Annex 10
An Appraisal of Extensive Tilapia Culture as an Alternative to Rice Cultivation

1. Introduction

The Government of Sri Lanka reportedly is considering a campaign to induce the farming community to cultivate, on irrigated land, crops other than rice. Fish is one such alternative. This annex analyses this possibility. The analysis is general and intends only to establish whether or not a more detailed feasibility study is worthwhile. The analysis rests on the hypotheses that water is more scarce than land, and considers the possibility of family operated ponds. These ponds are small and maintained as one activity, amongst many, that the farmer and his family are engaged in. The possibility of large-scale specialized tilapia farming will be dealt with in Annex 12.

2. Inputs

Fertilizers are cheap in Sri Lanka. It is argued elsewhere that fertilization of seasonal tanks would be likely to give as good a result as would the combination of fertilizer and feed; that is, better in economic terms. Thus, in this annex we will analyse the case of fish culture with artificial manuring of the ponds, but not with provision of supplemental feed.

Relatively little effort is required to convert a rice field into a fish pond. The bund needs to be heightened and strengthened, and a monk, or other arrangement for letting water out of the pond, must be provided.

Two fish ponds, totalling 0.5 ha, need bunds of at least 350 m. With a designed height of 1.2 m and slopes of 1:2, the volume of material needed for the bund is about 4.5 m³ per running metre of the bund. About 1 575 m³ will be needed all in all. It should be removed from the floor of the rice field, and transported to the site of the bund. This is equivalent to about 2 man-years of work (at the rate of 3 m³ per man/day). In total, the expenditures, assuming paid labour is used, for the two ponds are not likely to exceed Cey.Rs. 20 000. Given a length of life of 50 years, the annual depreciation amounts to Cey.Rs. 400.

Fertilizers are applied once, at the most twice, a month. Thus, the amount of actual physical labour is very small to tend to two 0.25 ha ponds. However, depending on the location, the culture may require somebody's continued presence in order to discourage poaching.

3. Management

Monoculture of Tilapia nilotica will be carried out, but not as a monosex operation. Each pond is harvested every six months, and production is staggered so that one harvest takes place every three months. The harvested pond is replenished with fingerlings and fry from the pond under culture. In the analysis we distinguish between two fertilizer application schemes. The first assumes application of urea (1 800 kg/ha/year) and superphosphate (4 800 kg/ha/year) leading to a production of 3 400 kg of Tilapia nilotica per hectare and year. The second scheme makes use only of cow dung (12 tons/ha/year) and gives a production of about 4 000 kg of Tilapia nilotica per hectare and year. The production rates assume that the pH of the pond water is above 7.

4. Return

Table 10.1 itemizes the details of direct expenditures associated with the production. At Cey.Rs. 4/kg (at pond side), it is evident that the net return can be quite substantial. When cow manure is used and 4 000 kg/ha/year are obtained, the net return is about Cey.Rs. 3 500 for half a hectare and a culture period of six months.

So far the analysis has assumed that all the water that is necessary will be available and at no cost. This may not be entirely realistic in the Sri Lanka context. For lack of water, land is not cultivated. Thus, any use of water should be considered in the light of what else could be produced with the quantity of water concerned.

5. Water supply as a constraint

While rice fields are covered with about 10 cm of water, a fish pond should have an average depth of 80 cm for the culture of T. nilotica. Thus, it would seem that culture of T. nilotica would need more water per hectare of land than would cultivation of rice. Therefore, if rice land is converted into fish ponds, the area “lost” to rice cultivation will not be that physically occupied by the pond, but also that which can not be cultivated because the necessary irrigation water has been used to fill up the fish pond.

Table 10.2 displays the water requirements for rice cultivation and for culture of T. nilotica respectively. It shows that about half as much water is needed to cultivate one ha of rice than to culture T. nilotica in fish ponds with a surface of 1 ha. It should be noted, however, that the periods of culture are not equally long - 135 days for the rice, and 180 days for the tilapia. It is also relevant that at the end of the culture of tilapia, (harvest through seining) the pond should, in theory, have 80 cm of water, while the normal rice field is not covered by water after the harvest. The water in the pond can be re-used. Table 10.3 shows three possible culture schemes, in a 12 months cycle, for a 1 ha plot: only rice, only fish, or fish followed by rice.

Fertility, water availability, culture practices and different input costs mean that the net return from rice cultivation is not uniform in Sri Lanka. Table 10.4 shows the net return projected by the World Bank for the 1979/80 Maha harvest in four Sri Lankan districts.

The information contained in Tables 10.1 to 10.4 is combined in Table 10.5. It shows net returns to 1 ha of rice land under the three different culture schemes identified in Table 10.3.

6. Conclusions

Table 10.5 indicates a number of things. First, that when the yield of fish is only 3 400 kg/ha/year (equivalent), then it is only in the Colombo region that tilapia culture might yield as high a net return as will rice culture. Also, under this yield assumption, fish culture, followed by rice, would yield a higher net return than does only fish culture. However, the main impression given by the top part of Table 10.5 is that from the point of view of net returns, there would be no attraction in tilapia culture using inorganic fertilizers.

When the yield assumption is increased to the equivalent of 4 000/kg/ha/year and cow manure is used instead of superphosphate and urea, benefits from culture of T. nilotica would exceed those from culture of rice. In Hambantota and Colombo exclusive fish culture could bring a slightly higher yield than fish culture followed by rice cultivation.

In fact, fish culture is likely to be more attractive than Table 10.5 indicates. The reason is that the cost of family labour has not been deducted from receipts. Rice cultivation demands at least some 5 times the number of man-days than does fish culture (once the ponds have been built), and a comparatively large proportion of the work in rice fields is carried out by family members (see Table 10.4). For the land owner (or pond owner), it is obvious that the less family labour he has to use to earn a certain net income from his lands the better. He would, therefore, be attracted by tilapia culture in rice fields. This same fact is, of course, detrimental to the interests of the landless labourer.

However, even if a more thorough analysis than this (based on field trials) would bear out these conclusions, it is not likely that the shift to tilapia culture would be massive or universal. The changes in supply of both rice and fish, which would result from any such change, would influence the market price of both products. This would, in turn, influence production decisions and an equilibrium would be approached.

The analysis in this Annex is partial and general. It is not intended, therefore, to be conclusive. However, it has served the purpose of identifying a possibility which should be pursued with trials in the field. Such trials should:

• establish the potential yield in converted rice fields for T. nilotica (and other species) using cow manure; and

• verify assumptions about water usage.

Table 10.1

Yearly costs and income for extensive culture of T. nilotica in former paddy fields

¹ Assuming pH of water is above 7

² Assuming that the average cost of converting 1 ha of paddy field into fish ponds is Cey.Rs. 20 000 and that this is depreciated over 50 years. This statement of costs and income considers the point of view of the pond owner. We assume that the operation has not required credit; therefore, no interest is charged

³ Assume that most production is retailed by the producer and that most producers only raise fish in plots much smaller than 1/2 ha

Table 10.2

Water requirements for culture of T. nilotica and for paddy cultivation

¹ 135 days

² 180 days

Table 10.3

Alternative usages of 1 ha of paddy land during a 12 months period, assuming that water availability decides the rice that can be cultivated

Table 10.4

Net returns on paddy cultivation in Sri Lanka

¹ World Bank, 1980

Table 10.5

Net returns¹ from 1 ha of paddy land, according to culture scheme, and level of production in fish ponds

¹ Net returns include the cost of hired unskilled labour but not that of family labour (see Table 10.4)