Decades of Nitrogen Pollution Quietly Rewire How Grassland Soils Store Carbon

On the wind-blown terraces of China’s Loess Plateau, where decades of grazing exclusion have slowly coaxed degraded grasslands back to life, an invisible experiment has been running for twelve years. Scientists added nitrogen to plots of restored grassland and watched, year after year, as the soil beneath them changed in ways that could reshape how we manage the world’s carbon reserves. The results, published in the journal Plant and Soil, reveal that chronic nitrogen enrichment does not simply add carbon to grassland soils or remove it. Instead, it reshuffles the very architecture of soil carbon, shifting it between forms that differ dramatically in how long they can lock carbon away.

The research team, led by Xiaoshan Zhang of Northwest A&F University together with colleagues from multiple Chinese institutions and Massey University researcher Hossein Ghanizadeh, focused on one of the most consequential questions in terrestrial ecology: how does sustained nitrogen input affect the fractions of soil organic carbon, or SOC, in grasslands at different stages of restoration? Their answer required patience. In 2013, they established nitrogen addition treatments at three sites on the Loess Plateau that had been fenced off from grazing for roughly ten, twenty, and thirty years. By the time they sampled the soils in 2024, those sites represented approximately twenty-one, thirty-one, and forty-one years of grazing exclusion, a natural laboratory spanning the full arc of grassland recovery.

The distinction the researchers drew between carbon fractions is central to the study’s significance. Soil organic carbon is not a homogeneous pool. Particulate organic carbon, or POC, consists of partially decomposed plant fragments and other relatively fresh organic debris that sits loose in the soil matrix. It is chemically vulnerable, turning over within years to decades and responding quickly to changes in plant input and microbial activity. Mineral-associated organic carbon, or MAOC, by contrast, is bound to fine soil particles and mineral surfaces, where chemical bonds and physical protection shield it from decomposers. This fraction can persist for decades to centuries, making it the true long-term vault of the terrestrial carbon cycle. Which fraction gains or loses carbon determines whether grassland soils become durable carbon sinks or temporary ones.

Across all three sites, the twelve-year nitrogen treatment produced a consistent and directional signature in the topsoil, the zero to twenty centimeter layer where most grassland carbon resides. Particulate organic carbon declined by between 0.4 percent and 37 percent, and the proportion of total SOC held as POC fell accordingly. Mineral-associated organic carbon moved in the opposite direction, rising by 0.6 percent to 32 percent, with the MAOC share of total SOC increasing. In other words, nitrogen addition pushed the soil’s carbon portfolio away from the fast, fragile pool and toward the slow, stable one. The net effect on total SOC depended on the site, but the rebalancing between fractions occurred everywhere.

To understand why, the team measured a suite of ecological variables at each site: aboveground plant biomass, plant diversity as captured by the Shannon index, soil pH, and the availability of inorganic nitrogen in the soil. The pattern of vegetation responses alone was instructive. Aboveground biomass peaked at the site with twenty years of enclosure history, while plant diversity was highest at the youngest, ten-year site. Nitrogen addition increased biomass at that youngest site but simultaneously reduced plant diversity, echoing a well-documented global pattern in which nutrient enrichment lets productive species shade out their neighbors and simplifies plant communities. Biomass and biodiversity, it turns out, respond along divergent paths as restoration matures.

The mechanistic core of the study came from structural equation modeling, a statistical technique that allows researchers to map networks of associations among variables rather than testing each link in isolation. The models pointed to two contrasting pathways connecting nitrogen addition to the carbon fractions. The first ran through the plants: nitrogen addition was associated with higher aboveground biomass, and shifts in productivity were in turn linked with changes in the particulate carbon pool, the fraction most directly fed by plant litter and root fragments. The second ran through the soil itself: nitrogen addition was associated with both lower soil pH, the familiar fingerprint of nitrogen-induced acidification, and greater inorganic nitrogen availability, and these edaphic changes were linked with the mineral-associated pool, whose formation depends heavily on soil chemistry and the fate of nitrogen as it cycles through the profile.

This plant-versus-soil dichotomy helps resolve a puzzle that has long bedeviled nitrogen experiments worldwide. Some studies report nitrogen-driven carbon gains in grassland soils, others report losses, and the variability is often attributed to climate or soil texture. The Loess Plateau findings suggest that restoration stage is another critical moderator. Enclosure duration itself was positively associated with SOC, POC, and MAOC across the sites, confirming that grazing exclusion builds carbon in both pools. But because nitrogen acts on those pools through different channels, its net effect on total carbon storage can flip depending on how far along the recovery trajectory a grassland happens to be.

The soil chemistry pathway deserves particular attention in a warming world. When nitrogen accumulates faster than plants and microbes can absorb it, ammonium oxidation and nitrification release hydrogen ions that acidify the soil. Acidification alters the surface charge of clay minerals and metal oxides, which can change how strongly organic molecules bind to mineral surfaces, the very mechanism that protects MAOC. At the same time, elevated inorganic nitrogen availability can suppress the microbial enzymes that microbes use to mine carbon from stable organic matter, potentially slowing decomposition of mineral-bound carbon. The structural equation models did not isolate a single one of these mechanisms, but they established that soil pH and inorganic nitrogen availability are the variables most tightly associated with shifts in the mineral-associated pool, giving future studies a precise target.

The findings arrive at a moment when grasslands are increasingly promoted as carbon sequestration assets. Grasslands store a substantial share of terrestrial soil carbon, and restoring degraded ones through grazing exclusion is among the cheapest, most scalable interventions available. Recent work in Science and other leading journals has championed this strategy. But the Loess Plateau study adds a cautionary nuance: grasslands are not being restored in a pristine atmosphere. Atmospheric nitrogen deposition has risen across much of the world over the past century, and it delivers a chronic, low-level fertilization that no fence can keep out. If deposition is quietly converting particulate carbon to mineral-associated carbon while acidifying restored soils, the carbon arithmetic of restoration projects may look different over a century than over a decade.

For land managers, the practical implications are twofold. First, the total carbon benefit of nitrogen enrichment cannot be read from biomass alone. Lusher aboveground growth can coexist with declining particulate carbon and shrinking plant diversity, meaning field inventories that measure vegetation may miss what is happening below. Second, restoration stage should inform nitrogen policy: sites in early recovery, like the ten-year enclosure site, responded to nitrogen with clear biomass gains and diversity losses, while older, more mature grasslands showed different carbon trajectories, suggesting that the same deposition load will not produce the same outcome across a landscape of differently aged restorations. The study’s twelve-year span, one of the longest continuous nitrogen treatments applied to restored grasslands on the Loess Plateau, is what gives the researchers confidence that these are not short-lived fertilization artifacts but durable shifts in the soil’s carbon architecture. Whether the gains in mineral-associated carbon can ultimately compensate for the losses in particulate carbon remains the critical open question, and one that the authors’ long-term dataset positions them uniquely to answer in the decades ahead.

Subject of Research: Effects of long-term nitrogen addition on soil organic carbon fractions in restored Loess Plateau grasslands

Article Title: Long-term nitrogen addition alters topsoil SOC fractions via contrasting plant- and soil-mediated pathways in restored grasslands

Article References: Long-term nitrogen addition alters topsoil SOC fractions via contrasting plant- and soil-mediated pathways in restored grasslands. (n.d.). https://doi.org/10.1007/s11104-026-09053-x

Image Credits: AI Generated

DOI: 10.1007/s11104-026-09053-x

Keywords: soil organic carbon, nitrogen addition, grassland restoration, particulate organic carbon, mineral-associated organic carbon, grazing exclusion, Loess Plateau, soil acidification, plant diversity, structural equation modeling, carbon sequestration, topsoil