In Tanzania numerous dams have been constructed for irrigation, domestic and livestock use as well as for flood control. About 50% of the resulting reservoirs have been stocked with fish. Baily in 1966 described a total of 490 reservoirs out of which the Nyumba ya Mungu Reservoir was the largest. In 1980 the Mtera dam in the Great Ruaha River was closed, which then resulted in the largest Tanzanian man-made lake (Vanden Bossche and Bernacsek, 1990a). The Mtera Reservoir is situated in the central part of Tanzania within the Rift Valley at the confluence of the Great Ruaha and Kisigo Rivers, filling most of the Pawaga and Logi plains that were originally grazing areas for livestock and wildlife (Mwalyosi, 1986). Positions of the principal reservoirs in Tanzania are presented in Figure 6.1.
The Great Ruaha River is the principal inflowing river feeding the Mtera Reservoir. The Great Ruaha has its origin in the Poroto Mountains and Kipengere Range; its length is about 750 km. Two dams have been constructed, the first (the Kidatu) in 1974, and the second (the Mtera) in 1980. The Mtera dam is found upstream of the Kidatu dam. Another inflowing river is the Kisigo River. The outflowing Great Ruaha discharges into the Rufiji. Figure 6.2 shows the shape of the lake in the area of the Mtera dam.
A thorough pre-impoundment study has been carried out by the Tanzania Electric Supply Company (TANESCO) in cooperation with the Swedish Consulting Group (SWECO). In their report (SWECO, 1976) various important matters are discussed varying from vegetation to filariasis surveys. Special attention was given to the multi-purpose planning of the reservoir. Petr (1976) studied the limnology of the Great Ruaha River as well as its fish fauna. He estimated a yield of the order of 15 000–30 000 t of fish/yr at the peak of the lake's productivity. He predicted the peak to occur between the third and sixth year after impoundment. Petr based this assumption on a rather limited number of fishing experiments in the Great Ruaha and Little Ruaha Rivers, the Nyumba ya Mungu and Hombolo Reservoirs. The fish family composition in the middle course of the Great Ruaha River was found to be as follows:
About 1% of the catches consisted of tilapias (mainly Tilapia urolepis). Petr recommended following the changes in fish population after impoundment and that, in case of failure of indigenous cyprinids and cichlids, introduction of other species should be considered. The various species observed in catches in 1971–1973 are listed in Table 6.1. Mwalyosi (1986) quoted a pre-impoundment study performed by Bernacsek (1980) who stated that the fishing potential was excellent with a production ranging from 3 000 to 7 000 tonnes per year for stabilized catches and from 4 800 to 31 400 per year for bloom catches. Prior to the closure of the dam an area of 141 km2 was cleared from vegetation (Mwalyosi, 1986).
Water physics and chemistry of the Great and Little Ruaha Rivers were studied by Petr in 1972 and 1973 (Petr, 1976). Relevant physico-chemical data for Mtera have been extracted from his tables and are presented in Table 6.2. The lake drains 68 000 km2 and inundates 610 km2 at its Full Supply Level (FSL). It receives an average of 450 mm of rainfall per year. Its surface at Minimum Supply Level is 190 km2 (SWECO, 1976; Mwalyosi, 1986).
The lake's altitude is 698.5 m above sea level, the dam's height is 45 m; the lake's depth was calculated to be 6.23 m at FSL. The maximum capacity is about 3.8 m3. Maximum values for length and width are 56 km and 15 km respectively (Vanden Bossche and Bernacsek, 1990a).
Mwalyosi (1986) visited 13 fishing camps in November 1983 and recorded about 500 fishermen. However, according to fisheries authorities the total number would be around 2 000. The fishing was conducted at that time mainly by professional fishermen who came from other lakes such as Nyasa, Victoria, Rukwa, Babati, and Tanganyika. Due to distance, accessibility and time limitations, not all fishing camps around the lake could be visited. The fishermen usually operate individually, using dugout canoes, according to Mwalyosi. The team recorded 450 boats; the total number of boats around the lake would be 1 800. Campbell and Moreni (1988) stated that the supply of suitable trees for the construction of new canoes was severely depleted.
Nylon gillnets are the principal gear. Fisheries regulations allow four inch mesh and above, but 2-inch meshed nets are used as well. Fishermen possess several gillnets, ranging in number from two to ten. A total of 3 720 gillnets were recorded at the camps; an estimate by the fisheries officials for the entire area was 14 000.
Lyimo (pers. comm.) summarized available information. The number of gear are known for the years 1986 and 1990. Although the number of fishermen increased in that period, the number of gillnets, longlines and traps decreased. Mesh sizes in 1986 ranged from 2 to 5 inches (stretched), while in 1990 the range was from 2.5 to 7 inches (stretched). The number of fishermen, boats and gear for years 1986, 1987, 1990, and 1991 are presented in Table 6.2.
126.96.36.199 Annual yield
Limited data are available on the fish production figures of the Mtera Lake. The total fish yield in 1983 was estimated by Mwalyosi (1986) at 23 000 t; this estimation was based on the ‘landing approach’, which depends on the level of fishing activity as given by Welcomme (1983), and on additional information from fisheries officers. Chale and Mwaya (1984) would have obtained independently a similar figure for the annual fish production, according to Mwalyosi. Vanden Bossche and Bernacsek (1990a) summarized existing data. The total annual catch in 1984 would have been of the order of 21 000 t in 1984. This figure is well in line with the prediction by Petr (1976). An enormous drop in production occurred after 1984; in 1986 the annual yield was estimated at 3 254 t. It should be observed that an annual production of 23 000 t at FSL means a production of 377 kg/ha; this value is quite high compared with production figures of other lakes. Campbell and Moreni (1988) expected that the sustained production would be around 5 000 t.
After 1986 the estimated production increased gradually to 5 025 t in 1991; the number of fishermen increased from 1 238 in 1986 to 1 815 in 1990 (Lyimo, pers. comm.). An annual production of 5 000 t would correspond to a yield of 82 kg/ha. This estimate seems more realistic than the 377 kg/ha. Yield estimates since 1986 are summarized in Table 6.2.
188.8.131.52 Catch rates
No catch rate data were available to the author.
184.108.40.206 Fishing effort
The number of fishermen and gear as mentioned under 6.2.4 are the only indication of effort available.
220.127.116.11 Mesh selectivity
No data available.
18.104.22.168 Species composition of catches
Tilapia urolepis and Hydrocinus vittatus appeared to be dominating in catches in 1983. Other species included Clarias sp., Labeo sp., Synodontis and Alestes. Adaptation to lacustrine conditions by Tilapia urolepis was quite successful (see Table 6.1, in which T. urolepis had an insignificant place).
22.214.171.124 Small clupeids
No information available.
No information available.