Agricultural nitrogen is essential for food production, but the same nutrient that supports crop growth can become a major source of pollution when it escapes from soil.
New global research suggests that the problem is not only how much nitrogen farmers add or how quickly individual soil reactions occur. What may matter just as much is whether the different stages of the soil nitrogen cycle remain coordinated with one another.
A study published in Nature Communications analysed 2,235 paired observations comparing organic substitution with synthetic fertilization across 210 experimental sites worldwide. On average, replacing at least part of synthetic fertilizer with organic amendments reduced nitrous oxide emissions by 21.8% and ammonia emissions by 40.6%.
The researchers argue that these reductions are best understood through what they call nitrogen-process synchrony. In simple terms, nitrogen appears less likely to accumulate in forms that can escape into the atmosphere when mineralization, immobilization, nitrification and denitrification operate in closer coordination.
The study looked beyond individual soil reactions
Much of agricultural nitrogen research examines individual processes. Mineralization releases plant-available nitrogen from organic matter. Immobilization incorporates available nitrogen into microbial biomass. Nitrification converts ammonium into oxidized forms such as nitrate, while denitrification can return nitrogen to the atmosphere.
Each process matters, but studying them separately can miss an important systems problem. If one stage produces a nitrogen compound faster than another stage can consume it, that intermediate can accumulate. The result can be greater opportunity for nitrogen to leave the soil as nitrous oxide or ammonia.
To test this idea at a global scale, Mengfei Li and colleagues assembled paired observations in which synthetic fertilization served as the comparison condition and organic fertilization as the treatment. Organic amendments included manure, returned crop straw, biochar and organic soil conditioners. The experiments ranged from a single cropping season to more than 50 years, although 675 observations, roughly 30% of the dataset, came from one-year experiments.
The team used a hierarchical mixed-effects meta-analysis to estimate treatment effects and then adapted a synchrony metric from community ecology. A synchrony deviation of zero represented the idealized state in which a nitrogen process changed in step with the other processes. Larger deviations represented greater mismatch.
This is an important distinction. The researchers were not simply asking whether nitrification rose or fell after organic substitution. They were asking whether its change remained proportionate to changes elsewhere in the nitrogen cycle.
Both major nitrogen pollutants declined
Across the global dataset, nitrous oxide emissions were 21.8% lower under organic substitution than under synthetic fertilization alone. The 95% confidence interval ranged from a 31.6% reduction to a 10.6% reduction, and the result was statistically significant at p<0.001.
The reduction in ammonia was larger. Emissions fell by 40.6%, with a 95% confidence interval from 46.9% to 33.5% lower and p<0.001.
Those averages are environmentally meaningful because the two gases create different problems. Nitrous oxide is a long-lived greenhouse gas, while atmospheric ammonia contributes to fine particulate matter formation and can disturb ecosystems after redeposition.
But the most revealing result was what did not explain the reductions. Several nitrogen transformations actually became faster after organic substitution. Mineralization increased by 34.0%, nitrification by 6.2% and denitrification by 24.4%. Immobilization showed the largest change, rising 2.24-fold.
If faster individual processes automatically meant greater pollution, those increases could appear contradictory. Yet changes in individual process rates did not reliably explain the observed reductions in gas losses. Ammonia changes were not significantly correlated with mineralization, while nitrous oxide changes were not significantly correlated with either nitrification or denitrification.
Coordination explained what individual rates could not
The synchrony analysis produced a different picture. Under organic substitution, synchrony metrics moved closer to the ideal coordinated state. The researchers reported improvements of 87.0% for mineralization, 55.2% for immobilization, 69.6% for nitrification and 83.6% for denitrification.
Structural equation modelling then showed that synchrony deviations in nitrification and denitrification jointly accounted for about 50.0% of the explained variation in nitrous oxide emissions. For ammonia, deviations involving mineralization and immobilization jointly explained 37.8% of the variation.
The interpretation is intuitive. When nitrification and denitrification become mismatched, reactive nitrogen intermediates can accumulate. Likewise, ammonia loss is strongly influenced by how the production of ammonium through mineralization balances against its capture through microbial immobilization and other pathways.
Organic amendments may help because they generally release nitrogen more gradually than a concentrated synthetic fertilizer pulse. They also supply carbon that can support a larger and more active microbial community. In the meta-analysis, soil organic carbon increased by 36.3% and total soil nitrogen by 17.8%. Microbial biomass carbon rose by 47.4%, while microbial biomass nitrogen increased by 29.1%.
These changes do not prove that every organic amendment will improve every field. They do, however, support the broader idea that nitrogen management can be evaluated as a connected biological system rather than a set of isolated reactions.
The potential benefits differed sharply by region
The researchers combined their meta-analytic results with global climate, soil and management data in a random-forest model at a spatial resolution of 0.5 degrees by 0.5 degrees. The projections indicated substantial regional variation.
European croplands showed the largest projected reduction in nitrous oxide emissions at 40.7%, followed by the Americas at 31.6%. For ammonia, the strongest projected benefit occurred in Africa, where emissions were estimated to fall by 46.1%. Europe followed with a projected 44.2% reduction.
The African result is particularly relevant because many agricultural soils on the continent have relatively low organic matter and limited nitrogen-retention capacity. The authors propose that slower nitrogen release from organic inputs and additional carbon for soil microorganisms could reduce short-lived ammonium accumulation that otherwise favours ammonia volatilization.
That should not be read as a recommendation to simply replace all synthetic fertilizer with manure or other organic material. The effects varied with climate, soil pH, soil carbon, precipitation, substitution intensity and experimental duration. Some specific substitution types also did not show the same reductions as the overall average.
A different way to think about fertilizer efficiency
The study shifts attention from a familiar question, how much nitrogen is entering the field, toward another question, how efficiently the entire soil system processes that nitrogen after it arrives.
That matters for both environmental modelling and farm management. Models that represent nitrogen processes as largely independent components may misestimate emissions when one process changes the substrate available to another. The authors argue that incorporating synchrony could improve predictions of nitrous oxide and ammonia losses.
For management, the findings suggest that the best intervention may depend on local conditions. A practice that improves coordination in one soil and climate may not produce the same response elsewhere. The regional projections therefore point toward targeted nitrogen strategies rather than a universal fertilizer recipe.
The results also complicate the assumption that reducing emissions necessarily requires slowing nitrogen cycling. Several transformation rates increased while gas losses fell. What appeared more important was whether the processes remained balanced enough to prevent reactive nitrogen from accumulating between stages.
Important limitations remain
The analysis combines experiments conducted under widely different soils, climates, crops, amendment types and durations. That diversity is valuable for identifying broad patterns, but the experimental sites were unevenly distributed around the world and cannot represent every farming system equally.
About 30% of the observations came from experiments lasting only one year. Short experiments may not capture slower changes in soil carbon, microbial communities or nutrient pools that emerge after repeated organic applications.
The synchrony metric was also calculated from several kinds of indicators, including measured process rates, functional gene abundances and enzyme activities. The researchers found broadly consistent patterns across these approaches, but they acknowledge that gene abundance and enzyme activity introduce additional uncertainty and that direct measurements of nitrogen transformation rates may provide stronger estimates.
Publication-bias tests detected potential funnel-plot asymmetry for several variables. The authors report that fail-safe analyses suggested this was unlikely to overturn the main conclusions, but it remains another reason to interpret the exact global effect sizes with appropriate caution.
Finally, the global maps are modelled projections rather than the results of continent-wide fertilizer experiments. The estimated 46.1% ammonia reduction for African croplands, for example, identifies mitigation potential under the model and should not be interpreted as a guaranteed reduction for an individual farm.
The broader lesson is about balance
Agricultural nitrogen pollution is often framed as a problem of excess. This study adds another dimension: imbalance.
Organic substitution was associated with substantially lower nitrous oxide and ammonia emissions, but the analysis suggests that the mechanism cannot be reduced to simply slowing one microbial process. Instead, the strongest explanatory signal came from how closely several processes moved together.
If that finding holds across future field experiments, nitrogen-process synchrony could become a useful way to evaluate fertilizer practices, improve agricultural emission models and identify where organic amendments are most likely to reduce pollution without treating every soil as if it behaves the same way.
Source Information
Study Title: Synchronizing nitrogen cycling processes reduces agricultural nitrous oxide and ammonia emissions
Authors: Mengfei Li, Yanzhong Yao, Bingbing Han and colleagues
Journal: Nature Communications
Published: 2026
Dataset: 2,235 paired observations comparing organic substitution with synthetic fertilization across 210 experimental sites worldwide
Method: Hierarchical mixed-effects meta-analysis, nitrogen-process synchrony metrics, structural equation modelling and global random-forest spatial projections using climate, soil and management data
Main finding: Organic substitution reduced nitrous oxide emissions by 21.8% and ammonia emissions by 40.6% on average. Synchrony deviations among nitrogen transformations explained substantially more emission variation than individual process changes, with the largest modelled mitigation potentials for nitrous oxide in European croplands and ammonia in African croplands.
DOI: 10.1038/s41467-026-76977-4
