Contribution of soil inorganic carbon to atmospheric CO₂: More important than previously thought

CO2 production from soil inorganic carbon (SIC) by neutralization of nitrogen (N) fertilization-induced acidity is globally relevant. Here we analyzed factors that may affect CO2 production from SIC after N fertilization: (1) buffering capacity of soil organic matter (SOM) and of clays, (2) increasing crop growth and C input belowground by N fertilization, (3) acidity localization at the fertilization point, (4) SIC localization in the sub-soil, (5) application of CaO and basic slag instead of lime, (6) inability of farmers in low income countries to apply lime. We conclude that our previous estimation of CO2 fluxes from carbonates by N fertilization (7.5 × 1012 g C year-1) and from liming of acidic soils (273 × 1012 g C year-1) is possibly an underestimation and consequently, the contribution of SIC to atmospheric CO2 is more important than previously thought. CO2 production from soil inorganic carbon (SIC) by neutralization of nitrogen (N) fertilization-induced acidity is globally relevant, but up to now was not considered as a CO2 source. We analyzed in detail the seven factors mentioned by Datta and Mandal (2018) to hone our estimation of CO2 efflux from SIC (Zamanian, Zarebanadkouki, & Kuzyakov, 2018). SOM and clays buffer the acidity only in soils with pH below 6.5—the soils without carbonates. The increased belowground C input by N fertilization can only be of very minor importance because the portion of belowground allocated C decreases with N availability (Figure 1) and the soil organic carbon (SOC) in most soils is under steady state. Consequently, the increase of root biomass by N fertilization is less than that of the shoots, because of decreasing root/shoot ratio. So, despite root C contributes 2.5–3.0 times more to the SOM formation (Rasse, Rumpel, & Dignac, 2005), the relative increase of root biomass is of minor importance. Further, the root/shoot ratio will decrease in future because the breeding is mainly aimed on the grain yield. The initial acidity induced by fertilization is limited to the point at which fertilizers are applied, but the produced H+ ions migrate easily to the CaCO3 depth and are neutralized with CO2 production. Furthermore, erosion (common in croplands even in semi-arid climates) places the sub-soil in contact to the atmosphere and thus to the applied fertilizers. When SIC is present, its amount is endlessly large compared to the applied N fertilizers (Figure 2). CaO and Ca2+-containing slag are produced by heating CaCO3 and metal ores, respectively, at temperatures exceeding 800°C. The production of CaO and basic slag leads to CO2 emission from CaCO3, but in a sector other than agriculture. The area amended with CaO and basic slag is minimal compared to the liming area. N fertilization in low-income countries is very low (see the N fertilization map: Figure 1 in Zamanian et al., 2018) and, accordingly, less acidification is expected. The annual release of 3.2–7.9 1012 g C from SIC as estimated by Perrin, Probst and Probst (2008) is based on a field trial and direct measurement of NO3− and HCO3− concentrations in surface water. They clearly showed the correspondence of alkalinity neutralization in surface water (not soil) with increasing nitrate concentrations. So far, our estimation based on the CaCO3 distribution in soils and N fertilization is the first approach for a global estimation of CO2 from SIC. Using more accurate and detailed maps accompanying local measurements will no doubt improve the estimation. Nonetheless, higher CO2 fluxes are expected because over-fertilization always takes place and N use efficiencies at local scales are much lower than the assumed averages. The wetting/drying of paddy soils, fertilizers other than urea (e.g., (NH4)2SO4, Ca(H2PO4)2), and mineralization of animal dung in pastures should also be included. Furthermore, not only N fertilization, but also global N pollution leads to acidification (Averill, Dietze, & Bhatnagar, 2018) and unaccounted CO2 efflux from CaCO3. We conclude that, considering the factors mentioned by Datta and Mandal (2018), our estimation (Zamanian et al., 2018) of CO2 fluxes from carbonates by N fertilization (7.5 × 1012 g C year−1) and from liming of acidic soils (273 × 1012 g C year−1) as accurate as it can be done to date is possibly an underestimation. We acknowledge German Research Foundation (DFG) for their support (KU 1184/34-1).