RESOLUTION OF THE SPATIAL VARIABILITY IN SEDIMENT COMPOSITION WITHIN AND BETWEEN WATER-STORAGE RESERVOIRS USING NONPARAMETRIC STATISTICAL TECHNIQUES

Non-parametric statistical techniques, specifically Multi-Dimensional Scaling (MDS), Analysis of Similarity (ANOSIM) and Similarity Percenta ges (SIMPER) are used to resolve the spatial variability of sediment c omposition based on major element distribution within and between two reservoirs in adjacent catchments which were to be used as the experim ental and control sites in a de-stratification experiment. MDS analysi s indicates that sediments separate into four groups, three from one d am and one from the second dam. ANOSIM analysis showed that there was a statistically significant difference between these four groups. SIMP ER analysis shows that there is a high, degree of similarity for sites within each of the four groups. In all cases, greater than 60% of the similarity within each group was explained by three elements--Si, Al and Fe. However, the elements responsible for the dissimilarity betwee n groups of sites were more variable. (C) 1998 Elsevier Science Ltd. A ll rights reserved.