Identification and classification of pure and adulterated honeys by multivariate analysis

Honey is a natural food with the high nutritional value and the most useful medicinal properties. The chief components of honey are simple carbohydrates and water, leading to the fraud in honey using cheaper commercially available sugar syrups. This is an important anxiety in quality control of honey. In this study, detection of pure natural honey from the honey adulterated by adding concentrations of 7, 15 and 30 percent of both beet invert sugar and date invert sugar was separately examined. For this purpose, multivariate analysis of physicochemical properties of honey samples were used. Applying principal component analysis (PCA), among 12 physicochemical properties measured, nine parameters were identified as principal components. Then the linear discriminant analysis (LDA) was utilized to discriminate the pure and adulterated honeys in three different batches according to these nine parameters. An optimum classification of 91.15% was reached using three physicochemical factors including pH, free acidity and ash. Correct classification of 85.84% were achieved by three others physicochemical properties, i.e. moisture content, electrical conductivity and reducing sugar, and of 82.30% by HMF, total sugar and Fructose/Glucose ratio. Results demonstrated that multivariate analysis in combination of the routine analysis of honey could be applied to rapid and easy detect of adulteration in honey.