Modeling the organic carbon dynamics in long-term fertilizer experiments of India using the Rothamsted carbon model

Abstract

Soil organic carbon (SOC) turnover simulation models have been widely used to predict SOC changes with changing climatic and management conditions. Rothamsted carbon model, RothC 26.3 is one of the most widely used C turnover simulation model, however, this model has not been extensively tested under the Indian conditions. Model prediction accuracy depends upon correct initialization procedure as wrong initialization can simulate the data incorrectly. We initialised the RothC model using the data set from three long-term fertilizer experiments (LTFE) of India. The model was parameterized for the pre cultivated soils of the three major soil orders (Vertisol, Alfisol and Inceptisol) of India for the treatments of nil fertilization (control), balanced fertilization (NPK) and NPK+ Farm Yard Manure (FYM). The RothC was successfully initialized (forward mode) by iteratively adjusting the C input and inert organic matter (IOM) stock of soil at steady state. The agreement between the modelled and measured data of SOC stocks across three different soil types was satisfactory, with root mean square error (RMSE) for LTFE treatments ranged from 2 to 14%. The results of the RothC simulations demonstrated that NPK and NPK+FYM increased SOC stocks at the 0–30 cm soil depth by 12–61 and 30–107%, respectively, over the initial value at the three sites. The SOC stocks reached steady state for the treatments of NPK and NPK+FYM between 48 and 95 and 68 to 116 years, respectively, in three different soil types. Therefore, the RothC model can successfully predict C dynamics under Indian conditions provided initialization and parametrization of the model is accurate.