IMIS | Flanders Marine Institute

Flanders Marine Institute

Platform for marine research


Publications | Institutes | Persons | Datasets | Projects | Maps
[ report an error in this record ]basket (0): add | show Printer-friendly version

Assessment of water quality using diatoms as bio-indicators in catchments of Lake Victoria, Kenya
Lung'ayia, H. B. O. (2002). Assessment of water quality using diatoms as bio-indicators in catchments of Lake Victoria, Kenya. PhD Thesis. VUB: Brussel. 202, annex (6) pp.

Thesis info:
    Free University of Brussels, Faculty of Science, Laboratory of General Botany and Nature management

Available in  Author 
Document type: Dissertation


Author  Top 
  • Lung'ayia, H. B. O.

    In this thesis a description is made of species composition and characteristics of epilithic diatoms in rivers Nyando, Kibos and Kisat draining into Lake Victoria (Kenya part). The water quality of the rivers is evaluated by examining the diatom data alone and in relation to environmental variables. The distributions of diatoms in surface waters of Lake Victoria were also examined in relation to environmental variables. The aim was to assess the potential of diatoms in determining the “ecological” water quality and in supporting management decisions and conservation strategies for these aquatic ecosystems. First, species composition, richness and diversity of the epilithic diatoms in rivers Nyando, Kibos and Kisat were investigated. 224 diatom taxa (218 species) belonging to 32 genera were recorded from the three rivers. Maximum species diversity was observed in less polluted river Kibos followed by Nyando with medium pollution levels and Kisat the most polluted had the lowest values of diversity. Species richness and diversity were significantly correlated with altitude, width of the river channel, depth, current velocity, volume of discharge and dissolved oxygen. Eutrophication, increase in ionic content and organic loading reduced diversity downstream where a few species tolerant to pollution, such as Nitzschia palea dominated the community especially in Kisat. Diatom diversity indices were found to be important in indicating changes in whole diatom assemblages in response to changes in water quality. Secondly, weighted means of known diatom ecological indicator values were used in determining the water quality in rivers Nyando, Kibos and Kisat. The ecological indicator values included Saprobity, Oxygen requirements, Trophic state, Nitrogen uptake metabolism, Moisture, pH and Salinity. Taxa with known ecological indicator values occurred consistently in high abundance in all stations sampled and throughout the sampling period. The ecological indicator values had strong correlations with the measured environmental variables, which they are known to reflect. Saprobity, Oxygen requirements, Trophic state and Nitrogen uptake metabolism tended to increase downstream showing increase in pollution in the same direction. The data from the indicator values was in agreement with the one of measured environmental variables in confirming that Kisat is more polluted than Nyando and Kibos. The diatom ecological indicator values used in this study were found to be suitable for assessing water quality in the three rivers. The diatom “Indice de polluo-sensibilite” (IPS) or pollution sensitivity index was evaluated and was found to give nearly the same information as the known diatom ecological indicator values. The results of the IPS showed that on average, weakly polluted to moderately polluted waters occurred in Kibos and moderate to heavily polluted waters occurred in Nyando. The whole of Kisat was heavily polluted and pollution levels were more acute downstream after the Kisumu industrial area. In the third investigation, the distribution of the epilithic diatoms was assessed in relation to environmental variables in rivers Nyando, Kibos and Kisat. Cluster analysis by Two-Way Indicator Species ANalysis separated the diatom community into two major groups comprising the less polluted waters of Nyando and Kibos together, from assemblages of the more polluted Kisat. The group of kibos and Nyando had Navicula exigua, N. schroeteri and Gyrosigma scalproides as indicator species. Subsequent splitting of the data resulted in assemblages also reflecting different water quality irrespective of the position of the sampling station. Ordination by Canonical Correspondence Analysis revealed that diatom species distributions were significantly influenced by overall effect of the measured environmental variables. Conductivity, alkalinity, turbidity, dissolved oxygen, silicate and altitude were identified as the main factors contributing to variation in diatom assemblages in the three rivers. Species reflecting various environmental gradients were identified. Finally, the distributions of diatoms in the surface waters of Lake Victoria were studied in relation to environmental variables. 101 taxa belonging to 29 genera were identified. Higher species richness and diversity occurred in the Nyanza gulf and bay areas when compared to the open lake. Conductivity and silicate were found to influence species richness, diversity and evenness. The diatom community was separated into two main groups comprising assemblages of the Nyanza Gulf and the ones from the open lake. These assemblages also reflected environmental gradients. The open lake was generally associated with higher abundance of Nitzschia acicularis that was also the indicator species for this group. Aulacoseira agassizii, Cyclotella meneghiniana, Nitzschia fonticola and Cyclostephanos dubius were indicator species for the Nyanza Gulf. Conductivity, alkalinity, dissolved oxygen and lake depth were identified as the main environmental variables that significantly explain variations in the diatom assemblages in Lake Victoria. j The results of this study are a contribution to knowledge on potential diversity and ecology of the present day diatoms in the Lake Victoria basin. Further, they provide evidence that diatoms can be useful indicators of water quality in the basin and that they can be employed in monitoring studies and for management purposes.

All data in IMIS is subject to the VLIZ privacy policy Top | Author