Testing of the applicability of European diatom indices in the tropical rift valley lake, Lake Hawassa, in Ethiopia

Abstract

Quantitative ecological monitoring of African lakes is needed to understand growing human pressures on ecosystems. Diatom-based indices are routinely used for this purpose elsewhere in the world, but have not yet been produced for the flora of African freshwater lakes. Here we tested the applicability of the European diatom indices on the biomonitoring system of Lake Hawassa, Ethiopia. Physico-chemical and benthic diatom sampling was done at nine sites with different degrees of human disturbance along the lakeshore area from February to November 2015 and 2016. A percentage disturbance score (PDS) was calculated at each site and categorised from no evident disturbance (0–25%) to high disturbance (75–100%). Based on this criterion and selected physicochemical parameters, the sampling sites categorized into minimal, moderate and high disturbance. Seventeen diatom indices were calculated using Omnidia software version 5.3. Out of a total of 17 indices that were calculated using the Omnidia software, six were selected as potential metrics. The diatom indices had a high discrimination efficiency and were significantly correlated with most the environmental parameters (r > 0.6; p < 0.05). Among these, the trophic diatom index (TDI) and generic diatom index (IDG) showed the best potential to discriminate the three clustered sites, based on their ecological classification. Accordingly, although robust locally based indices are needed, the TDI and IDG diatom indices could be used in monitoring of water quality in tropical African rift lakes.

Keywords: biomonitoring, discrimination efficiency, human disturbance, metrics selection, OMNIDIA