1. INTRODUCTION
This appendix has been included to illustrate the use of the models listed in the text in a real preimpoundment situation.
The proposed Mupata Gorge dam is on the Zambezi river, on the boundary of Zimbabwe and Zambia, downstream from Lake Kariba and upstream of Lake Cahora Bassa. Most of its water will come from the Zambezi via Kariba but about 30 percent would come from the Kafue. The lake will have many characteristics in common with Kariba and Cahora Bassa, but it would differ from the former by being very shallow and from both by having very small lake level fluctuations. The proposals to build this dam were highly controversial as the area to be flooded includes a fine wildlife area. Environmental and economic impact studies were made which attempted to quantify the effects of the dam (Du Toit, 1982; Stanning, 1982).
2. AVAILABLE DATA
Information about the dam is taken from Du Toit (1982), except for the shoreline length which was measured from a 1 : 250 000 map.
2.1 Physical Data
Altitude: 381 m
Latitude: 16°S
Shoreline length (Lo): c. 750 km
Surface area (Ao): 1 230 km2
Mean depth (z): 16 m
Volume (V): 19 800 m3 × 106
Catchment area (Ac): 900 000 km2
Mean annual flow, Zambezi river: 58.1 km3
Hydraulic retention time (Tw): 0.34 years
Calculated from the above:
Altitude factor: 1165
Shoreline development (DL): 6.03
2.2 Chemical Data
Conductivity (μS cm-1): 88
PO4 (mg l-1): 0.017
Calculated P. loading: 0.80 gP m-2yr-1
3. PREDICTIONS BASED ON AVAILABLE DATA
3.1 General Lake Characteristics
(a) Temperature
| Tmax | = 29.3° | (1) |
| Tmin | = 21.7° | (2) |
| T-m | = 25.5° | (3) |
These temperatures are reasonably close to those of Kariba and Cahora Bassa and this is what would be expected for the surface waters of Lake Mupata.
(b) Conductivity
| (Ac/V) × Tw = 15.5 | ||
| (i) | KL = 117 μS cm-1 | (4) |
| (ii) | KL = 152 μS cm-1 | (5) |
The first conductivity prediction seems most accurate since Kariba's conductivity = 80 μS cm-1 whilst Cahora Bassa = c. 100 μS cm-1. For this exercise, therefore, the prediction based on model (4) will be used. One possible reason for the high figure in (5) is that only two conductivity readings from the Zambezi were available; this is a major reason why models based on morphometric parameters may be more successful.
| TDS = 84 mg l-1 | (6) |
(c) Trophic status
Phosphorus load = 0.8 gP m-2yr-1 and z/Tw = 47.06, so according to the Vollenweider model that lake is likely to be oligotrophic. This is similar to Kariba.
(d) Lake phosphorus content
| P = 0.011 mg l-1 | (8) |
This again is fairly close to Kariba's phosphorus levels and seems to be a reasonably good prediction.
(e) Biological characteristics
With an input phosphate concentration = 17 mg m-3 Westlake's model suggests that:
Macrophyte penetration would be about 2.75 m. This is less than in Kariba where macrophytes penetrate to 12 m, although the bulk occur between 2 – 6 m (Kenmuir, 1983). This model probably indicates the region of greatest macrophyte density.
No phytoplankton blooms would be expected. This is consistent with its oligotrophic status.
Zooplankton biomass might reach 100 mg m-3 dry weight. This is higher than most figures from Kariba (Magadza, 1980) but greater production would be expected from Mupata because of its higher conductivity and shallower depth.
3.2 Fish Yields
The range of predicted fish yields from the Mupata lake are:
| (i) | Morphometric models: | Y = 7 920 tonnes = 64 kg ha-1 | (7) |
| Y = 33 kg ha-1 | (8) | ||
| (ii) | MEI (conductivity) models: MEI = 117/16 = 7.3 | ||
| Y = 36 kg ha-1 | (9) | ||
| Y = 46 kg ha-1 | (10) | ||
| Y = 57 kg ha-1 | (11) | ||
| Model (12) was not used because the lake's surface area exceeds 25 km2. | |||
| (iii) | MEI (TDS) models: | MEI = 84/16 = 5.3 | |
| Y = 40 kg ha-1 | (13) | ||
| Y = 52 kg ha-1 | (14) | ||
| Model (15) not used because z <25 m. | |||
Fish yields predicted by these models range from 33 – 64 kg ha-1 with a mean of 46.8 kg ha-1. This is very much higher than the present yield from Kariba but this is to be expected from a much shallower lake. Lake Mupata would be similar in morphometry and hydrology to Lake Kainji in Nigeria, except that conductivity might be rather higher. Kainji produces about 40 kg ha-1 so an overall prediction of 46.8 kg ha-1 for Lake Mupata would seem to be very reasonable.
4. CONCLUSIONS
It is hoped that this example will show how many predictions can be made with relatively little data. In this example there is obviously a great need for further chemical sampling and the chemistry-based predictions are weak. However, those based on physical parameters seem to be more realistic and could be used to plan fisheries development.
