Geographical discrimination of honeys by using mineral composition and common chemical quality parameters

Sixty honey samples from six different production zones of the provinces of Salamanca, Zamora and Cáceres (western Spain) were analysed for 13 common legal physicochemical parameters and 17 mineral elements (13 cations and four anions) in order to test for their geographical classification. Application of linear stepwise discriminant analysis to a number of variables made of a selection of analytical results and their simple mathematical functions allowed, firstly, discrimination between honeys from all six zones and, secondly, discrimination between honeys from the three zones of the province of Salamanca. The most discriminant variables selected for the six zones were 10 in number, a combination of three physicochemical parameters and nine elements, with a result of 91.38% of correctly classified samples; for the three zones of the province of Salamanca, regarded with special interest in our study, six variables were selected (made of eight elements) with a result of 97.07% of correctly classified samples. © 2000 Society of Chemical Industry