Chicken manure and cow manure were collected from local livestock farmers in Weifang, Shandong Province. The corn stover and wheat straw were also harvested from local cultivation areas and air-dried. The dried materials were ground using a grinder (FW135, Beijing Ever Bright Medical Treatment Instrument Co., Ltd., Beijing, China) and sifted using a 40-mesh sieve. The biogas slurry, used as inoculum, was obtained from an operational 6,000 m³ anaerobic digester (Shandong Luxi Dasheng Environmental Protection Technology Co., Ltd., Weifang) that primarily treats cow, chicken, and duck manure. Before use, the biogas slurry was pre-incubated under anaerobic conditions for 2 weeks at 37 °C ± 0.2 °C for the mesophilic trial and 55 °C ± 0.2 °C for the thermophilic trial, respectively. The characteristics of all materials are presented in Table 1.

The experiment was conducted using MultiTalent 203 (Nova Skantek Instruments Co., Ltd., Beijing, China) to automatically record the methane yield. The fully automatic methane potential tester consists of an AD unit, an acidic gas adsorption module, a gas flow rate and a data acquisition module (Figure 1). The AD unit includes an electric-heated thermostatic water bath and 500 mL sealed glass bottles equipped with separate mechanical stirrers. The acidic gas adsorption module comprises 100 mL glass bottles, each containing 80 mL of 3 mol/L sodium hydroxide solution.

The experiment comprised two treatments: a mesophilic group (MT, 37 °C ± 0.2 °C) and a thermophilic group (HT, 55 °C ± 0.2 °C) with three replicates per treatment. Each bottle was loaded with 350 g (wet weight) of the substrate-inoculum mixture, resulting in a working volume of approximately 350 mL. Based on experience in engineering practice, the substrate was formulated on a dry matter basis as a mixture of chicken manure, cow manure, wheat straw and corn stalks in a 1.6:1:1:1 ratio, with a total carbon-to-nitrogen ratio (C/N) of 20.2. The total solid (TS) content was adjusted to 12% (semi-dry) using biogas slurry. The feedstock contained naturally occurring quinolone antibiotics, with initial concentrations of ciprofloxacin (24.6 μg/kg), enrofloxacin (379 μg/kg) and ofloxacin (467 μg/kg). After sealing, the bottles were incubated in a thermostatic water bath for a 45-day AD period. The mixture was agitated intermittently at 80 rpm for 30 s, followed by a 3-min rest period.

Samples were collected on day 10 and day 45, with three replicates being tested. The pH value was measured using PHS-3E (INASE Scientific Instrument Co., Ltd., Shanghai, China) instantly after halting the AD process by rapid cooling in an ice box. Aliquots of fermentation liquid from each bottle were collected and pooled with the samples from the same group, rapidly frozen in liquid nitrogen, and stored at −80 °C until microbial community analysis. The fermentation liquid supernatant was obtained by centrifugation (20,000 × g, 15 min, 4 °C) and subsequently analyzed for ammonia nitrogen, chemical oxygen demand (COD), and volatile fatty acid (VFA). TS were determined by drying 20 mL of fermentation liquid at 105 °C for 48 h in a laboratory oven (WGL-45B, Tianjin Taisite Instrument Co., Ltd., China). Subsequently, volatile solids (VS) were determined by combusting the dried residues in a chamber furnace (SX-4-10, Tianjin Taisite Instrument Co., Ltd., China) at 550 °C for 5 h. The residual fermentation liquid of each group was pooled by group and submitted to Anqiu Agricultural Product Quality and Safety Management Service Center (Weifang, China) for the analysis of three antibiotics (ciprofloxacin, enrofloxacin and ofloxacin). The analysis was performed using high-performance liquid chromatography-triple quadrupole mass spectrometry, following the standardized protocol outlined in the Chinese National Standard HJ 1399-2024 (2024). Elemental analysis was performed on a Vario MAX cube system (Elementar Analysensysteme GmbH, Germany) to quantify total carbon and nitrogen content. The VFA concentrations were analyzed following the method of Sun et al. (2009). Ammonia nitrogen concentrations were quantified via Nessler’s reagent spectrophotometry (HJ 535-2009, 2009, China), and COD was assessed using the fast digestion-spectrophotometric method (HJ/T 399-2007, 2007, China).

Genomic DNA was extracted from the fermentation liquid samples using the E. Z.N.A.® Soil DNA Kit (Omega Bio-tek, Norcross, GA, United States). Following extraction, DNA quality was evaluated by measuring purity and concentration with a NanoDrop 2000 spectrophotometer (Thermo Scientific, Waltham, MA, United States), and assessing integrity through 1% agarose gel electrophoresis. Next-generation sequencing of the 16S rRNA gene was performed by Shanghai Biozeron Biotechnology Co., Ltd. (Shanghai, China). For microbial community analysis, the V3-V4 hypervariable regions of the bacterial 16S rRNA were amplified using primers 341F (5′-CCTAYGGGRBGCASCAG-3′) and 806R (5′-GGACTACHVGGGTWTCTAAT-3′), while archaeal communities were targeted using primers Arch519F (5′-CAGCCGCCGCGGTAA-3′) and Arch915R (5′-GTGCTCCCCCGCCAATTCCT-3′). Sequencing was carried out using NovaSeq PE250 (Illumina, Kapa Biosciences, Woburn, MA, United States) following library construction with the NEXTFlex Rapid DNA-Seq Kit (‌Bioo Scientific, Austin, TX, United States).

Data was analyzed using SPSS Statistics 22 software (IBM, Armonk, NY, United States). Measurements of ammonia nitrogen content, pH value, COD, TS, VS, VFA and cumulative methane yield on day 45 (n = 3) were subjected to an independent-samples T-test. Significant differences were considered at P ≤ 0.05. GraphPad Prism 6 (GraphPad Software, San Diego, CA, United States) was used to create the graphs.

The efficiency of methanogenesis in anaerobic digestion is governed by both the extent of substrate decomposition and the conversion of intermediate compounds into methane (Ali et al., 2025). As shown in Figure 2, the mesophilic (MT) digester’s daily methane fluctuated from 1.05 to 3.09 mL/g·VS during the first 17 days, after which it increased gradually to a first peak of 7.34 mL/g·VS on day 26. Conversely, the thermophilic (HT) digester achieved its first peak on day 14, but at a lower yield of 3.43 mL/g·VS. These trends suggest a rapid process initiation under mesophilic conditions, compared to a brief lag phase under thermophilic conditions. The second peak in the MT group occurred on day 36 (3.89 mL/g·VS) and then declined. Although a similar pattern was observed in the HT group, its daily methane production remained consistently lower than that of the MT group. These results indicate that the mesophilic AD of the four agricultural residues was more favorable for methane production than the thermophilic condition, a finding further supported by the cumulative methane production data (Figure 2). Ultimately, the cumulative gas production of the mesophilic group was significantly higher than that of the thermophilic group (p = 0.010), reaching approximately twice the total volume over the 45-day operational period. This result aligns with the findings of Yu et al. (2025), who reported that the methanogenic potential of corn stover was lower under thermophilic conditions (52 °C) than under mesophilic conditions (37 °C). However, it contrasts with a study by Hupfauf et al. (2018) on the anaerobic co-digestion of corn straw and cow manure, which observed a higher methane yield at 55 °C than at 37 °C. This discrepancy, in which the optimal temperature varies across studies, collectively underscores the importance of selecting an appropriate temperature regime for process optimization.

To identify the key drivers of methane yield, correlations with key physicochemical parameters and microbial taxa were evaluated (Supplementary Figure S1). Methanogens play a pivotal role in biogas production, and the diversity of their methanogenic pathways significantly influences yield variations (Sudiartha et al., 2024). In line with this, our analysis revealed a positive correlation between cumulative methane yield and the relative abundance of Methanoculleus, a hydrogenotrophic archaeon. Its prevalence was higher under mesophilic conditions (2.76% of total archaea) than under thermophilic conditions (1.48%) on day 45, consistent with the higher methane yield observed at mesophilic temperatures. This suggests that Methanoculleus was an important contributor to methane production in the present system, where hydrogenotrophic methanogenesis dominated, which is consistent with the reported positive association between its abundance and enhanced methane production (Sudiartha et al., 2024).

The extensive use of antibiotics to enhance feed efficiency, prevent infections, and treat diseases is a ubiquitous feature of intensive animal production systems (Chen et al., 2025). They are excreted in urine and feces, leading to high concentrations in manure, with a substantial fraction remaining in their original form or as transformed metabolites (Zubair et al., 2023). Figure 5 illustrates the degradation rates of three antibiotics during anaerobic fermentation under different temperature conditions. For enrofloxacin and ofloxacin, the degradation efficiency of AD under high-temperature treatment exceeded that of mesophilic treatment by 8.05% and 7.92%, respectively. After 45 days of anaerobic fermentation, ciprofloxacin was completely degraded in both the MT and HT groups, likely due to the relatively low initial concentration (<25 μg/kg). These antibiotics belong to quinolones, a class of synthetic antimicrobial agents with strong bactericidal activity and broad-spectrum antibacterial properties (Wu et al., 2024). Anaerobic fermentation markedly reduced their concentrations, supporting the standardized use of the resulting organic fertilizer and minimizing potential adverse impacts on soil and plants. Overall, thermophilic AD achieved greater removal of quinolone antibiotics, consistent with Zahedi et al. (2022), who found thermophilic conditions superior for removing various veterinary pharmaceuticals. This alignment suggests a potentially common, thermally enhanced degradation mechanism (Zubair et al., 2023). Zhang et al. (2022) stated that key enzymes involved in quinolone biodegradation in AD systems, such as hydrolases (deaminases and peptidases), ligases, transferases, and lyases, are positively associated with the phylum Bacillota. Their activity is sensitive to the accumulation of propionic acid. Consistent with this finding, our results demonstrated that the mesophilic group exhibited both a lower relative abundance of Bacillota and a significantly higher concentration of propionic acid compared to the thermophilic reactor at day 45. This combination of an unfavourable microbial community structure and inhibitory metabolic conditions likely suppressed the relevant enzymatic activity, thereby explaining the lower enrofloxacin and ofloxacin degradation efficiency observed under mesophilic conditions.

Studies have shown that anaerobic co-digestion systems of straw with livestock and poultry manure optimize the microbial community composition and structure (Huang et al., 2024). This is critical because the acid-forming and methane-forming microorganisms in these systems have distinct physiologies and growth requirements. Since different bacterial species synergistically produce a diverse array of enzymes, the composition and diversity of the bacterial community in AD systems significantly influence the extent of substrate degradation. Furthermore, the composition of methanogenic archaea strongly affects methane production, as acidolysis produces a variety of substrates that necessitate different methanogenic pathways for efficient methanogenesis (Niya et al., 2024).

In the present study, pretreatment and sequencing generated 727 bacterial operational taxonomic units (OTUs) and 612 archaeal OTUs across all samples. Among the alpha diversity indexes, the bacterial Chao1 index of the MT group was 1.93 and 2.52 times higher than that of the HT group on days 10 and 45, respectively (Figure 6). They exhibited a similar temporal pattern in bacterial richness, with higher species numbers in the early stage and lower richness in the late stage. It is widely recognized that greater microbial diversity in AD systems is associated with enhanced biogas production (Kong et al., 2018). Although the archaeal Chao1 indices were higher in the HT group, the greater bacterial diversity under mesophilic conditions likely played a more pivotal role. Diverse bacterial communities supply a broader spectrum of metabolic intermediates (e.g., hydrogen, formate, acetate) that are essential substrates for various methanogenic archaea (Finn et al., 2023). Therefore, the MT environment was more conducive to maintaining the complex syntrophic networks necessary for efficient methanogenesis, which may explain the higher methane production observed in the MT group.

Based on the bacterial Shannon index, the species evenness in the thermophilic group was higher than in the mesophilic group on day 10. However, by day 45, evenness in the HT group had decreased to a level below that in the MT group, which exhibited an increase over time. This pattern suggests that high temperature imposed a strong environmental filter on the bacterial communities, selecting for a subset of thermotolerant species. On the 10th day of AD, the archaeal evenness of the HT group showed slightly higher evenness than the MT group, but by day 45, the two groups exhibited nearly identical evenness. On day 10, the HT group harboured a community with high evenness and transient diversity. This was consistent with the reported decoupling between methanogenesis and biodiversity in AD (Finn et al., 2023). As fermentation progressed, however, sustained thermophilic conditions likely led to the gradual loss of less tolerant species, resulting in a community dominated by a few competitive, thermotolerant taxa. Additionally, the Simpson index values over 0.859 across all samples, confirming that several dominant species constituted a large proportion of the microbial community. Given that efficient methanogenesis relies on the participation of multiple microbial species, and higher microbial abundance and diversity can facilitate more robust methane conversion (Huang et al., 2024).

The analysis of diversity indices revealed distinct thermal responses between bacterial and archaeal communities. High temperature acted as an environmental filter, reducing bacterial richness while promoting archaeal diversity. This selective pressure ultimately shaped a functionally streamlined yet less resilient thermophilic community, dominated by specialist taxa adapted to the dual stresses of heat and potential ammonia. In contrast, the mesophilic system fostered a functionally robust and diverse community, characterized by a higher proportion of generalist taxa and complex interaction networks (Supplementary Figure S2), thereby explaining its superior and more stable methane production. To explore the specific differences in bacterial and archaeal communities at different taxonomic levels induced by temperature, the following sections provide a detailed analysis of community composition at both the phylum and genus levels for each domain.