Multivariate Statistical Approach for the Discrimination of Honey Samples from Galicia (NW Spain) Using Physicochemical and Pollen Parameters.

Abstract

Raw honey is a food with a close relation to the territory in which it is produced because of factors such as soil conditions, weather patterns, and plant communities living in the area together. Furthermore, beekeeping management affects the properties of honey. Protected Geographical Indication Miel de Galicia protects the honey produced in Galicia (Northwest Spain). Various types of honeys (362 samples) from this geographical area were analyzed using chemometric techniques. Principal component analysis was favorable to analyzing the physicochemical and pollen variables with the greatest weight in the differentiation of honey. The linear discriminant analysis correctly classified 89.8% of the samples according to the botanical origin using main pollen spectra and physicochemical attributes (moisture, pH, electrical conductivity, diastase content, phenols, flavonoids, and color). Regarding unifloral honey, blackberry, eucalyptus, and heather honeys were correctly grouped, while five chestnut honeys and fourteen samples of honeydew honeys were misclassified. The chestnut and honeydew honeys have similar physicochemical properties and frequently similar pollen spectra profiles complicating the differentiation. Experimental evidence suggests the potential of multivariate statistics in the characterization of honey of the same geographical origin. Therefore, the classification results were good, with electrical conductivity, total phenol content, total flavonoid content and dominant pollens Eucalyptus, Erica, Rubus and Castanea sativa as the variables of higher importance in the differentiation of botanical origin of honeys.