Development of a new multinutrient extraction method for macro- and micro-nutrients in arable land soil

A new multinutrient extraction method was developed for simultaneous extrac tion of macro- and micro-nutrients in arable land soil. Composition and con centration of the new multinutrient extractant were 0.2 mol L-1 CH3COOH, 0. 25 mol L-1 NH4Cl, 0.005 mol L-1 C6N8O7 (citric acid), and 0.05 mol L-1 HCl (pH 1.3). The extraction method consisted of continuous shaking for 30 min with a ratio of air-dried soil to extractant of 1 : 10 (w/v). For boron (B) extraction, the ratio of soil to extractant was 1 :2 (w/v), and then the e xtract was filtered. Correlation coefficients between the values of the nut rients extracted with the newly developed method and the values of those ex tracted with the conventional methods-i.e., NO3-N extracted with KCl, avail able P extracted with the Truog method, exchangeable K, Ca, and Mg extracte d with the Semimicro-Schollenberger method, easily reducible Mn extracted w ith hydroquinone-ammonium acetate, soluble Zn and Cu extracted with HCl, an d available B extracted with hot water-were equal to or higher than 0.91*** , respectively. The correlation coefficient for the soil samples, where ava ilable P values were below 400 mg P2O5 kg(-1) (Truog method), was as low as 0.783***. In the case of paddy soil samples, however, the correlation coef ficients were as high as 0.836***, and, in the case of upland soil samples, 0.878**. Values of Ca and Mg extracted with the newly developed method wer e higher than those of Ca and Mg extracted with the Semimicro-Schollenberge r method in one or two soil samples. They may have contained a large amount of Ca and Mg carbonates, because base-saturation percentages exceeded 100. For HCl-soluble Cu, the correlation coefficients were as high as 0.983*** for volcanic ash soil samples, and 0.990*** for nonvolcanic ash soil sample s. For available B, multiple correlation coefficient; (adjusted by degrees of freedom) increased from 0.925*** to 0.962***, when soil pH(H2O) was incl uded in a multiple regression equation between the values of a extracted wi th the newly developed method and those of B extracted with the hot water m ethod.