one publication added to basket [17523] | Assessment of environmental quality using diatoms in Nairobi River, Kenya
Ndiritu, G.G. (2001). Assessment of environmental quality using diatoms in Nairobi River, Kenya. MSc Thesis. VUB: Brussel. 100 pp.
|
Available in | Author |
|
Document type: Dissertation
|
Abstract |
Environmental variables and diatom data were collected from Nairobi River, Kenya to study the response of diatom assemblages on natural and artificial substrates to water pollution. Fifteen stations were sampled, covering a distance of 60 km from the source. The river was experiencing agricultural, domestic and industrial pollution. Environmental variables data were described by quartiles and standard error; explored using PCA and cluster analysis; and analysed using spearman correlation, ANOVA and multiple regression. Species data was analysed using, biodiversity indices (Pielou 1975, Hill 1973). Multivariate analyses used included the two way indicator species analysis (TWINSPAN) and cluster analysis for classification; and canonical correspondence analysis (CCA) and detrended correspondence analysis (DCA) for ordination. Ecological indicator values were also used (van Dam 1994). The multivariate analysis revealed three types of gradients in both substrates. The first gradient represented the upstream stations, which corresponded to high oxygen. The second gradient related to high agricultural pollution and had high NO3. The third group corresponded to both industrial and domestic pollution and included stations situated in lower parts of the river. Those stations had high measurernents of total dissolved solids, total suspended solids and PO4. Species richness, diversity and evenness tended to be higher in the second group while high densities were recorded in the third group. Our results showed that the response of diatom assemblages to environmental change and substrate type was observed in species composition variation. Multivariate analyses and ecological indicator values revealed this response and were preferred to species diversity measures. |
|