Chemometric Interpretation on the Occurrence of Endocrine Disruptors in Source Water from Malaysia

• Main Authors: Hafizan, Juahir, Veerasingan Armugam, Santhi, Ananthy, Retnam
• Format: Article
• Language: English; English; English
• Published: 2015
• Subjects: Q Science (General)
• Online Access: http://eprints.unisza.edu.my/5882/1/FH02-ESERI-15-03445.jpg; http://eprints.unisza.edu.my/5882/2/FH02-ESERI-15-03557.jpg; http://eprints.unisza.edu.my/5882/3/FH02-FBIM-15-02656.jpg; http://eprints.unisza.edu.my/5882/

Multivariate statistical techniques comprising of hierarchical cluster analysis (HACA), discriminant analysis (DA), and principal component analysis (PCA) were used to evaluate and interpret a large dataset for the rivers in Selangor, Malaysia. The dataset consists of analytical results for three physical parameters and nine endocrine disrupting compounds from 25 sampling sites representing source water in the state. HACA successfully grouped the sampling sites to three major clusters (low, moderate, and high pollution) while DA identified eight of the most significant variables that contributed to the high spatial variation to afford 94% correct assignations. Moreover, the application of DA to confirm the seasonal classification to dry and rainy season afforded 88% correct assignations using three variables. In addition, DA offers insight on the inadequacies of the current classification of river water by the Department of Environment based on physical variables. Meanwhile, PCA identified eleven variables to point out 67% of spatial variability. While significant data reduction was not achieved using PCA, it allowed grouping of the variables according to common features which was useful in identifying sources of contamination. This study highlights the usefulness of chemometric techniques for the effective management of source water and identifying appropriate pollution control strategies.