Of 13 reservoirs greater than 300 km2 in surface area, two are found in Côte d'lvoire: Lake Kossou and Lake Buyo (see Figure 3.1). Lake Kossou was studied extensively, especially in the early 1970s; results of those studies were compiled by Entz (1984) and updated in the present study.
Lake Kossou (Figure 3.2) is located in the savanna lowlands of the tropical belt. It covers large areas of alluvial soils and savanna with vegetation. As a consequence considerable quantities of organic material with a high level of nutrients were liberated from the flood areas of Lake Kossou and this resulted in an initially high biological production of the reservoir. As the nutrients became quickly exhausted, the peak was followed by a sharp decline.
The Bandama River is formed by the confluence of the Bandamas Rouge and Blanc. The White Bandama River was impounded in February 1971, which resulted in the Kossou Reservoir. The principal purpose of the reservoir was the generation of hydro-electric power. Another objective was to prevent the people of the savanna from leaving the area by the development of agricultural activities at the edge of the lake, by the settlement of farmers on the cleared land and by encouraging farmers to start fishing (Bearez, 1988). A dam was constructed in the Bandama River in 1978, forming the Taabo Reservoir, whose water level is regulated by discharges from Kossou upstream. The river has its origin in the northern part of the country between the towns Boundiali and Korhogo at an altitude of about 400 m. Before the periods of severe droughts the river was fed by an average annual rainfall of about 1 400 mm. The river's total length is approximately 950 km; it discharges in the Lahou Lagoon.
No data available.
The Full Supply Level (FSL) was foreseen to be 206 m above sea level. If the lake had reached its maximum level, the dimensions would have been 180 km long and 45 km wide with a total surface area of 1 740 km2 and a volume of 25.5 km3. After closure of the dam in February 1971 the water level was at 157 m and at the end of the first rainy season the level reached 183 m. The highest level recorded was 192.87 m in 1975; the corresponding surface then was of the order of 863 km2, approximately 50% of the potential surface. The lake never reached its maximum level due to the shortage in rainfall since 1973 and to the contruction of dikes upstream for irrigation purposes. Bearez (1988) summarized the available information on water levels of Lake Kossou which are presented in Table 3.1.
Physico-chemical parameters were reviewed by Bearez (1988) and his findings (mainly historical data) are summarized in Table 3.2. Entz (1984) summarized all relevant hydrological, physical, chemical, limnobiological and other data. A modification of this summary is presented in Appendix 1 of this study.
Fishermen from neighbouring countries were attracted to Kona by the news of new rich fishing areas. In the White Bandama region riverine fishing was not a custom. Fishermen from other parts of Côte d'lvoire were encouraged by the country authorities to start fishing in Lake Kossou (Entz, 1984).
The Bandama Valley Management Authority (AVB), assisted by the FAO/UNDP Fisheries Development Project, trained a considerable number of people in fishing techniques. By the end of the project 1 623 fishermen had been trained. In the meantime over 900 Ivorians were attracted to the lake to start fishing there. About 650 foreign fishermen fished in Kossou until 1975 when their access to the lake was refused. The number of AVB canoes increased sharply from 41 to 1 305 over the period 1971–1975 (Entz, 1984). The number of nets increased from 600 to approximately 20 900 and then dropped to 15 728 in 1975 with the departure of the foreign fishermen. More detailed information is presented in Table 3.3.
The dissolution of the AVB in 1982 meant the termination of the assistance to the fishermen; this resulted in a considerable decrease in the number of fishermen. The gear operated are traps, longlines (baited), castnets, gillnets and beach seines. Total numbers of various gear for two administrative regions are presented in Table 3.4.
Before the very first filling of the Kossou Reservoir on the Bandama River, two fishing operations using poison were performed, one upstream of the dam, the other downstream opposite Lamto. Fish population compositions, coefficients of apparent total mortality, biomasses and their variations during the low-water period were studied. The biomass downstream was found to be three times as great as that upstream from Kossou. Yields of 50 kg/ha and 177 kg/ha were observed upstream and downstream, dominant species being Labeo coubie and Alestes rutilus. However, in dead river-arms with stagnant water yields of 305 kg/ha were recorded, the dominant species being Tilapia zillii (Daget et al., 1973).
3.2.5.1 Annual yield
During the FAO/UNDP Fisheries Development project a start was made to estimate the annual fish production. After termination of the project the yield estimation continued for a number of years until 1978 and then stopped. The production was estimated again in 1987. The maximum annual production was recorded in 1977 when it reached a total of 9 315 t, six years after the closure of the dam. The lowest production was observed in 1973 when the lake yielded 4 695 t. Both Entz (1984) and Bearez (1988) summarized the available annual yield figures which are presented in Table 3.5 The estimates of yield per unit area are presented in the same table. The lowest figure of 67 kg/ha/yr was recorded in 1975; in 1987 this amount doubled to 133 kg/ha/yr. The low value for 1975 may be explained by the fact that in April of that year foreign fishermen were no longer allowed access to the lake.
Total fish landings for the period 1972–1975 summarized by Entz differ from those by Bearez, especially for the year 1975. Bearez estimated the fish production at 8 000 t in 1987.
Crul's estimate of potential yield (1992) of 7 378 t is based on a total surface of 1 600 km2. Since this level was never reached, and likely never will be reached, the estimation should be adjusted.
3.2.5.2 Catch rates
Average total catch rates for monofilament as well as multifilament gillnets were presented by Roest (1975) for different seasons in the period 1973–1975. The total average catch rate of 20 mm mesh monofilament nets (knot-to-knot) ranged between 5.5 and 8.5 kg/100 m2/night. Highest catch rates were observed in 25 mm meshed nets (up to 9.3 kg/100 m2/net/night). Nets of 40 mm mesh yielded between 1.5 and 2.5 kg/100 m2/net/night. Average catch rates for different mesh sizes are presented in Figures 3.3 and 3.4 for monofilament and multifilament gillnets respectively. From the above figures it may be concluded that catches of the 25 and 30 mm meshed nets were varying inter-annually. Highest catch rates were recorded in 1973 and lowest in 1975. This pattern was not evident from the larger meshed nets.
Roest presented catch rates for individual species as well; no distinction, however, was made between mesh sizes. As a result, trends are not identical for 1973 and 1974. Only catch rates of the Lates niloticus and Distichodus rostratus showed comparable patterns for the two years. Catch rates for Pellonula afzeliusi are presented in Figure 3.5. A clear annual peak may be observed around the months from November to January.
3.2.5.3 Fishing effort
The total number of fishermen, Ivorian and foreign, increased sharply over a short period of time. In 1972 a total of 1 371 fishermen was registered, which increased up to 2 970 fishermen in 1974 (see Table 3.3). In 1975 the foreign fishermen were denied access to the lake and, although the number of lvorian fishermen increased by approximately 10%, the total number of fishermen dropped from 2 970 to 2 530. As a result the total number of gillnets recorded dropped considerably from 20 902 to 15 728. Kapetsky (1986) considered the lake little exploited. This underexploitation would be largely a matter of economics and culture. Fish from other sources generally is available and the people inhabiting the area, farmers by tradition, have not taken well to fishing.
3.2.5.4 Mesh selectivity
The majority of the data were combined for gillnet fleets of various mesh sizes. No data could be traced for individual mesh sizes.
3.2.5.5 Species composition
In the period October 1972 to January 1975 experimental fishing was carried out by the UNDP/AVB/FAO Fisheries Development Project. A gillnet fleet of various mesh sizes was used, meshes ranging from 30 to 60 mm (knot-to-knot) until November 1973 and from 30 to 85 mm (knot-to-knot) in the remaining period of the project. Roest (1975) presented the evolution of the species composition and drew some interesting conclusions. In 1973 the dominating Alestes species, A. baremoze and A. nurse, were the main contributors to the catches (up to 50% in May 1973). Their importance decreased remarkably. On the other hand A. rutilus became more important; Distichodus rostratus showed an increasing trend as well. The contribution to the catches of Labeo coubie was variable, probably due to its gregarious behaviour which increased its vulnerability to the gear during the spawning season. The importance of the Tilapia species to the commercial fisheries were not reflected in the results of the research. Tilapia zillii and Sarotherodon galilaeus were the target species of the fishermen who set their nets in areas where Tilapia are abundant. The species composition of the project's catches is presented in Figure 3.6.
3.2.5.6 Small clupeids
According to Roest (1975) the microclupeid Pellonula afzeliusi was abundant in the northern part of the lake from September 1973 to April 1974 and subsequently the species appeared in the southern regions of the lake as well. The small clupeids were caught with small meshed gillnets during experimental fishing surveys; meshes used were 5, 10 and 12.5 mm (knot-to-knot). Catches by 5 mm meshed nets were very small. The 10 mm nets also caught Physailia pellucida and Alestes baremoze.
3.2.5.7 Fish population dynamics
Roest (1975) concluded that since the closure of the dam a number of changes had taken place, some of a temporary character, others more permanent. The temporary changes were caused by variation in the year class strength of a number of species such as Alestes baremoze, A. nurse, Distichodus rostratus, Eutropius mentalis and Labeo senegalensis. The more or less permanent changes were the reduction in abundance in the northern part of the lake (probably due to migration) of Lates niloticus and of the following families: Polypteridae, Clariidae, Notopteridae, Mormyridae and Mochokidae.
Jocque (1975; 1977) studied the biology of the African catfish, Clarias senegalensis, the dominant silurid of the man-made Lake Kossou, and some smaller impoundments in the same region. Comparison of data spread over a five-year period starting from 1970, some months before the dam was completed, reveals that the density of C. senegalensis dropped to the original level that existed in the river after reaching a peak in the years following the damming. This phenomenon, which coincides with the impoverishment of the lake waters, is explained by the fact that C. senegalensis has its optimal habitat in poorly oxygenated waters, rich in organic matter. It is able to colonize this kind of productive waters thanks to the suprabranchial organ. In these optimal conditions it mainly feeds on aquatic insects and plankton. The fecundity of C. senegalensis may reach 200 000 ripening eggs per female. Reproduction occurs during the rainy season in flooded savannas. Growth rates are very high and the age determinations show that, in the region of Kossou, C. senegalensis reaches an average standard length of 620 mm (2.7 kg) within three years.
3.2.6.1 Introduction of non-indigenous species
In 1971–1972 two species were introduced into Lake Kossou: Heterotis niloticus and Oreochromis niloticus (Kouassi, 1979). Kapetsky (1986) considered the economic success of the introductions poor.
3.2.6.2 Cage culture
The experimental rearing of Oreochromis niloticus in 1 m3 cages floating in Lake Kossou, Côte d'lvoire, was performed between March 1974 and April 1975. The fish were fed water-stable pellets (24.7% protein). Fingerlings (9–55 g each) were stocked at densities of 200–400 fish/m3. They reached commercial size (more than 200 g) within 4–6 months, depending on the biomass present in the cage (stocking density and average individual weight of fish), as well as on limnological conditions. The average annual mortality was 5.9%. Annual production varied from 36 to 64 kg/m3. The feed conversion rate averaged 2.8. Fingerlings should average at least 20–30 g in weight at initial stocking of the cage. Two crops a year can then be easily realized. If only male fish are reared, three harvests per year appear to be possible, increasing the annual cage production of fish close to 200 kg/m3 with proper management. The possible development of O. niloticus cage culture is actually hampered due to the lack of adequate fish feed available commercially, and to the lack of sufficient quantities of fingerlings throughout the year (Coche, 1977).
The Buyo Reservoir was formed by impoundment of the Sassandra River at Buyo. This river has its origin near Odienné; it has a length of 840 km and it drains an area of 66 000 km2.
Another inflowing river of the Buyo Reservoir is the N'zo River; the only outflowing river is the Sassandra, which discharges to the Atlantic Ocean.
No data available.
The Buyo dam was closed in March 1980 resulting in an artificial lake with a maximum length of 89 km, a maximum width of 23 km and a total surface area of 900 km2 at Full Supply Level, which is at 200 m above sea level. Its minimum level is at 186.5 m above sea level; the corresponding surface is then around 200 km2.
The ionic composition of the Sassandra water was determined by Molinier (1976) as quoted by Vanden Bossche and Bernacsek (1990b). The following concentrations were reported: Na, 5.1 mg/l; K, 2.4 mg/l; Ca, 4.2 mg/l; Mg, 2.3 mg/l; HCO3, 34.1 mg/l and Cl, 2.6 mg/l.
No data available.
3.3.5.1 Annual yield
The potential annual yield was estimated at 4 345 t by Crul (1992). No other yield data were available to the author.
3.3.5.2 Catch rates
No data available.
3.3.5.3 Fishing effort
Vanden Bossche and Bernacsek (1990b) reported 238 fishermen in 1983. No other data available.
3.3.5.4 Mesh selectivity
No data available.
3.3.5.5 Species composition
No data available.
3.3.5.6 Small clupeids
No data available. It is not known whether any small clupeids occur in the Sassandra and N'zo Rivers.
No information available.
