3. AQUACULTURE: STATUS AND DEVELOPMENT (Contd.)

3.3 Fish culture development with particular reference to the Lake Basin

3.3.1 Major reasons for past failures. Introduced in Kenya nearly 30 years ago, fish culture has remained until now practically undeveloped. A series of reasons for this have been advanced in various reports by Government and previous missions (e.g. Kenya, 1979c; FAO/UNDP, 1966; Coûteaux, 1981; Anon. 1980). Combined with our own observations, these major reasons for past failures of small-scale rural fish farming development may be summarized as follows, not necessarily in order of importance:

1. inadequate recurrent budget support resulting, for example, in reduced extension staff mobility and reduced staffing;

2. fish culture research of little practical importance, not providing the necessary information for extension support;

3. wrong fish species selection particularly in cooler regions unsuitable for warm-water fish culture development: even in suitable areas, selection of a slow-growing species such as T. zillii;

4. introduction of fish farming into regions where fish is not traditionally part of the diet, without a strong, long-term commitment to development;

5. wrong approach to rural development: free distribution of fingerlings, low enforcement and extension combined, individual initiatives undeveloped;

6. lack of adequate technical support to fish farming development resulting in bad site selection (acid soils, inadequate water supply), bad farm design (faulty water supply), bad construction records (excessive cost, lack of bank slopes), and bad pond management (stocking rates, organic fertilization, feeding, length of rearing period);

7. pond sites difficult of access (deep valleys) and very dispersed;

8. lack of properly trained personnel, both cadres staff and extension workers;

9. instability of staff due to frequent rotations, resulting in discontinued programmes/efforts;

10. lack of collaboration and coordination between planning, research and management staff, impairing the flow of information.

Some of these sectors will be analysed in more detail in the following sections, on the basis of available information.

3.3.2 Present effectiveness of fish farming extension personnel. Two sets of data are available to estimate the present effectiveness of fish farming extension personnel in the Lake Basin. The Kakamega District monthly reports of the fish scouts involved with such duties were summarized for 1979–1982 (Table 22) and a questionnaire survey of six Peace Corps Volunteers with at least one year of field experience in the Western and Nyanza Provinces was conducted during their visit to the Baobab Fish Farm in Mombasa, on 26 April 1982, (Table 23).

Results mainly point out to the fact that extension personnel effectiveness may be greatly increased by the availability of such transport means as a small motorcycle. On the average, a fish scout visits 6.2 – 8.3 ponds per month and the frequency of his visits to the same fish farmer does not exceed once a month. In comparison, a highly mobile volunteer visits 20 – 40 ponds per month and his visiting frequency may exceed twice a month. The volunteers' ponds are also fished more regularly (at 7 – 8 month interval) and average fish production may be thus increased from 0.3 to 1 – 2 t/ha/yr.

The data also show that most of the fish produced is consumed locally, benefitting on the average 8.2 – 14.8 people per pond. About 40 – 50 percent of the fish are sold by the farmer at a relatively high price (Ksh. 3.3 – 7.9/kg in 1981).

3.3.3 Cost of pond construction. Available records of DDC and Government funds allocated for fish pond development in Kisii, Siaya and South Nyanza Districts (Nyanza Province) show a very high and rapidly increasing average cost of pond units (Table 24 and Fig. 15). Since the period 1974–1977, the average cost per 400 m² pond has increased from Ksh. 13 000 to 54 000. The present average size of ponds does not exceed 0.04 ha which suggests an equivalent construction cost in 1981 of at least Ksh.1.25 million per hectare.

More than one-third of this cost consists in labour cost (Table 25). Construction is done by hand and it is often not well supervised. This results in slow progress and faulty design. One of the consequences is that out of 32 pond development projects (1980–1982), 43.7% of them remain now unfinished due to excessive costs and insufficient funds (Table 26).

Recent rises in pond construction cost have been drastic (Fig. 15). If a similar trend persists, future cost may be estimated from log Y = 474.3. log X - 1.559 (r = 0.99), where Y is cost in thousands of Ksh. and X is time expressed as years (e.g. 1989). On this basis, the average pond cost would reach Ksh. 360 000 by 1989.

3.3.4 Financial support for fish farming development. The Agricultural Finance Corporation (AFC), a commercial bank founded in 1963, is the main channel for small and medium-scale farmers to obtain financial loans. Only a couple of loans have been given for fish farming development (Anon. 1980).

More financial support for fish farming development originates from Government loans for demonstration ponds and from District Development Committees for communities such as churches and schools. In the Nyanza Province alone for example, Ksh. 1 726 500 have been made available in 1980–1982 through these two sources for the construction of 32 ponds (Tables 24 and 26), a substantial increase over the previous periods.

Table 22

Fish production statistics, Kakamega District, Western Province, 1979–82
(Acc. Kakamega District monthly report by fish scouts)

Table 23

Survey of Peace Corps volunteers on fish farming extension ^a

^a Volunteers with at least one year of field experience

Table 24

Average 400 m² pond cost in the Nyanza Province based on DDC Projects, 1974 – 1982

Table 25

Estimated cost breakdown of fish pond construction in Bungoma
(acc. Fisheries Department records)

Table 26

Construction cost of fish ponds based on DDC and Ministry of Finance grants/loans, 1980–82

^a D.F.P. = Demonstration Fish Ponds

Figure 15. Evolution of the construction cost of fish ponds in the Nyanza Province

It would therefore appear that money is available from public funds for this type of fish farming development. Its effective use is however greatly impaired for various reasons, as discussed in the previous sections. It would also appear that sound development and technical criteria for project selection are now generally lacking.

3.3.5 The World Bank assisted project for fish culture development. In 1979, a World Bank was designed to support fish farming development on a national scale (Anon. 1980).

During Phase 1 (Table 16) it was envisaged, as far as fish farming was concerned, to establish as an autonomous entity a Fish Farming Development Centre (FFDC). Sixty seven hectares of ponds and additional facilities for research/training would enable this centre: (a) to produce 300 t/yr of table fish (tilapia and common carp) for sale; (b) to produce fingerlings for distribution to fish farmers; (c) to train fishery personnel; and (d) to carry out research to develop commercial techniques applicable to the development of warm-water aquaculture in Kenya.

Phase I of the project, which was originally planned to last 6 years and to start field activities in July 1980, was estimated to cost Ksh.100 million net of which Ksh.12 million was for the FFDC establishment, Financing is shared by GOK (23%), IDA (75%) and cooperative fishermen (2%). After one year of preparatory actions, the establishment of FFDC was planned to last 2.5 years from July 1981 to late 1983. A tentative site has been proposed in Kabonyo (15 km south of Kisumu), next to the West Kano irrigation scheme. It is a flat area of about 100 ha located close to the lake shore and periodically flooded by the Miriu River. Black cotton soil seems predominant.

After completion of Phase I, it was envisaged that a second phase might directly assist further development through the provision of extension services and expanded training.

At present, implementation of Phase I by the Department of Fisheries has started, the administrative procedure for land acquisition now nearing completion. This should be soon followed by the recruitment of the specialized personnel and the selection of a consulting firm for the construction of the centre.

3.3.6 Miscellaneous foreign assistance for aquaculture development. Apart from the World Bank project, six other assistance projects have an aquaculture component (Table 16). National training facilities are being completed in Naivasha also with World Bank assistance. On the coast, a UNDP/FAO project assists the Department of Fisheries since 1978 to establish a pilot fish farm in mangrove swamps and to demonstrate the economic feasibility of a brackishwater tidal aquaculture system based on shrimp production.

In the Lake Basin, a brief survey was conducted in 1981 to determine the fish farming potential and to propose a rehabilitation scheme (Coûteaux, 1981). Applied research has been planned with IDRC assistance on the use of irrigated rice fields for fish culture and cage culture in irrigation canals. Finally a limited assistance is provided to the Department of Fisheries by foreign volunteers to develop small-scale rural fish culture through direct extension support.

The above projects are limited in scope and/or in means, particularly in the Lake Basin where most of the past efforts of the Kenya Government for fish culture development have been concentrated and where so little positive achievements have been realized to date.

3.4 Potential for future aquaculture development

A great potential exists in Kenya for aquaculture development, particularly for fresh-water fish farming. Subject to water availability and according to the proposed climatic zonation (Sec. 3.1.3), pond culture could be considerably developed to produce both warm-water and coolwater fish, such as tilapia, Chinese carps, common carp and trout. Previous proposals in this direction have been made long ago (FAO/UNDP, 1966). Although there is a need for revising these proposals to take full advantage, for example, of the latest developments in tilapia and carp culture under African conditions, the past emphasis put on applied research and specialized training particularly is still valid today. The major reasons for past failures (Sec. 3.3.1) are well known, not only from field experiences in Kenya, but also in most other developing countries of the world.

Future fish farming development plans should not limit themselves to the production of fish in ponds. Although pond culture still remains the best available system for small-scale rural development, other cultural systems are available. Some of them give the additional opportunity to integrate fish production (and fish feed production) with agriculture, such as rice-cum-fish culture, fish cage culture in irrigation canals, and tank culture. Fish-pen culture in shallow bays may be practised in natural lakes and particularly in man-made lakes with an indented shoreline. Intensive tilapia tank culture, developed in Kenya at the Mombasa Baobab Farm in brackishwater, could be adapted and further expanded to fresh-water environments. In all cases, however, sound feasibility studies are required before embraking on the implementation of the necessary pilot scale projects.

The above considerations apply also to the Lake Basin. Several large-scale irrigation schemes exist or will be developed (Tables 2 and 3) which provide opportunities for integrated fish production. For example, rice and fish could be cultivated either in rotation or in combination (with adequate pesticides) in the irrigated fields. It has been shown that the simultaneous stocking of fish among rice increases paddy yields by 5 – 15% (Coche, 1967) and this has been already observed in Ahero (Enderlein, 1974). A mean fish yield of at least 300 kg/ha/yr may then be obtained in addition to the paddy crop (Table 3). Suggestions have also been made by Baobab Farm Ltd. on the possible use of tank and cage culture for the intensive production of tilapia at the West Kano Pilot scheme (R. Haller, pers.com.).

A complementary feasibility study (particularly concerning local fish feeds availability) as well as the realisation of such two pilot scale projects should be considered in the future following the establishment of the Fish Farming Development Centre (Sec. 3.3.5). Several man-made lakes (Table 2) could offer additional potential sites for further development of cage culture, on the basis of its established feasibility in the West Kano pilot scale project and complementary local studies.

The potential for further development of pond culture within the Lake Basin development area appears to be limited. Restricted land availability (Sec. 2.5.1) might confine future large-scale pond culture enterprises to the swampy areas of the warmwater lake shores. Even there, conflicts might arise with agricultural development. Small-scale pond culture has a greater potential for development, particularly when one considers its important nutritional and socio-economical impacts in inland rural areas.

A close collaboration with the livestock development programme should ensure that one takes full advantage of the increased potential of integrated fish-animal husbandry systems. Such systems are particularly adapted to the farming of microphagous fish species such as T. nilotica, without any supplementary feeding. The most common integrated systems combine fish production with pig, chicken or duck rearing, but other animal manures (sheep, goat, cow) may also substantially increase fish productions, either directly or through composting.

3.5 Conclusions and recommendations for aquaculture development in the Lake Basin

3.5.1 Short-term development programme for aquaculture (up to 1986).

(a) It is planned to build a large-scale pilot farm near Kisumu, in the warmwater zone, with World Bank assistance. Although this project is already close to one year behind schedule, it is still hoped that farm construction will start in mid-1983. Fish rearing and seed producing facilities would then become available in 1986 (Fig. 17). In particular, tilapia and common carp fingerlings will be produced for distribution to the fish farmers (Sec. 3.5.2).

(b) Greatest potential exists for small-scale pond culture development in the future. Numerous individual ponds already exist in the inland rural areas, but their fish production is insignificant. It is estimated that at least 50% of them could be rehabilitated and that their production could thus be progressively increased by a factor of 6 to 7, to finally average 2 000 kg/ha/yr. In Phase I of the fish rehabilitation programme yields could be increased from 300 – 1 200 kg/ha/yr, maximum yields being attained during Phase II (Fig. 17).

(c) Selected Government demonstration ponds and DDC community ponds should become the focal points of future development, being used not only for positive demonstrations but also for fingerling production and for practical training. Particular efforts should be made to involve primary school children in fish production projects, for example through the existing “Wildlife Clubs”.

(d) Rehabilitation should be carefully planned on a regional basis, taking into account land availability and population density/growth. The selection of the ponds to be rehabilitated should be based on strict technical criteria, giving primary consideration to water temperature, water supply, and soil quality.

(e) Although fish pond rehabilitation should proceed in the whole Lake Basin, some priority should be given to the relatively warmer areas for ecological reasons. It is generally considered that the fish cropped in Lake Victoria is marketed at the most 25 – 30 km inland, as proposed in Figure 16. Some priority therefore should also be given to areas located further away from the lake shore. Combining these two priorities, it is recommended that particular emphasis be given to fish farming development in the following areas (Fig. 16): northern Siaya District, Busia District, southern Bungoma District and western Kakamega District, all situated within the fish farming Zones C and D. Some rehabilitation of fish ponds might also be envisaged in the South Nyanza District, where the water supply does not fall below 20°C for any long period of time.

(f) An adequate Extension Service, specialized in fish farming development, should be created for the Lake Basin. Its staff should include field coordinators at the district level (1/1400 – 1500 km²) and extension workers (1/300 km²), carefully selected from these areas, well trained, highly motivated and regularly monitored by executive officers in the field. Full mobility of all staff should be considered as essential, the extension workers and the field coordinators using for example small trail motocycles equipped with carrier and mud-grip tyres.

(g) New fish ponds should be developed only in priority areas and preferably for communities, until the rehabilitation programme of the existing ponds reaches the end of its first phase.

(h) The above policy of development will permit the continued use of tilapia as the primary farmed species while still obtaining acceptable yields. But all efforts should be made to favour the plankton eater T. nilotica rather than the plant eater T. zillii. Although the latter may show a better resistance to cooler temperatures, its growth is slower and yields are reduced. T. nilotica can perform much better, particularly in the presence of “green water”, rich in microscopic organisms on which it feeds constantly. Organic fertilization of the rural ponds should be strongly emphasized, using composting piles of vegetal material enriched if possible with animal manure. The pond stocking rate should be about two fishes per square meter (ab. 2/sq.yd). Selective fishing of the bigger fishes may start after 4 – 5 months, using traps, hook/line or nets. Complete harvesting should be done soon after harvesting, reconditioning the pond and repairing if necessary the water supply canal.

(i) The fingerlings necessary for pond stocking should either be regularly made available to the farmers (at a subsidized price) or preferably be produced by the farmers themselves, together with table fish. A rural trade of T. nilotica fingerlings could even be developed by some farmers, with Government assistance. The main local suppliers however should be Government demonstration farms and DDC community ponds, which should be integrated with the local rural development. Long transports of fingerlings should be avoided, as expensive and time-consuming. Fingerling production centres should be developed in centralized locations, in each area where fish farming development justifies it.

(j) The existing assistance of the U.S. Peace Corps Volunteers should be coordinated and integrated with the above development programme, both for in-service training and for extension work, mainly to avoid duplication of efforts and to ensure a unified approach to fish farming development in the Lake Basin.

Figure 16. Priority areas for fish farming development in the Lake Basin

Figure 17. Fish farming development in the Lake Basin: proposed schedule for implementation.

3.5.2 Medium-term development programme for aquaculture development (1986–90)

(a) By 1986, the Kabonyo FFDC should become functional, supplying tilapia and carp seeds (Fig. 17). From then on, FFDC should become the main support to fish farming development in the Lake Basin, both for large-scale and small-scale enterprises.

(b) A feasibility study should be conducted in 1987, in collaboration with FFDC, for the establishment of a pilot-scale intensive tank system to be integrated with the West Kano irrigation scheme. Both tilapia and common carp could be considered for rearing, according to availability of fish feeds ingredients.

(c) A feasibility study should be also conducted for the direct integration of fish culture and rice farming in the West Kano scheme, using the paddy fields for tilapia/carp production, either in combination or in rotation with rice. A pilot scale project should be initiated, in collaboration with FFDC.

(d) A third feasibility study and pilot scale project should study the potential of tilapia/carp intensive cage culture in the main irrigation canal of the West Kano Scheme, in collaboration with FFDC.

(e) Phase II of the rural pond rehabilitation programme (Sec. 3.5.1) should rapidly progress and reach completion, the maximum potential yield (2 – 2.5 t/ha/yr) being realised in most of the rehabilitated ponds using T. nilotica as the main species.

(f) A Phase III (1987–90) of the rural rehabilitation programme should then be started in the cooler areas (Fig. 16) of the Western Province (eastern Kakamega and northern Bungoma Districts) and of the Nyanza Province (Kisii District). Here the main fish species should be common carp, tilapia becoming secondary because of the lower water temperatures. Average production goal should be at least 1 – 1.5 t/ha/yr. Fingerlings should be produced in FFDC (warmer zone), from where they should be provided periodically to rural relay stations and distributed (at a subsidized cost) to the fish farmers.

(g) Development should intensify, with FFDC support, and new fish ponds should be developed, especially in well defined priority areas.

(h) The Fish Farming Extension Service should progressively grow, proportionally to the increasing importance of the fish farming development. It should become an administrative entity, separate from the Capture Fishery Extension Service, the training requirements and the development problems being totally different. Training of new staff and regular recycling of existing staff should take place according to a programmed schedule, securing in advance the necessary financial means. Staff training, motivation and close supervision (through frequent contacts in their own zone of activity), together with a high staff mobility through motorization, will remain essential requirements to ensure the success of the development programme. On the other hand, field staff rotation should be avoided and restricted to exceptional circumstances.

(i) Fingerling production and distribution (through relay centres) should keep up with the farmers' demand. They should be carefully planned and managed to avoid any break in their supply. Local T. nilotica fingerling production by privates should be encouraged and even financially supported through soft loans. Initially, the production of common carp seeds should be concentrated in FFDC where water temperature conditions will be optimum and where broodstock production/maintenance will be easier to manage. Later, when the technology is well established and when fully trained personnel exist, the creation of secondary stations (public or private) for fingerling production from transported embryonated eggs might be envisaged, to reduce transport costs. But carp egg production should remain concentrated in FFDC for future years, primarily to ensure broodstock quality control.

(j) As soon as FFDC becomes functional, it should obtain from Baobab Farm Ltd. (Mombasa) fingerlings of T. andersonii and T. aurea to develop a broodstock. As these are microphagous tilapia species known for their relatively good growth in the cooler waters of Zambia/Zimbabwe and Israel respectively, comparative trials should be organized in the cooler areas of the Lake Basin to assess their respective value for fish farming development, compared to T. nilotica. If the results show a definite possible improvement, the best of these two species should progressively replace T. nilotica in the fish ponds. Simultaneously, T. zillii should be eradicated from the rural fish ponds.