Indexing

Abstract:The aim of this work was to test the suitability of principal component analysis and cluster analysis for distinguishing 27 unifloral and multifloral honeys from Eastern Mau using melissopalynological parameters: Pollen density, pollen types, season and site of collection, honey type and Shannon Weaver diversity index of the pollen types. The extraction of six variables were well represented in the common factor space. Six principal components have their communalities explained. Principal component 1 explains upto 44.07% of the variance and has the highest Eigen value 2.64. Component 1 and 2 had Eigenvalues above 1.00 and both cumulatively accounted for 67.67% of the total variance. Pollen density, Season, pollen types, honey types are more correlated to the principal component 1, while sites and pollen Shannon Weaver diversity index were more correlated to principal component 2. Four of the six clusters created in cluster analysis had mixture of honey samples from various regions  and Botanical origin. All samples collected from Mariashoni in April formed an isolated single cluster, all the samples (MA-S1-AP, MA-S2-AP,MA-S3-AP) being unifloral honey. Honey samples KA-S2-DE, KA-S3-DE, NE-S1-DE, NE-S2-DE , all collected in December formed a cluster of honeys from adjacent mesoregions of Nessuit and Kapkembu.
Keywords:Cluster analysis, Eastern Mau, Honey, Principal component analysis, Honey, Pollen.
Title:Authentication of honey samples from Eastern Mau forest Kenya by principal component analysis and cluster analysis of quantitative melissopalynological variables
Author:Philip O. Onyango
ISSN 2349-7823
International Journal of Recent Research in Life Sciences (IJRRLS)
Paper Publications