5.1.1 The areas of Rwanda best suited for cultivation of fish are the flat valley bottoms with very gentle slopes, which lend themselves to easy construction of ponds. Most existing ponds are in such areas.
5.1.2 Water supply would not be a constraint. Records of air temperatures suggest that in some places the temperature of the water may drop below 10° to 12°C for a sufficiently long time to cause mortalities of fish of such species as Tilapia rendalli, T. mossambica and T. nilotica, although the Mission is not aware of any actual incidents of this kind. When water temperatures are below 17° to 18°C in June, July and August, tilapia will not spawn and growth rates will be poor. It will, therefore, have to be accepted that fingerlings will not be available for three months in the year and the main growing season will be from September to the end of May.
5.1.3 The peat soils are cold, acid, deficient in calcium and in other minerals, including phosphorus, and in available nitrogen. Biological productivity is therefore comparatively low, and corrective measures are necessary to achieve good results when cultivating fish or other crops. These measures include use of lime and fertilizers and are discussed in more detail later below.
In some cases drying out adversely affects the mechanical and physico-chemical properties of the peat: it becomes irreversibly compressed and hydrophobic; it would be necessary to maintain the water table as high as possible to avoid these problems.
In some cases the content of sulphur compounds is high enough to make the soil acid and unsuitable for nearly all sorts of cultivation, but if fish ponds were constructed in such soils, the condition could be corrected by repeated flushing of the ponds to leach out the sulphates.
The very fibrous peaty soils formed from reeds or Cyperacae are not suitable for fish farming.
An account of the various types of peaty soil encountered in Rwanda, the methods used to reclaim them for cultivation and the appropriate treatments to enhance or retain fertility can be found in Annex 3. Before constructing fish ponds, a detailed survey including soil analysis is advisable.
5.1.4 Both malaria and schistosomiasis occur in Rwanda. The marshalands provide habitats suitable for mosquito larvae and bilharzia snails and so do fish cultivation ponds, but in the latter case it is possible by various means to control the concentrations of these animals at a low level so that the risk of contracting disease is low. See 5.3 below.
5.1.5 The natural productivity of waters in Rwanda is between 100 kg/year and 300 kg/year/ha. To achieve a better production it is necessary to feed the farmed fish, or to fertilize the ponds, or both; species of fish that respond to this treatment by growing fast and converting the available food efficiently have to be chosen. This in turn means, in practice, that farmers would have to be supplied with fingerlings of these species, at least for initial stocking. These matters are discussed in detail later below. First, the systems of fish culture that appear worthy of consideration for small-scale rural enterprises in Rwanda will be briefly described.
5.2.1 Systems of fish cultivation appropriate to Rwandan conditions would include monoculture of Tilapia nilotica; polyculture with T. nilotica as the main species (the secondary species including other tilapias, Clarias spp. as predators, and snail-eating species such as Haplochromis mellandi and Astatoreochromis alluaudi); culture of fish in association with husbandry of animals such as pigs, ducks or chickens; perhaps rice-cum-fish culture; and perhaps also cage culture with fully artificial feeding. The considerations governing choice are set forth below.
5.2.2 Monoculture
At the initial stage of rehabilitation of the existing ponds, and as farmers are not yet trained and not well accustomed to fish farming, it is foreseen that they would first raise T. nilotica in monoculture, going progressively over to polyculture and/or integration of fish farming and animal husbandry (pig-cum-fish culture).
5.2.3 Polyculture
Natural food in ponds (increased by fertilization) and supplemental feeds are not completely consumed if only one species is cultured in the pond and better yields are obtained with polyculture. Choice of species is discussed in 5.3 below. Suggested combinations and stocking rates are in Annex 4.
5.2.4 Integration of fish farming and animal husbandry
An effective and economical system of fish production involves the raising of pigs, ducks or chickens in pens beside or over fish ponds to provide a source of continuous organic fertilization. It works very well in polyculture of T. nilotica and common carp and also with other combinations. This practice increases the efficiency and profitability of both animal husbandry and fish culture by utilizing animal and feed wastes and can produce yields higher than any other system of fish farming.
Details of systems of this kind are given in Annex 5. None have yet been tried in Rwandan conditions, and they should be the subject of experimental trials in due course when resources permit.
5.2.5 Rice-cum-fish culture and cage culture of fish with full artificial feeding have likewise not been tried in Rwandan conditions and should be the subject of experimental trials as and when sufficient resources and experience become available.
5.3.1 The main species cultured at present in Rwanda are Tilapia melanopleura (or T. rendalli) and T. macrochir. Tilapia nilotica and common carp are also raised in ponds, but on a very small scale. Tilapia macrochir and T. rendalli are not fast growing species and experience elsewhere in Africa suggests that incremental growth rates are generally only 0.5 g/day and never exceed 1 g/day, even with fertilization of the ponds and artificial feeding. Clarias carsonii occurs sometimes in ponds, but it is a relatively slow-growing species which never reaches more than 20 cm. These species should only be considered suitable for farming in polyculture with fast-growing species, such as T. nilotica, common carp (Cyprinus carpio), Clarias mossambica, C. lazera. Eventually, grass carp (Ctenopharyngodon idella) and silver carp (Hypophtalmichthys molitrix) can be raised if these Chinese carps can be acclimatized in Rwanda, and if adequate quantities of fingerlings can be produced.
5.3.2 As a predator on undesirable animal organisms C. mossambica can be used, but if natural spawning does not occur in ponds or if induced spawning is not successful, one could utilize instead C. lazera which has been cultured successfully in the Central African Republic and in the Ivory Coast.
5.3.3 Growth rate increments under good conditions (fertilization of the ponds and artificial feeding) can be expected from experience elsewhere in Africa to be between 1 and 2 g/day for T. nilotica; 2–4 g/day for C. lazera; 8–14 g/day for common carp; and 6–8 g/day for silver and grass carp.
5.3.4 As bilharzia (schistosomiasis) is a common disease in Rwanda all fish ponds, dams, rice fields and swamps should be stocked with a few of the malacophageous species Astatoreochromis alluaudi and Haplochromis mellandi (Vincke, 1972). Some other species, which are not snail eaters, but selective plankton eaters or feeding only on periphyton (biota growing on stems and leaves of submerged aquatic weeds) are also effective in reducing the number of snails because they destory the eggs which adhere to the stems of submerged plants, such as Potamogeton, Myriophyllum and Ceratophyllum, to the lower face of the floating leaves of Nymphea, Azolla, Pistia stratiotes or Salvinia natana, or to the roots and leaves of the water hyacinth (Eichornia crassipes). Species feeding on periphyton are T. nilotica, T. mossambica and T. zilii. Herbivores such as T. rendalli and grass carp also destroy eggs when they eat aquatic plants.
5.4.1 One of the main constraints to the expansion of fish production by aquaculture is the poor availability of fingerlings of the species most suitable for culture (5.3 above).
5.4.2 Tilapia nilotica fingerlings have been available only at the Ruganwa Fish Station near Kigali, but arrangements were in hand to transfer some to Kigembe in October 1980. A few fingerlings of T. rendalli and T. macrochir are available at Kigembe.
5.4.3 An outline of the techniques that would be required for successful and economical production of fingerlings can be found in Annex 6. The sizes of ponds in the planned breeding stations (4.3.6) seem too large. The requirements for fingerlings to support a planned programme of development are indicated in 6 below.
5.5.1 Supplementary feeding will enhance the growth rates of farmed fish and increase productivity per unit area of pond. Whether it is economic, however, depends upon the availability of suitable raw materials and their costs, including costs of transport and of any processing to which they have to be subjected to make them suitable as feed stuffs for fish.
5.5.2 The need to include animal protein in the supplementary feeds depends upon species and its inclusion is not considered essential when cultivating tilapias and Clarias.
5.5.3 Table 7 lists some agricultural and industrial waste products found in Rwanda that could possibly be used as raw materials for fish feeds. The table also indicates the quantities produced and where and at what time of year they are available. It is not clear, however, to what extent some or all of these by-products and wastes are already utilized as animal feeds, agricultural fertilizers, as domestic fuel or for other purposes. To the extent that they are already so used, their diversion to fish farming would bear an opportunity cost. However, to use them to produce animal protein should, in theory, be more efficient than using them as fertilizer or fuel.
Table 7
Availability of Agro-Industrial By-Products
| Agro-industrial by-products | Localities | Availability (tons/year) | Period of availability | Cost (Rw.F.) |
| Draff (wet) | Gisenyia | 1 400 | all year | free of charge |
| Beer yeast | Gisenyia | 200 000 litres | all year | free of charge |
| Draff from banana beer | all countries | 87 000 | all year | 2/kg |
| Draff from sorghum beer | all countries | 20 800 | all year | 2/kg |
| Blood meal | b | 390 | all year | free of charge |
| Stomach contents | b | 800 | all year | free of charge |
| Bone meal | b | c | all year | free of charge |
| Rice brans | Kigali | 74 | March, Sept.-Oct. | 2 000/ton |
| Rice polishings | Kigali | 149 | March, Sept.-Oct. | 2 000/ton |
| Wheat brans | Ruhengerid | 193 | all year | 2 000/ton |
| Remouldings (wheat) | Ruhengerid | 262 | all year | 2 000–2 500/ton |
| Bagasse | Kabuye | 2 100 | all year | free of charge |
| Molasses | Kabuye | 350 | all year | to be assessed |
| Filter press mud (cane) | Kabuye | c | all year | to be assessed |
| Maize bran | all countries | e | all year | 2 000/ton |
| Groundnut cake | f | c | all year | to be assessed |
| Soya oil cake | Kigali | 10 000 | all year | to be assessed |
| Coffee pulp | all countries | 100 000 | all year | to be assessed |
| Coffee hulls | all countries | 3 000 | all year | to be assessed |
| Cassava offals | all countries | c | all year | to be assessed |
| Household scraps | all countries | c | all year | to be assessed |
| Other wastes | all countries | c | all year | to be assessed |
Notes: a A new brewery will be established presently in Kigali.
b Available in the three abattoirs (Kigali, Butare and Rusumu),
c 28 slaughterhouses and 276 unimproved slaughter places.
d Figure not available. Quantities should be assessed.
e A new wheat mill will be installed soon at Adikongoro.
f Available in almost every village, but in small quantities.
g Available in Cyangugu, Kibuye, Gisenyi, Gitarama and Kigali prefectures.
Annex 7 discusses the treatment and utilization of these materials at some length. Annex 8 indicates how suitable diets might be formulated from raw materials available in Rwanda; the 25 percent protein diet should be used for fingerlings and adults; the 30 percent protein diet for fry.
The extent to which such diets can be or should be used in fish farming in Rwanda will be a matter for practical trial. One constraint may be local availability: if such materials have to be transported more than a certain distance (perhaps 50 km) their utilization in fish farming may be uneconomic.
5.6.1 Lime will have to be used to correct the acidity of peat soils and of the water in areas where they occur.
5.6.2 To obtain high yields with T. nilotica, T. macrochir and common carp (C. carpio), fertilization of the ponds has to be carried out regularly.
Either organic or inorganic fertilizers may be used, but recent increases in the price of chemical fertilizers may greatly limit their use in Rwanda, taking also into account local transport problems, especially in remote areas.
Inorganic fertilizers include superphosphates, ammonium compounds and urea. Organic fertilizers might include animal manures, vegetable compost and artificial manure; the last named is made by composting vegetable matter and adding urea or some other source of nitrogen.
Details of the use of various fertilizers and, in appropriate cases, their manufacture, are given in Annex 9.
5.7.1 In estimating the likely yield per hectare from small-scale rural ponds in Rwanda, experience has been taken into account from elsewhere in Africa and in other tropical regions where fish farming of the species under consideration is already established. Yields of 3 to 3.5 t/ha/year are attainable in rural small-scale aquaculture by skilled management cultivating the right species; an average yield when culturing T. nilotica might be 1.8 to 2.0 t/ha/year.
5.7.2 The assumption made later below (Chapter 6), namely an average yield of 1.5 t/ha/year, therefore seems conservative. However, it takes into account the fact that performance is found to vary from 0.7 to 0.8 t/ha/year or less with comparatively unskilled management of the ponds and no fertilization or feeding, to 5.5 t/ha/year and more for skilled management of large ponds with regular fertilization and daily feeding of fish.
5.7.3 It would take some years to acquire sufficient knowledge, experience and skill to attain such a high level of performance as 5.5 t/ha/year, and only a minority of farmers can be expected to do so - perhaps one in eight or one in ten. One-third to one-half of the farmers can be expected to achieve an average of 2.4 t/ha/year in due course; about the same number will probably never average more than 0.7 to 0.8 t/ha/year.
