Método non-targeted aplicado em méis de abelha sem ferrão: combinação de análises físico-químicas e eletroanalítica com quimiometria

Abstract

The so-called native or indigenous bees are included in a class called Meliponini, which includes around 400 species in the Neotropical region. Due to the great diversity, geographical origin and type of management, the honey produced by these bees can vary in taste and aroma and, consequently, in its composition. The diversity of the aromas and flavors of this honey broadens its culinary uses, but makes it difficult to monitor and control its quality. Honeys from 10 species of stingless bees were investigated by means of pH, conductivity, moisture, cyclic voltammetry and differential pulse voltammetry analyses. For voltammetry analysis, a Cu/CuO electrode was constructed and voltammograms were obtained in 0.1 mol L-1 NaOH as the supporting electrolyte. The results obtained were part of the database that was used in the chemometric study using the principal component analysis (PCA), hierarchical cluster analysis (HCA) and multivariate regression analysis methods. PCA was used to retain the maximum information in terms of total variation contained in the data, with the least possible loss of information. HCA was used to discover natural groupings of variables from the observed data, grouping the samples based on species similarities. The combination of PCA, HCA and multivariate regression methods showed that factors such as bee species, collection site and collection period play a crucial role in determining the physicochemical and electrochemical parameters of honey. It was observed that the location and collection period were decisive in grouping the samples by their similarities. The collection period was a determining factor in the electrochemical profile of honey, significantly influencing its properties throughout the year. Jataí bee honey showed the least variations in the analyzed characteristics and even samples collected at different times were grouped by their similar characteristics.