February 1988




A report prepared for the Fishculture Development Project


Michael B. New
Senior Aquaculturist
Aquaculture Development and Coordination Programme
Fisheries Department

This report was prepared during the course of the project identified on the title page. The conclusions and recommendations given in the report are those considered appropriate at the time of its preparation. They may be modified in the light of further knowledge gained at subsequent stages of the project.

The designations employed and the presentation of the material in this document do not imply the expression of any opinion whatsoever on the part of the United Nations or the Food and Agriculture Organization of the United Nations concerning the legal or constitutional status of any country, territory or sea area, or concerning the delimitation of frontiers.

Rome, 1988

Hyperlinks to non-FAO Internet sites do not imply any official endorsement of or responsibility for the opinions, ideas, data or products presented at these locations, or guarantee the validity of the information provided. The sole purpose of links to non-FAO sites is to indicate further information available on related topics.

This electronic document has been scanned using optical character recognition (OCR) software. FAO declines all responsibility for any discrepancies that may exist between the present document and its original printed version.




















The consultant, who was released from the UNDP/FAO Aquaculture Development and Coordination Programme (INT/86/017) for the purpose, visited Zambia 9–19 February 1988 to advise on the use of compound tilapia feeds in the project's farms at Chilanga, Mwekera and Chipata. During the consultancy a brief review of ingredient availability and cost was carried out and the results of earlier experiments with mixed feeds for tilapia production were assessed. A number of alternative formulations for tilapia feeds were provided, together with recommendations for suitable feed manufacturing equipment for the project farms.

The following report summarizes the results of this mission. The consultant's terms of reference are given in Annex I; the itinerary and a list of persons met is contained in Annex II and Annex III respectively.


A review of feedstuff ingredient availability for fish feeds in Zambia had been conducted during the previous UNDP/FAO project (Fish Culture Development, ZAM/79/005) in 1981 (Gopalakrishnan, 1986) but kept not updated. However some information on current availability in the Copperbelt and Chipata had been assembled by project staff prior to the consultant's arrival. This was supplemented with data obtained during visits made during the mission. The combined results are summarized in Annex IV.

Generally, ingredients for stockfeeds are in extremely short supply in Zambia. Shortages are particularly acute seasonally and during temporary problems (e.g, drought). For example, during the mission there was an acute nation-wide shortage of maize products including mealie meal itself, the staple human food in Zambia. In addition, foreign exchange for the purchase of non-indigenous ingredients is extremely limited. Since late 1986 all animal feed mills have been nationalized, although some livestock rearing companies manufacture their own stock feeds. The result is that the limited supplies of conventional indigenous ingredients available tend to be contracted to the state-run feed milling companies or other regular purchasers and there is no open feedstuff market as such. Processing factories, whether they be for the manufacture of animal stockfeeds or the processing of ingredients used in them, are often short of input materials. Soy Nutrients Co. Ltd., for example, advertise for soybeans and often can only obtain them if they sell the processed product back to the supplier.

The principal indigenous vegetable proteins are full-fat whole soybeans, expeller and extracted soybean, expeller groundnut and cottonseed, fuzzy cottonseeds, and extracted sunflower. Expeller soybean and groundnut cakes, and fuzzy cottonseeds are only available in small quantities at specific locations (Annex IV). Supplies of all vegetable proteins (and other ingredients) are subject to limited national supply and local and seasonal shortages, sometimes due to processing plant breakdowns. Current prices for vegetable protein, as well as other ingredients, are recorded in Annex IV but have been adjusted for formulation purposes where price increases are expected (Annex V). Compressed yeast is available but is normally used only in bread and chibuku production.

Processed animal proteins are almost non-existent. Very limited supplies of blood meal and meat and bone meal exist but are said to be of poor and inconsistent quality. However, fresh cattle blood is readily available and is free to those who will dispose of it. There is no fish meal available. The only sources of ‘marine’ protein would be kapenta fish (Lymnothrissa miodon), usually dried, so-called ‘trash’ fish (unwanted species from fish farming), and under-sized tilapia (available when fish ponds are harvested). Only kapenta is plentiful but it is difficult to justify its use as a feed ingredient except for fry feeds since it is part of the very limited available supply of indigenous animal protein for human consumption in Zambia.

Several maize products are among the carbohydrate sources available but they are currently, like mealie meal and bread themselves, in short supply. There is no corn oil production and therefore no maize gluten meal or maize gluten feed available for stock feeds. Wheat bran is scarce and more expensive than maize meal. Rice bran is only locally available (Annex IV) and has not been used in the fish feed formulae (section 5). Cassava meal is only available to those that farm cassava; it is not available for sale. Brewery wastes in an undried form are readily available, both from (imported) barely and maize (chibuku production). Like cattle blood, brewery wastes are either free (chibuku) or very cheap (barley) to those who will remove them from the breweries. Being low in solids they are, however, costly to transport and must be used fresh.

Forages are not available in dried form but some estates dry them for their own use. One, the Zambia Sugar Company, uses lucerne and leucaena meals in its home-made fish feeds.

Should supplemental lipids be required, refined soy and sunflower oils (food grade) are available but no sources of animal fat were located.

Dicalcium phosphate is available for calcium and phosphorus supplementation.

Other ingredients, such as animal protein concentrates, amino acids, vitamins, preservatives, additives, etc., are imported (when possible) by the nationalized milling companies but are not available to others on the Zambian market. Those with foreign exchange purchasing rights could import them direct, if appropriate.


Considering the prevaling conditions in the stockfeed industry in Zambia it is difficult to recommend the use of compound feeds in aquaculture at present. The decision of the project to concentrate on integrated aquaculture (pig-fish; duck-fish)1 and the use of manure was, in my view, very wise. Any use of conventional indigenous feedstuff ingredients by commercial aquaculture means that less of the already deficient supplies is available for terrestrial and avian stockfeed production. Similarly the use of imported ingredients, either to supplement those locally available or to widen the range (e.g., animal proteins) which can be used for aquaculture would necessitate the use of scarce foreign exchange. Foreign exchange may be available for this purpose to some commercial companies but is unlikely to be available to the small farmer or at the local community level. Para-statal bodies such as the feed milling companies, which are subsidiaries of INDECO, experience considerable problems with the import of concentrates, vitamins and other additives; consequential shortages result in quality problems in their feed products. From a national government viewpoint, any decision to encourage the use of compound feeds in aquaculture (either through the diversion of existing ingredient supplies from terrestrial and avian stockfeed production, by encouraging greater production of indigenous ingredients, or by increased imports) can only be taken if economic benefit analysis shows it to be advisable. This type of study was outside the scope of the present mission and no recommendation is therefore given on this topic.

1 One commercial farm in Zambia (Kalimba Farm) has integrated crocodile rearing into its pig and duck integrated aquaculture system

Despite the difficulty in justifying the use of compound feeds in Zambian commercial aquaculture, there is a good case for demonstrating simple techniques of small-scale fish feed production and the results which can be obtained from the application of feeds made by them on project farms.


A number of feeding trials, which were reported by L'Heureux (1985), were carried out by the project between 1981 and 1985. These trials and some data collected by L'Heureux (1985) from commercial farms are summarized in Annex VI.

Those conducted by the government fish farms at Chilanga and Mwekera involved simple mixtures of oilseed cake (sunflower) and maize bran, fed in conjunction with manuring. In only one case was a food conversion ratio reported by L'Heureux (1985), for a 1981–2 trial at Chilanga. In this case AFCR's (apparent food conversion ratios) of 3.3:1 to 4.5:1 were achieved.

These FCR's are apparent rather than real since the feed mixture presented was not the sole source of feed. The efficiency of feed conversion achieved was typical for the simple feed mixtures used and agrees well with other data in the literature (Balarin, 1979). However, it is poor compared to what is achievable with a well-balanced pelleted compound diet. AFCR's of 1:1 to 2.5:1 have been achieved in Zambia in intensive tilapia culture (Annex VI) and in fish culture generally ratios in the range 1:1 to 2:1 are normal. The superior performance of pelleted compound feeds is due to better digestibility, less wastage, and (usually) a better nutritionally balanced diet. The simple mixtures used in feeding trials at Chilanga and Mwekera and at commercial farms in Chingola and Kitwe (Annex V) act as much as fertilizers as feeds. Much of the feed will not be directly used by the target fish but will contribute to pond nutrients and thus to natural food supply.

The generally rapid increase in stockfeed costs in Zambia (for example, a ton of poultry layers mash sold at Zam.K. 265–338 in 1981, Zam.K. 396–446 in May 1984, Zam.K. 606–613 in January 1986 and Zam.K. 1 120 in December 1986) is reflected in the difference in the (raw material only) cost of the fish growers feed made by the Zambia Sugar Company (Annex VI) in 1985–6 (Zam.K. 404/mt) and in February 1988 (Zam.K. 1 250/mt).


The following formulations for tilapia feeds have been based on the conventional feedstuff ingredients currently available in Zambia (Annex IV); the analytical and cost data of the selected ingredients used in formulation are given in Annex V. Ingredients have been selected for their suitability in compound fish feeds; other Zambian ingredients, such as poultry manure, banana leaves and maize sweepings, would be better used as individual inputs to pond culture, where they would be utilized practically as feeds and partially as fertilizers, rather than as constituents of pelleted compound feeds.

Some formulations contain ingredients where quality is especially important. These include moist ingredients such as fresh cattle blood and the two types (barely and chibuku) of brewery wastes. Feeds containing these ingredients are prone to fungal and bacterial spoilage unless they are fed immediately (as a moist feed) or rapidly dried to about 10% moisture before storage. It is essential in all cases to ensure that the incoming material is fresh; these moist ingredients must be collected and compounded into feeds on the day of production - they must not be stored at the point of supply or on the fish farm before use. Mouldy ingredients of any type must not be used in fish feeds. Ingredients must be inspected on arrival for quality and properly stored. Detailed advice on quality control and storage is contained in New (1987) which should be carefully studied before ingredients are ordered.

Formulations have been provided for experimental broodstock, fry, fingerlings and rearing diets for Tilapia andersonii. Several versions of each have been formulated, partly to cover probable temporary shortages of ingredients and partly so that experiments can be conducted to see which gives the most effective result. It should be noted that overfeeding (see section 7) is as often the cause of poor conversion efficiency (high AFCR) as poor quality feeds. AFCR will depend not only on the formulation and the ingredients available but also on the quality of each batch of raw materials used. AFCR is also not the only criterion for a successful diet - these also include its unit cost (a high unit feed cost with a low AFCR may prove more effective than a cheap feed with poor feed conversion efficiency - it is the cost of feeding per unit weight of fish produced that counts), the growth rate achieved, and the proportion and quality of marketable fish in the harvest. All these topics are fully discussed in New (1987).

Within each type of diet formulated there is a wide range of unit feed costs which depend upon the ingredients used. All feeds containing dried kapenta are expensive but a source of fish protein is advisable in fry and fingerling diets and in good quality broodstock diets. A cheaper diet, without kapenta, has also been included in all the non-fry diets, however, so that trials can be conducted. None of the diets includes a vitamin mix because of cost and lack of availability and primarily because a vitamin mix is probably unnecessary in the project rearing system. One Zambian commercial fish farm even operates an intensive rearing system using a non-vitamin-supplemental diet (Annex VI). However it would be useful to demonstrate whether a vitamin mix is beneficial or not. The project should conduct experiments to examine the necessity for vitamin supplementation in semi-intensive aquaculture. A formulation for a vitamin supplement is given in Annex VII.

All diets formulated can be extruded by the moist feed production process (mincer-pelletizing) and can either be fed immediately in the moist form (approximately 45% moisture) or sun-dried to about 10% moisture for storage and later use. If fed in moist form, a greater feeding rate is of course necessary (see section 7). Excessive exposure to sun during the drying process may cause partial loss of vitamin potency which may explain why diets with vitamin supplementation surprisingly failed to give better results in a Zambian intensive commercial farm (Annex VI) than unsupplemented feeds. This problem is not expected to be severe with the diets formulated here, which are unsupplemented, provided feeds are not left in the sun longer than necessary to reduce the moisture content to about 10%.

Some formulations (those not containing moist ingredients) could also be used, without the addition of water, for the dry pelleting process (laboratory pelleter). These have been marked.

In general, the unit cost of the diets listed below reflects their anticipated nutritional quality. All diets have been balanced (within ± 10%) for crude lipid, crude protein, calcium and available phosphorus. However some might be deficient in one or both of the essential amino-acids, lysine and methionine. This should not prove a problem in the project rearing system currently in use and it is not advisable to supplement ‘deficient’ diets with synthetic amino acids in view of (a) the difficulties of adequately mixing in small components without premixing (b) their high cost and (c) the foreign exchange necessary for their purchase. Those formulations which are below suggested minima for lysine and/or methionine (which have been derived for diets which provide the sole food source in intensive systems) have been marked.

The estimated cost (on a dry-matter basis, for comparison) of each diet formulated has been indicated. This has been derived from ingredient costs, plus 30% to cover production and transport costs.

The formulations which have been derived from selected ingredients available in Zambia follow:

Table 1: Experimental broodstock diets of Tilapia andersonii culture in Zambia

IngredientsBrood 1Brood 2Brood 3Brood 4
Dried kapenta meal3535-  31
Full-fat soy   6---
Extracted sunflower meal   4-28-
Extracted soyabean meal-   930   6
Compressed baker's yeast--- 10
No. 3 (maize) meal  49---
Maize bran-  50  30.5 47
Refined sunflower or soybean oil--  2.5-
Cane molasses   6   66   6
Dicalcium phosphate--3-
 Add water to consistency (see New, 1987, pages 91 and 99 for details) during mixing to form a dough prior to moist diet extrusion. Brood 3 could alternatively be pelleted in the dry laboratory pelleter without addition of water. Moist diets must be fed immediately after manufacture or sun dried to 10% moisture before storage
Cost on dry matter basis (ingredient cost + 30%) in Zam.K./kg8.

The above diets were formulated to contain (on a dry matter basis) 7% crude lipid and 32% crude protein. All have at least 0.75% available phosphorus. Brood 3 is deficient in lysine and Brood 4 is deficient in methionine.

Table 2: Experimental fry(> 1 g) diets for Tilapia andersonii in Zambia

IngredientsFry 1Fry 2Fry 3Fry 4
Dried kapenta meal  57  60  59  50
Extracted soybean meal-  17  18-
Full-fat soy  24--  22
Fresh cattle blood---  10
No. 3 (maize) meal  13--  18
Maize bran-  13   13.3-
Refined sunflower or soybean oil-    4     3.7-
Cane molasses    6    6    6-
 Add water to consistency (see New, 1987, pages 91 and 99 for details) during mixing to form a dough prior to moist diet extrusion. The moisture content of all these diets is too high for use in the laboratory dry pelleter. After moist extrusion, sun-dry the feeds and grind to pass 1 mm hammermill screen before use or storage
Cost on dry matter basis (ingredient cost + 30%) in Zam.K./kg15.116.015.714.2

The above diets were formulated to contain (on a dry matter basis) 10% crude lipid and 50% crude protein. All have at least 1.1% available phosphorus. None are deficient in lysine but Fry 1, Fry 2 and Fry 3 are a little low in methionine.

Table 3: Experimental fingerling (1–20 g) diets for Tilapia andersonii in Zambia

IngredientsFing 1Fing 2Fing 3Fing 4
Dried kapenta meal  38  35  35-
Full-fat soy  13  10--
Wet undersized minced tilapia-  10-  10
Extracted sunflower meal--  16  28
Extracted soybean meal---  30
Refined sunflower or soybean oil--    1    3
No. 3 (maize) meal  43  39--
Maize bran--  42  26
Cane molasses   6   6    6-
Dicalcium phosphate---    3
 Add water to consistency (see New, 1987, pages 91 and 99 for details) during mixing to form a dough prior to moist diet extrusion. Fingerling 1 could alternatively be pelleted in the dry laboratory pelleter without addition of water. Moist diets must be fed immediately after manufacture or sun-dried to 10% moisture before storage
Cost on dry matter basis (ingredient cost + 30%) in Zam.K./kg9.810.58.83.9

The above diets were formulated to contain (on a dry matter basis) 8% crude lipid and 35% crude protein. All have at least 0.8% available phosphorus. Fing 4 is deficient in lysine. All diets are deficient in methionine and, if the diets were to be used as a sole source of food, they would be better balanced with 0.3% supplementary methionine.

Table 4: Experimental growers (20 g to market size) diets for Tilapia andersonii in Zambia

IngredientsGrow 1Grow 2Grow 3Grow 4Grow 5Grow 6
Expeller groundnut cake  29---  36-
Extracted sunflower meal  20  27-  48--
Full-fat soy-    5----
Extracted soybean meal--  48-  15  32
Fresh cattle blood-----  10
Moist brewery or chibuku waste-  35---  18
No. 3 (maize) meal  42  31--  41-
Maize bran--  42  47-  37
Refined sunflower or soybean oil--   1--   0.5
Cane molasses   6-   6   2   5-
Dicalcium phosphate   3   2   3   3   3   2.5
 Add water to consistency (see New, 1987, pages 91 and 99 for details) during mixing to form a dough prior to moist diet extrusion. Grow 1, Grow 3, Grow 4, and Grow 5 could alternatively be pelleted in the dry laboratory pelleter without addition of water. Moist diets must be fed immediately after manufacture or sun-dried to 10% moisture before storage.
Cost on dry matter basis (ingredient cost + 30%) in Zam.K./kg1.

The above diets were formulated to contain (on a dry matter basis) 6% crude lipid and 28% crude protein. All have at least 0.7% available phosphorus. Grow 1, Grow 2, Grow 4, and Grow 5 (i.e., the cheaper diets) are deficient in lysine. Grow 1, Grow 3, Grow 5 and Grow 6 are deficient in methionine. These deficiencies should only prove a problem if the feeds were used as sole sources of feed.

In relation to the amino acid deficiencies of some of the diets whose formulae are given above, it should be noted that the 50% maize bran/50% sunflower mixture used by the project earlier had only about 50% of the desirable lysine content although it was sufficient in methionine. It also contained only 36% of the desirable available phosphorus content.

All aspects of ingredient purchase, quality control, moist feed manufacture,storage, are covered in detail in the ADCF feeds manual (New, 1987) and are not repeated here. It is essential that this manual be studied carefully before feed manufacture and usage commences.


It is recommended that the three project farms, in their feed preparation and feeding trials and demonstrations, utilize the moist extrusion technique. The resultant moist product may either be fed immediately after manufacture or sun-dried and stored for subsequent use. A full description of all the techniques for feed preparation by this method, including grinding, mixing, extrusion, drying and quality control, together with information on the storage of ingredients and complete feeds are contained in an ADCP manual (New, 1987) and will not be repeated here. Several copies of this manual have been distributed to the project and to others, including the Zambia Sugar Co., the Animal Sciences Department of the University of Zambia and Kafue Fisheries. The specific equipment requirements for this type of fish feed manufacture at the project sites is given in this section, with a rough indication of cost. In addition, in response to a request from the Director of Fisheries, the names of manufacturers of laboratory-scale dry pelleting equipment have been provided. Purchase of one laboratory-scale pelleter for use at Chilanga, would enable the technique of conventional dry pellet production to be demonstrated. However, this technique is not recommended for general project use. An alternative method of demonstrating dry pelleting techniques would be to visit the facilities at the Zambia Sugar Company.

Fry feeds, which need to be of 500–1500u particle size, can be made by grinding dried pelleted feeds through the smallest diameter (1 mm) screen of the grinder. However, since fry are currently reared in the broodstock ponds in the current project rearing system, rather than in a hatchery, fry feeds will not be required, except for experimental purposes, for which formulations (section 5) have been provided. 2 mm diameter pellets are suggested for fingerlings and 3 or 4 mm diameter pellets for growers and broodstock tilapia.

The following equipment for moist feed manufacture will be required 1 for each project farm, and is capable of producing several tons of feed per day:

 Price Guide
One250 kg×100 g Avery Platform scale Model No. 3901 2 18 500
One10 kg Thornton scale, with set of weights 2   5 500
Onehammermill with range of screens from 2 mm down to 1 mm. A hammermill with a 5–7.5 HP motor would be ample. Locally available (Chilanga has one already)  12000
OneTaninaka3 1 HP fish-feed mixer, model No. UZU-12 GS (500 litre batch mixing capacity) 220V 50 cycle  40 000
OneFuji-Mizuho3 5 HP Chopper-pelletizer (for mincing wet ingredients and producing moist pellets) 220V 50 cycle, model No. 82 KM3-5GP, with 2 mm, 3 mm and 4 mm die plates  63 000
 Tentative total(US$ 17 375)139 000

1 The mention of a specific company does not imply a recommendation. Similarly, the absence of other companies' names should not be construed to mean that they are unsatisfactory.
2 Available in Cha Cha Cha Street, Lusaka
3 Obtainable from Yashima Bussan Co. Ltd., 6F Chugai Building, 5-13-5 Nishi-Tenma, Kita-Ku, Osaka 30, Japan Tel: (06) 313–2144. Telex 5234284 YASMAX J

Laboratory pellet mills can be obtained either from California Pellet Mill (CPM) Europe Ltd., West March, Daventry, Northants, NNll 4SA, England (Telephone 03 272 4721) or from Buhler Brothers Ltd., Engineering Works, CH-9240 Uzwil, Switzerland. Specifications for the Buhler version are not available with the consultant but CPM produce three versions of their laboratory-scale pelleter. Each has a main drive 2 HP motor and a complete range of die plate sizes from 1 mm up to 12.7 mm are available. The models differ only in the method of feeding mixed ingredients into the pelleter. Model CL Type 2 is suggested as the most appropriate; this has a ¼ HP variable speed, variable pitch, feeder screw. This is usable for ‘cold’ pelleting but also has an inlet for steam, should the project wish to experiment with small-scale steam pelleting (this would require local supply of a steam boiler). Being a US-made equipment, the standard model requires 3 phase, 60 cyle 230–460 V current. In ordering this equipment, it is essential that their English subsidiary be asked to re-wire the equipment to suit power supply before shipment. It can be adjusted on site but requires an electrician familiar with the equipment. Experience with this pelleter is that it will produce about 80 kg of 2.4 mm diameter pellets per hour. The production rate of larger pellets would be greater but, if smaller diameter pellets are required, the output would be significantly less. The Buhler laboratory pelleter is similar to the CPM equipment but detailed specifications should be requested from the manufacturers. Detailed CIF quotations should be sought from the manufacturers allowing for an expenditure of US$ 15 000–20 000 for the pelleter and a range of dies and spares.

The laboratory (or other dry) pellet mills cannot cope with input materials with a moisture content higher than 12–15%; the moist extruder (pelletizer/chopper) specified above requires an input moisture content of about 40–45% for optimum quality extrusion. Full details on the method to achieve the correct moisture content for moist feed extrusion are given in the ADCP feeds manual (New, 1987), pages 91 and 99.


Wild tilapia naturally feed more or less continuously during the day (Jauncey and Ross,1982) and fry require more frequent feeding than adults. Experiments in Kuwait (New, et al., 1984) showed that 16–23 day old, 9 mm tilapia (Oreochromis spilurus) fed continuously (automatically) or manually five times per day survived better than those fed three times per day. Jauncey and Ross (1982) suggest that fry should be fed a minimum of four times per day. Twice per day would be sufficient for adults. Basically it is always best to spread the daily feed for tilapia amongst as many feeds as economically feasible. The use of automatic feeders is one way to achieve this but it removes the advantages of manual feeding (more control, inspection of fish behaviour/disease, etc.). The topic of feeding frequency is discussed more fully in the ADCP feeds manual (New, 1987) pages 122–126, while the various types of automatic feeders, some of which can be ‘home-made’, are described in the same publication, pages 235–245.

The following table provides suggested feeding rates for tilapia. However, these must not be rigidly applied without daily judgement, based on the operator's observations on water quality, water temperature, fish behaviour, etc. Over-feeding will result in wastage of feed (and thus poor conversion rates and poor economic viability), while under-feeding will reduce growth rate and output per unit (pond) area.

Table: Suggested Guidelines for Feeding Rates for Tilapia andersonii

Fish SizeFeeding Rate (% of biomass per day)
Dry Feeds (10% H2O)Moist Feeds (45%-H2O)
<1to demandto demand
5–1010decreasing to816reducing to13
10–20  8"713"11.5
20–50  7"411.5"  6.5
50–75  4"3  6.5"  4.9
75–100  3"2.5  4.9"  4.1
100–150  2.5"2.0  4.1"  3.3
150–200  2.0"1.8  3.3"  2.9
200–300  1.8"1.5  2.9"  2.5
300–400  1.5"1.2  2.5"  2.0
400–500  1.2"1.1  2.0"  1.8
500–600  1.1"1.0  1.8"  1.6
>600 1.0  1.6 

If feeds with moisture contents between 10 and 45% are used, the appropriate feeding rate adjustment should be made.

Comments on feeding rates are given in the ADCP feeds manual (New, 1987), pages 118–122, and on the assessment of biomass in the same document, pages 126–128.


Development of Fish Culture in Zambia



The Consultant will:



February8Left Rome
 9Arrived Lusaka; FAO Representative's Office, Lusaka
 10Department of Fisheries; Chilanga Farm; National Resources Development College, Lusaka; University of Zambia Animal Husbandry Department, Lusaka; Soy Nutrients Co. Ltd., Lusaka; National Milling Company, Lusaka
 11ALCOM (Lusaka); National Milling Company (Feedmill), Lusaka; Chilanga Farm; Mount Makalu Central Agricultural Research Station
 12Kalimba Farm; Chilanga Farm; University of Zambia Animal Husbandry Department; sundry equipment suppliers in Lusaka
 13–14Data analysis
 15Chilanga; Statistcs Office; Ministry of Agriculture and Water Development; data analysis
 16Zambia Sugar Co. (Fish Farm), Mazabuka; Lee Yeast Factory, Kafue; Kafue Fisheries, Kafue
 17Formulation and draft report writing
 17Formulation and draft report writing; Mr. F. Flynn of Kafue Fisheries
 19Draft report finalization; FAO Representative's Office, Lusaka; Chilanga Farm (slide presentation);
 20Departed Lusaka 00.30; Arrived Rome 17.30



Mr. Sheikh Ahmad Tejan Wadda, FAO Representative, Lusaka

Mr. N.A.C. Watson, Programme Officer, Office of the FAO Representative in Zambia; Lusaka

Mr. E.D. Muyanga, Director, Fisheries Department, Chilanga Professor M.W. Siamwiza, Deputy Director-General, National Council for Scientific Research

Dr. D. Nyirenda, Head, Animal Science Department, School of Agricultural Sciences, University of Zambia, Lusaka

Dr. Aziz A. El-Danbouky, Head of Department of Animal Science, Natural Resources Development College, Lusaka

Dr. V. Gopalakrishnan, Chief Technical Adviser (GCP/ZAM/038/NET), Chilanga

Mr. E. Boma, Project Co-Manager, Development of Fish Culture in Zambia (GCP/ZAM/038/NET), Chilanga

Mr. K.P. Banda, Aquaculturist (GCP/ZAM/038/NET), Chilanga

Ms. E Cayron-Thomas, Aquaculturist (GCP/ZAM/038/NET), Chilanga

Mr. Arne Andreasson, Programme Manager, Aquaculture for Local Community Development Programme (ALCOM), GCP/INT/436/SWE, Lusaka

Mr. Boyd A. Haight, Aquaculturist, Aquaculture for Local Community Development Programme (ALCOM), GCP/INT/436/SWE, Lusaka

Mr. Hans Aase, APO Aquaculturist, ALCOM (GCP/INT/436/SWE)

Mr. Ulf Wijkström, Aquaculture Information Systems Consultant to GCP/INT/436/SWE

Mr. S.D. Madhu, Information Officer, Bay of Bengal Programme (Consultant to GCP/INT/436/SWE)

Dr. N.J. Siulapwa, Technical Superintendent (Stockfeeds), National Milling Company, Lusaka

Ms.R. Chabala Mwale, Chief Chemist, National Milling Co. Ltd., Lusaka

Mrs. B.M. Mwamba, Technical Representative (Stockfeeds), National Milling Company, Lusaka

Mr. Pramod Mahnot, Soy Nutrients Ltd., Lusaka

Mr. Isaik Chishiba, Livestock Manager, Zambia Sugar Co., Mazabuka

Mr. Cosmas M. Mbanacele, Livestock Officer, Zambia Sugar Company, Mazabuka

Mr. I. Suleman, Lee Yeast Factory, Kafue

Mr. S. Shah, Lee Yeast Factory, Kafue

Mr. W. Thomas, Kalimba (crocodile-pig-duck-fish) Farm

Mr. Fergus Flynn, Kafue Fisheries, Kafue

Mr. Domingo Tapiador, Project Coordinator, World Bank Fisheries Development Project, Lusaka



The following table (Table 1) shows the availability of raw materials for animal feedstuffs close to the three government fish farms (Chilanga, Mwekera and Chipata) in 1988. Where prices are shown, they are quotations obtained in early 1988. Prices quoted for the Lusaka area would generally apply also to the Copperbelt and vice versa; however, the price of ingredients in Chipata, if not already available and priced there, could be up to 25% higher than those in Lusaka due to transport costs.

Table 1: Cost of feedstuff ingredients in February 1988 1 2

IngredientLocation and Cost (Zam.K./mt)
(Copperbelt area)
Vegetable Proteins   
Full-fat soybean3 000–3 260  
Expeller soybean cake 31 100  
Extracted soybean meal2 800  
Expeller groundnut cake3  1 500
Extracted sunflower meal1 000–1 500 870
Expeller cottonseed cake1 500500 
Fuzzy cottonseeds 3  900
Compressed bakers/brewer's yeast3 120  
Animal Proteins   
Blood meal36 500  
Fresh cattle bloodFreeFreeFree
Carcass meal (meat and bone meal)35 500  
Dried kapenta fish3 500–20 000  
Trash fish3 4FreeFreeFree
Undersized tilapia35 0005 0005 000
Maize (whole) 5889  
Maize meal 5870  
No. 3 meal 6250 133
Maize residue (No. 3 meal?) 260 
Maize bran200 133
Maize sweepings  Free
Rice bran  133
Wheat bran210 222
Moist (barley) brewery waste 25 
Moist (maize) chibuku beerFreeFreeFree
Cane molasses100  
Cassava meal 3   
Leucaena meal 7   
Lucerne meal7   
Dried chicken manure  140
Wet chicken manure  40
Waste vegetables  20(drum)
Urea1 300  
Fats and Oils   
Soybean oil 59 800  
Sunflower oil 59 800  
Dicalcium phosphate2 740  
Limestone flour400  
Hydrated lime340  
Coarse salt820  
Amino Acids and Vitamins 8   
Pure vitamin E4 500  
Lysine29 100  
Methionine29 100  

1 Prices were obtained in early 1988. Actual prices used for feed formulation are different (see Annex V)
2 = available but no specific cost data (see introduction to the annex)
3 Extremely limited and localized availability
4 Minimal quantities (10–20 kg/week) of extraneous fish from project farm pond harvesting; these are normally given to staff
5 Normally for human consumption
6 Byproduct of maize milling (for cereal); contains bran, germ and endosperm
7 Only used by Zambia Sugar Company - grown and dried on company plantation; not available on market
8 National Milling Company intake prices - not available for general sale



IngredientAnalytical Characteristics (% on As-Fed Basis)Cost1
MoistureCrude FatCrude ProteinCrude FibreAsh
Blood meal10.01.476.5-0.56 500
Fresh cattle blood80.00.317.0-0.1free
Undersized tilapia fish80.05.412.0-2.65 000
Dried kapenta fish22.25.353.7-18.815 000
Soybean or sunflower oil1.099.0---9 800
Expeller undecorticated cottonseed cake11.04.523.519.11.01 000
Expeller groundnut cake11.07.441.44.97.01 500
Full-fat soya8. 000
Extracted soybean meal10. 700
Extracted sunflower meal10.02.043.522.06.91 500
Moist (barley) brewery waste85.
Moist chibuku brewery waste85.
Maize bran12.
No.3 meal12.
Wheat bran12.
Compressed brewery yeast60.00.520.00.63.43 120
Cane molasses25.0-2.8-8.3100
Dicalcium phosphate3.0---902 740
Lysine32.0-+--29 100
Methionine31.0-+--29 100

1 Analytical figures derived from NMC data, Chimwano (1978), Göhl (1981), etc.
2 Costs used were the highest quoted in February 1988 (see Annex IV); in the case of soybean products the costs used in formulation reflect expected price increase by May 1988
3 Assumed price



LocationDateFeed Composition (meal mixture unless otherwise stated)Stocking
Unit Feed
Chilanga government fish farm1981Maize bran/sunflower cake (with fertilization)3–4??1.94–4.93 (4.6)L'lleureux, 1985
"1981–250% maize bran/50% sunflower cake (with fertilization)2.5?3.3:1
"1983?40% maize bran/60% sunflower cake4.25 ?3.02"
"1985 (economic model)50% maize bran/50% sunflower cake 238  "
Mwekera government fish farm1984Maize bran/soybean cake2.5   L'Heureux, 1985
Lazy-A Ranching Co. Ltd. Chingola1984–5Maize bran, Chicken litter and calcium phosphate (with pig manure) 80–130 2.8L'Heureux, 1985
Copperbelt Power Co. Ltd., Kitwe1984–5Chicken litter sweepings (contains 20% maize bran) 60 4.1 (polyculture)L'Heureux, 1985
Zambia Sugar Co., Mazabuka1985–6Fish silage (fish meal & molasses), algae, blood, soya, cottonseed, sunflower, maize, wheat bran yeast, kapenta, dryfish, leucaena (proportions not stated - feed pelleted)3674041.5:1
67 kg/m2 (hybrid tilapia in running water)L'Heureux, 1985
1988Similar, but without algae and including Lucerne meal, blood meal and carcass meal 1 2501.0:1
 This mission (Annex VI)

1 AFCR = Apparent food conversion ratio. See New (1987), page 13



During the mission the consultant visited three tilapia farms with very diverse management systems. These were the farms run by Kafue Fisheries (a subsidiary of TAP building Products Ltd.), the Zambia Sugar Company and Kalimba Farm (a subsidiary of AMI).

The farm run by the Zambia Sugar Company produces about 1 mt/week (90% Tilapia nilotica; 10% common carp) for its estate workers from an intensive raceway system supplied from the estate's irrigation system. The farm has its own hatchery and is the only one with a feed pelleting plant of its own, which produces all its feed requirements. Feed manufacturing equipment includes a pre-grinding mixer, a hammer mill, a mixer, a screw-elevator and a complete Lister Farm Pelleting Press with two pelleters. 2.5 mm and 4 mm dies are used but, at the time of visiting, all pellets were 2.5 mm in diameter (which the staff find break more easily on handling) as the 4 mm die was broken. Production capacity is about fifteen times the feed requirement of the existing fish farm. The water stability of the pellets had not been tested and was not thought important because of the feeding habits of the fish. Some of the difficulties of operating a pelleter of this type were in evidence during the visit; a die blockage had occurred and severe over-heating of the pellets had resulted. The pelleter is a ‘cold’ press (without steam). Molasses are used in the feeds at up to 10% but are added during mixing rather than pelleting. Other feed ingredients (some of which are alternatives) include blood meal and carcass (meat and bone) meal obtained from the Cold Storage Board as meals or dried from the moist form on the farm. Raw fresh cattle blood is boiled for 10–15 minutes and the scum is sun-dried. Dried kapenta fish and under-sized tilapia (from Kafue Fisheries) are utilized. When dried fish are unavailable or fresh fish cannot be dried because of climatic conditions the wet fish is silaged so that it can be stored. 10% molasses and formic acid are added to the minced fish for the 3-day fermentation process. Other ingredients included expeller cottonseed cake, expeller soybean cake, wheat bran (now abandoned as scarce and too expensive), expeller sunflower cake, No. 3 meal, molasses, leucaena meal (used at up to 5%) and lucerne meal. The last three items are produced on the company estate. It was extremely interesting to note that the use of a very expensive (Zam. K. 342/kg in 1985, equivalent now to about Zam. K. 171/kg (US$21.5/kg) custom made vitamin mix, recommended by Stirling University, had been abandoned several years ago as unnecessary. Trials had shown that omitting the vitamin mix had not affected growth rate although some haemorrhaging had occurred. The consultant had already suggested that the GCP/ZAM/038/NET project might not need to add a vitamin mix for their semi-intensive pond culture rearing system but this would be expected to be essential in an intensive raceway system, such as that used by the Zambian Sugar Company, whose no ‘natural’ food is available to the fish. Generally, omitting vitamin mixes is anathema to nutritionists (and to vitamin manufacturers!) and it was significantly interesting to learn that at least one intensive fish farm has been able to successfully produce fish for several years without supplementary vitamins. Losses of the vitamin content of the feed ingredients used on this farm may (except when a die blockage occurs!) be less in the ‘cold’ press system used than in a steam pelleter. In addition, some of the ingredients used were rich in vitamins. The farm produces fry feed (mixed and finely ground only) and pelleted fingerling and growers feeds. The fry feed is about 45–50% crude protein (CP) and 10% crude lipid (CL) while the fingerling and growers diets are 35% CP and 28–30% CP respectively. Both fingerling and growers feeds are about 8% CL. The total raw material cost of the growers pellets was said to be Zam. K. 1 250/mt. This is similar to several of the diets formulated for growers during the mission (see section 5). Food conversion ration (FCR) of 0.7–0.9:1 were being obtained for fish weighing less than 30 g while those between 30 g and harvested size (150–200 g) achieved an FCR varying between 1:1 and 2.5:1. At an assumed average FCR of 2:1 the feeding cost (raw materials only) for 1 kg of production size fish could therefore be Zam. K. 2.5. If feed processing and related costs were taken into account the feeding cost would be about Zam.K. 3.25.

Kafue Fisheries is an integrated pig-fish (Tilapia andersonii) farm which has just started experimental hatchery work. Production of paddy rice has also recently commenced and it is intended to use rice bran as an input into the aquaculture system. The marketable fish are all sold to one supermarket in Lusaka. Fish are not sexed, but there are plans to commence mono-sex production using methyltestosterone. A sizeable proportion of the harvest is below acceptable size for the marketing outlet used. ‘Non-marketable’ fish are either sold to Zambia Sugar Company at Zam.K. 4–5/kg for use in their feed manufacture or bartered with the same company in exchange for fry feeds for Kafue Fisheries new hatchery system. Apart from this, no feed is used currently on this farm, except for the pigs. Details of the Kafue Fisheries and the Zambia Sugar Company fish farms are not given here because of the short time available and because they are recorded elsewhere (e.g., L'Heureux, 1985). The visit of Kafue Fisheries was particularly interesting to the consultant because, in 1978–9, he had prepared the original proposal with which the consultants (Landell Mills Associates) secured the contract to design and operate the original farming system. The integration of pigs into the system was a direct result of the various FAO-executed projects based at Chilanga, as was the species of Tilapia now being successfully used. The manager of Kafue Fisheries, Mr. Fergus Flynn, has prepared a practical manual on integrated pig-fish aquaculture and seeks publication. It was suggested that ADCP be contacted to see whether they could publish it withing the African series. However, the consultant was not shown the manual and no comment can be provided on its quality. The concept is useful.

The Kalimba Farm is unique (at least in Zambia) in that it integrates crocodile rearing into a duck-fish and pig-fish integrated farm. The tilapia are hand-sexed at 20 g. Females are fed to the crocodiles, which take two years to reach a size at which their skins are marketable. The farm gate value of Tilapia andersonii was said to be Zam. K. 15/kg. It is hoped that the manager, Mr. W. Thomas, will write an article on the operation of this farm, which would be of considerable interest both within FAO and to the general public through the fish farming press. Mr. Thomas' wife, Ms. E. Cayron-Thomas, who leaves the GCP/ZAM/038/NET project in March 1988 is already preparing a review of Zambian aquaculture and will bring the draft to Rome during her de-briefing session in May 1988.



The following vitamin mix formulation (designed by Hoffman La Roche for carp) is suggested for experimental use in GCP/ZAM/038/NET.

The mix should be added at 1% in dry diets or 0.6% in moist diets, replacing the same quantity of one of the carbohydrate sources such as No. 3 meal or maize bran.

(per kg of vitamin mix)
Vitamin A800 000 I.U
    "       D3180 000 I.U.
    "       E6 500 mg
    "       K31 000 mg
    "       B12 000 mg
    "       B22 500 mg
    "       B62 000 mg
    "       B12        2 mg
Vitamin C15 000 mg
Niacin10 000 mg
Pantothenic acid  6 000 mg
Folic acid     500 mg
Biotin       30 mg
Choline chloride80 000 mg
Inositol15 000 mg



Animal feed production in Zambia fluctuated around the 100 000 mt/year (± 20%) from 1976–81. Strenuous efforts were made to obtain figures for 1982 onwards but were unsuccessful. Following nationalization of the private milling companies in late 1986, all feed mills are now controlled by the para-statal organization INDECO. Only one will, that is operated by the National Milling Company (NMC) in Lusaka, was visited during the mission; however it is believed to be representative of other Zambian feed mills.

NMC Lusaka makes a range of meals for poultry, pigs, cattle and horses and, on (extremely limited) demand, cattle cubes and rodent pellets. Its Buhler pelleting press (which could not be seen) was said to be 12–15 years old and to have a capacity of 3 t/hour. As far as it could be ascertain, the only other pelleter in Zambia is operated by the Zambia Sugar Company (see Annex VII) for its own fish farm use only. Pellets or cubes are made by NMC only on order. The factory had once made some fish pellets but these had not been accepted by the customer who requested them, on quality grounds, and they remain on the stock list. The mill relies heavily on the manufacturers of its imported concentrates (containing animal proteins, vitamins and other additives) for feed formulation but does make its own local feed formulation adjustments when these concentrates are unavailable. Supplies of concentrates and vitamin mixes frequently run out due to purchasing/foreign exchange problems. In addition the company is unable to select its suppliers; open bidding often results in the purchase of concentrates from dubious sources and poor feed quality. These factors, and the frequent necessity to replace concentrates with local vegetable proteins without vitamin supplementation, cause severe feed quality problems. The latter result in poor growth rates, egg production, survival rate, etc., amongst the animals supplied, and a poor image for the mill. Least cost formulation adjustments on a daily basis are impossible due to the lack of control which the mill is permitted to have over its ingredient purchasing. In addition to the problems with imported ingredients, supplies of local ingredients are very limited and, in common with all other milling companies, NMC has severe supply problems.



Balarin, J.D. and Hatton, J.P., 1979. Tilapia. A guide to their biology and culture in Africa. Institute of Aquaculture, University of stirling, UK, 174 pp

Chimwano, A.M.P., 1978. Zambia's farm livestock feedstuffs: their average chemical composition and nutritive value. The University of Zambia School of Agricultural Sciences, Lusaka, 11 pp

Göhl, B., 1981. Tropical Feeds. FAO Animal Production and Health Series No. 12, FAO, Rome, 529 pp

Gopalakrishnan, V., 1986 Terminal Report (draft), Zambia - Fish Culture Development (ZAM/79/005), FAO, Rome, 131 pp

Jauncey, K. and Ross, B., 1982. A guide to tilapia feeds and feeding. Institute of Aquaculture, Stirling University, UK, 111 pp

Leung, W-T.W., Busson, F. and Jardin, C., 1968. Food Composition Table for use in Africa. FAO and U.S. Department of Health, Education and Welfare, Bethesda, Maryland 20014, USA, 306 pp

L'Heureux, R.L., 1985. Economic Feasibility of Fish-Culture in Zambia: Report preparted for the Republic of Zambia (Project TCP/ZAM/4405(A), 99 pp + 10 annexes

New, M.B., 1987. Feed and Feeding of Fish and Shrimp - A Manual on the Preparation and Presentation of Compound Feeds for shrimp and Fish in Aquaculture, FAO, Rome, ADCP/REP/87/26, 275 pp

New, M.B., Hopkins, K.D., and El-Dakour, S., 1984. Effects of feeding frequency on survival and growth of tilapia fry. Kuwait Institute for Scientific Research, Report No. KISR 1287, 8 pp

Rothius, A.J., 1988, Quarterly report of Associate Professional Officer (in preparation), GCP/ZAM/038/NET

Top of Page