Nitrogen Fertiliser Cuts Carbon Loss on Farms but Slows Recovery in Grassland Soils

Soil is one of the planet’s largest carbon reservoirs, and the way we farm it can either drain that reservoir or refill it. A remarkable 47-year field experiment on a calcareous chernozem in Hungary has now delivered one of the clearest long-term pictures yet of how nitrogen fertiliser interacts with land use to shape soil organic carbon. The study, published in the journal Plant and Soil, tracked the same experimental plots from continuous cropping through a dramatic switch to managed grassland, measuring both total soil organic carbon (SOC) and its fast-cycling, labile fraction known as permanganate-oxidizable carbon (POXC). The headline finding is strikingly two-sided: nitrogen fertiliser proved to be a powerful shield against carbon loss under arable farming, yet it actively slowed the recovery of carbon once the land was converted to grassland.

The experiment began in autumn 1973 at the Nagyhörcsök Research Station of the HUN-REN Centre for Agricultural Research, on a calcareous chernozem soil with a water pH of 8.24, a cation exchange capacity of 28.3 meq per 100 grams, and an initial organic carbon content of 3.04 percent in the top 20 centimetres. Plots received one of four annual nitrogen doses: zero, 100, 200 or 300 kilograms of nitrogen per hectare, applied as calcium ammonium nitrate and split between autumn and spring. For the first decades, the plots grew a rotating sequence of field crops under conventional ploughing, with stalks and residues ploughed back into the soil after harvest. Then, in autumn 2000, the entire experimental area was sown with a mixture of eight grass species, and the sward was mown once or twice a year with all above-ground biomass removed.

The researchers sampled the topsoil at three pivotal moments: in 1990 after 17 years of cultivation, in 2000 after 27 years of cropping, and in 2020 after two decades of grassland management. SOC was determined with the modified Walkley–Black method, while POXC was measured by shaking one gram of soil with 20 millimolar potassium permanganate solution and reading the absorbance at 550 nanometres, a protocol chosen because it has proven effective for distinguishing land-use effects in chernozem soils. Bulk density was measured with sampling cylinders, and SOC stocks were calculated for the 0–20 centimetre layer. The team then deployed repeated measures linear mixed models and structural equation modelling to disentangle the direct and indirect pathways linking fertiliser, plant biomass, biodiversity and soil carbon.

The cropland results tell a sobering story about the cost of ploughing. In the unfertilised control plots, SOC content fell significantly from 1.91 percent in 1990 to 1.74 percent in 2000, equivalent to a staggering loss of 471 kilograms of carbon per hectare every year. Even after 27 years of cropping, the soil had still not reached a new equilibrium; depletion was ongoing. But nitrogen fertiliser changed the picture dramatically. Across the fertilised treatments, the average annual SOC stock change dropped to just 41 kilograms per hectare per year, a roughly 90 percent reduction in the rate of loss. Notably, even the highest dose of 300 kilograms of nitrogen per hectare could not flip the trend into net carbon gain under continuous cropping, but a modest 100 kilograms per hectare was enough to essentially halt the decline.

The mechanism, the structural equation models suggest, runs almost entirely through plant biomass. Nitrogen fertiliser boosts crop growth, and the extra residues returned to the soil replenish some of the carbon that tillage strips away. Interestingly, in the early years of the experiment nitrogen had little effect on crop biomass on this naturally nitrogen-rich chernozem, which explains why SOC differences between treatments only emerged clearly by the 2000 sampling. The crops, with their relatively short growing seasons, simply could not use more than about 100 kilograms of nitrogen per hectare effectively, so higher doses brought no additional biomass and no additional carbon benefit.

Then came the grassland conversion, and with it a reversal of fortune. Twenty years of mown grassland lifted SOC content to an average of 2.24 percent, the highest of the three land-use phases, and in the unfertilised plots carbon accumulated at a rate of 425 kilograms per hectare per year, a figure consistent with other studies of cropland-to-grassland conversions in temperate zones, which typically report 400 to 580 kilograms per hectare annually. Yet here nitrogen fertiliser worked against the soil. At the 200-kilogram dose, the sequestration rate fell to just 164 kilograms per hectare per year. The authors point to several likely explanations: mowing removes most of the above-ground carbon, so soil inputs depend heavily on roots, and high nutrient supply can suppress root growth; meanwhile, abundant mineral nitrogen can stimulate microbial decomposition of existing organic matter. Even after two decades of grass, the SOC content of 2.30 percent in the best plots remained far below the original 3.04 percent measured in 1973, underscoring how slow carbon recovery really is.

The grassland also revealed a tension between two ecological goods. Nitrogen fertilisation cut plant species diversity by about 20 percent, eliminated legumes from all but the control plots, and shifted the sward toward nutrient-loving perennials. Ecological theory and prior studies suggest that higher biodiversity should accelerate carbon sequestration, and the researchers had hypothesised that diversity might mediate part of the fertiliser effect. The structural equation models told a different story: when biomass and biodiversity were tested as parallel mediators, the biodiversity pathway was negligible compared with the biomass pathway. In other words, on this intensively managed, fertilised grassland, the sheer quantity of plant production mattered far more for soil carbon than the variety of species producing it, and nitrogen’s suppression of diversity did not override its biomass-driven effects.

The labile carbon fraction, POXC, behaved less predictably. Averaged across the three sampling dates, POXC rose from 352 milligrams per kilogram in the unfertilised controls to 399 at the highest nitrogen dose, but the response flipped between time points: the 100-kilogram dose produced the highest POXC in 1990 yet the lowest in 2000 and 2020. Land use proved more decisive than fertilisation, with POXC dropping to a low of 308 milligrams per kilogram at the end of the cropping period and rebounding to 413 under grassland. Crucially, the correlation between POXC and total SOC weakened progressively as nitrogen doses increased, suggesting that elevated nitrogen accelerates the microbial mineralisation of labile, nitrogen-rich residues and decouples the fast carbon pool from the slow one.

That decoupling has practical consequences for the growing number of farmers and researchers who use POXC as a quick soil-health test. The study concludes that POXC works reasonably well as a proxy for SOC dynamics only at fertilisation rates of 100 kilograms of nitrogen per hectare or less; above that, its reliability breaks down. The POXC-to-SOC ratio, which fell during cropping and began climbing again under grassland, may still serve as an early-warning indicator of management-driven shifts, but even it responded inconsistently to fertilisation across land uses. The authors also caution that their interpretations are constrained by the 0–20 centimetre sampling depth, which captures the plough layer but misses deeper carbon dynamics, and by the fact that bulk density had to be measured decades after the original samplings.

The broader message is one of nuance in an era when soils are increasingly counted among climate solutions. Nitrogen fertiliser is neither hero nor villain: on tilled cropland it prevented nearly all of the carbon that would otherwise have been lost, while on restored grassland it quietly throttled the very sequestration the conversion was meant to deliver. For the world’s chernozem black soils, which are both major carbon stores and the backbone of global food production, the study suggests that the most effective carbon strategy is land use itself, with fertiliser playing a supporting role that must be tuned to the system. Leave residues on the field and fertilise moderately, and arable soil can hold its carbon. Convert to grassland and keep fertiliser low, and the soil will begin, slowly, to give it back.

Subject of Research: Effects of nitrogen fertilisation and land use on soil organic carbon and labile carbon in a calcareous chernozem

Article Title: Effect of nitrogen fertilisation under different land use on the total and labile organic carbon content in a calcareous chernozem under temperate climate

Article References: Effect of nitrogen fertilisation under different land use on the total and labile organic carbon content in a calcareous chernozem under temperate climate. (n.d.). https://doi.org/10.1007/s11104-026-09114-1

Image Credits: AI Generated

DOI: 10.1007/s11104-026-09114-1

Keywords: soil organic carbon, nitrogen fertilisation, chernozem, grassland, cropland, POXC, labile carbon, carbon sequestration, biodiversity, plant biomass, long-term field experiment, soil health