The palynological and physicochemical analysis of 62 honey samples produced in different biogeographical areas of Algeria was conducted. Results showed high variety in the botanical origin of samples and their physicochemical profile. Twenty-six samples were polyfloral honey, 30 were unifloral honey from different botanical sources such as Eucalyptus, Citrus, Apiaceae, Punica, Erica, Rosmarinus, Eriobotrya, or Hedysarum, and 6 were characterized as honeydew honey. Pollen analysis allowed the identification of 104 pollen types belonging to 51 botanical families, whereas the physicochemical profile showed important variations between samples. Multivariate techniques were used to compare the characteristics of samples from different biogeographical areas, showing significant differences between humid-area samples, located in the northeast of the country, and samples taken in semiarid, subhumid, and arid zones. Principal-component analysis (PCA) extracted nine components explaining 72% of data variance, being 30%, the sum of Component 1 and Component 2. The plot of both components showed samples grouped upon botanical and geographical origin. The results of this paper highlighted the great variability in honey production of Algeria, evidencing the importance of honey characterization to guarantee authenticity and to valorize local production.
This study aimed to characterize the honeys of Babors Kabylia through sensory, melissopalynological and physico-chemical parameters. Thirty samples of honey produced in this region were collected over a period of two years and analyzed. All the samples presented physico-chemical parameters in conformity with legislation on honey quality, with few exceptions, linked mainly to beekeeping management. The pollen spectrum revealed a great diversity with 96 pollen types. The main pollen types were spontaneous species as Fabaceae (Hedysarum, Trifolium, Genisteae plants), Asteraceae plants, Ericaceae (Erica arborea L.) or Myrtus and Pistacia. The sensory properties of samples showed a high tendency to crystallization, the colors were from white to brown, but most of them had gold color. Smell and odor corresponded mainly to vegetal and fruity families and in taste perceptions besides sweetness highlighted sourness and saltiness notes. Seventeen samples were polyfloral, one was from honeydew and twelve were monofloral from heather, genista plants, sulla, blackberry or Asteraceae. Heather and the honeydew samples showed the darkest color, the highest electrical conductivity and phenol and flavonoid content. A statistical analysis based on the most representative pollen types, sensory properties and some physico-chemical components allowed the differentiation of honey samples in terms of botanical origin.
Honey color and other physicochemical characteristics depend mainly on the botanical and geographical origin. The study of these properties could make easier a correct classification of unifloral honey. This work determined the palynological characteristics and some physicochemical properties such as pH, electrical conductivity, and color (Pfund scale and the CIELa*b* coordinates), as well as the total content of the bioactive compounds phenols and flavonoids of ninety-three honey samples. Samples were classified as chestnut, blackberry, heather, eucalyptus, and honeydew honey. The study showed a close relationship between the physicochemical variables and the botanical origin. The five types of honey presented different physicochemical properties among them. A principal component analysis showed that Hue, lightness, b*, and Chroma variables were important for the honey types classification, followed by Erica pollen, pH, Cytisus, and Castanea variables. A forward stepwise regression analysis was performed introducing as dependent variables the color (mm Pfund) and the Chroma and the Hue variables. The regression models obtained explained 86%, 74%, and 86% of the variance of the data, respectively. The combination of the chromatic and physicochemical and pollen variables through the use of multivariable methods was optimal to characterize and group the honey samples studied.