Some aspects of the ecology of the Groot Letaba River in the Northern Province,South Africa

The aim of this pilot study was to evaluate the current ecological status of the Groot Letaba River and to compare this information with historical data. The objective was to determine the effects of various impacts on the fish populations of the river. This was done by analysing the water quality and by considering the effect of weirs and dams, as well as various illegal angling activities, on the fish community.

The Groot Letaba River is not highly polluted and the decline in its flow seems to be the greatest threat to the system. During a preliminary study to develop the river's resource potential, it was stated that the annual water allocation from Tzaneen Dam was 103.9 million m³/annum for irrigation, 8.4 million m³/annum for domestic and industrial use and 14.7 million m³/annum for environmental purposes. However, the yield from Tzaneen Dam was only 98 million m³/annum, suggesting that more water had been allocated than was available. As a result only 20% of the simulated natural flow is observed at Letaba Ranch Weir at the lower end of the river.

Over the past few years many weirs and dams, none of which have fishways, have been constructed in the Groot Letaba River, impacting on the flow regime and on the migration potential of many fish species. Tiger fish (Hydrocynus vittatus) and the largescale yellowfish (Barbus marequensis) are two of the more prominent species influenced negatively by these barriers. This problem is aggravated by the illegal netting of fish stranded below these barriers during their spawning migrations.     

Reproductive biology of a riverine cyprinid,Labeo umbratus (Teleostei: Cyprinidae), in small South African reservoirs

The reproductive and recruitment characteristics of moggel, Labeo umbratus, populations were examined in four small South African reservoirs. Reproduction, characterised by an extended spawning season, high fecundity, short incubation time and rapid larval development, appears to be ideally suited to the highly variable environment of small reservoirs. Evidence suggested that L. umbratus spawns in the reservoirs. In two reservoirs where samples were conducted monthly, GSI (gonado-somatic index) was positively correlated with both water temperature and day length, whilst the CPUE (catch per unit effort) of juveniles was not related to any environmental variable. The success of moggel spawning appeared to increase when there was early spring and consistent summer rainfall.     

Bioaccumulation of selected inorganic substances in the tissue of Oreochromis shiranus from Bunda Dam,Malawi

A study to assess and monitor the concentrations of sulphate, sodium, magnesium, calcium, chloride, iron and nitrate in Bunda Dam was carried out from May to November 1998. The same parameters, except nitrate and chloride, were also determined in the liver, gills, intestines and muscle of the fish Oreochromis shiranus from the same dam. Results showed that whilst the concentrations of all parameters in general were higher in September and October in the water, there were variations from month to month in the fish. Comparison of the overall concentrations in the water and fish showed that, for each parameter, there were higher concentrations (P<0.05) in the water than in the fish. Iron and magnesium were more concentrated in the liver while sulphate, calcium and sodium were more concentrated in the intestine. The concentrations of the parameters in both the water and the fish were below the hazard limits. The results suggest that the intensive farming activities taking place around the dam and the presence of the sewage pond nearby have no significant effect on the levels of chemicals in the dam water or its fish.     

Potential environmental impacts associated with the proposed abstraction of water from the Okavango River in Namibia

The Namibian Department of Water Affairs has in the past faced considerable pressure to relieve the water shortages caused by recent droughts. One of the options considered in 1996, following poor runoff during the 1994/95 and 1995/96 seasons, was a proposal to abstract some 17Mm³ of water per year from the Okavango River at Rundu, and transfer this via a 260km long pipeline to the head of the Eastern National Water Carrier at the town of Grootfontein. Part of the overall evaluation of this scheme included an assessment of the potential environmental impacts that could arise. An initial environmental evaluation was conducted from a point approximately 40 kilometres upstream of Rundu in Namibia, to the distal end of the Okavango Delta at Maun in Botswana.

Hydrological studies showed that the proposed abstraction represented a reduction of approximately 0.32% in the mean annual flow of the Okavango River at Rundu. The abstraction represents 0.17% of the mean annual flow at Mukwe, downstream of the Cuito River confluence. The adverse effects of the proposed water abstraction scheme would be extremely small along the Okavango River in Namibia, whilst outflows from the lower end of the Okavango Delta to the Thamalakane River would be reduced by some 1.44Mm³/year (11%). Additional studies showed that these effects could be reduced by some 10–13% if abstraction was confined to the falling limb of the hydrograph.

Hydrological simulations have shown that the maximum likely loss of inundated area in the Okavango Delta would total approximately 7km² out of some 8 000km². This potential loss in inundated area would be concentrated in the lower reaches of the seasonal swamps and seasonally inundated grasslands, specifically in the lower reaches of the Boro, Gomoti, Santantadibe and Thaoge channels. However, these effects would most likely be expressed as a shoreline effect, with the loss in area spread out over the shoreline and periphery of islands and would not be restricted to a single specific area. This anticipated loss in inundated area is unlikely to have measurable impacts on environmental components.

Overall, the study found no 'fatal flaws' which would prevent the water abstraction scheme from proceeding and the anticipated effects on the Okavango system are more likely to be seen in the Okavango Delta, rather than along the Okavango River. The anticipated ecological implications of the scheme were small in spatial extent and are unlikely to be perceptible against the natural year-to-year variability in inundation of the Okavango Delta or outflows to the Thamalakane River. However, the public perceptions of the proposed water transfer project were strongly negative and appeared to be at least in part due to the very low water levels in the Okavango River and Okavango Delta during the past three years and during the study period. These low water levels were amongst the lowest on record and it is likely that the public would attribute any adverse effect recorded in the future to the abstraction scheme, whether this were true or not. These negative perceptions of the desirability and acceptability of the proposed scheme were strongly linked to potential adverse affects on the tourism industry along the Okavango River and in the Okavango Delta, with possible adverse economic effects on local residents.     

A comparison of several methods of assessing river condition using benthic macroinvertebrate assemblages

This paper attempts to address some of the problems that are faced in the endeavour to minimise the time and financial costs of environmental monitoring. At the same time, it aims to assess the usefulness to freshwater studies of some of the advances being made in the marine sciences. The paper examines the results obtained and the conclusions drawn using several recognised techniques for measuring aquatic macroinvertebrate communities. The physical, chemical and macroinvertebrate data used were collected upstream and downstream of a land-based trout farm situated on a tributary of the Elands River in the Molenaars River catchment, south-western Cape, South Africa. Five measures of community stress were examined to determine the extent to which they summarised community changes in response to the trout farm effluent. They ranged from those requiring quantitative sampling and species-level identification to those requiring only qualitative sampling and an ability to distinguish between taxa. The results obtained suggest that: the incorporation of an abundance measurement can greatly enhance the sensitivity of an index; combining the results from different microhabitats can artificially inflate the scores obtained at a site; change in community structure following a perturbation may be more marked at higher taxonomic levels, because the community response is more evident above the natural environmental noise; identification to resolution lower than family (e.g. order) may result in important information being missed, and increase the chances of the effects of a perturbation going unnoticed. It is concluded that quantitative or semi-quantitative family-level data probably represent the best compromise between the costs of collecting and processing the samples, and the information content of the resultant data, but that the use of several different techniques would enhance the reliability of the conclusions drawn. In addition, it is recommended that some of the advances being made in the marine sciences, such as the 'marine' ranked Abundance Biomass Curves (ABCs) should be developed and implemented, since they can provide a much needed functional underpinning for results obtained.