The Government of Nigeria, assisted by the United Nations Development Programme and the Food and Agriculture Organization of the United Nations has been engaged in a project whose main purpose is to assist in the comprehensive development of man-made lake resources through research and surveys, the results of which will be made available to all regions of Nigeria. Under the Fisheries Programme of the Work Plan of the Project Document concerning Lake Fisheries, studies are required to assess the possibility of cage aquaculture.
The predicted decline in fisheries in the relatively infertile lake and in the newly regulated river below the lake, led to this interest in possible fish production through aquaculture.
The project became operational on 12 August 1968 and, as part of the project operation, FAO assigned Mr. Mark Konikoff, an Aquaculture Consultant, from 4 May to 4 August 1974 with the following terms of references
“to investigate the feasibility of cage aquaculture in the lake and running water aquaculture in the vicinity of the River Niger.” Further terms of reference included selection of suitable species for aquaculture schemes deemed feasible, initiation of field tests, and preparation of a programme of work for development in this field.
Kainji Lake was created in 1968 by the completion of the first of two proposed dams on the River Niger. It is located in northwestern Nigeria between 9°50' and 10°50' N and is 1 280 km2 in area. The lake is mainly a hydroelectric project, but is also important in fisheries, flood control, navigation, and agriculture*
Kainji Lake Research Project is multidisciplinary and is concerned with the limnology, fisheries, wildlife, range ecology, agriculture, economics, sociology, and public health of the impoundment and the surrounding area. FAO assistance to the project started in the pre-impoundment period.
Kainji Lake is about 130 km long and has a maximum width of 24 km. The lake exhibits a high exchange ratio (ratio of volume of outflow to volume of lake) of 4:1 per annum. It also has a high annual drawdown of about 10 m. Average depth varies between 8.8 and 12.3 m, and maximum depth is about 50 m (Henderson, 1973).
Kainji Lake has two major periods of water inflow. The “White Flood” which begins in August, is derived from local rains and carries a large quantity of colloidal clay. This water is highly turbid. The “Black Flood” which reaches its peak in February, originates in the headwaters of the Niger and is relatively clear. In a normal year, maximum water levels are in December (the dry season), and minimum levels in August (the rainy season).
The lake is stratified from February to mid May. During this period, the hypolimnion becomes deoxygenated and hypolimnial water containing dissolved hydrogen sulphide is discharged into the river. In the lake, the upper 20 m contain dissolved oxygen all year long, and surface waters (1 m in depth; do not fall below 70 percent saturation. Temperature of surface waters is fairly constant, varying from about 25°C in January to about 30°C in May (Henderson, 1973).
Primary productivity in Kainji Lake is low, compared to other tropical lakes, due to low nutrient content and restricted transparency (inorganic turbidity) of the water (Karlman, 1973). Littoral vegetation is still undeveloped. This may be due to the large annual drawdown.
In high intensity aquaculture, fish are held at high densities (200 to 400 fish/m3) in raceways, tanks or cages, and given a high quality feed for rapid growth. All three systems require a species of fish that grows rapidly, accepts artificial feed, tolerates crowding, and has a high market value.
The systems differ only in the method of delivering fresh oxygenated water to the fish and removing wastes. In raceway and tank culture, this is accomplished in a steady high volume flush of fresh water, and sometimes, mechanical aeration. In cage culture, the cages are floated at the surface of a body of water and are constantly flushed by currents without mechanical aids.
In all three systems, there are savings in the cost of land and labour over that which would be required to raise a similar quantity of fish in intensive pond culture, although the cost of feed and the technology required is high.
Given suitable sites, cage culture can be the most attractive of the three systems* It requires less technology, has low capital costs in small-scale production, and requires virtually no land investment.
Raceways culture probably has the lowest capital costs in large-scale production, and requires only a moderate amount of technology. It is advantageous where a large quantity of flowing water is available. Tank culture has high capital costs for equipment and also requires advanced technology such as automatic backup systems for electrical power failures and continuous monitoring of water chemistry. It is advantageous in culture of highly valuable species where land values are also high.
Since the biological problems are identical in all three systems, it was decided to initiate a pilot study using cage culture in Kainji Lake. Information obtained about suitable species, growth rates, food conversion, etc, would be applicable to the other high intensity systems.
Six experimental, wooden framed, poultry mesh cages measuring 1 m × 1 m × 1.25 m were constructed (Figures 2 and 3). A sheet metal feed tray measuring 61 cm x 29 cm × 6.4 cm was installed in the bottom corner of each of five cages (Figure 4). A 5 cm × 150 cm aluminium feeding tube was installed vertically with the opening about 3 cm above the tray bottom. This was later replaced with a 7 cm diameter tube because the smaller tube became easily clogged with feed.
The cages were partly covered with metal roof sheeting leaving a 15 cm gap at one end. This opening was covered with scrap nylon netting which was easily moved aside for adding or removing fishes. One cage was left without feed tray, tube, or metal cover and was used as a holding pen for fish to be stocked.
Four of the cages were suspended by ropes from a floating pier (Figures 5 and 6). Two cages were floated with pieces of scrap polystyrene (styrofoam) tied in netting. All of the cages were suspended so that 1 m3 was in the water.
In addition to the wire mesh cages, several woven kagi reed cages were under construction at the time of the departure of the consultant. The method of weaving used is similar to that for the “gora” trap described by Reed (1969).
The cost of material for building the six 1 m3 cages was Naira 132 (U.S.$ 211)or about Naira 22 (U.S.$ 35) per cage (Table 1). This price did not include labour, rope used to suspend the cages, scrap polystyrene, scrap netting, or the aluminium irrigation pipes used for feeding tubes. Some of the materials could be replaced by inexpensive local ones. For example, bamboo could be used for feeding tubes, and raffia could be used for floatation. The cost of materials for construction of three kagi reed cages was Naira 21 (U.S.$ 34) or about Naira 7 (U.S.$ 11) per cage. This includes reeds and metal for feed trays.
The most important calculation in determining the cost of cages is the cost per unit weight of fish raised in the cage. A 1 m3 cage can be used to produce 100 kg of fish every six months. Assuming a cage life of one year, the cost of the poultry mesh cages would be Naira 110 (U.S.$ 176) per metric ton of fish produced. Longer cage life would reduce costs proportionally. Welded wire mesh cages used in fresh water will last over three years. Although poultry mesh is probably not as resistant to corrosion as welded wire, Kainji Lake water has a low conductivity (55 micromhos per centimetre) (Henderson, 1973) and this would tend to increase cage longevity. At the time of the departure of the consultant, cages which had been in the water for over six weeks showed no sign of deterioration.
Another factor influencing the economy of cage culture is the size of cage used. The cost per unit weight of fish is inversely proportional to the linear dimensions of the cage. Since commercially operated cages would be larger than the experimental cages by a factor of 2 to 4, costs would be reduced substantially. Furthermore, experimental cages were constructed from materials purchased at retail prices in small quantities at local markets. Buying larger quantities of materials at wholesale prices would reduce the costs by an additional 25 percent or more.
Fishes suitable for cage culture must tolerate crowding, accept artificial feed, exhibit good growth rates, be readily available for stocking, and have good market prices. The nutrition and natural habitat of a species are important. Highly carnivorous fish are usually less desirable than detritus feeders or herbivorous fish because of higher cost of feed. Strictly planktophagic fish would be difficult to feed. Highly mobile predatory fish would be difficult to confine.
A major break-through in cage culture was the discovery that although many fishes would fight when confined in moderate densities (25 to 75 fish per m3), these same fish would thrive at higher densities (200 to 300 fish per m3) (Lewis, 1969). Freshwater fishes successfully cultured in cages include catfishes, cichlids, cyprinids, and salmonids.
Of the fishes available at Kainji Lake, there are three groups that suggest good chances of economic production in cages. These are Tilapia sp., Clarias sp. and Citharinus sp. The catfishes Bagrus and Heterobranchus are also promising. All of these have high market prices (wholesale prices about Naira 60/kg - U.S.$ 96/kg).
Tilapias are the most widely cultured fishes in Africa and have been grown in cages (Pagan, 1969; Shehadeh, 1974). There are four species of Tilapia in Kainji Lake: T. galilaea, T. nilotica, T. melanopleura and T. zilli, T. melanopleura is probably the most desirable local fish for cage culture. Aquarium observations indicate that it is highly adaptable to artificial feed. Unfortunately, it is not abundant in Kainji Lake, and only a few individuals were captured. About 100 T. melanopleura were obtained from the government fish ponds in Ibadan, but these died in transport.
T. nilotica is also a desirable cage fish. These also were not numerous enough for stocking a cage. Two cages, however, were stocked with these along with other tilapias.
T.galilaea is the most abundant tilapia in Kainji Lake. Pish of all sizes were easily caught by seining and cast netting. Although it has been reported to be strictly planktophagic, aquarium observations indicate that it learns to accept pelleted feed in 10 to 14 days. These fish were stocked monospecifically in two cages and with other tilapias in two cages.
T. zilli is not attractive for cage culture. It is smaller than the other tilapias and is highly pugnacious. However, it is readily available and was used in one of the mixed species cages.
The primary problem in using tilapia for cage culture is their high sensitivity to handling. Tilapia captured by seining and cast netting, carefully handled, and placed in a holding pen exhibited mortalities as high as 90 percent in two days. They were highly excitable and tended to jump during short transport. Later, 50 to 100 ppm MS 222 was used to tranquilize the fish immediately after capture. This reduced mortality in the holding cage to about 30 percent. Similar nervous behaviour and mortality in confined tilapia was reported by Avault et al. (1966).
Capture of Clarias spp. for cage stocking was continuing at the time of departure of the consultant. These fish are very tolerant of handling, but had to be captured by electrofishing which led to some mortality
Citharinus spp. suitable for stocking were not available during the summer months, but should become readily available in early December (Dr. B. Blake, Personal Communication). It is planned to stock a cage with them at that time. Citharinus spp. are omnivorous and grow very rapidly, but their resistance to handling and crowding is not known,
Fingerlings of the catfishes (Bagrus and Heterobranchus) were not readily available. The production of stockable Tilapia, Clarias and other catfishes could, however, be accomplished in small breeding ponds.
The most severe constraint on cage culture is the necessity of using an economical but nutritionally complete feed. Feeds which may produce good growth in pond culture can lead to nutritional deficiencies, unthriftiness, and mortality in high intensity aquaculture. Formulation of an economic feed at Kainji Lake was attempted under a double handicap. Not only are the nutritional requirements of the fish unknown, but information on commodity prices in Nigeria is difficult to obtain.
Feed was prepared using ingredients available at the local market: guinea corn bran (60 percent), roasted groundnuts (19 percent), yam flour (10 percent) and dried clupeids (10 percent) (Table 2). A poultry vitamin/trace mineral supplement (1 percent) was also included. Feed analysis is given in Table 3.
Feed ingredients were dried, ground, screened (1.6 mm screen) and stored. Measured ingredients were mixed with water to form a paste and this was extruded through a plate with 6 mm holes using a kitchen food grinder (Figure 7).
The pellets were sun dried in sheet metal pans with glass covers at 50° to 60°C. A laboratory oven was occasionally used, but there is sufficient sunshine at Kainji to dry pellets during most of the year. At the time of departure of the consultant, two kilograms of feed were being produced daily.
The feed pellets were water stable for about ten minutes. Tilapia in two 166 litre aquaria were used to determine pellet acceptability. The fish ate the feed vigorously and were maintained in good health for six weeks. Underwater observations of the fish in cages during feeding were made by skin diving. These indicated good palatability and the fish fed voraciously. On calm days, the fish could be seen coming to the surface to feed on pellets that fell into the cage. Large numbers of Alestes were attracted to the cages by small pieces of feed that were washed out.
The experimental feed could be produced for Naira 151 per metric ton (U.S.$ 242) if ingredients were purchased in 50 kg sacks at local wholesale prices. Retail purchase increases the cost to Naira 222 per ton (U.S.$ 355) (Table 2). The actual cost of ingredients when purchased locally in small quantities was considerably higher (Naira 42/kg = Naira 420/ ton - U.S.$ 770/ton).
Costs could be reduced by replacing local ingredients with those available nationally. Prices of some live stock feed ingredients in Ibadan in May 1974t are given in Table 4. Current prices of agricultural commodities can be obtained from Nigerian Produce Marketing Company Ltd. (72 Campbell Street, Lagos).
A substantial reduction in cost could be made by catching and processing clupeids from the lake. Some of the experimental feed was produced this way and plans were to continue this practice after the departure of the consultant.
The economy of the experimental feed depends on the efficiency with which the fish will grow. For example, if they are able to grow with a conversion ratio of 3t1 (3 kg feed for each kg growth) the cost of producing 1 metric ton of fish in a commercial size cage would be about Naira 475 (U.S.$ 760). These fish could be marketed at about Naira 600 per ton (U.S.$ 960). Small changes in feed efficiency would greatly affect profitability.
The requirements for a site for cage culture in a lake are good water quality, enough water depth to float the cage during low water, and protection from wind and wave damage. It is also desirable to locate cages where they will be convenient to feed and maintain, where security is available, and where harvested fish can easily be transported to markets.
The critical factor at Kainji Lake is protection from waves* The direction of prevailing winds coincides with the north-south axis of the lake, and waves of 50 to 70 cm are common (Henderson, 1973). Many of the east-west bays also get substantial wave action.
Experimental cages were set up at the floating pier (“the jetty”) at the south end of the lake, and in a small bay at Shagunu on the western shore of the wide middle section of the lake (Figure 1).The winds during this period were mainly southerly and the jetty cages were not strongly affected. The cages at Shagunu were more exposed, but had withstood several periods of strong wave action without damage. These cages were floated with scrap styrofoam wrapped in netting and tied between dead trees. They were placed so that they would get Borne protection behind a point of land projecting from the side of the bay. Bays similar to this one are common on the irregular Kainji Lake shore, and suitable site for cages are readily available.
Cage culture can be conducted in backwaters of rivers where they will not be subjected to damage by currents. Other requirements are similar to those of lake sites. The River Niger was inspected for aquaculture possibilities by the consultant during a river float trip from the Dam Site to below Leaba, a distance of 35 to 40 km.
The river below Kainji Lake is probably unsuitable for cage culture. In the area immediately below the dam there are several quiet backwaters which could accommodate cages, but they are limited in area. Furthermore, during parts of the year the hypolimnial water released from the lake contains dissolved hydrogen sulphide. The distance that the river is affected by this is unknown, but almost certainly includes these backwaters.
Farther below the lake, the river is narrow and currents are strong. Although there are places among the rooky out-croppings that could accommodate a few cages, these are limited. Another problem is that many spots on the river which might be suitable for cages are isolated from villages and can be conveniently reached only by boat.
Two cages (1 m3 capacity) were set up at the floating pier near Kainji Lake Research Project Headquarters at the south end of the lake. Cage 1 was stocked with 160 tilapias (mixed species, but mostly T. galilaea) on 25 June, and another 96fish on 1 July, for a total of 256 fishes. Average weight was 52 grammes. The fishes exhibited high mortality in the first few days after stocking, and 75 fishes (30 percent of the total) died by 5 July. Only two additional fishes died in the next three weeks.
Cage 2 was stocked with 127 T. galilaea on 1 July and an additional 202 fish on 12 July. Average weight of 329 fish was 29 grammes,Again, there was high mortality in the first few days following stocking, and 102 fish (31 percent) died by 18 July. Only five additional fish died in the following week. A third cage was being stocked with Clarias at the time of departure of the consultant.
The fish were fed at 1 percent of body weight for the first 10 to 14 days after stocking and then 3 percent of the calculated weight of survivors. Underwater observations were made after 10 days. The fish appeared healthy and ate vigorously. Feed is presently being increased at rate of 10 percent every 10 days. Because of mortality problems in handling, the fish will not be subsampled. Harvest is expected in 3 to 4 months.
Two cages were set up in the lake near the Shagunu Research Station. They are being tended by the staff of the station in cooperation with Kainji Lake Research Project staff. Cage S1 was stocked with 263 T. galilaea (x = 32 grammes)on 4 July. Mortality in the first few days was high, and 91 fish (35 percent) died by 9 July. Mortality in the next two weeks was negligible. Cage S2 was stocked with 42 T. nilotica and 192 T. galilaea on 4 July. Average weight of the 234 fishes was 41 grammes. Again mortality was high with 91 fishes (39 percent) dying in the first week but negligible deaths in the weeks following. These cages are to be fed until November. They will then be harvested and restocked with Citharinus .
There is great potential for aquaculture in this area. Pish are an important part of the diet and are in demand. Water supply is plentiful and land, labour, and power are available. Pish marketing is haphazard and would greatly benefit from a dependable source of fish. There is a large variety of fishes from which to choose cultured species. Many of these, such as Heterotis, Channa, Citharinus, and Gymnarchus, have been little used in aquaculture and offer highly interesting possibilities for polyculture.
Estimates of profitability of fish farming in Nigeria vary widely. A 1972 report predicted only a 6.5 percent return on investment. This was based on a production estimate of 1 250 lb/acre (1 402 kg/ha) per annum in feed ponds. A more realistic estimate of 2 400 lb/acre (2 688 kg/ha) was made by Sivalingam (1974). He has estimated a return of 14 percent on investment in a 50 acre (20.23 ha) fish farm in Nigeria, and a 30 percent return on investment on a 2 acre (0.8 ha) homestead pond* Other workers in tropical ponds have been able to produce even higher yields, and annual production of 4 000 lb/acre (4 480 kg/ha) in polyculture ponds would not be an impractical goal.
Both cage culture in the lake and raceway culture on the banks of the river below the dam could be established at Kainji.Large quantities of “free” high quality water are available for both.The only constraints are the availability of cheap nutritionally complete feed, a supply of stockable fry, and a shortage of trained personnel.Testing of feed in cages was in progress at the time of the departure of the consultant.Production of fry could be accomplished in small ponds peripheral to the lake, and the training of personnel could be undertaken at the new Fisheries Training School now under construction at Monai.
Production of warmwater fishes in raceways is very efficient.Capacity depends on the amount of water available, and production can be as high as three metric tons of fish per annum per cubic metre per minute of water supplied.This could be done in 4 to 6 concrete raceways 1 m × 1.25 m × 8 m.
A good location for this would be on the west bank of the river about 50 to 100 miles south of the road at the base of the dam.Water could be supplied by a siphon pipe over the dam. At this point, there would be about30 miles of head available.Species cultured in raceways would probably be catfishes.
In addition to the general requirements for fish culture, pond culture requires land with poorly draining soil that will retain water. Unfortunately, most of the soil surrounding the lake is sandy with rocky out croppings and is unsuitable for pond culture. However, clayey, poorly draining soils are still plentiful. For example, an area of 1 000 to 1 500 acres (400 to 600 ha) of flat, clayey, poorly drained soil (Kurwassa Association) is found near Makawa, south of Kayala, on the eastern shore of the lake. Other areas with suitable soils are also found around the lake. Many of these areas are being considered for irrigated agriculture. Irrigation pumps and equipment could also be used for aquaculture in some places. Areas with poorly draining soils are usually marginal for agriculture, but may be used for rice. Rice-cum-fish culture or alternate culture of rice and fish are good possibilities in these areas.
Establishment of large fish farms (ponds) along the river is not attractive. Suitable sites are isolated and small. Small homestead ponds, however, offer good possibilities. The consultant noted several places along the river that would be ideal for small (0.5 to 2.0 ha) ponds. These could be filled by pumping from the river with an inexpensive, portable, gasoline powered pump. For example, a self priming, 3 horsepower, 2 inch (5.08 cm), 64 lb (29 kg) pump, sold in the U.S.A. for U.S.$ 119, can deliver 100 U.S. gpm (378.5 litres per minute) at 50 feet (15.2 metres) of head. This would fill a two acre-ft (2 467 m3) pond in about four days.
Homestead ponds can produce appreciable quantities of fish when fertilized with animal and household wastes and stocked with several species of fishes. Fishes that could be used include Tilapia spp., Citharinus, Gymnarchus, Clarias, and other catfishes. Channa and Heterotis are both able to use atomospheric oxygen (as are Clarias spp.) and these might be good fishes for small ponds. Heterotis is phytoplanktophagic, Channa is piscivorous, and Clarias spp. are omnivorous.
The harvest of fishes from flooded swamplands below Kainji Dam has been greatly reduced by the regulation of the River Niger. Reed (1969) suggested the use of check dams to maintain water levels in swamps after high water. These enclosures were tested by Reed and proved to be very productive. Unfortunately, this practice has been abandoned because there has been inadequate high water over the past few years. The check dam sites mapped out by Reed (below Jebba) could still be used, but considerable pumping might be required. Another problem with building enclosures in the swamps just below Kainji Lake is that the construction of a second dam on the river at Jebba will further alter the water level regimen, and flood some of the sites.
Above the dam, in Kainji Lake, there is good opportunity for enclosure culture. There are many shallow bays with relatively narrow mouths. Annual drawdown in the lake is about 10 metres and many of these bays are exposed at low water. A check dam across the mouths of these would create a pond that could be high productive. With fertilization and/or feeding, these enclosures could produce two to four metric tons per ha. Fishes used woud be the same as in pond culture, Tilapia, Citharinus, Gymnarchus, Clarias, and other catfishes. Permanent breeding ponds above high water would have to be established to supply fry for the enclosures.
(1) Experiments to determine the feasibility of cage culture in the lake should continue. Criteria for determining the suitability of various species for cage culture should be feed conversion rate, mortality, fingerling availability, and marketability. Especially interesting will be a comparison of Tilapia, Citharinus and Clarias . Promising-species should be raised in larger commercial sized cages.
(2) Composition of experimental feed should be varied and tested on fish in cages. Replacing the yam flour in the present formula with blood meal would not change the rice, but would increase the protein level from 18 percent to about 25 percent. Feed ingredients should be purchased in larger quantities to lower the cost.
(3) Five to ten small (0.01 ha) breeding ponds should toe constructed peripheral to the lake. These should be located near the research project. The ponds should be drainable, and water could be supplied with an inexpensive portable gasoline powered pump. If necessary, the ponds could be sealed by burying polyethylene sheets beneath them. These ponds should be used for breeding Tilapia and Clarias for experimental aquacultural schemes.
(4) A peripheral pond culture project should be initiated. This fish farm should be constructed in an area with clayey, poorly draining soil that is fairly level. The ponds should be just above the high water level of the lake. They should be drainable and 0.5 to 2.0 ha in size. The water supply system (pumped lake water) should be large enough to fill 10 percent of the ponds in four days while supplying maintenance water to the others. Smaller (0.1 to 0»4 ha) breeding ponds should also be constructed. This farm should be used to determine good combinations of local fishes for pond culture, and to demonstrate pond culture to area peoples. Promising local fishes include Tilapia, Citharinus, Gymnarchus, Heterotis, Channa, Clarias and other catfishes. Lates might also be suitable for pond culture.
(5) Check dams should be constructed at the mouths of shallow bays in the lake. Experimental groups of fishes should then be stocked in these enclosures. Economy of fertilization and/or feeding should be investigated. Fish stocked would be similar to those in ponds.
(6) Experimental homestead ponds should be built on the banks of the Niger. These would be filled by pumping from the river with an inexpensive portable gasoline powered pump. The ponds should be associated with poultry keeping and general farming and agricultural wastes should be added to the pond. Fish stocked would be similar to those in ponds near the lake.
(7) A small expandable pilot raceway culture plant should be constructed on the banks of the Niger. A good place for this is on the west bank just below the dam. Four to six 1 m × 1.25 m × 8 m concrete raceways would be supplied with water by a siphon pipe over the dam. A small pump would be used to fill the siphon. Fish stocked would probably be Clarias or other catfishes, and production of about three metric tons per cubic metre per minute of water per year. Fish produced would require a nutritionally complete feed.
(8) An aquaculture training programme should be established at the new Fisheries Training School at Monai.
(9) The continuation of cage culture experiments, the testing of experimental feeds, the construction of small breeding ponds, and the construction of check dams (recommendations 1, 2, 3, 5) could be carried out by the present staff of the project. The building and initial operation of a peripheral pond culture fish farm, homestead ponds, pilot raceway plant, and the teaching of aquaculture (recommendations 4, 6, 7, 8) will require the assistance of one or more technical experts and extension workers.
