Cereal Response to Nitrogen

This study examines various mathematical formulations of cereal response to nitrogen. Questions about the curvature of the response function and its shape after yield maximum have occupied scientists for decades. Available agronomic knowledge combined with statistical fit to field trial data were criteria for comparing and ranking ten different functional forms. A quadratic response function with a maximum yield plateau was judged to be the best performing model overall, although several other models seem worthy of further considerations. 

Another goal of the study was to estimate the extent of nitrogen mineralization from soil resources. Estimation was based on the functions' nitrogen axis intercept assuming a zero yield without nitrogen and that the fitted functions were representative outside the observable fertilizer domain. Available climate data could not satisfactorily explain mineralization variation among trial sites. Economic evaluation showed that incon-ect choice of response model entails significant costs, especially if the choice is between linear and non-linear specifications. Insurance against unpredictable mineralization variation in the form of a permanent over-fertilization seems profitable only if a linear response specification were correct. 

Mortensen 2000, pt one.pdf (4.06 MB)

Mortensen 2000, pt two.pdf (2.86 MB)

Author(s)

Publication Date