The field experiment was conducted in a paddy field on campus farms at Anhui Science and Technology University in Chuzhou City, Anhui Province, China (32°87′N, 117°56′E) in 2023. The experimental area is in a subtropical humid monsoon climate zone with a mean annual temperature of 14.9°C and total precipitation of 904.4 mm. This area of China is typically recognized for the rice-wheat rotation systems in which the wheat growing season is from November of the previous year to June of the following year and the rice growing season from June to November. This experimental site was established in 2018, where fertilization experiments have been conducted with rice-wheat rotations for six years. In the paddy field, the clay soil had the following nutrient contents in 2023, respectively: 8.05 pH, 0.49 g·kg^-1 total N, 63.52 mg·kg^-1 alkali-hydrolyzable N, 0.51 g·kg^-1total phosphorus, 3.83 mg·kg^-1 rapidly available phosphorus, and 13.51 g·kg^-1 organic matter.

The experiment included six treatments: low N (90 kg N ha^-1) as urea (N1), organic fertilizer (O1), and 80% N from organic + 20% N from urea (N1O1); and high N (270 kg N ha^-1) as urea (N2), organic (O2), and 80% N from organic + 20% N from urea (N2O2) ( Table 1 ). The plots were arranged in a randomized complete block experimental design. Each treatment had three replicates (18 plots in total), each plot was 3.75 m². To avoid interference through the exchange of water and fertilizer, each plot was isolated using concrete bricks.

The field trial was conducted from July 1 to October 15, 2023, using rice cultivar Yangdao 6 transplanted at 19 plants·m^-2. Uniform phosphorus and potassium applications were maintained across all treatments: calcium superphosphate (12% P₂O₅) at 75 kg·ha⁻¹ and potassium sulfate (60% K₂O) at 150 kg·ha⁻¹ annually. Organic fertilizer, P and K were applied basally, while urea was split: 50% basal, 30% tillering, 20% grain-filling.

CH₄ samples were collected at the early stage of rice growth using static closed chambers with length 50 cm, width 50 cm and height 50 cm, while at the late stage of rice growth with length 100 cm, width 50 cm and height 50 cm, respectively, from 08:00-11:00 a.m. Four corrosion-resistant steel enclosures were embedded 10 cm below the soil surface seven days preceding initial sampling, remaining installed throughout the study. Each chamber’s exterior received dual thermal regulation: rubber foam insulation coupled with reflective aluminum cladding to reduce radiative heat transfer during measurements. A 60 mL syringe facilitated timed gas extraction, capturing 40 mL chamber headspace at four intervals (closure initiation, +5, +15, and +30 min) with concurrent temperature monitoring. Acquired samples underwent immediate transfer to 12 mL vacuum-sealed containers, followed by < 24 h laboratory analysis via GC system (Agilent 7890, USA).

The calculation of CH₄ flux and its cumulative emission flux were described by Jin et al. (2024).

At rice physiological maturity, composite soil cores (0–20 cm depth) were obtained from experimental plots for physicochemical characterization. Three replicates surrounding each plot were well blended thoroughly as one sample. Repeat sampling 3 times for each of the above mixed soil samples. The samples used for the determination of soil nutrients were dried at room temperature, crushed, and sieved to pass through a 2 mm mesh. Another samples used for the determination of methyl-coenzyme M reductase alpha subunit (mcrA) and methane monooxygenase alpha subunit (pmoA) were frozen and stored at -80°C for subsequent determination and analysis. Elemental composition (C, N) quantification employed combustion analysis via a CHNS elemental analyzer (Elementar Vario EI, Germany). Mineral nitrogen speciation (NH₄ ⁺-N and NO₃⁻-N) determinations utilized continuous flow analysis technology (Skalar San++ System, Netherlands).

To quantify functional bacteria, genes of mcrA and pmoA were used as molecular markers to determine the copies of the above functional bacteria in rice soils during the harvest period. Quantitative PCR used the SYBR Green method with two primer pairs for mcrA and pmoA ( Table 2 ).

The determination method of effective panicle number, grains per spike, thousand-grain-weight, theoretical yields and quality (protein, gel consistency and amylose) was described by Jin et al. (2024). Theoretical yield were calculated according to (Jin et al., 2024), using Equation 1:

where, Y is the theoretical wheat yield (kg hm^-2), Pn is the spikes per hectare, Gn is the grains per spike, TKW is the thousand kernel weight”.

Statistical evaluations were conducted using IBM SPSS Statistics version 19.0 (IBM, Armonk, New York, NY, USA) and R software. The averages and standard errors were computed for both CH₄ flux and the associated environmental variables. A one-way ANOVA was employed to assess the significance of the observed data. Statistical significance was considered at P < 0.05. Additionally, both linear and nonlinear regression analyses were utilized to explore the relationships between CH₄ flux and environmental factors. Random forest models assessed the relationships between soil nutrient and CH₄ emission and functional gene using rfPermute package.

To evaluate the impacts of the application rate and type of N fertilizer on soil physicochemical properties, NH₄ ⁺-N, NO₃⁻-N, SOC and TN were measured. Except for NH₄ ⁺-N in O1 and O2, the content of NH₄ ⁺-N, NO₃⁻-N, SOC and TN in soil were significantly increased with the N application rate. Under equivalent nitrogen application rates, urea application exhibited significantly higher NH₄ ⁺-N, NO₃⁻-N and TN in soil (P < 0.05, Figure 1 ). In contrast, compared with urea, organic fertilizer alone significantly decreased the content of NH₄ ⁺-N, NO₃⁻-N and TN, while increased SOC content. Specifically, mean soil SOC content significantly increased by 77.6% and 68.9% in O1 and O2, respectively, relative to N1 and N2 (P < 0.05, Figure 1 ). Compared with urea or organic fertilizer alone, all measured parameters under organic fertilizer combined with urea fell between those of the above two treatments. These results suggest that organic fertilization may enhance the potential fertility of paddy soils but significantly reduce available nitrogen content during the current growing season.

To investigate the effects of N fertilizer practices on methanogens and methanotrophs potentials, the copy numbers of mcrA and pmoA were quantified via qPCR. Under low-N application rate, O1 significantly increased mcrA gene copies by 128.6% and 59.8% compared to N1 and N1O1, respectively (P < 0.05, Figure 2a ). Under high-N application, mean mcrA gene copies were 7.51×10⁶ copies g^-1 dry soil in O2, significantly higher by 149.9% and 68.7% than N2 and N2O2, respectively (P < 0.05). In addition, increased N application rates significantly enhanced mcrA gene copies under both low and high nitrogen conditions (P < 0.05). The response of pmoA to different N fertilizer practices was consistent with the trend of mcrA (P < 0.05, Figure 2b ).

During the whole rice growing season, CH₄ emissions ranged from 0.18 to 4.71 (mean of 1.33 ± 0.12) mg m^-2 h^-1, implying that the rice field soil was a net carbon source for atmosphere ( Figure 3a ). From July 8 to August 19, 2023, CH₄ emissions from paddy fields generally followed the first rising and then decreasing, and then to the harvest period, the variation trend of CH₄ emissions was the same as the above, with two peaks occurring at the tillering and heading stages. Further analysis of CH₄ emissions revealed that under low-N application, O1 increased CH₄ emissions by 4.42-fold compared to N1 and by 48.3% compared to N1O1 (P < 0.05, Figure 3b ). Under high-N application, O2 resulted in 3.37-fold and 64.4% higher CH₄ emissions than N2 and N2O2, respectively. Meanwhile, increased N fertilizer application also significantly increased CH₄ emissions (P < 0.05, Figure 3b ). In addition, increased N fertilizer application significantly stimulated soil CH₄ emission (P < 0.05, Figure 3b ).

Linear regression and random forest analyses were employed to explore associations between soil properties and mcrA gene copies. Linear regression showed that SOC was significantly positively correlated with mcrA gene copies (R² = 0.89, P < 0.001; Figure 4a ). In contrast, TN (R² = 0.47, P < 0.001), nitrate nitrogen (R² = 0.44, P < 0.001), and ammonium nitrogen (R² = 0.53, P < 0.001) were significantly negatively correlated with mcrA gene copies. The random forest model incorporating SOC, pH, nitrate nitrogen, ammonium nitrogen, and TN explained 93.2% of the variation in mcrA gene copies (P < 0.001; Figure 4b ), with SOC identified as the most influential factor. These results suggest that the elevated SOC content induced by organic fertilization was the primary driver of increased methanogenic potential and methane emissions.

To explore the effects of environmental factors on CH₄ emission fluxes in rice field, multiple statistical analyses were employed to identify the relationships between the soil SOC, TN, NH₄ ⁺-N, NO₃ ^–N, mcrA gene and CH₄ emission fluxes ( Figure 5 ). In the present study, linear regression analysis demonstrated a significant positive correlation between CH₄ fluxes and mcrA gene copies (R² = 0.96, P < 0.001) and SOC content (R² = 0.97, P < 0.001). Conversely, CH₄ fluxes were negatively correlated with TN (R² = 0.49, P < 0.01), NH₄ ⁺-N (R² = 0.53, P < 0.001) and NO₃ ^–N (R² = 0.43, P < 0.01). The above findings indicated that the N fertilizer type exerted a greater influence on CH₄ emissions than the N application rate.

To comprehensively evaluate the agronomic performance of different N fertilizers practices, rice yield and quality parameters were analyzed. Results indicated that organic fertilizer combined with urea achieved the highest rice yield across all yield-related metrics (e.g., grain filling rate, thousand grain weight, and theoretical yield) under equivalent nitrogen application rates ( Figure 6 ). Conversely, organic fertilization alone yielded the lowest production. Notably, rice yield under low-nitrogen mixed fertilization was comparable to that under high-nitrogen urea fertilization, suggesting the feasibility of maintaining high yields while reducing nitrogen inputs through optimized fertilization strategies.

Regarding grain quality, organic fertilizer combined with urea significantly increased amylose content and gel consistency compared to urea fertilization (P < 0.05; Figure 7 ). Protein content was significantly lower under organic fertilization but higher under mixed fertilization. These findings demonstrate that organic-urea mixed fertilization represents the optimal strategy for balancing rice yield and quality.

N fertilization management practices significantly influenced CH₄ emissions in rice fields (Liang et al., 2017; Li et al., 2018; Liu et al., 2019). In the present study, CH₄ emissions exhibited distinct seasonal patterns by N fertilization application and prominent CH₄ peaks were observed at the stages of heading and tillering, and similar results have been observed in previous N fertilizer applications (Wu et al., 2018; Li et al., 2018; Miao et al., 2020). Such seasonal patterns of CH₄ emission have been well explained by the fluctuations of soil oxidation reduction conditions regulated by floodwater depth caused by irrigation, and the variations of the activity of methanotrophs and methanogens increased by C and N availability for soil microbes caused by topdressing (Tang et al., 2017; Wu et al., 2018; Liu et al., 2019; Miao et al., 2020; Wu et al., 2020). At the tillering stages, N fertilizer application could accelerate the rapid growth of rice plants and roots, and CH₄ emission was closely related to rice growth (Kim et al., 2016; Li et al., 2018). CH₄ emissions in rice fields were highly dependent on rice plant growth (Li et al., 2018), produced CH₄ is mainly transported by the aerenchyma of rice leaves, sheath and stems (Zhang et al., 2019). Additionally, rice roots provide an organic substrate for methanogens by the exudates or dead root biomass, which stimulates CH₄ production, and approximately up to 90% of CH₄ emissions were produced by the above pathways (Jia et al., 2001; Conrad, 2007; Liu et al., 2019; Zhang et al., 2019). Numerous studies have indicated that CH₄ emission was also strongly affected by soil moisture in rice cropping systems (Wu et al., 2018; Liang et al., 2017; Li et al., 2018). Irrigation practices following N fertilizer application in rice fields can make the soil moisture from the wet to saturated soil conditions, which will provide a better anaerobic environment for methanogenesis, stimulating the enhancement of soil methanogenic bacteria activity, decreasing methanotrophic archaea activity and promoting the production of soil CH₄ (Li et al., 2018; Liu et al., 2019; Wu et al., 2020). Similarly, in this present study, we also found that CH₄ emissions were higher at the stages of tillering and heading under irrigation conditions. Compared to the above stages, the relatively lower CH₄ emissions during the other rice growing periods were likely due to the less pronounced anaerobic conditions under intermittent flooding patterns. Fluctuations O₂ availability and oxidation reduction conditions in soil regulated by soil moisture caused by irrigation might have shifted the balance between methanotrophic CH₄ consumption and methanogenic CH₄ production leading to differences in CH₄ emission (Fang et al., 2014; Ran, 2016; Liang et al., 2017; Miao et al., 2020; Wu et al., 2020). The previous results of Zhou et al. (2015); Liang et al. (2017) and Wu et al. (2020) confirm the above phenomenon.

Furthermore, levels and types of N fertilizer were also significantly influenced CH₄ emissions. Similar to the results from other rice cropping ecosystems (Kim et al., 2016; Tang et al., 2017; Liu et al., 2019), increased N fertilizer application rates stimulated CH₄ emissions by 25.6%-55.7% in our study. Meanwhile, we also found that CH₄ emission was significantly correlated with soil available N content. Soil NH₄ ⁺-N and NO₃ ^–N in urea plots increased by 36.5% and 51.8% respectively, which the results that the copy number of the methanogens gene increased by 18.3%, and stimulated CH₄ emissions by 55.7%. Increased N fertilizer application rate not only enhanced soil water-filled pore space, with higher soil moisture but also promoted soil mineralization rates and N availability for soil microbes, directly affecting the rates of methanotrophs and methanogens activity to further influence CH₄ emission (Fang et al., 2014; Yue et al., 2016; Wu et al., 2020). The increased NH₄ ⁺-N following N fertilizer application could provide substrate for soil CH₄ production, increase the availability of NH₄ ⁺ to nitrifiers and stimulate the activities of methanogenic archaea and inhibit methanogens activity (King and Schnell, 1994; Wang and Ineson, 2003; Tang et al., 2017; Liu et al., 2019; Wu et al., 2020). Moreover, NH₄ ⁺ could interfere with the oxidation of CH₄, and NH₄ ⁺-N oxidation precedes CH₄ oxidation, because NH₄ ⁺ competes with CH₄ for CH₄ monooxygenase (Schimel, 2000; Yang et al., 2015), which promotes CH₄ emission in rice field.

Apart from the application rate of N fertilizer, organic fertilizer application will generally stimulate the CH₄ emission by 45%-252% in paddy soil (Zou et al., 2003; Tang et al., 2017; Miao et al., 2020). Compared with urea, organic fertilizer application in the experiments of low N and high N significantly increased CH₄ emissions by 442.1% and 337.3%, respectively, in the present study. The application of organic fertilizer significantly increased soil organic carbon by 68.8% to 77.6%, and increased organic matter can not only effectively provide abundant substrate for soil methanogens, but also consume a large amount of oxygen during the degradation process and formed soil anaerobic environment (Tang et al., 2017), which enhanced copy number of methanogens by 128.6% in O1 plots and 149.9% in O2 plots, respectively. Applied organic fertilizer can not only provide continuous N support for rice plants, but also significantly increase soil temperature and moisture, reduce soil oxygen content, provide better survival environmental conditions and sufficient substrate for methanogens, and promote CH₄ emission in rice fields at the later stage of rice growth when the paddy soil is relatively dry (Tang et al., 2017; Miao et al., 2020; Jin et al., 2024). However, of special interest is that organic fertilizer combined with urea reduced CH₄ emissions by 39.2-48.4% compared to organic alone, possibly due to less anaerobic conditions and enhanced methanotroph activity. Soil anaerobic environment is difficult to form in the experiment of organic fertilizer combined with urea, but the above experiment treatment can significantly enhance the activity, diversity and abundance of soil methanogens, resulting in a large amount of CH₄ being oxidized to CO₂, and reducing CH₄ emissions in paddy fields (Zheng et al., 2008; Shao et al., 2022).

Organic fertilizer or combination with urea fertilizer has the potential to avoid the environmental damage caused by excessive urea fertilizer application (Kakar et al., 2020). Therefore, it is important to evaluate the impact of organic fertilizer alone or in combination with other fertilizers on yield potential and rice grain quality to secure food supply. In this study, compared with N1, O1 and N1O1, the additional application of N fertilizers significantly increased rice yield by 61.6%, 49.2% and 31.9%, respectively. In addition, rice yield in N1O1 plot was significantly increased by 49.8% and 128.5% compared with N1 and O1 plot, and N2O2 plot increased significantly by 22.3% and 102.1% compared with N2 and O2 ( Figure 6 ). Numerous preceding studies demonstrated increasing trends in grain yields with higher N application rates or organic fertilizer combined with urea (Miao et al., 2020; Zhang et al., 2020; Hu et al., 2021, 2022; Jin et al., 2024). Meanwhile, increased N application rates also significantly enhanced the soil available N content such as NH₄ ⁺-N and NO₃ ^–N ( Figure 1 ). The increased NH₄ ⁺-N and NO₃ ^–N in soil can improve effectively the net photosynthetic rate and chlorophyll content, enhance N supplies for the grain filling stage and promoted rapid the development of branches and roots, which beneficial for rapid growth rice plants and improving crop production (Li et al., 2018; Zhang et al., 2019; Tang et al., 2020; Zhang et al., 2020; Li et al., 2023; Jin et al., 2024). Chemical fertilizers and organic fertilizers exhibit different N release characteristics (Peng et al., 2010). Chemical fertilizers such as urea had a higher nutrient release rate in the early stage of application, but caused high nutrient loss that synchronized with the crop nutrient requirement (Peng et al., 2010), resulting in insufficient N supplies during the later stages of crop growth (Jin et al., 2024). On the contrary, organic fertilizer had a lower nutrient release rate during the growing season which may cause insufficient N supply to the grain filling stage of rice (Jin et al., 2024; Song et al., 2024). Organic-urea mixed fertilization can give full play to the advantage of the two fertilizers to ensure adequate nutrient supply during the critical period of rice production (Zhang et al., 2023; Jin et al., 2024). Additionally, productive tillers are responsible and critical elements for rice production which can be affected by the rate or type of application of N fertilizer. Previous studies showed that the combined application of organic and inorganic fertilizers increased tiller number, spikelet number, thousand kernel weight, and yield, which grain yield was controlled by the above components (Kim et al., 2016; Moe et al., 2019; Tang et al., 2020; Zhang et al., 2020; Hu et al., 2021; Li et al., 2023). Consequently, the above explanation strongly indicates that under equivalent N applications, the rice yield of organic fertilizer combined with urea application was significantly higher than that of urea or organic fertilizer alone, which was very consistent with the results of this study.

Similar to grain yield, rice quality (e.g. amylose content, gel consistency, protein, head rice percentage, chalkiness and hot viscosity) was comprehensive traits controlled by the rate and type of N fertilizer application, which the contents of amylose, gel consistency and protein were essential elements to define the grain quality and nutritional value of rice (Ju et al., 2018; Kakar et al., 2020; Tang et al., 2020; Zhang et al., 2020; Zhao et al., 2020; Hu et al., 2021, 2022). Organic fertilizer combined with inorganic fertilizer not only ensured the continuous supply of nutrients at each key growth stage of rice (Zhang et al., 2023; Jin et al., 2024), but also increased the absorption and accumulation of potassium and the transfer of potassium in rice, which improved the appearance and milling quality of rice (Nie et al., 2016; Li et al., 2023). More importantly, the addition of organic fertilizer could increase the chlorophyll content, improve the photosynthetic rate of rice plants, promote the generation of photosynthetic products and the efficiency of transport to rice grains, which nutrients were fully gathered in rice grains, and thus improve rice quality (Li et al., 2023). In the present study, compared with urea or organic fertilizer alone, organic fertilizer combined with urea significantly increased amylose content, gel consistency, and protein ( Figure 7 ), which is consistent with the results of Gao et al. (2024). Together, these results suggest that organic-urea mixed fertilization is an effective method both for rice yield production and quality.

In the present study, although optimizing organic fertilizer combined with urea practices enhanced rice grain yield, improved rice quality and mitigated CH₄ emissions, the impacts of optimizing organic fertilizer combined with urea on the microbial activity of methanogens and methanotrophs at the species or genus scale have not been explicitly addressed, in particular, abundance and diversity of the above functional genes. Additionally, the application of organic fertilizer changed soil water-filled pore space, reduced soil oxygen content, and increased soil moisture and pH, which significantly stimulated CH₄ emissions, especially in the fallow stage of rice fields. Therefore, the lack of monitoring of CH₄ emissions from organic fertilizer combined with urea experiments during the non-growing season seriously affects the estimation of greenhouse gas inventories in rice-cropping ecosystems. Targeted research is needed to clarify the microbial activity of methanogens and methanotrophs to reveal the molecular biological mechanism of CH₄ production, and evaluate CH₄ emissions budgets during the growing season and the fallow stage of rice fields under organic fertilizer combined with urea practices in rice-cropping systems.