Estimativa das incertezas para os teores dos nutrientes obtidos pelo procedimento Mehlich-1.

• Autor(a) principal: Rocha, Janilson Gonçalves da
• Data de Publicação: 2014
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: LOCUS Repositório Institucional da UFV
• Texto Completo: http://locus.ufv.br/handle/123456789/1666
• topic: Química do solo; Solos - Composição; Solos - Nutrientes; Incerteza; Mihlich; Robustez; Chemistry of soil; Soil - Composition; Soil - Nutrients; Uncertainty; Mihlich; Strength; CNPQ::CIENCIAS AGRARIAS::AGRONOMIA::CIENCIA DO SOLO

Uncertainty is a parameter pertinent to measurement data that characterize the random variability of an analytical result. Uncertainty can be a standard deviation (standard uncertainty) or a multiple thereof (extended uncertainty) or even half a confidence interval at the established level. The purpose was to determine the uncertainty of the availability of P, K, Cu, Mn, Fe, and Zn extractable by Mehlich-1. The study was initiated by using the measurement function of the cause-effect diagram to identify sources of uncertainties, based on the following input quantities of the measurement procedure: measured volumes of soil, extractor, aliquot and reading solution, and the dilution procedure and the calibration curve coefficients of the device. Uncertainties arising from the influence quantities, which do not participate directly in the measurement function but also affect the results, were determined. To this end, a study with a fractional factorial design was carried out, with the study factors: hydrochloric and sulfuric acid concentrations, stirring time, stirring speed, and settling time of the 22 analyzed soils. Regression equations were calculated for the estimation of the variability of the influence on the availability of P, Cu, Mn, Fe, and Zn, with significant effects by the t test (p < 0.05) of the sulfuric acid hydrochloric acid concentrations, stirring time, stirring speed, and settling time. For the availability of K, the same significant effects on the influence quantities of these nutrients were observed, except that the stirring time was not significant (p > 0.05) by the t test. The results showed that the sources that contribute most to the combined uncertainty of the available nutrients were those of the influence quantities followed by the calibration curve. Among the uncertainties of the influence quantities stirring speed was the most significant source for all nutrients, except for Mn, for which soil settling time was the greatest source. Therefore, to reduce the combined uncertainty in these nutrients, one can work with more accurate stirring systems (less uncertainty) or even work with calibration curves of devices at lower concentrations, with a lower uncertainty of this source.