To consolidate the analysis of production results obtained so far by the project at the Ngomeni aquafarm.
In order to understand the production data, it is necessary to give a summary of the present Ngomeni technology. A technology is a combination of engineering and management; it can be compared to the “hardware” and the “software” in the computer industry.
Generally speaking, aquaculture is to water what agriculture is to land; a way for man to increase the vegetal and animal biomass of his environment. That is why one can expect to find important differences in the technologies used from one country to another and even from one farm to its neighbour. In Ngomeni the basis of the work is a technology traditionally developed in South East Asia, using intertidal exchange of water and the growth of “lablab” in the pond to feed the shrimps.
The Ngomeni site was selected after a first unsuccessful try in a sandy creek, south of Malindi. The present site lies within the concession of a salt operator, but an agreement was found to permit the construction of the pilot station.
The first idea seems to have been to build a large farm on the 50 ha site available but difficulties in building the perimeter dike led to an early renunciation. Then small (1 ha) ponds were built. Over the area conceded to the project various quality of soils exist. Ponds 2,3,4 and 5 are built in a loam and clay area, dried and long stabilized. The soil analysis shows a correct pH with little organics. Ponds 1,6,7,8, and 9, NP are situated in a reclaimed mangrove area where the soil is a mixture of sand and mud, high in organics and in a phase of decomposition.
The production data were obtained from ponds 1,2,3,4,5 and 6 covering a total area of 8.2 ha. The plan of the aquafarm is presented in Annex I. The main characteristic of the pond design is that the exchange of water is done through a single gate. The Project Manager insists on many details of the design which he believes are important for a successful harvest, such as the presence of a central ridge, the depth and the slope of the bottom, the mean elevation of each pond in relation to the tide fluctuations, the presence of branches stuck in the bottom of the pond.
In Ngomeni, the housing facilities consist of a warehouse, an office and an outside shed used for processing the shrimps after harvest. The office for administrative work, the laboratory and the scientific equipment (pH meter, D.O. meter, balance, engineering tools,…), the freezers and the vehicles are located in Malindi.
During the evolution of the project, the personnel has changed from time to time and it is difficult to trace such movements, During the stay of the consultant, the number of staff and workers was estimated at 34. Strange as it may seem, it was not easy to find who was working for the project and even more difficult to find out who was doing what. The list of the personnel attached to the project is given in Annex 1. Additional workers were hired for collecting fry, for harvesting and for construction.
The different phases of culturing shrimps are the preparation of the pond, the stocking of juveniles, the grow-out, the harvest and the commercialization. The rest of the work to be executed in a shrimp farm consists in general maintenance and surveillance, in controlling the evolution of the culture by monitoring some pond and population parameters.
The juveniles are caught in the mangrove by a team of workers, the numbers are estimated and then they are stocked in the ponds. Details have been given in technical reports by the biologist associate expert on the methodology and on the numbers collected throughout the year.
During the grow-out phase, shrimps in the ponds are sampled every 2 weeks by scoopnet. The samples are weighed in Malindi.
On the day of the harvest additional manpower is hired, the shrimps are sorted eventually beheaded or peeled, iced and brought back to Malindi where they are frozen in racks. Later the shrimps are packed in 2 kg plastic bags.
Despite some fluctuations from one experiment to another, which are normal in a research and development phase, the general water management - as it has been understood by the local staff - can be summarized as such:
The Ngomeni staff fills in different record forms - daily record form, biological stocking form, sampling data during grow out, harvesting form - (presented in Annex 1). This data, some technical reports and the quarterly reports are the sources of information used for the present report.
The species selcted is Penaeus indicus, based on the fact that fry are readily available in the area and in any mangrove area along the coast.
Table 1 gives the summary of the 36 “commercial” trials that have been realized in 6 ponds from August 1981 until November 1984. Table 2 gives the average production by pond. Out of 36 grow-out trials, 6 550 kg of shrimps have been produced with an average production figure of 142 kg per hectare of 7.2 g shrimps at a rate of 3 harvest per year. This gives an annual production of 426 kg/ha.
In a production cycle, the pond is dried during 10 days, the lablab is prepared during 15 days and the grow-out period lasts 90 to 95 days.
A closer look at the results (table 2) shows that the average production figures vary from 60 kg/ha for pond no. 6 to 191 kg/ha for pond no. 5. There are two groups of ponds; the ponds (1,6) produced 60 to 65 kg/ha on average and the ponds (2,3,4,5) produced 165 kg/ha with an identical technology. The average weight of individual shrimps harvested in pond 5 reached 8.5 g.
Growth curves are presented in table 3. Curve 24-4-5 represents one of the best productions. It has been noted that some of the good productions and the bigger sizes were obtained over periods of more than 100 days; this may be due to the fact that lablab had had more time to develop in the ponds. The curve 83-2-4 represents a correct average curve; the shrimps start growing in the same way as in the preceeding case but the curve levels off after 2 months in the ponds. The last curve represents a poor growth in pond 1. It can be seen that the growth is slow from the very beginning of the experiment despite a low stocking density, indicating a lack of food. The average growth rate is 0.55 g per week and the best one is close to 1.0 g per week. It is difficult to obtain a deeper interpretation of the evolution of the biomass due to a lack of accuracy of the data.
The average stocking density fluctuates between 20 000 and 40 000 fry per hectare. The survival rate has not been reported on table 1 as it is often over 100 %. (This means that either the numbers of fry actually stocked were higher than recorded, or that wild stock found its way into the ponds through the filters during water exchanges). It is difficult to demonstrate any influence of the stocking density on the growth rates or on the total production recorded.
It is not posible to show any action of fertilizers as the best pond has never been fertilized and as the amount of fertilizers used was very low. The dissolved oxygen level seems never to have been a problem; this is due to the low level of biomass and organics in the water. The Secchi disk readings (50 or 60 cm) indicates a transparent water with little plankton.
The main conclusion that can be read from these data is that the limiting factor in culturing P. indicus in Ngomeni is the amount of food available in the pond. The shrimps are very hardy and seem to resist very well both manipulations and adverse conditions, providing they are correctly fed. So P. indicus is perfectly adapted to aquaculture even when using a very low technology in the Kenyan environment.
From a very strict scientific point of view the “experimental trials” should have been incorporated in the average as all the trials can be considered as experimental and as any shrimp farmer is expected to have a poor crop from time to time. However, in the opinion of the consultant a higher or a lower average number would not have changed the whole meaning of the results.
From a biological point of view, the general environment is suitable for shrimp aquaculture all year round. Soil and water are good and unpolluted. The sea-water is very clean and probably does not bring any significant amount of natural feed in the pond. In the consultant's opinion a water poor in organics is to be preferred to a richer water because the oxygen level of the incoming water is known to be saturated and because the amount of organics in the pond can be precisely controlled, thus the whole production can be secured in a repetitive manner.
The water circulated in the ponds for some 8 to 10 days each month. This is enough to restore the salinity to normal; it takes 2 to 3 days at each spring tide to completely exchange the pond water.
Penaeus monodon and Penaeus indicus, the two local species grow perfectly over 25 C to 35 C, which are the average temperatures recorded in the ponds. Temperatures in the ponds have reached 42 C but no adverse effect have been noted. The salinity is high, from 35 to 38 ppm at sea; lower salinity is usually preferred for shrimp rearing but experience shows that such high salinity, which is in fact the normal sea water salinity, permit good growth though probably slower growth. The average evaporation rate is about 1 cm per day in the warmer season but, at times, the salinity rose up to 60 ppm. The data records did not show any significant adverse reactions by P. indicus.
The technology selected has proved to be effective in Asia. However, as it is a traditional practice there, many improvements have been made and it is recommended that one should use directly such advances without meaning to reinvent something completely new. Many references were available from FAO, such as “Fish Culture in East Africa (1966) 1” about pond constructions and the project had received all the due recommendations in time but did not make much use of them.
It is understood that the accent was put on developing an easy technology with a limited investment. However, a pilot station should have a long term view and what is presented as “saving” sometimes appears later to be unnecessary expense. One example is the perimeter dike and the unused main gate. Both are useless in the present design of the ponds and it can be feared that a single main gate would not be sufficient to provide water for the full 50 ha area. Anyway, the main gate is not necessary for the present technology. Such errors are due to a poor initial planning and to the choice of using manpower not machines for pond constructions even when government machines were available at very reasonable prices.
At the time the operation started, a soil survey of Kenya for agricultural purposes was already completed. A contact with the team in charge of the survey would have been very useful and would have saved a lot of time and expense. The existing sites for aquaculture will be discused in chapter 5.
For a traditional farm, the engineering design is basically correct except for some details that will have to be fixed as soon as possible (chapter 2). It must be noted that the slope of the dikes does not permit to use a level of water higher than 50 cm without creating an important erosion problem.
It is not understood why the Project Manager is so attached to the central ridge system which does not seem to have any part to play in the circulation of water. In fact the production of the ponds may well be significantly enhanced by eliminating that ridge. It is calculated that the areas under 10 cm of water occupy 20 to 25% of the surface of the pond;, eliminating them would probably enhance the pond production by the same value.
At a first glance the results are really positive, 426 kg of shrimps per hectare and per year is really a good result, especially without fertilizers.
However, a closer look at the “technology” shows that the results are more due to a favourable combination than to a true controlled knowledge of what is succeeding in the ponds. The project has failed to gain experience from one experiment to the next, by lack of scientific approach, of basic biological knowledge, and by refusing to put in practice recommendations given by numerous experts.
The data available were correct for salinity, temperature and for the shrimps harvested, but the data about the numbers stocked and the grow-out are biased by manipulation errors. Most of the data were available in an untreated form which means that no interpretation or discussion had been attempted. This is strange as the analysis of the results of an experiment is the only way to improve the following one.
Farming procedures are detailed in chapter 2.
The very first remark concerning the staff and workers is that they are perfectly capable in their duties and that they are willing to learn more about aquaculture and to participate more deeply in the project.
One basic remark of the Accountancy Mission was that there were too many people involved in the project to make it credible and attractive to private investors. This is definitively true. Too many people and a lack of stability of the staff are two handicaps for a correct demonstration. It is understandable that the project should train workers and technicians but this should be done for specified training periods and with a prepared programme inspired by the work already done by FAO in other similar situations.
For the Project Manager, it is more difficult to organize the work of 34 people around 6 ponds when a private investor would probably use only 3 to 6 workers, as in the adjoining salt operation which covers 25 ha. It is ridiculous to pay 4 watchmen and 4 to 6 workers to change water from 2 to 4 a.m. some nights. Many technicians and workers are not busy full time, especially in the absence of the Project Manager. In contradiction to what was presented to the consultant, some of the staff also work part time outside of the project, for the Fisheries Department. Such a situation sounds normal if there is not enough work for every one in the project, but sometimes it results in a conflicting situation as the Project Manager does not know what exactly his team is doing. The extra staff should have been used to classify and to complete the available data.
It is not understood why a supervisor external to the project has to be hired for a harvest after more than 40 experiments. If the personnel is not in condition to accept such responsibility, a reason must be found to explain such a failure. It is a question of work organization and training.
The results show that by flooding dry land, behind the mangrove belt, it is possible to grow shrimps by stimulating the development of lablab, even without any fertilizer, and P. indicus resists very adverse conditions.
However, the technology employed failed to obtain regular production in all the ponds, showing that the quality of the soil was more important than the farming procedure. A true technology should be capable of recognizing and eventually correcting an experiment that went wrong for any reason.
As a conclusion the project has shown that shrimps can be cultured in Kenya but an economical demonstration remains to be done.
