عنوان مقاله [English]
Plant appropriate nutrition is one of the most important factors in improving the quality and quantity of product. Providing adequate and balanced of plant nutrients is essential for achieving maximum potential production. Potassium is one of the essential elements to plant’s growth. Therefore, knowledge of soil potassium status is so important in order to optimum usage of potassium fertilizers. There are several methods to measure soil K+. Ammonium acetate method to extracting the plant usable potassium is not efficient in all kinds of soils and high cost of sodium tetra-phenyl boron is also the negative point of this method for extracting numerous soil samples.
Considering this matter that Extraction method with ammonium acetate in Loess soils with high specific surface area in Golestan Province is less accurate compared to the method of sodium tetra-phenyl boron extraction; and on the other hand sodium tetra-phenyl boron extractio is time-consuming and costly; We have to select a different method to estimate K which is more accurate and time and costs are lower. The aim of current study was to determine the feasibility of a low-cost index as potassium excess in determination of soil available potassium in some rain-fed wheat in Golestan province with limited usability of ammonium acetate. Forecast wheat yield with using artificial neural network in a farm limited unit in order to achieve the ultimate goal, the application of potassium excess doneto estimate fertilizer requirements. This study was done in a piece of land (922m2) in Gorgan University of Agricultural Sciences and Natural Resources Farm one. Plots were divided into 40 plots and wheat (var., line17) seeds were drilled. During one stage (before planting) soil samples from each of 40 plots harvested and transferred to the laboratory for further analyzes. At the time of harvest, plants were harvested each plot separately. Potassium was measured by three methods of ammonium acetate, sodium overload and sodium tetra-phenyl boron. Artificial neural network model was used to estimate the predicated yield. In this model output yield and potassium intake was measured by three methods mentioned. According to the results, the correlation between yield and the three extractor ammonium acetate, sodium tetra-phenyl boron and potassium excess, are 0.62, 0.78 and 0.77 respectively. Accordingly, the extraction method with an overload of potassium, has higher correlation with grain yield, than ammonium acetate extraction method, and has close results with extractor sodium tetra-phenyl boron. The studied soil extraction method with an overload of potassium compared to the current (in the region) by ammonium acetate extraction has greater accuracy and efficiency. As a result of using less fertilizer and potassium excess not only increase the yield per unit area will be sought, but also will reduces environmental pollution due to excessive use of chemical fertilizers.