What if part of the carbon stored in America’s richest grassland soils depends on something as ordinary as calcium? This study links mineral weathering, agriculture, and soil carbon across the entire conterminous United States. Temperate grasslands feature fertile, dark-colored soils known as Mollisols, and Mollisols are the world’s most intensively farmed soil type.

Mollisols comprise only 7% of the Earth’s land area, yet they support 24% of agricultural land and 32% of all calories from corn, soy, and wheat production. They also contain 32% of the organic carbon held in Earth’s agricultural soils, making Mollisols critical to land-based climate mitigation efforts.

At the same time, Mollisols are disproportionately exposed to threats including soil erosion and have lost a significant fraction of their soil organic carbon to cultivation. Geologic factors can help explain the carbon richness of Mollisols: these soils commonly form from calcareous rocks or eolian deposits rich in calcium carbonate.

Mollisols are enriched in exchangeable calcium, and a growing body of evidence shows that calcium stabilizes soil organic carbon. Calcium facilitates sorption of organic molecules on clays and promotes bacterial biofilms that are retained on mineral surfaces. These facts suggest that calcium released by carbonate and silicate dissolution may link soil organic carbon storage to mineral weathering in Mollisols.

To address calcium’s role in Mollisol soil carbon storage, the study developed and applied a geochemical model across the conterminous United States using soil mineralogy data. The model tracked inputs and outputs of major cations and anions to the soil solution, while simulating soil pH, mineral weathering rates, and associated calcium release and retention.

The study estimated the effect of calcium availability on soil organic carbon statistically while accounting for vegetation, climate, and other confounding factors. The model separated carbonate weathering, silicate weathering, and atmospheric deposition, then quantified the sensitivity of carbon storage to weathering-derived calcium in Mollisols versus non-Mollisol soils.

The geochemical model estimated calcium availability across the conterminous United States using climate, atmospheric deposition, net primary productivity, agricultural element budgets, and soil mineralogy as input data. It linked solute inputs and outputs, biological release of carbon dioxide and organic acids, and mineral weathering through pH-dependent equilibrium reactions.

After calibrating two parameters related to mineral surface reactivity, the model explained 69% of the spatial variation in depth-averaged soil pH across the United States. For validation, it explained 37% of the variation in the soil exchangeable calcium fraction and 61% of the variation across all exchangeable cations.

The study used total A-horizon soil organic carbon, incorporating both topsoil thickness and carbon content, as an index of the primary carbon-storage mode in Mollisols. Soil organic carbon stock data came from the United States Department of Agriculture Rapid Carbon Assessment and the National Cooperative Soil Survey archives.

The relationship between A-horizon carbon stocks and modeled exchangeable calcium was quantified across the conterminous United States with multiple linear regression. The regression included eight potentially confounding factors, including vegetation, soil temperature, precipitation, texture, drainage, pH, net primary productivity, and non-calcium exchangeable cations.

In the coefficient plot, the NPP point is farthest from zero on the positive side, with its interval clearly separated from the negative MAT estimate. The exchangeable-calcium and grassland points sit near fourteen hundredths, while the remaining predictors cluster closer to zero.

The regression model explained 30% of the variation in A-horizon soil organic carbon stocks. Net primary productivity and soil temperature were the most important predictors, followed by exchangeable calcium, the presence of grasslands or savannas, and then other factors.

Modeled exchangeable calcium was positively correlated with A-horizon carbon storage, while the combined effect of other exchangeable ions was weak. To represent preindustrial conditions, the model used atmospheric deposition from 1850, eliminated agricultural processes, and ran simulations under preindustrial carbon dioxide levels.

The experiment set carbonate weathering, all silicate weathering, or atmospheric calcium deposition to zero, then used the updated model outputs to predict changes in A-horizon soil organic carbon stocks. A counterfactual scenario also assumed that all vegetation was non-grassland.

The model experiment found that both carbonate and silicate weathering help explain A-horizon soil organic carbon stocks across the United States. Across Mollisol regions, predicted preindustrial carbon storage was 5% lower when carbonate weathering was eliminated, 3% lower when silicate weathering was eliminated, and 15% lower when both were eliminated together.

Eliminating all weathering had an effect similar to eliminating grassland vegetation, which reduced A-horizon carbon storage by 20%. Mollisol carbon stocks were particularly sensitive to calcium inputs from carbonate weathering and to the combined action of carbonates and silicates.

Eliminating carbonate weathering alone had a modest effect on A-horizon carbon storage, but the effect became more substantial when silicate weathering was also eliminated. Likewise, eliminating calcium inputs from silicate weathering alone had no clear effect, but it had a substantial effect when both mineral sources were eliminated.

The geochemical model produced this pattern because carbonate and silicate weathering suppressed each other through their alkalizing effect on soil pH, so removing one calcium source increased release from the other. When silicates or carbonates are sufficiently abundant, soil calcium availability is high, conditions that appear to favor Mollisol formation and soil organic carbon accrual.

The model predicts that acid inputs from fertilizer and atmospheric deposition accelerated calcium release from carbonate weathering by 28% in Mollisols. Agricultural liming also massively increased calcium inputs to cropland soils.

Using modeled liming rates as the reference point, agriculture increased calcium inputs to Mollisol cropland soils by approximately fourfold relative to preindustrial levels. Using 1987 agricultural census liming data instead, the study estimates a fivefold increase for Mollisol cropland soils, assuming lime was 20% dolomite and 80% calcite.

In the left Mollisol pair, the recent gray liming distribution is much broader and taller than its preindustrial counterpart, while the yellow carbonate-weathering distributions shift more modestly. The same visual pattern appears on the right for non-Mollisols, where the recent liming estimate and uncertainty span are especially pronounced.

The results suggest that conservation practices increasing root inputs, increasing plant diversity, or reducing tillage may not be sufficient by themselves to preserve Mollisol soil organic carbon. Vegetation-focused strategies may need to be complemented with geochemical strategies that mimic the natural calcium cycle of these soils.

Reducing excess nitrogen can protect soil carbonates and preserve a calcium reservoir that likely helps protect soil organic carbon. Changing the cadence and quantity of lime applied to croplands could better simulate the natural weathering regime, but agricultural calcium budgets remain poorly quantified.

The study finds that calcium released by carbonate and silicate weathering is associated with organic carbon storage, especially in Mollisols, while agriculture has greatly increased calcium inputs. Soil-carbon conservation may therefore need both ecological and geochemical strategies.