Soil N 2 O emission intensity under different crop fertilization systems and the use of PGPR

1 Introduction

Nitrogen use efficiency from fertilizers by crops typically ranges between 35% and 65%, depending on crop species, soil properties, and climatic conditions (Vitousek et al., 1997; Cavigelli et al., 2012). The fraction of nitrogen not assimilated by plants poses a significant environmental threat, as mineral nitrogen compounds in soil are highly mobile and susceptible to various loss pathways, most notably nitrous oxide (N₂O) emissions and nitrate leaching (Akiyama et al., 2006; Robertson et al., 2013).

Nitrous oxide is recognized as a potent greenhouse gas and a major contributor to stratospheric ozone depletion (Ravishankara et al., 2009). At the same time, approximately 52% of total global N₂O emissions originate from agricultural soils (Qian et al., 2025). These emissions primarily result from two microbial processes: aerobic nitrification, during which ammonia is oxidized, and anaerobic denitrification, wherein nitrate serves as the terminal electron acceptor under low-oxygen conditions (Okereke, 1993; Norton, 2008; Moreau et al., 2019; Linton et al., 2020; Prosser et al., 2020, Myrold, 2021).

Fertilizer application is now widely acknowledged as a key driver of N₂O emissions from agricultural soils (Syakila and Kroeze, 2011). Numerous studies have shown a direct positive correlation between mineral nitrogen input levels and N₂O fluxes (Shcherbak et al., 2014; Shakoor et al., 2018).

In response to these environmental concerns, research efforts have been directed toward developing management strategies to mitigate N₂O emissions from croplands. These include investigations into the effects of fertilizer type, application rate, and tillage practices on N₂O dynamics (Bell et al., 2015; Geng et al., 2021; Min et al., 2021; Suzuki et al., 2021; Wang et al., 2021).

To this end, in recent years, the adoption of enhanced-efficiency nitrogen fertilizers, such as those with controlled-release formulations, as well as the use nitrification inhibitors, or urease inhibitors, has gained momentum. These technologies improve nitrogen use efficiency and partially attenuate nitrogen losses by reducing nitrate leaching, surface runoff, and gaseous emissions of nitrogen compounds (Akiyama et al., 2015; Feng et al., 2019; Bortoletto-Santos et al., 2020; Lawrencia et al., 2021). However, the use of chemical inhibitors increase production costs and pose non-target environmental risks. For instance, nitrapyrin, a widely used nitrification inhibitor, has been detected in stream water, indicating off-site transport (Woodward et al., 2016), while residues of dicyandiamide have been found in cow milk in New Zealand, prompting regulatory bans on its use in pasture systems (Thapa et al., 2016).

An alternative approach involves the use of biopreparations based on active strains of plant growth-promoting microorganisms (PGPM), particularly those capable of enzymatically reducing N₂O to environmentally benign N₂ via nitrous oxide reductase activity (Thomson et al., 2012; Butterbach-Bahl et al., 2013; Hu et al., 2015; Gao et al., 2016). Additionally, inoculated plants often exhibit enhanced nitrogen uptake, thereby reducing the availability of mineral nitrogen substrates for nitrifiers and denitrifiers in the rhizosphere (Calvo et al., 2016; Volkogon et al., 2021).

Another promising agronomic strategy to limit gaseous nitrogen losses involves supplying readily available carbon to soil microorganisms in stoichiometric proportion to unassimilated mineral nitrogen. Under such conditions, excess nitrogen is immobilized into microbial biomass and incorporated into organic pools (Kirkby et al., 2014; Mooshammer et al., 2014). These organic forms may then contribute to humic substance synthesis or undergo gradual mineralization, serving as a sustained nitrogen source for subsequent crops in the rotation.

Crop residues and the biomass of green manure crops can serve as an important carbon source to limit soil nitrogen losses under nitrogen fertilizer application, although opinions regarding the efficacy of these agricultural practices remain controversial.

To address these conflicting findings, the objective of our research was to evaluate the potential for mitigating soil N₂O emissions by combining mineral fertilization systems with straw and green manure biomass as a carbon source for microorganisms and the use of PGPR.

2 Materials and methods

The research was conducted over a period of three years (2019 – 2021) in a field experiment at the field research facility of the Institute of Agricultural Microbiology and Agro-Industrial Production of the National Academy of Agrarian Sciences of Ukraine located near Chernihiv (Ukraine). The field experiment was established in 2009. Prior to 2019, the research aimed to determine the effect of inoculation on the productivity of agrocenoses. Measurements of N₂O emissions began in 2019. It was considered that, over the 10-year period of systematic application of mineral, organic, and organo-mineral fertilizers, soil biological processes in the corresponding plots had acquired the distinct characteristics established by each fertilizer type. Experimental site coordinates: 51°33’51.0”N; 31°20’19.9”E. Elevation: 144 m above sea level. The climate is temperate continental with sufficient annual precipitation (500–600 mm), warm summers (+18 °C to +19.5 °C in July), and relatively mild winters (-6 °C to -8 °C in January). The period with temperatures above 10 °C lasts 150–160 days per year. Environmental conditions during the growing seasons, including temperature and soil moisture, were close to the long-term averages for the Ukrainian Polissia region.

The soil was characterized as leached light loam chernozem on loess-like loams (Phaeozem Chernic). Soil bulk density was 1.46 g/cm³. Agrochemical characteristics of the soil were determined by standard methods harmonized with EU standards (Fomin and Fomin, 2000): pH_KCl – 5.2; soil organic matter – 3.1%; easily hydrolyzed nitrogen – 109 mg/kg; labile forms of phosphates (as P₂O₅) – 168 mg/kg; exchangeable potassium (as K₂O) – 58 mg/kg of soil.

Standard regional management practices were applied consistently across all plots, including primary tillage in autumn along with the application of phosphorous and potassium mineral fertilizers and pre-sowing spring cultivation alongside nitrogen fertilizers application.

Mineral fertilizers were applied annually in form of ammonium nitrate (34.5% N), granulated superphosphate (19.5% P₂O₅), potassium chloride (60% K₂O), and potassium sulfate (56% K₂O) in potatoes. Phosphorus and potassium fertilizers were incorporated during main tillage (in autumn), while nitrogen fertilizers were applied during pre-sowing cultivation in spring and as top dressing.

The experiment was established using a randomized complete plot design with four replicates. The study area was organized into two parallel blocks one without microbial inoculation and the other with the application of PGPRs, with plots randomized within each block to minimize spatial bias. Each individual plots measured 43.2 m², resulting in a total area of 172.8 m² (43.2 m² × 4) per variant (see Supplementary Table 1). In total, the experiment comprised 320 plots (80 plots per crop).

The crop rotation used in the experiment consisted of four crops: potato – spring barley – pea – winter wheat. Since the start of the experiment in 2009, three complete crop rotation cycles have been completed. Fertilization treatments included mineral, organic, and organo-mineral systems with and without PGPR inoculation (see Table 1; Supplementary Table 1).

Table 1. Fertilization scheme and organic residue residual effects across the crop rotation.

• All crop residues were removed from all experimental plots except for the specified fertilization regimes.

• All fresh organic fertilizers in corresponding variants were applied prior to potato planting.

• Mineral fertilizers were applied annually according to the specified fertilization regimes.

• Each experimental layout evaluated in two parallel series: uninoculated and inoculated with plant growth-promoting rhizobacteria.

• The subscript numbers next to NPK here and in text denote the mineral fertilizer (nitrogen, phosphorous and potassium) application rates in kilograms per hectare.

• Effects of organic residues were observed in the year of application (direct effect) and in the 1st, 2nd, and 3rd years thereafter (residual effects) for each corresponding organic input.

• *Incorporation of 5 t/ha chopped straw into the soil immediately after winter wheat harvest (end of July).

• **Incorporation of chopped straw as described above, followed by an intermediate narrow-leaved lupine green manure crop incorporated into the soil in late fall (disked and shallowed plowed at the end of November).

Fresh organic matter was applied annually prior to potato cultivation. In designated treatment variants, chopped wheat straw (carbon content – 2100 kg, nitrogen – 27 kg, phosphorus – 20 kg, potassium – 30 kg; C/N ratio – 77:1) was incorporated into the soil at a rate of 5 t/ha immediately following winter wheat harvest (late July) via disking. In variants where straw application was not planned, it was removed from the plots. Following straw application in the designated treatments, narrow-leafed lupin was sown as an intermediate green manure at a seeding rate of 200 kg/ha. The lupine biomass, with an average over the years of study, ~2.6 t/ha of dry aboveground mass and ~2.7 t/ha of dry root biomass (carbon content – 1600 kg, nitrogen – 99 kg, phosphorus – 25 kg, potassium – 35 kg; C/N ratio – ~16:1), was disked into the soil in late autumn (end of November) followed by shallow plowing to a depth of 15 cm. Total carbon input from the application of straw and green manure biomass was 3600 kg/ha, nitrogen – 126 kg/ha, phosphorus – 25 kg/ha, and potassium – 65 kg/ha.

Given the temporal and spatial deployment of the experiment across three crop rotation cycles (2009 – 2021), each plot received straw and green manure applications three times. This allowed for the assessment of both direct effects and residual (first, second- and third year) aftereffects of organic fertilizers. Mineral fertilizers were applied annually according to the experimental design, ensuring consistent nutrient input across treatments. The systematic nature of fertilization over multiple rotations supports the interpretation that observed processes are attributable to the type and combination of fertilizers used. The systematic application of fertilizers over three crop rotations allows us to conclude that the studied processes reflect specific features determined by the type of fertilization.

Pre-sowing inoculation was performed for each crop using specific strains of agronomically beneficial microorganisms:

• potato (cv. Belaroza) and spring barley (cv. Nosivsky) were inoculated with Azospirillum brasilense 410 (deposited in the microorganism collection of the DK Zabolotny Institute of Microbiology and Virology NAS of Ukraine under accession number IMV B-7222). Application rates: 2 L/4 t of potato planting material/ha; 50 mL/200 kg of barley seed (bacterial suspension titer 3 × 10⁹ cells/mL).
• pea (cv. Gotivsky) seeds were treated with Rhizobium leguminosarum 348a (stored in the collection of the Institute of Agricultural Microbiology and Agro-Industrial Production of the NAAS under accession number B-42). Application rate: 100 mL/300 kg of seed bacterial suspension titer 2 × 10⁹ cells/mL).
• winter wheat (cv. Sonechko) was inoculated with Paenibacillus polymyxa KB (deposited in the microorganism collection of the DK Zabolotny Institute of Microbiology and Virology NAS of Ukraine under accession number IMV B-8059), a growth-promoting and phosphate-solubilizing bacterium. Application rate: 100 mL/200 kg of seed (3.5 × 10⁹ cells/mL).

For inoculation, bacterial suspensions were used after cultivation in liquid nutrient media on rotary shakers. The aforementioned volumes of bacterial suspensions were diluted with water (water volume was 2% of seed mass) to ensure uniform application to the planting material and applied to seeds during mixing. Seed material designated for uninoculated treatments was treated with water at 2% of seed mass to maintain experimental consistency. Following inoculation, seeds were air-dried in a thin layer in an outdoor shaded area for 1 hour prior to sowing. Because the same seeder/potato planter was used for sowing, uninoculated seeds were sown first, followed by the inoculated seeds/planting material.

The intensity of soil nitrous oxide emissions were measured using sealed chamber method (Kusa et al., 2008). For this after cutting the aboveground biomass cylindrical chambers (0.30 m diameter, 0.35 m height) were inserted into the soil to a depth of 5 cm. To ensure airtight conditions and prevent gas leakage, a water seal was made around each chamber. Because the chambers were inserted 5 cm into the soil, the headspace volume was 21195 cm³.

To inhibit the reduction of N₂O to N₂, cups of water containing 20 g of calcium carbide were placed inside chambers before exposure. Acetylene, produced by the reaction of CaC₂ with water, inhibits nitric oxide reductase, blocking the dissimilation of NO^3- and NO^{2^-} at the nitric oxide reduction stage (Duxbury, 1986). This enables quantification of the N₂O emission intensity formed during nitrification and denitrification.

The acetylene inhibition technique used in the study has specific methodological limitations (Felber et al., 2012). While less sensitive than 15N isotope tracing, it is substantially more cost-effective and feasible for large-scale field evaluations when appropriate instrumentation is available.

Samples were collected in quadruplicate, with measurements taken three times during each crop’s vegetation period, corresponding to main growth stages. Acetylene exposure lasted 3 hours, validated in previous studies (Volkogon et al., 2019) as sufficient for reliable N₂O measurement. Minor chamber microenvironment changes were assumed to be consistent across treatments.

After exposure, gas samples were collected with syringes and transferred into vacuum vessels sealed with rubber membranes for laboratory analysis. N₂O concentrations were determined using a gas chromatograph with an electron capture detector and 3 m steel sorption columns packed with Paropak Q sorbent (60–80 mesh; Waters Corporation, USA). Operating conditions: column temperature – 40 °C; evaporator temperature – 120 °C; detector temperature – 330 °C; carrier gas – argon with 5% methane, flow rate – 35 cm³/min.

N₂O emissions were calculated using the formula:

EN2O=E×V3

where, EN2O – is the amount of nitrous oxide in the chamber (nmol N₂O),

• E – is the amount of nitrous oxide in 1 cm³ injected into the chromatograph detector,
• V – is the chamber headspace volume (cm³),
• 3 – the chamber exposure duration (hours).
• N₂O emissions per 1 ha area were calculated using the formula:

where: TN2O – is the amount of N₂O, nmol/ha/24 h;

• S₁ – is the 1 ha area in m²;
• S₂ – is the area covered by the chamber, m²; and
• 24 – is the number of hours per day.

Gas chromatographic data were used to estimate nitrogen losses, accounting for the molecular weight of N₂O and extrapolating emission rates to a per-hectare-per-day basis.

Crop yield was determined by weighing the harvested produce from each plot, excluding border rows. For potato, total fresh tuber weight was recorded. Grain crop yields were adjusted to a standard moisture content of 14%. Yield data were standardized to a per-hectare basis.

Differences in N₂O emissions among treatments and crop yields across experimental treatments were evaluated using the analysis of variance (ANOVA). Statistical significance was determined at a level of p ≤ 0.05. The Least Significant Difference (LSD_0.05) was utilized as the primary measure of variability and for mean separation between variants, ensuring that all reported comparisons are statistically supported. Data analysis was performed using Statistica 6.0 software (StatSoft Inc., USA).

3 Results

The three-year measurements of N₂O emission intensity losses from leached chernozem soil across different crops and fertilization systems demonstrated consistent patterns across experimental years; therefore, the results presented below focus on the data from the final study year (2021). According to the findings presented in Table 2, during potato cultivation, straw incorporation alone did not yield significant differences in N₂O emissions compared to the unfertilized control, though a slight trend toward reduced flux was observed. Conversely, lupine green manure incorporation increased nitrous oxide emissions across all three sampling periods. Combining straw with green manure also increased emissions compared to the control, although emission rates remained slightly lower, except for the last study period, than those in the treatment utilizing green manure alone (Table 2).

Table 2. Soil N-N₂O emission rates across developmental stages of potato as affected by fertilization regime and PGPR inoculation (g N-N₂O ha^-1 day^-1).

1. Values stand for treatment means (n = 4).

2. LSD_0.05 values stand for the Least Significant Difference (p ≤ 0.05).

3. Color gradients indicate soil N-N₂O emission rates, ranging intensity from dark red (highest flux) to yellow (medium flux) to dark green (lowest flux).

Mineral fertilizer application significantly increased N₂O emissions, with flux rates escalating with the increase of the nitrogen application rates. Specifically, application of 40 kg N ha^-1 increased N-N₂O losses by 72-98-81% across observation periods relative to the control. At 80 kg N ha^-1, losses increased by 102-127-124%, while the highest rate of 120 kg N ha^-1 resulted in increases of 125-127-163% compared to control values. However, N₂O emission intensity dynamics shifted markedly when mineral fertilizers were applied over the background of straw and green manure, resulting in significant emission reductions across all treatments. Incorporating N₄₀P₄₀K₄₀ alongside straw and green manure biomass reduced N-N₂O losses by 7-12% relative to the equivalent mineral fertilizer dose applied alone.

Similarly, combining N₈₀P₈₀K₈₀ with straw and green manure reduced N₂O emissions by 7-14%. When the highest fertilization rate (N₁₂₀P₁₂₀K₁₂₀) was applied over the straw and green manure background, gas emissions decreased by 7-11% compared to mineral fertilization alone, depending on the plant organogenesis stage.

Pre-planting inoculation of potato tubers with A. brasilense 410 further reduced nitrous oxide emissions across nearly all fertilization backgrounds, with the most definite mitigation occurring in treatments receiving mineral fertilizers, including those combined with straw and green manure. Thus, comparing N-N₂O losses in the “N₄₀P₄₀K₄₀ + straw + green manure + inoculation” treatment to the “N₄₀P₄₀K₄₀” treatment alone revealed a reduction of 16 – 18.3% depending on the growth stages of the potato plants. In the “N₈₀P₈₀K₈₀ + straw + green manure + inoculation” treatment compared to “N₈₀P₈₀K₈₀”, the reduction ranged from 16% to 25.2%, depending on the stage of plant organogenesis.

When the highest fertilizer rate, N₁₂₀P₁₂₀K₁₂₀, was combined with the straw, green manure, and inoculation, N-N₂O losses decreased by 16 – 25% compared to the application of mineral fertilizers alone, varying by potato developmental stage.

In the study of nitrous oxide emissions during spring barley cultivation, gaseous nitrogen losses in the variants with the first year of the residual effect of straw were nearly identical to those observed in unfertilized control. A minor increase in N₂O emissions was observed during the residual action of lupine green manure (Table 3).

Table 3. Soil N-N₂O emission rates across developmental stages of spring barley as affected by fertilization regime and PGPR inoculation (g N-N₂O ha^-1 day^-1).

1. Values stand for treatment means (n = 4).

2. Superscript 'a' denotes first-year residual effects following incorporation of corresponding organic input

3. LSD_0.05 values stand for the Least Significant Difference (p ≤ 0.05).

4. Color gradients indicate soil N-N₂O emission rates, ranging intensity from dark red (highest flux) to yellow (medium flux) to dark green (lowest flux).

Combining organic inputs with mineral fertilizers significantly reduced nitrous oxide emissions compared to the N-N₂O losses seen under mineral fertilizers applied alone. Specifically, when the lowest fertilizer rate (N₃₀P₃₀K₃₀) was applied following the first-year incorporation of wheat straw and lupine green manure, N₂O emissions decreased by 6.6 – 13.9% across crop developmental stages compared to the mineral fertilizer treatment alone. Applying N₆₀P₆₀K₆₀ over the organic background reduced N-N₂O losses by 8 – 11%, while integrating organic residues with the highest fertilizer rate (N₉₀P₉₀K₉₀) resulted in an 8 – 13% reduction compared to mineral fertilization alone (Table 3).

Substantial reductions in nitrous oxide losses were also observed in variants with PGPR. Inoculation enhanced the mitigation effect achieved by combining mineral fertilization and organic inputs. In the variant with pre-sowing barley seed bacterization with Azospirillum applied over the N₃₀P₃₀K₃₀ background, N₂O emissions decreased by 10% at germination, by 13% at stem elongation, and by 5% at the milk-wax ripeness stage, compared to the same treatment without inoculation. When compared directly to the application of mineral fertilizers alone at the same rate (N₃₀P₃₀K₃₀), N₂O emissions decreased by 22% at germination, 19% at stem elongation, and 11% at milk-wax ripeness (Table 3).

Similar trends were observed for inoculated plants grown under higher mineral fertilizer rates, both alone and over the residual organic background. For the combination of N₆₀P₆₀K₆₀ with straw, green manure, and inoculation, N-N₂O losses decreased by 13%, 16%, and 15% across the respective developmental stages compared to the uninoculated organic-mineral treatment. When comparing “N₆₀P₆₀K₆₀ + straw + green manure + inoculation” to “N₆₀P₆₀K₆₀” applied alone, N₂O emission reductions reached 20% at germination, 23% at stem elongation, and 24% at milk-wax ripeness (Table 3).

In the “N₉₀P₉₀K₉₀ + straw + green manure + inoculation” treatment, nitrous oxide emissions decreased by 17%, 16%, and 12% across the three developmental stages compared to the corresponding uninoculated treatment, and by 28%, 23%, and 22%, respectively, compared to the “N₉₀P₉₀K₉₀” mineral treatment alone.

When growing the third crop in the rotation, pea, no significant changes in N₂O emission rates were observed in the second year of the residual effect of straw and lupine green manure alone. However, applying mineral fertilizers over the residual organic background reduced the intensity of N-N₂O losses (Table 4). Specifically, combining the lowest fertilizer rate (N₃₀P₃₀K₃₀) with the residual effect of straw and green manure reduced nitrous oxide emissions by 10 – 13% across pea developmental stages compared to mineral-only treatments. Applying N₆₀P₆₀K₆₀ over the second-year residual organic background reduced N₂O emissions by 17% at budding, 15% at flowering, and 18% at pod formation stages relative to the same mineral dose applied alone.

Table 4. Soil N-N₂O emission rates across developmental stages of pea as affected by fertilization regime and PGPR inoculation (g N-N₂O ha^-1 day^-1).

Notes:.

1. Values stand for treatment means (n = 4).

2. Superscript 'b' denotes a second-year residual effects following incorporation of corresponding organic input

3. LSD_0.05 values stand for the Least Significant Difference (p ≤ 0.05).

4. Color gradients indicate soil N-N₂O emission rates, ranging intensity from dark red (highest flux) to yellow (medium flux) to dark green (lowest flux).

When the highest mineral fertilizer rate (N₉₀P₉₀K₉₀) was applied, N-N₂O losses were the highest among all studied pea fertilization treatments (Table 4). However, applying this rate over the second-year residual background of organic inputs reduced N-N₂O losses by 13%, 16%, and 16% across the respective developmental stages.

Pre-sowing inoculation of pea seeds with Rhizobium leguminosarum 348a effectively suppressed gaseous nitrogen losses from the soil, particularly under crop cultivation on mineral and organo-mineral fertilization backgrounds. Inoculation under the N₃₀P₃₀K₃₀ background reduce N₂O emissions by 12%, 21%, and 8%, respectively.

The integrated pea fertilization system combining N₃₀P₃₀K₃₀ with the second-year residual effect of straw, green manure, and inoculation reduced N₂O emissions by 23%, 29%, and 21% across stages compared to the systemic application of N₃₀P₃₀K₃₀ alone (Table 4).

Combining mineral nitrogen, residual organic matter, and inoculation also significantly lowered nitrous oxide emissions at the medium fertilizer dose (N₆₀P₆₀K₆₀). At the budding stage, daily N-N₂O losses in this variant were 83.6 g ha^-1, compared to 118.1 g ha^-1 under mineral fertilization alone. At flowering, emissions were 87.3 g ha^-1 versus 104.7 g ha^-1, and at pod formation, emissions dropped to 66.9 g ha^-1 compared to 86.4 g ha^-1 in mineral-only control.

Similar mitigation trends were observed when comparing the highest fertilizer rate (N₉₀P₉₀K₉₀) applied over the second-year residual organic background with inoculation against the application of N₉₀P₉₀K₉₀ alone (Table 4).

In winter wheat cultivation, the mitigation efficacy of the third-year residual effect of straw and green manure combined with mineral fertilizers significantly declined (Table 5). Across nearly all developmental stages, differences in N₂O emission rates between integrated fertilization treatments and mineral-only applications were either statistically non-significant or marginal.

Table 5. Soil N-N₂O emission rates across developmental stages of winter wheat as affected by fertilization regime and PGPR inoculation (g N-N₂O ha^-1 day^-1).

1. Values stand for treatment means (n = 4).

2. Superscript 'c' denotes a third-year residual effects following incorporation of corresponding organic input

3. LSD_0.05 values stand for the Least Significant Difference (p ≤ 0.05).

4. Color gradients indicate soil N-N₂O emission rates, ranging intensity from dark red (highest flux) to yellow (medium flux) to dark green (lowest flux).

However, pre-sowing seed inoculation with the plant growth-promoting rhizobacterium P. polymyxa KB mitigated soil N₂O emissions across most fertilizer backgrounds. The results demonstrated either a statistically significant reduction in N₂O emission rates or distinct decreasing trends in gaseous losses (Table 5).

Overall, applying mineral fertilizers alongside straw and green manure reduced N-N₂O loss intensity. Nevertheless, the mitigating capacity of these organic amendments became negligible by the third year following soil incorporation.

Pre-sowing seed inoculation provides a substantial and persistent reduction in gaseous nitrogen losses. The integrated application of mineral fertilizers, organic residues, and microbial inoculation achieved the most pronounced reduction in N₂O emissions during potato, spring barley, and pea cultivation, while this synergistic effect was notably reduced during the third year after straw and green manure incorporation.

Crop yields varied considerably across the evaluated fertilization treatments (Table 6). Contrary to the initial hypothesis that straw incorporation might reduce crop yields through nitrogen immobilization, the experiment revealed a consistent trend toward enhanced productivity. Intermediate lupine green manure demonstrated a pronounced fertilizing effect on the first crop in the rotation, potato. However, for subsequent crops, the residual effect of green manure alone did not produce statistically significant yield differences compared to unfertilized controls. Combining straw with green manure exhibited a synergistic fertilizing effect, particularly during the cultivation of potato, spring barley, and pea. By the third year of the residual effect of organic inputs, no statistically significant impact on winter wheat yield was observed. In contrast, mineral fertilizer application consistently increased crop yields across the entire rotation. Notably, integrating mineral and organic inputs resulted in significant yield enhancements or strong positive trends across all crops (Table 6).

Table 6. Crop yields across the crop rotation in response to fertilization regimes and PGPR inoculation (t ha^-1, mean values for 2019–2021).

1. LSD_0.05 values stand for the Least Significant Difference (p ≤ 0.05) for the main effects (fertilizer and inoculation) and their interaction.

2. Superscript 'a' denotes a significant difference compared to unfertilized control.

3. Superscript 'b' denotes a significant increase resulting from the combined application of mineral and organic fertilizers compared to mineral fertilizers only.

4. Superscript 'c' denotes a significant increase resulting from the PGPR inoculation.

Significant yield increases were recorded in potato under the “N₈₀P₈₀K₈₀ + straw + green manure” treatment compared to “N₈₀P₈₀K₈₀” alone. Other combinations of mineral fertilizers with organic inputs showed a positive trend. In spring barley, statistically significant yield increases were recorded in the “N₃₀P₃₀K₃₀ + straw + green manure” and “N₆₀P₆₀K₆₀ + straw + green manure” treatments relative to their respective mineral-only controls. For pea, all evaluated mineral fertilizer doses applied over the second-year residual organic background produced statistically significant yield increases compared to the corresponding mineral-only treatments. Conversely, for winter wheat, applying mineral fertilizers over the third-year residual background of straw and green manure did not yield statistically significant changes in crop productivity (Table 6).

Pre-sowing seed or tuber inoculation with PGPR enhanced productivity across all fertilization backgrounds. The sole exception occurred with Azospirillum brasilense 410 applied over the straw background in potato cultivation, where only a non-significant trend toward increased yield was noted. The most pronounced yield gains in potato, barley, and pea were achieved through the integrated application of mineral fertilizers, organic residues, and PGPR inoculation. For winter wheat, although statistically significant yield increases were not established under this integrated treatment, a clear positive trend in grain yield was observed (Table 6).

4 Discussion

Mitigating nitrous oxide emissions from agricultural soils is still a critical priority for sustainable farming systems. Systematic and often excessive nitrogen fertilizer application contributes significantly to environmental degradation, primarily due to low crop nitrogen use efficiency, which leads to substantial nitrogen losses, including elevated N₂O emissions (Akiyama et al., 2006; Robertson et al., 2013). However, these environmental risks can be mitigated through several strategies, including supplying soils with sufficient microbial carbon substrates via the incorporation of crop residues. Sander et al. (2014) showed that plant residue incorporation can suppress N₂O emissions, with efficacy depending on soil characteristics and the carbon-to-nitrogen (C/N) ratio of the added organic material. The authors posited that high C/N ratios stimulate competition for ammonium between heterotrophic microorganisms and autotrophic nitrifiers, thereby reducing N₂O formation.

In our study, straw and lupin green manure biomass served as carbon sources for soil microorganisms. This combination was primarily intended to prevent excessive nitrogen immobilization, which typically occurs when straw with a wide C/N ratio is applied alone, potentially impairing crop yields. Furthermore, combining straw with lupin green manure biomass, residues characterized by contrasting mineralization rates, was designed to balance and optimize the release of microbially available carbon from these organic residues.

The existing literature presents conflicting evidence regarding the impact of straw incorporation on N₂O emissions. Wu et al. (2018) reported that rice straw mulching at rates of 5 and 10 t ha-1 reduced emissions by 23.1% and 33.5%, respectively. Similarly, Ma et al. (2007) observed that straw mulching decreased N₂O emissions by an average of 30%, even under high soil mineral nitrogen levels (200–270 kg N ha-1) that otherwise increased emissions by 5- to 14-fold. Wang et al. (2020) attributed this emission-reducing effect to structural shifts within the denitrifying microbial community. Furthermore, Rothardt et al. (2021) demonstrated in a three-year field trial in northern Germany that residues with high C/N ratios, such as winter wheat straw, temporarily immobilized mineral nitrogen and reduced direct N₂O losses. Conversely, a study at the Reinshof experimental farm in Germany indicated that oilseed rape straw had a negligible mitigating effect on N₂O emissions (Köbke et al., 2022), though the authors highlighted its broader ecological benefits. Moreover, Bai et al. (2020) documented increased N₂O emissions when straw was co-applied with mineral fertilizers. Crop residue incorporation has also been reported to stimulate N₂O flux by increasing carbon and nitrogen substrate availability, thereby promoting microbial activity. Additionally, residue addition can alter soil aeration by enhancing aggregation and accelerating microbial oxygen consumption, factors that critically govern the balance between the nitrification and denitrification pathways responsible for N₂O production (Gregorutti and Caviglia, 2017). Given these conflicting findings, further research is required to elucidate the effects of residues from different crops on N₂O emission dynamics, particularly in interaction with varying mineral nitrogen application rates, as evaluated in our study.

Green manures represent a distinct category of organic amendments that enhance soil organic carbon accumulation (Wang et al., 2025). During their vegetative phase, green manure crops assimilate nitrogen into biomass, lowering soil nitrate concentrations and scavenging residual inorganic nitrogen following cash crop harvest (Basche et al., 2014; Kladivko et al., 2014). Following termination and soil incorporation, green manure biomass gradually releases nitrogen back into the soil, potentially lowering synthetic nitrogen fertilizer requirements and regulating inorganic nitrogen dynamics. In support of this mechanism, Behnke and Villamil (2019) reported that cover cropping in Illinois (USA) reduced N₂O emissions and soil nitrate levels through active plant nitrogen uptake. Similarly, Bharati et al. (2020) demonstrated that green manure species such as cowpea, sesbania, and mung bean effectively suppressed N₂O emissions in paddy fields. Mineralization of cover crop biomass not only moderates nitrogen release but can also enhance subsequent crop yields. Huang et al. (2004) found that decomposing legume cover crops released nitrogen progressively, reducing mineral fertilizer requirements for the succeeding cash crop. However, a meta-analysis by Basche et al. (2014) revealed that cover crop inclusion was associated with increased N₂O emissions in 60% of evaluated studies, regardless of species. We hypothesize that these discrepancies derive from differences in residue C/N ratios, cropping system architecture (e.g., main vs. intermediate crops), and agronomic management practices.

In our experiment, narrow-leafed lupin cultivated as an intercrop green manure resulted in elevated N₂O emissions, particularly during the first-year post-incorporation (Table 2). This increase can be attributed to the relatively high nitrogen content of lupin biomass (1.87%) and its narrow C/N ratio (16.2). Additionally, this emission surge aligns with a short-term priming effect induced by the addition of fresh, labile organic matter to the soil, alongside a transient expansion of reactive nitrogen pools. This phenomenon, first described by Löhnis (1926), has since evolved into a fundamental concept in soil organic matter dynamics (Fontaine et al., 2003; Kuzyakov, 2010; Li et al., 2022; Daly et al., 2024).

When lupin green manure is co-applied with straw, the combined residue C/N ratio widens, promoting nitrogen immobilization. Notably, the most compelling finding of our study was the consistent mitigation of soil N₂O emissions following mineral fertilizer application in treatments receiving both straw and lupin green manure as supplemental carbon sources. This confirms the conclusion of Kuzyakov and Xu (2013) that microbially available carbon is the primary driver of metabolic assimilation of unutilized mineral nitrogen in soil. This mitigating effect persisted during the initial application year and through the first and second years of residual action. However, the second year represented the limit of residual efficacy in stabilizing these gaseous fluxes.

Differentiating the microbially driven processes governing N₂O generation in situ remains challenging. One plausible mechanism constraining N₂O emissions in our study is the substrate limitation imposed on nitrifying and denitrifying microorganisms through enhanced microbial nitrogen immobilization, driven by combined carbon inputs from straw and green manure alongside functional shifts in the soil microbiome.

A promising strategy for further suppressing N₂O emissions involves the application of biological preparations based on plant growth-promoting rhizobacteria. Several studies have documented the efficacy of pre-sowing inoculation with bacteria capable of expressing N₂O reductase and utilizing N₂O as an electron acceptor for reduction to dinitrogen (N₂) (Thomson et al., 2012; Butterbach-Bahl et al., 2013; Hu et al., 2015). For instance, Gao et al. (2016) showed that inoculating red clover with denitrifying strains from the genera Azospirillum and Herbaspirillum significantly reduced soil N₂O emissions under greenhouse conditions, while no such effect occurred with timothy grass. Similarly, Itakura et al. (2013) found that inoculating soybean seeds with Bradyrhizobium japonicum suppressed N₂O emissions. More recently, Stutzerimonas stutzeri (formerly Pseudomonas stutzeri) has shown strong potential for mitigating soil N₂O emissions (Gao et al., 2024).

In addition to direct biological denitrification pathways, PGPR can indirectly reduce gaseous nitrogen losses by stimulating plant root architecture and shoot growth, thereby increasing crop nitrogen uptake (Calvo et al., 2016; Volkogon et al., 2021; Volkogon et al., 2022). This plant-mediated mitigation mechanism is further supported by studies involving eukaryotic organisms such as Trichoderma guizhouens (Geng et al., 2021) and Trichoderma viride (Xu et al., 2018), which demonstrate similar crop uptake-driven emission reductions.

In our trials, the PGPR strains established robust rhizosphere populations in their respective host crops (Sydorenko et al., 2020) and exhibited high agronomic efficacy. Two evaluated strains, A. brasilense 410 and R. leguminosarum 348a, possess functional denitrification capabilities. In contrast, P. polymyxa KB, applied as a pre-sowing seed inoculant for winter wheat, lacks denitrifying capacity. Nonetheless, its application likewise led to a significant reduction in N-N₂O losses. This observation strongly supports the hypothesis that inoculation-enhanced plant nitrogen assimilation depletes the soil mineral nitrogen pool accessible to native nitrifiers and denitrifiers, thereby mitigating gaseous nitrogen losses.

Our results demonstrate that pre-sowing seed inoculation of pea with R. leguminosaru substantially suppressed N₂O emissions. This might appear unexpected, given that this bacterium is a symbiotic nitrogen fixer primarily responsible for meeting plant nitrogen requirements. However, this suppression can be explained by mechanisms shared with other PGPR. Rhizobia are not functionally restricted to atmospheric nitrogen fixation in legume symbiosis. Biological nitrogen fixation typically satisfies only 50-80% of grain legumes nitrogen demand (Salvagiotti et al., 2009), as grain protein synthesis requires substantial nitrogen quantities exceeding symbiotic supply capacity (Sanchez, 2019). Consequently, legume plants concurrently rely on soil mineral nitrogen uptake alongside symbiotic fixation (Lepetit and Brouquisse, 2023). Inoculation directly enhances this mineral nitrogen acquisition: as is known phytohormone production by rhizobial bacteria stimulates root system expansion, thereby accelerating nutrient flux from the bulk soil (Concha and Doerner, 2020; Habtewold and Goyal, 2023). Crucially, phytohormone-induced root growth stimulation occurs independently of soil nitrogen levels, unlike root nodulation (Garrido-Oter et al., 2018).

While mineral nitrogen fertilizer application remains critical for maintaining target yields and ensuring global food security. However, since the use of mineral nitrogen is associated with substantial gaseous losses via N₂O emissions agricultural practices should seek operational trade-offs that optimize crop productivity while minimizing soil nitrogen losses. Our findings indicate that integrating mineral nitrogen fertilizers with straw, intercropped green manures, and pre-sowing microbial inoculation provides an effective management strategy to suppress N₂O losses and improve yields within crop rotations. These practices directly align with the principles of sustainable agricultural intensification: achieving higher yields from existing arable land while minimizing environmental externalities and enhancing ecological services (Pretty, 2008).

Addressing these challenges requires re-evaluating certain established agronomic dogmas. Conventional management typically mandates co-applying straw with supplemental mineral nitrogen (generally at least 10 kg N per metric ton of straw) to avoid transient nitrogen immobilization (Hutchinson and Richards, 1921; Rathke et al., 2006). When organic residues are incorporated into soils depleted of nitrogen by long-term unfertilized management, microbial immobilization of scarce soil nitrogen inevitably compromises crop nutrition. However, our study proceeds from the assumption that supplemental nitrogen compensation is unnecessary under modern, high-input fertilizer regimes, where a fraction of applied mineral nitrogen routinely remains unassimilated by the crop. Rather than applying supplemental nitrogen, this residual pool should be biologically immobilized into organic forms. Straw and related crop residues should thus serve as an exogenously supplied carbon source to drive microbial assimilation of surplus fertilizer nitrogen, simultaneously advancing soil carbon sequestration (Xia et al., 2018).

Our experimental data demonstrate consistent crop yield increases with mineral fertilization alone, with substantially higher gains achieved when fertilizers were combined with straw and green manure biomass. Pre-sowing PGPR inoculation provided additional productivity gains across the crop rotation. Therefore, to effectively mitigate N₂O emissions while enhancing cropping system performance, mineral nitrogen applications should be integrated with organic carbon residues and targeted PGPR inoculation.

5 Conclusions

Integrated application of mineral nitrogen fertilizers with straw and narrow-leafed lupin green manure reduced soil N₂O emission rates compared to mineral fertilization alone. This mitigating effect was primarily observed in the year of direct incorporation of organic residues and persisted over the subsequent two years of residual action, resulting from the alignment of substrate availability with microbial stoichiometric carbon demand and enhancing nitrogen immobilization.

The capacity of PGPR to suppress gaseous nitrogen losses may result from inoculation-driven increases in plant fertilizer nitrogen uptake, as well as direct bacterial enzymatic reduction in specific cases. Overall, combining mineral fertilizers with organic residues and PGPR decreased N₂O emission rates by 16-29%. These findings highlight the strong potential of the evaluated agronomic practices for achieving sustainable nutrient management in crop production.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fagro.2026.1959485/full#supplementary-material

Supplementary material Supplementary material at the source

Footnotes

Edited by: Rajiv Kumar Srivastava, Texas A&M University, United States

Reviewed by: Prabu Padanillay Chidambaram, Tamil Nadu Agricultural University, India

Giulia Forghieri, Ca’ Foscari University of Venice, Italy