Carbon storage in an arable soil combining field measurements, aggregate turnover modeling and climate scenarios.

[Display omitted] • The CAST model well simulated the effect of fertilization practices on carbon dynamics in a chronosequence; • Free microaggregates in the Fluvaquent were the primary C sequestration fraction under RCP2.6 scenarios; • MNPK treatment promotes C storage in the study soil under current and future climates. Quantifying soil structural dynamics and aggregate turnover is important in understanding soil organic carbon (SOC) stocks, particularly over decadal and larger time scales. Until now it has remained unclear clear how soil aggregate size and its associated carbon respond to both long-term soil fertility and climate change. Here, we explore changes in soil structure and aggregate organic C (OC) stocks under different fertilization practices by combining field chronosequence SOC measurements with dynamic and process modeling in a long-term wheat-maize field experiment on the North China Plain. The fertilization practices comprise no fertilization (CK), chemical fertilization (NPK), and combined manure and NPK treatments (MNPK). The experimental measurements included the mass of OC stocks in different soil aggregate size classes. We used this information to calibrate parameters of the Carbon, Aggregation, and Structure Turnover (CAST) model and to predict future changes in aggregate structure and the resulting OC stocks using the RCP2.6 scenarios that were defined by the outputs of five future climate models from IPCC projection. With trends towards a wetter climate and increasing soil moisture under the RCP2.6 scenarios for the region, soil OC stocks will increase in all three treatments, with the strongest increase under MNPK due to exogenous C inputs. The CAST model output further suggests that changes in microaggregate (250–53 μm) OC stocks in the NPK and MNPK treatments accounted for 78.6 % and 75.3 % of the calculated change in total SOC stocks between the early and late 21st century. In conclusion, our combined data and modeling approach describes changes in soil aggregate C, identifies the primary soil aggregate size class of microaggregates involved in C sequestration in an agricultural soil, and predicts the role of Fluvaquent soils on the North China Plain as a future C sink. [ABSTRACT FROM AUTHOR]