In order to determine which culture system would be most suitable in a given situation, a number of selection criteria have to be balanced. The characteristics of each culture system, like species, management level, inputs, consumer preferences etc. are strongly related to each other.
From the ALCOM survey data it is clear that fish farming in Lubombo region is a low-level activity. Farmers with ponds constructed these often for other purposes and practise fish farming only as a side activity at a very extensive level.
Clarias is the most preferred species. The interviews and discussions however indicate that the size of fresh fish may be more important than the species.
From the species cultured in Swaziland, clarias, carp, and tilapia are adapted to the climatic conditions of the Lubombo region. The growth of tilapia and clarias however, will be restricted to the 7–8 warmer months of the year.
With good management, carp and clarias can reach an acceptable size, starting from fingerling, during this period. Monosex culture has to be practised for tilapia to reach an acceptable size. There are three ways of obtaining a monosex tilapia stock, which all have their complications and limitations. With manual sexing there is always the human error factor and an expensive wastage of female fish. With cross-breeding, a pure parental stock must be maintained that consistently produces 100% male off-spring, although low fertility and poor spawning can occur. Hormone treatments require facilities for holding large numbers of fry and trained technicians.
Since fish farming in Swaziland is at an early stage, and the Nyetane Fish Farm is not operational, import of fingerlings should be continued for the next few years.
The production of large fish requires a fairly intensive management. On-farm inputs are probably not sufficiently available to support this management level (no pigs, few chicken, small farms). Since those farmers involved in fish farming are already more oriented towards sale of livestock and crops rather than towards subsistence farming, some of them might be in a position to purchase fish feeds. These feeds are locally available. A high percentage of the population in the homesteads are job seekers or not employed full-time in farming activities. A commercially oriented form of fish farming could be an alternative income generating activity for this group.
The culture of clarias has been well documented from experiences in different regions of Africa. In South Africa, the eastern Transvaal Regions are considered to be the prime areas for the culture of clarias, because of ideal climate and water conditions and the proximity to large markets. Lubombo region has climatic and water conditions which are similar to these regions.
The Sand River Reservoir has been commercially fished from August 1975 until March 1979, and from September 1987 until 1989. The purpose of the fishing was to determine the potential of fresh water fish in the market. The result was that although there was some initial resistance in Swaziland, markets could be found. However, once there was irregularity in the supply buyers would divert to other sources of fish. It was also found that the sale of clarias was very difficult, due to its appearance. Whole clarias was bought predominantly by Mozambicans only. This information conflicts with the results from the ALCOM survey where clarias was mentioned as the most preferred species. Carp was not being caught in the reservoir. Further study on fish preferences and market opportunities is needed.
Clarias can be cultivated using different culture systems; intensively in tanks or cages; semi-intensively in ponds; in polyculture with extensive management; and extensively in monoculture.
4.1.1 Intensive culture
In intensive culture the fish are kept in high densities in tanks or cages and fed with a balanced feed which meets all the nutritional requirements of the fish. Swaziland Milling in Manzini imports a pellet which is suitable for clarias culture (see Table 14).
Table 14: Composition of Farmix Trout Pellets
| Ingredient | Concentration | |
| Protein | > 38% | |
| Moisture | < 12% | |
| Fat | > 6% | |
| Fibre | < 4% | |
| Calcium | < 1.2% | |
| Phosphorus | > 0.7% | |
| Price: E76.74/ 50 kg. El.54/kg. | ||
Although C. gariepinus tolerates a wide temperature range, good growth is achieved only in a narrow range. Smaller fish achieve their maximum growth between 27.5 and 32.5 °C, temperatures of 25 °C still give an acceptable growth. Larger fish have their optimum between 25 and 27.5 °C, growth reduces markedly at below 20 and above 35°C.
For intensive culture of clarias, fed with a high quality balanced feed, the optimum feeding levels and the expected corresponding specific growth rates are calculated by Hogendoorn et al (1983) and given in Table 15.
| Body weight (gr) | ||||||
| Temp | 1 | 5 | 25 | 50 | 100 | 200 |
| 20 | 2.9/3.1 | 1.9/2.6 | 1.2/1.5 | 1.0/1.1 | 0.9/0.6 | |
| 21 | 3.6/4.2 | 2.5/3.6 | 1.7/2.3 | 1.4/1.7 | 1.2/1.0 | 1.0/0.4 |
| 22 | 4.4/5.3 | 3.1/4.6 | 2.2/3.0 | 1.9/2.3 | 1.6/1.4 | 1.4/0.6 |
| 23 | 5.1/6.3 | 3.7/5.6 | 2.6/3.8 | 2.3/2.9 | 2.0/1.9 | 1.7/0.9 |
| 24 | 5.8/7.2 | 4.2/6.5 | 3.1/4.4 | 2.7/3.4 | 2.3/2.2 | 2.0/1.1 |
| 25 | 6.5/8.0 | 4.7/7.3 | 3.4/5.0 | 3.0/3.8 | 2.6/2.6 | 2.3/1.2 |
| 26 | 7.0/8.7 | 5.1/7.9 | 3.7/5.4 | 3.3/4.2 | 2.8/2.8 | 2.5/1.4 |
| 27 | 7.4/9.2 | 5.4/8.3 | 3.9/5.7 | 3.4/4.4 | 3.0/2.9 | 2.6/1.4 |
| 28 | 7.7/9.6 | 5.6/8.5 | 4.0/5.8 | 3.5/4.4 | 3.0/2.9 | 2.6/1.4 |
| 29 | 7.9/9.7 | 5.6/8.6 | 4.0/5.7 | 3.5/4.3 | 3.0/2.8 | 2.6/1.3 |
| 30 | 8.0/9.7 | 5.6/8.5 | 3.9/5.6 | 3.4/4.2 | 2.9/2.7 | 2.5/1.2 |
| 31 | 8.0/9.6 | 5.5/8.5 | 3.8/5.2 | 3.2/3.9 | 2.7/2.4 | 2.3/1.1 |
| 32 | 7.9/9.3 | 5.3/7.7 | 3.6/4.8 | 3.0/3.5 | 2.6/2.2 | |
| 33 | 7.8/8.8 | 5.1/7.2 | 3.4/4.3 | 2.8/3.1 | ||
| 34 | 7.5/8.2 | 4.8/6.5 | ||||
| 35 | 7.2/7.5 | 4.5/5.8 | ||||
(Source: Hogendoorn et al, 1983)
The water temperatures at the Nyetane Fish Farm are given in Table 1. From the figures of both tables the expected growth of C. gariepinus at the Nyetane Fish Farm can be calculated. Nyetane Fish Farm is situated in the lowveld which is the warmest region of Lubombo District. Some farmers have their farms in the lowveld but others live in higher, cooler, and thus for fish farming less favourable regions.
Assuming that the growth will be zero at water temperatures below 20°C, growth will take place only 8 months per year. In particular the growth of the smaller fish will be below the optimum throughout the year.
When the culture starts with fingerlings of 5 grammes in October a final weight of over 700 grammes can be expected in May. For calculations see Table 16.
The overall feed conversion rate of 1.81 is not very promising. It is caused by the high conversion rates when larger fish is cultured at low temperatures. If the culture is ended at the end of April at a final weight of 655 gr. the overall FCR is 1.53.
Table 16: Growth calculations of C. gariepinus at Nyetane Fish Farm with optimum feeding
| Temp | begin weight | growth rate | days | final weight | gain gr. | FCR* | feed gr. | |
| Oct | 20 | 5 | 2.6 | 30 | 11 | 6 | 0.73 | 4 |
| Nov | 22 | 11 | 3.8 | 30 | 34 | 23 | 0.73 | 17 |
| Dec | 24 | 34 | 4.4 | 10 | 52 | 18 | 0.70 | 13 |
| 52 | 3.4 | 20 | 101 | 49 | 0.79 | 39 | ||
| Jan | 25 | 101 | 2.6 | 30 | 218 | 117 | 1.00 | 117 |
| Feb | 26 | 218 | 1.4 | 30 | 330 | 112 | 1.79 | 201 |
| Mar | 25 | 330 | 1.2 | 30 | 472 | 142 | 1.92 | 273 |
| Apr | 24 | 472 | 1.1 | 30 | 655 | 183 | 1.82 | 333 |
| May | 21 | 655 | 0.4 | 30 | 738 | 83 | 4.00 | 332 |
| Total: | 240 | 733 | 1.81 | 1328 |
It should be noted that in Table 15 only growth rates and feeding levels are given for fish up to 200 grammes. In the growth calculation these figures have also been used for larger fish. The specific growth rate however decreases for larger fish, and the FCR increases. This is caused by a higher percentage of dry matter and fat in larger fish. This means that the fish weights after February are probably overestimated and the overall FCR underestimated.
The calculation is merely academic. Most farmers in Swaziland are not in a position to culture fish in a system which would require such intensive management, including optimal feeding, water quality control etc. The calculation however shows the effects of applying a balanced feed under Nyetane climatic conditions. Although a local farmer will get lower results the FCR of 1.53 between October and May can be assumed to be realistic, or even underestimated.
4.1.2 Semi-intensive culture in ponds. Viveen et al (1986) give an example of semi-intensive clarias culture in ponds, using a high quality feed (30% digestible protein) and stocking densities of 1000 fingerlings of 1–3 grammes per 100m2. After a period of 25 weeks he obtains a survival of 50%, and a final individual weight of 200 grammes. In total 140 kg. of feed was needed per are.
4.1.3 Polyculture
Clarias can be reared in polyculture with tilapia. The clarias predates on the offspring of the tilapia, through which it receives high protein food and the tilapia population does not get stunted. The primary production of the pond is important for both the tilapia and the clarias and can be increased by fertilizing the pond. In order to obtain good size fish extra feeding of the pond will still be necessary. However, since the fish receives natural food from the pond and the clarias predates on the tilapia fingerlings, feeds which are high in carbohydrates and fats, such as maize bran, are sufficient. The Dairy Board in Matsapha can supply maize bran and sunflower cake.
Viveen also gives an example of this culture system. A pond of 100m2 was stocked with 300 clarias fingerlings of 1–3 grammes, and 200 tilapia fingerlings of 5–15 grammes. Total feed provided was 75 kg of cotton cake, and fertilization was done with 150 kg of cattle manure. After 25 weeks he obtained a 65% survival with clarias and an individual weight of 184 grammes. The tilapia weighed 80 grammes with a survival of 100%, meaning that reproduction balanced mortality.
4.1.4 Extensive culture
Little is known about the results of this culture method, and it is questionable if this system is appropriate under Swaziland conditions, with moderate temperatures, limited input resources, and a market preference for large fish.
4.1.5 Economics
South Africa has an abundance of fish and part of its surplus is exported to Swaziland. Sample prices for cleaned fresh fish are given in Table 17. Hawkers buy the fish at the Sand River Reservoir and transport it to the towns, where bream weighing 250–350 gr sells at El. 50–3.00 each. At one time hawkers were buying 200–300 kg per week.
Any aquaculture project has to compete with these prices. Aquaculture in Swaziland was introduced at the subsistence level, the primary objective being household nutrition, not economic viability.
The Fisheries Section imported in 1990 clarias and carp fingerlings at a price of E0.20/piece plus E150 for transport to Mbabane. Transport costs from Mbabane to the farmers are high in those cases where the farmer needs only a limited amount of fingerlings. A price of E0.30 per fingerling delivered to the farmer is used in the following calculations.
Table 17: Fish prices in Swaziland
| Source | Fish | price E/kg. |
| OK Mbabane (April 1991) | Hake fillet | 24.65 |
| Ocean Bream Fillet | 8.50 | |
| Spar Mbabane (April 1991) | Kingklip steak | 32.40 |
| Kingklip fillet | 27.90 | |
| Haddock steak | 15.10 | |
| Hake fillet | 13.80 | |
| Hake cutlets | 10.50 | |
| Whole hake (200 gr/pc) | 4.60 | |
| Pilchards head off (200 gr/pc) | 4.10 | |
| Mixes fish (200 gr/pc) | 3.30 | |
| Sand River Reservoir Fisheries (1989) | Bream | 2.10 |
| Bream head off | 2.60 | |
| Yellow fish | 2.00 | |
| Clarias | 2.00 | |
| SSR Hawkers (1989) | Bream | 7.50 |
The influence of the cost for fingerlings on the production cost of one kilogramme produced fish increases when the final weight of the fish decreases (Table 18).
Table 18: Cost of fingerlings per kg of produced fish at different final weights
| final weight gr | survival % | cost fingerlings/kg produced fish (E) |
| 200 | 65 | 2.30 |
| 400 | 60 | 1.25 |
| 600 | 55 | 0.91 |
| 800 | 50 | 0.75 |
| 1000 | 50 | 0.60 |
The FCR, and consequently the cost for feed per kilogramme of produced fish, increases with increasing body weight of the fish. Table 19 gives the cost for feed, without transport cost, and the cost for feed plus fingerlings per kilogramme of fish produced in intensive culture under Nyetane Fish Farm conditions.
Table 19: Feed and fingerling cost per kg of produced fish at different final weights.
| final weight gr | overall FCR | cost feed E/kg fish | cost fing. E/kg fish | feed+fing E/kg fish |
| 100 | 1.00 | 1.54 | 4.29 | 5.83 |
| 200 | 1.00 | 1.54 | 2.30 | 3.84 |
| 400 | 1.32 | 2.03 | 1.25 | 3.28 |
| 650 | 1.53 | 2.36 | 0.83 | 3.19 |
| 740 | 1.81 | 2.79 | 0.81 | 3.60 |
The figures in this table are different for a large scale fish farm, which can produce its own fish feed, reduce the cost of fingerlings by producing into own in large quantities, and can sell its product at a high value market. The figures indicate that intensive and semi-intensive fish farming is not feasible for small scale farmers. It should be stressed that in intensive culture the costs for feed and fingerlings normally account for 40–50% of the total production costs only.
In a polyculture system of clarias and tilapia the farmer can produce his own tilapia fingerlings. The natural production of the pond can be increased by fertilization, and high quality feed is not a must. Therefore production costs can be reduced. Since the cost for fingerlings remains unchanged the final weight of the product should be well over 200 grammes, and therefore the stocking rate for clarias should be much lower than that proposed by Viveen et al.
4.2.1 Intensive culture
The optimum temperature for the culture of carp is lower than the one for the culture of clarias. Therefore under Nyetane climatic conditions the optimum feeding level and consequently the growth rate will be higher. From October until May carp can grow with optimum feeding from 2 up to 1000 grammes. Growth can also be expected during the cooler period of the year. Optimum feeding levels for a high value feed at 23°C are given in Table 20.
Table 20: Optimum feeding levels and expected growth for carp at 23°C
| weight gr. | feeding level % | period weeks |
| 2–7.5 | 10 | 0–4 |
| 7.5–17 | 8 | 4–6 |
| 17–30 | 7 | 6–9 |
| 30–100 | 5 | 9–14 |
| 100–1000 | 4 | 14–32 |
Although the growth of carp will be faster than that of clarias, the FCR will not be lower. For the growth range as mentioned in Table 20, an overall FCR of 1.7 can be assumed. The prices the Fisheries Section pays for fingerlings of carp and for clarias are the same. This means that for the culture of carp under intensive conditions the costs for fingerlings and feed will be similar to these costs for clarias culture. In view of the market prices for other fish in Swaziland it is questionable if intensive carp culture can be made economically feasible.
The cost for fingerlings can be reduced once they are produced in the vicinity of farmers, e.g. when the Nyetane Fish Farm is operational. Production of fingerlings is feasible only when large quantities are required. It can be expected that the break even point for the production of fingerlings is reached at a smaller number for carp than for clarias.
4.2.2 Extensive culture
In extensive culture the growth of carp will be less than the growth mentioned in Table 20. Since the price of carp fingerlings is equal to the price of clarias fingerlings, the cost for fingerling per kilogramme produced fish as mentioned in Table 18 can also be used for carp. In order to reach a high final weight it is necessary, in those cases where the amount of feed is limited, to culture carp over two seasons. An example of this culture is given in Table 21.
Table 21: Two year production of carp
| stocking/are | harvest/are | |||
| no | weight gr | no | weight gr | |
| year 1 | 100 | 2 | 60 | 250 |
| year 2 | 30 | 250 | 25 | 1000 |
For the second season the pond surface has to be twice the size of the first season. A farmer needs at least two ponds for this culture system.
The examples of the culture systems and production given above should be seen as guidelines only. In order to get realistic information about the feasibility of the culture systems, trials under field conditions are needed. For extensive or semi-intensive systems whereby the on-farm products are used for feeding and fertilization, trials managed by farmers are indispensable. These trials can be complemented by on-station trials.
It is proposed that, if enough interested farmers can be identified, different culture systems should be tested, e.g. a polyculture system with clarias and tilapia, and a monoculture system with carp. These trial protocols should be seen as examples. The actual trials should be identified with the farmer and modified according to the availability of inputs and wishes of the farmer. A trial with all three species could also be an alternative.
4.3.1 Polyculture tilapia-clarias
| Initial stocking: | 200 tilapia/are (10 gr.) 25 clarias/are (2 gr.) |
| Feeding: Fertilization: | to be identified with the farmer 10 kg. cattle manure/are/week or 7 kg. chicken or pig manure/are/week |
| Expected harvest: | 200 tilapia 80 gr. average individual weight 15 clarias average individual weight 500 gr. |
4.3.2 Monoculture carp
| Initial stocking: | as per Table 21 |
| Feeding: Fertilization: | to be identified with the farmer 10 kg. cattle manure/are/week or 7 kg. chicken or pig manure/are/week |
| Expected harvest: | as per Table 21. |
4.4.1 Farmers
From the survey information, farmers were identified who could be contacted for the trials using the criteria:
From the 46 pond owners interviewed during the survey, only 8 ponds of 8 farmers fulfilled the criteria. The Fisheries Assistant Officer in Siteki had additional information and eliminated three more ponds. These three ponds were not operational and could not be repaired. He also had information on ponds which were not included in the survey, and one such pond was added to the list suitable for the trials. A total of six ponds were identified (Table 22), however there might be more suitable ponds which can be identified later.
Table 22: Potential fish farmers for farmer-managed trials
| Name farmer | Area | survey no | pondsize m2 |
| Babe Mpungose | Vuvulane | PO 11 | 140 |
| Sem Mndzebele | Vuvulane | 100 | |
| Steven Simelane | Sitsatsaweni | PO 35 | 80 |
| Mandongolwan | Maphangwane | PO 9 | 144 |
| e Maziya | Lubuli | PO 30 | 180 |
| Soloman Mgometata | Lubuli | PO 37 | 500 |
| Zakewa Ngweya |
Some of the information on these farmers gathered by the survey team is given in Appendix 2.
4.4.2 Trials
If farmer-managed trials are to be successful, the involvement of the farmer is crucial in all stages of the trial. The farmers interested in the trials should be well informed about the different culture systems, the needed inputs, and the expected production. The purpose of fish farming, the desired harvest strategy, the available inputs, and the possibility of purchasing inputs elsewhere should be identified together with the farmer. The extension worker should use a participatory approach in his or her work.
It will be necessary to train the Fisheries Assistant Officer, and preferably the other staff of the Fisheries Section, in extension approaches. This training should be conducted by someone with a social science background, oriented towards extension. A two-week course, including at least one week of practical training and the preparation of a workplan for the coming 6 months, should be given before the field work is started. A one-week course 6 months after the work has started should review the experience gained and discuss the problems encountered during the field work.
