Nitrogen input, 15N balance and mineral N dynamics in a rice–wheat rotation in southwest China

A field experiment and farm survey were conducted to test nitrogen (N) inputs, 15N-labelled fertilizer balance and mineral N dynamics of a rice–wheat rotation in southwest China. Total N input in one rice–wheat cycle averaged about 448 kg N ha−1, of which inorganic fertilizer accounted for 63% of the total. The effects of good N management strategies on N cycling were clear: an optimized N treatment with a 27% reduction in total N fertilizer input over the rotation decreased apparent N loss by 52% and increased production (sum of grain yield of rice and wheat) compared with farmers’ traditional practice. In the 15N-labelled fertilizer experiment, an optimized N treatment led to significantly lower 15N losses than farmers’ traditional practice; N loss mainly occurred in the rice growing season, which accounted for 82% and 67% of the total loss from the rotation in farmers’ fields and the optimized N treatment, respectively. After the wheat harvest, accumulated soil mineral N ranged from 42 to 115 kg ha−1 in farmers’ fields, of which the extractable soil NO3 −–N accounted for 63%. However, flooding soil for rice production significantly reduced accumulated mineral N after the wheat harvest: in the 15N experiment, farmers’ practice led to considerable accumulation of mineral N after the wheat harvest (125 kg ha−1), of which 69% was subsequently lost after 13 days of flooding. Results from this study indicate the importance of N management in the wheat-growing season, which affects N dynamics and N losses significantly in the following rice season. Integrated N management should be adopted for rice–wheat rotations in order to achieve a better N recovery efficiency and lower N loss.