1. Introduction
Swine manure (SM), a nutrient-rich but poorly degradable organic waste, has garnered increasing attention as a renewable feedstock for anaerobic digestion (AD) [1]. However, its application in mono-digestion is hindered by several intrinsic limitations, including high ammonia concentrations, low hydrolyzability, and an imbalanced carbon-to-nitrogen ratio [2, 3]. The fibrous and proteinaceous nature of SM further contributes to low solubilization and slow microbial conversion rates, limiting the efficiency of energy recovery [4]. To address these challenges, co-digestion of SM with readily biodegradable substrates such as food waste (FW) has been widely studied [5, 6]. FW is rich in hydrolysable carbohydrates, proteins, and lipids that can complement the physicochemical properties of SM and promote synergistic effects [7]. Numerous studies have reported enhanced biogas production when food waste is incorporated into swine manure digestion. This synergy stems from an improved C/N balance and greater availability of readily degradable substrates, which together stimulate microbial co-metabolism during hydrolysis and acidogenesis [8, 9]. However, these benefits are highly ratio-dependent. Excessive FW input can lead to organic acid accumulation and ammonium inhibition, particularly under high organic loading conditions [10], thereby offsetting the potential gains in process performance. Achieving an optimal mixing strategy that balances synergy and stability remains a key challenge in FW–SM co-digestion systems [11].
Traditional co-digestion studies have investigated the effects of FW mixing ratios when combined with less readily biodegradable substrates such as yard waste, rice straw, and animal manures. While moderate FW additions (e.g., 10%) have been shown to significantly enhance methane production, severe volatile fatty acid inhibition has been observed when FW proportions exceed 20–26% of total feedstock [12]. However, recent attention has turned to the possibility that trace amounts of biodegradable organics may act not merely as substrates but as metabolic stimulants [13, 14]. This emerging concept, sometimes referred to as trace substrate triggering, suggests that minimal co-substrate inputs may enhance microbial syntropy, stimulate dormant hydrolytic populations, or alter pathway dominance in a direction favorable to methane production [15, 16]. This approach may help lower operational costs, simplify feedstock logistics, and improve flexibility in feedstock management [17]. Nonetheless, few studies have evaluated whether trace FW supplementation can yield meaningful performance improvement or induce significant shifts in microbial community structure and function [18].
This study employed two batch experimental frameworks. The co-digestion ratio optimization (CRO) test was designed to assess methane production, organic matter degradation, and process stability across a range of SM:FW mixing ratios. In parallel, the trace food waste addition (TFWA) test investigated whether minimal FW supplementation could stimulate microbial activity and enhance the conversion of recalcitrant SM, even at negligible loading rates [19].
These tests were coupled with detailed chemical and biochemical analyses, including organic fraction characterization and high-throughput 16S rRNA gene sequencing, to track microbial community dynamics at both class and species levels [20]. Beyond process performance metrics, microbial community analysis plays a crucial role in revealing the mechanistic basis for observed performance changes [21]. Identifying functional taxa shifts under trace-level interventions can provide insights into biological drivers and inform microbial-targeted process control strategies [22, 23]. In particular, the ability to selectively stimulate fermentative bacteria or acetoclastic methanogens with minimal FW input may offer a low-impact approach to optimizing performance in nutrient-rich but hydrolytically limited substrate systems like SM digestion [24].
Ultimately, this study aimed to elucidate the functional and microbial mechanisms underlying co-digestion synergy in SM–FW systems, with a particular emphasis on the biological effects of TFWA. By integrating process performance data with microbial community profiling, the work seeks to determine whether small doses of biodegradable co-substrate can drive meaningful metabolic restructuring and enhance AD performance without the operational risks associated with high-strength feedstock inputs.
2. Materials and Methods
2.1. Preparation of Substrates and Seed Sludge
SM was collected from a livestock wastewater treatment facility in Cheongju, South Korea, and used as the primary substrate. FW was obtained from a university cafeteria at Chungbuk National University in South Korea and homogenized into a uniform slurry using a commercial blender to ensure consistent volumetric dosing. Both SM and FW were stored at 4°C in sealed containers and used within 3 days prior to the experiments to minimize storage-related alterations, in accordance with a short-term storage approach previously applied to FW for AD [25]. Seed sludge was sourced from a full-scale anaerobic digester treating brewery wastewater (Cheongju, South Korea) and served as the inoculum. The physicochemical properties of SM, FW, and seed sludge are summarized in Table S1.
2.2. Experimental Design
Two batch experiments were conducted using 250 mL serum bottles to evaluate the effects of feedstock composition on AD performance. Anaerobic conditions were established by purging the headspace with high-purity nitrogen gas (99.999%) for at least two minutes before sealing. All experiments were carried out in duplicate under mesophilic conditions (38 ± 1°C) with continuous shaking at 150 rpm for 60 days. Two distinct experimental approaches were applied: the CRO test and the TFWA test.
2.2.1. CRO test
The CRO test aimed to evaluate the effect of varying SM:FW ratios on methane production and process stability in AD. SM and FW were mixed at COD-based weight ratios of 100:0, 80:20, 60:40, 40:60, 20:80, and 0:100 corresponding to approximately 100:0, 91:9, 78:22, 61:39, 37:63 and 0:100 on a VS basis. For these conditions, C/N ratios were 12, 12, 13, 14, 15, and 16, while the alkalinity values were 3,185, 2,770, 2,355, 1,940, 1,525, and 1,110 mg CaCO3/L, respectively. Each bottle had a working volume of 150 mL, with seed sludge accounting for 30% of the volume. The final substrate concentration was adjusted to 20 g COD/L in all conditions. The initial pH of each sample was adjusted to 8.0 ± 0.1 using 3 N KOH or 3 N HCl before sealing.
2.2.2. TFWA test
The TFWA test was designed to assess whether small additions of FW could enhance SM digestion by stimulating microbial and enzymatic activity without significantly increasing the organic loading. In this test, the SM loading remained constant, while FW slurry was added in trace amounts (i.e., 1, 2, 4, and 8 mL, corresponding to approximately 0.7%, 1.3%, 2.7%, and 5.3% of the working volume, or 0.9%, 1.8%, 3.6%, and 7.0% on a VS basis). Each bottle had a working volume of 150 mL, with seed sludge accounting for 30% of the volume, and the SM concentration was adjusted to 20 g COD/L, consistent with the conditions applied in the CRO test. Under these conditions, the final COD concentrations ranged from 20.0 to 21.6 g COD/L, indicating an incremental increase of less than 2 g COD/L even at the highest FW addition level, while C/N ratios varied slightly from 4.75 to 4.65 and Total Kjeldahl Nitrogen (TKN) ranged from 4.21 to 4.65 g N/L. The initial pH of each sample was adjusted to 8.0 ± 0.1 using 3 N KOH or 3 N HCl before sealing.
2.3. Analytical Methods
Biogas production was measured daily using a gas-tight glass syringe (LGC50CCT, Lab SciTech, CA, USA). Methane content was analyzed using gas chromatography (GC) (SRI 310, SRI Instruments, CA, USA) equipped with a thermal conductivity detector (TCD) and a HayeSep T column, using nitrogen (99.999%) as the carrier gas at a flow rate of 10 mL/min. Standard methods (APHA, 2005) were applied to measure COD, TS, VS, alkalinity, and pH. For the analysis of organic acids, ammonia, fluorescence excitation-emission matrix (FEEM), dissolved organic carbon (DOC), and molecular weight (MW) distribution, samples were collected at the end of the AD process, then centrifuged at 716×g for 10 minutes (FLETA-5, Hanil, Seoul, South Korea) and sequentially filtered through GF/C (1.2 μm) and 0.45 μm syringe filters. Organic acids were quantified using high-performance liquid chromatography (HPLC) (Nexera-i LC-2040C 3MT Plus, Shimadzu, Kyoto, Japan) with a UV detector at 210 nm and an Aminex HPX-87H column. Ammonia was measured by ion chromatography (IC) (790 Personal IC, Metrohm, Herisau, Switzerland). FEEM spectra were obtained using a spectrofluorophotometer (RF-6000, Shimadzu, Kyoto, Japan) over excitation wavelengths of 250–600 nm and emission wavelengths of 220–400 nm. DOC was quantified using a TOC analyzer (TOC-L, Shimadzu, Kyoto, Japan), and MW distribution was measured via HPLC-SEC (LC-20A series, Shimadzu, Kyoto, Japan).
2.4. Microbial Community Analysis
Microbial community analysis was conducted only for the TFWA test to investigate how trace FW supplementation affects bacterial and archaeal communities at the end of digestion. From each reactor, 50 mL of digestate was collected and centrifuged at 3,000 rpm for 10 minutes. Approximately 2 mL of the resulting pellet was used for genomic DNA extraction using the Fast DNA Spin Kit for Soil (QBioGene, Carlsbad, CA, U.S.A). The extracted DNA was further purified with the UltraClean Microbial DNA Isolation Kit (Mo Bio Laboratories, Carlsbad, CA, U.S.A) to ensure high-quality templates for downstream processing.
The 16S rRNA gene was amplified using the Illumina MiSeq™ sequencing platform (Illumina, San Diego, CA, U.S.A). Universal bacterial primers 27F (5′-GAGTTTGATCMTGGCTCAG-3′) and 518R (5′-WTTACCGCGGCTGCTGG-3′) were employed to target the V1–V3 hypervariable regions of the 16S rRNA gene. The PCR reaction was performed using 5 ng of template DNA, 5× reaction buffer, 1 mM of dNTPs, 500 nM of each primer, and Herculase II fusion DNA polymerase (Agilent Technologies, Santa Clara, CA, USA). Thermal cycling conditions included an initial denaturation at 95°C for 3 minutes, followed by 25 cycles of denaturation (95°C for 30 s), annealing (55°C for 30 s), and extension (72°C for 30 s), ending with a final elongation step at 72°C for 5 minutes.
The resulting amplicons were purified using AMPure XP magnetic beads (Agencourt Bioscience, Beverly, MA, U.S.A). For library indexing, a secondary PCR was conducted using Nextera XT Index primers under similar reaction conditions, except for 10 amplification cycles. The indexed products were purified once more using AMPure beads.
Quantification of the final library was performed using a qPCR-based KAPA Library Quantification Kit for Illumina platforms (Roche, Basel, Switzerland), and fragment size distribution and quality were verified using the TapeStation D1000 ScreenTape system (Agilent Technologies, Waldbronn, Germany). The prepared libraries were then subjected to paired-end sequencing on the MiSeq™ platform. Sequence data were processed and analyzed using the MOTHUR pipeline, following the standard operating procedures [26]. Quality filtering, taxonomic classification, and operational taxonomic unit (OTU) clustering were performed to assess microbial diversity and minimize sequencing artifacts.
2.5. Statistical Analysis
Methane production was calculated using a mass balance approach based on headspace gas composition and total biogas volume produced during both CRO and TFWA tests. The modified Gompertz equation, Eq. (1), was used to fit the cumulative CH4 production data, allowing for the prediction of CH4 production rates [27].
where M(t) is the cumulative CH4 production (mL) at time t (days), P is the ultimate gas production (mL), Rm is the maximum gas production rate (mL/day), λ is the lag period (days), and e is Euler's number (2.71828).
Under the neutral pH conditions typically observed in AD, the carbon dioxide–bicarbonate buffering system plays a primary role in maintaining pH stability. When organic acids accumulate, they partially neutralize bicarbonate ions, leading to a shift in the system’s buffering capacity. In such cases, the bicarbonate alkalinity can be estimated by subtracting the acidity contributed by the organic acids from the total alkalinity, as expressed in Eq. (2), following the method proposed by McCarty [28].
where BA is the bicarbonate alkalinity (mg/L as CaCO3), TA is the total alkalinity (mg/L as CaCO3), TVA is the total volatile acid concentration (mg/L as acetic acid).
After the batch test for methane production via AD, the final COD removal and alkalinity were measured. The COD removal efficiency was determined by comparing the initial and final COD concentrations for each experimental condition. Specifically, it was calculated by subtracting the final COD value from the initial COD, as shown in Eq. (3).
where CODinitial and CODfinal represent the COD concentrations (mg COD/L) measured at the beginning and at the end of the batch test, respectively.
The synergistic effect of TFWA was assessed for each mixture based on specific methane yield compared to the SM-only condition. The synergy index was calculated as the ratio of observed to theoretical yields, where SM yield was derived from the SM-only condition, while FW yield was estimated from COD concentration using the stoichiometric equivalent of 350 mL CH4/g COD, as the trace-level FW addition constrained direct yield measurement. Therefore, synergy indices <1 and >1 indicate antagonistic and synergistic effects, respectively.
3. Results and Discussion
3.1. Methane Production and Buffering Response at Various Mixing Ratios of SM and FW
The CRO test was conducted to evaluate how varying mixing ratios of SM and FW affect methane production and buffering capacity. Methane yield trends were well described by the modified Gompertz model (R2 > 0.98), confirming the reliability of the observed trends across different mixing ratios (Fig. 1).
Under the SM-only condition, methane production was relatively low at 132 ± 2 mL. However, the addition of 20% FW (i.e., SM:FW = 80:20) significantly increased cumulative methane production to 463 ± 12 mL. The highest methane yield was achieved at a 60:40 mixing ratio, reaching 242 ± 12 mL/g COD, an approximately 5.5-fold improvement compared to the SM-only condition (Table 1). Although methane yield differed between the 20% and 40% FW mixing ratios, both conditions maintained stable pH and moderate levels of organic acids, indicating that the inherent buffering capacity of SM effectively mitigated acidification within this range. Although the C/N ratio is widely recognized as an important factor influencing biogas production in AD, the C/N ratios of the tested mixtures (12–16) in this study did not differ substantially, suggesting that C/N optimization may not be the primary mechanism driving the observed enhancement. Instead, the increased methane production at moderate FW mixing ratios may be more directly associated with the inherently high biodegradability of FW itself. Given that FW contains a greater proportion of readily degradable organic fractions compared to SM, the enhanced methane yield could simply reflect the additive contribution of this more biodegradable substrate, rather than synergistic interactions or other complex mechanisms. In contrast, when the FW proportion exceeded 40%, the sharp decline in methane production at higher FW proportions was more closely associated with organic acid accumulation and the resulting pH decrease, which likely suppressed methanogenic activity.
In the SM-only condition, total alkalinity and bicarbonate alkalinity were maintained at 8,898 ± 18 mg CaCO3/L and 8,046 ± 117 mg CaCO3/L, respectively, with a stable pH of 7.7 ± 0.1. Even with 20–40% FW addition, bicarbonate alkalinity slightly decreased, remaining above 6,600 mg CaCO3/L and maintaining a pH above 7.0 that is favorable for methanogenesis. However, with FW content of 60% or higher, bicarbonate alkalinity dropped to nearly zero, and pH dropped to 5.7, reaching as low as 5.0 ± 0.2 under the FW-only condition. These findings underscore the critical role of SM-derived bicarbonate alkalinity in maintaining AD stability.
Interestingly, even when bicarbonate alkalinity was nearly depleted, total alkalinity remained relatively high. For instance, total alkalinity was 4,503 ± 110 mg CaCO3/L at 60% FW, and it was 1,830 ± 14 mg CaCO3/L at 100% FW. This is attributed to the presence of organic acid, particularly weak acids such as acetic and butyric acids, which can accept hydrogen ions and contribute to total alkalinity measurements, even when bicarbonate levels are low.
As shown in Fig. 1b and summarized in Table 1, changes in pH and alkalinity across different SM:FW mixing ratios were closely linked to the accumulation patterns of organic acids. When the FW proportion was between 0% and 40%, total organic acid concentrations remained below 2,000 mg COD/L, suggesting effective conversion of intermediates into methane. However, when FW content exceeded 40%, total organic acid concentrations rose sharply, surpassing 6,000 mg COD/L, with significant accumulation of acetic and butyric acids. For instance, under the 80% FW condition, the total organic acid concentration reached 7,410 ± 323 mg COD/L, composed of 63% acetic acid and 26% butyric acid, indicating the dominance of acidogenic fermentation pathways.
Co-digestion of SM with moderate FW levels (20–40%) substantially improved methane production while maintaining buffering capacity and pH stability. However, excessive FW input led to acid accumulation, bicarbonate depletion, and process failure. These results raised a critical question: could the co-digestion benefits observed at moderate FW levels be replicated or even improved by adding trace amounts of FW without altering the bulk composition? To explore this, the following section investigates whether trace FW additions could stimulate microbial activity and enhance AD performance through a metabolic activation mechanism.
3.2. Enhancement of Methane Production through Low-dose FW Addition
Building upon the findings from the CRO test, the TFWA test was conducted to evaluate whether small quantities of FW could stimulate the degradation of SM through metabolic activation rather than by simply increasing the organic load. The primary goal was to determine whether FW could act as a metabolic activator, enhancing methane production beyond the contribution of its own organic content.
As shown in Table 2, the methane yield improvement was observed with trace FW additions. Under the SM-only condition, the methane yield was 97 ± 3 mL/g COD. With the addition of 1, 2, 4, and 8 mL of FW, the methane yield progressively increased to 117 ± 4, 132 ± 3, 145 ± 3, and 170 ± 9 mL/g COD. This improvement was accompanied by enhanced COD removal efficiency, which increased from 63 ± 5% to 81 ± 7%. However, these metrics based on yield alone could not distinguish whether the observed enhancement arose primarily from the biodegradation of the added FW itself or from trace FW stimulating the degradation of SM. To address this question, cumulative methane production was analyzed to quantify the contribution from each substrate. As shown in Fig. 2a, methane production under the SM-only condition was approximately 290 ± 10 mL. Remarkably, the addition of just 1 mL of FW increased methane production to 351 ± 12 mL, and further increased to 509 ± 30 mL with the addition of 8 mL of FW.
A synergy analysis was conducted to separate the direct substrate contribution of FW from any enhancement in SM degradation (Table 3). The actual methane production (A) was compared with the theoretical sum of methane from SM alone (B, 290 mL) and the maximum potential methane from the added FW (C), with synergy calculated as A − (B + C). For the C term, the theoretical maximum conversion factor of 350 mL CH4/g COD was used, representing the upper limit of methane generation from fully biodegradable organic substrates under ideal conditions. Even under this assumption, synergy values of 50, 84, 104, and 135 mL were observed for 1, 2, 4, and 8 mL FW additions, respectively. To account for the increasing organic loading with higher FW additions (2.0–14.4% of total COD), the synergy index was additionally calculated as the ratio of observed to theoretically anticipated specific methane yields. The calculated synergy indices ranged from 1.17 to 1.36, indicating 17–36% enhancement above theoretical expectations based on weighted-average yields. It should be noted that the actual biodegradability of FW is typically lower than this theoretical maximum, which would further increase the calculated synergistic effect. These results indicate that the enhanced methane production was primarily due to stimulated degradation of SM, rather than direct contribution from FW.
Additional insights into the underlying mechanisms were obtained by analyzing indicators of hydrolysis and acidogenesis. Total organic acid concentrations increased from 1,282 ± 129 mg COD/L (SM-only) to 2,942 ± 81 mg COD/L (1.6 mL FW addition) with propionic acid as the dominant fraction (76–91%), yet remained below 3,000 mg COD/L without triggering metabolic inhibition or process failure. The pH remained stable at 8.0 ± 0.1 across all TFWA Samples, indicating that SM provided sufficient alkalinity to buffer the increased organic acid production. The total ammonia nitrogen (TAN)/TKN ratio increased from 63% in the SM-only sample to 86% with 8 mL FW addition, indicating intensified protein hydrolysis.
DOC increased from 2,450 mg/L in the SM-only condition to 4,555 mg/L following the addition of 8 mL of FW, indicating enhanced solubilization of organic matter (Fig. 3a). MW distribution analysis further revealed that, although the <0.5 kDa fraction increased from 39% to 42%, all MW ranges (<0.5, 0.5–1, and 1–5 kDa) showed absolute concentration increases, suggesting a broad-based breakdown of complex macromolecules into more soluble, microbially accessible forms. FEEM analysis (Fig. 3b) provided further insights into the composition of these solubilized compounds, showing intensified signals in the aromatic protein-like region (Ex/Em = 270–280/335–350 nm). This fluorescence signature is characteristic of tyrosine- and tryptophan-containing proteins, which can originate from various sources including substrate hydrolysis and microbial metabolites [29, 30]. Notably, protein-based soluble microbial products (SMPs) in anaerobic systems typically exhibit MW ranges of 1–10 kDa (small peptides and protein fragments), while the <1 kDa fraction primarily comprises fermentation intermediates such as volatile fatty acids and amino acids from substrate degradation [31, 32]. The increase in DOC and SMPs across a wide MW spectrum suggests improved substrate availability and potential formation of fermentation intermediates, which could be further utilized by methanogenic populations. These combined results indicate that trace FW additions promote metabolic activation and intermediate solubilization without triggering acidification or system overload, thereby supporting stable and enhanced AD performance.
3.3. Functional Restructuring of Bacterial and Archaeal Communities under Trace FW Addition
To investigate microbial community response to trace FW addition into SM, 16S rRNA gene sequencing was performed using the MiSeq™ platform as part of the TFWA test. Digestate samples were collected from both SM-only and SM + FW samples at the end of the AD process. A total of 197,416 high-quality reads were obtained, and OTUs were clustered at 97% sequence similarity. Taxonomic classification was conducted at both the class and species levels, and taxa below 1.0% relative abundance grouped into the “others” category.
The observed enhancement in biogas production following FW addition was further analyzed in relation to structural and functional shifts within the microbial community, focusing on key transitions across hydrolysis, acidogenesis, and methanogenesis, potentially driven by synergistic microbial interactions [33].
3.3.1. Bacterial community restructuring and metabolic response
Class-level community analysis revealed significant shifts in bacterial community composition under the FW-supplemented condition (Fig. 4a). In the SM-only condition, Anaerolineae (30.6%) and Clostridia (26.8%) were dominant. These taxa are commonly involved in the degradation of carbohydrates and proteins under anaerobic conditions [34, 35], but their activity may be limited when the substrate is already partially degraded during storage and pre-treatment.
Upon FW addition, the relative abundance of Anaerolineae and Clostridia decreased to 23.5% and 16.3%, respectively, while fermentative groups such as Synergistia (9.1%), Tissierellia (11.5%), and Bacteroidia (4.8%) increased. The coordinated enrichment of Tissierellia, typically found in systems degrading complex proteinaceous materials, alongside Bacteroidia, commonly associated with protein- and polysaccharide-rich environments, and Synergistia, known for propionate-oxidizing and amino acid-fermenting capabilities, may reflect the microbial adaptation to the altered substrate composition and availability [36, 37]. These shifts suggest that FW supplementation promoted a functional restructuring of the bacterial community, enriching organisms better suited to hydrolyzing and fermenting complex substrates
Species-level heatmap analysis provided further details (Fig. 4b). Clostridium saccharoperbutylacetonicum, prevalent in the SM-only condition (0.122), decreased to 0.084 with FW addition. In contrast, Keratinibaculum paraultunense increased from 0.022 to 0.085 and Bacteroides eggerthii emerged at 0.023 (previously undetected). These organisms are known for their ability to degrade protein- and polysaccharide-rich substrates [38], and their enrichment suggests potential enhanced hydrolytic capacity by FW. Specifically, Bacteroides spp. is reported to encode diverse carbohydrate-active enzymes, suggesting a potential to deconstruct a wide range of complex polysaccharides [39]. FW AD is generally characterized by fermentation-dominated metabolism during the hydrolytic/acidogenic stages, generating organic acids that serve as key intermediates for downstream methanogenesis. Fermentation-related pathways are frequently reported as dominant functions in FW AD; notably, fermentative taxa including Bacteroides-affiliated populations are often observed among the dominant bacteria during the hydrolytic/acidogenic stages and have been linked to carbohydrate breakdown and organic acid formation [40]. Additionally, amino acid–degrading species such as Aminithiophilus ramosus and Aminobacterium thunnarium showed increased abundances from 0.019 to 0.030 and 0.002 to 0.019, respectively – likely in response to the elevated levels of free amino acids and nitrogenous compounds in FW [41]. The taxa enriched following FW supplementation have been commonly reported in manure-based anaerobic systems, and their increase co-occurred with physicochemical shifts suggesting altered organic-matter transformation. While some of these changes may reflect direct microbial or substrate inputs from FW itself, the observed responses appeared relatively large compared with the limited FW addition under the tested conditions. For example, DOC increased from 2,450 mg/L to 4,555 mg/L (Section 3.2), which is not fully explained by the direct contribution of FW alone. Organic acid accumulation also increased (1,282 to 2,942 mg COD/L), and absolute concentrations increased across all molecular weight fractions. Given that shifts in bacterial hydrolysis and fermentation can alter the availability of key intermediates for methanogenesis (e.g., acetate and H2), archaeal community dynamics were therefore examined to assess whether downstream methanogenic assemblages responded in parallel under FW supplementation.
3.3.2. Archaeal pathway transition in response to fermentative substrate shift
FW addition also led to the compositional changes in the archaeal community, as shown in the class-level profiles (Fig. 5a). In the SM-only condition, methanogens affiliated with Methanobacteria and Methanomicrobia comprised 46.7% and 50.3% of the community, respectively. Under FW-supplemented conditions, the proportion of Methanobacteria decreased to 33.8%, while Methanomicrobia increased to 62.0%.
Species-level data showed that the hydrogenotrophic methanogens, Methanobacterium palustre and Methanobacterium petrolearium, dominant under SM-only conditions (0.166 and 0.259, respectively), decreased to 0.123 and 0.174, respectively with FW addition (Fig. 5b). These species utilize H2, CO2, formate, and alcohols as substrates [42] and may have been outcompeted in environments enriched with acetate and other fermentative by-products.
In contrast, the relative abundances of acetoclastic methanogens Methanothrix harundinacea and Methanothrix soehngenii increased from 0.028 to 0.037 and from 0.369 to 0.416, respectively. Methanosarcina barkeri, a metabolically versatile methanogen capable of utilizing CO2, acetate, and one-carbon compounds, appeared under FW conditions with a relative abundance of 0.051. These compositional changes suggest a shift toward a broader methanogenic community, likely reflecting shifts in the profile and conversion of fermentation intermediates produced by the enriched bacterial community [43]. Although fermentation appeared active (with propionate being the dominant organic acid), acetate concentrations remained comparatively low and did not show pronounced accumulation. This pattern suggests that acetate, once produced, was likely turned over rapidly rather than accumulating as a stable pool. In this context, the enrichment of Methanothrix—an acetate-specialist methanogen often associated with efficient acetate scavenging at low concentrations—is consistent with a community configuration supporting stable intermediate turnover and process robustness [44].
The present results were obtained from batch-scale experiments using FW derived from a single source. Although this design allowed the dose-dependent responses of process stability, metabolite dynamics, and microbial community composition to be examined under controlled conditions, the compositional heterogeneity commonly observed in FW streams may influence the magnitude and kinetics of these responses in practical systems. In addition, batch experiments primarily capture short-term biochemical responses and do not fully reflect the operational dynamics of continuous AD processes. These factors should be considered when interpreting the applicability of the observed effects to full-scale systems. Further investigation using continuous reactor configurations and diverse FW feedstocks will be necessary to evaluate the robustness of the trace FW co-digestion strategy.
4. Conclusions
This study systematically investigated (i) the effect of FW mixing ratios on AD performance via the CRO test and (ii) the stimulatory impact of trace FW addition through the TFWA test. The CRO test identified an optimal SM to FW ratio of 60:40, which achieved a maximum methane yield of 242 ± 12 mL/g COD and maintained stable pH conditions (7.6–7.7), supported by high bicarbonate alkalinity (>6,600 mg CaCO3/L). However, higher FW proportions (≥60%) resulted in acid accumulation, a drop in pH (as low as 5.0), and system instability. In contrast, the TFWA test revealed that even small additions of FW slurry (i.e., 1–8 mL) significantly boosted methane production, reaching up to 509 ± 30 mL, with observed synergy values up to 135 mL beyond theoretical expectations. COD removal efficiency improved from 63% to 81%, and DOC concentrations increased from 2,450 to 4,555 mg/L. Enhanced solubilization was further supported by increases across all low- to medium-MW fractions and intensified protein-like fluorescence in FEEM spectra, suggesting elevated levels of SMPs. These physicochemical changes suggest improved substrate accessibility and intermediate formation that may facilitate methanogenesis. Microbial community analysis revealed concurrent shifts, with selective enrichment of fermentative bacteria (e.g., Bacteroides, Keratinibaculum) and acetoclastic methanogens (e.g., Methanothrix, Methanosarcina), suggesting potential metabolic linkages that may have facilitated enhanced hydrolysis and methanogenesis. These findings offer practical insights for implementing low-impact co-digestion strategies to enhance AD of hydrolytically limited substrates such as SM. Future studies should explore the long-term stability and scalability of trace co-substrate supplementation under continuous and pilot-scale operations to validate its real-world applicability.
PDF Links
PubReader
Full text via DOI
Download Citation 
Supplement
Print 
