Effect of Conventional Nitrogen Fertilization on Methane Uptake by and Emissions of Nitrous Oxide and Nitric Oxide from a Typical Cropland During a Maize Growing Season

Abstract

Conventional nitrogen fertilization in a maize cropping system enhances the soil’s methane (CH₄) sink but exacerbates emissions of nitrous oxide (N₂O) and nitric oxide (NO). This study demonstrates that conventional nitrogen application (UN) increased CH₄ uptake by 154%, while elevating N₂O and NO emissions by 190% and 301%, respectively, compared to zero nitrogen plots (N0). Fertilization fundamentally reconfigured the regulatory mechanisms governing gas fluxes: under UN, fluxes were controlled by a complex interplay of nitrogen substrates, carbon availability, moisture, and temperature, whereas under N0, CH₄ uptake exhibited significantly enhanced temperature sensitivity (with Q₁₀ increasing from 1.06 to 7.54) and nitrogen oxide emissions became more dependent on native ammonium and extractable organic carbon. Crucially, nitrogen withdrawal reduced soil ammonium by 37.1% without altering non-nitrogen soil properties, including temperature, moisture, and labile carbon pools. Collectively, these findings are consistent with the concept of nitrogen saturation under conventional fertilization rates. Optimizing these rates presents a significant opportunity to mitigate greenhouse gas emissions and air pollution while improving nitrogen use efficiency, thereby aligning agricultural production with climate goals and public health objectives without destabilizing short-term soil function.

1. Introduction

Agriculture is a cornerstone of the national economy but also a significant contributor to greenhouse gas emissions [1]. Agricultural activities are responsible for approximately 50% of global methane (CH₄) emissions and around 60% of nitrous oxide (N₂O) emissions [2]. Non-carbon dioxide (non-CO₂) greenhouse gas emissions from agriculture are responsible for 10–12% of total anthropogenic greenhouse gas emissions worldwide [3], and this proportion is increasing annually [4]. In China, agriculture is a major source of non-CO₂ greenhouse gas emissions, accounting for more than 15% of the national total, with N₂O and CH₄ emissions contributing as high as 90% and 60%, respectively [5]. Addressing these emissions is therefore critical not only for environmental protection but also for achieving China’s ambitious “Dual Carbon” goals (carbon peak and carbon neutrality).

Mitigating these impacts requires sustainable agricultural practices that reduce greenhouse gas emissions without compromising food security. This complex challenge necessitates optimizing practices for both productivity and environmental performance. CH₄ and N₂O are among the most potent greenhouse gases after CO₂, with global warming potentials of 34 and 298 times that of CO₂, respectively [2,6]. Current research on CH₄ emissions in agroecosystems focuses on developing empirical models to predict emissions from rice paddies and other crop systems [7]. These models consider factors such as soil organic carbon, water management, and organic amendments. However, they often underestimate the magnitude of emissions due to the lack of sensitivity to site-specific variables like soil texture and planting methods. For N₂O emissions, research highlights the importance of microbial processes such as nitrification and denitrification in agricultural soils [8]. Researchers are exploring strategies to improve nitrogen use efficiency and reduce N₂O emissions through biological nitrification inhibitors and integrated nutrient management. Additionally, the use of organic fertilizers like vermicompost has shown potential in reducing N₂O emissions compared to conventional fertilizers [1]. Studies also emphasize the role of soil moisture, temperature, and pH in regulating N₂O emissions [9]. Regarding nitric oxide (NO) emissions, research is still emerging, but it is recognized that synthetic fertilizers and livestock management from agriculture elevate atmospheric NO levels [10]. Efforts to reduce NO emissions focus on optimizing fertilizer application and adopting conservation practices such as reduced tillage and cover cropping [11]. Overall, sustainable land management practices, including crop rotation, organic fertilization, and integrated pest management, offer significant potential to mitigate greenhouse gas emissions while maintaining agricultural productivity [1].

Nitrogen fertilizer application significantly influences the CH₄ exchange between agricultural soils and the atmosphere [12,13,14], with effects that vary and can even stimulate emissions [15]. For example, as evidenced by Li et al., the application of nitrogen fertilizer can stimulate CH₄ emissions in paddy fields, with increases ranging from 13% to 66% [16]. A recent meta-analysis conducted on agricultural land in China found that nitrogen fertilizer application increased CH₄ emissions and gross global warming potential by 32.5% [17]. Nitrogen fertilizers enhance plant growth and stimulate root growth, releasing more organic compounds into the soil, which serves as a substrate for CH₄-producing bacteria [15,18]. N fertilizers also alter the soil microbial community, promoting CH₄-producing microbes while reducing CH₄-consuming ones [19]. Additionally, they affect soil redox potential, creating anaerobic conditions that favor CH₄ production [20]. Finally, enhanced nutrient availability from nitrogen fertilizers boosts microbial activity and organic matter decomposition, leading to higher methane emissions. In others, CH₄ emissions are inhibited under nitrogen addition [21,22] through several mechanisms. For instance, in Chinese rice fields, Xie et al. demonstrated that typical application rates of ammonium-based fertilizers can inhibit cumulative seasonal CH₄ emissions by 28% to 30%, relative to a control with no nitrogen [23]. High levels of nitrogen can enhance the activity of CH₄-oxidizing bacteria, which convert CH₄ into CO₂, thus reducing emissions [22]. This nitrogen enrichment also fosters competition between CH₄-producing and CH₄-consuming microbes, with the latter often outcompeting the former [24,25]. Additionally, the promotion of plant and microbial growth by high nitrogen levels can create more aerobic conditions in the soil, which are unfavorable for methanogens that thrive in anaerobic environments [26]. Excessive nitrogen can also inhibit soil organic matter decomposition, reducing methanogen substrate availability [27]. This results in lower methane production. Finally, nitrogen-induced changes in soil pH can further inhibit CH₄-producing microbes [28]. In certain situations, there were no significant effects of nitrogen addition on CH₄ emissions [13,29]. In general, the effect of nitrogen addition on CH₄ emissions can vary based on factors such as soil type and texture, moisture content, temperature, organic matter content, the form of nitrogen used, plant type and growth stage, climate conditions, and existing soil nutrient levels [18]. These factors influence the interaction between soil nitrogen availability and microbes, affecting CH₄ production and oxidation. For instance, CH₄ emissions tend to increase under waterlogged soils and high temperatures [30], but may be reduced in well-drained soils or with certain nitrogen forms [31]. Understanding these variables is therefore crucial for developing nitrogen management strategies that mitigate CH₄ emissions from agricultural systems.

The use of nitrogen fertilizers alters the nitrogen cycle in the soil, which in turn affects the production of N₂O and NO [32,33]. The application of nitrogen fertilizers, such as ammonium sulfate or urea, directly increases emissions of both N₂O and NO from cultivated soils, with N₂O emissions rising proportionally to the application rate [34]. A global meta-analysis indicates that the relationship between nitrogen fertilizer and soil N₂O emissions is non-linear, following an exponential trend once inputs exceed crop demands [35]. This over-application represents a direct economic loss for farmers and contributes to a cascade of environmental and public health issues, including drinking water contamination from nitrate leaching and the formation of atmospheric particulates that impact respiratory health. However, the relationship between nitrogen fertilization and emissions of these gases is influenced by various factors that affect microbial processes in different ways. Soil physical properties, such as texture and structure, influence gas diffusion and water retention [36], thus impacting nitrification and denitrification. Climate conditions, including temperature and precipitation, affect soil moisture and microbial activity, with warmer temperatures accelerating these processes and increasing N₂O and NO emissions [37], thereby increasing N₂O emissions. Soil organic matter provides carbon for microbes, enhancing nitrification and denitrification, leading to higher emissions of these gases, especially when combined with nitrogen fertilization [38]. The form and application rate of nitrogen fertilizer are crucial, as ammonium-based fertilizers promote nitrification, while nitrate-based fertilizers facilitate denitrification, both increasing N₂O and NO emissions [32,39]. Agricultural practices such as tillage, crop rotation, and the use of nitrification inhibitors significantly impact emissions; reduced tillage helps maintain soil structure, while inhibitors slow down the conversion of ammonium to nitrate [40]. The type of crops grown and their growth stages also play a role; different crops and stages affect nitrogen uptake and root exudation, influencing soil nitrogen dynamics [41]. Additionally, existing soil nutrient levels determine how additional nitrogen impacts N₂O and NO emissions, with nutrient-rich soils exhibiting less pronounced effects compared to nutrient-poor soils as baseline nutrient levels determine the impact of additional nitrogen fertilization [42]. Collectively, these factors interact in complex ways to regulate N₂O and NO emissions from soils treated with nitrogen fertilizers.

As a cornerstone of national food security, the North China Plain ranks among China’s most vital grain-producing areas. It accounts for approximately 26% of the country’s total cultivated land, producing over 60% of its wheat and about 45% of its maize (analysis based on data from the National Bureau of Statistics, 2020; as summarized in Foley et al., 2011) [43,44]. The winter-wheat–summer-maize double cropping system is a classic, intensive production pattern in the North China Plain, characterized by high water and nitrogen inputs [45]. This system depends heavily on irrigation [46] (90–690 mm yr⁻¹) and synthetic fertilizers (550 to 600 kg N ha⁻¹ yr⁻¹) [47]. While historically supported by subsidies to ensure high yields, this level of input is financially burdensome and environmentally unsustainable. Excessive nitrogen use leads to environmental issues such as nitrate leaching, water pollution, soil salination, soil acidification, and elevated greenhouse gas emissions [48]. Despite its agronomic importance, the responses of CH₄ uptake, N₂O and NO emissions to N fertilization in wheat-maize rotation systems remain poorly quantified. Understanding these responses is essential for improving greenhouse gas emission inventories and reducing uncertainties in climate change projections. Moreover, quantifying these fluxes is a prerequisite for developing sustainable farming practices that can lower production costs for farmers, mitigate environmental harm, and align agricultural production with national climate targets.

Therefore, a field experiment comparing conventional nitrogen fertilization (usual nitrogen application rate, UN) with a zero-nitrogen control (zero nitrogen application rate for the current year, N0) was conducted in a typical irrigated maize field in northern China. This study provides a novel, concurrent assessment of the three dominant non-CO₂ greenhouse gases (CH₄, N₂O, and NO) and their underlying regulatory mechanisms within the same intensive cropping system. This integrated, multi-gas perspective is crucial for accurately evaluating the net climatic impact of management practices. Furthermore, by focusing on the maize phase of the dominant wheat–maize rotation in the North China Plain—a system critical to national food security but characterized by high nitrogen inputs—we address a significant regional knowledge gap. Based on the established role of nitrogen as a key substrate and regulator of soil processes, we hypothesized that: (i) conventional nitrogen fertilization would significantly enhance the soil’s CH₄ sink strength but would concurrently stimulate emissions of N₂O and NO; and (ii) the presence of fertilizer nitrogen would fundamentally alter the relationships between gas fluxes and soil factors. This study aims to (1) quantify the fluxes of CH₄, N₂O, and NO, and (2) identify the key soil environmental factors regulating these fluxes under the two contrasting nitrogen regimes to elucidate the mechanistic controls. By elucidating these dynamics, this research seeks to inform nitrogen management policies and practices that can mitigate agricultural greenhouse gas emissions, reduce the economic and environmental costs of farming, and support climate-resilient and sustainable food production in China.

2. Materials and Methods

2.1. Experimental Site

The study was conducted at the Jiyang Experimental Base of the Shandong Academy of Agricultural Sciences (36.984° N, 116.988° E), The site experiences a temperate climate with a mean annual air temperature of 14.7 °C and receives 671.1 mm of annual precipitation. The soil is classified as an aquic soil according to USDA Soil Taxonomy. The tested soil had a silty loam texture, composed of 18% clay (<0.002 mm), 69% silt (0.002–0.053 mm), and 13% sand (0.053–2 mm). Initial soil properties (0–20 cm depth) measured prior to the experiment were as follows: pH 8.62, bulk density 1.48 g cm⁻³, soil organic carbon 10.4 g kg⁻¹, total nitrogen 0.12%, total phosphorus 0.08%, available phosphorus 23.33 mg kg⁻¹, and available potassium 113.34 mg kg⁻¹. The experimental field had a long-term history of conventional fertilization (approximately 250–300 kg N ha⁻¹ yr⁻¹) under the winter wheat-summer maize rotation system. The area receives an atmospheric nitrogen deposition of 25 to 40 kg N ha⁻¹ yr⁻¹ [49,50].

The experiment employed a completely randomized design with two nitrogen fertilizer treatments: conventional fertilization (a usual nitrogen application rate, UN, 240 kg N ha⁻¹ yr⁻¹) and a zero-nitrogen application rate for the current year (N0, 0 kg N ha⁻¹ yr⁻¹). This binary design was selected to provide a clear, mechanistic understanding of the fundamental impact of fertilizer nitrogen addition on soil biogeochemical processes. While optimal for quantifying the nitrogen effect size on gas fluxes, it is acknowledged that this design does not define the full spectrum of production–environment trade-offs, which would require a multi-level nitrogen rate experiment. With four replications per treatment, the experiment comprised eight individual plots. Each plot (6 m × 8 m) was separated by a 1 m buffer zone to reduce cross-plot interference.

Maize (Zea mays L.) was sown on 16 June 2024, with a row spacing of 60 cm and a plant spacing of 20–22 cm, resulting in a planting density of 6–7 plants m⁻². Fertilization was carried out on 18 June 2024. Nitrogen in the form of urea was applied via banding into 15 cm deep furrows spaced 60 cm apart. A total of 13 furrows were created per plot, and the fertilizer was evenly distributed among them. Maize seeds were sown at a depth of 5 cm, positioned 5 cm away from and parallel to the fertilizer furrows. This was the sole nitrogen application for the growing season. Phosphorus, potassium, and zinc were co-applied with the urea at rates of 65 kg P₂O₅ ha⁻¹, 58 kg K₂O ha⁻¹, and 3 kg ZnSO₄ ha⁻¹, respectively, for all treatments. Irrigation was performed on 20 June 2024, and the maize was harvested on 14 October 2024.

2.2. Measurements of CH₄, N₂O and NO Fluxes

Gas fluxes were manually measured using gas chromatograph-based static opaque chamber method [51,52]. Measurements were taken daily for the first week after each fertilization event and then two to three times per week for the remainder of the study period (16 June to 14 October 2024).

Prior to seeding, stainless-steel base frames (length × width = 45 × 35 cm) were inserted approximately 20 cm deep in the center of each plot. The frames were only removed for tillage and seeding operations, remaining in place for the entirety of the study period. The chamber (length × width × height = 45 × 35 × 28 cm³), constructed from PVC material, was designed to accommodate varying crop heights.

To equilibrate headspace air pressure during deployment, each chamber was equipped with a 2 cm diameter aperture and a pressure balance tube. This tube was intermittently closed during gas sampling to prevent pressure imbalances and ensure accurate flux measurements [53,54]. Immediately before closing the chamber, ambient air was sampled as the initial time-zero reference. Chamber deployment times ranged from 20 to 40 min. During each deployment, five headspace gas samples were collected at 10-min intervals using 60 mL polypropylene syringes.

The CH₄ and N₂O concentrations were analyzed using an Agilent 7890A gas chromatograph (GC) (Agilent Technologies, Inc., Santa Clara, CA, USA) equipped with a flame ionization detector (FID, at 200 °C) for CH₄ and an electron capture detector (ECD at 330 °C) for N₂O [55]. High-purity nitrogen (N₂, 99.999%) served as the carrier gas. To prevent CO₂ interference during N₂O detection, the ECD cell was supplied with a make-up gas mixture containing approximately 10% CO₂ in N₂ [53,54]. The N₂O concentrations were calibrated hourly using a three-point series of certified reference gases (1.02 ppm N₂O in N₂, Air Products and Chemicals, Inc., Beijing, China).

For NO flux determination, two air samples (2.5–3.0 L each) were collected: one from ambient air immediately after chamber mounting and another from the chamber headspace after the final N₂O sampling [56]. Using a 12 V DC-powered air pump (N86KNDC, KNF Neuberger (Trenton, NJ, USA); flow rate: 2–3 L min⁻¹), samples were transferred to 5 L gas bags (Guangming Research & Design Institute, Liaoning, China) and analyzed for NO concentrations within one hour via a chemiluminescence NO–NO₂–NO_x analyzer (Model 42i, Thermo Environmental Instruments, San Diego, CA, USA) [57].

The fluxes of CH₄ and N₂O were calculated from concentration changes over time, using linear or non-linear regression of the five gas concentrations against chamber enclosure time [58]. NO fluxes were calculated based on the linear rate of concentration change within the chamber headspace, using a two-point linear method [59].

Based on the physical definition of a gas flux from a soil surface, the flux of a target gas was calculated using the following equation:

$$F=kr{\displaystyle \frac{V}{A}}{\displaystyle \frac{\rho P{T}_0}{P_{0}T}}$$

where F is the flux (μg C or N m⁻² h⁻¹) of a target gas, k = 60,000 is the dimension adaptation factor, r (=dC/dt or ΔC/Δt) is the concentration change rate within the chamber enclosure (μL L⁻¹ min⁻¹ or ppmv min⁻¹), V is the chamber headspace volume (m³), A is the soil surface area covered by the chamber (m²), ρ is the density (g L⁻¹) of the target gas at standard condition (273.15 K and 1013 hPa), P and P₀ are the headspace air pressure (hPa) at the actual and standard condition, and T and T₀ are the headspace air temperature (K) at the actual and standard condition, respectively.

When a non-linear regression curve of five concentrations (C, ppmv for CH₄ or N₂O) against sampling time (t) or a linear regression curve of 3 to 5 concentrations (C) against sampling time (t) was obtained, r = dC/dt was determined as the initial slope of the non-linear fit or the slope of the linear fit [58]. When only two concentrations were available, r was set equal to ΔC/Δt, assuming a linear concentration change over time within the chamber enclosure. For NO, we acknowledge that this linear assumption may on average underestimate a flux by up to 31% (ranging from 3% to 59% at the 95% confidence interval, CI) compared to a non-linear method [57]. Each hourly gas flux was considered representative of the daily value for a specific location replicate and was thus directly extrapolated to a daily total by multiplying by 24 [60].

2.3. Auxiliary Measurements

Soil temperature at 5 cm depth logged at 30 min intervals using HOBO sensors (UTBI-001, Onset, Bourne, MA, USA). Concurrently, topsoil moisture (0–6 cm depth) was measured manually during gas flux sampling with a portable frequency domain reflectometry probe (MPM-160, Jiangsu RDS Technology Co., Ltd., Changzhou, China). Then, 20 g soil samples were extracted with either 100 mL of 2 M potassium chloride solution to analyze the ammonium (NH₄⁺) and nitrate (NO₃⁻) or 100 mL of deionized water to determine the extractable organic carbon concentrations (EOC). The extracts were analysis with a continuous flow analyzer (San++ Continuous Flow Analyzer, Skalar Analytical B. V., Breda, The Netherlands) [52].

2.4. Statistical Analyses

All statistical analyses were performed using R [version 4.3.0]. The effects of conventional nitrogen fertilization (UN) versus the zero-nitrogen control (N0) on the fluxes of CH₄, N₂O, and NO, as well as on soil factors, were assessed using one-way analysis of variance (ANOVA).

To elucidate the relationships between gas fluxes and soil physicochemical factors, two complementary approaches were employed. First, linear and non-linear regression analyses were performed using the IBM SPSS Statistics 23 (SPSS Inc., Beijing, China) to model the dynamics of CH₄ uptake, N₂O emissions, and NO emissions as functions of soil temperature, moisture, NH₄⁺, NO₃⁻, and EOC concentrations for the UN and N0 treatments separately. The temperature sensitivity of gas fluxes was specifically quantified by fitting the data to the Arrhenius equation. The model was linearized via log-transformation of flux data to derive the Q₁₀ coefficient, which represents the fold-change in flux per 10 °C increase. This established method effectively isolates temperature effects from other variables and is validated for such biogeochemical processes [55,61]. For regression model selection, multiple candidate models (linear, exponential, Michaelis–Menten) were tested. The best-fitting model was selected based on a combination of biological plausibility, the highest coefficient of determination (R²) and statistical significance (p value), with preference given to models that provided the most parsimonious explanation of the observed relationships. Second, Pearson correlation analyses were performed using the cor function in the psych package (2.5.6) for R to examine the bivariate relationships between these variables within each treatment.

For all tests, statistical significance was evaluated at three levels: p < 0.1 was considered marginally significant, p < 0.05 was considered significant, and p < 0.01 was considered highly significant, where p denotes the probability of rejecting the true null hypothesis.

3. Results

3.1. Effects of Nitrogen Fertilization on CH₄, N₂O and NO Fluxes and Their Integrated Global Warming Potential

Conventional nitrogen fertilization (UN) significantly altered the fluxes of CH₄, N₂O, and NO compared to the zero-nitrogen treatment (N0) (

Table 1. Summary of gas fluxes and global warming potential under conventional nitrogen fertilization (UN) and a zero-nitrogen control (N0). Values shown are minimum, maximum, mean, and cumulative fluxes, the percentage change induced by fertilization (Effect %), and cumulative CO₂-equivalent emissions.

Gas	Treatment 1	N 2	Mean	Min	Max	Effect 3	p 4	Cum 5	Cumulative CO2-eq 6
			μg C or N m−2 h −1			%		kg C or N ha−1	kg CO2-eq ha−1
CH4	UN	50	33.62	0.39	92.51	154	<0.001	0.82 ± 0.07	−2.25 ± 0.19
	N0	51	13.25	1.72	46.56			0.30 ± 0.14	−0.82 ± 0.38
N2O	UN	56	12.07	1.71	81.90	190	<0.001	0.28 ± 0.02	74.20 ± 5.30
	N0	56	4.16	0.80	49.84			0.10 ± 0.04	26.50 ± 10.60
NO	UN	54	6.62	1.14	34.69	301	<0.001	0.15 ± 0.04	0.15 ± 0.04
	N0	57	1.65	0.15	9.81			0.04 ± 0.02	0.04 ± 0.02
Total Gaseous N Loss (N2O + NO)	UN							0.43 ± 0.04
	N0							0.14 ± 0.04
Net Total CO2-eq (CH4 + N2O + NO)	UN								72.10 ± 5.32
	N0								25.72 ± 10.61
Total CO2-eq Mitigation 7	N0 vs. UN								46.38 ± 11.89

¹ UN: Usual N application rate, N0: Zero N application rate for the current year. ² N denote sample size (number of observations. ³ Effect of conventional nitrogen fertilization determined by (F_UN − F_N0)/F_N0 × 100%, wherein the F_UN and F_N0 values are provided in the Mean column. ⁴ The significance levels (in p values), as shown in Figure 1, are resulted from one-way ANOVA tests for all the temporally and spatially replicated hourly fluxes. ⁵ Cumulated fluxes of a gas from UN and N0, respectively (unit: kg C or g N ha⁻¹), of which each is the arithmetic mean of four spatial replicates, and their standard errors. ⁶ Cumulative CO₂-equivalent emissions were calculated using 100-year global warming potentials: GWP for N₂O = 265, GWP for CH₄ = 28 (converting kg C to kg CH₄ by 16/12 gives a multiplier of ~34.1), and a conservative factor of 1 for the indirect forcing of NO. Negative values for CH₄ indicate uptake (a climate cooling effect). ⁷ Total CO₂-equivalent Mitigation represents the net reduction in global warming potential achieved by the N0 treatment compared to the UN treatment (i.e., Net CO₂-eq_UN−Net CO₂-eq_N0). The value is in kg CO₂-eq ha⁻¹.

, Figure 1). Specifically, UN enhanced the mean CH₄ uptake rate by 154% (p < 0.001). In contrast, it stimulated mean N₂O and NO emissions by 190% and 301%, respectively (both p < 0.001).

The temporal dynamics and cumulative fluxes further underscored these divergent effects (Figure 1 and Figure 2). The total gaseous nitrogen loss (as N₂O + NO) was threefold higher under UN (0.43 kg N ha⁻¹) than under N0 (0.14 kg N ha⁻¹).

Converting cumulative emissions to CO₂-equivalents revealed that the net greenhouse gas balance was significantly lower under N0 (25.72 kg CO₂-eq ha⁻¹) than under UN (72.10 kg CO₂-eq ha⁻¹) (Table 1). This difference of 46.38 kg CO₂-eq ha⁻¹ represents the total CO₂-equivalent mitigation achieved by forgoing nitrogen fertilizer. The overall global warming impact was overwhelmingly dominated by N₂O emissions in both treatments.

3.2. Effects of Nitrogen Fertilization on Soil Factors

Conventional nitrogen fertilization (UN) significantly altered soil nitrogen availability but had minimal impact on other measured dynamic soil factors (

Table 2. Soil physicochemical factors under conventional nitrogen fertilization (UN) and a zero-nitrogen control (N0). Values shown are minimum, maximum, and mean. The effect of fertilization (Effect %) and its significance are indicated.

Gas	Treatment 1	n	Mean	Min	Max	Effect (%) 2	p 3
Soil temperature (°C)	UN	46	29.09	23.20	39.53	−0.31	0.73
	N0	45	29.18	23.15	39.53
Soil moisture (%)	UN	55	27.72	6.15	42.40	−0.11	0.69
	N0	56	27.75	6.15	38.20
NH4+ (mg N kg−1)	UN	35	4.36	0.42	11.25	37.11	<0.01
	N0	37	3.18	0.42	11.15
NO3−	UN	38	2.49	0.11	13.94	−14.73	0.25
(mg N kg−1)	N0	37	2.92	0.09	9.09
EOC	UN	38	35.59	9.44	80.31	−0.25	0.97
(mg C kg−1)	N0	37	35.68	10.44	89.19

¹ UN: Usual N application rate, N0: Zero N application rate for the current year. ² Effect of conventional nitrogen fertilization determined by (F_UN − F_N0)/F_N0 × 100%, wherein the F_UN and F_N0 values are provided in the Mean column. ³ The significance levels (in p values), as shown in Figure 3, are resulted from one-way ANOVA tests for all the temporally and spatially replicated hourly fluxes.

, Figure 3). The most pronounced effect was a 37.1% increase in ammonium (NH₄⁺) concentrations under UN (4.36 mg N kg⁻¹) compared to N0 (3.18 mg N kg⁻¹), a difference that was highly significant (p < 0.01).

In contrast, NO₃⁻ concentrations and other key factors, including soil temperature, moisture, and EOC, remained stable across treatments, with no statistically significant differences (p > 0.05 for all). The temporal dynamics illustrated in Figure 3 reinforce these findings, showing consistent trajectories for non-nitrogen factors and a clear, sustained elevation of NH₄⁺ in the UN plots following fertilization.

3.3. Effects of Nitrogen Fertilization on the Relationship Between Gas Fluxes and Soil Factors

The presence or absence of conventional nitrogen fertilization fundamentally altered the relationships between gas fluxes and soil physicochemical factors, as demonstrated by the regression analyses (

Table 3. Linear or non-linear regressions of dynamical fluxes of methane (CH₄) uptake by and emissions of nitrous oxide (N₂O) and nitric oxide (NO) against soil factors from the investigated typical cropland during a maize growing season (All) with (UN) and without (N0) conventional nitrogen fertilization.

No. 1	Equations 2	Q10 3	N 4	R2 4	p 4
	All
(1)	$F_{{\mathrm{C}\mathrm{H}}_4}$ = $e^{-0.05{\mathrm{N}\mathrm{H}}_4-0.022{\mathrm{N}\mathrm{O}}_3+0.016\mathrm{E}\mathrm{O}\mathrm{C}+0.018M_{\mathrm{s}}+0.7}{e}^{0.059T_{\mathrm{s}}}$	1.80	49	0.07	<0.1
(2)	$F_{{\mathrm{N}}_2\mathrm{O}}$ = $e^{0.066{\mathrm{N}\mathrm{H}}_4-0.033{\mathrm{N}\mathrm{O}}_3+0.022\mathrm{E}\mathrm{O}\mathrm{C}+0.049M_{\mathrm{s}}-1.01}{e}^{0.022T_{\mathrm{s}}}$	1.25	53	0.15	<0.05
(3)	$F_{\mathrm{N}\mathrm{O}}$ = $e^{0.089{\mathrm{N}\mathrm{H}}_4-0.47{\mathrm{N}\mathrm{O}}_3+0.097\mathrm{E}\mathrm{O}\mathrm{C}-3.51M_{\mathrm{s}}+0.028}{e}^{0.058T_{\mathrm{s}}}$	1.79	53	0.27	<0.01
(4)	$F_{{(\mathrm{N}}_2\mathrm{O}+\mathrm{N}\mathrm{O})}$ = $e^{0.073{\mathrm{N}\mathrm{H}}_4-0.17{\mathrm{N}\mathrm{O}}_3+0.06\mathrm{E}\mathrm{O}\mathrm{C}-0.92M_{\mathrm{s}}+0.024}{e}^{0.031T_{\mathrm{s}}}$	1.36	53	0.20	<0.01
	UN
(5)	$F_{{\mathrm{C}\mathrm{H}}_4}$ = $e^{-0.05{\mathrm{N}\mathrm{H}}_4+0.062{\mathrm{N}\mathrm{O}}_3+0.019\mathrm{E}\mathrm{O}\mathrm{C}+0.034M_{\mathrm{s}}+1.9}{e}^{0.006T_{\mathrm{s}}}$	1.06	24	0.18	<0.01
(6)	$F_{{\mathrm{N}}_2\mathrm{O}}$ = $e^{0.025{\mathrm{N}\mathrm{H}}_4+0.048{\mathrm{N}\mathrm{O}}_3+0.023\mathrm{E}\mathrm{O}\mathrm{C}+0.051M_{\mathrm{s}}-0.30}{e}^{0.008T_{\mathrm{s}}}$	1.08	26	0.17	<0.01
(7)	$F_{{\mathrm{N}}_2\mathrm{O}}$ = 1.$11M_{\mathrm{s}}$${\displaystyle \frac{{\mathrm{N}\mathrm{H}}_4}{0.48+{\mathrm{N}\mathrm{H}}_4}}{{\displaystyle \frac{{\mathrm{N}\mathrm{O}}_3}{0.093+{\mathrm{N}\mathrm{O}}_3}}{\displaystyle \frac{\mathrm{E}\mathrm{O}\mathrm{C}}{15.08+\mathrm{E}\mathrm{O}\mathrm{C}}}e}^{-0.013T_{\mathrm{s}}}$	0.88	26	0.26	<0.01
(8)	$F_{\mathrm{N}\mathrm{O}}$ = $e^{0.05{4\mathrm{N}\mathrm{H}}_4-0.28{\mathrm{N}\mathrm{O}}_3+0.018\mathrm{E}\mathrm{O}\mathrm{C}+0.087M_{\mathrm{s}}-2.06}{e}^{0.038T_{\mathrm{s}}}$	1.46	26	0.29	<0.01
(9)	$F_{{(\mathrm{N}}_2\mathrm{O}+\mathrm{N}\mathrm{O})}$ = $e^{0.035{\mathrm{N}\mathrm{H}}_4-0.034{\mathrm{N}\mathrm{O}}_3+0.017\mathrm{E}\mathrm{O}\mathrm{C}+0.054M_{\mathrm{s}}+0.42}{e}^{0.008T_{\mathrm{s}}}$	1.08	26	0.20	<0.01
	N0
(10)	$F_{{\mathrm{C}\mathrm{H}}_4}$ = 0.87 $e^{0.089T_{\mathrm{s}}}$	2.44	38	0.20	<0.01
(11)	$F_{{\mathrm{C}\mathrm{H}}_4}$ = –0.57$M_{\mathrm{s}}$+ 29.33		50	0.14	<0.05
(12)	$F_{{\mathrm{C}\mathrm{H}}_4}$ = (0.89–0.001$M_{\mathrm{s}}$] $e^{0.089T_{\mathrm{s}}}$	2.44	38	0.20	<0.01
(13)	$F_{{\mathrm{C}\mathrm{H}}_4}$ = $e^{-0.18{\mathrm{N}\mathrm{H}}_4-0.11{\mathrm{N}\mathrm{O}}_3+0.016\mathrm{E}\mathrm{O}\mathrm{C}-0.018M_{\mathrm{s}}-2.87}{e}^{0.202T_{\mathrm{s}}}$	7.54	25	0.43	<0.001
(14)	$F_{{\mathrm{N}}_2\mathrm{O}}$ = 1.11$M_{\mathrm{s}}$${\displaystyle \frac{{\mathrm{N}\mathrm{H}}_4}{1.36+{\mathrm{N}\mathrm{H}}_4}}{{\displaystyle \frac{\mathrm{E}\mathrm{O}\mathrm{C}}{43.37+\mathrm{E}\mathrm{O}\mathrm{C}}}e}^{-0.03T_{\mathrm{s}}}$	0.74	27	0.30	<0.01
(15)	$F_{{\mathrm{N}}_2\mathrm{O}}$ = $e^{0.12{\mathrm{N}\mathrm{H}}_4+0.022{\mathrm{N}\mathrm{O}}_3+0.023\mathrm{E}\mathrm{O}\mathrm{C}+0.059M_{\mathrm{s}}-1.33}{e}^{-0.012T_{\mathrm{s}}}$	0.89	27	0.36	<0.001
(16)	$F_{\mathrm{N}\mathrm{O}}$ = 0.015$M_{\mathrm{s}}$${\displaystyle \frac{{\mathrm{N}\mathrm{H}}_4}{0.27+{\mathrm{N}\mathrm{H}}_4}}{{\displaystyle \frac{\mathrm{E}\mathrm{O}\mathrm{C}}{5.48+\mathrm{E}\mathrm{O}\mathrm{C}}}e}^{0.052T_{\mathrm{s}}}$	1.68	27	0.17	<0.01
(17)	$F_{\mathrm{N}\mathrm{O}}$ = $e^{0.04{\mathrm{N}\mathrm{H}}_4-0.03{\mathrm{N}\mathrm{O}}_3+0.012\mathrm{E}\mathrm{O}\mathrm{C}+0.067M_{\mathrm{s}}-4.56}{e}^{0.086T_{\mathrm{s}}}$	2.36	27	0.20	<0.01
(18)	$F_{{(\mathrm{N}}_2\mathrm{O}+\mathrm{N}\mathrm{O})}$ = $e^{0.089{\mathrm{N}\mathrm{H}}_4+0.019\mathrm{E}\mathrm{O}\mathrm{C}+0.058M_{\mathrm{s}}-1.69}{e}^{0.024T_{\mathrm{s}}}$	1.27	27	0.36	<0.001
(19)	$F_{{(Ln}_{{(\mathrm{N}}_2\mathrm{O}+\mathrm{N}\mathrm{O})})}$ = 0.044${Ln}_{{(\mathrm{N}\mathrm{H}}_4+{\mathrm{N}\mathrm{O}}_3)}$ + 0.016$M_{\mathrm{s}}$ + $0.003T_{\mathrm{s}}$		27	0.19	<0.01

¹ Equation ID referenced in results. ² $F_{{\mathrm{C}\mathrm{H}}_4}$, $F_{{\mathrm{N}}_2\mathrm{O}}$ and $F_{\mathrm{N}\mathrm{O}}$ denote the fluxes of methane (CH₄, μg C m⁻² h⁻¹) uptake and emissions of nitrous oxide (N₂O, μg N m⁻² h⁻¹) and nitric oxide (NO, μg N m⁻² h⁻¹). The involved factors include soil (5 cm depth) temperature (T_s, in °C), topsoil (0–6 cm depths) moisture content (M_s, in %) and concentrations (mg N or C kg⁻¹ d.s.) of soil (0–20 cm depth), ammonium (NH₄⁺), nitrate (NO₃⁻) and extractable organic carbon (EOC). ³ Q₁₀, a measure of temperature sensitivity, denotes the fold of a change in fluxes of a gas due to a 10-degree change in temperature. It is calculated from the Arrhenius exponent (β) using the equation: Q₁₀ = $e^{10\mathsf{\beta }}$. The error of an equation parameter or Q₁₀ indicates the uncertainty magnitude at the 95% confidence interval. ⁴ N, R² and p denote sample size (number of observations), determination coefficient and possibility to reject a null hypothesis for obtaining a regression curve.

) and correlation matrices (Figure 4).

For CH₄ uptake, the relationship was more complex under conventional fertilization (UN). A multi-factor exponential model incorporating NH₄⁺, NO₃⁻, EOC, moisture, and temperature provided the best fit (Equation (5), R² = 0.18, p < 0.01), with a low temperature sensitivity (Q₁₀ = 1.06). Notably, simpler models using only temperature or moisture were not significant for the UN treatment, confirming the necessity of a multi-factor approach. In contrast, under the N0 treatment, CH₄ uptake was most strongly predicted by a model incorporating NH₄⁺, NO₃⁻, EOC, moisture, and temperature (Equation (13), R² = 0.43, p < 0.001), which exhibited a very high temperature sensitivity (Q₁₀ = 7.54). This model was a substantially better fit than the simpler temperature or moisture-driven models (Equations (10)–(12), R² = 0.14–0.20), indicating that while temperature and moisture are key drivers in zero-nitrogen fertilized soil, they interact with other nitrogen and carbon substrates. This indicates that while multiple soil factors regulate CH₄ uptake in both treatments, the metabolism of methane-oxidizing microbes becomes acutely sensitive to temperature only when nitrogen fertilizer is absent.

The controls on N₂O emissions also shifted markedly. Under UN, emissions were best described by a multi-factor exponential model (Equation (6), R² = 0.17, p < 0.01, Q₁₀ = 1.08) or a more complex non-linear Michaelis-Menten-type model (Equation (7), R² = 0.26, p < 0.01). For the N0 treatment, the best-fitting model was a multi-factor exponential function (Equation (15), R² = 0.36, p < 0.001, Q₁₀ = 0.89), which incorporated NH₄⁺, NO₃⁻, EOC, moisture, and temperature. An alternative non-linear model for N0 (Equation (14) also performed well (R² = 0.30, p < 0.01), and both highlight that in the absence of fertilizer nitrogen, N₂O production is strongly governed by the availability of native soil nitrogen substrates and bioavailable carbon, exhibiting a muted response to temperature.

Similarly, the drivers of NO emissions differed between treatments. Under UN, a multi-factor exponential model was effective (Equation (8), R² = 0.29, p < 0.01, Q₁₀ = 1.46). For N0, the temperature sensitivity (Q₁₀) for NO was higher, as shown in Equation (17) (Q₁₀ = 2.36, R² = 0.20, p < 0.01). A non-linear model (Equation (16)) showed a comparable Q₁₀ of 1.68 but a lower R² (0.17), leading to the selection of the exponential model as more representative.

For the combined N₂O+NO emissions, the best model under N0 was a strong exponential function (Equation (18), R² = 0.36, p < 0.001), whereas the model for the entire dataset included NO₃⁻ as a negative regulator (Equation (4)). To gain deeper mechanistic insight into the controls on nitrogen oxide emissions, a theoretically grounded model was tested, integrating the core drivers of inorganic nitrogen availability, soil moisture, and temperature. This model was formulated based on the pore-space theory, which links gas diffusivity and substrate availability to soil moisture in driving nitrogen-oxide production [62,63], and the Arrhenius equation, which describes the temperature dependence of biochemical reaction rates [64] (Equation (19)). Although this theoretically derived linear model was statistically significant (p < 0.01), it explained a lower proportion of the variance (R² = 0.19) compared to the more empirical, non-linear exponential model (Equation (18), R² = 0.36). This suggests that the combined emissions of N₂O and NO are governed by more complex, non-linear interactions between soil factors than can be captured by this foundational theoretical framework alone.

The correlation analysis (Figure 4) provides a nuanced perspective that complements the regression findings. A key insight is that CH₄ uptake showed no significant linear correlations with any soil factors in the UN treatment, despite being part of a predictive multi-factor model. Under the N0 treatment, however, CH₄ uptake correlated positively with soil temperature and negatively with soil moisture. For nitrogen oxide emissions under N0, both N₂O and NO showed a significant positive correlation with EOC. This pattern underscores a fundamental shift in controlling factors, where in the absence of fertilizer, CH₄ uptake demonstrates clear, opposing dependencies on temperature and moisture, while N₂O and NO production is more closely linked to soil carbon availability.

In summary, conventional fertilization creates a system where gas fluxes are governed by a complex interplay of nitrogen substrates, carbon, and physical factors. Conversely, the absence of nitrogen fertilization simplifies the regulatory network, making the direct, opposing effects of temperature and moisture on CH₄ uptake, and the role of bioavailable carbon in nitrogen oxide emissions, more directly observable.

4. Discussion

4.1. Mechanistic Shifts in Gas Fluxes Induced by Nitrogen Fertilization

The transition from conventional (UN) to zero N application rate (N0) fertilization fundamentally restructured the control over gas flux, with distinct mechanistic shifts observed across CH₄ uptake and the emissions of N₂O and NO. The substantial reduction in mean CH₄ uptake under N0—by 154% compared to UN (Table 1)—underscores the critical role of nitrogen in sustaining methanotrophic activity, likely by supporting the synthesis of key enzymes such as methane monooxygenase or influencing microbial community composition [12,65]. Notably, it presents a divergence from the frequently reported suppression of methane oxidation in natural ecosystems following nitrogen addition [13]. The maize crop itself may resolve this apparent contradiction by mediating the soil physical environment through transpiration. Conventional nitrogen fertilization promoted vigorous crop growth, thereby enhancing this “bio-pump” effect that depleted soil moisture and improved soil aeration (as indicated by the lower mean soil moisture under UN, 27.72%, and N0, 27.75%, Table 2). This process is a well-established principle of plant–water relations [66] and is consistent with observations of vegetation-driven soil moisture dynamics in agroecosystems [67]. The resultant improvement in soil gas diffusivity, a key factor governing trace gas fluxes [68], would thereby facilitate the transport of atmospheric CH₄ and O₂ into the soil matrix, stimulating methanotrophic activity. This proposed mechanism is corroborated by our statistical models: under UN, CH₄ uptake showed no significant negative correlation with soil moisture (Figure 4b), and the coefficient for moisture in the multi-factor model (Equation (5)) was positive. This suggests that the fertilization-induced improvement in diffusion conditions effectively overrode the potential competitive inhibition between NH₄⁺ and CH₄ for the active site of methane monooxygenase, a suppression mechanism that predominates in many natural ecosystems [12].

Our regression analyses revealed a systematic simplification of the regulatory mechanisms—shifting from a multi-factorial control involving NH₄⁺, NO₃⁻, EOC, moisture, and temperature under UN (Equation (5), R² = 0.18) to a system where CH₄ uptake was predominantly governed by temperature under N0 (Equations (10) and (12), R² = 0.20). This shift was further reflected in the marked increase in temperature sensitivity (Q₁₀), which rose from 1.06 under UN to 2.44 under N0. The complex interactions under UN, including the absence of significant linear correlations between CH₄ uptake and any soil factor (Figure 4), are consistent with competitive inhibition between methanotrophs and ammonia-oxidizing microorganisms for oxygen or ammonia as a substrate for ammonia monooxygenase, an enzyme structurally similar to methane monooxygenase [12,69]. In contrast, the simplified and highly temperature-dependent kinetics under N0 suggest that nitrogen scarcity alleviates this competition, rendering the metabolic activity of methanotrophs more directly responsive to abiotic thermal conditions [19].

The pronounced reductions in N₂O (190%) and NO (301%) emissions under N0 (Table 1) empirically validate the central role of nitrogen fertilizer as the primary substrate for nitrification and denitrification processes [63,70]. Regression models further indicated a shift in the dominant controls between treatments.

Under UN, N₂O emissions were best described by a multi-factor exponential model (Equation (6), R² = 0.17) or a Michaelis–Menten-type model (Equation (7), R² = 0.26), indicating substrate-dependent nitrifier-denitrification or heterotrophic denitrification pathways commonly activated after fertilizer application [71]. Under N0, the best-fitting model for N₂O emissions was a multi-factor exponential function (Equation (15), R² = 0.36) that emphasized the importance of the native soil NH₄⁺ pool and EOC, with a lower Q₁₀ of 0.89. This suggests that in the absence of fertilizer, N₂O production becomes more reliant on mineralization of indigenous organic nitrogen and available carbon, with processes that are less sensitive to short-term temperature fluctuations [72].

Similarly, for NO emissions, the Q₁₀ increased from 1.46 under UN (Equation (8) to 2.36 under N0 (Equation (17). This elevated thermal sensitivity under nitrogen-limited conditions may indicate a relative increase in the importance of abiological processes such as chemodenitrification, which can become more prominent in low-N environments and exhibit strong temperature dependence [73]. Correlation analysis further supports these shifts, showing that under N0, both N₂O and NO emissions were significantly correlated with EOC (Figure 4c), highlighting the critical role of native soil carbon when fertilizer nitrogen is absent [74].

These findings affirm that nitrogen management is a key lever for mitigating agricultural greenhouse gas emissions. However, the increased temperature sensitivity of CH₄ uptake and NO emissions under N0 raises the hypothesis that the climate-regulating function of unfertilized systems could be more vulnerable to future warming—a critical consideration for long-term climate adaptation planning.

4.2. Stability of Non-N Soil Factors and Their Implications

This study demonstrates that zero-nitrogen application rate for the current year (N0) had no statistically significant impact on key non-nitrogen soil factors—namely, soil temperature, moisture, and EOC—compared to conventionally fertilized (UN) plots (Table 2, p > 0.05 for all). Soil temperature and moisture remained nearly identical between treatments (29.09 °C for UN vs. 29.18 °C for N0; 27.72% for UN vs. 27.75% for N0). This stability is consistent with previous findings that basic soil physical properties in intensively managed croplands are often buffered against short-term changes in nutrient inputs, as these parameters are primarily governed by plant canopy development and soil structure [27,75]. The absence of treatment effects on EOC (35.59 mg C kg⁻¹ for UN vs. 35.68 mg C kg⁻¹ for N0) further indicates that the labile carbon pool, largely driven by recent rhizodeposition and soil organic matter turnover, was not immediately altered by the absence of nitrogen fertilization over a single growing season [76].

The stability of these factors is critical for interpreting the observed gas flux changes. It indicates that the dramatic reductions in N₂O and NO emissions under N0 were a direct consequence of nitrogen substrate limitation for nitrification and denitrification, rather than being mediated by changes in soil physical conditions or labile carbon availability [70,77]. This stands in sharp contrast to the significant increase in NH₄⁺ under UN (4.36 mg N kg⁻¹ for UN vs. 3.18 mg N kg⁻¹ for N0, p < 0.01), indicating that nitrogen availability is the variable most directly and immediately impacted by the fertilization regime. The clear dissociation between nitrogen and non-nitrogen factors suggests a degree of functional resilience in the processes governing soil microclimate and labile carbon pools under short-term nitrogen shifts, possibly due to functional redundancy in microbial communities involved in carbon cycling [78,79].

However, the stability observed in this single-season study should not be equated with long-term resilience. Prolonged nitrogen limitation can indirectly affect these factors by altering plant biomass, root-shoot ratios, and litter quality, which subsequently affect soil moisture via evapotranspiration and soil organic matter dynamics [80,81]. For instance, sustained nitrogen deprivation often reduces plant productivity, potentially leading to lower soil organic carbon inputs and a gradual decline in EOC pools over time [82,83]. Furthermore, although soil temperature was unchanged, the significantly higher temperature sensitivity (Q₁₀) of CH₄ uptake under N0 (e.g., 7.54 in Equation (13) vs. 1.06 under UN in Equation (5)) indicates a fundamental shift in the metabolic response of the microbial community to temperature. This enhanced temperature sensitivity under nitrogen limitation may stem from a shift in the methanotroph community composition. This enhanced temperature sensitivity under nitrogen limitation may stem from a shift in the methanotroph community composition. Nitrogen limitation often favors K-strategist Type II methanotrophs [12], which could theoretically exhibit life-history strategies involving greater responsiveness to temperature fluctuations [65]. These findings underscore that while non-nitrogen soil factors remain stable in the short term—supporting reduced nitrogen fertilization as a viable strategy for immediate greenhouse gas mitigation—the long-term implications for soil biogeochemical cycling and climate feedback potential warrant careful, long-term evaluation.

4.3. Optimizing Nitrogen Management: Mitigation Potential, Economic and Social Benefits

While nitrogen fertilization is essential for sustaining high crop yields [84,85], excessive application reduces nitrogen use efficiency (NUE) [86], increases environmental pollution, and can lead to yield stagnation [87]. Our results directly substantiate these concerns: conventional fertilization (UN) significantly increased soil ammonium (NH₄⁺) concentrations by 37.11% (Table 2) and dramatically elevated emissions of the potent greenhouse gases N₂O and NO by 190% and 301%, respectively, compared to the N0 treatment (Table 1). This non-linear surge in gas fluxes in response to fertilization is consistent with the concept of nitrogen saturation under high input regimes [88,89]. Critically, the integrated CO₂-equivalent analysis reveals that this saturation creates a substantial climate cost: the net global warming potential of the UN treatment (72.1 kg CO₂-eq ha⁻¹) was 2.8 times greater than that of N0 (25.7 kg CO₂-eq ha⁻¹) (Table 1). This stark contrast, representing a mitigation potential of 46.4 kg CO₂-eq ha⁻¹, by simply avoiding fertilizer application, provides a direct and quantifiable pathway to reduce the climate footprint of maize production.

The environmental impact extends beyond climate forcing. The 301% reduction in NO emissions under the N0 treatment has important co-benefits for regional air quality and public health. NO is a key precursor to tropospheric ozone (O₃) and secondary fine particulate matter (PM_2.5), major pollutants associated with respiratory and cardiovascular diseases [90]. Therefore, reducing nitrogen fertilizer application to optimal levels can contribute to tangible improvements in air quality and human health in agricultural regions.

From an economic perspective, the conventional practice represents a dual financial loss. First, the nitrogen lost as N₂O and NO—0.43 kg N ha⁻¹ under UN, a threefold increase over N0 (0.14 kg N ha⁻¹)—is a direct waste of fertilizer input paid for by the farmer. Second, this wasted nitrogen is converted into a quantifiable climate cost, representing an unpriced environmental liability. The mitigation potential of 46.4 kg CO₂-eq ha⁻¹ provides a tangible metric that could form the basis for results-based policy instruments, such as integration into carbon credit markets or subsidies for low-emission practices. Furthermore, our empirically derived flux data and this quantified mitigation potential provide critical, region-specific emission factors that can directly enhance the accuracy of life-cycle assessments (LCA) and agricultural carbon-footprint analyses for maize production in the North China Plain. Reducing application rates would thus simultaneously lower input costs for farmers and generate a verifiable public good through climate mitigation and improved air quality.

Field trials in northern China’s intensive cropping systems have consistently demonstrated that current farmer practices can be significantly optimized. Studies show that nitrogen application rates can often be reduced by 30–60% without compromising maize yields [5,47,91]. Our findings provide a strong environmental rationale for such reductions. The sharp decline in nitrogen oxide emissions under N0, coupled with the stability of non-nitrogen soil factors (Section 4.2), indicates that moving away from conventional rates can immediately mitigate environmental harm without destabilizing the soil physical environment in the short term.

While a complete removal of nitrogen (N0) is agronomically unsustainable, our study robustly demonstrates that the upper bound of 240 kg N ha⁻¹ is environmentally and economically inefficient. The significant accumulation of NH₄⁺ under UN suggests that the conventional fertilization rate likely exceeds the maize crop’s physiological nitrogen requirement during the growing season. This observation aligns with established yield-response plateaus for maize in northern China, where optimal application rates commonly fall within the range of 150–200 kg N ha⁻¹ [47,92]. Precision nutrient management strategies—such as applying a reduced baseline amount coupled with in-season adjustments based on crop needs—can help achieve this balance, sustaining yields while improving NUE and reducing greenhouse gas emissions [93,94].

In summary, our findings confirm that conventional nitrogen application is supra-optimal, incurring unnecessary climate and economic costs. The demonstrated mitigation potential of 46.4 kg CO₂-eq ha⁻¹, combined with co-benefits for air quality and farm economics, provides a compelling rationale for policy-driven adoption of precision nitrogen management. This transition is critical for building climate-resilient agriculture, safeguarding productivity, and ensuring the long-term sustainability of food systems in the North China Plain.

4.4. Limitations and Future Research Directions

While this study provides valuable insights into the short-term effects of nitrogen fertilization on greenhouse gas fluxes and their driving factors, several limitations should be considered. First, the investigation was conducted over a single maize growing season. Therefore, the observed responses may not fully capture interannual variability or long-term trends in soil carbon and nitrogen cycling that could emerge under sustained N management regimes [95,96]. Second, a key limitation is the experimental design with only two nitrogen levels and the absence of grain yield data. This precludes a full economic and agronomic assessment of production–environment trade-offs and the calculation of critical sustainability metrics like emission intensity (yield-scaled emissions). Third, the regression models, though insightful, had low to moderate explanatory power (R² values), particularly for the combined dataset. This indicates that unmeasured factors, such as microbial community composition, enzyme kinetics, or finer-scale soil heterogeneity, also significantly influence CH₄ and N₂O fluxes. It is well established that gas production is not solely a function of bulk soil conditions but is strongly mediated by micro-scale microbial community structure and function [63,97]. The inability of measured abiotic variables to fully explain flux dynamics points to the critical need to integrate these biological factors into models [98]. Fourth, this study focused on gas fluxes and immediate soil physicochemical properties but did not include microbial or isotopic analyses. Such analyses are crucial for elucidating the underlying biological pathways, for example, differentiating between nitrifier-denitrification and fungal denitrification [70]. Accurately partitioning N₂O sources requires isotopic techniques to trace the origin of molecules [99], while understanding the microbial drivers necessitates linking fluxes to the abundance and activity of specific functional genes [63].

To build upon these findings, future research should pursue the following directions:

(i)

Agro-ecosystem trade-offs: Future studies should employ multi-level nitrogen rate experiments coupled with concurrent yield monitoring to establish the relationship between nitrogen input, crop productivity, and greenhouse gas emissions, thereby enabling the identification of rates that minimize emission intensity.

(ii)

Long-term temporal analysis: Multi-year studies are essential to assess the permanence of the observed gas flux reductions under N0 and to evaluate potential long-term consequences for soil fertility, organic matter dynamics, and the resilience of the methane sink [96].

(iii)

Mechanistic microbial investigations: Integrating molecular techniques (e.g., qPCR, amplicon sequencing, metagenomics) is needed to characterize how nitrogen fertilization regimes shape the abundance, diversity, and functional gene expression of methanotrophic, nitrifying, and denitrifying microbial communities [12,100].

(iv)

Process-level tracing: Employing stable isotope techniques (e.g., ¹⁵N tracing) would allow for the precise quantification of the contributions of different pathways (e.g., nitrification, denitrification) to N₂O and NO production, thereby moving beyond correlations to mechanistic understanding [63,101].

(v)

Evaluation of management strategies: Field trials should evaluate integrated management strategies—such as reduced nitrogen rates combined with enhanced-efficiency fertilizers (e.g., nitrification inhibitors, slow-release formulations), cover cropping, and precision split-application approaches—to identify the most effective methods for decoupling crop productivity from greenhouse gas emissions [92].

Addressing these limitations through a multidisciplinary approach—integrating agronomic, microbial, and climate sciences—is critical for developing predictive models and evidence-based policies that enable truly sustainable and productive agricultural systems in the face of global change.

5. Conclusions

This study indicates that conventional nitrogen fertilization (UN) in a typical maize cropping system leads to a state consistent with nitrogen saturation, creating a significant environmental and economic burden. While enhancing the soil’s methane (CH₄) sink, conventional management increased nitrous oxide (N₂O) and nitric oxide (NO) emissions by 190% and 301%, respectively, resulting in a net global warming potential 2.8 times higher than an unfertilized control. This quantifiable climate impact, alongside the substantial loss of applied fertilizer as gaseous pollutants, underscores an urgent need for reform.

Our findings provide a scientific basis for integrating agricultural nitrogen management into broader sustainability frameworks. Optimizing fertilizer use is not merely a technical agronomic issue but a critical strategy for achieving carbon neutrality in the agricultural sector. The significant mitigation potential—46.4 kg CO₂-eq ha⁻¹ for a single maize season—directly contributes to climate goals, while the concurrent reduction in NO emissions offers vital co-benefits for public health through improved air quality.

To realize this potential, a multi-pronged policy approach is essential. Implementation should be accelerated through: (i) Financially supported precision farming, including subsidies for enhanced-efficiency fertilizers and soil testing. (ii) Farmer training programs to promote site-specific nutrient management. (iii) Agri-environmental policies that create economic incentives, such as integrating verified emission reductions into carbon credit markets. Adopting such measures can make optimized nitrogen management a cornerstone of climate-resilient agriculture, which safeguards food security, farmer livelihoods, and environmental health simultaneously.