ANÁLISE ESTATÍSTICA MULTIVARIADA APLICADA À ESPECTROMETRIA DE EMISSÃO ÓPTICA DE PLASMA ACOPLADO INDUTIVAMENTE PARA A AVALIAÇÃO DE MICRONUTRIENTES EM TOMATES DE CULTIVO ORGÂNICO E CONVENCIONAL

• Ano de defesa: 2021
• Autor(a) principal: Thiago Gomes Ricci
• Orientador(a): Carlos Eduardo Domingues Nazario
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Fundação Universidade Federal de Mato Grosso do Sul
• País: Brasil
• Palavras-chave em Português: micronutrientes; ICO OES; análise estatística multivariada; OPLS-DA
• Link de acesso: https://repositorio.ufms.br/handle/123456789/4282

This study was carried out through the analysis of micronutrients by inductively coupled plasma optical emission spectrometry (ICP OES) in organic and conventional cultivation tomatoes, which, multivariate statistical analysis was applied to interpret the data. Samples of cherry tomatoes (Solanum lycopersicum var. cerasiforme) and Italian tomatoes (Solanum lycopersicum ‘Roma’) from organic and conventional crops were studied. The samples were obtained from markets, local producers and organic product fairs, from Campo Grande, Mato Grosso do Sul, and surrounding areas, in a set of 14 samples, 7 samples from organic cultivation and 7 samples from conventional cultivation. The data generated through the analysis by ICP OES were obtained through samples of dried and ground tomatoes and mineralized by microwave-assisted wet-acid digestion. The elements Al (167.079 nm), Cd (228.802 nm), Co (228.616 nm), Cr (283.563 nm), Cu (324.754 nm), Fe (259.940 nm), Mn (257.610), Ni (221.647 nm), V (309,311 nm) and Zn (213,856 nm) were analyzed by ICP OES, in order to help obtain the chemical identity of tomato samples, and thus correlate them to their type of cultivation. The ranges of analytical work were determined for each element, and thus the parameters of analytical reliability were calculated, correlation coefficient (R), between 0.9989 and 0.9998, coefficient of determination (R²), between 0.9978 and 0.9996, limit of detection (LD) and limit of quantification (LQ) for each element. Precision values were determined through the percentage relative standard deviation (DPR%) and accuracy through the percentage relative error, calculated in three levels n=3, high, medium and low, for each element. Data interpretation was performed by calculating the concentration of each element in mg kg-1 of fresh samples, data that were subjected to multivariate statistical analysis to interpret the correlations between the chemical identity of the samples and the type of culture. Multivariate statistical analysis was performed using the online software MetaboAnalyst 5.0, graphical representations of the principal component analysis (PCA) were generated, with cumulative explained variance for the principal components (PC1 vs. PC2) of 82.1%. The graphical representation clustergram, originated from hierarchical cluster analysis (HCA) combined with heatmap, indicated the hierarchical clustering of samples and elements studied. Three hierarchical clusters were obtained, (I) samples from organic products market, with the greatest contribution of the elements Al, Fe, Zn and Mn, with the highest concentration values, (II) samples from conventional cultivation, with the greatest contribution of the elements Al, Fe, Zn and Mn, but with only lower concentration values and (III) sample of organic cultivation from local and market producers, with greater contribution of the elements Cr and Ni. Through orthogonal partial least squares discriminant analysis (OPLS-DA) it was possible to obtain a class prediction model with moderate level of predictive accuracy: values of explain variance on the Y axis (R2Y) and predictive performance of the model (Q2), 0.523 and 0.457, respectively. In the analysis of the OPLS-DA model, the elements Cr, Fe, Ni, Zn and Al, presented greater contributions to the prediction and differentiation of classes. In the study, it was observed that the correction process of the soil hydrogen potential (pH) plays a fundamental role in determining the availability of the studied micronutrients in the soil.