended samples, it remained at 58.1 and 58.9%, whereas in magnetite-amended samples, it declined to 56.7 and 48.6%, respectively, indicating a suppression of acetoclastic methanogenesis under substrate overload, with a possible shift toward hydrogenotrophic pathways. Furthermore, Methanobacterium and Methanolinea, both hydrogenotrophic methanogens, exhibited decreasing abundance with increasing SIR. In control samples, their abundances decreased from 46.4 and 3.4% at SIR 0.1 to 25.2% and 1.0% at SIR 3.2. This trend indicates that hydrogen accumulation under high substrate conditions exceeded the consumption capacity of these methanogens, resulting in propionate accumulation and inhibited methane production. Magnetite-amended samples exhibited similar patterns, with Methanobacterium and Methanolinea decreasing to 33.0 and 0.53%, respectively, at SIR 3.2. Consistently, such archaeal community shifts were also reflected in methane yield performance at high SIRs. At SIR 1.6 and 3.2, the methane yields of the magnetite-amended samples (77 ± 9 and 39 ± 6 mL CH₄/g COD, respectively) were not significantly improved compared with those of the control groups (70 ± 10 and 45 ± 9 mL CH₄/g COD, respectively), indicating that the stimulatory effect of magnetite diminished under substrate-overloaded conditions.

These results demonstrate the close relationship between microbial community structure and functional performance in AD. Particularly, the balance between acetoclastic and hydrogenotrophic methanogens, influenced by DIET and substrate availability, plays a crucial role in process stability and methane conversion efficiency. The stimulatory effect of magnetite was evident under specific conditions, particularly via enhanced Syntrophobacterium activity and maintenance of acetoclastic methanogenesis. However, under substrate overload, alternative microbial pathways and genera, such as Acetonema and Methanosarcina, became more prevalent, suggesting a shift in community resilience. These findings highlight the importance of tailoring operational strategies based on microbial community traits and prevailing metabolic routes to optimize methane production under variable substrate conditions.

4. Conclusions

This study demonstrated that combined adjustment of SIR and magnetite supplementation significantly influenced methane production efficiency, effluent quality, and microbial community structure during anaerobic digestion of swine manure. The highest methane yield (262 ± 10 mL CH₄/g COD) was achieved at SIR 0.1; however, effluent quality indicators, particularly soluble COD and residual organic acid concentration, were more favorable at SIR 0.4. Therefore, considering both energy recovery and post-treatment stability, SIR 0.2–0.4 can be considered the most balanced operating window. Magnetite addition enhanced methane yield by up to 54.1% at moderate SIRs, promoted syntrophic acetate oxidation, and contributed to the maintenance of acetoclastic methanogenesis by increasing the relative abundance of key functional genera, such as Syntrophobacterium and Methanothrix. Additionally, under low SIR and magnetite-amended conditions, the MW distribution of dissolved organics shifted toward higher-MW fractions (e.g., >15 kDa), indicating the generation of EPS or microbial autolysis byproducts. Furthermore, microbial diversity patterns revealed that bacterial diversity decreased as SIR increased, whereas archaeal diversity exhibited a marginal increase. Community structure analysis further confirmed that magnetite influenced bacterial clustering at low-to-moderate SIRs; however, this effect diminished under substrate-overloaded conditions. These findings highlight the critical importance of balancing organic loading and electron transfer facilitation to optimize AD performance. However, this study has some limitations, including its batch-scale experimental design, which may not fully reflect the long-term operational stability in continuous systems. Therefore, future research should validate the SIR–magnetite interaction in continuous or pilot-scale operations, explore the potential of combining magnetite with other conductive materials such as granular activated carbon or biochar to achieve synergistic DIET enhancement, and assess the long-term stability of microbial syntrophy under operational fluctuations.