Methodology of nitrogen rate determination and its practical utilization

Nitrogen fertilization problems are corresponding with fertilization economics as well as with environment conservation. Mentioned causes lead to elaboration of various methodological approaches to nitrogen fertilizer rate determination. For nitrogen rate calculations besides total nitrogen consumption with planed yield, as a rule are recommended data on mineral or easily mineralizable nitrogen in soil and nitrogen content data in overhead plant mass (or plant organs). N min-method use in practice is connected with the problem of relatively high number of soil samples and analyses in relatively short time. Disadvantage of plant analyses is the fact they do not present variability of soil available nitrogen and its uptake conditions in various pedo-ecological conditions. These factors taking in account at the results of plant analyses is an inevitable condition for plant analyses practical use. Ideal combination for nitrogen rate determination in vegetation period is fusion of soil and plant analyses. Mathematical models are usually used at study and verification various processes and mutual interactions of various factors within given system. Several models in effort to attain some phenomenon complex characterization become oversophisticated with direct impact on approximation to reality. The possibility to calculate input data, instead of setting concrete results of measurement is introducing further element of inaccuracy. The model formation is based very often on existing submodels (growth submodel, nitrogen leaching submodel, etc.) that are incorporated. Mechanistic approach at a model construction and some phenomena description brings considerable inacuracy into the calculations. The submodels (e.g. for nitrogen leaching, organic nitrogen mineralization, etc.) are more simple, and they are included into some methods of nitrogen rates determination (e.g. balance methods). Objective shortage in a simulation model accuracy is (and probably will be) the problem of weather characteristics prediction. Any model, usable for the purposes of subsequent N-fertilization need calculation, must be simple, with relative small number of really obtainable input data