The rearing of the African catfish caught also attention in Kenya as it fetches a higher price as Nile tilapia and the project for the development of small scale fish farming in the lake basin area started to reproduce and raise the African catfish in April 1993 after the arrival of the new CTA. First experiments on the artificial reproduction of the African catfish were executed at Kibos fish farm and semi-artificial spawning techniques were used at first. This technique is rather simple, the results were promising and it was decided to produce also catfish fingerlings at the other fry production centres (FPC) of the project. At the FPC the existing facilities were modified into small hatcheries. In february 1994 all hatcheries were ready for operation and a summary of the facilities of the FPC's is presented in Table 1.
Table 1: The existing facilities at the fry production centres of the project
Station | Pond area (ha) | Water source | Facilities |
Kibos | 1.35 | drilled well/pumping | concrete basins |
Yala | 0.35 | spring/gravity | concrete basins |
Alupe | 0.35 | spring/gravity | concrete basins |
Rongo | 0.23 | spring/gravity | concrete basins & water storage/sedimentation tanks |
Chewele | 0.68 | spring/gravity | concrete basins & water storage/sedimentation tanks |
Mature females are injected with pituitaries of Catfish or Nile Tilapia and are placed together with a male in a concrete tank. The bottom of the concrete tank is covered with a layer of stone gravel which function as substrate for the released egg. The breeders will spawn in the night after the pituitaries are injected. The fertilized eggs will stick to the gravel and the following morning the breeders are removed from the tank. The fertilized eggs remain in the tank and hatching takes place and after 3–4 days approximately 5000 larvae per female can be collected from the tank. The disadvantages of this method are;
The disadvantage of semi-artificial reproduction was recognized in an early stage and at Kibos this technique was replaced by artificial reproduction. The artificial reproduction of the African catfish, as for all fin-fishes, is a chain of activities which is more or less similar to that of natural reproduction. It starts with the selection of broodstock from the broodstock ponds after which they are transferred to the holding tank in the hatchery. They are injected with hormones in order to induce the ovulation and the following day the eggs can be stripped from the female. The milt obtained from the male is mixed with the stripped eggs and fertilization take place. After 24–45 hours the eggs hatch and a small yolk sac hatchling is born. After approximate 3–4 day the yolk will be absorbed and further nursing of the larvae's takes place in earthen ponds. The advantage of this technique is:
All the “ripe eggs” can be obtained from a mature female.
The fertilization rate increases considerably.
A specific incubation technique for the fertilized eggs can be chosen which increases the produced number of hatchlings.
At Kibos the fertilized eggs are spread on the bottom of a concrete tank and the 4–5
day old larvae are siphoned off and placed in nursing ponds of 1300 m2. From March 1994
to July 1994 the catfish was reproduced almost weekly and approximately 250,000 larvae of
3–4 days old were harvested.
The artificial reproduction through stripping is carried well at Kibos and the farm managers
obtained a lot of experience. The following remarks are however to be made:
Incubating the fertilized on the bottom of a concrete basin results in the fact that hatchlings and dead eggs/egg shells can not be separated. The production decreases as the hatchlings will be infected with fungi which develops quickly on the dead eggs.
The number of stripped females is not in accordance with the needed number of three day old larvae. This results in under stocking of the nursing ponds.
2 Details on this technique can be found in the training manual written by the consultant.
Three factor are of importance once nursing takes place in earthen ponds;
The availability of large quantities of zooplankton.
The stocking density of the of the 3 day old larvae.
The duration of the rearing period.
A well fertilized pond, stocked with 100 larvae per square meter which is harvested after 30–40 days, produces 35–40 fingerlings per square meter (de Graaf et al., in press, see training manual). The production will decrease considerably if one of the factors is not well managed.
Table 2 presents the production figures of Kibos (the only station with production data3).
Pond no | Total no stocked | Total no harvested | No stocked (no/m2) | No harvested (No/m2) | Rearing period (days) | Weight (g) |
B1 | 3,000 | 240 | 2.3 | 0.2 | 30 | 10–12 |
B1 | 20,000 | 12,144 | 15.3 | 9.3 | 36 | 3–5 |
B1 | 10,000 | 8,467 | 7.6 | 6.5 | 44 | ? |
B2 | 6,000 | 1,231 | 4.6 | 0.9 | 38 | big |
B2 | 10,000 | 7,772 | 7.6 | 5.9 | 30 | 15 |
B3 | 3,000 | 1.320 | 2.3 | 1.0 | 26 | big |
B3 | 10,000 | 5,511 | 7.6 | 4.2 | 25 | ? |
B3 | 10,000 | 6,789 | 7.6 | 5.2 | 36 | ? |
B4 | 4,000 | 13,600 | 3.1 | 10.5 | 30 | ? |
B4 | 20,000 | 14,876 | 15.4 | 11.4 | 29 | ? |
B4 | 10,000 | 6,739 | 7.6 | 5.2 | 37 | ? |
B5 | 5,000 | 750 | 3.8 | 0.6 | 31 | ? |
B5 | 20,000 | 4 | 15.3 | 0 | 29 | big |
B5 | 10,000 | 1,410 | 7.6 | 1.1 | 37 | ? |
B6 | 4,000 | 312 | 3.1 | 0.2 | 37 | ? |
B6 | 20,000 | 28,994 | 15.3 | 22,3 | 34 | 2–3 |
B6 | 58,000 | 7,989 | 44,6 | 6.1 | 26 | ? |
At Kibos on the average the ponds are stocked with 10 larvae/m2, the rearing period is 33 days and 5.3 fingerlings are harvested per square meter. The harvested fingerlings are relatively large and the production is unreliable as it varies from 0 to 22,3 fingerlings per square meter. The results are typical for an “early stage” nursery operation with unexperienced farm managers and similar events were observed in Congo-Brazzaville and in Ivory Coast (de Graaf et al., in press, table 2, de Graaf and Galemoni, 1989). The main problems are;
The stocking density of the ponds is much to low and the method used for the estimation of the stocked numbers is in adequate.
The quantity of zooplankton is to low as the ponds are fertilized 1 day before stocking. With the actual used stocking density the production is still reasonable but large problems will be encountered once the stocking density is increased to 100 larvae/m2.
The ponds are rather large (1300 m2) which makes careful management difficult and it reduces the ratio dike length/pond area. The latter proofed to be of importance for the survival rate of the larvae as the larvae feed or hide in the submerged vegetation of the dike during the first week of the nursing phase.
At the FPC of Yala till present only semi-artificial reproduction has been carried out. The results were reasonable (5000–6000 larvae per reproduced female) but were not sufficient in order to stock the nursing ponds with the appropriate densities (100 larvae per square meter). The following was observed at Yala;
The water quality of the spring water is good (no siltation). However a large number of people are using the spring water in the hatchery for domestic use which should be avoided once the tanks are used for the incubation of eggs and larvae.
The nursery ponds are filled to late and not well fertilized
The nursery ponds are under stocked.
Stripping of eggs was demonstrated at Yala. Three females were injected and the following morning approximately 200 grams of eggs could be stripped and fertilized. The eggs were attached to submerged water weeds and incubated in a circular holding tank. Unfortunately the weeds were not cleaned thoroughly before use and a heavy silt load appeared when they were used and consequently the hatching was rather poor. During the training course it was proofed that the weeds can be used for the incubation of the eggs but a disadvantage is that the hatched larvae and the submerged weeds are difficult to separate.
At the FPC of Rongo the catfish has been reproduced semi-artificial and artificial the results until now were limited, high mortalities in the hatchery and during the nursing stage took place. Therefore it was decided that the consultant carried out a reproduction cycle at Rongo. Five females were injected with a mixture of Nile Tilapia pituitaries and Catfish pituitaries. The following morning the eggs were stripped and a happa placed in a concrete tank and nile cabbage was used for the incubation of the eggs. The incubated eggs were treated daily with malachite green (0.1 ppm) and hatching took place after three days. Table 3 presents the key figure of this reproduction trial.
Weight female (g) | Weight stripped eggs (g) | GSI (%) | Hatching percentage (%) |
400 | 50 | 12.5 | 60–70 |
485 | 60 | 14.1 | 60–70 |
390 | 40 | 10.2 | 60–70 |
780 | 30 | 3.8 | 0 |
375 | 25 | 6.6 | 0 |
During the reproduction several problems were encountered;
The main water source of the station is a spring, water is collected in a tank near the spring and transported through underground PVC pipes to the ponds. The pipe was blocked somewhere and had to be dug up and the blockage had to be removed. In general it can be stated that this kind of problems are not encountered in open channels, which could easily be made at Rongo.
The water temperature is rather fluctuation 22 °C at day time and 18–19 °C at night. The females are injected at day time, the temperature is recorded and from a table (see training manual) the needed time till stripping is taken. In our case 22 °C gives 15–16 hour. The temperature however drops at night and the whole ovulation process will slow down somewhat. This phenomenon causes that the females are stripped to early. During the reproduction at Rongo two females did no respond well after 16 hours. The eggs were difficult to strip, we were to early. The stripping was deliberately proceeded in order to demonstrate the results; low number of eggs and poor hatching (see Table 3). It is therefore recommended to strip later as expected from the table and to inject more females as needed (inject 5), as not all females respond properly.
The spring water contains a rather large amount of silt which make it unsuitable for the incubation of fertilized eggs. The silt will stick to the outside of the fertilized eggs and hatching will be very poor.
A second source of water is tap water which is collected in a large reservoir near the hatchery. This water has been used during the reproduction and it worked well. The reservoir and the concrete basin were the happa is place must be completely filled up before reproduction starts. Fertilized eggs are sticked to the roots of Nile cabbage and placed in the happa. The reservoir contains enough water to keep an adequate flow rate through the happa during the incubation of the eggs, which is the critical period at Rongo.
The spring contains wild fish and the hatchery drains in the nursing pond of catfish. In the past tilapia were found at harvest in the nursing pond. These Tilapia either escaped from the hatchery or the were introduced from the spring. Tilapia will eat your catfish larvae it is therefore essential that the two inlets are screened.
At Alupe reproduction has been carried out semi-artificially and the results were similar as to the other FPC's. The water quality seems all right (no siltation problem). It is recommended to use artificial reproduction, with Nile cabbage as an egg substrate and happa's as incubators at Alupe.
Semi artificial as well as artificial reproduction has been carried out with limited results at the FPC of Chewele. The major problems are;
Low and fluctuating water temperatures 18–23 °C. This can be solved by waiting longer before the females are stripped and by injected more females.
A heavy siltation of the used spring water. A sedimentation tank has been constructed near the hatchery and it is recommended to carry out artificial reproduction following the method as proofed to be successful in Rongo, where a similar problem exists.