The techniques for the artificial reproduction of the African catfish Clarias gariepinus are well known and have been practiced in Africa for the past several years (Micha, 1975; Hoggendorn et al, 1980a; Janssen, 1989, De Graaf et al, in press). Inducing the fish to spawn through the use of hormones is relatively straight forward; one of the simplest methods is to inject a fecund female with the crushed pituitary of a similar sized male or female of the same species suspended in a 9 ppt NaCl solution. Depending on the temperature, ovulation occurs 7 to 21 hours later, and the eggs can be easily stripped and fertilized (Viveen et al, 1985). Incubation of the eggs can be done successfully in concrete tanks or using the roots of water hyacinth (De Graaf et al, in press). While the mass production of larval Clarias is fairly simple and it is easy to obtain large numbers at any one time, there are still problems with rearing the larval Clarias to the fingerling stage, or a size of 5 to 10 g.
There are two basic options for rearing larval catfish to this size: pond rearing and in-hatchery systems. In Africa, the later systems have encountered technical and economic problems. Construction costs are relatively high, the necessary techniques are somewhat sophisticated, well balanced and complete feeds are necessary, and considerable time and effort are required for success. Initial feeding using Artemia has worked successfully in the Central African Republic and Cote d'Ivoire, however problems with the cost and irregular supply of encysted Artemia eggs and the resulting unit production cost of a fingerling have inhibited further expansion of the technique. (Janssen, 1989, Nugent, pers.comm, De Graaf, pers.comm). Elsewhere, larvae have been successfully raised in aquaria and small hatcheries using artificial feeds, Artemia, or a combination of both (Hogendorn, 1980, Hecht and Appelbaum, 1987).
The second option is to raise the fingerlings in a pond. With this technique, larvae are transferred to fertilized ponds 3 to 4 days after hatching when they are swimming freely and their yolk sac is about 2/3 absorbed. At this stage, they will accept external feed (Hecht and Appelbaum, 1987). The natural diet of the larvae is planktonic crustaceans (Greenwood, 1966, Holl, 1968, Bruton, 1979) which they find in the fertilized ponds. This technique is some what simpler and less expensive, however the major problem is that the survival rate is unreliable and can vary between 0 and 90 percent. This results in a variable production of 0 to 60 fingerlings/m2/cycle (Micha, 1975; Hogendoorn, 1979; Janssen, 1985; De Graaf et al, in press).
In Africa, even in hatchery based systems, it is usually not economical to raise the entire mass of clarias fingerlings to a size of 5 to 15 g in the hatchery because the requirements of space, flow rates, and feeding large quantities of a well balanced and complete diet. The larvae are usually fed on Artemia for the first 1 or 2 weeks or to a size of 0.2 to 1 g, and then the small fish are transferred to ponds where they are then grown to the required size. In actual field conditions, there is no evidence that initial rearing in the hatchery has any significant effect on survival in the ponds; highly variable mortality still occurs and the results obtained after rearing the fish for about 1 month in a pond are much the same if the larvae were put in the pond directly and feed on naturally occurring zooplankton (Jennsen, pers.comm., De Graaf pers.comm).
In actual practice in Africa, the problem of pond mortality is then common to both systems. This is considered to be due to several factors; predation by amphibians (frogs) and birds, competition for food amongst themselves and between phytophagous frog larvae and the catfish larvae, inappropriate or inadequate feed, and cannibalism among the fish themselves (Nugent, pers.comm.; De Graaf et al, in press; Janssen, 1985, Hecht et al, 1988).
One approach to solving the problems of high and extremely variable mortality is the use of heavy, continuous fertilization in large ponds. Theoretically, the high level of fertilization should create a continuous, abundant source of zooplankton. This should allow the larvae to grow at a more uniform rate, reducing cannibalism. In larger ponds, the increased surface area should hopefully decrease predation pressure.
This paper presents the results of 24 trials of producing Clarias gariepinus fingerlings using heavily manured, large ponds. It is primarily intended to assist those workers in the field of African aquaculture, both within Kenya and without, who may encounter similar situations. The work was carried out in 1994 in Kisumu, Kenya, under the auspices of The Lake Basin Development Authority and in collaboration with UNDP/BSF/FAO Project KEN/86/027, “Development of Small Scale Fish Farming in the Lake Basin Region, Phase II”. The actual field work was done by Mr S Obuya, Technical Officer, and Mr M Spoo, Kibos Farm Supervisor. The team was under the guidance of The Chief Technical Advisor to the Project (Mr D Campbell) and the National Project Coordinator (Mr J Ochieng).
