- Standardization of farm management practices: it is essential that all the fry production centres use the same farm management techniques for fry production and grow-out, including fish stocking density, liming, fertilization and supplementary diet feeding regimes.
- Weekend coverage: it is essential that the managers of the fry production centres rotate their staff in such a manner that weekends can be covered. Fish unlike government employees do not have a 5-day working week, and therefore need to be attended 7-days per week if maximum economic production efficiency is to be attained.
- Improved fertilizer application methods: all granular and powdered phosphate-based fertilizers (ie. DAP) should be applied to the fish ponds by either first dissolving in a bucket of water prior to distribution or through gradual dissolution within a floating perforated canister, suspended perforated sack or by placing onto underwater raised wooden-platforms (see Tacon, 1988). If powdered phosphate fertilizers are allowed to come into direct contact with the pond bottom they will become rapidly adsorbed by the soil particles, thus rendering the phosphate unavailable to the water column and consequently reducing its fertilization effect on the resident pond phytoplankton and zooplankton community. The optimum fertilization rate for DAP is generally between 50–60kg/ha applied every two weeks. Ideally, the optimum level should be determined for the LBDA pond conditions through experimentation with varying DAP application rates. However, it must be remembered that whilst the use of chemical fertilizers as pond inputs requires minimal labour and effort, they are extremely expensive commodities (see Table 4) and only act on the autotrophic and grazing food chain by directly stimulating phytoplankton production within the pond (the latter being limited by available solar energy). By contrast, organic manures are readily available and inexpensive, and act on the heterotrophic food chain by supplying organic matter and detritus to the pond ecosystem; the manure serving as a substrate for the growth of bacteria and protozoa and therefore not light limiting. For example, in Israel fish yields within ponds receiving only chemical fertilizer inputs range from 10 to 15kg/ha/day (equivalent to a fish yield of 3.65 – 5.47 tonnes/ha/year) and within ponds receiving both chemical and organic manures reach up to 32kg/ha/day (equivalent to a net fish yield of 11.68 tonnes/ha/year), with no supplemental feeding (fish stocking density 0.93/m2; for details see Schroeder, 1980). Clearly, in view of the universal use of organic manures by the private farmers in the Region, efforts should be made within the fry production centres to also use and test locally available organic manures as additional pond inputs. By so doing the fry production centres would also serve as demonstration centres for interested farmers. For example, although the extensionists from the Project usually advise the private farmers to use cribs within their ponds for manure application, cribs are not present within the fry production centres.
- Improved supplementary diet feeding methods: the success of supplementary diet feeding depends upon the feeding regime (level and frequency) and feed application method employed. For example, published feeding rates (expressed as % body weight/day) vary from 36% for first feeding tilapia fry, down to 12% by day 30, to 6% for fish above 1g, to 3% for 50g fish, and 2% for 100–250g fish when fed a high protein diet (ca. above 45% crude protein; Jauncey and Ross, 1982; see also New, 1988). Clearly, feeding levels and frequency need to be much higher for fast growing fry and fingerlings, and lower for the larger slower growing fish; metabolic rate decreasing with increasing fish size. However, published feeding tables should only be used as a very rough guide, and adjusted according to water temperature, natural food availability, and the response of the fish to the diet fed. Since all fish eat primarily to satisfy their energy requirements (assuming that the diet is palatable) it follows therefore that the feeding level for low energy feeds (ie. cereal based products such as dairy feed mash) should be higher than for high energy feeds such as Omena or Tilapia pellets. A constant feeding level of 3% body weight per day cannot be used for the entire culture operation from fry to market size! The situation is further complicated within the production centres because fry are usually reared within the same pond as their parents. Clearly, fry have to be removed and transferred to separate nursery rearing ponds if feeding rates are to be scientifically implemented. Finally, maximum benefit from diet feeding can only be achieved if the diet is ingested completely by the fish. Therefore to avoid feed wastage it is recommended that different feed presentation methods be tested, including 1) manual feeding by broadcasting over the pond surface (ie. the present feeding method), 2) manual feeding by placing floating feed items into a floating or fixed surface bamboo frame, 3) manual feeding by placing sinking feed items on a submersible feeding tray, and 4) using fixed demand bait-rod feeders (for details see New, 1987 and Tacon, 1988).
- Improved feed storage methods: since feed items are composed of perishable nutrients it is essential that the feed storage period on the farm be kept to a minimum and that adequate storage facilities are provided. Feed should be stored under clean dry ventilated (ie. cool) conditions, and used within two months of manufacture. Bags should be stored on wooden pallets and stacked in such a manner so as to facilitate good air circulation between sacks, and should never be allowed to rest directly against the floor or walls.
- Development of in-house dietary feeding regimes: the results obtained to date using the in-house prepared Tilapia pellet have generally been disappointing. To a large extent this has been due to the low fish stocking densities employed. In contrast to complete diet feeding, the formulation of a supplementary diet is dependent upon the standing crop (ie total biomass) of fish present and the consequent availability of natural food organisms within the pond. For example, no difference was observed in the growth of common carp (Cyprinus carpio) in earthen ponds in Israel when fed a cereal grain (sorghum) or a formulated pellet containing 22.5% protein up to a standing crop of 800kg fish/ha, or between a ration containing 22.5% protein and a ration containing 27.5% protein up to a standing crop of 1400kg/ha (Hepher, 1979). Similarly, no difference was observed in the growth of prawns (Macrobrachium rosenbergii) in outdoor concrete ponds in Thailand when fed a 35% protein diet, a 15% protein diet or a broiler starter feed (stocking density 5 animals/m2; Boonyaratpalin and New, 1982), or in the growth and survival of shrimp (Penaeus vannamei) in earthen ponds in Hawaii receiving no fertilizer and feed input, fertilization input only, or fed a commercial pelleted shrimp ration (stocking density 7.1 – 9.4 animals/m2; Lee and Shleser, 1984). It is important to mention at this stage that many fish species, including Tilapia, have the ability to filter fine particulate matter (ie. phytoplankton and detritus containing bacteria) directly from the water column (see Bowen, 1982). For example, silver carp (H. molitrix) and Tilapia (O. mossambicus) were reported to grow from 15g to 260g and from 10g to 110g, respectively, over a 190 day period without artificial feeding in floating cages within fertilized ponds (Gaiger and Krause, 1983). Similarly, Wannigama, Weerakoon & Muthukumarana (1985) found no significant difference in the growth rate and feed efficiency of cage reared fingerling Tilapia (O. niloticus) when fed a 29% protein diet or a 19% protein diet (containing 92% chicken mash) at stocking densities of 400, 600 and 800 fish/m3.
In view of the above discussion it is perhaps not surprising that the performance of Tilapia fed the 25% protein in-house pellet (costed at Ksh 11/kg) was no different to that of fish fed a 12% protein rice bran mash (costed at Ksh 2.1/kg; see report of Weaver Bird Women's Group Farm; 4.2). Clearly, in view of the low fish stocking densities employed within the fry production centres and private fish farms emphasis must be given on the use of simple low-cost agricultural by-products such as mill sweepings, brewery waste, rice bran, abattoir wastes, and alike.
However, if higher fish stocking densities are to be employed (as would be used for the envisaged cage culture operations and for the future mass production of sex-reversed all-male Tilapia fingerlings in concrete tanks within the centres) there is no doubt that pelleted feeds would be more beneficial and economic in terms of feed efficiency and growth. For example, Table 6 gives some recommended dietary formulations which could be employed and tested at the centres and proposed cage sites if higher fish stocking densities were to be employed (> 5/m2). For example, excellent experimental pond facilities exist at the Kibos fry production centre which could be used for the testing of in-house produced and purchased commercial pelleted Tilapia feeds; apart from the normal fry production ponds, Kibos has 8 ponds of 200m2, 6 of 300m2 and 3 of 500m2.
If experimental in-house pelleted feed is to be produced by the Project then considerable improvements will have to be made to the very basic equipment (ie. hand operated kitchen meat mincer/extruder and grinder) already present at the Kibos fry production centre. This would include the purchase of a disk mill/grinder (10 HP, screen sizes 0.6, 1.2 and 3mm), fish grinder/extruder (3HP, 1–4mm die plates), dough mixer with variable gears (2HP, Hobart), 50 kg weighing scales and top pan electronic balance (0.1–6kg), diet drying cabinet with connected electric air blower, voltage stabilizer, sieves, refrigerator, and chest freezer. The cost of such a laboratory scale feed plant (daily production capacity ca.100kg dry pellets/day) would be approximately US$ 20,000. Alternatively, if larger quantities of feed would be required for field testing (ie. a small-scale feed plant with a maximum production capacity of 500 tonnes/annum) a package farm pelleting system such as that produced by Philco Dierings (UK) Ltd could be purchased, at an estimated landed cost of approximately US $ 50,000.
Justification for the above purchases would, however, depend upon the implementation of higher farm fish stocking densities, the establishment of a rotating fund for the purchase of feed ingredients, the availability of laboratory facilities within LBDA or KMFRI for the chemical analysis of feed ingredients, and finally but not least the allocation by LBDA of a least one staff member to the Project on a full-time basis to cover all aspects of feed formulation, manufacture and feeding.
- In-house training: it is strongly recommended that an in-house training programme be organized by the Project on fish feeds and feeding for LBDA staff, including the managers of the fry production centres, district field coordinators and extensionists. It is essential that government staff be trained on all aspects of fish feed and feeding; food and feeding costs generally representing the largest single operating cost item of fish farming enterprises. In addition, it is recommended that the Project send at least one LBDA/Project staff member for overseas training in fish nutrition and feed manufacture.
- District training programmes: on the completion of the in-house training programme on fish feeds and feeding (including pond fertilization) it is recommended that the district field coordinators and extensionists organize training programmes for interested farmers within their respective districts. For example, of the 18 private fish farms visited during this consultancy, 17 said that they would like to attend a training course on fish feeds and feeding if organized by the Project (see questionnaire - Annex 1).
- Semi-intensive feeding strategy: a low cost semi-intensive feeding strategy using a combination of organic fertilization (either by direct application or through livestock integration) and supplementary feeding with locally available agricultural by-products is recommended for the rural farmer. This feeding strategy has the necessary flexibility in that fish growth is not dependent upon a single food item but on a combination of different feed types. It is essential that the feeding strategy chosen has this flexibility as fertilizer, feed, and labour inputs may vary over a growing season depending on availability, and the financial status of the farmer.
- Application of organic manures: farmers should be encouraged to apply fresh animal manures to their ponds on a daily basis either through livestock integration or by dissolution and broadcasting over the entire pond water surface if maximum pond fertility and fish production is to be achieved. Although anaerobic compost cribs are suitable for those farmers who can only attend their ponds on an irregular basis (due to other more pressing agricultural farming activities), the stimulatory effect of this type of application method on pond fertility is slow. Manures should therefore be added as frequently as possible (if available), at least daily, on a little and often basis. According to Schroeder (1980) the maximum amount of manure that a pond can safely digest without undesirable anaerobic effects is about 100–200kg manure dry weight/ha/day or 70–140kg organic matter/ha/day (for Israeli pond conditions). These values correspond approximately to the manure produced from 100–200 pigs weighing 100kg each/ha/day, 15–30 cows weighing 500kg each/ha/day, or 2000–4000 poultry each weighing 2kg each/ha/day (Edwards, 1982). As a general rule manuring rate can be computed as dry organic matter at 2–4% of the standing fish biomass daily (Schroeder, 1980; for example manure fertilization programmes in different parts of the world see Tacon, 1988).
- Aerobic composts: although this method of composting requires a higher labour input and a planned farming approach, it is ideally suited to those farmers close to sugar cane, coffee and rice processing centres; sugar cane bagasse, coffee pulp and rice straw being waste products available at no cost to the farmer other than transportation. For details of different composting methods see Tacon (1988). The advantage of composting is that agricultural by-products which would otherwise have little or no nutritional value to farmed fish can be nutritionally transformed and upgraded into a potentially usable commodity for aquaculture; either as an efficient pond fertilizer or a supplementary feed source through the growth of bacteria and invertebrate food organisms.
- Supplementary diet feeding: farmers with limited cash resources should be encouraged to use low-cost agricultural by-products such as slaughterhouse wastes, mills sweepings, rice bran and chopped green leafy material when ever possible. Ingredient selection should be based on the following criteria: in order of importance, these include 1) cost (if at all), 2) availability, 3) handling and processing requirement prior to feeding (including transportation), and finally 4) nutritional value. As with organic manures, these supplementary feeds should be fed on a little and often basis every day according to fish appetite by hand feeding or the use of raised underwater feeding platforms, floating feeding collars or bait-rod feeders. In addition, farmers with high pond fish yields and adequate cash funds should be encouraged to intensify their production by increasing fish stocking density and using supplementary pelleted diets. Several key farmers should be selected for testing commercial/in-house produced feeds on a trial basis; the diets should be provided by the Project to the farmer free of charge for the first production cycle. However, it is essential that good farm records are kept of fertilizer and feed inputs and fish production, and that the results of these feeding trials are analysed from an economic and biological viewpoint. In this respect, it is important to calibrate standard feeding containers used by private farmers for feeding; for example, what is the weight of a 2kg capacity tin full of mill sweepings or dried omena, and what is the average weight of one sack of chopped green leaves or a wheel-barrow load of cow manure.
- District agricultural feed surveys: it is strongly recommended that an agricultural feed survey be conducted for each district within the Lake Basin Region. The main aim of this survey is to assist the farmer to develop his or her own farm feeding strategy and consequently maximise fish production and the utilization of available agricultural by-products within the Lake Basin Region. Guidelines for conducting an agricultural feed survey are given in Tacon, Maciocci and Vinatea (1987).
