An investigation of the relationship between soil mineral nitrogen in the autumn or spring and optimum nitrogen rate for winter cereals

Data from 143 nitrogen (N) response trials on winter cereals carried out at sites across England during the period 1981-1988 were reviewed. Linear regression models for the relationship between optimum N requirements and soil mineral nitrogen (SMN) were evaluated. Multiple regression analyses and F tests were used to determine whether separate regression lines for different combinations of soil group and N Index, substituting for cropping history, were justified, and whether the additional terms used to account for SMN in the 30-60 cm and 60-90 cm depth bands significantly improved the regression model, used to fit the data. The results indicated that optimum N requirement can be predicted from SMN 0-30 cm and 30-60 cm data. However, although the slope of the regression line does not change for different combinations of soil group and N Index, the intercept does. If no allowance is made for differences between sites which alter this intercept, the amount of variation in optimum N requirement accounted for by the regression models is considerably reduced. Factors which should be considered are mineralization of previous crop residues, organic manures and indigenous organic N, and the depth to which SMN is accessible by plant roots.