Classification of honeydew and blossom honeys by discriminant analysis
2006
Bogdanov, S. | Gfeller, M. (Agroscope Liebefeld-Posieux (ALP), Posieux (Switzerland))
The objective of the present study is to use discriminant analysis for the classification of blossom and honeydew honeys by parameters which can be determined easily in routine honey control. Following parameters were determined in 113 blossom and 34 honeydew honeys of Swiss origin: the sugars glucose, fructose, turanose, saccharose, nigerose, maltose, isomaltose, erlose and melezitosepHfree acidityelectrical conductivity and protein content. The honeys were classified by discriminant analysis using single and multiple variables. The best single variable for honey classification was melezitose with 96 % correct classification, while the other variables had a weaker discriminatory power. Two canonical variable functions proved especially powerful for discrimination between blossom and honeydew honeys. The first used following variables: sum of fructose and glucose content, ratio of fructose and glucose content, pH and electrical conductivity. This function classified correctly 98.2 % of the blossom and 81.2 % of of the honeydew honeys. The other function using in addition melezitose classified correctly 88.2 % of the honeydew and all blossom honeys. The results were successfully validated using 180 different Italian unifloral honeys, 84 Swiss honeys of honeydew origin and 242 blossom honeys of various botanical origin.
Mostrar más [+] Menos [-]Palabras clave de AGROVOC
Información bibliográfica
Este registro bibliográfico ha sido proporcionado por Agroscope Reckenholz-Tänikon Research Station