Temporal Dynamics of Soil Organic Carbon Contents in Long‐Term Monitored Arable Sites in Lower Saxony, Germany

ABSTRACT Background Temporal changes in soil organic carbon (SOC) contents in agricultural soils are variable, and underlying drivers are not sufficiently understood. Aim The objectives were to describe the long‐term dynamics of SOC contents at 41 sites in Lower Saxony, Germany, and to analyze the relationships between estimated long‐term changes of SOC contents and climate, soil, and agricultural management data. Methods Mixed‐effects regression models with observation weights were calculated for each site to estimate long‐term SOC changes (ΔSOC), and the long‐term ΔSOC values were then described using climate, soil, and agricultural management variables. Results The fitted regression lines for the temporal trajectories of SOC dynamics fell within the 95% confidence intervals of the annual mean SOC values in 96% of the observations, consistent with the expected coverage under a well‐fitting model. Multiple linear regression modeling for the estimated long‐term ΔSOC values explained the estimated ΔSOC variation satisfactorily (coefficient of determination R 2 = 0.77) and identified four variables with significant contributions: the initial SOC content with a negative regression coefficient and the initial pH, initial proportion of years with catch crops, and the clay content with positive regression coefficients. Conclusions The results suggest that initial soil properties (SOC content and pH) and agricultural management, as well as the clay content may explain temporal changes in SOC contents to a large extent.