Principal Component Analysis (PCA) and Cluster Analysis of the Physicochemical Characteristics of Honey Application to Libyan Samples

Abstract

This study applies Principal Component Analysis (PCA) and Cluster Analysis to investigate the botanical and geographical classification of Libyan honey samples. The research examines physicochemical data from 28 honey samples collected across western Libya, representing six types: thyme, spring, sidr, tamarisk, harmal, and arugula. Ten parameters were measured per sample, including heavy metal concentrations (lead, zinc, cobalt, copper, iron) and quality indicators (reducing sugars, free acidity, total ash, electrical conductivity, absorbance). PCA results showed that the first three principal components captured 68.4% of total variance. PC1 (28.6%) was primarily associated with heavy metals and acidity, while PC2 (25.0%) linked to reducing sugars and electrical conductivity. Hierarchical cluster analysis using Ward's method grouped samples into three distinct clusters. Cluster 1 comprised spring, arugula, and tamarisk honeys, characterized by low ash content and moderate acidity. Cluster 2 exclusively included thyme honey, distinguished by notably high acidity and ash content. Cluster 3 contained sidr and harmal honeys, characterized by high reducing sugars and balanced remaining properties. The findings confirm the effectiveness of multivariate statistical techniques in discriminating and classifying honey samples according to their floral and geographical origins. This approach provides a robust methodological framework applicable to quality control and geographical origin authentication of honey. Additionally, the study contributes to the knowledge base on Libyan honey's physicochemical characteristics and opens avenues for future research incorporating broader environmental and climatic variables.