The aim of this work was to compare extracting solutions (Mehlich 1, Mehlich 3, DTPA-TEA) for iron and manganese in soils of the Amazon Region. The correlations between Fe and Mn in soils and their contents in the rice plants were determined. The technique of diagnosis by subtractions was used, in a randomized block design with split plots. The soil types were Podzols, Yellow Podzolic, Red-Yellow Podzolic, Yellow Latosol, Humic Latosol, and Alluvial under eight treatments: control, complete with all micronutrients, and omitting one by one of the following micronutrients B, Cu, Fe, Mn, Mo and Zn. Micronutrients and lime were applied before the first crop. Plants were harvested at 58, 68 and 70 days after emergence. Considering all soils, Mehlich 3 provided the highest correlation coefficients with the content of micronutrients in the plants. The best coefficients of determination were observed with Mehlich 1 and 3 solutions with respect to available Fe in the Podzols, Alluvial and Red-Yellow Podzolic soils. The several extracting solutions were equally efficient for exchangeable Mn determination, and presented significant determination among each other.
soil testing; analytical methods; heavy metals; cultivation; rice
