Using Oreochromis Andersonii (Kafue Bream) In Zambia

Background

Generally in Zambia, small-scale farmers use the polyculture system of mixing Oreochromis andersonii, O. macrochir and Tilapia rendalli. This is due to the different feeding habits of the fish and the availability of nutrients at village level. The larger farms practice the monoculture system using either O. andersonii or O. niloticus, and integration and use of waste from pigs or ducks is used to build up the biomass of the ponds. In the north of Zambia the use of O. niloticus has been strongly promoted at small-scale level by the Peace Corps of the United States of America Rural Aquaculture Programme (RAP, see http://zambia.usembassy.gov/zambia/rapp.html).

The Food and Agriculture Organisation of United Nations decided to sponsor trials in various parts of the country in order to clarify the best species for promotion through extension, especially to farmers in irrigation schemes.

Kalimba Farms (see, http://www.thebestofzambia.com/kalimba.html) was identified as a location for part of these trials and it was agreed that a comparison between O. andersonii and O. niloticus would be carried out within the normal farming system practiced on the farm.

This final report compiles the data and the observations during three years of trials (2004- 2007).

Description

Kalimba Farm is situated at an altitude of 1200 m, and is 20 km North East of Lusaka, the capital city of Zambia. It is adjacent to the Ngwerere River that supplies the water to the fishponds. The river drains the eastern domestic water effluents and rainfall from Lusaka and flows into the Chongwe River, which eventually flows into the Zambezi River below the Kariba Dam.

Lusaka, at an altitude of 1300 m, experiences fairly cold winters and otherwise mild temperatures for the rest of the year.

Water temperatures recorded at 50-cm depth in fishponds over the past 15 years are summarized in below.

Origin of the Breeding Stock

O. andersonii was stocked at Kalimba in 1985, originating from the Chilanga Government Fish Farm, which is located 20 km south of Lusaka. The fish were offspring of specimens collected from the Kafue River in 1981 under the auspices of the FAO Project ZAM/79/005 (http://www.fao.org/docrep/field/003/AC086E/AC086E00.HTM).

Since 1985, Kalimba has made a yearly selection of new breeders (+/- 400 males and 2000 females) from its own genetic stock, choosing the fastest growing fish and the morphologic characteristics specific to the original fish. The present strain is the result of 18 years of mass selection within Kalimba Fish Farm only.

O. niloticus was purchased from Chirundu Bream Farm in Chirundu by the Project (GCP/RAF/361/EA) and delivered to Kalimba in December 2003.

The following numbers were delivered: -

1. 20,000 fingerlings, 5 g each
2. 55 males, 200 g each
3. 280 females, 135 g each
4. 2,785 male fingerlings, 35 g each.

The O. niloticus from Chirundu Bream Farm came originally from Baobab Farm in Mombassa and Lake Turkana, Kenya, transiting via the Kariba Bream Farm that started in the early 1980s.

Kalimba Farm was established in 1985 and was initially conceived as a crocodile farm. Integrated aquaculture was also part of the project using both pig and duck natural waste to add nutrients to produce zooplankton and phytoplankton in the water. As of July 2010 there are 8.5 ha of fishponds comprising breeder, grower and production ponds. The farm operates using a mono-sex culture, which is hand selected. The total employment of the farm is around 50 people, split into three sections working on fish, pigs and crocodiles. Each section has a supervisor responsible to the Director. A small reptile park is also attached to the farm to generate tourist and leisure revenue. The duck section was discontinued in 2008 due to the difficulty of obtaining new bloodstock due to the outbreak of H5N1 avian influenza.

The fish farm complex practices integration with pig production and no direct alternative feeding of the fish is used. Three pigsties situated above the inlet channels provide the necessary manure to produce the nutrient base to feed the fish. The water containing the manure is directed manually to the different ponds at the time that the pigpens are being cleaned and is fed into the ponds by gravity. The number of pigs per pond varies due to seasonal variations in temperature that can affect the oxygen levels in the ponds as no artificial aeration is used. The pigs are mainly sold to processing companies but some are processed on the farm to provide food for the reptile park. Fish at present account for 13.8 per cent of farm turnover and pigs, 22.5 per cent, but this varies from year to year depending on the fluctuations in the market for crocodile skins, which is fashion driven. Approximately 32 tons of fish and 380 pigs are produced per annum.

The ponds complex is organized into breeding ponds, nursery ponds and production ponds and is designed to be self-sufficient in fish seed production all the year round.

The breeding ponds are stocked with farm-selected breeders every year in August. A sex ratio of five female to one male is used at a stocking density of 30 fish per 100 m^². The fry/fingerlings are harvested every month at a weight of around 3g from November until May of the following year. In June the ponds are drained and rested for two months before restocking for the following season. From time to time, usually around every 2 years, sediments are removed and the ponds are limed for hygienic reasons.

The nursery ponds accommodate the fry/fingerlings harvested from the breeding ponds and are stocked at a high density of 30,000 per 100 m^² until they reach a size at which they can be sexed, 15 to 20 g, which normally takes 2-3 months.

The production ponds are stocked with males at a rate of only 250 per 100 m^². The 20-g fingerlings are manually sexed.

This method has been used during the species trials of O. andersonii and O. niloticus. A couple of mixed-sex ponds for each species have been stocked for production as well, to compare production using this alternative method.

The trials took place in a complex of 16 ponds with a total area of 2.44 ha.

The trials were carried out over three years, from January 2004 to January 2007.

From May 2006 to October 2006, a final-year, University of Zambia student from the Animal Science Faculty (http://www.unza.zm/index.php?option=com_content&task=view&id=359&Itemid=452) closely monitored six production ponds. During that period, temperature readings, more regular sampling, and different feeding methods were used. These activities were reported in his thesis for the completion of a Bachelor Degree in Agriculture Science.

The following was achieved:

• Three breeding cycles for each species were completed between December 2003 and July 2006.
• Six nursery ponds for each species were stocked in December 2003 and November 2006. The monitoring of these ponds was not accurate and record keeping was poor.
• Thirteen complete production cycles were concluded for each of the species. These involved eleven mono-sex male ponds and two mixed-sex ponds. The ponds given to these trials were, as far as possible, stocked simultaneously with each of the two species. Regular samples were taken to monitor growth rates of each species.
• Towards the end of the trial period another student from the University of Zambia joined us to carry out closer monitoring. Six production ponds, three for each species, were assigned to the project to compare growth, but on this occasion three different feeding methods were used.
  • Pig manure only
  • Pig manure followed by supplementary feeding with floating pellets of 18 per cent protein.
  • Complete artificial feeding with floating pellets of 18 per cent protein.
  • All artificial feeds were donated by Tiger Feeds Ltd.
• During the period, the frequency of the test weighing sampling was increased from once-a-month to twice a month under the supervision of the student. The water temperature was checked on a daily basis. The UNZA student could not be present up to the conclusion of the cycle and his report and conclusions do not include the end results of these six ponds. However, a French student from Angers Agricultural College spent three months on the farm from September to December 2006 and at the final harvest she tabulated the results from all of the ponds in relation to length and weight of the two species.

Results

A greater number of fry were harvested from the O. andersonii ponds than from the O. niloticus ponds.

When the respective breeding ponds were being stocked every August, breeding activity was observed in O. andersonii, whilst there was no such activity observed in O. niloticus ponds.

This would appear to show that O. andersonii breeds earlier in the season and probably breeds at a lower temperature than O. niloticus. This would also explain the lower number of fry collected in the O. niloticus ponds as their breeding season would be shorter if they breed at a higher water temperature than O. andersonii.

The survival rate of the fingerlings indicates that the management of the nursery ponds was poor. The ponds were not drained for a complete year. The survival rate of O. niloticus was better than that of O. andersonii, which, to a certain extent, would compensate for their different breeding rate.

On a practical note, Kalimba staff found O. niloticus more difficult to sex manually than O. andersonii. The fish were more evasive and difficult to catch within the net.

The overall average yield in the production ponds for these three years was better for O.niloticus, 5322 kg/ha/year against 4920 kg/ha/year for O.andersonii. This result was not reflected in the interim samples, which generally showed O. andersonii ahead of O.niloticus. One of the major reasons was the higher loss of numbers (probably due to poaching) in the O. andersonii ponds compared with in the O. niloticus ponds.

These observations were confirmed in the past six months’ trials that were monitored by the two students. O. andersonii showed a better growth rate compared with that of O.niloticus all the way through the sampling that was done every two weeks, except for the comparison of manuring followed by pellet feeding in F13 with O. niloticus and F12 with O. andersonii. The main points from these last samples were:

1. O. niloticus was more difficult to sample as they were evading the net.
2. The samples were taken during the cold season and demonstrated that temperature had less affect on the growth rate of O. andersonii than on that of O. niloticus.
3. O. niloticus accelerated its growth rate once the water temperature increased above 20°C.
4. O. andersonii responded well to artificial feed during the cold weather. During that period, O. niloticus showed a zero growth rate.
5. A similar observation was made in the ponds with manure only.
6. The samples taken in F13 indicated that there might have been an accidental stocking by the staff, as there were large numbers of O.andersonii found in a pond where only O. niloticus should have been stocked.

The final harvest of these six ponds showed contrasting results. The overall productivity (net production in tons/ha/year) was higher in the O.niloticus ponds than in those of O.andersonii. During these harvests, a good number of fish were measured for their length and weight.

All fish are sold ex-farm, mostly to women marketers and as live fish for the restaurant trade.

Three levels of pricing are used, which are currently; Large (140g+) K12,000/ kg, Medium (50g) K8000/ kg, Small (20g-) K5000/ kg.

There appears to be little difference in the marketability of the two species but comments were made that O. andersonii was preferred for its taste and color.

Conclusion

The results of this study indicate that there is a definite potential for using O. andersonii for aquaculture in Zambia, particularly at higher altitudes, as its growth and breeding cycles perform better than O. niloticus at low temperature. In fact, its natural geographic distribution in Zambia is in the river systems of higher altitude (Kafue and Upper Zambezi for example). Its growth rate is equivalent to O. niloticus and it has an advantage in marketability. It is an interesting fact that the traditional ruler of the Western Province of Zambia, the region of the Upper Zambezi, is traditionally forbidden from eating any fish other than O. andersonii!

In addition, there is currently considerable focus on environmental issues worldwide and the introduction of non-indigenous species to native waterways has caused much discussion. The further introduction of O. niloticus to fish farms in Zambia will naturally cause an extensive invasion of a non-indigenous fish to the river systems that could have an unknown effect on the naturally occurring species.

There is little doubt that O. andersonii could benefit from a further selective breeding programme such as the GIFT process of collecting wild specimens from geographically different, naturally occurring populations and crossing them. In the process it is likely that O. andersonii could prove to be the most suitable fish for aquaculture in Zambia.