Abstract
Optimal rates of N fertilizer depend mainly on three factors, (1) the total N demand of crops, (2) the amount of available N from sources other than applied N, and (3) the efficiency of fertilization in order to increase plant available N. Predictability and variability of each of these factors affect the precision of N fertilizer recommendations.In a theoretical approach the change of [crop N + extractable inorganic soil N] (CSN) in N‐fertilized cereals can be ascribed to two components, firstly an N rate independent and secondly a strongly N rate dependent one. This approach fits well when applied to an actual data set of N field trials on sandy soils in Hessia, Germany.The N rate dependent change of CSN was interpreted as apparent recovery of applied N indicating the efficiency of fertilization. Apparent recovery rates of applied N varied considerably between field sites and ranged from 33 % to 96 %.The N rate independent change of CSN was dominated by N mineralization, especially during the grain filling period (range: 4 to 61 kg N ha−1), and was significantly (p 2 extraction; r = 0.51, EUF‐technique).The total N demand of the crops varied widely within each cereal species. For barley and rye the total N demand was closely correlated with the optimal grain yield.The results suggest that in order to improve N fertilizer recommendations the prediction of the total N demand as well as the prediction of the fertilizer efficiency is badly needed. The soil N mineralization appeared to be an important N source for cereals, which was independent from the applied N rate. Soil N tests, which provide an index of mineralizable soil N, merit therefore consideration for adoption into N fertilizer recommendations for cereals.