3. RESULTS

In the following section, rice production of all seven farmers in the two fields is discussed and compared; fish production in the rice-cum-fish fields is discussed. The economic viability of the two production systems is also discussed. However, due to the very small sample size, the results of the study can only be regarded as indicative.

3.1 Rice production

The trials show that the rice yields from the rice-cum-fish fields are on an average 11% higher than the yield from the rice-only field, namely 1,517 kg/ha compared to 1,371 kg/ha (Table 1). Similar production systems, also with traditional rice varieties, in other parts of the world show yields of 4,500 kg/ha after four months (in Madagascar⁷) and 3,130 kg/ha after a one year cycle (in the Philippines⁸).

Table 1. Rice yield from the two types of fields.

The average rice yield in Luapula Province in 1988 was estimated to be 1432 kg/ha⁹. The result from the rice-cum-fish plots compared to the provincial average is only slightly (6%) higher. This is possibly because rice cultivation is a new activity for six of the seven trial farmers.

These figures take into account the entire area used and needed for the cultivation activity. However, if the net area actually planted with rice is considered and the area utilized for trenches and refuge holes is disregarded, the rice yield from the rice-cum-fish field on average exceeds the yield from the rice-only field by 33%, benefitting from the presence of fish.

⁷ FAO, 1992.

⁸ NAGA, 1992.

⁹ Republic of Zambia, 1988.

However, the farmer considering taking up rice-cum-fish farming would take into account the potential production from the gross area, including the trenches and the refuge hole, since that is the area of land he has to allocate.

Considering that Chembe is a traditional rice producing area, it was surprising to find that these farmers did not secure the highest yield. There is a considerable difference between the two fields here: the rice yield from the rice-cum-fish field is on an average 33% higher than that from the rice-only field.

The highest rice yields -- considering both the rice-only field and the rice-cum-fish field -- is obtained in Monga, where the rice was cultivated a month longer than in the two other areas.

The trials in Kalasamukoso had negative results. While one farmer had very low rice yields from both fields (though the rice-cum-fish field yields are almost double the rice-only field), the other farmer faced a 39% lower yield from the rice-cum-fish field¹⁰.

3.2 Fish production

Advice was given to the trial farmers on fertilizing both fields and on feeding the fish. The inorganic fertilizer used by the trial farmers was Ammonium Nitrate and D Compound. The rate of application varied between 38 and 119 kg/ha and 0 and 119 kg/ha respectively, which is lower than the recommended rates of 100–200 kg/ha and 200–300 kg/ha. Throughout the trial period, farmers also applied organic manure (chicken, pig, goat, cattle) at a rate of between 41 kg/ha and 3150 kg/ha. Compost as well as grass and vegetable leaves were applied to a lesser extent.

Table 2. Management of and fish yield from the rice-cum-fish field.

¹⁾ No catch, as the fish had escaped.

²⁾ This trial was not monitored after 1 March due to serious illness of the farmer. Besides, most of the fish in the trial pond had died. The few fish harvested had an average weight of 100 grammes.,

¹⁰ A possible explanation for this low outcome is that the farmer was seriously ill for two months.

Feeds included maize and rice bran, vegetable and cassava leaves, grass, sugar cane and kitchen leftovers -- mainly nshima (maize porridge).

The average fish yield of the seven rice-cum-fish fields was 150 kg/ha. When the two farmers whose fish had escaped (farmer no. 1) or died (with only 7 fish left, farmer no. 7) are excluded, the average yield was 196 kg/ha, 4½ month after stocking and at a stocking density of 50 fingerlings per 100 m². In addition to adult fish, a substantial number of fingerlings was produced. These were counted by two farmers who actually captured the fingerlings (for restocking in other ponds), the numbers being 650 and “more than 200”. A considerable amount of fry could also be seen in the ponds. In the longer term the fingerlings produced could be an input for further enhancement of fish production and/or a source of income generation -- an activity which is difficult to value in economic terms here, because this aspect was not included in the study.

Rice-cum-fish culture in the Northern Philippines yields 200 kg/ha over a six-month (stocking density 100 fingerlings per 100 m², applying only organic manure)¹¹.

According to a manual for rice-cum-fish farming in Madagascar, the fish crop should yield 300 kg/ha for extensively managed fields after four months¹². Semi-intensive managed fields are supposed to yield 600 kg/ha after the four months¹³.

3.3 Economic viability of rice-cum-fish farming

The economic viability of rice-cum-fish farming is determined by whether the activity gives the farmer higher benefits than rice-only farming or other activities. In practical terms, net return is most commonly measured for the area of the land set aside for the activity or for the amount of labour that the farmer's household has allocated to the activity. These have been calculated by using June 1993 prices, or prices otherwise applicable¹⁴. Reliable data on labour hours was only available from the trials in Chembe.

The average net return to land from the rice-only field amounts to K 6/m² and K -31/m² for the rice-cum-fish production (see Table 3). The benefits of increased rice yield and the extra crop from the rice-cum-fish field do not outweigh the costs of the higher inputs needed for fish production, thus resulting in a lower net return to land.

¹¹ NAGA, 1992.

¹² FAO, 1992 Stocking density = 25 fingerlings (Oreochromis nilotocus and Cyprinus carpio) per 100 m², inorganic fertilizer (NPK and Urea) applied at 360 kg/ha combined with compost at 1000 kg/ha and no feeding.

¹³ ibid. Stocking density = 50 fingerlings per 100 m², the same species and same rate of fertilizing as above but with a rate of additional feeding on 200 kg/ha.

¹⁴ See Appendix 2.

If, however, one considers the construction of the plots as an initial investment made during the first year and the household labour hours needed for maintenance are 10% of the time spent for the construction, the difference in net return to land from the two fields will be smaller, i.e. K 6/m² and K -21/m² respectively. The difference is nonetheless distinct -- an issue that will be dealt with in the evaluation of the trial methodology.

Table 3 also gives an economic evaluation of the net return to household labour. The average figure for the rice-only field is K 58/hour, whereas the net return to household labour for the rice-cum-fish field is negative, namely K -45/hour. While the figure will remain the same the second year for the rice-only field, since there is no change in labour hours, a slight increase is seen for the rice-cum-fish field to K -41/hour, without changing the pattern of the unrewarding effort put into this activity. As the wage rate for unskilled hired labour is K 56/hour¹⁵, rice-only production gives only a slightly higher net return.

Table 3. Net return to land and household labour assuming that household labour has the value of hired labour (Chembe only).

Note: The following weights for household labour, based on information from ARPT-LP, have been applied:
1 hour female labour = 1 hour male labour;
1 hour child labour = 0.3 hour male labour.

One factor that might enhance the benefit of rice-cum-fish production is the fingerlings obtained during the trials. The value of the fingerlings, however, has not been a subject for investigation and has therefore not been included in the figures for the net returns.

¹⁵ In the analysis wage rate per hour is calculated from the normal working day of nine hours for which the hired labourer gets K 500. Even if he works only half day the usual payment is K 500, but this fact is disregarded here.

After an initial season the farmer should be able to either expand his own fish (or rice-cum-fish) production or start selling fingerlings to other farmers and thereby get a higher benefit.

As there is no sure way to estimate the real value of the household labour, a second set of calculations has been made, where household labour is assumed to have no alternative use and therefore no value (Table 4). In reality the value of household labour may be in the range of K 0–56/hour given the fact there is unemployment in the area.

The average net return to land under these assumptions has increased to K 11/m² for the rice-only field and K 8/m² for the rice-cum-fish field, eliminating the difference between the two. This becomes even more evident when figures for the second year are considered and the labour spent for pond construction is disregarded -- except 10% for pond maintenance. While the average net return to land for the rice-only field remains at K 11/m² it has increased slightly to K 10/m² for the rice-cum-fish field. No hired labour is used in year 2.

Assuming that household labour has no opportunity cost also changes the net return to household labour. The average net return to household labour is K 114/hour for the rice-only field and K 11/hour for the rice-cum-fish field. After the first year of construction the figure for the next year will double for the rice-cum-fish field to K 20/hour, while the net return to household labour for the rice-only field will remain at K 114/hour.

Table 4. Net return to land and household labour where the value of household labour equals 0 (Chembe only).

To ensure that the activity is economically viable for the farmer, the net return to land and to household labour for rice-cum-fish farming has to be at least equal to or higher than for other agricultural activities.

From the above results it appears that rice-cum-fish farming is less profitable than rice farming considering the net return to land as well as the net return to household labour.

It is possible that the farmer may have motives other than profitability for rice-cum-fish farming, i.e. nutritional value, taste, level of cash input and risk, easy maintenance, etc.

A comparison of labour consumption shows that the rice-cum-fish field in fact has demanded seven to ten times as much time as the rice-only field. Figures from Bangladesh suggest an increase of labour in rice-cum-fish farming compared to rice-only farming of 3%¹⁶. A similar comparison in the Philippines shows that a 11% increase in labour hours is the norm¹⁷.

This significant difference calls for a consideration of the trial and supervision methodology used in these trials. As the farmers had been asked to estimate the time spent for the individual tasks in connection with farming, including walking to and from the field, there may have been some double counting if the farmers -- as they often do -- went to the fields for a combination of reasons.

One of the assumed advantages of rice-cum-fish farming is that the farmer can take care of the fish when he is working with the rice. While this may be correct logistically, the day-to-day management, primarily feeding, actually takes 50–55% of the time spent in the total production cycle at the rice-cum-fish field¹⁸. The activity supposed to take up the most time is pond construction. In this trial, however, it took up only 19–38% (i.e. 31–39 hours) of the time spent on the entire production cycle.

Though no far-reaching conclusion should be drawn from a small sample like the present one, the results indicate that rice-cum-fish farming is not economically viable when evaluating the net return to household labour and land.

¹⁶ CARE/ODA/BRAC.

¹⁷ Vincke, P. et al, 1985.

¹⁸ As indicated on page 9, the figures on labour consumption are based on two trials only.