Multivariate statistical techniques for the assessment of seasonal variations in surface water quality of pasture ecosystems

• Published in: Environmental monitoring and assessment 2013-10, Vol.185 (10), p.8649-8658
• Main Authors: Ajorlo, Majid, Abdullah, Ramdzani B., Yusoff, Mohd Kamil, Halim, Ridzwan Abd, Hanif, Ahmad Husni Mohd, Willms, Walter D., Ebrahimian, Mahboubeh
• Format: Article
• Language: English
• Subjects: Agricultural pollution; Ammonia; Animal Husbandry - statistics & numerical data; Animal, plant and microbial ecology; Applied ecology; Assessments; Atmospheric Protection/Air Quality Control/Air Pollution; Biological and medical sciences; Biological Oxygen Demand Analysis; Chemical oxygen demand; Cluster Analysis; Conservation, protection and management of environment and wildlife; Datasets; Discriminant analysis; Dry season; E coli; Earth and Environmental Science; Ecology; Ecosystem; Ecosystems; Ecotoxicology; Environment; Environmental Management; Environmental Monitoring; Escherichia coli - growth & development; Factor analysis; Factor Analysis, Statistical; Feces; Fundamental and applied biological sciences. Psychology; Hydrogen-Ion Concentration; Malaysia; Monitoring/Environmental Analysis; Multivariate Analysis; Nitrates - analysis; Nitrogen; Oxygen - analysis; Pasture; Pastures; Pollutants; Principal Component Analysis; Principal components analysis; Rain; Rainy season; Rivers - chemistry; Rivers - microbiology; Sampling; Seasonal variations; Seasons; Statistical methods; Stream flow; Streams; Studies; Surface water; Temporal logic; Terrestrial ecosystems; Water Pollutants, Chemical - analysis; Water pollution; Water Pollution, Chemical - statistics & numerical data; Water Quality
• ISSN: 0167-6369; 1573-2959
• DOI: 10.1007/s10661-013-3201-8

This study investigates the applicability of multivariate statistical techniques including cluster analysis (CA), discriminant analysis (DA), and factor analysis (FA) for the assessment of seasonal variations in the surface water quality of tropical pastures. The study was carried out in the TPU catchment, Kuala Lumpur, Malaysia. The dataset consisted of 1-year monitoring of 14 parameters at six sampling sites. The CA yielded two groups of similarity between the sampling sites, i.e., less polluted (LP) and moderately polluted (MP) at temporal scale. Fecal coliform (FC), NO 3 , DO, and pH were significantly related to the stream grouping in the dry season, whereas NH 3 , BOD, Escherichia coli , and FC were significantly related to the stream grouping in the rainy season. The best predictors for distinguishing clusters in temporal scale were FC, NH 3 , and E. coli , respectively. FC, E. coli , and BOD with strong positive loadings were introduced as the first varifactors in the dry season which indicates the biological source of variability. EC with a strong positive loading and DO with a strong negative loading were introduced as the first varifactors in the rainy season, which represents the physiochemical source of variability. Multivariate statistical techniques were effective analytical techniques for classification and processing of large datasets of water quality and the identification of major sources of water pollution in tropical pastures.