Nitrogen's Effect on Soil Respiration Flips Over Time, Study Finds

How did the researchers reach their findings?

The team compiled data from manipulation experiments conducted in forests and grasslands worldwide. In such experiments, researchers deliberately add nitrogen to plots of land and measure the resulting soil emissions, creating a controlled record of cause and effect.

The analysis produced a clear pattern. The effect of nitrogen addition on Rs changed systematically over time, shifting from short-term stimulation of soil respiration to long-term suppression.

That shift followed a nonlinear temporal trajectory: Rs declined rapidly at first, and the rate of decline gradually leveled off, much as a ball rolling down a hill loses speed as it approaches the valley floor.

The pace of decline, the researchers found, did not depend significantly on the rate at which nitrogen was added. But higher nitrogen inputs did produce stronger suppression overall.

When the team classified treatments into low- and high-rate categories based on the median addition rate in the analyzed data set — 50 kilograms of nitrogen per hectare per year — the difference became measurable. High-rate addition cut Rs by roughly 11 percent. Low-rate addition left no significant overall imprint.

What does this change for climate models?

The study extends what the authors describe as the prevailing static-response paradigm, in which nitrogen's effect on soil carbon is assumed to hold steady. The new findings show instead that the effects of nitrogen enrichment on soil carbon fluxes depend strongly on two factors: how long the exposure lasts and how much nitrogen has accumulated in total.

This distinction carries practical weight. Current models of land–atmosphere carbon exchange, the researchers argue, need to capture these temporal dynamics rather than assume a constant response. Without that correction, predictions of how ecosystems absorb and release carbon under global change — including ongoing nitrogen deposition from industrial and agricultural emissions — risk systematic error.

The results remain, as with any synthesis of field experiments, dependent on the coverage of the underlying data. The study draws on forests and grasslands worldwide, but the researchers' conclusion about exposure duration and cumulative input is a model-level inference drawn from those pooled experiments rather than a single long-running trial.

The paper, by Aijun Xing et al., appears in the Proceedings of the National Academy of Sciences (2026), DOI: 10.1073/pnas.2622400123.

The ground beneath forests and grasslands, the data suggest, does not respond to nitrogen once and settle; it responds, slows, and settles into a new balance.