PART III
AQUACULTURE IN RIVER BASIN DEVELOPMENT/AQUACULTURE ET MISE EN VALEUR DES BASSINS FLUVIAUX (continued)

SESSION V   INTEGRATED DEVELOPMENT, ECONOMIC AND TECHNICAL ASPECTS OF AQUACULTURE/DEVELOPPEMENT INTEGRE ET ASPECTS ECONOMIQUES ET TECHNIQUES DE L'AQUACULTURE (continued)

CONSTRUCTION DE CAGES FLOTTANTES POUR UN ELEVAGE COMMERCIAL DE SAROTHERODON NILOTICUS

CONSTRUCTION OF FLOATING CAGES FOR COMMERCIAL REARING OF SAROTHERODON NILOTICUS

par

et

J.C.V. Arrignon
Directeur FAO du Projet
Développement de la Pêche et de la Pisciculture en Côte-d'Ivoire
B.P. 494
Bouaké, Côte-d-Ivoire

1. INTRODUCTION

Les cages décrites ci-après sont inspirées du type de cage utilisé à Kossou par l'AVB¹ (Campbell, 1978). On a également tenu compte des expérimentations faites par le CTFT² en lagunes côtières (Magnet, 1980).

L'objectif est de mettre au point des cages d'élevage robustes, coûtant le moins cher possible, utilisant au maximum les matériaux locaux, aisément maniables et exploitables par les aquiculteurs ruraux.

Cette mise au point a été faite dans le cadre du développement de la pèche, de la pisciculture, de l'aquaculture en Côte-d'Ivoire par le Projet PNUD/FAO/MINEFOR/IVC/77/003.

2. TYPES DE CAGES

Deux types de cages sont considérés suivant les caractéristiques limnologiques des plans d'eau supportant les cages:

• un type de cage offrant une grande surface d'échange à l'interface air/eau et une faible profondeur (300 × 400 × 185 cm);

• un autre type offrant une faible surface d'échange et une grande profondeur (275 × 275 × 310 cm).

Le premier type est utilisable dans les plans d'eau peu oxygénés en profondeur et à forte turbidité.

Le second type est utilisable dans les plans d'eau encore bien oxygénés 4 m en dessous du niveau de l'eau et à faible turbidité.

Deux types de filets sont utilisés afin d'en tester la solidité et la durée de vie (filet sans noeud = maille 14 mm, fil 210/18 ou 210/48). La différence entre les deux types de filets est la grosseur du fil (120/18 et 210/48) rendant le filet en fil 210/48 deux fois plus lourd que le filet en 210/18.

Nous avons donc construit 4 types de cages:

• type 1: dimensions 275 × 275 × 310 cm filet sans noeud, maille 14 mm, fil 210/18

• type 2: dimensions 300 × 400 × 185 cm filet sans noeud, maille 14 mm, fil 210/18

• type 3: dimensions 275 × 275 × 310 cm filet sans noeud, maille 14 mm, fil 210/48

• type 4: dimensions 300 × 400 × 185 cm filet sans noeud, maille 14 mm, fil 210/48

Le montage des filets en 210/18 se fait à 50 %, c'est-à-dire qu'une nappe de 100 m (mailles étirées) est montée sur une ralingue de 50 m.

Le montage des filets en 210/48 se fait à 71 %, c'est-à-dire que la nappe de 100 m (mailles étirées) est montée sur une ralingue de 71 m.

Ces différents montages permettent une plus grande solidité car les tractions sont réparties sur plusieurs mailles.

¹ AVB = Autorité pour l'Aménagement du Bandama
² CTFT = Centre Technique des Forêts Tropicales

3. CONSTRUCTION DES CAGES

3.1 Confection du cadre

Le cadre soutenant la cage (poche de filet) est constitué d'une armature de chevrons de bois blanc (Samba) de section 6 × 8 cm ou 8 × 8 cm suivant la disponibilité sur le marché.

Les flotteurs sont des fûts en plastique de 60 1, disponibles sur le marché local ou aux usines de filatures et tissages de Gonfreville/Bouaké.

L'assemblage est exécuté suivant les figures 1, 2 et 4.

Les montants du cadre en bois sont maintenus entre eux à l'aide de boulons de Æ 12 mm et de longueur 140 mm. Les montants sont montés les uns sur les autres et non assemblés par mortaises, ce qui permet une plus grande solidité (figure 4).

Le maintien du filet sur le cadre se fait avec des pitons à vis placés tous les 30–40 cm.

Les fûts en plastique sont fixés sur le cadre à l'aide de cordes en nylon ou en polypropilène (figure 4).

3.2 Montage des poches

Cage N^o 1

La découpe de la nappe de filet en fil 210/18 se fait selon la figure 6, afin de perdre le minimum de filet.

Si le montage se fait à 50 %, la maille a la forme d'un losange dont les dimensions sont les suivantes:

côte = c = 14 mm petite diagonale = c = 14 mm grande diagonale = cV3 = 24 mm

Sachant que l'on monte le filet dans le sens de la nappe sur la hauteur de la cage, et dans le contresens sur les autres dimensions de la cage, il faut tenir compte de la forme de la maille qui est un losange avec des diagonales différentes.

Côte à 275 cm (contresens du filet)

• nombre de mailles = = env. 196 mailles

Côté à 310 cm (sens du filet)

• nombre de mailles = = env. 130 mailles

¹ On entend par armement, une portion déterminée de la ralingue devant soutenir un certainnombre de mailles à l'aide d'un fil de même grosseur au minimum que le filet et que l'onattache sur la ralingue (figure 5)

Figure 1 Assemblage du cadre - cage types 1 et 3

Figure 2 Assemblage du cadre - cage types 2 et 4

Figure 3 Coupe transversale d'une cage

Montage d'un flotteur

Figure 4 Détails sur l'assemblage du cadre

A. Fixation du filet sur la ralingue par armements

B. Fixation du fil sur la ralingue

Figure 5 Fixations

Figure 6 Découpe de la nappe de filet 210/18 pour le montage de la cage (1)

Pour le fond, il faut tenir compte aussi du sens de la nappe car pour une même longueur le nombre de mailles y sera différent à cause de leur forme.

On aura donc, pour deux côtés du fond carré: = 114 mailles et, pour les deux autres côtés: = 196 mailles.

Pour les deux autres côtés à 114 mailles, il faut réaliser le montage des parois (196 mailles) avec le fond (114 mailles) de la manière indiquée ci-dessous:

NB: La ralingue verticale des 4 coins de la cage doit dépasser le fond de 50 cm afin de pouvoir accrocher le bloc en béton permettant de tendre le filet. Il n'est pas nécessaire de monter une ralingue sur le fond de la cage.

Cage N^o 2

La découpe de la nappe de filet 210/18 se fait selon la figure 7.

Comme pour la cage N^o 1, le montage sur la ralingue (Æ 6 mm) se fait à 50 %.

Le nombre de mailles sur les différents côtés est le suivant:

côté 300 cm (sens du filet)

• nombre de mailles = = env. 214 mailles

côté 400 cm (sens du filet)

• nombre de mailles = = env. 286 mailles

Côté 185 cm (contresens du filet)

• nombre de mailles = = env. 79 mailles

Pour le fond, il faut découper un morceau de filet de 286 mailles dans le sens du filet sur 125 mailles dans le contresens du filet (figure 7).

Le montage se fait de la façon suivante:

joindre maille à maille les 2 parois à 286 mailles avec les deux côtés du fond à 286 mailles

joindre les deux parois à 214 mailles avec les deux côtés du fond à 125 mailles comme indiqué ci-dessous:

Figure 7 Découpe de la nappe de filet 210/18 pour le montage de la cage (2)

Figure 8 Découpe de la nappe de filet 210/48 pour le montage de la cage (3)

Figure 9 Découpe de la nappe de filet 210/48 pour le montage de la cage (4)

Cage N^o 3

Les dimensions sont les mêmes que celles de la cage N^o 1, mais le filet utilisé est différent (filet sans noeud, maille 14 mm, fil 210/48).

Le montage sur la ralingue Æ 6 mm se fait à 71 %, c'est-à-dire que la maille a la forme d'un carré dont les caractéristiques sont les suivantes:

côté = c = 14 mm

diagonales = cV2 = 19,8 mm

La découpe de la nappe de filet 210/48 se fait selon la figure 8.

Le nombre de mailles sur les différents côtés est le suivant:

côté à 274 cm (contresens du filet)

• nombre de mailles = = env. 138 mailles

• armements 23 fois 12 cm avec 6 mailles

côté à 310 cm (sens du filet)

• nombre de mailles = = 156 mailles

Pour le fond, il suffit de rabattre le morceau de 138 sur 138 mailles et de joindre les 3 côtés restants de la paroi, maille par maille, avec les 3 côtés libres du fond.

Cage N^o 4

Les dimensions sont les mêmes que celles de la cage N^o 2 mais le filet utilisé est différent (filet sans noeud, maille 14 mm, fil 210/48).

Le montage sur la ralingue (Æ 6 mm) se fait à 71 %. La découpe de la nappe de filet 210/48 se fait selon la figure 9.

Le nombre de mailles sur les différents côtés est le suivant:

côté 300 cm (contresens du filet)

• nombre de mailles = = env. 152 mailles

côté 400 cm (contresens du filet)

• nombre de mailles == env. 202 mailles

côté 185 cm (sens du filet)

• nombre de mailles == env. 93 mailles

Pour le fond, il suffit de replier celui-ci comme indiqué sur la figure 9 et de joindre maille à maille les 3 côtés libres de la paroi et du fond.

4. COUT DES DIFFERENTS TYPES DE CAGES

4.1 Prix unitaire

Les nappes de filet vendues par COFIPECHE à Abidjan ont les dimensions suivantes:

longueur mailles étirées = 50 m

chute (hauteur de la nappe) = 400 mailles = 3,5 m ce qui donne une surface de nappe de 175 m².

4.2 Tableau récapitulatif - Prix coûtant des différents types de cages construites par le Projet PNUD/FAO/MINEFOR/IVC/77/003

5. ELEVAGE

Les données qui sont indiquées ci-après sont des données de base issues des travaux de Campbell (1978), à partir desquelles un certain nombre de variantes sont en cours d'expérimentation, tant en ce qui concerne la mise en charge que la formulation et la périodicité de distribution des aliments.

5.1 Mise en charge

Elle est basée sur une densité de 120 fingerlings de 20 à 30 g par m³ de cage avec une fluctuation possible en fonction de la teneur de l'eau en oxygène et du rapport interface/volume.

Elle varie donc présentement de 2 000 à 2 400 sujets par cage de 20 m³ représentant un coût de CFA.F. 12 000 par cage, soit US$ 57.

5.2 Nourriture

Elle est composée de granulés légers fabriqués par AFRIPORC à Dimbokro et, depuis 1980, par l'usine de Bouaké/Manikro.

La formulation de base est la suivante:

• farine basse de riz 65 %

• tourteaux de coton 18 % Coût variant de CFA.F. 50 à 75 le kg (US$ 0,23 à 0,35)

• rémoulage de blé 12 %

• farine de poissons 4 %

• os calcinés, broyés 1 %

Le QN varie suivant les situations, de 2,02 à 3.

5.3 Entretien et gestion

Entretien forfaitaire:

Par cage et par récolte, il faut compter: 1 chevron, 1 m² de filet, 1 flotteur, du fil de ramendage, soit CFA.F. 3 000 (US$ 14,25).

Gestion:

main-d'oeuvre pour l'alimentation: 2 h/jour, soit, par mois,
sur 25 jours = 50 heures pour l'alimentation
                          8 heures pour la vidange et l'entretien
                        58 heures

Côut de la main-d'oeuvre 58 × 70 = CFA.F. 4 060 (US$ 19,35)

5.4 Rendement

• Empoissonnement = 2 400 fingerlings de 25 g = 60 kg

• Mortalité = 5 %, reste à la vidange = 2 280 unités

• Durée du cycle = 5 mois (parfois moins)

• Gain de poids = 570 kg - 60 kg = 510 kg

• Poids moyen du poisson = 250 g

• QN moyen = 2,5

• Quantité d'aliments à prévoir, par cycle et par cage 510 kg × 2,5 = 1 275 kg

• Recettes moyennes à la vente = 570 kg × CFA.F. 300 = CFA.F. 171 000 (US$ 814,25)

BIBLIOGRAPHIE

Campbell, D., 1978 La technologie de construction des cages d'élevage de Tilapia nilotica (L) dans le lac de Kossou. AVB Rapp.Tech., (45)

Campbell, D., 1978 Formulation des aliments destinés à l'élevage de Tilapia nilotica (L) en cages dans le lac de Kossou, Côte-d'Ivoire. AVB Rapp.Tech., (46)

Centre Techniques des Forêts Tropicales, 1964 Confection, montage et réparation des filets de pêche. Publ.Cent.Tech.For.Trop., (25)

Magnet, E., 1980 Développement de l'aquaculture lagunaire. Ministère de la Production Animale. Rapport d'activités 8.10.80, Nogent sur Marne, Centre Technique des Fôrets Tropicales

ASPECTS OF FISH CULTURE DEVELOPMENT IN MALAWI AND ITS POTENTIAL ECONOMIC IMPACT ON A RURAL ECONOMY

ASPECTS DU DEVELOPPEMENT DES PECHES AU MALAWI ET SON IMPACT ECONOMIQUE POTENTIEL SUR L'ECONOMIE RURALE

by

O.V. Msiska
Fisheries Department
P.O. Box 44
Domasi, Malawi

1. INTRODUCTION

Fish culture in Malawi should be reappraised in line with current trends of fish production and its potential economic impact be re-examined. Fish production in Malawi fluctuates between 68 000 t and 84 000 t (FAO, 1978) of which only about 0.14 percent (100 t) are produced through fish culture. The overall fishery has shown a tendency to stagnate since the all time high of 1972 and in some specific fisheries (i.e., Clarias, Barilius and Labeo), notable declines have been reported (FAO, 1976).

Data based on the National Sample Survey of Agriculture of 1968/69 show that in Malawi at least 76.7 percent (wet weight basis) per caput animal protein consumption comes from fish. The consumption fluctuates between 14 and 16 kg. However, because the maximum sustainable yield of the natural fishery appears to have been reached, there is a pressing need to develop supplementary sources of fish protein, if consumption is to keep pace with the annual population growth of 2.6 percent (Malawi, National Statistics Office, 1977).

Experience gained since the 1950s in fish culture has shown the merits and failures of different development strategies. In ponds which are too small and where appropriate management skills are lacking little success has been obtained, while in bigger estate-type ponds there has been some measure of success. A fuller discussion of these may be found in reports by Jones (Unpublished UNICEF report, 1979) and Pruginin and Arad (1977).

In the present report some features arising from an economic analysis of a selected production trial are high-lighted using the monoculture of Sarotherodon shiranus chilwae. Attention is drawn to comparative economic returns with other agricultural crops to illustrate the potential role of fish farming in a rural economy in Malawi. Certain factors limiting fish production are also mentioned where these have a significant bearing on economic returns.

The use of the exotic fish species (Cyprinus carpio) in polyculture was found to be very promising, its acceptability among Malawians is excellent hence the author believes time is ripe for quasi-commercial and commercial fish farming units with emphasis on some form of integration.

2. FIELD STUDIES

2.1 Sites

Initially, Domasi (Zomba) was used as a government development and extension centre after Nchena-chena (Rumphi) was abandoned. At this site various sizes of ponds ranging from 0.06 to 0.75 ha have been constructed for fingerling production and grow-out. Pond construction has been manual utilizing the available labour force. This serves to demonstrate to local farmers the cheapest techniques used in construction. However, this is a painfully slow process and only a limited size of ponds can be constructed.

Kasinthula Fish Farm (Chikwawa) was started in 1970 to investigate estate-type fish farming through funding by FAO. It was hoped that pond productivity would be raised through the use of modern techniques of production, and later the site would be a base for training farmers. Eleven l-ha ponds were built; two 5-ha; two of 0.05 ha and ten of 0.1 ha, each pond was later added. Construction was by means of heavy farm machinery. Through demonstration, both the Government and FAO thought that commercial techniques of fish farming would spread to the surrounding area as in the ‘oil spot’ phenomenon. At Satemwa Tea Estate (Thyolo), six ponds of 0.15 to 0.40 ha were built by the Tea Company using available farm machinery. The site was carefully selected such that existing walls of a gorge were used and dykes were built across the walls. Here the fish were to be fed to employees as an incentive to attract the generally scarce labour force needed in tea plantations.

The OXFAM/Christian Service Committee began a pilot extension integrated fish farming project in 1975 to enable the construction of fish ponds at various selected sites including Chingale (Zomba). Hired manual labour was used in the construction work. The growout pond of 0.30 ha at Chingale cost Mlwi.K 178 and this was supported by other smaller ponds for breeding.

2.2 Pond Management

Stocking of grow-out ponds at the two government fish farms was normally around the optimum of 8 000–10 000 male Tilapia (Sarotherodon) per ha. The optimum stocking density was obtained through trials conducted at the respective government stations (Msiska, unpublished, research report, 1978). Males are easy to sex once they reach 20–30 g and depending on market demand, harvesting is done when the fish are at least over 150 g. This may be regarded as marketable size, however smaller fish are accepted when salted and smoked.

Satemwa and Chingale were both stocked by extension staff from Domasi. Satemwa now conducts its own stocking through the advice of the extension service from Domasi. At Chingale, government staff still supervise all stocking. It has been noted that mistakes made during stocking may later render results discouraging hence the need for close supervision at this stage of the production cycle. Evidently, this requires experience and is time consuming, if done on a big scale.

Depending on availability, manure and fertilizers have been added generally only as light applications. Feeding is carried out once daily (six days a week) using maize waste at 4 percent of the total fish body weight. It is still not known whether the food has any worthwhile nutritive value, or if it acts more as a fertilizer.

The major fish species used are Sarotherodon (or Tilapia) mossambicus and Sarotherodon shiranus chilwae. Tilapia rendalli is also occasionally used plus Clarias geriepinus, the latter may invade ponds from the wild. Sarotherodon mossambicus is raised only at Kasinthula and may eventually be used throughout the Shire Valley.

2.3 Data Analysis

Since stocking rates and feeding rates were the same it was felt that yields could be compared directly. The pumped water at Kasinthula requires that production be substantially raised before a meaningful economic appraisal is possible (Meecham, 1976), it was therefore left out of this analysis. The basic budget analysis developed for channel catfish by Crawford and McCoy (1977) was used for the Chingale operation with appropriate modifications. Rental for the seine net and cost of supervision by government staff were not included.

3. RESULTS AND DISCUSSION

3.1 Monoculture

Yields of up to 3 t/ha have been obtained at Kasinthula using Sarotherodon mossambicus However, because the logistics of the original plan did not materialize, i.e., lack of gravity fed water, the system is still far from viability. Elsewhere, the production of Sarotherodon shiranus chilwae were of the same order of magnitude as S. mossambicus at Kasinthula. The highest yields were attained at Satemwa Tea Estate where 1 014–5 863 kg/ha in 250 days were obtained. Incidentally, this is the only site where growth of over 1 g/day was recorded with S. shiranus chilwae. Domasi gave the lowest yields of 500–1 600 kg/ha. Yields from Chingale were intermediate between Domasi and Satemwa ranging from 1 366 to 2 262 kg/ha. Costs and returns for the operation at Chingale are given in Tables 1, 2 and 3.

Table 1

Costs/returns of fish production from a 0.30-ha pond at Chingale in 1977

Table 2

Costs/returns of fish production from a 0.30-ha pond for 1978

Table 3

Costs/returns of fish production from a 0.30-ha pond at Chingale in 1979

3.2 Polyculture: The Role of Cyprinus carpio vas specularis

Since the introduction of the Israel mirror carp to Malawi in 1976 a repeatable spawning technique has been developed.

Some preliminary growth trials have been conducted and are given in Table 4.

Schroeder (1974) reported rates of up to 5 000 fish/ha in Israel where growth was 25–100 percent faster in manured than in unmanured ponds. Carp of 18 g averaged 3.3 g/day in 40 days. The present studies were started with the understanding that both Sarotherodon and Cyprinus carpio may be raised on manure and that with rising energy and fertilizer costs, this may be a more viable alternative method of fish production in Malawi.

Indications were that even at light stocking densities of mirror carp fish production in ponds was augmented and that judicious addition of manure substantially increased production. Better planned studies are currently underway to confirm these observations, especially on the rate of manure application and on the influence of stocking density on growth of mirror carp.

Table 4

Yields of Cyprinus carpio vas specularis and Sarotherodon shiranus chilwae at Domasi Fish Farm

4. OTHER AGRICULTURAL CROPS

Richards (1978) used the gross margins (gross value of output minus its variable costs of production) as a criterion for evaluating returns of agricultural crops in the Henga/Kasitu Valley. He showed that per ha, with or without fertilization, respectively, maize produced Mlwi.K 42.39 and Mlwi.K 89.64, maize combined with pulses gave Mlwi.K 82.24 and Mlwi.K 69.83, while millet produced Mlwi.K 13.03 and Mlwi.K 3.70. Groundnuts alone produced between Mlwi.K 52.27 and Mlwi.K 71.98 and when combined with maize Mlwi.K 79.48 was realized. Other results did not vary much from these.

A gross economic analysis of Malawi tea by Tyler (1978) indicates that the price of the finished product (processed and packed) on the world market varied between Mlwi.K 272/ac in 1978 and Mlwi.K 597/ac in 1976. The raw product earns between Mlwi.K 100/ha and Mlwi,K 200/ha according to Jones (Unpublished UNICEF Report, 1979).

5. ECONOMIC ANALYSIS

The basic budget developed by Crawford and McCoy (1977) for determining expected expenditures and returns for culture of channel catfish in watershed ponds was, with minor modifications, suitable for estimating profits from single crops of Sarotherodon shiranus chilwae males at a subsistence scale (at Chingale) in 1977. Partial budget analysis, which adjusted for added cost of stock, feed and added receipts for increasing sales were conducted for 1978 and 1979 yields, as shown in Tables 2 and 3. The net gains per 0.3 ha were Mlwi.K 28, 81 and 158, respectively. Results of fish yields obtained by this local farmer indicated that subsistence fish farming can be economically viable if support services are readily available and if the pond system includes breeding and nursery ponds (500 m²). The increased returns in subsequent years were commensurate with the expected rise in management skills of the farmer and his family. The highest single cost was due to pond construction. For accounting purposes, the pond life was assumed to be 20 years, although it is well known that the pond does not have a defined lifetime. The Department is in future urged to provide a team of ox-drawn scoops and scrapers for rental to fish farmers as a means of reducing the costs of construction or renovation.

6. CONSIDERATIONS FOR PRACTICAL APPLICATION OF RESULTS

This study suggests that problems with low yields would be uncommon if effort is made to stock only male fish, a system which may require that breeding be conducted on site or alternatively that a reliable seed source be established. Particular attention should then be paid to the size of grow-out ponds because if it is conducted as a business then economies of scale become important. For instance, production of Sarotherodon mossambicus males suggests that fish may grow better in a 3.3-ha pond than in 1.0-ha ponds (Unpublished results). In general, this study goes some way in support of practical proposals put forward by Linn (1974), that:

1. Extension encourage proper pond management on a local level, and

2. Efforts be directed to make fish culture an economically viable investment and endeavour.

Although mirror carp has only been recently introduced, it promises to make significant contribution to pond yields. Table 4 shows that yields of up to 7 186 kg/ha are possible with increased use of manure (up to 5 704 kg/ha/month). The Malawian consumers cherish this fish for the table and therefore an added effort should be made to raise the fish. The author believes that apart from producing a steady yield of fish, Malawi could usefully develop commercial fish culture units utilizing Sarotherodon and Cyprinus carpio vas specularis. In support of this, it has been shown that natural fish stocks are not inexhaustible and there is a general rise in fish prices at main markets (at least there has been an addition of Mlwi.K 0.05/fish from 1979 to 1980). Furthermore, the culture production cycle is fairly simple utilizing available agricultural by-products (maize waste, rice bran, maize bran, cotton seed cake, groundnuts meal, etc.). Clearly at a subsistence level, some form of horizontal integration with other livestock (chickens, pigs, cattle, etc.), will be to the mutual benefit of the otherwise separate enterprises.

Where conditions are right fish culture may just be as economical as other agricultural crops like maize, millet, groundnuts and pulses as this survey shows. Probably what is also significant about fish culture is that people far removed from the main lakes are not adequately provided with fish.

While estate-type farming is the better direction for future development, use of pumped water for fish production is economically indefensible, and contrary to the view expressed by Meecham (1976), the present author believes that use of labour-intensive methods remains the better alternative in view of increasing energy costs.

7. RECOMMENDATIONS FOR FUTURE DEVELOPMENT

Like other agricultural sectors, there is a need to establish an infrastructure for research and development. Country-based experimental trials will have to be expanded to include other regions of Malawi. In the absence of an appropriate training institute in the country for professional and technical staff, financial help should be sought to train personnel in such areas as:

1. Seedling production and hatchery management
2. Water quality
3. Fish nutrition
4. Aquaculture economics

The Agriculture Rural Development strategy should provide growth centres where fish farming will eventually play a role in the economy, but not without careful consideration of sites in terms of availability of water, essential support services and markets.

Probably the programme outlined by Linn (1974) still remains the most promising course of action for the extension section of fish farming. However, there is now a stronger need for bilateral financial support to rationally develop this renewable resource and a project has been proposed to the Government for long-term development. Malawi may be particularly suited for this kind of development because it is endowed with a diverse freshwater system as at least 21 percent of the country is covered with water.

REFERENCES

Crawford, K.W. and E.W. McCoy, 1977 Budgetting for selected aquacultural enterprises. Bull.Agric. Exp.Stn.Auburn Univ., (495)

FAO, 1976 An analysis of the various fisheries of Lake Malawi. Report prepared for the Government of Malawi by FAO of the UN acting as executing agency for UNDP. Based on the work of J. Turner. Rome, FAO/UNDP, FI:DP/MLW/71/516, Technical report 1:73 p.

FAO, 1978 Yearbook of fishery statistics. Annuaire statistique des pêches. Anuario estadistico de pesca. Catches and landings. Captures et quantités débarquées. Capturas y desemborques. Yearb.Fish.Stat./Annu.Stat.Pêches/Anu.Estad.Pesca, (44): pag. var.

Jones, T., 1979 OXFAM Extension Fish Farming Project. Final report. Fisheries Department

Linn, D.W., 1974 Development of fish farming in Malawi. Fish Farming Int., (3)

Malawi, Fisheries Department, Marketing and Statistics Section, Annual report, 1976 and 1977. Annu.Rep.Fish.Dep.Malawi, (1976/7)

Malawi, 1970 National Statistics Office, National sample survey of agriculture, 1968/69. Zomba, Government Printer

Malawi, 1977 National Statistics Office, Population census, 1977. District population by traditional authorities. Preliminary report. Zomba, Government Printer

Meecham, K., 1976 Aquaculture in Malawi. CIFA Tech.Pap./Doc.Tech.CPCA, (4) Suppl. 1:66–72

Pruginin, Y. and A. Arad, 1977 Fish farming in Malawi (3). a report for the promotion of Integrated Fishery Development Project. Rome, FAO, FI:DP/MLW/71/516/10:22 p.

Richards, E.M., 1978 An exercise in the use of gross margins for the analysis of farm management survey data (Henga/Kasitu area). In Symposium on Social Aspects of Rural Development in Malawi. Zomba, Malawi, University of Malawi

Schroeder, G.L., 1974 Use of fluid cowshed in fish ponds. Bamidgeh, 26(3):83–96

Tyler, G.P., 1978 Malawi's interest in an international tea agreement. In Symposium on Social Aspects of Rural Development in Malawi, Zomba, Malawi, University of Malawi