Measuring and Modeling the Long‐Term Impact of Crop Management on Soil Carbon Sequestration in the Semiarid Canadian Prairies

Abstract

Agricultural management practices which promote soil organic carbon (SOC) sequestration can contribute to the long‐term productivity of soils, thus research must quantify and predict SOC dynamics in response to crop management. Using long‐term (1967–2009) data from 10 cropping systems on a Brown Chernozem (Aridic Haploboroll) in the Canadian semiarid prairies at Swift Current, Saskatchewan, we assessed the effect of fertilizer, cropping frequency, and crop type on SOC dynamics in the 0‐ to 15‐cm depth. Three models: Campbell, introductory carbon balance model (ICBM), and DayCent were evaluated, all of which produced fairly accurate predictions of SOC content and sequestration rates (R2 of 0.64–0.82); however, DayCent had the highest correlation and lowest errors of prediction and was deemed superior. Residue inputs of 0.87 to 1.13 Mg C ha−1 yr−1 maintained the SOC level, and SOC content was directly related to factors which increased C inputs. The SOC content and sequestration rates were lowest for wheat (Triticum aestivum L.)‐based rotations which were frequently fallowed and included flax (Linum usitatissimum L.), but highest for systems which were frequently cropped, well‐fertilized, and included rye (Secale cereale L.) or pulse crops in rotation. For systems with high C input, DayCent projected SOC gains of 12 Mg C ha−1 from 2009 to 2100, indicating that the soil at Swift Current had not reached maximum C capacity. This study was the first to rigorously test and demonstrate the strength of the DayCent for simulating SOC under different cropping systems on the Canadian prairies.