Clarias gariepinus, or African catfish, is widely distributed throughout Africa and parts of the Middle East (Teugels, 1984). The branchial organs allow it to use atmospheric oxygen. It can survive and grow in very poor environmental conditions. The fast growth rate, the omnivorous habits of the species, and consumer preference in many parts of Africa make it a strong candidate for aquaculture.
Research on the reproduction and rearing of larval C. gariepinus in Africa began in the early 1970's with simulation of natural spawning conditions in small ponds. This was followed by simple methods of semi-artificial reproduction. An extract made from a fish pituitary gland or DOCA (11 -desoxycorticosterone-acetate) was used to induce the female to spawn. The female is injected with the hormone and placed with a sexually mature male in a small pond or concrete tank and the fish are left to spawn on their own (Misha, 1973; Janssen, 1989; Coche, pers.comm.). Spawning is often partial as not all the ripe eggs are released. The number of fry obtained is relatively small, with a maximum of 10 000 to 15 000 per female (Janssen, 1989; de Graaf, pers.comm.). The breeders often injure or kill each other, However, the technique is relatively simple and easy to master.
In the 1980's, much of the work in Africa on the propagation of this species has been based on the establishment of specialized, centralized hatcheries using the more advanced techniques of artificial reproduction. This gives the aquaculturist more control over the entire egg laying and hatching process and yields far higher numbers of larvae; 40 000 is not uncommon from a single spawn. The female is treated with a hormone and upon ovulation, the eggs are striped by hand into a bowl, fertilized with macerated testes, and incubated in special containers until hatching (Hoggendorn et al, 1980a; Huisman and Richter, 1987; Janssen, 1989, de Graaf et a!, in press), For a variety of socio-economic and technical reasons, several of these specialized hatcheries have recently fallen into decline or have been abandoned (de Graaf et al, in press; Coche, pers.comm.; Janssen, pers.comm.; Nugent, pers.comm.). It is interesting to note that one can still find the original, simpler techniques being used in Government fry centres around Africa (abet on a very small scale) some 25 years later.
Recently, there has been a renewed interest in more simple techniques, aided by the knowledge gained in more intensive systems. Risch (pers.comm.) recently spawned catfish in Zaire by placing ripe breeders in an empty compost crib in a pond. In the early 1990's, C. gariepinus were reproduced in Congo using the roots of water hyacinth suspended in tanks as a substrate to hatch the eggs (de Graaf et al, in press). In 1993. trials were undertaken in Malawi to reproduce Clarias in oil drums with water syphoned from a fish pond (Brooks, pers.comm.).
The larval Clarias must be grown to fingerling size, or at least 0.5 to 1 g before stocking in grow out ponds. The fish develop the branchial apparatus after 10 to 14 days (temperature dependant) and at a size of between 100 and 200 mg, They can then use atmospheric oxygen which makes transport and handling far easier. Larger fingerlings of 1 to 5 g are preferable; less mortality and predation occurs in the grow out ponds if larger fish are stocked.
Much of the work done in the 1980's on larval rearing of Clarias centered on in-hatchery systems. The larvae are kept inside a hatchery building for several days or weeks. They are fed with an artificial feed such as Artemia or a specialized fish larval feed. although initially zooplankton may be collected from ponds and brought into the hatchery. This is a bit complicated and imported inputs can be expensive.
The other option is to use a small fertilized pond, usually 100 m2 The larvae are simply transferred to the pond 1 to 2 days after hatching when they can swim freely and their yolk sac is about 2/3 absorbed. The natural diet of the larvae is planktonic crustaceans and other zooplankton which occur in the ponds (Greenwood, 1966; Holl. 1968; Bruton, 1979; Hecht and Appelbaum, 1987), This technique is simpler, less expensive, and several of the African hatcheries that began with in-hatchery systems have since shifted to this method. The major problem is the unreliable survival rate which can vary between 0 and 90 % (Micha, 1975; Hogendoorn, 1979; Janssen, 1985; de Graaf et a!, in press),
This variable survival rate is due to several factors; predation by amphibians (frogs) and birds, predation by aquatic insects, competition for food amongst themselves and between phytophagous tadpoles and the catfish larvae, inappropriate or inadequate supplementary feed, and in later stages, cannibalism amongst the fish themselves. Workers experienced in the field conclude that in Africa, the most significant influence on the survival of pond-raised larval Clarias is the presence of frogs and tadpoles, either in direct predation from the frogs, and/or from competition for food from the tadpoles (Nugent pers.comm.; de Graaf et ah in press; Janssen. 1985, and pers.comm.; Hecht et a!, 1988). In Congo, surrounding small (100 m2) ponds with roofing sheets 0.8 m high effectively eliminated frogs from the ponds and the harvested number of fingerlings /m2 were on average 6 times higher than when using unprotected ponds (de Graaf et all, in press). In Cote d'lvoire, survival was greatly improved by placing the larvae in a large mosquito net enclosure, or “hapa”, in the middle of a pond (Nugent, pers.comm.).
In western Kenya, due to the fresh water Clupidae fisheries of Lake Victoria, there are inexpensive, 4 mm mesh (6 ply) nets available in the market, A piece of this net can be used to completely seal a pond from frogs by covering the entire surface. G. J. de Graaf (pers.comm.) mentions that there appears to be a positive correlation between the survival of the larvae and the ratio of pond dike to pond surface area. The shelter provided by the dike or the micro-environment of the dike/pond interface help; the greater the ratio/the higher the survival, A pond measuring 1x10m could easily be covered with a length of 4 mm mesh net, and at the same time have a high ratio of water surface to dike area.
Lake Basin Development Authority collaborated with the FAO Technical Cooperation Programme in 1995 to execute Project TCP/KEN/4551 “Support to Small Scale Rural Aquaculture In Western Kenya”. One aspect of this project was to define a simple, small-scale method of spawning and rearing larval Clarias gariepinus. The eventual aim is allow rural African farmers to propagate this fish themselves. This paper presents the preliminary results obtained from a simplified system for reproduction of Clarias using semi-artificial techniques. Secondly, a method of rearing relatively small numbers of larvae to fingerling stage in very small protected ponds is described and analyzed.
