Role of Fungi in N₂O Emissions from Nitrogen-Fertilized Lawn Soil

Abstract

Urban lawns are a predominant form of vegetation in sports grounds and greenbelts. Nitrogen (N) fertilization is widely used to sustain lawn productivity. However, it also promotes nitrous oxide (N₂O) emissions, a potent greenhouse gas. The microbial mechanisms underlying N₂O emissions from fertilized lawn soils remain poorly understood. In this study, we conducted a controlled incubation experiment with four N application rates [0 (N0), 100 (N100), 200 (N200), and 300 kg·ha⁻¹·yr⁻¹ (N300)] to investigate N₂O emissions and associated microbial processes in urban lawn soil. Biological inhibitors combined with high-throughput sequencing were used to quantify the inhibitor-sensitive fraction of fungi and bacteria contributing to N₂O emissions. Our results showed that N fertilizer significantly increased N₂O emissions, with the highest emission observed under N200. The fungi inhibitor-sensitive fraction accounted for ~45% of total N₂O emissions, significantly higher than that of bacteria (~31%). Dominant fungal phyla included Ascomycota, Basidiomycota, and Zygomycota, with N fertilization significantly increasing the relative abundance of Ascomycota and decreasing that of Basidiomycota. Redundancy analysis revealed strong positive correlations between Ascomycota abundance and N₂O emissions across N treatments. At the genus level, Pyrenochaetopsis, Myrothecium, and Humicola were positively associated with N₂O production and identified as key functional taxa. These findings demonstrate that moderate N fertilization can disproportionately stimulate fungal-driven N₂O emissions in urban lawns. The results provide a scientific basis for optimizing N fertilization strategies in green spaces, with implications for N policy and sustainable landscape management.

1. Introduction

Nitrous oxide (N₂O) is a potent greenhouse gas with a global warming potential nearly 300 times that of carbon dioxide and a long atmospheric lifetime [1]. It also contributes substantially to the depletion of the stratospheric ozone layer [2]. As a key intermediate in the nitrogen (N) cycle, N₂O emissions are strongly influenced by ecosystem type and land management practices. In particular, external N inputs—via fertilization or atmospheric deposition—can significantly alter soil N status and subsequently influence N₂O production [3,4,5].

Numerous studies have demonstrated that N fertilization enhances N₂O emissions [6,7,8,9]. For example, fertilization significantly increased N₂O fluxes from urban lawn soils in Baltimore and Phoenix, USA, likely due to substantial N inputs for turf maintenance [10,11]. However, the shape of the N input-N₂O response curve remains debated, with either linear or non-linear relationships reported [12,13]. These discrepancies suggest that the magnitude and direction of fertilizer effects are context-dependent, influenced by fertilizer type, soil characteristics, vegetation, and microbial activity [14,15]. A clearer understanding of the N₂O response to fertilization is therefore essential for developing effective N management strategies.

N₂O production in soils is primarily mediated by nitrification and denitrification processes [16]. Bacteria have traditionally been regarded as the main drivers of these processes [17]. Recent evidence, however, suggests that fungi also contribute substantially to N₂O production [18]. For instance, studies in the arid grasslands of the southwestern United States and on the Qinghai–Tibet Plateau in China identified fungi as the dominant contributors to N₂O emissions via heterotrophic nitrification and denitrification [19,20]. In contrast, bacterial dominance has been reported in intensively managed croplands in northern China [21]. These contrasting findings underscore that the relative roles of fungi and bacteria in N₂O emissions vary across ecosystems and vegetation types, highlighting the need for site-specific evaluations.

Urban lawns represent one of the most extensive types of managed green spaces. Dense turf roots, frequent irrigation, and anthropogenically altered urban soils create microsites with heterogeneous oxygen, carbon, and nitrate availability that may favor fungal processes such as heterotrophic nitrification and denitrification [22,23]. These conditions could enhance fungal activity relative to bacterial activity in lawns, yet empirical evidence remains scarce [19,24,25,26,27]. Our previous work showed that N fertilization significantly increased N₂O emissions from lawn soils, with autotrophic nitrification accounting for approximately 63% of the emissions [28]. However, the relative contributions of fungi and bacteria to these emissions remain unresolved.

In this study, we examined the effects of different N fertilizer application rates (0, 100, 200, and 300 kg ha⁻¹ yr⁻¹) on N₂O emissions and associated microbial communities in lawn soil using incubation experiments combined with biological inhibitors and high-throughput sequencing. We hypothesized that (1) N₂O emissions would increase with N application rate, and (2) fungi would contribute more strongly to N₂O emissions than bacteria. Our findings aim to provide new insights into the microbial mechanisms driving N₂O emissions from urban lawn soils and to inform strategies for emission mitigation.

2. Materials and Methods

2.1. Site Description and Experimental Design

This study was conducted at Nanhu Park, located in Shenyang, Liaoning Province, Northeast China (123°40′ E, 41°48′ N). The region has a temperate semi-humid continental climate, with a mean annual temperature of 6.5 °C and annual precipitation of approximately 700 mm. The soil is classified as brown soil. The baseline soil properties were measured as follows: soil organic carbon (SOC) 7.33 g·kg⁻¹, total nitrogen (TN) 0.93 g·kg⁻¹, ammonium N (NH₄⁺-N) 3.21 mg·kg⁻¹, nitrate N (NO₃⁻-N) 25.99 mg·kg⁻¹, available phosphorus (P) 12.36 mg·kg⁻¹, available potassium (K) 98.96 mg·kg⁻¹, and pH 7.0. The lawn was established on 7 May 2016, and consisted of Poa pratensis (Kentucky bluegrass).

Four N fertilization levels were applied based on standard lawn management practices in northern China: 0 (N0, control), 100 (N100), 200 (N200), and 300 (N300) kg·ha⁻¹·yr⁻¹. Each treatment had three replicates (n = 3), randomly distributed in 1 × 1 m plots (12 in total). Urea was applied twice annually (15 April and 15 August) over two consecutive years, followed immediately by sprinkler irrigation. This approach reflected typical lawn management practices in northern Chinese cities to sustain turf density and color [28].

Soil samples were collected on August 16, one day after the second fertilization. Five soil cores (0–20 cm depth) were randomly taken from each plot and composited. The samples were transported to the laboratory within 24 h, sieved (<2 mm), and split into two subsamples: one for physicochemical analysis and the other for microbial DNA extraction.

2.2. Potential N₂O Emissions

To assess the N₂O emission potential, 30 g of air-dried soil (adjusted to 50% water-holding capacity, WHC) were placed in 250 mL Erlenmeyer flasks sealed with silicone rubber stoppers fitted with the three-way valves. The soils were pre-incubated at 25 °C and 50% WHC to activate microbial processes. The temperature of 25 °C was chosen to approximate the mean summer soil temperature in the study region, while 50% WHC represented moderate moisture conditions. Flasks were aerated daily by removing the stoppers for 1 h to prevent complete oxygen depletion and to simulate field gas exchange associated with soil macropore connectivity and surface aeration after irrigation events.

The experiment included four fertilization treatments, each with three replicates (n = 3), totaling 12 soils. Gas and soil samples were destructively collected on days 1, 3, 5, 7, 9, 12, and 15 from the start of the incubation, resulting in a total of 84 (12 × 7) flasks for soil incubation. For gas samples, 50 mL syringes were used to draw gas from the flasks, transferred to gas sampling bags for subsequent N₂O analysis. After gas sampling, the remaining soil was used for NH₄⁺-N and NO₃⁻-N analysis.

N₂O concentrations were determined using an Agilent 7890 GC equipped with an Electron Capture Detector (ECD) and an HP-PLOT Q column (30 m × 0.32 mm × 20 µm; Agilent). High-purity N₂ served as the carrier gas (2.0 mL∙min⁻¹). Calibration was performed every 12 samples with 5, 20, and 200 ppm N₂O standards (National Center for Standard Matters, Beijing, China), yielding R² > 0.999. Ambient-air blanks were run at the start and end to check for carryover. The limit of detection (LOD) and the limit of quantitation (LOQ) were 0.05 and 0.15 ppm (S/N = 3 and 10). Data were accepted only when the coefficient of variation in standards was <1%. Peak integration and quantification were performed using Agilent ChemStation with manual baseline verification. N₂O fluxes were calculated as follows [29,30,31]:

$$F={\displaystyle \frac{(C-C_0)\times \mathrm{V}\times \mathrm{M}\times [273/(273+\mathrm{T}\left)\right]}{\mathrm{d}\times \mathrm{m}\times 22.4\times 1000}}$$

(1)

where F = N₂O flux (μg·kg⁻¹·d⁻¹); C = N₂O concentration at sampling (μL·L⁻¹); C₀ = Initial N₂O concentration (μL·L⁻¹); V = Headspace volume (mL); M = Molecular weight of N₂O (44 g·mol⁻¹); T = Temperature (°C); d = Incubation time (h); m = Dry weight of soil (g).

2.3. Quantifying the Inhibitor-Sensitive Fraction of N₂O Emissions

Only N0 soil was used in the next experiment. To distinguish the fungal and bacterial inhibitor-sensitive fractions of N₂O emissions, four treatments were established [19,32,33,34]: (A) no-inhibitor, control; (B) fungal inhibitor, cycloheximide at 1.5 mg g⁻¹; (C) bacterial inhibitor, streptomycin at 3.0 mg g⁻¹; (D) both fungal and bacterial inhibitor, cycloheximide at 1.5 mg g⁻¹ and streptomycin at 3.0 mg g⁻¹.

Cycloheximide (a eukaryotic protein synthesis inhibitor) was selected to suppress active fungal metabolism, while streptomycin (an aminoglycoside antibiotic targeting bacterial 30S ribosomes) was selected to suppress bacterial activity. These concentrations were chosen based on prior soil incubation studies and preliminary trials that balanced inhibitory efficacy and soil sorption effects [19,32,33,34].

It is important to note that no inhibitor is perfectly specific in soil matrices: cycloheximide primarily affects eukaryotic ribosomes and so may have limited impacts on some soil protists or fauna; streptomycin targets many bacteria but has variable efficacy across taxa and can be adsorbed to soil particles, reducing bioavailability. Both inhibitors may also indirectly affect non-target taxa through alteration of substrate competition and community interactions [33]. The combined-inhibition treatment (cycloheximide + streptomycin) indicates the maximum achievable reduction in biological N₂O production.

Due to potential non-specific effects and differential sorption, the inhibitor-based partitioning provides a relative estimate of the fungal versus bacterial inhibitor-sensitive fractions rather than an absolute measurement. Therefore, we interpret the inhibitor results in conjunction with high-throughput ITS community profiles (Section 2.5) to identify taxa associated with N₂O production [32,33,34].

The fungal and bacterial inhibitor-sensitive fractions of N₂O emissions were calculated using the following formula:

$$\mathrm{Fungal}\mathrm{inhibitor-sensitive}\mathrm{fraction}(\%)=100\times {\displaystyle \frac{\mathrm{A}-\mathrm{B}}{\mathrm{A}-\mathrm{D}}}$$

(2)

$$\mathrm{Bacterial}\mathrm{inhibitor-sensitive}\mathrm{fraction}(\%)=100\times {\displaystyle \frac{\mathrm{A}-\mathrm{C}}{\mathrm{A}-\mathrm{D}}}$$

(3)

$$\mathrm{Other}\mathrm{microbial}\mathrm{inhibitor-sensitive}\mathrm{fraction}(\%)=100\%-\mathrm{Equation}\left(2\right)-\mathrm{Equation}\left(3\right)$$

(4)

where A = N₂O flux in control; B = N₂O flux with fungal inhibitor; C = N₂O flux with bacterial inhibitor; D = N₂O flux with both inhibitors.

2.4. Soil Physicochemical Analyses

Soil NH₄⁺-N and NO₃⁻-N were extracted using 2 M KCl (1:5 soil-to-solution ratio) and measured with a SmartChem140 Auto Analyzer (AMS, Rome, Italy). SOC was determined via the potassium dichromate oxidation method, and TN was analyzed using the Kjeldahl digestion method. Soil pH was measured with a pH meter (pHS-3C) in a 1:2.5 soil-to-water suspension.

2.5. Microbial Community Analysis

Soils from 4 treatments were used to measure ITS community profiles. Total DNA was extracted from 0.5 g of fresh soil using the FastDNA SPIN Kit for soil (MP Biomedicals, Santa Ana, CA, USA) according to the manufacturer’s instructions. Amplicon libraries were prepared using tagged universal fungal primers (ITS5 and ITS4), which target the internal transcribed spacers 1 and 2 (ITS1&2). Each of the 18 DNA samples was amplified separately using the fusion primer pair ITS5 (5′-A (6 bp MID) ACCCGCTGAACTTAAGC-3′) and ITS4 (5′-B TCCTGAGGGAAACTTCG-3′), generating PCR fragments of approximately 700 bp. Here, A and B denote the two pyrosequencing primers, and MID denote the multiplexing barcode tags for post-sequencing reads. PCR reactions contained 1 × PCR reaction buffer, 1.5 mM MgCl₂, 0.4 mM dNTPs, 1 mM of each primer, 1 U Taq DNA recombinant polymerase (Takara Biotechnology Co., Ltd., Shiga, Japan), and 2 μL DNA template, all in a final volume of 25 μL. Amplifications were conducted in a GeneAmp PCR System 9700 thermal cycler (Applied Biosystems Inc., Foster City, CA, USA), starting with an initial DNA denaturation step at 94 °C for 4 min, followed by 27 cycles of denaturation at 94 °C for 30 s, annealing at 58.3 C for 45 s, extension at 72 °C for 1 min, and a final elongation at 72 °C for 7 min. The 18 tagged PCR products were purified using the MinElute PCR Purifi-cation Kit (Qiagen Gmbh, Hilden, Germany), separated by electrophoresis through a 1.5% agarose gel in 1 × TAE and purified from the gel using the Qiagen QIAquick Gel Extraction Kit (Qiagen Gmbh, Germany). These cleaned PCR products were quantified on a NanoDrop 1000 spectrophotometer (Thermo Scientific, Waltham, MA, USA) and pooled in equimolar concentrations to a final concentration of 10 ng μL⁻¹. The product pool was sequenced on one quarter of a sequencing plate on a GS-FLX sequencer (454 Life Sciences) at Personal Biotechnology Co., Ltd. (Shanghai, China).

Data were processed following the procedure described previously, using the Quantitative Insights into Microbial Ecology (QIIME) pipeline (Available online: http://qiime.sourceforge.net (accessed on 15 July 2025)). In brief, sequence reads with an average quality score of <25, read lengths of <200 bp after trimming the last 30 bps, ambiguous bases and homo-polymers of >7 bases were removed from the dataset. Sequences were clustered and assigned to operational taxonomic units (OTUs) using the QIIME implementation of cd-hit with a threshold of 97% pairwise identity. The longest sequences of the 20 most abundant OTUs were extracted and taken as representatives for taxonomic identification by BLAST 2.2.26 searches against the non-redundant GenBank sequence database. Rarefaction analyses were performed using Analytic Rarefaction v.1.3 (Hunt Mountain Software 10.1, Athens, GA, USA).

2.6. Statistical Analyses

Statistical analysis in this study was performed using SPSS 22.0 (SPSS Inc., Chicago, IL, USA). One-way analysis of variance (ANOVA) was used to test differences among the four treatments (n = 3 for each treatment), with multiple comparisons performed using the least significant difference (Tukey HSD) test at a 95% confidence interval. Data normality was tested using the Shapiro–Wilk test, and homogeneity of variances was evaluated using Bartlett’s test prior to ANOVA. Pearson correlation analysis was performed to examine the relationships between N₂O emissions and the fungal genus. Principal component analysis (PCA) was performed on the top 20 fungal genera (cumulatively accounting for 98% of sequences) to visualize differences in composition and structure of the fungal communities in different N treatments. Redundancy analysis (RDA) was used to examine correlations between fungal communities and environmental variables using CANOCO 5.0, with variable significance tested via Monte Carlo permutation (n = 499). Statistically significant differences were defined as having a p-value of 0.05 unless otherwise stated. To control for false positives arising from multiple testing when screening many taxa, we applied the Benjamini–Hochberg false discovery rate (FDR) correction to the p-values. Only correlations that remained significant after FDR correction were considered statistically significant.

3. Results

3.1. Effects of N Fertilization on Inorganic N Concentrations

Soil NH₄⁺-N concentrations remained low (~6.74 mg·kg⁻¹) across all treatments (

Table 1. Soil properties at 0–20 cm soil depth sampling (mean ± standard error, n = 3). Different lowercase letters denote significant differences among treatments (Tukey HSD, p < 0.05).

Treatment	Soil Organic C (g·kg−1)	Total N (g·kg−1)	C/N Ratio	NH4+ Content (mg·N·kg−1)	NO3− Content (mg·N·kg−1)	pH
N0	7.33 ± 0.51	0.93 ± 0.10	7.95 ± 0.35	6.19 ± 0.25	25.99 ± 2.14 b	7.10 ± 0.02 a
N100	7.42 ± 0.95	0.90 ± 0.06	8.21 ± 0.48	6.75 ± 0.16	36.46 ± 1.70 ab	7.01 ± 0.01 a
N200	7.44 ± 0.22	0.89 ± 0.05	8.41 ± 0.24	7.55 ± 0.08	38.22 ± 0.67 a	6.83 ± 0.02 b
N300	7.11 ± 0.30	0.93 ± 0.05	7.66 ± 0.25	6.51 ± 0.08	41.65 ± 4.78 a	6.73 ± 0.05 b

). During the first seven days of incubation, no significant differences in NH₄⁺-N levels were observed among treatments (Figure 1a). However, on days 9 and 12, NH₄⁺-N concentration in N200 was significantly lower than that in the control (N0) (p < 0.05). No significant differences in the C/N ratio were observed among treatments (p > 0.05).

Soil NO₃⁻-N concentrations increased consistently with increasing N application (p < 0.05). On day 15, NO₃⁻-N concentrations in N100, N200, and N300 treatments reached 36.46, 38.22, and 41.65 mg·kg⁻¹, respectively—approximately 1.4 to 1.6 times higher than that in the N0 treatment (Table 1). By the end of the incubation, NO₃⁻-N accumulation was higher in N200 than in N300 (Figure 1b).

3.2. Effects of Nitrogen Fertilization on N₂O Emissions

On day 1, the N₂O flux in N300 was significantly higher than in other treatments (p < 0.05), though not at peak levels. The highest daily N₂O flux occurred on day 5 in N200 (10.29 μg·kg⁻¹·d⁻¹) and N300 (9.78 μg·kg⁻¹·d⁻¹), which were 1.6 and 1.5 times higher than in N0 (Figure 2a). Differences between N200 and N300 were not significant (p > 0.05), and N100 did not significantly differ from the control (p > 0.05). After day 7, fluxes stabilized, with no significant differences among treatments.

Cumulative N₂O emissions over the 15-day period were significantly affected by N fertilization (Figure 2b). N200 yielded the highest cumulative emission (49.65 μg·kg⁻¹), significantly exceeding all other treatments (p < 0.05) and 9% higher than N0. N300 also increased cumulative emissions relative to N0 (p < 0.05), whereas N100 showed no significant difference. Notably, more than 60% of total N₂O emissions in N200 and N300 occurred within the first seven days, indicating a strong short-term stimulation of high N inputs.

3.3. Inhibitor-Sensitive Fraction of N₂O Emissions

Biological inhibitors significantly reduced N₂O emissions compared to the uninhibited control (p < 0.05). Specifically, emissions were reduced by 45.4% with fungal inhibition, 30.9% with bacterial inhibition, and 76.3% with combined inhibition, indicating active contributions from both groups. Emissions were significantly lower under fungal inhibition than under bacterial inhibition (p < 0.05), suggesting stronger fungal inhibitor-sensitive processes (Figure 3). These values were interpreted as inhibitor-sensitive fractions consistent with fungal or bacterial involvement, rather than absolute contributions at the taxonomic level.

3.4. Fungal Community Composition and Structure

A total of 491,602 high-quality fungal ITS sequences were obtained, clustering into 2698 operational taxonomic units (OTUs) at 97% similarity. The dominant phyla were Ascomycota (21.1–36.6%), Basidiomycota (6.7–10.1%), and Zygomycota (0.7–5.4%). Ascomycota abundance was highest in N200, whereas Basidiomycota was most abundant in N100 and least in N300 (Figure 4a). N fertilization significantly increased the relative abundance of Ascomycota while decreasing Basidiomycota.

At the genus level, Pyrenochaetopsis (2.1–10.8%), Chaetomium (1.0–7.0%), and Mortierella (0.7–4.3%) were dominant. Pyrenochaetopsis was significantly enriched in N200 (10.81%) compared with other treatments (p < 0.05) (Figure 4b). N fertilization also significantly increased the relative abundance of Pyrenochaetopsis, Penicillium, Talaromyces, Humicola, Guehomyces, and Thermomyces, while decreasing Chaetomium, Mortierella, Tuber, and Simplicillium.

Principal component analysis (PCA) of the top 20 genera showed clear separation among treatments, with the first two axes explained 85.44% of the total variation. Replicates within each fertilized treatment clustered tightly and were clearly separated from N0, indicating that N fertilization significantly altered fungal community structure (Figure 5).

3.5. Correlations Between N₂O Emissions, Fungal Communities, and Soil Properties

Redundancy analysis (RDA) showed that the first two axes explained 60.7% and 7.5% of the variation in fungal community composition at the phylum level (Figure 6a). N₂O emissions were strongly and positively correlated with Ascomycota (R = 0.726, p = 0.008), whose abundance pattern across treatments mirrored the trend of N₂O emissions.

At the genus level, RDA revealed that 73.2% of the variation in fungal community structure could be explained by the first two axes (46.6% and 26.6%) (Figure 6b). N₂O emissions were positively correlated with Pyrenochaetopsis, Myrothecium, Zopfiella, Humicola, Bullera, and Conocybe. Among these, Pyrenochaetopsis, Myrothecium, and Humicola (all belonging to Ascomycota) showed significant positive correlations with N₂O emissions based on correlation analyses: (1) Myrothecium (R² = 0.556, p = 0.005); (2) Pyrenochaetopsis (R² = 0.478, p = 0.013); (3) Humicola (R² = 0.372, p = 0.035). In contrast, Cryptococcus showed a significant negative correlation (R² = 0.551, p = 0.006) (Figure 7), while Zopfiella, Bullera, and Conocybe were not significantly correlated with emissions.

In addition, N₂O emissions were positively correlated with soil NO₃⁻-N concentrations and negatively correlated with soil pH and NH₄⁺-N (Figure 6). Among soil properties, pH exerted the strongest influence on fungal community structure, followed by NO₃⁻-N, NH₄⁺-N, SOC, and TN. These shifts in soil chemical properties under N fertilization indirectly influenced fungal dynamics and N₂O emissions.

4. Discussion

This study examined the effects of different N fertilization rates on N₂O emissions and the contributions of fungi and bacteria in lawn soils. We further examined shifts in fungal community composition and their associations with N₂O dynamics. Our results showed that N fertilization significantly stimulated N₂O emissions (Figure 2b), in agreement with earlier reports reporting that exogenous N inputs enhance N₂O fluxes across a range of ecosystems [35,36,37,38,39]. The emission factors (1.1–2.2%) observed here are compared to, or higher than, the IPCC Tier 1 default value of 1% of applied N [1]. This suggests that urban lawns may deviate from global defaults, emphasizing the need for regionally specific emission factors and further field-based measurements. It should be noted, however, that our data were derived from short-term, aerobic microcosm incubations using air-dried, rewetted soils. These conditions represent potential N₂O production but do not account for plant uptake, root–microbe interactions, fluctuating redox states, soil structure, or gas diffusivity. Therefore, we refrain from extrapolating directly to field-scale emission budgets without additional validation.

The fungal inhibitor-sensitive fraction of N₂O emissions (45.4%) was significantly higher than that of bacteria (30.9%) or other microbes (24%) (Figure 3), supporting the growing recognition of fungi as major contributors to soil N₂O production, particularly under fertilized conditions. These values represent inhibitor-sensitive fractions rather than direct taxon-level flux partitioning; they should be interpreted cautiously given potential off-target effects of cycloheximide/streptomycin and the lack of independent confirmation (e.g., qPCR or plate counts). Similar fungal dominance has been reported in Tibetan grasslands [20,40], New Zealand pastures [41], and croplands and agroforestry systems in China [42,43], where fungal contributions range from 40% to >50%. The relatively greater fungal role may be linked to their ecological tolerance. In our study, N fertilization increased NO₃⁻-N concentrations while reducing soil pH (Table 1), conditions that likely suppressed bacterial ammonia oxidation but favored fungal activity. Thus, fertilized urban lawn soils may create microenvironments conducive to fungal-driven N₂O production.

Interestingly, the highest N rate (N300) did not yield the highest N₂O emissions. Instead, N200 produced the peak emission (Figure 2b), indicating a non-linear response. Several mechanisms may explain this pattern. Moderate N inputs may provide optimal substrate availability for fungal nitrification and denitrification without causing osmotic or pH stress [44]. By contrast, excessive N inputs (N300) may lead to acidification, accumulation of toxic intermediates, or altered microbial community composition that constrain N₂O production [45,46]. In support, the relative abundance of Ascomycota was slightly lower under N300 than N200 (35.1% vs. 36.6%) (Figure 4a), paralleling the emission trend. These findings suggest that N₂O emission potential depends not only on N inputs but also on microbial community structure and function, which may differ substantially across ecosystems.

Compared with croplands and natural grasslands, lawn soils are characterized by dense turf root mats, frequent irrigation, and anthropogenic soil compaction, all of which create microsites favorable to fungal processes. While bacterial dominance has been observed in intensively managed croplands in northern China [17], our results show a relatively greater fungal contribution, consistent with findings from other urban soils [18]. This contrast underscores the uniqueness of lawn ecosystems and the need to consider them separately in urban N₂O inventories.

Ascomycota emerged as the dominant fungal phylum, showing a strong correlation with cumulative N₂O emissions (R = 0.726, p = 0.008) and consistent patterns across treatments (Figure 6a). Previous studies indicated that over 90% of known N₂O-producing fungi belong to Ascomycota, followed by Basidiomycota (7%) and Zygomycota (3%) [46,47]. At the genus level, Pyrenochaetopsis, Myrothecium, and Humicola were significantly enriched under N fertilization and positively correlated with N₂O emissions (Figure 6b and Figure 7). Among these, Myrothecium showed the strongest association (R² = 0.556, p = 0.005), consistent with earlier studies demonstrating its high N₂O-producing capacity [45]. Its highest abundance in the N200 treatment further supports its role as a key fungal driver of lawn soil N₂O emissions.

Temporal patterns of flux also aligned with microbial responses. Both N200 and N300 exhibited emission peaks on day 5 (Figure 2a), consistent with earlier reports reporting rapid N₂O surges within days of fertilization [38,42,44]. More than 60% of cumulative emissions occurred in the first week, highlighting the short-term stimulatory effect of N addition [48]. Notably, the low-rate N100 treatment did not significantly increase emissions, likely due to rapid microbial immobilization, sorption, or limited transformations pathways under the incubation conditions [6,49].

From an ecological perspective, fungal dominance in N₂O production carries trade-offs. Unlike bacteria, many fungal denitrifiers lack a functional N₂O reductase, resulting in higher N₂O:N₂ ratios [15,41]. Thus, fungal dominance may increase net N₂O emissions per unit nitrate reduced. Conversely, fungi also decompose complex organic matter, thereby influencing soil C cycling and CO₂ fluxes [40]. Management strategies aiming to favor bacterial pathways with higher N₂O reductase activity could mitigate emissions but may also alter C cycling and nutrient availability for lawn growth.

Our findings that fungal processes contributed more strongly to N₂O emissions have practical implications: fungi and bacteria respond differently to environmental conditions such as pH, moisture, and soil amendments. Targeted management practices—such as soil pH adjustment, biochar application, or the use of nitrification inhibitors—could preferentially suppress fungal pathways and thereby reduce N₂O emissions more effectively. For example, controlled-release fertilizers can regulate N supply and reduce post-fertilization peaks [27,44]. Biochar amendments may improve soil aeration, enhance nutrient retention, and lower substrate availability for N₂O-producing microbes [12,48,49]. Soil pH adjustment can also influence microbial community composition, potentially favoring denitrifiers with higher N₂O reductase activity [46]. While these strategies require testing under field conditions, they represent promising avenues for balancing turf quality and climate mitigation.

Taken together, our results underscore the central role of fungal pathways—especially Ascomycota genera such as Myrothecium, Pyrenochaetopsis, and Humicola—in shaping N₂O emission from fertilized lawns. They provide a microbial-level explanation for the observed emission dynamics and suggest that sustainable N management must account for microbial responses, not only N application rates. Future studies incorporating functional gene quantification (amoA, nirK, nirS, nosZ, p450nor) will be valuable to directly link community structure to process pathways. Moreover, as our study focused solely on N₂O, additional work is needed to include CO₂ and CH₄ to better capture the multi-gas contributions of lawns to urban greenhouse gas budgets.

5. Conclusions

Our incubation experiments demonstrate that N fertilization significantly stimulates soil N₂O emissions in urban lawn soils. The N200 treatment produced the highest cumulative N₂O emission, whereas further increase to N300 did not increase emissions, likely due to shifts in microbial community structure and soil chemistry. Biological inhibitor assays indicated that the inhibitor-sensitive fraction consistent with fungal contribution was 45% (inhibitor-sensitive estimate), with Ascomycota—particularly, Pyrenochaetopsis, Humicola, and Myrothecium—positively linked to these emissions.

These results suggest that moderate N fertilization (≤200 kg·ha⁻¹·yr⁻¹) can maximize fungal-driven N₂O emissions in urban lawns. Sustainable N management should therefore consider fungal community responses. Promising approaches include lowering N rates, using controlled-release formulations, adjusting fertilization timing relative to irrigation, and applying soil amendments such as biochar or pH adjustments to reduce fungal-driven pathways. The finding that >60% of cumulative emissions occurred within 7 days of fertilization further highlights the importance of fertilizer timing and the potential of split applications to mitigate emission pulses.

Finally, because this study was based on short-term laboratory incubations and focused only on N₂O, future field trials that include plant uptake, diurnal fluxes, and multi-gas (CO₂, CH₄) measurements are essential to scale results to urban lawn greenhouse gas budgets. Such work should also quantify the net climate impact of different management strategies while considering their implications for lawn health and nutrient runoff.